TOMSK POLYTECHNIC UNIVERSITY

A. I. Pushkarev, Yu. I. Isakova, R. V. Sazonov,

G. E. Kholodnaya

PHYSICS AND CHEMISTRY OF

LOW-TEMPERATURE PLASMA
FOR MATERIAL TREATMENT

It is recommended for publishing as a study aid

by the Editorial Board of Tomsk Polytechnic University

Tomsk Polytechnic University Publishing House

2011

1
 UDC 000000
BBC 00000
E00

A. I. Pushkarev
E00 Physics and Chemistry of Low-Temperature Plasma for Material
Treatment/ A. I. Pushkarev, Yu. I. Isakova, R. V. Sazonov, G. E. Kholod-
naya; Tomsk Polytechnic University. – Tomsk: TPU Publishing House,
2011. – 181 p.

The book describes some physical and technical aspects of plasma chemistry. The
general concepts of plasma physics and chemical kinetics as well as the methods and
sources for generation of nonequilibrium plasma are considered. The application of low-
temperature plasmas in the chemical industry include: the synthesis and decomposition
of inorganic compounds; synthesis of nanosized materials; conversion of hydrocarbon
gases; synthesis of hydrogen and supramolecular carbon clusters (fullerenes); treatment
of membranes and coatings, etc.
This book is recommended for students, studding on the specialty "Physics and
technology of low temperature plasmas, plasma chemistry and plasma technology"
course 140200 " Electroenergetics".

UDC 000000
BBC 00000

Reviewer professor of Tomsk State University

Kohanenko A.P.

© STE HPT TPU, 2011

© A. I. Pushkarev, Yu. I. Isakova,
R. V. Sazonov, G. E. Kholodnaya, 2011
© Design. Tomsk Polytechnic University
Publishing House, 2011

2
 TABLE OF CONTENTS
Preface ..............................................................................................................7
Introduction .....................................................................................................8
Сhapter 1 Introduction to plasma physics .................................................12
1.1 What is a plasma ....................................................................................12
1.2 How are plasmas made ..........................................................................13
1.3 Some applications of plasmas................................................................13
1.4 Plasma parameters .................................................................................14
1.4.1 The Degree of Ionization .................................................................16
1.4.2 Plasma Temperature ........................................................................17
1.4.3 Plasma Frequency ............................................................................20
1.4.4 Debye Shielding ..............................................................................21
1.5 The plasma parameter: strongly and weakly coupled plasmas .............22
1.6 Plasma types ..........................................................................................24
1.7 Coalitional processes in plasma .............................................................26
1.7.1 Collision cross sections, mean-free paths and collision frequencies
..................................................................................................................26
1.8 Coronal equilibrium ...............................................................................28
1.9 Penetration of neutrals into plasmas ......................................................33
1.10 Collisions with neutrals and with charged particles: relative
importance ...................................................................................................36
Chapter 2. Introduction to gas discharge physics .....................................38
2.1 What Is the Subject of Gas Discharge Physics? ....................................38
2.2 Typical Discharges in a Constant Electric Field ...................................38
2.3 Brief History of Electric Discharge Research .......................................41
2.4 Mechanism of electrical discharges .......................................................43
2.4.1 Cathode processes ...........................................................................43
2.4.2 Photoelectric emission .....................................................................44
2.4.3 Thermionic emission .......................................................................45
2.4.4 Schottky effect .................................................................................45
2.4.5 Field emission ..................................................................................46
2.5 Electrical breakdown .............................................................................47
2.5.1 Electron avalanche ...........................................................................47
3
 2.5.2 Formation of a streamer...................................................................49
2.5.3 Propagation of the streamer discharges ...........................................52
2.6 Electrical breakdown in very small gaps – Townsend’s breakdown
mechanism ...................................................................................................54
2.6.1 Townsend’s experiment...................................................................54
2.6.2 Townsend’s theory of electrical breakdown ...................................56
2.6.3 Townsend’s electrical breakdown criterion ....................................60
2.6.4 Townsend’s mechanism in the presence of electron attachment ....61
2.7 Paschen’s law.........................................................................................62
2.7.1 Physical interpretation of the shape of the Paschen curve ..............64
2.7.2 Validity of Paschen’s law ................................................................65
2.8 Classification of Discharges ..................................................................65
Chapter 3 Basic principles of plasma chemical processes ........................68
3.1 Chemical kinetics...................................................................................69
3.1.1 Rates of reactions ............................................................................69
3.1.2 Dependence of rates on concentration.............................................72
3.1.3 Dependence of Rate on Temperature ..............................................78
3.2 The object and the main features of the plasma chemistry ...................81
3.2.1 Quasi-equilibrium plasma-chemical processes ...............................82
3.2.2 Non-equilibrium plasma-chemical processes..................................83
3.3 Elementary Plasma-Chemical Reactions ...............................................84
3.3.1 Ionization processes .........................................................................85
3.3.2 Classification of Ionization Processes .............................................86
3.3.3 Direct Ionization by Electron Impact ..............................................87
3.3.4 Stepwise ionization by electron impact ...........................................88
3.3.5 Ionization in collisions of heavy particles: adiabatic principle and
Massey parameter .....................................................................................89
3.3.6 Penning ionization effect and associative ionization ......................90
3.4 Elementary plasma-chemical reactions of positive ions .......................91
3.4.1 Different Mechanisms of Electron–Ion Recombination in Plasma.91
3.5 Elementary plasma-chemical reactions involving negative ions ..........92
3.5.1 Dissociative electron attachment to molecules as a major
mechanism of negative ion formation in electronegative molecular gases
..................................................................................................................92

4
 3.5.2 Three-body electron attachment and other mechanisms of formation
of negative ions .........................................................................................93
3.6 Special features of plasma-chemical reactions, their dynamics and
kinetics .........................................................................................................95
3.7 Initiation of plasma-chemical processes by electron-beams ...............102
Chapter 4 Chain gas-phase processes in the external action ..................105
4.1 Classification of chain processes .........................................................105
4.1.1 Unbranched chain reactions ..........................................................107
4.1.2 Chain reactions with quadratic branching .....................................108
4.1.3 Chain reactions with degenerate branching...................................108
4.1.4 Electron-induced chain reactions ..................................................109
4.2 Chain chemical processes under external action .................................109
4.2.1 Investigation of the induction period of ignition of the oxygen -
hydrogen mixture ....................................................................................109
4.2.2 Investigation of the displacement of the ignition limits of a
stoichiometric oxygen-hydrogen mixture under external action............115
4.3 Radiation-thermal cracking of methane ..............................................119
4.4 Chain oxidation of methane under external action ..............................122
4.4.1 Oxidation of methane in the equilibrium conditions at a low
pressure ...................................................................................................124
4.4.2 The initiation of methane oxidation by an external influence.......126
4.5 The conversion of carbon disulfide CS2 in the atmospheric air ..........129
Chapter 5. Applied plasma-chemistry ......................................................134
5.1 Nonequilibrium synthesis of nanosized oxide powders ......................134
5.1.1 Comparison of the available processes for production of nanosized
TiO2 particles ..........................................................................................134
5.1.2 An overview of the techniques to synthesize nanosized
(TiO2)x(SiO2)1-x .......................................................................................137
5.1.3 Non-equilibrium plasmochemical synthesis of nanosized particles
of metal oxides........................................................................................140
5.1.4 Research of composite nanosized oxides (TiO2)x(SiO2)1-x
synthesized using a non-equilibrium plasmochemical process ..............146
5.1.5 Synthesis of composite oxides Si-C-Ox.........................................153
5.2. Plasmochemical conversion of methane ............................................156
5
 5.2.1 Plasma pyrolysis of methane .........................................................156
5.2.2 Partial oxidation of methane ..........................................................159
Conclusion ...................................................................................................172
References ....................................................................................................176

6
 Preface

Plasma-chemical technologies are widely used in world practice, name-

ly, in large-tonnage chemistry; synthesis of powder-like materials, deposition
of coatings; semiconductor industry, microelectronics; in addressing the envi-
ronmental issues and scientific research. This book mainly focuses on the
plasma-chemical processes in which the non-equilibrium excitation of inter-
nal degrees of freedom of the reacting molecules makes a decisive influence
on the kinetics of chemical reactions. It is shown that for the initiation of
chemical reactions, the non-equilibrium excitation of vibrational degrees of
freedom of the molecules is the most effective. Plasma-chemical processes
occurring in these conditions have several advantages, which is important in
traditional industry, allowing for a reduced energy costs and increased
productivity. The conditions, occurring within the pulsed excitation of gas
mixtures, are also favorable for the organization of chain chemical processes.
The book starts with a chapter devoted to the general concepts of plasma
physics (describing a plasma Debye length, the ionization of plasma, etc.)
and chemical kinetics (rate and reaction order, rate constant, Arrhenius equa-
tion, etc.). The types of the internal energy of molecules in the quantum elec-
tronic state as well as the basic laws of non-equilibrium excitation of the
molecules are considered. The physical and chemical processes which are ac-
companied by high vibrational non-equilibrium reaction products, including
the excitation of molecules by electron impact are studied. The results of ex-
perimental studies of gas-phase chain processes under external influence are
also summarizes. Most of the studies of gas-phase chain processes under non-
equilibrium conditions were conducted by Russian scientists, who continued
a study of the Nobel laureate N.N. Semenov.

7
 Introduction

The modern large-tonnage chemical industry, which uses a tradition-

al approach of thermal activation of the chemical processes, is now faced
with the problem of energy conservation. Further development of the in-
dustrial base leads to an unjustified cost of resources to build the equip-
ment and the depletion of minerals, metals and fuels.
A natural way to help this situation, obviously, should be a transition
to new technological solutions in metallurgy, chemistry, energy and other
industries. The qualitative changes are noticed if the performance of
equipment will be improved, namely, the productivity per unit volume of
the reaction zone. This requires a significant increase in temperature in
the reaction zone, since this chemical process, within the framework of
classical kinetics, is accelerated exponentially in accordance with the Ar-
rhenius law. The heating of the reactor and the reactants to high tempera-
tures also requires increased energy consumption, and therefore needs the
new ways to increase productivity while specific energy consumption is
reduced.
The combination of the reaction zone with the discharge can locally
heat the reactants to high temperatures without heating the reactor walls,
which significantly decreases the energy consumption. These conditions
are easily realized by excitation of the reagent gas mixture by a continu-
ous electron beam or in arc discharge, etc. The reduction of the energy
barrier of the reaction is also achieved through participation of free radi-
cals and atoms, which are effectively produced in gas discharges due to a
dissociative recombination and relaxation of excited molecules.
The next step is to reduce the energy costs for conducting a chemical
process, which can be achieved by the following: using non-equilibrium
plasma-chemical processes, characterizing by a significant excess energy
of in the internal degrees of freedom of molecules (mainly vibrational)
compared with the thermodynamic equilibrium state. In this case, the gas
temperature may not exceed 300-400K, which reduces the energy loss in
heating the reactor walls, as well as the original components of the gas
mixture, which facilitates the hardening (stabilization) of the chemical
products.
Over the last 30-40 years in Russia and abroad many studies have
been devoted to the use of low-temperature plasma for gas-phase chemi-
cal processes. The formation of a low-temperature plasma by a pulsed
electron beam, in contrast to many other methods, is showed to occur
with a significant reduction in energy, consumed on the conversion of
gaseous compounds. An analysis of experimental works devoted to the

8
decomposition of impurities of different compounds (NO, NO2, SO2, CO,
CS2, etc.) in the air by a pulsed electron beam showed that the energy
used to decompose one molecule of the gas is lower than its dissociation
energy. This is due to the fact that some favorable conditions for the oc-
currence of chain processes are formed when an electron beam is in-
volved [1]. At low temperatures, when thermal initiation of the reaction
does not occur, some active centers are formed due to the influence of
plasma. These active centers are free radicals, ions or excited molecules,
which can start a chain reaction. This chain reaction occur at a tempera-
ture of 150-200 degrees below the temperature of a normal thermal pro-
cess, but with the same speed, since the plasma facilitates the most energy
intensive stage - thermal initiation of the reaction. With a sufficient length
of chain an electro physical set-up provides a small part of total energy
consumption to the chemical process. The main source of energy in this
case is the thermal energy of the original gas or exothermic energy of the
chemical reactions of chain process (e.g. oxidation or polymerization). It
is important to note that a chemical process, occurring at a temperature
below the equilibrium, allows one to synthesize compounds which are un-
stable at higher temperatures and the selectivity of the synthesis of which
at high temperatures is low. The increase in the temperature of chain
chemical processes with the radioactive impact is analogues to a catalytic
effect. But the chain process can entirely occur in the gas phase, which
greatly increases the reaction rate in comparison with the heterophase cat-
alytic process. A high speed of the reaction required for industrial tech-
nologies is achieved by the use of branched chain processes. However,
their major drawback is associated with the occurrence of an explosive
process, which significantly increases the risk of production. This disad-
vantage is eliminated by initiating of the chain process outside the scope
of self-ignition under external influence. The above features gas-phase
chemical processes with the electron beam involved indicate the prospects
of their use in large-scale chemical production [2]. Most studies of chain
gas-phase processes, including those under the external action, were per-
formed by Russian scientists, who continued the works of the Nobel lau-
reate N.N. Semenov. In his Nobel speech, he explained that the penetrat-
ing radiation will be used to initiate chain processes.
In plasma-chemical processes the energy transfer from an external
source (a source of electrical energy) to the reacting molecules occur in
several steps (see Fig. 1).

9
 Fig. 1. Stages of energy transfer in a gas discharge

In quasi-equilibrium plasma chemical processes, chemical transfor-

mations occur only after the transfer of the energy from an external
source to the kinetic energy of molecules of the reactive gas. In non-
equilibrium processes the chemical transformations begin after the excita-
tion of internal degrees and dissociation of these molecules.
Nowadays, plasma chemistry is one of the most promising directions
in low-temperature plasma physics and high energy chemistry. The ad-
vantages of plasma chemistry as a new type of chemical technology are
explained by a high level of energy introduced to a chemical system. This
leads to a significant increase in the rates of chemical transformations,
minimizing the size of a reactor, reduction of costs for the implementation
of the processes and organization of chemical production. The processing
of raw materials, which are poorly recyclable by conventional methods,
as well as various waste facilities in order to protect the environment, be-
came possible. In some cases the materials obtained by plasma-chemical
methods have unique physical and chemical properties.
The transition to plasma chemical technology means a transition to
higher levels of temperatures and higher velocities of chemical reactions.
In many cases, the rates of chemical reactions are compared with the rates
of transitions between the energy levels and even exceed it. In addition,
the plasma-chemical systems, used in the laboratories and technological
plants, are of an open type in the thermodynamic sense. An intense heat
and mass exchange with the environment and the energy exchange with
external fields (electric field, radiation field, etc.) are involved in plasma-
chemical reactions. This leads to the creation of the following: physical
non-equilibrium, temperature gradients, violation of the Maxwell- Boltz-
mann distribution of velocities and levels of the internal degrees of free-
dom. For example, in the plasmas of electrical discharge at low pressures
or high values of electric field strength, and low degrees of ionization (in
electric arcs, glow, high-frequency and microwave discharges at pressures
below atmospheric, as well as in the corona, barrier and pulse discharges),
the average electron energy exceeds the average energy of heavy particles
and the electron distribution function in the velocities differ from the

10
equilibrium Maxwellian. A non-equilibrium concentration of charged par-
ticles and the particles which are excited by different degrees of freedom -
rotational, vibrational and electronic levels is observed.
The object of plasma chemistry is a low-temperature plasma in mo-
lecular gases. Plasma chemistry studies the kinetics and mechanism of
chemical reactions and physical-chemical processes in low-temperature
plasmas. It should be emphasized that in plasma chemistry, the term
"plasma" is understood as a substance containing not only the charged
particles, but molecules in the ground electronic state with the excess
(non -equilibrium) internal energy.
The object of plasma chemistry as a scientific discipline is a study of
the relationship between physical and chemical phenomena in chemical
reactions in the plasma, the possibility of plasma usage for various prob-
lems in applied chemistry. As a consequence, there is a conventional divi-
sion into theoretical and applied plasma chemistry.
The terms "plasma chemistry", "plasma-chemical reactions" and
"plasma-chemical processes" were introduced in Russian scientific litera-
ture since the publication of the book entitled "The kinetics and thermo-
dynamics of chemical reactions in low-temperature plasmas", edited by
L.S. Polak (Moscow: Nauka, 1965). They reflect the fact that the object
of consideration is the specific chemical objects, which feature lies in the
fact that chemical reactions are initiated or occur in plasma. It has been
more than two hundred years since the existence of electrical discharge
chemistry as a science. It has come a long way from a phenomenological
description of phenomena to the creation of a new research area with its
own apparatus, including theoretical and experimental techniques. The
mechanisms of many plasma-chemical reactions have been studied and
understood. Besides the theoretical aspects of plasma-chemistry, there are
numerous applications of plasma-chemistry for the solution of practical
problems that are grouped under the title "applied plasma chemistry.

11
 Сhapter 1 Introduction to plasma physics

1.1 What is a plasma?

First and foremost, a plasma is an ionized gas. When a solid is heat-

ed sufficiently that the thermal motion of the atoms break the crystal lat-
tice structure apart, usually a liquid is formed. When a liquid is heated
enough that atoms vaporize off the surface faster than they recondense, a
gas is formed. When a gas is heated enough that the atoms collide with
each other and knock their electrons off in the process, a plasma is
formed: the so-called ‘fourth state of matter’. Taking into consideration
the energy of the particles constituting it, plasma is energetically the
fourth state of the matter, apart from the solid, liquid, and gas states. Fig.
2 presents schematically the ranges of temperature, or particle energy, in
which each of the four forms of matter occur in nature. For the plasma
state, the temperature range reflects only the energy of the heavy particles
(not of the electrons) for reasons to be explained later.

Fig. 2. State of matter versus temperature

The important point is that an ionized gas has unique properties. In

most materials the dynamics of motion are determined by forces between
near-neighbor regions of the material. In a plasma, charge separation be-
tween ions and electrons gives rise to electric fields, and charged particle
flows give rise to currents and magnetic fields. These fields result in ‘ac-
tion at a distance’, and a range of phenomena of startling complexity, of
considerable practical utility and sometimes of great beauty.

12
 Irving Langmuir, the Nobel laureate who pioneered the scientific
study of ionized gases, gave this new state of matter the name ‘plasma’.
The reader who is more interested in plasma physics is referred to an
excellent book written by Robert J. Goldston and Paul H. Rutherford [3]
and the lectures of Professor Richard Fitzpatrick (The University of Texas
at Austin) [4].

1.2 How are plasmas made?

A plasma is not usually made simply by heating up a container of

gas. The problem is that for the most part a container cannot be as hot as a
plasma needs to be in order to be ionized-or the container itself would va-
porize and become plasma as well.
Typically, in the laboratory, a small amount of gas is heated and ion-
ized by driving an electric current through it, or by shining radio waves
into it. Either the thermal capacity of the container is used to keep it from
getting hot enough to melt-let alone ionize during a short heating pulse, or
the container is actively cooled (for example with water) for longer-pulse
operation. Generally, these means of plasma formation give energy to free
electrons in the plasma directly, and then electron-atom collisions liberate
more electrons, and the process cascades until the desired degree of ioni-
zation is achieved. Sometimes the electrons end up quite a bit hotter than
the ions, since the electrons carry the electrical current or absorb the radio
waves.

1.3 Some applications of plasmas

There are all sorts of uses for plasmas. To give one example, if we
want to make a short-wavelength laser we need to generate a population
inversion in highly excited atomic states. Generally, gas lasers are
‘pumped’ into their lasing states by driving an electric current through the
gas, and using electron-atom collisions to excite the atoms. X-ray lasers
depend on collisional excitation of more energetic states of partially ion-
ized atoms in a plasma. Sometimes a magnetic field is used to hold the
plasma together long enough to create the highly ionized states. A whole
field of ‘plasma chemistry’ exists where the chemical processes that can
be accessed through highly excited atomic states are exploited. Plasma
etching and deposition in semiconductor technology is a very important
related enterprise. Plasmas used for these purposes are sometimes called
‘process plasmas’. Perhaps the most exciting application of plasmas such
is the production of power from thermonuclear fusion.

