Professional Documents
Culture Documents
(Springer Series in Advanced Microelectronics Vol. 10.1007 - 978!3!319-56561-3) Egorov, Nikolay
(Springer Series in Advanced Microelectronics Vol. 10.1007 - 978!3!319-56561-3) Egorov, Nikolay
(Springer Series in Advanced Microelectronics Vol. 10.1007 - 978!3!319-56561-3) Egorov, Nikolay
Nikolay Egorov
Evgeny Sheshin
Field Emission
Electronics
Springer Series in Advanced Microelectronics
Volume 60
Series editors
Kukjin Chun, Seoul, Korea, Republic of (South Korea)
Kiyoo Itoh, Tokyo, Japan
Thomas H. Lee, Stanford, CA, USA
Rino Micheloni, Vimercate (MB), Italy
Takayasu Sakurai, Tokyo, Japan
Willy M.C. Sansen, Leuven, Belgium
Doris Schmitt-Landsiedel, München, Germany
The Springer Series in Advanced Microelectronics provides systematic information
on all the topics relevant for the design, processing, and manufacturing of
microelectronic devices. The books, each prepared by leading researchers or
engineers in their fields, cover the basic and advanced aspects of topics such as
wafer processing, materials, device design, device technologies, circuit design,
VLSI implementation, and subsystem technology. The series forms a bridge
between physics and engineering and the volumes will appeal to practicing
engineers as well as research scientists.
123
Nikolay Egorov Evgeny Sheshin
Saint Petersburg State University MIPT
St. Petersburg Dolgoprudny, Moscow region
Russia Russia
Translation from the Russian language edition: Avtoelektronnaya emissiya. Principy i pribory by Nikolay
Egorov and Evgeny Sheshin, © Intellect Publishers 2011. All Rights Reserved.
© Springer International Publishing AG 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
v
vi Preface
In the course of their research many original models, techniques and devices
were introduced, and many of the most interesting results are presented in this book.
The authors extend their thanks to their colleagues and students whose collab-
oration over the years, common research and discussion of results were the prin-
cipal motive behind this book.
Chapters 1, 3, 4, 7 are written by N.V. Egorov. The introduction, Chaps. 2, 5, 6,
8 are written by E.P. Sheshin. The conclusion is written by the authors with con-
tributions from K.A. Nikiforov.
The seventh chapter and conclusion was based on work of N.V. Egorov with the
Saint Petersburg Electrotechnical University “LETI,” Russia. The work has been
funded by Russian Science Foundation (Grant No. 15-19-30022), SPETU LETI.
vii
Contents
1 Basic Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1
1.2 Fowler–Nordheim Theory of Field Emission from Metals . . . . . .. 11
1.3 Classical Morgulis-Stratton Theory of Field Emission
from Semiconductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 20
1.4 Basics of Müller Field Emission Microscope Theory . . . . . . . . . .. 27
1.5 Disadvantages of Classical Field Emission Theories
and Limits to Their Applicability . . . . . . . . . . . . . . . . . . . . . . . . .. 35
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 40
2 Experimental Equipment and Technique . . . . . . . . . . . . . . . . . . . . . . 43
2.1 Field Emission Microscopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.1.1 Design of Field Emission Microscopes . . . . . . . . . . . . . . . . 47
2.1.2 Special Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2 Field-Electron Total Energy Analyzers. . . . . . . . . . . . . . . . . . . . . . 57
2.3 Field Emission Microscopes of Universal Design . . . . . . . . . . . . . 67
2.4 Electronic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.5 Emitter Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.5.1 Metal Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.5.2 Machining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
2.5.3 Foils, Sheets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
2.6 Some Technological Cases of Application of a Field Emission
Microscope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.6.1 First-Type Migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.6.2 Second-Type Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.6.3 Third-Type Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
ix
x Contents
Among the most relevant problems of modern electronics, creation of stable field
emission cathodes able to function for extended periods of time in high technical
vacuum (10−6–10−7 mm Hg) occupies an important place. There are well-known
advantages of field electron cathodes compared to other types of free electron
sources. Those are as follows: lack of heating, high emission current density,
resistance against temperature fluctuations, low sensitivity to external radiation,
lack of inertia, and exponential slope of volt-ampere characteristics. Those prop-
erties in conjunction make field cathodes invaluable for use in various electronic
devices such as cathode ray devices, and flat screens.
The main obstacle for creation of stable field emission cathodes is extremely
high sensitivity of field emission process to any changes of cathode geometry or its
surface structure. Function of a field emission cathode in any real device is always
accompanied by various processes occurring on its surface such as: ion bombard-
ment; ponderomotor stress; adsorption and desorption of molecules of residual
gases; and surface migration. Depending on specifics of a particular construction
and mode of exploitation of the cathode, those processes individually or cumula-
tively lead to a series of effects that ultimately change the structure of a cathode or
mode of its operation: cathode dissipation of material, changes in the shape of
emitter surface, changes in number or position of microelevations on emitter sur-
face, changes in work function, heating of the cathode, mechanical stresses, etc.
Field emission is the most conservative type of free electron emission which
allows creation of new efficient electron devices with new consumer properties.
In order to provide better understanding of phenomena and processes emerging
during field emission, this book is separated into theoretical and experimental parts.
Theoretical chapters consider properties of field emission based on modern
physical and mathematical concepts and models.
Experimental chapters are dedicated to experimental techniques and methods of
field emission experiments as well as practical applications of field cathodes. This
part also includes assessment of field cathode manufacturing methods.
A particular attention is paid to perspectives of future development of field
emission electronics.
xiii
xiv Introduction
1.1 Introduction
i.e. emission of electrons from warmed up solids. There is yet another type of
electron emission that does not require applying of additional energy, which is thus
called field-electron emission (FEE).1
The effect of FEE appears when an extremely strong electric field (with field
strength E * 107 V/cm) is applied to the surface of a solid resulting in transfor-
mation of the potential threshold into a potential barrier which is low and, most
importantly, narrow enough to allow the electron tunneling i.e. the electrons are
able to phase through this barrier and leave the solid without loss of energy.
The phenomenon of FEE was not theoretically explained up until the discovery
of quantum mechanical tunneling effect (e.g., [1, 2]). Soon after this discovery,
about 1928–1929, Fowler and Nordheim have constructed the first well-founded
theory of FEE in metals [3, 4]. Further evolution of FEE theory was mostly based
around development of new models and computation techniques.
As to the theory of semiconductor FEE, one has to admit that up to now such a
theory that would be general enough, logically sound and mathematically and
physically consistent, does not exist [5–7]. The same applies to FEE from carbon
structures [8].
In order to understand the phenomenon of autoelectronemission (same as in case
of thermoelectron and other types of emission) it is necessary to understand the
properties of the system of electrons in solid crystals that serve as emitters. The
most important such properties are the characteristics of the electrons’ energy
spectrum (zonal structure of the crystal), defined by laws of quantum mechanics,
and the energy distribution of electrons in permitted zones that are defined by
statistical laws. These questions are thoroughly studied is a series of excellent
monographs and textbooks (such as, e.g. [1, 5, 6, 9–12]), and thus we are not going
to pursue those problems in too much detail within confines of this book.
Instead we are going to focus mostly on studying and analysis of phenomena on
phase boundaries between metal (or semiconductor) and vacuum and between
metal (semiconductor) and dielectric, that define the phenomenon of FEE itself and
its circumstantial processes.
Basic dependence of emission current density on intensity of the
emission-causing impact (temperature, lighting, electric field intensity etc.) for any
kind of emission can be calculated if we manage to evaluate the integral
Zþ 1 Zþ 1 Zþ 1
j¼e dpy dpz f ðpx ; py ; pz ÞDðpx ; EÞpx m1 dpx ; ð1:1Þ
1 1 1
where e is the electron charge, f is the function of electron distribution in the metal
by pulse components ðpx ; py ; pz Þ; Dðpx ; EÞ is the potential barrier penetrability
factor (penetrability) in presence of an external electric field at the metallic surface
1
Other names for the effect of field electron emission (FEE) include autoelectron emission (AEE),
cold emission, electrostatic emission etc.
1.1 Introduction 3
for the electrons falling on the surface of metal with impulses px along the normal to
the surface coinciding with axis x, and E is electric intensity. If D depends only on
px, then integration of (1.1) over py and pz reduces the problem under consideration
to one-dimensional one and transition from variable px to Ex results in
Zþ 1
j ¼ jej DðEx ; E; UÞnðEx ÞdEx ; ð1:2Þ
0
where nðEx ÞdEx is an electron flux with energies in the interval dEx about Ex from
within the cathode; Ф is the work function known to be caused by the energy spent
to remove an electron from the solid (or the fluid) in the vacuum.
The value of nðEx Þ is calculated from Fermi-Dirac distribution from the ratio
4pmkT E
nðEÞ ¼ ln 1 þ exp ;
h3 kT
I II U III
0 a b x
4 1 Basic Principles
barrier even when it does not get any supplementary energy (tunnel effect). Classic
statistical physics often represents thermal energy exchange between particles as an
energy source necessary to overcome the potential barrier. The exchange is similar
to the one taking place, for example, in thermionic emission. If at the “summit” of
the potential barrier E ¼ Umax , the total energy of the particle E is only an
insignificant part of the total energy of the entire closed system in the thermal
equilibrium state, the probability of penetration through the potential barrier is
directly proportional to the Boltzmann factor exp½ðUmax EÞ=kT; where E is the
average energy of the particle, T is the absolute temperature, k is the Boltzmann
constant [14].
In quantum mechanics [15], penetration through a potential barrier can take
place even in a purely conservative system not located in a thermostat nor
exchanging energy with any other objects. In the “classically unattainable” region
U [ E; the wave function of the particle is distinct from zero (although it decreases
exponentially towards the potential barrier region). If the region dimensions are
finite, the particle can penetrate through it. In a one-dimensional case, when the
“classically unattainable” (the so-called subbarrier) region is limited with two
points x1 and x2, whose locations are identified by the equation UðxÞ ¼ E; the
probability of penetration of a particle/electron through a potential barrier
(barrier-transmission factor, barrier penetrability) is defined as
8
x 9
Z 2
< =
D ¼ exp 2 kðxÞdx ;
ð1:3Þ
: ;
x1
where
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2m½E UðxÞ
k¼
h
is the wave number of the particle, m is its mass and ħ is the Planck constant.
Expression (1.3) holds true on condition that within the distance of one particle
wave length from any cusp decomposition of potential energy U into the difference
ðx x1; 2 Þ can be done through linear approximation. Away from the cusps,
quasi-classical approximation of quantum mechanics is applicable everywhere [1].
Given these conditions, the transmission factor D is exponentially small (in clas-
sical mechanics it is zero). For this type of potential barrier, the particle wave
properties can also be significant in the subbarrier region of values E [ U. In
classical mechanics, this region would demonstrate free passage of particles over
the barrier in the direction of their movement and the reflectivity factor R ¼ 1 D
would become zero. In quantum mechanics, it remains finite for any value of E,
though, generally, it is exponentially small in the suprabarrier region.
Let us go back to Fig. 1.1. Obviously, the shape of the potential barrier shown
there is nowhere near the actual situation at the solid-vacuum interface. In absence
1.1 Introduction 5
Fig. 1.2 Field electron emission. Change the width of the potential barrier with increasing electric
field strength. U(x)—the potential energy of the electron
of an external electric field at the boundary there is a potential threshold rather than
a potential barrier (cf. Fig. 1.2a). One can also imagine that in this case, electrons
are retained in the metal by a semi-infinite potential barrier. Electron emission only
appears when electrons can leave the metal by passing the threshold due to the
energy received before. In particular, the energy can be received as a result of
heating the metal in the same way it happens in thermionic emission. The situation
is radically different if the metal/semiconductor surface is subjected to a strong
electric field. The potential threshold becomes a potential barrier. The higher the
electric field intensity, the narrower is the barrier (Fig. 1.2b, c). As a result, metal
electrons have a theoretical possibility to leave the metal due to the tunnel effect [1].
The triangular potential barrier represented in Fig. 1.2b, c, is closer to the actual
situation, and Fowler and Nordheim even used it in developing their first theory to
evaluate potential barrier penetrability [3]. Nevertheless, it is also highly idealistic.
Just like the electrons in the shell of any atom cannot leave the atom under usual
conditions (i.e. without an external action) due to the forces of attraction to the
nucleus, free electrons of a solid cannot (as we already mentioned) leave the solid
under usual conditions due to the decelerating field forces at the boundaries of the
solid. As established, these forces have a dual origin. Firstly, they are caused by the
fact that the electron leaving a solid polarizes electron gas (the term of polarization
has been used in a general sense, i.e. it includes both electrostatic induction for
conduction electrons and intra-atomic polarization for bound electrons). Secondly,
these forces are caused by double electric layers on clean surfaces of solids. From
the quantum-mechanical point of view, the double layers are explained by
“buckling” of w-functions on the crystal surface. From the classical point of view
double layers are caused by the fact that some free electrons leave the crystal
surface and develop an equilibrium electron cloud near the surface, thus creating a
responding deficiency of electrons on the surface.
Double layers also emerge when alien substances are adsorbed onto the surface
of a solid. Depending on their nature, they can form electric double layers of
various values and signs.
As escape of an electron obviously needs expenditure of energy to deform the
electron gas, there is less energy at the solid-vacuum interface before than after the
escape of the electron irrespective of the real shape of the potential barrier.
6 1 Basic Principles
The value characterizing the height of the potential barrier at the solid boundary
is the work function U.
Work function is defined as the minimum work necessary to remove an electron
from a metal. In this definition, the work function equals the difference in energies
of two states of the entire crystal. In the initial state, an electrically neutral metal
with N electrons is in its basic state with energy EN . In the final state, one electron
has been removed from the metal and has only electrostatic energy corresponding to
the level of vacuum Evac . The metal with remaining N 1 electrons passes into a
new state EN1 . Hence, the work function can be described as
U ¼ EN1 Evac EN :
For thermodynamic change of the state, the difference EN1 EN can be rep-
resented as a derivative of free energy F with respect to the number
@F of particles at
constant temperature and constant volume V. The derivative @N T;V is an elec-
trochemical potential Г of electrons. At nonzero temperatures it is equal to the
Fermi level EF :
@F
EN1 EN ! ¼C
@N T;V
or
@F
¼ EF
@N T;V
U ¼ Evac EF :
The work function from clean surfaces mostly comes down to work against
polarization forces. As for double layers, they are either small (e.g. in alkaline
metals) or compensated by high enough kinetic energy of electrons in the substance.
As for semiconductors or dielectrics, we have not found any discussion of double
layers on their surfaces in the literature yet. From the classical point of view, the
relatively low concentration of “free” electrons makes the double layers very small.
The quantum-mechanical calculations made for metals show that the polariza-
tion forces outside the metal come down to the classical “mirror force” (polarization
force) introduced by Schottky. According to him, the force retaining an electron in a
metal is caused by the fact that the escaping electron induces a distributed positive
charge in the metallic surface, and the field thus generated between the electron and
the induced charge is fully equivalent to the field between the electron and the
positive charge equal to the electron charge and located in the point of the electron
2
image (Fig. 1.3), i.e. F ¼ 4xe 2 : However, according to Schottky considers this only
1.1 Introduction 7
+e 0 -e
x x
holds starting with some distance x a nearly equivalent to the constant of the
metal lattice. At the distances 0 < x < a, the expression does not hold because of
the influence of the atomic structure of the surface. Schottky believes that at dis-
tances 0 < x < a, the polarization force can be regarded as constant and equal to the
image forceR in the point x = a. Therefore, the total work function of the metal
1
U ¼ eu ¼ 0 Fdx turns out to be equal to
e2
U ¼ eu ¼ : ð1:4Þ
2a
e2 e 1
U ¼ eu ¼ : ð1:5Þ
2a e þ 1
8 1 Basic Principles
Here the value of dielectric constant e shows only the electronic part of polar-
ization because the orientation part does not have enough time to manifest itself
within the electron escape period. The expression (1.5) for semiconductors and
dielectrics determines the value of their electron affinity. As it will be clear from the
subsequent, it is this value that matters for consideration of field emission.
If the surface of a metal or a semiconductor is covered with an adatom layer, and
the adatom is polarized and supplied with a dipole moment either due to interaction
between adatoms and the main substance or for some other reasons, then the work
function is changed by the value of potential energy surge caused by the dipole
layer (Fig. 1.4).
In that case, the work function can be expressed as the sum of two summands.
The first summand is connected with the bulk properties of the metal and deter-
mined by the electrostatic potential in the depth of the bulk. In this case, the Fermi
level counts as zero. In terms of a potential diagram (Fig. 1.5), it is U(−∞). The
second summand is a term sensitive to the surface properties and expressed by the
difference of electrostatic potential energies:
Z1
DU ¼ Uð1Þ Uð1Þ ¼ 4pe x½nðxÞ n þ ðxÞdx;
1
eφ
+
D
1.1 Introduction 9
Energy, eV
Fermi level
where n(x) is the electron charge density distribution, and n+(x) is the positive
charge density distribution.
This term is referred to as an electrostatic dipole barrier because it corresponds to
the work that has to be done to transfer an electron through the surface dipole layer.
The dipole moment value is a surface characteristic, which changes from surface to
surface. In one and the same metal, closely packed (atomically smooth) faces
usually have a larger dipole moment than loosely packed (atomically-rough) faces
do. It is understandable at the qualitative level if we imagine a rough surface as
containing some frame ions biased to the region of the charged negative layer,
which reduces the dipole moment, and consequently, the work function.
As a rule, the above also means that electron emitter surfaces are not homogeneous
in terms of work function. Inhomogeneity is caused either by the fact that the surface
is formed by different faces of a single crystal, or the different degree of coverage with
adatoms, or the polycrystalline nature of the surface. As a result, the contact potential
difference causes development of patches of accelerating and decelerating fields (spot
fields) on the surface. However, when a strong enough electric field is applied to the
emitter surface, relatively long-range spot fields cease to matter, while the local work
functions of particular surface patches remain significant.
Field emission Field emission is emission of electrons in the case when the emitter
boundary has a strong external electric field applied to it, which accelerates elec-
trons from the surface. A potential wall at the solid-environment interface is turned
into a barrier whose thickness depends on the field applied (the stronger the field,
the thinner is the barrier) (see Fig. 1.1). In this case, non-excited electrons can
tunnel this barrier, generating current whose density is a characteristic of this
10 1 Basic Principles
process. It is clear that the field emission current density largely depends on the
field intensity at the surface.
Thermal-field emission Strictly speaking, emission of non-excited electrons in the
course of tunneling takes place at zero temperature. If T [ 0, some electrons have
levels of energy E [ E0 For excited electrons, the potential barrier becomes lower
and, respectively, narrower, which increases their barrier-transmission factor. This
leads to growth of field emission current in comparison with the previous case. So
thermal-field emission is caused by both the processes of transmission of electrons
through a potential barrier, and the processes of their thermal excitation.
Photo-electron emission By analogy with thermal-field emission, electrons can be
additionally excited by irradiation of the emitter surface with light quanta. This is
extremely important in case of a semiconductor emitter since this excitation
transfers valence-band electrons to the conduction band. Imposition of the electric
field results in a combined photo-field emission.
Having considered the main types of electron emission, we can now present a
potential barrier model closest to the actual situation and, consequently, most often
used to calculate potential barrier penetrability before calculation of the basic
dependence of field emission j = j(E) [integral (1.2)].
As noted above, electron emission without pre-excitation (field emission) can
only take place as a result of a tunnel effect if a strong electric field is applied to the
surface, whereby the potential threshold (an infinite potential barrier) turns into a
finite-width potential barrier. A triangular potential barrier is based on the speci-
ficity of field emission, i.e. availability of an electric field. However, this barrier
model is too rough. By now, the most complete theory of field emission from metals
Energy
vac
=0 Distance from surface, x
F
Φ /eE
metal
Fig. 1.6 The diagram of the potential energy of the electron near the metal surface in the presence
of an external electric field with electric field strength E. The total potential (solid line) is the sum
of the image potential (dotted line) and the potential of the applied field (dashed line). Ф—work
function in the absence of an applied field. Reducing the potential for value dФ because the
Schottky effect. x0—position of the maximum total potential
1.1 Introduction 11
has been developed for the so-called polarization force or electrical image-force
barrier. A potential diagram explaining this model is presented in Fig. 1.6.
In subsequent paragraphs we will take it upon ourselves to develop a strict
quantum-mechanical assessment of the field emission phenomenon. In conclusion
of this paragraph, however, we have to mention that this phenomenon can also be
explained in terms of classical physics.
Actually, the external-field force acting on an electron is accompanied by an
attraction force from its electrostatic image in the metal. Effective potential energy
including interaction with the image is formulated as
e2
Ueff ðxÞ ¼ U0 eEx ;
4x
where x is the distance to which the electron moves from the metallic surface, the
distance between the electron and its image being 2x. The function of Ueff(x) peaks
pffiffiffiffiffiffiffiffiffiffi
at the point x0 ¼ e=4E, the potential energy at the peak being
pffiffiffiffiffiffiffiffi
Umax ¼ Ueff ðx0 Þ ¼ U0 e3 E :
or
pffiffiffiffiffiffiffiffi
e3 E U;
where EF is the Fermi level, U is the work function, then the electrons will be free
pffiffiffiffiffiffiffiffi
to leave the metal. According to classical views, the ratio e3 E U gives us the
electric field intensity at which field emission (e.g. for silver) is E 108 V/cm.
Meanwhile, experimental data (e.g. [17]) show that field emission can be observed
for fields at least an order of magnitude smaller. Hence, classical views cannot
adequately explain the field emission phenomenon.
The classic Fowler–Nordheim theory of field emission from metals (e.g. [3–6, 8,
913–15, 17–20, ]) is not only of historical interest. Mainly, this theory describes the
process of field emission from metals adequately in terms of quality and precisely
enough in terms of quantity despite considerable simplifying assumptions included
into the physical and mathematical models of the phenomenon.
12 1 Basic Principles
Before drilling down the central tenets of the Fowler–Nordheim theory, let us
state that any physical theory is a compromise between a strict physical model,
adequately selected calculation method and a possibility to obtain a result in the
simplest way and an analyzable form and turn our attention to its basic physical
prerequisites and assumptions used to simplify evaluations:
(1) the metal—vacuum interface is considered an ideal plane: the evaluation
problem becomes one-dimensional, the potential function (energy) U(x) de-
pends only on the x-coordinate, the external field is homogeneous;
(2) inside the metal U1(x) = const = − U0; outside the metal, the potential barrier
is caused only by activity of image forces U2(x) = e2/4x, where e is electron
charge;
(3) barrier penetrability D is computed by the semi-classic Wentzel-
Кramers-Brillouin (WKB) method; subsequent improvements of the theory of
field emission from metals were mostly connected to application of penetra-
bility computation methods exceeding the WKB method in validity;
(4) the metal model selected was the Sommerfeld model of free electrons in a
potential box evolving degenerated gas obeying Fermi-Dirac statistics;
(5) the Fowler–Nordheim theory was developed for the temperature T = 0 K, i.e.
for mere field emission.
The current density in the case of field emission can be calculated by the usual
formula that is also suitable for calculating current density for other types of
emission, such as thermionic, photo-electron etc.:
Z1
j¼e nðEx ÞDðEx ; EÞdEx ; ð1:6Þ
0
where nðEx Þ d Ex is the number of electrons with energy ranging between Ex and
Ex dEx falling from the inside of the metal onto the surface unit per second;
EðEx ; EÞ is a potential-barrier transmission factor at the metallic surface equal to the
ratio of intensity of the electronic wave passing through the barrier to the incident
wave intensity. Naturally, D depends on both Ex and the field value E:
spectrum into several regions and apply particular mathematical methods to each
region. Figure 1.7 shows an energy-level diagram for a metal and a potential barrier
at the metal-vacuum boundary in absence and presence of electric field. When
T = 0, energies of all electrons are concentrated in Region A. T > 0 causes elec-
trons to appear in the Regions B, C and D. Region B electrons take part in field
emission along with Region A electrons. Within Region B, there is a small sub-
group B1, where electron energies get close to the maximum of potential barrier.
Obviously, barrier penetrability for Region B electrons is higher than that for
Region A electrons. The electron emission caused by electrons in Region B is
referred to as thermal-field emission. Region C and D electrons cause thermionic
emission, Region C electrons are responsible for the Schottky effect and Region D
electrons cause emission without an external electric field. (For the sake of sim-
plicity, Figs. 1.7 and 1.8 show a potential energy curve passing through the point
−U at x = 0. Actually, it passes through this point at xk determined from the
2
equality U ¼ Exk þ 4xe k ).
Figure 1.8 shows the image force barrier in presence of an external accelerating
electric field E. Then, for a potential electron energy V in the presence of an external
field, the following ratios are valid
V ¼ U at x\0;
e2 ð1:8Þ
V ¼ 4x eEx at x a:
C (Ee3)1/2
B1
-eEx
-eEx-e2/4x
-(Ee3)1/2 (x>a)
-W
0 x1 x+ x0 x++ x2
In this case, the barrier summit is located below the vacuum level by the value of
pffiffiffiffiffiffiffiffi pffiffiffi
e3 E and at a distance from the metal; x0 ¼ 12 Ee ; x1 and x2 are the distances from
the metallic surface at which the set electron energy E is equal to potential energy.
The Schrödinger equation for the barrier shown in Fig. 1.8 is given by
2
d2w 2m e
þ E eEx w ¼ 0: ð1:9Þ
dx2 h2 g2 4x
where ai, ar, at are the amplitudes of incident, reflected and transmitted waves,
respectively.Let
2
e
UðxÞ ¼ k02 E eEx ; ð1:11Þ
4x
d2w
þ UðxÞw ¼ 0 ð1:12Þ
dx2
14 Zx
UðxÞ i 1
w¼ e ½UðxÞ2 dx ð1:14Þ
k02
This limiting form of solution for (1.12) is valid if the barrier (Fig. 1.8) contains
points x+ and x++ meeting the following conditions:
0 1 9
U ðx1 Þ 1 3 >
>
U00 ðx1 Þ jx þ x1 j U0 ðx Þ; >
>
1 >
>
3 =
jU0 ðx þ Þj
jUðx þ Þj2 ;
0 1 ð1:15Þ
U ðx2 Þ 1 3 >
U00 ðx2 Þ jx2 x þ þ j U0 ðx ;>
>
>
2 Þ >
>
0 þþ þþ 2
3 ;
jU ðx Þj
jUðx Þj :
Zx2
1 4
jUðxÞj2 dx : ð1:16Þ
3
x1
Rx 1
i ½UðxÞ2 dx
14
w ¼ aUðxÞ e ; ð1:17Þ
where the constant factor a and the lower limit of integration are chosen arbitrarily.
The solution (1.17) is valid given the following conditions:
2
1 U00 5 U0
UðxÞ [ 0;
1: ð1:18Þ
4U U 4 U
It can be shown that for the barrier described by (1.8), this condition is met well.
Calculating the left-hand side, we get:
3a0 1 þ 32 Ex
1: ð1:19Þ
3 e2
8x 1 þ 4 Ex2 3
e
The expression (1.20) differs from (1.17) by the fact that in the former the lower
limit of integration is defined and another wave phase is introduced for conve-
nience: (1.20) represents an electronic wave leaving the metal. It is invalid in the
negative kinetic energy field x1 < x < x2, since Ф < 0 [cf. (1.18)].
According to Kramers [24], in this region, the
individual function grows
R 1
exponentially from x2 to x1 if it behaves as U4 cos U2 dx þ p4 at x > x2 and
1
decreases along the same direction if cos is replaced with sin. So, in the field
x1 < x < x2
2 0 1 0 x 13
Zx2 Z2
14 4 1 1 1
wðxÞ ¼ ajUj exp@ U2 dxA þ i exp@ U2 dxA5: ð1:21Þ
2
x x
It is well-known that at x < 0 inside the metal, the function w has the wave nature
again. Therefore, the expression (1.21) can be represented as follows:
1.2 Fowler–Nordheim Theory of Field Emission from Metals 17
20 x 13
Z
1 p
1 þ e2Q exp4i@ U2 dx þ A5
14 1
wðxÞ ¼ aU eQ
4 4
x1
2 0 x 139 ð1:22Þ
Z =
1 p
þ 1 e2Q exp4i@ U2 dx þ A5 :
1
4 4 ;
x1
where
Zx2
1
Q¼ U2 dx: ð1:23Þ
x1
then
a 1 1
wðxÞ ¼ pffiffiffi eQ 1 þ e2Q eiðvx þ cÞ þ 1 e2Q eiðvx þ cÞ ; ð1:25Þ
x 4 4
i.e. inside the metal, the function w is the sum of the incident wave and the reflected
wave of almost identical amplitude. By definition,
2
1 14 e2Q e2Q 2Q
D¼1R¼1 2 ¼ 2 e ð1:26Þ
1 þ 14 e2Q 1 þ 14 e2Q
It follows from (1.27) that the barrier penetrability depends on the area of the
potential curve located above this energy level E.
The integral in the exponential curve (1.27) is a complete elliptic integral cal-
culated by Nordheim through its transformation to normal elliptic integrals.
If the expression for D is applicable to the case of a triangular barrier, which is in
general case not necessarily true, then it is easy to get:
pffiffiffiffiffiffi Zx2 pffiffiffiffiffiffi 3
2p 2m pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 2p 2m jEj2
Q0 ¼ jE þ eExjdx ¼ : ð1:28Þ
h 3 h eE
0
18 1 Basic Principles
Q ¼ Q0 hð yÞ; ð1:29Þ
where
pffiffiffiffiffiffiffiffi
e3 E
y¼ ð1:30Þ
jEj
is the ratio of reduction of the work function due to the Schottky effect (i.e. the
mirror forces) to the “work function of an electron with given energy”. Function
h(y) was tabulated by Nordheim (cf. Table 1.1), which makes it easy to calculate Q,
and, thus, D. Graph of h(y) is plotted in Fig. 1.9.
y
0 0.5 1
1.2 Fowler–Nordheim Theory of Field Emission from Metals 19
Z1 2
4pm2 ekT mvx
j¼ ln 1 þ exp E0 =kT Dðvx ; EÞvx dvx : ð1:31Þ
h3 2
0
mv2x
Let E ¼ 2 E0 . When T ! 0, the expression becomes much simpler because
E
E
ln 1 þ e kT ¼ at E\0; ð1:32Þ
kT
Z0
4pme
j¼ 3 DðE; EÞEdE: ð1:33Þ
h
1
To count the electrons emitted not only from the Fermi level E0 , but also from the
levels close to it, we expand the expression (1.29) into series in terms of E.
Eventually, that results in the expression
D ¼ D0 ekE ; ð1:34Þ
where
" pffiffiffiffiffiffi 3 pffiffiffiffiffiffiffiffi!#
8p 2m u2 e3 E
D0 ¼ exp h : ð1:35Þ
3 h eE U
The expression (1.37) is the basic formula of field emission, i.e. the Fowler
−Nordheim formula. If work function is measured in eV and electric field in V/cm,
then
" 3 pffiffiffiffi#
6 E2 6:85 107 U2 3:62 104 E
j ¼ 1:55 10 exp h : ð1:38Þ
U E U
Formula (1.38) shows that field emission depends on the electric field in the
same way as thermionic emission depends on temperature, i.e. it is greatly sensitive
to the value of the field. Just as lg Tj2 ¼ f T1 is a straight line in case of thermal
emission, the relation lg Ej2 ¼ f E1 here is an almost straight line as well. Little
deviations from the straight line are caused by presence of function h(y) in the
exponential quantity, but usually for the experimentally studied fields values of
h(y) vary insignificantly.
Field emission currents are already quite considerable at rather small D (ap-
proximately 10−10). When fields E = 6 107 V/cm and U = 4.5 eV,
j 107 A/cm2.
, eV Semiconductor Vacuum
4.0
ψ δφ
U0-ρ2e2/4x
U0-eEx
U0-eEx-ρ2e2/4x
0
δR
C
ζn
g
-1.2
υ
V x
x=0
Fig. 1.10 The diagram of the potential energy of an electron near the surface of the
semiconductor in the non-degenerate case fo electron gas. The value Ev ¼ ðE þ Eg Þ—energy,
measured down from the top of the valence band Ev on the surface. Numerical values are given for
silicon
U0
C
ζn δR
F
0
x
x=0
22 1 Basic Principles
value is affected by both the electric field penetrating into the semiconductor from
the outside and the surface charges. This effect is usually measured by the band dR
bending value. In this case, the distribution function is given by
E þ fn þ dR 1
f ðEÞ ¼ 1 þ exp ; fn [ 0; ð1:39Þ
kT
where fn is the energy-level separation from the Fermi level to the lower boundary
of the conduction band Ec in the bulk of the semiconductor. Both the
Morgulis-Stratton theory and later theoretical papers usually assumed that the bands
bend only in respect of the Fermi level, and the Fermi level remains invariable up to
the semiconductor surface (Figs. 1.10 and 1.11). Band bending is positive when the
bands on the surface are bent upwards, i.e. the electron concentration in the surficial
region of the semiconductor space charge is less than inside the bulk.
If Fermi level lies in the conduction band (Fig. 1.11 shows the degenerate case),
the Fermi distribution function is used (1.39). If Fermi level lies in the forbidden
band up to the surface (Fig. 1.10, non-degenerate case), one can use a simplified
expression (the Boltzmann distribution function):
E þ fn þ dR
f ðEÞ exp ; fn þ dR kT; ð1:40Þ
kT
where energy E is counted from the lower conduction band boundary on the
semiconductor surface. Formulas (1.39) and (1.40) regard dR as a parameter. The
band bending calculation methods have been written up in [6].
In case of semiconductors with a large specific resistance and at high field
emission current densities, thermodynamic equilibrium may be disturbed. In these
cases, we cannot use the distribution function given by (1.39).
Another important issue in development of the theory is selection of expression
for penetrability D in the equation similar to the one used to calculate dependence
j = j(E) for metals (1.2) and taking account of influence of the semiconductor band
structure on field emission.
The Morgulis-Stratton theory has been developed only for n-type low-resistance
semiconductors and for the region of relatively weak fields and currents in the way
quite similar to the one for metals. It is based on the assumption of thermodynamic
stationarity of the emission process and, hence, constancy of the chemical potential
level in the sample. The latter is only valid if just a few electrons falling on the
potential barrier tunnel in vacuum. The summed current through the sample is much
less than the drift electron current to the surface and the diffusion electron current
from the surface into the depth of the sample, so the respective approximation is
referred to as zero current approximation. In particular, zero current approximation
means that voltage drop IR of the cold emitter is zero, I being the total emission
current, R being resistance of the pointed emitter. Let us mention that for metals,
zero current approximation is valid in arbitrary strong fields, but for semiconduc-
tors, the region of its applicability is considerably narrower.
1.3 Classical Morgulis-Stratton Theory of Field Emission from Semiconductors 23
With zero current approximation, to solve the problem of the structure of the
surficial region of the space charge and field distribution in the sample means to
solve the Poisson equation with set boundary conditions.
However, when using the ratio (1.39) to calculate field current from the semi-
conductor, one has to bear in mind that the Nordheim function argument is to be
further multiplied by
U ¼ v þ eVs þ ðEc EF Þ;
where v is electron affinity, eVs describes band bending, and the term ðEc EF Þ
denotes the difference in energies between the Fermi level and the conduction band
bottom in the bulk. The dipoles caused by adsorbates affect the first two terms,
while semiconductor doping affects the third term.
The integrand in (1.2) has a sharp maximum located either in the neighborhood
of E ¼ EF or near the conduction band bottom. In the neighborhood of this
maximum, the integrand is expanded into the Taylor series and integrated term by
term. Taking into account that electron energy distribution of the conduction band
NðEÞ can be determined in several approximations, Stratton calculation results can
be summed up as follows:
24 1 Basic Principles
1=2
1 vaf E
NðEÞ ¼ n1 exp ; ð1:41Þ
2pmkT kT
where n∞ is the equilibrium concentration of electrons in the bulk, vaf is the affinity
barrier.
For this approximation, we can get:
" pffiffiffiffiffiffi 3=2 ( )#
kT 1=2 8p 2mvaf e 1 1=2 e3=2 E1=2
j ¼ en1 exp H : ð1:42Þ
2pm 3heE eþ1 vaf
4. Stratton also derived the emission current formula with account of surface states.
We are not going to present it in its complexity, but rather illustrate the main
results of the Stratton theory, including those attained with account of emission
from surface states (Fig. 1.13). Figure 1.13 prompts that, firstly, current voltage
characteristics of density of field emission current lg j = f (l/E) from an electron
semiconductor are shown with straight lines in all approximations, except for
the case when surface states have to be taken into account. Figure 1.13 shows a
few calculated current-voltage characteristics for SiC (curve 1 according to
(1.42), curve 2 according to (1.44), curve 3 according to the formula taking
surface states into account). The bending of current-voltage characteristics
arising in the presence of surface states is caused by the following. When fields
are small, field emission becomes lower due to the inner decelerating barrier. In
the process of cancellation of this barrier by a penetrating field, emission begins
growing in an accelerated way. Along with the external field growth, the
emission growth causes a simultaneous increase in both potential-barrier pen-
etrability and electron concentration in the surface layer.
According to (1.42), the field emission current density is proportional to the
electron concentration in this layer, i.e. j * n, if the electron gas of the conduction
band in the surficial layer is non-degenerate.
If field penetration is disregarded, n = n∞. While if we take it into account,
n depends on n∞ according to (1.43), i.e. in a more complex way. According to
(1.43), field emission current also depends on temperature [mainly, due to the
dependence n = f(T)]. For degenerate electron gas according to (1.44), j is not an
explicit function of the temperature.
To conclude this brief summary of the results derived in zero-current approxi-
mation, we should say that for the case of field emission from low-resistance n-
semiconductors and in the field of weak fields and currents, the Morgulis-Stratton
theory is satisfactory for description of experimental facts (linearity of
current-voltage characteristics, absence of photosensitivity and weak thermosensi-
tivity) and allows calculation of emission-current density. However, experimental
research has not identified any dependences lg j = f(1/E), e.g. curve 3 in Fig. 1.13.
The straight current-voltage characteristic slope constructed in coordinates lg j,
E−1 (Fig. 1.13) is determined by affinity to the electron, the bandgap width and
degeneration parameter. In a considerable concentration of surface states, the
current-voltage characteristics of n-type semiconductors can display characteristic
bending (Fig. 1.13). Their effect is that when intensity of external field E is low, the
26 1 Basic Principles
lgJ
0
2
-1
-2
-3
1
-4
-5
-6
-7
-8
3
-9
-10
2.5 3.0 3.5 4.0 4.5 5.0
Fig. 1.13 The calculated current-voltage characteristics for SiC: 1—of formula (1.42), 2—of
formula (1.45), 3—calculation that takes into account the surface states
surface-charge field bends the allowed energy bands upwards, and emission initially
comes from the valence band. As soon as the field reaches the value when the
surface-charge field is canceled, the bands start bending contrariwise, conductance
electron concentration goes up and field emission from the conduction band
becomes prevailing. On the current-voltage curve, emission moves from the valence
band to the conduction band, which is accompanied by upsurge of current.
First attempts to develop a theory of field emission from n-type high-resistance
semiconductors and p-type semiconductors were not made until after it had been
established that current-voltage characteristics in this case are non-linear, and
emission is photo- and heat-sensitive (e.g. [5–7, 29, 30]).
As for the theory of thermofield emission (TFE) from semiconductors, which
was most fully developed by Vasiliev [31], it has no practical sense, as the
1.3 Classical Morgulis-Stratton Theory of Field Emission from Semiconductors 27
experiment shows (e.g. [29]) that heating of a semiconductor field emission cathode
to quite low temperatures * 200 400 °C results in significant distortion of its
initial properties. For this reason, presentation of even some particular elements of
the theory of thermal-field emission from semiconductors is of no special sense,
which is why we are not going to do that.
(4) A strong external electric field and high temperature change the emitter shape,
and that is easy to detect both by emission pattern and by change of the
emission current. By the emission pattern one can directly see the exit of
impurities out of the bulk of the crystal.
(5) A field emission microscope can be used to study the phenomena of epitaxy,
crystal irregularities, emergence of dislocations etc.
(6) Field emission microscopy allows to evaluate the work functions, i.e. to make
quite accurate measurements of relative changes in work functions for various
crystallographic planes, to evaluate surface tension factors, activation energy
for migration, energies of adsorption and desorption and to determine
energy-distribution half-widths with high resolution etc.
However, all those advantages of the field emission method do not come without
drawbacks:
(1) Field emission results from the tunnel effect in electric fields of about
107 V/cm. As a rule, the emitter is shaped as a very thin tip with a curvature
radius of 10 100 nm on the top to provide the electric field intensities
necessary for emission. That makes it impossible to take measurements of the
same object by other methods and compare the results.
(2) The phenomena taking place on the field-emitter surface can only be observed
under very strong electric fields. Extrapolation of the results to the case of
infinitely small field is most often incorrect.
Figure 1.14 shows a diagram of a model of the most primitive field emission
microscope (a field emission diode). The anode is almost spherical as the shape of
the anode is of no importance. The emitter tip (terminus) is situated in curvature
center of the anode. Electrons emitted from the termination have very little initial
kinetic energy and their paths are determined by spatial distribution of the applied
acceleration field, this distribution being almost spherical in immediate proximity to
the hemispherical surface of the tip. That is the case for a metal emitter, whose
surface is an equipotential of an electric field and is directed perpendicular to its
power lines. The field value decreases in a close proportion to the squared radial
distance from the terminus. So the electrons acquire their terminal velocity at a
distance of several peak radii from the emitting surface, going further to the anode
along the radial paths [23, 36]. The anode is phosphor-coated, which turns it into a
luminescent screen. Falling electrons activate the screen and it generates a mag-
nified image of the terminus surface. The magnification factor is proportional to the
ratio of the distance between the terminus and the screen to the tip radius; its
practical value is usually in an order of 105.
In Fig. 1.15 one can a simple representation of a field electron motion trajectory
as a continuous curve. Let us note that this diagram is not in a true scale. In reality,
the distance from the emitter terminus to the screen equals several centimeters,
which is almost 105 times as long as the radius of the terminus. The radial and
tangential components of the initial electron velocity are designated with иr and иt.
Since иt is non-zero, the electron reaches the screen in a point biased by the value of
1.4 Basics of Müller Field Emission Microscope Theory 29
Tip emitter
D/2
Ut
rt
δ/2 Ur
D/2 from the radial projection of the point of field-electron escape from the surface
of the cathode. The calculations made by R. Homer [37] showed that
where 〈ut〉 is a mean tangential component of the initial velocity of emitted elec-
trons and t is given as
t xð2eV=mÞ1=2 ; ð1:47Þ
where x is the distance from the terminus to the screen, V is the voltage applied
between the emitting cathode and the anode. 〈ut〉 can be estimated as
1.0
оо ++
оо
0.8
++
0.6
M(θ)/M0 2
β(θ)/β0
0.4
0.2
2 20 40 60 80 100 θ, degrees
Fig. 1.16 Dependence of the local field factor b (circles) and the multiplication M (crosses) of the
azimuthal angle h at h = 0, b = b0, M = M0. The dashed curves are calculated for two models of
the sphere on the orthogonal cone, simulating two different emitter profiles: with a pronounced
(1) and small (2) compression. The solid curve corresponds to the average emitter
hEt i ¼ 0:433Et1 3:79F 1=2 =U U1=14 ; ð1:49Þ
where t is the elliptic function, U is the effective work function, E is the field on the
tip surface (in volts per ångström), which can be found from the ratio
E ¼ bV V=krt ; ð1:50Þ
k 5 at the emitter terminus and grows with increase of the polar angle
(Fig. 1.16).
Figure 1.14 shows that the observer looking at the screen perceives an arbitrary
small segment on the tip surface as stretched by the value of
M ¼ x=crt ð1:52Þ
1.4 Basics of Müller Field Emission Microscope Theory 31
where c 1.5. From (1.47) to (1.52) we can find that the resolution of the field
emission microscope (in ångströms) is
1=2
d 2; 62c rt =ktU1=2 ; ð1:53Þ
where rt is the tip radius (in ångströms). It is estimated that for a typical emitter it is
about d 25 Å.
Along with (1.53), there is another condition that constraints the resolution of a
field emission microscope. It is caused by the Heisenberg uncertainty principle,
from which follows that if an electron is emitted from the region of width d0 it has
to have the minimum tangential component of velocity of at least ut ħ/2тd0. This
principle imposes an irremovable limitation on resolution of any microscope.
However, the estimates show that in most cases this limitation can be neglected as
the resolution estimate is not much different from the estimate derived from the
ratio (1.53).
As to identification of the basic crystallographic faces displayed on the lumi-
nescent screen of the field emission microscope (Fig. 1.17), it is carried out with
standard stereographic (or pseudo-stereographic) projections [18]. For the basic
stereographic structures (BCC, FCC, HCP), they can be found in monographs and
textbooks (e.g. [38]) or manuals on crystallography. When deciphering structure
images, they have to be calculated on computer. In the field emission microscope, a
sample (a tip field emission cathode) is shaped and deformed in the process of
surface image acquisition and the deformation nature can be of two kinds, both
elastic and plastic. If the field meeting the conditions of the best mapping of the
sample surface equals E0, then the crystalline lattice of the sample is elastically
extended and deformation
Fig. 1.17 Examples of typical field emission images of the tip surface of tungsten (a, b) and
niobium (c): a—the surface is smoothed high-temperature heating (T0 ≅ 0.7 Tmelting) without a
field; b, c—surface reconstructed with a slight warming (T0 ≅ 0.2 Tmelting) in the presence of an
electric field
32 1 Basic Principles
where K is the volumetric modulus of elasticity usually equal to a few percent. The
respective value of the elastic energy accumulated in the sample per atom equals
where Xa is atomic volume. Since Eat is usually of the order of 0.1 eV, i.e. com-
parable with activation energy of low-temperature diffusion processes, impression
of an electric field in the field ion microscope can considerably change the nature of
pin-hole defect annealing.
Approximating the tip shape with a lathe object having symmetry axis z and
neglecting particular atomic and electronic surface structures acting on the plane
z = z0, we can express force Fz as
Z Z
Fz ¼ ½F 2 =ð8p cos aÞdSe ; ð1:56Þ
where dSe is an element of the tip surface, a is an angle between the normal line to
this element and axis z. Integration is performed over the entire surface, i.e. from the
tip terminus to the chosen section plane. As a result, (1.56) can be represented as
follows:
Zr0 ZF0
Fdr F2
Fz ¼ 2pdl ¼ rdr; ð1:57Þ
8pdl 8p
0 0
where dl is the linear element of the surface, r0 is the distance from the tip centre to
z 0.
Another “concurrent” process is bombardment of the sample surface with
low-energy (several dozen kiloelectronvolts) ions of residual gases. In absence of
an electric field, the tip surface experiences n0 gas atom impacts per time unit
caused by their mere thermal motion:
1=2
n0 ¼ S0 Pg 2pMkTg ; ð1:58Þ
where S0 is the tip surface area; M is the gas particle mass; Pg is the gas pressure
and Tg is its temperature. In the electric field of intensity E, the gas-molecule flow
from the tip surface increases п* times, which is connected with polarization of the
molecules. n* can be approximated from the equation
n pag E 2 = 2kTg ; ð1:59Þ
X X Pi IFE
nu ¼ nu;i ¼ 1; 44 R0 ri ðWÞ ; ð1:60Þ
i i
kT0 e
where Pi is the partial pressure of the residual gas components; k is the Boltzmann
0 is the mean radius of
constant; T0 is the tip temperature; e is the electron charge; R
the tip surface curvature; ri (W) is the section of ionization equal to
Ne pa20 W1
ri ðWÞ ¼ 2
2:66 lnð1:25WÞ; ð1:61Þ
Vion W2
where Ne is the number of electrons in the outer shell of the ionizing atom of the
residual gas, a0 is the first Bohr orbit radius, Vion is the atom ionization potential,
and W ¼ Ee =Vion , where Ee is the electron energy equal to the value of the potential
in the atom ionization point.
The bombardment of the field-emitter surface with low-energy ions of residual
gases can have both negative and positive value. In the practice of operation of
cathode devices on the basis of field emission (in the same way as in the case of
research with field emission microscope), it can result in a change of the surface
structure, pulverization or redistribution of the component (film) activating electron
emission, vacuum breakdown and a total outage of the emitter (sample). At the
same time, the very possibility to irradiate the sample in the field emission
microscope looks undoubtedly fruitful and gives a wide range of opportunities.
Speaking about the bombardment of tip surfaces with low-energy gas ions
functioning in the field emission mode, it has to be noted that such a bombardment
is among the main causes of low-frequency emission noises. As such, let us also
dwell on the issue of the noises a bit.
The current emitted by the sample in the field emission microscope is somehow
labile: in addition to the usual “schrot” noise component natural for any kind of
particle emission, the field emission current has another relatively low-frequency
noise component. The change of the spectral power density W(f) of the latter
component is close to dependence 1/fc, where c ranges between 0.8 and 1.2. That is
the so-called flicker noise of field emission.
As is known, when studying fluctuation of the current, the main values to be
analyzed were the correlation function (auto-correlation) f*(s) and the noise-power
spectral density function W(f):
Zþ 1
1
Wðf Þ ¼ f ðsÞ expð2pIf sÞds: ð1:63Þ
2p
1
In addition, the ratio between the mean square of the current intensity fluctuation
ðDIÞ2 , the correlation function and the noise-power spectral density function is
given by the expression
Z1
2
DI ¼ f ð0Þ ¼ Wðf Þdf : ð1:64Þ
0
It is widely agreed that the main cause of field emission flicker noise is the
presence of adsorbed particles on the field-emitter surface and their irregular surface
movement [39].
Another “concurrent” process that has already been mentioned is the
thermal-field change of the tip surface shape. The joint action of heat and the
electric field (the latter manifesting itself through surface atom polarization and
movement of the resulting dipoles in the locally-enhanced field region, whereas
heating of tips enables, facilitates and boosts such mobility) can significantly
change the tip surface shape. This change has to be controlled and taken into
account when studying these or those phenomena in the field emission microscope.
Real surface geometry. Various authors successfully resorted to field emission
microscopy to analyze both the equilibrium shape of the metal microcrystal surface
at different temperatures in and out of an electric field, and to study the proper atom
condensation processes on an approximately hemispherical substrate. Below we’ll
give a consistent consideration of the most important results attained in these
aspects.
To begin with, the field emission microscopic image itself easily enables to
establish the mean curvature radius R 0 of the tip terminus. In case of field emission
microscopy, R 0 is determined by measuring of dependence of field emission current
IFE on tip potential V0. Having expressed the Fowler−Nordheim equation in terms
of IFE and V0 as
" !#
IFE a 0 cV01=2
bR
¼ 2 exp t ; ð1:65Þ
V02 R 0 V0 1=2
R 0
where parameters a, b and c are constant for this material and connected with the
value of the mean work function U, t is a tabulated function, we can see that
2
dependence ln(IFE/V0) of 1/V0 is close to the straight line up to ln a=ðR 0 Þ2 and its
slope equals bR 0 t. At t 1, value R
0 can be approximately determined from the
graph ln(IFE/V20) = f(1/V0) by the line slope. The determined value R0 can be later
1.4 Basics of Müller Field Emission Microscope Theory 35
used to calculate the argument of function t and then value t can be determined
from the table. Having inserted the new value t in the ratio bR 0 t equal to the
straight line slope, we can determine a more precise value R 0 etc. Precision of this
method of finding R 0 is ±15%. It cannot be improved in view of essential varia-
tions of the work function and, consequently, IFE along the surface of the tip
terminus.
As shown in [40], impact of temperature leads to spheroidisation in tip apex.
Let us note that the mass transfer leading to evolution of the tip terminus profile
is most probably caused [18] by surface self-diffusion. This seems feasible given
the small size of tip apex.
At the same time, we can see an absolutely different pattern of evolution of tip
terminations when they are heated in the presence of an external electric field [18].
In this case, apart from the capillary forces we have to take into account the
ponderomotive ones as well. In this case, the chemical potential of atoms at the
needle tip can be given by:
The minimum value of voltage V0 (tip potential) at which the electric field can
cause a considerable change in the tip shape against the background of capillary
forces can be estimated with the following expression:
V0 ð16pcR=vÞ1=1:4 : ð1:67Þ
Dependence V0* = f(R) means that in the course of needle elongation under the
influence of the field, the termination curvature radius decreases and, consequently,
the difference between the set value of potential V0 and its critical value increases,
which means that elongation of the tip has to be a self-accelerating process.
Analysis of thermal-field and condensation growth patterns of metal crystal tips
by field emission microscopy is presented in more detail in [39–41].
distinctions, and, at the same time, it is extremely interesting from the standpoint of
some important applications of field emission. Because of this, we are going to
analyze it in greater detail.
In addition, the classical Fowler–Nordheim theory considers only the case of
“pure” field emission (which, strictly speaking, only holds for T = 0 K). It does not
analyses theoretically interesting and practically important cases of thermofield
emission for T > 0 K nor for T 0 K.
Noteworthy is the following often neglected circumstance. The basic formula of
the Fowler–Nordheim theory does not link the voltage and current, but rather the
electric intensity and current density, i.e. basically unmeasurable values, especially
if the tip dimensions are microscopic. The most common workaround for this
probles is to assume that the field intensity value is proportional to voltage and the
current value is proportional to its density, i.e.
where b is the so-called field factor and S is the emitting area. The first relation does
not usually cause doubt if values b and E are understood as their local values in an
emitting surface point. As for the second equation, it is more precisely written down
as
Z
I¼ jðx; y; zÞdS; ð1:69Þ
ðSÞ
doubts that are not entirely unbased [45, 46]. Analysis of this type of criticism is
beyond the purpose of our study, though some issues concerning the physical
prerequisites of tunneling calculation with the WKB method will be considered in
further chapters. Secondly, correctness of substitution of the elliptic integrals as
components of the full formulas with their approximate values is open questionable
and improved expressions are being proposed [47, 48]. Thirdly, doubt is caused by
the formula parameter determination methods based on construction of current-
voltage characteristics in coordinates {ln(I/V2), l/V} with approximation by the least
square method [44].
The critics’ arguments are actually quite feasible. Indeed, application of the least
square method is connected with a particular (Gaussian) kind of random error
distribution [49]. It is known that any transformation of variables violates the initial
formula for the distribution function [50]. However, it still remains to be seen how
significant this violation is. This question is even more important if we note that
there are a lot of approximations initially built into the model, which may cause
much more serious errors. It is also important that according to the Neumann-David
theorem, the estimate of the linear-regression coefficient by the least-square method
is asymptotically unbiased and effective irrespective of a particular formula of the
error-distribution function [51]. Certainly, modification of the formula of the
error-distribution function causes change in the estimates of confidence intervals.
However, in this case, we should speak not about the biased estimate of the physical
model parameters, but rather about determination of the confidence interval for this
estimate. We cannot but note that the latter is very seldom, if ever, done.
The papers we know of have not payed enough attention to influence of an error
in the independent variable. The classical least square method assumes that a
random error burdens the measurements only in case of a dependent variable, while
the independent one is considered to be measured in an absolutely precise way.
Certainly, that is not the case. In practice, voltage measurements also bear a random
error comparable with a current-measurement error (of course, in relative terms).
Thus the usual formula of the least square method is mathematically unreasonable
here. What we need is a more complex model that’s still based on minimization of
residual sum of squares but taking into account the errors of both variables. So the
first stage of work should be analysis of applicability of Fowler−Nordheim for-
mulas or their improved analogues, comparative estimate of efficiency in terms of
computer time consumption and, finally, definition of the physical sense of the
values obtained. The last stage is also important if the results attained are not the
end in itself or a reference point for relative comparisons but the initial data for
further mathematical or natural simulation or even practical use.
For such analysis, we developed an information expert system [20] and carried
out a natural experiment on measuring of dependence I(V) in a wide range of
current voltages and densities, the latter exceeding five orders of magnitude. The
results attained were processed in various ways:
38 1 Basic Principles
10-3
10-4
10-5
ln(I, A)
10-6
10-7
10-8
6000 8000 10000 12000 14000 16000 18000 20000
V, V
Fig. 1.18 The dependence of the field emission current from the voltage: the experimental points
and the curve of non-linear regression
1.5 Disadvantages of Classical Field Emission Theories … 39
Table 1.2 Regression-curve parameters ln(I/V2) = A(I/V) + B obtained for the available
experimental points by various methods and their confidence intervals with 95% confidence level
Type of regression A DA B DB
Classical −1.158 105 0.007 105 −18.85 0.07
With error V −1.161 105 0.007 105 −18.81 0.08
Non-linear −1.052 105 0.02 105 −19.67 0.09
Non-parametric −1.143 10′ 0.03 105 −19.0 0.1
(4) The influence of the surface charges and field penetration on concentration of
electrons in the surface region was described in terms of band bending. Possible
quantization of electron states in the bulk-charge band near the surface was not
taken into account.
(5) We discussed diversions from thermodynamic equilibrium. The results fur-
nished were attained in the assumption that it is only acoustic-phonon scattering
that matters. We considered field emission from n-type semiconductors without
account of surface states. However, the experiments found out hot-electron
emission from p-type semiconductors as well [6, 7].
(6) The influence of effective mass was studied in the assumption that emission
preserves the tangential component of the wave vector. Essentially, the latter
matters when band boundaries in the space of wave vectors k lie at k 6¼0, e.g. in
the case of germanium and silicon. Preservation of the transverse wave-vector
component was admitted both in the semiconductor photoemission theory [9]
and in the general consideration of penetrability. Silicon photoemission
experiments [9] justify such an assumption for at least a large number of
emitted electrons.
For the sake of theory we’d prefer to preserve the transverse wave-vector
component unless the translation symmetry of the crystal is broken on the surface.
That is observed in the case of a flat boundary, when external forces are perpen-
dicular to the surface [5–7].
Silicon field emission experiments [6] indicate possible non-preservation of the
transverse wave-vector component. We gave the following prospective reasons
thereof:
• diffused scattering of electrons on the surface due to its roughness;
• violation of periodicity of the crystal potential on the surface;
• interaction with surface phonons.
Stratton explored this issue and found out that only simultaneous presence of
first and second factors can result in non-preservation of the transverse wave-vector
projection.
It is clear from what has been said that the classical Morgulis-Stratton semicon-
ductor field emission theory, however improved and enlarged in a number of later
works it might be, is still connected with a lot of essential simplifications. Therefore,
it is no wonder that the theory is still unable to explain a lot of experimental results in
terms of quantity. In a lot of cases, qualitative explanation is constrained as well.
References
1. N. Mott, J. Sneddon, Wave Mechanics and Its Applications. M.: Nauka, (1966), 427 p
2. A. Messia, Quantum Mechanics (in 2 volumes). M.: Nauka (1979)
3. R.H. Fowler, L. Nordheim, Electron emission in intense electric field. Proc. Roy. Soc. A. 119
(781), 173–181 (1928)
References 41
42. M. Lampert, P. Mark, Injection Currents in Solids. M.: Mir (1973), 416 p
43. R.G. Forbes, Refining the application of Fowler–Nordheim theory. Ultramicroscopy. 79,
11–23 (1999)
44. E. Hantzsche, Beitrage zur Plazma-physik, 22, 325 (1982)
45. G.D. Yakovleva, Tables of Airy functions and their derivatives. M.: Nauka (1969)
46. Y.V. Kryuchenko, V.G. Litovchenko, Computer simulation of the field emission from
multilayer cathodes. JVST. B14, 1934–1937 (1996)
47. M. Hollander, D. Wolfe, Non-parametric statistical methods. M.: Finansy i Statistika (1983),
518 p
48. S.G. Rabinovich, Measurement Errors. M.: Energia (1978), 261 p
49. Y. Linnik. The method of least squares and the foundations of mathematical and statistical
evaluation of parameters. M.: Fizmatgiz (1962), 350 p
50. A. Brunetti. A fast and precise genetic algorithm for a non-linear fitting problem. Comp. Phys.
Comm. 124, 204 (2000)
51. Enrico Fermi. Quantum mechanics. M.: Mir (1965). 242 p
52. Bohr. Atomic Physics and Human Knowledge. M.: GIIL (1961)
53. M. Kendall, A. Stewart, The Theory of Distributions. M.: Nauka (1966), 520 p
54. J.N. David, Probability Theory of Statistical Methods (Cambridge, 1951)
Chapter 2
Experimental Equipment and Technique
Abstract This chapter presents the field emission electron microscope, its tech-
nical characteristics and various modern construction of microscopes. A particular
attention is paid to analyzing full energies of the electrons and universal con-
structions of field electron microscopes that are the main branch of development of
microscopy and spectroscopy. The most popular practical techniques of manufac-
ture of tip field emitters from different materials.
A field emission microscope, invented by Müller in 1937 [1, 2], is a powerful and to
a certain degree one-of-a-kind tool to research emission features of a field-emitter
surface with high-resolution. First of all, those features include adsorption and
desorption kinetics, distribution of field electrons by energies, distribution of an
electron work function over an emitter surface, surface diffusion and eventually,
testing of field emission cathodes.
Some functions of structural research of field emission microscopy have grad-
ually moved to field-ion microscopy (e.g., surface diffusion research), which has
atomic resolution. However, emission and energy characteristics are still the
domain of field emission microscopy.
The basic circuit of a field emission microscope has not changed since the its
invention (Fig. 2.1). Current leads (1) are welded to a molybdenum or tungsten arc
(2), to which, in turn, the sample (3) is attached. This design makes it possible to
heat the sample up to the necessary temperature. The visor window of a glass bulb
has a luminescent screen (5), where the field-electrode image of the emitting surface
of the sample is formed. The luminescent screen is a transparent current-conducting
coating applied to glass with a thin phosphor layer. High positive voltage is applied
to the screen (5) through a current lead (6). The high-energy electrons emitted from
the emitter bombard the phosphor layer, causing its luminescence. Thereby, the
screen displays a pattern of distribution of electrons from sharp bands of the emitter
surface or the regions with a relatively low electron work function. Gas is evacuated
from the flask through an exhausting tube (7). The microscope bulb can work both
under continuous gas evacuation and as a sealed-off device.
The exponent of Fowler–Nordheim equation has the factor of 6.8 107 in the
numerator. Since the electron work function u of most materials, including
adsorbed systems, belongs to the range of 2–6 eV, it is clear that field emission can
be appreciable if the fields are about 106–107 V/cm. The easiest way to generate
such fields is to make a surface with large curvature, e.g. a needle with a very small
apex curvature radius. In this case, very high intensity of the electric field is
generated at the tip of the tip under moderate voltages. Electric intensity for an
insulated charged sphere is determined as
V
F¼ ; ð2:1Þ
r
V
F¼ ; ð2:2Þ
kr
these fluxes are perpendicular to the apex surface and hence radiating outward.
Hence, the increase is determined as
D
M¼ ; ð2:3Þ
br
A much more important factor that actually determines the resolution is the
statistical spread of electrons in tangential velocities in relation to the apex-screen
axis. For tangential velocity V at the emitter surface, distance d from the initial
electron motion direction on the screen is
2.1 Field Emission Microscopes 47
d ¼ 2Vst; ð2:4Þ
where t is the time for an electron to travel from the emitter to the screen.
Factor 2 appears because the tangential motion “throws” the electron away from
the electric flux, so the electron gets tangential acceleration and an additional bias
from the initial electric flux. The electron transit time can be sufficiently well
approximated as
t ¼ Dðm=2VeÞ1=2 ; ð2:5Þ
where V is the voltage applied because the entire electron path meets its total
kinetic energy Ve.
Half-breadth of resolution of a field emission microscope is defined by the
following formula [4]
1=2
d=2 ¼ 1:31 104 b r=kru1=2 ; ð2:6Þ
The designs of field emission microscopes are numerous because they are not issued
commercially. Which is most important, each experimenter develops the micro-
scope design according to the purpose of his/her experiment and proceeding from
his/her resources.
Figure 2.4 shows classification of the basic structural components of field
emission microscopes. Being relatively conventional, this classification describes
only the most important of them.
As a rule, the design of any field emission microscope contains from one to five
elements specified.
This section gives some examples of designs of field emission microscopes,
disclosing interrelation between the elements aimed at achievement of specific
research goals and giving an idea of general trends in designing of field emission
microscopes.
For a long time, since the very invention of field emission microscope by Müller
in 1937 [1] and nearly up to the late 1920s, most designs of field emission
microscopes were constructed out of glass, as shown in Fig. 2.1. Usually the flask
was made of high-melting glass S53, which is a so-called molybdenum glass, i.e.
the linear expansion factor is almost equal to the molybdenum linear expansion
factor and electrical outputs are made of molybdenum. Since such devices are used
48 2 Experimental Equipment and Technique
to research apex emitters, the working vacuum in the device should be at least
10−9 mm Hg. Such pressure is achieved by warming to the point 450 °C for more
than 8 h [5] and use of getters that maintain vacuum in the course of operation. For
adsorption research, glass sealed off field emission microscopes are usually sup-
plemented with adsorbate sources (e.g. [6, 7]). With development of vacuum
technique and gas evacuation equipment, fully sealed-off devices were replaced first
by glass metal-flanged devices connected to evacuation systems and then by fully
metallic demountable constructions.
2.1 Field Emission Microscopes 49
the heavy current of field emission destroys the conductive coating and the lumi-
nescent layer. Even replacement of the conventional glass with an
yttrium-aluminum garnet plate with an addition of cerium, that has a much better
thermal stability, does not solve the problem. For these purposes, a special
microscope (Fig. 2.6), was designed on the basis of DN50 flange (2).
All elements of the microscope are mounted on four rods of the ceramic current
leads (1). The anode system is a luminescent screen (5) fixed between two rings (4)
and a tungsten grid (6). A reliable electric contact between the conductive coating
and the lower ring (4) is provided by an Aquadag layer. A pre-annealed grid is
strained between the two rings (7). The grid is additionally strained by three screws
(9) and a special ring (8). The distance between the grid and the luminescent screen
is 1.5 mm. The field emission cathode (10) fixed in the holder (11) is at the distance
of 1.5 mm from the grid.
In such a design, the grid assumes more than 90% current draw without essential
distortions in the field emission image of the sample surface.
Tenacity of the emitters and their structural components can be determined with a
special design intended for other purposes as well (Fig. 2.7). The device can hold as
many as four samples at the same time (3). High voltage from the high-voltage
rectifier is applied to the anode (2), which is a transparent disk with a conductive
2.1 Field Emission Microscopes 51
Fig. 2.7 Field emission projector for stress tests in an electric field: 1 Kovar ring; 2 luminophor
screen; 3 field cathode; 4 nickel holder; 5 ceramic insulator; 6 current lead-ins; 7 Kovar adapter;
8 housing; 9 Viton laying; 10 fixture flange; 11 flange with current lead-ins; 12 contact brush;
13 glass window; 14 Kovar adapter; 15 setup flange
coating and a luminophor. In this design, the role of the anode (2) is different from
that of the observation window (13). The entire anode system is electrically insu-
lated from the chamber case (8) by a high-voltage insulator (5) rated for 30 kV.
Performance of the flange (11) with current leads (6) is improved due to Viton
sealing (9).
Tenacity of structural components of carbon fibers is determined by the method
of ponderomotive electric field loads [8, 9]. In this case, mechanical tension is
calculated by the formula
E2
r¼ ; ð2:7Þ
8p
Then the screen is fed with a negative voltage pulse. After that, a new value of
field emission current and a new field emission image are recorded under direct
voltage with the amplitude being constant during the experiment. A control sample
is put under the microscope. The sample is fed with pulses of only positive polarity
and a constant amplitude. It is intended for accurate differentiation of destruction of
the fibers caused by ionic bombardment under direct voltage and ponderomotive
loads under reverse voltage. The sample tested is fed with a series of pulses of direct
and reverse polarity. The amplitude of the reverse-polarity pulses is increased by
200–400 V per pulse. For every direct-polarity pulse, a new value of field emission
current and a respective field emission image are recorded. Supply of pulses ceases
when the field emission current of the sample reaches the range of 1–5 µA.
The value of the electric field at the microelevation to be destroyed is determined
through the density of the field emission current from the microelevation before the
destruction. The electron work function of the carbon-fiber material with different
temperatures of thermal treatment is approximately the same. It amounts to about
4.6–4.8 eV. The change in the current caused by destruction of the microelevation,
which is identified by field emission images, can be calculated from the flow chart
of the currents as a difference between the currents corresponding to the moments
before and after the destruction.
Then, determination of the microelevation emitting-surface area necessitates
making a number of assumptions. The most important among them is that the size
of the emitting microelevations by the order of magnitude is that of the fibrils. Such
an assumption is quite justifiable because the ionic bombardment results in dis-
closure of the fibril structure of the fiber, and the fibril size is known to amount to
30–100 Å. Knowing the microelevation size, we can express the emitting area for
each particular microelevation as
S ¼ r2 ; S ¼ r2 ;
E2 2
r¼ k ð2:8Þ
8p
where E is the value of the electric field at the microelevation, k is the factor equal
to the ratio of the negative and the positive pulse-voltage polarity amplitudes at the
moment of destruction of the microelevation.
The field emission current is lowly dependent on temperature; however, its value
is among the key operational parameters of electronic devices. Therefore, it is an
urgent problem to study the operating temperatures of field emission cathodes.
2.1 Field Emission Microscopes 53
The temperature of the apex field emitters welded to the arc of a high-melting
metal (W, Mo) can be measured by change of electrical resistance of the particular
arc section [10]. A more precise temperature value of the emitting apex tip can be
calculated by the design formulas [11]. If the cathode is fixed in a different way or,
which is even more important, more massive cathodes are used, such methods of
measuring temperature are improper.
The temperature of graphite field emission cathodes with a well-developed
effective surface of tens of square millimeters can be determined [12] with a direct
method of measuring by means of thermocouple reference junctions.
The structure of the experimental device (Fig. 2.8) [13] is mounted on DN50
flange (5). The cathode (7) (a rod 12 mm long, cross-section 1.5 1.5 mm) is
mounted at the distance of 0.1 mm from the anode. The cathode is equipped with
two thermocouples (the experiment involved copper constant thermocouples
manufactured from the wires of 30 lm in diameter). One thermocouple (8) is
mounted in the groove filled up with Aquadag at the distance of 0.5 mm from the
effective surface of the field emission cathode, and another one (11), intended to
study distribution of temperature along the length of the cathode, at the opposite
end face of the cathode. The anode (1), as well as the field emission cathode, is
made of MPG-6 graphite. As the anode in our experiments is under high voltage,
Fig. 2.9 Structure of an experimental device: 1 tip field emission cathode from the recording side;
2 lattice substrate of the target with signal plate; 3 accumulating and semiconductor layers of the
target; 4 collector of secondary electrons; 5 scanning electron projector with field emission
cathode; 6 electron beam focusing and diversion device
2.1 Field Emission Microscopes 55
diameter and deflect the beam from the target at the velocity of 2000 m/s. The
reading is done by charge exchange with an unmodulated focused beam of electrons
of 1 keV. The tube resolution is 1200 lines per target diameter. The signal-noise
ratio is 10 for the read-out beam current of 0.4 µA. Such device can record signals
under various modes of operation of the field emission cathode: a single pulse,
infrequent and periodic pulses, a continuous current-draw mode. The signals
recorded can be stored and read-out within a period of at least a few tens of seconds,
the erasing time being a few seconds. Independence of operation of the field
emission cathode and the reading gun permits recording in parallel with reading.
Measuring is done at the installation diagrammed in Fig. 2.10. Reading is done
by scanning of the target with an electron beam swept as a standard television raster
in step with the TV display raster. The signal of reading from the target is amplified
by a preamplifier and supplied to the TV display. The screen of the display shows a
field emission image of the carbon fiber cathode in the course of its forming and
operation. The amplitude of the signal of reading from the selected target region is
measured by a horizontal separator (HS) displaying the information on the screen of
an oscillograph.
Illustratively, Fig. 2.11 shows the oscillograms made in the course of forming of
a field emission cathode from a polyacrylonitrile carbon fiber. The forming usually
takes 30 min. The oscillograms are made with a horizontal separator (HS) along
three parallel lines, perpendicular to the target diameter and dividing it into four
equal parts. The oscillograms show spatial distribution of emission current of a field
emission cathode in selected target regions in arbitrary units. Figure 2.11b oscil-
logram is angular distribution of the field emission cathode current of this direction
after recalculation of linear coordinates into angular ones.
In the central part of the target, measurements are made with angular resolution
2 mrad, where L 50 mm is the distance from the end face of the field emission
cathode to the target, dx = 0.1 mm is the minimum distance along the target set by
the diameter of the focused electron beam of the reading gun.
The oscillograms show that heterogeneity of spatial distribution of the field
emission cathode peaks at the boundaries of emission spots at the beginning of
forming of the field emission cathode. After forming has been over, the hetero-
geneities decrease to the acceptable level in connection with a considerable inter-
ception of electron fluxes from a great number of emission centers.
In this way, the installation described makes it possible to measure the charac-
teristics of spatial and angular distributions with angular resolution up to 2 mrad of
the emission current of the field emitters.
A field-electron total energy analyzer is among the promising field emission devices
used to research the internal structure of a solid body and the phenomena taking
place on the emitter surface [17].
All analyzers share the field emission microscope geometry, which means that
the emission pattern given by the electrons having tunneled from the apex is pro-
jected to the fluorescent screen. The screen has a small hole, the so-called trial hole,
transmitting the electrons subjected to further analysis [18]. To select the surface
region to be analyzed, the emission pattern is positioned so that the image of this
region aligns with the trial hole [19, 20].
There are several ways to position an emission image in respect of the trial hole.
They can be divided into three categories: (a) electromagnetic deflection, (b) elec-
trostatic deflection, (c) mechanical travel.
Irrespective of the deflection type applied, the apex image should be maintained
on the optical axis of the system.
The magnetic deflection system [21] has two pronounced disadvantages when
used in the energy analyzer: difficulty to screen the analyzer from the magnetic field
and a bias of the pattern caused by extraction-voltage changes when volt-ampere
characteristics are read.
The electrostatic deflection system [22, 23] solves both of these problems, but
introduces those of a technical nature. It is necessary to mount small insulated
deflecting plates in the immediate proximity of the sample and apply control electric
signals through the vacuum chamber. In addition, such systems distort the field
emission image of the apex surface.
The third type, the mechanical movement, is rather difficult and cumbersome,
but preferable under certain conditions because it does not distort the field-electron
energy spectrum. The only condition is that the traveling apex tip should not widely
deviate from the analyzer axis.
In measuring energy distribution of field electrons, it is assumed that tunneling
does not change the electron energy. Essentially, there are some regions of accel-
eration and deceleration of electrons in the field-electron analyzer (Fig. 2.12a).
Figure 2.12b shows a potential diagram appropriate to the occasion when electron
emission comes from the Fermi level and the collector is at the vacuum zero level.
If the collector is maintained at the zero potential, the field emitter potential
should be negative. Field electrons reach the collector, if their energy
E [ ucoll þ EF Ucoll ;
where ucoll denotes the collector work function, EF is Fermi energy, Ucoll is the
delay potential.
In this way, research of the emitter-material electron energy distribution by the
moment of emergence of collector current can be supplemented with determination
of the collector work function value.
58 2 Experimental Equipment and Technique
The first design of the field-electron energy analyzer with a retarding potential is
described in [24, 25] (Fig. 2.13). It consists of a flask containing a collector, i.e. a
thin conducting film laid on glass, and a shield electrode. Near the emitter, there is a
2.2 Field-Electron Total Energy Analyzers 59
screened anode with an hole of 1.5 mm and the focusing cup. The foot is con-
nected with a flexible glass bellow for convenience of positioning of the necessary
emitter face to the hole. For visual observation of the image, the anode is coated
with a phosphor layer (at 106 magnification, resolution is about 10 Å).
The anode design assumes the field in the hole to be zero, i.e. there is no lens
effect. Appropriately focused electron beam converges in the sphere center and then
moves perpendicularly to the spherical collector surface. To improve stability and
eliminate the background, anode voltage is supplied to the microscope from the
battery packages. The retarding-potential source permits changing the delay
potential through a millivolt of up to 1 V. The analyzer design also includes
electrostatic and magnetic screens. Use of thin apices (down to 300 Å) to decrease
anode voltage permits improving the analyzer resolution by diminishing the
influence of geometrical defects and heterogeneity of the collector on the work
function. The analyzer resolution is *30 mV. Figure 2.14 depicts a delay curve
produced by this device.
Collector current is
Z1
Icoll ¼ PðEÞdE;
ucoll u þ Ucoll
i.e. differentiation of the delay curve gives us total-energy distribution proper. The
fall in the delay curve at high voltages of the collector in the structure [24] is caused
by reflexion of electrons from the collector. That is the central disadvantage of the
early design, aggravated by poor sensitivity and long-time constant of the device.
This design serves as the basis for energy analyzers intended to research field
emission properties of semiconductor films [26–32].
Efforts to raise the analyzer resolution resulted in replacement of the spherical
collector with a hemispherical one [21, 33, 34].
Fig. 2.15 Construction of an analyzer with retarding potential and hemispherical collector:
1 measuring wires; 2 arch; 3 tip; 4 anode cylinder with a probing hole; 5 focusing lens; 6 Faraday
cylinder cover; 7 fixation «beads»; 8 hemispherical collector; 9 Faraday cylinder; 10 Faraday
cylinder base; 11 mounting leg; 12 electrode lead-outs; 13 collector lead-out; 14 glass bulb;
15 conductive covering; 16 conductive covering lead-out
foundation (9). The foundation hosts grid (7) manufactured by electroplating and
stretched on a ring (400 lines/cm). The grid is attached to the foundation with three
insulators (8). The grid pitch, negative allowance and equipotentiality of its surface
62 2 Experimental Equipment and Technique
have the most powerful effect on the analyzer resolution. The entire structure is
attached to the chamber flange with three screws (10).
Behind the grid, there is microchannel plate (6), which is used as a collector.
([20] replaces the microchannel plate with an electronic multiplier.) Sample (11) is
welded to arc (12) secured in holders (13) on insulator plate (14). To align the
sample with the analyzer probe hole, plate (14) is secured on bar (15) rigidly bound
with flange (17), which can move with respect to the chamber case by means of
bellow displacement node (16) (Fig. 2.17 shows only the bellow).
The lens potentials are calculated on the computer to optimize the motion tra-
jectories in the retarding and focusing system of the analyzer. The values of lens
potentials and electron trajectories in the analyzer are given in Fig. 2.18. The design
values u1/u0 = 0.11, u2 = 15/17u1, u3 = 13/17u1 etc. The potentials are set with a
divider, which permits changing them by 10% design value to compensate for
lens size deviations from those of the design. The real working value of u1/u0 for
this analyzer is 0.117.
The electrical part of the analyzer is represented in Fig. 2.19. It is assembled on
the basis of a personal computer and contains high-voltage source PwrSrc [38]
controlled by digital-to-analogue converter DAP, voltmeter B7-16 measuring this
voltage 1/1000 divider, controlled delay-potential source DAP X, voltmeter B7-21,
measuring analogue-to-digital voltage converter ADP, sample warm-up unit
SpecimenHeater, electrometer B7-30 measuring current in the analyzer collector
circuit and an x/y-plotter. Automatic functioning of the analyzer is provided by
purposefully-developed programs making it possible to take, digitize and process
delay curves in mathematical terms. The measuring system permits carrying out
research of the energy spectrum in the range from 0 to −12 V with the accuracy
limited only by noise in electric circuits.
To take the delay curve, delay voltage is supplied from digit-to-analogue con-
verter DAP X through range extender BandExtender (scanning ranges from 0 to
5 V) and through smooth bias control unit OffsetFineTuner, which permits smooth
biasing of the initial point of scanning within the range from 0 to −7.5 V.
2.2 Field-Electron Total Energy Analyzers 63
Fig. 2.17 Construction of a cylindrical analyzer. 1 Luminophore, 2 anode with probing hole and a
protective cylinder, 3 cylindrical lenses, 4 tightening rods, 5 screws, 6 microcanal plates, 7 lattice,
8 insulators, 9 base, 10 fixing screws, 11 the sample, 12 arch, 13 holder, 14 ceramic plate, 15 rod,
16 bellows assembly, 17 flange, 18 ceramic insertions
Fig. 2.19 Schematic of the measuring device. A is the analyzer, Ctrllr is the controller, Iface is the
interface, BtryComm-r is the battery commutator
Fig. 2.20 Electrostatic dispersion analyzer of field electron energies. 1 Anode with a probing hole
and a layer of luminophore; 2–7 cylindrical lenses; 8 insulating rings of quartz; 9 electron
multiplier
microscope. It also permits observing the emission pattern of the field emission
cathode to be researched, controlling of its surface state and determining of the
crystallographic direction. The analyzer anode is supplied with high voltage. The
electron analyzer consists of six electrostatic lenses (2–7) formed by cylindrical
elements electrically isolated from each other by quartz rings (8). The anode and the
lenses are made of molybdenum.
The electrons of the region probed come to the energy analyzer through the input
diaphragm, slow down to initial energies on the first lenses and disperse by energies
by means of chromatic aberration on the last lenses [42]. Through the output
diaphragm, electrons get to the input of the channel secondary-electron multiplier
SEM-6 (9) working in the electron-count mode, which improves analyzer sensi-
tivity. From the output of the secondary-electron multiplier, the electric impulses
come to the input of high-frequency emitter follower EF and move on to broadband
pulse amplifier PA-29. High-speed comparator-inverter HSC converts the amplified
impulses into square impulses of the necessary amplitude coming to the
transistor-transistor logic input of CAMAC-crate counter M401 connected to the
computer.
The energy analyzer element potentials are set by the resistance divider
according to the computations [42]. If R is the total bleeder resistance, the bleeder
resistances fit the following ratios:
divergence of the electron beam in respect of the plane insignificant. The voltage
difference is usually 0.145 V.
The luminescent screen to observe the image in field emission electrons is 6 mm
in diameter. There is a hole of 1 mm in diameter in the center. The distance between
the electron emitter apex and the screen is about 5 cm. The hole in the screen is
covered with a thin wire grid supplied with the same voltage as the screen. As a
result, there is no field in this region and the hole is not functioning as an electron
lens.
For the entire system to fit a small ultra-high-vacuum chamber, it is made quite
compact. All parts of the analyzer are made of electrolytic copper. Insulation and
accurate positioning of the parts was achieved due to rods and rings of aluminum
oxide. The system is mounted on a flange of 20.3 cm in diameter. The system is
protected from magnetic fields and scattered electrons and ions with a screen of
permalloy.
High positive voltage necessary to draw the electrons is applied to the grid, the
first element of the lens and the deflecting plates. In this case, field intensity in the
area of the grid on the screen hole is small, so the field does not practically deflect
the electrons in the grid area. Having entered the lens system through the screen
hole, the electrons are retarded in three stages down to the energy necessary to be
laced up by the analyzer and focused into an almost parallel beam on the entrance
slit. Both analyzers lace the electrons of identical energy. The electrons laced up
through the analyzers are accelerated in two stages and received by the electron
multiplier.
A detailed electric circuit of measuring is given in [44]. The resolution deter-
mined by the Young-Kuyatt criterion [40] is 15 meV at the temperature of 150 K.
There are also modifications of such field-electron total energy analyzers, e.g.
[45, 46].
Under current conditions, most interesting results can come from improved per-
formance of research in the first place and application of universal devices com-
bining several techniques supplementing each other in the second.
The first condition is met by application of field emission microscopes, which
makes it possible to research several samples, e.g. a cassette intended for several
samples, as well as linking devices able to replace the samples without deterioration
in the working vacuum in the primary chamber of the microscope. The linking
devices are widespread in transmission and scanning electron microscopes.
However, they are designed for relatively low vacuum and have no rigid fixation of
samples necessary to provide electric contacts. Therefore, such linking devices
rated at the pressure of at least 10−7 mm Hg are designed on a case-to-case basis or
the available devices are modernized, e.g. [47, 48].
68 2 Experimental Equipment and Technique
Fig. 2.22 Device schematics: 1 vacuum chamber; 2 luminescent screen; 3 upper flange with drive
units for samples; 4 VN-2MG forevacuum pump; 5, 8 nitrogen traps; 6 VMN-150 turbomolecular
pump; 7 Trion-150 magnetic charge pump; 9 additional volume
In the deplenishing system, the total pressure is measured with vacuum gauge
VTB-1, the low and medium vacuum with vacuum gauge VMB-8 and high and
ultra-high vacuum with vacuum gauge VI-14.
The imaging gas necessary to form a field-ion image or process the cathodes in a
special way is puffed to the chamber through leak valve B2, and the chamber gas
components are measured by partial-pressure gage IPDO-2.
High voltage is supplied from the high-voltage (0–40 kV) unit to the lumines-
cent screen of the microscope and measured by a digital voltmeter through a
calibrated voltage divider. The work function change is measured with a stabilized
70 2 Experimental Equipment and Technique
power source [54]. The potential of the samples is close to the earth potential, which
permits their warming up from the filament voltage unit of a conventional design in
case of need and essentially simplifies sample control kinematics. The emission
current is measured by electrometric voltmeter VK2-16 or VK2-20, depending on
the range of the current to be measured.
The volt-ampere characteristics are recorded with an x/y-plotter or a computer
with built-in ADC plate connected to outputs VK2-16 and a high-voltage unit
divider. Some particular experiments involved different measuring devices as well.
The luminescent screen design is shown in Fig. 2.23. This design [55] is special
in its ability to supply high voltage to the luminescent screen (2) itself by parti-
tioning of functions of the screen (2) and the observation glass (8), which essentially
simplifies the design and improves reliability of the cathode assembly of the
microscope.
The screen (2) is a flat glass plate of 100 mm in diameter having a conductive
coating and a phosphor layer.
The screen (2) is fixed between two rings (3, 4) and separated from the case with
three ceramic insulators (5) attached to the flange (1) through screw pairs (6) able to
relocate the screen, thereby changing the anode–cathode distance, i.e. magnification
of the microscope. High voltage is supplied to the screen (2) through a ceramic
high-voltage current lead (9) protect with the screen (10).
The luminescent screen designed like that can be promptly swapped, and par-
titioning of functions of the observation window and the screen makes it possible to
fabricate screens of larger dimensions irrespective of the observation-glass size.
Microscope kinematics. The universal field emission microscope kinematics
should meet the following conditions:
Fig. 2.24 Kinematic schematics of the microscope: 1 upper flange of the chamber; 2, 3, 6 pulleys;
4 cross-beam; 5 brace; 7 sample; 8, 9 gears; 10 screw drive; 11 cassette with samples; 12 Dewar
flask; 13 moving and rotation input; 14 electric vacuum inputs; 15 rotation inputs
72 2 Experimental Equipment and Technique
The design includes a cassette (11) simultaneously holding six samples (7) with
independent electrically insulated leads and general cooling of samples through a
flexible heat conductor from a Dewar vessel (12). Every sample can be placed into
the working position opposite the luminescent screen and turned as stated above.
The sample positioned is rotated about the vertical axis of the chamber with one
of the bellow rotary-motion feedthrough (15) placed on flange DN-50.
A cable transmits rotation from the feedthrough (15) via the pulley (2) to the
pulley (3), which is rigidly connected with the balance beam (4). Gyration of the
balance beam (4) provides rotation of the positioned sample about the vertical axis
within 0–360°. The angle of rotation is only limited by the danger of twisting the
heat conductor and the interconnecting electric wires.
The sample positioned is rotated about the horizontal axis of the chamber by the
rotary-motion and repositioning feedthrough (13).
To perform the rotation, the feedthrough (13) is engaged with the joint box,
which is rigidly connected with the pulley (6). A cable transmits rotation from the
pulley (6) to the pulley set on the worm spindle. The worm wheel (10) is set on the
cramp axle (5), where the sample cassette (11) can run in bearings. The cramp and
the cassette are designed so that the end of the sample (7) lies exactly on the
interception of the axes and the rotation does not change its position in space. The
worm gear can solve the problem of fixing the sample after the rotation has been
over.
Application of a rotary-motion and re-positioning feedthrough is caused by a
necessity of kinematic unlocking of axial rotations because if the interlocking were
rigid, rotation about the vertical axis would transmit a torque to the cramp (5),
which would cause rotation about the chamber axis as well. The samples are
changed by rotation of the cassette (11) with gear wheels (8, 9) and one of the
rotary-motion feedthroughs (15).
The samples are electrically connected through ceramic current leads (14) by
flexible wires (missing from the illustration) tubed in ceramics to prevent short
circuit.
The samples are cooled by a flexible heat conductor from the double Dewar
vessel (12), cooling the sample down to helium temperatures. In this case, the
internal tank is filled up with helium, and the external tank, separated from the
internal tank by a vacuum gap, is filled up with fluid nitrogen.
It is expedient for such large and complex high-vacuum systems as a field
emission-and-field-ion microscope to use cassettes able to load the vacuum
chamber with several samples at once, which essentially reduces the research time.
Operation in ultra-high vacuum necessitated development of a rotary sample cas-
sette [60] (Fig. 2.25). It can hold 6 samples with independent electrical connection
at a time, heat the sample up to the temperature of fusion and cool down to helium
temperature. Every sample can be placed into the working position opposite the
luminescent screen and turned in respect of the two mutually perpendicular axes in
such a way that the sample vertex is always in the interception of the axes, which is
necessary when working with probe holes. It seems optimal to have six samples in
the cassette. A larger number would not only essentially complicate the sample
2.3 Field Emission Microscopes of Universal Design 73
design, but also have mutual effect on emission characteristics because of a shorter
distance between them.
The samples (1) secured on bails (2) are screwed (3) in the holes (16) of copper
segments (5) separated from each other by vacuum gaps. The segments (5) are
pressed to the heat conductor (19) through a thin sapphirite plate (15) by screws (6).
The screws (6) are separated from the segments with a ceramic plate (8). The heat
conductor (19) is pressed to the foundation (14) by screwing (6). The foundation
can freely rotate in the ball-bearing (12) secured in the holder (13) and, therefore, in
the cramp (4). To reduce heat losses, the heat conductor and the foundation contact
only in the screwing areas (6).
The sample (1) is fixed in the working position by a ball (20) pressed to the
foundation (14) with a spring (22) through a steel insert (21).
The foundation (14) has six grooves located along the same radii with the
samples. The groove size is about 2/3 diameter of the ball, which does not let the
74 2 Experimental Equipment and Technique
ball slip into the groove, on the one hand, and secures the foundation on the slit, on
the other hand.
To change the samples, the gear wheel (11) rigidly secured on the foundation
axis (14) is engaged with the gear wheel (10) turned by the bellow rotary-motion
feedthrough (9). When the gear wheel (10) rotates, the gear wheel (11) turns
together with the foundation (14), the ball (20) leaves the fixing groove and rolls
along the lateral surface of the foundation to the next groove. The number of
gear-wheel teeth (10) provides rotation of the gear wheel (11) through 60°, i.e. the
value of the shift to the next groove and, therefore, the next sample. This system can
change samples in any direction from the first to the last sample and in the reverse
order.
The samples are electrically connected with external circuits by flexible wires
(18) tubed in ceramics (17). The samples are cooled from the Dewar vessel (po-
sition (12), Fig. 2.24) through a flexible heat conductor (19) pressed into a rela-
tively massive copper heat conductor.
The gear wheels (10) and (11) are only engaged at the moment of change of the
sample. That is why the sample placed into the working position can freely rotate
about the two mutually perpendicular axes. The angles of rotation are only limited
by the wires (18) and the heat conductor (19).
Certainly, that is not a full list of field emission microscopes. This section shows
only the general basic principles of design of such microscopes.
Fig. 2.26 Schematic of the field cathode emission characteristic measurement system
76 2 Experimental Equipment and Technique
Practically, most well-proven and handy workpieces for metal-apex emitters are
thin wires of 20–200 lm in diameter, though nearly all workpiece shapes will do.
For many metals, they are produced on the industrial scale or fabricated under
experimental conditions, which is to be featured below.
In case of massive materials, a workpiece with a square or round cross-section is
fabricated. Usually it is less than 0.5 mm in diameter and about 1 cm in length. As a
rule, the workpieces are fabricated by ordinary mechanical methods, such as
low-speed diamond sawing, electric spark machining, centerless grinding, wire
drawing, warm forming or extrusion [67].
Selection of the method depends on plasticity and ductility of the material with
account of the effect of the processing technology on the material microstructure.
E.g. wire drawing, warm forming or extrusion are very effective methods to obtain a
2.5 Emitter Technology 77
Fig. 2.27 Review diagram of the main classes of field electron emitters
round cross-section, but they extremely increase dislocation density and change the
grain shape. Polishing removes external layers of the material, thereby eliminating
some injuries caused by the previous machining, but leaving the defects caused by
the operations of a wire-drawing type. If the fabrication method gives a
rectangular-section workpiece, it is important for the sides to be 1:1. Otherwise, the
sample would be shaped as a knife. Right angles usually round in the course of
polishing, resulting in the workpieces with a round cross-section. The samples of a
definite shape or made from particular materials are manufactured with special
methods described in the following sections.
If the materials are not ductile enough for the appropriate operations of drawing
or needed in a metastable state, the wires can be fabricated by rapid-solidification
methods [68].
In fabrication of a liquid-rolled band, the basic drawing process can often give a
by-product of thin metal threads, which can be used as workpieces for field-emitter
samples, despite their rather irregular cross-section [69]. The best method is
drawing of a thread from a melt into a rotating liquid. In this process, the melted
78 2 Experimental Equipment and Technique
material of the sample is extruded through a hole into a cooling liquid, forming a
continuous wire [70, 71] (Fig. 2.28).
The drum is rotated at a velocity of about 250 rev/min, and the cooling liquid is
introduced into the internal part of the drum, where the material is shaped as a film.
The metal is melted in a high-frequency induction furnace and extruded by pressure
of argon through a quartz or ruby nozzle of 50–300 lm in diameter into the rotating
layer of the coolant. Continuous threads develop when the cooling-surface tan-
gential velocity is 1.15–1.3 times higher than the metal-jet feeding rate. The tan-
gential velocity, the layer thickness and the cooling-liquid rate, the extrusion rate,
the temperature and the metal-jet incidence angle, the nozzle size and the distance
between the nozzle and the cooling liquid, as well as the drum rotation rate are
extremely important experimental parameters [72, 73]. If they are properly selected,
a large amount of a small-diameter wire can be fabricated within one work cycle.
This method also creates fine or amorphous materials due to rapid solidification at
high cooling velocities (103–105 K/s). That is the process to fabricate threads from
the alloys of Fe–C–Si, Pd–Cu–Si, Ni3Al, Cu and Al.
A continuous wire can also be fabricated directly from the liquid phase [74]. The
sample material melts and gradually runs out to the rotating cold cylinder
(Fig. 2.29). Unlike the band fabrication process, the flat cross-section of the cold
cylinder slightly changes in that it has a small prominence where the metal lands.
The wire section size is usually 40–80 lm. The prominence makes it non-circular.
The method advantage is in production of a great deal of very fine material.
2.5.2 Machining
Many materials, in particular single crystals, molding materials, such as e.g. gra-
phites or high-temperature ceramics, are output as massive workpieces. Therefore, a
high-quality apex comes from a thin workpiece. At the first stage, the machining is
usually mechanical, e.g. by skive, and results in studs of 0.3 0.3 10 mm. The
dimensions are approximate and depend on the nature of the material and the
research objectives. Electric-spark machining is not desirable for blanking because
it can occasionally result in doping of the stud surface with reaction products.
Paper [75] describes the fabrication technology of field tip emitters from Ge and
Si. At the first stage, rectangular workpieces are cut out of Ge single-crystal ingots
oriented in the direction of (100) with the accuracy of ±1° (Fig. 2.30a). Resistance
heating of the emitter is provided by ultrasonic cutting approximately to the middle
of the sample on one of its ends so that the sample is fork-shaped (Fig. 2.30b). After
that, the workpiece is luted to the holder with quartz putty and placed in the chuck
of the jeweler’s lathe. The uncut end of the workpiece is ground with a diamond
wheel to form a cylinder of 0.38 mm in diameter (Fig. 2.30c). The bridge is made
slightly thinner on the fork than on the ends, which enables to heat up the emitter
almost to the melting point, whereas at the clamps, the sample remains cold. That
makes it possible to minimize temperature migration of impurities from the holder
to the apex of the emitter. The processed workpieces are cleaned in methanol under
gentle ultrasonic mixing. After that, the surface pollution is removed by the etching
agent, consisting of 25 parts of smoking HNO3, 15 parts of HF and 15 parts of ice
acetic acid by volume. Emitters are formed by electrolytic etching of cylindrical
workpieces in cold etching agent at the temperature of 10 °C (cf. below). To
preserve the apex shape after the scattering, the cone-apex angle should be about 5°.
