Professional Documents
Culture Documents
(Handbook of Pressure-Sensitive Adhesives and Products) Istvan Benedek, Mikhail M. Feldstein - Fundamentals of Pressure Sensitivity-CRC Press (2009)
(Handbook of Pressure-Sensitive Adhesives and Products) Istvan Benedek, Mikhail M. Feldstein - Fundamentals of Pressure Sensitivity-CRC Press (2009)
(Handbook of Pressure-Sensitive Adhesives and Products) Istvan Benedek, Mikhail M. Feldstein - Fundamentals of Pressure Sensitivity-CRC Press (2009)
PRESSURE
SENSITIVITY
FUNDAMENTALS OF
PRESSURE
SENSITIVITY
EDITED BY
ISTVÁN BENEDEK
MIKHAIL M. FELDSTEIN
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher can-
not assume responsibility for the validity of all materials or the consequences of their use. The
authors and publishers have attempted to trace the copyright holders of all material reproduced
in this publication and apologize to copyright holders if permission to publish in this form has not
been obtained. If any copyright material has not been acknowledged please write and let us know so
we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copy-
right.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that pro-
vides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.
TP971.F86 2009
668’.3--dc22 2008012198
6 Probe Tack
Costantino Creton and Kenneth R. Shull .................................. 6-1
7 Peel Resistance
Hyun-Joong Kim, Dong-Hyuk Lim, and Young-Jun Park ...........7-1
8 Shear Resistance
Sergey V. Antonov and Valery G. Kulichikhin ........................... 8-1
In recent years, pressure-sensitive products (PSPs) have reached a maturity that warrants
a detailed and critical examination of their science and technology. Based on our experi-
ence in both scientific activity and industrial areas, as well as on the special knowledge
of outstanding scientists and technologist as contributors, we have addressed all aspects
of pressure-sensitive adhesives (PSAs) in the form of a handbook. The huge volume of
data accumulated in this field over the past decade presents a delicate problem due to
the gap between the fundamentals of pressure-sensitive materials and their practice.
The application of PSAs requires a thorough knowledge of basic rheological and visco-
elastic phenomena. Adhesive and polymer scientists, however, are not often employed
as industrial managers or machine operators. Therefore, a need exists to investigate and
summarize the most important features of PSA technology and explain the phenomena
scientifically. This book covers the fields of manufacturing, conversion, application and
end uses of PSAs using a classic approach to compile a treatise based on the work of
various experts, theoreticians, chemists and engineers. The volume and diversification
of the data, as well as the boundary between theory and application, imposed the need
to impart our treatise in three books.
The destination of this handbook is twofold. On one hand, it is addressed to scien-
tists focusing on the fundamental processes underlying the complex phenomenon of
pressure-sensitive adhesion; on the other hand, it is intended for industrial researchers
who are involved in the practical application of these fundamentals for the develop-
ment of various products and specialists working in various end-use domains of PSPs.
Fundamentals of Pressure Sensitivity contains a detailed characterization of the pro-
cesses occurring in PSA materials at all stages of the life of an adhesive joint: its forma-
tion under compressive force, under service as the bonding force is removed, and under
adhesive bond fracture when the major type of deformation is extension.
Technology of Pressure-Sensitive Adhesives and Products describes particular fea-
tures of different classes of PSAs, such as rubber–resin-based adhesives, acrylics, and
silicones, and presents a discussion of the synthesis of pressure-sensitive raw materials,
their formulation, and the manufacture of PSAs and PSPs.
Applications of Pressure-Sensitive Products describes the main classes and represen-
tatives of PSPs, their competitors, end use, application domains, application technol-
ogy, and tests.
vii
followed by relaxation of the adhesive material as the bonding force is removed, and
then debonding as a detaching stress is applied. The mechanisms of PSA deformation
at each of these stages are different, and the contribution of PSA relaxation to adhesion
is also different. The strength of the adhesive joint requires the contribution of slow
relaxation processes, which imply the specific importance of both longer relaxation time
and large-scale structural rearrangements in the PSA material for proper adhesion.
We suggest that our readers use the list of abbreviations and acronyms in the end
of this book to facilitate the comprehension of various symbols, whenever they are not
sufficiently clear. The role of this book is to provide comprehensive and convenient up-
to-date information for users in both industry and academia.
We were pleased to see the participation of scientists and industrial experts, working
in very different areas of the field, on this book. We thank our contributors for their
efforts.
The Editors
Mikhail M. Feldstein, one of the world’s leading experts in the development of new
polymeric composites with tailored performance properties that span pressure-sensitive
adhesives and other materials designed for medical and pharmaceutical applications,
was born in 1946 in Moscow. In 1969 he graduated with honors from M.V. Lomonosov
Moscow State University, Faculty of Chemistry, and in 1972 he earned his PhD in poly-
mer science from the same university for the investigation of polyelectrolyte complexes
with ionic surfactants and lipids. His early research interests were associated with the
mechanisms of the formation and molecular structure of interpolymer complexes. Since
1972 he has worked in the industry of polymers for medical usage as a developer of
hydrophilic pressure-sensitive adhesives for skin application in transdermal therapeutic
systems and wound dressings. He received international recognition comparatively late:
his earliest contacts with colleagues beyond the borders of former Soviet Union date
to 1994 only. In 1999, a famous scientist and vice president of the Russian Academy
of Sciences, academician Nicolai A. Platė, invited him to join A.V. Topchiev Institute
xi
of Petrochemical Synthesis of the Russian Academy of Sciences, one of the most well-
known academic institutes in polymer science. Later that year, Feldstein established
long-term and large-scale research cooperation with a leading pharmaceutical company,
Corium International, Inc. (CA). In 2005, Feldstein earned his DrSc in polymer science
from the A.V. Topchiev Institute of the Russian Academy of Sciences.
Since the second half of the 1990s, Feldstein has focused on the molecular origins of
pressure-sensitive adhesion and the interrelationship between adhesion and other prop-
erties of polymer blends. Based on gained insight into the phenomenon of adhesion at a
molecular level, he has developed the first-ever technology for obtaining numerous
novel pressure-sensitive adhesives of controlled hydrophilicity and performance prop-
erties by the simple mixing of nonadhesive polymer components in certain ratios.
Feldstein is the author of nearly 200 research papers, 7 book chapters, and 25 patents.
He is a member of Adhesion Society and Controlled Release Society. Feldstein is also an
associate editor of the Journal of Adhesion.
xiii
1.1 Introduction
In accordance with the definition given by the Pressure Sensitive Tape Council, “pressure
sensitive” is a “term commonly used to designate a distinct category of adhesive tapes
and adhesives which in dry form (solvent/water free) are aggressively and permanently
tacky at room temperature and that firmly adhere to a variety of dissimilar surfaces
upon mere contact without the need of more than finger or hand pressure” [1]. This defi-
nition contains some uncertainties regarding the conditions of formation of the adhe-
sive joint, namely, the exact pressure and duration of its action. It is clear, however, that
1-1
where ρ is the density of the wetting liquid and ΔPc is the difference between the values of
capillary pressure inside the drop at the beginning and at the end of the inertial stage.
If the drop upon spreading has the shape of a spherical segment, ΔPc can be calcu-
lated as
1 1
Pc 2 (1.3)
Ri R*
where σ is the surface tension of the liquid and R i and R* are the curvature radii of the
drop surface at the beginning and at the end of the inertial stage of spreading, respec-
tively. It is convenient to express the capillary pressure difference, ΔРс, as a function of
the dynamic contact angle Θ. For a spherical segment of a constant volume,
1
3
Ri 4 2
sin 1 2 cos
R*
(1.4)
2 2
Expression of the spreading rate in the inertial stage can be derived from a combination
of Equations 1.2 through 1.4.
1
3
4 4 2
v
2
1 sin 1 2 cos (1.5)
2 2
in
Ri
The inertial spreading regime is usually observed upon the contact of drops of low-
viscosity liquids (hydrocarbons, water) with solid surfaces at ambient temperature.
0.8 180
1
150
0.6
120
Θ (degree)
r (mm)
0.4 90
2
60
0.2
30
0 0
0.000 0.005 0.010 0.015 0.020 0.025 0.030
t (s )
FIGURE 1.1 Dependence of the radius of the wetted area (1) and contact angle (2) on time for
water drop spreading on a steel surface.
At the initial stage of the spreading process, the surface of the drop is unstable and the
drop height changes unpredictably, which is reflected by a “zigzag” dependence of the
contact angle on spreading time. An example of such dependence is illustrated in Figure
1.1 for spreading of the water drop on a steel surface.
The duration of the inertial stage is very short—about 10 −2 s; therefore, for its reg-
istration high-speed shooting [several thousands frames per second (fps)] is needed.
At this stage the spreading rate does not depend on the viscosity of the spreading
liquid.
With the increased radius of the wetted area and reduction in the height of a drop, the
role of viscous resistance in a volume of the drop increases. The inertial stage transforms
into the viscous stage [7,8]. According to the hydrodynamic model, we can evaluate the
kinetic features of viscous spreading, such as the change in radius of the wetted area r
with time, by comparing the driving force of spreading,
where Θ 0 and Θ are the equilibrium and current values of the contact angle, respec-
tively, with the force of viscous resistance,
f r 2 grad v (1.7)
(v = dr/dt; η is viscosity). If the spreading drop has the shape of a spherical segment of
height h, the whole volume is involved in the spreading process and
1 dr
gradv (1.8)
h dt
If, otherwise, the drop spreads with a precursor fi lm, the main resistance is located in
the layer with thickness δ. As the first approximation, this thickness can be accepted as
constant and Equation 1.8 can be rewritten as
1 dr
gradv (1.9)
dt
0.25
16V
r A1t 0.25 t 0.25 (1.10)
0. 5
8 0.5
r A2t 0.5 t (1.11)
r ∼t
1
n
(1.12)
where n = 8–12. For polymer droplets n can be even higher [7,12,13]. The following
expression of slow spreading was proposed in [11].
0. 1
r A3t 0.1 V 0.3t 0.1 (1.13)
where κ is a correction factor, usually equal to ~10, that describes the increase of viscous
friction in a layer with the shape of a spherical segment in comparison with a flat layer.
The kinetic laws of the contact angle Θ changing can be deduced from the basic relaxa-
tion and rheological models of the spreading process [14]. Various versions of relaxation
models are based on the assumption that the rate of wetting d cosΘ/dt is proportional
to the deviation of the three-phase substrate/liquid/vapor line from the thermodynamic
equilibrium conditions that correspond to the equilibrium contact angle, Θ 0. In accord-
ance with this assumption,
cos 0 cos
e
t
(1.14)
cos 0 cos i
where τ is the retardation time of the wetting process and Θi is the initial contact
angle.
Upon application of this model to experimental results, the dependence, Θ(t), should
be linear if plotted in coordinates ln(cosΘ0 − cosΘ) − t. Such linear dependences were
experimentally obtained for spreading of dodecane–vaseline oil and water–glycerol
mixtures on steel plates. τ values were calculated from the slope of the experimental
curves [10].
The Kelvin–Voight model, with parallel connection of viscous and elastic elements
[15], can be used to calculate the retardation time of spreading. If such a system is loaded
and deformed up to relative deformation εi, the kinetics of deformation after release of
the load can be described by the equation
0
e
t
(1.15)
i 0
(1.16)
G
(1.17)
where λ is a characteristic linear dimension that is close to the radius of droplet Ri for
high-viscosity liquids. For low-viscosity liquids, the λ values are rather high (several
centimeters), which requires additional modification of the spreading model. So, the
rheological model of spontaneous spreading can describe the real process for high-
viscosity liquids only.
3.0
1
2.5
2
2.0
r/r0
1.5
1.0
0.5
0.0
0 10 20 30
t (s)
FIGURE 1.2 Spreading kinetics of PEG drops on the hydrophilic (1) and hydrophobic (2) sur-
faces of the release liner.
0.0
−0.5
Hydrophilic surface
ln(cosΘ0 − cosΘ)
Hydrophobic surface
−1.0
−1.5
−2.0
−2.5
0.0 0.5 1.0 1.5 2.0
(a) t (s)
0.0
−0.5
ln(cosΘ0 − cosΘ)
−1.0
−1.5
−2.0
−2.5
0 20 40 60 80
(b) t (s)
FIGURE 1.3 Evolution of contact angles upon spreading of PEG (a) and PI (b) drops on the
surface of the PET release liner.
Drops of PEG and PI spread on both surfaces (Figures 1.5 and 1.6). Amphiphilic PEG
easily wets both surfaces, whereas hydrophobic PI wets the hydrophobic surface much
better. The contact angles reach equilibrium values after several hours for the hydro-
philic surface and after more than 1 day for the hydrophobic surface (Figure 1.7).
(a) (b)
FIGURE 1.4 Water drops on the hydrophilic (a) and hydrophobic (b) surfaces of the release liner.
(a) (b)
FIGURE 1.5 PEG-400 drops on hydrophilic (a) and hydrophobic (b) surfaces.
(a) (b)
Using Equation 1.17, it is possible to estimate the viscosity, η, of the spreading liquid
during the first stage of the spreading process. The calculated values are presented in
Table 1.1. The values are quite reasonable, although they are essentially higher than the
values measured by conventional rotational viscometry (~0.1 Pa s for PEG and ~30 Pa s
for PI). This difference is understandable taking into account the adhesion contact of the
PSA with a solid surface. Regarding the real multicomponent, multiphase adhesives, we
80 80
Hydrophilic Hydrophobic
Θ (degree) 60 60
Θ (degree)
PEG-400
PEG-400
40 40
PI PI
20 20
0 0
0 500 1000 1500 0 500 1000 1500
(a) t (min) (b) t (min)
FIGURE 1.7 Time dependences of PEG and PI contact angles on hydrophilic (a) and hydrophobic
(b) surfaces of the release liner.
can expect much slower spreading, which is indeed observed upon PSAs application. Some
insight into a possible role of individual PSA components in the wetting–spreading proc-
esses can be derived from the analysis of the model situation for multicomponent liquids.
For precursor solutions with dissolved adhesive, the role of evaporation is significant
upon fabrication of fi lm on release liner or backing surface.
As mentioned previously, surfactants can play an important role in the development of
multiphase formulations. To understand their effect on wetting–spreading processes, the
consideration of model low-molecular-weight systems may be useful and important.
quartz, or sapphire). First, the liquid spreads due to the higher surface tension on a solid/
air interface compared with the sum of surface tensions at the liquid/air and liquid/solid
interfaces. Then, an adsorption layer of surfactant molecules at a solid/liquid interface
is formed, and conditions of wetting change. The liquid does not wet its own adsorption
layer and the reverse movement of the wetting front is observed.
P PC PH (1.18)
2
PC (1.19)
R
where σ is the surface tension of the liquid and R is the radius of the curvature of the
liquid surface.
This pressure causes deformation of the substrate under the drop, with the formation
of a dimple in the shape of a spherical segment. The depth of this dimple is proportional
to the ratio P/E, which is very small. An appreciable deformation can therefore be observed
only for soft materials with a low modulus of elasticity. The contact angle, measured as the
slope of the drop’s surface relative to the horizontal plane of the substrate, decreases over
the course of dimple deepening.
Another type of deformation is the formation of a ridge near the three-phase contact
line. As demonstrated by Bikerman [29] and later in a number of other papers [30–35],
this deformation is caused by the vertical component of the surface tension (i.e., σ sin Θ).
Although this deformation is typically small (<1 μm), for some systems (e.g., upon wet-
ting of agar or gelatin gels) the height of the ridge can reach 0.1 mm. The value h can be
calculated from the surface tension, contact angle, and elastic modulus of the substrate
[36] (Figure 1.8).
3sin
h (1.20)
E
Wetting of the agar gel by low-molecular-weight and polymeric liquids (water, PI, and
PEG-400) is illustrated in Figures 1.9 through 1.15.
h
2b
FIGURE 1.8 Gel deformation upon wetting in the vicinity of the three-phase line.
2 3
r (mm)
2
1
0
0.0 0.5 1.0 1.5 2.0
t (s)
FIGURE 1.9 Spreading of water (1) and PEG-400 (2, 3) drops on thin (1, 2) and thick (3) agar gel
plates. Agar content is 0.3%.
The initial stage of the wetting process was investigated using a digital camera (12 fps).
The time dependence of the wetted area and the shape of the interface was calculated
using a PC program for the analysis of video images. Static contact angles were mea-
sured using a horizontal microscope.
Agar gels with a concentration of 0.1–0.5 wt % are hydrophilic and rather soft . Water
drops coalesce upon these gels. The spreading rate of the water on the gel surface is high
enough—more than 32 mm/s. PEG drops spread slower compared with water and form
2.0
2
1.5 1
r (mm)
1.0
0.5
0.0
0 2 4 6 8 10
t (s)
FIGURE 1.11 Spreading of water drops on agar thin (1) and thick (2) gel plates. Agar content
is 1%.
FIGURE 1.12 Water drop on the surface of a thick agar gel plate. Agar content is 1%.
2.5
1
2.0
2
1.5
r (mm)
3
1.0
0.5
0.0
0 1 2 3 4 5
t (s)
FIGURE 1.13 Spreading of water (1), PEG-400 (2), and PI (3) drops on 1.5% agar gel.
(a) (b)
FIGURE 1.14 PEG (a) and PI (b) drops on the surface of agar gel. Agar content is 1.5%.
1.6 PEG-400
1.2 PI
r (mm)
0.8
0.4
0
0 10 20 30 40 50
t (min)
FIGURE 1.15 Spreading kinetics of PEG and PI drops on the hydrophilic surface of the
release liner.
nonzero contact angles (Figures 1.9 and 1.10). In all cases, the soft gel deforms under the
drops, forming intrinsic dimples and ridges.
The time dependence of the radius of the wetted area for the spreading of water drops
on thick and thin plates of gels is presented in Figure 1.11. During contact with thick
plates of gels, the volume of water drops decreases (by ~15%), whereas on the fi lm-like
gels the volume of the drops remains constant (within 1 min after initial contact).
The kinetics of water, PEG, and PI drops spreading on 1.5% thick gel plates is presented
in Figure 1.13. The unstable spreading process observed for PI drops can be explained by
the slow relaxation of the drop surface and by experimental difficulties connected with
drop detachment from a pipette tip.
Figure 1.14 represents the images of PEG and PI drops on the thick 1.5% gel (~10 min
after moment of the initial contact). The wetting processes for PEG (or PI) drops on thick
and thin gel plates are similar.
6sin a
a a 0 (1.21)
bE
6sin r
r r 0 (1.22)
bE
where Θa0 and Θr0 are the advancing and receding contact angles for the solid surface
with negligible deformation (the true contact angles) and 2b is the width of the ridge
(see Figure 1.8). As determined in Ref. 37, upon wetting of natural and butadiene rub-
ber by water and ethylene glycol Θa − Θr = 50 ÷ 70°. For harder polyolefin surfaces
[polyethylene(PE), polypropylene (PP), polystyrene (Pst) Θa − Θr = 10 ÷ 30°.
The spreading of drops on gels surfaces requires additional energy to overcome a vis-
cous dissipation of the substrate [32–40]. Spreading of the same liquids on rigid substrates
therefore proceeds essentially faster than on the gel surface; see, for example, the spread-
ing kinetics of water drops on steel (Figure 1.1) and agar gel (Figure 1.11) surfaces.
A similar result was obtained in Ref. 38: the contact angle on a quartz surface
reached its equilibrium value in 20 s, whereas on the soft epoxized natural rubber
(Е = 1.1 MPa), it takes 1 h for the drop to reach the equilibrium state.
The difference in spreading rate on soft and rigid surfaces is more pronounced for low-
viscosity liquids than for high-viscosity liquids. Thus, as discussed previously, spreading
water on a gel proceeds much more slowly than on a rigid surface. Figures 1.13 and 1.15
illustrate that PI and PEG-400 do not demonstrate a significant decrease in spreading
rate on the gel surface.
Data on PEG-400 spreading kinetics on gelatin and agar gels were used to estimate the
contact angle retardation time and PEG-400 viscosity, utilizing a previously described
approach—see Equations 1.14 and 1.17. The calculated values are presented in Table 1.2.
The comparison of PEG-400 viscosity values in Tables 1.1 and 1.2 illustrates that spread-
ing on a gelatin gel surface proceeds more slowly, with substantially higher apparent
viscosity.
Thus, spreading kinetics on deformable substrates (gels) is controlled not only by the
viscosity of the spreading liquid, but also by the rheological properties of the gel. Wet-
ting of the deformable substrates proceeds more slowly than that of substrates with a
high modulus of elasticity. The wetting process can be complicated by diff usion of a
liquid into a gel (e.g., for a system water–agar gel), which also slows down spreading.
Work Time
Normalized _ action (1.23)
Elementary _ work Elementary _ time
We assume that the extremums of this invariant product determine the range of PSA
application.
Let us explore the capabilities of the described approach for determining the influ-
ence of contact time on the durability of adhesive joints. A hydrophilic formulation
based on poly(N-vinylpyrrolidone) (PVP) and PEG with a PVP/PEG ratio of 64/36 was
chosen as the test system in our experiments. Th is ratio corresponds to the polycomplex
formed by hydrogen bonding between the terminal hydroxyls of PEG and carbonyls of
PVP [2,42]. Oligomeric PEG acts as a plasticizer (solvent) for PVP, enabling tackiness
of the formulation. Formation of the PVP–PEG polycomplex strengthens the structure
of the hydrophilic adhesive and improves its elasticity (recoverable strain) and cold
flow behavior. It is logical to assume that the durability of the adhesives cross-linked by
weak H-bonds is determined mainly by the kinetics of the H-bond dissociation, which
leads to flow of the adhesive.
To determine the intrinsic time characteristics of the formulation, creep and recovery
were measured at room temperature (Figure 1.16). At the creep stage, the sample was
squeezed from the gap between two parallel plates with constant load applied to the
upper plate. Then the load was released and the thickness of the sample was partially
recovered (recovery stage). The gap between the plates, h, as a function of time, t, is reg-
istered in this experiment.
Because the rubbery state of the PVP–PEG formulation can be characterized by two
processes (movement of polymer chain segments and relaxation of the H-bonds net-
work), it is reasonable to describe the creep function as a sum of two exponents:
h h0 A1e 1 A2e
t t
2 (1.24)
where h0 is the initial gap, A1 and A2 are constants, and τ1 and τ2 are characteristic times.
The obtained experimental data can be successfully described using Equation 1.24.
The values of calculated characteristic times are as follows: τ1 = 8 s and τ2 = 327 s.
1.1
σ = 69 kPa σ = 0 kPa
1.0
0.49
0.9
h/h0
0.8
0.7
0.6
Contact time Recovery
0.5
0 500 1000 1500
t (s)
400
h (µm)
200
Durability
0
0 50 100 150
Debonding time (s)
It can be assumed that the shorter time (τ1) characterizes segment movement, whereas
the longer time (τ2) should be related to the contribution of the H-bonds network.
To measure the durability of the adhesive joints, a probe-tack-type setup with a quartz
rod was used. The device allowed us to apply constant compressive or tensile force. The
adhesive fi lm was reliably fi xed on the lower (steady) plate. Then the rod was pressed to
the fi lm with a definite load for a specified time. Constant tensile force was applied to the
rod until failure of the adhesive joint. The testing conditions were as follows: tempera-
ture: 23°C; pressing stress: 4 kPa; apparent tensile stress: 75 kPa. The contact time under
pressure varied from 5 to 3600 s.
As illustrated by the deformation curves (Figure 1.17), at contact times comparable
with τ1, the failure of the adhesive joint proceeds via elastic mechanism without flow-
ing of the adhesive. At longer contact times the flow of the adhesive with formation and
breakage of fibrils is observed.
Durability, which was determined as time to failure of the adhesive joint under standard
detaching force, increases with contact time up to a virtually constant value. Therefore,
it can be concluded that there exists a critical contact time after which durability is no
longer sensible to this parameter. To describe this phenomenon, let us convert durability
θ and dwell time td to dimensionless values normalized by the highest relaxation time, τ2 .
These values are plotted in Figure 1.18, which demonstrates that for given experimental
conditions the durability of the adhesive joints under fi xed detaching force reaches its
maximum value at td ≥ 4τ2, that is, approximately 20 min. The maximum durability at
said conditions is approximately 0.5τ2 .
It can be concluded from the presented data that the least action principle sets restric-
tions on the time characteristics of formation and failure of adhesive joints: contact time
and durability.
0.5
0.4
0.3
θ/τ2
0.2
0.1
0.0
0 4 8 12
t d /τ2
FIGURE 1.18 Normalized durability of PVP/PEG formulation versus normalized contact time.
log τ (Pa)
8.0
log η 4.44
7.5 −2
4.40
7.0
log τ (Pa)
6.5 −3 4.36
5.5 −4
0 200 400 600 800 1000
Time (s)
FIGURE 1.19 Dynamics of the shear stress, shear rate, and apparent shear viscosity behavior
for PVP–PEG (36%) adhesive hydrogels over the time of squeezing under a compressive force of
1 N. (Feldstein, M.M., Kulichikhin, V.G., Kotomin, S.V., Borodulina, T.A., Novikov, M.B., Roos,
A., and Creton, C., J. Appl. Polym. Sci., 100, 522–537, 2006.)
compressive force at application, but to reach the maximum strength of adhesive contact
we must use a contact time comparable with the highest relaxation time.
The thickness of the PVP/PEG fi lm does not reach zero but tends toward a definite
constant value that is determined, as noted above, by the yield stress of the system.
The viscosity tends to infi nity upon approaching the yield stress value. The results pre-
sented in Figure 1.19 are of fundamental importance for the rheology of pressure-sensitiv-
ity because they demonstrate the change of shear stress, shear rate and PSA’s viscosity
under compressive load imitating bonding pressure.
References
1. Pressure Sensitive Tape Council. Glossary. http://www.pstc.org/technical/glossary.
php#glossary.
2. Feldstein M.M., Creton C. 2006. Pressure-sensitive adhesion as a material prop-
erty and as a process. In Pressure-Sensitive Design, Theoretical Aspects. Volume 1,
ed. I. Benedek, pp. 27–62. Leiden, VSP.
3. Good R.J., Chaundry M.K., van Oss C.J. 1991. Theory of adhesive forces across
interfaces (2). Interfacial hydrogen bonds and acid/base phenomena as factors
enhancing adhesion. In Fundamentals of Adhesion, ed. L.H. Lee. New York,
Plenum Press.
4. Good R.J., Chaundry M.K., 1991. Theory of adhesive forces across interfaces (1).
The Lifshitz – van der Waals forces of interactions and adhesion. In Fundamentals
of Adhesion, ed. L.H. Lee. New York, Plenum Press.
5. Chaundry M.K. 1996. Interfacial interaction between low-energy substrates.
Mater. Sci. Eng. R16:97–159.
6. de Gennes P.G. 1985. Wetting: statics and dynamics. Rev. Mod. Phys. 57:863.
7. Summ B.D., Goryunov Yu.V. 1976. Fiziko-khimicheskie osnovy smachivaniya i ras-
tekaniya (in Russian). Physical-Chemical Basis of Wetting and Spreading. Moscow,
Khimia.
8. Eustathopoulos N., Nicholas M.G., Drevet B. 1999. Wettability at high temperatures.
Amsterdam, Pergamon.
9. Soboleva O.A., Summ B.D., Raud E.A. 1989. Transition from inertial to viscous
spreading of a drop. Colloid Journal. 51:1204–1207.
10. Soboleva O.A., Raud E.A., Summ B.D. 1992. Rastekanie smesey uglevodorodov
po poverkhnosti stali (in Russian). Spreading of hydrocarbon mixtures on steel
surface Vestnik MGU Ser 2 33:42–46.
11. Soboleva O.A., Raud E.A., Summ B.D. 1991. The initial stage of drops spreading on
solid surface. Colloid Journal. 53:1106–1110.
12. Arslanov V.V., Ivanova T.I., Ogarev V.A. 1971. Kinetika rastekaniya polymerov na
tverdykh poverkhnostykh (in Russian). Spreading kinetics of polymers on rigid
surfaces. Dokl. Akad. Nauk SSSR. 198:1113–1116.
13. Vavkushevskii A.A., Arslanov V.V., Ogarev V.A. 1984. Rastekanie kapel’ poly-
merov po gladkim tverdym poverkhnostyam (in Russian). Spreading of a liquid
droplet over a solid horizontal surface. Kolloid. Zh. 46:1076–1081.
14. Raud E.A., Summ B.D. 1984. Adgeziya rasplavov i paika materialov (in Russian)
Adhesion of melts and soldering of materials. N12. p. 3.
15. Leaderman H. 1958. Viscoelasticity phenomena in amorphous high polymeric
systems. In Rheology. Theory and Applications. Volume II, ed. F.R. Eirich, pp. 1–61,
New York, Academic Press.
16. Cottington R.L., Murphy R.S., Singleterry C.R. 1964. Effect of polar-nonpolar
additivities on oil spreading on solids, with application to nonspreading oil. In
Contact Angle, Wettability and Adhesion. Adv. Chem. Ser. N43. pp. 341–354.
17. Bascom W.D., Cottington R.L., Singleterry C.R. 1964. Dynamic surface phenom-
ena in spontaneous spreading of oil on solids. In Contact Angle, Wettability and
Adhesion. Adv. Chem. Ser. N43. pp. 355–379.
18. Neogi P. 1985. Tears-of-wine and related phenomena. J. Colloid and Interface Sci.
105:94–101.
19. Pesach D., Marmur A. 1987. Marangoni effects in the spreading of liquid mixtures
on solid. Langmuir. 3:519–524.
20. Kulichikhin V., Antonov S., Makarova V., Semakov A., Tereshin A., Singh P. 2006.
Novel hydrocolloid formulations based on nanocomposites concept. In Pressure-
Sensitive Design, Theoretical Aspects. Volume 1, ed. I. Benedek, pp. 351–401. Leiden,
VSP.
21. Zorin Z.M., Romanov V.P., Churaev N.V. 1979. Influence of surfactants on quartz
wetting by electrolytes solutions. Colloid Journal. 45:1066–1073.
22. Princen H.V., Cazabat A.M., Cohen S.M.A., Heslot F., Nicolet S. 1988. Instabilities
during wetting process. Wetting by tensioactive liquids. J. Colloid and Interface
Sci. 126:84–92.
23. Zhmud B.V., Tiborg F., Hallstensson K. 2000. Dynamic of capillary rise. J. Colloid
and Interface Sci. 228:263–269.
24. Soboleva O.A., Summ B.D. 2001. Reverse flow of mixed surfactant solutions after
their capillary rise. Colloid Journal. 63:769–773.
25. Frank B., Garoff S. 1995. Origins of the complex motion of advancing surfactant
solutions. Langmuir. 11:87–92.
26. Kumar N., Varamasi K., Tilton R.D., Garoff S. 2003. Surfactant self-assembly
ahead of the contact line on a hydrophobic surface and its implication for wetting.
Langmuir. 19:5366–5373.
27. Afsar-Siddiqui A., Luckham P.F., Matar O.K. 2003. Unstable spreading of aqueous
anionic surfactant solutions on liquid fi lms.1. Sparingly soluble surfactant. Lang-
muir. 19:696–702.
28. Zisman W.A. 1964. Relation of equilibrium contact angle to liquid and solid con-
stitution. In Contact Angle, Wettability and Adhesion, Ј. Advances in Chemistry
Series. Am. Chem. Soc. Washington. N43. pp. 1–50.
29. Bikerman J.J. 1950. Sliding of drops from surfaces of different roughness. J. Colloid
Sci. 5:349–359.
30. Michaeles A.S., Dean S.W., Jr. 1962. Contact angle relationships on silica aquagel
surfaces. J. Phys. Chem. 66:1790–1798.
31. Rusanov A.I. 1975. Theory of the wetting of elastically deformed bodies. Colloid
Journal. 37:614–618.
32. Yuk S.H., Jhon M.S. 1986. Contact angle on deformable surface. J. Colloid Interface
Sci. 110:252–257.
33. Shanahan M.E.R., Carre A. 1994. Anomalous spreading of liquid drops on an
elastomeric surface. Langmuir. 10:1647–1649.
34. Carre A., Shanahan M.E.R. 1995. Direct evidence for viscosity-independent
spreading on soft surface. Langmuir. 11:24–26.
35. Shanahan M.E.R., Carre A. 1995. Viscoelastic dissipation in wetting and adhesion
phenomena. Langmuir. 11:1396–1402.
36. Extand C.W. 2006. Hysteresis in contact angle measurements. In Encyclopedia of
Surface and Colloid Science. Volume 4, ed. P. Somasundaran, pp. 2876–2891.
37. Extand C.W., Kumagai Y. 1996. Contact angles and hysteresis on soft surfaces.
J. Colloid Interface Sci. 184:191–200.
38. Carre A., Gastel J.-C., Shanahan M.E.R. 1996. Viscoelastic effects in the spreading
of liquids. Nature. 379:432–434.
39. Long D., Ajdari A., Leibler L. 1996. Static and dynamic wetting properties of thin
rubber fi lms. Langmuir. 12:5221–5230.
40. Chan L.W., Chow K.T., Heng P.W.S. 2006. Investigation of wetting behavior of
nonaqueous ethylcellulose gel matrices using dynamic contact angle. Pharm. Res.
23:408–421.
41. Chang E.P. 1997. Viscoelastic properties of pressure-sensitive adhesives. J. Adhesion.
60:233–248.
42. Feldstein M.M., Kulichikhin V.G., Kotomin S.V., Borodulina T.A., Novikov M.B.,
Roos A., Creton C. 2006. Rheology of poly(N-vinyl pyrrolidone) – poly(ethylene
glycol) adhesive blends under shear flow. J. Appl. Polym. Sci. 100:522–537.
2.1 Introduction
Unlike simple fluids, polymers have very slow dynamics so that even relatively far away
from their glass transition temperature, highly entangled polymer melts can diff use dis-
tances of the order of the coil size in minutes or more. Th is was recognized some time
ago1 and spurred the creation of the so-called interdiff usion theory of adhesion, which
stated that the adhesion between two polymers put in contact in melt increased with the
square root of the contact time. However, almost all the studies on that topic focused on
a situation in which the interface is formed in the melt state, but its mechanical strength
is tested after the system is cooled, in its glassy state.2–5
In many industrial processing applications such as coextrusion, molding, elastomer
processing prior to vulcanization, or pressure-sensitive bonding by tack, both interface
formation and some mechanical strength build-up must occur in the melt state.6–9 For
pressure-sensitive adhesives (PSAs), the substrate surface is often rigid and impenetrable
to polymer chains, but in certain cases the PSA can be bonded to an elastomer surface,
where chain interpenetration is a possibility.
2-1
Hamed and coworkers10–14 studied self-adhesion of elastomers using the classic peel
test technique. This technique allows the measurement of adhesive properties, but only
for long contact times. The technique is well adapted to filled elastomers, which have
extremely long relaxation times, but would not be the technique of choice for unfi lled
polymer melts due to its poorly defi ned geometry and difficulty in achieving short con-
tact times of the order of seconds. To overcome that limitation, Gent and Kim15 devel-
oped another apparatus to study the tack at short contact times. They studied the tack
of elastomers by measuring the impact and rebound velocity of a rigid pendulum with
an uncross-linked rubber sample at its tip impacting another elastomer sample. With
this experiment, very short contact times are accessible, but the contact time and the
debonding velocity are not independent and the contact area is not exactly known.
An alternative technique that combines the versatility of the peel test with the accessi-
bility of short contact times in a well-defined geometry is the probe test commonly used in
the PSA industry to evaluate the adhesive properties of sticky materials.16 As described in
detail in Chapter 6, a flat steel probe approaches the adhesive layer (which has been depos-
ited on a glass slide) at a constant velocity, applies a controlled compressive force during
a set contact time, and is then removed at a constant debonding velocity while a charge-
coupled device (CCD) camera allows observation of the debonding mechanism.17
With a few technical modifications, Schach and colleagues18,19 adapted this clas-
sic PSA technique to study the tack of uncross-linked elastomers. Two recent experi-
mental studies focusing on the self-adhesion and adhesion of a series of model linear,
high-molecular-weight polymers will be summarized in this chapter. Although these
polymers are not, strictly speaking, PSAs, it is nevertheless instructive and relevant to
examine their tack properties as an example of the effect of polymer chain interdiff usion
on tack. This chapter will be divided into two parts: the first part will focus on tack and
self-diff usion, that is, the gradual build-up of tack as chains mutually cross the interface
and entangle on the opposite side.
109 1000
SBR240K
G
SBR160K
108 tan (δ)
SBR80K
100
107
106 10
G ′ (Pa)
tan (δ)
105
1
104
103
0.1
102
101 0.01
FIGURE 2.1 Master curves of G′ and tan δ for the three SBR at a reference temperature of 20°C.
(Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
master curves. This is a critical parameter during the process of interface healing, as
demonstrated in neutron reflectivity studies.20 The reptation times, as well as other
important molecular or rheological parameters of the polymer, such as the viscosity at
vanishing rate or the plateau modulus, are listed in Table 2.2.
These master curves are essential for estimating the type of behavior expected from
materials during a tack experiment. Using the fi lm thickness and the debonding velocity
of the probe, an average initial normal strain rate for the test (dε/dt∼vd/h) can be esti-
mated, which indicates whether the entangled polymer has a typical elastic solid behav-
ior or a liquid-like behavior at the debonding velocity. Of course, this simple method
only gives an approximate estimate of the range of strain rates seen by the polymer layer
during the debonding process. In reality, the strain rate is highly inhomogeneous, both
spatially and as a function of time.
The probe test experiments were performed on a custom-designed apparatus based on
an MTS (Materials Testing Systems, Minneapolis, USA) 810 hydraulic testing machine.17
The typical probe test for this elastomer–elastomer situation can be divided into three
TABLE 2.2 Mechanical Properties of the Three SBR Determined Using the Master Curves
at Tref = 20°C
Molecular Weight
Polymer (g/mol) τd (s) η0 (MPa s) GN0 (MPa)
SBR80K 74,200 13 ± 2 1 ± 0.1 0.715 ± 0.15
SBR160K 144,900 140 ± 15 11 ± 2 0.75 ± 0.1
SBR240K 235,400 1170 ± 200 120 ± 20 0.73 ± 0.07
stages. In the first stage, a flat, stainless-steel probe with a silicon wafer coated with a
∼1-µm-thick elastomer layer glued on its end approaches a 200-µm-thick elastomer
layer grafted on a microscope glass slide. When a contact force of 70 N is reached
(corresponding to a nominal contact pressure of 1 MPa), the probe stops during a
contact time varying from 1 to 1000 s. The probe is then removed during stage 3 at a
constant debonding velocity varying from 1 to 100 µm/s.
Probe tack results are typically presented as tensile stress–strain curves measured
during debonding. The stress is a nominal stress obtained by normalizing the tensile
force by the maximum area of contact obtained during the compression stage, and the
strain is given by the displacement of the probe normalized by the initial thickness of
the fi lm. The adhesion energy (J/m2) is then calculated by multiplying the area under the
stress–strain curve (which is equal to the total dissipated energy per unit volume) by the
total thickness of the sample.
To obtain reliable results for polymer–polymer adhesion, the two polymer layers must
be strongly attached to the underlying rigid substrates. Th is problem was overcome by
using a reactive mercaptosilane layer on the rigid substrates, which could react with the
polymer and form covalent bonds. A thick layer was used on the glass slide and a thin
layer on the probe for two reasons. First, it was easier to prepare a thin layer by spin coat-
ing on a small surface of 1 cm2 because of edge effects. Second, the thin layer on the steel
probe deforms very little and effectively acts as a boundary condition for the debonding
and deformation of the thick layer. This becomes particularly important when different
polymers are used for the thin layer.
2.0
SBR240K, tc = 100 s, v = 100 µm/s
SBR80K, tc = 100 s, v = 1 µm/s
1.5 SBR240K, tc = 1 s, v = 100 µm/s
Stress (MPa)
1.0
0.5
0.0
0 1 2 3 4
Strain
FIGURE 2.2 Different types of tack curves observed with three different sets of experimental
conditions. (Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
2 tc = 1 s
1 tc = 10 s
6 tc = 60 s
4 tc = 1000 s
Stress (MPa)
0.1
6
4
0.01
0 1 2 3 4 5 6
Strain (ε)
FIGURE 2.3 Evolution of the tack curves with contact time in the high-velocity regime
(V = 100 µm/s) for the SBR240K polymer. (Schach, R. and Creton, R., J. Rheol. 52, 749, 2008.
With permission.)
regime, in which the polymer responds as a viscoelastic solid, and a low-velocity regime,
in which the polymer behaves as a viscoelastic liquid. These distinctions lead to very
different deformation patterns for the polymer during the failure process.
Figure 2.3 illustrates the evolution with contact time of the tack curves for the SBR
of molecular weight 240 kg/mol at a debonding velocity of 100 µm/s (viscoelastic solid
1.0
tc = 50 s
0.8
tc = 100 s
tc = 140 s
Stress (MPa)
0.6
0.4
0.2
0.0
0.0 0.5 1.0 1.5 2.0 2.5
Strain (ε)
FIGURE 2.4 Evolution of the tack curves with the contact time in the low-velocity regime;
SBR80K. (Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
regime). At short contact times, the work of adhesion is low, with a symmetric tack
curve characterized by a maximum stress of 1 MPa and a failure strain of 0.8. As the
contact time increases, the mutual diff usion process of the polymer at the interface pro-
ceeds, the interface is able to transfer higher stresses, and the observed peak stress and
failure strain both increase as the adhesion energy significantly increases. Furthermore,
the tack curves are not self-similar: at short contact times, the failure is very sharp, and
as the contact time increases they become asymmetric, characteristic of the bulk defor-
mation by cavitation of a thin, confined elastic fi lm.21 Thus, in the high-velocity regime,
there is a transition from an interfacial failure at short contact times to a bulk failure at
long contact times.
Figure 2.4 illustrates the evolution of the tack curves in the low-velocity regime
(molecular weight 80 kg/mol, vd = 1 µm/s). In this liquid-like regime, there is no effect
of contact time on the adhesion properties: the debonding rate is so slow that the poly-
mer chains had time to diff use to the point where the interface is indistinguishable from
the bulk, even for relatively short contact times. These liquid-like tack curves are very
similar to the tack results obtained by Poivet et al.22,23 on silicone oils. Figure 2.5 illus-
trates the evolution of the tack curves at long contact times (100 s for a molecular weight
of 80 kg/mol, i.e., a reptation time of 15 s) with debonding velocity. At low debonding
velocity, the tack curve is fluid-like (low maximum stress, followed by a stress plateau),
and when the velocity increases, there is a transition from this fluid-like behavior to a
solid-like behavior, as described for the high-velocity regime (asymmetric curves with
very high maximum stress and maximum strain).
1.2
v = 1 µm/s
v = 5 µm/s
1.0
v = 10 µm/s
v = 30 µm/s
0.8 v = 50 µm/s
Stress (MPa)
v = 100 µm/s
0.6
0.4
0.2
0.0
0 1 2 3 4
Strain
FIGURE 2.5 Evolution of the tack curves with the debonding velocity at long contact time (100 s)
for the SBR80K polymer. (Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
• At short contact times and high debonding rate, the failure mechanism is
interfacial and can be described by the propagation of a crack at the interface
between a rigid surface and a viscoelastic medium.
• At long contact times and high debonding rates, failure occurs in the bulk of the
fi lm by forced disentanglement, initiated by cavitation. This regime is indistin-
guishable from the fracture under tension of a thin confi ned layer of the same
polymer.
• At low debonding velocity, the layer fails by a fluid flow mechanism, fully
described using fluid mechanics tools and equations.
Figure 2.6 illustrates the variations in adhesion energy with time of contact for differ-
ent debonding velocities for the three SBR tested. These curves are directly correlated
with the transition between debonding mechanisms. Indeed, the variations of adhesion
energy with contact time are very different depending on both the debonding rate and
the molecular weight of the polymer, suggesting the existence of reduced parameters
describing the observed failure mechanisms.
In the high-velocity regime, Wadh increases until it reaches a maximum for contact
times of the order of magnitude of the reptation time. The adhesion energy is clearly
related to the degree of interdiff usion at the interface, and the saturation corresponds
to a fully healed interface. Figure 2.7 illustrates the variation in adhesion energy for the
three SBR at 100 µm/s. Figure 2.7 clearly demonstrates the trade-off between a good
6
5
4 V = 100 µm/s
3 V = 10 µm/s
V = 1 µm/s
2
6
5
4
3
10
1 10 100 1000
(a) tc (s)
6
5
4
3
2
Wadh (MPa)
100
6
5
4
3
V = 100 µm/s
2
V = 10 µm/s
V = 1 µm/s
10
100
6
5
4
3
10
2 3 4 56 2 3 4 56 2 3 4 56
1 10 100 1000
(c) tc (s)
FIGURE 2.6 Adhesion energy, Wadh, as a function of contact time for the autohesion of
(a) SBR240K, (b) SBR160K, and (c) SBR80K for three probe-debonding velocities. (Schach, R.
and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
1000
8 SBR80K
6 SBR160K
SBR240K
4
2
Wadh (MPa)
100 SBR240K
0.3
8 SBR160K
Stress (MPa)
6 SBR80K
0.2
4
0.1
2
0.0
0.0 0.5 1.0 1.5 2.0
Strain
10
1 10 100 1000
tc (s)
FIGURE 2.7 Adhesion energy as a function of contact time at high debonding velocity (Vd =
100 µm/s) for the three SBR. Comparison with the stress–strain curves of the three SBR in a tensile
test at constant cross-head velocity. The correlation between equilibrium adhesion energy and elon-
gational properties is evident. (Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
adhesion at short contact times, requiring sufficiently short and mobile polymer chains,
and a good adhesion at long contact times, requiring long polymer chains for cohesive
strength when the interface is fully healed. For the reported system the intermediate
molecular weight seems to strike the best compromise.
In the low-velocity regime, Wadh is independent of contact time because the debond-
ing is so slow that the interface is always fully healed (diff usion continues to occur dur-
ing debonding). This is the case for the SBR80K at 1 and 10 µm/s and for the SBR160K
at 1 µm/s. Figure 2.8 illustrates the variation of Wadh for these experimental systems. It
is worthwhile to note that the adhesion energy of the SBR80K at 10 µm/s and that of the
SBR160K at 1 µm/s is identical. For fluid systems,23 the relevant parameter controlling
the behavior of the thin fi lm should be the product Vdη, which is identical for SBR80K at
10 µm/s and SBR160K at 1 µm/s.
It is useful to define two reduced parameters to describe more generally the effect of
polymer interdiff usion on tack between polymer melts: first, the ratio of the contact time
to the reptation time (tc/τd), and second, the Deborah number, which is defined here as
the product of the average initial strain rate of the probe test and the reptation time of
the polymer,
vd
De d (2.1)
h0
with vd the debonding velocity and h0 the initial thickness of the sample.
Stress (MPa)
6 0.8
5
4
0.4
3
0.0
2
0.0 0.5 1.0 1.5 2.0
Strain
Wadh (J/m2)
100
6
5
4
2 SBR160K, V = 1 µm/s
SBR80K, V = 10 µm/s
SBR80K, V = 1 µm/s
10
1 10 100 1000
tc (s)
FIGURE 2.8 Adhesion energy as a function of contact time in the low-velocity regime.
Comparison of the tack curves obtained for the debonding of SBR80K at 10 µm/s and SBR160K at
1 µm/s. (Schach, R. and Creton, R., J. Rheol., 52, 749, 2008. With permission.)
Using a CCD camera to identify the failure mechanism for each set of material and
experimental parameters, a mechanism map can be constructed and is illustrated in
Figure 2.9 as a function of the two reduced parameters. For De > ∼3, failure is in the
high-velocity regime, where the polymer behaves as a viscoelastic entangled solid.
The transition between bulk failure and interfacial crack propagation depends also
on the De: the higher De is, the higher the critical ratio, tc/τd, becomes. For De < ∼3,
the polymer behaves as a viscoelastic fluid. In this low-De regime, tc/τd is no longer
relevant to describe the interface, which is fully healed during the debonding.
This study clearly demonstrates the significant effect of polymer interdiff usion on self-
tack. It is also clear that short mobile chains can diff use quickly but do not impart much
strength to the interface, whereas long chains are slow to diff use but strengthen the inter-
face much more significantly. Of course, these model studies were carried out on linear
chains and PSAs are always partially cross-linked. However, mobile chains exist in PSA
formulation and if there is some degree of miscibility, for example, between the PSA and
a release liner, blocking problems can arise. Also, when PSA need to be bonded to elasto-
meric substrates, some small degree of miscibility may help to increase adhesion.
The second conclusion that can be drawn from the study is that it is very difficult to clearly
attribute a mechanical reinforcement to an interdiff usion effect if one has not checked the
changes in macroscopic rheological response of the PSA. Highly viscoelastic materials such
as polymer melts can deform very differently when strained at different rates.
10
Bulk failure
Liquid-like failure
1
0.1
tc /d
0.01
Interfacial crack
propagation
0.001
FIGURE 2.9 Failure mechanism map as a function of the reduced contact time and the Debo-
rah number. ❍, Interfacial crack propagation; ❏, bulk failure; ∆, liquid-like failure. ●, liquid/
interfacial transition; ■, bulk/interfacial transition; ▲, bulk/liquid transition. (Schach, R., Tran,
Y., Menelle, A., and Creton, C., Macromolecules, 40, 6325, 2007. With permission.)
nonzero χ parameter, the interfacial width will be fi nite and at thermodynamic equilib-
rium will be given by25
1
w w ( N ∞) (2.2)
1 1
1 2 ln 2
N A N B
where
a
w ( N ∞) (2.3)
c
and a is the segment length and c is a constant that has a value of 6 or 9, depending on
whether the interface is in the weak or strong segregation limit.
The width of the polymer–polymer interfaces is usually measured by neutron reflec-
tivity, which is an ideal technique to measure interfacial widths27 ranging between
2 and 30 nm with a resolution of the order of magnitude of several angstroms. This tech-
nique is sensitive to gradients of the scattering length density, which depends directly on
the composition of the layers. Owing to the big difference in scattering length densities
between hydrogen and deuterium, it is possible to obtain very good contrast between
two polymers using isotopic substitution of hydrogen by deuterium.
2.3.1.1.1 Materials
The study was carried out mainly with a linear polybutadiene (PB; M = 420 kg/mol,
more than 80% 1,4) with a well-defined micro- and macrostructure (PB420K-H). The
adhesion properties of several elastomers with this PB were then investigated: the same
three linear SBR described in Figure 2.1, one linear SBR with another type of micro-
structure (36% styrene) and a molecular weight of 160 kg/mol, and three “industrial”
rubbers with some degree of branching: an ethylene–propylene–diene copolymer
(EPDM), a polyisobutylene (PIB), and a poly(dimethyl siloxane) (PDMS). The molecular
characteristics of the polymers are shown in Table 2.3. For neutron reflectivity experi-
ments, the interfacial width between each of these polymers and a deuterated PB with
the same chemical structure as PB420K-H but with M = 120 kg/mol (PB120K-H) was
measured. The relevant characteristics and nomenclature of the materials are summa-
rized in Figure 2.3.
1.0
PDMS
0.8 PIB
EPDM
SBR160K
SBR 36% Sty
φPB120K-D
0.6
0.4
0.2
0.0
−200 −100 0 100 200
(a) z (Å)
1.0
SBR160K
0.8 SBR80K
SBR240K
φPB120K-D
0.6
0.4
0.2
0.0
−150 −100 −50 0 50 100 150
(b) z (Å)
FIGURE 2.10 Volume fraction of deuterated monomer as a function of distance along the inter-
face for interfaces between a deuterated PB and various polymers: (a) various polymers with differ-
ent monomer composition and (b) three SBR polymers with identical monomer composition but
different molecular weights. (Schach, R., Tran, Y., Menelle, A., and Creton, C., Macromolecules,
40, 6325, 2007. With permission.)
characterizing the interfacial width, the interpenetration distance of the polymer chains
can be calculated by subtracting the broadening due to capillary waves18,29 and the final
results are summarized in Table 2.4.
The expected variations of interfacial width with molecular weight are observed for
the three SBR polymers with the same chemical structure (Figure 2.10b): the higher the
molecular weight, the sharper the interface.
Figure 2.4 summarizes the relevant characteristics of the interfaces studied here.
The sharpest interface is the PDMS/PB interface, with an interpenetration width, w,
of 16 Å and a very high Flory parameter value of 0.15. The PIB forms a wider inter-
face with PB (with w = 30 Å), giving a monomer–monomer interaction parameter of
0.04. The order of magnitude of the EPDM interface (82 Å) is closer to the radius of
gyration of the polymers (order of magnitude of 150 Å), with a χ of 0.006. SBR rub-
bers have the widest interfaces, with interpenetrations between 150 and 200 Å. The χ
parameter for the three SBR 40% polymers with PB is 0.0033, two orders of magnitude
less than the PDMS/PB. Finally, the SBR with a lower styrene content has the widest
interface, with 204 Å, and a Flory parameter with PB of 0.0023. Th is result is consis-
tent with the fact that the immiscibility between PB and SBR comes from the styrene
part of the SBR.
2
PDMS
PIB
1
8 EPDM
6 SBR160K
4 SBR 36% Sty
σ (MPa)
0.1
8
6
4
0.01
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
ε
FIGURE 2.11 Stress–strain curves obtained with the probe test at room temperature and
Vdeb = 100 µm/s for the interfaces between PB420K-H and polymers with different monomer
compositions. (Schach, R., Tran, Y., Menelle, A., and Creton, C., Macromolecules, 40, 6325, 2007.
With permission.)
times: at equilibrium the tack curves became independent of contact time. Equilibrium
contact time values varied between 300 and 2000 s, depending on the polymer–polymer
system.
Figure 2.11 illustrates the nominal tensile stress–strain curves obtained for the adhe-
sion measurement at thermodynamic equilibrium for a series of different interfaces. As
discussed earlier, for a probe debonding velocity of 100 µm/s, the thick PB layer behaves
like a viscoelastic solid and two types of debonding mechanisms can be observed: inter-
facial crack propagation (leading to a very sharp decrease in stress after the maximum)
and bulk cavitation (which leads to a more progressive decrease in stress).
The adhesion energy of PDMS on PB is very low, <1 J/m2. The PIB has a better adhe-
sion (22 J/m2), but the fracture remains brittle and apparently completely interfacial,
as determined by video observation, with very little bulk deformation of the PB layer
(maximum strain of 0.2). The fracture of the EPDM/PB interface is less brittle and the PB
layer is completely detached from the EPDM surface for a strain of ∼60%, with a begin-
ning of bulk fracture and an adhesion energy of 77 J/m2. Finally, the four SBR/PB sys-
tems are characterized by a bulk fracture behavior: fracture occurred in the PB layer, not
at the interface. The adhesion energy measured is very close to the fracture energy of the
PB, 120 J/m2. Note that the high-molecular-weight SBR and the SBR 36% Sty curves
demonstrate a pronounced tail in the deformation curve. Although it is tempting to
attribute this difference to the interface, one should remember that it is the PB420K-H
that deforms in these experiments. The difference is much more likely due to a slight
difference in the temperature at which the test was carried out and should not be inter-
preted further.
140
Cohesive fracture energy of PB420K-H
120
100
Wadh (J/m2)
80
60
40
20
0
0 50 100 150 200 250
w (Å)
FIGURE 2.12 Adhesion energy, Wadh, of interfaces between PB420K-H and different polymers
as a function of the interpenetration width at the interface w. (Schach, R., Tran, Y., Menelle, A.,
and Creton, C., Macromolecules, 40, 6325, 2007. With permission.)
