Handbook Triboluminiscence

David O. Olawale · Okenwa O.I.
Okoli
Ross S. Fontenot · William A. Hollerman
Editors
Triboluminescence
Theory, Synthesis, and Application
Triboluminescence
David O. Olawale • Okenwa O.I. Okoli
Ross S. Fontenot • William A. Hollerman
Editors
Triboluminescence
Theory, Synthesis, and Application
Editors
David O. Olawale Okenwa O.I. Okoli
Industrial and Manufacturing Industrial and Manufacturing
Engineering Department Engineering Department
FAMU-FSU College of Engineering FAMU-FSU College of Engineering
Tallahassee, FL, USA Tallahassee, FL, USA
Ross S. Fontenot William A. Hollerman

Naval Surface Warfare Center Department of Physics
Bethesda, MD, USA University of Louisiana
Lafayette, LA, USA
ISBN 978-3-319-38841-0 ISBN 978-3-319-38842-7 (eBook)

DOI 10.1007/978-3-319-38842-7
Library of Congress Control Number: 2016944358
© Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
Preface
The triboluminescence phenomenon, since its first recorded discovery in 1605, has
enjoyed extensive studies targeted at understanding the underlying mechanism,
discovery, and synthesis of new materials, their characterization, and applications
in civil and aerospace engineering systems. Significant progress has been made in
these areas since the last text that was written over 35 years ago. The need to
concisely document the significant progress that has been made in this field in the
last 35 years necessitated this book.
The book expounds on progress made over the last 35 years in the theory,
synthesis, and application of triboluminescence for creating smart structures. It
presents in detail research into the utilization of the triboluminescent properties of
certain crystals as new sensor systems for smart engineering structures. These
triboluminescence-based sensor systems have the potential to enable wireless, in
situ, real-time and distributed (WIRD) damage, stress, and impact sensing in civil
and aerospace systems like bridges, aircrafts, space crafts, and wind blades.
Furthermore, the book is divided into three sections according to the covered
areas which are the theory, synthesis, and application of the triboluminescence
phenomenon. In order to ensure depth and breadth in the coverage of these key
areas, the editors worked with leading experts in the field from all over the world to
author the very insightful chapters in the book. The book is written to present
information on triboluminescence relevant to engineers and scientists across a
range of fields, including aerospace, defense, civil infrastructure, and wind energy.
The goal is to facilitate readers’ understanding with concise treatments of the topics
covered in the text.
In conclusion, we would like to express our profound gratitude to all our
contributing authors for the great depth and expertise they have brought to this
book. You have helped in documenting the various advances made in the field for
the benefits of the present and coming generations. Thanks for all the hard work and
v
vi Preface
timely submission of the manuscripts. We are particularly grateful to our publishing

team at Springer. Special thanks to Michael Luby and Ms. Brinda Megasyamalan
for their support, patience, and guidance throughout the project.
Tallahassee, FL David O. Olawale

Tallahassee, FL Okenwa O.I. Okoli
Bethesda, MD Ross S. Fontenot
Lafayette, LA William A. Hollerman
March, 2016
Contents
1 Introduction to Triboluminescence . . . . . . . . . . . . . . . . . . . . . . . . . 1
David O. Olawale, Ross S. Fontenot, Md Abu S. Shohag,
and Okenwa O.I. Okoli
2 Nature of the Electronic Charge Carriers Involved
in Triboluminescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Friedemann T. Freund
3 Mechanoluminescence of Coordination Compounds . . . . . . . . . . . . 39
Ercules Epaminondas Sousa Teotonio, Wagner Mendonça Faustino,
Hermi Felinto Brito, Maria Claudia França Cunha Felinto,
Jandeilson Lima Moura, Israel Ferreira Costa,
and Paulo Roberto Silva Santos
4 Luminescence of Triboplasma: Origin, Features,
and Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Roman Nevshupa and Ken’ichi Hiratsuka
5 Triboluminescence of Liquid Dielectrics: On a Way
to Discover the Nature of Sonoluminescence . . . . . . . . . . . . . . . . . . 95
D.A. Biryukov and D.N. Gerasimov
6 Functional Triboluminescent Nanophase for Use
in Advanced Structural Materials: A Smart Premise
with Molecular and Electronic Definition . . . . . . . . . . . . . . . . . . . . 125
M. Jasim Uddin, David O. Olawale, Jin Yan, Justin Moore,
7 Europium Tetrakis Dibenzoylmethide Triethylammonium:
Synthesis, Additives, and Applications . . . . . . . . . . . . . . . . . . . . . . 147
Ross S. Fontenot, Kamala N. Bhat, William A. Hollerman,
and Mohan D. Aggarwal
vii
viii Contents
8 Mechanoluminescence Induced by Acoustic Cavitation . . . . . . . . . 237

Nathan C. Eddingsaas
9 Triboluminescence of Inorganic Lanthanide Salts . . . . . . . . . . . . . 273
G.L. Sharipov and A.A. Tukhbatullin
10 Triboluminescent Sensors for Polymer-Based Composites . . . . . . . 305
Kunal Joshi, Margaret Scheiner, David O. Olawale,
and Tarik J. Dickens
11 Detection of Low-Velocity-Impact Triboluminescent
Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Shawn M. Goedeke, William A. Hollerman,
Stephen W. Allison, and Ross S. Fontenot
12 3D Sensing Using Solid-State Wire-Shaped Photovoltaic
Sensor in TL-Based Structural Health Monitoring . . . . . . . . . . . . . 351
Jin Yan, M. Jasim Uddin, David O. Olawale, Tarik J. Dickens,
13 Triboluminescent Sensors for Cement-Based Composites . . . . . . . . 379
David O. Olawale, Jasim M. Uddin, Jin Yan, Tarik J. Dickens,
14 Triboluminescence at Speeds Greater than 100 m/s . . . . . . . . . . . . 411
Ross S. Fontenot, William A. Hollerman,
and Noah P. Bergeron
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Chapter 1
Introduction to Triboluminescence
David O. Olawale, Ross S. Fontenot, Md Abu S. Shohag,

1.1 Luminescence
Light emission in a material can either be by incandescence or luminescence.

Figure 1.1 gives an overview of some the various forms by which light can be
produced [1]. Incandescence is light produced by heating an object to such a high
temperature that the atoms become highly agitated leading to the glowing of the
bulk matter [2]. It is an inefficient way of producing light as most of the energy is
dissipated as heat. Incandescence is explained by Planck’s black body emission
theory. On the other hand, luminescence, sometimes referred to as “cold light,” is
light produced at normal and lower temperatures [3]. It involves an energy source
elevating an electron of an atom out of its “ground” state into an “excited” state; and
the sequential release of energy in the form of light when the electron returns to its
ground state [2]. It is a more efficient mechanism of light emission.
Wiedemann identified six kinds of luminescences based on the method of
excitation [4]. These are photoluminescence, thermoluminescence, electrolumines-
cence, crystalloluminescence, triboluminescence, and chemiluminescence
(Fig. 1.1). Photoluminescence of solids is excitation by light and is subdivided
D.O. Olawale (*)

Nanotechnology Patronas Group Inc., 914 Railroad Ave., Tallahassee, FL 32310, USA
Department of Industrial and Manufacturing Engineering, FAMU-FSU College
of Engineering, 2525 Pottsdamer Str., Tallahassee, FL 32310, USA
e-mail: doo07@my.fsu.edu
R.S. Fontenot
Carderock Division, Naval Surface Warfare Center, Code 6301, West Bethesda,
MD 20817, USA
M.A.S. Shohag • O.O.I. Okoli
© Springer International Publishing Switzerland 2016 1

D.O. Olawale et al. (eds.), Triboluminescence, DOI 10.1007/978-3-319-38842-7_1
2 D.O. Olawale et al.
Light emission
With heating?
yes no
Incandescence Luminescence
From mechanical forces?

yes no
Triboluminesce With excitation by light?
Yes no
Immediately? Inanimate?
yes no yes no
Fluorescence Phosphorescence Chemiluminescence Bioluminescence
Fig. 1.1 Overview of different light emission mechanisms [1]
into fluorescence and phosphorescence. Thermoluminescence is light produced

from excitation by gentle heating. Electroluminescence is caused by excitation of
gases in electrical fields. Crystalloluminescence is from excitation resulting from
solution crystallization while triboluminescence is from excitation of crystals when
they are stressed or broken. Chemiluminescence occurs during chemical reaction
and all bioluminescence, occurring in living organisms, are examples of
chemiluminescence [4].
Light-induced luminescence can be classified based on the time delay between
the completion of the excitation process and the start of photon emission for a
luminescent material [5]. Fluorescence is luminescence that occurs when photons
emission is due to a direct transition of less than 10 ms decay time. It is the process
where light emission is ruled by the lifetime of the emitting center. Materials that
emit fluorescence are known as fluors and are used for applications where timing is
important. Increasing the temperature of a substance generally reduces its fluores-
cence. Phosphorescence is luminescence that persists for more than 100 ms after
cessation of irradiation [5]. Phosphorescent materials can emit luminescence many
hours after cessation of excitation. During phosphorescence, atomic transitions
through intermediate metastable states determine the exact duration of the phos-
phorescence. Materials that emit phosphorescence are known as phosphors and are
used in powdered form for lamps, television screens, and related purposes. How-
ever, the light from phosphors in these applications exhibit decay times less than
10 ms [5]. The remaining part of this chapter focuses on triboluminescence.
1 Introduction to Triboluminescence 3
Fig. 1.2 Stressed SrAl2O4:Eu and the triboluminescence image under a compressive load of
1000 N [21]
1.2 Triboluminescence
Triboluminescence, also known as fracto- [6], piezo- [7], mechano- [8], crystallo-,
or sono-luminescence [9, 10], is the emission of light by solid materials when they
are stressed or fractured [3, 11, 12]. The term triboluminescence was coined by
Wiedemann in 1888 [13] and it basically means light from friction, as the term
comes from the Greek word tribein, meaning “to rub,” and the Latin prefix lumin,
meaning “light” [5]. Triboluminescence has been observed in various luminescent
processes such as [9] emission during breaking of adhesive bonds [14, 15]; shaking
of mercury in a glass vessel [16]; rapid crystallization of certain substances [17];
collapse of small gas bubbles in a liquid [18]; excitation of a laser-induced shock
wave in solids [19]; elastic and plastic deformation of solids; scratching; milling;
and fracture [13, 20]. The phenomenon was first studied by Sir Francis Bacon with
sugar as recorded in The Advancement of Learning in 1605 [5, 13]. Figure 1.2 shows
the stressed SrAl2O4:Eu sample and the TL image under a compressive load of
1000 N. The stressed sample emitted intense visible green light from its two ends.
Although triboluminescence phenomenon has been studied for centuries, it
remains an enigma at the conceptual and theoretical levels [9]. The phenomenon
is closely related to both friction and wear and all three require an understanding of
the highly nonequilibrium processes occurring at the molecular level [22]. These
processes are significantly different depending on the tribological conditions,
environment and materials [9]. The emission spectrum for sugar indicates that the
light comes from the atmospheric nitrogen that fills the gap during fracture;
the same source of light as lightning or touching a doorknob on a winter day. On
the other hand, spectra for other samples show emission from the material as well as
the nitrogen lines, suggesting a secondary energy process. Furthermore, other

materials show a spectrum characteristic of the material without the nitrogen
lines. While the abrupt charge separation is the same in all cases, the mechanism
(s) of emission depend on the material [5].
1.3 Elastico-, Plastico-, and Fracto-Triboluminescence
Triboluminescence may be divided in to three types: namely (1) elastico-,

(2) plastico- and (3) fracto-triboluminescence [3, 23].
1.3.1 Elastico-Triboluminescence
Elastico-triboluminescence is luminescence produced during the elastic deforma-

tion of solids, where neither fracture nor plastic deformation is required [3]. This
can be by mechanical or electrostatic interaction of dislocations with defect centers,
or by thermal excitation in the stressed regions of crystals that exhibit elastico-
triboluminescence such as x- or γ-irradiated alkali halides and ZnS:Mn [23].
Chandra et al. [24] provided experimental evidence supporting the suitability of
a piezoelectrically stimulated electron detrapping model as being responsible for
the elastic and plastic triboluminescence of ZnS:Mn. The suggested steps for
elastico-triboluminescence in ZnS:Mn based on the piezoelectric mechanism [24]
are as follows: The deformation of ZnS:Mn crystals produces piezoelectric field
because the crystal structure of ZnS is non-centrosymmetric [25]. The piezoelectric
field results in decrease in the trap-depth which causes detrapping of electrons from
filled-electron traps, with the electrons reaching the conduction band as illustrated
in Fig. 1.3. The electrons may recombine with the holes trapped in the defect
centers or they may fall to the valence band with the energy being released
non-radiatively. The energy released non-radiatively may be transferred to the Mn2+
ions to cause their excitation [26–29]. The de-excitation of excited Mn2+ ions gives
rise to the light emission characteristic of the Mn2+ ions.
Elastico-triboluminescence in ZnS:Mn starts at a pressure of about 1 MPa
(106 Nm2) [30]. The piezoelectric charge density γ generated at this pressure
in the crystal with a piezoelectric constant (d33 ¼ 3.3 1011 CN1) [31] will be
3.3 105 Cm2. This will result in an electric field F given by the relation
F ¼ γ=ε0 ð1:1Þ
The permittivity ε0 of the crystal is 8.85 1012 CNm2; consequently, an

electric field of about 3.7 106 Vm1 or 3.7 104 V cm1 will be developed
near the crystal surface [30]. The internal electric field will however be about one
order less than the external field on the crystal surface because the dielectric
Fig. 1.3 Schematic of

elastico-triboluminescence Conduction Band
in ZnS:Mn [3]
Shallow
traps
Holes
hv
Valence Band
constant of ZnS crystals is 8.8 [30]. An electric field of the order of 106 V cm1 is
however needed to cause electron detrapping from traps or the impact excitation of
Mn2+ centers [30]. The local electric field near Mn2+ ions may however be higher
because of the local change in the crystals’ structure near Mn2+ sites [32, 33]. This
may result in a higher piezoelectric constant near the Mn2+ sites [34, 35] to generate
an electric field of the order of 105 V cm1 that may cause sufficient decrease in the
trap-depth but not sufficient to cause the impact excitation of Mn2+ centers
[30]. Subsequent electron–hole recombination may release energy non-radiatively
for the excitation of Mn2+ centers.
1.3.2 Plastico-Triboluminescence
Plastico-triboluminescence is luminescence produced during plastic deformation of

solids where fracture is not required. It can be excited by the mechanical or
electrostatic interaction of dislocations with defect centers; electrification of crystal
surfaces by the movement of charged dislocations; or thermal excitation in the
stressed regions of solids like colored alkali halides, II–VI compounds, alkaline-
earth oxides, and metals [23]. Chandra et al. [36] reported on the luminescence
arising from the plastic deformation of colored alkali halides using pressure steps.
In the elastic region, the strain increases linearly with the stress, and the tribolumi-
nescent intensity also increased linearly with stress. In the plastic region, the strain
and TL intensity increased with stress according to the power law [36].
The suggested steps involved in the TL of x- or γ-irradiated alkali halide crystals
are as illustrated in Fig. 1.4: (1) Plastic deformation causes movement of disloca-
tions. (2) The moving dislocations capture electrons from the interacting F-centers
lying in the expansion region of dislocations. (3) The captured electrons from
F-centers move with the dislocations and they also drift along the axes of
Fig. 1.4 Energy level

diagram of Conduction Band
mechanoluminescence
of colored alkali halide
crystals [37] Interacting F
Shallow
traps centers
Holes
hv
V2
Valence Band
dislocations. (4) The recombination of dislocation-captured electrons with the holes

lying in the dislocation donor band gives rise to the light-emission characteristic of
the halide ions in hole centers [38].
The triboluminescent (TL) intensity I was found to be directly related to η, the
efficiency of radiative electron–hole combination, and nd, the number of electrons
in the dislocation band at any time t as follows:
ηP0m mξpf r f nf
I ¼ ηβnd ¼ ½expðфtÞ expðξtÞ ð1:2Þ
K m bλðξ фÞ
where β ¼ 1=τd , τd is the lifetime of electrons in the dislocations band, P0 is the

final value of pressure, ξ ¼ 1=tr , tr is the time constant for rise of pressure, K is
referred to as the strength coefficient, n is the work-hardening exponent, m ¼ 1=n,
pf is the probability of capture of interacting F-center electrons by the dislocations,
rf is the radius of interaction between the moving dislocations and F-centers, nf is
the density of F-centers in the crystals, ф ¼ 1=τp , τp is the pinning time of the
dislocations, b is the Burgers vector, and λ is the mean free path of the moving
dislocations. The maximum triboluminescent intensity Im is given as
P0m mξpf r f nf
Im ¼ ð1:3Þ
K m bλ
The total intensity is given as
Z1
P m mξp r f nf
I t ¼ Idt ¼ 0 m f ð1:4Þ
K bλф
0
The Im and tm increase according to the power law with an increase in the applied
pressure. It has also been shown that the TL intensity depends on many parameters
including strain rate, stress, density of F-centres, size of crystals, temperature, and
luminescence efficiency [13, 23, 39, 40]. The following relationships show the
effect of temperature on the triboluminescent response [36]:
P0m mξpf0 r f nf
Im ¼ expðEa =KT Þ ð1:5Þ
K m bλ
The total intensity is given as
Z1
P m mξp r f nf
I t ¼ Idt ¼ 0 m f expðEa =KT Þ ð1:6Þ
K bλф
0
where Ea is the energy gap between the bottom of the dislocation band and the
average ground state of the interacting F-centers. The Im and It increase with
increase in temperature because at low temperature, expðEa =KT Þ increases with
temperature. However, at higher temperature, Im and It decrease because nf
decreases as a result of thermal quenching. Hence, Im and It are optimum for a
particular temperature of the crystals.
1.3.3 Fracto-Triboluminescence
Fraco-triboluminescence is luminescence produced due to the creation of new

surfaces during the fracture of solids. During fraco-triboluminescence, there is
creation of charged surfaces at fracture (Fig. 1.5) due to processes such as
piezoelectrification, movements of charged dislocations, and charged defect
barodiffusion [23, 41]. There is neutralization of these surface charges by the
charge carriers or ions produced from the dielectric breakdown of the intervening
gases and solids. This results in the production of luminescence that resembles a gas
discharge (e.g., sucrose, tartaric acid, Rochelle salt) or luminescence of a solid (e.g.,
coumarin, resorcinol, and phenanthrene) or one that combines the characteristics of
both the intervening gases and solids (e.g., uranyl nitrate hexahydrate, impure
saccharin, and chlorotriphenyl-methane [12, 23].
If a crystal with thickness H is cleaved along a plane parallel to its width W, with
velocity of crack propagation being v, provided α3 t < 1, then the TL intensity may
be given as [23]
I ¼ 2ðη1 α1 þ η2 α2 Þγwvt ð1:7Þ
where γ is the charge density of the newly created surfaces, α1 and α2 are the rate
constants for the relaxation of charges on the newly created surfaces, α3 ¼ α1 þ α2 ,
and η1 is the luminescence efficiency associated with the movement of carriers
produced by the dielectric breakdown of the crystals while η2 is the efficiency
+ + −
− −
+ − d +
+ + −
− −
+ + − +
− −
+ − + − +
+ + −
− + −
+ + −
− +
+ −
c + −
+ + −
− −
+ − +
+ −
+ −
Fig. 1.5 A schematic of the piezoelectric theory illustrating TL phenomena upon cleavage [41]
associated with the movement of electrons and ions produced by the dielectric
breakdown of intervening gases. According to Eq. (1.7), when a crystal is cleaved,
the TL intensity should rise linearly with time t. At the end of cleavage (at t ¼ tm,
v ¼ 0), the TL intensity may be expressed as
I ¼ ½η1 α1 þ η2 α2 Q0 exp½α3 ðt tm Þ ð1:8Þ
where Q0 is the surface charge at t ¼ tm.

Equation (1.8) shows the exponential decay of the TL intensity after the cleav-
age of the crystals.
Triboluminescence can be used to determine the velocity of crack propagation in
crystals [23] as follows:
v ¼ H=tm ð1:9Þ
The peak TL intensity at t ¼ tm is given as
I m ¼ ðη1 α1 þ η2 α2 ÞγA ð1:10Þ
where A ¼ 2WH is the area of the newly created surfaces. Equation (1.10) implies
that Im should increase linearly with A and γ. The total intensity IT is given as
Z
I T ¼ I dt ¼ ðη1 α1 þ η2 α2 ÞγA=α3 ð1:11Þ
When α2 ¼ 0, that is no gaseous discharge as is the case with ZnS:Mn, the emission
will be primarily bulk triboluminescence and α3 ¼ α1 . When α2 ¼ 0, that is, no bulk
triboluminescence—emission is primarily by gaseous discharge, and α3 ¼ α2 .
Based on Eqs. (1.10) and (1.11), Im and IT should decrease with increase in
temperature because η1, η2, and γ decrease with temperature increase. However, a
specific temperature may be reached above which γ may decrease to the point that
the breakdown of gases and solids becomes impossible, and triboluminescence no
longer occurs [23].
1.4 Triboluminescent Materials
According to Virk [4], the year 1603 was the beginning of modern luminescent
materials because the first artificial phosphor described in Western literature dates
from this year. In an effort to create gold, the Italian shoemaker and alchemist,
Vincenzo Cascariolo, heated the natural mineral barite (BaSO4) to create a persis-
tent luminescent material called Bolognian stone. The host material was BaS
making it the first sulfide phosphor ever synthesized [11] and the first scientifically
documented material to show persistent luminescence [4]. In the same vein, Francis
Bacon in 1605 made the first recorded observation of triboluminescence in his
writings, “Advancement of Learning,” about sparkling light from hard sugar when
scrapped with a knife [41]. In 1684, Waller reported that when substances such as
white sugar, loaf sugar, and rock salt were crushed in a mortar, they gave such
intense light that the sides of the mortar and shape of the pestle could distinctly be
displayed [23]. Friedrich Hoffmann synthesized CaS as a phosphor in 1700 while
J. F. John synthesized SrS in 1817 [4]. The luminescent properties of ZnS, one of
the most important luminescent hosts in the twentieth century, were not recognized
until 1866, when the Sidot blend (hexagonal ZnS) was developed by Theodor Sidot
in France [8].
The main contribution of early nineteenth-century research was the compilation
of an extensive list of TL materials using highly subjective visual observation of
the TL response as a function of time and quantity [13]. The development of the
photomultiplier tube (PMT) in the 1930s and its application in triboluminescence
studies in 1952 introduced a quantitative technique for detecting, measuring, and
comparing TL emissions objectively.
The discovery of phosphor-based high-field electroluminescence in solids in
1936 is credited to Destriau [42]. Between late nineteenth and early twentieth
centuries, Philip E.A. Lenard and his colleagues used different rare earth ions in
addition to heavy metal ions as luminescent ions in different host materials to create
new phosphors [43]. Hurt et al. synthesized europium dibenzoylmethide
triethylammonium (EuD4TEA), also known as europium tetrakis, in 1966
[44]. EuD4TEA is an organic material in which triboluminescence can be observed
in daylight [44, 45]. It has been estimated that 30 % of organic crystals and 50 % of
inorganic crystals are triboluminescent [13].
Triboluminescent (TL) materials are being used to develop sensor systems for
engineering structures [46–49]. Triboluminescence-based sensor systems have the
potential for wireless, in situ, real-time, and distributed (WIRD) sensing that can
enable continuous monitoring of civil and aerospace structures [3]. Researchers
have investigated the application of TL in damage detection [11, 48]. They can be
used as stress, fracture, and damage sensors [24, 50]. They have also been proposed
for visualizing the stress field near the crack-tip, stress distribution in solids, and
quasi-dynamic crack-propagation in solids [47, 49–53].
Any candidate triboluminescent material for this purpose should however have a
large triboluminescent emission yield, be readily available, be environmentally
Table 1.1 Triboluminescence of various materials [49]

Group Sample TL intensity (cps)a
Hexagonal ZnS 60
ZnS–Mn0.05 2800
ZnS–Cu0.01 1100
Zn2SiO4:Mn0.01 57
ZnO 3
SiC 4
X2O3 (X ¼ Al or Y) α-Al2O3 10
α -Al2O3:Mn0.01 60
Y2O3 9
Y2O3:Eu 20
MgAl2O4 31
CaAl2O4 14
SrAl2O4 36
Fluorite ZrO2 8
HfO2 3
CeO2 3
Perovskite YBa2Cu4O8 1
PbZr0.52 Ti0.48 O3 3
Pb0.93La0.07Zr0.60Ti0.40O3 3
a
TL intensity was measured as a function of friction applied to the material by a brass rod of 1 mm
diameter under a load of 5 N and slide speed of 6.3 cm/s (2.5 rps)
benign, and be compatible with the host material [3, 45]. Xu et al. [49] compared
the triboluminescent performance of different inorganic materials under identical
mechanical stress conditions and obtained similar results for thin film and bulk
materials (Table 1.1). The ZnS-based materials showed much higher TL intensities
while the ZnS doped with Mn exhibited the highest luminous intensity of all the
materials tested (Fig. 1.6a). It was shown that a manganese doping level of about
5 % gives the highest TL response (Fig. 1.6b) making ZnS:Mn a strong candidate
for TL-based sensor systems.
The hexagonal crystal structure of ZnS:Mn is considered “loose” because it
emits light with very little stress applied [45] such as by simply scratching it with a
nail or any sharp object. Materials such as ZnS:Mn are usually made up of a
semiconductor host (ZnS) and an impurity (Mn) called the dopant. The dopant
concentration is usually a small fraction of the composition but it plays the critical
function of changing the band structure of the crystal, thereby narrowing the energy
gap between conduction and valence bands. With narrower energy gaps, transitions
that emit light are more probable and this increases the opportunity for light to be
emitted during excitation or relaxation of electrons [45].
In a more recent study, Hollerman et al. [45] measured the relative tribolumi-
nescent emission yields for 27 candidate triboluminescent materials powders under
impact loading using a low energy drop tower. The goal was to compare the
a b
1200 3000
1000 2500
800
TL intensity (a.u.)
2000
TL intensity (a.u.)
600 1500
400 1000
200 500
0 0
Al2O3 ZrO2 Y2O3 ZNS ZNS:Mn ZNS:Cu Al2O3:Mn
0 5 10 15 20
TL Flim Mn additive amount (at%)
Fig. 1.6 (a) TL intensity of various inorganic thin films under the same friction conditions and (b)
effect of Mn additive amount on TL intensity [49]
triboluminescent emission yields for ZnS-based materials as a function of grain size

and dopants (Table 1.2). The study included three organic Europium based lumi-
nescent materials (LM 9, 181, 194). The result shows that the EuD4TEA sample
doped with dimethyl methylphosphonate (DMMP) (LM- 194) has the largest tested
triboluminescent yield for all the tested materials [45]. It had over 3.19 times the
triboluminescent yield compared to the baseline 7.5 μm ZnS:Mn. In addition, the
19.8 μm ZnS:Mn (LM-33) has the largest triboluminescent yield among all the
tested inorganic materials. It was proposed that the method used to prepare LM-33
in terms of grain size, trap concentration, and dopant concentration, was responsible
for the increased triboluminescent yield. The low yield from LM-88 and LM-89
ZnS:Mn may be accounted for by the 40 % manganese in the formulation. On
average, ZnS:Mn,Cu tends to have a larger triboluminescent yield compared to
ZnS:Mn. No triboluminescence was observed with impact loading of the 5 nm sized
ZnS:Mn powder. It is likely that the small 5 nm ZnS:Mn particles were trapped in
surface imperfections and were not subject to sufficient force to produce tribolu-
minescence [45]. Compared to ZnS:Mn, EuD4TEA appears to be more easily
damaged than ZnS:Mn. The triboluminescent yield for EuD4TEA decreases by a
factor of three from drops one to five whereas, the triboluminescent yield for ZnS:
Mn only decreases by only about 10 % for the same number of drops.
Finally, the compatibility of the TL material with the host matrix is vital to
manufacturing and implementation of TL-based sensor systems [3] because the TL
material cannot function as a sensor on its own. For integration into composite
material systems, it might be necessary that the host matrix have a melting point
greater than the cure temperature of the composite and also need to be chemically
compatible with it [54]. Sage et al. [54] provided information about the melting
Table 1.2 Comparison of the triboluminescent yields for the 27 luminescent materials (LM) as
measured using the drop tower [45]
Grain
Sample LM Lot size Yield
Base material mass (g) number number Manufacturer (μm) ratioa
ZnS:Mn 1 171 1 Center for Integrated 0.005 0.000
Nanotechnologiesb
1/0.1 34 15,248 Phosphor technology 7.5 1.000/
1.000a
1 7 17,112 Phosphor technology 8.5 1.223
ZnS:Cu 1 28 19,017 Phosphor technology 2.9 0.005
ZnS:Mn,Cu 1 96 19,010 Phosphor technology 21.9 1.130
1 177 20,267 Phosphor technology 1.585
ZnS:Cu,Pb 1 97 15,027 Phosphor technology 19.0 0.034
ZnS:Cu,Pb,Mn 1 95 17,002 Phosphor technology 19.3 1.017
MgF2:Mn 1 138 09,147 Phosphor technology 0.029
La2O2S:Eu 1 15 10,185 Phosphor technology 0.004
Y2O2S:Eu 1 90 19,145 Phosphor technology 0.000
EuD4TEA 0.1 9 None Sandia National 0.960
Laboratoriesc
0.1 181 3 Alabama A&M 2.063
University
EuD4TEA + 0.1 194 10 Alabama A&M 3.196
1.25 mL DMMP University
a
Ratio based on the TL light yield for both the 1 and 0.1 g samples of 7.5 μm ZnS:Mn (LM-34) set
equal to 1.000
b
The Center for Integrated Nanotechnologies (CINT) is located in Albuquerque, New Mexico
c
Sandia National Laboratories is located in Livermore, California
point and chemical compatibility of some highly efficient TL materials with Ciba
resin systems (MY750/HY956 or MY750/HY917) as illustrated in Table 1.3.
A good interface is also needed for effective load transfer to the crystal for adequate
TL excitation.
Table 1.3 Melting points and chemical compatibility of a range of highly triboluminescent
materials [54]
Material Melting point ( C) Chemical compatibility
Eu complex 170–250 Y
Tb complex 155–290 Y
Mn complex I 230 Y
Mn complex II 300 Y
U complex 60 -
Ester 195 Y
Acetyl complex 185 Y
Aspirin derivative 135 Y
1.5 Conclusion
Although extensive work has been done in the synthesis and characterization of
triboluminescent materials, there still remains significant gaps in the understanding
of the underlying mechanisms responsible for the phenomenon. Work is ongoing to
develop triboluminescence-based sensors as load, damage, and impact monitoring
systems in engineering systems such as civil and aerospace structure. The key
requirements for triboluminescent materials for such applications include large
triboluminescent emission yield, be readily available, be environmentally benign,
and be compatible with the host material. The ZnS:Mn and EuD4TEA are the
leading materials for these applications because of their high triboluminescent
yields. The ZnS:Mn, because it is inorganic, has however exhibited higher durabil-
ity property compared to the organic EuD4TEA.
References
1. O’Hara, P. B., Engelson, C., & St Peter, W. (2005). Turning on the light: Lessons from
luminescence. Journal of Chemical Education, 82, 49–52.
2. Vishwakarma, K., Ramrakhiani, M., & Chandra, B. P. (2007). Luminescence and its applica-
tion. International Journal of Nanotechnology and Applications, 1, 29–34.
3. Olawale, D. O., Dickens, T., Sullivan, W. G., Okoli, O. I., Sobanjo, J. O., & Wang, B. (2011).
Progress in triboluminescence-based smart optical sensor system. Journal of Luminescence,
131, 1407–1418.
4. Virk, H. S. (2015). History of luminescence from ancient to modern times. Defect and
Diffusion Forum, 361, 1–13.
5. Goedeke, S., Allison, S., Womack, F., Bergeron, N., Hollerman, W. (2003). Tribolumini-
nescence and its application to space-based damage sensors. Proceedings of the Propulsion
Measurement Sensor Development Workshop. Huntsville, Alabama.
6. Kawaguchi, Y. (1998). Fractoluminescence spectra in crystalline quartz. Japanese Journal of
Applied Physics: Part 1-Regular Papers Short Notes and Review Papers, 37, 1892–1896.
7. Reynolds, G. T. (1997). Piezoluminescence from a ferroelectric polymer and quartz. Journal
of Luminescence, 75, 295–299.
8. Chandra, B. P., Elyas, M., & Majumdar, B. (1982). Dislocation models of mechanolumi-
nescence in [gamma]- and X-irradiated alkali halides crystals. Solid State Communications,
42, 753–757.
9. Chandra, B. P., & Shrivastava, K. K. (1978). Dependence of mechanoluminescence in
rochelle-salt crystals on the charge-produced during their fracture. Journal of Physics and
Chemistry of Solids, 39, 939–940.
10. Nevshupa, R., & Hiratsuka, K. (2015). Triboluminescence. In E. Gnecco & E. Meyer (Eds.),
Fundamentals of friction and wear on the nanoscale. Cham: Springer International Publishing.
11. Bergeron, N. P., Hollerman, W. A., Goedeke, S. M., Hovater, M., Hubbs, W., Finchum, A.,
et al. (2006). Experimental evidence of triboluminescence induced by hypervelocity impact.
International Journal of Impact Engineering, 33, 91–99.
12. Sweeting, L. M. (2001). Triboluminescence with and without air. Chemistry of Materials, 13,
854–870.
13. Walton, A. J. (1977). Triboluminesence. Advances in Physics, 26, 887–948.
14. Derjaguin, B. V., Krotova, N. A., & Toporov, Y. P. (1981). Emission of high-speed electrons
and other phenomena accompanying the process of breaking adhesion bonds. In J. M. Georges
(Ed.), Tribology series. Amsterdam: Elsevier.
15. Miura, T., Chini, M., & Bennewitz, R. (2007). Forces, charges, and light emission during the
rupture of adhesive contacts. Journal of Applied Physics, 102, 103509.
16. Licoppe, C. (2013). La formation de la pratique scientifique: le discours de l’expérience en
France et en Angleterre (1630–1820), La découverte.
17. Weiser, H. B. (1918). Crystalloluminescence II. The Journal of Physical Chemistry, 22,
576–595.
18. Brenner, M. P., Hilgenfeldt, S., & Lohse, D. (2002). Single-bubble sonoluminescence. Reviews
of Modern Physics, 74, 425.
19. Tsuboi, Y., Seto, T., & Kitamura, N. (2008). Laser-induced shock wave can spark tribolumi-
nescence of amorphous sugars. The Journal of Physical Chemistry. A, 112, 6517–6521.
20. Butyagin, P. Y., Yerofeyev, V., Musayelyan, I., Patrikeyev, G., Streletskii, A., & Shulyak,
A. (1970). The luminescence accompanying mechanical deformation and rupture of polymers.
Polymer Science U S S R, 12, 330–342.
21. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Direct view of stress
distribution in solid by mechanoluminescence. Applied Physics Letters, 74, 2414–2416.
22. Urbakh, M., Klafter, J., Gourdon, D., & Israelachvili, J. (2004). The nonlinear nature of
friction. Nature, 430, 525–528.
23. Chandra, B. P. (1998). Luminescence of solids. New York: Plenum Press.
24. Chandra, B. P., Baghel, R. N., & Chandra, V. K. (2010). Mechanoluminescenct glow curve of
ZnS:Mn. Chalcogenide Letters, 7, 1–9.
25. Lu, H.-Y., & Chu, S.-Y. (2004). The mechanism and characteristics of ZnS-based phosphor
powders. Journal of Crystal Growth, 265, 476–481.
26. Grmela, L., Macku, R., & Tomanek, P. (2008). Near-field measurement of ZnS:Mn nanocrys-
tal and bulk thin-film electroluminescent devices. Journal of Microscopy (Oxford), 229,
275–280.
27. Manzoor, K., Vadera, S. R., Kumar, N., & Kutty, T. R. N. (2004). Multicolor electrolumines-
cent devices using doped ZnS nanocrystals. Applied Physics Letters, 84, 284–286.
28. Suyver, J. F., Wuister, S. F., Kelly, J. J., & Meijerink, A. (2001). Synthesis and Photolumi-
nescence of Nanocrystalline ZnS:Mn2+. Nano Letters, 1, 429–433.
29. Wood, V., Halpert, J. E., Panzer, M. J., Bawendi, M. G., & Bulovic, V. (2009). Alternating
current driven electroluminescence from ZnSe/ZnS:Mn/ZnS nanocrystals. Nano Letters, 9,
2367–2371.
30. Chandra, B. P., Xu, C. N., Yamada, H., & Zheng, X. G. (2010). Luminescence induced by
elastic deformation of ZnS:Mn nanoparticles. Journal of Luminescence, 130, 442–450.
31. Kobyakov, I. B., & Pado, G. S. (1968). Investigation of electrical and elastic properties of
hexagonal zinc sulfide in temperature range 1.5-300 degrees K. Soviet Physics Solid State, 9,
1707.
32. Chandra, B. P., & Rathore, A. S. (1995). Classification of mechanoluminescence. Crystal
Research and Technology, 30, 885–896.
33. Sage, I., & Bourhill, G. (2001). Triboluminescent materials for structural damage monitoring.
Journal of Materials Chemistry, 11, 231–245.
34. Chudacek, I. (1966). Influence of pressure on recombination centres in piezoelectric
luminophores. Czechoslovak Journal of Physics, 16, 520–524.
35. Chudacek, I. (1967). Kinetics of triboluminescence of zinc sulphide I. Czechoslovak Journal of
Physics, 17, 34–42.
36. Chandra, B. P., Bagri, A. K., & Chandra, V. K. (2010). Mechanoluminescence response to the
plastic flow of coloured alkali halide crystals. Journal of Luminescence, 130, 309–314.
37. Allison, S. W., & Gillies, G. T. (1997). Remote thermometry with thermographic phosphors
instrumentation and applications. The Review of Scientific Instruments, 68, 2615–2650.
38. Chandra, B. P., Baghel, R. N., Singh, P. K., & Luka, A. K. (2009). Deformation-induced
excitation of the luminescence centres in coloured alkali halide crystals. Radiation Effects and
Defects in Solids, 164, 500–507.
39. Chandra, B. P., Singh, S., Ojha, B., & Shrivastava, R. G. (1996). Mobile interstitial model and
mobile electron model of mechano-induced luminescence in coloured alkali halide crystals.
Pramana: Journal of Physics, 46, 127–143.
40. Molotskii, M. I., Poletaev, A. V., & Shmurak, S. Z. (1989). Dislocation-induced sensibilization
of photoexoelectronic emission. Fizika Tverdogo Tela, 31, 14–20.
41. Chakravarty, A., & Phillipson, T. E. (2004). Triboluminescence and the potential of fracture
surfaces. Journal of Physics D: Applied Physics, 37, 2175–2180.
42. Destriau, G. (1936). Recherches sur les scintillations des sulfures de zinc aux rayons@,
R. Bussière.
43. Shionoya, S., Yen, W. M., & Hase, T. (1999). Phosphor handbook. Boca Raton, FL: CRC
Press.
44. Hurt, C. R., Mcavoy, N., Bjorklund, S., & Filipescu, N. (1966). High intensity tribolumines-
cence in europium tetrakis (dibenzoylmethide)-triethylammonium. Nature, 212, 179–180.
45. Hollerman, W. A., Fontenot, R. S., Bhat, K. N., Aggarwal, M. D., Guidry, C. J., & Nguyen,
K. M. (2012). Comparison of triboluminescent emission yields for 27 luminescent materials.
Optical Materials, 34, 1517–1521.
46. Olawale, D. O., Dickens, T., Lim, A., Okoli, O., Wang, B., & Sobanjo, J. O. (2010).
Characterization of the triboluminescence (TL) performance of ZnS:Mn under repeated
mechanical loading for smart optical damage sensor system. NDE/NDT for highways and
bridges: Structural materials and technology (SMT) 2010. New York, USA: American Society
of Non-Destructive Testing (ASNT).
47. Sage, I., Humberstone, L., Oswald, I., Lloyd, P., & Bourhill, G. (2001). Getting light through
black composites: Embedded triboluminescent structural damage sensors. Smart Materials
and Structures, 10, 332–337.
48. Womack, F. N., Goedeke, S. M., Bergeron, N. P., Hollerman, W. A., & Allison, S. W. (2004).
Measurement of triboluminescence and proton half brightness dose for ZnS:Mn. IEEE Trans-
actions on Nuclear Science, 51, 1737–1741.
49. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Preparation and characteristics
of highly triboluminescent ZnS film. Materials Research Bulletin, 34, 1491–1500.
50. Chandra, B. P., & Zink, J. I. (1980). Triboluminescence and the dynamics of crystal fracture.
Physical Review B, 21, 816–826.
51. Kim, J. S., Kwon, Y. N., Shin, N., & Sohn, K. S. (2007). Mechanoluminescent SrAl2O4: Eu,
Dy phosphor for use in visualization of quasidynamic crack propagation. Applied Physics
Letters, 90, 241916.
52. Sohn, K. S., Seo, S. Y., Kwon, Y. N., & Park, H. D. (2002). Direct observation of crack tip
stress field using the mechanoluminescence of SrAl2O4:(Eu, Dy, Nd). Journal of the American
Ceramic Society, 85, 712–714.
53. Xu, C. N., Zheng, X. G., Akiyama, M., Nonaka, K., & Watanabe, T. (2000). Dynamic
visualization of stress distribution by mechanoluminescence image. Applied Physics Letters,
76, 179–181.
54. Sage, I., Badcock, R., Humberstone, L., Geddes, N., Kemp, M., & Bourhill, G. (1999).
Triboluminescent damage sensors. Smart Materials and Structures, 8, 504–510.
Chapter 2
Nature of the Electronic Charge Carriers
Involved in Triboluminescence
Friedemann T. Freund
2.1 Introduction
Luminescence describes the emission of light from matter, mostly solids, that is not
generated by heat. Different actions can cause luminescence. Triboluminescence,
for instance, is caused by mechanical action. Some major questions arise: How does
mechanical action produce luminescence and why can it be observed in some
materials but not in others?
When solids are subjected to compressive or tensile forces, they deform. At first,
within the elastic range, deformation is proportional to the applied force. At higher
stresses deformation becomes non-linear. Eventually the solids will fracture creat-
ing new surfaces.
Microscopically, fracturing means that interatomic bonds are broken as a fracture
propagates through the solid medium creating two opposite surfaces, which begin to
separate. Because of the stochastic nature of any fracture event these fracture
surfaces will contain atoms/ions with dangling bonds and will carry patches of
charges of opposite signs. As the surfaces separate, electric (E) fields develop,
which can be high to very high. These E fields momentarily accelerate electrons
and ions emitted from the fracture surfaces. The E fields are particularly high at the
tip of propagating cracks—high enough to accelerate electrons and ions over short
distances to such high energies that they impact-ionize gas neutrals, creating addi-
tional electrons and ions, and causing avalanche electric discharges. Such discharges
produce bursts of light with spectra extending through the visible (VIS) into the
ultraviolet (UV) and even into the X-ray region [1]. The high-energy portion of this
light will surely interact with the bulk of the solid and initiate secondary reactions.
F.T. Freund (*)

GeoCosmo Science Center – NASA Ames Research Park, Bldg 19,
Suite 1070-I, Moffett Field 94035, CA, USA
e-mail: friedemann.t.freund@nasa.gov

18 F.T. Freund
Interesting from the viewpoint of triboluminescence are cases when the emitted
light is colored and does not only come from the narrow gap between the fracture
surfaces or from the fracture surfaces themselves, but from the inside of the bulk.
Often triboluminescence is produced by relatively gentle mechanical actions such
as rubbing or tumbling. Though rubbing or tumbling may cause spallation and
microfractures on very small scales, they are a far cry from the highly energetic
processes that take place during macroscopic fracture. If there is persistent light
emission coming from within the bulk during rubbing or tumbling, we have to
consider mechanisms, by which relatively large bunches of energy are created with
very small input of mechanical energy, flowing from the surface or near-surface
region, where the mechanical action had taken place, into the underlying bulk.
This chapter deals primarily with oxide materials. It reports on a specific type of
point defects in oxide materials that has been consistently overlooked by the
scientific community: peroxy defects, where oxygen changes its valence from 2
to 1. Peroxy defects occur in many oxides and in silicates. They occur in rocks. It
will be argued here that peroxy defects and the highly mobile positive hole charge
carriers, which they engender, play a major role in energy transfer processes that
allow triboluminescence to occur at sites, which are some distance away from the
sites of the mechanical action. At the end of this chapter the insight gained from
studying peroxy defects and positive holes in oxide materials will be extended to
non-oxide materials.
2.2 Peroxy—The Stealth Defects
2.2.1 Band Structure and Electronic Charge Carriers
Any form of luminescence requires the formation of excited states, commonly

electronically excited states, and their de-excitation via radiative transitions. If
the light emission falls into the visible (VIS) range, their photon energies range
from 1.65 eV in the red (750 nm) to 3.1 eV in the violet (400 nm). Since kT, the
mean thermal energy1 at 300 K, is only 25 meV, the energy emitted over this
spectral region is 25–65 times the amount of energy available through kT. Two
questions thus arise: (1) where does the energy come from that can lead to the
emission of photons in the VIS region and (2) how does this energy travel through
the solid medium from the site of mechanical action to the site of photon emission.
To start it may be good to look at some fundamental issues regarding electronic
charge carriers in solids. Figure 2.1a and b sketches the valence and conduction
bands for semiconductors and insulators. respectively. The difference between the
two classes of materials lies in the width of the bandgap Eg. In a semiconductor Eg is
sufficiently narrow for some electrons from the fully occupied valence band to be
1
With k being the Boltzmann constant and T the absolute temperature.
2 Nature of the Electronic Charge Carriers Involved in Triboluminescence 19
Fig. 2.1 (a, b) Schematic representation of the band structure of (a) a semiconductor and (b) an
insulator, both with fully occupied valence bands. Eg is the width of the band gap. The arrows
indicate that electrons can be thermally activated (by kT) into the conduction band or into impurity
levels in the band gap, respectively. Figure b indicates that, when Eg is much wider than the
thermal energy kT as is the case in insulators, electrons cannot be thermally activated to the
conduction band. However, if impurities create energy levels in the band gap that can act as
acceptors, electrons from the valence band can be promoted to these impurity levels. Unless their
density is so high that their wave functions partly overlap, electrons on those impurity levels will
be localized and do not contribute to the electrical conductivity
thermally promoted into the conduction band due to the thermal energy kT, where
k is the Boltzmann constant and T the absolute temperature. The promotion of an
electron into the conduction band leaves a defect electron in the valence band, also
known as a hole [2]. In the intrinsic case the ratio of the number of electrons in the
conduction band n0 and the number of holes in the valence band n• is 1. Electrons
and holes are both mobile, albeit with different mobilities, μ0 and μ•, respectively,
with electrons being typically more mobile than holes.
2.2.2 Peroxy Defects in Oxide Materials, Minerals,

and Rocks
The O2 anions in oxide materials are commonly assumed to be in the 2 valence
state and in the 2 valence state only. However, oxygen can exist in two valence
states, 1 and 2. A peroxy defect consists of two oxygen anions oxidized from
20 F.T. Freund
Fig. 2.2 (a) Schematic representation of the effect of a peroxy defect on the energy surface of the
valence band of an oxide insulator. The majority 2 oxygen anions are shown in dark gray. The
white dumbbell represents a dip in the energy surface of the valence band due to the presence of a
peroxy defect. (b) When a peroxy defect becomes activated, it generates two states: an electron
that remains trapped in the broken peroxy bond and a hole that delocalizes over many neighboring
O2 and becomes as mobile charge carrier
2 to 1. Figure 2.2a and b sketches a situation, where a peroxy defect is introduced
into an oxide material. It is equivalent to two holes trapped on two adjacent O2 sites.
Since the 1 valence state primarily affects energy levels of O 2sp symmetry at the
upper edge of the valence band, peroxy defects manifest themselves by a local dip in
the energy surface of the valence band as indicated in Fig. 2.2a.
When the O–O bond breaks, an electron can be transfered into the broken
peroxy bond from an outside O2. This electron gets trapped in the broken peroxy
bond, occupying an energy level slightly below the edge of the valence band as
indicated by lower dashed line in Fig. 2.2b. By symmetry an empty energy level is
created slightly above the edge of the valence band, as indicated by the upper dashed
line in Fig. 2.2b. The donor O2 turns into an O, i.e., into a defect electron in the
O2 sublattice, e.g., a hole. As will be argued further below this hole state tends to
delocalize over many O2 neighbors as indicated by the reddish hue in Fig. 2.2b.
Because of this and of other rather remarkable properties associated with this hole
state in the oxygen anion sublattice, it has been given the name “positive hole” [3].
Much of what follows from here onward derives from the noteworthy properties
of positive holes, which the peroxy defects release when they break up. A major
point will be to examine how peroxy defects and positive holes contribute to
luminescence phenomena and specifically to triboluminescence.
2.2.3 Nature of Electronic Charge Carriers, Electrons,

and Holes, in Insulating Oxide Materials
Peroxy defects are a family of point defects in oxide materials that have not
received the attention they probably deserve. Though they seem to be ubiquitous
across a wide range of oxide materials and rock-forming minerals, their presence
has been largely overlooked.
Peroxy defects consist of pairs of oxygen anions, which have changed their
valence from the usual 2 state to 1, where O is more oxidized than O2. The
two O are covalently bonded, forming a very short O–O bond, only ~1.5 Å as
compared to the usual 2.8–3.0 Å distances between adjacent O2. Though the
activation energy to dissociate peroxy bonds is relatively high, on the order of
2.4 eV in MgO [4] and probably similar in other oxide matrices, they are also quite
labile. They break up, when an electron is transferred from some nearby O2 into
the peroxy bond and becomes trapped, leaving one O with the broken peroxy
bond. This O is stationary, while the donor O2 turns into O and becomes a
mobile positive hole charge carrier, which can move away from its point of origin.
There is strong evidence that the wave function associated with this mobile O
state, e.g. with the positive hole, is highly delocalized over the O 2sp-type energy
levels that form the upper edge of the valence band [5, 6]. The peroxy dissociation
can of course reverse by the mobile hole recombining with another mobile hole or
with a defect-bound O.
O have a strong propensity to take over an electron to return to the O2 state.
Thus they act as oxidizing agents. When positive holes roam through the bulk, they
can interact with transition metal cations and participate in charge transfer pro-
cesses. This is the reason why positive holes are interesting actors in the context of
luminescence, including triboluminescence.
Since peroxy defects and positive holes have not been widely discussed in the
literature yet—except in papers by the present author—the work presented here and
the references quoted will primarily tap this source of information. Magnesium
oxide, MgO, played a major role in the discovery of peroxy defects and positive
holes. Subsequently peroxy defects and positive hole-type charge carriers have
been shown to also exist in other oxide materials, including silicates. There is
evidence that similar defects also exist in non-oxide materials.
2.3 Discovery and Validation of Peroxy Defects
2.3.1 Formation of Oxygen Anions in the 1 Valence State
Peroxy defects were first observed in MgO, the structurally simplest oxide, face-
centered cubic. As a main group element Mg has only one chemically stable
22 F.T. Freund
Fig. 2.3 During thermal

decomposition of Mg(OH)2,
made from ultrahigh-purity
MgO, H2O is not the only
gas that evolves but also H2
and atomic O [7]
oxidation state, 2þ. The thermal decomposition of Mg(OH)2 should produce MgO
plus H2O as the only gaseous product:
2 MgðOHÞ2 , 2 MgO þ H2 O ð2:1Þ
However, during the thermal decomposition of ultrahigh purity Mg(OH)2 substan-

tial amounts of H2 were observed [7]. Figure 2.3 shows that the reaction started with
H2O evolving first, around 200 C. After the decomposing Mg(OH)2 had begun to
recrystallize to nano-sized MgO [8], H2 began to evolve around 300 C with
maxima around 450 C and 750 C, the latter accompanied by the evolution of
atomic O starting at 600 C. Note: The H2 recorded above 900 C does not come
from the MgO sample but from the walls of the fused silica tube in which the
decomposition reaction was conducted.
Here are more details. The dehydroxylation of the Mg(OH)2 begins with the loss
of H2O:
MgðOHÞ2 , Mg1δ ðOHÞ2δ O12δ þ δ H2 O ð2:2Þ
During this initial step the hexagonal structure of Mg(OH)2 is preserved, while H2O
molecules are removed from between the layers, probably up to 90 % of the total
OH [9]. The resulting highly defective, hexagonal structure then becomes unstable
and collapses to form nanosized cubic MgO, still retaining a high concentration of
residual OH [10]. As Fig. 2.3 shows the H2 evolution begins after the cubic MgO
nanocrystal had started to form [9].
The face-centered cubic MgO structure offers only a few sites for OH to be
incorporated as depicted on the left of Fig. 2.4 [11]. There are three types of
“impurity” OH:
Fig. 2.4 Projection of the different crystallographically allowed “impurity” OH onto the (1 1 0)
plane of MgO. Left: before redox conversion of OH pairs. Right: After redox conversion of OH
pairs and formation of a peroxy anion and molecular H2. Top row: Kr€ oger-Vinck point defect
designation for the different defects formed
I. OH pairs adjacent to Mg2þ vacancy sites (expected to be most abundant).

II. Single OH at Mg2þ vacancy sites (expected to be less abundant).
III. OH with the O–H vector pointing into an interstitial site (also less abundant).
To describe these impurity OH we use the Kr€oger-Vinck point defect nomen-
clature [12],2 where OH pairs at Mg2þ vacancy sites (I) are written as [OH
00
VMg HO]x, with two hydroxyl Hþ at O2 sites compensating for the 2– charge
associated with the Mg2þ vacancy. This charge-compensated OH pair defect
(I) should be most abundant. It can dissociate into defect types II and III, carrying
a single-negative and single-positive charge, respectively:
h 00
ix h 00
i0
OH VMg HO , OH VMg þ OHi ð2:3Þ
I II III
Analysis of the IR absorption spectra of MgO single crystals containing residual

OH however, indicated low concentrations of the OH pair defect I [11]. Instead
an IR band was observed, characteristic of the H–H stretching mode νHH of
molecular H2 in solid matrices [13]. The assignment of this band to molecular H2
on interstitial sites has been confirmed by the νHD band in MgO single crystals
grown from a melt saturated with a D2O–H2O mixture [11].
2
Square brackets outline the essential parts of the point defects. V stands for vacancy. Subscript i
means interstitial and subscripts identify the crystallographic site (except for oxygen sites, where
subscripts are omitted). Superscripts prime, dot, and x designate single-negative, positive, and
neutral charges, respectively, while double prime and double dot designate double-negative and
positive charges, respectively, relative to the unperturbed crystal structure.
24 F.T. Freund
The disappearance of most of the OH pair defects I from the IR spectrum and
appearance of molecular H2 in the MgO matrix can only be accounted for by
assuming a redox conversion:
h ix x
00 O 00
OH VMg HO , ðH2 ÞMg ð2:4aÞ
O
Equation (2.4a) indicates that OH pairs at Mg2þ vacancy sites change into a
peroxy anion, O22, plus molecular H2. Peroxy defects can also be described as
two adjacent O2 having trapped two holes, converting them into O, which then
tie a peroxy bond, O–O.
As H2 molecules are diffusively mobile, even in structurally dense MgO, they
can move away, leaving the Mg2þ vacancy behind, chargewise compensated by the
peroxy anion, O22:
h ix x
00 O 00
OH VMg HO ) VMg þ ðH2 Þi x ð2:4bÞ
O
In the case of nanocrystalline MgO, the distance for the interstitial H2 molecules to
diffuse of the nanograins is short, allowing the H2 to easily degas as demonstrated in
Fig. 2.3:
h ix x
00 O 00
OH VMg HO ) VMg þ ðH2 Þgas ð2:4cÞ
O
Equation (2.4b) leads to cation-deficient Mg1–δO with most OH converted to H2

on interstitial sites:
Mg1δ ðOHÞ2δ O12δ , Mg1δ ðH2 Þδ O ð2:5aÞ
With H2 diffusing out as of Eq. (2.4c), a non-stoichiometric MgO is left behind:
Mg1δ ðH2 Þδ O ) Mg1δ O þ δ H2 ð2:5bÞ
If δ 1, we may also write the cation-deficient Mg1–δO as an MgO with excess-

oxygen, MgO1þδ.
2.3.2 Break-Up of Peroxy Defects and Formation of Positive

Holes
Peroxy defects are not unique to MgO. They exist across a range of oxide materials
and even seem be ubiquitous—so much so that it is surprising that their very
existence has escaped the attention of the scientific community. It is most likely due
to their stealth nature, which has made it difficult to detect them and to study their
effects on various physical properties of the materials. In order to gain further
insight into peroxy defects and how they generate highly mobile positive hole
charge carriers, MgO can continue to serve as a model, followed by a discussion
of peroxy defects in other matrices, in particular silicates.
The O–O distance in the peroxy defect is only 1.5 Å as compared to 3.0 Å
between regular O2 in MgO. With the two O spins antiparallel and tightly
coupled, the O22 entity is diamagnetic. The shortness of the O–O distance
suggests a very strong peroxy bond. There are several ways to activate and break
it: (1) by heating, (2) by mechanical stress, and (3) by UV light.
Upon heating the O–O bond first loosens up, starting below 200 C, while
staying diamagnetic []. The persistence of diamagnetism suggests that the antipar-
allel spin coupling is maintained producing a bipolaronic state [5]. Starting around
300 C paramagnetism begins to develop, following second order kinetics, indicat-
ing that the O spins start to flip. In combination these two processes comprise
Phase I of the two-step peroxy break-up [6]:
x h ix
O 00 00
I VMg , O VMg O ð2:6aÞ
O
The subsequent Phase II, which sets in around 430 C, leads to the dissociation of
the O–O bond and generation of unbound O, a mobile positive hole charge
carrier, also following a second-order kinetics:
h 00
ix h 00
i0
II O VMg O , O VMg þ O ð2:6bÞ
A condition sine qua non for the dissociation of the peroxy defect is the transfer of
an electron into the already partly broken peroxy bond. The source of this electron
is not known, but it can only come from an O2 outside the peroxy defect. The
donor O2 thereby turns into O, i.e., a positive hole h•. This positive hole is not
spatially close to the Mg2þ cation site. Due to slight lattice rearrangements around
00
the [OVMg ]0 site, which serve to screen its associated negative charge [], this h• is
able to turn into a mobile charge carrier that is free to move away from its
“birthplace” and to roam through the solid matrix.
With a band gap as wide as 7.8 eV [14] MgO should be a near-perfect electrical
insulator. Indeed, in accordance with the sequence of reactions described by the
Eqs. (2.4a, 2.4b, 2.4c)–(2.6a, 2.6b, 2.6c), the electrical conductivity of nominally
highest purity MgO single crystals remains low up to about 430 C, the temperature
at which mobile h• charge carriers appear as a result of the dissociation reaction as
given by Eq. (2.6b). The activation energy for the dissociation reaction is ~2.4 eV,
while the activation energy for the h• thus generated is around 1.0 eV [4].
26 F.T. Freund
If the dissociation of the peroxy bond requires ~2.4 eV, then energies as high as
00
~2.4 eV can be expected to become available when [OVMg ]0 sites recapture h•,
reconstituting the peroxy defects. Similar amounts of energy may be released
through other recombination reactions involving h• charge carriers.
In the context of luminescence and triboluminescence heating is not of primary
interest. More relevant is mechanical action at different levels of stress and different
stress rates, including crushing and fracturing. An alternative way to break peroxy
bonds is by photodissociation using UV photons of sufficiently high energy. In both
cases, mechanical and UV, Step I and Step II coalesce into a single step:
x h i0
O 00 00
V , O VMg þ O ð2:6cÞ
O Mg
Figure 2.5 (right) depicts the one-step dissociation of a peroxy defect in MgO
according to Eq. (2.6c). It leaves the Mg2þ vacancy with one O and an effective
00
charge of 1. The [OVMg ]0 is depicted as an Mg2þ vacancy with the hole state
equally spread over its nearest neighbors. This is the V center, a paramagnetic site
that has been studied extensively by electron spin resonance, ESR, in alkaline earth
oxides [15] as well as in many minerals [16].
The positive hole identified as the unbound O state, h•, is shown to the right as a
delocalized charge spread over many O2 as indicated by different hues of pink.
Evidence for extensive delocalization of the positive hole wavefunction was first
derived from the anomaly in the thermal expansion behavior of high purity MgO
single crystals and by measurements of the dielectric polarization in an electric field
gradient at the limit of 0 Hz [6].
2.3.3 Peroxy Defects in Silica and Silicates
In silica and silicates, peroxy defects are introduced by the same basic redox
conversion as in MgO, starting from hydroxyl pairs, O3Si–OH HO–SiO3, which
form peroxy links O3Si–OO–SiO3 plus molecular H2 as illustrated in Fig. 2.6.
While discussing the formation and activation of peroxy defects in MgO, heating
was the method of choice to investigate some of the basic properties of peroxy
defects. In the context of triboluminescence, the application of stress is more
relevant. Stresses can activate peroxy defects by changing the O3Si–OO–SiO3
bond angle as illustrated in Fig. 2.7.
To illustrate this further we look at the energy levels of the peroxy entity from a
simplified molecular orbital (MO) perspective as illustrated in Fig. 2.8, where we
plot the energy as a function of the angle of deformation of the peroxy bond, α. Each
O has one electron less than the closed-shell configuration of O2. Hence, a
characteristic feature of the O–O bond is that, in the peroxy bond, the strongly
antibonding σ-type orbital is empty, which controls the interaction of two adjacent
Fig. 2.5 Representation of a peroxy anion (left) and a V center plus a positive hole (right) in the
(1 1 0) plane of MgO
Water splits Hydroxyl pair Peroxy bond forma

Si-O-Si bond rearranges electronically plus H2
e- e-
H2O + O3Si-O-SiO3 O3Si-OH HO-SiO3 O3Si-OO-SiO3 + H2
Fig. 2.6 Hydrolysis of an O3Si–O–SiO3 bond to form an O3Si–OH pair, which undergoes a redox
conversion to a peroxy link, O3Si–OO–Si O3 plus H2. Each step is reversible unless the H2 diffuses
away from the site where it was generated
Si-OO-Si bond Neighboring O2- Peroxy w/trapped electron

is being bent transfers electron plus positive hole
e-
O3Si-OO-SiO3 + O2- O3Si-O:O-SiO3 + O- O3Si-O•O-SiO3 + O-
Si-OO-Si bond Transition state Electron e' plus hole h-
Fig. 2.7 Changing the O3Si–OO–Si O3 bond causes it to break and for an electron transfer to take
place from a nearby O2. The donor O2 turns into O, which becomes a mobile positive hole,
while the transferred electron becomes trapped in the broken peroxy bond
O2 and causes them to repel each other. This empty orbital is indicated by a dotted
line in Fig. 2.8. Therefore the highest occupied level is now the non-bonding π-type
MO formed by the overlap of the O 2px and 2py atomic orbitals, if z defined as the
direction between the two O. This π-type MO is fully occupied with 4 electrons,
all of the same energy, as indicated by the solid line at α ¼ 0 of the left of Fig. 2.8.
Changing the angle α shifts the energy levels of all MO’s. Assuming that the
plane of the drawing in Fig. 2.9 is the x–z plane, the MO which derives from O 2px
will destabilize, i.e., its energy will be raised, while the MO which derives from O
2py will stabilize, i.e., its energy will decrease. At some critical angle α, the
28 F.T. Freund
Fig. 2.8 Simplified energy diagram for the highest occupied and the lowest unoccupied molecular
orbitals describing the O3Si–OO–SiO3 bond. As the Si–OO–Si bond angle changes, the relative
positions of the energy levels change. Eventually the occupied nonbonding π-type MO crosses the
empty antibonding σ-type MO, causing the break-up of the peroxy bond
occupied π-type MO will cross the empty σ-type MO and the electrons will flow
from the non-bonding π-type MO into the antibonding σ-type MO, causing the size
of the two O to suddenly increase. This is the stage equivalent to partly broken
peroxy bond as described above by Eq. (2.6a) for O22 in MgO.
The right side of Fig. 2.8 retains for illustration purposes the same MO diagram,
with the same relative positions of the energy levels, even though there will
certainly be substantial changes due to near-instant structural rearrangements and
changes in the local geometry around the broken peroxy bond. The right side
indicates that an transfer of an electron is taking place from an O2 outside the
peroxy defect as described by Eq. (2.6c) for O22 in MgO and by Fig. 2.7 for the
break-up of the O3Si–OO–SiO3 bond.
2.3.4 Propagation of the Positive Hole Charge Carriers
As defect electrons in the O2 sublattice, positive holes reside in the O 2sp-type
energy levels that form the upper edge of the valence band. Though MgO is among
the most ionic oxide materials with the ionicity of the Mg2þ–O2 interaction in the
90 % range [17], the residual covalency provides an energetic continuum along
which hole states can propagate, roaming through the oxide matrix and even cross
grain boundaries.
Fig. 2.9 Proposed phonon-assisted electron hopping mechanism for the propagation of positive
holes through the bulk of an oxide or silicate material
The structural rearrangements around the broken peroxy bond will affect the
reversibility of the reaction and therefore the lifetime of the positive holes states,
once they are generated and free to roam through the matrix of the material. The
mode of propagation of the positive holes is probably via a phonon-assisted electron
hopping mechanism as has been suggested for the delocalization of the hole state
over the nearest neighbors of the V center in MgO [18].
Figure 2.9 illustrates the phonon-assisted electron hopping mechanism. Assum-
ing that the positive hole can be represented by a point charge, its propagation, say
from left to right as indicated in Fig. 2.9, occurs by electrons hopping from right to
left. Taking the phonon frequency at ambient temperature to be 1012 Hz and the
transfer distance at each jump to be 2.8 Å (2.8 1010 m), the maximum speed
would be ~280 m/s. If the propagation is by random walk, the speed in any one
direction will be about 1/3 of this value, ~100 m/s. The measured speed of
propagation of stress-activated positive hole charge carriers in laboratory rock
samples indicate values in the range [19].
Next the question arises: Where are the peroxy defects located in any given
sample so that they can be activated by different levels of stress?
When hydroxyls are introduced into materials, for instance at high temperatures
during crystallization from a melt or recrystallization during thermal treatments
such a sintering, hydroxyls become invariably incorporated in the host matrix but
they are “impurities.” Hence, as thermodynamics requires, during cooling, there
will always be a driving force to segregate those impurities and to remove them
from the matrix. In polycrystalline materials the most favored sites for segregation
are grain boundaries, subgrain boundaries, and dislocations. They all provide local
stress release accommodating impurities for which the structure does not provide
sites. Hence, impurities such as hydroxyls must segregate.
Figure 2.10 illustrates for a polycrystalline material that, when hydroxyls accu-
mulate on the grain boundaries, many of them will occur in the form of pairs on
adjacent grains, directly facing each other. Those hydroxyl pairs will undergo the
same type of redox conversion depicted in Fig. 2.6, forming peroxy bonds that
straddle the grain boundary. Obviously, such peroxy defects will be extremely
sensitive to ever to slight mechanical stresses that tend to shift grains relative to
each other. Even sub-Å shifts are expected to cause an instant break-up of peroxy
30 F.T. Freund
Fig. 2.10 Schematic representation of a polycrystalline silicate material with hydroxyls segre-
gated to the grain boundaries, where many of them occur as hydroxyl pairs. Those hydroxyl pairs
will undergo the redox conversion to form peroxy defects, which straddle the grain boundaries and
are prone to activation by the slightest shifts between grains
defects and the release of positive hole charge carriers. Something similar would be
expected for hydroxyls having segregated to subgrain boundaries and dislocations
in single-crystal materials though the required stress levels will be higher.
As mentioned above, due to local structural rearrangements in the neighborhood
of broken peroxy bonds, the positive holes are expected to not be able to recombine
as readily as they had broken apart. As a result, once activated, positive holes will
have wide range of lifetimes allowing them to roam through the bulk of the
material.
This can be demonstrated by setting up an experiment as sketched in the inset in
Fig. 2.11, taking a finely grained rock tile, here a gabbro, 30 cm 30 cm 0.9 cm,
fitting it with Cu electrodes around the periphery and in the center, and applying
stress at the center via a pair of 5 cm diameter steel pistons [19]. As soon as a load is
applied, a positive hole current starts to flow from the center to the periphery. At
low loads an initial burst of current is observed, arising from the activation of
peroxy defects along grain boundaries, which release their positive holes. While
these charge carriers flow out of the stressed rock volume, they also recombine.
Thus the magnitude of the initial outflow current is controlled by the balance
between the activation and recombination of the positive holes. The faster the stress
is applied, the less time is available for the positive holes to recombine. Hence, the
outflow current is highest at the fastest stress rates. Figure 2.11 plots the actual
currents measured at different stress rates, changing the stress rates over 5 orders of
magnitude, while Fig. 2.12 plots the peak current.
Because of differences in the local microscopic environments the lifetimes of the
positive holes differ depending on the way stresses are applied. If the stress level is
raised suddenly and then kept constant, average lifetimes up to 2 months have been
observed. When the stresses are removed, the outflow currents return to zero
Fig. 2.11 Initial burst of current flowing out of the stressed rock volume, due to the activation of
peroxy defects sitting on or straddling grain boundaries and, hence, most sensitive to small stress
changes. Inset: Sketch of experimental setup
Fig. 2.12 Peak intensity of

the initial outflow current at
different stress rates,
increasing the stress rates
over 5 orders of magnitude,
indicating a wide spread of
lifetimes
relatively rapidly, though some mobile charge carriers remain active for at least
several hours [19].
No information is currently available about triboluminescence during stressing
of gabbro, but triboluminescence has been observed during stressing of granite up
to near-failure. In this case bursts of light have been reported, increasing in intensity
with increasing stress rates and followed by a long afterglow [20].
32 F.T. Freund
2.3.5 Trapping and Detrapping of Positive Hole Charge

Carriers
As positive holes can leave the site, where they have been created and roam through
the solid matrix, they can interact with various lattice constituents. If transition
metal cations are present in reduced oxidation states such as Fe2þ on Mg2þ sites in
MgO, they can trap the positive holes and turn into Fe3þ, if this oxidation reaction is
energetically favorable:
x
FeMg þ O , FeMg þ Ox
ð2:7aÞ
uncharged þ positive positive þ uncharged
where Fex and Fe• stand, respectively, for ferrous (Fe2þ) and ferric (Fe3þ) on Mg2þ
sites.
This reaction can be experimentally demonstrated by heating MgO doped with
Fe2þ and 3d transition metal cations using electron spin resonance (ESR) spectros-
copy to measure the amount of Fe3þ produced by the reaction given by Eq. (2.7a)
proceeding from left to right. In order to oxidize Fe2þ to Fe3þ MgO containing Fe2þ,
Mn2þ and Cr3þ was heated in a stepwise fashion. Above 430 C, when peroxy
defects become unstable and dissociate according to Eqs. (2.6b) and (2.6c),
positive holes are released. Those positive holes are then trapped by Fe2þ on
Mg2þ sites, oxidizing them to Fe3þ according to Eq. (2.7a) as illustrated by the
solid line in Fig. 2.13. In order to maintain the Fe3þ in the oxidized state and to
be able to record it by ESR spectroscopy, the MgO samples had to be quenched
fast to liquid nitrogen temperature, 77 K. Otherwise, upon slow cooling or
stepwise cooling, the oxidation reaction reverses as illustrated by the dotted
line in Fig. 2.13 with Fe3þ self-reducing back to Fe2þ [21]. This reaction
requires the reconstitution of peroxy bonds according to Eq. (2.7b), which
may contribute as much as 2.4 eV to the total energy balance:
h i0 h ix
00
O 00
O VMg þ O , O VMg ð2:7bÞ
negative þ positive uncharged
Nominally pure MgO single crystals exhibit a bright orange luminescence around
600 nm (2.0 eV), when excited with UV light or high-intensity lasers in the visible
region. It has been proposed [22] that this luminescence arises from hole trapping
by V centers as shown by Eq. (2.7b). Though Duley et al. do not mention the
formation of a peroxy entity at the Mg2þ vacancy site, the quoted reaction describes
the reconstitution of peroxy bonds that had been broken by incident UV photons of
sufficiently high energy to cause a direct electron transition from the valence band
into the conduction band or by non-linear effects during the absorption of VIS
photons during high intensity illumination in the VIS region similar to the Raman
effect [23].
Fig. 2.13 Increase in Fe3þ concentration due to trapping of positive holes by Fe2þ on Mg2þ lattice
sites during stepwise heating beyond 430 C, followed by quenching to 77 K, as indicated by the
solid line. During slow or stepwise cooling Fe3þ self-reduces to Fe2þ as indicated by the dotted
line due to peroxy defects able to form again below 430 C [21]
Similarly, during frictional excitation, triboluminescence of MgO starts at

600 nm (2.0 eV) and peaks around 730 nm (1.7 eV) [24]. Triboluminescence has
also been reported to occur during the low temperature and ambient temperature
deformation of MgO, CaO, and SrO crystals. Though the nature of the emitting
center has not been determined, excitonic transitions from cation vacancy sites are
suspected to be involved [25], also pointing at the reconstitution of peroxy defects
according to Eq. (2.7b).
Nanosized MgO thin films can be made ferromagnetic by doping them with
oxygen, resulting in stoichiometries between MgO and MgO2 [26]. These films
exhibit a strong luminescence at approximately 460 nm (2.7 eV) believed to be due
to holes trapped at Mg2þ vacancy sites [27]—a suggestion that again points to the
reconstitution of peroxy bond according to Eq. (2.7b).
Changing from MgO to CaO, also in the face-centered cubic NaCl-type struc-
ture, allows for larger transition metal cations to be incorporated on Ca2þ sites.
CaO:Sn2þ exhibits photoluminescence at 490 nm (~2.5 eV) with a very long
afterglow when excited at room temperature [28], possibly due to positive holes
being trapped and detrapped by Sn2þ and the reconstitution of peroxy bonds.
Mechanical deformation of nominally pure CaO single crystals introduces an
absorption band at 270 nm (4.6 eV), which thermally anneals above 330 C, and
a luminescence band around 450 nm (2.75 eV), which anneals already around room
temperature [29]. Any CaO crystals grown from the melt by the carbon fusion
34 F.T. Freund
technique will contain peroxy defects analogous to MgO crystals grown by the
same technique [11]. Hence, the capability of mechanically deformed CaO and
MgO [30] to support luminescence is most likely related to the activation of
positive holes and their deactivation through reconstitution of peroxy bonds.
2.4 Luminescence and Triboluminescence in Sulfides

and Other Materials
Another way to allow for the incorporation of large transition metal cations is to
move from oxide to sulfide materials where the S2 anions have significantly larger
ionic radii than O2. MgS, for instance, crystallizing in the same face-centered
cubic NaCl-type structure as MgO, can incorporate rare earth metal cations,
providing a rich field of study for optically and thermally stimulated luminescence
[31]. The same is true for ZnS-based systems [32–34]. In sulfides the role of the
peroxy entity is taken over by disulfide anions, S22, though this analogy is hardly
ever mentioned.
However, because the energy levels at the top of the molecular orbital scheme of
the S22 disulfide and the O22 peroxy entities are quite similar, it is to be expected
that, in solid sulfides, disulfide entities will be present and also dissociate, releasing
hole states into the S2 sublattice, chemically equivalent to S. In this case the S
states will act as charge carriers similar to the O states, able to deliver oxidizing
potential to distant sites in the sulfide matrix, where transition metal cations such as
rare earths reside in low-oxidation states, able to capture the hole and emit photons
as part of a luminescent excitation/de-excitation process.
When it comes to organic luminescent materials [35–38], the concept of hole
states as mobile charge carriers may also be applied. After all, there are only two
basic requirement: (1) that sites should exist, which—upon activation—generate
electron-hole pairs, and (2) that the electrons should become trapped and rendered
immobile, while the holes would be able to move along the upper edge of the
valence band.
2.5 Summary
Triboluminescence is caused by mechanical actions. They may include crushing

and fracturing. In this case, highly energetic processes take place on very short time
scales during propagation of cracks, leading to the emission of bursts of light from
the fracture surfaces and from the space between them. This emission covers a
broad spectral range, extending into the UV and possibly into the X-ray region.
If the mechanical actions are more gentle such as during rubbing and tumbling,
but nonetheless cause luminescence emerging from within the bulk, the question
arises as to how does such weak mechanical action can lead to luminescence and
how does the necessary electronic excitation flow from the surface into the bulk.
The answer may lie in the presence of certain defects in the matrix of oxide
materials, which have not been paid the attention they probably deserve: peroxy
defects, which consist of pairs of oxygen anions that have changed their valence
from 2 to 1. Under normal conditions the two O are tightly coupled and strictly
localized, forming a very short O–O bond. However, this peroxy bond is also
very labile. Slight variations in the local environment can cause the peroxy bonds to
break, even transient perturbations by sound waves. When this happens, electron-
hole pairs are generated with the electrons trapped in the broken peroxy bonds and
the holes turning into highly mobile charge carriers, “positive holes.” Those
positive holes have the ability to flow away from the sites where they had been
generated, propagating via a phonon-assisted electron hopping mechanism at
speeds up to about 100 m/s, traveling far afield.
Chemically, the positive holes represent O in a matrix of O2. They have a
strong propensity to take over one additional electron. This means that they are
strongly oxidizing, able to oxidize transition metal cations, which they may
encounter as they roam through the matrix of the oxide materials. Such oxidation
reactions are often strongly exothermal, causing the transition metal cations to
become electronically excited. If they radiatively de-excite, luminescence will
occur.
Positive holes can also recombine with broken peroxy bonds, reconstituting the
peroxy defects. It is known that the activation energy required to dissociate peroxy
bonds in MgO is ~2.4 eV [4] and probably of similar values in other oxide matrices.
Hence, the recombination of a positive hole with a broken peroxy bond may yield
energies up to ~2.4 eV. If the recombination takes place at a site of a transition
metal cation, this energy might be transferred from the newly forming peroxy
defect onto the transition metal cation, causing it to become electronically excited,
able to subsequently de-excite by emitting a luminescence photon.
Acknowledgments The early part of this work was supported by the Deutsche Forschungsge-
meinschaft in Germany, and the latter part primarily by NSF and NASA. Current support by
NASA is provided by the grant #NNX12AL71G. I thank many students and coworkers who
participated in this research and helped me develop the knowledge base, on which this chapter is
based. Their names are listed as coauthors in the references given below.
References
1. Langford, S. C., & Dickinson, J. T. (2009). Emission of particles and photons from the fracture
of minerals and inorganic materials. In Spectroscopic characterization of minerals and their
surfaces (pp 224–244). ACS Symposium Series, Vol. 415, Chapter 12, doi: 10.1021/bk-1990-
0415.ch012.
2. Kittel, C. (1996). Introduction to solid state physics (7th ed.). New York: Wiley.
3. Griscom, D. L. (1990). Electron spin resonance. Glass Science and Technology, 48, 151–251.
36 F.T. Freund
4. Kathrein, H., & Freund, F. (1983). Electrical conductivity of magnesium oxide single crystal
below 1200 K. Journal of Physics and Chemistry of Solids, 44, 177–186.
5. Batllo, F., LeRoy, R. C., Parvin, K., Freund, F., & Freund, M. M. (1991). Positive hole centers
in magnesium oxide—correlation between magnetic susceptibility, dielectric anomalies and
electric conductivity. Journal of Applied Physics, 69, 6031–6033.
6. Freund, F., Freund, M. M., & Batllo, F. (1993). Critical review of electrical conductivity
measurements and charge distribution analysis of magnesium oxide. Journal of Geophysical
Research, 98(B12), 22209–22229.
7. Martens, R., Gentsch, H., & Freund, F. (1976). Hydrogen release during the thermal decom-
position of magnesium hydroxide to magnesium oxide. Journal of Catalysis, 44, 366–372.
8. Freund, F., Scheikh-ol-Eslami, N., & Gentsch, H. (1975). Formation of O centers by
homolytic decomposition of OH groups on magnesium oxide. Angewandte Chemie, Inter-
national Edition, 14(8), 568–569.
9. Freund, T., Martens, R., & Scheikh-ol-Eslami, N. (1975). Recrystallization effect during the
dehydration of magnesium hydroxide. Journal of Thermal Analysis, 8(3), 525–529.
10. Freund, F., & Sperling, V. (1976). A magnesium oxide defect structure of hexagonal symme-
try. Materials Research Bulletin, 11, 621–630.
11. Freund, F., & Wengeler, H. (1982). The infrared spectrum of OH-compensated defect sites in
C-doped MgO and CaO single crystals. Journal of Physics and Chemistry of Solids, 43, 129–
145.
12. Kr€oger, F. A. (1964). The chemistry of imperfect crystals. Amsterdam: North-Holland.
13. Kriegler, R. J., & Welsh, H. L. (1968). The induced infrared fundamental band of hydrogen
dissolved in solid argon. Canadian Journal of Physics, 46(10), 1181–1189. doi:10.1139/p68-
151.
14. Taurian, O. E., Springborg, M., & Christensen, N. E. (1985). Self-consistent electronic
structures of MgO and SrO. Solid State Communications, 55(4), 351–355.
15. Henderson, B., & Wertz, J. E. (1977). Defects in the alkaline earth oxides. London: Taylor &
Francis.
16. Marfunin, A. S. (1979). Spectroscopy, luminescence and radiation centers in minerals (pp.
257–262). New York: Springer Verlag.
17. Pacchioni, G., Sousa, C., Illas, F., Parmigiani, F., & Bagus, P. S. (1993). Measures of ionicity
of alkaline-earth oxides from the analysis of ab initio cluster wave functions. Physical Review
B, 48, 11573–11582.
18. Shluger, A. L., Heifets, E. N., Gale, J. D., & Catlow, C. R. A. (1992). Theoretical simulation of
localized holes in MgO. Journal of Physics: Condensed Matter, 4(26), 5711–5722.
19. Scoville, J., Sornette, J., & Freund, F. T. (2015). Paradox of peroxy defects and positive holes
in rocks: Part II. Outflow of electric currents from stressed rocks. Journal of Asian Earth
Sciences, 114(Part 2), 338–351.
20. Maeda, I. (1986). Activities of triboluminescence at sample failure of granite. Journal of the
Faculty of Science, Hokkaido University. Series 7, Geophysics, 8(1), 65–81.
21. Kathrein, H., Freund, F., & Nagy, J. (1984). O-ions and their relation to traces of H2O and
CO2 magnesium oxide: an EPR study. Journal of Physics and Chemistry of Solids, 45, 1155–
1163.
22. Duley, W. W., & Rosatzin, M. (1985). The orange luminescence band in MgO crystals.
Journal of Physics and Chemistry of Solids, 46(2), 165–170.
23. Harris, D. C., & Bertolucci, M. D. (1989). Symmetry and spectroscopy: An introduction to
vibrational and electronic spectroscopy. New York: Dover Publications.
24. Miura, T., Hosobuchi, E., & Arakawa, I. (2009). Spectroscopic studies of triboluminescence
from a sliding contact between diamond, SiO2, MgO, NaCl, and Al2O3 (0001). Vacuum, 84(5),
573–577.
25. Williams, G. P., & Turner, T. J. (1979). Triboluminescence in single crystal alkaline earth
oxides. Solid State Communications, 29, 3.
26. Mahadeva, S. K., Jincheng Fan, A. B., Sreelatha, K. S., Lyubov, B., & Rao, K. V. (2013).
Magnetism of amorphous and nano-crystallized Dc-sputter-deposited MgO thin films.
Nanomaterials, 3, 486–497.
27. Martı́nez-Boubeta, C., Martı́neza, A., Hernández, S., Pellegrino, P., Antony, A., Bertomeu, J.,
et al. (2011). Blue luminescence at room temperature in defective MgO films. Solid State
Communications, 151(10), 751–753.
28. Donker, H., Smit, W. M. A., & Blasse, G. (1987). On the luminescence of CaO:Sn2+. Physica
Status Solidi, 145(1), 333–342.
29. Gonzalez, R., Chen, C. B. Y., Liu, H., Williams, G. P., Jr., Rosenblatt, G. H., Williams, R. T.,
et al. (1993). Luminescence properties of deformed CaO crystals. Physical Review B, 47, 4910.
30. Melton, R., Danieley, N., & Turner, T. J. (1980). Luminescence of MgO during mechanical
deformation. Physica Status Solidi A, 57(2), 755–764.
31. Chakrabarti, K., Mathur, V. K., Rhodes, J. F., & Abbundi, R. J. (1988). Stimulated lumines-
cence in rare‐earth‐doped MgS. Journal of Applied Physics, 64, 1363–1366.
32. Aich, N., Appalla, A., Saleh, N. B., & Ziehl, P. (2013). Triboluminescence for distributed
damage assessment in cement-based materials. Journal of Intelligent Material Systems and
Structures, 24(14), 1714–1721.
33. Bergeron, N. P., Hollerman, W. A., Goedeke, S. M., Hovater, M., Hubbs, W., Finchum, A., et
al. (2006). Experimental evidence of triboluminescence induced by hypervelocity impact.
34. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2012). Comparison of the
triboluminescent properties for europium tetrakis and ZnS:Mn powders. Journal of Theoret-
ical and Applied Physics, 6, 15.
35. Blasse, G., & Grabmaier, B. C. (1994). Luminescent materials. New York: Springer.
36. Chandra, B. P., & Zink, J. I. (1980). Mechanical characteristics and mechanism of the
triboluminescence of fluorescent molecular crystals. Journal of Chemical Physics, 73(12),
5933–5941.
37. Wang, J., Zhang, M., Zhang, Q., Ding, W., & Su, Q. (2007). The photoluminescence and
thermoluminescence properties of novel green long-lasting phosphorescence materials Ca8Mg
(SiO4)4Cl2:Eu2+,Nd3+. Applied Physics B, 87(2), 249–254.
38. Teotonio, E. E. S., Fett, G. M., Brito, H. F., Faustino, W. M., de Sa, G. F., Felinto, C. F. C.,
et al. (2008). Evaluation of intramolecular energy transfer process in the lanthanide(III) bis-
and tris-(TTA) complexes: Photoluminescent and triboluminescent behavior. Journal of Lumi-
nescence, 128, 193–198.
Chapter 3
Mechanoluminescence of Coordination
Compounds
Ercules Epaminondas Sousa Teotonio, Wagner Mendonça Faustino,

Hermi Felinto Brito, Maria Claudia França Cunha Felinto,
Jandeilson Lima Moura, Israel Ferreira Costa, and
Paulo Roberto Silva Santos
Abbreviations
acac Acetylacetonate
acbz 1,4-Diacetylbenzene
acetbz 1,4-Diacetoxybenzene
APY 4-Aminopyridinium
bipy 2,20 -Bipyridine
BIPYPO 3,30 -Bis(diphenylphosphoryl)-2,20 -bipyridine
BL Bioluminescence
CL Chemiluminescence
dbm Dibenzoylmethanate
dia 4,5-Diazafluoren-9-one
dmbp 4,40 -Dimethyl-2,20 -bipyridinate
dmdpy 4,40 -Dimethyl-2,20 -bipyridyl ligand
DMPY 1,4-Dimethylpyridinium
dmtph 1,4-Dimethyltherephthalate
E.E.S. Teotonio (*) • W.M. Faustino • J.L. Moura • I.F. Costa • P.R. Silva Santos
Department of Chemistry, Federal University of Paraiba, Cidade Universitária, Jo~ao Pessoa,
PB 58051-970, Brazil
e-mail: ercteot@gmail.com; wmfaustino@quimica.ufpb.br; jandeilsonlimamoura@hotmail.
com; israelqi@hotmail.com; roberto.santos.p.s@outlook.com
H.F. Brito
Department of Fundamental Chemistry, University of S~ao Paulo, S~ao Paulo,
SP 05508-000, Brazil
e-mail: hefbrito@iq.usp.br
M.C.F.C. Felinto
Institute of Nuclear Energy and Research, Av. Prof Lineu Prestes 2242 Cidade Universitária,
S~ao Paulo, SP 05508-000, Brazil
e-mail: mfelinto@ipen.br

40 E.E.S. Teotonio et al.
DPEPO Bis[2-(diphenylphosphino)phenyl]ether oxide

dpm Pivaloylmethanate
EL Electroluminescence
FL Fractoluminescence
hfac Hexafluoroacetoacetonate
hmpa Hexamethylphosphoramide
EuD4TEA HNEt3[Eu(dbm)4]
HPPI 3-Phenyl-4-propanoyl-5-isoxazone
LEDs Lighting-emitting devices
LMCT Ligand-to-metal charge transfer states
Ln3þ Trivalent lanthanide ions
LRR ()-4,5-Pinene bipyridine
LSS (þ)-4,5-Pinene bipyridine
ML Mechanoluminescence
MLCT Metal-to-ligand charge transfer
MLQ N-Methylisoquinolium
MOFs Metal-organic frameworks
N2 Nitrogen
NO3 Nitrate ion
PD Photodiode
phen 1,10-Phenantroline
PL Photoluminescence
PyO Pyridine N-oxide
PZL Piezoluminescence
S Singlet state
T Triplet state
TL Triboluminescence
tppo Triphenylphosphine oxide
tta Thenoyltrifluoroacetonate
3.1 Introduction
3.1.1 A Brief on Luminescence
The luminescence exhibited by matter has been sparked the human curiosity for
centuries and has been called as “cold light” to differ from the emission processes
owing to the incandescence “hot light” that is well described by the Planck’s black
body radiation theory [1, 2]. The incandescence is usually produced by a direct
heating of matter at high temperatures, which results in emitting light over a
continuous range of energies depending on the temperature. Whereas, lumines-
cence is characterized by emission of radiations at its normal lower temperatures
when the luminescent material is submitted to external stimuli. These stimuli are
3 Mechanoluminescence of Coordination Compounds 41
Fig. 3.1 Illustration of different types of luminescent phenomena: PL (photoluminescence), EL

(electroluminescence), CL (chemiluminescence), ML (mechanoluminescence) and BL (biolumi-
nescence). The pictures for EL and BL were reproduced with permission from Refs. [10] and [11],
respectively
the driving forces that create excited electronic states, which may after some
relaxation processes undergo radiative decays. Among the processes associated to
these stimuli, some of the most investigated ones are photoluminescence, electro-
luminescence, radioluminescence, cathodoluminescence, thermoluminescence,
chemiluminescence, and mechanoluminescence (ML) (Fig. 3.1) [3, 4].
In photoluminescence (PL) the absorption of photons excites the samples, which
can exhibit emission by either fluorescence (transitions involving states of same
spin angular momenta: ΔS ¼ 0, e.g., S1 ! S0) or phosphorescence (transitions
involving states of dissimilar spin momenta: ΔS 6¼ 0, e.g., T1 ! S0) processes.
Over the last several decades PL materials have found applications as markers,
sensors, probes, and labels for immunoassays [5–7].
In electroluminescence a passage of an electric current results in the emission of
light due to the recombination of electrons and holes into the material. The main
applications of this phenomenon is in lighting-emitting devices (LEDs) for lighting
or display technologies.
Radioluminescence is concerned to the excitation of a material via bombardment
of ionizing radiations, finding applications in scintillators, X-ray imaging for
medical diagnosis [3]. Whereas, cathodoluminescence is the excitation of the
sample by a fast electrons beam (Cathode rays), traditionally used in screens of
television, oscilloscopes, electron microscopes, etc.
In thermoluminescence the sample excitation occurs due to the heating, which

results in re-emitting radiations after a prior energy storage. Owing to the propor-
tionality between the luminescence intensity and the received radiation dose, this
phenomenon is used in dosimetry and dating.
Chemiluminescence (in general) and bioluminescence (particularly in living
systems) are luminescent processes underlined by special chemical reactions that
form products in excited states. The former process is usually employed in analyt-
ical procedures for determination of chemical and biological species, whereas the
last one is exhibited by living organism, such as fireflies, fungi, fishes and so on,
which use this ability to counter-illumination camouflage, attraction, defense,
warning, mimicry, and communication [8].
Mechanoluminescence (ML) comprehends all the phenomena in which lumines-
cence is induced by mechanical action on solids. A wide varieties of mechanical
stimuli on solid, such as compressing, stretching, static loading (pressure-step),
pressure pulse, cleaving, rubbing, and grinding, might induce the excitation of
solids, eventually generating luminescent phenomena [9]. Some of these stimuli
are explored in details in the following sections.
The mechanism associated with each type of luminescence has been intensively
investigated through both experimental and theoretical viewpoints. For some of
these phenomena, nowadays, it is possible to design new materials with optimized
luminescent properties for practical applications.
3.1.2 General Aspects of Mechanoluminescence
Several terms have been employed in literature as a synonym of mechanolumi-

nescence (ML) phenomenon, among them, the triboluminescence (TL), fractolumi-
nescence (FL), and piezoluminescence (PZL) are the most commonly used
[12]. Chandra and co-workers classified ML phenomenon for solid samples as
deformation ML and tribo ML. The former one is based on the deformation stress
limit required for light emission takes place, which was subclassified as elastico-,
plastico- and fracto-ML [9]. The first two processes are non-destructive, whereas the
fracto-ML is destructive, in the sense of crystal cleavage [13, 14] (Figs. 3.2 and 3.3).
On the other hand, the tribo ML is assigned to the ML light emission induced by the
friction (or rubbing) between two dissimilar materials, being subclassified as electri-
cally, chemically, and thermally induced tribo ML, which depend on the excitation
mechanism resulting from the friction [9]. For instance, tribo ML may be either
produced when the sticky sides of the adhesive tapes are separated from each other or
by speeling adhesive tapes from metal or glass surfaces (triboelectric field) [15–17].
In a history point of view, the first light emission has been reported from crystal
cleavage, which was made by the English scholar Francis Bacon in 1605 [19]. It
was noticed when sugar crystals are crushed under darkness and after a certain time
of eye adaptation, a blue light sparks could be observed. In the end of 1920, a
hundreds of inorganic and organic ML materials were known. Thereafter, this
number was increased hugely due to the development of photographic techniques
Fig. 3.2 Schematic flow chart for mechanoluminescence classification phenomenon
Fig. 3.3 Different mechanoluminescence phenomena. (a) Elastico—ML (reprinted with permis-
sion from [18]. Copyright 2009 Elsevier; (b) electrically induced—ML; (c) fractoluminescence
[20–22] and employing photoelectronic detectors [23], being the first ML spectra
were recorded by Longchambon in 1922 [20]. Recently, it is estimated that 50 % of
all known organic and inorganic solids display luminescence when they undergo
cleavage, which have resulted in a great increase in number of peer-reviewed
articles published up to date [9, 12].
Unfortunately, the ML intensity is largely dependent on several experimental
parameters such as size, quality and particularly orientation of the single crystal.
This fact has challenged the quantitative investigation of ML intensity. However,
instrumental techniques have been developed in order to overcome this problem.
Among apparatus the most important are air-driven [24, 25], the impacting needle
[26, 27], the crystal milling [27, 28], the air-driven piston [29], the falling weight
[21], and variant of the falling weight [30]. Other methods to determine the ML
intensities were used by Wick [31] as well as Zink and coworkers [32–34]. More-
over, in term of detectors for ML measurements some special features are required,
for instance, high speed of collection, low cost, portability and also high resolution
and sensitivity [12]. Among these detectors, fiber-coupled multichannel miniatur-
ized spectrometers have been largely applied for this purpose, although their
resolution and sensitivity are compromised [12].
Fontenot and co-workers [35] developed a simple low-cost equipment capable of
quantifying the ML properties such as decay time and light yield with low velocity
impact. The equipment is consisted of a drop tower, steel ball, detector photodiode
a 1.25” Diameter b 1.25” Diameter

130g Steel Ball 130g Steel Ball
42 inches 42 inches
LAPTOP
36 inches 36 inches
PVC Pipe PVC Pipe
24 inches SPECTROMETER
24 inches
12 inches 12 inches
Triboluminescent Material Triboluminescent Material
Under Examination Under Examination
Impact Impact
Chamber Chamber
Drawer Containing Plexiglass Drawer Containing Plexiglass
Photodiode Photodiode
RECORDING
OSCILLOSCOPE
Mirror Mirror
at 45º at 45º
AMPLIFIER
Fig. 3.4 Schematic diagram of the specially designed drop tower used to measure the tribolumi-
nescence: (a) integrated light yield, and (b) visible spectrum. Reprinted with permission from Ref.
[35]. Copyright 2012 Elsevier
(PD) that is positioned under the plexiglass plate, Melles Griot large dynamic range
linear amplifier and a Tektronix 2024B oscilloscope (Fig. 3.4).
ML materials have been proven to be suitable candidates for designing new high
emission intensity compounds that have found potential applications as stress,
fracture [36], and real-time structural damage sensors. In addition, these materials
may be applied in a great number of distinct areas, ranging from commercial to
military markets [12]. Based on the structural damage onset, these sensors can
detect severity of damage as an indirect mesurement of emitted light.
A huge amount of work has been dedicated to both organic and inorganic (salts
and oxides) compounds. The ML properties in coordination compounds were
reported for the first time in 1960s by Cotton and Goodgame [37, 38], and since
then, the studies have been concentrated on transition (mainly Mn2þ and Cu2þ) and
lanthanide compounds [9], for which the obtained ML spectra have shown the
characteristic emission bands generally assigned to the metal ions centered transi-
tions. Nowadays, it is known a great number of ML lanthanide complexes with
different classes of organic ligands, such as β-diketonates, carboxylates, and phos-
phine oxides. However, the ML complexes of Mn2þ ion are generally based on
halides, phosphine, phosphine oxide, and arsine oxide ligands [39].
The principal aim of this chapter is to describe the recent investigations on the
mechanisms of mechanoluminescence and their correlation with structural proper-
ties. A concise overview on the latest progress made on the current variety of ML
coordination compounds or materials based on transition and lanthanide metals is
also presented.
3.2 Mechanoluminescence Mechanisms in Coordination

Compounds
The large number of investigations over the emission induced by mechanical stress
suggest that ML phenomenon may not occur by a single mechanism. The key fact to
understand, which mechanism takes place in coordination compounds is directly
related to the nature of the excited state responsible for both the mechanolumi-
nescence and the pathway itself [12]. The main research studies have been focused
on lanthanide, manganese, and copper coordination compounds as mentioned
above [9].
One of the proposed mechanisms for ML involves a charge recombination,
which can be described as following processes: (1) The cleavage process induces
an opposite charges separation on newly crystal surfaces; (2) The charge recombi-
nation excites the organic ligand to their singlet (S1); (3) Excited ligand transfer its
energy to the metal ion; and finally (4) radiatively decays to its ground state,
emitting the excess of energy as photons. Another mechanism involves gas dis-
charge by the following pathways: (1) The electrical discharge on newly opposing
crystal surfaces under a nitrogen atmosphere promotes the excitation of N2 mole-
cules in gas phase. (2) The excited N2 molecules emit photons due to the 3Πu ! 3Πg
transition in the range of 300–420 nm. (3) The electromagnetic radiation from the
gas discharge is absorbed by the ligand, which is excited to its excited singlet state
(S0 ! S1). (4) The ligand undergoes an intersystem crossing into a triplet state
(S1 ! T). (5) Afterwards, the ligand transfers its energy to the metal ion and finally
goes to the ground state emitting photons [40, 41].
Rheingold and King [41] investigated the role of gas in ML intensities and
studied the growing and fracturing of crystals under noble gas or nitrogen atmo-
spheres. According to their result, the trapped nitrogen gas during the crystal lattice
growth plays an important role in the ML phenomenon in the complexes [41]. Addi-
tionally, the authors also observed the unique lanthanide coordination compound
that displays both gas discharge and ML emission, which was tris(2-thenoyltri-
fluoroacetonate)phenantrolineeuropium(III) complex [12, 42].
ML properties have been largely reported in the literature [9, 12, 43]. It was
believed for a long period that the necessary condition for a transition metal
coordination compound, presenting ML phenomenon is its crystal structure that
belongs to a non-centrosymmetric space group. The main reason was that the
generation of piezoelectric charge could not take place in crystals of symmetric
space groups. However, some centrosymmetric trivalent lanthanide coordination
compound crystals display ML spectra that are characteristic of the metallic ion.
This fact is commonly attributed to the presence of impurities or, more frequently,
intrinsic structural disorders. Nevertheless, the perfect relationship between struc-
ture and crystal symmetry is not completely understood yet [44, 45].
The assignment of space group to coordination compounds may be
compromised due to the quality of X-ray data, which may lead to inaccurate
conclusions. For instance, Europium tetrakis(dibenzoylmethide)triethylammonium
C22 C21 C66 C62 C58

C37 C36 C59
C23
C65 C57
C35 N1 C60
C19 C20 C56
C38 C61
C24 C55
C39 08
C16 03 C63
C34
C64 C15 C14
C48
C17
C31 05 02 C47 C13
C32 C18 Eu C46
C30 04 C10
C25 C12
C29 C33 07 C3 C54

06
C40 01 C2 C53
C1 C52
C28
C41 C51
C9 C4
C44 C42
C5
C43
C8 C6
C7
Fig. 3.5 Displacement ellipsoid plot of HNEt3[Eu(dbm)4], less hydrogen atoms, showing the
disorder and the strong hydrogen bond between the cation and the complex. Ellipsoids are drawn at
the 30 % probability level. Disordered atoms are shown as arbitrarily sized spheres. Reprinted with
permission from Ref. [44]. Copyright 2001 Elsevier
is highly mechanoluminescence and has been reported as crystallizing in the

centrosymmetric I2/a space group [46], which prevents the possibility of an oper-
ating ML mechanism underlined by piezoelectric charges. For more than a decade,
the literature has reported the EuD4TEA as an exception of a centrosymmetric
system, displaying ML properties [46]. In this case, the structural disorders in the
aromatic moieties of ligand were justified as responsible for the charge separation
process. Nevertheless, important issues have been addressed by Cotton and
co-workers, whom pointed out the EuD4TEA crystal structure indeed belongs to
the non-centric space group Ia instead of the I2/a [44]. In addition, Sage and
co-workers [12] confirmed the asymmetry of the space group of a EuD4TEA ML
crystal by observing strong second harmonic generation when 1064 nm laser was
passed throw the sample (Fig. 3.5).
It is noteworthy to emphasize that the vast majority of ML materials reported in
literature also display PL properties. Although, ML and PL originate from two
distinct stimuli, however, their spectra for coordination compounds show similar
profiles. This behavior is very common for lanthanide compounds. In which, the
intraconfigurational-4f transitions are almost insensitive to the chemical
Fig. 3.6 Schematic of

chelating molecules that
serve a dual function as both
the protective lanthanide
and the sensitizing antenna.
Adapted with permission
from Ref. [48]. Copyright
2014 American Chemical
Society
environment and also the long lived-emission of millisecond order minimize the
effects of pressure-induced Franck-Condon factors [12]. Moreover, in contrast to
the organic and transition metals materials, the lanthanide compounds indicate that
both types of luminescence may come from the same excited state and no distortion
in the punctual group occurs during the breaking [12, 47]. In this context, two
mechanisms have been suggested for ML process involving the antenna effect. This
effect is associated with the intramolecular energy transfer from the excited ligand
to the central ion, which exhibits characteristic luminescence (Fig. 3.6).
3.3 Mechanoluminescence of Lanthanide Compounds
It is well known that lanthanide ions (Ln3þ) display unique spectroscopic proper-
ties, owing to the 4f-shielding effect from the chemical environment by filled 5s and
5p subshells. As a result, Ln3þ exhibit low molar absorption coefficients
(<10 L mol1 cm1) and long-lived narrow emission bands [49–52]. This is also
the reason for all lanthanide ions, undergoing low light absorption, resulting in
weak luminescence from 4f energy levels (Fig. 3.7).
On the other hand, organic molecules containing chromophores generally dis-
play high values of absorption coefficient. Therefore, these molecules are used as
ligands in luminescent coordination compounds in order to overcome the low
absorption coefficient of the metal ions, acting as an “antenna” which absorbs
efficiently radiation and then transfers the excitation energy to the lanthanide ion by
intramolecular energy transfer process [4, 6, 54] (Fig. 3.8). However, high emission
intensities centered on the Ln3þ ions may be only achieved by indirect excitation
[55]. Thus, it should be mentioned that antenna effect also plays an essential role in
order to explain ML emission phenomenon in transition metal coordination com-
pounds, as will be discussed later [56, 57].
Among the ML lanthanide coordination compounds, those based on Eu3þ and
3þ
Tb ions have received special attention. Some of them are known as the brightest
ML materials reported so far, which might be explained by the large energy gap
between the emitting levels (5DJ) and the highest energy ground levels (7FJ) of
Fig. 3.7 Partial energy diagrams for the lanthanide aqueous ions. The main luminescent levels are
drawn in red, while the fundamental level is indicated in blue. Reproduced from Ref. [53] with
permission of The Royal Society of Chemistry
Fig. 3.8 Schematic energy transfer mechanism in trivalent lanthanide ion coordination com-
pounds, where the broad and dashed arrows indicate radiative and non-radiative processes,
respectively
these ions; therefore, reducing the non-radiative decay rates. The ML emission
from these coordination compounds can be easily observed, even in daylight with
the naked eye [58, 59]. The Eu3þ and Tb3þ complexes generally show red and green
characteristic ML emissions arising from their intraconfigurational 5D0 ! 7FJ and 5

D4 ! 7FJ transitions (J ¼ 0, 1, 2, 3, 4, 5, and 6), respectively [57, 60].
One of the most important types of luminescent lanthanide compounds are those
ones, where β-diketonate ligands act as efficient antennae [54, 61]. Based on the
theoretical mechanisms, this behavior is due to the almost resonance condition
between the energy donor low lying excited triplet state (T) of ligand and an
acceptor excited level (2Sþ1LJ) of the lanthanide ion [49, 52, 55]. In many cases,
it has been possible to design highly luminescent compounds by tuning the triplet
state energy via appropriate choice of the substituent groups in the ligands. It is well
known that the β-diketonate ligands, containing aromatic, a combination of ali-
phatic and aromatic groups or even alkyl fluoride moieties, such as dibenzoyl-
methanate (dbm), thenoyltrifluoroacetonate (tta), and hexafluoroacetoacetonate
(hfac) are very efficient sensitizers for the Eu3þ ion [55]. On the other hand, the
strongly luminescent Tb3þ-β-diketonate compounds are those ones with aliphatic
substituent groups, for instance, acetylacetonate (acac), and pivaloylmethanate
(dpm) [62–64].
Hurt and co-workers [56] reported for the first time the mechanoluminescence in
the tetrakis(dibenzoylmethanate)europium(III) complexes with several aliphatic
and aromatic organic amine counter-ions. In particular, the complex containing
triethylammonium cation (EuD4TEA) has been reported as the most brightest ML
coordination compound so far, which has opened great perspective for its applica-
tions in smart damage sensors. This system is also known for its high luminescence
intensity, presenting quantum yield value at around 75 % [47].
Since the last few years, doping has been adopted as a strategy for improving ML
properties. In this regard, the role of different small organic molecule, polymer, and
metal ion dopants on the EuD4TEA ML intensities and prompt decay time were
investigated by Fontenot and co-workers [65–71]. Therefore, in the majority of the
cases, it was noticed a considerable increase in the ML intensities for doped systems
when compared with the undoped ones. For example, an increase of 55 % was
observed by using dimethyl methylphosphanate as a dopant [65]. A qualitative
investigation has also shown ML intensity enhancement for uranyl acetate [67],
dysprosium [68], and poly(methyl methacrylate) [69] dopants. On the other hand, a
deleterious effect in the intensities was observed when Sm3þ ion [70] and dibutyl
phosphate [71] were applied as dopands in EuD4TEA material (Fig. 3.9).
The ML properties of tris- and tetrakis-thenoyltrifluoroacetonate europium
complexes have been also extensively investigated [R] so far, and considerable
attention has been devoted to correlate ML and structural properties of the Eu
(TTA)3L complexes, where L is an ancillary heteroaromatic ligand, such as, 2,2-
0
-bipyridine (bipy) or 1,10-phenantroline (phen) or their derivatives [42, 72–
76]. The high ML intensities have been reported for [Eu(tta)3(phen)] and [Eu
(tta)3(dmdpy)] complexes that crystallize in asymmetric Pca21 space group,
where dmdpy is the 4,40 -dimethyl-2,20 -bipyridyl ligand.
Similar behavior has been observed for highly ML bis-diketonate lanthanide
complexes of formula [Ln(tta)2(NO3)(tppo)2] which crystallize as Pca21 space
group, where Ln: Eu3þ, Tb3þ and Gd3þ, where tppo corresponds to
Fig. 3.9 Triboluminescence of the EuD4TEA complex
triphenylphosphine oxide. In contrast to the ML spectra of Eu3þ and Tb3þ com-

pounds, showing only narrow bands attributed to the intraconfigurational-4fN, 5
D0 ! 7FJ and 5D4 ! 7FJ transitions, respectively, the spectrum for [Gd(tta)2(NO3)
(tppo)2] complex exhibits a broad band due to the intraligand T ! S0 transition or
radiative decay [57]. The latter result indicates that antenna effect is also an
operative mechanism in the ML phenomenon for lanthanide coordination
compounds [57].
Li et al. [77] synthesized the europium complexes represented by the formulas
[Eu(dbm)LSS] (1) and [Eu(dbm)3LRR] (2), containing the chiral organic ligands
(þ)-4,5-pinene bipyridine and ()-4,5-pinene bipyridine, respectively, and
dbm ¼ dibenzoylmethanate as photosensitizer. Both complexes were crystallized
in monoclinic chiral space group P21 and displayed an intense red light emission
under mechanical perturbation. This behavior was attributed to the material ability
of generating opposite charges on newly surfaces, which excite the β-diketonate
and pyridine derivative ligands that transfer energy to the europium ion, resulting in
its luminescence.
Although the vast majority of the ML materials reported in the literature reflect
the piezoelectric properties of the crystals, which might be presented only in the case
of asymmetric space groups, but ML phenomenon is indeed not only restricted to
this kind of compounds. A number of lanthanide complexes with β-diketonate
ligands crystallize in centrosymmetric space groups and exhibit high ML intensities.
Chen and co-workers [72] investigated the ML properties of tris-tta europium
complexes of formulas [Eu(tta)3(bipy)], [Eu(tta)3(dmbp)]H2O and [Eu(tta)3(dia)]
where dia: 4,5-diazafluoren-9-one, dmbp: 4,40 -dimethyl-2,20 -bipyridinate. How-
ever, the first two complexes crystallized in centrosymmetric P21/n space group.
According to this study, the observed ML activities might be ascribed to disorders of
the thienyl and CF3 substituents for [Eu(tta)3(bipy)] and the disordered lattice water
molecule in [Eu(tta)3(dmbp)]H2O. This argument have been used to explain the ML
properties presented by large number of centrosymmetric tetrakis-β-diketonate
lanthanide compounds such as DMPY[Eu(TTA)4] (P21/n) [78], MLQ[Eu(TTA)4]
Fig. 3.10 Triboluminescence (TL) spectra (uncorrected) of crystals of [Dy(PPI)3(DPEPO)] and

[Tb(PPI)3(DPEPO)] recorded at 298 K. Adepted with permission from Ref. [79]. Copyright 2013
American Chemical Society
(P21/n), and APY[Eu(TTA)4] ( P 1 ), where DMPY: 1,4-dimethylpyridinium,