To provide effective heat transfer to the apex, the emitter is 1 mm long. Formation
of the apex is usually followed by urgent measures of neutralization and removal of
etching residues. The final stage of emitter fabrication is shown in Fig. 2.30d.
A more sophisticated technology can result in making samples of rather thick sheets.
One of such methods is suggested in [76]. It can fabricate sample workpieces of the
Fig. 2.31 An illustration of forming sample blanks out of a thin band. a after the band is
thoroughly cleaned its surface is covered with a thin film of photoresistant material; b the
photoresistant film hardens forming two masks on upper and lower surface, and the construction is
exposed to UV radiation in order to transfer the image to the photoresist; c the masks are removed
and the photoresist is developed by dissolving the films that were not exposed to UV. The
construction is then dried off to harden the photoresist film; d the construction is electropolished to
remove the parts of surface not covered with photoresist; e the electropolishing continues until the
band is etched through and the sample blanks are obtained (1 photoresist; 2 sample; 3 mask)
2.5 Emitter Technology 81
sheets both as thick as more than 2 mm and as thin as 0.2 mm. Figure 2.31 repre-
sents the operating sequence of sample fabrication with this method.
With this end in view, the sheets of e.g. tungsten or stainless steel are degreased
for 30 min in the concentrated alkali of NaOH, followed by coating with a pho-
toresist. Then they are repeatedly washed with water, dried up and then sunk in the
photoresist solution and re-dried at t = 100 °C for 30 min. This process is repeated
several times to remove possible defects caused by air bubbles or dust present is the
photoresist solution. The unmasked photoresist part is exposed to ultraviolet radi-
ation for 2 min with a high-pressure mercury arc lamp placed 0.14 m away from
the object surface. After unmasking, the sheet is kept in the etching solution to
remove the unexposed photoresist part. To solidify the remaining photoresist, the
sheet is warmed up for 30 min at t = 140 °C. When the photoresist film is abso-
lutely dry, the upper and lower surfaces are covered with photomasks.
The photomask pair is usually made of two films with appropriate printed
drawings. The photomasks should be installed accurately so that the upper and the
lower exposed regions were properly oriented. The lines of the two photomasks can
be located either opposite each other or with a shift. The drawing pitch changes to
meet the band thickness, which results in an approximately square cross-section.
If the photomasks are placed with a shift, the drawing pitch should be approx-
imately equal to two thicknesses of the sample. The photoresist layer is exposed to
UV radiation to fix the print. The photoresist film is developed, solidified and then
removed from the regions that were not exposed to light. In these places, the sheet
is etched through to shape the workpieces. One operation can produce a great
number of the workpieces. This method can give the samples of metal glass [77],
stainless steel and tungsten [78].
The photoresist method is also applied to fabricate the workpieces of emitters
from the surface layer of a thick material (Fig. 2.32) [76]. After the surface is coated
with a photoresist, exposed in ultraviolet through the photomask and developed, the
grooved sides are masked and etching goes on until the grooves on each side of the
photoresist are etched through and a workpiece is formed.
The simplest way to fabricate apex emitters from wire workpieces includes
pointing of the wire in the gas-burner flame [79–81].
This method assumes that the wire material is oxidized at high temperature in the
presence of air oxygen, the oxide being continuously removed. For purely metallic
wires, it is difficult to fabricate high-quality apices in this way. However, this method
is good for the materials containing highly-volatile oxides. E.g. the same method can
shape tips from some materials such as glass carbon [77]. The tip is shaped as a tip of a
vitreous carbon material, e.g. furan, phenol, pyrrole or vinyl resins obtained from
divinylbenzene and carbonization of the solidified and shaped resin in vacuum or an
inert-gas medium. Carbonization takes place under rise in temperature with particular
velocities within particular temperature ranges. Eventually, the apex tip is etched in
the flame of three symmetrical gas burners to obtain the radius of 1000–3000 Å.
Chemical etching was widely used in the earliest days of field emission
(cf. [82, 83]). However, chemical etching is rather difficult to control and the resulting
tips have a wide spread of curvature radii. This method is only up to a highly-qualified
experimenter, which is why it is mostly applied to low-conductance materials.
82 2 Experimental Equipment and Technique
form of the applied voltage can largely change the apex curvature radius and the
shape of the tip, e.g. the angle and the form of the cone.
The fifty-years-plus history of development of electrochemical etching methods
have brought a great number of designs of electrochemical waves, electrolyte
compositions and apex-fabrication techniques, e.g. [89–92].
Briefly, the essence of electrochemical etching is as follows [93, 94].
Direct or alternating voltage is applied to the wire workpiece immersed in the
electrolyte and the auxiliary electrode (Fig. 2.34a). Its value depends on the
workpiece diameter, the dimensions and configuration of the device used for
etching, the etching-agent composition and temperature. Therefore, it is expedient
to select its value on the experimental basis for each particular case and use the
values given in this section for reference only.
When voltage is applied, the wire workpiece conducts current and creates a
potential difference UB > UA along its length. As etching-current density is directly
proportional to voltage and inversely proportional to the curvature radius of the
surface stripped, the etching-current density in point B is higher than in point A. In
the course of etching, the current falls in ratio to decrease in the wire diameter, and
resistance of the region stripped increases in inverse proportion to the diameter
squared (pos. 2, 3 in Fig. 2.34b). The drop in voltage along the region immersed in
the electrolyte increases, and the vicinity of point B experiences more intensive
etching. Quicker reduction of the diameter in this place accelerates the etching rate,
which leads to development of a constriction (pos. 4 in Fig. 2.34b).
Fig. 2.34 Schematic of volumetric etching of the tips (a) and steps (1–7) of electrolytic etching of
the tip (b) along with the time curves of etching current and its derivative: 1 arch; 2 tip blank;
3 tub; 4 electrolyte; 5 additional electrode
2.5 Emitter Technology 85
nd 2 rB
\P; ð2:9Þ
4
where q is the specific gravity of the material, the weight force of the displaced
liquid included, dl is the average diameter of the lower part of the workpiece, Ll is
an equivalent length of the lower part, which is determined from the condition that
the volume of the lower part at the tearing-off moment equals
Ll pdl2
Vl ¼ : ð2:11Þ
4
decreases until the lower part of the workpiece tears off (pos. 5 in Fig. 2.34b), after
which the current value slumps. At that moment, the current-derivative diagram
shows a surge of a rather high amplitude. To fabricate an apex with the minimum
curvature radius, it is necessary to stop the etching process either mechanically
(using the design features of the etching systems) or electrically (when the elec-
tronic circuit fixes a current drop or a peak in the current derivative and disconnects
supply, e.g. [95]) from the moment of tearing-off of the lower part of the workpiece.
Efficiency of such systems and, accordingly, the tip curvature radius are determined
by the response time from the moment of tearing-off of the lower part.
Currently, there are quite a lot of patterns and modifications of tip
electrolytic-etching systems.
Below we give three examples of implementation of the main classes of this
process. It is worth noting that the simplest electrolyte for fabrication of
samples/apices from carbon fibers is the 1N solution of NaOH or KOH at the
alternating voltage of *1 3 V.
One of the most conventional and simplest tips etching methods is the method of
volume etching (Fig. 2.34).
When positive voltage is applied to a metal sample and the sample is immersed
in an electrolyte of the opposite polarity, voltage build-up first causes increase in
current, which meets the course of electric etching processes. With further increase
in voltage, current reaches the plateau and the etching process turns into
electropolishing.
At even higher voltages, the sample surface blisters with gas, which can cause
pitting etching of the sample. The electrolyte usually contains an oxidizer and a
solvent to remove the anode-etching products. Electropolishing often produces a
viscous layer denser than that of the electrolyte. It flows down the wire and insu-
lates its lower end. In the course of this process, the material is most effectively
removed from the region located above the wire tip, thereby gradually reducing the
cone angle in the immediate proximity from the developing apex. Under these
conditions, the rising bubbles can prevent further impact on the upper parts of the
sample, thereby causing an undesirable increase in the cone angle.
Plastic deformation changes the mechanical properties of the surface layer near
which the wire is cut. Therefore, tip fabrication follows removal of the damaged
area by electropolishing.
Simple dipping in an electrolyte produces apices of practically any cone angle
depending on the electrolyte depth the workpiece is immersed at. A small cone
angle results from immersion at the depth of 3–5 mm. A large angle results from
immersion at the depth of 1 mm. Such a method can be successfully used to
fabricate tips from easily etchable materials. However, this method is not good at
sample-shape reproducibility and needs much training, which is especially
important.
The most widespread ways to fabricate apex samples include etching in an
electrolyte droplet [79] diagrammed in Fig. 2.35a. A glass tray (1) is filled with a
current-conducting solution which is usually a saturated NaCl solution. The latter
holds an electrode (6) shaped as a plate of the material non-reactive with the
2.5 Emitter Technology 87
Fig. 2.35 Schematic of electrolytic etching in a drop (a) and wire loop (b): 1 tub; 2 conductive
solution; 3 emitter blank; 4 diaphragm; 5 drop of electrolyte; 6 additional electrode; 7 wire loop
solution, e.g. nickel. A workpiece (3) is put in the center of a small hole in the plate
(4). An electrolyte droplet (5) is pipetted in the hole and retained inside by surface
tension forces. When voltage is supplied, electric current flows through the elec-
trolyte droplet, where it initiates electropolishing.
Another good thing about this method is that etching automatically ceases when
a workpiece is over-etched since its lower part falls off and breaks the electric
circuit.
Formation of tips depends not only on spatial distribution of the concentration
layer consisting of reaction products, but also on the composition, concentration
and temperature of the electrolyte and the sample material [96]. It has also been
demonstrated that gassing that accompanies etching and is to a great extent the
cause of poor reproducibility of the tips shape can be practically eliminated by
adequate selection of the etching mode.
This method prevents the effect of plastic deformation caused by cutting of wire
into workpieces since the wire tip does not belong to the etching area and, con-
sequently, the tip area.
Very thin samples, as well as the apices subject to re-polishing, are etched in an
electrolyte droplet [92] (Fig. 2.35b) filling a loop of 3–5 mm in diameter made of a
thin wire, such as a tantalum or platinum wire of 0.2 mm in diameter.
The loop is filled up with the electrolyte and moved along the needle at the
moment of voltage application usually in a form of a series of manually controlled
impulses lasting for about 0.2–2 s. [97]. This procedure results in desirable shaping
of the needle-apex area by change in duration of polishing or the impulse number.
The desirable accuracy of positioning is obtained by mounting of the sample or the
wire loop on the micromanipulator. The process is usually controlled with a light
88 2 Experimental Equipment and Technique
microscope. The small volume of the electrolyte necessitates its frequent substi-
tution. As a rule, the electrolytes are the same as at the second stage of usual
electropolishing.
Originally, the micropolishing technique was developed to taper filamentary
crystals [98]. However, it is mostly used to taper the samples that cannot be formed
by usual electropolishing or re-taper the samples blunted as a result of exploitation.
The technique can also be used to fabricate samples from unusually-shaped
workpieces. E.g. tips can be fabricated from singular droplets of fast-curing alloys
[99].
Good reproducibility of polishing of sharp tips during their fabrication is pro-
vided by regulation of the following conditions of polishing: (1) electrolyte solu-
tion; (2) electrolyte temperature; (3) distance between the cathode and the tip;
(4) the value of alternating or direct voltage; (5) frequency of alternating voltage;
(6) number of alternating-voltage periods or duration of constant-voltage impulses.
Paper [100] suggests a power source able to work in the following four modes:
(1) electropolishing with alternating current; (2) electropolishing with direct cur-
rent; (3) automatic polishing with alternating or direct current; (4) charge integra-
tion in any of the three modes.
To reduce the tip-radius spread and the curvature radius down to *100 Å, paper
[101] suggests an advanced method of «a droplet in a loop». Figure 2.36 demon-
strates a simple experimental needle-fabrication device. The metal-wire segment of
a typical thickness ranging from one to several hundreds of micrometers is placed
so that it crosses the region of interface between two liquids. Lower liquid (2) is
denser. It does not conduct current. Usually, that is CCl4. Upper liquid (1) is an
electrolyte. One electrode is wire (6) subject to etching. Electrode (4) has a ring on
the end. When the diameter of the etched wire (6) area reaches a critical value, a
part of wire (5) held in a non-conducting liquid gets detached under action of
gravity. Due to a clear-cut interface between liquids (1) and (2), the detached part of
wire (5) that gets into the SF trap has an extremely sharp tip with the curvature
radius down to *10 nm.
A design similar in principle but without a loop is presented in Fig. 2.37 [102].
Fig. 2.37 Etching device. 1 blank; 2 electrolyte layer; 3 hole in the bottom of the tub; 4 receiving
glass; 5 prepared electrode; 6 pipe for supply of the electrolyte for etching; 7 hole for draining the
electrolyte out
Wire workpiece (1) is placed vertically to pass through electrolyte layer (2) and
hole (3) in the bottom of the bath. The hole has a diameter sufficient to hold the
electrolyte due to surface tension. The largest diameter of the hole can be evaluated
by the formula
4c
D¼ ;
nd
the electrolyte and the fluid contact causes etching of the workpiece in the area of
Formation of tip (3). At the same time, voltage is additionally applied between the
fluid contact and the sample holder. Etching of area (3) causes development of a
constriction. The density of the current passing through it increases with decrease in
its diameter and reaches the value at which the electroplasticity effect manifests
itself, usually about several thousand amperes per square mm. Due to dramatic
increase in plasticity of the tip-formation area, falling off of the lower part of the
workpiece causes drawing of the tip with a small-curvature radius. The latter can be
regulated both by the value of voltage applied between the workpiece holder and
the fluid contact, and the length/weight of the part of the wire workpiece immersed
in the electrolyte.
The disadvantages of those methods include overly large droplet or overly thick
electrolyte layer, which results in increase in length of the etching part of the wire
and rather rapid exhaustion of the electrolyte. Localization of the etching area can
be improved by application of the device [105] shown in Fig. 2.39. The electrolyte
is poured between two plates (1) with coaxial holes. The lower end of the work-
piece of field emission cathode (3) inserted from above into the holes of plates (1)
touches current-conducting liquid (4), where the second electrode (6) is held.
Voltage for electrolytic etching comes to electrodes (1) and (6) from source (7).
The cone angle and the length of the cone part of the field emission cathode is
regulated by change of thickness of the current-conducting film of the electrolyte.
Concave area (2) of a plate serves for increase in the total quantity of the electrolyte
between the electrodes, which makes it possible to avoid exhaustion of the elec-
trolyte in the course of etching without increase in thickness of the etching film.
The design of the double bath for automatic etching of tips is shown in Fig. 2.40
[106]. This design is a considerable improvement of two-layer and three-layer baths
[99]. Polarity of voltage between electrodes (1) and (2) shown in Fig. 2.40 induces
positive potential in workpiece (3) in respect of electrode (1). The reason is that
workpiece (3) causes short circuit of the electrolyte area in hole (6) as the workpiece
has a smaller electric resistance in comparison with the electrolyte in a small-section
channel. A certain current between electrodes (1) and (2) makes voltage on
2.5 Emitter Technology 91
workpiece (3) sufficient for over-etching of the wire material in the vicinity of point
A. At that, the lower area of wire AB falls off, causing the upper part of the
workpiece to lose positive potential, and workpiece etching ceases, which provides
safety of the upper tip. Screen (4) and electrolyte pumping are necessary to destruct
the bubbles that intensively stand out in the vicinity of point B and electrodes (1)
and (2). The apex-cone height is mostly determined by the electrolyte-mark height
in vessel (7), electrolyte flow-rate in hole (6) and current between electrodes (1) and
(2). The apex shape is also determined by the dimensions of hole (6). E.g. if the
sectional area of this hole is about 3 mm2, the tip-cone length reaches 2–3 mm.
Some typical materials and conditions of their electrochemical or chemical
polishing are presented in Table 2.1 [67, 107].
92 2 Experimental Equipment and Technique
Sometimes two stages of polishing can be combined into one operation. The first
stage, when an electrolyte layer usually as thin as 6–8 mm floats on the surface of a
denser inert liquid e.g. CCl4 or glycerine, is intended for prompt removal of the
material and development of a constriction in the workpiece center. The inert liquid
should not mix or react with the electrolyte.
At the second stage, polishing proceeds until the weight of the lower part of the
workpiece exceeds the tenacity of the thin constriction. The starting material can be
shaped as a wire of a square or round section. The opposite electrode is usually
made of a precious metal e.g. gold or platinum.
2.5 Emitter Technology 95
This method is especially good for thick wires or the workpieces cut out of
massive samples e.g. aeroturbine blades or semiconductor single crystals.
Fabrication of apices from thick wires and workpieces can also involve a device
for etching of apex samples in an electrolyte flow (Fig. 2.41 [108]). Wire workpiece
(1) fixed in collet lock (3) is over-etched inside ring electrode (2). The electrolyte
flow-rate appreciably affects the apex-surface quality, the shape and the
apex-curvature radius. E.g. tungsten is electropolished in an alkaline solution in the
alternating-current mode at the optimum electrolyte flow-rate of 0.2 m/s. The
electrolyte is pumped over a closed circuit.
The current used in electropolishing should be controllable to prevent heating or
even melting of the thin area of the sample constriction, which is quite possible if
the constriction is extended enough. When alternating current is used, control of
such parameters as voltage, wave form, phase angle, frequency and cycle number is
useful for fabrication of high-quality samples [109, 110]. It is also possible to
control current and interrupt polishing automatically as soon as the lower half [78]
gets detached.
The research carried out to determine the ultimate strength of very thin tungsten
wires established dependence between the final cross-section measured with a
transmission electron microscope and the wire-end mass that fell off as a result of
electropolishing. Thus, the method cannot be used to fabricate an unimpaired
sample. It deforms the sample and the field-ion image shows all signs of plastic
deformations and cold working. That takes place even when the end of the wire
immersed in a non-conductive liquid is reduced in diameter and made very short as
a result of preliminary etching to reduce the weight of the lower area of the wire.
The lifting force of gas bubbles flowing to the lower end of the wire just before the
separation can deform the wire by bending the etched part of the sample.
To obtain an undeformed apex with a small cone angle, it is necessary to reduce
etching in the vicinity of the tip apex by elimination of the steep current gradient
through its gradual reduction due to partial screening of the sample apex. After
preliminary electric etching, the sample apex is brought up to the necessary
dimensions by immersion of the wire tip in a protective tube of 1–2 mm in diameter
(Fig. 2.42). For normal circulation and uniform distribution of the dissolved sub-
stance, the tube should be open from both ends and made of a coloured glass for
optimum visual observation. Any necessary cone angle can be obtained without
separation of the lower part of the wire by immersion of the sample at different
depths relative to the upper edge of the tube.
The tip polish quality is usually examined with an optical microscope. For this
purpose, a 500-fold magnification with the aperture of 0.6 is quite enough. The
tip-curvature radius ranges from 100 to 500 Å, which is much higher than the
maximum resolution of an optical microscope. However, achievement of a uniform
change of the cone angle in the immediate proximity of the tip apex is a quite
sufficient criterion of its suitability as emitter samples.
In some cases, it is not expedient to use chemical etchants, especially for more
porous, e.g. carbon-based, materials.
Fig. 2.43 Schematic of tip manufacturing in glow discharge (a): 1 blank; 2 conductive plate;
3 insulating screens. Schematic of sharpening of the graphite fibers (b): 1 field cathode blank
(graphite fiber); 2 holder; 3 rod; 4 micrometric screw; 5 anode; 6 microscope; 7 observer
98 2 Experimental Equipment and Technique
Variation of voltage or the distance with the micrometer screw (4) can change the
apex shape. For the fiber of 7 lm in diameter, the discharge voltage is about 5 kV
at the electrode-separation distance of 5 mm. The tapering can take place with
exposure to air or in technical vacuum, and voltage can be pulsing if necessary.
Accurate processing, i.e. tapering of new and sharpening of old apices, involves
the systems of focused ion beams of 1–2 mm in diameter [112, 113]. The angle
between the sample axis and the ion beam is usually about 90°.
If an tip workpiece is mounted in the holder of a transmission electron micro-
scope, the end of the apex-fabrication process can be fixed with high accuracy. At
that, fabrication of a uniform cone results from rotation of a sample workpiece
about the axis in the ion flux until the tip radius is reduced to a suitable value.
It has been established that ionic bombardment lasts approximately as long as
usual electropolishing. (Though the name of the paragraph places a priority on ionic
bombardment, any particles e.g. electrons can be used [114].)
The time necessary for ion etching can be reduced by preliminary chemical
etching of the sample cf. [115].
At first a workpiece with the curvature radius of 260 Å fabricated by the elec-
trolytic method is placed in a vacuum chamber. After pumping-out down to the
pressure of 10−5 Pa, the chamber is filled with neon up to the pressure of 10−3 Pa.
Voltage of 800 V is applied between the cathode and the anode. At that, the
electron-current density is 106 A/cm2. At this potential difference, the cathode is
kept for 5 s. Over this period, the field emission current increases due to ion etching
and then stabilizes at the density of 107 A/cm2. After that, the processing voltage is
10% reduced and the cycle is repeated until the running voltage reaches 300 V. At
that, the apex-curvature radius is 100 Å. Reproducibility of emitter parameters is
very high, only about 2% are rejected.
Fabrication of apices from epitaxial layers in the semiconductors of type III–V is
suggested in two techniques [67, 116].
The samples containing epitaxial layers can be made by preliminary cutting of
square-section workpieces from a multicomponent assembly consisting of a sub-
strate material plus a layer. These workpieces are mechanically polished from two
or three sides up to formation of a cone with a apex in the layer. Then the remaining
substrate is chemically removed by selective etching. The resulting conical work-
piece is chemically polished up to the tip shape. In this case, the sample axis is
parallel to the initial surface.
The samples whose needle axis is perpendicular to the surface can be prepared
by surfacing of a ball of about 1 lm in diameter made from polystyrene or quartz in
the ethanol suspension containing a droplet of a surface-active substance followed
by ionic bombardment as shown in Fig. 2.44. A sample with a suitable cone angle
is fabricated with a ball that is more resistant to bombardment than the surface
layer. The material lying under the ball is protected from the bombardment. In the
course of bombardment, the ball diameter is decreased, which gradually opens the
entire layer for ionic treatment. As a result, the sample is tapered.
Most metals can develop thin microelevations on the surface of blunt tips due to
the surface diffusion of tip material atoms. That is usually done by the electric field
2.5 Emitter Technology 99
Fig. 2.44 Manufacture of field ion microscopy sample out of epitaxial layer
of the field-ion sign with voltage of about 2 V/Å combined with the elevated
temperature of >1000 K. Such microelevations have a size of 40–80 Å [117].
A lot of research projects need formation of a surface layer on the apex.
The most common method is thermal evaporation [118–126], though there is
also activated thermal evaporation [127, 128], chemical reactions [129, 130] and
radio-frequency sputtering [131–134].
There are also some other methods with immersion into a liquid, including
deposition [135], anode treatment or passivation [136, 137] and immersion into a
molten metal [138]. Thickness of the films layered should be controlled because the
sample-curvature radius increases over the homogeneous layer, which brings the
research to increase in running voltage.
Moreover, application of an electric field can strip thick films.
Another type of the surface layer can result from application of biological
molecules [139–142]. These experiments need special techniques to provide opti-
mum application of molecules on an tip.
Tip quality control in a light microscope after preparation and cleaning is a
routine procedure. Though the imaged part of the sample is not resolved at small
magnifications 200–500, a light microscope is helpful in examination of the end
radius, surface processing and selective etching.
A smooth conical rod and presence of diffraction fringes on the needle tip
indicate satisfactory quality of the sample. It is not necessary to use large magni-
fications needing thin immersion oils because a slight increase in magnification is
not enough to obtain more information, whereas the liquid surface tension can
destroy the thin needle. The details of surface processing and selective polishing are
better determined at larger magnification in a scanning electron microscope.
A small apex-tip curvature radius can be examined in a transmission electron
microscope.
100 2 Experimental Equipment and Technique
Below you can find some practical examples of application of field emission
microscopy.
First of all, it is field emission microscopy of different materials and composites,
which has been developing since 1936. Its fundamentals were given in Sect. 2.1.
The maximum brightness of a field emission image is produced by the regions
with a lower value of the electron work function. So crystallographic directions of
single crystals in emission patterns can be decoded on the basis of stereographic
projections [79, 150]. Application of field emission microscopy is a great advance
in comparison with widespread macroscopic diffusion-research methods due to
2.6 Some Technological Cases of Application … 101
x2 =t Q=kT
Ds ¼ ¼ D0e ð2:13Þ
Ds
where x2 is the root mean displacement; t being the time; Q being the activation
energy; T being the absolute temperature; k being the Boltzmann constant; D0 being
the diffusion constant.
According to [153],
D0e ¼ a2 m;
where a is the «jump distance» of the migrating atom, which is approximately equal
to the crystal-lattice constant; m is the frequency factor (*1013 s−1).
The motion of the adsorbed atom in diffusion can be regarded as transition
between the adsorption centers with different binding energy. The centers are
separated with potential barriers, whose height, depending on the geometrical
structure of the surface, can be determined by dependence of the diffusion coeffi-
cient on temperature.
The idea of a potential surface texture is largely substantiated in the «pair-wise
interaction» theory [154], which represents the metal surface as consisting of
densely packed balls featuring neutral atoms pair-wisely interacting with energy
1
u ;
r6
where r is the distance between the ball centers. The energy binding the adsorbed
atom to the surface is calculated by summation of interactions with the nearest and
the farther neighbours, its value being determined by the number of neighbours.
The energy of adsorption on any face can be calculated from the known facet
structure and the atom size.
The energy of migration on different faces is also calculated by change of the
number of the nearest and other neighbours when transiting from the basic state to
the «on-the-seat» state between «wells».
Certainly, the energy of migration differs, depending on the face structure and
the physical properties of adsorbate and adsorbent atoms. Experimental determi-
nation of migration energies for the widest possible set of crystal faces, adsorbates
and adsorbents (if possible) is an important objective in studying of a solid surface.
Solution of this problem will give information on energy «texture» of the surface,
substantiate the pair-wise interaction theory and enable evaluation of the binding
energies of adsorbed atoms in different areas of the surface.
102 2 Experimental Equipment and Technique
These problems became solvable only with the advent of a field emission
microscope, enabling a researcher to observe movement of an adsorbate on a clean
surface of an adsorbent under the conditions of ultra-high vacuum with the enor-
mous magnification of M 105 times and the resolution of −20 Å. Such resolution
does not differentiate between individual atoms, but allows accurate observation of
the diffusion front, measurement of the rate of its movement and determination of
the diffusion coefficient of Ds and the activation energy of Qm with (2.13).
Visual observation of the diffusion edge is based on the emission contrast
between the surface coated with adsorbed atoms and the clean surface of the
adsorbent, which is caused by their different emissive capacity. Adsorption changes
the transmission factor of surface barrier D (Ex, F) through change of its height (e.g.
change of the work function) or width (e.g. enhancement of the local field), which
significantly affects the field emission current.
If the diffusion front is clear cut, it is not difficult to measure diffusion parameters
to a high accuracy. Otherwise, measured is the time of transition between the two
surface states fixed visually or by the field emission current.
To study diffusion processes with a field emission microscope, the latter is
charged with a source evaporating the adsorbate under investigation to the side of
the apex. The evaporator is usually a tungsten or molybdenum spiral charged with a
substance or a wire holding an adsorbate as a weight. The designs of more
sophisticated sources are also detailed in original articles. The mutual positioning of
the evaporator and the apex should allow the evaporated adsorbate to get only to a
part of the tip surface (Fig. 2.45).
When the tip is heated, the edge between the surface coated with the adsorbate
and the clean part of the surface begins moving. The field emission microscope
flask surface is usually marked to measure the time for the edge to move between
the two marks. As a result, path x in (2.13) remains constant and progression along
this path at different temperatures is timed. Migration energy Qm can be determined
by the Arrhenius equation
1 Qm
t¼ ¼ A exp ; ð2:14Þ
t kT
where t is the diffusion rate, and Qm can be determined by the slope of the line
1 1
lg ¼ f :
t T
U 1=2
r ¼ 0:15 ; ð2:15Þ
u2
If the original coating is significantly thicker than the monolayer, migration takes
place with a clear-cut edge progressing uniformly over the tip surface. The edge
begins moving at T = 20 K for hydrogen [156], T = 27–77 K for oxygen [157],
T = 40 K for nitrogen [158]. Movement of the edge roughly agrees with law (2.13).
If the evaporated portion is insufficient, the edge movement ceases and a part of the
apex surface remains uncoated. Further diffusion becomes possible only at a con-
siderable rise in temperature. Gommer [156] suggests the following mechanism
explaining the first-type migration.
A multi-layer coating resulting from condensation of a sufficient gas portion on a
lateral surface of a cooled apex consists of a chemosorbed first layer and subsequent
poorly bound layers. The chemosorbed layer is motionless at the above-specified
migration temperatures. The upper-layer particles adsorbed physically are the more
poorly bound the further they are from the surface. They can move over the
underlying chemosorbed layer. When they come to the end, they are captured at the
edge and chemosorbed as well, spreading the layer edge and creating an illusion of
its movement. The edge stops moving when the entire material of the second layer
is exhausted (Fig. 2.46).
This mechanism provides filling of all adsorption centers on the surface. The
chemosorbed layer formed in this way is monatomic though the layer of the
roughest surface areas may be significantly thicker than one atom. This type of
migration is often compared with an unfolding carpet or a drawing curtain.
If the original portion is about twice as little as that necessary for full coverage of
the apex in the first-type migration, the edge stops and its movement resumes at
higher temperatures: T = 180–240 K for hydrogen [156], 500–530 K for oxygen
[157], T = 400–650 K for nitrogen [158].
Though in the second-type migration, diffusion also proceeds with a clear-cut
edge, this edge is very anisotropic. All the above-listed adsorbates most promptly
migrate over densely packed faces {011} and channels connecting faces {011} and
{211}. These channels mostly consist of face areas {011}. These adsorbates
migrate more slowly over areas {111} and most slowly over the vicinals around the
faces {001}. Activation energies are different for different areas and average
5.9 kcal/mole for hydrogen and 25.5 kcal/mole for oxygen and nitrogen.
Undoubtedly, there is migration of the chemosorbed layer atoms in this case. The
clear-cut edge indicates the diffusive capture mechanism theoretically substantiated
in [159].
The clear-cut nature of the edge is caused by the fact that the diffusing atoms are
captured by the adsorption centers, «traps», spaced shorter than the resolution of a
field emission microscope, i.e. less than 20 Å. If there is enough adsorbate, the
diffusing atoms captured by the traps facilitate the diffusion of the subsequent
particles.
If the concentration of adsorbed atoms is very low, the traps are not filled because
the diffusion front can only progress if the adsorbed atoms have enough energies to
escape from the traps. The higher the temperature, the deeper traps become vacant.
The activation energy is determined by the energy of binding of the adsorbed atoms
at the adsorption centers with a high binding energy. The diffusion goes on without
a clear-cut edge because the distance between the similar adsorption centers
exceeds the microscope resolution. The activation energies measured grow with
reduction of coverage h. In particular, hydrogen is reported [160] to have the
following activation energies for diffusion without an edge: 9.5 kcal/mole at
h = 0.1 − 0.3; 16 kcal/mole at h = 0.01 and 64 kcal/mole at h 0.
Papers [156–158], as well as [159, 161–163], dedicated to different gas-metal
systems, give substantial reasons to the point of view that increase in the binding
energy, i.e. desorption heat, with reduction in the concentration is connected not so
much with interaction of the adsorbed particles with each other, as with presence of
various adsorption centers not only on the surface having a number of crystal faces,
but also on a separate face of single crystals, except for the most densely packed
ones [164].
Presence of gas affects the rate and the activation energy of the surface diffusion.
In particular, Müller [162] found that oxygen and water vapors increase the rate of
the tungsten-lattice surface diffusion. Brenner [160] discovered the same effect
when iridium diffused over its own lattice.
At the same time, a number of authors note an increase in the activation energy
for the surface diffusion of adsorbed atoms with gas present. In particular, [165]
shows that oxygen and nitrogen increase the activation energy of copper-tungsten
diffusion from 17 to 25 kcal/mole. It is assumed that gases form compounds both
with the substrate and the atoms of the adsorbate. Depending on the energy binding
the atoms of the adsorbent and the adsorbate with the appropriate gas, it is possible
for the compound to migrate.
It should be pointed out that there is a budding tendency for gases to accelerate
diffusion of metals over their own lattice and decelerate diffusion of adsorbate
atoms over the tungsten surface. As of today, the papers investigating the effect of
106 2 Experimental Equipment and Technique
gases on the surface diffusion with a field emission microscope are too few to
warrant a particular judgment on this issue.
Some authors [166–169] found an effect of the value and the sign of the external
electric field on the equilibrium concentration of adsorbed atoms at the apex vertex
and on the activation energy of the surface diffusion.
The importance of field emission has increased with enhancement of the applied
component, i.e. development of effective field emission cathodes, especially those
made from carbon materials. In particular, [170] described the experiments with
freshly prepared samples of polyacrylonitrile carbon fibers and stated [170, 171]
that their volt-ampere characteristics plotted on the Fowler−Nordheim coordinates
have breaks/deflections towards smaller currents. A typical characteristic is given in
Fig. 2.47. It also shows the snapshots of field emission images of a carbon fiber
viewed on the screen.
Analysis of the images suggests that the effective surface of the emitter is
obviously irregular with chaotic heaping of emitting microelevations [172].
The breaks of the characteristics can be explained by the fact that manual
measurement of cathode characteristics during voltage boost causes reconstruction
of the surface at heavy currents and, accordingly, voltage, which is well illustrated
by the photos in Fig. 2.47. Such reconstruction shifts the volt-ampere characteris-
tics towards heavier voltages.
Reconstruction of the carbon-fiber surface is accompanied by its destruction, i.e.
removal of poorly bound crystallites, affected by ionic bombardment and pon-
deromotive loads.
In those cases, the characteristics were taken at voltage boost and were not
reproducible.
As a rule, some operation and formation make the characteristics forward. That
assumes such a range of currents and voltages that do not destroy the fiber. Voltage
boosting does not change the fiber effective-surface geometry, but increases the
emission current due to increase in the field intensity at the working
microelevations.
The emitting-surface area can be evaluated either by field emission images
viewed in a field emission microscope [173, 174] or by volt-ampere characteristics
as suggested by Oostrom [175]. If the average area of the single-fiber end face is
assumed as 4 10−7 cm2, the field emission images in Fig. 2.47 suggest that
growth of the drawn current from 2 10−8 to 5.22 10−5 A causes the
emitting-surface area also to increase from 3.2 10−11 to 1.2 10−9 cm2, i.e.
about 250 times. If each fibril is further assumed to have a diameter of 50 Å and its
end face emits uniformly, the number of the working fibrils increases in this case
from 160 to 4000 pieces, while the density of the current from each fibril increases
from 6 102 to 6.5 104 A/cm2. However, such evaluations of the emitting
surface can be exaggerated due to the electron-optical effect. Evaluation of the
emitting-surface area by its emission characteristics suggests that the
emitting-surface area remains nearly the same and equals 3.5 10−12 cm2
throughout the range of the current drawn. As each fibril end-face area is about
2 10−13 cm2, it is easy to estimate that a stable state of the field emission cathode
involves functioning of up to 15 fibrils, each providing the maximum current
density of 107 A/cm2.
A vast number of standardized fibrils, 106 pieces on the end face of a single
fiber, provide high stability of the field emission cathode emission. Ionic bom-
bardment destroys some fibrils and replaces them with identical new ones, and
stability of such a huge statistical fibrillar system is preserved until shrinkage of the
end face results in decrease in electric intensity and drop of field emission.
In the course of operation, the field emission cathode is exposed to intensive
ionic bombardment, and the ion energy is several orders of magnitude higher than
the thermal energy of atoms (3/2 kT). That is why the bombardment breaks equi-
librium in the adsorption process towards increase in desorption. When the cathode
functions, it prevents the residual-gas molecules from adsorption near the
microelevations on the effective surface of the polyacrylonitrile fiber, where
emission mostly comes from. In the adsorption mode, the residual-gas molecules
are also first adsorbed not on microelevations but in checks and pores of the fiber,
and only after they are filled, the molecules can settle down near the microeleva-
tions. That is why residual-gas adsorption scarcely affects the electron work
function from PAN fibers.
The current step-up taking place after the adsorption mode can be explained by
different mechanisms:
– change of the work function or barrier penetrability;
– change of the surface geometry;
– change of the emitting-surface area.
108 2 Experimental Equipment and Technique
changes in current, it’s the residual gas adsorption. The field emission images
obtained at the beginning of the work period are somehow different from those
obtained at the end of the previous work period; firstly, the image becomes brighter;
secondly, the previous image looks powdery, i.e. it is characterized by magnifi-
cation of the image details, which seems to be explained by additional emission
from other fiber-surface areas. Particularly interesting are the cases when fiber
cathodes are under the pressure of p = 10−1 mm Hg for a long time, usually more
than a month.
Here the original image at the beginning of the next work period after such a
long absorption period is utterly different from that obtained at the end of the
previous and the current period. That may be because the adsorbed molecules are
oriented so that the barrier becomes more penetrable. We believed that residual-gas
adsorption decreases the work function. Actually, it can both decrease and increase.
E.g. a polar molecule of water can be adsorbed by both the positive and the negative
end of a dipole to the outside; in the first case, the work function decreases, in the
second case, it increases.
References
15. S.V. Denbnovetskiy, A.B. Leshinin, G.F. Semenov, Transformation of information on the
new memorable EBT (M.: Energoizdat, 1984)
16. S.V. Denbnovetskiy, B.H. Mosienkov, T.V. Temirazeva, other, Selection and information
transfer. Kiev: Sci. Thought 45, 114 (1978)
17. V.A. Kyznetsov, E.P. Sheshin, Technique and application of analyzers of a power range of
field emission electrons. PTE 5, 7–16 (1972)
18. E.W. Mueller, Z. Physik. 120, 261 (1943)
19. A.L. Suvorov, A.F. Bobkov, B.Y. Kyznetsov, S.B. Zaytsev, Field emission microscope for
research of emission properties of faces of monocrystals. PTE 4, 248–250 (1979)
20. E.W. Plummer, R.D. Yong, Field emission studies of electronic energy levels of absorbed
atoms. Phys. Rev. B1, 2088–2109 (1970)
21. L.W. Swanson, L.C. Crouser, Phys. Rev. 163, 622 (1967)
22. D.R. Bowman, R. Gomer, K. Mutelib, M. Tringides, Surf. Sci. 138, 581 (1984)
23. R. Gomer, A. Auerbach, Surf. Sci. 167, 493 (1986)
24. R.D. Young, E.W. Mueller, Experimental measurement of the total-energy distribution of
field-emitted electrons. Phys. Rev. 113, 115–120 (1959)
25. K. Kowasaki, K. Senzaki, Y. Kumashiro, A. Okada, Energy distribution of field-emitted
electron from TiC single crystal. Surf. Sci. 62, 313–316 (1977)
26. M.I. Elinson, Radio Technician Electron. 4, 140 (1959)
27. A.G. Ghdan, M.I. Elinson, Radio Technician Electron. 6, 671 (1961)
28. A.G. Ghdan, M.I. Elinson, A.B. Sandomirsky, Radio Technician Electron. 7, 670 (1963)
29. I.L. Sokolskay, G.P. Sherbakov, Solid State Phys. 3, 167 (1961)
30. G.P. Sherbakov, I.L. Sokolskay, Solid State Phys. 4, 3526 (1962)
31. I.V. Mileshkina, I.L. Sokolskay, Solid State Phys. 5, 2501 (1963)
32. G.N. Fursey, P.G. Shlahtenko, Solid State Phys. 12, 2645 (1970)
33. F.G.I. Van Oostrom, Validity of the Fowler–Nordheim model for field electron emission.
Philips Res. Rept. 21(1), 1–102 (1966)
34. V.A. Kyznetsov, E.P. Sheshin, Power range of the field emission electrons emitted from
microledges on tungsten, an edge. Radio Eng. Edektronik. 20(7), 1550–1553 (1975)
35. G.L.R. Mair, D.C. Grindrod, M.S. Mousa, Deam-energy distribution measurements of liquid
gallium field-ion sources. J. Phys. D: Appl. Phys. 16, 209–213 (1983)
36. V.A. Ivanov, T.S. Kirsanova, T.A. Tumareva, Field emission spectroscopy of the tungsten
covered with layers of barium and an oxide of barium. Phys. Solid Body 23(3), 664–668
(1981)
37. C.I. Workowski, I.J. Czyzewski, Field emission spectrometer whis an electron multiplier
operating in the phase-sensitive detection system. Acta Physia Polonica. A39(5), 523–529
(1971)
38. A.S. Kypryshkin, L. Rybakov Yu, E.P. Sheshin, The analyzer of total energies of field
emissions electrons with the retarding potential. PTE 1, 151–153 (1990)
39. A.S. Kupryashkin, L. Rybakov Yu, E.P. Shehin, Field-electron total energy analyzer with
retarding potential. Instr. Experim. Tech. 33(1), 155–157 (1990)
40. R.D. Young, C.E. Kuyatt, Resolution determination in field emission energy analysers. Rev.
Scient. Instrum. 39(10), 1477–1480
41. R.Z. Bahtizin, U.M. Ymagyzin, Analyzer of a energy spectrum of field emission electrons.
PTE 3, 212–216 (1984)
42. R.Z. Bahtizin, V.M. Lobanov, K.G. Yu, M. Umagyzin Yu, Calculation of trajectories of
electrons in electrostatistical lenses. Radio Technician Electron. 4, 1556–1558 (1988)
43. C. Kuyatt, E. Plammer, The reflecting analyzer energy for research of field emission
electrons. Devices Sci. Res. 1, 122–126 (1972)
44. N. Kempin, K. Klapper, G. Ertl, The analyzer for removal of energy distributions of field
emission electrons. Devices Sci. Invest. 9, 52–55 (1978)
45. S.I. Shkyratov, S.H. Ivanov, S.N. Shilimanov, Field emission laboratory—the field emission
electronic spectrometer combined with a field ionic/electronic microscope. PTE 4, 126–134
(1999)
References 111
73. J. Liu, L. Arnberg, N. Backstrom, S. Savage, Mater Sci. Eng. 98, 21 (1988)
74. R. Maringer, C.E. Mobley, in Proceedings of 3rd International Conference on Rapidly
Quenched Metals, vol. 1, ed. by B. Cantror (Metals Society, London, 1978), p. 49
75. Y. Jashiro, T. Terao, A new method of specimen preparation for FIM and FEM. Surf. Sci. 67
(2), 605–610 (1977)
76. A.R. Bhatti, B. Cantor, D.S. Joag, G.D.W. Smith, Phil. Mag. B. 52, 63 (1985)
77. Pat. USA 414 3292 Cl/ 313-336, 25.06.76. Field emission cathode of glassy carbon and
method of preparation. S. Hosoki, H. Okano
78. R.J. Morgan, Sci. Instrum. 44, 808 (1967)
79. M.I. Elinson, G.F. Vasiliev, Field Electron Emission (GIFML, Moscow, 1958)
80. I.S. Andreev, Gyrn. Tehnich. Fiz. 22(9), 1428–1441 (1952)
81. E.W. Mueller, Zs. F. Phys. 108(9–10), 668–680 (1938)
82. E.W. Mueller, Zs. F. Phys. 102, (11–12), 734–761 (1936)
83. R. Haefez, Zs. F. Phys. 116(9–10), 604–609 (1940)
84. M. Benjamin, R.O. Jenkins, Nature 143, 599 (1939)
85. M. Benjamin, R.O. Jenkins, Phil. Mag. 26(12), 1049–1062 (1938)
86. M. Benjamin, R.O. Jenkins, Proc. Roy. Soc. 176(10), 262–279 (1940)
87. W.J. Tegart, The electrolytic and Chemical Polishing of Metals (Pergamon Press, Oxford,
1959)
88. P.A. Jacquet, Metall. Rev. 1, 157 (1961)
89. B.M. Tsarev, Achievements Phys. Sci. 36(2), 181–209 (1948)
90. M.I. Elinson, V.A. Gorkov, G.F. Vasiliev, Mag. Radio Eng. Electron. 2(2) (1957)
91. M.B. Benyminovich, B.G. Smirnov, G.N. Shyppe, Mag. Phys. Phys. (release 10) 23, 1690–
1699 (1953)
92. A.L. Syvorov, Devices Technol. Exp. 5, 5 (1969)
93. E.P. Sheshin, Structure of a Surface and Field Emission Properties of Carbon Materials
(MIPT-fizmatgiz Publishing House, 2001)
94. G.H. Freyberg, Production of thin field emitters. PTE 6, 176 (1967)
95. H.A. Hubner, Field emitter etching facility with good reproducibility. Optik 63(2), 179–183
(1983)
96. G.N. Freyberg, The improved semiconductor scheme of etching tips. PTE 5, 233–234 (1970)
97. A.J. Melmed, J.J. Carroll, Vac. Sci. Technol. 2, 1388 (1984)
98. A.J.J. Melmed, Chem. Phys. 38, 607 (1963)
99. R.-I.G. Garber, G.I. Dranova, N.A. Mansyrov, I.M. Mihaylovsky, Devices Technol. Exp. 1,
196 (1969)
100. A.S. Kypryshkin, A.G. Shahovskoy, E.P. Sheshin, The stabilized high-voltage power
supply. PTE 4, 238–239 (1991)
101. H. Lemke, T. Goddenbenrich, H.P. Bochem, U. Hartmann, C. Heiden, Improved microtips
for scanning probe microscopy. Rev. Sci. Instrum. 61(10), 2538–2541 (1990)
102. V.Y. Fridman, Production field emission tips by etching. PTE 1, 227 (1974)
103. G.N. Freybrg, The improved semiconductor scheme of etching tips. PTE 5, 233–234
104. A.c. №293515 USSR. Method of production of needle cold emitters. N.I. Komyak, V.G.
Pavlov, A.A. Rabinovich, V.N. Shrednik, Publication in B.I. 1975.31
105. A.c. №568981 SSSR. The device for etching field emitters. D.M. Pautov, A.V.
Kocheryzhnikov. C. HO1 J 9/02 of 04.01.76. Publication in B.I., 1977, No. 30
106. G.N. Freyberg, Double bath for automatic etching tips. PTE 4, 244–245 (1972)
107. E. Mueller, T. Tsong, Field Jon Microcopy (American Elsevier Publishing Company, Juc.,
New York, 1969)
108. V.I. Afanasev, V.M. Tybaev, T.I. Karpenko, T.I. Ivchenko, Electrochemical receiving thin
needles in a stream of electrolyte. PTE 1, 195–196 (1983)
109. H. Morikawa, K. Goto, Rev. Sci. V. Instrum. 59, 2195 (1988)
110. H. Morikawa, K. Goto, F. Iwatsu, T.J. de Terao, Physique 46(6), 589 (1987)
111. A.C. USSR 213200 Kl.HO1 J 1/30 of 8.06.66. A way of production of needle fieldemitters.
R.-I.G. Garber, V.I. Afanasyev, I.M. Mikhaylovsky
References 113
112. J.M. Walls, H.N. Southworth, B.J. Rushton, The preparation of Field electron/field ion
emitters by ion etching. Vacuum 24(10), 475 (1975)
113. M. Hellsing, Mater. Sci. Technol. 4, 824 (1988)
114. M. Drechsler, J.P. Prulhiere, Procedes et dispositifs pour usiner, reaffuter, chanffer et
nettoger des pointes par bombardment electronigue, patent of France No. 2098954, kl.
B 23 K 15/00 of 30.07.70
115. G.I. Dranova, B.V. Kylko, I.M. Mihaylovsky, Method of production of the needle field
emission cathode, a.s. 630669, Cl. HO1 J9/02, of 27.06.77
116. J.A. Liddle, A. Norman, A. Cerezo, C.R.M.J. de Grovenor, Physique 49(6), 509 (1988)
117. V.A. Kyznetsov, E.P. Sheshin, Energy spectrum of the field electrons emitted from
microledges on a tungsten tip. Radio Technician Electron. XX(7), 1550–1553 (1975)
118. T. Gurney, F. Hutchinson, R.D.J. Young, Chem. Phys. 42, 3939 (1965)
119. J.P. Jones, Nature 211, 479 (1966)
120. G.D.W. Smith, J.S. Anderson, Surf. Sci. 24, 459 (1971)
121. W.R. Graham, R.A. Reed, F.J. Hutchinson, Appl. Phys. 43, 295 (1972)
122. J.P. Jones, A.D. Martin, Surf. Sci. 41, 559 (1974)
123. A.P. Janssen, J.P. Jones, Surf. Sci. 41, 257 (1974)
124. G.L. Kellogg, Surf. Sci. 192, 879 (1987)
125. O. Nishikawa, Y. Tsunahima, E. Nomura, M. Wada, S. Horie, M. Shitata, T. Yoshimara,
R. Uemori, Surf. Sci. 126, 529 (1983)
126. T.T. Tsong, S.C. Wang, F.H. Liu, H. Chung, M. Ahmad, J. Vac. Sci. Technol. B1, 915
(1983)
127. E.D.J. Boyes, Comm. Metals 26, 207 (1972)
128. A.L. Emmanue, A.R. Moore, H.M. Pollock, Phys. Stat. Sol. 28a, 511 (1975)
129. R.D. French, M.H. Richman, Phil. Mag. 18, 471 (1968)
130. M.H. Richman, W.D. Sproul, Metallography 2, 149 (1969)
131. N. Ohmae, A. Nakamure, S. Koike, M.J. Umeno, Vac. Sci. Technol. A5, 1367 (1987)
132. S.V. Krishnaswamy, R. Messier, Y.S. Ng, T.T. Tsong, S.B.J. McLane, Non Cryst. Solids. 35
(36), 531 (1980)
133. S.V. Krishnaswamy, R. Messier, S.B. Mclane, Y.S. Ng, T.T. Tsong, Thin Solid Film 79(21)
(1981)
134. S.V. Krishnaswamy, R. Messier, C.S. Wu, S.B. McLane, T.T. Tsong, J. Vac. Set Technol.
18, 309 (1981)
135. K. Rendulic, E.W. Mueller, J. Appl. Phys. 38, 550 (1967)
136. C.C. Schubert, C.L. Page, B. Ralph, Eiectrochima Acta 18, 33 (1973)
137. G. Bauer, M.J. de Keisch, Physique 47(7), 189 (1986)
138. O. Nishikawa, T. Utsumi, J. Appl. Phys. 44, 955 (1983)
139. W.R. Graham, F. Hutchinson, D.A. Reed, J. Appl. Phys. 44, 5155 (1973)
140. E.S. Machlin, A. Freilich, D.C. Agrawal, J.J. Burton, C.L. Briant, J. Microsc. 104, 127
(1975)
141. J.A. Panitz, Rev. Sci. Instrum. 56, 572 (1985)
142. J.A. Panitz, I. Giaever, Ultramicroscopy 6, 3 (1988)
143. H. Norden, K.M. Bowkett, J. Sci. Instrum. 44, 238 (1967)
144. B. Loberg, H. Norden, Phil. Mag. 16, 1147 (1967)
145. B. Loberg, H. Norden, Proceedings of 4th European Regional Conference on Electron
Microscopy, Rome, p. 251 (1968)
146. D.A. Smith, K.M. Bowkett, Proceedings of 4th European Regional Conference on Electron
Microscopy, Rome, p. 261 (1968)
147. K.M. Bowkett, D.A. Smith, Field-Ion Microscopy (North-Holland, Amsterdam, 1970),
p. 215
148. M.J. Goringe, D.A. Smith, Jernkont. Ann. 155, 502 (1971)
149. R.M. Hatapova, L.L. Demskay, V.H. Romanova, Manufacturing techniques of carbon field
emitters. PTE 3, 205-2-7 (1985)
114 2 Experimental Equipment and Technique
150. G.N. Shyppe, Electron Emission of Metal Crystals (SAGA Publishing House, Tashkent,
1959)
151. I.L. Sokolskay, Application of an field emission microscope for studying of surface diffusion
and self-diffusion. “Surface diffusion and spreading” under the editorship of Y.E. Geguzin,
M. Nauk of pp. 108–148
152. V. Zayt, Diffusion in Metals (M, Il., 1958)
153. J.E. Lennard-Jones, Proc. Phys. Soc. 49, 140 (1937)
154. I.N. Staranski, R. Suhrmann, Ann. Phys. 6 Folge 1, 153 (1947)
155. M. Drechsler, H.-Z. Angew, Phys. 6, 341 (1954)
156. R. Gomer, R. Wortman, R. Lundy, J. Chem. Phys. 26, 1147 (1957)
157. R. Comer, J.K. Hulm, J. Chem. Phys. 27, 1363 (1957)
158. G. Ehrlich, F.G. Hudda, Chem. Phus. 35, 1421 (1961)
159. H. Fujita, J. Chem. Phys. 21, 700 (1953)
160. S.S. Brenner, Surl. Sci. 2, 496 (1964)
161. R. Gomer, R. Wortman, R. Lundy, J. Chem. Phus. 27, 1009 (1957)
162. E.W. Mueller, Z. Elektrochemie 59, 372 (1955)
163. R. Klein, L.B. Leder, J. Chem. Phys. 38, 1866 (1963)
164. R. Gomer. Field Emission and Field Ionisation (Harvard University Press, 1961)
165. A.J. Melmed, J. Appl. Phys. 37, 275 (1966)
166. L.W. Swanson, R.W. Strayer, F.M. Charbonnier-, Surf. Sci. 2, 177 (1964)
167. A.G. Naymovec, Solid State Phys. 6, 2088 (1964)
168. H. Utsugi, R. Gomer, J. Chem. Phys. 37, 1706 (1962)
169. Y.V. Zybenko, I.L. Sokolskay, Radio Technician Electron. 9, 1467 (1962)
170. B.V. Bondarenko, V.Y. Raevsky, E.P. Sheshin, Field electron emission carbon fibers.
Theses of reports of ShM of the All-Union symposium on not heated cathodes, Tomsk,
pp. 47–48 (1980)
171. E.P. Sheshin, in Emission characteristics carbon fibers. Physical Processes in Devices
Electronic and Laser (Tekhnikiyu-M: MFTI, 1980), pp. 6–10
172. E.P. Sheshin, Field emission of carbon fibers. Ultramicroscopy 79, 101–108 (1999)
173. B.V. Bondarenko, Y.L. Rybakov, E.P. Sheshin, Field electron emission of carbon fiber.
Radio Technician Electron. Eng. 27(8), 1593–1597 (1982)
174. B.V. Bondarenko, Y.L. Pybakov, E.P. Sheshin, Field electron and field ion microscopy of
the emitting surface of carbon fiber. Theses of reports of III conference on field ionic
microscopy, Sverdlovsk, p. 30 (1982)
175. F.G.L. Van Oostrom, Validity of the Fowler–Nordheim model for field electron emission.
Philips Res. Rept. 21(1), 1–102 (1966)
176. A.L. Suvorov, E.P. Sheshin, V.P. Babaev, Influence of adsorption of residual gases on
emission properties of carbon materials. ZhTF 66(9), 164–169 (1966)
Chapter 3
Modern Developments in Theoretical
Research of Field Emission
3.1 Introduction
or
j B
lg ¼A ; ð3:2Þ
E2 E
is also to be linear. However, the E value can only be determined on the basis of
factor b. But this factor would have a constant value of 1/r all over the cathode
surface only for a ball-shaped cathode. For an arbitrarily-shaped apex cathode, b
differs between surface elements, i.e. b = b (n, η), where n and η are apex-surface
coordinates, and function b(n, η) depends on the apex shape. A formula for function
b(n, η) can be calculated from the potential theory only for some
axially-symmetrical apex shapes, namely a rotation hyperboloid, a rotation para-
boloid and a ball on a cone centered in its vertex. Therefore, the apex is usually
approximated by one of these shapes, which is referred to as hyperbolic approxi-
mation, parabolic approximation etc. However, real apices merely approximate
these shapes. Therefore, the error of determination of b(n, η) and, consequently,
E = bV, turns out essential, usually ΔE/E 30%. Moreover, the real apex surface,
even a single-crystal one, reveals different crystal faces having different work
functions Фikl, i.e. Ф = Ф(n,η). Generally speaking, inhomogeneity of the electric
field above the apex surface in different points of this surface and inhomogeneity of
the apex surface by work functions prevent from determination of field emission
current J from the apex by the Fowler–Nordheim formula derived for a cathode
with a definite work function on condition of homogeneity of the electric field
above the cathode surface. J can be determined by summing currents dJ coming
from particular elements dS of the cathode surface. Taking into account that
E = bV, current dJ is
" #
3=2
k 2
ðn; gÞ 0:965 S U ðn; gÞ
dI ¼ B0 V 2 exp 0
dS; ð3:4Þ
U0 ðn; gÞ V bðn; gÞ
consequently,
Z " #
3=2
0 k2 ðn; gÞ 0:965 S U0 ðn; gÞ
I ¼BV 2
exp dS: ð3:5Þ
U0 ðn; gÞ V bðn; gÞ
3.1 Introduction 117
It may seem that (3.4) assumes that dependence lg(I/V2) = f(1/V) cannot be
represented as a straight line and that it is difficult to verify the field emission theory
in such a way. However, a large body of experimental research of field emission has
arrived at linear dependence (3.3) is a wide range of V values and, consequently, the
E value. In particular, Hoffmann [1] measured that dependence (3.3) is linear within
18 orders of field emission current, and only very low and very high electric field
intensities cause appreciable deviations from linearity (Fig. 3.1). So the linear
dependence (3.3) experimentally determined in a wide range of voltages and cur-
rents seem to prove that the Fowler–Nordheim field emission theory is qualitatively
true. The linearity can be attributed to strong dependence of j on the value of Ф3/2/
b, so the measured real current comes from small local areas of the apex, where the
value of Ф3/2/b is minimal. It is this minimum that determines the slope of char-
acteristic (3.3).
Linearity deviations of dependence (3.3) in the area of high current densities, i.e.
at high field intensities, can be attributed to either
(a) effect of the space charge of the field-electron flow, which makes field intensity
E dependent on the current density and, consequently, non-linearly dependent
on potential difference V, i.e. value b is not constant if V changes;
(b) possible deviation of the law of the forces acting on the electron near the
surface at very small distances from the surface, from the electrical-image force
law.
No justified hypotheses of the nature of deviations of dependence lg(I/V2) = f(1/V)
from linearity in the area of small currents has been suggested yet.
Quantitative verification of the theory necessitates accurate determination of the
work function by an experiment. Comparison of the experimental and theoretical
curves shows that their slopes coincide in the order of magnitude. However, this
-3
-7
-11
-15
-19
0,5 1 2 3 4 5 6 104/V, V-1
118 3 Modern Developments in Theoretical Research of Field Emission
Z1 " ( )#1
4pmekT 8p21=2 m1=2 jEj3=2 hðe3=2 E 1=2 Þ
j¼ 1 þ exp
h3 3heE jEj
1
EþU
ln 1 þ exp dE; ð3:6Þ
kT
(a) 10
(b) 1,8
9 1,6
T2=2000 K
8 1,4
I 1=
10
1,2
5
7 lg(I2/I1)
A/
cm
lgI2
) 2
6 1,0
I =1
lg(I2,A/cm2)
lg(I2/I1)
10
lgI1
6
I=
A/
5 0,8
10 7
cm
1
2
A/
)
cm
4 0,6 2
)
T1=1000 K
3 0,4
2 0,2
= 3,23 eV
1
1 2 3 4 5 6 2,0 2,5 3,0 3,5 4,0 4,5
104/V,V-1 eV
Fig. 3.2 Illustration of the method for determining the work function with field emission
measurement data used by Elinson et al. [2, 3]. a Experimental dependences lgI from 1/V for two
temperatures T1 and T2. b Theoretical dependences g(I2/I1) from Ф for three selected current
densities
I
S ¼ 4 1017 p2 ; ð3:7Þ
V2 1=V¼0
U3=2
p ¼ 2:82 109 ;
b
the current being expressed in amperes, the length in meters and the work function
in electron-volts. The first ratio is based on assuming that value of Ф2 exp
(9.81 Ф−1/2) is roughly constant. For the second one one needs to know value Ф
itself, which results in decrease in precision as the local work function value is
measured with great difficulty and low precision.
The Fowler–Nordheim experimental line determines the emitter parameters,
namely geometrical field factor b for given work function values and the emitting
surface area S.
If the emitter surface is coated with an adsorbate, the expressions (I = jS) and
(3.7) remain true for this new surface. Supposing that adsorption does not seriously
affect the b factor value, causing changes only in the work functions, we can
evaluate this change from formula (3.7):
120 3 Modern Developments in Theoretical Research of Field Emission
2=3
Uads pads
¼ ; ð3:8Þ
Ucln pcln
where Ucln and Uads are the work function values before (a clean surface) and after
adsorption, respectively; pcln, pads are the Fowler–Nordheim line slope ratios before
(a clean surface) and after adsorption, respectively [4].
Equations (3.7) and (3.8) result from the assumption that the local electric field
and the work function are constant over the emitter surface and the emission area
does not change with voltage variation. Such assumptions are only true if the
electron emission current is measured from the individual face of the emitter
[5, 6].
The authors [7] have shown that the experimental work function value obtained
with the above-described technique is close to the minimum of the real work
function heterogeneously spread over the surface.
Most research papers are physically unreasonable when assuming the work
function to be constant over the emitter surface. However, the 10% scatter of work
functions over the surface and the 30% scatter of field intensities increase the scatter
of current densities more than twice [8].
On the other hand, ordinary observations of the emitter surface and the
emission area in the field emission microscope show changes of the emission area
even for a small interval of applied voltages V (ΔV/V = 10%) necessary to plot the
Fowler–Nordheim equation. In the course of increase in the applied voltage,
heterogeneity of the work function manifests itself in increase in the emission area
as emission from the areas with high work function becomes non-negligible in
case of strong fields [9]. However, the emission area all the same goes up with
increase in the field due to heterogeneity of distribution of the field over the
emission area near the apex foundation even in a simulation case at Ф = const
over the emitter surface [10].
There is one more method of experimental determination of the emission area
having no those disadvantages. The experimental value of the emission area S of a
tungsten cathode can also be evaluated from the emission image [7] according to
the formula
2
r
S ¼ Sim : ð3:9Þ
rim
Knowledge of the apex shape is necessary but not enough to determine the field
over its surface. It is difficult to determine the field under real conditions, i.e. for a
real apex shape and with account of the anode geometry and the arm supporting the
apex etc. To find distribution of the field from the potential theory, researchers
usually resort to approximations.
The field calculated for a model shape of the surfaces of the emitter and the
anode is macroscopical and adequate only if the distance from the surface exceeds
*3/4 size of atomic-size surface irregularities [24]. It is about 1–2 nm [23]. The
local electric field immediately at the surface often considerably exceeds the
macroscopical field because of atomic-size surface irregularities. Calculation of the
local field from electrostatics can be inadequate. It is necessary to make
quantum-mechanical calculations of electron-charge distribution. This challenge
was only met for some special cases, which is why there were no theoretical
estimates of the field enhancement factor,
E
c¼ ; ð3:10Þ
Emacro
where E is the electric field at the emitter surface, Emacro is the macroscopic electric
field.
Value c is used as an adjustable parameter to compare the results of simulation
and experiment. It ranges *1–3.
122 3 Modern Developments in Theoretical Research of Field Emission
is a function depending only on this coordinate in the area limited to two coordinate
surfaces connected with one common curvilinear coordinate qi, i = 1, 2, 3…
As a result, if
V ¼ V1 for q1 ¼ n1 ; 8 q2 ; q3 ð3:13Þ
and
V ¼ V2 for q1 ¼ n2 ; 8 q2 ; q3 ð3:14Þ
then the Laplace equation solution (3.11) depending on coordinate q1 can be rep-
resented as
3.1 Introduction 123
Z
1
V ¼ C1 dq1 þ C2 ; ð3:16Þ
Fðq1 Þ
where C1, C2 are the constants that can be found from boundary conditions (3.13),
(3.14).
Having found out the electric potential, one can find the intensity of electric field
E as a gradient of function V:
X3
1 @V
E ¼ grad ðVÞ ¼ pffiffiffiffiffi ~
ei ; ð3:17Þ
i¼1
gii @qi
C1
E ¼ pffiffiffiffiffiffi ~
e1 : ð3:18Þ
g11 Fðq1 Þ
The Laplace equation solution for potential V and electric field intensity E for the
coordinate surfaces of the second order in the respective curvilinear coordinates is
illustrated by a two-sheeted hyperboloid of revolution.
Potential V and electric field intensity E for a revolution hyperboloid, a revo-
lution paraboloid and flattened and extended revolution ellipsoid [25] in the
respective curvilinear coordinates are specified in Table 3.1.
According to the authors of the paper [22], who microscopically probed a
considerable number of emitters viewed from different sides at different processing
stages, a real emitter is shaped closely to one of equipotential surfaces of an electric
field generated by a charged orthogonal cone topped with a sphere. This equipo-
tential at the very top is well approximated by the sphere. It is followed by a
transitional area, which eventually turns into the orthogonal cone.
Potential distribution for an insulated charged orthogonal cone with a spherical
end is given by expression [22]
Table 3.1 The electric field potential and intensity in curvilinear coordinates. Here C1 is a
constant determined from the respective boundary conditions. It is individual for every revolution
surface
Revolution surface Potential V Field strength E
Two-sheet hyperboloid C1 ln
tg g
þ C2
2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
C1
eg
ch2 lcos2 g0
a sin g0
Elongated ellipsoid C1 ln
th l2
þ C2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
C1
el
a sh l0 ch2 l0 cos2 g
Paraboloid C1 ln ðlÞ þ C2 pCffiffiffiffiffiffiffiffiffiffi
1 ffie
2 l
l0 2 l0 þ k
VR n
V¼ ðr a2n þ 1 r n1 ÞPn (cos h). ð3:19Þ
Rn
Naturally, the potential distribution will not change if one equipotential surface
is taken as an emitter and the other as an anode. In (3.19), V is the potential counted
from the cone, r and h are ordinary polar coordinates, R is the distance to the anode,
VR is the applied voltage, a is the sphere radius, Pn is the Legendre first-kind n-order
function of the, n being selected such that if h = a where a is an external semiangle
of the cone, Pn = 0 (n = 0 for a = 0, n = 1 for a = p/2). The equation of the
theoretical anode is derived from the equation U = UR In this case, the summand
a2n þ 1 r n1 r n is neglected and
Rn
rn ¼ : ð3:20Þ
Pn ðcos hÞ
Analyzing the equipotential surface shapes for orthogonal cones with different
angles and different sphere diameters, it is quite easy to select the equipotential that
is good at approximation of the experimental emitter.
Let us turn to Fig. 1.2. Fowler and Nordheim (cf. Chap. 1) calculated the density of
field emission current from energy band A at T = 0. If T > 0, emission involves not
only these electrons, but also the electrons of bands B, B1 and those higher in
comparison with the case when T = 0. The problem becomes much more compli-
cated: firstly, we can no longer use the limiting expression for distribution of electrons
by energies; and, secondly, for energies of the electrons lying close to the top of the
potential barrier, we can no longer use the expression for transparency D found in
Chap. 1. Gas and Mallin [37] were the first to find an analytical solution of calculation
of D for the case of T > 0. Later Andreev [42] applied the same scheme to make
calculations giving essentially similar, but somewhat simpler expressions for fun-
damental thermal field emission regularities. That was how the first thermal-field
emission theory was developed. Later, more general assumptions excluding the
restrictions accepted in paper [37] enabled numerical solution of the problem [39, 43].
Let us give a brief calculation of transparency of the barrier and the current for
band B1. Schrödinger equations for areas x < 0 and x > 0 can be written down as
d2 w
dx2 þ Ew
¼0 ðx\0Þ;
d2 w ð3:21Þ
dx2 þ EWþ 1
2x þ x
2x20
w¼0 ðx [ 0Þ:
1 ðx x0 Þ2
V ¼W þ : ð3:22Þ
x0 2x30
This expression shows quite good approximation of the barrier near the maximum.
Some transformations and subsequent insertion of expression (3.22) into (3.21)
126 3 Modern Developments in Theoretical Research of Field Emission
turns the latter into a parabolic cylinder equation. The integration methods for such
an equation are well-known.
Linking the two solutions as usual
dw dw
w þ 0 ¼ w0 ; ¼ ;
dx þ 0 dx 0
suitable for C-band electrons with the energies close to the top of the barrier and the
expression
ð3:24Þ
true for a high-temperature case, after which it is easy to calculate the density of
emission current from band B1 and band C near the top of the potential barrier.
Density of current from band B1 and C
(
u0
2 kT
X
1
ð2n þ 1Þd
jB1 þ C ¼ BðkTÞ e ð1Þn
n¼0 nðn þ 1Þd2 þ d 1
2
1
2
W W 6
2 cos UC
1 2 4n o12
2kT 4x0 2 ð3:26Þ
pa þ kT1 þ b2 a20
39
>
=
cos UB1 7
þn o 15 :
2 2 >;
2pa 1 þ b2 a2 kT 0
pffiffiffiffiffiffiffiffi
Here U0 ¼ U e3 E; where U is the work function; d = 2pa0kT;
3 12
x
a0 ¼ 0 :
2
In the limiting case of high temperatures and weak fields, the current-density
expression turns into the famous Schottky formula if we neglect a small periodic
member, assume d 1 and use the first member of the sum
0
jB1 þ C ¼ B ðkTÞ2 eU =kT : ð3:27Þ
A better understanding of the temperature effect of field emission can come from the
cases of higher fields (107–108 V/cm) and low and average temperatures (300–
1500 K). Under these conditions, electrons from the B-band, mostly those coming
from its lower part, start playing an increasingly more important role in emission.
The total density of current of bands A and B can be expressed by the final
formula
(
E2 X1
ð1Þn þ 1
j¼ 1:55 106 þ 120; 5T 2
h2 U n¼1
n
!)
1 1
1 þ 1 ð3:28Þ
n þ 8:813 103 hU2 TE n 8:813 103 hU2 TE
" 3
#
6:85 107 U2 h
exp A/cm2 ;
E
In the presence of external electric field E, potential (3.28) gets an addition eEx.
As a result, the potential of the field that meets the electron in electron-emission
experiments is given by the approximate expression:
EF þ U e2 =4x eEx; x [ xc ;
VðxÞ ¼ ð3:30Þ
0; x [ xc ;
where xc is determined from the condition that V(xc) = 0. Figure 3.3 shows a
potential barrier for typical value EF þ U and external electric field intensity
characteristic for field emission experiments. Please note that potential (3.29) peaks
at
2 Metal Vacuum
0
0 10 20
z,Ǻ
3.2 Thermal-Field Emission (TFE) from Metals 129
In the version considered, the TFE theory is built as follows. Let us assume
(which is quite reasonable as electric field inside the metal can be neglected) that the
number of electrons falling on the barrier surface from the metal bulk with energy
ranging from W to W + dW does not change during emission and equals its equi-
librium value of
Z1
mdW mkB T W EF
n ðW; TÞ dW ¼ 2 3 f ðEÞdE ¼ ln 1 þ exp dW: ð3:33Þ
2p h 2p2 h3 kB T
W
The probability for every electron to pass through the superficial potential barrier
is equal to D (W). Consequently, emission current density, i.e. the number of
electrons emitted from the unit area per unit time multiplied by electron charge can
as before be expressed by the ratio
Z1
j ðE; TÞ ¼ e n ðW; TÞ D ðWÞ dW; ð3:34Þ
0
where T and E designate temperature and the field applied, as usual. Murphy and
Good showed that one can obtain quite a simple expression for j(E,T) that is true at
any values of E and T.
The potential barrier-transmission coefficient is given by
where
Zz2
Q ðEÞ ¼ 2i kðnÞdn; ð3:36Þ
z1
1=2
k ðnÞ ¼ ð2m=h2 Þ ðWEF U þ e2 =4n þ eEnÞ ; ð3:37Þ
k2 ðxÞ ¼ 0: ð3:38Þ
For the case of W < Vmax, roots x1 and x2 are both valid and x1 < x2. For the case
of W > Vmax, roots x1 and x2 are mutually complex conjugate with Im(x1) > 0 and
130 3 Modern Developments in Theoretical Research of Field Emission
Im(x2) < 0. Please note that according to (3.37), function k(n) has singularity in a
point n = 0. It does not affect applicability of (3.33) in cases where W < Vmax, but it
cannot be ignored at the energies significantly exceeding Vmax when Re(x1) = Re
(x2) approaches zero. If
pffiffiffi
E [ EL Emax þ ð1 1= 2Þ ðe3 EÞ1=2 ; ð3:39Þ
then (3.35) can turn out unsuitable for application. However, for the energies
W > WL, the transmission coefficient is very close to zero, which is why, for the
current density to be assessed, it can be assumed that
D ðW; EÞ 1: ð3:40Þ
Zp=2
K ½k ¼ ð1 k2 sin2 hÞ1=2 dh; ð3:45Þ
0
Zp=2
E ½k ¼ ð1 k2 sin2 hÞ1=2 dh: ð3:46Þ
0
Please note that y > 1 corresponds to the energies above the barrier maximum
(W > Vmax), a y < 1 corresponds to the energies below this maximum (W < Vmax).
Table 3.2 gives the values of v(y) for y
1 calculated in [35].
3.2 Thermal-Field Emission (TFE) from Metals 131
Z1
j ðE; TÞ ¼ e N ðW; TÞ D ðW; EÞdW
0
8W
ZL
emkB T < W EF dW
¼ ln 1 þ exp ð3:47Þ
2p2 h3 : kB T 1 þ exp QðWÞ
0
9
Z1 =
W EF
þ ln 1 þ exp dW :
kB T ;
WL
the Fermi level. So the second integral in (3.47) can be omitted, and the limits in the
first integral can be replaced with −∞ and +∞. In addition, for typical values of U
and E, U = 3 eV, E < 6 107 V/cm:
so the one in the denominator of the first integral can be omitted in (3.47). Then we
get
Z1
emkB T W EF
jðE; TÞ ¼ exp½QðWÞ ln 1 þ exp dW: ð3:49Þ
2p2 h3 kB T
1
The main contribution to this integral is given by the vicinity of the Fermi level.
So the exponent of the transmission coefficient can be replaced with the first two
members of Taylor expansion in powers of WEF :
where
1=2 3=2
b0 ¼ ð4=3Þ 2m=h2 U =eE vðy0 Þ; ð3:51Þ
b0 ¼ 0:683v 3:79F 1=2 =U U3=2 =E: ð3:52Þ
1=2
f0 ð1=2Þ 2m=h2 vðy0 Þ ½eEU1=2 ð1e3 E=U2 Þ1 ; ð3:55Þ
where
Table 3.2. shows the numerical values of function t(y). Please note that under the
conditions of field emission, y < 1. Having kept only the first two members in
(3.50) and inserted them in (3.49), we will obtain
Z1
emkB T W EF
j ðE; TÞ ¼ exp ðb0 Þ exp ½c0 ðW EF Þ ln 1 þ exp dW:
2p2 h3 kB T
1
ð3:60Þ
Please note that at c0kBT < 1, the integrand in (3.60) decreases exponentially at
W\EF , behaves in the same way as the initial expression under the integral sign in
(3.49) at W [ EF and coincides with it near the Fermi level. The integral in (3.60)
can be calculated analytically to produce the field emission current density
" #
e3 E 2 pc0 kB T 4 2m 1=2 U3=2
j ðE; TÞ ¼ exp vðy0 Þ : ð3:61Þ
16p2 hUt2 ðy0 Þ sin ðpc0 kB TÞ 3e
h2 E
It can be shown that the expression (3.61) for emission current density holds true
if the following two conditions are met:
1=2 1=4
U e3 E [ p1 h4 = m2 e2 ðeEÞ3=4 þ kB T ð1 c0 kB T Þ1 ; ð3:62Þ
Figure 3.4 shows the field emission and TFE bands for U = 4.5 eV determined
by ratios (3.62) and (3.63). For the points lying on the border of the field emission
band, the error in the current-density value (3.61) changes within the range of 15–
40% in comparison with the exact evaluation in (3.47).
At very low temperatures, i.e. when pc0 kB T 1, it is reduced to the
well-known Fowler–Nordheim equation,
where
h 1=2 i1
A0 ¼ e3 16p2 hUt2 e3 E =U ; ð3:65Þ
1=2 3 1=2
B0 ¼ ð4=3eÞ 2m=h2 v eE =U : ð3:66Þ
134 3 Modern Developments in Theoretical Research of Field Emission
n
ctro
ele
3000
ion ermo
h
em a of t
iss
Are
2000
0
0,2 0,4 0,6 0,8 1,0 E, V/Å
Having inserted the respective numerical values for different constants, we can
find that
" ! #
1:537 1010 E 2 0:683U3=2 3:79E 1=2
j ðEÞ ¼ 2 exp v ; ð3:67Þ
Ut ð3:79E 1=2 =UÞ E U
where
The numerical values of s(y) are given in Table 3.2. The table shows that s(y) is
almost constant in the narrow range of variation of the field applied. Consequently,
it is theoretically possible to determine the work function from the slope of the
Fowler–Nordheim plot. In practice, determination of the work function is impeded
by impossibility of sufficiently precise determination of the absolute value of the
local field on the emitting plane. In addition, please note that the factor of pro-
portionality preceding U3/2 in (3.68) depends on the selection of the shape of the
surface potential barrier.
3.2 Thermal-Field Emission (TFE) from Metals 135
In conclusion of the paragraph, please note that Murphy-Good formula (3.61) for
the TFE-current density can after some transformations be compacted as follows:
ðT6¼0Þ ðT¼0Þ px
jTFE ¼ jTFE ; ð3:70Þ
sin px
E A C
φe
φc
je φc
jc
Fig. 3.5 The potential distribution in the three electrode system [emitter (E)—anode (A)—
collector (C)] suitable for measuring the distribution of field emission electrons energy
Let us now turn to the theoretical aspects of the method [32, 44, 46, 51, 52].
Energy distribution of field electrons. The number of electrons emitted from the
unit area per unit time with full energy in the range from E to E þ dE can be written
down as J ðEÞ dE. Function J ðEÞ is known as full-energy electron distribution. Let
us designate the full-energy distribution calculated within the framework of the
theory of metals in free-electron approximation as J0 ðEÞ. What we get is
ZE
J0 ðEÞ ¼ N ðE; Ex ÞDðEx ; EÞdW; ð3:72Þ
0
where NðE; Ex Þ and DðEx ; EÞ are determined by the ratios given in Chap. 1 and the
previous sections of this chapter. According to the assumptions made above, the
potential barrier can be given as (3.30) and the barrier-penetrability value can be
found according to (3.35)–(3.41). In the field emission band, the bracketed one in
(3.35) can be omitted according to (3.62) and (3.63) in the same way as has already
been done when calculating the emission-current density. What we will as a result
will be
ZE
J0 ðEÞ ¼ f ðEÞðm=2p h Þ2 3
exp ½QðEx ÞdEx ; ð3:73Þ
0
where Q ðEx Þ is known from (3.41). QðEx Þ can be substituted by the first two
members of Taylor expansion in the vicinity of Ex .
As a result, we have
3.3 Field Emission Spectroscopy. Energy Distribution … 137
v(yE) and t(yE) are the functions given in Table 3.2. If (3.74) is inserted into (3.73),
the resultant integrand is small everywhere except for the vicinity of point Ex ¼ E.
Therefore, the lower limit of the integral can be substituted with −∞. Then ana-
lytical integration produces
" #
mdE f ðEÞ 4 2m 1=2 ðU þ EF EÞ3=2
J0 ðEÞ ¼ exp vðyE Þ : ð3:78Þ
2p2 h3 3 h2 eE
j0 expðM=d0 Þ
J0 ðMÞ ¼ ; ð3:79Þ
ed0 1 þ expðM=pd0 Þ
where
M E EF ; ð3:80Þ
p kB T=d0 ; ð3:81Þ
d01 c0 ; ð3:82Þ
0,4
8
7
0,2 6
5
4
2 3
0
-4 -3 -2 -1 0 1 2 E/d0
Zþ 1
j ðE; TÞ ¼ e J0 ðMÞdM: ð3:83Þ
1
Figure 3.6 gives a set of theoretical curves of the full-energy distribution cal-
culated by Swanson and Krauser [54] on the basis of (3.79) for different values of
parameter p.
1,2
Т = 1800 К
1500 К
0,8
1200 К
900 К
600 К
300 К
0,4 78 К
0,31
0,26
0,21
0,16
0,13
0,1
-2 -1 0 1 2 3 , eV
140 3 Modern Developments in Theoretical Research of Field Emission
the top of the barrier has to be sensitive to the shape of the barrier itself. For the case
of stronger fields (Fig. 3.8), coordinate xm (3.31) of the barrier maximum counted
from the metal surface is rather small, usually x 3–4 Å. That is why any dis-
crepancy between the theory and the experiment caused by inadequacy of the
image-force potential at small distances should in principle manifest itself as a
divergence from the theoretical full energy distribution curve in the area of the
strong fields. Certainly, such quantitative analysis assumes independent knowledge
of the work function and precise measurements of field E on the emitter surface. At
this point of our reasoning, when we proceed from the data taken from Fig. 3.8, we
can only say that the experimental results correlate with our assumption that the
potential image-force barrier holds true for the distances of 3–4 Å from the
metal-vacuum boundary.
In addition to the energy spread, the electrons approaching the surface barrier
also have some angular dependence. The foregoing consideration assumes energy
E to be the energy connected with the momentum component orthogonal to the
surface. However, it is also necessary to consider the issue of how the results
obtained change with account of angular dependence. The internal discrepancy of
the situation comes from the fact that the surface barrier carries out a kind of energy
analysis by the momentum component orthogonal to the surface, while the spec-
trometer does that by the total energy. If the electron subsystem is slightly over-
heated, the above-described models hold adequate because the angular distribution
of electrons is narrow-beamed to the interface [58, 59].
If the electron subsystem is considerably overheated, i.e. if the electron energy
reaches several tenths of electron-volts, which is equivalent to several thousand
degrees, the particles can be located in an upper valley of the semiconductor. In
particular, when the gallium-arsenide electrons reach the energy of 0.35 eV, they
can quasi-thermalized in the L-valley. This process generates a compact group of
electrons having not only a narrow energy spectrum but also a narrow angular
localisation correlating with the position of the valley in the k-space in terms of the
crystal-lattice direction (111).
However, in this case, the changes of curves of energy distribution in the vac-
uum caused by variation of the work function are to have a prominent feature.
Indeed, let the probability of escape of this-group electrons at a work function is
B. Because of narrow localisation on the scale of energies, the energy interval
occupied by this group in the spectrum is far away from zero. It has already been
mentioned that for gallium arsenide, it is at least 0.35 eV. With decrease in the
work function, probability B goes up. The typical changes on the distribution curves
occur not at the energies of zero or close to it, but in the area separated from the
zero by the interval characteristic for a material.
This raises an extremely interesting question of preservation of the tangential
component of the electron momentum at transition through the semiconductor-
vacuum boundary. Simultaneous observance of the energy and momentum con-
servation laws imposes rather severe restrictions on the cone of electron escape in
the vacuum or the medium with a different effective mass in general (an analogue of
total internal reflection in optics). As this issue is topical in a lot of applications, it is
3.3 Field Emission Spectroscopy. Energy Distribution … 141
under lively discussion. Unfortunately, the measurements usually have the afore-
mentioned integrated nature, which is the reason why the results obtained are
sophisticatedly and often inadequately interpreted. What makes the issue even more
sophisticated is that a lot of semiconductor surfaces undergo faceting, which is why
the normal line to the macrosurface can be far from the normal line to the micro-
surface, i.e. the emitting surface. That makes the experimental angular dependences
of electron emission obtained by some authors untrustworthy, either. As a result,
while some authors claim that the results they have obtained can be well explained
subject to tangential momentum conservation law, others have to admit possibility
that it is violated, at least on the incoherent boundary, i.e. the boundary where the
long-range crystallographic order does not hold. Of course, this does not refute one
of the fundamental laws of nature. The point at issue is that when the incoherent
boundary is crossed there may be some exchange of momentum between the
electron and the crystal as a whole, which is perceived as an inconsistency of
particle momentum.
It has already been mentioned in Chap. 1 that the R. Stratton classical theory of
field emission from semiconductors is true in quantitative and qualitative terms only
as applied to field emission from n-type low-resistance semiconductors, which
follows from comparison of theoretical and experimental volt-ampere characteris-
tics. At the same time, field emission from n-type high-resistance semiconductors
and p-type semiconductors has a number of peculiarities [2, 32, 33, 44, 60–62].
Figure 3.9a shows the most complete experimental non-linear volt-ampere
characteristic of a semiconductor field emission cathode having eight characteristic
sections of change of field current with growth of the running voltage [60].
In this case, the basic features of field emission as compared to the similar
process for metals become clear through analysis of the presented volt-ampere
characteristic for field emission cathodes from p-Si (q = 3000 Ohm cm) [60], as
well as Figs. 3.10, 3.11, and 3.12. These features are listed below.
1. As a rule, the limiting current densities are much less than in case of emission
from metals;
2. The volt-ampere characteristics of lgI = f(l/U) are non-linear, i.e. they can show
the areas of saturation and proliferation of current;
3. In the saturation area, the emission current demonstrates thermo- and
photosensitivity;
4. The emission image in the saturation area tends to shrink [63], which is
indicative of a field-configuration change near the emitter surface (Fig. 3.10);
5. Pulsed emission excitation makes relaxation effects, i.e. the current-pulse shape
depends on the amplitude and duration time of anode voltage (Fig. 3.9b);
142 3 Modern Developments in Theoretical Research of Field Emission
(a)
lg I 8
-5 6
-6 5
4 C
-7 3
2
-8
5а
1
-9 B
A
D
1 2 3 4 5 6 7 8 9 104/V, V-1
(b) V
τd
I, А
1
3
Increasing of the action pulse voltage
t,μs 8
Fig. 3.9 Semiconductor field emission current according to the cathode of the p-Si
(q = 3000 Ohm cm). a linear (A) and nonlinear (B—straightforward course, C—reverse)
current-voltage characteristics, taken at T = 300 K; 1–8 portions of presented characteristics; I–III
most commonly recorded portions; D curve plotted at T = 77 K; b the waveform of the current
pulses, registered under the action of the voltage rectangular pulses at the cathode and
corresponding to the points marked on the current-voltage characteristic of arabic numerals;
rectangular pulse duration is 200 ms
R1 /R2
0,8
-0,6
0,4
0,2
2 3 4 5 6 7 U, kV
Fig. 3.10 Illustration of the compression effect of the emission images at portions 2, 3 of the
current-voltage characteristics (see. Fig. 3.9a)
(a) (b)
1,0 0,3
f,r. u.
0,2
f,r. u.
0,5
0,1
Fig. 3.11 a energy distribution curve for germanium [68], b energy distribution curve for silicon
with two maxima. The distance between the peaks of 1.1 eV [69]
(a) (b)
N ×103,pulses
1,0
0,5
15 30 15 30 45 , keV
Fig. 3.12 Statistical field emission spectra for different portions of a nonlinear current-voltage
characteristics (see Fig. 3.9a—the curves B and D): 1–4 portions for (a) and the portion 5a for (b)
144 3 Modern Developments in Theoretical Research of Field Emission
7. The experimental evidence of heating of the electron gas in the near-surface area
of the space charge of the p-type semiconductor cathode is also the field emission
statistic data from p-Si at the liquid-nitrogen temperature (curve D in Fig. 3.9a)
obtained in paper [62], which leads to the conclusion (Fig. 3.12), that the area of
fields and currents correlating to sect.1–4 in Fig. 3.9a, generates one-electron
emission acts and sections 5 generates many-electron emission acts;
8. It is also necessary to note that it follows from the experiment that electron
emission for the field ranges correlating with the volt-ampere dependence sec-
tions with a large slope (2, 3, 7 in Fig. 3.9a) weakly depends on temperature and
light, i.e. these conditions cause the semiconductor cathode to behave as a
quasi-metal one.
All these features result from relatively low concentration of electrons in the
conductivity band of semiconductors as compared with metals. The external electric
field applied to the sample to accelerate electrons distorts the energy bands of the
semiconductor and near its surface (cf. Chap. 1), shapes the screen layer of the
space electron charge. If penetrability of the potential barrier represented by
intra-crystal and external electric fields is high enough, electrons from the screen
layer tunnel into the vacuum. With growth of anode voltage due to increase in the
barrier penetrability, the tunnel emission-current density also grows up. In this part,
the pattern coincides with the one observed in metals in terms of quality.
But unlike metals, further increase in voltage and the electron-takeoff level makes
density of emission current from semiconductors depend not only on barrier pene-
trability, but also on the rate of electron inflow from the sample bulk to the surface.
For lack of multiplication of charge carriers in the internal crystal field, this flow is
limited. It is determined by the rate of their generation in the near-surface area and
stationary values of drift and diffusion. So the current density grows saturated, i.e.
the volt-ampere characteristic gets a characteristic step (Fig. 3.9a—section 4).
Further increase in anode voltage in the steady state can result in emergence of a
current-surge section (section III in Fig. 3.9a) preceding the explosive rupture of the
emitter. Many authors [46] usually attribute presence of this section to the begin-
ning of avalanche-type multiplication of current carriers in the near-surface space
charge area. When current is taken off in a quasi-steady state, e.g. when reading the
characteristic represented in Fig. 3.9a (reading time 1 s), a more sophisticated
pattern takes place in which an additional volt-ampere dependence section appears.
Analysis of the data represented in Figs. 3.9, 3.10. 3.11 and 3.12 allows the
authors of papers [43, 60–62, 64–67] to conclude that every section of volt-ampere
dependence of the p-type semiconductor cathode or the n-type high-resistance one
meets the classical Fowler–Nordheim linear form lgj = A – B/E. At that, the
electron concentration in the near-surface area of the space charge affects only value
A, and the slope of each section, i.e. value B, is determined by some effective work
function Фeff. So the first section of the volt-ampere characteristic (Fig. 3.9a) is
caused by tunnel electron emission from the surface states; the second one is caused
by tunnel emission from the valence band; the third one, i.e. the first saturation
section, is caused by tunnel emission of “hot” electrons from the conduction band
3.4 Phenomenological Theories of Field Emission from Semiconductors 145
because of penetration of the electric field into the near-surface area of the space
charge, the rate of generation and recombination of carriers looking approximately
identical; the fourth section is caused by tunnel emission of “chilled” electrons,
wherein penetration of the field into the near-surface area of the space charge results
in the field “throwing” of electrons from the valence band to the conduction band,
i.e. increase in concentration of electrons, and consequent displacement of the field
from the near-surface area; the fifth one is caused by mixed emission: one part of
“hot” electrons penetrates the barrier though tunneling, while the other part goes
that over the barrier; in the sixth section, i.e. the second saturation section, all
electrons emit over the barrier; the seventh section may be caused by thermal
breakdown. It is worth noting that it is not always that thermal breakdown results in
destruction of the field emission cathode. Reduction of time to take the volt-ampere
characteristic causes transition to the eighth section, i.e. the third saturation section,
which is apparently caused by mere thermionic emission.
Elinson model. The above-presented qualitative, hypothetical pattern of field
emission from p-type semiconductors and n-type high-resistance semiconductors
has had no rigorous theoretical substantiation yet though the theoretical research
carried out later than Stratton’s papers had been published was particularly focused
on theoretical volt-ampere dependences similar to the experimental ones at least in
terms of quality.
In our opinion, the papers most physically grounded and mathematically correct
were those by Elinson et al. [2, 38, 65, 66], even though they aimed at solution of a
rather specific problem, i.e. an attempt to answer the question as to why the
experimental field emission characteristics in the area of heavy current densities are
usually steeper than those in the area of low current densities. The authors of the
papers specified explained this phenomenon by deflections from thermodynamic
equilibrium in energy distribution of electrons of the conduction band. These
deflections are more pronounced when internal electric fields are large. With growth
of the internal electric field, the average electron energy grows up, and the prob-
ability of their penetration through the surface potential barrier increased. Further
on, large internal electric fields can increase charge-carrier concentration due to
internal field emission, ionisation by collision etc. Theoretical consideration of field
emission with large current densities is rather difficult. Elinson et al. [65, 66]
attempted to solve this problem, and the way they selected seems to be most
adequate to the actual state of things.
The rudest approximation that could affect correctness of the theoretical con-
clusions was that spatial dependence of electron concentration near the semicon-
ductor surface was not taken into account. This dependence can result in the
respective spatial dependence of electron temperature. If penetration of the field
makes electron concentration in the surface area much higher than that inside the
crystal, it is quite reasonable that electron temperature on the surface is much less
than that calculated for the crystal bulk. This effect was taken into account in paper
[67]. Before featuring the findings of this paper, let us make some comments on the
most precise account of spatial dependence.
146 3 Modern Developments in Theoretical Research of Field Emission
with electron temperature T*(x), then we obtain the following system of equations:
dS 12mn S2
jE þ ndðT TÞ ¼ 0; ð3:85Þ
dx Tl2
dE 4pe0
¼ ðn n1 Þ; ð3:86Þ
dx j
dn n dT
j ¼ ne0 lE þ e0 d þ ; ð3:87Þ
dx 2T dx
where
1=2 1=2
2kB T dT T T
S¼ j 2kB d ; l ¼ l0 ; d ¼ d0 : ð3:88Þ
e0 dx T T
The values used in these equations stand for the following. S is the density of the
energy flow transferred by the electrons, T is the lattice temperature, E is the electric
field intensity, j is the electric-current density, l is the electron free path, n is the
electron concentration, d is the electron diffusion constant, l is electron mobility.
Application of the concept of electron temperature T* assumes presence of the
3.4 Phenomenological Theories of Field Emission from Semiconductors 147
• at x = 0
j ¼ je ; ð3:91Þ
where
1=2
kB T kB T
E0 ¼ ; x0 ¼ l ;
e0 x0 12mn s2
Here
e0 n1 kB T 1=2
j0 ¼ e0 n1 l0 E0 ; jT ¼ ;
2 2mn
m0 ¼ mðx ¼ 0Þ;
s0 ¼ sðx ¼ 0Þ; x0 ¼ xðx ¼ 0Þ; ð3:94Þ
!
tðuÞ
D ðtÞ ¼ exp k1 1=2 :
x0 u3=2
148 3 Modern Developments in Theoretical Research of Field Emission
dg E 1=2 b
u¼ ; t¼ ; dg ¼ ax0 ; k1 ¼ ;
wkB T t kB T a1=2
3=2 1=2
4c0 e0 2m e0 F0 j1=2
b¼ ; c20 ¼ ; a¼ :
3ðkB TÞ 2
h2 kB T
-3
a
-5
-7
0,02 0,03 0,04 ω0-1
3.4 Phenomenological Theories of Field Emission from Semiconductors 149
-4 50 I
2
-6
25
4
II
-8
0 0
25 50 75 100 ω0
10 2
5 1 II
I
0,900 0 0
2 4 6 8 10 ξ
Area b. In this area, the electron temperature as the external-field function surges
(curve I in Fig. 3.14). What happens is transition to thermoemission, which reduces
electron concentration on the semiconductor surface (curve II in Fig. 3.14).
The field emission characteristic is presented in Fig. 3.13. It clearly shows three
areas a, b and c.