13
 1.4 Plasma parameters

A plasma, especially one sustained in a mixture of molecular gases,

contains a multitude of different neutral and charged particles. A group of
identical particles in a plasma is commonly referred to as a species.
The plasma is broadly characterized by the following basic parame-
ters:
1 The density of the neutral particles, nn.
2 The densities of the electrons and ions, ne and ni . In the quasi-
neutral state of plasma the densities of the electrons and of the ions are
usually equal, ni =ne =n (n is called the plasma density.
3 The energy distributions of the neutral particles, fn (W) ions, fi (W);
and electrons, fe(W).
The plasma density is an important parameter in plasma processing
because the efficiency of the processes occurring in the plasma and their
reaction rates are generally dependent directly on the density of the
charged particles. The electrons are the main factor responsible for the
transfer of the energy from the external electric field to the discharge gas.
Being electrically charged, both electrons and ions interact with the ap-
plied external electric field and are accelerated by absorbing energy from
it. Because the electrons are the lightest particles in the plasma, they are
easiest accelerated and absorb the largest amount of energy from the ex-
ternal field. The electrons then transfer through collisions energy to the
molecules of the gas and cause their ionization and dissociation. The ef-
fectiveness of these processes increases with increasing electron density.
Ions, too, play a significant role in the chemical reactions taking place in
the plasma. Many of the reactions occurring in a plasma are controlled, or
affected, by ion chemistry. It is therefore important to achieve high ion
densities to increase the rates of reactions involving the ions.
As in any gaseous system, particles in the plasma are in continuous
motion, inducing collisions between them. The collisions which take
place between the particles in the plasma are of two types, elastic or ine-
lastic. Collisions between electrons and heavy targets (i.e., neutral or
charged particles) that do not result in an excitation of the target are called
elastic collisions, whereas those collisions that leave the target in an ex-
cited state are called inelastic collisions. Collisional processed in plasma
will be discussed in more details in section 2.
The energy transfer WTr in an elastic collision between an electron
and a heavy target is determined by the mass ratio of the particles
2me
WTr  W (1)
M

14
 where M–mass of the heavy particle;
W–energy of the electron;
me–mass of electron.
For an elastic collision of an electron with an argon atom, the frac-
tion of transferred energy is therefore very small, about
WTr 1
 (2)
W 40000
On the other hand, a significant amount of energy is transferred in a
collision between two electrons.
The electrons gain energy through acceleration by the electric field,
which sustains the plasma and transfers that energy by inelastic collisions
with the neutral gas molecules. The inelastic collisions between energetic
electrons and the heavy species of the plasma result in excitation, ioniza-
tion, or dissociation of the target if it is multiatomic. Energy transfer in an
inelastic collision is not controlled by the mass ratio of the colliding parti-
cles. In an inelastic collision between two particles, the fraction of trans-
ferred energy is given by
WTr M
 (3)
W min  M
where min is the mass of particle losing energy.
According to Eq. (1.3), in an inelastic collision between an electron
and a heavy particle (min = me «M), the electron can transfer almost all its
energy to the heavy particle, creating an energetic plasma species. The in-
elastic collisions therefore sustain the plasma by producing the particles
that form it and giving the plasma its special features. Inelastic collisions
involve energy transfer in amounts that vary from less than 0.1 eV (for
rotational excitation of molecules) to more than 10 eV (for ionization).
Electron-electron collisions can also play a significant role in the en-
ergy transfer processes in the plasma. Their importance depends on the
degree of ionization prevalent in the plasma. For degrees of ionization be-
low 10-10 the contribution of the electron-electron collisions to the energy
transfer is negligible.
However, in electron cyclotron resonance (ECR) plasmas, where the
degree of ionization can be above 10 -3, electron-electron collisions domi-
nate[5].
The relative contribution of each type of collision to the processes
taking place in the plasma depends on additional plasma parameters,
which will be discussed next and which derive from the previously de-
scribed parameters.

15
 1.4.1 The Degree of Ionization

The parameter that defines the density of the charged particles in the
plasma is the degree of ionization of the gas. It specifies the fraction of
the particles in the gaseous phase which are ionized. The degree of ioniza-
tion, , is defined as
ni
 (4)
n
There are two basic processes of ionization, satisfying the conditions
for conservation of momentum and energy: (a) impact ionization, where
an electron strikes an atom, so that an ion and two electrons come off and
(b) radiative ionization, where a photon with sufficient energy (often in
the ultraviolet range) is absorbed by an atom, dissociating it into an ion
and an electron. Ions can recombine into atoms by the reverse of these
processes: (a) three - body recombination, where two electrons and an ion
join to make a neutral atom plus a free electron; and (b) radiative recom-
bination, where an electron and an ion combine into an atom, and a pho-
ton is emitted. These processes are illustrated in Fig. 3.

Fig. 3. Ionization and recombination processes: (a) electron-impact ioni-

zation, and (b) radiative recombination

The processes of three-body recombination and radiative ionization

are the inverse of these processes and are obtained by reversing the direc-
tion of the arrows in (a) and (b), respectively. Neutral atoms are repre-
sented by black dots, while electrons and protons are represented by open
circles labeled (-) and (+), respectively.
For plasmas sustained in low-pressure discharges, the degree of ioni-
zation is typically 10-6 to 10-3. However, if the electrical discharge is as-
sisted and confined by an additional magnetic field, the degree of ioniza-
tion can reach values of 10 -2 or higher, as for example, in an ECR plasma.
Table 1 presents the range of values of the degree of ionization encoun-
tered in different low-pressure plasmas used for processing of solids.

16
 Table 1. Ranges of Parameters for Various Low-Pressure Plasmas

Plasma Type Pressure Ion Density (cm-3) Degree of Ion-

(torr) ization
Deposition/etching < 10 < 1010 10-6
Reactive ion Etch- 10-2 – 10-1 1010 10-6 – 10-4
ing Magnetron sput- 10-3 1011 10-4 – 10-2
tering <10-4 – 10-2 1012 <10-1
Electron cyclotron
resonance

The degree of ionization in a plasma is a function of the elements

contained in the plasma. For example, in plasmas used in magnetron sput-
tering, the degree of ionization of the sputtered metal is higher than that
of the process gas employed for the sputtering.
The value of the critical ionization is defined by [6]
  1.73 1012   ва  Tв2 (5)
where a is the electron-atom collision cross section at the average
electron velocity, expressed in cm 2 ;Te is the electron temperature of the
plasma, expressed in eV. The electron temperature will be defined later,
in Sec. 1.4.2. If the degree of ionization is much bigger than the critical
ionization value, the charged particles behave as in a fully ionized gas.

1.4.2 Plasma Temperature

One of the physical parameters defining the state of a neutral gas in

thermodynamic quilibrium is its temperature, which represents the mean
translational energy of the molecules in the system. A plasma contains a
mixture of particles with different electric charges and masses. At a first
approximation, the plasma may be considered, thermally, as consisting of
two systems: the first containing only electrons and the second containing
the heavy species, that is, neutral atoms or molecules, ions, and neutral
molecular fragments.
The electrons gain energy from the electric field, which energizes
the plasma, and lose part of it by transfer to the second system through
elastic or inelastic collisions. The system of heavy particles loses energy
to the surroundings, either by radiation or by heat transfer to the walls of
the vessel containing the plasma.

17
 The electrons and the heavy species in the plasma can be considered
approximately as two subsystems, each in its own thermal quasi-
equilibrium.
The ions and electrons in the plasma can therefore be characterized
by their specific different average temperatures: the ion temperature, Ti
and the electron temperature, Te. Actually in some cases additional tem-
peratures may characterize the particles in the plasma. For example, in the
presence of a magnetic field, even a single plasma species, for example,
the ions, is characterized by two different temperatures, one representing
the translation of the ions parallel to the magnetic field, T||, and one repre-
senting the translation perpendicular to the magnetic field, . This is
caused by the fact that the forces acting on the species parallel to the
magnetic field are different from those acting perpendicular to it [7].
The situation is even more complicated, as the heavy species in the
plasma can be characterized by several temperatures at the same time,
even in the absence of a magnetic field [8]: the temperature of the gas, Tg,
which characterizes the translatory energy of the gas; the excitation tem-
perature, Tex, which characterizes the energy of the excited particles in the
plasma; the ionization temperature, Tion; the dissociation temperature, Td,
which characterize the energy of ionization and dissociation; and the radi-
ation temperature, Tr which characterizes the radiation energy. Thermo-
dynamic equilibrium will exist in the plasma only if the following equa-
tion is satisfied:
Tg  Tвх  Tion  Td  Tr  Tв (6)
Complete thermodynamic equilibrium cannot be achieved in the en-
tir9e plasma because the radiation temperature, Tr at the envelope of the
plasma cannot equal the temperature in the plasma bulk. However, under
certain laboratory conditions, it is possible to achieve local thermodynam-
ic equilibrium in plasma in volumes of order of the mean free path length
(The reader who is unfamiliar with the term is referred to the Sect. 2.1).
If this happens, the plasma is called a local thermodynamic equilibrium
(LTE) plasma. In low pressure plasmas, produced by direct current glow
discharge or radio frequency excitation, the LTE conditions are generally
not achieved. These plasmas are therefore called non-LTE plasmas.
In non-LTE plasmas the temperatures of the heavy particles are
normally too small to promote chemical reactions in thermodynamic equi-
librium. The electron temperature is therefore the most important temper-
ature in non-LTE plasmas, among all those different temperatures men-
tioned previously. The fraction of electrons that will cause the different
reactions in the plasma, the overall efficiency of the plasma processes,

18
and the processing rates increase with increasing electron temperature.
The electron temperature is discussed in further detail in the following
section.

1.4.2.1 Electron Temperature

The velocity distribution for a system of particles is defined as

the density of particles in the velocity space that satisfies the equation

n(cm3 )  4   f (v)v 2 dv (7)
0

where v is the velocity; f(v) is the velocity distribution function (den-

sity in velocity space); n is the the density of the particles in the geomet-
rical space.
If it is assumed that the velocity distribution of the electrons in the
plasma is isotropic, that the effects of inelastic collisions act only as a per-
turbation to the isotropy, and that the effects of the electric fields are neg-
ligible, then the velocity distribution is Maxwellian. The Maxwellian dis-
tribution assumes that the temperature of the electrons equals the tem-
perature of the gas, Te = Tg.
If the distribution of the electron velocities can be considered Max-
wellian, then it can be described by [3,9]
mв 3 2 m v2
f (v)  nв ( ) exp( в ) (8)
2kTв 2kTe
The electron energy distribution function is related to the ve-
locity distribution function f(v) through the relation [5]
4
f (W )  v  f (v ) (9)
mв
Therefore, the Maxwellian energy distribution function for the elec-
trons is given by
3 1   1.5W 
f (W )  2.07Wav 2 W 2
exp  (10)
 Wav 
where Wav is the average energy of electrons.
It can be shown that the average energy of the electrons is related to
their temperature by
3
Wav  kTв (11)
2
Due to the simplifying assumptions, the Maxwellian distribution
provides only a first approximation of the electron energy (or velocity)
distribution in the plasma. The assumptions made for the Maxwellian dis-

19
tribution can be replaced in low-pressure plasmas by the following as-
sumptions:
1 The electric field strength in the plasma is sufficiently low such
that one can neglect the inelastic collisions, but large enough for the elec-
tron temperature to be much higher than the ion temperature, Tв >> Ti
2 The electric field is of sufficiently low frequency, that is, it is of a
frequency ω much lower than the frequency of collisions .
3 The collision frequency is independent of the electron energy.

1.4.3 Plasma Frequency

The plasma frequency,

nв 2
 2
(12)
p
 0m
is the most fundamental time-scale in plasma physics. Clearly, there
is a different plasma frequency for each species. However, the relatively
fast electron frequency is, by far, the most important, and references to
“the plasma frequency” in text-books invariably mean the electron plasma
frequency.
It is easily seen that ωp corresponds to the typical electrostatic oscil-
lation frequency of a given species in response to a small charge separa-
tion. For instance, consider a one-dimensional situation in which a slab
consisting entirely of one charge species is displaced from its quasi-
neutral position by an infinitesimal distance δx. The resulting charge den-
sity which develops on the leading face of the slab is . An equal
and opposite charge density develops on the opposite face. The x-directed
electric field generated inside the slab is of magnitude

Ex    en x /  0 . Thus, Newton’s law applied to an individual particle
0
inside the slab yields
d 2x
m  eEx  m p2x (13)
dt 2

giving x  (x) 0 cos( p t )

Note that plasma oscillations will only be observed if the plasma
system is studied over time periods longer than the plasma period τ
≡1/ωp, and if external actions change the system at a rate no faster than
ωp. In the opposite case, one is clearly studying something other than
plasma physics (e.g., nuclear reactions), and the system cannot not useful-
ly be considered to be a plasma. Likewise, observations over length-

20
scales shorter than the distance vtτ p traveled by a typical plasma particle
during a plasma period will also not detect plasma behaviour. In this case,
particles will exit the system before completing a plasma oscillation. This
distance, which is the spatial equivalent to τp, is called the Debye length,
and takes the form
D  T (14)
1
m p
Note that
 0T
D  (15)
nв 2
is independent of mass, and therefore generally comparable for dif-
ferent species. Clearly, our idealized system can only usefully be consid-
ered to be a plasma provided that
D
<<1 (16)
L
p
<<1 (17)


Here, τ and L represent the typical time-scale and length-scale of the

process under investigation.
It should be noted that, despite the conventional requirement (16),
plasma physics is capable of considering structures on the Debye scale.
The most important example of this is the Debye sheath: i.e., the bounda-
ry layer which surrounds a plasma confined by a material surface.

1.4.4 Debye Shielding

Plasmas generally do not contain strong electric fields in their rest

frames. The shielding of an external electric field from the interior of a
plasma can be viewed as a result of high plasma conductivity: i.e., plasma
current generally flows freely enough to short out interior electric fields.
However, it is more useful to consider the shielding as a dielectric phe-
nomena: i.e., it is the polarization of the plasma medium, and the associ-
ated redistribution of space charge, which prevents penetration by an ex-
ternal electric field. Not surprisingly, the length-scale associated with
such shielding is the Debye length.
Let us consider the simplest possible example. Suppose that a quasi-
neutral plasma is sufficiently close to thermal equilibrium that its particle
densities are distributed according to the Maxwell-Boltzmann law,
n s  n0 e  в ф / T
s
(18)

21
 where Φ(r) is the electrostatic potential, and n0 and T are constant.
From ei= −ee= e, it is clear that quasi-neutrality requires the equilibrium
potential to be a constant. Suppose that this equilibrium potential is per-
turbed, by an amount δΦ, by a small, localized charge density pext . The
total perturbed charge density is written
 p  pexp  e(ni  nв )  pexp  2e 2 n0Ф / T (19)
Thus, Poisson’s equation yields

p  pexp  2e n0Ф / T  2

 Ф    
2  (20)
0   
 0 
which reduces to

 2 2  p
   2 Ф   exp (21)
 D  0
If the perturbing charge density actually consists of a point charge q,
located at the origin, so that , δpext=qδ(r) then the solution to the above
equation is written
q  2 r / D
Ф(r )  e (22)
4 0 r
Clearly, the Coulomb potential of the perturbing point charge q is
shielded on distance scales longer than the Debye length by a shielding
cloud of approximate radius consisting of charge of the opposite sign.
Note that the above argument, by treating n as a continuous function,
implicitly assumes that there are many particles in the shielding cloud.
Actually, Debye shielding remains statistically significant, and physical,
in the opposite limit in which the cloud is barely populated. In the latter
case, it is the probability of observing charged particles within a Debye
length of the perturbing charge which is modified.

1.5 The plasma parameter: strongly and weakly coupled plasmas

Let us define the average distance between particles,

1
rd  n 3 (23)
and the distance of closest approach,
e2
rd  (24)
4 0T
Recall that rc is the distance at which the Coulomb energy

22
 1 2 e2
U (r , v)  mv  (25)
2 4 0 r
of one charged particle in the electrostatic field of another vanishes. Thus,
U(rc,vt)= 0.
The significance of the ratio rd/rc is readily understood. When this
ratio is small, charged particles are dominated by one another’s electro-
static influence more or less continuously, and their kinetic energies are
small compared to the interaction potential energies. Such plasmas are
termed strongly coupled. On the other hand, when the ratio is large,
strong electrostatic interactions between individual particles are occasion-
al and relatively rare events. A typical particle is electrostatically influ-
enced by all of the other particles within its Debye sphere, but this inter-
action very rarely causes any sudden change in its motion. Such plasmas
are termed weakly coupled. It is possible to describe a weakly coupled
plasma using a standard Fokker-Planck equation (i.e., the same type of
equation as is conventionally used to describe a neutral gas). Understand-
ing the strongly coupled limit is far more difficult, and will not be at-
tempted in this course. Actually, a strongly coupled plasma has more in
common with a liquid than a conventional weakly coupled plasma.
Let us define the plasma parameter
  4n3D (26)
This dimensionless parameter is obviously equal to the typical num-
ber of particles contained in a Debye sphere. However, Eqs. (8), (16),
(17), and (19) can be combined to give
D 1 rd 3 2 4 03 / 2 T 3 / 2
  ( )   1/ 2 (27)
rC 4 rC в3 n
It can be seen that the case «1, in which the Debye sphere is
sparsely populated, corresponds to a strongly coupled plasma. Likewise,
the case » 1, in which the Debye sphere is densely populated, corre-
sponds to a weakly coupled plasma. It can also be appreciated, from Eq.
(1.28), that strongly coupled plasmas tend to be cold and dense, whereas
weakly coupled plasmas are diffuse and hot. Examples of strongly cou-
pled plasmas include soliddensity laser ablation plasmas, the very “cold”
(i.e., with kinetic temperatures similar to the ionization energy) plasmas
found in “high pressure” arc discharges, and the plasmas which constitute
the atmospheres of collapsed objects such as white dwarfs and neutron
stars. On the other hand, the hot diffuse plasmas typically encountered in
ionospheric physics, astrophysics, nuclear fusion, and space plasma phys-
ics are invariably weakly coupled.

23
 Table 2. Lists the key parameters for some typical weakly coupled plas-
mas

n(m-3) T(eV) ωp (sec-1) λD(m) λ

Interstellar 106 10-2 6×104 0.7 4×106
Solar Chromo- 1018 2 6×104 5×10-6 2×103
sphere
Solar Wind (1AU) 107 10 2×105 7 5×1010
Ionosphere 1012 0.1 6×107 2×10-3 1×105
Arc Discharge 1020 1 6×1011 7×10-7 5×102
Tokamak 1020 104 6×1011 7×10-5 4×108
Inertial 1028 104 6×1015 7×10-9 5×104
Confinement

In conclusion, characteristic collective plasma behaviour is only ob-

served on time-scales longer than the plasma period, and on length-scales
larger than the Debye length. The statistical character of this behaviour is
controlled by the plasma parameter. Although ωp, λD, and λ are the three
most fundamental plasma parameters, there are a number of other pa-
rameters which are worth mentioning.

1.6 Plasma types

The plasma state exists in natural form in the cosmos or is created

under unique conditions for specific purposes. The plasmas found in na-
ture cover a very large range of electron densities and temperatures. As
shown in Fig. 4, the plasma density, ne, spans the range between 1 and
10-20 cm-3 , while the electron temperature, Te can vary between 10 -2 and
105 eV.

24
 Fig. 4. Plasma types by electron density and temperature

Solar winds are a continuous stream of charged particles with

ne = 5 cm-3 and Te = 50 eV. The interstellar material contains a hydrogen
plasma with a density of 1 cm-3. The ionosphere, which extends approxi-
mately from 50 km upward from the earth's surface, is populated by a
weak plasma with a density varying up to 10 6 cm-3 and an electron tem-
perature of 0.1 eV, while the sun and stars have surface temperatures
ranging from 5000 to more than 70,000 °K (0.5 to 7 eV). They consist en-
tirely of plasma, the outer layer being partially ionized and the interior hot
enough to be completely ionized. The temperature at the center of the sun
is at about 2 keV.
In the quest for controlled thermonuclear fusion, it is necessary to
create plasmas with electron temperatures above 10 keV and with ion
densities of (1–2) x 1014 cm-3. These values are required in order to obtain
the nuclear fusion reaction between deuterium and tritium atoms, because
reasonable crosssections for the fusion reactions are obtained only for en-
ergies above 5 KeV. Central ions temperature of Ti « 35 KeV and electron
temperature of Te = 15 KeV have been reached in Tokomak fusion reac-
tors [10].
Taking into account the wide ranges of parameters, the plasmas are
classified into several categories:
– Plasmas in complete thermodynamic equilibrium (CTE plasmas).
CTE plasmas exist only in stars or during the short interval of a strong

25
explosion. They have no practical importance because they do not exist in
controlled laboratory conditions.
– Plasmas in local thermodynamic equilibrium (LTE plasmas).
These are plasmas in which all temperatures, except the radiation temper-
ature, Tr, are equal in each small volume of the plasma.
– Plasmas that are not in any local thermodynamic equilibrium – non
– LTE plasmas. These plasmas, also named cold plasmas.
The plasmas produced for research or manufacturing purposes are
either LTE or non – LTE type plasmas, designated in daily use, respec-
tively, as thermal and cold plasmas.