MLQ: N-methylisoquinolium, and APY: 4-aminopyridinium [74]. The
homodinuclear bridging μ2-(ONC5H5), [Eu2(TTA)6(PyO)2] ( P 1 ), where
PyO ¼ pyridine N-oxide [38] and [Eu(NO3)3(hmpa)3](P21/n) complexes, where
hmpa ¼ hexamethylphosphoramide are centrosymmetric mechanoluminescent
systems.
Bulgakov and co-workers [40] studied the ML phenomenon of trivalent lantha-
nide acetylacetonates complexes, represented as [Ln(acac)3H2O], where Ln ¼ Tb
and Eu, which exhibited brightness centered on the metal ion when they were
fractured under the nitrogen atmosphere. Nevertheless, the most intense ML was
emitted by [Tb(acac)3H2O] which could be seen in a slightly shaded room as green
sparkles. Further evidence for the ML centered on the Ln3þ ions were given by Biju
and co-workers [79], whom synthesized heterocyclic β-diketonate coordination
compound: [Ln(PPI)3(DPEPO)], where Ln ¼ Tb3þ and Dy3þ, HPPI ¼ 3-phenyl-4-
propanoyl-5-isoxazone and DPEPO ¼ bis(2-(diphenylphosphino)phenyl)ether
oxide (Fig. 3.10).
Recently, the ML properties of bis(dipivaloylmethanate)nitrato-bis(triphenyl-
phosphineoxide)lanthanide(III) complexes: [Eu(dpm)2(NO3)(tppo)2] and [Tb
(dpm)2(NO3)(tppo)2] were investigated by fracturing their crystals at room and
liquid nitrogen temperatures. Whereas, the Tb-complex exhibited intense green
emission, that can be observed even in daylight, at both temperatures. On the other
hand, the ML phenomenon for Eu-complex was detected only at low temperature.
This result was ascribed to the luminescence quenching owing to the low lying
Fig. 3.11 Qualitative energy level diagram of the intramolecular energy transfer mechanism for
the [Eu2(dpm)6] and [Eu(dpm)2(tppo)2(NO3)] complexes. Reprinted with permission from Ref.
[58]. Copyright 2015 Elsevier
ligand-to-metal charge transfer states (LMCT), which is minimized at liquid nitro-

gen temperature [58] (Figs. 3.11 and 3.12).
Despite the high photoluminescence intensity presented by lanthanide coordi-
nation polymers and metal-organic frameworks (MOFs) only an optimum number
of works have been reported about their ML properties. It is important to mention
that the photoluminescence quantum yields of this type of materials are usually
high due to their structural arrays, which contribute to low luminescence quenching
via non-radiative decays. In particular, the europium materials based on haxafluor-
oacetonate and 3,30 -bis(diphenylphosphoryl)-2,20 -bipyridine (BIPYPO) are among
the most brilliant ML coordination polymer reported so far [59]. The high ML
intensity observed in this compound has been explained as a consequence of its
non-centrosymmetric structure (space group Cc) and low-vibrational frequencies
assigned to the β-diketonate and phosphane oxide ligands. A series of
one-dimensional coordination polymers of formula [Ln(hfa)3(Q)], where Ln ¼ Eu,
Gd, Tb, and Lu as well as Q ¼ 1,4-diacetylbenzene (acbz), 1,4-diacetoxybenzene
(acetbz), or 1,4-dimethyltherephtalate (dmtph) were investigated. The complexes
with acbz and acetbz ligands crystallize in the centrosymmetric triclinic space
group P-1, while those ones with dmtph in the non-centrosymmetric monoclinic
space group P21, which explain the fact that the latter compounds exhibit the
strongest mechanoluminescence when their crystals are fractured [80] (Table 3.1).
Fig. 3.12 Tribolumine-

scent spectra of the [Eu
(dpm)2(NO3)(tppo)2] and
[Tb(dpm)2(NO3)(tppo)2]
recorded at liquid nitrogen
temperature. Reprinted with
permission from Ref.
[58]. Copyright 2015
Elsevier
3.4 Mechanoluminescence of Transition Metal Compounds
The ML of the transition metal coordination compounds was first reported by

Cotton and Goodgame [37, 38]. Most of the studies reported so far have been
concerned to the manganese(II) complexes. Only very limited works have been
reported about ML properties of Cu(I) complexes with arenethiolate ligands
[83, 84]. This kind of ligands present unoccupied antibonding (π*) orbitals that
may accept electron densities from central metal ions. Considering that Cu(I) shows
filled electronic configuration (3d10), the main electronic transitions in its com-
plexes with arenethiolate ligand are arisen from metal-to-ligand charge transfer
(MLCT) processes, which correspond to broad emission bands in the green or red
visible spectral regions (Fig. 3.13).
The electronic levels structure of the transition metal compounds are
characterized by the 2Sþ1Γ states due to the ligand field perturbation (Δ 6¼ 0) on
54
Table 3.1 Chemical structures of some highly triboluminescent coordination compounds

Space
Coordination compound group Structural formula
Europium tetrakis(dibenzoylmethide)triethylammonium Ia [44]
Europium tris(2-thenoyltrifuoroacetone)phenanthroline P21/n [72]
Hexakis(antipyrine)terbium triiodide R 3 [81]
Tris(1,3-di-tert-butyl-b-propanedione)terbium-p-dimethylaminopyridine P21 [82]

E.E.S. Teotonio et al.
Tris(1,3-di-tert-butyl-b-propanedione)samarium-p-dimethylaminopyridine P21 [82]
Poly[3,30 -bis(diphenylphosphoryl)-2,20 -bipyridine][tris(hexafluoroacetyla- [59]

cetonate)]europium
[Eu(hfa)3(dmtph)]1 P21 [80]
Adapted from Ref. [12] with permission of The Royal Society of Chemistry
3 Mechanoluminescence of Coordination Compounds
55
Fig. 3.13 Qualitative

energy-level diagram for Cuþ
coordination compounds
the free ion 2Sþ1L energy levels, where Γ refers to an irreducible representation of
the point group. In this case, the PL and ML properties of these complexes can be
rationalized by using mainly assigned Tanabe-Sugano diagrams. These diagrams
take into account the mutual interactions between ligand field and d electrons of the
transition metal ions. It is noteworthy to emphasize that the d-intraconfigurational
transitions are forbidden by parity selection rule; however, they may be relaxed by
either the vibronic coupling (for centrosymmetric point groups) or mixing opposite-
parity wave functions (for non-centrosymmetric point groups). In contrast to the
lanthanide ions, transition metal ion compounds generally display ML broad bands
due to the stronger interactions between the d electrons and the chemical
environments [3].
Transition metal complexes may exhibit distinct colors which are dependent on
the symmetry and the ligand field strength. The Mn2þ tetrahedral complexes
present ML green emission, whereas octahedral ones red emission, due the spin
forbidden 4T1 ! 6A1 transition [85–87].
The most well-accepted mechanoluminescence mechanism for transition metal
complexes also involves the intramolecular energy transfer from organic molecules
excited by a mechanical stimulus to the excited 2Sþ1Γ ligand field states of the
central metal ion [85]. Most of the ML Mn2þ complexes reported so far are either
tetrahalide or a mixed complex of two different ligands, such as phosphine,
phosphine oxide or amine together with halide ligands (Table 3.2). These organic
ligands present excited triplet states with energy slightly higher than the ligand field
excited states, 4A1, 4T2 and 4T1, arising from the 4G spectroscopic term of the Mn2þ
(Fig. 3.14). Although this sensitization process plays the major role in ML phe-
nomenon, the energy transfer from the ligand excited singlet state to the other levels
of higher energies cannot be neglected. Despite the considerable number of
Table 3.2 Properties of tetrahedral complexes of manganese(II)

Compound ML Spectrum measured Space group
1. (PyH)2MnCl4 No P 1
2. (Me3PhP)2MnCl4 Yes Yes P213
3. (Me4N)2MnCl4 No P21/n
4. (Et4N)2MnBr4 Yes Yes P 4 21 m
5. (Bu4N)2MnI4 No ?
6. (Ph3PO)2MnCl2 at 80 K Yes Fdd2
7. (Ph3PO)2MnBr2 Yes Yes P1
8. (Ph3PO)2MnI2 Yes No P1
9. (Ph3AsO)2MnCl2 No P21/n
10. (Ph3AsO)2MnBr2 at 80 K No Pca21
11. (Ph3AsO)2MnI2 No P 1
Adapted with permission from Ref. [39]. Copyright 2001 American Chemical Society
Fig. 3.14 Representation of the energy adsorption, migration, and emission process of the Mn
(II) complexes. Adapted from Ref. [85] with permission of The Royal Society of Chemistry
publications, dealing with ML in transition metal coordination compounds, sensi-

tization process has not been theoretically investigated yet. Consequently, the
transfer and back transfer energy efficiencies for these compunds are not exten-
sively well known as they are for lanthanide complexes.
Hardy et al. [88] investigated the ML and the effect of increasing pressure on PL
of tetrahedral manganese (II) complexes, such as, Mn(L)2X2 or Mn(L)2X4 where
X ¼ Cl, Br and I as well as L ¼ Ph3PO, MePh3P, Et4N, Bu4N, or PyH. In this case,
the highest ML intensities were observed for Mn(Ph3PO)2Cl2 and Mn(Ph3PO)2Br2
complexes, while the Mn(MePh3P)2Cl4 displayed a very weak intensity. The PL
intensity for the latter complex under high pressure was also investigated [88, 89],
and no sign of PL was detected. Taking into account the intramolecular excitation
mechanisms proposed for tetrahedral manganese (II) compounds, which involve
pre-excitations of low-lying excited ligand states, there are no clear reasons for this
behavior, still reported [88, 90].
Chandra and co-workers [91, 92] investigated the ML excitation process pro-
moted by a piston impact on crystals of the [(C2H5)4]MnBr4, (Ph3PO)2MnBr2,
(Et4N2)2MnBr4, (MePh3P)MnCl4 and (Ph3PO)2MnCl2 complexes. According to
this study, ML intensity is largely dependent on the impact velocity and tempera-
ture of the crystals.
Cotton and co-workers [39] synthesized and solved the molecular structures of
11 tetrahedral coordination Mn(II) compounds in order to investigate the structural
origin of ML phenomenon (Table 3.2). According to their results, all ML manga-
nese compounds crystallize in non-centrosymmetric space groups, which were
considered a necessary condition for observing ML phenomenon, at that time [39].
Recently, new brilliant green ML tetrahalogenomanganese(II) complexes
containing bis[2-(diphenyl-phosphino)phenyl]ether oxide (DPEPO) [85] that crys-
tallize in centrosymmetric space groups were synthesized. The obtained materials
presented values of photoluminescence quantum yield between 30 and 70 %,
indicating that the chelating bidentate ligand DPEPO acts as a good sensitizer.
Based on the non-ionic structures and lack of disordered groups for these manga-
nese complexes, the ML properties cannot be supported by the previously proposed
mechanism and a coherent explanation is still required.
3.5 Final Remarks
In spite of the extensive number of organic and inorganic compounds that exhibit
mechanoluminescent phenomenon, the main studies have been focused on the
academic level. As compared with devices that work on the basis of photolumi-
nescence, electroluminescence, and chemiluminescence phenomena, the develop-
ment of devices for practical applications using ML coordination compounds have
found additional difficulties mainly owing to the lack of ML properties when these
compounds are incorporated in polymer hosts and the unrepeatable ML intensities
measurements. However, recently there are great efforts to apply ML coordination
compounds as smart optical sensors for real-time damage detections and monitor-
ing civil and military structures, luminescent probes for biological applications,
displays and lightning systems, and so on [93, 94]. The most important advantage of
ML devices is that there is no need of an external electric source. Besides, the
straightforward synthesis procedure, the high ML brightness together with narrow
emission bands found for some lanthanide coordination compounds with
β-diketonate ligands fulfill the most important requirements for damage sensor
applications [12].
However, additional studies still deserve to be performed in order to provide a

clear understandable and underlying ML phenomenon in coordination compounds,
considering that many of these materials exhibit unexpected properties based on the
current proposed mechanism. Furthermore, experimental data about ML intensities
as function of synthetic route, crystallization and fracture atmospheres as well as
crystal size for those complexes that exhibit unusual ML properties need to be
reported in order to give information of each parameter separately, which affects
the luminescence intensity. These results may produce a more general ML mech-
anism for coordination compounds, which can contribute to design more efficient
ML materials for practical applications.
References
1. Harvey, E. N. (1957). A history of luminescence: From the earliest times until 1900. Phila-
delphia: American Philosophical Society.
2. Atkins, P., & Friedman, R. (2005). Molecular quantum mechanics (4th ed.). New York:
Oxford University Press.
3. Blasse, G., & Grabmaier, B. C. (1994). Luminescent materials. Heidelberg: Spriger-Verlag.
4. Bünzli, J.-C. G., & Choppin, G. R. (Eds.). (1989). Lanthanide probes in life, chemical and
earth sciences: Theory and practice. Amsterdam: Elsevier.
5. Hemmilä, I., Dakubu, S., Mukkala, V. M., Siitari, H., & L€ ovgren, T. (1984). Europium as a
label in time-resolved immunofluorometric assays. Analytical Biochemistry, 137(2), 335–343.
6. Binnemans, K. (2009). Lanthanide-based luminescent hybrid materials. Chemical Reviews,
109(9), 4283–4374.
7. Bünzli, J.-C. G., & Eliseeva, S. V. (2011). Lanthanide luminescence: Photophysical, analyt-
ical and biological aspects (Springer series on fluorescence). Heidelberg: Spriger-Verlag.
8. Bechara, E. J. H., & Viviani, V. R. (2015). Living light in the darkness: Facts and stories.
Revista Virtual de Quimica, 7(1), 3–40.
9. Jha, P., & Chandra, B. P. (2014). Survey of the literature on mechanoluminescence from 1605
to 2013. Luminescence: The Journal of Biological and Chemical Luminescence, 29(8),
977–993.
10. Lima, P. P., Paz, F. A. A., Brites, C. D. S., Quirino, W. G., Legnani, C., Costa e Silva, M.,
et al. (2014). White OLED based on a temperature sensitive Eu3þ/Tb3þ β-diketonate
complex. Organic Electronics, 15(3), 798–808.
11. Oliveira, A. G., Stevani, C. V., Waldenmaier, H. E., Viviani, V., Emerson, J. M., Loros, J. J.,
et al. (2015). Circadian control sheds light on fungal bioluminescence. Current Biology, 25(7),
964–968.
Journal of Materials Chemistry, 11(2), 231–245.
13. Xu, C. N., Li, C., Imai, Y., Yamada, H., Adachi, Y., & Nishikubo, K. (2006). Development of
elastico-luminescent nanoparticles and their applications. Advances in Science and Technol-
ogy, 45, 939–944.
14. Wang, W. X., Matsubara, T., Takao, Y., Imai, Y., & Xu, C. N. (2011). A new smart damage
sensor using mechanoluminescence material. Materials Science Forum, 675–677, 1081–1084.
15. Zhenyi, M., Jiawen, F., & Dickinson, J. T. (1988). Properties of the photon emission accom-
panying the peeling of a pressure-sensitive adhesive. Journal of Adhesion, 25(1), 63–77.
16. Camara, C. G., Escobar, J. V., Hird, J. R., & Putterman, S. J. (2008). Correlation between
nanosecond X-ray flashes and stick–slip friction in peeling tape. Nature, 455(7216),
1089–1092.
17. Lee, S., Jensen, L. C., Langford, S. C., & Dickinson, J. T. (1995). Electrical transients
generated by the peel of a pressure sensitive adhesive tape from a copper substrate. Journal
of Adhesion Science and Technology, 9(1), 1–26.
18. Zhang, H., Terasaki, N., Yamada, H., & Xu, C.-N. (2009). Development of mechanolu-
minescent micro-particles Ca2MgSi2O7:Eu, Dy and their application in sensors. Thin Solid
Films, 518(2), 610–613.
19. Bacon, F. (1605). The advancement of learning. Oxford: CreateSpace Independent Publishing.
20. Longchambon, H. (1922). Study of the spectrum of the light emitted in the triboluminescence
of sugar. Journal of the Franklin Institute, 195, 269–270.
of Luminescence, 75(4), 295–299.
22. Hagan, A. K., & Zuchner, T. (1982). Lanthanide-based time-resolved luminescence immuno-
assays. Analytical and Bioanalytical Chemistry, 400(9), 2847–2864.
23. Sweeting, L. M., & Guido, J. L. (1985). An improved method for determining tribolumines-
cence spectra. Journal of Luminescence, 33, 167–173.
24. Longchambon, H. (1925). Research experiments on the phenomenons of triboluminescence
and of crystalloluminescence. Bulletin de la Société française de minéralogie, 48, 130–214.
25. Sodomka, L. (1971). Triboluminescence of silicon carbide and other uncommon materials.
Physica Status Solidi A, 7, K65–K66.
26. Meyer, K., & Polly, F. (1965). Zum Mechanismus der tribolumineszenz in alkalihalogeniden.
Berichte der Bunsengesellschaft f€ur Physikalische Chemie, 69, 244.
27. Meyer, K., Obrikat, D., & Rossberg, M. (1970). Progress in triboluminescence of alkaline
halides and doped zinc sulphides (II). Kristall und Technik, 5, 181.
28. Nowak, R., Krajewska, A., & Samoc, M. (1983). Efficient triboluminescence in
N-isopropylcarbazole. Chemical Physics Letters, 94(3), 270–271.
Physical Review B, 21(2), 816–826.
30. Chandra, B. P. (1995). Preparation and characterization of gold doped (Zn, Cd)S mixed
phosphors for mechano-optical transducers. Bulletin of Materials Science, 18(5), 503–515.
31. Wick, F. G. (1940). Triboluminescence of sugar. Journal of the Optical Society of America A,
30, 302.
32. Tabushi, I., Yamamura, K., & Nishiya, T. (1979). Pulsed infrared laser induced visible
luminescence. Journal of the American Chemical Society, 101(10), 2787–2788.
33. Angelos, R. A., Zink, J. I., & Hardy, G. E. (1979). Triboluminescence spectroscopy of
common candies. Journal of Chemical Education, 56, 413.
34. Zink, J. I. (1975). Triboluminescence of uranyl nitrate. Inorganic Chemistry, 14(3), 555–558.
35. Fontenot, R. S., Hollerman, W. A., Aggarwal, M. D., Bhat, K. N., & Goedeke, S. M. (2012). A
versatile low-cost laboratory apparatus for testing triboluminescent materials. Measurement,
45(3), 431–436.
36. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2012). Comparison of the
triboluminescent properties for europium tetrakis and ZnS:Mn powders. Journal of Theoret-
ical and Applied Physics, 6(1), 1–9.
37. Cotton, F. A., Goodgame, D. M. L., & Goodgame, M. (1962). Absorption spectra and
electronic structures of some tetrahedral manganese(II) complexes. Journal of the American
Chemical Society, 84, 167–172.
38. Goodgame, D. M. L., & Cotton, F. A. (1961) Phosphine oxide complexes. Part IV. Tetrahedral,
planar, and binuclearcomplexes of copper(II) with phosphine oxide, and some arsine oxide
analogues. Journal of the Chemical Society 2298–2305.
39. Cotton, F. A., Daniels, L. M., & Huang, P. (2001). Correlation of structure and tribolumines-
cence for tetrahedral manganese(II) compounds. Inorganic Chemistry, 40(14), 3576–3578.
40. Bulgakov, R. G., Kuleshov, S. P., Zuzlov, A. N., & Vafin, R. R. (2004). Triboluminescence of
lanthanide acetylacetonates. Russian Chemical Bulletin, 53(12), 2712–2714.
41. Rheingold, A. L., & King, W. (1989). Crysta structures of three brilliantly triboluminescent
centrosymmetric lanthanide complexes: Piperidinium Tetrakis(benzoylacetonato)europate,
Hexakis(antipyrine)terbium triiodide, and Hexaaquadichloroterbium chloride. Inorganic
Chemistry, 28(9), 1715–1719.
42. Takada, N., Sugiyama, J., Katoh, R., Minami, N., & Hieda, S. (1997). Mechanoluminescent
properties of europium complexes. Synthetic Metals, 91(1–3), 351–354.
131(7), 1407–1418.
44. Cotton, A. F., Daniels, L. M., & Huang, P. (2001). Refutation of an alleged example of a
disordered but centrosymmetric triboluminescent crystal. Inorganic Chemistry Communica-
tions, 4(6), 319–321.
45. Cotton, F. A., & Huang, P. (2003). Further observations on the non-rigorous relationship
between triboluminescence and crystal centricity. Inorganica Chimica Acta, 346, 223–226.
46. Sweeting, L. M., & Rheingoldf, A. L. (1987). Crystal disorder and triboluminescence:
Triethylammonium Tetrakis(dibenzoy1methanato)europate. Journal of the American Chemi-
cal Society, 109(9), 2652–2658.
47. Bourhill, G., Palsson, L. O., Samuel, I. D. W., Sage, I. C., Oswald, I. D. H., & Duignan, J. P.
(2001). The solid-state photoluminescent quantum yield of triboluminescent materials. Chem-
ical Physics Letters, 336, 234–241.
48. Heffern, M. C., Matosziuk, L. M., & Meade, T. J. (2014). Lanthanide probes for bioresponsive
imaging. Chemical Reviews, 114(8), 4496–4539.
49. Brito, H. F., Malta, O. L., Felinto, M. C. F. C., & Teotonio, E. E. S. (2009). Luminescence
phenomena involving metal enolates. Chichester, UK: John Wiley & Sons.
50. Bünzli, J.-C. G. (2015). On the design of highly luminescent lanthanide complexes. Coordi-
nation Chemistry Reviews, 293–294, 19–47.
51. Binnemans, K. (2015). Interpretation of europium(III) spectra. Coordination Chemistry
Reviews, 295, 1–45.
52. Wang, Z., Ananias, D., Carné-Sánchez, A., Brites, C. D. S., Imaz, I., Maspoch, D.,
et al. (2015). Lanthanide-organic framework nanothermometers prepared by spray-drying.
Advanced Functional Materials, 25(19), 2824–2830.
53. Bünzli, J.-C. G., & Piguet, C. (2005). Taking advantage of luminescent lanthanide ions.
Chemical Society Reviews, 34(12), 1048.
54. Sabbatini, N., Guardigli, M., & Lehn, J.-M. (1993). Luminescent lanthanide complexes as
photochemical supramolecular devices. Coordination Chemistry Reviews, 123(1–2), 201–228.
55. de Sá, G. F., Malta, O. L., de Mello Donegá, C., Simas, A. M., Longo, R. L., & da Silva Jr, E. F.
(2000). Spectroscopic properties and design of highly luminescent lanthanide coordination
complexes. Coordination Chemistry Reviews, 196, 165–195.
56. Hurt, C., McAvoy, N., Bjorklund, S., & Filipescu, N. (1966). High intensity triboluminescence
in europium tetrakis (Dibenzoylmethide)–triethylammonium. Nature, 212, 179–180.
57. Teotonio, E. E. S., Fett, G. M., Brito, H. F., Faustino, W. M., de Sá, G. F., Felinto, M. C. F. C.,
et al. (2008). Evaluation of intramolecular energy transfer process in the lanthanide(III) bis-
and tris-(TTA) complexes: Photoluminescent and triboluminescent behavior. Journal of Lumi-
nescence, 128(2), 190–198.
58. Miranda, Y. C., Pereira, L. L. A. L., Barbosa, J. H. P., Brito, H. F., Felinto, M. C. F. C., Malta,
O. L., et al. (2015). The role of the ligand-to-metal charge-transfer state in the
dipivaloylmethanate-lanthanide intramolecular energy transfer process. European Journal of
Inorganic Chemistry, 18, 3019–3027.
59. Hasegawa, Y., Hieda, R., Miyata, K., Nakagawa, T., & Kawai, T. (2011). Brilliant tribolumi-
nescence of a lanthanide coordination polymer with low-vibrational-frequency and
non-centrosymmetric structural networks. European Journal of Inorganic Chemistry, 32,
4978–4984.
60. Hüfner, S. (1978). Optical spectra of transparent rare earth compounds. London: Academic
Press.
61. Crosby, G. A., Whan, R. E., & Alire, R. M. (1961). Intramolecular energy transfer in rare earth
chelates. Role of the triplet state. Journal of Chemical Physics, 34(3), 743.
62. Sato, S., & Wada, M. (1970). Relations between intramolecular enegy transfer efficiencies and
triplet state enegies in rare earth diketonate chelates. Bulletin of the Chemical Society of Japan,
43(7), 1955–1962.
63. Dawson, W. R. (1966). Internal-energy-transfer efficiencies in Eu3þ and Tb3þ chelates using
excitation to selected ion levels. Journal of Chemical Physics, 45(7), 2410.
64. Tsaryuk, V., Zolin, V., & Legendziewicz, J. (2003). The structure of ligands and effects of the
europium luminescence excitation. Journal of Luminescence, 102–103, 744–750.
65. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2012). Synthesis and
characterization of highly triboluminescent doped europium tetrakis compounds. Journal of
Luminescence, 132(7), 1812–1818.
66. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2012). Innovative
triboluminescence study of multivitamin doped europium tetrakis. Crystal Research and
Technology, 578(5), 573–578.
67. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2013). Effects of added
uranium on the triboluminescent properties of europium dibenzoylmethide triethylammonium.
Journal of Luminescence, 134, 477–482.
68. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., Alapati, T. R., & Aggarwal, M. D. (2013).
Triboluminescent properties of dysprosium doped europium dibenzoylmethide
triethylammonium. ECS Journal of Solid State Science and Technology, 2(9), P384–P388.
69. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., Aggarwal, M. D., & Penn, B. G. (2013).
Incorporating strongly triboluminescent europium dibenzoylmethide triethylammonium into
simple polymers. Polymer Journal, 46(2), 111–116.
70. Bhat, K. N., Fontenot, R. S., Surabhi, R., Hollerman, W. A., Aggarwal, M. D., & Alapati, T. R.
(2014). Measurement of the triboluminescent properties for europium and samarium tetrakis
dibenzoylmethide triethylammonium. Electronic Materials Letters, 10(6), 1149–1153.
71. Fontenot, R. S., Bhat, K. N., Owens, C. A., Hollerman, W. A., & Aggarwal, M. D. (2015).
Effects of added dibutyl phosphate on the luminescent properties of europium tetrakis
dibenzoylmethide triethylammonium. Journal of Luminescence, 158, 428–434.
72. Chen, X., Zhu, X., Xu, Y., Raj, S. S. S., Fun, H., & You, X. (1999). Triboluminescence and
crystal structures of non-ionic europium complexes. Journal of Materials Chemistry, 9,
2919–2922.
73. Takada, N., Hieda, S., Sugiyama, J., Katoh, R., & Minami, N. (2000). Mechanoluminescence
from piezoelectric crystals of an europium complex. Synthetic Metals, 111–112, 587–590.
74. Chen, X., Duan, C., Zhu, X., You, X., Sundara, S. S., Fun, H., et al. (2001). Triboluminescence
and crystal structures of europium(III) complexes. Materials Chemistry and Physics, 72,
11–15.
75. Zheng, Z., Wang, J., Liu, H., Carducci, M. D., Peyghambarian, N., & Jabbourb, G. E. (2001). A
triboluminescent europium(III) complex. Acta Crystallographica Section C, 58(1), m50–m52.
76. Chen, X.-F., Zhu, X.-H., Xu, Y.-H., Raj, S. S. S., Fun, H.-K., Wu, J., et al. (2002). Crystal
structure and triboluminescence spectrum of Mu2-(ONC5H5) bridging dinuclear europium
(III) complex. Journal of Coordination Chemistry, 55(4), 421–428.
77. Li, X.-L., Zheng, Y., Zuo, J.-L., Song, Y., & You, X.-Z. (2007). Synthesis, crystal structures
and triboluminescence of a pair of Eu(III)-based enantiomers. Polyhedron, 26(18),
5257–5262.
78. Chen, X.-F., Liu, S., Duan, C., Xu, Y.-H., You, X., Ma, J., et al. (1998). Synthesis, crystal
structure and triboluminescence spectrum of 1,4-dimethylpyridinium tetrakis. Polyhedron, 17
(97), 1883–1889.
79. Biju, S., Gopakumar, N., Bunzli, J.-C. G., Scopelliti, R., Kim, H. K., & Reddy, M. L. P. (2013).
Brilliant photoluminescence and triboluminescence from ternary complexes of DyIII and TbIII
with 3-Phenyl-4-propanoyl-5-isoxazolonate and a bidentate phosphine oxide coligand. Inor-
ganic Chemistry, 52, 8750–8758.
80. Eliseeva, S. V., Pleshkov, D. N., Lyssenko, K. A., Lepnev, L. S., & Bunzli, J.-C. G. (2010).
Highly luminescent and triboluminescent coordination polymers assembled from lanthanide
β-diketonates and aromatic bidentate O-donor ligands. Inorganic Chemistry, 3, 9300–9311.
81. Clegg, W., Bourhill, G., & Sage, I. (2002). Hexakis(antipyrine-O)terbium(III) triiodide at
160 K: Confirmation of a centrosymmetric structure for a brilliantly triboluminescent complex.
Acta Crystallographica, 58(4), m159–m161.
82. Clegg, W., Sage, I., Oswald, I., Peter, B., & Bourhill, G. (2000). Two isostructural tribolumi-
nescent lanthanide complexes. Acta Crystallographica, C56, 1323–1325.
83. Knotter, D. M., van Maanen, H. L., Grove, D. M., Spek, A. L., & van Koten, G. (1991).
Synthesis and properties of trimeric ortho-chelated (Arenethiolato)copper(I) complexes. Inor-
ganic Chemistry, 30(17), 3309–3317.
84. Knotter, D. M., Blasse, G., Vliet, J. P. M., & van Koten, G. (1992). Luminescence of copper
(I) arenethiolates and its relation to copper(I) luminescence in other complexes. Inorganic
Chemistry, 31(7), 2196–2201.
85. Chen, J., Zhang, Q., Zheng, F.-K., Liu, Z.-F., Wang, S.-H., Wu, A.-Q., et al. (2015). Intense
photo- and tribo-luminescence of three tetrahedral manganese(II) dihalides with chelating
bidentate phosphine oxide ligand. Dalton Transactions, 44(7), 3289–3294.
86. Orive, J., Mesa, J. L., Balda, R., Fernańdez, J., Fernańdez, J. R., Rojo, T., et al. (2011).
Enhancement of the luminescent properties of a new red-emitting phosphor, Mn2(HPO3)F2,
by Zn substitution. Inorganic Chemistry, 50, 12463–12476.
87. Lin, J., Zhang, Q., Wang, L., Liu, X., Yan, W., Wu, T., et al. (2014). Atomically precise doping
of monomanganese ion into coreless supertetrahedral chalcogenide nanocluster inducing
unusual red shift in Mn2þ emission. Journal of the American Chemical Society, 136(12),
4769–4779.
88. Hardy, G. E., & Zink, J. I. (1976). Triboluminescence and pressure dependence of the
photoluminesnce of tetrahedral manganese(II) complexes. Inorganic Chemistry, 15(21),
3061–3065.
89. Chandra, B. P., & Balakrishna, R. K. (1982). Mechanoluminescence, electroluminescence and
high-pressure photoluminescence of Mn(Ph3 P0)2Br2. Journal of Luminescence, 27, 101–107.
90. Chandra, B. P., Jaiswal, A. K., Chandraker, T. R., & Kaza, B. R. (1981). Mechanolumi-
nescence and electroluminescence of non-photoluminescent Mn(Ph3PO)2Cl2 crystals.
Physica Status Solidi A, 68, K207–K210.
91. Chandra, B. P., Deshmukh, N. G., Sahu, R. B., & Verma, A. K. (1986). Deformation-induced
after-glow in [(C2H5)4N]2MnBr4 crystals. Crystal Research and Technology, 21(12),
1559–1565.
92. Chandra, B. P., Khokhar, M. S. K., Gupta, R. S., & Majumdar, B. (1987). Tetrahedral
manganese(II) complexes—Intense and unique type of mechanoluminophores. Journal of
Physics, 29(4), 399–407.
93. Terasaki, N., Zhang, H., Yamada, H., & Xu, C.-N. (2011). Mechanoluminescent light source
for a fluorescent probe molecule. Chemical Communications, 47(28), 8034–8036.
94. Akiyama, M., Xu, C.-N., Matsui, H., Nonaka, K., & Watanabe, T. (1999). Recovery phenom-
enon of mechanoluminescence from Ca2Al2SiO7:Ce by irradiation with ultraviolet light.
Applied Physics Letters, 75(17), 2548.
Chapter 4
Luminescence of Triboplasma: Origin,
Features, and Behavior
Roman Nevshupa and Ken’ichi Hiratsuka
4.1 Introduction
Nowadays, scientific research into triboluminescence embraces luminous

phenomena of different physical and chemical nature occurring during some kind
of tribological action, e.g., rubbing, scratching, impact, fretting, indentation, or
deformation [1]. Roughly these phenomena can be classified into two main groups
[2]: intrinsic, or solid, and extrinsic triboluminescence. While the former refers to
the luminous phenomena originating from the volume of the material subjected to
mechanical action the latter basically occurs in the surrounding gaseous or liquid
medium. In some cases these processes are not mutually exclusive. The examples of
intrinsic luminescence are the shock wave luminescence in organic crystals and
metals [3–5], luminescence induced by dislocation motion in transition metals
during deformation [6, 7], luminescence under elastic, plastic deformation or
fracture of crystals (2-carboxyphenyl 2-hydroxybenzoate, triethylammonium
tetrakis(dibenzoylmethanato)europate, nicotinium salicylate, N-acetylanthranilic
acid, and many others) [2, 8–10], etc. Extrinsic luminescence is frequently associ-
ated with triboplasma that generally could be of two kinds: Thiessen’s magma-
plasma [11] and tribomicroplasma [12]. In 1967 Thiessen [11] postulated that
magma-plasma is an “energy bubble” in the sub-microscopic deformation zone
that is formed as a result of a quasi-adiabatic energy accumulation at a contact
surface subjected to collision with a particle or a microscopic protuberance from the
rough mating surface. It leads to high excitation states, strong lattice loosening and
R. Nevshupa (*)
Spanish National Research Council, Institute of Construction Sciences
“Eduardo Torroja”, Madrid, Spain
e-mail: r.nevshupa@csic.es
K. Hiratsuka
Faculty of Engineering, Chiba Institute of Technology, Chiba, Japan

66 R. Nevshupa and K. Hiratsuka
structural disruptions for very short periods which are connected with the detach-
ment of lattice components, photons and electrons. A highly excited plasma-like
state around the impact point exists for less than 107 s whereas local heating may
generate an activated states and produce thermal luminescence in the neighborhood
of the impact site for about 104 s [13]. Although an existence of magma-plasma
has been proved in highly loaded contacts and the model magma-plasma was useful
for the explanation of various observations including light and particle emission
during the propagation of a crack [14], it cannot be adopted as an universal
conceptual framework of triboluminescence phenomena.
As early as 1792 Saussure demonstrated that in certain cases light could be
obtained under such conditions that a pure temperature radiation was altogether
unlikely [15]. Further systematic researches gave plenty of evidence that the
average mechanical energy fed in per atom of a solid as it is strained to the
point of fracture is insufficient to excite an atom directly into emitting light [1],
but high energy excited states can be developed during dissipation of mechanical
energy [16]. Various authors put forward the hypothesis that deformation, bond
breaking, and stress-induced chemical transformations resulting from quasi-
adiabatic energy release excite energy-rich states [17–19], whereas subsequent
de-excitation promotes cascade of multiple emission processes. Radiation
de-excitation can lead to luminescence while radiationless one can cause emission
of electrons and ions, chemical transformations, generation of electric field, and so
on. Furthermore, collision of the emitted particles and photons with atoms and
molecules can lead to the secondary excitation and decay processes and so on
[20, 21]. Thus, it is considered that energy dissipation at a tribo-contact is an
extremely complex process involving a large variety of elementary physical
processes and chemical reactions including tribo-induced electroluminescence,
thermoluminescence, chemoluminescence, recombination luminescence, and
ionoluminescence [11].
An example of such mechanical energy dissipation is tribomicroplasma that
occurs during fracture or rubbing of dielectric materials. Tribomicroplasma [22]
as well as fracture-induced luminescence of non-photoluminescent crystals
[23, 24] have been established as electrical phenomena related with charge
generation and electric discharge (lightning) and it manifested in hundreds of
organic and inorganic substances [1, 2, 9, 25–28]. In nature, triboluminescence
related with electric discharge exists in many different forms spreading from tiny
sparks between a comb and dry hair to huge luminous phenomena in
planet atmospheres associated with tectonic slip [29], lightning displays in sand-
storms [30] and pyroclastic clouds of volcanoes. A distinctive feature of
tribomicroplasma is that energy dissipation occurs through a cascade of various
excitation-relaxation processes, both in gas phase and on gas-solid interface
[20, 31, 32]. The present work summarizes recent investigations into characteris-
tics, behavior, and mechanisms of tribo-induced luminescence associated with gas
discharge triboplasma.
4 Luminescence of Triboplasma: Origin, Features, and Behavior 67
4.2 Phenomenology, Spatial Configuration, and Spectral

Characteristics of Luminescence Associated
with Triboplasma
From the end of the nineteenth century the term “triboluminescence” has been used
to describe the ability of some solids to emit light when they are fractured
[24]. Although fracture can occur in a tribo-contact, especially under scratching
or abrasion, the complexity of mechanical phenomena in a tribo-contact cannot be
reduced entirely to whatever single process. Apart from the variety of types of
relative motion in which the mating surfaces can be involved, tribological interac-
tion, in both natural and artificial objects, is usually distinct from fracture in the
following three features: (1) definite geometry of interacting solids on various
dimensional scales, (2) regular time behavior, and (3) contact of (normally) dis-
similar materials. Accordingly, triboluminescence differs from fracture-induced
luminescence in characteristics and behavior, although both phenomena are the
combinations of the same elementary processes.
Fracture-induced luminescence usually occurs as a series of single nonrecurrent
events (lightning). The overall shape and the size of the crystal subjected to
fracturing are irrelevant for fractoluminescence since it localizes at a microscopic
regions of fresh surfaces (Fig. 4.1a) where the charge separation takes place. On the
contrary, tribological interaction is normally a repetitive or cyclic process, thus
allowing accumulation of much higher electric charge. Moreover, regular or quasi-
regular macroscopic geometry of the contacting solids enables development of
quasi-static distribution of electric field at much larger distances from the contact
zone, where the charge is being generated, so it leads to a more complex spatial
configuration of luminescent regions (Fig. 4.1b).
Configuration of luminescent region of triboplasma was studied in several works
using a pin-on-disk setup schematically shown in Fig. 4.2 [12, 33–37]. Tribolumi-
nescence was observed through a transparent disk using a microscope. Variation of
Fig. 4.1 Schematic drawing of fracture-induced luminescence (a) and luminescence of

triboplasma (b)
Fig. 4.2 Schematic

drawing of an experimental
setup for study of
triboluminescence under
dielectric/dielectric sliding.
A pin-on-disk configuration
with a transparent disk
allowed observation 2D
distribution of
luminescence around a pin
the triboluminescent region configuration at various pressures of Ar gas is shown in

Fig. 4.3. The tip and the real contact area are shown in Fig. 4.3a by dashed and
dotted circles, correspondingly. Luminescent zone usually surrounded the tip, but
its center shifted towards the tail, i.e., the region behind the contact with respect to
the relative motion. A bright region (BR) can usually be observed behind the
contact area. This region with a circle shape is slightly brighter than surrounding
regions. By using various optical filters Nakayama and Nevshupa [12, 38] showed
that in N2 and dry air the BR has the maximum luminescence in ultraviolet
(UV) spectral range. Two semicircle lobes connected the bright region with the
frontal part of the contact zone. In addition, some weakly luminescent regions could
be seen behind the contact area and on it. As the gas pressure decreased from the
atmospheric down to 10 Pa the lobes expanded from about 0.2 mm to more than
1 mm. For Ar, He, and other gases that are luminescent in visible range tribolumi-
nescent zones could be easily observed by an eye in a dark room. At gas pressures
below approx. 10 Pa the luminescence steeply decreased in intensity.
Different shape of a triboluminescent zone was reported by Miura and
Nakayama [34] (Fig. 4.4). In addition to a UV bright region β corresponding to a
BR in Fig. 4.3 another bright UV spot γ was found in front of the contact zone. The
lobes were not well defined. Although the reasons for different configurations of
triboluminescent regions are not well-studied, there are some experimental evi-
dences that the spatial distribution of triboluminescence around a tribo-contact
could depend on the overall pin geometry including both the diamond tip and its
holder, especially at lower pressures when triboplasma spreads over large distance
from the contact zone. Simulation [39] has revealed that for pin-on-disk configu-
ration the gas discharge is governed mainly by the electric field of the pin.
Therefore, variations of the shapes of the tip holder that have not usually been
considered and documented by the researchers might affect the triboplasma and
could be a reason for variation of geometrical features of triboluminescence.
With varying gas pressure the total intensity of triboplasma luminescence
followed the behavior shown in Fig. 4.5. This can be interpreted in terms of
Townsend’s model of electrical breakdown. The pressure at which maximum
luminescence occurred corresponds to the optimum electron multiplication
Fig. 4.3 Real-color images of triboluminescence at decreasing Ar gas pressure from 100 kPa (a)
down to 5 Pa (l). The dotted circle in (a) shows a center of the contact zone, whereas the dashed
circle is the projection of the pin
conditions defined by Stoletov’s constant [40]. These conditions depend on both the
type of the gas and the tip geometry [38, 39]. The increase in the normal load
resulted in nearly proportional scaling up of the peak, whereas the general shape of
the graph remained the same. When a sharper pin was used the peak shifted to
Fig. 4.4 (a) Unfiltered pseudo-color image of the TL during sliding in air; (b) band-filtered
pseudo-color image of the TL in the range 610–780 nm; (c) band-filtered pseudo-color image of
the TL in the range 300–420 nm. The size of all images is 250 300 μm2. Red color corresponds to
the maximum intensity indicated in each figure. A circle shows the position and the size of the
contact area (reprinted with permission from [34] © 2001, AIP Publishing LLC)
FN
Luminescence intensity (a.u.)
200
150 0.98 N
100 Diamond tip

0.49 N
0.20 N r =1000 μm
50
102 103 104 105

Ar pressure (Pa)
Fig. 4.5 Triboluminescence intensity at sliding of a diamond pin on a sapphire disk at various
normal loads as function of Ar gas pressure. The radius of curvature of the tip r ¼ 1000 μm
higher pressures and got more intensive [39]. The increase in luminescence inten-
sity with sharper pins and higher loads can be confidently ascribed to higher rate of
tribocharging under higher contact stress, whereas a shift of the peak could be due
to modification of the electric field configuration. For sharper tips the collision
probability of electrons accelerated by an electric field of the pin should be lower
than for blunt tips due to shorter gap between the tip and the disk surface (see
Sect. 4.4). Thus, at lower pressures, when the mean free path of electron gets larger
than the characteristic length of the gap electric discharge is hindered and tribolu-
minescence is dimmed.
Spectral features of triboplasma luminescence are the clear manifestation of its
electric nature. For non-photoluminescent solids the spectra of triboluminescence
measured in the range of gas pressure from 100 kPa to 100 Pa were hardly
distinguishable from the spectra of gas discharge luminescence [22, 24, 41].
ArIII & ArII ArI
Intensity
300 400 500 600 700 800 900

λ (nm)
Fig. 4.6 Spectrum of triboluminescence in Ar gas under 200 Pa pressure
This feature was confirmed for air, N2, O2, H2, He, Ar, and C1 to C3 alkanes
[22]. Although almost all spectral bands of gas discharge luminescence could be
found in triboluminescence spectra, proportions between various bands were
slightly different. For example, in the spectrum of triboluminescence measured
during sliding of a diamond pin on a sapphire (0001) disk in Ar gas at 200 Pa
(Fig. 4.6) the bands of the ionized species (ArII and ArIII) [42] were notably less
intensive than the bands corresponding to neutral Ar, whereas in a glowing dis-
charge spectrum of Ar the bands ArII and ArIII dominated [22]. In the tribolumi-
nescence spectrum the proportion of each component decreased in the following
order: ArI > ArIII > ArII.
In addition to the spectral lines of gas discharge different spectral features were
also reported in literature. Presence of these features manifested a crossover of gas
discharge luminescence with other luminescent processes. These processes can be
independent of the triboplasma or associated with it and may include
fractoluminescence, photoluminescence, etc. [24, 33, 43, 44]. Nakayama [43]
investigated into the effect of the disk material on spectral characteristics of
triboluminescence produced by scratching of single-crystal Al2O3, MgO, and
quartz disks by a diamond pin. For sapphire disk he found a band at approx.
696 nm that could correspond to R-line luminescence of Cr3þ impurities in sapphire
(Fig. 4.7). Such a solid luminescence has been referred to as
tribophotoluminescence [2]. It has been proposed that secondary luminescence
was a photoluminescence excited by a primary UV–vis emission of nitrogen gas
discharge [23, 24]. However, other excitation mechanisms, e.g., ion and electron
bombardment, cannot be excluded.
Evidences for it come from the analysis of the triboluminescence images in a
visible spectral range that revealed presence of irregular strip patterns extending for
several hundreds of micrometers behind the contact zone (Fig. 4.8) [43, 45]. The
luminescence lifetime estimated from the strip length and the sliding velocity was
quite large exceeding several tens of millisecond that is consistent with R emission.
Quite surprising was to find that the strips originated from the spots on the sides just
in front of the contact zone (shown by magenta arrows in Fig. 4.8) and not from the
regions of maximum UV emission situated behind the contact zone. The strips were
Fig. 4.7 Energy spectrum of photons emitted during sliding of a diamond pin on a rotating Al2O3
(a), MgO (b), and SiO2 (c) disks in ambient air with a relative humidity of RH ¼ 40 %, 72 %, and
72 %, respectively, under FN ¼ 1.0 N and V ¼ 32 cm/s and that emitted from air-discharge plasma
between a steel needle and a steel flat disk at an electrode distance of 0.3 mm in ambient air with
RH ¼ 40 % (d). (With kind permission from Springer Science þ Business Media Fig. 13 from [43]
© Springer Science þ Business Media, LLC 2009)
Fig. 4.8 Pseudo-color images (scaled by intensity from blue to red) for UV, visible and IR
emission of triboplasma under scratching a sapphire disk by a diamond pin in Ar gas atmosphere
under various pressures. The points of origin of the strips are shown by magenta arrows (reprinted
with permission from [45] © 2002 JAST)
negligible under atmospheric pressure, but intensified between 10 kPa and 500 Pa.
This behavior can be related with variation of electron mean free path with the gas
pressure. Considering that mean free path of electrons at 100 kPa was only 2.7 μm
that is much smaller than the size of triboplasma [39, 40], only few electrons could
reach the contact zone where the strips originated from. As pressure decreased the
mean free path proportionally increased. These results suggested that secondary
solid luminescence could not be plausibly explained by photoluminescence excited
by gas discharge luminescence as it is generally considered in fractoluminescence
[2, 23]. Other processes such as cathodoluminescence and electron-impact lumi-
nescence could be relevant.
Another manifestation of the importance of electron and ion bombard to which
the surfaces are subjected in triboplasma was the observation of deposits formed on
the sides of the diamond pin when microscopic Al debris were introduced into the
contact zone. Chemical composition of the deposits analyzed by energy dispersive
spectrometry was dominated by aluminum and a small amount of oxygen. The
deposits had a shape of two oblong blots with their sharp ends coinciding with the
positions of the spots where luminescent strips originated from. The interference
pattern observed on the deposit was characteristic for thin films obtained by sputter
deposition and indicated that thickness of the deposit was the highest along the
vertical axis and decreased to its edge. Aluminum deposits were strongly adhered to
the pin surface and could not be removed by scratching. Appearance of these
deposits can be ascribed to ion sputtering of Al debris in triboplasma. Possible
reasons for localization of sputtering at two spots in front of the contact zone will be
discussed in Sect. 4.4. These findings suggest that ion bombardment could also
explain excitation of secondary solid luminescence in sapphire in absence of debris.
Electron bombardment of the disk is unlikely because it obtains negative charge
when is rubbed by diamond [39, 43]. In turn, electron bombardment is inevitable on
a positively charged pin and may cause recombination luminescence on diamond.
For a quartz disk, a broad band in the spectral range from 450 to 700 nm
(Fig. 4.7) was attributed to luminescence of nonbridging oxygen hole centers
[43, 46]. Similar luminescence was reported by Miura et al. [33] under scratching
a quartz disk in vacuum (Fig. 4.9). Taken together these findings suggest that solid
luminescence of quartz should be induced by mechanical deformation and fracture
rather than by UV–vis gas discharge luminescence. Finally, a broad band in the
spectral region from 700 to 1000 nm was related to triboluminescence of diamond
[33]. This band correlated with a cathodoluminescence band centered at 520 nm
and ascribed to recombination of an electron and a hole at a donor–acceptor pair or
H3 defect centers in diamond.
In contrast to fractoluminescence, there is much less information on lumines-
cence of triboplasma that has been studied for only limited number of solids most of
them being ionic inorganic crystals. Considering similarity of fundamental mech-
anisms of both phenomena it is reasonable to expect that triboluminescence of
materials that have not been investigated so far will have the same spectral features
as fractoluminescence. The excitation mechanisms involving a discharge for
fractoluminescence should be also valid for triboluminescence. These mechanisms
were classified by Sweeting [2] into three groups: (a) excitation of the surrounding
gas by electron bombardment; (b) excitation of the photoluminescence of the
material by electron bombardment; and (c) excitation of the photoluminescence
of the material by absorption of the UV emissions of the excited gas. In addition, as
specific characteristics of triboluminescence, one should consider the possibility of
luminescence of both solids and luminescence induced by ion bombardment.
Further studies are still necessary to clarify triboluminescence characteristic of
fractoluminescent crystals.
4.3 Triboluminescence Time Series
Another specific feature of triboluminescence being frequently ignored in studies of

fractoluminescence is a complex behavior. Nevshupa et al. [39, 47] investigated
into behavior patterns of luminescence time series in the range of time scales from
104 to 104 s and under various gas pressures. Four different behavior patterns were
identified (Fig. 4.10). At low gas pressures ( p < 20 Pa) triboluminescence was
3 2.5 2 1.5 [eV]
100
a Fused silica
80
60
40
20
20 b
MgO
Energy density [a. u.]
10
0
c
NaCl
3
0
400 500 600 700 800 900 1000
Wavelength [nm]
Fig. 4.9 Spectra of triboluminescence measured in vacuum (103 Pa) for (a) synthetic fused
silica, (b) MgO, and (c) NaCl. The intensities were normalized by the contact area, the sliding
velocity, and the normal force (reprinted from [33] © 2009, with permission from Elsevier)
weak, but with some irregular intensive bursts. In the pressure range from 20 to
300 Pa a regular pattern appeared. It was characterized by a series of groups of
almost equally spaced bursts with a period between the bursts from 20 to 40 ms.
This regular behavior maintained for thousands of disk revolutions [47]. Some of
these bursts appeared as doublets with submillisecond interval between two spikes.
The group frequency corresponded to the disk rotation. In the pressure range from
300 Pa to 10 kPa the bursts still occurred but the periodic behavior vanished. In
turn, a weak uniform triboluminescence was found at the bottom of the bursts.
×10 ×1 ×2 ×10
A B C D
Imean
10-2 10-1 100 101 102 103 104 105

p (Pa)
Fig. 4.10 Schematic drawing for illustration of the triboluminescence behavior patterns in four
pressure regions. The main plot shows the mean triboluminescence intensity as a function of the
gas pressure. The insets show typical behavior patterns on different vertical scales indicated in
each graph. Region A—very weak triboluminescence with chaotic and sparse bursts; region B—
periodic intensive bursts; region C—mixture of the uniform and burst components of tribolumi-
nescence with chaotic bursts; region D—the uniform triboluminescence component (© IOP
Publishing. Reproduced from [39] with permission. All rights reserved)
Finally at pressures higher, than 10 kPa, triboluminescence was completely

uniform.
Figure 4.11 presents results of statistical analysis of luminescence time series.
Axis of ordinates represents the number of photon counts acquired by a
photomultiplier tube in a time bin of 1.3107 ms, whereas the abscissa shows the
percentage of the bins with the certain number of counts in the whole record
containing 214 bins. As can be seen in the inset (b) a histogram of burst lumines-
cence had an exponential-like distribution. In contrast, the uniform luminescence at
pressures above 1 kPa had a bell-shape distribution. Variation of the mean values of
the burst and uniform components of luminescence are shown in Fig. 4.12.
The burst component has been represented by the graph of the maximum number
of counts per bin in a record, Imax. It reached the maximum at the center of region B
that did not coincide with the maximum of the mean luminescence intensity shown
by the dash line. The bursts were intensive but sparse. Therefore their contribution
to the mean emission intensity was not as significant as that of the uniform
component. In region C, in which burst component coexisted with the uniform
one (half-solid dots), the mean intensity was the highest, although the bursts were
less intensive than in region B.
This apparently complex behavior has certain correlation with variation in
electron mean free path with gas pressure. Mean free path (MFP) of electrons in
a b c
300 3k 400
2k
Number of counts per bin
250 170 Pa
300
170 Pa
770 Pa
770 Pa
1k
200
-3
0 x10
200
0.0 1.0
150
4.5 kP a
17 kPa
100 100
93 kP a
10 Pa
-2
I mu
50
0 x10-3
0 5 10 15
0 Percentage
0.0 0.3
Percentage
Fig. 4.11 (a) Density distribution functions of the time bins, Ni/N, on the number of counts per
bin, Ib, at various gas pressures. The histograms are normalized by the total number of time bins in
each record. (b) Enlarged view of the graph at p ¼ 170 Pa. (c) Enlarged view of the graph
component at p ¼ 770 Pa
Fig. 4.12 Dependence of the parameters and components of the triboluminescence on the gas
pressure: Imax—the maximal triboluminescence intensity, Imean—the mean intensity, Imu—the
centroid of the bell-shaped component of the uniform component, Imax,exp—the expected maximal
intensity of the burst component. The shadowed regions α and β denote the gas pressure intervals
where the mean free path for electrons is approximately equal to the size of the zone of triggering
electron emission and the total luminescent zone, respectively (© IOP Publishing. Reproduced
from [39] with permission. All rights reserved)
gas is inversely proportional to the gas pressure, p, and the momentum transfer
collision cross section, σ:
λ ¼ kT=pσ; ð4:1Þ
where k is Boltzmann’s constant, and T is the gas temperature. The values of σ for
elastic collision and ionization of Ar were 8.0 1020 m2 and 2.5 1020 m2,
correspondingly [48]. At p ¼ 1.6 kPa and slightly below (shadowed region α in
Fig. 4.12) MFP approximately equals the distance from the zone where the trig-
gering electrons are emitted to the contact zone (for details see [39]). These
conditions are optimal for electron multiplication in self-sustained gas discharge
because electrons expend minimum energy for gas ionization. This pressure range
correlated with the uniform component of luminescence that reached it maximum at
p ¼ 1.6 kPa and vanished at lower pressures. In the pressure range from 0.35 to
1.6 kPa MFP was larger, than the optimal, but yet smaller, than the overall
dimension of the zone nearby the contact where the gas discharge can occur. At
p ¼ 350 Pa MPF was about 500 μm (shadowed region β) that was approximately the
maximum possible spread of the luminescent zone for given geometry and condi-
tions [38]. Gas discharge still could occur, but discontinuously. These conditions
corresponded to burst luminescence. Finally, at even lower gas pressures the
electrons could cross large distances with almost no collisions with gas molecules,
thus luminescence was concentrated on the contact zone and associated with
different mechanisms (thermal radiation, etc.).
4.4 Fundamental Processes of Triboplasma
When embarking on a study of luminescence related with triboplasma one should

address the questions of charge generation at the contact zone, transport/diffusion
of the charge in the bulk and/or on the surface of dielectric solids subjected to
friction and wear, configuration of electric field in the space surrounding a contact
zone, emission of triggering electrons, gas discharge, charge recombination on
dielectric surfaces facing triboplasma as well as secondary luminescence excitation
mechanisms including photoluminescence of solids induced by UV–vis emission of
triboplasma, ionoluminescence, and electron bombardment-induced luminescence.
General scheme of various relevant processes and luminescence phenomena related
with triboplasma is shown in Fig. 4.13.
Triboelectrification is the cornerstone of triboplasma. Despite being studied
from the ancient times triboelectrification remains quite a puzzling phenomenon
that depends on material properties, environmental and frictional conditions
[49–52]. Although the rate of electric charge generation is very difficult to measure
in triboplasma, it can be relatively easy assessed from surface electrical potential
under high vacuum. At gas pressures below 103 Pa gas discharge is inhibited, thus
charge recombination is blocked. On the other hand, charge leakage through
conductivity can be neglected for solids with very high resistivity, e.g., 1011 Ωm
for soda lime glass, and 1016 Ωm for sapphire [53]. Therefore, both the charge and
the surface electric potential gradually increase with sliding duration. It was
reported that for a sapphire disk rubbed by a diamond pin under 0.96 N normal
load and 4.4 cm/s sliding velocity the surface potential at the wear track increased
Emission of
Electric field
Solids triggering
build up
electrons
Friction
Triboelectrification Gas discharge
and wear
Frictional Charge
heating recombination
Iono- Luminescence
Incadescence
luminescence of gas discharge
Electron Photo-
-induced luminescence
luminescence
Fig. 4.13 General scheme of some relevant processes and luminescence phenomena related with
triboplasma
from 0 to 300 V in 60 s [39]. The rate of charge generation estimated from these
data was Q1 ¼ 3.39 1011 C/s that is close to the values deduced from the
theoretical considerations [36]. Considering highly asymmetric pin-on-disk config-
uration, it should be noted that while negative charge spreads over the large surface
area on the disk, the positive charge is concentrated on a small area on the tip. So,
for a typical experimental set-up the charge surface density on the disk can be two
to four orders of magnitude smaller, than on the pin. This finding is crucial for
understanding the configuration of electric filed around a contact zone that, in turn,
is determinative for gas discharge characteristics.
Electric field was simulated using a finite element method for above loading and
sliding conditions [39]. The tribological contact was modelled by a diamond
hemisphere of 1 mm radius being in contact with a glass block with dimensions
1 mm 1 mm in plane and the height equal to the disk thickness. Contact area was
determined from Hertz model. The density of positive charges on the pin was
assumed σ p ¼ 3.33 102 C/m2 that is consistent with the limiting charge density
for metal–dielectric contact [54]. In this rough model σ p is likely to be
overestimated since no charge leakage from the contact area has been taken into
account. Linear charge density along the wear track was λd1 ¼ 7.71 1010 C/m
and the surface charge density σ d ¼ 2.14 105 C/m2. Figure 4.14a shows dis-
tribution of surface electric potential on the pin and on the disk. The maximum
potential is found at the contact zone due to concentrated positive charge. Potential
distribution has radial symmetry with respect to the vertical axis z. Distribution of
Fig. 4.14 (a) Model of surface potential on the disk and the pin; (b) profile of electric potential
across the wear track 2 mm above the disk surface
a b
3 1 Ezypin
1,2 Exxdisk
Exxpin
0
1,0 Eyydisk
Ex, Ey (MV/m)
2 Ezxpin
Ez(MV/m)
Eyypin Ezydisk
0,8 -1
0,20 0,22 0,24
1 Ezxdisk
-2
0 -3
0,1 0,2 0,3 0,4 0,0 0,1 0,2 0,3 0,4
x, y (mm)
Fig. 4.15 (a) Radial components of electric field along x and y axes; (b) axial components of
electric field along x- and y-axes
the potential across a wear track 2 mm above the disk surface is shown in
Fig. 4.14b. In order to evaluate the effect of non-uniformity of the wear track on
surface potential, simulation was done for various widths of the wear track. It was
found that the profile of surface potential depended only on the linear charge
density, but not on the geometry.
The components of electric field on the surfaces of the pin and the disk are shown
in Fig. 4.15. The components Exx and Eyy, i.e., radial components along axes x and
y, match reasonably well both on the disk and the pin surfaces revealing almost
perfect axial symmetry of electric field.
Both radial components are positive and gradually increase when approaching
the contact zone. This behavior is typical for electric field of a point positive charge.
Near the contact zone both radial components reach very high values that can be
physically unreasonable since possible dielectric breakdown of solids has been
ignored in this simple model. Nevertheless, previous experimental studies have
revealed that electric field as strong as 107 V/m can be readily generated at the local
surface regions [55]. The radial component along the wear track (y direction) was
slightly weaker than in x direction because of low-density negative charge on the
wear track.
The axial components of the electric field, Ezx and Ezy (Fig. 4.15b), have also
very similar behavior that underpins the argument for radial symmetry of electric
field. However, the axial components are negative close to the contact zone and get
positive at the distances from 48 to 80 μm from the contact zone. This configuration
can be attributed to the dielectric displacement field induced in the disk material. At
larger distances from the contact zone all axial component tend to converge. Only a
minor difference is maintained between Ez on the disk and the pin along y-axis
because of negative charge on the wear track.
These findings have highlighted that the negative charge has only a minor effect
on the configuration of electric field, but it can be relevant to seeding electron
emission from the disk. Electric field is similar to that of the point positive charge,
i.e. with the main component in a radial direction. This conclusion contrasts with
the common belief that electric field is oriented across the gap between the pin and
disk, predominantly along z axis [34, 36, 37, 43, 56]. It should be stressed that the
developed model corresponds to undistorted electric field in vacuum in absence of
electric gas discharge. Volume charge of electrons and ions in triboplasma and
recombination processes on the surfaces may influence the initial electric field
configuration.
A considerably non-uniform configuration of the electric field with a highly
stressed anode (pin) and a low field cathode (disk) is typical for positive corona
[57]. Furthermore, triboplasma should have certain similarity with dielectric barrier
discharge because of dielectric materials of the electrodes and with asymmetric
point-to-plane discharge because of the overall geometry. However, triboplasma
does not match any of the known types of gas discharge. It can be tentatively
defined as positive corona DC barrier discharge.
Although further simulation is necessary for deeper understanding of
triboplasma, the obtained results are useful for deciphering some of its fundamental
processes such as triggering and impact ionization. Emission of triggering electrons
from negatively charged disk has been associated with liberation of electrons from
traps of different types associated with zero-to-three dimensional defects, local
deviation from oxides stoichiometry, presence of alkali atoms, growing cracks, etc.
[58, 59]. Taking into account the configuration of electric field, it is reasonable to
suggest that the triggering electron emission (TEE) has to occur predominantly in
the regions in front and behind the contact area, where the axial component of the
electric field is negative. At distances further away than 80 μm from the contact area
only electrons with relatively high energies can be liberated from the cathode
because of the positive axial component of electric field. This hypothesis has
been corroborated through spatially resolved visualization of electron triboemission
[60]. The most common distribution was approximately circular in shape coincid-
ing with the geometry of the wear track with the maximum intensity close to the
front of the contact. Sporadic postcontact emission from the wear track has been
detected as well [60, 61]. Emitted electrons have energy ranging from several eV up
to 48 eV [62].
After being liberated the initial electrons are accelerated by the radial component
of electric field towards the contact zone and experience elastic and inelastic
collisions with gas molecules during travelling along radial direction in the gap
between the pin and the disk. Bearing in mind that for positive corona the cathode
does not furnish a secondary mechanism of electron emission through ion collisions
[63], secondary mechanisms must supply seed electrons that start electron ava-
lanches at the remote regions, where initial liberation of electrons from the cathode
is hindered. These secondary processes for Ar include metastable atoms collision
[64] and photoionization. The energy required for excitation of metastable levels of
Ar is around 11.62 eV [63], while for ionization it should be above 100 eV.
Considering mean energy gained by electrons and ions as function of gas pressure
in the pressure range from 100 to 10 kPa the conditions are favorable for effective
coupling between electron impact ionization, secondary processes and triggering
electron emission. So, gas discharge is stable and luminescence is uniform. How-
ever, ion energy is not sufficient to produce ionoluminescence because of very short
Arþ MFP being in the submicrometer range. At gas pressure below 2 kPa corona
pulsations can be associated with charging-discharging oscillations rather than with
intrinsic corona instabilities that usually have frequency higher than 100 kHz
[40]. This behavior can be explained considering that ionization probability
decreases with decreasing gas pressure below the Stoletov’s point that requires
higher voltage for electric breakdown than under optimal conditions (Paschen’s
law). When the rate of charge generation due to triboelectrification is not sufficient
for compensation of charge recombination at the pin during discharge the potential
decreases below the breakdown value and the discharge ceases. Then the potential
increases again and the cycle repeats. Since surface charges are not completely
recombined between the bursts a highly non-uniform distribution of surface poten-
tial can be formed as it was experimentally observed [47].
More clues for understanding the spatial features of luminescence can be
obtained from the analysis of the configuration of electric field in the plane of the
disk surface (Fig. 4.16). Potential distribution inside the contact area obeys axial
symmetry with respect to y, so the gradient of electric field on the sides of the
contact zone is larger, than at the front and behind the contact. Larger gradient of
electric field enhances corona discharge and can be related with higher lumines-
cence intensity. Potential steps on the sides of the wear track that roughly coincide
with the positions of the spots of ionoluminescence can promote extraction of the
ions from triboplasma and their acceleration towards the disk surface. These
findings have manifested that luminescence of triboplasma, as contrast to solid
triboluminescence, is specific to the overall geometry of mating solids.
Fig. 4.16 (a) Potential distribution at the disk surface; (b) electric field in yz central cross section
4.5 Triboluminescence from Polymers
As most polymers are electrical insulators, when they are in rubbing or rolling
contact with other materials, electric charges are accumulated on their surfaces
[65–67]. It results in subsequent discharge through surrounding gas or liquid. The
discharge is often observed by naked eye when clothes, if made of non-conductive
polymers, are rubbed together with each other or with our skin. The phenomenon of
triboluminescence is the manifestation of triboplasma on polymer surfaces.
Although it is quite common in daily life, the luminescence characteristics have
not been fully investigated yet. On the other hand, the luminescence from organic
compounds associated with oxidation has been studied as chemiluminescence
[68]. Bearing in mind that friction enhances material oxidation through
tribochemical activation, luminescence due to the tribochemical oxidation would
also be expected. Accordingly, luminescence from polymers has to have the
following two possible causes: oxidation reaction and triboplasma that are related
to chemical activity and triboelectrification, respectively.
To identify which factor is dominant, researches on triboluminescence have
been conducted by utilizing a chemiluminescence test rig [69]. In that study the
measurement of triboluminescence was carried out simultaneously with that of
triboelectrification among eight kinds of polymers and the effect of type of friction
(sliding or rolling) and humidity on both triboluminescence and triboelectrification
was clarified. Using this test rig, further experiments have been done with focusing
on the origin of triboluminescence from polymers in rubbing contact [70].
The experimental setup is schematically drawn in Fig. 4.17. A twin-ring
tribometer [71] was installed in a dark box chamber where photons were detected
by the photomultiplier. The polymer ring was subjected to abrasion by the teeth of
the knurling die. Both specimens rotated in the same direction, so that their
cylindrical peripheries were in sliding contact at the contact interface. During the
rubbing experiment, a pure gas such as nitrogen, oxygen, argon or nitrous oxide was
Fig. 4.17 Twin-ring tribometer enclosed in a chemiluminescence chamber to measure photons

from the sliding contact between polymer (20 mm in diameter) and knurling die (30 mm in
diameter) or another polymer (20 mm in diameter). The knurling die was made of steel (SKH 51).
The applied load was 2 N between polypropylene (PP) and knurling die, or 1 N for the rubbing
between polymers. The relative sliding velocity was 400 mm/s (reprinted with permission from
[70] © 2008 JAST)
Fig. 4.18 The edge of the knurling die after the rubbing against polypropylene (reprinted with
permission from [70] © 2008 JAST)
blown to the interfacial area through the nozzle, so that residual air inside the
chamber was gradually displaced by the selected gas.
Figure 4.18 shows the cutting blade of the knurling die after the test. A number
of wear particles could be seen on its surface. However, the edge of the blade
Fig. 4.19 (a) Effect of gas blow on triboluminescence of PP rubbed against knurling die. The load
was 2 N; sliding velocity was 400 mm/s. (b) Enlarged portion of the plots shown in figure (a). A
plot in orange represents mean luminescence intensity for N2O gas. These recent data do not form
a part of the original figure from [70] and are given for comparison (reprinted with permission
from [70] © 2008 JAST)
remains the same as original without wear nor transfer layer from polypropylene
(PP). It means that abrasive wear continued constantly until the end of the test.
Figure 4.19 shows the effect of gas blow on the photon count from the rubbing
between the knurling die and polypropylene. The vertical axis is the luminescence
intensity represented by the total number of photon counts in the whole spectral
range of the photomultiplier. When the rubbing was continued for 1000 s with no
gas blow, the photon intensity slightly decreased but remained nearly constant at
the rate about 1500 counts per second.
After the initial rubbing in air for 200 s that was used to stabilize the photon
emission intensity, the gas was blown to the contact interface. Nitrogen and argon
led to increase in the registered intensity of photon emission. The rate of photon
emission in nitrogen finally reached steady value about 14 kilocounts per second
after 6000 s. This value was almost tenfold greater than in the air. On the contrary,
introduction of oxygen and nitrous oxide led to decrease in the rate of photon
emission. The signals of emission intensity in oxygen and nitrous oxide reached
steady values after 200 s from the start of the gas blow that was much earlier as
compared with nitrogen.
Considering the possible causes for luminescence from polymers in a tribo-
contact the following have been identified:
1. Radical reactions of hydrocarbons produced through mechanical actions
2. Carbonyl group formations associated with the oxidation of hydrocarbons
3. Excitation of atmospheric molecules due to the attachment of electrons emitted
from the rubbing surface
4. Fluorescence or phosphorescence of polymers
The experiments with different gases shown in Fig. 4.19 clarified which mech-
anism is predominant. Radical reactions in polymers were discarded as the origin of
the light emission since they cannot explain variation in luminescence intensity
with change of the surrounding gas. If the formation of carbonyl groups had
something to do with the luminescence, the oxygen blow would be the most
effective in increasing the emission intensity [72]. Nevertheless, the oxygen blow
suppressed the emission of light. The finding that luminescence intensity gradually
varied when a certain gas was being blown into the chamber, e.g., the rate of photon
emission gradually increased during introduction of nitrogen and argon, indicates
that the mechanisms of luminescence has to be electric discharge of these gases.
The common property of nitrogen and argon is their zero or slightly negative
electron affinity. It is highly contrasted with the positive electron affinity for oxygen
(0.45 eV) and nitrous oxide (1.46 eV). Therefore, for nitrogen and argon electron
avalanches could be formed easier under electron bombardment that promoted gas
discharge in these gases. Stabilization of luminescence intensity can be related with
stabilization of gas composition in the chamber. On the other hand, the positive
electron affinity of oxygen and nitrous oxide led to capture of electrons and
generation of negative ions. This process quenched electron avalanches and
suppressed gas discharge. Accordingly, photon emission also decreased.
By replacing the knurling die with a polymer disk, adhesive wear mode was
accomplished. It is known that polyamide (PA66) and polytetrafluoroethylene
(PTFE) are the most distant polymers in the triboelectric series. Combination of
these polymers may induce high potential difference when they are rubbed together
(tribocharging). Therefore, they were used in these experiments. In addition,
polymer specimens of both PTFE and PA66 were separately heated up to 150 C
in laboratory air without rubbing in order to evaluate a contribution from chemilu-
minescence and contrast these results with luminescence under rubbing.
Figure 4.20 shows a comparison of chemiluminescence of PA66 and PTFE and
triboluminescence induced by rubbing of these polymers. When PA66 was heated
up, the photon emission rate slowly increased with a transitional valley at 400 s
before it stabilized. No light was detected from the heating of PTFE. On the
contrary, in the triboluminescence from the rubbing between PTFE and PA66,
photons were emitted immediately after the beginning of rubbing, then emission
rate gradually decreased and finally stabilized. The abrupt increase in tribolumi-
nescence just after the start of rubbing means that the cause for the luminescence
should not be frictional heating, because the specimen temperature must be the
lowest at the beginning and gradually increase with rubbing.
After triboluminescence stabilized at the minimum value at the end of the test,
rubbing was stopped. Then the specimens were allowed to rest in the test rig. After
24 h, the friction test was resumed. An increase was observed at the onset of sliding,
as demonstrated in Fig. 4.21. This sudden increase and gradual decrease was found
every time when the test was resumed after a certain rest time. During the initial
preparation prior to the first test and during the rest time the surfaces were exposed
to the atmosphere. Therefore, the interaction between the surfaces and surrounding
Fig. 4.20 Time evolutions of photon count in chemiluminescences of PA66 and PTFE, and
photon count in triboluminescence from the rubbing between them. In chemiluminescence mea-
surements, both specimens were heated up to 150 C. In triboluminesence measurement, the load
was 1 N. Relative sliding velocity was 400 mm/s
Fig. 4.21 Initial peak at the

onset of sliding between
PA66 and PTFE after the
rest of 24 h. Load was 1 N.
Relative sliding velocity
was 400 mm/s
gas molecules, such as adsorption or condensation, or any kind of interaction would

be the causes for the initial peak in photon intensity.
The effect of surface water on the triboluminescence was further tested using a
wet specimen. For this purpose PA66 sample was specially prepared by immersing
in distilled water for 15 h. The results of the rubbing test of this PA66 and normal
PTFE are shown in Fig. 4.22. From the start of sliding and through 1200 s there was
no luminescence detected. Then the luminescence suddenly emerged until the
maximum peak was recorded. It gradually decreased and stabilized in the time
interval between 8000 and 12,000 s. The sudden increase indicates that a minute
difference in surface water concentration could significantly change the lumines-
cence, while there should be an optimum concentration of water on the surface to
18000
16000
Photon Intensity, a.u.

14000
12000
10000
8000
6000
4000
2000
0
0 2000 4000 6000 8000
Time, s
Fig. 4.22 Triboluminescence in the drying up process of water on the surface of PA66 rubbed
against PTFE. Load was 1 N. Relative sliding velocity was 400 mm/s
induce maximum triboluminescence. This luminescence behavior should also be

related with the wear behavior of polymers, because polymer wear is greatly
influenced by humidity. It has been generally believed that surface water inhibits
triboluminescence. These experimental results, however, indicate that traces of
surface water would enhance triboluminescence. Most probably, the surface
water causes the surface fracture and wearing resulting in the electron emission
and luminescence. Another reason could be related with surface charge redistribu-
tion due to dissociation of adsorbed water [55].
4.6 Triboplasma on Coated Dielectrics
It is generally believed that application of thin conductive layer on the top of a

dielectric solid can completely inhibit triboplasma and suppress associated emis-
sion of photons and charged particles. In fact, the experiments with nanometer thick
Al films deposited on soda lime glass demonstrated that 30 nm film was effective in
cancellation of triboemission in the whole range of applied normal load (Fig. 4.23)
[73]. However, a film of 5 nm showed complex behavior. Emission was negligible
at the beginning of sliding. After an incubation period, emission intensity increased
until it reached a steady value. The steady value increased with the normal load but
remained three- to fourfold smaller than the emission from bare glass. Duration of
the incubation period spanned from several tens to several thousands of disk
revolutions.
Analysis of the wear tracks showed considerable wear of the film in all cases
(Fig. 4.24). The thicker coating was completely gone in the center of the wear track
so that glass substrate was in direct contact with the diamond pin, but the coating
remained continuous on the sides. Therefore the charged generated in the center of
a b
250 FN =0.2 N 400 bare glass
bare glass
Emission rate (pC/s)

Emission rate (pC/s)
200
199±10.4 pC/s 300
150
Al covered glass 200
100 (Δ=5 nm) Al (Δ 5 nm)/glass
59±15 pC/s
50 100
Al (Δ 30 nm)/glass
0 0
0 500 2500 3500 0.0 0.2 0.4 0.6 0.8 1.0
Time (s) Normal load (N)
Fig. 4.23 (a) Time behavior of the triboemission under scratching bare glass substrate and 5 nm
thick Al film on glass. Normal load FN ¼ 0.196 N; (b) effect of normal load on the emission rate
from bare glass as well as from substrates covered with 5 and 30 nm thick Al films (reprinted from
[73], © 2002, with permission from Elsevier)
Fig. 4.24 Microscopic images of the scratched zones of the samples after performing experi-
ments. Bright—aluminum, dark—glass substrate (reprinted from [73], © 2002, with permission
from Elsevier)
the wear track dissipated through the coating. For the thinner coating, only isolated
spots of Al could be found in the wear track. The number and size of the spots
decreased with increasing normal load. Despite the metal spots triboplasma could
appear. From these findings it has been concluded that presence of discontinuous
conductive coating can damp triboplasma but not inhibit it. The effect of the metal
spots can be ascribed to reducing the local strength of electric field and partial
suppression of triggering electron emission.
4.7 Concluding Remarks
Luminescence of triboplasma is a complex phenomenon that remains poorly

understood. In contrast to fractoluminescence, it was experimentally studied for
only limited number of materials. However, its significance in tribology can hardly
be overestimated because of its influence on durability of advanced tribological
materials and lubricants [60, 74]. Gas discharge at a sliding contact differs from
other types of gas discharge and can be tentatively defined as positive corona DC
barrier discharge. Triboluminescence was found in tribo-contacts under both liquid
[75] and solid lubrication [76]. Similarity of the triboluminescence spectrum
obtained under oil lubrication and the spectrum of electric discharge in air
suggested that triboplasma can emerge in air bubbles inside the oil or in voids
that can be formed behind the contact zone for certain combination of the oil
viscosity, thickness of the oil layer, and sliding velocity.
Triboluminescence and triboemission can also be used as a probe for
non-destructive characterization of damage of materials in the contact zone (patent
application JP20020366302). Onset of triboemission for coated glass (Fig. 4.23)
correlates with the damage of the metal coatings. Intensity of triboluminescence is
sensitive to geometry of the solids subjected to friction and surface roughness.
References
1. Walton, A. J. (1977). Triboluminescence. Advances in Physics, 26, 887–948.

854–870.
3. Abramova, K. B., Vettegren’, V. I., Shcherbakov, I. P., Rakhimov, S. S., & Svetlov, V. N.
(1999). Mechanoluminescence and submicrorelief of a copper surface. Technical Physics, 44,
1491–1493.
4. Banishev, A. F., Panchenko, V. Y., & Shishkov, A. V. (2003). Nonthermal glow of thin metal
plates and films exposed to Pulsed Laser Radiation. Technical Physics, 48, 612–615.
5. Tsuboi, Y., Seto, T., & Kitamura, N. (2008). Laser-induced shock wave can spark tribolumi-
nescence of amorphous sugars. The Journal of Physical Chemistry A, 112, 6517–6521.
6. Abramova, K. B., Semenov, A. A., & Shcherbakov, I. P. (2001). Dynamics of photon emission
due to strains in metals. Technical Physics, 46, 1396–1400.
7. Abramova, K. B., Vettegren’, V. I., Shcherbakov, I. P., & Svetlov, V. N. (2002). The emission
of photons and the dynamics of submicrodefects on the surface of noble metals. Technical
Physics, 47, 268–271.
8. Bhat, K., Fontenot, R., Surabhi, R., Hollerman, W., Aggarwal, M., & Alapati, T. (2014).
Measurement of the triboluminescent properties for europium and samarium tetrakis
dibenzoylmethide triethylammonium. Electronic Materials Letters, 10, 1149–1153.
9. Chandra, B. P. (1998). Mechanoluminescence. In D. R. Vij (Ed.), Luminescence of solids.
New York: Springer.
10. Zhang, J.-C., Long, Y.-Z., Wang, X., & Xu, C.-N. (2014). Controlling elastico-
mechanoluminescence in diphase (Ba, Ca)TiO3:Pr3þ by co-doping different rare earth ions.
RSC Advances, 4, 40665–40675.
11. Heinicke, G. (1984). Tribochemistry. Munchen: Carl Hanser Verlag.
12. Nakayama, K., & Nevshupa, R. A. (2002). Plasma generation in a gap around a sliding contact.
Journal of Physics D: Applied Physics, 35, L53–L56.
13. Takacs, L. (2013). The historical development of mechanochemistry. Chemical Society
Reviews, 42, 7649–7659.
14. Urakaev, F. (2007). Mechanodestruction of minerals at the crack tip. Overview: 1. Experiment.
Physics and Chemistry of Minerals, 34, 351–361.
15. Weiser, H. B. (1917). Crystalloluminescence II. The Journal of Physical Chemistry, 22,
576–595.
16. Rusanov, A. I. (2002). Thermal effects in mechanochemistry. Russian Journal of General
Chemistry, 72, 327–344.
17. Butyagin, P. Y. (1971). Kinetics and nature of mechanochemical reactions. Russian Chemical
Reviews, 40, 901–915.
18. Craig, S. L. (2012). Mechanochemistry: A tour of force. Nature, 487, 176–177.
19. Fischer, T. E. (1988). Tribochemistry. Annual Review of Materials Science, 18, 303–323.
20. Nevshupa, R. A. (2004). Triboemission: an attempt of generalized classification. In C. Kajdas
(Ed.), Tribology: Science and applications. Vienna: PAS.
21. Varentsov, E. A., & Khrustalev, Y. A. (1995). Mechanoemission and mechanochemistry of
molecular organic crystals. Russian Chemical Reviews, 64, 783–797.
22. Nakayama, K., & Nevshupa, R. A. (2003). Characteristics and pattern of plasma generated at
sliding contact. Journal of Tribology-Transactions of the ASME, 125, 780–787.
23. Longchambon, H. (1925). Etude spectro-photographique des phénomènes de tribolumines-
cence et de cristalloluminescence. Bulletin de la Societe Francaise de Mineralogie et de
Cristallographie, 48, 130–214.
24. Sweeting, L. M., Cashel, M. L., Dott, M., Gingerich, J. M., Guido, J. L., Kling, J. A.,
et al. (1992). Spectroscopy and mechanism in triboluminescence. Molecular Crystals and
Liquid Crystals Science and Technology: Section A. Molecular Crystals and Liquid Crystals,
211, 389–396.
25. Butyagin, P. Y. (1970). The luminescence accompanying mechanical deformation and rupture
of polymers. Polymer Science (Visokomoleculiarnie Soedinenia), A12, 290–299.
26. Kricka, L. J., Stroebel, J., & Stanley, P. E. (1999). Triboluminescence: 1968–1998. Lumines-
cence, 14, 215–220.
27. Orel, V. E., Kadyuk, I. N., Dzyatkovskaya, N. N., Danko, M. I., & Mel’nic, Y. I. (2000).
Mechanoluminescence: Lymphocyte analysis after exposure to ionizing radiation. Lumines-
cence, 15, 29–36.
28. Sweeting, L. M., Cashel, M. L., & Rosenblatt, M. M. (1992). Triboluminescence spectra of
organic crystals are sensitive to conditions of acquisition. Journal of Luminescence, 52,
281–291.
29. Miura, T., Chini, M., & Bennewitz, R. (2007). Forces, charges, and light emission during the
rupture of adhesive contacts. Journal of Applied Physics, 102, 103509. 6 pages.
30. Gu, Z., Wei, W., Su, J., & Yu, C. W. (2013). The role of water content in triboelectric charging
of wind-blown sand. Scientific Reports, 3, 1337.
31. Nevshupa, R. (2009). The role of athermal mechanisms in the activation of tribodesorption and
triboluminisence in miniature and lightly loaded friction units. Journal of Friction and Wear,
30, 118–126.
32. Nevshupa, R. A., Roman, E., & DE Segovia, J. L. (2010). Model of the effect of local frictional
heating on the tribodesorbed gases from metals in ultra-high vacuum. International Journal of
Materials and Product Technology, 38, 57–65.
33. Miura, T., Hosobuchi, E., & Arakawa, I. (2009). Spectroscopic studies of triboluminescence
from a sliding contact between diamond, SiO2, MgO, NaCl, and Al2O3 (0001). Vacuum, 84,
573–577.
34. Miura, T., & Nakayama, K. (2001). Two-dimensional spatial distribution of electric-discharge
plasma around a frictional interface between dielectric surfaces. Applied Physics Letters, 78,
2979–2981.
35. Muto, J., Nagahama, H., Miura, T., & Arakawa, I. (2007). Frictional discharge at fault
asperities: Origin of fractal seismo-electromagnetic radiation. Tectonophysics, 431, 113–122.
36. Muto, J., Nagahama, H., Miura, T., & Arakawa, I. (2008). Frictional discharge plasma and
seismo-electromagnetic phenomena. Physics of the Earth and Planetary Interiors, 168, 1–5.
37. Nakayama, K. (2010). Microplasma generation in gap of sliding contact through discharging
of ambient gas due to triboelectrification. International Journal of Plasma Environmental
Science & Technology, 4, 148–153.
38. Nakayama, K., & Nevshupa, R. A. (2004). Effect of dry air pressure on characteristics and
patterns of tribomicroplasma. Vacuum, 74, 11–17.
39. Nevshupa, R. (2013). Effect of gas pressure on the triboluminescence and contact electrifica-
tion under mutual sliding of insulating materials. Journal of Physics D: Applied Physics, 46,
185501.
40. Raizer, Y. P. (1991). Gas discharge physics. New York: Springer.
41. Miura, T., & Nakayama, K. (2000). Spectral analysis of photons emitted during scratching of
an insulator surface by a diamond in air. Journal of Applied Physics, 88, 5444–5447.
42. Kaufman, V., & Edlén, B. (1974). Reference wavelengths from atomic spectra in the range
15 Å to 25000 Å. Journal of Physical and Chemical Reference Data, 3, 825–895.
43. Nakayama, K. (2010). Triboplasma generation and triboluminescence: Influence of stationary
sliding partner. Tribology Letters, 37, 215–228.
44. Nevshupa, R., & Hiratsuka, K. (2015). Triboluminescence. In E. Gnecco & E. Meyer (Eds.),
Fundamentals of friction and wear on the nanoscale. New York: Springer.
45. Nakayama, K., & Nevshupa, R. (2002). Pattern and spectral characteristics of
microtriboplasma. In JAST Tribology Conference, Tokyo (pp. 1–2). Tokyo: JAST.
46. Tohmon, R., Shimogaichi, Y., Munekuni, S., Ohki, Y., Hama, Y., & Nagasawa, K. (1989).
Relation between the 1.9 eV luminescence and 4.8 eV absorption bands in high‐purity silica
glass. Applied Physics Letters, 54, 1650–1652.
47. Nevshupa, R. A., & Nakayama, K. (2003). Triboemission behavior of photons at dielectric/
dielectric sliding: Time dependence nature at 104–104 s. Journal of Applied Physics, 93,
9321–9328.
48. Biagi, S. F. (2012). Momentum transfer cross-section Biagi-v8.9. Plasma Data Exchange
Project.
49. Harper, W. R. (1967). Contact and frictional electrification. Oxford: Clarendon.
50. Ireland, P. M. (2008). The role of changing contact in sliding triboelectrification. Journal of
Physics D: Applied Physics, 41, 025305.
51. Kluev, V. A., Vladikina, T. N., Toporov, Y. P., Anisimova, V. J., & Derjaguin, B. V. (1978).
Emission phenomena accompanying the triboelectrification process in vacuum. IEEE Trans-
actions on Industry Applications, IA-14, 544–546.
52. Kornfeld, M. I. (1976). Frictional electrification. Journal of Physics D: Applied Physics, 9,
1183–1192.
53. Bach, H., & Neuroth, N. (1995). The properties of optical glass. New York: Springer.
54. Labadz, A. F., & Lowell, J. (1991). Contact charge density and penetration depth. Journal of
Electrostatics, 26, 251–260.
55. Gouveia, R. F., & Galembeck, F. (2009). Electrostatic charging of hydrophilic particles due to
water adsorption. Journal of the American Chemical Society, 131, 11381–11386.
56. Muto, J., Nagahama, H., Miura, T., & Arakawa, I. (2006). Frictional discharge plasma from
natural semiconductor/insulator junctions: Origin of seismo-electromagnetic radiation. Phys-
ics and Chemistry of the Earth, Parts A/B/C, 31, 346–351.
57. Samsonenko, S., Samsonenko, N., & Timchenko, V. (2010). Dislocation electrical conductiv-
ity of plastically deformed natural diamonds. Semiconductors, 44, 1140–1144.
58. Blaise, G. (2001). Charge localization and transport in disordered dielectric materials. Journal
of Electrostatics, 50, 69–89.
59. Mclellan, G. W., & Shand, E. B. (1984). Glass engineering handbook. New York: McGraw-
Hill.
60. LE Rouzic, J., & Reddyhoff, T. (2014). Spatially resolved triboemission measurements.
Tribology Letters, 55, 245–252.
61. Molina, G. J., Furey, M. J., Ritter, A. L., & Kajdas, C. (2001). Triboemission from alumina,
single crystal sapphire, and aluminum. Wear, 249, 214–219.
62. Molina, G. J. (2000). Triboemission from ceramics: Charge intensity and energy distribution
characterizations. Ph.D. Thesis, Virginia Polytechnic Institute and State University.
63. Loeb, L. B. (1965). Electrical coronas, their basic physical mechanisms. Berkley: University
of California Press.
64. Sigmond, R. S. (1997). The oscillations of the positive glow corona. Le Journal de Physique
IV, 7, 383–395.
65. Baytekin, H. T., Patashinski, A. Z., Branicki, M., Baytekin, B., Soh, S., & Grzybowski, B. A.
(2011). The mosaic of surface charge in contact electrification. Science, 333, 308–312.
66. Lacks, D. J., & Sankaran, R. M. (2011). Contact electrification of insulating materials. Journal
of Physics D: Applied Physics, 44, 453001.
67. Williams, M. W. (2012). Triboelectric charging of insulating polymers—some new perspec-
tives. Aip Advances, 2, 010701 (p. 9).
68. Beutel, J. (1971). Chemiluminescence in oxidation reactions. Oxidation mechanism of
dimedone. Journal of the American Chemical Society, 93, 2615–2621.
69. Hiratsuka, K. I., & Hosotani, K. (2012). Effects of friction type and humidity on
triboelectrification and triboluminescence among eight kinds of polymers. Tribology Interna-
tional, 55, 87–99.
70. Hosotani, K., & Hiratsuka, K. (2008). Triboluminescence in the rubbing of some polymers. In
JAST Tribology Conference 2008–9, Nagoya. JAST (pp. 527–528).
71. Hiratsuka, K., & Yoshida, T. (2011). The twin-ring tribometer—Characterizing sliding wear of
metals excluding the effect of contact configurations. Wear, 270, 742–750.
72. Blakey, I., & George, G. A. (2001). Simultaneous FTIR emission spectroscopy and chemilu-
minescence of oxidizing polypropylene: Evidence for alternate chemiluminescence mecha-
nisms. Macromolecules, 34, 1873–1880.
73. Nevshupa, R. A., & Nakayama, K. (2002). Effect of nanometer thin metal film on
triboemission of negatively charged particles from dielectric solids. Vacuum, 67, 485–490.
74. Nakayama, K. (2007). The plasma generated and photons emitted in an oil-lubricated sliding
contact. Journal of Physics D: Applied Physics, 40, 1103–1107.
75. Nakayama, K. (2011). Mechanism of triboplasma generation in oil. Tribology Letters, 41,
345–351.
76. Nakayama, K., & Nguyen, S. (2000). Triboelectromagnetic phenomena in a diamond/hydro-
genated-carbon-film tribosystem under perfluoropolyether fluid lubrication. Applied Surface
Science, 158, 229–235.
Chapter 5
Triboluminescence of Liquid Dielectrics:
On a Way to Discover the Nature
of Sonoluminescence
D.A. Biryukov and D.N. Gerasimov
5.1 The Mystery of Sonoluminescence
We understand (or we think we understand, at least) the processes which occurred

at the very first second of our Universe’s existence. We know enough about the
neutrino oscillations. We can observe the surface of Pluto, but we cannot under-
stand the simplest processes in the jar of liquid. Or not the simplest?
5.1.1 Light from Sound
Liquids emit light under effect of ultrasonic pressure. Exactly, liquids emit light
under the influence of non-ultrasonic pressure too: for instance, in our experiments
we treat luminescence of glycerol in the “ultrasound” field of 20.5 kHz; according
to ordinary scale, this frequency belongs to the aural range (despite on the fact that
not every human can really hear such a tone).
This phenomenon—sonoluminescence (SL)—was discovered in 1934 by
Frenzel and Schultes [1]: they observed luminescence of water in ultrasound field
at 500 kHz. There were many studies on SL since those times, but until the present
days there is no clear understanding of the nature of such light. Or, in more accurate
words, there is no common view on it: many researchers have their own, quite solid
D.A. Biryukov
Joint Institute for High Temperatures of the Russian Academy of Sciences, Izhorskaya,
13, Bd. 2, Moscow 125412, Russian Federation
D.N. Gerasimov (*)
National Research University Moscow Power Engineering Institute, Krasnokazarmennaya,
14, Moscow 111250, Russian Federation
e-mail: deniger@mail.ru

96 D.A. Biryukov and D.N. Gerasimov
Fig. 5.1 Experimental setup: 1—titanium waveguide on the magnetostrictive drill, 2—jar with a
liquid, 3—ultrasonic wave generator, 4—temperature monitor with thermocouple, 5—tank for
cooling water, 6—pump, 7—fan, 8—radiator; arrows show the flow in the cooling system
opinions on the nature of SL, but these opinions contradict opinions of their
colleagues. Some researchers believe that this light originates from electrical
discharges in bubble, and others suppose that this luminescence is a kind of thermal
radiation; there also exists an opinion that different kinds of SL have different
explanations [2].
SL can be observed as a couple of dimly shone bubbles (so-called multi-bubble
sonoluminescence, MBSL; this regime was treated by Frenzel and Schultes), or as a
single bright flashing bubble (single-bubble sonoluminescence, SBSL, discovered
in 1992 by Gaitan et al. [3]). This is the traditional classification; however, the first
term—MBSL—implies very different regimes of luminescence.
In this chapter we present our images—patterns of multi-bubble luminescence;
different pictures were obtained on a single experimental setup (ultrasonic wave
generator IL10-2.0 with magnetostrictive drill: 2 kW power unit with titanium
waveguide with diameter 3.5 cm; experimental setup is presented on Fig. 5.1)
with a single liquid (mixture of 90 % glycerol þ 10 % water; room temperatures).
The only regulated parameter was the frequency of ultrasonic generator: magneto-
strictive drill has a single resonant frequency (~20.5 kHz), but we were slightly
changing the input frequency, adjusting it out of resonance; this is some dangerous
procedure for equipment, by the way—fraught with grave consequences for it.
Multi-bubble sonoluminescence is a very dim light. MBSL can be observed only
in a complete darkness, and even in such conditions—where you cannot see your
own hand—an imperfect vision can miss SL: one of our colleagues did not see any
sonoluminescence except when observed on the computer monitor after video
camera (we used Advert AD-9346V faint object camera with Fujinon 2.8–8 mm
F0.95 lens). All our images of SL are actually frames from videos.
5 Triboluminescence of Liquid Dielectrics. . . 97
Fig. 5.2 Multi-bubble

sonoluminescence as it is
Fig. 5.3 According to the

traditional classification,
this is MBSL too. But we
prefer the term “filamentous
sonoluminescence”
In Fig. 5.2 one can see a plane, ordinary MBSL: a cloud of luminous bubbles.
But this not the only regime of MBSL. There also exists “filamentous regime” of
SL: numerous “sparks,” often organized in regular structures (Fig. 5.3) on ultra-
sonic waveguide; we use a special term to point out the particularity of this pattern.
The origin of this strange configuration can be explained by Fig. 5.4: actually,
sometimes cavitation bubbles organize regular structures (kind of foam); structures
of bubbles on ultrasonic waveguide was mentioned in [4], where authors used the
term “fractional clusters” for them. Thus, bubbles on Fig. 5.4 actually emit light,
but we cannot see it at the external light. Note that every “filament” that we have
seen—spark-similar pattern of SL, including “sparks” from ultrasonic waveguide
into the bulk of liquid—was a couple of bubbles, at a steadfast attention.
Fig. 5.4 Structure from

micro-bubbles (a foam) on
the ultrasonic waveguide;
this is not a luminescence:
image has been made with
external (violet) lightning.
Bubbles are organized in a
regular structure (compare
this image with Fig. 5.3)
Fig. 5.5 Light emission

from liquid on the ultrasonic
waveguide
Light emission of liquid directly on ultrasonic waveguide (Fig. 5.5) is a very

simple and the most frequent case of SL; this is the basic pattern for strict resonance
of magnetostrictive drill. Single luminous bubbles can be observed in the surround-
ings of the bubble cloud (Fig. 5.6), so we can say—with some stretch—that MBSL
and SBSL can coexist. On the other side, considering the fact that many people
believe that MBSL and SBSL have different physical nature, one can say that light
emission from luminous single bubble is not SBSL (sic!).
Cavitation is a key aspect of sonoluminescence. It was shown many times in
many works that light is let out from the cavitation bubbles, so the processes inside
vapor phase define this phenomenon. What are these processes? We consider them
below, but all such considerations are speculative: all that we have are
Fig. 5.6 MBSL (center)

and SBSL (upper left)
Fig. 5.7 Multi-bubble

sonoluminescence in liquid
(bottom) and light emission
from the ultrasonic
waveguide (upper right)
“circumstantial evidences” of the nature of SL. The intriguing part of these “evi-
dences” are tracks of SL on a waveguide.
As we know, cavitation may destruct objects (propeller blades, etc.) due to hard-
pressure effects on their surfaces. What are connections inside the “surface–
cavitation–light” triad? On Fig. 5.7 one can see the interesting pattern of light
emission: usual MBSL in the bulk of liquid and the bright “flash” on the waveguide.
Actually, this “flash” is continuous luminescence: the light was emitted from the
stationary location on waveguide. Why this light was emitted from that position?
Possibly, there was a trace of SL on the metallic surface? On Fig. 5.8 we can see
that trace (dark trail from the cavern to the edge of waveguide); the structure of the
Fig. 5.8 Place of intensive light emission on the waveguide (Fig. 5.7) after experiment
Fig. 5.9 The cavern on the waveguide—the source of bright luminescence from Fig. 5.7
cavern is presented on Fig. 5.9. Photos on Figs. 5.8 and 5.9 have been made after
~3 min. of work (cavitation and SL); the structure of the cavern before this series of
experiments is presented on Fig. 5.10. Thus only 3 min. of intensive cavitation
(with intense luminescence) made rather dramatic changes in the titanium surface.
We can conclude that not only cavitation leads to the destruction of surface;
there exists feedback too: decayed surface—the cavern on Fig. 5.10—plays a role
Fig. 5.10 The cavern on the waveguide before the experiment shown on Fig. 5.7
of a “cavitation center,” and the light emission from this center is incredibly bright
(Fig. 5.7). So, the connection is the following: cavitation þ SL ! cavern !
cavitation þ brighter SL ! larger cavern etc. Looking at Figs. 5.8–5.10, one may
say that “something burnt on this surface.” And we know what it was:
sonoluminescence.
In this chapter we have abused qualitative description. In the next one we will
compensate it by the scientific point of view.
5.1.2 Different Ways to the Unachieved Point
The nature of SL is intriguing, especially considering the fact that it was

undiscovered for more than 80 years. Here we suppose that the verb “to discover”
means “to explain main issues by conclusive theory”: there had been no indisput-
able explanation of light emission during cavitation found yet. We did not even
pretend to make a complete review of theoretical explanations of SL in this chapter;
there were many reviews published only in XXI century [2, 5–10]. All that we want
is to show main ways to the desirable point—a complete theory of
sonoluminescence.
The light of SL is very interesting by itself. SL spectra contain both bands and
continuum: see Fig. 5.11 from [10] and Fig. 5.12 from [11]. There are many reasons
for continuous spectra to be shown (see Sect. 5.3 below), but for any of them some
excitation of media is needed.
What kind of excitation may produce light? The first reason reads: the temper-
ature. Gas inside a bubble can be heated due to compression; bubble dynamics can
be described by the well-known Rayleigh–Plesset equation:
Fig. 5.11 Spectrum of MBSL [10]; argon saturated dodecane, 4 C. Bands correspond to Swan
series of C2
Fig. 5.12 Spectrum of SBSL [11]; water, 22 C
€ þ 3 R ¼ Δp ;
RR ð5:1Þ
2 ρ
where R is the bubble radius, and ρ is the density of the liquid, Δp—the difference
of pressures inside the bubble and in the liquid; the latter one contains ultrasonic
pressure. Equation (5.1) can be obtained, for instance, from the continuity equation
and the Navier–Stokes equation for spherically symmetric bubble.
To use the Eq. (5.1) we must compose a term for the gas (and vapor) pressure
inside the bubble. In simple simulations three assumptions are used: (1) constant
mass of gas, (2) saturation of vapor, corresponding to the constant temperature of
liquid, and (3) total adiabatic conditions. Enormous temperatures of gas ~104–5 K
can be obtained by such a way; in fact, the answer depends on the initial pressure of
gas, which is uncertain. Of course, for extraordinary conditions—for rapidly heated
bubble—all three assumptions can be easily targeted, especially (2), and instead of
the condition ps ¼ const a more complicated method can be used: the mass of vapor
can be calculated through the evaporation flux and the condensation one (according
to the Hertz–Knudsen formulae). This—molecular kinetic theory (MKT)—
approach is preferred for these spatial and temporal scales, but even in MKT-
approach there are many open questions: it was shown in [12] that usually the
Hertz–Knudsen relation is not valid. We agree with authors of [9]: actually, all
significant processes for SL are out of macroscopic, equilibrium, hydrodynamic
level of theoretical description. Probably, the most difficult part of SL theory is to
find proper starting point for it.
Both the Rayleigh–Plesset equation and common sense predicts that the speed of
walls of a collapsing bubble can exceed the speed of sound; that is, shock wave may
be generated in the vapor phase. Focusing in the center of the bubble, such wave
may produce incredible increase in temperature [13].
Another way to explain light emission from cavitation is calling up electricity.
Probably, historically the first work with electrical theory of SL was [14]: Frenkel
supposed that SL is the light emission from the electric discharges in bubbles;
electric charges are collected on the bubble’s wall due to balloelectric effect
(ionization of liquid at pulverization). On qualitative level, this hypothesis was
discussed previously in [15, 16]. It is more interesting that in his monumental book
Frenkel described, as a matter of fact, luminescence of liquid on ultrasonic wave-
guide: charges are separated on the waveguide (electric double layer), and ultra-
sonic oscillations provide electric field of up to ~107 V/cm between these two layers
of charges [17]. In case when the gas-layer thickness exceeds the electron mean
free-path length (~105 cm at atmospheric pressures), microdischarges would
occur in the gas layer. Frenkel did not know any confirmation of existing of such
discharges (it was only a hypothesis for him), but we can see corresponding pattern
of luminescence in Fig. 5.5. Of course, this is an indirect indicator, so as the damage
pattern of the waveguide (Figs. 5.8–5.10), which is similar to the damage caused by
an electrical discharge. Strictly, even direct electrical effects from [18] verify only
the existence of the electric double layer on a waveguide and its dynamics during
SL, but not the electrical discharges themselves.
In conclusion, we must say that the dominant description of SL is the “shock-
wave theory” now, despite the fact that this theory cannot explain every issue of
SL. One of the reasons behind this situation is, in our opinion, the following:
“shock-wave theories” are more exciting and are more promising than “boring”
electrical theories. But the final answer to the question about the nature of SL is still
far away.
5.2 Non-Ultrasonic Mechanoluminescence of Liquids
There are many ways to produce luminescence of substances. Possibly, some of

them are relative to sonoluminescence?
5.2.1 Mechanical Effects: Triboelectricity

and Triboluminescence
The fact that friction causes electrization was known in ancient Greece. Thales of
Miletus observed that amber rubbed with fur attracts lightweight objects; this
phenomenon was the first manifestation of electricity even from the terminological
point of view: the term “electricity” originated from Greek “ηλεkτρoν” (amber).
Despite of such a long history, physics of triboelectricity is still a disputable
question [19]. The main issue of theory remains open: what sort of charged particles
produces triboelectricity—electrons, ions or both? Moreover, even experimental
facts are unclear: there exist different reports even about the sign of charges
produced by friction of the same materials [19]. Thus, sonoluminescence is in the
noble company; it is rather difficult to explain (i.e., to provide a complete theory of)
the light emission from cavitation bubble through electrical effects when even such
simple electrization effects have no complete theory.
The closest relative of triboelectricity is electrokinetics: a number of phenomena
connected with charged solid surfaces and tangential fluid motion along
(or through) them [20]. Flow electrification in pipelines is the manifestation both
of triboelectricity and of electrokinetics.
Both friction and electricity can cause luminescence. In general, luminescence is
the radiative transition—a deexcitation after excitation of some kind:
• Photoluminescence—due to external irradiation.
• Mechanoluminescence—due to various mechanical effects: triboluminescence
caused by friction, fractoluminescence caused by fractioning, acoustolumi-
nescence (or sonoluminescence) caused by acoustic waves.
• Electroluminescence—due to electrical effects, especially due to external elec-
tric field.
• Chemiluminescence—due to chemical reactions, etc.; for example, some authors
distinguish radioluminescence (scintillation)—light emission at external ioniz-
ing irradiation.
Thus, if we are interested only in the mechanism of light emission, we can
examine simpler yet relative processes. If sonoluminescence is complicated by
many factors, then other kinds of mechanoluminescence might possibly provide
clearer results. In other words, what are the mechanisms of luminescence of liquids
in common?
5.2.2 Another Way of Luminescence: Hydroluminescence
We can produce luminescence of liquid not only by ultrasonic wave irradiation.

Another way to create a light emission is the propulsion of liquid through a narrow
channel; we will refer to a luminescence from such object, i.e. light emission of
liquids in absence of ultrasound, as hydroluminescence (HL, after [21]).
Hydroluminescence has been studied in many works (some of them are [21–24]);
sometimes HL may be accompanied by strange effects: Koldamasov reported about
gamma radiation at energy ~0.3 MeV [25].
Here we present our results [26, 27].
Channel configuration is shown on Fig. 5.13 and the whole experimental setup is
shown on Fig. 5.14; we used I-40A industrial oil in our experiments. Under
pressures ~30–40 atm. at the entrance into the channel the oil emits bright blue
light—much brighter than multi-bubble sonoluminescence; this radiation can be
registered by conventional photo and video cameras.
Typical luminescence pattern is presented on Fig. 5.15 (in external light) and on
Fig. 5.16 (in darkness; to prove that it is own radiation, not a reflection of an
external light). As one can see on these pictures, light emission is produced in the
narrowest part of throttle—this is a common case but we observed the important
exception (shown on Fig. 5.17).
Fig. 5.13 Narrow channel

(throttle); the narrowest part
is the source of
luminescence
Fig. 5.14 3D scheme of

experimental setup: 1—
reservoir of the working
liquid, 2—frequency
converter, 3—three-phase
asynchronous motor, 4—
reduction gear, 5—narrow
channel (presented on
Fig. 5.13), 6—pressure-
gauge, 7—gear pump
Fig. 5.15 Hydroluminescence of oil in narrow channel. Liquid flows from the right to the left
Fig. 5.16 Light emission from the throat of the throttle (liquid flows from the right to the left)
The regime presented on Fig. 5.17 can be named—by analogy with SL—as “a
single-bubble hydroluminescence.” It is intriguing that there is no other light
emission (in the throat). Therefore this light is not a reflection: sometimes, really,
it is hard to determine whether this light emission is the intrinsic light of substance
or not.
Why this bubble emits light, i.e., what kind of energy is converted into the light
there? Can we assume—in view that there are no other sources of light in the
throttle—that luminescence has a long decay, so long that this separated bubble
emits light without any energy feeding at the exit of channel, only due to the energy
which was reserved previously by the flowing substance (liquid and gas) in the
throat beforehand?
Fig. 5.17 “Single-bubble hydroluminescence”; a very rare regime. Liquid still flows from the
right to the left, so this luminous bubble nucleates at the exit of the channel
Fig. 5.18 Fast Fourier transform of the brightness of hydroluminescence for typical regime (such
as on Fig. 5.15). There are two distinctive peaks here: at 26 Hz (frequency of hydroluminescence
pulsations) and at 100 Hz (frequency of external light pulsations)
Fast Fourier transform of the brightness of typical pattern of hydroluminescence

(intensity of blue color after RGB decomposition of video with 240 frames per
second) is presented on Fig. 5.18. The “test” frequency of 100 Hz (induced by
Fig. 5.19 Gamma spectrum of hydroluminescence (1) and background (2): there is no difference
pulsation of external light; the electrical network frequency in Russia is 50 Hz)

shows that our method is correct, and frequency of 26 Hz corresponds with the
pulsations in gear pump. The fact that we can distinguish these pulsations means
that:
• The length of one pulsation is less than 0.04 s
• There is no long inertia in the source of light emission—the whole emission
cycle, including excitation of media and relaxation processes, lasts no longer
than ~102 s.
We obtain the same result for the characteristic frequency with various video
cameras at various frequencies of frames (from 100 to 480 frames per second).
As for gamma radiation (according to [25]), we tested emission from channel
with scintillation spectrometer: there is no any radiation exceeding the background,
see Fig. 5.19.
Thus, there are no extraordinary characteristics of luminescence concerning
emission decay (with time delay of recombination caused by the ion hydration as
a possible reason of such decay) or high-energy radiation. However, the main
question remains open: what is the hydroluminescence?
5.2.3 Lightened Version of Sonoluminescence?
It is impossible to avoid comparison between sonoluminescence and hydrolumi-

nescence. Both of them:
• are luminescence of liquids with cavitation bubbles under mechanical effects:

ultrasonic pressure for SL and friction in channel for HL;
• have the most probable source of light in the gas phase: besides other reasons,
“single-bubble regimes”—which exists both in SL and in HL—give clear
confirmation to this statement.
In [24] authors state that all significant characteristics of sonoluminescence and
of hydroluminescence are the same:
• Temperature dependence of the intensity of light emission.
• Dependence of luminescence on dissolved salts.
• Dependence of luminescence on intensity of mechanical effect (on the speed
of flow).
• Distribution of the magnitudes of light impulses.
Note that in previous studies the electrification of media was observed [21, 22],
i.e., electrokinetic effects play crucial role in these processes. In [28] authors
suppose that hydroluminescence is the light emission in photorecombination
(or photoattachment) acts.
In further chapters we will establish whether electrical effects cause light
emission or hydroluminescence has its origin in absolutely different physical
processes.
5.3 Spectroscopy Is the Clue
Every person has his/her own unique fingerprints. Similarly, every atom or mole-
cule has its own unique spectrum of electromagnetic radiation. It is easy to measure
the emission spectrum, but its interpretation is a hard and delicate work.
5.3.1 Method of Spectroscopy
Spectrum is a function of intensity of electromagnetic radiation in relation to a

wavelength (or frequency, or energy of quants). Fortunately, atoms and molecules
have lines in optic range of electromagnetic radiation (according to various defini-
tions, optic range corresponds to wavelengths from ~100 nm up to ~1 mm; optic
range is wider than visible range—from 400 to 750 nm), thus simple spectrometer
schemes can be used to decompose light on a spectrum: with a diffraction grating or
with an optical prism.
Spectrum of electromagnetic radiation may contain information about temper-
ature and pressure of the medium, number of charged particles, etc. [29, 30]. Actu-
ally, today spectroscopy became an independent area of science. Many methods
were advanced to investigate various objects of different physical nature, but
plasma is an undisputable leader in application of spectroscopy. Spectral diagnos-

tics of arc plasma (object with temperature ~104 K or higher) is a well-developed
method because of a simple reason: it is the only way to measure parameters of such
kind of plasma (probes can be used only in cold gas discharges). We have similar
problem with sonoluminescence: the only information about conditions of the light
emission is the radiation itself.
There is another situation for hydroluminescence. Many works demonstrate the
electrical effects during this kind of light emission. However, we are interested in
the following question: how does electricity affect the light emission? Is the light
emission a consequence of electrical effects or, for example, this kind of radiation
has the thermal nature and electricity is the side effect issue? To answer this
question we consider the basic kinds of spectra.
5.3.2 Electron Transition and Relative Spectra
In common, electromagnetic waves (including optical quants) can be emitted by

three different ways:
(a) At bound—bound electron transition: Electrons remain in bound state inside
atoms or molecules. This type of emission leads to linear spectrum: one
electron transition leads to one optical line in spectrum.
(b) At free—bound electron transition: Free electrons are grasped by atoms, and
their excess energy is let out as optical quants (so-called photorecombination).
This kind of emission leads to continuous spectrum.
(c) At free—free electron transition: Electrons are decelerated by ions (or even by
neutral atoms or molecules due to polarization), and their bremsstrahlung
provides continuous, non-fractured spectrum.
For atomic plasma, bound–bound transitions are the main source of information
about its parameters.
Let the rate of optical transition from the given state m to the lower state n of
atom is Amn [s1], and number density of atoms in upper state is Nm [cm3]. Thus
the energy
hc
ε¼ Amn N m ð5:2Þ
λ
is emitted out from the unit volume of gas (or plasma) on the given wavelength λ
per second; here h is Planck’s constant, and c is the speed of light.
Since we believe in Boltzmann’s distribution, the number density of atoms in the
state m with the energy Em correlates with the total number density N0:

gm Em
Nm ¼ N0 exp ; ð5:3Þ
Z ðT Þ kT
where gm is degeneracy (statistical weight), k is the Boltzmann constant, T is the

temperature, and Z(T ) is the partition function:
X
Ej
Z ðT Þ ¼ gj exp : ð5:4Þ
j
kT
In the experiment we measure not exactly the energy (5.2) but the irradiance
(energy per second per square unit per space angle unit at the distant point out from
the emitting object):

hc gmn Emn
I ¼ φ Amn N 0 exp ; ð5:5Þ
λ Z ðT Þ kT
Factor φ contains all additional information, including self-absorption, geomet-

ric parameters etc. By taking Amn, gmn, and Emn from literature (we use NIST
reference data) and measuring I in an experiment (by calibrated spectrometer)
one can obtain temperature T from Eq. (5.5) if one also knows φ and N0. But this
is a difficult task, especially concerning φ, so the method of irradiance ratio can be
used. Assuming φ ¼ const 6¼ f ðλÞ we can measure intensities Eq. (5.5) for two
lines, and corresponding ratio contains only well-defined parameters:

I 1 λ 2 A1 g 1 E2 E1
¼ exp : ð5:6Þ
I 2 λ 1 A2 g 2 kT
Here we used simplification for notation (indexes “mn” are omitted). For the first
glance, we may easily obtain temperature T from Eq. (5.6)—using the information
only about two lines. Theoretically, this is possible, but practically the temperature
obtained from one pair of lines would differ from the temperature obtained from
another pair of lines (difference can be up to ~10 % or even larger for cold plasma at
T < 104 K). There are many reasons of this discrepancy:
• Methodical problem: the Boltzmann distribution (5.3) is not carried out strictly.
• Methodical problem: for non-opaque plasma φ depends on λ.
• Methodical problem: usually, temperature depends on coordinates, so spectrom-
eter collects radiation from the sites with different temperatures; thus Radon’s or
Abel’s transforms are needed.
• Calculation and instrumentation problem: often difference between two upper
energy levels is very small; for example, for bright lines of Arat λ ¼ 650
950
I 1 λ1 g2 A2
nm energies E1, 2 ¼ 12:907 13:480 eV, and logarithm ln I 2 λ2 g1 A1 is very
sensitive to its argument, i.e., to the values of transition probabilities A1,2 and
measured irradiances I1,2. In other words, inaccuracy of input parameters would
make strong impact on inaccuracy of temperature.
For these reasons, another implementation of irradiance ratio method is used in
practice. Constructing variables
Fig. 5.20 The tangent of the angle of slope of this line to y-axis is the temperature of the electric
arc in argon; in ideal case—without errors (methodological, instrumental, etc.)—all dots would fit
that line

E2 E1 I 1 λ 1 A2 g 2
xi ¼ and yi ¼ ln ; ð5:7Þ
k i I 2 λ 2 A1 g 1 i
(where “i” denotes all available pairs of lines) we have a dependence
1
xi ¼ yi ; ð5:8Þ
T
and the temperature can be obtained as a coefficient of linear approximation.