Area c. The current almost reaches saturations. Almost all electrons are emitted
thermally. Electron concentration n = m0 n∞ can be even less than n∞; nevertheless,
electron temperature as the external-field function goes on growing (curve I in
150 3 Modern Developments in Theoretical Research of Field Emission
90 III
II
85 2
I
90,600 80
1
90,590 75
90,580 70 0
1 2 3 4 ξ
Fig. 3.14). In this case, spatial dependence of the internal-field intensity is inap-
preciable (Fig. 3.16). Figure 3.16 shows that the enhanced electron concentration
on the surface causes reduction of electron temperature.
Under almost invariable conditions, increase in electron affinity w reduces cur-
rent density, thus causing smaller deflections from thermodynamic equilibrium.
Our opinion is that the theoretical current-voltage diagram (sections a, b, c,
Fig. 3.13) obtained by M.I.Elinson et al. at least qualitatively meets sections 6, 7
and 8 of the experimental current-voltage diagram (Fig. 3.9a).
Theory of A.F.Yatsenko
In his research Yatsenko [76] made an attempt to solve the problem of com-
putation of volt-ampere characteristics of semiconductor emitter that would corre-
spond with the most common observed volt-ampere characteristics (VAC)
consisting of three cardinal areas of change in the current with increase of voltage
(cf. Fig. 3.9a).
Additionally, one of the goals of his paper was to provide a theoretical expla-
nation of experimentally determined high photosensitivity of p-type semiconductor
samples under influence of fields corresponding to area 2 of VAC (cf. Fig. 3.9a).
Probably because of this Yatsenko named his theoretical paper “On modeling of
photo field emission out of p-type semiconductors”.
The object of modeling is lightly doped p-type semiconductor (i.e. the acceptor
concentration Na < 1019 cm−3) at relatively low temperature, but still with enough
heat so that small acceptors would be fully ionized. A flat field emission diode is
approximated with a metal-insulator-semiconductor (MIS) structure (Fig. 3.17) that
is different from the usual in that the insulator prat is represented by vacuum with field
strength (in working mode) up to 107 V/cm. Leaking of charge from the semicon-
ductor into the insulator is an unintended and often harmful phenomenon in most
MIS-devices, but in given case such leakage (i.e. emission) is the base of its function.
The area of volume charge of the emitter can be separated into three parts:
1. Base
2. Area of exhaustion
3. Inverse layer
3.4 Phenomenological Theories of Field Emission from Semiconductors 151
(b)
ty
jes
jed Nd
For each part a problem of field, potential and current distribution is solved.
When considering the states of electrons in inverse layer one should take in
account the quantization of their movement along the field direction. Results of
those computations are presented in Table 3.3, from which one can see that due to
quantization the electron gas in inverse layer is not always degenerated.
The states of electrons of inversion layer were measured in effective mass
approximation.
The charge density in inversion layer is defined by level of balance
dnS nS
¼ gðd d0 Þ ; ð3:96Þ
dt se
where se is the electron lifetime in inversion layer relative to emission into vacuum;
d is the depth of exhausted area; g is the total speed of generation of current carriers;
d0 is the Debye radius of screening.
The emission time se is by definition dependent on the potential barrier trans-
parency and thus on field strength.
In stationary case the emission current density can be written as:
enS E
je ¼ ¼ : ð3:97Þ
se 4pðse þ g=Na Þ
In areas with weak fields the emission currents increases with increase of field
with relation close to the Fowler–Nordheim law:
h i
je
E 5=3 exp ðw1=2 U3=2 =w2 Þtðw=UÞ : ð3:98Þ
se Na =g; ð3:99Þ
je ¼ gE=4pNa : ð3:100Þ
ðmÞ
jðmÞ
e
eNa td ;
ðmÞ
where td is the limit drifting speed of the electrons.
From (3.97)–(3.100) one can see that it is only with sufficiently strong fields
when se = Na/g is emission density considerably increasing with increase of elec-
tron generation speed in the area of exhaustion:
Theoretical VAC calculated in [77, 78] for p-type Ge is shown on Fig. 3.18. The
two curves on this figure correspond to different equations for motility presented in
the caption. The authors of [77, 78] pointed out that the experimental values of
saturation current must be greater that ones calculated due to conical geometry of
real-life emitter which leads to increase of saturation current density up to one or
two orders of magnitude as compared to theoretical estimates (Fig. 3.18). The
following relationship between mobility and the field strength inside the emitter are
assumed:
8
< ln ð0Þ;
> E\102 V/cm;
curve 1 : ln ðEÞ ¼ ln ð0Þð102 =EÞ1=2 ; 102 \E\104 V/cm;
>
:
l ð0Þ 103 E; E 104 V/cm:
( n
ln ð0Þ; E\102 V/cm;
curve 2 : ln ðEÞ ¼
ln ð0Þð102 =EÞ1=2 ; E [ 102 V/cm:
Figure 3.19 presents the results of calculation of length l (defined as the distance
between points where the concentration is twice the minimum value) of an area of
field emission cathode exhausted by carriers depending on E0, performed in those
studies. Lvov et al. also noted that increase of l with growth of E0 in stable current
only emerges due to field entering the semiconductor. Let’s remark that the cal-
culated depth of penetration of the field into the semiconductor reaching 1 mm at
the saturation area is large relative to the tip radius. This means that the
one-dimensional theory would not work very well for quantitative estimates
whenever current saturation occurs. This theory overestimates both length of
exhausted area and length of area of field penetration. One has to note yet another
result of authors of those papers. They found that the concentration of electrons in
the zone of permissibility first slowly increased with increase of E0, and then is
almost constant in a rather wide range of E0, and finally drops abruptly with further
101
100
2,2 2,6 3,0 3,4 Е0-1, Å/V
3.4 Phenomenological Theories of Field Emission from Semiconductors 155
10-2
10-3
10-4
10-5
0 3,4 3,8 4,2 Е0, 107 V/cm
increase of the field, when the emission current is saturated and becomes practically
stable.
Results similar to ones presented on Fig. 3.19 were also obtained for n-type
silicon. In this case, unlike p-type semiconductors, the saturation current density
was significantly dependent on the concentration of impurities.
The transfer phenomena changing the shape of the heated crystals of pure metals
were theoretically studied by Hering [82, 83] with application of the principles of
thermodynamics. The equations featuring variations in the shape as a result of
surface and bulk diffusion, evaporation-condensation and viscous flow for the
isotropic crystal was solved ad hoc by Mullins [84, 85].
The Mallins equations also give a criterion to evaluate the role of each process in
particular cases. As shown in paper [86], the basic transfer process in field emission
emitters is surface diffusion.
According to Hering, mass transfer by diffusion can only take place if accom-
panied by reduction of the total free energy, which necessitates a gradient of
chemical potential. For a curved surface, Hering derived an expression linking the
difference between chemical potentials of atoms and gaps with the free surface
energy:
@2a @2a
l lh ¼ l0 þ X0 R1 a þ þ R1
a þ P xx ; ð3:102Þ
1
@n21 2
@n22
D0 eQ=kT
JM ¼ ðrlÞM ; ð3:103Þ
A0 kT
JM
r
158 3 Modern Developments in Theoretical Research of Field Emission
(a) (b)
Y α=30°
16
10 30
12
10 20
8 15
Y
6
20 10
20695 4 8
10 0,8 2525 10454 2 5
4
3
0 10 20 30 40 50 60 70 80 90 z 0 2 4 6 8 10 12 14 16 z
Fig. 3.21 Sharp profiles obtained by numerical calculation: a a/2 = 2—terminus spheroidization
occurs, b stationary forms with a/2 3°
3.5 Theoretical Aspects of Phenomena and Processes … 159
between the potential energies of the atom embedded into the edge of the flat face
and the planarly adsorbed atom. In the two-dimensional gas state, atoms can only
move upon energization with Qm, which is determined by the difference between
the two potential minima for this face. Consequently, surface self-diffusion will be
an activated process with the activation energy of Q = Qbind + Qm, and the
mass-transfer rate proportional to diffusion constant DS will be determined by the
following expression:
Qbind þ Qm
v
DS ¼ D0 exp : ð3:106Þ
kT
where C0 0,5 is a unitless coefficient, (dz/dt)0 is the shortening rate beyond the
field determined according to (3.105). From (3.104) one can see that the apex
blunting rate is determined by the ratio of electrostatic pressure and the
surface-tension forces.
The direction of the electric field applied does not matter because of quadratic
dependence. Expression (3.107) makes it possible to determine the field intensity at
which (dz/dt)E = 0 and the vertex point reaches equilibrium
160 3 Modern Developments in Theoretical Research of Field Emission
1=2
8pa
E01 ffi : ð3:108Þ
rc
In other surface points, the migratory processes still result in diffusion of atoms
to the periphery, so equilibrium is labile. The field intensity providing equilibrium
in the local area of the hemisphere-shaped apex vertex is determined by the fol-
lowing ratio:
1=2
16pa
E02 ffi : ð3:109Þ
rc
It can be expected that if E > E2, the migratory fluxes will be re-directed.
According to [82], that provides for the processes of ordinary or complete reshaping
based on delivery of the substance to the vertex. I.L. Sokolskaia figured that this
effect can be of secondary importance [80]. In her opinion, an important part in
diffusion in the field is played by the atoms of the plane-adsorbed phase abundant
on the surface of the heated rounded apex. The faces are completed and the faceted
equilibrium shape is reached by the reason that the field pulls the polarized atoms to
the face edges, where the field is the most intense and where they integrate into the
crystal plane (Fig. 3.22).
At that, I.L. Sokolskaia definitely figured that the mentioned bright spots meet
the qualitatively new objects, so-called “microelevations”, which grow on the
emitter surface affected by the threshold field of
1:6 10 V=cm. As the
microelevations easily disintegrated when heated beyond the field, I.L. Sokolskaia
considered them structureless, a “chaotic pile of atoms”. However, I.L. Sokolskaia
also assumed that there are certain conditions, including relatively low temperatures
and the fields that are high but insufficient for desorption by the field, under which it
is possible to grow up irregular bumps resistant to heating at low temperatures
beyond the field.
M
3.5 Theoretical Aspects of Phenomena and Processes … 161
It was shown earlier that for the apices subjected to critical reshaping, the field
emission current can be more than doubled and followed by the preexplosive
phenomena, in particular, a spontaneous current growth.
The effect observed can be explained on the basis of the solution of the
thermal-conductivity equation with account of the Nottingham effect [94, 95].
It is possible to consider a non-reshaped rounded emitter as an analogue of the
cylinder (Fig. 3.23a). The problem of heating of such an emitter by the flowing
current was considered in paper [25].
At that, the losses of heat due to radiation or the Nottingham effect were
neglected. Neither was the dependence of physical constants of the emitter material
on temperature taken into account. Then the conductivity equation [25] is given by
@ 2 T mc @T j2 q
¼ ; ð3:110Þ
@l2 k @t k
where a = j2 q/k, C′ and C″ are the integration constants determined from boundary
conditions
Tjl¼0 ¼ 0; ð3:112Þ
@T
¼ 0: ð3:113Þ
@l
l¼L
I I
T/Tmax
Fig. 3.23 Profiles of “smoothed” (a) and the “rearranged” (b) emitter and the temperature
distribution (c) at a steady current
162 3 Modern Developments in Theoretical Research of Field Emission
Condition (3.112) means that the emitter base is supported at a particular tem-
perature that is assumed to be zero. Condition (3.113) means that there is no heat
flowing through the emitter vertex at l = L (Fig. 3.23a). Under the boundary con-
ditions, the temperature peaks at the emitting vertex [25]:
It has already been noted above that at high temperatures, the Nottingham effect
results in cooling of the emitter vertex. At that, the peak temperature area biases
deep into the apex (Fig. 3.24). Vertex cooling is equivalent to the heat flux through
it.
To take this circumstance into account, we will demand fulfillment of the fol-
lowing condition instead of (3.113):
@T
¼ P; ð3:115Þ
@l
l¼L
2000 1
1000
2
1 2 3 4
r ×104, cm
3.5 Theoretical Aspects of Phenomena and Processes … 163
P ¼ ah: ð3:116Þ
T ¼ 0:5al2 þ a ðL hÞ l: ð3:117Þ
The emitter profile is given in Fig. 3.23a. The qualitative temperature distribu-
tion graph is shown in Fig. 3.23c, curve 1 (cf. with Fig. 3.24).
The emitting vertex temperature is determined by the following expression:
ð3:118Þ
@ 2 T1 j21 q
¼ ; ð3:119Þ
@l2 k
At that, it is assumed that field emission current I goes through the vertex of the
cylindrical bump. Strictly speaking, such problem cannot be treated as
one-dimensional, but we neglect this circumstance for qualitative evaluations.
For the combined (3.110) (given that ∂T/∂l = 0) and (3.119), we demand ful-
fillment of the following boundary conditions:
Tjl¼0 ¼ 0; ð3:120Þ
@T
¼ 0; ð3:121Þ
@l
l¼Lh
where
a1 ¼ j21 q=k;
i.e. functions T and T1, as well as their derivatives with respect to L, are continuous.
The qualitative graph of distribution of temperature T1 is shown in Fig. 3.23c, curve
2.
If l = L, expression (3.125) gives us the temperature of reshaped emitter apex:
T1a h2 a
1
¼1 2 1 \1: ð3:127Þ
Ta L 2Lh a
Ta Ta
given h ! 0 T1 ! 1; h ! L T1 ! aa1 , which has to be the case.
a a
Figure 3.23c, as well as expression (3.127) imply that there are particular modes
when a set full current I supplied to the reshaped field emission cathode apex heats
it less than the filtered current. So the preexplosive effects will manifest themselves
at heavy field emission currents, which is the case in the experiment.
Qualitative reduction of heating of the pitted emitter vertex can be explained as
follows. The field emission current density through the apex has somewhat
increased. So Joule heat can be expected to heat the near-apex area up to a higher
temperature. However, increase in current density intensifies cooling due to the
Nottingham effect. If the apex temperature is higher than the inversion temperature,
this process prevails over heating, which gives this effect.Obviously, the following
condition is to be met:
Qout
1; ð3:128Þ
Qrel
I 2 qsp
where Qrel ¼ p2 r 4 is the volumetric density of the energy released in the near-apex
WIs
area of the emitter (Fig. 3.23a) within the time of flowing of current I; Qout ¼ epr2ph is
the volumetric density of the energy carried away from the near-apex area. W is the
energy carried away during the emission of one electron, e is the electron charge.
Ratio (3.128) can be reduced to
Wpr 2
1: ð3:129Þ
eqhI
The volumetric density of the energy released and carried away from the
near-apex area for the reshaped cathode (Fig. 3.23b) can be determined from the
following expressions:
I 2 qsp WIsp
Q0rel ¼ 4
; Q0out ¼ :
p2 ðr DÞ epðr DÞ2 h
Qout
[ 1;
Qrel
166 3 Modern Developments in Theoretical Research of Field Emission
where DQrel ¼ Q0rel Qrel ; DQout ¼ Q0out Qout .After transformation we obtain
DQout Wp 1
¼ :
DQrel ehIq ðr DÞ2 þ r 2
As D r,
DQout Wpr 2
ffi 1: ð3:130Þ
DQrel 2ehIq
Comparison of expressions (3.129) and (3.130) shows that the field emission
current making the prevailing Nottingham effect for a reshaped cathode can be fully
twice as high as the current for a rounded emitter. This result is in good agreement
with the experiment.
It is worth adding that the model considered does not give a full description of
non-stationary heating processes of reshaped field emission cathodes. Nevertheless,
its approximate nature does not prevent it from qualitatively and conditionally even
quantitatively correct description of behavior of a reshaped apex cathode as its
vertex is heated up less.
References
15. L.N. Kantorovich, A.S. Foster, A.L. Shluger, A.M. Stoneham, Role of image forces in
non-contact scanning force microscope images of ionic surfaces. Surf. Sci. 445, 283–299
(2000)
16. N.M. Miskovsky, S.H. Park, J. He, P.H. Cutler, Energy exchange processes in field emission
from atomically sharp metallic emitters. J. Vac. Sci. Technol. B 11(2), 366–371 (1993)
17. S. Georgieva, D. Vichev, K. Drandarov, Computer simulation of the emission process of
some field emission alloy ion sources. Vacuum 47(10), 1143–1144 (1996)
18. N.V. Egorov, E.M. Vinogradova, Solution of boundary-value problem in bispherical
coordinates. in Proceedings of 3-th International Workshop: BDO-96, St. Petersburg,
pp. 274–278 (1996)
19. G. Mesa, E. Dobado-Fuentes, J.J. Saenz, Image charge method for electrostatic calculations in
field emission diodes. J. Applied Phys. 79(1), 39–43 (1996)
20. Y. Ohkavara, T. Naijo, T. Washio, S. Oshio, H. Ito, H. Saitoh. Field emission properties of
AlZnO whiskers modified by amorphous carbon and related films. Jpn. J. Appl. Phys. 40(12),
7013–7017 (2001)
21. K.L. Jensen, J.E. Yater, Advanced emitters for next generation rf amplifiers. J. Vac. Sci.
Technol. B, 16(4), 2038–2049 (1998)
22. M.S. Yermoshina, Mathematical modeling of electron emission from a point cathode of
complex configuration. Kand. Diss. SPb. 112 p (2004)
23. R.G. Forbes, C.J. Edgcombe, U. Valdre, Some comments on models for field enhancement.
Ultramicroscopy 95, 57–65 (2003)
24. P.J. Birdseye, D.A. Smith, G.D.W. Smith, Analogue investigation of electric field distribution
and ion trajectories in the field ion microscope. J. Phys. D: Appl. Phys. 7 (1974)
25. W.W. Dolan, W.P. Dyke, J.K. Trolan, The field emission initiated vacuum area. II. The
resistively heated emitter. Phys. Rev. 91(5), 1054–1057 (1953)
26. J.R. Barbour, W.W. Dolan, J.K. Trolan et al., Space-charge effects in field emission. Phys.
Rev. 92(1), 45–51 (1953)
27. J.K. Trolan, J.R. Barbour, E.E. Martin, W.P. Dyke, Electron emission from a lattice step on
clean tungsten. Phys. Rev. 100(6), 1646–1649 (1955)
28. W.P. Dyke, F.M. Charbonnier, R.W. Straer et al., Electrical stability and life of the heated
field emission cathode. J. Appl. Phys. 31(5), 790–805 (1960)
29. J.R. Barbour, F.M. Charbonnier, W.W. Dolan et al., Determination of surface tension and
surface migration constants for tungsten. Phys. Rev. 117(6), 1452–1459 (1960)
30. F.M. Charbonnier, J.R. Barbour, L.F. Garret, W.P. Dyke, Basic and applied studies of field
emission at microwave frequencies. Proc. IEEE 51(7), 991–1004 (1963)
31. W.P. Dyke, Field emission. A new practical electron source. IRE Trans. Mil. Electron. 38–45
(1960)
32. A. Modinos, Field, thermionic and secondary electron emission spectroscopy. M. Nauka, 320
(1990)
33. M.I. Elinson (ed.), The cold cathodes. Sov. Radio 386 (1974)
34. W.W. Dolan, W.P. Dyke, Temperature and field emission of electrons from metals. Phys.
Rev. 95, 327–332 (1954)
35. E.L. Murphy, R.H. Good, Thermionic emission, field emission and transition region. Phys.
Rev. 102(6), 1464–1473 (1956)
36. R.H. Good, E.W. Mueller, Field emission. In Handbuch der Physik (ed. By S. Flugge), Bd.
21, 176–231 (Springer, Berlin, 1956)
37. E. Guth, C.J. Mullin, Electron emission of metals in electric field. Phys. Rev. 61(5–6), 339–
348 (1942)
38. M.I. Elinson, The emission of electrons under the influence of strong electric fields. Dokt.
Diss., L: LPI n. MI Kalinin (1961)
39. S.G. Christov, General theory of electron emission from metals. Phys. Stat. Sol. 17(1), 11–26
(1966)
40. F.I. Itskovich, On the theory of field emission from metals. Part I. ZETP, 50(5), 1425–1437
(1966)
168 3 Modern Developments in Theoretical Research of Field Emission
41. F.I. Itskovich, On the theory of field emission from metals. Part II. ZETP 52(6), 1720–1735
(1967)
42. I.S. Andreev, The study of electron emission from the metal in its transition from cold to
thermionic. J. Tech. Phys. 22, 1428–1441 (1952)
43. S.C. Miller Jr., R.H. Good Jr., Phys. Rev. 92, 13–67 (1953)
44. M.I. Elinson, G.F. Vasiliev, Field emission. M. Fizmatgiz, 272 (1958)
45. L.N. Dobretsov, M.V. Gomoyunova, Emission elekcronics. M. Nauka, 364 (1964)
46. R. Fischer, H. Neumann, Field emission from Semiconductors. M. Nauka, 215 (1971)
47. V.T. Cherepin, M.A. Vasiliev, Methods and tools for the analysis of material surface
Handbook (Naukova Dumka, Kiev, 1982), p. 400
48. J.E. Henderson, R.E. Badley, Phys. Rev. 38(3), 590 (1931)
49. J.E. Henderson, R.K. Dahlstrom, Phys. Rev. 55(5), 473 (1939)
50. J.E. Henderson, R.K. Dahlstrom, F.R. Abott, Phys. Rev. 41(1), 261 (1932)
51. L. Feldman, D. Meyer, Basics of analysis of surfaces and thin films. M. Mir, 344 (1989)
52. J.D. Caret, B. Feyrbah, B. Heaton et al. The use of electron spectroscopy for surface analysis.
H. Ubach (ed.). Riga Zinatne, 315 (1980)
53. D.A. Orlov, M. Hoppe, U. Weigel, D. Schwalm, A.S. Terechov, A. Wolf, Energy distribution
of electrons emitted from GaAs (Cs, O). Appl. Phys. Lett. 78, 2721 (2001)
54. L.W. Swanson, L.C. Crouser, Phys. Rev. 163, 632 (1967)
55. A.B. El-Karen, J.C. Wolfe, J.E. Wolfe, J. Appl. Phys. 48, 4749 (1977)
56. J.W. Gadzuk, E.W. Plummer. Phys. Rev. 3, 2125.58 (1971)
57. A.E. Bell, L.W. Swanson, Phys. Rev. B. 19, 3353 (1979)
58. V.A. Korablev, Y. Kudinov, MSh Sugainov, T.A. Baranova, Photoelectron spectroscopy with
angular resolution of GaAs with a negative electron affinity. Raditehnika i Elektronika. 32,
321 (1992)
59. G. Vergara, A. Herrera-Gomez, W.E. Spicer, Electron transverse energy distribution in GaAs
negative electron affinity cathodes: Calculation compared to experiment. J. Appl. Phys. 80,
1809 (1966)
60. N.V. Egorov, V.R. Tolstyakov, Investigation of the effect of the surface state on the emission
characteristics of semiconductor photo field cathodes. Surface 8, 23–33 (1996)
61. D.A. Ovsyannikov, N.V. Egorov, Mathematical modeling of systems for formation of
electron and ion beams (Publishing of St. Petersburg State University, Russia, 1998). 276 p
62. N.V. Egorov, V.R. Tolstyakov, The effect of multi-particle tunneling in the field electron
emission from semiconductors. Povrhnoct’ 9, 10–13 (1996)
63. N.V. Egorov, On the possibility of narrow collimated electron beams. ZTP. 52(12), 2440–
2442 (1982)
64. N.V. Egorov, A.G. Karpov, Diagnostic information and expert systems. SPb. (St. Petersburg
State University Publishing House, Russia, 2002) 472 p
65. M.I. Elinson, Effect of internal electric fields in the semiconductor at its field emission.
Raditehnika i Elektronika. 4,140–142 (1959)
66. A.G. Zhdan, M.I. Elinson et al., Raditehnika i Elektronika. 7, 570 (1962)
67. M.I. Elinson et al., Raditehnika i Elektronika. 10, 1288 (1965)
68. Y.A. Frenkel, ZETP. 8, 1893 (1938)
69. W. Franz, Ergeb. exapt, Naturwiss. 27, 1 (1953)
70. W. Franz, Handb. Phys. 17, 155 (1956)
71. R.J. Hodgkinson, Proc. Phys. Soc. 82(58), 1010 (1963)
72. Y.A. Frankel. ZETP. 7, 1069 (1937)
73. R. Stratton, Phis. Rev. 126, 2002 (1962)
74. M. Sanchez, Helv. Phys. Acta. 36, 1 (1963)
75. H. Frohlich, B.V. Paranjape, Proc. Phys. Soc. 69(21), 866 (1956)
76. A.F. Yatsenko, On a model photo-field emission from p-type semiconductors. Phys. Stat.
Solid. 1(2), 169–175 (1970)
77. G.N. Fursei, M. Kaplan, O.I. L’vov, On the theory of field emission from semiconductor of
p-type. Vestnik Leningrad. Universiteta. Ser. Fisiki I Himii. 16, 167–170 (1968)
References 169
78. L.M. Baskin, O.I. Lvov, G.N. Fursey, Generation features of field emission from
semiconductors. Phys. Stat. Sol. B. 47, 49–62 (1971)
79. V.M. Zhukov, Processes on the surface under field emission. SPb (VVM, Russia, 2007), 295
p.
80. I.L. Sokolskaya, Application of a field emission microscope to study the surface diffusion and
self-diffusion. ed. by Y.E. Geguzina, Procedinds of “Surface diffusion and spreading” Nauka,
108–148 (1969)
81. B. Honigman, Growth and shape of the crystals. M. IIL, 224 (1961)
82. C. Herring, Structure and properties of solid surface. ed. By R. Gomer, S.S. Smith (University
of Chicago Press, USA, 1953), pp. 5–72
83. C. Herring, The physics of powder metallurgy. ed. by W.E. Kingston (McGrow Book Co., N.
Y., 1953), p. 143
84. W.W. Mullins, Theory of thermal growing. J. Appl. Phys. 28(3), 333–339 (1957)
85. W.W. Mullins, Flattening of nearly plane solid surface to capillarity. J. Appl. Phys. 30(1), 77–
83 (1959)
86. J.L. Boling, W.W. Dolan, Blunting of tungsten needles by surface diffusion. J. Appl. Phys. 29
(3), 556–559 (1958)
87. E.W. Mueller, Oberflachenwanderung von Wolfram auf dem eigenen Kristallgitter. Z. f. Phys.
126(7–9), 642–665 (1949)
88. R.C. Sanwald, J.J. Hren, Surf. Sci. 52, 697 (1975)
89. F.A. Nichols, W.W. Mullins, Morphological changes of a surface of revolutions due to
capillarity-induced surface diffusion. J. Appl. Phys. 36(6), 1826–1835 (1965)
90. Y.I. Frenkel, On the surface atoms crawling and natural facet roughness. ZTP. 16(1), 39–50
(1947)
91. E.W. Mueller, Weitere Beobachtung mit dem Feldelektronenmicrosko. Z. Phys. 108, 668–
680 (1938)
92. M. Benjamin, R.O. Jenkins, The distribution of autoelectronic emission from single crystal
metal points. I. Tungsten, Molibdenium, Nickel in the clean state. Proc. Roy. Soc. (A) 95,
262–279 (1940)
93. I.L. Sokolskaya, Surface migration of tungsten atoms in an electric field. ZTP. 26, 1177–1184
(1956)
94. V. Zhukov, A.A. Almazov, The spontaneous growth of the field electron emission
(FEE) current to rebuild tip emitters. XXI All-union Conf. Emission Electron. L. 1, 241
(1990)
95. N.V. Egorov, V.M. Zhukov, The effect of increasing the current to rebuild emitting surface of
the field electron cathode. Poverhnost. Phys. Chem. Mech. 3, 48–53 (1995)
96. S.C. Miller Jr., R.H. Good Jr., A WKB-Type approximation to the Schrodinger equation.
Phys. Rev. 91(1), 174–179 (1953)
Chapter 4
Simulation of Structure and Parameters
of Field Emission Cathodes
The review methods of computation of D (cf. Chap. 1) used to develop the classical
theories of field emission, i.e. of Fowler–Nordheim for metals and of Stratton for
semiconductors, shows that the mathematical assumptions simplifying computa-
tions do not result in statistically significant deviations. However, it stands to reason
that these assumptions, especially when they are not correct in terms of physics, can
manifest themselves when it is necessary to take “subtle effects” into account.
That is, there is an obvious need to develop new penetrability computation
methods chiefly taking into account a need to use them in solution of the problems
unrelated to defining the potential in the near-surface region. It is a question of
penetrability-function recovery by a set of experimental data, i.e. the energy spectrum
of outgoing electrons obtained as a result of computer or natural experiment. The
small work function variation method to be described below is specific in the respect
that if its conditions are met, penetrability can be found disregarding the barrier shape.
There is one more advantage from the mathematical point of view, as this method is
not bound with the conditions of one-dimensionality or stationarity. According to the
purpose designated, this paragraph will consider the basic results of papers [1–14].
Figure 4.1 shows the barrier-penetrability recovery results.
Dependence of the solution on the regularization parameter was studied with
different values. The pictures specify the bracketed value as recommended. The
4.0×10-4
2.0×10-4
0.0
linearity assumption (Fig. 4.1) seems to have failed for the barrier penetrability,
which should have shown exponential dependence.
In our opinion, the best result came from the assumption of quadratic depen-
dence of the solution (Fig. 4.1). A comprehensible result came for a wide range of
regularization parameters.
Let us make a mathematical model of electron transport through a potential
barrier that combines the internal distribution of electrons by energies, a potential
barrier of an unknown shape and detected external distribution of electrons. Let
distribution of electrons inside the metal follows the law of
pffiffiffiffi E
nint ðEÞ ¼ N E exp ; ð4:1Þ
kT
It is this value that we will regard as the detector observed. Having replaced the
work function and the electron energy with the DФ value, we can find the changed
distribution of electrons n0ext ðEÞ after the outflow into vacuum.
Under the conditions of our experiment, let us employ the expression
~ ðEm Þ ¼ 1:
DðEm Þ 1; or D
4.1 Simulation of a Potential Barrier and Barrier Permeability … 173
~ ðE1 Þ; i ¼ 1; m. Basing on
and recover the function D
next ðEÞ
~nint ðEÞ ¼ ;
~ ðEÞ
D
we can find the distribution n~int ðEÞ. The sought-for characteristics are shown in
Figs. 4.2 and 4.3, where the difference between the values simulated and recovered
as a result of the numerical experiment does not exceed 15%.
Correct experiment simulation seems to need allowance for an error in the data to
be measured, i.e. the so-called “noise”. The small work function variation method
was studied for stability in relation to raise of noise. For this purpose, normally
distributed errors of two types, namely ones with constant and progressing variance,
were introduced into the external (detected) distribution. It was shown that the
0,6 1
2
0,4
0,2
0,0
0 1 2 3 4 5 6
Energy, eV
5,0×10 -13
0,0
Penetrability
measuring errors, solid line
with measuring errors 0,6
0,4
0,2
0,0
0 2 4 6 8 10
Energy, eV
constant-variance error makes only initial bias as compared with the penetrability
recovered without noise. The progressing error behaves in a sufficiently stable way,
which makes it possible to consider the method flexible in terms of measurement
uncertainty. An example of introduction of a measurement error is shown in Fig. 4.4.
An electric field applied to a solid surface causes changes in the shape of a
potential barrier at the interface. If the electric field intensity is high enough, it
induces the current of emitted electrons. That makes it possible to research barrier
penetrability by change of value of the electric field applied. This research tech-
nique has an advantage of a possibility to apply the modulation measurement
technique cardinally reducing the noise by narrow-band recording. It also counts in
favor of this procedure that the field emission current changes very quickly even at
small field-intensity variations [15], which can result in an intolerably large error
when measurements are made under static conditions. When modulation voltage is
applied, the error can be considerably reduced.
Modulation methods are widely used in experimental research. As a rule,
interpretation of the results obtained assumes a linear or a quasilinear relationship
between the parameters under study. However, the foregoing proves that in field
emission studies, the linearity assumption seems unreasonable. This circumstance
necessitates careful analysis of the method suggested by mathematical simulation.
A physical analogue suggested is a semi-infinite metal sample in vacuum with an
electric field applied to the solid surface and a particle energy distribution detector.
Without loss of generality, let us consider a case of an external electric field
changing in time according to the harmonic law of intensity:
an electron and the cathode surface is neglected, the penetrability can be expressed
as follows.
DðE; E0 ; E1 ; x; tÞ
" #
4ð2mÞ1=2 1=2 1 a
¼ exp ðU þ EF EÞ ¼ exp ;
hjejE0
3 1 þ EE12 cos xt 1þb ð4:4Þ
4ð2mÞ1=2 1=2 E1
a¼ U þ EF E ; b ¼ cos xt:
3hjejE0 E2
where nint ðEÞ is the density of internal distribution of electrons by energies con-
nected with their velocity components normal to the cathode surface. Insertion of
(4.5) into (4.6) makes it possible to obtain expansion of density of external dis-
tribution of electrons (signal) into degrees of cosine:
E1
next ðE; E0 ; E1 ; x; tÞ ¼ n0ext ðE; E0 Þ þ n1ext ðE; E0 Þ cos xt
E0
2 3
E 1 E1
þ n2ext ðE; E0 Þ cos2 xt þ n3ext ðE; E0 Þ cos3 xt þ . . .;
E0 E0
ð4:7Þ
8 0
> next ðE; E0 Þ ¼ nint ðEÞ exp½aðE; E0 Þ;
>
>
>
< n1 ðE; E0 Þ ¼ nint ðEÞ exp½aðE; E0 Þ aðE; E0 Þ;
ext
>
> n2ext ðE; E0 Þ ¼ nint ðEÞ exp½aðE; E0 Þ 12 a2 ðE; E0 Þ aðE; E0 Þ ;
>
>
: 3
In case of a metal cathode, external electric field variations do not make any
appreciable effect on internal distribution of electrons by energies [17]; which
means that
Having differentiated both parts of (4.6) in terms of time and neglected the
addends of about and higher in (4.7), we can obtain the following ratio based on
equality (4.8):
E0
0 0 0 0 n1ext E; E00 E10
dt D E; E0 ; E1 ; x; t dt next E; E0 ; E1 ; x; t
d d
¼ ¼ 0
: ð4:9Þ
dt DðE; E0 ; E1 ; x; tÞ dt next ðE; E0 ; E1 ; x; tÞ n1ext ðE; E0 Þ EE10
d d
Calculation of the barrier penetrability makes it possible to find the density of the
external distribution of electrons next ðE; E0 ; E1 ; x; tÞ by formula (4.6). The variation
of intensity of an external electric field results in modification of the
electron-distribution density next E; E00 ; E10 ; x; t after the escape into vacuum.
Quantum-barrier penetrability can be determined on the basis of solution of the
Schrödinger stationary equation [20] (the time for an electron to tunnel the barrier is
much less than the modulation period [21]):
4.1 Simulation of a Potential Barrier and Barrier Permeability … 177
d 2 w 2m
¼ 2 ½U ðx; tÞ E w; ð4:11Þ
dx2 h
where w is the wave function of the electron. Equation (4.11) can be solved by the
Runge-Kutta-Nyström method for different energy values and different time
moments. That enables numerical integration by e.g. a trapezium method to find
expansion of the potential-barrier penetrability into the Fourier series for a partic-
ular electron-energy value and evaluate the contribution of the addends that we
neglect when deriving formula (4.9).
The machine-assisted implementation of numerical algorithms for the mathe-
matical model presented above in operation [9] employed the cathode parameters
typical for tungsten (Ф = 4.5 eV, EF ¼ 11:7 eV). The density of internal distri-
bution of electrons by energies was calculated for the temperature of 300°K.
Figures 4.5 and 4.6 represent the densities of external distribution of electrons by
4,0×1024
2,0×1024
0,0
2,0×1024
0,0
6,0×10-19
4,0×10-19
Let us consider the function system p0 ðxÞ; . . .pn ðxÞ. Let us represent the potential
as a linear combination of these functions:
X
n
UðxÞ ¼ ai pi ðxÞ: ð4:12Þ
i¼0
In this case, we are in the space of linear-combination coefficients and the func-
tional minimum
ZEmax Zb
Q 2
JðuÞ ¼ ðDðEÞ dðEÞÞ2 dE þ uðtÞ U þ Et þ dt
t
Emin a
(here u is the control function, t being the time) should be sought in this space:
where a ¼ ða0 ; . . .
an Þ affords a minimum to the functional J(U) given as
For the system of functions p0 ðxÞ; . . .; pn ðxÞ, physical considerations can help to
parcel out a limited subspace of the linear combination coefficients, where we can
find the minimum quality-functional with descent methods. Naturally, there is a
possibility of hitting a local minimum. This problem can be avoided by step-by-step
scanning of all coefficient values on the desired hypercube. If the minimum point
gets on a hypercube facet, its center can be transferred to that point and the scanning
can go on. However cumbersome this procedure might be, it allows for a solution in
the desired class of functions.
Nevertheless, solution of the task at hand needs additional research of possible
classes of functions suitable for potential approximation. Analysis of the research
findings suggests stopping at the polynomial system as most wide-spread. The
potential reduction for rectangular and triangular barriers results in accurate coin-
cidence because it is a question of the zero- and the first-degree polynomials,
respectively, in this case; the barrier penetrability being considered as known
without an error. In case of a triangular barrier with account of image forces, there is
a necessity to carry out additional research of the possibility of barrier approximation
by polynomials from the point of view of the problem of the barrier penetrability.
The results obtained show that the small work function variation method can
successfully be applied to calculation of the transmission coefficient under the
conditions of a real experiment.
180 4 Simulation of Structure and Parameters of Field Emission Cathodes
The mathematical model of the method contains the assumptions that are directly
or indirectly justified by the results of the numerical experiment.
Unlike the wide-spread methods of direct solution of the penetrability problem,
the method does not need any information on the kind of the potential electron
function, which is essential because the latter can be written out for a particular
model only with certain assumptions and restrictions.
The method does not stipulate any space dimensionality or time dependence,
which shows its universality.
Introduction of normally distributed errors into the simulation results showed
tolerance of the method to their occurrence.
In conclusion of the paragraph, let us re-write the inverse penetrability problem
in terms of the optimum-control construction problem.
Let there be a system of
where v denotes the phase coordinates, u is the control function, t being the time.
In our case,
v ¼ Y; u ¼ f ; t ¼ x: ð4:16Þ
And let there be a set of constraints gj ðv; u; tÞ 0. In our case, such restrictions
can be posed by the inequalities of the following form:
It is required to find the control u0 ðxÞ that would afford the minimum to the
following functional:
Zb
JðuÞ ¼ pðvðaÞÞ þ G0 ðv; u; tÞdt; ð4:19Þ
a
i.e. JðuÞ
Jðu0 Þ.
Let us construct such a functional for our problem. We know penetrability d for
each point E. At the desired control, the estimated value follows the formula of
k3 ðEÞ 1
DðEÞ ¼ : ð4:20Þ
k1 ðEÞ jAðEÞj2
4.1 Simulation of a Potential Barrier and Barrier Permeability … 181
Wave vectors are known. Therefore, we have to determine value A through the
finite value of the phase coordinates, i.e.
A expðik1 aÞ þ B expðik1 aÞ ¼ x1 ðaÞ;
ð4:21Þ
ik1 ðA expðik1 aÞ B expðik1 aÞÞ ¼ x2 ðaÞ:
Wherefrom
1 1
A ¼ expðik1 aÞ v1 ðaÞ þ v2 ðaÞ : ð4:22Þ
2 ik1
Consequently,
k1 ðEÞk3 ðEÞ
DðEÞ ¼ 4 : ð4:23Þ
k1 ðEÞv21 ðaÞ þ v22 ðaÞ
2
ZEmax
2
pðxðaÞÞ ¼ kDðEÞ dðEÞk ¼ ðDðEÞ dðEÞÞ2 dE: ð4:24Þ
Emin
There is also a component part of the functional, namely its time integral. For a
more reasonable control, function G0 can be selected as a diversion from the
theoretically predicted shape, i.e.
Q 2
G0 ðx; u; tÞ ¼ uðtÞ U þ Ft þ : ð4:25Þ
t
We have set all the necessary conditions to establish the optimum control of our
system, which will coincide with the required potential. The notation of the func-
tional makes it possible to apply the V.I. Zubov’s method of successive approxi-
mation [17] to this problem.
Derived for the flat geometry of a field emission cathode, the Fowler–Nordheim
equation relates emission current density to the external field intensity. It is obvious
that the natural experiment measures the total emission current and the applied
182 4 Simulation of Structure and Parameters of Field Emission Cathodes
and though the squared voltage V 2 can be taken outside the integral sign, we cannot
do the same with the exponent, which means that from the theoretical point of view,
neither dependence IðVÞ can be obtained in the form of (4.27), nor can we speak
about linearity of (4.29) in the Fowler–Nordheim coordinates. However, it has
already been noted that numerous experimental research of field emission lead to a
linear dependence lgðI=V 2 Þ ¼ f ð1=VÞ.
The generally accepted real-emitter shape approximation models include the
second-order rotation surfaces [24–26, 28–32]. The axial symmetry of such models
saves from extra mathematical difficulties in calculation of the emission charac-
teristics connected with tridimentionality of the problem.
In curvilinear coordinates of a prolate ellipsoid of revolution, the emission
current density can be represented as
2 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi3
AC12 Ba sin g0 ch2 l cos2 g0
j¼ exp4 5: ð4:30Þ
a2 sin2 g0 ch l cos2 g0
2 C1
It is obvious that
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
dI ¼ jdS ¼ 2pa j sin g0 shl sin2 g0 þ sh2 ldl:
2
ð4:31Þ
Then the total electron-emission current from the cathode surface approximated
by a sheet of the two-sheeted hyperboloid of rotation is expressed as
Za pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sin2 g0 þ sh2 l
2pAC12 shl Ba sin g0
I¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi e C1
dl: ð4:32Þ
sin g0
0 sin2 g0 þ sh2 l
Za pffiffiffiffiffiffi2 1 þ s0 pffiffiffiffiffiffiffi
ffi
r2 s2
2pAV 2 1
aB 1s ln
0 1s0 0
I ¼ pffiffiffiffiffiffiffiffiffiffiffiffi2ffi 2 1 þ s pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffie V dr: ð4:33Þ
1 s0 ln 1s00 r2 s20
1
Total emission current IðVÞ was obtained in numbers for all second-order sur-
faces considered in the paper. Voltage-current characteristics approximated by
straight lines were constructed for a hyperboloid, a paraboloid and the ellipsoids of
184 4 Simulation of Structure and Parameters of Field Emission Cathodes
rotation. Linear approximation was carried out by the least-squares method, LSM
[33], the correlation factor falling in the range of 0:99999 R 0:99998
(Fig. 4.8). The calculations for all the surfaces of rotation considered were made on
the parameters typical for a classic field emission diode with r0 ¼ 1 µm, d = 1 cm.
What is more, characteristic curves for a hyperboloid and an oblong ellipsoid of
rotation were constructed on the parameters characteristic for a scanning tunnel
microscope, where r0 ¼ 1 nm, d = 2 nm.
Figure 4.9 shows the voltage-current characteristics of all the surfaces of rotation
considered.
A hyperboloid, a paraboloid and an oblong ellipsoid of rotation are suggested as
approximations of apex emitters, while an oblate ellipsoid of rotation is to
-20
-25
-30
-10 Φ = 4,5 eV
-15
-20
-25
-30
-1 -1
0,0 3,0×10-5 6,0×10-5 9,0×10-5 1,2×10-4 1,5×10-4 V ,V
4.2 Computation of Theoretical Characteristics of a Metal Field … 185
approximate edge-type cathodes (emitters of such a shape are also widely used in
microwave electronics [34]).
The graphs in Figs. 4.8 and 4.9 show that in the analyzed range of voltage values
ð5 103 B V 1 105 BÞ, voltage-current characteristics are linear in the
Fowler–Nordheim coordinates, which 1 was also mentioned in many other experi-
mental papers, e.g. [23], as lg V 2 ; V .
I
But from the theoretical point of view, linearity of expression (4.33) in the
Fowler–Nordheim coordinates can only be proved through further mathematical
studies of this area within the framework of the problem in hand.
The mathematical models suggested can be applied to real apex emitters
(Fig. 4.10) [28]. The approximating model for a particular apex was selected with
the least-squares method, LSM.
The apex shown in Fig. 4.10 was suggested to be approximated with two
models: a sheet of a two-sheet hyperboloid of rotation and a paraboloid of rotation.
Analysis of the resulting curve showed that the curve is best of all approximated by
a hyperbola (Fig. 4.11) expressed by
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðX X0 Þ2
YðXÞ ¼ 1 þ b þ Y0 ; ð4:34Þ
a2
2,9×10-6
2,8×10-6
2,7×10-6
The parameters obtained from mathematical models turned out different from
and determined from the voltage-current characteristic in the Fowler–Nordheim
coordinates by the formulas
I
b ¼ 2:82 109 U3=2 =p; S ¼ 4 1017 p2 ;
V2
V ¼0
1
The mathematical simulation results we have presented show that the values
determined in that way need substantial adjustment. That may turn out essential
because parameters b and S have not only theoretical, but also technological
importance as variable objectives of vacuum micro- and nanoelectronic devices.
Though the technique used to compare the theoretical and experimental results
on the basis of the Fowler–Nordheim formula with insertion of constant Ф and S, is
generally accepted, it can hardly be considered as quite correct. Making such
comparison, one should pass from the experimentally measured values I and V to
characteristics j and E, which is suggested to be done by mathematical simulation of
the real configuration of the field emission cathode whose apex surface consists of
flat areas meeting well determined crystallographic facets.
The images obtained in a field-ion microscope reconstruct the atomic structure of
the emitter surface, i.e. spot the position of each surface atom. Sizing up the
reconstructed model and the ideal three-dimensional surface-structure model
(Fig. 4.12), one can identify adsorbed particles and other defects on the surface and
determine the local emitter-surface curvature radii [35].
There are two approaches in this area, namely the “thin shell model” and the
“broken bonds model” also referred to as the local atomic environment model [40].
Basically, similar approaches can be applied to simulation of a field emission
emitter surface as well; however, nobody has done that yet. It is not improbable that
the reason thereof is that computations of this case become rather complicated
because the sizes of a typical metal electron emitter can be one or two order of
magnitude larger than those of an ionic emitter. Papers [37, 39] present the field
emission image models; however, the real shape of the apex sample is actually
ignored and replaced with approximation by an abstract polyhedron.
This paragraph focuses on the apex-emitter surface modeling procedure within
the framework of the thin shell and broken bonds approaches to field emission. The
simulation is meant to analyze distribution of the atomic-packing density and the
work function over the surface, determination of arrangement and sizes of the
external single-crystal facets and model-emission imaging. The surface character-
istics listed are critical for the processes of field emission, adsorption, heating, etc.
E.g., the surface-facet sizes indicate not only amplification of the exterior electric
field during field emission, but also the surface area contacting with exterior
molecules or adsorbed particles. In its turn, the density and the atomic-packing
structure determine stability and sustainability of the surface, as well as the most
important emission characteristic, namely the electron work function.
The emitter surface structure is determined by arrangement of surface atoms. An
atom is understood as belonging to the surface if it has at least one broken bond
with an any-order neighbor, i.e. if it has an incomplete number of neighboring
atoms:
where Ni is the number of the neighboring atoms of the ith order, Nimax is the
maximum number of the ith order neighbors in the body, M is the maximum order
of the neighbors analyzed.
The surface atom coordinates are constructed in two stages. At the first stage, the
thin-shell approach serves to localize surface atoms, i.e. the atoms lying in a
near-surface layer of a set thickness limited to the geometrical surface determining
the shape of the crystal. At the second stage, the broken-bond approach is used to
identify surface atoms, i.e. the number of the closest neighbors of the near-surface
layer atoms are determined and the surface atoms having at least one broken bond
are identified.
Geometrically, the stage 1 algorithm implies representation of the emitter body
as a set of axial crystalline planes/crystal sections perpendicular to the
surface-rotation axis and determination of the sites lying in the near-surface layer of
each plane. The initial data are the crystal-lattice structure, the surface shape and the
crystal orientation.
In the model involved, the geometrical surface limiting the emitter crystal is
approximated by the sphere-on-the-cone model [53] (Fig. 4.13) and the
body-centered cubic lattice structure is considered. However, the algorithm works
on any axially symmetrical crystal shape, as well as the face-centered cubic lattice
structure. It is obvious that the thickness of the thin shell calculated at the first stage
should be at least the length of the link to the neighbors of the Mth order.
The arrangement of the surface atoms determines the crystallographic facets
opening out into the emitter surface. It is necessary to determine the normal line of
the facet containing each surface atom. The problem is solved on the basis of the
fact that the atoms having an identical geometrical environment, i.e. an identical set
of neighbors, are equidistant from the separating plane of the crystal surface [36].
So, a particular surface atom with radius vector rð1Þ and the number of neighbors
ð1Þ
Ni has to be matched with two nearest surface atoms with radius vector rð2Þ and
rð3Þ having the same set of neighbors, and if these three atoms do not lie on one
190 4 Simulation of Structure and Parameters of Field Emission Cathodes
y, 10-7 m
4
3
2
r
1 θ0 θ
0
a
-1
-2
-3
-4
-10 -8 -6 -4 -2 0 2 x, 10-7 m
Fig. 4.13 The equal potential line of electric field created by charged sphere on the cone (showed
with dashed line). The equal potential line approximates the emitter shape. Symbols are explained
in (4.36): a ¼ 120 109 m, curvature radius of emitter apex r0 ¼ 239 109 m, R ¼ 0:04 m,
n ¼ 0:14 ðh0 177 Þ
straight line, then the plane resting on these atoms is the one determining the
required crystallographic facet and its normal line is calculated by formula
on condition that
8 ð3Þ ð3Þ
< r r \r r þ r r ;
> ð2Þ ð1Þ ð1Þ ð2Þ
where
ð2Þ
r rð1Þ ¼ min r rð1Þ ; ð4:37Þ
ð1Þ
Ni ðrÞ¼Ni i¼1;...;M
ð3Þ
r rð1Þ ¼ min r rð1Þ : ð4:38Þ
ð1Þ
r6¼rð2Þ N i ðrÞ¼Ni i¼1;...;M
Here Ni ðrÞ is the number of the neighboring atoms of the ith order for the atom with
radius vector r.
Though the identical geometrical environment of the atoms means that the atoms
have an identical number of the nearest neighbors up to the arbitrary order; in
practice, the geometrical environment of an atom can be analyzed only up to some
limited order of neighbors set by finite value M (formulas (4.35), (4.37), (4.38)). In
this connection, we computed the emitter facets with account of the nearest
neighbors of only the first order (M = 1), then only the first and the second order
4.3 Simulation of the Metal Field Emission Cathode Surface Structure 191
(M = 2), then etc. up to the 5th order (M = 5). The computations covered
hemisphere-shaped crystals with the radii ranging from 100 to 1000 lattice
parameters and a body-centered cubic structure. For all the radii, the nature of
dependence NNhkl ðMÞ between the relative number of surface atoms of the particular
facet (hkl) and order of the nearest neighbors M taken into account in construction
of the facets was identical, being the number Nhkl of atoms belonging to facet (hkl)
and being the total number of surface atoms N.
Figure 4.14 presents the computation result for the radius of 722 lattice
parameters. The computations show that simulation of hemispheric or nearly
hemispheric crystallographic emitter facets can only be limited to the neighbors of
up to the 5th order inclusive because account of the neighbors of a higher order
makes a correction for less than of atomic number of facets (Fig. 4.14), i.e.
Nhkl ðM þ 1Þ Nhkl ðMÞ
NðM þ 1Þ NðMÞ 0:01; M
5:
25
M=2
M=3 M=5 (111)
M=4
M=5
20
(211)
N(hkl)/N, %
15 (221)
(411) (321)
(521)
10 (721)
(100)
(310) (210) (110)
5
0
110
321
521
721
421
210
631
543
332
732
510
511
331
831
311
832
654
731
111
921
530
751
951
320
443
211
310
100
411
531
431
542
821
432
652
653
632
641
221
532
741
433
621
541
322
643
754
711
522
941
(hkl)
Fig. 4.14 The relative number of surface atoms on emitter facets calculated with taking in account
neighbors of different order. Facet locations in unit stereographic triangle are significant for five
nearest neighbor’s approximation. Crystal has a body centered cube structure, hemispherical shape
and radius of 722 lattice parameter
192 4 Simulation of Structure and Parameters of Field Emission Cathodes
where g is the average facet radius, a is the lattice parameter, r is the local
surface-curvature radius, h, k, l are Miller indices of the crystallographic facet, the
formula being true for the body-centered cubic structure only if (h + k + l) is even.
So the facet measured by the hemispherical-shaped emitter turn out oversized as
compared with the field-evaporated emitter model (Fig. 4.15).
The simulated and experimental emission images can be compared in both
qualitative and quantitative terms. In quantitative terms, we compare the curves of
photometric measurement of emission images for different crystallographic
directions.
The emission characteristics were studied in a conventional way of observation
of field emission from the apex cathode under the conditions of ultrahigh vacuum.
The studies were carried out in an ultrahigh-vacuum cabinet with continuous
pumping-out at the pressure in the region of 10−9 mm Hg. The apex installed on the
manipulator could move rotationally and transitionally. It was positioned at the
distance of 35–40 mm from the anode. The anode was represented by an earthed
luminescent screen with a diameter of 45 mm. The apex emitter was made of a
polycrystalline tungsten wire of 0.1 mm in diameter by electrochemical etching in
an alkaline medium. The experiment involved generation of projection pictures on
the screen and measurement of volt-ampere characteristics, i.e. dependences of the
total emission current on the voltage applied represented in Fowler–Nordheim
coordinates.
Figure 4.16 shows an emission image model constructed as aforesaid. We can
see qualitative correlation between the calculated (Fig. 4.16, upper right) and the
4.3 Simulation of the Metal Field Emission Cathode Surface Structure 193
(a) (b)
(c) (d)
Intensity, relative units Intensity, relative units
60000 70000
50000 60000
50000
40000
40000
30000 30000
20000 20000
10000 10000
0
0
Fig. 4.16 Experimental (a) and model (b) emission images for tungsten emitter, voltage 3885 V,
apex curvature radius r0 = 239 10−9 m; c and d are photometric curves of emission images for
tungsten emitter. Brightness (in relative units) was plotted for dots of (011)–(001) direction (c) and
for dots of (011)–(100) direction (d)
experimental (Fig. 4.16 upper left) emission images of the tungsten emitter. The
images were photometrically measured along the two lines connecting the centers
of facets (011) and (001) (Fig. 4.16, lower left) and the centers of facets (011) and
(100) (Fig. 4.16, lower right).
As the local changes of electric intensity on the emitter surface are only taken
into account on average, by means of integrated parameter c, the maximums of the
photometric curve of the simulated image is somewhat different in value from the
maximums of the experimental curve. The biased position of the maximums of the
simulated curve means that the facets measured by the smooth sphere-on-the-cone
emitter model were oversized.
It is likely that computation of facets by the initial surface structure in a form of
an unsmoothed model of the emitter evaporated by the field can produce more
adequate emission images. Finally, please note that the suggested mathematical
194 4 Simulation of Structure and Parameters of Field Emission Cathodes
model features a surface structure of the apex-shaped field emission metal emitter,
taking into account the arrangement of different crystallographic facets at the
emitter apex.
The emitter shape is approximated by the smooth equipotential surface, and the
crystalline lattice is assumed to be ideal.
In reality, the ideal emitter surface undergoes changes on adsorption, relaxation,
reconstruction and temperature flattening, and account of these phenomena is a
separate problem basically solvable within the framework of this approach [38].
Disregard of those structural details results in oversizing of the facets as compared
with those of the field-evaporated emitter.
On assumption of the ideal surface structure and application of the integrated
parameter for account of the local changes of the field on microscopic inhomo-
geneities of the surface, the model adequately features the data of the natural
experiment.
The broken bonds approach used to simulate the surface structure makes it
possible to apply the model to the case of adsorption on the emitter involving
account of coordination surface-atom numbers [54].
Among the techniques to study the processes taking place on the surface of metals
and semiconductors, field emission research techniques and especially field emis-
sion microscopy are noted for importance of the information delivered and deep
penetration into the micromechanism of the phenomena studied. The reason is a
possibility of visual observation of the course of adsorption, diffusion and migration
at the resolving ability of 20–30 Å (cf. Chap. 1). Moreover, this technique involves
delicate measurements delivering valuable data on the major surface characteristics
along with a possibility of active influence on the subject of enquiry. It is also
important that probation in a field emission microscope can be carried out in
ultrahigh vacuum along with a possibility of attainment and maintenance of atomic
cleanliness of the surface for the entire period of experimental measurements.
However, studying of the surface processes should not be reduced to qualitative
visual observation, especially as far as comparison of the observational results with
theoretical representations is concerned, which is why quantitative studies should
be based on the data of mathematical simulation of the emitting surface. According
to what has been said, the simulation of distribution of the work function over the
emitting surface is of a doubtless interest.
Let us consider an apex of tungsten wire extended along the crystallographic
direction [001]. Then the cathode apex will have the work functions meeting the
reduced Miller indices. Tungsten was selected for the reason of availability of a
great deal of experimental information.
4.4 Simulation of Distribution of the Work Function Over the Field … 195
The crystallographic plane with Miller indices (hi, ki, li) agrees with a particular
value of work function Фi. This information can be treated as a database filled up as
a result of natural [39, 55] or numerical [52] experiments. Surma’s paper [52] is
taken as a basis because of a high correlation factor with the experimental data.
Normal line ni is found for each work function value from the database by formula
ni ¼ s1 1
hkl ðhi ; ki ; li Þ ¼ shkl hi ex þ ki ey þ li ez ; ð4:40Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where shkl ¼ h2i þ ki2 þ l2i . All components of vector ni are one-octant-oriented
(x
0, y
0, z
0). In other words, u 2 [0, p/2].
Let us construct normal line n = (nx, ny, nz) in an arbitrary point of a hyper-
boloidal apex surface (l, u):
Here all components of the normal line are positive, i.e. the cathode apex is
reoriented along axis Oz for convenience. The angle u condition is caused by the
fourth-order symmetry axis in cubic crystals.
So if normal line (4.41) is close to one of normal lines (4.31), e.g. with index i,
the work function value in the (l, u) point can be assumed as equal to Фi.
Let us determine the proximity measure of normal lines (4.40) and (4.41). It can
be a component-by-component comparison or determination of the nearest direction
by the following method. Let us consider the scalar product of
ni ðhi ; ki ; li Þ; n nx ; ny ; nz ¼ cos #;
where # is the angle between the vectors. The smaller the angle between the
vectors, the more cos # tends to one. As a result, the nearest normal line can be
looked for among the minimums of the following expressions:
J1 ðhi ; ki ; li Þ ¼ max jhi nx shkl j; ki ny shkl ; jli nz shkl j ; ð4:42aÞ
2
J2 ðhi ; ki ; li Þ ¼ ðhi nx shkl Þ2 þ ki ny shkl þ ðli nz shkl Þ2 ; ð4:42bÞ
J3 ðhi ; ki ; li Þ ¼ jhi nx shkl j þ ki ny shkl þ jli nz shkl jg; ð4:42cÞ
J4 ðhi ; ki ; li Þ ¼ nx hi þ ny ki þ nz li shkl ; ð4:42dÞ
Equation (4.42a) being the minimax functional; (4.42b) being the congruence;
(4.42c) being the comparison by the sum of the moduli; (4.42d) being angular
proximity.
The functional minimum was searched for by looking through admissible values
in the database. Let us define the acceptable-value concept. As a high-index
196 4 Simulation of Structure and Parameters of Field Emission Cathodes
hi þ ki þ li imax ; ð4:43cÞ
Equation (4.43a) the normal vector ends get into a limiting cube; (4.43b) the normal
vector ends get into the one-eighth of a ball; (4.43c) the normal vector ends get into
a simplex.
The uniform distribution of points over the cathode surface can be used to
determine the relative fraction of different facets in the surface. For this purpose, it
is enough to construct a histogram of distribution of points over the surface with
assigned Miller-index values (Fig. 4.17). The histogram was constructed by M =
1,118,146 points for the algorithm of N = 106. The total number of points M is
proportional to area S of the hyperboloid sector from l = 0 to l = lmax. As a result,
the histogram gives an idea of contribution of Si into total area S of different
crystalline facets. Let us note absence of several low-index facets, which seems to
be caused by the small value of lmax.
The average electron work function can be determined as follows. Let Mi be the
number of occurrences of a point on the surface. The point has been assigned the
work function value of Фi. Then the average value of U is determined as
1X 1X
U¼ U i Si U i Mi :
S i M i
cathode surface
4.
34
4.
45
5 · 10 4
4.
48
4.
4.
91
42
4.
42
4.
54
4.
4.
29
72
4.
43
4.
74
5.
4.
4.
4.
4.
4.
4.
64
44
41
44
70
68
62
0
12
13
11
11
11
12
01
02
13
22
23
33
00
01
01
01
11
11
12
4
1
2
2
3
3
4
1
1
3
6
Face
4.4 Simulation of Distribution of the Work Function Over the Field … 197
Sequence li (i = 0.1…, Nl) breaks radius rmax into the segments of identical length
l. At that,
il
ri ¼ a0 shli ; li ¼ Arsh :
a0
Nu;i ¼ ½2pi:
198 4 Simulation of Structure and Parameters of Field Emission Cathodes
50
50
100
100 50 0 50 100
In this simulation, angle #rnd assumes the desired values in interval [0, p/2]. If r =
0, the normal line will not vary. The larger value r, the more probable is deflection
of the normal line by an essential angle. The approach suggested results in a bundle
of vectors distributed round axis Oz (Fig. 4.21). Simulation of random variable n
involved the following algorithm:
qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi
2 ln R2 2 ln R2
n1 ¼ t 1 R2 ; n2 ¼ t 2 R2 ;
t1 ¼ 2c1 1; t2 ¼ 2c2 1; R2 ¼ t21 þ t22 ;
if R2 > 1, the pair of n1, n2 does not suit us. This approach gives two realisations of
value n by two realisations of random variable c. Simulation of uniformly dis-
tributed value c involved the pseudorandom-number generator Mersenne Twister
MT19937 (period 219937 – 1 106001, uniformity in ℝ623).
The random-component simulation algorithm performance in the normal line is
presented in Fig. 4.21. The normality of distribution of random variable n causes
large-angle deflections to occur much less often and depend on parameter r. Within
the framework of simulation of surface inhomogeneities, that may indicate absence
of sharp turns.
To set the necessary direction corresponding to normal line (nx, ny, nz), the
random vector is turned by normal angles uп and #n:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
n2x þ n2y
un ¼ u; #n ¼ arctg
nz
by matrix transformations
0 1 0 1
1 0 0 cos un sin un 0
T#n ¼ @0 cos #n sin #n A; T un ¼ @ sin un cos un 0 A:
0 sin #n cos #n 0 0 1
Finally, the simulated normal line with a random component is then of the fol-
lowing form:
When a random component is introduced into the normal vector, the emission
pattern blurs because of the noise and visually resembles the images developed as a
result of a real experiment.
The physical processes affecting the field emission current density limits were
thoroughly analyzed in [16, 56, 57]. One of the basic causes limiting the field
emission current density is heating of the field emission cathode with this current.
The spatial source of energy-release in the emitter [27, 56] under field emission
current flow is well-studied Joule heat. Nottingham was the first to identify one
more source of heat release [58]. The phenomenon he noticed is caused by dif-
ference of medium energies of emitted electrons from those replacing them, which
leads to release or absorption of field emission heat. The Nottingham effect at low
emitter temperatures leads to its heating because most of the electrons tunnel from
the energy levels lying below the Fermi level. At rather high temperatures, the
medium energy of the emitted electrons exceeds the energy of the Fermi level, i.e.
the Nottingham effect becomes a cooling factor. In this respect, it is similar to the
so-called calorimetric effect [59] accompanying thermionic emission.
The first attempt to compute the field-emitter temperature mode at limiting field
emission currents was made by Dyke et al. [60]. At that, a real field emission
cathode shaped like a truncated cone with a spherically rounded apex was replaced
with a part of a cone limited with a spherical surface of radius rm orthogonal to the
elements of the cone (Fig. 4.22). Section r = rm was selected for its ability to
provide the maximum current density. The distance from this section to the emitter
apex was approximately equal to the section radius. Thermal constants were rep-
resented by intermediate values corresponding to some medium temperature of the
cathode.
It was also assumed that the temperature at the cone base is always equal to the
initial, preliminarily set temperature T0, and the length of the conical part of the
emitter rl rj, which is only fair for an emitter with small a.
Though paper [60] did not take thermal emission caused by the Nottingham
effect into account aggravated by inadequate emitter geometry, the arbitrary choice
of specific resistance q and specific heat capacity C brought theoretical variables
quite close to experimental values of stationary field emission current density limits.
202 4 Simulation of Structure and Parameters of Field Emission Cathodes
rm
rl
2,0
3
8
9×10
1,5
5
1,0
1,0
4 A
3 8
2 1×10
1 2 B
1
1,2 2,4
5
t ×10 , s 3 6 t, s 10 50 90 α°
Fig. 4.23 The calculated dependences of the emitter temperature on the current flow time at
different densities jf-emis = const (a) and for different a (b): a jFEE, A/cm2 (1 1.13 108, 2 1.42
108, 3 1.55 108, 4 1.70 108, 5 1.85 108, 6 2.00 108, 7 2.15 108); b a (1 90°, 2 70°, 3
50°, 4 30°, 5 10°); c theoretical (o) and experimental dependences jcrit = f(a)
E ≅ 108 V/cm. Accordingly, there are large, close to the maximum permissible,
tensile stresses caused by ponderomotive forces. Moreover, huge current densities
cause heating of the emitter to high temperatures and emergence of large temper-
ature gradients up to *107 108 K/cm, which essentially changes the nature of the
diffusion processes both in the cathode body and on its surface.
All these factors generate a number of effects observed in experiments, such as
growth of the field emission current with time at the squared voltage impulse and
generation of a bright ring bordering the field emission image (cf. [56] as an
example).
Development of a bright ring on the periphery of an emission image of a metal
field emission cathode (Fig. 4.24) is known to predict its explosive rupture or
transition of field emission into explosive one, which is the same. That is the reason
why the “ring effect” research, going on for several decades, still attracts attention
of the experts interested in conditions of stable operation of a field emission cathode
when limiting emission currents are taken off from it.
W. Dyke et al. were the pioneers of deliberate analysis of the ring effect in their
studies of operating conditions of metal field emission cathodes with removed
currents of extreme densities. Thus, one of the first explanations of generation of a
ring was suggested, assuming movement of the temperature maximum towards the
periphery of the emitting surface of the cathode because of inversion of the
Nottingham effect in this operating mode.
In 1964, Sokolskaia et al. [65] published an article where the ring generation was
attributed to transition of a pure field emission into thermal field one. Unlike Dyke
et al., the authors of publication [65] are notable for registration of ring generation
in the stationary field emission mode, having provided the thermal field emission
204 4 Simulation of Structure and Parameters of Field Emission Cathodes
mode with exterior heating of the cathode. Finally, the recent research attempted to
substantiate the ring effect by wave and diffraction mechanisms [66].
It is obvious that correct interpretation of field emission pre-explosive effects,
including the ring effect, can be based only on consistent concepts disclosing the
physical mechanism of these effects. However, the above-said makes it clear that
there is no single opinion in this respect even when the researchers analyzing the
experimental facts and developing the hypotheses deal with the same experimental
results.
In our opinion, papers [67, 68] suggested a physical analogue and a mathe-
matical model accompanying these phenomenon effects and consistently explaining
the ring effect on the basis of the experimental data on pre-explosive effects.
When constructing the physical analogue, authors [67, 68] proceeded from the
quite definitely established experimental fact that in the pre-explosive operating
mode of the field emission cathode, its surface is heated up by through emission
current to the temperatures close to the surface-melting point, this temperature
being usually 2/3 of the material bulk-melting temperature, which results in a
liquid-metal film on the metal surface (Fig. 4.25a). After development of the film
according to our model, the electric field applied between the emitter and the anode,
i.e. extractor, develops the so-called Taylor cone [68] and draws a nanotip with the
curvature radius of about 10−9 m on the apex from the liquid metal (Fig. 4.25b),
which explodes in the field enhanced by it up to *1010 V/m, thus generating anode
plasma near the emitter apex, which is a source of both electrons and positive ions.
The resulting ions are focused on the cathode apex. They press on the liquid metal
and push it to the periphery of the emitting surface, thus forming a ledge on it
(Fig. 4.25c). This ledge increases electric intensity near its apex, thus amplifying
emission from this area, which cannot but manifest itself in the emission image as a
bright ring.
4.5 Theoretical Studies of the Physical Processes Affecting … 205
Fig. 4.25 Model of emission surface evolution for metal field emission cathode that explains the
development of a ring on the emission image: a apex with a liquid-metal film; b development of
the Taylor cone and nanotip on the apex; c forming a ledge on emission surface; point O area of
the best ionization
As mentioned above, the torus section of the ledge is approximated by the Gauss
curve, taking into account that S = const, where S is the area outlined by the Gauss
curve, which can be written down in the following way:
pffiffiffi
y ¼ A a expðax2 Þ; ð4:44Þ
where A is a stationary value depending on the quantity of the liquid metal; a is the
parameter shown further to depend on the value of the voltage applied to the
emitter. As R r, R being the torus radius, we believe that there is only one
curvature radius r, namely the curvature radius of the Gauss curve
pffiffiffi
½1 þ ð2Aa ax expðax2 ÞÞ2 3=2
r¼ pffiffiffi :
2Aa að1 2ax2 Þ expðax2 Þ
1
r0 ¼ rjx¼0 ¼ pffiffiffi : ð4:45Þ
2Aa a
The surface of the liquid metal has to meet the following condition of balance for
pressure forces:
e0 E 2 Y I 0 pa
¼ þ ; ð4:46Þ
2 y eXR2
where E is the electric intensity on the surface of the liquid ledge; Y is the surface
tension; I0 is the field-ionization ion current bombarding the emitter; pa is the
impulse of the ions bombarding the emitter; X is the solid angle of the ionic beam;
R is the torus or crater radius.
In the area of field ionization, the electric intensity should be constant and equal to
Vi U0
E0 ¼ ffi ;
d R
where U0 is the voltage initiating the pre-explosive field emission from the ledge
area; R is the torus or crater radius approximately equal to the virtual-emitter radius.
For the voltage between the anode and the cathode we will assume the following
expression true:
U0
U¼ h; ð4:47Þ
R
where h is the distance from the emitter to the field-ionization area, i.e. the point
O in Fig. 4.25c. The energy-conservation law and ratio (4.47) give us the impulse of
the ions bombarding the emitter:
4.5 Theoretical Studies of the Physical Processes Affecting … 207
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pa ¼ 2eU0 mðn 1Þ;
U ð4:48Þ
n¼ :
U0
or
!
n
X ¼ 2p 1 pffiffiffiffiffiffiffiffiffiffiffiffiffi
n2 þ 1
E1 r1
Effi þ E2 ; ð4:49Þ
r
Here, E1, r1 are the electric intensity and the curvature radius pf the ledge apex at
initial moment of pre-explosive emission, i.e. emergence of the liquid ledge,
respectively; r is the resulting curvature radius of the ledge; E2 = E0n is the electric
intensity at the bottom of the ledge.
With stepping up of the voltage between the cathode and the anode, the
ledge-curvature apex radius becomes smaller, and the electric intensity at the ledge
apex builds up according to (4.49), which leads to current growth according to the
Fowler–Nordheim formula:
109
j ¼ 107 ðE1 r1 KðnÞ þ E0 nÞ2 exp :
E1 r1 KðnÞ þ E0 n
Assuming that stepping up the voltage does not change the area of emission, the
emission current from the ledge apex can be expressed as
I ¼ jðnÞpRr0 :
taking place in the pre-explosive operating mode of the field emission cathode
and, consequently, on the evolution of the emitting surface in superstrong fields.
2. Despite the estimate nature of the computations made, the resulting theoretical
volt-ampere characteristic constructed in Fowler–Nordheim coordinates is
qualitatively similar to those experimentally observed for the area of superstrong
fields and extremely heavy currents, in particular the theoretical curve, as well as
the experimental one, deviate towards smaller currents.
If the physical analogue suggested is true, the experiment can be expected to
generate a ring on the emission-image periphery under take-off of a considerably
smaller emission currents; firstly, with increase in the pressure of residual gases
(when the vacuum of the experimental device softens); and secondly, during
application of a film of a fusible metal, such as Au, In, Ga etc., on the field emission
cathode made of high-melting metal, such as W, Mo, Ta etc.
Almost direct confirmation of the mechanism suggested was really obtained [56]
by coating of the emitter made of a high-melting material, usually tungsten, with a
layer of a substance with a significantly lower melting temperature. Then emer-
gence of rings caused the layer to change into a liquid state. At that, the initial
temperature of the tungsten apex can be varied.
Let the pulse mode and initial temperature T0 of the tungsten-apex vertex coated
with a layer of a lower-melting material generate the ring at emission current I, and
temperature T00 does the same at current I′. Let us believe that the ring generation
causes the coating to change from a solid to disordered phase due to additional
Joule heating by the flowing emission current. Then disordered-phase temperature
Tliq, and, actually, the thin surface-film melting temperature can be determined as
ðI 0 =IÞ2 T0 T00
Tliq ¼ : ð4:50Þ
ðI 0 =IÞ2 1
For particular computations, let us first take advantage of all data of papers [56,
66]. Their authors saw the ring on the barium- or thorium-coated tungsten apex. The
pulse mode and the apex temperature of T0 = 1000 K generate the ring in the
emission image at current I ≅ 650 µA. Termination of the heating or reduction of
the current make the ring disappear. Without heating, i.e. at the initial temperature
of T0 = 300 K, the ring re-appeared at I ≅ 1300 µA, the coating being thorium.
Insertion of numerical values I, I′, T0 and T00 into expression (4.50) results in Tliq ≅
1233 K. According to [66], the film-melting temperature Tfilm is (2/3) Tbulk, where
Tbulk is the bulk-material melting temperature. For thorium, Tbulk = 2023 K, so Tfilm
≅ 1350 K. Taking into account the inexactitude of values I and I′, coincidence of
Tliq and Tfilm can be considered quite good. Iteration of the experiments in a more
careful way showed the value of the current ratio being I/I′ = 1.78. So Tliq = 1320 K,
which is about 98% coincidence with Tfilm. It is clear that if the initial temperature is
lowered e.g. by 200–300 K, in any ring-generation mode, then the surface layer
cannot change into the disordered phase, and the rings in the emission image have
to disappear, which is really the case.
4.5 Theoretical Studies of the Physical Processes Affecting … 209
An obvious way to reduce the self-heating effect of the field emission cathode is
its cooling to ultra-low temperatures, e.g. that of liquid helium.
There is a comparatively inconsiderable number of experimental papers on field
emission of metals at low temperatures that need theoretical substantiation or
evaluation. The research was mostly focused on field emission from the super-
conductive (s) state and detection of the energy gap in transition into the normal
(n) state [69]. According to paper [70], transition from the superconductive into
normal state connected with the change of energy of about kTs, Ts being the
superconductor state transition temperature, could manifest itself in changes of the
field emission current. Dependence of field emission current on voltage was studied
at helium temperatures ranging within 4.4–2.2 K.
It was shown there was no difference in field emission current for a tantalic
emitter of the temperature both above and below Ts within the experimental
accuracy estimated as nearly 0.2%. The Fowler–Nordheim dependence looked in a
way usual for that current density. The tantalic apex of 0.57 µm in radius developed
current density jfe = 80 A/cm2. That was whence we concluded that the potential
barrier on the surface of a superconducting metal and the chemical potential of
conduction electrons do not vary up to the values exceeding kTs.
Measurement of field emission current increment for a niobium emitter in a
superconductive (at 4.2 K) or normal (at temperatures above Ts) state was attempted
in paper [69]. The current increase caused by a rise in temperature having been
excluded, the current increase connected with s–п transition turned out less than the
measurement error. Therefore it was suggested that field emission might make the
field penetrate inside the metal by about 1 Å, which terminates the superconductive
state on the surface.
Septiers and Bergere [71] underscored the importance of the Nottingham effect
in thermal phenomena at low temperatures. A superconducting apex was shown to
change to a normal state with enhancement of energy generated due to the
Nottingham effect at rather small emission current, as small as 90 µA.
Alexeevskiy [72] studied the influence of emission current on the critical
parameters of a superconductor. A Ta-wire cathode was energized with current Js,
after which voltage drop V on the wire was measured. Voltage U was applied
between the cathode and the niobium-disk anode and emission current Je was
running. At Je = 0, the cathode was in a superconductive state and V = 0.
Achievement of some emission current was seen to carry the cathode into a normal
state, value V becoming finite. Further build-up of U increased V up to the maxi-
mum value Vn, leading to the entire cathode being in a normal state. Dependence of
ratio V/Vn on value U at different temperatures is shown in Fig. 4.26. N.E.
Alexeevskiy attributed emergence of the normal-phase area both to local heating
and some other possible reasons. In particular, the latter include emission-caused
reduction of the number of carriers taking part in superconducting current or the
mechanical voltages reducing the critical current.
So, research of field emission from the cathodes that initially had the temperature
below critical in stable experimental environment detected no specificity in
voltage-current characteristics or emission images suggesting that field emission
210 4 Simulation of Structure and Parameters of Field Emission Cathodes
0,5
0
4,5 5,0 5,5 6,0 6,5 7,0 V, kV
current takeoff is provided by the apex emitter in the superconductive state. At the
same time, no data on field emission current density limits at low temperatures is
available in the literature.
As already noted, the effect of increase in j0 at low temperatures is connected
with expansion of the Nottingham heat-release area referred to as “smearing”. At
that, it is necessary to differentiate between the essence of the Nottingham effect and
the mechanism of its implementation. The essence of this effect is known to be
connected with a difference between the average energy of the emitted electrons and
the average energy of the electrons coming from the external circuit, which can
result in both cooling and heating of the field emitter as the case may be. In this
context, let us consider the Nottingham heat-release mechanism coming down to
the following. Field emission injects a hot hole into the cathode (Fig. 4.27a). The
hole scatters the electrons that pass to the states lying above the Fermi level
(Fig. 4.27b). The electrons thermolize over the period see, which is a typical time of
electron-electron interaction. It is worth noting that the excess energy has not been
transmitted to the lattice yet, i.e. the Nottingham effect as a whole has not been
implemented by the time of thermolization yet. To transmit the excess energy to the
lattice, the electrons should interact with phonons, which is really the case as a
result of electron-phonon interaction over the characteristic period of sef and the
characteristic length of kef. Electrons are thermolized only if the condition of see
(a) (b)
λef
− + +
sef is true. At low temperatures, the area where energy is transmitted to the lattice
(kef * tФsef, tФ being the velocity of electrons on the Fermi surface) can be as thick
as a few parts per millimeter in pure metals. It also suggests that the Nottingham
heat release is voluminous under these conditions (delocalization of the Nottingham
effect).
Total theoretical analysis of such a complicated situation can only be carried out
within the framework of the kinetic theory. But assuming that the condition of see
sef has been met, metal can be considered as a set of two subsystems, an electron
and a lattice one.
The energy-balance equations for these subsystems looks as follows:
@Te
Ce ¼ rðje rTe Þ AðTe ; Ti Þ þ jf ðrÞ; ð4:51Þ
@t
@Ti
Ci ¼ rðji rTi Þ þ AðTe ; Ti Þ: ð4:52Þ
@t
The Te, Ce, je of these equations are the temperature of electrons, the electron
heat capacity and the thermal conduction, respectively; Ti, Ci, ji are the respective
values for the lattice subsystem; A(Te, Ti) is the function featuring the nonlocal
transmission of energy from the electron subsystem to the lattice. The last term of
(4.51) takes account of Nottingham heat-release in the electron subsystem over
length kee.
Despite some incorrectness of such an approach, appropriate selection of func-
tions A and f can be expected to give adequate representation of the basic features of
the phenomenon of “smearing” of electron-electron interaction over length kee and
electron-lattice interaction over length kef. Functions A(Te, Ti) and f(r) were selected
as
Z
j r xj 3
AðTe ; Ti Þ ¼ a U½Te ðxÞ; Ti ðrÞ exp d x; ð4:53Þ
kef
l
f ðrÞ ¼ b exp ; ð4:54Þ
kee
Ti ðr ! 1Þ ¼ Te ðr ! 1Þ ¼ T0 ; ð4:55Þ
212 4 Simulation of Structure and Parameters of Field Emission Cathodes
@Ti @Te
¼ ¼ 0; f jr ¼ jðTe ; EÞ ð4:56Þ
@n r @n r
Ti ðt ¼ 0Þ ¼ Te ðt ¼ 0Þ ¼ T0 ; ð4:57Þ
Dj
T e Ti ¼ expðx=kee Þ;
mkee
Dj
Te Ti ¼ expðx=dÞ:
md
These ratios show that if kee d, the characteristic electron-gas heating area is
*d. If kee d, the heating area is dramatically expended, and the electron-gas
temperature is much lower than in the first case. Consequently, the lattice-heating
velocity plummets:
Dj t
Ti T0 : ð4:59Þ
Ci kee þ d
8
10
0 4 8 t, ns
cannot considerably exceed rj/a0, rj being the radius of the emitting surface, a0
being the angle of the emitter cone.
For T0 = 300 K, jfiel-em does not continuously increase over the impulse of 10 ns,
which seems to be caused by heating of the emitter (Fig. 4.28). jfiel-em begins to
grow at the end of the impulse for initial temperature T0 = 4.2 K, which can also be
connected with the fact that the cathode achieved a certain temperature that seems
to be not very much different from 300 K. That is basis for Graph 1 in Fig. 4.28 to
evaluate the ratio of su =su as about 13.6% for sи = 10 ns, which matches the
experimental results of su =su = 13.3%. Once again, this circumstance proves
adequacy of the model suggested.
It is worth noting that the solution of a purely thermal problem shows that for
tungsten the current density limit cannot be increased more than 1.45 times without
account of size effects due to cooling of the emitter.
Increase in current density limit strongly depends on selection of a right emitter
material. The cathode material usually includes a basic chemical element and some
additives, as well as some imperfections/defects of the crystal lattice. At ultra-low
temperatures, it becomes crucially important that electrons are scattered by crystal
lattice defects, namely isolated pin-hole defects, such as vacancies, interstitial
atoms, chemical additives, isotopes etc.
In case of a single-crystal field emitter, the major factor influencing the free-path
length kef of electrons in the cathode at helium temperatures are isolated pin-hole
defects. At that, calculation shows that
cathode to destroy the cathode upon reaching a certain value, originated not near the
surface of the cathode, but in the depth of its bulk, the phenomenon being referred
to as smearing of the Nottingham effect. This condition is met if the free-path length
of electrons in the cathode at the working temperature is 10 times longer than the
cathode radius. Actually, this requirement is fair for apex cathodes with very small
cone-vertex angles. Apex cathodes with arbitrary cone-vertex angles should meet a
softer constraint of
rj
kef
; ð4:61Þ
a0
Special properties of field emission in presence of strong electric fields and high
current densities were relatively fully analyzed in the introductory chapter as well as
in book by Zhukov [73] and an overview by Fursey (cf. [56], Chap. 6, Sect. 6.4)
and possible influence of simultaneous influence of spatial charge and proper
magnetic field or spatial charge and external magnetic field on field emission
characteristics was assessed by Egorov theoretically [74] and experimentally [75] as
well as addressed in [17, 76, 77].
The beginning of experimental research of field emission in presence of strong
electric fields and high current densities are connected with work of Dyke et al.
[44–60, 78–80]. In presence of those fields Dyke and Trolan [44] discovered sig-
nificant inconsistencies with the Fowler–Nordheim law in form of much weaker
than predicted dependence of emission current on applied potential difference
(Fig. 4.29).
Dyke and Trolan [44] interpreted the violation of linearity of volt-ampere
characteristics as an effect due to influence of spatial charge of emitted electrons
that weakens the field strength at the cathode surface.
The first attempt to theoretically assess the possible influence of spatial charge
onto field emission was made by Stern, Gossling and Fowler [81]. Calculation of
the theoretical relation lg(j) = f(V) (where j is the current density and V is voltage) is
4.6 Special Properties of Field Emission in Presence of Strong … 215
A
-16
-20 4 -1
1.0 2.0 3.0 10 /V, V
integrating (4.63) for boundary conditions (4.66) the authors of [81] obtained the
relation between field strength, current density and potential:
h i
E ¼ ðUa =d Þ 1 16=3pðme =2eÞ1=2 jUa1=2 =E 2 : ð4:67Þ
216 4 Simulation of Structure and Parameters of Field Emission Cathodes
Without taking into account the spatial charge E = Ua/d, were Ua is equal to voltage
on the anode, criterion of smallness of influence of this charge according to (6.21)
can be written as
In more detail this process is analyzed in [82]. Same as before, the authors based
their conclusions on (4.63). Validity of chosen one-dimensional approximation of
real tip-cathode diode configuration is justified by the fact that the spatial charge
influencing the field close to emitting surface can be viewed as concentrated at a
very small distance from the cathode. Solving (4.67) together with Fowler–
Nordheim equation,
j ¼ AE 2 expðB=EÞ ð4:69Þ
Together with (4.69) this equation allows to calculate the current density for several
fixed values of E depending on the applied voltage. With small E, (4.70) leads to a
normal interdependence of field and potential for zero current (E = Ua/d). With very
big E the exponential parts become dominant and it’s easy to see that (4.70) is thus
reduced to (4.65).
The parameter d in [82] was chosen as equal to 1/b, where b is the geometrical
factor connecting the field strength with voltage E = bV.
0
A
-2
0 2 4 6
108/βV, cm/V
4.6 Special Properties of Field Emission in Presence of Strong … 217
r2 U ¼ 4pqe ; ð4:71Þ
D~j ¼ 0 ð4:73Þ
where rk is the cathode radius and ra is the anode radius; DUm same as (4.70) is an
error taking in account the influence of spatial charge on field emission, but this
time in a relativist conditions and η eErk/mec2.
Equation (4.74) same as (4.70) is obtained assuming that ra rk and its
applicability is limited with condition
where U0 is the potential of spherical anode without accounting for spatial charge, it
is connected with field strength E by relation
In accordance with (4.69) and (4.74–4.76), for each defined value of E there is a
defined value of current density je and a defined value of anode voltage Va. Using
this relations one can construct non linear volt-ampere characteristics of spherical
diode in coordinates lg(j)(1/Va). Same as before, the approximate character of the
solution limits the construction to the initial part of disconnection from Fowler–
Nordheim line.
Assuming the value of current that differs twofold from the one calculated by
Fowler–Nordheim formula (without accounting for spatial charge and relativist
effects) with the same applied voltage as the criterion of start of disconnection, one
can obtain explicit definition of disconnection current density from (4.74–4.76).
" #
ln 2E 3=2 ð2e=me Þ1=2 1 ra
ln jd ¼ ln ln rk ln ln 4 1 : ð4:77Þ
4pð2 þ Bu3=2 Þ=E 2 rk
10 4
10-4 10-3 rc, cm
4.6 Special Properties of Field Emission in Presence of Strong … 219
computations show that the influence of relativist effects is necessary to take into
account with emitter radius greater that 10−3 cm. It is interesting to note that
whenever a large amount of emission current flows through a spherical diode
from which follows that the amount of flowing current is not dependent on the
cathode material work function. Field emission is fully limited with the spatial
charge.
Volt-ampere characteristics calculated in [71] are presented on Fig. 4.32. As one
can see from the figure there is a rather large area of the volt-ampere characteristics
where growth of the current with increase of the voltage is considerably slower than
on the initial linear area (increase of characteristic slope with large voltages is
connected to the scale effects).
The slowing of the current growth is especially noticeable with rk
10−3 cm
and current densities je > 107 A/cm2. Dampening of emission by spatial charge is so
large in these conditions, that the maximum current densities the given field
emission cathode can potentially provide with small influence of spatial charge
10 6
10 5
10 4
0 0.5 1 1.5 108/βV, cm/V
220 4 Simulation of Structure and Parameters of Field Emission Cathodes
become difficult to achieve (108–109 A/cm2). These current densities were exper-
imentally observed in impulse modes on thin tip cathodes (with rk = 10−4…10−5
cm). For example, for rk = 10−2 cm, in order to obtain je = 108 A/cm2 one has to
increase the voltage by more than an order of magnitude as compared to the value
of voltage computed without taking into account the spatial charge.
As the relation between the magnetic component of the Lorentz force and forces
of electrostatic repulsion of electrons in the beam is of the order v2/c2 and velocities
of electrons with megavolt-scale voltages are comparable to the speed of light,
accounting for the magnetic field can lead to grave corrections to the calculations of
volt-ampere characteristics of real emitter.
U=U0
U=0
r1 H
0
z
E
cathode anode
y
4.7 Modeling of Elements of Beam Formation and Control Systems … 221
thus solving this problem can be reduced to solving the equation for potential. In
order to account for the influence of spatial charge in given model case during field
analysis, the Poisson equation in variables of spherical coordinate system is used.
Given the symmetry of the problem, it can be written down as:
1 @U 2 @U qðrÞ
r ¼ ð4:80Þ
r 2 @r @r e0
qðrÞ ¼ jðrÞ=v;
where F = −e{∇U(r) + [v, H]}; e, m are the charge and rest mass of the electron;
v is its velocity; H is the vector sum of the external and proper magnetic field
strengths of the beam. U(r) is the potential of electric field; c is the speed of light.
After differentiating of the left hand part of (4.82) and several elementary trans-
formations, it can be rewritten as:
@2r
þ AðtÞ ¼ cf; ð4:83Þ
@t2
where
rffiffiffiffiffiffiffiffiffiffiffiffiffi
@r v2 e v2 @v
f ¼ b rU þ ;H ; c ¼ 1 2; b¼ ; AðtÞ ¼ v:
@t c m c2 c2 @t
ð4:84Þ
When calculating the proper magnetic field of a particle beam only the azimuthal
component was taken into account. When finding trajectories of electron motion in
222 4 Simulation of Structure and Parameters of Field Emission Cathodes
the nodes of mesh enveloping the area through which the beam passes, the spatial
distribution of currents Iij is memorized in it. Based on the assumption that the
current density is constant for each ij-th elementary cell, the formulas for proper
magnetic fields were constructed:
8 lI
>
< 2pð0;5hr Þ2 rT ;
0 1j
rT 0; 5hr1 ;
1
H0 ¼ l I1j r 2 ri1=2
>
: 2pr
0
Ii1;j þ ðIij Ii1;j Þ r2 T r 2 ; rT
0; 5hr1 ;
T i þ 1=2 i1=2
where ri ± 1/2 = 0.5(ri + ri ± 1); hr1 is the step of mesh close to the axis of symmetry;
Iij is the total current, i.e. the sum of all the currents
of all trajectories passing
through the j-th section of DSij ¼ p ri2þ 1=2 ri1=2
2
(close to the axis of symmetry
r 2
DSij ¼ 0:25p h1 ). In the initial moment the coordinates and velocities of a
boundary electron are defined:
Ujr¼r1 ¼ 0; Ujr¼r2 ¼ U0 ;
where U0 is the initial distribution of potential as found at the first step. For q(z) the
values found at the second step are used.
At the fourth step, same as at the second, we yet again find the family of
trajectories by solving the system of motion equations. This time, however, we use
U(r) (and thus also E(r)) found at the third step in the motion equations.
4.7 Modeling of Elements of Beam Formation and Control Systems … 223
Repeating those steps enough times one can compute the potential distribution U
(r) with any arbitrary precision. Using the known U(r) one can find the electric field
strength E0 at the cathode surface and thus also the current density j0 of the field
electron emission.
Solving the self-consistent problem of integrating of a system of differential
equations (steps 2 and 4) was performed using the Runge-Kutta methods as noted
above. It was assumed that the initial velocities of electrons were equal to the
thermal velocities. The step of integration for each step was connected to the step of
corresponding area of the mesh, and the origin of spherical coordinate system was
positioned at the emitter plane. The process of constructing the trajectory assumes
that for each next trajectory the place of the “zeroth” one in the computer memory is
occupied by the first trajectory, which leads to eventual growth of volume. Thus we
were able to avoid introduction of additional working arrays for storing the infor-
mation on the trajectories being constructed.
Results of computation and numerical experiment. Numerical experiment was
performed for both v(y) = 1 (i.e. tunneling of electrons through the mode triangular
barrier) and for v(y) 6¼ 1 (i.e. the barrier of “imaging forces”).
No qualitative differences between the cases v(y) 6¼ 1 and v(y) 1 were
observed: in both cases there were “pulsating” relationships lg j = f (1/E) (cf.
Fig. 4.34). If in the “limit” case the magnetic field strength was assumed zero, the
“pulsation” disappeared. In the same time there were noticeable quantitative dif-
ferences: for v(y) 6¼ 1 the pulsation amplitudes were significantly larger than for v
(y) 1.
3
1 2 E, 102V/cm
224 4 Simulation of Structure and Parameters of Field Emission Cathodes
This is probably connected to the fact that while considering the values of
Nordheim function not equal to 1, the influence of proper spatial charges of elec-
trons (and thus also the proper magnetic field) in the vacuum cathode-anode gap
begin to manifest themselves at smaller values of emission-exciting electric fields.
In the same time, resulting influences of proper and external magnetic fields onto
the spatial charge distribution for the case of tunneling of the electrons through the
barrier of “imaging forces” must be more efficient.
Analyzis of the obtained results has shown that apart from that, the trajectory
slope and the size of “anode spot” vary dependent on the relation of E and
H. Character of non-monotonousness (e.g. the amplitude of pulsation of current
density) is significantly dependent on the angle h0, i.e. from the side of emitter
surface (or the emission area). There is a certain minimum threshold value of h0
(let’s call it “critical zone”), surpassing which leads to complete disappearance of
magnetic effects: h0 7–9°.
One also has to note that the results of numerical experiment are also dependent
on the chosen numerical method, its precision etc. For example, when integrating
the initial system of equation, due to closeness of all the characteristics of the beam,
it is very easy to cross the line where there will be variance of numerical solutions
from iteration to iteration rather than any “physical” pulsation effect. The other
possible reason of observed pulsations can be instability of the iterated process due
to monoenergeticity of the inbound beam which leads to emergence of defects of
q(r) at the area of virtual cathode. As such, even small fluctuations in this functions
cause large-scale changes in the density of electron flow, i.e. the problem becomes
unstable (i.e. incorrect by Tikhonov: the third condition of correctness in [88] is
broken). Thus all the results of mathematical modeling and numerical experiments
always need experimental confirmation. It is even more important given that in a
mathematical model it is all but impossible to consider all the different phenomena
that take place in real devices based on charged particle beams.
Experimental studies of influence of magnetic fields on field emission from
metals were undertaken in [75–77] and analyzed in great detail in an appendix to
[17]. For an instance, authors of [76] studying the influence of magnetic field on
field emission, claimed based on their results that they were able to observe
multi-electron field emission. However further research showed that “pure” field
emission from metals is always single-electron, i.e. no more than one electron can
be tunneled in one elementary act of field emission, and the multi-electron phe-
nomena observed in [76] are connected to the secondary effects such as superim-
posing of secondary electron and ion-electron emission onto field emission [75].
However absence of real-life effects (such as “pulsation” of field emission cur-
rent) in experimental observations of influence of magnetic field on the emission
process probably has to do with the fact that the magnetic field intensity in those
experiments did not exceed a few tenths of Tesla or sometimes a few Tesla. One
can’t overlook the possibility of emergence of the pulsation effects predicted in [74]
if stronger magnetic fields are to be applied.