Although two different polymers were used for neutron reflectivity experiments
and adhesion energy measurements, the chemical structure of the two PB was exactly
the same, the only difference being the molecular weight. Knowing the molecular
weights of all polymers and assuming that χ is independent of molecular weight,
neutron reflectivity results on the PB120K-D can be used to calculate the theoretical
interfacial width of the different polymers with the PB420K-H using Equation 2.2.
The measured adhesion energy can then be represented as a function of the calculated
interpenetration distance.
This comparison is illustrated in Figure 2.12 and clearly demonstrates that the adhe-
sion energy between the polymer layers depends strongly on the degree of interpenetra-
tion at the interface. Adhesion energy increases from a value of several joules per square
meter for a very sharp interface (10 or 20 Å) to the PB fracture energy for interfaces
larger than 150 Å. It is important to note for the significance of the comparison that
all samples except PDMS were tested in these experiments at a strain rate, which cor-
responds to their rubbery plateau, as is the case for PB.
2.4 Discussion
It is worthwhile to discuss these results more generally in light of the state-of-the-art
techniques regarding adhesion at soft polymer–polymer interfaces. It has been known
for quite some time that for fully miscible identical polymers, mutual interdiff usion at
the interface would lead to higher mechanical strength.2–5,33–36
Wrev
1
2
12 (2.4)
where G c is the critical energy release rate and φ(aTv) is a multiplicative factor depend-
ing on the dissipative properties of the soft polymer. Th is classic result obtained for
elastomers was later modified through the use of two important concepts: the role
played by the interpenetration of polymer brushes and the possible existence of
interfacial slippage. If the solid surface is functionalized with a layer of end-tethered
chains that is fully miscible with the soft polymer, the extraction of the chain from
the polymer during debonding contributes significantly both to the measured G c and
to the dissipative component φ(aTv), which is, however, still represented as a multi-
plicative factor.42,43
Interfacial slippage is important because the surface acts as a boundary condition
and if the surface is unable to sustain shear forces, the dissipative shear deformation in
the polymer layer during debonding is greatly diminished, reducing the multiplicative
factor φ(aTv) for the same values of Wrev and the same rheological properties of the poly-
mer. The groups of Chaudhury and Leger, in particular, demonstrated that resistance to
interfacial slippage could be more important than surface energetics and dominates the
macroscopic adhesive behavior44–48 and, therefore, Gc.
For very soft elastomers or polymer melts, however, the measurement of Gc as a purely
interfacial property becomes difficult because failure does not occur by crack propaga-
tion, but by a much more complex deformation pattern involving fi ngering instabilities
and the formation of fibrils49 that cannot be easily separated from the experimental
geometry. In this regime, the best way to compare the adhesive strength of different
interfaces is therefore to measure the total work of detachment in the same experi-
mental geometry. This parameter is defined here as Wadh. Few studies of these very soft
polymers exist and either focus on surface thermodynamics changes50,16 or report only
qualitative results.51
In the recent study summarized here, the interpenetration of polymer chains at the
interface acts by modifying the boundary condition at the interface: the deeper the inter-
penetration is, the higher the interfacial stresses that can be sustained by the interface.
1/ 2
* c
G
(2.6)
Eh
with E representing Young’s modulus of the material and h the thickness of the layer. It
can be used here to estimate Gc for three of our samples, PDMS, PIB, and EPDM, which
indicate an interfacial failure mechanism.
Using Equation 2.4 and the fact that the interfacial tension, γ1,2, is directly related to
the interface width, one obtains the following.
kTa2
1,2 kTa (2.7)
6 6w
Wrev can be estimated by the neutron reflectivity measurement to calculate the inter-
facial tension of the polymer–polymer interfaces and data from the literature for the
surface energy of the polymers. Figure 2.13 illustrates the calculated Gc value versus the
thermodynamic work of adhesion Wrev . There is no obvious direct correlation between
the thermodynamic work of adhesion and the measured Gc.
A second point that is interesting to discuss is the extent of interpenetration neces-
sary to obtain a high mechanical strength. The fracture toughness between glassy poly-
mers increases with interpenetration until it reaches the bulk fracture energy for an
interpenetration value, which varies with the experimental system from 0.5 to 1.5 times
the average distance between entanglements, de, in the bulk polymers.36,52
Figure 2.12 demonstrates that, for polymer melts, the increase in adhesion energy
with interpenetration figure is more progressive and the maximum adhesion energy is
14
12
10
Gc (J/m2)
0
0 10 20 30 40 50 60 70
Wrev (mJ/m²)
FIGURE 2.13 Gc estimated from the probe test curves and Equation 2.6 versus Wrev. (Schach, R.,
Tran, Y., Menelle, A., and Creton, C., Macromolecules, 40, 6325, 2007. With permission.)
achieved for a deeper interpenetration than de: the order of magnitude of the entangle-
ment spacing is 20 Å for PB and 30 Å for the other polymers and the maximum value is
achieved for an interpenetration of 150 Å equivalent to four to five entanglement lengths
and comparable to the radius of gyration of the polymers.
These results are also consistent with the theoretically proposed argument 33,37 that,
for polymer melts, the adhesion energy should reach its saturation value when the inter-
penetration distance becomes of the order of the radius of gyration of the polymers and
the degree of entanglement at the interface is the same as that in the bulk. This predic-
tion follows from the fact that the force to extract a polymer chain in the melt should
increase continuously with molecular weight because entanglements play a lesser role in
transferring stress than in the glassy state.
For very high-molecular-weight polymers or temperatures close to Tg, chain fracture
rather than chain extraction may occur and modify this result, defining then a critical
interpenetration distance for optimum toughness that is lower than the radius of gyra-
tion of the chains. The situation would then become closer to that of polymer glasses.
2.5 Conclusion
The tack of uncross-linked polymer melts depends strongly on the De, defined as the
ratio of the average strain rate applied to the polymer over its terminal relaxation time.
Failure mechanisms can vary from a viscoelastic fluid fracture mechanism, where frac-
ture is initiated by cavitation of only one or two large cavities and can then be modeled
with a fluid mechanics approach, to an elastic, rubber-like mechanism in which fracture
is initiated by cavitation in an elastic medium (at a stress well correlated to the elastic
modulus of the rubber) followed by melt fracture in the filaments formed during separa-
tion. The transition from one type of mechanism to the other is observed when De > ∼3.
For the increased tack between identical uncross-linked polymers due to molecular
interdiff usion, the following results can be emphasized.
1. If the polymer is a viscoelastic fluid, the strength of the interface does not play a
very important role and the fracture energy is independent of contact time.
2. For De > 3, two situations arise:
a. For contact times much shorter than the reptation time of the polymer, the
interface is the weak point in the assembly and the failure mechanism is cavita-
tion at the interface, followed by the propagation of these cavities as interfacial
cracks. This mechanism can be well described by a viscoelastic fracture mecha-
nism analogous to the failure of interfaces between cross-linked rubbers.
b. For contact times of the order of the reptation time of the polymer, the chains
have time to interdiff use and the interface chains by bulk fracture, giving a
higher value of tack if the molecular weight of the polymer is higher.
In regime 2, the key aspect to understanding polymer tack is the compromise between
the mechanical strength of the interface (optimum for long, well-entangled chains) and
the fast build-up of strength (optimum for short very mobile chains).
The situation that is more relevant for PSA interfaces is that of partially miscible
polymers. The degree of interpenetration no longer depends on interdiff usion times (for
times longer than the reptation time), but is a thermodynamic property.
For these interfaces, the most important factor controlling tack is the degree of inter-
penetration at the interface. This result can be connected with theories based on wetting,
in which surface energies of the two polymers chiefly determine the work of adhesion.
Whenever the Flory χ parameter between two polymers is below 0.05, the mechanical
strength of the interface between the polymers in the molten state will roughly increase
with χ−1/2 until the interpenetration width reaches a value of the order of the radius of
gyration of the polymers.
The second important result from the experimental study relates to the mechanical
strength one can expect from entanglements in the melt. The level of interpenetration
necessary to retrieve the bulk strength of the interface between two polymer melts is of
the order of several entanglements, rather than one entanglement, as in polymer glasses.
This result is not unexpected because chains can be extracted much more easily from
polymers above their glass transition temperature than from glasses.
References
1. Voyutskii, S. S. and V. L. Vakula, The role of diff usion phenomena in polymer to
polymer adhesion, J. Appl. Polym. Sci. 7: 475–491, 1963.
2. Jud, K., H. H. Kausch and J. G. Williams, Fracture mechanics studies of crack
healing and welding of polymers, J. Mater. Sci. 16: 204210, 1981.
3. Prager, S. and M. Tirrell, The healing process at polymer-polymer interfaces,
J. Chem. Phys. 75(10): 5194–5198, 1981.
4. Kausch, H. H. and M. Tirrell, Polymer interdiff usion, Annu. Rev. Mater. Sci. 19:
341–377, 1989.
5. Wool, R. P., B. L. Yuan and O. J. McGarel, Welding of polymer interfaces, Polym.
Eng. Sci. 29(19): 1340–1367, 1989.
6. Wool, R. P., Molecular aspects of tack, Rubber Chem. Technol. 57: 307–319, 1984.
7. Wool, R. P. (1991). Welding, tack, and green strength of polymers. L. H. Lee (eds).
in: Fundamentals of Adhesion. Plenum Press.
8. Arzondo, L. M., N. Pino, J. M. Carella, J. M. Pastor, J. C. Merino, J. Póveda and
C. Alonso, Sequential injection overmolding of an elastomeric ethylene-octene
copolymer on a polypropylene homopolymer core, Polym. Eng. Sci. 44(11): 2110–2116,
2004.
9. Zhang, J. B., T. P. Lodge and C. W. Macosko, Interfacial slip reduces polymer-
polymer adhesion during coextrusion, J. Rheol. 50(1): 41–57, 2006.
10. Hamed, G. R., Tack and green strength of elastomeric materials, Rubber Chem.
Technol. 54: 576–595, 1981.
11. Hamed, G. R. and C. H. Shieh, Flow criterion for elastomer tack, Rubber Chem.
Technol. 55(5): 1469–1481, 1982.
12. Hamed, G. R. and C. H. Shieh, Relationship between the cohesive strength and the
tack of elastomers, J. Polym. Sci., Polym. Phys. Ed. 21: 1415–1425, 1983.
13. Hamed, G. R. and C. H. Shieh, Relationship between the cohesive strength and
the tack of elastomers: part II, contact time effects, Rubber Chem. Technol. 58:
1038–1044, 1985.
14. Hamed, G. R. and C. H. Shieh, Relationship between the cohesive strength and
tack of elastomers: part III, various elastomer types, Rubber Chem. Technol. 59(5):
883–895, 1986.
15. Gent, A. N. and H. J. Kim, Effect of contact time on tack, Rubber Chem. Technol.
63: 613–623, 1990.
16. Zosel, A., Adhesion and tack of polymers: influence of mechanical properties and
surface tensions, Colloid Polym. Sci. 263: 541–553, 1985.
17. Lakrout, H., P. Sergot and C. Creton, Direct observation of cavitation and fibrilla-
tion in a probe tack experiment on model acrylic pressure-sensitive-adhesives,
J. Adhesion 69(3/4): 307–359, 1999.
18. Schach, R. and C. Creton, Adhesion at interfaces between highly entangled poly-
mer melts, J. Rheol. 52(3): 749–767, 2008.
19. Schach, R., Y. Tran, A. Menelle and C. Creton, Role of chain interpenetration in the
adhesion between immiscible polymer melts, Macromolecules 40: 6325–6332, 2007.
20. Karim, A., G. P. Felcher and T. P. Russell, Interdiff usion of polymers at short times,
Macromolecules 27: 6973–6979, 1994.
21. Gent, A. N., Elastic instabilities in rubber, Int. J. Non-Linear Mech. Special Issue in
Honour of C.O. Horgan 40(2–3): 165–175, 2005.
22. Poivet, S., F. Nallet, C. Gay and P. Fabre, Cavitation-induced force transition in
confined viscous liquids under traction, Europhysics Lett. 62(2): 244–250, 2003.
23. Poivet, S., F. Nallet, C. Gay, J. Teisseire and P. Fabre, Force response of a viscous
liquid in a probe-tack geometry: fingering versus cavitation, Eur. Phys. J. E 15:
97–116, 2004.
24. Helfand, E. and Y. Tagami, Theory of the interface between immiscible polymers II,
J. Chemical Phys. 56(7): 3592–3601, 1971.
25. Broseta, D., G. H. Fredrickson, E. Helfand and L. Leibler, Molecular weight and
polydispersity effects at polymer-polymer interfaces, Macromolecules 23: 132–139,
1990.
26. Guckenbiehl, B., M. Stamm and T. Springer, Interface properties of blends of
incompatible polymers, Phys. B 198: 127–130, 1994.
27. Stamm, M. and D. W. Schubert, Interfaces between incompatible polymers, Annu.
Rev. Mater. Sci. 25: 325–356, 1995.
28. Agrawal, G. and R. P. Wool et al., Interdiff usion of polymers across interfaces,
J. Polym. Sci.: Part B: Polym. Phys. 34: 2919–2940, 1996.
29. Jones, R. A. L. and R. W. Richards, Polymers at Surfaces and Interfaces. Cambridge,
Cambridge University Press, 1999.
30. Webber, R. E., K. R. Shull, A. Roos and C. Creton, Effects of geometric confine-
ment on the adhesive debonding of soft elastic solids, Phys. Rev. E 68: 021805,
2003.
31. Josse, G., P. Sergot, M. Dorget and C. Creton, Measuring interfacial adhesion
between a soft viscoelastic layer and a rigid surface using a probe method, J. Adhe-
sion 80(1–2): 87–118, 2004.
32. Lakrout, H., C. Creton, D. Ahn and K. R. Shull, Influence of molecular features on
the tackiness of acrylic polymer melts, Macromolecules 34: 7448–7458, 2001.
33. Wool, R. P., Polymer Interfaces. Munich, Hanser Verlag, 1995.
34. Schnell, R., M. Stamm and C. Creton, Direct correlation between interfacial width
and adhesion in glassy polymers, Macromolecules 31: 2284–2292, 1998.
35. Schnell, R., M. Stamm and C. Creton, Mechanical properties of homopolymer
interfaces: transition from simple pullout to crazing with increasing interfacial
width, Macromolecules 32(10): 3420–3425, 1999.
36. Creton, C., E. J. Kramer, H. R. Brown and C. Y. Hui, Adhesion and fracture of
interfaces between immiscible polymers: from the molecular to the continuum
scale, Adv. Polym. Sci. 156: 53–136, 2002.
37. Aradian, A., E. Raphael and P. G. de Gennes, A scaling theory of the competition
between interdiff usion and cross-linking at polymer interfaces, Macromolecules
35(10): 4036–4043, 2002.
38. Gent, A. N. and J. Schultz, Effect of wetting liquids on the strength of adhesion of
viscoelastic materials, J. Adhesion 3: 281–294, 1972.
39. Andrews, E. H. and A. J. Kinloch, Mechanics of adhesive failure I, Proc. R. Soc.
London, Ser. A: Math. Phys. Sci. 332: 385–399, 1973.
40. Andrews, E. H. and A. J. Kinloch, Mechanics of adhesive failure II, Proc. R. Soc.
London, Ser. A: Math. Phys. Sci. 332: 401–414, 1973.
41. Maugis, D. and M. Barquins, Fracture mechanics and the adherence of viscoelas-
tic bodies, J. Phys. D: Appl. Phys. 11: 1989–2023, 1978.
42. Creton, C., H. R. Brown and K. R. Shull, Molecular weight effects in chain pullout,
Macromolecules 27: 3174–3183, 1994.
43. Deruelle, M., L. Léger and M. Tirrell, Adhesion at the solid-elastomer interface:
influence of the interfacial chains, Macromolecules 28: 7419–7428, 1995.
3.1 Introduction
There exist only a few physicochemical phenomena that are as multiform as the phenom-
enon generally referred to as adhesion. Sufficiently strong and stable interfacial adhesive
bonds between adhesive and substrate layers ensure successful adhesive performance of
various composite materials, reinforced plastics, glues, and paint and varnish protective
coatings, as well as polymer blends. Numerous fundamental and applied studies, as well
as experimental and theoretical investigations, demonstrate that the differences in the
strength of adhesive joints can be accounted for by the nature of contacting phases, the
composition and the structure of the substrate and adhesive, the thickness of the adhe-
sive layer, the contact area, and the roughness of substrate surface, as well as bonding
and debonding conditions [1,2].
In his works [3,4], Deryaguin repeatedly emphasized the duality of the meaning of
the term of adhesion. He wrote, “On the one hand, adhesion is generally understood as
a process resulting in establishment of a bond between two bodies, and the bond failure
requires application of external force. On the other hand, a debonding process is often
considered, and the energy required to separate the contacting bodies is taken as a quan-
titative measure of the strength of adhesive interaction.”
To eliminate this ambiguity, Deryaguin suggested the term sticking when referring to
the “process of establishment and progressive development of interfacial intermolecular
bond with time, whereas the term ‘adhesion’ should be used to designate the achieved
3-1
In summary, it can be claimed that the transition zone, formed as the result of
substrate–adhesive interaction, is one of the key elements of the adhesive joint. It
accounts for the state of the substrate surface, the phase state of the adhesive joint, and
the ability of the components to interpenetrate each other and form chemical or inter-
molecular interfacial bonds, etc. It is within the transition zone that various defects
are generated; its phase composition and supramolecular structure define the strength
of adhesive bonds, the mechanism of adhesive material deformation under debonding
stress, and the path of crack propagation.
However, none of the suggested classifications of adhesives takes into account the
structure and phase nature of transition zones. This is understandable because the tasks
of material science and engineering fi rst require determination of the type of adhesive
that is defined by the nature of bonded or welded materials, as well as the conditions
of adhesive joint formation and service. The tasks concerned with the identification of
transition zone structure usually gain priority at the second stage of the research, when
it becomes necessary to develop the design of the adhesion joint, forecast its behavior
under various processing and operation conditions, and optimize a formulation.
Attempts to classify transition zones of adhesive joints have been undertaken repeat-
edly. Lipatov [20] suggested the use of morphological characteristics of the adhesive–
substrate interface as the key feature of faceted classification, with many objects divided
into two major groups. The first group should include the systems with the transi-
tion zone represented by two layers, divided by a boundary surface, which differ by
morphology from the polymer in the bulk and have uniform chemical composition. The
second group should be formed of systems with a transition layer of nonuniform com-
position, represented by a spontaneously forming emulsion of one polymer within the
other polymer. According to Lipatov [20], the mechanism of transition zone formation
in the systems of first group should be defined as thermodynamic, whereas the mecha-
nism in the second group is colloid chemical.
Vakula and Pritykin [34] also discussed faceted classification of transition zones,
but in doing so, they noted three independent features—composition diff usivity, con-
formational inhomogeneity, and layer-by-layer distribution of chain packing density.
The authors of Ref. 35 took virtually the same point of view in discussing “structure
heterogeneities at the melt–substrate interface that propagate to the depth exceeding
400 µm; the orienting impact of the substrate; increase in concentration of more sta-
ble conformer; redistribution of copolymer units and admixtures near the interphase
boundary.”
However, when describing a vast variety of adhesive systems and materials, the speci-
fied classification features are insufficient. Therefore, in Refs 36–38, a complicated fac-
eted hierarchical classification was developed, in which parallel grouping of a vast array
of objects into independent classes is supplemented by sequential division of materials
into subordinate subclasses.
The structural and morphological characteristics of transition zones represent
the basic classification features of independent classes. Researchers suggested distin-
guishing the following types of systems: structure-mechanical, structure-gradient,
concentration-gradient, and systems of complex or composite structure. Within each
of these classes, one can make further divisions according to other criteria, such as the
phase or physical state of adhesive joint elements, the number of components, topologi-
cal and conformational parameters of macromolecular chains, the degree of nonequi-
librium, types of defects and distribution across the transition zone, etc.
This classification has two important properties. First, it takes into account the basic
processes that occur during the formation of adhesive joints in many systems. Second,
it presents information on specific features of the internal structure of the adhesion
contact zone that are eventually responsible for the deformation and strength param-
eters of adhesive joints.
~
~ 5 nm
~ 40 nm
~ 5 nm
~
~
Substrate Substrate
(a) (c)
Adhesive
Oxide layer
Substrate Substrate
(b) (d)
FIGURE 3.1 Schematic drawing of the oxide layer (a) at the surface of the aluminum alloy (sub-
strate) and consecutive stages of adhesive joint formation (b, c), and breaking (d) during peeling.
* For example, macroscopic oxide layers, specially formed at substrate surface [36].
TABLE 3.1 Energy of Fracture (G) of Adhesive Joints between Electrochemically Deposited Foil
and Epoxy Laminate during Peeling
Surface Topography of Copper Foil
Relief Schematic Profile G (kJ/m2)
The basic mechanism of transition zone formation for this class of substrates involves
rheological fi lling of the porous structure at the substrate surface using viscous-flow
adhesive. The stages of this process are schematically illustrated in Figure 3.1. It was
reported in Ref. 41 that effective concentration profi le over the section of transition
zone can be experimentally obtained with electron microscopy using the “wide screen”
technique at comparatively small magnification and large observation area. Contrary
to the transition zones of diff usion-gradient type, in which each iso-concentration
plane changes its space position, in the structure-mechanical transition zones a similar
change can be observed at the front-face part of the concentration profi le only. The rest
of the profi le remains invariable with time. This feature seems to play a major role in the
identification of the mechanism of contact formation between conjugated phases rather
than the growth kinetics of adhesive joint strength.
It is conventional to assume that fi lling a porous structure with adhesive results in
its “mechanical coupling, seizure with rough edges of substrate surface (Figure 3.1 and
Table 3.1)” and “increase in surface area of interphase contact” [42]. According to the
mechanical interlocking theory, these effects represent a major strength-controlling
factor in the adhesive interaction between joint components. Within the framework of
the Washborn approach, Equation 3.1 was proposed [42] to describe the kinetics of the
change of contact area, S, during joint formation between the substrate and adhesive,
Pt
S 2d 2 (3.1)
d d (1 t )2 (3.2)
where α and τ are numerical constants and d∝ is the pore size at t∝. It follows from these
equations that the change in the strength of adhesive joint, A, is determined by the
expressions
Pt
A ≅ 2d 2
∑ (Eini ) and A A0 A A et (3.3)
where Ei is the adhesive bond energy and ni is the number of ith type links per unit area
of interphase contact. Multiple studies demonstrate that Equations 3.3 present a satis-
factory description of a large body of experimental data, involving both mechanical and
chemical treatment of the substrate surface [5,42]. Equation 3.1 underlies the so-called
rheological theory of adhesion, whereas Equations 3.3 represent the basis of a “rheo-
adsorption” theory of adhesion. The latter theory is capable of predicting trends in the
behavior of adhesive strength under the change in both environmental conditions and
the nature of interphase surfaces.
Nevertheless, other points of view exist on the role of mechanical treatment of the
substrate surface, grade of roughness, and interlocking in adhesive–substrate interac-
tions. First, observed strengthening of the adhesive joint relates to the removal of weak
surface layers, impurities, and defects [5,23]. Second, the grade of roughness manifests
itself in the formation of a “microfibrillar” structure in the adhesive layer that results, in
turn, in increased energy dissipation within the zone of plastic deformation during the
fracture of this class of adhesive joints [44,45] (see also Applications of Pressure-Sensitive
Products, Chapter 7). Finally, it is time to abandon predominant erroneous ideas of
porosity as “the cause of infinitely high polymer adhesion towards porous substrates” [5].
This concept contradicts the results obtained in Ref. 33, which demonstrated that upon
certain limiting depth of pore impregnation, or porosity exceeding some critical value,
subsequent impregnation makes no sense, because with depths greater than the critical
L cr value, cohesive failure (Аcoh) occurs in bulky polymers. Arslanov [36] offered a crite-
rion allowing the definition of the relationship between the strength and the geometric
parameters of this type of adhesive joints:
Acoh Lcr
(3.4)
A R
where R is the pore radius and A is the shear strength of a system polymer–pore wall.
As follows from Equation 3.4, at Аcoh ≅ А, a pillar of interlocked adhesive pulled out
of the pore will have length of the order of the pore radius. The size of the structure–
mechanical transition zone seems to be confi ned to the length indicated.
120 µm
Crack
TABLE 3.2 Crystallinity (φ), Surface Tension (γ), and Phase State of Epitaxial Layers
of Crystalline Polymers
Source: Wu, S. in Polymer Blends (Paul, D.R. and Newman, S., Eds.), Vol. 1, Academic Press,
New York, 1978. With permission.
It was determined in Refs 43, 46, and 47 that the structure of transcrystalline layers
was inhomogeneous. Epitaxial layers that are in direct contact with substrate surfaces
have the most ordered structure (the highest crystallinity). As illustrated in Table 3.2,
different surfaces can produce different crystallinity degrees. Thus, for low-energy sur-
faces [PTFE and polyethylene terephthalate (PET)] the crystallinity degree tends to zero,
whereas for high-energy substrates (glass, metals) the crystallinity turns out to be very
large for many polymers of different chemical composition.
As the distance from the surface increases, the orthogonal orientation of lamellar lay-
ers and flat spherulites degenerates, their sizes decrease, and their structure transforms
to spherulitic (Figure 3.2). There are two approaches to explaining the formation mecha-
nism of transcrystalline and epitaxial layers. The first is based on information regard-
ing the change in the conformational set of macromolecular chains in the adsorption
monolayer, followed by the formation of specifically oriented crystallites → lamellas →
flat spherulites that form the macroscopic structure-gradient transition zone [35]. The
second approach implies that transcrystalline layers appear in cases when the substrate
surface initiates the formation of a large number of nucleation centers close to each other,
which results in the intergrowth of crystalline areas at right angles to the surface [43].
The efficiency of substrate surfaces is assessed, in the first case, in terms of their adsorp-
tion activity (member Σ(Eini) in Equation 3.3), whereas in the second case it is assessed
in terms of its nucleating capability [47]. For example, a transcrystalline structure can be
formed if the rate of heterogeneous nucleation is higher than the nucleating capability of
the polymer itself. Otherwise, these spherulites are formed in subsurface layers.
Fracture-mechanical studies present evidence that the failure of adhesive joints
with transcrystalline transition layer occurs at a significant distance from the inter-
face (see Figure 3.2). Crack propagation most often takes place in the interface region
(a)
0
1.0
1 2 3
log P/P0
P/P 0 (%)
0.6 (b)
−1
0.2
52 60
ϕ (% vol.)
−1 0 1 2
log t
FIGURE 3.3 (a) Kinetics of the changes in adhesion strength during the fracture of adhesion joint
tetrafluoroethylene-hexafluoropropylene copolymer–copper at various annealing temperatures:
443 (1), 473 (2), and 503 K (3). (b) The copolymer crystallinity−strength relationship.
between the columnar and spherulitic structure of the adhesive. As reported in Refs 46
and 47, it is in this region that internal stresses reach their highest values. The position
of this region depends on the adhesive joint history. Thus, as illustrated in Figure 3.3,
thermal annealing of structure-gradient transition zones in the tetrafluoroethylene–
hexafluoropropylene copolymer–copper system [48] leads, on one hand, to the growth
of adhesive crystallinity due to the decrease in transcrystalline structure area and the
shift of the disruption zone toward the interface. On the other hand, it leads to a drop
in adhesive joint strength during peeling. From the thermodynamic point of view, these
results mean that the structure-gradient transition zones are in a nonequilibrium state,
which must be taken into account when estimating the service life for this type of adhe-
sive joint. The process of secondary crystallization is likely to be responsible for gradual
transformation of the transcrystalline structure into a spherulitic structure. Calculation
of the kinetics of the changes in adhesion strength of these joints can be based on the
calibration curve of peeling strength against crystallinity, as illustrated in Figure 3.3b.
Similar structure-gradient transition zones are also formed if adhesives are in the
liquid crystalline (Figure 3.4a) or amorphous state (Figure 3.4c). In these cases, one must
speak of a different nature of structural unit that “imprints” information from a solid
surface [36–38]. For amorphous polymers, such a structural unit is a domain, and for
liquid crystals it is a one- or two-dimensional mesophase element.
Heteroepitaxy can be observed in such systems as crystalline polymer–amorphous
adhesives (Figure 3.4d) and amorphous polymer–liquid crystalline adhesives (Figure 3.4a).
This is especially pronounced in the case of random and block copolymers. For exam-
ple, Figure 3.4 illustrates the formation of geometrically different epitaxial structures
in ethylene–propylene copolymers and ethylene–vinyl acetate (EVA) copolymers lami-
nated with biaxially oriented polyethylene (PE) and polypropylene (PP) substrates. Each
of these structure-gradient transition zones has its own specific features, which include
zone width, degree of ordering, and behavior with the distance from the interface.
Polysulphone
LC polyester
200 nm
(a)
PE
PEU
500 nm PP
(b)
EVA 20
PE
500 nm
(d)
T (a) wi
T1 A B
UCST
(b)
Temperature
w1′ w1″
T2 A B
w1″
(c)
w1′
w1′ w1″
T3
A w1″ B
(d)
w1′
A wi B −x 0 +x
Composition Diffusion coordinate
FIGURE 3.5 Diagram of amorphous phase separation of the A–B binary system (a) and
concentration profi les in transition zones of adhesive joints at bonding temperatures: Т1 (b),
Т2 (c), and Т3 (d). А is the adhesive and В is the substrate.
If conjugated phases are above the upper critical solution temperature (UCST; Т1 >
UCST), as illustrated in Figure 3.5b, and observation time is restricted, then the transi-
tion zone of the adhesive joint coincides with the diff usion zone (DZ) of component
mixing. For such DZ a continuous change with time of the extension and composition
distribution is featured, with the changes controlled by the rate of translational diff usion
of macromolecules. It was reported in Refs 48 and 51 that in the whole concentration
range, structure organizations, formed within these DZs, are similar to the structure
organizations of polymer solutions.
If adhesive–substrate bonding occurs at Т2 < UCST (Figure 3.5c), then, depending
on the observation time, or to be more exact, on the ratio of the observation time (t) to
the time of diff usional relaxation t D (t D = L2/D, where L is the thickness of adhesive and
D is the diff usion coefficient), the transition zone may have various patterns in terms of
structure and concentration. If the interfacing phases have finite size and t >> t D, the
system has achieved its equilibrium state and the interface separating the coexisting
phases is part of the transition zone.
If the interfacing phases are represented by semi-infinite samples and t << t D, the
transition zone, as illustrated in Figures 3.5b and 3.5d, consists of at least four regions:
two DZs corresponding to component solutions in each other, the interface, and the
regions of epitaxial structures in the A and B phases.
T (a) wi
Melt
A B
T1
(b)
Tm
Temperature
w1′
T2 B
A III
w1″ w1′ (c)
I II
T3
Crystal
A B
III (d)
w1″ II
I
A wi B −x 0 +x
Composition Diffusion coordinate
FIGURE 3.6 Phase diagram of crystalline equilibrium in a binary system (a) and concentration
profi les in transition gradient zones at temperatures T1 (b); Т2 (c), and Т3 (d). A is the amorphous
adhesive and B is the crystalline substrate.
includes the zone of crystalline polymer solubility in an amorphous polymer (I), the
zone of amorphous polymer solubility in the amorphous phase of a crystalline polymer
(II), and the interface (III). At temperatures higher than the melting point, Tm, the pro-
fi le of the concentration distribution corresponds to the above-described profi le for the
systems with amorphous phase separation at T > UCST. For systems featuring liquid
crystalline equilibrium, there are two types of structure organization to be expected for
the transition zone. At temperatures higher than the isotropization point, the transition
zone is identical to the mutual diff usion zone. At Тm < T < Тiso, the transition zone is
represented by a superposition of three zones—the zone of LC polymer solubility in
contacting amorphous component, the zone of solubility of this component in the LC
phase, and the interface.
(a)
T
T1
Melt
w ′′′ Tm
w′ w ′′
T2 wi
Temperature
Crystal
A B
(b)
A wi′′′ B
wi′′
(c)
wi′
A wi B −x 0 +x
Composition Diffusion coordinate
FIGURE 3.7 Complex amorphous–crystalline equilibrium. Phase state diagram of a binary sys-
tem (a) and concentration profi les in transition gradient zones at temperatures T1 (b), and Т2 (c).
220
T (°C) 180
140
1
2
3
100
60
0 0.2 0.4 0.6 0.8 1
w
(a)
EVA PVC
(1/T ) × 103
2.0 2.2 2.4 2.6 2.8
−9
2
−10
In D (cm2/s)
1
Tg PVC
−11
−12
−13
(b)
FIGURE 3.8 Phase state diagram of the PVC−EVA system (a) and the temperature relation-
ship of partial interdiff usion coefficients (b). Designations: (а) 1, EVA 70; 2, EVA 30; 3, EVA 20.
Numbers correspond to the percentage of VA monomer units in the EVA copolymer. (b) 1, PVC
diff usion into EVA copolymer; 2, EVA copolymer diff usion into PVC. Arrows mark the positions
of PVC glass temperature and the change in diff usion coefficients.
(a)
200 nm
(b)
ω Cl
1
PVC BNR 40
0.5
−16 −8 0 8 16
(a) x (µm)
z
2.5
II
0.5 ∼ω ′1
−16 −8 I 8 16
−0.5 x (µm)
∼ω ′′1
III
(b) −2.5
FIGURE 3.10 Diff usion profi les in the PVC−BNR system in coordinates w − x (а) and z − x (b),
plotted using KαCl. Annealing time is 25 min; annealing temperature is 433 K.
Second, the above-described phenomena are most clearly demonstrated if the con-
centration profi les are presented in the coordinates of Equation 3.5 [50],
1
w i [1 erf ( z )] (3.5)
2
which is commonly used to describe the evolution of the concentration diffusion profiles in
two conjugated semi-infinite media. Here, wi is the concentration
___ expressed in a weight or
volume fraction, erf(z) is the Gaussian integral, z = x/√Dt is the tabulated value of the inte-
gral, D is the diffusion coefficient, t is the time of contact, and x is the diffusion coordinate.
After these transformations are performed, the concentration profiles of transition zones
take the shape of three intersecting straight lines (Figures 3.10 and 3.11). The slope angle
of the outer regions (II and III) of the concentration profi les changes with annealing time,
indicating the penetration of macromolecular diff usion fluxes into individual phases.
ω Cl
1
PVC EVA 30
0.5
−16 −8 0 8 16
(a) x (µm)
z
2.4
II 1.2
∼ω1′
−16 −8 8 x (µm)
16
I
∼ω1′′
−1.2
III
(b) −2.4
FIGURE 3.11 Diff usion profi les in the PVC−EVA 30 system in coordinates w−x (а) and z−x (b),
plotted using KαCl. Annealing time is 60 min; annealing temperature is 433 K.
The slope and experimental points of the middle profi le region remain unchanged for a
specified annealing temperature.
Third, of principal importance is the repeatedly confirmed experimental fact that the
coordinates of the points formed by intersections of straight lines II and III with inter-
phase boundary stay constant over time, varying only with change in annealing temper-
ature (Figure 3.11b). In addition, the change in temperatures of bonding and annealing
might always result in the same values of wi′ and wi″ at the interface. This implies the
reversibility of the process and equilibrium composition of the phases, which coexist at
the interface in the course of the interdiff usion of adhesive and substrate components.
Fourth, the described character of the composition evolution in diff usion zones of
contacting PVC and copolymer phases is sufficiently general. Similar dependences
have been described earlier in PVC−butadiene–nitrile rubber (BNR) (Figure 3.10),
PVC−polymethyl methacrylate (PMMA), polychloroprene−BNR, and others [49,50].
Each individual system displays its specific values of the compositions of coexisting
phases, the rates of isoconcentration plane movement, the width of the interphase
boundaries, and the entire transition zones.
It is also of fundamental importance that the diffusion coefficients are invariable with
diffusion time, contrary to the suggestion by Vasenin [21], according to whom the trans-
lation mobility of the macromolecules of the adhesive within the substrate phase should
decrease with observation time (Figure 3.12). The constant slope of the curves in Figure 3.12
is a direct indication of the invariability of the diff usion coefficients with time.
Thus, the PVC–EVA systems can be referred to as partially compatible systems with
UCST. The phase structure that is formed within the zone of adhesive–substrate contact
is quite complex. It includes the interface region, where a jumpwise concentration pro-
fi le is observed that describes the compositions of coexisting phases, and two diff usion
zones, formed due to the solubility of adhesive and substrate phases in each other.
12 1
x (µm)
8
2
0 4 8
(a) t (min1/2)
12
2
x (µm)
0 5 10 15
(b) t (min1/2)
FIGURE 3.12 Kinetics of the increase in transition zone width for the PVC–BNR 40 (a) and PVC–
EVA 70 (b) systems at annealing temperature of 453 K (PVC–BNR 40) and 393 K (PVC–EVA 70).
(1) Diff usion of BNR 40 macromolecules into the PVC phase and (2) diff usion of PVC macromol-
ecules into the copolymer phase.
4
2000
Cohesive
failure
1500 3
A (N/m)
1000 2 Adhesive
failure
1
500
FIGURE 3.13 Effect of dwell time and temperature on 180° peel strength of a PVC−EVA 30 sys-
tem. Contact pressure is 0.5 MPa; contact time is 20 min. After removal of bonding pressure, the
sample is annealed at a fi xed temperature for a specified dwell time. Тtest = 295 K. The temperature
of annealing is (1) 373, (2) 393, (3) 413, and (4) 433 K.
2000
Mixed Adhesive
Adhesive failure Cohesive failure
1500 failure failure
A (N/m)
3
1000
2
4
500
1
5
0 6
0 20 40 60 80 100
VA content (% wt)
FIGURE 3.14 Dependence of 180° peel force versus the content of VA monomer units (% wt)
for various bonding temperatures: (1) 373, (2) 393, and (3) 413 K. (4) EVA–steel coated with epoxy,
(5) EVA−steel, and (6) EVA−glass. Тtest = 295 K.
Applications of Pressure-Sensitive Products, Chapters 7 and 8). The peel strength reaches
its limit at a specified value of dwell time. The dwell time required to establish the maxi-
mum strength of adhesive joints decreases with the rise in temperature. Thus, for 373 K
it amounts to 350 min, whereas for 433 K it is 60 min. A similar relationship is valid for
copolymers of various composition.
For this group of adhesives, information on the relationship between adhesive prop-
erties and copolymer composition is of fundamental importance. The relevant data
are generalized in Figure 3.14. As the percentage of VA units in the polymer backbone
grows, the peel strength goes through a maximum for all adhesive joints obtained at
various temperatures and with various substrates. The position of the maximum cor-
responds to copolymers containing 70–80% wt VA monomer units.
All adhesive joints, especially those obtained at high temperatures, exhibit a similar
change in the mechanism of debonding with increased content of VA units. In the vicin-
ity of the maximum of adhesion, cohesive failure is always observed, whereas the copo-
lymers overloaded with either ethylene or VA units demonstrate decreased adhesion
with the adhesive mechanism of debonding (this phenomenon is used in the practice
of EVA-based, hot-laminated, self-adhesive fi lms; see also Applications of Pressure-
Sensitive Products, Chapter 7).
Information on the type of deformation and the mechanism of debonding is of great
importance for the interpretation and comprehension of the phenomenon of pressure-
sensitive adhesion. The results of two groups of studies are presented below. The first
group deals with the analysis of substrate and adhesive surfaces upon adhesive joint
failure and the second group with in situ analysis of deformation zones of these joints
during debonding (Figures 3.15 and 3.16).
Cohesive failure (Figures 3.15a and 3.15b) is associated with significant residual adhe-
sive material at the substrate surface. The morphology of adhesive residual upon debond-
ing is indicative of a plastic deformation of adhesive copolymers under peel stress. Upon
adhesive failure (Figure 3.15d), there are no traces of adhesive remaining at the substrate
surface. In both cases there must exist interpenetration of adhesive and substrate mate-
rials within the transition zone, confirming the conclusions made in Chapter 2.
As illustrated in Figure 3.16, large extension strain achieving 3000–4000% for EVA
70 and 500% for EVA 30 is observed in the course of the debonding process. Negligible
plastic deformation is observed for EVA 14.
The measurements of concentration gradients and determination of diff usion fluxes
of adhesive joint components through the interface, identification of the locus of failure,
and observation of the kinetics of peel strength growth suggest the following mecha-
nism to explain how diff usion affects adhesion. Figures 3.17 and 3.18 demonstrate that
adhesive strength relatively quickly achieves its limiting value, despite rather intense
continuation of the interdiff usion processes in the system. The results of morphological
studies demonstrate that adhesive joint failure is located in a specific area of the con-
centration gradient (Figure 3.18), the position of which is independent of the length of
diff usion path of PVC macromolecules within the EVA phase. This finding allows us to
conclude that kinetics of the growth of adhesive joint strength relates not only to the dif-
fusion processes, but also to the processes of Bikerman’s weak zone formation. Numerous
(a) (b)
(c) (d)
FIGURE 3.15 Microphotographs of the PVC surface upon peeling of PVC−EVA joints of
various compositions: (a) EVA 70, (b) EVA 40, (c) EVA 30, and (d) EVA 14.
Adhesive
EVA 70
Substrate
PVC
Stress
direction
FIGURE 3.16 Optical microphotograph of the deformation of EVA 70 adhesive under peeling
from the PVC substrate.
investigations performed by our group with similar systems allow us to make the fol-
lowing conclusion. Kinetics of the growth of adhesive joint strength is only informative
for the mechanism of adhesive joint formation when the kinetics of weak zone formation
is controlled by the interdiff usion process.
2500
A (N/m)
1500 3
500
2 2
Q (% wt/min)
1
1
t
0
0 200
Time (min)
FIGURE 3.17 The correlation relationship between kinetics of the growth of adhesive joint
strength (1) and kinetics of mass transfer of EVA 70 into PVC (2), and PVC into EVA 70 (3).
The amounts of interpenetrated EVA adhesive and PVC substrate are measured using an x-ray
microanalysis technique.
PVC 1 w Cl EVA
0.5
"Weak zone"
−30 0 30
x (µm)
Crack
FIGURE 3.18 Concentration distribution profi le in the gradient zone of the EVA–PVC
adhesive joint. An arrow marks the position of a “weak zone,” where crack propagation and
debonding occur.
T (K) T (K)
1
3
2 2
450 425
T∗ 1
3 400
400
375
350 350
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
BNR 40 w PFO BNR 40 w PFO
(a) (b)
FIGURE 3.19 Phase state diagrams of PFO−BNR (a) and BNR 40–PFO (b) systems at various
degrees of oligomer conversion (α) in the polymerization reaction. Designations: (а) 1, BNR 18; 2,
BNR 26; and 3, BNR 40; (b) conversion degrees 0 (1), 40 (2), and 50% (3). T* is the vulcanization
temperature. The numbers indicate the content of nitrile groups.
and time conditions, providing for adhesive setting and substrate vulcanization. In
addition, mixtures of phenol–formaldehyde Novolac oligomers with BNR are charac-
terized by phase state diagrams with UCST (Figure 3.19). The PFO–BNR 40 system is
of particular interest because its critical temperature is below the curing temperature
of the adhesive joint. Second, due to the fact that interdiff usion coefficients in these
systems are ∼5 × 10 −7 cm 2/s at the temperatures of structure formation within BNR,
it can be expected that mass exchange processes between the elements of adhesive
joint will be rather fast, resulting in the production of a concentration gradient transi-
tion zone.
In terms of geometric dimensions, a nitrile substrate represents an infi nite medium,
whereas the adhesive is a finite-size body. Several types of transition region can be
formed depending on adhesive joint contact time (Figure 3.20). At short contact times
(t), when L > (Dt)1/2, a diff usion zone with phase interface is formed in the PFO layer. At
longer observation times, when L << (Dt)1/2, the diff usion frontline reaches the PFO–
steel boundary. Further, the elastomer concentration in the adhesive layer increases,
implying PFO dissolution in the elastomer.
In actual systems, the process of PFO–elastomer bonding is accompanied by the for-
mation of three-dimensional networks within the adhesive and substrate phases. Pre-
liminary experiments demonstrated (Figure 3.19b) that elastomer cross-linking and
PFO curing lead to decreased mutual solubility [50]. Figure 3.19b demonstrates that,
with the increase in oligomer molecular weight (conversion degree), the heterogeneous
w1
L
w1′′
2
Steel
PFO BNR 40
w1′
1
(a) −2 0 2 ×10−2 (µm)
w1 w1
1 1
Interface Interface
Steel
Steel
0 0
(b) 0 x (c) 0 x
w1 w1
1 1
Steel
Steel
0 0
(d) 0 x (e) 0 x
FIGURE 3.20 The concentration distribution in the adhesive joint region steel–PFO–BNR 40
(a) at T = 423 K; t = 5 (1) and 50 min (2); w′ and w″ are compositions of coexisting phases. The
stages of adhesive joint formation include initial contact (b), saturation of the PFO layer with
elastomer (c), establishing a limiting rubber solubility w″ in BNR (d) at the steel interface, and the
formation of dilute PFO solutions in rubber (e).
region of the phase state diagram becomes more extended and UCST becomes higher.
This means that the setting reaction results in phase segregation if it occurs in the dif-
fusion zone in regions of component concentrations exceeding the limiting values of
mutual solubility. In bulky elastomer, PFO phase segregates, whereas in bulky PFO-
dispersed BNR phase arises. The schematic drawing in Figure 3.21 illustrates the pattern
of the relevant structure-morphological organization of the transition zone.
Experimental studies of transition zones in these systems provide support for the
described regularities in their structure formation. As follows from the data of local
x-ray microprobe analysis, in the cross-linked elastomer layer the PFO phase exists
and, alternatively, in cured PFO there are particles of the BNR dispersed phase. Frac-
tographic studies of the failure of steel–PFO–elastomer adhesive joints demonstrated
that the initial stages of transition zone formation can be associated with cohesive brittle
failure involving the phase of phenol formaldehyde adhesive (region I in Figure 3.22).
After long-time contact, the BNR elastomer diff usion frontline reaches the metal bound-
ary. As a consequence, specific interaction at the PFO–steel interface is disrupted [39],
1
PFO BNR 40
Steel
0.5
1 2
BNR
PFO
Steel
Crack
30
20
I II III
A (kN/m)
10
30 60
t (min)
FIGURE 3.22 The relationship between peel resistance and the contact time of joint formation
(the rate of adhesive curing) in rubber–metal products. Locus of failure: I, within the PFO
layer; II, within rubber; and III, along the interface with metal.
Steel Crack
PFO BNR 40
(a)
I (KαBr) 1
0.5
(b) 0 x (µm)
20 60 100
FIGURE 3.23 Crack propagation in a real rubber–metal adhesive joint (a) and the distribution
elastomer (contrasted with Br), recorded on a Br L-line (b). The dotted line indicates the initial
thickness of the adhesive layer. The arrow marks crack localization within the concentration pro-
fi le. In this region, the PFO phase strengthens the elastomer by a dispersing mechanism.
followed by adhesive peeling (region III in Figure 3.22). The highest value of the strength
of the adhesive joint is achieved at intermediate times of adhesive contact when crack
propagates in the elastomer phase (region II). Hence, the ratio of diff usion rate to the
rate of structure transformations in adhesive and elastomer substrate acts as a factor
that controls the structure and strength of the adhesive joint in the combined transi-
tion zone occurring in the steel–PFO–BNR adhesive joint. The best case occurs when
the phenol formaldehyde oligomer has fully reacted with functional groups at the metal
surface, the particles of the rubber elastomer phase have been formed in its bulk, and
the particles of the cured oligomer are formed in the bulky elastomer. This makes it pos-
sible to realize extensive adsorption interaction between the components of the adhesive
joint and provide plasticization of the glassy phase of the Novolac phenol formaldehyde
polymer, accompanied by dispersion strengthening of the BNR phase (Figure 3.23).
3.3 Conclusions
Classification of the transition zones in adhesive joints involving the adhesives of dif-
ferent types and phase state was presented. The fact that the kinetic curve of adhesive
strength reaches a state of saturation (see Figure 3.17) does not imply the completion of
mass transfer processes, as assumed in the diff usion theory of Voyutskii and Vasenin
[16,21]. This is indicative of the formation and stabilization of a weak zone, which is
responsible for adhesive joint failure under debonding stress.
The concept of transition zones in full measure presented in this chapter can be used
for the analysis and design of PSA joints using various substrates.
References
1. Deryaguin B.V., Churaev N.V., and Myller V.M. Surface Force. New York: Consul-
tants Bureau, Plenum Publishing Corporation. 1987.
2. Mittal K.L. The role of the interface in adhesion phenomena. Polym. Eng. Sci. 17(7),
467–473, 1977.
3. Deryaguin B.V., Krotova N.A., and Smilga V.P. Adhesion of Solids. New York:
Cons. Bur. 1978. 320p.
4. Deryaguin B.V. and Krotova N.A. Adhesion [in Russian]. Moscow: Izd. Akad.
Nauk USSR. 1949. 244p.
5. Kinloch A.J. Adhesion and Adhesives Science and Technology. London: Chapman &
Hall. 1987. 441p.
6. Adamson A. Physical Chemistry of Surfaces. New York: John Wiley. 1990. 770p.
7. Wake W.C. Adhesion and the Formulation of Adhesives. London: Applied Science
Pub. 1982. p. 89.
8. Feldstein M.M. and Creton C. Pressure-sensitive adhesion as a material prop-
erty and as a process, in: Pressure-Sensitive Design, Theoretical Aspects. vol. 1,
I. Benedek (ed.) Leiden: VSP. 2006. Chap. 2.
9. Arslanov V.V. and Ogarev V.A. Adhesive joints of light metals with polymers. Prog.
Org. Coatings 15(1), 1–31, 1987.
10. Rudoi V.M. and Ogarev V.A. Certain methods for investigation of surface layers of
polymers, in: Modern Physical Methods for Investigation of Polymers [in Russian].
Moscow: Khimiya. 1982.
11. Zubov P.I. and Sukhareva L.A. Structure and Properties of Polymer Coatings [in
Russian]. Moscow: Khimiya. 1982.
12. Deryaguin B.V. and Toporov Yu.P. Role of the electric double layer in adhesion.
Russ. Chem. Bull. 31(8), 1544–1548, 1982.
13. Voyutskii S.S., Kamenskii A.N., and Fodiman N.M. Direct proofs of self- and
mutual diff usion in the formation of adhesion bonds between polymers. Mech.
Composite Mater. 2(3), 279–283, 1966.
14. Morozova L.P. and Krotova N.A. Dokl. Akad. Nauk. USSR 115. 747, 1957.
15. Krotova N.A. and Morozova L.P. Dokl. Akad. Nauk. USSR 127. 141, 1959.
16. Voyutskii S.S. Autohesion and Adhesion of High Polymers. New York: Wiley Inter-
science. 1963. 212p.
17. Boiko Yu.M. Self-adhesion of amorphous polymers and their miscible blends.
Mech. Composite Mater. 36(1), 79–82, 2000.
18. Bikerman J.J. The Science of Adhesive Joints. 2nd ed. New York: Academic Press.
1968.
19. Basin V.E. Adhesion Strength [in Russian]. Moscow: Khimiya. 1981. 208p.
20. Lipatov Yu.S. Interfacial Phenomena in Polymers [in Russian]. Kiev: Naukova
Dumka. 1980. 260p.
21. Vasenin, R.M. Adhesion, Fundamentals and Practice. London: McLaren and Son.
1969.
22. Voyutskii S.S. and Vakula V.L. The role of diff usion phenomena in polymer-to-
polymer adhesion. J. Appl. Polym. Sci. 7(2), 475–491, 1963.
23. Kuleznev V.N. Polymer Blends [in Russian]. Moscow: Khimia. 1980. 303p.
24. Dillard D.A. and Pocius A.V. (eds.), Adhesion Science and Engineering: The
Mechanics of Adhesion. vol. 1. Amsterdam: Elsevier. 2002.
25. Pocius A.V. Adhesion Science and Engineering: Surfaces, Chemistry and Applications.
Vol. 2. Amsterdam: Elsevier. 2002.
26. Petrova A.P. in: Adhesive Materials Handbook. E.N. Kablov and S.V. Reznichenko
(eds.) [in Russian] Moscow: K i R. 2002. 196p.
27. Satas D. (ed.), Handbook of Pressure Sensitive Adhesive Technology. 3rd ed.
Warwick, RI: Satas & Associates. 1999. 1017p.
28. Kardashov D.A. Structural Adhesives [in Russian]. Moscow: Khimiya. 1980.
29. Adams R.D., Comyn J., and Wake W.C. Structural Adhesive Joints in Engineering.
2nd ed. London: Chapman & Hall. 1997. 359p.
30. Petrie E.M. Handbook of Adhesives and Sealants. New York: McGraw-Hill. 2000.
880p.
31. Pritykin L.M., Kardashov D.A., and Vakula V.L. Monomer Adhesives [in Russian].
Moscow: Khimiya. 1988. 172p.
32. Freidin A.S. Strength and Durability of Adhesive Joints [in Russian]. Moscow:
Khimiya. 1971. 352p.
33. Benedek I. Pressure Sensitive Adhesives and Applications. New York: Marcel Dekker.
2004. 747p.
34. Vakula V.L. and Pritykin L.M. Polymer Adhesion: Basic Physico-Chemical Principles.
New York: Ellis Horwood Ltd. 1991. 350p.
35. Povstugar V.I., Kodolov V.I., and Mikhalova S.S. Structure and Properties of the
Surface of Polymer Materials [in Russian]. Moscow: Khimiya. 1988. 192p.
36. Arslanov V.V. Doctor of Science thesis: Physical chemistry of forming and debond-
ing processes of composite transition zones in adhesive joints polymer−metal.
1989. 425p.
37. Chalykh A.E. in: Surface Phenomena in Polymers [in Russian]. Kiev: Naukova
dumka. 1982. p. 123.
38. Arslanov V.V. and Chalykh A.E. Current status and future prospects of the theory
of adhesive joints. Zashch. Met. 25(4), 547–554, 1989.
39. Rubtsov A.E. PhD thesis: Transition zones in polymer systems. 1992. 228p.
40. Pluedemann E.P. in: Industrial Adhesion Problems. D.M. Brewis and D. Briggs
(eds.) Oxford: Orbital Press. 1985. p. 148.
41. Chalykh A.E., Aliev A.D., and Rubtsov A.E. Electron Probe Microanalysis in Polymer
Investigation. Moscow: Nauka. 1990. 192p.
42. Gul V.E. and Kuleznev V.N. Structure and Mechanical Properties of Polymers
[in Russian]. Moscow: Labirint. 1994. 367p.
43. Wu S. Polymer Interface and Adhesion. New York: Marcel Dekker Inc. 1982. 630p.
44. Evans J.R. and Packham D.E. J. Adhesion 10(1), 39–47, 1979.
45. Pickham D.E. in: Developments in Adhesives: 2nd ed. A.J. Kinlock (ed.) London:
Applied Sciences Pub. 1984. p. 315.
46. Schultz J. Effect of orientation and organization of polymers at interfaces on
adhesive strength. J. Adhesion 37(1–3), 73−81, 1992.
47. Wu S. in: Polymer Blends. D.R. Paul, and S. Newman (eds.) Vol. 1, New York :
Academic Press. 1978.