1.7 Coalitional processes in plasma

Collisions between charged particles in a plasma differ fundamental-

ly from those between molecules in a neutral gas because of the long
range of the Coulomb force. In fact, it is clear from the discussion in Sect.
1.7 that binary collision processes can only be defined for weakly coupled
plasmas. Note, however, that binary collisions in weakly coupled plasmas
are still modified by collective effects – the many-particle process of De-
bye shielding enters in a crucial manner. Nevertheless, for large Λ we can
speak of binary collisions, and therefore of a collision frequency, denoted
by νss′.
Vs   S  v ss (28)
Here, νss′ measures the rate at which particles of species s are scat-
tered by those of species s′. When specifying only a single subscript, one
is generally referring to the total collision rate for that species, including
impacts with all other species. Very roughly,

1.7.1 Collision cross sections, mean-free paths and collision fre-

quencies

Before proceeding with quantitative treatments of these effects, we

must introduce the idea of a collision ‘cross section’. A cross section can
be defined for any kind of collision, but for present purposes it is suffi-
cient to consider the case of an electron colliding with a neutral atom.
Even in this restricted case there can be two types of collisions: (1) ‘elas-
tic collisions’ in which the electron essentially ‘bounces’ off the atom,
with the two particles retaining their identities as electron and atom, and
the atom remaining in the same energy state; and (2) ‘inelastic collisions’,
such as ionization or excitation, in which one or more of the particles

26
changes its identity or internal energy state. In the first case, the electron
may lose any fraction of its initial momentum, depending on the angle at
which it rebounds. The probability of momentum loss can be expressed in
terms of the equivalent cross section that the atoms would have if they
were perfect absorbers of momentum. In the second case, the probability
of ionization, for example, can be expressed in terms of the equivalent
cross section that an atom would have if it were ionized by all electrons
striking within this cross sectional area.
In Fig. 5, electrons are incident upon a thin slab of thickness dx con-
taining n, neutral atoms per unit volume. The atoms are imagined to be
opaque spheres of cross sectional area a: i.e. every time an electron strikes
the area blocked by the atom, either it loses all of its momentum (elastic
collision) or it ionizes the atom (inelastic collision).

Fig. 5. Electrons incident on a thin slab of thickness dx containing neutral

atoms of density n

The number of atoms per unit area of the slab is nndx and the fraction
of the slab blocked by atoms is nnσdx. If a flux r of electrons is incident
on the slab, the flux emerging on the other side is Г+dГ = Г(1-nσdx), so
that the change of flux r with distance is given by
dГ
 nnГ (29)
dx
which has the solution
Г  Г 0 exp( nnx)  Г 0exp(  x /  mfp ) (30)

where

27
 mfp  (nn ) 1 (31)

The quantity Amfp is called the mean-free path for collisions. In a dis-
tance Amfp, the flux would be decreased to 1/e of its initial value. In other
words, an electron travels a distance mfp before it has a reasonable prob-
ability of colliding with an atom. For electrons of velocity , the mean
time between collisions is given by
  mfp / v (32)
The ‘collision frequency’, namely the inverse of , is usually defined
in terms of an average over all velocities in the Maxwellian distribution
(which may have different individual collision frequencies), namely
v   1  nn v  (nn / n e )  f e (v) (v)vd 2 v (33)
As is implied by this formula, for more complex collisional process-
es than that illustrated in Fig. 4, the cross section σ is often itself a func-
tion of the velocity of the incident particle.

1.8 Coronal equilibrium

In the case where the collision of the electron with the atom results
in ionization of the atom, we may calculate the rate of production of new
electrons per unit volume simply by multiplying the ionization collision
frequency of the electrons, equation (34), by the electron density, ne. This
‘source rate’ Se of electrons is given by
S в  nв nn  ion vв (34)
where ion is the cross section for electron-impact ionization and
where we assume that the electron velocities e greatly exceed the neutral
velocities n so that the velocity of impact comes mainly from the elec-
tron’s motion. This cross section is definitely a strong function of electron
velocity, at least below energies of about 30 eV, so the averaging over the
Maxwellian distribution of electrons is necessary. There is, of course, an
equal and opposite ‘sink rate’ for neutral atoms, i.e. neutral atoms are lost
by ionization at the same rate per unit volume, Se.
The dependence of the ionization cross section σion for hydrogen at-
oms on the energy of the bombarding electron is shown in Fig. 6, and the
ionization rate (σion ve) averaged over a Maxwellian distribution of elec-
trons is shown in Fig. 7. The maximum cross section σion reached for elec-
trons with energies somewhat above Ei (the Rydberg ionization energy,

28
which is about 13.6eV for hydrogen) and is in the neighborhood 1020 m2,
the ‘size’ of the hydrogen atom. However, the ionization rate is signifi-
cant even for electron temperatures well below Ei, because a Maxwellian
distribution still contains a few energetic electrons that are efficient ioniz-
ers. A good approximation to the data is given by the simple formula
2.0 10 13 Tв (вV ) 12 13.6
 ion vв  ( ) exp( )m 3 s 1 (35)
T (вV ) 13.6 Tв (вV )
6.0  в
13.6
The source rate for neutrals (corresponding to a sink term for elec-
trons) in a plasma in coronal equilibrium is given by
S n  nв ni  rвв vв (36)

Fig. 6. Ionization cross section for hydrogen atoms as a function of the

energy of the bombarding electron

29
 Fig. 7. Ionization rate (σion ve ) of electron-impact ionization of hydrogen
atoms averaged over a Maxwellian distribution of electrons, temperature Te

where σrec is the cross section for radiative recombination. For a neu-
tral hydrogen plasma, ni=ne. A good approximation to the data on radia-
tive recombination in the relevant temperature regime is given by the
simple formula
13.6 12 3 1
 ion vв  0.7 10 19 ( ) ms (37)
Tв (вV )
The degree of ionization of a homogeneous hydrogen plasma in cor-
onal equilibrium is given by balancing the source of electrons by colli-
sional ionization against the sink of electrons by radiative recombination.
We find that, at an electron temperature of approximately the ionization
potential, i.e. 13.6 eV, the plasma is almost fully ionized so that the neu-
trals constitute only about one part in105. Only at electron temperatures
below about 1.5 eV is the plasma less than 50 % ionized. Fig. 8 shows the
degree of ionization, i.e. ne /nn, against electron temperature for the coro-
nal equilibrium model, and also for higher-density plasmas where three-
body recombination has been included.

30
 Fig. 8. Ionization equilibrium for hydrogen in the coronal equilibrium
model, and at higher electron densities with three-body recombination included

The concept of coronal equilibrium can be generalized to the case of

a plasma composed of, or containing an admixture of, high-Z ions. In
such cases, depending mainly on the electron temperature, the ions will be
stripped of their outer-shell electrons but will retain some bound inner-
shell electrons. An equilibrium distribution among the various ionization
states arises which, in the solar corona for example, is determined by bal-
ancing the processes of impact ionization and radiative recombination for
each ionization state. A particular case is illustrated in Fig. 9, which
shows the fractional abundances in the various ionization levels of oxy-
gen as a function of electron temperature in coronal equilibrium. We see
that oxygen ions are stripped of all six outer-shell electrons (giving an ion
with charge – number Z = 6) at electron temperatures of about 30eV, but
to remove the final two inner-shell electrons to produce fully stripped ox-
ygen with Z = 8 requires temperatures in excess of about 200 eV. The va-

31
lidity of any coronal equilibrium model depends on the time-scale for
reaching ionizationhecombination balance (for the slowest such process,
generally at the highest relevant ionization state, in the case of high-Z
ions) being much shorter than the timescale on which particles are intro-
duced into, or lost from, the plasma. If this ‘confinement’ time begins to
be comparable to the slowest atomic processes, the ionization balance
shifts towards lower charge states. If hydrogen neutrals are present, there
is also the possibility of a ‘charge-exchange’ event, in which the electron
of a neutral hydrogen atom is captured by a high-Z ion; this process also
lowers the charge-state balance of the high-Z ions. Both electrons then
decay to the ground state, emitting photons. Dielectronic recombination
has not been included in calculating the charge-state distribution shown in
Fig. 9.

Fig. 9. Fractional abundances in the various ionization levels for oxygen

ions as a function of electron temperature in coronal equilibrium. Fully ionized
oxygen has Z = 8

For high-Z ions, another process known as ‘dielectric recombina-

tion’ can play a significant role in the charge – state balance. In this re-
combination process, a free electron is captured into an excited state and
the excess energy that is available is invested in excitation of a different
bound electron to a higher state.

32
 1.9 Penetration of neutrals into plasmas

To complete our discussion of neutrals in plasmas, we should con-

sider what happens at the edge of a hot dense plasma that is enveloped by
neutral gas. This situation arises in many laboratory plasmas-magnetically
confined fusion plasmas or low-pressure arc discharges, for example. In
such cases, the plasma is often hot and dense enough to be fully ionized,
but the electrons and ions that diffuse out of the plasma recombine into
neutral atoms when they strike the containing vessel. The neutral atoms
thus formed are often reflected back into the plasma (or other neutrals are
desorbed from ‘saturated’ vessel walls), where they are ionized again.
Depending on the surface material of the containing vessel (and whether
its surface is already saturated with a layer of hydrogen molecules), this
process of recycling can be almost ‘perfect’, i.e. the plasma density is
maintained almost indefinitely despite diffusive losses of charged parti-
cles, because the lost particles reappear one – for – one as neutrals which
are readily ionized again by the plasma. For hot dense laboratory plasmas,
this recycling process occurs entirely at the plasma edge, because the
main body of the plasma is ‘opaque’ to neutrals, i.e. a neutral atom has
almost no chance of reaching the center of the plasma before being ion-
ized. Recombination in the plasma (as distinct from at the vessel surface
itself) is usually unimportant in this situation: the neutral density in the
edge region of the plasma is set by a balance between the influx from out-
side and the ionization within the plasma.
Neutral atoms entering the plasma with velocity , will penetrate a
distance given by the neutral ‘mean-free path’ for ionization, i.e.
vn
n  (38)
nв ( ion vв )
This can be derived by noting that the volumetric ionization rate is
given by nвnn ( σion ve) for ve >> vn, implying that the effective ‘collision
frequency’ of the neutrals must be nв (σion ve). The thermal velocity of
neutral hydrogen `atoms at ‘room temperature’ is about 2 x 103 m s-1. If
the electron temperature in the edge – region of the plasma reaches (10 –
20) eV, the ionization rate ( ion e) is about 10-14 m3 s-1. Thus, we have
λn(m) = 2 x 1017 / ne(m-3). For example, if the density of the edge-region
plasma is about 1019 m-3, typical of many magnetically confined fusion
plasmas, the neutrals will only penetrate about 2 cm into the plasma. In
many practical cases, such as neutrals re-emerging from a saturated sur-
face, the hydrogen appears initially in molecular, rather than atomic,
form. In such cases, the first effect of electron impact is molecular disso-

33
ciation, which produces two atoms with equal and opposite momenta and
each with energy of about 3eV; the atom with momentum directed toward
the plasma can penetrate somewhat further into the plasma.
A second atomic process-known as ‘charge exchange’-allows much
deeper penetration of hot dense plasmas by neutrals. In hydrogen charge
exchange, an energetic plasma proton captures the electron from a lower-
energy neutral. As a result, it can escape from the plasma, or move further
into the plasma, as an energetic neutral, as illustrated in Fig. 10.

Fig. 10. Charge-exchange process in which an energetic ion takes an

electron from a cold neutral, thereby becoming an energetic neutral

In Fig. 10 time just before the charge-exchange collision is shown to

the left of the thick black arrow; a time just after the collision is shown to
the right.
Not much energy is exchanged by the charge – exchange collision it-
self the emerging neutral has about the same energy as the incident plas-
ma ion.
The cross section for charge-exchange of hydrogen atoms bombard-
ed by protons of various energies is shown in Fig. 11. In the energy range
of most interest for laboratory plasmas and the edge-region of fusion
plasmas ((10 – 100) eV), the cross section is seen to be quite large ( 4 x
10-19 m2), almost a hundred times larger than the ionization cross section.
(The cross section is large because charge exchange is a resonant process,
where the initial and final quantum mechanical states have no difference
in energy.) For a plasma with Te ≈ Ti the charge-exchange rate (vi) is usu-
ally two-to-three times larger than the ionization rate (σion ve). The process
of charge exchange essentially prevents a hot-ion plasma from ever being
formed with an appreciable neutral-gas density in the hot region. The
cross section for charge exchange is so large that, if this were to occur,
each energetic ion would readily turn into an energetic neutral, which

34
would escape, so that the hot plasma would quickly be converted into
cold plasma.

Fig. 11. Cross section for charge exchange in hydrogen against the ener-
gy of the bombarding ion

A low-energy neutral atom injected into the edge-region of a plasma

has a somewhat higher probability of undergoing a charge-exchange
event than of being ionized. Thus, it might appear that charge exchange
would reduce the penetration of neutrals into hot dense plasmas. In fact
the opposite is true, because charge exchange produces a second genera-
tion of more energetic neutrals, with energies comparable to the ion ener-
gies in the region of the plasma where charge exchange occurs. While
some of these more energetic second-generation charge-exchange neutrals
will escape from the plasma, others will penetrate much more deeply into
the plasma interior than did the first generation neutrals, until these neu-
trals themselves are ionized or produce a third generation of charge-
exchange neutrals with even higher energies. Two neutral trajectories-one
ionized and one charge-exchanged – are illustrated in Fig. 12. The pro-
duction of multiple generations of increasingly penetrating neutral atoms
by charge exchange is primarily responsible for the presence of any neu-
trals at all in the center of a hot dense plasma. In the core of the plasma,
these neutrals have ‘thermalized’ with the plasma ions-i.e. they have

35
about the same average energy. However, charge-exchange transport still
provides an avenue for ion energy loss from the plasma.

Fig. 12. Trajectories of two neutrals incident (thick arrows on the left)
upon a plasma of increasing density

Neutral (a) is ionized. Neutral (b) undergoes charge exchange, pro-

ducing a more energetic neutral that penetrates (thick arrows on the left)
further into the plasma before being ionized.

1.10 Collisions with neutrals and with charged particles: relative

importance

Finally, we return to the question raised of the beginning of this

Chapter: what is the relative importance of collisions of charged particles
in a plasma with other charged particles versus collisions with neutral par-
ticles?
The cross section for elastic scattering of an electron by a neutral at-
om may be estimated very roughly as
 n  a02  10 20 m 2 (39)
At the distance , an incoming electron has a substantial chance of
undergoing a large-angle collision. On the other hand, when an electron

36
comes within a distance of a singly charged (e.g. hydrogen) ion, it expe-
riences an attractive Coulomb force:
Fr  e 2 / 4 0 r 2 (40)
which tends to deflect the electron orbit toward the ion. When the
angle of deflection is as much as 900, the electron’s initial momentum is
mostly lost.
Thus, from the viewpoint of momentum exchange, a ‘close encoun-
ter’ with the Coulomb force of another charged particle is essentially the
same as a ‘collision’. The angle of deflection will be large when the po-
tential energy of the Coulomb interaction equals the kinetic energy of the
colliding electron, i.e.
e 2 / 4 0 b  mv 2 / 2  Tв (41)
where m and U are the mass and velocity of the electron, and where
b is the distance of closest approach of the electron to the ion. This serves
to define an effective ‘Coulomb cross section’ of the ion, namely
в 4
 i  b 2   10 17 / Tв (eV ) 2 m 2 (42)
(4 0 ) Tв
2 2

where Tв (eV ) denotes the electron temperature measured in eV.

Comparing n with i, simply using equation (40) for the latter, and
consulting Fig. 12 to relate the degree of ionization to Te, we see that
Coulomb collisions will dominate over collisions with neutrals in any
plasma that is even just a few per cent ionized. Only if the ionization level
is very low (<10-3 ) can neutral collisions dominate. Moreover, a plasma
becomes almost fully ionized at electron temperatures above about 1 eV.
Thus, the case of collisions with neutrals is not of much concern to the
physicist interested in high-temperature plasmas. Not only are high-
temperature plasmas almost fully ionized, but the dynamical behavior of
charged particles even in partially ionized plasmas with more than very
small ionization levels tends to be dominated by Coulomb collisions with
other charged particles, rather than by collisions with neutrals. Of course,
the various inelastic scattering processes involving high – Z ions dis-
cussed in the previous Section, i.e. ionization, recombination and excita-
tion, are still more important than Coulomb collisions in determining the
radiation from high – temperature plasmas, provided only that there re-
mains a sufficient fraction of these ions in partially stripped ionization
levels.

37
 Chapter 2. Introduction to gas discharge physics

2.1 What Is the Subject of Gas Discharge Physics?

The term "gas discharge" originates with the process of discharge of

a capacitor into a circuit incorporating a gap between electrodes. If the
voltage is sufficiently high, electric break down occurs in the gas and an
ionized state is formed. The circuit is closed and the capacitor discharges.
Later the term "discharge" was applied to any flow of electric current
through ionized gas, and to any process of ionization of the gas by the ap-
plied electric field. As gases ionized to a sufficient degree emit light, it
has become customary to say that a discharge "lights up", or is "burning".
As a rule, the flow of electric current is associated with the notion of
a circuit composed of conductors. Actually, a closed circuit or electrodes
are not needed for a directed motion of charges (electric current) in rapid-
ly oscillating electric fields, and even less so in the field of electromagen-
tic radiation. However, quite a few effects observed in gases subjected to
oscillating electric fields and electromagnetic waves (breakdown, main-
taining the state of ionization, dissipation of energy of the field) are not
different, in principle, from dc phenomena. Nowadays all such processes
are referred to as discharges and included within gas discharge physics.
The fact that electric current flow in open circuits in the field of electro-
magnetic waves is of no general significance. In such cases, the dissipa-
tion of the energy of the field is described not as the release of the Joule
heat by electric current, but as the absorption of radiation.
The modern field of gas discharge physics is thus occupied with pro-
cesses connected with electric currents in gases and with generating and
maintaining the ability of a gas to conduct electricity and absorb electro-
magnetic radiation. Gas discharge physics covers a great variety of com-
plex, multi-faceted phenomena; it is full of an enormous amount of exper-
imental facts and theoretical models.
The reader who is interested in this field of gas discharges is referred
to an excellent book written by Yuri Raizer [11].
Before we begin their analysis, it is expedient to single out the main
types of discharge processes and clarify them.

2.2 Typical Discharges in a Constant Electric Field

A relatively simple experiment introduces us to several fundamental

types of discharge. Two metal electrodes connected to a dc power supply
are inserted into a glass tube (Fig. 13). The tube can be evacuated and

38
filled with various gases at different pressures. The quantities measured in
the experiment are the voltage between the electrodes and the current in
the circuit. This classical device served the study of discharge processes
for nearly 150 years, and still remains useful.

Fig. 13. Typical gas discharge tube

If a low voltage is applied to the electrodes, say several tens of volts,

no visible effects are produced, although a supersensitive instrument
would record an extremely low current, on the order of 10 ~15 A. Charg-
es are generated in the gas by cosmic rays and natural radioactivity. The
field pulls them to the opposite - sign electrodes, producing a current. If
the gas is intentionally irradiated by a radioactive or X-ray source, a cur-
rent of up to 10 ~6 A can be produced. The resultant ionization is never-
theless too small to make the gas emit light. A discharge and an electric
current that survive only while an external ionizing agent or the emission
of electrons or ions from electrodes is deliberately maintained (e.g. by
heating the cathode) are said to be non-self-sustaining. As the voltage is
raised, the non-self-sustaining current first increases because most of the
charges produced by ionization are pulled away to electrodes before re-
combination occurs. However, if the field manages to remove all new
charges, the current ceases to grow and reaches saturation, being limited
by the rate of ionization.
As the voltage is raised further, the current sharply increases at a cer-
tain value of V and light emission is observed. These are the manifesta-
tions of breakdown, one of the most important discharge processes. At
pressure P ~ 1 Torr and interelectrode gap L ~ 1 cm, the breakdown volt-
age is several hundred volts. Breakdown starts with a small number of
spurious electrons or electrons injected intentionally to stimulate the pro-
cess: The discharge immediately becomes self- sustaining. The energy of
electrons increases while they move in the field. Having reached the
atomic ionization potential, the electron spends this energy on knocking
out another electron. Two slow electrons are thus produced, which go on
to repeat the cycle described above. The result is an electron avalanche,
and electrons proliferate. The gas is appreciably ionized in 10-7 to 10-3 s,
which is sufficient for the current to grow by several orders of magnitude.