Figure 5.20 illustrates the dependence yi(xi) for arc plasma in argon: as one can
see, dispersion is significant. Certainly, linear regression cannot correct methodo-
logical stretches, but it helps to find the mean value of temperature and its
inaccuracy.
Every emitted and measured line has finite width. There are many reasons for
broadening:
• Finite lifetime of excited levels
• Collisions
• Doppler’s effect
• Apparatus effect
The last one is usually described through the “full width half maximum”
(FWHM) parameter, which represents the optical resolution of spectrometer. For
our AvaSpec-2048 spectrometer FWHM is 2.4 nm—much wider than any “phys-
ical” broadening of emitted line.
Calculated spectrum of arc plasma (considering apparatus broadening in Gauss-
ian form) is presented in Fig. 5.21.
Fig. 5.21 Spectra of arc plasma in argon: 1—calculated, 2—measured with AvaSpec-2048
spectrometer
Bound–bound transitions in molecules lead to more complicated spectra. In

comparison to atoms, there are two additional transitions here: between vibrational
levels and between rotational ones. We may expect that Boltzmann’s distribution is
suitable both for vibrational energy and for rotational energy, but in general case
these distributions have different temperatures Tvib and Trot, and both these tem-
peratures differ from the temperature of gas Tgas and from the temperature of
electrons Te. In non-equilibrium gas discharges, where electrons gain energy from
electric field and give it back to molecules in collisions, usually
T e >> T vib > T rot T gas : ð5:9Þ
But there are no absolute rules here. Temperatures depend on kinetic of occu-
pying of energetic levels, and relations between different temperatures may con-
siderably differ from Eq. (5.9).
Spectrum of a molecule can be obtained from Eq. (5.2). Boltzmann’s distribu-
tion function allows not only electron energy Eel, but also vibrational energy for
vibrational level v
Evib ðvÞ ¼ ωe ðv þ 1=2Þ ωe xe ðv þ 1=2Þ2 ; ð5:10Þ
(ωe and ωexe are vibrational constants; second term is the correction on
anharmonicity) and rotational energy for rotational level J
Erot ðJ Þ ¼ ðBe αe ðv þ 1=2ÞÞJ ðJ þ 1Þ þ ðDe þ βe ðv þ 1=2ÞÞJ 2 ðJ þ 1Þ2 ; ð5:11Þ

(Be, De, αe, and βe are constants). Thus, number density of molecules in the given
state is

gel gvib grot Eel Evib Erot
N ¼ N0 exp exp exp : ð5:12Þ
Zel Zvib Zrot kT el kT vib kT rot
Practically, probability of the vibrational-rotational transition can be represented

independently (i.e., as a production of factors) for the probability of vibrational
transition (band strength, Franck–Condon factor qv0 v00 ) and for the probability of
rotational transition (line strength, H€onl–London factor SJ0 J 00 ) [31].
Theoretical methods of molecular spectroscopy allow us to reconstruct spectra
of various interesting objects, such as corona discharge (Figs. 5.22 and 5.23) in the
air; the calculated spectrum consists of bands of N2 and OH.
Free–free transition provides continuous radiation spectrum. Spectral intensity
of radiation is
Z1
I ν dν ¼ hνN i N e f ðve Þvedσ ν dve : ð5:13Þ
v0e
Here Ni and Ne are number densities of ions (or atoms) and electrons corre-
spondingly, f(ve) is the electron distribution function on velocities and dσ ν—cross
section of emission; v0e ¼ ð2hν=me Þ1=2 is the minimal velocity of electron to emit
quant hν.
Fig. 5.22 Single-electrode discharge (corona discharge)

Fig. 5.23 Spectrum of corona discharge (Fig. 5.22): 1—calculated, 2—measured with AvaSpec-
2048 spectrometer
According to Kramers, for ions with charge Ze
Z2
dσ ν ¼ C1 dν: ð5:14Þ
v2e ν
Using Maxwellian distribution function for electrons

mv2
f ðve Þ ¼ C2 v2e exp e ; ð5:15Þ
kT e
we obtain bremsstrahlung spectrum

Z2 N i N e hν
I ν dν ¼ C3 pffiffiffiffiffi exp dν; ð5:16Þ
Te kT e
or

Z2 N i N e c hc
I λ dλ ¼ C3 pffiffiffiffiffi 2 exp dλ: ð5:17Þ
Te λ λkT e
Thus, spectrum Iλ has maximum at
hc
λm ¼ : ð5:18Þ
2kT e
Fig. 5.24 Recombination and bremsstrahlung spectra (in arbitrary units, but in the same scale) for
hydrogen-like atom; Z ¼ 10, Te ¼ 104 K. Bremsstrahlung part dominates in far infrared range
Typical spectrum for free–free transition is presented on Fig. 5.24. Note that its
‘hump’ is wider than the one for the black-body spectrum, due to λ2 in denominator
in Eq. (5.17) instead of λ5 in usual Planck’s law.
As free–free transitions, free–bound transitions lead to continuous spectrum. For
hydrogen-like atoms this problem can be considered by analogy. Namely, if
electron is captured on level n with energy En,
me Z 2 e4
En ¼ ð5:19Þ
32π 2 ε20 h2 n2
(ε0 ¼ 8.85 1012 F/m, h ¼ h=2π, me is the electron’s mass), cross-section is
Z4
dσ ν ¼ C1 dν: ð5:20Þ
v2e νn3
Using Eq. (5.13), we get irradiance as

Z2 N i N e 2xn hν
I νn dν ¼ C3 pffiffiffiffiffi expðxn Þexp dν; ð5:21Þ
Te n kT e
where xn ¼ En =kT e . Formula (5.21) corresponds to the capture on level n; to obtain

a complete spectrum we must add up Eq. (5.21) for all n at En hν. The
corresponding spectrum is presented in Fig. 5.24 for Iλ.
In the next chapter we will see, what sort of spectra is suitable for
hydroluminescence.
5.4 The Electric Nature of Hydroluminescence
As we stated above, spectroscopy may provide the answer to questions regarding

the nature of light emission and, consequently, regarding the conditions in the given
medium. Luminescence of dielectric fluid in a narrow channel has a complicated
spectrum, but all parts of this spectrum can be explained from a single point
of view.
5.4.1 Hydroluminescence Spectrum
In our experiments we observed the luminescence of the I-40A oil (lubricating and
cooling liquid). Its composition [32]:
• Alkanes: 4,6-dimetildodecane, pentadecane, hexadecane, heptadecane,
octadecane, nonadecane, 2,6,10,14-tetramethylhexadecane, 10-methylno-
nadecane, 9-methylnonadecane, heneicosane, tricosane, octacosane.
• Alkenes: 1-heptadecene, 1-nonadecene, 1-docosene, 11-tricosene, cis-12-
pentacosene, 1-hexacosene, 9-hexacosene.
• Cycloalkanes: 1,2-dimethyl-3-pentyl-4-propylcyclohexane, ethylcyclooc-
tadecane, 1-(1,5-dimethylhexyl)-4-(4-methylpentyl)cyclohexane, 1,7,11-
trimethyl-4-(1-methylethyl)cyclotetradecane.
• Alcohols and phenols: 2,4,6-tri(1-methylethyl)phenol, 4,6-dibutyl-m-cresol,
2-hexyl-1-decanol, 2-octadecylhydroxyethanol, 1-tricosanol, 1-hexacosanol.
• Ester: 11,13-dimethyl-12-tetradecene-1-olacetate.
• Sulfur-containing organic compounds: butylheptadecylic sulfuric-acid ester,
butyloctadecylic sulfuric-acid ester.
• Halogen-containing organic compounds: 1-chlornonadecane,
1-iodinehexadecane, 1-bromine-11-iodinehendecane, 1-iodineoctadecane,
1-brominedocosane, 1-chlorheptacosane.
• Nitrogen-containing organic compound: 6-nitrohendecyne-5.
The experimental setup has been described in Chap. 2. AvaSpec-2048 diffrac-
tion spectrometer was used to measure the spectrum; its main parameters:
• Wavelength range: 180–1100 nm.
• Resolution (FWHM): 2.4 nm.
The most interesting thing: more than one spectrum of hydroluminescence had
been registered. We have seen different spectra from the same substance (I-40A) in
our experiments (Fig. 5.25). It was possible to observe different spectra even during
the course of the same experiment while observing different parts of the tube.
“The most UV” spectrum (marked black on Fig. 5.25) contains strongly
expressed bands. This is 2þ-series of N2 ( C3 Π B3 Π ; see also Fig. 5.23); in
addition, several peaks of 1-series of N2þ ( B2 Σ þ 2 þ
u X Σ g ) are also there, the
Fig. 5.25 Hydroluminescence spectra
brightest of them are on 391 nm (of (0,0) vibrational transition) and on 428 nm
(of (0,1) vibrational transition). The source of these bands was the air dissolved in
the working liquid (oil was used without outgassing).
For some, spectrum of N2 can be a proof of the electrical nature of radiation—
not the thermal one—by itself for the following reasons:
• The temperature of 99 %-dissociation of N2 is about 104 K, thus the gas was not
too hot.
• 2þ-series of N2 is the important feature of the radiation in air-discharge plasma,
such as corona discharge (see Sect. 5.3) or even lightning.
• Absence of atomic lines in the spectra.
However, the presence of N2 emission bands provides not only qualitative
description of the substance’s state. Series of emission lines is convenient for the
diagnostics of parameters of the luminescent medium. As it was described in the
Sect. 5.3, one can determine vibrational and rotational temperatures of N2 [26]: T vib
¼ 4000 K and T rot ¼ 300 K with the ~10 % margin of error. Note that the direct
simulation of the whole spectrum was impossible because of the attenuation of
radiation in the liquid and in the wall of the tube (made of organic glass); moreover,
the band spectrum “lays” on the continuous one. Thus, only closely spaced adjacent
lines were expedient to compare under the assumption that the absorption coeffi-
cient remains approximately constant within short spectral intervals.
The rotational temperature of gas provides an estimation of the transitional one,
i.e., the gas temperature. Thus, the gas inside the bubbles is cold and the lumines-
cent medium is strictly non-equilibrium: it is rather similar to the plasma of the
glow discharge than the thermal plasma.
5.4.2 Hydroluminescence as Triboluminescence
Thus, one part of the light emission from the narrow channel is the luminescence of
gas (nitrogen from the air within the oil). But there is another part of lumines-
cence—the continuous spectrum. It is a more intriguing component of emission
than bands, and it is more difficult to interpret.
Of course, these continuous spectra are not the thermal radiation of liquid: they
are “too narrow” for Planck’s curves; consequently, they are “too narrow” for
bremsstrahlung or recombination emission too. Note also that it is difficult to expect
blackbody radiation from the sub-millimeter bubble: the layer must be optically
thick to produce such kind of radiation. It is also hard to believe that liquid can
produce thermal radiation with λmax 400 nm, i.e., (according to the Wien dis-
placement law) at temperature T 7500 K.
The most possible reason for continuous spectrum is the luminescence of oil: it is
well known that organic oils are luminophores. Organic luminophore emits light
due to the transition in a single molecule; this molecule can be excited in many
ways. The simplest method to excite the molecule is UV radiation; we used a
UV-emitter with lines in ultra-violet and visible ranges at 253, 265, 313, 365,405,
436, 546, and 578 nm. One can expect that the spectrum of photoluminescence
(induced by UV emitter) and continuum of hydroluminescence spectrum will prove
to be similar.
The luminescence of I-40A oil is presented in Fig. 5.26 (as illustration); in the
experiment the UV-source was placed over the quartz cell and emitted light was
detected through a lateral surface with the AvaSpec-2048 spectrometer. Therefore,
the photo- and hydroluminescence spectra were measured using the same method:
according to [33], the main reasons of difference between photo- and tribolumi-
nescence spectra are different conditions of spectra collecting.
Spectra of hydro- and photoluminescence are presented on Figs. 5.27, 5.28, 5.29:
we can see an almost ideal coincidence. So what does it mean? We can obtain all
parts of hydroluminescence spectra in other ways:
• The band component: by electrical discharge in the air,
• The continuous component: by photoluminescence of oil.
There are no other types of light emission of the liquid in narrow channel. Thus,
there are no special conditions in throttle and no extraordinary processes here (for
instance, chemical reactions due to high temperatures caused by shock waves in
collapsing bubbles).
All parts of the HL spectrum can be induced by separation of charges on the
liquid–solid interface. After the separation, strong electrical field (~104–105 V/cm)
[34] causes electrical breakdown in gas phase; this breakdown leads to the appear-
ance of plasma with excited atoms and even ions—light emission from them
provides the band spectrum. Electrical field, charged particles, and UV photoemis-
sion from the discharges (in bubbles) is a classic set for electroluminescence
(or photoelectroluminescence) as it was described in [35]. This kind of light
emission provides continuous part of the spectrum.
All components of hydroluminescence have electric nature.
Fig. 5.26 Photolumine-

scence of I-40A oil
Fig. 5.27 1—spectrum of hydroluminescence, 2—spectrum of photoluminescence (spectrum #1)

5.5 Sonoluminescence and Hydroluminescence: Relatives

or Strangers?
Sonoluminescence (SL) is a light emission from cavitation bubble in a bulk of a

liquid. There are many theories and ideas about the nature of this process; most of
them can be divided in two groups:
• “Thermal” theories explain luminescence with an increase in temperature; these

theories can be subdivided on “static” models (temperature increases due to
quasi-adiabatic compression of bubble) and “dynamic” ones (temperature
increases due to shock waves inside the collapsing bubble).
• “Electric” theories explain luminescence by various electrical effects; electrical
charges appear on interfacial surfaces.
Actually, there is no indisputable theory of sonoluminescence, but “thermal”
theories dominate now.
Hydroluminescence (HL) is a light emission that occurs during a flow of a liquid
in a narrow channel. There were many reports about electrical effects in such kind
of a flow motion; our research (presented in this chapter) show that light emission
can be completely explained by electricity: both the band part of the spectrum and
the continuous one evidence the triboluminescence caused by electrization. We do
not see any sign of another source of radiation, we do not see any trace of high
temperature inside a bubble; thus, Occam would possibly cut “thermal” theories out
from the hydroluminescence phenomena by his razor forever. We prefer to con-
clude that hydroluminescence can be explained by electrical effects at high prob-
ability (we might have missed some effects which could turn over our notion about
this type of luminescence).
How can the nature of hydroluminescence be connected with the nature of
sonoluminescence? Both phenomena are the examples of mechanoluminescence
of liquids; in early works any type of light emission of liquid was referred as SL
despite the presence or the absence of ultrasound irradiation. However, correlation
between SL and HL is not so clear.
HL and SL are relatives because:
• Gas discharge inside a bubble is an important part of HL; thus the gas phase
plays a significant role in light emission for HL, as it was established for SL.
• Moreover, even a single bubble can luminescent in such a way (see Fig. 5.17); so
we see a complete analogy of the regimes of SL and HL.
HL and SL are strangers because:
• For HL, there exists a clear mechanism of charge separation on the liquid–solid
interface during the flow motion in narrow channel (triboelectricity); there are
no such clear processes for charge separation during SL (except for multi-bubble
SL; Frenkel predicted the existence of gas discharges in proximity of a
waveguide).
• Spectra of SL qualitatively differ from spectra of HL.
We have rendered habitable territory close to the unknown one, but the way out
from this foothold to the wilderness is still questionable. Anyway, now we have
complete physical picture for an interesting phenomenon—the hydroluminescence.
Acknowledgments We are grateful to M.I. Vlasova for her contribution to the preparation of this
chapter for publication.
References
1. Frenzel, H., & Schultes, H. (1934). Zeitschrift f€ur Physikalische Chemie, 27, 421.
2. Margulis, M. A. (2000). Physics-Uspekhi, 43, 259.
3. Gaitan, D. F., Crum, L. A., Church, C. C., & Roy, R.A. (1992). The Journal of the Acoustical
Society of America, 91, 3166.
4. Skokov, V. N., Reshetnikov, A. V., Vinogradov, A. V., & Koverda V. P. (2007). Acoustical
Physics, 53, 136.
5. Borisenok, V. A. (2015). Acoustical Physics, 61, 308.
6. Brenner, M. P., Hilgenfeldt, S., & Lohse, D. (2002). Reviews of Modern Physics, 74, 425.
7. Hammer, D., & Frommhold, L. (2001). Journal of Modern Optics, 48, 239.
8. Margulis, M. A., & Margulis, I. M. (2002). Ultrasonics Sonochemistry, 9, 1.
9. Putterman, S. J., & Weninger, K. R. (2000). Annual Review of Fluid Mechanics, 32, 445.
10. Suslick, K. S., & Flannigan, D. J. (2008). Annual Review of Physical Chemistry, 59, 659.
11. Hiller, R., Putterman, S. J., & Barber, B. P. (1992). Physical Review Letters, 69, 1182.
12. Gerasimov, D. N., & Yurin, E. I. (2015). High Temperature, 53, 502.
13. Suslick, K. S., Eddingsaas N. C., Flannigan, D. J., Hopkins, S. D., & Xu, H. (2011). Ultra-
sonics Sonochemistry, 18, 842.
14. Frenkel, L. I. (1940). Acta Physicochimica URSS, 12, 317.
15. Harvey, E. N. (1939). Journal of the American Chemical Society, 61, 2392.
16. Levshin, V. L., & Rzhevkin, S. N. (1937). Comptes rendus de l’Académie des sciences de
l’URSS, 16, 399.
17. Frenkel, J. I. (1946). Kinetic theory of liquids. Oxford: Oxford University Press.
18. Birykov, D. A., Gerasimov, D. N., & Sinkevich, O. A. (2014). Technical Physics Letters, 40,
138.
19. Galembeck, F., Burgo, T. A. L., Balestrin, L. B. S., Gouveia, R. F., Silva, C. A., &
Galembeck, A. (2014). RSC Advances, 4, 64280.
20. Delgado, A. V., Gonzalez-Caballero, F., Hunter, R. J., Koopal, L. K., & Lyklema, J. (2005).
Pure and Applied Chemistry, 77, 1753.
21. Margulis, M. A., & Pil’gunov, V. N. (2009). Russian Journal of Physical Chemistry A, 83,
1414.
22. Gertsenshtein, S. Y., & Monakhov, A. A. (2009). Fluid Dynamics, 44, 430.
23. Peterson, F. B., & Anderson, T. P. (1967). Physics of Fluids, 10, 874.
24. Verbanov, V. S., Margulis, M. A., Demin, S. V., Korneev, Y. A., Klimenko, B. N., Nikitin, Y.
B., & Pogodaev, V. I. (1990). Zhurnal Fizicheskoi Khimii, 64, 3357.
25. Koldamasov, A. I. (1991). Zhurnal Tekhnicheskoi Fiziki, 61, 188.
26. Biryukov, D. A., Gerasimov, D. N., & Sinkevich, O. A. (2012). Technical Physics Letters,
38, 81.
27. Biryukov, D. A., Vlasova, M. I., Gerasimov, D. N., & Sinkevich, O. A. (2013). Optics and
Spectroscopy, 114, 704.
28. Polyanskii, V. A., & Pankrat’eva, I. L. (2005). Doklady Physics, 50, 397.
29. Ochkin, V. N. (2009). Spectroscopy of low temperature plasma. Weinheim: Wiley-VCH
Verlag GmbH & Co.
30. Raizer, Y. P. (2011). Gas discharge physics. Berlin, Heidelberg: Springer.
31. Kovacs, I. (1969). Rotational structure in the spectra of diatomic molecules. Budapest:
Akademiai Kiado.
32. Ermolaeva, N. V., & Golubkov, Y. V. (2011). Oborudovanie i tekhnologii dlya neftegazovogo
kompleksa, 3, 49.
33. Duignan, J. P., Oswald, I. D. H., Sage, I. C., Sweeting, L. M., Tanaka, K., Ishihara, T., Hirao,
K., & Bourhill, G. (2002). Journal of Luminescence, 97, 115.
34. Biryukov, D. A., Vlasova, M. I., Gerasimov, D. N., & Sinkevich, O. A. (2013). High
Temperature, 51, 566.
35. Curie, D. (1963). Luminescence in crystals (pp. 237–287). New York: Wiley.
Chapter 6
Functional Triboluminescent Nanophase
for Use in Advanced Structural Materials:
A Smart Premise with Molecular
and Electronic Definition
M. Jasim Uddin, David O. Olawale, Jin Yan, Justin Moore,

6.1 Introduction
The term “triboluminescence” is commonly used to refer to light emission from

certain materials when stressed or fractured [16, 21–23, 33, 35, 40, 52]. The
application of this type of luminescence (light emission) has occurred for centuries,
but under names only differing by their prefix: triboluminescence, mechanolumi-
nescence, or fractoluminescence, which result from applied friction, mechanical
actions, or the fracturing of solid materials, respectively. Luminescence in materials
or solids is usually caused by electricity, chemical reactions, stress on their crystal
structure, or subatomic motions, which then emit light based on their atomic
characteristics [9, 17, 40, 41]. It is believed that triboluminescent emission is
associated with an asymmetric crystal structure. Crystal bonds are broken along
planes with opposing charge, and when the bonds re-connect, light is emitted as the
charges pass through the separation created from the fracture [18, 40, 42].
According to literature estimates, 36 % of inorganics, 19 % of organics, 37 % of
M.J. Uddin (*) • J. Moore

Department of Chemistry, College of Science, University of Texas Rio Grande Valley,
Brownsville, TX 78520, USA
e-mail: mohammed.uddin@utrgv.edu
D.O. Olawale
Department of Industrial and Manufacturing Engineering, High-Performance Materials
Institute, FAMU-FSU College of Engineering, 2525 Pottsdamer Street, Tallahassee, FL 32310,
USA
Nanotechnology Patronas Inc., Tallahassee, FL 32310, USA
J. Yan • O.O.I. Okoli
Department of Industrial and Manufacturing Engineering, High-Performance Materials
Institute, FAMU-FSU College of Engineering, 2525 Pottsdamer Street, Tallahassee, FL 32310,
USA

126 M.J. Uddin et al.
Fig. 6.1 TL emission of triboluminescent crystals [17] © John Wiley & Sons, Inc.
aromatic compounds, 70 % of alkaloids, and perhaps 50 % of all crystalline mate-

rials are triboluminescent [16, 21–23, 40, 52] (Fig. 6.1).
Luminescence, which is light emission not caused by heating, can be distin-
guished from incandescence, which is light emission as a result of heating [51]. An
unusual example of gas excitation is found in triboluminescence (TL), which
translates as light from friction or fracture. In a miniaturized version of the effect
of lightning, gas excitation of nitrogen in the air produces a blue-white light. When
a high enough potential difference is created, electrons jump across the fractures in
the crystal colliding with nitrogen molecules in the air. These excited nitrogen
molecules produce positively charged nitrogen ions, which then emit blue light.
Some ultraviolet emission is also produced to fluoresce with a particularly intense
blue-light production. There are numerous common materials that produce tribo-
luminescence. Diamonds, for example, sometimes emit light during the polishing
process and when separating certain sticky surfaces, such as band-aid wrappers or
postal envelopes [1].
Hence, this chapter further elucidates the relationship between deformations
with applied force and contacting stress, and the photoemission behavior of differ-
ent phosphors materials using a TL measuring system. Stress determination and the
influence of experimental parameters such as applied frictional or external force on
mechanoluminescence related to structural applications and monitoring are also
discussed.
6 Functional Triboluminescent Nanophase for Use in Advanced Structural. . . 127
6.2 The Principle of Triboluminescence: An Electronic

Excitation of TL Materials
TL emission spectra of different triboluminescent materials contained the line

spectrum characteristic of an electrical discharge through air (C3πu ! C3πg bands,
or second positive group of N2 characteristic of lightning: the light emission from
nitrogen molecules; N2* ! N2 þ hν1) [21–23, 35]. Note that TL materials are
excited by either discharge by electron bombardment of the material or secondary
excitation by the UV emissions of the surrounding gas. Several triboluminescent
materials may even be excited without a discharge, probably by the defect recom-
bination characteristic of deformation luminescence [40]. The TL spectra of each
solid contained varying fractions of triboluminescent emissions from photolumi-
nescence (PL) and dinitrogen emission modified by absorption by the solid and
dinitrogen emission from static discharge to the apparatus [43]. Electrorheology
[15] is concerned with field enhanced dipole-dipole interaction in relation to the
mechanical stress required to achieve electronic charge separation. Less well
understood are the materials whose TL spectrum approximates to the PL spectrum
but where no nitrogen discharge spectrum is evident. Even more puzzling are
coumarin (and other large organic molecules) whose TL spectra contain features
not seen in their PL spectra. It is suggested that these features may arise through
changes in the Franck-Condon factors brought about by the high stresses existing at
the tips of growing cracks [53]. Figures 6.2 and 6.3 show TL emission spectra of
triboluminescent polymeric PL fiber core doped with menthyl-9-anthracenecar-
boxylate (MAC) and rhodamine 6G fiber core, respectively [35]. The blue line is
the emission spectrum and the red line is the energy it absorbs to produce TL
emission (green line). These complex molecules show multiple bands overlapping,
which result in a complex and broad spectrum [40, 41, 44].
In Fig. 6.2, it is shown that the peak triboluminescent emission wavelength of
MAC occurs at approximately 440 nm, which is closely matched to the peak of the
Fig. 6.2 The spectral overlap of the MAC triboluminescence (centered at ~450 nm) with the
absorption spectrum of the polymeric PL Fiber core. The PL emission spectrum from the fiber core
is shown for completion [35] © IOP Publishing
Fig. 6.3 The spectral overlap of the terbium TL (multiple bands at 480, 545 and 570 nm) with the
absorption spectrum of the rhodamine 6G Fiber core. The PL emission spectrum from the
rhodamine 6G fiber core is shown for completion [35] © IOP Publishing. The rhodamine 6G
was chosen since the triboluminescent emission spectrum of the terbium complex significantly
overlapped the absorption spectrum of rhodamine 6G.
absorption spectrum of the PL fiber at 435 nm. The PL core then re-emits the
absorbed energy with a peak emission wavelength of approximately 475 nm
[35]. The peak of the terbium TL occurred at 548 nm whereas the peak absorption
of rhodamine 6G occurred at 537 nm (Fig. 6.3). The terbium TL was efficiently
collected/absorbed by the rhodamine 6G, which would reemit this energy by
photoluminescing at 565 nm. At a low- to moderate-energy region, luminescent
pulses consist of two distinguishable peaks that can be assigned to a loading and
unloading sequence, respectively. Transient characteristics are found to be inde-
pendent of the external excitation. An increment of peak emission intensity with
higher impact energy and smaller tip diameter indicates a significant relationship of
TL to the geometry of contact. Intense emission deviating from the linearity and an
absence of the secondary peak are suggested to be the consequence of
fractoluminescence [20].
6.3 Spectroscopy of TL Materials
The solid-state PL lifetime has been determined for all samples and the implication
that these values have for observing pressure-induced (Franck-Condon) lumines-
cent spectral changes are discussed. In all cases where a significant difference was
observed between TL and PL spectra, either self-absorption of the TL emission or
fracture-induced perturbations of the local field have been evoked to explain the
difference (Figs. 6.4 and 6.5).
Figure 6.5 shows that the TL spectrum for samples 1–3 are identical (at least to
the resolution used) to the PL spectrum. This spectral agreement may not
be surprising since the emitting species in these materials is a trivalent lanthanide
(Laþ3). The intra f-shell transition, responsible for light emission, is relatively
Fig. 6.4 The PL emission

spectrum of 1 as collected
by the optical multichannel
analyser (A) and a high-
resolution (0.25 nm)
conventional
spectrofluorometer (B) [9]
© Elsevier Science
Fig. 6.5 The TL (data points) and PL (solid line) spectra of samples 1–3. The TL spectra of all
samples did not exhibit any spectral shift compared with the PL spectra [9] © Elsevier Science
insensitive to crystal field effects. Therefore, symmetry changes due to the creation
of fractured surfaces are expected to play a minor role, at best, in modifying the TL
spectrum relative to the solid-state PL emission spectrum [9]. During structural
fracture, the crack propagation velocity can be estimated. Assuming a crack tip
diameter of the order of a few nanometers [3, 9], the emitting species will be
subjected to significant pressures only for a few picoseconds. The pressure-induced
changes to Franck-Condon factors could alter the intensity and bandwidth of the
luminescence during the stress operation period. The TL crystal is comprised of
molecules with two different symmetries, which leads to the double exponential.
This explanation is consistent with reports that this compound may exist in various
polymorphic forms [3]. It should be noted that the slow component of the decay is
very similar to the solid-state PL lifetime reported by others [45]. It is evidenced
that the Mn:ZnS shows non-exponential decays [37].
Bourhill et al. [9] described that the TL spectra are sometimes less intense than
PL spectra at the short wavelength side of the peak emission wavelength. The
reason frequently given for this is that solid-state PL is to a first approximation a
surface effect, whereas TL can be generated within the bulk of a solid sample
(Figs. 6.2 and 6.3). Any resulting TL light has to propagate through the sample, and
thus suffer absorptive attenuation, to reach the detector. In other words, self-
absorption of the TL light can lead to spectral differences compared with the
solid-state PL spectrum. The large Stokes shift (between the absorption and
emission maxima) for 1–3 precludes spectral differences due to self-absorption,
consistent with the results shown in Fig. 6.5. The TL spectra in samples 9–12
display significant spectral shift compared with the PL spectrum (Fig. 6.6) although
the TL intensity is significantly unchanged. A solid-state PL excitation spectrum
(thus solid-state absorption) with a monitoring wavelength of 525 nm, highlighted
that the reduction in TL intensity is more than likely due to self-absorption
(Fig. 6.6). The normalized solid-state PL excitation (absorption) spectrum
(Fig. 6.6, squares) shows that solid-state absorption becomes significant for
Fig. 6.6 Normalized solid-state PL emission (line) and normalized solid-state PL excitation, thus
absorption, (data points) of the hexacelsian samples 9–12 [9] © Elsevier Science
wavelengths below 450 nm and is very strong for wavelengths below 400 nm. This
is consistent with the data presented in the literature [9], where the TL intensity
starts to deviate from that observed for the PL at approximately 480 nm, with the
difference between TL and PL intensity strongest at wavelengths around 380 nm.
6.4 Interaction of Impact and Electronic Transformation
Stress-induced TL is defined as a visible light emission. TL spectrum of ZnS:Mn

was reported as being similar to a characteristic emission of 4 T t ! 6 At energy
transition of luminescence centers created by doping manganese [19, 20]. The
bright TL of lanthanide based organic complexes (Lanthanide acetylacetonates
Ln(acac)3 H2O: Ln ¼ Ce, Pr, Gd, Eu, Tb) at solid state certainly exhibit the lines
that appear for all the crystals in a discharge in molecular nitrogen N2(C3 Пu – B3
Пg), except for the shortest wavelength line of 316 nm, which is observed in the TL
spectra of a few other materials [38]. The crystalline component of the TL is only
registered for ions Eu3þ and Tb3þ whose radiative levels are situated lower than the
triplet level of the acacT1 ligand. The Table 6.1 lists the relative TL intensity of
the nitrogen and solid-state crystalline salts of different lanthanides. It is clear
that the highest total TL intensity is characteristic of sulfates and hardly present
in nitrates [11].
Note: I s.st/I nitr are the ratios of intensities of the solid state and nitrogen
components; they are given in the table cells.
Figure 6.7 shows the TL spectra of some typical lanthanide crystalline materials
with molecular nitrogen excitation. It is evident that the radical increase in the
intensity of the nitrogen component in the TL materials in comparison with the
conventional TL that was observed in organic crystals is not characteristic of
inorganic salts of lanthanides [10, 11, 38]. The contribution of the different com-
ponents to the specific luminescence band varies in different experiments and
extensively depend on the exposure time and other experimental conditions. The
nitrogen component under atmospheric conditions plays an important role in this
excitation process (vide supra). The resolution power of the lines of this component
Table 6.1
Fig. 6.7 TL spectrum of

Tb2(SO4)3 8H2O [38]
© Springer International
Publishing AG
are defined by the nature of the lanthanide [38]. In the case of terbium salts
(Fig. 6.7), the spectral ranges of emission of nitrogen and the lanthanide ion are
well separated. It is possible to obtain the nitrogen component spectrum with high
resolution, not withstanding that its intensity can be approximately 10–100 times
less than the intensity of the solid state component (chloride, carbonate, terbium
acetylacetonate) [38].
6.5 TL Materials with Integrated Structural Health

Monitoring
6.5.1 Advanced Polymer Composites
A remarkable amount of research work has been done in last two decades in
applying of the TL phenomenon for structural and engineering application such
as damage detection of structural materials. A TL-based ubiquitous damage sensor
can detect material deformation within a structure such as aircraft wings or fuse-
lage, or a bridge. Sage and Geddes [36] developed an intrinsic structural health
monitoring (SHM) system with supporting of a triboluminescent material
connected via light-guiding fibers or layers to one or more detectors. The sensor
may be embedded within the structure or mounted on its surface. Impact of objects
on the structure causes a physical damage to the triboluminescent material. Such
damage causes light emission which is detected and recorded for later observation.
The intensity of emission may be measured to give an estimation of nature of the
materials structure. Several different triboluminescent materials may be arranged in
different location within the structure so that location of damage may be detected by
a single detector sensitive to the different wavelengths of light emitted by the
several materials. Light from the triboluminescent material may be detected
directly by the detectors. Alternatively, material doped with suitable photo excit-
able dyes may be caused to photo excite, and the resultant light output detected [36]
(Scheme 6.1).
Scheme 6.1 A typical novel TL material [34] (wherein x is 1 or 2; M can be europium, terbium,
dysprosium or samarium; H and C1-C6 branched- or straight-chain alkyl, thiophene and C1-C6
fluorinated alkyl wherein the fluorination may be in 1 or all positions or any intermediate value,
substituted phenyl wherein the substituents are independently selected for C1-C4 straight- or
branched-chain alkyl, Cl, Br, F, I and the phenyl group may be substituted in 1, 2 or 3 positions;
R4 and R5 are independently of each other selected from phenyl, tolyl, naphthyl, C1-C6 branched-
or straight-chain alkyl and substituted phenyl wherein the substituents are independently selected
from C1-C4 straight- or branched-chain alkyl and the phenyl group may be substituted in 1, 2 or
3 positions; R6 is selected from phenyl, tolyl, naphthyl, C1-C6 branched- or straight-chain alkyl)
The various attempts were applied in detection of impact damages in compos-

ites, civil and aerospace structures. Because of ubiquitous dispersibility of
TL-based nanocrystals throughout the entire materials, it could have a great poten-
tial for the distributive sensory system with integrated structural health monitoring.
Okoli and co-workers [5, 7, 8, 27, 29, 32] at Florida State University designed and
developed of state-of-the-art system in application of TL-based sensor systems for
transportation, civil and aerospace structures. The detail understanding of this cold
emission phenomenon at optoelectronic level and efficient application is still under
great challenge. The increasing use of advanced materials in critical structures,
where subsurface damage initiation may go unnoticed, further highlights the
urgency in developing efficient online monitoring technologies. Manufacturing of
advanced composite laminates that have been doped with various concentrations of
a triboluminescent material (ZnS:Mn ~10 %wt.) could provide an easy and smart
self-monitoring system. Laminates were manufactured using a vacuum infusion
process. Dispersing the ZnS:Mn particulates was cumbersome because their density
was higher than the resin that caused settling during resin infusion (Fig. 6.8). The
dispersion of ZnS:Mn is critical to their use in the health monitoring of the host
structure. As such, a method for mechanical agitation using a rotational vacuum
infusion apparatus was developed employing centrifugal motion. The TL materials
emit light when stressed or broken makes them ideal candidates for the ubiquitous
and in situ monitoring of structures [7]. Of course, the concentration ZnS:Mn in the
laminates or hybrid composites effects it mechanical properness and workability.
The high-concentration TL materials on the laminates reduced the tensile strength
[6]. It was seen that increasing ZnS:Mn content resulted in increased TL emission
intensity, that is, up to 10 % concentration, beyond which a notable decrease is
observed. However, as the ZnS:Mn content is increased beyond 25 wt%, the
emission intensity begins to decrease abruptly (Fig. 6.9).
Fig. 6.8 Micrographs of (a) ZnS:Mn phosphor, (b) specimen fracture and (c) ZnS:Mn sheared
components at 50 % concentration [6] © Elsevier Science
Fig. 6.9 Double plot of intensity and specific bending stiffness versus concentration loading for a
loading rate of 40 mm min1 [6] © Elsevier Science
A smart organic–inorganic composites, unreinforced vinyl ester resin and dis-

continuous glass-fiber-reinforced structural materials, etc., were fabricated incor-
porating 10 (wt.)% of triboluminescent material (ZnS:Mn). The photoemissions of
TL with organic–inorganic composites and its evaluation at the onset of damage
and crack propagation were reported. Triboluminescent microemissions of the
incorporated TL microstructures were seen in both two- and three-phase composite
systems throughout the failure loading cycle [5]. The TL emissions occur at various
intensities corresponding to initial notch-length and imminent micro-matrix frac-
ture due to external applied force. The fracturing or deformation energy was
estimated by an experimental method, where a lower threshold for excitation was
found to be approximately less than 0.5 J.m2. The respective critical composite
fracture energy is ~3–7 J.m2. Initiation of micro-cracks was observed for
reinforced samples and were subjected to three-point bend tests in lieu of the
multiple signatures of the transient signal response (Figs. 6.10, 6.11, and 6.12) [5].
Fig. 6.10 (a) MTS flexural fixture, (b) specimen geometry and (c) harness and measurement
apparatus [5] © Elsevier Science
Fig. 6.11 Micrographs of (a) concentrated composite showing evidence of micro- fracture and
direction of fracture in view of crack-tip (10–50 μm), and (b) an excerpt of a minute micro-damage
site revealing sites of possible TL emissions [5] © Elsevier Science
The smart composite matrix was synthesized using thermosetting polymer, glass
-fiber (diameter and lengths of 13 μm and 6.4 mm, respectively) and various amount
of TL crystals. During real time crack investigation test, multiple irregular emis-
sions were observed throughout the damage cycle of mechanically reinforced TL
based hybrid composite [5]. In view of catastrophic failure and deformation, the
Fig. 6.12 Comparison of the signal emission (blue line) and load–displacement (green dashed
line) plots of the (a) two-phase (unreinforced) and (b) three-phase TL concentrated composite
flexural beams with respect to time. The representative trend for brittle fracture of concentrated
resulted in one peak event. Flexural load indicates multiple emissions along the failure cycle for
toughened composites [5] © Elsevier Science
threshold energy is needed to rupture a composite beam and a lower energy

threshold exist for excitation of TL particulates. The base emissions, the tallest
peaks, are indicative of structural failure. In view of micromechanics and fracture
mechanics, the failure mechanisms in discontinuous short fiber and continuous
composites are commensurate with several modes of propagation. Fiber–matrix
debonding and delamination usually occur thereafter. Matrix cracking can be
assumed dominate, if a lower weight and volume fraction of fiber exist. Nguyen
[25, 26] describes two main damage mechanisms related to short fiber composites
as matrix cracking and fiber–matrix debonding leading to fiber pull out. In his
simulations of short fiber composites under tension, nonlinear behavior was
observed as a result of matrix cracking [5]. The macroscopic response was a
resultant of damage accumulations that lead towards initialization and propagation
of cracking. Fiber and matrix properties, aspect ratio, volume fraction and orienta-
tion distribution of short fibers were said to strongly influence the mechanical
results [24]. Meraghni et al. [24] classified damage progression as a sequential
event corresponding to the strain level that process events leading from matrix
cracking to fiber pullout and then fiber breakage.
Bourhill and co-workers [35] developed a viable route to real time structural
damage sensing. The specially designed sensors can be externally attached to the
surface of metals or composites, or embedded within composites. In situ monitoring
of structural damage in composites is particularly relevant since severe internal
damage can exist with little indication of this damage on the composite surface.
Earlier work relied on side-coupling of the triboluminescent light into a curved
conventional silica fiber and/or end-coupling into the silica fiber, if the damage to
the host structure had also broken the fiber. Both these light collection methods are,
of course, inefficient and resulted in an optical damage signal with a very poor
signal-to-noise ratio. By using novel PL polymeric and silica fibers, it is possible to
efficiently capture and guide the optical damage signal from an embedded tribolu-
minescent sensor to a remote detector. The light collection/guiding technique can
even be used effectively with triboluminescent sensors embedded within black

carbon fiber-reinforced plastic (CFRP) composites [35]. The resins and glass
fiber-reinforced plastic (GFRP) composites, triboluminescent sensors act as truly
global damage sensors, whereas for CFRP composites, the damage sensing is
localized to areas close to the embedded PL fiber [33].
Resin systems can be considered similar to fiber CFRP composites that allow
light to propagate. Therefore triboluminescent sensors in GFRP panels can be
viewed as being ‘global’ indicators of damage, i.e., the damage does not have to
occur close to the light collection/guiding PL fibers since the optical damage signal
can propagate in the GFRP composite until it is collected by the nearest PL fiber
[33, 35]. The optical damage signal detected upon impacting the resin doped with a
silica PL fiber is shown in Fig. 6.13a, whereas the signal obtained from impacting a
CFRP panel is shown in Fig. 6.13b. The CFRP panel post-impact is shown in
Fig. 6.13c. It is reported [35] that the PL silica fibers are an efficient way of
capturing and guiding the optical damage signal from embedded triboluminescent
crystals to a remote detector (Fig. 6.13a, b). This is true even for black CFRP panels
(Fig. 6.13b, c).
Fig. 6.13 (a) The optical damage signal upon impacting a resin and (b) a CFRP panel containing a
silica PL fiber. Both structures contained the terbium triboluminescent complex and a silica
capillary containing a solution of rhodamine 6G in benzyl alcohol. The CFRP panel post-impact
is also shown (c), where two PL silica fibers can be seen [35] © IOP Publishing
6.5.2 Cementitious Smart Structures
While significant work is being done in developing such integrated SHM systems
for aerospace structures, several recent works are being done in developing
TL-based sensor systems for the critical and aging civil infrastructure systems. A
ZnS:Mn-based cementitious patch was manufactured [28–30] that emits light when
stressed or fractured was developed and its TL performance under loading charac-
terized. The results show that a ZnS:Mn concentration level of ~10 % gives the best
TL response without adversely affecting the mechanical strength of the structures
while also minimizing the use of the expensive ZnS:Mn crystals. The TL response
increased as the concentration of ZnS:Mn in the system increased. The highest TL
response was obtained at a concentration level of 25 % but resulted in significant
reduction in the system’s compressive strength.
This prospective triboluminescent damage sensor (Figs. 6.14 and 6.15) of
cementitious structures comprising highly efficient triboluminescent materials
could allow simple and real-time monitoring of both the magnitude and location
of damage. The inability to effectively capture and transmit the triboluminescent
optical signals generated within opaque composites like concrete has, however,
limited their damage monitoring applications. The ITOF sensor has been developed
to enable the detection and transmission of damage-provoked triboluminescent
emissions without having to position triboluminescent crystals in the host material.
The intrinsic triboluminescent signals generated in the beams under loading were
successfully transmitted through the optical fibers to the photomultiplier tube by
side coupling. The optical sensor coupling makes a truly distributed integrated
Fig. 6.14 An in situ triboluminescent optical fiber (ITOF)-based sensing system for SHM of
bridges: (a) damage-sensitive region with ITOF sensor, (b) TL (damage) signal from ITOF sensor
due to crack, (c) signal transmission by ITOF sensor and signal detection by PMT, and (d) signal
analysis and damage identification [29] © SAGE journals
Fig. 6.15 SEM image: (a) ITOF sensor coating with randomly distributed TL crystals, (b) ITOF
sensor cross section showing interface between sensor coating (with dispersed ZnS:Mn crystals)
and polymer optical fiber, and (c) fractured surface of sensor coating with fractured crystals and
some debonding [29] © SAGE journals
triboluminescent optical fiber sensor possible when the entire length of the sensor is
mostly covered with the triboluminescent composite coating. The results show the
viability of the in situ triboluminescent optical fiber sensor for the structural health
monitoring of cementitious composites. Real-time failure detection was demon-
strated in unreinforced mortar beams, while real-time damage (crack) detection was
demonstrated in reinforced concrete beams. The work on reinforced concrete
beams showed that the integrated in situ triboluminescent optical fiber sensor was
able to detect multiple cracks caused by loading, thereby providing early warning of
structural degradation before failure [29].
6.5.3 Organic–Inorganic Hybrid Composite
Fontenot et al. [2, 12, 13] incorporated of the triboluminescent europium

dibenzoylmethide triethylammonium (europium tetrakis) into poly(methyl meth-
acrylate), etc. in view of manufacturing a durable triboluminescent polymer for
small-area sensors. Specially designed nearby sensors were designed to monitor

these structures for structural failure or impacts. The effect of polymer encapsula-
tion on the luminescent properties of TL materials is well understood [12]. The TL
particulates would need to be embedded into an inert and optically accessible
structure or composite which was the focus of the work presented here. The
luminescent behavior of ZnS:Mn powder has not been discernibly affected by
incorporating the polymer network into a clear elastomeric medium. The tribolu-
minescent emission from impact scales with phosphor concentration and was not
affected by the encapsulating medium.
The measured PL emission spectra for the ZnS:Mn-doped PDMS is shown in
Fig. 6.16. The bright emission from the ZnS:Mn-doped PDMS sample shows in the
inset. This yellow emission had a center wavelength of 587 nm, which originates
from the Mn2þ transition. The highest concentration of phosphor particles (50 %)
leads to a compound material with an increased stiffness compared to lower doping
levels where a significant degree of elasticity can still be observed when compared
to the neat polymer [12]. The elastomers are generally hydrophobic, and the ZnS:
Mn particles are statically charged, clustering does occur during the doping step.
The PDMS thickness and TL dopant concentration define the physicochemical and
spectral behavior of the smart composites.
Fig. 6.16 PL emission spectra for ZnS:Mn-doped PDMS:Photo- luminescent emission spectra for
the ZnS:Mn-doped PDMS originating from the Mn2þ transition. The PDMS samples were
irradiated using a UV lamp (365 nm). The inset shows the 50 % ZnS:Mn-doped PDMS while
under UV irradiation [12] © Elsevier Science
6.6 Further Materials Design for Delayed and Strong

Luminescence
Nature continues to provide ways for living organisms to develop sensing and real-
time monitoring abilities, with pain sensitivity being at the forefront. Manmade
structural materials could be provided with similar features that would achieve the
next-generation material status, which will further the technology gains of the
twenty-first century. Monitoring damage by conventional sensors is limited as
those methods result in expensive maintenance downtime. Most conventional
SHM systems [14, 39], such as acoustic emission sensors etc., are handicapped as
localized monitoring devices (Fig. 6.17) with high cost materials [4]. They are also
handicapped by their need for external power sources. Detection by measures of
energy distortion of an elemental body of material is a missing feature in current
techniques. This rise or change in energy transformation and exchange is the vital
measurement, as this is the basis for all dynamic systems. The change in energy
levels is the lowest form of material interactions and should be the target of sensing
technologies.
Figure 6.17 shows an example of the current state of the art of the recent
nondestructive testing feature of composite/spacecraft structures. This approach
requires a bundle of external electrodes and or costly materials, such as gold wire. It
is imperative to develop intelligent and integrated smart and low cost systems for
real-time sensing. This may provide for wireless transmission and situation assess-
ment and may result in truly ubiquitous and intrinsic SHM. An in-situ SHM system
could monitor, diagnose, and make real-time critical go-no-go decisions. It is
critical to understand the effects of service conditions with respect to the determi-
nation of health status, required maintenance, and retirement of aging structures
[6, 31]. This future SHM system aims to develop lightweight multiscale advanced
Fig. 6.17 Acoustic emission sensors with Au/Cr electrode/metallic wires [4] © John Wiley &
Sons, Inc.
composites with TL-based sensing networks yielding a paradigm shift in SHM.

Damage is detected and monitored based on the sensitivity to energy changes
within the structures or materials resulting from a damage-inducing event, such
as unexpected impact, wave pounding, and lightering.
Okoli and co-workers [46–50, 55, 56] developed structural sensory materials to
be embedded with infused triboluminescent crystals and embedded integrated self-
powered 3D sensor network will concurrently work as damage/impact sensor and
real-time transmitter respectively. Upon impact, the onset of structural damage is
indicated by light emission (TL phenomenon), the severity of damage is indicated
by the overall light intensity, and the location of damage is obtained by the
wavelength of emitted light [33]. The integrated sensor that consists of enormous
number of MWCNTs’ network, sense and convert the TL emission in real time to
electrical signals and transmit as a function of structural damage to the onboard
monitoring system. The focus is bio-mimicry–enabled pain sensitivity in critical
structural systems. The structural damage monitoring is a function of internal
energy-level change resulting in TL emissions. This means that the SHM system
is independent of any external energy sources [27, 48, 54].
6.7 Conclusion
Triboluminescent cementitious composites, GFRC, organic–inorganic hybrid com-

posite, etc. with in situ real-time failure detection capability have been reported and
demonstrated. In particular, the aerospace industry spends 27 % of its average life
cycle cost on damage detection/inspection. Additionally, present damage monitor-
ing systems consist of conventional strain gauges and or microprocessor based
electronics which need external power supply. Hybrid composites with ubiquitous
TL phase and 3D sensor are prospective self-monitored advanced materials for
future structural application. We are currently in need of unique SHM system,
which when incorporated into advanced composite components will provide crit-
ical information enhancing safety and operational cost-effectiveness.
Acknowledgment The authors gratefully acknowledge the support of the National Science
Foundation (NSF) under NSF Award (CMMI-0969413) and The Welch Foundation.
References
1. Armelao, L., Bottaro, G., Pascolini, M., Sessolo, M., Tondello, E., Bettinelli, M., & Speghini,
A. (2008). Structure–luminescence correlations in europium-doped sol–gel ZnO nanopowders.
The Journal of Physical Chemistry C, 112, 4049–4054.
2. Bhat, K. N., Fontenot, R. S., Surabhi, R., Hollerman, W. A., Aggarwal, M. D., & Alapati, T. R.
(2014). Measurement of the triboluminescent properties for europium and samarium tetrakis
dibenzoylmethide triethylammonium. Electronic Materials Letters, 10, 1149–1153.
3. Chapman, G. N., & Walton, A. J. (1983). Triboluminescence of fluorites. Journal of Physics C:

Solid State Physics, 16, 5543.
4. Chen, X., Li, J., Zhang, G., & Shi, Y. (2011). PZT nanoactive fiber composites for acoustic
emission detection. Advanced Materials, 23, 3965–3969.
5. Dickens, T., Armbrister, C., Olawale, D., & Okoli, O. (2015). Characterization of tribolumi-
nescent enhanced discontinuous glass–fiber composite beams for micro-damage detection and
fracture assessment. Journal of Luminescence, 163, 1–7.
6. Dickens, T. J., Breaux, J., Olawale, D. O., Sullivan, W. G., & Okoli, O. I. (2012). Effects of
ZnS:Mn concentrated vinyl ester matrices under flexural loading on the triboluminescent yield.
7. Dickens, T. J., & Okoli, O. I. (2011). Enabling damage detection: Manufacturing composite
laminates doped with dispersed triboluminescent materials. Journal of Reinforced Plastics and
Composites, 30, 1869–1876.
8. Dickens, T. J., Olawale, D. O., & Okoli, O. I. (2011). Toward triboluminescent sensor
realization for SHM. SPIE.
9. Duignan, J. P., Oswald, I. D. H., Sage, I. C., Sweeting, L. M., Tanaka, K., Ishihara, . . .
Bourhill, G. (2002). Do triboluminescence spectra really show a spectral shift relative to
photoluminescence spectra? Journal of Luminescence, 97, 115–126.
10. Eddingsaas, N. C., & Suslick, K. S. (2006). Light from sonication of crystal slurries. Nature,
444, 163.
11. Eddingsaas, N. C., & Suslick, K. S. (2007). Intense mechanoluminescence and gas phase
reactions from the sonication of an organic slurry. Journal of the American Chemical Society,
129, 6718.
12. Fontenot, R. S., Allison, S. W., Lynch, K. J., Hollerman, W. A., & Sabri, F. (2016). Mechan-
ical, spectral, and luminescence properties of ZnS:Mn doped PDMS. Journal of Luminescence,
170(Part 1), 194–199.
13. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., Aggarwal, M. D., & Penn, B. G. (2014).
Incorporating strongly triboluminescent europium dibenzoylmethide triethylammonium into
simple polymers. Polymer Journal, 46, 111–116.
14. Frangopol, D. M., Bocchini, P., Deco, A., Kim, S., Kwon, K., Okasha, N. M., & Saydam, D.
(2012). Integrated life-cycle framework for maintenance, monitoring, and reliability of naval
ship structures. Naval Engineers Journal, 124, 89–99.
15. Freeman, G. R., & March, N. H. (1999). Triboelectricity and some associated phenomena.
Materials Science and Technology, 15, 1454–1458.
16. Gross, G., Stranski, I. N., & Wolff, G. (1955). Zeitschrift f€
ur Elektrochemie, 59, 346.
17. Halliday, D., Resnick, R., & Walke, J. (1970). Fundamentals of Physics (6th ed.). New York,
NY: Wiley.
18. Hollerman, W. A., Fontenot, R. S., Bhat, K. N., Aggarwal, M. D., Guidry, C. J., & Nguyen, K.
M. (2013). Review of triboluminescence impact research at projectile speeds of 1 m/s to
6 km/s. Procedia Engineering, 58, 392–400.
19. Leelachao, S., Muraishi, S., Sannomiya, T., Shi, J., & Nakamura, Y. (2016). Correlation of
triboluminescence and contact stresses in ZnS:Mn/polymeric matrix composite. Journal of
Luminescence, 170(Part 1), 24–29.
20. Leelachao, S., Muraishi, S., Sannomiya, T., Shi, J., & Namamura, Y. (2015). Mechanolumi-
nescence of ZnS:Mn phosphors and its correlation to impact energy and contact geometry.
Optics Letters, 40, 4468–4471.
21. Longchambon, H. (1922). Comptes Rendus Hebdomadaires des Séances de l’Académie des
Sciences, 174, 1633.
22. Longchambon, H. (1923). Comptes Rendus Hebdomadaires des Séances de l’Académie des
Sciences, 176, 691.
23. Longchambon, H. (1925). Bulletin de la Societe Francaise de Mineralogie, 48, 130.
24. Meraghni, F., Desrumaux, F., & Benzeggagh, M. L. (2002). Implementation of a constitutive
micromechanical model for damage analysis in glass mat reinforced composite structures.
Composites Science and Technology, 62, 2087–2097.
25. Nguyen, B. N., Gao, F., Henager, C. H., Jr., & Kurtz, R. J. (2014). Prediction of thermal
conductivity for irradiated SiC/SiC composites by informing continuum models with molec-
ular dynamics data. Journal of Nuclear Materials, 448, 364–372.
26. Nguyen, B. N., & Kunc, V. (2010). An elastic-plastic damage model for long-fiber thermo-
plastics. International Journal of Damage Mechanics, 19, 691–725.
131, 1407–1418.
28. Olawale, D. O., Dickens, T., Uddin, M. J., & Okoli, O. I. (2012). Triboluminescence
multifunctional cementitious composites with in-situ damage sensing capability. SPIE, San
Diego, CA.
29. Olawale, D. O., Kliewer, K., Okoye, A., Dickens, T., Uddin, M. J., & Okoli, O. I. (2014).
Getting light through cementitious composites with in situ triboluminescent damage sensor.
Structural Health Monitoring, 13, 177–189.
30. Olawale, D. O., Kliewer, K., Okoye, A., Dickens, T. J., Uddin, M. J., & Okoli, O. I. (2014).
Real time failure detection in unreinforced cementitious composites with triboluminescent
sensor. Journal of Luminescence, 147, 235–241.
31. Olawale, D. O., Sullivan, G., Dickens, T., Tsalickis, S., Okoli, O. I., Sobanjo, J. O., & Wang, B.
(2012). Development of a triboluminescence-based sensor system for concrete structures.
32. Olawale, D. O., Kliewer, K., Dickens, T., Uddin, M. J., & Okoli, O. I. (2013). Tribolumines-
cent optical nerves for smart concrete structures. Structural Health Monitoring, 1–2,
1376–1383.
34. Sage, I., Bourhill, G., & Oswald, I. (2007). Triboluminescent materials and devices. US Patent,
US7270770 B2.
black composites: Embedded triboluminescent structural damage sensors. Smart Materials
and Structures, 10, 332.
36. Sage, I., Badcock, R., Humberstone, L., Geddes, N., Kemp, M., Bishop, S., et al. (1999).
Triboluminescent damage sensors. Smart Structures and Materials Technologies, 3675, 169.
37. Selle, B. (1964). Das Abklingen der Lumineszenz von ZnS: Mn bei Anregung im Gebiet der
Mn-Eigenabsorption. Physica Status Solidi B, 5, 649–656.
38. Sharipov, G. L., Tukhbatullin, A. A., & Abdrakhmanov, A. M. (2011). Triboluminescence of
crystals and suspensions of inorganic salts of lanthanides. Protection of Metals and Physical
Chemistry of Surfaces, 47, 13–19.
39. Sielski, R. A. (2012). Ship structural health monitoring research at the office of naval research.
JOM, 64, 823–827.
40. Sweeting, L. M. (2001). Triboluminescence with and without Air. Chemistry of Materials, 13,
854–870.
41. Sweeting, L. M., Cashel, M. L., Dott, M., Gingerich, J. M., Guido, J. L., Kling, J. A., . . .
Spence, R. A. (1992). Spectroscopy and mechanism in triboluminescence. Molecular Crystals
and Liquid Crystals Science and Technology. Section A. Molecular Crystals and Liquid
Crystals, 211, 389–396.
42. Sweeting, L. M., Cashel, M. L., Dott, M., Gingerich, J. M., Guido, J. L., Pippin, R. F., III, . . .
Spence, R. A. (1992). Molecular Crystals and Liquid Crystals, 211, 389.
43. Sweeting, L. M., Cashel, M. L., & Rosenblatt, M. M. (1992). Triboluminescence
spectra of organic crystals are sensitive to conditions of acquisition. Journal of Luminescence,
52, 281–291.
44. Sweeting, L. M., Rheingold, A. L., Gingerich, J. M., Rutter, A. W., Spence, R. A., Cox, C. D.,
& Kim, T. J. (1997). Crystal structure and triboluminescence .2. 9-anthracenecarboxylic acid
and its esters. Chemistry of Materials, 9, 1103–1115.
45. Takada, N., Sugiyama, J.-I., Katoh, R., Minami, N., & Hieda, S. (1997). Mechanoluminescent
properties of europium complexes. Synthetic Metals, 91, 351–354.
46. Uddin, M. J., Daramola, D. E., Velasquez, E., Dickens, T. J., Yan, J., Hammel, E., . . . Okoli, O.
I. (2014). A high efficiency 3D photovoltaic microwire with carbon nanotubes (CNT)-quantum
dot (QD) hybrid interface. Physica Status Solidi (RRL) - Rapid Research Letters, 8, 898–903.
47. Uddin, M. J., Davies, B., Dickens, T. J., & Okoli, O. I. (2013). Self-aligned carbon nanotubes
yarns (CNY) with efficient optoelectronic interface for microyarn shaped 3D photovoltaic
cells. Solar Energy Materials & Solar Cells, 115, 166–171.
48. Uddin, M. J., Dickens, T., Yan, J., Chirayath, R., Olawale, D. O., & Okoli, O. I. (2013).
Solid-state dye sensitized photovoltaic micro-wires (DSPM) with CNT yarn as counter
electrode: Synthesis and characterization. Solar Energy Materials & Solar Cells, 108, 65–69.
49. Uddin, M. J., Dickens, T. J., Yan, J., Olawale, D. O., Okoli, O. I., & Cesano, F. (2012).
Solid-state dye sensitized optoelectronic carbon nanotube-wires: An energy harvesting dam-
age sensor with nanotechnology approach. ASME 2016 Conference on Smart Materials,
Adaptive Structures and Intelligent Systems, 8200, 19–21.
50. Uddin, M. J., Dickens, T. J., Yan, J., Olawale, D. O., Okoli, O. I., Cesano, F., & ASME. (2013).
Solid-state dye sensitized optoelectronic carbon nanotube-wires: An energy harvesting
damage sensor with nanotechnology approach.
51. Valeur, B., & Berberan-Santos, M. N. (2011). A brief history of fluorescence and
phosphorescence before the emergence of quantum theory. Journal of Chemical Education,
88, 731–738.
52. Walton, A. (1977). Journal of Advanced Physics, 26, 887.
Measurement of triboluminescence and proton half brightness dose for ZnS : Mn. IEEE
Transactions on Nuclear Science, 51, 1737–1741.
55. Yan, J., Uddin, M. J., Dickens, T., Olawale, D. O., & Okoli, O. I. (2013). 3D photovoltaic
sensors for in-situ structural health monitoring of advanced composites. Structural Health
Monitoring, 1 & 2, 1645–1653.
56. Yan, J., Uddin, M. J., Dickens, T. J., Daramola, D. E., & Okoli, O. I. (2014). 3D wire-shaped
dye-sensitized solar cells in solid state using carbon nanotube yarns with hybrid photovoltaic
structure. Advanced Materials Interfaces, 1, 1400075.
Chapter 7
Europium Tetrakis Dibenzoylmethide
Triethylammonium: Synthesis, Additives,
and Applications
Ross S. Fontenot, Kamala N. Bhat, William A. Hollerman,

and Mohan D. Aggarwal
7.1 Introduction
There are a number of techniques currently being used for damage detection and
monitoring of civil, aerospace, military structures, and aircraft [1]. However, the
major drawbacks of the current techniques are that they do not provide in situ and
distributed sensing [1–4]. In addition, the associated cost resulting from the down-
time required for periodic nondestructive inspections can be very high for aero-
space structures like aircrafts and civil structures such as bridges [1].
The emission of triboluminescence (TL) could be used as the active element for
smart structural sensors [5]. In 1888, Wiedemann and Schmidt defined TL as the
emission of light produced by mechanical action [6–9]. Triboluminescent based
sensor systems may be able to overcome these challenges as they have the potential
for wireless, in situ, and distributed sensing that can enable real time continuous
monitoring [1]. In addition, triboluminescent based sensor systems comprising of
triboluminescent materials could allow for real-time monitoring of both the mag-
nitude and location of damage with respect to the host structure [10, 11]. They can
also be used as stress, fracture, and damage sensors [8, 12].
R.S. Fontenot (*)

Carderock Division, Naval Surface Warfare Center, Code 632, West Bethesda 20817, USA
e-mail: rsfontenot@hotmail.com
K.N. Bhat • M.D. Aggarwal
Department of Physics, Chemistry, and Mathematics, Alabama A&M University,
P.O. Box 1268, Normal, AL 35762, USA
W.A. Hollerman
Department of Physics, University of Louisiana, P.O. Box 44210, Lafayette, LA 70504, USA

148 R.S. Fontenot et al.
In 1999, Sage and Geddes used the property of TL to patent a design for a sensor
capable of discerning the locations of impacts [11, 13, 14]. Their design involved
coating a material with a triboluminescent crystal or creating a composite tribolu-
minescent object [11, 13, 14]. A sensor would then be embedded within the
structure or mounted on its surface [14]. Impacts to the structure would produce
light which would be recorded and analyzed to determine the location [14]. In
addition, Sage et al. proposed that several different triboluminescent materials
could be used and arranged at various locations [11, 13, 14]. The advantage is
that when an impact takes place, its location could be determined by the wavelength
emitted [14]. For example, by placing two different triboluminescent materials at
known distances from the detector, it is possible to determine the approximate
location of the impact by measuring the emitted wavelength.
In 2000, Xu et al., were the first to propose that manganese doped zinc sulfide
(ZnS:Mn) and strontium aluminum oxide doped with europium (Sr0.975Al2O3.985:
Eu0.01) could be used for mechanical stress sensors [15–17]. They fabricated a film
of ZnS:Mn on various ceramic substrates [17]. The substrates were then subjected
to mechanical stress and various impacts. This caused the ZnS:Mn to emit a bright
light clearly visible with the naked eye [17]. In experiments with Sr0.975Al2O3.985:
Eu0.01, Xu et al. applied a known stress to plastic objects with a 0.1 mm layer of the
phosphor [16]. Their results showed that the light output of the phosphors was
proportional to the applied strain [16, 17].
In 2003, Goods et al. proposed a triboluminescent polymer foam for mechanical
sensing [18]. Typically, triboluminescent materials have been incorporated into
fully dense polymer resins for damage sensors [18, 19]. However, this method
produces a small signal since the triboluminescent response was localized only to
the failure location of the resin. None of the TL material away from the cracks of the
polymer could increase the triboluminescent output due to little or no mechanical
damage [18, 19]. To combat this problem, Goods et al. proposed to incorporate
triboluminescent powder materials into the cellular structure of foam. This creates a
uniform distribution that allows for the entire structure to collapse and all the
triboluminescent material to respond to the destruction [18]. Besides the increase
in triboluminescent output, this method has other benefits including the host foam
being able to be fabricated into any arbitrary shape, be molded in situ to fill the free
volume of existing structures, lower failure points than resins (can detect lower
energies), and finally the foam can be incorporated in the mechanical design of
components [18]. Their results showed that europium tetrakis dibenzoylmethide
triethylammonium (EuD4TEA) could be incorporated into a brittle foam formula-
tion that yields TL when stressed and strained [18].
If these sensors are to become reality, however, the TL emitted must be bright.
Of the hundreds of known triboluminescent materials, only a few are bright enough
to be seen in daylight. This chapter explores a triboluminescent material known as
EuD4TEA. Its synthesis, effects of additives, and incorporation into simple poly-
mers are discussed.
7 Europium Tetrakis Dibenzoylmethide Triethylammonium. . . 149
7.2 Background
7.2.1 Luminescence
Luminescence is defined to be the emission of nonthermal electromagnetic

(EM) radiation [20–22]. It is observed in all phases of matter for both organic
and inorganic materials. These luminescent materials are usually composed of a
host matrix crystal and a small amount of intentionally added impurity atoms
[23–27]. These impurities are usually known as activators or dopants. It is these
dopants that are responsible for the characteristic light emission [22–27].
In 1900, Max Planck postulated that an oscillator could emit or absorb EM
radiation only in integral multiples of energy. The energy of these packets (photons)
has discrete levels as shown in Fig. 7.1. The absorption and emission of light by an
atom is a discrete instantaneous transition between a ground state (E0) and an
excited state (E1) [20, 23, 28–32]. The transition energy (ΔE) of the absorbed or
emitted photon is
ΔE ¼ E1 E0 ¼ hυ ð7:1Þ
where ν is the frequency and h is Planck’s constant.

As Fig. 7.1a shows, atoms gain energy by absorbing photons. The absorbed
energy must correspond to the difference between the two levels ΔE as given in
Eq. 7.1 [33]. Light with an energy that is not equal to ΔE will be reflected by the
material. Atoms can de-excite either by radiative (photon) or non-radiative (pho-
non) emission. Phonons are quanta of lattice vibrations that often result in the
release of thermal energy [33, 34]. If the energy is emitted as photons (light), then
the wavelength will correspond to the energy lost by the atom [33]. This relation-
ship is given by
hc
ΔE ¼ ð7:2Þ
λ
Fig. 7.1 (a) Absorption and (b) emission of photons for a two-level system
where λ is the wavelength and c is the speed of light. The time the atom remains
excited is known as the state lifetime and is a characteristic of the atom [33].
Once the excitation has ceased for many materials, the fluorescence light yield
(I) decays according to the equation
I ¼ I 0 eαt ð7:3Þ
where I0 is the initial light yield, α is a constant, and t is the time. The time needed to
reduce the light intensity to e1 (36.8 %) of its original value is defined as the
fluorescence decay time (τ) and is equal to τ1. Luminescence can be divided into
two categories: fluorescence and phosphorescence. Phosphorescence, also known
as delayed emission, has a much longer decay lifetime than fluorescence. Typical
phosphorescence lifetimes can vary from 100 ms up to many seconds in duration,
while the fluorescence usually ranges from 1 ns to 10 ms [35]. More specifically,
phosphorescence is the emission of light from a triplet excited state, which is where
the electron in the ground state and excited state has the same spin orientation
[35]. The triplet state is due to the fact that the spin flips take more time to complete.
Fluorescence occurs in singlet states, with the excited electron having the opposite
spin of the paired ground state, resulting in a short emission time for photons
[35]. The fluorescence decay time is unique to each material. It might be possible
to use the fluorescence decay time as an indicator to gauge the production of TL
from impacts. Some materials such as ZnS:Mn,Cu have multiple decay times
[36, 37]. Equation 7.3 is valid for a large number of luminescent materials [36, 37].
7.2.2 Triboluminescence
Luminescence is usually classified by the mechanism that causes its excitation

[21, 24, 38, 39]. The most widely known type of luminescence is likely photolumi-
nescence (PL), which is excitation by optical energy [38]. Other types include
electroluminescence, cathodoluminescence which are excitation by electrical
energy, and triboluminescence (TL) which is caused by the fracture of crystals
[6, 38]. While TL is not as familiar to the public at large, common examples include
the blue light emitted from real Wint-O-Green lifesavers [40] and the peeling of
tape [41]. The most dramatic examples, however, occur during an earthquake. TL is
produced by the friction and fracture in Earth’s crust during and just prior to the
earthquake [38, 42, 43].
Luminescence induced by any mechanical action is known as mechanolumi-
nescence (ML) [44]. This category can generally be divided into two more sub-
divisions: (1) the light emitted during elastic and plastic straining known as
deformation luminescence (DL) and (2) the light emitted during fracture known
as TL. The mechanoluminescent intensity depends on a large number of factors,
which includes stress, strain, strain rate, pressing rate, temperature, atmospheric
pressure, dielectric constant, mobility, size of particles, size of crystals, mass of
samples, trap-depth, piezoelectric constant, concentration of traps, radiation dose,
dopants, dopant concentration, firing temperature of the sample, sintering

temperature of the sample, annealing temperature, firing time, sintering time,
capture cross section of charge carriers, number of successive pressings, efficiency
for the conversion of mechanical energy into light energy, materials used, crystal-
structure, crystallographic direction, piezoelectric constant, dislocation density,
surface area of newly created surfaces, dislocation charge, hardness of samples,
and age of the sample [44–63]. As a result of this complexity, the exact cause of ML
is unknown [6, 44]. Despite this limitation, great progress has been made experi-
mentally and theoretically in the understanding of ML [6, 44]. However, since
1888, several properties or characteristics have been discovered [6]:
1. TL occurs more frequently in crystals whose structure lacks a center of symme-
try than in crystals possessing center symmetry.
2. If a compound is optically active and one of the isomers is triboluminescent, the
other one may or may not be triboluminescent.
3. With certain compounds, the triboluminescent activity may fade over a period of
weeks or months after the crystals have been grown. Such types of materials
exhibit temporary TL. Thus far, only arsenic trioxide has been shown to unam-
biguously possess this property.
4. The triboluminescent intensity produced depends on the physical conditions of
the crystallization.
5. For every material, there is a crystal size below which the substance is not
triboluminescent.
6. Elements by themselves are not triboluminescent.
7. Few amorphous substances are triboluminesccent, which indicates that a peri-
odic crystal structure is required.
8. Triboluminescent emission can arise from the dielectric breakdown of the
surrounding gas during fracture.
Since the 1930s, it has been known that alkali halide crystals irradiated by X- or
γ-rays experience DL [6]. The amount of light varies between materials; however, as
an estimate approximately 10 % of the light occurs in the elastic region, 75 % from
the plastic region, and 15 % from TL. In addition, it has been discovered that
generally no DL is produced from alkali halide crystals which are either unirradiated
or merely additively covered. However, TL is observed in both cases [6].
It is believed that DL comes from the recombination of free electrons (e) with
luminescence centers (L ). The free electrons are produced when a mobile disloca-
tion (D) interacts with an F-center (color center). This can be expressed as [6]:
D þ F ! e; ð7:4Þ
0
e þ L ¼ L þ f: ð7:5Þ
In 1970, Senchukov and Shmurak discovered that the DL hysteresis existing

under both elastic and plastic deformation is consistent with electrons being released
from F-centers. Under low stresses, dislocation lines bend elastically about
their pinning points (point defects). As they bend, electrons are released from the
F-centers [6]. The electrons will then combine with luminescent centers to produce
DL. In 1969, Shmurak showed that the total number of photons emitted from
irradiated KCl is proportional to the area swept out by the dislocation lines [6]. It
was also discovered that after removing the samples from low stress, the dislocation
lines return to their original positions by the tensions acting along these lines [6]. As
they do, they move through regions of the crystal cleared of F-centers. To produce
new luminescence, a stress must be applied which will bend the dislocation lines
further than in the previous cycle. At even higher stresses, a dislocation line will
unpin itself [6]. This freely moving dislocation produces a glide in the crystal which
is consistent with the observation that the DL peaks which occur during plastic
deformation do so in the glide region of the stress-strain curve [6].
In 1956, Bohun and by Pirog and Sujak (1968) observed that electrons are
emitted from the surface of X-ray irradiated alkali halides when they are stressed.
This emission was attributed to the presence of F-centers in the surface regions of
the crystal [6]. It should be noted that no emission of electrons was detected when
the crystal had been bleached prior to stressing. In addition, Gyulai and Hartly
discovered that plastic deformation produced a large transient increase in the
conductivity of natural rock salt crystals. The stress required to produce the first
substantial rise in conductivity agreed roughly with the yield stress [6]. If the load
was returned to its original value, removal and reapplication of the same load had
no further effect. A new transient effect was produced only if the previous total load
was exceeded. This result was attributed to the electrons being released by the
motion of the dislocation [6].
In 1966, Butler studied the luminescence produced when straining γ-irradiated
alkali halides. When deformed plastically, the crystals glowed [6]. The maximum
wavelength at which the main peak occurred in each salt’s DL spectrum was found
to fall on a single line on a Mollwo–Ivey plot. In this particular type of plot, the
maximum wavelength is graphed against the square of the lattice constant of the salt
under consideration [6]. The results from this indicated that the luminescence
produced from each salt was similar. In another experiment, Butler measured the
fraction of F and M-centers remaining after various bleaching times when the KCl
crystal was flooded with light absorbed by the F-band [6]. Their data indicated that
the DL curve goes through a maximum at the same bleaching time as does the M-
center concentration [6]. Therefore, both the F and M-centers were involved in the
luminescence process. Because additively colored crystals containing both F and
M-centers do not exhibit DL, a second type of center known as the recombination
center must also be involved in the luminescence process [6].
7.2.3 Mechanisms for Triboluminescence
The oldest method of triboluminescent excitation is by fracturing crystals [64]. By

crushing grains in a bowl, Alzetta et al. showed that the emitted light yield was
directly proportional to the mass of the triboluminescent material [64]. By using a
device that blew phosphors against a glass plate, it was discovered that the total
light emitted was proportional to the total surface of the grain [64]. In 1970, Alzetta
et al. also discovered that the triboluminescent intensity is directly proportional to
the rate of change of the applied pressure [64]. Similarly, Alzetta et al. reported that
the ML produced by ZnS:Mn occurs 6 % in the elastic region, 80 % in the plastic
region, and 14 % during fracture [64]. For the current research, the bowl has been
replaced with a custom built drop tower. By using a set amount of crystals and a
fixed height, the TL can be accurately measured. Thus, the effects of dopants and
different materials can also be studied and measured.
7.2.4 Review of Triboluminescent Research
Since 2001, Hollerman et al. have been investigating the triboluminescent proper-
ties of ZnS:Mn [15, 65–69]. Using a drop tower, Bergeron et al. measured the low
velocity (few m/s) triboluminescent spectrum of ZnS:Mn [65, 70]. For this
research, the ZnS:Mn powder was placed on a sample tray with a tube mounted
on top. A 130 g ball was placed a set distance above the powder using a release pin
placed through one of several holes in the drop tube [71]. This allowed the user to
vary the drop height and thus the impact velocity. By placing a fiber optic under-
neath the drop tube, Bergeron et al. measured the triboluminescent spectrum of
ZnS:Mn as shown in Fig. 7.2 [71]. The triboluminescent emission was greater for
higher speed (higher energy) drops indicating a relationship between intensity of
emitted light and the impact velocity.
Later in 2005, Bergeron et al. quantified the relationship between the tribolumi-
nescent intensity and impact velocity for ZnS:Mn by placing a photodetector
140
4.5 m/s
120
Arbitrary Light Intensity
100
80
3.9 m/s
60
3.2 m/s
40
20
0
500 550 600 650 700
Wavelength (nm)
Fig. 7.2 ZnS:Mn triboluminescent emission spectra for low speed impacts [71]
Fig. 7.3 Photomultiplier detector response for ZnS:Mn [65, 69–71]
Fig. 7.4 Plot of the detector potential versus impact velocity for ZnS:Mn [65, 69–71]
underneath the drop tube [65, 69–71] A typical triboluminescent emission of ZnS:
Mn captured by the photodectector and recorded on an oscilloscope is shown in
Fig. 7.3. Notice the orange TL produced by ZnS:Mn [65, 69–71].
Bergeron et al. determined that the intensity of the triboluminescent response
appears to be a function of impact velocity with two regions of interest as shown
in Fig. 7.4. The first region is in the velocity range below 2 m/s. Here, the
production of triboluminescent light appeared to have a threshold of approximately
Velocity (km/s)
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
12
10
8
Light Ratio
0
2 4 6 8 10 12 14 16 18 20
Impact Energy (J)
Fig. 7.5 Light yield ratio versus impact energy for ZnS:Mn [69–71]
0.5 m/s [65, 69–71]. Above this threshold, the projectile has sufficient
velocity (or energy) to break ZnS:Mn crystals and to produce triboluminescent
light [65, 69–71].
Bergeron et al. also determined that the triboluminescent intensity increases
rapidly until about 2 m/s [65, 69–71]. The second region begins at 2 m/s and
appeared to be a saturation state, where the slope was shallow. This was likely due
to the ball breaking as many crystals as possible in the impact area. Increasing the
energy above this threshold cannot generate more light due to the lack of crystals in
the impact area thus indicating less sensitivity to impact velocity [65, 69–71].
In 2006, Bergeron et al. measured the hypervelocity (>1 km/s) impacts using the
two-stage light gas gun at the NASA Marshal Space Flight Center (MSFC) in
Huntsville, AL [69–71]. Their goal was to determine the relationship between impact
energy and triboluminescent emission. As such, shots were taken to intentionally
vary the velocity of the projectile. ZnS:Mn data was collected at impact velocities of
3.3, 4.7, and 5.6 km/s. To analyze this data, the plots of time-dependent light intensity
were numerically integrated from the peak of triboluminescent intensity for one
millisecond beyond the peak. The resulting integrated light yield ratios for ZnS:Mn
as functions of both impact energy and velocity are shown in Fig. 7.5 [69–71].
The “light ratio” is defined as the integrated triboluminescent yield at a given
impact speed divided by the equivalent integrated emission at 3.3 km/s. The
average integrated light intensity of the 4.7 km/s impacts was measured to be
about 6.2 times greater than the 3.3 km/s impacts, and the average integrated
light intensity of the 5.6 km/s impacts was measured to be about 4.8 times greater
than the 3.3 km/s impacts as shown in Fig. 7.5. Both the 4.7 and 5.6 km/s impacts
were statistically more luminous than the 3.3 km/s impact. However, it was not
Fig. 7.6 GSFC one stage light gas gun [72]
statistically possible to determine the exact nature on the dependence of tribolumi-

nescent emission with respect to impact energy or velocity. These impacts produced
a triboluminescent decay time of 292 58 μs [71]. In addition, Bergeron
et al. determined that the integrated light intensity of the impacts shows that the
4.7 km/s impact was approximately 6.2 times more luminous than the 3.3 km/s.
However, the error bars determined by taking the standard deviation of the mean
made the 5.6 and 4.7 km/s impact statistically indistinguishable. Their result
suggests that emitted TL increases with kinetic energy [71].
In 2007, Malespin et al. measured the triboluminescent decay time from ballistic
(0.1–1 km/s) impacts using the one-stage light gas gun at the NASA Goddard Space
Flight Center (GSFC) in Greenbelt, MD [72]. The GSFC gun was capable of
launching projectiles with velocities between 0.1 and 1 km/s and was comprised
of four main parts: breech, barrel, target chamber, and catch tank, shown in Fig. 7.6
[72]. In order to fire this gun, first high pressure gas must be loaded into the breech
chamber. The gas was then released to push the sabot down the barrel and impact
the sabot catcher. The data is then collected using light detectors and oscilloscopes.
Once the shot has finished, air was bled back into the gun raising the pressure back
to atmospheric level before the tank is opened. This entire process took about an
hour [72].
The work of Malespin et al. determined that ZnS:Mn has a triboluminescent
lifetime of approximately 300 μs. While triboluminescent light can be produced
Fig. 7.7 First ballistic triboluminescent impact spectrum for ZnS:Mn [75]
during ballistic impacts, it was difficult to separate TL from the other sources of
light, such as debris impact, sparks, or flashes caused by thermal effects or phase
changes [72]. In addition, the triboluminescent light was sometimes not intense
enough to overcome the other sources which cause decay times to be much different
depending on the combinations of projectile, materials used, and concentration of
triboluminescent phosphor [72]. As a result of the research, Malespin
et al. concluded that a spectrum would yield better data for the identification of
impacts. However, at that time, an impact spectrum could not be measured due to
the limitations in the setup of the light gas gun [72].
While the work of both Malespin and Bergeron expanded the knowledge base of
triboluminescent detection, no triboluminescent spectrum was recorded. In 2011,
Fontenot et al. were the first to investigate the triboluminescent spectra of ZnS:Mn
at ballistic velocities [73, 74]. The first TL ballistic spectrum for ZnS:Mn ever
recorded is shown in Fig. 7.7 [75].
The research began by machining a hole in the 9 mm and 30 caliber projectiles
[73, 74]. The hole depths were 12 mm for the 9 mm and 10 mm for the 30 caliber
projectiles [73, 74]. The ZnS:Mn powder was then placed inside the projectile and
the hole was covered using epoxy [73, 74]. After the epoxy cured, the 9 mm and
30 caliber projectiles were fired into an aluminum and steel target, respectively
[73, 74]. The triboluminescent spectrum was then recorded for each impact. The
recorded 30 caliber spectra are shown in Fig. 7.8.
Fontenot et al. results showed that there was approximately a 1 nm shift in the
triboluminescent peak wavelength between the 30 caliber and 9 mm projectiles
augmented with ZnS:Mn powder [73, 74]. The higher energy 30 caliber projectiles
had a peak wavelength that was about 1 nm longer (lower in energy) than the
equivalent data measured for the lower energy 9 mm projectiles augmented with
a b
40 40
589.4 nm 587.1 nm
30 30
Light Yield
Light Yield
20 20
10 10
0 0
-10 -10
400 500 600 700 800 400 500 600 700 800
Wavelength (nm) Wavelength (nm)
Fig. 7.8 ZnS:Mn triboluminescent spectra for 30 caliber projectiles with speeds of (a) 932 m/s
and (b) 897 m/s [74, 75]
ZnS:Mn powder [73, 74]. This research indicated that ZnS:Mn could be used as a
pressure sensor as well as an impact sensor capable of discerning the impact energy
by investigating the triboluminescent wavelength emission.
7.3 Europium Tetrakis Dibenzoylmethide

Triethylammonium
The purpose of this chapter is to explore the TL properties of EuD4TEA for its use
in impact sensors. In 1966, Hurt et al. synthesized the first EuD4TEA material
[76]. To put it into perspective, EuD4TEA emits 106 % more TL compared to the
more commonly known ZnS:Mn [37]. The synthesis of EuD4TEA began by
preparing a solution of 2 mmol of anhydrous europium chloride in 10 mL of pure
ethyl alcohol. This was then added to a hot solution containing 8 mmol of
1,3-diphenyl-1,3-propanedione also known as dibenzoylmethane (DBM) and
8 mmol of triethylamine (TEA) in 60 mL of absolute ethyl alcohol. The crystals
that formed were collected by filtration under suction and repeatedly washed with
ethanol until no chloride ion was detected by testing the filtrate with silver nitrate.
The crystalline product was then air dried.
In 1987, Sweeting et al. studied the crystal structure for EuD4TEA [77] using the
synthesis described above. Purification by recrystallization was accomplished by
room temperature evaporation of methanol and dichloromethane [77]. After testing
both materials for triboluminescence by grinding each of the crystals with a glass
rod or steel spatula, it was discovered that the methanol based material exhibited TL
while the dichloromethane lacked TL [76].
To understand the structure of EuD4TEA, the final products were analyzed using
X-ray diffraction (XRD). From these results, Sweeting et al. determined that the
material made with methanol showed no evidence of solvent incorporation and had
Fig. 7.9 Plot of the excitation and emission spectra for EuD4TEA with its associated electronic
transitions [81]
a centric symmetry belonging to the I2/a space group [77]. However, in 2001
Cotton et al. showed that EuD4TEA does not belong to the I2/a space group, but
instead to the non-centric (and polar) space group Ia [78].
The measured photoluminescent emission and excitation spectra for EuD4TEA
are shown in Fig. 7.9. The excitation spectrum was measured using the 612 nm
(2.03 eV) emission peak, while the emission spectrum was measured using the
412 nm (3.01 eV) excitation wavelength. The spectrum indicates that the lumines-
cence from the europium compound is due to its Eu3þ ion that produces the excited
state Eu3þ centered transitions from the 5D0 levels to the lower 7F0–4 levels
[79, 80]. Due to overwhelming numbers, the wavelengths corresponding to each
measured transition are shown in Table 7.1 [81].
Further it was observed that in Fig. 7.9, the excitation energy is always from the
ground 7FJ levels to the higher excited states. In order to save space and make the
graph more readable, each energy transition has been color coded. Thus the 3.01 eV
(312 nm) energy transition would be the blue 7F1 ! 5F2. Also, notice that the
majority of the excitation electronic transitions are from the 7FJ ! 5DJ and 5GJ
levels. The most probable excitation transition is the 3.97 eV 7F4 ! 5G3 transition
while the most favorable emission transition is the 2.03 eV 5D0 ! 5F2 transition.
This indicates that the europium compound can only be excited by ultraviolet to
blue light, which yields a red luminescence. The energy level diagrams
corresponding to the Fig. 7.9 transitions that are listed in Table 7.1 are shown in
Fig. 7.10. The energy levels were determined from the measured values in Fig. 7.9
and from Carnall et al. [79].
Table 7.1 The electronic transitions for EuD4TEA with the associated wavelengths and
energies [81]
Excitation Wavelength Energy Emission Wavelength Energy
transition (nm) (eV) transition (nm) (eV)
7
F ! 5F 312.0 3.97 7
D ! 7F 591.0 2.10
7
F! H5 326.0 3.80 7
D ! 7F 611.5 2.03
614.0 2.02
7
F ! 5D 362.4 3.42 7
D ! 7F 651.8 1.90
7
F ! 5D 378.4 3.28 7
D ! 7F 702.1 1.77
7
F! D5 405.1 3.06
7
F ! 5D 410.4 3.02
7
F ! 5G 420.7 2.95
424.9 2.92
7
F ! 5D 433.8 2.86
7
F ! 5G 441.5 2.81
7
F ! 5G 447.6 2.77
7
F! G5 460.4 2.69
7
F! D5 465.8 2.66
7
F ! 5D 484.6 2.56
7
F ! 5D 488.0 2.54
7
F! D5 526.4 2.36
7
F! D5 536.6 2.31
7.4 Experimental Methods and Analytical Techniques
7.4.1 Design and Fabrication of a Drop Tower

for Testing TL Materials
The TL emission of the phosphor has to be bright, so that the light from the fracture
can reach the detector with a sufficiently strong signal. However, the majority of the
known triboluminescent materials do not emit light with sufficient intensity to
allow detection with compact and inexpensive detectors [82]. To test these candi-
dates, a specially designed drop tower, shown in Fig. 7.11, was fabricated in the
laboratory [37]. This drop tower system allows the user to determine the light yield,
triboluminescent spectrum, and light duration for hundreds of materials. The tower
was designed and fabricated using a 0.75 in. (19 mm) medium density fiberboard,
0.25 in. (6 mm) Plexiglass plate, and a 1.25 in. diameter (32 mm) piece of PVC pipe
as shown in Fig. 7.11. In the set up the pipe has holes drilled in measured
increments. These holes are used to place a pin that holds a 1.25 in. (32 mm)
diameter steel ball (0.13 kg), and serves as the release mechanism [37].
The experiment begins by placing a sample material on the Plexiglass plate as
shown in Fig. 7.11. The material is arranged so that it is aligned about the center of
the tube with a minimum height. Next, a 0.13 kg steel ball is placed at a set distance
Fig. 7.10 Excitation and emission energy level diagrams for the EuD4TEA ions [81]
(normally 42 in.). The ball is then released and falls with a known velocity
impacting the powder. After each test, the drop tube is removed, the ball is cleaned,
and the powder redistributed near the center of the target area [37].
To determine TL intensity for each material, a photodiode is placed under the
Plexiglass plate as shown in Fig. 7.11. An amplifier increases the signal into an
oscilloscope which records the triboluminescent data in single sequence mode.
Table 7.2 shows the oscilloscope settings that have been used for most of the tested
samples. Once the signal was acquired, it was analyzed using custom LabVIEW
programs specifically written for this type of research. These programs have a
function that integrates the area under the curve there by giving the total light
output by each material per drop and calculates the decay time [37].
Knowing the decay time and relative intensities of materials is useful; however,
how does one characterize materials with similar triboluminescent yields or decay
Fig. 7.11 Schematic diagram of the specially designed drop tower used to measure tribolumi-
nescent light yield [37]
Table 7.2 Typical Vertical scale (V/div) 2 or 4

oscilloscope settings used
Time scale (μs) 500
to measure triboluminescence
in the drop tower [37] Center position (ms) 1.680
Probe attenuation 20
Trigger (mV) 160
times? One way to solve this dilemma was to measure the triboluminescent
spectrum emitted at impact. To do this, the photodetector was replaced with a
cosine corrector connected to a spectrometer. Following the same procedures as
described above, a triboluminescent spectrum was recorded with integration time
set to 100 ms.
For troubleshooting or demonstration purposes, the detectors can be removed
and a mirror placed underneath the drop tower at a 45 angle. This allows a user to
observe and record the TL emission with a camera or other such device. Figure 7.12
shows an example of three ZnS:Mn TL images taken from the drop tower using a
NAC GX-1 high speed video camera at a frame rate of 1000 Hz. Triboluminescence
from ZnS:Mn was visible in 19 frames of the camera video. It should be noted that
the NAC GX-1 high-speed camera is similar to one used on the popular
Mythbusters® television show [37].
Fig. 7.12 Three still pictures of the TL emission from ZnS:Mn powder in the drop tower recorded
using a NAC GX-1 high-speed camera at a frame rate of 1000 Hz [37]
Fig. 7.13 Diagram of the LabVIEW program needed to determine the light yield and spectrum
from data [37]
7.4.2 Specially Written LabVIEW Software
As described above, the particular property being measured would determine which
custom LabVIEW virtual instrument (VI) program would be run as shown in
Fig. 7.13. If the triboluminescent light yield is measured, then the Area Under
Curve and Decay Time VI would be run separately to determine the amount of light
produced at impact and the duration of the light. However, it should be noted that
these two programs are independent of each other, and it will not matter which
program is run first or at all. The only thing that matters is what particular property
the user wants to analyze [37].
If the user wants to measure the spectrum of a drop using the Ocean Optics
spectrometer, then Spectrum Finder and Spectrum Processor VIs must be run as
shown in Fig. 7.13. As the figure shows, Spectrum Finder must be run before using
graphing software. This is due to the fact that Spectrum Finder will tell the user the
file containing the spectrum they need and will not require importing hundreds of
files one after another to find and graph the needed file. Also, if the centroid location
is required, then Spectrum Processor can be run to determine the emission centroid
and light yield produced upon impact [37].
Fig. 7.14 Front panel for the LabVIEW Light Yield Finder VI [37]
Emission files were analyzed using a custom LabVIEW program called Light
Yield Finder as shown in Fig. 7.14. This application is used to calculate the light
emission yield by measuring the area under the detector output signal versus time
curve. Running the application begins with the user setting the Amplifier Correction,
which takes into account the varying amplifier settings required to record a signal.
By setting the detector amplifier to 20 μA instead of 200 μA, the user can record a TL
flash that is ten times dimmer. Dividing the resulting light yield by ten, will allow
for the comparison of ZnS:Mn set at 200 μA and ZnS:Cu set at 20 μA [37].
When executed, the VI asks the user for the location of the TL emission data files
in comma space delimited (CSV) format. After selecting the folder, the program
calculates the area under the curve using the Uneven Numerical Integration virtual
instrument, which is standard component of the LabVIEW software package. Due
to the fact that LabVIEW only writes text files to the thousandths place, the
resulting relative light yield was multiplied by 107. The results are then written to
a text file set by the user. Once all the files have been analyzed, the mean and
standard deviation of the mean is calculated and written to the last line of the
file [37].
In order to calculate the triboluminescent decay time, a specially written
LabVIEW VI known as Decay Time was developed. The front panel for Decay
Time (version 7.0) VI is shown in Fig. 7.15. It reads a file and formats the data into
time in seconds versus light yield. This particular version was written specifically to
read files generated by a Tektronix 2024B oscilloscope.
To run this VI, the user must first set the location of where to save the data file.
The VI then begins by asking the user which file they would like to analyze. Unlike
the previous virtual instruments, the user must sequentially analyze each data file.
Some materials such as ZnS:Mn,Cu or ZnS:Cu,Pb,Mn have multiple decay times,
which makes it difficult to calculate anything without intervention.
Fig. 7.15 Front panel display picture for the LabVIEW Decay Time VI [37]
After the data file is selected, the user moves two adjustable cursors to the region
of interest as shown in Fig. 7.15. The positions of the lower and upper bounds of the
fit are fixed by pressing the Set Cursor Ok button. This button push generates an
exponential fit results graph in Arrhenius form as shown in Fig. 7.15. The user can
then change the fit method (least squares is default) if it is does not match the data.
In order to estimate accuracy, the R-squared value, decay constant (ms1), and
decay time (ms) is displayed on the front panel. If the fit is accurate, the user can
press the Fit Complete button. Then the user will be asked if they would like to save
the data, defined fit region, exponential fit, decay time constant (ms1), calculated
decay time (ms), and R-squared value. If the fit curve does not accurately match the
data, then this procedure can be repeated using different values for the lower and
upper bounds [37].
Usually TL spectra are collected using an integration time of 100 ms. In order to
ensure that useful TL spectral data is recorded, hundreds of separate files are
recorded. To maximize efficiency, a specialized LabVIEW VI known as Spectrum
Finder was developed as shown in Fig. 7.16. This VI is used to locate the
SpectraSuite-generated files containing the sought after TL spectra. The VI begins
by asking the user the location of the folder containing the SpectraSuite data files.
Spectrum Finder then reads the files and shows the file number and displays the data
graphically as shown in Fig. 7.16. Then the user scrolls through the files by pushing
the up or down arrows on the keyboard as shown in Fig. 7.16. Once the tribolumi-
nescent spectrum is located, the file number is recorded, and the text file is opened
in SigmaPlot® to plot the spectral emission [37].
7.4.3 Photoluminescence
In addition to measuring the TL, the photoluminescence (PL) of each material was
determined. This allowed for the determination of the cause of luminescence and
identifies the luminescent center. The europium complexes have an emission
Fig. 7.16 Front panel display picture for the LabVIEW Spectrum Finder VI [37]
mechanism consisting of: the ligand absorbing energy, undergoing intersystem

crossing into a triplet state, and then transferring its energy to the trivalent lantha-
nide ion [83]. The phosphorescence of the majority of the materials was detected by
excitation using an UV transilluminator manufactured by UVP, Inc. The
photoluminescence emission spectra were recorded using an Avantes AvaSpec
2048 USB2 spectrometer, which has a wavelength range of 445–680 nm and a
resolution of 0.13 nm. The integration time was controlled using the AvaSoft
software program. The integration time was increased by the software until the
photoluminescent emission spectrum reached its maximum that was just below the
saturation limit of the spectrometer or until the background could no longer be
completely subtracted due to the large integration time. Once this was determined,
one hundred (100) spectra )were recorded and averaged.
7.5 Results and Discussion
7.5.1 Synthesis Using a New Recipe for Europium Tetrakis
The problem with the synthesis method described by Hurt et al. is that they use
anhydrous europium chloride. Anhydrous europium (III) chloride is very expen-
sive. In addition, chlorine quenches the luminescence. If EuD4TEA will be used as
an impact sensor, then the synthesis cost must be reduced and the synthesis process
must be simplified. As a result, in 2011, Fontenot et al. developed a new synthesis
method for EuD4TEA [5]. The europium (III) chloride was replaced with 99.999 %
Fig. 7.17 Comparison of the TL signal generated from one drop on EuD4TEA made with
europium chloride and nitrate [5]
europium (III) nitrate, purchased from Metall Rare Earth Limited. The synthesis of
the EuD4TEA was carried out using a 2 mmol ratio of Europium nitrate to 8 mmol
each of DBM and TEA in 25 mL of ethanol. The mixture was heated to obtain a
clear solution and kept aside to crystallize by slow evaporation under ambient
conditions.
The results from the batches of europium chloride and nitrate indicated that the
europium nitrate produces 82 % more TL compared to europium chloride, as shown
in Fig. 7.17. In addition, the new method produced more material due to the lack of
washing required and took less time. As a result, the cost of synthesizing EuD4TEA
is nearly cut in half making it a great first step for its use in sensors [5].
7.5.1.1 Comparison of TL Light Yield for EuD4TEA
EuD4TEA is a relatively new material with very little research. As a result, it was
decided to compare this material and 26 other materials to manganese doped zinc
sulfide (ZnS:Mn). After performing 5 drops on each material, the relative tribolu-
minescent emission yield ratios for the 27 tested luminescent materials are shown in
Table 7.3. This ratio was calculated for each material and mass by dividing the
average triboluminescent emission yield by the equivalent value for ZnS:Mn (lot
15248), which is the 7.5 μm ZnS:Mn and has been extensively tested since 2003
[15, 36, 65, 67–71, 84, 85]. The results shown in Table 7.3 have an error estimated
to be less than 9 % [36].
Table 7.3 Comparison of the triboluminescent yields for the 27 luminescent materials as mea-
sured using the drop tower [36]
Sample Lot Grain Yield
Base material mass (g) number Manufacturer size (μm) ratioa
ZnS:Mn 1 1 Center for Integrated 0.005 0.000
Nanotechnologiesb
1/0.1 15248 Phosphor Technology 7.5 1.000/
1.000a
1 17112 Phosphor Technology 8.5 1.223
ZnS:Cu 1 19017 Phosphor Technology 2.9 0.005
ZnS:Mn,Cu 1 19010 Phosphor Technology 21.9 1.130
1 20267 Phosphor Technology 1.585
ZnS:Cu,Pb 1 15027 Phosphor Technology 19.0 0.034
ZnS:Cu,Pb,Mn 1 17002 Phosphor Technology 19.3 1.017
MgF2:Mn 1 09147 Phosphor Technology 0.029
La2O2S:Eu 1 10185 Phosphor Technology 0.004
Y2O2S:Eu 1 19145 Phosphor Technology 0.000
EuD4TEA 0.1 None Sandia National 0.960
Laboratoriesc
0.1 3 Alabama A&M 2.063
University
EuD4TEA þ 1.25 mL 0.1 10 Alabama A&M 3.196
DMMP University
a
Ratio based on the TL light yield for both the 1 and 0.1 g samples of 7.5 μm ZnS:Mn (LM-34) set
equal to 1.000
b
The Center for Integrated Nanotechnologies (CINT) is located in Albuquerque, New Mexico
c
Sandia National Laboratories is located in Livermore, CA
The average triboluminescent yield for the nitrate based EuD4TEA was 2.06
times larger than what was measured for the baseline 7.5 μm ZnS:Mn powder. In
fact, the EuD4TEA materials were the brightest materials tested. However, as
EuD4TEA is an organic compound, it was much more sensitive to damage com-
pared to the inorganic ZnS:Mn as shown in Fig. 7.18. The TL of EuD4TEA
decreased exponentially very quickly. In fact, the EuD4TEA crystals had most of
80000 20
18
Maximum PMT Output (V)

16
60000
14
Light Yield
12
40000
10
8
20000
6
4
0 2
0 2 4 6 8 10 12 0 10 20 30 40
Number of Drops Number of Drop
Fig. 7.18 Plot of triboluminescent yield as a function of number of drops for (a) EuD4TEA and
(b) ZnS:Mn [86]
their TL extinguished after approximately 10 drops. ZnS:Mn, on the other hand, can
withstand over 40 drops, though still decreasing exponentially as shown in
Fig. 7.18b [15].
Table 7.3 shows the EuD4TEA sample doped with DMMP has the largest tested
triboluminescent yield for any of the tested materials. The EuD4TEA doped with
DMMP had more than 3.19 times the triboluminescent yield compared to the
baseline 7.5 μm ZnS:Mn. Conversely, the 19.8 μm ZnS:Mn emitted the most TL
of any of the tested inorganic materials. It is likely that the method used to prepare
ZnS:Mn (lot 09029) in terms of grain size, trap concentration, and dopant concen-
tration, was responsible for the increased triboluminescent yield. On average, it
appears that ZnS:Mn,Cu has a larger triboluminescent yield compared to ZnS:Mn.
As Fig. 7.18a shows, taking the average of the first five drops will not provide
reliable data if we want to determine the effect of impact energy on tribolumines-
cent emission. The reason was due to the fact that a majority of the EuD4TEA TL
was extinguished after five drops. As a result, the TL yield becomes not a result of
the drop height, but instead a function of the material itself. Interestingly, it appears
as though the total TL light produced over all drops and heights remains relatively
constant with the average total TL light output being 33,736.54 1274.869 a.u. The
error was estimated using the standard deviation of the mean of all the average
values [86].
In order to determine how the TL light yield changes with impact energy, only
the first two values of each drop height were averaged for the EuD4TEA and the first
five drops were averaged for ZnS:Mn samples. In addition, the TL change between
the two EuD4TEA drops was smaller and more comparable to the change exhibited
by ZnS:Mn over five drops.
The results from this test are shown in Fig. 7.19. The data presented in Fig. 7.19
was normalized to the 8.5 μm Zns:Mn triboluminescent yield collected from a drop
height of 6 in. (152 mm). A 6 in. drop corresponds to an impact energy of 0.1944 J.
The experimental uncertainty was estimated to be 7 % for the light yield and 5 % for
the impact energy. The ZnS:Mn triboluminescent results measured here are
completely consistent with data collected by Bergeron and Hollerman
Fig. 7.19 Plot of triboluminescent light yield as a function of impact energy for EuD4TEA and
ZnS:Mn [86]
et al. [65, 69–71]. The triboluminescent data for ZnS:Mn exhibits a region where
the emission yield experiences a rapid increase from 0.06 to 0.19 J, which corre-
sponds to drop heights of 2–6 in. (51–152 mm). Due to experimental uncertainty,
the triboluminescent yield does not appear to be a function of ZnS:Mn grain size for
energies less than or equal to 0.19 J. Above 0.19 J, the triboluminescent emission
yield continues to increase; however, it is at a much slower rate. The reason for this
could be that for the lower energy impacts, there is not enough energy to break all
the grains of powder. However, above 0.19 J, there is sufficient energy to break all
the grains in the area of impact. For impacts with energies above 0.8 J, the TL yield
for 8.5 μm ZnS:Mn powder is statistically larger than the equivalent data for the
24.1 and 30 μm samples. It is possible that the methods used to manufacture these
samples could play a significant role in the triboluminescent yield [86].
Results from Fig. 7.19 also indicate the triboluminescent yield from impacts
with the EuD4TEA samples are significantly brighter than those observed from the
ZnS:Mn powders over any of the tested kinetic energies. In addition, the increase in
TL does not appear to be linear, but instead appear to be directly proportional to the
impact energy, E1/3 [86].
To determine the minimum amount of energy needed to produce TL, the 0.13 kg
steel drop ball was replaced with a 10 g wooden ball (made for a drawer handle) and
1 g styrofoam ball. Both of these new balls had similar diameters as the original
0.13 kg steel bearing. Using balls of smaller mass allowed the impact energy to
decrease proportionally. The styrofoam ball dropped from a height of 2 in. had the
available kinetic energy of 5 104 J (500 μJ) at impact. By contrast, the wooden
ball had an available kinetic energy of 5 103 J (5 mJ) at impact [86].
When tested with the wooden ball dropped from a height of 2 in., all three ZnS:
Mn and the one EuD4TEA samples produced a strong triboluminescent signal. By
using the styrofoam ball dropped from a height of 2 in., both the 8.5 and 30 μm ZnS:
Mn samples produced small amounts of detectable light. As a result, it can be safely
assumed that the realistic limit for the 8.5 and 30 μm grain sized ZnS:Mn is 500 μJ.
It is possible that the detection kinetic energy might be lower if more expensive
photomultiplier tubes were used to measure the emitted light in complete
darkness [86].
Interestingly, the 24.1 μm ZnS:Mn sample did not produce light when the
styrofoam ball was dropped below a distance of 12 in. (305 mm), which corre-
sponds to a kinetic energy of about 3.0 mJ. Light was also observed when the ball
was dropped from a height of 18 in. (kinetic energy of 4.5 mJ). No other measure-
ments were attempted between 12 and 18 in. Therefore, light is produced when a
minimum of 3.0–4.5 mJ of energy is applied to the 24.1 μm ZnS:Mn powder. This
result is likely caused by oxygen defects being added to the 24.1 μm powder during
synthesis. Additional research is needed to understand this phenomenon [86].
The most intriguing result is the fact that EuD4TEA appears to have an activation
energy that is significantly less than was measured for 8.5 μm ZnS:Mn (500 μJ). A
strong light emission was observed when a 1 g styrofoam ball was dropped over a
distance of 2 in. on the EuD4TEA powder. In fact, when the amplifier gain was
reduced by a factor of 10, the resulting measured light yield from EuD4TEA was
found to be similar to what was measured for ZnS:Mn. From this observation it can
be concluded that the activation energy for EuD4TEA is approximately ten times
less, or 50 μJ, than what was measured for ZnS:Mn [86].
The measured triboluminescent emission spectra for are shown in Fig. 7.20.
Notice that both emission peaks are identical in wavelength location but not
intensity. The spectra indicates that the luminescence from EuD4TEA comes
from the typical Eu3þ centered transitions from the 5D0 levels to the lower 7F0–4
levels [87] as shown in Fig. 7.21. Due to the high resolution of the spectrometer,
multiple peaks were observed for each transition Due as shown in Table 7.4. The
main emission occurred from the 5D0 ! 7F2 transition. In addition, the 5D0 ! 7F4
transition was not observed due to the transition occurring around 705 nm, which is
outside the limit of the Avantes spectrometer. This peak has been observed using an
Ocean Optics S2000 spectrometer with a larger wavelength range and is shown by
Fontenot et al. [88].
The measured photoluminescent emission spectra for all three grain sizes are
shown in Fig. 7.22. The spectra indicates that the luminescence from ZnS:Mn
comes from the Mn2þ centered transitions from the 6D9/2 level to the lower 6S5/2
level [87] as shown in Fig. 7.21. Statistically the photoluminescent emission peak
for each grain size of ZnS:Mn are the same. The average emission peak for all three
grain sizes was 585 1 nm. Due to the limits of the experimental setup, a sufficient
amount of light was not generated to measure the triboluminescent emission spectra
from each ZnS:Mn material using the Avantes spectrometer.
Fig. 7.20 Measured EuD4TEA photoluminescent and triboluminescent spectra [86]
Fig. 7.21 Energy level diagrams for Eu3þ and Mn2þ [86]
Table 7.4 Photoluminescent 5