48. Chalykh A.E. and Raisin I.B. J. Polym. Sci. Ser. A. USSR 16(5), 1068, 1974.
49. Chalykh A.E. Diff usion in Polymeric Systems [in Russian]. Moscow: Khimiya.
1987. 312p.
50. Chalykh A.E., Gerasimov V.K., and Mikhailov Yu.M. Phase Diagrams of Polymer
Systems. Moscow: Yanus-K. 1998. 216p.
51. Zagaytov, A.I., Chertkov V.G., and Chalykh A.E. Comparative study of kinetics of
structure formation in binary polymer systems and inorganic glasses J. Mol. Liq.
93(1–3), 177–180, 2001.
52. Bikerman J.J. Physical Surfaces. New York: Academic Press. 1970. 476p.
53. Chalykh A.E. and Gerasimov V.K. Phase equilibria and phase structures of polymer
blends. Russ. Chem. Rev. 73(1), 59–74, 2004.
54. Gorbunov A.D. Master of Science thesis: Effect of miscibility and diff usion on
adhesion performance in PVC−EVA co-polymers systems. 2004. 144p.
55. Aliev A.D., Vokal M.A., Chalykh A.E., and Gerasimov V.K. Phase diagrams of
ethylene-vinyl acetate copolymers. Polym. Sci. Ser. A 48(12), 1281–1286, 2006.
56. Kalnin’ M.M. Changes in the interface in the adhesive reaction of polyethylene
with steel. Mech. Composite Mater. 26(5), 571–576, 1991.
57. Kalnin’ M.M. Adhesive Interaction of Polyolefins with Metal [in Russian]. Riga:
Zinatne. 1990.
4.1 Introduction
The use of pressure-sensitive adhesives (PSAs) has been steadily increasing in a large
domain of industrial applications within the past 30 years [1] (see also Applications of
Pressure-Sensitive Products, Chapter 4). This evolution is particularly due to the steadily
4-1
improving design of efficient and custom adhesive formulations resulting from a better
knowledge of the relationship between the relevant adhesive properties (such as rheo-
logical or interface properties) and the composition of the adhesive (primary polymer
components, resins, and additives; see also Technology of Pressure-Sensitive Adhesives
and Products, Chapter 8). At the stage of bond formation, soft (highly viscoelastic) adhe-
sives mainly possess shear deformation, whereas upon debonding, tensile strains may be
dominant, when the elongation of the adhesive layer along the direction of the applied
detaching force may achieve levels from a few hundred to a few thousand percent. The
viscoelastic behavior of the adhesive then plays a leading role during bonding and
debonding stages.
Recent experimental developments have focused on measuring the physical proper-
ties of the adhesive and the visualization of its detachment from the substrate [2,3],
which helped to understand more precisely the individual and coupled effects of the
main components of the adhesive. The use of model adhesives allows the defi nition of
a predictive relationship with their linear viscoelastic properties using relevant models
of molecular dynamics. This approach allows the rationalization of PSA formulation
through the use of analytical models that take into account the individual and cou-
pled effects of the various components (“virtual formulation”) [4,5]. Despite important
breakthroughs in industrial applications, some fundamental points are not yet resolved.
In the case of strong affinity between a PSA and its substrate, the rheological behavior
of soft adhesives governs, to a large extent, its adherence performance. However, when
the PSA presents a low affinity with the substrate, the rheological properties of the PSA
are no longer the governing factor of the adherence properties [6]. What are the driving
forces governing detachment? These are interesting new research areas for the adhesion
science arena in which fundamental studies take place. Some important features are
as follows:
1. In the case of large adhesive deformation, the rheological behavior in the nonlin-
ear domain of viscoelasticity is a major factor governing the mode of detachment
when cavitation takes place, creating a fibrillar structure in the bulk of the PSA [7].
Is it possible to measure, understand, and finally predict the rheological behavior
in the nonlinear domain for adhesives? Some examples will be presented here,
specifically in the field of PSAs.
2. The structure formation of surfaces may improve and help to control adhesion
properties, sometimes by mimicking the natural stickiness of plants or animals.
Structure formation of surfaces at the micro- or nanoscale level seems to be a
promising research domain to understand and fine-tune adhesion phenomena
in practical applications. This is what we may define as modulation of adhesion.
The nano or micro structure formation may also be obtained by chemical modi-
fication [8].
In this chapter we will first describe the general features of the relationships between
rheological and adherence properties and will explain how a better understanding of
these features can be used to improve PSA formulations for a wide range of applica-
tions. We will demonstrate that tack measurements, particularly the visualization of the
1. The thermodynamic work of adhesion (W0), which is the energy required to sepa-
rate reversibly the interface between two bodies in contact, from their equilib-
rium state to infinity [9].
2. The fracture energy (G), which can be defined as the energy required to create a
unit surface of fracture.
The relationship between W0 and G is particularly complex for soft adhesives because
of the energy dissipated irreversibly during the debonding stage. A general expression
derived from Ref. 10 is often proposed to link W0 and G:
where the function, Φ, is an amplifying factor that depends on the temperature, the
rate of debonding, and more generally on all parameters that modify the viscoelastic
properties of adhesives. G may be 100–10,000 times larger than W0 in the case of soft
adhesives.
In this chapter, we will distinguish between adhesion performance with respect to
surface properties and adherence performance with respect to the energy (or force)
necessary to break an assembly.
General trends of the relationship between rheological behavior and adherence prop-
erties have been reported in a large number of cases. The pioneering vision of the dif-
fusion and relaxation processes of flexible macromolecular chains of de Gennes [11]
provided polymer scientists with effective and predictive models of the viscoelasticity
of polymer melts. As a consequence, some authors proposed a kind of “predictive for-
mulation” in the case of polymer-based soft adhesives by inverting molecular models of
viscoelasticity [12,13].
In 1996, de Gennes proposed a simple theoretical picture to explain the strong adhe-
sion properties of weakly cross-linked rubbers, known as the trumpet model [14]. Based
on what Gent and Petrich describe as “the first transition,” this model qualitatively
describes the transition between the cohesive fracture domain and the first interfacial
fracture domain observed in peeling experiments of cross-linked butadiene–styrene
rubbers adhering on a polyethylene terephthalate polyester fi lm [15]. Even if this model
may seem crude in comparison with the complex behavior observed during the detach-
ment stage, it describes the observed general features and indicates clearly that the rheo-
logical behavior may be dominant when fracture takes place. The trumpet model details
the failure process for a soft model adhesive that presents a single relaxation time and
two levels of storage modulus, as presented in Figure 4.1a. We do have the general fea-
tures of this type of signature for lightly cross-linked elastomers typically used for soft
adhesive applications. When failure takes place within this type of adhesive, the stresses
relax along the crack that opens. De Gennes postulates that the form of the crack (the
opening of the failure, U) depends on two main viscoelastic parameters of the adhesive:
the relaxation time (τ) and the ratio between the two characteristic (hard/soft) storage
moduli. This model demonstrates that the total length of the crack that totally relaxes
the stresses may correspond to the thickness of the adhesive (W). This model refers to
soft solid like adhesive with a single relaxation time. In that case the form of the crack
looks like a trumpet as shown in Figure 4.1b.
G inf
G (au)
G0
U (w)
O o
V X
l1
V.τ
V.τ.Ginf/G0
(b) W
FIGURE 4.1 Crack shape during the detachment stage for a model soft adhesive. (a) Relaxation
modulus versus time. (b) Opening of the crack. (From de Gennes, P.G., Langmuir, 12, 4497–4500,
1996.)
The viscoelastic properties are also responsible for the high energy losses during the
debonding process. After we present some examples correlating the rheological proper-
ties with adherence performance, we will focus on rheological properties and identify
the relevant rheological parameters in the bonding and debonding stages.
140
CRC_59378_C004.indd 6
120
Cohesive failure Stick-slip failure
Interfacial failure
100
80
F.Tref/T(N)
60
40
Glassy failure
20
0
0 1 2 3 4 5 6 7 8 9
(a) log (aT .V ) (mm/min−1)
Rigid aluminum
substrate
Force (au)
Force (au)
Flexible aluminum
substrate
(b) (c) Length (au) (d) Length (au)
FIGURE 4.2 (a) Peeling master curve (floating roller test, Tref = 20°C) for a model formulation based on a homopolymer (polybutadiene) and a tackifying
resin. Peeling force is plotted (F) as a function of the logarithm of the reduced peeling rate (aT · V). The data were obtained at different temperatures (From
Derail, C., Allal, A., Marin, G., and Tordjeman, P., J. Adhesion, 61, 123–157, 1997.) and shifted along the x-axis to build the master curve. In this example, the
volume fraction of resin is equal to 70%. (b) Scheme of the floating roller test. (c) Force versus peeling length at constant rate in cohesive or interfacial domain.
Fundamentals of Pressure Sensitivity
(d) Force versus peeling length at constant rate in the stick–slip domain.
8/14/2008 1:48:04 PM
Role of Viscoelastic Behavior of Pressure-Sensitive Adhesives 4-7
6
log(G′, G ″) (Pa)
2
−3 −2 −1 0 1 2 3 4 5
log(a T. ω) (rad/s)
FIGURE 4.3 Rheological master curve [G′ () and G″ ()] as a function of circular frequency,
Tref = 20°C) for a model formulation based on a homopolymer (polybutadiene) and a tackifying
resin. (From Derail, C., Allal, A., Marin, G., and Tordjeman, P., J. Adhesion, 61, 123–157, 1997.) The
volume fraction of resin is 70%.
when the polymer is well defi ned. In that case, the maximum relaxation time, measured
in the terminal domain of mechanical relaxation, is the relevant reduction parameter
in the peel rate scale for peeling curves plotted in the cohesive failure zone [4] (see also
Chapter 11).
In the case of semicrystalline polymers for HM applications, the characteristic time
of transition between a viscoelastic liquid behavior (amorphous) and a viscoelastic solid
behavior (semicrystalline) by mechanical spectroscopy is of the same order of magni-
tude as the time necessary to observe the transition from cohesive to interfacial failure
measured by a peel test at very low peel rate and room temperature [19].
A typical domain of application for which the rheological behavior of the adhesive
is particularly relevant is the domain of medical applications (see also Applications
of Pressure-Sensitive Products, Chapter 4). The deformation of the substrate (skin) is
of the same order as the adhesive fi lm deformation during the debonding stage
[23,24].
Finally, some authors have presented models that estimate the peeling force values
from the rheological properties of the adhesive. The model proposed by Gent and
Petrich [15] (see Equation 4.2, where F is the peel force, W is the thickness of the
adhesive, σ is the tensile stress, and ε is the extensional deformation.
max
F W ⋅ ∫ ⋅ d (4.2)
0
leads to a fracture criterion that is so large that this model cannot reasonably describe
the physical fracture phenomenon.
On the basis of the previous examples, one can conclude that rheological properties
are relevant during the debonding stage, particularly for soft adhesives such as PSAs
or HMPSAs. If the linear viscoelastic properties are directly related to the adherence
properties in the case of soft adhesives, as presented above, the physics of the debond-
ing process itself is not clearly understood. Other parameters must also be taken into
account in relation to the adhesion processes at the interface/interphase itself. The previ-
ous cases dealt with strong adhesion (high-energy surfaces in general). When adhesion
is poor, the relationships are more complex and the rheological properties of the bulk of
the adhesive are not the relevant properties because of the competition between adhe-
sion at the interface and rheological properties of the bulk, as well as the interphase [6]
(“modulation of adhesion”).
We may partly conclude that the linear viscoelastic parameters are relevant to
formulate PSAs. Derail and colleagues [12,13] demonstrated that it is possible to
design molecular structures to fit properties dedicated to a given application: adhe-
sives based on triblock/diblock blends have been improved by designing tetrablock-
based or star/diblock blend-based “calculated” formulations. Table 4.1 presents some
adherence performances of various adhesives derived from this concept. When a
similar rheological behavior is mimicked using molecular design, adherence perfor-
mances are also similar. Some of these features were discussed in detail by Yarusso
in Ref. 25.
TABLE 4.1 Comparison of Adherence Performances for PSA Blends Based on Different Types
of Block Copolymers with Different Architectures
Note: SIS + SI is a blend of a triblock copolymer and a diblock copolymer, SISI is a tetrablock copolymer,
and SI4 + SI is a blend based on a radial copolymer and a diblock copolymer. All copolymers have been
synthesized with the aim of obtaining a similar rheological behavior. The structural parameters of
the copolymers were calculated from the model detailed in Derail, C., Cazenave, M.N., Gibert, F.X.,
Kappes, N., Lechat, J., and Marin, G., J. Adhesion, 80, 1131–1151, 2004.
Recently, Teisseire et al. [33,34] performed tack experiments on highly viscous silicone
oils and describe the observed phenomena on the basis of Maxwell rheological behavior.
They observed with these model fluids that, at high velocities, cracks appear before
bubbles because of cavitation. An important conclusion of this work is that interfacial
cracks can be observed even with liquid materials in a new regime where bubbles and
cracks are observed simultaneously.
According to these examples, it seems that all typologies of detachment have not been
explored yet. An important conclusion is that the use of model materials helps to defi ne
more specifically the various fracture modes as a function of the viscoelastic properties
according to the time of deformation (i.e., the rate of detachment in the case of peeling
or tack measurements).
We will discuss further under modeling some of the examples presented above.
1.6
1.4
1.2 Percent SI increases
Stress (MPa)
1.0
0.8
0.6
0.4
0.2
0.0
0 5 10 15
(a) Strain
0.8
Percent SI increases
Normal stress (MPa)
0.6
0.4
0.2
0.0
0 5 10 15 20
(b) Deformation
FIGURE 4.4 (a) Stress–strain curves for different diblock content (0–54%) in SIS + SI blends and
(b) tensile stress–strain curves on the same samples. (From Creton, C., Roos, A., and Chiche, A., in
Adhesion: Current Research and Applications, W.G. Possart, Ed., Wiley-VCH, Weinheim, 2005.)
Creton et al. propose an analysis based on the comparison of the nominal stress obtained
from tensile tests or probe–tack tests and conclude that the stress level obtained during
the probe–tack test can be derived from uniaxial tensile data obtained at high strain rates.
When parallel fibrils are formed during elongation of the adhesive, each fibril exhibits the
same type of mechanical behavior observed in tensile testing. In this way, an estimate of
the stress level is possible from tensile testing at high strains. A basic question remains,
however: how does each fibril detach from the surface?
The last point in Ref. 7 concerns the ability to form a fibrillar structure. Probe–tack tests
have been performed on different surfaces (steel and ethylene–propylene copolymer, see
Figure 4.5) with similar adhesives. Whatever the surface, stress increases rapidly. Cavities
are nucleated in the two cases but, for the low-adhesion surface, cavities propagate as inter-
facial cracks coalesce and the stress drops rapidly to zero. For the high-adhesion surface,
cavity walls are extended and foam is created with fibrils (Figure 4.6). The key parameter
in this case, where there is a balance between interfacial propagation and bulk dissipation,
is the ratio Gc/E, where Gc is the critical energy release rate and E is the elastic modulus.
When Gc is lower (e.g., ethylene–propylene copolymer), interfacial crack propagation is eas-
ier. Marin and Derail [6] qualitatively observed the same behavior by peeling on different
substrates. The ratio Gc/E seems to be a key parameter in controlling detachment.
0.6
Detachment on steel
0.5
0.4
Stress (MPa)
0.3
0.2
Detachment on copolymer EP
0.1
0.0
0 2 4 6 8 10 12
Strain
FIGURE 4.5 Probe–tack data for a SIS + SI (19% SI) on steel and an EP copolymer. (From
Creton, C., Roos, A., and Chiche, A., in Adhesion: Current Research and Applications, W.G.
Possart, Ed., Wiley-VCH, Weinheim, 2005.)
FIGURE 4.6 Fibrils during a probe–tack test can be deformed with a very high extension.
adhesive layer. For example, the time of contact or the force applied during the bonding
stage can modify adherence properties [3,33].
The nature of the surface can be changed by making physical or chemical modifica-
tions. Lakrout et al. [3] demonstrated the strong coupling between surface properties
(surface energy, roughness, etc.) and the bulk properties of the adhesive, the coupling
governing the nature and location of the fracture (fibrillation, cavitation, etc.).
In the particular case of HMPSAs, the deposition of the fi lm is performed at high
temperature with a liquid-like adhesive. After this stage, the temperature of the fi lm
of adhesive decreases and the variations in viscoelastic properties of the bulk adhesive
during the process may be studied in laboratory experiments using thermomechanical
analysis (see Figure 4.7). As thermomechanical analysis allows one to follow the evolu-
tion of the complex shear modulus as a function of temperature at a given heating or
cooling rate and a given frequency. It is also possible to establish a relation between the
temperature scale of thermomechanical curves and the time scale in a given process by
knowing the evolution of temperature as a function of time within the adhesive during
that process. According to the nature of the polymer, this variation may differ. The inset
in Figure 4.7 presents the evolution of the temperature of an adhesive along the same
process for two typical cases: a semicrystalline polymer and an amorphous polymer.
Crystallization of the material modifies largely the time–temperature dependence by
increasing the open time. For more details, please refer to Ref. 38.
As illustrated in Figure 4.7, the adhesive becomes harder and increasingly elastic as
the temperature decreases. Because adhesives are viscoelastic materials, one can observe
various viscoelastic relaxation domains upon cooling. In the first stage (application of the
adhesive), the relevant parameter (step 1 in Figure 4.7) is the viscosity at the application
2
G′
Temperature
Temperature
Semi crystalline
3 2
1
Time
Setting time Amorphous
Open time
FIGURE 4.7 Deposition of the fi lm of adhesive for the HM case. Three relevant parameters
exist for the bonding stage: (1) a viscosity level, (2) a modulus level (Dahlquist criteria), and (3) a
compliance level. The inset graphs temperature versus time within the adhesive according to the
type of adhesive.
1. The first polymer (dots) exhibits a low secondary plateau in the terminal region
(lowest frequencies). The level of the corresponding modulus, which lies below
the Dahlquist criterion, is obtained by adjusting the morphology of block copoly-
mers. This plateau defines a solid-like behavior that “eliminates” the flow domain
observed at large times (low frequencies) for viscoelastic liquids. The glass transi-
tion temperature may be adjusted according to the temperature of use and type
of application, using tackifying resins with relevant glass transition temperatures.
This is the thermodynamic effect of adding resins, as opposed to the topologic
effect on elasticity described above. With regard to processing, a low viscosity is
obtained above the order–disorder temperature for copolymers.
G ′ and G ″ (au)
Frequency (au)
FIGURE 4.8 “Reference” rheological behavior [G′ () and G″ () at Tref = 20°C] for a PSA for
a given family of applications. One distinguishes between the behavior of block copolymer for-
mulations (secondary plateau for G′) and the behavior of an acrylic-based adhesive with a gel-like
behavior in the terminal domain of relaxation (full lines).
2. The second polymer exhibits the typical signature of a critical gel in a wide fre-
quency range. In the terminal relaxation domain, the elastic modulus varies slowly
and its value is slightly higher than that of its viscous counterpart. The signature is
typically obtained with acrylic emulsions for PSA applications.
Flat punch
Displacement
Force
0
Stress
(b) Strain
FIGURE 4.9 (a) Probe–tack test and experimental parameters. (b) Different stages of the defor-
mation on a typical stress–strain curve with pictures of the various detachment states of the
adhesive.
Yamaguchi and Doi [44] proposed a block model which accounts for the main factors
appearing during the debonding stage: viscoelasticity of the PSA, cavitation, and slip-
page at the surface. Although the authors make some simplifications, particularly for
the cavity expansion (a spherical expansion is chosen in the model), the results of this
model allow the reproduction of the stress–strain curve obtained with a probe–tack test.
The model does not consider interfacial fracture, however, and the viscoelastic behavior
of the adhesive seems oversimplified.
Other models have been proposed in the literature. The two models described above
demonstrate that the phenomena appearing during the debonding stage are quite
complex. The viscoelastic behavior, in both the linear and the nonlinear regimes, is
highly relevant. Therefore, we will focus on the measurement of viscoelastic properties
of adhesives in the last part of this chapter.
The complex shear modulus G*(ω) is measured using a rotational rheometer. The flow
experienced by the material must be as close as possible to simple shear fl ow to generate
well-defined experimental data in terms of material functions. Different flow geometries
can be used, as summarized in Figure 4.10.
In the case of soft adhesives, the most typical rotational geometries used are cone
and plate, parallel plates, and rectangular torsion for hard materials (glassy domain).
FIGURE 4.10 Measurement of viscoelastic properties with basic rotational geometries. (a) parallel
plates, (b) cone and plate, (c) rectangular torsion, and (d) concentric cylinders (couette flow).
G ′ and G ′′ (au)
FIGURE 4.11 Typical rheological behavior of a PSA. (a) Master curve at Tref = Tuse of a PSA
(block copolymer + tackifying resins). (b) Master curve at Tref = Tprocess of the same adhesive.
The choice of the geometry depends on the behavior of the adhesive (from liquid-like to
solid-like). Cone and plate or parallel plate geometries were used for the experimental
data presented in this chapter. Figure 4.11 presents the typical rheological behavior of
a PSA. A master curve is obtained at a reference temperature using time–temperature
equivalence. Figure 4.11a exhibits a solid-like behavior with a secondary elastic plateau
due to either a nanostructure formation for block copolymers or crystallization in the
case of semicrystalline polymers. As discussed previously, the level of the storage modu-
lus is the key parameter for “instantaneous” adhesive properties. The secondary plateau
at low frequencies insures a no-flow condition.
Figure 4.11b exhibits a liquid-like behavior: this adhesive could flow at long times.
The time for application of the adhesive would have to be smaller than the time to flow.
The reptation concept introduced by de Gennes [11] and its subsequent improvements
set up a general framework describing the linear viscoelastic behavior of polymers and,
hence, the evolution of the complex shear modulus as a function of frequency for homo-
polymers. Benallal et al. [45], proposed an analytical approach to describe the relaxation
function G(t) of linear homopolymers melts as a function of polymer structure. In this
approach, G(t) can be considered the sum of different independent relaxation processes.
For a more detailed model describing each relaxation domain, readers could refer to
Refs 5, 45, and 46. This model can be modified and extended to the case of diblock and
triblock copolymers, particularly by taking into account the effect of the nanostructure
in the terminal zone of relaxation at the lowest frequencies as well as the specific relax-
ation of the elastomeric sequence of the diblock copolymers. In the case of diblock copo-
lymers, the elastomer sequence is considered to relax like the branch of a star polymer
[5]. As a consequence, by inverting these molecular models of viscoelasticity [10,12,46],
it is possible to “design” new molecules to obtain tailored viscoelastic properties and
improved adherence and process properties, as discussed previously (see Table 4.1). Fig-
ure 4.12 presents the rheological data obtained with three different types of block copo-
lymer formulations designed (“calculated”) to achieve a similar rheological behavior:
+ = +
+ =
G ′ for SIS + SI
G ′′ for SIS + SI
G′ and G ′′ (au)
G′ for SI4 + SI
G ′′ for SI4 + SI
G ′ for SISI
G ′′ for SISI
Frequency (au)
FIGURE 4.12 Molecular design. Structural parameters of new molecules or molecular topolo-
gies may be calculated to obtain an expected rheological behavior. Master curves have been
vertically shifted. (From Derail, C., Cazenave, M.N., Gibert, F.X., Kappes, N., Lechat, J., and
Marin, G., J. Adhesion, 80, 1131–1151, 2004.)
= /Sr (4.5)
materials. The viscosity increases or decreases with the strain rate, so the strain rate of
the viscosity measurement must be relevant to the process.
4.4.1.3 Compliance
The compliance of adhesives is certainly an important feature during the bonding stage
in which the adhesive flows under low pressure (see also Applications of Pressure-Sensitive
Products, Chapter 4). The compliance function, J(t), obtained from creep experiments,
can be defined as follows:
J(t) =
(t)/0 = JG + t/0 + JR(t) (4.6)
where JG is the instantaneous compliance that describes a purely elastic behavior; t/η 0 is a
transient term that describes the partial flow of the material, and JR(t) is associated with
the retardational creep behavior typical of a viscoelastic material.
According to Equation 4.6, under low pressure (i.e., stress σ0), the PSA may be more
or less deformed; depending on (i) the value of the newtonian viscosity (η 0) and (ii) the
retardational compliance transient function, JR(t), the behavior upon bonding can be
quite different. Figure 4.13 presents an example of the variation of the compliance func-
tion with the different terms described in Ref. 47. For more details on the viscoelastic
behavior of polymers, one can refer to Ref. 48, and Ref. 49 contains more details regard-
ing calculations for the various flow geometries.
1.E−2
J(t ) calculated
Retardation term
Degenerated term
1.E−3 J(t ) experimental
Compliance (Pa−1)
1.E−4
1.E−5
1.E−6
1.E−2 1.E −1 1.E 0 1.E +1 1.E+2
Time (s)
FIGURE 4.13 Transient creep function: the compliance is the sum of different terms as a
function of time: t/η 0, viscous term; JR(t), retardational term.
1.E+7
0.0506 s−1
0.1088 s−1
0.2782 s−1
1.E+6
0.9483 s−1
Linear viscoelasticity
ηel (Pa s)
1.E+5
1.E+4
1.E+3
1.E−2 1.E −1 1.E 0 1.E +1 1.E +2
Time (s)
FIGURE 4.14 Transient elongation viscosity at start-up of flow for a linear polymer (polypro-
pylene, Mw = 370 kg/mol) at different strain rates. The solid line indicates the Trouton transient
viscosity calculated from linear viscoelasticity.
1.E+8
0.01 s−1
0.0277 s−1
0.0941 s−1
0.2966 s−1
1.E+6
ηel (Pa s)
Linear viscoelasticity
1.E+4
FIGURE 4.15 Elongation viscosity versus time for long-chain branched (LCB) polyolefi ns at
different strain rates. Strain hardening is a specific signature of LCB.
Figures 4.14 and 4.15 demonstrate that branched polymers (combs, trees, pom-pom
structures, and random long-chain branched polymers) exhibit a strain hardening that
can be controlled by the molecular weight and topology of the materials. In the case of
adhesives, the behavior at high strain can be improved by the addition of triblock or star
polymers, which govern strain-hardening effects. The description of this type of model
is beyond the scope of this chapter, but for more information, one can fi nd examples of
applications in Refs 55 and 56.
Experiments on adhesives in elongational flow will certainly yield important infor-
mation on the high viscoelastic losses one can obtain (and control) with these materials.
More specifically, we believe that the control of the topology of polymers designed for
adhesive applications may lead to breakthroughs in the control of adherence of PSAs
and hot-melts, as polymer chain topology may lead to exotic shear and elongational
properties in the bulk and in the interface/interphase. This feature, already used for
improving the process and physical properties of thermoplastics, could lead to interest-
ing applications in the case of adhesives.
4.5 Conclusions
This chapter focused on the effect of the viscoelastic properties of PSAs on their adher-
ence properties, as measured using peeling or probe–tack experiments. The probe–tack
test brings new insight to the physics of adherence with visualization of the detach-
ment stage. One can identify, in particular, different stages such as cavitation, fibril-
lar extension, and interfacial cracking. The debonding stage is indeed quite complex
and there is not yet a comprehensive understanding or model for all observations.
Crude rheological models are often used to set up quantitative predictions, leading to
unrealistic calculations.
As indicated in this chapter, the viscoelastic properties govern, to a large extent,
adherence behavior. The rheological behavior in the linear viscoelastic domain provides
the main information and allows understanding and improvement of the main features
in the formulation of adhesives. This may be directly applied in the case of practical/
industrial applications (formulation of adhesives, design of molecules, improvement of
processes, etc.). Some authors recently indicated that nonlinear viscoelastic properties
may be particularly relevant during the detachment stage, in which fibrils stretch out.
Two points should be highlighted for better knowledge regarding the adherence of
highly dissipative adhesives.
1. Correlation for the understanding of strain hardening in branched or structured
polymers along elongation and tack experiments. Star polymers can improve per-
formance. Could exotic topologies lead to a better control of adhesion/adherence?
2. Chemical or topological modification of surfaces allows the modulation of adhe-
sion at the surface/interface/interphase. A large number of studies now deal with
this aspect of adhesion. Some original ideas could come from examples derived
from nature (gecko, mussel, fly, etc).
Acknowledgments
The authors thank Costantino Creton from ESPCI for the data and figures on tack
experiments. Frédéric Léonardi from IPREM-EPCP is also acknowledged for help-
ful discussions on the structure–property relationships and for the data and figures
obtained using the Meissner rheometer.
References
1. Benedek, I., In Pressure Sensitive Design and Formulation, Application, I. Benedek,
Ed. VSP, Leiden, 2006, pp. 1–23.
2. Creton, C., Fabre, P., In The Mechanics of Adhesion, D.A. Dillard and A.V. Pocius,
Eds. Elsevier, Amsterdam, 2002, pp. 535–575.
3. Lakrout, H., Sergot, P., Creton, C., J. Adhesion, 69, 307–359, 1999.
4. Derail, C., Allal, A., Marin, G., Tordjeman, P., J. Adhesion, 61, 123–157, 1997.
5. Gibert, F.X., Marin, G., Derail, C., Allal, A., Lechat, J., J. Adhesion, 79, 825–852,
2003.
6. Marin, G., Derail, C., J. Adhesion, 82, 469–485, 2006.
7. Creton, C., Roos, A., Chiche, A., In Adhesion: Current Research and Applications,
W.G. Possart, Ed. Wiley-VCH, Weinheim, 2005, pp. 337–363.
8. Lamblet, M., Ph.D. Thesis, Université de Paris VI, 2005.
9. Wu, S., In Polymer Interface and Adhesion, Marcel Dekker, Inc., New York,
1982.
10. Gent, A.N., Schultz, J., J. Adhesion, 3, 281–294, 1972.
46. Derail, C., Marin, G., In Adhesion: Current Research and Applications, W.G.
Possart, Ed. Wiley-VCH, Weinheim, 2005, pp. 229–248.
47. Léonardi, F., Ph.D. Thesis, Université de Pau et des Pays de l’Adour, 1999.
48. Ferry, J.D., Viscoelastic Properties of Polymer, 3rd edition, John Wiley & Sons,
New York, 1980.
49. Marin, G., Oscillatory Rheometry, in Rheological Measurement, Chapman & Hall,
London, 1998, Chap 1.
50. Weissenberg, K., Nature, 159, 310–311, 1947.
51. Meissner, J., Rheol. Acta, 10, 230–242, 1971.
52. Münstedt, H., J. Rheol., 23(4), 421–436, 1979.
53. Schulze, J.S., Lodge, T.P., Macosko, C.W., Rheol. Acta, 40, 457–466, 2001.
54. Schweizer, T., Rheol. Acta, 39, 428–443, 2000.
55. Sarrazin, J., Ph.D. Thesis, Université de Pau, 2004.
56. Bourrigaud, S., Ph.D. Thesis, Université de Pau, 2004.
5.1 Introduction
Pressure-sensitive adhesives (PSAs) are polymeric materials that can form a physi-
cal bond with another material upon brief contact and with light pressure. Examples
of applications of PSAs include office, price marking, and electronic data processing
labels, as well as office, packaging, diaper, auto/masking tapes, bandages, decals, etc.
The general technical requirements for these types of materials are tack, peel resis-
tance, and shear resistance. A wide range of PSA products have been designed based
on the balance of these properties (see Applications of Pressure-Sensitive Products,
Chapter 4).
From a technical viewpoint, all PSA applications involve bond formation and debond-
ing steps. Bond formation is the result of a polymeric material being able to flow and
wet under light pressure and thereby is capable of establishing a contact area with a
substrate. The debonding step involves deformation of the polymeric material under
stress (typically extension), followed by separation from the substrate. Both the bond-
ing and the debonding processes are related to the rheological properties of the PSA
5-1
8
XLQ
Peel strength (20 min) (ppli)
r 2 > 0.9897
2
0
−0.6 −0.4 −0.2 0.0 0.2 0.4
Dissipation (log tan δ at 127°C)
FIGURE 5.1 Correlation of 20 min peel strength with log tan δ at 127°C. (From Dale, W.C.,
Paster, D.M., and Haynes, J.K., Mechanical Properties of Acrylic PSAs and Their Relationship to
Industry Standard Testings, Taylor & Francis, London, 1989. With permission.)
3
XLQ
r 2 >0.87
−1
−2
3.2 3.4 3.6 3.8 4.0 4.2 4.4
Log storage modulus G ′ (Pa) at 127°C
FIGURE 5.2 Correlation of shear hang time with log-log storage modulus at 127°C. (From Dale,
W.C., Paster, D.M., and Haynes, J.K., Mechanical Properties of Acrylic PSAs and Their Relationship
to Industry Standard Testings, Taylor & Francis, London, 1989. With permission.)
T = P0 BD (5.1)
where T is the adhesion performance and P0 is an intrinsic interfacial failure energy
(either the energy required to open up a unit area of PSA–substrate interface in the
absence of viscoelastic energy loss or the thermodynamic work of adhesion, which
is substrate dependent). B is the bonding function, assumed to be constant when
the Dahlquist contact criterion is satisfied (i.e., the plateau modulus is lower than
3.3 × 105 Pa); D is the debonding function, which is the viscoelastic loss component. It
is strongly dependent upon the characteristic debonding frequency (i.e., the separation
speed of the PSA test).
Although viscoelastic measurements involve low strains, whereas PSA adhesion tests
involve high strains, Tse’s study confirms many earlier findings of good correlations of
viscoelastic properties with PSA adhesion.
The noteworthy features and contributions of this work are as follows:
107
G′ (dyn/cm2) at 25°C
106
105
0.1 1.0 10 100
Bonding Frequency sweep data (ω, rad/s) Debonding
shear resistance tack peel
FIGURE 5.3 Room temperature modulus requirement at various frequencies. Shaded area indi-
cates the window with good PSA properties. (From Chu, S.G., Adhesive Bonding, Plenum Publish-
ing, New York, 1991. With permission.)
145
107 25°C
Tape
Label PSA
14.5
106
Modulus (psi)
G′ (dyn/cm2)
105 1.45
104 104
0.1 1.0 10 100
ω (rad/s)
FIGURE 5.4 Room temperature modulus values of label and PSA tape at various frequencies.
(From Chu, S.G., Adhesive Bonding, Plenum Publishing, New York, 1991. With permission.)
1010
ω = 10 rad/s
109
108
G′ (dyn/cm2)
107
106
105
104
Melt
processing
103
−10 25°C 80 120 150
Temperature (°C)
FIGURE 5.5 Modulus (G′) requirements (windows) for good PSA tapes. (From Chu, S.G., Adhe-
sive Bonding, Plenum Publishing, New York, 1991. With permission.)
VWs were constructed by plotting the four coordinates: (1) G′ at 10−2 rad/s, G″ at 10−2
rad/s; (2) G′ at 102 rad/s, G″ at 10−2 rad/s; (3) G′ at 10−2 rad/s, G″ at 102 rad/s; and (4) G″
at 102 rad/s, G″ at 10 2 rad/s on the log-log cross plot of G′ and G″. Chang [16,17] reported
that for most PSAs, the range of G′ and G″ at room temperature within the selected fre-
quencies fell between 103 and 106 Pa. In addition, there was a unique correlation between
the adhesion performance of the PSAs and the location of their VWs. A four-quadrant
concept was therefore adopted to categorize different types of PSAs. The location of
different VWs is illustrated in Figure 5.7 [42], along with their corresponding operative
viscoelastic regions. To illustrate the consistency of the VW concept for the different
types of PSAs, several key materials in each group are listed below to ascertain their
viscoelastic uniqueness.
Quadrant 1 (top left-hand quadrant): high G′ and low G″. This quadrant corresponds
to high modulus, low dissipation. The bonding and debonding frequencies, in this case,
both occur at the plateau region of the rheological master curve. No PSA can be found
in this quadrant because of the high bonding modulus (i.e., G′ at 10−2 rad/s) and highly
elastic nature (lack of flow) of the material, making the bonding step unfavorable. Some
elastomers and release coatings occupy this quadrant.
Figure 5.8 illustrates the VWs of a polydimethyl siloxane (PDMS), PDMS + 40%
control release agent (CRA, an additive to increase the release force, which is a tri- or
tetrafunctional methylsilicate resin; see also Technology of Pressure-Sensitive Adhesives
107 145
Vistanex
Reichhold (TYLAX)
SBR Polysar
Latex Union chemical (AMSCO)
Dow
208
NR SBR1011
L120
Hartex 8226
3892 30678 6130
G ′, 25°C (dyn/cm2)
222
Modulus (psi)
8244
Good
106 444 14.5
PSA
8277
Good
LM-MS label
105 1.45
−60 −50 −40 −30 −20 −10 0 10
Tg (°C), tan δ maximum temperature
FIGURE 5.6 Viscoelastic properties of Piccotac HM2162L/Kraton 1107/oil blends and empirical
windows required for various labels and PSA tapes. (From Chu, S.G., Adhesive Bonding, Plenum
Publishing, New York, 1991. With permission.)
and Products, Chapters 6 and 10), and PDMS + 60% CRA. Because of the high bond-
ing modulus (G′ at 10−2 rad/s) and low dissipation (G″), PDMS is not a PSA, but rather
a release coating. However, with the incorporation of an increasing amount of CRA, the
modulus is progressively reduced, accompanied by an increase in dissipation or flow.
Such a progressive decrease in G′ and increase in G″ results in a progressive increase in
the tackiness of the samples. Thus, PDMS, when modified with 60% CRA, is tacky. It can
be anticipated that with further increase in CRA concentration, the VW of the resulting
sample will move toward the central region, which, as described in a later section, is the
location for general purpose PSAs.
Quadrant 2 (top right-hand quadrant): high G′ and high G″. This quadrant corre-
sponds to high modulus and high dissipation. The bonding frequency corresponds to
106
Quadrant 1 Quadrant 2
Rubbery region Transition−plateu region
High modulus High modulus
Low dissipation High dissipation
105
Release−Non PSA Transition−flow region High shear PSA
G ′ (Pa)
Medium modulus
Medium dissipation
Plateau−flow region General purpose PSA
Low modulus Flow−flow
104
Low dissipation Low modulus
High dissipation
FIGURE 5.7 VWs of PSAs as related to different regions on the rheologic master curves. (From
Chang, E.P., Viscoelastic Windows of PSAs, Taylor & Francis, London, 1991. With permission.)
106
105
G ′ (Pa)
104
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.8 VWs of release coatings. (From Chang, E.P., Viscoelastic Windows of PSAs, Taylor
& Francis, London, 1991. With permission.)
106
105
G ′ (Pa)
104
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.9 VWs of high-shear PSAs. (From Chang, E.P., Viscoelastic Windows of PSAs, Taylor
& Francis, London, 1991. With permission.)
the plateau region, whereas the debonding frequency corresponds to the glass transi-
tion region in the rheological master curves for high-shear PSAs. The high bonding
modulus, compensated by the high dissipation or flow, makes the bonding marginally
efficient. Shear is high because of the high G′ or cohesive strength of the material.
Figure 5.9 illustrates the VWs of three high-shear PSAs: HSPSA 1, 2, and 3. All
of the VWs occupy the top right-hand corner, which means that these adhesives
have comparatively high modulus (G′) and high dissipation (G″) within the applica-
tion rates (i.e., between 10 −2 and 102 s−1). In general, these adhesives have compara-
tively high Tg’s and are comparatively highly cross-linked to achieve the high shear
performance.
Quadrant 3 (bottom left-hand quadrant): low G′ and low G″. This quadrant corre-
sponds to low modulus, low dissipation. The bonding frequency corresponds to the
onset of the flow transition, whereas the debonding frequency corresponds to the plateau
region in the rheological master curves for removable PSAs. Bonding is facilitated by the
low modulus despite the low flow characteristics. Peel values are usually low because of
the comparatively low debonding strength and low dissipation. Removable and medi-
cal-type PSAs fall within this quadrant.
Figure 5.10 illustrates the corresponding VWs for removable adhesives RPSA 1, 2,
and 3. The distinct characteristics of this type of adhesives are as follows:
• Low bonding modulus so that the adhesive is very contact-efficient
• Low dissipation, which implies more elasticity or better removability
106
105
G ′ (Pa)
104
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.10 VWs of removable PSAs. (From Chang, E.P., Viscoelastic Windows of PSAs,
Taylor & Francis, London, 1991. With permission.)
Figure 5.11 illustrates the VWs of some of the removable PSAs used in medical appli-
cations (e.g., bandages). Comparing the VWs of these medical removables with those
removables in Figure 5.10, one notes that they tend to occupy the lower (better conform-
ability) and farther right (better flow) area of Quadrant 3. Some of the notable differ-
ences between the removable and the bandage adhesives are as follows:
• The reference temperature for the medical adhesive is the body temperature,
37°C, rather than 23°C in the removable case. This makes the bonding modulus
of the medical adhesives even lower (i.e., more conformable) than that of the
removables because of the higher reference temperature. This is desirable for
contact area considerations because of the rough, frequently varied, and con-
taminated nature of the skin.
• The debonding moduli (top right-hand corner of the window) are usually higher
than those of the removables. This, again, is necessary to prevent lift or detach-
ment because of frequent flexing of the skin, especially on curved areas, such as
knees and elbows.
Quadrant 4 (bottom right-hand quadrant): low G′ and high G″. This quadrant corre-
sponds to low modulus, high dissipation. The bonding frequency corresponds to the
flow region, whereas the debonding frequency corresponds to the onset of the flow
region in the rheological master curves for very quick or cold-stick PSAs. The low bond-
ing modulus coupled with high flow makes bonding very efficient, thus permitting the
material to stick even at low temperatures or very short contact time.
106
105
G ′ (Pa)
104
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.11 VWs of medical PSAs. (From Chang, E.P., Viscoelastic Windows of PSAs, Taylor
& Francis, London, 1991. With permission.)
Three adhesives that have a significant portion of their VWs in the fourth quadrant
should have low bonding modulus, G′, as well as good flow or highly dissipative nature,
G″. So far, no good example of a PSA has been found with its VW located right in the
fourth quadrant.
Central area: medium G′ and medium G″. This central area corresponds to medium
modulus, medium dissipation. The bonding frequency corresponds to the onset of the
flow region, whereas the debonding frequency corresponds to the onset of the glass tran-
sition region in the rheological master curves (i.e., usually characterized by the absence
of a distinct plateau region) of general purpose PSAs.
Figure 5.12 illustrates the corresponding VWs for three general purpose acrylic PSAs,
GPPSA l, 2, and 3. They all occupy the central region (overlapping part of the four quad-
rants), illustrating the general purpose nature of this type of PSAs.
106
105
G ′ (Pa)
104
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.12 VWs of general purpose PSAs. (From Chang, E.P., Viscoelastic Windows of PSAs,
Taylor & Francis, London, 1991. With permission.)
106
G ′ (100) G ′ (100)
105 G ′′ (.01) Removability G ′′ (100)
G′ (Pa)
(elastic)
Flowability
(viscous)
104
G ′ (.01) G ′ (.01)
G ′′ (.01) G ′′ (100)
103
103 104 105 106
G ′′ (Pa)
FIGURE 5.13 Relationship of the VW with the Dahlquist contact criteria and tan δ. (From
Chang, E.P., Viscoelastic Windows of PSAs, Taylor & Francis, London, 1991. With permission.)
107
106
Storage modulus (G ′/Pa)
105
8400
= 21
MW
104
000
203 00 0
46 80
103 16 3
12
102
10−5 10−4 10−3 10−2 10−1 100 101 102 103 104 105
(a) Frequency (ω/rad/s)
6
Loop tack (N/cm2)
0
0 2 4 6 8
(b) 1/G ′ × 105/Pa
FIGURE 5.14 (a) Storage modulus, G′, as a function of frequency. (b) Loop tack as a frequency
of inverse of G′. Stainless-steel substrate (○) and polyethylene substrate (△) for a series of acrylic
copolymer samples having the same composition (75 wt % 2-ethylhexyl acrylate, 23 wt % ethyl
acrylate, 2 wt % acrylic acid) but different molecular weights. (From Yang, H.W.H. and Chang,
E.P., The Role of Viscoelastic Properties in the Design of PSAs. Elsevier/Springer, New York, 1997.
With permission.)
modulus (i.e., the base of the application window) much below the Dahlquist line (which
means good conformability). In other words, by comparing the position of the base of the
window with the Dahlquist line, we immediately know whether the material is a PSA.
loss modulus, G″ (i.e., tan δ < 1 and tan δ > 1, respectively). The portion of the window
to the left of the line (i.e., tan δ < 1) indicates the more elastic region. In other words,
the closer the window is to the top left-hand corner of the G′–G″ plot, the more elastic
(or better removeablity) the material characteristics. Conversely, the closer the window
is to the lower right-hand corner of the plot, the more viscous (or cohesive failure prone)
the material characteristics.
Assuming that the adhesive is the only variable in the construction and that the sur-
face effect is negligible, the following adhesion and convertibility performance can be
correlated with the shape and location of the VW.
1. It can be obtained by just making measurements at two frequencies (i.e., 10−2 and
102 rad/s at the specified temperature). Such simple and rapid measurements will
immediately identify the nature and type of the adhesive.
The limitation of Chang’s VW method is that it only gives the G′ and G″ values at those
two frequencies. There is no information for G′ and G″ at the in-between frequencies
(e.g., 10−1, 100, and 101 rad/s) or other frequencies as in a master curve, which would be
recommended for more quantitative information.
∆E = πG″L2 (5.4)
where ∆E is the dissipation energy per cycle of the oscillatory shear deformation, and L
is the applied strain amplitude of the deformation.
By combining the effect of bond formation and separation, a simplified correlation for
the PSA adhesive strength was proposed [24],
P ∼ I × G″(ω1)/G′(ω2) (5.5)
where G″ is measured at the peeling frequency, ω1, and G′ is measured at the bonding
frequency, ω2. Equation 5.5 provides a basis for correlating PSA adhesive strength, mea-
sured by 180° peel or loop tack, with the linear viscoelastic properties of the adhesives.
In a standard PSA peel test according to Ref. 30, the 180° peel is conducted at a rate
of 30.48 cm/min with an adhesive layer thickness of 0.037 mm. The corresponding
debonding frequency has been calculated to be 435 rad/s. However, the bonding is car-
ried out at a much lower frequency, typically at ∼1 rad/s.
To verify the validity of the above statements, a series of acrylic copolymers was syn-
thesized [47] to test the validity of Equation 5.5. The first series of samples have the
same composition and Tg but different molecular weights, as illustrated in Figure 5.14a.
These samples have the same G″ at the debonding frequency region but a different G′
at the bonding (1 rad/s) region. The result of the adhesive strength test, as illustrated in
Fig ure 5.14b, indicates that the loop tack is inversely proportional to G′. The slope of the
straight line correlates with the intrinsic surface energy, as indicated in Equation 5.5. In
Figure 5.14b, we see a steeper slope with a stainless-steel substrate than with a polyethyl-
ene substrate. This is evidently because of the higher intrinsic surface energy associated
with the more polar stainless-steel substrate.
Another set of acrylic copolymers with a wider range of compositions was also synthe-
sized. In this case, G″ and G′ at the corresponding debonding and bonding frequencies
were different. The test results indicate that 180° peel resistance is indeed proportional to
G″/G′, as illustrated in Figure 5.15, again supporting the proposal stated in Equation 5.5.
2.0
1.5
Peel (lb/in.)
1.0
0.5
0
0 0.4 0.8 1.2 1.6 2 2.4 2.8
G′′(ω1)/G′(ω2)
FIGURE 5.15 180° peel strengths (PSTC-1 procedure 37 for a wide range of acrylic copolymers,
G″ measured at debonding frequency of 435 rad/s, G′ measured at a bonding frequency of 1 rad/s). □,
ethyl acrylate (EA)–ethylhexyl acrylate (EHA); +, methyl acrylate–EA–EHA; ●, methyl meth-
acrylate–EHA; △, butyl acrylate; ○, EHA–EA–acrylic acid (various molecular weights). (From
Yang, H.W.H. and Chang, E.P., The Role of Viscoelastic Properties in the Design of PSAs. Elsevier/
Springer, New York, 1997. With permission.)
TABLE 5.1 Factors Affecting Dynamic Shear Storage Modulus, G′, and Dynamic Shear
Loss Modulus, G″
G′ G″
Increase polymer molecular weight ++ +
Increase polymer entanglement molecular weight −− −−
Increase polymer glass transition temperature, Tg + ++
Add hydrocarbon resin or rosin −− ++
Add plasticizer or oil −− −−
Cross-link polymer chains ++ +
Note: G′ measured at 1 rad/s. G″ measured as 435 rad/s. ++ and −− refer to a significant increase
or decrease. + and − refer to a minor increase or decrease.
Gn ∼= ρRT/Me(Vp)2 (5.6)
where Vp is the volume fraction of the polymer in the polymer–resin blend, Me is entan-
glement molecular weight of the polymer, ρ is the density of the blend, R is the gas
constant, and T is the temperature in Kelvin. Hydrocarbon-based resins have been very
effective in tackifying SIS and SBS block copolymers for hot-melt adhesives [14,15] (see
also Technology of Pressure-Sensitive Adhesives and Products, Chapters 3 and 8). Water-
based hydrocarbon resins are also available to tackify acrylic, NR, and SBR latices [7].
The effects of resin and polymer compatibility on the viscoelastic behavior of the finished
products and their PSA properties have also been investigated [7,13,21,48,49,51,52].
Class and Chu [7–9] observed a pronounced shift in the tan δ peak (toward a higher
temperature as a result of higher Tg), together with a decrease in the plateau modulus,
when the resin and the rubber were compatible. As discussed earlier, the depression in
the modulus at low frequencies facilitates the bonding process. The shift of the tan δ
peak to a higher temperature also results in higher G″ in the debonding region, as indi-
cated in Table 5.1.
Another important parameter affecting G′ is the Me of the polymer. To have good
bonding characteristics, we need polymers with low G′ at the bonding region. Low G′
normally associates with high Me. For acrylic polymers, Me increases with increasing
monomer side chain length. Unfortunately, the Tg for these polymers also decreases with
increasing side chain length. It is, therefore, necessary to employ some higher Tg como-
nomers, such as methyl methacrylate or methyl acrylate, to counterbalance the low Tg
monomers such as butyl acrylate or 2-ethylhexyl acrylate. The entanglement molecular
weight (Me) is closely related to the structure of polymer chain segments. A summary of
Me values for various polymers can be found in a review by Fetters et al. [50].
The use of DMA to formulate and evaluate complex PSA systems is demonstrated by
Chang and Holguin in Applications of Pressure-Sensitive Products, Chapter 3 (see also
Applications of Pressure-Sensitive Products, Chapter 8).
References
1. Dahlquist, C. A. (1966), Proc. Nottingham Conf. on Adhesion, Part I, Maclaren &
Sons Ltd., London, Chap. 5, p. 134.
2. Kaelble, D. H. (1969), J. Adhesion, 1, 102.
3. Gent, A. N. and Petrich, R. P. (1969), Proc. Roy. Soc. (London), A, 310, 433.
6.1 Introduction
Pressure-sensitive adhesive (PSA) properties are typically characterized by three types
of standard tests, peel tests, shear tests, and tack tests (see also Applications of Pressure-
Sensitive Products, Chapter 8). The latter type of test is designed to probe the ability of
the PSA to stick on a surface under a light applied pressure. The test is often carried out
with a loop of tape (loop tack test; see also Applications of Pressure-Sensitive Products,
Chapter 8) or by testing the rolling resistance of a steel ball on the adhesive1 (see also
Applications of Pressure-Sensitive Products, Chapter 8).
An alternative way to test the tackiness of a PSA is by applying a rigid steel punch
under controlled conditions on the surface of the adhesive and subsequently removing
it at constant velocity, as illustrated schematically in Figure 6.1. Although this type of
test has never been really adopted as a standard industry test for tack, it has emerged
in recent years as a very powerful and sensitive analytical tool to evaluate the adhesive
properties of PSAs2–4 (see also Applications of Pressure-Sensitive Products, Chapter 8).
This success is due, first of all, to the high sensitivity of the test to small changes in chem-
ical structure of the adhesive or interfacial interactions and then to the development
of a detailed understanding of the specific deformation mechanisms observed during
6-1
Debonding
Force
Approach
Time
Contact
FIGURE 6.1 Schematic of a probe tack test as normally performed with a texture analyzer or
analogous equipment. The compressive force is applied until a set force is reached and then the
displacement is kept constant, allowing the force to relax. The probe is then removed from the
adhesive fi lm at constant velocity.
the test.5–9 The deformation of a soft, confined polymer disk is very sensitive to the detailed
rheological properties of the material, not only in the linear regime of small strains but
also in the highly nonlinear regime of large strains. Furthermore, when the probe test is
carried out with a relatively large compressive force, it does not really test tackiness, but
rather adhesive properties as measured in a peel test with the added advantage of being
able to vary contact time, contact pressure, and debonding rate independently.
The purpose of this chapter is not to review all previous work in this area but, rather,
in the first section, to review the basis for the interpretation of probe test stress–strain
curves as they can be obtained experimentally with a probe tester and, in the second
section, to present some examples of interpretation.
2a 0
2a h0
FIGURE 6.2 Schematic of the contact between a flat cylindrical probe and a thin fi lm and
definitions of the layer.
Stress σ = F
(MPa) A0
Cavitation 0.8
σmax
1mm
0.4
εmax
W adh = h0 σ(ε)d ε
0.2
0
Detachment εmax
Homogeneous of fibrils
1mm deformation
0
0 2 4 6 8 10
h − h0
Strain ε =
h0
FIGURE 6.3 Example of a stress–strain curve obtained with a probe tester equipped with a
flat-end probe. The various stages of deformation of the layer are illustrated by images in low and
high magnification of the deformed layer. The appearance and growth of cavities under tensile
stress are apparent.
interface between the probe and the adhesive, lateral expansion of the cavities, and,
finally, growth of a fibrillar structure.6 The first two stages are nearly always observed,
whereas the last two stages depend on the interplay between adhesive properties and
deformability of the layer.
r2
p (r ) zz (r ) p0 2 N 1 2 (6.1)
a
where r is the radial distance from the axis of symmetry, p0 is the external pressure
(typically equal to 1 atm), λ is the extension ratio, and σ N is the nominal tensile stress.
≡ h / h0 N ≡ P/a02 (6.2)
Two underlying assumptions are involved in the derivation of Equation 6.1. The first is
that the layer is thin enough that the pressure is nearly uniform throughout the thickness
of the layer. This assumption is valid whenever a/h is large, regardless of the mechanical
properties of the layer itself. The second assumption underlying Equation 6.1 is that the
shear stress increases linearly with the radius, that is, σrz(r) ∝ r. This assumption is valid
provided that the material is incompressible and that a no-slip boundary condition is
observed at the confining surfaces of the deformed material. This no-slip condition is
nearly always met for liquids, but is not always met for materials with a more solid-like
character, such as PSAs.11
For large values of a 0/h0 and for elastic layers, the nominal stress σ N is proportional to
strain and for sufficiently small strains (λ close to 1) it can be written as
N 3
(a0 / h0 )2 (6.3)
2
3. The material at the edge of the contact is deformed well into the nonlinear
regime so that linear elastic theory is no longer appropriate to describe the elastic
behavior.15
6.2.2 Cavitation
Cavitation in highly deformable elastic materials is generally a heterogeneous process,
corresponding to the expansion of an existing cavity. For a sufficiently low elastic
modulus, cavity expansion is determined by the pressure needed to overcome the inter-
nal Laplace pressure. The cavity is assumed to have an initial radius of curvature R0
and to be fi lled with an ideal gas at a pressure of p0. The presence of gas in the initial
defect corresponds to the entrapment of atmospheric gas from the external environment
(where the pressure is equal to p0 initially), within regions with a characteristic size that
is defined by the roughness of the probe.9,20 Mechanical equilibrium requires that the
applied pressure (p) is equal to the sum of the Laplace pressure (−2γ/R0) and the internal
pressure (p 0V0/V),
p0 2
p (6.4)
r r R0
3
where the extension ratio, λ r, is equal to Rc/R0. We have made the simplifying assump-
tion that the shape of the bubbles does not change with volume, so that V ∝ λr3.