References 225
References
26. A. Seidl, M. Takai, A. Hosono, S. Yura, S. Okuda, Geometry effects arising from anodization
of field emitters. J. Vac. Sci. Technol. B. 18(2), 929–932 (2000)
27. M.I. Elinson, G.F. Vasiliev. Field emission. M.: Fizmatgiz (1958), 272 p
28. N.V. Egorov, V.M. Zhukov, C.A. Polezhaev, Metal tip producing for scanning tunneling
microscopy, in Proceedings of 148. WE-Heraeus-Seminar: STM-Related Spectroscopies of
Semiconductor Interfaces, Bad Honnef (1995), pp. 5–7
29. N.V. Egorov, E.M. Vinogradova, Solution of boundary-value problem in bispherical
coordinates, in Proceedings of 3-th International Workshop: BDO-96, S.-Petersburg (1996),
pp. 274–278
30. G. Mesa, E. Dobado-Fuentes, J.J. Saenz, Image charge method for electrostatic calculations in
field emission diodes. J. Appl. Phys. 79(1), 39–43 (1996)
31. Y. Ohkavara, T. Naijo, T. Washio, S. Oshio, H. Ito, H. Saitoh, Field emission properties of
AlZnO whiskers modified by amorphous carbon and related films. Jpn. J. Appl. Phys. 40(12),
7013–7017 (2001)
32. K.L. Jensen, J.E. Yater, Advanced emitters for next generation rf amplifiers. J. Vac. Sci.
Technol. B. 16(4), 2038–2049 (1998)
33. K.A. Nikiforov, Mathematical modeling of the field electron emission from a metal-dielectric
systems. Dissertation for the degree of Candidate of Sciences, St. Petersburg (2005)
34. C.A. Spindt, C.E. Holland, P.R. Schwoebel, I. Brodie. J. Vac. Sci. Technol. B. 16, 758 (1998)
35. E. Mueller, T. Tson, Ion microscopy. M.: Metallurgy (1972), 360 p
36. J. Rena, S. Ranganathan (eds.), The field-ion microscopy. M.: Mir (1971), 210 p
37. A.L. Suvorov, T.L. Razinkova, V.A. Kuznetsov, Computers in field ion microscopy. Phys.
Stat. Sol. 61A, 11 (1980)
38. A.L. Suvorov, Structure and properties of the surface atomic layers of metal. M.:
Energoatomizdat (1990), 296 p
39. A.L. Suvorov, T.L. Razinkova, V.A. Kuznetsov, Computer simulation of field electro images.
Surf. Sci. 52, 697 (1975)
40. K.A. Nikiforov, N.V. Egorov, Modelling of the surface structure and the numerical
calculation of a current density of field emission metal cathode. Vestnik St. Petersburg State
University, Ser.10. Applied mathematics, computer science, management processes. 2, 39–45
41. I.M. Gribkova, Mathematical modeling of the image field in an electron microscope. Bachelor
work, St. Petersburg (2009), 37 p. I.M. Gribkova, Mathematical modeling of the image field
in an electron microscope. The Control Processes and stability, ed. by N.V. Smirnov, G.S.
Tamasyan, in Proceedings of XXXIX Scientific Conference of Applied Mathematics and
Control Processes Faculty of Saint Petersburg State University, Publishing House of St.
Petersburg state University, St. Petersburg (2008), pp. 116–121
42. N.V. Egorov, On the possibility of narrow collimated electron beams. ZTP. 52(12), 2440–
2442 (1982)
43. M. Drechsler, E. Henkel, Feldemissions-Stromdichten und Oberflachenfeldstarken bei
Feldemissionsmikroskopen sowie Methoden zur Bestimmung des Spitzenrauius, der
Spitzenform, der Vergrosserung und des Auflosungsvermogens. Z. Angew. Phys. 6, 341–
346 (1954)
44. W.P. Dyke, J.K. Trolan, Field Emission: large current densities, space charge and vacuum arc.
Phys. Rev. 89(4), 799–808 (1953)
45. W.W. Dolan, W.P. Dyke, Temperature and field emission of electrons from metals. Phys.
Rev. 95, 327–332 (1954)
46. G.E. Vibrans, Vacuum voltage breakdown as thermal instability of the emitting protrusions.
J. Appl. Phys. 35(10), 2855–2857 (1964)
47. V.M. Zhukov, Stability of the surface of the tip emitters when exposed to thermo and field.
Stability and control processes, ed. by D.A. Ovsyannikov, L.A. Petrosyanm in Proceedings of
the International Conference Volume 1: Section 1–5, 11. St. Petersburg State University,
Research Institute of the VM and PU, OOO VVM, St. Petersburg (2005), pp. 161–164
48. Y.V. Andrianov, V.N. Bazdyrev, D.A. Borisov, V.M. Zhukov, Spontaneous growth of field
emission current in a microwave field. ZTP. 61(9), 183–185 (1991)
References 227
49. V.L. Egorov, Y.V. Chentsov, Field emission cathodes in modern electron microscopes. In:
Proceedings of the State Optic Inst. L. Publishing House of the State Optic Inst, vol. 58,
pp. 68–87 (1985)
50. N.V. Egorov, K.A. Nikiforov, C-C. Shen, Reconstruction of the surface of the field electron
emitter. Poverhnost. X-ray Synchrotron Neutron. (10), 100–106 (2009)
51. K. Nikiforov, Simulation of the crystal structure of the surface of a metal cathode. The control
processes and stability, ed. by N.V. Smirnov, V.N. Starkov, in Proceedings of XXXV
Scientific Conference of Applied Mathematics and Control Processes Faculty of Saint
Petersburg State University, Publishing House of St. Petersburg state University, St.
Petersburg (2004), pp. 244–247
52. V.S. Fomenko, The Emission Properties of the Materials. Directory, 4th edn. Revised and
complementary (Naukova Dumka, Kiev, 1981), 340 p
53. L. Feldman, D. Meyer, Basics of analysis of surfaces and thin films. M.: Mir (1989), 344 p
54. L.A. Rudnicki, On the work function of the electron and the donor-acceptor properties of
non-ideal adsorption particles and coated metal surfaces. Math. USSR Acad. Sci. Ser. Nat. 26,
1240–1246 (1982)
55. V. Linnik, The method of least squares, and the theoretical foundations of mathematical and
statistical processing of observations. L.: Fizmatgiz (1962), p. 352
56. M.I. Elinson (ed.), The cold cathodes. M.: Sov. radio (1974), 386 p
57. T.I. Lewis, Theoretical interpretation of field emission experiments. Phys. Rev. 101(6), 1694–
1698 (1956)
58. W.B. Nottingham, Remarks on energy losses attending thermionic emission of electrons from
metals. Phys. Rev. 59(11), 907–908 (1941)
59. O.W. Richardson, Extraction of electrons from cold conductors in intense electric fields. Proc.
Roy. Soc. 117, 719–730 (1928)
60. W.W. Dolan, W.P. Dyke, J.K. Trolan, The field emission initiated vacuum area. II. The
resistively heated emitter. Phys. Rev. 91(5), 1054–1057 (1953)
61. V.A. Gorkov, M.I. Elinson, G.L. Yakovlev, Theoretical and experimental research in the
fore-arc phenomena of field emission. Radiotehnika i Elektronika. 7(9), 1501–1510 (1962)
62. A. van Oostrom, Dependence of the critical field strength for vacuum breakdown of tungsten
on the field emitting area, in Proceeding III International Symposium on Discharges and
Electrical insulation in Vacuum, Paris (1968), pp. 174–180
63. D. Lee, The influence of Nottingham effect on temperature of field emitting surface, in
Proceedings of Insulation of High Voltages in Vacuum, Boston (1965), pp. 96–98
64. A.S. Kompaneets, Effect of space charge on the field emission. DAN USSR 128(6), 1160–
1162 (1959)
65. Y.V. Zubenko, I.L. Sokolskaya, G.N. Fursey. ZTP. 34(5), 911–912 (1964)
66. N.V. Egorov, V.M. Zhukov, On the mechanism of formation of “rings” on the field emission
cathodes images. Abstracts, in 21st All-Union. Conference for emission electronics, L. (1990),
p. 240
67. N.V. Egorov, A.G. Karpov, Diagnostic Information and Expert Systems (St. Petersburg State
University Publishing House, St. Petersburg, 2002), 472 p
68. N.V. Egorov, B.V. Yakovlev, Evolution of the surface of liquid metal in a strong electric
field. Poverhnost. (8–9), 36–41 (1994)
69. R. Klein, L. Leder, Filed emission from Niobium in the normal and Superconducting State.
Phys. Rev. 124(4), 15–18 (1961)
70. H. Bergeret, A. Septier, Finding of cathode heating owing to Nottingham effect. CR Acad.
Sci. 277(17), 489–492 (1973)
71. L.M. Baskin, V.A. Godyak, O.I. Lvov et al., Effect of space charge of relativistic electrons in
the field emission. ZTP. 42(6), 1282–1287 (1972)
72. N.E. Alexeevskiy, The impact of the emission current on the critical parameters of tantalum.
DAN USSR. 242(4), 816 (1978)
73. V.M. Zhukov, Processes on the Surface Under Field Emission (VVM, St. Petersburg, 2007),
295 p
228 4 Simulation of Structure and Parameters of Field Emission Cathodes
Abstract In the beginning of the chapter the main problems and obstacles to
widespread practical use of field emission cathodes are considered. Then it proceeds
to discuss the features, advantages and disadvantages of various types of field
cathodes known as of today.
cathode accompanied by heating can extend its service life by emitting surface
reconstruction.
In practice, that operating mode of a field emission cathode is extremely
inconvenient because it involves expansion of the field emission cathode design and
the device in general and deprives it of some advantages over hot cathodes.
Surface migration of adsorbed residual-gas atoms in a strong electric field also
tapers the emitter, which results in growth of E and j up to breakdown. However,
the electrical breakdown of this case is special. It does not result in loss of the field
emission cathode. What the breakdown destroys here is adsorbed-atom micropiles,
which causes field emission current surges in its turn [13].
Migration of adsorbed residual-gas atoms can be excluded by heating of the field
emission cathode to clean its surface. In this case, tungsten has to be heated to
*1920 K [14], rhodium *1320 K and palladium *1120 K [15].
The area of research of field emission current stabilization in heater field
emission cathodes is not prospective because of the above-mentioned inconve-
niences of practical implementation.
3. Let us consider the field emission current stabilization potential coming from
application of ion traps and electrostatic gates.
Such device designs were first suggested by M.I. Elinson et al. and thoroughly
considered in monograph [6]. The research shows that such devices can consid-
erably reduce the number of ions bombarding the working section of a pointed
emitter, thereby extending its service life. However, it is not always possible to
resort to ion traps and gates in practice even if the inter-electrode distances are
relatively large because they significantly complicate the design of the cathode
assembly. Such designs have not been widely used yet.
4. The pulse operation mode considerably reduces average current loading of the
field emission cathode and the thermal loading of the anode. With other things
being equal, the mean power released under the pulse operation mode is sq
times less than under the continuous one, being the pulse length in seconds and
q being the number of pulses per second.
Numerous experiments testify that field emission current is much more stable
under the pulse mode than under continuous one. However, use of field emission
cathodes only under the pulse operation mode prelimits the potential area of their
application.
5. Sputter-resistant substances suitable for making field emission cathodes are
looked for continuously, e.g. [16–21].
Especially careful research covered high-strength metals, usually tungsten,
rhenium, tantalum and metal-like compounds of the carbide and boride group.
The research shows that field emission cathodes made of a series of substances,
e.g. rhenium, lanthanum hexaboride, provide a more stable field emission current
than those of tungsten.
232 5 Field Emission Cathodes
(*10−2 m), change of their dimensions can fluctuate within (5–20) µm. Obviously,
field emission current is instable in designs with a very small value of R, in order of
several µm, without thermostating of the electrodes exceeding one centimeter.
Apparently, if R is very small, one can only make the diodes having small linear
dimensions of electrodes. However, development of any current-controlled elec-
tronic device involves additional electrodes to be located in the cathode-anode
space volume. Development of controlled electronic devices needs R of at least a
few tens of microns, which seems to be a reasonable practical limit.
8. Let us consider one more area of research of field emission current stability
improvement. This area is connected with a specific feature of emission char-
acteristic of field emission current for semiconductors [9, 10]. This characteristic
i(u) has a saturation region in external-voltage area u or external-field intensity
E. In this region, field emission current does not depend much on
potential-barrier penetrability, i.e. the field emission cathode surface state. In
other words, the adsorbed-atom layer on the semiconductor field emission
cathode surface resulting from residual-gas adsorption scarcely affects the value
of field emission current, the latter being stabilized thereby.
An author of this book has carried out a great deal of experimental research of
field emission from semiconductor materials, mostly silicon [10]. His papers show
that the prerequisites and conclusions of the semiconductor field emission theory
are in reasonable agreement with the results of experimental research.
Paper [10] points out «… that semiconductor materials can become the basis
for… cardinally new solution of field emission cathode stability improvement».
However, the semiconductor area of research has a series of restrictions.
A relatively high specific resistance of semiconductors entails a considerable Joule
heating of such field emission cathodes with flowing current and limits the possible
values of field emission current. However, the restrictions caused by technical
properties of field emission cathode materials are worsened by a more serious
physical reason constraining development of research in this area. The matter is that
the saturation region makes current uncontrollable or slightly controllable by
external voltage.
That is the critical weakness of semiconductor field emission cathodes: they
practically cannot serve as the basis for current-controlled electrovacuum devices
without a complex cathode assembly.
9. Creation of a statistically stable system of emitting microelevations became
possible with development of carbon field emission research [11]. Such mate-
rials can develop the maximum number of emission centers uniformly spread
over the surface of the field emission cathodes and making nearly the same
contribution to total emission current, i.e. the materials develop an emitting
surface. At that, the internal structure of some carbon materials causes some
emitting microelevations destroyed as a result of ion bombardment to be
replaced with emission centers of similar parameters emerging from the material
structure. The result is a considerable long-term stability of the field emission
234 5 Field Emission Cathodes
cathodes made of such materials under the conditions of high technical vacuum
generated in sealed-off electronic devices. The field emissivity of carbon
materials and designs of field emission cathodes on their basis are featured in
detail in Chap. 6.
So the brief analysis of the identified areas of research of field emission current
stability improvement has shown that the progress achieved in some of them has still
been insufficient to provide field emission cathodes with a stable electron emission in
technical vacuum or to maintain some obvious advantages of field emission cathodes
over hot cathodes, which cannot but constrain the device developers’ interest to such
research. Therefore, it would be expedient to supplement those areas of research with
some new ones supposedly able to improve field emission current stability, which is
the central problem in this area of physical electronics.
The chapter continues with brief overview of the main types of field emission
cathodes and their specific features.
restructuring of the emitter (Fig. 5.1c) and subsequently is reshapes so that the
entire emission takes place only from the facet [100] (Fig. 5.1d).
Emitter restructurization is accompanied by stepping up of the emission current.
At this stage of processing of the cathode, emission is stabilized at constant values
of temperature and voltage [34].
The field emission cathodes produced in this way are capable of a long service
life at T 1200–1700 K with current takeoff of 1–300 µA at residual-gas pressure
of q 10–6 10−5 Pa. Pressure reduction to q 10−9 mm Hg makes it possible
to step down the temperature of the field emission cathode by 570 K.
Paper [35] focuses on application of two-component coatings for field emission
current localization in a small solid angle. Boride field emission cathodes are pro-
duced by consecutive etching in the solutions of sulphuric and phosphoric acids [36].
The limiting current value that can be taken off from a field emission cathode is
limited to heating of the emitter with Joule heat and the Nottingham effect and
ultimately determined by the heat-conducting path conditions [37], the apex
geometry [38] and the operating mode [39]. In practice, the volt-ampere charac-
teristic biases towards current reduction, which limits the emission localization
possibilities. Paper [40] explains this phenomenon by influence of spatial charge.
There are also some other explanations, e.g. those connected with difference
between the true shape of the potential barrier and the classical image-force law as
adopted to derive the barrier-transparency expression [41]. Influence of the space
charge on field emission was studied in paper [42]. The estimates made are in good
agreement with the respective experimental data. So a particular current density of
(1 5) 10−6 A/cm2 causes emission in the localization region to grow much
more slowly than in the surface region unaffected by the space charge. The latter
equalizes emission over the entire emitting surface, practically delocalizing it and
making further reduction of the emitting area inexpedient.
(a) (b)
310
100
310 310
100
310 310
310
(c) (d)
100 100
310 310 310 310
Fig. 5.1 Schematic of rearrangement of emitting surfaces for reduction of the electron beam solid
angle: a initial monocrystal and its Miller indices; b emission in initial state; c beginning of
thermo-field rearrangement process; d final state
5.2 Pointed and Multiple-Apex Field Emission Cathodes 237
Fig. 5.2 Diagram of multi-tip field cathode manufacture out of metal foil
Fig. 5.4 Design of multi-tip field cathode with breakdown-proof electrode: a cathode design; 1
tip field cathodes; 2 additional electrode; 3 isolator; 4 contact electrode; b flux distribution between
field cathodes and the protective electrode at zero potential between electrodes; c at positive
potential on electrode 2
semiconductor substrate through the gaps between the island regions up to devel-
opment of microcones with pin-point tips. The rest parts of the mask were removed
from the tips to strip off pointed microstructures.
Etching was done by a chemical-dynamical method in the etchants consisting of
the mixtures of nitric, hydrofluoric and acetic acids. Depending on the
plate-conductivity type, orientation and dimensions of the etched structures, the
equivalence ratio in the etchants varied.
The etching nature was essentially determined by the plate orientation and the
shape of the image elements in the photo mask.
240 5 Field Emission Cathodes
Fig. 5.5 Final stages of manufacturing of a field cathode out of metal particles: 1 substrate; 2
solvent; 3 metal particles; 4 thin film on the solvent surface; 5 microledge; 6, 7 holes in the sides of
microledges
deep etching of facets (100), (110) and (111) of single-crystal silicon polished by
the chemical-dynamical method. The papers describe testing of several types of
masking materials [57, 61].
The geometrical homogeneity achievable with the etching method throughout
the matrix area depended on many factors. The major of them were the uniform
exposure of the photoresist, the target-revolution speed at etching and the etchant
composition. Naturally, some apex-etching cessation skills were also necessary.
A wrong selection of the revolution speed at etching caused the matrix center or
edges to be etched too deeply, which prevented those regions from emission.
Moreover, the method based on liquid chemical etching suggested the resolving
ability limited to the effect of underetching under the masking coating. That method
could not make apex packing at the given height more dense because the initial size,
i.e. that of the etching-resistant mask, and the final size of the resulting apex differed
by more than two orders of magnitude, otherwise, the apices could not achieve the
necessary height. So deviation of the mask dimensions during image transmission
or a local speed variation conduced to geometrical heterogeneity of the resulting
structures. That was why there were some requirements to the etching process.
Nevertheless, chemical etching cannot provide reproducibility of fabrication of
pointed field emission cathodes.
It was suggested to increase effectiveness of multiple-apex cathodes through
apex-packing densification at the given height. That method differs from previous
ones in greater reproducibility of the technological process along with minimum
contamination of the semiconductor substrate surface and provides uniform emis-
sion of surface-structure elements. Ion-plasma etching is included in the flowchart
and the surface decontamination system for all stages of the technological process is
developed. Ion-plasma etching could provide definite-height bumps, which makes it
possible to begin controlling quality of multiple-apex structures as early as at the
initial stage of fabrication.
Papers [62–65] suggest fabrication of multiemitter semiconductor structures by
application of a Ta2O5 film as an etch-resistant mask, the method being now used
for germanium. The emitter-matrix fabrication flowchart is shown in Fig. 5.6.
The germanium substrate decontaminated by magnetron deposition was layered
with tantalum, after which projection photolithography created a tantalum disk
matrix subsequently oxidized to Ta2O5 (stage I). The disks were 10–15 lm in
diameter and 0.1–0.2 lm thick. At stage II, ion-plasma etching created some
definite-height bumps with etch-resistant masks of Ta2O5. Durable oxide Ta2O5
with good dielectric properties makes it possible to reach high etching coefficients
during ion-plasma sputtering at direct current. The resulting bumps were chemically
etched in the mixture of HNO3:CH3COOH:HF = 5:3:3 up to formation of a pointed
structure (stage III). Having good adhesion properties, the Ta2O5 mask stays up to
final formation of apices and remains almost unaffected by the acid mixture.
Etching results in formation of apices under the Ta2O5 layer. It is especially
important to control the structure resulting from etching at the final stage of etching.
The control is visually assisted with an optical microscope because the Ta2O5 layer
is transparent. Finally, the Ta2O5 mask is removed by a high-frequency discharge in
242 5 Field Emission Cathodes
the argon medium (stage IV). Figure 5.7 shows a typical microphoto of the
resulting structures.
Field photocathode matrices were fabricated from single-crystal plates of ger-
manium with specific resistance ranging from 3 to 300 X m, orientation (111), the
area of a few cm2 and thickness of 0.3–0.8 mm.
Homogeneity of emission is largely predetermined by geometrical homogeneity
of the cathode matrix. However, research often identifies some uncontrollable
centers of emission absolutely unrelated to the resulting structure, which used to be
the case for silicon emitters. Auger analysis confirms [66] the earlier-made con-
clusion that the uncontrollable centers are connected with admixture micro inclu-
sions in the initial semiconductor plate, as well as contamination of its surface and
defects in workmanship.
Therefore we removed the etch-resistant dielectric mask by high-frequency
discharge in argon [64], which made it possible to exclude the operation of removal
of the waste photoresist by boiling in monoethanolamine.
It was a large step forward in development of multiple-apex field emission
cathodes when Spindt et al. suggested a field emission cathode fabrication tech-
nique combining a thin-film technology and electron-beam microlithography [67–
69]. The basic fabrication diagram is presented in Fig. 5.8.
5.2 Pointed and Multiple-Apex Field Emission Cathodes 243
The initial structure is shown in Fig. 5.8a. Insulating substrate 1 is covered with
a thin layer of metal 2, namely molybdenum, which serves as a cathode electrode.
Over it, there is dielectric layer 3 and control electrode 4, which is also made of
molybdenum. Layer 3 includes cathode cavity 5 as shown in Fig. 5.8a.
Then the revolving substrate is sprayed with an Al2O3 dielectric layer at the
acute angle of about 15°. The angle value was determined by the necessity to
prevent Al2O3 from penetrating into cavity 5 (Fig. 5.8b). When the hole diameter
decreases down to some value determined by the parameters of sputtering, the
molybdenum sputtering source located completely perpendicularly to the structure
surface is turned on (Fig. 5.8c). Due to continuous reduction of the hole diameter,
simultaneous sputtering of Al2O3 and Mo shapes apex 6 on substrate 2.
Some molybdenum that deposits on the Al2O3 layer surface forms mixture 7,
which is chemically dissolved in boiling phosphoric acid and removed. The result is
a thin-film field emission cathode with a control electrode (Fig. 5.8d). The resulting
structure has the following characteristic dimensions, insulating layer 3 as thick as
0.35 1.5 µm, the control electrode as thick as 0.4, 1.5 µm diameter of the hole in
the control electrode, *100 Å rounding-off radius of the resulting pointed emitter.
Such field emission cathodes are shaped in large matrices with the number of
emitters ranging from 5000 [70] to 50,000 [71]. The distance between the centers of
the emitters forming a rectangular matrix range from 12.7 [70] to 4 µm [71]. The
apices are packed as densely as 108 cm−2.
Such analysis of field emission cathodes is given in [70]. Recent achievement
under the pulse mode is currents of 300 mA lasting 100 ls, which is 6 µA per
emitter on average.
244 5 Field Emission Cathodes
Fig. 5.8 Schematic of manufacturing of Spindt field cathodes: 1 substrate; 2 field cathode
electrode; 3 insulating layer; 4 control electrode; 5 hollow part of the cathode; 6 field emitter;
a base blank; b spraying on the protective layer; c forming the tip; d finished structure
Over the years that have passed since invention of this technology, its basic
development has been aimed at densification of apices and involvement of different
materials for the main electrodes of the emitter, the substrate, the control electrodes
etc.
Though the running voltage of such structures is hundreds of volts, short dis-
tances between the cathode and the control electrode cause desorption effects that
can conduce to development of electric arc 2, which puts the field emission cathode
out of order.
So there were attempts to reduce the running voltage by coating of apices with
materials reducing the electron work function and thereby the voltage. E.g., [72]
describes a pointed field emission cathode coated with cesium under the conditions
of the ready device. The amount of the adsorbate should be optimum to minimize
the running voltage and at the same time it should not cover the dielectric layers to
minimize leakage currents. However, such application of cesium prevents from
taking the cathode structure outside, e.g. to mount the devices.
5.2 Pointed and Multiple-Apex Field Emission Cathodes 245
It seems prospective to process the cathode structure with hydrogen plasma [73,
74]. The chemical sorption of hydrogen on molybdenum reduces the electron work
function from *0.3 eV at the processing time of 10 min to *0.92 eV at the
processing time of 100 min. That increases the emission current by 15% under
constant voltage.
Another example is sputtering of ZrC on molybdenum apices [75]. ZrC is well
known to have good chemical stability and resistance to cathode sputtering.
Sputtering took place at the pressure of 10−6 mm Hg. The field emission trials
showed threshold-voltage reduction by about 40% in comparison with pure
molybdenum apices, which is explained by reduction of the electron work function
of the field-emitter material.
A natural desire to simplify fabrication of such field emission cathodes and
improve their operational characteristics, especially the ability to work under the
conditions of serial sealed-off devices, resulted in a considerable number of sug-
gestions in this area.
It is well known that stabilization of the field emission cathode can be improved
by inclusion of a high-ohmic resistor comparable with internal resistance of the
control electrode—field emission cathode gap, in its circuit. An example of
implementation of such a solution is given in [76], which illustrates a standard
approach to this problem. It involves sputtering on the substrate and cathode
contacts of the resistive layer.
Practical purposes are more often attained with well-focused electron beams.
However, the positive potential of the extending electrode causes expansion of the
electron beam. The electrons are drawn to positively-charged ring electrodes. They
will have both radial and axial component of the speed depending on the emission
angle. Therefore use of the second, focusing electrode with a potential lower than
that of the control electrode gives a good focusing effect [77]. This idea was tested
on silicon pointed cathodes located below the extraction electrode by 400–0.7 µm,
respectively [78].
For these dimensions, a negative-resistance region is created on the focusing
electrode at the voltage of 4 12 V. The emission current is stepped up with
voltage increase at the extraction electrode; however, the number of emitted elec-
trons reaching the anode is thereby decreased due to their repulsion by the focusing
electrode. Further increase in extraction voltage switches the anode current off. For
the voltage at the focusing electrode >12 V, increase in the extraction voltage
makes the anode current scale up. That shows absence of negative resistance in the
emission circuit.
It has already been pointed out that a huge number of papers is dedicated search
of alternative materials suitable for Spindt structure. This section is not to mention
all such attempts, but give a few examples. Many metals were tested as a material of
field emission cathodes, e.g. Pt [79], silicon [78] and carbon nanotubes [80, 81].
The technologies of fabrication of metal and silicon apices were featured above.
Carbon nanotubes are deposited on the bottom of the ready structure of each
element by the CVD method, where a metal catalytic agent is preliminarily sput-
tered. The number of the grown-up nanotubes is uncontrollable is this case. To
246 5 Field Emission Cathodes
reduce the number of nanotubes per element, there were attempts to grow them up
on molybdenum apices of Spindt cathode structures. The structure combining the
thin-film technology and growing-up of whiskers is given in [82]. A low-resistant
n-type silicon substrate was sputtered with a SiO2 layer as thick as 1 µm. Then a
layer of molybdenum as thick as 150 nm was applied by electron sputtering. The
control-electrode aperture was made by the usual technology involving pho-
tolithography and reactive ion etching. The sputtered structure was plunged into a
diluted solution of hydrofluoric acid (water:HF = 6:1) to remove oxides and form
an aperture. The end structure had an aperture of 2.5 µm diameter and *1 µm
depth.
Tungsten nanofibers were applied by the electron-induced deposition method in
WF6 vapor. That material was selected for two reasons: (1) its boiling temperature
is 17.5 °C, which mitigates the condensation problem; (2) absence of carbon and
oxygen should conduce to higher purity of the coating applied. The microfiber
radius essentially depends on the conditions of the technological process such as
pressure, current, time etc., but identical parameters of the technological process
result in nearly identical radii. Paper [82] mentions fabrication of tungsten nano-
fibers 1 µm high, 40 nm thick and having the apex diameter of less than 15 nm.
The field emission measurements were made in a vacuum chamber at the pressure
of <6.7 10−5 mm Hg. The emission characteristics were measured with an
earthed control electrode. The anode located 6.7 mm from the structure was sup-
plied with the voltage of +1000 V, and the cathode with the voltage ranging from
−60 to 250 V. The threshold voltage determined as the cathode-voltage value
necessary to generate the emission current of 20 nA was 139 V at the moment of
energization. Then it dropped down to 90 V within a few minutes. That drop of
control voltage at constant current can be accounted for by contamination of the
tungsten apex in the air.
It was discovered that the current of more than 40 nA released by the apex
destroys it, which seems to be connected with very small rounding-off radius of the
apex. At less 20 nA, the emission current remained stable.
The present-day production demands increase in productivity, which in our case
means increase in the number of simultaneously fabricated apices. It is especially
important in fabrication of large display screens. However, there are numerous
technological difficulties in this way. Growing of field emission cathodes of iden-
tical height nearly equal to the thickness of the dielectric layer requires homoge-
neous sputtering over the entire surface of the substrate. For this purpose, an
ordinary sputtering plant needs a large distance between the sputtering source and
the substrate. E.g., the substrate of 10 cm in diameter needs the working distance of
1 m, i.e. the angular spread of the flowing particles to be sputtered not more than
10°. Moreover, reduction of the ratio of the field emission cathode diameter to its
height is accompanied by reduction of the necessary diameter of the hole in the
control electrode from more than 1 µm to less than 0.3 µm without respective
reduction of thickness of the dielectric layer. That also makes it necessary to reduce
the angular spread of the current of particles to be sputtered to prevent early closure
of the cathode cavity and reduction of the cone height. Further increase in the
5.2 Pointed and Multiple-Apex Field Emission Cathodes 247
The fluctuations of the field emission current measured in ultrahigh vacuum with
pressure <10−10 mm Hg, at the current density of 1.46 mA/cm2 are less than 15%.
In addition to sputtering processes, multiple-apex field emission cathodes can be
formed by pulsed laser bombardment of the silicon surface in the atmosphere of
reacting gases, e.g. SF6 or O2. The bombardments conduce to formation of
self-organized quasi-periodic structures consisting of apices with high aspect ratio
[88]. That is a simple one-step technology to fabricate quasi-periodic structures
without vacuum equipment or contaminant-free rooms. Such structures termed as
black silicon can be sufficiently effective field emission cathodes [89].
Among the pulsing laser parameters, namely wave length, pulse duration, pulse
energy, pulse count, it is pulse duration that is most important for formation of an
apex array.
Paper [90] considers the effect of pulsed-laser bombardment duration on the
surface morphology and field emissivity of silicon irradiated by KrF laser pulses in
the presence of SF6. The KrF laser wave length is 248 nm. The pulse duration used
in experiments is 15; 5; 0.5 ns. The process is detailed in [91]. The field emission
cathode blanks are made of n-type phosphorus-doped Si (100) of 2 2 mm2. The
resulting emitter samples are tested in a vacuum chamber at the pressure better than
10−5 mm Hg in the diode mode. Before electrical tests, the samples are treated in a
10%-water solution of HF to remove oxides from the apex surface. The distance
between the apices and the anode is 150 ± 10 µm for the samples irradiated with
the pulses of 0.5 and 5 ns and 300 ± 10 µm for the pulses of 15 ns. The main
findings of the experiments are given in Table 5.1.
The geometrical parameters are calculated from the cathode-structure image
analyses in a raster-type electron microscope. It is worth noting that the laser pulses
of 15 ns produce the apices as high as 65 µm having the rounding-off radius of
*670 nm. At the same time, reduction of pulse duration reduces the height and the
5.2 Pointed and Multiple-Apex Field Emission Cathodes 249
Table 5.1 Morphological characteristics and field emissivity of the microstructured Si surface
The pulse Typical radii The The average The Threshold
duration of tips height of distance density of field
k = 248 nm curvature tips between apices apices (V/m km)
(nm) (m km) (m km) (1/cm2)
15 ns 670 ± 120 65 ± 5 11 ± 4 4.3 106 2.5
5 ps 470 ± 110 10 ± 2 5.5 ± 1.5 4.33 107 6.0
0.5 ps 530 ± 200 8±1 5.0 ± 1.3 5.53 107 8.0
rounding-off radius of the apices. However, a better combination of the height and
the rounding-off radius in the first case give a higher form factor, which reduces the
threshold field.
Field emission formulas show that there are only two ways to step up total emission
current:
1. Densification of the field emission current.
2. Increase in the emitting-surface area.
Field emission current can be densified by both reduction of electron work
function u of the cathode material and increase in field intensity E.
A change of the u value of the emitter causes a change of anode voltage
U necessary to obtain a given current value i. E.g., if the given current density is
*105 A/cm2, reduction of u from 4.6 to 4.0 eV reduces the anode voltage by
almost 18%, whereas increase of u from 4.6 to 6.3 eV necessitates increasing the
anode voltage by almost 60%.
It follows that low u-value materials are quite useful in terms of both densifi-
cation of field emission current and stepping-down of anode voltage.
Emission electronics knows a lot of materials having rather low u reaching (1.5–
2.5) eV [92]. However, so low u values are only characteristic for man-made
metal-film systems based on high-melting transition metals and surface films of
atomic cesium, barium, etc. The ideal conditions for long-term existence of such
films are very low temperatures and perfect vacuum. Under real conditions, these
films fail rather rapidly, and their perpetual recovery comes from the reserves of the
active substance in the cathode body, which is only possible at elevated tempera-
tures sufficient to activate thermodiffusion processes.
Application of effective metal-film systems can hardly be successful in case of
the field emission cathodes working at room or even lower temperatures when the
diffusion rates are negligibly low. E.g., the lanthanum hexaboride surface film
cannot to be recovered through diffusion of its atoms from the cathode body at
room temperature. Instability of field emission current of the preactivated
250 5 Field Emission Cathodes
lanthanum hexaboride cathode is obvious in this case. Such and similar cathodes
need heating for normal diffusion processes and recovery of active surface film of
adsorbed atoms. A promising way to solve the problem of stability of field emission
current is search of the materials adsorbing residual gases in the way hardly
affecting the electron work function. E.g., oxygen adsorption on clean surfaces of
niobium and tantalum increases the u value from (4.0–4.2) V, i.e. by as little as
14%.
At that value of u, field emission current can be further densified through
increase in E. However, increase in E is in its turn limited by the mechanical
strength of the field emission cathode material. The theoretical strength of metals is
very high, *(103–104) kg/mm2.
However, the defects of crystal lattices of metals and, first of all, dislocations
reduce these values by several orders of magnitude. Therefore, the data on the
strength of real samples of metals should be treated with care since they charac-
terize only this or that technical sample of metal with its technological history.
It was already mentioned that Joule heating of a field emission cathode is notably
manifested at j 107 A/cm2. This field emission current density value can be
accepted as a benchmark upper limit of this value, and the respective value of E can
be accepted as the electric-intensity upper limit. A particular value of E depends on
the electron work function of the field emission cathode material, the value of
E being less, the less is the value of u.
Actually, such high current-density values can only be obtained in ultrahigh
vacuum in absence of ion bombardment. In technical vacuum, field emission is only
stable at the current density of (102–103) A/cm2, when ion bombardment is rela-
tively small. Current stepping-up conduces to avalanche development of secondary
processes in the inter-electrode gap and the anode of the device [93], resulting in
destruction and loss of the field emission cathode within a few minutes at contin-
uous operation in technical vacuum.
Summing up the discussion of potential increase in the total field emission
current through its densification, we can see that this potential is quite high. The
possible field emission current density values exceed those for hot cathodes by
several orders of magnitude. Nevertheless, they are limited by field emission
cathode material characteristics, such as their electron work function and
mechanical strength. These limitations are of the fundamental physical nature. On
the contrary, the potential for increase in the total current through increase in the
area of the emitting surface is not fundamentally limited. This way of increase in
total current has no physical limitations and seems to be very promising. Blade,
multiple-blade and similar systems enable substantial growth of the emitting surface
area of the field emission cathode up to 10−2–10−1 cm2. Dike suggested increasing
the emitting surface of field emission cathodes with keen blades [94], and E.G.
Shirokov suggested a keen disk chamfered along the perimeter of the disk and
rounded off with the radius of about a micron [95, 96]. A blade field emission
cathode can be exemplified with a multiple-blade field emission cathode assembled
from steel safety-razor blades [97] having a relatively constant rounding-off radius
5.3 Blade and Wire Field Emission Cathodes 251
of the apex from sample to sample. This cathode has shown itself off as a reliable
and convenient emitter for development of electronic devices.
Figure 5.10 presents a design of a prototype diode with a multiple-blade field
emission cathode. The system involved nine elements cut out of the whole blade by
spark cutting. Each element was 10 mm long. The experiments tested different
kinds of standard blades. The best results, both in terms of reproducibility of
voltage-current characteristics and the limiting values of anode current, were shown
by «Perma Super» blades of high-strength steel with the 18°–21° angle of factory
sharpening. The rounding-off radius of such a blade was 1.0 µm. The anode and the
fixing details of the cathode were made of stainless steel. The working surface of
the anode was ground. The insulator was ceramic. The distance of 0.2 mm was set
between the electrodes by measurement probes. The prototype diode was fixed on
molybdenum lead-ins of a dismountable glass device.
The experiments involved continuous pumping-out with an oil-diffusion pump at
the pressure of *10−7 Torr.
The emission characteristics of field emission cathodes were taken in a pulse
mode with a generator of squared high-voltage pulses lasting 60 µs at the recur-
rence rate of 100 pps. Three diodes with voltage-current characteristics that coin-
cided with each other within the possible current and voltage measurement
variations, were fabricated and examined.
The research was limited with the extreme anode-current values not exceeding
10−2 A because increase of this value typically conduced to emission-current
instability, deviation of Fowler–Nordheim voltage-current characteristics in the area
of lower current values and then to vacuum breakdown and loss of the cathode as a
result of destruction of its emitting edge.
Shear blade field emission cathodes turned out more convenient in fabrication
and application. Paper [98] exemplifies fabrication of such field emission cathodes
of tantalum, niobium and copper. The disks of 13 mm in diameter were made of
50 µm foil. They were electrochemically etched in a special installation similar to
the one featured in paper [98]. The disk to be etched was fixed with two Teflon
Fig. 5.10 A model of a diode with a multi-blade cathode. 1 Blade (cathode element); 2 anode; 3
ceramic; 4 base for fastening on the blades; 5 fastening rod (cathode input)
252 5 Field Emission Cathodes
washers on the engine shaft. The axial play of the engine did not exceed 50 µm.
The rotation rate was 100 rpm. The etching was visually controlled with a binocular
microscope.
Tantalum was etched in a mixture of concentrated hydrofluoric acid and nitric
acid concentrated up to 38% with addition of water (15:80:5). The etch-tank
electrode voltage was 1.5 V. The circuital current was 20 mA. The etching time
was 20–80 min.
Niobium was etched in the mixture of concentrated hydrofluoric and nitric acids
in the ratio of 25:70:5 at the voltage of 2 V and the current of 20–40 mA. The
negative electrode used in etching of both metals was an iridium-wire grid of
0.2 mm in diameter. The disk to be etched was immersed at the depth of 1 mm.
Copper disks were etched to the cutting-edge sharpness in a chromic mixture.
The cathode was represented by a lead plate. The disk rotated at the rate of 10 rpm.
The etching time was 30 min. Copper was etched well in case of both the elec-
trochemical method (u = 1.5–20 V) and the chemical method, i.e. without voltage.
After the etching, the blade field emission cathode was formed by heating with
high-frequency currents in vacuum with simultaneous exposure to electric field of
reverse polarity, i.e. the positive of high voltage was supplied to the disk cathode,
and the negative went to the anode. Joint action of surface-tension forces and
ponderomotive forces was supposed to result in leveling and reduction of the radius
along the perimeter.
The field emission cathodes were tested in a sealed-off cylinder-anode device at
the pressure of *10−7 mm Hg. The anode-cathode distance was 0.5 mm. The
current of 100 µA was generated at the voltages of 5 kV for tantalum, 6 kV for
niobium, 7 kV for copper in a steady state.
For all the three field emission cathodes, emission was instable in time and
heterogeneous along the disk edge. It is especially characteristic for a copper field
emitter. In particular, its exposure with 5 minute-current takeoff made the current
drop from 100 to 20 µA.
For the emitters involved, the attempts of large-current takeoff by voltage
increase conduced to transition of field emission into a vacuum arc in some regions
of the blade as a result of which these regions of the emitter failed and melted off.
This phenomenon can be accounted for by the spread of curvature radii in different
regions of the blade and different electron work function of crystal facets opening
out into the metal surface.
Unlike tungsten and molybdenum, which have no stable oxide of their own,
application of silicon turned out most preferable for the field emission cathodes
working in technical vacuum (P * 10−7 mm Hg).
Papers [99, 100] suggest a design of matrix blade field emission cathode of
silicon.
The cathode assembly is a matrix of silicon field emission cathodes containing
6400 pointed cathodes or 3200 blade cathodes spaced 5 µm from each other. The
blade cathode is 5 µm long. The pointed and blade field emission cathodes are
packed in the matrix at the density of 4 106 and 2 106 cm2, respectively.
5.3 Blade and Wire Field Emission Cathodes 253
p3 dDl
S¼ h d:
3 h
The working capacity of the field emitter was tested with a prototype emitter
having the following parameters: rod mount (molybdenum, tungsten) D = l mm,
l = 1 cm, d = 50 µm, h = 150 lm.
The helix ends are attached to the rod mount by spot welding. The system is
hardened by coating of the rod-mount surface with a thin layer of Aquadag water
solution, the layer being as thick as 0.5–0.6 diameter of the wire. A smaller Aquadag
thickness value would degrade the mechanical strength of the field emission cathode,
while a larger one would erode field emission current stability. The field emitter is
installed in the diode coaxial system with the anode-cathode distance of 2 mm. The
prototype diode is pumped out up to the pressure of 10–7 mm Hg with heating of the
entire system up to 450 °C. The emitter is heated up to 1500 K by passage of current
through the rod mount. Some aging at the voltage of 15 kV makes it possible to
generate the current of 35 mA.
The field emission current stability is 1.5–2 times as high as the cylindrical field
emission cathode stability with the emitter service-life of 300 h.
The emitter is highly reliable as the field emission cathode does not decay even
after inception of an arc discharge in the diode, while the cylindrical cathode
usually breaks after such a discharge.
Summing up, it may be said that the emission produced by blade field emission
cathodes is extremely heterogeneous.
The spread in the values of the rounding-off radius in different places of the apex
of such cathodes is as wide as *40%, and the emitting surface area is only about
10−4–10−3 of the average surface area of the blade as determined by its geometrical
dimensions.
The wedge shape of blade field emission cathodes is their fundamental weakness
because ion bombardment blunts the keen edge of the blade and reduces field
emission current.
The emitting centers of blade and wire field emission cathodes are the
microelevations on their surface formed as a result of preliminary mechanical and
electrochemical processing of the latter.
Film field emission cathodes appeared due to two factors. Firstly, cathode sput-
tering noticeably change the rounding-off radii of the tip of pointed and blade field
emission cathodes in the course of operation, which changes electric field intensity
and field emission current accordingly. Secondly, stepping up of field emission
current necessitates increase in the emitting-surface area. Such field emission
cathodes were suggested in [104, 105]. The potential stability of field emission
current is attributable to the fact that change of form factor b in the course of
operation was much less than this value [106]
256 5 Field Emission Cathodes
db b dR
;
dt 2R dt
Fig. 5.13 Schematics of diodes with single-film a and multi-film b field cathodes: 1 field cathode;
2 anode; 3 substrate; 4 an area of the field cathode that is destroyed in the process of cathodic
evaporation; arrows show the directions of field electrons (–e) to the anode and positively charged
ions (+) to the cathode
5.4 Film Field Emission Cathodes 257
Teflon bath, a bath vertical-shift mechanism and a binocular microscope for visual
control of etching. The disk plunged into the etching solution to the depth of 1 mm.
The etching went on with continuous revolution of the disk at the rate of 60 rpm.
The axial play of the shaft did not exceed 50 µm. The etching took about 30 min.
The result was a chromium film with the end face uniformly projected over the
copper substrate along the entire perimeter of the disk at the height of l = 50–
100 µm.
Such field emission cathodes can be not only disk-shaped, but also flat, tubular
etc. The experimental research of polycrystalline film field emission cathodes made
of chromium [108], titanium [110], niobium [111, 112], hafnium [113] and other
metals confirmed higher stability of field emission current in comparison with that
of the pointed ones at the residual-gas pressure of 10–6 10–7 mm Hg.
5.4 Film Field Emission Cathodes 259
Field emission tests were limited with the emission current of 100–500 µA in the
continuous mode and 4–5 mA in the pulse mode [114], which was caused by
overheating of the anode.
Film field emission cathodes open up broad possibilities for fabrication of
electronic devices with microelectronic technologies, which encourages gives a
new impetus to development of such field emission cathodes.
At the initial stage of the research, the film of a metal or a semiconductor was
considered as perfectly smooth, and form factor b was determined by the effective
rounding-off radius of the film end face. Further studies found out that the emitting
surface of the films had some microroughness, which reduces the running voltage, on
the one hand, and queries stability of such field emission cathodes, on the other hand.
The attempts to look into the problem were made in [115] and developed in
[116]. Any device manufactured in the microelectronic technology is based on
reduction of the running voltage, which mostly depends on form factor b.
As for the multifilm field emission cathode [116], running voltage largely
depends on the geometrical macroscopical parameters of the device design, namely
the anode-cathode distance, the cathode-cathode distance, the cathode height, the
film thickness and the field emission cathode tip curvature radius, accordingly.
The experiments were carried out with the system of silicon films formed by the
methods of integrated technology located normally to the monocrystal substrate of
orientation (110) (Fig. 5.15). The samples were boron-doped to the concentration of
*1020 cm−3. The researchers focused on two types of experimental samples of film
field emission cathodes. The first-type films were as high as 5 µm and the distance
between them was 5 µm, too. The second-type films were as high as 1 µm and the
distance between them was 10 µm. Studying of the samples in a raster-type electron
microscope showed that the field emission cathode fabrication technology was able
to produce the films of cylindrical tips with quite the same curvature radius. The
emissivity characteristics were studied with the cathodes filmed with the thickness
ranging from 40 to 400 nm. The samples of d = 0.1 µm showed experimental
dependences of the form factor on the cathode-anode distance (Fig. 5.16). The gap
varied from 20 to 165 µm for the first-type samples and from 0.2 to 165 µm for the
second-type samples.
ECE = 3.8 µA/mm. Figure 5.18 shows that total stoppage of current involves almost
twice as little EGE as ECE. Gate-current measurements showed that it is less than
10−9 A/mm at 0 EGE 5 V/µm. The results attained show that the emitter
involved can be used as a basis for a complete analogue of the vacuum valve in
micro- and nano-implementation and, consequently, hardware components for
emission electronics.
A planar field emission cathode and a device on its basis are presented in
Fig. 5.19 [122].
Dielectric substrate 1 develops three-layer conducting structure 2–4 that cannot
but include layer 3 of catalytic materials such as Fe, Co, Ni etc., and upper layer 4
of a non-catalytic material. All end faces of the three-layer structure are coated with
dielectric 5, except for the end face overlooking anode 6. The end face of catalytic
layer 3 develops carbon nanotubes 7. The thickness of layer 3 is selected within the
range of l–20 nm.
The fixed thickness of the catalytic layer limits the diameter of the nanotubes to
be developed. Unlike an ordinary vertical emitter, its flat version produces the tubes
of the diameter not exceeding the thickness of layer 3.
Unlike the vertical design, the flat version provides a dense growth of nanotubes,
which does not severely limit the emissivity of the emitter since the nanotubes form
a thin blade that develops at the length of the nanotubes equal to 10–25 thicknesses
of layer 3, which steps up the electric field intensity near the edge of the blade. To
direct the growth of nanotubes in the plane of the catalytic layer, the latter can be
selectively underetched from the end face. In this case, the developing nanotubes
Fig. 5.19 Construction of a planar field electron cathode based on carbon nanotubes: 1 dielectric
substrate; 2, 4 conductive layers; 3 a layer of cathalytic material (Fe, Co, Ni etc.); 5 a layer of
dielectric; 6 anode; 7 carbon nanotubes; 8 shutter
5.4 Film Field Emission Cathodes 263
begin their growth in the plane of the catalytic layer, and the blade projects over the
end face, thereby sticking to the growth direction. This property is preserved if the
nanotube length does not exceed *10 thicknesses of the catalytic layer. If the
nanotube length considerably exceeds this limit, the blade turns into a fringe
consisting of nanotubes. At the catalytic-layer thickness of more than 40 nm and
the nanotube length of *3–10 µm, this fringe is well identified as a set of nan-
otubes coming out of the end face. In this case, it is easy to estimate the tube
diameter with an ordinary electron microscope and compare their diameter with the
catalytic-layer thickness. Structures with the catalytic-layer thickness of *40 nm
and the nanotube length of *5–10 µm were used as test structures to control
observance of the modes of normal formation of carbon nanotubes.
For wider functional use of the emitter, the three-layer structure is coated with
dielectric 5 with subsequent underetching from the open end face before devel-
opment of nanotubes. This is the way to fabricate the vacuum triode. With this end
in view, ordinary electron- or photolithography has to be used to form gate 8 on the
dielectric surface. In the design involved, the anode and the gate can be located at
the distance ranging from a tenth of a micron to a few microns from the blade,
which makes it possible to turn to development of vacuum micro- and nanoelec-
tronic devices.
Four emitters of 500 µm located in the corners of the square with a side of 5 mm
were measured in terms of voltage-current characteristics. The emission currents
spread ±0.25 µA/mm at the average current of 4.25 µA/mm and the applied field
intensity of 5 V/µm. The resulting spread shows that today it is quite realistic to
develop the technology of fabrication of emission integrated circuits in the area of
more than l cm2.
The technological implementation of such a cathode in form of carbon nan-
otubes grown up on a catalytic layer, combined with control electrodes was sug-
gested in [23]. In this version, electrons emitted from the nanotube field emission
cathode move out into the external space, e.g. to the anode of the display screen.
@n n @2n
¼A I Bn þ Ds 2 ; ð5:1Þ
@t Z @x
where n(x, t), cm2 is the surface concentration of adsorbed atoms on the lateral
surface of the growing whisker in point x at moment t, determined by the exchange
vapor-phase processes, on the one hand, and inflow/outflow of the substance during
surface diffusion, on the other hand; Z is the lattice-point density (cm−2); (I − n/Z)
is a defect fraction, so A − (I − n/Z) is the adsorption rate (A being the
particle-surface collision number (cm−2 s−1); Bn is the desorption rate; Ds ∂2n/∂x2
is the increment of the substance density due to surface diffusion, Ds being the
surface diffusion coefficient (cm−2 s−1).
The equation is solved for a stationary process when ∂п/∂t = 0 and the lateral
surfaces of growing whiskers are free from substance accumulation, which is
possible at small cover degrees and an inconsiderable flow of diffusing adsorbed
atoms. Then (5.1) is reduced to
@2n n A
D s 2 ¼ A I 1 ; n1 0 ¼ ; ð5:2Þ
@x n0 A=Z þ B
where n∞0 is the Langmuir adsorption isotherm, which is the solution of (5.1) at the
whisker length of l ! ∞. This equation admits elementary integration. If we take
into account that the stationary approach assumes the number of adsorbed atoms
getting to the tip and embedding into the lattice to be equal to the difference
between the number of the particles adsorbed and desorbed within the same time in
the region from x = 0 to x = l, then we will get that
dn
¼ n1
0 nl b thðblÞ; ð5:3Þ
dx x¼l
where
rffiffiffiffiffiffiffiffiffiffiffiffi
A 1
b¼ n : ð5:4Þ
Ds 0
5.5 Field Emission Cathodes Made of Whiskers 265
The material embedded into the whisker tip comes from two flows; one of them
is made of the particles condensed directly from the vapor at the end face, while the
other is caused by surface diffusion on the lateral crystal faces. In case of a suffi-
ciently thin apex, the direct condensation at the end face can be neglected; then the
growth rate of the whisker is inversely proportional to its radius, i.e.
dL 2J
thðblÞ; ð5:5Þ
dt rb
where J is the deposition rate during direct condensation on the crystal surface.
At the adsorption balance, the middle of a sufficiently long crystal should contain
the number of adsorbed atoms equal to that of the desorbed ones, i.e. n∞ 0 /ss. In case
of a thin adsorption layer, we have A = n∞ 0 /s s , so
rffiffiffiffiffiffiffiffiffiffiffiffi
I Ds 1 pffiffiffiffiffiffiffiffi
¼ n ¼ Ds s ¼ ks m; ð5:6Þ
b A 0
dL ks
¼ 3I ð5:7Þ
dt r
ks
L ¼ 2I t; ð5:8Þ
r
lg P ¼ 11:538 3886=T;
where T is the temperature of the vapor source in Kelvin degrees, and P (Torr) is the
W(CO)6 vapor pressure.
When the pressure selected within the range from 10−2 to 10−3 Torr is achieved
at the microduct outflow and a power supply, voltage rising to the discharge
ignition point (Uign), is connected to the electrodes via control resistance (Rctrl),
current is generated in the cathode-anode gap. Growth of the system necessitates an
apex to concentrate the discharge on the apex/substrate, there being a feeble glow in
the apex area. The type of glowing in the apex area suggests that the discharge
seems to be coronal. However, the range of pressures used in this case and strong
dependence of the current on the control resistance in the external circuit are
essentially different from the living conditions of the corona discharge. The main
5.5 Field Emission Cathodes Made of Whiskers 269
where n = 1−6, formed when thermal field electrons emitted by the pointed sub-
strate make an electron impact on the W(CO)6 vapor molecules. This process can be
represented by the equation
þn
WðCOÞ6 ! WðCOÞ6n þ nðCOÞ;
where n = 1−6; such ionization was established during mass spectrometer analysis
of products of tungsten hexacarbonyl decomposition by an electron impact [147,
148]. The way the material is fed to the growing apices is featured in Fig. 5.22. The
ions formed above the emitting apices/system branches move to them along the
electric field lines and deposit on the apices.
The apices are heated by the current running through them, as well as the
bombarding ions and the crystallization heat released when the metal atoms embed
into the branch. Herewith, the fragment tungsten-carbonyl ions deposited on the
apex/substrate decompose as a result of either thermal dissociation during
adsorption on the heated substrate accompanied by release of kinetic energy, or
bombardment by new ions.
Hence, the underpan forms excess free tungsten continuously fed to the sub-
strate. It can be presented as highly supersaturated metal vapor or a liquid metal
layer on the surface with a large supply of potential energy of crystallization
expressed by the difference of chemical potentials Δl = ll + lcr. That creates
conditions for growth of whiskers of the main trunk or bush branches.
The growth takes place on surface inhomogeneities, i.e. dislocations opening out
into the surface, a foreign-atom group or just microelevations of the substrate.
270 5 Field Emission Cathodes
Fig. 5.21 Images of growth of the system of tips at different points in time (in minutes) a 1, b 2,
c 3, d 4, e 5, f 6, g 7, h 8, i 9, j 10, k 15 min of growth
somewhat larger thickening at the tip, which usually generates two new whiskers at
a great rate. Their emergence is similar to original discharge of a bush trunk from
the substrate; however, we cannot determine this rapid growth rate even with
high-speed filming. The phenomenon of abnormally rapid growth rate of a whisker
has nowhere been mentioned before and needs more detailed studying within the
framework of the problem of the multiple-apex whiskerization and arborization
mechanism.
The whisker apices formed in the course of growth conduce to development of
new field emission centers, additional ionization of vapor molecules and domi-
nating deposition of tungsten atoms on these centers. The emitting apices grow in a
linear fashion. After achievement of the length per-selected for the chosen condi-
tions, the linear growth is slowed down and the tip accumulates the excess building
material. As a rule, this aggregation originates crystallization centers, which rapidly
develop into 2–3 new apices.
In their turn, these new apices also begin to emit, which causes the material to be
locally fed to them. At that, the whiskers formed earlier and serving later as a
substrate begin thickening, i.e. their radial growth begins. A trunk with new
branches growing from its tip is formed. The course of this process results in strong
ramification of the developing dendritic system (Fig. 5.21d–f), which in its turn
increments the total emitting area. The finite-size whisker (Fig. 5.21l) is *100 µm
in diameter and *50 µm high.
The effective emitting area calculated by Fowler–Nordheim volt-ampere char-
acteristics in the assumption of the average work function of eu = 4.5 eV grows
from 10–7 µm2 for initial microelevations, to 10–3 µm2 for a completely formed
dendritic structure.
272 5 Field Emission Cathodes
The total current from the dendritic structure reaches 500 µA at the anode
voltage of 400–500 V. Let us note that the total current from this structure is
comparable with the similar parameter of multiple-apex field emission cathodes of
much bigger sizes.
In the course of growth, the electron microscopical image of some thinnest
apices appears washed out (Fig. 5.21), which is most likely connected with their
oscillating motion. Apparently, origination of oscillations can be accounted for by
the electromechanical effect found out for whiskers in [151]. In our case,
mechanical oscillations of dendrites can be observed when current is taken off from
them and they are exposed to the same electric field as in [151]. When the anode
voltage is taken off, the oscillations cease, and the electron-microscopical image
becomes sharp and clear-cut.
It is worth noting that the most intensive emission of 100–200 µA and the
anode-cathode potential difference of 300–400 V, is inherent in the multiple-apex
systems formed under optimum conditions of formation of multiple-apex field
emission cathodes of the dendritic type Rctrl = 10–15 MX; Ip = 200–250 µA;
W(CO)6 vapor pressure of 1–2 10–3 Torr; growth time of 8–10 min.
Low voltage of dendritic field emitters is achieved by making a rather
branched-out bush of thin apices, which is in its turn achieved by the optimum
molecular flow of the building material during formation of microgeometry of
tungsten dendrites. At the same time, the discharge-burning voltage at the end of
formation correlates with the field emission excitation potential during the current
takeoff from the cathode. The reason thereof is that W(CO)6 vapors are ionized by
field electrons from the growing apices and the ions are fed to the most intensively
emitting regions of the dendritic system, providing development of these emitting
whiskers through sprouting of new apex branches, so-called emission centers, and
consequently, their subsequent progressive growth to a certain limit; that is why it is
the power failure in the discharge gap that is the field emission threshold. The
cathode-formation time is determined by the moment of achievement of the voltage
drop to 250–300 V, which amounts to the above-mentioned 8–10 min.
The growth controllability, expressed by direct interrelation of the formation
conditions, i. e. the discharge parameters at the end of the formation, and the
emissivity of the cathodes, provides achievement of high technological repro-
ducibility. Technological reproducibility of emissivity characteristics is provided
under the optimum conditions of formation irrespective of the formation-cell
geometry. The control-voltage spread for constant field emission current of 100 µA
was 10–30 V at the voltage of 350 V, i.e. 5–10%.
The cathode emissivity was tested with the cathode-circuit control resistance of
5–10 MX to prevent development of a vacuum arc. Emission instability, i.e. field
emission current oscillations, when the cathode functions in technical vacuum at the
fixed control voltage of 350 V and the anode current of 100 µA was 3.3%. At this
level, the emission remained steady throughout vacuum testing, *10−6 Torr for
*200 h, further testing being ceased.
5.5 Field Emission Cathodes Made of Whiskers 273
Figure 5.24 shows that the pore diameter depends on the electrolyte temperature
and linearly grows with the formation voltage in the examined voltage interval from
60 to 210 nm on average.
The idea to fabricate field emission nanoemitters by the galvanic method sug-
gested in papers [174, 175] involves use of porous anodic aluminum oxide with a
remote barrier layer as a mold for embedding of metal nanowires into the
pointed-emitter formation pores.
As porous anodic aluminum oxide is an almost perfect nanostructured dielectric
matrix with vertically-oriented cylindrical nanochannels/pores, a lot of researchers
focused on engineering of carbon nanotube growth in anodic aluminum-oxide
pores. Some of them [175, 176] demonstrated a possibility to grow
vertically-oriented carbon nanotubes in the pores of ordered structures of nano-
porous anodic aluminum oxide without specification of their emissivity
characteristics.
Another researcher [176] synthesized carbon nanotubes in an oxide membrane
with embedded nanowires. The feed material was aluminum foil, where porous
oxide was formed. Thereafter, the barrier anodic aluminum-oxide layer was
removed by dissolution in 20% sulphuric acid. Then the oxide membrane was
separated from the aluminum. After that, one side of the oxide membrane was
covered with a gold layer used as an electrode, and then the membrane was placed
in an electrolytic cell to electro-deposit nickel into the open pores on the other side
of the membrane. I.e., the initial-sample preparation technique was similar to that
mentioned before in [174, 175]. Then the substrate was cleaned in ionized water by
the ultrasonic method. After that, nickel nanowires embedded into the oxide
membrane were used as catalytic agents to grow carbon nanotubes by chemical
276 5 Field Emission Cathodes
test field emission images shows the effect of the carbon film on the
emission-current density homogeneity over the cathode surface [179]. However, it
is worth noting that despite the inducement to apply carbon films in field emission
microdevice fabrication, their practical use is hindered by high sensitivity of the
carbon films to any subsequent operations, which results in their loss of all the
above-stated properties. Another disadvantage of the carbon films is their high
deposition temperature, generally about 850 900 °C, which determines certain
requirements, most importantly high-temperature stability, to the involved sub-
strates and materials of the functional layers of the field emission devices.
The samples intended for outdoor testing under atmospheric pressure were
subjected to electrochemical polishing aimed at enlargement of the pores by etching
of the incoherent part of the oxide layer surrounding the pores in the oxide cells. As
a result of this operation, the thickness of the dielectric layer of the porous oxide
was reduced to 150 nm. The subsequent vacuum annealing of the surface was
followed by deposition of a double-layer thin-film structure consisting of a tantalum
substrate 10 of (10–20) nm in thickness and an aluminum layer 11 of 1 µm in
thickness. Sputtering of the pores during deposition of the double-layer thin-film
structure on the porous anodic oxide surface generates pointed emitters 12 with the
apex tips directed towards the porous anodic oxide (Fig. 5.25e). The experimental
estimates of the pore-sputtering depth showed that the deposited material runs into
the pores for the distances comparable with their diameter. Then the
aluminum-layer surface is covered with a photoresistant mask meeting the topo-
logical pattern of the upper electrode 13. The aluminum layer unprotected by the
photoresist is etched, the photoresist is removed, and the structure generated in this
way is heat-treated in the atmospheric air. As a result, the tantalum substrate
unprotected by the aluminum layer turns into thermal tantalum oxide 14
(Fig. 5.25f).
This metal-deposition flowchart involved the magnetron vacuum sputtering
method, which can be used along with any other method.
The nanostructural field emission matrices fabricated according to the flowchart
given in Fig. 5.25 were tested outdoors under atmospheric pressure as components
of a nanostructural two-electrode device. When the test device having an active area
of 1 mm2 was under the continuous current loading of 0.5 mA for 10 h, the
observable volt-ampere characteristic did not show any variation. In this case, the
emission-current oscillations at the running voltage of 2.7 V did not exceed 1%.
After outdoor storage under natural, i.e. normal climatic, conditions for one year,
the test device confirmed the initial volt-ampere characteristics.
At the inter-electrode dielectric thickness of 150 nm, the effective field emission
current begins at the running voltages of *2.5 V.
The inter-electrode dielectric resistance measured was about 5 1010 X. As a
result, the conductivity current through the dielectric layer of the device was about
0.2 nA, and its effect on volt-ampere characteristics can be neglected.
Some more details of fabrication of the devices based on field emission cathodes
with anodic aluminum oxide can be found in [153].
5.7 Other Field Emission Cathode Types 279
Vacuum border
such a structure by chloroform does not result in complete removal of the polymer
from the surface layer of the nanotubes, field emission failing to manifest itself even
at the fields of 12.5 V/µm. The chloroform is assumed to dissolve the polymer on
the surface, and then the polymer coats the nanotubes again after drying.
The best results were achieved when the samples were tested according to
process 2. The threshold field was 0.8 V/µm while pure nanotubes resulting from
process 1 generated the threshold field of 2.2 V/µm.
This additional increase in field emission is accounted for by the triple junction
between the nanotube/semi-metal, polymer/semiconductor and vacuum. A similar
mechanism was suggested to account for field emission of diamond films [182] and
nanocrystal graphite [183]. Figure 5.26 the surface potential changes the curvature
at the junction from the nanotube to the polymer due to the difference of the
electron work function of these materials. This non-uniform potential changes the
potential lines directly above the contact of the two materials. This jump in potential
above the junction between the materials results in a very heavy electric intensity.
So this case involves no morphological necessity to create the additional conditions
of increase in the electric field. Moreover, the strong electric field in the immediate
proximity from the surface junction runs into the conducting channels of the matrix,
which causes additional tunneling of electrons.
Paper [184] suggested a triode design using carbon nanotubes and a conductive
polymeric composite. The researchers resorted to electrochemical polymerization of
pyrrole and electrophoretic application of carbon nanotubes. The other elements of
the triode structure were fabricated by standard thin-film processes.
Field emission cathodes were fabricated from one-wall carbon nanotubes of
about 1.6 nm in diameter and 0.5–2.0 µm in length. The cleaned nanotubes were
mixed with pyrrole and lithium perchlorate used as polymeric monomer and
electrolyte, respectively.
Lithium perchlorate intensifies electrochemical polymerization and functions as
a conductive additive in the polymeric matrix. Concentration of carbon nanotubes
5.7 Other Field Emission Cathode Types 281
Spacer
Controlling
electrode
from Cr
ITO-
SiO2
cathode
a-Si
was 7.5 10−2 mg/ml, and that of pyrrole and electrolyte was 7 10−4 and
7 10−5 mol/ml, respectively.
The cross-section of the experimental structure is shown in Fig. 5.27.
The structure consists of an emitting spot and a control electrode of Cr. The
diameter of the composite apex made of a nanotube and a polymer, located also on
the conductive cathode, was 20 µm, and the diameter of the hole in the extraction
electrode was 30 µm. The insulator SiO2 is 1.5 µm thick. The amorphous silicon
layer (a–Si) as thick as 2500 Å functions as a resistive layer for the field emission
cathode.
The field emissivity was measured in a vacuum chamber at the pressure of
5 10–6 mm Hg in the pulse mode. At the anode voltage of 1000 V and the
extraction electrode voltage of 60 V, the emission current was 35 µA.
Fabrication of field emission cathodes from vertical tube bundles in a matrix with
cyanoacrylate was suggested in [185].
The carbon nanotube array was synthesized from the ethylene-argon mixture
with the iron catalyst by chemical deposition on the porous-silicon surface. Silicon
was lithographed over the iron-film surface from the electron beam through a mask.
The nanotubes were synthesized in the Ar/C2H4 mixture with the volume ratio of
3:1 at 700 °C. Figure 5.28 shows the sample photographed in a raster-type electron
microscope. The nanotube array was perpendicular to the substrate surface. The
total nanotube array area was about 2.3 cm2, and the nanotube height was con-
trolled by synthesis effect variation. Every cathode contains about 100 nanotube
bundles, the total diameter being about 100 µm (Fig. 5.28b).
Carbon nanotubes grow directly on a substrate to provide electrical contact.
The nanotubes are confused on the bundle surface and oriented on the lateral
surface.
Contact of carbon nanotubes with cyanoacrylate is provided by sticky
cyanoacrylate dripped on top of the nanotube array.
Capillary forces make cyanoacrylate promptly soak throughout the array.
Cyanoacrylate adhesion is not durable. Cyanoacrylate shrinks in the course of
282 5 Field Emission Cathodes
Analysis of atom power topography of the polymer film surface showed that the
polymer surface was considerably modified after studying of the field emission
effect, in particular apices of 400 nm in height appeared. Formation of polymeric
apices in a quite intense electric field can be understood as manifestation of local
electrization of polymer due to self-injection of charge carriers in a quite elastic
polymeric matrix from the metal electrode [191].
A reversing volt-ampere characteristic, VAC, measured at small voltages agrees
with the Fowler–Nordheim theory, which indicates the field emission nature of
work of the polymer-coated cathode. However, the voltage growth up to more than
6000 V reduces the current measured.
According to X-ray research, after an emitter has worked at the field intensity of
E > 10 kV/mm in the inter-electrode gap and the emission currents of I > 1 mA for
a few hours, the polymeric film develops a large number of carboniferous areas due
to polymer thermal degradation. That indicates the local heating of the polymeric
coating by emission currents of more than 700 °C. The researchers detected an
appreciable mass transfer of the carboniferous material to the anode, as well as
intermittent luminosity of the luminescent screen. Some emitting centers stopped
existing, others were formed again.
Paper [192] suggested organic nanowires for field emission cathodes of a larger
area.
They are represented by the organic complexes of Ag TCNQ and Cu TCNQ,
which are good at conducting electric charges (TCNQ = 7, 7, 8,
8-tetracyanoquinodimethane). They have a high density of charge carriers, which
results in high conductivity. The electrical resistance of these materials switches from
high to low within a few nanoseconds by application of the electric field. That
suggests that they can become a new class of materials for field emission cathodes.
The authors of [192] prepared the field emission cathode samples of a large area
(8 3.6 cm2, 14.3 3.4 cm2) on the substrates of silver and copper foil by the
organic vapor-solid body method. This method gave good results and easy uniformity
control.
Practically, the TCNQ powder is loaded into a ceramic boat and placed in the
center of the quartz tube in the horizontal furnace. The copper or silver foil is placed
on the top of the ceramic boat.
The furnace temperature is stepped up to 120 °C for copper and 150 °C for
silver and then maintained for 2 h. The TCNQ is sputtered on the foil surfaces
where the films of Cu TCNQ and Ag TCNQ are formed, respectively. The length of
nanowires is a few tens of microns and the diameter is about 40–150 nm. The
diameter of the resulting nanofibers is one and a half times more at the foundation
than at the top.
Both types of the field emission cathodes have good field emissivity. In par-
ticular, the threshold electric field values necessary to generate the current density
of 10 mA/cm2 were 2.58 V/µm for Ag TCNQ and 3.13 V/µm for Cu TCNQ. The
difference in the threshold field value can be accounted for by different electron
affinity between Ag TCNQ and Cu TCNQ.
5.7 Other Field Emission Cathode Types 285
References
1. E.E. Martin, J.K. Trolan, W.P. Dyke, Stable, high density field emissiom cold cathode.
J. Appl. Phys. 31(5), 782–789 (1960)
2. W.P. Dyke, F.M. Charbonnier, R.W. Strayer, R. Floyd, J.P. Barbour, J.K. Trolan, Electrical
stability and lie of the heated field emission cathode. J. Appl. Phys. 31(5), 790–805 (1960)
3. M.I. Elinson, G.F. Vasiliev, Way of reduction of ionic bombing, emitting electrons metal
more sharply. A. S. USSR. No. 107388 of Cl.H01j 1/30 of 24.09.55
4. W.P. Dyke, Progress in electron emission at high fields. Proc. IRE 43(2), 162–167 (1955)
5. M.I. Elinson, G.A. Kydinceva, Field emission cathodes on the basis of metal similar
refractory connections. Radio Technician Electron. 7(9), 1511 (1962)
6. M.I. Elinson, G.V. Vasiliv, Field emission. GIFML (1958)
7. V.V. Chesnokov, Electron lamps with field emission cathodes. A. S. USSR. No. 314454 of
27.12.65
8. V.V. Chesnokov, V.M. Gayles, N.A. Morozova, Receiving-amplifying lamp with the cold
field emission cathode. A.S. USSR. No. 339986
9. R. Fisher, H. Noyman, Field emission of semiconductors. Science (1971)
10. N.V. Egorov, Experimental dissertation research of field emission of silicon. Abstract of the
thesis. I. LIE. Leningrad (1973)
11. E.P. Sheshin, Structure of a surface and field emission properties of carbon materials.
M. MFTI. Fizmatgiz (2001)
12. D. Alpert, R.S. Burits, Ultra-vacuum. D. limiting factors on the attainment of very low
pressures. J. Appl. Phys. 25(2), 202 (1954)
13. I.L. Sokolskay, Surface migration of atoms of tungsten in electric field. ZhTF 26(6), 1177
(1956)
14. G.N. Fursey, G.K. Karcev, The stability of field emission issue and migratory processes
preparing development of a vacuum arch. ZhTF 40(2), 310 (1970)
15. P. Staib, K. Ulmer, Maintenance in vacuum of clean metallic surfaces. Phys. Stat. Solidi. 32
(2), K163–K165 (1969)
16. G. Krause, D. Stark, Field emission of chromium, cobalt and hafnium (zur feldemission top
cjarom, ilobalt und hafnium), Z. Phys. 201(1), 69 (1967)
17. A.P. Komar, U.N. Talanin, Pictures of field emission of crystals of carbides of tungsten and
molybdenum. Jnt. AS Suur. Ser. fiz. 22(5), 580 (1958)
18. M.I. Elinson, G.F. Vasiliev, Research of field emission of a geksaborid of lanthanum. Radio
Technician Electron. 3(7), 945 (1958)
19. M.I. Elinson, G.F. Vasiliev, Experemental research of field electronic issue of a geksaborid
of lanthanum. Radio Technician Electron. Eng. 2(3), 348 (1957)
20. YuV Zybenko, I.L. Sokolskay, Field electronic emission of carbonized and thoriated
carbonized tungsten. Radio Technician Electron. Eng. 5(8), 1327 (1960)
21. G.A. Kydinceva, Application of refractory connections as field electronic cathodes. Question
Radionics S1 Electron. 3, 75 (1961)
22. V. Espe, Technology of electrovacuum materials, Energy 2 (1968)
23. E.I. Davydova, F.D. Kirichenko,V.A. Shishkin, Stability of field emission tip cathodes
passivated by films of transitional metals. ZhTF 49(11), 2344–2348 (1979)
24. M.I. Elinson, A.G. Gdan, Field electronic emitter. A. S. USSR. №118913. Kl. HOlj 1/30 of
28.02.58
25. M.S. Mousa, Characteristics of tungsten substrate with Al2O3 coatings under UHV
conditions. Vacuum 38(8–10), 835–838 (1988)
26. M.S. Mousa, Field electron emission studies on zinc oxide coated tungsten microemitters.
Surf. Sci. 266, 110–120 (1992)
27. V.V. Zhirnov, W.B. Choi, I.I. Cuomo, I.I. Hren, Diamond coated Si and Mo field emitters:
diamond thickness effect. Appl. Surf. Sci. 94/95, 123–128 (1996)
References 287
28. W.B. Choi, I.I. Cuomo, V.V. Zhirnov others, Field emission from silicon and molybdenum
tips coated with diamond powder by dielectrophoresis. Appl. Phys. Lett. 68(5), 720–722
(1996)
29. A.V. Crewe, Electron gun using a field emission source. Rev. Sci. Inst. 39(4), 576 (1968)
30. N. Shogo, K. Tsukasa, Remolding of tungsten field emitter. Jap. J. Appl. Phys. 11(4), 604–
605 (1972)
31. I.W. Wiggins, I.A. Zubin, M. Beer, High-resolution scanning transmission electron
microscope at Ions Hopkins. Rev. Sci. Instrum. 50(4), 403–410 (1979)
32. G.N. Fursey, S.N. Shakirova, To a question of possibility of localization of an field emission
in small space angles. ZhTF 33(6), 1125 (1966)
33. L.W. Swanson, Comparative study of the zincroniated and built-up W. thermal-field
cathode. J. Vac. Sci. Tech. 12(6), 1228 (1975)
34. T.A. Baker, M.M. Balsiger, K.T. Considine, H.E. Litsjo, Electron-beam tube including a
thermonic-field emission cathode for a scanning electron microscope. Pat. №3809899 (USA)
35. S. Hosoki, S. Yomamoto, H. Todokoro others, Field emission cathode and method of
fabricating the same. Pat. USA. №4379250. CL. 313/336 (HOl l/16) of 5.04.1983
36. E.I. Davydova, A.D. Karpenko, V.A. Shishkin, B.I. Pab, A.c. №439028 (USSR). Way of
production of field emission cathodes. 1974. No. 29
37. G.A. Kudintseva, V.P. Savchenko, B.I. Pab, A.c. №424254 (USSR). Way of production of
needle field emitters. 1974 No. 14
38. E.A. Litvin, A.A. Starobinec, Limit currents of field emission. ZhTF 47(10), 2032 (1977)
39. A.D. Karpenko, U.S. Lobachev, V.A. Shishkin, About limit current of the one tip field
emission cathode. Radio Technician Electron. Eng. 21(1), 207 (1976)
40. M.I. Elinson, V.A. Gorkov, A.A. Yaciopolskay , G.A. Kydinceva, Research of pulse field
emission at the high density of currents. Radio Technician Electron. Eng. 5(8), 1318 (1960)
41. J.P. Barbour, W.W. Dolan, J.K. Trolan, E.E. Martin, W.P. Dake, Space-charge effects in
field emission. Phys. Rev. 22(1), 53 (1953)
42. T.J. Lewis, Theoretical interpretation of a field emission periments. Phys. Rev. 101(6), 1694
(1956)
43. E.G. Shirokov, Field emission of system tungsten tips. Radio Technician Electron. T9(7),
p1320 (1964)
44. A.A. Golov, G.N. Fursey, I.D. Ventova others, Field emission cathode type “comb”. A.S.
USSR. No. 342241. C. HOlj 1/30 of 14.06.72
45. V.M. Gykov, D.M. Paytov, S.A. Polegaev, other, Way of production of the Many tip field
emission cathode. A.S. USSR. No. 1001225. Cl.HOlj 9/02 of 28.02.83
46. P. Genequand, Pat. Switzerland. №603005. C. HOlj 1/30 of 15.08.78
47. I.D. Ventova, L.E. Valuyeva, B.I. Pab, A. S. №416784 (USSR), Way of production of many
tip field emission cathode. 1974, No. 7
48. V.G. Pavlov, A.S. Rabinovich, V.L. Savchenko, V.N. Shrednik, B.I. Pab, A. S. №493834
(USSR), Way of blunting tip cathode. 1975. No. 44
49. B.V. Okylov, I.Z. Gleyzer, L.T. Dronova, Manufacturing techniques of multiemitter cold
cathodes. PTE, (6), 124–125 (1973)
50. J. Shelton, Field effect electron emitter. Pat. USA. №3745402. Cl. 313/309 (HOlj 1/30) of
10.07.1973
51. I. Shelton, Field effect electron gun having at least a million emitting fibers per square
centimeter. Pat. USA. №3783325. C. 313/336 (HOlj 1/02) of 1.01.1974
52. I. Shelton, Edge effect elimination and beam forming designs for field emitting arrays. Pat.
USA. №434181. Cl. 313/309 (HOlj 1/02) of 17.08.82
53. W.L. Less, Method and apparatus for liming-emission current. Pat. USA. №3671798. Cl.
313/336(HOlj 1/16) of 20.06.1972
54. K. Inoue, K. Bersui, T. Nakatani, O. Toyoda, Field emission cathode and process for
producing the same. Pat. USA 2003/0184203 Kl. 313/309 (Hoij 9/04 of 2.10.2003)
55. V.F. Bibik, P.G. Borzak, A.F. Yshenko, Photofield cathode. A. S. USSR. №584667. Cl.
Ho1j 9/09
288 5 Field Emission Cathodes
56. R.N. Thomas, R.A. Wickstrom, D.K. Schroder, H.C. Nathanson, Fabrication and some
applications of large-area silicon field emission arrays. Solid State Electron. 17, 155 (1975)
57. A.I. Klimin, A.A. Mostovsky, I.A. Pystylnik, D.A. Sakseev, L.P. Titova, N.M. Eden, Field
emission research one and many tips silicon catode. Izv. Acad. Sci. USSR. Ser. fiz. 40(8),
1575 (1976)
58. V.F. Bibik, P.G. Borzak, A.F. Ycenko, Germanium and silicon field emission photo cells.
Ukr. Phys. Zh. 13(5), 868 (1968)
59. X.S. Natanson, Y. Goldberg, Matrixes tips for field emission systems.:fizik of thin films, 8
(M.:mir, 1978), p. 264
60. D.N. Schroder, R.N. Thomas, J. Vine, H.C. Hathanson, The semiconductor field emission
photocathodes. JEEE Trans. Electr. Devices 21(12), 785–798 (1974)
61. R.N. Thomas, H.C. Hathanson, Photosensitive Field emission from silicon point arrays.
Appl. Phys. Lett. 21(8), 384–389 (1972)
62. R.Z. Bahtizin, E.K. Ratnikova, V.P. Petrakov, Way of production of the many tips field
emission cathodes. A. S.. USSR No. 1215536. Cl. HOij 1/30
63. R.Z. Bahtizin, E.K. Ratnikova, Edektronny equipment. Ser. 2. Semicond. Device 7(173),
124 (1984)
64. R.Z. Bahtizin, E.K. Ratnikova, G.G. Kylishova, A.F. Yshenko, V.P. Petrakov, Production
germanium many tips structures and research of their emission characteristics. News Acad.
Sci. USSR. 49(9), 1738–1740 (1985)
65. R.Z. Bakhtizin, Field emission from semiconductor point arrays. I. De Physique (Colloque
P. 6). 49(11), 155–160 (1988)
66. P.G. Borzak, E.I. Givargizov,G.G. Kylishova, I.E. Lifshic, A.N. Stepanova, A.F. Yshenko,
News of academy of sciences of the USSR. Ser. Fiz. 40(9), 1571 (1976)
67. C.A. Spindt, K.R. Shoulders, L.N. Heynick, Field emission cathode structures and devices
utilizing such structures. Pat. USA. №3755704. Cl. 313/309 (Hoij 1/16 of 28.08.1973)
68. C.A. Spindt, K.R. Shoulders, L.N. Heynick, Field emission cathode structures and devices
utilizing such structures, and methods of producing such structures. Pat. USA. №3789471.
Cl. 29/25.17 (Hoij 9/02 of 5.02.74)
69. C.A. Spindt, L.N. Heynick, Method of producing field ionizer and field emission cathode
structures. Pat. USA, №3812559 Kl. 29/25.18 (Hoij 9/02 of 28.05.74)
70. C.A. Spindt, I. Brodie, L. Humphrey, E.R. Westerberg, Physical properties of thin-film field
emission cathode with molybdenum cones. J. Appl. Phys. 47(12), 5248–5263 (1976)
71. C.A. Spindt, in Field emission arrays for high-current density, and high frequency
operation. Teth. Digest of 18th. IVNS. (Oxford, UK, 2005), p. 20–21
72. I.M. McCaulay, C.A. Spindt, C.E. Holland, I. Brodie, Field emission cathode array coated
with electron work function reducing material, and method. Pat. USA. №5089292. Kl.
427/78. (BO5D 5/12 of 18.02.1992)
73. M.S. Mousa, C.E. Holland, J. Brodie, C.A. Spindt, The effect of hydrogen and acetylene
processing on microfabricated field emitter arrays. Appl. Surf. Sci. 67, 218–221 (1993)
74. M.S. Mousa, A study of the effect of hydrogen plasma on microfabricated field- emitter
arrays. Vacuum 45(2/3), 235–239 (1994)
75. H. Li, M.Q. Ding, J.J. Feng, X. Li, G. Bai, F. Zhang, in Fabrications and characterization of
spindt-type field emission arrays coated with ZrC thin films. Thech. Digest IVNC. (Oxford,
UK, 2005), p. 237–238
76. C.C. Chang, Field emission display cathode plate with an internal via and the fabrication
method for the cathode plate. Pat/ USA 2002/0105261. Cl. 313/495 H01j 63/04 of 8.08.2002
77. L. Dvorson, A.J. Akinwande, Double-gated Spindt emitters with stacked focusing electrode.
J. Vac. Sci. Tech. B 20(1), 53–59 (2002)
78. L.Y. Chen, A.I. Akinwande, in Negative resistance region in double-gated silicon field
emission arrays. Tech. Digest JVNC. (Oxford, UK, 2002), p. 15–16
79. Y. Gotoh, M. Nagao, D. Nozaki others, Electron emission properties of Spindt-tipe platinum
field emission cathodes. J. Appl. Phys. 95(3), 1537–1549 (2004)
References 289
80. M.Q. Ding, X. Li, G. Bai other, Spindt-type cathodes with aligned carbon nanotube emitters.
JVESC 117–119 (2004)
81. L.R. Baylor, D.H. Lowndes, M.L. Simpson others, J. Vac. Sci. Technol. B. 20, 2646 (2002)
82. X. Yang, Simpson M.L., Rondolph S.J others, Integrated tungsten nanofiber field emission
cathodes selectively grown by nanoscale electron beam-induced deposition. Appl. Phys. Let.
86, 183106 (2005)
83. A. Jankowski, J.P. Hayes, Process system and method for fabricating field emission cathode.
U.S. patent №5746634 of 5.05.1998
84. A. Jankowski, J. Hayes, J. Morse, J. Ferreira, Large area deposition of field emission
cathodes for flat panel displays. Thin Solid Films 355–356, 194–198 (1999)
85. Y-F. Bai, S. Ren, S.Z. Deng, J. Chen, N-S. Xu, in Preparation of ZnO nanowires and there
field emission properties. Tech. Digest of JVNC. (Oxford, UK, 2005), p. 178–179
86. C.Y. Li, X. Liu, J. Chen, S.Z. Deng, N.S. Xu, S. Yang, in Uniform field emission from CuO
nanowires prepared by thermal oxidation method. Tech. Digest JVNC. (Oxford, UK, 2005),
p. 213–214
87. P. Feng, X.Q. Fu, S.Q. Li, Y.G. Wang, T.H. Wang, Stable electron field emission from
triangular-shaped ZnO nanoplate arrays with low local heating effect//Nanotechnology.2007.
V.18.165704
88. A.J. Pedzaza, J.D. Fowlkes, D.H. Lowndes. Appl. Phys. Lett. 77, 3018 (2000)
89. A.V. Karabutov, V.D. Frolov, E.N. Loubnin, A.V. Simakin, G.A. Stapheev. Appl. Phys.
A 76, 413 (2003)
90. V. Zorba, P. Tzanetakis, C. Fatakis others, Silicon electron emitters fabricated by ultraviolet
laser pulses. Appl. Phys. Lett. 88(08), 103 (2006)
91. Zorba V., I. Alexandrou, I. Zergioti others, Thin Solid Films. 453–454, 492 (2004)
92. V.S. Fomenko, Emission properties of materials. Naukova Thought, Kiev (1981)
93. I.N. Slivkov, Electrical insulation and the blunting in vacuum. Atomizdat (1972)
94. W.D. Dyke, Field emission, a newly practical electron source. IRE Trans. Mil. Electron.
MIZ 4(1), 38 (1960)
95. E.G. Shirokov, Field emission cathode. C.S. №180712. in Opening, Inventions, Industrial
Samples, Trademarks, 8, 52 (1966)
96. E.G. Shirokov, Technique of preparation and control emission surface of a tungsten edge of
a disk-shaped form. Изв. Acad. Sci. USSR Ser. Tekh. Nauk. 6(2), 45 (1965)
97. B.V. Bondarenko, V.I. Makyha, The field emission cathode from steel edges. PTE (4), 235–
236 (1965)
98. B.V. Bondarenko, V.I. Makyha, A.S. Gaydarov, Research field electronic edge of emitters of
a disk-shaped form. Radio Eng. Electron. 17(12), 2634–2637 (1972)
99. V.I. Mahov, N.A. Dugev, A.I. in Kozlov, Cathodic Knot With Silicon Matrix Field Emission
Cathodes. The xx vsesoyuzn. konf. on emission electronics: Tez.dokl. Kiyev, (1987) p. 232
100. A.A. Vasenkov, N.A. Dughev, V.I. Mahov, Low-voltage silicon matrix field 1 emission
cathodes. Electron. Indus. (8), 48–49 (1987)
101. N.A. Dughev, A.I. Kozlov, V.I. Mahov, V.M. Serovayskiy, Speed of operation microelec-
tronic triode structures with field emission cathodes. Microelectronics 19(5), 478–485 (1990)
102. V.I. Volosov, V.N. Lazarev, V.E. Teraev, About some features of field emission of
cylindrical cathodes. ZhTF 40(4), 855 (1970)
103. V.A. Kyznetsov,A.N. Kyrochkin, Field e emitter. A. S. USSR №1078492. Cl. H01j 1/30, of
07.03.84
104. V.V. Chesnov, Electronic equipment. Series. (4), 3 (1968)
105. R.L. Akopyn, B.V. Bondarenko, V.I. Makyha, Radio Technician Electron. 16(11), 2236
(1971)
106. B.V. Bondarenko, R.L. Akopyn, Calculation of emission characteristics of film field
emission cathodes. Radio Technician Electron. Eng. 17(5), 1059 (1972)
107. B.V. Bondarenko, Ways of increase of stability of emission and life of field emission
cathodes. Electronic equipment. Series 1(6), 74–82 (1973)
290 5 Field Emission Cathodes
108. B.V. Bondarenko, V.I. Makyha, About durability of the film emission cathode from chrome.
Radio Technician Electron. Engineer. XVII(8), 1683–1686 (1972)
109. E.A. Gulbranson, K. Andern, J. Electrochem. Soc. 99, 402 (1959)
110. B.V. Bondarenko, V.I. Makyha, A.C. Gaydarov, The film emission cathode from the titan.
Radio Technician Electron. 18(10), 2212 (1973)
111. B.V. Bondarenko, L.A. Kirichenko other, Research of field emission foil and films and the
mechanical durability of microinhomogeneities on their surface. Summary of reports of
All-Union conference on issue electronics. Makhachkala (1976)
112. B.V. Bondarenko, L.A. Kirichenko, N.D. Konovalov, Field emission thin foil refractory
metals. Theses of reports of the All-Union symposium on not heated cathodes. Tomsk
(1980)
113. B.V. Bondarenko, V.I. Makyha, A.C. Gaydarov, Research of pulse field electronic emission
of film cathodes from chrome. Radio Eng. Elektronika 17(8), 1771–1772 (1972)
114. V.V. Chesnokov, Ways of production of field emission film electrode system. A. S. USSR
No. 174727. Cl. H01j/30, of 08.07.64
115. N.A. Dughev, A.I. Kozlov, A.A. Kadnikov, V.I. Mahov, I.V. Panaev, Influence of
geometrical parameters of silicon film field emission cathodes on their emission character-
istics. Radio Technician Electron. (11), 2385–2389 (1990)
116. L.D. Karpov, Field-effect emitter device. European patent №0681311. Cl. H01j/30, of
08.11.95
117. L.D. Karpov, Edge emitter display device. Pat. USA. №5965971. Cl. 313/309 (H0Ij 01/02),
of 12.10.1999
118. L.D. Karpov, Edge emitter with secondary emission display. Pat. USA. № 6023126. Cl.
313/310(H01j 1/05, of 8.02.2000
119. S.A. Gavrilov, E.A. Ilichev, E.A. Poltorackiy, N.V. Syetin other. The lateral emitter as basic
Element of integrated emission electronics. Lett. ZhTF 30(11), 48–53 (2004)
120. S.A. Gavrilov, N.N. Dzbanovsky, V.V. Dvorkin et al., in Proceedings of 11th International
Symposium, Nanostructures: Physics and Technology, p. 234–236, St. Petersburg, Russia.
23–28 June 2003
121. S.A. Gavrilov, N.N. Dzbanovsky, V.V. Dvorkin et al. in Abstract International Conference
«Micro-and nanoelectronics-2003», p. 01–27, Moscow-Zvenigorod, Russia, 6–10 October
2003
122. S.A. Gavrilov, E.A. Ilichev, E.A. Poltorackiy, N.V. Syetin, Emitters from carbon nanotubes
for planar emission vacuum micro and a nanoelectronics. Lett. ZhTF 30(14), 75–81 (2004)
123. Givargizov E.I. Growth of threadlike and lamellar crystals from steam. Science, 304c (1977)
124. G.V. Bereghkova, Threadlike crystals. M.: Science, 160c (1969)
125. S.A. Ammer, V.S. Postnikov, Threadlike crystals (Polytechnic University, Voronezh, 1974),
p. 284c
126. V.G. Syrkin, Gas-phase metalliztion through carbonyls. Metallurgy, 264c (1985)
127. V.G. Syrkin., Carbonyls of metals. Khimiya, 200c (1983)
128. B.G. Gribov, G.A. Domrachev, B.V. Ghyk other. Sedimentation of films and coverings
decomposition of metalloorganic connections of M of a.:nauk, (1981), p. 322
129. A.A. Nosov, T.A. Poshehonova, P.V. Poshehonov, About the education mechanism the
filiform of monocrystals on electrodes of electronic devices. RVE, 18(9), 199–1999 (1973)
130. L.V. Poshehonov, D.K. Nosova, A.A. Nosov, in Growth of Filiform Crystals of an Oxide of
Chrome. Materials of the All-Union Conference “Filiform Crystals for New Equipment”.
Voronezh VPK, 73 (1979)
131. R. Gomer, Field emission from mercury whiskers. J. Chem. Phys. 28(3), 457–469 (1958)
132. A.A. Nosov, T.A. Poshehonova, P.V. Poshehonova, Investigation of conditions of
destruction of filiform monocrystals in high-voltage vacuum devices at influence strong
electric fields. 15(10), 2156–2161 (1970)
133. P.V. Poshehonov, V.M. Gennadiev, N.P. Ovsynnikov, Methods of a increasing of efficiency
of training of field emission cathodes on the basis of threadlike monocrystals. RIE 21(12),
2568–2573 (1976)
References 291
134. P.V. Poshehonov, N.P. Ovsannikov other, Investigation of process of production of effective
field emission cathodes when forming many tips system of monocrystals of gold out of the
working volume of the device. In kN.:tex. rep. An all-Union seminar on not heated cathodes,
Tomsk, p. 33 (1977)
135. P.V. Poshehonov, N.P. Ovsannikov, V.M. Genadiev, N.N. Chadaev, Influence of contact
with the gas environment on emission ability of the field emission cathode formed by system
of filiform crystals. Vacuum and gas-discharge electronics, Ryazan, RRTI, p. 73–77 (1981)
136. E.I. Givargizov, G.A. Gydingeva other, Many tips field emission cathodes from a
single-crystal geksaborid of lanthanum. In kN.:tes.rep. Hush Vsesoyuznyu Konferentsii on
emission electronics, M.:nauka, p. 234 (1981)
137. N.G. Borsak, E.I. Givargizov other, Emission properties the many tips of field emission
cathodes six Gei-x from the made crystallization from a steam phase. Izv. Acad. Sci. USSR
40(7), 1571–1580 (1976)
138. F. Okuyama, Vapor-grown tungsten whiskers induced by vacuum discharges. J. Appl. Phys.
45(10), 4239–4241 (1974)
139. H.B. Linden, H.D. Beckey, F. Okuyama, On the mechanism of cathodio growth of tungsten
needles by decomposition of hexacarbonyl tungsten under high field conditions. Appl. Phys.
22(1), 83–87 (1980)
140. F. Okuyama, T. Shibata, I. Yasuda, tungsten needle produced by decomposition of
hexacarbonyl tungsten. Appl. Phys. Lett. 35(1), 6–7 (1979)
141. F. Okuyama, Growth of metallio w wiskers crystals incorporated with field electron
emission. Appl. Phys. Iott. 36(1), 46–47 (1980)
142. F. Okuyama, Cathodio needle growth from Mo (Co)6 and Cr(Co)6 vapors at lower electric
fields. Appl. Phys. 22(1), 39–46 (1980)
143. F. Okuyama, Cathodio needle growth from Mo(Co)6 vapors at higher electric fields. Appl.
Phys. 27(1), 57–64 (1982)
144. F. Okuyama, Growth of Cr needle crystals induced by field electron emission. J. Appl. Phys.
22(2), 245–251 (1984)
145. A.A. Nosov, N.P. Ovsannikov, G.N. Shyppe, Direct electron microscopy supervision of
dynamics of formation of the many tips field emission cathode on the basis of tungsten.