D0 ! 7F1 5
D0 ! 7F2 5
D0 ! 7F3
emission peak wavelengths
578.8 611.7 652.1
(nm) for EuD4TEA
591.7 613.5 653.7
594.9 616.5
597.8
Fig. 7.22 ZnS:Mn photoluminescent spectra for various grain sizes [86]
7.5.1.2 Comparison of TL Decay Times for EuD4TEA and ZnS:Mn
Using decay time software as developed by Fontenot et al. [37], the decay times
were determined for each drop. From the data as shown in Table 7.5, it appears as
though the impact energy does not change the TL decay time. The average
EuD4TEA triboluminescent decay time over all drops was 584.151 35.736 μs.
The average 8.5 μm ZnS:Mn decay time was 297.7 36.5 μs. The average 24.1 μm
ZnS:Mn decay time was 272.8 50.4 μs, and the average 30 μm ZnS:Mn decay
time was 310.7 51.1 μs which matches previous results by Hollerman et al. The
error for each was estimated using the standard deviation of all the drops [86].
7.5.2 Estimating the Batch Synthesis and Drop Tower Error
If EuD4TEA will be used in sensors, then the synthesis batch error must be
determined. In order to estimate the error between batches of EuD4TEA and the
Table 7.5 Decay times (1/e) as a function of impact energy for EuD4TEA and ZnS:Mn [86]
Drop height Impact energy ZnS:Mn

(in) (J) EuD4TEA 8.5 μm 24.1 μm 30 μm
2 0.0648 576.6 2.8 293.2 6.3 264.8 29.9 263.9 22.6
3 0.0972 590.5 9.9 286.4 7.4 295.0 40.6 301.3 23.2
4 0.1296 616.5 19.4 334.8 12.1 237.3 15.1 337.5 23.2
5 0.1620 593.4 9.9 286.3 7.9 227.3 15.6 316.9 9.0
6 0.1944 594.2 10.0 323.3 14.2 274.4 43.7 315.2 32.6
12 0.3887 631.2 19.5 327.6 18.6 308.1 11.5 322.3 12.3
18 0.5831 545.7 7.8 295.4 10.2 288.9 9.9 344.8 27.6
24 0.7774 553.8 10.1 271.9 9.4 270.7 15.4 294.8 8.4
30 0.9718 563.4 7.7 284.0 19.0 315.0 32.1 286.4 9.8
36 1.1661 601.9 17.0 311.2 18.0 268.4 6.6 348.1 30.2
39 1.2633 562.4 15.0 285.4 29.1 245.9 7.3 313.0 30.0
42 1.3605 580.2 13.3 273.2 10.4 279.4 16.0 284.3 16.5
Table 7.6 Relative triboluminescent light yields (Yrel) for EuD4TEA synthesized with ethyl
alcohol and acetone [89]
Yrel for EuD4TEA made with two solvents
Trial batch number Ethyl alcohol Acetone
1 1.000 1.000
2 1.083 0.966
3 1.282 1.276
4 1.238 1.365
5 0.896 1.127
Average 1.100 0.072 1.147 0.077
triboluminescent light yield from the drop tower, five batches of EuD4TEA were
synthesized using acetone and ethyl alcohol solvents. These samples were made
over many months using different batches of 99.999 % europium (III) nitrate
hexahydrate from Metall Rare Earth, Limited, DBM, TEA, laboratory grade ace-
tone, and pure ethyl alcohol from Sigma-Aldrich [89].
Table 7.6 shows the measured relative triboluminescent emission yields for the
five EuD4TEA trial batches synthesized with both acetone and pure ethyl alcohol.
The relative triboluminescent emission yield (Yrel) for a given trial batch was
calculated by:
Yt
Y rel ¼ ð7:6Þ
Y1
where Yt and Y1 are the average triboluminescent emission yield measured over five
drops for a given trial batch and the trial batch number one respectively. The
individual Yrel values shown in Table 7.6 are normalized to the average yields
measured for the first trial batch of EuD4TEA synthesized with acetone and ethyl
alcohol. In other words, the relative triboluminescent emission yield for the
first trial batches of EuD4TEA synthesized with acetone and ethyl alcohol were
set to 1 [89].
Table 7.6 also shows the average relative triboluminescent yield and its uncer-
tainty calculated over the five trial batches. This uncertainty was estimated using
the standard deviation of the mean from the five individual batch results. The
average relative triboluminescent yield over the five batches of EuD4TEA synthe-
sized with ethyl alcohol was 1.10 0.07. The corresponding relative yield for five
batches of EuD4TEA synthesized with acetone was 1.15 0.08. As a result, there
appears to be less than a 7 % measured batch-to-batch variance in triboluminescent
yield for EuD4TEA. The data also showed that EuD4TEA samples synthesized with
acetone emit slightly more TL than those made with ethyl alcohol. It should be
noted that this 7 % batch-to-batch yield variance also includes the measurement
uncertainty of the drop tower [89].
To determine the measurement uncertainty of the drop tower, five additional TL
drops were recorded using EuD4TEA synthesized in trial batch number two with
ethyl alcohol. Once each TL measurement was completed, the drop plate was
cleaned and a new 0.1 g pile of EuD4TEA synthesized during trial batch number
two was placed in the center. A total of five drops, each taken with a new 0.1 g
sample of the trial batch number two powder, were completed. The average of the
relative triboluminescent emission yield (Yrel) for the trial batch number two drops
was 0.99 0.03. The error in these measurements was estimated using the standard
deviation of the mean taken over the five drops [89].
During the initial part of this research as shown in Table 7.6 Yrel for trial batch
number two was measured to be 1.08, which is about 9 % larger than the 0.99
average value measured here. For this reason, the measurement uncertainty for the
drop tower appears to have a maximum variance of about 9 %, which was similar in
magnitude to the 7 % batch-to-batch yield variance for synthesized EuD4TEA. For
this reason, it can be stated that no statistically significant batch-to-batch variations
in the triboluminescent emission yield were observed for the synthesized
EuD4TEA [89].
7.5.3 Effects of Solvents
If the cost of EuD4TEA is to be reduced, then the solvents used during synthesis
must also be inexpensive. Pure ethyl alcohol is expensive so other solvents must be
investigated to reduce the cost even further. However, precipitation of tribolumi-
nescent materials from solvents may introduce imperfections to the crystal struc-
ture, thus affecting the emission yield. In 1988, Sweeting and Rheingold found that
europium tetrakis dibenzoylmethide recrystallized from dichloromethane did not
initially produce TL when physically excited [90]. However, if the europium
tetrakis dibenzoylmethide was heated for 18 h at 50 C or was allowed to stand at
~20 C for several months, TL was observed when the sample was excited. During
their research, it was also found that the crystal structure for europium tetrakis
dibenzoylmethide recrystallized from methyl alcohol and dichloromethane was
different when this material was prepared using other solvents [90].
In 1992, Hocking et al. found that triphenylphosphine recrystallized from ethyl
alcohol, methyl alcohol, and carbonyl-based solvents produced TL when excited
[91]. Their research also showed that triphenylphosphine recrystallized from aryl
solvents such as benzene and cyclic solvents such as cyclohexane produced no TL
when excited [91].
In order to study the effect of solvents on the TL property of EuD4TEA
compounds, solvents were chosen to provide a wide range of dielectric properties
as well as different functional groups such as alcohols, ketones, aryl and alkyl
solvents to name a few. The synthesis of the europium tetrakis compound was
carried out by the modified method by replacing ethanol with the various solvents
as shown in Table 7.7. The products obtained were cooled, filtered and allowed to
dry at ambient temperature [92].
The rate of dissolution was highest in the solvents acetone and acetonitrile as
they required no heating. Interestingly, while the acetonitrile showed the same rate
of dissolution for the europium salt and DBM, the EuD4TEA crystals formed
instantly. In fact, complete precipitation occurred almost immediately once the
TEA was added and the solution was mixed. However, the EuD4TEA crystals that
precipitated were smaller and had a lighter yellow color compared to the crystals
synthesized from the rest of the solvents. Synthesis carried out using acetone as a
solvent produced a small crystalline product with sizes of approximately a few
millimeters as shown in Fig. 7.23a and large crystals that are a few centimeters as
shown in Fig. 7.23b. The larger crystals were obtained by setting aside the mother
liquor after the initial filtration of the small product formed over a few days [92].
EuD4TEA crystals did not precipitate from three of the four carbonyl containing
solvents, methyl ethyl ketone (MEK), propionaldehyde, and cyclohexanone. Only
samples synthesized with the carbonyl solvent ethyl acetate produced crystals. The
aliphatic solvents, octane, hexane, and cyclohexane did not dissolve the europium
even when the solvent volume was tripled and the temperature was increased to the
boiling point. However, when the DMB and TEA was added, a yellow paste was
obtained. This paste produced a bright photoluminescence under ultraviolet irradi-
ation; however, no TL was observed when the sample was physically excited. It
should be noted that triboluminescent measurements were recorded a week after
synthesis of the crystalline products and a few months when the paste dried to the
consistency of hard clay [92].
Alcoholic solvents required heating during the synthesis of EuD4TEA. Com-
pared to ethyl alcohol, methyl alcohol quickens the formation of EuD4TEA yielding
a sparkling crystalline structure. In addition, the EuD4TEA crystals synthesized
from methyl alcohol were larger in size than that obtained from the solvent ethyl
alcohol. There is no apparent measurable size difference between the EuD4TEA
crystals produced when using either the 200 proof or 95 % denatured ethyl alcohol
Table 7.7 A description of the solvents used in the synthesis of EuD4TEA samples in this
research [92]
CAS
Solvent Grade Manufacturer Product number number TL?
Carbonyl containing solvents
Acetone Laboratory Sigma-Aldrich 179973 67-64-1 Yes
reagent 99.5 %
Certified ACS Fisher A184 Yes
Cyclohexanone Baker analyzed J.T. Baker 9210 108-94-1 No
reagent
Ethyl acetate Certified ACS Fisher E145 141-78-6 Yes
Alcoholic solvents
Methyl alcohol Anhydrous Mallinckrodt 3016 67-56-1 Yes
Ethyl alcohol 200 proof Sigma-Aldrich 459836 64-17-5 Yes
anhydrous
Laboratory 95 % Scholar 9506406 Yes
denatured Chemistry
Propyl alcohol GR grade Matheson PX1815 71-23-8 Yes
Coleman and
Bell
1-butanol Certified ACS Fisher A399 71-36-3 Yes
3-pentanol 98 % Sigma-Aldrich P8025 584-02-1 Yes
Miscellaneous solvents
Acetonitrile HPLC Sigma-Aldrich 270717 75-05-8 Yes
Grade 99.93 %
Ethyl ether HRGC/HLPC— Pharmco 37300DISC54L 60-29-7 No
Trace Grade
Benzene Certified ACS Fisher – 71-43-2 Yes
Chloroform Certified ACS Fisher C298 67-66-3 Yes
Methylene Certified ACS Fisher D37 75-09-2 Yes
chloride
Squalane 99 % Sigma-Aldrich 234311 111-01-3 No
Oleic acid Technical 90 % Alfa Aesar A16663 112-80-1 No
Octane Reagent Fisher O39801 111-65-9 No
Hexane ReagentPlus 99 % Sigma-Aldrich 139386 110-54-3 No
Cyclohexane ACS reagent Sigma-Aldrich 179191 110-82-7 No
solvent. However, on changing the solvent to 1-butanol, the structure of the

EuD4TEA changed, making it larger, more flaky and clumpy crystals [92].
Interestingly, the methylene chloride and benzene solvents did not dissolve the
europium salt. However, the addition of DBM and TEA, appeared to dissolve the
europium salt. The use of a benzene solvent produced solid crystals. When meth-
ylene chloride was used as a solvent, crystals with an elongated structure were
formed as shown in Fig. 7.24. On the other hand, when chloroform was used as the
starting solvent solid crystals were produced that was very sticky and paste-like.
Fig. 7.23 Close-up pictures of the EuD4TEA crystals produced with an acetone as solvent during
the synthesis and crystallization process that are: (a) a few millimeters and (b) centimeters in
size [92]
Fig. 7.24 Picture showing

the elongated structure of
the crystals produced with
methylene chloride
solvent [92]
Crystals obtained from the oleic acid solvent were transparent white, square in
shape, and produced no TL [92].
A comparison of the average normalized triboluminescent emission yields for
the compounds from different solvents are shown in Fig. 7.25. The tribolumines-
cent emission yield for EuD4TEA synthesized using the anhydrous ethanol is set to
one. The uncertainty in triboluminescent yield was estimated to be 7 %, which is the
standard deviation of the mean for the emission data if the sample was synthesized
again using the same procedure [89]. These 7 % error bars are shown in Fig. 7.25.
Results indicate that acetone was useful for the synthesis of EuD4TEA since it
was inexpensive to procure and can precipitate EuD4TEA in only a few days. In
addition, the large crystals that formed using the acetone produced the most TL of
any of the listed solvents. This result was completely consistent with the notion that
grain boundaries between many small crystallites attenuate the TL more compared
to samples with few boundaries between large crystals. Effectively, the size of the
precipitated crystals appears to be an important consideration when designing any
TL-based impact sensor system [92].
Fig. 7.25 Comparison of the average triboluminescent yield emitted by each material using
different solvents [92]
Further, triboluminescent emission from the small crystal of EuD4TEA synthe-

sized using methyl alcohol and acetone was found to emit about half as much TL
compared to the large crystal samples synthesized with acetone. It is not clear from
these results what parameters in the procedure needs to be controlled in order to
maximize crystal size and triboluminescent yield. Additional research is needed to
develop procedures to maximize EuD4TEA crystal size, which results in the largest
possible triboluminescent emission [92].
Interestingly, although the chloroform and methylene chloride samples pro-
duced solid crystals, the relative triboluminescent yield was small. This result
was likely due to the chlorine present in both solvents. The presence of chlorine
in a sample typically quenches the triboluminescent yield of EuD4TEA. When
solids were added to squalane, the rate of dissolution was very small. This does not
include the tested toluene and ethylene glycol solvents which did not dissolve any
of the DBM or europium nitrate. It took 35 mL of squalane to dissolve all the DBM
and europium. However, once it was heated and dissolved, two separate layers
formed. The first layer was a yellow layer paste stuck to the bottom of the beaker.
On top of this was a white wet crystalline product. After 3 months of air drying, this
crystalline product was still wet. As a result of this and the two layers formation, it
was not tested for TL. The crystals that formed with oleic acid were a transparent
white and square in shape. After testing several of these crystals, none exhibited
TL [92].
The results showing the average decay time data over five drops are shown in
Table 7.8. Results show that the measured decay times appear to depend more on
crystal size than on the type of solvent used to make the samples. There was no
statistical difference between the decay times measured for small EuD4TEA crys-
tals made with any type of acetone or alcohol [92].
Table 7.8 Triboluminescent Solvent Average (μs)

decay time (1/e) results for
Fisher acetone large crystal 565.9 14.2
EuD4TEA made with a
selection of solvents [92] Fisher acetone small crystal 545.2 11.6
200 proof ethyl alcohol 544.8 7.4
95 % denatured ethyl alcohol 549.0 15.9
Methyl alcohol 535.1 9.2
Laboratory grade acetone large crystal 594.3 13.9
Laboratory grade acetone small crystal 546.5 16.1
Ethyl acetate 523.6 5.9
Acetonitrile 468.8 15.5
Benzene 527.6 22.0
Chloroform 372.7 19.7
Propyl alcohol 516.0 11.5
1-butanol 534.5 20.6
3-pentanol 493.5 7.0
Methylene chloride 485.9 23.6
Fig. 7.26 Photoluminescent emission spectra from each material synthesized using the solvents
shown in Table 7.7 [88]
The measured photoluminescent emission spectra for each EuD4TEA sample is

shown three dimensionally in Fig. 7.26. The data shown in Fig. 7.26 is typical of the
Eu3þ centered transitions from the 5D0 levels to the lower 7F0–4 levels. The results
shown in Fig. 7.26 indicate that most of the samples synthesized using the solvents
shown in Table 7.7 have an emission spectrum that is consistent with what has been
previously reported for EuD4TEA. In two cases, however, the intensity of 611.7 and
613.5 nm peaks were different. The intensity of the 611.7 nm peak was much higher
than the 613.5 nm peak from all the solvents except 1-butanol and methylene
chloride. In the case of the 1-butanol and methylene chloride, the 613.5 nm peak
was larger than the 611.7 nm peak, and the 616.5 nm peak was also more intense. In
addition, the 1-butanol shows another distinctive peak located at 624.2 nm. This,
however, does not appear to be a new peak present in only this solvent. The other
solvents have what could be a peak located around 624.2 nm, however, it was not
intense enough to distinguish it from the background. The 1-butanol seems to be
amplifying this peak by possibly embedding some of the solvent into the crystal.
More research is needed to fully understand these phenomena [92].
7.5.4 Effects of Grain Size
To determine if the size of the EuD4TEA has any role in the triboluminescent yield,
the amount of acetone solvent was varied to change the size of the crystals. The
synthesis process was the same except this time the amount of acetone solvent was
varied from 10 to 40 mL in 5 mL increments. Pictures of the resulting EuD4TEA
crystal samples are shown in Fig. 7.27a–f. Note the crystals appear to be monoclinic
in shape [93]. This finding is consistent with earlier research results as reported by
Sweeting and Rheingold [77].
Initially, an attempt was made to synthesize EuD4TEA with 5 mL of carrier
acetone. However, there was not enough liquid solvent present to dissolve the
europium and DBM. As a result, the amount of carrier acetone was increased
Fig. 7.27 Close-up pictures of synthesized yellow EuD4TEA crystals using: (a) 10, (b) 15, (c)
20, (d) 30, (e) 35, and (f) 40 mL of carrier acetone. Note the crystal grain size is proportional to
acetone carrier volume [93]
Table 7.9 Comparison of US Standard opening size

US standard mesh
standard sieve number and
sieve number in. mm
standard opening sizes for
several common meshes [94] 120 0.0049 0.125
80 0.0070 0.177
60 0.0098 0.250
45 0.0139 0.355
35 0.0197 0.500
25 0.0278 0.710
slowly until everything was dissolved. It appears that the minimum amount of
solvent required to synthesize EuD4TEA using this method is 10 mL. Small grains
of EuD4TEA precipitated almost immediately when 10 mL of carrier acetone was
used in this process. The crystals were light yellow in color as shown in Fig. 7.27a.
As the quantity of carrier acetone was increased beyond 10 mL, the synthesized
EuD4TEA crystals were darker yellow in color. As shown in Fig. 7.27b–f, the grain
size for the precipitated EuD4TEA crystals also appears to be proportional to the
volume of carrier acetone that is used in the process. Synthesizing larger crystals of
EuD4TEA is possible, but is not probable based on the quantities of required
precursor chemicals [93].
To determine the crystal size, the assorted EuD4TEA samples were sifted using
120, 80, 60, 45, 35, and 25 US standard mesh sieves [94]. Table 7.9 shows the
relationship between US standard sieve number and the mesh opening size
expressed in both inches and millimeters. For example, a crystal that has a maxi-
mum size of 0.30 mm would pass through a number 45 US standard mesh sieve
(0.355 mm opening size), but would be caught and not passed through a number
60 standard mesh sieve (0.250 mm). Based on earlier research [77], it was likely the
crystal structure for these EuD4TEA samples was monoclinic. Therefore, the
maximum “grain size” that would pass through a sieve opening was likely equiv-
alent to a “body diagonal” for the monoclinic system and would be somewhat larger
than the actual crystallographic dimensions for the EuD4TEA samples.
After all the samples were synthesized, the resulting EuD4TEA crystallites of
different sizes were mixed together and the entire sample was sifted using the
number 25 sieve. This process caused all EuD4TEA crystallites smaller than
0.710 mm to pass through the number 25 sieve and into a clean container. The
crystallites larger than 0.710 mm were left on the sieve and were removed. The
pass-through crystallites were then sifted again with the number 35 sieve, causing
the particles smaller than 0.500 mm to be separated. Particles with sizes larger than
0.500 mm and smaller than 0.710 mm were left on the number 35 sieve. This
process was repeated for the rest of the sieves shown in Table 7.9.
In all cases, the EuD4TEA crystallites left on sieves 45, 60, 80, and 120 had sizes
between 0.355 and 0.500, 0.250 and 0.355, 0.177 and 0.250, and 0.125 and
0.177 mm respectively. This sifting process generated five samples ranked by
grain size. Due to the small quantity of size-sorted crystallites, 0.030 g of EuD4TEA
Table 7.10 Measured triboluminescent yield results for EuD4TEA samples sorted by grain
size [93]
EuD4TEA grain size limits (mm) Average grain size (mm) Normalized TL yield
0.125–0.177 0.151 0.026 0.576 0.040
0.177–0.250 0.213 0.037 0.817 0.057
0.250–0.355 0.303 0.052 1.000 0.070
0.355–0.500 0.427 0.073 1.544 0.108
0.500–0.710 0.605 0.105 2.743 0.391
Fig. 7.28 Plot of the normalized TL yield versus the grain size for the tested EuD4TEA samples
[93]
was used for each TL drop trial. In addition, three drops were completed on each
size-sorted EuD4TEA sample to give an average TL emission yield. The tribolu-
minescent emission yield for each size bin was normalized to the 0.250–0.355 mm
value [93].
Table 7.10 shows the results of this analysis. The normalized triboluminescent
yield appears to increase as a function of the average grain size. The measurement
uncertainties for the triboluminescent emission yields were estimated to be 7 %,
based on earlier research [89]. The estimated uncertainties shown in the square of
the average grain size from Table 7.10 were calculated using standard error
propagation techniques [95].
A plot of the normalized triboluminescent yield versus the average grain size for
EuD4TEA is shown in Fig. 7.28. It appears that the triboluminescent yield was
directly proportional to the square of the average grain size as shown by the square
fit [93].
What does this mean? It is logical to assume that the number of luminescent
centers that generate TL should be proportional to the average effective surface
area of the individual EuD4TEA crystallites. During these measurements, each
Table 7.11 Fluorescence Amount of acetone (mL) Average (μs)

decay times (e1) as a
10 516.6 18.0
function of acetone volume
used during synthesis of 15 548.6 21.1
EuD4TEA [93] 20 518.3 13.0
25 546.5 16.1
30 518.1 15.2
35 527.4 13.0
40 566.0 13.9
size-sorted sample of EuD4TEA was spread evenly on the clear Plexiglass plate in
the drop tower. When the ball bearing was dropped, a relatively uniform layer of
EuD4TEA was available at the impact location to produce TL. Since each sample
was sorted for size, the “average effective surface area” for the individual
EuD4TEA crystallites was relatively constant. Since the exact crystallographic
information on each EuD4TEA crystallite is not yet known, it was not possible to
directly measure the surface area for all the grains in each sample. This process was
made more difficult by the fact that the crystallites are positioned randomly on the
drop tower plate and have no preferred orientation. However, the average surface
area for the crystallites in each sample should be proportional to the square of the
effective grain size. Remember this grain size was equivalent to the body diagonal
for the monoclinic EuD4TEA crystallites. Although Fig. 7.28 does not directly
show that the normalized triboluminescent yield for EuD4TEA is proportional to
the average effective surface area, it does strongly suggest this is likely the case.
The square of the data is remarkable, with a calculated goodness-of-fit or R2 value
equal to 0.997 [93].
The prompt fluorescence decay times for the acetone-synthesized EuD4TEA
samples were measured as shown in Table 7.11. From this data, it appears as though
increasing the volume of acetone used during synthesis does not change the
triboluminescent decay time for EuD4TEA [93]. Over all drops, the average
measured triboluminescent decay time for EuD4TEA was found to be
534.5 7.3 μs, which is consistent with earlier results [86, 88, 92, 96].
7.5.5 Synthesis of Lanthanide Compounds
To determine if other triboluminescent emission wavelengths (colors) could be

created for sensors, a series of lanthanide (Ln) compounds were synthesized. Ln
compounds have distinctive optical properties that include long luminescence
lifetimes that range from microsecond to milliseconds and sharp emission bands
with a full width half maximum that rarely exceeds 10 nm. These properties make
them useful for organic light emitting diodes, laser materials, and sensors [97]. To
increase luminescence excitation, the lanthanide ion is usually coordinated to the
ligands of β-diketone and aromatic amine derivatives. This typically increases the
quantum efficiency through “synergistic effects” and prevents the coordination of

solvent molecules that quench the emissions [97].
Besides their outstanding photoluminescent properties, lanthanide β-diketone
complexes containing chiral groups have additional properties that include second
harmonic generation, ferroelectricity, TL, chirality sensing, circularly polarized
luminescence, symmetric catalysis, and enantiomer selective synthesis [97].
The synthesis of each Ln dibenzoylmethide triethylammonium compound was
based on the procedures and methods used for EuD4TEA using 35 mL of anhydrous
denatured ethyl alcohol. Table 7.12 shows the color of the resulting solution or
product during each step of the synthesis process. Table 7.12 also shows the color of
the photoluminescent emission when the synthesized crystallites were exposed to
ultraviolet light. The Ho, Ce, and La compounds do not emit fluorescence when
excited with ultraviolet irradiation [81].
Table 7.12 Colors of the resulting solution or product during each step of the synthesis process
for a selection of lanthanide tetrakis dibenzoylmethide triethylammonium compounds [81]
Color of the solution after the
material below is added
Ln Ln Color and condition of Photoluminescence
compound nitrate DBM TEA the finished solid product color
Eu Clear Dark Dark Light yellow sparkly Bright red
yellow yellow crystalline
Ho Pink Light Dark Orange or yellow flakes None
orange orange
Dy Clear Clear Dark Crystalline sparkly yel- Green
yellow yellow low green
Er Light Peach Dark Flaky peach color Red
pink orange
Sm Clear Clear Dark Orange flaky crystalline Red
yellow yellow
Nd Clear Light Dark Pinkish red flaky Green
purple clear orange clumped crystals
yellow brown
Gd Clear Clear Dark Yellow-orange Red
yellow yellow crystalline
Tb Clear Clear Dark Yellow crystalline Green
yellow yellow
Pr Light Yellow Dark Small green flakes None
green green green
yellow
Yb Clear Light Dark Yellow hair-like clumped Green
yellow yellow particles
La Clear Clear Dark Pale yellow crystalline None
pale yellow
yellow
Ce Clear Dark Dark Black crystalline None
yellow blood red
All lanthanide elements form trivalent cations whose chemistry is largely deter-
mined by the ionic radius, which decreases steadily from lanthanum to lutetium. For
this reason, the notation LnD4TEA can be used to represent a generic lantha-
nide tetrakis dibenzoylmethide triethylammonium compound. For individual com-
pounds, Ln will be replaced with the symbol used for the specific lanthanide
element. For example, TbD4TEA is shorthand for the terbium-based tetrakis
dibenzoylmethide triethylammonium compound [81].
Observations as recorded in Table 7.12 indicated that EuD4TEA remained clear
when the europium nitrate was added, then turned dark yellow with the addition of
DBM and TEA. This synthesis created a light yellow crystalline structure that
sparkles. However, this was not true for all Ln compounds. The holmium nitrate
compound turned the ethanol solution pink. When DBM was added to this solution,
it turned light orange. When TEA was added, this solution turned dark orange.
Figure 7.29 shows the holmium compound sample has a flaky crystal structure and
exhibits a color that depends on the nature of ambient light. When viewed under
indoor fluorescent light, the holmium compound flakes are light orange in color as
shown in Fig. 7.29b. However, when the fluorescent lights are turned off and the
sample was exposed to standard daylight, the holmium compound crystallites were
light yellow as observed in Fig. 7.29a. The holmium compound crystallites emitted
no luminescence when excited.
The cerium tetrakis dibenzoylmethide triethylammonium compound also
showed some interesting properties. The cerium nitrate and ethanol solution was
completely clear. The addition of DMB turned this solution dark yellow; subse-
quently when TEA was added to complete the reaction, the solution became blood
red as shown in Fig. 7.30a. Figure 7.30b shows that small black crystallites are
produced during synthesis. The cerium compound crystallites emit no lumines-
cence when excited [81].
Results of the triboluminescent drop tower measurements for the synthesized
lanthanides are shown in Fig. 7.31. Since this was a relative measurement, the
resulting emission light yield for each compound was normalized to the previously
Fig. 7.29 Pictures of the flaky holmium tetrakis dibenzoylmethide triethylammonium compound
crystallites when illuminated by: (a) daylight and (b) fluorescent light [81]
Fig. 7.30 Pictures taken during the cerium tetrakis dibenzoylmethide triethylammonium com-
pound synthesis process showing the resultant: (a) blood red solution after the TEA was added, and
(b) black crystalline structure [81]
Fig. 7.31 Comparison of the average normalized triboluminescent light yield for the synthesized
Ln compounds. All the collected results were normalized to the EuD4TEA light yield [81]
measured EuD4TEA value. In other words, the relative light yield for EuD4TEA is
equal to one [81]. The error in emission yield was estimated to be 7 % which is the
combination of the uncertainties from material synthesis and drop tower operation
[89]. As shown in Fig. 7.31, it is evident that none of the tested Ln compounds
produce as much TL as EuD4TEA. The next largest triboluminescent yield was
emitted by the Sm compound, which is 1.8 % that measured for EuD4TEA [81].
Once the average light yields were determined, the triboluminescent decay times
for each compound were measured as shown in Table 7.13. It is evident that the
triboluminescent decay time was unique for each Ln compound. EuD4TEA had the
longest decay time while the Yb compound had the shortest decay time.
Table 7.13 Decay time (1/e) Ln tetrakis dibenzoylmethide Average decay

results for the Ln triethylammonium compound time (μs)
dibenzoylmethide
Sm 63.8 4.9
triethylammonium
compounds [81] Gd 56.8 3.3
Nd 36.3 4.0
Pr 432.0 25.9
Tb 494.4 7.9
Yb 49.5 0.5
Eu 519.2 9.2
Interestingly, it appears as though the decay time and triboluminescent emission

yield were not correlated. While EuD4TEA had the longest decay time and is the
brightest material, the Tb compound had the second longest decay time and one of
the smallest TL emission yields. Likewise, the Sm compound had the second
highest TL yield, but it had one of the fastest decay times [81].
7.5.6 Effects of Different Europium Salts
To further reduce the synthesis cost of EuD4TEA, europium acetate was tried
instead of europium nitrate as the precursor material. Europium acetate was chosen
because it can be dissolved in ethanol, it is the same europium ion state as the
common europium nitrate, and is significantly less expensive. The synthesis pro-
cess began by pouring 25 mL of HPLC grade anhydrous denatured ethanol (Alfa
Aesar, 22930) into an Erlenmeyer flask and placed onto a hot plate. Afterwards,
4 mmol of 99.9 % (REO) europium (III) acetate hydrate (Alfa Aesar, 42878) was
added to the hot solution. Then, 13 mmol of 99 % (GC) 1,3-diphenyl-1,3-
propanedione (Chem-Impex, 00693) also known as dibenzoylmethane (DBM)
was added. This addition caused the clear solution to turn dark yellow. Once the
DBM was dissolved, the flask was removed from the hot plate and 14 mmol of
99.5 % triethylamine (TEA) (Sigma-Aldrich, 471283) was added. The solution
was then kept aside to cool at ambient temperature. The EuD4TEA powder that
formed were light yellow in color, clumpy, and hard as shown in Fig. 7.32a. The
same procedure was then repeated using 99.9 % (REO) europium (III) nitrate
hexahydrate (Alfa Aesar, 15290), which produced a light yellow powder that was
very sparkly as shown in Fig. 7.32b [98].
Using the drop tower and LabVIEW VI tools, the effects of europium acetate on
the triboluminescent properties of EuD4TEA is shown in Fig. 7.33. Due to this
being a relative measurement, all the data was normalized to the europium nitrate
result. The error in normalized triboluminescent light yield was estimated to be 7 %
which is the error from synthesizing the material and the drop tower apparatus
[89]. From these results, it is evident that europium acetate provides about half of
Fig. 7.32 Pictures of the finished EuD4TEA samples synthesized with (a) europium acetate and
(b) europium nitrate. These pictures were taken under standard white room light [98]
1.2
Normalized Triboluminescent Light Yield
1.0
0.8
0.6
0.4
0.2
0.0
Acetate Nitrate
Fig. 7.33 Comparison of the TL produced by EuD4TEA synthesized with europium acetate and
nitrate precursors [98]
the triboluminescent yield as the nitrate version. It should be noted that while the
triboluminescent yields were different, the decay times are statistically identical.
The measured triboluminescent decay time for the acetate based EuD4TEA
was 522.7 10.1 μs, while the nitrate based EuD4TEA decay time was
524.3 17.7 μs [98].
To understand why the acetate provides a lower triboluminescent yield, we first
used a standard optical microscope to investigate the crystals. As you can see in
Fig. 7.34a, the crystals formed by the acetate are cloudy. Meanwhile, the crystals
produced by the nitrate are clear as shown in Fig. 7.34c [98]. This could be
explained using the solubility. When acetate was used to create EuD4TEA, the
acetate had a very low solubility with ethyl alcohol (it was completely insoluble
Fig. 7.34 Optical microscope pictures of EuD4TEA crystallites made with europium acetate and
nitrate at a magnification of 40. EuD4TEA made with europium acetate is shown under (a) regular
white light and (b) ultraviolet light. EuD4TEA made with europium nitrate is shown under (c)
regular white light and (d) ultraviolet light [98]
with acetone). This in turn caused a quick precipitation and rougher crystal.
Conversely, the europium nitrate has a higher solubility with ethyl alcohol, which
slows the precipitation and allows for a better crystal to form [98]. This is verified
by the SEM images shown in Fig. 7.35. Notice how the acetate based EuD4TEA
looks dirty, i.e., it is typically a rod shape with many smaller rods growing around
it [Fig. 7.35a]. Conversely, the nitrate based EuD4TEA tends to be more plate-like
and cleaner [Fig. 7.35b]. More research is needed to fully understand these
results [98].
The measured photoluminescent and triboluminescent emission spectra for
EuD4TEA are shown in Figs. 7.36 and 7.37, respectively. Notice that the PL and
TL excited emissions for both the acetate and nitrate precursor samples have the
same peak wavelengths and are very bright. The spectra indicate that the lumines-
cence from EuD4TEA originates from the typical Eu3þ centered transitions from
the 5D0 levels to the lower 7F0–4 levels [79]. The main emission occurred from the 5
D0 ! 7F2 transitions.
Fig. 7.35 Scanning electron microscope (SEM) images of EuD4TEA crystallites made with (a)
europium acetate and (b) europium acetate precursors at a magnification of 300 [98]
Fig. 7.36 Photoluminescent emission spectrum of EuD4TEA synthesized with europium

acetate and nitrate precursors. The inset pictures show the PL emitted from the samples under
the UV light [98]
7.5.7 Additives or Dopants for EuD4TEA
As this chapter has shown thus far, europium is obviously the best candidate for the
use in TL based sensors. No other rare earth can produce the triboluminescent
intensity of EuD4TEA. In addition, the synthesis process has become much more
inexpensive with the use of europium nitrate and other solvents such as acetone or
denatured ethanol. By increasing the amount of solvent, the amount of TL can also
4000 Acetate
5D ®7F
Nitrate 0 2
3500
3000
Intensity (a.u.)
2500
2000
Acetate Nitrate
1500
5D ®7F
0 1
5
1000 D0®7F4
5D ®7F
0 3
500
0
550 575 600 625 650 675 700 725 750
Wavelength (nm)
Fig. 7.37 Triboluminescent emission spectrum of EuD4TEA synthesized with europium acetate
and nitrate precursors. The inset pictures show the TL emitted upon impact [98]
increase. Thus, the optimization of the EuD4TEA synthesis is complete. There is

little that can be done to optimize the synthesis further. The only other way that the
triboluminescent properties can be enhanced even further is by incorporating
additives or dopants into the EuD4TEA crystal structure.
7.5.7.1 Morphine
The only known research into doping of EuD4TEA before 2013 was done in 2000
by Zeng et al. The authors in this case studied the effects of morphine on the TL of
EuD4TEA [99]. The synthesis was based on the method first described by Hurt
et al. [76, 99]. Their study began by dissolving europium (III) chloride, DBM, TEA,
and morphine in absolute ethanol. Once the materials were filtered, the TL was
recorded using a Hitachi 850 fluorospectrometer [99]. The results shown in
Fig. 7.38 indicated that the addition of morphine to the synthesis made EuD4TEA
eight times brighter [99].
7.5.7.2 Piperine Doping
Using morphine as a dopant was not possible in the United States due to severe
regulatory restrictions on its use. As a result, piperine was used due to its similar
properties to morphine. Piperine is the alkaloid responsible for the characteristic
smell and taste of black pepper. It has been used in some forms of traditional
Fig. 7.38 Comparison of

the triboluminescent
spectrum of (A) EuD4TEA
and (B) morphine doped
EuD4TEA [99]
Fig. 7.39 Chemical structure of (a) morphine and (b) piperine [88]
medicine and as an insecticide. Piperine forms monoclinic needles and is slightly

soluble in water and alcohol. It is sometimes used to simulate the chemical
properties of morphine. Piperine and morphine have the same stoichiometric
formula and molar mass, but different chemical structure. Morphine is a benzyliso-
quinoline alkaloid with two additional ring closures as shown in Fig. 7.39a
[100]. Conversely, piperine is a peperinic acid with a piperidide structure as
shown in Fig. 7.39b [101]. As a result, both food grade (W290904) and 97 %
analytical reagent grade (P49007) piperine from Sigma-Aldrich was used as a
dopant for EuD4TEA during this research [88].
Two sets of synthesis were carried out to study the effect of the purity of the
dopant used on the TL property of EuD4TEA. In order to determine the effects of
piperine on TL yield for EuD4TEA, nine different trials using food grade piperine
(masses of 0.0053, 0.3617, 0.5558, 0.8797, 1.1085, 2.1396, 2.8135, 3.5207, and
5.2576 mmol) were dissolved separately using 25 mL of heated anhydrous ethanol.
To study the effect of reagent grade Piperine on EuD4TEA a second set of piperine
experiments was synthesized using ten different trials using analytical 97 % reagent
grade piperine (masses of 0.0084, 0.0382, 0.3484, 0.5576, 0.8877, 1.7670, 2.3267,
2.6694, 2.9870, and 3.5169 mmol) were dissolved using 25 mL of heated anhydrous
ethanol [88]. The remainder of the synthesis process for each trial was identical to
what has been described by Fontenot et al. [5].
The addition of food grade piperine significantly decreased the amount of time
required for crystallization. While the undoped EuD4TEA took about 30–60 min
before crystals start precipitating out of solution, complete precipitation for
undoped EuD4TEA took almost one day. The presence of piperine in EuD4TEA
reduced the start of precipitation to a few minutes with the full process taking no
more than a few hours. The fastest precipitation occurred using 0.5558 mmol of
piperine where complete crystallization took less than 5 min [88].
It also appeared that crystal size for the food grade piperine doped EuD4TEA can
be controlled by the amount of piperine added during synthesis. Figure 7.40 shows
pictures of the resulting EuD4TEA crystals doped with 0.5579 mmol of food grade
piperine excited by standard room light (a) and UV light (b). EuD4TEA crystals
containing less than about 1 mmol of food grade piperine are relatively small
similar to what is shown in Fig. 7.40. Figure 7.41 shows pictures of the resulting
EuD4TEA crystals doped with 3.5207 mmol of food grade piperine excited by
standard room light (a) and UV light (b). When the amount of piperine added to the
Fig. 7.40 Pictures showing the EuD4TEA crystals doped with 0.5579 mmol of food grade
piperine illuminated by: (a) standard room light and (b) UV light [88]
Fig. 7.41 Pictures showing EuD4TEA crystals doped with 3.5207 mmol of food grade piperine
illuminated by: (a) standard room light and (b) UV light [88]
Fig. 7.42 Pictures of EuD4TEA doped with: (a) 0.3484 mmol, and (b) 2.3267 mmol of 97 %
reagent grade piperine showing differences in crystal size [88]
EuD4TEA is slightly greater than 1 mmol, the crystals are larger and exhibit a flaky
structure that significantly quenches the resulting TL [88].
As was observed for the crystals doped with food grade piperine, there was a
noted decrease in time for the complete precipitation of EuD4TEA crystals from the
solution when the 97 % reagent grade piperine was added. However, changing
the mass of added piperine did not significantly change the precipitation time.
Full precipitation of the 97 % reagent grade doped EuD4TEA crystals took less
than 2 h [88].
Similar to what has been observed above, the crystal size and triboluminescent
behavior depend on the amount of piperine added during synthesis. EuD4TEA
crystals containing less than about 1 mmol of 97 % reagent grade piperine are
relatively small as shown in Fig. 7.42a. When the mass of 97 % reagent grade
piperine is more than 1 mmol, the doped EuD4TEA crystals are larger and exhibit a
flaky structure, as shown in Fig. 7.42b, which quenches the resulting TL [88].
Fig. 7.43 Plot of the normalized TL light yield versus the amount of added food grade (black
points) and 97 % reagent grade (red points) piperine [88]
Figure 7.43 shows a plot of the normalized triboluminescent yield versus the
amount of added dopant for both food grade and 97 % reagent grade piperine. The
triboluminescent yields for all measured samples were normalized against undoped
EuD4TEA, which was set to one. In general, increasing the amount of added
piperine decreased the triboluminescent light yield. Smaller quantities of added
food grade reagent appeared to quench the triboluminescent yield compared to
97 % reagent grade piperine. Both reagents reduced the triboluminescent yield to
about 15 % of the undoped value when 3 mmol of piperine was added to the
EuD4TEA. However, a small quantity (0.0364 mmol) of 97 % reagent grade
piperine appears to produce 8.5 % more TL compared to the undoped material.
This small increase in light yield was within the statistical uncertainty of this data.
As the amount of 97 % reagent grade piperine increased beyond 0.0364 mmol, the
TL yield decreased. The dashed lines shown in Fig. 7.43 were added to show the
general trend in the data and are for comparison purposes only [88].
An average decay time was calculated from drop tower triboluminescent data
collected during five drops. Pure EuD4TEA was found to have a decay time of
544.8 7.4 μs. Decay time results for EuD4TEA doped with food grade piperine is
shown in Table 7.14. Notice that adding more than about 2 mmol of food grade
piperine significantly reduced the decay time [88].
Decay time results for EuD4TEA doped with 97 % reagent grade piperine is
shown in Table 7.15. As before, the measured decay times significantly decreased
when approximately 2 mmol of 97 % reagent grade piperine was added.
Table 7.14 Measured Amount of added Average decay

triboluminescent decay times piperine (mmol) time (μs)
for EuD4TEA doped with
0.0053 599.6 22.9
food grade piperine [88]
0.3617 578.9 16.1
0.5558 577.4 27.5
0.8797 529.0 13.0
1.1085 566.0 36.4
2.1396 319.0 42.1
2.8135 493.2 38.0
3.5207 452.4 17.4
5.2576 519.0 13.5
Table 7.15 Measured Amount of added Average decay

triboluminescent decay times piperine (mmol) time (μs)
for EuD4TEA doped
0.0084 566.2 11.4
with 97 % reagent grade
piperine [88] 0.0382 583.1 23.9
0.3484 547.2 10.7
0.5576 524.2 26.6
0.8877 517.1 12.3
1.7670 515.9 14.6
2.3267 506.3 9.3
2.6694 316.0 18.3
2.9870 307.0 41.9
3.5169 476.6 23.9
7.5.7.3 Multivitamin Doping
A common multivitamin contains a broad selection of elements. It was decided to

use a commercial-over-the-counter multivitamin as a dopant to study the multi-
element effect on TL. For it is possible that one or more of these added elements
could increase triboluminescent yield. Results from this research could yield a
roadmap as to how to best proceed in the selection of possible dopants for EuD4TEA.
Production of the multivitamin doped EuD4TEA began by crushing a single
One-A-Day Energy® multivitamin (produced by Bayer) with a mortar and pestle.
The powdered vitamin was then dissolved in 5 mL of anhydrous ethanol, and kept
aside to allow the undissolvable solids to settle down. Next, 4 mmol of europium
nitrate hydrate was dissolved in 25 mL of heated 200 proof anhydrous ethanol. The
vitamin solution was added to the europium nitrate solution making sure that the
insoluble solids (starch and filler) did not enter. To this resulting solution, 13 mmol
of DBM followed by 14 mmol of TEA were added. The solution was heated until
the mixture was clear and then allowed to cool slowly overnight at ambient
temperature. The precipitated EuD4TEA crystals were collected under suction.
The crystalline product was then dried in air under ambient conditions [102].
Fig. 7.44 Pictures of the multivitamin doped EuD4TEA illuminated with: (a) Standard room
light, and (b) Ultraviolet light [102]
4000 5 7
1800
D0→ F2 5 7
1600 D0→ F2
1400
3000
1200
Intensity
Intensity
1000
2000
800
600
1000
400
5 7
D0→ F1 5 7 5 7
5
D0→ F3
7 D0→ F4 200 D0→ F1 5 7 5 7
D0→ F3 D0→ F4
0 0
550 600 650 700 750 550 600 650 700 750
Fig. 7.45 Measured TL spectra for: (a) pure EuD4TEA, and (b) Multivitamin doped
EuD4TEA [102]
Pictures of the EuD4TEA crystals illuminated with (a) standard room light and
(b) ultraviolet light are shown in Fig. 7.44. As the figure shows, the doped
EuD4TEA emits the same red luminescence that is present in the undoped version.
The crystals appear to be dark yellow under standard room light and are relatively
hard. Pure EuD4TEA is lighter in color and is more soft and fluffy. The doped
crystals also appear to be more spherical in shape and produce TL with low or light
applied pressure [102].
The triboluminescent spectrum was measured for undoped and doped EuD4TEA
as shown in Fig. 7.45. A S2000 fiber optics spectrometer was used to make these
measurements. The resulting spectra are typical of the Eu3þ centered transitions
from the 5D0 levels to the lower 7F0–4 levels. Interestingly, the addition of the
multivitamin dopant in EuD4TEA did not shift the triboluminescent or
photoluminescent emission peaks. Instead the dopant acted as a coactivator by
only changing the intensity of the emission peaks [102].
Inductively coupled plasma optical emission spectrometry Perkin Elmer
(ICP-OES) DV2400 model was used to determine the elemental concentration for
the both pure and doped EuD4TEA samples. This was accomplished by taking
0.100 g of each sample and dissolving it in a solution of 10 mL of nitric acid and
deionized water. A nebulizer converts the liquid into a fine aerosol, which is
injected into an argon-supported inductively coupled plasma. The formation of
excited atoms for ICP-OES requires plasma temperatures, between 5000 and
10,000 C. ICP-OES is similar to particle induced X-ray emission (PIXE), except
plasma is used instead of particles as an excitation source. Visible light is emitted as
part of the ICP-OES analysis. Conversely, inner shell X-ray photons are emitted
during the PIXE process. ICP-OES can detect parts-per-billion (ppb) levels for
more than 70 elements [102].
Table 7.16 shows the results of this ICP-OES analysis. The measured quantities
for each detected element are normalized to europium having a ratio of 100 %.
Results indicated that small quantities of elemental calcium, iron, magnesium,
copper, and manganese were present in the doped EuD4TEA. For example, the
doped sample contains about 0.10 % of elemental iron compared to europium [102].
Figure 7.46 shows the TL emission yield ratio for both the pureand doped
EuD4TEA samples as a function of the number of drops from the ball in the drop
tower. The triboluminescent emission yield ratio is normalized to the first ball drop
on pure EuD4TEA, which were numerically set to 1.00. Error bars for both the
undoped and doped EuD4TEA samples were set to about 10 % of the yield ratio
values. This level of measurement uncertainty was based on similar data that has
been collected by authors over the last 2 years [36]. The solid lines shown in
Fig. 7.46 are exponential curve fits, which are similar to what was used with earlier
collected data for EuD4TEA [36].
Results from Fig. 7.46 show the doped EuD4TEA powder statistically emits
more TL during the first three drops. After the fourth drop, the pure EuD4TEA
sample emits more TL, which shows the multivitamin-doped material is more
easily damaged. Table 7.17 shows the relative TL decay time results for the pure
and doped EuD4TEA samples. Results from Table 7.17 show there was no statis-
tical difference in decay time between the pure and doped EuD4TEA
powders [102].
7.5.7.4 Uranium Doping
Uranium is known for its strong luminescent properties [103, 104]. By incorporat-
ing uranium with EuD4TEAs, the triboluminescent properties and the effects of
long term radiation exposure will also be determined. This introduces a novel
dimension completely different from the usual surface effects that are normally
investigated [105].
The synthesis of uranium doped EuD4TEA began by pouring 75 mL of 95 %
denatured ethanol into a beaker. The ethanol was heated to 250 C and the stirring
set to maximum. Then, 4 mmol of europium nitrate and the desired amount of
uranyl acetate were added to the hot solution. The quantity of added uranium for the
doped EuD4TEA samples ranged from 0 to 100 mol%. The uranium concentration
Table 7.16 Results from Mass ratio of given

the ICP-OES analysis Element element to europium (%)
for the vitamin doped
Europium 100
EuD4TEA [102]
Calcium 3.66
Iron 0.10
Magnesium 0.62
Copper 0.01
Manganese 0.01
2.0
1.5 Doped
TL Emission Ratio
1.0
Undoped
0.5
0.0
1 2 3 4 5
Drop Number
Fig. 7.46 Plot of triboluminescent emission ratio versus the drop number for: (a) Undoped, and
(b) Doped EuD4TEA. The triboluminescent emission intensity for drop number one for the pure
EuD4TEA is normalized to 1.000 [102]
Table 7.17 Decay time EuD4TEA form Decay time (μs)

results for the pure and doped
Pure 544.8 7.5
EuD4TEA [102]
Doped 538.4 4.5
in each sample was determined using the U4þ/Eu3þ ratio. The uranyl acetate source
material used in this research was manufactured by Mallinckrodt, Incorporated and
has an activity of 0.2 μCi/g. This activity is due to the depleted uranium that was
used to synthesize the acetate [105].
Fig. 7.47 Pictures of EuD4TEA doped with uranium showing the light yellow color of the
low concentration of uranium doping observed under: (a) Standard room light, and (b) UV
irradiation [105]
Fig. 7.48 Pictures of EuD4TEA doped with uranium showing the: (a) orange uranium dots that
formed “uranium pox,” and (b) the effects of the increased concentration creating a more orange
color [105]
Once the uranyl acetate and europium salts are dissolved, 13 mmol of DBM was
added to the hot solution. A funnel was then placed on top of the beaker and the
solution was left for 20 min. After 20 min, the stirrer was removed and 14 mmol of
TEA was added. The solution was then kept aside to cool at ambient temperature.
The EuD4TEA crystals that formed were filtered and air dried at room
temperature [105].
For low concentrations of added uranium (<10 mol%), the material was light
yellow in color as shown in Fig. 7.47a. Figure 7.47b shows the emission intensity
when the uranium-doped samples were exposed to UV light. By comparison, as the
concentration of uranium increased, dark orange spots (“uranium pox”) started to
form as shown in Fig. 7.48a. A further increase in concentration caused the whole
Fig. 7.49 Plot of the normalized triboluminescent light yield ratio versus the uranium dopant
concentration [105]
material to turn dark orange in color as shown in Fig. 7.48b. It is likely that it
becomes increasingly more difficult to incorporate uranium into the EuD4TEA. In
addition, as the uranium concentration became larger, the fluorescence yield tended
to decrease. There appeared to be a direct relation between the reduction in
fluorescence yield and sample color [105].
In addition to using uranium to dope EuD4TEA, a sample of uranium tetrakis
dibenzoylmethide triethylammonium (UD4TEA) was synthesized using a similar
procedure. In this case, no europium nitrate was used. The synthesis began by
adding 4 mmol of uranium acetate to 25 mL heated 95 % denatured ethanol. To this
hot solution, 13 mmol of DBM was added. Once the DBM and uranium was
completely dissolved into the solution, 14 mmol of TEA was added. The solution
was heated until the mixture was clear, and then kept aside to cool slowly overnight
at ambient temperature. The UD4TEA crystals that formed were filtered and air
dried at room temperature. The material that formed was a dark orange throughout
and showed no luminescence under UV irradiation [105].
The triboluminescent yield for the doped samples was normalized in relation to
the undoped EuD4TEA. The experimental uncertainty was also estimated to be
about 7 % which is a combination of the product consistency between batches and
the testing method [89]. The results of the triboluminescent light yield measure-
ments taken 120 days after synthesis are shown in Fig. 7.49. Notice a uranium
concentration of 4 mol% was found to have the largest triboluminescent emission
yield. In fact, the 4 mol% uranium sample emitted 80 % more TL than undoped
EuD4TEA. For dopant concentrations less than 4 mol%, the TL ratio was similar to
undoped EuD4TEA. Above 4 mol%, the yield starts to decrease. The reason for this
Fig. 7.50 Three dimensional photoluminescent emission spectra for uranium doped
EuD4TEA [105]
increase or decrease could be due to the uranyl ion increasing or decreasing the
absorption lines [105].
The triboluminescent yield ratio above 15 mol% started to fluctuate between
0.25 and 0.75, which is likely due to the uranium not completely being incorpo-
rated into the sample. It is likely that this extra uranium formed isomorphic
products, which may be either located on the surface or as a separate compound,
forming a mixture of the two. This may be the case as uranium pox appears at
certain locations, and also as the concentration of uranium increases the deep
orange product on the surface may be responsible for the reduction in the emission
of PL. Instead, some of the uranium possibly gets inside the sample with the rest
remaining in the excess liquid, thus forming an orange particulate that exhibits no
PL or TL. These yield fluctuations occur at dopant concentrations of between
15 and 75 mol%. At 100 mol% uranium concentration, the sample did not exhibit
any TL or PL. Here, the uranium appears to completely quench the luminescence.
As a check, UD4TEA was synthesized and was found to not exhibit any TL upon
impact. The excess of electrons from the uranium could be causing the highly
doped EuD4TEA samples to act like a conductor and therefore quenching the
TL [105].
The photoluminescent emission spectra for the uranium doped EuD4TEA
samples are shown in Fig. 7.50. The spectra indicate that the luminescence was
due to the Eu3þ excited transitions. As the figures indicate, the addition of uranium
did not shift the wavelengths for any of the measured emission peaks. The Eu3þ
transitions are from the 5D0 levels to the lower 7F0–4 levels with the main 614 nm
emission occurring from the 5D0 ! 7F2 transition. It should be noted that as the
concentration of uranium was increased, the PL decreased once the optimal amount
of uranium was reached. As a result, the integration time had to be raised signif-
icantly compared to the lower concentration. This created light leakage problems
with the spectrometer. Due to this and a desire to make Fig. 7.50 more readable,
only the uranium concentrations up to 20 % are shown. The peak locations however
remained constant despite the increase in uranium. In addition, it should be noted
that the 100 % uranium doping concentration, and UD4TEA samples exhibited no
PL. From the above observations, it appears that low concentrations of uranium act
as a coactivator by changing the intensity of the emission peaks. The addition of
uranium in low concentrations does not shift the peaks in the emission
spectra [105].
The triboluminescent emission spectra for pure EuD4TEA and 4 mol% uranium
doped EuD4TEA are shown in Fig. 7.51a, b respectively. Similar results were
observed for PL, doping EuD4TEA with uranium does not appear to change the
triboluminescent emission spectrum. Both spectra are virtually identical, which
means the uranium simply increases triboluminescent yield [105].
In addition to determining the light yield, the prompt fluorescent decay times
were measured for each sample. The results indicated the decay time does not
depend on the amount of uranium added to the sample. The average decay time for
all of the runs was 499.5 56.0 μs [105].
The effects on the impact energy on the TL emitted from 4 mol% doped
EuD4TEA was also investigated. This was accomplished by varying the initial
drop height of the ball from 2 to 42 in. (5–107 cm). Thus the normalized TL
yield could be measured as a function of kinetic energy of the drop ball at the time
of the impact. The results of this are shown in Fig. 7.52. All data presented in
Fig. 7.52 was normalized to the average TL yield for the 2 in. (5 cm) data point. The
error in triboluminescent light yield was estimated to be 7 %. The uncertainty in
kinetic energy was estimated to be 5 % based on conservation considerations [105].
The data shown in Fig. 7.52 was completely consistent with earlier results as
reported by Fontenot et al. [86]. The data in Fig. 7.52 indicated that the tribolumi-
nescent yield was proportional to kinetic energy for low speed impacts. Plots
similar to Fig. 7.52 make it relatively simple to calibrate a TL-based impact
sensor [105].
Using the software as developed by Fontenot et al. [37], the decay times were
measured for the 4 mol% doped sample for each drop height. The data indicates that
the decay time was not dependent on drop height. This was completely consistent
with earlier results as reported by Fontenot et al.[86]. The average 4 mol% doped
EuD4TEA triboluminescent decay time over all the drop heights was
539.2 63.1 μs. The measurement uncertainty was estimated using the standard
deviation over all drops [105].
Fig. 7.51 Measured triboluminescent emission spectra for the: (a) undoped and (b) 4 mol%
uranium doped EuD4TEA [105]
All the measurements to this point were completed just after the doped
EuD4TEA samples were synthesized. It has been known for quite some time that
ionizing radiation reduces luminescence emitted from both inorganic and organic
materials [106, 107]. In 1951, Birks and Black showed experimentally that
the fluorescence efficiency of anthracene bombarded by alphas varied with total
fluence [106]. Since 1990, Hollerman et al. have measured the reduction in light
yield for several single crystals, polycrystalline paint, and pressed tablet forms of
selected rare earth materials prepared at ambient temperature. These results showed
Fig. 7.52 Plot of the normalized TL yield versus impact kinetic energy for 4 mol% uranium doped
EuD4TEA [105]
that fluences of 3 MeV protons between (0.1–4.0) 1014 mm–2 will reduce the
emission yield to half its original value [66, 68, 84, 85, 106, 108–110]. It is clear
that exposure to ionizing radiation reduced the fluorescence yield. As time passed,
the doped EuD4TEA samples were exposed to radiation from depleted uranium
with an activity of 0.2 μCi/g [105].
As was stated before, doping EuD4TEA with 4 mol% uranium increases the
initial triboluminescent yield over undoped EuD4TEA by about 80 %. However,
gains in triboluminescent yield were found to decrease with time based on the
emission of radiation from the depleted 238U in these samples. It is likely that
alpha particles emitted from the decay of 238U and its corresponding daughter
products are causing this reduction in emission yield. The reduction in fluores-
cence intensity upon exposure to ionizing radiation from heavy charged particles
follows the Birks and Black relation as described by Birks and Hollerman
et al. [84, 107]. These alpha particles break chemical bonds, thus reducing the
radiative emission in doped EuD4TEA. After 120 days, the drop tower was again
used to measure the TL emission yield for the 4 mol% doped uranium samples.
Results showed the triboluminescent yield was reduced by about 20 % over the
initial value measured when the sample was synthesized. At this rate, it should
take about 335 days for the triboluminescent yield to be reduced to half of its
original value [105].
7.5.7.5 Dimethyl Methylphosophonate Doping
DMMP is also known as phosphonic acid and methyl-dimethyl ester [111]. It is

commonly used as a flame retardant, a nerve agent simulate, a preignition additive
for gasoline, an antifoam agent, a textile conditioner, and an antistatic agent
[88, 111–113]. It is used experimentally to mimic the physical and spectroscopic
but not biological properties of anticholinesterase agent [111]. In addition it is used
in applications where high phosphorus content, good solvency, and low viscosity
are desired [112]. It lowers viscosity of epoxy resins and unsaturated polyesters
filled with hydrated aluminum oxide [112].
The synthesis of doped dimethyl methylphosphonate (DMMP) began by adding
4 mmol of europium (III) nitrate hexahydrate to a hot solution of 25 mL of
denatured anhydrous ethanol. Next, DMMP was added to the hot solution. Sixteen
different trials were conducted to determine the effects of DMMP. The amount of
DMMP added was 0.00, 2.32, 4.61, 6.92, 9.23, 11.53, 13.84, 18.46, 23.07, 27.68,
32.30, 36.91, 46.14, 55.37, 64.60, and 73.82 mmol. After the desired amount of
DMMP was added, DBM was added followed by the addition of 14 mmol of TEA.
The solution was then left at ambient temperatures to cool. The crystals that formed
are shown in Fig. 7.53 were collected under suction [114].
The addition of DMMP had a significant change in the precipitation time. Full
precipitation occurred in less than 2 h with crystallization starting almost immedi-
ately for all dopings less than or equal to 32.30 mmol. Beyond the dopant concen-
tration of 32.30 mmol, crystallization became very slow. After nearly 2 months,
only a half liquid half crystal product was formed. This material was more like a
yellow orange sticky watery piece of goo that did not exhibit any TL and only a
small amount of PL [114].
One interesting effect noticed with the doping of DMMP was that, as the
concentration increased, the particles that formed started becoming a flat snowflake
and more sparkly as shown in Fig. 7.53. The normal crystalline spherical product
usually observed with the pure EuD4TEA is shown in Fig. 7.53a. It should be noted
that as the DMMP increased, so did the snowflake structure as shown in
Fig. 7.53d [114].
Since DMMP dissolved both europium and DBM, one trial run was attempted
using DMMP as a solvent. No heating was required for this run. However, after
nearly 3 months, nothing formed. As a result, DMMP was not useful as a solvent as
the synthesis time would be too long to manufacture EuD4TEA in a cost effective
manner if any material manages to form at all [114].
The results of the effects of DMMP are shown in Fig. 7.54. Due to this being a
relative measurement, all the data was normalized to the baseline 0 mmol of
DMMP [114]. The error was estimated to be 7 % [89].
From Fig. 7.54, it appears as though DMMP can significantly increase the TL of
EuD4TEA. A second run of the two outliers 6.92 and 9.23 mmol DMMP were done.
Fig. 7.53 Pictures of the (a) 0, (b) 9.23, (c) 18.46, and (d) 32.30 mmol of DMMP doped
EuD4TEA [114]
Fig. 7.54 Effects of DMMP on the TL of EuD4TEA [114]

Fig. 7.55 Effects of DMMP on the PL spectra of EuD4TEA [114]
There was no change in the observed results. As a result, it appears as though adding
a small amount of DMMP can double the output of TL. Afterwards, the TL is
quenched to only about 25 % increase, then it increases until the TL is nearly
tripled. This increase in TL was significantly brighter than any other dopant tested.
Further, the addition of DMMP appears to be similar to a Geiger counter voltage
curve. At low concentrations, the TL starts to increase, then it reaches a plateau for
a wide range of DMMP concentrations, afterwards the TL starts to increase
again [114].
The measured photoluminescent and triboluminescent emission spectra for
EuD4TEA are shown in Figs. 7.55 and 7.56 respectively. It was observed that
both emission peaks have the same peak wavelengths but differ in intensity yield.
The spectra indicated that the luminescence from EuD4TEA comes from the typical
Eu3þ excited transitions from the 5D0 levels to the lower 7F0–4 levels.
It was observed that the peak locations remain constant even as the amount of
DMMP was increased. As a result, it appears as though the added DMMP acts as a
coactivator by changing the intensity of the emission peaks, but does not shift the
electronic spectra locations in any way. The triboluminescent emission spectra are
shown in Fig. 7.56. The resulting spectra are identical to the photoluminescent
spectra. This likely means that the process involved for the PL is the same process
involved for the TL of EuD4TEA [114].
Using decay time software as described by Fontenot et al. [37], the decay times
were determined for each drop. From the data as shown in Table 7.18, it appears as
though DMMP does not change the triboluminescent decay time for EuD4TEA.
Fig. 7.56 Measured triboluminescent spectra of the plain and DMMP doped EuD4TEA [114]
Table 7.18 Decay time (1/e) DMMP added (mmol) Average (μs)
results for DMMP doped
0 531.1 21.7
EuD4TEA [114]
2.307 527.1 19.0
4.614 539.4 7.9
6.921 514.2 9.4
9.228 512.9 6.6
11.535 492.5 8.7
13.842 524.3 11.4
18.456 538.7 8.9
23.070 550.2 8.0
27.684 543.8 9.6
32.298 532.1 13.4
Average 529.1 30.4
The average decay time for all the dopings was 529.1 30.4 μs. The error was
estimated using the standard deviation of all the drops [114].
Figure 7.57 shows the effects of impact energy on the TL of EuD4TEA and
EuD4TEA doped with 32.3 mmol of DMMP. Each material was normalized to their
6 in. (0.19 J) value. Both sets of data were fit using a two parameter power equation
in SigmaPlot®. Interestingly, it appears as though the TL of the pure version of
EuD4TEA is directly proportional to the E1/3. However, when DMMP was added,
the TL is directly proportional to E1/4. This indicates that there was a single process
involved which is in contrast to the more widely known ZnS:Mn which experiences
2.5
EuD4TEA
EuD4TEA+DMMP
2.0
Normalized TL Light Yield
1.5
1.0
0.5
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Fig. 7.57 Effects of impact energy on the TL of EuD4TEA þ DMMP [114]
Table 7.19 Effects of impact energy on EuD4TEA doped with 32.3 mmol of DMMP [114]
Drop height (in.) Drop ball material Impact energy (J) Measured decay time (μs)
2 Styrofoam 0.0005 583.3 11.8
2 Wood 0.0050 576.1 5.6
2 Steel 0.0648 610.7 10.8
3 0.0972 598.2 10.5
4 0.1296 611.5 8.6
5 0.1620 608.8 10.0
6 0.1944 604.2 7.1
12 0.3887 578.4 5.0
18 0.5831 533.2 7.1
24 0.7774 578.7 20.2
30 0.9718 535.5 10.2
36 1.1661 537.5 10.8
39 1.2633 517.3 11.7
42 1.3605 531.3 10.8
plastic and elastic deformation [64, 69, 70, 86, 115]. These extra processes create
two regions of interest with two linear fit lines [114].
Table 7.19 shows the effect of impact energy on the decay time for EuD4TEA
with 32.3 mmol of DMMP. The results indicate that the impact energy has little
effect on the triboluminescent decay time.
7.5.7.6 Dibutyl Phosphate
Due to the success of DMMP, dibutyl phosphate (DBP) was investigated as a possible
additive because it is similar to DMMP. DBP is a pale-amber colorless medium strong
acid that is used as a catalyst in organic synthesis [116–118]. However, DBP can be a
dangerous material since it reacts with strong oxidizers and attacks many metals
forming flammable and explosive hydrogen gas [117, 118]. DBP also reacts with the
moisture in the air by releasing acid fumes due to the formation of phosphoric acid with
water [117]. As a result, it is often used as a catalyst in the manufacture of phenolic and
urea resins and in the separation and extraction of metals [116, 117].
The synthesis of EuD4TEA began by pouring 25 mL of reagent grade anhydrous
denatured ethanol (Sigma-Aldrich, 277649) into an Erlenmeyer flask. The desired
amount of DBP, which ranged from 0 to 5 mL, was then added to the ethanol. The
DBP and ethanol mixture was then heated on a hot plate. Afterwards, 4 mmol of
europium (III) nitrate hexahydrate was added to the hot solution. This addition
caused the solution to turn from clear to white. Once the solution became clear
again, 13 mmol of DBM was added. Finally, once the DBM was dissolved,
14 mmol of TEA was added. The solution was then kept aside to cool at ambient
temperature. The pure EuD4TEA powder that formed were light yellow in color and
very sparkly as shown in Fig. 7.58a. It should be noted that violent reactions occur
when DBP is mixed with bases, amines, and amides. As such, the mixing order is
vital to avoid such reactions [119].
For small concentrations of DBP (0.5 mL), the final EuD4TEA product looked
and behaved like the pure material, i.e., a fluffy sparkly light yellow powder, as
shown in Fig. 7.58a. However, when the DBP concentration was increased above
this limit, changes in the precipitation time and crystal structure were observed. For
example, when 0.5–1.0 mL of DBP was added to the mixture, small yellow hard
crystals formed (less than about 1 cm), as shown in Fig. 7.58b [119]. When the DBP
concentration was increased over this limit, i.e., >1.0 mL, significant changes
started to occur. The first change was a significant increase in the precipitation
Fig. 7.58 Standard pictures of EuD4TEA doped with (a) 0 and (b) 1 mL of DBP [119]
Fig. 7.59 Pictures of EuD4TEA with high concentrations of DBP added. Notice the changes in the
structure caused by the DBP, such as the rod-like structure in (a) with 3 mL of DBP and peach fur
balls in (b) with 1.5 mL of DBP [119]
time. Generally, precipitation starts within minutes for EuD4TEA using this method
with a complete precipitation occurring within an hour. This was true for all
concentrations below 1.5 mL. However, when the concentration of DBP reached
1.5 mL, a white film started to form in the mixture, which caused the precipitation
to start after approximately 24 h with full precipitation occurring in 48 h. This
produced a yellow flaky structure with some large orange balls of varying size
beneath some of the flakes [119]. At 2.0 mL of DBP, the final product was a mixture
of whitish yellow and whitish peach flakes. Above this DBP concentration, the
changes in the crystal structure and film started to become more pronounced. For
example, the film changed to a thick yellow white film when 2.5 mL of DBP was
added. This produced whitish peachy yellow colored rod like clumpy structure
shown in Fig. 7.59a. When the DBP concentration was increased even further to
3.0 mL, a clumpy white product formed with peach fur balls underneath, as shown
in Fig. 7.59b [119].
Using a custom built LabVIEW VI, the TL produced from each trial was
analyzed. Due to this being a relative measurement, all the light yield data was
normalized to the pure EuD4TEA light yield result. The error in yield was estimated
to be 7 %, which is the error from synthesizing the material and the drop tower
[89]. From Fig. 7.60, it is evident that DBP can significantly increase the TL of
EuD4TEA. The optimal amount of DBP was determined to be 0.75 mL as shown in
Fig. 7.60a. This concentration yields a fivefold increase in the total amount of TL
emitted compared to the pure EuD4TEA sample. Furthermore, after investigating
the decay times for each drop, it was evident that DBP does not have any effect on
the decay time. The decay time for EuD4TEA with DBP added was measured to be
664.69 65.57 μs [119], which matches the measured decay times for EuD4TEA
discussed in refs. [5, 37, 81, 86, 88, 92, 93, 102, 105, 114, 120–123].
Due to this significant increase in TL, we synthesized another batch by again
varying the amount of DBP between 0 and 1.5 mL so that we could ensure the
accuracy of these results. The only difference between the batches was that we
a b
7 7
6 6

5
5
4
4
3
3
2 2
1 1
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Amount of Dibutyl Phosphate (mL) Amount of Dibutyl Phosphate (mL)
Fig. 7.60 (a, b) Effects of adding measured quantities of DBP on the TL light yield for EuD4TEA
[119]Part labels (a, b) is inserted in Fig. 7.60 caption to match artwork. Please check if appropriate.
It is fine
used regular denatured ethanol instead of anhydrous denatured ethanol. Drop tests
were performed on this batch as well; the results of which are shown in Fig. 7.60b.
Notice the significant increase, i.e., a sixfold increase, was repeated this time with
1.00 mL of DBP. It is evident that DBP can significant increase the total amount
of TL emitted from EuD4TEA. The optimal amount of DBP is between 0.75 and
1.00 mL, and this will produce between a five- and sixfold increase in TL. The
reason for this variance between batches could be due to some of the DBP getting
stuck on the side of the flask instead of becoming incorporated into the
mixture [119].
Furthermore, it should be noted that the products exceeding 1.5 mL of DBP, i.e.,
the products with a white top, were not triboluminescent. These results are consis-
tent with the previous findings in Refs [105, 114]. Figure 7.61 shows a comparison
of the triboluminescent yields for a selection of EuD4TEA additives. From these
results, it is clear that DBP is the best additive thus far. In fact, it produces nearly
twice the amount of TL than DMMP, which increased the triboluminescent light
yield by nearly 200 % [114].
To obtain a better understanding of why DBP is having such a significant effect
on the TL of EuD4TEA, several samples from each batch were placed under a
standard optical microscope with a CCD camera attached. Under 10 magnifica-
tion, it appears as though that pure EuD4TEA is comprised mostly of rods with a few
flattened rod crystals, as shown in Fig. 7.62a. However, as the DBP concentration
increased, the rods started to become more flattened and joined together, as shown in
Fig. 7.62b–d. This was further verified by the 40 images as shown in Fig. 7.63.
Notice that the pure EuD4TEA is a collection of rod like crystals. However, when
DBP is added, the structure starts to flatten out with 0.5 mL of DBP as shown in
Fig. 7.63b. Notice the flat clumped structure for concentrations greater than 0.5 mL
of DBP, as shown in Figures 7.63c, d. At 40 magnification, it is evident that these
flat structures have formed on top of each other.
Fig. 7.61 Comparison of the triboluminescent emission yields for a selection of EuD4TEA
additives. All results were normalized to the pure version of EuD4TEA [114]. The results from
each additive were obtained from refs. [28, 30–32, 42, 44, 45]
Fig. 7.62 Digital images of the crystal structure for EuD4TEA with: (a) 0, (b) 0.25, (c) 0.50, and
(d) 1.00 mL of DMP obtained under 10 magnification [119]
Fig. 7.63 Digital images of the crystal structure for EuD4TEA with: (a) 0, (b) 0.25, (c) 0.50, and
(d) 1.00 mL of DMP obtained under 40 magnification [119]
The measured photoluminescent emission spectra for EuD4TEA are shown in

Fig. 7.64. These photoluminescent results indicate that DBP does not change the
electronic emission energy transitions. The slight variations in the 1.50 mL DBP
results is likely from the large recording time required to obtain the emission
spectra. It should be noted that once the optimum amount of DBP is reached, it
starts having adverse effects on the PL, i.e., the PL decreases significantly once
1 mL is reached. In fact, we were barely able to observe the photoluminescent
emission for 2.0 mL of DBP, and no PL was observed for DBP concentrations
greater than 2.0 mL [119].
The spectra indicate that the luminescence from EuD4TEA (both pure and DBP
added) originates from the typical Eu3þ excited transitions from the 5D0 levels to
the lower 7F0–4 levels with the main emission occurring from the 5D0 ! 7F2
transition [79]. Notice that the peak locations remain constant even as the amount
of DBP is increased. As a result, it appears as though the added DBP acts as a
synthesizer by changing the intensity of the emission peaks, but does not shift the
electronic spectra locations in any way [119].
Figure 7.65 shows the effects of impact energy on the TL of EuD4TEA and
EuD4TEA with 1 mL of DBP added. Each material was normalized to its own 6 in.
Fig. 7.64 Effects of DBP on the PL of EuD4TEA [119]
Fig. 7.65 Effects of impact energy on the total triboluminescent yield of pure EuD4TEA and 1 mL
of DBP added to the EuD4TEA synthesis [119]
(0.19 J) value. Both sets of data were fit using a two parameter power equation in
SigmaPlot®. Interestingly, it appears as though the TL of pure version of EuD4TEA is
directly proportional to the E1/3. However, when DBP was added, the TL is directly
proportional to E1/4 or E1/5. Owing to the previous varying height results of
EuD4TEA discussed in refs. [114, 120, 122], we believe that E1/4 is the correct
value. This indicates that there is a single process involved, which is in contrast to the
more widely known ZnS:Mn, which also experiences plastic and elastic deformation
[64, 69, 70, 86, 115]. These extra processes create three regions of interest with three
linear fit lines. Moreover, these results indicate that the triboluminescent theory
proposed in ref. [115] is not valid for all triboluminescent materials and in particu-
larly EuD4TEA [119].
7.5.8 Polymers Containing EuD4TEA
Now that it has been shown that the TL of EuD4TEA can be increased by nearly
200 % and be made relatively inexpensively, the only task remaining is determining
how to incorporate it into sensors or structures. This chapter will explore three types
of polymers and mounting media for EuD4TEA.
Cytoseal® is clear and colorless and is formulated from high-quality advanced
acrylic resins. It dries rapidly, does not become brittle and crack or discolor with
age, and preserves original slide specimen quality almost indefinitely. Slides
mounted with Cytoseal will not stick together, and if tissue requires remounting,
it can be dissolved in toluene or xylene. Cytoseal is available in two viscosities:
60 for use with cover glass, and 280 for thicker coating or minimal spreading.
Poly (methyl methacrylate), or sometimes called PMMA, is a clear plastic used
as a shatterproof replacement for glass [124]. The clear barrier that keeps hockey
pucks from flying in the faces of hockey fans is made of PMMA. The chemical
company Rohm and Haas makes windows from it and calls it Plexiglas®. Ineos
Acrylics also manufactures PMMA and calls it Lucite®, which is used to make the
surfaces of household items such as hot tubs, sinks, and one piece bathtub and
shower units. PMMA is a member of a family of polymers which chemists call
acrylates, shown in Fig. 7.66a, but the rest of the world calls acrylics [124].
Phthalocyanine (Pc) is a symmetrical 18 p-electron aromatic macrocycle,
closely related to the naturally occurring porphyrins as shown in Fig. 7.66b
[125]. Like the porphyrins, the Pc macrocycle can play host to over seventy
Fig. 7.66 Chemical structures of (a) PMMA and (b) Pc

different metal ions in its central cavity. Since its discovery over seventy years ago,
Pc and its derivatives have been extensively used as colorants (dyes or pigments)
[125]. More recently they have been employed in several applications such as the
photoconducting material in laser printers and the light-absorbing layer in record-
able compact disks. They are also used as photosensitisers in laser cancer therapy,
as nonlinear optical materials and as industrial catalysts [125] .
7.5.8.1 Synthesis of EuD4TEA for Cytoseal Mounting
To determine the effects of a mounting media such as Cytoseal on the tribolumi-

nescent properties of EuD4TEA, a new batch was created using the same synthesis
procedures as described above. Once the EuD4TEA was dry, 0.10 g was placed in
the center of a 1 in. Plexiglass drop tower slide. Cytoseal was then poured over the
EuD4TEA and left to dry for a week. Cytoseal is a clear colorless toluene-based
mounting media formulated from acrylic resins that will not crack and discolor with
age. For this experiment, Cytoseal 280 and 60 was used. Cytoseal 280 is a high
viscosity medium for application where minimum spreading is desired or in cases
where a cover glass is not used. Cytoseal 60 is a low viscosity medium that dries
quickly and allows for a rapid even spreading which nearly eliminates air bubbles.
Pictures of the EuD4TEA mounted in Cytoseal 280 and 60 are shown in Fig. 7.67.
Notice that Cytoseal 280 created a thick layer coating whereas Cytoseal 60 created
a very thin layer coating. It should be noticed that the feel of the Cytoseal 60 coating
still leaves the familiar powder feeling that was present for EuD4TEA. In the case of
Cytoseal 280, the layer was so thick that all that was felt was the Cytoseal
coating [126].
Fig. 7.67 Close-up pictures of 0.10 g of EuD4TEA mounted in Cytoseal (a) 280 and (b) 60 [126]
7.5.8.2 Synthesis of EuD4TEA and PMMA Compounds
The synthesis of EuD4TEA and PMMA was based on the procedures and methods
previously described. For these experiments, the PMMA (Sigma-Aldrich® 182265)
was crystalline and had an average molecular weight of 996,000. The first test
synthesis began by dissolving 4 mmol of the europium nitrate in 25 mL of
laboratory grade acetone. Then 13 mmol DBM and 14 mmol of TEA were added
to the solution. The EuD4TEA solution was then poured into a small icetray until
each section was filled about half way with the liquid. Afterwards, PMMA was
added until the EuD4TEA solution was almost overflowing into the next section of
the icetray. This created about a 50 % volume ratio of EuD4TEA solution and
PMMA. After each section was filled, the solution was left to form over the course
of a week. The end product that formed was a hard yellow object as shown in
Fig. 7.68a. This EuD4TEA PMMA compound was extremely hard to break; how-
ever, when broken it showed a quick flash of red light. In addition, this compound
exhibited a strong luminescence as shown in Fig. 7.68b. This was the same type of
PL that was present in pure EuD4TEA. Due to its extreme hardness and large
amount of energy required for fracture, no further testing was completed.
The EuD4TEA PMMA compound in this form would likely be good for use in
ballistic or hypervelocity impact conditions [126].
The synthesis of a second EuD4TEA and PMMA test compound was prepared in
a similar manner. For this synthesis, 4 mmol of the europium nitrate was dissolved
in 25 mL of laboratory grade acetone. Then 4.0 g of PMMA and 13 mmol DBM was
added to the solution. Once dissolved, 14 mmol of TEA were added. The solution
was then poured into the ice tray sections and left to dry. The product that formed
consisted of a brittle yellow-bottomed polymer with completely clear sides as
shown in Fig. 7.69. One interesting observation was that the clear sections that
formed appeared to be a thin coat that glowed bright red in sunlight as shown in
Fig. 7.69. This red glow was more pronounced and easily visible when brought
outside around noon. Every noticeable clear section of the EuD4TEA and PMMA
compounds glowed bright red under sunlight and no other forms of excitation. In
Fig. 7.68 Block of PMMA under (a) standard room light and (b) UV irradiation [126]
Fig. 7.69 EuD4TEA and

PMMA compound under
natural lighting conditions
[126]
addition, this compound yielded the same bright red triboluminescent flash as the
plain EuD4TEA [126].
In order to determine the effects of PMMA on the TL of EuD4TEA, a final run
was completed. This run began by placing a varying amount of PMMA into the
different sections of the ice tray. The amounts of PMMA used were 0.0096, 0.0148,
0.0203, 0.0408, 0.0524, 0.0999, 0.2044, 0.3006, 0.5012, and 1.0632 g. The solution
of EuD4TEA was doubled using 8 mmol of the europium nitrate dissolved in 50 mL
of laboratory grade acetone with the addition of 26 mmol of DMB and 28 mmol of
TEA. Once the products were dissolved, 5 mL of the EuD4TEA solution was poured
into each of the sections of PMMA. The EuD4TEA solution and PMMA were mixed
until everything was dissolved and then left to dry over the course of a week [126].
The products that formed are shown in Fig. 7.70. As the figure indicates, PMMA
had a significant effect on the structure of the finished compound. At low concen-
trations (<0.0203 g) the end product is a pasty sticky yellow compound as shown in
Fig. 7.70a. As the PMMA increased, the compound started forming one solid layer.
However, there was not enough PMMA to keep the compound all together when
removing it from the tray as Fig. 7.70b indicates. Instead, generally one solid block
and a flaky compound formed. The amount of flakes reduced as the PMMA
increased until 0.5012 g was reached. At this concentration, a solid block com-
pound was formed as shown in Fig. 7.70c, d. This compound had a smooth side with
internal pimples (Fig. 7.70c) produced by air bubbles and a rough uneven side as
shown in Fig. 7.70d. All of these materials exhibited PL and TL [126].
7.5.8.3 Synthesis of EuD4TEA and Phthalocyanine Compounds
The synthesis of EuD4TEA and phthalocyanine (Pc) was similar to PMMA. Pc is an

intensely colored macrocyclic compound that is used as dyeing agent. In addition, it
forms coordination complexes with most elements of the periodic table [127]. The
synthesis began by putting various amounts of the Pc into the various sections of the
ice tray. The amounts used were 0.0033, 0.0074, 0.0128, 0.0251, 0.0670, 0.2058,
0.4431, 1.3050, 1.3320, and 1.9953 mmol. The solution of EuD4TEA was doubled
Fig. 7.70 End product of the 5 mL EuD4TEA solution with (a) 0.1048, (b) 0.0203, and (c, d)
1.0632 g of PMMA [126]
using 8 mmol of the europium nitrate was dissolved in 50 mL of laboratory grade

acetone to which 26 mmol of DMB and 28 mmol of TEA were added. Once the
products were dissolved, 5 mL of the EuD4TEA solution were poured into each of
the sections of Pc. The solutions were then stirred to try to dissolve all of the Pc and
left to dry over the course of a week [126].
Similar to PMMA, Pc had a significant effect on the structure of the EuD4TEA
and Pc compound as shown in Fig. 7.71. Small concentrations of Pc produced a
compound that was a dark yellow with blue green spots as shown in Fig. 7.71a.
However, as the concentration increased so did the amount of blue in the compound.
In fact, as the concentration increased, the yellow started to turn to brown and the
blue started to turn into a dark navy color as shown in Fig. 7.71b. Once the
concentration reached approximately 1.3 mmol of Pc, a very brittle blue/brown
material formed as shown in Fig. 7.71c. As the Pc was increased even more, the Pc
did not dissolve as is evident by the blue powder shown in Fig. 7.71c. One interesting
observation was that as the amount of Pc increased, the PL also decreased. This was
due to the Pc absorbing red light. In fact, the 1.9953 mmol of Pc compound had a
Fig. 7.71 End product of the 5 mL EuD4TEA solution with (a) 0.0033, (b) 0.2058, and (c)
1.9953 mmol of Pc [126]
near 100 % quenching of luminescence. The only observed luminescence under UV

irradiation was at the edges of the compound. As a result, Pc can be used to create
compounds or thin films that do not emit PL, but still emit TL [126].
7.5.8.4 Photoluminescent Results
The measured photoluminescent emission spectra for EuD4TEA compounds are

shown in Figs. 7.72, 7.73, 7.74 where the compounds are Cytoseal, PMMA, and Pc
respectively. The figures show that the photoluminescent spectra did not change when
the PMMA or Pc amounts are increased. The only slight differences between the
spectra of the different media were due to some compounds having a lower PL. As a
result, the lower probability transitions were buried in the noise. The spectra from all
three compounds indicate that the luminescence from EuD4TEA compounds came
from the typical Eu3þ excited transitions from the 5D0 levels to the lower 7F0–4 levels.
7.5.8.5 Triboluminescent Results
Using the drop tower and LabVIEW VI tools, the TL produced from each trial was
analyzed. Due to these measurements being a relative value, the data was normal-
ized to a reference value. In the case of Cytoseal, the reference value was the first
drop for each type of Cytoseal. The error in the triboluminescent yield in these
experiments was estimated to be 7 %. From Fig. 7.75, it was apparent that as the
number of drops increase, the TL decreases exponentially. This data is consistent
with previous findings [86]. In addition, it appears that the type of Cytoseal did not
have any effect on the triboluminescent light yield [126].
The LabVIEW VI tools were used to measure the decay time for each drop. The
average triboluminescent decay time for the EuD4TEA in Cytoseal 60 was
637.8 66.9 μs. The error in this case was estimated to be the standard deviation
of all the drops. This decay time was consistent with previous measurements for
EuD4TEA. The average triboluminescent decay time for EuD4TEA in Cytoseal
280 was 685.2 70.2 μs. This measurement does fall within the error of previous
Fig. 7.72 Measured photoluminescent spectra of EuD4TEA and Cytoseal [126]
Fig. 7.73 Selected measured photoluminescent spectra of EuD4TEA polymers for the varying
PMMA (g) precursor quantities. The orange red inset is a picture of the photoluminescent emission
produced by the EuD4TEA polymers under UV irradiation [126]
Fig. 7.74 Selected measured photoluminescent spectra of EuD4TEA polymers for the varying Pc
(mmol) precursor quantities [126]
Fig. 7.75 Triboluminescent light yield for EuD4TEA in Cytoseal [126]

Fig. 7.76 Triboluminescent light yield for EuD4TEA PMMA compound [126]
measurements. The reason for the longer decay time was likely due to the fact that
the TL is very dim for Cytoseal 280. As a result, the noise is increased and the amp
had to be set to a high setting. This likely caused an increase in the triboluminescent
decay time [126].
The effects of PMMA on the TL of EuD4TEA are shown in Fig. 7.76. For these
measurements, the data was normalized to the 0.1627 vol.% of PMMA/EuD4TEA.
The error in vol.% was estimated to be 5 % and the error in TL yield was estimated
to be 7 % which is the error from the synthesis process and drop tower [89]. As
Fig. 7.62 shows, the TL is dependent on the amount PMMA. The TL starts out small
due to the elastic nature of the PMMA. Then it reaches a maximum tribolumines-
cent yield at about 3.4 % PMMA to EuD4TEA solution. Once this concentration
was reached the TL decreases rapidly. This was due to the fact that it was becoming
harder for the ball to break the EuD4TEA PMMA compound. In fact, when the
concentration reaches about 8.5 % it becomes nearly impossible to break the
polymer compound. This was the case for the 18 % PMMA. Here it took two
drops just to split the compound in half. As a result, the amount of PMMA can be
increased or decreased to optimize the impact energy under investigation. The
decay time for the EuD4TEA PMMA compound was 570.9 39.6 μs which is
consistent with previous research [126].
The effects of Pc on the triboluminescent properties of EuD4TEA are shown in
Fig. 7.63. The triboluminescent light yield was normalized to the 0.0033 mmol light
yield value. The error in light yield was estimated to be 7 % [89]. As Fig. 7.77
shows, the triboluminescent decays exponentially as the Pc was increased. This was
likely due to the fact that Pc has an absorption energy that is similar to the emission
of EuD4TEA. As a result, the larger the amount of Pc that was present the greater
Fig. 7.77 Triboluminescent light yield for EuD4TEA Pc compound [126]
the quenching. In addition, as the amount of Pc increased, the compound became

very brittle. As a result, most of the compound was destroyed with only one drop,
thereby completely quenching the TL, and thus indicating that Pc would not be a
suitable material for impact sensors [126].
7.6 Conclusions
An extensive quantity of good research has been completed on the triboluminescent

properties of EuD4TEA and related compounds. This research has shown that the
EuD4TEA synthesis could be optimized to increase the yield, TL light, and reduce
the cost to approximately $50 per gram. The results from this research also indicate
that EuD4TEA can be used as the active element for structural health monitoring
and radiation sensors. For the first time, a real time radiation dose sensor consisting
of only a light source such as a laser or LED, photodiode detector, and a coating of
EuD4TEA can be utilized. This detector can be placed in high radiation zones where
standard radiation badges cannot be placed without shutting down equipment and
waiting for the radiation levels to decrease.
One of the most significant findings from this research was the fact that the
addition of dopants can greatly enhance the triboluminescent properties of
EuD4TEA. Clearly DBP was the best dopant from this research showing an increase
in the TL nearly sixfold.
Appendix 1: Synthesizing EUD4TEA
The materials required to synthesize EuD4TEA include the following:

1. Europium (III) nitrate hydrate, powder, 99.999 % (Metall Rare Earth
Limited, 6371).
2. 1,3-diphenyl-1,3-propanedione (dibenzoylmethane), powder, 98 % (Sigma-
Aldrich, D33454).
3. Triethylamine (TEA), liquid, 99.5 % (Sigma-Aldrich, 471283).
4. 25 mL of solvent.
(a) Ethanol, 200 proof, anhydrous, 99.5 % (Sigma-Aldrich, 459836).
(b) Ethyl alcohol, denatured (Fisher Scientific, A407).
(c) Acetone, laboratory reagent, 99.5 % (Sigma-Aldrich, 179973).
(d) Methanol, anhydrous (Sigma-Aldrich, 322415).
(e) Ethyl alcohol, denatured laboratory grade, 95 % (Scholar Chemistry,
9506406).
5. 125 mL Erlenmeyer flask.
6. Whatman ashless, grade 42 filter paper, 125 mm diameter (Sigma-Aldrich,
Z241113).
7. Powder funnel, polypropylene, 100 mm diameter (Ward Scientific, 6187107).
8. Hot plate.
9. Storage container, 1 oz clear round wide mouth jars (Uline, S-14487).
The common manufacturers for making EuD4TEA described in this chapter are
listed. It should be noted that other manufacturers could be used as long as the
purity is similar. The user must be careful, however, when using solvents. While
most solvents are similar between manufacturers, denatured ethyl alcohol varies
greatly. For example, using the Fisher brand denatured ethyl alcohol will yield
similar results to the pure ethyl alcohol, acetone, and methanol. However, the
Scholar Chemistry version will not yield similar results. These variations were
observed in both the denatured and denatured anhydrous versions of ethyl alcohol.
As a result, it is imperative that one solvent be used throughout the same experi-
mental run in order to do batch comparison.
The steps to synthesize EuD4TEA are as follows:
1. Set the hot plate to a low heat, between 2 and 3, under a vent hood.
2. Transfer 25 mL of the desired solvent to an Erlenmeyer flask.
3. Place the Erlenmeyer flask on the hot plate.
4. Add 4 mmol (1.44 g) of europium nitrate to the solvent.
5. Add 13 mmol (2.93 g) of DBM to the solution.
6. Place a funnel on top of the Erlenmeyer flask to prevent solvent loss by
evaporation.
7. Mix the slurry over a hot plate until completely dissolved.
8. Remove from the hot plate.

9. Add 14 mmol (2 mL) of TEA to the solution.
10. Place a paper towel over the Erlenmeyer flask to ensure that nothing is added
while crystallizing.
11. Set aside to cool overnight.
12. Once precipitation is complete, fold the filter paper in half twice.
13. Make a funnel out of the filter paper and place it inside a funnel.
14. Swirl the solution inside the Erlenmeyer flask so that the crystals break free
from the Erlenmeyer flask.
15. Pour the solution into the funnel.
16. Let the solution pass through the paper so that only the crystals remain.
17. Repeat steps 12–14 until all the crystals have been transferred to the beaker into
the filter paper.
18. Let dry for a few days.
19. Once dry, transfer to a container for storage.
Dopants can enhance the luminescence of many compounds. Depending on the
type of dopant, a new synthesis procedure is required. For compound dopants such
as caffeine or DMMP, the synthesis method is modified slightly. Step five above is
replaced with add the required amount of dopant. The rest of the synthesis process
remains unchanged. If the results from the added dopant does not take the shape of a
Gaussian curve or the data is scattered, then the solution must be heated. It has been
found that in the case of elemental dopants such as dysprosium, samarium, or
uranium requires heating. The steps to synthesize EuD4TEA doped with elements
are as follows:
1. Set the hot plate to 250 C under a vent hood.
2. Set the stirrer to maximum.
3. Add 75 mL of the desired solvent to the Erlenmeyer flask.
4. Place the Erlenmeyer flask on the hot plate.
5. Add 4 mmol (1.44 g) of europium nitrate to the solution.
6. Add the desired amount of dopant.
7. Add 13 mmol (2.93 g) of DBM to the solution.
8. Place a funnel upside down on top of the Erlenmeyer flask to ensure that most
of the solvent does not evaporate completely.
9. Let mix for 20 min.
10. Remove from the hot plate.
11. Remove the stirrer.
12. Add 14 mmol (2 mL) of TEA to the solution.
13. Place a paper towel over the many Erlenmeyer flask to ensure that nothing is
added while crystallizing.
14. Set aside to cool overnight.
15. Once precipitation is complete, fold the filter paper in half twice.
16. Make a funnel out of the filter paper and place it inside a funnel.
17. Swirl the solution inside the Erlenmeyer flask so that the crystals break free
from the Erlenmeyer flask.
18. Pour the solution into the funnel.