The addition of elasticity complicates the problem, because one must take into
account an appropriate energy balance that can be addressed analytically in some limit-
ing cases. For a spherical void of initial radius R0 in an infinite, incompressible material,
the strain state around the void is fully specified by λ r = R/R0. One must now account
for an additional “elastic” inflation pressure, pel, that must be applied to maintain this
elastic deformation of the material. For a spherical void in an isotropic medium, the
inflation pressure becomes
p0 2
p pel ( r ) (6.5)
r r R0
3
The specific form of pel(λ r) depends on the constitutive model that is used to describe the
strain energy density of the deformed material. The problem involves large strains, and
linear elasticity is no longer sufficient. Models are needed that relate the overall strain
energy to λ1, λ 2, and λ 3, the principal extension ratios characterizing the strain state of
the material. The simplest constitutive model describing the elasticity of rubber at fi nite
strains is the neo-Hookean material, for which the elastic strain energy density, Uel, can
be written in the following form:
(
U el ∝ 12 22 32 3 ) (6.6)
1 4 1
pel Ef el ( r ) f el ( r ) 5 4 (6.7)
6 r r
p0 2
p Ef el ( r ) (6.8)
r r R0
3
Elasticity enters the problem only through the elastic modulus in the third term on the
right-hand side of Equation 6.8. The magnitude of this term relative to the other two
terms therefore determines the extent to which elasticity affects the cavitation behav-
ior.4,9 This relative balance can be quantified by a dimensionless ratio, γ/ER0, which
defines the surface-controlled and elastic-controlled limits for the cavitation stress. The
distinct features of these different regimes are discussed in the following sections. The
quantity γ/ER0 is proportional to the ratio of the Laplace pressure to the elastic expan-
sion pressure for a sample with a characteristic defect size of R0. In situations where γ/E is
small in comparison to the initial defect size, R0, surface deformations can be ignored,
and the response of the system is determined by bulk elasticity. For small values of
γ/ER0, the cavitation stress is therefore dominated by the elastic expansion pressure. In
this case, Equation 6.7 for a neo-Hookean material predicts that a preexisting cavity in
an infinite elastic medium under a constant tensile stress will expand indefi nitely when
the magnitude of the negative hydrostatic pressure approaches 5E/6.
If the material becomes softer and the interface is prepared more carefully, the Laplace
pressure term in Equation 6.8 may exceed the elastic expansion term. The effect of the
surface energy can be illustrated in a qualitative sense by assuming that p0/E is small and
combining Equations 6.7 and 6.822
p 1 12 (
/ ER0 ) 4 1
5 4 (6.9)
E 6 r r
We define the pressure as negative because these negative values correspond to the state
of hydrostatic tension that is relevant to our experiments. For γ/ER0 → 0, the size of
the initial void is predicted to increase continuously with increasing hydrostatic ten-
sion. For values of γ/ER0 larger than a critical value of 1/3, the situation is more compli-
cated and the cavity should expand rapidly once the applied tensile pressure exceeds the
Laplace pressure corresponding to the original defect of size R0.23 The final size of the
cavities is typically comparable to the thickness of the elastic layer itself, because this
thickness controls the length scale from which elastic energy is available to drive the
cavity growth.24,25
Th is picture is qualitatively consistent with experimental results illustrated in Fig-
ure 6.4.9 Small cavities that were optically invisible at the beginning of the test expand
more rapidly and at higher stress levels than larger cavities that were optically visible at
the beginning of the test. Furthermore, for these soft rubbery PSAs, the pressure neces-
sary to observe cavities can be as large as 10 times Young’s modulus (E), while remain-
ing much larger than the atmospheric pressure, suggesting that the growth of defects is
140
120
Cavity radius (µm)
100
80
60
Cavity 1
40
Cavity 2
20
0
0 2 4 6 8 10
Time (s)
t=0s t=7s
2
1
1
500 µm
FIGURE 6.4 Time dependence of cavity growth and the contact images at early and late stages
of the cavity formation process. Cavity 1 initiates at the early stages of the deformation process
from a visible defect, at a relatively low value of normal stress. Cavity 2 initiates from a very
small defect at a large value of stress and expands very rapidly. (From Shull, K.R. and Creton, C.,
J. Polym. Sci. Part B: Polym. Phys. 42, 4023–4043, 2004. With permission.)
2Rd
θ
hc
(a) Rc
2Rp
(b) 2Rd
FIGURE 6.5 Schematic illustrations of various cavity geometries: (a) an interfacial cavity
and (b) ellipsoidal interfacial cavity with θ < 90, illustrating the distinction between debonded
radius, Rd, and the projected radius, Rp. (From Shull, K.R. and Creton, C., J. Polym. Sci. Part
B: Polym. Phys. 42, 4023–4043, 2004. With permission.)
introduces in a rather natural way the familiar balance between released elastic energy
(G) and critical energy release rate, Gc. If one applies Griffith’s criterion, the cavity will
grow/expand when G > Gc.
The overall picture can be summarized by considering the response of an initial
penny-shape interfacial crack (hc/Rd << 1) to an increasing hydrostatic tension, pel. For
pel/E << 1, we recover the crack-driving force from standard linear elasticity theory33
2
3Rd E pel
G (6.10)
2 E
As pel/E increases, several things happen. The crack begins to inflate in the verti-
cal direction, and the value of hc in Figure 6.5a increases. The energy release rate also
increases in accordance with Equation 6.10, which remains valid for values of pel/E less
than about 0.4. For pel/E ≈ 1, the defect becomes hemispherical (hc ≈ Rd) and the energy
release rate increases nonlinearly to a much higher value that is determined by the
large-strain response of the material.34 The specific value of pel/E corresponding to this
nonlinear increase depends on the details of the strain energy function that is used to
describe the material.34 The example in Figure 6.6 is for a neo-Hookean material. If this
large increase in G corresponds to an increase from a value that is below Gc to a value
that is above Gc, cavities will grow. The fact that this sudden increase in G can be rela-
tively large leads to a criterion for cavitation that is coupled to the elastic modulus of the
material and is relatively insensitive to the specific value of the interface toughness, Gc.30
In other words, one expects that, if Gc/ER0 > 1, an interfacial defect will grow in the bulk
at a value of stress roughly independent of the nature of the interface.
100
10
1
G /ER
0.1
0.01
0.001
0.1 1
−p /E
FIGURE 6.6 Energy release rate for an interfacial cavity of debonded radius R0 as a function
of p/E. (From Shull, K.R. and Creton, C., J. Polym. Sci. Part B: Polym. Phys. 42, 4023–4043, 2004.
With permission.)
The length scale defined by Gc/E plays an important role in determining the overall
behavior of the system, even when fibrillation occurs. The materials parameters that
determine Gc are complex and include the linear and nonlinear viscoelastic properties
of the adhesive layer35,36 and the frictional properties of the probe/layer interface.37–39
Our purpose here is not to elucidate all of the factors that determine the value of Gc for
a given system, but to describe in qualitative terms how the overall deformation of the
adhesive layer proceeds for different values of Gc/E.
Low Gc /E
FIGURE 6.7 Schematic tack curves for low values of Gc/E, where the stress remains below the
cavitation stress.
High Gc /E
FIGURE 6.8 Schematic tack curves illustrating the case where the cavitation stress is exceeded.
The arrows in each part of the figure denote increasing Gc/E.
CRC_59378_C006.indd 13
Upper actuator
Probe Tack
Spring
Steel ball
Video
Mirror
Microscope slide
Adhesive layer
( : Contact)
Probe
Mirror
Microscope slide
Adhesive layer
Extensometer
Lower actuator
FIGURE 6.9 Schematic of the experimental design and setup used for the probe tests. (From Creton, C., Roos, A., and Chiche, A., Adhesion: Current
Research and Applications, W.G. Possart, Ed., Wiley-VCH, New York, 2005. With permission.)
6-13
8/14/2008 10:49:09 PM
6-14 Fundamentals of Pressure Sensitivity
deformable part in the measuring chain is the adhesive itself. Therefore, great care must
be taken to have a very stiff fi xture and possibly also a very stiff load cell. It must be
kept in mind, however, that it is impossible to completely avoid any deformation of the
instrument during a test and this deformation must be corrected for the results to be
true material properties.40
Although prototypes with very stiff loading fi xtures and load cells may be desirable
in the academic lab, commercial instruments such as the texture analyzer are perfectly
capable of providing reproducible and reliable data if the sample holder is designed care-
fully. It is, however, important to ensure that the compliance of the apparatus does not
greatly exceed that of the fi lm one wishes to test. This may lead to large differences in the
curves that are measured, as illustrated in Figure 6.10.47
It is fi nally useful to discuss briefly the effect of the shape of the probe itself. In
1997, Chuang at Avery advertised a probe test protocol using a spherical probe.48 The
main advantage of the spherical probe is the reproducibility because alignment is no
longer a problem.7,49,50 However, the detachment of a spherical probe for a fibrillating
PSA more closely resembles a peel test than a parallel plates test and imposes a less
homogeneous stress field once cavities are fully formed. On the other hand, for weak
adhesive forces or interfacial fracture, the spherical geometry is better defi ned and the
so-called JKR analysis of the contact problem provides more meaningful data. Several
detailed analyses exist for the spherical probe geometry and have been discussed in
detail in a recent review. 36
10
5
Force (N)
−5
Sample
Motor
−10
FIGURE 6.10 Measured uncorrected force–displacement curve (dashed line) and corrected
adhesive displacement (solid line). In this case, the deformation in the initial stage of the test
(compression and tension) is mainly due to the apparatus and sample holder rather than to the
adhesive layer itself. Only after the layer fails do the two curves join each other. (From Josse, G.,
De l’Adhérence à l’Anti-Adhérence à travers le Probe Tack, Paris, Université Paris VI, 2001. With
permission.)
FIGURE 6.11 Images of cavities obtained with SIS-based PSA but three different layer thick-
nesses. The size of the cavities clearly depends on the thickness of the layer that is being tested.
(From Chiche, A., Décollement d’Adhésifs Souples: Cavitation et Fracture, Paris, Université
Paris VII, 2003. With permission.)
0.7 0.15
A
0.6 C D E
B 0.10
0.5
C
Stress σ (MPa)
0.4
0.05
D A B
0.3
E
0.00
0.2 0.6 0.8 1.0 1.2
0.1
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Strain ε
100 6
4
50
2
0 0
A B C D E
Polished probe identity
layer starts to increase faster due to cavitation than the hardening of the material itself.
The process of cavity growth is rather complex, as clearly demonstrated by some recent
theoretical arguments and numerical simulations.51–53 Hence, a direct and quantitative
interpretation of the peak stress as a cavitation stress should be avoided. Figure 6.13a
illustrates the stress–strain curves of three probe tests performed on the same probe
surface with three different materials with increasing elastic moduli. The difference in
1.4
1.2 2AA
4AA
Stress (MPa)
1.0 8AA
0.8
0.6
0.4
0.2
0.0
0 2 4 6 8
(a) Strain
107
2 AA
4 AA
106 8 AA
G (MPa)
105
104
103
10−3 10−1 101 103 105
(b) ω*a T
FIGURE 6.13 (a) Nominal stress as a function of strain for acrylics with variable acrylic acid
content. The first digit corresponds to the acrylic acid content (wt %). Probe tests were performed
at Vdeb = 100 µm/s and room temperature; (b) master curves of storage modulus G′ as a function
of reduced shear rate for three different adhesives with increasing acrylic acid content. Reference
temperature: 25°C. (From Lindner, A., Lestriez, B. et al., J. Adhesion 82(3), 267–310, 2006. With
permission.)
peak height and plateau stress can be directly attributed to the differences in the moduli,
as illustrated in Figure 6.13b. These results are in agreement with the theoretical argu-
ments presented in Section 6.2.
Examples of cavities shown in probe tests are given in Figure 6.14. One example (Fig-
ure 6.14, left) illustrates the dense population of cavities that is observed for a strong,
FIGURE 6.14 Images of cavitation of three different PSAs on the same steel surface. An SIS-
based PSA forming very small cavities that do not grow beyond a size comparable with the thick-
ness of the layer (left). An acrylic all-purpose PSA where substantial lateral propagation occurs
(center). A sample of 100 kg/mol linear poly(n-butyl acrylate) as an example of cavitation in a
liquid (right).
permanent PSA where the lateral growth of cavities is nearly precluded. When the
resistance to interfacial crack propagation is lower, the cavities can grow significantly
in the plane of the layer from their original size. An example if this is given next (Fig-
ure 6.14, center). Finally, for very soft adhesives that are liquid-like, the cavities tend
to nucleate toward the center of the probe and assume irregular shapes that resemble
Saff man–Taylor fi ngers observed in fluids (Figure 6.14, right).54,55 As discussed in Sec-
tion 2.2, it is important to remember that the pressure inside the cavities is nearly zero
until air can penetrate from the outside of the layer.11,56 Th is air penetration is accom-
panied by a drop in nominal stress corresponding roughly to 0.1 MPa. Th is drop is
generally not visible for normal PSA (see Figures 6.3 and 6.13a) where detachment of
the fibrils occurs before the air pressure can equilibrate, but it nevertheless contributes
to the measured stress. For very liquid PSA, on the other hand, such as that shown
on Figure 6.14 (right), the probe test curve can indicate a distinctive double plateau,
which is characteristic of the breakup of cavity walls without actual detachment of the
fibrils.18,57
0.30
0.25
0.20
σ (MPa)
0.15
Increasing G c
0.10
0.05
0.00
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
ε
FIGURE 6.15 Probe test curves of a removable acrylic adhesive on different surfaces. The arrow
indicates increasing Gc values for the surfaces. In all cases the propagation is mostly or completely
interfacial, with little deformation of the layer and no clear fibrillar structure. The surfaces are
three different release coatings and a steel surface (the most adhesive surface). (Data from Josse,
G., Sergot, P., Dorget, M., and Creton, C., J. Adhesion, 80(1–2), 87–118, 2004 and Josse, G., De
l’Adhérence à l’Anti-Adhérence à travers le Probe Tack, Paris, Université Paris VI, 2001.)
It is clear from Figure 6.15 that whereas the peak stress is at about the same level for
the three surfaces, the decrease in stress after the peak is very different. This difference is
due to the very different rates at which the crack propagates.40 In essence, the initiation
occurs for the same level of stored elastic energy, but the rate at which this elastic energy
can be released back to the system varies and depends on Gc. For very low values of Gc
the release rate is very fast and the stress drops almost instantaneously to zero.
This rate dependence of Gc is analogous to what is always observed in the field of
adhesion of well-cross-linked rubbers, where there is a unique relationship between Gc
and crack velocity. One can typically write35,59
v
n
Gc G0 1
v * (6.11)
Any variation in Gc can come either from a change in G 0, the component due to inter-
facial interactions, or from the viscoelastic dissipation component, which is typically
more dependent on the dissipative properties of the rubber.60,61 For adhesion of PSA on
silicone release liners, the level of adhesion is so weak that changes in the dissipative
properties of the PSA do not change Gc much, which is therefore dominated by the value
of G0. That is why release liner manufacturers tend to think of adhesion as a phenom-
enon controlled by the interface.
FIGURE 6.16 Images of the different stages of propagation of interfacial cracks at the interface
between a PSA layer and a silicone-coated probe surface. Time proceeds from top left to bottom
right. (From Léger, L. and Creton, C., Adhesion Mechanisms at Interfaces between Soft Polymers,
Philosophical Transactions of the Royal Society of London Series a-Mathematical and Physical
Sciences, 2008. With permission.)
1.6
1.4
1.2
0% SI
1.0 19% SI
σ (MPa)
0.8 42% SI
0.6
0.4
0.2 54% SI
0.0
0 5 10 15
(a) ε
0.8
0% SI
0.6
σnominal (MPa)
19% SI
42% SI
0.4
54% SI
0.2
0.0
0 5 10 15 20
(b) λ
FIGURE 6.17 (a) Probe tests of the debonding of four PSAs based on SI block copolymers
from a steel probe. (b) Tensile stress–strain curves of the same adhesives. The four adhesive con-
tain the same amount of tackifying resin but variable diblock(SI)/triblock(SIS) ratios, as speci-
fied. (From Creton, C., Roos, A., and Chiche, A., Adhesion: Current Research and Applications,
W.G. Possart, Ed., Wiley-VCH, New York, 2005. With permission.)
on steel for a series of PSAs made from SIS and styrene–isoprene (SI) block copolymers
is illustrated in Figure 6.17. The tensile curves demonstrate a distinctive strain harden-
ing that is quantitatively reproduced by the probe test curves.
The shape of probe tack curve is typical of strong adhesion and, in this case, the probe
test can be used as an approximate characterization of the tensile properties of the PSA.
Obviously, in this case the work of adhesion that is measured is rather sensitive to the
nonlinear elongational properties of the PSA and not very sensitive to the details of the
adhesive interactions, as long as they are sufficient to cause the fibril formation and stop
interfacial crack growth.
Finally, the fibrillar structure will detach from the surface. The quantitative crite-
rion for this process to occur remains relatively poorly understood, but qualitatively
the detachment of the fibrils is controlled by the strain hardening of the polymer in
the fibrils.44 The more pronounced the strain hardening, the less the fibrils can extend
before detaching from the surface of the probe. This can be seen in Figure 6.17 through a
comparison of the fibril extension at detachment with the tensile stress–strain curves.
0.6
0.5
0.4
σ (MPa)
0.3
0.2
0.1
0% SI 19% SI 42% SI 54% SI
0.0
0 2 4 6 8 10
ε
FIGURE 6.18 Probe tests of the debonding of four PSAs based on SI block copolymers from
an ethylene–propylene surface. The four adhesives contain the same amount of tackifying
resin but variable diblock(SI)/triblock(SIS) ratios, as specified. (From Creton, C., Roos, A., and
Chiche, A., Adhesion: Current Research and Applications, W.G. Possart, Ed., Wiley-VCH, New York,
2005. With permission.)
sign of a mixed failure mode (detachment and extension) is a plateau in stress that
decreases with increasing ε. This is typically never observed in tensile tests for elastic
materials and demonstrates unambiguously that the fibril-stretching process in this
case is not analogous to a tensile test, where the material cannot escape the grips of the
tensile tester and typically hardens.
In this transition regime between strong and weak adhesion, PSAs are very sensitive
to subtle changes in rheological properties. The value of Gc becomes very sensitive to the
dissipative properties of the PSA. A recent study indicated that better adhesion can be
obtained with a dissipative layer close to the surface, backed by a more elastic layer.31
6.5 Conclusion
If executed and analyzed carefully, the probe test can be a powerful analytical tool to
help interpret standard adhesive property tests of PSAs because of its sensitivity to small
changes in the molecular structure of the PSA. We have described theoretically the
process of cavity nucleation and growth and the transition from interfacial crack propa-
gation to fibril growth. Most information comes from the shape of the stress–strain
curve obtained during the debonding stage at different strain rates. A single peak with
a sharp drop in force after the peak is indicative of weak adhesion and interfacial crack
propagation. As the level of adhesion increases, the stress decrease more progressively
after the peak and forms a distinct shoulder, which then turns into a plateau in stress.
If adhesion further increases, the plateau in stress will become a second peak at higher
elongation, which immediately precedes fibril detachment.
The transition from one mechanism to the other can be described by the elastic length
Gc/E, which represents the ratio between the critical energy release rate for crack prop-
agation and the elastic modulus. When Gc/E < R0, failure occurs by crack propaga-
tion only. At the other extreme, when Gc/E > h0, failure occurs by fibril formation and
growth and the adhesion energy is essentially controlled by the nonlinear properties of
the adhesive in tension. The transition from one mechanism to the other occurs progres-
sively and in this regime adhesion energy is particularly sensitive to the value of Gc/E.
References
1. PSTC test methods for pressure sensitive adhesive tapes, Pressure Sensitive Tape
Council. 2000.
2. Zosel, A., Adhesion and tack of polymers: influence of mechanical properties and
surface tensions, Colloid Polym. Sci. 263: 541–553, 1985.
3. Zosel, A., Fracture energy and tack of pressure sensitive adhesives, Adv. Pressure
Sensitive Adhesive Technol. 1: 92–127, 1992.
4. Shull, K. R. and C. Creton, Deformation behavior of thin compliant layers under
tensile loading conditions, J. Polym. Sci.: Part B: Polym. Phys. 42: 4023–4043,
2004.
5. Zosel, A., Adhesive failure and deformation behaviour of polymers, J. Adhesion
30: 135–149, 1989.
6. Lakrout, H., P. Sergot and C. Creton, Direct observation of cavitation and fibril-
lation in a probe tack experiment on model acrylic pressure-sensitive-adhesives,
J. Adhesion 69(3/4): 307–359, 1999.
7. Crosby, A. J., K. R. Shull, H. Lakrout and C. Creton, Deformation modes of
adhesively bonded elastic layers, J. Appl. Phys. 88(5): 2956–2966, 2000.
8. Webber, R. E., K. R. Shull, A. Roos and C. Creton, Effects of geometric confinement
on the adhesive debonding of soft elastic solids, Phys. Rev. E 68: 021805, 2003.
9. Chiche, A., J. Dollhofer and C. Creton, Cavity growth in soft adhesives, Eur.
Phys. J. E 17: 389–401, 2005.
10. Gent, A. N., Compression of rubber blocks, Rubber Chem. Technol. 67(3): 549–558,
1994.
11. Lindner, A. and B. Lestriez et al., Adhesive and rheological properties of lightly
crosslinked model acrylic networks, J. Adhesion 82(3): 267–310, 2006.
12. Creton, C. and H. Lakrout, Micromechanics of flat probe adhesion tests of soft
viscoelastic polymer fi lms, J. Polym. Sci.: Part B: Polym. Phys. 38: 965–979, 2000.
13. Kakavas, P. A. and P. J. Blatz, Effects of voids on the response of a rubber poker
chip sample .3, J. Appl. Polym. Sci. 43(6): 1081–1086, 1991.
14. Laun, H. M., M. Rady and O. Hassager, Analytical solutions for squeeze flow with
partial wall slip, J. Non-Newtonian Fluid Mech. 81: 1–15, 1999.
15. Roos, A., Sticky Block Copolymers: Structure, Rheological and Adhesive Properties.
Paris, Université Paris VI, 2004.
16. Chiche, A., Décollement D’adhésifs Souples: Cavitation et Fracture. Paris, Université
Paris VII, 2003.
17. Tirumkudulu, M., W. B. Russell and T. J. Huang, On the measurement of “tack”
for adhesives, Phys. Fluids 15(6): 1588–1605, 2003.
18. Poivet, S., F. Nallet, C. Gay, J. Teisseire and P. Fabre, Force response of a viscous
liquid in a probe-tack geometry: Fingering versus cavitation, Eur. Phys. J. E 15:
97–116, 2004.
19. Brown, K., J. C. Hooker and C. Creton, Micromechanisms of tack of soft adhesives
based on styrenic block copolymers, Macromol. Mater. Eng. 287(3): 163–179,
2002.
20. Gay, C. and L. Leibler, Theory of tackiness, Phys. Rev. Lett. 82(5): 936–939,
1999.
21. Gent, A. N. and P. B. Lindley, Internal rupture of bonded rubber cylinders in ten-
sion, Proc. R. Soc. London, Ser. A: Math. Phys. Sci. 249 A: 195–205, 1959.
22. Gent, A. N. and D. A. Tompkins, Surface energy effects for small holes or particles
in elastomers, J. Polym. Sci., Part A-2 Polym. Phys. 7: 1483–1488, 1969.
23. Dollhofer, J., A. Chiche, V. Muralidharan, C. Creton and C. Y. Hui, Surface energy
effects for cavity growth and nucleation in an incompressible neo-Hookean mate-
rial—modeling and experiment, Int. J. Solids Struct. 41(22–23): 6111–6127, 2004.
24. Chikina, I. and C. Gay, Cavitation in adhesives, Phys. Rev. Lett. 85: 4546–4549,
2000.
25. Lin, Y. Y., C. Y. Hui and H. D. Conway, A detailed elastic analysis of the flat punch
(tack) test for pressure sensitive adhesives, J. Polym. Sci.: Part B: Polym. Phys.
38(21): 2769–2784, 2000.
26. Chiche, A., P. Pareige and C. Creton, Role of surface roughness in controlling the
adhesion of a soft adhesive on a hard surface, Comptes Rendus Acad. Sci. Paris Ser.
IV 1: 1197–1204, 2000.
27. Lau, A. W. C., M. Portigliatti, E. Raphaël and L. Léger, Spreading of latex particles
on a substrate, Europhys. Lett. (5): 717, 2002.
28. Gent, A. N. and C. Wang, Fracture mechanics and cavitation in rubber-like solids,
J. Mater. Sci. 26: 3392–3395, 1991.
29. Fond, C., Cavitation criterion for rubber materials: a review of void-growth
models, J. Polym. Sci.: Part B: Polym. Phys. 39: 2081–2096, 2001.
30. Creton, C., J. C. Hooker and K. R. Shull, Bulk and interfacial contributions to
the debonding mechanisms of soft adhesives: Extension to large strains, Langmuir
17(16): 4948–4954, 2001.
31. Carelli, C., F. Déplace, L. Boissonnet and C. Creton, Effect of a gradient in
viscoelastic properties on the debonding mechanisms of soft adhesives, J. Adhe-
sion 83(5): 491–505, 2007.
32. Williams, M. L. and R. A. Schapery, Spherical flaw instability in hydrostatic
tension, Int. J. Fracture Mech. 1: 64–71, 1965.
33. Lawn, B., Fracture of Brittle Solids. Cambridge, Cambridge University Press.
1993.
34. Lin, Y. Y. and C. Y. Hui, Cavity growth from crack-like defects in soft materials,
Int. J. Fracture 126(3): 205–221, 2004.
35. Maugis, D. and M. Barquins, Fracture mechanics and the adherence of viscoelas-
tic bodies, J. Phys. D: Appl. Phys. 11: 1989–2023, 1978.
36. Shull, K. R., Contacts mechanics and the adhesion of soft solids, Mater. Sci. Eng. R,
Report 36: 1–45, 2002.
37. Zhang Newby, B.-M., M. K. Chaudhury and H. R. Brown, Macroscopic evidence of
the effect of interfacial slippage on adhesion, Science 269: 1407–1409, 1995.
38. Chaudhury, M. and B.-M. Zhang Newby, Friction in adhesion, Langmuir 14:
4865–4872, 1998.
39. Amouroux, N., J. Petit and L. Léger, Role of interfacial resistance to shear stress
and adhesive peel strength, Langmuir 17: 6510–6517, 2001.
40. Josse, G., P. Sergot, M. Dorget and C. Creton, Measuring interfacial adhesion
between a soft viscoelastic layer and a rigid surface using a probe method, J. Adhe-
sion 80(1–2): 87–118, 2004.
41. Creton, C., Pressure-sensitive-adhesives: An introductory course, MRS Bull. 28(6):
434–439, 2003.
42. Creton, C., A. Roos and A. Chiche Effect of the diblock content on the adhesive
and deformation properties of PSAs based on styrenic block copolymers. in: W. G.
Possart (eds). Adhesion: Current Research and Applications. Wiley-VCH, 2005.
43. Roos, A. and C. Creton, Nonlinear elastic properties of elastomeric block copoly-
mers, Macromolecules 38: 7807–7818, 2005.
44. Glassmaker, N. J., C. Y. Hui, T. Yamaguchi and C. Creton, Detachment of stretched
viscoelastic fibrils, Eur. Phys. J. E 25: 253–266, 2008.
45. Hammond, F. H., Polyken probe tack tester, ASTM Spec. Technical Publ. 360:
123–134, 1964.
46. Zosel, A., The effect of bond formation on the tack of polymers, J. Adhesion Sci.
Technol. 11: 1447–1457, 1997.
47. Josse, G., De l’Adhérence à l’Anti-Adhérence à travers le Probe Tack. Paris, Université
Paris VI, 2001.
48. Chuang, H. K., C. Chiu and R. Paniagua, Avery adhesive test yields more perform-
ance data than traditional probe, Adhes. Age (September): 18–23, 1997.
49. Crosby, A. J. and K. R. Shull, Adhesive failure analysis of pressure-sensitive
adhesives, J. Polym. Sci.: Part B: Polym. Phys. 37: 3455–3472, 1999.
50. Crosby, A. J., K. R. Shull, Y. Y. Lin and C. Y. Hui, Rheological properties and
adhesive failure of thin viscoelastic layers, J. Rheol. 46(1): 273–294, 2002.
51. Yamaguchi, T. and M. Doi, Debonding dynamics of pressure-sensitive adhesives:
3D block model, Eur. Phys. J. E 21(4): 331–339, 2006.
52. Yamaguchi, T., H. Morita and M. Doi, Modeling on debonding dynamics of
pressure-sensitive-adhesives, Eur. Phys. J. E 20(1): 7–17, 2006.
53. Teisseire, J., F. Nallet, P. Fabre and C. Gay, Understanding cracking versus
cavitation in pressure-sensitive adhesives: The role of kinetics, J. Adhesion 83(7):
613–677, 2007.
54. Saff man, P. G. and G. Taylor, The penetration of a fluid into a porous medium
or Hele-Shaw cell containing a more viscous liquid, Proc. R. Soc. London, Ser. A:
Math. Phys. Sci. 245: 312–329, 1958.
55. Derks, D., A. Lindner, C. Creton and D. Bonn, Cohesive failure of thin layers of
soft model adhesives under tension, J. Appl. Phys. 93(3): 1557–1566, 2003.
56. Poivet, S., F. Nallet, C. Gay and P. Fabre, Cavitation-induced force transition in
confined viscous liquids under traction, Europhys. Lett. 62(2): 244–250, 2003.
57. Lakrout, H., C. Creton, D. Ahn and K. R. Shull, Influence of molecular features on
the tackiness of acrylic polymer melts, Macromolecules 34: 7448–7458, 2001.
58. Léger, L. and C. Creton, Adhesion Mechanisms at Interfaces between Soft Polymers,
Philosophical Transactions of the Royal Society of London Series a-Mathematical
and Physical Sciences 366: 1425–1442, 2008.
59. Shull, K. R., D. Ahn, W. L. Chen, C. L. Mowery and A. J. Crosby, Axisymmetric
adhesion tests of soft materials, Macromol. Chem. Phys. 199: 489–511, 1998.
60. Ahn, D. and K. R. Shull, Effects of methylation and neutralization of carboxylated
poly(n-butyl acrylate)on the interfacial and bulk contributions to adhesion, Lang-
muir 14(13): 3637–3645, 1998.
61. Ahn, D. and K. R. Shull, Effects of substrate modification on the interfacial
adhesion of acrylic elastomers, Langmuir 14: 3646–3654, 1998.
7.1 Introduction
According to a common definition, pressure-sensitive adhesives (PSAs) can be adhered
to various surfaces with light pressure within a few seconds. In certain cases, they also
can be removed without leaving any residue or contaminating the substrate. Such behav-
ior is due to their ambiguous, solid-like, as well as liquid-like behavior, that is, their vis-
coelasticity (see also Chapters 4 and 5). The former behavior (called cohesion) gives high
strength in the debonding process, whereas the latter behavior (called adhesion) allows
wetting of the surface during the bonding process.
The adhesive performances of PSAs can be evaluated as tack, peel resistance, and
cohesive strength, such as holding power and shear adhesion failure temperature
(SAFT) (see also Applications of Pressure-Sensitive Products, Chapter 8). Among these
performances, peel resistance is an especially important parameter for PSAs. Peel resis-
tance is the force required to peel off a pressure-sensitive product (label, tape, etc.) from
a substrate. According to ASTM D 907 [1], the peel resistance is the average load per
unit width of bondline required to progressively separate a flexible member from a rigid
member or another flexible member.
Control of peel resistance in PSA products is very important for both manufacturers
and customers. For instance, in the label industry, the value of peel resistance is one of
7-1
Source: Czech, Z., J. Appl. Polym. Sci., 97, 886–892, 2005. With
permission.
the criteria used to classify applications. Labels are usually classified by peel resistance
as permanent or removable (see also Applications of Pressure-Sensitive Products, Chapter 1).
PSAs used for permanent labels possess high peel resistance. Such labels are definitively
laminated on the substrate. On the other hand, PSAs for removable labels have low peel
resistance, which allows their slight detachment from substrates. A common classifi-
cation of PSAs according to the value of peel resistance is given in Table 7.1 [2]. (Such
classification is realistic in connection with the bulk and surface strength of the carrier
material only. For instance, a peel resistance value of 14 N/25 mm causes paper-tear; that
is, for paper, such PSA can ensure an excellent permanent bond. However, this behavior
depends on the paper strength as well. Removability is peel value dependent, but reposi-
tionability is peel build-up dependent. That means that peel values of 2–4 N/25 mm, as
given in Table 7.1, can impart or not impart repositionability.)
In this chapter, we will examine the research regarding peel resistance, both to defi ne
it and to illustrate its importance. An understanding of the origin of peel resistance, the
mechanics of peel resistance, and its theory is essential. Moreover, the failure mechanism
must be also considered (see also Applications of Pressure-Sensitive Products, Chapter 8)
Finally, correlation between the PSAs’ bulk/surface properties and peel resistance will
be discussed. The bulk/surface properties and peel resistance of PSAs are affected by
various factors, such as chemical structure, tackifier type and content, cross-linking
density, and viscoelastic properties.
peel angle and peel rate. Kaelble [5] extended the elastic analysis to include viscoelastic
peeling. Kaelble described his peel resistance theory in early papers in 1959 [4]. The
descriptive assumptions for an idealized type of peel resistance (simple stripback) are
listed as follows:
1. A steady-rate debonding proves
2. Properties of the flexible member
a. Of slender rectangular cross-section
b. Elastically deformed
c. Bonded and unbonded portions of semi-infinite length
3. Properties of the forces
a. Applied force—acts in tension along the central plane of flexible members
b. Shearing forces—distributed over the bonded interface of the flexible member
c. Cleavage forces—a highly localized parallel array, perpendicular to the bond
plane
d. All forces uniformly distributed across the bond width
Under these assumptions, the deformation curve of the flexible member under tension
was characterized using the infi nite analysis method. Figure 7.1 illustrates the peeling
behavior for cellophane/rubber–resin-based adhesive tape detached from the cello-
phane surface at various peeling angles. If the summation of moments of tensional and
compressional forces in the bond are independent of the peel angle and cleavage is the
controlling failure mechanism, then peel strength should vary inversely as (1 − cosω).
Figure 7.1 presents steady-state peel strength data as a function of peel angle [5]. At low
stripping rates and accompanying lower stripping force, experimental data confirm the
theoretical inverse P versus (1 − cosω) relationship over a wide range of angles. At higher
10 44.48
Rate (in./min)
20
2.0
5 0.2 22.24
Peel strength (LB)
0.02
Peel strength (N)
2 8.90
1 4.45
0.5 2.22
FIGURE 7.1 Peel force (P) versus angle (w) for aluminum foil tape. Peeling rate: 0.02, 0.2, 2.0,
and 20 (in./min). (From Kaelble, D.H., Trans. Soc. Rheol., 4, 45–73, 1960. With permission.)
stripping rates, the deviation becomes more marked due to the dual effect of tensile
and flexural strains upon Young’s modulus of the backing. At low rates of peeling, a
significant fall-off of jog in value of the peel force was observed when the peeling angle
was in the range of 20∼40°. Kaelble [5] suggested that cleavage stress and shear stress
interacted in some way to cause the unusual fall-off in peel force. This phenomenon was
explained by the fact that the jog was associated with a relatively rapid change in deco-
hesion (fracture) mechanism, according to Williams and Kauzlarich [6]. When peeling a
flexible tape from a solid surface, there can exist a transition from a mechanism of deco-
hesion relying on cleavage to one with a much greater component of shear. This change
in mechanism requires that the specific energy for decohesion is less than decohesion
by cleavage. This switch in mechanism is apparent only at low peeling angles, typically
<40°, when the component of the applied tape tension acting parallel to the adhesive
interface reaches some critical value. This critical value of peel force is influenced by
the residual stress generated within the tape by the process of attachment to the rigid
substrate.
Figure 7.2 illustrates the diagrammatic view of the separation zone during peeling [7].
Gent and Meinecke [8] demonstrated that the stiff ness of such flexible sandwiched
blocks depends on a “shape factor,” S, which is defined as the ratio of area of one of the
loaded faces of the material to the area of the stress-free block. In peeling, the adhesive
in advance of the peel front is constrained and exhibits an enhanced stiff ness, which
may well be drawn out into fibrils. Figure 7.3 presents the value of S by ratio of the area.
Peel resistance does not measure the adhesion bond strength directly, but the sum
of the energy required to break the bond and to deform the backing and adhesive, as
demonstrated in the following equation [9]:
Gc G0 (7.1)
Backing Unconstrained
adhesive forming
fibrils
ha
L
Constrained
adhesive
Peel front Substrate
FIGURE 7.2 Diagrammatic view of the zone of separation in peeling. (From Williams, J.A.,
and Kauzlarich, J.J., J. Adhesion Technol., 21(7), 515–529, 2007. With permission.)
90°
Peeling
direction Foam of Substrate:
strengthening cataphoresis
PSA
A
O
Bending Start of
zone removing
FIGURE 7.3 Schematic representation of the PSA peel test. (From Horgnies, M., Darque-
Ceretti, E., and Felder, E., Int. J. Adhesion Adhesives, 27, 661–668, 2007. With permission.)
where Gc is the adhesive fracture energy, G 0 is the energy required to propagate a crack,
which is a direct measure of the bonding forces, and ψ is the energy dissipated visco-
elastically within the adhesive and backing. The value of ψ is the major contributor to
the value of Gc and it is highly dependent on the rate and temperature of testing. We can
measure Gc, but we cannot measure G 0. We can only estimate its value, which is fairly
modest compared with the measured peel force values. This indicates that ψ is indeed
an important factor in peel measurements. However, ψ is a function of G 0. If G 0 = 0, ψ
is also equal to zero.
The facture energy dissipation during debonding in a peel test seems to be connected
to the formation and growth of fibrils during bond separation. Fracture energy dissipa-
tion is observed for polymers with a molecular mass between entanglements of about
104 g/mol, whereas below this limit debonding occurs by homogeneous deformation.
Figure 7.3 illustrates the schematic representation of the 90° peel test [10]. The bending
line (located at point O) is characteristic of the propagation of the crack between the
adhesive and substrate. The start of removing line (located at point A) corresponds to the
specific zone where the PSA fibrils are beginning to elongate. The facture energy, G, can
be acquired by adhesion force, P, peel angle θ, and the bandwidth using Equation 7.2.
P
G (1 cos ) (7.2)
b
Deformation of the PSA and the real contact area during peeling can be investigated using
a fast charge-coupled device (CCD) camera and optical microscopy. Figure 7.4 presents
Fibrils of 15 mm Fibrils of 3 mm
elastomer acrylic
PSA "Be" PSA "Ba"
Foam of
strengthening
(a) (b)
FIGURE 7.4 Fibril elongation of elastomer PSAs (a) and acrylic PSAs (b) observed using a fast
CCD camera during the stationary regime of peeling. (From Horgnies, M., Darque-Ceretti, E.,
and Felder, E., Int. J. Adhesion Adhesives, 27, 661–668, 2007. With permission.)
30
20
15
10
0
0 1 2 3 4 5 6 7 8 9 10
Distance from the beginning of the bending zone (cm)
FIGURE 7.5 Fibril length distribution compared with the initial thickness of PSAs and accord-
ing to the distance from the beginning of the bending zone. (From Horgnies, M., Dargue-Ceretti,
E., and Felder, E., Int. J. Adhesion. Adhesives 27, 661–668, 2007. With permission.)
pictures taken every 8 ms during the stationary regime of peeling (peel resistance is
almost constant) [10]. Figure 7.4 highlights fibrils that appeared during the (a) peel test
of elastomer and (b) acrylic PSAs. Figure 7.5 illustrates the fibril length distribution of
the PSA according to the distance (see Figure 7.3). The profiles of adhesive deformation
and measured forces were constant during the peel test. Horgnies et al. [10] developed an
original peeling procedure that can detect the local fracture energy and investigated the
relationship between PSA deformation and local fracture energy. The interface between
the PSA and glass substrate was observed by optical microscopy to measure the con-
tact area. The interface between the PSA and the glass substrate is composed of adhesive
and gaseous bubbles that are induced from the noncontact area. The increase in fracture
energy is in direct relation to the contact ratio between the PSA and substrate.
A schematic diagram representing the side profi le of a PSA tape undergoing peel is
illustrated in Figure 7.6 [11]. Below the view of the bond is a typical stress distribution
profi le as recorded by the measuring instrument, a newly designed bond stress ana-
lyzer. The instrument is a true accessory to the Instron tester in that the force-sensing
substrate uses an Instron load cell as the force transducer. The weighing system is also
utilized to measure and recorded the bond forces. As indicated in Figure 7.6, the bond
stress is highly localized at the boundary of the bond. Tensional stress, which character-
izes the micromechanisms of unbonding, is observed in the adjacent boundary region.
Maximum compressional stress and compressional zone length are produced due to
leverage of the flexible member. A low tensional stress zone appears as one proceeds
farther away from the boundary and then the normal stress effectively drops to zero.
The stress distribution displayed in Figure 7.6 strongly indicates that an important step
during the fracture process is microcavitation of the adhesive layer, which is similar to
the cavitation observed in the course of the probe tack test (see Chapters 4 and 6). The
stress profi le of Figure 7.6 indicates a very rapid change in bond stress in the region
between σc and σ t. In this narrow zone of the bond, it is evident that small spherical
cavities are formed within the adhesive phase of the bond. Williams and Kauzlarich [7]
have recently employed the finite element method for the calculation of stress distribu-
tion in the adhesive layer along with the corresponding strain from 90° peel test results.
Obtaining the stress–strain curves allows us to gain an insight into the viscoelastic
behavior of PSAs during peeling.
A fundamental method of defining bond strength is expressed in energy or work.
The definition of peel work (W) by Kaelble [4,5] isolates two separate contributions, as
follows,
W WT WD (7.3)
Backing material
Adhesive
Adherend
6
5 σt
4
Tension
Stress
3
2
σx (kg/cm2)
1
0
−1
Compression
−2
Stress
−3
−4
−5
σc
−6
−0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Bond distance X (mm)
FIGURE 7.6 Schematic representation of the peel profi le and experimental boundary dis-
tribution of normal stress. (From Kaelble, D.H., Trans. Soc. Rheol., 9(2), 135–163, 1965. With
permission.)
of PSAs, tapes were peeled at temperatures above and below room temperature to deter-
mine the usability of a product at various temperatures. In other cases, PSA tapes were
peeled and reapplied several times to determine the product’s release ability (see also
Applications of Pressure-Sensitive Products, Chapter 8). The properties and performance
of a PSA are affected by several factors, including [13] the following:
1. Coating weight
2. Adhesive composition
3. Substrates to which it is applied
4. Temperature
5. Bonding pressure
6. Residence time before breaking the bond
7. Type and characteristics of the surface where the adhesive is applied
(a) (b)
(c) (d)
FIGURE 7.7 Various peel test methods: (a) 180° peel, (b) 90° peel, (c) drum peel (tape unwind),
and (d) T-peel. (From Satas, D., Handbook of Pressure Sensitive Adhesive Technology and Applica-
tions, D. Satas, Ed., Satas & Associates, Warwick, RI, 2002. With permission.)
Chapter 8). Many researchers test peel strength using 90° or 180° peel strength. The
drum test measures the property of the unwinding process, which is an important
property of tapes (see Applications of Pressure-Sensitive Products, Chapters 4 and 8). The
T-peel test measures the force required to break the bond between two adhesive layers.
This test is also used to measure the force needed to separate the tape from a flexible
adherend (see Applications of Pressure-Sensitive Products, Chapter 8). Figure 7.8 illus-
trates tools for (a) 180° and (b) 90° peel tests (see Applications of Pressure-Sensitive Prod-
ucts, Chapter 8). Another representative method for 90° peel is the mandrel peel test [14].
This method involves peeling around a circular mandrel while applying an alignment
load to the base of the laminate to ease abutment of the peel arm to the mandrel.
There are two major categories in peel testing—dynamic and static. For the purposes
of definition, dynamic peel is measured by a force applied at some given rate of peeling,
and the unit is the force per unit area. Most of the standard peel testing methods fall
into this category. In the case of static peel testing, the force is applied by a fi xed weight
and the time to fail is recorded. ASTM D2860 and PSTC 14 fall into this latter category.
Industry organizations (ASTM, TLMI, and PSTC and FINAT and AFERA in Europe)
have devised standard test methods that cover peel testing. For convenience, they are
listed for comparison in Table 7.2 [15] (see also Applications of Pressure-Sensitive Prod-
ucts, Chapter 8).
The standard specimen for peel strength testing is a 1 in. wide strip of tape. The speci-
men is adhered to a clean stainless-steel panel without air bubbles and is pressed by a
constant weight of roller. After sufficient bonding time, the peel specimen is loaded on
the test machine. One side of the strip is placed in one side jaw, whereas the other side
jaw fastens to the test panel [16,17].
P
(a) (b)
FIGURE 7.8 Tools for (a) 180° peel test and (b) 90° peel test. (From Kawashita, L.F., Moore,
D.R., and Williams, J.G., J. Adhesion, 80, 147–167, 2004; ASTM D3330. With permission.)
Organization Method Peel Angle (°) Test Speed Bonding Time Notes
ASTM D903 180 12 in./min Open P1
ASTM D1000 180 12 in./min 20 min P6
ASTM D2860 90 Static 3 min P2
ASTM D3330 180 12 in./min <1 min P3
TLMI L-IA 1 180 12 in./min Open P1
PSTC 1 180 12 in./min <1 min P3
PSTC 2 90 12 in./min <1 min P4
PSTC 3 180 12 in./min <1 min P5
PSTC 14 90 Static 3 min P2
FINAT FTM 1 180 300 mm/min 20 min, 24 h —
FINAT FTM 2 90 300 mm/min 20 min, 24 h —
AFERA 4001 180 300 mm/min 10 min —
P1 ASTM D903 and TLMI L-IA 1 describe the method, but do not specify residence times prior to peel-
ing. The TLMI method suggests residence times ranging from 30 min to 24 h.
P2 These methods are similar and describe static peel test from a standard linerboard material.
P3 These methods are virtually identical.
P4 These methods are similar.
P5 Peel test taking into consideration the needs for testing double-faced PSA product.
P6 There are two procedures in this method. Method A uses flat stainless plates, whereas Method B uses
stainless drums.
Source: Urahama, Y., Tokunaga, Y., and Tanaka, Y., J. Adhesion Soc. Jpn., 23(5), 171–177, 1990. With
permission.
Variances of the standard peel test methods, product-specific peel test methods, and
other methods for tests of peel resistance as well the dependence of peel resistance on
different experimental parameters are described by Benedek (see also Applications of
Pressure-Sensitive Products, Chapter 8).
C1(T Ts )
log aT (7.4)
C2 T Ts
a reference temperature, Ts, which was allowed to be an adjustable parameter, but gener-
ally fell about 50°C above the glass transition temperature, Tg, as follows:
8.86(T Ts )
log aT (7.5)
101.6 T Ts
17.5(T Tg )
log aT (7.6)
51.6 T Tg
The existence of such a useful universal equation for the shift factor formed the basis
for many experiments and theories regarding the viscoelasticity of polymers. Based on
these ideas, the time–temperature superposition principle states that for viscoelastic
materials, time and temperature can be superimposed onto data at another temperature
by shift ing the curves along the time axis. In other words, time–temperature superposi-
tion allows prediction of peeling energies: (1) at a lower rate by shift ing data at T > T0 to
the left and (2) at high rates by shifting data at T < T0 to the right. The curve obtained
after the shifts is called a master curve. However, the range of peel rates for which some-
one can perform experiments is limited, because extremely low or high peel rates are
hard to maintain and difficult to measure. If we conduct a peel test for a small rate
range and measure the strength over a wide range of measurable temperatures, we can
obtain the family of curves shown in Figure 7.9. The temperatures are in the order of
T1 > T2 > T3 > T4 > T5 > T6. Examination of these generic data demonstrates that shift-
ing the curves for temperatures T1 through T5 to the left from the standard temperature, T6,
by some amount will cause all of the curves to form a smooth curve. This is the case for
most PSAs. The result is known as a master curve and the amount by which a segment is
T6
T5
Peel strength
Peel strength
T4
T3
T2
T1
Rate log Va T
Temperature
FIGURE 7.9 The making of a master curve. (From Pocius, A.V., Adhesion and Adhesives
Technology, A.V. Pocius, Ed., Hanser, Munich, 1997. With permission.)
shifted is known as a shift factor. Thus, a master curve is a plot of the peel strength as a
function of a reduced peel rate. The reduced variable is obtained by multiplying the peel
rate by the shift factor. Thus, the set of curves illustrated in Figure 7.9a can be reduced to
a single curve, illustrated in Figure 7.9 [18]. Although direct testing of peel energy may
be limited to a relatively narrow range of test speeds, time–temperature superposition
allows prediction of peeling energies. Therefore, we can overcome the limit of the peel-
ing tester and predict the peel resistance in an extended test range.
Derail et al. [19] demonstrated that the shift factors derived from peel adhesion mas-
ter curve construction were equivalent to those derived via rheologic measurement (see
Chapter 4). Peel master curves are therefore examples of rheologic master curves. Peel-
ing behavior through the construction of peel master curves can reveal the influence of
adhesive molecular features, including molecular weight and composition.
Failure mode location in peeling tests depends only on the rheological properties of
the PSA at the temperature considered. This is a very important statement for industrial
application. The fact that fracture does not depend on the location of crack initiation
means that within a controlled range of temperature and deformation rates, the user
may trust the adhesive even when some defect initiates a crack in fracture mode, which
is different from what is expected.
B-C
A B C
P
Peel strength (N)
B
A
B-C
FIGURE 7.10 Failure modes as a function of peeling rate. (From Urahama, Y., Tokunaga, Y.,
and Tanaka, Y., J. Adhesion Soc. Jpn., 23(5), 171–177, 1990. With permission.)
3.5
3.0
2.0
1.5
1.0
0 1 2 3 4 5 6 7
Length of tape peeled (cm)
FIGURE 7.11 Peel strength–strain curve demonstrating stick–slip peeling. (From Aubrey,
D.W., Welding, G.N., and Wong, T., J. Appl. Polym. Sci., 13, 2193–2207, 1969. With permission.)
adhesive residue on the test panel. If the adhesive is not fi rmly anchored to the backing
material, it may transfer from the backing material to the test panel, leaving at least
part of the backing bare. In the case of transfer tapes, such behavior is designed inten-
tionally; otherwise, it denotes product failure (see Applications of Pressure-Sensitive
Products, Chapter 4). The facture shapes of four failure modes in peeling experiments
are illustrated in Figure 7.10. Urahama et al. [15] observed that at relatively low peel-
ing speeds below 10 mm/min, the stringiness conformation of adhesives of the porous
backing had a honeycomb structure, whereas that of the nonporous backing had a saw-
tooth-shape structure.
In the region of intermediate pulling rate, the mode of failure is stick–slip, which is a
regular, jerky peel in which the observed peel force oscillates between well-defi ned lim-
its. An autographic recording obtained in this region is illustrated in Figure 7.11 [20]. In
this region, the alternative failure between adhesive and glassy failure was observed. The
alternations of force were in step with the alternations of failure mode and, furthermore,
the rising and falling parts of the autographic trace corresponded with cohesive and
adhesive separations, respectively. The steady rate of pulling is insufficient to sustain
peeling with the fast mechanism, so the mechanism reverts to the slow mechanism. This
completes the cycle of stick–slip oscillation.
In the case of adhesive failure, a general judging standard is the existence of adhesive
residue. As previously mentioned, the test method must be selected considering the
application of the adhesive (see Applications of Pressure-Sensitive Products, Chapter 4).
For example, a special PSA tape is used to fabricate semiconductor chips [21]. Ultra-
violet (UV)-curable dicing tape used in the die-bond process for semiconductors must
hold tightly upon mounting of a silicon wafer. During the pick-up process, the adhe-
sion strength of the dicing tape must be reduced with no residue so that the diced chip
can be picked up easily. In this case, human sight is not enough to determine adhesive
failure mode with no residue. Additional instrumental analysis like optical micros-
copy, Fourier transform infrared spectroscopy, or X-ray photoelectron spectroscopy
(XPS) may be carried out for an exact determination. Th is is why with the recent rapid
development in integrated semiconductor technology, higher reliability is required for
the manufacturing process of electronic devices.
Preparation of
PSA
Coating onto
backing material
Bonding and
sample stabilization
Peeling
Data analysis
FIGURE 7.12 The flow chart of sample preparation and peel testing.
TABLE 7.3 Parameters of Processing and Testing that Affect Peel Resistance
Stage of Processing
and Testing Parameters Affecting Peel Resistance
Preparation of PSA Chemical composition and cross-linking nature and density
Viscoelastic properties
Miscibility between PSA and other formulation components
Coating onto carrier material Modulus of carrier material
PSA thickness
Thickness of backing material
Surface properties of carrier material
Preparation of sample Sample width
Cleaning substrate Surface properties of substrate (surface energy and roughness)
Surface treatment
Degree of pollution
Bonding step and sample Bonding pressure
stabilization Bonding time
Peeling test Peeling angle or geometry and peeling tool
Peeling rate
Temperature and humidity
1600 100
Peel strength
1200
60
SAFT (°C)
40
800
20
400 0
5 10 15
Acrylic acid contents (wt %)
FIGURE 7.13 Adhesion properties such as probe tack, peel strength, and SAFT with variation
in AA content. (From Joo, H.S., Do, H.S., Park, Y.J., and Kim, H.-J., J. Adhesion Sci. Technol.,
20(14), 1573–1594, 2007. With permission.)
Comonomers, such as methyl methacrylate and styrene, which increase the Tg of the
adhesive, can be used in limited amounts to improve cohesive strength. Much research
exists regarding the formulation of acrylates on the properties of PSAs [23–26]. Gower
and Shanks [27–29] discussed the formulation for polymerization on adhesive perfor-
mance and peel master curve. They explained the relationship between comonomers
and PSA performance.
Figure 7.13 illustrates the effect of chemical composition of PSAs on peel resistance
[30]. PSAs with varying amounts of acrylic acid (AA) were investigated, and the expected
effects of AA addition were good adhesion to substrates, enhancement of cohesion, and
increased viscosity. The Tg of the PSAs increased with increasing AA concentration,
because AA has the highest Tg among monomers such as 2-EHA, vinyl–acetate (VAc),
and AA. Peel strength slightly increased as AA concentration increased, although PSAs
with an AA content of 10 and 15 wt % demonstrated somewhat similar peel strengths.