ZhTF 54(2), 372–374 (1984)
146. V.G. Syrkin, Carbonyl metals. M. Metalurgizdat, p. 255 (1978)
147. V.E. Ivanov other, Crystallization the refractory of metals from a gas phase. M. Atomizdat of
1974, p. 212
148. T.A. Razyvaev other, Metalloorganic connections in electronics. M. Sci. 479 (1972)
149. V. Barton, N. Rabrera, F. Frank, Elementary processes of growth of crystals (1959)
150. R. Vagner, Growth of crystals on the PKR mechanism. Monokristalny fibers and materials
reinforced by them. M.:Mir, pp. 42–57 (1973)
151. M.M. Pogorelskiy. Lett, ZhTF 4, 1068 (1978)
152. A.A. Nosov, D.I. Nosova, N.P. Ovsannikov, N.N. Chadaev, Limit density of field emission
current of tip emitter of submicron dimension which are grown up in various ways. Radio
Technician Electron. XXXI(7), 1466–1467 (1986)
153. N.I. Tatarenko, V.F. Kravchenko. Field emission nanostructures and devices on their basis.
M. Fizmatgiz (2006)
154. V.A. Labynov, B.A. Sokol, N.I. Tatarenko, The electrovacuum device. A. S. USSR
No. 713386. H01j 21/10 of 05.10.1979
155. N.I. Tatarenko, in Development of an Integrated Technique for Fabricating Thin-Film Field
Emission Microelectronic Devices. The Third International scientific and technical
conference: Modern technology of hybrid integrated chips, including elements of
superconductor electronics. The Belorussky state univrsitt informatics and radio electronics.
Naroch, Conference Materials, pp. 159–160 (1994)
156. N.I. Tatarenko, Way production of the thin-film vacuum microdevice. Patent USSR
No. 1729243. HO1j 21/10 of 22.12.1991
292 5 Field Emission Cathodes
157. F. Keller, M.S. Hunter, D.L. Robinson, Structural features of oxide coatings on aluminum.
J. Electrochem. Soc. 100(9), 411–419 (1953)
158. J.P. OSullivan, G.C. Wood, The morphology and mechanism of formation of porous anodic
films on aluminium. Proc. R. Soc. London A 317, 511–643 (1970)
159. G.E. Thompson, Porous anodic alumina: fabrication, characterization and applications. Thin
Solid Films 297, 192–201 (1997)
160. A.P. Li, F. Mueller, A. Birner, K. Nielsh, U. Gosele, Hexagonal pore arrays with a 50–
420 nm interpore distance formed by self-organization in anodic alumina. J. Appl. Phys. 84
(11), 6023-6026 (1998)
161. H. Masuda, M. Nakao, T. Tamamura, H. Asoh, Fabrication of ordered nanostructure based
on anodic porous alumina. The SPIE Conference on Micromachining and Microfabrication
Process Technology IV.-SPIE Vol.3511,1998.-P.74–81
162. A.P. Li, F. Mueller, A. Birner, K. Nielsh, U. Gosele, Fabrication and mi-crostructuring of
hexagonally ordered two-dimensional nanopore arrays in anodic alumina. Adv. Mater. 11
(6), 483–487 (1999)
163. D. Crouse, Y.H. Lo, A.E. Miller, M. Crouse, Self-ordered pore structure of anodized
aluminum on silicon and pattern transfer. Appl. Phys. Lett. 76(1), 49–51 (2000)
164. L. Ba, W.S. Li, Influence of anodizing conditions on the ordered pore formation in anodic
alumina. J. Phys. D: Appl. Phys. 33, 2527–2531 (2000)
165. H. Assoh, K. Nishio, M. Nakao, T. Tamamura, H. Masuda, Conditions for fabrication of
ideally ordered anodic porous alumina using pretextured Al. J. Electrochem. Soc. 148(4),
B152-B156 (2001)
166. H. Masuda, M. Yotsuya, M. Asano, K. Nishio, M. Nakao, A. Yokoo, T. Tamamura,
Self-repair of ordered pattern of nanometer dimensions based on self-compensation
properties of anodic porous alumina. Appl. Phys. Lett. 78(6), 826–828 (2001)
167. C.Y. Liu, A. Datta, Y.L. Wang, Ordered anodic alumina nanochannels on
focused-ion-beam-prepattened aluminum surfaces. Appl. Phys. Lett. 78(1), 120–122 (2001)
168. A. Cai, H. Zhang, H. Hua, Z. Zhang, Direct formation of self-assembled nanoporous
aluminium oxide SiO2 and Si substrates. Nanotechnology 13, 627–630 (2002)
169. W.L. Xu, M.J. Zheng, S. Wu, W.Z. Shen, Effects of high temperature annealing on structural
and optical properties of highly ordered porous alumina membranes. Appl. Phys. Lett. 85
(19), 4364–4366 (2004)
170. N.V. Myung, J. Lim, J.P. Fleurial, M. Yun, W. West, D. Choi, Alumina nanotemplate
fabrication on silicon substrate. Nanotechnology 15, 833–838 (2004)
171. N.I. Tatarenko, T.N. Andrushenko, Influence of conditions of anodizing on a microstructure
of porous anode oxide films of aluminum. Prot. мeтaллoв (3), 499–501 (1984)
172. V.A. Labynov, N.I. Danilovich, A.C. Yksysov, V.E. Minaychev, Modern magnetron
sputtering devices. Foreign Electron. Equipment 10, 3–62 (1982)
173. A.N. Govyadinov, S.A. Zakhvitcevich, in Field Emitter Arrays Based On Natural
Seiforganized Porous Anodic Alumina. 10th International Vacuum Microelecronics
Conference. Kyongju, Korea, pp. 735–737 (1997)
174. S. Zakhvitcevich, A. Govyadinov, Nanodimensional field emitter arrays based on porous
anodic alumina. Phys., Chem. Appl. Nanos-truct. 297–300 (1997)
175. E.R. Holland, M.T. Harrison, M. Huang, P.R. Wilshaw, Nonlithographic technique for the
production of large area high density gridded field emission sources. J. Vac. Sci. Technol.
B 17(2), 580–582 (1999)
176. R.R. Greene, H.F. Gray, P-N junction controlled field emitter array cathode. United States
Patent № 4 513 308, Int. Cl. H01L 29/06- H01L 29/34; H01L 27/12, Apr.23,1985
177. N.I. Tatarenko, A.A. Blyablin, N.V. Suetin, in Field Emission Nanostructures Formed By
Self-Organization In Anodic Porous Alumina. 12th International Vacuum Microelectronics
Conference.-Darmstadt, Germany, (1999), pp. 136–137
178. A.A. Blyablin, N.V. Suetin, D.B. Suyatin, E.S. Soldatov, N.I. Tatarenko, in STM u ATM
Study of a Nanoscale Field Emission Array Formed by Self-Organization in Porous Anodic
References 293
Aluminium Oxide. 2nd Europen Field Emission Workshop. Seqovia, Spain, (2000), pp. 176–
177
179. N.I. Tatarenko, A.O. Vorobyev, Volt-ampere characteristics and differential parameters
nanostructural field emission of microdevices. Electromag. Waves Electron. Sistemy 10(9),
68–74 (2005)
180. X.B. Yan, S. Xu, H.W. Liu, S.R. Yang, Field emission properties of polymer-converted
carbon films by heat treatment. Solid State Commun. 133(2), 113–116 (2005)
181. J. Alexandrou, E. Kymakis, G.A.J. Amazatunga, Polymer-nanoyube composites: burying
nanotubes improves their field emission properties. Appl. Phys. Lett. 80(8), 1435–1437
(2002)
182. J.Y. Shim, E.J. Chi, H.K. Baik, K.M. Song. Thin Solid Films 355–356, 223 (1999)
183. H. Busta, D. Furst, A.T. Rakhimov, V.A. Samorodov, B.V. Seleznev, N.V. Suetin, A.
Silzars. Appl. Phys. Lett. 78, 3418 (2001)
184. Y.W. Jin, J.E. Jung, Y.J. Park others, Triode-type field emission array using carbon
nanotubes and a conducting polymer composite prepared by electrochemical polymerization.
J. Appl. Phys. 92(2), 1065–1068 (2002)
185. Y. Liu, S. Fan, Enhancement of field emission properties of cyanoacrylate-carbon nanotube
arrays by laser treatment. Nanotechnology 15, 1033–1037 (2004)
186. A.N. Ionov, E.O. Popov, V.M. Svetlichny, A.A. Pashkevich, Influence of a thin polymeric
covering on field emission properties of flat metal cathodes. Lett. ZhTF 30(13), 77–82
(2004)
187. A.N. Ionov, E.O. Popov, V.M. Svetlichnyi, M.N. Nikolaeva, A.A. Pashkevich, Field
emission from metal/polymer construction. Suf. Interface Anal. 39, 159–160 (2007)
188. A.N. Ionov, V.A. Zakrevskii, V.M. Svetlichnyi, R. Rentzsch. In Abstract of 10th
International Symposium Nanostructure: Physic and Technology. (St.Petersburg, Russia,
2002), June 17–21.. p. 531
189. J. Yilgor, B.C. Johnson. Amer. Chem. Soc. Polym. Prepr. 27, 54 (1986)
190. J. Lowell, A.C. Rose-Innes, Adv. Phys. 29, 947 (1980)
191. H. Liu, Q. Zhao, Y. Li, F. Lu others, Field emission Properties of Large-Area Nanowires of
organic Charge Transfer Complex. J. Am. Chem. Soc. 127, 1120–1121 (2005)
192. J. Kymissis, A.J. Akinwande, Field emission from a patterned organic conduction
composite, Appl. Phys. Lett. 82(14), 2347–2349 (2003)
193. W. Eccleston, G.A.J. Amaratunga, J. Musa, Field emission cathode and field emission
display. Pat.USA. №6853127 of 8.02.2005. Cl. H01j 1/62(313/495)
194. L. Tonks, A theory of liquid surface rupture by uniform electric field. Phys. Rev. 48, 562–
568 (1935)
195. Y. Frenkel. On a theory of liquid surface instability in electric field in vacuum. Zh. Tekh.
Fiz.6, 347–351 (1935)
196. K. Hata, R. Ohya, S. Hishigaki, H. Tamura, T. Noda, Stable field emission of electrons from
liquid metal. Jpn. J. Appl. Phys. 26(6), L896–L898 (1987)
197. J.W. Beams, Field electron emission from liquid mercury, Phys. Rev. 44(15), 803–807
(1933)
198. J. Metterauer, Field emission from thin liquid metal films. Appl Surf. Sci. 94–95, 161–170
199. L.M. Baskin, A.A. Kantonistov, G.N. Fursey, L.A. Shirochin, Peculiarities in explosive
emission of liquid metals in microwave field. Rep. USSR Acad. Sci. (USSR) 296(6), 1352
(1987)
200. A.A. Kantonistov, I.N. Radchenko, G.N. Fursey, L.A. Shirochin, Field emission of liquid
metals in alternating fields. Collogue de Physique, Colloque C8, supplement au Tome 50.
C.8. P.203–207.1989
Chapter 6
Carbon-Based Field-Emission Cathodes
Abstract This chapter reviews the main carbon-based materials that, by the
authors’ opinion, are suitable for use in the manufacturing of field emission cath-
odes. Those are, first and foremost, polyacrylonitrile carbon fibers, carbon nan-
otubes, fullerenes, graphenes etc. The ways to improve the emission characteristics
of carbon-based field emission cathodes and techniques of measurement of
volt-ampere characteristics out of such materials are also discussed.
Recent years have seen a great deal of research dedicated to carbon field emission
and carbon-based field emission cathodes. Monograph [1], issued in 2001, was
supplemented with a lot of new data, so it was decided to dedicate a separate
chapter to carbon-based field emission cathodes to append the data published in [1].
It should be noted that all carbon materials used as field-emission cathodes fall
within the category of nanostructured materials. Field emission of such materials
comes from the micro-elevations sized from 0.1 to 100 nm.
This chapter reviews the today’s areas in development of carbon-based field
emission cathodes.
bundle fibers are unequally spaced both in terms of their height and the field
emission cathode end-face area. Such inequality in arrangement of the fibers causes
even greater inequality of field-emission current from them due to the exponential
dependence of the current on the electric field intensity, and some fibers carrying
the basic current loading are destroyed. The clouding plasma [8] short-circuits other
emission centers, thereby causing a series of electric discharges on the cathode,
which finally degrades the cathode.
Graphite field emission cathodes with a well-developed emitting surface could
generate the emission current of 10 mA [9] at the average current density of less
than 10−2 A/cm2 from the cathode area of about 1 cm2.
Fiber structure. Polyacrylonitrile carbon fiber is known [10] to consist of
deeply intertwined filamentary fibrils reaching 1 µm in length and 1–5 nm in
diameter. Being the emitting surface of the field emission cathode, the end surface
of fibers is a set of chaotically located microelevations of different curve radii,
height and configuration (Fig. 6.1).
The surface structure results from mechanical cleave of a carbon fiber, when the
fibrils making up the fiber occasionally break off, thus forming a set of microele-
vations. When electrical voltage is applied, the numerous microelevations are sure
to include a few where electric intensity at the vertexes peaks. Such microelevations
are the initial emission centers. With further increase in voltage, electric intensity is
also sufficient for emission from other, less pointed microelevations.
Initial distribution of the emitting centers over the end surface of the fiber is
absolutely arbitrary and does not coincide even for quite closely-located sites of the
same fiber because of the spread of mechanical properties along the fiber axis. This
fact accounts for a wide spread of initial emission characteristics of carbon-based
field emission cathodes, which can amount to several orders of magnitude.
Heat treatment temperature. Field-emission characteristics of carbon materials
essentially depend on heat-treatment temperature because the latter affects the
mechanical characteristics, thermal conduction and electric conductivity of these
materials [10]. That accounts for the wide spread of the emission-current values
mentioned in different references, e.g., [11–13].
Measurement of the limiting field-emission current of polyacrylonitrile carbon
fiber, viz. the emission current that the cathode can steadily endure for 1 h, shows
that increase in heat-treatment temperature from 1500 to 2600 °C increases the
limiting field-emission current from 120 to 190 µA [14] (Fig. 6.2). This depen-
dence, based on research of more than 40 samples, is evaluative because some
particular field emission cathodes developed the emission current as high as 230–
300 µA. However, field emission cathodes are only failsafe in operation if the
values are lower than those specified in Fig. 6.2.
Electronic work function. The electronic work function of different carbon
materials, viz. carbon fiber, natural and technical graphite, hovers within
4.75 ± 0.05 eV [15]. Variations of the electronic work function of carbon materials
in a wide range of residual-gas pressures, viz. from ultrahigh vacuum to atmo-
spheric pressure, are relatively insignificant and amount to a few percent of the
initial value; whereas similar variations of the electronic work function of a number
of metals, such as tungsten, amount to a few tens of percent, which causes abrupt
changes in the emission current, reaching several orders of magnitude. The latter is
very important for practical use of carbon materials as field emission cathodes
because it accounts for their stable and long functioning under different vacuum
conditions, including high technical vacuum at the pressures of about 10−7–
10−7 mm Hg. Moreover, the carbon materials studied show that their electronic
work function remains almost the same after the samples are ground and aged in air.
Service life. Service-life testing of carbon-based field emission cathodes shows
their high durability in high technical vacuum. In particular, the operating time of
carbon polyacrylonitrile fiber is 7500 h [16] at the current draw of 60 µA and that
of fiber bundles is more than 500 h in the cyclic mode. Carbon fibers were
repeatedly tested for 1000 h at the currents of 20, 50, 100, 150, 200 µA [17] to
show stabilization of field-emission characteristics of carbon fibers in long-term
operation. Dependence of the instability parameter on field-emission current for
298 6 Carbon-Based Field-Emission Cathodes
different treatment periods is given in Fig. 6.3. The instability coefficient is deter-
mined for 5 min by the formula of
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
N 2 2
r¼ I ðI Þ ð6:1Þ
N1
Fig. 6.4 Emission images of the field emission cathode out of bunches of carbon fibers in process
of molding
of the formed fibers has not changed throughout the experiment, so it has been
impossible to determine the shrinkage rate of the formed-cathode fiber.
Equally, these conclusions pertain to other carbon materials.
Reproducibility of characteristics. Forming is closely connected with one of
the main problems related to practicality of field-emission characteristics in the
system of such cathodes.
Analysis of the carbon-material structure shows that it is almost impossible to
achieve geometrical reproducibility of the form-factor for field emission cathodes of
such materials. So the main efforts should be aimed at achievement of identical
current-voltage characteristics at an uncontrollable form-factor. The characteristics
are achieved by purposeful shift of the current-voltage characteristics to a higher
voltage area up to falling within the tolerance limits. This operation is usually
carried out with supervisory computer control systems by taking a series of
current-voltage characteristics up to the currents exceeding the original value for
forming, followed by reforming of the field emission cathode. On completion, its
current-voltage characteristic practically does not vary in the field of heavy currents
(in Fowler–Nordheim coordinates), but shifts up to falling within the desired tol-
erance in the field of minimum currents (Fig. 6.5).
Parallel switching-on of the field emission cathodes treated in this way causes full
addition of the currents in the achieved multiple-emitter system; i.e., the total current is
equal to the sum of emission currents of each cathode within the limits of fluctuations.
300 6 Carbon-Based Field-Emission Cathodes
Fig. 6.5 Volt-ampere characteristics of field emission cathodes before molding (a), after
individual molding (b), after shift of characteristics (c)
are pierced with graphite-filament bundles, the filaments being vitrified by heating of
the glass tubes pierced with graphite-filament bundles to the temperature close to the
melting point of glass with subsequent elongation and compression. The carbon
material is arranged in the glass substrate by respective alternation of vitrified
graphite-filament bundles and continuous glass rods, the block being fused and cut
into separate matrices before machining. The matrix is ground and polished until
graphite-filament parts stick up over the matrix surface and the effective area is flat
enough for the matrix to be mounted into the field-emission device.
After the grinding and polishing, the matrix can be etched with, inter alia,
hydrofluoric acid to improve control over the graphite-filament parts sticking up
over the matrix surface.
The graphite-filament bundles passing through the glass tubes can be simultane-
ously fed from one and/or several bobbins into one and/or several glass tubes,
respectively, which not only speeds up fabrication of vitrified graphite-filament bun-
dles, but also provides uniformity of the vitrification mode. The glass selected has the
coefficient of thermal expansion that is close to the coefficient of thermal expansion of
graphite-filament bundles. The purpose is to avoid cracking of the glass and rupture of
graphite filaments during their vitrification, fusing and temperature conditioning.
Carbon-fiber bundles are not particularly rigid, which is why they mix up and/or
break off during attempted insertion into finite-size capillaries. This feature makes it
necessary to pre-spool the bundles, feed them from the spools into the glass tube/billet,
heat up the bottom of the glass tube/billet to a ductile condition, weigh it down and pull
out at a certain rate. Difference in the rates of feeding and pulling generates a vitrified
graphite-filament bundle. The cross-section shape of the resulting item is similar to
that of the initial billet, but has somewhat smaller geometrical dimensions. The
technology suggested makes it possible to vitrify several graphite-filament bundles at
the same time, their number depending, inter alia, on the parameters of the preheater
used in this technology. For this purpose, the graphite-filament bundles are fed from
several spools at the same al respective glass tubes/billets.
302 6 Carbon-Based Field-Emission Cathodes
The cohesion between the carbon fibers and the glass can be improved, inter alia,
by application of a two-layer glass tube, resulting from insertion of a tube of one
glass type into a tube of another glass type. In this case, the brittle-to-ductile tran-
sition temperature of the internal glass layer is lower than that of the external glass
layer by about 100 °C. Subsequent heating makes the external layer ductile and
shapes the resulting item, while the internal glass layer liquates, which improves the
cohesion between the graphite filaments and the glass after cooling.
The necessary arrangement of graphite filaments in a glass substrate is provided
by formation of a fiber glass block by alternation of vitrified graphite-filament
bundles and continuous-glass rods. The resulting block is fused and cut into matrices.
To provide an equal distance between the cathode and the extraction electrode
when mounting the matrix cathode in the field-emission device, it is necessary to have
a flat effective area of the matrix. For this purpose, the matrix effective area is ground
and polished. To improve connection of the insulated contacts to each of the cathodic
elements, it is also desirable to polish the back surface of the matrix. The macro-
scopical effect of the field near the cathode steps up if the graphite filaments stick up
over the glass surface at constant running voltage. This effect is made by abrasive
machining of the matrix or its selective chemical etching. Polyacrylonitrile
graphite-filament bundles (Fig. 6.7) are coiled on one or several spools 2, one of
which can be seen on Fig. 6.7. Every bundle goes through a respective glass tube
(billet) 3. The most suitable glass type is make C93 because its coefficient of thermal
expansion is close to that of the glass used in device-case assembly. Billet 3 is fixed in
the collet closer of feed mechanism 4, which is ball-screw assembly 5. The ball-screw
assembly is driven by a worm-gear electric motor (not shown on the drawing). Then
the lower part of the billet is dropped into preheater 6 and heated until the glass
becomes ductile. For the glass used, this temperature is 550 °C. The ductile billet is
weighed down, fed into pulling mechanism 7 and pulled out at a certain rate.
Difference in the rates of feeding and pulling produces an item similar to the initial
billet in the shape of the cross-section, but having somewhat smaller geometrical
dimensions. The pulled-out vitrified graphite-filament bundles are cut into separate
billets of the targeted length, their width and height being determined by the ratio of
the feeding rate to the pulling-out rate of the preheater, and the optimum length being
given by the preheater dimensions. The carbon material is properly arranged in a glass
substrate during formation of the block by pertinent alternation of vitrified
graphite-filament bundles and continuous-glass rods. The block is assembled in a
metal case called “die”. The assembled block is fastened by tying up at the ends with a
thin nickel-chromium thread, followed by removal of the “die”. The laid and fastened
block is put into the preheater and fused under the pressure of 2–3 atmospheres. The
ready block is cut into matrices, their thickness being selected depending on the design
of the vacuum device where they are to be used. The matrices cannot be too thin, i.e.
less than 2 mm, otherwise cohesion between the glass and the graphite filaments can
fail with the result that the graphite filaments can drop out of the matrix. Then the
matrices are ground and polished. For the graphite filaments to stick up over the matrix
surface, the matrix is treated either mechanically or chemically. Difference between
the mechanical rigidity of the graphite filaments and the glass substrate makes the
graphite filaments stick up over the matrix surface as a result of mechanical abrasive
treatment by trying abradants and the rotation rates of the polishing-machine spindle.
The length of the prominent part is finer controlled by chemical etching in a solution
that does not affect the graphite filaments, but dissolves the glass, e.g., hydrofluoric
acid. Duration of the etching is selected on an experimental basis depending on the
glass type and the matrix area.
That makes the method suggested more technological and able to provide any
desired arrangement of the graphite-filament bundles and package density.
Moreover, the matrix produced with this method is quite durable in mechanical
terms and suitable for further treatment if necessary. Use of monolithic graphite
filaments improves stability to ion bombardment of the cathode when working in
technical vacuum and extends the life capability of the device, as well as prevents
premature failure caused by separation of a layer from the emitter. Use of glass to
fabricate the carrier matrix provides electrical isolation of elements of the cathode
and allows to drive each element separately.
Paper [29] suggests field emission cathodes for high-voltage accelerating devices
with voltage of about 70 kV. The carbon fibers of 10 µm in diameter and 0.5 mm in
length were grown with a plate catalyst on a graphite-fabric substrate. The fibers
were grown perpendicularly to the surface consisting of a conducting carbon com-
posite coating the carbon fabric. Paper [30] describes a combination of the Spindt
thin-film technology with carbon fibers. In this case, a fiber-like carbon structure is
vertically sputtered on a molybdenum cone. The growth of carbon nanofibers is
controlled by the emission current. The longer the fiber, the higher is the emission
current. The process comes to the end when the fiber length is a few tens of µm.
However, the technologies of direct carbon nanofiber growing seem to be very
promising. In particular, paper [31] suggests a method of fabrication of
vertically-aligned carbon nanofibers in situ as a field-emission element. The method
is based on the processes recognized in microelectronics, e.g., electronic lithogra-
phy of high-resolution and plasma CVD (PECVD) methods. In brief, the techno-
logical process of fabrication of a field emission cathode structure is shown in
Fig. 6.8. The technological process is detailed in [31].
Nowadays nitrogen-containing carbon nanofibers cause great interest.
Application of carbon-nitrogen materials in field emission cathodes is very
promising because replacement of carbon atoms with those of nitrogen in graphite
layers causes the structural defects facilitating emission of electrons when an electric
field is applied [32, 33]. In addition, a number of theoretical papers shows that the
nitrogen atoms embedded in the carbon nanotube wall graphite layers act as electron
sources, which can also intensify electron emission from such objects [34, 35]. None
the less interesting is the fact that nitrogen-containing carbon nanotubes and nano-
fibers are essentially more inert than pure carbon nanotubes or nanofibers in
chemical terms [36, 37]. That is why the field emission cathodes based on
carbon-nitrogen nanomaterial are less tending to be affected by residual-gas
adsorption-desorption processes, which cannot but improve their functional stability.
The review of research done by now shows that the maximum concentration of
nitrogen in carbon nanofibers does not exceed 5% [35–37]. This concentration of
nitrogen does not produce a high structure-defect density in graphene layers. Nor
does it distort the electron density or essentially increase the emission efficiency.
Paper [38] suggests a new method of synthesis of carbon-nitrogen nanomaterial in a
high-pressure apparatus, as well as a method of fabrication of highly-effective
field-emission cathodes on its basis.
Carbon-nitrogen nanofibers were produced with a high gas pressure apparatus
[32, 39]. The processes of carbon-nitrogen nanofiber synthesis in both nitrogen and
the argon-nitrogen mixture are well developed. Preliminary testing has shown
prospective viability of this material for cold cathodes with the lowest ever initial
electric field intensity [32]. The basic methods of fabrication of field emission
cathodes from this material is screen printing of carbon-nitrogen powder paste,
deposition of carbon-nitrogen nanofibers on the substrate during synthesis in the
working chamber, catalytic growing of carbon-nitrogen material directly on the
substrate.
6.1 Carbon Fibers 305
Fig. 6.8 Stages of the process of manufacturing a field emission cathodic structure out of vertical
carbon fibers: a 4-in. silicon substrate out of n-type Si; b direct electronic lithograph for applying
the plasma CVD catalyst consisting of a 200 Å Ti layer and a 200 Å Ni layer; c drawing an 1 lm
layer of SiO2; d a photolithograph for applying the control electrode; e creation of the image in the
layer of photoresist; f removal of the photoresist coating; g applying of a 1000 Å Cr film; h,
i reactive ion etching in order to remove the layer of exposed resist coating and the SiO2 layer;
j in situ plasma growth of vertically oriented carbon nanofibers, directly in the operating structure
306 6 Carbon-Based Field-Emission Cathodes
The control samples of 0.5 cm2 were tested in a vacuum chamber at the pressure
of 10−6 mm Hg. At the current density of 10 µA/cm2, the macroscopical threshold
field of turn-on of the field emission cathodes is 1.2–1.3 V/µm. The emission
current density of 1 µA/cm2 is produced by the electric field of 2.3–2.4 V/µm. The
results attained show that carbon-nitrogen nanofibers are very effective emitters
with threshold electric field intensity much less than that of field emission cathodes
based on multiple-wall carbon nanotubes.
The long-term tests were carried out for 80 h in the current stabilization mode.
The current was fixed at 250 µA. At the same time, the dependence of the applied
anode-cathode voltage necessary to attain the preset emission-current value, on time
was measured.
The field emission cathodes produced from carbon-nitrogen nanofibers by screen
printing have a quite high emissivity stability in time. It is worth noting that
long-term stability of field-emission current depends not on nanofiber properties but
on the field emission cathode fabrication method, i.e., the strength of cohesion
between the fiber and the substrate.
Fig. 6.9 Crystallographic structure of graphite. The elementary cell is defined by unit vectors a,
b, c
Fig. 6.10 Formation of carbon nanotubes by twisting of the graphene plane and connecting the
points A and A′
308 6 Carbon-Based Field-Emission Cathodes
ch ¼ na þ mb; ð6:2Þ
where п, m are real numbers, and a, b are unit vectors in the graphene planes.
A carbon nanotube results from folding of the graphene layer and connection of
points A and A′. Then it is only determined by vector ch.
Tube diameter d is equal to
jch j a pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
d¼ ¼ n2 þ nm þ m2 ; ð6:3Þ
p p
pffiffiffiffiffiffiffiffiffiffiffi
where a ¼ 3dcc ; the constant of the lattice, and angle
pffiffiffi m
h ¼ arctan 3 ð6:4Þ
2n þ m
quantitatively determine the chirality of the nanotube. There are two extreme cases
of carbon nanotubes, viz. armchair- and zigzag-type ones. Zigzag is determined by
angle h = 0°, which meets the (n, 0) vector. In this structure, C–C bonds are parallel
to the tube axis. The armchair-type structure is characterized by angle h = 30°,
which meets the (п, п) or (2n, –n) vector. In this group of tubes, the C–C bonds are
perpendicular to the tube axis. Other tube combinations belong to a chiral type with
angles 0°< h < 30°. A single tube is usually termed as a single-wall nanotube,
SWNT. The nanotubes consisting of coaxial graphite cylinders are termed as
multiple-wall nanotubes, MWNT.
We have reviewed the types of ideal 1D-carbon structures. In practice, the
carbon nanomaterials resulting from synthesis have a great variety of structures.
Unlike ideal cases, real carbon structures are characterized by defects of different
types. A failure in the sequence of layers causes packing defects. A large number of
such defects results in a complete disorder of layers about the tube axis. Another
type is lattice-bond defects caused by alien atoms or a change in valence of some
carbon atoms.
Accordingly, there may be compound forms of carbon transitional from nan-
otubes to carbon fibers. They are usually termed as graphene nanofibers [40, 41].
These formations have a structure similar both to nanotubes and carbon fibers, but
the internal walls are not very smooth and usually include graphene layers.
Let us note here that hereinafter the term “carbon nanomaterials” mostly refers to
carbon nanotubes and nanofibers.
Now there are a lot of publications about field-emission properties of different
carbon nanostructures, e.g., review [42]. Almost all the authors mention a dis-
crepancy between the available experimental data about the geometrical shape and
field-emission parameters, viz. emission-current density, threshold field, when
analyzing the simplest model of the emitter with properties characteristic for a metal
apex, e.g., papers [43, 44]. Analyzing the available data based on the regularities for
metal apices, the authors of paper [45] received the geometrical amplification
coefficient exceeding 104, which is only possible if the emitting area is comparable
6.2 Carbon Nanotubes 309
Fig. 6.11 Scheme image of the curved graphene planes arising on edges of nanocrystallites of
graphite
The assumption of resonant tunneling through the surface states of carbon field
emitters is directly proved by the profile of the energy spectrum of the emitted
electrons. In particular, paper [64] shows an additional maximum in the power
distribution of electrons with the energy lower than the basic maximum by 0.45–
0.5 eV when studying field emission from carbon nanocrystals and nanotubes. The
authors establish that the additional maximum is generated by the conditions of the
experiment; in particular, further increase in the emission current eliminates the
additional maximum. The conclusion is that there are surface electronic states at the
Fermi level.
Previously, a similar field-emission model allowing for the effect of the surface
layer of a dielectric or a wide-gap semiconductor was also considered in [63, 65].
As a rule, these models have the surface layer of a dielectric or a wide-gap semi-
conductor that is much thicker than sp2–sp3 transition considered in [53, 54].
However, effective resonant tunnelling necessitates a very thin dielectric layer
approximately equal to the barrier width at the boundary with vacuum.
For metal or semiconductor emitters, production of such thin layers from tra-
ditional materials, viz. W, Mo, Si, etc., is a rather difficult problem in terms of
technology; while for carbon nanostructure based films such a layer is a conse-
quence of bending of atomic layers, which makes them a unique material for
field-emission cathodes.
two-region furnaces by the thermal CVD method. The resulting nanotubes are
about 20 nm in diameter and 1 µm in height. In this case, the threshold electric field
is 3 V/µm and the current density is 0.3 A/cm2 at the field intensity of 4 V/µm.
The author of papers [74–76] managed to achieve further reduction in the
deposition temperature. He introduced an additional layer of CaF2 between the Ni
catalyst and the substrate, which reduced the process temperature to 500 °C, and
produced the field emission cathode samples differing in distribution of the catalyst
over the substrate surface. The experiments show that the cathodes spotted with a
catalyst of 50 µm in diameter have the best field-emission characteristics.
The latest result is of not only theoretical but also great practical importance in
terms of fabrication of field emission cathodes for flat display screens since they
need a large array of insulated field emission cathodes. Some authors suggest
fabrication of matrix field emission cathodes from carbon nanotubes sized
50 50 µm2 [77] and 30 30 µm2 [78]. Paper [77] suggests fabricating field
emission cathodes of 10 10 pixels consisting of multiple-wall nanotubes of
10 102 nm in diameter and a few µm in length at 550 °C, the nanotubes being
interwoven to form a “felt” structure. Though the threshold electric field is 1.8 V/
µm, the total current from 400 emitters is 73 µA at the field of 10 V/µm.
The vertical well-aligned nanotubes are grown on a Si substrate at 800 °C on a Fe
catalyst [78] (Fig. 6.12). This vertical position is caused by Van der Waals interac-
tion between the neighboring nanotubes. The nanotube density is determined by the
density of Fe nanoparticles on the silicon substrate. The current density is indicated to
be 1 MA/cm2 at the field of 2 V/µm and as low as 80 mA/cm2 at 3 V/µm.
Fig. 6.12 The photo in SEM of vertically focused nanotubes on the catalyst from Fe film on Si to
a silicon substrate: a 30 30 lm blocks of nanotubes at distance 125 mm from each other; b the
increased image of the block on which the orientation of tubes is well visible
6.2 Carbon Nanotubes 313
Fig. 6.13 Flowchart of manufacturing process of field emitters out of carbon nanotubes
314 6 Carbon-Based Field-Emission Cathodes
nanotubes. In the groove array, a pixel of 300 300 µm2 square can have different
sizes, namely 5 5, 10 10, 20 20, 30 30 or 40 40 µm2.
The prepared substrate is thermally oxidized to make an insulating layer between
the substrate and the would-be nanotubes (Fig. 6.13a). Then a photoresist and Fe of
25–50 nm in thickness is deposited (Fig. 6.13b). Further on, the photoresist is
removed together with the iron, while the iron film remains where the nanotubes are
grown up (Fig. 6.13c). What follows is carbon-nanotube growing by the thermal
CVD method at 900 °C (Fig. 6.13d). The process is detailed in [80].
The silicon substrate with grown nanotubes is coated with a thin layer of a
composite material based on glass (Fig. 6.13e). Having dried up, it provides a
strong adhesion of the nanotubes to each other and to the substrate. What follows is
mechanical polishing with diamond paste. At the first stage, the structure is graded
(Fig. 6.13f). At the second stage, properly selected diamond powder removes the
excess of glass composite material to release the nanotubes; in addition, it removes
the tops of the closed nanotubes, thereby making them open (Fig. 6.13g).
The emission tests carried out in the diode mode at the pressure of 10−7 mm Hg
show high stability of the emission current for 2 h and a threshold electric field of
*1.7 V/µm.
Another problem of fabrication of field emission cathodes with planar tech-
nologies is current leakage through the electron-electron gap of the field emission
cathode—the extraction electrode. The reason thereof is shortness of the gap. The
gap can be lengthened by sputtering of an additional internal protective layer [81,
82]. This measure alone reduces the leakage current more than 7 times at the
extraction voltage of 100 V.
The temperature of carbon nanotube growth decreases due to reduction of the
catalyst particles down to <10 nm [83]. It is also necessary to remove amorphous
carbon that accompanies deposition of carbon nanotubes at low temperatures.
Vertical position of carbon nanotubes is improved at supply of extraction voltage
between the sample surface and the plasma generator case during the deposition.
Finally, removal of defective nanotubes with gas-mixture plasma gives the most
perfect structure. That is the way to fabricate a field emission cathode with a 1 V/
µm threshold field at *500 °C. A flat display screen with size of 5.4″ produced
with this technology shows a very good uniformity of emission over the screen
area, viz. *3.3%.
Field emission cathodes made out of carbon nanotubes with improved substrate
adhesion produced via methods of powder metallurgy were suggested in [84]. Such
cathodes are made of a mixture of multiple-wall carbon nanotubes and powdered
metal (zinc in case of the paper). Nanotubes of 1–10 µm in length and 50–60 nm in
diameter are made by arc discharge. At first, the zinc powder and the nanotubes are
mixed in a mortar and the resulting mixture is pressed at 0.5 MPa. Fusing takes
place at 420 °C in a vacuum furnace. The final size of the sample is
ø 5 mm 2 mm. Field-emission properties of the surface of the sample fused are
improved by polishing followed by etching in an acid to remove a thin surface zinc
layer. So the ends of carbon nanotubes develop the microelevations effectively
6.2 Carbon Nanotubes 315
aggravating the electric field, while the main part of the nanotubes is in zinc, which
essentially improves their cohesion with the substrate.
It is noted that stability of field-emission current of a field emission cathode
made by powder metallurgy is higher than that of the field emission cathodes grown
by the CVD method (Fig. 6.14).
Further development of this approach leads to application of organotitanium or
organometallic compounds [85], e.g., Ti (O–n–C3H7)4, Al(IC4H9)3, Si(O–n–C4H9)4
and other metals of periodic groups 3, 4, 5 and 6. First comes the paste consisting of
40–95 wt% of an organometallic compound, the rest being multiple-wall carbon
nanotubes. The paste is applied on substrates by the printing method and fused at
the temperature ranging from 250 to 600 °C. The temperature and the treatment
conditions are determined by a necessity of all-burn of the organics, on the one
hand, and the minimum oxidation of nanotubes, on the other hand. During the
fusing, the organic carrier is completely removed with evaporation, and the layer
surface microcracks by 0.1–100 µm forming grooves where the free ends of nan-
otubes are put. The rest of them are securely fixed in the metal layer. A smaller
amount of the organic compound results in worse microcracks, and the content of
more than 95% essentially decreases the emission centers in number.
316 6 Carbon-Based Field-Emission Cathodes
It has long been known [6, 20] that stable operation of carbon-based
field-emission cathodes needs current forming or aging. Right selection of tech-
niques and modes largely determines the practicality of carbon-based field emission
cathodes in general and carbon nanotubes in particular.
Recent technologies include pre-treatment of the surface of field emission
cathodes made of carbon nanotubes.
Classical ion bombardment is supplemented with laser exposure [86] and glow
discharge [87]. In the former case, carbon nanotubes are applied by a standard
printing, dried at 400 °C for 2 h and are irradiated with a KrF laser of 248 nm in
wavelength in an open air, energy density amounting to 2–4 mV/cm2 or 40–
80 mJ/cm2. The laser bundle of 9 9 mm2 in area is homogeneous. The homo-
geneity of the bundle turns out to result in considerable homogeneity of the
field-emission image. Addition of a glass filler to the original printing composition
and subsequent laser treatment reduce the threshold field to 0.34 V/µm. Paper [87]
suggests an inexpensive method of ageing of field emission cathodes made of
carbon nanotubes in glow discharge in a hydrogen medium. Carbon nanotubes are
grown on a stainless-steel plate by the thermal CVD method with involvement of a
catalyst. The system intended for aging of the resulting field emission cathode and
measuring of its field-emission characteristics was located in a vacuum chamber.
The pilot installation diagram is shown in Fig. 6.15. The metal anode and the field
emission cathode are spaced 100 µm, while the anode and the glow-discharge
cathode are *3 cm apart. For the hydrogen plasma to cover the whole of the field
Fig. 6.15 The scheme of experimental setup for training the field emission cathode out of carbon
nanotubes in abnormal glow-discharge
6.2 Carbon Nanotubes 317
emission cathode surface, the discharge conditions are similar. The best results are
attained at the hydrogen pressure of less than 1 mm Hg. Under these conditions, the
running voltage of the field emission cathode is seen to go down.
Active search of the ways to create new carbon materials is still going on. Most of
them are based on the technologies already developed for nanotubes. The following
summarises the technologies and structures based on them.
Diamond-like carbon is a form of amorphous carbon (a-C) or hydrogenated
amorphous carbon (a-C–H) with a considerable number of sp3-bonds [88]. The
number of sp3-bonds and hydrogen in any diamond-like carbon depends on the
deposition process involved and can be represented by a three-phase diagram
(Fig. 6.16). The most popular method for synthesis of a-C:H is plasma CVD.
However, this technology yields a lot of hydrogen.
The larger number of sp3-bonds in a-C is plasma CVD. However, this tech-
nology yields a lot of hydrogen.
The larger number of sp3-bonds in a-C, which is designated as tetrahedral
amorphous carbon (ta-C), results from use of a highly ionized ion beam of the
respective energy.
Electronic structure a-C depends on the content of sp3-states. The sp3-states
develop only strong bonds, while s sp2-states develop both a bonds and looser p
bonds [89]. p bonds lie close to the Fermi level, thereby determining the forbidden
bandwidth. It has been experimentally found that impurity-doping incorporation of
ta-C and ta-C:H with nitrogen in the course of sputtering does not reduce the
forbidden bandwidth.
Field emission from diamond-like structures is mostly limited by intrinsic
conductivity of the structure. Diamond-like structures are polycrystals, which is
why their grain boundaries function as conducting tracks.
Fig. 6.17 Scheme of an element of an address-controlled matrix of field emission cathodes on the
basis of diamond-like structures
catalyst layer. Then the resist layer is removed and a nanostructured carbon layer of
a methane-hydrogen mixture is applied at the temperature of 680 °C by the thermal
CVD method [93, 94]. It is worth noting that the carbon layer is synthesized only in
the central part of the hollow because the catalyst grains are applied through the
hole of *10 µm. Small control current can result from adjustment of the sticking
up of the modulator electrode over the silicon-dioxide film. Both a simulation and
an experiment show that practically zero current of the modulator electrode can be
achieved if the chromium film is at the same level with SiO2.
At present, the designers of field emission cathodes pay much attention to
amorphous structures based on nanotubes and similar structures. Different authors
give them have different names, viz. felt, mesh, velveteen, velvet etc. However their
essence is the same: chaotically interwoven carbon nanofibers without any common
orientation. Let us give some examples of implementation of these structures
below.
Paper [95] gives an example of such a structure, which the authors termed as
carbon felt. It appearance is given in Fig. 6.19. The total thickness of the layer is
*70 µm. The material, single-wall carbon nanotubes, is 50–70% pure. The
impurities are mostly leftover nickel catalyst in form of nanoscale particles. The
average diameter of the carbon nanotubes in the filaments is *20 nm, which
accounts for a small radius of the field-emission centers.
Fig. 6.19 SEM microphotograph of chaotically intertwined nanotubes (so-called “carbon felt”)
320 6 Carbon-Based Field-Emission Cathodes
This section understands foils as carbon materials of small thickness and large
length and width, which are not applied on a substrate.
The first material to be used as field emission cathodes were thin pyrolytic-
graphite plates.
Pilot experiments show that pyrographite is quite promising as a field emitter.
The emission centers of field emission cathodes made of pyrolytic graphite are thin
strong microplates forming the pyrolytic graphite structure (Fig. 6.20). So
field-emission properties of foil made of pyrolytic graphite are close to those of
polyacrylonitrile carbon fibers. Further research is aimed at determining the optimum
treatment temperature and thickness of pyrographite plate. To this end, the research
focuses on the emission characteristics of the field emission cathodes made of pyro-
graphite plates of 30 lm in thickness processed at 1400, 2000, 2500 °C, as well as the
plates of 5 and 150 lm in thickness processed at 2000 and 1700 °C, respectively.
The measurements are preceded by a standard forming procedure in the linear
mode up to the maximum possible current of 6–8 mA at the length of the working
edge of 2 mm, though the field emission cathodes made of such materials show
some elements of a self-forming process.
The current-voltage characteristics of these field emission cathodes taken every
10 min during forming determine the dependence of the current at which the
effective cathode area begins restructuring on the heat-treatment temperature of the
pyrolytic graphite as thick as 30 µm and on the thickness of the sample plates at
Ttrtm = 2000 °C (Fig. 6.21).
The diagrams show that the limit current of these field emission cathodes goes
up with increase in the material heat-treatment temperature. The most likely reason
Fig. 6.21 Dependence of current at which pyrographite field emission cathode surface
reorganization begins on the plate thickness at T = 2000 °C (a) and on temperature at plate
thickness of 30 lm (b)
thereof is that the material tenacity improves with growth of the treatment tem-
perature. A similar dependence also takes place for carbon-based field emission
cathodes [14]. The growth in the sample-plate thickness increases the total number
of emitting microelevations on the cathode surface, thereby increasing the current
value at which the effective area restructures.
The field-emission current instability level is constant for all samples at small
currents, viz. down to 1–10 µA, and goes down with growth of the current at its
higher values, viz. more the 1 mA. At the currents of less than 1 µA, the pyro-
graphite field emission cathodes of 30 lm in thickness showed the lowest insta-
bility value at Ttrtm = 2000 °C (r = 9%), and the highest one at Ttrtm = 2500 °C
(r = 27%). At the currents of 3–6 lA, the instability of all samples is about 1–2%.
Analysis of field-emission current fluctuations shows that the largest number of
the emitting centers with other things being equal belongs to the pyrographite field
emission cathode of 30 lm in thickness and the treatment temperature of 2000 °C.
As the maximum current drawing, long service life and uniformity of field emission
over the cathode surface directly depend on the number of the emitting centers,
pyrographite with such parameters is most preferable for electronic devices.
The main weakness of pyrolytic carbon field emission cathodes is difficulty of
their fixation and fabrication of cathode assemblies or cathode-modulation assem-
blies with reproducible field-emission characteristics, respectively. The reason is
impossibility to achieve field emission from a foil flat surface.
This conclusion refers not only to pyrolytic graphite, but also to any carbon-
based foil.
The following are the data on a successful attempt to fabricate a planar field
emission cathode of graphite foil. In particular, thermally-expanded graphite
(TEG) can be used. The central idea of the method is to form field-emission centers
on the graphite-foil surface by pulsing laser light [102, 103].
The pulsing laser light forms a crater on the TEG foil surface. Figure 6.22 shows
a REM-image of a typical crater. The crater falls into two areas: the central area
6.4 Carbon Foils 323
(I) and the peripheral area (II). The central area is a hollow in the foil developed
during intensive evaporation of graphite caused by laser. The peripheral area is
formed by bulged graphite foil layers. The foil is bulged by the forming vapors
penetrating into the pores. This process does not crumple up the initial
microstructure of thermally-expanded graphite but even additionally unfolds it.
The REM-images of the craters show that the crater-surface microrelief has a
characteristic size of 20–60 nm. In the relief structure, the TEG-particles are uni-
formly distributed over the crater surface and the space between the neighboring
particles exceeds their size by an order of magnitude. Moreover, the TEG-particles
are seen to form the crater by orienting themselves at right angle to the crater
surface. In this case, most probable is field emission from the end faces of these
particles; i.e., the emission can proceed from the graphite-layer end faces, i.e. layer
packages positioned at right angle to crystallographic axis c. The bibliographic-data
review made in Chap. 1 mentions that this structure is optimal for a field-emission
cathode. It is also worth noting that the characteristic sizes of the resulting irreg-
ularities are close to the nanotube size. Hence, the crater surface is expected to
produce intensive field emission comparable with that of the carbon-nanotube
cathode surface, which is confirmed by the data given below.
Such emitters have to meet the following requirements:
• the crater has to be formed without a through hole in the TEG-foil;
• the crater has to be as small as possible;
• the crater has to be axially symmetrical;
• the near-crater area has to be free from TEG impurities breaking the crater
symmetry.
324 6 Carbon-Based Field-Emission Cathodes
1 4 d3 1 4 d3
E ¼ mL0 ¼ p L0 ¼ p Cq; ð6:5Þ
23 8 23 8
where v is the hemisphere volume, L0 stands for the energy necessary to sublime a
substance volume unit at 0 °C, d is the hemisphere diameter, q is the substance
density, C stands for the energy necessary to sublime a substance volume unit at
0 °C, E is the laser-pulse energy. Then the diameter of the evaporated area can be
expressed by the following formula, viz.
sffiffiffiffiffiffiffiffiffiffiffiffi
3E
d¼2 : ð6:6Þ
3
2pCq
where d0 is the diameter of the focused laser bundle, Y is the laser-bundle con-
vergence half-angle, C standing for the specific energy necessary to sublime a
substance at 0 °C. Our further calculations are based on Formula (6.7).
Adequate determination of the external size of the crater necessitates consider-
ation of TEG-foil bulging. The crater peripheral-area size depends on how deeply
the graphite vapours generated by laser light have penetrated into the graphite-foil
pores. A characteristic giving an idea of the foil porosity is the ratio of the graphite
density to that of the TEG foil. Accordingly, the linear coefficient of vapor pene-
tration into the pores is proportional to the cube root of this ratio multiplied by
empirical coefficient k.
6.4 Carbon Foils 325
Proceeding from the assumptions made, the external diameter of the crater,
which depends on the power of laser pulse E and TEG-foil density q, is expressed
by the following formula, viz.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rffiffiffiffiffi 3
q 3 d0 3EtgY
d ¼ 2k 3 0 þ ; ð6:8Þ
q 2 pCq
The total test period is about 800 h. Within the first 70 h, the current is 200 µA,
the current density drawn from the surface of each crater being 0.4 mA/cm2. At the
expiration of the 70 h, the current is 100 µA, the current density being 0.2 mA/cm2.
Figure 6.24 shows the diagrams of current and voltage variation in the device.
Analysis of current-voltage characteristics shows that degradation of emissivity
is caused by reduction of the total number of emission centers on the cathode
surface. The principal cause of the degradation is bombardment of the cathode with
residual-gas ions. The more intensive the ion bombardment, the heavier is the
current drawn. It is also confirmed by the data of the long-term operating time.
Figure 6.24 shows that at the first stage of the operating time, when the cathode
current is 200 µA, the running voltage of the cathode doubles within 70 h, and
when the drawn current is reduced down to 100 µA, the running voltage doubles
within about 700 h.
Research of field-emission properties of flat field emission cathodes based on
graphite foils discovered a ring effect.
6.4 Carbon Foils 327
The rings are a diffusely flared area of the luminophore surrounding the main
field-emission image. The ring diameter often exceeds the size of the main image.
The ring brightness increases with increase in the emission current.
The reason thereof is as follows. The electrons get to the anode and knock
secondary electrons out of its surface. The electric field makes the secondary
electrons come back to the anode and form a ring.
Let us consider a flat-diode model with the anode and cathode spaced L apart
(Fig. 6.25). At the anode voltage of U, the field in the anode-cathode gap is
E = U/L. Considering the motion of the secondary electrons near the anode, we
neglect existence of emission centers on the cathode. They cause local enhancement
of the field near the cathode, but do not essentially affect the electron behaviour near
the anode. Let us consider the motion of the secondary electron having the initial
energy of eS and escaping from the anode surface at the angle of a.
Simple calculations of the electron path in a homogeneous constant electric field
make it possible to find the distance from the secondary-emission point in the anode
plane as
eS
r¼2 L sin 2a; ð6:9Þ
eU
As a result, the external radius of the ring formed by electrons with the initial
energy of eS is
eS
rmax ¼ 2L : ð6:11Þ
eU
A lot of applications, such as flat display screens, light sources, some SHF-devices
etc., need large-size field-emission cathodes, i.e., those of the area ranging from a
few cm2 to several hundreds of cm2. Now there is a great number of flat field
emission cathode fabrication methods, some of which we have already considered
in previous sections. They include thermal vapour-phase deposition (CVD) [109,
110], CVD in an electric arc [111], CVD in microwave plasma [112], laser-pulse
sputtering [113].
Each of these methods has a disadvantage of its own; e.g., all kinds of vapor
phase deposition are technologically sophisticated, which results in a high cost price
and a narrow range of suitable carbon materials mostly represented by different
kinds of diamond-like films and carbon nanotubes. Use of such promising carbon
materials as carbon fibers and pyrolytic carbon is thus out of question.
330 6 Carbon-Based Field-Emission Cathodes
6.5.1 Electrophoresis
where e is the dielectric permeability of the fluid, η is the fluid viscosity, E is the
electric field intensity, 1 is the electrokinetic potential.
The fundamental value affecting the particle velocity is electrokinetic potential 1,
which is determined by the interphase-potential jump and the double-layer
structure.
There are different theories of the double-layer structure, the mechanisms of its
formation and the effect on the electrokinetic potential. Papers [114–118] feature
and analyze these models in details.
The schematic diagram of electrophoretic application of carbon materials on the
starting-sheet surface is presented in Fig. 6.28 [119].
A glass vessel is filled up with a dispersion medium, viz. isopropyl alcohol, ISA,
with a salt of a high-valence metal dissolved in it and functioning as a charger, and
a carbon powder. In this case, the charger was La(NO3)3. The basic function of the
charger is intensification of application, as well as improvement of adhesion of
carbon particles to the substrate due to formation of a cementing film consisting of
the charger-metal oxide and hydroxide on the surface of the cathode during the
electrolytic reactions. Then the resulting solution assumes electrodes. The anode is
represented by a platinum plate. An advantage of this metal is its inertness in most
chemical processes. Due to this property, platinum is not subject to etching by
6.5 Large-Size Flat Field Emission Cathodes 331
Fig. 6.28 Process scheme of electrophoretic application of carbon powder onto the substrate
to Van der Waals forces and formation of a thin film consisting of the charger
metal, its oxide and hydroxide.
The electrophoresis having been over and the cathode substrate having been
coated with an emitting layer, the cathode substrate is dried.
The above-mentioned operations result in a cathode coated with ground elon-
gated carbon-fiber particles and a given emitter-arrangement geometry, the particles
being oriented at right angle to the cathode surfaces and having a high adhesion to
the substrate.
This cathode-fabrication method provides the vertical orientation of particles on
the substrate, which greatly improves the emissivity of the resulting cathode in
comparison with the one where carbon-coating particles are located on the substrate
at random. Due to anisotropy of field-emission properties of carbon fibers, emission
from the particle end face is much better than from its lateral face. As a result, the
running voltages of the cathode based on vertically-oriented particles on the sub-
strate drop twice or thrice in comparison with those of the cathode based on
horizontally-arranged carbon-coating particles (Fig. 6.29). Moreover, the drop in
the running voltages is provided by increase in the field-enhancement factor at the
ends of the fibers in comparison with such enhancement on polydisposed fibers.
The large value of the emission current is in line with the cathode produced in
presence of the magnetic field. It is nearly five times as large as the current of the
cathode produced without the magnetic field (Fig. 6.29a). As the slope angles of the
curves plotted in Fowler–Nordheim coordinates practically coincide, this difference
cannot be caused by reshaping of the emission centers or the electronic work
function (Fig. 6.29b). It is most likely caused by increase in the number of the
emission centers. The emission centers have increased due to orientation of the
carbon particles affected by the magnetic field during electrophoretic deposition.
Fig. 6.29 Volt-ampere characteristics of the field emission cathodes manufactured by the method
of electrophoretic coating: a in current-voltage coordinates; b in Fowler–Nordheim coordinates; 1
characteristics of the field emission cathode manufactured without magnetic field; 2 characteristics
of the field emission cathode manufactured in presence of a magnetic field
334 6 Carbon-Based Field-Emission Cathodes
d ¼ gbh; ð6:13Þ
Fig. 6.30 Volt-ampere characteristics of the field emission cathodes manufactured at various
temperatures of annealing: a in current-voltage coordinates; b in Fowler–Nordheim coordinates; 1
temperature of annealing is 460 °C; 2 temperature of annealing is 500 °C; 3 temperature of
annealing is 520 °C
336 6 Carbon-Based Field-Emission Cathodes
time is about 60 min. After that, the second current-voltage characteristic is mea-
sured. The measurement shows the effect of the aging on emissivity of printed field
emission cathodes.
The lowest-voltage characteristic pertains to the cathode annealed at 500 °C
(Fig. 6.30). The characteristic of the field emission cathode annealed at 460 °C lies
in a higher-voltage area. The characteristics plotted in the Fowler–Nordheim
coordinates make it possible to recover the form factor and the area of the emitting
surface. The form factor of the field emission cathode annealed at 500 °C is the
largest among the form f actors of field emission cathodes. The reason is that the
annealing temperature of 500 °C brings out the fiber structure best of all other
temperatures. Apparently, the temperature of 460 °C does not provide complete
evaporation of the binder, which considerably affects the topography of the
resulting cathode. The annealing temperature of 520 °C is uprated anyway as the
carbon-fiber particles experience considerable oxidation. Field-emission research
confirms the assumptions made after the topographic investigation of field emission
cathodes exposed to thermal annealing under different conditions.
A side benefit can be thermal annealing in a special-gas atmosphere; e.g., paper
[133] attempts to anneal the coated plate on the basis of carbon nanotubes in the
argon atmosphere. However, the experiments show deterioration of emissivity as
compared with annealing in air. The reason is most likely the rests of the binder in
case of annealing in the argon atmosphere.
As the effective functioning of field-emission cathodes is widely known to need
their aging, development of field emission cathode aging techniques may be of
crucial importance for the wide practical use of printed carbon-based field emission
cathodes in electronic devices; e.g., paper [134] reports annealing of the applied
paste at 350 °C for 1 h followed by plasma etching of a 20 20 mm2 sample in a
gas mixture, the composition being undisclosed. The wattage is 300, the time being
40 and 160 min. Some increase in the emission current is only observed at increase
in the treatment duration.
According to the authors [135], pulsed plasma treatment triplicates the service
life of the field emission cathode. The treatment is carried out in the diode mode in
the Ne/Xe and Ne/Kr gas mixture at the pressure of *20 mm Hg. The variation of
plasma-treatment parameters, viz. the gas-mixture composition, voltage, pressure
and treatment time, makes it possible to find the optimum treatment of printed field
emission cathodes.
The field-emission properties of the field emission cathodes produced by
printing can be improved by addition of a dilatant material [136]. When carbon is
annealed, the film is fractionated, which increases the form factor, thereby
increasing the emission current.
The special areas that can bring success are a combination of printing with
different technologies. Let us give some examples of such combinations. It is worth
noting at once that such work has just begun. Paper [137] suggests the technology
combining photolithography and ink-jet printing. The flowchart of fabrication of
such a field emission cathode is shown in Fig. 6.31. Photolithography is used to
fabricate an open-hole array structure (Fig. 6.31b). Then the surface is
6.5 Large-Size Flat Field Emission Cathodes 337
water-proofed. That prevents from carbon-ink bleeding over the substrate and
reduces the leakage current between the modulator and the field emission cathode.
The next stage is carbon ink jet printing and drying (Fig. 6.31d) of the resulting
structure. Finally, the extraction electrode and the field emission cathode are
properly spaced by ion etching of the cathode structure in pure oxygen.
Paper [137] reports of fabrication of a structure containing 51 51 elements,
each of 100 µm in diameter, spaced at a pitch of 200 µm.
emission cathodes made of structural materials can be very useful in some cases.
Certainly, the greatest benefit expected from them is achievement of heavy
field-emission currents. However, some further improvement can make them more
universal.
The first field emission cathodes of this type are those made of high-strength
graphite, e.g., MPG-6 and MPG-7 [138–141]. These cathodes are produced in the
simplest process known that mostly stipulates only machining of suitable billets
[138–140], which is naturally followed by current aging. The microroughness can
also be further improved by other methods, e.g., radiation treatment [141].
Porous glassy carbon is also of interest because field emission cathodes based on
it show promising field-emission parameters [142–144]. This material consists of a
large number of pores widely ranging in size, the manufacturer reporting the
variation from 10 to 100 pores/in. The edges of these pores are sharp blade-like
structures, which potentially makes them good field emitters.
Paper [145] suggests forming nanoporous carbon on a conductive substrate, the
nanopore size ranging from 0.6 to 4.0 nm. Such nanoporous carbon results from
thermochemical treatment of the periodic table group III–IV carbide emitters with
chlorine at 400–1100 °C. Chlorination provides complete removal of the
carbide-forming element from the carbide composition. The chlorination is fol-
lowed by removal of chlorine from the nanoporous carbon lattice by treatment with
inert gases, hydrogen or ammonia at 500–700 °C. That is the way to form almost
any area of a field emission cathode.
Due to their structural features, carbon materials cannot provide identical form
factors over the entire surface of the extended field emission cathode or different
specimens of field emission cathodes. In other words, finding the way to provide
reproducibility of carbon field-emission cathodes is crucial for wide practical
application of such field emission cathodes. Now there are several ways to solve
this problem, which have to be closely integrated.
Forming (ageing) is among the main ways to provide stable field emission from
carbon field emission cathodes [20]. Forming results from the effect of residual-gas
ions and ponderomotive loads of the applied electric field on the effective area of a
field emission cathode. Practically, it results from controlled stage-up or scale-up in
the emission current.
Field emission cathodes are formed on the basis of carbon to create the maxi-
mum number of emission centers uniformly spread over the surface of the field
6.6 Emissivity Improvement 339
emission cathodes and making nearly the same contribution to total emission
current, i.e. to create a well-developed emitting surface.
Apparently, field emission cathodes are formed according to the following
mechanism. Ion bombardment easily destroys the amorphous filler between the
crystallites, which enhances the field effect on them, thereby increasing pondero-
motive loads. That is accompanied by increase in the current from this region and
the density of the ion flow on it. The enhanced field can destroy the
cathode-material conglomerates coupled to the starting sheet in the weakest way.
So the physical meaning of forming of carbon field emission cathodes is
exposure of the internal skeletal structure of the cathode material with formation of
a dynamically equilibrium configuration of the emitting surface. At the same time,
forming to the level of the rated operating current does not result in exposure of the
skeleton in case of low currents. In this case, forming should proceed at increase in
the current up to its threshold value, when the exposure of the skeletal structure has
been over and the emission characteristics of the cathode have become stable.
As the ion concentration of residual gases directly depends on the pressure,
reduction of the pressure can reduce the forming time.
Development of the forming method assumes de-energisation if the vacuum
deteriorates to the threshold value.
The essence of the aging method is as follows. A high-voltage power unit is set
to the current stabilization mode. The value of the current is selected so that
functioning of the cathode over a short period of time deteriorates the chamber
vacuum. That is accompanied by intensive cathode destruction by ion bombard-
ment and forming of the optimum cathode surface structure. However, this process
cannot go on for a long time as increase in gas concentration in the vicinity of the
cathode intensifies bombardment and causes a snowballing process, leading to
disruption of the anode-cathode gap and irreversible destruction of the cathode. To
avoid that, the high voltage has to be switched off at a certain time until the vacuum
has improved; i.e., all evolved gases have been pumped away. After that, the
voltage is switched back on and the cycle is repeated until the cathode has stopped
gassing while functioning.
The value of the current suitable for aging is selected depending on the cathode
area and the method of its fabrication. For a particular cathode, this value is
determined by experiment.
6.6.2 Pre-treatment
Reduction of the forming time often necessitates pre-treatment of the field emission
cathodes produced. That somehow improves their geometrical characteristics and
increases the form factor. Let us give some examples of such a treatment below.
First of all, it is ion and plasma treatment of produced flat field emission cath-
odes, which has been detailed in previous sections. Carbon-fiber bundles can be
plasmachemically treated by corona discharge in air, the method having been
340 6 Carbon-Based Field-Emission Cathodes
developed and approved. Such treatment can essentially improve emission char-
acteristics of the carbon-fiber cathode [146, 147].
When a field-emission cathode consisting of a carbon-fiber bundle functions,
electrostatic forces have been established to make the peripheral fibers divert. These
diversions result in oscillation of the emission current under pulse operation.
The matter is that the mechanical oscillations can tear off some fibers from the
bundle. The torn-off fibers can short-circuit the interelectrode cathode-modulator
gap. That used to be the case at a large amplitude of emission-current oscillation.
This situation is unacceptable because the purpose of the work done is development
of a reliable design of a cathode-modulator assembly with stable characteristics. So
it is necessary to somehow decrease or completely eliminate the diversion of the
peripheral bundle fibers.
The effect of electrostatic forces diverting the peripheral bundle fibers can be
reduced by shaping the fiber bundle so that it can provide maximally identical
electric field of all fibers in the bundle; i.e., cut some fibers projecting from the
bundle, as well as shorten the peripheral fibers. It is technologically difficult to do
that by machining of the fibers. To achieve this, a new method to treat a cathode
made of a carbon-fiber bundle has been suggested. It is plasmachemical etching of
the cathode carbon-fiber bundle by corona discharge in air as corona discharge is
closest to field emission in its nature.
The effect of the corona discharge on the carbon fibers [148, 149] is that
bombardment of the cathode with oxygen ions O2 causes oxidation of carbon C,
which leads to etching of the cathode material. So the length of a particular fiber
projecting from the fiber bundle decreases until its electric field enhancement factor
is less than or equal to other fibers from the bundle. Then the corona process
involves other fibers as well.
Accordingly, that equalizes the electric field on the surface of the carbon-fiber
bundle. In its turn, the equalization affects functioning of the field-emission cathode
in vacuum. So when the cathode is switched on, a considerable part of fibers begins
emitting because the unetched structure of the bundle begins electron emission from
particular fibers projecting from the overall mass.
Figure 6.32 represents some photos of the cathode carbon-fiber bundle taken
with an optical microscope. The bundle has been etched by corona discharge in air
under the conditions described above.
Figure 6.32 shows that corona-discharge etching makes the carbon-fiber bundle
round, the projecting fibers having been trimmed and the peripheral fibers
shortened.
The carbon-fiber bundle is etched away for a quite long distance (Fig. 6.32a, b),
and it is safe to say that almost all fibers of the bundle have taken part in the corona
process.
Under the above-stated conditions of plasmachemical treatment, the carbon
material of the bundle consisting of *300 fibers is etched in air at the rate of
*5 µm/min.
6.6 Emissivity Improvement 341
Fig. 6.32 Bunch of carbon fibers before (a, b) and after (c, d) corona discharge etching in air
Field-emission tests of the resulting samples are carried out in a diode structure.
The residual pressure in the chamber is *10−6 Torr. This paper only refers to the
field-emission tests of barium-doped samples. At first, the study focuses on the
features of MPG-6 pure graphite.
Figure 6.33 gives the current-voltage characteristics of one of such samples and
a series of current-voltage characteristics for a pure cathode under the same con-
ditions of field-emission tests.
The Ba-implanted sample demonstrates a considerable bias of the feature into
the area of low voltages, which may indicate increase in effectiveness of the
cathode. Here cathode efficiency is understood as the anode-cathode voltage nec-
essary to attain the preset emission-current value.
Similar results have also come for other barium salts [157], as well as for carbon
nanotubes [158] and graphite foil [159]. The latter case involves impregnation with
BaBr2 barium salt in water for 1 min in a saturated solution. According to the
measurements taken with an atomic-force microscope, the electronic work function
has gone down from 4.32 (cf. points 5–8 in Fig. 6.34) to 3.05 (cf. points 1–4).
Figure 6.34 shows these points with a white line. As a result, the current-voltage
characteristics have considerably biased into the low-voltage area.
I B
ln ¼ Aþ ; ð6:15Þ
U2 U
its slope ratio giving us parameter B = −B′, and its crossing with the vertical axis
indicating parameter A = lnA′.
So in practice, it is coefficients A and B instead of cathode physical properties u,
bmean, Stotal that are really determined. However, the theoretical equation for total
emission current can relate experimentally determined parameters to real ones in the
following way
b2
A ¼ ln A0 ¼ ln Stotal a mean ; ð6:16Þ
u
u3=2
B ¼ B0 ¼ b : ð6:17Þ
btotal
does not exceed the typical time of variation of cathode parameters, viz. Tcat.
Figure 6.35 shows the characteristics taken in different modes.
Unfortunately, turbo characteristics are more sensitive to imperfections of
experimental equipment [161]. There are a few principal causes that affect the CVC
in Fowler–Nordheim coordinates. Let us consider them in more detail.
In practice, the turbo speed of measurement of characteristics is usually
implemented as follows. Right after the control signal has been fed to the power
unit, current and voltage are measured as frequently as possible. However, there is
always a Tcap characteristic measurement time-limit below which we cannot go. It is
6.7 Features of Measurement and Analysis … 347
connected with availability of capacitive currents in the device. On the other hand,
we are limited by the characteristic cathode-parameter variation time, viz. Tcat, so
measurement time Dt has to be selected from the ratio Tcap Dt Tcat. The time
is usually about a few seconds, so if the number of points per characteristic totals
about one thousand, the measurement period takes some milliseconds. To meet
these requirements, the signal has to be digitised by analogue-to-digital converters,
ADC. As the signal at the ADC output is discrete, there are noises that interfere
with the real characteristic.
The ideal current-voltage characteristic of a field emission cathode in Fowler–
Nordheim coordinates is known to be a straight line (Fig. 6.36a) and follow (6.15).
The digitization of the signal with an ADC discretizes the current signal; i.e., the
current is represented in the integers meeting the number of the ADC counts.
Differently, if the count value is DI, and the current value is Ireal, the measured
current value, viz. IADC, is
Fig. 6.37 Accounting of non-ideality of the device model: a accounting for non-ideality of
contacts; b accounting for leaks in the device
Ireal
IADC ¼ DI : ð6:18Þ
DI
where sign D designates truncation. Taking this fact into account, we can express
the ideal characteristic as shown in Fig. 6.36b.
The ADC noise characteristics studied at the zero current suggest that the ADC
readings can vary within several counts.
In practice, when an anode and a cathode are mounted in a vacuum chamber or a
probe, there are always some imperfections in contacts (e.g. either high contact
resistances, or the cathode is highly resistive in itself) and the device itself (e.g.
there are current leakages between the anode and the cathode). Their effect on the
measuring circuitry can be estimated by adding stray resistance R in the circuit.
Addition of a series-connected resistance describes the situation when there is a
stray resistance in contacts (Fig. 6.37a). Addition of a parallel-connected resistance
describes the situation when there is a current leakage (Fig. 6.37b).
In this case, it is obvious that the real measures of voltage Umeas and current Imeas
differ from the values of current Ipr and voltage Upr of the device.
For serial connection the following holds:
Umeas ¼ Upr þ Ipr ðUpr ÞR;
ð6:19Þ
Imeas ¼ Ipr :
high-voltage power unit in the event of short circuit. It is clear that the value of this
resistance should be selected so that it does not essentially distort the characteristic.
In case of a parallel connection, we have
Umeas ¼ Upr ;
ð6:20Þ
Imeas ¼ Ipr ðUpr Þ þ Upr =R:
In this case, the characteristic is computed in the same way as in case of a series
connection. Figure 6.38 shows the really measured characteristic as affected by
leakages.
The diagrams show that the shortcomings of the measuring bench can essentially
distort the current-voltage characteristics, which can later result in incorrect cal-
culation of approximation coefficients A and B in the Fowler–Nordheim coordi-
nates. In turn, the incorrect evaluation of parameters A and B can lead to
misidentification of the reasons of the cathode degradation. So it is necessary to pay
much attention to correctness of the measurements taken.
350 6 Carbon-Based Field-Emission Cathodes
Research shows that a skill of proper measurement of the field emission cathode
CVC is often insufficient for an adequate evaluation of real field emission cathode
parameters, including the emitting surface area, the form factor and the electronic
work function. The bibliographic-data review shows [162, 163] that calculation of
these parameters on the basis of the measured A and B interpolation coefficients is a
difficult problem since the simple Fowler−Nordheim theory does not take account
of re-shaping or changing the height of the potential barrier as affected by the field,
considering the triangular barrier only. The Murphy–Good theory [164] gives a
more exact expression for field-emission current, viz.
0 pffiffiffiffiffi 1
bU
bv c u 3=2
ab2 U 2 B u C
I ¼ S pffiffiffiffiffi expB
@ C;
A ð6:21Þ
bU bU
t c u u
2
pffiffiffiffiffiffiffiffiffiffi
where t, v are Nordheim special functions and c = 3.79469 10−5 m=V eV.
pffiffiffiffiffiffiffi
Let us designate y ¼ c bU =u as the relative decrease in the height of the
potential barrier. In this case, the equation expressed in the Fowler–Nordheim
coordinates takes the following form, viz.
I Sab2 bvðyÞu3=2
ln ¼ ln 2 : ð6:22Þ
U2 t ðyÞu bU
This equation can be written in the previous form (6.15), the only difference
being that now coefficients A′ and B′ depend on the value of the field. However,
these coefficients must not be equated with the values derived as a result of
interpolation of the experimental curve (6.15).
To account for the difference between these parameters, let us consider
Fig. 6.39. It shows a straight line approximating the experimental dependence and
two straight lines with coefficients and A′ and B′ plotted for two values of voltage
U1 and U2, assuming that parameter A′ does not depend on the field because of a
weak dependence of function t on y. The parameter values are clearly seen to differ
greatly. So if the real parameters of a field emission cathode have to be evaluated
precisely enough, the analysis to be carried out is to be more sophisticated.
In practice, the field emission cathode parameters are usually calculated as fol-
lows. As function t(y) weakly depends on the value of the field, this dependence is
neglected. In this case, the equation for coefficient B takes the following form, viz.
bu3=2 1 dv
B¼ vðyÞ y : ð6:23Þ
b 2 dy
6.7 Features of Measurement and Analysis … 351
1 dv
sðyÞ ¼ vðyÞ y ; ð6:24Þ
2 dy
pffiffiffiffiffiffiffi
bu3=2 bU
B¼ s c : ð6:25Þ
b /
At present there are two main ways to study stability of field-emission current over
a long period of time. The first of them [165] is periodic measurement of current at
fixed voltage. However, this technique is not good to study emissivity of
carbon-based FEC because of strong time instability of parameters of such cathodes
352 6 Carbon-Based Field-Emission Cathodes
Fig. 6.40 Emergence of area of acceptable values as a result of dispersion of experimental points:
a experimental points (the best line I and two other lines (II and III) that also describe this
dependence with a smaller coefficient of correlation); b areas of acceptable values A and B (the set
of points corresponding to the approximating lines. Points corresponding to I, II and III are
marked)
way. If the requirements to the correlation factor are mitigated, these points can be
connected with not one, but more straight lines, which more or less precisely
describe this dependence (Fig. 6.40a). The result is that this characteristic is rep-
resented in coordinates A–B with not one point, but an area of acceptable values
(Fig. 6.40b). Then it is clear that if we have a set of such areas, there may be
situations when it is impossible to say unequivocally whether parameters A and/or
B vary in time at all (Fig. 6.40).
In this case, it is expedient to resort to the following approach. Let there be not
much variation in one parameter, e.g., B, during the experiment. That means that the
projections of the areas of acceptable values to axis B intersect. Having fixed this
parameter at an average level, we can recalculate parameter A according to the
formula A ¼ hY iBh X i; i.e., if the time variation of coefficient B is less than the
scattering in its determination, B can be considered invariable and only dependence
of coefficient A can be analyzed. As a result, the two-parametric problem is reduced
to one-parametric.
The authors have tried this testing technique. The test specimen is carbon
nanotube based FEC. The cathode is tested in the following mode (Fig. 6.41). The
testing time totals to about 600 h. The cathode works in a direct-current mode of
100 µA throughout the experiment, the current and voltage being measured at the
device every 10 s. In addition, current-voltage characteristics of the device are
measured every 30 min. The characteristics measured determine coefficients A and
B of the Fowler–Nordheim equation.
To check if the prediction of cathode behavior in the same operating mode
within long-term periods is true, the first 100 h are cut out from the operating time
and the cathode behavior is calculated for the next 500 h according to the technique
354 6 Carbon-Based Field-Emission Cathodes
b0
A ¼ lnða1 a2 tÞ; B¼ ; ð6:27Þ
b1 t
where a1, a2, b0 and b1 are constant, t being the time. Figure 6.42 (curve II) shows
the predicted data. It demonstrates that curve II is in good agreement with the
experimentally observed dependence. So adequate application of the technique
suggested depends on preliminary identification of the main degradation mecha-
nisms of the FEC to be tested.
In addition to high performance, long service life and large area of planar field
emission cathodes, an important parameter is uniformity of emission-center dis-
tribution over the surface. These FEC are usually applied in design of light-emitting
indicators and information displays, where the problem of uniform flare of one
element/pixel is especially pointed. Moreover, this parameter is also indirectly
connected with stable cathode operation. In particular, if only a few emission
centers work efficiently, destruction of even one of them drastically changes the
characteristics of the field emission cathode. If uniform emitting comes from the
entire surface of the cathode having a great number of centers, destruction of a few
of them does not cause any considerable fluctuations of the emission current.
In practice, uniformity of emission from the surface of a planar field emission
cathode is evaluated with transparent anodic plates coated with a thin luminophor
layer to transform the flow of incident electrons into visible light. Accordingly,
when the field emission cathode functions, the so-called emission image appears at
the anode. It characterizes distribution of emission centers over the field emission
356 6 Carbon-Based Field-Emission Cathodes
cathode surface, as well as represents their efficiency, i.e. their contribution to the
total current.
Uniformity of distribution of emission centers over the FEC surface is suggested
to be measured with the following procedure. High voltage is applied between the
cathode and the anode to generate the current heavy enough to photograph the
emission image formed at the anode by a digital camcorder in a monochromic, viz.
black-and-white, mode. The resulting digital photo is kept in a personal computer,
and the appropriate software coverts it into a histogram of distribution of luminosity
over the anode surface, an isocandle histogram, according to the following formula
X
hk ¼ yk xi;j ; ð6:28Þ
i;j
where i is the coordinate of the image point in the line, j is the coordinate of the
image point in the column, k is the number showing the light-intensity value,
0 k 255, 0 and 255 being the minimum and maximum luminosity, respec-
tively, hk is the number of image points of luminosity k, xi,j is the luminosity of the
image point with coordinates (i, j), yk is the function of the following type
1 x ¼ k;
yk ðxÞ ¼ ð6:29Þ
0 x¼6 k:
X
255
u¼ hk =M; ð6:30Þ
k¼P
Fig. 6.44 Experiment scheme. Characteristic values of times are specified: a implementation
scheme; b measurement scheme
great number of values of current and voltage, usually *5000, are taken within as
short a period of time as possible, usually just a few seconds. This way to measure
CVC makes it possible to avoid considerable variations of cathode parameters
during the measurement period.
The software developed also preprocesses the data obtained. The preprocessing
procedure is as follows. The values of current and voltage frames taken over one
measurement are averaged to find the average value and variance at the moment of
measurement. So every measurement gives 4 figures: the average current and
voltage values and their variance. These data are filed with specification of the
measurement time, the time specified being that returned by the system timer; so
even if there is a measurement bug, every couple of current-voltage values meets
the true time of measurement. Some ordering information is also filed, namely the
implementation number, the number of measurement within the implementation,
the time elapsed from the beginning of the implementation. These data make it
possible to test the software at the stage of development and promptly fix the
defects.
The software also records the files with volt-ampere characteristics, so-called
second-type files, on the hard disk of the computer. The file also keeps the
characteristic-measurement time and the data themselves, i.e., the voltage column
and the respective current column. Before being filed, the data are also reprocessed.
The current values below 5 ADC counts are cut off to eliminate the effect of ADC
noise discretization. Then the characteristic obtained is recalculated in the Fowler–
Nordheim coordinates and approximated by a straight line to find coefficients A and
B. These parameters are also filed. To accelerate the recording, viz. reduce the
errors of data desynchronization in between the implementations, not all the data
are filed but only *500 values uniformly selected from the entire array.
Before the beginning of the long-term test, a test circuit for the particular cathode
is laid. With this end in view, a file is created to contain (1) t1 as the period of time
when high voltage is switched on; (2) t2 as the period of time when high voltage is
switched off; (3) U as the maximum voltage; (4) I as the peak cathode current,
depending on which of these restrictions works up earlier, there is either current or
voltage stabilization; (5) N1 and N2 as the number of volt-ampere characteristics
measured at the beginning and at the end of the implementation, respectively.
The selected experimental-arrangement algorithm makes it possible to carry out
long-term tests of almost any complexity; e.g., it is possible to carry out both simple
long-term tests in the constant voltage or current maintenance mode and sophisti-
cated tests with periodic measurements of cathode characteristics. One more
advantage of this experimental arrangement is the simplicity of adaptation of
measuring to specific targets; i.e., long-term test arrangement can be amended
without a radical change of the control program, the only amendment necessary
being that of the file containing the experimental arrangement.
References 361
References
1. E.P. Sheshin, Structure of a Surface and Autoissue Properties of Carbon Materials (M of a.:
MFTI-fizmatbook, 2001)
2. E.P. Sheshin, Field emission of carbon fibers. Ultramicroscopy 79, 101–108 (1999)
3. E.P. Sheshin, Properties of carbon materials, especially fibers, for field emitter applications.
Appl. Surf. Sci. 215, 191–200 (2003)
4. E.P. Sheshin, Modern ways of formation of field electronic cathodes from carbon materials.
Achievements Mod. Radio Electron. 5–6, 36–40 (2004)
5. Kypryshkin, E.P. Sheshin, A.A. Shyka, Methods of production of field electronic cathodes
from carbon materials. Nano Microsc. Equip. 3, 26–31 (2005)
6. E.P. Sheshin, Possibility of receiving big autoissue currents from autocathodes from carbon
volokn. Electron. Equip. Ser. 4. 2, 58–62 (1988)
7. E.P. Sheshin, L. Rybakov Yu, Field electronic cathodes from carbon volokn. Theses dokl.
xviii Vsesoyuzn. Konf. on issue electronics, pp. 213–214, M.: nauka (1981)
8. V.A. Nevrovsky, V.I. Rahovsky, To a question of time of development of thermal instability
of microledges on the cathode at vacuum breakdown. ZhTF 50, 2127–2135 (1980)
9. B.V. Bondarenko, V.I. Makyha, V. Titov Yu, E.P. Sheshin, Autocathodes with a big
working area. Electron. Equip. (It is Gray. SVCh Electron) 4, 47–51 (1986)
10. A.S. Fialkov, Carbon and Graphite Materials (M.: energiya, 1979), p. 320
11. E. Braun, J. Smith, D. Sykes, Carbon fibers as field emitters. Vacuum 25(9/10), 425–426
(1975)
12. E.P. Sheshin, Emission characteristics carbon fiber, in Physical Processes in Devices of
Electronic Equipment (M.: MFTI, 1980), pp. 6–10
13. R.M. Hatapova, V.H. Romanova, About issue stability of carbon autocathodes in the
unsoldered devices. Theses дoкл. IV Vsesoyuzn. A symposium on not heated cathodes,
Tomsk, ISE, p. 51 (1980)
14. B.V. Bondarenko, V.A. Sileverstov, E.P. Sheshin, Issue properties carbon вoлoкн various
temperature of processing. Radio Technician Electron. Eng. 30(8), 1601–1605 (1985)
15. B.V. Bondarenko, V.I. Makyha, E.A. Tishin, E.P. Sheshin, About work of an exit of
electrons of carbon materials, in The Physical Phenomena in Devices of an Electronic and
Laser Teznika (M.: MIPT, 1983), pp. 13–18
16. B.V. Bondarenko, V.I. Makyha, E.P. Sheshin, Stability of issue and durability of some
options of autocathodes. Radio Technician Electron. Eng. 28(8), 1649–1652 (1983)
17. B.V. Bondarenko, A.G. Shahovskoy, E.P. Sheshin, Stabilization of autoissue characteristics
carbon вoлoкн during the long work, in Physical Yaleniye in Devices of Electronic and
Laser Equipment (M.: MIPT, 1985), pp. 4–9
18. A.S. Baturin, T. Kelly, M.S. Mousa, E.P. Sheshin and others, Lifetime and emission stability
of carbon fiber cathodes. Mat. Sci. Eng. 353, 22–26 (2003)
19. B.V. Bondarenko, V.I. Makyha, E.P. Sheshin, Field electronic cathodes from graphite.
Theses dokl. iv Vsesoyuzn. A symposium but to not heated cathodes, Tomsk, pp. 49–50
(1980)
20. V.I. Makyha, E.P. Sheshin, About possibility of receiving big autoissue currents from
graphite, in The Physical Phenomena in Devices of Electronic and Laser Equipment (M.:
MIPT, 1983), pp. 22–25
21. E.P. Sheshin, A.C. Batyrin, Stability field electron emission uglerodnovolokonnykh of
autocathodes. Mater. The All-Russian symposium on Amis. To electronics, Ryazan,
pp. 141–142 (1996)
22. E.P. Sheshin, A.B. Stolarov, A.V. Anashenko, A.H. Kydravcev, N.B. Dyakonova, Studying
of issue properties and structure carbon вoлoкн various types. Mater. VII meetings
“Radiation physics of a solid body”, Moscow-Sevastopol, pp. 78–82 (1997)
23. E.P. Sheshin, A.V. Anaschenko, S.G. Kuzmenko, Field emission characteristic research of
some type of carbon fibers. Ultramicroscopy 79, 109–114 (1999)
362 6 Carbon-Based Field-Emission Cathodes
24. D.M. Paytov, V.A. Tkachenko, V.S. Neshpor, E.I. Podolskay, Mnogoostriyny cold cathode.
Ampere-second. USSR, No. 767858. Cl. H01j 1/30 of 14.06.78
25. D.M. Paytov, V.A. Tkachenko, V.H. Fedorov, E.V. Gorbachevskiy, Mnogoostriyny cold
cathode. Ampere-second. USSR, No. 1019518. Cl. H01j 1/30 of 15.01.82
26. A.G. Chaknovskoi, E.P. Sheshin, A.S. Kupryashkin, V.A. Seliverstov, Method of fabrication
of matrix carbon fiber field emission cathode structure for flat-panel indicators. J. Vac. Sci.
Techn. 11(2), 511–513 (1993)
27. T.N. Bredihina, V.K. Esina, A.U. Cherepanov, A.G. Shohovskoy, E.P. Sheshin, Way of
production of the multiemitter autocathode. Ampere-second. USSR, No. 151606. Cl. H01j
9/02 of 15.10.87
28. A.S. Batyrin, V.I. Beloglazov, V.F. Lebedev, N.B. Skibina, A.V. Sherbakov, E.P. Sheshin,
Way of production of the matrichesky autocathode. Stalemate. Russia No. 2183362. Cl.
H01j 1/14 of 04.04.2001
29. R.R. Bessette, M.G. Madeiros, C.J. Patrissi, C.M. Deschenes, C.N. La fratta, Development
and characterization of a novel carbon fiber based cathode for semi-fuel cell applications.
J. Power Sour. 96, 240–244 (2001)
30. T. Keesmann, W.H. Grosse, Field emission cathode using carbon fibers. Pat. USA
2004/0036402. Cl. H01j 1/05 (313/311), of 8.04.2003
31. M.A. Guillorn, M.L. Simpson, G.J. Bordonaro, V.J. Merkulov, L.R. Baylor, D.H. Lowndes,
Fabrication of gated cathode structures using an in situ grown vertically aligned carbon
nanofiber as a field emission element. J. Vac. Sci. Tech. 19(2), 573–578 (2001)
32. V.D. Blank, E.V. Polyakov, D.V. Batov, B.A. Kulnitskiy, U. Bangert, A. Gutierrez-Sosa, A.
J. Harvey, A. Seepujak, Diam. Rel. Mater. 12, 864–869 (2003)
33. Y. Wada, Y.K. Yap, Y. Mori, M. Yoshimura, T. Sasaki, Diam. Rel. Mater. 9, 620–624
(2000)
34. R. Kurt, J.M. Bonard, A. Karimi, Thin Solid Films 398, 193–198 (2001)
35. A.G. Kydashov, A.V. Okotryb, N.F. Udanov, A.I. Romanenko, L.G. Bylysheva, A.G.
Abrosimov, A.L. Chyvilin, E.M. Pazhetov, A.I. Boronin. FTT 44(4), 626–629 (2002)
36. M. Terrones, XV International Winterschool on Electronic Properties of Novel Materials
(Austria, 2001), p. 63
37. C. Kimura, Y. Yamamuro, H. Aoki, T. Sugino, 17th European Conference on Diamond,
Diamond-Like Materials, Carbon Nanotubes, and Nitrides (Portugal, 2006), p. 380
38. V.S. Bormashov, E.P. Sheshin, D.V. Baton, V.D. Blank, S.G. Buga, Novel method of flat
cold cathode production from carbon-nitrogen nanofibers. Technical Digest of 19th IVNC
and 50th IFES, Guilin, China, p. 293 (2006)
39. V.D. Blank, E.V. Polyakov, B.A. Kulnitskiy, A.A. Nuzhdin, L. Alshevskiy Yu, U. Banget,
A.J. Harvey, Thin Solid Films 346, 86–92 (1999)
40. P.M. Ajayan, M. Terrones, A. De La Guardia, V. Hue, N. Grobert, B.Q. Wei, H. Lezec, G.
Ramanath, T.W. Ebbe, Science 296(5568), 705 (2002)
41. N.G. Shang, F.C.K. Au, X.M. Meng, C.S. Lee, I. Bello, S.Y. Lee, Chem. Phys. Lett. 358,
187–191 (2002)
42. Z.Y. Kosakovskay, L.A. Chernozatonskiy, E.A. Fedorov, Lett. ZhETF 56(1), 26–30 (1992)
43. S. Fan, M.G. Chapline, N.R. Franklin, T.W. Tombler, A.M. Cassell, H. Dai, Science 283,
512–514 (1999)
44. A.G. Rinzler, J.H. Hafner, P. Nikolaev, L. Lou, S.G. Kim, D. Tomanek, P. Nordlander, D.T.
Colbert, R.E. Smalley, Science 269, 1550–1553 (1995)
45. Q.H. Wang, T.D. Corrigan, J.Y. Day, R.P.H. Chang, A.R. Krauss, Appl. Phys. Lett. 70,
3308–3310 (1997)
46. J.-M. Bonard, J.-P. Salvetat, T. Stockli, W.A. De Heer, L. Forro, A. Chatelain, Appl. Phys.
Lett. 73, 918–920 (1998)
47. H. Schmid, H.-W. Fink, Appl. Phys. Lett. 70, 2679–2680 (1997)
48. W.A. De Heer, A. Chatelain, D. Ugarte, Science 270(5239), 1179 (1995)
49. V. Gulaev Yu, N.I. Sinicin, G.V. Torgashov, L.A. Chernozatonsky, Z.Y. Kosakovskay, F.
Zaharchenko Yu, Microelectronics 26(2), 84–88 (1997)
References 363
50. V. Filip, D. Nicolaescu, M. Tanemura, F. Okuyama, J. Vac. Sci. Technol. 21, 382–390
(2003)
51. T.C. Choy, A.H. Harker, A.M. Stoneham, J. Phys. Matter. 16, 861–880 (2004)
52. T.C. Choy, A.M. Stoneham, A.H. Harker, J. Phys. Condens. Matter 17, 1505–1528 (2005)
53. A.N. Obraztsov, I. Pavlovsky Yu, A.P. Volkovv, J. Vac. Sci. Technol B. 17, 674–678 (1999)
54. A.N. Obrazcov, A.P. Volkov, I. Palovskiy Yu, A.L. Chyvilin, N.A. Rydina, V.L. Kyznetsov,
Lett. ZhETF 69(5), 381–386 (1999)
55. A.N. Obraztsov, J. Pavlovsy, A.P. Volkov, V.L. Kuznetsov, A.L. Chuvilin, Oriented carbon
nanotubes growth for field emission application. Mat. Res. Soc. Symp. Proc. 558, 111–116
(2000)
56. H. Hiura, T.W. Ebbesen, J. Fujita, K. Tanigaki, T. Takada, Nature 148–151 (1994)
57. F. Salver-Disma, J.-M. Tarascon, C. Clinard, J.-N. Rouzaund, Carbon 37, 1947–1959 (1999)
58. J.K. Huang, H. Yasuda, H. Mori, Chem. Phys. Lett. 303, 130–134 (1999)
59. Y. Omaru, M. Nagamine, J. Appl. Phys. 88, 6389–6377 (2000)
60. K.E. Robinson, D.D. Edie, Carbon 34, 13–36 (1996)
61. S.-H. Hong, Y. Korai, I. Mochida, Carbon 38, 805–815 (2000)
62. M.I. Elinson, Not Heated Cathodes (M.: Soviet Radio, 1971)
63. V.T. Binh, C. Adessi, Phys. Rev. Lett. 85, 864–867 (2000)
64. V.M. Lobanov, ZHTF 75(11), 92–96 (2005)
65. V.D. Frolov, A.V. Karabutov, S.M. Pimenov, V.I. Konov, V.P. Ageev, Diam. Rel. Mater.
10, 1719–1726 (2001)
66. M.S. Dresselhaus, G. Dresselhauss, P. Eklund, The Science of Fullerenes and Carbon
Nanotubes (Academic, 1996)
67. T.W. Ebbesen (ed.), Carbon Nanotubes, Preparation and Properties (CRC Press, 1996)
68. R. Saito, G. Dresselhaus, M.S. Dresselhaus, Physical Properties of Carbon Nanotubes
(World Scientific, 1998)
69. A.V. Eleckiy, Carbon nanotubes and their issue svostvo. UFN 172(4), 401–438 (2002)
70. Y. Zhang, S.Z. Deng, J. Cheng, N.S. Xu, Synthesis and Field Emission Properties of
Large-Area Uniform Carbon Nanotubes Films (Techn Didgest JVNC, Oxford, UK, 2005),
pp. 324–325
71. C. Zhu, C. Lou, W. Lei, X. Zhang, Fabrication and Characterization of High Current
Density Carbon Nanotubes Cold Cathodes (JVESC, Beijing, China, 2004), pp. 237–238
72. N.Y. Huang, J.C. She, J. Chen, S.Z. Deng, N.S. Xu, H. Bishop, S.E. Huq, L. Wang, D.Y.
Zhong, E.G. Wang, D.M. Chen, Phys. Rev. Lett. 93(7), 075501-1 (2004)
73. T. Shiroishi, T. Sawada, A. Hosono, S. Nakata, Low Temperature Growth of Carbon
Nanotubes by Thermal CVD with FeZrN Catalyst (Techn Didgest JVMC, Osaka, Japan,
2003), pp. 13–14
74. A.N. Rodokin, L.V. Malarevich, Receiving carbon nanofibre and nanotubes by method of
ultrahigh heating of vapors of ethanol. Inorg. Mat. 39(4), 433–437 (2003)
75. A.M. Lamanov, R.M. Ibragimov, K.N. Nikolskiy, A.N. Redkin, R.G. Chesov, E.P. Sheshin,
The autoissue cathodes manufactured by method of low-temperature gas sedimentation of
ethanol pa. Nano Microsyst. Equip. 9, 34–37 (2005)
76. A.M. Lamanov, E.P. Sheshin, A.N. Redkin, New Technique of Field Emission Cathodes
Preparation by Low Temperature Deposition From Ethanol Vapor. Hydrogen Materials and
Chemistry of Carbon Nanomaterials (Springer, 2007), pp. 265–268
77. Z. Sun, Y.J. Li, G.Y. Chen, S.P. Lau, B.K. Tay, J.S. Chen, L.K. Chean, Fabrication of
carbon nanotubes film arrays film arrays for field emission flat panel display application.