19. Let the solution pass through the paper so that only the crystals remain.
20. Repeat steps 12–14 until all the crystals have been transferred to the Erlen-
meyer flask into the filter paper.
21. Let dry for a few days.
22. Once dry, transfer to a many container for storage.
Appendix 2: Directions for Drop Tower
The directions on how to perform the drop tower test are as follows:
1. Get a new Plexiglas.
2. Set the Photodiode 2.25 cm away from the bottom of the Plexiglas.
3. Set the Amplifier to the appropriate settings (200 μA for ZnS:Mn and
EuD4TEA).
4. Place 1.068 0.005 g of material in the center of the drop tower.
5. Place the pin and ball 42 in. from the sample.
6. Set the oscilloscope to the required settings.
(a) Channel 1: 2.00 V (varies between materials).
(b) M Time Scale ¼ 500 μs (EuD4TEA), 250 μs (ZnS:Mn).
(c) Center Position M ¼ 780 μs.
(d) CH 1 trigger ʃ ¼ 80 mV
(e) Probe attenuation ¼ 20
7. Release the ball (130.276 g).
8. Save the light duration and repeat until you have five good data sets.
9. Clean out the drop tower and place the powder and Plexiglas into a storage bag
making the date, material, and number of drops.
10. Copy the data to the computer.
11. Run the Area under curves2 LabVIEW program to determine the light yield.
References
Journal of Luminescence, 131, 1407.
2. Childs, P., Wong, A. C., Terry, W., & Peng, G. D. (2008). Measurement Science and
Technology, 19, 65301.
3. Chong, K. P., Carino, N. J., & Washer, G. (2003). Smart Materials and Structures, 12, 483.
4. Merzbacher, C. I., Kersey, A. D., & Friebele, E. J. (1996). Smart Materials and Structures, 5,
196.
5. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2011). Materials Today,
14, 292.
6. Walton, A. J. (1977). Advances in Physics, 26, 887.
7. Chandra B. (1998). In D. R. Vij (Ed.), Luminescence of solids (pp. 361–389). New York, NY:
Plenum.
8. Chandra, B. P., & Zink, J. I. (1980). Physical Review B, 21, 816.
9. Meyer, K., Obrikat, D., & Rossberg, M. (1970). Krist and Tech, 5, 5.
10. Sage, I. C., & Bourhill, G. (2001). Journal of Materials Chemistry, 11, 231.
11. Sage, I. C., Badcock, R., Humberstone, L., Geddes, N. J., Kemp, M., & Bourhill, G. (1999).
Smart Materials and Structures, 8, 504.
12. Chandra, B., Baghel, R. N., & Chandra, V. (2010). Chalcogenide Letters, 7, 1.
13. Sage, I. C., Humberstone, L., Oswald, I., Lloyd, P., & Bourhill, G. (2001). Smart Materials
and Structures, 10, 332.
14. Sage, I. C. & Geddes, N. J. (1999). US Patent 5905260.
15. Womack, F. N. (2004). Development of a drop tower to study the triboluminescence of ZnS:
Mn with attention to possible applications to spacecraft. Lafayette, LA: University of
Louisiana at Lafayette.
16. Xu, C. N., Zheng, X. G., Akiyama, M., Nonaka, K., & Watanabe, T. (2000). Applied Physics
Letters, 76, 179.
17. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Applied Physics Letters, 74,
1236.
18. Goods, S. H., Dentinger, P. M., & Whinnery, L. L. (2003). FLASHFOAM: a triboluminescent
polymer foam for mechanical sensing. Livermore, CA: Sandia National Laborato.
19. Dickinson, J. T., Jensen, L. C., & Bhattacharya, S. K. (1985). J Vac Sci Technol A Vacuum
Surfaces Film, 3, 1398.
20. Leverenz, H. (1950). Introduction to luminescence of solids. New York, NY: Wiley.
21. Gribkovskii, V. P. (1988). In D. R. Vij (Ed.), Luminescence of solid (1st ed., pp. 1–40).
New York, NY: Plenum.
22. Williams, F. (1966). Luminescence of inorganic solids (1st ed.). New York, NY: Academic.
23. Nakazawa, E. (2006). In Yen, W. M., S. Shionoya, & H. Yamamoto (Eds.) Phosphor
handbook (2nd ed., pp. 11–70). Boca Raton, FL: CRC Press.
24. Ronda, C. R. (2007). In C. R. Ronda (Ed.), Luminescence: From theory to applications (1st
ed., pp. 1–34). Weinheim: Wiley.
25. Nakazawa, E. (2006). In W. M. Yen, S. Shionoya, & H. Yamamoto (Eds.) Fundamentals of
phosphors (1st ed., pp. 1–129). Boca Raton, FL: CRC Press.
26. Yen, W. M., & Weber, M. J. (2004). Inorganic phosphors: Compositions, preparation and
optical properties (1st ed.). Boca Raton, FL: CRC Press.
27. Yen, W. M., Shionoya, S., & Yamamoto, H. (2006). Practical applications of phosphors (1st
ed.). Boca Raton, FL: CRC Press.
28. Pascoe, K. J. (1973). Chemical and physical properties of materials for electrical engineers.
New York, NY: Wiley.
29. Serway, R. A., Moses, C. J., & Moyer, C. A. (2004). Modern physics (3rd ed.). Belmont, CA:
Brooks Cole.
30. Csele, M. (2004). Fundamentals of light sources and lasers (1st ed.). Hoboken, NJ: Wiley.
31. Newbacher, C., Jatla, A. (2003). In R. A. Charvat (Ed.), Color. Plast (2nd ed., pp. 242–257).
New York, NY: John Wiley & Sons, Inc., New York, NY.
32. Pringsheim, P., & Vogel, M. (1943). Luminescence of liquids and solids and its practical
applications. New York, NY: Interscience Publishers, Inc.
33. Hitz, C. B. (1991). Understanding laser technology: An intuitive introduction to basic and
advanced laser concepts (2nd ed.). Tulsa, OK: Pennwell Corp.
34. Sze, S. M. & Ng, K. K. (2006) In Physics of semiconductor devices (3rd ed., pp. 7–16).
New York, NY: John Wiley & Sons.
35. Lakowicz, J. R. (2006). In Principles of fluorescence spectroscopy (3rd ed., pp. 1–25).
New York, NY: Springer.
36. Hollerman, W. A., Fontenot, R. S., Bhat, K. N., Aggarwal, M. D., Guidry, C. J., & Nguyen,
K. M. (2010). Optical Materials (Amsterdam), 34, 1547.
37. Fontenot, R. S., Hollerman, W. A., Aggarwal, M. D., Bhat, K. N., & Goedeke, S. M. (2012).
Measurement, 45, 431.
38. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Materials Research Bulletin,
34, 1491.
39. Lakshmanan, A. (2007). In Lumin. disp. phosphors phenom. appl. (1st ed., pp. 9–31).
New York, NY: Nova Science Publishers Inc.
40. Sweeting, L. M., Cashel, M. L., Dott, M., Gingerich, J. M., Guido, J. L., Kling, J. A.,
et al. (1992). Molecular Crystals and Liquid Crystals, 211, 389.
41. Camara, C. G., Escobar, J. V., Hird, J. R., & Putterman, S. J. (2008). Nature, 455, 1089.
42. Xu, C. N., Zheng, X. G., Watanabe, T., Akiyama, M., & Usui, I. (1999). Thin Solid Films,
352, 273.
43. Freund, F. T. (2003). Journal of Scientific Research, 17, 37.
44. Chandra, V., & Chandra, B. (2012). Journal of Luminescence, 132, 858.
45. Chandra, B., & Bisen, D. P. (1992). Physica Status Solidi, 132, K101.
46. Chandra, B., Mahobia, S. K., Jha, P., Kuraria, R. K., Kuraria, S. R., Baghel, R. N.,
et al. (2008). Journal of Luminescence, 128, 2038.
47. Akiyama, M., Nishikubo, K., & Nonaka, K. (2003). Applied Physics Letters, 83, 2003.
48. Xu, C. N., Yamada, H., Wang, X., & Zheng, X. G. (2004). Applied Physics Letters, 84, 3040.
49. Chandra, B., Xu, C. N., Yamada, H., & Zheng, X. G. (2010). Journal of Luminescence, 130,
442.
50. Kim, J. S., Kibble, K., Kwon, Y. N., & Sohn, K.-S. (2009). Optics Letters, 34, 1915.
51. Chandra, B., Baghel, R. N., Luka, A. K., Sanodiya, T. R., Kuraria, R. K., & Kuraria, S. R.
(2009). Journal of Luminescence, 129, 760.
52. Jia, J., Yei, M., & Jia, W. (2006). Optical Materials (Amsterdam), 28, 974.
53. Xu, C. N. (2002). In Encycl. Smart Mater (pp. 190–201). New York, NY: John Wiley & Sons,
Inc.
54. Reynolds, G. T., & Austin, R. H. (2000). Journal of Luminescence, 92, 79.
55. Chandra, B. (2011). Journal of Luminescence, 131, 1203.
56. Sohn, K.-S., Seo, S. Y., Kwon, Y. N., & Park, H. D. (2002). Journal of the American Ceramic
Society, 85, 712.
57. Kim, J. S., Kwon, Y.-N., Shin, N., & Sohn, K.-S. (2007). Applied Physics Letters, 90, 241913.
58. Akiyama, M. (2002). Journal of Luminescence, 97, 13.
59. Akiyama, M., Xu, C. N., Matsui, H., Nonaka, K., & Watanabe, T. (1999). Applied Physics
Letters, 75, 2548.
60. Chandra, B., Sonwane, V. D., Haldar, B. K., & Pandey, S. (2011). Optical Materials
(Amsterdam), 33, 444.
61. Zheng, H., Terasaki, N., Yamada, H., & Xu, C. N. (2009). Japanese Journal of Applied
Physics, 48, 04C109.
62. Reddy, D. R., & Reddy, B. K. (2002). Applied Physics Letters, 81, 460.
63. Li, C., Adachi, Y., Imai, Y., Nishikubo, K., & Xu, C. N. (2007). Journal of the Electrochem-
ical Society, 154, J362.
64. Alzetta, G., Chudáček, I., & Scarmozzino, R. (1970). Physica Status Solidi, 1, 775.
65. Goedeke, S. M., Allison, S. W., Womack, F. N., Bergeron, N. P., & Hollerman, W. A. (2001).
Triboluminescence and its application to space-based damage sensors. Oak Ridge, TN: Oak
Ridge Nation Laboratory.
66. Hollerman, W. A., Allison, S. W., Goedeke, S. M., Boudreaux, P., Guidry, R., & Gates,
E. (2003). IEEE Transactions on Nuclear Science, 50, 754.
67. Hollerman, W. A., Goedeke, S. M., Bergeron, N. P., Muntele, C. I., Allison, S. W., Ila,
D. (2005). Nuclear Instruments and Methods in Physics Research Section B: Beam Interac-
tions with Materials and Atoms, 241, 578.
IEEE Transactions on Nuclear Science, 51, 1737.
69. Bergeron, N. P., Hollerman, W. A., Goedeke, S. M., & Moore, R. J. (2008). International
Journal of Impact Engineering, 35, 1587.
et al. (2006). International Journal of Impact Engineering, 33, 91.
71. Bergeron, N. P. (2006). Detection of triboluminescence from ZnS:Mn and ZnS:Cu. Lafayette,
LA: University of Louisiana at Lafayette.
72. Malespin, C. A. (2007). Detecting triboluminescent light from frontal meso-velocity impacts.
Lafayette, LA: University of Louisiana at Lafayette.
73. Fontenot, R. S. (2010). Measuring the triboluminescence characteristics of ZnS:Mn gener-
ated during ballistic impacts. Lafayette, LA: University of Louisiana at Lafayette.
74. Fontenot, R. S., Hollerman, W. A., & Goedeke, S. M. (2011). Materials Letters, 65, 1108.
75. Fontenot, R. S., Hollerman, W. A., Steuart, M. S., & Broussard, B. M. (2009). AIP Confer-
ence Proceedings, 1195, 1413.
76. Hurt, C. R., Mcavoy, N., Bjorklund, S., & Flipescu, N. (1966). Nature, 212, 179.
77. Sweeting, L. M., & Rheingold, A. L. (1987). Journal of the American Chemical Society, 109,
2652.
78. Cotton, F. A., Daniels, L. M., & Huang, P. (2001). Inorganic Chemistry Communications, 4,
319.
79. Carnall, W. T., Fields, P. R., & Rajnak, K. (1968). The Journal of Chemical Physics, 49,
4450.
80. Martin, W. C., Zalubas, R., & Hagan, L. (1978). Natl Stand Ref Data Syst NSRDS-NBS 60
(pp. 1–411). United States National Bureau of Standards, Washington, DC.
81. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., Allison, S. W., & Aggarwal, M. D. (2013).
Journal of Theoretical and Applied Physics, 7, 30.
82. Aggarwal, M. D., Penn, B. G., Miller, J., Sadate, S., & Batra, A. K. (2008). Triboluminescent
materials for smart optical damage sensors for space applications. Huntsville, AL: Marshall
Space Flight Center.
83. Takada, N., Peng, J., & Minami, N. (2001). Synthetic Metals, 121, 1745.
84. Hollerman, W. A., Bergeron, N. P., Womack, F. N., Goedeke, S. M., & Allison, S. W. (2004).
IEEE Transactions on Nuclear Science, 51, 1080.
85. Hollerman, W. A., Goedeke, S. M., Moore, R. J., Boatner, L. A., Allison, S. W., & Fontenot,
R. S. (2007). In 2007 I.E. Nucl. Sci. Symp. (pp. 1368–1372). Honolulu, HI: IEEE.
86. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2012). Journal of
Theoretical and Applied Physics, 6, 15.
87. Ralchenko, Y, Kramida, A. E., & Reader, J. (2011).
Luminescence, 132, 1812.
89. Hollerman, W. A., Fontenot, R. S., Bhat, K. N., & Aggarwal, M. D. (2012). Metallurgical and
Materials Transactions A, 43, 4200.
90. Sweeting, L. M., & Rheingold, A. L. (1988). Journal of Physical Chemistry, 92, 5648.
91. Hocking, M. B., Vandervoort, M., Francies, W., McKiernan, J., & Zink, J. I. (1992). Journal
of Luminescence, 51, 323.
92. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., Aggarwal, M. D., & Nguyen, K. M. (2012).
CrystEngComm, 14, 1382.
93. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., Aggarwal, M. D. (2013). Adv. Mater. Lett.
(Accepted) (2013).
94. Yang, W. C. (2003). Handbook of fluidization and fluid-particle systems. Boca Raton, FL:
CRC Press.
95. Bevington, P. R. & Robinson, D. K. (2003). Data reduction and error analysis for the
physical sciences (3rd ed.). McGraw-Hill Science/Engineering/Math.
96. Bhat, K. N., Fontenot, R. S., Hollerman, W. A., & Aggarwal, M. D. (2012). International
Journal of Chemistry, 1, 100.
97. Li, D. P., Li, C. H., Wang, J., Kang, L. C., Wu, T., Li, Y. Z., et al. (2009). European Journal
of Inorganic Chemistry, 2009, 4844.
98. Fontenot, R. S., Hollerman, W. A., Meche, J. A., Williams, S. A., McRaie, C. M., & Baltz,
G. L. (2016). Solid State Sciences.
99. Zeng, X. R., Xiong, R. G., You, X. Z., & Cheung, K. K. (2000). Inorganic Chemistry
Communications, 3, 341.
100. Khan, M. T. H. (2007). In M. T. H. Khan (Ed.) Bioact heterocycles IV (1st ed., pp. 75–98).
Heidelberg: Springer.
101. Krishnamoorthy, B., & Rema, J. (2012). Handbook of herbs and spices, volume 1 (2nd ed.).
Boca Raton, FL: CRC Press.
102. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2012). Crystal Research
and Technology, 47, 573.
103. Burrows, H. D., & Kemp, T. J. (1974). Chemical Society Reviews, 3, 139.
104. Lermontov, A. S., Lermontova, E. K., & Wang, Y.-Y. (2009). Inorganica Chimica Acta, 362,
3751.
106. Hollerman, W. A., Glass, G. A., & Allison, S. A. (1999). MRS Online Proceedings Library,
560, 335.
107. Birks, J. B. (1951). Proceedings of the Physical Society Section A, 64, 874.
108. Hollerman, W. A., Fisher, J. H., Holland, L. R., & Czirr, J. B. (1993). IEEE Transactions on
Nuclear Science, 40, 1355.
109. Hollerman, W. A., Fisher, J. H., Ila, D., Jenkins, G. M., & Holland, L. R. (1995). Journal of
Materials Research, 10, 1861.
110. Hollerman, W. A., Goedeke, S. M., Bergeron, N. P., Moore, R. J., & Allison, S. W. (2005).
and L. A. Lewis. In E. W. Taylor (Ed.), Photonics Sp. Environ. X (pp. 58970F–10). San
Diego, CA: SPIE.
111. Hathaway, G. J., & Proctor, N. H. (2004). Proctor and Hughes’ chemical hazards of the
workplace (5th ed.). New York, NY: Wiley Interscience.
112. Milne, G. W. (Ed.). (2005). Gardner’s commercially important chemicals: Synonyms, trade
names, and properties (1st ed.). New York, NY: Wiley Interscience.
113. Allen, D. W., Tebby, J. C., & Van de Grampel, J. C. (2005). In D. W. Allen, J. C. Tebby, &
D. Loakes (Eds.), Organophosphorus chemistry (Vol. 34, p. 385). Cambridge: Royal Society
of Chemistry.
114. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2013). Sensors Trans-
ducers Journal, 149, 109.
115. Chandra, B., Chandra, V., Jha, P., Patel, R., Shende, S. K., Thaker, S., et al. (2012). Journal of
116. Koren H (2004) Illustrated dictionary and resource directory of environmental and occupa-
tional health, 2nd edn. Boca Raton, FL: CRC Press.
117. Pohanish, R. P., & Greene, S. A. (2009). Wiley guide to chemical incompatibilities (3rd ed.).
New York, NY: Wiley.
118. Pohanish, R. P. (2011). Sittig’s handbook of toxic and hazardous chemicals and carcinogens
(6th ed.). Berlin: Elsevier.
119. Fontenot, R. S., Bhat, K. N., Owens, C. A., Hollerman, W. A., & Aggarwal, M. D. (2015).
Journal of Luminescence, 156, 428.
120. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., Alapati, T. R., & Aggarwal, M. D. (2013).
ECS Journal of Solid State Science and Technology, 2, P384.
121. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., Aggarwal, M. D., & Penn, B. P. (2014).
Polymer Journal, 46, 111.
122. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2013). In M. S.
A. Moustafa (Ed.), Eur. Synth. Charact. Potential Appl. (1st ed., pp. 85–160). Hauppauge,
NY: Nova Science Publishers Inc.
123. Bhat, K. N., Fontenot, R. S., Hollerman, W. A., Alapati, T. R., & Aggarwal, M. D. (2012).
International Journal of Chemistry, 2, 165.
124. Mantia, F. L. (2002). Handbook of plastics recycling. Shrewsbury: Smithers Rapra
Technology.
125. Zollinger, H. (2001). Color chemistry (3rd ed.). New York, NY: Wiley-VCH.
126. Fontenot RS, Bhat KN, Hollerman WA, Aggarwal MD (2013) Polym Adv Technol
127. Miessler, G. L., & Tarr, D. A. (2010). Inorganic chemistry (4th ed.). Upper Saddle River, NJ:
Prentice Hall.
Chapter 8
Mechanoluminescence Induced by Acoustic
Cavitation
Nathan C. Eddingsaas
8.1 Introduction
The first published observation of mechanoluminescence (ML) was by Sir Francis

Bacon in 1605 in “On the Advancement of Learning,” where he noted that lumps of
sugar emitted light when scraped [1]. Since then the phenomenon has been studied
by a number of groups to determine which compounds exhibit ML, what causes
ML, what occurs during ML, as well as more recent studies of how this phenom-
enon can be harnessed for use in mechano-optical devices such as stress, damage,
fracture, and impact sensors (see ref. [2] for a comprehensive survey of ML
literature) [2]. The emission spectrum of ML is an excellent resource for the
study of the conditions that occur during ML; unfortunately, the photon emission
is normally weak from a given ML event, making it difficult to study. To better
understand the conditions during ML, brighter ML must be produced—either from
a greater number of events or greater intensity of a given event.
One way to increase the emission intensity of ML is to initiate ML via acoustic
cavitation of a slurry of mechanoluminescent crystals in a long chain alkane [3–
5]. To do this, an ultrasonic horn, or sonic disrupter, is immersed into the long chain
hydrocarbon and vibrated at 20 kHz. This vibration sends sound waves through the
liquid creating thousands of micron sized bubbles that collapse at ultrasonic speeds.
This launches a shockwave that will fracture organic crystals upon impact, as well
as pick up the mechanoluminescent crystals and violently collide them into each
other, causing many fracture events during the high speed interparticle collisions.
This increases the ML emission in two ways—first by sheer numbers of fracture
events, as thousands of collisions are possible each second, as well as by the great
N.C. Eddingsaas (*)

School of Chemistry and Materials Science, Rochester Institute of Technology,
85 Lomb Memorial Dr., Rochester, NY, USA
e-mail: ncesch@rit.edu

238 N.C. Eddingsaas
intensity of the fracture event. By using this technique, ML has been produced with
intensity greater than 1000 times that available from mechanical grinding. The
increase in intensity has allowed us to learn much greater detail about the ML event
itself.
This chapter focuses on the nature of the discharge produced during ML
discovered through the use of initiation by acoustic cavitation. We first discuss
the properties of acoustic cavitation that produces the intense ML emission. We
then discuss what was learned from the emission products observed, including the
reactive nature of the discharge as well as the properties of the discharge plasma
including temperature, electron density, and electron energy.
8.2 Background
The chemical effect of ultrasound does not arise from the direct interaction of the
acoustic field with chemical species on a molecular level [6–10]. Instead, the effects
are a consequence of acoustic cavitation: the formation, growth, and implosive
collapse of bubbles in a liquid irradiated with high-intensity ultrasound [11]. The
growth of each bubble is relatively slow, allowing for the accumulation of gas and
vapor within the bubble while the collapse in its final stage becomes very rapid,
with velocities of up to 1500 m s1, i.e., faster than the speed of sound within the
liquid [12, 13]. This high speed collapse effectively concentrates the diffuse energy
of sound by compressing the contents of the bubble. During the rapid compression,
the heating of the contents is faster than thermal transport, creating a short-lived
localized hot spot in an otherwise cold liquid. The high temperatures and pressures
generated within the bubble are the cause of homogeneous sonochemical reactions,
with some analogies to both pyrolysis and radiolysis. Along with sonochemical
reactions, the compressional heating of the bubble contents creates a ~50 ps to
nanosecond flash of light referred to as sonoluminescence; this is a concentration of
energy by 12 orders of magnitude [14, 15]. The conditions of acoustic cavitation
have been elucidated, with temperatures up to 15,000 K, pressures of hundreds to
thousands of bars, and cooling rates on the order of 1012 K s1 [16–20].
Acoustic cavitation also has effects on solids within a liquid exposed to ultra-
sonic irradiation. Near an extended surface, the collapse of the bubble is
nonspherical and a high-speed jet of liquid is driven into the surface. The high
speed jet deposits heat at the surface and causes damage and fracture, exposing new
surfaces and altering morphology [11, 21]. In addition, the supersonic collapse of
the bubble launches a shockwave into the liquid at speeds up to 5000 m s1
[22–25]. In a liquid–powder slurry, the shockwave produced will cause sonofrag-
mentation due to direct interaction with the crystals, as well as causing the small
particles to be accelerated to half the speed of sound, resulting in high velocity
interparticle collisions. These interparticle collisions generate temperatures of up to
3000 K at the point of impact [26, 27]. The sheer force of the shockwave and
8 Mechanoluminescence Induced by Acoustic Cavitation 239
interparticle collisions can cause dramatic changes in particle surface size, surface
morphology, composition, and reactivity.
8.2.1 Acoustic Cavitation
The irradiation of a liquid with ultrasound (which spans the frequency of roughly
15 kHz to 1 GHz) can cause the formation of cavities or bubbles within the liquid on
the rarefaction phase of the sound field and cause these cavities to collapse violently
during the compressional phase of the sound field; this is known as acoustic
cavitation.
The oscillation of a bubble in a liquid due to the presence of an acoustic field is
dependent on many variables including the applied acoustic pressure and fre-
quency, the concentration of the dissolved gas, the vapor pressure and viscosity
of the liquid. The behavior of a bubble is often described by two categories—stable
or transient cavitation [11, 28]. During stable cavitation, usually under very low
acoustic pressure, a bubble oscillates linearly about an equilibrium size for many
acoustic cycles. In contrast, a bubble undergoing transient cavitation typically
expands to a large size during the rarefaction phase of the sound field. As the
sound field switches from rarefaction to compression, the bubble still grows in size
against the pressure of the surrounding liquid. At a certain critical point in the
compression phase the bubble begins a runaway, or nonlinear, collapse. During the
nonlinear collapse, the wall velocity and acceleration are large enough and the
collapse time is short enough that there is almost no heat transfer between the
contents of the bubble and the bulk liquid; thus the contents of the bubble are heated
nearly adiabatically. One of the main differences between stable and transient
cavitation is the amount of energy that is concentrated within the bubble. Sonolumi-
nescence, sonochemistry, and ultrasonic damage are due to transient cavitation.
When an acoustic field of sufficient amplitude is propagated through a liquid,
putting the liquid under dynamic tensile stress, any preexisting inhomogeneities can
act as nucleation sites for liquid rupture. This results in the production of many
individual cavitation events (multibubble cavitation) [11, 28]. The sound field will
not be homogeneous throughout the liquid and different bubbles are acted upon by
different acoustic pressures leading to a multitude of bubble sizes, and subsequently
a variety of conditions produced within the bubbles. In addition, the bubbles will be
affected by the transverse wave moving through the liquid and the bubbles will
interact with each other, resulting in a collapse that is less than spherical. The large
amplitude of the sound field can cause the bubbles to expand to such a size that the
inertia keeps it growing, preventing complete collapse. When the bubble does
eventually collapse to its minimum radius, the large acoustic pressure and the
interaction with other bubbles causes catastrophic collapse resulting in the destruc-
tion or fragmentation of the bubble.
When an acoustic standing wave is applied to a liquid, pressure nodes and
antinodes are generated. A bubble within this standing wave will be forced toward
240 N.C. Eddingsaas
either the node or antinode depending on whether the bubble is larger (node) or
smaller (antinode) than the resonance size. Under the appropriate conditions, the
acoustic force can balance the buoyancy force and a bubble can be trapped just
above the pressure antinode resulting in the bubble being “acoustically levitated.”
A single-bubble can be acoustically levitated by the application of a fairly mild
sound pressure—in water, typically below 1.4 bar [29–31]. Transient cavitation can
still occur during single-bubble cavitation, but unlike during multi-bubble cavita-
tion, the bubble collapses in phase with the sound field each and every cycle and
therefore is sometimes instead referred to as “synchronous cavitation”. Single-
bubble cavitation is a very controlled environment and can therefore be examined
in considerable detail. On the other hand, it is multi-bubble cavitation that is
implemented for practical applications of sonochemistry and acoustically caused
mechanochemistry, and is what causes ultrasonic damage.
8.2.2 Acoustic Cavitation Near an Extended Surface
When a bubble collapses near an extended surface, the collapse becomes asym-
metric and a liquid microjet is formed that will impinge on the surface, causing
damage (Fig. 8.1) [11, 21, 23, 32]. The microjet can reach speeds of hundreds of
meters per second, and it often deposits enormous energy densities at the site of
impact. This large energy deposition can cause severe damage to the surface,
including to steel and titanium. In fact, damage caused by microjetting was one
of the first observations of cavitation in the late 1800s by Sir John Thornycroft and
Sydney Barnaby, as well as Sir Charles Parsons. Microjets from cavitation was
found to be the source of the damage to boat propellers that rendered them useless
after a short time [33].
The high pressure generated in the final stage of bubble collapse causes a shock
wave to be radiated outward from the bubble (Fig. 8.2) [22–25]. The shock wave
Fig. 8.1 Photograph of

liquid microjet formation
and high speed impact
during cavitation near a
solid surface, photograph
courtesy of L.T. Crum
Fig. 8.2 Images of the shock wave launched from the cavitation of a laser-generated bubble, time
between each image is ~96 ns, frame size is 1.5 1.8 mm2 and the exposure time was 5 ns.
Figure adapted from ref. [23]
from a single isolated bubble in water can be launched at velocities of ~5000 m s1
and the pressure of the shock front can reach in excess of 60 kbar. The shock wave
loses most of its energy shortly after it is launched, but it still travels through the
liquid at ~1500 m s1. It is the combined effects of microjets and shock waves that
are responsible for ultrasonic cleaning, many heterogeneous sonochemical events,
and cavitational erosion [34–36].
8.2.3 Acoustic Cavitation of Heterogeneous Slurries
For a bubble’s symmetry to be distorted by a surface during cavitation, the surface

must be several times larger than the resonance bubble size. Therefore, small solid
particles (<200 μm at an acoustic frequency of 20 kHz) within a liquid cannot
induce a microjet, yet damage and surface modification still occurs. There are two
ways in which cavitating bubbles cause damage and modify surfaces of small
particles within liquid. The first is high-speed interparticle collisions that result
from the shock wave radiated from the collapsed bubble [26, 27, 37–43]. The
second is fracture resulting directly from a propagating shockwave in the solution,
a condition known as sonofragmentation [44–47].
Suslick and coworkers have found that the turbulent flow and shock wave
produced by high intensity ultrasound can drive metal particles together at suffi-
ciently high speed to induce melting and fusing at the point of impact [26, 27]. A
series of transition-metal powders with increasing melting points were used to
probe the maximum temperature reached during the interparticle collision. The
irradiation of Cr (mp ¼ 1857 C) and Mo (mp ¼ 2617 C) powders in dodecane at
20 kHz and 50 W cm2 resulted in agglomeration of the particles, while irradiation of
W (mp ¼ 3410 C) did not cause agglomeration to occur. Using the melting points
of these metal powders, the effective transient temperature reached at the point of
impact during interparticle collision was found to be roughly 3000 C. In addition
the energy generated during impact was determined from the volume of the melting
region of impact. It was determined that the velocity of interparticle collision as a
result of acoustic cavitation had a lower bound of half the speed of sound.
242 N.C. Eddingsaas
Interparticle collision is the primary way in which fine metal particles within a
cavitation field are damaged, agglomerate, and are smoothed.
Organic crystals are more brittle than micron sized metal particles and as such
can be broken by the sheer force of the propagating shockwave launched from the
collapsing bubbles [45]. In a series of studies by the Suslick group they determined
the dominant cause of fracture of acetylsalicylic acid (aspirin) crystals. There are
four possible mechanisms for fragmentation of organic crystals within a cavitation
field: interparticle collision mentioned above, fracture by the sheer force of the
propagating shockwave, collisions of the particles with the walls of the container,
and collisions with the vibrating horn. Interparticle collisions were ruled out as the
major fragmentation route due to the fact that the rate of fragmentation was first
order and the insensitivity of the final crystal size to crystal loading level. Collisions
with the wall and horn were ruled out as the dominant cause of sonofragmentation
by studying the rate of fragmentation in experiments where the crystals were
isolated from the horn and vessel walls. The rate was not reduced by an appreciable
amount under isolation conditions. While all four mechanisms probably play a role
in sonofragmentation of organic crystals, direct interaction with the propagating
shockwave is the dominant pathway.
High intensity ultrasound has a profound effect on solids within a liquid, for both
extended surfaces and small particles. The external effects of microjets and shock
waves are the cause of many heterogeneous interactions causing erosion, modifi-
cation of surfaces, scission of polymers, ultrasonic cleaning, high speed
interparticle collisions, and fracture of brittle material and crystals.
8.2.4 Mechanoluminescence from the Fracture of Crystals
Mechanoluminescence is the general term for light that is produced from the
mechanical action on a solid. There are a number of different types of mechanolu-
minescence based on the mechanism of light production. For instance, tribolumi-
nescence is light produced from rubbing two dissimilar objects (the prefix “tribo”
refers to friction). There are other physical processes that lead to the production of
ML, including plastic or elastic deformation of materials such as γ and X-ray
irradiated alkali halides [48] and rubber [49], the separation of certain materials
such as peeling tape off a substrate [50], rubbing dissimilar materials together such
as the movement of mercury over glass [51], and from the fracture of many crystals
such as sucrose [52]. It is this last process, called fractoluminescence or fractome-
chanoluminescence (fracto-ML) that will be the focus of this chapter. The irradi-
ation of a slurry of organic crystals in a liquid causes fracture of crystals due to
direct interaction with the propagating shockwave and high intensity interparticle
collisions.
There are a number of methods that have been used to produce fracto-ML,
include crushing, scraping, and cleaving. The most common fracto-ML generation
method currently used is striking a sample with a known amount of impact energy
using a drop tower [53] or an air driven piston [54]. There are a number of more
exotic methods that have also been used; for example, high power laser pulses [55],
thermal shock [56, 57], and extreme pressures applied to a sample causing the
sample to fracture and be destroyed [58] have all been used to initiate fracto-ML.
The best example of ML from fracture is from piezoelectric crystals. When a
piezoelectric crystal is stressed perpendicular to its piezoelectric axis, polarization
of electrical charge is induced. At fracture, the two faces of the crack will have
opposite charges; as the faces separate, the local electric field created is sufficient to
cause dielectric breakdown of the solid and the intervening gases [59–61]. The two
most common types of emission from ML are from excited states of the crystal
itself and from excited states of the surrounding gas, and in many instances a
combination of both.
The intensity of fracto-ML is proportional to the force of the fracture event.
Chandra and Zink have studied crystals that exhibited ML only during a fracture
event, which was confirmed by graphing the ML signal versus the applied force,
and showed that there was an ML signal only when there was a sharp reduction in
force indicating that a fracture event had occurred [54, 62–64]. They also showed
that the greater the reduction in force (i.e., the larger the fracture) the greater the ML
intensity. The lifetime of the ML is dependent on the velocity of the piston used to
crush the crystal; the faster the impact, the shorter the ML lifetime and the greater
the peak intensity. In addition, they found a positive correlation between the total
intensity of the fracto-ML and the impact velocity of the piston [54, 63]. However,
when one of the crystals, coumarin, was cleaved rather than crushed, the total ML
intensity was independent of the piston velocity [62]. These results indicate that one
factor that governs the intensity of the ML is the total new surface area created. It
should be noted that the ML spectra for all of the crystals studied showed only gas
discharge or only crystal luminescence, never a combination of the two.
There are a few examples in the literature that examine the intensity of ML from
crystals that exhibit both gas discharge and crystal luminescence when the crystals
are fractured [65, 66]. Sweeting took several crystals whose ML spectrum consists
of both gas discharge and crystal luminescence and looked to see if the shape of the
spectrum was dependent on conditions of excitation, such as the material used to
grind the crystals and the forcefulness of the grinding. It was found that for all
crystals examined, the more forceful the grinding the greater the contribution from
gas discharge was to the ML spectrum, as can be seen in Fig. 8.3 of coumarin
ground gently and forcefully [66].
Subsequent studies by Sweeting observed that coumarin absorbed UV N2 emis-
sion lines, indicating that the N2 discharge lines originated within the propagating
crack from the fracture event [65]. When coumarin was recrystallized under other
gas environments other than N2, including He, CO2, and SF6, strong crystal lumi-
nescence ML was observed. As He, CO2, and SF6 have faint or no UV emission
lines it was determined that the crystal luminescence is a result of electron bom-
bardment of the crystal during fracture. These results indicate that the ratio of gas
discharge to crystal luminescence can be used as an indicator of the forcefulness of
a fracture event.
244 N.C. Eddingsaas
a b
Intensity (au)
300 400 500 600 300 400 500 600

Wavelength (nm)
Fig. 8.3 ML spectra of coumarin ground (a) genteelly and (b) forcefully. Note the appearance of
N2 emission lines, in addition to the crystal luminescence, when ground forcefully. Figure adapted
from ref. [66]
8.3 Bright Mechanoluminescence Induced by Acoustic

Cavitation
The intensity of fracto-ML will increase with increased number of fracture events
as well as the strength of the mechanical action imparted onto the solid that caused
the fracture event to occur. The creation of a cloud of cavitating bubbles in a liquid
containing mechanoluminescent crystals results in bright fracto-ML, due to fracture
from the propagating shock wave and high-intensity interparticle collisions. The
production of fracto-ML using high-intensity ultrasound has produced bright ML,
up to 1000-fold greater in intensity that from mechanical grinding. By varying the
solution properties, mainly vapor pressure, ML intensity is changed, indicating that
the intensity of the collapse and therefore propagating shockwave is important to
the overall intensity of the ML event. The bright ML has also revealed many new
emission lines that show that the plasma produced is also reactive.
8.3.1 Increase in Overall Emission Intensity
Sucrose is a very well-known mechanoluminescent crystal. Upon fracture in air the

swan bands of molecular nitrogen (C3Πu B3Πg) and bands from N2þ (B2Σþu X2
Σþg) are observed [52, 65, 67]. When the same mass of sucrose crystals are crushed
with a glass rod in dodecane rather than in air, the same emission lines, but weaker,
are observed (Fig. 8.4) [3]. The reduction in ML intensity can be due to a number of
causes: rescued nitrogen gas in contact with the crystal, decreased force of grinding,
and the possibility of liquid entering the crack, quenching the discharge, and
shortening the lifetime of the ML event. The reduction in N2 surrounding the
crystals is due to the low solubility of nitrogen in dodecane of (mole fraction
solubility of 1.31 103) [68]. The reduction in grinding force is caused by the
Fig. 8.4 ML spectra of

sucrose sonicated in Sonicated (dodecane, N2)
dodecane sparged with N2
(black line), crushed in air Crushed (air)
(red line), and crushed in Crushed (dodecane, N2)
ML Intensity
dodecane sparged with N2
(blue line), all at 300 K. The
emission spectra consist of
the N2 (C3Πu B3Πg) and
N2þ (B2Σþu – X2Σþg)
progressions. The intensity
from sonication is an order
of magnitude more intense
than from crushing by hand.
320 360 400 440

Wavelength (nm)
high viscosity of dodecane, which will slow down the glass rod, increase the
buoyancy of the sucrose crystals, and increase the turbulent flow within the liquid
caused by the motion of the glass rod.
When crystals of sucrose are sonicated in dodecane saturated with N2, the same
emission lines are observed, but are an order of magnitude more intense (Fig. 8.4)
[3]. The increase in intensity could be due to greater number of fracture events or an
increase in impact on the crystals—both a result of acoustic cavitation. The speed of
the glass rod during crushing was measured during the experiment, and was found
to be moving at ~0.1 m s1, while the speed of the shockwave generated during
sonication is thousands of meters per second. The speed at which the shockwave hit
the sucrose crystals is many orders of magnitude greater than from grinding.
Similar results are observed from other crystals that exhibit only gas discharge
ML, such as sodium chloride. These results alone cannot definitively show if the
increased ML intensity from acoustic cavitation is due to increased number of
events or greater impact velocity.
To distinguish between the possible causes of the increase in ML during
sonication, a crystal that exhibits both gas discharge and crystal luminescence
during ML can be used. If the increase in ML is due to increased number of events
then the intensity of both the gas discharge and crystal luminescence will increase
proportionally, while if the increase is due to increased force of fracture, the ratio of
gas discharge to crystal luminescence will increase, as has been observed from
coumarin [65, 66].
Resorcinol is a highly mechanoluminescent crystal that exhibits both gas dis-
charge and crystal luminescence. The crystal luminescence of resorcinol is from
290 to 360 nm. For the crystal luminescence to originate from fluorescence,
resorcinol would have to absorb photons with wavelengths less than 280 nm
(peak excitation wavelength for resorcinol is 276 nm). There have been no gas
discharge lines at these wavelengths detected from ML in air or nitrogen
246 N.C. Eddingsaas
environment, and therefore the crystal luminescence observed from resorcinol must
be a result of electron bombardment during the fracture event. When crushed in air
or in dodecane sparged with N2, crystal luminescence dominates the ML spectrum
along with weak N2 emission (Fig. 8.5) [4]. However, when resorcinol is sonicated
in dodecane sparged with N2, very strong N2 and N2þ emission is observed, and
upon first inspection no crystal luminescence is observed (Fig. 8.6a). Figure 8.6b
shows mechanoluminescence from resorcinol sonicated in dodecane and crushed
with a glass rod in dodecane, zoomed in to be able to observe the emission from
grinding. As can be seen, the crystal luminescence is still observable when the ML
is induced by acoustic cavitation; it is just dwarfed by the N2 emission lines. In fact,
Fig. 8.5 ML spectra of

resorcinol ground with a 800
Crushed (air)
glass rod in air (red line) Crushed (dodecane)
and dodecane sparged with
Intensity (counts)
N2 (black line). The

spectrum consists of both 600
crystal luminescence as
well as week N2 and N2þ
emission lines 400
200
0
280 320 360 400 440
Wavelength (nm)
a 120000
b 1000
Intensity (counts)
100000
800
80000
600
60000
400
40000
20000 200
0 0
280 320 360 400 440 280 320 360 400 440
Fig. 8.6 (a) ML from sonication of a slurry of resorcinol in dodecane sparged with N2. (b) ML
from resorcinol crushed in dodecane and sonicated on the same scale. There is a large increase in
N2 gas discharge lines, but there is negligible increase in crystal luminescence when ML is induced
by acoustic cavitation. Figure adapted from ref. [4]
Fig. 8.7 Near UV to near IR spectrum of ML discharge from sonication of resorcinol in hexadecane
sparged with N2. The spectrum contains N2 (C3Πu B3Πg) and N2þ (B2Σþu X2Σþg) progressions
(from 300 to 500 nm). The insert (~1000-fold intensity enlargement) shows very weak emission
from N2 (B3Πg A3Σþu), which has not been reported for any prior mechanoluminescent system
(presumably due to its inherently low emissivity). The slurry was 10 %w/w resorcinol in hexadecane,
and was irradiated with ultrasound at 20 kHz, 6 W/cm2 with a liquid temperature of 300 K.
the N2 and N2þ emission lines are over 1000-fold more intense from sonicating than
from crushing in dodecane.
Very strong N2 and N2þ emission is observed from the sonication of a slurry of
resorcinol in long chain hydrocarbons. Looking at the spectrum from 300 to 850 nm
as seen is Fig. 8.7, other N2 emission features are observed between 600 and 800 nm
belonging to the B3Πg A3Σþu progression of N2 [4]. These emission lines are
common in nitrogen discharges but had not previously been reported from a
mechanoluminescent event. Presumably they would be observed in other ML
events if the gas discharge was sufficiently intense.
The increase in gas discharge lines due to sonication could also arise from
frictional electrification caused by rubbing two dissimilar materials together (i.e.,
static charging). To determine whether crystal fracture or frictional electrification is
responsible for the observed ML, the sonication cell was equipped with replaceable
windows. The windows were varied from insulator (plexiglass) to semiconductor
(quartz and glass) to conductor (ITO coated glass); in otherwise identical experi-
ments, no change in the intensity of ML was observed [4]. This makes frictional
electrification an untenable explanation because no electrification could have
occurred with the conductive ITO window [69]. A large enhancement of the gas
discharge emission lines without subsequent enhancement in the crystal lumines-
cence indicates that the increase in ML during sonication is due to the forcefulness
of the fracture event.
248 N.C. Eddingsaas
8.3.2 Identification of Gas Discharge Lines and Their

Origins
ML has been observed in many gaseous environments other than air or N2. Light
from the fracture of many crystals has been seen in Ar, Ne, He, N2, O2, and even
SF6 (which has a very high dielectric strength and is used in high-voltage electrical
devices to prevent dielectric breakdown) [58, 67, 70–78]. For crystals whose ML
spectrum consists at least partially of crystal luminescence, crystal luminescence
still persists in other gases and under vacuum [65]. For crystals whose ML spectrum
consists, at least partially, of gas discharge, the emission lines are of the gas
atmosphere that is present, as long as the gas is highly luminous such as N2, Ar,
Ne, and He. However, if the gas emission lines have low luminous efficiencies
(such as CO2 and SF6) and the ML spectrum consists only of gas discharge, no ML
has been observed [79].
Nakayama has recorded the ML spectra from scratching a sapphire disk with a
diamond tip under several gaseous environments, and compared them to the plasma
created by a discharge in the same environments between Al electrodes [80]. While
this is not ML from fracture, the ML spectra are consistent with those produced
from ML upon fracture. The ML spectra in each of the environments resemble that
of the discharge plasma, with a few notable exceptions; the ML in a N2 atmosphere
lacks the N2 (B3Πg A3Σþu) (first positive system) that is typically observed in
normal discharges, and many of the weak He and O lines are not observed in the
ML spectra. The absence of the weak He and O lines in the ML spectra are due to
the low intensity of the signal, but the absence of the N2 first positive system has
never been explained. It could also be due to low intensity, but there appears to be
enough signal that it would be visible if it were present.
The intense gas discharge emission ML during acoustic cavitation has allowed
for detailed studies of the ML discharge event in gaseous environments other than
air or N2. The fact that the ML occurs within a liquid allows for good control of the
gaseous environment, as the solution can be continuously saturated with a given gas
thus providing a singular gaseous environment. The gas environment can be varied
from inert gases such as Ar, He, or N2 to reactive ones such as O2, CO2, or CH4.
This technique has been used to observe a large number of new emission lines from
ML, as well as provide a better understanding of the discharge event.
As previously mentioned, the ML from crushing resorcinol in air or in dodecane
gives crystal luminescence as well as N2 (C3Πu B3Πg) and the N2þ (B2Σþu X2Σ
þ
g) emission lines. When resorcinol is crushed in hexadecane sparged with He,
however, only the crystal luminescence is observed (Fig. 8.8) [4]. Also shown in
Fig. 8.8 is the ML induced by acoustic cavitation of resorcinol in hexadecane
sparged with He; strong He lines along with the crystal luminescence, as well as
weak N2 lines can be seen. The N2 lines are presumably from nitrogen trapped
within the crystal. Once again, the observation of the He lines from sonication is an
indication of the intense fracture caused by interparticle collision during acoustic
cavitation, whereas no He lines are detected from crushing. As controls, sonication
Sonicated (hexadecane, He)

Crushed (air)
Crushed (hexadecane, He) He
Normalized Intensity
Solid state PL
crystal
luminescence
N2
280 320 360 400

Wavelength (nm)
Fig. 8.8 Comparison of ML from Resorcinol sonicated in hexadecane sparged with He (black
line), crushed in air (green line), crushed in hexadecane sparged with He (red line) and its
photoluminescence (blue line) at 300 K. Strong He gas discharge is observed along with the
crystal luminescence when resorcinol is sonicated in hexadecane sparged with He; N2 emission is
also observed from trace N2 contamination. No He discharge is observed from grinding in the
presence of He gas. Figure adapted from ref. [3]
of hexadecane itself or of a hexadecane slurry with hydroquinone

(non-piezoelectric, non-mechanoluminescent analog of resorcinol) at the same
acoustic intensity and saturated with He gives no observable luminescence. When
other mechanoluminescent crystals are sonicated as slurries, intense gas discharge
and crystal luminescence are observed, similar to resorcinol.
The sonication of a slurry of resorcinol in hexadecane sparged with He not only
shows intense He discharge lines and crystal luminescence, but also several new
atomic and molecular emission series are observed (Fig. 8.9). The new emissions
include Heþ (~77 eV above the 1 s state of He), multiple transitions from the Swan
band of C2 (d3Πg a3Πu), the CH (A2Δ X2Π) transition, and the Hα line.
When the same slurry is sparged with O2, the same crystal luminescence, C2 and
CH discharges are still present, and in addition, new emissions from O, CO (b3Σ a
3
Π), (C1Σ A1Π) and (B1Σ A1Π), COþ (B2Π X2Σ), and CO2þ (A2Π X2Π) are
observed (Fig. 8.10) [4]. The emissions of Heþ, C2, CH, CO, COþ, and CO2þ have
not been reported before from any ML event. It should be noted that the Swan band
of C2 as well as the CH transition at 431 nm are commonly observed in simple
sonoluminescence (SL) of hydrocarbons [81]. However, the high temperature and
high pressure conditions within the bubble when the emission of C2 and CH are
observed from SL result in the lines being substantially broader and very much
weaker than the ones observed from ML. When hexadecane sparged with O2 is
sonicated without resorcinol, with all other conditions held constant, weak
sonoluminescence is observed. The sonoluminescence consists of broad CO2 emis-
sion and is subtracted from the raw emission spectrum to reveal only the ML.
250 N.C. Eddingsaas
Fig. 8.9 (a) ML from a

Sonication of resorcinol in
hexadecane sparged with
ML Intensity
He. The spectrum consists
of very strong emission
from He excited states
(marked with red square).
(b) A tenfold intensity
enhancement of intensity
which show extensive C2
(d3Πg – a3Πu), CH (A2Δ
X2Π), and Hα line emission.
Figure adapted from ref.
[4] 300 400 500 600 700
b
b
CH
ML Intensity
Crystal C2 H
luminescence
N2
300 400 500 600

Wavelength (nm)
The multitude of new features arising in the ML spectra could have arisen from
either the liquid or the solid. To determine the origin of these emissions, the spectra
of slurries of resorcinol in dodecane and perdeutero-dodecane were compared. In
perdeutero-dodecane, if the emission lines originate from the crystal, then the CH
(A2Δ X2Π) progression will be observed with the band head at 431.4 nm, while if
the emission lines arise from the liquid then the CD (A2Δ X2Π) progression will
be observed with the band head at 430.8 nm. The sonication of the dodecane slurry
only showed CH emission, whereas only CD emission is observed from the
perdeutero-dodecane slurry [4]. This indicates that the carbon-containing small
molecules in the ML spectra are from the liquid and not the crystal.
The low vapor pressures of dodecane and hexadecane, however, present a
paradox because it seems unlikely that the new emitting species can arise directly
from the vapor of the liquid itself. It is known, however, that the high temperature
and pressure inside the collapsing bubbles cause the decomposition of any vapor
that accumulates inside, resulting in the production of many radicals that will
recombine to form new molecules [82]. For instance, the sonication of water
produces hydroxyl radicals as well as hydrogen and oxygen atoms. These species
react to reform water, hydrogen gas, and hydrogen peroxide. Previous work in the
Suslick group has shown that the sonication of long chain alkanes produces a
substantial quantity of small hydrocarbons such as CH4 and C2H2 along with H2
b c
Intensity (a. u.)
‡
e
a *
200 400 600
200 300 400 500 600 700 800

Wavelength (nm)
Fig. 8.10 Spectrum of the ML from the sonication of a slurry of resorcinol in hexadecane sparged
with O2, showing both O atom emission (marked with red circle) and extensive molecular
emission. The inset shows a blow up of the region from 200 to 650 nm show the molecular
features, including (a): COþ (B2Π X2Σ), (b): CO (b3Σ a3Π) (280–350 nm), (c): CO (C1Σ A1
Π) (368–466 nm) and (d): CO (B1Σ A1Π) (450–610 nm), (e): CO2þ (A2Π X2Π), asterisk: C2 (A
3
Πg X3Πu), double dagger: CH (A2Δ X2Π). Figure adapted from ref. [4]
Table 8.1 Sonolysis Product Rate, nmol/min Product Rate, nmol/min

products from the sonication
H2 1610 1-C5H10 135
of dodecanea
CH4 350 C5H12 21
C2H2 260 1-C6H12 68
C2H4 345 C6H14 28
C2H6 46 1-C7H14 44
C3H4 22 C7H16 16
C3H6 297 1-C8H16 42
C3H8 32 C8H18 9
1-C4H8 188 1-C9H18 21
C4H10 47 C9H20 3
a
Data from ref. [82]
via radical reactions, as seen in Table 8.1 [82]; moreover, when a reactive gas is
present (such as N2 or O2), other sonolysis products are formed, such as CN and
CO2 [81–83].
To show that the new emissions from fracto-ML from acoustic cavitation arise
from sonolysis products, a slurry of resorcinol and dodecane was placed in a steel
cell whose gas inlet and outlet could be sealed off, the sample was saturated with He
252 N.C. Eddingsaas
gas, and then the cell was sealed. The slurry was sonicated while the cell remained
sealed, with no further gas allowed to enter or leave the cell. Upon subsequent
sonication of the slurry in the sealed cell, the intensity of the CH and H emission
increased in comparison to the He lines. On the other hand, upon multiple sonica-
tions of the same slurry that was continuously sparged with He, the ratio of intensity
of the CH (A X) and Hα emissions compared to the He emission remained
constant. This result is consistent with the hypothesis that the volatile sonolysis
products of the liquid play a major role in the discharge emission. In addition, the
ratio of CH and H to He was observed to increase as the vapor pressure of the alkane
decreased as a result of lower temperature or longer chain length; two conditions
known to increase the amount of sonolysis products [82].
The origin of the CO, COþ, and CO2þ emissions observed during sonication in
the presence of oxygen must arise from the chemical reaction induced by the
dielectric breakdown either of CO2 produced during sonolysis or of O2 with small
hydrocarbon sonolysis products (e.g., CH4 or C2H2). When resorcinol in
hexadecane saturated with CO2 was sonicated, only CO2þ (A2Π X2Π) and (A2Σ
þ
X2Π) emission bands were observed. There was no visible emission from CO,
COþ, or O, which are the main components of the ML from resorcinol slurries
sparged with O2 (Fig. 8.10). The production of CO and COþ is therefore due to
discharge reactions of small hydrocarbons with O2 and not from CO2. Consistent
with these results, electric discharge through CH4 or C2H2 and O2 (used for the
production of diamond films) gives emission from CH, C2, H, CO, COþ, and O
[84, 85].
The reactive nature of the ML discharge has also been observed by Sharipov and
coworkers [86]. They observed the emission of OH and oxygen atom during the
fracture of hydrated cerium/terbium sulfates. The emission of O was determined to
be a result of the decomposition of O2 during the ML discharge and OH from the
decomposition of water within the crystal caused by the ML discharge.
8.3.3 Mechanoluminescence as a Probe of Intensity

of the Shockwave Launched During Acoustic
Cavitation
As shown in Sect. 8.3.2, the fracto-ML is sensitive to the intensity of the fracture
event. In particular, the ratio of gas discharge to crystal luminescence increases the
more intense the fracture event is. This has provided a way to verify that the
shockwave launched during acoustic cavitation is sensitive to liquid properties,
namely vapor pressure. The lower the vapor pressure of the liquid, the harder the
collapse (due to fewer polyatomics in the bubble effectively increasing the
polytropic ratio of heat capacities) resulting in a more intense shock wave.
Previous work by Doktycz demonstrated the effect of vapor pressure on the
shockwave by monitoring heterogeneous reactions from acoustic cavitation at
variable vapor pressure [87]. Doktycz showed that both the Ullman coupling
reaction and the Reformatsky reaction were sped up and produced higher yields
as the vapor pressure of the liquid was reduced. The enhancement of these hetero-
geneous reactions was shown to be caused by an increase in the shockwave
launched into the liquid. Few other studies have looked at the effect of slurry
composition on heterogeneous interactions during acoustic cavitation. The domi-
nant pathway to fragmentation of organic crystals is direct interaction with the
shockwave; therefore, ML of resorcinol is an excellent candidate to show the effect
of vapor pressure on the intensity of the shockwave launched during acoustic
cavitation.
Slurries of 150–250 μm resorcinol crystals in alkanes from heptane to
hexadecane were sparged with He and irradiated with ultrasound at 6 W/cm2. As
seen in Fig. 8.11, the total ML intensity increases as the alkane chain length
increases. It can be seen that there is no He gas discharge line visible from
sonication of a slurry composed of an alkane with a smaller chain length than
decane (vapor pressure of 2.7 Torr). As the alkane chain length is increased to
dodecane (vapor pressure 0.40 Torr) and finally to hexadecane (0.0070 Torr) the He
discharge increases in intensity by a factor of ~10, while the intensity of the crystal
luminescence only increases by a factor of ~3. This result is indicative of a more
intense fracture event. On the other hand, when resorcinol was crushed in each of
the liquids saturated with He, no enhancement of the gas discharge is observe in any
of the solutions.
a b
He
CH
ML Intensity
280 320 360 400 440 280 320 360 400 440
Hexadecane (vp: 0.007 torr) Nonane (vp: 7.28 torr)
Dodecane (vp: 0.404 torr) Octane (vp: 21.8 torr)
Decane (vp: 2.72 torr) Heptane (vp: 61.6 torr)
Fig. 8.11 ML from the sonication of resorcinol in alkanes of varying chain length from 7 to
16 carbons saturated with He. The vapor pressure decreases from 61.6 Torr for heptane to
0.007 Torr for hexadecane while the viscosity increases from 0.387 to 3.454 mPa s. Notice that
gas discharge is not observed until sonication of the slurry in decane or longer chain length
254 N.C. Eddingsaas
While this is only a qualitative analysis of the effect of increase in impact

velocity (and therefore intensity of the fracture event), it highlights the ability of
ML to not only show where and when stress and fracture occur, but also how
forceful the event was. Mechanoluminescent material has been widely researched
as damage sensors, but in each case they have only been used for visual indication
of damage or fracture [53, 88–90]. Using mechanoluminescent crystals that exhibit
both gas discharge and crystal luminescence (such as resorcinol or coumarin) into
damage sensors would reveal much more attainable information.
8.4 Characteristics of Plasma Generated During

Mechanoluminescence
The application of mechanical energy to a system causing a chemical reaction to

occur due to the rupture of a covalent chemical bond is known as mechanochem-
istry or tribochemistry [91–95]. Mechanochemistry is important in many chemical
and industrial processes and has been summarized by Heiniche [93]. Along with the
chemical reactions, many other processes occur at the point of mechanical stress.
These other processes are known as mechanoemission, fracto-emission, or
triboemission and they play a vital role in the mechanochemical reactions. These
mechanoemissions include electrons, ions, photons, UV radiation, IR radiation,
acoustic emission, and even X-rays [92, 96–99]. The different emissions have been
studied to better understand the chemical processes that occur upon addition of
mechanical energy into a system. While the different mechanoemissions have
provided a great deal of insight into mechanochemistry, one type of emission not
yet fully explored is the emission of photons—mechanoluminescence. The ML
discharge could provide insight into the conditions generated upon wear or fracture
of materials, but remains largely unexplored because the intensity of ML emission
is inherently weak compared to traditional means of producing ML, such as
grinding and cleaving.
While the phenomenon of ML has been known for hundreds of years [1] and the
cause of the discharge is well formulated [53, 59, 60, 65, 100], the characteristics of
the resultant discharge plasma have never been fully studied. The fact that produc-
ing ML by acoustic cavitation results in an enhancement of the ML emission
intensity and the observation of new emission products including Heþ, Hα, C2,
CH, CO, and COþ will allow for the better characterization of the discharge plasma
[3, 4].
Plasmas are an electrically neutral ionized gas, the fourth state of matter.
Plasmas are the most common phase of matter in the universe in terms of both
mass and volume, and their physical properties are very diverse in terms of
temperature, density, electron density, and electron energy. For example,
Fig. 8.12 compares several plasmas in terms of electron density and degree of
ionization and describes the different plasmas as being reactive, radiative, or
Fig. 8.12 Description of a number of plasmas based on the electron density (ne) and degree of
ionization (ne/ng). The plasmas are also grouped as “reactive,” “radiative,” and “thermonuclear”
based on these parameters. The types of plasmas listed here are; (gd) glow discharge, (μd)
microwave discharge, (dbd) dielectric barrier discharge, (ag) anomalous glow, (ECR) electron
cyclotron resonance discharge, (ICP) inductively coupled plasma, (va) vacuum arc, (t) thermal
plasma, (ma) magnetized arc. Figure adapted from ref. [101]
thermonuclear based on these two parameters [101]. The diversity of plasma is

immense; in fact, the electron density of plasmas spans over 30 orders of magnitude
and electron energy varies over seven orders of magnitude [102]. In order to
256 N.C. Eddingsaas
understand and characterize a plasma, the temperature, pressure, and electron

density must be fully measured.
In the rest of this section, plasma diagnostics will be applied to an ML discharge
produced from the sonication of a slurry of resorcinol in dodecane sparged with
He. Spectroscopic plasma diagnostics are used to characterize a plasma where
direct probes cannot be used, or under conditions where probes will be destroyed.
Plasma spectroscopy is a field that is well developed and is the subject of numerous
books [103, 104]. The physics and mathematics used to determine pressure, tem-
perature, electron density, and electron energy from the ML discharge will be
presented in each sub-section.
8.4.1 Pressure
To apply plasma diagnostics to the ML discharge the pressure first needs to be

known or estimated. In the experiments of ML from acoustic cavitation, the
pressure was not explicitly measured but a good approximation was obtained. To
approximate the pressure of the gas within the fracture site, first the rate of fracture
and duration of ML event was calculated based on impact velocity and rate of crack
propagation [63, 64], and then gas diffusion into the resulting void was determined
using Fick’s law of mass diffusion. Fick’s Law of mass diffusion describes the
transport of dissolved gas within a liquid and is represented by:
AD
V gas ¼ ðC1 C2 Þ ð8:1Þ
T
where A is the area of the interface, D is the diffusivity, T is the thickness of the
interface and C1 and C2 are the gas concentrations on either side of the interface.
Using Eq. 8.1, a reasonable estimate as to the pressure of gas within the fracture site
during the discharge event is ~0.1 bar.
To apply plasma diagnostics it is also important to know the composition of the
gas present when the discharge occurs. The solution used to determine the plasma
conditions during ML discharge by acoustic cavitation was dodecane (vapor
pressure ¼ 0.13 Torr) saturated with He. In this solution the vast majority of gas
available to enter the void of the fracture is He with minor small hydrocarbons such
as methane and ethane formed from sonolysis of dodecane. Due to this fact it is
assumed that collisions of gas molecules within the fracture will be with He in order
to simplify the calculations and discussion.
8.4.2 Temperature
When a system is at thermodynamic equilibrium, a single temperature characterizes

all internal energy modes (vibrational, rotational, and electronic). Under these
circumstances the temperature of the system can be determined from the absolute
intensity of any atomic or molecular feature or from Boltzmann plots of relative
population distributions of rotational, vibrational, or atomic lines. However, in a
nonthermal plasma the system is not represented by a single temperature due to
departure from equilibrium in the distribution of internal energy state population. In
nonthermal plasmas the gas temperature is usually best inferred from the intensity
distribution of rotational lines.
The emission intensity of each rovibrational line is proportional to the steady
state population (Nν0 J0 ) by:
0 0 0 0
I νν00JJ00 ¼ KAνν00JJ00 N ν0 J0 ð8:2Þ
0 0
where Aνν00JJ00 is the Einstein coefficient for the rovibrational transition, and K is an
experimentally determined constant dependent on optical configuration and detec-
tion electronics. The steady state population in the excited state is given by the
equilibrium between the excited state production and destruction rates, where
electronic quenching dominates at near atmospheric pressure. Prior to the removal
of electrons from the excited state by radiative emission or electronic quenching,
collisions of the molecules or atoms with other atoms or molecules can move the
electron population among the molecules rotational and vibrational levels; the
higher the pressure the more collisions will occur pushing the population closer
to equilibrium. Since the difference in energy between rotational levels is much
smaller than the other internal energy modes, they are moved toward equilibrium at
the fastest rate. This relaxation of rotational distribution toward equilibrium is why
the heavy atom temperature is best represented by the rotational distribution.
A thermalized population is defined as a distribution following Maxwell–
Boltzmann statistics, which is a function of the level degeneracy (gJ0 ), its energy
(Ey0 J0 ), and the temperature (T ):
N ν0 J0 ¼ gJ0 eðEν0 J0 =kT Þ ð8:3Þ
where k is the Boltzmann constant, and the rotational degeneracy is gJ0 ¼ 2 J0 þ 1.

When looking at only a single vibrational level, the temperature is obtained from the
slope of the line of the plot of the rotational population versus the rotational energies:
NJ0 E0
ln 0 ¼ J þC ð8:4Þ
2J þ 1 kT
258 N.C. Eddingsaas
Fig. 8.13 (a) CH (A2Δ X2Π) emission from ML of resorcinol used to determine heavy atom
temperature during ML discharge (solid black line). In addition, the simulated spectrum obtained
using LIFEBASE is also included (red dashed line). (b) Boltzmann plot of rotational population
distribution of CH (A2Δ X2Π) determined from simulated spectrum. The CH ML emission is
from the sonication of resorcinol in dodecane sparged with He. Figure adapted from ref. [5]
In a nonthermal plasma the rotational, vibrational, and electronic distribution can

all be represented by different temperatures. In addition, it is not uncommon for
there to be complex rotational distributions where the population is not represented
by a single temperature. Rather, some of the lower energy rotations will have
reached equilibrium, while the population within the higher rotational energy levels
will have a distribution of energies representing a much hotter temperature more
representative of the excitation energy [105–108]. In this case the heavy atom
temperature can still be inferred from the plot of the rotational population versus
the rotational energy. The population will be fit to a bimodal distribution with the
distribution from the lower energy rotational states representing the heavy atom
temperature.
Using these principles the heavy atom temperature during an ML discharge can
be determined. The CH (A2Δ X2Π) emission from ML of resorcinol in dodecane
sparged with He is shown in Fig. 8.13a [5]. In addition, Fig. 8.13a also shows a
simulated CH spectrum obtained using LIFEBASE software, a program used to
simulate spectra of diatomic molecules. Very good agreement between experimen-
tal emission spectrum and simulated spectrum is observed.
To obtain temperatures from the simulated spectrum, the rotational population at
each rotational energy level was extracted from LIFEBASE and plotted in a
semilogarithmic plot also shown in Fig. 8.13b. As can be seen, the data shows a
bimodal distribution that is fit with two separate linear regressions indicating two
temperatures, Tr1 ¼ 405 22 K for the low energy rotational transitions
(representing 80 % of emitting CH) and Tr2 ¼ 4015 730 K for the higher energy
rotational transitions. A bimodal distribution of rotational population could arise
from a couple of mechanisms. First, there could be two excitation mechanisms
resulting in emission at two different temperatures. Second, there could be one
excitation mechanism with partial rotational relaxation as a result of collisional
energy transfer. The former is highly unlikely while the latter is reasonable due to
the pressure in the system being near atmospheric pressure, allowing for a number
of collisions prior to radiative emission. This bimodal distribution arises from the
fact that high rotational quantum numbers have spacings much larger than the
thermal collision energy, so many more collisions are necessary for full thermal
relaxation, resulting in emission from CH which is still hot after formation by
electron impact on hydrocarbon vapor. The larger error in the high temperature Tr2
arises in part from the fact that >80 % of the rotational population is in the
thermalized state and in part from the wide distribution of hot CH that Tr2 actually
represents. Similar bimodal populations of rotational levels of CH have been
observed in electric discharges of alkanes at pressures from a few mbar up to an
atmosphere [105, 106, 109].
8.4.3 Electron Density
When a Langmuir probe cannot be used to measure the electron density of a plasma,
passive spectroscopy is used. The well-known Thomson scattering can be used to
measure electron density [110, 111]; however, this technique requires a powerful
laser system, is difficult to implement, and does not work on small or confined
plasmas. A simpler and well developed technique is to use the Stark broadening of
atomic emission lines. Another indicator of electron density is the presence of the 41
F–21P forbidden transition near the 41D–21P transition of He at 492.2 nm; emission
from this forbidden transition is induced by the electric microfield in the plasma
which causes a breakdown of the parity selection rule [112–114]. Both methods
have been used to determine the electron density in a ML discharge that result from
the sonication of resorcinol in dodecane [5].
8.4.3.1 Electron Density from Stark Broadening
Stark broadening is due to the collisions of the emitter with charged particles in
close proximity to the emitter according to the Stark effect. Stark broadening is a
result of Coulombic interactions between the radiating species and the charged
particles present in the plasma. Both electrons and ions induce Stark broadening,
but the electron contribution dominates due to their higher relative velocity. A
number of atomic emission lines such as Ar [115], He [116], and most often the
Balmer series of hydrogen [117–120] have been used to determine electron density.
The most often used emission line from the Balmer series of hydrogen is the Hβ
line. This is due to the fact that the Hβ line undergoes negligible self-absorption,
exhibits strong Stark broadening, and is not appreciably broadened by ion dynam-
ics. In addition, Stark broadening is highly insensitive to temperature, and so it may
still be applied to nonthermal plasmas. The extent of Stark broadening increases as
the energy increases within the series, but in many cases only the Hα and Hβ emit
260 N.C. Eddingsaas
with enough intensity to be observed. The Stark broadening of the Hβ has been well
studied and the parameters and constants needed to determine the contribution of
the Stark effect to the overall atomic line broadening have been well developed.
These values have errors associated with them that are less then 10 % [117–120].
In order to determine the electron density from the Stark broadening of the Hβ
line, first the contribution to the broadening of the line by the Stark effect must be
extracted from the overall line width. To do this, the contribution of all other
broadening mechanisms must be determined and proper deconvolution of these
other broadening mechanisms must be performed so that only the Stark broadening
is left.
The lineshape of the Hβ line is a convolution of the Gaussian (instrument,
Doppler) and Lorentzian (natural, resonance, van der Waals, Stark) broadening
mechanisms resulting in a Voigt profile [117, 119, 121, 122]. The full width at half
maximum (FWHM) of the Gaussian component (ΔλG) is given by:
1=2
ΔλG ¼ Δλ2inst þ Δλ2Dopp ð8:5Þ
while the FWHM of the Lorentzian component (ΔλL) is a linear combination of its
parts:
ΔλL ¼ Δλnat þ Δλres þ ΔλVdW þ ΔλStark ð8:6Þ
and the two can be combined together to obtain the Voigt FWHM (ΔλV) using the
method derived by Whiting [123]:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 ffi
ΔλL ΔλL
ΔλV ¼ þ þ Δλ2G ð8:7Þ
2 4
Once the linewidth of the non-Stark broadening mechanisms is determined math-

ematically from well-established plasma diagnostics [104, 117–122], these equa-
tions can be used to extract the contribution to the line broadening from the Stark
effect. A full description of contributions of the other effects to line broadening is
beyond the scope of this chapter and can be found in the above references.
The line shape from Stark broadening can be approximated by a Lorentzian
profile, except at the line center where electrostatic interactions with ions can cause
a dip. For the Hβ line, it has been found that the dip is very small and the depth is a
function of the ion dynamics [124, 125]. Ion dynamic calculations have also
confirmed that the Stark width of the Hβ is mostly a function of the free electron
concentration and a weak function of the temperature [126]. The Stark effect on the
Hβ line has been extensively studied and the broadening parameters are well
developed. The FWHM (in Å) for Stark broadening can be represented [122] by:
Fig. 8.14 High resolution ML spectrum used to determine electron density. (a) ML spectrum
from the sonication of a slurry of resorcinol in dodecane sparged with He. The spectrum consists of
He I (41D–21P) transition, the forbidden He I (41F–21P) transition, and the Hβ line at 486.1 nm.
(b) Zoom in on Hβ emission line along with Voigt fit (red long dash) with a full width half
maximum of 0.718 Å and simulation of the instrumental broadening (orange short dash) with full
width half maximum of 0.342 Å. The x-axis is wavelength from line center in angstroms. (c) ML
spectrum of 41F–21P forbidden and 41D–21P allowed transition of the He I 492.1 nm line. The
separation between the allowed and forbidden transition of 1.8 Å correlates to an electron density
between 1014 and 1015 cm3. Figure adapted from ref. [5]
ΔλStark ¼ 2:0 1010 n2=3

e ð8:8Þ
Where ne is the electron density in cm3.

The high resolution spectrum of resorcinol ML in dodecane sparged with He is
shown in Fig. 8.14a, which contains the Hβ line as well as the He I 41D–21P and 41
F–21P transitions. The He lines can also be used to determine electron density and
will be discussed below. As long as the electron density is greater than ~5 1013
cm3, the electron number density can be determined from the broadening of the Hβ
line at 486.1 nm. The excited state H atoms in this system result from the discharge
262 N.C. Eddingsaas
through H2, CH4, and C2H2 (sonolysis products of dodecane) [4, 82], so no external
H2 source needed to be added.
Figure 8.14b shows a high resolution spectrum of the Hβ line observed during
ML along with the pseudo-Voigt fit (FWHM 0.72 Å), and the Gaussian profile fit to
the instrument broadening, which has a linewidth of 0.34 Å. Other atomic emission
lines did not show broadening beyond instrument broadening. The contributions to
the line broadening other than the Stark effect are as follows. Natural broadening is
very small and represents only 6.2 105 Å of the Lorentzian component of the
line width. Resonance broadening is even less than natural broadening as the main
collisional partner is He and therefore is ignored. The contributions to the line
broadening by van der Waal and Doppler interactions were calculated to be 0.047
and 0.070 Å respectively. Using the known values for the other mechanisms of line
broadening, the broadening by Stark effects was calculated to be 0.55 Å resulting in
calculated electron density of 1.3 0.13 1014 cm3. The Hβ linewidth is highly
sensitive to the electron density. For comparison, the calculated Stark linewidth
would be 0.41 Å at an electron density of 1014 cm3 and 1.97 Å at 1015 cm3
[117, 118].
8.4.3.2 Electron Density from Forbidden He I Emission
Electron density can also be derived from the line shape of the 4D–2P He I lines at
447.1 and 492.1 nm [112–114, 127]. Both of these He lines show strong 4F–2P
forbidden components about 1.5 Å to the blue of the allowed transition line,
induced by the electric microfield in a plasma which results in a breakdown in
the parity selection rule. The intensity of the forbidden component and the separa-
tion from the allowed component are highly dependent on electron density, there-
fore, the measurement of these parameters is another method for determining
electron density. Since He has only two electrons, theory does an excellent job of
predicting the effect of all of the broadening parameters and equations have been
formulated that take in account the influence of electrons and ions on the line
broadening process [128]. The position and intensity of the 4F–2P forbidden
transition in relation to the 4D–2P transition at 492.1 nm has been studied by a
number of groups, over a range of electron densities of 1014–1016 cm3 [112, 129,
130]. The literature results can be used to qualitatively determine electron density
within the ML event.
Figure 8.14c shows the emission spectrum of He I 41D–21P and 41F–21P from
the sonication of resorcinol in dodecane sparged with He. By comparison to
literature results, the electron density during the ML event was estimated to be
between 1014 and 1015 cm3, which is fully consistent with the value determined by
the Hβ line.
8.4.4 Electron Energy
At high electron densities (ne > 1018 cm3), line ratios will adhere to the Saha–
Boltzmann formula, and the electrons will be in thermodynamic equilibrium. When
a system is at thermodynamic equilibrium, the electron temperature will be the
same as the heavy atom temperature and can be obtained from simple Boltzmann
statistics. At lower electron densities, thermodynamic equilibrium is not reached,
but the electron energy can still be determined through spectrographic methods. A
few of the spectrographic methods that can be employed are—the ratio of two lines’
intensity, the ratio of a line to continuum intensity, and the ratio of two parts of
continuum intensity. Many of these techniques are dependent on the mechanism of
electron interaction, and four models have been formulated including: local-
thermal-equilibrium model, steady-state corona model, time dependent corona
model, and collisional-radiative model. For a plasma with electron densities
found during a ML event, the collisional-radiative model must be used. This
model is mathematically more intensive as stepwise excitation and redistributive
processes are included.
One method for determining electron energy that is not model dependent is
comparing intensities of emission lines from the same element which have excited
states of suitably large energy difference, such as the He I line at 587.6 nm (excited
state energy is ~23 eV above the ground state) and the He II line at 468.6 nm
(excited state energy is ~77 eV above the ground state of He I) [104, 131]. Mewe
expanded on the work of Griem to determine the electron temperature based on the
ratio of these two lines using the equation:

Ia gf =λ3 a bp ð1Þ 1 54:4 1 1

¼ 6 10 21
3=2
ðkT e Þ exp 1 ð8:9Þ
Ib gf =λ3 b bq ð0Þ b1 ð1Þ ne kT e p2 4q2
where g is the statistical weight, f is the absorption oscillator strength, λ is the

wavelength of the line in Å, bp(z) is the ratio between rates of collisional ionization
and recombination to level p (constants found in ref. [50]), ne is the electron density
in cm3, and kTe is the electron energy in eV. This equation is applicable for
electron densities between 1010 and 1019 cm3 and electron energies between 2 and
11 eV.
Figure 8.15 shows the portion of the ML emission spectrum from the sonication
of resorcinol in dodecane sparged with He that contains the He I and He II lines
along with Gaussian fits to the two lines. From these fits it was found that the
HeII468.6/HeI587.6 ratio is 3.5 102. Using Eq. 8.9 the electron energy is calculated
to be 3.5 eV (i.e., an effective Te ~41,000 K). Because our gas pressure is near the
lower limit for the applicability of Eq. 4.12, the electron energy estimate is a lower
limit [131].
264 N.C. Eddingsaas
Fig. 8.15 Comparison of

the He I (587.6 nm) and He
II (468.6 nm) ML discharge
emission lines produced
from sonication of
resorcinol in dodecane
sparged with He. ML (black
line), Gaussian fit (red
dashed line). The ratio of
area of the He II/He I line is
0.035. Figure adapted from
ref. [5]
Table 8.2 Characteristics of different plasmas

Heavy atom Electron density Electron energy
Plasma temperature (K) (cm3) (eV)
Corona discharge 300–500 ~1010 1–2
Mechanoluminescent 405 22 1.3 0.13 1014 3.5
discharge
Dielectric barrier 300 1014–1015 1–10
discharge
Inductively coupled 5000–15,000 1013–1014 6–100
plasma
Laser plasma 13,000 ~1020 102–103
Center of the sun 1.5 107 ~1025 104–105
8.4.5 Comparison to Other Discharge Plasmas
The characteristics of the plasma generated during the ML discharge are summa-
rized in Table 8.2 along with the characteristics of a number of other plasmas. The
ML discharge resembles most closely atmospheric microdischarges such as direct
current microplasmas [122, 132], microplasma jets [121], and dielectric barrier
discharges [105, 133, 134]. Typical microdischarge plasmas are nonthermal with
heavy atom temperatures near ambient, electron density of 1014–1015 cm3, and
electron energy of 1–10 eV [133]. These types of discharges are highly reactive
plasmas and have been used for many applications, such as ozone production,
remediation of chemical toxins, sterilization, surface treatment, and lighting.
The plasma was characterized under one condition—fracture of resorcinol in the
presence of He. Subsequent studies have compared spectral features from sonica-
tion of resorcinol in dodecane sparged with He with that of resorcinol sparged with
Ar as well as other mechanoluminescent crystals sparged with He and Ar. In all
cases the spectral features that could be compared (CH, C2, and H emission) were
identical, indicating that the same plasma conditions are generated during ML. The
comparison of the ML discharge to atmospheric pressure microdischarges again

confines the reactive nature of the ML discharge noted earlier when emission lines
were observed from the decomposition of gases by the ML discharge.
8.5 Summary
Mechanoluminescence is an intriguing phenomenon that has garnered scientific

attention for over 400 years and is now being utilized in materials for damage and
strain sensors. It has been known for some time that the light emission was due to an
electrical discharge that neutralizes charges on opposing faces of a propagating
crack, and that this electrical discharge is able to excite both gas molecules as well
as the crystal itself. In addition, many other emissions occur during the fracture of a
mechanoluminescent crystal including electrons, ions, and acoustic emission. The
conditions within the discharge itself remained largely unanalyzed for many years.
By increasing the frequency of mechanoluminescent events and, more importantly,
increasing the force of fracture by initiating fracto-ML with acoustic cavitation the
conditions within the mechanoluminescent discharge have been more thoroughly
characterized. The large enhancement of the gas discharge ML from sonication of a
slurry of piezoelectric crystals such as sucrose, resorcinol, and coumarin in long
chain alkanes has revealed a number of new emission lines stemming from atomic,
molecular, and ionic origins. The new emission lines have shown that the
mechanoluminescent discharge is a highly reactive discharge very similar to a
dielectric barrier discharge that is used for chemical reactions and lighting. This
indicates that the mechanoluminescent discharge probably plays a role in mecha-
nochemistry, and knowing the conditions can help in understanding the outcomes
of the reactions. Finally, the use of acoustic cavitation to induce ML has highlighted
the effect of the intensity of the fracture has on a number of crystals that exhibit
both gas discharge and crystal luminescence. The fact that the intensity of the gas
discharge ML relative to crystal luminescence ML increases as the fracture event
becomes more forceful can be utilized in damage sensors to not only show where
damage is occurring but the extent of the force as well.
References
1. Bacon, F. (1915). Of the advancement of learning. London: J. M. Dent and Sons.

2. Jha, P., & Chandra, B. P. (2014). Survey of the literature on mechanoluminescence from 1605
to 2013. Luminescence, 29(8), 977–993. doi:10.1002/bio.2647.
3. Eddingsaas, N. C., & Suslick, K. S. (2006). Light from sonication of crystal slurries. Nature,
444(7116), 163–163. doi:10.1038/444163a.
129(21), 6718–6719.
266 N.C. Eddingsaas
5. Eddingsaas, N. C., & Suslick, K. S. (2007). Plasma characteristics of the discharge produced
during mechanoluminescence. Physical Review Letters, 99(23), 234301. doi:10.1103/
PhysRevLett.99.234301.
6. Suslick, K. S. (Ed.). (1988). Ultrasound: Its chemical, physical, and biological effects.
New York, NY: VCH.
7. Suslick, K. S. (1990). Sonochemistry. Science, 247, 1439–1445.
8. Suslick, K. S., Didenko, Y., Fang, M. M., Hyeon, T., Kolbeck, K. J., McNamara, W. B., III,
et al. (1999). Acoustic cavitation and its chemical consequences. Philosophical Transactions
of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences,
357(1751), 335–353.
9. Suslick, K. S. (1989). The chemical effects of ultrasound. Scientific American, 260(2), 80–86.
10. Mason, T. J., & Lorimer, J. P. (Eds.). (1988). Sonochemistry: Theory, applications and uses
of ultrasound in chemistry. Chichester: Ellis Horwood.
11. Leighton, T. G. (1994). The acoustic bubble. San Diego, CA: Academic.
12. Barber, B. P., Wu, C. C., Lofstedt, R., Roberts, P. H., & Putterman, S. J. (1994). Sensitivity of
sonoluminescence to experimental parameters. Physical Review Letters, 72(9), 1380–1383.
13. Barber, B. P., Hiller, R. A., Loefstedt, R., Putterman, S. J., & Weninger, K. R. (1997).
Defining the unknowns of sonoluminescence. Physics Reports, 281(2), 65–143.
14. Young, F. R. (2005). Sonoluminescence. Boca Raton: CRC Press.
15. Gompf, B., Gunther, R., Nick, G., Pecha, R., & Eisenmenger, W. (1997). Resolving
sonoluminescence pulse width with time-correlated single photon counting. Physical Review
Letters, 79(7), 1405–1408.
16. McNamara, W. B., III, Didenko, Y. T., & Suslick, K. S. (1999). Sonoluminescence temper-
atures during multi-bubble cavitation. Nature, 401(6755), 772–775.
17. McNamara, W. B., III, Didenko, Y. T., & Suslick, K. S. (2003). Pressure during sonolumi-
nescence. Journal of Physical Chemistry B, 107(30), 7303–7306.
18. Flannigan, D. J., & Suslick, K. S. (2005). Plasma formation and temperature measurement
during single-bubble cavitation. Nature, 434(7029), 52–55.
19. Flannigan, D. J., Hopkins, S. D., Camara, C. G., Putterman, S. J., & Suslick, K. S. (2006).
Measurement of pressure and density inside a single sonoluminescing bubble. Physical
Review Letters, 96(20), 204301/204301–204301/204304.
20. Suslick, K. S., & Flannigan, D. J. (2008). Inside a collapsing bubble: sonoluminescence and
the conditions during cavitation. Annual Review of Physical Chemistry, 59, 659–683. doi:10.
1146/annurev.physchem.59.032607.093739.
21. Blake, J. R., & Gibson, D. C. (1987). Cavitation bubble near boundaries. Annual Review of
Fluid Mechanics, 19, 99–123.
22. Pecha, R., & Gompf, B. (2000). Microimplosions: Cavitation collapse and shock wave
emission on a nanosecond time scale. Physical Review Letters, 84(6), 1328–1330.
23. Ohl, C.-D., Kurz, T., Geisler, R., Lindau, O., & Lauterborn, W. (1999). Bubble dynamics,
shock waves and sonoluminescence. Philosophical Transactions of the Royal Society of
London, Series A: Mathematical, Physical and Engineering Sciences, 357(1751), 269–294.
24. Holzfuss, J., Ruggeberg, M., & Billo, A. (1998). Shock wave emissions of a sonoluminescing
bubble. Physical Review Letters, 81(24), 5434–5437.
25. Reisman, G. E., Wang, Y. C., & Brennen, C. E. (1998). Observation of shock waves in cloud
cavitation. Journal of Fluid Mechanics, 355, 255–283.
26. Doktycz, S. J., & Suslick, K. S. (1990). Interparticle collisions driven by ultrasound. Science,
247(4946), 1067–1069.
27. Prozorov, T., Prozorov, R., & Suslick, K. S. (2004). High-velocity interparticle collisions
driven by ultrasound. Journal of the American Chemical Society, 126(43), 13890–13891.
28. Walton, A. J., & Reynolds, G. T. (1984). Sonoluminescence. Advances in Physics, 33(6),
595–660.
29. Gaitan, D. F., Crum Lawrence, A., Church, C. C., & Roy Ronald, A. (1992). Sonolumi-
nescence and bubble dynamics for a single, stable, cavitation bubble. Journal of the Acous-
tical Society of America, 91, 3166–3183.
30. Brenner, M. P., Hilgenfeldt, S., & Lohse, D. (2002). Single-bubble sonoluminescence.
Reviews of Modern Physics, 74(2), 425–484.
31. Putterman Seth, J., & Weninger, K. R. (2000). Sonoluminescence: How bubbles turn sound
into light. Annual Review of Fluid Mechanics, 32, 445–476.
32. Lauterborn, W., & Ohl, C. D. (1997). Cavitation bubble dynamics. Ultrasonics
Sonochemistry, 4(2), 65–75.
33. Burrill, L. C. (1951). Sir Charles Parsons and cavitation. Transactions of the Institution of
Marine Engineering, 63, 149–167.
34. Krefting, D., Mettin, R., & Lauterborn, W. (2004). High-speed observation of acoustic
cavitation erosion in multibubble systems. Ultrasonics Sonochemistry, 11(3-4), 119–123.
35. Feller, H. G., & Kharrazi, Y. (1984). Cavitation erosion of metals and alloys. Wear, 93,
249–260.
36. Preece, C. M. (1980). A comparison of liquid impact erosion and cavitation erosion. Wear,
60, 269–284.
37. Suslick, K. S., & Doktycz, S. J. (1989). The sonochemistry of zinc powder. Journal of the
American Chemical Society, 111(6), 2342–2344.
38. Casadonte, D. J., Jr., & Sweet, J. D. (1998). Tribochemical applications of high-intensity
ultrasound: sonochemical production of intermetallic coatings in heterogeneous media.
Journal of Tribology, 120(3), 641–643.
39. Suslick, K. S., Casadonte, D. J., & Doktycz, S. J. (1989). Ultrasonic irradiation of copper
powder. Chemistry of Materials, 1(1), 6–8.
40. Suslick, K. S., & Doktycz, S. J. (1990). Effects of ultrasound on surfaces and solids. Advances
in Sonochemistry, 1, 197–230.
41. Suslick, K. S., & Casadonte, D. J. (1987). Heterogeneous sonocatalysis with nickel powder.
Journal of the American Chemical Society, 109(11), 3459–3461.
42. Prozorov, T., Prozorov, R., Snezhko, A., & Suslick, K. S. (2003). Sonochemical modification
of the superconducting properties of MgB2. Applied Physics Letters, 83(10), 2019–2021.
43. Prozorov, T., McCarty, B., Cai, Z., Prozorov, R., & Suslick, K. S. (2004). Effects of high-
intensity ultrasound on Bi2Sr2CaCu2O8 þ x superconductor. Applied Physics Letters, 85(16),
3513–3515.
44. Gopi, K. R., & Nagarajan, R. (2008). Advances in nanoalumina ceramic particle fabrication
using sonofragmentation. IEEE Transactions on Nanotechnology, 7(5), 532–537. doi:10.
1109/tnano.2008.2002985.
45. Zeiger, B. W., & Suslick, K. S. (2011). Sonofragmentation of molecular crystals. Journal of
the American Chemical Society, 133(37), 14530–14533. doi:10.1021/ja205867f.
46. Sander, J. R. G., Zeiger, B. W., & Suslick, K. S. (2014). Sonocrystallization and sonofrag-
mentation. Ultrasonics Sonochemistry, 21(6), 1908–1915. doi:10.1016/j.ultsonch.2014.02.
005.
47. Kim, H. N., Sander, J. R. G., Zeiger, B. W., & Suslick, K. S. (2015). Spray sonocrystal-
lization. Crystal Growth & Design, 15(4), 1564–1567. doi:10.1021/acs.cgd.5b00072.
48. Chandra, B. P. (1996). Luminescence induced by moving dislocations in crystals. Radiation
Effects and Defects in Solids, 138(1-2), 119–137.
49. Grabec, I. (1974). Triboluminiscence of rubber. Journal of Polymer Science, Polymer Letters
Edition, 12(10), 573–576.
50. Dickinson, J. T. (1988). Photon-emission from peeling pressure sensitive adhesives. Pro-
ceedings of SPIE The International Society for Optical Engineering 910 (pp. 13–20), Fluo-
rescence Detection II).
51. Budakian, R., Weninger, K., Hiller, R. A., & Putterman, S. J. (1998). Picosecond discharges
and stick-slip friction at a moving meniscus of mercury on glass. Nature, 391(6664),
266–268.
268 N.C. Eddingsaas
52. Zink, J. I., Hardy, G. E., & Sutton, J. E. (1976). Triboluminescence of sugars. The Journal of
Physical Chemistry, 80(3), 248–249.
Physical Review B, 21(2), 816–826.
55. Hardy, G. E., Chandra, A. P., Zink, J. I., Adamson, A. W., & Fukuda, R. (1979). Pulsed
infrared laser induced visible luminescence. Los Angeles, CA: Department of Chemistry,
University of Southern California.
56. Wick, F. G. (1940). Triboluminescence of sugar. Journal of the Optical Society of America,
30, 302–306.
57. Angelos, R., Zink, J. I., & Hardy, G. E. (1979). Triboluminescence spectroscopy of common
candies. Journal of Chemical Education, 56(6), 413–414.
58. Brady, B. T., & Rowell, G. A. (1986). Laboratory investigation of the electodynamics of rock
fracture. Nature, 321, 488–492.
59. Chandra, B. P. (1998). Mechanoluminescence. In D. R. Vij (Ed.), Luminescence of Solids
(pp. 361–389). New York, NY: Plenum.
60. Walton, A. J. (1977). Triboluminescence. Advances in Physics, 26(6), 887–948.
61. Zink, J. I. (1978). Triboluminescence. Accounts of Chemical Research, 11(8), 289–295.
62. Chandra, B. P., Khan, M. S., & Ansari, M. H. (1998). Cleavage mechanoluminescence in
crystals. Crystal Research and Technology, 33(2), 291–302.
triboluminescence of fluorescent molecular crystals. The Journal of Chemical Physics, 73
(12), 5933–5941.
64. Chandra, B. P. (1977). Kinetics of triboluminescence in crystals. Journal of Physics D
Applied Physics, 10, 1531–1538.
65. Sweeting, L. M. (2001). Triboluminescence with and without air. Chemistry of Materials, 13
(3), 854–870.
66. Sweeting, L. M., Cashel, M. L., & Rosenblatt, M. M. (1992). Triboluminescence spectra of
organic crystals are sensitive to conditions of acquisition. Journal of Luminescence, 52,
281–291.
67. Harvey, E. N. (1939). The luminescence of sugar wafers. Science, 90, 35–36.
68. Hesse, P. J., Battino, R., Scharlin, P., & Wilhelm, E. (1996). Solubility of gases in liquids. 20.
Solubility of He, Ne, Ar, Kr, N2, O2, CH4, CF4, and SF6 in n-alkanes n-ClH2lþ2 (6 < l < 16) at
298.15 K. Journal of Chemical & Engineering Data, 41, 195–201.
69. Nakayama, K. (1999). Triboemission of charged particles and resistivity of solids. Tribology
Letters, 6, 37–40.
70. Chandra, B. P., & Zink, J. I. (1982). Triboluminescence of crystals containing the triphenyl
group. The Journal of Physical Chemistry, 86(21), 4138–4141.
71. Donaldson, E. E., Dickinson, J. T., & Shen, X. A. (1986). Time and size correlations of
photon and radiowave bursts from peeling pressure sensitive adhesives in air. Journal of
Adhesion, 19(3-4), 267–286.
72. Chandra, B. P., & Zink, J. I. (1981). Triboluminescence of triclinic crystals. Journal of
Luminescence, 23(3-4), 363–372.
73. Chandra, B. P., & Zink, J. I. (1980). Triboluminescence of inorganic sulfates. Inorganic
Chemistry, 19(10), 3098–3102.
74. Hoff, M. C., & Boord, C. E. (1950). Triboluminescence in cis-4-octene. Journal of the
American Chemical Society, 72, 2770–2771.
75. Sweeting, L. M., & Rheingold, A. L. (1988). Crystal structure and triboluminescence.
1. 9-anthryl carbinols. The Journal of Physical Chemistry, 92(20), 5648–5655.
76. Hurt, C. R., McAvoy, N., Bjorklund, S., & Filipescu, N. (1966). High-intensity tribolumi-
nescence in europium tetrakis(dibenzoylmethide)triethylammonium. Nature, 212(5058),
179–180.
77. Belyaev, L. M., & Martyshev, Y. N. (1969). Triboluminescence of some alkali halide
crystals. Physica Status Solidi, 34(1), 57–62.
78. Walton, A. J., & Botos, P. (1978). The application of an image intensifier spectroscope in
triboluminescent studies. Journal of Physics E: Scientific Instruments, 11, 513–514.
79. Chandra, B. P. (1982). Crystalloluminescence and mechanoluminescence spectra of barium
chlorate hydrate (Ba(ClO3)2.H2O) and potassium sulfate.sodium sulfate (2K2SO4.Na2SO4)
crystals. Crystal Research and Technology, 17(4), 505–508.
80. Nakayama, K., & Nevshupa, R. A. (2003). Characteristics and pattern of plasma generated at
sliding contact. Journal of Tribology, 125(4), 780–787.
81. Suslick, K. S., & Flint, E. B. (1987). Sonoluminescence of non-aqueous liquids. Nature, 330
(6148), 553–555.
82. Suslick, K. S., Gawienowski, J. J., Schubert, P. F., & Wang, H. H. (1983). Alkane
sonochemistry. The Journal of Physical Chemistry, 87(13), 2299–2301.
83. Flint, E. B., & Suslick, K. S. (1989). Sonoluminescence from nonaqueous liquids: emission
from small molecules. Journal of the American Chemical Society, 111(18), 6987–6992.
84. Durrant, S. F., & Bica de Moraes, M. A. (1995). Conventional and dynamic actinometry of
discharges of hydrocarbon-oxygen-argon mixtures. Journal of Vacuum Science & Technol-
ogy, 13(5), 2513–2518.
85. Benndorf, C., Joeris, P., & Kroeger, R. (1994). Mass and optical emission spectroscopy of
plasmas for diamond synthesis. Pure and Applied Chemistry, 66(6), 1195–1205.
86. Sharipov, G. L., Tukhbatullin, A. A., & Abdrakhmanov, A. M. (2012). Detection of OH
radical and O atom during triboluminescence of hydrated cerium/terbium sulfates. Journal of
Luminescence, 132(1), 175–177. doi:10.1016/j.jlumin.2011.08.010.
87. Doktycz, S. J. (1990). The chemical effects of ultrasound. Ph.D., University of Illinois at
Urbana-Champaign, Urbana, IL.
Triboluminescent damage sensors. Smart Materials and Structures, 8, 504–510.
89. Mastro, S. A., Mathur, V. K., Jarrett, A. W. (2004). Fiber optic composite damage sensor.
United States Patent US 6,710,328 B1
90. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Artificial skin to sense
mechanical stress by visible light emission. Applied Physics Letters, 74(9), 1236–1238.
91. Beyer, M. K., & Clausen-Schaumann, H. (2005). Mechanochemistry: The mechanical acti-
vation of covalent bonds. Chemical Reviews, 105(8), 2921–2948.
92. Kajdas, C. K. (2005). Importance of the tirboemission process for tribochemical reaction.
Tribology International, 38, 337–353.
93. Heinicke, G. (1984). Tribochemistry. Berlin: Carl Hanser Verlag Munchen.
94. Boldyrev, V. V. (1994). Mechanochemistry and sonochemistry. Ultrasonics Sonochemistry,
2(2), S143–S145.
95. Gutman, E. M. (1998). Mechanochemistry of materials. Cambridge: Cambridge Int Science
Publishing.
96. Kajdas, C., Furey, M. J., Ritter, A. L., & Molina, G. J. (2002). Triboemission as a basic part of
the boundary friction regime: A review. Lubrication Science, 14(2), 223–254.
97. Dickinson, J. T., Jensen, L. C., & Jahan-Latibari, A. (1984). Fracto-emission: The role of
charge separation. Journal of Vacuum Science & Technology, 2(Pt 2), 1112–1116.
98. Freeman, G. R., & March, N. H. (1999). Technical note: Triboelectricity and some associated
phenomena. Materials Science and Technology, 15(12), 1454–1458.
99. Nakayama, K., & Hashimoto, H. (1995). Triboemission of charged particles and photons
from wearing ceramic surfaces in various hydrocarbon gases. Wear, 185(1-2), 183–188.
100. Zink, J. I. (1981). Squeezing light out of crystals: Triboluminescence. Naturwissenschaften,
68(10), 507–512.
101. Lelevkin, V. M., Otorbaev, D. K., & Schram, D. C. (1992). Physics of non-equilibrium
plasmas. Amsterdam: North-Holland.
270 N.C. Eddingsaas
102. Peratt, A. L. (1997). Advances in numerical modeling of astrophysical and space plasmas.
Astrophysics and Space Science, 242, 93–163.
103. Cooper, J. (1966). Plasma apectroscopy. Reports on Progress in Physics, 29, 35. doi:10.1088/
0034-4885/29/1/302.
104. Griem, H. R. (1997). Principles of plasma spectroscopy. New York, NY: Cambridge Uni-
versity Press.
105. Luque, J., Kraus, M., Wokaun, A., Haffner, K., Kogelschatz, U., & Eliasson, B. (2003). Gas
temperature measurement in CH4/CO2 dielectric-barrier discharges by optical emission
spectroscopy. Journal of Applied Physics, 93(8), 4432–4438.
106. Cooper, J. L., & Whitehead, J. C. (1993). Rotational and vibrational energy transfer in
methylidyne(A2D). Journal of the Chemical Society, Faraday Transactions, 89(9),
1287–1290.
107. Brennen, W., & Carrington, T. (1967). Chemiluminescence of CH in the O þ C2H2 reaction:
rotational relaxation and quenching. The Journal of Chemical Physics, 46(1), 7–18.
108. Dixon, R. N., Newton, D. P., & Rieley, H. (1987). Collisionally induced rotational energy
transfer within the A 2D state of methylidyne. Journal of the Chemical Society, Faraday
Transactions 2: Molecular and Chemical Physics, 83(4), 675–682.
109. Nagata, T., Suzuki, M., Suzuki, K., Kondow, T., & Kuchitsu, K. (1984). Λ-Doublet
populations in methylidyne (CH)(A2D) produced in the 193 nm multiphoton dissociation
of 2-propanone, 2-propanone-d6, thiobismethane, and nitromethane. Chemical Physics, 88
(1), 163–170.
110. Kito, Y., Sakuta, T., & Kamiya, A. (1984). Thomson scattering of laser light from a high
pressure air arc discharge and its applicaiton to electron density measurement. Journal of
Physics D, 17, 2283–2290.
111. Glenzer, S. H., Rozmus, W., MacGowan, B. J., Estabrook, K. G., De Groot, J. D.,
Zimmerman, G. B., et al. (1999). Thomson scattering from high-Z laser-produced plasmas.
Physical Review Letters, 82(1), 97.
112. Burgess, D. D., & Cairns, C. J. (1971). Experimental studies of helium line broadening in a
plasma. II. Low electron densities (lines with forbidden components). Journal of Physics B:
Atomic and Molecular Physics, 4(10), 1364–1376.
113. Barnard, A. J., Cooper, J., & Smith, E. W. (1975). Stark broadening tables for helium (He I) l
4922 Ang. Journal of Quantitative Spectroscopy & Radiative Transfer, 15(6), 429–437.
114. Schoning, T. (1994). Stark broadening of He I lines. Journal of Physics B: Atomic, Molecular
and Optical Physics, 27(19), 4501–4511.
115. Dong, L., Ran, J., & Mao, Z. (2005). Direct measurement of electron density in
microdischarge at atmospheric pressure by Stark broadening. Applied Physics Letters, 86
(16), 161501/161501–161501/161503.
116. Griem, H. R., Baranger, M., Kolb, A. C., & Oertel, G. (1962). Stark broadening of neutral
helium lines in a plasma. Physical Review, 125(1), 177.
117. Griem, H. R. (1974). Pure and Applied Physics (Vol. 39: Spectral line broadening by
plasmas). New York: Academic.
118. Kepple, P., & Griem, H. R. (1968). Improved Stark profile calculations for the hydrogen lines
Hα, Hβ, Hγ, and Hdelta. Physical Review, 173(1), 317–325.
119. Laux, C. O., Spence, T. G., Kruger, C. H., & Zare, R. N. (2003). Optical diagnostics of
atmospheric-pressure air plasmas. Plasma Sources Science & Technology, 12(2), 125–138.
120. Yubero, C., Calzada, M. D., & Garcia, M. C. (2005). Using the Stark broadening of the Ha,
Hb and Hg lines for the measurement of electron density and temperature in a plasma at
atmospheric pressure. Journal of the Physical Society of Japan, 74(8), 2249–2254.
121. Yanguas-Gil, A., Focke, K., Benedikt, J., & von Keudell, A. (2007). Optical and electrical
characterization of an atmospheric pressure microplasma jet for Ar/CH4 and Ar/C2H2
mixtures. Journal of Applied Physics, 101(10), 103307/103301–103307/103308.
122. Wang, Q., Koleva, I., Donnelly, V. M., & Economou, D. J. (2005). Spatially resolved
diagnostics of an atmospheric pressure direct current helium microplasma. Journal of Physics
D Applied Physics, 38(11), 1690–1697.
123. Whiting, E. E. (1968). An empirical approximation to the Voigt profile. Journal of Quanti-
tative Spectroscopy & Radiative Transfer, 8, 1379–1384.
124. Kelleher, D. E., & Wiese, W. L. (1973). Observation of ion motion in hydrogen stark profiles.
Physical Review Letters, 31(24), 1431.
125. Wiese, W. L., Kelleher, D. E., & Helbig, V. (1975). Variations in Balmer-line Stark profiles
with atom-ion reduced mass. Physical Review A, 11(6), 1854.
126. Gigosos, M. A., & Cardenoso, V. (1996). New plasma diagnosis tables of hydrogen Stark
broadening including ion dynamics. Journal of Physics B: Atomic, Molecular and Optical
Physics, 29, 4795–4838.
127. Adler, H. G., & Piel, A. (1991). Stark-broadening of the helium lines 447 and 492 nm at low
electron densities. Journal of Quantitative Spectroscopy & Radiative Transfer, 45, 11–31.
128. Barnard, A. J., Cooper, J., & Smith, E. W. (1974). The broadening of He I lines including ion
dynamic corrections, with application to 4471 A*. Journal of Quantitative Spectroscopy &
Radiative Transfer, 14, 1025–1077.
129. Burgess, D. D., & Jenkins, J. E. (1971). Experimental studies of helium line broadening in a
plasma: I. High electron densities. Journal of Physics B: Atomic and Molecular Physics, 4,
1353–1363.
130. Birkeland, J. W., Bacon, M. E., & Braun, W. G. (1971). Profiles of the He I 4471.5 and 4922
A lines in a dc arc. Physical Review A: Atomic, Molecular, and Optical Physics, 3, 354–358.
131. Mewe, R. (1967). Relative intensity of helium spectral lines as a function of electron
temperature and density. British Journal of Applied Physics, 18(1), 107–118.
132. Becker, K. H., Schoenbach, K. H., & Eden, J. G. (2006). Microplasmas and applications.
Journal of Physics D Applied Physics, 39(3), R55–R70.
133. Xu, X. (2001). Dielectric barrier discharge – Properties and applications. Thin Solid Films,
390(12), 237–242.
134. Kogelschatz, U., Eliasson, B., & Egli, W. (1999). From ozone generators to flat television
screens: history and future potential of dielectric-barrier discharges. Pure and Applied
Chemistry, 71(10), 1819–1828.
Chapter 9
Triboluminescence of Inorganic Lanthanide
Salts
G.L. Sharipov and A.A. Tukhbatullin
9.1 Introduction
In recent years, lanthanide compounds characterized by high luminescence yields

have been included in the range of triboluminescent materials promising for the
design of sensors for damage monitoring in construction materials. These com-
pounds are mainly trivalent lanthanide complexes with complicated organic che-
lating ligands [1–5]. Meanwhile, it was found that simple inorganic lanthanide salts
exhibit fairly intense triboluminescence having, moreover, a multi-emitter nature
[6–9]. These compounds proved to be suitable for investigation of many previously
unknown details of mechanisms giving rise to triboluminescence and the effects of
additives (gases, solids) and for elucidation of the mechanisms of activating and
quenching effects of the additives important for the future practical applications of
triboluminescence. This chapter mainly deals with generalization of these studies.
9.2 Gas and Solid-State Components of Triboluminescence

Spectra
Although triboluminescence (TL) has long been known, it has been studied less
thoroughly than other luminescence phenomena such as photoluminescence,
chemiluminescence, or electroluminescence. Data on the spectral composition of
TL appeared only by the mid-twentieth century after the development of instru-
ments able to reliably detect weak short-term flashes. Currently, it has been well
G.L. Sharipov (*) • A.A. Tukhbatullin

Institute of Petrochemistry and Catalysis of Russian Academy of Sciences,
141 Prospeckt Oktyabrya, 450075 Ufa, Russia
e-mail: glus@anrb.ru

274 G.L. Sharipov and A.A. Tukhbatullin
established that TL spectra comprise two components observed either together or

separately depending on the sort of crystals, namely, the nitrogen gas component in
the UV range that coincides with the emission spectrum of molecular nitrogen
induced by electric discharge in air and the solid-state component caused by
emission directly from the crystal.
The emission from nitrogen is recorded as narrow lines in the wavelength range
of 290–450 nm. The lines of molecular nitrogen are located very close to one
another; therefore, in early studies, most of the lines in TL spectra were merged
together, and even their exact number could not be determined [10, 11]. These lines
were properly recorded for sugar [12] and for triboluminescence of lanthanide
acetylacetonates [7]. They are attributable to the second positive system of N2
(C3Пu B3Пg) [13]. It was also shown [14–16] that lines of the first negative
system of N2þ (B2∑u X2∑g) are also often observed. As has already been noted,
the nitrogen component is found in the emission spectra of many compounds
exhibiting TL. It does not disappear upon evacuation or displacement of nitrogen
from the surrounding atmosphere by other gases. This fact attests that emission of
nitrogen is largely due to the N2 molecules adsorbed on the crystal surface.
However, in another study [11], no nitrogen component was observed upon
mechanical trituration of sugar in solvents.
The solid-state component of the TL spectrum mainly resembles the photolumi-
nescence (PL) spectrum of crystals or fully coincides with it [12]. However, for
some compounds, the difference between the TL and PL spectra is quite significant.
For example, considerable differences between the positions of emission peaks in
the TL and PL spectra were found for N-acetylanthranilic acid and methyl
9-anthracenecarboxylate [17, 18]. The spectral pattern can be affected by several
factors: first, the geometric conditions of TL and PL experiments, that is, different
positions of the crystals during recording the spectrum [19, 20]; second, TL occurs
throughout the crystal (this can lead to additional absorption and reemission of
light), whereas PL is mainly emitted from the crystal surface, being excited by
surface irradiation by a directed light beam [18]; and, third, crystal fracture gives
rise to new crystal surfaces changing the crystal symmetry [21], which can also
change the emission spectrum.
Among other specific features of TL spectra of various materials, the following
can be noted. The presence of nitrogen component together with the solid-state one
is most typical of inorganic materials [22]. In this study, we considered a number of
inorganic salts among which TL was detected only for piezoelectric crystals.
The authors of another work [23] demonstrated that the organic crystals they
studied exhibit TL upon formation of new surfaces and cracking. The duration of a
single TL pulse, apparently related to the formation of one crack, is about several
microseconds. The time of continuous TL is much longer; this implies that contin-
uous TL is the overall emission caused by the formation of numerous microcracks.
The generation of TL upon destruction of organic crystal suggests that the lumi-
nescence mechanism is related to the appearance of oppositely charged surfaces
and the subsequent breakdown. The electric breakdown arising between the sur-
faces is sufficient to excite the adsorbed nitrogen and to produce, at least, the gas
component of the TL spectrum.
9 Triboluminescence of Inorganic Lanthanide Salts 275
While passing to characteristics of the TL spectra of inorganic lanthanide salts,

we would like to note that lanthanide compounds produce the most intense tribo-
luminescence. Furthermore, lanthanide TL is distinguished by specific features.
The TL spectra of many Ln3þ ions (Sm3þ, Eu3þ, Gd3þ, Tb3þ, Dy3þ, etc.) exhibit
specific narrow bands [24]. This feature is due to f-f transitions between the terms
within the electron shell.
The TL spectra of organic lanthanide compounds do not necessarily contain the
nitrogen component. However, in the TL spectra of inorganic lanthanide salts, both
solid-state and nitrogen components are clearly seen [8]. Nevertheless, the contri-
butions of spectral components to the total emission intensity and their ratio can
considerably change from one experiment to another and depends on the exposure
time and other experimental conditions. This is especially true for the nitrogen
component. The possibility of line resolution for this component is determined by
the nature of the lanthanide. For example, in the case of terbium salts, the spectral
ranges of nitrogen and the lanthanide ion are clearly separated, and a well-resolved
spectrum of the nitrogen component can be recorded, although its intensity can be
10–100 times lower than the emission intensity of the lanthanide ion (terbium
chloride, carbonate, or acetylacetonate) [8].
Considering the influence of the nature of inorganic anions on the TL intensity,
note that lanthanide sulfates exhibit the highest TL intensity, while the least
intensity of TL is found for nitrates (Table 9.1).
We will start the detailed discussion of the TL of salts of individual lanthanides
from the cerium ion.
The emission band of the Ce3þ ion, having moreover a high intensity, is located
in the 300–420 nm range with a maximum at about 355 nm and a shoulder at
365 nm. The form of this rather broad band is due to transitions from the lower 2D
level to the 2F7/2 and 2F5/2 levels of the 2F manifold split by spin–orbit coupling
[24–27] (Fig. 9.1). Most of emission lines of nitrogen occur in this TL spectral
range. This superimposition interferes with the separation of components of the TL
spectra of cerium salts and precludes recording of well-resolved nitrogen lines
[8, 22]. At low resolution, the TL spectrum virtually coincides with the PL spectrum
Table 9.1 Estimate of relative intensity of solid-state and nitrogen components of spectrum TL
(Isolid/Igas) of lanthanide salts
Element
Salt Ce Pr Eu Gd Tb Dy
Sulfate ILn3þ 10,000 1000 2000 100 5000 800
IN2 1000 1000 1000 1 ~1000 800
Carbonate ILn3þ 300 – 100 1 30 ~0
IN2 ~30 – 100 10 0–30 10
Chloride ILn3þ 0 0 0 0 10 0
IN2 ~0 ~0 ~0 ~0 1 0
Nitrate ILn3þ – – – – <1 <1
IN2 – – – – 0 0

Ce2(SO4)3 8H2O,
Δλ ¼ 12 nm
Fig. 9.2 TL spectrum

praseodymium sulfate,
Δλ ¼ 6 nm
of cerium crystals, the differences being only in the line intensity. Evidently, the
high TL intensity in the 355 nm maximum is caused not only by the high lumines-
cence quantum yield of Ce3þ but also by the fact that the TL spectrum shows the
sum of Ce3þ and nitrogen emissions.
The TL spectra of praseodymium salts (Fig. 9.2) contain a solid-state component
coinciding with the photoluminescence spectrum of this ion [28, 29] and the
nitrogen component [8]. The TL spectrum in the Pr3þ emission range is the first
example of TL being excited in the far (<300 nm) UV spectral region [8].
The Eu3þ luminescence as a red ring arising upon grinding of europium sulfate
crystals by a stirring rod against the cell bottom can be observed by naked eye in a
slightly darkened room (Fig. 9.3).
The TL of europium salts includes the nitrogen and solid-state components
(Fig. 9.4). The PL spectrum and the solid-state part of the TL spectrum of

scence photo of europium
sulfate
Fig. 9.4 TL spectra Eu2(CO3)3 3H2O (a) and Eu2(SO4)3 8H2O (b), Δλ ¼ 6 nm. The europium
bands in Eu2(SO4)38H2O are shown with dotted line because of their higher intensity
(see right axis)

scence photo of terbium
sulfate
europium(III) exhibit bands corresponding to transitions from two excited states, 5

D0 and 5D1, to the 7F0–6 ground multiplet levels [30]. Usually the intensity of
luminescence bands caused by transitions from the 5D1 level is weaker than the
intensity of the bands associated with transitions from the 5D0 level, as fast
nonradiative deactivation from the 5D1 to 5D0 level takes place and thus the
population of the 5D1 level markedly decreases. Owing to high luminescence
intensity, the TL spectra of europium(III) sulfates (Fig. 9.4b) show bands
corresponding to transitions from the 5D1 level: 5D1 ! 7F1 at 535 nm and 5D1 ! 7
F2 at 550 nm [31]. The bands corresponding to the 5D0 ! 7F0,1 (580 and 590 nm), 5
D0 ! 7F4 (λ 700 nm), and 5D0 ! 7F2 transitions (a maximum at 612 nm) are the
most intense PL and TL bands in the visible region.
The TL spectrum of gadolinium(III) salts, like their PL spectrum, has an intense
ultraviolet band with a maximum at 311 nm corresponding to the 6 P = ! 8 S = 7
2
7
2
transition of the Gd3þ ion [32].

The green emission from terbium(III) salts induced by mechanical treatment is
observable by naked eye (Fig. 9.5). The TL spectrum of terbium salts contains
clearly resolved nitrogen and solid-state components (Fig. 9.6). The most intense
maxima are located in the visible region and coincide with the PL maxima of the
crystals. The PL and TL spectra of terbium(III) are due to transitions from the 5D4
level to the 7Fj ( j ¼ 0. . .6) ground multiplet levels.
For some crystals of terbium salts, for example, Tb(NO3)3∙5H2O, TL cannot be
detected using monochromators, although the PL of these crystals is ~4 times more
intense than the PL of terbium chloride crystals [8, 33]. In turn, the TL spectrum of
terbium chloride does not contain a nitrogen component [8].
The TL spectrum of dysprosium(III) salts consists of the nitrogen component
and solid-state component, which resembles the PL spectrum of the crystals.
However, in the case of TL, apart from the bands at 480 and 570 nm present
in the PL spectrum, low-intensity bands at 665 nm (4 F !6 H transition) and 750 nm
(4 F !6 H transition) are also readily registered. In the TL spectrum, the maximum
Fig. 9.6 TL spectra

Tb2(SO4)3 8H2O (a) and
TbCl3 6H2O, Δλ ¼ 4 nm
(a) and 10 nm (b)
Fig. 9.7 Spectra of crystals

dysprosium sulfate TL
(solid line) and PL (short
dash dot) λex ¼ 350 nm,
Δλ ¼ 8 nm for TL and PL
spectra
at 570 nm is more intense than that at 480 nm (Fig. 9.7) to a higher extent
compared to this intensity ratio in the PL spectrum. Evidently, the intensity
distribution among the transitions in the TL spectrum of Dy3þ differs from that
in the PL spectrum [34].
9.3 Activation and Quenching of Triboluminescence

in Gas Atmosphere
9.3.1 Effect of Monatomic Inert Gases
Purging of samples with inert gases, which displace air from the cell, affects both
TL components. The positions of bands of Ln3þ ions and nitrogen lines barely
change in the atmosphere of these gases or upon evacuation of the cell. However,
the spectral composition of the gas component and the emission intensity depend on
the particular gas used and the purging conditions [8, 35].
For example, while discussing the gas component of the TL spectrum resulting
from electrostatic charging of crystals by mechanical treatment and from accom-
panying discharges, most researchers treat this component only as the emission
from molecular nitrogen. Meanwhile, in some cases, it was noted that the TL
intensity changes in the atmosphere of various gases and that emission from these
gases may also be involved [36, 37]. However, clear spectral evidence for these
facts in the literature devoted to TL is scarce. Nevertheless, the discharge of various
gases was reliably detected for a new variety of TL called sonotriboluminescence
(STL) produced by sonolysis of crystal suspensions [14].
The introduction of inert gases into the cell increases the integrated TL intensity
of lanthanide salts approximately 30- to 40-fold compared to the intensity in air
[35]. Figure 9.8 shows the variation of the integrated intensity of TL of terbium
carbonate during injection of argon recorded using an interference filter with a
Fig. 9.8 The integrated intensity of TL of terbium carbonate during the Ar injection
transmission maximum at 545 nm. The insets show the photo images of the
emission looking like green rings consisting of indiscernible tiny terbium salts
crystals being rubbed by the stirring rod against the cell bottom. As can be seen
from the images, an increase in the Tb3þ TL intensity during injection of argon can
be easily detected visually.
The lines corresponding to the transitions of N2 (C3Пu B3Пg) and N2þ (B3
∑u X3∑g) play a special role in TL. Previously, in a study of the TL of sugar [11],
it was also noted that they do not disappear upon nitrogen displacement from the
atmosphere by other gases or upon moderate (down to 101 Torr) evacuation. These
facts attest to a considerable contribution of adsorbed N2 molecules to TL.
Apart from these features of the gas component of TL spectra, it was found that
after purging with various gases, the lines of these gases are present in the gas
component spectrum together with the nitrogen lines [35]. Indeed, 12 separate lines
for the 4p–4s transitions are observed for argon in the 670–850 nm range
(Fig. 9.9a).
Approximately the same picture is found when the cell is purged with helium:
the TL spectrum contains helium and nitrogen lines (Fig. 9.9b). Helium lines by
themselves do not make a considerable contribution to the increase in the overall
TL intensity, as He emission is of low intensity.
Figure 9.10 shows a TL spectrum of cerium sulfate in the region of 550–800 nm
upon neon injection. The principal lines of neon corresponding to the 3p–3s electron
Fig. 9.9 (a) TL spectra Tb2(SO4)3 8H2O during the Ar injection, Δλ ¼ 2.5 nm, (b) TL spectra
Ce2(SO4)3 8H2O during the He injection, Δλ ¼ 4 nm

cerium sulfate in the region
550–800 nm during the Ne
injection, Δλ ¼ 4 nm
Fig. 9.11 (a, b) TL spectra Ce2(SO4)3 8H2O during the Xe injection, Δλ ¼ 4 nm
transitions can be clearly seen. The spectral pattern is in good agreement with the
reported spectra obtained in an electric discharge in a neon atmosphere [38].
The principal xenon lines corresponding to the 6p–6s electron transition are
located in the near-IR wavelength range (800–1050 nm) (Fig. 9.11a). In the visible
spectral range, a low-intensity Xe line at 475 nm is observed. The TL spectrum of
cerium sulfate in a xenon atmosphere was also found to contain lines corresponding
to the excited Xeþ (Fig. 9.11b). It is noteworthy that as for Ar and He, in the
presence of xenon or neon, the TL spectra of terbium and cerium sulfates exhibit the
lines of these gases together with the nitrogen lines.
It is known from the literature that the electric field strength between
the oppositely charged surfaces formed upon crystal fracture to induce TL can
reach 107–108 V/cm [39]. The ionization potential of xenon I ¼ 12.13 eV, the
threshold electric field strength (Ethr) where xenon starts to be ionized is 107 V/
cm. At this Ethr value, ionization of helium, neon, and argon atoms does not occur as
yet [40]. Since no lines of ions of these gases were detected in the TL spectra,
presumably, the electric field strength in the reported experiments [35] did not
exceed 107 V/cm.
9.3.2 Effect of Polyatomic Gases
Purging with polyatomic gases (O2, SO2, CO2) tends to quench the TL. However,
oxygen and carbon dioxide barely affect the intensity of the solid-state TL compo-
nent (Fig. 9.12) but all gases suppress the emission of nitrogen.
The most pronounced quenching of nitrogen lines is induced by purging the cell
with O2 (Fig. 9.12) and CO2. Figure 9.12b shows how the introduction of oxygen
affects the TL intensity of Tb2(SO4)38H2O in the UV region isolated by the color
filter, which transmits the 260–400 nm wavelengths covering the principal nitrogen
lines. It can be seen that oxygen quenches the nitrogen component almost
completely. Evidently, these findings reflect the fact that the TL gas component
involves both atmospheric gases that surround the crystal and the molecules
adsorbed on the crystal surface. Thus, obviously, oxygen strongly quenches the
emission from nitrogen. Evacuation or purging with helium or argon weakens the
Fig. 9.12 (a) Quenching of nitrogen by oxygen in TL of Tb2(CO3)3 3H2O: TL spectrum before
the O2 injection (1), during the O2 injection (2), in 5 min after the cessation of feeding O2 (3).
Δλ ¼ 8 nm. (b) Quenching of nitrogen line (λ ¼ 260–400 nm) by oxygen in TL Tb2(SO4)3 8H2O
Fig. 9.13 Spectrum of

cerium sulfate TL in the
region of 630-800 nm in O2
atmosphere, Δλ ¼ 2 nm
quenching effect. However, despite the sharp decrease in the nitrogen concentration
in the surrounding atmosphere, rather intense nitrogen luminescence in an inert gas
atmosphere is provided by adsorbed N2 molecules, which are desorbed and elec-
tronically excited upon electric discharge. Inert gases do not suppress nitrogen
luminescence in the atmosphere.
Apart from quenching of the gas component, TL in the O2 atmosphere was
characterized by a clear-cut line at 777 nm [9] (Fig. 9.13) corresponding to atomic
triplet oxygen (3P5P 3S5S transition according to reported data [41]). The excited
O atom is evidently a product of the mechanochemical reaction, namely, cleavage
of the oxygen molecule, O2 ! O þ O*, detected upon destruction of lanthanide
salts.
Purging of the cell with SO2 results in complete quenching of both nitrogen and
solid-state TL components of Tb2(SO4)38H2O [42]. Figure 9.14a shows that,
unlike the effect of O2, quenching induced by SO2 does not occur immediately
after gas injection but develops over several tens of seconds. Sulfur dioxide also
quenches the photoluminescence of terbium sulfate crystals (Fig. 9.14b).
Since the Tb3þ ions are in the solid phase, while SO2 molecules are in the gas, it
is obvious that quenching is due not only to luminophore and quencher collisions in
the atmosphere but also to sorption of the quencher on the crystal surface and to
migration of the electronic excitation energy in the solid phase. Therefore, the
emission slowly decreases, because in the beginning of cell purging, the concen-
tration of sorbed SO2 molecules is low but it increases with time together with the
probability of quenching.
In the sulfur dioxide quenching of the TL of terbium sulfate, the efficiency of
quenching of the solid-state component somewhat predominates over quenching of
the nitrogen component; this can be seen from the presented Stern-Volmer plots for
the TL intensity versus the quencher concentration (Fig. 9.15). The quenching is
already noticeable at a quencher concentration of about 104 mol/l.
Fig. 9.14 (a) Quenching of TL of Tb2(SO4)3 8H2O by SO2; (b) Quenching of PL of

Tb2(SO4)3 8H2O by SO2
Fig. 9.15 Quenching of the solid-state (1) and nitrogen (2) component of TL Tb2(SO4)3∙8H2O
by SO2
The rate constants for TL quenching by sulfur dioxide calculated from the Stern-
Volmer equation
I0
¼ 1 þ k τ0 ½SO2 ;
I
where τ0 is the lifetime (640 μs for Tb3þ* [43, 44] and 40 ns for N2 [45]), I0, I are
the initial and current TL intensities, respectively, were as follows: k(Tb3þ) ¼
1.3 107 l mol1 s1, k(N2) ¼ 6 1011 l mol1 s1. Thus, the rate constant
Fig. 9.16 Scheme of the

energy levels and
deactivation of the electron-
excited Tb3þ ion and the N2
molecule in the presence of
SO2 as a quencher
for the quenching of terbium triboluminescence are almost an order of magnitude

higher than those for sulfur dioxide quenching of the photoluminescence of these
ions, k(Tb3þ) ¼ 5.5 105 l mol1 s1 [42]. It is also worth noting that we are
comparing some effective rate constants for quenching, because the quencher
concentration has been determined only for the gas phase, while *Tb3þ ions are
located in the solid phase.
The intensity of the nitrogen component of the triboluminescence of
terbium sulfate decreases by 80 % when the sulfur dioxide concentration in the
reactor is 103 ;mol/l. In our opinion, the decrease in the intensity of TL gas
component is mainly due to dynamic quenching upon gas-phase collisions of
excited nitrogen molecules with sulfur dioxide molecules accompanied by
nonradiative energy transfer (Fig. 9.16).
Quenching of the solid-state component of the TL spectrum of terbium sulfate
cannot be attributed to energy transfer from the excited *Tb3þ ions to upper energy
levels of SO2 (Fig. 9.16). Apparently, quenching follows a more complex mecha-
nism, similar to the mechanism of electronic-vibration nonradiative excitation
energy exchange in aqua complexes caused by the distribution of the electronic
excitation energy of central ion to the O–H bond vibrations of water molecules
surrounding Tb3þ in aqueous solutions or hydrated crystals [46]. Upon contact of a
sulfur dioxide molecule with the hydrated crystal surface, this molecule can enter
the coordination sphere of terbium. When the electronic excitation quantum energy
of terbium ion is equal to the sum of energies of the vibrational levels of the
coordinated molecules (H2O and SO2), resonance energy transfer to these levels
becomes possible. Thus, the excited terbium ions are deactivated at the interface
(crystal surface).
h i
Tb3þ ðH 2 OÞX * þ SO2 ! Tb3þ ðH2 OÞX SO2 * ! Tb3þ# ðH2 OÞX # SO2

! Tb3þ ðH2 OÞX SO2 þ Δ ! Tb3þ ðH2 OÞX þ SO2
In this scheme, # designates the vibrational excitation and Δ means heat.