These results can be explained, because the increase in AA concentration affected the
viscosity and cohesive strength of synthesized PSA through hydrogen bond formation.
Therefore, the peel strength of PSAs increased with AA concentration.
The relationship between peel resistance and pulling rate has been examined using a
range of poly(butyl acrylate) homopolymer adhesives with different molecular weights
[20]. Results are illustrated in Figure 7.14. In the slow peeling region, the peel resistance
increases with molecular weight, as would be expected if viscous flow of the adhesive is
the controlling factor. The rate at which the transition to stick–slip occurs increases with
decreasing molecular weight, although with the two lowest molecular weight adhesives
this transition did not occur, even at the highest cross-head speed obtainable. In the fast
5
e d
Force range at stick−slip peeling
4 c
Peel strength (kg/25 mm)
Range of force maxima
a b
3 Range of force minima
d
b
2 a e
e
1 a
b
f
0
−1.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0
Log rate of jaw separation (cm/min)
FIGURE 7.14 Effect of molecular weight of poly(n-butyl acrylate) on peel strength at various rates
of jaw separation. (Molecular weight: a > b > c > d > e > f, ○, cohesive failure; ●, adhesive failure
between adhesive and glass surface (substrate); □, adhesive failure between adhesive and backing;
△, mixture of two types of adhesives.) (From Aubrey, D.W., Welding, G.N., and Wong, T., J. Appl.
Polym. Sci., 13, 2193–2207, 1969.)
peeling region, the amount of data available is small but gives no indication of appre-
ciable dependence of peel resistance on molecular weight.
To study the relationship between cross-linking density and peel resistance, UV-
cross-linkable PSAs were synthesized using the solution polymerization method using
2-EHA, VAc, AA, 2-hydroxyethyl methacrylate, and 4-acryloyloxydiethoxy-4-chloro-
benzophenone (P-36) as polymerizable photoinititators with double bonds. The effect
of photoinitiator content on the PSAs is illustrated in Figure 7.15 [31]. As expected, the
higher photoinitiator content in SH3P2 produced a decrease in peel resistance of 69.8%
at a UV dose of 210 mJ/cm2, whereas SH3P05 demonstrated only a 9.1% reduction in peel
resistance with the same dose. The peel resistance is also influenced by the Tg. If the Tg of
the PSA is sufficiently low, the wettability will be high after application to the substrate
and, as a result, the peel resistance will also increase.
Czech [2] synthesized removable and repositionable water-borne acrylic PSAs. In this
study, the effect of copolymerizable emulsifier and plasticizer was investigated. Figure
7.16 illustrates how the influence of vinyl-unsaturated emulsifiers on peel resistance on
steel depends on the type of investigated internal emulsifiers and their concentration.
The relatively low peel adhesion is observed for the vinyl-unsaturated emulsifier SPMK.
The increased amount of SPMK above 5 wt % has very little influence on removability.
One possibility of reducing peel resistance in the water-based PSAs lies in the varia-
tion of the plasticizer used. The plasticizer, such as di-n-butyl phthalate, diethylhexyl
phthalate, and di-n-octyl phthalate, was used in an amount between 10 and 30 wt %. Its
influence is illustrated in Figure 7.17.
4000
PI content (phr)
0.5
1.0
Peel resistance (g/25 mm)
3000
2.0
2000
1000
0
0 500 1000 1500 2000 2500
UV dose (mJ/cm2)
FIGURE 7.15 Change in peel resistance of UV cross-linkable PSAs with varying UV dose (PI,
photoinitiator) (*phr:part). (From Do, H.S., Park, Y.J., and Kim, H.-J., J. Adhesion. Sci. Technol.,
20(13), 1529–1545, 2006. With permission.)
7
Peel adhesion (N/2.54 cm)
5
AMPSNa
4
SPIK
3 Removable
2
1 SPMK
0
0 1 2 3 4 5 6 7 8
Internal emulsifier concentration (wt %)
FIGURE 7.16 Influence of internal emulsifiers on peel adhesion on steel (SPMK, potassium
salt of sulfopropyl methacrylate; SEMNa, sodium salt of sulfoethyl methacrylate; SPIK, bis-
(3-sulfopropyl)-itaconic acid ester dipotassium salt; AMPSNa, sodium salt of 2-acrylamido-
2-methylpropyl sulfonic acid). (From Czech, Z., J. Appl. Polym. Sci., 97, 886–892, 2005. With
permission.)
5 Di-n-butyl phthalate
Peel adhesion (N/2.5 cm)
Diethylhexyl phthalate
4
3 Removable
Di-n-octyl phthalate
2
1
10 20 30
Plasticizer concentration (wt %)
FIGURE 7.17 Influence of plasticizer concentration on peel adhesion. (From Czech, Z., J. Appl.
Polym. Sci., 97, 886–892, 2005. With permission.)
One effective PSA bulk property control methods is the addition of the tackifier into
the elastomer. Adhesion properties such as peel resistance and tack can be easily con-
trolled by the addition of tackifier into a styrenic block copolymer. To investigate the
effect of tackifiers on peel resistance, PSAs were blended with various tackifiers such as
GA-100 (rosin ester), Hikorez A-1100S (aliphatic hydrocarbon), Regalite R-125 (hydro-
genated aromatic hydrocarbon), Quintone U-185 (Modified C5), and Sukorez SU-100
(hydrogenated dicyclopentadiene). Figure 7.18 illustrates the peel resistance of the SIS/
tackifier blends as a function of tackifier contents [32]. The PSAs blended with Sukorez
SU100, Hikorez A1100S, and Quintone U-185 exhibit maximum peel resistance at
60 wt % of tackifier content, whereas other PSAs with GA-100 and Regalite R-125 exhibit
a peak at 40–50 wt % of tackifier content. The GA-100 and R-125 have a higher soft-
ening point and Tg than the other tackifiers. Thus, the maximum peel strength shifts
with lower tackifier content, depending on the softening point. Similar behavior is
illustrated in Figure 7.19. The hot-melt PSAs (HMPSAs) were prepared by SIS copo-
lymer and tackifier with various Tg: SU-90 (Tg, 39.1°C), SU-110 (Tg, 64.3°C), SU-130
(Tg, 75.4°C). As noted, the peak location of peel resistance shifts to lower tackifier con-
tent with increasing tackifier Tg [33].
Considering the fact that the PSAs are a mixture of elastomeric polymers and low-
molecular-weight tackifier resins, it is important to investigate the degree of miscibility
between the components, because the phase structures are governed by miscibility; that
is, when the components are miscible with each other, the blends must be in a uniform
one-phase structure, but when they are immiscible with each other, phase separation
occurs. The physical properties and the practical performance of the PSA are dependent
on the phase structure of the materials.
15000
12500
7500
5000
2500
0
40 50 60 70
(a) Wt % of tackifier
15000
12500
Peel strength (g/25 mm)
10000
7500
5000
2500
0
40 50 60 70
(b) Wt % of tackifier
FIGURE 7.18 Peel strength of SIS/tackifier blends on SUS substrate (peeling rate, 300 mm/min):
(a) Kraton D1107 blends and (b) Vector 4111 blends (○, Hikorez A 1100S; △, Regalite R 125; ▽,
Quintone U 185; ◇, Sukorez SU 100; □, GA-100). (From Kim, D.J., Kim, H.-J., and Yoon, G.H., Int.
J. Adhesion Adhesives, 25, 288–295, 2005. With permission.)
In the acrylic PSA/tackifier system, Kim and Mizumachi [34] studied systemically
the effect of miscibility on adhesion properties. The peel resistance of PSAs is low at a
low separation rate, and it gradually increases and decreases again as the separation rate
is increased. In the case of a miscible system, incorporation of a tackifier resin results
in modification of the bulk properties of the blend; therefore, the peel resistance mas-
ter curves shift along the rate-axis, whereas the tackifier content changes, as illustrated
in Figure 7.20. On the other hand, in the case of a system where the components are
15000
Vector + SU-90
Vector + SU-110
Vector + SU-130
10000
Peel strength (g)
5000
0
30 40 50 60 70
Tackifier content (wt %)
FIGURE 7.19 Peel strength of vector/tackifier as a function of tackifier content. (From Lim,
D.H., Do, H.S., and Kim, H.-J., PSA J. Appl. Polym. Sci., 102, 2839–2846, 2006. With permission.)
3000
Tackifier content (%)
: 0
: 10
: 20
: 30
Peel strength (gf/cm)
2000 : 40
1000
0
10−5 10−3 10−1 101
Va T (cm/s)
FIGURE 7.20 Peel master curve dependence on peeling rate in miscible blends. (From Kim, H.-
J. and Mizumachi, H., Advances in Pressure Sensitive Adhesive Technology-3, D. Satas, Ed., Satas &
Associates, Warwick, RI, 1998. With permission.)
1500 : 0
: 10
: 20
: 30
: 40
Peel strength (gf/cm)
1000
500
FIGURE 7.21 Peel master curve dependence on peeling rate in immiscible blends. (From Kim,
H.-J. and Mizumachi, H., Advances in Pressure Sensitive Adhesive Technology-3, D. Satas, Ed.,
Satas & Associates, Warwick, RI, 1998. With permission.)
not miscible, physical properties of the two phases are not modified, even if the blend
ratio is varied, and the master curve does not shift along the rate-axis, as illustrated
in Figure 7.21, because the peel resistance depends mainly upon the properties of the
matrix phase. The peak height of the master curve in peel resistance decreases as the
amount of the dispersed phase increases.
Fujita et al. [35] investigated the effects of miscibility on the peel resistance of natu-
ral rubber-based PSAs. In the case of miscible PSAs, the peak position of the pulling
rate-peel resistance curve shifted to a lower velocity as the tackifier content increased.
Immiscible PSA blends had lower peel resistances than miscible blends and did not
exhibit any apparent peak shifts.
Viscoelastic properties play a vital role in the study of adhesion behavior, including
the application of PSAs, diff usion through interfaces, and internal stress (see Chapter 4).
Viscoelastic properties deal with the problem of the action of mechanical forces on the
adhesion system, depending on the velocity and loading frequency.
Figure 7.22 illustrates the viscoelastic spectrum of typical PSAs [36]. Peeling is a
high-rate process and takes place at a speed between 100 and 1000 rad/s. Peel rate can
Tack Shear
zone zone
Log G′, Log G′′ (Pa) 7
6 Peel
zone
G′′
G′
4
−3 −2 −1 0 1 2 3 4
Log frequency (rad/s)
FIGURE 7.22 Viscoelastic spectrum of a typical PSA (G′, storage modulus; G″, loss modulus).
(From Satas, D., Handbook of Pressure Sensitive Adhesive Technology and Applications, D. Satas,
Ed., Satas & Associates, Warwick, RI, 2002. With permission.)
cover both the plateau and the transition zones. If the rate of peel is not too fast, then
entanglements will dominate and control the peel strength of PSAs. If peeling occurs at
higher rates, then small molecular structures, such as tackifying resins and short chain
pendant groups attached to the base elastomer, can contribute to the peel strength of
the PSA.
The characteristics in a master curve will be described in reference to the character-
istics of natural rubber/tackifier blends, as illustrated in Figure 7.23 [37]. In the master
curve, three regions of steady peeling and one region of oscillatory (stick–slip) peeling
may be distinguished. At high temperatures or very low peel rates, the adhesive experi-
ences predominantly viscous flow, which allows it to be drawn out into long fibrils in
the region where the tape curls away from the substrate. These eventually fail in tension,
leaving traces of the adhesive on both the tape and the rigid surface, which is known as
cohesive failure (see Figure 7.10).
At higher peel rate or less elevated temperatures, the adhesive becomes detached from
one of the solid surfaces, usually remaining attached to the flexible tape, either before
any fibrils have formed or at least before they are well developed. This is characteristic of
an interfacial or adhesion failure (see Figure 7.10). Both of these detachment modes are
stable and if the peeling rate is kept constant, the peel force will be also. At either high
peel rates or low temperatures, the adhesive may undergo transformation into its glassy
or brittle form, which now allows deadhesion to proceed by the advance of a crack. Th is
may be either stable or unstable and may run either cohesively through the adhesive
layer or along the adhesive/substrate interface.
20 246
B
240
B- 258
10
C 246
A 240
C
0
−8 −6 −4 −2 0 2 4
Log Ra ′T (m/s)
FIGURE 7.23 Master curve of peel force against pulling rate at 296 K for the system 6:4 NR:
Piccolyte S115, showing superposition of experimental data (A, cohesive failure; B, adhesive fail-
ure between adhesive and glass (substrate); C, adhesive failure between adhesive and baking; B-C,
stick–slip). (From Satas, D., Handbook of Pressure Sensitive Adhesive Technology and Applications,
D. Satas, Ed., Satas & Associates, Warwick, RI, 2002. With permission.)
Figure 7.24 illustrates the rheologic behavior at different temperatures and the cor-
responding peeling curves [38]. The terminal domain of relaxation corresponds to an
interfacial fracture with a crack localized between the adhesive and the rigid substrate
(Figure 7.24a). In the intermediate domain (transition region), one can observe stick–slip
behavior (Figure 7.24b). Finally, at very low temperatures, the crack is localized between
the adhesive and the flexible substrate with a very low peel force (Figure 7.24c). One
can conclude that the propagation of the crack, as well as the value of the peeling force,
depends largely on the rheological behavior of the adhesive, which is linked to the tem-
perature and the peeling rate. In the case of EVA-based PSAs, Gibert et al. [39] demon-
strated that the cohesive-to-interfacial transition observed on a peeling curve appeared
at the same time as the liquid-like to solid-like transition induced by crystallization of
EVA (Figure 7.25) [39]. Figure 7.25 provides a schematic diagram of the relationship
between viscoelastic and peel properties (see Chapter 4).
The failure mode does not always follow the four main types: cohesive, adhesive,
stick–slip, and glassy failure. The peeling behavior illustrated in Figure 7.26 demon-
strates differences. Figure 7.26a exhibits all four fracture domains already described
(cohesive, interfacial 1, stick–slip, and interfacial 2), whereas the interfacial 1 fracture
domain has disappeared in Figure 7.26b. Cohesive fracture can be shifted directly to
a stick–slip behavior as the peeling rate increases. Aubrey and Sherrif [37] derived the
same conclusion with another formulation. The addition of resin to the polymer not
only diminishes the width of the elastic plateau region but also decreases the plateau
9
G′
8 G′′
(a) (a[i])
8 T = 20°C 5 G ′ T = 20°C
6
5 4
4
3 3
2
1
0 2
0 2 4 6 8 10 −1 0 1 2
Time (s) Log (ω/[rad/s])
(b) (b[i])
G ′ T = −15°C
8 T = −30°C 6
7 G′′
Log (G′ and G′′/[Pa])
6
Peel force (N)
5 5
4
3
4
2
1
0 3
0 1 2 3 4 5 6 −1 0 1 2
Time (s) Log (ω/[rad/s])
(c) (c[i])
8 T = −30°C 9 G′ T = 20°C
Log (G ′ and G′′/[Pa])
7 G ′′
Peel force (N)
6
8
5
4
3 7
2
1
0 6
0 2 4 6 8 10 −1 0 1 2
Time (s) Log (ω/[rad/s])
FIGURE 7.24 G′, G″ versus frequency for SIS + SI blend and peeling curves measured on the
tensile machine at various temperatures: (a) terminal domain (rubbery plateau); (a[i]) adhesive
fracture; (b) transition region; (b[i]) stick–slip; (c) glassy domain; and (c[i]) glassy fracture. (From
Marin, G. and Derail, C., J. Adhesion, 82, 469–485, 2006. With permission.)
F
F
Crack
Crack
F
Peel force
Crack
Glassy
Cohesive Interfacial Stick-
fracture
fracture fracture slip
Peel rate
G′
G ′′
Terminal
Shear modulus
zone
Glass Glassy
Rubbery
transition state
state
Frequency
FIGURE 7.25 Schematic diagram of the relationship between viscoelastic and peel properties.
(From Gibert, F.X., Allal, A., Marin, G., and Derail, C., J. Adhesion Sci. Technol., 13(9), 1029–1044,
1999. With permission.)
modulus value. The gradual disappearance of the type 1 interfacial fracture is probably
due to both effects. Changing the resin content not only changes the plateau modulus
value and extent, but also changes the Tg, so all these combined effects do not lead to a
simple relationship.
Interfacial 1 Interfacial 2
160
140
Cohesive Stick-slip
120
PF TO/ T (N)
100
80
60
40
20
0
0 1 2 3 4 5 6 7 8 9
(a) log (a T·V) (mm/min)
60
50
PF TO/ T (N)
40
30
20
10
0
1 2 3 4 5 6 7 8 9
(b) log (a T·V) (mm/min)
FIGURE 7.26 Peel force as a function of reduced peel rate: (a) PB:tackifier = 170:30; (b) PB:
tackifier = 270:30. (From Aubrey, D.W. and Sherrif, M., J. Polym. Sci. Polym. Chem. Ed., 18, 2597–
2608, 1980. With permission.)
adherends, indicating that both the rate of the bonding process and the failure criterion
concerning the interfacial failure are closely related to the critical surface tension of the
adherend. Performance of the natural rubber-based PSA was more complicated than
that of the acrylic PSA.
Kim et al. [32] reported the peel strength of the SIS/tackifier blends at various peel-
ing rates with many substrates having different surface tension values (Figure 7.28).
Table 7.4 presents the surface tension of various substrates. High peel strength was
PET = 10.3
10 = Cohesive 44.48
failure
5 22.24
0 0
8 9 10 11 12 13
Solubility parameter (δ) of adhesive
observed for stainless steel (SUS 304) and glass, medium peel strength for Bakelite,
polyvinyl chloride (PVC), and polypropylene (PP), and low peel strength for polyethyl-
ene (PE) and Teflon. A similar classification was obtained using surface tension as the
distinguishing factor. Although PE exhibits a surface tension similar to that of medium-
peel substrates, it has low peel strength due to the different failure mode. Interfacial fail-
ure was observed for PE, whereas cohesive failure was observed for other medium-peel
substrates. This may be due to differences in the characteristics of the substrates.
In the peel tests performed using PP substrates, cohesive failure occurred at various
test temperatures. However, in the peel test performed using the PE substrate, adhesive
failure occurred for the Kraton D1107 blends, but stick–slip failure was observed for Vec-
tor 4111 blends. After aging at 100°C, cohesive failure occurred in all blends except for the
Regalite R-125 blends because PSAs transfer to the substrates during the aging process.
A typical stress–strain curve of the Kraton D1107/Sukorez SU-100 blends with the
PE substrate is illustrated in Figure 7.29. In the Kraton D1107/Sukorez SU-100 (40/60)
blend, a stick–slip type of stress–strain curve was observed (Figure 7.29a), whereas
cohesive failure (Figure 7.29b) was observed in the Vector 4111/SU-100 (40/60) blend.
Although the curves in Figure 7.29a correspond to stick–slip failure, the adhesive was
stripped cleanly from the substrate, leaving no visually residue. Therefore, the surface
was clear after the test because of the high cohesion of the SIS-based, HMPSAs with a
high SIS content.
Not only is the peel force decreased for a higher energy surface, but also the positions
of the transitions from cohesive to adhesion and from adhesion to stick–slip failure are
changed. That is, the activation of the stiffening behavior of a PSA depends on the sub-
strate to which the adhesive is bonded. Marin and Derail [38] reported similar effects of
substrate nature on peel strength and mechanisms of fracture (see Chapter 4).
15000
12500
7500
5000
2500
0
0 200 400 600
(a) Rate (mm/min)
15000
12500
Peel strength (g/25 mm)
10000
7500
5000
2500
0
0 200 400 600
(b) Rate (mm/min)
FIGURE 7.28 Peel strength of SIS/Hikorez A 1100s (40/60) blends as a function of peeling rate:
(a) Kraton blends and (b) Vector blends (□, SUS; ○, PE; △, PP; ▽, PVC; ◇, Bakelite; ◁, Teflon;
and ▷, glass). (From Kim, D.J., Kim, H.-J., and Yoon, G.H., Int. J. Adhesion Adhesives, 25, 288–295,
2005. With permission.)
Note: SS, stainless steel; PE, polyethylene; PP, polypropylene; PVC, polyvinylchloride.
a Determined by contact angle measurement.
b Not available.
Source: Czech, Z., J. Appl. Polym. Sci., 97, 886–892, 2005. With permission.
6000
Peel strength (g/25 mm)
4000
2000
0
0 20 40 60 80
(a) Distance (mm)
6000
Peel strength (g/25 mm)
4000
2000
0
0 20 40 60 80
(b) Distance (mm)
7.5 Conclusion
In the evaluation of PSA performance, the measurement of peel resistance is one of the
most important characteristics because various factors, such as properties of backing
materials, surface of the adherend, peeling speed, and test temperature, affect the peel
strength. Failure mode detection and drawing up the master curves for peel resistance
provide a great deal of information. Therefore, to gain insight into the peeling mecha-
nism it is helpful to design PSAs and evaluate their performance.
References
1. ASTM D 907.
2. Czech, Z., 2005. Synthesis of removable and repositionable water-borne pressure-
sensitive adhesive acrylics. J. Appl. Polym. Sci. 97:886–892.
3. Chen W. T. and T. F. Flavin, 1972. Mechanics of fi lm adhesion: Elastic and elastic–
plastic behavior, IBM J. Res. Dev. 16(3):203–213.
4. Kaelble, D. H., 1959. Theory and analysis of peel adhesion: Mechanisms and
mechanics, Trans. Soc. Rheol. 3:161–180.
5. Kaelble, D. H., 1960. Theory and analysis of peel adhesion: Bond stress and distri-
butions, Trans. Soc. Rheol. 4:45–73.
6. Williams J. A. and J. J. Kauzlarich, 2004. Peeling shear and cleavage failure due to
tape restrain, J. Adhesion, 80:433–358.
7. Williams J. A. and J. J. Kauzlarich, 2007. Application of the bulk properties of an
acrylic pressure sensitive adhesive to peeling, J. Adhesion Technol. 21(7):515–529.
8. Gent A. N. and E. A. Meinecke, 1970. Compression, bending, and shear of bonded
rubber blocks, Polym. Eng. Sci. 10(1):48–53.
9. Satas, D., 2002. Peel. In Handbook of Pressure Sensitive Adhesive Technology and
Applications, ed. D. Satas, Satas & Associates, Warwick, RI, pp. 62–86.
10. Horgnies, M., E. Darque-Ceretti, and E. Felder, 2007. Relationship between the
fracture energy and the mechanical behavior of pressure-sensitive adhesives, Int.
J. Adhesion Adhesives 27:661–668.
11. Kaelble, D. H., 1965. Peel adhesion: Micro-fracture mechanics of interfacial
unbonding of polymers, Trans. Soc. Rheol. 9(2):135–163.
12. Kinloch, A. J., C. C. Lau, and G. J. Williams, 1994. The peeling of flexible lami-
nates. Int. J. Fracture 94:79–88.
13. Muny, R. P., 2002. Testing pressure sensitive adhesives. In Handbook of Pressure
Sensitive Adhesive Technology and Applications, ed. D. Satas, Satas & Associates,
Warwick, RÌ, pp. 139–152.
14. Kawashita, L. F., D. R. Moore, and J. G. Williams, 2004. The development of a
mandrel peel test for the measurement of adhesive fracture toughness of epoxy-
metal laminates, J. Adhesion 80:147–167.
15. Urahama, Y., Y. Tokunaga, and Y. Tanaka, 1990. Morphology at peeling pressure
sensitive adhesives (II), J. Adhesion Soc. Jpn. 23(5):171–177.
16. ASTM D3330.
17. PSTC-1.
18. Pocius, A. V., 1997. Basic physico/chemical properties of polymers. In Adhesion
and Adhesives Technology, ed. A.V. Pocius, Hanser, Munich.
19. Derail, C., A. Allal, G. Marin, and Ph. Tordjeman, 1997. Relationship between
viscoelastic and peeling properties of model adhesive. Part1. Cohesive fracture,
J. Adhesion 61:123–157.
20. Aubrey, D. W., G. N. Welding, and T. Wong, 1969. Failure mechanisms in peeling
of pressure-sensitive adhesive tape, J. Appl. Polym. Sci. 13:2193–2207.
21. Do, H. S. and H.-J. Kim, 2006. UV-curable pressure-sensitive adhesives. In Pres-
sure-Sensitive Design and Formulation, Application Volume 2, ed. I. Benedek, VSP,
Leiden, Boston, pp. 251–290.
22. Auchter, G., O. Aydin, A. Zettl, and D. Satas, 2002. Acrylic adhesives. In Handbook
of Pressure Sensitive Adhesive Technology and Applications, ed. D. Satas, Satas &
Associates, Warwick, RI, pp. 445–514.
23. Mayer, A., T. Pith, G. Hu, and M. Lambla, 1995. Effect of the structure of latex par-
ticles on adhesion. Part II: Analogy between peel adhesion and rheological prop-
erties of acrylic copolymers, J. Polym. Sci. Part B: Polym. Phys. 33(12):1793–1801.
24. Mayer, A., T. Pith, G. Hu, and M. Lambla, 1995. Effect of the structure of latex
particles on adhesion. Part I: Synthesis and characterization of structured latex
particles of acrylic copolymers and their peel adhesion behavior, J. Polym. Sci.
Part B: Polym. Phys. 33(12):1781–1791.
25. Shen, H., J. Zhang, S. Liu, G. Liu, L. Zhang, and X. Qu, 2008. Effect of the chain-trans-
fer-agent content on the emulsion polymerization process and adhesive properties of
poly(n-butyl acrylate-co-acrylic acid) latexes, J. Appl. Polym. Sci. 107(3):1793–1802.
26. Rana P. K. and P. K. Sahoo, Synthesis and pressure sensitive adhesive performance
of poly (EHA-co-AA)/silicate nanocomposite used in transdermal drug delivery,
J. Appl. Polym. Sci. 106(6):3915–3921.
27. Gower M. D. and R. A. Shanks, 2004. The effect of varied monomer composition
on adhesive performance and peeling master curves for acrylic pressure-sensitive
adhesives, J. Appl. Polym. Sci. 93(6):2909–2917.
28. Gower M. D. and R. A. Shanks, 2004. The effect of chain transfer agent level on
adhesive performance and peel master-curves for acrylic pressure sensitive adhe-
sives, Macromol. Chem. Phys. 205(16):2139–2150.
29. Gower M. D. and R. A. Shanks, 2006. Acrylic acid level and adhesive performance
and peel master-curves of acrylic pressure-sensitive adhesives, J. Polym. Sci. Part B:
Polym. Phys. 44(8):1237–1252.
30. Joo, H. S., H. S. Do, Y. J. Park, and H. –J. Kim, 2007. Adhesion performance of
UV-cured semi-IPN structure acrylic pressure sensitive adhesives, J. Adhesion Sci.
Technol. 20(14):1573–1594.
31. Do, H. S., Y. J. Park, and H.-J. Kim, 2006. Preparation and adhesion performance
of UV-cross-linkable acrylic pressure sensitive adhesives, J. Adhesion Sci. Technol.
20(13):1529–1545.
32. Kim, D. J., H.-J. Kim, and G. H. Yoon, 2005. Effect of substrate and tackifier on
peel strength of SIS (styrene-isoprene-styrene)-based HMPSAs, Int. J. Adhesion
Adhesives 25:288–295.
33. Lim, D. H., H. S. Do, and H.-J. Kim, 2006. PSA performances and viscoelastic
properties of SIS-based PSA blends with H-DCPD tackifier, J. Appl. Polym. Sci.
102:2839–2846.
34. Kim, H.-J. and H. Mizumachi, 1998. Miscibility of acrylic pressure sensitive
adhesives. In Advances in Pressure Sensitive Adhesive Technology-3, ed. D. Satas,
Satas & Associates, Warwick, RI, pp. 77–128.
8.1 Introduction
From the rheological viewpoint, pressure-sensitive adhesives (PSAs) can be described
in general as viscoelastic materials with a specific complex of rheological properties.
Under shear load they demonstrate creep behavior that can lead to failure of the adhe-
sive joint. Shear resistance characterizes a PSA’s ability to resist shearing forces. Shear
resistance is also sometimes termed holding power [1] (see also Applications of Pressure-
Sensitive Products, Chapter 8).
Shear resistance is one of the most significant characteristics of PSAs. The requirements
for shear resistance come from the application of PSAs (see also Applications of Pressure-
Sensitive Products, Chapter 4). Shear resistance is very important, for example, for packag-
ing tapes and labels, upholstery, films for floor advertising, automobile industry, etc. In
some applications PSAs are exposed to shearing forces at elevated temperatures (see also
Applications of Pressure-Sensitive Products, Chapter 4) for relatively short times, whereas
in other applications the material must withstand shear load for longer periods of time
8-1
3 2
1 6
FIGURE 8.2 Schematic design of the apparatus for a modified static shear test. (1) steady steel plate,
(2) movable plate or backing, (3) adhesive layer, (4) load, (5) block, and (6) temperature chamber.
different authors using the static shear method can hardly be directly compared due to
the differences in testing parameters and thickness of the specimens.
Although this method simulates the real behavior of PSA tapes under shear loads, it has
a drawback of determining only the “end point” (holding time), thus providing very lim-
ited information on the mechanism and kinetics of failure. Several authors [8–11] used a
modified version of this test based on a device that is more sophisticated from a rheological
viewpoint, known as a shear plastometer. The schematic design of such a device is illus-
trated in Figure 8.2. The constant shear load is applied via a block to the sample (PSA tape
with backing) adhered to the substrate. The backing is sometimes reinforced by a second
steel plate [8] to prevent the sample from bending. The sample can be placed in a heated
chamber, permitting measurements at different temperatures. This equipment allows the
displacement of the one of plates to be registered as a function of time at constant load
during the experiment with good resolution (i.e., it is a creep measurement as well).
As mentioned previously, devices of this kind have often been used by rheologists for
the study of creep behavior and are known as shear (parallel plate, sandwich type, etc.)
plastometers [12,13] (see also Chapter 4). In fact, the shear test is a creep test carried to
high deformation [14]. Later in this chapter we will discuss a rheological description of
the creep test in the shear plastometer.
Zosel [8] reports 20% standard deviation of the experimental results obtained for a
large number of similar specimens tested on such a device, explained by deviations in
sample preparation. Imprecision of the tape’s geometric parameters, especially thickness
and incomplete stress relaxation, can be mentioned among the possible reasons for data
scattering.
The main drawback of the static shear method is that the experiments are quite time-
consuming. Therefore, a “dynamic” test has been proposed (see also Applications of Pressure-
Sensitive Products, Chapter 8). A sample similar to that used for the static test is deformed
and separated from the substrate, not at constant load but at constant strain rate [8,15]. The
experiment can be performed with a tensile tester. The cross-head speed depends on the
intended application of the PSA. The maximum force determined during the experiment
can be regarded as a measure of shear resistance. The stress–strain curves obtained using
this method can provide useful information about the failure of the adhesive joint.
In some applications (e.g., packaging, insulation, mounting, or splicing tapes) it is
necessary to know the shear resistance at elevated temperatures and the maximum per-
missible temperature. In the shear adhesion failure temperature (SAFT) method [1,16,17]
a sample similar to one described above is loaded with a static weight while the temperature
is ramped at rate of 0.4 C/min until the specimen fails. The testing conditions are generally
the same as in the static shear test (except for the temperature) and the experiment can be
carried out in the same machine if it is equipped with a temperature-controlled chamber. In
this method the temperature of the failure of the adhesive joint is considered as a character-
istic of shear resistance (see also Applications of Pressure-Sensitive Products, Chapter 8).
Reliability of the results obtained in different shear tests depends on many factors.
The most common errors in measuring shear resistance include nonuniform thick-
ness of the adhesive layer, the presence of nonrelaxed stresses, the substrate surface is
not clean, and the direction of the applied force is not parallel to the interface. Careful
inspection of the surface of the adhesive joint after separation to determine the failure
mode (cohesive, adhesive, or mixed adhesive–cohesive) can provide valuable information
that can be useful for improving performance of the tested PSA.
F = const.
(creep)
C F=0
(elastic recovery)
B
D
B′
Strain
Unrecoverable
A
strain
O
Time
FIGURE 8.3 Typical plot of plate displacement versus time in a shear plastometer (small
deformations).
F
(8.1)
A
where F is the force applied to the adhesive tape and A is the overlapping area.
Shear rate is
(8.2)
d
where v is the current linear speed of the adhesive tape and d is the thickness of the
adhesive layer. The overlapping area decreases with time during the test in accordance
with Equation 8.3,
l
A A0 1 (8.3)
l0
where ∆l is the shift of the adhesive tape from the initial position and l 0 is the initial
length of the overlapping area. It is, therefore, clear that shear stress grows with time.
Taking into account that
dl
(8.4)
dt
it is possible to solve this system and prolong the analysis to the end point when the
adhesive tape slides off the substrate. Th is was done by Zosel [8], who tried to predict
the holding time from the flow curves of the adhesives. Unfortunately, such a model
assumes sliding off of the tape governed by the viscous flow of adhesive as the only
mechanism of failure and completely neglects the elasticity of the adhesive. All these
equations are valid only when there is no slippage between the tested material and
the substrate. It is not surprising, therefore, that this model was able to predict the
holding time more or less correctly only for rather low-viscosity non-cross-linked
adhesives.
By extrapolating the steady segment to the strain axis it is possible to separate elastic
and plastic deformations: segment AB′ in Figure 8.3 represents the elastic portion of the
total deformation developed in segment AB.
If the load is released, then elastic recovery starts with a fast stage (segment CD), fol-
lowed by slower recovery (DE). The plastic deformation cannot be recovered (unrecov-
erable strain). Of course, everything noted regarding plastic deformations is applicable
only to noncross-linked polymers.
All stages of the creep process in polymers can be qualitatively described using Burgers’s
model [19,15], which consists of four elements, connected as illustrated in Figure 8.4,
where springs elastic constants (moduli) are G 0 and G1 and dashpot viscosities are η 0
and η1. This model combines Maxwell (index 0) and Kelvin–Voigt fluids (index 1). The
total stress in this model is equal to the stress of each element:
0 1 (8.5)
η0
G1
η1
G0
0
1 (8.6)
and, consequently,
d
d
0 d
1
(8.7)
dt dt dt
An expression for stress can be derived from Equations 8.5 through 8.7 as follows:
d
1 d
G1
1 1 1 (8.8)
dt G0 dt 0
0 0
t
(t ) 1 e 1 0 t (8.9)
G0 G1 0
In Equation 8.10, t > t1, and t1 is the moment of load release. The plastic deformation
τ0
developed in the dumper of the Maxwell element during creep stage, __ η 0 t1, is
unrecoverable.
Although Burgers’s model describes qualitatively the whole creep–recovery cycle,
several authors proposed some modifications of this model to enable better compliance
with experimental results obtained in shear tests of PSAs. Geiss and Brockmann [15]
investigated the creep behavior of SIS-based PSA in cycling creep–recovery tests. The
slope of the strain curve increases from cycle to cycle as if the viscosity of the PSA is
decreasing. To describe this feature, a time-dependent nonlinear factor was introduced
to the reaction of the G1 spring. Kano et al. [10] added a combination of friction and
damper elements to extend the application of the model to adhesive failure. The friction
element simulates slippage between the adhesive tape and the substrate.
The overlapping area diminishes with time, thus increasing the shear stress (Equa-
tions 8.1 and 8.3). Therefore, simple viscoelastic models derived for constant
stress conditions become inapplicable for creep description.
The shear stress field becomes nonuniform. For example, the material at the edge
of the moving plate (point E in Figure 8.2) is exposed not only to shear stresses
but also to elongation.
Cross-linked materials with limited ability for deformation would lose contact
with either of the plates at small strain.
The large deformations are, however, inherent for real shear tests. Figure 8.5 illustrates
the typical dependence of the displacement of the adhesive tape on time. The diminish-
ing overlapping area noted above has several consequences. First, if the viscosity does not
depend on shear stress (Newtonian liquid), this leads to acceleration of the plate move-
ment (see Figure 8.5). For many polymers, however, viscosity decreases with increased
shear stress (viscosity anomaly), which means, in this case, even more prominent accel-
eration. Moreover, as determined by Vinogradov et al. [23], high shear stresses can be
easily reached for very viscous systems such as fi lled rubbers, even at small strains, which
can induce the forced transition of the polymer melt to a rubbery state, resulting in the
loss of adhesion to the substrate and transition to the adhesive mode of joint failure. The
elasticity of the adhesive plays a very important role in this process.
It can be concluded from information noted previously that the shear resistance of
non-cross-linked adhesives, as long as they remain adherent to the substrate, depends
mainly on their viscosity and elasticity and, consequently, on the factors influencing the
rheological parameters of the adhesive. Elasticity becomes more important at the fi nal
stages of the creep test.
As for the cross-linked adhesives, their limited ability to deform in the shear field
depends mainly on parameters of cross-linking (cross-linking density and their nature).
Plate displacement
Time tc
FIGURE 8.5 Typical plot of plate displacement versus time in a shear tester (large deforma-
tions). The rectangular area in the left lower corner represents the region of small deformations
illustrated in Figure 8.3.
The thickness of the specimen is considerably smaller than its diameter; it is larger
than the plates and the area under compression is constant. The excessive
amount of material can flow freely from the gap.
The thickness of the specimen is considerably smaller than its diameter; it is smaller
than the plates and the volume of the specimen under compression is constant.
A h
FIGURE 8.6 Scheme of a squeeze tester. (A) movable plate, (B) steady plate, and (C) specimen.
We can also consider the situation of a constant applied compressive force and constant
squeezing rate. We will, however, limit our discussion in this section to the case of a
constant area under compression and constant compressive force because this situa-
tion is closer to that observed in practice. Figure 8.6 illustrates the scheme of a squeeze
tester. The upper (movable) plate has diameter d. The kinetics of plate movement and the
factors influencing it will be the matters of interest. Several assumptions must be made
to derive the corresponding equations.
For a Newtonian liquid the kinetics of the distance change can be presented in the form
of Stefan’s equation:
1 1 4Ft
(8.11)
h2 h02 3R 4
In this equation, h is the current distance between plates, h 0 is the initial distance, F is
the applied force, t is time, R is the radius of the smaller plate, and η is the viscosity of
the tested liquid. As seen from Equation 8.11, a reduction in the distance between plates
slows down in time but never stops completely. In fact, however, even for low-viscosity
liquids complete squeezing is impossible due to capillary forces.
For a power-law fluid that can be characterized by the expression
k
n (8.12)
(where k and n are constants), the change in distance between the plates in time is deter-
mined by Scott’s equation:
2kRn3 2n 1 (dh/dt )n
n
F (8.13)
n3 n h2n1
If n = 1 and k = η, Equation 8.13 becomes Equation 8.11. The field of shear stresses in
this operating unit is nonuniform. Expressions of shear stress and shear rate at the rim
of the smaller plate are as follows:
n 3 hF
R (8.14)
2 R3
2n 1 ( h )R (8.15)
R
n h2
where ḣ = dh∕dt. Unlike in the static shear test, squeezing at constant force proceeds
with decreasing shear stress, and for systems with yield behavior even if the initial shear
stress is greater than the yield stress, sooner or later (depending on viscosity) the shear
stress will approach the yield stress value. Th is approaching is accompanied by further
slowing of the plate’s movement and there is a definite minimum distance at every given
value of the applied force that corresponds to shear stress equal to the yield stress.
As mentioned previously, Equations 8.11 through 8.15 were derived by neglecting the
elasticity of the tested liquid. Several attempts to take into account the elastic proper-
ties of the material resulted in very sophisticated models [25]. Laun [26], based on the
results of squeezing polyethylene (PE) melts, proposed an empirical criterion to elimi-
nate experimental points with significant elasticity:
t
40 (8.16)
8.5.2 Load
Kano et al. [10] demonstrated that the holding time decreases significantly with
increased dead load in the static shear test for poly(butyl acrylate) and its blends with
poly(vinylidene fluoride-co-hexafluoro acetone). Similar results were obtained by Geiss
and Brockmann [15] for a styrene–isoprene–styrene (SIS)-based hot-melt adhesive and
a commercial high-performance adhesive tape.
If a material possesses yield stress and the applied load does not exceed its level, then
the creep is very limited [22].
8.5.3 Temperature
Because an increase in temperature usually leads to a drop in viscosity, it is under-
standable that temperature reduces shear resistance and decreases drastically the time
to failure [14,15]. Kim [28] investigated the temperature dependence of the holding
power in adhesive joints of an SIS-based hot-melt adhesive with different substrates:
stainless steel, glass, Bakelite, polyvinyl chloride, PE, polypropylene, and Teflon. The
holding power decreases with temperature for all adhesive joints though the slope
of this dependence and absolute values of the holding power were different for each
adhesive–substrate pair.
strength. A similar effect was also observed by Demarteau and Loutz [32]. Indeed, it is well
known from polymer synthesis that AA is used as a comonomer to increase the cohesion
of various acrylic formulations.
Park et al. [33] investigated the change in holding power with composition in quater-
nary copolymers of 2-ethylhexyl acrylate (2-EHA), BA, ethyl acrylate (EA), and vinyl
acetate (VA). Increased BA and 2-EHA content leads to a significant decrease in holding
time, probably because of the higher flexibility and lower intermolecular interaction
introduced by these monomer sequences.
Tobing and Klein [29] measured the shear resistance of poly(BA-stat-AA) and
poly(2-EHA-stat-AA) copolymers. In all cases, poly(BA-stat-AA) demonstrates higher
holding power than (2-EHA-stat-AA), regardless of the method of polymerization. This
effect was attributed to a higher number of entanglements in poly(BA-stat-AA) per unit
weight or volume.
Gower and Shanks [34], however, determined that in BA/2-EHA/methyl methacry-
late (MMA)/AA copolymers the holding power increases with 2-EHA concentration.
It should be noted that in this case the shear properties were compared not at equal gel
content, and the result can be attributed to the increased gel content in formulations
with 2-EHA.
References
1. Lim, D.H., H.J. Kim. 2006. General performance of pressure-sensitive adhesives. In
Pressure-Sensitive Design, Theoretical Aspects. Volume 1, ed. I. Benedek, Leiden–
Boston, VSP, pp. 291–317.
2. Eveloy, V., P. Rodgers, M.G. Pecht. 2004. Reliability of pressure-sensitive adhesive
tapes for heat sink attachment in air-cooled electronic assemblies. IEEE Trans.
Device Mater. Reliability 4:650–657.
3. ASTM D3654/D3654M-06. Standard Test Methods for Shear Adhesion of Pressure-
Sensitive Tapes.
4. PSTC-7. Holding Power of Pressure-Sensitive Tape (revised ed.), 1985.
5. FINAT Test Method 8 (FTM8). 1985. Resistance to shear from a standard surface,
1985.
6. AFERA Test Method 4012. Measurement of shear adhesion, 1979.
7. Brockmann, W., P.L. Geiß. 1997. Creep Performance of Mounting Tapes Based on
Hot Melt Pressure Sensitive Adhesives. Proc. 1997 TAPPI Hot Melt Symposium,
TAPPI Press, Atlanta, pp. 153–158.
8. Zosel, A. 1994. Shear strength of pressure sensitive adhesives and its correlation to
mechanical properties. J. Adhesion 44:1–16.
9. Miyagi, Z., K. Yamamoto. 1987. Viscoelastic analysis of shear adhesion test for
pressure-sensitive adhesive tape. J. Adhesion 21:243–250.
10. Kano, Y., S. Akiyama, Z. Miyagi. 1998. Analysis of holding power in the blends of
poly(butyl acrylate) with poly(vinylidene fluoride-co-hexfluoro acetone). J. Appl.
Polym. Sci. 68:727–738.
11. Fujita, M., A. Takemura, H. Ono, M. Kajiyama, S. Hayashi, H. Mizumachi. 2000.
Effects of miscibility and viscoelasticity on shear creep resistance of natural-
rubber-based pressure sensitive adhesives. J. Appl. Polym. Sci. 75:1535–1545.
12. Malkin, A.Ya., A.A. Askadsky, V.V. Kovriga, A.E. Chalykh. 1983. Experimental
Methods of Polymer Physics. Mir Publishers, Moscow.
13. Dealy, J.M. 1982. Rheometers for Molten Plastics. A Practical Guide to Testing and
Property Measurement. Van Nostrand Reinhold Company.
14. Kim, H.-J., H. Mizumachi. 1995. Miscibility and shear creep resistance of acrylic
pressure-sensitive adhesives: acrylic copolymer and tackifier resin systems. J. Appl.
Polym. Sci. 58:1891–1899.
15. Geiss, P.L., W. Brockmann. 1997. Creep resistance of pressure sensitive mounting
tapes. J. Adhesion 63:253–263.
16. ASTM D4498-00. Standard Test Method for Heat-Fail Temperature in Shear of
Hot Melt Adhesives.
17. Do, H.-S., Y.-J. Park, H.-J. Kim. 2006. Preparation and adhesion performance of
UV-crosslinkable acrylic pressure sensitive adhesives. J. Adhesion Sci. Technol.
20:1529–1545.
18. Satas, D. 1989. Handbook of Pressure Sensitive Adhesive Technology. Van Nostrand
Reinhold, New York.
19. Burgers, J.M. 1935. Mechanical considerations—model systems—phenomenological
theories of relaxation and of viscosity. In: First Report on Viscosity and Plasticity,
ed. J.M. Burgers, New York, Nordemann Publishing Company.
20. Schramm, G. 1994. A Practical Approach to Rheology and Rheometry. HAAKE
GmbH, Karlsruhe, Germany.
21. Barnes, H.A. 1999. The yield stress–a review or ‘παντα ρει’–everything flows?
J. Non-Newtonian Fluid Mech. 81:133–178.
22. Kulichikhin, V., S. Antonov, V. Makarova, A. Semakov, A. Tereshin, P. Singh. 2006.
Novel hydrocolloid formulations based on nanocomposites concept. In Pressure-
Sensitive Design, Theoretical Aspects. Volume 1, ed. I. Benedek, Leiden–Boston,
VSP, pp. 351–401.
23. Vinogradov, G.V., A.I. Elkin, S.E. Sosin. 1978. Fracture of uncured linear flexible-
chain polymers of narrow molecular mass distribution in triaxial stress (behaviour
of elastomers as adhesives). Polymer 19:1458–1464.
24. Oka, S. 1960. The principles of rheometry. In Rheology. Theory and Applications.
Volume 3, ed. F.R. Eirich, Academic Press, New York, pp. 17–82.
25. Leider, P.J. 1974. Squeezing flow between parallel discs II. Ind. Eng. Chem. Fundam.
13:342–349.
26. Laun, H.M. 1992. Rheometers towards complex flows: squeeze flow technique.
Macromol. Chem., Macromol. Symp. 56:55.
27. Feldstein, M.M., V.G. Kulichikhin, S.V. Kotomin, T.A. Borodulina, M.B. Novikov,
A. Roos, C. Creton. 2006. Rheology of poly(N-vinyl pyrrolidone)-poly(ethylene
glycol) adhesive blends under shear flow. J. Appl. Polym. Sci. 100:522–537.
28. Kim, D.-J. 1995. Ph. D. Thesis. Performance of SIS-based hot-melt pressure sensitive
adhesives. Seoul National University.
29. Tobing, S.D., A. Klein. 2001. Molecular parameters and their relation to the
adhesive performance of acrylic pressure-sensitive adhesives. J. Appl. Polym. Sci.
79:2230–2244.
30. Benedek, I. 2006. Principles of pressure-sensitive design and formulation. In
Pressure-Sensitive Design, Theoretical Aspects. Volume 1, ed. I. Benedek, Leiden–
Boston, VSP, pp. 131–289.
31. Krenceski, M.A., J.F. Johnson. 2004. Shear, tack and peel of polyisobutylene: effect
of molecular weight and molecular weight distribution. Polym. Eng. Sci. 29:36–43.
32. Demarteau, W., J.M. Loutz. 1996. Rheology of acrylic dispersions for pressure
sensitive adhesives. Progr. Organ. Coatings 27:33–44.
33. Park, W.-H., S.W. Kim, S.C. Shim, I.R. Jeon, K.H. Seo. 2003. Properties of pressure
sensitive adhesive made of acrylic quaternary copolymers and their blends with
vinyl chloride copolymers. Mat. Res. Innovat. 7:172–177.
34. Gower, M.D., R.A. Shanks. 2004. The effect of varied monomer composition on
adhesive performance and peeling master curves for acrylic pressure-sensitive
adhesives. J. Appl. Polym. Sci. 93:2909–2917.
35. Naruse, S., H.-J. Kim, T. Tsukatani, M. Kajiyama, A. Takemura, H. Mizumachi.
1994. Miscibility and PSA performance of acrylic copolymer and tackifier resin
systems. J. Adhesion 47:165–177.
36. Everaerts, A., K. Zieminski, L. Nguyen, J. Malmer. 2006. Cross-linking of hot-
melt-processible acrylic pressure-sensitive adhesives using acid/base interaction.
J. Adhesion 82:375–387.
37. Feldstein, M.M., G.W. Cleary, P. Singh. 2006. Pressure-sensitive adhesives of con-
trolled water-absorbing capacity. In Pressure-Sensitive Design and Formulation,
Application. Volume 2, ed. I. Benedek, Leiden–Boston, VSP, pp. 181–230.
38. Kulichikhin, V.G., L.A. Tsamalashvili, E.P. Plotnikova, A.A. Barannikov,
M.L. Kerber, H. Fischer. 2003. Rheological properties of liquid precursors of
polvnronvlene-clay nanocomnosites. Polym. Sci. (Russia). 45:564–572.
9.1 Introduction
Adhesive strength is one of the main characteristics of pressure-sensitive adhesives
(PSAs), but how long the adhesive joint can provide contact to the substrate over time
and under various external conditions (debonding force, temperature, and relative
humidity) is also important. Several methods exist to test the adhesive strength of
PSAs, including tack, peel, and shear tests (see Chapters 6 through 8 and Applications
of Pressure-Sensitive Products, Chapter 8). These are carried out under standard
conditions in accordance with ASTM D1876-01 (Method for Peel Resistance of
Adhesives (T-Peel Test) and probe tack test (according to ASTM D2979-01) or using
other methods. For long-term durability of PSAs, there is only one standard test—the
method for determining the durability of adhesive joints stressed in shear loading
(ASTM D2919-01 or PSTC-7). In most publications, discussions of durability proper-
ties include the analysis of behavior under various environmental conditions [1,2] (see
also Applications of Pressure-Sensitive Products, Chapter 8) or fatigue tests in cyclic
mode [3].
The static shear experiment is similar to a creep measurement with a constant shear
stress [4–6] (see also Applications of Pressure-Sensitive Products, Chapter 8). In the
shear resistance test the contact area is not constant; thus, at a constant load force the
shear stress gradually increases in time and there is no control of thickness and defor-
mation of the adhesive layer in the normal direction. These factors can cause large
scatter and discrepancies in test results (see Chapter 8). Any slight inclination in the
9-1
direction of the tear force vector would cause a normal force component, which may
influence fracture development for PSAs.
The adhesive elastic fi lm under extension in the normal direction is in a triaxial
stress state (or under negative hydrostatic pressure) [7]. The phenomenon of internal
rupture of tensioned cylindrical rubber samples starting with the formation of micro-
cavities was described 50 years ago by Gent and Lindley [8]. Twenty years previously,
Yerzley (1939) first described “yield-point” in the load–extension relation of similar test
pieces and attributed it to internal rupture [9]. The recent works of Creton and col-
leagues [10–12] analyzed the mechanism of cavity formation during fracture of a PSA
in a probe tack test. The works of Yamaguchi and co-workers [13,14] are examples of a
model approach for the analysis of the cavitation process during the same test using
some assumptions.
∆
H
I
G
E
J
D
t * B()m (9.1)
U
t * B1()m exp M (9.2)
kT
where U is the activation energy of the rupture process and B1, m, and α are the
constants characterizing the material and type of fracture. The physical meaning of
coefficients B, B1, and m is explained in Ref. 18; for epoxy resins at T > Tg constant
m = U0/3kTg [19].
Values of the coefficient B and the exponent m according to Ref. 20 are very sensi-
tive to fracture type and decrease by several orders of magnitude as the fracture type
changes from cohesive to adhesive.
Equation 9.1 noticeably differs from Zhurkov’s classic equation for the durability of
solid materials [18],
U 0
t * 0 exp (9.3)
kT
where τ0 is the thermal oscillation time of atoms (for polymers τ0 ∼ 10−12 to 10−13 s); U0 is
the activation energy of self-induced rupture of polymer chains at σ = 0. The difference
between Equations 9.1 and 9.3 might be explained by the standpoint of fracture mecha-
nisms for brittle solids and elastic materials. In the first case the standpoint is a surface
crack propagation; in the second case it is an internal cavity nucleation inside the material.
Figure 9.2 illustrates that the t* versus σ* curves for PB in contact with a steel sub-
strate exhibit two regions of durability values with different magnitudes of the indi-
cated constants. These regions represent the different modes of fracture of PB fi lms.
In the region of lower values of B and –m (indicated by fi lled symbols in Figure 9.2),
cohesive fracture was observed. For cohesive fracture, similar to uniaxial extension, the
long time durability reduces with decreasing Mw of the adhesive.
Miscellaneous cohesive–adhesive fracture was observed at higher stresses (see sym-
bols ∆ and ◊ in Figure 9.2). In this case, only a few areas of the contact surfaces were
covered with the polymer.
At stresses above 0.9 MPa the fracture becomes adhesive for two specimens of PB
in contact with steel, but tear-off occurs very fast and it is difficult to measure the time
6
4
4
3
2
log t ∗ (C)
2
1
−2
−1.2 −0.8 −0.4 0.0
log σ∗ (MPa)
FIGURE 9.2 Durability t* versus stress σ for PB in contact with PTFE (curves 1 and 2) and steel
(curves 3 and 4) at 20°C. For 1 and 2 the unfi lled symbols relate to adhesive failure; for 3 and 4,
the black symbols denote cohesive fracture; whereas unfi lled symbols indicate a miscellaneous
cohesive–adhesive type of fracture. Mw = 1.5 × 105 g/mol (1, 3) and 6.4 × 105 g/mol (2, 4).
to fracture (<0.l s) The effect of Mw on the long-term durability in the case of adhesive
fracture has been studied for PB in contact with PTFE. The results of these studies are
presented in Figure 9.2 (open symbols). The calculated values of the equation constants
B and m are given in Table 9.1.
The fracture mode was adhesive and the long-term durability in this case was, in fact,
independent of the Mw of the adhesive.
The influence of temperature on long-term durability for the joint of PB (Mw =
6.4 × l05) with PTFE is illustrated in Figure 9.3, and the calculated coefficients in
Equation 9.2 are also listed in Table 9.1.
At 5 and 20°C the fracture was adhesive over the entire range of stresses; at 60°C
a transition from cohesive–adhesive to adhesive failure was observed. In Figure 9.3
the region of cohesive–adhesive fracture is indicated by fi lled symbols. The transition
from one type of fracture to another is clearly seen on the change in the curve’s slope.
As follows from the experimental data, three characteristic regions may be distin-
guished on the t* versus σ* curve, which correspond to the cohesive, miscellaneous
cohesive–adhesive, and adhesive fractures. Upon the transition from cohesive to
cohesive–adhesive fracture, a sharp change in the dependence of long-term durability
on stress is observed, so that in the case of adhesive fracture the long-term durability
drops drastically with increasing stress, as for a brittle fracture.