39
Several conditions determine how the process develops at higher voltage.
At low pressure, say 1 to 10 Torr, and high resistance of the external cir-
cuit (it prevents the current from reaching a large value), a glow dis-
charge develops. This is one of the most frequently used and important
types of discharge. It is characterized by low current, i ~ 10-6-10-1 A in
tubes of radius R ~ 1 cm, and fairly high voltage: hundreds to thousands
of volts. A beautiful radiant column, uniform along its length, is formed
in sufficiently long tubes of, say, L ~ 30 cm at P ~ 1 Torr. (This is how
glowing tubes for street advertisements are made.) The ionized gas in the
column is electrically neutral practically everywhere except in the regions
close to the electrodes; hence, this is a plasma. The glow discharge plas-
ma is very weakly ionized, to x = 10-8-10-6 (where x denotes the fraction
of ionized atoms), and is non-equilibrium in two respects. Electrons that
_
get energy directly from the field have a mean energy   1eV and a tem-
perature Te  10 4 K . The temperature T of the gas, including the ions, is not
much higher than the ambient temperature of 300 K. This state, with
widely separated electron and gas temperatures, is sustained by a low rate
of Joule heat release under conditions of relatively high specific heat of
the gas and high rate of its natural cooling. Also as a result of the high
rate of charge neutralization in a cold gas, its degree of ionization is many
orders of magnitude lower than the thermodynamic equilibrium value cor-
responding to the electron temperature.
If the pressure in the gas is high (about the atmospheric level) and
the resistance of the external circuit is low (the circuit allows the passage
of a high current), an arc discharge usually develops soon after break-
down. Arcs typically burn at a high current (i > 1 A) at a low voltage of
several tens of volts; they form a bright column. The arc releases large
thermal power that can destroy the glass tube: Arcs are often started in
open air! Atmospheric-pressure arcs usually form thermodynamic equilib-
rium plasmas (the so-called low-temperature plasma), with Te  T  10 4 K
and the ionization of x  103  101 corresponding to such temperatures.
The arc discharge differs essentially from the glow discharge in the
mechanism of electron emission from the cathode, which is vital for the
flow of dc current of the arc. In the glow discharge, electrons are knocked
from the surface of the cold metal by impacts of positive ions. In the arc
discharge, the high current heats up the cathode, and thermionic emission
develops.
If P ~ 1 atm, the interelectrode gap L > 10 cm, and the voltage is suf-
ficiently high, sparking occurs. The breakdown in the gap develops by
rapid growth of the plasma channel from one electrode to another. Then

40
the electrodes are as if short - circuited by the strongly ionized spark
channel. Lightning, whose "electrodes" are a charged cloud and the
ground, is a giant variety of the spark discharge.
Finally, a corona discharge may develop in strongly nonuniform
fields that are insufficient for the breakdown of the entire gap: A radiant
corona appears at sharp ends of wires at sufficiently high voltage and also
around power transmission line conductors.

2.3 Brief History of Electric Discharge Research

Leaving lightning aside, man's first acquaintance with electric dis-

charges was the observation, dating back to 1600, that friction-charged
insulated conductors lose their charge. Coulomb proved experimentally in
1785 that charge leaks through air, not through imperfect insulation. We
understand now that the cause of leakage is the non-self-sustaining dis-
charge.
Occasional experiments were conducted in the 18th century with
sparks produced by charging a body by an electrostatic generator, and
with atmospheric electricity, experiments with lightning sometimes hav-
ing tragic consequences.
Sufficiently powerful electric batteries were developed at the begin-
ning of the 19th century to allow the discovery of the arc discharge. V.V.
Petrov, who worked in the Saint Petersburg Medical Surgery Academy in
Russia, reported the discovery in 1803. The arc was obtained by bringing
two carbon electrodes connected to battery terminals into contact and then
separating them. Several years later Humphrey Davy in Britain produced
and studied the arc in air. This type of discharge became known as "arc"
because its bright horizontal column between two electrodes bends up and
arches the middle owing to the Archimedes' force. In 1831-1835, Faraday
discovered and studied the glow discharge. Faraday worked with tubes
evacuated to a pressure P ~ 1 Torr and applied voltages up to 1000 V. The
history of physics of gas discharges in the late 19th and early 20th centu-
ries is inseparable from that of atomic physics. After William Crookes's
cathode ray experiments and J.J. Thomson's measurements of the e/m ra-
tio, it became clear that the current in gases is mostly carried by electrons.
A great deal of information on elementary processes involving electrons,
ions, atoms, and light fields was obtained by studying phenomena in dis-
charge tubes.
Beginning in 1900, J.S.E. Townsend, a student of J.J. Thomson and
the creator of a school in the physics of gas discharges discovered the
laws governing ionization and the gaseous discharge (known as the

41
Townsend discharge) in a uniform electric field. Numerous experimental
results were gradually accumulated on cross sections of various electron-
atom collisions, drift velocities of electrons and ions, their recombination
coefficients, etc. This work built the foundations of the current reference
sources, without which no research in discharge physics would be possi-
ble. The concept of a plasma was introduced by I. Langmuir and L. Tonks
in 1928. Langmuir made many important contributions to the physics of
gas discharge, including probe techniques of plasma diagnostics.
As regards different frequency ranges, the development of field gen-
erators and the research into the discharges they produce followed the or-
der of increasing frequencies. Radio frequency (rf) discharges were ob-
served by N. Tesla in 1891. This kind of discharge is easily produced if
an evacuated vessel is placed inside a solenoid coil to which high-
frequency voltage is applied. The electric field induced by the oscillating
magnetic field produces breakdown in the residual gas, and discharge is
initiated. The understanding of the mechanism of discharge initiation
came much latter, in fact, after the work of J.J. Thomson in 1926-1927.
Inductively coupled rf discharges up to tens of kW in power were ob-
tained by G.I. Babat in Leningrad around 1940.
The progress in radar technology drew attention to phenomena in
microwave fields. S.S. Brown in the USA began systematic studies of mi-
crowave discharges in the late 1940s. Discharges in the optical frequency
range were realized after the advent of the laser: A spark flashed in air
when the beam of a ruby laser producing so-called giant pulses (of more
than 10 MW in power) was focused by a lens, this success being achieved
in 1963. Continuously burning optical discharges, in which dense steady-
state plasma is sustained by the energy of light radiation, were first initi-
ated in 1970 by a cw CO2 laser. Optical discharges (this term reflects a
large degree of similarity with conventional discharges) immediately at-
tracted considerable attention. Both microwave and optical discharges
have by now been studied with at least the same thoroughness that the
discharges in constant electric fields has been during nearly 100 years of
research.
The physics of the glow discharge, one of the oldest and, presuma-
bly, best- studied fields, has lived through an unparallel revival in the past
15-20 years, and numerous new aspects of this phenomenon have been
revealed. This surge of attention was stimulated by the use of glow dis-
charges in electric-discharge CO2 lasers developed for the needs of laser
technologies. Likewise, the application of plasmatrons (generators of
dense low-temperature plasma) to metallurgy, plasma chemistry, plasma
welding and cutting, etc. provided a stimulus for new extensive, detailed

42
studies of arc plasma at P ~ 1 atm, T ~ 104 K, and of similar discharges in
all frequency ranges. These, and many other practical applications of gas
discharge physics place it within the range of sciences that lie at the foun-
dation of modern engineering.

2.4 Mechanism of electrical discharges

2.4.1 Cathode processes

In general, laboratory discharges are created between electrodes and

these electrodes can supply the discharge with electrons through various
physical processes. Under normal conditions electrons in a metal are pre-
vented from leaving that metal by electrostatic forces between the elec-
trons and the ions in the metal lattice. As shown in Fig. 14 a the electrons
in the metal are trapped in a potential well. The energy necessary to re-
move an electron from the top of the Fermi energy levels is known as the
work function of the metal. This is denoted by φ in Fig. 14 a [12].

Fig. 14. Energy diagram: a) electrons in a metal; b) when an electric field is

applied

43
 Note that the electric field not only reduces the height of the barrier
but it also reduces the thickness of the barrier. This makes it easier for the
energetic electrons to jump over the barrier (Schottky effect) and a certain
percentage of the electrons to tunnel through the barrier
The work functions of typical metals are tabulated in Table 3 [13].

Table 3. Work function for typical elements

Element Ag Al Cu Fe W
Wα(eV) 4.74 2.98- 4.07- 3.91- 4.35-
4.43 4.7 4.6 4.6

Electrons can be removed from the metal either by giving the elec-
trons sufficient kinetic energy to surmount the potential barrier, or the
work function, at the surface, by reducing the height of the barrier so that
the electrons can overcome it or by reducing the thickness of the barrier
so that the electrons can tunnel through it. The first can be achieved by the
application of heat to the electrode, through impact of photons on the sur-
face of the electrode or by the incidence of particles such as other elec-
trons, positive ions, neutral molecules, metastable atoms onthe electrode.
The reduction in the potential barrier height or its thickness can
beachieved by the application of an electric field in the correct direction
so that the electrons will experience a force directed out of the metal sur-
face (see figure 14 b).Let us consider different physical processes that can
cause emission of electrons from metals.

2.4.2 Photoelectric emission

When a photon is incident on a metal it can transfer all its energy to

an electron in the metal so that the latter can surmount the potential barri-
er at the surface. Thus, an electron at the Fermi level after absorbing the
energy of the quanta can escape with energy Ee given by:
Ee  hv   (43)

Where hv is the energy of the photon. For a clean Ag surface

φ=4.74eV (Table 3) and the threshold frequency corresponds to a wave-
length of 2616 Å.

44
 2.4.3 Thermionic emission

In metals at room temperature the energy of the conduction electrons

is not sufficient for them to surmount the potential barrier at the surface of
metals. However, heating a metal will increase the kinetic energy of its
electrons. With increasing temperature one may expect the number of
electrons surmounting the barrier to increase. Since this process is caused
by the heating of the electrode it is called thermionic emission. The ther-
mionic current density, Jt , is given by the following equation [14]:
J t  AT 2 exp(  / kT ) (44)

Where k is the Boltzman constant, φ is the work function of the

metal, T is the absolute temperature and A is a constant equal to 120 A
cm – 2 deg−2.
Actually, the experimentally obtained current densities are smaller
than those predicted by his equation because this equation does not take
into account the wave nature of electrons and the possibility that some of
the electrons will be reflected at the barrier even if they have enough en-
ergy to overcome it.

2.4.4 Schottky effect

As one can see from the equation given in the previous section, the
thermionic current depends on the height of the barrier that the electrons
have to surmount in order to come out of the metal. The height of this
barrier can be decreased by the application of an electric field in such a
way that the electrons in the metal experience a force out of the metal.
This is illustrated in Fig. 14 b.
In the presence of such an electric field the work function is effec-
tively reduced to [15]:
e3 E
1    (45)
4 0

This reduction in the barrier height will lead to a change in the ther-
mionic emission current. Thus, if J0 is the thermionic emission current
density for zero electric field at temperature T then the current density at
the same temperature in the presence of an electric field E is given by:
J s  J 0exp( 4.4 E / T ) (46)

45
 where the electric field E is given in V/cm and T in Kelvins. This was
shown to be valid for electric fields up to about 106 V/cm. This process of
enhancement of the thermionic emission current due to the reduction in
the barrier height is called Schottky emission.

2.4.5 Field emission

Calculations done with the Schottky equation show that the thermi-
onic emission current at T = 293K for values of φ about 4.5eV is negligi-
ble even when the electric field reaches values as high as106 V/cm. How-
ever, experiments show that electrodes in vacuum do emit appreciable
currents, in the range of μA, at such electric fields. The reason for this is
the quantum nature of the elementary particles. The illustration in Fig. 14
b shows that the application of the electric field not only reduces the
height of the barrier but will also decrease its thickness. In the absence of
the electric field the barrier is infinitely thick but its thickness decreases
with increasing electric field.
Electrons incident on the barrier can be represented by a wave; dur-
ing the inter-action part of the wave will be reflected and the other part of
the wave will be transmitted. The transmitted wave attenuates rapidly
when moving into the barrier, but if the thickness of the barrier is not
large a small fraction of the wave may be able to penetrate it. Of course,
in the case of electrons, the reflection and transmission coefficients have
to be regarded as probability functions, so there is a certain probability
that an electron incident on the barrier will penetrate it. If the number of
electrons incident on the barrier per unit time is known, quantum mechan-
ical calculations can be performed to evaluate the number of electrons
coming out of the barrier. Fowler and Norheim [16] analysed this process
in detail and obtained the following expression for the field emission cur-
rent density, Jf , for pure metallic surfaces in vacuum.
 1/ 2 E 2
J f  61.6  10 7 exp(6.8  107  3 / 2 / 3E ) (47)
(   ) 1/ 2

Where  =5eV (typical value for metals),φ is the work function in

eV, E is the electric field in V/cm and the current density is given in
A/cm2.
This equation indicates that measurable currents should be obtained
for electric fields of the order of 107 V/cm. This has been found to be val-
id for very clean surfaces. Experiments show, however, that an apprecia-
ble electric field dependent emission current can be obtained for electric
fields one to two orders of magnitude smaller than this in the presence of

46
surface contamination. The reason for this is that surface contamination
causes a reduction in the width of the barrier thus enhancing the field
emission process. Moreover, if there are protrusions on the surface the
electric field at the tip of these protrusions can reach very high values
leading to field emission from them. The field emission process is very
important in providing initiatory electrons in the creation of electrical dis-
charges.

2.5 Electrical breakdown

The physical processes that take place during the formation of an

electric discharge can be summarised as follows. The electrical break-
down in a gas starts with a single electron which will lead to an avalanche
of electrons created through the electron collision ionisation. As the ava-
lanche grows the electric field created by charges concentrated at the ava-
lanche head starts to modify the electric field in the vicinity of the head of
the avalanche. When this space charge electric field reaches a critical val-
ue the avalanche will convert itself to a streamer discharge. If the gap is
short then the streamer discharge may bridge the gap and, after streamer
to spark transition, the complete breakdown of the gap may take place. If
the gap is long many streamers may start from the electrode having their
origin at a common streamer stem. The heat generated by the streamer
currents will increase the temperature of the streamer stem and when the
temperature reaches a critical value the thermal ionisation sets in the
stem, the conductivity of the stem increases and it will convert itself to a
leader discharge. Since the leader channel is a good conductor the poten-
tial of the electrode is now transferred to the head of the leader channel
and the resulting high electric field will cause streamers to grow from the
head of the leader channel. The leader elongates in the gap through the
action of streamers that forge further and further into the gap. When the
leader discharge reaches the grounded electrode the current in the channel
increases, and the applied voltage collapses, leading to the formation of a
spark. In the following sections the processes mentioned above are de-
scribed.

2.5.1 Electron avalanche

Consider a free electron originated at x=0 in space and moving under

the influence of a background electric field directed in the x direction. If
the background electric field is larger than the critical value necessary for
cumulative ionisation the electron may produce another electron through

47
ionisation collisions and these two electrons in turn will give rise to two
more electrons. In this way the number of electrons increases with in-
creasing x.
Assume that the number of electrons at a distance x from the origin
is n x . Let  be the number of ionising collisions per unit length made by
an electron travelling in the direction of the electric field and  be the
number of attachments over the same length. Here  is Townsend’s ioni-
zation coefficient and  is the attachment coefficient. Consider an elemen-
tary length of width dx located at a distance x from the origin. In travel-
ling across the length dx, n x number of electrons will give rise to dn addi-
tional electrons:
dn  nx (  )dx (48)
The solution of this equation is:
nx  e(  ) x (49)
This equation shows that the number of electrons increases exponen-
tially with dis-tance. This exponential growth of electrons with distance is
called an electron avalanche. Fig. 15 shows a photograph of an electron
avalanche obtained in a cloud chamber [17].
The equation also shows that cumulative ionisation is possible only
if (α−η) >0. The quantity (α−η) is known as the effective ionisation coef-
ficient and denoted by  .

Fig. 15. Cloud chamber photograph of a single electron avalanche

(from [18])

It is important to note, however, that the value of nx given in eqn. 49

is a mean value and it is subject to considerable variations due to the sta-

48
tistical nature of the collision process. The probability that one electron at
the origin results in an avalanche of total number n at a distance x is given
by
n 1
1  1 
Р(n, х)  1   (50)
nmean  nmean 
with a standard deviation given by:
1/ 2
 1 
  nmean 1   (51)
 nmean 
Where
nmean  е (  ) x (52)

For large values of n this becomes:

1
Р(n, х)  е n / nmean (53)
nmean

With   nmean . That is, the size of an electron avalanche originating

from a single electron follows an exponential distribution. It is important
to note that this result is based on the assumptions that the space charge of
the avalanche does not have significant influence on the background elec-
tric field and the probability of an electron ionising a gas molecule is con-
stant and is independent of the distance it has travelled in the electric field
direction since the last ionising collision.

2.5.2 Formation of a streamer

A schematic description of the formation of a positive streamer is

shown in Fig. 16 [19]. As the electron avalanche propagates towards the
anode low mobile positive space charge accumulates at the avalanche
head. When the avalanche reaches the anode, the electrons will be ab-
sorbed into it leaving behind the net positive space charge. Due to the re-
combination of positive ions and electrons, the avalanche head is a strong
source of high energetic photons. These photons will create other ava-
lanches in the vicinity of the positive space charge. If the number of posi-
tive ions in the avalanche head is larger than a critical value the electric
field created by the space charge becomes comparable to the background
electric field and the secondary avalanches created by the photons will be
attracted towards the positive space charge. The electrons in the second-
ary avalanches will be neutralised by the positive space charge of the
primary avalanche leaving behind a new positive space charge, a little bit

49
closer to the cathode. The process repeats itself and the positive space
charge head travels towards the cathode as a consequence. This discharge
is called a cathode directed streamer or a positive streamer.

Fig. 16. Schematic diagrams showing the formation a positive streamer

A: an external photon triggers an avalanche. B: a positive ion strikes

the cathode and starts an avalanche C,D: the avalanche tip reaches the an-
ode. E: photons originating from the avalanche produce free electrons
both from the cathode and in the gas. F: the positive space charge close to
the anode increases the electric field and a streamer is just about to be
formed. G: plasma of positive ions and electrons forms the streamer
channel. H: streamer tip. I: production of free electrons by photons. J:
streamer close to the cathode (from [7]).

50
 The formation of a negative streamer or an anode directed streamer
is shown in Fig. 17. The electrons of the avalanche move into the gap
leaving behind positive charge close to the cathode. When the avalanche
reaches the critical size the secondary avalanches extend the positive
space charge towards the cathode (as in a cathode directed streamer).
When the positive channel reaches the cathode both the field enhance-
ment associated with the proximity of positive space charge to the cath-
ode and the collision of positive ions on the cathode lead to the emission
of electrons from the latter. These electrons will neutralize the positive
space charge creating a weakly conducting channel that connects the neg-
ative head of the electron avalanche to the cathode. The high electric field
at the head of the avalanche pushes the negative space charge further into
the gap while the positive space charge left behind is neutralized by the
electrons supplied by the cathode and travelling along the weakly con-
ducting channel connecting the streamer head and the cathode.

Fig. 17. Schematic representation of the formation of a negative streamer

and the physical processes taking place at the streamer head
(adapted from [20])

As the streamer propagates it may lead to the formation of branches.

The avalanche to streamer transition takes place when the number of
charged particles at the avalanche head exceeds a critical value, Nc. From
cloud chamber photographs of the avalanches and streamers, Raether [21]
estimated that an avalanche will convert to a streamer when the number

51
of positive ions in the avalanche head reaches a critical value of about
108. A similar conclusion is also reached independently by Meek [22].
Thus the critical avalanche length for transition to a streamer is given by
e x  10 8
c
(54)
or
axc  18 (55)

Note that the critical avalanche length decreases with increasing

electric field.

2.5.3 Propagation of the streamer discharges

A schematic representation of the propagation of a positive streamer

is shown in Fig. 18. The local but strong electric field resulting from the
concentration of positive charge at the streamer head attracts the second-
ary avalanches towards it. These avalanches neutralise the positive space
charge of the original streamer head leaving behind an equal amount of
positive space charge at a location slightly ahead of the previous head.
The repetition of the process leads to an effective forward propagation of
the streamer head connected to the anode by a weakly conducting chan-
nel. Since the electron multiplication in the active region is supported by
the space charge electric field of the streamer head, the streamer can
propagate in electric fields which are much smaller than the critical elec-
tric field necessary for cumulative electron ionisation. Indeed, Dawson
and Winn have shown that a spherical space charge containing 108 elec-
trons confined within a radius of 30 μm can propagate in zero electric
field for a short time [23].

Fig.18. Schematic presentation of the propagation of positive streamers (from

[11])

52
 The mechanism of propagation of negative streamers is a little bit
more complicated than that for positive streamers. This is shown in Fig.
19.