It is noteworthy that, apart from the indicated mechanisms, sulfur dioxide
molecules can not only merely deactivate the excited ions and molecules but also
directly affect the formation of excited species. Sulfur dioxide is a polar dielectric,
i.e., its molecules are dipoles. According to the electric theory of TL, charge
separation takes place on microcrack walls. Sulfur dioxide molecules should be
readily adsorbed on the microcrack surfaces owing to Coulomb interactions. As a
result, the dielectric permittivity of the microcrack space changes and the effective
charge on the walls decreases. All these processes are expected to reduce the
probability of electric breakdown, and, hence, decrease the emission intensity.
9.4 Mechanochemical Decomposition of the Water

Crystallization
Figure 9.17 shows the low-resolution TL spectrum of terbium sulfate in an

argon atmosphere under pressure. The spectrum exhibits the overall emission
from the Tb3þ ion (solid-state component) and from argon, OH radical, and
molecular nitrogen, which form the gas component of the TL [44].
Fig. 9.17 TL spectrum of terbium sulfate in the cell filled by Ar the 1.3 atm pressure. The inset
shows argon lines
Fig. 9.18 Spectra of

cerium sulfate TL: in the air
(1), in Ar atmosphere
without pressure (2), under
Ar pressure 1.3 atm (3). N2
lines are observed on the
background band of Ce3þ
ion luminescence
In the low-resolution spectrum, the OH maximum at about 309 nm

corresponding to the 0–0 transition of the A2Σþ ! X2Π band is superimposed on
the nitrogen line at 316 nm. As can be seen from Fig. 9.18, this OH maximum is
absent from the TL spectra recorded in air. When the cell is purged with He, N2, or
O2 it is not detected either. Only upon purging with Ar, this maximum and a less
intense maximum at about 283 nm (1–0 transition of the OH band) become
detectable. The intensity of these peaks increases with pressure rise, which is
consistent with published data on the dependence of the OH band intensity on the
argon pressure for direct or alternating current discharge plasma in an argon
mixture with water vapor [47]. The assignment of the new maxima that appear in
the TL spectrum in the presence of argon to the OH radical is based on comparison
of the high-resolution spectrum in the 275–315 nm range with the OH spectrum in
the same range recorded for water vapor in the plasma discharge [47] (Fig. 9.19).
The Figure demonstrates high similarity of the two spectra, which show the fine
structure of the 1–0 (283 and 290 nm maxima) and 0–0 (308.4 and 309.6 nm
principal maxima) transitions.
A somewhat different description of the OH triboluminescence band was given
in an early publication [49]. According to the authors, the OH maximum at about
310 nm is clearly observable even in the TL of LiF in air. The discrepancy between
these results and our data might be due to the use of different subjects of investi-
gation. Also, examination of the TL spectrum given in [49] and comparison of this
spectrum with the nitrogen luminescence spectrum in this spectral range [7, 8, 12,
36] implies that the TL spectrum recorded in air may show the N2 maximum at
316 nm, which is close to the OH maximum, whereas the TL maximum at 310 nm is
registered, as in our case, only in argon atmosphere.
It was assumed that water vapor present in the atmosphere could serve as a
source of OH for the TL in air [49]. In order to verify this assumption, special
experiments were performed. Passing of air, He, N2, or Ar through water heated to
50 C did not give rise to the OH maximum (or did not increase its intensity in the
Fig. 9.19 Spectra under Ar pressure 1.3 atm. for the OH band upon TL Tb2 (SO4)3∙8H2O (1) and
the OD band at TL Tb2 (SO4)3∙nD2O (2). MDR-23, Δλ ¼ 0.5 nm. The emission spectrum of the
plasma of glow discharge of the direct current in a mixture of He and H2O (3) [48]
case of Ar). Meanwhile, when argon was passed through the coil trap cooled to
80 C (with a mixture of liquid nitrogen and chloroform) in order to freeze out the
water impurity, only enhancement of OH luminescence was observed (Fig. 9.18).
Besides, for more detailed proof of mechanochemical decomposition of crystal-
lization water, additional experiments were performed, namely, the H2O molecules
of crystallization were replaced by D2O molecules and the TL spectra of terbium
sulfate obtained under argon were compared [44]. The replacement of light water
by heavy water in the terbium sulfate crystals had a crucial effect on the gas
component of the TL spectrum (Fig. 9.19). The TL spectrum of deuterated crystals
recorded under Ar pressure shows emission from the OD radical. The luminescence
maxima in the 1–0 transition range of the A2Σþ ! X2Π band are markedly shifted to
longer wavelengths on going from the OH to OD radical. Furthermore, the intensity
of the A2Σþ ! X2Π TL band for heavy-water samples is twice higher than for light-
water crystals.
Thus, water vapor entering the cell together with the saturating gas does not
induce OH luminescence in the TL and, obviously, the only source of OH is water
of crystallization contained in the solid. The OH radical is likely to arise in the
decomposition reaction H2O ! OH* þ H, which is induced by electron impacts
resulting from electric discharges caused by mechanical treatment of the crystals.
Yet another possible channel of water decomposition under electron bombard-
ment is the formation of excited hydrogen atom H2O ! OH þ H*. However, no H
line was observed in our TL experiments, which can be attributed to several factors.
It is known [43] that H luminescence in the discharge plasma is much weaker than
the OH luminescence; unlike OH luminescence intensity, the intensity of H lumi-
nescence decreases rather than increases with argon pressure rise in an H2O–Ar

cerium sulfate in the region
of 630–800 nm in the air,
Δλ ¼ 2 nm. The arrow
shows the Hα line
mixture. In addition, around the most intense Hα line at 656.3 nm, there are two
groups of luminescence lines of N2 (B3Πg A3Σþu) at 630–780 nm, which interfere
with the detection of the possible but perhaps weak H line (Fig. 9.20).
Previously, these nitrogen lines were not recorded in the TL spectra. They were
recorded only for the case of STL [15], the intensity of these groups of lines being
approximately 1000 times lower than the intensity of the principal group
(300–500 nm). We observed a different intensity ratio: the “short-wavelength”
group was only approximately 20 times more intense than the “long-
wavelength” one.
9.5 Influence of Solid Additives on Triboluminescence
9.5.1 Quenching of Triboluminescence in Mixed Salts
Along with the effect of gases, TL dependence on the addition of crystalline

compounds also deserves attention. It is known, for instance, that nitrite ion is a
strong quencher of the PL of lanthanides, including terbium [50]. The dynamic
quenching of Tb3þ* upon the reaction with NO2 is attributable to electron
exchange resonance energy transfer from the 5D4 state of Tb3þ to the triplet level
of NO2, which is energetically possible [34]. The NO2 ion, which quenches the
terbium(III) PL in solutions in 105 104 mol/l concentrations, also efficiently
quenches multibubble sonoluminescence of terbium by penetrating into the cavi-
tation bubbles [51]. The effect of not only NaNO2 but also NaNO3 on the intensity
of multibubble sonoluminescence of TbCl3 was studied. As it turned out, nitrate ion
quenches sonoluminescence of terbium as nearly effective as the nitrite one,
although the nitrate ion does not quench PL of Tb3þ ions. This effect is also due
to formation of NO2 from NO3 upon sonolysis of solutions. The rate constant for
Fig. 9.21 The dependence of the intensity of nitrogen (1, 2) and the solid-state (3, 4) TL
component Tb2(SO4)3·8H2O adding Na2SO4 (1, 3) and NaNO2 (2, 4). Recorded with an color
filter (260–400 nm) (1, 2), and an interference filter (λ ¼ 546 nm) (3, 4)
the bimolecular quenching of Tb3þ* ions by nitrite ions in a 0.1 mol l–1 solution of
TbCl3 is 3.5 108 l mol1 s1 [51]. Apparently, one of the reasons for the
absence or very low intensity TL salt Tb(NO3)3∙5H2O [8], may be the presence of
impurities in the salt nitrite ion. On the other hand, electrization and appearance of
strong local electric fields upon destruction of inorganic lanthanide salts are crucial
processes in TL [8, 12], so that quenching TL with dielectric terbium nitrate crystals
is probably due to lack of effective charge, which is accumulated on its surface
upon mechanical effect. This leads to decrease in strength of electric fields, to
probable electrical breakdown in microcracks, and thus to decrease in TL intensity
or complete absence of light.
NaNO2 influences both the solid-state and nitrogen TL components [52]
(Fig. 9.21). For example, it can be seen from Fig. 9.21 that the addition of
50–60 mg of NaNO2 to 200 mg of terbium sulfate induces a decline of the TL
intensity. The addition of 125 mg of NaNO2 suppresses the TL almost completely.
The addition of the same amounts of Na2SO4 does not affect the TL of
Tb2(SO4)38H2O. Most likely, the effect of NaNO2 is due to the action of the NO2

group.
Quenching of the solid-state luminescence is indicative of the migration of
excitation energy not only by the exciton mechanism within crystals but also
between contacting surfaces of crystals. Similar energy transfer processes between
silicon nanocrystals were reported in the literature [53, 54]. In addition, quenching
by NO2 groups of not only the solid-state but also nitrogen TL attests that a
quenching process takes place on the crystal surface:
Fig. 9.22 Scheme of the

energy levels and
deactivation of the electron-
excited Tb3þ ion and the N2
molecule in the presence of
NO2 as a quencher
½N2 *gas þ ½NaNO2 solid ! ½N2 gas þ ½NaNO2 solid þ Δ;
where Δ is the thermal energy transferred to the crystal.

Since the results point to the crucial role of the NO2 ions in quenching, one can
assume that quenching involves a nonradiative deactivation mechanism with
triplet-triplet energy transfer. Indeed, the emission from molecular nitrogen in the
TL is caused by transitions between triplet levels (the C3Пu B3Пg second positive
system [8, 12, 13]). The upper level of this system is located at about 11 eV [55],
and energy transfer from this level is possible not only to the 3B1 level of NO2 but
also to higher triplet levels (Fig. 9.22).
9.5.2 Sonotriboluminescence
and Sonocrystallochemiluminescence in Suspensions
Triboluminescence induced by acoustic cavitation in crystal suspensions called

sonotriboluminescence (STL) was found only recently [14–16]. These references
describe the STL of organic suspensions: resorcinol, saccharose, and m-
aminophenol in higher alkanes—dodecane and hexadecane. The authors concluded
that STL arises during sonolysis of suspensions as a result of fracture of crystal
microparticles upon collisions during movement at high velocity (up to hundreds
m/s) under the action of cavitation shock waves.
It was noted previously that for the nitrogen component to be present in the TL
spectrum, the air atmosphere is not required; in some cases, it is present also in the
atmosphere of other gases or in liquids being provided by the N2 molecules

adsorbed by crystals in the air [7, 12]. The ratio between these components depends
on various conditions, and no correlations determining their contributions to the TL
were identified as yet. The described STL of resorcinol includes a predominant
contribution of the nitrogen component in N2-saturated suspensions. The enhance-
ment of this component in the STL of organic suspensions was attributed [14–16] to
particularly the high velocities of the colliding particles. Usually, TL is induced by
mechanical means that create crystal or deforming impact movement at velocities
of not more than several tens of m/s. In this case, the efficiency of excitation of
nitrogen lines is moderate. For example, in the case of resorcinol TL, these lines are
barely visible against the broad band for the solid-state luminescence of crystals,
which has much higher intensity than the nitrogen lines [15].
The STL was demonstrated only for organic crystals; however, in more recent
studies [33, 55, 56], this method for generating the emission was considered in
relation to other known triboluminescent materials: inorganic crystals and crystals
of metal complexes.
STL arises upon the sonolysis of suspensions of crystalline TbCl3.6H2O or Tb
(acac)3.H2O [33] or other lanthanide salts [55] in undecane, perfluorodecalin, or
water. However, first, the STL intensity of terbium compounds is relatively low,
being about 103 times lower than the resorcinol STL intensity. Second, the changes
in the emission intensity and spectral composition observed on going from TL to
STL of terbium crystals were different from those found for resorcinol. Indeed,
terbium acetylacetonate crystals were reported [7] to exhibit very bright TL
containing both nitrogen component and the solid-state component of intrinsic
emission from the crystals, i.e., Tb3þ* ions. The spectrum of this TL is shown in
Fig. 9.23. It can be seen that the solid-state component is approximately 100 times
Fig. 9.23 TL spectra: (a) Tb(acac)3∙H2O; (b) TbCl3.6H2O, Δλ ¼ 3 nm (a) and 8 nm (b)
more intense than the nitrogen component. Nevertheless, even the latter is an order
of magnitude brighter than the TL of resorcinol under the same conditions. The TL
spectrum of terbium acetylacetonate contains clear-cut lines of N2 (C3Пu B3Пg)
and Nþ2 (B2Σþu X2Σþg) apart from the 316 nm line recorded for the resorcinol
and saccharose TL [12, 14–16].
As opposed to the cited publications [14–16] dealing with the STL of organic
crystals, the nitrogen lines in the STL spectra of lanthanides do not become more
intense. For example, in the STL of terbium sulfate, the intensities of the nitrogen
and solid-state components are nearly equal and in some cases, the nitrogen lines
almost disappear [8]. This is surprising because nitrogen lines are very efficiently
excited, at least, in the TL of terbium acetylacetonate complex and terbium sulfate,
although they are 100 times less intense than the terbium lines. However, no 1000-
fold increase in the intensity of these lines, like that observed by the authors cited
[14–16], takes place on going from TL to STL.
Indeed, the STL of terbium acetylacetonate (Fig. 9.24b), which is 103 times less
intense than the TL of Tb(acac)3∙H2O under the described experimental conditions
[33], exhibit only 490 and 545 nm lines, while the other terbium lines, like nitrogen
lines, are not observed, evidently, due to low intensity. The intensity ratio between
the 490 and 545 nm lines of this spectrum is higher than this ratio in the spectra of
solid-state PL and TL.
Apparently, unlike the STL of resorcinol, in the case of STL of terbium
acetylacetonate, no enhancement of the nitrogen component takes place. Nitrogen
saturation of terbium chloride suspensions also did not exert a significant influence
on the STL: no nitrogen component was detected and the solid-state component of
STL remained invariable. As in the case of terbium acetylacetonate, the relative
intensity of the 490 nm line is enhanced (Fig. 9.24b). In the STL spectrum, this line
is even more intense than the 545 nm line. A similar increase in the intensity ratio
between the two principal lines of terbium in favor of the 490 nm line with respect
to the intensity ratio of these lines in the PL spectrum is also typical of sonolumi-
nescence of TbCl3 in aqueous solutions [57].
The sonoactivation method of TL excitation characterized by high velocities of
colliding particles and formation of gas-filled cavitation pockets in the suspensions
that are in contact with crystal surfaces during sonolysis substantially enhances the
emission of organic crystals that refer to the “discharge” triboluminescent materials
but has a much smaller effect on the luminescence of crystals with intense solid-
state component, among which are the crystals of lanthanide compounds.
Apart from STL, the enhancement of crystalloluminescence under the action of
sonication was also detected. Crystalloluminescence (CRL), i.e., the emission from
crystals that precipitate from supersaturated solutions, has been known for almost
200 years [58]. The mechanism of this phenomenon has long become obscure. It
was demonstrated [59] that in most cases of crystallization from organic solutions,
crystalloluminescence is nothing but chemiluminescence catalyzed by juvenile
(freshly formed) crystal surfaces during thermal decomposition of impurity perox-
ides. These peroxides are easily formed in organic solvents on contact with O2 and
in the light. They are adsorbed on the juvenile surface (L) of the arising crystal
Fig. 9.24 (a) TL spectrum, (b) STL spectrum of TbCl3∙6H2O, and (c) STL spectrum of Tb
(acac)3∙H2O; Δλ ¼ 11 nm. STL spectra were obtained by subtracting of SL spectrum of
perfluorodecalin
nuclei, in particular of lanthanide salts, and then rapidly decompose on the surface
emitting photons (Fig. 9.25).
The key features of the catalyzed thermal decomposition of peroxides were
described [60] in relation to model peroxides—1,2-dioxetanes. Briefly, they are
as follows. In the absence of metal compounds, there is no catalysis and direct
chemiluminescence arises: for example, 1,2-dioxetane unimolecularly decomposes
into two ketone molecules, one of which is excited and emits light. Some metals
enhance the emission without catalysis—via nonradiative energy transfer from the
excited ketone to the metal followed by high quantum yield luminescence [61, 62].
The catalytic ways of enhancement of dioxetane decomposition and lumines-
cence are more diverse. Complex formation without electron transfer accelerating
Fig. 9.25 Scheme of probable processes upon CRL appeared in acetone solutions of metal salts
with peroxide impurities after adding chloroform [60]
the dioxetane decomposition either gives rise to complexation-initiated chemilu-

minescence or leads to dark catalysis of decomposition (without emission)
[63]. Examples of complex formation with electron transfer include photodecom-
position of dioxetane under the action of excited trivalent cerium ion and
chemically induced electron exchange luminescence, i.e., activation of the chemi-
luminescent catalytic dioxetane decomposition by ruthenium(II) [64, 65].
The catalytic decomposition mechanism is disclosed by energy diagram
(Fig. 9.26). Upon complex formation through the electron density withdrawal
from the oxygen lone pairs to the vacant metal ƒ orbitals, the O–O bond is weakened
and can be cleaved with a lower activation energy. This populates the triplet nπ*
antibonding orbitals of the ketone, from which intracomplex energy transfer can
occur to the lower levels of metal ions (europium) [66] but cannot occur to higher
levels (gadolinium) [63]. In the latter case, dark rather than chemiluminescent
catalysis is involved.
Evidently, the complexation centers present on the juvenile surfaces actively
appearing upon sonication of crystal suspensions can also catalyze the decomposi-
tion of peroxides. This assumption was verified in special experiments. The CRL
intensity observed during precipitation of terbium nitrate crystals after the addition
of a hot acetone solution of Tb(NO3)3∙5H2O and isopropylbenzene hydroperoxide
(IPB HP) to chloroform increases upon sonication (Fig. 9.27) [67]. Thus, the glow
observed during sonolysis of a slurry of terbium crystals can be called sonocrys-
tallochemiluminescence (SCRCL).
As can be seen from Fig. 9.27, SCRCL is observed at the background of
sonoluminescence (SL) of the suspension, which is mainly caused by the SL of
the solvent mixture, its intensity being four times higher compared to CRL
observed without sonication. Sonication is known to promote the formation of
juvenile surfaces of disperse crystals [68]. This intensifies the peroxide adsorption
Fig. 9.26 Disintegration of 1,2-dioxetane in the catalyst complex with the metal ion
200 4
150
Intensity (arb. un.)
100 170
1 2 3
50
30 32 35
Fig. 9.27 The luminescence intensity: (1) upon SL of 1:4 chloroform–acetone (I) and (I)/IPB
HP mixtures (II); (2) SL of suspension of terbium nitrate crystals in (I) mixture; (3) CRL of suspension
of terbium nitrate in (II) mixture; (4) SCRCL of suspension of terbium nitrate in (II) mixture.
The concentration of the crystals Tb(NO3)3∙5H2O and IPB HP in acetone was 0.1 mol/l,
temperature of suspension mixtures was 54 1 C, volume was 10 ml, and specific power ultrasound
was 20 W/cm2
Fig. 9.28 (1) CRL spectrum and (2) SCRCL spectrum. The spectra were recorded with the use of
color filters, which were mounted on device for registration of weak luminescence
and catalytic decomposition, the energy of which excites the luminescence centers
of crystals, namely, Tb3þ ions (Fig. 9.28), which are CRL emitters. Note that the
observed SCRCL is not sonotriboluminescence, as it is not characteristic of terbium
nitrate crystal suspensions.
9.6 Conclusion
Currently, it has been ascertained that the TL spectra of inorganic lanthanide salts
exhibit, together or separately, two components: the gas (nitrogen) and solid-state
ones. The spectrum of the latter component is similar to or coincides with the
photoluminescence spectrum of the crystal [12]. The nitrogen component contains
lines in the range of 300–450 nm typical of nitrogen molecule emission upon gas
electric discharge: the second positive system of N2(C3Пu B3Пg) and partially the
first negative system of N2þ(B2∑u X2∑g) [13–16].
The contribution of the components to the emission intensity and the TL spectral
pattern can highly vary from one experiment to another, depending on the action
time and other experimental conditions. Poor reproducibility is especially typical of
the nitrogen component. Analysis of the published data attests to a substantial
contribution of nitrogen molecules adsorbed on the crystals.
The nitrogen component is believed to have an electric discharge nature and to
appear as a result of charge separation on the walls of microcracks formed in
crystals or between the crystal surface and the contacting surface that acts upon
the crystal. Gas phase breakdown is possible between the oppositely charged
surfaces. The mechanism of formation of the solid-state component has not yet
been ultimately clarified.
In order to determine the key TL mechanisms, most promising is probably to

study the TL of lanthanides as a good model system for elucidating the regular
patterns of formation and deactivation of the electronically excited states in het-
erogeneous systems, which are important for the understanding of the mechanisms
of heterogeneous catalysis, energy evolution, and the subsequent development of
new technologies possibly employing the excited species (in particular, sorption,
desorption, evaporation, condensation, and sublimation processes). The long-lived
excited states of lanthanide ions are best suited for this type of studies.
The spectroscopic investigation of the TL of lanthanides is especially preferable
for elucidating the excitation mechanisms of the solid-state components of the TL
spectrum. Comparative studies of the TL spectra of lanthanide ions, together with
the spectra obtained upon other types of luminescence excitation (photo-, electro-,
chemi-, radio-, thermo-, and sonoluminescence [24, 60, 63, 66, 68–74]) in crystals
or solutions can help to get a deeper insight into the mechanistic details of TL
generation.
References
1. Takada, N., Sugiyama, J., & Katoh, R. (1997). Mechanoluminescent properties of europium
complexes. Synthetic Metals, 91, 351–354.
2. Olawale, D. O., Dickens, T., Sullivan, W. G., Okoli, O. I., Sobanjo J.B., & Wang, B. (2011).
131, 1407–1418.
3. Eliseeva, S. V., Pleshkov, D. N., Lyssenko, K. A., Lepnev, L. S., Bunzli, J.-C. G., & Kuzmina
N. P. (2010). Highly luminescent and triboluminescent coordination polymers assembled from
lanthanide β-diketonates and aromatic bidentate O-donor ligands. Inorganic Chemistry, 49,
9300–9311.
4. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2012). Synthesis and
Luminescence, 132, 1812–1818.
5. Fontenot, R. S., Hollerman, W. A., Bhat, K. N., & Aggarwal, M. D. (2013). Effects of added
uranium on the triboluminescent properties of europium dibenzoylmethide triethylammonium.
6. Chandra, B. P., Chandra, V. K., & Piyush, J. (2013). Models for intrinsic and extrinsic fracto-
mechanoluminescence of solids. Journal of Luminescence, 135, 139–153.
7. Bulgakov, R. G., Kuleshov, S. P., Zuzlov, A. N., & Vafin, R. R. (2004). Triboluminescence of
lanthanide acetylacetonates. Russian Chemical Bulletin, 53, 2712–2714.
8. Sharipov, G. L., Tukhbatullin, A. A., & Abdrakhmanov, A. M. (2011). Triboluminescence of
crystals and suspensions of inorganic salts of lanthanides. Protection of Metals and Physical
Chemistry of Surfaces, 47, 13–19.
9. Sharipov, G. L., Tukhbatullin, A. A., & Abdrakhmanov, A. M. (2012). Detection of OH radical
and O atom during triboluminescence of hydrated cerium/terbium sulfates. Journal of Lumi-
nescence, 132, 175–177.
10. Zink, J. I. (1978). Triboluminescence. Accounts of Chemical Research, 11, 289–295.
11. Zink, J. I., Gordon, E. H., & Sutton, J. E. (1976). Triboluminescence of sugars. The Journal of
Physical Chemistry, 80, 248–249.
12. Sage, I., & Bourhil, G. (2001). Triboluminescent materials for structural damage monitoring.
13. El’yashevich, M. A. (2007). Atomnaya i molekulyarnaya spektroskopiya Molekulyarnaya

spektroskopiya. (Atomic and molecular spectroscopy: Molecular spectroscopy). Moscow:
KomKniga.
14. Eddingsaas, N. C., & Suslick, K. S. (2006). Mechanoluminescence: Light from sonication of
crystal slurries. Nature, 444, 163.
129, 6718–6719.
16. Eddingsaas, N. C., & Suslick, K. S. (2007). Plasma characteristics of the discharge produced
during mechanoluminescence. Physical Review Letters, 99, 234301.
17. Duignan, J. P., Oswald, I. D. H.,Sage, I. C., Sweeting, L. M., Tanaka, K., Ishihara, T.,
et al. (2002). Do triboluminescence spectra really show a spectral shift relative to photolumi-
nescence spectra? Journal of Luminescence, 97, 115–126.
18. Sweeting, L. M., Rheingold, A. L., Gingerich, J. M., Rutter, A. W., Spence, R. A., Cox,
C. D., et al. (1997). Crystal structure and triboluminescence 2 9-anthracenecarboxylic acid and
its esters. Chemistry of Materials, 9, 1103–1115.
19. Leyrer, E., Zimmermann, F., Zink, J. I., & Gliemann, G. (1985). Triboluminescence,
photoluminescence, and high-pressure spectroscopy of tetracyanoplatinate salts. Determina-
tion of the pressure at triboluminescent sites. Inorganic Chemistry, 24, 102–106.
20. Zink, J. I., & Klimt, W. (1974). Triboluminescence of coumarin, fluorescence and dynamic
spectral features excited by mechanical stress. Journal of the American Chemical Society, 96,
4690–4692.
21. Chapman, G. N., & Walton, A. J. (1983). Triboluminescence of fluorites. Journal of Physics C:
Solid State Physics, 16, 5543–5551.
22. Chandra, B. P., & Zink, J. I. (1980). Triboluminescence of inorganic sulfates. Inorganic
Chemistry, 19, 3098–3102.
triboluminescence of fluorescent molecular crystals. The Journal of Chemical Physics, 73,
5933–5942.
24. Poluektov, N. S., Kononenko, L. I., Efryushina, N. P., & Bel’tyukova, S. V. (1989). Spektrofo-
tometricheskie i lyuminestsentnye metody opredeleniya lantanidov (Spectrophotometric and
luminescent methods for detection of lanthanides). Kiev: Naukova Dumka.
25. Svetashev, A. G., & Tsvirko, M. P. (1984). Luminescence of Ce3þ ion in aqueous solution.
Optika i Spektroskopiya, 56, 842–846.
26. Zaidel, A. N., & Larionov, Y. I. (1940). Absorption and fluorescence spectra of solutions of
salts of cerium and praseodymium. Izvestiya Akademii Nauk SSR, Seriya Fizicheskaya, 4,
25–27.
27. Zaidel, A. N., Larionov, Y. I., & Filippov, A. N. (1938). The fluorescence of solutions of
cerium salts. Doklady Akademii Nauk SSSR, 20, 351–353.
28. Zaidel, A. N., Larionov, Y. I., & Novikova-Minash, O. B. (1938). About fluorescence and
absorption of salt solutions of praseodymium. Doklady Akademii Nauk SSSR, 21, 330–333.
29. Svetashev, A. G., & Tsvirko, M. P. (1981). Luminescence of rare-earth ions in the excitation of
4f-5d absorption band. Aqueous solutions chloride Tb and Pr. Optika i Spektroskopiya, 51,
1034–1037.
30. Sinha, A. P. B. (1971). Fluorescence and laser action on rare earth chelates. Spectroscopy in
Inorganic Chemistry, 2, 255–288.
31. Tukhbatullin, A. A., Muftakhutdinov, M. R., & Sharipov, G. L. (2013). Triboluminescence of
lanthanide sulfates. Izvestia Ufimskogo Nauchnogo Tsentra RAN, 4, 16–22.
32. Kondrat’eva, E. V., & Lazeeva, G. S. (1960). A study of the duration and intensity of
luminescence of trivalent gadolinium and terbium ions in solutions. Optika i Spektroskopiya,
8, 132–134.
33. Sharipov, G. L., Abdrakhmanov, A. M., & Tukhbatullin, A. A. (2009). Sonotriboluminescence
in suspensions of trivalent terbium compounds. Technical Physics Letters, 35, 452–455.
34. Tanaka, K., Fujita, K., Taniguchi, T., Hirao, K., & Ishihara, T. (2000). Triboluminescence of
alkaline earth aluminate polycrystals doped with Dy3þ. Journal of Applied Physics, 88,
4069–4074.
35. Tukhbatullin, A. A., Sharipov, G. L., Abdrakhmanov, A. M., & Muftakhutdinov, M. R. (2014).
Mechanoluminescence of terbium and cerium sulfates in a noble-gas atmosphere. Optics and
Spectroscopy, 116, 691–694.
854–870.
37. Brady, B. T., & Rowell, G. A. (1986). Laboratory investigations of the electrodynamics of rock
fracture. Nature, 1321, 488–492.
38. Zaidel’, A. N., Prokof’ev, V. K., & Raiskii, S. M. (1952). Tablitsy spektral’nykh linii (tables of
spectral lines). Moscow: GosTekhIzdat.
39. Chandra, B. P., Mahobia, S. K., Kuraria, R. K., & Chaudhary, V. (2007). Fracto-
mechanoluminescence produced during slow deformation of solids. IJEMS, 14, 443–446.
40. Voronov, G. S., & Delone, N. B. (1965). Ionization of the xenon atom by the electric field of
ruby laser emission. JETP Letters, 1, 66–68.
41. Antonov, G. G., Kovshechnikov, V. B., & Gnedovskiy, A. V. (2011). Comparison between the
parameters of a one-phase arc with copper and steel electrodes ignited in nitrogen and in air.
Technical Physics, 56, 65–71.
42. Sharipov, G. L., Tukhbatullin, A. A., Abdrakhmanov, A. M., & Muftakhutdinov, M. R. (2011).
Influence of gases and additives on solid-state and nitrogen components of terbium sulfate
triboluminescence. Vestnik Bashkirskogo Universiteta, 16, 658–661.
43. Mamykin, A. A., Mamykin, A. V., Ostakhov, S. S., & Kazakov, V. P. (2010). Chemilumines-
cence during thermal degradation of sodium persulfate in the presence of terbium sulfate:
2. The effect of water of crystallization. High Energy Chemistry, 44, 109–112.
44. Sharipov, G. L., Tukhbatullin, A. A., Muftakhutdinov, M. R., & Abdrakhmanov, A. M. (2014).
Luminescence of OD radical as an evidence for water decomposition under destruction of the
deuterated terbium sulfate crystal hydrate. Journal of Luminescence, 148, 79–81.
45. Gilmore, F. R., Laher, R. R., & Espy, P. J. (1992). Franck–Condon factors, r‐centroids,
electronic transition moments, and Einstein coefficients for many nitrogen and oxygen band
systems. Journal of Physical and Chemical Reference Data, 21, 1005–1107.
46. Ermolaev, V. L., Bodunov, E. N., et al. (1977). Bezyzlychatel’nyi penenos energii
elektronnogo vozbuzdeniya (Nonradiative energy transfer of electronic excitation). Leningrad:
Nauka.
47. Vul’, A. Y., Kidalov, S. V., Milenin, V. M., Timofeev, N. A., & Khodorkovski, M. A. (1999).
Investigation of a glow discharge in a mixture of Ar and OH. Technical Physics Letters, 25,
321–323.
48. Shuaibov, A. K., General, A. A., Kel’man, V. A., & Shevera, I. V. (2008). The emission
characteristics of low-pressure water vapor discharge UV emitters. Technical Physics Letters,
34, 588–590.
49. Belyaev, L. M., & Martyshev, N. Y. (1969). Triboluminescence of some alkali halide crystals.
Physica Status Solidi, 34, 57–62.
50. Tanaka, F., Ishibashi, T., & Okamoto, M. (1993). Kinetics and mechanism for luminescence
quenching of lanthanide ions by the nitrite ion in aqueous solution. Journal of Photochemistry
and Photobiology A: Chemistry, 1, 15–19.
51. Sharipov, G. L., Abdrakhmanov, A. M., Tukhbatullin, A. A., Gareev, B. M., & Zagretdinova,
L. R. (2009). Action NO3– on sonoluminescence Tb3þ in aqueous solution. In Collection of
Works (MarGTU), 16(3), 252–257.
52. Sharipov, G. L., Tukhbatullin, A. A., Abdrakhmanov, A. M., & Muftakhutdinov, M. R. (2013).
Quenching of tribo- and photoluminescence in a Tb2(SO4)3 8H2O þ NaNO2 crystalline mix-
ture. Technical Physics, 58, 1518–1521.
53. Gusev, O. B., Prokofiev, A. A., Maslova, O. A., Terukov, E. I., & Yassievich, I. N. (2011).
Energy transfer between silicon nanocrystals. JETP Letters, 93, 147–150.
54. Bolesta, I. M., Vistovskii, V. V., Gloskovskaya, N. V., Panasyuk, M. R., & Yaritskaya, L. I.
(2011). High-energy Frenkel cation exciton and specific features of its self-trapping in the
CdI2-PbI2 crystal system. Physics of the Solid State, 53, 799–803.
55. Virot, M., Pflieger, R., Ravaux, J., & Nikitenko, S. I. (2011). Sonoluminescence of Tb(III) at
the extended solid-liquid interface. Journal of Physical Chemistry C, 115, 10752–10756.
56. Virot, M., Pflieger, R., Skorb, E., Ravaux, J., Zemb, T., & M€ ohwald, H. (2012). Crystalline
silicon under acoustic cavitation: From mechanoluminescence to amorphization. Journal of
Physical Chemistry C, 116, 15493–15499.
57. Sharipov, G. L., Gaı̆netdinov, R. K., & Abdrakhmanov, A. M. (2006). Anomalous isotopic
effect in multibubble sonoluminescence of aqueous solutions of terbium chloride. JETP
Letters, 83, 493–496.
58. Harvey, E. N. (1957). A history of luminescence from earliest time until 1900. Philadelphia,
PA: American Philosophical Society.
59. Tolstikov, G. A., Sharipov, G. L., Voloshin, A. I., & Kazakov, V. P. (1984). Chemilumines-
cence under catalysis decay of peroxides on the surface of crystalline substances as the cause
crystalloluminescence. Doklady Akademii Nauk SSSR, 274, 658–661.
60. Sharipov, G. L., Kazakov, V. P., & Tolstikov, G. A. (1990). Khimiya i khemilyuminestsentsiya
1,2-dioksetanov (Chemistry and chemiluminescence of 1,2-dioxetanes). Moscow: Nauka.
61. Sharipov, G. L., Kazakov, V. P., & Tolstikov, G. A. (1987). Chemiluminescent reactions
involving singlet oxygen and dioxetanes. Izvestiya Akademii Nauk, Seriya Fizicheskaya, 51,
559–567.
62. Tolstikov, G. A., Sharipov, G. L., Voloshin, A. I., Ostakhov, S. S., & Kazakov, V. P. (1986).
Mechanism of sensitization of chemiluminescence in the thermolysis of diadamantylidene-
dioxethane. Russian Chemical Bulletin, 35, 715–721.
63. Sharipov, G. L., Ableeva, N. S., & Kukovinets, A. G. (1994). Chemiluminescence initiated by
complex formation in solutions of 1,2-dioxetane and lanthanide perchlorates. Russian Chem-
ical Bulletin, 43, 1500–1504.
64. Ostakhov, S. S., Sharipov, G. L., Voloshin, A. I., Kazakov, V. P., & Tolstikov, G. A. (1986).
Catalysis and sensitization of chemiluminescence at thermal decomposition of
adamantylidene-adamantane-1,2-dioxetane by Tb, Eu and Ru compounds. Doklady Akademii
Nauk SSSR, 287, 1165–1168.
65. Voloshin, A. I., Sharipov, G. L., Kazakov, V. P., & Tolstikov, G. A. (1991). Chemically
initiated electron-exchange luminescence of Ru(bpy)3Cl2 in catalytic reaction with
1,2-dioxetane. Russian Chemical Bulletin, 40, 1158–1162.
66. Sharipov, G. L., Voloshin, A. I., et al. (1990). Luminescence with 5d level (3-diketonate Eu
(III)) in chemiluminescence system Eu (III) -1,2-dioxetane. Doklady Akademii Nauk SSSR,
315, 425–429.
67. Sharipov, G. L., Abdrakhmanov, A. M., Tukhbatullin, A. A., & Khorolzhav, R. (2014).
Sonocrystalloluminescence in suspensions of terbium nitrate. Vestnik Bashkirskogo
Universiteta, 19, 434–437.
68. Wang, H., Zhu, J.-J., Zhu, J. M., Liao, X.-H., Xu, S., Ding, T., et al. (2002). Preparation of
nanocrystalline ceria particles by sonochemical and microwave assisted heating methods.
Physical Chemistry Chemical Physics, 4, 3794–3799.
69. Heller, A., French, K. W., & Haugsjaa, P. O. (1972). Formation of electronically excited ions
in electrode processes: Electroluminescence of trivalent rare earth ions in liquid solutions. The
Journal of Chemical Physics, 56, 2368–2377.
70. Sharipov, G. L., & Kazakov, V. P. (1980). Radioluminescence of aqueous solutions of Tb3þ.
Concentration dependence of radioluminescence intensity. Optika i Spektroskopiya, 48,
69–74.
71. Mirochnik, A. G., Polyakova, N. V., Karasev, V. E., & Kuryavyi, V. G. (2000). Influence of
ligands on thermoluminescence properties of re complexes. Russian Journal of Coordination
Chemistry, 26, 720–723.
72. Elbanowski, M., Staninski, K., & Kaczmarek, M. (1993). Chemiluminescence used in bio-
chemical investigations – An application of the lanthanide ions as a chemiluminescent probe.
Acta Physica Polonica A, 84, 993–1002.
73. Sharipov, G. L., Gaı̆netdinov, R. K., & Abdrakhmanov, A. M. (2003). Sonoluminescence of
aqueous solutions of lanthanide salts. Russian Chemical Bulletin, 52, 1969–1973.
74. Pflieger, R., Schneider, J., Siboulet, B., Mohwald, H., & Nikitenko, S. I. (2013). Luminescence
of trivalent lanthanide ions excited by single-bubble and multibubble cavitations. Journal of
Physical Chemistry B, 117, 2979–2984.
Chapter 10
Triboluminescent Sensors for Polymer-Based
Composites
Kunal Joshi, Margaret Scheiner, David O. Olawale, and Tarik J. Dickens
10.1 Introduction
10.1.1 NDT Overview
Monitoring the structural integrity of advanced materials can give an indication of

life cycle estimation of critical structures in aerospace, civil, and military applica-
tions [1]. Modern detection systems rely on non-destructive assessments that are
associated with high cost and non-value added processes. The major challenges of
using non-destructive evaluation (NDE) techniques in health monitoring of struc-
tures are the service conditions, cost, size, and geometry of a component. Currently,
in aerospace applications, 70–80 % of NDE is performed on the airframe and the
minority inspections are carried out on related components [2].
This chapter is aimed at providing information on the current state of develop-
ment of TL sensor systems for composite and concrete structures, with emphasis on
aerospace and civil applications. Longchambon [3] provided a detailed review on
TL research from early records to 1925 [4]. A review by Walton [4] covered
principal work from approximately 1930–1975. Olawale et al. [5] reviewed
K. Joshi • D.O. Olawale • T.J. Dickens (*)

Florida A&M University, Tallahassee, Florida, USA
FAMU-FSU College of Engineering, High Performance Materials Institute,
Tallahassee, Florida, USA
e-mail: dickens@eng.fsu.edu
M. Scheiner
Florida State University, Tallahassee, Florida, USA
FAMU-FSU College of Engineering, High Performance Materials Institute,
Tallahassee, Florida, USA

306 K. Joshi et al.
mechanism for TL sensors in 2011. Since then, various completed research have
focused on gaining a better understanding of the TL phenomenon and currently
little work has focused on the application for damage sensing. This chapter provides
an overview and reports on the progress of such efforts from 1980 up to the present
date with features of structural health monitoring (SHM).
10.1.2 Motivation: Eight Levels of Health Monitoring
10.1.2.1 Biological Mimicry
Nature holds within itself unique innate solutions for real-world technological
problems. A majority of problems encountered in science and engineering over
the centuries most likely have a preexisting solution or equivalent counterpart in
nature [6]. Imitation of biological systems will aid in achieving SHM capabilities.
By sampling the constructs of nature, borrowing ideas and patterns of function,
novel techniques can be realized. For example, the human nervous system com-
prises various components and inner working that are electrochemical in nature:
primarily the brain (CPU), central vertebrae (temporary CPU), and the vast array of
nerves (detectors). By this unique form of mimicry, a structure from whence to map
out the intricacies of an advanced material nervous system, although not organic
is known as structural health monitoring. Most biological systems behave in this
manner. Biological systems account for varying events such as pressure, damage,
humidity, strain, fatigue, and moisture [7]. Similarly, all of these elements are
essential parameters engineers have deemed significant to diagnosing material
systems.
Figure 10.1 gives an overall view of the progress that has been made in the field
of health monitoring. The graph covers the progress made from 1912 to 2015 and a
prediction of where the field will be by 2020. Various groups have contributed
towards developing and enhancing the capabilities of a health monitoring system
but have mainly achieved health monitoring up to “level 3” (Type of damage).
Fiber-optic sensor technology from 1980 to 1992 has come closest in achieving
level 3, and due to its favorable attributes, fiber-optics are widely used in health
monitoring systems. Fiber Bragg Grating (FBG) is the most popular fiber-optics
technology implemented in composites. Triboluminescent material along with
fiber-optic technology was used by DERA group and the High-Performance Mate-
rials Institute (HPMI) Group at the Florida State University. These groups are
currently able to achieve monitoring up to “level 1” (Damage detection). Other
works from Ihn et al. [8] and Kessler [9] have outlined features of intrinsic sensing
and self-healing capabilities. Although these groups have so far reached only up to
level 1, the use of fiber-optic technology as damage sensors in composites using the
triboluminescence phenomenon is the most promising system to be able to achieve
monitoring up to “level 7–8.”
10 Triboluminescent Sensors for Polymer-Based Composites 307
Fig. 10.1 Progress in health monitoring [10]
10.1.2.2 Damage and Health Monitoring
SHM is an evolving field of technology aimed at interpreting data from sensing

devices to provide critical information concerning the adverse effects of internal or
external damage of base components. SHM strives to enhance reliability and reduce
costs of assessing structural integrity. Current SHM involves the monitoring of
structural integrity over elapsed periods of time where catastrophic and residual
stresses can arise. Health monitoring also relies on historical data to make infer-
ences or critical judgments concerning life cycle assessment and economical sensor
placement. These decisions are made in near real-time and rely heavily on statis-
tical analysis and complex algorithms for detection of out-of-control post-damage
events as well as time-consuming detection instrumentation. A structural review of
this type is known as non-destructive inspections (NDI), non-destructive evaluation
(NDE), or non-destructive testing (NDT). In general, the development of a struc-
tural health monitoring system covers four major areas: integrated sensor networks,
signal generation, signal processing, and signal interpretation [2]. Other critical
components of a complete SHM system include sensor architecture, communica-
tions, computation, power consumption/contribution, and intervention. The main
objectives of SHM for third generation advanced composites that will include
multifunctional aspects are to obtain subsets of the following directives to present
extreme safety measures for critical structures. Dickens [10] modified the listed
levels of health monitoring by Farrar and Worden to include a multifunctional
spectrum [2, 11]. The representative order dictated here is usually associated with a
308 K. Joshi et al.
progressing feature of technology, where the advancement of the previous level

overlaps with similar features.
1. Damage existence (Detection).
2. Location of damage (Localization).
3. Type of damage (Assessment).
4. Extent of damage (Assessment).
5. Diagnosis (Assessment).
6. Prognosis (Prediction).
7. Self-healing/energy harvesting (Regeneration).
8. Self-awareness/autonomy.
The above list is a grading scale for detection technology encompassing modern
and future condition-based or structural health monitoring of critical components.
The distinction between diagnostics and prognostics arises between detection and
damage quantification in service versus upon known damage assessment in order to
ascertain remaining useful service life [12]. Damage assessment of various systems
is unique to the particular structure under observation (i.e., damage to the rudder of
an airplane). Thus, damage can be defined as changes that are inflicted on a system
that affect its continual performance. Damage occurrences in composite systems
are quite different from the modern materials that have been utilized for decades.
Generally, all materials decrease in performance (i.e., mass, strength, stiffness, etc.)
as flaws are introduced, making outfitted structures less reliable over time. For the
branding of composites, issues of nonlinear structural mechanics makes for an even
more complex material failure system. For all systems, it is known that minute
flaws are usually the culprit for gradual damage propagation, which can result in
catastrophic failure. Identifying damage, location, and the severity of those minute
defects is the overall issue and the precursor for detecting and predicting macro-
scopic fatigue or catastrophic events.
Furthermore, systems of SHM can be classified into two categories: passive or
active sensing. For a passive sensing system, only sensors (i.e., transducers, piezo-
electric patches, strain gauge, etc.) are installed in the structures. Sensor measure-
ments are constantly taken in real time and compared with a set of pre-damage
referenced base data. A passive system will estimate the condition of the structure
based on a comparison of past and present information. Hence, the technique of
data comparison for interpretation of structural conditions is crucial for the reli-
ability of the system and most efforts are exhausted in this area computationally.
The system would require a data bank with a history of pre-stored data to store large
data sets for featured comparisons [13]. For an active sensing system, known
external excitation mechanisms are intentionally inputted into the structures
through built-in calibrated devices such as transducers or actuators. Since the inputs
are known, comparable quantities, they can be strongly related to a physical change
in the structural condition such as onset of damage. Active sensing systems are
more technologically compatible with distributed and online monitoring systems
than passive systems, but are not developed enough for continual, large-scale
monitoring.
Table 10.1 Emerging NDE Modern practices Advanced technologies

for composites [14]
Ultrasonic scanning X-ray tomography
Acoustic emission Laser ultrasonic
Tap test Holography
Resonance X-ray Visual optical Laser-optical
Thermal Vibro-thermography
Eddy current Acousto-ultrasonic
D-sight
Interface integrity
Structure problems
(geometry)
10.1.2.3 Non-destructive Evaluation: Modern and Emerging
Current techniques for structural damage monitoring of in-use structures are known
as non-destructive evaluations or non-destructive inspections. The forename
non-destructive implies inspection of structures by evasive means that do not
require the physical testing by caustic assessment. Table 10.1 is a set of classical
and advancing inspection technologies [14]. With modern technologies and several
advancements, NDE methodology has been the standard practice for industry for
some time now. Emerging technologies have been derived from the database of
modern NDE practices. Its segue from metal inspection into composites has been
slow to address specific material differences, but seeks to provide reliable detection
techniques for deducing micro-sized flaws.
Only ultrasonic methods provide the adequate resolution needed to assess
mechanical condition and structural integrity [14]. However, practical implemen-
tation of ultrasonic testing requires point-to-point measurements, although comput-
erized C-scans have made it possible to increase throughput. Structural complexity,
however, presents tremendous operational difficulties. C-scans are sound in finding
impact damage and fiber fatigue, and the elusive delamination defects but lack
prognosis of mechanical state (giving no indication of life cycle state). This is the
popular tool of choice for aerospace industries.
10.1.3 Triboluminescent Overview
Triboluminescence phenomenon was discovered in the sixteenth century by Sir

Francis Bacon and first stated by Curie [15] as a form of luminescence arising from
mechanical action on solids (mechanoluminescence). As a consequence of the
mechanical action, a certain amount of mechanical energy must be converted into
light. Luminescence by deformation and fracture has long been purported to have
unique stress and strain-sensing capabilities. Examples of triboluminescent emis-
sion may be seen in Fig. 10.2 [16].
310 K. Joshi et al.
Fig. 10.2 (a). TL emission from an impact event [93] (b). Example of typical light emission from
highly efficient triboluminescent materials [16]
In the early nineteenth century, research was oriented towards compilation of

triboluminescent materials using subjective visual observation of the TL response
as a function of time and quantity. The discovery of photomultiplier tube (PMT) in
the 1930s and its utility in triboluminescent studies in 1952 introduced a quantita-
tive technique for detecting and comparing TL emissions. This excitation is initi-
ated by mechanical duress in excess of 1 MPa [17, 18]. The excitation of single
crystal by deformation was first studied by Alzetta et al. [19] in 1970. Their
experiments showed that the characteristics light peaks of triboluminescence are
intimately related to the trends occurring in the stress–strain curves. They reported
results for crystals of ZnS:Mn, NaCl, X-ray-irradiated KI and KCl, and sugar.
Triboluminescence can be divided into three types: namely
(1) elasticoluminescence, (2) plasticoluminescence, and (3) fractoluminescence
[20] as shown in Fig. 10.3 [21]. Though there is still debate as to the underlying
mechanisms behind triboluminescence for most TL materials [22–34], known TL
materials can be reliably synthesized. Since the first scientific investigations into
the TL phenomenon, it has been known that there are many factors which affect the
TL yield [35]. For example, the triboluminescence of potassium sulfate crystals
requires some amount of the parent solution to be present: neither the solution nor
the crystals triboluminesce alone.
General crystal synthesis methods are well understood. The steps of a basic
precipitation reaction are as follows [36–38]: parent materials dissolve in solvent,
solution components react, and solid forms through crystal nucleation and
growth. Synthesis conditions from choice of parent materials [39, 40], solvent
[41–44], and dopants [45–48] to timing [49] affect crystal properties like size,
shape particle dispersion, densification, and electrical conductivity, as well as
other properties. Even the level of humidity can prevent triboluminescence of
otherwise-TL materials [35]. Such factors must be kept in mind when synthesiz-
ing and testing crystals.
Light emission
With heating?
yes no
Incandescence Luminescence
From mechanical forces?

yes no
Triboluminesce With excitation by light?
yes no
Immediately? Inanimate?
yes no yes no
Fluorescence Phosphorescence Chemiluminescence Bioluminescence
Fig. 10.3 Overview of different light emission mechanisms [21]
Triboluminescent materials are being investigated to address the gaps in struc-

tural health monitoring. Their ability to emit light under stress or at rupture makes
them a viable option for monitoring structures.
10.1.4 Composite Damage Sensing with Triboluminescence
Although there has been generation of literature concerning TL and its proposed
use, not much experimental work has been done on near application with structural
materials.
In the elastic region, strain increases linearly with the stress, and the TL intensity
also increases linearly with stress. Therefore, Elastico-TL does not require any
plastic deformation or fracturing of the luminescent crystals. The luminescence
produced during the elastic deformation occurs from the solid and heterogeneous
phase of the material constituents. In the plastic region, the strain and TL intensity
increase with stress according to the power law [50]. TL has been reported as low as
0.7 MPa [18]. Plastico-TL can be excited by the mechanical or electrostatic
interaction of dislocations with defect centers; electrification of crystal surfaces
by the movement of charged dislocations; where the successful
mechanoluminescence is directly related to the efficiency of radiative electron–
hole combinations associated with the dislocation bands. This intermediate lumi-
nescent capability extends to pressures above 10 MPa [50]. Larger pressures and
stresses (~100 MPa) that result in fracture and photonic emissions are known as
Fracto-TL, which involve the creation of new surfaces. This has been related to
charging of surfaces at the fracture faces due to processes such as piezoelectrification,
charged dislocations movements, and charged defects [51]. In relation to fracture
312 K. Joshi et al.
mechanics, the speed at which TL is observed can be consistent with the failure rate
of a host material. However, deciphering the regime and mode of failure-to-excitation
is lacking in the literature and reported experimental base.
Efforts in TL damage sensing with composite experimentation were first per-
petuated by Sage and Bourhill of the Defense Evaluation Research Agency (DERA
located in the UK, from which this group has generated numerous works on the
practicality of triboluminescent sensing [16, 52, 53]. Their results were greatly
magnified once suitable materials were realized.
To date they have experimented with menthyl-9-anthracenecarboxylate (MAC)
and a terbium complex [53], through impacting at low-velocities. This group was
also the first to propose and demonstrate effective reception and transmission by use
of optical fibers. Optical fibers were used as a transport medium considering optical
fibers are able to intercept emissions and internally refract rays along the fiber core.
This is thought to be due to side coupling, where the fiber deforms sufficiently to
allow light leakage since attenuation is natural occurring when bended in commu-
nication applications. This however presented unique problems to the myriad of
composite materials, as the integrity of the optical fiber must be preserved during
manufacturing and operation. In addition, if the composite system is not transparent
or near translucent, the TL can be diminished.
Using embedded TL particles as failure indicators, the DERA group [52]
reported improved light transmission of embedded TL crystals emission by the
procurement of polymeric optical fibers. In addition to utilizing solid silica fibers,
hollow optical fibers were doped with a photon-absorbing laser dye (Rhodamine
6G), making light detection possible. This is known as a fluorescent “down con-
version,” which is aimed at efficiently shifting the wavelength of the tribolumines-
cent emissions. In essence, doping the hollow optical fibers (silica capillaries with
photoluminescent (PL) materials will reduce the loss of light created by host
absorption. PL materials will absorb the surrounding ambient light produced by
the TL crystals, thus hindering all other noise influences. The system set up by the
DERA team is composed of bulky liquid-based sensors which undoubtedly lead to
bulky detectors as optical fibers only supply a medium for transport and could pose
a hazard. Hadzic et al. explain that precautions must be exercised with use of optical
fibers embedded in composites [54]. If the density of fibers is too high, it can affect
the desired physical performance of the laminate itself. The results of Hadzic’s
experimentation indicated medium optical fiber embedment will not degrade the
mechanical properties of the host composite, although many researchers have
highlighted the negative effects of embedding bulky materials [54, 55]. Neverthe-
less, the DERA group [53] successfully demonstrated structural damage monitoring
in both glass-fiber and (optically black) carbon-fiber composites through use of
polymeric positioned optical fibers. Upon impact, the onset of structural damage is
indicated by light emission; the severity of damage is indicated by the overall light
intensity of two embedded photoluminescent fibers. The laser dyes have proved
light detection is possible in the commonplace CFRCs which act as an opaque
medium.
Fig. 10.4 Luminescence

intensity response of the
ZnS:Mn film to friction [56]
10.1.4.1 Artificial Skin to Sense Mechanical Stress
An artificial skin was made from a self-diagnosing piezoelectric thin film which can
reproducibly emit strong visible light under stress [56]. They found that the TL
signal generated from the ZnS:Mn crystals was reproducible [56]. This reproduc-
ibility was consistent with compression and impact loads.
Figure 10.4 shows that the response curve of the luminescence intensity of ZnS:
Mn to friction is reproducible. The characteristic TL based of ZnS:Mn gives these
crystal a distinct advantage over other reported piezoluminescent materials like
gamma colored halide crystals and rubber wherein the photons normally decrease
greatly during cyclic stress. Bhargava et al. [57] realized that the TL of ZnS:Mn is
from the Mn emitting center transitioning from 4 T1 to 6 A1.
10.1.4.2 Impact Sensor
Womack et al. [58] impacted ZnS:Mn crystals with a steel ball falling under gravity
in a custom-built impact test rig with increasing impact energy. Impact velocities
greater than 0.5 m/s were found to produce quantifiable TL light. The fluorescent
decay time of ZnS:Mn to detect TL from lab-generated impacts was studied by
Hollerman et al. [59] for meso-velocity impact (100m/s–1km/s). ZnS:Mn was
observed to have a short decay time of about 300 μs. The reduction in light intensity
is a simple decaying exponential for many luminescent materials and is given by the
relation [60]:
t
I ¼ I 0 expf g ð10:1Þ
T
They concluded that TL can be produced during meso-velocity impacts of

around 400 m/s and because TL is sometimes not intense enough to time dependent
sources it might be difficult to separate TL from other light sources, such as debris
impact, sparks, flashes caused by thermal effects or pause changes [59].
314 K. Joshi et al.
Breaux [61] proved that transmission of TL emissions from ZnS:Mn and vinyl
ester resin matrices has great potential to serve as a sensory device in laminated
reinforced structures. They used a plain optical fiber embedded within a vinyl ester
plate to transmit the TL emissions. The biggest impact of their work was the
discovery that the force and energy of impact could be measured by using the
time between impact, and the rebound impact. Micro optical fiber sensors doped
with ZnS:Mn crystals can be used for wireless and distributed sensing within the
composite once fully developed [5]. The ZnS:Mn crystals display superior tribolu-
minescence property in comparison to other triboluminescence materials [56],
which is described in the literature by Sage and Bourhil [16]. Olawale et al. [62]
developed a bio-inspired triboluminescent optical fiber sensor which has an inte-
grated sensing and transmitting system. Long lengths of this sensor have the
potential for real-time monitoring and distributed sensing. Specifically, if embed-
ded in a composite, it holds promise in detecting internal failures in composite
materials like glass fiber-reinforced polymer (GFRP) or carbon fiber-reinforced
polymer (CFRP) [62].
10.1.4.3 TL Material Coated on Optical Fiber
The triboluminescent coating acts similar to the sensory receptor in the human
nervous system [62]. The optical fiber is the transport medium of the TL optical
signals to photodetectors which translates the signals to voltage. The resin used was
UV-curing resin that served as the adhesive between the crystals and the optical
fiber. Olawale et al. [63] tested mortar beams embedded with the optical fiber
sensor in the tension region. The beams were tested under a 4-point bend setup. At
the instance of failure, the integrated sensor generated triboluminescence emissions
that were successfully coupled into the polymer optical fiber and transited to the
photodetector. Thus, TL signals generated by failure and damage such as cracks in
mortar beams were successfully detected as shown in Fig. 10.5 [63]. The in situ
a b
1.0 12.5 1.0
TL signal (V) TL signal (V) 200
Load (kN) Strain (microstrain)
0.8 10.0 0.8
Strain (microstrain)
150
TL signal (V)
TL signal (V)
Load (kN)
0.6 7.5 0.6

100
0.4 5.0 0.4
50
0.2 2.5 0.2
0
0.0 0.0 0.0
0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
Fig. 10.5 Typical distinctive single TL signal at the instant of brittle failure in mortar beams as
indicated by (a) sudden drop in load carrying capacity and (b) sudden jump in strain value [63]
triboluminescent optical fiber (ITOF) sensor shows great promise as a viable

structural health monitoring technique with distributed damage-sensing capability
when fully developed [63]. Research conducted by Sage et al. [52] exploited the
triboluminescent phenomenon as fracture-initiated damage response. They devel-
oped a series of highly efficient triboluminescent materials with sufficient thermal
and chemical properties to allow doping into composites. Triboluminescent sensors
can monitor damage to both metal and polymeric composite structures when
surface mounted. However, composites also allow the opportunity for embedding
the sensor during the manufacturing process.
10.2 Application in Composites:

Triboluminescence-Enhanced Polymer Systems
10.2.1 Fabrication and Integration
Enabling an in situ damage detection scheme using triboluminescent crystals for a

composite can be currently achieved by two ways. One way is to fabricate the
composite with TL materials integrated in the resin matrix to enable ubiquitous
distribution, thus making the composite a self-sensing component. The other way is
to integrate a sensor, typically a fiber-optic sensor, within the composite. These two
methods require two different manufacturing schemes.
10.2.1.1 TL-Doped Composites
The inclusion of differing types of fillers in composites has been studied widely
[64]. While the incorporation of TL materials inside composites has many ben-
efits, there is a cost of inclusion: since TL materials are necessarily weak, their
inclusion weakens the composite. To identify the extent of this parasitic effect a
number of tests can be conducted. The general idea is to measure the material
property of interest as a function of TL material loading (typically considered by
weight percent). A regression analysis of the property versus loading allows one
to create a simple model of the parasitic effect. For example, Dickens investigated
the effect of ZnS:Mn crystal inclusions on the elastic modulus of vinyl ester resin
using three-point bend tests [65]. He found that a loading of 5 wt% ZnS:Mn is
sufficient to generate a visual signal in response to mechanical deformation.
However, as ZnS:Mn loading is increased from 5–25 wt%, TL intensity increases
by a factor of just 1.1 while the composite strength decreases by a factor of 1.4 and
the elastic modulus decreases by a factor of 4.0. The conclusion is that practical
composite design should consider lower TL crystal concentration levels around
5 % or under.
316 K. Joshi et al.
(a) SEM image Energy Dispersive Scanning (EDS)
(b) Silica trace
(c) Zn trace
(d) BW image
Major Assumptions:
•Morphology is a planar
•Traces are particle centers
•If Zn exist, then also there is S and Mn
Fig. 10.6 EDS mapping of elements in SEM micro-graph [66]
The synthesized material form of TL phosphor crystal allow for homogenization

in resin, thus enabling the use of vacuum assisted resin transfer molding (VARTM
processes for the manufacturing of doped laminates. Sage et al. [53] used hand
lay-up procedures to manufacture composites for damage sensing in glass and
carbon FRCs. A critical issue with the manufacturing of composite laminates
doped with TL crystals is the settling of the crystals during resin transfer and
wetting, inhibiting uniform dispersion. Overall, in regard to damage initiation, a
deficiency in dispersion will have an effect on the luminescent properties of the
layered doped composite. For precise and accurate measurement of the TL phe-
nomena, adequate dispersion must be clarified. Dickens and Okoli [66]
manufactured laminates using a vacuum infusion process. A method for mechanical
agitation using a rotational vacuum infusion apparatus was developed employing
centrifugal motion. They determined the degree of dispersion in the resulting
laminates using energy dispersive scanning (EDS) for element mapping shown in
Fig. 10.6. The rotational apparatus prevented the particulate fillers from settling
while the system was still in a pre-gel state of viscous flow. The resulting centripetal
force was seen to improve particulate dispersion in glass fiber-reinforced composite
(GFRC) panels, and prevent onset of massive particle settling. The successful
integration of ZnS:Mn in resin systems during infusion is strongly affected by
resin viscosity, duration of infusion, and cure kinetics [66].
Frketic et al. [67] used a two-phase manufacturing process wherein a thin
plate composed of vinyl ester resin (VER) and ZnS:Mn (18 % by weight) and 2 %
methyl ethyl ketone peroxide (MEKP) as a catalyst was manufactured in a hot press
at 37 C.
Fig. 10.7 C-scan of a pristine sample [67]
This plate is called VER-plate in their work. The second phase consisted of
manufacturing a glass fiber composite plate using the VARTM process by
sandwiching VER-plate between glass fiber fabric. The two phases are shown in
Fig. 10.8.
In addition, to allow for the emissions to be transported to a photodetector, an
optical fiber was secured on top of the TL VER plate. The method introduced by
Frketic et al. [67] can be used to replace the method used by Dickens and Okoli [66]
The integrity of the composite produced by this method was verified using a C scan,
which confirmed a very low to negligible void content. The c-scan is shown in
Fig. 10.7.
10.2.1.2 Integration of Sensors
Sage et al. [53] presented the issues with embedding triboluminescent sensors
within composites, one of which is how to efficiently access the optical signal
generated upon structural damage and how to efficiently guide this optical damage
signal to a remote detector. Their earlier work relied on side coupling of the TL into
a curved silica fiber and/or end coupling into the silica fiber, if the damage to the
host structure had also broken the fiber. However, these methods result in a poor
signal-to-noise (S/N) ratio [53]. In order to overcome this issue, Sage et al. [53]
used polymeric optical fiber and found that the signal-to-noise ratio (S/N) was
318 K. Joshi et al.
Fig. 10.8 The two-phase process (a, b) of sensor mechanism embedment in a composite plate
[67]. (a) TL VER plate and sensory level mechanism (b) Schematic and experimental setup of the
vacuum assisted resin transfer molding: (A) Vacuum bagging layers (B) Pre-infusion vacuum bag
setup. (C) Resin infusion
increased by a large magnitude. The polymeric optical fiber used was procured
from Bicron, with an outside diameter of 250 μm. A glass fiber-reinforced polymer
(GFRP) panel was incorporated with such fibers. The triboluminescent crystals
used for their experiment were menthyl-9-anthracenecarboxylate (MAC), a
a b
40
35
Intensity (arb units)

Triboluminescent
30
25
20
15
10
0
-1000 -500 0 500 1000 1500 2000
Fig. 10.9 (a) GFRP panel incorporating triboluminescent damage sensors and a network of
standard optical fibers; (b) the signal obtained from GFRP panel upon impact [53]
moderately efficient triboluminescent material. The GFRP panel and the signal
obtained upon impact is shown in Fig. 10.9. Plain optical fiber of 1 mm diameter,
manufactured by Mitsubishi Rayon Company, was coated with ZnS:Mn crystals to
form sensors. ZnS:Mn crystals (type GL25/N-U1) from Phosphor Technology, UK,
were added to an epoxy polymer (50 % by resin weight fraction) and the mixture
was dispersed mechanically. Continuous sensors were produced and were allowed
to cure before cutting them to the required length.
GFRP laminates were manufactured using the VaRTM manufacturing process
before which glass fiber fabric was woven with μ-sensors. The μ-sensors were
interwoven typical to a plain weave as shown in Fig. 10.10a. A holder (Fig. 10.10b)
was manufactured of wood to hold the fabric in place to facilitate consistent
weaving process and the same tool was used to hold the μ-sensors in tension during
the manufacturing process. Two μ-sensor-woven glass fiber fabric layers were
placed at the center thickness of the composite (Fig. 10.10c) and were surrounded
by layers of non-woven glass fiber fabric in [0 /90 ] orientations. The two μ-sensors
were woven with a measured distance apart for each sensing layer, by placing the
next sensing layer at a 90 angle to the lower sensing layer to form a square target
area in the composite laminate. As part of a designed experiment, the μ-sensors
were then treated with a conformal coating from Performix to eliminate possible
external refractions. An illustration of the infused panel with the conformal coated
sensors is shown in Fig. 10.10d indicating transmission via side coupling. Josh
et al. [68] provided a system to monitor damage as shown in Figure 10.11 on the
CFRP retrofits through optical fiber sensors which were woven into the composite
structure to provide damage sensing. Fiber-reinforced polymer (FRP) materials as
thin laminates or fabrics are appearing to be an ideal alternative to traditional repair
technology in civil infrastructure, because of their high strength-to-weight ratios
and stiffness-to-weight ratios as well as because of their cost-effectiveness and
durability. Real-time failure detection of FRP strengthened beams was successfully
achieved and the retrofit damage-monitoring scheme aims at providing a tool to
reduce the response time and decision making involved in maintenance of deficient
structures. The method used to manufacture the composite is traditionally called as
320 K. Joshi et al.
Fig. 10.10 (a) A weaved GF Panel (b) Solu-Mold being implemented to fabricate the
composite (c) A schematic of the laminate (4 ply laminate shown) (d) A manufactured glass
fiber composite [70]
“manual lay-up” [69]. Joshi et al. [68] used a modified manual lay-up method in
which the optical fibers were laid between the carbon fiber fabrics before epoxy was
applied for adhesion. The method was shown to produce inconsistency in laminate
integrity as the optical fibers hindered complete adhesion between the carbon fiber
fabrics. Thus, to use the ZnS:Mn coated optical fiber in a manual lay-up method of
laminate manufacturing, innovative techniques of fiber integration need to be
researched.
10.2.2 Structural Characterization (Static and Dynamic

Testing)
Damage identification and characterization is the primary goal of any health

monitoring system. Although the ITOF system is relatively in its nascent phase, it
has been shown by researchers at the FAMU-FSU College of engineering [70, 71]
that when the sensor system is fully developed, it will provide in situ and distributed
health monitoring of large composite structures like aircraft and wind blades as well
as health monitoring in other applications where composite laminates are used, such
as civil infrastructure. Several preliminary studies have indicated TL-composites as
a unique failure indicator.
Fig. 10.11 Reinforced concrete beam strengthening [70] (a) Cross-sectional schematic of
reinforced concrete beam strengthening. (b) RC beam strengthening with μ-sensor
Recently, the team conducted a study using non-reinforced and fiber-reinforced

short beams for initial observations of the TL emissions during three-point bending
shown in Fig. 10.12 [72]. This study looked at the ubiquitous and distributive sensor
(i.e., micro-sensor) cases. The preliminary results indicated that both cases allow
for sensing during applied loading and bending, with appreciable signal response
when multiple damage modes are present. These damage modes are matrix crack-
ing and matrix-fiber debonding. The TL signal obtained was compared to the
stress–strain plot from the bending tests. The TL signal generated was observed
to have a driving frequency based on which a baseline can be generated in order to
eliminate noise. The comparison showed good correlation between the damage
modes and the obtained TL signal [73]. It was concluded that further study is
required to quantify and distinguish the damage modes based on TL signal.
Ongoing research showcases the mechanoluminescence property of ZnS:Mn as a
potential low-cost stress sensor. Initial SEM revealed the presence of microdamage
322 K. Joshi et al.
Fig. 10.12 View of TL response with micro-fracture during loading cycle (incomplete failure) [65]
propagation along the through-thickness direction without total composite failure.

The corresponding multiple TL emissions associated with the microstructural
damage evolution can be seen in Fig. 10.12. It is important to note the external
energies needed to cause damage in polymer composites classically differs due to
the individual constituents. In literature, impacts are mostly based on impacts by a
small ball mill at a few feet on dry powders or resin encapsulated samples [74].
Very few experiments have been conducted on fiber-reinforced composites with
damage sensing [53, 66]. Preliminary results with low-velocity impact testing (e.g.,
3–4 m/s) reveal emissions correlate well with impact area and induced composite
delamination [66]. The emissions show the extent of bruising from initial impact as
seen in Fig. 10.13. Ryu et al. [75] conducted impact tests on polydimethylsiloxane
(PDMS) embedded with EuD4TEA pellets, which is to be applied towards auton-
omous impact monitoring of aerospace structures. The brightness of the emitted
light was captured using a monochrome high speed camera. The impact energies
applied ranged from 7.78 to 62.9 J. The brightness of light was quantified by
counting the number of bright pixels in the captured images. They reported the
emissions were a characteristic of the impact energy and exhibited two distinct
peaks. The obtained peaks were a result of the impact load as well as the PDMS
deformation.
Recently, Leelachao et al. [76] studied the contact stresses associated with
sliding and friction of an impact or on dispersed ZnS-polymer matrix composite.
They identified internal stress controlling factors, namely reduction of contact area
and initial composite mechanical properties, as main contributors to the stress-
induced luminescent behavior. The ZnS:Mn powders were provided by Kojund
Chemical Laboratory, Japan, with a steady distribution of particle sizes around
5 μm. The researchers experimented with volume fractions of 5.57 and 22.8% as
calculated by Hosford [77]. To their novelty, the team experienced an enhancement
Fig. 10.13 Composite layer effect and increase in TL response show bruising of GF plates
in material strength at the higher concentration of particulate inclusion. Mechanical

and optical characterizations were carried out by in-house gaged impactor, with a
spherical optical tip for sampling luminescence. They reported on the depth-
dependent function of sliding-induced or essentially compression of ZnS:Mn
as the main progenerator to initial luminescent emissions observed in Xu
et al. [78, 79], Zhang et al. [80], Someya et al. [81], Chandra and Chandra [82],
Jia et al. [83]. However, even with the nuance of an excitation stimulus the intensity
is proportional to the stress compounded by the stress, similar to Walton [4],
Zink [34]. Therefore, the intensity of the luminescent event can be greatly
enhanced, irrespective of excitation principle, by higher volumetric concentrations.
The pattern of repeated impact strikes from the experiment shows ZnS:Mn is a
durable material in the polymer matrix composite.
Olawale et al. [71] identified that the magnitude of impact and its location in
relation to the embedded sensor determine the intensity of the observed TL signals
This has been found to be consistent with sensors fabricated using 30 and 50 %
crystals by weight, as shown in Fig. 10.14 [71]. The regression coefficient for the
sensor with 30 % crystals by weight and 50 % crystals are 0.816 and 0.870 respec-
tively. The linearity in the sensor response showcases consistency in sensor per-
formance though it can be said that the sensor with 50 % crystals by weight is
relatively more sensitive than the sensor with 30 % crystals by weight. This might
be a result of higher ZnS:Mn content per unit volume around the optical fiber
[71]. The location of impact is one of the critical parameters more than the
magnitude of impact. As explained by Joshi et al. [68] and Olawale et al. [71],
324 K. Joshi et al.
Fig. 10.14 (a) Scatterplot of TL signal and impact load for ITOF sensor with 30 % (wt) ZnS:Mn
content (b) Scatterplot of TL signal and impact load for ITOF sensor with 50 % (wt) ZnS:Mn
content [71]
the farther the location of impact from the sensor, the lower the intensity of TL
signal even if the impact force is similar. The maximum TL signal is 3.75.
AU was detected when the impact event was directly over the sensor location
even though the recorded impact load of 10.14 kN was significantly less than the
14 kN and 17.23 kN recorded for the latter two impact events that occurred at
locations away from the sensor. The corresponding TL signal values were 0.8 and
1.43 AU respectively (Figs. 10.15 and 10.16).
A recent study was conducted by Joshi et al. [68] to identify an optimal sensor
alignment scheme in order to bridge the gap in the study presented by Olawale
et al. [71]. Glass fiber composite (GFC) laminates were impacted using a
low-velocity impact hammer. Three impact loads were applied and the damage
was inspected employing the self-sensing μ-sensors. The sensor response was
verified against the images obtained from the (Ultrasonic) UT scanning system.
Impacts from a height of 121 cm resulted in delamination while impacts from drop
height 76 and 30 cm produced internal cracking. The damage produced by high
impact energies was reported by the μ-sensors and the array system was capable of
representing the quadrant location of the induced damaged. The damage created
with lower impact energy on 6 ply and 8 ply GFC laminates, however, did not reach
the μ-sensors and hence were not reported by the monitoring system. The damage
envelope spread over an area of 3.72 cm2 on the impacted surface and linearly
reduced to 1.11 cm2.
The total volumetric damage on the composite was 0.841 cm3, which indicates a
high tensile resistance of the 6 ply GF composite. From the 12 tested panels,
2 panels demonstrated real-time sensing. The results extrapolated from this study
suggest that at low-velocity impacts, the sensors array area should be restricted to
6.5 cm to achieve indirect damage sensing ability. The segmental and nodal
analysis provides a good representation of the induced damage on a composite
and can be used for real-time preliminary health screening of a composite subjected
to impact. The results obtained reveal that the abilities and limitations of the self-
sensing sensors are as follows:
Fig. 10.15 SEM micrographs of ITOF sensor cross section (a) 30 % (wt) ZnS:Mn dispersed in
polymer matrix of ITOF coating, (b) Closer view showing ZnS:Mn crystals in the polymer matrix,
(c) 50 % (wt) ZnS:Mn dispersed in polymer matrix of ITOF coating, (d) Polymer optical fiber and
triboluminescent composite coating of the ITOF [71]
Fig. 10.16 (a) C-Scan and observed damage of GF4 panel, (b) Segmental analysis result for a
GF4, (c) Junction analysis of a GF4 panel, (d) C-Scan and observed damage envelope of GF6.5 , (e)
Segmental analysis result for a GF6.5, (f) Junction analysis of a GF6.5 panel [70]
326 K. Joshi et al.
• Damage such as a delamination crack under or close to the μ-sensors can be

successfully detected.
• Damage away from the μ-sensors cannot be monitored when a combination of
low-velocity impact and large array areas is implemented.
The intensity of triboluminescent emissions from sensors were log transformed
and averaged over the eight channels (S1 through S8). The damage evidently
propagates to μ-sensor segments S3, S4, S5, S6, S7, and S8 which is well reported
by the segmented analysis. Junction analysis as shown in Fig. 10.16c was
performed by averaging the peak intensities in each quadrant. Junctions J3 and J4
projected peak TL emissions which suggests that the damage propagation is towards
the third and the fourth quadrant. Similar diagnosis was performed on a GF6.5 6 ply
panel induced with low-velocity impact from a drop height of 76 cm. Delamination
failure mode was observed which propagated towards segments S2, S3, S4, and S5
which is well supported by the segmental (Fig. 10.16e) and junction (Fig. 10.16f)
analysis. The sensors were thus able to monitor internal microcracking in a glass fiber
composite. It can be fairly stated that the sensor is capable of recording composite
rupture (perforation failure and delamination or microcracking).
10.3 Conclusion
Multifunctional composite materials for advanced structural systems are gaining

attraction in the defense and private sector for wide-scale adoption [2, 84–87] and
are set to replace conventional materials for the foreseeable future. In order for the
continued adoption, composite structures must be deemed “safe” for critical appli-
cations. This has necessitated need for third-generation composite structures that
comprise smart and intelligent components therein to carry out multiple functions
during processing, fabrication, and in-service. The realization of third generation
composites allow for unique multifunctionality to achieve systemic levels of active
SHM and prognostic health monitoring (PHM) capabilities beyond event detection,
location, and severity [11].
Structures can include vessels in aerospace, marine, machinery, and civil and
energy infrastructure. Industry has relied much on visual identification and testing
inspections that are non-value added, labor intensive and are extremely costly over
the operational life cycle. In addition, mechanical equipment via non-destructive
testing has been utilized to perform offline scans and inspections [88, 89]. Current
techniques involving popular methods like radiography, eddy current, and acoustic
emission sensing enable the detection of cracks; however, they also require large
external power and can be extremely expensive due to the number of sensors
required to monitor an entire structure. In particular, the commercial aerospace
industry spends 27 % of its average life cycle cost on damage detection and
inspection, while the defense industry spends nearly 40%. These scanning systems
are susceptible to misdetection and misdiagnosis. With more and more composite
usage, modern inspection technologies are becoming obsolete in the characteriza-

tion of current damage states. The next generation of safety mechanical systems
requires more sophisticated technologies and methods to interface with accumulat-
ing large datasets and deciphering causations. This is mostly due to the complex
phenomena surrounding failure and degradation of composite structures. Diagnosis
is an estimation of the extent of a problem to the condition of the component-
structure, which can lead to appropriate mitigation. Thus, realizing “four levels” of
health monitoring will pay dividends if achieved practically.
TL-containing composites need to be tougher and stronger, with no parasitic
effect on mechanical properties of the composite, to be used in large structures.
TL-based sensing systems need to have a high-enough signal-to-noise ratio for
precise data interpretation. Sensing distribution will have to be rigorously investi-
gated to ensure sufficient data collection. Methods of data analysis will have to be
developed based on the distribution. For distributed, location-specific sensing,
analysis systems will have to accept and interpret multichannel signals quickly
and precisely.
Very little is stated in this chapter about “level 7” (Self-healing). This is because,
though there has been a significant amount of research into achieving self-healing
and even autonomous self-healing in recent years [90], the self-healing materials
are not yet able to serve as structural materials in commercial applications. For a
material to self-heal, certain material-specific conditions must be met. For example,
coumarin-derived polymers exhibit healing under ultraviolet radiation, but only for
radiation-induced damage [91], and poly(ethylene co-methacrylic acid) exhibits
healing of puncture wounds, but only when the damage generates sufficient heat to
generate localized melting [92], and many more self-healing polymers can only
heal if damage breaks the appropriate chemical bond. For reliable self-healing, the
healing-enabling component is required to be weaker than the rest of the material.
Incorporating weak materials into a composite creates a weaker composite. Struc-
tural capability will be reached once self-healing materials have specific strength
and toughness comparable to today’s structural materials. Self-healing is an excit-
ing prospect, but a greater understanding of the virgin structural properties and
relevant healing efficiency must be achieved before these materials can be
widely used.
References
1. Measures, R. (2001). Structural monitoring with fiber optic technology. San Diego, CA:
Academic.
2. Farrar, C. R., & Worden, K. (2007). An introduction to structural health monitoring. Philo-
sophical Transactions of the Royal Society of London A: Mathematical Physical and Engi-
neering Sciences, 365(1851), 303–315.
3. Longchambon, H. (1925). Recherches expe´rimentales sur les pheńome‘nes de tribo lumi-
nescence et de cristalloluminescence. 2e the‘se. Propositions donneés par la Faculte´. impr.
Bussie‘re, 74, rue Lafayette.
328 K. Joshi et al.
4. Walton, A. J. (1977). Triboluminescence. Advances in Physics, 26(6), 887–948.

131(7), 1407–1418.
6. Bar-Cohen, Y. (2005). Biomimetics: Biologically inspired technologies. Boca Raton, FL: CRC
Press.
7. Greenberg, D., Aminoff, M., & Simon, R. (2002). Clinical neurology. New York: McGraw-
Hill.
8. Ihn, J. B., Chang, F. K., & Speckmann, H. (2001). Built-in diagnostics for monitoring crack
growth in aircraft structures. In Key engineering materials, (vol 204, pp. 299–308). Trans Tech
Publishers.
9. Kessler, M. (2007). Self-healing: a new paradigm in materials design. Proceedings of the
Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 221(4),
479–495.
10. Dickens, T. J. (2013). Assessment of triboluminescent materials for intrinsic health monitoring
of composite damage. Ph.D. thesis, Florida State University.
11. Farrar, C. R., & Lieven, N. A. (2007). Damage prognosis: The future of structural health
monitoring. Philosophical Transactions of the Royal Society of London A: Mathematical,
Physical and Engineering Sciences, 365(1851), 623–632.
12. Lopez, I., & Sarigul-Klijn, N. (2010). A review of uncertainty in flight vehicle structural
damage monitoring, diagnosis and control: Challenges and opportunities. Progress in Aero-
space Sciences, 46(7), 247–273.
13. Hurlebaus, S., & Gaul, L. (2006). Smart structure dynamics. Mechanical Systems and Signal
Processing, 20(2), 255–281.
14. Djordjevic, B. B. (2009). Nondestructive test technology for the composites. In The 10th
International Conference of the Slovenian Society for non-destructive testing (pp. 259–265).
Citeseer.
15. Curie, D. (1963). Luminescence in crystals. ∎: Methuen.
17. Chandra, B., Bagri, A., & Chandra, V. (2010). Mechanoluminescence response to the plastic
flow of coloured alkali halide crystals. Journal of Luminescence, 130(2), 309–314.
18. Kozytskyi, S., Rimashevskyi, O., & Cheska, T. Y. (1995). Ultrasonic investigation of mechan-
ical properties of polycrystalline ZnS-Mn obtained by self propagating high-temperature
synthesis. Materials Science, 30(4), 501–505.
19. Alzetta, G., Chudacek, I., & Scarmozzino R. (1970). Excitation of triboluminescence by
deformation of single crystals. Physica Status Solidi A. 1(4), 775–785.
20. Chandra, B. (1998). Luminescence of solids edited by DR Vij. ∎: ∎.
21. O’Hara, P. B., Peter, W. S., & Engelson, C. (2005). Turning on the light: Lessons from
luminescence. Journal of chemical education, 82(1), 49.
22. Auerbach, F. (1899). Kanon der Physik: di Bergriffe, Principien, Satze, Formeln
Dimensionsformeln und Konstanten der Physik nach dem neuesten Stande der. Veit &
Comp: Wissenschaft systematisch dargestellt.
23. Chandra, B. (1985). Squeezing light out of crystals: triboluminescence. Nuclear Tracks and
Radiation Measurements (1982), 10(1), 225–241.
24. Chandra, B. (1996). Luminescence induced by moving dislocations in crystals. Radiation
Effects and Defects in Solids, 138(1-2), 119–137.
25. Chandra, B., Chandra, V., Jha, P., Pateria, D., & Baghel, R. (2015). Is the fracto
mechanoluminescence of ZnS: Mn phosphor dominated by charged dislocation mechanism
or piezoelectrification mechanism? Luminescence, 31(1), 67–75.
26. Chandra, B., Tiwari, S., Ramrakhiani, M., & Ansari, M. (1991). Mechanoluminescence in
centrosymmetric crystals. Crystal Research and Technology, 26(6), 767–781.
27. Eddingsaas, N. C., & Suslick, K. S. (2006). Mechanoluminescence: Light from sonication of
crystal slurries. Nature, 444(7116), 163–163.
129(21), 6718–6719.
29. Garcia-Guinea, J., & Correcher, V. (2000). Luminescence spectra of alkali feldspars: influence
of crushing on the ultraviolet emission band. Spectroscopy Letters, 33(1), 103–113.
30. Nishida, J. I., Ohura, H., Kita, Y., Hasegawa, H., Kawase, T., Takada, N., et al. (2015).
Phthalimide compounds containing a trifluoromethylphenyl group and electron-donating aryl
groups: color-tuning and enhancement of triboluminescence. The Journal of Organic
Chemistry
31. Tauer, K. J., & Lipscomb, W. N. (1952). On the crystal structures, residual entropy and
dielectric anomaly of methanol. Acta Crystallographica, 5(5), 606–612.
32. Tekalur, S. A. (2010). Triboluminescence in sodium chloride. Journal of Luminescence, 130
(11), 2201–2206.
33. Trout, G. J., Moore, D. E., & Hawke, J. G. (1975). Triboluminescence and associated
decomposition of solid methanol. The Journal of Physical Chemistry, 79(15), 1519–1527.
34. Zink, J. I. (1978). Triboluminescence. Accounts of Chemical Research, 11(8), 289–295.
35. Heinrich, J. P. (1820). Die Phosphorescenz der Ko¨rper nach allen Unsta¨ nden un- tersucht
und erlautert.
36. Chernov, A. A. (2012). Modern crystallography III: Crystal growth (Vol. 36). ∎: Springer.
37. De Yoreo, J. J., & Vekilov, P. G. (2003). Principles of crystal nucleation and growth. Reviews
in Mineralogy and Geochemistry, 54(1), 57–93.
38. Lö¯ffelmann, M., & Mersmann, A. (2002). How to measure supersaturation? Chemical Engi-
neering Science, 57(20), 4301–4310.
39. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2011). Triboluminescent
materials for smart sensors. Materials Today, 14(6), 292–293.
40. Hurt, C., McAvoy, N., Bjorklund, S., & Filipescu, N. (1966). High intensity triboluminescence
in europium tetrakis (dibenzoylmethide)-triethylammonium.
41. Fontenot, R., Bhat, K., Hollerman, W., Aggarwal, M., & Nguyen, K. (2012). Comparison of
the triboluminescent yield and decay time for europium dibenzoyl methide triethylammonium
synthesized using different solvents. Cryst Eng Comm, 14(4), 1382–1386.
42. Li, J.-G., Zhu, Q., Li, X., Sun, X., & Sakka, Y. (2011). Colloidal processing of Gd2O3: Eu3þ
red phosphor monospheres of tunable sizes: Solvent effects on precipitation kinetics and
photoluminescence properties of the oxides. Acta Materialia, 59(9), 3688–3696.
43. Souza, E., & Muccillo, E. (2009). Effect of solvent on physical properties of samaria-doped
ceria prepared by homogeneous precipitation. Journal of Alloys and Compounds, 473(1),
560–566.
44. Xiu, Z., Li, J.-G., Li, X., Huo, D., Sun, X., Ikegami, T., et al. (2008). Nanocrystalline scandia
powders via oxalate precipitation: The effects of solvent and solution pH. Journal of the
American Ceramic Society, 91(2), 603–606.
45. Hollerman, W., Fontenot, R., Bhat, K., Aggarwal, M., Guidry, C., & Nguyen, K. (2012).
Comparison of triboluminescent emission yields for 27 luminescent materials. Optical Mate-
rials, 34(9), 1517–1521.
46. Owens, C., Fontenot, R. S., Bhat, K. N., & Aggarwal, M. D. (2014). Effects of added dopants
on various triboluminescent properties of europium dibenzoylmethide triethylammonium
(EuD4TEA). APS March Meeting Abstracts, 1, 3003.
47. Uthayarani, K., Sankar, R., & Shashidharan Nair, C. (2008). Growth, spectral and thermal
properties of KAP single crystals in the presence of DL-Alanine and L-Methionine amino acid
dopants. Crystal Research and Technology, 43(7), 733–739.
48. Zink, J. I. (1981). Squeezing light out of crystals: Triboluminescence. Naturwis Senschaften,
68(10), 507–512.
330 K. Joshi et al.
49. Scheiner, M., Dickens, T. J., & Okoli, O. (2015b). Synthesis conditions of europium tetrakis
dibenzoylmethide triethylammonium crystals. Crystal Research and Technology.
50. Chandra, B., Xu, C., Yamada, H., & Zheng, X. (2010). Luminescence induced by elastic
deformation of ZnS: Mn nanoparticles. Journal of Luminescence, 130(3), 442–450.
surfaces. Journal of Physics D: Applied Physics, 37(15), 2175.
Triboluminescent damage sensors. Smart Materials and Structures, 8, 504.
black composites: embedded triboluminescent structural damage sensors. Smart Materials and
Structures, 10, 332.
54. Hadzic, R., John, S., & Herszberg, I. (1999). Structural integrity analysis of embedded optical
fibres in composite structures. Composite Structures, 47(1), 759–765.
55. Czarnek, R., Guo, Y., Bennett, K., & Claus, R. (1989). Interferometric measurements of strain
concentrations induced by an optical fiber embedded in a fiber reinforced composite. In
OE/Fiber LASE’88, pp. 43–54. International Society for Optics and Photonics.
56. Xu, C., Watanabe, T., Akiyama, M., & Zheng, X. (1999a). Artificial skin to sense mechanical
stress by visible light emission. Applied Physics Letters, 74(9), 1236–1238.
57. Bhargava, R., Gallagher, D., Hong, X., & Nurmikko, A. (1994). Optical properties of
manganese-doped nanocrystals of ZnS. Physical Review Letters, 72(3), 416.
58. Womack, F. N., Goedeke, S. M., Bergeron, N.P., Hollerman, W., Allison, S.W. et al. (2003).
Measurement of triboluminescence and proton half brightness dose of ZnS:Mn. In Nuclear
Science Symposium Conference Record, 2003 IEEE, (Vol 1, pp. 625–628). IEEE.
59. Hollerman, W., Melespin, C., Fontenot, R., & Wasilewski, P. (2008). Measuring tribolumi-
nescence generated by meso-velocity impacts. In Lunar and Planetary Science Conference,
(Vol 39, pp. 1018)
60. Goedeke, S., Allison, S., Womack, F., Bergeron, & Hollerman, W. (2003).
Tribolumininescence and its application to space-based damage sensors. In Proceedings of
the Propulsion Measurement Sensor Development Workshop, Huntsville, AL.
61. Breaux, J. A. (2011). Manufacturability of triboluminescent composites: Towards a sensory
level component.
62. Olawale, D. O., Kliewer, K., Okoye, A., Dickens, T. J., Uddin, M. J., & Okol, O. I. (2014b).
63. Olawale, D. O., Kliewer, K., Okoye, A., Dickens, T., Uddin, M. J., & Okol, O. I. (2014).
Structural Health Monitoring, 13(2), 177–189.
64. Chamis, C. C. (1974). Analysis of the three-point-bend test for materials with un equal tension
and compression properties. ∎: National Aeronautics and Space Administration.
65. Dickens, T., Armbrister, C., Olawale, D., & Okoli, O. (2015). Characterization of tribolumi-
nescent enhanced discontinuous glass-fiber composite beams for micro-damage detection and
fracture assessment. Journal of Luminescence, 163, 1–7.
66. Dickens, T. J. & Okoli, O. I (2011). Enabling damage detection: manufacturing composite
laminates doped with dispersed triboluminescent materials. Journal of Reinforced Plastics and
Composites, 0731684411413490.
67. Frketic, J. B., Ariza, N., Olawale, D., Okoli, O., Dickens, T. (2015, December). Measurement
of impact force for triboluminescent-enhanced composites by modified impulse method.
Accepted.
68. Joshi, K., Frketic, J., Raley, M., Olawale, D., Dickens, T., & Okoli, O. (2015). Screening
failure detection of structural composite systems: Embedded triboluminescent structronic
wires. Structural Health Monitoring.
69. Mertz, D. & Gillespie J. (1996). Rehabilitation of steel bridge members through the applica-
tion of advanced composites. Final Report. Technical report, NCHRP-93-ID01.
70. Joshi, K., Frketic, J. B., Olawale, D., & Dickens, T. (2015). Damage monitoring of CFRP
retrofits using triboluminescent optical fiber sensors. In SPIE smar structures and materialsþ
nondestructive evaluation and health monitoring (pp. 943520–943520). International Society
for Optics and Photonics.
71. Olawale, D. O., Bhakta, D., Hammel, E., Yan, J., Carey, D., Dickens, T., et al. (2015).
Triboluminescent composite with in-situ impact sensing capability. Dallas, TX: CAMX.
72. Dickens, T., Breaux, J., Olawale, D., Sullivan, W., & Okoli, O. (2012). Effects of ZnS: Mn
concentrated vinyl ester matrices under flexural loading on the triboluminescent yield. Journal
of Luminescence, 132(7), 1714–1719.
73. Joshi, K., Campbell, C., Mishra, S., Vanli, A., Dickens, T. (2016). Correlating acoustic and
triboluminescent emissions with flexural failure modes in ZnS:Mn doped composite. In
Preparation.
74. Hollerman, W., Goedeke, S., Bergeron, N., Moore, R., Allison, S., & Lewis, L. (2005).
Emission spectra from ZnS: Mn due to low velocity impacts. In Optics & Photonics 2005,
(pp. 58970F–58970F). International Society for Optics and Photonics.
75. Ryu, D., Casta~no, N & Vedera, K. (2015). Mechanoluminescent composites towards autono-
mous impact damage detection of aerospace structures. Structural Health Monitoring.
76. Leelachao, S., Muraishi, S., Sannomiya, T., Shi, J., & Nakamura, Y. (2016). Correlation of
triboluminescence and contact stresses in ZnS: Mn/polymeric matrix composite. Journal of
Luminescence, 170, 24–29.
77. Hosford, W. F. (2004). A brick-wall model for calculating Young’s modulus of a particulate
composite. Metallurgical and Materials Transactions A, 35(7), 2191–2192.
78. Xu, C., Ueno, N., Terasaki, N., & Yamada, H. (2012). Mechanoluminescence and novel
structural health diagnosis. Tokyo: NTS Shuppan.
79. Xu, C., Watanabe, T., Akiyama, M., & Zheng, X. (1999b). Preparation and characteristics of
highly triboluminescent ZnS film. Materials Research Bulletin, 34(10), 1491–1500.
80. Zhang, J.-C., Xu, C.-N., Kamimura, S., Terasawa, Y., Yamada, H., & Wang, X. (2013). An
intense elastico-mechanoluminescence material CaZnOS:Mn2þ for sensing and imaging mul-
tiple mechanical stresses. Optics EXPRESS, 21(11), 12976–12986.
81. Someya, S., Ishii, K., Saeki, M., & Munakata, T. (2013). Lifetime-based measurement of stress
using mechanoluminescence of SrAl2O4: Eu2þ. Optics Letters, 38(7), 1095–1097.
82. Chandra, V., & Chandra, B. (2011). Suitable materials for elastico mechanoluminescence-
based stress sensors. Optical Materials, 34(1), 194–200.
83. Jia, Y., Yei, M., & Jia, W. (2006). Stress-induced mechanoluminescence in SrAl2O4: Eu2þ, Dy
3þ
. Optical Materials, 28(8), 974–979.
84. Deodatis, G., Ellingwood, B. R., & Frangopol, D. M. (2013). Safety, reliability, risk and life-
cycle performance of structures and infrastructures. Boca Raton, FL: CRC Press.
85. Dugnani, R. & Malkin, M. (2003). Proceedings of the 4th International Workshop on Struc-
tural Health Monitoring. Stanford University, Stanford, CA (Un published).
86. Gibson, R. F. (2010). A review of recent research on mechanics of multifunctional composite
materials and structures. Composite Structures, 92(12), 2793–2810.
87. Okoli, O., & Smith, G. (1998). Failure modes of fibre reinforced composites: The effects of
strain rate and fibre content. Journal of Materials Science, 33(22), 5415–5422.
88. Okoli, O. & Olawale D. (2012, October 19). Triboluminescent optical fiber sensor US Patent
App. 13/656,246.
89. Okoli, O., Wang, B., & Dickens, T. J (2010, March 5). Systems, methods, and apparatus for
structural health monitoring. US Patent 8,387,469.
90. Scheiner, M., Dickens, T. J., & Okoli, O. (2015a). Progress towards self-healing polymers for
composite structural applications. Polymer.
91. Aida, T., Meijer, E., & Stupp, S. (2012). Functional supramolecular polymers. Science, 335
(6070), 813–817.
332 K. Joshi et al.
92. Yuan, C. E., Rong, M. Z., Zhang, M. Q., Zhang, Z. P., & Yuan, Y. C. (2011). Self healing of
polymers via synchronous covalent bond fission/radical recombination. Chemistry of Mate-
rials, 23(22), 5076–5081.
93. Breaux, J. A., Harmon, N., Joshi, K., & Dickens, T. (2015). Correlation study on impact force
and TL emissions in enhanced glass fiber plates. Technical report. Florida State University
Chapter 11
Detection of Low-Velocity-Impact
Triboluminescent Emissions
Shawn M. Goedeke, William A. Hollerman, Stephen W. Allison,

and Ross S. Fontenot
11.1 Introduction
In 1888, Wiedemann and Schmidt defined triboluminescence (TL) as the emission

of light produced by mechanical action [1–12]. TL is a specific form of mechanolu-
minescence, and is the production of non-thermal, or “cold,” light from any type of
mechanical action or application of stress. Over a century later, the scientific world
has yet to fully understand triboluminescent emission. It is believed that tribolumi-
nescent emission is associated with an asymmetric crystal structure. Crystal bonds
are broken along planes with opposing charge. When they re-connect, light is
emitted as the charges pass through the separation created from the fracture
[1–12]. Other potential sources of TL exist, but they are not well understood.
It is possible to state that abrupt charge separation occurs during tribolumines-
cent emission; however, the exact mechanism is material dependent and not well
understood [1–12]. Triboluminescent emission is observed in approximately 30 %
of known organic and 50 % in inorganic crystals respectively [1–12]. Old-fashioned
Wint-O-Green® Lifesavers are a common example of a triboluminescent material.
If these lifesavers are crushed in the mouth in a darkened room, it is possible
to observe a blue flash of light. The triboluminescent spectrum measured by
S.M. Goedeke • W.A. Hollerman (*)

Department of Physics, University of Louisiana, P.O. Box 44210,
Lafayette, LA 70504, USA
e-mail: wah0320@louisiana.edu
S.W. Allison
Emerging Measurements, Collierville, TN 38017, USA
R.S. Fontenot
Carderock Division, Naval Surface Warfare Center, Code 632,
West Bethesda, MD 20817, USA

334 S.M. Goedeke et al.
Sweeting et al. indicates a 400–500 nm broad band emission from the

methyl salicylate [4, 5]. In addition, nitrogen lines due to sparking occur near
350 nm [4, 5]. It should be noted that modern Wint-O-Green® Lifesavers use
artificial means to flavor the candy. As a result, TL is not observed in today’s
Wint-O-Green® Lifesavers or similar candies.
The term “triboluminescence” comes from the Greek tribein, meaning “to rub,”
and the Latin prefix lumen, meaning “light.” A good history and review of TL
research to 1977 can be found in Walton [1]. Since then, much work has been done
to characterize materials, mechanisms, and possible applications of TL. Of partic-
ular note are the works of Chandra and Zink [2, 3] and Sweeting [4, 5] on TL
characterization and mechanisms. Sage and Bourhill [6] and Xu et al. [7] conducted
research on the use of TL as a stress and impact damage sensor. Recently, Stevens
et al. [8] have used the optical emissions from cerium-doped lutetium orthosilicate
coatings to measure shock wave velocity and to image shock fronts.
Though many types of materials exhibit triboluminescent phenomenon, phos-
phors or the same materials in their crystalline forms have been the primary
triboluminescent materials investigated here [9–12]. This owes in part to the
attention they have already received for temperature sensing applications in turbine
engines where experience has shown they are generally able to endure in turbine
environments.
In 1999, Sage and Geddes patented a design for a sensor capable of discerning
the locations of impacts [13]. To sense damage, triboluminescent materials are
embedded in a composite host structure for a variety of applications. When the
damage/fracture occurs in the host composite structure, it will lead to the fracture of
the triboluminescent crystals resulting in light emission. This will warn in real time
that a structural damage has occurred. The triboluminescent emission of the can-
didate phosphor has to be sufficiently bright, so that the light signal reaching from
the point of fracture to the detector through a fiber optics cable is sufficiently strong
to be detected.
Based on a desire to utilize TL as the active element for impact sensors, the
authors in 2003 started to investigate the emission properties for 7.5 μm ZnS:Mn,
since it emits copious amounts of light at impact and a large quantity was readily
available [14–21]. TL was observed during impacts on ZnS-based materials with
projectile speeds ranging from about 1 m/s to more than 5 km/s [14–21]. Results
from this research showed that TL has great potential to be used as the active
element of an impact sensor.
11.2 Drop Tower Testing Methodology
Figure 11.1 shows a schematic diagram for specially designed drop tower that was
originally constructed for less than $20 in 2003 [20–22]. Figure 11.2 shows a
picture of the drop tower apparatus that was similar to what has been used in
previous research [14–22]. It was made from 0.75 in. (19 mm) medium-density
11 Detection of Low-Velocity-Impact Triboluminescent Emissions 335
Fig. 11.1 Schematic diagram of the current drop tower used to collect triboluminescent emission
yields for selected materials [20–22]
fiberboard, 0.25 in. (6 mm) plexiglass sheet, and 1.25 in. (32 mm) PVC pipe. The
pipe has holes drilled in measured increments of 2 in. (5 cm). These holes are used
to place a pin that holds a 1.25 in. (32 mm) diameter steel ball bearing (130 g), and
serves as the release mechanism. For these measurements, balls were dropped from
a distance of 42 in. (1.1 m).
A small quantity of a sample powder was placed in a small pile on a plexiglass
plate about the center of the tube. After each test, the tube is removed, the drop mass
is cleaned, and the powder is redistributed near the center of the target area. A small
piece of PVC in the tower base is used to form the powder back into an even layer.
To measure the triboluminescent emission yield, a photodiode or other light
detector is placed under the plexiglass plate. A linear amplifier is used to increase
the magnitude of the resulting photodiode signal. An oscilloscope then records the
data in single-sequence mode. A custom LabVIEW code was used to calculate the
relative emission yield by integrating the triboluminescent oscilloscope trace with
time. Differences in sample mass between the organic and inorganic samples were
corrected in LabVIEW. An average triboluminescent yield was calculated from
data collected over five drops at a set height.
An example of the resulting plot of the triboluminescent light yield for a 7.5 μm
ZnS:Mn sample is shown in Fig. 11.3. Notice that significant TL is emitted for
Fig. 11.2 Three pictures of an early drop tower used to collect triboluminescent emission yields
showing: [a] A side view of the apparatus, [b] the ample tray, and [c] detector positioned under the
window [14–22]
Fig. 11.3 Example plot of the triboluminescent yield for 7.5 μm ZnS:Mn powder versus time
[15–20]
Table 11.1 Selected TL intensities from Ref. [7] with added comments
Phosphor TL emission yield Comments
ZnS:Mn 2800 Used for high-speed impact tests
ZnS:Cu 1100 Used for high-speed impact tests
SrAl2O4 36 Similar to many long-lived or persistent phosphors
Y2O3:Eu 20 Useful thermophosphor
ZrO2 8 Component of thermal barrier coatings
about 2 ms after the ball hits the powder. An inset picture of the yellow–orange TL
emission taken at impact is shown in the top right of Fig. 11.3.
The graph in Fig. 11.3 is total light output versus time. The peak signal is 25 V as
measured on the oscilloscope. Judging by the shape of the pulse as it falls, it appears
the signal has exceeded the threshold above which the detector responds
nonlinearly. When the time base is set to longer times, it should be pointed out
that light signal from the subsequent ball bounces can be observed on the
oscilloscope.
11.3 Material Selection Parameters
11.3.1 Emission Yield
There are a variety of triboluminescent materials from which to choose. They may
be divided into two categories: organic and inorganic. The temperature encountered
by fan blades will be at times as high as 200 F. This limit rules out the use of some
TL materials because increasing temperature quenches fluorescence and eventually
destroys the material. It could also cause TL-containing plastic like PMMA or
binder to melt and flow off. Reference [8] describes a methodology for evaluating
the relative triboluminescent efficiency of a number of materials. Table 11.1 lists
some of those materials and the relative triboluminescent emission yields of special
interest. It may be inferred that most photoluminescent materials will produce at
least some TL. The emission characteristics for a triboluminescent material usually
depend on composition, inherent physical properties such as grain size, and dopant
concentration.
11.3.1.1 Resent Research
Since 2003, the authors have concentrated on measuring the triboluminescent

emission yield from inorganic ZnS compounds [14–19]. It was found that the
ZnS compounds doped with Mn, Cu, and Pb had the largest measured tribolumi-
nescent emission yield of inorganic compounds [7, 14–22]. ZnS:Mn has a Wurtzite
(hexagonal diamond) structure that is considered “loose,” meaning that they give
off light with very little stress applied [14–22]. It can be seen emitting tribolumi-
nescent light simply by scratching it with a nail or other sharp object. Materials
usually consist of a semiconductor host and an impurity. For ZnS:Mn, manganese is
the impurity and is called the dopant or activator. The dopant concentration is
usually a small fraction of the composition, changes the band structure of the crystal
by narrowing the energy gap between conduction and valence bands. With
narrower energy gaps, transitions that will emit light are more probable and this
increases the opportunity for light to be emitted during excitation or de-excitation
of electrons. Sometimes luminescent materials contain more than one dopant to
give the desired set of emission characteristics.
11.3.1.2 Micro-Grained ZnS Compouunds
To complete a comparison, sample powders of ZnS:Mn, ZnS:Cu, ZnS:Mn, Cu,

ZnS:Cu,Pb, Mn, and ZnS:Cu,Pb were purchased from Phosphor Technology,
Limited of Great Britain (PT-GB). These microscale grain size samples are com-
mercially available in quantities from a few grams up to several tons. The grain size
and lot number for each sample was recorded for comparison purposes. It should be
noted that ZnS:Mn,Cu was made by PT-GB specifically for this research. Samples
of three other common inorganic luminescent materials (MgF2:Mn, La2O2S:Eu,
and Y2O2S:Eu) were used for comparison purposes and were also obtained from
PT-GB.
11.3.1.3 Nano-Grained ZnS:Mn
In 1994, Bhargava et al. showed that 3 nm nanocrystals of ZnS:Mn have a

photoluminescent quantum efficiency as high as 18 % at room temperature
[22]. Luminescent measurements also showed that the light production efficiency
increases with decreasing particle size [23]. This phenomenon is likely caused by
the fast transfer of the excited electron-hole pairs into the Mn2þ impurity, and a
subsequent efficient and fast radiative recombination of the D-electron
excitation [23].
In 2010, the authors decided to test the idea that light yield might be greater for
nano-grained particles compared to those commercially available micro-grained
materials. Two samples of ~5 nm diameter ZnS:Mn powder was synthesized at the
Center for Nanotechnologies (CINT) in Albuquerque, New Mexico. CINT is a joint
partnership between Sandia National Laboratories and Los Alamos National Lab-
oratory to complete a variety of research projects to grow and characterize nano-
grained materials.
The first 2 g sample of nano-grained ZnS:Mn was synthesized at CINT in
September 2010. Figure 11.4a shows a picture of a 2 g nano-grained sample sitting
on an ultraviolet (UV) illuminator. The resulting yellow light that is visible
Fig. 11.4 Images of the yellow emission from powdered ZnS:Mn samples sitting on a UV
illuminator, including the CINT nano-grained material (a) and three micro-grained materials
labeled (b) through (d), respectively
Fig. 11.5 Close-up image

of the glowing nano-grained
ZnS:Mn powder on a UV
illuminator
from Fig. 11.5a is characteristic of photoluminescent emission from ZnS:Mn [15–