Surf. Rev. Lett. 8(5), 505–508 (2001)
78. J.I. Sohn, S. Lee, Y.-H. Song, S.-Y. Choi, K.I. Cho, K.S. Nam, Patterned selective growth of
carbon nanotubes and large field emission from vertically well-aligned carbon nanotubes
field emitter arrays. Appl. Phys. Lett. 78(7), 901–903 (2001)
79. G.S. Choi, K.H. Son, D.J. Kim, Fabrication of high performance carbon nanotubes field
emitters. Microelectron. Eng. 66, 206–212 (2003)
364 6 Carbon-Based Field-Emission Cathodes
80. G.S. Choi, S.S. Cho, S.Y. Hong, J.B. Park, K.H. Son, D.J. Kim, J. Appl. Phys. 91, 3847
(2002)
81. Y.T. Jang, C.H. Choi, B.K. Ju, J.H. Ahn, Suppression of leakage current via formation of a
sidewall protector in the microgated carbon nanotubes emitter. Nanotechnology 14, 497–500
(2003)
82. Y.H. Lee, Y.T. Jang, D.J. Kim, J.H. Ahn, B.K. Ju, Realization of gated field emitter for
electrophotonic applications using carbon nanotubes line emitters directly grown into
submicrometer holes. Avd. Mater. 13(7), 479–482 (2001)
83. S. Kang, C. Bae, W. Son, M. Kim, J. Yi, A. Chang, J. Kim, C. Lee, Low Temperature
Carbon Nanotubes for Field Emission Display (Techn Didgest JVMC, Osaka, Japan, 2003),
pp. 51–52
84. C. Lou, X. Zhang, W. Lei, C. Qi, New method to fabricate field-emission cathode of carbon
nanotubes. Appl. Surf. Sci. 251, 254–257 (2005)
85. J.W. Moon, Electron emission source composition for flat panel display and method of
producing electron emission source for flat panel display using the same. Pat. US
2006/0066200 of 30.03.2006. Cl. 313/311 (H01j 1/14)
86. T. Honda, W. Rochanachirapar, K. Murakami, K. Ohsumi, N. Shimizu, S. Abo, F. Wakaya,
M. Takai, KrF Laser Surface Treatment of CNT Cathode (Tech Digest IVNC, Oxford, UK,
2005), pp. 300–301
87. X.H. Liang, N.S. Xu, I. Chen, S.Z. Deng, Enhanced Field Emission from Carbon Nanotube
Films by Hydrogen Glow Discharge Treatment (Tech Didgest IVNC, Oxford, UK, 2005),
pp. 302–303
88. J. Robertson, Diamond Relat. Mater. 3, 361 (1994)
89. J. Robertson, Amorphous carbon cathodes for field emission display. Thin Solid Films 296,
61–65 (1997)
90. V.J. Merkylov, D.H. Lowndes, L.R. Baylor and others, An addressable field emission array
for E-beam lithography using planar, pulsed-laser deposited amorphous diamond cathodes.
IVMS 178–179 (1998)
91. X. Zhang, Z. Lu, B. Zhang, N. Yao, B. Ma, Y. Zhao, Preparation of Nanostructure
Amorphous Carbon Film and Its Field Emission Properties (Techn Digest IVNC, 2004),
pp. 293–295
92. S. Bae, K.H. Park, S. Lee, K.H. Koh, Triode Emitters with Well-Structure Cathode (Techn
Diest IVNC, 2004), pp. 26–27
93. K.H. Park, K.M. Lee, S. Choi, S. Lee, K.H. Koh, J. Vac. Sci. Tehnol. 19, 946 (2001)
94. K.H. Park, W.J. Seo, S. Lee, K.H. Koh, Appl. Phys. Lett. 81, 358 (2002)
95. W. Knapp, D. Schleussner, Field emission characteristics of carbon buchypaper. J. Vac. Sci.
Tehnol. 21(1), 557–561 (2003)
96. W.J. Zhao, R. Wasu, M. Takai, Field Emission from Carbon Nanotube Material (Techn
Digest IVNC, 2003), pp. 57–58
97. W. Weibiao, X. Yuxue, Z. Chuanping, Electron Emission of Carbon Nanonets (Techn
Digest IVNC, 2003), pp. 155–156
98. H. Busta, Z. Tolt, J. Montgomery, A. Feinerman, Field Emission from Teepee-Shaped
Carbon Nanotube Bundles (Techn Digest IVNC, 2004), pp. 30–31
99. B.C. Halloway, M. Zhu, X. Zhao, J. Wang, R. Outlaw, Milliamp-Class Field Emission
Devices Based on Free-Standing, Two-Dimensional Carbon Nanostructures (Techn
Digest IVNC, 2005), pp. 24–25
100. V. Gulyaev Yu, V.N. Koro, A. Grigorev Yu, G.A. Rehn, Research of the Field Emission of
Fractal Blads Patterns of Large Length (Techn Digest IVNC, 2004), pp. 78–80
101. A. Kastalsky, S. Shokhor, J. Hou, S. Naar, N. Abianshin, B. Gorfinkel, Thin Film
Edge-Emitter Field Emission Flat Panel Display (Techn Digest of SiD, 2001), pp. 201–203
102. S.K. Gordeev, A.M. Lamanov, R.M. Jbragimov, K.N. Nikolskiy, E.P. Sheshin, Pyrolitic
Carbon Cathodes Prepared by Low Temperature Vapor Deposition (Techn Digest IVNC,
2005), pp. 176–177
References 365
103. S.K. Gordeev, S.B. Korchagina, A.M. Lamanov, E.P. Sheshin and other, New technique of
production of cathodes from a graphite foil. Nano Microsyst. тexникa. 12, 33–36 (2005)
104. V.P. Vereyko, M.M. Libenson, A.M. Miluchev, Laser technology. Electronics 68(137), 114
(1970)
105. M.F. Stelmakh, Lasers in technology under the editorship. Energy (1975)
106. V. Polezhaev Yu, F.B. Yurivech, Thermal Protection (1975)
107. K.N. Nikolskiy, A.C. Batyrin, A.C. Kyprashkin, R.G. Chesov, E.P. Sheshin, “Effect of a
ring” at intensive field electronic issue. Microsyst. Equip. 7, 8–10 (2003)
108. K.N. Nikolskiy, A.C. Batyrin, A.I. Knazev, R.G. Chesov, E.P. Sheshin, Formation of rings
round primary field emission image and possibility of their practical use. Mag. Tech. Phys.
74(2), 110–113 (2004)
109. Y.J. Li, S.P. Lau et al., Oriented carbon microfibers grown by catalytic decomposition of
acetylene and their field emission properties. Diamond Relat. Mat. 10(1), 878–883 (2001)
110. L.C. Jin, P. Jeunghee, Growth and field electron emission of vertically aligned multiwalled
carbon nanotubes. Chem. Phys. Latters 326(1), 175–181 (2000)
111. Q.H. Wang, R.P.H. Chang, M. Yam, Flat panel display prototype using gated carbon
nanotube field emitters. Appl. Phys. Lett. 78(9), 1294–1297 (2001)
112. Z. Lan, Z. Binglin et al., A Flat Panel Display Device Fabricated by Using Carbon
Nanotubes Cathode. IEEE, IVMC 2001 University of California, Davis, California, USA,
12–16 August 2001
113. Z. Wei, G. Kochanski et al., Large current density from carbon nanotube field emitters.
Appl. Phys. Lett. 75(6), 873–876 (1999)
114. S. Gleston, Introduction to Electrochemistry (M.: Foreign Literature, 1951)
115. E.M. Lifshic, L.P. Pitaevskiy, Theoretical Physics of T. 10. Physical Kinetics (M.; Science,
1979)
116. K. Fetter, Electrochemical Kinetics (M.: Chemistry, 1967)
117. G.A. Golikov, Guide to Physical Chemistry (M.: The Highest Shkola, 1988)
118. D.S. Dykin, B.V. Deragin, Electrophoresis (M.; Science, 1976)
119. D.A. Kurosov, A.S. Baturin, K.N. Nikolskiy, R.G. Tcheov, E.P. Sheshin, Influence of the
interelectrode distance in electrophoretic cold cathode fabrication of the emission uniformity.
Appl. Surf. Sci. 215, 232–236 (2003)
120. W.B. Choi, J.J. Cuomo, Field emission from silicon and molybdenum tips coated with
diamond powder by dielectrophoresis. Appl. Phys. Lett. 68(5) (1996)
121. V.V. Zhirnov et al., Diamond coated Si and Mo field emitters: diamond thickness effect.
Appl. Surf. Sci. 94(95), 123–128 (1996)
122. A.N. Alimova, N.N. Chubin et al., Electrophoresis of nanodiamond powder for cold cathode
fabrication. J. Vac. Sci. Technol. B. 17(2) (1999)
123. A.F. Pal, N.V. Suetin et al., Emission Properties of Electrophoretically Deposited
Nanodiamond Film. IEEE, IVMC 2001 Davis, California USA, 12–16 August 2001
124. Yoshikazu Nakayama, Seiji Akita, Field-emission device with carbon nanotubes for a flat
panel display. Synth. Met. 117, 207–210 (2001)
125. A.N. Obraztsova, A.P. Volkov et al., Single-Wall Carbon Nanotube Electron Emitters
Formed by Liquid Electron Emitters Formed by Liquid Electrophoresis. ITM-FEECM 2001,
Moscow, Russia, 2–4 July 2001
126. W.B. Choi, Y.W. Jin et al., Electrophoresis deposition of carbon nanotubes for triode-type
field, emission display. Appl. Phys. Lett. 78(11), 12 (2001)
127. H. Zhao, H. Song, Z. Li, G. Yuan, Y. Jin, Electrophoretic deposition and field emission
properties of patterned carbon nanotubes. Appl. Surf. 251, 242–244 (2005)
128. S.J. Oh, J. Zhang, Y. Cheng, H. Shimoda, O. Zhou, Liqid-phase fabrication of patterned
carbon nanotube field emission cathode. Appl. Phys. Lett. 84(19), 3738–3740 (2004)
129. E.P. Sheshin, A.S. Baturin, K.N. Nikolskiy, R.G. Tchosov, V.B. Sharov, Field emission
cathodes based on milled carbon fibers. Appl. Surf. Sci. 251, 196–200 (2005)
130. A.N. Alimova, N.N. Chubin, P.J. Belobrov, V.V. Zhirnov, Electrophoreses of nanodiamond
powder for cold cathode fabrication. Proc. IVMC 222–223 (1998)
366 6 Carbon-Based Field-Emission Cathodes
131. A.S. Batyrin, D.A. Kyrnosov, K.N. Nikolskiy, E.P. Sheshin, R.G. Chesov, Way of
production of the autoissue cathode. Stalemate. The Russian Federation No. 2225052 from
2002. Cl. H01j1/14
132. E.P. Sheshin, A.S. Baturin, K.N. Nikolskiy, R.G. Tchosov, in Field Emission Cathodes
Based on Milled Carbon Fibers. Proceedings of IVESC, Beijing, China, pp. 114–116 (2004)
133. J.-W. Nam, J.H. Choi, J.-H. Han, J.-B. Yoo, C.-Y. Park, Enhanced Emission from Carbon
Nanotube Paste After Firing (Techn Digest IVNC, Osaka, Japan, 2003), pp. 225–226
134. H. Ren, S.Z. Deng, J. Chen, J.C. She, N.S. Xu, Plasma Etching Treatment for Improving the
Field Emission Properties of Carbon Nanotubes Composite Emitters (Techn Digest IVNC,
Osaka, Japan, 2003), pp. 261–262
135. J.W. Nam, S.H. Cho, S.H. Choi, Y.C. Choi, J.S. Ha, J.H. Park, D.H. Choe, J.B. Yoo, Long
Lifetime of Field Emitters Fabricated Using Carbon Nanotube Paste (Techn Digest of
IVNC, Oxford, UK, 2005), pp. 292–293
136. L.-T.A. Cheng, D.H. Roach, Electron field emitter and composition related thereto.
Stalemate. USA 2004/0017141. 2004 Yu CL. H01j19/06
137. A. Boda, Y. Jshida, K. Hakiai, T. Asano, Fabrication of Gated Carbon Black Field Electron
Emitter Using Inkjet (Techn Digest of IVNC, Oxford, UK, 2005), pp. 42–43
138. B.V. Bondarenko, V.I. Makyha, E.P. Sheshin, Autoelectronic emitters with the developed
working surface. Electron. Equip. (It is gray. 1: Elektronika Microwave Oven) 10, 44–47
(1984)
139. B.V. Bondarenko, V.I. Makyha, L. Rybakov Yu, V.B. Sharov, E.P. Sheshin, Vlyany
roughnesses of a surface of autocathodes on their issue characteristics. Radio Eng.
Elektronika 32(12), 2606–2610 (1987)
140. B.V. Bondarenko, V.I. Makyha, A. Rybakov Yu, E.P. Sheshin, Field Electronic Emission of
Rod Graphite Cathodes. Physical Phenomena in Devices of Electronic and Laser Equipment
(M.: MFTI, 1981), pp. 11–15
141. A.L. Syvorov, E.P. Sheshin, V.V. Protasenko and other, The microrough flat autoissue
cathodes from graphite received in the radiftsionny way. ZhTF 66(7), 156–160 (1996)
142. A.G. Chakhovskoi, C.B. Hunt, Improved Image Uniformity in Light Sources with Carbon
Field Emitters (Techn Digest Lith JVMC, NC, USA, 1998), pp. 190–191
143. A.S. Baturin, E.P. Sheshin, Field Emission Characteristics of Reticulated Vitreous Carbon
(Tecchn Digest of IVNC, Guangzhou, China, 2000), pp. 42–43
144. C.E. Hunt, A.G. Chakhovskoi, Field emission cathode fabricated from porous carbon foam
material. Patent WO 99/43870 of 2.09.99. Cl. C25B3/00. H01j1/30
145. P.G. Gabdyllin, S.N. Davydov, V.V. Korablev, A.E. Kravchik, A. Kykyshkina Yu, V.V.
Sokolov, V.S. Travnikov, Cold emitter of electrons. Stalemate. Russian Federation
No. 2249876 of 6.11.2003. Cl. H01j1/30
146. M. Leshukov Yu, E.P. Sheshin, Plasmachemical Processing of Field Emission Cathodes
Made of Carbon Fibers Bundles (Techn Digest of IVNC, Guilin, China, 2006), pp. 333–334
147. A.S. Leychenko, M. Leshykov Yu, N.V. Lyparev, E.P. Sheshin, in Formation of the
Emitting Surface of Autocathodes from Bunches Carbon вoлoкн the Crown Category. Sb.
Tezisov of dokl. mezhdunarodny Conference “Carbon: Fundamental Problems of Science,
Materials Science, Ekhnologiya”, Moscow, p. 115 (2006)
148. L.B. Loeb, Electrical Coronas (University of California Press, 1965), p. 760
149. R. Morrow, Theory of negative corona in oxygen. Phys. Rev. A. 32, 1799–1809 (1985)
150. A. Cao, X. Zhang, C. Xu, J. Liang, D. Wu, B. Wei, Thinning and diluting aligned carbon
nanotube film for uniform field emission. Appl. Phys. A. 74, pp. 415–418 (2002)
151. V.V. Zhirnov, O.M. Kuttel, O. Gruning, A.N. Alimova, P.Y. Detkov, P.J. Belobrov, E.
Mailadr-Schaller, L. Schlapbach, Characterization of field emission cathodes with different
forms of diamond coatings. J. Vac. Sci. Thechnol. B. 17(2), 666–669 (1999)
152. V. Gilyaev Yu, N.J. Sinitsyn, G.V. Torgashov, S.G. Saveliev, Fundamental and applied
directions of field emission electronics using nanicluster carbon materials. Appl. Surf. Sci.
215, 141–148 (2003)
References 367
153. Y. Choi, W.W. Lee, B.-H. Ryu, Application of Carbon Nanofiber Coated with Pt
Nanoparticles to FED (Techn Digest of Euro Display, Moscow, 2007), pp. 318–320
154. R.B. Sharma, D.J. Late, D.S. Joag, A. Govindaraj Rao, CNR Field Emission Studies of
Boron and Nitrogen Doped Carbon on Pointed and Flat Substrates (Techn Digest of IVNC,
Oxford, UK, 2005), pp. 310–311
155. V.D. Yu, J.H. Zhang, X. Wang, X.M. Li, X.D. Gao, Characterization and field emission
property of aligned porous carbon nanotube film by hydrogen-ion implantation. Appl. Phys.
A. 81, 169–172 (2005)
156. A.C. Batyrin, K.N. Nikolskiy, A.I. Knazev, R.G. Chesov, E.P. Sheshin, Introduction of
shchelochnozemelny metal in structure of graphite for the purpose of decrease in work of an
exit. Mag. Tech. Fiziki 74(3), 62–64 (2004)
157. A.C. Batyrin, A.G. Borisov, V.S. Korsakov, L.A. Plavich, I. Tishin Yu, E.P. Sheshin, Way
of drawing an issue covering on the cathode. The Stalemate of the Russian Federation
No. 2177657. Cl. H01j1/14 of 22.01.2001
158. K.N. Nikolsky, A.S. Batyrin, V.S. Bormashov, E.P. Sheshin, D.V. Shyr, S. Zaginaychenko
Yu, V.E. Myradan and other, Field Emission Researches Dopirovannykh of Carbon
Nanotubes (Techn Didgest of Cont «Hydrogen Materials Science and Chemistry of Carbon
Nanomaterials», Sudak-Crimea, Ukraine, 2003), pp. 824–825
159. P.V. Sherstnev, A.S. Baturin, V.S. Bormashov, K.N. Nikolskiy, E.P. Sheshin, Investigation
of local work function of field emission of field emission cathode from carbon materials
doped by Ba with the help of AFM/SEM/X-ray. Phys. Chem. Solids 62, 159–163 (2004)
160. A. Bobkov, E. Davidov, S. Zaitsev, S. Karpov, M.A. Kozodaev, I.N. Nikolaeva, M.O.
Popov, E.N. Skorokhodov, A.L. Suvorov, N. Cheblukov Yu, J. Vac. Sci. Technol. B 19, 32
(2000)
161. V.S. Bormashov, A.S. Baturin, K.N. Nikolskiy, E.P. Sheshin, Tech Digest of 15th IVMC
and 48th IFES, vol. 2, no. 2, 64 (2002)
162. R.G. Forbes, J. Vac. Sci. Technol. B 17, 534 (1999)
163. R.G. Forbes, J. Vac. Sci. Technol. B 17, 526 (1999)
164. E.L. Murphy, R.H. Good, Phys. Rev. 102, 1464 (1956)
165. J.-M. Bonard, J.-P. Salvetat, T. Stockli, L. Forro, A. Chatelain, Appl. Phys. Lett. 69, 245
(1999)
166. W. Zhu, G. Koshanski, J. Sungho, C. Bower, O. Zhou, Appl. Phys. Lett. 75, 873 (1999)
167. V.S. Bormashov, A.S. Baturin, K.N. Nikolsky, E.P. Sheshin, Appl. Surf. Sci. 215, 178–184
(2003)
168. V.S. Bormashov, A.S. Batyrin, Materials of the international conference «Educational,
scientific and engineering applications in the environment LabVIEW and technologies
National Instruments» Moscow, p. 130 (2003)
Chapter 7
Computation of Field-Emission
Cathode-Based Electron Guns
7.1 Introduction
boundary point system. The first stage of the method is to find the charge distri-
bution over the electrodes, the second being determination of the potential in the
entire space. The basic difficulty of this method is caused by a necessity to solve the
integrated equation linking the boundary potentials with the surface charge-density
distribution. The analytical solution of such integrated equations being impossible
for most lenses, they are replaced with linear algebraic equation systems, which
makes it possible to get an approximate solution. There are also some other surface
charge-density approximation methods. Recently, wide application of the
integral-equation method is caused by a possibility to use high-speed high-volume
computers providing adequate accuracy within quite a short period of time.
The variable-separation method is among the general methods to solve equations
in partial differential coefficients. This method can only find potential distribution
for quite simple boundary conditions when the electrode surfaces belong to coor-
dinate planes, e.g., cylinders or diaphragms, which is an important constraint. If an
adequate coordinate system happens to be selected, the Laplace equation is rela-
tively easy to solve. In this case, it functions as a harmonic-polynomial potential
approximation method.
Numerous electron- and ion-optical devices employ the focusing lens system not
to form an electron-optical image, but make the greatest possible number of par-
ticles escaping from the source at different angles concentrate on the detector. In
particular, examination of a substance surface involves collection of the charged
particles escaping from the surface at as wide an angle as possible, their conduction
along the path including a mass spectrometer and registration with an information
receiver/detector, thus bringing up an issue of functioning of an electron-optical
system with essentially off-axis beams.
In electron-beam devices, electron flows are formed by coordinate-dependent
and time-independent, i.e. static, fields. The position of formed electron beams is
controlled by the fields varying in time. However, the variations within the con-
trolled beam electron transit time are not large, and these fields can be considered as
quasistatic. In paraxial optics, when the particle pathlength from the system axis can
be taken as a small parameter, electron-optical computations usually operate
axisymmetric electron-optical system fields set in a tabular way in a form of axial
distribution of the electric potential calculated by this or that method. At interme-
diate points, the values are usually calculated by the nearest neighbor interpolation
method, which can result in low accuracy of derivatisation of the second-and
higher-order potential by interpolation-polynomial differentiation.
Above-mentioned general methods to compute electron-optical systems are
characteristic for any system. However, computation of such systems encounters
specific difficulties as far as cathode lenses are concerned. It is inherent in cathode
lenses that the field starts forming the electronic image immediately at the cathode
surface. That is why solution of general electron-optical problems through com-
putation of cathodic systems needs special methods. First of all, they are
electrostatic-field computation methods. Each specific cathodic electron-optical
system is usually simplified to a model able to provide an approximate solution of
the problem. When computing the characteristics of pointed systems represented by
7.1 Introduction 373
cathode lenses, e.g., field-emitter electron guns, the main difficulty encountered is
that the sizes of the emitter and the focusing electrodes of the system differ by a few
orders of magnitude.
As a rule, specific consideration is given to both the cathode area, the approx-
imating functions necessary to determine the field in the apex neighbourhood being
formed, and the focusing system, followed by combination of the solutions arrived
at. Computation of the field near the pointed cathodes involves approximate
replacement of their surfaces with paraboloids or hyperboloids of revolution,
extremely oblong ellipsoids, as well as equipotential surfaces created by a charged
conductive cone with a conductive sphere on its apex, i.e. sphere-on-cone.
Well-developed general methods of theoretical examination of electron-optical
systems have supposedly lost their generality because of a qualitative change of
up-to-date electron-optical problems. The change is accounted for by increase in the
role of the elements whose electron-optical theory cannot rely on the classical
paraxial approximation, e.g., an unconstrained increase in the charged-particle path
slope to the main optical axis of these systems in the cathode neighborhood area.
That is why complication and widening of a circle of electron-optical problems
make it necessary to develop new methods of accurate computation of the field in
electron-optical systems.
This chapter is dedicated to the systems of forming and controlling of electron
guns based on the electron beams generated by pointed field-emission cathodes.
The focus of attention is on the forming and controlling systems based on
immersion electrostatic lenses. The primary reason thereof is that the operational
experience in beam formation and transportation systems shows sufficient effec-
tiveness of these lenses and their superiority over magnetic lenses for the
beam-intensity ranges considered. Moreover, practical implementation of electro-
static lenses is much cheaper than that of magnetic ones.
At the same time, their computation, simulation and numerical experiment for
the systems of forming and controlling electron sources and guns based on
field-emission cathodes is much more difficult. The primary reason thereof is that
specific characteristic of field emission that emission is excited by an electric field,
and current density considerably depends on its intensity directly at the surface of
such cathode. The field can be generated if at least one more electrode in addition to
the cathode is available. Frequently, a charged-particle beam cannot be conditioned
with only one electrode, and the forming and controlling systems have to be sup-
plemented with some other electrodes that are basically able to focus and transport
the beam, as well as control it in an efficient way. Unlike hot cathode-based forming
and controlling systems and guns, even the slightest change of potentials at the
system electrodes based on field emission cathodes can cause drastic changes in the
emissive capacity of the latter. And the change in the emissive capacity of the
cathode causes reconditioning of the emitted-electron beam, which entails a change
in its focusing and transporting conditions. In other words, the problem of com-
putation of field-emission cathode-based guns has to be brought up with account of
the cross-impact of these cathodes and the additional electrodes both on the cathode
emissivity and the characteristics of the forming and controlling systems, which
374 7 Computation of Field-Emission Cathode-Based Electron Guns
presents severe difficulties. Apparently, it is these difficulties that account for lack
of essential successes in computation of guns and forming and controlling systems
based on field-emission cathodes.
DU ¼ 0
UjX ¼ w0 ; ~
UjX~ ¼ w; ð7:1Þ
where Ω is the cathode-apex surface; w0 is the potential at the apex; UjX~ is the
anode surface; w ~ is the potential at the anode. The most important characteristic of
such systems is intensity of field E = –grad U at the apex peak. This value defies
detailed analysis because it depends on the geometrical dimensions of the apex. In
practice, it is thin apices that are most interesting. As a result, there is a small
parameter arising in the problem considered, viz. a ratio of the apex thickness to its
typical longitudinal dimension. To study such apices, it is enough to expand the
field intensity at the apex peak (E0) in an asymptotic series in the vicinity of the
ideally thin apex. A few first terms of this series would enable calculation of the
0 z
-L -δ
7.2 General Problem of Computation of the Optimum Structure of Electron Guns … 375
field intensity in a quite precise way. To construct such asymptotic expansion, let us
apply the following approach. Let us represent the sought-for solution as
ZL
qðz0 Þ
U0 ðr; zÞ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dz0 ; ð7:3Þ
d r 2 þ ðz z0 Þ2
DU1 ¼ 0; ~
U1 jX~ ¼ w; ð7:4aÞ
ZL
U2 ðr; zÞ ¼ u2 ðr; z; z0 Þqðz0 Þdz0 ; ð7:5Þ
d
where function u2(r, z; z′) is defined through the Green function of problem (7.3),
viz.
h i1=2
u2 ðr; z; z0 Þ ¼ 4pGðr; z; 0; z0 Þ r 2 + ðz z0 Þ2 ; ð7:6Þ
ZL
2ðL zÞ qðnÞ qðzÞ
Uj ¼ qðzÞ ln qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ dn
r!0 j n zj
d z þ r 2 þ ðd zÞ2 d ð7:7Þ
d\z\L
þ U1 ð0; zÞ þ U2 ð0; zÞ:
376 7 Computation of Field-Emission Cathode-Based Electron Guns
Taking into account (7.7), the equipotential surface close to the apex axis fol-
lows the ratio of
ZL
2ðL zÞ qðnÞ qðzÞ
qðzÞ ln qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ dn
j n zj ð7:8Þ
d z þ r 2 þ ðd zÞ2 d
+ U1 ð0; zÞ þ U2 ð0; zÞ ¼ w0 :
where r0 is the equipotential surface section radius taken as section height function
z. As surface shape r0(z) is known, ratio (7.8) is a charge-density determination
equation. Equation (7.8) includes value d, which is still unknown. Let us define it
from the condition that U(0, 0) = w0, i.e. we believe that apex peak is at the origin
of coordinates. Let us underscore that value d for a thin apex should be small. In the
latter case, condition U(0, 0) = w0 can be represented as
ZL
L qðnÞ qðdÞ
w0 ¼ qðdÞ ln þ dn þ U1 ð0; 0Þ þ U2 ð0; 0Þ: ð7:9Þ
d n
d
ZL
2ðL zÞ qðnÞ qðdÞ
w0 ¼ qðdÞ ln þ dn þ U1 ð0; dÞ þ U2 ð0; dÞ: ð7:10Þ
r0 ðdÞ nd
d
2d ¼ r0 ðdÞ: ð7:11Þ
Let us pay attention to the fact that function r20 for any apex shape can be
represented as
r02 ðzÞ ¼ 2R0 z + 0 r 2 ;
where R0 is the apex-peak curvature radius. Let us assume that the apex peak is
smooth enough. Moreover, we will consider the condition of
as met. With its account, (7.11) can give us an expression to calculate value d, viz.
d ¼ R0 =2: ð7:13Þ
This ratio can be also obtained in a different way, viz. it is necessary to calculate
the equipotential surface curvature radius going through the origin of coordinates
within the range of d ! 0 with involvement of (7.2), (7.3). Using the formula
1
R0 ¼ Uz0 ð0; 0Þ Urr00 ð0; 0Þ
for the equipotential surface curvature radius, let us transform the (7.13) as 2d = R0.
If condition (7.12) is met, (7.8) can be somehow reduced. Let us underscore that
if z d.
2
2ðL zÞ 4zðL zÞ r0 ðzÞ d
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 1 þ 0 þ 0 ;
d z þ r02 ðzÞ þ ðd zÞ2 r0 ðzÞ
2 z 2 z
ZL ZL
4zðL zÞ qðnÞ qðzÞ
qðzÞ ln þ dn þ u2 ð0; z; nÞqðnÞdn ¼ w0 U1 ð0; zÞ:
r02 ðzÞ j n zj
d d
ð7:14Þ
Using (7.14), let us construct the desired expansion for value q(z). To this effect,
it is enough to seek solution of equation (7.14) as an iterative series. Iterating (7.14),
we can obtain the solution in the form of
X
1
qðzÞ ¼ qk ðzÞ; ð7:15Þ
k¼0
w0 U1 ð0; zÞ
q0 ðzÞ ¼ ; ð7:16Þ
ln½4zðL zÞr02 ðzÞ
378 7 Computation of Field-Emission Cathode-Based Electron Guns
2 3
1 ZL ZL
4zðL zÞ 4 qk ðnÞ qk ðzÞ dn þ
qk þ 1 ðzÞ ¼ ln u2 ð0; z; nÞqk ðnÞdn5:
r02 ðzÞ jn zj
d d
ð7:17Þ
ZL
@U qðnÞ @
E0 ¼ ¼ dn þ ðU 1 þ U Þ
2 : ð7:18Þ
@z z¼0 n 2 @z z¼0
r¼0 d r¼0
Obviously, the main contribution into expression (7.18) for thin apices is given
by the first term. Calculating the integral asymptotically at d ! 0, we find the field
intensity as
In (7.20), the integral is calculated over the apex surface, viz. E(x) is the field
intensity at the apex surface at point x. At B E0 integral (7.20) can be calculated
asymptotically by the Laplace method, viz.
2pAE04 B
I¼ 00 ð0Þ
exp :
BErr E0
2pAE02
Sem ¼ IJ01 ¼ 00 ð0Þ
:
BErr
7.2 General Problem of Computation of the Optimum Structure of Electron Guns … 379
00
Using model potential (7.2), it is not difficult to calculate the value of Err ð0Þ:
However, the respective expression looks cumbersome. It is essentially reduced at
d ! 0, i.e. for a thin apex, and taking into account (7.13) and (7.19)
Let us note that the linear dependence of the emission area on the field intensity
was stated in a number of papers, e.g. [42, 43]. However, they did not determine the
coefficient of proportionality for that dependence. Processing the results of the
experiments aimed at determination of field intensity at the apex peak often
involves the Homer formula, viz. E0 = U0/(kR0), where U0 is the potential differ-
ence in the anode–cathode space; R0 is the apex-peak curvature radius, and k is the
geometrical factor amounting to about 5. Application of formulas (7.19) and (7.15)–
(7.17) can make a more accurate calculation of this form factor. For a thin-enough
apex, let us use the first expansion term (7.15). Having inserted (7.16) into (7.19),
we obtain
w0 U1 ð0; 0Þ
E0 ¼ ; ð7:21Þ
kR0
where
1 2L
k¼ ln : ð7:22Þ
2 R0
For a thin apex, the integral in square brackets is small because the q0(n) value is
small. Therefore, (7.23) can be represented as
2 3
ZL
1 2L 4 1 q0 ðnÞ q0 ðdÞ
k ¼ ln 1þ dn þ u2 ð0; 0; nÞq0 ðnÞ 5 þ :
2 R0 w0 U1 ð0; 0Þ n
d
ð7:24Þ
380 7 Computation of Field-Emission Cathode-Based Electron Guns
ZL
1 2L 4nðL nÞ 1 w0 U1 ð0; nÞ 1
k ¼ ln dn ln þ u 2 ð0; 0; nÞ : ð7:25Þ
2 R0 r02 ðnÞ w0 U1 ð0; 0Þ n
d
4nðL nÞ 2L
ln ¼ ln ; U1 ð0; nÞ ¼ 0; u2 ð0; 0; nÞ ¼ ð2D þ nÞ1 :
r02 ðnÞ R0
The integral of (7.25) can be calculated accurately. For the form factor, we
obtain
1 4LD
k ¼ ln : ð7:26Þ
2 R0 ð2D þ LÞ
Zhd X 1
U ¼ E1 z dnqðnÞ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi;
h þ d
k jr ak j þ ðz nÞ2
2
qðzÞ ¼ qðzÞ;
where E∞ is the field intensity far from the cathode; h is the apex height; ak is the
vector connecting the apex peak at the origin of coordinates with k-apex peak.
Having performed the computation above, one cannot but be sure that the apex-peak
field intensity is determined by the ratio similar to (7.19) in this case as well, i.e.
2qðh R0 =2Þ
E0 ¼ bE1 ; b¼ ;
R0 E1
where the sum applies to all apices except for the one located in the origin of
coordinates. Integration of (7.27) can help to make up a series similar to series (7.14).
It is not difficult to verify that the individual-apex emission area is given with
expression (7.20a) in both a multi-apex system and a one-apex case. However, it
should be noted that research of the respective asymptotic expansion for (7.27)
when studying multi-apex systems is difficult because, on the one hand, the
asymptotic expansion parameter of (ln4h/R0)−1 can be insufficiently small for such
real systems; on the other hand, even the second term of the series considered is
rather cumbersome, which prevents from efficient research. Therefore, the
multi-apex systems can be conveniently studied directly with (7.27). The approx-
imate solution of this equation seems to be equivalent to the approximate sum-
mation of the asymptotic series generated by this equation.
Let us note that numerical solution of (7.27) needs some caution because its
second term is a singular-nucleus integral, which may cause the computing circuit
to lose its stability.
For numerical computation, we selected a system consisting of semi-ellipsoidal
apices located in the nodes of a flat square grid. The equation was solved with the
Ritz method, the solution having been sought-for as odd polynomials to the 2N − 1
degree. The focus of research was on dependence of reduced field enhancement
factor b/b0, b0 being the intensity enhancement factor of an individual apex, on the
packaging parameter of ah−1, a being the distance between the two next apices, for
the apices with different peak-curvature radii.
The numerical computation showed that the solution stability was lost even for
N = 4 at ah−1 > 0.5. For N > 4, the stability was lost at even less close-packed
arrangement of the apices. But for N = 2 and N = 3 the numerical method worked
well even at ah−1 ≅ 1 and produced the results differing by about 5% at the worst.
The computation results are shown in Fig. 7.2. One can see that the apices begin
affecting each other at ah−1 ≅ 0.3. If the apices are even more close-packed, the
field-enhancement factor plummets down. The computation results suggest that the
apex packaging of more than ah−1 ≅ 3 4 is not effective.
0.6
0.4
0 4 6 ah-1
382 7 Computation of Field-Emission Cathode-Based Electron Guns
To determine field intensity at the apex peaks of the field-emission cathode, the
field potential of the system is represented as
X Z
hd
qðz0 Þdz0
Uðr; zÞ ¼ E1 z qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi;
k jr ak j2 þ ðz nÞ2 ð7:28Þ
h þ d
qðzÞ ¼ qðzÞ;
where E∞ is the field intensity far from the apices, h is the apex height, ak is the
vector connecting the k-apex foundation with the origin of coordinates.
Expression (7.28) meets the Laplace equation everywhere except for the points
lying on the apex axis and the boundary condition of u(r, 0) = 0. If values q(z) and
d meet the following conditions, viz.
d ¼ R0 =2; ð7:29Þ
Zh 0 Z
þh
4ðh2 z2 Þ qðz0 Þ qðzÞ 0 X qðz0 Þdz0 z
qðzÞ ln þ dz þ q ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ ; ð7:30Þ
r02 ðzÞ jz zj
0 h
h
k
h a2k þ ðz z0 Þ2
1. R0 H 1 ¼ 0:01; b0 ¼ 50:1;
2. R0 H 1 ¼ 0:02; b0 ¼ 178:5:
2qðh R0 =2Þ
E0 ¼ bE1 ; b¼
R0 E1
The total current from the apex peak in the field electronic cathode follows the
expression of
I ¼ Sem J0 ;
7.2 General Problem of Computation of the Optimum Structure of Electron Guns … 383
E0
Sem ¼ 2p R20 :
B
4ðh2 z2 Þ 4h
ln ¼ ln ;
r02 ðzÞ R0
x
h
y
384 7 Computation of Field-Emission Cathode-Based Electron Guns
(7.27) gives
4h
~ðnÞ ln
q þ K1 ð~
qðnÞÞ þ K2 ð~
qðnÞÞ ¼ n;
R0
0 Z
þ1
X q~ðn0 Þdn0
K1 ð~
qðnÞÞ ¼ l qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi;
k
1 S2k þ l2 ðn n0 Þ2
Z1
~ðn0 Þ q
q ~ðnÞ 0
K2 ð~
qðnÞÞ ¼ 0 dn :
jn nj
1
1 1 1 ðn n0 Þ2 3 4 ðn n0 Þ4
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ l2 þ l þ Oðl6 Þ;
0 2 jSk j 2 jS j3 8 jS j 5
Sk þ l ðn n Þ
2 2 k k
Z1 Z1
0 0 3 0 0
K1 ¼ C3 l 3
~ðn Þdn C5 l5
nn q ðn3 n0 þ nn 3 Þ~
qðn0 Þdn0 þ Oðl7 Þ;
2
1 1
where C3 and C5 are the coefficients stipulated by the apex-grid configuration only,
viz.
X
0
1 X
0
1
C3 ¼ ; C5 ¼ :
k jSk j3 k jSk j5
K1 ¼ C3 l3 P1 ðnÞ ~
hP1 ; qi
3 6 2
C5 l 5
~i þ ðP3 ðnÞhP1 ; q
P1 ðnÞhP1 ; q ~i þ P1 ðnÞhP3 ; q
~iÞ þ Oðl7 Þ;
2 5 5
here, h; i is the scalar product; P1(n) = n, P3(n) = 5n3/2 − 3n/2 are the Legendre
polynomials. Here let us focus on the relation of orthogonality, viz.
7.2 General Problem of Computation of the Optimum Structure of Electron Guns … 385
2 2
hP1 ; P1 i ¼ ; hP3 ; P3 i ¼ ; hP1 ; P3 i ¼ 0:
3 7
77
~i
K2 ¼ 3P1 ðnÞhP1 ; q ~i:
P3 ðnÞhP3 ; q
6
4h 77
ln q ~i P3 ðnÞhP3 ; q
~ðnÞ 3P1 ðnÞhP1 ; q ~i þ C3 l3 P1 ðnÞhP1 ; q
~i
R0 6
3 6 2
C5 l5 P1 ðnÞhP1 ; q ~i þ ðP3 ðnÞhP1 ; q
~i þ P1 ðnÞhP3 ; q
~iÞ þ Oðl7 Þ ¼ P1 ðnÞ:
2 5 5
Consequently,
1
4h
~ð1Þ ¼ a þ c ¼ m~
q q0 ð1Þ; ~0 ð1Þ ¼
q ln 2 ; ð7:32Þ
R0
386 7 Computation of Field-Emission Cathode-Based Electron Guns
1 1 !
2C3 l3 6C5 l5 4h 4h
m¼1 þ ln 2 þ 3 ln 11 þ Oðl6 Þ:
3 lnð4h=R0 Þ 6 5 R0 R0
m ¼ 1 d3 l3 þ d5 l5 þ Oðl6 Þ ð7:33Þ
d3 and d5 being obvious from comparison between (7.32) and (7.33), it is not
difficult to obtain
J
l2 ð1 Dðd3 l3 d5 l5 ÞÞ þ Oðl8 Þ; ð7:34Þ
where
B
D¼ þ 3:
b0 E1
Having differentiated (7.34), equated the expression obtained to zero and taken
away the trivial root of µ = 0, we obtain the equation stipulating the optimum
parameter of µ:
2
5d3 l3 7d5 l5 þ Oðl6 Þ ¼ : ð7:35Þ
D
7.2 General Problem of Computation of the Optimum Structure of Electron Guns … 387
Studies of (7.35) show that it has a single root arbitrary close to zero at D ! ∞.
Other roots are separated from zero at any value of D. Definitely, it is this solution
of the problem that is physically sensible, and other roots appeared due to the
approximate nature of (7.35).
In the higher order, (7.35) produces
1=3
2
l¼ ;
5d3 D
2
5d3 l4 7d5 l5 ¼ þ OðD2 Þ: ð7:36Þ
D
2C3 l3
d3 ¼ ; ð7:38Þ
3 lnð4h=R0 Þ 6
" 1 1 #
6C5 l5 4h 4h
d5 ¼ ln 2 þ 3 ln 11 : ð7:39Þ
5 R0 R0
B
D¼ þ 3 1: ð7:41Þ
b0 E1
Expressions (7.40) and (7.41) show that the optimum cathode-apex packaging
depends on the operating point, i.e. on E∞ of the emission diode, and we can see
that in high operating voltages in the anode-cathode gap more close-packed
arrangement of apices becomes optimal. The reason is that the curvature of the
volt-ampere emission-diode dependence decreases with field-intensity growth at the
cathode surface, which means that the emission-current density becomes less
sensitive to drop in the field intensity at the apex peak caused by tightening of the
388 7 Computation of Field-Emission Cathode-Based Electron Guns
apex structure. Expressions (7.37)–(7.39) give the first two terms of the µ value
expansion (which is equivalent to aopt) in parameter D.
a þ ib
z þ ir ¼ icctg ;
2
0 a p; 1\b\1:
β1 α2
V1 0
α1
α0
0 a a0 ; b ¼ b0 ;
a ¼ a0 ; b2 \b\b0 ;
• is the modulator
a1 a a0 ; b ¼ b1 ;
390 7 Computation of Field-Emission Cathode-Based Electron Guns
0 a a0 ; b ¼ b2 ;
The geometrical sizes of the simulated system electrodes are determined by the
following formulas, viz.
• is the radius of the sphere on the apex peak
1 þ chb0
R0 ¼ c ;
shb0
1 þ chb0
L¼c ;
shb0
The potential at the cathode is equal to 0, the modulator potential being equal to
V1, the anode potential being set by function f(a), viz.
(
V 2 ; 0 a a2 ; b ¼ b2 ;
f ðaÞ ¼ ð7:43Þ
0; a2 a a1 ; b ¼ b2 :
7.3 Mathematical Simulation of Model Triode Electron-Optical Systems 391
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb0 bÞ
U1 ða; bÞ ¼ chb cos a An Pmn ðcosðp aÞÞ; ð7:44Þ
n¼1
shðm n þ 1=2Þðb0 b1 Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb b2 Þ
U2 ða; bÞ ¼ chb cos a An
shðm n þ 1=2Þðb1 b2 Þ
n¼1
ð7:45Þ
shðmn þ 1=2Þðb1 bÞ
þ Bn Pm ðcosðp aÞÞ;
shðmn þ 1=2Þðb1 b2 Þ n
"
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X 1
P1=2 þ isn ðcos aÞ pnðb0 bÞ
U3 ða; bÞ ¼ chb cos a Cn sin
n¼1
P 1=2 þ isn ðcos a 1 Þ b0 b 2
# ð7:46Þ
X1
shðlm þ 1=2Þðb0 bÞ
þ Dm Pl ðcos aÞ ;
m¼1
shðlm þ 1=2Þðb0 b2 Þ m
where Pk ðcos aÞ is the Legendre function of the first kind; mn ; lm are the roots of
equations, respectively,
pn
Pkn ðcos a0 Þ ¼ 0; Plm ðcos a1 Þ ¼ 0; sn ¼ :
b0 b2
Let us insert the respective value at the boundary into the potential distribution in
the third area b = b2:
"
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X 1
P1=2 þ isn ðcos aÞ pn ðb0 b2 Þ
U3 ða; b2 Þ ¼ ch b2 cos a Cn sin ;
n¼1
P1=2 þ is n
ðcos a 1 Þ b0 b 2
#
X1
shðlm þ 1=2Þðb0 b2 Þ
þ Dm Pl ðcos aÞ
m¼1
shðlm þ 1=2Þðb0 b2 Þ m
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X
1
V2 ¼ ch b2 cos a Dm Plm ðcos aÞ; 0 a a2 ;
m¼1
V2 X
1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ Dm Plm ðcos aÞ
chb2 cos a m¼1
Z1
Pm ðxÞPn ðxÞdx ¼ 1; n ¼ m;
1
we get
Zp pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2 ½sin aPln ðcos aÞ= 2ðch b2 cos aÞda ¼ Dm ;
0
Z1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2 ½Pn ðxÞ= 2ðch b2 xÞdx ¼ Dm ;
1
ð7:47aÞ
Z1
V2 Plm ðxÞ
Dm ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dx:
Mm ch b2 x
cos a2
Z a1
cos
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb b2 Þ shðmn þ 1=2Þðb1 bÞ
U2 ða; bÞ ¼ chb cos a An þ Bn
n¼1
shðmn þ 1=2Þðb1 b2 Þ shðmn þ 1=2Þðb1 b2 Þ
Pmn ðcosðp aÞÞ:
Z1
Pm ðxÞPn ðxÞdx ¼ 1; n ¼ m;
1
get
Zp pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2 ½sin aPmn ðcosðp aÞÞ= 2ðchb2 cos aÞda ¼ Bn ;
0
Z1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2 ½Pn ðxÞ= 2ðchb2 xÞdx ¼ Bm ;
1
ð7:47bÞ
Z1
V2 Pmn ðxÞ
Bn ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dx;
Nn chb2 x
cos a2
394 7 Computation of Field-Emission Cathode-Based Electron Guns
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb0 b1 Þ
U1 ða; b1 Þ ¼ chb1 cos a An Pm ðcosðp aÞÞ;
n¼1
shðmn þ 1=2Þðb0 b1 Þ n
"
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X 1
P1=2 þ isn ðcos aÞ pnðb0 b1 Þ
U3 ða; b1 Þ ¼ chb1 cos a Cn sin
n¼1
P 1=2 þ isn ðcos a 1 Þ b 0 b2
#
X1
shðlm þ 1=2Þðb0 b1 Þ
þ Dm Pl ðcos aÞ ;
m¼1
shðlm þ 1=2Þðb0 b2 Þ m
(
U1 ða1 ; bÞ; b1 b b0 ;
U3 ða1 ; bÞ ¼
U2 ða1 ; bÞ; b2 b b1 :
ð7:49Þ
"
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X 1
P1=2 þ isn ðcos a1 Þ pnðb0 bÞ
U3 ða1 ; bÞ ¼ chb cos a1 Cn sin
n¼1
P 1=2 þ isn ðcos a1 Þ b0 b2
#
X1
shðlm þ 1=2Þðb0 bÞ
þ Dm Pl ðcos a1 Þ ;
m¼1
shðlm þ 1=2Þðb0 b2 Þ m
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb0 bÞ
U1 ða1 ; bÞ ¼ chb cos a1 An Pm ðcosðp a1 ÞÞ;
n¼1
shðmn þ 1=2Þðb0 b1 Þ n
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X1
shðmn þ 1=2Þðb b2 Þ shðmn þ 1=2Þðb1 bÞ
U2 ða1 ; bÞ ¼ chb cos a1 An þ Bn Pmn ðcosðp a1 ÞÞ:
n¼1
shðmn þ 1=2Þðb1 b2 Þ shðmn þ 1=2Þðb1 b2 Þ
1 X 1 h i
An Ck P1=2 þ isk ðcos a1 ÞP1mn ðcos a1 Þ P11=2 þ sk ðcos a1 ÞPvn ðcos a1 Þ
Nn k¼1
0 b1 Þ
sin pkðb
b b sin a1 1
0 21
¼ ðT1;n þ T2;n Þ;
P1=2 þ isk ðcos a1 Þ s2k þ ðmn þ 1=2Þ Nn
ð7:50Þ
2 X 1
mn þ 1=2 shðmn þ 1=2Þðb0 b1 Þ
Cn An
b0 b2 n¼1 s2k þ ðmn þ 1=2Þ2 shðmn þ 1=2Þðb0 b1 Þshðmn þ 1=2Þðb1 b2 Þ
pkðb0 b1 Þ 2
sin Pmn ðcos a1 Þ ¼ T3;k ;
b0 b2 b0 b2
ð7:51Þ
where
Z a0
cos
Pmn ðxÞ
T1;n ¼ V1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dx; ð7:52Þ
chb1 x
cos a1
X
1
shðlm þ 1=2Þðb0 b1 Þ sin a1 Plm ðcos a1 ÞPmn ðcos a1 Þ
1
T2;n ¼ Dm ; ð7:53Þ
m¼1
shðlm þ 1=2Þðb0 b2 Þ ðlm þ mn Þðlm þ mn þ 1Þ
0 b1 Þ
!
X
1
Pmm ðcos a1 Þ kþ1
sin pkðb
b b
T3;n ¼ Bm sk ð1Þ ðmm þ 1=2Þ 0 2
:
m¼1 s2k þ ðmm þ 1=2Þ2 shðmm þ 1=2Þðb1 b2 Þ
ð7:54Þ
V1 1 cos a1 mn
T1;n pffiffiffiffiffiffiffiffiffiffi Pmn 1 ðcos a1 Þ
chb1 mn þ 1 2chb1 ðmn 1Þðmn þ 2Þ
cos a1 sin2 a1 þ mn cos2 a1
þ Pmn ðcos a1 Þ þ ð7:55Þ
mn þ 1 2chb1 ðmn 1Þðmn þ 2Þ
1 cos a0 mn
Pmn 1 ðcos a0 Þ :
mn þ 1 2chb1 ðmn 1Þðmn þ 2Þ
U
1.0
0.8
0.6 «test»
0.996
0.4 0.998
0.96
0.2
0.98
0 0.96
0.94
0.995
0.2
0.15
r 0.1 1.3 1.4
1.1 1.2
0.05 0.9 1.0
0 z
0.8
Fig. 7.5 The potential distribution for an apex with a sphere-on-cone surface
7.4 Calculation of Distribution of the Electric Field in the Forming … 397
The previous paragraph reviewed the mathematical model of an electron gun with a
system of focusing electrodes represented by diaphragms. When the
diaphragm-aperture radii are much smaller than the distances between them, viz.
X kðZi Zi1 Þ
Kii1 ðx; tÞ ¼ 8xt :
k¼1;3;5... ððkðZi Z1 ÞÞ þ ðt xÞ2 Þ ððkðZi Zi1 ÞÞ2 þ ðt þ xÞ2 Þ
2
ð7:57Þ
On the basis of (7.56) and due to the fact that 0 t; x Ri , kernels Ki j ðx; tÞ
(7.57) assume a very simple form, viz.
1
Kii1 ðx; tÞ8C1 xt ;
ðZi Zi1 Þ3
" # ð7:58Þ
1 1
Kii ðx; tÞ ¼ 8C2 xt þ ;
ðZi þ 1 Zi Þ3 ðZi Zi1 Þ3
where
X 1 X 1
C1 ¼ 1:0517998; C2 ¼ 0:1502571:
1;3;5;...
K3 2;4;6...
k3
398 7 Computation of Field-Emission Cathode-Based Electron Guns
ðE K Þ/ðxÞ ¼ UðxÞ; ð7:59Þ
where E is the identity operator; /ðxÞ ¼ ð/1 ðxÞ; . . .; /N ðxÞÞ is the sought-for
function vector; UðxÞ ¼ ðU1 ðxÞ; . . .; UN ðxÞÞ is the free-function vector. K ¼ ðK ~ i;j Þ
is the matrix operator, viz.
8 RRi1
>
>
>
>
1
Ki i1 ðx; tÞdt; j ¼ i 1;
>
>
2p
>
>
0
>
< RRi
~ i;j ¼ 2p
1
Ki i ðx; tÞ dt; j ¼ i;
K ð7:60Þ
>
>
0
>
> RR
iþ1
>
> 1
Ki þ 1 i ðx; tÞdt; j ¼ i þ 1;
>
>
>
:
2p
0
0; j 6¼ i 1; i; i þ 1:
Matrix operator K maps CN in CN. Clearly, if y ¼ K/ðxÞ,
X Z
Rj þ 1
Then
X Z
Rj þ 1
maxjyi ðxÞj max K:
Ki j ðx; tÞ dt max /j ðtÞ /
x x x
j
Rj
In virtue
of assumption (7.56), stationary value K is less than 1 (K < 1), then
: So the K-operator norm is less than 1, and there is an inverse operator
k yk K /
of (E − K)−1 and
7.4 Calculation of Distribution of the Electric Field in the Forming … 399
ðE KÞ1 ¼ E þ K1 þ K2 þ :
¼ ðdi xÞ;
K/
4 n
di ¼ C1 bi1 ðRi1 =ðZi Zi1 ÞÞ3 þ C1 bi þ 1 ðRi þ 1 Zi þ 1 Zi Þ3
3p h io
C2 bi ðRi Zi þ 1 Zi Þ3 þ ðRi Zi Zi1 Þ3 ; i ¼ 1; N:
X
1
/ðxÞ ¼ K n U:
n¼0
/ðxÞ ¼ ðai xÞ; ð7:61Þ
where
( 3 3
1 2 Ri1 Ri þ 1
ai f i þ C1 fi1 þ C 1 fi þ 1
2 3p Zi Zi1 Zi þ 1 Zi
" 3 3 #)
Ri Ri
C2 fi þ ; i ¼ 1; N;
Zi þ 1 Zi Zi Zi1
2 Ui þ 1 Ui Ui Ui1
fi ¼ :
p Zi þ 1 Zi Zi Zi1
X
N
pi j aj ¼ f i ;
j¼1
400 7 Computation of Field-Emission Cathode-Based Electron Guns
8
3
>
> 3p C1 ZiRZ ; j ¼ i 1;
>
>
8 i1
>
>
3
i1
3
>
<
2 þ 3p
8
C2 Zi þR1 iZi þ Zi Z Ri
; j ¼ i;
pi j ¼ i1
>
>
>
>
3
>
> 3p
8
C1 Zi þR1 iZi ; j ¼ i þ 1;
>
:
0; j 6¼ i; i 1; i þ 1;
fi are determined in (7.61). So, the solution of this system is existent and unique.
It can be derived as the approximate analytical solution for a boundary-value
problem of the field-potential distribution in the small-aperture diaphragm system
without account of the apex. The solution is given by
Ui þ 1 Ui
V1 ðr; zÞjZi z Zi þ 1 ¼ Ui þ ðz Zi Þ
Z Z
( i þ 1
iqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X1
t þ iak
þ Im ai r 2 þ ðbk i tÞ2 r 2 þ ðak i tÞ2 i ak arcsin
k¼0
r
Ri
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
t þ ibk
þ ibk arcsin ai þ 1 r 2 þ ðb0 itÞ2 þ r 2 þ ða0 i tÞ2
r k k
Ri þ 1 #)
0
t þ iak t þ ibk
þ ia0k arcsin ib0k arcsin ;
r r 0
where
ak ¼ 2k ðZi þ 1 Zi Þ þ ðz Zi Þ;
bk ¼ ð2k þ 1ÞðZi þ 1 Zi Þ þ ðZi þ 1 zÞ;
a0k ¼ 2k ðZi þ 1 Zi Þ þ ðZi þ 1 zÞ;
b0k ¼ ð2k þ 1ÞðZi þ 1 Zi Þ þ ðz Zi Þ:
Ui þ 1 Ui
V1 ð0; zÞjZi z Zi þ 1 ¼ Ui þ ð z Zi Þ
Zi þ 1 Zi
X1
Ri Ri Ri þ 1 Ri þ 1
þ ai ak arctg bk arctg þ ai þ 1 a0k arctg ; b0k arctg 0 :
k¼0
ak bk a bk
7.5 Calculation of Optimum Characteristics in the Forming … 401
I ¼ Sem j0 ; ð7:62Þ
Sem being the emission area, j0 being the current density at the apex peak. Earlier
we have given the formulas to calculate the emission area and the density of the
current from the apex peak, viz.
E0 ¼ 2R1
0 qð0Þ; ð7:63Þ
where
4ð2mÞ1=2 U3=2
B¼ hðaÞ; ð7:65Þ
3he
80 40
60 20
40 0
0.105
20 0.100
0.095 z, cm
0 0.090
r, cm 0.085
0.010
0.008 0.080
0.006
0.004
0.002
0
(b)
r, cm
0.012
78.1374
43.9523
48.8359
53.7195
58.6031
63.4867
68.3703
73.2538
83.0210
87.9046
0.010
0.008
39.0687
0.006
0.004
29.9015 34.1851
0.002
19.5340 24.4179
9.7818
0
0.080 0.085 0.090 0.095 0.100 0.105 0.110
z, cm
The following parameters were taken as given/ known: l = 0.1 cm is the apex
length; R0 = 10−6 cm is the apex-peak curvature radius; z2 = l + 100R0,
r2 = 100R0 are the anode-diaphragm coordinates, viz. the position and the aperture
radius; U2 = 100 V is the potential at the anode diaphragm.
The apex shape depends on the radius of its foundation (rbase), which is an
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffi
ellipsoid with lR0 rbase \ 2lR0 , a paraboloid with rbase ¼ 2lR0 , and a hy-
pffiffiffiffiffiffiffiffiffi
perboloid of revolution with rbase [ 2lR0 .
An elliptic apex is the thinnest, so the physical considerations prompt that it is
this apex shape that makes the field change more rapidly. Actually comparing
Fig. 7.7a, b, we can see that the field potential increases more rapidly for the elliptic
rather than hyperbolic apex in the neighbourhood of the peak.
Figure 7.8 shows that the elliptic apex is characterized by the value of
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi
rbase \ 0:2 103 cm, while the hyperbolic one by rbase [ 0:2 103 cm. It
also shows that increase in the foundation radius from 0.3 10−3 to 5 10−3 cm
404 7 Computation of Field-Emission Cathode-Based Electron Guns
25
20
15
10 0.100015
0.100010
5 0.100005
0.100000
0 z, cm
12 0.099995
10
8 6 0.099990
r·10-6, cm 4 2
1 0
(b)
r·10-6, cm
10
8 17.4929
10.4957
24.4900
3.4986
4
0
0.099990 0.100000 0.100010 z, cm
causes the current value to decrease three times, viz. from 0.75 10−3 to
0.25 10−3A.
The pattern of the constraining electron paths emitted from the apices of different
shape (Fig. 7.9), shows that the revolution-hyperboloid apex is least preferable.
Let us study the effect of other parameters on the value of the total current from
the apex. Figure 7.10 demonstrates some nomograhic charts resulting from a
numerical experiment, viz. the current logarithm—reverse anode voltage curves.
The slopes and arrangement of these straight lines meet the known experimental
volt-ampere dependences, which validates theoretical computations. Table 7.1 give
the values of field intensity at the apex peak (E0 ), current density at the apex peak
(j0 ), emission area (Sem) and total current (I) as dependent on voltage at the anode
diaphragm (U2 ) for work function U = 4.5 eV.
Let us consider the effect of the anode-diaphragm arrangement on the emission
characteristics. Figure 7.11 shows dependence of the total current on the
7.5 Calculation of Optimum Characteristics in the Forming … 405
(a)
I·10-4, A
7
0 1 2 3 4r ·10-3, cm
base
(b)
I·10-4, A
7.40
7.36
7.35
7.28
7.24
3.0 3.4 3.8 4.2 r ·10-3, cm
base
Fig. 7.8 Dependence of full current from rbase for elliptic a and hyperbolic b apex
r· 10-5, cm
18
14
10
8
6
4
2
Fig. 7.9 The pattern of the constraining electron paths: solid line for electron emitted from the
elliptic apex, dashed line for electron emitted from the hyperbolic apex
ln(I/U22)
-5
-15 Φ=2.0 eV
-25 Φ=4.5 eV
Φ=6.0 eV
-35
-1 -1
0.003 0.004 0.005 0.006 0.007 0.008 0.008 U2 , kV
I·10-4, A
7.430
7.426
7.422
7.418
7.414
2 4 6 8 r ·10-5, cm
2
anode-diaphragm radius. The effect of the radius can be seen as small and
expectedly decreasing with its reduction.
The effect of the anode-diaphragm arrangement (z2) is shown in Table 7.2.
It shows that the anode-diaphragm arrangement has a great impact on the total
current, viz. its values change by one and a half order of magnitude when the
distance between the anode diaphragm and the apex peak changes from r0 102 to
r0 104 i. e. by two orders of magnitude.
The emission characteristics are greatly effected by introduction of a cathode
diaphragm with coordinates (r1 ; z1 ) at 0\z1 \l, l being the apex length, which is
shown in Tables 7.3 and 7.4. The potential of the cathode diaphragm coincides with
the potential of the apex/substrate. Table 7.3 presents r1 values as selected at the
distance r0 0:1 of from the apex surface. As the cathode-diaphragm aperture is
7.5 Calculation of Optimum Characteristics in the Forming … 407
within the tenth part of the apex-peak curvature radius from its surface, this table
practically shows the dependence of the emission characteristics on the apex length
if expressed as l – z1. Table 7.4 shows dependence of total current (I) on the
cathode-diaphragm coordinates. Table 7.4 demonstrates, firstly, an ambiguous
dependence of the total current on the cathode-diaphragm coordinates, and sec-
ondly, that this diaphragm can change the current within ten orders of magnitude.
Figure 7.12 shows the total-current dependence on the arrangement of the third
diaphragm, viz. the one following the anode diaphragm (r3 ; z3 ). The curves show a
weak dependence of the current on the parameters of the third diaphragm.
Figure 7.13 shows independence of the current values from the arrangement and
the potentail of the fourth diaphragm. So if the anode voltage is given, the most
profound impact on the total-current value is made by the arrangement and the
radius of the cathode-diaphragm, viz. r1 ; z1 . Therefore, let us assume the x1 ¼
r1 ; x2 ¼ z1 as optimization parameters.
Selection of the optimization method should be based on the assumption that we
are dealing with a constrained problem because physical considerations prevent us
from arbitrary/unconstrained variation of the parameters. Moreover, it is expedient
to resort to the methods based on the search only by the criterion function and the
constraints, i e. the direct search methods, because the function itself and the
constraints have a rather complicated dependence on the parameters.
I· 10-4, A
U2/U3=1
7.416
7.412
U2/U3=0.1
7.408
7.404
7.400
0.10 0.15 0.20 0.25 0.30 0.35 (r3, z3), cm
Fig. 7.12 The total-current dependence on the arrangement of the anode diaphragm (the third)
I·10-4, A
7.40
7.39
0.15 0.20 0.25 0.30 0.35 0.40 z4, cm
Fig. 7.13 Dependence of the current values from the arrangement of the fourth diaphragm
7.5 Calculation of Optimum Characteristics in the Forming … 409
Let us stop on the method of complexes [44]. It has a quite effective algorithm
making it possible to apply a direct search by simplex to solution of
inequality-constrained problems.
The general problem of conditional optimization has constraints represented by
inequalities or equalities, as well as higher and lower values of variables.
As the cathode-diaphragm aperture radius should exceed r0 ðz1 Þ, i.e. the distance
from the axis to the apex surface, let us write out the constraints on variables x1 ; x2 :
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
u" 2 #
u z þ a l
x1 \br t
1 z
; 0 x2 l;
az
where
R0 l2 R0 l
az ¼ ; br ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi :
2lR0 r 2 2lR0 r 2
Every resulting point is checked for admissibility, and if any constraint is vio-
lated, the point is moved to the center of gravity of already plotted points until the
admissible point results.
The set of points P having been plotted, each of them is calculated for the
criterion function, and the point with the maximum function is rejected. A new
point is obtained by mapping of the excluded point through the center of gravity of
other points. If xR is the excluded point and x is the centre of gravity of other points,
then a new point is determined as
xk ¼ x þ a x xR :
Let us assume that initial strictly admissible point x0, mapping parameter a and
calculation-termination parameters e and d have been determined.
Step 1 Plotting of the initial complex consisting of P admissible points. For
every point p ¼ 1; . . .; P 1, it is necessary to
(a) determine coordinates xpi in a random way;
(b) if xpi is an inadmissible point, find the center of gravity x of already found points
and put xpi ¼ xpi þ 1=2ðx xp Þ; re-apply the procedure until xpi is admissible;
(c) if xpi is an admissible point, re-apply (a) until p = P;
(d) calculate f ðxp Þ for p ¼ 1; . . .; P 1.
Step 2 The complex is mapped as follows.
(a) select point xR meeting the condition of f ðxR Þ ¼ max f ðxp Þ Fmin ;
(b) find center of gravity x and new point xk ¼ x þ aðx xR Þ;
(c) if xk is an inadmissible point and f ðxk Þ
Fmax , halve the distance between xk
and x centre of gravity until f ðxk Þ\Fmax ;
(d) if xk is an admissible point and f ðxk Þ\Fmax , move to Step 4;
(e) if xk is an inadmissible point, move to Step 3.
Step 3 Admissibility is provided by the following adjustment, viz.
(a) if xki \x1i , then put xki ¼ x1i ; if xki [ x2i , put xki ¼ x2i ;
(b) if xk is an inadmissible point, halve the distance to the center of gravity;
re-apply the procedure until xk is admissible.
Step 4.
Check of the calculation-completion conditions
(a) Let
X 1X p
f ¼ 1 f ðxp Þ; x ¼ x ;
P P
P P
(b) if ðf ðxp f Þ2 e and ðxp xÞ2 d; stop the calculations; other-
wise, move to Step 2(a).
Paper [44] recommends selecting a = 1.3 and P 2N, N being the problem
order. If the new point is beyond the boundaries of the variables, the respective
coordinate is considered to be equal to the boundary value. Selection of a > 1
compensates for the complex compression caused by halving the distance to the
center of gravity. The large number of peaks is used to prevent degeneration of the
complex, when the search takes place near the boundary of the admissible area.
Results of optimization of the investigated system at different anode voltages and
positions of the anode diaphragm are given in Table 7.5.
7.6 Computation of Electron Paths in a Field-Emission Cathode-Based System 411
gkk ¼ gll ¼ a20 ðsin2 k þ sh2 lÞ; guu ¼ a20 sin2 ksh2 l:
x
x
t
ns
co
=
0
0 ϕ
=
co
ns
t
a0
0 0
0
z
y
z
Fig. 7.15 To calculation of
cathode geometric parameters
r0
c
=
= a = 2
0
0
x
special coordinates is that two non-zero components of a metric tensor are equal to
each other. Simulation of the emission system leaves such parametres as
cathode-surface coordinate kc and focal distance obscure a0. They can be calculated
if distance from the anode to the apex peak d and apex-peak curvature radius r0
have been set, d and r0 of a real emission system being determinable by experiment
(Fig. 7.15).
To calculate the hyperboloid-surface curvature, let us write out the coefficients of
the first and second basic quadratic forms, viz.
7.6 Computation of Electron Paths in a Field-Emission Cathode-Based System 413
1 2 sin2 kc þ sh2 l
HðlÞ ¼ ctgkc : ð7:67Þ
2a0 ðsin2 kc þ sh2 lÞ3=2
Let us equate expression (7.67) at the cathode apex to the reciprocal value of the
curvature radius, viz.
cos kc 1
Hð0Þ ¼ ¼ :
a0 sin2 kc r0
If we assume that kc 2 (0, ka) and consider half-space z
0, the distance from
the cathode apex to the anode appears equal to
d ¼ zjl¼0 ¼ a0 cos kc :
The parameters of the simulated emission system were represented with the
following values: d = 2 cm, r0 = 1 µm (Fig. 7.16). The curvature distribution over
the cathode surface meeting the parameters taken is shown in Fig. 7.17. The cur-
vature can be seen to quite rapidly decrease with growth of curvilinear coordinate.
With the parameters specified, the system is a classical field-emission diode, which
makes it possible to consider the electron-path plotting problem. For its solution
and determination of the electric field intensity near the apex, we have to know the
potential-energy distribution generating the force field in the gap between the
cathode and the anode.
Let us assume that the gap between the electrodes contains such a low volume
concentration of electrons that potential distribution follows the Laplace equation:
Du ¼ 0
z, cm
z, cm
3 3
2
2
1
0
−1 1
−0. m 0
5 0.5 y,
0 0
0.5 5 −2 −1 2
−0. 0 1
x, 1 1 x , cm
m −
1 1
Fig. 7.17 Distribution of H , m H, m
curvature and its radius over 12.5 1.0
the cathode surface
10 0.8
7.5 0.6
5 0.4
2.5 0.2
r0
0 0.0
0 1 2 3 4
(
0; if k ¼ kc ;
u¼
u0 ; if k ¼ ka :
The boundary condition meets the equipotentiality of the cathode and anode
surfaces, i.e. the electrodes are metals. In the coordinate system selected, the
equation is as follows:
@ @u @ @u sin2 kc þ sh2 l @ 2 u
sin kshl þ sin kshl þ ¼ 0:
@k @k @l @l sin kshl @u2
7.6 Computation of Electron Paths in a Field-Emission Cathode-Based System 415
k
uðkÞ ¼ c1 ln tg þ c2 ;
2
tg k/2 > 0 at k 2 [kc, ka]. The boundary conditions determine the integration
constants
" #
ln tg k2
uðkÞ ¼ u0 1 : ð7:70Þ
ln tg k2c
1 du u0 1
Eðk; lÞ ¼ pffiffiffiffiffiffi ek ¼ k
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ek : ð7:71Þ
gkk dk a0 ln ctg 2 sin k sin2 k þ sh2 l
10
15
20
1
3
0.5 m
0
2 z, c
0.5 1
x, 0
m 1
416 7 Computation of Field-Emission Cathode-Based Electron Guns
2 1 0 1 2
x , cm
The electron that has got into the gap between the cathode and the anode finds
itself in force field − eE (Fig. 7.19; u0 = 20000 V, equipotential lines being drawn
through 1 kV). The direction of the field along vector –ek indicates that the field
lines result from intersection of the planes passing through axis Oz and ellipsoids,
i.e. they are the arcs/quarters of the ellipses with same focal distances a0.
On the cathode surface at k = kc, the intensity module assumes the following
values:
u0 1
EðlÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi : ð7:72Þ
a0 sin kc ln ctg k2c sin2 k þ sh2 l
The electric field intensity on the cathode surface is critical for field emission of
electrons. Its dependence on curvilinear coordinate l is shown by Fig. 7.20 for
different apex-peak curvature radii at the same voltage between the cathode and the
anode. The results show that value E quite rapidly decreases from the cathode peak
to its periphery. The intensity distribution gives an idea of the geometrical factor of
field b.
The result is the mathematical model of a pointed cathode-based field-emission
system. Within the framework of the model, it is relatively easy to get the distri-
bution of the potential energy of an electron in the interelectrode space. The geo-
metrical parameters specified make it possible to consider the electron-path plotting
problem. In addition, we get the distribution of electric field intensity over the apex
surface.
To obtain the particle paths, the Lagrange equations of the 2nd kind is formu-
lated and solved in numerical terms. The solution is based on the Runge–Kutta
method of the 8th order.
7.6 Computation of Electron Paths in a Field-Emission Cathode-Based System 417
2.5
2.0
1.5
1.0
0.5
0 1 2 3 4
2
0 1.5 m
0.5 1
1 0.5
1.5 0
m
Figure 7.21 shows the paths of the particles that have left the cathode under
different starting conditions, as well as the field line. The starting conditions are
selected so that to show the effect of change of the generalized velocities on the path
behavior, no establishment of the physical sense of these values having been
planned within the framework of the paper. Obviously, the emission-system
parameters considered do not make it possible to approximate the paths with field
lines.
Let us vary the relation of cathode-curvature radius r0 to electrode spacing d to
determine what system parameters make it possible to replace the paths with force
lines. It has already been indicated in the preliminary survey to the paper that it is
important to monitor the intensity on the cathode surface because the metallic
418 7 Computation of Field-Emission Cathode-Based Electron Guns
Table 7.6 Error of path d (lm) r0 (lm) r0/d lmax l0 (kV) Error (%)
replacement with force lines
5 50.0 10.0 0.60 16.5 3
5 25.0 5.0 0.57 15.5 6
5 5.0 1.0 0.45 10.9 21
5 2.5 0.5 0.37 8.2 32
bonding in the sample is broken at the values of about 1010 V/m thereby destroying
the sample.
Relation between the electric field intensity at the apex peak and the voltage
between the cathode and the anode is given by formula (7.72). Let us fix the
intensity value at E = 3.5 109 V/m. The parameters used, viz. d = 2 cm and
r0 = 1 µm determine the emission-image area boundary (lmax = 0.005) that meets
the reduction of E by about 1.2 times. Proceeding from this information, we
recalculate voltage u0 and boundary value lmax when varying relation r0/d.
Table 7.6 shows the error of path replacement with force lines calculated as a
relative deviation of the finishing points of extreme path (kc, l, u) at zero initial
velocities. The electrodes are spaced 5 µm apart. The error can be seen to become
admissible only at r0/d = 10.0. Actually, that brings us to flat geometry.
The computations made show that the projections of cathode-surface point (kc,
l, u) to the anode by force lines and the paths differ no more than twice under
rather adverse conditions. Such a result can only be achieved by increase in the
cathode-curvature radius with a consequent necessity to step up the voltage between
the cathode and the anode, which is undesirable. Another way out is to reduce the
space between the cathode and the anode, which reduces the significance of the
path-problem solution since small linear dimensions of the system, in the order of
fractions of micrometer, suggest solving such problem by quantum-mechanics
methods, i.e. in terms of probabilities.
The operation of the path-plotting algorithms suggested can be demonstrated by
simulation of an emission image in a field-electron microscope. Then it is necessary
to have an idea about current-density distribution over the cathode surface. Let the
image brightness be proportional to the current density. Hence, we face the problem
of projection of the current-density distribution to the anode with the paths. Let us
compute the current density by the following formula:
Zþ 1
j¼e mðEf ÞDðEf Þdef :
ðEF þ UÞ
–50
–100
–150
–200
–250
–300
0 20 40 60 80 100
(d – z ), nm
The distribution of the curvature over the cathode surface with the parameters
selected shows (Fig. 7.17) that the surface segment is considered to be flat for the
tunnel effect. Let us introduce local coordinate f normally counted off from the
cathode surface. Then the potential barrier along the electron path can be described
as the following dependence:
e2 1
UðfÞ ¼ eEf :
16pe0 f
ð7:73Þ
Here Ф depends on the point where an electron escapes from the cathode sur-
face, viz. its curvilinear coordinates; E is determined by formula (7.72).
Smoothness of the potential energy in point f = 0 is achieved automatically.
Constants f1 and G are derived from the equations, viz.
e2 1 e2 1
EF U þ Gf21 ¼ eEf1 ; 2Gf1 ¼ eE þ :
16pe0 f1 16pe0 f21
420 7 Computation of Field-Emission Cathode-Based Electron Guns
e2 EF þ U
Q¼ ; c¼ :
16pe0 eE
ðEF þ UÞ2
eE\ ;
3Q
which holds true in all physically implemented cases. Even if the total value EF þ U
of makes 5 eV, the field intensity has to be kept below 2 1010 V/m to prevent the
emission from displaying its explosive nature. Then we select the smaller root; the
larger root yields a negative value for G, whereas parabola EF Uðl; uÞ þ Gf2
has to have upward branches (Fig. 7.23). The result is the sought-for values:
"sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi# " #
3Q 1 Q
f1 ¼ c 1 1 ; G¼ eE :
c ðEF þ UÞ 2f1 f21
–2 Φ
–4
–3
–6 – ξF – Φ + Gζ 2
ξF
–8
–11
–10 –0.1 0 0.1
0 2 4 6 8 10
ζ , nm
7.6 Computation of Electron Paths in a Field-Emission Cathode-Based System 421
the force field for the electron can be considered as homogeneous with a relative
error under e.
Let us note that values f1 and G depend on coordinates l and u; f2 depending on
l and error e.
To get the current-density distribution over the cathode surface, we integrate the
external distribution of electrons, which is intrinsically the product of the delivery
function vint by the transparency of the potential barrier considered. The areas under
the curves being equal to one, the dependence of the normalized external distri-
bution vext on the energy for different field intensities at the cathode E at ambient
temperature and temperature T at intensity E = 3.5 109 V/m is presented in
Fig. 7.24. Normalization was based on the compound formula of trapezoids.
To find the current density, the external distribution of electrons has to be
integrated over all possible values of energy in the same way as in the course of
normalization. But here the integration was based on the Monte–Carlo method. This
method provides a great number of implementations of a random variable formed
so that its mathematical expectation coincides with the exact value of the solution of
the problem at hand. The variance of the probabilistic assessment of the problem
solution depends on the number of statistical tests. In our case, random points were
generated in the area considered and the function values in them were averaged.
Selection of this integration method makes it possible to simulate the
0
EF Φ
2 0K
293 K
1 1000 K
2000 K
3695 K
0
–10 –8 –6 –4 –2 0
ξ,
422 7 Computation of Field-Emission Cathode-Based Electron Guns
2.4
j =1.31·10 6A/m 2
2.0
1.6
1.2
0.8
0.4
0
10 10 2 10 3
N
ZEmax
eðEmax Emax Þ XN
j¼e mext ðEÞ dE mext ðEi Þ;
N i¼1
Emin ð7:74Þ
Ei 2 ðEmin ; Emax Þ;
Ei ¼ Emin þ ðEmax Emin Þ c; c 2 ð0; 1Þ:
50 0.9
0 0.6
-50 0.3
-100 0.0
-100 -50 0 50 100
x, μm
Zþ 1
j¼e minx ðEf ÞDðEf ÞdEf
ð EF þ U Þ
References
1. V.P. Ilyin (ed.), Algorithms And Methods for Calculating the Electron-Optical Systems
(Novosibirsk, 1983), p. 190
2. L.A. Baranova, S.Y. Yavor, The electrostatic electron lenses. M.: Nauka, 192 (1986)
3. V.P. Ilyin , Numerical methods for solving the electro-optics problems. M.:Nauka, 202 (1974)
4. N.D. Devyatkova (ed.), Methods for calculating the electron-optical systems. M.: Nauka, 174
(1977)
5. V.P. Ilyin (ed.), in The Methods of Calculating Electron-Optical Systems: Proceedings of IV
All-Union Sem (Novosibirsk, 1982), p. 202
6. A.P. Drum, V.V. Bulavinov, P.P. Konorov, Non-equilibrium electronic processes in layers on
the surface of silicon. in Molecular and Electronic Processes on the Interphase Boundaries.
(L.: Publishing of Leningrad University, 1989), p. 3–27
7. V.A. Syrovoy, Calculation of forming the electrodes in the optical planar electron beams.
Radiotehnika i Elektronika 39(3), 481–502 (1994)
8. V.A. Syrovoy, Calculation of forming the electrodes in optics rotationally symmetric electron
beams. Radiotehnika i Elektronika 39(4), 666–687 (1994)
9. L.E. Tsyrlin. Selected problem for electric magnetic fields. M.: Soviet Radio 319 (1977)
10. L.A. Neganov, V.A. Syrovoy, V.N. Tskhai, Calculation and experimental investigation of the
electron gun for technological purpose. Radiotehnika i Elektronika 35(10), 2146–2155 (1990)
11. V.V. Plohov, V.A. Syrovoy. On calculation of multi beams injector with a relativistic electron
beam. Radiotehnika i Elektronika 35(12), 2582–2593
12. V.A. Syrovoy, On the synthesis of nonparaxial relativistic electron beams at emission limited
by temperature, and the injection of a non-zero velocity. Radiotehnika i Elektronika 42(3),
348–360 (1997)
13. Y. Xuebiao, X. Zhangchcheng, H. Guoguang et. al. Emission characteristics of the
molybdenum-coated silicon field emitter array. J. Phys. D: Appl. Phys. 29(3), 506–510 (1996)
14. M.I. Yavor, E.V. Strigova, Field distribution and electrical properties of electrostatic conical
slit lenses. Nucl. Instr. Meth. Phys. Res. A 363(1/2), 445–450 (1995)
15. L.N. Baranova, N.S. Ulyanova, S.Y. Yavor, The study of spatial and temporal characteristics
of the charged particle beam focused by axisymmetric electrostatic lens. ZTP 61(2), 144–148
(1991)
16. I.M. Gordion, I.D. Tokman, The problem of electrostatics to an oblate spheroid in the point
charge field. ZTP 67(2), 121–122 (1997)
17. G.Ch. Shushkevich, The electrostatic problem for the disk and torus. ZTP 67(4), 123–126
(1997)
References 425
18. V.V. Rassadin. Calculation of the electric field in periodic structures of linear ion accelerators.
in A system of linear accelerators and the use of beams of charged particles. (M.: Publishing
House of -the Moscow Engineering Physics Institute, 1987), p. 41–45
19. L. Klemar, E. Kasper, On the numerical design of elecnron guns. Optic. 72(1), 29–30 (1985)
20. S.K. Godunov, E.I. Romenskii, G.A. Chumakov, in Construction of numerical grids in
complex domains using quasi-conformal mappings. (Novosibirsk: Nauka, 1990), pp. 75–83
21. A.Taflove, Emerging applications for FD-TD computation. IEEE on Comp. Sci. Eng. 2(4),
24–34 (1995)
22. S. Roques, M. Denizart, F. Sonier, Tetrode field emission guns for electron microscopy.
Optic. 61(1), 51–66 (1983)
23. M.I. Yavor, E.V. Strigova, Field distribution and electrical properties of electrostatic conical
slit lenses. Nucl. Instr. Meth. Phys. Res. A 363(1/2), 445–450 (1995)
24. G.V. Lebedev, N.A. Timchenko, Complex of programs for design of electron-optical systems
“Probe.”. Sci. Instrum. Electron-ion Optics L.: Nauka, 111–116 (1989)
25. N.V. Egorov. Calculation and optimization of the characteristics of the field emission electron
source. in Modeling and Control in the Electrical and Mechanical Systems, ed. by D.A.
Ovsyannikov (St. Petersburg: St. Petersburg State University Publishing House, 2002),
pp. 111–127
26. E.M. Vinogradova, N.V. Egorov. Calculation of the electrostatic field systems with coaxial
axially symmetric electrodes. Radiotehnika i Elektronika 52, 225 (2007)
27. A.A. Almazov, N.V. Egorov, Optimization of multi-tips field source. Vacuum 52, 295 (1999)
28. A.A. Almazov, N.V. Egorov, Optimization of multi-tips emission systems. Radiotehnika i
Elektronika 40(4), 638–643 (1995)
29. E.M. Vinogradova, Calculation of electron-optical systems based on field cathodes, in
Modeling and Control in the Electrical and Mechanical Systems, ed. by D.A. Ovsyannikov
(St. Petersburg: St. Petersburg State University Publishing House, 2002), pp. 99–110
30. E.M. Vinogradova, N.V. Egorov, R.Y. Baranov, in Calculation Of The Electrostatic Potential
In Single Tip And Multi-Tips Emission Systems, 1 (Vestnik of St. Petersburg State University,
Ser. 10, 2007), p. 39
31. E.M. Vinogradova, N.V. Egorov, R.Y. Baranov, Mathematical model of the cathode
assembly of the electron gun field, (Vestnik of St. Petersburg State University. Ser. 10, 2006),
p. 3
32. E.M. Vinogradova, N.V. Egorov, Calculation of the electrostatic field of spherical segments.
ZTP 78(8), 128 (2008)
33. K.A. Krimskaya, Mathematical modeling of the formation of the electron beam on the basis of
the field cathode. Diss. Cand. SPb, p. 85 (2009)
34. N.V. Egorov, E.M. Vinogradova, Mathematical model of electron gun on the field emission
electron cathode basis. Vacuum 57, 267–281 (2000)
35. N.V. Egorov, E.M. Vinogradova, Mathematical modeling of electron beam formatting
systems on the basis of field emission cathodes with various shapes. Vacuum 72, 103–111
(2004)
36. D.A. Ovsyannikov, N.V. Egorov, Mathematical modeling of systems of formation of electron
and ion beams (St. Petersburg State University Publishing House, St. Petersburg, 1998),
p. 276
37. N.V. Egorov, A.G. Karpov. Diagnostic information and expert systems (SPb.: Publishing
house of St. Petersburg University, 2002), p. 470
38. E.P. Shurina, M.U. Velikaya, M.P. Fedoruk, Algorithms for solving Maxwell’s equations on
unstructured grids. Comput. Technol. 5(6), 99–116 (2000)
39. A.A. Almazov, N.V. Egorov, By the method of calculating the field-emission systems.
Radiotehnika i Elektronika 31(12), 2452–2458 (1986)
426 7 Computation of Field-Emission Cathode-Based Electron Guns
40. Almazov A.A., Egorov N.V. Mathematical model of field-emission diode. Mathematical
methods of modeling and analysis of controlled processes.( L.: Publishing of Leningrad
University, 1989), pp. 20–28
41. A.A. Almazov, E.M. Vinogradova, Egorov N.V. Mathematical model of the electron gun
with field cathode. Abstracts of 12th All-Union. Sem. on linear particle accelerators. Kharkov,
1991. p.39
42. P.F. Bovey, A 100 kV high resolution analytical STEM. Mirrosc. Spectrosc. Electron. 1,
507–508 (1976)
43. G.F. Vasiliev, Influence of the form of the potential barrier at the emitter-vacuum and electric
field distribution on the surface of the emitter in the form of current-voltage characteristics of
field emission. Radiotehnika i Elektronika 5(11), 1857–1861 (1960)
44. G. Rekleytis, A. Reyvidran, K. Regsdel. Optimization in Technology. M.: Mir 2, 1986
45. N.S. Demchenko, A.Y. Antonov, Construction of the electron trajectories in the emission
system. , 3 (Vestnik of St. Petersburg State University, Ser. 10, 2010), pp. 1–10
46. M. Matsumoto, T. Nishimura, Mersenne Twister: A 623-dimensionally equidistributed
uniform pseudorandom number generator. ACM Trans. Model. Comput. Simul. (Special
Issue on Uniform Random Number Generation) 8(1), 3–30 (1998)
47. I.M. Gribkova, in Mathematical Modeling of the Images in a Field Electron Microscope. eds.
by N.V. Smirnov, G.S. Tamasyan. The Control Processes and stability. Proceedings of XXIX
Scientific Conference of Applied athematics and Control Processes Faculty of Saint
Petersburg State University (Publishing House of St. Petersburg State University, St.
Petersburg, 2008), pp. 116–121
Chapter 8
Field Emission Cathode-Based Devices
and Equipment
Abstract In this chapter the main features of field cathode-based devices are
reviewed. Such devices include various lighting sources, microwave appliances,
X-ray tubes and other electronic devices.
Light sources are an integral part of our life. The need for artificial illumination and
data-display devices steadily grows. A large number of light systems designed with
account of specific requirements to color rendition, luminosity and illumination
levels are under development. We know a wide class of light sources, viz. con-
ventional incandescent, halogen and fluorescent lamps intended to illuminate pre-
mises and advertisements, high-pressure gas discharge lamps intended to illuminate
large spaces and streets, semiconductor light-emitting diodes, i.e. LED, and organic
light-emitting diodes, i.e. OLED, including illumination systems, video-wall light
elements and cathodoluminescent lamps. Nevertheless, every light source has an
inherent weakness, e.g., an imperfect emission spectrum, a long setup time or an
insufficient efficiency factor.
The latest review of cathodoluminescent light sources it presented in [1].
Table 8.1 gives comparative characteristics of a small-button glass field emis-
sion cathode-luminescent light source (cf. Sect. 8.2.2) compared to industrially
manufactured light sources.
Summary: a field emission light source is comparable with the best up-to-date
light sources in terms of its parameters, but surpasses them in the working-
temperature range, luminescence uniformity, low production price, use of inex-
pensive, available and naturally widespread materials, as well as high ecological
compatibility.
Table 8.1 Comparative characteristics of a small-button glass field-emission cathodeluminescent light source
Basic parameters Field Incandescent Luminescent Halogen light bulbs Metal-halogen light bulbs LED lamps
emission light bulbs light bulbs
light source
Lamp type No Incandescent Gas-discharge Incandescent Gas-discharge/arc Semiconductor
incandescent
vacuum
Efficiency (lm/W) 30–60 10–12 40–60 22 80 *30
Energy conversion efficiency 20–30 5 *20 10 *30 20–30
8
(%)
Power-up time (s) 0.0005 0.1 1–3 0.3 3–4 0.0005
Voltage drop sensitivity Low Low High Medium Medium High
Stability of light current High Medium High Medium High High
Environmental damage High −196 to Low Medium Low Medium High −40 to
resistance +150 °C +85
Needs special voltage source Yes No Yes Yes Yes Yes
Lifetime (h) Up to 1000 10,000 180–600 15,000 50,000a
100,000a
Estimated costs for mass Medium Low Medium Medium High High
production
Materials used in production, Glass, steel, Glass, steel, Glass, steel, Glass, steel, tungsten, Glass, steel, tungsten, Indium,
exploitation and utilization carbon tungsten mercury bromine/iodine/chlorine bromine/iodine/chlorine, arsenic,
mercury gallium, glass
a
Lifetime was estimated using the common engineering method
Field Emission Cathode-Based Devices and Equipment
8.1 Light Sources 429
dQ
U¼ ;
dt
Z780
U ¼ 683 Uek ðkÞV ðkÞdk;
380
where n is the number of spectral lines; Uek ðkÞ is the spectral-flux density function;
(780–380 nm) is the visible-light wavelength range, nm.
Pointed-source luminous intensity is a space light-flux density,
dU
I¼ :
dX
where [R], [G], [B] are the unit quantities of the primary colors of the measurement
system; R, G, B are the fractions of the unit primary colors providing color equality,
i.e. the color coordinates.
LAW II: Color continuously changes with a continuous change of spectral-power
distribution.
LAW III: The mixture color depends only on the colors of blending agents and does
not depend on their spectral distributions, the emission property being termed
metamerism.
Grossmann’s laws generate an idea of the color vector having three components
of the same origin and different directions in space. The aggregate of the tridi-
mensional color vectors makes color space (Fig. 8.1).
Every color corresponds to only one vector in color space. The colors lying in
one line coming from the origin of coordinates, i.e. differing among themselves
only by intensity, are characterized by the same chromaticity. Thus, chromaticity
indicates the direction of the color vector in space and depends on two coordinates.
The chromaticity coordinates are described by a color triangle, which represents
a section of a three-coordinate space passing through the unit colors of the mea-
surement system selected. The chromaticity coordinates are relative values deter-
mining the position of a point in the triangle.
Basically, the existing light-source versions can be divided into two main classes
(Fig. 8.2) [3], viz. those functioning in the mode of transmission and reflection.
Generally, the basic classical version is similar to a usual field emission microscope.
Fig. 8.2 Principles of light source designs: a, b based on clearance; c based on reflection. 1 Core
field emission cathode; 2 modulator; 3 flow electrons; 4 phosphor; 5 transparent conductive
coating; 6 output glass; 7 visible light; 8 aluminum coating; 9 grid; 10 cathode matrix; 11 substrate
It consists (Fig. 8.2a) of a rod-type field emission cathode 1. Such a cathode can be
represented with any configuration, be it a graphite rod, a carbon fiber bundle or
nanostructures at the end face of an appropriate rod. Electrons are extracted with a
metal orifice plate. The cathode-modulator angle configuration is selected so that
three mutually exclusive factors are optimized, viz. the maximum current trans-
mission through the modulator, the minimum control voltage and the maximum
uniformity of electron flow over the screen surface.
The latter is most easily achieved by overlapping of field emission flows from
the sufficient number of emission centers. That is achieved by application of several
rod-type field emission cathodes and respective modulator orifices arranged uni-
formly in a circle to form a round-section light source.
In such classical light sources, luminophore brightness is used only partially (up
to 30%), because some photons go back to the envelope. Better brightness can be
obtained if one would switch the positions of luminophore and conductive coating
and if the conductive coating is made of aluminum (Fig. 8.2b). An aluminum
coating increases brightness up to 70% of maximum possible value.
The maximum efficiency is achieved when the light source is in a reflection
mode [4] (Fig. 8.2c). In this case, electrons 3 travel to luminescent layer 4 deposited
on aluminum mirror 8. In this case, the loss of light is only stipulated by negligible,
viz. less than 10%, light absorption in target glass 6 and transparency of matrix
cathode 10. The latter can be increased up to the level of less than 10%.
The particular technological requirements of practical application stipulate
selection of one of these principles or their modification.
The below covers different light-source versions.
432 8 Field Emission Cathode-Based Devices and Equipment
Fig. 8.4 a Light source design with the field emission cathode from a bundle of carbon fibers: 1
bulb; 2 high-voltage input; 3 phosphor; 4 layer of aluminum; 5 modulator with a diaphragm; 6
glass rods; 7 emitting part of the bundle of carbon fibers; 8 justified disks; 9 vitrified carbon fiber;
10 adjustment disks; 11 contact unit of the field emission cathode; 12 terminals; b appearance of
the lamp
The working voltage at the modulator ranges from 1.5 to 2.0 kV for this
structure. Similar results are given in [8, 9], where field emission cathodes are also
represented with PAN carbon fibers. The efficiency of the electron gun is improved
by modernization of the cathode-modulator unit of the lamp. The electron gun
(Fig. 8.6b) [10, 11] is redesigned so that current hogging is quite negligible, less
than 1%, unlike the cathode-modulator unit of the previous design, where some
electrons are sure to be retained by the control electrode (Fig. 8.6a).
A series of volt-ampere characteristics taken (Fig. 8.7) for a new design of the
electron gun shows that the anode voltage of +10 kV, which is the operating mode
of cathodoluminescent lamps, causes the maximum modulator voltage of 1200–
1300 V at the cathode current of 100 lA [12].
To increase light-emission power, one has to increase the emission current of the
field emission cathode. As the current from one carbon-fiber bunch is limited for
long-term continuous operation with the value of about 100–150 µA, the current
increase assumes that the field emission cathode contains several carbon-fiber
bunches. Particularly, papers [13, 14] mention a field emission cathode of 10
carbon-fiber bunches arranged in a circle (Fig. 8.8). In this case, the lamp dimen-
sions remain almost similar to those of the one-fiber field emission cathode.
Availability of a large number of fiber bunches in the field emission cathode makes
it possible to increase emission current up to 10 mA. However, such a heavy
current can overheat the anode. Another effect of application of a multi-bunch field
emission cathode is increase in screen glow uniformity. In this design, the modu-
lator is represented by a thin molybdenum grid stretched taut onto a ring.
The most effective design of a vacuum-tube light source based on reflection and
consisting of many carbon-fiber bunches is presented in Fig. 8.9.
Fig. 8.7 Dependence of the cathode current on the voltage on the modulator for the CMU of the
first design (a) and the optimized design (b) at various values of anode voltage: (1) Ua = +10 kV,
(2) Ua = +9 kV, (3) Ua = +8 kV, (4) Ua = +7 kV
A large size of the anode and a sufficient quality of carbon fibers make it possible
to generate a heavy emission current and an appropriate light flux.
The light flux can be increased even more by making a convex-concave structure
or a structure in a form of convex collecting lenses on the anode surface. That
increases the area coated with the luminophore and, accordingly, the light flux [15].
Along with carbon fibers as field emission cathodes, light sources employ other
materials as well, e.g., carbon nanotubes. Essentially, such lamps are designed in
the way similar to the one in Fig. 8.8. Nanotubes are laid on flat bases of a few
square millimeters either with the CVD method [13] or the print method [14]. At
the cathode-control grid distance of 0.2 mm, control voltage is 300 V. At the anode
voltage of 10 kV, the anode current is −60% total emission current of the field
8.1 Light Sources 437
emission cathode. Brightness of such a source is 6.3 104 cd/m2 for a green
luminophore (ZnS:Cu) at the anode current of 200 µA, the anode having been
aluminized. Efficiency of such a source is 70 lm/W.
The authors have elaborated skeleton pattern [16] into a high-brightness light
source [17] shown in Fig. 8.10. The field emission cathode is made of special
multiple-wall nanotubes termed as nanofibers fabricated in an electric arc in the
hydrogen atmosphere. One design provides the cathode-current density of
100 mA/cm2 at the grid voltage of 2.9 kV and the cathode-grid distance of about
1.5 mm. Brightness of 1,000,000 cd/m2 is achieved for a green luminophore (ZnS:
Cu) in a continuous mode at the anode voltage of 30 kV and the anode current of
400 lA. The light flux exceeds 1000 lm. The light-source envelope is protected
against Roentgen rays with lead-glass coating 11 and the space between envelope 6
and protection 11 is filled up with cooling shell 10 to prevent the light source from
overheating by the high power released at the anode. Auxiliary electrode 5 placed
around the cathode-grid unit is necessary to protect the field emission cathode from
unexpected high-voltage discharges.
Papers [16, 17] suggest diode light-sources with cylindrical field emission
cathodes made of nanotubes. The circuit of such a light source is presented in
Fig. 8.11. The carbon nanotube paste is applied on the internal surface of
cathode-protective cylinder 3.
No advantages of these designs have been reported so far.
The luminophore can be applied both on one [18], and the two opposite sides of
the envelope [19].
Note should be made of the designs with field emission cathodes made of
nanograhite [20], nanocoils [21] nanotubes turned into carbon filaments [22, 23].