Fig. 19. Schematic presentation of the propagation of negative streamers

(from [24])

Note that there are two main differences between the negative and
positive streamers. In the negative streamer, the electrode has to supply
the electrons necessary for the neutralisation of the positive space charge
left behind by the avalanches whereas in the positive streamers the anode
absorbs the extra electrons generated by the secondary streamers. The lat-
ter is a much easier process than the former. Second, in the positive
streamers the electrons propagate towards the positive charge head of the
streamer and therefore into an increasing electric field. In the case of neg-
ative streamers the electrons move into the low electric field region and
some of them will be captured by electronegative atoms which create an
immobile negative space charge region that will impede the streamer
propagation. Both these features make the propagation of positive stream-
ers easier than that of negative streamers.
We have seen before that for the inception of a streamer the number
of charged particles at the avalanche head should reach a critical value.
Similarly, for the continuous propagation of a streamer the number of
charged particles in the streamer head has to be larger than a critical num-
ber, Nstab. The value of Nstab partly depends on the background electric
field, Eb, and can be calculated from the following equation [25]:

N stab  0.558  10 8  0.231  10 3 Eb forEb  2  10 5V / cm (56)

5
N stab  3.34  108 e 1.61410 Eb
forEb  2  10 5V / cm (57)

53
 2.6 Electrical breakdown in very small gaps – Townsend’s
breakdown mechanism

As described previously, avalanche to streamer transition requires

that the avalanche grows to about 108 electrons and the space charge in
the avalanche tip creates an electric field that significantly adds to the
background electric field in the vicinity of the avalanche tip. This ava-
lanche to streamer transition has been observed when the product of the
pressure and the electrode spacing in plane uniform gaps exceeds about
0.5 bar.cm. Below this limit the space charge of the avalanche is not large
enough to change the background field significantly. This will inhibit the
avalanche to streamer transition. Under such conditions the breakdown
takes place according to the Townsend mechanism.

2.6.1 Townsend’s experiment

In Townsend’s experiment (Fig. 20) a plane parallel electrode gap

was located in a cell, the gas pressure of which was of the order of a few
torr [26]. The cathode was illuminated with a steady beam of ultraviolet
radiation which led to a steady stream of electrons from it. The current
flowing across the gap was measured as a function of the voltage. Town-
send found that the voltage and the current vary in a manner shown in
Fig. 21.

Fig. 20. Schematic representation of the apparatus used by Townsend in

his experiment. V is the voltage source and T is the vacuum tube

54
 The discharge gap is located in the vacuum tube and the cathode is
illuminated by ultraviolet radiation.

Fig. 21. Variation of the current flowing across the discharge tube of
Townsend’s experiment as a function of the applied voltage

Let us investigate the features of this curve in detail. Initially, the

current in the gap increases with increasing voltage. The reason for this is
that some of the electrons emitted by the cathode diffuse back into it and
some of the electrons are lost to the walls. This diffusion and the losses to
the walls of the tube decrease with increasing voltage. When the voltage
increases to a certain level almost all the electrons emitted by the cathode
are collected by the anode, thus producing the saturation in current. In re-
ality, however, the current will only approach but not sustain a saturation
level. The reason for this is that even though the back diffusion decreases
with increasing voltage not all the emitted electrons will reach the anode
even when the voltage is close to the sparking potential. As the voltage
increases further, Townsend found that the current starts to increase ex-
ponentially with the applied voltage. This exponential growth of the cur-
rent was sustained over a certain range of applied voltage but with further
increase in voltage the current started to increase faster than the exponen-
tial growth. Further increase in voltage resulted in the electrical break-
down of the gap.

55
 2.6.2 Townsend’s theory of electrical breakdown

Townsend assumed that the initial exponential growth of the current

in the discharge tube is caused by the production of secondary electrons
through the collision of primary electrons, generated from the cathode,
with gas atoms. The second phase of the discharge in which the current
grows faster than exponential is assumed to be caused by the ionization of
the atoms through the collision of ions. However, today we know that the
correct explanation is the additional production of electrons by the colli-
sion of positive ions with the cathode. Let us consider the mathematical
treatment of the discharge process taking place in Townsend’s experiment
and come up with an equation which matches the experimental data.

2.6.2.1 Primary ionization stage

nd  n0ed (58)
Consequently, the current inside the tube is given by:
I d  I 0ed (59)

Where d is the gap length, I d  nd e , I0  n0e ,  is Townsend’s first

ionization coefficient and e is the electronic charge. This equation ex-
plains the exponential growth of current with increasing voltage. Note
that even though the voltage does not appear explicitly in this equation it
indirectly appears in  which increases with increasing electric field (and
hence with increasing voltage when d is constant). Note that I 0 is the cur-
rent generated by the ultraviolet radiation at the cathode. Consequently, if
the source of ultraviolet radiation is removed (i.e. I 0 = 0) then the current
in the discharge tube will go to zero. Thus, the discharge is not self-
sustained, that is, it needs the support of the external agency for its con-
tinuation.

2.6.2.2 Secondary ionization stage

Townsend observed that with increasing voltage the current in the

discharge tube begins to increase with voltage at a rate faster than that
predicted by eqn. 59. This departure from eqn.59 is shown in Fig. 22
where log(I) is plotted as a function of the gap spacing, d. The upcurving
occurs when the current departs from this equation. Townsend assumed
that the cause of this departure of the experimental data from that predict-
ed by eqn.59 is the ionization of the gas atoms due to the collision of

56
positive ions. But, in reality, the energy gained by the positive ions at
electric fields encountered in Townsend’s experiment is not sufficient to
create significant ionization. However, one process that may cause this
departure from eqn. 59 is the generation of electrons from the cathode by
positive ion bombardment. As the voltage increases the positive ions gain
more and more energy and this energy is released at the cathode. With in-
creasing energy a stage will be reached in which these positive ions will
start liberating electrons from the electrode. In order to explain the varia-
tion of current with voltage one has to take this effect into account. Let us
now derive a mathematical expression for the current in the discharge
tube taking into account the electron current created by the bombardment
of positive ions on the cathode.

Fig. 22. Variation of the logarithm of the current flowing across the
discharge tube in Townsend’s experiment as a function of the logarithm of the
electrode spacing for different values of (E/p)

Here E is the electric field in the gap and p is the pressure.

Let n0 be the number of electrons emitted by the cathode per second
due to ultra-violet illumination and n+ be the number of electrons released
from the cathode per second due to bombardment of positive ions. Then
the number of electrons reaching the anode at steady state per second, n,
is equal to:
n  (n0  n )ed (60)

57
 The number of positive ions created by the electrons reaching the
anode per second is equal to n  (n0  n ) and at steady state the number of
positive ions reaching the cathode per second is equal to this number.
Consequently, the number of electrons released by the positive ion bom-
bardment at the cathode per second is given by:
n  n  (n0  n ) (61)
Where  is the average number of electrons released by each posi-
tive ion striking the cathode. This parameter is called Townsend’s sec-
ondary ionisation coefficient. Substituting this expression in eqn. 60 one
obtains:
n0ed
n (62)
1   (ed  1)
The current in the discharge tube is given by:
I 0ed
I (63)
1   (ed  1)
This equation predicts a faster current growth than eqn. 60 with in-
creasing electric field (or voltage) providing an explanation for Town-
send’s experimental results.
In deriving the above equation we have considered the bombardment
of positive ions on the cathode as the only secondary ionization process.
Let us now consider the other secondary ionization processes.
(I) Ionization of gas by positive ions: Townsend in his original deri-
vation assumed that the secondary ionization mechanism is due to the ion-
ization in the gas by positive ions. However, as mentioned above, positive
ions cannot produce significant ionisation in electric fields at which elec-
trical breakdown is observed in Townsend’s experiment. Nevertheless, if
we assume that the ions will contribute to the ionization the resulting
equation for the current will take the form [12]:
e(   ) d
i  i0 (64)
( /(   ))  (  /(   ))e(   ) d
Where β is the ionisation coefficient of positive ions. Since electrons
ionize more readily than positive ions one can replace (   ) by  and
the equation will reduce to:
e ( ) d
i  i0 (65)
(1  (  /  )e ( ) d
This equation has the same form as that of the expression obtained
for positive ion bombardment at the cathode (i.e. eqn. 63).
(II) Photo emission from the electrode: Another secondary emission
process that one may take into account is the interaction of photons in the

58
discharge with the electrodes. If the incident photon has an energy larger
than the work function of the electrode then the interaction may lead to
the liberation of an electron. If this process is taken in to account as a sec-
ondary mechanism then the expression for the current will take the form
[6]:
e( ) d
i  i0 (66)
(1  (g /    ))[e(   ) d  1]
Where g is the fraction of photons emitted in the gas that are
headed towards the cathode,  is the coefficient of absorption of photons
in the gas and  is the probability of photoelectric emission from the pho-
tons incident on the electrode (note that only a fraction of the incident
photons will have sufficient energy to cause photo ionization).
In general α >> μ and the equation will reduce to:
e( ) d
i  i0 (67)
(1  (g /  )[e( ) d  1]
This again has the same form as that of the eqn. (68).

(III) Collision of meta-stable ions on the cathode: Collision of meta-

stable atoms on the cathode can also liberate electrons from the cathode.
If the meta-stable level has a higher energy than the work function of the
metal, such an interaction can lead to the emission of electrons. If this is
taken into account [27]:
e( ) d
i  i0
 
1   m e ( ) d  1
(68)

where  m is the number of electrons liberated by the incidence of a

meta-stable atom on the cathode.
(IV) Ionisation of the gas by photons: Another process that may con-
tribute to electron production in the gas is the ionisation of the gas by
photons. If this is taken into account, one can show that the resulting
equation will have the form [13]:
e( ) d
i  i0 (69)
(1  (gf /    ))[e(   ) d  1]
where  is the coefficient of absorption of photons,  is the fraction
of absorbed photons that will cause ionisation, f    where  is the
number of excited states or photons created per unit path in the electric
field direction per electron and g is a geometrical factor that will describe
the fraction of photons directed in a given direction.
(V) Final expression for the current in the presence of secondary
processes: The results presented above show that irrespective of the sec-

59
ondary ionisation process under consideration the final expression for the
current has the same form. Indeed one can include all of them in a single
formula as follows:
e ( ) d
i  i0 (70)
(1  ( i )[e ( ) d  1]

Where
gf
 i    (  )  g /    m  ( ) (71)
 

2.6.3 Townsend’s electrical breakdown criterion

The final expression for the current given in eqn. 69 shows that the
discharge is still non self-sustained. That is, the discharge current goes to
zero if the UV illumination on the cathode is removed (i.e. as I0 → 0).
However, as the voltage continues to increase a stage will be reached at
which the discharge will transform itself from a non self-sustained dis-
charge to a self-sustained discharge. At this stage the discharge will con-
tinue to burn between electrodes even after removing the background ul-
traviolet radiation (i.e. when I0 = 0). This change of state of the discharge
is accompanied by a several orders of magnitude increase in the current
(provided that the voltage supply can sustain such an increase in the cur-
rent) in the discharge gap. This is the stage of electrical breakdown in the
gap. Townsend defined the electrical breakdown condition as the condi-
tion which makes the current in the discharge gap go to infinity.
From eqn. 69 one can see that I0→ ∞ when the denominator of the
expression for the current goes to zero. That is when:

1   i [(e( ) d  1)]  0 (72)

This condition is known as Townsend’s breakdown criterion.

One can indeed show that this criterion has a physical significance.
Assume that  is the dominant secondary process. Assume that n0 prima-
ry electrons leave the cathode per second. These electrons will give rise to
n0 ( e( ) d  1 ) positive ions in the gap, and these positive ions on incidence
on the cathode produce  i n0 (e( ) d  1) secondary electrons. When Town-
send’s breakdown criterion is satisfied the number of secondary electrons
is equal to the original number of electrons which has been drawn away
from the cathode and later passed into the anode. Consequently, each ava-
lanche will give rise to another avalanche through secondary processes

60
and so cause a repetition of the avalanche process. That is, the discharge
process becomes self-sustained.
An alternative expression for Townsend’s breakdown criterion can
be obtained by rewriting the above equation as:
1
d  ln(1  ) (73)
i
The value of  i is greatly affected by the cathode surface and gas
pressure. However,  i is a very small number (<10−2–10−3) so 1/  i is very
large. Therefore, ln(1+(1/  i )) does not change too much and is of the or-
der of eight to ten in a Townsend’s discharge.

2.6.4 Townsend’s mechanism in the presence of electron attach-

ment

In the experiments conducted by Townsend the gases under investi-

gation were noble gases. In this case electron attachment to atoms can be
neglected. However, in air, one cannot neglect the electron attachment.
Let us consider the effect of electron attachment in Townsend’s equation.
In the presence of attachment, the number of electrons reaching the cath-
ode per second is given by:
n0  n0e (  ) d (74)
Consider an elementary length dx located at distance x from the
cathode. Let nx be the number of electrons reaching x in a unit time. In
travelling across dx these electrons will generate dn− number of negative
ions per second. Then:
dn  nxdx (75)
substituting for n x one obtains:
dn  n0e(  ) x dx (76)
Since the number of negative ions at the cathode is equal to zero the
solution of this equation is given by:
n0 (  ) x
n 
 
e  1 (77)
Where n is the number of negative ions crossing a plane at a dis-
tance x from the cathode per second. Thus the number of negative ions
reaching the anode per second, nd , is given by:
n0 (  ) d
n d 
 
e  1 (78)
The total current reaching the anode, both due to the electrons and to
negative ions, is given by:

61
 i
i0
 

e(  ) d    (79)
In the presence of secondary ionisation due to bombardment of posi-
tive ions on the cathode one can show using the procedure outlined in the
section 2.6.2.2
e(  ) d  
i  i0
     e(  ) d  1
(80)
This reduces to eqn.80 in the absence of attachment (i.e.  =0). From
this equation the breakdown condition in the presence of electron attach-
ment is given by:
   (  ) d
1  [(e  1)]  0 (81)
 (   ) 
This criterion shows that if    then electrical breakdown is possi-
ble irrespective of the values of  , and  provided that d is large
enough. That is, for a given electric field there is a particular value of d at
which the gap breaks down. For    , with increasing d, the above equa-
tion approaches an asymptotic form:

 
 1 or   (82)
(   ) (1   )
This defines the limiting condition at which electrical breakdown is
possible in an electronegative gas. This condition depends only on E/p.
Noting that the value of  <<1,the limiting value of E/p which can cause
electrical breakdown in electronegative gases can be obtained from the
relationship    (see section 2.5). This point is illustrated in the plot
given in Fig. 23.

2.7 Paschen’s law

If a slowly increasing voltage is applied across two plane parallel

electrodes the electrical breakdown of the gap occurs at a certain critical
voltage. The experimental data shows that the breakdown voltage, Vs, is
only a function of the gas pressure, P (or gas density), multiplied by the
gap length, d . That is:
Vs  f (Pd ) (83)

62
 Fig. 23. Variation of the logarithm of the current flowing across the dis-
chargetube in Townsend’s experiment as a function of the logarithm of the elec-
trode spacing for different values of (E/p) for electronegative gases

This is known as Paschen’s law. The Paschen curve for air is shown
in Fig. 24; the data points correspond to measurements by several authors
and the solid black dots are generated from the equation
Vs  6.72 Pd  24 .4( Pd )) (P in bar and d in mm).

Fig. 24. Paschen curve for air in log–log scale at temperature 20◦C [28]

In this Fig. the results are given in terms of Pd corresponding to a

temperature of 293 K. Note that the sparking potential is high both in the

63
case of low and high values of Pd but there is a minimum at a certain val-
ue of pd. This minimum is called the Paschen minimum. The Paschen
minimum in air is about Pd=10−2 bar.mm. One can show that the break-
down voltage estimated using either Townsend’s or Raether and Meek’s
criterion adheres to Paschen’s law. The electrical breakdown criterion of a
uniform gap of length d is given by
d  K (84)
Where K is a constant. Depending on the value of K this equation
represents both Townsend’s breakdown criterion and the streamer break-
down criterion. Substituting the expression for  given in eqn. 80 we ob-
tain:
K  Apde  Bpd / V s
(85)
Where Vs is the voltage at which electrical breakdown is observed.
Note that in deriving this equation we have used E=Vs/d. Rearranging the
above equation we find that
Bdp
Vs  (86)
lnApd / K 
This equation shows that Vs is a function of pd. The general shape of
this equation is in agreement with the Paschen curve.

2.7.1 Physical interpretation of the shape of the Paschen curve

In order to give a physical interpretation for the shape of the Paschen

curve let us rewrite eqn. 86 as follows
B1 (d /  )
Es  (87)
lnA1 (d /  ) / K 

Where A1 and B1 are constants,  is the mean free path of the elec-
trons Es=Vs/d ,and we have used the relationship that the mean free path
is inversely proportional to the pressure. This equation has the same form
as eqn. 87 except that Pd is now replaced by d /  . It thus predicts that the
breakdown electric field has a minimum corresponding to a certain value
of d /  , say ( d /  )min, and it increases when the value of d /  moves away
from this minimum. The reason for the existence of this minimum can be
explained qualitatively as follows. Let Em be the electric field in the gap
corresponding to ( d /  )min. At ( d /  )min an electron crossing the gap will
make a certain number of ionisation collisions. Consider the case d /  >
( d /  )min. Now, the number of collisions made by an electron in crossing
the gap, and hence the energy lost in collisions in crossing the gap, is
higher than at ( d /  )min. If the background electric field remained at Em the

64
total number of ionisation collisions made by an electron crossing the gap
would be less than the corresponding value at ( d /  )min. Consequently, the
electric field should be increased in order to compensate for the losses and
to increase the probability of ionisation. For d /  < ( d /  )min the number of
collisions, and hence the number of ionisation collisions, made by an
electron in crossing the gap is less than the corresponding value at
( d /  )min. In this case the only way to increase the number of ionisation
collisions is to increase the probability of ionisation in each collision.
This can be achieved only by increasing the energy gained by electrons
within a mean free path. This requires a higher electric field than the one
which corresponds to ( d /  )min.

2.7.2 Validity of Paschen’s law

In section 2.7 it was shown that Paschen’s law follows directly if the
dominant collision processes, as is often the case, are such that the coef-
ficients representing them, for example α, are directly proportional to p at
a given value of E/p. When this is the case the processes are said to obey
similarity. In general, the experimental data obeys Paschen’s law and any
deviations are relatively small and arise from the existence of collision
processes in the gas which do not conform to similarity. The deviations
from Paschen’s law can occur at high pressures and at temperatures above
about 3000–4000 K. At high pressures exceeding a few atmospheres the
processes such as field emission may play a significant role in the break-
down process. The role of field emission at high pressures and its
influence on Paschen’s law is clearly demonstrated by the observation
that when very clean molybdenem electrodes are used Paschen’s law
holds up to very high pressures. The clean electrodes do not have oxide
layers that generate field emission at low electric fields. At high tempera-
tures experimental data departs from Paschen’s law partly due to the dom-
inant role of thermal ionisation and partly due to the gradual change of
chemical composition of the gas, for example, by dissociation. Paschen’s
law may also break down at low pressures because the breakdown process
is governed by preionisation processes caused by electrodes such as ther-
mionic emission and the break down phenomena has to be described by
vacuum breakdown processes.

2.8 Classification of Discharges

Discharges in a dc electric field can be classified into (a) non-self-

sustaining and (b) self-sustaining types. The latter are more widespread,

65
more diversified, and richer in physical effects. Steady and quasi-steady
self-sustaining discharges contain A) glow and B) arc discharges. We
have already mentioned in Sect. 2.2 that the cathode processes of two
types differ in principle. A close relation of the glow discharge is C)
Townsend's dark discharge. It proceeds with a cold cathode and at very
weak current. The D) corona discharge, also self-sustaining and also at a
low current, is a special case.
Typical conditions under which each of the combinations can be ob-
served are summarized in Table 4.