23]. Figure 11.4b, c, d shows UV illumination of jars containing ZnS:Mn powder
with grain sizes of 8.5, 24.1, and 30 μm, respectively. Samples marked (b) through
(d) were obtained from PT-GB. A close-up image of the nano-grained sample under
UV illumination is shown in Fig. 11.5. Notice from simple observation that the UV
illuminated ZnS:Mn nano-grained sample does not emit as much yellow light
compared to the more traditional micro-grained materials. Of course, we were
using a smaller quantity of the nano-grained material due to limited availability.
In addition, about 50 % of the nano-grained ZnS:Mn contains micro-grained
organic ligand clumps that are by-products of the chemical process used to synthe-
size the sample. These ligands appear to not emit fluorescence when excited by UV
radiation. Researchers at CINT estimate that the ZnS:Mn grains in the CINT sample
have a 5 nm particle size [24]. Later analysis showed that the ligands could be
removed by heating and crushing the clumps to about 300 C. A second 2 g batch
containing 25 % ligands was synthesized at CINT in October 2010. About 0.1 g of
the second batch nano-grained ZnS:Mn (with ligand clumps removed) was used in
this research. It should also be noted that the 24.1 μm ZnS:Mn sample shown in
(c) appears to emit the most yellow fluorescence of all the four materials on the UV
illuminator.
11.3.1.4 Europium Tetrakis
In 1966, Hurt et al. synthesized europium tetrakis dibenzoylmethide

triethylammonium (EuD4TEA) [25]. EuD4TEA is an organic material where tribo-
luminescence can be observed in daylight [25]. A small sample of EuD4TEA
synthesized using europium chloride was obtained from Sandia National Labora-
tories (SNL) in California in 2008. In early 2011, a small sample of EuD4TEA was
synthesized using europium nitrate hexahydrate by the authors at Alabama A&M
University (AAMU) in Huntsville, Alabama. In addition, a sample of EuD4TEA
doped with 1.25 mL of dimethyl methylphosphonate (DMMP) and 4 mol% uranyl
acetate were also synthesized at AAMU. EuD4TEA prepared using europium
nitrate was found to produce 82 % more TL than material synthesized using
europium chloride [17]. In addition, the new method produces larger quantities of
EuD4TEA in less time due to the lack of ethanol washing [17]. Detailed information
on synthesizing EuD4TEA can be found in Ref. [17].
11.3.1.5 Triboluminescent Light Yield
The relative triboluminescent emission yield ratios for 28 tested luminescent

materials are shown in Table 11.2. This ratio was calculated for each material
and mass by dividing the average triboluminescent yield by the equivalent value for
the 7.5 μm ZnS:Mn and has been extensively tested since 2003 [13–22].
The EuD4TEA powder synthesized at AAMU was found to emit more than twice
as much TL compared to the sample prepared by Sandia National Laboratories in
California. The average triboluminescent yield from the AAMU synthesized
EuD4TEA was 2.06 times larger than what was measured for the baseline 7.5 μm
ZnS:Mn powder. This relative yield for EuD4TEA is an average calculated over five
drops. Figure 11.6 shows a plot of the triboluminescent yield for the AAMU
synthesized EuD4TEA as a function of the number of drops. TL emitted during
the first drop was more than three times larger than the ZnS:Mn baseline. The
triboluminescent yield decreases by nearly a factor of three from the first to the fifth
drop. For comparison purposes, the TL yield from the baseline ZnS:Mn is reduced
less than 10 % from the first to the fifth drop [26]. Due to its organic nature,
EuD4TEA appears to be more easily damaged than ZnS:Mn.
Table 11.2 shows that the EuD4TEA sample doped with 4 mol% of uranyl
acetate emits about 1.8 times the TL compared to the baseline 7.5 μm ZnS:Mn.
The EuD4TEA doped with DMMP had the largest yield ratio, emitting about 3.2
times more TL compared to the baseline 7.5 μm ZnS:Mn. The 19.8 μm ZnS:Mn
emitted the most TL of any of the tested inorganic materials. Both the ZnS:Mn
formulations with 40 % manganese, were found to emit small quantities of TL. On
average, it appears that ZnS:Mn,Cu has a larger triboluminescent yield compared to
ZnS:Mn.
Table 11.2 Comparison of the triboluminescent yields for the 28 luminescent materials as
measured using the drop tower [21]
Sample Lot Grain size Yield
Base material mass (g) number Manufacturer (μm) ratioa
ZnS:Mn 1 1 CINTb 0.005 0.000
1/0.1 15248 PT-GB 7.5 1.000/
1.000a
1 17112 PT-GB 8.5 1.223
1 19252 PT-GB 8.5 1.092
1 20223 PT-GB 10.5 1.107
1 20056 PT-GB 11.5 0.004
1 20054 PT-GB 16.2 0.127
1 09029 PT-GB 19.8 1.766
1 20131 PT-GB 24.1 1.023
1 19275 PT-GB 30.0 0.982
ZnS:Cu 1 19017 PT-GB 2.9 0.005
1 14159 PT-GB 9.0 0.056
1 19018 PT-GB 30.0 0.019
ZnS:Mn,Cu 1 19010 PT-GB 21.9 1.130
1 19010 PT-GB 22.0 1.519
1 20267 PT-GB 1.585
1 20268 PT-GB 1.337
1 20269 PT-GB 1.038
1 20270 PT-GB 1.496
ZnS:Cu,Pb 1 15027 PT-GB 19.0 0.034
ZnS:Cu,Pb,Mn 1 17002 PT-GB 19.3 1.017
MgF2:Mn 1 09147 PT-GB 0.029
La2O2S:Eu 1 10185 PT-GB 0.004
Y2O2S:Eu 1 19145 PT-GB 0.000
EuD4TEA 0.1 None SNLc 0.960
0.1 3 AAMU 2.063
EuD4TEA þ 1.25 mL 0.1 10 AAMU 3.196
DMMP
EuD4TEA þ 4 mol% Ura- 0.1 75 AAMU 1.796
nyl Acetate
a
Ratio based on the TL light yield for both the 1 g and 0.1 g samples of 7.5 μm ZnS:Mnset equal to
1.000
b
The Center for Integrated Nanotechnologies (CINT) is located in Albuquerque, New Mexico,
USA
c
This Sandia National Laboratories (SNL) branch is located in Livermore, California, USA
No TL was observed when the steel ball bearing was dropped on the 5 nm sized
ZnS:Mn powder synthesized at CINT. Photoluminescence was observed when
these particles were illuminated with an ultraviolet light. TL was observed when
the EuD4TEA powder was scratched using a metal spatula inside its container. The
steel drop ball and plastic strike plate are considered to be “rough” over the
Fig. 11.6 Plot of the triboluminescent light yield for EuD4TEA powder versus the number of
drops from the 130 g ball in the drop tower [21]
nanoscale. The small 5 nm ZnS:Mn particles were likely trapped in surface imper-
fections and were not subject to sufficient force to produce TL. Additional research
should be conducted to quantify these results.
11.3.2 TL Emission Duration
For many luminescent materials, the reduction in light yield from the cessation of
excitation can be written as
n to
I ¼ I 0 exp ; ð11:1Þ
τ
where:
I ¼ Fluorescence light intensity (arbitrary units)
I0 ¼ Initial fluorescence light intensity (arbitrary units)
t ¼ Time since cessation of excitation source (s)
τ ¼ Fluorescence decay time (s)
The time needed to reduce the light intensity to e1 (36.8 %) of its original value
is defined as the fluorescence decay time (τ). Luminescence can be divided into two
groups: fluorescence and phosphorescence. Phosphorescence, also known as
delayed emission, has a much longer decay lifetime than fluorescence. Typical
phosphorescence lifetimes can vary from 100 ms up to many seconds in duration,
while the lifetimes of fluorescent materials usually range from 1 ns to 10 ms. More
specifically, phosphorescence is the emission of light from a triplet excited state,
Table 11.3 Comparison of measured fluorescence decay times for a selection of triboluminescent
materials as measured using the drop tower
Lot Decay time
Material Manufacturer number (μs)
ZnS:Mn Phosphor Technology, Limited of 15248 334.1 22.7
Great Britain
ZnS:Mn,Cu 19010 247.5 14.2
(τ1)
443.7 41.9
(τ2)
ZnS:Cu 14159 266.1 29.8
EuD4TEA [made with Eu 3 544.8 7.5
(NO3)3]
EuD4TEA [made with None 501.8 25.8
EuCl3]
which is where the electron in the ground state and excited state has the same spin
orientation. The triplet state is due to the fact that the spin flips take more time to
complete. Fluorescence occurs in singlet states, with the excited electron having the
opposite spin of the paired ground state, resulting in a short emission time for
photons. The fluorescence decay time is unique to each material. It might be
possible to use the fluorescence decay time as an indicator to gauge the production
of TL from impacts. Some materials like ZnS:Mn,Cu have multiple decay times.
Equation (1) would be valid over a limited range of times.
Table 11.3 shows the measured fluorescence decay times measured by the
authors for a selection of triboluminescent materials [26]. The method used to
calculate the decay times is discussed in detail in Ref. [26]. It should be noted that
ZnS:Mn,Cu was observed to have two decay curves (τ1 and τ2). All of the other
tested materials were found to have only one decay curve. The inorganic ZnS:Mn
powder with a 7.5 μm grain size was found to have an average triboluminescent
decay time of 334 23 μs, which is statistically identical to data collected in earlier
experiments [14–19]. The decay time for ZnS:Cu was statistically the same as was
measured for the first decay curve for ZnS:Mn,Cu. The second decay curve for ZnS:
Mn,Cu was measured to be about 444 42 μs. The decay curves for the organic
EuD4TEA samples were longer than was measured for the inorganic ZnS derivative
materials. Interestingly, changing the europium chloride to europium nitrate source
material during the synthesis process for EuD4TEA increased the decay time to
544.8 8 μs [26]. Additional research is needed to further quantify these results.
In addition to measuring the light yield and decay times, the triboluminescent
emission spectra of ZnS:Mn and EuD4TEA were also measured. ZnS:Mn has a
broad emission peak centered about 590 nm. Depending on how it was synthesized,
the typical peak range for ZnS:Mn is from 585 to 590 nm. EuD4TEA has four
narrow emission peaks centered at 593, 612, 653, and 702 nm. The 612 nm peak
(europium line) has the highest intensity. These important characteristics are unique
to the materials and can be used as a signature for impact sensors.
Fig. 11.7 Three still pictures of the TL emission from ZnS:Mn powder in the drop tower recorded
using a NAC GX-1 high-speed camera at a frame rate of 1,000 Hz [22]
For troubleshooting or demonstration purposes, the detector under the drop plate
(see Fig. 11.1) can be removed and a mirror placed underneath the tower at a 45
angle. This allows a user to observe and record the TL emission with a camera or
other such device. Figure 11.7 shows an example of three ZnS:Mn triboluminescent
images taken from the drop tower using a NAC GX-1 high-speed video camera at a
frame rate of 1,000 Hz. TL from ZnS:Mn was visible in 19 frames of the camera
video. It should be noted that the NAC GX-1 high-speed camera is similar to one
used on the popular Mythbusters® television show. A camera with a frame rate of
about 1000 Hz could be used to image light generated during impacts with a
triboluminescent coating.
11.3.3 TL Emission Wavelength (Color)
Each triboluminescent material has a characteristic emission with a unique spec-

trum. Typically they are in the visible region of the spectrum, though certainly
ultraviolet and infrared triboluminescent materials are possible. If background
sunlight is a problem for the detection system, blue or UV phosphor emission
would present an advantage as background sunlight would be lower in this region of
the spectrum. It is also the case that the emission may be broadband, as the case for
ZnS and the aluminate phosphors or it may be narrowband, as is normal for rare-
earth-doped materials, Y2O3:Eu for example. Narrowband emission is also easier to
distinguish from broadband background sunlight.
11.3.4 Environmental Exposure
The active triboluminescent material will be subjected to extremes in moisture,

temperature, changes in pressure/stress, and wind load during normal operation.
The aluminates are somewhat moisture sensitive but many other materials are not.
Testing of aluminates with polyurethane has shown quite good results. Samples left
on the roof of a building for a significant fraction of a year, exposed to sun and rain,
showed little degradation if any. This should not be an issue for controlled envi-
ronment situations such as engine test cell or wind tunnel applications.
It appears that this possible problem has a simple fix. And, it is easy to test (make
coatings) and expose for six months or longer. Illumination with black light and
comparison with witness samples gives the desired information.
If a moisture resistant or sun /ultraviolet blocking coating are degraded by
foreign object impact, what will happen? It would be expected that moisture
seeping in at the point of the breach would degrade/prevent luminescence in a
circular region about the impact. The diameter of the luminescence spot would
increase with time and exposure to the elements. Illumination with a black light
would reveal the number of impacts. This would therefore function as a secondary
impact detection approach.
11.4 Coating Methodology
There are three main approaches for producing a triboluminescent coating: plasma
spray, sputtering, and painting. Examples of sputter coating are radio frequency and
electron beam sputtering. To paint a surface, the phosphor is mixed with a binder
material and applied with airbrush or paint brush. This latter technique is easiest,
least expensive and it works with any material. The main drawback for this
approach is lack of coating uniformity. It is not known at this point in the devel-
opment of this technique how important uniformity will be. For situations where
one simply counts impact events, this may be the most fitting approach. Further-
more, for convenience, painting of a surface can be performed in-situ.
Plasma spray and sputtering provide the best adherence. And, no binder material
is necessary. But, because of their chemistry, not all materials can be deposited in
this way. For example, oxysulfide phosphor coatings cannot be formed with this
method.
11.5 Coating Properties
How does the coating nature affect its triboluminescent properties? For a painted
coating, the phosphor particles will be dispersed/suspended in another medium.
Will this improve brightness or be deleterious? These are questions that need
additional exploration. A less dense coating could have voids. This could be
advantageous as air may enhance the triboluminescence.
A structured coating that has layers of different materials could expand func-
tionality. In such a case, a bottom layer would require a higher energy threshold in
order to emit light. This could provide for an energy threshold/discrimination
capability. Illumination with a black light could reveal aspects of coating health.
This could function similar to the prospective thermal barrier health monitoring
approaches that are based on luminescence in the thermal barrier coating. A
triboluminescent coating that wears, as evidenced by a decrease in thickness and

luminescence brightness, could be an effective impact dosimeter.
11.6 Signal-Level Estimation
The utility of TL as an impact sensor is at an early stage and more investigations of

several aspects will be necessary. One area of importance concerns whether or not
triboluminescent emission will be sufficiently strong. In order to establish the
viability of the technique, the brightness of the luminescence must be sufficient
for reliable detection of impact events. The following analysis is an initial assess-
ment of detectability based on a simple lab test at low speed impact. The results are
used to scale what might be measured from impact of smaller objects of higher
velocity.
Beginning in 2003, the authors conducted research to induce TL from a
low-energy impact using a drop tower. Again a 130 g ball bearing was dropped
through varying distances on to a small pile of 7.5 μm diameter ZnS:Mn powder. A
series of drops would start with one from the maximum height (42 in. or 1.1 m).
Figure 11.8 shows the measured variation in TL yield characterized by the output
voltage from a photomultiplier detector as a function of the drop kinetic energy
[15, 16]. The triboluminescent emission was viewed from underneath by a
photomultiplier, producing a strong signal as large as 25 V. The measurement
uncertainty is approximately 5 %, as shown by the error bars in Fig. 11.8 The
intensity of the TL appears to be a function of impact energy with two regions of
Fig. 11.8 Plot of triboluminescent yield as a function of impact kinetic energy for the ZnS:Mn
powder [15, 16]
Radius (cm)
0.01 0.10 1.00
104
103
Speed (m/s)
102
101
100
10-5 10-4 10-3 10-2 10-1 100 101 102 103
Mass (g)
Fig. 11.9 Plot of the projectile speed versus its mass and effective steel ball bearing radius for
projectiles with a total kinetic energy of 16 mJ. FOD objects having kinetic energies above this line
would produce detectable TL
interest as shown in Fig. 11.8. The first region is in the energy range below 0.2 J.
Here the production of TL appears to have a threshold of approximately 16 mJ.
Above this threshold, the projectile has sufficient energy (or velocity) to break ZnS:
Mn crystals, producing light. The intensity increases rapidly until about 0.2 J. The
second region begins at 0.2 J and appears to be more like a saturation state: here the
slope is shallow which indicates that the sample is less sensitive to impact energy.
This is almost certainly due to the ball breaking as many crystals as it can in the area
of impact. Increasing the energy above this threshold cannot generate more light
because there are no more crystals in the impact area. For 7.5 μm diameter ZnS:Mn,
TL can be detected from drops with kinetic energies as low as 16 mJ.
It will be assumed for now that the triboluminescent yield will be the same
regardless of particle mass and velocity as long as the total kinetic energy is the
same. Thus, there are an infinite number of combinations of particle mass and
velocity that can have a kinetic energy of 16 mJ and can produce a detectable signal.
Figure 11.9 shows a plot of speed versus mass for projectiles having a total kinetic
energy of 16 mJ. The axis on the top portion of Fig. 11.9 represents the equivalent
radius of a stainless steel bearing to give a kinetic energy of 16 mJ. From above,
research has already shown that kinetic energies of as low as 16 mJ can produce TL
from 7.5 μm diameter ZnS:Mn energy [15, 16]. From Fig. 11.9, a 0.35 mg object
traveling at this speed should, under these assumptions, produce 16 mJ, which is
sufficient to be detected with current electronics. Given that a minimum detectable
signal for this setup was likely about 1 mV, much smaller particles of considerably
less kinetic energy should be detected. This assessment was based on an idealized
laboratory test. In a more realistic situation, there will be other factors that will
affect the signal level and which will give guidance as to other issues that need
investigation. The question of how the signal is affected by coating thickness, a less
efficient triboluminescent material, binder type, binder to phosphor ratio need to be
determined. Also, in an application the distance and viewing angle of a detector
may not be as good.
11.7 Possible Applications
Through the years there has been a lot of research into the materials that produce TL,
mechanisms in these materials that that produce Tl but there have been few applica-
tions presented. To date, Sage and Bourhill [6] and Xuet et al. [7] have attempted to
create stress and impact sensors. Even to the point of sage patenting an active armor
impact sensor. Goedeke et al. expanded this concept to the extreme to determine
impacts on spacecraft [27]. Both of these cases are in the higher velocity regions.
In 2011, Hollerman et al. suggested the use of TL for the detection of foreign
object debris (FOB) inside turbine engines [28]. This was inspired by a NATO
report titled Best Practices for the Mitigation and Control of Foreign Object
Damage-Induced High Cycle Fatigue in Gas Turbine Engine Compression System
Airfoils a need is expressed for determining how turbine blades respond following
FOD impact [29]. The purpose is for defining safe maintenance procedures and for
blade design input. The goal is utilize FOD-tolerant surfaces. In order to follow
through with desire, it is important to have instrumentation that can identify when
FOD occurs, its severity and the nature of the impacting material. A diagnostic
should provide information regarding which types of materials comprise FOD and
their sources. Armed with such knowledge, mitigating steps can be taken.
Figure 11.10 below is an illustration of the basic approach to fielding a system to
detect the debris striking a turbine engine vane or blade. At the point of impact,
emission will be brightest. The light that is produced will emanate from a tribolu-
minescent coating. A detector at a suitable, protected location will capture the light.
The FOD diagnostic approach described here will be most suitable for hard-body
foreign objects. This will include objects such as rocks, small metallic objects, or
runway debris. Reference [29] mentions that there is not one typical kind of debris.
Nonetheless, it is important to have some starting point for obtaining a quantitative
basis for damage research. For example, the report gives results of damage pro-
duced by 1 mm glass beads, each with a mass of about 2 mg used at speeds of
around 300 m/s.
At this point, TL remains a solution without a standard application. As has been
shown in the examples above, the application is only limited by the imagination of
the developer/researcher. Many of the fundamentals are understood as least at a
minimal level. Further refinement and transitioning to application will require
further research.
Camera or other detector
Foreign object debris

strikes blade
Triboluminescence
impact. Brightness correlates within
certain bounds with energy of impact.
Fig. 11.10 Simple illustration of FOD diagnostic approach using TL
11.8 Summary
This chapter presents the fundamentals of triboluminescent detection from

low-speed impacts. A drop tower device is described that can be used for initial
characterization of new materials, using varying impact energies. The flexibility of
this device is seen in the number of different detection methods including light
sensors, spectrometers, and high-speed camera in low light. Additionally, different
materials are evaluated and grain size is evaluated as a possibility for increased TL
yield. Basic parameters in the development of sensors are considered including
coatings, environment, and emission wavelength. Finally, several possible applica-
tions of TL are discussed.
Acknowledgements This work was performed in part by the authors under grants and contracts
from the State of Louisiana, State of Alabama, Department of Defense, National Aeronautics and
Space Administration, Department of Energy, and the National Science Foundation.
References
1. Walton, A. J. (1977). Advances in Physics, 26(6), 887–948.

2. Chandra, B. P., & Zink, J. I. (1980). Physical Review B, 21, 816.
3. Chandra, B. P., & Zink, J. I. (1980). Journal of Chemical Physics, 73(12), 5933–5941.
4. Sweeting, L. (2001). Chemistry of Materials, 13(3), 854–870.
5. Sweeting, L. M., et al. (1992). Molecular Crystals & Liquid Crystals, 211, 389.
6. Sage, I., & Bourhill, G. (2001). Journal of Materials Chemistry, 11, 231–245.
7. Xu, C. N., & Zheng, X. G. (1999). Thin Solid Films, 352, 273–277.
8. Stevens, G. D., Lutz, S. S., Turley, W. D., Adams, C. D., Boat, R. M., & Hull, L. M. (2004).
Review of Scientific Instruments, 75, 462–466.
9. Goedeke, S. M., Allison, S.W., Womack, F.N., Bergeron, N.P., & Hollerman, W.A. (2003).
Tribolumininescence and its application to space-based damage sensors. Proceedings of the
Propulsion Measurement Sensor Development Workshop, Huntsville, Alabama, May 12–15,
2003.
Tribolumininescence and its application to space-based damage sensors. Technical report,
Engineering Science and Technology Division, Oak Ridge Nation Laboratory, 2003.
11. Womack, F., Goedeke, S., Bergeron, N., Hollerman, W., & Allison, S. (2004). IEEE Trans-
12. Hollerman, W. A., Goedeke, S. M., Bergeron, N. P., Muntele, C. I., Allison, S. W., & Ila,
D. (2005). Nuclear Instruments and Methods in Physics Research, B241, 578–582.
13. Sage, I., & Geddes, N. (1999). Triboluminescent damage sensors, U.S. Patent 5,905,260.
14. Hollerman, W. A., Goedeke, S. M., Bergeron, N. P., Moore, R. J., Allison, S. W., & Lewis, L.
A. (2005). Photonics for space environments X (pp. 5897–5915). San Diego, CA: Society
of Photo-Optical Instrumentation Engineers. doi:10.1117/12.613570. Retrieved from http://
proceedings.spiedigitallibrary.org/proceeding.aspx?articleid=868665.
15. Bergeron, N. P., Hollerman, W. A., Goedeke, S. M., Hovater, M., Hubbs, W., Fichum, A.,
et al. (2006). International Journal of Impact Engineering, 33, 91–99.
16. Bergeron, N. P., Hollerman, W. A., Goedeke, S. M., & Moore, R. J. (2008). International
Journal of Impact Engineering, 35, 1587–1592.
17. Fontenot, R. S., Bhat, K. N., Hollerman, W. A., & Aggarwal, M. D. (2011). Materials Today,
14, 292–293.
18. Fontenot, R. S., Hollerman, W. A., & Goedeke, S. M. (2011). Materials Letters, 65,
1108–1110.
19. Fontenot, R. S., & Hollerman, W. A. (2011). Journal of Instrumentation, 6, T04001.
20. Fontenot, R. S., Hollerman, W.A., Bhat, K. N., & Aggarwal, M. D. (2011) Synthesis and
Luminescence (in progress).
21. Hollerman, W. A., Fontenot, R. S., Bhat, K. N., Aggarwal, M.D., Guidry, C. J & Nguyen,
K. M. (2011) Comparison of triboluminescent emission yields for twenty-seven luminescent
materials. Optical Materials (in progress).
22. Fontenot, R. S., Hollerman, W. A., Aggarwal, M. D., Bhat, K. N., & Goedeke, S. M. (2011) A
simple low-cost laboratory apparatus for testing triboluminescent materials. Measurement.
(accepted for publication).
23. Bhargava, R. N., et al. (1994). Physical Review Letters, 72, 416–419.
24. Ivanov, S. A., Center for Integrated Nanotechnologies, Sandia National Laboratories, Albu-
querque, NM 87185 (personal communication, September 2010).
25. Hurt, C. R., McAvoy, N., Bjorklund, S., & Filipescu, N. (1966). Nature, 212, 179–180.
26. Womack, F. A study of the triboluminescent properties of ZnS:Mn using a custom built drop
tower. Master’s Thesis, University of Louisiana at Lafayette, 2004.
Tribolumininescence and its application to space-based damage sensors. Proceedings of the
Propulsion Measurement Sensor Development Workshop, Huntsville, Alabama, May 12–15,
2003.
28. Hollerman, W. A., Fontenot, R. S., Allison, S. W., Cates, M. R., & Goedeke, S. M. (2011).
Detecting blade impacts from FOD using triboluminescence. ∎: NATO.
29. Best practices for the mitigation and control of foreign object damage-induced high cycle
fatigue in gas turbine engine compression system airfoils. RTO Technical Report
TR-AVT-094.
Chapter 12
3D Sensing Using Solid-State Wire-Shaped
Photovoltaic Sensor in TL-Based Structural
Health Monitoring
Jin Yan, M. Jasim Uddin, David O. Olawale, Tarik J. Dickens,

12.1 Introduction
Intrinsic structural health monitoring (iSHM) systems have been studied by

engineers and researchers for years, with designations in various journals, etc.
[1–3]. Current techniques are mostly based on acoustic emissions, thermography,
X-rays, and strain gages. In practice, their application may be laborious, time
consuming and expensive [4, 5]. New research based on the triboluminescent
(TL) phenomenon may provide a relatively efficient and inexpensive way for
ubiquitously monitoring the health of structures in real time [6]. ISHM combines
diverse sensing technologies with an embedded measurement controller to capture,
log, and analyze real-time data. In 2007, Schulz et al. emphasized the problem of
detecting damage in complex geometries and structural joints because damage
often initiates at joints and locations where section properties change [7]. TL is
an optical phenomenon in which light is generated when material is pulled apart,
ripped, scratched, crushed, or rubbed through the breaking of chemical bonds in the
J. Yan
High Performance Materials Institute, Industrial and Manufacturing Engineering, FAMU-FSU
College of Engineering, 2005 Levy Ave., Tallahassee, FL 32310, USA
M.J. Uddin
Department of Chemistry and Environmental Sciences, The University of Texas at
Brownsville, 1 W University Blvd, Brownsville, TX 78520, USA
D.O. Olawale • T.J. Dickens
High Performance Materials Institute, Industrial and Manufacturing Engineering, FAMU-FSU
College of Engineering, 2005 Levy Ave., Tallahassee, FL 32310, USA
Nanotechnology Patronas Group Inc., Tallahassee, FL 32310, USA
O.O.I. Okoli (*)
Nanotechnology Patronas Group Inc., Tallahassee, FL 32310, USA
e-mail: okoli@eng.fsu.edu

352 J. Yan et al.
material [8]. Although there is some controversy, there are two common mecha-
nism that lead to TL, http://scienceworld.wolfram.com/physics/Tribolumines
cence.html. The first results from the storage of electrons which have been
ejected by the penetrating gamma radiation of 40K decay in lattice defects. A
small mechanical shock is then sufficient for these electrons to overcome their
energy barrier and cascade down to ground state. The second type results from the
breaking of certain bonds, and this property is exhibited by materials such as
calcium fluorite (CaF2), sphalerite ((Zn, Fe)S), and wintergreen Life Savers™
[9]. This breakage creates free bonds, which immediately absorbs and ionize
nitrogen from the atmosphere, producing a characteristic green or blue-green
flash. Therefore, a TL material embedded in or attached onto a composite structure
will act as a real-time damage indicator, and therefore a sensor [10]. This can be
addressed through a biomedical inspired approach known as passive structural
neural system (SNS). Figure 12.1a, b illustrates the SNS in both a human body
and the body of an aircraft structure. The sensors would be attached to the inner
surface for continuous health monitoring.
In Fig. 12.1c, wire-shaped PV sensors are weaved into the fiber structure which
forms composite materials. These sensors are connected to form the SNS that is in
turn connected to the computer for data acquisition, storage, and analysis. The SNS
reduces the required number of data acquisition channels necessary to monitor a
Fig. 12.1 Structural neural system (SNS) in both (a) a human body and (b) the body of an aircraft
structure; (c) 3D PV sensors weaved in simulated composite fiber structure
12 3D Sensing Using Solid-State Wire-Shaped Photovoltaic Sensor in TL-Based. . . 353
Fig. 12.2 (a) Basic principle of the PV effect; (b) application in solar cell
structure, which will also allow for different combinations of sensor incorporation
(i.e. TL sensor). After an impact happens, PV sensor (or wire-shaped solar cell) will
receive the light from TL crystals in resin and transmit it as an electrical signal
which can be further analyzed by the processing terminal.
To introduce light sensing with PV sensors systematically, we have to mention
the history and principle of PV. The photovoltaic (PV) effect was first observed by
French scientist Alexandre-Edmond Becquerel in 1839, which is an energy con-
version effect of solar cell to convert light energy to electrical energy (Fig. 12.2a),
with an overall efficiency between 16 and 18 % for a single monocrystalline
photovoltaic cell, http://www.scienzagiovane.unibo.it/english/solar-energy/3-pho
tovoltaic-effect.html. In the case of solar photovoltaic, sunlight is composed of
photons—packets of solar energy. When photons strike a PV cell, they may be
reflected or absorbed, or they may pass right through. And the absorbed photons
generate electricity. After applied light on the photovoltaic device (Fig. 12.2b), we
can detect significant current in the external circuit, because the extra energy of
those excited electrons generates a potential difference, or electromotive force
Jenny Nelson, The Physics of Solar Cells, Imperial College Press, 2003. This
kind of force leads electrons go through a load (bulb in Fig. 12.2b) in the external
circuit for the electrical work. The solar cell is the basic building block of solar
photovoltaic. Here in this chapter, dye-sensitized solar cell (DSSC) is discussed as
PV sensor.
DSSCs’ low cost and environmental sustainability lay a good foundation for
practical applications. Recently, carbon nanotubes (CNTs) have been introduced to
DSSC for their superior advantages, such as light weight, low electrical resistance,
flexibility, excellent electrocatalytic activity, and mechanical integrity. In this
chapter, different types of solar cells using incorporated CNTs as an electrode
material are reviewed. CNTs’ outstanding electrocatalytic functions in enhancing
wire-shaped solar cells’ photo-energy conversion performance are also
emphasized.
354 J. Yan et al.
12.2 Principle of Solar Energy Conversion with Support

of Carbon
DSSCs were first discovered in the late 1960s, and a 12 % energy conversion
efficiency was reached in single platinized FTO (fluorine doped tin oxide) mono-
crystalline PV cells based on a 2D configuration [11–13]. However, constraints
from using FTO glasses in application were found, and the invention of flexible
wire-shaped PV cell greatly enhanced the cells’ energy conversion performance
and decrease costs [14, 15]. Since 2008, wire-shaped PV cells have attracted the
attention of researchers because of their variety of anticipated future technology,
such as being embedded into structures without influencing the mechanical prop-
erties. Liquid electrolytes and glass cladding were widely used to achieve high fill
factors and fast carrier transport. However, replacing liquid electrolyte with solid
ones can benefit the fabrication process and rule out the disadvantages of leakage
and short lifetime [16, 17]. On the other hand, flexibility of cells and the use of solid
electrolyte may also facilitate the development of roll-to-roll (R2R) manufacturing
to realize large-scale production.
The principle of PV sensors shown in Fig. 12.3 is based on DSSCs which can
efficiently convert light to electrical signal. The basic configuration of PEC dye
sensitized nanostructured solar devices consist of semiconductor oxides, assembly
with the WE, a CE, and assembly with a transparent conducting substrate (e.g., FTO
glass, glass cladding) [18–20]. The operating cycle can be summarized in chemical
reaction terminology [21]. Under the light, the incoming photon hits the dye
molecule and is absorbed on the surface of porous nanocrystalline TiO2. This is
Fig. 12.3 (a) P-N junction in solar cell. (b) The working principle of DSSC. (c) Electrons’
transportation routine in dye-sensitized TiO2 electrode [13]
the progress the dye-sensitized TiO2 gain hv, which results in that an electron is
excited from ground state to the higher excited state (S0 ! S*). Then, the excited
electron arrives at the conduction band of TiO2, while dye molecules go to an
oxidized state S+ meantime. After passing through the porous nanocrystalline TiO2
and reaches the TCO layer, this electron will keep going in to the external circuit,
then through an external load to the other TCO layer (right one in Fig. 12.3b), which
makes the first TCO layer the anode and the second the cathode of the current loop.
The electron goes inside of the cell again and transfers to triiodide (I3) in the
electrolyte to yield iodine. The cycle is closed by reduction of the oxidized dye by
the iodine in the electrolyte which is known as redox mediator.
So, the procedure can be simplified as below:
(a) hv + Dye ! Dye*
(b) Dye* + TiO2 ! e(TiO2) + Dye+
(c) e(TiO2) + FTO ! TiO2 + e(TCO) [this electron will go into the TCO and
then through a load]
(d) 3I + 2Dye+ ! I3 + 2Dye [I is in electrolyte]
(e) I3 + 2e(C.E.) ! 3I + (C.E.)
12.2.1 Solar Conversion Efficiency Measurement
Energy conversion efficiency is not defined uniquely, but instead depends on the
usefulness of the output, typically denoted by the Greek letter small Eta (η). The
solar conversion efficiency (η) of a DSC can be estimated using the conversion
efficiency formula:
Pmax
η¼
Pin
where Pmax and Pin are the maximum output and input power, respectively. Since a
DSC usually contains a series resistance and a shunt resistance, the fill factor (FF) is
introduced to count both effects:
Pmax
FF ¼
I sc V oc
where Voc is the open-circuit voltage and Isc is the short-circuit current. Then, the
presentation of solar conversion efficiency of a DSC can be
I sc V oc FF
η¼
Pin
In practical calculation, efficient light area plays an important role in final

performance, since Isc is displayed as milliwatt per centimeter square.
356 J. Yan et al.
I sc ðmA=cm2 Þ V oc ðV Þ FF
η¼ 100 %
Pin ðmW=cm2 Þ
12.3 Flat PV Cell with Support of Carbon
A conventional DSSC comprises a porous nanostructured TiO2 layer in combina-

tion with sensitizer and electrolyte. The light-facing side of the cell is usually
covered with a transparent electrode, called transparent conducting oxide (TCO),
which carries the charge carriers away from the TiO2 and out of the solar cell.
Unfortunately, TCO is brittle and crack easily; moreover, it’s also expensive and
could incur up to 60 % of the total cost of the dye-sensitized solar cell. To decrease
the cost and enlarge surface area, carbon materials were introduced into flat solar
cells. In 2004, single-wall carbon nanotube (SWCNT) film was applied to construct
solar cell with a conversion efficiency of 0.15 % [18]. Kongkanand et al. [19]
reported that the CNTs incorporated with TiO2 structures could play an important
role in transportation of photogenerated electron toward the conductive support
(Fig. 12.4). TiO2 particles were not deposited on carbon fiber electrode (CFE)
directly (Fig. 12.4b); instead, they were modified after conducting SWCNT film
was deposited on CFE (Fig. 12.4d). Figure 12.4e shows the device assembled
SWCNT-TiO2 clearly exhibits approximately two times higher photocurrent den-
sity than that of composed with bare TiO2 of same morphological characteristics.
Also, the large disparity between curves b and d in Fig. 12.4f indicates the
important electrocatalytic function of SWCNT in improving cells’ performance.
With the participation of SWCNT, the incident photon conversion efficiency
(IPCE) was increased from 7.36 to 16 % [19]. This approximately twice increase
indicate that SWCNT acts a valuable role to facilitate charge collection of electrode
and electron transportation in oxides film.
Lee et al. [22] also investigate CNT’s conducting properties and applied multi-
wall carbon nanotube (MWCNT) to speed the transmission of chargers from TiO2
to FTO glass (WE). Cells with different composition of TiO2-CNTs were fabricated
to test the function of CNTs. The SEM image of TiO2-coated MWCNTs shown in
Fig. 12.5a indicates that CNTs enlarge the surface area of oxides sensitizer. In
Fig. 12.5b, J–V curve of cell with TiO2-CNTs (0.1 wt%) has highest short-circuit
current (Isc) and open circuit voltage (Voc). And the conversion efficiency of TiO2-
CNTs (0.1 wt%) contained TiO2 electrodes reached up to 4.97 %, which is about
50 % higher than 0 wt% TiO2-CNTs cell [22]. It is inferred that the TiO2-CNT
hybrid nanostructures exhibit better photo-to-energy conversion efficiency than that
of bare TiO2.
Besides the function as mediator, CNTs were widely used as counter electrode in
solar cells. Han et al. [23] reported a structure of the monolithic DSSC, shown in
Fig. 12.6a. In this approach, one side of TCO glasses has been replaced by
platinized carbon layers. According to the Brunauer-Emmet-Teller theory, the
surface area of platinized carbon layer is 42 m2/g and the unplatinized carbon
Fig. 12.4 SEM image of (a) unmodified carbon fiber (b) carbon fiber modified with TiO2 particles
(c) carbon fiber deposited of SWCNT (d) TiO2 particles deposited on SWCNT film. (e) Photo-
current relies on the amount of TiO2 coated on carbon fiber (blue) or SWCNT film (red) (f) J–V
curve of cells when placing under light or in dark [19]
358 J. Yan et al.
Fig. 12.5 (a) SEM image

of TiO2-coated MWCNTs.
(b) J–V curve of solar cells
with different TiO2-CNTs
wt% [22]. Copyright
belongs to Thin Solid Film
Fig. 12.6 (a) The schema of monolithic DSSC’s structure. (b) SEM image of porous carbon film
[23] (Copyright belongs to American Institute of Physics). (c) Cross-sectional SEM image of CNT
layer (CVD method). (d) Cross-sectional SEM image of CNT paste (screen printing method). (e)
Cross-sectional SEM image of CNT CE (directly formed) [24] (Copyright belongs to Elsevier)
layer can be calculated as 36 m2/g. Compared to flat Pt CE, carbon layers’ larger
surface area allows more contacts to electrolyte, which ends up with the improve-
ment of average charge transfer. The scanning electron microscopy (SEM) image of
its porous structure is shown in Fig. 12.6b, in which the white particles observed in
the image are Pt and ZrO2. Pt has the function to lower carbon’s resistance, while
ZrO2 acts as a binder of the carbon particles and ensures electrolyte’s uniform
infiltration into carbon layer. At last, this cell achieved conversion efficiency up to
3.65 % and after a 3192 h exposure to air, the results still shows a slight increase.
Nam et al. [24] synthesized CNTs by two methods: screen printing and chemical
vapor deposition (CVD) (Fig. 12.6c–e). The conventional DSSC in combination
with CNTs exhibits over 10 % conversion efficiency. Also, CVD-CNTs demon-
strated better performance than that of printed CNTs under the same illumination
condition due to its highly aligned CNTs enlarged the surface area [24].
The conventional rigid DSSCs consist of TCO glass, which has restricted adapt-
ability during transportation, installation and application [15]. That is to say, these
disadvantages hold it back from the applications in SHM system. In order to apply
DSSC in detecting damages in composite materials, 3D wire-like DSSC is needed.
12.4 Invention of 3D Wire-Shaped Flexible DSSC
Before real 3D wire-shaped DSSC coming out, the most influential development is
to convert flat rigid cell to flexible cell [25]. Firstly, lightweight flexible solar cell
has the advantage for transportation and photovoltaic power-supply system
360 J. Yan et al.
Fig. 12.7 Configuration of

a flexible DSSC irradiated
through Pt-counter
electrode (back
illumination) [25]
equipment, such as SHM system. Secondly, replacing rigid substrate (TCO-less)

allows low-cost fabrication. Also, back illumination always plays ~4 % more
efficiency than front-side illumination, which provides more possibilities for fabri-
cating 3D wire-shaped DSSC. Ito et al. [25] reported a new method of fabricating
DSSC with Ti metal foil (0.2 mm thickness) and this method has been proposed in
order to use high-temperature TiO2 sintering (Fig. 12.7), which results in high
photo-conversion efficiency (7.2 %). The flexible PV cells are sensitive to TiO2
thickness. Ito and co-workers also optimized required TiO2 thickness (14–15 μm).
This invention of flexible cell with Ti metal foil provides the baseline for further
Ti wire based 3D DSSC (wire-shaped), which is also ensured by better efficiency
from back illumination. To further develop a 3D DSSC device, weaveable wire-
shaped flexible dye-sensitized solar cells (WSF-DSSCs) can be a good solution,
wherefore Fan et al. [15] reported a new method to produce WSF-DSSCs in 2008.
Both working electrode (WE) and counter electrode (CE) are fiber-shaped wire
since WE is stainless steel fiber coated with TiO2 and CE is conducting fiber such as
Pt wire. The most obvious advantage of WSF-DSSC is that the twisting structure of
two electrodes supports light to reach the WE easily, which also make sure the fine
workmanship and flexibility of cell. In Fig. 12.8, both optical photo and SEM
images are provided to illustrate the twisted structure of WSF-DSSC. This approach
was quickly adopted by other researchers [26–30].
12.5 CNTs Enrich 3D Solar Cells
12.5.1 Photovoltaic Carbon Fiber
In 2011, the initial year of carbon-based wire-shaped DSSC, Hou et al. [31] pro-
posed using low-cost and light-weight carbon fiber (CF) (Toray M40J) to replace Pt
wire as counter electrode. Ti wire was immersed into TiO2 and worked as WE, and
then CFs were wrapped around coated Ti wire. The whole electrodes are flexible
and light can reach the surface of dyed TiO2 nanoparticles through the gaps of CFs.
To finalize the cell, whole electrodes were embedded into a glass capillary with full
of electrolyte. The cyclic voltammetry (CV) study of platinized CF and original CF,
Pt-CF’s waveform shows that Pt-CF has better catalytic activity than that of original
Fig. 12.8 (a) Optical image of twisted WSF-DSSC. (b) SEM image of WSF-DSSC (general
view). (c) SEM image of WSF-DSSC (top view) (d) SEM image of WSF-DSSC (sectional view)
[15]
CF. In further experiments, the “Pt-CF” cell reaches 5.8 % energy conversion
efficiency, which is approximately two time higher than that of original CF cell
(2.7 %) [31].
The same group reported another new method to fabricate wire-shaped carbon-
based DSSC in 2012 [32]. In their research, carbon fiber yarns (Toray M40J) were
still used as CE, and two kinds of CF treatments were proposed. In Fig. 12.9,
bended cell can still retain integrity, and the highly aligned CFs are wrapped with a
thin film of poly(3,4-ethylene dioxythiophene)-polystyrene sulfonate (PEDOT:
PSS). PEDOT:PSS aqueous solution was used as comparative trial for Pt-CF and
original CF cells, and all of them chose TiO2-coated Ti wire as WE. This study
362 J. Yan et al.
Fig. 12.9 (a) Optical image of Ti-PEDOT-CF twisted structure and the inset is the according
SEM image. (b) SEM image of PEDOT-CF electrode. (c) Cross-sectional SEM image of PEDOT-
CF electrode [32]
demonstrated that PEDOT-CF cell performed better than both Pt-CF and original
CF cells due to its multiple core/shell structure. Moreover, PEDOT-CF-composited
CE has enhanced advantages in reducing cost and increasing cell flexibility.
Besides carbon fiber (CF), flexible carbon nanotube (CNT) film is also a
prospective candidate as CEs in solar cell. Despite using polymeric materials or
dye-sensitized oxides with wide band-gap, Zhang et al. [33] chose CdSe as the
photosensing material. First, the cell used a Ti wire as WE, and CdSe was directly
deposed on Ti wire through chemical vapor deposition (CVD) method (Fig. 12.10).
A piece of CNT film was wrapped onto Ti/CdSe composite electrode, and then
electrolyte was infiltrated into the cell to finalize the fabrication progress. At last,
the cell showed conversion efficiencies of 1–2 % with stable perform during
bending and rotation. In the same year of 2012, another Zhang’s group applied
Ti-TiO2-CNT film wire-shaped structure into DSSC and gained up to 2.6 % with the
assist of silver wire [34], shown in Fig. 12.11. In addition, a Ti-TiO2-Pt structure
was also applied and resulted in 2.79 % [34], which is comparable to Liu’s work on
infiltrate
Ti wire grow CdSe wrap CNT film electrolyte
CdSe CNT film electrolyte
Ti
Ti-CdSe Ti-CdSe-CNT Ti-CdSe-CNT-electrolyte
b c
100 mm 10 mm
d e
Ti-CdSe Ti-CdSe-CNT
200 mm 500 nm
Fig. 12.10 (a) Fabrication process for Ti-CdSe-CNT structured solar cell. (b) SEM image of
uniform CdSe layer on the surface of Ti wire. (c) Enlarged SEM image of CdSe nanowires with
length of 15–25 μm. (d) SEM image of CNT film partially covering on Ti-CdSe wire. (e) Enlarged
SEM image of CNT fibers’ porous surface [33]
the similar structure (2.78 %) [35]. However, conversion efficiency of Ti-TiO2-Pt

cell quickly degraded to under 0.5 % when exposed to air, while CNT-based cell
can still remain 1.63 % after 50-min exposure.
364 J. Yan et al.
Fig. 12.11 (a) SEM image of TiO2-coated Ti wire (WE) wrapped with thin CNT film (CE). (b)
Enlarged view of (a). (c) SEM image of CNT network deposited on the surface of WE. (d)
Enlarged view of CNT network. (e) Schema of Ti-TiO2-CNT film-structured DSSC [34]
Fig. 12.12 SEM image of (a) a steel wire coated by a ZnO layer (right) and polymer layers
(P3HT:PCBM and PEDOT) (left). (b) A steel wire is covered by both ZnO and polymer layers
(WE), then CNT film (CE) was wrapped around the wire. (c) Enlarged area of CNT film surface.
(d) The interface of CNT film and polymer layer. (e) Cross-sectional view of a steel wire covered
by polymer layers. (f) Enlarged view of a steel wire covered by polymer layers [36]
12.5.2 Photovoltaic Carbon Nanotube Yarn (CNY) Solar

Cells
Liu et al. [36] compared carbon nanotube yarn (CNY) and CNT film’s functions as
CEs in solar cells. During the synthesis and fabrication, ZnO-coated steel wire
(Fig. 12.12a) was sensitized as WE. A poly(3-hexylthiophene) (P3HT) and [6,6]-
phenyl-C61-butyric acid methyl ester (PCBM) (weight ratio 1:0.8) layer and a
PEDOT:PSS layer (40–60 nm in thickness) were in sequence. CNY and CNT
film were used as two kinds of CEs to fabricate solar cells. The cell with CNY as
CE can reach up to 2.11 % which is comparable to the efficiency (2.31 %) of solar
cell using CNT film. Under the same experimental condition, the cell with silver
wire can only gain 0.8 % and cannot last too long in the air. Moreover, carbon-based
(CNY/CNT film) cells have better properties to bear rotation and bending than
silver wire cell due to its rigidness. The CNT could be a potential electrode material
for solar energy harvesting due to its low cost, lightweight, and high efficiency.
A research focus on the classic Ti-TiO2-Pt/CNY structured DSSC was published
by Zhang et al. [37]. He reported a Pt nano-particle-adsorbed CNY made by
solution adsorption and yarn spinning processes, with uniformly dispersed Pt
nanoparticles through its porous nanotube network [37]. Twisted with a TiO2-
coated Ti wire, the cell was finalized in liquid electrolyte, and SEM images and
J–V curves are shown in Fig. 12.13. The outcome is that Ti-TiO2-Pt/CNY-struc-
tured DSSC can reach up to 4.85 % conversion efficiency, which is comparable to
the efficiency of Ti-TiO2-Pt-structured DSSC (4.23 %). Also, an optimum CNY
diameter of 90 μm has been proved (Fig. 12.13f, g). Thus, CNYs have the function
366 J. Yan et al.
Fig. 12.13 (a) SEM image of CNY with a diameter of 20 μm. (b) Enlarged SEM image of CNY.
(c) SEM image of carbon nanotubes contained in CNYs. (d) SEM image of TiO2-Ti-CNY-
structured DSSC. (e) SEM image of the interface of WE and CE (f). J–V curves for different
CNYs’ diameters. (g) Plots of current densities (J ) and conversion efficiencies (η) relying on
CNY’s diameter [37]
to enhance the interaction between Pt and electrolyte by decreasing the transfer

resistance of electrons.
Some articles demonstrated that liquid electrolyte and glass cladding may
improve the conversion efficiency, but this feature potentially causes the drawback
in transportation and manufacturing of the flexible cells [26, 30, 31, 35, 38, 39]. The
solid-state wire-shaped DSSC can be a prospective solution. The stability and
reproducibility of solid state electrolyte could lead practical advantage during
synthesis and installation when comparing to liquid [40]. A final result reports on
solid-state PEC cell discussed the fabrication of wire-shaped DSSC using metal
wire as WE and displayed energy conversion efficiency up to 0.06 %
[14, 15]. Recently, this efficiency was updated to 0.1959 % by Uddin et al. [40]
using CNYs as CEs. In this article, they applied solid state electrolyte to wire-
shaped solar cell and pointed out that the materials and diameter of CEs play an
important role for the cell performance. The optimization of CE and the TiO2-WE
CE
a
WE d e
Solid Electrolyte CNYs TiO2 film Ti wire
b c Intensity (a.u.) Anatase
1)
2)
200 400 600 800 1000
50 mm
Raman shift (cm-1)
f
200 nm
100 mm 200 nm
Fig. 12.14 (a) Schema of flexible DSSC’s structure. (b) Optical image of wire-shaped DSSC. (c)
SEM image of TiO2-coated Ti wire (WE). (d) Enlarged view of TiO2 layer. (e) SEM image of
CNY (CE). (f) Enlarged view of CNTs composite CNY surface. [40]
interface has been reported in their article. The PEC cell with all solid state
electrolyte showed the alternation percentage of photocurrent was less than 2 %
throughout a large mirror of on-off cycles of 5 s. It is highlighted that the PEC cell
demonstrated the photocurrent signal with remarkable reproducibility without
losing energy conversion efficiency [40]. Therefore, the solid-state PEC cell
could be a prospective device to develop multifunctional composites with inherent
light sensing capability for structural or aerospace safety and e-textile by weaving
the aforesaid wires into reinforced fabric (Fig. 12.14).
12.5.3 Wire-Shaped CNT-CNT Solar Cells
Carbon nanotubes (CNTs) possess great electrical and mechanical properties due to
their high alignment [41]. In addition, because of its semiconducting properties,
CNTs can be also used to replace metal wires and act as WE in DSSC
[42, 43]. Recently, the approach using CNTs (~170 μm) as both WEs and CEs
was proposed by Cai et al. [44]. One CNT fiber coated with dyed nanostructured
molecules (CNT/N719) was used as WE and the other CNT fiber (CNT/PVDF) was
regard as CE. Here, PVDF’s function is to prevent short circuit between two CNTs,
and a perfect PVDF film is synthesized when the concentration of PVDF in
368 J. Yan et al.
Fig. 12.15 SEM images of

a CNT-TiO2-CNT
structured DSSC and
CNT-fibers’ textile. (a and
b) Middle part of solar cell
in 100 μm and 20 μm
magnifications. (c) End part
of solar cell in 40 μm
magnification [45]
Dimethylformamide (DMF) is higher than 10 mg/mL. Using this approach, an

efficiency of 3.9 % was reported in liquid electrolyte.
Another CNT-based DSSC research was reported by Chen et al. [45], in which a
similar twisted structure was applied in this research (Fig. 12.15). TiO2-coated CNT
was used as WE, while a bare CNT fiber served as CE. To verify the good electrical
properties of CNT, a typical cyclic voltammogram test was conducted between
CNT and Pt wire. And the peak current suggested that CNT possess an excellent
electrocatalytic activity, which is because of its larger surface area than that of Pt
wire. At last, the CNT-CNT cell gained 2.60 % under standard illumination
(100 mW/cm2), which was increased to 2.94 % by placing a mirror under the cell.
Also, for a control study, a conventional Pt wire-based cell was tested under the
same condition and only reached an energy conversion efficiency of 0.16 %. This
low efficiency is caused by the less flexibility of Pt wire, and the gaps between WE
and Pt wire was the holdback for better interaction. Uddin et al. [46] reported an
innovative solid state dye sensitized photovoltaic carbon nanotube yarn (CNY),
which exhibits excellent electrical conductivity, thermal stability, and
electrocatalytic property. And the CNY-CNY solid-state device shows a 1.5 %
energy conversion efficiency.
Building off of research on all carbon-based DSSC, Okoli’s group produced a
cell that utilized all-carbon electrodes in solid-state and it yielded an efficiency of
7.39 % [17]. To the best of our knowledge, the reported efficiency is the highest
demonstrated with an all carbon wire-shaped solid-state PV cell. The structure of
this novel cell is shown in Fig. 12.16. Additionally the interfacial charge transfer
efficiency and the electron-hole pair separation efficiency both increased The self-
Fig. 12.16 3D wire-shaped photovoltaic sensor prepared by using a platinized CNY as CE, a
braid of 7-twisted CNYs with hybrid coating as the WE. Details are shown in SEM images and
sketch maps [17]
aligned CNYs (materials for both electrodes) demonstrate a better surface with
interaligned and entangled CNTs than carbon fibers (CFs) and CNT films, which
can enhance the reactions with electrolyte [45, 47–49]. In addition to CNTs,
quantum dots (QD) have recently drawn attention in photoconversion systems
due to their high photon absorption coefficient, tunable band gap, and multiple
exciton generation (MEG) effects [50, 51]. Therefore, in this paper, CNYs were
used as both WEs and CEs, and two kinds of QDs were doped to porous TiO2
structure for multiple exciton generations (MEG).
It has been proven that the usage of hydrofluoric acid (HF) as a morphology
controlling agent can improve the percentage of (001) facets [52]. This theory can
be explained by the force balance of O-O/O-F repulsion and Ti-O/Ti-F attraction,
which could stabilize Ti and O atoms on the surface [52]. Figure 12.3 shows that the
sol-gel method can reach the goal of creating more (001) facets of anatase TiO2
crystals. Figure 12.17a is an image with low-magnification for anatase TiO2
particles, in which we chose three points to analyze the crystal structure. Fig-
ure 12.17b is a simple sketch of a Wulff construction of the equilibrium crystal
shape (tetragonal bipyramidal class) [53]. The bipyramidal structure has exposed
faces that are (101) and are capped with (001) faces. The analysis and
corresponding crystal lattices of other two areas are shown in Fig. 12.17c–f. The
370 J. Yan et al.
Fig. 12.17 (a) TEM image of TiO2 nanoparticles in low magnification. (b) Schematic drawing of
truncated octahedral shape largely dominated with (101) and (001) facets. (c–f) TEM images of
enlarged parts of TiO2 nanoparticles and corresponding lattice pattern [17]
estimated distance between two bars in both area is d ¼ 2.9 0.1 Å which repre-
sents the (001) facets (Fig. 12.17d, f). The more (001) facets in the sensitizer
structure, the better the performance of cells.
Figure 12.18a shows the scheme of the TiO2/CdS/CdSe/N719 hybrid structure
and different particles dispersed well in iodide solid electrolyte. For example, when
the wire-shaped DSSCs receive low light from some special materials in the
structure, small band gaps are needed to facilitate the electron transmission
[13]. In this case, hybrid sensitizers may enhance cell performance to a large extent.
When light hits dye particles on the cell surface, electrons are generated and
transferred to either CdS or CdSe QDs [33, 50, 51]. Then, these electrons will be
taken from QDs to nanoporous TiO2 phase which has a larger band gap (3.2 eV).
Simultaneously, the electrons will possibly be transferred from dye particles to
TiO2 phase directly. These two methods are utilized together to accomplish the
hybrid electron transmission pass. This way, more charges were generated, and the
multiple excitation effect was achieved. In Fig. 12.18b, particular elements are
detected using EDX spectroscopy combined with SEM. Ti element in the graph has
the highest density, while other QD elements are shown much less on the surface.
Figure 12.18c shows the SEM and EDX images for certain hybrid surface, in which
the difference in the density of various elemental dots shows the quantity of them.
However, the QDs (CdSe and CdS) have been dispersed well in the hybrid struc-
ture, which can enhance the performance of cell and help it to reach higher energy
conversion efficiency.
More recently the effects of Ag doping on the physical and chemical properties
of the substrate have been investigated [54–57]. A research conducted by Chao
et al. used TiO2 doped with Ag to improve the layers’ photocatalytic activity
[58]. When doped with about (2–6 mol%) Ag, the TiO2 anatase grain size decreased
and the specific surface area increased. More research with Ag nanowires are also
under preparation, other more research about wire-shaped PV cell can be found in
the review paper of Zou’s group [59].
12.6 CNY Photovoltaic for Engineering Applications
Due to their efficient conversion efficiency and 3D bendable structure, CNT-based

solar cells have been investigated further. Chen et al. have woven their CNT-CNT
PV fibers into textiles via a convenient weaving technology [45]. In Fig. 12.19,
CNT-CNT solar cells were woven into both CNT and aramid fiber textiles. The
PEC yarns showed good obdurability, high integrity, and flexibility. These proper-
ties enable a large amount of applications, such as SHM system and clothing-
integrated photovoltaics [45].
More similar research on CNT-based PV cell weaving process were also
reported recently [59–61]. In 2014, Zhang et al. reported weaving efficient polymer
solar cell wires into flexible power textiles [61], shown in Fig. 12.20. The reported
372 J. Yan et al.
Fig. 12.18 (a) 3D sketch map of hybrid sensitizer. (b) EDS spectrum for hybrid surface. (c) SEM
and EDS mapping images for hybrid surface [17]
Fig. 12.19 SEM images of (a) CNT-based solar cell was woven into a CNT textile (red arrow).
(b) CNT-based solar cell was wove into aramid fibers textile (red arrow) [45]
wire-shaped polymer PV cells that had good performance were developed by

introducing a thin layer of Ti nanoparticles that increased the charge separation
and transportation and an aligned CF electrode that increased their stability. These
wire-shaped PV cells were easily woven into flexible clothes that served as light-
weight and deformable power sources for portable electronic devices [61].
Another important application was mentioned by Olawale et al. [62]. It is
valuable that wire-shaped carbon-based solar cells can be regarded as a structural
health monitoring (SHM) system. When embedded into triboluminesence
Fig. 12.20 Wire-shaped polymer solar cell textile being bent into different structures, which can
form a panel on shirt to charge ipod [61]
multifunctional cementitious composites (TL-based composites), solar cells can

receive light information and react electrical signals to identify structural damages
or defects. Each year, aerospace manufacturing spends about 27 % of its production
cost on structural damage inspection, and nowadays damage detectors require
conventional strain gauges and external power supply. In this case, flexible solar
cell enables spontaneous inspection and may realize wireless detection someday.
12.7 Conclusions and Future Trend
In conclusion, we have shown a novel approach and demonstration of a PV detector

for potential TL transmission. It senses damage by converting the mechanical
energy of the damage causing event like impacts and crack propagation into TL
optical signals. These flexible PV sensors are potentially proposed to be embedded
into advanced composites in association with TL materials for integrated and real-
time SHM. Further research is needed for improving the cell efficiency and the
scope of application, in which longer cell length and cell lifetime would be realized.
374 J. Yan et al.
Acknowledgement This work is supported by the US National Science Foundation through grant
ID CMMI-0969413 and Air Force Research Laboratory.
References
1. Balageas, D., Fritzen, C. P., & Guemes, A. (2010). Structural health monitoring. London:
ISTE.
2. Chang, F.-K. (1999). Structural health monitoring: A Summary Report on the First Interna-
tional Workshop on Structural Health Monitoring. Proceedings of the Second International
Workshop on Structural Health Monitoring, Stanford, CA (pp. xix–xxiv).
3. Lynch, J. P. (2005). Design of a wireless active sensing unit for localized structural health
monitoring. Structural Control & Health Monitoring, 12, 405–423.
4. Farrar, C. R., & Worden, K. (2007). An introduction to structural health monitoring. Philo-
sophical Transactions of the Royal Society A: Mathematical, Physical and Engineering
Sciences, 365, 303–315.
5. Worden, K., Farrar, C. R., Manson, G., & Park, G. (2007). The fundamental axioms of
structural health monitoring. Proceedings of the Royal Society A: Mathematical, Physical
and Engineering Sciences, 463, 1639–1664.
Triboluminescent damage sensors. Smart Materials & Structures, 8, 504–510.
7. Kirikera, G. R., Shinde, V., Schulz, M. J., Ghoshal, A., Sundaresan, M., & Allemang,
R. (2007). Damage localisation in composite and metallic structures using a structural neural
system and simulated acoustic emissions. Mechanical Systems and Signal Processing, 21,
280–297.
9. Chandra, B. P., Chandra, V. K., & Jha, P. (2013). Models for intrinsic and extrinsic fracto-
mechanoluminescence of solids. Journal of Luminescence, 135, 139–153.
10. Aggarwal, M. D., Penn, B. G., Miller, J., Sadate, S., & Batra, A. K. (2008). Triboluminescent
materials for smart optical damage sensors for space applications. Hanover, MD: NASA
CASI.
11. Gerische, H., Michelbe, M., Rebentro, F., & Tributsc, H. (1968). Sensitization of charge
injection into semiconductors with large band gap. Electrochimica Acta, 13, 1509–1515.
12. Oregan, B., & Gratzel, M. (1991). A low-cost, high-efficiency solar-cell based on
dye-sensitized colloidal TiO2 films. Nature, 353, 737–740.
13. Yan, J., Uddin, M. J., Dickens, T. J., & Okoli, O. I. (2013). Carbon nanotubes (CNTs) enrich
the solar cells. Solar Energy, 96, 239–252.
14. Fan, X., Chu, Z. Z., Chen, L., Zhang, C., Wang, F. Z., Tang, Y. W., et al. (2008). Fibrous
flexible solid-type dye-sensitized solar cells without transparent conducting oxide. Applied
Physics Letters, 92, 113510.
15. Fan, X., Chu, Z. Z., Wang, F. Z., Zhang, C., Chen, L., Tang, Y. W., et al. (2008). Wire-shaped
flexible dye-sensitized solar cells. Advanced Materials, 20, 592–595.
16. Uddin, M. J., Daramola, D. E., Velasquez, E., Dickens, T. J., Yan, J., Hammel, E., et al. (2014).
A high efficiency 3D photovoltaic microwire with carbon nanotubes (CNT)-quantum dot
(QD) hybrid interface. Physica Status Solidi (RRL): Rapid Research Letters, 8(11), 898–903.
17. Yan, J., Uddin, M. J., Dickens, T. J., Daramola, D. E., & Okoli, O. I. (2014). 3D wire-shaped
dye-sensitized solar cells in solid state using carbon nanotube yarns with hybrid photovoltaic
structure. Advanced Materials Interfaces. doi:10.1002/admi.201400075.
18. Barazzouk, S., Hotchandani, S., Vinodgopal, K., & Kamat, P. V. (2004). Single-wall carbon
nanotube films for photocurrent generation. A prompt response to visible-light irradiation.
Journal of Physical Chemistry B, 108, 17015–17018.
19. Kongkanand, A., Dominguez, R. M., & Kamat, P. V. (2007). Single wall carbon nanotube
scaffolds for photoelectrochemical solar cells. Capture and transport of photogenerated elec-
trons. Nano Letters, 7, 676–680.
20. Nwanya, A. C., Ezema, F. I., & Ejikeme, P. M. (2011). Dyed sensitized solar cells: A
technically and economically alternative concept to p-n junction photovoltaic devices. Inter-
national Journal of the Physical Sciences, 6, 5190–5201.
21. Matthews, D., Infelta, P., & Gratzel, M. (1996). Calculation of the photocurrent-potential
characteristic for regenerative, sensitized semiconductor electrodes. Solar Energy Materials
and Solar Cells, 44, 119–155.
22. Lee, T. Y., Alegaonkar, P. S., & Yoo, J.-B. (2007). Fabrication of dye sensitized solar cell
using TiO2 coated carbon nanotubes. Thin Solid Films, 515, 5131–5135.
23. Han, H., Bach, U., Cheng, Y.-B., Caruso, R. A., & Macrae, C. (2009). A design for monolithic
all-solid-state dye-sensitized solar cells with a platinized carbon counterelectrode. Applied
Physics Letters, 94, 103102–103102-3.
24. Nam, J. G., Park, Y. J., Kim, B. S., & Lee, J. S. (2010). Enhancement of the efficiency of
dye-sensitized solar cell by utilizing carbon nanotube counter electrode. Scripta Materialia,
62, 148–150.
25. Ito, S., Ha, N. L. C., Rothenberger, G., Liska, P., Comte, P., Zakeeruddin, S. M., et al. (2006).
High-efficiency (7.2%) flexible dye-sensitized solar cells with Ti-metal substrate for nano-
crystalline-TiO2 photoanode. Chemical Communications, 42, 4004–4006.
26. Fu, Y. P., Lv, Z. B., Wu, H. W., Hou, S. C., Cai, X., Wang, D., et al. (2012). Dye-sensitized
solar cell tube. Solar Energy Materials and Solar Cells, 102, 212–219.
27. Huang, S. Q., Guo, X. Z., Huang, X. M., Zhang, Q. X., Sun, H. C., Li, D. M., et al. (2011).
Highly efficient fibrous dye-sensitized solar cells based on TiO2 nanotube arrays. Nanotech-
nology, 22, 315402.
28. Lv, Z. B., Fu, Y. P., Hou, S. C., Wang, D., Wu, H. W., Zhang, C., et al. (2011). Large size, high
efficiency fiber-shaped dye-sensitized solar cells. Physical Chemistry Chemical Physics, 13,
10076–10083.
29. Wang, H., Liu, Y., Li, M., Huang, H., Zhong, M., & Shen, H. (2009). Hydrothermal growth of
large-scale macroporous TiO2 nanowires and its application in 3D dye-sensitized solar cells.
Applied Physics A, 97, 25–29.
30. Weintraub, B., Wei, Y. G., & Wang, Z. L. (2009). Optical fiber/nanowire hybrid structures for
efficient three-dimensional dye-sensitized solar cells. Angewandte Chemie International Edi-
tion, 48, 8981–8985.
31. Hou, S. C., Cai, X., Fu, Y. P., Lv, Z. B., Wang, D., Wu, H. W., et al. (2011). Transparent
conductive oxide-less, flexible, and highly efficient dye-sensitized solar cells with commer-
cialized carbon fiber as the counter electrode. Journal of Materials Chemistry, 21,
13776–13779.
32. Hou, S. C., Cai, X., Wu, H. W., Lv, Z. B., Wang, D., Fu, Y. P., et al. (2012). Flexible, metal-
free composite counter electrodes for efficient fiber-shaped dye-sensitized solar cells. Journal
of Power Sources, 215, 164–169.
33. Zhang, L. H., Shi, E. Z., Li, Z., Li, P. X., Jia, Y., Ji, C. Y., et al. (2012). Wire-supported CdSe
nanowire array photoelectrochemical solar cells. Physical Chemistry Chemical Physics, 14,
3583–3588.
34. Zhang, S., Ji, C. Y., Bian, Z. Q., Liu, R. H., Xia, X. Y., Yun, D. Q., et al. (2011). Single-wire
dye-sensitized solar cells wrapped by carbon nanotube film electrodes. Nano Letters, 11,
3383–3387.
35. Liu, Z. Y., & Misra, M. (2010). Dye-sensitized photovoltaic wires using highly ordered TiO2
nanotube arrays. ACS Nano, 4, 2196–2200.
36. Liu, D. Y., Zhao, M. Y., Li, Y., Bian, Z. Q., Zhang, L. H., Shang, Y. Y., et al. (2012). Solid-
state, polymer-based fiber solar cells with carbon nanotube electrodes. ACS Nano, 6,
11027–11034.
376 J. Yan et al.
37. Zhang, S., Ji, C. Y., Bian, Z. Q., Yu, P. R., Zhang, L. H., Liu, D. Y., et al. (2012). Porous,
platinum nanoparticle-adsorbed carbon nanotube yarns for efficient fiber solar cells. ACS
Nano, 6, 7191–7198.
38. Liu, Y., Li, M., Wang, H., Zheng, J. M., Xu, H. M., Ye, Q. H., et al. (2010). Synthesis of TiO2
nanotube arrays and its application in mini-3D dye-sensitized solar cells. Journal of Physics D:
Applied Physics, 43, 205103.
39. Wang, Y. H., Liu, Y., Yang, H. X., Wang, H., Shen, H., Li, M., et al. (2010). An investigation
of DNA-like structured dye-sensitized solar cells. Current Applied Physics, 10, 119–123.
40. Uddin, M. J., Dickens, T., Yan, J., Chirayath, R., Olawale, D. O., & Okoli, O. I. (2013). Solid
state dye-sensitized photovoltaic micro-wires (DSPMs) with carbon nanotubes yarns as coun-
ter electrode: Synthesis and characterization. Solar Energy Materials and Solar Cells, 108,
65–69.
41. Chen, T., Wang, S., Yang, Z., Feng, Q., Sun, X., Li, L., et al. (2011). Flexible, light-weight,
ultrastrong, and semiconductive carbon nanotube fibers for a highly efficient solar cell.
Angewandte Chemie International Edition, 50, 1815–1819.
42. Peng, H., Jain, M., Peterson, D. E., Zhu, Y., & Jia, Q. (2008). Composite carbon nanotube/
silica fibers with improved mechanical strengths and electrical conductivities. Small, 4,
1964–1967.
43. Peng, H. S., Sun, X. M., Cai, F. J., Chen, X. L., Zhu, Y. C., Liao, G. P., et al. (2009).
Electrochromatic carbon nanotube/polydiacetylene nanocomposite fibres. Nature Nanotech-
nology, 4, 738–741.
44. Cai, F. J., Chen, T., & Peng, H. S. (2012). All carbon nanotube fiber electrode-based
dye-sensitized photovoltaic wire. Journal of Materials Chemistry, 22, 14856–14860.
45. Chen, T., Qiu, L. B., Cai, Z. B., Gong, F., Yang, Z. B., Wang, Z. S., et al. (2012). Intertwined
aligned carbon nanotube fiber based dye-sensitized solar cells. Nano Letters, 12, 2568–2572.
46. Uddin, M. J., Dickens, T. J., Yan, J., Olawale, D. O., Okoli, O. I., & Cesano, F. (2012). Solid-
state dye sensitized optoelectronic carbon naotube-wires: An energy harvesting damage sensor
with nanotechnology approach. Proceedings of the ASME 2012 Stone Mountain, Georgia,
USA (pp. 19–25).
47. Murakami, T. N., & Gratzel, M. (2008). Counter electrodes for DSC: Application of functional
materials as catalysts. Inorganica Chimica Acta, 361, 572–580.
48. Papageorgiou, N. (2004). Counter-electrode function in nanocrystalline photoelectrochemical
cell configurations. Coordination Chemistry Reviews, 248, 1421–1446.
49. Wu, M. X., Lin, X. A., Hagfeldt, A., & Ma, T. L. (2011). A novel catalyst of WO2 nanorod for
the counter electrode of dye-sensitized solar cells. Chemical Communications, 47, 4535–4537.
50. Lee, Y.-L., Chi, C.-F., & Liau, S.-Y. (2009). CdS/CdSe co-sensitized TiO2 photoelectrode for
efficient hydrogen generation in a photoelectrochemical cell. Chemistry of Materials, 22,
922–927.
51. Lee, Y.-L., Huang, B.-M., & Chien, H.-T. (2008). Highly efficient CdSe-sensitized TiO2
photoelectrode for quantum-dot-sensitized solar cell applications. Chemistry of Materials,
20, 6903–6905.
52. Dozzi, M., & Selli, E. (2013). Specific facets-dominated anatase TiO2: Fluorine-mediated
synthesis and photoactivity. Catalysts, 3, 455–485.
53. Yang, H. G., Sun, C. H., Qiao, S. Z., Zou, J., Liu, G., Smith, S. C., et al. (2008). Anatase TiO
(2) single crystals with a large percentage of reactive facets. Nature, 453, 638–U4.
54. Naoi, K., Ohko, Y., & Tatsuma, T. (2004). TiO2 films loaded with silver nanoparticles: Control
of multicolor photochromic behavior. Journal of the American Chemical Society, 126,
3664–3668.
55. Sondergaard, R. R., Espinosa, N., Jorgensen, M., & Krebs, F. C. (2014). Efficient
decommissioning and recycling of polymer solar cells: Justification for use of silver. Energy
& Environmental Science, 7, 1006–1012.
56. Uddin, M. J., Cesano, F., Bertarione, S., Bonino, F., Bordiga, S., Scarano, D., et al. (2008).
Tailoring the activity of Ti-based photocatalysts by playing with surface morphology and
silver doping. Journal of Photochemistry and Photobiology A: Chemistry, 196, 165–173.
57. Zhao, D., Liu, T., Park, J. G., Zhang, M., Chen, J.-M., & Wang, B. (2012). Conductivity
enhancement of aerosol-jet printed electronics by using silver nanoparticles ink with carbon
nanotubes. Microelectronic Engineering, 96, 71–75.
58. Chao, H. E., Yun, Y. U., Xingfang, H. U., & Larbot, A. (2003). Effect of silver doping on the
phase transformation and grain growth of sol-gel titania powder. Journal of the European
59. Peng, M., & Zou, D. (2015). Flexible fiber/wire-shaped solar cells in progress: Properties,
materials, and designs. Journal of Materials Chemistry A, 3, 20435–20458.
60. Yun, M. J., Cha, S. I., Seo, S. H., Kim, H. S., & Lee, D. Y. (2015). Insertion of dye-sensitized
solar cells in textiles using a conventional weaving process. Scientific Reports, 5, 11022.
61. Zhang, Z., Yang, Z., Wu, Z., Guan, G., Pan, S., Zhang, Y., et al. (2014). Weaving efficient
polymer solar cell wires into flexible power textiles. Advanced Energy Materials. doi:10.1002/
aenm.201301750.
62. Olawale, D. O., Dickens, T., Uddin, M. J., & Okoli, O. O. (2012). Triboluminesence
multifunctional cementitious composites with in situ damage sensing capability. Proc. SPIE
8345 (pp. 834538–834538).
Chapter 13
Triboluminescent Sensors for Cement-Based
Composites
David O. Olawale, Jasim M. Uddin, Jin Yan, Tarik J. Dickens,

13.1 Introduction
Cementitious composites like concrete are one of the most widely used materials
for civil infrastructural systems like bridges, tunnels, and dams. Other than earth-
quakes, the most severe damage to large concrete structure arises from the corro-
sion of the embedded steel reinforcements especially due to deicing salt [1, 2]. The
direct cost of metallic corrosion in the USA is estimated at $300 billion per year
[3]. The degradation of concrete structures occurs through the formation and
propagation of cracks [4]. Crack openings above 0.2–0.4 mm (depending on
environmental exposure) may result in durability problems because of steel rein-
forcement corrosion. Larger crack openings (1–2 mm) that may be caused by
excessive loading during natural hazards like earthquake are signs of severe
damage that may require immediate closing of the facility [4]. Hence, monitoring
cracks in concrete is an effective way to assess the structural condition or “health”
of a concrete structure [5, 6].
D.O. Olawale (*)

Nanotechnology Patronas Group Inc., 914 Railroad Ave., Tallahassee, FL 32310, USA
e-mail: doo07@my.fsu.edu
J.M. Uddin
Department of Chemistry and Environmental Sciences, The University of Texas
at Brownsville, 1 W University Blvd, Brownsville, TX 78520, USA
J. Yan • T.J. Dickens • O.O.I. Okoli

13.1.1 In Situ Damage-Sensing System
There are currently a number of techniques being used for damage detection and
monitoring of cementitious composites like concrete structures [2, 7]. These
include imaging techniques such as ultrasonic C-scan, X-ray, and thermography.
The major drawback of these techniques is that they do not provide for in situ
sensing. This inhibits real-time monitoring of the health of the structure. Further-
more, the associated costs resulting from the downtime required for periodic
non-destructive inspections can be very high for civil structures like bridges
[8]. The acoustic emission technique is promising, but it suffers from low signal-
to-noise ratio. Surface-mounted resistive foil strain gages have potential for in situ
and continuous monitoring. They are however less effective in monitoring internal
damage and are vulnerable to electromagnetic and electrical interference, as well as
physical damage [7].
Triboluminescence-based sensor systems however have the potential to enable
in-situ and distributed structural health monitoring of composite structures [9]. The
chapter describes some progress made in the development and application of
triboluminescence-based sensor systems for cementitious composites (CC).
13.2 Crack Propagation in Cement-Based Composites

and the Monitoring Challenges
13.2.1 Crack Initiation and Propagation in Concrete
Visible cracking occurs in concrete when the tensile stresses developed in the
concrete system exceed the relatively low tensile strength of concrete [10]. This
can be as a result of any or a combination of factors such as volume changes due to
temperature and humidity gradient or crystallization pressure of salts in pores;
exposure to temperature extremes; structural loading; and chemical reactions
involving formation of expansive products such as in corrosion of the steel rebar
[11–13]. These are illustrated in Fig. 13.1. The presence of small cracks provides
easy access for moisture and chemicals, thereby accelerating concrete deterioration
and structural failure.
During contraction, a concrete component within a structure will be subjected to
some restraints from either the foundation, other part of the structure, embedded
reinforcing steel, or nonuniform shrinkage of the concrete [14]. These cause tensile
stresses in the concrete that results in to cracking because of the inherent low tensile
strength of concrete (Fig. 13.2). A typical value for the final shrinkage strain of
concrete in structures is 600 106 while the concrete tensile-strain capacity can
be 150 106 or less [14]. Likewise, the freezing of the infiltrated water to ice and
the accompanying expansion causes deterioration either of the hardened paste,
13 Triboluminescent Sensors for Cement-Based Composites 381
Cracking
Exposure to Chemical reactions

Volume changes Structural loading temperature with expansive
extremes products
Normal temperature
Overloading and Freeze-thaw Corrosion of steel
and humidity
impact action rebar
gradients
Crystallization
pressure of salts Cyclic loading Fire Sulfate attack
in pores
Alkali-aggregate
reaction
Fig. 13.1 Physical and chemical causes of cracking in concrete [11]
Fig. 13.2 Cracking of

concrete due to drying
shrinkage [14] ORIGINAL LENGTH
UNRESTRAINED
SHRINKAGE
RESTRAINED SHRINKAGE
DEVELOPS TENSILE STRESS
IF TENSILE STRESS IS
GREATER THAN TENSILE
STRENGTH, CONCRETE CRACKS
aggregate, or both due to the hydraulic and osmotic pressure developed in the pores
as the water freezes and expands [10].
The corrosion of steel in concrete is an electrochemical process [12]. It requires
the presence of moisture, oxygen, and an electrolyte to occur and is accelerated by
Table 13.1 Guide to reasonablea crack for reinforced concrete under service loads [14]
Crack width
Exposure condition In. mm
Dry air or protective membrane 0.016 0.41
Humidity, moist air, soil 0.012 0.30
Deicing chemicals 0.007 0.18
Seawater and seawater spray, wetting, and drying 0.006 0.15
Water-retaining structuresb 0.004 0.10
a
It should be expected that a portion of the cracks in the structure will exceed these values. With
time, a significant portion can exceed these values. These are general guidelines for design to be
used in conjunction with sound engineering judgment
b
Excluding non-pressure pipes
wetting and drying cycles. The chemical reaction at the anode and cathode are as
follows [12]:
Anode: Fe ! 2e þ Fe2þ
Cathode:
1=
2 O2 þ H2 O þ 2e

! 2ðOHÞ Fe2þ þ 2ðOHÞ ! FeO ðH2 OÞnx ½rust
The resulting rust occupies a larger volume than the iron from which it was
formed may result in a volume increase up to 600 % of the original metal [12]. This
volume increase results in internal stresses that cause cracking of the concrete and
debonding between the steel rebar and the concrete [10–12]. Preexisting cracks and
sustained loads can accelerate corrosion initiation and propagation, and can lead to
creep [15, 16]. The localized corrosion at the cracked areas lead to further longi-
tudinal surface cracking, delamination, and debonding, ultimately resulting in a
reduction in the strength capacity and stiffness of the structure [10, 17]. Table 13.1
provides recommended acceptable crack sizes for different scenarios.
13.2.2 Challenge of Crack Monitoring in Cement-Based

Composites
Structural health monitoring (SHM) is the process of implementing a damage

identification strategy for aerospace, civil and mechanical engineering infrastruc-
ture [18]. It involves the observation of a structure or mechanical system overtime
using periodically spaced measurements, the extraction of damage-sensitive fea-
tures from these measurements and the statistical analysis of these features to
determine the current state of the system’s health [18]. SHM usually involves
online global damage identification in structural systems. Global health monitoring
however does not give sufficiently accurate information to determine the extent of
the damage and are mostly based on either finding shifts in resonant frequencies or
changes in structural mode shapes. Modal information is a reflection of the global
system properties while damage such as crack initiation and propagation is a local
phenomenon [19, 20]. Effects of environmental changes such as temperature can

also mask subtler structural changes caused by damage [21, 22]. Hence, global
SHM may not be effective for crack detection and monitoring.
Two major factors further make crack monitoring in concrete structure very
challenging: exact crack locations cannot be determined a priori because of con-
crete’s high inhomogeneity; and there are many possible crack locations along the
concrete due to the load redistribution effect of the steel reinforcements [6]. Fur-
thermore, all existing electrical instruments and most fiber-optic sensors are “point”
sensors that can only detect changes at local points in a structure [6]. Such mea-
surements are not effective for health monitoring of concrete structures since
degradation of such structures occurs through the formation and propagation of
localized cracks that have little effect on structural stiffness [6, 23]. Consequently,
measuring the displacement, strain, or acceleration at a particular sensing point is
insensitive to cracking unless the point is very close to the damage location or by
using very large number of sensors which is impractical [6, 23].
13.3 Triboluminescence
The triboluminescence phenomenon has generated extensive research interest over

the years because of its potential application for damage detection [24]. Tribolumi-
nescence, also known as fracto- [25], piezo- [26], or mechano-luminescence [27], is
the emission of light by solid materials when they are stressed or fractured
[28, 29]. About 50 % of all crystal compounds exhibit a range of TL [30]. Although
TL was reportedly discovered in the sixteenth century by Sir Francis Bacon, serious
research into the phenomenon only began in the twentieth century. The main
contribution of early nineteenth century research was the compilation of an exten-
sive list of TL materials using highly subjective visual observation of the TL
response as a function of time and quantity [30]. The development of the
photomultiplier tube (PMT) in the 1930s and its application in triboluminescence
studies in 1952 however introduced a quantitative technique for detecting, measur-
ing, and comparing TL emissions objectively.
Triboluminescence may be divided into three types namely [9, 31] (1) elastico-,
(2) plastico-, and (3) fracto-triboluminescence. Elastico-TL is luminescence pro-
duced during the elastic deformation of solids, where neither fracture nor plastic
deformation is required. This can be by mechanical or electrostatic interaction of
dislocations with defect centers, or by thermal excitation in the stressed regions of
solids [31]. Examples of crystals that exhibit elastic-TL include X- or g-irradiated
alkali halides and ZnS:Mn [31]. Plastico-TL is luminescence produced during
plastic deformation of solids where fracture is not required. It can be excited by
the mechanical or electrostatic interaction of dislocations with defect centers;
electrification of crystal surfaces by the movement of charged dislocations; or
thermal excitation in the stressed regions of solids such as colored alkali halides,
II–VI compounds, alkaline-earth oxides, and metals [31]. Fracto-TL is
luminescence produced due to the creation of new surfaces during the fracture of
solids. During fracto-TL, there is creation of charged surfaces at fracture due to
processes such as piezoelectrification, movements of charged dislocations, and
charged defect barodiffusion [31, 32]. There is neutralization of these surface
charges by the charge carriers or ions produced from the dielectric breakdown of
the intervening gases and solids. This results in the production of luminescence that
resembles a gas discharge (e.g., sucrose, tartaric acid, and Rochelle salt) or lumi-
nescence of a solid (e.g., coumarin, resorcinol, and phenanthrene) or one that
combines the characteristics of both the intervening gases and solids (e.g., uranyl
nitrate hexahydrate, impure saccharin, and chlorotriphenyl-methane [29, 31].
13.3.1 Triboluminescent (TL) Sensor
Triboluminescence-based sensor systems have the potential for wireless, in situ,

real-time, and distributed (WIRD) sensing that can enable continuous monitoring of
civil and aerospace structures [9]. A Triboluminescence-based sensor system com-
prising highly efficient triboluminescent materials could allow simple, real time
monitoring of both the magnitude and location of damage with minimal parasitic
influence to the host structure [33, 34]. They can be used as stress, fracture, and
damage sensors [35, 36]. They have also been proposed for visualizing the stress
field near the crack-tip, stress distribution in solids, and quasi-dynamic crack-
propagation in solids [36–41].
13.3.2 Triboluminescent Materials for Sensing Applications
The principal requirements for any candidate triboluminescent (TL) material are
that it should be highly efficient (allowing detection using inexpensive, compact
detectors) and that it should emit light over a discrete and thus characteristic
spectrum (allowing integration into an array of sensors to monitor damage location)
[34, 42]. The TL efficiency of different TL materials (Table 13.3) have been
measured using the photoluminescence quantum based on the assumption that
both phenomena originate from the same excited state for the compounds studied
[44]. The ZnS–Mn has the highest TL efficiency among the TL materials studied. In
addition, metal complexes are among the most intense TL materials known
[34]. Metal complexes include europium tetrakis (dibenzoylmethide)
triethylammonium [45], tris (1,3-di-tert-butyl-B-propane dione) terbium-p-
dimethylamino pyridine [44] and dibromobis (triphenylphosphine oxide) manga-
nese [40, 46]. Most metal complexes contain a trivalent lanthanide as the emitting
center [34] (Table 13.2).
Table 13.2 Solid-state photoluminescent quantum yields of a number of triboluminescent com-

pounds [34, 44]
Compound Name Φ
1 Europium tetrakis (dibenzoylmethide) triethylammonium 0.75
2 Europium tris (2-thenoyltrifluoroacetone)-phenanthroline 0.85
3 Hexakis(antipyrine) terbium triiodide 0.53
4 Tris(1,3-di-tert-butyl-β-propanedione) terbium-p-dimethylaminopyridine 0.71
5 Tris(1,3-di-tert-butyl-β-propanedione) samarium-p- 0.04
dimethylaminopyridine
6 Tris(1,3-di-tert-butyl-β-propanedione) dysprosium-p- 0.07
dimethylaminopyridine
7 Terbium trichloride hexahydrate 0.33
8 Mn:ZnS 1.00
9 Dibromobis(triphenylphosphine oxide) manganese 0.19
10 Europium (2+) hexacelsian >0.57
11 Samarium (2+ and 3+) hexacelsian 0.62
12 Cerium (3+) hexacelsian 0.44
13 Cholesteryl salicylate 0.13
14 N-Acetylanthranilic acid >0.28
15 Salicylsalicyclic acid 0.17
16 Menthyl-9-anthracenecarboxylate >0.43
13.4 Ubiquitous Damage Sensing with Randomly

Distributed Triboluminescent Materials
The first approach investigated in developing a cementitious composite system with

a triboluminescence-based damage detection system involved the direct dispersion
of ZnS:Mn crystals in a mortar system. The approach is based on the assumption
that as the CC system is loaded and crack formation occurs, the stress field
developed around and along the crack path will cause the excitation of the crystals
along the crack path. Such a system will exhibit ubiquitous damage detection
[43]. It will be able to detect any crack provided there is a TL crystal along the
crack’s path. In order to have an effective system, there is the need to determine the
concentration level of the ZnS:Mn crystals required to ensure that any crack at any
point in the system is detected without adversely affecting the mechanical proper-
ties of the cementitious composite.
The triboluminescent cementitious optical sensor (TLCOS) was prepared by
physically mixing fly ash, water, cement, and ZnS:Mn crystals. The various prop-
erties of the constituents are given in Table 13.3. The ZnS:Mn is the tribolumines-
cent material that emits light when stressed or fractured. The samples produced had
dimensions 50 50 6.25 mm (2 2 0.25 in.).
The fabricated TLCOS samples were tested in a custom-built impact rig
(Fig. 13.3) where they were impacted with a pneumatically controlled piston at
specific pressure values. The TL signals from the TLCOS was measured by
Table 13.3 Main constituents of the triboluminescent cementitious optical sensor (TLCOS) [43]
Bulk specific Particle size
Material gravity distribution (μm) Proportion (by mass)
Type I/II Portland cement 3.15 1–50
Boral class F fly ash 3.00 25 25 % of cement
ZnS:Mn 4.10 5–20 0, 5, 7.5, 10, 15, 20, 25 %
of cementitious mixture
Water 1.00 Approx. 50 % of cement
Fig. 13.3 Schematic of the custom-built impact rig for assessing TL performance of sensor patch
samples [43]
connecting the plastic optical fibers (POF) from the fabricated samples to the
photomultiplier tube (PMT) which converted the TL optical signals in to electrical
signals (voltage). A National Instrument data acquisition device linked the PMT to
a personal computer for real time data collection.
The compressive strength of samples were determined using a CM-3000-DB
compression rig from Test Mark Industries (Fig. 13.4) according to ASTM C109/
C109M—08 standard test methods for compressive strength of hydraulic cement
mortars, using cube specimens of size 50 mm (2 in.).
Test results (Fig. 13.5) show that there tends to be an increase in the tribolumi-
nescent response of the TLCOS samples with increasing ZnS:Mn concentration
levels. Triboluminescent signals were detected at concentration levels as low as 2.5
and 5 % although not all the samples tested at these low concentration levels
produced a triboluminescent response on impact [43]. The crack paths may have
missed the scantily dispersed ZnS:Mn crystals at these low ZnS:Mn concentration
level of the TLCOS samples. Triboluminescent responses were however detected
Fig. 13.4 Experimental

setup for determining the
compressive strengths of
samples [43]
for all samples tested at concentration levels of 7.5 % and higher. Consequently, a
ZnS:Mn percolation level of 7.5 % is required for an effective TLCOS [43].
At concentration levels of 7.5 % weight and above, there are more ZnS:Mn
crystals dispersed within the composite system that are being activated by the
impact energy during testing. These higher levels of ZnS:Mn concentration also
make it possible for sufficient number of crystals to be located near to the trans-
mission component (polymer optical fiber) of the TLCOS for effective TL signal
coupling into the POF and for detection. The opaque property of the cementitious
component of the TLCOS kept the amount of TL signal successfully coupled into
the POF fairly constant for all ZnS:Mn concentration levels of 7.5–20 % [43]. Sig-
nificant increase in the TL response was however observed in TLCOS samples with
a concentration level of 25 %. The mean triboluminescent signal increased by about
2.5 times from a value of 0.16 V at 7.5 % concentration level to a value of 0.42 V at
25 % concentration level [43]. It is believed that at this high ZnS-Mn concentration
level, there were much more crystals in the TLCOS and around the POF to reduce
the effect of the opaque cementitious component. Hence, the significantly higher
TL signal generation and coupling.
Figure 13.6 shows the effect of ZnS:Mn concentration levels on the compressive
strengths of 28 days-cured samples. There was a reduction in the compressive
strength of the pristine cementitious patch from about 29.5 MPa to about 26 MPa
at ZnS:Mn concentrations of 5 and 7.5 %. Enhancement of the compressive strength
0.6
0.5
0.4
TL response (V)
0.3
0.2
0.1
0.0
0.0 2.5 5.0 7.5 10.0 15.0 20.0 25.0

ZnS:Mn concentration (%weight)
Fig. 13.5 Effect of ZnS:Mn concentration level on TL output of composite [43]
32000
Compressive strength (KPa)
30000
28000
26000
24000
22000
20000
0.0 5.0 7.5 10.0 15.0 20.0 25.0

ZnS:Mn concentration (%weight)
Fig. 13.6 Effect of ZnS:Mn concentration levels on the compressive strengths of 28-day cured
samples [43]
was however observed at ZnS:Mn concentration levels of 10 % and 15 % with mean

values of about 31 MPa and 30 MPa, respectively, compared to a mean value of
about 29 MPa for the mortar system with no ZnS:Mn content. The triboluminescent
crystals seem to provide some form of particle reinforcement at these concentration
levels [43]. The compressive strength (28 MPa) measured at 20 % ZnS:Mn con-
centration level is still greater than those at 5 and 7.5 % concentration levels. There
was however A drastic loss of strength at a ZnS:Mn concentration of 25 %. There
was a reduction of about 29.31 % in the compressive strength of the TLCOS with
25 % ZnS:Mn concentration compared to the control [43].
Fig. 13.7 SEM micrographs of 28-day cured cementitious composite (a) 0 % ZnS:Mn by wt, (b)
25 % ZnS:Mn by wt [43]
The products from the curing of cement (cement hydration) are the weaker link
in the TLCOS system because about 100 MPa is required for the cleavage of a ZnS:
Mn particle [31, 47] while the maximum value of the compressive strength obtained
for the samples tested was about 32 MPa. The ZnS:Mn can therefore act as
particulate reinforcement in the TLCOS. Sample fabrication process may have
caused the coagulation of the ZnS:Mn crystals resulting in the formation of more
microvoids that made the TLCOS system weaker [43].
Scanning Electron Microscope (SEM) micrograph of 28 days-cured samples
with 0 % wt ZnS:Mn (Fig. 13.7a) showed little or no ettringites while the 28-day
cured sample with 25 % wt ZnS:Mn showed the presence of ettringites (Fig. 13.7b).
The conclusion was that at the high ZnS:Mn concentration level of 25 %, the crystal
content may affect the hydration process by slowing down the conversion of the
needle-shaped crystals of calcium sulfoaluminate hydrate (ettringites) into the
monosulfate hydrate with hexagonal plate morphology [43]. The monosulfate
hydrate makes concrete vulnerable to sulfate attack. Similar behavior was observed
in samples cured for a shorter period (10 days) and with lower ZnS:Mn (5 % wt)
concentration (Fig. 13.8).
Similarly, Fig. 13.9 clearly shows ZnS:Mn crystals with fractured surfaces in the
TLCOS. The ZnS:Mn crystals of require about 1 MPa (106 Nm2) for elastico-
triboluminescence to commence with an elastic limit of about 30 MPa (30 106
Nm2), and about 100 MPa (1 108 Nm2) for fracto-triboluminescence
[36, 47]. In addition, coagulation of the crystals within the cementitious composite
system was observed and it may be possible to reduce the TL crystals concentration
requirement by improving the dispersion of the coagulated ZnS:Mn crystals
throughout the composite system [43]. Reduced crystal content is beneficial for
reducing any negative effect the presence of the TL crystals may have on the
structural properties of the cementitious composites [43].
Fig. 13.8 SEM micrographs of 10-day cured cementitious composite: (a) 0 % ZnS:Mn by wt, (b)
5 % ZnS:Mn by wt [43]
Fig. 13.9 SEM micrograph showing the interface between ZnS:Mn crystals and the products of
hydration (5 % wt ZnS:Mn, 10-day cure): (A) ettringite, (B) flyash, (C) ZnS:Mn, (D) calcium
silicate hydrate, (E) calcium hydroxide [43]
13.5 Damage Sensing with the In Situ Triboluminescent

Optical Fiber (ITOF) Sensor
A major challenge militating against triboluminescence-based sensing systems

involve the need effective capture and transmission of the optical signal generated
within opaque composites like concrete [9]. Direct dispersion of TL materials in the
cementitious composite introduces some undesirable effects. The concentration
level (weight fraction) of the triboluminescent (TL) crystals in the host matrix
Fig. 13.10 A schematic of

the ITOF sensor [48] Triboluminescence Sensory
Receptor (TSR) POF with jacket
Polymer optical fiber

(POF)
required for good TL response is usually high. This introduces parasitic weight
effect that can adversely affect the properties and performance of the host material.
The in-situ triboluminescent optical fiber (ITOF) sensor overcomes most of these
challenges. The (ITOF) sensor mimics the sensory receptor of the human nervous
system (HNS) by converting the energy from damaging events like impact and
fracture into optical signals that are indicative of the magnitude of damage [48]. The
ITOF sensor combines the highly desirable properties of optical fibers such as
lightweight, small size, immunity to electromagnetic interference, and capacity
for distributed sensing [2, 7], with the triboluminescent property of ZnS:Mn.
The ITOF sensor (Fig. 13.10) consists of a polymer optical fiber (POF) with a
highly sensitized section, the triboluminescent sensory receptor (TSR) [48]. The
TSR consists of an epoxy system with dispersed ZnS:Mn crystals which is one of
the most efficient triboluminescent materials known [44]. The ITOF sensor senses
damage in the composite structure by converting the mechanical energy of the
damage causing event like impacts and crack propagation into triboluminescent
(TL) optical signals. The TL signals are then transmitted through the POF to a
photodetector such as a photomultiplier tube (PMT) or a photodiode. The photo-
detector converts the TL optical signals into electrical signals that can be analyzed
by a computer system for damage characterization.
Figure 13.11a shows some of the ZnS:Mn crystals used in this study. The ITOF
coating, the interface between the coating and POF, and the POF are shown in
Fig. 13.11b. The smooth morphology of the POF is easily distinguishable from the
ZnS:Mn crystals-loaded ITOF coating. The ZnS:Mn crystals are randomly distrib-
uted in the epoxy matrix of the coating (Fig. 13.11c). Except for a few larger ZnS:
Mn crystals, the crystals are uniformly dispersed throughout the epoxy film giving it
the capability to produce triboluminescent signals during impact, cracking, or
flexing. There are also voids and crystals agglomeration in the sensor coating as
highlighted in Fig. 13.11d.
The development of the ITOF sensor will make it possible to create
multifunctional cementitious composites with in-situ damage detection capabili-
ties. Such a system will mimic the human nervous system with the integrated ITOF
sensors acting as the nerves with both sensing and signal transmission capabilities.
Figure 13.12 is a schematic illustration of a triboluminescent multifunctional
cementitious composite (TMCC) bridge girder with in situ crack monitoring capa-
bilities for protecting the steel reinforcement from corrosion through early crack
detection.
Fig. 13.11 Scanning electron micrographs: (a) ZnS:Mn crystals. (b) Cross section of an ITOF
sensor showing the ZnS:Mn crystals loaded coating, interface between the coating and the smooth-
surfaced POF. (c) ITOF sensor coating with ZnS:Mn crystals randomly dispersed in the epoxy
matrix. (d) Void and crystal agglomeration in the ITOF coating [49]
ITOF 2
Steel
Rebar
ITOF 1
Crack A Crack B
Fig. 13.12 Schematic of double-layer sensor TMCC girder for early crack detection [49]
13.5.1 Real-Time Failure Detection in Unreinforced

Cementitious Composite
The mix design for the fabricated triboluminescent cementitious composite (TCC)
is given in Table 13.4. After thorough mixing in an electric-driven concrete mixer,
the CC mix was poured into prepared silicon molds. The ITOF sensors were placed
in the molds (Fig. 13.13a) before pouring the CC mix (Fig. 13.13b). The CC beams
were de-molded after 24 h and immersed in water in a curing basin at room
temperature to cure for 28 days.
The TCC beams were instrumented (Fig. 13.14a) and subjected to three point
bend tests (3PBT) as shown in Fig. 13.14b. The TL signals generated by the TCC
during crack propagation were transmitted through the optical fibers to the
photomultiplier tube (PMT) where they were converted into electrical signals
(voltage) and quantified.
The typical responses obtained at the instance of failure of the unreinforced
cementitious beams are as shown in Fig. 13.15a, b. Failure was successfully
detected in real time, as indicated by the coincidence between the TL signal
generated when the sample fractured, and by the loss in the load carrying capacity
(Fig. 13.15a), as well as the sudden increase in the strain value (Fig. 13.15b). The
integrated ITOF sensor gave no false alarm and only generated TL signals at the
instance of brittle failure (Fig. 13.15a, b) [50]. The crack initiation and propagation
was so rapid in the unreinforced mortar beams that they exhibited brittle and
catastrophic failure [50].
Table 13.4 Constituents Proportion to cement

of triboluminescent Ingredient content (by wt)
cementitious composite
Portland cement 1.000
(TCC) [50]
Fly ash 0.300
Sand 3.575
Plasticizer 0.012
Water 0.585
Fig. 13.13 (a) Silicon mold with single-layer ITOF sensor ready for CC pouring, (b) fabricated
TCC beams ready for de-molding from silicon mold [50]
Fig. 13.14 (a) TCC beams with strain gages installed. (b) Three-point bend tests (3PBT) set
up [50]
a b
1.0 10.0 1.2
TL signal (V) TL signal (V) 200
Stress (MPa) 1.0 Strain (ue)
0.8
7.5
TL signal (V)
150
TL signal (V)
0.8
Stress (MPa)
Strain (ue)
0.6
5.0 0.6
100
0.4
0.4
2.5 50
0.2
0.2
0.0 0.0 0.0 0

0 10 20 30 40 0 10 20 30 40
Time (s) Time (s)
c d
4 4
3
TL signal (V)
TL signal (V)
2
2
1
1
0
0
3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4
Ultimate stress (MPa) Ultimate stress (MPa)
Fig. 13.15 (a) TL signals generated at the instance of brittle failure of unreinforced mortar beam,
(b) large increases in strain value at the instance of brittle failure corresponding to the instance of
TL signal emission, (c) TL signals at failure in TCC beams with sensor located about 10 mm from
beam base (tension side), (d) TL signals at failure in TCC beams with sensor located about 40 mm
from beam base (tension side) [50]
13.5.1.1 TCC Performance with ITOF Sensor Positioned Within

the Tension Zone
The performance of the TCC depends on the position of the ITOF sensor in the TCC
beam relative to the neutral axis. Based on the classical elastic beam theory, the
neutral axis lies in the midpoint of the beam. Any sensor position (y) below 44 mm
lies within the tension zone while any position above that lies within the compres-
sion zone (Fig. 13.16). Figure 13.15c, d show the TL responses obtained at failure
of ten (10) TCC beams tested with the integrated ITOF sensors positioned at two
different depths within the tension zone of the beam (Fig. 13.16) [50].
The TL signal intensity tended to increase with increase in the failure or
ultimate stress for beams with ITOF sensors located close to the base of the
beam (point of brittle failure initiation) as shown in Fig. 13.15c. The farther the
sensor is from the point of brittle failure initiation (in this case the tension face of
the beam), the lesser the correlation between the TL signal detected and the
ultimate stress at failure (Fig. 13.15d) [50]. The inhomogeneous nature of the
cementitious composites (CC) and the longer distance between sensor location
and beam’s tension face must have been responsible for the poor correlation
between the TL signal value and ultimate stress for sensors positioned about
40 mm from the base of the beams [50]. Failure usually occurs as crack initiation
and propagation from the tension face of a concrete beam under flexural loading.
The ultimate stress measures the stress at this outer fiber/layer at the instant of the
brittle failure.
Based on this study, it may be stated that the farther away an ITOF sensor is
located from the point of failure (crack) initiation, the less accurately the sensor can
sense the stress level responsible for failure. Sensors located farther away from the
point of failure initiation are however able to sense failure initiation and crack
propagation over a larger distance starting from the point of failure initiation to
shortly after the crack passes the sensor location.
Compression
Neutral
axis
End view Tension
Fig. 13.16 Position of integrated ITOF sensor relative to the neutral axis of the TCC beam [50]
Fig. 13.17 TL emission 55

duration tends to increase
with distance of sensor 50
Emission duration (ms)

location from the point of 45
failure initiation [50]
40
35
30
25
20
15
10 20 30 40 50
Sensor location from beam base (mm)
13.5.1.2 TL Emission Profile Analysis
The TL emission duration was measured up to the time at which the TL signal level
dropped to a value of about 0.0358 V. This value was chosen because multiple
peaks, indicative of multiple emissions, were observed in the emission profiles.
Consequently, the usual convention of estimating the emission duration by deter-
mining the time for the TL emission to fall to e1 value of its peak value could not
be used. The result indicates that the farther away the sensor is located from the
point of failure initiation, the larger the crack propagation distance and the longer
the TL emissions durations observed (Fig. 13.17).
The ITOF sensors located nearer to the base of the beam had a mean TL emission
duration of 18.90 ms with a standard deviation of 4.12 ms. The sensors located
farther away from the base of the beam however had a mean TL emission duration
of 35.98 ms with a standard deviation value of 11.13 ms. The result suggests that
sensors located farther away from the tension face of the beam were able to detect
the rapid crack propagation over the longer distance the crack has to propagate
before crossing the sensor. This capability may be of great value in studying crack
propagation in cementitious composites [50].
Figure 13.18a shows a typical TL emission profile obtained from the ITOF
sensor during the brittle failure of a TCC beam. The fracture plane crossed the
sensitized region of the integrated ITOF sensor and four distinct peaks in the TL
emission profile was observed. The first peak (I) has a TL intensity value of 0.30 V
while the second peak (II) has the largest TL intensity value of 1.09 V. The third
(III) and fourth (IV) peaks had TL intensity values of 0.70 V and 0.75 V, respec-
tively [50]. The TL profile indicates that the event lasted for about 18 ms. While
multiple peaks could be observed when the ITOF sensor was integrated into a CC,
only a single peak was observed when the sensor was loaded directly (without
integrating into a CC) to failure under 3 PBT loading condition (Fig. 13.18b).
Furthermore, the emission duration from the ITOF when directly loaded was
about 3 ms which is typical for ZnS:Mn crystals. Earlier work by Hollerman
a b
1.2
II 1.4
1.0 1.2
1.0
III IV
TL signal (V)
0.8
TL signal (V)
0.8
0.6
0.6
0.4 0.4
I
0.2 0.2
0.0
0.0
53.473 53.479 53.486 53.493 53.500 117.980 117.997 118.014 118.031
Time (s) Time (s)
Fig. 13.18 Typical TL emission profile from (a) integrated ITOF sensor at failure of CC beam, (b)
ITOF sensor at failure under direct 3 PBT loading [50]
et al. [36] indicates that the TL emission of ZnS:Mn lasts for about 3 ms and that ZnS:
Mn is strongly triboluminescent with a prompt fluorescence decay time of about 300 μs
[8, 36]. The fluorescence decay time is unique to each material and can be defined as
the time needed to reduce the light intensity to e1 (36.8 %) of its original value
[36]. The results clearly indicate that the observed multiple peaks and the extended
emission durations were from multiple TL emissions caused by the rapid crack
initiation and propagation detected by the integrated sensor as the beam failed [50].
In addition, there are some differences in the number of emissions between the
TL responses of the CC beams based on the ITOF sensor location [50].
Figure 13.19a, b show TL emission profiles from beams with the ITOF sensor
positioned about 10 mm from the tension side of the CC beams while the emission
profiles in Fig. 13.19c, d are from CC beams with the sensor located about 40 mm
from the tension face of the beams. The sensors positioned farther from the point of
failure initiation (about 40 mm from base of beam) experienced higher number of
TL excitations and emissions as indicated by the higher number of peaks
(Fig. 13.19c, d) than sensors positioned nearer to the point of failure initiation
(about 10 mm away from the tension face of the beams). The longer the distance the
crack had to travel from the point of crack initiation before reaching the ITOF
sensor, the more the numbers of TL excitations that occur in the sensor as the sensor
continually detects the rapidly propagating crack. This capability may be employed
in the study of crack propagation in cementitious composites like concrete.
13.5.1.3 ITOF Sensor Behavior in Compression Zone
The number of peaks, indicative of the number of TL emissions, is much larger in a

TCC beam with the ITOF located within the compression zone (Fig. 13.20a) than in
TCC beams with sensors located within the tension zone [50]. The TCC beams with
a b
1.8 3.5
1.6
3.0
1.4
TL intensity (V)
2.5
TL signal (V)
1.2
1.0 2.0
0.8 1.5
0.6
1.0
0.4
0.2 0.5
0.0 0.0
69.734 69.741 69.748 69.755 69.762 51.880 51.885 51.890 51.895 51.900
Time (s) Time (s)
c d
1.4
4
1.2
1.0 3
TL signal (V)
TL signal (V)
0.8
2
0.6
0.4
1
0.2
0.0 0
44.552 44.560 44.568 44.576 44.584 60.090 60.097 60.105 60.112 60.120 60.127
Time (s) Time (s)
Fig. 13.19 TL emission profile from ITOF sensor at failure of CC beams located (a) 10 mm, (b)
11 mm, (c) 38 mm, (d) 40 mm, from the base of the beams [50]
0.14 a b
0.12
0.10
TL signal (V)
0.08
0.06
0.04
0.02
0.00
58.46 58.48 58.50 58.52 58.54 58.56

Time (s)
Fig. 13.20 (a) TL profile from ITOF sensor in compression zone of mortar beam. (b) Larger crack
width in the tension zone than in compression zone at failure of mortar beam [50]
sensors positioned at about 10 mm and 40 mm had about four (Fig. 13.19a, b) and
seven (Fig. 13.19c, d) peaks respectively. On the other hand, the TCC beam with
the sensor in the compression zone had about 13 peaks (Fig. 13.20a). In addition,
the observed peak TL intensity value of 0.13 V at an ultimate stress of 6.02 MPa is
significantly lesser than that observed in the TCC beam with the sensor positioned
in the tension zone, with a minimum peak TL value of 1.08 V at an ultimate stress
value of 4.53 MPa.
The significantly low peak TL intensity value clearly shows that the sensor
located in the compression zone (50 mm from the beam base) did not experience
the high failure initiation and crack propagating energy experienced by the sensors
positioned within the tension zone of the beams. Furthermore, there is a lesser
probability of the ITOF sensor experiencing fractoluminescence due to the fracture
of the TL crystals within the ITOF coating because the sensor is under compression
loading. The fracturing of the coating results in fractoluminescence that usually
leads to the generation of much higher TL intensity values. The crack widths that
could well exceed 3 mm and the sudden caving in of the two fractured surfaces
definitely promote the fracturing of the coating particularly for sensors located in
the tension zone of the beam (Fig. 13.20b).
The compression state of the beam in the compression zone tended to slow down
the crack propagation speed as evidenced by the prolonged and multiple peaks
(emissions) [50]. The delayed separation of the cracked beam surfaces compared to
the quick separation of the cracked surfaces in the tension zone is believed to be
responsible for the prolonged and multiple peaks (emissions) observed in the TCC
beam with the ITOF within the compression zone [50]. The TL emission duration
was longer, over 50 ms. Consequently, it may be possible to gain insight into the
loading condition within a structure, whether it is in tension or compression, from
an analysis of the TL profile generated from the integrated ITOF sensor. Further
work is needed before this approach becomes a practical reality.
13.5.2 Real-Time Damage Monitoring in Reinforced