Practically, they coincide with the designs given in Figs. 8.4 and 8.6.
JNA NOV website [24] reports development of tiny light-sources with a field
emission cathode made of carbon nanotubes of 3 mm in diameter.
The section concludes that the few attempts to design lamp sources have not
been a success so far. First of all, that is a necessity to generate quite heavy field
emission current in vacuum in a lamp of about 10−6–10−7 mm Hg.
It is a light source with pointed field emission cathodes of about 10,000 apices
per substrate of 1 mm in diameter [25], as well as a light source with nanowires of
MoO2 [26].
8.1 Light Sources 439
Fig. 8.12 Flat light sources: a diode design of the device; b triode design of the device; 1 anode
covered with a phosphor layer; 2 glass spacer; 3 exhaust stem; 4 getter volume; 5 flat field
emission cathode; 6 modulator
The device works in a continuous mode. The anode area is 4.8 cm2. If the
voltage applied is 1200 V, the electric field being 2 V/µm, the anode-current
density is 0.2 µA/cm2. The device is about 3000 cd/m2 bright.
The basic methods of fabrication of large-area field emission cathodes include
screen printing. A field emission cathode based on diamond-dusted carbon nan-
otubes is exemplified in [29, 30]. The printing paste contains carbon nanotubes, a
binder, glass powder, a dispersing agent, an organic solvent and diamond dust at a
rate of 0.1–20 wt%.
After stirring, the paste obtained is applied on the conductive substrate by screen
printing and annealed at 330–470 °C for about one hour. The resulting structure of
the light source is shown in Fig. 8.15.
A mixture of carbon nanotubes and diamond dust produces a relatively high
density of emission current.
Moreover, such a structure is quite advantageous in terms of technology and
service properties of the field emission cathode.
The devices based on the reflection effect and ruled field emission cathodes can
be even more economical. In particular, papers [32, 33] mention such a light source
(Fig. 8.16).
Fig. 8.15 Structure of a flat light source with a field emission cathode printed from carbon
nanotubes and diamond powder. 1 Glass plate; 2 ITO layer; 3 phosphor; 4 diamond powder; 5
carbon nanotubes; 6 power supply
Fig. 8.16 The light source with line field emission cathodes based on the reflection effect: 1
bottom glass plate for light radiation lead; 2 sealant; 3 transparent conductive covering (ITO); 4 top
glass plate; 5 luminescent layer; 6 black cover; 7 aluminum anode; 8 semicylindrical groove; 9 line
cathode (silver paste); 10 nanotubes; 11 getter
442 8 Field Emission Cathode-Based Devices and Equipment
Fig. 8.17 Flat source with wire field emission cathodes. 1 Transparent glass spacer; 2 field
emission cathodes; 3 bearing wire field emission cathodes; 4 bottom base; 5 exhaust stem; 6
conductive anode cover; 7 convex surface of the luminous element; 8 phosphor; 9 top plate
8.1 Light Sources 443
emission from both plates adds up, which cannot but improve efficiency of the light
source.
The wires can be replaced with different grids represented by both perforated
plates [36] and mesh grids [37]. The grid is usually placed between two lumino-
phore coated anodes. The grid surface overlooking the anodes is coated with an
emission layer. It consists of numerous carbon nanotubes, metal conductive parti-
cles and a getter powder dispersed in a glass matrix.
The nanotubes are from 5 to 15 lm in length, the diameter ranging from 1 to
100 nm. Metal particles are powdered silver (Ag) or indium-tin oxide (ITO). These
particles provide electric contact between the nanotubes and the base, i.e. the
conductive mesh. The getter powder consists of a non-evaporating getter material,
i.e. the one containing titanium (Ti), zirconium (Zr), hafnium (Hf), thorium
(Th) and their alloys. The average size of the getter powders ranges from 1 to
10 µm.
In addition to nanotubes, field emission cathodes can be made of other materials
both carbon- and non-carbon-based; e.g., nanostructured carbon materials [38] and
carbon nanotube [39], microgranules of 12SaO; 7AL2O3 or 12SrO7AL2O3 with
particle of 0.1–100 µm [40].
The simplest structures employ field emission cathodes made of constructional
graphites [41]. The structure and the milestones of manufacturing of a diode-based
light source using a field emission cathode of constructional graphite are presented
in Fig. 8.18. At the first stage, a blank of 1 1.5 cm is cut out of a lump of
constructional graphite, viz. fine-grain high-density graphite MPG-6. The blank is
grooved 0.5 mm at a pitch of 1.5 mm according to the spark-erosion technology
(Fig. 8.18a). The grooves are added with POLYCOR (or Lucalox) type ceramic
plates as thick as 0.5 mm. The resulting structure is cast in glass-fiber reinforced
cement (Fig. 8.18b).
After the glass-fiber reinforced cement crystallizes in a muffle furnace at 400 °C,
it is ground off together with the ceramics until the graphite surface appears. Then
the surface is ground with an abrasive powder of a rated grain size, viz. *28 µm.
As Lucalox is close to abradants in hardness, while graphite is much softer, the
graphite surface is ground off below the Lucalox plate end-faces by the value close
to the size of the abradant particles. The simple technological means create a small
uniform clearance, viz. 30 µm, over the entire cathode plane (Fig. 8.18c). The
Fig. 8.18 Diagram for production of a diode light source based on constructional graphite:
a milling grooves in the graphite cathode workpiece; b inserting ceramic plates and molding with
glass frit; c completed structure. 1 Graphite field emission cathode; 2 glass frit; 3 ceramic sheets; 4
cathode-anode gap; 5 anode with phosphor and conductive cover
444 8 Field Emission Cathode-Based Devices and Equipment
grind field emission cathode structures can produce diode structures with a running
voltage of 200–300 V.
One of the most promising directions of development of flat light sources is the
lateral construction, i.e. such arrangement of the device where the field cathode and
its controlling electrode (modulator) lay in the same plane. In Fig. 8.21 one can see
one of the most well-turned examples of such design, a field emission cathode light
source presented in [42]. This construction together with modern developments in
microelectronics allowed to simultaneously achieve rather high level of homo-
geneity and strong light beam.
The conceptual design with an axial arrangement of elements is shown in Fig. 8.22.
This design employs a cathode based on fibers, nanotubes or other field emission
materials.
An important advantage of a cylindrical structure in comparison with a planar
one is that the same applied voltage produces the microscopic electric field value of
E ¼ Ur ‘n Rr, which is more than in the case of a planar structure, where R and r are
the radii of the anode and the cathode, respectively. Consequently, it is possible to
use a diode light-source structure, which is considerably cheaper in manufacturing
than the triode structure.
A cylindrical triode-based light source with field emission cathodes made of
carbon-fiber bunches is presented in Fig. 8.23 [43–45].
This light source has a conventional configuration of anode 1, i.e. a glass
cylinder-shaped base covered with a conductive coating with a luminophore layer.
Modulator 2 is a cylinder of metal gauze or etched metal foil. Field emission
cathode 3 is made of carbon fibers. There are two versions of cylindrical field
Fig. 8.23 Cylindrical light source with field emission cathodes made of bundles of carbon fibers.
a Design of a cylindrical light source; b light source cutting; c diagram of the lamp-chimney brush
field emission cathode d the field emission cathode with a radial arrangement of bundles of carbon
fibers. 1 Anode cylinder with the coating conductive and luminescent layers; 2 modulator; 3 field
emission cathode from carbon fibers; 4 bundles of carbon fibers; 5 wire at the base of the field
emission cathode; 6 clamping disk; 7 bearing disk
emission cathodes. In the first version, carbon fibers 4 are fixed between two or
three twisted wires 5 to form a brush-type structure. The structure is very easy in
manufacturing, but deformation of carbon fibers can result in their breakage and
subsequent extraction from the cathode by the electric field. Moreover, it is very
difficult to maintain the same height of the field emission cathode fibers and
respective equal spacing between the cathode and the modulator. In the second
version of the field emission cathode structure (Fig. 8.23d), carbon fibers are fixed
with conducting glue or soldered between two disks 6 and 7 made of a conducting
material. The assembled units are hafted on a bearing rod of the length desired. This
structure provides a more accurate coaxiality of the field emission cathode and the
modulator, as well as a higher structural efficiency. All those factors improve
uniformity of the field emission current over the surface of the field emission
cathode and reduce the likelihood of cathode-modulator gap bridging.
A lot of papers [46–50] are dedicated to the classical diode structure, where the
functions of a field emission cathode are performed by either a tungsten wire or a
metal rod of 2 mm in diameter coated with carbon nanotubes or ZnO nanoledges
[51]. In all the papers, the length was several centimeters, viz. from 3 to 10, and the
diameter was 15–30 mm. The nanotubes were deposited in tubular reactors made of
quartz tubes at 680–850 °C for 10–30 min. The brightness achieved was about
10,000 cd/m2 at anode voltage of 5.4 kV [46].
Efficiency of such lamps depends mostly on the nature of the luminophore and
the applied voltage and amounts to 21 lm/W for a white luminophore and 37 lm/W
for a green one [49].
8.1 Light Sources 447
Fig. 8.24 The schematic diagram of a cylindrical light source with a reflecting anode [46, 47]. 1
Aluminium semi-cylindrical covering lay phosphor; 2 field emission cathode; 3 glass bulb
Fig. 8.25 Cylindrical light source with a face arrangement of field emission cathodes.
a Diagram of the source; b a photo of a working model
Similarly to cylindrical lamps in [58, 59], a spherical envelope can be fitted with
a nanotube wire based field emission cathode. The wire base is waved or
saw-toothed and placed in the middle of the envelope. The nanotubes can be
applied both after and before the wire base is waved or saw-toothed. Application of
a reticle is also envisaged.
Finally, patent [60] finishes publication of a series of designs using a field
emission cathode of glass with metal and getter particles dispersed by nanotubes in
the same way as [37, 54].
Fig. 8.26 Miniature UV lamp: 1 evacuated case made of UV-permeable dielectric; 2 anode; 3 UV
phosphor; 4 aluminum layer; 5 modulator; 6 terminal lead; 7 nanostructured carbon field emission
cathode; 8 alignment disk; 9 contact node; 10 field emission cathode terminal leads
long operational time (50,000 h), almost zero time to turn on, a high resistance to
mechanical vibrations and voltage fluctuations, and the absence of polluting sub-
stances. Phosphors of various chemical compositions may be utilized, depending on
the necessary UV radiation wave-length [65–69]. For example, high-performance
Bin Ym Al2(BO3)4 compositions are suitable for the purpose at a wavelength of
300 nm. The radiation spectrum of a UV lamp with one such phosphor is presented
in Fig. 8.27. The efficiency of these phosphors is high enough for practical appli-
cations (27 mW at an overall power of 900 mW; U = 8 kV, J = 112 mA).
Phosphors of a different composition, e.g., ZnAl2O4, are needed if the wavelength is
to be decreased to 245–260 nm [68–70]. Only the first steps have been taken to
design vacuum luminescent lamps using field emission cathodes and to search for
the optimal chemical compositions of phosphors. There- fore, considerable effort to
continue development of this technology can be expected in the near future.
Therefore, field emission light sources, which have low power consumption and
more uniform glow, are in increasingly better demand.
Some designs of cathodoluminescent back lights using carbon-based field
emission cathodes are given below. Some of them have something in common with
flat light sources (Sect. 8.2.3). In particular, Fig. 8.28 presents a back light circuit
version. The lamp consists of anode plate 1 bearing a transparent coating layer 2
and a luminophore layer 3. Spacer 4 sets the gap between the anode and the
cathodes located on lower substrate 5.
The parallel bars of transparent conductive ITO coating 6 bear thin metal layers
7 serving as the contacts of field emission cathodes 8, 9. Field emission cathodes 8,
9 are formed by well-known above-described methods, e.g., screen-printing with
the carbon-nanotube paste. It is peculiar that electrodes 8 and 9 of this design can
alternatively function as a field emission cathode and a modulator. This feature
nearly doubles the service life of such a lamp at the same brightness and increases
uniformity of its glow, wherefore they are connected with each other skipping the
nearest one.
Control with the neighboring parallel electrodes in a way similar to [71] is also
applied in [72, 73]. The quality of the backlight is also determined by luminophore
flare uniformity. With this end in view, the modulator is shaped as a ruler
(Fig. 8.29) [74].
As the picture shows, control electrodes 6 are placed on linear insulators 5
located at right angle to cathodic electrodes 4. Cathodic electrodes 4 can be made of
any conducting material; e.g., a metal or a transparent conducting coating (ITO).
Insulators 5 are 3–10 µm high, the distance between the rulers ranging from 10 to
30 µm. Field emission cathodes 8 are made of the carbon nanotube paste by
conventional screen printing. This design has a wider electron-divergence angle,
Fig. 8.28 Backlight with printed field emission cathodes made of carbon nanotubes [54]: 1 anodic
plate; 2 transparent conductive cover; 3 layer of phosphor; 4 spacer; 5 cathodic plate; 6 thin layer
of metal; 7 ITO layer; 8, 9 field emission cathodes manufactured using paste which contains
carbon nanotubes
8.1 Light Sources 453
shown in Fig. 8.30 [76, 77]. The basic framework is passive gate LCD matrix 2
with a polaroid 1. The matrix is combined with light filter 3 having areas of three
different colors, viz. red (R), blue (B) and green (G). The LCD matrix is lit up with
white-color backlight 4 working in a continuous mode.
Light flux Y generated by backlight Y after the light filter is equal to ŋcf, after the
LCD matrix to Yŋcfŋm, at the front panel of the screen Yŋcfŋmŋp (ŋcf, ŋm, ŋp the
average transparency of the light filter, the matrix and the polaroid, respectively).
Every color element of the light filter passes less than 1/3 of the spectrum. The
exact value is determined by the spectral characteristic of the backlight and the
desired relation between the intensity of red, blue and green at the screen output. As
the average transparency of the light filter is ŋcf * 0.33, a considerable part of the
light flux is wasted in the light filter.
The light efficiency of the screen can be improved by removal of the light filter
and employment of a polychromatic backlight able to switch-on/off the color ele-
ments in a synchronous mode. The arrangement of such a screen is shown in
Fig. 8.31. This screen uses the same matrix 2 and polaroid 1 as the screen in
Fig. 8.30. However, the white backlight is replaced with trichromatic backlight 3
able to switch on each color in an independent way. The basic requirement to a
polychromatic lamp is matrix-lighting uniformity for each color. The sizes of the
backlight color-element do not have to coincide with the matrix-pixel size. At the
given flare non-uniformity criterion, the size of the backlight color-element deter-
mines the gap from the backlight to the LCD matrix.
Every backlight color-element switches on after the LCD matrix develops the
image appropriate to this color. At that, the effective screen resolution, viz. the
number of image elements, trebles as compared with the screen resolution in
Fig. 8.30 because every pixel of the matrix is engaged.
At the frame frequency of 100 Hz, each of the 3 color elements of the backlight
is switched on for not more than 3.3 ms at a power-on time 10 ms. Accordingly, the
lamp-lighting time should be less than 0.1 ms. Fluorescent backlights widely used
now cannot meet these switch-on/off rate requirements. The most acceptable ver-
sion providing high control speeds and high brightness is cathodoluminescent light
sources with a field emission cathode. The overall efficiency and brightness of such
lamps are determined by the luminophore and good luminophore can provide them
at nearly the same level as fluorescent lamps. A field emission cathode used makes
it possible to control the lamp with a frequency from 10 kHz upwards. The theo-
retical limit is determined by the phosphor-persistence time and can last a mere few
microseconds. The trichromatic backlight control is exemplified in Fig. 8.32.
Fig. 8.32 Graphic display (right) of the behavior of an LCD matrix and backlight over time: 1
image on the LCD corresponding to the color red (red backlight); 2 image on the LCD
corresponding to the color green (green backlight); 3 image on the LCD corresponding to the
color blue (blue backlight)
The red, green and blue lamps switch on in series, and the liquid-crystal matrix
elements switch on in step with them and in an appropriate configuration with the
video image.
As the sizes of lamp-color elements do not have to coincide with the dimensions
of the LCD matrix, one of the first devices for dynamic light consisted of a set of
small-button glass light sources of three colors (Sect. 8.1d) [77]. A photo of the
dynamic-light unit is presented in Fig. 8.33.
456 8 Field Emission Cathode-Based Devices and Equipment
The light sources of each color were switched on in series. The frame rate was
100 Hz.
Cathodoluminescent lamps are to have the switching rate of less than 1 ms. The
switching rate of lamps mostly depends on run-up time and luminophore persis-
tence. For the lamps used, the respective periods meet the requirement specified.
The power consumption of the video-wall unit is within 20 W or 100 mW/cm2
in a dynamic mode at the duty cycle of each color 3. The individual-source
brightness being 104 cd/m2 and the density of their packing in the unit being 2/5,
the white-light unit brightness is 1.3 103 cd/m2 in the dynamic mode and the
image brightness at the LCD screen is 300 cd/m2.
Another promising area to develop color dynamic-light lamps is low-resolution
display screens, i.e. those having the pixels of a few millimeters [78, 79].
Figure 8.34 shows the structure of a trichromatic unit based on carbon-fiber field
emission cathodes.
The anode of a conventional configuration has trichromatic luminophore-coated
sections of 10 10 mm in size. The sections are partitioned to prevent the electron
flows from mixing.
The electron beams are formed by a modulator with appropriate apertures, the
modulator being the same for all the cathodes, and controlled by the control system
through the cathode circuits.
In some cases, the flare-spot uniformity can be improved by holographic dif-
fusers [80]. A holographic diffuser is a surface relief hologram providing more
effective control over the angular distribution of light, (up to 100°) and higher
transmittance (more than 85%), than the conventional methods, e.g., clouded glass.
One of the promising areas of application of light sources is their use in the video
walls of large size and, consequently, low resolution [81]. The video walls can be
used in group displays necessary for many functions such as TV broadcasting,
advertising etc.
The units are manufactured in different sizes. Some of them are based on light
sources with field emission cathodes made of carbon-fiber bundles [82] as shown in
Fig. 8.33.
The unit has the following characteristics [83]:
– number of lamps 36,
– power consumed up to 40 W,
– luminous efficiency 18 lm/W,
– brightness up to 12,000 cd/m2,
– viewing angle 180 °C,
– pixel size 40 40 mm2 RGB,
– contrast 200:1.
Every light-emitting element of the unit is controlled by a controller.
Large full-color video walls need the units to be able to be connected into a
screen. Such a unit is to display at least one sign/letter/digit and have a power
supply and control system of its own. It is necessary to streamline the electrical
interconnections and the screen-control system in general. In this case, the unit is
only connected to the mains-supply source and control wires from the computer.
The outward appearance of the unit is presented in Fig. 8.35 [84].
Structurally, the unit is a complete sign space containing 8 8 = 64 full-color
pixels. Each pixel consists of three lamps of red, blue and green. The colors are
mixed up by pulse-width modulation of emission current. The total number of
lamps is 192. The unit structure enables to make up a panel of any size without loss
of resolution of the screen. The rated power of the unit when all the lamps work at
the same time is about 200 W. Other parameters are similar to those given above.
Designing and manufacturing of the unit solved the principal challenge of field
emission electron sources, namely identity of emission characteristics [82] in
mass-produced devices.
A new approach to lighting in industrial and office buildings is presented in [85].
In this scheme (Fig. 8.36) the electricity is supplied via a single-wire line S from
a high-frequency resonance Tesla transformer 4 (frequency *1–100 kHz).
Transformer 6 is connected through capacitors 2, 3 from converter 1. As a primary
energy source one can employ the alternating current network or autonomous
energy cells such as batteries. Field emission lamps 6 can be connected to the
ground 8 through either limiting resistors 7 or proper natural capacity (if the lamps
are suspended at some height). A variant of construction of a cathodoluminescent
lamp with a single-wire line is presented on Fig. 8.37.
The lamp on Fig. 8.37 is a transparent glass sphere 1 with electrode holders 2, 3
entering it through welded-in glass “bubbles”. One of them holds (conserving the
electrical contact) one of the field electron emitters 4.
Emitter 4 generates free electrons in one of the light sources 6. The other
electrode 2 is connected to the metallic covering of the luminophore-anode 6. The
other of the electrodes 2 holds, also conserving electrical contract, the emitter 5.
The emitter 5 generated free electrons for another light source 8. Metallic covering
9 of luminophore 10 serves as an anode for light source 8. Anode 9 is connected to
the external electrical circuit through electrode 3. Thus electrode 2 is connected to
electron emitter 5 which serves as the field emission cathode for light source 8 and
metallic covering 7 of luminophore 10 which is the cathode for light source 6. As
such, the light sources in the lamp are connected antiparallelly. So in the scheme of
electric lighting on Fig. 8.37 the cathodoluminescent lamps contain two parallelly
connected light sources 6, 8 (antiparallel relative to power supply) in each of the
bulbs. In one of the half-phases of the current the light is emitted from source 6, and
in the other half phase from source 8.
Fig. 8.36 Single-line coupling diagram for cathode luminescence field emission lamps: 1
converter; 2, 3 capacitors; 4 tesla transformer; 5 single-wire circuit; 6 cathodoluminescent lamp; 7
resistor; 8 ground
8.2 Flat Display Screens 459
coated with luminophores of three primary colors, viz. red, blue and green. The key
difference between field emission displays and electron-beam displays is that the
latter have one electron gun for all the pixels, while the display screens with field
emission cathodes have an individual electron gun for each subpixel. Such an
approach does not involve any cumbersome focusing systems like those in
electron-beam displays and enables us to make the display compact and comparable
in thickness with liquid crystal displays. That makes the technology of flat display
screens based on field emission cathodes look very promising for development of a
perfect flat screen [86].
However, it is only Futaba and PixTech that manufacture FED for commercial
application now. Their competitive but expensive displays are issued in small
quantities for special purposes.
The main reasons thereof are faults of available materials and structural concepts
of such displays. The up-to-date achievements in electronics make it possible to
develop steering circuits for the arrays of field emission cathodes, so the work is
focused on the field emission cathodes themselves and the materials to be used in
them. The materials for field emission displays develop along two main directions:
development of new luminophores with better glow and longer life, and search of
new emission materials.
The guidelines of development of field emission materials considered in this
book can help in designing mass display screens with field emission cathodes. In
terms of internal arrangement, field emission displays fall into several groups. Their
classification is shown in Fig. 8.38 [86].
Each structure solves this or that problem, but none solves all of them. The
below describes the types of structures with particular examples.
Flat display screens with field emission cathodes originate from the field emission
vacuum microtriode suggested by Shoulders [87] and the technology developed by
Spindt [88]. The specific features of the vacuum microdevices with field emission
cathodes are considered in [89].
One of the first displays of 110 90 mm, i.e. 6″ in diagonal size, was
demonstrated by Meyer [90] in 1991. Its brightness ranged from 150 to 300 cd/m2.
A pixel was as large as 0.12 mm2. The apex density was 104. The anode-cathode
gap was 200 lm. The control voltage was 80 V. The anode voltage was 400 V.
Futaba Corporation improved this technology and commercialized the displays
[91, 92]. The operating principle of such a display is shown in black and white in
Fig. 8.39a, while Fig. 8.39b demonstrates a photo of a field emission cathode array.
The control electrode has a lot of apertures of about 1 µm in diameter.
Application of voltage between the cathode and the control electrode generates a
heavy electric field in the vicinity of the top of the case. The field makes the
electrons emit from the cathode and accelerates them on their way to the anode. The
anode is reached by more than 98% of the electrons emitted. The electrons excite
the luminophore, and the light can be seen through the ITO layer.
The arrangement drawing of the display screen is shown in Fig. 8.40 [93].
By now, Futaba Corporation has developed a wide range of both monochrome
and color displays based on conical field emission cathodes. However, their sizes
are not large enough yet. There have already been reports about the display image
size of about 20 cm. The reason is a considerable difficulty to maintain high vac-
uum in the display case. The vacuum problem can be solved if the design is based
on up-to-date field emission materials that are more resistant to ionic bombardment,
viz. carbon nanotubes or graphite [94]. In addition, pointed field emission cathodes
can be controlled by different systems and fabricated by various methods [95–97].
The expertise in designing flat displays based on field emission cathodes has
shown that the wide spreading angle of field electrons can cause a simultaneous
flare of several elements of the image, which reduces the image contrast and makes
a turbidity effect on the image.
Fig. 8.39 Schematic diagram and operating principle of a flat display screen: a diagram and
operating principle; 1 anode substrate; 2 anodic electrode (ITO); 3 phosphor; 4 cold field emission
cathode; 5 control electrode; 6 insulating layer; 7 resistive layer; 8 cathodic electrode; 9 cathode
substrate. b General layout
462 8 Field Emission Cathode-Based Devices and Equipment
Fig. 8.41 The schematic design of a field emission flat display screen with increased contrast
[82]: 1 cathode substrate; 2 insulating layer; 3 control electrode; 4 cold Spindt cathode; 5 cathodic
electrode; 6 focusing electrode; 7 anodic electrode; 8 dividing barrier; 9 matrix is black; 10
phosphor; 11 anode substrate; 12 cathode circuit controller; 13 control circuit controller; 14
focusing circuit controller; 15 anode circuit controller
8.2 Flat Display Screens 463
separating the electron flows from different areas of luminophores. The same
purpose is attained by focusing electrode 6 located somewhat higher than control
electrode 3. This electrode is introduced to completely focus the pixels (P) released
from the field emission cathode on the respective luminophore spot 10.
The overall control of screen electrodes is exercised by respective controllers
12–15 delivering the necessary power and steering signals. When an ITO layer on
the anode is used as anodic electrode 7, it is enough to have accelerating voltage of
+5 kV in respect of cathodic electrode 5. As a rule, light intensity of the screen is
improved by using an aluminum layer as anodic electrode 7. The anode-cathode
plates are spaced about 1 mm apart. In order to increase the reaction speed of the
system, a control system with reduced proper capacitance is proposed in [101].
Chapter 5 has already mentioned that it is not only pointed field emission cathodes
of molybdenum that are under intensive development now but also pointed field
emission cathodes of different materials. The first one worth mentioning is silicon.
Silicon apices can be fabricated both by electrochemical etching [102] and
amorphous-silicon crystallization initiated by laser light [103].
Field emission cathodes for flat displays can also be made of composite mate-
rials; e.g., paper [104] suggests placing a niobium cone on the SiO2 columns of 5–
50 nm in diameter. However attractive, the pointed field emission cathodes have
already been noted above as unable to work in high technical vacuum at the
pressure 10−6–10−7 Torr in the device.
Therefore, most of our further descriptions of displays are dedicated to carbon
field emission cathode-based designs.
The diode structures are simplest and cheapest flat display screens. That is why their
development has still been attracting great attention. Especially active attempts to
develop such screens started with application of the screen-printing technique and
the electrophoretic method in production of field emission cathodes. The
screen-printing technique produced monochrome [105, 106] and color [107, 108]
display diode screens ranging from 4 to 9 inches in size.
In the former case, the screen prototype contained 128 128 graphite cathode
lines and a luminescent anode with the points as large as 300 300 µm. A pixel
was as large as 600 600 µm. In the latter case, there were 256 64 lines of
cathodes and anodes with the points as large as 450 450 µm. The pixels were as
large as 650 650 µm.
In both cases, brightness was 300 cd/m2 at 750 V, being able to reach 500 cd/m2
at 800 V. The displays were fed by impulse signals of 5 ms in duration at the
frequency of 1 kHz, the current density being 80 mA/cm2, the anode–cathode
voltage being 750 V. However, the technical characteristics of the display are noted
to be improvable by optimization and boost of the technological parameters, e.g.,
pica type and formation of surface morphology of the graphite cathode. The greatest
464 8 Field Emission Cathode-Based Devices and Equipment
successes have been noted in design of flat display screens based on finely ground
nanotubes. Paper [109] presents a version of a conventional matrix-addressed diode
display. The principle of design and control is shown in Fig. 8.42. The field
emission cathode is made as follows. A glass was etched to form grooves of
200 µm in width and 100 µm in depth spaced 300 µm apart. The emission material
which is a mixture of nanotubes and epoxy (50/50), was pressed into the grooves.
After solidification, the excess material was removed, and the surface was polished.
A test showed that the bar surface had about two apices per micrometer. The anode
was fabricated in the following way. A transparent conductive layer applied on
glass was cut into bars of 200 µm in width spaced 300 µm apart. Then these bars
were coated with Al2O3 doped with europium by electrophoresis. The anode–
cathode gap was 30 lm. The emission current necessary for the display to glow,
viz. 76 lA/mm2, was reached at the anode–cathode voltage of 230 V. The current
was studied for stability over 12 h at the constant voltage of 300 V. Fluctuations
did not exceed 8%, and no degradation could be seen.
The display was controlled by two impulses of ±150 V fed to the anode and the
cathode, respectively. A pixel was only switched on when the respective anodic and
cathodic plates were fed with simultaneous impulses. The one-pixel current was
limited to 5 lA, that constraint being established by voltage.
In order to increase efficiency of field emission current and reduction of running
voltage, one should prefer designs with nanotubes directed at right angle to the
effective area of the field emission cathode. The easiest way to achieve this is
electrophoretic deposition [110]. In this case, the threshold voltage reaches 180 V.
However, it should be noted that electrophoresis by no means always provides
verticality of the nanotubes along with their good adhesion to the substrate.
Paper [111] suggests achievement of good adhesion of nanotubes to the base by
deposition of the nanotube in the assembled device. Figure 8.43 presents the
arrangement drawing and the industrial process. A field emission display consists of
cathodic 1 and anodic 2 assemblies divided by spacers 3.
The glass base of anode 4 bears conductive coating 5 of ITO or aluminum. The
cathodic glass plate 6 bears conductive layer 7. The layer is made of an alloy with a
melting point decreasing in vacuum. The alloy is made of copper, aluminum, silver
and chromium. The first conductive layer bears nanotube suspension 8.
8.2 Flat Display Screens 465
Fig. 8.43 The schematic design for roasting of nanotubes in the ready field emission display: 1
cathode unit; 2 anode unit; 3 spacer; 4 anodic plate; 5 conductive cover; 6 cathodic plate; 7
conductive cover (low-melting-point alloy); 8 layer of carbon nanotubes; 9 spring clip; 10 vacuum
furnace; 11 heater; 12 contact wires; 13 power supply
The anodic and cathodic structures are connected with a spacer by spring clips 9.
To be sintered, the design assembled is seated in vacuum furnace 10, cathodic plate
1 directly adjoining infrared heater 11, the anode and the cathode being connected
to power supply 13 with flexible refractory-metal conductors 12.
The caking goes as follows. The furnace pressure of about 10−5 mm Hg triggers
the heating element. The furnace temperature is raised to 350 °C and maintained for
10 min. At the same time, layer 7 melts and the nanotubes from layer 8 begin
sinking into the melt. Simultaneously, the anode-cathode gap is energized. If the
anode-cathode gap is 150 µm, the voltage of 300 V is enough to orient the nan-
otubes along the electric lines of force. In a few minutes of soaking, the furnace
goes off, and the ready device is taken out.
The operational stability of the field emission display directly depends on the
residual-gas pressure inside. That is what causes a necessity in different gettering
schemes. Thereto, paper [112] suggests a separate assembly of surface
non-saturated nanotubes functioning as getters.
It has already been noted that polyacrylonitrile carbon fibers have good field
emission properties. That is why we cannot but consider some designs of displays
based on such field emission cathodes.
Figure 8.44 [113] shows a flat field emission imaging device consisting of a
hermetic glass case (unlabeled in Fig. 8.39a) shaped as a parallelepiped with
insulating optically-transparent front panel 1 made, e.g., of glass or another
visible-light transparent dielectric, the internal surface of which bears a series of
anode buses 2 parallel to each other, field emitters 3 spaced about 200 µm apart in
dielectric quartz-glass substrate 4 serving as the back panel of the case and sepa-
rated from front panel 1 with vacuum gap 5. Each of emitters 3 is a carbon
polyacrylonitrile fiber bundle (unlabeled in Fig. 8.44). The crosswise dimensions of
the bundle are about 100 µm. The bundle is extended along its axis and directed
towards corresponding anode bus 2 along the normal line to the internal surface of
466 8 Field Emission Cathode-Based Devices and Equipment
Fig. 8.44 Design of a display screen based on PAN carbon fiber. a Screen design; b emitting unit;
1 front panel; 2 anode bus; 3 field emitters; 4 dielectric substrate; 5 interelectron interval; 6
emitting surface; 7 spacer; 8 cathode buses
front panel 1. The bundle is a group of emitting elements 6 (Fig. 8.44b) opening out
into flat end-face of emitter 3, which is its emitting surface overlooking front panel
1. In this display version, emitting elements 6 are the threadlike fibrils extended
along the axis of each fiber. The fibrils are up to 1 µm in length and 1–5 nm in
diameter. The fiber bundles are put into cylindrical glass shells located in the
apertures of dielectric substrate 4 so that their ends are leveled with the shell edges
and form the end face of emitter 3 overhanging substrate 4 by about 40–400 µm,
the space between front panel 1 and back panel 4 being 0.5–1.4 mm. The spaces are
fixed by spacer 7 shaped as a rectangular projection on lateral case walls 8 standing
on both sides of panels 1 and 4 and tightly glued or welded to them. The opposite
substrate-surface 4 furthest from panel 1 bears parallel bus bars 9 of emitter
power-supply 3, each welded to the other end faces of respective emitters 3 and
crossed with anode buses 2 at a right angle. Anode buses 2 represented by con-
ductive aluminum bars as thick as a few hundreds of nanometers have cylindrical
cavities in the areas opposite respective emitters 3. Cavities are of the same size
with the image pixel, viz. the emitter size when the space between the emitter and
the anode bus is small. To make the cavities, bus bars 2 are sputtered first through
the masks round-screened in the place of the cavities and then through the same, but
unscreened masks. As a result, the cavity bottoms, which are as thin as just a few
hundreds of angstroms, become optically transparent. It is only the cavity bottoms
that the luminescent layer of a luminophore is applied to. Other buses 2 are free
from the luminescent layer. Electrical leads to buses 2 represented by a 50-µm thick
aluminum wire (absent from Fig. 8.44) are ultrasonically welded from one side.
The rectangular shape of the device case makes it possible to gang up large-area
screens intended, e.g., for advertising boards.
The device works as follows. After packaging, gas-ion treatment of the emitting
end face of all emitters 3 by delivery of voltage difference to all buses 2 and 9 for
8.2 Flat Display Screens 467
Despite their simplicity, diode screens generally have low screen glow because of a
necessity to reduce control voltage. That is why the greatest efforts are focused on
development of triode display screens. One section does not provide enough space
to cover the whole range of the structures suggested. So we have to specialize our
study with particular structures illustrating different approaches to solution of the
flat display screen problem.
A flat field emission cathode for a display screen based on graphite field
emission cathodes at rather low, viz. 100 V, extraction voltages was implemented
468 8 Field Emission Cathode-Based Devices and Equipment
on the basis of a graphite powder [115]. The graphite powder with the particles of
about 10 lm in size was mixed up with an organic binder to a paste-like state. Then
the paste was mechanically pressed into the orifices of a glass plate.
The orifices were made in a photosensitive glass with the ultraviolet pho-
tolithography technology. The orifices were 0.6–1.5 mm apart for the glass of
*1.5 mm in thickness at the diameter of 200–250 lm at the outer and 100 lm at
the inner edges of the orifices. After the paste had dried, the structure was annealed
at the temperature close to that of softening to remove the organic binder. The
protruding graphite material was removed by mechanical abrasion and polishing. In
this case, the modulator was shaped by molybdenum deposition on the structure
obtained with subsequent explosive etching. That resulted in a self-consistent
alignment of the field emission cathodes and the modulator orifices. The distance
between the edges of the graphite cathode and the modulator was about 3–5 lm.
The element of a flat display based on that structure is shown in Fig. 8.45. To
reduce leakage currents between the cathode and the modulator, the glass is etched
to form a cavity of 3–5 lm in depth. The anode used had a conventional structure
of ITO layers and a luminophore. Then a conventional reverse-contact technology
was applied.
The modulator voltage was 20–50 V. The anode voltage was *1000–1500 V.
A specific feature of this structure is that field emission takes place from those
cathode angles where electric intensity is higher. That is why the light spots on the
anode are concentric rings. In addition, the beam divergence angle was 60–90°. To
avoid overlapping of beam couplings of the adjacent cathodes, the cathode-anode
distance was not longer than the distance between the cathodes.
Since it is quite easy to get powders from different graphite materials, fabrication
of flat display screen was further improved by technology simplification.
Paper [116] suggests reducing the space between the emitters. The structure is
fabricated by the process depicted in Fig. 8.46.
At the first stage, the apertures of 30–100 µm in diameter are made in photo-
sensitive glass by ultra-violet photolithography and etching in 10% hydrofluoric
acid, the glass plate being as thick as 250 µm. The space between the apertures can
vary from 100 µm to 1 mm depending on the purpose of the device.
At the second stage, the apertures/microchannels are filled up with standard
photoresist, and then the plate is polished to remove the excess photoresist from its
surface. At the third stage, the surface is sputtered with a thin molybdenum layer of
up to 3000 Å. At the fourth stage, removal of the photoresist forms a control
electrode. The apertures develop an insulating layer followed by filling up with the
nanotube paste. The insulating layer is necessary to insulate the cathode from the
control electrode. The insulating layer is developed by deposition from an appro-
priate solution. The process consists of repeated dipping, drying and annealing. The
Fig. 8.46 Diagram of process of production of triode structures with nanotubes [97]: 1–7 voltage
on control electrode
470 8 Field Emission Cathode-Based Devices and Equipment
glass-coating thickness can be controlled by the number of cycles; e.g., the layer
thickness of 4 µm needs 6 cycles.
At the sixth stage, the apertures are filled up with the carbon nanotube paste, and
the plate undergoes heat treatment and polishing down to the metal layer. And
finally, at the seventh stage, the plate is etched in 10% NaOH solution to form a gap
between the emitter and the control electrode. The uniformity of field emission over
the cathode area is improved by microdischarge aging.
The latter involves gradual increase in impulse voltage at the electrode under low
vacuum, viz. 10−5 mm Hg.
The aging was established to have increased the number of working field
emitters from 60 to 95%. The typical parameters of such a structure are given in
Fig. 8.47.
In addition to embrocation of the carbon paste, there are also a few examples of
its application in fabrication of field emission cathodes by printing, viz. [117, 118].
A modification of this method is pulverization of the binder solution [119].
However, this method has not come into wide-spread acceptance because of its
relative complexity and marked heterogeneity of field emission over the field
emission cathode surface.
Screens are also suggested to be fabricated by electrophoresis [120, 121]. This
method is based on an electrolytic bath consisting of 600 g nickel sulphide, about
5 g nickel chloride, 30 g sodium hypophosphite, about 40 g boric acid and about
1 g saccharin per liter of distilled water. This solution is mixed in with about 40 g
carbon nanotubes [121]. Electrophoresis goes on at the temperature of 50 °C.
It has to be noted that this flat-screen fabrication method has not come into
wide-spread acceptance, either.
What attracts implementers best of all is direct carbon nanotube growing.
Such growing is usually done on already prepared structures at the final stage
[122, 123].
Nanotubes are grown according to the following well-known technology,
namely by low-temperature CVD on a pre-sputtered catalyst layer [124]. The
pre-sputtering of the latest version is shown in Fig. 8.48 [125]. Substrate 1 is one by
one coated with the following components, viz. cathodic electrode 2 usually made
surface of the cathode aperture and the surface of control electrode 5 and dielectric
3. Moreover, the layer also blocks the generated ultraviolet rays that can cause
leakage current between the field emission cathode and the control electrode.
Short circuits between the field emission cathode elements, e.g., nanotubes, and
the control electrode are eliminated by increase in running voltage [133]. The effect
of this voltage generates an increased current through the defect, which eliminates
the latter; i.e. the nanotube is burnt off. The voltage is increased due to heating of
the ballast resistor, the latter reducing its resistance when heated.
Large-size displays are impossible without a large number of spacers between
anodic and cathodic plates [134].
However, the electron bombardment can charge the spacer surface, which
increases the noise of the device and even puts it out of operation. The disadvantage
is eliminated by mounting of some special electrodes [135] connected to the control
circuit on the lateral surface of the spacers, the latter being as high as 700–
3000 µm. They change the spacer potential in such a way as to liquidate its
charging by electron bombardment.
The display screens considered above had an emitter system well-defined in the
plane. That can cause serious technological difficulties. Paper [136] suggests
forming a pixel or an image element by a large number of chaotically located
cathode-modulator assemblies. The assembly is fabricated as diagrammed in
Fig. 8.50.
First metal layer 2, which serves as the contact of the field emission cathode, is
applied on substrate 1. It is usually made of metals, viz. Ni, Cu, Ag, Co, Fe.
Subsequently, metal layer 2 is sputtered with masking material 6, viz. the par-
ticles of glass, plastic, metal or ceramics. The particles of the masking-material can
be solid or liquid. That is followed by application of the insulating layer and the
second metal layer 4 intended to form a triode structure. Then the masking material
is removed 6. That can be done in water or a solvent with involvement of an
ultrasonic bath. The final process is application of the field emission cathode
material 5 by any known method. The result is chaotically arranged emitters 1
within one pixel 2 (Fig. 8.51).
8.2 Flat Display Screens 473
Naturally, introduction of every additional electrode into the screen structure pushes
up the cost. However, the resulting or desired improvement of image brilliance,
brightness and sharpness compensates for these efforts and costs over and above. In
parallel, most structures employ another electrode with orifices coaxial with the
control electrode. It is designed to focus the beam coupling because a wide angle of
divergence of field electrons emitted is well known. The modulator orifices usually
reiterate those of the control electrode and the field emission cathode [137, 138].
The best results can be obtained by focusing with a quadrode system [139]. It
includes three apertures formed by the structural components of the cathode plate.
The focusing electrode can be used not only for its intended purpose, i.e. focusing,
but also to protect the field emission cathode from discharges [140]. In this case, the
aperture of the focusing electrode is smaller than that of the control electrode. In
this case, it is not only focusing that is possible, but also interception of some
electrons by the focusing electrode.
However, it is totally unnecessary to place the control electrode near the field
emission cathode. This electrode can be represented by a focusing one [141]. In
particular, the impulse anode voltage of 1.5 kV, the earthed cathode, the
focusing-electrode voltage of −20 V and the control-electrode voltage of +80 V
produce the electron-beam diameter of 11 µ, the anode-cathode space being
1.1 mm.
The focusing electrode can also be located in the same plane with the control
electrode [142]. In this case, it is expedient to design all the electrodes as bars.
The idea of focusing and protection of the field emission cathode is most
completely implemented in more sophisticated structures, e.g., [143]. Figure 8.52
presents a circuit of such structures. Cathodic I and modulator II electrodes are
fabricated by conventional processing techniques. Thickness of the elements is as
follows: substrate 1 is 0.5–5 lm; cathode lead 3 is 0.1–1.0 µm; the dielectric layer
is 0.5–20 µm; modulator 5 is 0.1–1.2 µm.
A specific feature of this structure is rather sophisticated design of the focusing
electrode consisting of cone-hole metal gauze 6, the internal surface of the holes
being coated with thin dielectric layer 7. Structurally, the diameter of outlet D2 is
from one to three times larger than the total thickness of grid 6 and dielectric layer
7. That screens the field emission of cathode 2 from the effect of the anode
potential. In addition, dielectric layer 7 makes an additional focusing effect.
The difference between diameters D1 and D2 additionally densifies the electron
flow. The parameters of this field emission device are further improved by intro-
duction of a thin-film control transistor into each field emitter [144] (Fig. 8.53).
The display-screen element scheme shown in Fig. 8.53 is clear enough and does
not need any comments. Technicals are detailed in [144].
8.2 Flat Display Screens 475
Fig. 8.52 Schematic design of a display screen with a focusing electrode [124]: Structural
components: I cathode; II control electrode; III focusing system; IV anode. Design elements: 1
cathode substrate; 2 field emission cathode; 3 cathode lead; 4 dielectric layer; 5 modulator; 6 metal
gauze; 7 dielectric layer; 8 anode substrate; 9 conductor layer of the anode. D1 and D2 are
diameters of apertures of the modulator and the focusing electrode, respectively
Fig. 8.53 Diagram of the cathode part of a display screen with control of current of the emitter of
the thin-film transistor: 1 field emission cathode; 2 lead of a source; 3 source of the transistor; 4
control electrode; 5 active layer; 6 drain of the transistor; 7 base; 8 isolating layer; 9 electric drain
terminal; 10 internal isolating layer; 11 dielectric layer; 12 metal gauze
and control electrode 9 are placed on the plane of glass support 1. Their leads 7 and
10 are located on the same plane. The end faces of electrode leads 7, 10 are coated
with resistive materials 11, 12.
In this case, the specific resistance lies within 10−7–10−4 X m. Electric resis-
tance is improved by introduction of some doping substances, e.g., B, P or poly-
tetrafluoroethylene (PTFE), into metal coating (11, 12). The Ni–B–P system can be
based on Ni or P, the concentration ranging from 3 to 40% or from 7 to 40%,
respectively. If doping with Ni involves PTFE, the PTFE concentration is 0.1–30%.
In itself, field emission cathode 13 is carbon nanotubes fixed in metal layer 11.
Control electrode 9 is almost the same as field emission cathode 8 in terms of
structure. That makes it possible to swap the cathode and the control electrode by
simple switching when the field emission cathode becomes less active.
As shown before, lateral designs are characterized by reduction of the
cathode-control electrode gap. Its value can range from 500 µm [147] to µm units
[146, 148]. It is very useful to combine different control circuits. That complicates
the device design, but can seriously improve the image quality. E.g., paper [149]
suggests a display design combining the lateral and the classical method of control
of the emitted electron beam (Fig. 8.55).
Field emission cathodes 2 consisting of carbon nanotubes rest on metal lead-out
bars 1. Control electrodes 6 lie in immediate proximity from the field emission
cathodes and are electrically connected to the external terminals through bus 4. The
electron beams are additionally focused by electrode 5, which works in a con-
ventional way (Sect. 8.2.3). The anode of this design has two elements improving
the image quality. It is, firstly, black mask 14 improving the image contrast; and
secondly, thin aluminum layer 10, which increases the image brightness. The image
8.2 Flat Display Screens 477
Fig. 8.55 The schematic design of a display with a combined focusing system: 1 cathode lead; 2
field emission cathode; 3 bottom (cathode) base; 4 control electrode lead; 5 focusing electrode; 6
control electrode; 7, 8, 9 layers of dielectrics; 10 metal layer (e.g. Al); 11 anode; 12 anode base; 13
phosphor; 14 black matrix
quality can also be improved if the image element is formed with involvement of
either a few field emission cathodes or a few control electrodes [150].
The position of the control electrode behind the field emission cathode under an
insulator layer sets this structure apart from others. First of all, it excludes a dis-
charge of electricity between the field emission cathode and the control electrode
through the vacuum gap.
One of the first structures of such a screen was presented in papers [151–153].
Figure 8.56a shows the arrangement drawing of the cathode part of the screen,
while Fig. 8.56b demonstrates the distribution of electric potential and the electron
path of such a cathode-modulator assembly.
Glass substrate 1 is coated with an aluminum layer of 1500 Å in thickness, and
then a standard photolithographic process produces control electrodes 2 of 400 lm
in width. Then insulator layer 3 comes. In this case, it is a polyimide layer of 13 lm
in thickness. The latter is coated with a cathode layer of aluminum 4 by
electron-beam vaporization, and the cathode layer of aluminum then develops into
cathode bars of 390 lm in width. The cathode bars are smeared with a paste with
single-walled carbon nanotubes through a metal grid with a mesh of 20 lm.
Subsequent thermal treatment at 350 °C enables the nanotubes to project above the
surface of the field emission cathode coating due to annealing of the binder.
Figure 8.56b shows the estimated data of electric potential and electron paths for
different voltages at the control electrode. The picture shows that the electric field
concentrates in the corners of the cathodic electrode. That is why electrons emit
478 8 Field Emission Cathode-Based Devices and Equipment
mostly from the corners of the cathode rather than its central part. This effect can be
minimized by reduction of width of the cathode assemblies [154].
The problem of edge effects can be solved by application of rounded field
emission cathodes [155]. The arrangement of a display based on such field emission
cathodes is shown in Fig. 8.57.
Glass substrate 1 bears control electrodes 5 represented by ITO buses. Dielectric
layer 2 coats the entire surface of the cathode plate. The dielectric bears leads 3 and
field emission cathode 4 laid at right angle to control electrodes. The cathode leads
are made of two layers. The first layer 3a is made of aluminum to provide a good
electric contact and protect the dielectric from cracking. Second layer 3b is made of
chromium to protect the first layer from destruction by possible electrical break-
down and improve reliability of the cathode lead. Field emission cathode 4 is made
of carbon nanotube containing paste and applied by screen printing. Anode 6–9 is
fabricated in a conventional way.
If further focusing is necessary, a special electrode reducing the field-electron
spreading angle can be applied to the dielectric plane [156].
Thin-film and blade field emission cathodes have an advantage over pointed ones
(Sect. 5.3). That is why implementers wish to design a display screen based on
them. Figure 8.58 presents the design of one screen element based on edge field
emission cathodes [157]. The basic technological stages of such a screen are
– formation of a thin layer of control electrode 2 (Nb);
– application of a photoresist, distribution of control electrodes, etching; further
layers are applied in nearly the same sequence, the dielectric layer consisting of
SiO2 and the field emission cathode being a metal film of Nb or Mo.
This design is simpler in fabrication than that based on pointed emitters. Its
emission characteristics display quite good field emission parameters (Fig. 8.59).
This technology can be assumed as the basis for home television screens of the
color pixel ranging from 0.6 to 1.0 mm, the screen ranging from 20 to 30 in.
The display-screen element design based on thin-film carbon field emission
cathode is shown in Fig. 8.60 [158]. Cathode-assembly glass substrate 1 bears the
layers of cathode leads 2 of molybdenum (Mo). Then the cathode leads are coated
with resistive layer 3 of a–Si, which levels the field emission characteristics of
particular cathodes. The subsequent layer is a-carbon, its end face functioning as a
field emitter. The cathode-anode space being 20–30 µm, field emission can be
initiated with about 600 V. This design can be further developed by introduction of
Fig. 8.58 Design of an element of a flat display screen with a blade field emission cathode
(cathode part) [138]: 1 glass substrate; 2 control electrode; 3 layer of dielectric; 4 field emission
cathode film; 5 passivated layer; 6 field emission cathode
480 8 Field Emission Cathode-Based Devices and Equipment
a control electrode below the level of the field emission cathode [158]. An inter-
esting combination of a thin-film field emission cathode with a reverse anode is
suggested in [159, 160]. The circuit diagram of the device and the pixel structure
are given in Fig. 8.61.
The cathodoluminescent screen has a vacuum shell made of two flat glass faces:
transparent plate and a substrate with electrodes, a flat, thin end-face emitter and an
anode coated with a conducting layer and a luminophore forming a light-emitting
cell, viz. a pixel, located at a small controllable distance from each other in the
parallel planes connected through anode buses and emitter buses with terminals
intended to connect to electric-voltage sources, a dielectric frame hermetically
connecting the face plate and the substrate. In such a screen, the anodes and the
emitters are shaped as two combs enclosed into each another, the emitters with
connectors being located in the upper plane of the substrate, the anodes and the
anode buses being located at the bottom of the grooves of predetermined depth in
the substrate from the side of the upper plane, the free space of the anode-bus
grooves being filled up with a dielectric to form a floor-space area hosting con-
nection emitter buses.
8.2 Flat Display Screens 481
Fig. 8.61 Schematic diagram of the device (a) and structure of the pixel (b): 1 glass base; 2
anode; 3 phosphor; 4 emission surface; 5 cathode lead; 6 anode bus; 7 protective resistor; 8
cathode bus
In addition, the anodes and the emitters can have a form of any geometric figures
enclosed into each other, e.g., in a form of spirals; a cathodoluminescent screen can
be additionally supplied with a protective resistor connecting the emitter bus with
each pixel emitter in series; the selected luminophore has sub-micron scale grains;
the selected low-voltage luminophore has the running voltage of no more than
50–70 V.
To this end in view, the well-known cathodoluminescent screen fabrication
method, including consecutive formation of the anode bus and the emitter bus, the
active pixel structure, installation of the intermediate dielectric stringer, hermetic
sealing of the device along the stringer perimeter, is modified by formation due to
grooving of the substrate by wet etching, coating of the groove bottom with a
delivery layer and filling up of the free space of the grooves with a dielectric, e.g.,
fusible glass, to form a planar structure. Then the entire surface is coated with a
resistive layer followed by formation of the protective-resistor topology by pho-
tolithography. Then the entire surface is coated with emitting and conducting
layers, the latter being used to form emitter buses orthogonal to the former anode
ones, by photolithography. Subsequent photolithography forms an active structure
of anodes and emitters in a pixel by etching in the windows of the conducting layer,
the emitting layer and the glass cavity at the predetermined depth, after which beam
sputtering deposits a conducting layer contacting with the anode bus. Then the
photoresist is removed together with the conducting layer sputtered upon it, and a
layer of low-voltage fine-grain luminophore is applied.
Application of high-precision glass-grooving technology, involvement of
high-precision fine-grain luminophores and consequent small gaps between the
emitter end-face and the anode makes it possible to use low-voltage luminophores
with electron energy below 100 V, thereby providing a long cycle of the device. If
the luminophores used are low-voltage, there is no need in triode structures. Diode
structures at low voltages provide a high resolving ability. Fabrication of
482 8 Field Emission Cathode-Based Devices and Equipment
X
1 X
1
2pnx jbmz
Vðx; zÞ ¼ anm cos e ; ð8:1Þ
n¼0 m¼1
d
XN XM
1
1 ¼ 0; ð8:5Þ
b
n¼0 m¼M nm
where
" 2 #
den 2pn
bnm ¼ b2 þ þ ZY
2ZYs m d
" #
d 2 en 2pm 2 2pm 2
¼ jc þ þ þ ZY : ð8:6Þ
2ZYs d d
The numerical solution of this equation gives transmission factor c for some
frequency and the space between the emitters. If the space between the emitters is
much less than the wave length, term 1/B in (8.4) is much more than the others. So
B 1. As a result, the transmission factor is
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c¼ Z ½Y þ ðYs =d 2 Þ ¼ ðR þ j-LÞ j-C þ Ys -=d 2 : ð8:7Þ
8.3 Microwave Devices 485
Comparison between (8.2) and (8.7) shows that the load connected with the
emitters simply increases the average conductance of the transmitting line per unit
length. If the space between the emitters is not substantially less than the wave
length, the quotient approaching 1 THz, (8.5) and (8.6) are used.
The equivalent shunt admittance caused by conical emitters can be considered as
infinite multiple-link distribution of the resistors shunted by capacities (Fig. 8.62b).
The differential elements of the circuit connected with the differential increase in the
emitter height can be calculated by the finite-difference method.
The capacitance structure is made up of three components (Fig. 8.62c), viz.
ring-type capacitor Cer, cone capacitor Cec and stud capacitor Cep. The values of
these capacitors, their efficient arrangement within the structure can be calculated
by analytical approximations or from the experimental measurements of the
capacitance in scale models. The resistance and inductivity of the cylindrical or
conical part between the efficient positions of the capacitors can be calculated
analytically on the basis of the skin layer in the emitter material. Having found out
these values, one can calculate shunt admittance as
1 1
¼ Zep þ ; ð8:8Þ
Ys Yep þ Zec þ
1
1
Yer
Yec þ
1 þ Zer Yer
where a = 2b/d and b is the radius of the aperture in the control electrode.
In addition, the apertures in the control electrode increase the inductivity per unit
length:
2t3 d1 d2
L ¼ lo þ l1 þ l2 : ð8:10Þ
x ð1 þ f - Þ 2f- - 2-
provides reasonable approximation of this effect. Secondly, the emitters increase the
capacitance due to complex shunt admittance Ys. Addition of these two effects
result in
- - Cs
C ¼ Cpp þ Im ð Y s Þ jRe ð Y s ÞCpp þ C s j ; ð8:12Þ
- d2 d2 -
where Ys ¼ Gs þ j-Cs :
The emitter losses caused by the distributed capacitance become complex.
The initial estimates of the field emission cathode arrays at 10 GHz are given
below.
A good example is Spindt field emission cathode-based systems, which consist of
molybdenum cones of 0.3 µm in height, a SiO2 dielectric of 1 µm in thickness and a
control electrode represented by an M0 film as thick as 0.3 µm. In addition, let us
take account of the aperture radius of 0.4 µm, the basic rounding-off radius of
0.45 µm and the cone height of 1.3 µm, which provide that the apex peak is above
the control electrode surface. In this case, the parasitic shunt capacitance of this
structure is analytically estimated as Cer = 4.8 10−17 F and Cec = 1.8 10−17 F,
while the resistance is Rer = 0.01 X and Rec = 0.08 X. Parasitic inductance is about
0.2 GHz and negligible at 10 GHz.
Such as amplifier is diagrammed in Fig. 8.64. The value of the microwave field
and the voltage of the high-voltage source are selected so that the total electric field
intensity is sufficient to excite field emission from the cathode. The field emission
cathode has a ring configuration consisting of 3 concentric rings made of pointed
field emission cathodes, the comb rings being attached to the concentric
piezocrystals connected to the rectangular-pulse generator 8. It controls the position
of the field emission cathodes in the resonator, which contributes to increase in the
average power of the amplifier.
The amplifier works as follows. Low-power microwave generator 2 in input
resonator 1 modulates field emission current from cathode 7 that goes to collector 4.
The amplified signal in the output resonator 3 goes to load 5.
488 8 Field Emission Cathode-Based Devices and Equipment
8.3.2 Magnetrons
filament-supply sources are also comparable or even exceeding the weight and
dimensions of high-voltage sources. Reduction of these parameters is especially
important for independent and portable radio transmitters. Application of field
emission cathodes removes all the heat problems connected with power-supply
sources.
The first magnetron structures were disk edge field emission cathodes. However,
more up-to-date magnetron structures employ disk film field emission cathodes
[168].
The magnetron is diagrammed in Fig. 8.69. Guide core 1 hosts components with
secondary electronic emission (secondary electronic emitters 2) represented by
tungsten-based impregnated materials added with barium, or alloys and compounds
based on platinum-group metals, such as platinum, palladium, iridium, osmium,
having addition agents of a material of alkaline-earth, alkaline or rare-earth (Ba, Ir,
Ca, La) elements. Between emitters 2, there are field emitters 3 projecting above
them by 5–20% interelectrode gap. They are shaped as beads consisting of a
conductive film and dielectric film 4 surrounding it from the sides.
The dielectric film protects the lateral surface of the field emitter not only
mechanically, but also electrically, viz. when treated with positive ions, it is charged
with homogeneous potential that generates a field repelling the inbound ions.
The dielectric film can be away from the lateral emitter surface by the distance
equal to the conductive film end-face radius. According to computations, that
provides additional increase in the resistance to ion-charge drain when the lateral
emitter surface is totally protected against the direct hit of residual-gas ions. If the
gap exceeds the end-face radius of the conductive film, its lateral surface can be
8.3 Microwave Devices 491
bombarded. If the gap is smaller than the radius, the dielectric film provides total
geometrical and electrical screening of the emitter surface.
Resistance can be increased by elongation of the ion-charge drain path due to
dielectric-film crimping. The computations made by mathematical simulation show
that the ion flow to the field emitter can be totally excluded if the dielectric films
project over the conductive-film surface by the value equal to the radius of the film
end-face.
Increase in the projecting part of the dielectric film by the value exceeding the
radius of the conductive-film end face reduces the field intensity and current at the
emitter.
Film 4 can be made of boride-, nitride-, oxide-type dielectrics, e.g., Si3N4, SiO2,
BN, AlN, Al2O3; or the dielectrics with alkaline and alkaline-earth additives (Ba, K,
Cs, Li, Na). Base 1 with secondary-electron and field emitters is located in cylin-
drical anode 5 with a working clearance, dielectric film 4 of the projecting part of
the field emitter being as far away from the lateral emitter surface as the radius of
the emitting end-face of the conductive film.
Film 4 can project over the field-emitter surface. The number of the triggering
field emitters in a device is determined by the value of the necessary starting
current, i.e. the device parameters. Field emission is provided by a constant field
generated by the anode near the cathode. Electrons collide with residual-gas atoms
in the cathode-anode space to form ions. Positively charged ions accelerate them-
selves to the field emitter along the equipotential power lines and move towards the
lateral field-emitter surfaces, thereby bombarding the dielectric protecting films.
The ions impact the dielectric film, thereby generating a charge equal to the
potential of the falling ions on the surface. As the resulting charge is of the same
492 8 Field Emission Cathode-Based Devices and Equipment
potential as the falling ions, it bends the ion path, repelling the ions from the
field-emitter surface.
The main advantage of the device is the extended life cycle of the cathode.
Fig. 8.71 Schematic diagram of the electronic part of a travelling-wave lamp with electron beam
focusing: 1 substrate; 2 cathodic electrode; 3 field emission cathode from made of diamond
powder; 4 insulating layers; 5 control electrodes; 6 electron beam; 7 anode
The principle circuit of the field emission cathode-based reflex klystron is dia-
grammed in Fig. 8.72 [171].
The operating principle of this device is based on modulation of the velocity of
the electrons emitted by the electron source 1. After emission, the electrons pass a
pair of metal grids forming resonator 2. Having left the second grid, the electrons
enter the drift region, where they are reflected back along the initial paths. That is
what negatively charged repulsive electrodes do.
Electron-current fluctuations amplify the oscillations of the electromagnetic field
in the resonator, which results in fluctuations in the potential difference between
grids 4. These changes in the potential slow the fast electrons and accelerate the
slow ones when they pass through the resonator, causing formation of electron
beams in the drift zone. If the reflected electron beams come back to the resonator at
494 8 Field Emission Cathode-Based Devices and Equipment
the appropriate potential, i.e. when the grid potential is positive, the power is
transmitted to the resonator and the beaming effect goes up. There is feedback under
the circumstances, and the resonance frequency of the resonator causes the variable
power to increase until the state becomes steady, viz. when increase in the gener-
ation power becomes equal to the loss power. Naturally, it is necessary for the
power transmitted from the electron beam to the resonator to be enough to com-
pensate for the losses in the resonator and to transmit the useful power to the
external load. According to such calculations, the devices can used up to
sub-millimeter wavelengths.
X-ray tubes, requiring high voltage to function, are convenient objects to use field
emission cathodes. Such attempts have been made from long ago [172], the field
emission cathode being the segments of a differently-configured graphite cloth.
Application of pointed molybdenum field emission cathodes in an
ultrahigh-vacuum structure (1010 mm Hg) of a reflective anode X-ray tube shows
the prospects of such field emission cathodes [173]. The cathodic part is a
50,000-apex field emission cathode array of 1 mm in diameter [174].
The voltage being 25 kV and the beam current being 25 mA, the X-ray spot
diameter is 0.3 mm. More than 90% emission intensity is accounted for by the
diameter of 0.2 mm, electrostatic lenses being geometrically adjustable to reduce
the spot diameter to less than 0.1 mm.
The X-ray tube resolution depends on the X-ray source dimensions. That is why
it is very important to design microfocus X-ray tubes. Modern materials for field
cathodes are based on various carbon nanostructures [175, 176]. Nowadays the
most widespread of such materials for X-ray tube field cathodes are carbon nan-
otubes. Figure 8.73 diagrams an X-ray tube based on a field emission cathode made
of a multiple-wall nanotube bundle [177]. The multiple-wall nanotube bundle 2 is
stuck down to the end face of the tungsten wire 1 with a conductor paste. The
potentials at cathode Vc and copper anode 4 Va are −6.5 kV and +6.5 kV,
respectively. So the accelerating voltage is 13 kV. Electrical lens 3 of three parallel
diaphragms is put between the cathode and the anode.
The first diaphragm of the lens is also an extraction electrode. Since the distance
from the end of the field emission cathode to the second electrode is 13 mm and to
the anode 26 mm, increase in the electron-optical system is equal to one. So the
focal dimension of the electron beam at the anode is equal to the dimension of the
nanotube based field emission cathode, its diameter being a few tens of nanometers.
The X-rays of the focused spot are withdrawn through beryl window 5 to the
atmosphere. Then the emission passes sample 6 and records itself on film 7, the
field emission current being 440 lA.
8.4 X-ray Tubes 495
Fig. 8.73 Schematic diagram of a microfocal X-ray tube with a field emission cathode made of a
bundle of multiwall nanotubes: 1 tungsten wire; 2 field emission cathode made of multiwall carbon
nanotubes; 3 electrostatic lens; 4 copper anode; 5 window made of Be; 6 sample; 7 film. Power
supplies of: Vc field emission cathode; Vy extraction electrode; Va anode; Vl control lens
To increase the power of the X-ray emitter, one has to increase the emission
current mostly by increase in the area of the field emission cathode. In this case, it is
necessary to use the electrostatic lenses of a more sophisticated design [178].
An X-ray bundle scanning usually involves an anode- or cathode-positioning
mechanism. That considerably complicates the design and increases its dimensions.
To remove these disadvantages, paper [179] suggests using several cathode
assemblies and the respective number of electrical lenses. Paper [179] describes 5
cathode assemblies arranged in a line. The field emission cathodes are controlled by
MOS transistors.
The simple structure of the X-ray tube field emission cathode assembly based on
a carbon nanotube field emission cathode is shown in Fig. 8.74 [180].
Carbon nanotubes 2 are applied on 1-mm thick stainless-steel substrate 1. The
layer of nanotubes is 70 lm thick. Spacer 3 separating the field emission cathode
and control electrode 4 is Micatex of 56 µm in thickness. The functions of a control
electrode are performed by a molybdenum microgrid of 100 lines/in. in trans-
parency and 25 µm in thickness with an orthogonal structure. Focusing electrode 5
is a steel plate of 2 mm in thickness with an aperture of 3 mm in diameter.
496 8 Field Emission Cathode-Based Devices and Equipment
The threshold electric field intensity for this cathode is 3 V/µm, and the
emission-current density is about 10 µA/cm2, the emitter area being 1 mm. In this
electron source, the factor of electron-beam transmission through the focusing
electrode is 0.5–0.8, depending on the spacer thickness and the anode voltage. For
the X-ray tube of this design to function, it is enough to have the total power of
about 0.1–0.4 W.
The development of field cathodes based on carbon nanotubes (CNT) led to
multiple inventions in construction and manufacture of miniature X-ray tubes with
CNT-based field cathodes, e.g. [181–186].
Nowadays we can see intensive development in the field of tiny through-target
X-ray tubes. A version of such a tube with a field emission cathode made of a
carbon-fiber bunch is suggested in [187–189] (Fig. 8.75).
The X-ray tube (Fig. 8.75) consists of vacuum envelope 1 containing electron
gun 2, anode 3 shaped as a chamber in this case, and X-ray ejection window 4. As
for electron gun 2, it contains field emission cathode 5, cap 6 with electrical lead 7,
bead 8, field emission cathode connector assembly 9.
Appearance and scheme of the cathode-modulator node of a miniature X-ray
tube is presented on Fig. 8.76.
Carbon-fiber bunch 10 is put into shell 11 to maintain orientation of carbon
fibers, their mechanical mounting and vibration strength. Shell 11 is made of a
conducting or semiconducting material, e.g. a conductor, semiconducting glass or a
metal-coated dielectric, i.e. a material excluding accidental potential shift of the
shell against the carbon-fiber bunch when the X-ray tube is in operation.
Fig. 8.76 Appearance (a) and schematic diagram of the cathodic modulator unit of a miniature
X-ray tube: 1 modulator; 2 cathode lead; 3 slug; 4 lamp stem; 5 carbon fibers; 6 glass insulator
Carbon-fiber bunch 10 projects above shell 11 on the emitter side. For field elec-
trons to transit from the cathode to the anode, the end wall of cap 6 functioning as a
control electrode of the tube has a diaphragm represented by orifice 12.
Contact field emission cathode assembly 9 is made of conducting material 13;
e.g., a conducting paste based on the powder of Ag, Al or other metals or con-
ducting materials. Above-mentioned material 13 is applied on the carbon-fiber
bundle end opposite the emitter and a part of the shell pressed out by metal ring 14,
which can be shaped as a cylinder or a cylinder cut along the plane passing through
its axis. Ring 14 has electrical lead 15 (cathode) from vacuum envelope 1 and is
rigidly connected with dielectric bead 8 with, e.g., glass-fiber reinforced cement,
glue or amber, the central axes of ring 14, carbon-fiber bundle 10 wrapped into shell
11 and the aperture of bead 8 being coaxial with the electron-optical system of the
X-ray tube.
The tube is assembled so that the internal walls of the aperture in bead 8, shell 11
and the upper end face of contact assembly 9 form cavity 16. The cavity is nec-
essary to prevent generation of surface conduction between cathode 5 and control
electrode 6 caused by formation semiconductive and conductive films and paths on
the shells of cathode 11, bead 8 and control electrode 6. The conductive and
semiconductive films emerge from thermovacuum treatment of the X-ray tube, as
well as evaporation of the cathode material exposed to ion bombardment and
breakdowns between the X-ray tube electrodes when aging and operating the X-ray
tube.
To prevent development of these films, cavity 16 can be supplemented with long
or traverse electrically-protective slots in bead 8 to increase the surface electrical
498 8 Field Emission Cathode-Based Devices and Equipment
The tests of the tiny X-ray tube claimed confirm the effectiveness of its structural
solutions. At the supply voltage of 40 kV and the current of 100 lA, the tube has
worked for more than 1000 h without failures, which proves its high reliability and
electrical strength. At the same time, the instability of characteristics of the X-ray
tube has been less than 1%.
A promising approach to manufacture with a shoot-through anode is proposed in
[190]. In this construction of X-ray tube with field cathode the external (incl.
magnetic) electrodes are used for control and focusing of the beam. This allows to
make the technology of manufacturing of X-ray tubes simpler and cheaper.
A scheme of the tube is presented on Fig. 8.77. Its function principle is obvious
from the schematic.
A more detailed review of X-ray tubes with low-strength field emission cathodes
is presented in [191].
Electronic devices mostly employ a narrow directed electron bunch, i.e. an electron
beam. Most devices employ a round-section beam, but sometimes shaped beams,
i.e. ribbon beams, oval-section beams and otherwise shaped beams are applied.
One of the basic units of an electron-beam tube is the device generating an
electron beam, i.e. a focused electron bunch with energy and current density nec-
essary for the tube. Such a device is referred to as an electron gun.
The requirements made to an electron gun can vary over a wide range depending
on the type of the device. E.g. beam electron energy can amount to several kilo-
electronvolt in small oscillograph tubes and some tens or even hundreds of kilo-
electronvolts in electronic microscopes.
In the same way, the beam current of some image pick-up tubes is several deciles
of a microampere, while it can reach several milliamperes in lithographs. However
wide the range of variation of energy and current might be, the beam section in a
particular plane, usually in the receiver plane, i.e. a screen or a target, should be as
small as possible. The beam section is usually estimated by the spot diameter, a spot
being understood as a trace of an electron beam on the receiver surface. In many
types of contemporary devices, the spot diameter is as short as several deciles of a
millimeter or less.
Most devices imply an ability to control the beam current value in a wide range,
from zero, i.e. gun blocking, to the maximum value depending on the purpose of
the device.
The designers of electron-beam devices should also take constructive reasons
into consideration the electron gun should be well agreed with the entire device; the
technology of its fabrication should not be too difficult; the materials of the electron
gun designed should meet the conditions of operation in vacuum. Moreover, the
gun should be quite economic in operation and long-lasting, as its failure can
destroy a sophisticated and usually expensive device.
The electrons emitted by the cathode are accelerated by the electric field gen-
erated in the cathode region by gun electrodes. As the gun should generate not only
accelerated, but also focused beam-forming electron flow, it is expedient to select
the configuration of the accelerating field so that acceleration went along with
electron flow focusing. For this purpose, the cathode is equipped with an acceler-
ating electronic lens. As the accelerating field should reach the cathode surface, the
cathode should be dipped into the lens field.
It has already been mentioned that most guns can control the beam current.
8.5 Electron Guns 501
There are a lot of applications needing localization of the electron beam in a small
solid angle; e.g., raster-type electron microscopes, X-ray microanalyzers, some
types of electron-beam tubes, microwave devices etc.
As a spike field emission cathode is spherically symmetrical, the emission cone
opening angle is very large, viz. about 60° for emission at an average current
density of 104–105 A/cm2 and about 100° for high densities (107–108 A/cm2).
Nowadays, there are several beam emission localization methods. The angular
limit can be achieved by cutting the most intensively emitting region of the apex
surface with a diaphragm and increase in its emission abilities through restructuring
[192]. The restructuring is carried out by expansion of certain crystallographic
planes by application of back voltage.
A lot of papers have been dedicated to research of adsorption of different
stranger atoms on the surface of tungsten or other materials, which are selectively
adsorbed on certain crystallographic planes, thus reducing the work function in the
502 8 Field Emission Cathode-Based Devices and Equipment
The alloy can result from zone melting by an electron beam in two passes at the
velocity of 2–4 mm/min. along crystal axis 001. The samples of 0.3 0.3 mm in
cross-section are cut out along this axis by the electric-spark method.
The apex is etched in the solution of 1.5 N NaOH at the voltage of 5 V and the
current of 20 mA.
The field emitter is mounted in a field emission microscope and exhausted down
to the pressure of 10−7 Pa, which is the working vacuum in electron-probe devices.
The field emitter is activated by preheating at 1460 K for 3 min. in the field of
0.2 V/Å, which makes the atoms migrate from the matrix to the apex surface and
form a stable aggregation on facet (001).
An apex with a small solid angle of an electron beam for heater field emission
cathodes is usually produced by thermal field processing [198]. In this case, the
basis is a tungsten field emission cathode with the rounding-off radius of about
2 µm and the axis oriented towards [98]. At first, the field emission cathode is
cleaned by heating up to 2300 K. Then the temperature is reduced to 1900 K and
high voltage is supplied and stepped up until the emission current reaches 1–
100 lA, depending on the emitter rounding-off radius. At the first moment, the
cathode emits perpendicularly to the facet [310] as shown in Fig. 5.1b. Growth of
the voltage triggers the restructuring of the emitter (Fig. 5.1c), which is subse-
quently reshaped so that the entire emission takes place only from the facet [98]
(Fig. 5.1d). The emitter restructurization is accompanied by stepping up of the
emission current. At this stage of processing of the cathode, emission is stabilized at
constant values of temperature and voltage. The field emission cathodes produced
in this way are capable of a long service life at T 1200–1700 K with current
takeoff of 1–300 µA at the residual-gas pressure p of 10−6–10−3 Pa. Pressure
reduction by an order of magnitude down to p 10−7 Pa makes it possible to step
the temperature of the field emission cathode down by 570 K. A face of W(310) is
still considered a classic field emission cathode for high-dimension microscopy, as
evidenced by the latest studies in this field [199].
The limiting current value that can be taken off from the field emission cathode is
limited to heating of the emitter with Joule heat and the Nottingham effect and
ultimately determined by the heat-conducting path conditions [200], the apex
geometry [201] and the operating mode [202]. In practice, the volt-ampere char-
acteristic biases towards current reduction, which limits the emission-localization
possibilities. Paper [203] suggests an explanation of this phenomenon based on the
difference between the true shape of the potential barrier and the classical
image-force law as adopted to derive the barrier-transparency expression.
Low-power electron guns are very widely used in electron probe devices, such as
focused-beam microscopes and transmission-electron microscopes [204, 205].
504 8 Field Emission Cathode-Based Devices and Equipment
To present day there has appeared a large number of focused-beam and trans-
mission microscope constructions that employ field emission cathodes for
increasing the image resolution. Some of those designs of previous years are still of
interest at this time.
The first electron gun based on a field emission cathode was developed by
Crewe [206] and calculated by Butler [207] (Fig. 8.78).
Voltage U1 applied between the apex and the first anode determines the total
emission current. Voltage U0 stipulates the energy of electrons. The electrical lens
developing between the anodes fixes the electron beam and forms a real image of
the field emitter in the plane of the sample. This structure was further developed in
[208].
The camera–trap inlet is 1–2 mm away from the field emitter point. It is supplied
with the voltage of 5–8 kV against the emitter. As a result, the gun works with
development of a false image of a fictional electron source. Paper [209] suggests
converting the fictional-source false image generated by the electron gun into a real
one, using an additional magnetic condensing lens for this purpose.
The paper [210] features the lens system of a field emission gun consisting of
three anodes and a magnetic lens. The system is more efficient in terms of current
than the accelerating-lens system of two anodes. Paper [211] features the optical
parameters of such a system in detail.
Magnetic-lens electron guns are featured in [212, 213]. The electron gun consists
of an apex, a reference electrode and an anode put in a magnetic lens, the focal
distance being a few millimeters. Such a system forms a real image beyond the pole
shoes. The reference electrode can control the field at the apex, the emission
conditions being maintained constant in the range of anode voltages from 2 to
10 kV. The gun structure shown in Fig. 8.79, is a diode system where the anode is
inside the magnetic-lens pole shoes. The lens is optimized in respect of the cathode
and forms a real image [213].
Paper [214] shows that the emission rate of a field emission gun is 3–4 orders of
magnitude higher than that of a thermal-emission gun with other things being equal.
Great difficulties in designing field emission cathodes able to work in high
technical vacuum have given rise to preheated field emission cathodes that can
provide thermal dependence or clean the surface of a field emission cathode.
Paper [215] suggests cathode assemblies of electron gun-based transmission
microscopes [216].
Field emission cathode-based electron guns working in ultrahigh vacuum of
10−8–10−7 Pa have a problem of supplementary degassing of accelerating-electrode
surfaces. The reason is that emission from a field emission cathode brings some
emitted electrons to the surface of these electrodes and desorbs the residual-gas
molecules on their surface, thereby increasing the pressure near the apex, which
degrades the electron-beam parameters and reduces the cathode life cycle. It is
suggested to eliminate this disadvantage with an electron gun able to degas the
accelerating electrons [217] (Fig. 8.80).
Degassing is performed by auxiliary ring-shaped thermionic emitter 5 sym-
metrical in respect of field emission cathode 3. First anode 4 is also symmetrical to
it. For second anode 6 to be degassed, the first anode is circumferentially punched
to form apertures 9.
506 8 Field Emission Cathode-Based Devices and Equipment
During the process of degassing, the current passing through auxiliary cathode 5
causes emission of electrons, the accelerating potential between cathode 5 and
anode 4 making the electrons bombard first anode 4. Some emitted electrons pass
through apertures 9 and bombard second anode 6 supplied with a higher positive
potential than the first anode.
To prevent destruction of apex 3, the latter is supplied with a potential equal to
that of auxiliary cathode 5. The symmetrical shape of the electrodes minimizes
perturbation to the emitted electrons.
If the anode surface bombarded by the cathode-emitted electrons is fabricated or
coated with a getter material, e.g., titanium, zirconium, thorium or barium; the
electron beam becomes even more stable and the current fluctuation diminishes
[218]. Before the electron gun starts functioning, the vacuum chamber housing it is
pumped out to the ultrahigh vacuum of 10−8–10−7 Pa. After that, the anode is
heated up to the temperature of 1070 K along with continuous pumping out,
evaporating the residual-gas molecules adsorbed on the anode surface, degassing
the anode surface and activating the getter. After the degassing, very few molecules
are desorbed from the anode surface due to electron bombardment from the field
emission cathode because these molecules are connected by the getter. Moreover,
the residual-gas molecules are forcefully absorbed by the activated getter. As a
result, the cathode, protected from ion bombardment, produces a stable electron
beam.
A two-chamber electron microscope [219] makes it possible to maintain different
pressures in the cathodic and sample compartments and replace the parts of one of
them without breaking vacuum in the other. The two-chamber electron microscope
is diagrammed in Fig. 8.81. Field emission cathode 1 is in compartment 3 of tubular
case 2. High-vacuum pump 4 mounted in the cathodic compartment maintains high
8.5 Electron Guns 507
Bombardment of the anode-surface by the emitted electrons not used in the main
beam has been calculated to cause formation of secondary electrons with the energy
ranging from 5 to 50 eV, most of them having the energy of 7–15 eV. After the
passage, the electrons have the energy of e (U3–U1). As the potential difference in
the electron microscope is high, the object should be considered as lit up with the
main beam and the diffused electron beam, which certainly complicates the
research. The secondary-electron suppressor gun is diagrammed in Fig. 8.82.
Emission from field emission cathode 1 is provided by high-voltage source U1, its
voltage being supplied between cathode 1 and anode 2. Emission of electrons
generates two electron beams, viz. main beam 4 and secondary-electron beam 3.
However, the secondary electrons generated on the surface of anode 2 do not reach
second electrode 5 because the voltage negative in respect of anode 2 is supplied
between anode 2 and electrode 5. The voltage generates a field pushing secondary
electrons 3 off from electrode 5, so they do not get to the accelerating field region
between electrodes 5 and 6. Voltage U2 is about 20 V, so most secondary electrons
of 7–15 eV are retained. It is worth mentioning that voltage U2 should be selected
so that it did not essentially affect the primary-beam electrons. Accordingly, it is
necessary for the voltage to be less than the accelerating voltage of U3 by two
orders of magnitude. There are a lot of devices intended to stabilize the current of
field emission guns, e.g., [221].
Paper [222] suggests a design of a field emission stabilization gun.
Figure 8.83a shows the dynamics of the electron-beam current for a conven-
tional electron gun. In this gun, the pointed field emission cathode is firstly cleaned
and somehow smoothed by a flashout and then supplied with high voltage in
vacuum of about 10–7 Pa without heating up the cathode. The beam current
plummets in the first period, stabilities in period B, but grows in period C. The
reason is that infection of the apex surface with adsorbed molecules smooths the
Fig. 8.83 Oscillograms of current in an electron gun over time: a change of current for an
ordinary electron gun; b change of current for an electron gun with stabilization; c change of
cathode heating current
electrode. The focused beam gets out of the aperture of the accelerating electrode.
The focusing field depends on the U2/U1 ratio, the voltages applied between the
accelerating electrode and the anode and the cathode, respectively. The surface state
of cathode 1 varies over time. Unless something compensates for these changes, the
emission current is unstable. On the other hand, the field emission current density
depends both on field intensity and the surface state, which makes it possible to
8.5 Electron Guns 511
maintain constant emission by variation of the electric field. However, it has already
been stated above that no variation of voltage U1 is allowed to prevent changing the
U2/U1 ratio and consequent variation of the electron-beam focusing. Variation of
U1 makes the beam leaving the aperture of accelerating electrode 3 vary over time
for the above-stated reasons even if the emission is maintained stable.
To avoid that variation, the electron gun is equipped with reference electrode 5
and the appropriate potential source 13. Reference electrode 5 is put not far from
cathode 1, which makes it possible to change the electric field at the cathode by
variation of the voltage of source 13, leaving voltage U1 constant.
Stability is automatically controlled by comparator 7. It consists of resistor 8,
reference voltage source 9 and a difference amplifier.
When sample 6 is lit by the electron beam, the current passes through resistor 8,
causing a drop of potential. The voltage is compared with the reference voltage
from source 9. As a result, the output of difference amplifier 10 develops a dif-
ference signal that comes to control unit 12. Control unit 12 gives out an actuating
signal to source 13, which changes its output voltage until the difference signal
disappears.
A sudden failure of the vacuum in the electron gun rapidly destroys the field
emitter as a result of disruption between the anode and the cathode. Paper [225]
suggests a number of circuits able to protect a field emission vacuum gun against
disruption. The operating principle of all the circuits is based on variation of
resistances of the circuit components, which leads to shunting the discharge gap and
eliminating the discharge.
Paper [226] suggests a design of a field emission gun combining some means of
stabilization and protection of the field emission cathode. Figure 8.86 diagrams a
gun used in a raster-type electron microscope. Field emission gun 1 issues a bright
focused electron beam that lights up sample 2. The focused beam is scanned with
As the current from one field emitter is very low, generation of high-ampere cur-
rents needs simultaneous functioning of many hundreds and thousands of field
emitters. The core of this problem has not been solved yet. However, the attempts to
meet the challenge are continuously under way.
Some other approaches to development of high-current field emission cathodes
is given in Chaps. 5, 6. In this section, we are going to consider just a few aspects
of this problem.
Classical high-current field emission cathodes can be exemplified with arrays of
pointed molybdenum field emission cathodes (look issue 5.2 and 5.8). Such arrays
employ more than 100,000 apices at the working pressure better than 10−9 mm Hg.
As already pointed out, field emission cathodes are manufactured from
carbon-based materials to reduce the working pressure.
In particular, paper [232] does so with an array of pointed field emission cath-
odes made of glassy carbon manufactured by the photolithography technology and
thermo-chemical etching. The emitters are 14 lm high. The vertex curve radius is
about 1–3 lm. The emitter density is 2.5 105 cm−2. If the cathode-plate diameter
is 3 mm, the emission current is 100 mA. In addition, there are field emission
cathodes of a large area, viz. more than 1 cm2, made of carbon-fiber bunches [233]
or high-tenacity graphite, e.g. fine-grained dense graphite MPG-6 [234] with a
high-transmission control grid positioned near the cathode plane. This electron-gun
structure makes it possible to obtain an electron beam of any configuration
depending on the shape and the size of the effective area of the field emission
cathode.
An example of such a structure is shown in Fig. 8.88.
Fig. 8.88 Electron gun with a graphite field emission cathode: a main circuit of the gun: 1 field
emission cathode made of MPG-6 graphite; 2 work surface of the cathode; 3 extraction electrode (a
tungsten wire gauze with a diameter of 0.01 mm and transparency of 80%); 4 electron beam
boundary; 5 anode. b ribbon-type field emission cathode; c matrix-type field emission cathode
8.5 Electron Guns 515
The field-electron gun is field emission cathode 1 with grooved end surface 2. In
close proximity to the effective area of the field emission cathode, there is extraction
electrode 3 represented by a grid made of a high-melting material. Application of an
electric field to the extraction electrode from the microelevations on the effective
area of the field emission cathode causes field emission.
The electron beam, whose configuration depends on that of the cathode, passes
the extraction electrode and gets to anode 5, which collects electrons.
The end surface of the field emission cathode made of a nanostructured
carbon-based material consists of a great number of microelevations statistically
uniformly distributed over the effective area. In the course of emission under
bombardment of the surface with residual-gas ions, the emission centers are
damaged and at the same time new microelevations are exposed to become new
emission centers. After the effective area of the cathode has aged, this process
stabilizes in time, which provides high time stability of the emission current drawn
from such a cathode. Increase in the working area of the cathode by more than 4–
6 mm2 aimed to increase the cross-section of the emitted beam causes spatial
instability of the beam because of heterogeneity of the cathode surface of such size.
A large-section beam can be generated by a set of separate stable electron beams
emitted by the structural elements of the end surface of a small-size cathode
commensurable with the cathode-extraction electrode gap.
If the dimensions of the structural element of the effective area are larger than the
cathode-extraction electrode gap, each element can work as an independent cathode
issuing an almost parallel electron beam. Moreover, such geometry improves
homogeneity of emission from such an element due to more homogeneous electric
field. Almost the same situation is true when the ratio of dimensions is
1 2D. The ratio has been selected experimentally.
When 1 < 2D, the electron beams from the neighboring microelevations
superimpose on one another. Reduction of D is mostly determined by the me-
chanical strength of the grid and the technology of its tensioning, which determines
a possibility to mount the device, less D causing less extraction voltage. These
ratios are assumed for the identical dimensions of the grooves and the projected
parts. The ratio can vary towards increase in the projected part. However, the
increase can reduce the overall performance of each element edges because other
elements can shield them. To prevent the bottom of the grooves from contributing
into the cathode-field emission, i.e. for the electric field intensity at the bottom to be
much lower than it is necessary to excite field emission, the groove depth should
meet ratio 2. The ratio was selected by experiment. h/D = 2 is the minimum
giving a positive effect. Practically, this ratio was obtained as 10–15, which is
mostly determined by the mechanical strength of the material used.
The problem of the extraction electrode is that it should have a high
electron-beam transmission factor, on the one hand, and provide uniformity of the
electric field at the cathode effective area, on the other hand. The structures like thin
foils or orifice plates do not meet the transmission-factor criterion. To provide
uniformity of the electric field of the extraction electrode represented by a wire
mesh of a high-melting conductor, the size of the bridges should be more than 0.02–
516 8 Field Emission Cathode-Based Devices and Equipment
0.1D with transparency of 70–90%. Such a grid provides a uniform potential at the
cathode and the maximum transmission of an electron beam. Increase in the size of
the bridge or the wire diameter at invariable D exceeding 0.1D increases
non-uniformity of distribution of the electric field, which results in non-uniform
emission from the surface, the minimum value of 0.02D being determined by the
mechanical strength of the grid material.
The field-emission cathode can be made of different materials, e.g., carbon
nanotubes, carbon foils, constructional graphites etc. [235, 236].
8.6.1 Heaters
The characteristic features of field emission can substantially improve the opera-
tional parameters of electronic devices. That is why the above-listed guidelines of
development of field emission cathode-based devices do not exhaust the possibil-
ities to use field emission cathodes in a wide range of devices and appliances.
This section outlines some possible types of devices where field emission
cathodes improve the operational parameters.
Some constructional elements of electronic devices need heating when in
operation. They include thermionic cathodes, gas leaks, getter pumps, getter-ion
pumps etc. The operating principle of such heaters is that emitted electrons are
accelerated in the anode-cathode gap and decelerated in the anode material with
heating of the latter.
The optimum heater is a field emission cathode.
When working, the field emission cathode is at the ambient temperature, which
removes the power expenditure due to heat emission by the heated cathode. That
makes the efficiency of the pump nearly two times higher. Unlike hot cathodes,
decrease in the operating currents and absence of a special heater makes it possible
to reduce the number of current collectors to two and cut the power losses due to
heating and heat transfer along them.
A lot of sophisticated electronic devices, especially those equipped with special
glued-in windows, employ embedded tiny getter pumps or getter-ion pumps when
the capacity of conventional flash getters is not enough [237].
As a getter material, such small pumps use lanthanides, e.g., erbium, gadolinium,
etc. This type of hot cathode-based pumps [237, 238] does not provide a short
triggering time and have relatively high power consumption and dimensions.
Figure 8.89 shows the designs of small pumps with field emission cathodes of
fine-grain high-density graphite MPG-6 type based on lanthanides [239].
The operating principle of the pumps is as follows. Voltage is applied between
anode 1 and cathode 2 so that the electrons emitted from the elevations of the field
emission cathode bombard the lateral surface of tantalum cup 6 or thin-walled
tantalum cylinder 7. Dissipation of the field emission electron energy heats the anode
8.6 Other Types of Devices 517
Fig. 8.89 Design of getter pumps with field emission cathodes: a cylindrical cathode with an
internal work surface; b flat cathode; c cylindrical cathode with an external work surface. 1
lanthanide metal; 2 graphite field emission cathode; 3 insulator between the anode and cathode; 4,
5 field emission cathode and anode holders; 6 tantalum cup; 7 thin-walled tantalum cylinder
and sublimes its material. All the three pump designs are essentially identical and
differ only in the heater design and the spreading angle of the getter material. So let
us consider one design in more details and note the distinctive features of the others.
Pump (Fig. 8.89a) consists of an evaporation anode, including cylindrical tan-
talum cup 6 with a pressed-in lanthanide tablet 1. Cup 6 is welded to the bottom of
the structure through supporting arm 5. The functions of graphite field emission
cathode 2 are performed by shape-fitted elevations on the interior of the cathode
cylinder.
The field emission cathode is fixed in back-up ring 4 soldered to insulators
coaxially with anode cup 6. In this design, the cathode envelops the anode, which
reduces the triggering time and the spreading angle of the getter material.
The spreading angle of the material is increased in the design detailed in
Fig. 8.89b.
These two designs can be used in the devices with a limited getter-sputtering
area. An unlimited area makes it possible to sputter round, which provides best
pumping rates (Fig. 8.83c). In this case, cathode cylinder 2 is inside the anode, the
latter being a thin-walled tantalum cylinder 7 with a lanthanide foil either laser- or
contact-welded thereon.
Such pumps have the pumping rate of l l/s. and the volume of 1 cm3, the
designs themselves being sufficiently rigid for vacuum electronic devices.
The field emission cathodes of such pumps should have a number of design
properties. First of all, the effective area of the field emission cathodes is laid down
with elevations of 0.3–1 mm in width. They work effectively if the space between
the emitting microelevations on the effective areas of the elevations and the anode is
substantially smaller than the working part of the elevations. In this case, the
microelevations located on the effective area of the elevations make the best con-
tribution to the field emission. If the space increases, the electric field intensity on
the microelevations decreases and the elevation works as a whole, i.e., mostly by
the perimeter, which substantially weakens the overall performance of the field
emission cathode.
518 8 Field Emission Cathode-Based Devices and Equipment
Experiments have established that the effect of the groove bottoms on the field
emission from the elevations is eliminated if the height of the elevation exceeds the
space between the anode and the effective area of the emitting elevations.
Except for the power consumption directly providing field emission, the former
is reduced due to special design properties, e.g., thermal-leakage reduction, parasitic
electron-scattering reduction etc., which are clear enough and do not need any
consideration in this section.
Testing involves a few experimental samples of each of the above-mentioned
designs. The evaporation-anode material is erbium. The tantalum cup is 5 mm in
diameter and 3 mm in thickness. The power consumption is 3 W at the pumping
rate of 1 l/s, which is about 2.5 times less than that of the analogues [201, 238].
Field emission heating is used in a light valve, which can function as a
high-resolution light projector [240], the design and the operating principle being
presented in Fig. 8.90.
The device suggested contains a vacuum shell represented by flat glass envelope
1 with entrance window 2 in a form of a plane-parallel plate, target 3 with lobes 4 of
an intelligent material and a system of emissive field emission cathodes 5 as an
electron source (Fig. 8.90a). Target 3 includes a substrate represented by metalized
ceramic bars 6 (Fig. 8.90b) located in one plane in parallel to each other with a gap
in between and a band of an intelligent-material, e.g., nickel-titanium, each having
one series of cut T-shaped lobes 4 connected at base 7 with a film band of the same
shape as bar 6 and fixed on it with an electric contact. Base 7 bears dielectric layer
8. Every lobe 4 is set a reversible memory of two shapes, namely a rectilinear shape
in the target plane and a curved shape at an acute angle to the target, while lobes 4,
coated with a reflective material (missing from Fig. 8.90), form a mirror matrix.
The lobe is as long as the space between bars 6 and wider than the bar, the gap is
wider than the length of the bar, the gap is wider than the length of lobe stem 9, and
lobe body 10 is above the next bar. Field emission cathodes 5 can be made of
fine-grain high-density graphite MPG-6 in a form of combs isolated from each other
and mounted at right angle to bars 6 in the plane parallel to the plane of the
substrate. Comb elevations 11 are located in the gap between the bars opposite lobe
stems 9 at the distance shorter than the linear stem surface dimension, and the
dimensions of effective area 12 of the cathode elevation are smaller than the lobe
stem surface dimensions.
Every base 7 and every comb of field emission cathode 5 is connected to
terminals 13 sputtered on the bottom of envelope 1. The device has projection
optical system 14 consisting of two lenses, viz. 01 and 02, and diaphragm D, the
aperture center of the latter coinciding with the lens focuses.
The light valve works as follows.
When power supply is off, all lobes 4 are rectilinear and pressed to bars 6. As a
result, collimated light beam 15, which has passed through entrance window 2 of
envelope 1, is reflected from the mirror coating applied on the lobes and gets to the
entry to projection optical system 14. Supply of base 7 with a potential positive in
respect of the comb of field emission cathode 5 results in field emission from comb
elevation 11 located at their intersection. An electron beam affects stem 9 of lobe 4
8.6 Other Types of Devices 519
Fig. 8.90 Design and operation diagram of a light-valve device a; its target with a system of field
emission cathodes. 1 Glass bulb; 2 input window; 3 target; 4 lobes made of material with shape
memory; 5 field emission cathodes; 6—metallized ceramic sliders; 7 base; 8 dielectric layer; 9 lobe
stem; 10 lobe body; 11 ledges of the field emission cathode; 12 working surface of a cathode
ledge; 13 leads; 14 optical projection system; 15 light beam
and causes heating and curving of the latter due to a structural-phase transition in
the lobe-stem material. The collimated light beam reflected from the lobe does not
pass through projection optical system 14. When the control voltage is turned off,
520 8 Field Emission Cathode-Based Devices and Equipment
the lobe is pressed to the substrate. The entire array of lobes has the information
recorded in the same way.