Table 4. Classification of discharge processes

Breakdown Nonequilibrium Equilibrium

plasma plasma
Constant elec- Initiation of Positive column Positive column
tric field glow discharge of glow discharge of high- pres-
in tubes sure arc
Radio frequen- Initiation of rf Capacitivcly cou- Inductively cou-
cies discharge in pled rf discharges pled plasma
vessels filled in rarefied gases torch
with rarefied
gases
Microwave Breakdown in Microwave dis- Microwave
range waveguides and charges in rare- plasmatron
resonators fied gases
Optical range Gas breakdown Final stages of Continuous op-
by laser radia- optical break- tical discharge
tion down

Corona has common features with glow and dark discharges. Among
transient discharges, the E) spark discharge stands out sharply, among
others. Many features of purely plasma processes, characterizing break-
down in a dc electric field, as well as the glow and arc discharges, are typ-
ical for discharges in rapidly oscillating fields, where electrodes are not
necessary at all. It is therefore expedient to construct a classification
avoiding the attributes related to electrode effects, and the following two
properties will be basic for the classification: the state of the ionized gas
and the frequency range of the field. The former serves to distinguish be-
tween A) breakdown in the gas, B) sustaining non-equilibrium plasma by
the field, and C) sustaining equilibrium plasma. Frequency serves to clas-
sify fields into A) dc, low-frequency, and pulsed fields (excluding very

66
short pulses), B) radio-frequency fields (f ~ 105 - 108Hz), C) microwave
fields (f ~ 109 - 1011 Hz, λ~ 102 – 10-1 cm), and D) optical fields (far from
infrared to ultraviolet light). The field of any subrange can interact with
each type of discharge plasma. In total, we have 12 combinations. All of
them are experimentally realizable, and quite a few are widely employed
in physics and technology.

67
Chapter 3 Basic principles of plasma chemical processes

Plasma chemistry has developed naturally from the low temperature

plasma physics of molecular gases, and especially of their multicompo-
nent mixtures. Plasma chemical reactions (particularly in electrical dis-
charges) usually occur under non-equilibrium conditions and themselves
create non-equilibrium. Reactions of charged particles and excited mole-
cules play a very important if not principal role; chemical reactions under
plasma conditions result in excited particles which have non-equilibrium
distributions. From this the need arose for new developments of certain
theoretical ideas and experimental methods of plasma chemistry.
Chemically active plasma is a multi-component system highly reac-
tive due to large concentrations of charged particles (electrons, negative
and positive ions), excited atoms and molecules (electronic and vibration-
al excitation make a major contribution), active atoms and radicals, and
UV photons. Each component of the chemically active plasma plays its
own specific role in plasma-chemical kinetics. Electrons, for example, are
usually first to receive the energy from an electric field and then distribute
it between other plasma components and specific degrees of freedom of
the system. Changing parameters of the electron gas (density, tempera-
ture, electron energy distribution function) often permit control and opti-
mization of plasma-chemical processes. Ions are charged heavy particles,
that are able to make a significant contribution to plasma-chemical kinet-
ics either due to their high energy (as in the case of sputtering and reac-
tive ion etching) or due to their ability to suppress activation barriers of
chemical reactions. This second feature of plasma ions results in the so-
called ion or plasma catalysis, which is particularly essential in plasma-
assisted ignition and flame stabilization, fuel conversion, hydrogen pro-
duction, exhaust gas cleaning, and even in the direct plasma treatment of
living tissue.
The vibrational excitation of molecules often makes a major contri-
bution to plasma-chemical kinetics because the plasma electrons with en-
ergies around 1 eV primarily transfer most of the energy in such gases as
N2, CO, CO2, H2, and so forth into vibrational excitation. Stimulation of
plasma-chemical processes through vibrational excitation permits the
highest values of energy efficiency to be reached. Electronic excitation of
atoms and molecules can also play a significant role, especially when the
lifetime of the excited particles is quite long (as in the case of metastable
electronically excited atoms and molecules). As an example, we can men-
tion plasma-generated metastable electronically excited oxygen molecules
O2 (1  g) (singlet oxygen), which effectively participate in the plasma-

68
stimulated oxidation process in polymer processing, and biological and
medical applications.
The contribution of atoms and radicals is obviously significant. As
an example, we can point out that O atoms and OH radicals effectively
generated in atmospheric air discharges which play a key role in numer-
ous plasma-stimulated oxidation processes. Plasma-generated photons
play a key role in a wide range of applications, from plasma light sources
to UV sterilization of water.
Plasma is not only a multi-component system, but often a very non-
equilibrium one. Concentrations of the active species described earlier
can exceed those of quasi-equilibrium systems by many orders of magni-
tude at the same gas temperature.
The successful control of plasma permits chemical processes to be
directed in a desired direction, selectively, and through an optimal mech-
anism. Control of a plasma-chemical system requires detailed understand-
ing of elementary processes and the kinetics of the chemically active
plasma. The major fundamentals of plasma physics, elementary processes
in plasma was discussed in Chapters 1 and 2, and plasma kinetics are to
be discussed in Chapters 3.

3.1 Chemical kinetics

Kinetics is the study of the rates of chemical processes in an effort to

understand what it is that influences these rates and to develop theories
which can be used to predict them. A knowledge of reaction rates has
many practical applications, for example in designing an industrial pro-
cess, in understanding the complex dynamics of the atmosphere and in
understanding the intricate interplay of the chemical reactions that are the
basis of life.
At a more fundamental level we want to understand what happens to
the molecules in a chemical reaction – that is what happens in a single re-
active encounter between two reagent molecules. By understanding this
we may be able to develop theories that can be used to predict the out-
come and rate of reactions.
For more detail on this section Principles of Chemical Kinetics by J.
E. House (Elsevier 2007) [29] is a good source.

3.1.1 Rates of reactions

The rate of a chemical reaction is expressed as a change in concen-

tration of some species with time. Therefore, the dimensions of the rate

69
must be those of concentration divided by time (moles=liter/sec,
moles=liter/min, etc.). A reaction that can be written as
A→B (88)
has a rate that can be expressed either in terms of the disappearance
of A or the appearance of B. Because the concentration of A is decreasing
as A is consumed, the rate is expressed as -d[A]/dt. Because the concen-
tration of B is increasing with time, the rate is expressed as +d[B]/dt. The
mathematical equation relating concentrations and time is called the rate
equation or the rate law. The relationships between the concentrations of
A and B with time are represented graphically in Fig. 25 for a first-order
reaction in which [A]o is 1.00 M and k =0:050 min-1.
If we consider a reaction that can be shown as
aA + bB → cC + dD (89)

Fig. 25. Change in concentration of A and B for the reaction A → B

The rate law will usually be represented in terms of a constant times

some function of the concentrations of A and B, and it can usually be
written in the form

Rate = k[A]x[B]y (90)

where x and y are the exponents on the concentrations of A and B,
respectively. In this rate law, k is called the rate constant and the expo-
nents x and y are called the order of the reaction with respect to A and B,
respectively. As will be described later, the exponents x and y may or may
not be the same as the balancing coefficients a and b in Eq. (89). The
overall order of the reaction is the sum of the exponents x and y. Thus, we
speak of a second-order reaction, a third-order reaction, etc., when the

70
sum of the exponents in the rate law is 2, 3, etc., respectively. These ex-
ponents can usually be established by studying the reaction using differ-
ent initial concentrations of A and B. When this is done, it is possible to
determine if doubling the concentration of A doubles the rate of the reac-
tion. If it does, then the reaction must be first-order in A, and the value of
x is 1. However, if doubling the concentration of A quadruples the rate, it
is clear that [A] must have an exponent of 2, and the reaction is second-
order in A. One very important point to remember is that there is no nec-
essary correlation between the balancing coefficients in the chemical
equation and the exponents in the rate law. They may be the same, but
one can not assume that they will be without studying the rate of the reac-
tion.
If a reaction takes place in a series of steps, a study of the rate of the
reaction gives information about the slowest step of the reaction. We can
see an analogy to this in the following illustration that involves the flow
of water,

If we study the rate of flow of water through this system of short

pipes, information will be obtained about the flow of water through a 1cm
pipe since the 3 cm and 5 cm pipes do not normally offer as much re-
sistance to flow as does the 1cm pipe. Therefore, in the language of chem-
ical kinetics, the 1cm pipe represents the rate-determining step.
Suppose we have a chemical reaction that can be written as
2A + B → Products (91)
and let us also suppose that the reaction takes place in steps that can be
written as
A + B → C (slow) (92)
C + A → Products (fast) (93)

The amount of C (known as an intermediate) that is present at any

time limits the rate of the overall reaction. Note that the sum of Eqs. (92)
and (93) gives the overall reaction that was shown in Eq. (91). Note also
that the formation of C depends on the reaction of one molecule of A and
one of B. That process will likely have a rate that depends on [A]1 and
[B]1. Therefore, even though the balanced overall equation involves two
molecules of A, the slow step involves only one molecule of A. As a re-
sult, formation of products follows a rate law that is of the form Rate =

71
k[A][B], and the reaction is second-order (first-order in A and first-order
in B). It should be apparent that we can write the rate law directly from
the balanced equation only if the reaction takes place in a single step. If
the reaction takes place in a series of steps, a rate study will give infor-
mation about steps up to and including the slowest step, and the rate law
will be determined by that step.

3.1.2 Dependence of rates on concentration

In this section, we will examine the details of some rate laws that
depend on the concentration of reactants in some simple way. Although
many complicated cases are well known, there are also a great many reac-
tions for which the dependence on concentration is first-order, second-
order, or zero-order.

3.1.2.1 First-Order

Suppose a reaction can be written as

A→B (94)
and that the reaction follows a rate law of the form

d [ A]
Rate  k[ A]1   (95)
dt
This equation can be rearranged to give
d [ A]
  kdt (96)
[ A]
Equation (96) can be integrated but it should be integrated between
the limits of time =0 and time equal to t while the concentration varies
from the initial concentration [A]o at time zero to [A] at the later time.
This can be shown as
[ A] t
d [ A]
   k  dt (97)
[ A]0
[ A] o

When the integration is performed, we obtain

[ A]0 [ A]0 kt
ln  kt or log  t (98)
[ A] [ A] 2.303
If the equation involving natural logarithms is considered, it can be
written in the form
ln[ A]0  ln[ A]  kt (99)
or

72
 ln[ A]  ln[ A]0  kt (100)
y  b  mx (101)
It must be remembered that [A]o, the initial concentration of A, has
some fixed value so it is a constant. Therefore, Eq. (100) can be put in the
form of a linear equation where y=ln[A], m=-k, and b=ln [A]o. A graph of
ln[A] versus t will be linear with a slope of -k. In order to test this rate
law, it is necessary to have data for the reaction which consists of the
concentration of A determined as a function of time. This suggests that in
order to determine the concentration of some species, in this case A, sim-
ple, reliable, and rapid analytical methods are usually sought. Additional-
ly, one must measure time, which is not usually a problem unless the re-
action is a very rapid one.
It may be possible for the concentration of a reactant or product to be
determined directly within the reaction mixture, but in other cases a sam-
ple must be removed for the analysis to be completed. The time necessary
to remove a sample from the reaction mixture is usually negligibly short
compared to the reaction time being measured. What is usually done for a
reaction carried out in solution is to set up the reaction in a vessel that is
held in a constant temperature bath so that fluctuations in temperature will
not cause changes in the rate of the reaction. Then the reaction is started,
and the concentration of the reactant (A in this case) is determined at se-
lected times so that a graph of ln[A] versus time can be made or the data
analyzed numerically. If a linear relationship provides the best fit to the
data, it is concluded that the reaction obeys a first-order rate law. Graph-
ical representation of this rate law is shown in Fig. 26 for an initial con-
centration of A of 1.00 M and k=0:020 min-1. In this case, the slope of the
line is - k, so the kinetic data can be used to determine k graphically or by
means of linear regression using numerical methods to determine the
slope of the line.

73
 Fig. 26. First-order plot for A → B with [A]o = 1:00 M and
k = 0:020 min -1

The units on k in the first-order rate law are in terms of time-1. The
left-hand side of Eq. (98) has [concentration]/[concentration], which
causes the units to cancel. However, the right-hand side of the equation
will be dimensionally correct only if k has the units of time-1, because on-
ly then will kt have no units
The equation
ln[ A]  ln[ A]0  kt (102)
can also be written in the form

[ A]  [ A]0 e  kt (103)
From this equation, it can be seen that the concentration of A de-
creases with time in an exponential way. Such a relationship is sometimes
referred to as an exponential decay.
Radioactive decay processes follow a first-order rate law. The rate of
decay is proportional to the amount of material present, so doubling the
amount of radioactive material doubles the measured counting rate of de-
cay products. When the amount of material remaining is one-half of the
original amount, the time expired is called the half-life. We can calculate
the half-life easily using Eq. (98). At the point where the time elapsed is
equal to one half-life, t=t1/2, the concentration of A is one-half the initial
concentration or [A]o=2. Therefore, we can write

74
 A0  ln
A0  kt1 / 2  ln 2  0.693 (104)
ln
A A0
2
The half-life is then given as
0.693
t1 / 2  (105)
k
and it will have units that depend on the units on k. For example, if k
is in hr-1, then the half-life will be given in hours, etc. Note that for a pro-
cess that follows a first-order rate law, the half-life is independent of the
initial concentration of the reactant. For example, in radioactive decay the
half-life is independent of the amount of starting nuclide. This means that
if a sample initially contains 1000 atoms of radioactive material, the half-
life is exactly the same as when there are 5000 atoms initially present. It
is easy to see that after one half-life the amount of material remaining is
one-half of the original; after two half-lives, the amount remaining is one-
fourth of the original; after three half-lives, the amount remaining
is one-eighth of the original, etc. This is illustrated graphically as shown
in Fig. 27.

Fig. 27. Half-life determination for a first-order process with

[A]o = 1:00 M and k = 0:020 min-1

While the term half-life might more commonly be applied to pro-

cesses involving radioactivity, it is just as appropriate to speak of the half-

75
life of a chemical reaction as the time necessary for the concentration of
some reactant to fall to one-half of its initial value.

3.1.2.2 Second-Order

A reaction that is second-order in one reactant or component obeys

the rate law

d A
Rate  k A  
2
(106)
dt
Such a rate law might result from a reaction that can be written as

2 A  Pr oducts (107)
However, as we have seen, the rate law cannot always be written
from the balanced equation for the reaction. If we rearrange Eq. (106), we
have

 d A
 kdt (108)
A2

If the equation is integrated between limits on concentration of [A]o

at t =0 and [A] at time t, we have

d A
 A t

 A A2  k 0 dt (109)
0

Performing the integration gives the integrated rate law

1 1
  kt (110)
A A0

Since the initial concentration of A is a constant, the equation can be put

in the form of a linear equation,
1 1
 kt  (111)
A A0

As shown in Fig. 28, a plot of 1/[A] versus time should be a straight

line with a slope of k and an intercept of 1/[A]o if the reaction follows the
second-order rate law. The units on each side of Eq. (111) must be

76
1/concentration. If concentration is expressed in mole/liter, then
1/concentration will have units of liter/mole. From this we find that the
units on k must be liter/mole time or M-1 time-1 so that kt will have units
M-1.

Fig. 28. A second-order rate plot for A → B with [A]o = 0:50 M and k =
0.040 liter/mol min

The half-life for a reaction that follows a second-order rate law can
be easily calculated. After a reaction time equal to one half-life, the con-
centration of A will have decreased to one-half its original value. That is,
[A] =[A]o/2, so this value can be substituted for [A] in Eq. (110) to give

1 1
A0 A0  kt1/ 2
 (112)
2
Removing the complex fraction gives

2 1 1
  kt1 / 2  (113)
A0 A0 A0
Therefore, solving for t1/2 gives

77
 1
t1 / 2 
k Ao
Here we see a major difference between a reaction that follows a
second-order rate law and one that follows a first-order rate law. For a
first-order reaction, the half-life is independent of the initial concentration
of the reactant, but in the case of a second-order reaction, the half-life is
inversely proportional to the initial concentration of the reactant.

3.1.3 Dependence of Rate on Temperature

In order for molecules to be transformed from reactants to products,

it is necessary that they pass through some energy state that is higher than
that corresponding to either the reactants or products. For example, it
might be necessary to bend or stretch some bonds in the reactant molecule
before it is transformed into a product molecule.
The essential idea is that a state of higher energy must be populated
as a reaction occurs. This is illustrated by the energy diagram shown in
Fig. 29. Such a situation should immediately suggest that the Boltzmann
Distribution Law may provide a basis for the explanation because that law
governs the population of states of unequal energy. In the case illustrated
in the figure, denotes the high-energy state, which is called the transition
state or the activated complex. The height of the energy barrier over
which the reactants must pass on the way to becoming products is known
as the activation energy.

Fig. 29. The energy profile for a chemical reaction

78
 The example illustrated in the figure represents an exothermic reac-
tion because the overall energy change is negative since the products have
a lower energy than the reactants. When the various rate laws are inspect-
ed, we see that only k can be a function of temperature because the con-
centrations remain constant or very nearly so as the temperature changes
only a small amount. Therefore, it is the rate constant that incorporates
information about the effect of temperature on the rate of a reaction.
There are several types of behavior exhibited when the rates of re-
actions are studied as a function of temperature. Three of the most com-
mon variations in rate with temperature are shown in Fig. 30.
The first case shows the variation followed by most reactions, that
of an exponentially increasing rate as temperature increases. The second
shows the behavior of some material that becomes explosive at a certain
temperature. At temperatures below the explosive limit, the rate is essen-
tially unaffected by the temperature. Then, as the temperature is reached
at which the material becomes explosive, the rate increases enormously as
the temperature is increased only very slightly. In the third case, we see
the variation in rate of reaction that is characteristic of many biological
processes. For example, reactions involving enzymes (biological cata-
lysts) frequently increase in rate up to a certain temperature and then de-
crease in rate at higher temperatures. Enzymes are protein materials that
can change conformation or become denatured at high temperatures.

Fig. 30. Some of the ways in which reaction rates vary with temperature

Svante August Arrhenius suggested in the late 1800s that the rates
of most reactions vary with temperature (as shown in Fig. 30 a) in such a
way that
k  Ae  E / RT
a
(114)
where k is the rate constant, A is the frequency factor (or pre-
exponential factor), R is the molar gas constant, Ea is the activation ener-
gy, and T is the temperature (K). This equation is generally referred to as
the Arrhenius equation. If we take the natural logarithm of both sides of
Eq. (114), we obtain

79
 Ea
ln k    ln A (115)
RT
By rearrangement, this equation can be put in the form of a straight
line,

Ea 1
ln k     ln A (116)
R T
y  m x  b
Therefore, a plot of ln k versus 1/T can be made or linear regression
performed after the rate constants have been determined for a reaction
carried out at several temperatures. The slope of the line is - Ea/R and the
intercept is ln A. Such a graph, like that shown in Fig. 31, is often called
an Arrhenius plot. It is from the slope, determined either numerically or
graphically, that the activation energy is determined.
For a particular reaction, the following rate constants were obtained
when the reaction was studied at a series of temperatures which yielded
the data shown on the next page.

Fig. 31. An Arrhenius plot constructed using the data in the text

In determining the activation energy from an Arrhenius plot, it is

important to observe several precautions. For example, if the reaction is
being studied at 300 K, the value of 1/T will be 0:00333 K-1. If the reac-
tion is then studied at 305 K, the value of 1/T will be 0:00328K-1.
Such small differences in 1/T make it very difficult to determine the
slope of the line accurately, especially if the temperature has not been
controlled very accurately. Consequently, it is desirable to study a reac-
tion over as large a range of temperature as possible and to use several
temperatures within that range in order to minimize errors. For most reac-

80
tions, the temperature range in which the reaction can be studied is rather
limited because at low temperatures the reaction will be very slow and at
high temperatures the reaction will be very fast. Therefore, it is generally
desired to study a reaction over a range of at least 20–250C.
If the rate constant for a reaction is determined at only two tempera-
tures, it is still possible to evaluate the activation energy, but such a case
is not nearly as desirable as the procedure described earlier. Small errors
in the rate constants will cause inaccuracy in the activation energy deter-
mined, because all of the errors in placing the line are present in only two
points. More data would be needed to ‘‘average out’’ the error in the val-
ue of any one rate constant. If k1 is the rate constant at T1 and k2 is the rate
constant at T2, we can write the Arrhenius equation as

Ea
ln k1  ln A  (117)
RT1
E
ln k2  ln A  a (118)
RT2
Subtracting the equation for ln k2 from that giving ln k1 gives

Ea E
ln k1  ln k 2  (ln A  )  (ln A  a ) (119)
RT1 RT2

We can simplify this equation to obtain

k 2 Ea (T2  T1 )
ln  (120)
k1 RT1T2

3.2 The object and the main features of the plasma chemistry

Plasma chemistry studies the kinetics and mechanisms of chemical

reactions and physicochemical processes in low-temperature plasmas. It
should be emphasized that in the plasma chemistry, the term "plasma" is
understood as a substance containing not only the charged particles, but
molecules in the ground electronic state with an excess (non-equilibrium)
internal energy.
In plasma chemistry it is particularly important to discriminate low-
temperature plasma on the quasi-equilibrium that exists at pressures
above atmospheric, and characterized by a common temperature for all
particles, and the non-equilibrium in which the temperature of free elec-
trons is much higher than the temperature of heavy particles (molecules,
ions). This division is due to the fact that the kinetics of the quasi-

81
equilibrium plasma chemical processes is determined only by a high tem-
perature of the interacting particles, whereas the specific of the non-
equilibrium plasma-chemical processes is due to a large contribution of
the internal energy of the molecules into their chemical activity.
A low-temperature plasma is plasma with a temperature of 103-105 K
and the degree of ionization of 10-6-0.1, obtained from electric, high-
frequency and microwave gas discharges, in shock tubes, adiabatic com-
pression machines and by other methods.
Plasma chemistry today is a rapidly expanding area of science and
engineering, with applications widely spread from micro-fabrication in
electronics to making protective coatings for aircrafts, from treatment of
polymer fibers and films before painting to medical cauterization for
stopping blood and wound treatment, and from production of ozone to
plasma TVs.
Applications of plasma technologies today are numerous and involve
many industries. High-energy efficiency (energy cost with respect to the
minimum determined by thermodynamics), high specific productivity
(productivity per unit volume of reactor), and high selectivity may be
achieved in plasmas for a wide range of chemical processes. As an exam-
ple, for CO2 dissociation in non-equilibrium plasma under supersonic
flow conditions, it is possible to selectively introduce up to 90 % of the
total discharge power in CO production when the vibrational temperature
is about 4000 K and the translational temperature is only about 100 K.
The specific productivity of such a supersonic reactor achieves 1,000,000
L/h, with power levels up to 1 MW. This plasma process has been exam-
ined for fuel production on Mars, where the atmosphere mostly consists
of CO2. On the Earth, it was applied as a plasma stage in a two-step pro-
cess for hydrogen production from water.