In the case of cohesive–adhesive and adhesive fracture mechanisms, the t* versus
σ* curves at various temperatures have a more complicated character than those
obtained for cohesive fracture. Fracture of the polymeric layer between two solid sur-
faces and its separation from the solid surface depends largely, according to the litera-
ture [21], on the state of the adhesive, namely, whether it behaves as an elastic, a
viscoelastic, or a plastic body.
The dependence of long-term durability on the inverse absolute temperature under
conditions of cohesive–adhesive fracture and adhesive fracture is illustrated in
3
4 2
3
1
2
log t ∗ (s)
−1
log σ∗ (MPa)
FIGURE 9.3 Durability versus stress σ for PB (Mw = 6.4 × 105 g/mol) in contact with PTFE
at different temperatures. The unfi lled labels denote adhesive failure; the fi lled labels represent
cohesive–adhesive fracture at temperature (1) 5, (2) 20, and (3) 60°C.
7
3
6
8
3
log t ∗ (s) 2 5
4
1
2
0
1
−1
3.2 3.6
T −1 × 10−3 (K −1)
FIGURE 9.4 Cohesive–adhesive fracture. The long-term durability t* versus inverse temperature,
T –1, at stress values: (1) 0.8, (2) 0.75, (3) 0.71, (4) 0.67, (5) 0.63, (6) 0.595, (7) 0.56, and (8) 0.53 MPa.
8
4
log t ∗ a (s)
log t ∗ (s)
6
5 2
2 4
2
3
0 0
3.0 3.2 3.4 3.6
T −1 × 103 (K−1)
FIGURE 9.5 Adhesive failure. The long-term durability t* versus inverse temperature at differ-
ent stresses (solid lines). σ: (1) 0.071, (2) 0.075, (3) 0.08, (4) 0.084, (5) 0.089, (6) 0.094, (7) 0.1, and
(8) 0.106 MPa. The dashed line represents the master curve of log t*aσ versus (T –1 + a σ).
PB (Mw = 6.4 × 105 g/mol) in contact with PTFE.
Figures 9.4 and 9.5. The relationship between log t* and T−1 is reduced to the specific
stress along the abscissa and the ordinate. The master curve of (log t*)aσ, versus
(T−1 + aσ) for adhesive fracture is represented by a dashed line in Figure 9.5 (the reduc-
tion was made to stress σ0 = 0.106 MPa).
t * f ( M ) (9.4)
C1(T T0 )
log aT (9.5)
C2 (T T0 )
The values of the constants, C1 = 1.26 and C2 = 31, were calculated assuming that the
relationship log αT versus (T – T0) is linear, and a reduction was made to T = 20°C.
Figure 9.6 illustrates the master curve representing the relation between log t*aTaM
and log σ* for PB with Mw = 6.4 × 105 and 1.5 × 105 g/mol, reduced to 20°C and Mw =
6.4 × 105 g/mol.
6
log t ∗ a Ta M (s)
2
−1.0 −0.5 0.0
log σ∗ (MPa)
FIGURE 9.6 Master curve for the dependence of durability versus fracture stress (reduced to
20°C) for PB (Mw = 6.4 × 105 and 1.5 × 105 g/mol). The signs are the same as in Figures 9.4 and
9.5; (+) for PB (Mw = 1.5 × 105 g/mol).
A comparison of the obtained results with the data in the literature [7,23,24] indi-
cates that the long-time durability of thin fi lms in the triaxial stressed state is 1.0–1.5
decimal orders of magnitude higher and the fracture stresses are 2–3 times greater
than the corresponding values obtained under conditions of uniaxial extension.
In triaxial stress the fracture begins with the formation of microcavities in the bulk
of the adhesive. There is some delay from the moment of loading to the formation of
the first cavities. The fracture process gradually develops upon increasing the size and
the number of cavities for a certain period of time. At the fast fi nal stage, crack forking
is observed. It is important that at all stages of the fracture process removal of the load
does not lead to the recovery of continuity (for several hours), which indicates unre-
coverable plastic deformation.
Thus, the results of experiments carried out under triaxial stress proved the validity
of Equations 9.1 and 9.2 for elastomer PB and are comparable with the data obtained
for the same samples under simple shear and uniaxial extension [20].
In Vinogradov’s work [15], the rupture time of different types of PB was studied
without the analysis of fracture kinetics, and the adhesive contact had the shape of a nar-
row ring to provide a uniform normal stress distribution. It was of practical importance
to check these results using a test cell with the geometry of probe tack tester, where the
adhesive fi lm was placed between flat surfaces. Research [16,17,25] presents the results
of the study of durability and fracture kinetics of PSAs composed of blends of poly(N-
vinylpyrrolidone) (PVP) with poly(ethylene glycol) (PEG), along with an uncured poly-
butadiene (PB) also used in Vinogradov’s experiments.
In this study a simple experimental procedure, the squeeze–recoil technique, was
developed. This technique provides comprehensive characterization of the viscoelastic
and adhesive properties of the material with a small sample in the course of a single test
cycle [25]. For experiments, blends of PVP (Mw = 106 g/mol) with PEG (Mw =
400 g/mol) and monodisperse PB (Mw = 120,000 g/mol) were used. Previously it was dem-
onstrated that viscosity and the long-term relaxation properties (recovery compliance) of
the system PVP (64%) + PEG and that of PB were similar [26]. Adhesive PVP–PEG
hydrogels were prepared and tested as described previously [27] and other details may be
found in Technology of Pressure-Sensitive Adhesives and Products, Chapter 7.
The polymer sample is placed between flat silica surfaces formed by a loading rod
and a supporting plate of the measuring cell of the thermomechanical analyzer DTMD,
made by a special design workshop (SKB UP) of the Russian Academy of Sciences.
The schematic representation of the test cell is shown in Figure 9.7.
Sample loading under this test is similar to the process of normal compression for
PSAs. The polymer sample squeezed out between flat silica surfaces under constant load
(Figure 9.8, stage I), leading to a decreased gap between the upper and lower plates of the
tester that is equal to the adhesive fi lm thickness.
A sample loading under this test is the same as normal compression for PSAs. More
details about the squeeze flow of viscous fi lm under compression can be found in
Chapter 8.
Removing the load after the squeezing stage leads to a partial recovery (II), whereas
the change of direction of the applied force produces a gradual separation of the plates
and, consequently, an extension of the sample (III).
1 2
FIGURE 9.7 The scheme of the squeeze–recoil test under normal force, F: (1) upper plate,
(2) polymer sample, and (3) bottom plate. (Redrawn from Kotomin, S.V., Borodulina, T.A.,
Feldstein, M.M., and Kulichikhin, V.G., in Proceedings of the XIIIth International Congress on
Rheology, Cambridge, U.K., 4, 44–46, 2000.)
1
I
Squeezing
II
Gap (mm)
Recovery
III
0.1 Debonding
FIGURE 9.8 The stages of the squeeze–recoil test: (1) squeezing, (2) creep recovery, and
(3) debonding for a PVP (64% wt)–PEG hydrogel. (Redrawn from Kotomin, S.V., Borodulina,
T.A., Feldstein, M.M., and Kulichikhin, V.G., in Proceedings of the XIIIth International Congress
on Rheology, Cambridge, U.K., 4, 44–46, 2000.)
1 2 3 4 5 1.0
0.8 6
Transparency (a.u.)
0.8
Gap (mm) 0.6
0.6
0.4
0.4
0.2 0.2
10 100 1000
t (s)
FIGURE 9.9 Gap change (1-5) and light transmittance (6) versus time during plate separation
for PB (2,3) and for the PVP–PEG blends (1,4,5,6) with PVP content (1) 34%; (4) 50%; and (5,6)
64%. The debonding force is (1,4) 0.3 N, (2) 0.5 N, and (3,6) 1 N at ambient temperature.
Using this test, the kinetics of plate separation under constant detaching force and
the strength of the adhesive joint were studied in terms of long-time durability (the time
to debonding). The nucleation of microcavities in the sample and further fibrillation
were observed with a vertical optical microscope and were also monitored by laser light
transmission through the transparent adhesive layer.
The procedure of plate detachment under fi xed tensile load is very similar to the
conventional probe tack test (see Chapter 6), but the distinctive feature of the proce-
dure is that the debonding occurs under constant force, and the time required for the
rupture of adhesive bond characterizes the durability of the joint. The peculiarities of
loading and recovery stages in these experiments were discussed in Ref. 28.
The kinetics of adhesive debonding are illustrated in Figure 9.9 for the PVP–PEG
blends of different compositions.
For these polymeric blends the debonding time differs by several orders of magnitude.
The slower stage of the debonding process results in fibril nucleation, followed by the
faster stage of fibril elongation and fracture of the adhesive film. The first (slower) stage
involves an orientation of polymer chains under applied tensile stress and the second
(much faster) stage includes the elongation flow of polymer chains inside the fibrils until
their cohesive fracture occurs. As demonstrated by many authors, at the early stage the
formation of microcavities is observed [7–14].
The moment of fibril formation corresponds to the border between the slower and
the faster debonding stages. The onset of global fibrillation is visible in Figure 9.9
with an abrupt fall in optical clarity, measured by the intensity of the transmitted light
through the sample in a tension direction.
Changes in the structure of the adhesive fi lm prior to fibrillation have been recently
discussed in detail by Creton et al. [10–12] in probe tack testing of other PSAs. These
results are also in accordance with the peel test data obtained by Chalykh et al. [27]
for the PVP–PEG PSAs. The PVP (64 wt %)–PEG H-bonded stoichiometric complex
5000 500
Peel force (N/m)
Durability (s)
4000 400
3000 300
2000 200
1000 100
0 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7
PEG weight fraction
FIGURE 9.10 Peel adherence and durability for PVP blends with PVP–PEG-400 adhesive.
(Redrawn from Feldstein, M. and Creton, C. Pressure-Sensitive Adhesion as a Material Property
and as a Process in Pressure-Sensitive Design, Theoretical Aspects, VSP, Leiden, 2006.)
demonstrates the best peel and probe tack adhesion and exhibits the longest durability
of the adhesive joints (Figure 9.10) [29].
As Figure 9.10 illustrates, the composition relationship of the long-term durability of
PVP–PEG blends follows the pattern shown by peel strength.
On the microphotographs in Figures 9.11 and 9.12 one can see the fi nal stage of
plate separation for PVP–PEG adhesive composition at the maximal gap between
plates of 3 mm. Fibrillation development for the PVP–PEG PSA depends on viscosity.
For the viscous PVP (64%)–PEG system the fibril structure retains up to the moment of
debonding (Figure 9.11), but for the low-viscosity blend PVP (50%)–PEG, only a single
pillar is observed (Figure 9.12). In the latter case, the pillar formation proceeds most
likely through very fast merging of “liquid” fibrils at extension flow resembling neck for-
mation. For both systems the rupture has miscellaneous adhesive–cohesive character.
As demonstrated in Ref. 25, in contrast to PB, the PVP (64%)—PEG hydrogel is a vis-
coplastic liquid with a defi nite yield stress. Thus, its behavior upon stretching is strain
hardening. Probably for this reason, no PB fibrillation was observed and neck formation
took place only at the lowest stress. At higher stresses (>6.6 × 103 Pa) the plate separa-
tion led to compete PB fracture, which is in accordance with the results obtained by
Vinogradov et al. [15].
The experimental data proved the validity of Equations 9.1 and 9.2 for the durability
of the PB joint for and for the PSA PVP(64%)–PEG, as illustrated in Figure 9.13.
The calculated values of the coefficients in Equation 9.1 and activation energy in
Equation 9.2 at the indicated temperature and stresses values are presented in Table 9.2.
As follows from these data, the activation energy of cohesive fracture for adhesive
PVP–PEG hydrogel is much higher than that for PB.
FIGURE 9.11 Fibrillization of PVP (64%)–PEG PSA under separation of tester plates.
FIGURE 9.12 Neck formation under plate separation for PVP (50%)–PEG adhesive.
log σ (Pa)
5.8 6.0 6.2 6.4 6.6
3 1/T
log σ
log t ∗ (s)
0
2.8 3.0 3.2 3.4 3.6
1000/T (K−1)
FIGURE 9.13 Durability of PVP–PEG adhesive joints versus debonding stress at fi xed tem-
perature and versus inverse temperature at fi xed tensile stress. (Redrawn from Kotomin, S.V.,
Borodulina, T.A., Feldstein, M.M., and Kulichikhin, V.G., in Proceedings of the XIIIth International
Congress on Rheology, Cambridge, U.K., 4, 44–46, 2000.)
The durability of adhesive joints was studied using the setup illustrated in Figure 9.14.
The set of weights renders the control of the compression or extension of the sample
located between flat butt-end surfaces of two steel rods. Varying the weight allows to
squeeze the adhesive fi lm between the rods and then to detach them.
To compare durability for different PSAs, one must use similar test conditions, such
as the load and time of loading. The higher the load, the thinner the adhesive layer and
the higher the long-time durability.
The dependence of durability on loading time at a pressure of 350 KPa is illustrated
in Figure 9.15. Long-time durability for every type of PSA increases gradually and
Weights
500
5
t d (s) 4
250 3
10 20 30
t l (min)
FIGURE 9.15 Detachment time, t*, vs loading time, t l, for PSAs: (1) acrylic Gelva 3011, (2) PIB,
(3, 4) PB, and (5) acrylic DuroTak-87-900A at (1–3, 5) 350 KPa and (4) 490 KPa. (Redrawn from
Sadykova, I., B.Sci. thesis: Long-time durability of acrylic PSAs, Moscow, State Academy of Fine
Chemical Technology (MITHT), 2007.)
3.5
4
3.0
log t (s)
2 3
2.5
1
2.0
FIGURE 9.16 Durability versus stress for (1) Gelva, (2) PB, (3) PIB, and (4) DuroTak PSAs.
reaches a certain limit with time. The loading time was 20 min at a normal stress value
of 350 KPa and at the same (but negative) extension stress. The dependence of PSA
durability on extension stress is illustrated in Figure 9.16 and Table 9.3.
The calculated values of coefficients B and m are presented in Table 9.3, along with
the data obtained for the PVP–PEG hydrogel, which were taken from Figure 9.11.
TABLE 9.3 Parameters B and m in Equations 9.1 and 9.2 Calculated from Data Presented
in Figure 9.16 along with the Values of Practical Work of Adhesion (W) Derived from the
Probe Tack Test
Maximum Work of
Temperature Film Thickness Stress Adhesion,
Adhesive (°С) (mm) B −m (kPa) W (J/m2)
Gelva 30 0.13 46.5 7.8 140 1330
DUROTAK
26 0.09 68.3 11.5 1020 300
87-900A
PIB 20 0.13 56.8 9.6 280 240
PB 29 0.11 80.0 13.7 300 7.5
PVP (64%)–PEG 25 0.10 20.0 0.43 1200 100
Source: Feldstein, M., Creton, C. 2006. Pressure Sensitive Adhesion as a Material Property and as a
Process in Pressure-Sensitive Design, Theoretical Aspects. VSP, Leiden, pp. 22–67 and Sadykova, I.
2007. B.Sci. thesis: Long-time durability of acrylic PSAs, Moscow, State Academy of Fine Chemical
Technology (MITHT).
1.2
1.0
0.8
4
0.6
0.8 0.4
4
Stress (MPa)
0.2 3
0.0 1 2
0 5 10 15 20
0.4
2
1
0.0
0.01 0.1 1 10
Distance (mm)
FIGURE 9.17 Probe tack curves (time–semilog scale): (1) PB, (2) Gelva, (3) PIB, (4) DuroTak [30].
In the upper right corner of the image the same figure is presented in linear time scale. The bonding
force is 10 KN, the contact time is 1 s, and the probe detachment rate is 0.1 mm/s.
Table 9.3 also contains the results of the probe tack test derived from Figure 9.17 [29,30].
In Figure 9.17 the probe tack curves are presented as stress versus probe displacement,
both in the linear and in the logarithmic scale, to discern the curves relating to the
deformation of various adhesives. The constant rate of debonding was 0.1 mm/s.
Table 9.3 indicates that the highest value of the practical work of debonding W is
obtained for Gelva adhesive, but for the same PSA, as seen in Figure 9.16, the value of
durability t*, as well as the apparent tensile stress, is the lowest. Th is contradiction
can be understood if we take into consideration the low value of coefficient m, which
provide for Gelva a good durability at higher stresses. The PVP–PEG PSAs have the
lowest value of m, which means the best durability at high stresses. These results also
mean that the relationship among the durability of various PSA may vary, depending on
the value of apparent tensile stress.
9.4 Conclusions
The dependence of long-term durability of thin viscoelastic polymer fi lms on tempera-
ture and stress is described by equations similar to those derived for cured rubbers,
indicating the similarity between the effect of entanglement networks in noncured
elastomers and that of a network of chemical bonds in vulcanized rubbers. The elastic-
ity and the viscosity of the polymer play a decisive role in the fracture mechanism of
thin adhesive fi lms.
The study of adhesion characteristics of noncured polymers demonstrated that in
the case of adhesive and miscellaneous cohesive–adhesive fractures, long-term durability
is very sensitive to temperature. This feature is associated with the specific temperature
dependence value of total deformation and its recoverable and plastic components for
noncured polymers in the region of transition from the fluid to the rubber-like state.
The behavior of viscoelastic adhesives drastically depends on the rate of deforma-
tion. In the case of constant stress, different types of fracture—cohesive, miscellaneous
cohesive–adhesive, and adhesive—may be observed, depending on the stress value.
The results of durability experiments carried out under triaxial stress proved the
applicability of the simple phenomenological equations for the prediction of PSA behav-
ior under various stresses and temperatures. Characteristic coefficients in Equation 9.1
for various PSAs can be determined both for adhesive fracture and for cohesive debond-
ing. Using these coefficients, one can forecast the behavior and long-term durability of
PSA at any stress and temperature. At the same time, the comparison of peel and probe
tack adhesion with the value of long-term durability requires further study.
References
1. Kinloch A.J. 2002. The durability of adhesive joints. In Adhesion Science and
Engineering, vol. 1. eds. D.A. Dillard, A.V. Pocius, Elsevier, Amsterdam, pp. 661–98.
2. Temiz, S., Ozel, A., Aydin, M.D. 2004. A study on durability of joints bonded with
pressure-sensitive adhesives. J. Adhesion Sci. Technol. 10:1187–1198.
3. Sohn, S. 2003. A new method based on application of cyclic strain to evaluate the
durability of pressure sensitive adhesives. J. Adhesion Sci. Technol. 17:1039–1053.
4. Lim, D.H., Kim, H.J. 2006. General performance of pressure-sensitive adhesives.
In Pressure-Sensitive Design, Theoretical Aspects, ed. I. Benedek, Boston, VSP,
Leiden, pp. 310–311.
5. PSTC-7.
6. ASTM-D3654/D 3654M-027.
25. Kotomin, S.V., Borodulina, T.A., Feldstein, M.M., Kulichikhin, V.G. 1999. Squeeze-
recoil analysis of adhesive hydrohels and elastomer. Polym. Mater. Sci. Eng.
81:425–428.
26. Feldstein, M.M., Lebedeva, T.L., Shandryuk, G.A., Igonin, V.E., Avdeev, N.N.,
Kulichikhin, V.G. 1999. Polym. Sci., Ser. A 41:867–882.
27. Chalykh, A.E., Chalykh, A.A., Feldstein, M.M. 1999. Effects of composition and
hydration on adhesive properties of poly(N-vinyl pyrrolidone)—poly(ethylene
glycol) hydrogels. Polym. Mater. Sci. Eng. 81:456–466.
28. Novikov, M.B., Borodulina, T.A., Kotomin, S.V., Kulichikhin, V.G., Feldstein, M.M.
2005. Relaxation properties of pressure-sensitive adhesives upon withdrawal of
bonding pressure. J. Adhesion 81:77–107.
29. Feldstein, M., Creton, C. 2006. Pressure-Sensitive Adhesion as Material Property
and as a Process, in Pressure-Sensitive Design, Theoretical Aspects. Ed. Benedek I.,
VSP, Leiden-Boston, pp. 22–67.
30. Sadykova, I. 2007. B.Sci. thesis: Long-time durability of acrylic PSAs, Moscow,
State Academy of Fine Chemical Technology (MITHT).
10-1
The diff usion theory of adhesion, originally advocated by Vojutskii14–17 and further
developed by Vasenin18–20 and Chalykh and colleagues, 21–23 states that the intrinsic adhe-
sion of viscoelastic polymers is due to interdiff usion of polymer molecules across the
interface. This requires that the macromolecules, or chain segments, possess sufficient
mobility and are mutually soluble. The latter requirement may be restated by the condi-
tion that they possess similar solubility parameter values. However, when the solubility
parameters of the materials are not similar or one polymer is highly cross-linked, crys-
talline, or below its glass transition temperature, then the interdiff usion is an unlikely
mechanism of adhesion. Most recent advances in the diff usion theory of adhesion are
reviewed in this book by Creton.24
Deryaguin’s electronic theory of adhesion25 treats the adhesive–substrate joint as a
capacitor that is charged due to the contact of two different materials. Separation of the
parts of the capacitor leads to a separation of charge and a potential difference, which
increases until discharge occurs. Adhesion is presumed to be due to the existence of
attractive forces across the electrical double layer. However, it is now established13 that
for typical adhesive–substrate interfaces, generated the electrical double layer does
not contribute significantly to intrinsic adhesion. Further, any electrical phenomena
observed during the joint fracture process arise from the failure event, rather than from
intrinsic interfacial forces.
Nowadays, the adsorption theory has the widest applicability, whereas each of the other
theories may be appropriate only in certain circumstances. The adsorption theory proposes
that the materials adhere because of the interfacial interaction forces, which are established
between the atoms and molecules in the surfaces of the adhesive and substrate.26 Although
for most adhesive–substrate joints the main mechanism of adhesive bond formation is
outlined by adsorption theory, this theory cannot be regarded as universal either, because
the majority of PSAs adhere to substrates of different chemical nature.
All these theories deal with various aspects involved in the attainment of intimate
molecular contact at the adhesive–substrate interface. The attainment of such interfa-
cial contact is invariably a necessary first stage in the formation of a strong and stable
adhesive bond. The next stage is the generation of intrinsic adhesion forces across the
interface, which hold the surfaces of the adhesive and substrate together. Both stages in
combination represent the first stage of the complex phenomenon of adhesion, adhesive
bond formation, or tack. The second and final stage of the adhesion phenomenon is
adhesive joint fracture under applied detaching force that has recently been shown to
represent a multistage process in itself.8
Much of the current confusion in the literature concerns the meaning of the term
adhesion. According to classic definition noted previously, the phenomenon of adhesion
relates rather to instantaneous adherence or tack, a process of adhesive bond formation.
Adhesion, therefore, remains an illusive property, and it is suggested that tack is a mea-
sure of the capability of the adhesive to form a bond with the substrate.27 Information
regarding the magnitude of the intrinsic adhesion forces due to adsorption, diff usion,
mechanical interlocking, or electronic mechanisms and the resulting adhesive bond
formation may be obtained indirectly, based on geometric factors and surface energy.
Direct methods used to measure tack and adhesion (in the classic meaning of the term)
rely mainly on measuring the energy required to break the adhesive bond. PSAs are
viscoelastic materials that deform under attached debonding force, attaining large strain
values. Thus, although the intrinsic adhesion forces affect the strength of the adhesive
joint, their contribution is usually significantly obscured by the work of viscoelastic
deformation of the adhesive material until the adhesive bond breaks. As demonstrated
below, the contribution of viscoelastic deformation of the PSA in the course of debond-
ing prevails against the strength of interfacial interaction forces. The importance of the
rheological behavior of PSAs recently received wide recognition. The very name of the
phenomenon of pressure-sensitive adhesion implies its rheological nature and the con-
tribution of shear deformation under compressive load to the behavior of PSAs. Thus,
although a viscoelasticity theory of pressure-sensitive adhesion seems to be more universal
and provides the most productive approach to the comprehension of this phenomenon, it
conflicts with the classic definition of adhesion.8
FIGURE 10.1 Chemical structure of PVP (left) and PEG (right). m ≈ 10,000; n = 9–10.
PVP
PEG
FIGURE 10.2 Simplified scheme of the proposed molecular structure of a PVP–PEG H-bonded
complex.
400
300
Peel force (N/m)
200
100
60
50
40
)
(%
00 30
2O
10 20 20
H
30 40 10
PEG c 50
ontent 60 0
(%)
FIGURE 10.3 Peel adhesion of PVP–PEG blends as a function of PEG concentration and content
of absorbed water (percentage of water absorbed per 100% PVP + PEG).
composition of PVP–PEG H-bond network complex has been established. 36,39,40 The
phase state of PVP–PEG blends has been considered with DSC in a series of research
papers. 33,35,42–44 The free volume in PVP–PEG blends is evaluated using positron
annihilation.45 Interdiff usion and PVP–PEG miscibility have been characterized
with optical microinterference techniques, 30,32,46 whereas self-diff usion of polymer
components and absorbed water has been studied using pulsed-field gradient (PFG)
nuclear magnetic resonance (NMR).47,48 Rheological and mechanical properties are
also described in detail.49–51 All structures and the properties have been related to
adhesion and evaluated with peel 34 and probe tack 51,52 tests. The main results were
reviewed. 8,11,29
Backing film
Substrate
FIGURE 10.4 Microphotographs of the failure of an adhesive bond in the course of a probe tack
(acrylic PSA, top)8 and a peel test (hydrophilic PVP–PEG adhesive, bottom).34 The intact adhesive
layer of 0.25-mm thickness is seen in the bottom panel as a light band at the border between
the backing film and the PE substrate. (From Feldstein, M.M. and Creton, C., Pressure-Sensitive
Design, Theoretical Aspects, Vol. 1, I. Benedek, Ed., VSP, Leiden, 2006; Chalykh, A.A., Chalykh,
A.E., Novikov, M.B., and Feldstein, M.M. J. Adhesion, 78(8), 667, 2002.)
3. Rapid lateral growth of the cavities during the steep decrease of nominal σ; then,
if there is a plateau in the stress–strain curve, slow growth of these cavities in the
direction parallel to the tensile direction; and fi nally
4. Elongation of the walls between cavities (fibrillation).
These fibrils eventually either break cohesively or detach from the probe surface, caus-
ing complete debonding. As demonstrated by the data in Figure 10.5, the behavior of
the PVP–PEG PSA model follows this general description fairly reasonably.51
12
10
8
Force (N)
0
0 2 4 6 8 10 12
Strain
FIGURE 10.5 Direct observation of the debonding mechanisms and force-versus-strain curve
in the course of probe tack test for a PVP blend with 36 wt % PEG at a debonding rate of 1 µm/s.
Each image corresponds to a specific point in the stress–strain curve.. (From Roos, A.,
Creton, C., Novikov, M.B., and Feldstein, M.M., J. Polym. Sci. Polym. Phys., 40, 2395, 2002. With
permission.)
1.0 31%
0.6
36%
0.4
39%
0.2
41%
0.0
0 4 8 12 16 20 24 28
Tensile strain
FIGURE 10.6 Tensile stress–strain curves to break the PVP–PEG blends, containing 31, 34, 36, 39,
and 41 wt % PEG-400 at 8–9% degree of hydration. The drawing rate is 20 mm/min. (From Novikov,
M.B., Roos, A., Creton, C., and Feldstein, M.M., Polymer, 44(12), 3559, 2003. With permission.)
28
100
24 εb
80
20
W b (MJ/m3)
εb, σb (MPa)
16 wb 60
12
σb 40
8
20
4
0 0
30 32 34 36 38 40 42
PEG content (%)
FIGURE 10.7 The total work of viscoelatic deformation to break the PVP–PEG fi lm, Wb, the
ultimate tensile strength, σ b, and the break elongation, ε b, as a function of PEG concentration
in blends. The extension rate is 20 mm/min. (From Novikov, M.B., Roos, A., Creton, C., and
Feldstein, M.M., Polymer, 44(12), 3559, 2003. With permission.)
600
100
550
90
500
80
450 70
W (MJ/m3)
P (N/m)
400 60
350 50
300 40
30
250
20
200
10
30 32 34 36 38 40 42
PEG (%)
FIGURE 10.8 Effects of PVP–PEG composition on 180° peel force, P, and the work of viscoelas-
tic deformation of the adhesive fi lm up to break, W, under uniaxial extension. The peel and draw-
ing rates are 20 mm/min. (From Feldstein, M.M., Developments in Pressure-Sensitive Products,
2nd ed., I. Benedek, Ed., CRC-Taylor & Francis, Boca Raton, 2006.)
stress–strain curve that defines the value of the total work of viscoelastic deforma-
tion to break the PVP–PEG adhesive blends, Wb, correlates well with both peel34 and
probe tack 51 adhesion and reveals a maximum at 36% PEG concentration for the blend
demonstrating the best adhesion (Figure 10.8).
800
SIS
700
Peel force (N/m)
600
36%
500
400 39%
300 34%
41% PEG
200 31%
FIGURE 10.9 The contribution of the work of viscoelastic deformation of the PVP–PEG model
adhesives and SIS-based PSA (Duro-Tak 34-4230, National Starch & Chemical Corp.) into their
peel adhesion toward the PET substrate. The contents of PEG in the blends with PVP (in wt %)
are indicated. (From Feldstein, M.M., Developments in Pressure-Sensitive Products, 2nd ed.,
I. Benedek, Ed., CRC-Taylor & Francis, Boca Raton, 2006.)
but only the latter blend belongs to the class of PSAs, whereas the former is, in essence,
a tacky liquid. To be a PSA, a tacky material should dissipate an appreciable amount of
energy in the course of debonding, and the value of the work of viscoelastic deformation
to break the tacky fi lm under its uniaxial extension may be taken as a measure of the
dissipated energy (60 MJ/m3 and higher).
Second, the linear relationship between the peel force and the work of deformation
in Figure 10.9 has, most likely, a general character, spanning not only hydrophilic PVP–
PEG, but also hydrophobic rubber-based PSAs. Indeed, the values of peel adhesion and
deformation work for traditional PSAs, based on a styrene–isoprene–styrene (SIS) tri-
block copolymer, are aligned with those for PVP–PEG adhesives.
The linear relationship in Figure 10.9 can be described by Equation 10.1,
b
P kbl ∫ d (10.1)
0
where b and l are the width and thickness of the adhesive fi lm, σ and ε are the tensile
stress and relative elongation, εb is the maximum elongation of the film at the break, and
k is a constant that takes into account the contributions of backing fi lm deformation and
interaction between the adhesive and the substrate. If we compare the peel adhesion of
various adhesives using the same backing fi lm and a standard high-energy substrate,
we can accept k ≈ 1. Assuming further that the deformation of the adhesive fi lm in the
course of both debonding and uniaxial drawing follows the linear elastic law, Equation 10.1
can be written as
b l b2
P (10.2)
E
where the σ b is the ultimate tensile strength and E is an approximate tensile modulus of the
adhesive material. For a PSA, this is not a bad approximation because they usually soften
and then harden at large strains. Equation 10.2 holds for PSAs in the linear elastic region of
deformation, whereas for the PVP–PEG blend the deviation of the measured relationship
from the law presented by Equation 10.2 was earlier shown to achieve 20%.34
Equation 10.2 is similar to the well-known Kaelble equation,60
b l f2
P (10.3)
4E
where σf is a critical value of the ultimate stress upon fracture of PSA material under
debonding from a substrate with a fixed rate. The implication of the similarity of Equations
10.2 and 10.3 is that the Kaelble equation, Equation 10.3, holds for any type of PSA, including
the hydrophilic PVP–PEG. Thus, the rule described by Equations 10.2 and 10.3 is universal.
Equation 10.6 is, of course, only qualitatively illustrative, because it makes many crude
approximations, including ignoring the existence of the spectrum of relaxation times.
It is inappropriate for quantitative calculations of peel force, because it includes immea-
surable terms like a (the size of the diff using polymer segment). Nevertheless, it pre-
dicts qualitatively the significance of diff usion and relaxation processes (both of which
require molecular mobility) for the adhesive behavior of polymers when their debond-
ing is dominated by the formation of fibrils. Equation 10.6 was derived on the basis of
the analysis of the deformation contribution to peel adhesion without resorting to the
so-called diff usion theory of adhesion.14 Thereby, the rheological approach based on the
analysis of viscoelastic deformation of the adhesive material under the debonding pro-
cess, described here, has more universal character than others mechanisms of adhesion
considered previously in this chapter.
B
D A exp (10.7)
fv
where A and B are constants. A specific feature of all PSAs is that they should com-
bine a high energy of cohesive interaction with a large free volume. Most commonly,
the strong cohesive interaction between macromolecules causes a drastic decrease in
free volume, which explains why pressure-sensitive adhesion is a comparatively rare
phenomenon.
In the model PVP–PEG system these apparently confl icting properties are never-
theless conciliated due to the location of reactive hydroxyl groups at the opposite ends
of PEG chains of appreciable length and flexibility (see the scheme in Figure 10.2). In
other PSAs of different chemical compositions, these confl icting properties may be
conciliated in a variety of alternative ways. High cohesion energy may result from
intermacromolecular cross-linking (both covalent and noncovalent and the entangle-
ments of long chains), the addition of tackifiers with high Tg, or to the hydrophobic
association of side groups of polymer chains. Large free volume is most frequently
provided by the usage of elastomers with a low glass transition temperature, Tg.
The glass transition temperature relates to the energy of cohesion and free volume by
the equation63
z D0
Tg 0.445 (10.8)
R
where z is the coordination number, a value that is inversely proportional to the free
volume, and 〈D 0〉 is the total interaction energy of atoms forming a polymer segment. In
acrylic PSAs containing neutralized carboxylic groups, the increase in free volume may
result from electrostatic repulsion of the carboxylate anions. In un-cross-linked poly-
isobutene (PIB) PSA, cohesive strength is a consequence of the presence of a network
of long-chain entanglements of high-molecular-weight fraction and is also due to the
van der Waals interaction between nonpolar functional groups. In SIS-based triblock
copolymers and other thermoplastic elastomers, cohesion is provided by the physical
cross-links of high-Tg polystyrene blocks, whereas the free volume is provided by blocks
of lower molecular weight polymers.
The combination of high cohesion energy and large free volume, featured for all PSAs,
is also embedded in such fundamental characteristics of adhesive materials as specific
values of solubility parameters, δs, defined as the cohesive energy density or the ratio of
the energy of cohesion to the total volume,
1
H v RT 2
s (10.9)
Vm
where ∆Hv is the molar heat of vaporization, R is the gas constant, T is the tem-
perature (K), and Vm is the molar volume. The heat of vaporization, ∆Hv, is the direct
measure of cohesion energy because it is defi ned as the total amount of energy needed
to overcome intermolecular attractive forces and transfer a molecule to the vapor phase,
in which no intermolecular interaction occurs. The solubility parameter relates to the
enthalpic component of the Flory interaction parameter χ between monomer units of
two polymer chains i and j (or between a polymer and a solvent) by the equation
Vm
ij (i j)2 (10.10)
RT
The smaller the difference between the solubility parameters of an adhesive and a sub-
strate, the greater the peel adhesion.9
Viscoelastic
Glasses Liquids
materials
σb
FIGURE 10.10 Directions of varying the values of relaxation time, τ, self-diff usion coefficient, D,
and ultimate tensile strength, σ b, under the transition from glassy to liquid state.
from glassy polymer to viscous liquid, for example, with the increase of PEG plasti-
cizer content in blends with PVP. Indeed, the longest relaxation times are featured for
glasses (years and centuries), whereas low-molecular-weight liquids relax almost instan-
taneously. In contrast, the lowest diff usion coefficients are observed for glasses, whereas
the highest diff usion coefficients are demonstrated in liquids and gases. According to
Equation 10.6, maximum peel strength, P, relates to the maximum magnitude of the τD
product. Evidently, this product achieves its maximum magnitude in a certain range of
values of relaxation time and diff usion coefficient, which are intermediate between those
inherent for liquids and glasses. The materials coupling the properties of the liquids and
the solids are in a viscoelastic state, which is why all PSAs are viscoelastic materials.
8.0 600
7.5 500
7.0 400
P (N/m)
f v (%)
6.5 300
6.0 200
5.5 100
5.0 0
0 10 20 30 40 50
PEG (wt %)
FIGURE 10.11 The fraction of free volume ( fv, %), and peel adhesion (P, N/m) of PVP blends
with various amounts of PEG-400 at 100% relative humidity of the surrounding atmosphere.
where p1 are the fractions (populations) of resonating species possessing the self-diff usion
coefficient D1.
For the PVP–PEG blend that contains 36 wt % PEG-400 and exhibits the best adhe-
sion, only a single population of diffusing species (D = 2.3 10−13 m2/s) was observed 2 days
after blend preparation. Within the following 32 days, the diff usion decay evolved into
a biexponential behavior with a slower self-diff usion coefficient of 1.5 10 −13 m 2/s and
p = 0.87.65 After this time, the diff usion properties of the system were unchanged. Such
an aging process is typical for self-assembling supramolecular structures involving the
hydrogen bonding and the stoichiometric complex formation.66 The established varia-
tion in molecular mobility has only a very slight effect on adhesion. The implication is
that the best adhesion in the PVP–PEG system relates to the values of the self-diff usion
coefficient ranging from 1.5 10−13 to 2.3 10−13 m2/s.
To what component of the PVP–PEG blend may the established self-diff usion coefficient
be attributed? The binary PVP–PEG blend contains three components: high-molecular-
weight PVP with an extremely low self-diff usion coefficient, a short-chain PEG with
a comparatively high diff usion coefficient, and the product of their interaction—the
stoichiometric H-bonded network PVP–PEG complex, whose formation underlies the
occurrence of pressure-sensitive adhesion. The PVP–PEG blends may be also treated as
concentrated solutions of glassy PVP in liquid PEG-400. Two populations of diff using
molecules in the PVP–PEG system were detected. As the PEG concentration in blends
decreases, tending to a stoichiometric complex composition ([PEG]:[PVP] = 0.15), the
portion of slow-diff using protons, p, approaches unity (Figure 10.12). Such behavior
demonstrates that the D value of the slow-diff using component characterizes the seg-
mental mobility in the PVP–PEG complex. As [PEG]:[PVP] → 1, p → 0. Within this
composition range, only the fast-diff using component (PEG) is observed. This area of
the composition may be defined as a solution of the stoichiometric PVP–PEG complex
in PEG. However, at [PEG]:PVP > 1, the fraction of slow-diff using component increases
again. This effect may be treated as increasing involvement of PEG chains in the motion
of PVP segments, whose self-diff usion coefficient increases in dilute solutions, approach-
ing a value of D ≈ 10−12 m2/s.
Figure 10.13 compares the peel adhesion (P) of the PVP–PEG model adhesives with
their diff usion behavior expressed in terms of the self-diff usion coefficient of slowly
diff using component, measured using PFG NMR techniques,47,48 and the partial diff u-
sion coefficient of PEG-400 in blends with PVP (DPEG ). The latter has been evaluated
[PEG-400]/[PVP]
0.15 1.0
−11.2 0.9
−11.4 0.8
Dfast
−11.6 0.7
−11.8 0.6
log D (m2/s)
Dslow
−12.0 0.5
−12.2 P
−12.4 pslow 0.3
−12.6 0.2
−12.8 0.1
−13.0 0
30 40 50 60 70 80 90 100
PEG-400 (wt %)
FIGURE 10.12 The effect of PEG content (wt %) (bottom axis) in terms of the number of PEG
molecules per one PVP repeating unit, [PEG-400]:[PVP], top axis, on the self-diff usion coefficient
(D) and population (p) of the fast and slow diff using components at 20°C. The diff usion time is
103 ms. The composition in which the stoichiometric PVP–PEG complex is fully formed and that
demonstrates best adhesion is denoted by the dashed line on the left.
1E-11 600
550
DPEG
P 500
1E-12 Dslow 450
D (m2/s)
P (N/m)
400
350
1E-13
300
250
200
1E-14
20 30 40 50 60 70 80 90 100
PEG (%)
FIGURE 10.13 Partial diff usion coefficient (D PEG) of PEG-400 in blends with high-molecular-
weight PVP of different compositions, the self-diff usion coefficient of the slow diff using component,
and peel adhesion (P) at room temperature.
from the values of the PVP–PEG interdiff usion coefficient (D v) with optical microinter-
ferometry.30,32 DPEG relates to D v by the equation21
Dv
DPEG (10.12)
PVP
where φPVP is the volume PVP fraction in the binary blend. As established, the values of
DPEG exactly meet those of the self-diff usion coefficient of the fast-diff using component
by PFG NMR (Figure 10.12). Thus, the fast-diff using component in PVP–PEG blends
most likely represents the PEG that is uninvolved in the network complex with PVP.
As follows from the data presented in Figure 10.13, the self-diff usion coefficient of
PEG-400 within the H-bonded PVP–PEG network at 36% PEG concentration in the
blend is DPEG = 1.3 10−13 m2/s. The independent data of the microinterferometry and
PFG NMR are in good agreement and outline diff usivity of the model PVP–PEG adhe-
sive that provides the best adhesion on the border between 10−12 and 10−13 m2/s. Such
high self-diff usion coefficients are most likely due to low molecular weight and high
flexibility of PEG short chains. Needless to say, such high diff usivity is also favorable for
fast adhesive bond formation. Thus, the liquid-like diff usion mobility of PSAs is needed
not only to provide deep penetration of the PSA polymer into a substrate under slight
pressure, but also to develop large tensile strain of the adhesive layer under a detach-
ing force. As Equation 10.6 predicts, high adhesion is associated with high diff usion
coefficient values. As direct measurements of diff usivity in the PVP–PEG model PSA
demonstrate, the PEG self-diff usion coefficient in this PSA relates to the border between
elastomers and viscous liquids, ~10−13 m2/s.
80 400
Activation energy
70 350
Peel force
for self-diffusion (kJ/mol)
60 300
Activation energy
50 250
40 200
30 150
20 100
10 50
0 0
0.2 0.4 0.6 0.8 1.0
PEG weight fraction
FIGURE 10.14 Relationships between PEG content in the blends with PVP, free energy for self-
diff usion, and peel adhesion.
55.3 kJ/mol. As the activation energy increases to 69.7 kJ/mol with the decrease in
PEG concentration to 30 wt %, the locus of adhesive joint failure of the PVP–PEG PSA
with PE substrate changes from miscellaneous (in the maximum of peel adhesion)
to adhesive. In contrast, at higher PEG concentrations (40–60% PEG) the activation
energy varies between 52.3 and 48.2 kJ/mol. For these blends the locus of failure is
always cohesive.
Because debonding occurs after a certain time under a fi xed tensile force, the time, t*,
required to rupture the adhesive bond characterizes the durability of the joint. The dura-
bility of an adhesive joint is a fundamental quantity that characterizes pressure-sensitive
adhesion, as described in Chapter 9.67 Temperature dependence of the logarithm of
durability follows the Arrhenius relationship fairly reasonably, allowing evaluation of
the activation energy for the fracture process of PVP–PEG adhesives under detaching
stress. The latter value is plotted versus the composition of PVP–PEG adhesives in
Figure 10.15, along with the durability of the adhesive joint and the activation energy
for PEG self-diff usion, determined using the PFG NMR method.47,48
As the data in Figure 10.15 demonstrate, the activation energy for adhesive bond
failure follows the pattern of the activation energy of the self-diff usion coefficient,
climbing sharply with a decrease in PEG concentration below 36%, when maximum
adhesion has been achieved. For a blend with maximum adhesion, the activation
energy for adhesive debonding is 10–15 kJ/mol higher than the activation energy
for self-diff usion. The translational mobility measured in terms of the self-diff usion
coefficient takes an appreciable part of the activation energy for adhesive debonding
in the blends, which exhibit a miscellaneous (adhesive–cohesive) mechanism of adhe-
sive joint failure.34 For low PEG weight fractions, the activation energy of the debond-
ing process reaches a value of 210 kJ/mol. Within this composition range the lack in
5000
ED self-diffusion
200
Ea adhesive debonding 4000
Activation energy (kJ/mol)
Adhesive durability
150 3000
2000
100
1000
50 0
FIGURE 10.15 Relation of adhesive joint durability and activation energy for adhesive debond-
ing and for PEG self-diff usion to the composition of PVP–PEG adhesive blends. (From Feldstein,
M.M., Developments in Pressure-Sensitive Products, 2nd ed., I. Benedek, Ed., CRC-Taylor & Francis,
Boca Raton, 2006.)
molecular mobility makes a cohesive fracture of the joint impossible and the fracture
proceeds mainly by an interfacial mechanism of crack propagation.
If one examines the monomer unit level, the activation energy for self-diff usion, ED,
should be a function of the product of the cohesive energy density (CED) and the volume
of a mole of cylindrical cavities required for diff usion of a polymer chain segment of
diameter d over a jump length λ,
d 2
ED CED (1 f v )Ec (10.13)
4
where Ec is the cohesive energy and fv is the fractional free volume in the polymer.
Thus, a phenomenological analysis of the relationship between pressure-sensitive
adhesion and the molecular mobility of the PSA, outlined by Equations 10.6 and 10.13,
suggests that the two important parameters controlling pressure-sensitive-adhesion and
diff usion are the energy of favorable intermolecular interactions (cohesion) and molecu-
lar free volume. The difference between adhesion and diff usion is rather quantitative:
adhesion occurs only within a very narrow range of the ratio of cohesion energy to free
volume, and in this case both the former and the latter magnitudes are to be high. In
contrast, diff usion takes a place at any value of cohesion energy and free volume if only
the gradient of concentration is available.
velocity on the strength of adhesive joints.6 However, Equation 10.6 poses at least two
new questions:
Because the values of both the relaxation times and the moduli depend on the type of
deformation of adhesive fi lm, which is different under compressive bonding force and
at the stage of debonding, more extensive research is necessary before the significance
of the relaxation processes for the adhesion can be properly established. For a detailed
discussion of this problem, see Chapter 11.68
TABLE 10.1 Glass Transition Temperatures and Relevant Values of Heat Capacity Jumps
at the Glass Transition Featured for Typical Pressure-Sensitive Adhesives
200 400
Upper Tg
150 Lower Tg
Tm 300
100 Peel force
200
50
0
100
−50
0
−100
0 20 40 60 80 100
PEG (wt %)
FIGURE 10.16 Peel adhesion and phase state of PVP–PEG systems. Tg is the glass transition
temperature and Tm is the melting temperature of PEG in blends with PVP. (From Feldstein,
M.M., Developments in Pressure-Sensitive Products, 2nd ed., I. Benedek, Ed., CRC-Taylor & Francis,
Boca Raton, 2006.)
Figure 10.16 compares adhesive behavior with the phase state of PVP–PEG blends.11
The PVP–PEG system is plausibly among the first and most illustrative examples of
miscible single-phase polymer blends, which reveal two distinct relaxation transitions.
These transitions are seen on DSC scans as the heat capacity jumps, resembling glass
transitions, and the corresponding glass transition temperatures demonstrate coherent
compositional behavior.40 The behavior of the upper Tg in PVP–PEG blends obeys the
rule of homogeneous PVP–PEG mixing or glassy PVP dissolution in liquid PEG due to
PEG chains H-bonding to PVP through one terminal group only.40 This un-cross-linked
and labile PVP–PEG complex requires a small amount of heat for dissociation and is
comparatively unstable. In contrast, lower Tg is due to the formation of a hydrogen-
bonded PVP–PEG network complex (gel), which behaves like a new chemical entity.38–40
In this stoichiometric complex, nearly 20% of PVP repeat units are cross-linked by
PEG terminal OH-groups (via H-bonding). The lifetime of the cross-linked PVP–PEG
complex is much longer due to multiple hydrogen bonds involved in its formation.
Another example of a single-phase system with two Tg is partly denaturated proteins,
where a lower Tg relates to glass transition and an upper Tg corresponds to the relaxation
of denaturated protein.70
The data presented in Figure 10.16 demonstrate that the Tg values for adhesive PVP–PEG
blends range between –55 and –71°C for blends containing 12 wt % of absorbed water.40
The effects of PEG and absorbed water contents on the Tg in adhesive PVP–PEG blends is
illustrated in Figure 10.17. If the amount of absorbed water in the PVP–PEG adhesive is
from 4 to 8 wt %, the Tg values have been found to vary from –45 to –52°C (Figure 10.17).35
Thus, the Tg values for the model PVP–PEG hydrophilic adhesives represent no exceptions
to the rule that was established previously for typical, hydrophobic PSAs.
PVP dry
150
Tg (°C)
50
−50
0
20 15
40 PEG
10
PE 60
G 5
(%
)
80 (%)
100 0 H 2O
FIGURE 10.17 The glass transition temperature of PVP–PEG adhesive blends as a function of
PEG and water content. (From Feldstein, M.M., Kuptsov, S.A., Shandryuk, G.A., Platé, N.A., and
Chalykh, A.E., Polymer, 41(14), 5349, 2000.)
30 35 40
30
7.0
25
20 6.5
fv (%)
15
εb
6.0
10
5.5
5
0 5.0
0 10 20 30 40 50
PEG in blend (%)
FIGURE 10.18 Fractional free volume, fv, and maximum elongation of adhesive film at the break
of uniaxially stretched material, εb, as functions of PVP–PEG composition. (From Feldstein, M.M.,
Developments in Pressure-Sensitive Products, 2nd ed., I. Benedek, Ed., CRC-Taylor & Francis, Boca
Raton, 2006.)
9 σb 30
σb, σb / εb (10−1), σb εb (%)
6 σb εb 20
εb
εb
3 10
σb / εb
0 0
30 32 34 36 38 40 42
PEG (wt %)
FIGURE 10.19 Relationship of ultimate tensile stress (σ b) and maximum elongation at break
(εb), their product, and the ratio to the content of PEG-400 in the blends with PVP. (From Feldstein,
M.M., Developments in Pressure-Sensitive Products, 2nd ed., I. Benedek, Ed., CRC-Taylor & Francis,
Boca Raton, 2006.)
fractional free volume and maximum elongation at break. Consequently, the εb value,
which characterizes a process of elongational flow, can be taken as an indirect measure-
ment of the free volume in PVP–PEG blends.
With the rise in PEG concentration, the cohesive strength of PVP–PEG adhesives,
embedded by the σ b quantity, grows until 36% PEG concentration is achieved and then
decreases (Figure 10.19). This implies a twofold role of PEG in the blends. In PVP-
reached blends, where an H-bonded complex is forming, PEG acts simultaneously both
1
6
4 80
20
10
2 5
σ (MPa)
2 1 µm/s
0.1
6
4
0.01
0 2 4 6 8 10 12 14
FIGURE 10.20 Typical stress–strain curves obtained in probe tests for different probe debond-
ing velocities at a constant PEG content (36 wt % PEG). The maximum extension measured (the
point where the force drops to zero) increases monotonously with decreasing debonding velocity.
The velocities used are (from right to left) 1, 2, 5, 10, 20, and 80 µm/s. (From Roos, A., Creton, C.,
Novikov, M.B., and Feldstein, M.M., J. Polym. Sci., Polym. Phys., 40, 2395, 2002. With permission.)
1.6
100 mm/min
1.4
1.2
Nominal stress (MPa)
1.0
50 mm/min
0.8
0.6
20 mm/min
0.4
0.2
10 mm/min
0.0
0 5 10 15 20 25
Tensile strain
FIGURE 10.21 Stress–strain curves to break the PVP–PEG hydrogel, containing 36 wt % PEG-
400 and 8–9% water, at drawing rates ranging from 10 to 100 mm/min. (From Novikov, M.B.,
Roos, A., Creton, C., and Feldstein M.M., Polymer, 44(12), 3559, 2003. With permission.)
strain-hardening region, where the final rupture of H-bonded cross-links between the
PVP chains occurs (Figure 10.21). In contrast, at higher extension rates the H-bonds
have insufficient time for rearrangement at new places and behave like pseudo cross-
links, which must be ruptured to deform the polymer. The narrow transition from
ductile to tight stretching with the rise of drawing rate in Figure 10.21 corresponds to
a well-defined rate of rearrangement of the H-bonded network under drawing of the
PVP–PEG adhesive. Assuming that breakup and reformation of hydrogen bonds form-
ing a PVP–PEG network can occur below the critical deformation rate of 0.05 s–1, we
can identify the characteristic time for this process to occur at about 20 s.51 The effects
of deformation velocity on adhesion and large-strain viscoelastic behavior of the model
PVP–PEG adhesive provide a highly illustrative example of how the relaxation processes
control the performance properties of PSAs.
900
10 mm/min
800
SIS
Nominal tensile stress (kPa)
700
600
500
400
300
200
PVP−PEG
100
PIB
0
0 5 10 15 20 25
Tensile strain
1200
Nominal tensile stress (kPa)
1000
600
400
200
PIB
0
0 2 4 6 8 10 12 14 16 18 20 22
Tensile strain
FIGURE 10.22 Stress–strain curves to break the PVP–PEG (36%), PIB, and SIS block copolymer
(Duro-Tak 34-4230) adhesives at extension rates of 10 and 100 mm/min.
fact that the 10-fold increase in drawing rate affects the ductile–tight behavior of the
hydrogen bonded PVP–PEG adhesive to a much greater extent than in conventional
hydrophobic adhesives. In other words, the ductile–tight transition in the SIS and PIB-
based adhesives is much wider than that in PVP–PEG blends. Thus, the sharp transi-
tion between the ductile and tight deformation modes with the change of extension
rate is featured only for the H-bonded PVP–PEG system and is atypical of SIS- and
PIB-based adhesives.
As follows from the comparison of tensile stress–strain curves presented in Figures 10.6
and Figures 10.21 through 10.23 and Table 10.2, according to their tensile behaviors
20
18 10 mm/min
PIB
16
14
12
σn (kPa)
10
4 Acrylic
0
0 10 20 30 40 50 60 70
ε
60
100 mm/min
50
40
Acrylic
σn (kPa)
30
20
10 PIB
0
0 10 20 30 40 50 60
ε
FIGURE 10.23 Stress–strain curves to break the PIB and acrylic (Duro-Tak 87-900A) adhesives
at extension rates of 10 and 100 mm/min.
all PSAs can be classified into two main groups. The PSAs of the first group, exemplified
by un-cross-linked PIB and acrylic adhesives, represent very soft, compliant materi-
als with low tensile modulus (E, defined as the slope of initial part of the stress–strain
curve), which deform like viscous liquids, demonstrating negligible or zero values of
ultimate tensile strength (σ b), comparatively low work of viscoelastic deformation up to
break (W), and extremely high values of maximum elongation at break (εb). In fact, such
adhesives never break in the course of tensile test at the employed values of extension
Extension Rate
PSA (mm/min) E (MPa) σb (MPa) σy (kPa) εb W (MJ/m3)
PVP–PEG 10 0.56 4.2 16 17.9
SIS 1.42 19.77 37.9 23 114
PIB 0.12 0.00006 21.4 7.5 0.14
Acrylica 0.024 ∞
PVP–PEG 100 3.63 10.4 8.6 24.6
SIS 3.51 28.57 84.1 20 156
PIB 0.18 0.063 143.4 15.8 0.88
Acrylica 0.083 ∞
a Duro-Tak 87-900A.
velocities because the distance between tester cross-heads achieves its limit before the
breakpoint. Therefore, the characteristic value of ultimate tensile stress, σ b, cannot serve
as a measure of cohesive strength for these materials. Instead of σ b, the value of yield
stress (the maximum on stress–strain curve), σy, can be used for comparative character-
ization of the cohesive strength of fluid adhesives (Figure 10.23).