Cementitious Composite
The reinforced concrete beams had three sections of the integrated ITOF sensors
coated with the triboluminescent (TL) composite in an attempt to enhance distrib-
uted damage sensing along the length of the ITOF (Fig. 13.21a, b). Unlike the
unreinforced CC systems that exhibited brittle failure, the reinforced CC system
exhibited a more ductile behavior characterized by multiple cracks formation
(Fig. 13.21c, d) from the flexural loading [51].
Figure 13.22a shows the TL signal and load profiles obtained during the flexural
loading of the beam shown in Fig. 13.21d. Similar profiles were obtained from other
tested beams. Although four cracks were observed on the tested beam, only three
TL signal jumps were detected. The first TL signal (0.14 V) was detected when the
beam was loaded to approximately half of its strength (50 kN) at time 109 s. The
second TL signal (0.09 V) was detected at time 123 s at a load value of 57 kN.
The third and highest TL signal value of 0.18 V was detected shortly before
failure at time 207 s and at a load of 95 kN. The beam failed at time 236 s and at a
Fig. 13.21 (a) Schematic showing the location of the coated sections of the ITOF within the
beam. (b) Fabricated ITOF sensor with three sections coated with TL composite for enhanced
distributed damage sensing. (c) Multiple cracks from 4-point bend test. (d) Close view of multiple
cracks on tested RC beam [51]
Fig. 13.22 (a) Multiple TL signals from multiple cracks providing early warning before structural
failure of RC beam. (b) Base (tension side) of tested beam with cracks not being detected by strain
gage but detected by integrated ITOF (black horizontal lines indicate position of integrated ITOF
sensor) [51]
load of 105 kN [51]. The continued loading of the beam after the third TL signal
was detected resulted in increase in the crack width. This caused more ambient light
to be detected resulting in the gradual increase in the TL noise level from time 206 s
(Fig. 13.22a). Visual inspection of the sample after the test showed that only two of
the four cracks passed through the sensitized section of the integrated ITOF sensor
(Fig. 13.22b). The second TL signal (with the least TL signal value of 0.08 V) is
most likely due to the non-contact excitation of the ITOF sensor by a crack that
propagated about 20 mm away from the end of the nearest coated section [51]. The
fourth crack was not detected because it propagated farther away from the ITOF
sensor (about 81 mm from the end of the nearest coated section) [51]. The detection
of localized damage (cracks) by the ITOF sensor is a good indicator of the viability
of the ITOF sensor for structural health monitoring applications [51]. The ITOF
sensor also provided warning of impending structural failure. There was a big jump
in the TL signal to 0.18 V at time 207 s before the failure of the beam. There was
also an increase in the noise level, a clear indicator of crack opening that may lead
to structural failure. This observation holds promise for crack width monitoring
which is critical for corrosion prevention [51]. Further work is necessary to fully
exploit this possibility.
The magnitude of the TL signal levels detected in the reinforced concrete beams
are also much lower than that detected at the failure of the mortar (unreinforced CC)
beams. The difference could be explained by the fact that the detected TL signals in
the reinforced CC were from TL emissions caused by multiple localized damages
(cracks) in the RC beams and not from the RC beam failure [51]. The ability to get
signals before failure in the RC beams further provide evidence of the potential of
the ITOF sensor as a viable structural health monitoring technique that can provide
early warning of system overload or deterioration before structural failure. When
compared to conventional point sensors like strain gages, the ITOF sensor offers
superior advantages as a distributed sensor when its entire length is coated with the
TL composite coating [51]. As shown in Fig. 13.23, point sensors like strain gages
may not be able to provide continuous monitoring because of ease of damage. The
crack (damage) that occurred in the beam (Fig. 13.23a) was detected by both the
ITOF sensor and installed strain gage as indicated by the jump in the TL signal and
strain value at the instance of crack propagation (Fig. 13.23b). The crack damaged
the strain gage at about time 80s thereby ending its damage monitoring function
while the ITOF sensor continued to provide real-time damage monitoring of the
structure until time 255 s when the beam eventually failed.
Fig. 13.23 (a) Crack propagating through location of installed strain gage and coated section of
integrated ITOF sensor. (b) Crack detection by strain gage and ITOF sensor with damage to strain
gage [51]
13.5.3 ITOF Sensor for Retrofitted Concrete Structures
Fiber-reinforced polymer (FRP) materials in form of thin laminates are being used
as ideal alternatives to traditional repair technology for aging and damaged concrete
structures. FRP possess many desirable properties such as high strength to weight
ratios, and high stiffness to weight ratios. They offer significant potential for
lightweight, high strength, cost-effective, and durable retrofitting systems for con-
crete structures like bridges. A major drawback to the use of FRP retrofitting is its
brittle-type failure behavior because of its nearly linear elastic stress-strain behavior
[52]. This can cause significant reduction in the strength of the retrofitted member.
Consequently, the health of the retrofit applied on the concrete structure becomes
equally important so as to sustain the serviceability of the structure. The ITOF
sensors were used to monitor the failure (debonding) of carbon fiber-reinforced
polymer (CFRP) retrofit on loaded precracked reinforced concrete beams.
The beams used in this study were chosen based on the failure mode from
experiments conducted by Olawale et al. [51]. The beams were reinforced in
tension using #2 steel bars with a cover of 0.5 in. The reinforcing bars were of
grade 40 and cut to a length of 20 in. before placing them in position. The beams
were subjected to four-point bending, and the test was terminated when cracks were
established on the tension face of the beam (Fig. 13.24). The beam was loaded to
70 % of the estimated strength, before unloading and the load recorded before
unloading was 70 kN [51].
The cracks on the surface of the precracked beams were injected with two-part
epoxy polymer compound and the surface prepared by smoothening with a mechan-
ical grinder and cleaning with acetone to ensure a suitable surface [53] for applying
the carbon fiber plies [52]. The critical step in secondary bonding is surface
preparation, which creates chemically active sites that increase the bond
strength [52].
Three carbon fiber plies were cut to overlap the tension region of the beam to
overlap the cracks. The plies were designed according to the ACI 440.2R-02 and
NCHRP report 655 guidelines [54, 55]. The plies were cut to a length of 15 and 6 in.
Fig. 13.24 Tension face of

a precracked beam [52]
Fig. 13.25 Control sample

with CFRP retrofit and no
integrated ITOF sensor [52]
Fig. 13.26 Triboluminescent optical fiber sensor: (a) schematic, (b) uncoated and coated optical
fiber [52]
wide. Figure 13.25 shows the control sample which is a retrofitted precracked
concrete beam with no integrated ITOF sensor.
For instrumented retrofitted precracked concrete beams, continuous sensors
(ITOF with entire length coated with the TL composite film) were fabricated
(Fig. 13.26). Two optical fiber sensors were placed between the adhesive layer
and the first carbon fiber layer, to detect the adhesive failure mode. The other two
sensors were placed between two carbon fiber layers to detect failure by delamina-
tion or failure within the composite. A 40 % epoxy by weight was applied on the
concrete beam at room temperature by brushing at a rate of 50–150 sp.ft/gal., after
each carbon fiber layer. Rollers were used to remove excessive epoxy and to evenly
apply the carbon fiber. The ITOF sensors layup and the beam schematic is shown in
Fig. 13.27. The optical fibers extending from the retrofit were coated with a rubber
coating to provide insulation from external light so as to reduce the noise level.
The retrofitted precracked concrete beams were subjected to four-point bending
test according to the guidelines from the ASTM C78-09 [56] standard test methods
21"
6"
4 ITOF @ 1.2" c/c
17"
Fig. 13.27 Sensor layup on the tension face [52]
Fig. 13.28 Data acquisition for sensor response: (a) sensor nomenclature, (b) data acquisition
[52]
for flexural strength of concrete. The TL signals from the integrated ITOF sensors
were detected with a multichannel photomultiplier tube (PMT) manufactured by
Hamamatsu Corporation, Japan (H10722-01). The integrated ITOF sensors in the
retroffited precracked RC beams were identified as shown in Fig. 13.28. The sensor
labeled “4-I” was positioned between the first and the second carbon fiber layer to
detect possible emissions resulting from internal failure in the composite. The TL
emissions from this sensor were detected in the fourth channel of the PMT
(Fig. 13.28b). Similarly, the sensor labeled “2-o” was positioned within the adhe-
sive interface to detect the bond failure between the composite and the concrete.
Debonding failure mode (Fig. 13.29) was observed in beam-II. A change in
stiffness and discontinuity of beam curvature created a stress concentration in the
concrete initiating cracks that lead to the debonding [52]. Debonding initiated at a
load of about 50 kN while failure occurred at 85 kN after the debonding propagated
to the middle third of the beam. Table 13.5 provides a summary of each tested
beam’s performance.
Figure 13.30 shows the TL signal and stress profiles obtained during the flexural
loading of beam-II. The FRP retrofit debonding commenced at a load of 50 kN
(or 9 MPa) and was detected by the ITOF sensor labelled “7-0” installed within the
adhesive layer (Fig. 13.30). The two ends of the optical fiber were inserted into two
separate channels (channel 7 and channel 8) of the PMT and both successfully
detected TL signals during the beam loading. Channel 7 (Fig. 13.30) recorded the
first TL signal value of 2 mV at a load of 50 kN and at time 78 s. The second TL
Fig. 13.29 Debonded FRP composite at a stress of 16 MPa [52]
Table 13.5 Beam performances [52]

Sample Strength (MPa) Deflection (mm) Failure mode
Control beam 18.1 7.2 Debonding and flexure
Beam-I w/ sensors 14 8.7 Flexure
Beam-II w/ sensors 16.1 7.3 Debonding and flexure
×10-3
6 18
TL Emission Concrete Failure
Beam-II Stress 16
5
14
4
12 Stress (Mpa)
3 Debonding
Intensity (V)
10
2
8
1
6
0
4
-1 2
-2 0
0 20 40 60 80 100 120 140
Time (Sec)
Fig. 13.30 TL emissions at debonding and failure [52]
signal of 57 mV was obtained at a stress value of about 16 MPa and at time 105 s.
The second emission represents complete peel-off of the composite from the
concrete surface about 15 s before the beam failed. The beam finally failed at
time 125 s at a peak stress of 16 MPa due to the loss of concrete-composite action
caused by the retrofit debonding. The sensor is thus capable of detecting the
debonding failure mode and was also able to provide distributed sensing through
the composite [52]. Compared to conventional point sensors such as strain gages,
the optical fiber sensors promise to offer superior advantage as a distributed sensor.
In addition, it is not possible to install strain gages within the adhesive layer as it
will be detrimental to the concrete-composite action.
13.6 Conclusion
A triboluminescence-based sensor system comprising highly efficient tribolumi-

nescent materials could allow simple, real time monitoring of both the magnitude
and location of damage with minimal parasitic influence to the host structure. The
first approach investigated in developing a cementitious composite system with a
triboluminescence-based damage detection system involved the direct dispersion of
ZnS:Mn crystals in a mortar system. Test results indicate that there tends to be an
increase in the triboluminescent response (signal) of the TLCOS samples with
increasing ZnS:Mn concentration levels. Triboluminescent signals were detected
at concentration levels as low as 2.5 and 5 % although not all the samples tested at
these low concentration levels produced a triboluminescent response on impact.
Triboluminescent responses were however detected for all samples tested at con-
centration levels of 7.5 % and higher. Consequently, a ZnS:Mn percolation level of
7.5 % is required for an effective TLCOS.
There was a reduction in the compressive strength of the pristine cementitious
patch from about 29.5 MPa to about 26 MPa at ZnS:Mn concentrations of 5 and
7.5 %. Enhancement of the compressive strength was however observed at ZnS:Mn
concentration levels of 10 % and 15 % with mean values of about 31 MPa and
30 MPa, respectively, compared to a mean value of about 29 MPa for the mortar
system with no ZnS:Mn content. The compressive strength (28 MPa) measured at
20 % ZnS:Mn concentration level is still greater than those at 5 and 7.5 % concen-
tration levels. There was a reduction of about 29.31 % in the compressive strength
of the TLCOS with 25 % ZnS:Mn concentration compared to the control. Conse-
quently, a ZnS:Mn concentration level of 7.5 % is recommended for a TLCOS with
consistent damage sensing capability and minimal parasitic weight effect.
The in-situ triboluminescent optical fiber (ITOF) sensor was developed to
overcome most of the challenges associated with the direct dispersion of TL
materials in the cementitious composite. Such undesirable effects include the
parasitic weight effect that can adversely affect the properties and performance of
the host material. Real time failure detection in unreinforced cementitious compos-
ites has been successfully demonstrated. The integrated ITOF sensor gave no false
alarm and only generated TL signals at the instance of brittle failure. The results
show that the farther away an ITOF sensor is located from the point of failure
(crack) initiation, the less accurately the sensor can sense the stress level respon-
sible for failure. The TL emission profile also depends on the position of the
integrated ITOF sensor relatively to the neutral axis of the beam. The TL emission
profile analysis promises to be a viable tool for understanding crack initiation and
propagation in CC.
In addition, real time damage monitoring in reinforced cementitious composite

with integrated ITOF sensors has been successfully demonstrated. Unlike the
unreinforced CC systems that exhibited brittle failure, the reinforced CC system
exhibited a more ductile behavior characterized by multiple cracks formation. The
ITOF sensor was able to detect all the cracks that propagated across the sensitized
(coated) sections of the integrated sensor. The ITOF sensor was also able to provide
warnings about crack propagation and crack opening before structural failure. The
superiority of the ITOF sensor over conventional strain gages was also
demonstrated.
Finally, the ITOF sensor was used to detect failure in the FRP retrofit system
used to strengthened precracked reinforced concrete beams. Real-time failure
detection (debonding failure mode) was demonstrated in beams retofitted with
FRP composite under flexure loading. The ITOF sensor successfully detected
debonding of the FRP retrofit and in future work, the ITOF sensor will be used to
determine the location of the damage by introducing patterned layups. The ITOF
sensor holds great promise as a viable composite retrofit health monitoring system
with distributed damage sensing capability when fully developed.
Acknowledgment The authors gratefully acknowledge the support of the National Science
Foundation (NSF) under NSF Award (CMMI-0969413).
References
1. Measures, R. M. (2001). Structural monitoring with fiber optic technology (pp. 1–50). San
Diego: Academic Press. ISBN 0124874304.
2. Hota, G. V. S., & Hota, S. R. V. (2002). Advances in fibre-reinforced polymer composite
bridge decks. Progress in Structural Engineering and Materials, 4, 161–168.
3. Sloan, J. (2008). Composites fly high around the world, in Sourcebook 2008: The international
guide to products and services for the composites industry (pp. 6–45). Cincinnati: Gardner
Publications Inc.
4. Leung, C. K. Y., Elvin, N., Olson, N., Morse, T. F., & He, Y. (2000). A novel distributed
optical crack sensor for concrete structures. Engineering Fracture Mechanics, 65, 133–148.
5. Carpinteri, A., & Lacidogna, G. (2006). Damage monitoring of an historical masonry building
by the acoustic emission technique. Materials and Structures, 39, 161–167.
6. Leung, C. K. Y. (2001). Fiber optic sensors in concrete: The future? NDT&E International, 34,
85–94.
7. Zhou, G. A., & Sim, L. M. (2002). Damage detection and assessment in fibre-reinforced
composite structures with embedded fibre optic sensors—Review. Smart Materials and Struc-
tures, 11(6), 925–939.
8. Chong, K. P., Carino, N. J., & Washer, G. (2003). Health monitoring of civil infrastructures.
Smart Materials and Structures, 12, 483–493.
131, 1407–1418.
10. Transportation Research Board. (2006). Control of cracking in concrete-state of the art.
Washington, DC: Transportation Research Board.
11. Mehta, P. K., & Gerwick, B. C., Jr. (1982). Cracking-corrosion interaction in concrete exposed
to marine environment. Concrete International, 4, 45–51.
12. Mehta, P. K., & Monteiro, P. J. M. (1993). Concrete, structure, properties, and materials.
Englewood Cliffs, NJ: Prentice-Hall.
13. Portland Cement Association. (2009). Frequently asked questions. Cement and concrete
technology [Online]. Portland. Retrieved Feburuary 2, 2010, from www.cement.org/tech/
faq_cracking.asp
14. American Concrete Institute. (2001). Control of cracking in concrete structures. ACI Commit-
tee 224r-01.
15. Marcotte, T. D., & Hansson, C. (2003). The influence of silica fume on the corrosion resistance
of steel in high-performance concrete exposed to simulated sea water. Journal of Materials
Science, 38, 4765–4776.
16. Yoon, D. J. W. J., Weiss, W. J., & Shah, S. P. (2000). Assessing corrosion damage in
reinforced concrete beams using acoustic emission. Journal of Engineering Mechanics, 126,
273–283.
17. Attanayaku, U., Duyar, O., & Aktan, H. M. (2004). Early-age cracking potential of slipformed
concrete bridge barriers. In 83rd Annual Meeting Of The Transportation Research Board.
Washington, DC: Transportation Research Board.
18. Farrar, C. R., & Keith, W. (2007). An introduction to structural health monitoring. Philosoph-
ical Transactions of the Royal Society A, 365, 303–315.
19. Banks, H. T., Inman, D. J., Leo, D. J., & Wang, Y. (1996). An experimentally validated
damage detection theory in smart structures. Journal of Sound and Vibration, 191, 859–880.
20. Carden, E. P., & Fanning, P. (2004). Vibration based condition monitoring: A review.
21. Ko, J. M., & Ni, Y. Q. (2005). Technology developments in structural health monitoring of
large-scale bridges. Engineering Structures, 27, 1715–1725.
22. Sohn, H., Dzwonczyk, M., Straser, E. G., Law, K. H., Meng, T. & Kiremidjian, A. S. (1998).
Adaptive modeling of environmental effects in modal parameters for damage detection in civil
structures. Smart Systems For Bridges, Structures, And Highways. SPIE (pp. 127–138).
23. Wan, K. T., & Leung, C. K. Y. (2007). Applications of a distributed fiber optic crack sensor for
concrete structures. Sensors and Actuators A, 135, 458–464.
Measurement of triboluminescence and proton half brightness dose for ZnS:Mn. IEEE Trans-
25. Kawaguchi, Y. (1998). Fractoluminescence spectra in crystalline quartz. Japanese Journal of
Applied Physics Part 1, 37, 1892–1896.
of Luminescence, 75, 295–299.
27. Chandra, B. P., Elyas, M., & Majumdar, B. (1982). Dislocation models of mechanolumi-
nescence in [gamma]- and X-irradiated alkali halides crystals. Solid State Communications,
42, 753–757.
et al. (2006). Experimental evidence of triboluminescence induced by hypervelocity impact.
854–870.
30. Walton, A. J. (1977). Triboluminesence. Advances in Physics, 26, 887–948.
31. Chandra, B. P. (1998). Luminescence of solids. New York: Plenum Press.
surfaces. Journal of Physics D: Applied Physics, 37, 2175–2180.
33. Sage, I., Badcock, R. A., Humberstone, L., Geddes, N., Kemp, M., Bishop, S., et al. (1999b).
Squeezing light out of crystals: Triboluminescent sensors. Smart Structures and Materials
1999: Smart Materials Technologies, Newport Beach, CA. SPIE (pp. 169–179).
35. Chandra, B. P., Baghel, R. N., & Chandra, V. K. (2010). Mechanoluminescenct glow curve of
Zns:Mn. Chalcogenide Letters, 7, 1–9.
Physical Review B, 21, 816–826.
37. Kim, J. S., Kwon, Y. N., Shin, N., & Sohn, K. S. (2007). Mechanoluminescent SrAl2O4: Eu,
Dy phosphor for use in visualization of quasidynamic crack propagation. Applied Physics
Letters, 90, 241916.
black composites: Embedded triboluminescent structural damage sensors. Smart Materials &
Structures, 10, 332–337.
39. Sohn, K. S., Seo, S. Y., Kwon, Y. N., & Park, H. D. (2002). Direct observation of crack tip
stress field using the mechanoluminescence of SrAl2O4:(Eu, Dy, Nd). Journal of the American
40. Xu, C. N., Watanabe, T., Akiyama, M., & Zheng, X. G. (1999). Preparation and characteristics
of highly triboluminescent ZnS film. Materials Research Bulletin, 34, 1491–1500.
41. Xu, C. N., Zheng, X. G., Akiyama, M., Nonaka, K., & Watanabe, T. (2000). Dynamic
visualization of stress distribution by mechanoluminescence image. Applied Physics Letters,
76, 179–181.
Triboluminescent damage sensors. Smart Materials & Structures, 8, 504–510.
43. Olawale, D. O., Sullivan, G., Dickens, T., Tsalickis, S., Okoli, O., Sobanjo, J. O., et al. (2012).
Development of a triboluminescence-based sensor system for concrete structures. Structural
Health Monitoring, 11, 139–147.
44. Bourhill, G., Pålsson, L. O., Samuel, I. D. W., Sage, I. C., Oswald, I. D. H., & Duignan, J. P.
(2001). The solid-state photoluminescent quantum yield of triboluminescent materials. Chem-
ical Physics Letters, 336, 234–241.
45. Chandra, B. P., Singh, S., Ojha, B., & Shrivastava, R. G. (1996). Mobile interstitial model and
mobile electron model of mechano-induced luminescence in coloured alkali halide crystals.
Pramana, 46, 127–143.
46. Sweeting, L. M. (1990). Spectroscopic characterization of minerals and their surfaces. In L. M.
Coyne, S. W. S. McKeever, & D. F. Blake (Eds.), Acs Symposium Series 415 (pp. 245–260).
Washington, DC: American Chemical Society.
47. Chandra, B. P., Xu, C. N., Yamada, H., & Zheng, X. G. (2010). Luminescence induced by
elastic deformation of Zns:Mn nanoparticles. Journal of Luminescence, 130, 442–450.
48. Olawale, D. O., Sullivan, W., Dickens, T., Okoli, O., & Wang, B. (2011). Mimicking the
human nervous system with a triboluminescence sensory receptor for the structural health
monitoring of composite structures. Proc. SPIE 7981, Sensors and Smart Structures Technol-
ogies for Civil, Mechanical, and Aerospace Systems, San Diego. SPIE (pp. 798125-798125).
49. Olawale, D. (2013). In-situ triboluminescent optical fiber sensor for real-time damage mon-
itoring in cementitious composites. PhD Electronic Theses, Treatises and Dissertations, Flor-
ida State University.
52. Joshi, K., Breaux Frketic, J., Olawale, D., & Dickens, T. (2015). Damage monitoring Of CFRP
retrofit using optical fiber sensors. 943520–943520-11.
53. Zhou, H., Attard, T. L., Zhao, B., Yu, J., Lu, W., & Tong, L. (2013). Experimental study of
retrofitted reinforced concrete shear wall and concrete-encased steel girders using a new
carbonflex composite for damage stabilization. Engineering Failure Analysis, 35, 219–233.
54. ACI Committee 440. (2002). Guide for the design and construction of externally bonded FRP
systems for strengthening concrete structures.
55. TRB. (2010). Recommended guide specification for the design of externally bonded FRP
systems for repair and strengthening of concrete bridge elements. Transportation Research
Board.
56. ASTM. (2009). Standard test method for flexural strength of concrete (using simple beam with
third point loading).
Chapter 14
Triboluminescence at Speeds Greater
than 100 m/s
Ross S. Fontenot, William A. Hollerman, and Noah P. Bergeron
14.1 Introduction
A phosphor is an inorganic crystalline material composed of a host insulating or

semiconducting material to which an impurity (or dopant) is introduced. It is
important to note that phosphors are considered a subset of luminescent materials,
such as bioluminescent insects and naturally occurring minerals. The dopant added
to the host introduces metastable states between the conduction and valence
band, thereby narrowing the band gap and better permitting light-producing
transitions. As a result, phosphors are materials that, when excited, emit “cold,”
or non-thermally induced, light. This production of light is known as luminescence
and is commonly encountered in such applications as cathode ray tube televisions
and fluorescent light bulbs. Luminescence is generated by first exciting the elec-
trons of the dopant in the phosphor materials to a higher energy state. As the
electrons return to ground state, they emit photons, thereby generating
luminescence [1].
Luminescence is categorized by excitation source and emission lifetime. Many
types of excitation sources exist, such as photons (photoluminescence), electrons
(cathodoluminescence), and ionizing radiation (radioluminescence). Of particular
interest to this research is luminescence induced by mechanical stress, which has
alternately been referred to as mechanoluminescence (ML). This phenomenon can
R.S. Fontenot (*)

Carderock Division, Code 632, Naval Surface Warfare Center,
West Bethesda, MD 20817, USA
e-mail: rsfontenot@hotmail.com
W.A. Hollerman • N.P. Bergeron
Department of Physics, University of Louisiana, P.O. Box 44210, Lafayette, LA 70504, USA

Fig. 14.1 Triboluminescent spectrum of a real Wint-O-Green® Lifesavers [9]
be further subdivided into triboluminescence (TL), luminescence induced

by crystal fracture, or deformation luminescence (DL), luminescence induced by
mechanical deformation of crystal structure without fragmentation. DL may also
be elastic or plastic in nature. DL has been observed for materials such as colored
alkali halides, quartz, semiconductors, polymer crystals, and rubber. DL tends to
have a much dimmer luminescent intensity than FL, reducing its potential appli-
cation [2–8].
TL is a commonly encountered phenomenon, with approximately 50 % of
known crystals exhibiting such tendencies [2]. The most commonly known exam-
ple of a TL material is real Wint-O-Green® Lifesavers. If they are crushed in the
mouth or in a darkened room, it is possible to observe a blue flash of light from the
crushing of the crystals. The spectrum of Wint-O-Green® Lifesavers emitting TL as
observed by Sweeting, et al. is shown in Fig. 14.1. The 400–500 nm broad band
emission is from the methyl salicylate oil, while the line emissions near 350 nm are
from sucrose sparking in a nitrogen environment [9].
TL has many different mechanisms that are not completely understood. While
up to 50 % of all crystals may emit TL, the necessary qualities a crystal must posses
to TL are not agreed upon, and spectral evidence exists that suggests multiple
mechanisms may be important in many crystals [10–13].
If stressed, a triboluminescent crystal produces local regions with charge imbal-
ances. Several mechanisms may then occur to produce light. It is possible for the
charge imbalance to be sufficient to cause a dielectric breakdown of the surrounding
gas, producing light. It is also possible for the emission of light from a dielectric
breakdown to excite the triboluminescent material. A charge imbalance can also
generate electroluminescence. Other potential sources of triboluminescence exist
that are not well understood or researched. It is possible to state that abrupt charge
separation occurs during triboluminescent emission, but the exact mechanism is
material dependent and not well understood.
Impacts occur over wide ranges of masses and velocities. To properly under-
stand and characterize impacts, a certain classification must be employed. Due to
14 Triboluminescence at Speeds Greater than 100 m/s 413
30
25
20
PMT Output (Volts)
15
10
-5
-2 0 2 4 6 8 10
Time (ms)
Fig. 14.2 Typical photomultiplier detector response for ZnS:Mn
the lack of universal projectiles in impact studies, the most common classifications
are in terms of impact velocity or specific energy. As such, for the purpose of this
work, all impacts occurring at 100 m/s or less will be considered low velocity,
impacts between 100 m/s and 1 km/s will be considered mesovelocity, and impacts
occurring at greater than 1 km/s will be considered hypervelocity.
The existence of materials that produce light when subjected to mechanical
stress has indicated the possibility of constructing an impact sensor with tribolu-
minescent phosphors as the active element. Beginning in 2003, research was
conducted to induce TL from a low specific energy impact [14, 15]. To this end,
several drop towers were constructed to investigate low velocity TL. In these
experiments, a 130 g steel ball was dropped on loose powder, and later, on
aluminum cards coated with phosphor. To vary the impact velocity, the height at
which the ball was dropped was changed [14, 15].
A typical low-velocity triboluminescent response from the ZnS:Mn powder is
shown in Fig. 14.2 for a drop height of 1 m [15]. A series of drops would start with
one from the maximum height of 1 m. Subsequent drops were done from progres-
sively decreasing heights until no light was observed. Notice in Fig. 14.2 that
triboluminescent decay appears to follow the standard exponential decay curve of
luminescence [1, 16]. The orange color inset shown in Fig. 14.2 is a picture of
actual triboluminescent light produced when a 130 g steel ball bearing was dropped
on a small quantity of ZnS:Mn powder.
Figure 14.3 shows the variation in the triboluminescent emission characterized
by the output voltage from a photomultiplier detector as a function of the drop
kinetic energy [15]. The measurement uncertainty is approximately 5 %, as shown
Fig. 14.3 Plot of the peak TL intensity as a function of impact kinetic energy for ZnS:Mn
powder [15]
by the error bars. The intensity of the TL response appears to be a function of

impact energy with two regions of interest. The first region is in the energy range
below 0.2 J. Here, the production of TL light appears to have a cut-on threshold of
approximately 0.05 J. Above this threshold, the projectile has sufficient energy to
break ZnS:Mn crystals, producing light. The intensity increases rapidly until about
0.2 J. The second region begins at 0.2 J and appears to be more of a saturation state:
here, the slope is shallow, which indicates that the sample is less sensitive to impact
energy. This is almost certainly due to the ball breaking as many crystals as it can in
the area of impact. Increasing the energy above this threshold cannot generate more
light because there are no more crystals in the impact area. From this work, it is
possible to discriminate impact energy/velocity based on peak light intensity to
within a factor of about 2 for energies greater than 0.2 J.
Experiments were also conducted at ORNL in the summer of 2005 to investigate
TL from mesovelocity impacts. Experiments were conducted with a 9 mm bore
gun, and consisted of firing 10.3 g projectiles at velocities from 114 to 275 m/s.
Circumstantial evidence of TL was observed, but no definite results were
obtained [17].
Research into electrical signals produced by impacts on rock has demonstrated
evidence of TL. In these experiments, 0.13–1.30 g steel spheres were fired onto
granite at velocities up to 100 m/s. While it was the production of electrical current
that was of most interest to the researchers, it was also found that impacts on
granite produced TL [18]. This research implies it is possible that granite in rock
during earthquakes and meteoroid impacts could produce TL. Astronomers have
observed impacts on the lunar surface that have produced light that cannot be
adequately reproduced by models only incorporating thermal effects [19]. As a
result, there is interest in studying the lunar environment for triboluminescent
materials.
14.2 Hypervelocity Guns
Guns used to conduct research involving hypervelocity impacts were first devel-
oped in the 1950s. During that time, the primary goal of such devices was to
simulate flight environments of high-speed missiles [20]. Applications of such
devices have grown from this initial requirement and now also include determining
equations of state for materials. These guns operate in a manner somewhat similar
to conventional firearms in that gas compressed to high pressure and used to
accelerate a projectile. However, conventional firearms burn energetic
nitrocellulose-based molecules to generate the gas pressure used to accelerate the
impacting projectile. Hypervelocity impact guns usually compress low-molecular-
mass gases such as helium or hydrogen to accelerate the impacting projectile. Low-
molecular-mass gases are more compressible and are able to generate higher
pressures than gases with larger, heavier molecules [17].
Hypervelocity impact guns are usually classified according to the number of
stages used in their construction. In a single-stage gun, gas is compressed by
conventional pumps in a high-pressure chamber and released by quick release
valves to accelerate the projectile to impact. These guns are typically capable of
maximum velocities in the 1–2 km/s. For impact velocities greater than 2 km/s,
two-stage light gas guns are used. Figure 14.4 is a diagram of the two-stage light gas
gun at NASA MSFC and is representative of the type [17].
Two-stage guns fire an initial projectile by either compressed gas or combusti-
bles to act as a piston to compress a second stage chamber filled with low-atomic-
mass gas. Once the gas reaches a certain pressure, it ruptures a seal, allowing the
highly compressed second-stage gas to accelerate the final projectile. This proce-
dure allows for a greater pressure to be achieved than by the pumps used in single
stage guns. Two-stage guns can typically achieve velocities of up to 15 km/s. In
three-stage guns, the final projectile of a two-stage gun is incident upon a third
projectile of lesser mass. In this collision, the third projectile, by benefit of its lesser
Fig. 14.4 Diagram of NASA MSFC two-stage hypervelocity impact gun

mass, is accelerated to a higher velocity. The maximum velocity of these guns is

highly variable, with typical maximums twice those of the two stage guns, approx-
imately 20–30 km/s. This is also the theoretical maximum for elastic collisions.
Other, more exotic designs are in use, but are beyond the scope of this
discussion [20].
14.3 Experimental Design and Materials
14.3.1 Phosphors
The primary phosphor investigated for the hypervelocity research was ZnS:Mn
purchased from Phosphor Technology, Limited, of Hertfordshire, Great Britain.
This phosphor was chosen because it is one of the most efficient, commonly
available triboluminescent phosphors known, with loose crystals giving off light
when subject to relatively small stresses, such as being scratched by a person’s
fingernail [3, 5, 14].
ZnS:Mn has a hexagonal crystal structure with an average grain size of less than
10 μm [21]. The loose crystal form of this phosphor is a fine pink powder. In this
phosphor, manganese dopant replaces some of the sulfur atoms, with dopant
concentrations estimated to be 4 % for the sample used in this research [22]. As
can be seen from Fig. 14.5, the emission spectra of ZnS:Mn is a broad emission
peak with a 585 nm centroid and a full width at half maximum (FWHM) of 65 nm
[17]. It is also useful to note that the spectrum appears to be independent of
excitation source [17].
Previous work has been conducted to investigate the emission spectra of ZnS:
Mn with different excitation sources, and the results of these experiments are shown
in Fig. 14.5. For the UV excitation, 254 nm light was incident on an aluminum plate
coated with ZnS:Mn. In the proton excited spectrum, an aluminum plate coated
with ZnS:Mn was irradiated by 3 MeV protons, while the spectrum was being
recorded; the same technique was used to record the spectrum of pressed ZnS:Mn
pellets excited by 20 keV electrons. For the triboluminescent spectrum, a drop
tower with a clear polycarbonate base had loose powder placed in it, and a spectrum
was recorded during impact. The photoluminescent (UV) spectrum was taken as a
baseline comparison since this is a standard technique to investigate phosphor
emissions. The proton and electron spectra were taken during tests to determine
the radiation hardness of ZnS:Mn relative to its suitability for use in hostile
environment sensing [23–25]. The triboluminescent data was collected to confirm
that the ZnS:Mn emission spectra was independent of the excitation source.
To further establish and validate the production of TL, a second phosphor was
used in this study, ZnS:Cu. This phosphor was produced from Phosphor Technol-
ogy located in Great Britain. It is a green phosphor with a centroid peak at 530 nm
and a FWHM of 85 nm; its emission spectrum is shown in Fig. 14.6.
800 350
700 UV Light 300 Protons

600 (3 MeV)
250
Light Intensity
500
200
400
150
300
200 100
100 50
0 0
500 550 600 650 700 500 550 600 650 700
140 2500
4.5 m/s
TL Electrons
120
Three
2000 (20 keV)

100 measured
drop
80 1500
velocities
3.9 m/s
60 1000
3.2 m/s
40
500
20
0 0
500 550 600 650 700 500 550 600 650 700
Fig. 14.5 Emission spectra for ZnS:Mn for different excitation sources: 355 nm ultraviolet light,
3 MeV protons, TL at three different impact velocities, and 20 keV electrons
Fig. 14.6 Emission spectrum of ZnS:Cu

14.3.2 Target Design and Construction
Targets were machined from 0.5 cm thick aluminum at the UL Lafayette Physics
Department machine shop into plates 29 cm long and 10 cm wide. A paint
composed of ZnS:Mn and binder was applied to a 10 cm 10 cm area on one
side of each plate. An example of a painted target plate in visible and UV light can
be found in Fig. 14.7. The pink ZnS:Mn paint shown in the visible light picture
corresponds to the bright yellow fluorescence illuminated with ultraviolet light.
The ZnS:Mn paint presented a unique problem. Initially, it consisted of a ZnS:
Mn and Resbond™ Cotronics 793 paint with 25 % phosphor by mass. This paint
was sprayed on to the aluminum plate using a commercial airbrush. This binder was
chosen for its high temperature survivability and ease of application. However,
Cotronics 793 is a ceramic and is quite brittle [26]. Initial hypervelocity impacts
with this binder caused the entire phosphor/binder to shatter and completely
delaminate from the aluminum plates as can be seen from Fig. 14.8.
Further investigations concluded that poly (phenyl methyl) siloxane (PPMS)
gave the toughest and most wear resistant paint and was sufficiently resistant to
hypervelocity impacts [27]. PPMS is cured at 160 C for 20 min. Curing causes the
PPMS to cross-link and gain mechanical strength. Using a ZnS:Mn and PPMS paint
(with 25 % phosphor), the loss of coating would only occur at the site of impact for
roughly twice the crater diameter as shown in Fig. 14.9.
It is important to note that the type of binder, paint composition, and application
technique proved to be more influential in these studies than has been the case in
previous non-impact related phosphor sensor research. In order to emit TL light,
crystals of the phosphor must be broken or stressed. A binder that is too soft will
deform the binder without breaking phosphor crystals and dissipate energy of the
impact throughout the paint. Additionally, a binder that is too brittle, such as
Resbond, will break apart the binder without breaking or stressing phosphor
crystals. If the paint had too much binder or if it was applied at an air gun pressure
Fig. 14.7 Picture of an

example painted target plate
taken in visible and
ultraviolet light
Fig. 14.8 Picture of impact chamber excited by UV light with Resbond binder
Fig. 14.9 Picture (UV light) of the painted side (reverse from impact) of an aluminum target plate
showing the effects of the projectile and sabot strikes
less than 0.25 GPa, a binder rich paint was produced that was too soft. If the paint
was too binder rich and soft, the polymer would deform at the impact site without
breaking phosphor crystals. This greatly decreased the emitted triboluminescence of
the paint. While not as relevant to other luminescence studies, it is crucial to impact
related research that the coating have proper mechanical properties Fig. 14.9.
14.3.3 Gun Description
The light gas gun located at NASA MSFC in Huntsville, Alabama, was used for this
research at the gracious invitation of Dr. Dave Edwards [27]. This gun is a
two-stage device constructed by Southwest Research, Incorporated. The first
stage is constructed from a Sako .22-250 bolt action rifle. Instead of the typical lead
copper-jacketed projectile fired by this rifle, a high-density polyethylene cylinder is
the initial projectile. The cylinder travels down a section of 0.57 cm diameter barrel
until it encounters a second barrel that narrows progressively. As the cylinder travels
down these two barrel sections, it compresses hydrogen or helium gas in the high
pressure chamber of the gun. Both hydrogen and helium have low atomic masses,
allowing them to be compressed to higher pressures than is possible with gasses
composed of elements with higher atomic masses. A 76.2 μm Mylar burst disk holds
the high-pressure gas until it ruptures. At that point, the expansion of light gas then
drives the impact projectile down the third and final barrel section. The impact
projectile is a 1 mm aluminum sphere positioned inside a plastic sabot that provides a
gas seal against the bore of the second stage. The projectile then enters the flight tube
of the gun. The gas pressure of the flight tube separates the sabot and the projectile
travels downrange to the impact chamber. The MSFC gun has no system to stop or
catch the sabot in the flight tube and instead has a scatter shield as part of the target
holder system. This scatter shield is designed to catch the sabots and prevent them
from entering the target holder. A picture of the NASA MSFC gun can be seen in
Fig. 14.10 and a diagram of the gun is also shown in Fig. 14.4.
Prior to this research, the instrumentation in the MSFC gun consisted of two
silicon light detectors used to determine projectile velocity to within 10 % of the
actual value. In addition, the backplane of the impact chamber was constructed of
clear Plexiglas and lacked any feed throughs. As such, a new backplane was
required. Specifications for the backplane were: use of multiple BNC and fiber
optic feed throughs, ability to keep out ambient light, and vacuum tight to about
10 Torr. Figure 14.11 shows pictures of the original Plexiglas and revised
instrumented backplanes.
Several other laboratories associated with hypervelocity impact studies provided
invaluable technical assistance in the proper setup of the experiment and analysis of
data. AEDC gave technical assistance regarding the presence and characteristics of
the muzzle flash, and Sandia National Laboratories and NASA Johnson WSTF
provided expertise with timing, triggering, and instrumentation issues by demon-
strating their light detection and imaging systems.
Fig. 14.10 NASA MSFC hypervelocity light gas gun

Fig. 14.11 Pictures of the (a) new instrumented backplane and (b) original MSFC gun backplane
14.3.4 Target Holder and Detector Setup
The target holder system consists of an aluminum plate that has two 6.4 mm
diameter threaded rods spaced 28 cm apart. Stop plates, targets, and the scatter
shield are attached to the holder through the threaded rods. During the initial shots,
there was a tendency for both the sabots and the projectile to impact the scatter
shield. To prevent this occurrence, the scatter shield was removed until several of
the gun components could be re-machined, preventing unwanted impact of the
projectile on the scatter shield.
To capture a triboluminescent event, silicon detectors were placed inside the
impact chamber, with their output running into amplifiers before the signal was
displayed on an oscilloscope. This presented several challenges to the acquisition
of a signal that could be positively identified as a TL event. For the first test series,
conducted in April 2004, two detectors were placed 2.5 cm behind the target plate,
with one detector directly behind the coated section and the second facing bare
aluminum. In this setup, the projectile would impact the bare aluminum, generat-
ing a shock wave that would travel through the aluminum and stress the ZnS:Mn
coating on the target reverse of the impact. This technique was chosen to eliminate
the thermal light production caused by the bare aluminum impact. The lower
detector functioned as a trigger for the detector facing the coating, and served to
record a background signal. Figure 14.12 demonstrates that when the data from
this test series was analyzed, it was found that there was no measurable difference
in signal between the lower detector and the one facing the coating. Upon
conferring with AEDC, it was discovered that light gas guns produced significant
blackbody radiation from their muzzles when fired. This fact made eliminating the
muzzle flash a crucial step in attempting to measure TL inside the impact
chamber.
Fig. 14.12 Sample data from April 2004 test series
Fig. 14.13 Single-shielded detector mounted to a holder assembly for the MSFC gun [27]
A solution to properly shield the silicon detectors from muzzle flash was found
in a local hardware store. The silicon detectors were placed into 5 cm diameter
holders. It was found that soft PVC pipe adapters would fit around the detectors to
shield them from muzzle flash. An example of a single shielded detector mounted to
a holder assembly is shown in Fig. 14.13. The second unshielded detector was used
to trigger the oscilloscope. Figure 14.14 shows an example of a typical impact and
how the detectors are mounted to the holder [27]. A validation of the shielded
detector system occurred during the August 2005 test series, when a shot was fired
that impacted the aluminum plate above the shielded area of the detector.
Fig. 14.14 Pictures of an example impact showing (a) the front unpainted side of plate with center
projectile and two surrounding sabot hits in visible light, (b) back painted side of plate showing
center projectile and two surrounding sabot hits in ultraviolet light, (c) back side of plate showing
two silicon detectors looking at the paint behind the impact with rubber shield, (d) and inside of the
silicon detectors and black rubber shield showing residue paint powder after the hypervelocity
impact
This impact generated no detectable light for the shielded detector, while the
unshielded detector functioned normally and recorded the typical signature of the
muzzle flash.
Once the proper target, target holder, and detector designs were established, it
was then necessary to determine proper settings for the amplifiers and oscilloscope.
This was conducted through trial and error until captures of the triboluminescent
event were recorded that did not exceed the detection range of input or output
devices. While work commenced in April 2004, it was not until December 2004
when the experimental setup was finalized. A schematic of the experimental setup
is given in Fig. 14.15. It is important to note that the setup was chosen for several
reasons. Photodiodes were wired to the signal amplifiers as the TL light was
insufficiently bright to be recorded by the photodiodes without amplification.
Photomultiplier tubes (PMTs) were not able operate in the vacuum required by
the impact chamber, and by sending the muzzle facing detector signal to the
oscilloscope, it was possible to use it as a trigger.
To properly characterize the muzzle flash of the MSFC gun, a fiber-optic feed
through was placed on the backplane of the impact chamber in March 2005. The
fiber was routed into an Ocean Optics S2000 spectrometer and the spectrum of the
muzzle flash was taken. Results of that experiment showed that the muzzle flash
had a significant contribution of light at 585 nm, which is the same emission
wavelength as ZnS:Mn. Muzzle flash spectra of the MSFC gun is given in
Fig. 14.16.
Target
Phosphor Coating
Shielded Photodiode
Projectile Melles Griot

Large Dynamic
Velocity Range Amplifier
Digital Recording
2-6 km /s In Out
Oscilloscope
Ch1 Ch2 Trig
Melles Griot
Large Dynamic
Range Amplifier
In Out
Unshielded Photodiode
Fig. 14.15 Schematic of experimental design
Hydrogen Helium
80
40
60
40
Intensity
20
20
0
0
-20 -20
200 400 600 800 200 400 600 800
Fig. 14.16 Sample spectrum of NASA MSFC muzzle flash for hydrogen and helium
The peak wavelength for the muzzle flash spectra can be calculated using Wien’s
displacement law:
λmax ¼ b=T; ð14:1Þ
where λmax is the peak wavelength in meters, b is Wien’s displacement constant of

value 2.8977685 106 K—nm, and T is the temperature in Kelvin. Using
Eq. (14.1), the temperature of the muzzle flash was estimated to be 4500 K.
Considering that nonluminescent light was produced by both the gun’s muzzle
flash and the impact, a proper shielding system became a necessity; otherwise it
would not be possible to separate the muzzle flash from the TL.
Also, prior research in orbital debris characterization has indicated that there is a
nontrivial production of light due solely to the hypervelocity impact [17]. To
minimize the detection of this light by the phosphor-facing detector, all targets
were mounted such that the coating was opposite of the impact, and no detector
faced the region of impact. In addition, the rubber shield of the photodetector served
to block any potentially reflected light from the impact. To prove this concept,
several shots were taken against targets with no phosphor coating to serve as
background data. In all of these shots, no light above baseline was measured by
the shielded detector. Finally, there were two shots wherein the projectile impacted
a phosphor plate above the target area. For these two shots, no light was detected
above the baseline by the shielded detector. As such, no light produced by the
impact was measured. These shots served to validate the integrity of the shielding
system [27, 28].
The time duration intensities of each of the three sources of light during these
impacts is important in collecting and properly analyzing results. For each of the
sources of light, TL, bare aluminum impact, and muzzle flash, each had a unique
time duration. The muzzle flash was measured to have a time duration greater than
750 μs. The bare (no phosphor) impact had a time duration on the order of
0.1–0.01 μs. The fluorescent decay time of ZnS:Mn is known to be approximately
300 μs, a value between those of the aluminum on aluminum impact and muzzle
flash [27–31]. By choosing a TL phosphor with fluorescent decay time different
than the other sources of light, it was possible to state with greater certainty the
presence or absence of TL.
14.4 Results
14.4.1 Hypervelocity Impact-Induced Triboluminescence

Detection
In the course of the seven experiments conducted in support of this research,

49 hypervelocity shots were completed. Table 14.1 shows the breakdown of these
49 shots according to their outcome.
Of the 14 impacts where data was collected, 10 had their entire TL light
production recorded in both intensity and time duration with the corresponding
impact velocity. Table 14.2 categorizes each of these shots by date, phosphor used,
impact velocity, measured fluorescent decay time, and integrated TL light produc-
tion. If the statistics of the recorded data did not permit a reliable measurement of
Table 14.1 Number of Purpose Number

hypervelocity shots
Setup 8
completed at MSFC
Misses 16
Attempted TL spectrum 2
Background 5
Muzzle flash spectrum 4
TL data acquired 14
Total 49
Table 14.2 Velocity, decay time, and integrated TL light production of selected shots
Fluorescent Integrated light
Date Phosphor Velocity (km/s) decay time (μs) (arbitrary units)
Dec 04 ZnS:Mn 5.55 0.55 336 5.835
4.76 0.48 274 11.65
5.55 0.55 315 1.923a
Mar 05 4.76 0.48 Insufficient statistics 0.714
Aug 05 ZnS:Mn 5.55 0.55 175 4.348
ZnS:Cu Insufficient statistics 0.698
0.485
Jan 06 ZnS:Mn 3.33 0.33 143 1.311
Insufficient statistics 0.828
4.76 0.48 508 1.711
Data taken with 589 10 nm filter
a
the decay time, this is noted as “Insufficent Statistics” in the table. Data shown in
Table 14.2 are those shots where impact velocity and integrated light production
was successfully collected.
From March 2004 to December 2005, the focus of the research was to properly
understand the function and characteristics of the light gas gun. By so doing, it was
then possible to instrument and analyze data from impacts. In December 2005, a
series of preliminary measurements was completed to determine if tribolumines-
cent light could be detected as a result of a hypervelocity impact. A ZnS:Mn and
PPMS painted plate with two light detectors (one shielded and one unshielded)
were positioned in the impact chamber of the MSFC gun as shown in Figs. 14.14
and 14.15. Results from shot one of this test series are shown in Fig. 14.17. The
projectile velocity for shot one was approximately 5.6 0.5 km/s. The light inten-
sity from both the unshielded and shielded photodetectors saturates the identical
amplifiers. However, the light intensity from the unshielded detector increases
earlier and has a different rise time to saturate when compared to the shielded
detector. Light from the unshielded detector is an indicator of the muzzle flash of
the MSFC gun. Conversely, light from the shielded detector is TL from the ZnS:Mn
and PPMS paint. Only TL generated inside the shield can be detected. The impact
of the 1 mm aluminum projectile generated triboluminescent light, which was
detected by the photodetector on the painted side of the aluminum plate. The
rapid increase in light intensity from the shielded detector shows the tribolumines-
cent rise time for ZnS:Mn.
Additional evidence can be found in Fig. 14.18 for shot five of the December
2004 test series. A 1 mm projectile with a velocity of 5.6 0.5 km/s impacted an
aluminum target plate and detector assembly as shown in Fig. 14.13. The time-
dependent light intensities for the shielded and unshielded photodetectors are
shown in Figs. 14.18 and 14.19. As before, the shielded photodetector measured
only TL from ZnS:Mn produced by the hypervelocity impact. For shot five during
the December 2004 test, the corresponding triboluminescent decay time for ZnS:
0.4
Shot One Unshielded
Detector Shielded
(Muzzle Flash) Detector
0.3 (TL Light)
Relative Light Intensity
0.2
0.1
0.0
-40 -20 0 20 40
Time (µs)
Fig. 14.17 Light intensity comparison from unshielded and shielded detectors as a result of a
hypervelocity impact on a ZnS:Mn-coated plate for shot one during the December 2005 test series
at MSFC
Mn was calculated to be 336 μs, which is consistent with what was previously
measured using ultraviolet excitation and is also in agreement with others who have
measure the prompt fluorescent decay time of ZnS:Mn [27–31].
14.4.2 Signal Analysis
For the hypervelocity impacts, two pieces of information were desired. The first
was a measurement of the fluorescent decay time. These values were measured by
fitting a two parameter exponentially decaying function of the form
I ¼ I o et=τ ; ð14:2Þ
where Io is the initial intensity and τ is the fluorescent decay time, to the output data
of time, t, and intensity, I. As can be seen from Figs. 14.18 and 14.19, there appears
to be an undefined spike in light intensity signal before the exponential decay. It is
probable that this spike is due to either piezoelectric shock to the detector or flexing
from the shield that would permit muzzle flash light to leak. It is important to note
that misses and background shots produced no measurable light for the shielded
detector. As such, only the exponential component of the recorded TL light was
analyzed. The time domain of the fit was conducted from the maximum along the
decay to 1 ms after the maximum. A 1 ms integration time was chosen as it was
nearly three fluorescent decay lifetimes, and it is also approximately the time
0.6
Unshielded
Detector
0.5 (Muzzle Flash)
Shielded
0.4 Detector
(TL Light)
0.3
0.2
0.1
0 1 2 3 4
Time ( ms )
hypervelocity impact on a ZnS:Mn-coated plate for shot five during the December 2005 test series
at MSFC
1.4
Unshielded
1.2 Detector (Muzzle Flash)
1.0
0.8
Shielded
Detector
0.6 (TL Light)
0.4
0.2
0.0
0 1 2 3
Time (ms)
hypervelocity impact on a ZnS:Mn-coated plate for shot six during the December 2004 test series
at MSFC
0.7
0.6 Fit Limits

Light Intensity (Arb. Units)
0.5
0.4
0.3
Exponential
0.2
Decay Fit
0.1
0.0
0 1 2 3
Time (ms)
Fig. 14.20 Exponential decay fit of shot six, December 2004
required for ZnS:Mn to decay to a tenth of its initial value [21]. A plot of an
example fit superimposed over the signal is shown in Fig. 14.20.
To establish a figure of merit allowing comparison of triboluminescent light,
data points of light intensity were integrated with respect to time over the same
domain as the exponential decay fit. This figure of merit is defined as the tribolu-
minescent light yield, and to permit easier interpretation is normalized to the
average triboluminescent light yield at the minimum measured impact velocity of
3.3 0.3 km/s as the triboluminescent light yield ratio. It is this quantity that will be
referred to in discussion of impact discrimination.
14.4.3 Fluorescent Decay Time Analysis
A table of measured fluorescent decay times measured in this research is given in

Table 14.3. The average decay time of the TL produced by ZnS:Mn induced by
hypervelocity impact was measured to be 292 58 μs. This result is consistent with
values previously measured [27–31]. These results serve to confirm that the events
measured were indeed triboluminescent.
14.4.4 Spectral Analysis
Once the hypervelocity triboluminescent event was consistently observed, it was

desired to record the spectrum and compare it to the data previously gathered on the
Table 14.3 Measured Date Fluorescent decay time (μs)

fluorescent decay times of
December 2004 336
ZnS:Mn excited by
hypervelocity impact 274
315
August 2005 175
January 2006 143
January 2006 508
Average fluorescent decay time 292 58
emission spectrum of ZnS:Mn as shown earlier in Fig. 14.5. This was first
attempted by instrumenting the impact with a shielded fiber optic facing the
phosphor coating of the targets. However, given the limitations of fiber optic
diameter, loss of signal at the feed through, and variation of impact location, it
was not possible to measure a spectrum with sufficient signal/noise ratio that would
permit analysis.
In December 2005, a 589 10 nm pass filter was placed in front of the light
detector. When a shot was fired at 5.6 0.5 km/s, the resultant plot was similar in
shape and had a decay time of 274 μs, consistent with other impact data. The
integrated light production of this shot had a value 38 % of that without the filter in
place. These results demonstrated that emitted light had a decay time that was
consistent with other TL events for ZnS:Mn. Also, light was emitted at wavelengths
that ZnS:Mn is known to fluoresce. As the emission peak is much wider than the
filter pass window, the measured light intensity is less than was observed when no
filter was in place. This evidence served as to further verify that TL induced from
hypervelocity impact was recorded.
14.4.5 Impact Energy/Velocity Discrimination
Once hypervelocity impact-induced TL was reliably observed and measured, it was

desired to observe the nature between impact energy/velocity and light production.
Such dependency would suggest the possibility of discriminating impact energy/
velocity by observing TL. As such, during the last test series in January 2006, shots
were taken to intentionally vary the velocity of the impact. Data was collected for
ZnS:Mn at impact velocities of 3.3, 4.7, and 5.6 km/s.
To analyze these shots, the plots of time dependent light intensity were numer-
ically integrated by computer from the peak of triboluminescent intensity for 1 ms.
When plotted, the resulting integrated light as functions of both impact velocity and
energy are given in Fig. 14.21. The average integrated light intensity of the 4.7 km/s
impacts was measured to be about 6.2 times greater than the 3.3 km/s impacts, and
the average integrated light intensity of the 5.6 km/s impacts was measured to be
about 4.8 times greater than the 3.3 km/s impacts. It can be stated that the both the
Velocity (km/s)
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
12
10
Triboluminescent Yield Ratio
0
2 4 6 8 10 12 14 16 18 20
Impact Energy (J)
Fig. 14.21 Light yield ratio versus impact energy and velocity for ZnS:Mn
4.7 and 5.7 km/s impacts were statistically more luminous than the 3.3 km/s impact.
However, the 4.7 and 5.7 km/s impacts were not distinguishable in luminosity.
These results indicate the current state of the art in impact energy/velocity discrim-
ination using phosphors in the hypervelocity regime. While it is not possible to
determine the exact nature of any dependency of TL light production on impact
energy/velocity, that increased light was generated as the impact energy was
increased.
14.4.6 ZnS:Cu
While the majority of this research has concerned itself with ZnS:Mn as the active
phosphor element, this is by no means the only material that may be considered. As
previously stated, many materials exist with the proper characteristics to make them
suitable for use in such experimentation. To demonstrate this, two shots were taken
with ZnS:Cu prepared similarly as the ZnS:Mn samples at 5.6 0.5 km/s. When
their triboluminescent yields were compared with ZnS:Mn at the same impact
velocity, it was found that ZnS:Cu emitted 11 % the light of ZnS:Mn at the same
impact energy. A plot of ZnS:Cu triboluminescent emission at an impact velocity of
5.6 0.5 km/s is given in Fig. 14.22.
0.20
Unshielded
Detector
Phoshor
(Muzzle Flash) Muzzle Flash
0.18
Light Intensity
0.16
0.14
Shielded
0.12 Detector
(TL Light)
0.10
0.0010 0.0012 0.0014 0.0016 0.0018 0.0020 0.0022 0.0024
Time (s)
Fig. 14.22 Triboluminescent light production from ZnS:Cu at 5.6 0.5 km/s
14.4.7 Hypervelocity Conclusions
Results clearly show that hypervelocity impact-induced TL has been observed and
recorded for both ZnS:Mn and ZnS:Cu. The first verifiable recorded instance of TL
induced by hypervelocity impact occurred in December 2005. This was the event
that permitted further exploration into the characteristics of hypervelocity impact
induced TL. The hypervelocity impact-induced TL decay time was found to be
292 58 μs, which is very consistent with previously measured values for ZnS:Mn.
Further, the emission of TL from ZnS:Mn undergoing hypervelocity impact has
been demonstrated to have a significant component at the known peak emission
wavelength of ZnS:Mn of 585 nm.
TL from hypervelocity impact has been demonstrated with another phosphor,
ZnS:Cu, known to have TL characteristics. This permits one to note that hyperve-
locity impact TL is not specific to only one phosphor and implies its tendency to
occur with any phosphor that is known to generate TL.
The most intriguing conclusion from this research is that it may be possible to
discriminate impact energy/velocity by measuring the time integrated light pro-
duction of TL phosphors. While this phenomenon has been observed at lower
impact energies/velocities, the impact energies/velocities investigated in this
research have a much greater physical significance with respect to impacts likely
to occur in a space environment. An ability to measure the energy/velocity of an
impact is a significant indicator of the potential usefulness of a proposed impact
sensor.
The characteristics of hypervelocity impact induced TL imply its potential use as
the active element in an impact sensor. Phosphors may be applied to a very wide
range of surfaces and materials in a non-invasive fashion. They also provide a well
characterized signal and demonstrate the ability to discriminate impact energy/
velocity beyond binary (impact or no impact) sensing. Additionally, as many
phosphors have TL properties, experimentation with these other materials may
permit a more useful phosphor than the ZnS:Mn and ZnS:Cu used in this research.
A great deal of work by many has gone into the research and development of TL
phosphors that has taken them from an interesting and accidental phenomenon in
candy to this demonstration as a potential active element in a hypervelocity impact
detection system.
14.5 Measuring the Triboluminescent Spectrum

of ZnS:Mn at Meso-Velocities
While the work of Bergeron et al. was successful at measuring the triboluminescent
signature of ZnS:Mn at hypervelocities, they were not able to measure a tribolu-
minescent spectrum of ZnS:Mn owing to the experimental limitations. After all,
triboluminescent emission generated at meso-velocities typically takes place far
away from the recording equipment. In order for a spectrometer to measure dim TL,
a telescope must be used. This allows for the possibility of capturing tribolumines-
cent light generated by the impact while keeping the fiber-optic cables and spec-
trometer out of the line of fire.
14.5.1 Telescope
For this research, three telescopes were tested: (1) 10 in. Meade LX2000GPS
Schmidt-Cassegrain telescope, (2) Orion Observer 60 mm Alt-azimuth refractor
telescope, and (3) Celestron Regal 65F-ED finder scope. After observing several
objects approximately 15 m away from the Meade telescope, it was determined
that it would not be useful for this research. While the Meade telescope would
capture the most light, its field of view (FOV) at 50 m was only 2 cm wide.
The small FOV for the Meade makes it impractical for detecting TL 15 m away
[29, 32, 33].
The Orion Observer 60 mm refractor telescope has a larger FOV (about 8 cm
wide at 16 m) and was used to capture the first triboluminescent emission spectrum
for ZnS:Mn at ballistic speeds. However, the Orion telescope does not have the
light gathering power of the Meade. Additionally, the practical FOV was still too
small to capture a triboluminescent flash 15 m away with any consistency.
If the spectrum is to be measured for each impact, then a telescope that has a
FOV that covers the entire 30 cm target stand is needed. As a result, a Celestron
Regal 65F-ED spotting scope was used for this research. This finder scope features
a 65 mm diameter aperture, 23 m linear FOV at 1000 m with 48 magnification,

near focus of 4.6 m at 16 magnification, focal length of 386 mm with a focal ratio
of f/6, an eyepiece zoom of 8–24 mm, angular field of view of 2.6 at 16 and 1.3
at 32, and magnification range between 16 and 48 [29, 32, 33]. With these
properties, the Celestron Regal telescope gives researchers the best chance to
measure TL produced during ballistic impacts.
14.5.2 Spectrometer
To connect a fiber optic spectrometer to a telescope, a 600 μm SMA-based fiber-

optic cable connects the spectrometer to the spotting scope using a C-mount adapter
purchased from Ocean Optics. This adapter was attached to the top of the eyepiece
of the spotting scope. The Celestron Regal spotter scope was aligned before the
C-mount adapter was attached. The focus of the spotter scope was adjusted to
maximize the intensity of the spectral emission from a small flashlight placed on the
target stand.
In order to determine if the spotting scope and S2000 spectrometer could detect a
ZnS:Mn triboluminescent spectrum, a final lab test was performed. Pellets (2.00 cm
long and 0.88 cm in diameter) of red and green phosphorescent materials were
placed 8 m from the spotting scope. These pellets had approximately the same
luminescent intensity as a ZnS:Mn triboluminescent flash observed at a distance of
16 m. After charging the pellets under white light for a few seconds, the source was
removed and the spectra of the green and red phosphors were recorded using the
S2000. The samples were then placed 16 m away from the Celestron Regal spotter
scope and the experiment was repeated. This distance test showed that the S2000
could record dim light with intensity similar to that emitted by a TL flash. The
equipment was moved to the field to begin the actual TL measurements.
14.5.3 Complete System
Figure 14.23 shows a diagram of the experimental system used to measure the TL
emission spectrum from ZnS:Mn. Figure 14.24 shows a picture of the actual setup.
A Celestron Regal 65F-ED finder scope was used to collect triboluminescent light
at a distance of 16.8 m. An Ocean Optics S2000 spectrometer and laptop computer
running SpectraSuite software was used to image and record the emission spectrum
from the TL flash. A CED M2 infrared chronograph was used to measure the
velocities of individual ballistic projectiles. Later, the emission spectrum and
total emitted triboluminescent yield were determined using a specially written
LabVIEW virtual instrument (VI).
Fig. 14.23 Diagram of the experimental apparatus used to measure the triboluminescent spectral
shift using 9 mm ammunition augmented with ZnS:Mn powder [29, 32, 33]
Fig. 14.24 Picture of experimental apparatus used to measure the triboluminescent spectral shift
using 9 mm ammunition augmented with ZnS:Mn powder [29, 32, 33]
Fig. 14.25 Picture of a

typical copper-jacketed
9 mm projectile similar
to what was used in this
research [29, 32, 33]
14.5.4 Ballistic Projectiles
For this research, 9 mm projectiles were augmented with ZnS:Mn powder. The term
“9 mm” describes both the ammunition type and size. The base 9 mm projectiles
used in this study, shown in Fig. 14.25, have a mass of 124 grains (8.03 g) and were
manufactured by Rainier Ballistics. These bullets feature a flat nose and a copper
plated jacket. These projectiles are 13.2 mm long and 9 mm across. The process of
machining these projectiles began by first setting a 3/16 in. (4.8 mm) drill bit into a
lathe as shown in Fig. 14.26a. Using a collet, a 0.478 in. (12.1 mm) deep hole was
cut in the rounded end of each 9 mm projectile as shown in Fig. 14.26b.
The ZnS:Mn powder used in this research had an average grain size of 7.5 μm
(lot 15248) and was manufactured by Phosphor Technology, Limited of Great
Britain. Approximately 5 grains (0.33 g) of ZnS:Mn powder was used in each
9 mm round as shown in Fig. 14.26c. Each projectile was then capped with two-part
epoxy as shown in Fig. 14.26d. After the epoxy cured, the projectiles were prepared
for firing using standard reloading procedures [34]. For this research, 200 rounds of
augmented 9 mm projectiles were manufactured.
14.5.5 Analysis Software
Once the emission spectra were recorded in the field, the SpectraSuite-generated
text files were processed using a specialty written LabVIEW VI called Spectrum-
Processor (version 1.0). The flow of SpectrumProcessor is shown in Fig. 14.27. This
program was designed to analyze large amounts of SpectraSuite data files effi-
ciently. As Fig. 14.27 shows, once the files are generated, the user first must decide
on the spectral range to perform the analysis. For example, the triboluminescent
emission of ZnS:Mn has a spectral range of 500–700 nm. Then, the location to save
the generated analysis is set. Afterwards, the user puts the two conditions to save the
data. The first is the minimum spectra amplitude. By setting this number greater
than 0, all of the dark spectra (contain no light) are ignored. The maximum full
Fig. 14.26 Pictures of the machining process for 9 mm triboluminescent rounds showing: (a) A
hole being cut in a projectile using a collet and lathe. (b) Appearance of the projectiles after
machining and engraving. (c) Projectiles containing the ZnS:Mn powder. (d) Completed pro-
jectiles ready to be reloaded for firing [29, 32, 33]
User sets Text file generated

requirements containing
Spectral Range File containing
SpectraSuite SpectrumProcessor spectra
Location to Save Data
generates text files performs the Light yield
Minimum spectra
from S2000 amplitude analysis Emission centroid
location
Maximum FWHM Centroid amplitude
FHWM
Fig. 14.27 Diagram of the flow from SpectrumProcessor [29, 32, 33]
width half maximum (FWHM) must then be set. By setting this to a high number
(~90 nm), the white light used to see the target just before shooting will be thrown
out. For ZnS:Mn, its FWHM is approximately 65 nm.
Once the conditions are set, the program starts by asking the user to choose
which file(s) to analyze. This can be either an entire folder or individual files within
a folder. The files are then sent to a file list VI that sends all the files into a
while loop that continues to run until all the files have been read and processed
[29, 32, 33].
While in the loop, the files are modified to contain only the selected wavelength
range. LabVIEW subroutines were used to calculate the relative light yield and
emission spectrum centroid using the uneven numerical integration VI (trapezoid

rule) and least squares method Gaussian fit VI, respectively. Only the spectral data
satisfying the two conditions are saved for later review. The rest of the data is
discarded and removed from memory. It takes anywhere from about 1 s (20 files or
less) to about 30 min (1200 files) to completely analyze all the data [29, 32, 33].
Later versions of SpectrumProcessor replaced the WHILE loop with a FOR loop,
which significantly reduced processing time. The updated version allows for more
than 30,000 spectra files to be analyzed in slightly over 2 h. In addition, the code
was modified to analyze any spectrum text file, not just SpectraSuite files. Also,
some MATLAB code was embedded into the block diagram to speed up the code as
well as an override button so that all spectra files will be saved regardless of the
condition [29, 32, 33].
14.6 Meso-Velocity Results
All data was collected during the summer of 2009 at the Lafayette Parish, Louisiana
Sheriff’s Department Training Academy in Lafayette, Louisiana. Augmented
9 mm projectiles were fired into an aluminum target using a Masterpiece Arms
971 firearm.
14.6.1 ZnS:Mn Triboluminescent Spectral Data
Fifty 9 mm augmented projectiles were fired to obtain a triboluminescent spectrum

from ZnS:Mn. From this, nine triboluminescent emission spectra were recorded. An
example triboluminescent emission spectrum for ZnS:Mn is shown in Fig. 14.28.
The maximum TL emission wavelength for ZnS:Mn was found to be was
586 1 nm, which is consistent with earlier TL and other photo-excited measure-
ments [17, 24, 25, 27, 31].
14.6.2 Increasing Material Mass
In order to test the customized LabVIEW software, it was decided to measure how
the TL yield changes with increasing ZnS:Mn powder mass. To make this mea-
surement, one to five augmented projectiles would be fired into an aluminum target
over the 5 s integration time of the S2000 spectrometer. One augmented projectile
fired in a 5 s window would mean that about 5 grains of ZnS:Mn powder would
impact with the target. Conversely, when four augmented projectiles are fired in 5 s,
a total of about 20 grains of ZnS:Mn powder would impact with the target. The
spectrometer integration time would be set to 5 s so it would record all impacts in
Fig. 14.28 Example TL ZnS:Mn emission spectrum generated from an impact of an augmented
9 mm projectile moving at a velocity of about 410 m/s with an aluminum target [29, 32, 33]
Table 14.4 Increasing ZnS: Projectiles fired Projectiles Cumulative

Mn material mass shot Set per integration fired per number of fired
matrix [29, 33] number time set projectiles
1 1 10 10
2 2 20 30
3 3 30 60
4 4 40 100
5 5 50 150
that interval. Multiple projectiles would be fired during this interval to achieve the
desired ZnS:Mn mass hitting the target. The increasing material mass shot matrix is
shown in Table 14.4. A total of ten trials were completed for each set of projectiles.
For Set 1, a single 9 mm augmented round were fired per integration time.
Conversely, Set 4 shows that four augmented 9 mm projectiles is fired per integra-
tion time. A total of 150 rounds of augmented 9 mm ammunition were fired for this
measurement [29, 33].
This test began by firing a single augmented 9 mm projectile into the aluminum
target within the 5 s integration window. The resulting triboluminescent emission
spectrum was saved to a computer. This process was repeated until ten trials were
recorded, each with one triboluminescent spectrum being generated per round.
Next, two augmented 9 mm projectiles were fired into the target during the 5 s
spectrometer window. Thus producing one triboluminescent spectra for the run.
This process was repeated until a total of ten runs were completed. This process
then was repeated by firing three, four, and five round sets of augmented projectiles
within the open spectrometer window.
Table 14.5 Results for the 9 mm increasing ZnS:Mn mass experiment. The spectrometer inte-
gration time was set to 5 s for each set [29, 33]
Average Average Average
Number of ZnS:Mn Average total emission
impacts per Projectiles mass on projectile projectile peak Average total
integration fired per target velocity kinetic wavelength relative light
time set (grains) (m/s) energy (J) (nm) yield
1 10 5.3 0.1 407 4 515 11 586.1 0.1 8596 1383
2 20 10.2 0.1 412 3 1059 16 586.1 0.1 16,299 2952
3 30 15.4 0.1 416 3 1625 21 585.8 0.1 17,230 1678
4 40 20.9 0.2 408 2 2094 24 585.8 0.1 26,218 4288
5 50 22.9 0.3 411 1 2658 12 586.2 0.1 22,515 2788
40000
30000
Total TL Yield
20000
10000
0
0 1 2 3 4 5 6
Number of Impacts
Fig. 14.29 Plot of the average total relative light yield versus the number of augmented 9 mm
projectile impacts [29, 33]
After processing the data with the LabVIEW software, the total relative light
yield was calculated for each set as shown in Table 14.5. Note the velocity for the
150 augmented 9 mm projectiles was about 410 m/s. The average peak wavelength
is about 586 nm, which is also consistent with earlier results [17, 25, 27–29, 31, 33,
35]. Table 14.5 also shows that the total triboluminescent light yield is proportional
to the mass of ZnS:Mn on target. Figure 14.29 shows a plot of the average total
relative light yield versus the number of 9 mm impacts per 5-s integration time.
As expected, the relative triboluminescent light yield is proportional to the
number of impacts since each projectile was loaded with nearly the same mass of
ZnS:Mn [29, 33].
14.6.3 Wavelength Shift of ZnS:Mn
In 1966, Koda et al. began investigating the photoluminescent properties of ZnS:

Mn powder under hydrostatic pressure [36]. Using a Drickamer cell, the ZnS:Mn
was subjected to pressures from 5 to 52 kbar with intervals of about 5 kbar. During
each interval, the high pressure ultraviolet source was turned on, and the
photoluminescent emission spectrum was measured. These results showed that
ZnS:Mn has a hydrostatic pressure redshift of about 2.3 meV/kbar, which indicates
that the emission wavelength increases with pressure [36].
The 9 mm projectiles only produce approximately 500 J upon impact. In order to
determine the effects of higher impact energies on the triboluminescent emission
spectra, several 150 grain (9.71 g) 30 caliber projectiles were made in a similar
manner, which resulted in impact energies ranging from 750 to 4000 J [29, 32].
Due to the penetrating power of a high speed 30 caliber projectile, a 1 in. thick
steel plate was placed in front of the aluminum target. To protect the equipment and
users from impact back splatter, plastic and wood shields where placed in front of
the equipment and personnel as shown in Fig. 14.30.
A total of ten projectiles were fired for this test. Only two ZnS:Mn tribolumi-
nescent spectra where recorded during the 30 caliber test as shown in Fig. 14.31.
The average emission spectrum for these two projectiles was 588.3 1.2 nm as
shown in Table 14.6 [29, 32]. During this part of the experiment, the temperature
passed through the dew point. This caused a thick layer of condensation to appear
over the plastic shields and the optics. In addition, the 30 caliber projectiles where
Fig. 14.30 Picture of the setup for higher energy experiment [29, 32]
a 40 b
589.4 nm 40
587.1 nm
30
30
Light Yield
Light Yield
20 20
10 10
0 0
-10 -10
400 500 600 700 800 400 500 600 700 800
Fig. 14.31 ZnS:Mn triboluminescent spectra for 30 caliber rounds with speeds of (a) 932 m/s and
(b) 897 m/s [29, 32]
Table 14.6 Resulting 9 mm and 30 caliber projectile data and peak triboluminescent wavelength
comparison [29, 32]
Average
Number Average fired projectile Average ZnS: Average peak
Projectile of projectile kinetic energy Mn kinetic wavelength
type projectiles mass (g) (J) energy (J) (nm)
9 mm 10 5.83 0.03 514.6 9.1 27.3 1.0 586.1 0.1
30 caliber 2 8.32 0.01 3481.9 137.3 121.3 12.7 588.3 1.2
destroying the steel plate. Due to these limitations, testing was stopped
[29, 32]. However, the obtained results, shown in Table 14.6, indicate that there
is an approximate 1 nm shift in the triboluminescent peak wavelength between the
30 caliber and 9 mm projectiles augments with ZnS:Mn powder. This result is
statistically significant and is analogous to the shift measured by Koda et al. [36].
It should be noted that more data was attempted with the 30 caliber projectiles.
The pointed 30 caliber projectiles where replaced with a Hornady blunt nose
30 caliber (110 grain) projectile similar to the 9 mm. After preparing and loading
30 test projectiles with just epoxy, they were fired into a 1 in. A572 grade 50 high
carbon steel plate. The results from this test indicated that these bullets
would provide the same speeds as a 9 mm projectile. As a result, it was determined
that air rifles should be acquired for future experiments, to better vary the impact
energy [29, 32].
14.6.4 Meso-Velocity Conclusions
Results from this research show that an apparatus can be designed, tested, and
utilized to measure the triboluminescent emission spectrum and total light yield
from materials like ZnS:Mn from ballistic impacts. This apparatus was used to
record the first triboluminescent emission spectrum from ZnS:Mn generated during
a ballistic impact at a velocity of about 410 m/s. In addition, this research has shown
that the triboluminescent emission peak does shift with increasing velocity. More-
over, this shows that it is also possible to determine the impact energy by measuring
not only the light emission, but also the spectrum peaks. As such, an impact sensor
can now be designed that can discern energy upon impact.
References
1. Shionoya, S. (2006). “Fundamentals of Luminescence”, in Phosphor Handbook, 2nd ed., Yen,

W. M. and Yamamoto, H., (Eds.), (pp. 3–8), CRC Press.
2. Walton, A. J. (1977). Triboluminescence. Adv. Phys. 26(6), 887–948, Taylor & Francis.
doi:10.1080/00018737700101483.
3. Xu, C. N., et al. (1999). Preparation and characteristics of highly triboluminescent ZnS film.
Mater. Res. Bull., 34(10–11), 1491–1500.
4. Xu, C. N., et al. (2000). Dynamic visualization of stress distribution by mechanoluminescence
image. Appl. Phys. Lett., 76(2), 179. doi:10.1063/1.125695.
5. Xu, C. N., et al. (1999). Enhancement of adhesion and triboluminescence of ZnS:Mn films by
annealing technique. Thin Solid Films, 352(1–2), 273–277.
6. Xu, C.-N. et al. (2001). Visualization of stress distribution in solid by mechanoluminescence.
In: Mercer, C. R. Cha, S. S. and Shen, G. (Eds.), Optical Diagnostics for Fluids, Solids, and
Combustion 4448(1), (pp. 398–407), SPIE, San Diego, CA, USA.
7. Yang, H., Holloway, P. H., & Ratna, B. B. (2003). Photoluminescent and electroluminescent
properties of Mn-doped ZnS nanocrystals. J. Appl. Phys., 93(1), 586. doi:10.1063/1.1529316.
8. Xu, C. N., Xu, C.-N., Yamada, H., Wang, X., Zheng, X.-G., et al. (2004). Strong elasticolu-
minescence from monoclinic-structure SrAl2O4. Appl. Phys. Lett., 84(16), 3040. doi:10.1063/
1.1705716.
9. Sweeting, L. M. et al. (1992). Spectroscopy and Mechanism in Triboluminescence. Mol. Cryst.
Liq. Cryst. Sci. Technol. Sect. A. Mol. Cryst. Liq. Cryst. 211(1), 389–396, Taylor & Francis.
doi:10.1080/10587259208025838.
10. Akiyama, M., Xu, C.-N., & Nonaka, K. (2002). Intense visible light emission from stress-
activated ZrO[sub 2]:Ti. Appl. Phys. Lett., 81(3), 457–459. AIP.
11. Chandra, B. (1998). Mechanoluminescence. In: Vij, D. R. (Ed.), Luminescence of Solids, (pp.
361–389), Plenum Press, New York, NY.
12. Chandra, B. P. and Zink, J. I. (1980). Triboluminescence and the dynamics of crystal fracture.
Phys. Rev. B 21(2), 816–826, American Physical Society. doi:10.1103/PhysRevB.21.816.
13. Chandra, B. P. and Zink, J. I. (1980). Triboluminescence of inorganic sulfates. Inorg. Chem. 19
(10), 3098–3102, American Chemical Society. doi:10.1021/ic50212a054.
14. Womack, F. N., et al. (2004). Measurement of triboluminescence and proton half brightness
dose for ZnS:Mn. IEEE Trans. Nucl. Sci., 51(4), 1737–1741.
15. Womack, F. N. (2004). Development of a drop tower to study the triboluminescence of ZnS:
Mn with attention to possible applications to spacecraft, University of Louisiana at Lafayette.
16. Allison, S. W. and Gillies, G. T. (1997). Remote thermometry with thermographic phosphors:
Instrumentation and applications. Rev. Sci. Instrum. 68(7).
17. Bergeron, N. P. (2006). Detection of triboluminescence from ZnS:Mn and ZnS:Cu, University
of Louisiana at Lafayette.
18. Allison, S. et al. (2003). Advances in High Temperature Phosphor Thermometry for Aerospace
Applications, in 39th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit,
American Institute of Aeronautics and Astronautics. doi: 10.2514/6.2003-4584.
19. Heiken, G. H., Vaniman, D. T., & French, B. M. (Eds.) (1991). Lunar sourcebook: A user’s
guide to the moon, Cambridge University Press.
20. Charters, A. C., pat Denardo, B., & Rossow, V. J. (1977). Development of a Piston-Compres-
sor Type Light-Gas Gun for the Launching of Free-Flight Models at High Velocity. Moffet
Field, CA.
21. Yen, W. M. & Weber, M. J. (2004). Inorganic Phosphors: Compositions, Preparation and
Optical Properties, 1st ed., CRC Press.
22. F. H. S., B. S. M., Z. L. F., K. D., G. H. L., & W. Chen (2002). Temperature behaviour of the
orange and blue emissions in ZnS:Mn nanoparticles. J. Phys. Condens. Matter 14(47), 12657.
23. Hollerman, W. A. et al. (2007). Unusual fluorescence emission characteristics from
europium-doped lead phosphate glass caused by 3 MeV proton irradiation. In: 2007 IEEE
Nuclear Science Symposium, pp. 1368–1372, IEEE, Honolulu, HI. doi:10.1109/
NSSMIC.2007.4437255.
24. Hollerman, W. A., et al. (2005). Effects of proton irradiation on triboluminescent materials
such as ZnS:Mn. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater.
Atoms, 241(1–4), 578–582. doi:10.1016/j.nimb.2005.07.072.
25. Hollerman, W. A. et al. (2005) Emission spectra from ZnS:Mn due to low velocity impacts. In:
Photonics for Space Environments X 5897(1), Taylor, E. W. (Ed.), p. 58970F–10, SPIE, San
Diego, CA, USA. doi:10.1117/12.613570.
26. Goedeke, S. M. et al. (2006). Study of Resbond® Ceramic Binders Used for High Temperature
Non-Contact Thermometry. In: Advances in Photonic Materials and Devices, (pp. 115–125),
John Wiley & Sons, Inc. doi:10.1002/9781118407233.ch10.
27. Bergeron, N. P., et al. (2008). Triboluminescent properties of zinc sulfide phosphors due
to hypervelocity impact. Int. J. Impact Eng., 35(12), 1587–1592. doi:10.1016/j.ijimpeng.
2008.07.007.
28. Bergeron, N. P., et al. (2006). Experimental evidence of triboluminescence induced by hyper-
velocity impact. Int. J. Impact Eng., 33(1–12), 91–99. doi:10.1016/j.ijimpeng.2006.09.079.
29. Fontenot, R. S. (2010). Measuring the Triboluminescence Characteristics of ZnS:Mn Gener-
ated During Ballistic Impacts, University of Louisiana at Lafayette.
30. Fontenot, R. S., et al. (2012). Comparison of the Triboluminescent Properties for Europium
Tetrakis and ZnS:Mn Powders. J. Theor. Appl. Phys., 6(1), 15. doi:10.1186/2251-7235-6-15.
31. Hollerman, W. A., et al. (2012). Comparison of Triboluminescent Emission Yields for
Twenty-Seven Luminescent Materials. Opt. Mater. (Amst), 34(9), 1547–1521. doi:10.1016/j.
optmat.2012.03.011.
32. Fontenot, R. S., Hollerman, W. A., & Goedeke, S. M. (2011). Initial evidence of a
triboluminescent wavelength shift for ZnS:Mn caused by ballistic impacts. Mater. Lett.
65(7), 1108–1110, Elsevier B.V. doi:10.1016/j.matlet.2011.01.043.
33. Fontenot, R. S., & Hollerman, W. A. (2011). Measuring triboluminescence from ZnS:
Mn produced by ballistic impacts. J. Instrum., 6(04), T04001. doi:10.1088/1748-0221/6/04/
T04001.
34. Lee, R. (2003). Modern reloading, 2nd ed., A.C. Kerman.
35. Hollerman, W. A. Glass, G. A., & Allison, S. A. (1999). Survey of Recent Research Results for
New Fluor Materials. MRS Online Proc. Libr. 560, 335–341, Cambridge Journals Online.
doi:10.1557/PROC-560-335.
36. Koda, T., et al. (1966). Effect of pressure on the luminescence of zinc sulphide phosphors. J.
Phys. Chem. Solids, 27(10), 1577–1586. doi:10.1016/0022-3697(66)90235-6.
Index
A Cementitious composites, 380–383

Acoustic cavitation civil infrastructural systems, 379
definition, 239 crack propagation (see Cracks)
extended surface, 240–241 in situ damage-sensing system, 380
heterogeneous slurries TL, 383–384
interparticle collision, 241 TLCOS, 385–387, 389
microjets, 242 Center for Nanotechnologies (CINT), 338
organic crystals, 242 Central vertebrae (temporary CPU), 306
homogeneous sonochemical Cerium tetrakis dibenzoylmethide
reactions, 238 triethylammonium compound,
interparticle collisions, 238 186, 187
liquid–powder slurry, 238 CF. See Carbon fiber (CF)
stable cavitation, 239 CFE. See Carbon fiber electrode (CFE)
synchronous cavitation, 240 CH (A2Δ X2Π) emission, 258
transient cavitation, 239 Chemical vapor deposition (CVD), 359, 362
ultrasonic irradiation, 238 Chemiluminescence, 2, 42, 83
Active sensing system, 308 CNTs. See Carbon nanotubes (CNTs)
Aerospace and civil applications, 305 CNY. See Carbon nanotube yarn (CNY)
Antenna effect, 47, 50 Composite layer effect, 323
Concrete
crack initiation and propagation,
B 380–382
Bimodal distribution, 259 ITOF sensors, 399
Bioluminescence, 2, 42 monosulfate hydrate, 389
Boltzmann plot, 258 Coulomb interactions, 287
Crack
challenges, 382–383
C drying shrinkage, 380, 381
Carbon fiber (CF), 360 electrochemical process, 381
Carbon fiber electrode (CFE), 356 physical and chemical causes, 380, 381
Carbon nanotube yarn (CNY), 365, 366 Crystal synthesis methods, 310
Carbon nanotubes (CNTs), 353 Crystalloluminescence (CRL), 2, 294, 296
Catalytic decomposition mechanism, 296 CVD. See Chemical vapor deposition (CVD)
Cathodoluminescence, 41 Cytoseal®, 218, 219, 224, 225
Celestron Regal 65F-ED finder scope, 434 Cytoseal mounting, 219

D.O. Olawale et al. (eds.), Triboluminescence, DOI 10.1007/978-3-319-38842-7
446 Index
D E
Defense Evaluation Research Agency Elastico-triboluminescence, 4–5, 383
(DERA), 312 Electrokinetics, 104, 109
Detector photodiode (PD), 43 Electroluminescence, 2, 41
Dibenzoylmethanate (dbm ), 49 Electron spin resonance (ESR)
Dibutyl phosphate (DBP) spectroscopy, 32
effects, 214 Electronic excitation, 127–128
concentrations, 212 Energy dispersive scanning (EDS), 316
drop tests, 214 Energy transfer, 291, 292, 295
flattened rod crystals, 214 Europium salts
formation, 212 acetate, 188
high concentrations, 213 drop tower and LabVIEW VI tools, 188
impact energy, total TL yield, 216, 217 ethanol, 188
LabVIEW VI, 213 europium acetate, 188
luminescence, 216 lower triboluminescent yield, 189
organic synthesis, 212 SEM, 191
outcomes, 214 TL emission spectrum, 192
PL, 217 UV light, 191
photoluminescent emission Europium tetrakis
spectra, 216 anhydrous europium (III) chloride, 166
standard optical microscope, 214 nitrate, 167
synthesis, EuD4TEA, 212 synthesis method, 166
Dielectrics, 88–90 TL decay times, 173, 174
Dimethyl methylphosphonate TL light yield, EuD4TEA
(DMMP), 11 activation energy, 171
crystalline spherical product, 207 DMMP, 169
crystal structure, EuD4TEA, 215, 216 energy level, 172
decay outcomes, 210 grain sizes, 173
decay time software, 209 impact energy, 169, 170
EuD4TEA, 208 measures, 171
europium and DBM, 207 minimum energy, 170
impact energy, 210, 211 nitrate, 168
measured TL spectra, 210 outcomes, 170
methyl-dimethyl ester, 207 photoluminescent and
outcomes, 207 triboluminescent spectra, 172
phosphonic acid, 207 photoluminescent emission spectra, 171
photoluminescent and triboluminescent testing, 171
emission spectra, 209 ZnS:Mn, 167
PL spectra, 209 Europium tetrakis dibenzoylmethide
precipitation, 207 triethylammonium (EuD4TEA),
synthesis, 207 147, 149–150, 173–180, 184–190,
TL, 208, 209 207, 212, 218–227
Dopant, 10 additives, 215
3D solar cells crystalline product, 158
CNY, 365–367 DBM, 158
CNY photovoltaic, engineering DBP (see Dibutyl phosphate (DBP))
applications, 371–373 DMMP doping (see Dimethyl
photovoltaic carbon fiber, 360–363 methylphosphonate (DMMP))
wire-shaped CNT-CNT solar cells, drop tower test, 230
367–371 electronic transitions, 160
DSSC. See Dye-sensitized solar cell excitation and emission spectra, 159
(DSSC) excitation energy, 159
Dye-sensitized solar cell (DSSC), 353 food grade piperine, 194, 195, 197
Index 447
grain size, 181–184 lanthanide compounds, 184–188

health monitoring and radiation light yield, 187
sensors, 227 Ln, 188
ICP-OES analysis, 199 materials, 228
LabVIEW Software, 163–165 steps, 228–230
luminescence (see Luminescence) techniques, 147
material, 158 TL (see Triboluminescence (TL))
morphine, 192, 193 uranium doping, 200–206
multivitamin doping, 197–200
photoluminescent emission and
excitation spectra, 159 F
piperine doping, 192–196 Fabrication and integration, 315–320
PL, 165, 166 Fiber-optic technology, 306
polymers (see Polymers, EuD4TEA) Fiber-reinforced polymer (FRP), 319, 402
quantity, research, 227 Fractoluminescence (FL), 242, 243, 383
reagent grade piperine, 195, 197 Fractomechanoluminescence (fracto-ML),
salts (see Europium salts) 242, 243
solvents effects Fracto-triboluminescence, 7–8, 383
acetone, synthesis and crystallization FRP. See Fiber-reinforced polymer (FRP)
process, 176, 178 Full width half maximum (FWHM), 112
acetonitrile, 176 FWHM. See Full width half maximum
alcoholic, 176 (FWHM)
benzene, 177
carbonyl, 176
chloroform and methylene chloride, 179 G
crystal grain size, 181 Gas discharge lines
cyclohexane, 176 acoustic cavitation, 248
decay time data, 179 CH and H emission, 252
ethyl alcohol, 175 crystal luminescence, 248
material, TL yield emission, 179 He and O lines, 248
methylene chloride, 177, 178 He gas discharge, 248, 249
outcomes, 178 perdeutero-dodecane, 250
photoluminescent emission spectra, 180 SL, 249
PL emission spectra, 180 sonication, 249–252
research, 177 sonolysis products, 251
TL-based impact sensor system, 178 GFRP. See Glass fiber-reinforced plastic (GFRP)
TL decay time, 180 Glass fiber composite (GFC), 324
TL emission, 178 Glass fiber-reinforced composite (GFRC), 316
TL property, 176 Glass fiber-reinforced plastic (GFRP), 137
triboluminescent materials, 175 Guns, 415
triphenylphosphine recrystal, 176
structure, 158
synthesis, 166–173 H
cerium, 187 Hexafluoroacetoacetonate (hfac ), 49
decay time outcomes, 200 Hexamethylphosphoramide (hmpa), 51
drop tower error, 173–175 High-Performance Materials Institute
elements, 186 (HPMI), 306
ethyl alcohol and acetone, 174 Holmium compound crystallites, 186
europium acetate and nitrate, 189, 190 Homogeneous sonochemical reactions, 238
europium tetrakis (see Europium Hydroluminescence, 105–108
tetrakis) sonoluminescence, 121–122
fluorescence decay, 184 spectrum, 117–118
holmium, 186 triboluminescence, 119
448 Index
Hypervelocity Low-velocity-impact, TL
characteristics, 432 applications, 348
fluorescent decay time analysis, 429 coating methodology, 345
guns, 415–416 coating properties, 345–346
impact energy/velocity discrimination, damage/fracture, 334
430–431 decay time, 343
signal analysis, 427–429 drop tower testing methodology, 334–337
spectral analysis, 429–430 emission wavelength, 344
ZnS:Cu, 431 emission yield
Hypervelocity guns, 415–416 dopant/activator, 338
europium tetrakis, 340
inorganic ZnS compounds, 337
I micro-grained ZnS compounds, 338
In situ triboluminescent optical fiber (ITOF), 314 nano-grained ZnS:Mn, 338–339
HNS, 391 TL intensities, 337
POF, 391 triboluminescent light yield, 340–342
real-time damage monitoring, 399–401 environmental exposure, 344–345
real-time failure detection fluorescence, 342, 343
compression zone, 397, 399 luminescent materials, 342
TCC, 393, 395 mechanoluminescence, 333
TL emission, 396–397 Mythbusters®, 344
retrofitted concrete structures, 402–406 NAC GX-1 high-speed camera, 344
TMCC, 391, 392 phosphorescence, 342, 343
TSR, 391 signal-level estimation, 346–348
ZnS:Mn crystals, 391, 392 triboluminescent emission, 333, 343
Incident photon conversion efficiency triboluminescent spectrum, 333
(IPCE), 356 ZnS-based materials, 334
Insulators, 18, 19 Luminescence, 1–2, 17, 34, 40–42
Interparticle collision, 241 absorption and photons emission, two-level
Intrinsic structural health monitoring system, 149
(iSHM), 351 activators/dopants, 149
IPCE. See Incident photon conversion emission, EM radiation, 149
efficiency (IPCE) excitation, 184
iSHM. See Intrinsic structural health fluorescence, 150
monitoring (iSHM) intensity, 297
ITOF. See In situ triboluminescent optical materials, 149
fiber (ITOF) mechanical action, 150
phosphorescence, 150
photoluminescence (PL), 150
L
LabVIEW software
Decay Time VI, 164, 165 M
Light Yield Finder VI, 164 Mechanochemical decomposition
materials, 164 Ar pressure, 289
program, 163 cerium sulfate TL, 288
Spectrum Finder VI, 165, 166 H luminescence, 289
spectrum processor, 163 OH
TL emission data files, 164 band intensity, 288
virtual instrument (VI) program, 163 and OD luminescence, 289
Lanthanide aqueous ions, 48 terbium sulfate, 287
Lanthanide complexes, 44, 49, 50, 57 water vapor, 289
Light emission mechanisms, 2, 311 Mechanoluminescence (ML), 6, 383
Lighting-emitting devices (LEDs), 41 acoustic cavitation
Index 449
alkane chain length, 253 Non-destructive inspections, 309

emission intensity, 245–247 Non-destructive testing (NDT), 307
fracto-ML, 244 Nonradiative energy transfer, 286
gas emission lines, 248, 249, 251, 252 Nonthermal plasma, 257, 258
He discharge, 253
resorcinol, 245, 246
shockwave, 252, 254 O
sonication, 247 Organic–inorganic hybrid composite, 139–140
sucrose crystals, 244–247 Oxide materials, 18, 19, 21, 24, 28, 35
coordination compounds, 45–47
coumarin ground, 243, 244
emission intensity, 237 P
fracto-ML, 242, 243 Paschen’s law, 82
general aspects, 42–44 Passive sensing system, 308
of lanthanide compounds, 47–52 Peroxy
near UV to near IR spectrum, 247 band structure and electronic charge
piezoelectric crystals, 243 carriers, 18–21
plasma (see Plasma characteristics) defects, 19–20
shock wave, 241 break-up and formation, positive
of transition metal compounds, 53–58 holes, 24–26
types, 242 oxygen anions in 1 valence state,
Menthyl-9-anthracenecarboxylate (MAC), 21–34
312, 318 propagation, positive hole charge
Meso-velocity carriers, 28–31
ballistic projectiles, 436 in silica and silicates, 26–28
caliber test, 441, 442 trapping and detrapping, positive hole
hydrostatic pressure, 441 charge carriers, 32–34
LabVIEW software, 440 nature of electronic charge carriers,
material mass, 439 electrons, 21
software analysis, 436–438 Peroxy bond, 26
spectrometer, 434 Peroxy defects, 18, 20, 21, 25, 26, 29–35
telescope, 433–434 Phosphor Technology, Limited of Great
Metal-to-ligand charge transfer (MLCT) Britain (PT-GB), 338
processes, 53 Phosphorescence, 2
MKT. See Molecular kinetic theory (MKT) Photoluminescence (PL), 1, 41, 165, 166, 274
Modern detection systems, 305 emission, 173
Molecular kinetic theory (MKT), 103 materials, 312
Monatomic inert gases properties, 185
cerium sulfate, 281, 282 quantum, 384
STL, 280 Photomultiplier tube (PMT), 9, 310, 383, 384
terbium carbonate, 280 Photovoltaic carbon fiber, 360–363
xenon lines, 282 Phthalocyanine (Pc), 218
Morphine, 193 Piezoelectric crystals, 243
Multi-bubble cavitation, 240 Piezoelectric theory, 8
Multi-wall carbon nanotube (MWCNT), 356 Piezoelectrification, 311
MWCNT. See Multi-wall carbon nanotube Piezo-luminescence, 383
(MWCNT) Pivaloylmethanate (dpm), 49
Mythbusters®, 344 PL. See Photoluminescence (PL)
Planck’s black body emission theory, 1
Planck’s black body radiation theory, 40
N Plasma characteristics
NASA MSFC, 415, 419, 420, 424 center of the sun, 264
Non-destructive evaluation (NDE) corona discharge, 264
techniques, 305 dielectric barrier discharge, 264
450 Index
Plasma characteristics (cont.) R

electron density Radioluminescence, 41
He I emission, 262 Real-time damage monitoring
stark broadening, 259, 260, 262 crack propagation, 401
electron energy, 263 cracks formation, 399, 400
ICP, 264 RC beam, 399, 400
laser, 264 system overload/deterioration, 401
mechanoluminescent discharge, 264 Real-time failure detection
microplasma jets, 264 compression zone, 397–399
parameters, description, 255 TCC, 393, 395
pressure, 256 TL emission, 396–397
temperature, 257–259 Reformatsky reaction, 253
types, 255 Reinforced concrete beam strengthening, 321
Plastico-TL, 5–7, 311, 383 Resonance broadening, 262
Plexiglass plate, 335 Resorcinol, 245
PMMA. See Poly (methyl methacrylate) Retrofitted concrete structures
(PMMA) beam performances, 405
PMT. See Photomultiplier tube (PMT) carbon fiber plies, 402
POF. See Polymer optical fiber (POF) debonding failure mode, 404, 405
Poly (methyl methacrylate) (PMMA) FRP, 402
chemical structures, 218 precracked beam, 402
lighting conditions, 221 Rochelle salt, 7
luminescence, 220
TL light yield, 226
Poly (phenyl methyl) siloxane (PPMS), 418 S
Polyatomic gases Saha–Boltzmann formula, 263
microcrack surfaces, 287 Scanning electron microscope (SEM), 359,
nitrogen lines, 283 389, 390
O2 atmosphere, 284 SEM. See Scanning electron microscopy
quencher concentration, 284 (SEM)
SO2, quencher, 286 Semiconductors, 18
terbium sulfate, 286 Sensor, 390–406
Polydimethylsiloxane (PDMS), 322 ITOF (see In situ triboluminescent optical
Polymer optical fiber (POF), 391 fiber (ITOF))
Polymers, EuD4TEA, 223–227 TLCOS, 385
cytoseal, 218 triboluminescence, 380, 384
and mounting media, 218 SHM. See Structural health monitoring (SHM)
outcomes Single-bubble cavitation, 240
photoluminescent emission Single-wall carbon nanotube (SWCNT), 356
spectra, 223 SL. See Sonoluminescence (SL)
TL, 223–227 Solar energy conversion
photoluminescent spectra, 224 DSSCs, 354
phthalocyanine (Pc), 218, 221, 225 3D wire-shaped flexible DSSC, 359–360
PMMA, 218 flat PV cell, 356–359
Positive holes, 18, 20, 21, 25–30, 32, 33, 35 flexibility, 354
PPMS. See Poly (phenyl methyl) siloxane FTO glasses, 354
(PPMS) liquid electrolyte, 354
Prognostic health monitoring (PHM), 326 measurement, 355–356
porous nanocrystalline TiO2, 354
PV sensors, 354
Q redox mediator, 355
QD. See Quantum dots (QD) Sonication, 296
Quantum dots (QD), 369 Sonoactivation method, 294
Index 451
Sonocrystallochemiluminescence He lines, 261

(SCRCL), 296 Lorentzian profile, 260
Sonofragmentation, 242 nonthermal plasmas, 260
Sonoluminescence (SL), 249 Structural health monitoring (SHM), 306,
cavern, waveguide, 99, 100 307, 382
cavitation, 98, 101, 103 catastrophic failure and deformation, 135
characteristics, 109 cementitious smart structures, 138–139
filamentous, 97 CFRP panel, 137
intensive light emission, 99, 100 fiber–matrix debonding and
light emission, 98 delamination, 136
liquids emit light, 95 fracturing/deformation energy, 134
MBSL and SBSL, 98, 99 GFRP, 137
MKT, 103 in situ monitoring, 136
multi-bubble sonoluminescence, 96, 99 intensity of emission, 132
Navier–Stokes equation, 102 iSHM, 351
photorecombination, 109 light collection/guiding technique, 136
Rayleigh–Plesset equation, 101, 103 mechanical agitation, 133
shock-wave theory, 103 mechanical shock, 352
spectrum optical damage signal, 137
MBSL, 101, 102 optical phenomenon, 351
SBSL, 101, 102 organic–inorganic composites, 134
ultrasonic waveguide, 97, 98 organic–inorganic hybrid composite,
ultrasound, 95 139–140
Sonotriboluminescence (STL), 280, 292 PL silica fibers, 137
cavitation shock waves, 292 PV effect, 353
organic crystals, 293 resin systems, 137
resorcinol, 294 sensor, 132
sonolysis, 293 smart composite matrix, 135
terbium acetylacetonate, 294 SNS, 352
SpectraSuite software, 434 solar cell, 353
Spectrographic methods, 263 state-of-the-art system, 133
Spectroscopy triboluminescent material, 132
electron transition and relative spectra vacuum infusion process, 133
arc plasma, 112 wire-shaped PV sensors, 352
atomic plasma, 110 ZnS:Mn, 133
Boltzmann’s distribution, 110, 113 Sucrose, 244
bound–bound transitions, 113 SWCNT. See Single-wall carbon nanotube
electromagnetic waves, 110 (SWCNT)
free–free transition, 114, 116 Synchronous cavitation, 240
FWHM, 112
irradiance ratio method, 111
linear approximation, 112 T
Maxwellian distribution function, 115 Target holder system
optical transition, 110 muzzle flash, 425
single-electrode discharge, 114 orbital debris characterization, 424
spectrum, corona discharge, 114, 115 photodiodes, 423
temperatures, 113 silicon detectors, 422
vibrational-rotational transition, 114 TCC. See Triboluminescent cementitious
method, 109–110 composite (TCC)
SpectrumProcessor, 436–438 TCO. See Transparent conducting oxide (TCO)
Stark broadening TEE. See Triggering electron emission (TEE)
Coulombic interactions, 259 Testing, 171
electron density, 261, 262 Tetrahedral complexes of manganese(II), 57
452 Index
The Advancement of Learning in 1605, 3 intensity, 169

Thenoyltrifluoroacetonate (tta ), 49 internal energy-level, 142
Thermalized population, 257 γ-irradiated alkali halides, 152
Thermodynamic equilibrium, 263 KCl crystal, 152
Thermoluminescence, 42 light emission, 142
Thomson scattering, 259 light gas gun, 419
TLCOS. See Triboluminescent cementitious light yield, 170
optical sensor (TLCOS) light yield vs. amount of added
TMCC. See Triboluminescent multifunctional food grade, 196
cementitious composite (TMCC) living organisms, 141
Transient cavitation, 240 luminescent material, 168
Transparent conducting oxide (TCO), 356 magnitude and location of damage, 147
Triboelectric field, 42 manganese doped zinc sulfide, 148
Triboelectricity, 104 masses and velocities, 412
Triboelectrification, 78 materials, 9–12, 148, 311
Triboluminescence (TL), 3–4, 8, 10, 17, 34, M-centers, 152
104, 162, 274–278, 292, 309, mechanisms, 152–153
425–427, 433 mechanoluminescent, 150
acoustic emission sensors, 141 meso-velocities (see Meso-velocity)
alkali halide crystals, 151, 152 monatomic inert gases, 280–283
cerium sulfate, 290 monitoring damage, conventional
characteristics, 151 sensors, 141
coordination compounds, 54 NASA MSFC gun, 420
design and fabrication, drop tower, nitrogen molecules, 126
160–162 oscilloscope settings, drop tower, 162
destruction, 148 outcomes, 148
dopant, 411 phosphors, 416, 417
drop kinetic energy, 413, 414 piezoelectrification, 384
drop tower measurements, 186 plastico-TL, 5–7
elastico-TL, 4–5 polyatomic gases, 283, 284, 286
electrical signals, 414 polymer foam, 148
emission, 147 powder materials, 148
emission ratio vs. drop number, 199 principle, 127–128
emission wavelengths, 184 quenching, mixed salts, 290–292
emission yields, 188 sensor systems, 147, 384
F-centers, 152 signal generation, 167
fracto-TL, 7–8 solid-state photoluminescent quantum
fracture of crystals, 150 yields, 385
friction, 126, 150 spectroscopy, 128–131
gas and solid-state components STL (see Sonotriboluminescence (STL))
Ce3+ ion, 275 target design and construction, 418–419
dysprosium(III) salts, 278 target holder system, 421, 423, 425
europium salts, 276, 277 TLCOS, 384, 386
europium sulfate, 277 triboluminescent emission, 125
gadolinium(III) salts, 278 types, 383
inorganic lanthanide salts, 275 Wint-O-Green® Lifesavers, 150, 412
nitrogen gas component, 274 wireless transmission, 141
PL, 274 ZnS:Mn (see ZnS:Mn)
praseodymium salts, 276 Triboluminescent cementitious composite
terbium(III) salts, 278 (TCC), 393, 394
hybrid composites, 142 elastic beam theory, 395
hypervelocity impact (see Hypervelocity) failure/ultimate stress, 394, 395
impact and electronic transformation, Triboluminescent cementitious optical
131–132 sensor (TLCOS)
Index 453
cement hydration, 389 magma-plasma, 66

CM-3000-DB, 386 mechanical energy, 66
custom-built impact rig, 385, 386 metal–dielectric contact, 79
mortar system, 385 metastable atoms collision, 82
properties, 385, 386 nitrogen gas discharge, 71
SEM, 389, 390 Paschen’s law, 82
ZnS:Mn crystals, 386, 388 photoionization, 82
Triboluminescent multifunctional cementitious pin-on-disk configuration, 68
composite (TMCC), 391 polymers
Triboluminescent sensors chemical activity, 83
artificial skin, 313 chemiluminescence, 83
biological systems, 306 dielectrics, 88–90
damage monitoring, 307–309 drying up process, 87, 88
delamination, 326 effect of gas, 85
emissions, 326 electrical insulators, 83
external excitation mechanisms, 308 electron emission and
fabrication and integration, 315–320 luminescence, 88
grading scale, 308 nitrogen and argon, 86
health monitoring, 307–308 oxygen and nitrous oxide, 85
impact sensor, 313–314 photon emission intensity, 85
integration of sensors, 317–320 polyamide (PA66), 86
optical fiber, 314–315 polypropylene, 84
SHM, 307 PTFE, 86
structural characterization, 320–326 radical reactions, 86
TL-doped composites, 315–317 tribo-contact, 85
ZnS:Mn, 314, 321 triboelectrification, 83
Triboluminescent sensory receptor (TSR), 391 twin-ring tribometer, 83, 84
Triboplasma potential distribution and electric
aluminum deposits, 74 field, 82, 83
BR, 68 radial and axial components, 80, 81
cathodoluminescence band, 74 real-color images, 68, 69
chemical composition, 73 regular/quasi-regular macroscopic
configuration, 67 geometry, 67
contact zone, 71 semicircle lobes, 68
dielectric solids, 78 spectral characteristics, 71
electric field, 79–81 spectral features, 70
electron bombardment, 74 spectrum, 71
electron triboemission, 81 sporadic postcontact emission, 82
emission, triggering electrons, 81 surface electric potential, 79, 80
emitted particles and photons, 66 Townsend’s model, 68
energy bubble, 65 tribo-contact, 67
energy spectrum, 72, 74 triboelectrification, 78
features, 67 triboluminescence time series, 74–78
interference pattern, 73 tribomicroplasma, 66
luminescence UV–vis emission, 78
cathodoluminescence, 73 vacuum, 74, 75
electron-impact, 73 Triggering electron emission (TEE), 81
extrinsic, 65 TSR. See Triboluminescent sensory
fracto, 73, 74 receptor (TSR)
fracture-induced, 67
gas discharge, 71
intensity, 70 U
intrinsic, 65 Ullman coupling reaction, 253
solid, 71 Ultrasonic methods, 309
454 Index
Uranium WIRD. See Wireless, in situ, real-time, and

concentration, 201 distributed (WIRD)
data, 204 Wireless, in situ, real-time and distributed
decay, 206 (WIRD) sensing, 9, 384
dibenzoylmethide triethylammonium, 202
drop height, 204
emission intensity, 201 Z
low concentration, 201 ZnS:Mn
luminescence, 205 ballistic triboluminescent impact
luminescent properties, 200 spectrum, 157
measures detector potential vs. impact velocity, 154
decay times, 204 GSFC one stage light gas gun, 156
triboluminescent emission light yield ratio vs. impact energy, 155
spectra, 205 low speed impacts, triboluminescent
outcomes, 206 emission spectra, 153
photoluminescent emission spectra, 203 photomultiplier detector response, 154
reduction, fluorescence intensity, 206 powder, 163
synthesis, 200 TL
three dimensional photoluminescent ballistic spectrum, 157
emission spectra, 203 identification of impacts, 157
TL-based impact sensor, 204 impact energy and emission, 155
TL yield vs. impact kinetic energy, 206 impact velocity, 153
triboluminescent light yield ratio vs. intensity, 153–156
dopant concentration, 202, 203 light detectors and oscilloscopes, 156
uranyl acetate and europium salts, 201 light ratio, 155
one-stage light gas gun, 156
photodetector, 153, 154
V powder, 157
Vacuum assisted resin transfer molding recorded, impacts, 157
(VARTM), 316 spectrum, 153
thermal effects/phase changes, 157
wavelength emission, 158
W triboluminescent spectra, caliber
Wint-O-Green® Lifesavers, 412 projectiles, 158

Handbook Triboluminiscence

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Handbook Triboluminiscence

Uploaded by

Copyright:

Available Formats

David O. Olawale · Okenwa O.I.

Ross S. Fontenot William A. Hollerman

ISBN 978-3-319-38841-0 ISBN 978-3-319-38842-7 (eBook)

Library of Congress Control Number: 2016944358

© Springer International Publishing Switzerland 2016

Printed on acid-free paper

This Springer imprint is published by Springer Nature

timely submission of the manuscripts. We are particularly grateful to our publishing

Tallahassee, FL David O. Olawale

8 Mechanoluminescence Induced by Acoustic Cavitation . . . . . . . . . 237

David O. Olawale, Ross S. Fontenot, Md Abu S. Shohag,

Light emission in a material can either be by incandescence or luminescence.

D.O. Olawale (*)

© Springer International Publishing Switzerland 2016 1

From mechanical forces?

Fluorescence Phosphorescence Chemiluminescence Bioluminescence

Fig. 1.1 Overview of different light emission mechanisms [1]

into fluorescence and phosphorescence. Thermoluminescence is light produced

the nitrogen lines, suggesting a secondary energy process. Furthermore, other

1.3 Elastico-, Plastico-, and Fracto-Triboluminescence

Triboluminescence may be divided in to three types: namely (1) elastico-,

Elastico-triboluminescence is luminescence produced during the elastic deforma-

The permittivity ε0 of the crystal is 8.85 1012 CNm2; consequently, an

Fig. 1.3 Schematic of

Plastico-triboluminescence is luminescence produced during plastic deformation of

Fig. 1.4 Energy level

dislocations. (4) The recombination of dislocation-captured electrons with the holes

where β ¼ 1=τd , τd is the lifetime of electrons in the dislocations band, P0 is the

The total intensity is given as

The total intensity is given as

Fraco-triboluminescence is luminescence produced due to the creation of new

I ¼ 2ðη1 α1 þ η2 α2 Þγwvt ð1:7Þ

I ¼ ½η1 α1 þ η2 α2 Q0 exp½α3 ðt tm Þ ð1:8Þ

where Q0 is the surface charge at t ¼ tm.

The peak TL intensity at t ¼ tm is given as

I m ¼ ðη1 α1 þ η2 α2 ÞγA ð1:10Þ

1.4 Triboluminescent Materials

Table 1.1 Triboluminescence of various materials [49]

triboluminescent emission yields for ZnS-based materials as a function of grain size

F.T. Freund (*)

© Springer International Publishing Switzerland 2016 17

2.2 Peroxy—The Stealth Defects

2.2.1 Band Structure and Electronic Charge Carriers

Any form of luminescence requires the formation of excited states, commonly

2.2.2 Peroxy Defects in Oxide Materials, Minerals,

2.2.3 Nature of Electronic Charge Carriers, Electrons,

2.3 Discovery and Validation of Peroxy Defects

2.3.1 Formation of Oxygen Anions in the 1 Valence State

Fig. 2.3 During thermal

2 MgðOHÞ2 , 2 MgO þ H2 O ð2:1Þ

However, during the thermal decomposition of ultrahigh purity Mg(OH)2 substan-

MgðOHÞ2 , Mg1δ ðOHÞ2δ O12δ þ δ H2 O ð2:2Þ

I. OH pairs adjacent to Mg2þ vacancy sites (expected to be most abundant).

Analysis of the IR absorption spectra of MgO single crystals containing residual

Equation (2.4b) leads to cation-deficient Mg1–δO with most OH converted to H2

Mg1δ ðOHÞ2δ O12δ , Mg1δ ðH2 Þδ O ð2:5aÞ

With H2 diffusing out as of Eq. (2.4c), a non-stoichiometric MgO is left behind:

Mg1δ ðH2 Þδ O ) Mg1δ O þ δ H2 ð2:5bÞ

If δ  1, we may also write the cation-deficient Mg1–δO as an MgO with excess-

2.3.2 Break-Up of Peroxy Defects and Formation of Positive

2.3.3 Peroxy Defects in Silica and Silicates

Water splits Hydroxyl pair Peroxy bond forma

H2O + O3Si-O-SiO3 O3Si-OH HO-SiO3 O3Si-OO-SiO3 + H2

Si-OO-Si bond Neighboring O2- Peroxy w/trapped electron

If δ 1, we may also write the cation-deficient Mg1–δO as an MgO with excess-