3.2.1 Quasi-equilibrium plasma-chemical processes

A plasma in which the particle states (the velocity distribution and

internal states) can be well described by the Maxwell-Boltzmann distribu-
tion with a single temperature is called a quasi-equilibrium plasma. In
equilibrium plasmas this is fulfilled very strictly. This distribution is often
called an equilibrium distribution. In this case, the plasma composition
(concentrations of ions, electrons, atoms, radicals, etc.) is equilibrium and
determined, except for the initial conditions, by a single parameter - the
temperature T, which is the same for all particles. Usually, quasi-
equilibrium plasmas are formed at pressures close to atmospheric. Since
in the gas-discharge plasmas the energy from an external source is deliv-

82
ered to a heavy component by the electrons, the full balance in such a sys-
tem cannot be achieved. To transfer energy to the gas the electron energy
(temperature) is always greater than the temperature of heavy particles
(unlikely to the plasma which is produced in adiabatic compression).
Plasma-chemical processes carried out in such conditions, are considered
as quasi-equilibrium processes.
Despite the equilibrium character of the physical and chemical pro-
cesses in plasmas formed in the temperature range of 3000-5000 K the
chemical reactions occur with a very high speed, so that their characteris-
tic times are of the same order that the times of heat and mass transfer. As
a result, the processes can be transformed from the kinetic region of per-
colation to the diffusion. The reaction mechanisms are undergoing signif-
icant changes, in particular, can enter the reaction with the walls of a reac-
tor. The plasma in this case is considered only as an effective energy car-
rier.

3.2.2 Non-equilibrium plasma-chemical processes

As it was shown above in case of thermodynamic equilibrium the

energy supplied, used to initiate chemical reactions, is very inefficient.
There is a wide distribution of the total energy of molecules and a broad
distribution of the internal energy of molecule.
A fundamentally different situation takes place when the initiation of
chemical reactions occurs in plasma. The energy of the electric field of
the gas discharge is transferred to the electrons, which give it to other
plasma particles while colliding with them. In elastic collisions due a low
mass of electrons the efficiency of electron energy transfer into kinetic
energy of heavy particles is small and proportional to the ratio of the elec-
tron mass to molecule mass, i.e., does not exceed 0.05 % for hydrogen at-
om.
Therefore, most of the electron energy is transferred to internal de-
grees of freedom of the molecules, creating an unbalanced distribution of
the energy input. In turn, the ions and excited particles (especially in the
metastable electronic states) may have a decisive influence on the mecha-
nism and kinetics of plasma chemical reactions. The rates of reactions
with participation of excited particles, ions and radicals is much higher
than the rate of formation of these particles, so the overall rate of chemi-
cal reactions is limited by the stage of formation.
Low-temperature plasma is bounded by solid walls of the vessel in
which it is generated, and at reduced pressures, the characteristic time of

83
particle diffusion to the walls of the vessel approaches the characteristic
times of chemical reactions. As a result, the role of heterogeneous (on the
walls), physic-chemical processes is increasing to such an extent that they
should be considered when analyzing the mechanisms and kinetics of
plasma chemical reactions.
In non-equilibrium plasma the distribution of velocity and the inter-
nal states are different, in general, from that in Maxwell-Boltzmann dis-
tribution. In this situation we should use a set of Boltzmann equilibrium
distributions with different temperatures for different particle states. The
internal state of the particles are divided into rotational, vibrational, and
electronic, the populations of which are described by the equilibrium
Boltzmann distribution with temperatures Tr, Tv, Te, respectively. If the
electron velocity distribution is Maxwellian with the temperature Te, and
the distribution of atoms and molecules by velocities (a translational de-
gree of freedom) is also Maxwellian with the temperature Tg (gas temper-
ature), then, for a non-equilibrium plasma
Tg ≠ Tr ≠ Tv ≠ Te
Typically, this situation is observed in the plasma at low pressures
(less than 100 Torr). It often happens that Tg ≈ Tr <Tv <Te, and if Tg and
Tr are close to room temperature (300K), the electron temperature corre-
sponds to the average energy of several electron volts (1 eV = 11600K).
Non-equilibrium thermodynamics is a consequence of "openness" of the
system. Plasma-chemical processes carried out in such conditions, are de-
fined as non-equilibrium processes.

3.3 Elementary Plasma-Chemical Reactions

The yield of a total plasma-chemical process is due to synergistic

contributions of numerous different elementary reactions taking place
simultaneously in a discharge system. The sequence of transformations of
initial chemical substances and electric energy into products and thermal
energy is usually referred to as the mechanism of the plasma-chemical
process. Elementary reaction rates are determined by the micro-kinetic
characteristics of individual reactive collisions (like, for example, reaction
cross sections or elementary reaction probabilities) as well as by relevant
kinetic distribution functions (like the electron energy distribution func-
tion [EEDF], or population function of excited molecular states).
The elementary reaction rate is actually a result of integration of the
reaction cross section or probability over the relevant distribution function
and characterizes the energy or excitation state of reactants. We will focus

84
in this chapter mostly on the micro-kinetics of the elementary reactions –
on their cross sections and probabilities.

3.3.1 Ionization processes

Plasma is an ionized gas. The key process in plasma is ionization,

which means conversion of neutral atoms or molecules into electrons and
positive ions. Thus, ionization is the first elementary plasma-chemical
processes to be considered.
Usually the number densities of electrons and positive ions are equal
or close in quasineutral plasmas, but in “electronegative” gases (like O2,
Cl2, SF6, UF6, TiCl4, etc.) with high electron affinity, negative ions are al-
so effectively formed. Because ionization is the first plasma-chemical re-
action to consider, let us first briefly discuss the main features of elemen-
tary charged particles in plasma and their elastic and inelastic collisions.
Electrons are first in getting energy from electric fields, because of their
low mass and high mobility. Electrons then transmit the energy to all oth-
er plasma components, providing energy for ionization, excitation, disso-
ciation, and other plasma-chemical processes. The rates of such processes
depend on how many electrons have enough energy to do the job. It can
be described by means of the electron energy distribution function
(EEDF) f (ε), which is the probability density for an electron to have en-
ergy ε.
The EEDF strongly depends on electric field and gas composition in
plasma (especially in non-thermal discharges) and often can be very far
from the equilibrium distribution.
Sometimes, however (even in non-equilibrium plasmas), the EEDF
is determined mostly by the electron temperature Te and, therefore, can be
described by the quasi-equilibrium Maxwell-Boltzmann distribution func-
tion:
3
f ( )  2  /  (kTe ) exp( / kTe ) (121)
where k is a Boltzmann constant (when temperature is given in ener-
gy units, then k = 1 and can be omitted). The mean electron energy, which
is the first moment of the distribution function, in this case is proportional
to temperature in the conventional way:

3
   f ( )d  Te (122)
0
2
Numerically, in most plasmas under consideration, the mean electron
energy is from 1 to 5 eV.

85
 Atoms or molecules lose their electrons in the ionization process and
form positive ions. In hot thermonuclear plasmas, the ions are multi-
charged, but in relatively cold plasmas of technological interest their
charge is usually equal to +1e (1.6 · 10−19 C). Ions are heavy particles, so
usually they cannot receive high energy directly from electric fields be-
cause of intensive collisional energy exchange with other plasma compo-
nents. The collisional nature of the energy transfer results in the fact that
the ion energy distribution function at lower pressures is often not far
from the Maxwellian one (120), with an ion temperature Ti close to neu-
tral gas temperature T0.
Electron attachment leads to formation of negative ions with
charge−1e (1.6 10−19 C). Attachment of another electron and formation of
multi-charged negative ions is actually impossible in the gas phase be-
cause of electric repulsion. Negative ions are also heavy particles, so usu-
ally their energy balance is due not to electric fields but to collisional pro-
cesses. The energy distribution functions for negative ions, similar to that
for the positive ones, are not far from the Maxwellian distributions (121)
at pressures that are not too low. Their temperatures in this case are also
close to those of a neutral gas.

3.3.2 Classification of Ionization Processes

Mechanisms of ionization can be very different in different plasma-

chemical systems and may be subdivided generally into the following five
groups.
1 Direct ionization by electron impact is ionization of neutral and
previously unexcited atoms, radicals, or molecules by an electron whose
energy is high enough to provide the ionization act in one collision. These
processes are the most important in cold or non-thermal discharges, where
electric fields and therefore electron energies are quite high, but where the
excitation level of neutral species is relatively moderate.
2 Stepwise ionization by electron impact is ionization of preliminary
excited neutral species. These processes are important mostly in thermal
or energy-intense discharges, when the ionization degree ne/n0 as well as
the concentration of highly excited neutral species is quite high.
3 Ionization by collision of heavy particles takes place during ion–
moleculeor ion–atom collisions, as well as in collisions of electronically
or vibrationally excited species, when the total energyof the collision
partners exceeds the ionization potential. The chemical energy of collid-
ing neutral species can also contribute to ionization in the so-called asso-
ciative ionization processes.

86
 4 Photo-ionization takes place in collisions of neutrals with photons,
which result in the formation of an electron–ion pair. Photo-ionization is
mostly important in thermal plasmas and in some mechanisms of propa-
gation of non-thermal discharges.
5 Surface ionization (electron emission) is provided by electron, ion,
and photon collisions with different surfaces or simply by surface heating.
This ionization mechanism is quite different from the first four.

3.3.3 Direct Ionization by Electron Impact

Direct ionization is a result of the interaction of an incident electron,

having a high energy ε, with a valence electron of a preliminary neutral
atom or molecule. Ionization occurs when the energy  transferred to
the valence electron exceeds the ionization potential I. Analysis of the el-
ementary process obviously requires a quantum mechanical considera-
tion, but a clear physical understanding can be obtained from the classical
Thomson model (1912). The valence electron is assumed to be at rest in
this model, and interaction of the two colliding electrons with the rest of
the atom is neglected. The differential cross section of the incident elec-
tron scattering with energy transfer  to the valence electron can be
defined by the Rutherford formula:
1 e 4
d i  d ( )
(4 0 ) 2  ( ) 2 (123)
When the transferred energy exceeds the ionization potential,  ≥ I,
direct ionization takes place. Thus, integration of (120) over  ≥ 1 gives
an expression for the ionization cross section by direct electron impact,
known as the Thomson formula:
1 e 4 1 1
i  (  )
(4 0 ) 2  I  (124)

Expression (121) should be multiplied, in general, by the number of

valence electrons, Zv. At high electron energies, ε » I, the Thomson cross
section (121) is falling as σi ∼1/ε. Quantum-mechanical treatment gives a
more accurate but close asymptotic approximation: σi ∼ln ε/ε. When ε=2I,
the Thomson cross section reaches the maximum value:
1 e 4
 max i  (125)
(4 0 ) 2 4 I 2
The Thomson formula can be rewritten taking into account the kinet-
ic energy εv of the valence electron:

87
 1 e 4 1 1 2 v 1 1
i  ( _  (  )) (126)
(4 0 ) 2   I 3 I2 2
The Thomson formula (124) agrees with (123), assuming the valence
electron is at rest and εv =0. An interesting variation of the Thomson for-
mula (124) can be obtained assuming a Coulomb interaction of the va-
lence electron with the rest of the atom and taking εv = I:
1 e 4 5 1 2 I
i  (   ) (127)
(4 0 ) 2  3I  3 2
All the modifications of the Thomson formula can be combined us-
ing the generalized function f (ε/I) common for all atoms:
1 e 2 
i  Zv f ( ) (128)
(4 0 ) 2
I 2
I

Fig. 32. Molecular and ionic terms illustrating dissociative ionization of

molecules AB

3.3.4 Stepwise ionization by electron impact

When the plasma density and, therefore, the concentration of excited

neutrals are high enough, the energy (I) necessary for ionization can be
provided in two different ways. First, like in the case of direct ionization,
it could be provided by the energy of plasma electrons. Second, the high

88
energy of preliminary electronic excitation of neutrals can be converted in
the ionization act, which is called stepwise ionization. If the level of elec-
tronic excitation is high enough, stepwise ionization is much faster than
direct ionization, because the statistical weight of electronically excited
neutrals is greater than that of free plasma electrons. In other words, when
Te « I, the probability of obtaining high ionization energy I is much lower
for free plasma electrons (direct ionization) than for excited atoms and
molecules (stepwise ionization).
In contrast to direct ionization, the stepwise process includes several
steps to provide the ionization event. At first, electron–neutral collisions
prepare highly excited species, and then a final collision with a relatively
low-energy electron provides the actual ionization event.
In thermodynamic equilibrium, the ionization process
+
e+A→A +e+e is inverse to the three-body recombination
A++e+e→A∗+e→A+e, which proceeds through a set of excited states.
According to the principle of detailed equilibrium, the ionization process
e+A→A++e+e should go through the set of electronically excited states
as well, which means the ionization should be a stepwise process. How-
ever, this conclusion can only hold true for quasi-equilibrium or thermal
plasmas.
A description of the stepwise ionization and the calculation of rate
coefficient can be found in Fridman (2008) [1]

3.3.5 Ionization in collisions of heavy particles: adiabatic princi-

ple and Massey parameter

An electron with a kinetic energy only slightly exceeding the ioniza-

tion potential is often quite effective in producing the ionization event;
however, this is not true for ionization by collisions of heavy particles –
ions and neutrals. Even when they have enough kinetic energy, they actu-
ally cannot provide ionization, because their velocities are much less than
those of electrons in atoms. Even if it has enough energy, a heavy particle
is very often unable to transfer this energy to an electron inside an atom
because the process is far from resonant. Such slow motion may be
termed “adiabatic” and is unable to transfer energy to a fast-moving parti-
cle. The adiabatic principle can be explained in terms of the relationship
between low interaction frequency ωint =α  and high frequency of elec-
tron transfers in atoms, ωtr =ΔE/  . Here 1/α is a characteristic size of the
interacting neutral particles,  is their velocity, and ΔE is a change of
electron energy in the atom during the interaction. Only fast Fourier com-
ponents of the slow interaction potential between particles with frequen-

89
cies about ωtr =ΔE/  provide the energy transfer between the interacting
particles. The relative weigh to probability of these fast Fourier compo-
nents is very low if ωtr » ωint; numerically it is about exp(−ωtr/ωint). As a
result, the probability PEnTr of energy transfer processes (including the
ionization process under consideration) are usually proportional to the so-
called Massey parameter:
tr E
PEnTr  exp( )  exp( )  exp( PMa ) (129)
int 
where PMa  E /  is the adiabatic Massey parameter. If PMa »1,
the process of energy transfer is adiabatic and its probability is exponen-
tially low. It takes place in collisions of heavy neutrals and ions.

3.3.6 Penning ionization effect and associative ionization

If the electronic excitation energy of a metastable atom A∗ exceeds

the ionization potential of another atom B, their collision can lead to an
act of ionization, the so-called Penning ionization.
The Penning ionization usually proceeds through the intermediate
formation of an unstable excited quasi molecule in the state of auto-
ionization; cross sections of the process can be very high. Cross sections
for the Penning ionization of N2, CO2, Xe, and Ar by metastable helium
atoms He(23S)with an excitation energy of 19.8eV reach gas-kinetic val-
uesof about 10−15 cm2. Similar cross sections can be attained in collisions
of metastable neon atoms (excitation energy 16.6 eV) with argon atoms
(ionization potential 15.8 eV).
An exceptionally high cross section of 1.4 10−14 cm2 (Smirnov,
1974) can be achieved in Penning ionization of mercury atoms (ionization
potential 10.4 eV) by collisions with the metastable helium atoms He(23S,
19.8 eV). If the total electronic excitation energy of colliding particles is
not sufficient, ionization is still possible when heavy species stick to each
other, forming a molecular ion. Such a process is called the associative
ionization; for example,
Hg (6 3 P1 , E  4.9eV )  Hg (6 3 P0, E  4.7eV )  Hg 2  e (130)
The total electronic excitation energy here (9.6 eV) is less than the
ionization potential of mercury atoms (10.4 eV) but higher than ionization
potential for a Hg−2 molecule. This is actually the main “trick” of the as-
sociative ionization. Cross sections of the associative ionization (similar
to the Penning ionization) can be quite high and close to the gas-kinetic
value (about 10−15 cm2). Associative ionization A∗+B→AB++e takes place
effectively when there is a crossing of the electronic energy term of the

90
colliding particles with an electronic energy term of the molecular ion
AB+ and, as a result, the process is non-adiabatic. Such a situation takes
place only for a limited number of excited species.

3.4 Elementary plasma-chemical reactions of positive ions

Positive ions are obviously major players in plasma-chemical pro-

cesses. Their exothermic reactions with neutrals usually have no activa-
tion energy, which makes their contribution significant in many specific
plasma-chemical processes, particularly in plasma catalysis.
In addition to high chemical activity, the ions can have significant
kinetic energy, which determines their contribution, for example, in reac-
tive ion etching.

3.4.1 Different Mechanisms of Electron–Ion Recombination in

Plasma

Electron–ion recombination is a highly exothermic process. There-

fore, the process should have a specific channel of accumulation of the
energy released during the neutralization of a positive ion and anelectron.
Most of these channels of recombination energy consumption are related,
either to dissociation of molecules or to three-body collisions or to radia-
tion, which determines the following three major groups of mechanisms
of electron–ion recombination.
1 The fastest electron neutralization mechanism in molecular gases,
or in the presence of molecular ions, is dissociative electron–ion recom-
bination:
e  AB   ( AB)*  A  B* (131)
Recombination energy in these processes goes into dissociation of
the intermediately formed molecule ion and to excitation of the dissocia-
tion products. These processes are common for molecular gases, but they
can also be important in atomic gases because of the formation of mo-
lecular ions in the ion conversion processes:

A   A  A  A2  A (132)
2 Electron–ion neutralization is due to a three-body electron–ion re-
combination in atomic gases in the absence of molecular ions:
e  e  A  A*  e (133)
Energy excess in this case is going into the kinetic energy of a free
electron, which participates in the recombination act as “a third-body
partner”. We must remark that heavy particles (ion and neutrals) are una-

91
ble to accumulate electron recombination energy fast enough in their ki-
netic energy and, therefore, are ineffective as the third- body partner.
3 Finally, the recombination energy can be converted into radiation
in the process of radiative electron–ion recombination:
e  A  A*  A   (134)
The cross sectionof this process is relatively low and can com-
petewith the three-body recombination only when the plasma density is
not high.

3.5 Elementary plasma-chemical reactions involving negative

ions

3.5.1 Dissociative electron attachment to molecules as a major

mechanism of negative ion formation in electronegative molecular
gases

The dissociative attachment is effective when the products have pos-

itive electron affinities
e  AB  ( AB  )*  A  B  (135)
The mechanism of the process is similar to the dissociative recombi-
nation (Section 2.2.2) and proceeds by intermediate formation of an auto-
ionization state (AB−)∗. This excited state is unstable and decays, leading
either to the reverse reaction of auto-detachment (AB+e) or to the dissoci-
ation (A+B−). An electron is captured during the attachment and is not
able to provide the energybalance of the elementary process. Therefore,
the dissociative attachment is a resonant reaction requiring a quite definite
electron energy.Typical potential energy curves are presented in Fig. 33.
The process starts with a vertical transition from the AB ground state to a
repulsive state of AB− following the Frank-Condon principle [30]. Dur-
ing the repulsion, before (AB−)∗ reaches the inter section point of AB and
AB− electronic terms, the reverse auto-detachment reaction (AB+e) is
very possible. But after passing the intersection, the AB potential energy
exceeds that of AB−and further repulsion results in dissociation (A+B−).
Some halogen compounds have an electron affinity of a product ex-
ceeding the dissociation energy; corresponding potential energy curves
are presented in Fig. 34.