The strain-hardened PSAs of the second group, to which physically cross-linked SIS and
PVP–PEG adhesives belong, deform like typical rubber-like networks,60,72–74 exhibiting
much higher values of tensile modulus (E), work of extension up to break (W), and appre-
ciable resistance to strain, expressed in terms of high tensile strength, σ b (see Table 10.2).
Covalent cross-linking of fluid acrylic adhesives with Ti-chelate (namely with Ti-acetylace-
tonate) has been shown75 to change the type of tensile stress–strain curves to the deforma-
tion with pronounced strain hardening that is typical of the adhesives in the second group.
In contrast to such quantities as glass transition temperature, the characteristic val-
ues of tensile test do not generate any universal range of magnitude for various PSAs.
This implies that although the values of ultimate tensile stress and maximum elongation
at break are, respectively, indirect characteristics of cohesive strength and free volume,
this is true only for an individual PSA. The σ b and εb values cannot be utilized for com-
parison of cohesive strength and free volume of different PSAs that belong to different
groups. The reasoning behind this conclusion is that most likely that the structures of
the strain-hardened (cross-linked) and viscous adhesives at large tensile strain are too
different to be characterized in terms of tensile test parameters.
The storage modulus, G′, measures the elasticity of the adhesive. High G′ values are
typical for hard polymers with great cohesion energy. The loss modulus, G″, is associated
with energy dissipation during deformation. The factor underlying the G″ value at the
molecular level is free volume. The greater the G″ value relative to G′, the more dissipative
the adhesive. The G″/G′ ratio, defined as the loss tangent, tan δ, is the balance of viscous/
elastic behavior. tan δ = 1 is a limiting value. Above this value the adhesive is generally
considered a viscoelastic fluid, whereas below this value, the adhesive can be considered a
viscoelastic solid. Once tan δ > 1, the free volume dominates the cohesion energy. Con-
versely, if tan δ < 1, the contribution of cohesion energy overrides that of free volume.
This suggests that all PSAs should possess the specific range of G′, G″, and tan δ values.
The DMA technique allows us to estimate indirectly the ratio of cohesion energy to
free volume in terms of tan δ values and provides indirect separate characteristics of the
cohesion energy and free volume contributions in terms of G′ and G″, respectively. In
this section we focus on outlining the range of G′, G″, and tan δ values, which are typi-
cal for both conventional hydrophobic and hydrophilic PVP–PEG model adhesives.
Based on DMA data, one of the most important critera for a material to display PSA
behavior is Dahlquist’s criterion of tack, which states that the elasticity modulus of a PSA
should be below 105 Pa.5,6 Typically, at a deformation rate of 1 Hz and ambient tempera-
ture, the values of G′ for various PSAs fall in the range 0.01–0.1 MPa (Figure 10.24).80
Typical PSAs are used in a temperature range corresponding to the beginning of the
high-temperature rubbery plateau or the end of the transition region (Figure 10.24).
Typical values of tan δ for conventional PSAs range from 0.7 to 1.0, implying that either
the contributions of free volume and cohesion energy are counterbalanced or the inter-
molecular cohesion slightly dominates. The temperatures relating to the loss tangent
peak vary most frequently between −50 and +5°C.
9.5
tan δ 1.2
9.0 PSA
8.5 1.0
log G′, log G′′ (Pa)
8.0
0.8
7.5
tan δ
7.0 0.6
6.5
0.4
6.0 G ′′
5.5 0.2
G′
5.0
0.0
4.5
−120 −80 −40 0 40 80 120 160
T (°C)
FIGURE 10.24 Idealized plot of the storage modulus, G′, loss modulus, G″, and tan δ for a typical
PSA as a function of temperature.80 ωref = 1 Hz.
Until recently, Dahlquist’s criterion of tack described fairly reasonably the interrelation-
ship between the adhesive and rheological properties of all PSAs known to date. Recently a
new class of PSAs has been obtained, exemplified by a part of triblock thermoplastic elas-
tomers81 (Figure 10.25)82 and hydrophilic PVP–PEG model adhesive (Figure 10.26),8,11,29,49
the behavior of which sometimes disobeys appreciably Dahlquist’s criterion.
G′ 10
1E9
1E8 G ′′
G ′, G ′′ (Pa)
1E7 1
tan δ
1000000
100000
0.1
10000
tan δ
1000
FIGURE 10.25 Temperature sweep curves of the storage modulus, G′, loss modulus, G″, and
tan δ for a SIS-based hot-melt PSA (NSC 12602-60H1). (Courtesy of Dr. Y. Hu, National Starch
and Chemical.)
9 2.0
G′
8
1.5
log G′, log G ′′ (Pa)
G ′′
7
tan δ
1.0
6
tan δ
5 0.5
4
0.0
−40 0 40 80 120 160
T (°C)
FIGURE 10.26 Temperature dependence of dynamic shear moduli G′, G″, and tan δ for PVP–
PEG (36 wt %) PSA at a reference deformation frequency of 1 Hz.
As illustrated by the curves in Figure 10.25, the values of G′ for SIS-based PSA lie in
the range from 1.94 to 0.40 MPa with an increase in temperature from 15 to 40°C. SIS-
based PSA is used in a temperature range corresponding to the beginning of the high-
temperature rubbery plateau and the end of the transition region (Figure 10.25). Only in
this connection does its behavior not deviate too much from that of other PSAs. The val-
ues of tan δ for the SIS PSA range from 3.1 to 0.4 between 15 and 40°C. The temperature
of loss tangent peak is 15°C, tan δ = 3.1. At ambient temperature (20°C), G′ = 0.96 MPa,
G″ = 2.16 MPa, and tan δ = 2.2.
At increased temperatures, the value of G′ for the PVP–PEG adhesive (Figure 10.26)
decreases in the range from 2.95 (15°C) to 0.09 MPa (40°C). At 20°C, G′ = 1.23 MPa, a
value clearly incompatible with Dahlquist’s criterion for tackiness, which specifies that
an adhesive loses its tack if its elastic modulus at 1 Hz is higher than about 0.1 MPa. The
PVP–PEG blend is tacky in a region corresponding to the center of the transition region,
where the loss tangent peak occurs (25°C, tan δ = 1.6).
The DMA data in Figures 10.25 and 10.26 were measured in the linear elastic regime,
that is, at very small strains. Let us refer now back to large strain data for the PVP–PEG
adhesive presented in Figure 10.19. The ratio of ultimate tensile stress to maximum
elongation at break, σ b/ε b, can be interpreted physically as an average modulus of the
adhesive at the moment of fracture under uniaxial extension.49 As the data in Fig-
ure 10.8 indicate, the maximum energy that must be expended to draw and break a
unit volume of the PVP–PEG adhesive (~40 to 90 MJ/m3) corresponds to the apparent
tensile modulus σ b/ε b ≈ 2 – 9 × 105 Pa. The latter values correspond exactly to those
specified by Dahlquist’s criterion, taking into account that E = 3G. Thus, Dahlquist’s
criterion holds for the PVP–PEG systems at large strains, rather than within the linear
elastic region of deformation. In other words, the PVP–PEG blend is very stiff at small
strains but undergoes a pronounced softening above 10–20% deformation, which puts
it in the range to be a PSA.
Similar yet less pronounced inconsistencies between small and large strain behaviors
of PSAs are also observed for polymers with more complicated architectures and very
pronounced elastic effects.83 For such polymers the elongational properties, involving
the orientation of polymer chains, cannot be readily predicted from the linear elastic
properties in shear. One such class of polymers is that of diblock and triblock copoly-
mers of styrene and isoprene, which serve as base components of PSA formulations. As
recently established by Roos and Creton,73 much like PVP–PEG adhesives, the blends
of triblock and diblock copolymers of polystyrene (PS) and polyisoprene (PI) display
marked nonlinear viscoelastic behavior and are much more dissipative at low strains
than at high strains. Let us turn refer to Figures 10.22 and 10.23 and recall that, in
terms of large strain behavior, the PVP–PEG and SIS adhesives form a separate group
of PSAs, which elongate as lightly cross-linked rubbers with pronounced strain harden-
ing, whereas other (un-cross-linked) PSAs (PIB, acrylic) deform as viscous fluids. The
implication of these observations is that Dahlquist’s criterion of tack holds only for
un-cross-linked adhesives with random structure. To become more universal, it must be
now extended by taking into consideration the values of storage modulus obtained for
cross-linked adhesives and supplemented by the inclusion of loss tangent magnitudes
and peak temperatures.
10.7 Conclusions
In a manner quite similar to that of many other properties of materials, such as glass
transition, solubility, or diff usion, pressure-sensitive adhesion results from the coun-
teraction of molecular cohesion and free volume. The length scale where these two
factors are important is different: a PSA requires a low density of strong bonds that
are relatively far apart and a high density of weak bonds with a high molecular mobil-
ity. A large free volume, resulting in liquid-like fluidity, is necessary to facilitate the
formation of good adhesive contact as compressive bonding force is applied and to
develop large deformations of a PSA under detaching force. High cohesion energy
of the PSA material is required to resist the fracture of adhesive bond and dissi-
pate much mechanical energy in the course of debonding. Once the adhesive bond
is properly formed, viscoelasticity theory describes the PSA behavior under detach-
ing force fairly reasonably. This viscoelasticity theory claims that the strength of the
adhesive joint is controlled by the amount of viscoelastic energy needed to stretch and
break the adhesive fi lm cohesively. The viscoelasticity theory seems to be much more
universal than other theories, such as mechanical interlocking, diff usion, electronic,
and adsorption theories, and it involves the contributions of high cohesive strength, high
diff usion coefficient, and long relaxation time. Because the cohesive strength, diff usion
coefficient, and relaxation time vary in opposite directions under the transition from glassy
materials to liquids, viscoelasticity theory requires that the values of cohesive strength,
diff usion coefficient, and relaxation time should be in a specific and rather narrow range
of magnitude to make the value of their product and, consequently, adhesion as high
as possible. The specific value of the ratio between high cohesion energy and large free
volume, featured for the PSAs of various chemical composition and structure, may be
expressed in terms of the glass transition temperature, diff usion coefficient, relaxation
time, elasticity modulus, and loss tangent. The glass transition temperatures of PSAs fall
in the range between −10 and −115°C, the values of diff usion coefficients should be on
the border between those typical of elastomers and viscous liquids, ~10−13 m2/s, and the
appropriate values of the relaxation times are identified in the next chapter.
According to their large strain behavior, all PSAs can be classified into two groups.
The majority of PSAs belong to the group of low-modulus, soft polymer materials that flow
upon attainment of yield stress in the course of uniaxial drawing. The second group of
PSAs consists of comparatively high-modulus, cross-linked covalently (or noncovalently)
and highly ordered polymer materials that exhibit a rubber-like type of deformation
with pronounced strain hardening. Dahlquist’s criterion of tack, which states that the
elasticity modulus of the PSA must be in the range 0.01–0.1 MPa, holds for adhesives in
the first group, whereas the PSAs in the second group, such as thermoplastic elastomers
and PVP–PEG hydrophilic PSA, based on the network of hydrogen bonds, demonstrate
at 20°C the values of storage moduli in the range of 0.6–1.5 MPa, tan δ ≈ 1.2–2.2, and
loss tangent peak temperatures around 20°C.
The insight gained into the molecular structures responsible for the occurrence of
pressure-sensitive adhesion opens the way to the molecular design of new PSAs with
optimized performance properties by blending nonadhesive polymers. This approach is
illustrated in Technology of Pressure-Sensitive Adhesives and Products, Chapter 7.
References
1. Satas D. (Ed.) Handbook of Pressure-Sensitive Adhesive Technology, 3rd ed., Satas &
Associates, Warwick, RI, 1999.
2. Benedek I., Chemical basis of pressure sensitive products, in: Benedek I. (Ed.)
Developments in Pressure-Sensitive Products, 2nd ed., CRC-Taylor & Francis, Boca
Raton, 2006, chap. 5.
3. Benedek I., Physical basis of pressure sensitive products, in: Benedek I. (Ed.)
Developments in Pressure-Sensitive Products, 2nd ed., CRC-Taylor & Francis, Boca
Raton, 2006, chap. 3.
4. Dillard D.A. and Pocius A.V. (Eds.), The Mechanics of Adhesion, Elsevier, New York,
2002.
5. Dahlquist C.A., in: Patrick R.L. (Ed.), Treatise on Adhesion and Adhesives, vol. 2,
M. Dekker, New York, 1969, 219 p.
6. Creton C. and Leibler L., How does tack depend on time of contact and contact
pressure, J. Polym. Sci.: Part B: Polym. Phys. 34, 545–554, 1996.
7. Zosel A., Structure property relations in pressure sensitive adhesives: a review, in:
Proceedings of the 5th European Conference on Adhesion (EURADH’2000), SFV,
Lyon, France, 2000, pp. 149–153.
8. Feldstein M.M. and Creton C., Pressure-sensitive adhesion as a material prop-
erty and as a process, in: Benedek I. (Ed.) Pressure-Sensitive Design, Theoretical
Aspects, vol. 1, VSP, Leiden, 2006, chap. 2.
9. Berg J.C., Semi-empirical strategies for predicting adhesion, in: Chaudhury M.
and Pocius A.V. (Eds.) Adhesion Science and Engineering: Surfaces, Chemistry and
Applications, Elsevier, Amsterdam, 2002, chap. 1.
10. Zisman W.A., Adhesion and bonding, in: Mark H.F., Gaylord N.G., and Bikales
N.M. (Eds.) Encyclopedia of Polymer Science and Technology, Interscience Publishers
(J. Wiley & Sons), New York, 1964, pp. 445–477.
11. Feldstein, M.M., Molecular fundamentals of pressure-sensitive adhesion, in:
Benedek I. (Ed.) Developments in Pressure-Sensitive Products, 2nd ed., CRC-Taylor &
Francis, Boca Raton, 2006, chap. 4.
12. Creton, C., Pressure-sensitive-adhesives: an introductory course, MRS Bull. 28(6):
434–439, 2003.
13. Kinloch A.J., Adhesion and Adhesives: Science and Technology, Chapman & Hall,
London, 1987, chap. 3.
14. Voyutskii S.S., Autohesion and Adhesion of High Polymers, Wiley Interscience,
New York, 1963.
15. Voyutskii S.S., Adhesive Age 5(4), 30, 1962.
16. Voyutskii S.S., Markin Yu.I., Gorchakova V.M., and Gul V.E., How temperature
dependence and activation energy affects the adhesion of polymers to metals,
Adhesive Age 8(11), 24, 1965.
17. Voyutskii S.S., J. Adhesion 3, 69, 1971.
18. Vasenin R.M., Adhesion, Fundamentals and Practice, McLaren and Son, 1969,
p. 29.
19. Vasenin R.M., Adhesion of high polymers, Adhesive Age 8(5), 18, 1965.
20. Vasenin R.M., Adhesion of high polymers: predicting adhesion, Adhesive Age 8(6),
30, 1965.
21. Chalykh A.E., Diffusion in Polymer Systems, Chemistry, Moscow, 1987 (in Russian).
22. Chalykh A.E. and Shcherbina A.A., Transition zones in adhesive joints, in: Benedek
I. and Feldstein M.M. (Eds.) Fundamentals of Pressure-Sensitive Adhesion, Taylor
& Francis, Boca Raton, 2009, chap. 3.
23. Arslanov V.V. and Chalykh A.E., Current state and prospects of the development
of the theory of adhesion joints, Prot. Met. (USSR). 25(4), 425–431, 1989.
24. Creton C., Diffusion and adhesion, in: Benedek I. and Feldstein M.M. (Eds.) Funda-
mentals of Pressure-Sensitive Adhesion, Taylor & Francis, Boca Raton, 2009, chap. 2.
25. Deryaguin B.V. and Smilga V.P., Adhesion, Fundamentals and Practice, McLaren
and Son, London, 1969, p. 152.
26. Good R.J., Chaudhury M.K., and van Oss C.J., Theory of adhesive forces across
interfaces, in: Lee L.H. (Ed.) Fundamentals of Adhesion, Plenum Press, New York,
1991, chaps. 3 and 4.
27. Satas D., Tack, in: Satas D. (Ed.) Handbook of Pressure-sensitive Adhesive Technology,
3rd ed., Satas & Associates, Warwick, RI, 1999, chap. 4.
28. Bϋhler V., Kollidon®: Polyvinylpyrrolidone for the Pharmaceutical Industry, BASF,
Ludwigshafen, 1996, p. 20.
29. Feldstein M.M., Adhesive hydrogels: structure, properties and application, Polym.
Sci., Ser. A. 46(11), 1265, 2004.
30. Bairamov D.F., Chalykh A.E., Feldstein M.M., Siegel R.A., and Platé N.A., Dis-
solution and mutual diff usion of poly(N-vinyl pyrrolidone) in short chain
poly(ethylene glycol) as observed by optical wedge microinterferometry, J. Appl.
Polym. Sci. 85, 1128, 2002.
31. Painter P.C. and Coleman M.M., in: Paul D.R. and Bucknall C.B. (Eds.) Polymer
Blends, Vol. 1: Formulation, John Wiley & Sons, New York, 2000, chap. 4.
32. Bairamov D.F., Chalykh A.E., Feldstein M.M., and Siegel R.A., Impact of molecu-
lar weight on miscibility and interdiff usion between poly(N-vinyl pyrrolidone)
and poly(ethylene glycol), Macromol. Chem. Phys. 203(18), 2674, 2002.
33. Feldstein M.M., Lebedeva T.L., Shandryuk G.A., Kotomin S.V., Kuptsov S.A.,
Igonin V.E., Grokhovskaya T.E., and Kulichikhin V.G., Complex formation of
poly(N-vinyl pyrrolidone) with poly(ethylene glycol), Polym. Sci. 41A(8), 854, 1999.
34. Chalykh A.A., Chalykh A.E., Novikov M.B., and Feldstein M.M., Pressure-sensitive
adhesion in the blends of poly(N-vinyl pyrrolidone) and poly(ethylene glycol) of
disparate chain lengths, J. Adhesion 78(8), 667, 2002.
35. Feldstein M.M., Kuptsov S.A., Shandryuk G.A., Platé N.A., and Chalykh A.E.,
Coherence of thermal transitions in poly(N-vinyl pyrrolidone)–poly(ethylene
glycol) compatible blends. 3. Impact of sorbed water upon phase behaviour, Polymer
41(14), 5349, 2000.
36. Feldstein M.M., Lebedeva T.L., Shandryuk G.A., Igonin V.E., Avdeev N.N., and
Kulichikhin V.G., Stoichiometry of poly(N-vinyl pyrrolidone)-poly(ethylene glycol)
complex, Polym. Sci. 41A(8), 867, 1999.
37. Lebedeva T.L., Kuptsov S.A., Feldstein M.M., and Platé N.A., Molecular arrange-
ment of water associated with poly(N-vinyl pyrrolidone) in the first hydrate shell,
in: Iordanskii A.L., Starzev O.V., and Zaikov G.E. (Eds.) Water Transport in Syn-
thetic Polymers, Nova Science Publishers, Inc., New York, 2003, chap. 4.
38. Feldstein M.M., Shandryuk G.A., and Platé N.A., Relation of glass transition tem-
perature to the hydrogen-bonding degree and energy in poly(N-vinyl pyrrolidone)
blends with hydroxyl-containing plasticizers. Part 1. Effects of hydroxyl group
number in plasticizer molecule, Polymer 42(3), 971, 2001.
39. Feldstein M.M., Kuptsov S.A., Shandryuk G.A., and Platé N.A., Relation of glass
transition temperature to the hydrogen-bonding degree and energy in poly(N-
vinyl pyrrolidone) blends with hydroxyl-containing plasticizers. Part 2. Effects of
poly(ethylene–glycol) chain length, Polymer 42(3), 981, 2001.
40. Feldstein M.M., Roos A., Chevallier C., Creton C., and Dormidontova E.D., Rela-
tion of glass transition temperature to the hydrogen bonding degree and energy
in poly(N-vinyl pyrrolidone) blends with hydroxyl-containing plasticizers: 3.
Analysis of two glass transition temperatures featured for PVP solutions in liquid
poly(ethylene glycol), Polymer 44(6), 1819, 2003.
41. Gerasimov V.K., Chalykh A.A., Chalykh A.E., Razgovorova V.M., and Feldstein
M.M., Thermodynamic potentials of mixing in poly(N-vinyl pyrrolidone)–
Polyethylene glycol system, Polym. Sci. 43A(12), 2141, 2001.
42. Feldstein M.M., Shandryuk G.A., Kuptsov S.A., and Platé N.A., Coherence of
thermal transitions in poly(N-vinyl pyrrolidone)–poly(ethylene glycol) compat-
ible blends. 1. Interrelations among the temperatures of melting, maximum cold
crystallization rate and glass transition, Polymer 41(14), 5327, 2000.
43. Feldstein M.M., Kuptsov S.A., and Shandryuk G.A., Coherence of thermal transi-
tions in poly(N-vinyl pyrrolidone)–poly(ethylene glycol) compatible blends. 2. The
temperature of maximum cold crystallization rate versus glass transition, Polymer
41(14), 5339, 2000.
44. Feldstein M.M., Peculiarities of glass transition temperature relation to the
composition of poly(N-vinyl pyrrolidone) blends with short chain poly(ethylene
glycol, Polymer 42(18), 7719, 2001.
45. Li Y., Zhang R., Chen H., Zhang J., Suzuki R., Ohdaira T., Feldstein M. M., and
Jean Y.C., The depth profi le of free volume in a mixture and copolymers of poly(N-
vinyl-pyrrolidone) and poly(ethylene glycol) studied by positron annihilation
spectroscopy, Biomacromolecules 4, 1856, 2003.
46. Bairamov D.F., Chalykh A.E., Feldstein M.M., Siegel R.A., and Platé N.A., Mutual
diff usion of poly(N-vinyl pyrrolidone) and water, in: Iordanskii A.L., Starzev O.V.,
and Zaikov G.E. (Eds.) Water Transport in Synthetic Polymers, Nova Science
Publishers, Inc., New York, 2003, chap. 3.
47. Vartapetian R.S., Khozina E.V., Karger J., Geschke D., Rittig F., Feldstein M.M.,
and Chalykh A.E., Self-diffusion in poly(N-vinyl pyrrolidone)–poly(ethylene glycol)
system, Colloid Polym. Sci. 279(6), 532, 2001.
48. Vartapetian R.Sh., Khozina E.V., Karger J., Geschke D., Rittig F., Feldstein
M.M., and Chalykh A.E., Molecular dynamics in poly(N-vinyl pyrrolidone)–
poly(ethylene glycol) blends by pulsed field gradient NMR method: effects of
aging, hydration and PEG chain length, Macromol. Chem. Phys. 202(12), 2648,
2001.
49. Novikov M.B., Roos A., Creton C., and Feldstein M.M., Dynamic mechanical and
tensile properties of poly(N-vinyl pyrrolidone)–poly(ethylene glycol) blends, Polymer
44(12), 3559, 2003.
50. Feldstein M.M., Kulichikhin V.G., Kotomin S.V., Borodulina T.A., Novikov M.B.,
Roos A., and Creton C., Rheology of poly(N-vinyl pyrrolidone)-poly(ethylene
glycol) adhesive blends under shear flow, J. Appl. Polym. Sci. 100, 522, 2006.
51. Roos A., Creton C., Novikov M.B., and Feldstein M.M., Viscoelasticity and tack
of poly(N-vinyl pyrrolidone)–poly(ethylene glycol) blends, J. Polym. Sci., Polym.
Phys. 40, 2395, 2002.
52. Novikov M.B., Borodulina T.A., Kotomin S.V., Kulichikhin V.G., and Feldstein
M.M., Relaxation properties of pressure-sensitive adhesives upon withdrawal of
bonding pressure, J. Adhesion 81(1), 77, 2005.
53. Zosel A., Adhesive failure and deformation behaviour of polymers, J. Adhesion 30,
135, 1989.
54. Zosel A., The effect of fibrillation on the tack of pressure-sensitive adhesives, Int. J.
Adhes. Adhesives 18, 265, 1998.
55. Christensen S. F., Everland H., Hassager O., and Almdal K., Int. J. Adhesion Adhe-
sives 18, 131, 1998.
56. Chiche A., Zhang W. H., Stafford C. M., and Karim A., Meas. Sci. Technol. 16(1),
183, 2005.
57. Lakrout H., Sergot P., and Creton C., Direct observation of cavitation and fibril-
lation in a probe tack experiment on model acrylic pressure-sensitive adhesives,
J. Adhesion 69, 307, 1999.
58. Creton C. and Fabre P., Tack, in: Dillard D.A. and Pocius A.V. (Eds.) Adhesion
Science and Engineering, Vol. I: The Mechanics of Adhesion, Elsevier, Amsterdam,
2002, chap. 14.
59. Kaelble D.H., Theory and analysis of peel adhesion, in: Satas D. (Ed.) Handbook of
Pressure-sensitive Adhesive Technology, 3rd ed., Satas & Associates, Warwick, RI,
1999, chap. 6.
60. Ward I.M., Mechanical Properties of Solid Polymers, Wiley Interscience, London,
1983, chap. 2.
61. De Gennes P.G., J. Chem. Phys. 55, 572, 1971.
62. Fujita H., Fortsch. Hochpolym. Forsch., Bd. 3, 1, 1969.
63. Askadskii A.A. and Matveev Yu.I., Chemical Structure and Physical Properties of
Polymers, Chemistry, Moscow, 1983, pp. 24–48.
64. Dupasquier A. and Mills A.P., Jr. (Eds.) Positron Spectroscopy of Solids, IOS Press,
Amsterdam, 1995.
65. Feldstein M.M., Borodulina T.A., Vartapetian R.Sh., Kotomin S.V., Kulichikhin
V.G., Geschke D., and Chalykh A.E., Correlations between activation energy for
debonding and that for self-diff usion in pressure-sensitive adhesive hydrogels,
Proceed. 24th Annual Meeting Adhesion Soc. Williamsburg, VA, 137, 2001.
66. Kabanov V.A. and Zezin A.B., Makromol. Chem. (Suppl. 6), 259, 1984.
67. Kotomin S.V., Durability of adhesive joints, in: Benedek I. and Feldstein M.M.
(Eds.) Fundamentals of Pressure-Sensitive Adhesion, Taylor & Francis, Boca Raton,
2009, chap. 9.
68. Feldstein M.M., Significance of relaxation for adhesion, in: Benedek I. and Feld-
stein M.M. (Eds.) Fundamentals of Pressure-Sensitive Adhesion, Taylor & Francis,
Boca Raton, 2008, chap. 11.
69. Shandryuk G.A., personal communication.
70. Bessmertnaya L.Ya., Goncharova A.I., Rumsh L.D., Grokhovskaya T.E., and
Feldstein M.M., Thermostability and phase behaviour of enzymes in solid state
and within polymer matrices, Proceed. Intern. Symp. Control. Release Bioactive
Mater. 26, 383, 1999.
71. Ferguson J., Reilly B., and Granville N., Extensional and adhesion characteristics
of a pressure-sensitive adhesive, Polymer 38(4), 795, 1997.
72. Meissner B., Tensile stress–strain behavior of rubberlike networks up to break.
Theory and experimental comparison, Polymer 41, 7827, 2000.
73. Roos A. and Creton C., Effect of the presence of diblock copolymer on the non
linear elastic and viscoelastic properties of elastomeric triblock copolymers, Mac-
romolecules 38, 7807, 2005.
74. Creton C., Roos A., and Chiche A., Effect of the diblock content on the adhesive
and deformation properties of PSAs based on styrenic block copolymers, in: Possart
W.G. (Ed.) Adhesion: Current Research and Applications, Wiley-VCH: Weinheim,
2005, pp. 337–364.
75. Lindner A., Lestriez B., Mariot S., Creton C., Maevis T., Luhmann B., and
Brummer R., Adhesive and rheological properties of lightly cross-linked model
acrylic networks, J. Adhesion 82(3), 267, 2006.
76. Creton C., Block copolymers for adhesive applications, in: Matyjaszewski K.,
Gnanou Y., and Leibler, L. (Eds.) Structure-Property Correlation and Character-
ization Techniques, vol. 3, 1st ed., Wiley-VCH, Weinheim, 2007, pp. 1731–1752.
77. Satas D., Dynamic mechanical analysis and adhesive performance, in: Satas D.
(Ed.) Handbook of Pressure-sensitive Adhesive Technology, 3rd ed., Satas & Associ-
ates, Warwick, RI, 1999, chap. 10.
78. Chu S.G., Viscoelastic properties of pressure-sensitive adhesives, in: Satas D. (Ed.)
Handbook of Pressure Sensitive Adhesive Technology, 2nd ed., Van Nostrand Rein-
hold, New York, 1989, pp. 158–203.
79. Dale W.C., Paster M.D., and Haynes J.K., Mechanical property–performance
relations of acrylic pressure-sensitive adhesives, in: Satas D. (Ed.) Advances in
Pressure Sensitive Adhesive Technology, Satas & Associates, Warwick, RI, 1995,
pp. 65–111.
80. Benedek I., Pressure-Sensitive Adhesives and Applications, 2nd ed., Marcel Dekker,
New York, 2004, p. 9.
81. Derail C. and Marin G., Rheology of hot-melt PSAs: Influence of polymer struc-
ture, in: Possart W. (Ed.) Adhesion–Current Research and Application, Wiley-VCH,
Weinheim, 2005, chap. 16.
82. Courtesy of Dr. Y. Hu, National Starch and Chemical. Unpublished data.
83. Christensen S.F. and McKinley G.H., Int. J. Adhesion Adhesives 18, 333, 1998.
11-1
mat
nD (11.1)
exp
which is the ratio of the time scale of the material rearrangements, τmat, to the time scale
of experimental observation, τexp (see Ref. 10, p. 35). Even some apparent solids “flow”
if they are observed long enough. The origin of the name is the line “The mountains
flowed before the Lord,” in a song by prophetess Deborah recorded in the Bible (Judges
5:5). When the Deborah number approaches zero, the material may appear purely
viscous to the observer and purely elastic when it approaches infi nity. Real materials
fall in between and are viscoelastic. In particular, polymeric materials have Deborah
numbers around unity and are the viscoelastic materials par excellence. In fact, methods
for dealing with time dependence in mechanical properties have developed largely with
the development of polymeric materials.
Time-dependent mechanical properties are traditionally characterized in terms of
so-called response times and almost always by a distribution of such times.1 If the mate-
rial is strained to a fi xed value, which is then held constant, the corresponding stress
relaxes and the response times are called relaxation times. On the other hand, if a stress
is applied and kept fi xed, the strain is retarded and the response times are called retarda-
tion times. Because the context nearly always makes it clear which are under discussion,
both relaxation and retardation times are customarily designated by the same symbol, τ.
However, the relaxation and the retardation times are not identical. When ranking
both sets in ascending or descending order, they alternate (i.e., the retardation times are
intercalated between the relaxation times).10
Each response time is associated with a corresponding spectral strength that may be
a modulus, Gi, or a compliance, Ji. The time dependence of a material is thus revealed in
a finite, discrete set of response times and their associated spectral strengths. When a
shear strain is fi xed and a stress relaxation occurs, this set is {Gi, τi},3,11
in
Gt Geq ∑ Gi exp(t/i ) (11.2)
i1
where Geq is the equilibrium relaxation modulus, τi is the-relaxation time, and Gi is the
relaxation modulus associated with τ.
When a stress is removed and a strain recovery occurs, this set is {Ji, τi},3,11
in
J J 0 ∑ J i (1 et / i ) (11.3)
i1
where Ji is the compliance (Pa−1) in the i element of a structure and τi is the retardation
time (s). As t → ∞, J0 → 0. The corresponding value of relaxation modulus can be evalu-
ated with Equation 10.3 as the reciprocal of the compliance, Gi = 1/Ji.
Most well-known examples of time dependence in pressure-sensitive adhesion are
the effects of contact bonding time12–15 and debonding velocity16,17 on the strength of
adhesive joints. The former effect is often explained by the molecular diff usion of the
PSA into the substrate.18–20 However, as Equation 10.6 (Chapter 10) predicts, both dif-
fusion and relaxation processes contribute to the work of viscoelastic deformation of
the adhesive fi lm and failure of the adhesive bond that control peel adhesion.21 In the
following discussion we focus on this contribution.
The well-known Dahlquist’s criterion of adhesion specifies that the values of elas-
ticity moduli of various PSAs should be in the range between 0.01 and 0.1 MPa.22 As
demonstrated previously, 22,23 this criterion is incapable of predicting adhesive behavior
from the linear elasticity data at small strains for the PSAs, which are either cross-
linked (covalently or physically) or based on ordered supramolecular structures. One
such PSA is a hydrogen-bonded stoichiometric network complex of high-molecular-
weight poly(N-vinyl pyrrolidone) (PVP) with a short-chain polyethylene glycol (PEG),
which serves in our study as a model PSA.24,25 Another type of PSA departing from
Dahlquist’s criterion (although to a less pronounced extent) are the adhesives based on
diblock and triblock copolymers of styrene and isoprene, which are physically cross-
linked by the glassy domains composed of polystyrene chains with a higher glass tran-
sition temperature (Tg; see Technology of Pressure-Sensitive Adhesives and Products,
Chapter 3). For both such PSAs the large strain elongational properties, involving the
orientation of polymer chains, cannot be readily predicted from the linear elastic prop-
erties in shear at small strains. As demonstrated in Chapter 10, 22 for these PSAs, it is the
large strain modulus (the derivative of the stress–strain curve at strains above 100%)
that is in agreement with Dahlquist’s criterion, rather than the linear elastic properties
at small strains.
The implication of this fact is that the supramolecular structure of PVP-PEG and
styrene–isoprene–styrene (SIS) adhesives under an applied detaching stress under-
goes transformation in such a way that the behavior typical of a PSA is a result of the
large deformation behavior. This means that the interpretation of the phenomenon of
pressure-sensitive adhesion solely as an equilibrium material property cannot always
be adequate and, to gain further insight into this phenomenon, adhesion should be also
treated as a process.
Adhesion is traditionally defined as the phenomenon in which surfaces of contacting
materials are held together by interfacial forces.26 In this chapter, the term phenomenon
is treated as any event, series of experimental facts, or experience that is detectable by
our senses and measuring instruments and that can be scientifically described. In turn,
the term property means a characteristic quality regarded as being possessed by a mate-
rial or a group of materials and that is common to all the members of this group. Process
is defined as a continuing development involving many changes or transformations of
the structure and properties of the material. Within this framework, the phenomenon is
a conjunction of the property and the process.
In other words, the term property can be defi ned as an equilibrium feature of the
molecular structure independent of time. To reduce the concept of pressure-sensitive
adhesion to such a defi nition would be a gross simplification, because adhesion is a
phenomenon that is due to a nonequilibrium process. Th is phenomenon consists of a
series of transformations of the structure of the adhesive material under an applied
bonding and detaching stress, that is, it involves the process of evolution of mate-
rial structure, geometry (e.g., cavitation and formation of fibrils), and properties. The
order and arrangement in time of these transformations has great importance for
the perception of the phenomenon as a whole. In the following section we present a
rheological and structural description of all stages of the process of pressure-sensitive
adhesion.
1 I II III
Compression
(bonding)
F
F F=0
h (mm)
0.1
Relaxation
Debonding
FIGURE 11.1 Typical protocol of squeeze–recoil testing of PVP–PEG (36%) adhesive fi lm.
Compressive and debonding forces are 0.2 N. (From Feldstein, M.M., in Benedek, I. Ed., Develop-
ments In Pressure Sensitive Products, 2nd ed., Chap. 4, CRC-Taylor & Francis, Boca Raton, 2006.)
and the third (debonding) stages cannot be studied separately if the adhesive joint in the
course of the first stage is never formed. The PSA behavior at each subsequent stage is
affected by the conditions of earlier stages, providing a “memory effect.” In other words,
at each stage of the process the PSA “remembers” the way as an earlier stage has been
performed. In fact, only the second stage, relaxation, does not always occur if compres-
sive bonding load is immediately followed by application of a detaching force, avoiding
the stage of PSA relaxation (II; Figure 11.1). As illustrated below, the third stage, adhe-
sive debonding, is a multistage process itself, as evidenced by structural and geometric
transformations of adhesive material in the course of a probe tack test.
In the following discussion we describe the relaxation properties of PSAs at each
subsequent stage of the process and compare them with adhesive behavior to identify
the values of relaxation times and corresponding moduli that produce strong pressure-
sensitive adhesion.
t (s)
σ (MPa)
Compression Debonding
(a)
t (s)
σ (MPa)
Compression
Debonding
Relaxation
(b)
FIGURE 11.2 Nominal stress versus time (displacement) curves of a typical probe test. (a) The
compressive stress remains constant during the contact time and (b) the strain of the adhesive
layer is kept constant during the contact time and the adhesive is allowed to relax in the course
of bond formation.
Of course, the more a PSA is viscoelastic and the thicker the layer is, the more the details
of the bonding stage will affect the debonding stage.
0.8
500 s
0.6
300 s
σ (MPa)
0.4
50 s
0.2 1s
0.0
0 2 4 6
ε
FIGURE 11.3 The effect of contact time on the curves of probe separation for the acrylic adhe-
sive Gelva 3011. Bonding pressure is constant (0.8 MPa) and debonding velocity is 0.1 mm/s.
dependence of the debonding process on the layer thickness itself.30 The relaxation
curves for PIB (Oppanol B15), silicone (BIO-PSA® 7-4302), and SIS-based Duro-Tak®
34-4230 PSAs are illustrated in Figure 11.4.31 By fitting the exponential relaxation curves
in Figure 11.4 with the sum of exponents (Equation 11.2), it has been established that
three populations of relaxation times and corresponding moduli can approximate the
curves of stress relaxation in Figure 11.4. Values are presented in Table 11.1. For all adhe-
sives examined, the following ranges of relaxation times and moduli have been estab-
lished: Geq = 0–0.6 MPa, G1 = 0.08–0.8 MPa, τ1 = 1–6 s, G 2 = 0.06–0.5 MPa, τ2 = 5–50 s,
G3 = 0.02–0.2 MPa, τ3 = 30–800 s.
Figures 11.5 through 11.8 illustrate the effects of contact bonding time on the tack
behavior for a variety of PSAs.31 All employed adhesives can be classified into two
0.35
0.30
0.25
0.20
σ (MPa)
SIS
0.15
0.10
0.05 Silicone
PIB
0.00
0 20 40 60 80 100 120 140 160 180 200
t (s)
FIGURE 11.4 Compressive stress relaxation curves for PIB (Oppanol B15), silicone (BIO-PSA
7-4302), and SIS-based (Duro-Tak 34-4230) adhesives.
TABLE 11.1 Relaxation Times and Corresponding Moduli for Acrylic (Gelva 3011), PIB
(Oppanol), Silicone (BIO-PSA 7-4302), and SIS-Based (Duro-Tak 34-4230) Adhesives
Obtained by Fitting the Compressive Stress Relaxation Curves with Equation 11.2
Adhesive
Type Geq (MPa) G1 (MPa) τ1 (s) G2 (MPa) τ2 (s) GB (MPa) τ3 (s)
BIO-PSA 0.025 ± 0.0007 0.8 ± 0.02 3.3 ± 0.1 0.5 ± 0.02 21.4 ± 0.8 0.2 ± 0.005 150 ± 5.4
7-4302
Oppanol 0 0.57 ± 0.01 1.2 ± 0.03 0.4 ± 0.01 7.2 ± 0.2 0.07 ± 0.003 48 ± 1.8
B15
Oppanol 0 0.2 ± 0.01 1.8 ± 0.04 0.17 ± 0.01 5 ± 0.4 0.02 ± 0.005 35 ± 3.4
B12
Duro-Tak 0.6 ± 0.005 0.2 ± 0.006 5.3 ± 0.3 0.1 ± 0.004 44.5 ± 2.9 0.2 ± 0.004 770.7 ± 56.1
34-4230
Gelva 3011 0.035 ± 0.0006 0.08 ± 0.003 5.1 ± 0.3 0.06 ± 0.002 43.4 ± 3.04 0.05 ± 0.001 355.7 ± 20.3
0.8
0.6
0.4
σ (MPa)
0.2
0.0
−0.2
−0.4
0 200 400 600 800
Time (s)
FIGURE 11.5 Effect of contact time on probe tack curves of the silicone adhesive BIO-PSA
7-4302. The initial bonding pressure is 0.8 MPa and the velocity of probe separation is 0.1 mm/s.
0.15
0.10
0.05
σ (MPa)
0.00
−0.05
−0.10
−0.15
0.3
0.2
0.1
σ (MPa)
0.0
−0.1
−0.2
−0.3
0 50 100 150 200 250 300 350 400
Time (s)
FIGURE 11.6 Kinetics of nominal compressive stress relaxation during adhesive bond forma-
tion, followed by the debonding process for PIB Oppanol B12 (top) and Oppanol B15 (bottom).
0.3
0.2
0.1
σ (MPa)
0.0
−0.1
−0.2
−0.3
0 200 400 600 800 1000 1200
Time (s)
FIGURE 11.7 Probe tack curves of Duro-Tak 34-4230 obtained under different contact times.
0.12
0.08
0.04
Stress (MPa)
0.00
−0.04
−0.08
−0.12
0 200 400 600 800 1000 1200
Time (s)
FIGURE 11.8 Effect of contact time on probe tack of acrylic adhesive Gelva 3011.
groups: (1) fully relaxing PIB adhesives (Figure 11.6), such as Oppanol B15 and Oppanol
B12 PSAs, described in Technology of Pressure-Sensitive Adhesives and Products,
Chapter 4, and (2) adhesives that are able to store energy in the course of deformation
and exhibit a residual (unrelaxed) stress during the contact time. SIS-based (Duro-Tak-
34-4230, Figure 11.7) and, to a somewhat lesser extent, the acrylic Gelva® 3011 PSA
(Figure 11.8), belong to the second group. Silicone adhesive BIO-PSA 7-4302 (Figure 11.5)
takes an intermediate position.
As a rule, the appearance of the second maximum on probe tack curve results from
a network structure of adhesive material. For instance, in the SIS-based Duro-Tak-34-
4230 adhesive, the network structure is provided by the glassy domains of polystyrene
blocks. Polymer networks frequently reveal an apparent yield stress that is defi ned as
the minimum pressure at which the material flows extremely slowly under an applied
stress.28 Such behaviors are typical for SIS-based Duro-Tak 34-4230 (Figure 11.7) and,
to a less pronounced extent, for acrylic Gelva 3011 (Figure 11.8) adhesives. As follows
from the data in Table 11.1, 31 the relaxation times of SIS DURO-Tak 34-4230 and acrylic
Gelva 3011 PSAs are appreciably higher than those for PIB Oppanol B12, Oppanol B15,
and silicone BIO-PSA 7-4302. Thus, the longer relaxation processes are mostly associ-
ated with the appearance of a pronounced plateau on the debonding curves (BIO-PSA
7-4302, Figure 11.5), or even a second maximum (SIS Duro-Tak 34-4230, Figure 11.7,
and acrylic Gelva 3011 PSA, Figure 11.8). Fully relaxing, soft viscous adhesives such as
Oppanol B12 or B15 have low relaxation times and a liquid-like debonding mechanism
typical of fluid adhesives. The values of characteristic modulus associated with shortest
relaxation times, G1, are generally higher than G 2 and G3. The equilibrium relaxation
modulus, Geq, is a direct measure of the stored elastic energy in a polymeric material.
Fluid polyisobutene (PIB) adhesives (Oppanol B12 and Oppanol B15) do not reveal any
apparent yield stress and, consequently, no equilibrium modulus.
The complex geometry of the probe tack test does not allow us to present a straight-
forward physical meaning of the two shorter relaxation time values, because in addi-
tion to the viscoelastic response of the adhesive material and small-scale recovery of
the material structure, these values account for the formation of adhesive contact and
instrument compliance. However, let us take into consideration that the longer relax-
ation time has been earlier demonstrated to contribute to high adhesive strength (see
Equation 10.6 in Chapter 10).
1.0
700
W
650 0.8
σmax (MPa)
W (J/m3)
600 σmax
0.6
550
0.4
500
450 0.2
0 100 200 300 400 500
Time (s)
FIGURE 11.9 Debonding energy (W) and maximum stress values (σmax) versus the contact time
for silicone adhesive BIO-PSA 7-4302.
120
Oppanol B12
100
80 Oppanol B15
W (J/m2)
60
40
20
0 50 100 150 200 250 300
Time (s)
FIGURE 11.10 Practical work of adhesion (W) versus the contact time for PIB Oppanol B12 and
Oppanol B15 adhesives.
than a true relaxation of the material per se, although the relaxation process contrib-
utes to the driving force for such annealing. At best, this example is illustrative of the
fact that, in general, it is not wise to ignore the effect of the bonding stage when analyzing
the debonding stage. This is particularly true when the PSA is very fluid-like with an
important viscous component.
0.40
550
0.36
W 500
0.32
Stress (MPa) 450
W (J/m2)
0.28
400
0.24 350
200
0 200 400 600 800 1000
Time (s)
FIGURE 11.11 Effect of contact time on the work of debonding and the values of two stress
peaks for SIS-based Duro-Tak 34-4230 PSA.
0.15 1000
W
0.14 900
0.13
Stress (MPa) 800
W (J/m2)
0.12
700
Peak 2
0.11
600
0.10 Peak 1
0.09 500
0.08 400
0 200 400 600 800 1000
Time (s)
FIGURE 11.12 The contact time dependence of the work of debonding and the values of two
stress peaks on probe tack curve for acrylic Gelva 3011 adhesive.
0.25
W 700
0.20
0.15
W (J/m2)
σ (MPa)
600
Intermediate
0.10 Slow
Fast
0.05
500
0.00
0 200 400 600 800 1000
Time (s)
FIGURE 11.13 Effect of contact time on the practical work of adhesion compared with the
bonding stress relaxation curve for silicone BIO-PSA 7-4302 adhesive.
the process of slow relaxation (since 125–150 s). Shorter relaxation times dominate within
the period of fast relaxation, whereas the longer relaxation times govern the steady-state
(equilibrium) adhesion of both types of PSAs. In both cases, achieving maximum adhesion
falls the end of the intermediate relaxation period and the onset of slow relaxation.
Thus, in full agreement with Equation 10.6 presented in the Chapter 10, high adhe-
sion is associated with longer relaxation time. The relaxation mechanism provides the
links between all the stages of the process of pressure-sensitive adhesion.
600
W
0.20
500
W (J/m2)
σ (MPa)
400
0.15
Intermediate
Fast
300
Slow
0.10 200
0 200 400 600 800 1000
Time (s)
FIGURE 11.14 Comparison of the bonding stress relaxation curve with the change in the prac-
tical work of adhesion as a function of contact time for SIS-based Duro-Tak 34-4230 PSA.
900 900
800 τ 800
700 700
600 600
W (J/m2)
500 W 500
τ3 (s)
400 400
300 300
200 200
100 100
0 0
23 k ®
11 a ®
02 A ®
l
5 no
2 no
− 4 -Ta
30 elv
B1 ppa
B1 ppa
43 PS
0
G
34 uro
7− IO-
O
B
D
FIGURE 11.15 Longer relaxation times and practical work of adhesion for SIS-based Duro-Tak
34-4230, acrylic Gelva 3011, silicone BIO-PSA 7-4302, and two grades of PIB adhesives (Oppanol
B12 and B15). Observation time is 1000 s.
800
700 SIS
600
500
τ3 (s)
400 Acrylic
300
200
100 Silicone
FIGURE 11.16 Effect of the observation window on the values of longer relaxation time, τ3, for
SIS-based Duro-Tak 34-4230, acrylic Gelva 3011, and silicone BIO-PSA 7-4302 adhesives.
60
55
50
45
Oppanol B15
40 nD = 0,06
3 (s)
35
30
Oppanol B12
25
20
15
100 150 200 250 300
t obs (s)
FIGURE 11.17 Effect of the observation window on the values of longer relaxation time, τ3, for
PIB Oppanol B12 and B15 adhesives.
TABLE 11.2 Values of Deborah Numbers, nD, Determined for SIS-Based (Duro-Tak 34-4230),
Acrylic (Gelva 3011), Silicone (BIO-PSA 7-4302), and PIB (Oppanol B15) Adhesives by Treating
the Relationships between the Longer Relaxation Times and Observation Times
tend to deviate from the linear relationship for the most elastic, cross-linked SIS Duro-Tak
34-4230 and acrylic Gelva 3011 adhesives. Interestingly, however, the slopes of the linear
parts of the curves presented in Figures 11.16 and 11.17 are constant and can be related to
the Deborah number, nD, of the PSAs. The values of the Deborah numbers are listed in
Table 11.2, along with parameters of linear regression, R2.
All PSAs are viscoelastic materials that couple the properties of solids and liquids.
Values of the Deborah number are informative regarding the liquid-like and solid-like
contributions. As nD = 1, a material is 50% solid and 50% liquid. Th is is the case of phys-
ically cross-linked SIS-based adhesive characterized by the magnitude of the Deborah
number that approaches unity (Table 11.2). For all other adhesives considered, the
liquid-like viscous contribution dominates the solid-like elastic contribution. Whereas
the elastic contribution is appreciable for the chemically cross-linked acrylic Gelva 3011
adhesive (nD = 0.388), silicone BIO-PSA 7-4302 and PIB adhesives are typical viscous
liquids. For most liquid-like, lower-MW PIB fractions (Oppanol B12), the fitting of
stress relaxation curves with Equation 11.2 at observation times less than 200 s does not
yield plausible values of relaxation times (Figure 11.17). Cross-linked adhesives (SIS-
based and acrylic), as well as silicone PSA, which exhibit the best adhesion, demonstrate
Deborah numbers between 0.15 and 1.
The values of the Deborah number described in this section are only approximate
and are calculated with longer relaxation times. They are only used to give an idea of the
type of behavior (liquid or solid) of the PSA examined. In reality, the spectrum of the
relaxation times in the linear viscoelastic region is a well-defined material property that
does not depend on observation time. However, such viscoelastic properties as moduli
G′ and G″ depend on both the relaxation time and the observation time.
0.45 44
W 40
0.40 σmax
36
0.35
σmax (MPa)
W (J/m2)
32
0.30
28
0.25
24
0.20 20
24 28 32 36 40 44
TEC (wt %)
FIGURE 11.18 Values of maximum stress and the practical work of adhesion versus plasticizer
(TEC) concentration for a model PSA adhesive based on a polybase–polyacid complex. Contact
time, 1 s. Debonding rate, 0.1 mm/s. (From Novikov M.B., Kiseleva T.I., Anosova J.V., Singh P.,
Cleary G.W., and Feldstein M.M., Proceed. 30th Annual Meeting Adhesion Soc. Tampa, FL, 2007.)
1.6
25% TEC
1.4
30% TEC
1.2
σ (MPa)
1.0
0.8
35% TEC
0.6
40% TEC
0.4
45% TEC
0.2
FIGURE 11.19 Relaxation curves obtained in the course of adhesive joint formation for the
blends of an interpolymer complex with different amounts of plasticizer (TEC). Contact time,
200 s. (From Novikov M.B., Kiseleva T.I., Anosova J.V., Singh P., Cleary G.W., and Feldstein M.M.,
Proceed. 30th Annual Meeting Adhesion Soc. Tampa, FL, 2007.)
160
140
120
100
Time (s)
80
60
40
20
0
25 30 35 40 45
TEC (wt %)
processes within the polymer system. However, using the exponential Equation 11.2 and
the Kelvin–Voigt model, we are able to define the short and large-scale relaxation pro-
cesses within this time region. As illustrated in Figure 11.19, there is a sharp transition
in the relaxation behavior of the blends that contain 30 and 35 wt % TEC. Although
the interpolymer complexes containing 25 and 30 wt % TEC demonstrate a highly pro-
nounced residual stress on the relaxation curves, which is a characteristic feature of cross-
linked and ordered structures, the relaxation curves of the blends with 35 wt % TEC and
more demonstrate a gradual decrease in stress that is rather typical of viscous liquids.
Adequate fitting of the relaxation curves in Figure 11.19 with Equation 11.2 is possible
using a sum of three exponents. The effect of TEC concentration on relaxation times is
illustrated in Figure 11.20. The longer relaxation time is a decreasing function of plasti-
cizer content. Accordingly, the values of the equilibrium relaxation modulus reduce with
the increase in TEC concentration (Table 11.3). Faster relaxation processes, τ1 and τ2,
are unaffected by the concentration of the plasticizer (Figure 11.20) that controls the
adhesive properties (compare with the data in Figure 11.18). Based on this observation, a
logical deduction can be drawn that the large-scale relaxation processes, characterized
by the value of the longer relaxation time, contributes more to PSA performance.
0.8
0.4
0.0
Stress (MPa)
−0.4
−0.8
−1.2
−1.6
0 50 100 150 200
Time (s)
FIGURE 11.21 Effect of contact time on the curves of bonding stress relaxation, followed by
probe separation from adhesive fi lm surface under detaching force for the model PSA based on an
interpolymer polybase–polyacid complex containing 35 wt % TEC.
25% 35%
0.6
30%
0.5
40%
σmax (MPa)
0.4
0.3 45%
0.2
FIGURE 11.22 Effect of contact time on the maximum values of probe detaching stress, σmax,
for a plasticized interpolymer polyacid–polybase complex. (From Novikov M.B., Kiseleva T.J.,
Anosova J.V., Singh P., Cleary G.W., and Feldstein M.M., Proceed. 30th Annual Meeting Adhesion
Soc. Tampa, FL, 2007.)
100 35%
30%
80
W (J/m2)
40%
60 25%
45%
40
20
FIGURE 11.23 Effect of contact time on the practical work of adhesion (W) for the plasticized
interpolymer complex. (From Novikov M.B., Kiseleva T.J., Anosova J.V., Singh P., Cleary G.W., and
Feldstein M.M., Proceed. 30th Annual Meeting Adhesion Soc. Tampa, FL, 2007.)
to note that the dependence of adhesion parameters on contact time for the blend that
contains 45 wt % TEC is less pronounced than that for blends with a lower plasticizer
content. Indeed, the relaxation of the most fluid blend containing 45 wt % TEC occurs
much faster than for other blends. The values of practical work of adhesion for the blend
containing 45 wt % TEC are much lower than for the blends containing 30 or 35 wt %
TEC (Figure 11.23). On the other hand, the interpolymer complexes with still lower
160
100 Contact time 200 s
60
80
Contact time 1 s
40
40
20
25 30 35 40 45
TEC (wt %)
FIGURE 11.24 Effect of TEC content on the values of longer relaxation time, τ3, and practical
work of adhesion, W, for the model PSA based on an interpolymer complex.
3
900
80
800
W (J/m2)
500
40 W
400
300
20 200
2
100
1
0 0
e
30 lva ic
S
B
87 uro ylic
on
G cryl
PI
SI
00 k
lic
-9 -Ta
D Acr
A
A
Si
e
11
FIGURE 11.25 Comparison of relaxation times at observation time of 200 s and practical work
of adhesion for a series of commercial adhesives (SIS, silicone, cross-linked acrylic Gelva 3011,
un-cross-linked acrylic Duro-Tak® 87-900A, and PIB).
longer retardation times (150–800 s) are significant for high adhesion of various PSAs
and they relate mainly to the energy-dissipating processes and chain entanglements,
which in turn are associated with translational movement (self-diff usion) of polymer
segments and entire macromolecules in the course of large-scale structural rearrange-
ments. The minimum values of longer relaxation times are typical of fluid adhesives.