The intelligent material lobes and data recording with field emission make it
possible for the device to be flat, need no electron beam focusing system or cathode
heating (unlike the prototype), provide the maximum power consumption of an
electron beam, which contributes to downsizing and performance-gaining of the
device. At the same time, it has an increased contrast and vibration resistance.
These advantages scale up application of light valves in data processing systems.
8.6.2 Sensors
Field emission cathodes can essentially increase sensitivity of sensor devices, i.e.
those responding to even a small change in any physical parameter.
In particular, paper [241] describes and rates a high-sensitivity magnetic sensor
based on a lateral field emission device, the arrangement drawing and the electric
connection diagram being given in Fig. 8.91.
Fig. 8.92 a Sectional drawing of a memory element based on field emission cathodes made of
carbon nanotubes. The element consists of two triodes N1 and N2, marked by dashed outlines.
Triodes are assembled separately on two substrates and later connected as a single unit.
b Equivalent electric circuit of an element of a memory element. Cf is the main capacity, Rb is the
ballast resistor, Cbl is stray capacity
522 8 Field Emission Cathode-Based Devices and Equipment
This element can be made according to the conventional micro- and nanotech-
nology. Note that the memory cell is made up of an upper part and a lower part
showed on two separate silicon bodies.
Body I is ion-implanted to form, firstly, a lightly-doped region functioning as
limiting resistor RB, and secondly, a heavily-doped layer specified in Fig. 8.92a as a
WL for the memory cell. Installation of vertical carbon nanotubes (CNT) and
reduction of capacitance of the control electrode needs relatively thick, viz.
1 µm, layer of silicon oxide. Subsequently, the oxide layer is sputtered with
metal and the coating is properly configured by the photolithographic method (cf.
the intermittent BL). The next stage is formation of modulator apertures through the
layers of oxide and metal (BL) by electron-beam lithography.
After that, an appropriate diffuse layer, e.g., TiN, and a catalyst, e.g., Ni, Fe, Co,
are sputtered through the modulator aperture, and vertical nanotubes are grown by a
CVD method. The grown-up carbon nanotubes are field emission cathodes of triode
N1. The BL area functions as the control electrode for triode N1 and the anode for
triode N2 (Fig. 8.92b).
Triode N2 nanotubes, i.e. the second part of the memory cell, are fabricated on
substrate II. At first, a thin layer of silicon oxide, viz. 20 nm, is grown up on
substrate II. The metal is sputtered into the oxide and etched away into islands,
forming bias capacitor CF (Fig. 8.92b). This metal layer serves as anode for triode
N1. A thick oxide layer is coated with a metal. This layer is configured by a
photolithographic method (SL line) and the aperture etched in the oxide lets the
triode—N1 electron beam reach the anode. Electron-beam lithography is used to
form an aperture in the control electrode and a place to grow up triode—N2
nanotubes.
It may be noted that substrates I and II can be fabricated both simultaneously and
separately.
Paper [245, 246] suggests field emission solid-state memory cells. The basic
principle applied is that the electric field between the emitter contact and the col-
lector electrode of the condenser-type film becomes high enough for the charge
carriers to get sufficient energy for tunneling through the collector into the insulated
retaining tank below the collector. The basic circuit of the cell is presented in
Fig. 8.93.
Flat substrate 1 can be made of silicon or glass. Silicon is most preferable since
its chemical, mechanical and electrical properties have been well studied.
Insulating oxide layer 2 is applied on substrate 1. It is at least 100 Å thick. This
layer 2 is coated with a layer of polysilicon 3 as thick as 500–100 Å coated with a
layer of oxide or nitride 4. The layer is selected depending on the subsequent
middle-band metal layer 5. The thickness of the layer ranges from 10 to 50 Å.
Metal layer 5 to be sputtered is as thick as a few nanometers. The thickness and
structure of this metal layer can be reproduced in production of semiconductor
chips.
The term of middle band means that the Fermi level of the collector material lies
in the middle between the conduction band and the valence band of the
charge-storage tank, i.e. polysilicon 3.
8.6 Other Types of Devices 523
Fig. 8.93 Schematic diagram of a solid-state field emission element of memory: 1 substrate; 2
isolating oxidic layer; 3 polysilicon; 4 layer of nitride or oxide; 5 metal layer; 6 layer of nitride or
oxide; 7 conductor layer; 8 emitter contact; 9 first electric contact; 10 virtual capacitor; 11 earth
contact
After that, metal layer 5 is sputtered with the second oxide or nitride layer 6
similar to layer 4 in properties. The next stage is sputtering conductive layer 7
consisting of metal or polysilicon. It is as thick as layer 5. All these layers form an
island structure shown in Fig. 8.87.
First electric contact Vp 9 is connected to the middle-band layer 5. Second
emitter voltage 8 is connected to conductive layer 7. Two voltages applied to two
conductive layers through the insulator layer also function as a capacitor. While the
voltage of emitter VE is negative in respect of the collector voltage, the electrons
tunnel from emitter layer 7 to the middle-band metal layer 5. At the voltage of
VE 4 V, electrons reach first nitride layer 4, and some of them overcome the
potential barrier and move to tank 3 to be caught in this way.
Tank 3 has a capacitor connected to substrate 1 specified in Fig. 8.93 as virtual
capacitor 10. Substrate 1 is earthed through contact 11.
When the positive voltage of 6 V is applied to contact 8, the holes are injected
into area 3, which neutralizes the electrons.
A similar version of the memory cell with a silicon field emission cathode is
given in [247].
A very similar principle is the basis for a transistor, which can also be considered
as a memory cell.
A high initial velocity of field electrons accounts for their good ionization ability.
This feature initiated the first attempts to design them 50 years ago [248], and the
work has still been under way.
Ion sources can be based on both pointed [249, 250] and carbon-fiber [251, 252]
field emission cathodes.
An ion-source structure with field emission cathodes made of carbon fibers is
presented in Fig. 8.94.
524 8 Field Emission Cathode-Based Devices and Equipment
Fig. 8.94 Low-background ionic source with a field emission cathode from carbon fibers: 1
copper plate; 2 ion lead opening; 3 ion beam; 4 field emission cathode; 5 metal disk; 6 bundles of
carbon fibers; 7 extraction electrode; 8 retarding grid; 9 Faraday cylinder; 10 ion lens; 11
molecular beam; 12 external thermal screen; 13 pipe; 14 liquid helium; 15 radiation screen
Original copper plate 1 has central aperture 2 to transmit ion beam 3 to the mass
spectrometer (missing from the picture). Field emission cathodes 4 rest on metal
disk 5, carbon-fiber bundles of 8 fibers 6 arranged symmetrically around the disk
circle, so that fiber tips were directed to the chamber center. Every bundle is about
1 mm in diameter.
Cylindrical extraction grid 7 is mounted inside field emission cathode 4. The
extraction voltage of 2–4 kV makes the electrons pass accelerating grid 7 and move
to decelerating grid 8. Grid 8 is under the potential of a few hundred volts. The
internal part of Faraday cup 9 is the ionization area where the decelerated electrons
oscillate and ionize gas molecules. A series of diaphragms 10 is an ionic lens
intended to take the ions out of the ionization area.
The left sides of grids 8 and 9 have central apertures intended to transmit
molecular beam 11. Faraday cup a is under the earth potential. Copper plate 1 is in a
good thermal contact with external screen 12 cooled by liquid nitrogen. An
ion-source electrode system is ducted in cylindrical tube 13 of a cryogenic pump or
a cryostat cooled by liquid helium. Tube 13 and other parts of the cryogenic pump
are made of stainless steel and connected with each other by welding.
8.6 Other Types of Devices 525
primary spectral colors, thereby changing the wavelength of the light emitted.
Variation of the inter-pulse period changes the brightness of the light flux.
Source 1 of the collimated light flux with a variable emission wavelength can
simultaneously replace a few elements providing an intensive light flux with the
necessary spectrum.
Based on the above principles, variable-wavelength source 1 works in the
highlight pulsed mode of 30,000 cd/m2. Every impulse makes source 1 generate
collimated light emission corresponding to red, green or blue colors. The light-flux
size meets the light-modulator 2 size that forms a polychromatic highlight tiny
image of no more than 20 20 mm projected by lens 3 to external screen 4. The
entire system is controlled from control unit 5.
The light flux emitted should have the maximum intensity in the desired spectral
intervals along with the minimum electrical-energy expenditure. It should also be
incoherent as otherwise there would be light interference destructive towards the
image formed and violating correct image transmission.
The device employs the field sequential color (FSC) system, when a full-color
image frame is formed of three basic components, wherefore the synchropulse
arriving at light source 1 makes the latter emit a light flux sequentially consisting of
three impulses of primary colors, viz. red, green and blue.
As an electron source, light source 1 employs a composite field emission cathode
based on carbon, viz. graphite, carbon, one-wall or multiple-wall carbon nanotubes
etc. Carbon-based field emission cathodes are able to work in high technical vac-
uum of 10−6–10−7 mm Hg achieved in conventional sealed-off devices and provide
high emissive capacity.
Up-to-date luminophores make it possible to achieve high light intensity, viz.
*30,000 cd/m2, which makes it possible to curtail the light-source dimensions by
making its emitting part *20 20 mm, i.e. equal to light modulator 2 in size.
The function of light modulator 2 can be performed by any controllable matrix
of microelements if their variable configuration can modulate the light flux from
light source 1.
Employment of field emission light sources in this type of devices provides the
following advantages:
– contrast improvement in the system based on a pulsing changeable-wavelength
light source;
– increase in the image-forming frequency in the system based on use of the above
light source and, respectively, non-use;
– a necessity to apply a composite light filter with an electromechanical principle
of light-filter displacement;
– reduction of dimensions of the image-forming device caused by lack of a
composite light filter based on an electromechanical principle of light-filter
displacement, a liquid-crystal shutter and a joined optical system consisting of
the first and the second lens arrays, a collimator lens and heat-removal fans;
– reduction of noise and undesirable vibrations connected with lack of elec-
tromechanical light-filter devices and heat-removal fans;
8.6 Other Types of Devices 527
References
1. A.S. Bugaev, V.B. Kireev, E.P. Sheshin, A.Y. Kolodyzhnyj, Сathodoluminescent light
sources: status and prospects. Physics-Uspekhi v 58(2), 792–818 (2015)
2. Y.B. Ayzenberg, Lighting Engineering (Energoatom, M., 1983)
3. E.P. Sheshin, Structure of a Surface and Field Emission Properties of Carbon Materials
(MIPT, M., 2001)
4. E.P. Sheshin, A.L. Suvorov, A.F. Bobkov, D.E. Dolin, Light source on the basis of carbon
field electron cathodes: design and parameter, in Abstract of 7th International Vacuum
Microelectronics Conference, Grenoble, France (1994), pp. 423–426
5. A.S. Baturin, V.S. Kaftanov, S.G. Kyzmenko, E.P. Sheshin, Field emission device. Patent R.
F. No. 2180145, Cl. H01J/28 of 15.02.2000
6. M.Y. Leshukov, A.S. Baturin, N.N. Chadaev, E.P. Sheshin, Characterizations of light
sources with carbon fiber cathodes. Appl. Surf. Sci. 215, 260–264 (2003)
7. M.Y. Leshukov, N.N. Chadaev, A.S. Baturin, E.P. Sheshin, Vacuum cathode luminescence
lamps for systems of display of information, in Proceedings of Conference on Modern
Television of Moskow (2004), pp. 36–38
8. M.O. Popov, S.Y. Bulakhon, A.V. Karpov, S.A. Shiriaev, E.N. Skorokhodov, A.L. Suvorov,
Influence of parameters of field emission cathodoluminescence light sources on their
technical properties. Appl. Surf. Sci. 215, 253–259 (2003)
9. E.P. Sheshin, A.Y. Kolodyazhnyiy, A.S. Raufov, Sources of general lighting with field
emission cathodes from carbon materials. Izvestiya Vysshikh Uchebnykh Zavedenii; Seriya
Khimiya I Khimicheskaya Tekhnologiya 58(7), 69–73 (2015)
10. M.Y. Leshukov, E.P. Sheshin, Optimization of an electron-optical system of a cathode
luminescent light source with the field emission from carbon fibers, in Proceedings of
Conference on Stability and Management Processes, vol. 1, St. Petersburg (2005), pp. 202–
208
11. M.Y. Leshukov, E.P. Sheshin, New design of electron gun for field emission light sources
with carbon fibers cathode. Hydrogen Materials Science and Chemistry of Carbon
Nanomaterials. NATO Security Through Science Series A: Chemistry and Biology, Special
issue, XLI (Springer Science, 2007), pp. 255–258
12. M.Y. Leshukov, N.N. Chadaev, E.P. Sheshin, Field emission light source with carbon fibers
bundle cathode. Technical Digest of IVNC, Oxford, UK (2005), pp. 340–341
13. W. Knapp, O.F. Kieler, D. Schleussner, E.P. Sheshin, I.N. Yeskin, Fieldemission von
kohlenstoff-fasern und deren anwendung in einem CRT-lightelement. Konferenzband
Electronic displays, Berlin (2000), pp. 92–98
528 8 Field Emission Cathode-Based Devices and Equipment
14. W. Knapp, D. Schleussner, A.S. Baturin, I.N. Yeskin, E.P. Sheshin, CRT lighting element
with carbon field emitters. Vacuum 69, 339–344 (2003)
15. H. Murakami, M. Hirakawa, C. Tanaka, H. Yamakawa, Field emission from well-aligned,
patterned, carbon field emitters. Appl. Phys. Lett. 76(13), 1776–1778 (2000)
16. Y. Saito, K. Hamagychi, R. Mizushima, S. Uemura, T. Nagasako, J. Yotani, T. Shimojo,
Field emission from carbon nanotubes and its application to cathode ray tube lighting
elements. Appl. Surf. Sci. 146, 305–311 (1999)
17. Y. Saito, K. Hata, A. Takakura, J. Yotani, S. Uemura, Field emission of carbon nanotubes
and its application as electron sources of ultra-high luminance light-sours devices. Physica B
B323, 30–37 (2002)
18. Y.-C. Yang, J. Tang, L. Liu, S.S. Fan, Field emission device. Patent USA №US
2008/0007153 of 30.11.2006. Cl 313/497 (2006)
19. Y.-C. Yang, J. Tang, L. Liu, S.S. Fan, Field emission device. Patent USA №US
2008/0007153 of 30.11.2006. cl.313/496 (2006)
20. A.N. Obraztsov, V.J. Kleshch, E.A. Smolnikova, A nano-graphite cold cathode for an
energy-efficient catholuminescent light source. Beilstein J. Nanotechnol. 4, 493–500 (2013)
21. K.-J. Chung, N.N. Pu, M.J. Youh, Y.-M. Liu, M.-D. Ger, W.-K. Huang, Improvement of
lighting uniformity and phosphor life in field emission lamps using carbon nanocoils.
J. Nanomater. 2015, ID373549 (2015) 9 p
22. Y.-C. Yang, Y. Wei, L. Liu, K.-L. Jiang, S.S. Fan, Field emission device. Patent USA №US
2008/0012466 of 16.11.2006, cl 313/495 (2006)
23. Y.-C. Yang, Y. Wei, L. Liu, K.-L. Jiang, S.S. Fan, Pixel tube for field emission device.
Patent USA № US 2008/0030123 of 12.07.2007, cl.313/495 (2007)
24. www.inanov.fr
25. Y. Kondo, J. Matsuura, H. Kimura, Characteristics of field emitter arrays for light source.
Technical Digest IVMC 91, Nagahama (1991), pp. 56–57
26. Y.L. Ke, J. Zhou, J. Chen, S.Z. Deng, N.S. Xu, Luminescent tubes based on MoO2 nanowire
cold cathode. Technical Digest JVNC, Oxford, UK (2005), pp. 362–363
27. V.S. Kaftanov, A.L. Suvorov, E.P. Sheshin, Field emission cathode and methods in the
production thereof. Patent USA. №5588893 cl.445/6(H01 J 9/02) of 6.06.1995
28. V.S. Kaftanov, A.L. Suvorov, E.P. Sheshin, A field emission cathode and methods in the
production thereof. Patent RST № WO 96/25753 of 15.02.1995
29. V.S. Kaftanof, A.L. Suvorov, E.P. Sheshin, Metod vid tillverkning av en faltemissions
cathode samt faltemissions cathode. Patent Sweden №504603 of 02.15.1995 cl.H01 J/30
30. V.S. Kaftanov, A.L. Suvorov, E.P. Sheshin, Field emission cathode and methods in the
production thereof. Patent USA № 5973446 of 15.02.1999 cl.313/310 (H01 J 1/05)
31. G. Forsberg, C.-H. Andersson, Method of producing a field emission cathode and a light
source. Patent RST WOOO/77813 A1 of 10.06.1999
32. W.-H. Tsai, M.-C. Fan, J.-C. Chang, Flat field emission illumination module. Patent USA №
US 2008/0036361 of 14.02.2008 cl.313/496 (H01j 1/62)
33. W-H. Tsai, M.-C. Fan, J.-C. Chang, Flat field emission illumination module. Patent UK GB
2441618, of12.03.2008. cl. H01j 63/02
34. P.J. Chen, P. Liu, L.M. Sheng, Y. Wei, L. Liu, Z.F. Hu, C.L. Guo, S.S. Fan, Light source
apparatus using field emission cathode. Patent USA US 2006/0022574 of 2.02.2006,
cl.313/495 (H01J 9/02)
35. J.H. Choi, A.R. Zoulkarneev, Y.T. Park et al., Cathodoluminescent flat lamp using the
suspend wire structure. Technical Digest of JVNC 2005, Oxford, UK (2005), pp. 350–35
36. K.W. Cheng, S.H. Lee, C.Y. Hsiao, Field emission type planar lamp and method for the
same. Patent USA US 2008/0079348 of 3.04.2008 cl.313/495 (H01 J 1/02)
37. L. Qian, L. Lin, P. Liu, J. Tang, Y. Wei, S.S. Fan, Field emission double-plane light source
and method for making the same. Patent USA US 2007/0222363 of 27.09.2007 cl.313/496
(H01 J 63/04)
References 529
38. A.N. Obraztsov, J.Y. Pavlovsky, A.P. Volkov, Prototype of light emitting device ith thin
film cold cathode. Technical Digest of Euro Display 99, Berlin, Germany (1999), pp. 229–
231
39. A. Gorecka-Drzazga, B.J. Cichy, P. Szczepanska, R. Walczak, J.A. Dziubun, Field emission
light sources for lab-on-a-chip microdevices. Bull. Polish Acad. Sci. Techn. Sci. 60(1), 13–
17 (2012)
40. Y. Kuroiwa, S. Narushima, S. Ito, Electron emitter, field emission display unit, cold cathode
florescent tube, flat type lighting device, and electron emitting material. European patent EP
1876628 29.01.2008 H01 J 1/304 (2008)
41. V.B. Sharov, E.P. Sheshin, A.A. Shyka, Field emission cathodes from graphite for diode
light sources. Nano Microsyst. Equip. 3, 17–19 (2005)
42. M.-J. Youh, C.-L. Tseng, M.-H. Jhuang, S.-C. Chiu, L.-H. Huang, J.-A. Gong, Y.-Y. Li, Flat
panel light source with lateral gate structure based on SiC nanowire field emitters. Sci.
Rep. 5, 10976, c1–34 (2015)
43. V.S. Kaftanov, A.L. Suvorov, E.P. Sheshin, J. Olsford, Field emission cathode and a light
source including a field emission cathode. Patent USA №5877588 of 5.03.1999, cl.313/491
(H01 J 1/30)
44. V.S. Kaftanov, A.L. Suvorov, E.P. Sheshin, J. Olsford, Light source including a field
emission cathode, and a field emission cathode. Patent USA 6008575 of 28.12.1999, cl.
313/484 (H01 J 1/30)
45. E.P. Sheshin, A.L. Suvorov, V.S. Kaftanov, J. Olstors, Field emission cathode and a light
source including a field emission cathode. Patent Canada CA 2293269 cl, H01 J 001/30 of
17.12.1998
46. A.A. Blyablin, A.V. Kandidov, A.T. Rakhimov, V.A. Samorodov et al. Divergence of
electrons emitted from a carbon cold cathode cold cathode. Technical Digest of IVMC 1999,
Darmstadt, Germany (1999), pp. 346–347
47. J.-M. Bonard, T. Stockli, O. Noury, A. Chatelain, Field emission from cylindrical carbon
nanotube cathodes: possibilities for luminescent tubes. Appl. Phys. Lett. 78(18), 2775–2777
(2001)
48. X.X. Zhang, C.C. Zhu, X. Li, W.Z. Cui, The fabrication of novel structure of field emitting
light tube with carbon nanotubes as cathode. Technical Digest of IVNC, Oxford, UK (2005),
pp. 342–343
49. J.X. Huang, J. Chen, S.Z. Deng, J.C. She, N.S. Xu, Optimization of carbon nanotube
cathode for a fluorescent lamp. Technical digest of IVNC, Oxford, UK (2005), pp. 284–285
50. J. Kjellman, M. Lindmark, Light source, and a field emission cathode. Patent USA 6873095
of 29.03.2005 313.336 (H01 J 1/16)
51. Y. Yun, G. Tailiang, J. Yadong. Application of ZnO nanopillars and nanoflowers to field
emission luminescent tubes. J. Semiconductors, 33(4), 043003, 1–5 (2012)
52. A.N. Obraztsov, A.P. Volkov, A.A. Zakhidov, D.A. Lyashenko, Y.V. Petrushenko, O.
P. Satanovskaya, Field emission characteristics of nanostructured thin film carbon materials.
Appl. Surf. Sci. 215, 214–221 (2003)
53. A.N. Obraztsov, Cathodoluminescent light source. Patent USA US 005/0174059 of
11.08.2005 cl. 313/634 (H01 J 63/04)
54. L. Qian, L. Lin, P. Liu, J. Tang, Y. Wei, S.S. Fan, Field emission lamp and method for
making the same. Patent USA US 2007/0228919 of 4.10.2007, 313/310, (H01 J 9/02)
55. A.S. Leychenko, M.Y. Leshukov, N.N. Chadaev, E.P. Sheshin, Effective lamp for
LCD-backlighting with the field emission cathode. Technical Digest IVNC IFEC, Guilin,
China (2006), pp. 383–384
56. A.S. Leychenko, M.Y. Leshukov, N.N. Chadaev, E.P. Sheshin, Field emission lamp
illuminations for l.c. displays, in Proceedings of Modern Television Conference, Moscow
(2006), pp. 30–31
57. A.L. Suvorov, E.P. Sheshin, N.E. Lazarev, N.N. Chubun, Vacuum luminescent light source
with carbon fibres field emission cathode. Technical Digest of IVMC, Portland USA (1995),
pp. 516–521
530 8 Field Emission Cathode-Based Devices and Equipment
58. P. Liu, Y. Wei, L.M. Sheng, L. Qian, J. Tang, L. Liu, C.L. Guo, C.L. Du, B.C. Du, S.S. Fan,
Field emission luminescent light source. Patent USA US 2006/0091782, of 04.03.2006
cl.313–496 (H01 J 63/04)
59. K.J. Chung, C.C. Chiang, Y.M. Liu, N.W. Pu, M.D. Ger, The study of fabricating the field
emission lamps with carbon nano materials. Int. J. Chem. Mol. Nucl. Mater. Metall. Eng. 6
(5), 437–439 (2012)
60. L. Qian, L. Liu, P. Liu, J. Tang, Y. Wei, S.S. Fan, Field emission lamp and method for
making the same. Patent USA US 2007/0247071 of 25.10.2007. cl.313/634 (H01 J 61/30)
61. C. Li, K. Song, C. Lan, Field emission excited UV light source structure and preparation
method there of. Patent CN 103 400919 от 20.11.2013 кл.HOIj-063/06
62. S. Ono, T. Suyama, K. Fukura, S. Ishizu, N. Kawaguchi, T. Nagami, A. Yoshikawa, T.
Yanagida, Y. Yokota, Device for emitting vacuum ultraviolet light. Patent W0201127881 от
10.03.2011,кл.Н01о-061/38,Н01j-063/00
63. M. Yanagihara, M.Z. Yusop, M. Tanemura, S. Ono, T. Nagami, K. Fukuda, T. Suyama, Y.
Yokota, T. Yanagida, A. Yoshikawa, Vacuum ultraviolet field emission lamp utilizing
KMgF3 thin film phosphor. APL Mater. 2, 046110 (2014)
64. V.D. Blank, S.G. Buga, I.V. Ekhmenina, N.N. Chadaev, E.P. Sheshin, Patent RF
№2529014, kl.N01j 61/06, H05B33/12 ot 27.09.14. Lampa vakuumnaya ul’trafioletovogo
diapazona spektra
65. I.V. Ekhmenina, E.P. Sheshin, N.N. Chadaev, Problemy sozdaniya ul’trafioletovyh
istochnikov na osnove nanostrukturirovannyh avtoehmissionnyh katodov. ZHurnal
«Nano- i mikrosistemnaya tekhnika» 2, str.39–45 (2010)
66. I.V. Ekhmenina, E.P. Sheshin, N.N. Chadaev, Istochniki izlucheniya na osnove nanostruk-
turirovannyh avtokatodov. Zhurnal «Nano- i mikrosistemnaya tekhnika» 4, str.45–48 (2010)
67. I.V. Ekhmenina, E.P. SHeshin, N.N. CHadaev, Avtoehmissionnyj istochnik ul’trafiole-
tovogo izucheniya s avtokatodom iz nanostrukturirovannogo uglerodnogo materiala. Vestn.
S.-Peterburg, un-ta. Ser.10. Prikl.matem. Inform.Proc.upr. 1, str.3–8 (2011)
68. I.V. Ekhmenina, E.P. SHeshin, Issledovanie vliyaniya razlichnyh faktorov na ehffektivnost’
katodolyuminescencii s cel’yu sozdaniya konkurentosposobnogo avtoehmissionnogo
istochnika izlucheniya N ZHurnal «Trudy Moskovskogo fiziko-tekhnicheskogo instituta»,
tom 5, 1(17), str.36–43 (2013)
69. I.V. Ekhmenina, E.P. SHeshin, Issledovanie harakteristik izlucheniya avtoehmissionnyh
lamp s avtokatodami iz nanostrukturirovannogo uglerodnogo materiala. Izvestiya vysshih
uchebnyh zavedenij. Seriya: Himiya i himicheskaya tekhnologiya. T.56(5), S.74–76 (2013)
70. I.V. Ekhmenina, E.P. Sheshin, N.N. Chadaev, Source of ultraviolet radiation with field
emission cathode made of nanostructured carbon materials.
Springer_Book_BWF_2164446_Carbon Nanomaterials in Clean Energy Hydrogen
System-II (2011), pp. 299–303
71. H.S. Kang, J.T. Han, Y.W. Jin, M.J. Bae, Y.J. Park, Field emission backlight unit, method of
driving the backlight unit, and method of manufacturing lower panel. Patent USA US
2008/0106221 of 8.05.2008 cl.315/334 (H01 J 19/24)
72. B.N. Lin, C.H. Fu, Field emission backlight unit and scanning driving method. Patent USA
US 2008/0100235 of 1.05.2008 cl.315–349 (H05B 37/02)
73. Y.J. Jung, J.H. Park, J.-S. Jeong, J.W. Nam, A.S. Berdinsky, J.B. Yoo, C.Y. Park,
Fabrication and characteristics of flat lamp with CNT based triode structure for back light
unit in LCD. Technical Digest of IVNC 2005, Oxford, UK (2005), pp. 202–203
74. S.H. Park, C.W. Baik, J.H. Lee, Y.W. Jin, Field emission type backlight unit and method of
manufacturing the same. Patent USA US 2007/0229003 4.10.2007 315/324 (H01J 1/02)
75. J.H. Choi, B.G. Song, M.J. Shin, A. Zoulkarneev, D.S. Chung, M.J. Bae, Field emission
backlight device and method of fabricating. Patent USA US 2005/0179363 of
18.08.2005 cl.313/497 (H01J 1/62)
76. A.S. Batyrin, N.N. Chadaev, E.P. Sheshin, N.A. Duzhev, Y.I. Tishin, Development of a
lamp of illumination for liquid crystal screens, in Proceedings of Conference on “Modern
Television” Moscow (2001), pp. 21–23
References 531
77. A.S. Baturin, N.N. Chadaev, M.Y. Leshukov, A.J. Trufanov, E.P. Sheshin,
Cathodo-luminescent dynamic backlighting for color liquid crystal displays, in
Proceeding of Conference on Displays and Vacuum Electronics, Garmisch-partenkirchen,
Germany (2004), pp. 321–323
78. J. Gorog, P.M. Ritt, Liquid crystal display having a field emission backlight. Patent PCT
WO 2008/002321of 03.01.2008 cl. (H01J 9/227)
79. M.Y. Leshukov, N.N. Chadaev, E.P. Sheshin, Three-colored cathode luminescent lamp for
systems of display of information, in Proceedings of Conference on “Modern Television”,
Moscow (2005), pp. 37–38
80. M.Y. Leshukov, N.N. Chadaev, A.S. Baturin, K.H. Nikolskiy, R.G. Chesov, Illumination of
liquid crystal screens cathode luminescent lamps with field emission cathodes from carbon
fibers, in Proceedings of Modern Television Conference, Moscow (2002), pp. 30–31
81. N. Abanshin, A. Ezhkov, P. Ivashkin, S. Kashyrin, A. Vrotov, Y. Timofeev, Cathode
luminescent radiators for display units of collective use. Electron. Compon. 4, 114–116 (2007)
82. A.I. Tryfanov, A.S. Batyrin, M.Y. Leshukov, N.N. Chadaev, E.P. Sheshin, Emission
characteristics of a light source with field emission cathode on the basis of a bunch carbon
fibers. Microsyst. Equip. 3, 32–35 (2004)
83. A.S. Leychenko, M.Y. Leshukov, N.V. Luparev, P.A. Starikov, N.N. Chadaev, E.P. Sheshin,
Element of the big video screen with field emission cathode luminescent light sources as
pixels, in Proceedings of “Modern Television” Conferences, Moscow (2007), pp. 45–46
84. A.V. Kudryashov, E.P. Sheshin, N.N. Chadaev, M.M. Kustikov, The element of a big size
screen based on light emitters with field emission cathodes, in Abstract of International
Vacuum Electron Sources Conference (IVESC), London (2008)
85. D.S. Strebkov, V.Z. Trubnikov, A.V. Pastuhov, E.P. SHeshin, N.N. CHadaev, Sistema
ehlektricheskogo osveshcheniya (varianty). Patent RF №2505 744, ot 27.01.2014, kl.
F2159/00
86. S. Groznov, A.S. Leychenko, E.P. Sheshin, A. Shyka, Flat display screens on a basis the
field emission cathodes. Chip News 7, 21–25 (2008)
87. K.R. Shoulders, Microelectronics using electron-beam-activated machining techniques, in
Advances in Computers, vz, (1961), pp. 135–197
88. S.A. Spindt, A thin film field emission cathode. J. Appl. Phys. 39(7), 3504–3505 (1968)
89. C.A. Spindt, J. Brodie, L. Humphrey, E.R. Westarber, Physical properties of thin film field
emission cathode with molybdenum cones. J. Appl. Phys. 47(12) (1976)
90. R. Meyer, Recent development on «microtips» display at LETI. Technical Digest of IVMC
1999, Hagahama, Sapan (1999), pp. 6–9
91. S. Jtoh, T. Watanabe, T. Yamaura, K. Yano, A challenge to field emission displays. Asia
Displays 95, 617–620 (1995)
92. S. Jtoh, Current status of the field emission display. ASET International Forum on Low
Power Displays, Shinagawa (2000), pp. 59–68
93. S. Stoh, M. Tanaka, T. Tonegawa, Development of field emission display. Technical Digest
of IVMC 2003, Osaka, Japan (2003), pp. 19–20
94. B.E. Russ, J. Barger, J. Wang, Field emission cathode structure using perforated gate.
Patent USA US 2003/0193297 of 16.10.2003, cl.315/169 (H05B 39/04)
95. S. Kubota, K. Kikuchi, H. Sata, Cold cathode field emission device, process for the
production there of, and cold cathode field emission display. Patent USA US 2002/0050776,
of 2.05.2002, cl.313/309 (H01J 9/02)
96. Y. Kamide, S. Kubota, H. Sata, K. Kikuchi, Cathode panel for a cold cathode field emission
display, and cold cathode field emission display, and method of producing cathode panel for
a cold cathode field emission display. Patent USA US 6917155 of 15.07.2005, cl.313/495
(H01J 1/62)
97. Y. Kamide, S. Kubota, H. Sata, K. Kikuchi, Cathode panel for a cold cathode field emission
display and cold cathode field emission display, and method of producing cathode panel for
a cold cathode field emission display. Patent USA US 2005/0236964 of 27.10.2005,
cl.313/495(H01J 1/62)
532 8 Field Emission Cathode-Based Devices and Equipment
98. M. Konishi, K. Jida, Cold cathode electric field electron emission display device.
Patent USA US 2006/0087248 of 27.04.2006, cl.315/169, 3 (G09G 3/10)
99. E. Negishi, Method for treating a cathode panel, cold cathode field emission display device,
and method for producing the same. Patent USA US 2008/0012467 of 17.01.2008,
cl.313/495 (H01J 1/62)
100. M. Konishi, Cold cathode electric field electron emission display device. Patent USA US
2005/0082964 of 21.04.2005, cl.313/497 (H01J 31/12)
101. D. Sarrasin, Method of driving a matrix display device having an electron source with
reduced capacitive consumption. Patent USA, 8477156B2, 2.07.2013, cl.345/690
(G09G5/10)
102. M. Yamamoto, K. Koga, A. Shiota, S. Kanemaru, M. Nagao, Field emission electron source.
Patent USA US 2005/0001536 of 6.01.2005, cl.313/497, (H01J 1/02)
103. S.-T. Yan, Field emission display and method for manufacturing same. Patent USA US
2008/0074031 of 27.03.2008 cl.313/496, (H01J 63/04)
104. G.-L. Chen, Field emission display device. Patent USA US 6646282 of 11.11.2003,
cl.257/10 (H01L 29/06)
105. K.S. Choi, S.J. Lee, J.M. Kim et al., FED devices containing a novel graphite cathode
prepared by a screen printing process, in Digest of 12th JVMC, Darmstadt, Germany (1999),
pp. 32–33
106. K.S. Choi, S.J. Lee, J.M. Kim et al., Field emission display devices containing a novel
graphite cathode prepared by a screen printing process. Digest SID 00 (2000), pp. 671–673
107. W.B. Choi, D.S. Chung, J.H. Kang, H.Y. Kim et al., Fully sealed, hight-brightness
carbon-nanotube field emission display. Appl. Phys. Lett. 75(20), 3129–3131 (1999)
108. J.M. Kim, W.B. Choi, N.S. Lee, J.E. Jung, Field emission from carbon nanotubes for
displays. Diam. Relat. Mater. 9, 1184–1189 (2000)
109. Q.H. Wang, A.A. Setlur, J.M. Lauerhaas, J.Y. Dai, E.W. Seelig, R.P.H. Chang, A
nanotube-based field emission flat panel display. Appl. Phys. Lett. 72(22), 2912–2913
(1998)
110. Y. Nakayama, S. Akita, Field emission device with carbon nanotubes for a flat panel display.
Synth. Metals 117, 207–210 (2001)
111. C.C. Kuo, W.S. Hsu, C.Y. Hsao, Sintering method for carbon nanotube cathode of field
emission display. Patent USA US 2006/0009110 of 12.01.2006. cl.445/50 (H01J 9/12)
112. Y.H. Chien, C.P. Peng, C.H. Fu, W.Y. Lin, L.H. Chan, Field emission system and method
for improving its vacuum. Patent USA US 2008/0042547 of 21.02.2008, cl.313/495 (H01J
63/04)
113. E.P. Sheshin, A.M. Kyrnosov, Flat field emission device to display the image. Patent PF ru
2178598 from 20.01.2002 CL.H01J 31/12
114. A.S. Baturin, N.N. Chadaev, K.N. Nikolski, R.G. Tchesov, E.P. Sheshin, I.N. Yeskin,
Application of cathode materials for field mission display, in Proceedings of Conference
Displays and Vacuum Electronics, Garmisch-partenkirchen, Germany (2004)
115. A.Y. Tcherepanov, A.G. Chakhovskoi, V.B. Sharov, Flat panel display prototype using
low-voltage carbon field emitters. J. Vac. Sci. Technol. B 13(2), C.482–486 (1995)
116. Q.H. Wang, M. Yan, R.P.H. Chang, Flat panel display prototype using gated carbon
nanotube field emitters. Appl. Phys. Lett. 78(9), 1294–1296 (2001)
117. L. Yukui, Z. Changchum, L. Xinghui, Field emission display with carbon nanotubes cathode
prepared by a screen-printing process. Diam. Relat. Mater. 11, 1845–1847 (2002)
118. B.E. Russ, J. Barger, Field emission display using line cathode structure. Patent USA US
2003/0193296 of 16.10.2003 cl.315/169.3 (G09 G3/10)
119. A. Okamoto, K. Konuma, Y. Tomihari, F. Ito, Y. Okada, CNT film and field emission cold
cathode comprising the same. Patent USA US 2004/0104660 of 3.01.2004 cl.313/346 R
(H01j 19/06)
References 533
120. W.B. Choi, Y.W. Jin, H.Y. Kim, S.J. Lee et al., Electrophoresis deposition of carbon
nanotubes for triode-type field emission display. Appl. Phys. Lett. 78(11), 1547–1549 (2001)
121. M. Nakamoto, Field emission type cold cathode device, manufacturing method thereof and
vacuum micro device. Patent USA US 2004/0265592 of 30.12.2004 cl.428/408 (B32B
15/04)
122. X. Xu, C.P. Beetz, G.R. Brandes, R.W. Boerstler, J.W. Steinbeck, Carbon fiber-based field
emission devices. Patent USA, №5973444 of 26.10.1999 cl.313/309 (H01J 1/30)
123. M.A. Gruillom, M.L. Simpson, V.J. Merkulov, A.V. Melechko, D.H. Lowndes, Gated
fabrication of nanostructure field emission cathode material within a device. Patent USA US
6858455 of 22.02.2005. cl.438/20 (H011 21/00)
124. Z. Zheng, S.-S. Fan, Method for manufacturing cathode assembly of field emission display.
Patent USA US 2007/0287350 of 13.12.2007 cl.445/24 (H0iJ 9/00)
125. H.J. Kim, J.T. Han, Y.C. Choi, K.S. Jeong, Method of forming a carbon nanotube structure
and method of manufacturing field emission device using the method of forming a carbon
nanotube structure. Patent USA US 2008/0003733 of 3.01.2008 cl.438/197 (H01L 21/8234)
126. M. Muroyama, T. Yagi, K. Jnoue, S. Saito, Electron emitter and its production method.
Patent USA US 2004/0108515 of 10.06.2004, Cl.257/144 (H01L 21/00)
127. S. Kang, C. Bae, J. Kim, Barrier metal layer for a carbon nanotube flat panel display.
Patent WO 2004/064099 of 29.07.2004
128. A.A. Blyablin, A.T. Rakhimov, V.A. Samorodov, N.V. Suetin, Method for producing an
addressable field emission cathode and an associated display structure. Patent USA US
2003/0143321 of 31.07.2003. Cl 427/77 (B 05 D5/12)
129. Z. Sun, B.K. Tay, S.P. Lau, Y. Li, Field emission device and method of fabricating same.
Patent USA US 2005/0077811 of 14.04.2005 Cl.313/495(H01j 1/62)
130. M. Okai, T. Muneyoshi, T. Yaguchi, N. Hayashi, Flat panel display. Patent USA US
2006/0049743 of 09.03.2006, cl. 313/496 (H01j 1/62)
131. F. Ito, Field emission type cold cathode and method for manufacturing the same and method
for manufacturing flat display. Patent USA US 2003/0080663 of 1.05.2003. CL.313/336
(H01j 1/16)
132. H. Lee, S. Lee, S. Park, Triode structure field emission display device using carbon
nanotubes and method of fabricating the same. Patent USA US 2005/040752 of 24.02.2005,
Cl.313/495 (H01j 1/02)
133. K.A. Dean, B.F. Coll, E.M. Howard, L.L. Tisinger, Method for reducing leakage current in a
vacuum field emission display. Patent USA US 2007/0097567 of 03.05.07, cl.361/42
(H02H9/08)
134. W. Gu, W. Lei, X. Zhang, Influence of the dielectric layer on the performance FED.
Technical Digest of IVNC, Oxford, UK (2005), pp. 330–331
135. S.V. Johnson, Discharge of a field emission display based on charge accumulation.
Patent USA US 2008/0048570 of 28.02.2008, cl.315/169.3 (G09 G3/10)
136. M.J. Youh, C.L. Tseng, Y.H.J. Huang, M. Liu, Triode field emission cold cathode device
with random distribution and method. Patent USA US 2005/0104506 of 19.05.2005,
cl.313/496 (Ho1j 1/62)
137. N.S. Xu, Z.S. Wu, S.Z. Deng, J. Chen, High-voltage triode flat-panel display using field
emission nanotube-based thin films. J. Vac. Sci. Technol. B19(4), 1370–1372 (2001)
138. T. Yaguchi, T. Muneyoshi, M. Okai, N. Hayashi, T. Nakamura, Flat panel display device.
Patent USA US 2005/0057178 of 17.03.2005 of 315/169.4 (G 09 G 3/10)
139. T. Oh, Field emission display. Europatent EP 1542258 of 15.06.2005 cl.H01j 31/12
140. K. Oono, Field emission display device and method of manufacturing same. Patent USA US
2004/0239235 of 2.12.2004, cl.313/497 (H01j 1/62)
141. P.S. Ahn, H.W. Lee, Field emission device with focusing control electrode and field
emission display. Patent USA US 2005/0189868 of 1.09.2005. Cl 313/497 (H01j 1/02)
142. J.T. Han, Field emission display and method of manufacturing the same. Patent USA US
2005/0110393 of 26.05.2005, cl.313/495 (H01j 1/62)
534 8 Field Emission Cathode-Based Devices and Equipment
143. Y.H. Song, J.H. Lee, K.Y. Kang, Field emission device and field emission display device
using the same. Patent USA US 2006/0290259 of 28.12.2006 cl.313/495 (H01j 63/04)
144. Y.H. Song, J.H. Lee, C.S. Hwang, Field emission display. Patent USA 2005/0248256 of
10.11.2005, cl 313/495 (H01j 1/62)
145. M. Nakamoto, Field emission cold cathode device of lateral type. Patent USA US
2004/0183421 of 23.09.2004 cl.313/309, (H01j 1/62)
146. M. Nakamoto, Field emission cold cathode device of lateral type. Patent USA US
2006/0061257 of 23.03.2006. Cl.313/496 (H01j 63/04)
147. Y. Takeuchi, T. Nanataki, J. Ohwada, T. Horiuchi, Electron-emitting device and field
emission display using the same. Patent USA US 7088049 of 8.08.2006 cl.315/169.1 (G 09
G3/10)
148. H. Kawamura, S. Kafo, T. Maki, T. Kobayashi, Fabrication of planar diamond electron
emitters for flat panel displays. Mat. Res. Soc. Symp. Proc. 558, 155–160 (2000)
149. T.S. Oh, Field emission display. Patent USA US 2004/0222734 of 11.11.2004 cl.313/497
(H01j 1/62)
150. H. Moon, Field emission display device. Patent USA US 2005/0093424 of
5.05.2005 cl.313/495 (H01j 1/62)
151. J.H. Kang, Y.S. Choi, W.B. Choi, N.S. Lee et al., Under-gate triode type field emission
displays with carbon nanotube emitters. Mat. Res. Soc. Symp. Proc. 621, P.R 5.2.1–R 5.2.5
(2000)
152. Y.S. Choi, J.H. Kang, Y.J. Park, W.B. Choi et al., An under-gate structure field emission
display with carbon nanotube emitters. Diam. Relat. Mater. 10, 1705–1708 (2001)
153. Y.S. Choi, J.H. Kang, H.Y. Kim, B.G. Lee et l, A simple structure and fabrication of
carbon-nanotube field emission display. Appl. Surf. Sci. 221, 370–374 (2004)
154. V.P. Mammana, G.E. Mc Guire, O.A. Shenderova, Back-gated field emission electron
source. Patent USA, US 2005/0116214 of 02.06.2005 cl.257/10 (H01 L 29/06)
155. S.H. Ahn, S.J. Lee, Field emission display and method of manufacturing the same.
Patent USA US 2005/0093426 of 05.05.2005 cl.313/496 (H01j 1/62)
156. S.H. Moon, Field emission display device. Patent USA US 2005/019 4880 of
08.09.2005 cl.313/311 (Ho 0j 1/02)
157. M. Taniguchi, M. Kitada, K. Nakamura, S. Kawata, Field emission element. Patent USA US
7067971 of 27.06.2006 cl.313/497 (H01j 1/62)
158. A. Kastalsky, S. Shokhor, F.J. Disanto, D.A. Krusos, B. Gorfinkel, N. Abanshin, Pixel
structure for an edge-emitter field emission display. Patent USA US 2007/0046165 of
1.03.2007.cl.313/309 (H01j 1/02)
159. B.I. Gorfinkel, N.P. Abanshin, Cathodoluminescent screen and its manufacturing method.
Patent PF RU 2152662 of 10.07.2000. cl.H01j 1/62
160. N.P. Abanshin, B.J. Gorfinkel, Thin-film planar edge-emitter field emission flat panel
display. Patent USA 6 590320 of 08.07.2003 cl.313/309 (H01j 1/02)
161. L.D. Karpov, Patent USA №5 965971 of 12.10.1999, cl.313/309 (H01j 01/02), Pat USA №6
023126 of 08.02.2000, cl.313/310 (H01j 1/05)
162. K.L. Jensen, Field emitter arrays for plasma and microwave source applications. Phys.
Plasmas 6(5), 2241–2253 (1999)
163. S.P. Morev, N.P. Abanshin, B.J. Gorfinkel, A.N. Darmaev, D.A. Komarov, A.E. Makeev, A.
N. Yakunin, Electron-optical systems with planar field emission cathode matrices for
high-power microwave devices. J. Commun. Techn. Electron. 58(4), 357–365 (2013)
164. J.P. Calame, H.F. Gray, J.L. Shaw, Analysis and design of microwave amplifiers. J. Appl.
Phys. 73(3), 1485–1504 (1993)
165. A.G. Rozhnev, N.M. Ryskin, D.V. Sokolov, D.J. Trubetskov, S.T. Han, J.I. Kim, G.S. Park,
Novel concepts of vacuum microelectronic microwave devices with field emitter cathode
arrays. Phys. Plasmas 2(2), 4020–4027 (2002)
166. J.C. She, O.H. Ha, N.S. Xu, S.Z. Deng, S. Chen, S.E. Hug, L. Wang, Arrays of vacuum
microdiodes using uniform diamondlike carbon tip apexes. Appl. Phys. Lett. 89, 233518
(2006)
References 535
167. P.A. Ananiev, Y.V. Andrianov, V.A. Bashilov, M.M. Bogatskiy, D.A. Borisov, L.A.
Kontonistov, G.N. Fursey, HFF-amplifier. Patent USSR №1072145, of 07.02.84. cl.H01j
25/00
168. V.I. Mahov, B.V. Bondarenko, M.F. Kopylov, SVCH-devices M-type. Patent RF
№2040821 of 27.07.1995 cl.H01j 1/30
169. N.M. Ryskin, S.T. Han, K.H. Jang, G.S. Park, Theory of the microelectronic traveling wave
klystron amplifier with field emission cathode array. Phys. Plasmas, 14, 093106 (2007)
170. J.E. Graebner, S. Jin, G.P. Kochanski, W. Zhu, Microwave vacuum tube devices employing
electron sources comprising activated ultrafine diamonds. Patent USA, №5796211 of
18.08.1998. cl.315/3.5 (H01j 23/04)
171. J. Garcia-Garcia, F. Martin, R.E. Miles, D.P. Streenson, J.M. Chamberian, J.R. Fletcher, J.R.
Thorpe, Parametric analysis of micromachined reflex klystrons for operation at millimeter
and submillimeter wavelengths. J. Appl. Phys. 92(11), 6900–6904 (2002)
172. T.C. Cunningham, Field emission X-ray tube having a graphite fabric cathode. Patent USA
№3883760, 313/55 (H01J 35/06) of 07.04.71
173. P.R. Schwoebel, Field emission arrays for medical X-ray imaging. Appl. Phys. Lett. 8,
113902 (2006)
174. C.A. Spindt, C.E. Holland, R.D. Stowell, Recent progress in low-voltage field emission
cathode. J. Phys. t.45(Col.C9, suppl. 12), C9-9–C9-278 (1984)
175. B. Diop, V.T. Binh, Quasi-monochomatic field emission X-ray source. Rev. Sci. Instrum.
83, 094704 (2012). doi:10.1063/1.4752406
176. A.N. Bodrov, B.P. Merkulov, Y.V. Nikolyukin, Impul’snaya rentgenovskaya trubka, rat. RF
№ 2521436, ot 27.06.2014. kl.N01j 35/02
177. K. Kawakita, K. Hata, H. Sato, Y. Saito, Development of micro-focused X-ray source by
using carbon nanotubes field emitter. Technical Digest 18th IVNC, Oxford, UK (2005),
pp. 192–193
178. Z. Liu, O.Z. Zhou, S. Lu, Micro-field emission X-ray sources and related methods.
Patent US 2008/0043920, of 21.02.2008, cl.378/138 (H01J 35/14)
179. J. Zhang, G. Yang, Y. Cheng, B. Gao, Q. Qiu, Y.Z. Lee, J.P. Lu, O. Zhou, Stationary
scanning X-ray source based on carbon nanotube field emitters. Appl. Phys. Lett. 86,184104
(2005)
180. W. Knapp, D. Schleubner, S. Bjeoumikhov, H. Wolff, N. Langhoff, X-ray sources with
carbon field emitter cathodes, in IVESC (2004), pp. 309–311
181. S.H. Heo, H.J. Kim, J.M. Ha, S.O. Cho, A vacuum-sealed miniature X-ray tube based on
carbon nanotube field emitters. Nanoscale Res. Lett. 7, 258–265 (2012). doi:10.1186/1556-
276x-7-258
182. Y. Sun, The design and fabrication of carbon-nanotube-based Field emission X-ray cathode
with ballast resistor. JEEE Trans. Electr. Dev. 60(1), 464–470 (2012)
183. J.W. Hwang, C.B. Mo, H.K. Jung, S. Ryu, S.H. Hong, Field emission behavior of carbon
nanotube yarn for micro-resolution X-ray tube cathode. J. Nanosci. Nanotechnol. 13(11),
7386–7390 (2013)
184. Y. Jwai, K. Muramatsu, S. Tsuboi, A. Jyouzuka, T. Nakamura, Y. Onizuka, H. Mimura,
X-ray tube using a graphene flower cloth field emission cathode. J. Appl. Phys. Express 6,
105102 (2013)
185. J.-W. Jeong, J.-T. Kang, S. Choi, J.-W. Kim, S. Ahn, Y.H. Song, Digital miniature X-ray
tube a high-density triode carbon nanotube field emitter. Appl. Phys. Lett. 102, 023504
(2013)
186. G. Phillip, K. Timothy, W. Scotta, A sealing cold cathode X-ray tube for use in small X-ray
sources is provided. Patent USA 2014 0226791, от 14.08.2014. кл.Н01j-035/06
187. E.P. Sheshin, N.N. Chadaev, A.S. Baturin, A.I. Tryfanov, X-ray tube cathodes. Patent RF N
8248643,cl.Ho1J 35/02 of 35/02 of 20.03.05
188. A.S. Baturin, A.I. Trufanov, N.N. Chadaev, E.P. Sheshin, Field emission gun for X-ray
tubes. Technical Digest of Display and Vacuum Electronics, Garmisch-Partenkirchen
(2004), pp. 407–409
536 8 Field Emission Cathode-Based Devices and Equipment
189. A.S. Baturin, A.I. Trufanov, N.N. Chadaev, E.P. Sheshin, Field emission gun for X-ray
tubes. Nuclear Instrum. Methods Phys. Res. A, 558, 253–255 (2006)
190. J.W. Jeong, Y.-H. Song, J. Kang, J.-W. Kim, Field emission X-ray tube and method of
operating the same. Patent US 2012/0148027, от 14.06.2012, кл.378/106 (Н05G 1/22)
191. A.S. Bugaev, P.A. Eroshkin, V.A. Romanko, E.P. Sheshin, Low-power X-ray tubes the
current state. Physics-Uspekhi 56(7), 691–703 (2013)
192. A.V. Crewe, Electron gun using a field emission source. Rev. Sci. Jnsx. 39(4), 576 (1968)
193. G.N. Fursey, S.A. Shahirova, To the question of possible localization of the bodily field
emission in small angles-ZhTF t.33(6), 1125 (1966)
194. S. Hosoki, S. Yamamoto, H. Todokoro, S. Kawasa, Y. Hirai, Field emission cathode and
method of fabricating the same. Patent USA N 4379250, cl. 313/336 of 05.04.1983
195. L.W. Swanson, L.C. Grouser, Angular confinement of field electron and ion emission.
J. Appl. Phys. 40(12), 4741 (1969)
196. Z.I. Dranova, V.A. Ksenofontov, V.B. Kylko, B.G. Lazarev, L.S. Lazarev, I.M.
Mihaylovskiy, A method of manufacturing the tip emission field emitter with localized
emission. USSR №828261 cl H01 J9/02 of 07.05.1981
197. V.A. Kyznetsov, B.N. Vasichev, Y.L. Rybakov, Field emitter with localized emission a.s.
USSR №1069029 cl.Ho1J 1/30, 23.01.1984
198. T.A. Bakeer, M.M. Balsiger, K.T. Considine, H.E. Litsjo, Electron-beam tube including a
thermonic field emission cathode for a scanning electron microscope. Patent №3809899
(USA)
199. B. Cho, K. Shigeru, C. Oshima, W(310) cold-field emission characteristics reflecting the
vacuum states of an extreme high vacuum electron gun. Rev. Sci. Instrum. 84, 013305
(2013)
200. E.A. Litvinov, A.A. Starobinec, The limiting currents of field emission. ZhTF 47(10), c.2032
(1977)
201. A.D. Karpenko, Y.S. Lobachev, V.A. Shishkin, On the limiting current of field emission
electronic cathode tip. Radio-electronics t.21(1), s.207 (1976)
202. M.I. Elinson, V.A. Gorkov, A.A. Yasnopolskaya, G.A. Kydinceva, A investigation of
pulsed field emission at high density of currents. Radio Eng. Electron. 5(8), 1318 (1960)
203. J.P. Barbour, W.W. Dolan, J.K. Trolan, E.E. Martin, W.P. Dake, Space-charge effects in
field emission
204. Jeol. J. Electron Microscope
205. http://www.hitachi-hitec.com
206. Patent N3191028 (USA), Scanning Electron Microscope/Crewe A.V
207. J.W. Butler, Digital computer techniques in electron microscopy, in 6th International
Congress Electron Microscopy (Kyoto) (1966), p. 191
208. M. Okai, K. Hidaka, M. Hayashibara, S. Watanabe, Field emission electron gun and electron
beam applied device using the same. Patent USA US 2008/0029700 of 07.02.2008, 250/310
(Ho1J 37/073)
209. D.I. Swan, D. Kynastou, The development of a field Emission SEM, in Proceedings of the
6th Annual Scanning Electron Symposium, N 1 (1973), p. 57
210. K. Kuroda, T. Suzuki, High current efficiency accelerating lens system of field emission
scanning electron microscope. J. Appl. Phys. 46, N1 (1975)
211. K. Kuroda, H. Ebisui, T. Suzuki, Three-anode accelerating lens system for the field emission
scanning electron microscope. J. Appl. Phys. 45(N5) (1974)
212. R.J. Taylor, P.J. Swan, An experimental scanning electron microscope. In: Proceedings of
the Scanning Electron Microscope System and Application Symposium (1973), p. 36
213. H.P. Kue, B.M. Siegel, A field emission probe forming system with magnetic pre-accelerator
lens. In: Proceedings of Electron Microscope Society of America, 34th Annual Meeting
(1976)
214. A.V. Crewe, Electron microscope using field emission source. Surf. Sci. 48(N1), 152–160
(1975)
References 537
215. R.I. Garber, A.K. Malik, I.M. Fishman, Use of needle emitters in electronic microscopy.
PTE 4, 188–190 (1974)
216. A.K. Malik, I.M. Fishman, S.M. Hazan, A.c.№389564 (USSR). Cathodic unit for electronic
gun. №29 (1973)
217. Pat.№1426509 (USSR), Improvements in field emission electron guns
218. Y. Sakitani, Patent №4019077 (USA). Field emission gun
219. T.A. Baker, M.M. Balsiger, K.T. Considine, H.E. Litsjo, Separable-chamber electron-beam
tube including means for puncturing. Pat.№3881125 (USA)
220. L.M. Welter, Field emission gun improvement. Pat №3946268 (USA)
221. V.A. Vasin, V.I. Zaporozhchenko, V.I. Rahovskiy, A.c. №594540 (SSSR). The device for
stabilization of current of an field emission source, №7 (1978)
222. S. Nomura, Patent №3786268 (USA). Electron gun device of field emission type
223. K.C.A. Smith, I.R.A. Cleaver, Electron gun. Patent №3872351 (USA)
224. T. Someya, M. Kobayshi, T. Goto, Feldemissions-Electronenguelle. Patent №2221138
(FRG)
225. R. Aihara, S. Ota, N. Kabayshi, Field emission type electron gun. Patent №3810025 (USA)
226. V.J. Coates, L.M. Welter, Patent №1355365 (England). Improvements in and relating to
electron gun
227. Siemens atteiengesellschaft Co., Electron-beam microanalyzer apparatus. Patent №1389119
(England)
228. R.V. Latham, D.A. Wilson, The development of s high-definition cathoderay tube using a
carbon-fibre field emission electron source. J. Phys. E: Sci. Instrum. 15(10), 1083–1092
(1982)
229. B.V. Bondarenko, V.A. Siliverstov, E.P. Sheshin, Use of field emission cathodes from
carbon fibers in an electronic gun. Theses of Reports of the 5th Symposium to Not Heating
Cathodes, Tomsk, pp. 81–83 (1985)
230. V.I. Tihtilo, E.P. Sheshin, Electrovacuum device. A.c. 1345935 SSSR, Cl.Ho1J 1/30 of
21.10.1985
231. F. Houdeller, L. Knoop, G. Gatel, A. Masseboeuf, S. Manishin, Y. Taniguchi, M. Delmas,
M. Monthioux, M.J. Hytch, E. Snoeck, Development of TEM and SEM high brightness
electron guns using cold-field emission from a carbon nanotip. Ultramicroscopy 151, 107–
115 (2015)
232. Y.A. Grigoriev, A.I. Petrosya, V.V. Penzyakov, V.G. Pimenov, V.I. Rogovin, V.I.
Shesterkin, V.P. Kudryashov, V.C. Semyonov, Experimental study of matrix carbon field
emission cathodes and computer aided design of electron guns for microwave power
devices, exploring these cathodes. J. Vac. Sci. Technol. B v 15(2), 503–506 (1997)
233. A.S. Baturin, I.N. Yeskin, A.I. Trufanov, N.N. Chadaev, E.P. Sheshin, R.G. Tchesov,
Electron gun with field emission cathode of carbon fiber bundle. J. Vac. Sci. Technol. B 21
(1), 1–4 (2003)
234. B.V. Bondarenko, V.I. Makyha, E.P. Sheshin, Y.V. Kydrashov, V.N. Ylacyuk, Y.V.
Petryshenko, Field emission electronic gun. A.c. 1294188 SSSR,Cl.H01J 1/30 of 11.03.85
235. A.G. Chakhovskoi, A.V. Galdetski, A.N. Obraztsov, A.P. Volkov, Experimental investi-
gation of electron gun wih nanostructured carbon cathode. Technical Digest of IVNC,
Oxford, UK (2005), pp. 204–205
236. N.N. Dzbanovskiy, P.V. Minakov, A.A. Pilevskiy, A.T. Rahimov, B.V. Seleznev, N.V.
Syetin, A.Y. Yuriev. High current electronic gun on the basis of an field emission cathode
and a diamond grid. Zh.T.F. t.75(10), 111–114 (2005)
237. V.S. Nazarov, A.G. Gryshnikova, M.B. Silaev, O.G. Varnakova et al., A.c. 490211 SSSR,
Cl. H01J 41/00 of 4.11.72. Ion getter pump/
238. A.c. 1240279 SSSR, Cl H01 J 41/16 of 12.07.84. Ion getter pump/ Varnakova O.G.,
Grinchenko V.T., Konovalov N.D., Pogibelskay N.B
239. A.V. Anashenko, A.S. Batyrin, I.N. Eskin, E.P. Sheshin, A.A. Shyka et al., Field emission
cathodes for heaters of getter-ion pumps. Electron. Ind. 3−4, 80–81 (1998)
538 8 Field Emission Cathode-Based Devices and Equipment
240. V.A. Antonov, Y.A. Bykovskiy, A.I. Larkin, V.N. Ylasuk, A.V. Shelakov, E.P. Sheshin,
Light-valve equipment. a.s. SSSR №1609371 of 24.06.1988, Cl. H01J 31/24
241. Y. Sugiyama, J. Itoh, S. Kanemaru, Vacuum magnetic sensor with comb-shaped field emitter
arrays, in The 7th International Conference on Solid-State Sensors and Actuators,
Yokohama, Japan (1993)
242. D. Hong, M. Aslam, Diamond field emitter pressure sensor. Technical Digest of JVMC,
Oregon, USA (1995), pp. 335–337
243. T. Asano, R. Kajiwara, Fabrication of a tunnel sensor with cantilever structure. Technical
Digest of JVNC, Nagahama, Japan (1991), pp. 204–205
244. J.P. Hollingsworth, P.R. Banlaru, Carbon nanotube based nonvolatile memory. Appl. Phys.
Lett. 87, 2333115 (2005)
245. B. Lojek, Solid state field emission charge storage. Patent USA US 2008/0105946 of
2.05.2008.Cl.257/505 (H01L 29/00)
246. B. Lojek, Solid state field emission charge storage. Int. Pat. WO 2008/051675 of 02.05.2008
Cl.H01 L21/336
247. S.S. Furkay, D.V. Horak, C.H. Lam, H.-S.P. Wong, Field emission phase change diode
memory. Patent USA US 2005/0127350 of 16.06.2005, cl.257/4 (H01L 47/00)
248. C. Goodman, Neutron sources. Patent USA №2816242 of 10.12.57 cl.313/61
249. A. Otuka, M. Yoshino, C. Ohima, Application a Spindt emitter to an ionization gauge for an
extremely high vacuum. Tecnical Digest of IVMC, Nagahama, Japan (1991), 202–203
250. B.E. Barrington, A.L. Floras, W.L. Lees, Field-ionization electrodes. Patent USA, 3562881,
of 16.02.71, cl.29/25.18 (H01j 9/16)
251. M. Faubel, W. Hobler, J. Toennies, Field emission cathode. Patent England №2021854 of
5.12.79, cl.H01j 1/30
252. M. Faubel, W.M. Hobler, J.P. Toennies, Electron impact ion source with field emission
cathode. Patent USA №4272699 of 9.06.1981 cl.313/309 (H01j 37/073)
253. P.J. Traynor, R.G. Wright, Carbon nanotube electron ionization sources. Patent RST WO
2005/048290 of 26.05.2005, cl.H01j 49/00
254. O.A. Velikodnaya, V.A. Gurin, V.A. Ksenofontov, I.M. Mihaylovskiy, E.V. Sadanov et al.,
Multiemitter field source of ions on a basis the nanostructure of carbon materials. Letters
ZhTF t.33(13), c.90–94 (2007)
255. G.F. Byrchak, S.P. Chervonobrodov, E.P. Sheshin, Method of forming of a light stream on
the external screen for full-color system of display of a video information and the equipment
for its implementation. Patent RF.RU 2265964 of 05.08.2003, cl.H04 №9/30
256. R.Z. Bahtizin, S.S. Goc, R.F. Zaripov, R.R. Faizov, Noise generator. a.s. USSR №1157642
of 23.05.1985, cl.H03 B 29/00
257. Y.M. Wong, W.P. Kang, J.L. Davidson, V.K. Choi, D.V. Kerns, J.H. Huang, Design and
fabrication of single-chip carbon nanotubes vacuum field emission differential amplifier.
Technical Digest of IVNC, Oxford, UK (2005), pp. 35–36
258. A. Zoulkarneev, J.-H. Choi, Field emission RF amplifier. Patent USA US 2005/0184675 of
25.08.2005 cl.G3/10 (H01j 315/169.3)
259. B.J. Kampherbeek, M.J.-J. Wieland, P. Kruit, Emission photo cathode array for lithography
system and lithography system provided with such and array. Patent USA.US 2003/0178583
of 25.09.2003., cl.250/492.3 (H01j257/10)
260. D.A. Shiffler, Method of making a field emission cold cathode. Patent USA 2004/0202779
of 14.10.2004, cl.427/78 (B05D 5/12)
261. K.W. Cheng, Mirror having a field emission information display. Patent USA US
2008/0012727 of 17.01.2008, cl.340/933 (G08G 1/01)
262. F.G. Rudenauer, Field emission devices for space applications. Surf. Interface Anal. 39,
116–122 (2007)
263. V. Mironov, Fundamentals of the scanning probe microscopy. M. Technosphere (2004)
Conclusion (Prospective Development)
Abstract As previously noted, this book is at its core the corrected edition of a
Russian-language publication from the year 2011. However, although the chapters
1 to 8 of the translated edition are only partially updated and corrected, the con-
clusion chapter was written from scratch. This is mostly due to the fact that in last
few years appeared a significant number of original research papers on the book's
topic that requires a new analysis. Without reviewing them the authors would
consider their mission incomplete.
Given that field emission has indisputable advantages as compared with other free
electron sources, there’s little doubt that work for adapting different devices to
practical usage is to continue. This book lists a number of prospective fields of use
and branches of development for different field emission cathodes and devices based
on them, but below we’d like to mention again some of the most important of them.
Those fields are shown as a diagram on Fig. 1. The directions of development
can be subdivided into three basic categories: further development of traditional
technologies, introduction of new materials and nanotechnology. Different types of
treatment of field cathodes, increasing the homogeneity of electron current and
focusing of electron beam can be used for each of those categories. Naturally, all
these methods are aimed at practical use of field cathodes in various electronic
devices. Qualitative leap in development of field cathodes can be achieved by both
developing new methods and further development of traditional technologies at the
new level of understanding of physical processes.
Prospects of development of
field emission cathodes
Practical applications
of field emission cathodes
different from ones obtained at the end of preceding work stage: the image
brightness increases and the previous image seems more “grainy” i.e. its details
look larger. The latter can perhaps be explained by emission from different areas of
the surface of polyacrylonitrile fiber.
According to the model of adsorption-desorption processes described above, the
first and the third cases cannot be considered as separate, as changes of barrier
transparency due to adsorption phenomena occur not so much at the microtips
themselves as in pores, holes, capillary etc, i.e. those areas where there was possibly
no emission at all before. Thus, along with reduction of the potential barrier,
increase of the emitting area also occurs at this areas.
Figure 2 shows the transitional characteristics of a sample annealed at temper-
ature of 2600 °C for different adsorption times. Analysis of these characteristics
leads us to conclusion that the dynamic equilibrium between adsorption and des-
orption in absence of ion bombardment of the cathode (and thus with filling of
capillary) is achieved in about 10 h.
From Fig. 2 one can conclude that a rather large decrease of field emission
current occurs in first minutes or even seconds of cathode function. This can
probably be explained with that that intensive ion bombardment rips off atoms and
molecules that are adsorbed close to emission centers (microtips) and thus have
small adsorption energy. As a result, close to the tips the work function grows and
the current decreases abruptly. Further removal of the adsorbed atom layer goes
significantly slower. This is caused, firstly, by a stronger bond of adsorbed atoms
and molecules of the residual gas at the surface of polyacrylonitrile fiber and,
secondly, the fact that the adsorbed molecules come to the surface out of capillary
and replace ones removed with ion bombardment. Speed if this process is defined
by diffusion speed and thus has the same order of magnitude. Those processes go
on until the dynamic equilibrium between adsorption and desorption is achieved.
Additionally, the analysis of electron characteristics shows weak dependence of
desorption speed on the intensity of ion bombardment of the field electron surface.
Those dependencies are common for all carbon materials. They lead to a
practical conclusion: field cathodes out of carbon materials produce larger current in
impulse mode. Additionally this regime is characterized with increased stability of
field emission current and longevity of the cathode.
As noted above, the increase of overall level of current taken from field electron
cathodes can be achieved by increasing the total emitting surface of cathodes in
case they work in parallel, i.e. with cumulative impact of a large number of separate
emitters. Nevertheless simple increase of emitters quantity does not proportionally
increase of emission current due to large variance of emitter curves in terms if
height and radius as well as because of mutual screening of different emitters.
Increase of working voltage leads to increase of current out of several particular and
the sharpest emitters, and eventually to their evaporation or explosion. This in its
turn leads to emergence of an ark between the anode and cathode that spoils the
vacuum in the interelectrode gap, which leads to further ark etc.
Thus field cathode systems use limiting resistance. Those can come in form of
external resistors or in form of constructive layers with relatively high resistance.
Conclusion (Prospective Development) 543
This resistor serves as negative feedback, which stabilized the field emission
current and doesn’t allow disruptions.
With all this in mind we can conclude the optimal construction of field cathodes
is one that involves mutually insulated fragments each with individual limiting
resistors. This technique allows, among else, to straighten the currents of individual
fragments. Finally, there is a way that can be employed already at the current stage.
It involves separation of a large field cathodes into small insulated fragments,
serially to each of which a transistor is added to play a role of controlled limiting
resistor. In this case one can achieve near-perfect homogeneity of emission current
along the field cathode surface. The same transistors can be used to control the
emission current.
New Materials
Fig. 4 Structure of the composite field emission cathode on the basis of carbon nanotubes: 1 glass
substrate; 2 silver paste; 3 carbon nanotubes
Nanotechnology
In the last few years the problems of low-field electron emission out of nanos-
tructured materials attract growing attention of researchers all over the world.
Low-field emission occurs out of nano-scale conductive structures (s.c. nano-
phase of a substance) surrounded with insulating phase or vacuum. High emission
properties of such nano-objects are defined not only by geometric factors of electric
field growth but also by lowered potential barrier for electron tunneling into vac-
uum out of those areas. The latter is connected to one of the fundamental properties
of a solid body, namely the difference of physical characteristics of nanoparticles
(electric, magnetic, optical, acoustic etc) from “macroscopic” properties of the
substance.
Conclusion (Prospective Development) 545
When quantum-scale effects start to play a dominating role the energy levels of
electrons of the substance become discrete and emitted electrons have rather narrow
energy spectrum. By estimation this usually occurs when nano-object has size
5 nm. When distance between nanoemitters is comparable to the length of
coherency mutual influence of nanoemitters and interference of emitted electron
beams should manifest themselves (at distance of 10 nm or matrix density of
approximately 1012 elements/cm2). Those are so-called quantum points [14].
It is important to note that “self-arranging” carbon nanostructures (e.g.
nanocrystal diamond films) that are to a greater extent non-ordered systems for-
mation of nano-objects on any particular area of surface depends on a number of
uncontrollable factors. This means that creation of electron nanoemitters with given
parameters in this case is all but impossible.
It seems advisable and perspective to solve the problem of creation of
nanoemitters that can serve as monoenergetic electron sources and matrix of
nanoemitters based on ordered carbon nanostructures using a new approach
including directed formation of carbon nano-objects with different given sizes,
compositions, configurations and locations on the substrate.
In current time one can name two technological techniques that can implement
such nanostructures. The first of them is forming nanostructures using nano-scale
modification (i.e. nanophase transformations) of carbon materials. Works in this
field are performed all over the world. In terms of several criteria (composition,
sizes of nano-objects, electric properties) the following types of nanostructures with
pre-defined topological structures are the closest to nanoemitters:
• carbon nanostructures on substrate surface
• nanostructures in diamond and diamond-like films
• nanostructures in polymer films
One of the principal problems of this approach is guarantying stable and re-
producible properties of the created nanostructures. From the point of creation of
nanoemitters, this problem can by formulated based on the following criterif: the
formed body needs to be reliably chemically bonded with the substrate, it must be
conductive and be able to increase its mass by forming new chemical bonds.
The second technological approach is based on growing carbon nanotubes
(CNT) on the substrate surface [15, 16]. Formation of CNT (single or multi-wall)
arrays is an important problem studied in many of leading research centers and
private companies. There are test samples of flat screen displays based on CNT
arrays, in which the elements of cathode matrix include CNTs positioned perpen-
dicular to the substrate surface. The maximum achieved matrix density is about 106
elements per cm2. The main problem with creation of CNT-based nanoemitters is in
creating metals catalytic platforms (Ni, Fe, Co) with sizes of several nanometers for
growth of singular quantum-scale objects i.e. single-wall CNTs on substrate
surface.
Scanning probe nanolithography (SPL) seems to be the most efficient method of
forming a new type of field electron cathodes. SPL is one of the principal fields of
modern nanotechnology methods and is beyond any competition for the task of
546 Conclusion (Prospective Development)
Research of field emission fro graphenes warrant particular interest as the most
thin film field cathode. Despite that this area of research is relatively new [19, 20],
the volume of study is steadily increasing and one can reasonably hope that it would
lead to a new class of field cathodes.
Almost in every chapter of this book while describing field cathodes we mentioned
various methods of treatment of cathodes after they are already produced. This is
particularly applicable to cathodes out of carbon materials. Methods of treatment
include bombardment with ions of residual gases, plasma, thermal treatment etc.
Paper [21] proposed a method of cathodes by method of radiation.
The cathodes were prepared out of flat billets made of highly resilient fine-grain
graphite MPG-6. There were billets with round and rectangular cross-section with
diameters and sides, respectively, of 1 cm and thickness of 3 mm. After being cut
the billets were positioned onto the target of ion injector. Different samples were
bombarded with ions of Ar+, Xe+ and La+ in technical vacuum (about 10–4 Torr).
Other varied values included ion energies (50, 100 or 200 KeV), radiant exposures
(1016, 1017, 1018 or 1019 ions/cm2 with roughly constant radiation intensity at about
1014 ions/cm2 s) and angles of impact of the beam (angles between the beam axis
and billet surface were 90, 65 or 45°).
As shown by profiling of the manufactured samples, the most developed surface
containing the most potential emission centers is formed at radiant exposures of
about 1017–1018 ions/cm2 and angle of impact of ion beam of 45° to the surface.
Resulting density of emission centers was over 108 cm−1 and their curvature varied
between 10 and 100 nm. There is another notable fact: one can expect that this
548 Conclusion (Prospective Development)
method of field cathode production one can pick a mode of radiation that would
reduce height variance of emission centers to no more that 100–1000 nm. This
would allow move the cathode much closer to the fluorescent surface thus con-
siderably reducing working voltage, up to several hundred Volts and less.
Let us also note that in order to assess emission parameters all the produced
cathodes were installed into the vacuum system regardless of CTM image of their
surface. In process of measuring the volt-ampere characteristics of the cathodes
(with cathode-anode distance kept unchanged at about 5 mm) the previous con-
clusion on optimal parameters of radiation was confirmed. It were those cathodes
that allowed to obtain constant electron emission at voltages of 1–10 kV, while the
maximum current achieved the values of several mA. There is also a notable
dependence of cathode longevity on voltages at which they are used. Starting at
certain threshold voltage (usually around 5 kV) cathode lifetime dropped abruptly
from hundreds of hours to several minutes. A logical explanation for this phe-
nomenon is destruction of cathode emission centers due to high outbound current
(and thus heating of the cathode) and intensive ion bombardment of its surface.
Finally, measurements of cathode longevity and stability of emitted electron
current demonstrated that method of bombardment with heavy gas ions shows more
promise for field cathode production than bombardment with ions of metals. This
may be connected to the fact that in the former case we are dealing with emission of
pure carbon materials while in the latter one the carbon (graphite) is partially
metalized and implanted with metals.
This example shows that treatment of field cathodes has an important and often
critical role in defining the functional properties of the whole electronic device.
Thus, development of principles and modes of cathode treatment is a very
promising field.
For further increase of characteristics use of two such place with opposite ori-
entation of holes was propose (Fig. 7). In this case the controlling field may be
reduced to 1.5 V/lm. Besides, energy homogeneity of electron current is also
increased which leads to a considerably higher homogeneity of electron current
along the field cathode surface.
Development of this approach led to invention of a system of secondary-electron
electrodes for a flat field emission cathode out of carbon nanotubes [24, 25].
Construction of an element of a flat display is presented on Fig. 8. It’s based on a
funnel-like hole in a glass plate (4) acting like a secondary emission cathode. The
second insulation layer (7) serves for focusing the electron beam. Electronic sys-
tems similar to electron multipliers are proposed in [26, 27].
Fig. 9 The scheme of the display with metal secondary emission an amplifier: 1 resistive layer of
the field emission cathode; 2 substrate; 3 cathodic conclusion; 4 insulator; 5 control electrodes;
6 metal plates with big coefficient of secondary emission; 7 insulator; 8 masking layer; 9 layer
ITO; 10 anode plate; 11 phosphor
Figure 9 shows a schematic for one such device. The main distinguishing feature
is use of several electrons made on ferro-nickel or silver-magnesium alloys that are
known to have a high secondary emission coefficient [27]. The other options for the
coating materials include beryllium oxide, strontium oxide, barium oxide and other
metal oxides having high secondary emission coefficients [25]. The other substance
with high secondary emission is diamond [28].
Using this effect it is also possible to develop highly efficient microelectronic
vacuum devices.
Such devices can be based on polycrystalline diamond membranes able to
increase the electron beam of given distribution in a plane perpendicular to the
direction of electron spreading.
One of possible methods of increase of secondary emission is based on use of
porous membranes. In this case the secondary electrons passing through the surface
as in “reflection” scheme pass through the opposite surface through the pores due to
influence of the external field. Given that energies of secondary electrons is small
they can be “pulled in” into membrane pores. Of course, the multiplicative coef-
ficient Kn is considerably smaller that in “reflection” scheme as certain part of
Conclusion (Prospective Development) 551
primary electrons passes through the pores without taking part in birthing of sec-
ondary electrons. In this case Kn = (1 − Sn/S)KD, where Sn and S are total area
covered by pores and total work area of the membrane, respectively; KD is the
coefficient of secondary emission in “reflection” mode.
In order to test this idea experimentally we assessed energies of the secondary
electrons and created porous membranes. Direct measurements show that the
energy of secondary electrons does not exceed 3–5 eV.
We’ve prepared porous membranes with pores of 8 lm and distance between
them of also 8 lm (Fig. 10).
The proposed construction has quite serious practical applications. It all but
negates the dependence of amplification coefficient on membrane thickness and
allows to work in “shoot-through” mode with small energies of primary electrons.
The problem of focusing for field electrons is still not completely solved due to
great initial energies of electrons and thus dispersion in terms of speeds and
directions.
Reduction of the electron speeds e.g. via transforming them into secondary
emission ones allows to significantly reduce their energies and thus helps to more
efficiently focus them.
This book lists several rather promising constructions of focusing systems.
In our opinion the most efficient system for field electron focusing is reducing
their energies to the order of kT i.e. to the thermoelectron levels. When this is
achieved, it is possible to use already well-developed focusing systems.
However, the technologies of focusing field electrons with their initial energies
are by no means irrelevant or redundant. For an instance, [29] proposes a flat
552 Conclusion (Prospective Development)
system of turning off the field electron beam and [30] presents a field electron
display scanning the beam. In the latter case the focusing system consists of three
lenses.
The current state of affairs in the science of field electron emission can be char-
acterized as a some transition stage, or according to Forbes, reconstruction phase
[31–33]. Under this representation, the development of fundamental theory of field
electron emission can be viewed as consisting of several phases. Main theoretical
breakthroughs happened in the Fowler–Nordheim phase (1928 to mid-1950s), when
the theory of electron tunneling was initially formulated, and Murphy-Hood phase
(mid-1950s to mid-1990s), when a Fowler–Nordheim-type expression based on
Schottky-Nordheim barrier theory was introduced, which in turn led to develop-
ment of theory of Schottky emission (i.e. emission from the barrier top; later Jensen
[34] would build upon this idea and develop a general theory of field and photo-
electron emission) [33]. The current stage mostly consists of attempts at system-
atization and collating of data to make the mainstream science clearer and to
develop effective theory of specialized field electron emission topics.
This is evidenced by the character of activities of the international associations
and conferences on the topic. The International Field Emission Society holds
regular symposiums (IFES) since 1952. The main focus of attention has since
shifted from electron field emission towards field-ion microscopy and atom-probe
tomography, as well as their various applications. The Society and the symposiums
have always been in a sense practice-oriented. That is exactly why the symposiums
have demonstrated a wide interest in fundamental science of both electron and ion
field emission, surface effects of strong electric field, as well as in development of
new methods and apparatuses based on those effects, understanding of their prin-
ciples of function and their practical use. In the previous years proceedings of the
symposiums included a lot of papers on new sources of field electron emission and
related scientific fundamentals. Other consistent topics of discussions within the
Society included liquid-metal ion sources and many aspects of strong surface field
theory.
Lately the proceedings of the symposium (which, after 2002, takes place bien-
nially) are showing a noticeable decline in the amount of papers on electron field
emission: usage of keywords and references to sources having to do with field
electron emission in the papers of the last few years [35–37] is down about 85%
relative to the previous average. In 2014 the event was altogether renamed into the
Conference on Atom Probe Tomography and Microscopy [37] and its proceedings
contained hardly any mention of field electron emission at all. One can say that the
fundamental theory of electron field emission is waiting for a new theoretical
breakthrough.
Conclusion (Prospective Development) 553
References
1. C.A. Spindt, Field Emitter Arrays for High-Current, High-Current Density, and High
Frequency Operation (Tech. Digest JVNC, Oxford, 2005), pp. 20–21
2. V.J. Shesterkin, Y.A. Grigoriev, The multiple Beam electron gun with cold field emission
cathode made from glassy carbon, in Proceedings of JVESC-JCEE (2014), p 250
3. A.L. Suvorov, E.P. Sheshin, V.P. Babaev, Influence of adsorption of residual gases on
emission properties carbon properties. J. Technol. Phys. 66(9), 164–169 (1996)
4. A.V. Karabutov, V.D. Frolov, V.J. Konov, V.G. Ralchenko, S.K. Gordeev, P.J. Belobrov,
Low-field electron emission of diamond/pyrocarbon composites. J. Vac. Sci. Tech. B. 19(3),
965–970 (2001)
5. A.V. Karabutov, V.I. Konov, V.G. Ralchenko, V.D. Frolov, S.K. Gordeev, Advanced
diamond/sp2-bonded carbon nanocomposite materials as low threshold field electron
emitters. New Diam. Front. Carbon Technol. 11(4), 265–275 (2001)
6. S.K. Gordeev, A.I. Kosarev, A.N. Andronov, A.Y. Vinogradov, Field emitter of electrons.
Patent R.F. Ru 2149477 of 20.05.2000, cl.Ho1j 1/02, 1/30
7. S.K. Gordeev, V.G. Ralchenko, M.A. Negodaev, A.V. Karabytov, P.I. Belobrov, Field
emitter of electrons and way of its production. Patent RF Ru 2150154, of 27.05.2009, cl.
Ho1j 1/02
8. G.Z. Otarashvili, A.S. Rautov, E.P. Sheshin, Diamond-carbon nanocomposites as material
for cold cathode planar-type emitters. Technology Digest of JVNC (2011), p. 156
9. E.P. Sheshin, A.L. Shornikova, The field emission properties of different carbon-contained
powders, in Proceedings of JVESC-JCEE (2014), p. 252
10. R.B. Rakhi, A.I. Reddy, M.M. Shaijumon, K. Sethupathi, S. Ramaprabhu, Electron field
emitters based on multiwaled carbon nanotubes decorated with nanoscale metal clusters.
J. Nanopart Res. 10, 179–189 (2008)
11. K.-W. Cheng, G.-Y. Hsiao, J.-T. Hsu, Methode of fabricating cathode structure of field
emission display. Patent USA US 2006/0009111 of 12.01.2006, cl.445/51 (H01j 9/04)
12. R. Ryoo, S.H. Joo, M. Krur, M. Jaroniec, Ordered mesoporous carbons. Adv. Mater. 13(9),
677–681 (2001)
13. S.V. Lobanov, J.A. Fedorov, E.P. Sheshin, Researching dependencies of thermonic and field
emission characteristics of intercalated pyrographite cathodes from ingredients composition
and pressing conditions, in Proceedings of XXII conference Vacuum Science and Technology
(2015), pp. 208–210
14. R.-Z. Wang, W. Zhao, H. Wang, B. Wang, Y. Hui, Field emission from AlGaN/GaN
quantum-well nanotip arrays. Technology digest of JVNC (2011), pp. 100–101
Conclusion (Prospective Development) 555
15. P. Serbun, S. Putkowski, A. Navitski, N.T. Hong, S. Lee, Field emission from carbon
nanotube (CNT) arrays and triode test of single CNT columns. Technology digest of JVNC
(2011), pp. 38–39
16. F.-G. Zeng, H. Ma, V. Ru, H. Huo, F. Xiang, G. Wang, A configuration of substrate surface
for CNT cold cathode, in Proceedings of JVESC-JCEE (2014), p. 78
17. V.D. Frolov, V.L. Konov, S.M. Pimenov, V.J. Kuzkin, The low-dimensional effect in single
carbon-based nanoemitters of electrons. Appl. Phys. A 78, 21–23 (2004)
18. V.D. Frolov, E.V. Zavedeev, S.M. Pimenov, V.I. Konov, E.N. Loubnin, G.G. Kirpilenko,
Nanocones on (a-C:H): Si composite film: termal stability, growth dynamics and electrical
properties. Diam. Relat. Mater. 16, 1218–1221 (2007)
19. H. Tsurumi, T. Higuchi, Y. Yamada, M. Sasaki, Field emission from dense graphene edges.
Technology digest of JVNC (2011), pp. 226–227
20. Y. Zeng, Z. Chen, Z. Xie, J. Zhang, H. Li, Stable field emission from carbon nanotube grown
on grapheme substrate, in Proceedings of JVESC-JCEE (2014)
21. A.L. Suvorov, E.P. Sheshin, V.V. Protasenko, N.E. Lazarev, A.F. Bykov, V.P. Babaev, The
microrough flat field emission cathodes from graphite received in the radiation method.
ZhTF. T. 66(7), 156–160 (1996)
22. W. Lei, X. Zhang, X. Zhang, X. Zhou, Z. Zhu, C. Lou, Characteristic of the cold cathode
with secondary electron emission. Technology-digest IVESC (2004), pp. 52–54
23. W. Lei, X. Zhang, X. Zhou, Z. Zhu, C. Lou, H. Zhao, Characteristics of a cold cathode
electron source combined with secondary electron emission in a FED. Appl. Surf. Sci. 251,
170–176
24. X. Zhang, W. Lei, M. Liu, L. Zhang, D. Engelsen, X. Zhou, Q. Wang, Analysis of the
transverse energy distribution of HOPFED. Technology digest JVESC (2004), pp. 103–164
25. X. Zhang, W. Lei, M. Liu, L. Zhang, D. Engelsen, X. Zhou, Q. Wang, Analysis of the
transverse energy distribution of hopping electron through a glass funnel. Appl. Surf. Sci.
251, 182–190 (2005)
26. J.M. Chiou, J.L. Kwo, K.J. Chang, S.W. Lai, A.K. Li, H.J. Lai, M. Yokoyama, Field
emission display device. Patent USA US 2005/0099112 of 12.05.2005, cl.313/496 (H01J 21/
10)
27. T.H. Yang, J.M. Chiou, H.Y. Li, C.H. Lin, S.H. Lee, Field emission display device. Patent
USA 2007/0114910 of 24.05.2007, cl.313/495 (H01J 1/00)
28. S.A. Gavrilov, N.N. Dzbanovskiy, E.A. Ilichev, P.V. Minakov, E.A. Poltorackiy, G.S.
Rychkov, N.V. Syetin, Amplification of a flow of electrons by means of a diamond
membrane. ZhTF. T. 74(1), 108–114 (2004)
29. Z. Cui, Simulation of planar deflection system for field emission device. J. Vac. Sci. Technol.
B 19(3), 1077–1081 (2001)
30. H.S. Kim, B.J. Kim, Scanning field emission display. Patent WO 2006/016771 of
16.02.2006, cl.H01J 1/30
31. R.G. Forbes, Progress in establishing field electron emission science, in 10th International
Vacuum Electron Sources Conference, St Petersburg, June–July 2014. doi: 10.1109/
IVESC.2014.6891979
32. R.G. Forbes, Progress in field electron emission fundamentals, in 10th International Vacuum
Electron Sources Conference, St Petersburg, June–July 2014. doi: 10.13140/
RG.2.1.3846.8567
33. R.G. Forbes, Field electron emission theory: Present state and future needs, Lecture, St.
Petersburg, 9 Nov 2015. doi: 10.13140/RG.2.1.1736.2008
34. K.L. Jensen, in Advances in Imaging and Electron Physics, ed. by P.W. Hawkes. Electron
Emission Physics, vol 149 (Elsevier, Amsterdam, 2007)
35. IFES’2010, ed. by Y. Suchorski, G. Schmitz. in 52th International Field Emission Symposia,
Sydney, 2010. Ultramicroscopy, 111(6) (2011)
36. IFES’2012, ed. by B.D. Deconihout, G. Schmitz, G. Thompson. in 53th International Field
Emission Symposia, Tuscalossa, 2012. Ultramicroscopy, 132 (2013)
556 Conclusion (Prospective Development)
62. R.G. Forbes, Z. Li, Emission reference level: A missing concept in emission theory, in
IVNC’2011, 24th International Vacuum Nanoelectronics Conference, Wuppertal (2011),
6004592, p. 121
63. R.G. Forbes, Progress in reconstructing field emission theory to make it more transparent, in
IVNC’2015, 28th International Vacuum Nanoelectronics Conference, Guangzhou (2015),
7225514, p. 20
64. R.G. Forbes, A more scientific approach to describing Fowler-Nordheim theory, in
IVNC’2015, 28th International Vacuum Nanoelectronics Conference, Guangzhou (2015),
7225534, p. 66
65. R.G. Forbes, improved methods of extracting area-like information from CFE current-
voltage data, in IVNC’2015, 28th International Vacuum Nanoelectronics Conference,
Guangzhou (2015), 7225536, p. 70
66. R.G. Forbes, Comments on the voltage scaling of field electron emission current-voltage
characteristics, in IVNC’2014, 27th International Vacuum Nanoelectronics Conference,
Engelberg (2014), 6894781, p. 116
67. R.G. Forbes, Progress in establishing field electron emission science, in IVESC-ICEE’2014,
10th International Vacuum Electron Sources Conference, and 2nd International Conference
on Emission Electronics, ed. by N.V. Egorov, St. Petersburg, 6891979 (2014)
68. R.G. Forbes, Development of an integrated theory of field emitter optics, in BDO’2014, 20th
International Workshop on Beam Dynamics and Optimization, ed. by D.A. Ovsyannikov, St.
Petersburg, 6890017 (2014)
69. R.G. Forbes, Conceptual error in the international definition of voltage, and implications for
theories of patch fields and electron emission, in IVESC-ICEE’2014, 10th International
Vacuum Electron Sources Conference, and 2nd International Conference on Emission
Electronics, ed. by N.V. Egorov, St. Petersburg, 6891978 (2014)
70. R.G. Forbes, Use of a spreadsheet to test for lack of field emission orthodoxy, in IVESC-
ICEE’2014, 10th International Vacuum Electron Sources Conference, and 2nd International
Conference on Emission Electronics, ed. by N.V. Egorov, St. Petersburg, 6891980 (2014)
71. A.G. Kolosko, S.V. Filippov, P.A. Romanov, E.O. Popov, R.G. Forbes, Real-time
implementation of the “orthodoxy test” for conformity of current–voltage characteristics
with classical field electron emission theory. J. Vac. Sci. Tech. 34, 041802 (2016)
72. Sh-D Liang, Quantum Tunneling and Field Electron Emission Theories (World Scientific
Publishing, Singapore, 2014)
73. M.A. More, R.G. Forbes, in Encyclopedia of Nanotechnology, ed. by B. Bhushan. Field
Electron Emission from Nanomaterials (Springer Science Business Media Dordrecht, 2015)
Index
Boundary condition, 163, 215, 374, 375, 382, Chemosorbed layer, 104
388, 392, 414, 415 Chloroform, 280
Boundary problem, 375, 391 Chromaticity, 430
Boundary-value problem, 122, 400 Chromium, 259
Brightness, 357, 429, 456, 474 Classical diode, 446
Broken bonds, 188 Coating, 235, 442
Build-up, 345 Coating material, 260
Coaxiality, 446, 450
C Cold emission, 2
Cantilever, 548 Cold emitter, 22
Carbon-based, 296 Collector, 58, 59
Carbon-based material, 515 Color, 430
Carbon-carbon composites, 546 Color displays, 461
Carbon Composite, 546
fiber, 56, 298, 302, 332, 341, 434, 439, 445, Computer-controlled, 75
496 Concentration of impurities, 155
material, 81, 233, 297, 480, 544 Conditional optimization, 409
nanocrystals, 311 Conduction band, 21
nanofiber, 304 Conductive coating, 444
nanomaterials, 308 Cone-apex angle, 79
nanostructures, 306 Cone-shaped, 449
nitrogen, 304 Conformal mapping, 371
polyacrylonitrile, 465 Conical emitters, 485
powder, 331, 334 Contact resistances, 348
Carbon nanotube (CNT), 263, 282, 306, 309, Continuous mode, 441
313, 320, 325, 332, 342, 344, 353, 354, Control, 180
441, 452, 496, 547 electrode, 245, 263, 474
Cartesian coordinate, 411 function, 179
Catalytic growing, 304 signal, 359
Catalytic layer, 262 voltage, 431, 467
Cathode, 116, 142, 214, 352, 354, 373, 390 Controlling systems, 373
array, 461, 494 Copolymer, 283
coating, 477 Copper, 252, 257
current, 435 Corona discharge, 340, 341
degradation, 354 Crater, 323, 325, 326
design, 239 Criterion function, 409
diaphragm, 407 Cross-beam, 71
forming, 298 Crossing fields, 488
lifetime, 370 Crystal
parameters, 358 faces, 105
sputtering, 229, 234, 498 irregularities, 28
stability, 233 plates, 238
surface, 175, 196, 199, 414, 417 tips, 35
Cathodoluminescent, 427 Crystalline facets, 196
Cathodoluminescent lamp, 458 Crystalline lattice, 39
Center of gravity, 410 Crystalline plane, 118
Ceramic insulation, 73 Crystallization, 271
Charge-carrier concentration, 145 Crystallization method, 267
Charge density, 9, 370, 402 Crystallographic
Charge distribution, 122 axis, 323
Charging, 472 faces, 76
Chemical etching, 241, 257, 303 facets, 188, 194
Chemical polishing, 91, 98 planes, 235
Chemical vapor deposition, 311 Crystallography, 31
Index 561
U Y
Ultra-high vacuum, 67, 69, 72, 230, 247, 494, Yanoacrylate, 282
505 Young-Kuyatt criterion, 67
Ultrasonic cutting, 79
Ultraviolet, 472 Z
Ultraviolet lamps, 448 Zero-current approximation, 26
Ultra-violet photolithograph, 469
Uminescent screen, 436