Whether chemically or physically cross-linked, network adhesives reveal much greater
values of longer relaxation times. Compressive stress relaxation in the course of adhe-
sive bonding defines the mechanism of adhesive joint failure during debonding under
a tensile detaching force. The adhesives exhibiting complete stress relaxation debond
mainly as fluids, with or without a pronounced plateau on the probe tack stress–strain
curves. In contrast, network adhesives, such as SIS and covalently cross-linked acrylic
PSAs, are capable of storing mechanical energy during the bonding stage. They demon-
strated the occurrence of residual, unrelaxed stress and typically have two peaks on the
debonding stress–strain curves. For all PSAs, the practical work of adhesion achieves
its maximum value as the contact time becomes comparable with the longer relaxation
time or, more precisely, as the mechanism of slow relaxation is rendered dominating.
If the stress during adhesive bonding is not allowed to relax, the mechanism and the
energy of debonding are independent of the contact time. However, if the bonding stress
can relax, the contribution of the contact time to the work of adhesive debonding is
appreciable. Correlation between the adhesion and relaxation time for all PSAs exam-
ined can be described fairly reasonably with Equation 10.6 of Chapter 10, which relates
peel adhesion to the relaxation time and translational mobility of adhesive polymer.
Sample Composition
SIS + I SIS Vector 4111 (57% wt)/Isolene 400 (43% wt)
SIS + R SIS Vector 4111 (50% wt)/Regalite R9110 (50% wt)
SIS + R + I SIS Vector 4111 (36.4% wt)/Regalite R9110 (27.2% wt)/Isolene 400 (36.4% wt)
B + PIB BR 065 (60% wt)/PIB Vistanex LM-MH (40% wt)
Viscous flow
1a
3b
Strain / deformation 2a 2b
Elastic recovery
3a
1b
t1 Time
FIGURE 11.26 Typical view of the squeeze–recoil profi le according to Burger’s model. (From
Novikov, M.B. et al. J. Adhesion, 81, 77–107, 2005.)
0.00 SIS
−0.04
PVP-PEG 36%
h − h0 /h0
−0.08
−0.12
−0.16
FIGURE 11.27 Squeeze–recoil profi les for SIS and PVP blends with 36% PEG under stepwise
increasing compressive force of 0.5, 1, 2, and 5 N, respectively. (From Novikov, M.B. et al. J. Adhe-
sion, 81, 77–107, 2005.)
When the compressive stress is removed, Burger’s model recovers in a two-step manner:
1b. Strain reduces instantaneously by the elastic response.
2b–3b. Strain reaches a value that is equal to the permanent, nonrecoverable strain
and represents the viscous flow of the Maxwell dashpot.
Agreement between the behaviors of the idealized Burger model (Figure 11.26) and
real adhesives is illustrated in Figure 11.27 and provides qualitative evidence that the
examined adhesives behave like linear viscoelastic systems, at least for intermediate
times, implying the applicability of the squeeze–recoil test for characterization of the
relaxation properties of adhesives upon the removal of compressive force.
0.0
31% PEG
36% PEG
−0.1
h − h0 /h0
−0.3
FIGURE 11.28 Effect of PEG content on squeeze–recoil profi les of PVP–PEG blends under
stepwise increasing compressive force of 0.5, 1, 2, and 5 N. (From Novikov, M.B. et al. J. Adhesion,
81, 77–107, 2005.)
PEG content, the greater the contribution of plastic deformation. Th is then requires
a longer time to recover the equilibrium thickness of the hydrogel upon removal of
compressive stress, indicating that the retardation time increases with the rise in PEG
concentration.37
0.08
0.07 5N
0.06
2N
0.05
h − h0 /h0
0.04
1N
0.03
0.02
0.01
0.00
0 100 200 300 400 500 600 700 800 900 1000
Time (s)
FIGURE 11.29 Impact of compressive force on the kinetics of strain recovery upon the removal
of compressive force for PVP–PEG (36 wt %) adhesive. (From Novikov, M.B. et al. J. Adhesion, 81,
77–107, 2005.)
0.35
0.30
41% PEG
0.25
39% PEG
h − h0 /h0
0.20
0.15
0.00
0 500 1000 1500 2000 2500 3000
Time (s)
FIGURE 11.30 Effect of PEG concentration on the kinetics of strain recovery for PVP–PEG
blends. (From Novikov, M.B. et al. J. Adhesion, 81, 77–107, 2005.)
6 100
5
80
4
60
G1 (MPa)
τ1 (s)
3
40
2
20
1
0 0
30 32 34 36 38 40 42
PEG content (%)
FIGURE 11.31 Shorter retardation time and corresponding modulus as a function of the con-
centration of plasticizer (PEG) in PVP–PEG adhesive blends. The data are averaged for two values
of compressive force (1 and 2 N). (From Novikov, M.B. et al. J. Adhesion, 81, 77–107, 2005.)
The coefficients of regression with Equation 11.3 (G1 and G2 moduli) for the squeeze–
recoil profiles, illustrated in Figures 11.29 and 11.30, are illustrated in Figures 11.31
and 11.32 as the functions of PEG content in adhesive blends with PVP. Two reliably
different retardation times are determined for the PVP–PEG adhesives, which differ
5 600
4 500
3 400
G2 (MPa)
τ2 (s)
2 300
1 200
0 100
30 32 34 36 38 40 42
PEG content (%)
FIGURE 11.32 Longer retardation time and corresponding modulus plotted versus PEG
concentration in PVP–PEG adhesive blends. The data are averaged for two values of compressive
force (1 and 2 N). (From Novikov, M.B. et al. J. Adhesion, 81, 77–107, 2005.)
in their magnitudes by about 1 decade: the shorter time, τ1, is in the range of 10–110 s
and the longer time, τ2, is ∼120–950 s. Within the framework of Burger’s model of visco-
elasticity, the shorter retardation or relaxation time is mainly associated with an elastic
contribution of the spring element into strain recovery, whereas the longer time charac-
terizes the behavior of a coupled dashpot and spring elements of the model and the rate
of strain recovery during the dissipation of the initially applied energy. The mechanism
of the latter process involves large-scale rearrangement of the structure of the polymeric
material via long-range motion (diff usion) of polymer segments and entire macromole-
cules. This process takes normally a much longer time than the elastic recovery of polymer
chain conformations. The longer process may also be associated with the entanglements
between polymer chains. The shorter retardation time relates mainly to the restoration of
the original conformation of polymer segments between entanglements.
As demonstrated in Figures 11.31 and 11.32, with increasing PEG concentration in
blends, the shorter retardation time is nearly constant, whereas the longer time increased
gradually and the relevant values of corresponding moduli decreased gradually with
PEG content. The increase in the longer retardation time as a function of PEG content
is also easily observable in Figure 11.28 and reflects the slowing down of strain recovery
upon the removal of compressive force.
Under relatively moderate compressive forces the retardation times and correspond-
ing moduli are practically independent of the applied compressive stress.37 However, at
comparatively high shear stresses (compressive force of 5 N and higher) both the retar-
dation times and the corresponding moduli tend to increase. The higher the stress, the
larger the molecular rearrangements occurring in the strained material and the longer
the time required for relaxation. The values of retardation times and moduli at relatively
moderate compressive forces of 1 and 2 N can be considered material characteristics.
G1, τ1
5 31
31 G2, τ1
4 34
31 34
31
36
G (MPa)
3
34 31 36
36
2
34
41 41
36
1 39 39
39
39
41
0
0 100 200 300 400 500
τ (s)
FIGURE 11.33 The relationship between retardation times and the corresponding moduli for
PVP–PEG blends. The PEG contents (wt %) are indicated. (From Novikov, M.B. et al. J. Adhesion,
81, 77–107, 2005.)
In a PSA formulation of that type, the rubbery polymer provides the elastic component,
whereas a low-molecular-weight tackifying resin and a plasticizer constitute the viscous
components. Most parent elastomers per se do not have the proper rheology to be PSAs.
Typically, the addition of a tackifier raises the glass transition temperature, Tg, lowers the
plateau modulus by diluting the chain entanglements of the elastomer, and increases
the ratio of viscous to elastic response of the elastomer/tackifier blend, improving
both the bond-making and the bond-breaking processes. Plasticizers demonstrate simi-
lar effects on rheology, but cause reduction in Tg. It is, therefore, of particular interest to
trace how the formulation process affects the relaxation properties of a composite PSA.
In this work we use a SIS block copolymer and BR as base elastomers. A hydrocarbon
resin (Regalite R9100, R) has been employed as a tackifier. R is a partially hydrogenated
resin with a specific balance of aliphatic and aromatic groups. As a plasticizer for
the SIS, a low-molecular-weight polyisoprene rubber, Isolene (I), has been used. The
pressure-sensitive adhesion in BR is provided by mixing the BR with a low-molecular-
weight PIB (Vistanex). The compositions of the samples examined in this work are
presented in Table 11.4.
As evident from the data in Figures 11.34 and 11.35, mixing elastomers (SIS and BR)
with plasticizers (I for SIS and low-molecular-weight PIB for BR) and tackifying resin (R)
results in an appreciable increase in retardation times (Figure 11.34) and a corresponding
decrease in the corresponding moduli (Figure 11.35). However, their effects on the values
of the shorter retardation time are less marked compared with the dramatic changes in
the longer retardation time. The joint effect of the plasticizer and tackifying resin on the
retardation time and corresponding modulus of SIS are much more pronounced than the
700 P
Retardation times (s); P (N/m)
600
500
400
300
200
τ2
τ1
100
0
SIS SIS+I SIS+R SIS+I+R DT BR BR+PIB
FIGURE 11.34 Effects of plasticizers [isolene (I) and low-molecular-weight PIB] and tackifier
resin [regalite (R)] on retardation times and peel adhesion (P) of SIS and BR compared with Duro-
Tak 34-4230 (DT), used in this work as a typical hydrophobic PSA. (From Novikov, M.B. et al.
J. Adhesion, 81, 77–107, 2005.)
700 P
600
400
300
200 G2
G1
100
0
SIS SIS+I SIS+R SIS+I+R DT BR BR+PIB
FIGURE 11.35 The impact of plasticizers (I and low-molecular-weight PIB) and tackifier resin
(R) upon the retardation moduli and peel adhesion (P) of SIS block copolymer and BR compared
to Duro-Tak 34-4230 (DT), used in this work as a typical hydrophobic PSA. (From Novikov, M.B.
et al. J. Adhesion, 81, 77–107, 2005.)
separate effects of the plasticizer–tackifier mixture (compare the SIS + I + R system with
the SIS + I and the SIS + R blends in Figures 11.34 and 11.35).
600
P
500
P (N/m) τ (s)
400
300
τ2
200
100
τ1
0
30 32 34 36 38 40 42
PEG content (%)
FIGURE 11.36 180o peel adhesion, P, and retardation times of PVP–PEG adhesives as a function
of PEG concentration at 50% relative humidity of the surrounding atmosphere. (From Novikov,
M.B. et al. J. Adhesion, 81, 77–107, 2005.)
Let us compare now the values reported for the hydrophobic adhesives with values
featured for hydrophilic PVP–PEG PSAs. The PVP–PEG system provides an appropri-
ate model to illustrate the effect of the molecular structures underlying pressure-sensi-
tive adhesion, because its adhesive behavior can be related to the changes in structure,
interaction mechanism, phase state, and other physical properties as the PEG content is
varied. The effects of PEG concentration on adhesive and relaxation properties of PVP–
PEG blends expressed in terms of 180° peel force, 39 retardation times, and correspond-
ing moduli are presented in Figures 11.36 and 11.37. Whereas the shorter retardation
time is nearly constant, the longer time is a monotonously increasing function of PEG
content, and the corresponding moduli decrease with the increase in PEG concentra-
tion between 31 and 41% PEG, peel adhesion comes through a maximum at 36% PEG
concentration in the blends. The maximum adhesion relates to the PEG concentration
at which a stoichiometric PVP–PEG H-bonded complex is completely formed within
the PVP–PEG blends.22,24,25,40 This complex demonstrates properties that are not typi-
cal of both parent polymers. As evident from the data in Figures 11.36 and 11.37, the
maximum adhesion in PVP blends with 34–39% PEG is observed when the shorter and
longer retardation times range within 10–65 and 120–450 s, respectively, whereas the
values of the corresponding moduli vary between 0.5–2.6 and 0.8–3.8 MPa, respec-
tively. The relaxation properties of the blend containing 36% PEG and providing the
best adhesion are characterized by τ1 = 26–46 s, G1 = 1.3–2.17 MPa, and τ2 = 325–430 s,
G 2 = 2.94–3.3 MPa.
Within the PEG concentration region (31–34%), where debonding occurs through a
predominantly adhesive type of bond failure, 38 the gain in adhesion is always associ-
ated with an appreciable rise in the value of the longer retardation time (Figure 11.36).
8 600
G1
G2
6 500
P
400
4
300
2
200
0
30 32 34 36 38 40 42
PEG content (%)
FIGURE 11.37 Effect of PEG concentration on 180o peel adhesion, P, and moduli corresponding
to shorter (G1) and longer (G 2) retardation times for PVP–PEG adhesives at relative humidity of
50%. (From Novikov, M.B. et al. J. Adhesion, 81, 77–107, 2005.)
However, as the PEG concentration reaches 36% and higher, the type of debonding
becomes miscellaneous (adhesive–cohesive) and this rule no longer holds. For these
blends, the longer retardation time continues to increase more smoothly, whereas peel
adhesion begins to decrease.
In the same manner as that established above for hydrophobic blends based on SIS and
BR (Figures 11.34 and 11.35), for hydrophilic adhesives greater adhesion is associated with
the values of longer retardation time ranging from 325 to 445 s (Figures 11.36 and 11.37).
It is evident that further work is needed to demonstrate whether the established values
of retardation times and relevant moduli are also typical of the entire variety of PSAs
currently available. Furthermore, more data should be obtained to trace quantitative
correlations between the adhesion and relaxation characteristics within the window
outlined by the above criteria. As follows from the data in Figures 11.34 and 11.35, the
retardation times and G1 modulus for SIS + I adhesive fall within the relaxation criteria
for PSAs, yet the SIS + I adhesive exhibits a relatively low peel force. Th is is most likely
due to the fact that the modulus G 2 = 1.61 MPa for the SIS + I blend is below the lower
limit outlined by the relaxation criterion (G 2 = 2.5 MPa). It implies also that there exist
different combinations of retardation times and corresponding moduli that are either
favorable or unfavorable for high adhesion.
0.8 σmax
Stress σ (MPa)
0.6
0.4 250 µm
σbf
0.0
0 1 2 3 4
Strain ε 1 mm
FIGURE 11.38 Direct observation of the debonding mechanisms and stress versus strain curve
in the course of probe tack test for SIS-based PSA and corresponding images illustrating trans-
formations of the PSA structure in the plane normal to the adhesive layer at different stages of the
debonding process. Debonding velocity 0.1 µm/s. T = 22°C, contact time 1 s. (From Creton, C.
et al. in Adhesion: Current Research and Applications, Possart, W.G., Ed., Weinheim, Wiley-VCH,
2005. With permission.)
Figure 11.39 illustrates the displacement of motors and a typical force curve as a func-
tion of time during a relaxation experiment carried out in a probe tack setup. The
displacement of the motors driving the probe was stopped at a given moment during
different stages of the debonding process. The system, under tensile stress, was then left
to relax for a given time, tstop. During this time, relaxation of the force occurs. At the
end of the stop the displacement of the probe was resumed until complete debonding of
the adhesive occurred.44 The model adhesives used in this study consisted of a series of
acrylic copolymers based on 2-ethylhexyl acrylate as a base monomer. As comonomers,
they contained increasing amounts (2, 4, and 8 wt %) of acrylic acid (AA). Adhesives are
referred to as 2AA, 4AA, and 8AA, respectively. The advantage of using polyacrylates is
that the pure polymers exhibit PSA properties without any need for additional formula-
tion ingredients such as tackifying resins.
Figure 11.40 illustrates the nominal stress as a function of time for a nonstop test
and for three relaxation tests on a steel probe. The relaxation tests were always per-
formed during the increase in force at the beginning of the debonding process before
catastrophic failure was observed. The stops from stop 1 to stop 3 were performed
at increasing values of stress at the beginning of the relaxation. The level of tensile
stress at which the test is stopped and the adhesive allowed to relax is referred to as σ 0.
Because this early loading stage of the debonding process is essentially elastic at the
Displacement
(motors)
Vdeb
Vdeb tstop
Time
Force
Stop
Time
FIGURE 11.39 Force and displacement of the motors for a relaxation test as a function of time.
The motors are stopped at a given moment, and the adhesive is allowed to relax for a given time,
tstop. (From Lindner, A., Maevis, T., Brummer, R., Lűhmann, B., and Creton, C., Langmuir 20,
9156, 2004. With permission.)
0.30
Without stop
0.25
0.20
Stress (MPa)
0.15
Stop 3
0.10 Stop 2
Stop 1
0.05
0.00
0 50 100 150 200 250
t (s)
FIGURE 11.40 Stress, σ, as a function of time for 2AA acrylic PSA. A test without stop and
three tests with stops at different initial values of σ 0: Stop 1, σ0 = 0.08 MPa; Stop 2, σ 0 = 0.16 MPa;
Stop 3, σ0 = 0.25 MPa. For all stops, tstop is 180 s. (From Lindner, A., Maevis, T., Brummer, R.,
Lűhmann, B., and Creton, C., Langmuir 20, 9156, 2004. With permission.)
Figure 11.41 (top) illustrates the stress relaxation of the 2AA adhesive on a high-energy
steel probe for the three different stops. Figure 11.41 (bottom) demonstrates the same
relaxation process for 2AA but on an apolar low-energy poly(ethylene-co-propylene)
(EP) substrate. As evident from these data, relaxation of the tensile stress is strongly
affected by the nature of the substrate. This implies that stress relaxation is not only a
property of the adhesive material, but also the property of the adhesive–adherent pair.
In other words, the boundary conditions at the probe–adhesive interface contribute
greatly to the relaxation of stress in the course of the debonding process.
As follows from Figure 11.41 (top), the decrease in stress slows down toward the end of
the stop and finally reaches a nearly constant value, as occurs for viscoelastic materials
0.25
0.20
0.15
Stress (MPa)
Stop 3
0.10
Stop 2
0.05 Stop 1
0.00
0 50 100 150 200
t (s)
0.25
0.20
Stress (MPa)
0.15
Stop 3
0.10
Stop 2 Stop 1
0.05
0.00
0 50 100 150 200
t (s)
FIGURE 11.41 Stress relaxation during probe stops for 2AA acrylic adhesive on high-energy
steel (top) and low-energy EP substrates (bottom). (From Lindner, A., Maevis, T., Brummer, R.,
Lűhmann, B., and Creton, C., Langmuir 20, 9156, 2004. With permission.)
possessing a yield stress. When visualizing the debonding process, it is obvious that
most of the cavity growth and cavitation takes place during the first 60 s. The further
comprehension of structural evolution process is possible if the characteristic times of
the structural evolution are taken into consideration.
To describe precisely the deformation mechanism during the debonding process
observed in the probe tack test, we must define the terms cavitation and propagation.
Confined fi lms of PSA under a tensile stress develop cavities. These cavities appear where
surface defects were initially present,44 and we can optically detect them when their size
becomes of the order of a few micrometers. The cavities then grow rapidly to a size of the
order of the thickness of the fi lm. At this stage of the debonding process, two different
mechanisms can be observed on video images made in the course of deformation and
relaxation: either the cavity stops growing and new cavities progressively nucleate, even-
tually fi lling the space previously occupied by the adhesive fi lm, or the cavity continues
growing laterally in a disk-like shape until it comes in contact with an adjacent disk-like
cavity.45,46 The process of nucleation of new cavities under constant stress is called cavi-
tation, whereas the growth of existing cavities in the plane perpendicular to the tensile
direction is referred to as propagation.
Video captures taken during the force relaxation process (Figures 11.42 and 11.43)
reveal whether cavitation or propagation takes place. These observations reveal whether
the cavities eventually coalesce or whether individual cavities persist and form a foam-
like structure when the adhesive layer is stretched further. This last question is of partic-
ular importance to the long-term durability of the adhesive bond, because a coalescence
of individual cavities leads to rapid complete detachment of the adhesive.
1a 1b 2a 2b
FIGURE 11.42 Snapshots taken during the relaxation process for 2AA (left) and 8AA (right) on
steel.44 The pictures at the top are always taken at the beginning of the stop (t = 0) and the pictures
at the bottom are taken at the end of the stop (t = 180 s). (1a) A stop at a low initial stress level
(σ0 = 0.08 MPa, σ0/σmax = 28%). (1b) A stop at higher initial stress (σ0 = 0.25 MPa, σ0/σmax = 85%)
for 2AA. One observes little cavitation, but some growth of the existing cavities. (2a) A stop at a
low initial stress level (σ0 = 0.14 MPa, σ 0/σmax = 25%). (2b) A stop at a higher initial stress for 8AA
(σ0 = 0.45 MPa, σ0/σmax = 81%). In this case one observes the nucleation of new cavities. (From
Lindner, A., Maevis, T., Brummer, R., Lűhmann, B., and Creton, C., Langmuir 20, 9156, 2004.
With permission.)
a b c
FIGURE 11.43 Snapshots taken during the relaxation process of 2AA acrylic adhesive on a low-
energy EP substrate. The images at the top are always taken at the beginning of the stop, whereas
the images at the bottom are made at the end of the stop (t = 180 s for a and b, t = 15 s for c).
(a) Stop at a low initial stress level (σ0 = 0.13 MPa, σ 0/σmax = 51%). (b) Stop at intermediate initial
stress (σ0 = 0.15 MPa, σ 0/σmax = 59%). (c) Stop at high initial stress level (σ0 = 0.22 MPa, σ 0/σmax =
86%). On the low adherence surface, no cavitation but substantial growth is observed. (From
Lindner, A., Maevis, T., Brummer, R., Lűhmann, B., and Creton, C., Langmuir 20, 9156, 2004.
With permission.)
When the adhesives were bonded to a steel surface, both mechanisms of cavity nucleation
and growth were observed, whereas on low-energy EP surfaces, only crack propagation
occurred. This result demonstrates that unless the sensitivity of these specific mechanisms
to the molecular structure of the adhesive is understood, there will be no hope for predict-
ing the lifetime of the bond. Because cavitation and crack growth are processes that entail
locally large strains of the adhesive, it is highly unlikely that linear viscoelastic properties
alone will be able to predict the nucleation or growth of these cavities.
The sharp difference in behavior between the two substrates highlights the impor-
tance of adhesive interactions. On steel, resistance to crack propagation is high, so the
most important property, that the adhesive must have, is good resistance to the forma-
tion of cavities. On the contrary, on EP surfaces, resistance to cavitation is not very
important because failure occurs by crack propagation. It is the balance between these
two properties that must be optimized.
stress, when a fibrillar foam-like structure is not yet formed. Here, we consider the
results of a similar relaxation test performed by stopping the driving motor of the probe
tack apparatus and monitoring the force relaxation under constant displacement con-
ditions while the fibrillar foam is well formed, the peak stress is covered, and a strain-
hardening effect is observed, indicating the formation of strong fibrils (Figure 11.44).47
The adhesive used in this case is based on SIS triblock copolymer.
As indicated by the data in Figures 11.44 and 11.45, stress does not relax to zero, as one
would expect for a viscous liquid capable of flow, but to ∼70% of its initial value before
remaining constant. A similar relaxation curve with appreciable residual unrelaxed stress
was obtained for an SIS-based adhesive at the conditions imitating adhesive bond for-
mation under bonding pressure (compare with Figures 11.4 and 11.7). This relaxation
behavior is typical of viscoelastic adhesives possessing an apparent yield stress and clearly
demonstrates that the fibrils are able to store elastic energy during their formation and
stretching; only during detachment or fracture is this elastic energy released.47
The video images of the debonding process, illustrated in Figure 11.46, clearly indi-
cate that the plateau stress for the SIS adhesive corresponds to the formation and elon-
gation of the fibrillar foam.48 The images demonstrated that the average cell size and
the amount of cavities are fairly independent of the relaxation time so that observed
differences in measured stress cannot be due to a different microscopic structure of the
foam. Comparatively negligible changes in the structure of SIS adhesive occur during
the relaxation process, emphasizing the very elastic nature of the polymer in the fibrils.
Such an elastic fibrillating behavior has also been reported in an even more pronounced
way for soft physical gels, which are composed of block copolymers swollen in a prefer-
ential solvent for the midblock.44
1.0
0.8
0.6
σ (MPa)
0.4
0.2
0.0
0 50 100 150 200 250 300
Time (s)
FIGURE 11.44 Stress-versus-time curves for probe tack tests with intermediate stops at differ-
ent levels of initial stress for SIS adhesive. The test is stopped for 120 s and then resumed. Note
that the stress–strain curves are almost identical for all tests. (From Roos, A. and Creton, C.,
Macromol. Symp. 214, 147, 2004. With permission.)
0.5
0.4
0.3
σN (MPa)
0.2
0.1
0.0
0 100 200 300 400 500 600
t − t stop (s)
FIGURE 11.45 Stress-versus-t − tstop curves obtained by stopping the driving motors in the
fibrillation regime at three different stress levels for the SIS adhesive. The test was stopped here for
600 s. (From Roos, A. and Creton, C., Macromol. Symp. 214, 147, 2004. With permission.)
FIGURE 11.46 Snapshots taken during the relaxation process for SIS adhesive. The left image
was taken at the beginning of the stop, whereas the right image relates to the end of the stop. The
horizontal full scale is 1.5 mm. (From Roos, A., Ph.D. Thesis, Universite Paris VI, Paris, 2004.
With permission.)
modulus was measured as a function of time G(t) for 103–104 s. In elongation, the
relaxation tests were performed at a deformation of 500%. It took 9 s for the cross-head
to go to this elongation (Vt = 500 mm/min). Tests were performed at room temperature,
and the relaxation of stress was measured as a function of time for 1000 s.
The relaxation moduli normalized by the moduli at the beginning of the relaxation
(G 0) are presented in Figure 11.47 for the four model resin blends.49 G 0 is independent
of the PS–PI content, but always lower than G′. The material relaxes more when there is
more diblock in the blend, namely, from 30 to 85% of the stress after an imposed defor-
mation of 2%. In all cases, the relaxation seems to be over after 1000 s.
Figure 11.48 illustrates the relaxation of stress normalized by the level of stress at
the beginning of relaxation for both pure SIS blends and resin blends.49 Unlike in the
linear regime, the levels of stress at the beginning of relaxation decrease with increas-
ing diblock and resin contents. This is not surprising because in the tensile geometry
the stretch to 500% deformation takes several seconds, and some relaxation can take
place already in this loading stage. Similar to what is observed in the linear regime, the
more diblock in the blend, the more rapid and pronounced the relaxation. However, the
similarity ends here. After 1000 s, the material only relaxes 18–27% of the initial applied
stress for pure polymer blends and 18–34% of the stress for adhesive formulations. In
addition, in all cases the relaxation process is not over after 1000 s.
These two experiments demonstrate that there is a remarkable difference in the kinet-
ics of relaxation at small and large strains. When the sample is stretched to large strains,
the relaxation of polymeric chains occurs in a strongly oriented network at a rate that
probably depends mainly on the relaxation of the triblock chains bridging between PS
domains. On the other hand, at small strains the network of polymer chains is not ori-
ented to such an extent and most of the stress, and then the relaxation, occurs between
entanglements in the polyisoprene domains, which relax faster than the central PI
blocks of the triblocks.49 The SIS blends with a PS–PI diblock copolymer display marked
1
9
8
7
6
5
4
54→0
G(t)/G0
0.1
10−1 100 101 102 103 104
t (s)
FIGURE 11.47 Relaxation modulus in shear for SIS blends with 60 wt % resin after a deformation
of 2%.49 The content of PS–PI diblock copolymer varies between 54 and 0 wt %. (From Roos, A.
and Creton, C., Macromolecules, 38, 7807, 2005. With permission.)
1.0
0.9
σN(t )/σ0
54→0
0.8
0.7
0.6
2 4 6 8 2 4 6 8 2 4 6 8
(a) 1 10 100 1000
t (s)
1.0
0.9
σN(t )/σ0
54→0
0.8
0.7
0.6
2 4 6 8 2 4 6 8 2 4 6 8
(b) 1 10 100 1000
t (s)
FIGURE 11.48 Relaxation of the stress in elongation after a deformation of 500% for (a) SIS blends
without resin and (b) blends with 60% resin. The content of the PS–PI diblock copolymer varies between
54 and 0 wt %. (From Roos A. and Creton C., Macromolecules 38, 7807, 2005. With permission.)
nonlinear viscoelastic behavior and are much more dissipative at low and intermediate
strains than at very high strains. This suggests that free diblock chains dangling from PS
domains may have very long relaxation times, which dominate the viscoelastic behavior
at low and intermediate strains, whereas at high strains the behavior is dominated by the
bridging chains provided by the triblocks.
The significance of the relaxation data presented in this section for pressure-sensitive
adhesion emerges from the fact that the adhesion of SIS blends increases with increasing
contents of resin and PS–PI diblock copolymer. As evident from Figures 11.47 and 11.48,
both tackifying resin and PS–PI plasticizer accelerates the stress relaxation, both in the
linear elastic deformation region and at large strains. For a more insightful analysis, the
values of relaxation times and corresponding moduli are to be taken into consideration.
With this purpose in mind, the relaxation properties of another physically cross-linked
model PSA based PVP–PEG H-bonded complex have been studied.
6
log H (Pa)
−12 −10 −8 −6 −4 −2 0 2 4 6 8
log τ (s)
FIGURE 11.49 Relaxation spectrum featured for the PVP adhesive blend with 36% PEG at
20°C. (From Feldstein M.M. et al., J. Appl. Polym. Sci., 100, 522–537, 2006. With permission.)
and increase in extension velocity (Figure 10.21, between 20 and 50 mm/min). The narrow
transition from ductile to tight stretching with the increased drawing rate in Figure 10.21
corresponds to a well-defined rate of rearrangement of the H-bonded network during the
deformation of the PVP–PEG adhesive. Assuming that the cooperative breakup and ref-
ormation of hydrogen bonds forming this PVP–PEG network can only occur below the
critical deformation rate of 0.05 s−1, we can identify the characteristic time for this process
to occur at about 20 s.53 What is of particular importance for the following discussion is
the equivalence between the effects of the increase in plasticizer (PEG) content and the
decrease in stretching rate on the tensile and, consequently, relaxation properties.
Figure 11.50 represents the effect of drawing rate on tensile stress relaxation curves
for the most tacky PVP blend with 36 wt % PEG containing 8% of absorbed water at a
fi xed elongation, ε = 3. The relaxation follows the exponential law described by Equa-
tion 11.2 and written as
in
Et Eeq ∑ Ei exp(t /i ) (11.4)
i1
where Ei are the tensile relaxation moduli. Similar to what is observed for SIS adhesives in
the shear regime (Figures 11.4 and 11.47), the PVP–PEG relaxation in Figure 11.50 illus-
trates the occurrence of a residual unrelaxed stress, Eeq. The lower the drawing rate in the
loading stage, the more rapid and pronounced the relaxation and the lower the equilib-
rium value of the corresponding modulus, Eeq. The curves relating to a comparatively fast
relaxation process can be satisfactorily described by Equation 11.4 with two characteris-
tic times of elastic recovery. This is the case for blends containing 41 wt % PEG at low
(10 mm/min) loading rates. In contrast, the blends containing 36 wt % PEG and less, as well
as the samples deformed with higher extension rates, have relaxation curves that can be
340
300
260
220
E (kPa)
100
180
140 50
20
100
60 10
20
0 200 400 600 800 1000 1200
Time (s)
FIGURE 11.50 Relaxation curves recorded for the PVP blend with 36 wt % of PEG-400 that is
stretched to a deformation of 300% (ε = 3) with a rate ranging from 10 to 100 mm/min. The con-
tent of absorbed water in the blend is 8 wt %.
adequately fitted with Equation 11.4, including three populations of characteristic recovery
times, τ. For this reason, in the following discussion we consider the effects of relative elon-
gation on relaxation, which have been measured for the elastic PVP blend with 36% PEG,
in terms of two characteristic recovery times, whereas the effects of stretching velocity and
PEG concentration are treated with three different populations of recovery times.
80
70 100 mm/min
60
50
τ1 (s)
40
30
20
10 mm/min
10
500
100 mm/min
450
400
τ2 (s)
350
300
250
10 mm/min
200
150
FIGURE 11.51 Effects of relative elongation and extension velocity on characteristic times of
elastic recovery for the PVP–PEG (36 wt %) model PSA.
160
140
100 mm/min
120
10 mm/min
100
Eeq (kPa)
80
60
40
20
0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
ε
FIGURE 11.52 Impacts of relative elongation, ε, and drawing rate on the values of equilibrium
characteristic recovery modulus, E eq, for the PVP–PEG (36%) model PSA.
By fitting the relaxation curves with Equation 11.4, one can extract two characteris-
tic times of recovery, which increase linearly with increasing elongation (Figure 11.51).
The increase in strain rate during loading makes this relationship more pronounced.
At a comparatively low velocity of stretching (10 mm/min), the shorter characteristic
time varies between 18 and 75 s, whereas the longer time ranges from 200 to 520 s. The
10-fold increase in strain rate results in an appreciable decrease and narrowing of the
characteristic time ranges: τ1 = 7–15 and τ2 = 145–230 s. The values of characteristic
equilibrium moduli are presented in Figure 11.52. At the low strain rate of 10 mm/min
the Eeq is an increasing function of elongation. The most growth is observed in the range
of elongations between ε = 1–2, wherein the Eeq increases from zero to an almost con-
stant value. At high strain rate (100 mm/min) the equilibrium characteristic modulus
tends to increase linearly with tensile strain. Within a strain range ε = 2–4, the values
of the equilibrium characteristic moduli are almost identical and independent of strain
rate (Figure 11.52).
The E eq value characterizes the amount of mechanical energy stored by the adhesive
material in the process of loading. At low tensile rates the material has enough time to
relax and the amount of stored energy is negligible. This behavior is typical of viscous
liquids and un-cross-linked linear polymers. In contrast, at strain rates of 100 mm/min
the material behaves as a cross-linked elastomer, storing elastic energy in the course of
stretching. The transition from the liquid-like to rubber-like behaviors occurs in very
narrow range of tensile rates that is due to the well-defined relaxation time of the net-
work of hydrogen bonds in the PVP–PEG complex.
350
300
250
200
E1 (kPa)
100 mm/min
150
100
10 mm/min
50
280
240
200
160
E2 (kPa)
100 mm/min
120
80
10 mm/min
40
0
(b) 1.0 1.5 2.0 2.5 3.0 3.5 4.0
ε
FIGURE 11.53 Relation of the characteristic moduli of elastic recovery to the relative elonga-
tion and drawing velocity of an adhesive PVP blend with 36 wt % PEG.
500
τ3
400
300
200
100 τ1
τ2
0
0 20 40 60 80 100
Tensile rate (mm/min)
FIGURE 11.54 Effect of loading tensile rate on the values of characteristic recovery times of
PVP–PEG PSA; ε = 3.
45
40
35
Eeq (kPa)
30
25
20
0 20 40 60 80 100
Tensile rate (mm /min)
FIGURE 11.55 Equilibrium characteristic recovery modulus as a function of the tensile rate in
the course of loading of a PVP–PEG (36%) model PSA; ε = 3.
140
120 E3
100
80 E2
E (kPa)
60
E1
40
20
0
0 20 40 60 80 100
Tensile rate (mm/min)
FIGURE 11.56 The effect of extension velocity on the characteristic moduli of elastic recovery
of a PVP–PEG PSA; ε = 3.
curves was performed using Equation 11.4 with three populations of characteristic
recovery times. In agreement with the data in Figures 11.51 through 11.53, the 10-fold
increase in strain rate from 10 to 100 mm/min does not affect the values of shorter
and intermediate characteristic times (Figure 11.54), whereas the longer recovery time
tends to decrease with the increase in strain rate. The equilibrium characteristic recov-
ery modulus increases at lower deformation rates, achieving its limiting value since the
velocity of 50 mm/min (Figure 11.55). Characteristic recovery moduli corresponding
to shorter and intermediate recovery times, E1 and E2 , are increasing functions of
the tensile rate, whereas the value of E3 decreases with the increase in loading tensile
velocity (Figure 11.56).
The results presented above suggest that the conditions of adhesion testing, such as
strain rate and strain amplitude, affect appreciably the relaxation properties of adhesive
materials. The effect of maximum strain on relaxation is in synergy with the impact of
strain rate. The relaxation behavior is controlled by two major factors: the longer charac-
teristic time of elastic recovery and the value of the equilibrium characteristic recovery
modulus. The former value increases with strain and decreases with increasing strain
rate; the latter increases in both cases.
300
τ3
250
200
τ (s)
150
100
50 τ2
τ1
0
32 36 40
PEG (wt %)
FIGURE 11.57 Characteristic recovery times plotted versus the composition of PVP–PEG PSA.
The composition displaying maximum adhesion is marked by a dotted line. The extension rate is
100 mm/min; ε = 3.
80
60
Eeq (kPa)
40
20
0
30 32 34 36 38 40 42
PEG (wt %)
FIGURE 11.58 The effect of PEG content in blends with PVP on the value of the equilibrium
characteristic recovery modulus. The dotted line designates the composition exhibiting maxi-
mum adhesion. The extension rate is 100 mm/min; ε = 3.
50
40 E2
30
E (kPa)
E1
20
E3
10
0
30 32 34 36 38 40 42
PEG (wt %)
FIGURE 11.59 Values of the characteristic recovery modulus for PVP–PEG adhesives of vari-
ous composition. The extension rate is 100 mm/min; ε = 3.
the corresponding characteristic times (Figure 11.59). This behavior implies that PEG is a
good plasticizer of PVP.
It is instructive to compare the effects of PEG concentration in PVP–PEG blends on
relaxation properties, illustrated in Figures 11.57 through 11.59, and on the retardation
time and relaxation modulus, presented in Figures 11.36 and 11.37. With the increase
in PEG content, the relaxation and recovery moduli decline, but the longer retardation
time increases in contrast to the behavior of the characteristic recovery time that is,
in essence, the relaxation time. This distinction accounts for a fundamental difference
between the relaxation and the retardation processes, defined by the Equations 11.2 and
11.4 on the one hand and Equation 11.3 on the other.
All PVP–PEG blends considered in this section demonstrate some level of pressure-
sensitive adhesion. However, the PVP blend with 36% PEG exhibits the best adhesion.
Its relaxation properties indicate which values of the characteristic recovery times and
corresponding moduli are associated with a high level of pressure-sensitive adhesion.
In turn, the third stage, adhesive debonding, is a multistage process itself, including the
steps of homogeneous deformation under a tensile stress, followed by cavitation and
fibrillation of the adhesive material. Relaxation mechanisms accompany all stages of the
adhesion process and provide the links between the stages.
In terms of relaxation properties, all PSAs can be classified into two groups: (1) fully
relaxing (fluid) adhesives and (2) elastic adhesives that are able to store energy in the
course of bond formation and exhibit residual (unrelaxed) stress during the contact
time. In terms of molecular structure, the group of fluid adhesives includes un-cross-
linked, entangled polymers with low glass transition temperatures, whereas the elastic
PSAs are most often networks of polymers that are cross-linked covalently or physically.
The elastic PSAs of the second group demonstrate a much slower relaxation than the
fluid PSAs. The longer relaxation processes are mostly associated with the appearance
of a pronounced plateau or even of a second maximum on the probe tack curves. The
equilibrium relaxation modulus is a direct measurement of the stored elastic energy in a
polymer material and relates to its yield stress. Fluid adhesives do not have a yield stress
or, consequently, an equilibrium modulus.
As the comparison of the relaxation and adhesive properties of various PSAs indi-
cates, high adhesion is associated with longer relaxation times. This conclusion is in full
agreement with Equation 10.6 in Chapter 10, which predicts improved adhesion with
the increase in longer relaxation time of the adhesive material. In turn, the longer val-
ues of relaxation time are typical of large-scale entangled and network supramolecular
structures. This implies that large-scale rearrangements of molecular structures in the
course of relaxation govern a high level of pressure-sensitive adhesion. The relationship
between peel adhesion and the molecular structure of the adhesive material, predicted
by Equation 10.6 in Chapter 10, also suggests that the greater the size of the relaxing
molecular structures in adhesive material, a, the higher the adhesion.
As predicted using Equation 10.6 in Chapter 10, for high adhesion a compromise
must be reached among the values of cohesion energy, diff usion coefficient, and relax-
ation time of PSAs. It is, therefore, not surprising that a direct correlation between the
practical work of adhesion and the values of longer relaxation times has been estab-
lished for the examined PSAs. Adhesion appears with the rise of longer relaxation
times values above 50 s and increases, passing through a maximum at τ3 = 330–380 s.
A further increase in the longer relaxation times results in a gradual decline in adhe-
sion. Good adhesion occurs when the longer relaxation time varies in the range of 150
to 800 s.
Relaxation properties of the PSAs depend on the observation window. If the observa-
tion time is long enough, a PSA seems to be more fluid than at shorter observation time.
PSAs are viscoelastic materials that combine the properties of solids and liquids. Relative
contributions of the viscous and elastic behaviors can be estimated in terms of the Deborah
numbers that relate the time scale of structural rearrangement in the material to the time of
experimental observation. Because the viscous and elastic contributions are perfectly coun-
terbalanced, the Deborah number is unity. Very fluid, quickly relaxing materials possess
values of Deborah number tending to zero. Useful PSAs have demonstrated Deborah num-
bers between 0.15 and 1 if the longer relaxation time is chosen as the characteristic relax-
ation time of the adhesive.
The relaxation behavior during the stage of strain relaxation upon withdrawal of the
bonding pressure has been characterized in terms of retardation times. Retardation
times are not identical to relaxation times. Whereas the relaxation times increase with
plasticization, the retardation times decline. Two populations of retardation times have
been determined for PSAs: shorter retardation times of the order 10–70 s and longer
retardation times in the range of 300–660 s. The longer retardation times relate mainly
to the energy-dissipating processes, which, in turn, are associated with large-scale rear-
rangements of supramolecular structures involving polymer chain entanglements and
translational movements of polymer segments and entire macromolecules through a
self-diff usion mechanism. The observed correlation between high adhesion and the val-
ues of longer retardation times supports the validity of Equation 10.6 in Chapter 10.
The large-scale rearrangements of the structure of adhesive materials during relax-
ation can be easily visible and characterized if the probe tack tester is stopped at different
stages of the debonding process, when the main mode of PSA deformation is in tension.
When a soft adhesive is kept under a tensile load during the debonding process, two
important failure mechanisms are observed: nucleation of new cavities and the growth
of existing cavities as cracks at the interface between the adhesive and the adherent. The
plateau stress in probe tack curves corresponds to the formation and elongation of a
fibrillar foam. The stop of the probe at this stage for elastic adhesives demonstrated that
the size and number of cavities remain the same in the course of relaxation.
To describe quantitatively PSA relaxation in the course of debonding, when large
tensile strain dominates, the method of stretching adhesive fi lms is useful. The work
of viscoelastic deformation up to fi lm fracture under uniaxial extension has been
established to control peel adhesion. If the relaxation properties of PSAs under a large
tensile strain are expressed in terms of characteristic times of elastic recovery and
corresponding characteristic moduli, the effect of debonding conditions, such as load-
ing rate and tensile strain amplitude, on relaxation becomes evident. The effect of
maximum tensile strain on relaxation is in synergy with the effect of strain rate. The
relaxation behavior at large strains is controlled by two major factors: the longer char-
acteristic time of elastic recovery and the value of equilibrium characteristic recovery
modulus. The former value increases with strain and decreases with strain rate; the
latter increases in both cases.
The data presented in this chapter demonstrate the significance for pressure-sensitive
adhesion of relaxation properties typical of PSAs, as well as the development and trans-
formations of these properties in the course of adhesion testing.
References
1. Tshoegl N.W., Time dependence in material properties: An Overview, Mechan.
Time-Depend Mater., 1, 3, 1997.
2. Moonan W.K. and Tschoegl N.W., The effect of pressure on the mechanical prop-
erties of polymers. 4. Measurements in torsion, J. Polym. Sci., Polym. Phys. Ed., 23,
623, 1985.
3. Ferry J.D., Viscoelastic Properties of Polymers, 3rd ed., Wiley & Sons, New York,
1980.
4. Aklonis J.J. and MacKnight W.J., Introduction to Polymer Viscoelasticity, 2nd ed.,
Wiley & Sons. New York, 1980.
5. Christensen R.M., Theory of Viscoelasticity, 2nd ed., Academic Press, New York,
1982.
6. Matsuoka S., Relaxation Phenomena in Polymers, Oxford University Press, New
York and Carl Hauser Verlag, Munich, 1992.
7. Satas D. (ed.), Handbook of Pressure-Sensitive Adhesive Technology, 3rd ed., Satas
& Associates, Warwick, RI, 1999.
8. Dillard D.A. and Pocius A.V. (eds.), The Mechanics of Adhesion, Elsevier Science,
New York, 2002.
9. Zosel A., The effect of bond formation on tack of polymers, J. Adhesion Sci. Tech-
nol. 11, 1447, 1997.
10. Tschoegl N.W., The Phenomenological Theory of Linear Elastic Behavior, Springer-
Verlag, Heidelberg, 1989, p. 126.
11. Schramm G., A Practical Approach to Rheology and Rheometry, Haake Rheom-
eters, Karlsruhe, 1994, p. 104.
12. Zosel A., The effect of fibrillation on the tack of pressure-sensitive adhesives, Int. J.
Adhesion Adhes. 18, 265, 1998.
13. Zosel A., Fracture energy and tack of pressure-sensitive adhesives, in: Satas D (ed.)
Advances in Pressure Sensitive Adhesive Technology, vol. 1, Satas & Associates,
Warwick, RI, 1992, chap. 4.
14. Zosel A., Adhesion and tack of polymers: Influence of mechanical properties and
surface tension, Colloid Polym. Sci., 263, 541, 1985.
15. Pickering J.P., Van Der Meer D.W., and Vancso G.J., Effects of contact time, humid-
ity, and suface roughness on the adhesion hysteresis of polydimethylsiloxane,
J. Adhes. Sci. Technol., 15(12), 1429, 2004.
16. Lakrout H., Sergot P., and Creton C., Direct observation of cavitation and fibril-
lation in a probe tack experiment on model acrylic pressure-sensitive adhesives,
J. Adhesion 69, 307, 1999.
17. Creton C. and Fabre P., Tack, in: Dillard D.A. and Pocius A.V. (eds.) Adhesion
Science and Engineering, Vol I: The Mechanics of Adhesion, Elsevier, Amsterdam,
2002, chap. 14.
18. Creton C. and Leibler L., How does tack depend on time of contact and contact
pressure, J. Polym. Sci.: Part B: Polym. Phys. 34, 545–554, 1996.
19. Creton C. and Schach, R., Diff usion and adhesion, in: Benedek I. and Feldstein
M.M. (eds.) Fundamentals of Pressure Sensitivity, Taylor & Francis, Boca Raton,
2008, chap. 2.
20. Voyutskii S.S., Autohesion and Adhesion of High Polymers, Wiley Interscience,
New York, 1963.
21. Dahlquist C.A., in: Patrick R.L. (ed.) Treatise on Adhesion and Adhesives, vol. 2, M.
Dekker, New York, 1969, p. 219.
22. Feldstein M.M., Molecular nature of pressure-sensitive adhesion, in: Benedek I.
and Feldstein M.M. (eds.) Fundamentals of Pressure-Sensitive Adhesion, Taylor &
Francis, Boca Raton, 2008, chap. 10.
23. Feldstein M.M. and Creton C., Pressure-sensitive adhesion as a material property
and as a process, in: Benedek I. (ed.) Pressure-Sensitive Design, Theoretical Aspects,
vol. 1, VSP, Leiden, Boston, 2006, chap. 2.
24. Feldstein M.M., Molecular fundamentals of pressure-sensitive adhesion, in:
Benedek I. (ed.) Developments in Pressure-Sensitive Products, 2nd Edition,
CRC-Taylor & Francis, Boca Raton, 2006, chap. 4.
25. Feldstein M.M., Adhesive hydrogels: structure, properties and application, Polym.
Sci., Ser. A., 46(11), 1265, 2004.
26. Zisman W.A., Adhesion and bonding, in: Mark H.F., Gaylord N.G., and Bikales
N.M. (eds.) Encyclopedia of Polymer Science and Technology, Interscience Publish-
ers (J. Wiley & Sons), New York, 1964, pp. 445–477.
27. Kotomin S.V., Durability of adhesive joints, in: Benedek I. and Feldstein M.M.
(eds.) Fundamentals of Pressure Sensitivity, Taylor & Francis, Boca Raton, 2009,
chap. 9.
28. Feldstein M.M., Kulichikhin V.G., Kotomin S.V., Borodulina T.A., Novikov M.B.,
Roos A., and Creton C., Rheology of poly(N-vinyl pyrrolidone)-poly(ethylene gly-
col) adhesive blends under shear flow, J. Appl. Polym. Sci., 100, 522, 2006.
29. Satas D., Tack, in: Satas D. (ed.) Handbook of Pressure-Sensitive Adhesive Technol-
ogy, 3rd ed., Satas & Associates, Warwick, RI, 1999, chap. 4.
30. Chiche, A., Dollhofer J., and Creton C., Cavity growth in soft adhesives, Eur. Phys.
J. E, 17, 389–401, 2005.
31. Novikov M.B., Gdalin B.E., Anosova J.V., and Feldstein M.M., Stress relaxation
during bond formation and adhesion of pressure sensitive adhesives, J. Adhesion
84, 164, 2008.
32. Bartenev G.M. and Barteneva A.G., Relaxation Properties of Polymers, Chemistry
Publishers, Moscow, 1992, p. 384.
33. O’Connor A.E. and Willenbacher N., The effect of molecular weight and temperature
on tack properties of model polyisobutylenes, Int. J. Adhes. Adhes., 24, 335, 2004.
34. Feldstein M.M., Cleary G.W., and Singh P., Pressure-sensitive adhesives of con-
trolled water-absorbing capacity, in: Benedek I. (ed.) Pressure-Sensitive Design and
Formulation, Application, vol. 2, VSP, Leiden, Boston, 2006, chap. 3.
35. Feldstein M.M., Cleary G.W., and Singh P., Hydrophilic adhesives, in: Benedek I.
and Feldstein M.M. (eds.) Technology of Pressure-Sensitive Adhesives and Products,
Taylor & Francis, Boca Raton, 2008, chap. 7.
36. Novikov M.B., Kiseleva T.I., Anosova J.V., Singh P., Cleary G.W., and Feldstein
M.M., Contribution of relaxation processes into pressure-sensitive adhesion of
interpolymer complexes, Proceed. 30th Annual Meeting Adhesion Soc., Tampa,
Florida, USA, 54, 2007.
37. Novikov M.B., Borodulina T.A., Kotomin S.V., Kulichikhin V.G., and Feldstein
M.M., Relaxation properties of pressure-sensitive adhesives upon withdrawal of
bonding pressure, J. Adhesion, 81(1), 77, 2005.
38. Rohn Ch., Rheology of pressure-sensitive adhesives, in: Satas D. (ed.) Handbook of
Pressure-Sensitive Adhesive Technology, 3rd ed., Satas & Associates, Warwick, RI,
1999, chap. 9.
39. Chalykh A.A., Chalykh A.E., Novikov M.B., and Feldstein M.M., Pressure-
sensitive adhesion in the blends of poly(N-vinyl pyrrolidone) and poly(ethylene
glycol) of disparate chain lengths, J. Adhesion, 78(8), 667, 2002.
40. Feldstein M.M., Roos A., Chevallier C., Creton C., and Dormidontova E.D., Rela-
tion of glass transition temperature to the hydrogen bonding degree and energy
in poly(N-vinyl pyrrolidone) blends with hydroxil-containing plasticizers: 3.
Analysis of two glass transition temperatures featured for PVP solutions in liquid
poly(ethylene glycol), Polymer, 44(6), 1819, 2003.
41. Zosel A., Adhesive failure and deformation behaviour of polymers, J. Adhesion,
30, 135, 1989.
42. Shull K.R. and Creton C., Deformation behavior of thin, compliant layers under
tensile loading conditions, J. Polym. Sci., Polym. Phys., 42, 4023, 2004.
43. Creton, C., Roos A., and Chiche A., Effect of the diblock content on the adhe-
sive and deformation properties of PSAs based on styrenic block copolymers,
in: Possart W.G. (ed.) Adhesion: Current Research and Applications, Weinheim,
Wiley-VCH, 2005, 337p.
44. Lindner A., Maevis T., Brummer R., Lűhmann B., and Creton C., Sub-critical fail-
ure of soft acrylic adhesives under tensile stress, Langmuir 20, 9156, 2004.
45. Shull K.R., Flanigan C.M., and Crosby A., Fingering instabilities of confined elas-
tic layers in tension, J. Phys. Rev. Lett., 84, 3057, 2000.
46. Creton C., Hooker J.C., and Shull K.R., Bulk and interfacial contributions to the
debonding mechanisms of soft adhesives, Langmuir, 17, 4968, 2001.
47. Roos A. and Creton C., Linear viscoelasticity and nonlinear elasticity of block
copolymer blends used as soft adhesives, Macromol. Symp., 214, 147, 2004.
48. Roos A., Ph.D. Thesis, Universitė Paris VI, Paris, 2004.
49. Roos, A. and Creton, C., Effect of the presence of diblock copolymer on the non
linear elastic and viscoelastic properties of elastomeric triblock copolymers, Mac-
romolecules, 38, 7807, 2005.
50. Osako K., in: Nagasawa M. (ed.) Molecular Conformation and Dynamics of Macro-
molecules in Condensed Systems, Elsevier, Amsterdam, 1988, p. 175.
51. Isono Y., and Nishitake T., Stress relaxation and change in entanglement structure
of polyisobutylene in large shearing deformations, Polymer, 36(8), 1635, 1995.
52. Novikov M.B., Roos A., Creton C., and Feldstein M.M., Dynamic mechanical and
tensile properties of poly(N-vinyl pyrrolidone)–poly(ethylene glycol) blends, Poly-
mer, 44(12), 3559, 2003.
53. Roos A., Creton C., Novikov M.B., and Feldstein M.M., Viscoelasticity and tack
of poly(N-vinyl pyrrolidone)–poly(ethylene glycol) blends, J. Polym. Sci., Polym.
Phys., 40, 2395, 2002.
A-1
BR butyl rubber
D self-diff usion coefficient
E elasticity tensile modulus
Eeq equilibrium elasticity tensile modulus
EP poly(ethylene–co–propylene) substrate
F force
G0 relaxation moduli normalized by the moduli at the beginning of the
relaxation
Geq equilibrium relaxation shear modulus
h rod displacement (a gap between the upper and bottom plates of
squeeze-recoil tester), adhesive fi lm thickness
H relaxation spectrum
I isolene, polyisoprene rubber
J compliance
nD Deborah number
P 180° peel force, resistance, adhesion
PFG pulsed-field gradient
PS polystyrene
QSPR quantitative structure–property relationship
R regalite R9100 hydrocarbon resin
R2 regression parameter
t* durability of adhesive joint
TEC triethylcitrate
Tm melting temperature
Vdeb debonding rate
W practical work of adhesion (the area under probe tack curve)
I-1
E G