92
Fig. 33. Elementary process of dissoci- Fig. 34. Elementary process of disso-
ative attachment in the case of low ciative attachment in the case of high
electron affinity electron affinity

In this case, in contrast to Fig. 33, even very-low-energy electrons

provide dissociation. The intersection point of AB and AB− electronic
terms (Fig. 34) is actually located in this case inside of the so-called geo-
metric size of the dissociating molecule. As a result, during the repulsion
of (AB−)∗, the probability of the reverse auto-detachment reaction
(AB+e)is very low and the cross section of the dissociative attachment
reaches the gas-kinetic value σ0 of about 10−16 cm2.

3.5.2 Three-body electron attachment and other mechanisms of

formation of negative ions

Electron collisions with two heavy particles also can result in the
formation of negative ions:
e  A  B  A  B (136)
The three-body electron attachment can be a principal channel for
electron losses when electron energies are not high enough for the disso-
ciative attachment, and when pressure is elevated (usually more than 0.1
atm) and the third-order kinetic processes are preferable.
In contrast to the dissociative attachment, the three-body process is
exothermic and its rate coefficient does not depend strongly on electron
temperature. Electrons are usually kinetically less effective as a third
body B because of a low degree of ionization. Atmospheric-pressure non-
thermal discharges in air are probably the most important systems, where
the three-body attachment plays a key role in the balance of charged par-
ticles:
e  O2  M  O2  M (137)
This process proceeds by the two-stage Bloch-Bradbury mechanism
(Bloch&Bradbury, 1935; Alexandrov, 1981) starting with the formation

93
of a negative ion (rate coefficient katt) in an unstable auto-ionization state
(τ is the time of collisionless detachment):

e  O2 (O2 ) *
k
att ,
(138)
The second stage of the Bloch-Bradbury mechanism includes colli-
sion with the third-body particle M (density n0), leading to relaxation and
stabilization of O−2 or collisional decay of the unstable ion:

(O2 ) *  M 
k st
O2  M (139)
 *
(O )  M O2  e  M
2
kdec
(140)
Taking into account the steady-state conditions for number density
of the intermediate excited ions (O−2)∗, the rate coefficient for the total at-
tachment process (138) can be expressed as
k att k st
k 3M  (141)
1
 (k st  k dec )n0

Usually, when the pressure is not too high, (kst +kdec)n0«τ−1, and
(141) can be simplified
k3M  k att k st (142)
Thus, the three-body attachment (137) has a third kinetic order and
depends equally on the rate coefficients of formation and stabilization of
negative ions. The latter strongly depends on the type of the third particle:
the more complicated molecule a third body (M) is, the easier it stabilizes
the (O−2 )∗ and the higher the rate coefficient k3M. Numerical values of the
total rate coefficients k3M are presented in Table 5; dependence of the total
rate coefficients on electron temperature is shown in Fig.35. For simple
estimations, when Te =1eV and T0 =300K, one can take k3M≈10−30
cm6/s.

94
 Fig. 35. Rate coefficients of electron attachment to different molecules in
three-body collisions as a function of electron temperature; molecular gas is
assumed to be at room temperature

The rate of the three-body process is greater than for dissociative at-
tachment (ka) when the gas number density exceeds a critical value n0
>ka(Te)/k3M. Numerically in oxygen, with Te =1eV,T0 =300K, the rate re-
quires n0 >1018 cm−3, or in pressure units p>30 torr. Three other mecha-
nisms of formation of negative ions are usually less significant. The first
process is polar dissociation
e  AB  A  B  e (143)

This process includes both ionization and dissociation; therefore, the

threshold energy is quite high. On the other hand, an electron is not cap-
tured here, and the process is non-resonant and is effective over a wide
range of high electron energies. For example, in molecular oxygen, the
maximum value of the cross section for polar dissociation is about 3 ·
10−19 cm2 and corresponds to an electron energy of 35 eV.
To stabilize the formation of a negative ion during electron attach-
ment, the excessive energy can be in principal emitted. As a result, the
negative ion can be formed in the process of radiative attachment:
e  M  (M  )*  M    (144)
Such an electron capture can take place at low electron energies, but
the probability of the process is very low: 10−5–10−7.Corresponding values
of attachment cross sections are about 10−21–10−23 cm2.
Finally, some electronegative polyatomic molecules like SF6 have a
negative ion state very close to a ground state (only 0.1 eV in the case of
SF6). As a result, the lifetime of such metastable negative ions is fairly
long. Such a direct one-stage attachment process is resonant and for very
low electron energies has maximum cross sections of about 10−15 cm2.

3.6 Special features of plasma-chemical reactions, their dynam-

ics and kinetics

Chemical reaction dynamics are described both by the frequency of

reacting particle collisions and by the fields affecting them. The total re-
sult of the study in the field of reaction dynamics can be presented as the
function,  (kl / ij; ) [31].
The 'elementary act' (i.e. collision dynamics) of a chemical process
in 'classical' Arrhenius chemistry occurs under the following conditions:
the mean energy of the relative motion of colliding particles (molecules)

95
is not above 0.1 eV; the population of the first vibrational level of the
ground electronic state predominates over the others; reactions of the fol-
lowing type: e  M , M   M , e  M  etc. are practically absent; reaction prod-
ucts are not usually excited; elastic collisions predominate over non-
elastic ones. In general molecules may be represented as hard elastic
spheres.
In the field of interest (plasma chemistry, radiation chemistry, etc.)
the situation is quite different: mean molecule energies are  0.1 eV; ex-
cited molecules are frequently in collision, and chemical reaction prod-
ucts produced by collisions appear often to be in excited states; collisions
of molecules with electrons and ions (as well as collisions between mole-
cules) are essential, and there is usually wide difference between transla-
tional rates and energy (in the laboratory coordinate system) of the
charged particles of small mass (~1-50 eV) and those of heavy particles
( ~ 0.03-2.3 eV), as well as in the vibrational energy of the latter
(~0.2-0.8 eV). The model of elastic spheres is therefore inapplicable [32].
The kinetics of chemical reactions describes the particular type of
molecule ensemble behaviour, and this description has a statistical char-
acter. Because of this, the kinetic description of ensemble behaviour of
various particles is defined, assuming the dynamics of collisions (for in-
stance binary ones) to be known, by molecule quantum level populations
and the function for particle distributions by translational energy as well
as by ratios of flux probabilities over various channels, for instance
M (i)  M ( j ), M (i)  M ' (k ),
where i, j, k is the totality of molecule quantum state parameters.

In classical chemistry this is not so obvious, as it produces a distribu-

tion of a Maxwell type occurring at the initial stage, which at later stages
is not (or is only slightly) disturbed by a chemical reaction; the population
is considered to be of a Boltzmann type with ground level populations
predominating.
The transitions of M (i)  M ( j ), type are therefore disregarded; all the
reacting particles have the same mean energy (temperature) giving uni-
form Maxwell distribution. In plasma chemistry the situation is different;
the initial distribution may not be of Maxwell type, and the population
may not be Boltzmannian.
Even when the initial distribution is a Maxwell type it is disturbed by
a chemical reaction, and the new distribution produced affects the reac-
tion rate; various reaction components may have distribution functions
differing in form and mean energy, which can change in the course of

96
time with various relaxation times. The transitions of M (i)  M ( j ), type can
no longer be neglected in chemical reaction descriptions; non-elastic col-
lision frequencies are not small by comparison with elastic collision fre-
quencies.
Naturally, the chemical reactions possible in a molecule, and their
rates, depend on the structure of the molecule (and for non-
monomolecular reactions also on other reaction component structures).
This, in particular, prevents the introduction of the term 'reactivity', with
respect to molecules. The concept of molecular structure can be formulat-
ed differently. For our purpose the following definition is believed to be
the best: the structure of a molecule, comprising several atoms, is a sys-
tem of its quantum levels and space distribution of the component parti-
cles.
The problem of connecting reactivity with molecular structure is of-
ten considered of paramount importance in 'classical' chemical kinetics.
However, it cannot be solved within the limits of the molecular model
used. Automatically, according to the character of the reactions described
using generalized chemical kinetics, the problem is being solved within
the latter. This, however, takes place when the vague term 'reactivity' is
substituted for ki.
2. Plasma-chemical reactions may be of the non-equilibrium or qua-
si-equilibrium types. Various types of plasma-chemical reactions are
enumerated in Table 5.

Table 5. Types of plasma-chemical reactions

Processes occurring Characteristics of the Process Charac-

in the system system teristics
Chemical reactions in Difference between vi- Stationary or re-
Non-equilibrium

non-equilibrium brational and transla- laxing

systems tional temperatures
Two subsystems, one
Maxwellian and the
other non-Maxwellian
Non-equilibrium in Two subsystems with Stationary or
equi-
libri-
Qua-

the system due to the different relaxing

um
si-

occurrence temperatures

97
 of chemical reactions Both subsystems are
non-Maxwellian (and
non-Boltzmannian)

The kinetics of plasma-chemical reactions, as seen from Table 5, can

be considered as a particular case of non-equilibrium chemical kinetics.
Plasma-chemical reactions in the gas phase (only these reactions will
be discussed further) are characterized by the following:
1. Non-elastic collisions1 are more important in 'hot' gases than in
'ordinary' (monatomic) gases, the properties of which are described well
by the kinetic theory based upon the consideration of only elastic colli-
sions. The kinetic theory also explains some properties of polyatomic and
reacting gases, assuming the non-elastic collision frequency to be small;
such a treatment fails when the gas temperature is increased sufficiently.
2. The difference between the characteristic times of various physi-
cal and chemical processes decreases and it becomes impossible to distin-
guish between the various processes as is possible in classical chemical
kinetics.
3. Chemical kinetics cannot be considered independently of physical
kinetics in the field of interest. If the reaction rate exceeds the collision
frequency for the momentum-transfer collisions, deviations from equilib-
rium conditions occur. These can lead to violations of the equilibrium
Maxwell-Boltzmann energy distribution and to a relative decrease in den-
sity of the highly energetic molecules, owing to the occurrence of chemi-
cal reactions.
The highly energetic 'tail' of the distribution cannot be maintained at
the expense of momentum-transfer collisions since the frequency of such
collisions is too small. Hence, the basic principles of the simple kinetic
theory of chemical reactions and of absolute reaction rate theory are vio-
lated.
A general solution for the distribution function disturbed by chemi-
cal reactions has not yet been found. However, this problem has been
studied in some particular cases. The effect of a fast chemical reaction
upon the Maxwell-Boltzmann distribution and the effect of the distribu-
tion thus produced upon the reaction rate were studied in Ref. [33] using
the Monte Carlo method. There are well known conditions in which trans-
1
Non-elastic electron collision cross sections for many molecular gases may be con-
sidered to have maxima both at energies of 1-2 eV and above 4--b eV due to excita-
tion of vibrational and rotational, and electronic levels, respectively.

98
lational, vibrational, and rotational temperatures differ and/or various
components of a system (electrons, ions, and neutral molecules, for in-
stance) have different temperatures, and/or a system cannot be described
at all using the concept of temperature * (non-equilibrium, stationary, and
relaxing systems).
Strictly speaking, Arrhenius-type kinetics cannot be used in these
cases and the ordinary expression for the rate of a chemical reaction is in-
applicable.
4. Multichannel processes should be considered in plasma-chemical
kinetics. From the quantum mechanical point of view (we shall consider
only two channels for simplicity), in a certain energy range two pairs of
particles (A1, B1) and (A2, B2) can exist, so that there are two independ-
ent wave functions which satisfy boundary conditions for a given system.
It is known that in a one-channel problem the S-matrix contains all the in-
formation about interaction properties of the system, while in a multi-
channel case a similar theorem has not yet been proved6. Moreover, in
order to compose the Hamiltonian using the results of scattering all com-
ponents of the S-matrix should be known for all energies.
5. The barrier-type nature of the reactions (the existence of an energy
barrier opposing the reaction), being peculiar to the Arrhenius classical
kinetics, can be completely changed in reactions under plasma conditions.
It is necessary to allow for the occurrence of chemical reactions from dif-
ferent quantum levels of the system, and for one of the subsystems as a
whole to have above-barrier energy.
6. These five special features of plasma-chemical reactions are pecu-
liar to non-equilibrium chemical processes, and chemical reactions under
low-temperature plasma conditions are particular (but still important) cas-
es of such processes. In this section some special features of reactions un-
der plasma conditions are discussed briefly. The degree of ionization is
the most important characteristic of a plasma under equilibrium condi-
tions; given the temperature and pressure the degree of plasma ionization
may be found using the Saha equation.
Ionization processes result from such factors as mutual collisions of
heavy particles at high temperatures (energies), their collisions with elec-
trons, photo-ionization, and ion-molecular reactions. A detailed descrip-
tion of all these processes under non-equilibrium conditions necessitates a
generalization of the kinetic gas theory for the plasma state. Such a gener-
alization is still under development. In low-temperature plasmas long
range collisions occur, due to electromagnetic interaction between
charged particles, in addition to ordinary short-range collisions, such as
take place in gases. Thus, very small scattering angles and consequently

99
many interactions resulting in small momentum transfer must be consid-
ered. Moreover, interaction of charged particles with external electro-
magnetic fields should also be taken into account [34,35,36]
When electron, ion, and molecular gases are not in thermal equilibri-
um there is energy exchange between electrons and neutral particles and
between electrons and ions. The latter occurs through Coulomb-type col-
lisions. Both processes may lead, either directly or through successive
steps of excitation, to chemical reactions. The mechanisms and probabili-
ties of these reactions may be studied by plasma-chemical kinetics on the
basis of a detailed analysis of all interactions occurring in a system, and
using the basic principles of non-equilibrium chemical kinetics. At pre-
sent the limited usefulness of Arrhenius kinetics, which are valid only
close to equilibrium (for example under small perturbations when a sys-
tem can still be considered in quasi-equilibrium), is evident. In other
words, this type of kinetics can be used when a single value of tempera-
ture (being the parameter of the Maxwell-Boltzmann distribution) can be
defined for the system.
Non-equilibrium distribution of reactants takes place, for instance,
when the energy is injected pulsewise into the equilibrium system, pro-
vided that the condition  i   Ch.r is satisfied, where  i is the duration of
the pulse and  Ch.r is the chemical relaxation time (pulse electrical dis-
charges, shock tubes, flash photolysis). The treatment of kinetic problems
which depend more upon the molecular interaction mechanism and mo-
lecular quantum level populations* requires far more information about
the system [37,38,39]. The importance of the effect of vibrational excita-
tion upon the rate of chemical reaction has been confirmed by the results
of experiments[40]. As the vibrational temperature of N2 molecules is in-
creased from 1000 K to 6000 K (under Ttrans = 300K) the rate constant for
the reaction O   N 2  N 2  O increases by forty times (see Fig. 36). The
same change in the translational temperature (under Tvibr = const) would
lead, in accordance with the Arrhenius equation, to a rate constant in-
crease of sixty times. The effects are comparable.

100
 Fig.36. Experimental rate constant dependence on vibrational level num-
bers up to 10

Non-equilibrium effects may be observed in a system initially in

equilibrium if any processes are occurring at finite rates within the sys-
tem. Such processes inevitably result in a violation of the Maxwell distri-
bution. In particular, such a violation may be due to non-elastic collisions
resulting in mass exchange and internal energy redistribution of colliding
particles. Hence, the occurrence of chemical reactions always produces
non-equilibrium in a system. This non-equilibrium can be neglected (alt-
hough with a rather vague degree of accuracy) in the classical Arrhenius
kinetics when reaction rates are comparatively small. However, at high
temperatures in plasma-chemical processes, in radiation chemistry, shock
tubes and other conditions, the effects of these reactions must be taken in-
to account. For instance, thermal dissociation of diatomic molecules is
known not only to violate the Boltzmann distribution but also to decrease
the vibrational temperature in relation to the translational temperature.
These changes affect both dissociation rates and vibrational relaxation
processes. Non-equilibrium due to a chemical reaction and further occur-
rence of the reaction under non-equilibrium conditions is the problem to

101
be studied for non-equilibrium, generalized chemical kinetics. The Arrhe-
nius kinetics appears to be an extreme case and thus are believed to be
rigorously founded, their applicability limits being clearly defined. The
solution of this problem must be based upon a consideration of the Pauli
equation or the Boltzmann equation (or any equation of the Liouville
type). Using one of these equations and taking into account chemical re-
actions and relaxation processes of the internal degrees of freedom of re-
acting molecules, specific solutions under various conditions can be ob-
tained, including ordinary chemical kinetic equations and expressions for
the rate factors (constants) for chemical reactions.
Three solutions to the problems under discussion may be suggested,
allowing for the possibility that a set is found of cross sections  ( ) for
the processes involved; all the required information about the system can
be obtained through averaging.

3.7 Initiation of plasma-chemical processes by electron-beams

A significant advance in the development of pulsed electron beams

sources in recent years [41] encouraged scientists to study plasma for-
mation initiated by high-energy electrons and their energy dissipation
channels. In contrast to the space discharge, the use of pulsed electron
beams can easily solve the problem of excitation of large volumes of re-
active gas at high pressure, which is important for creating high-
performance processing facilities.
However, as it was shown by in many articles, the pulsed electron
beam is inefficient for direct plasma chemical processes in which the en-
ergy of electron impact is the only source for the initiation of chemical
reactions. In case with a plasma produced by electron beams, the low-
energy electrons (as compared with the energy carried by the electrons in
a beam) is only to play a role in plasma chemistry. These low-energy
electrons are formed due to relaxation of an electron beam during the
formation of ionization cascade of secondary emitted electrons. The exci-
tation of vibrational levels of molecules consumes a small fraction of the
energy of a pulsed electron beam, as the electron energy must exceed 20-
30 keV in order to overcome the anode foil between the beam generation
region and the reactor (see Fig. 37). Calculations performed by the au-
thors [42] showed that with the electron kinetic energy of more than 30
eV, the energy of degradation cascade in nitrogen is spent mostly on the
ionization and excitation of electronic levels of the molecule. The energy
balance of the electrons cascade with energies above 30 eV is shown in
Fig. 38. An electron may loose energy through the excitation of rotation-

102
al, vibrational and translational degrees of freedom about the ground state
of the molecule. At low energy these losses do not exceed 17% and are
reduced to 10% with increasing electron energy (curve 5). As a result, the
energy of the electron beam consumed on the decomposition of one mol-
ecule of the original gas mixture is significantly (10-100 times) higher
than its dissociation energy.
In case of high-current beams a significant role is played by the elec-
tromagnetic fields and collective electrodynamics processes which have
an affect both on the interaction of an electron beam with gases, and on
the kinetics of plasma chemical processes taking place.

Fig. 37. The energy distribution of the electron with energy Te on ioniza-
tion levels and excitation in nitrogen: 1 – total loss on excitation, 2-total loss
on ionization, 3 - loss pn excitation of electronic levels, 4 - loss on dissociative
ionization, 5 - other losses

The change in the energy distribution of a pulse electron beam on

different dissipation channels, taking into account the collective effects is
shown in Fig. 38 [43].

103
 Fig. 38. Redistribution of the energy spent on ionization and ohmic losses
of the electron beam at different degrees of freedom of molecular gas plasma:
ionization (1), dissociation (2), electron (3) and vibrational (4) excitation and
heating of the gas (5). Solid lines - without electric fields, dotted line - in the
presence of electric fields

In assessing the effectiveness of plasma-chemical processes in the

case of excitation of the reaction gas mixture by a pulsed electron beam
one must also take into account that the efficiency of the accelerator (the
ratio of the electron beam energy to the total energy consumed by the ac-
celerator) does not usually exceed 50%.

104
 Chapter 4 Chain gas-phase processes in the external action

Over the past 30-40 years in Russia and abroad extensive studies on
gas-phase chemical processes initiated by the gas discharge have been
carried out. It was found that some favorable conditions for the occur-
rence of chain processes are formed when the gas discharge has an effect
on gas-phase medium. With a sufficient length of the chain an electro-
physical installation makes it possible to reduce the total energy con-
sumption required for the chemical process. The primary source of energy
in this case is the thermal energy of the initial reactionary gas or the ener-
gy of the exothermic elementary reactions of chain process (e.g., reactions
of oxidation or polymerization). This can greatly reduce the power inputs
of the electrophysical installation. The reduction of the temperature of
chain chemical processes at radiation exposure is similar to the catalytic
effect. But the chain process can occur entirely in the gas phase, which
greatly increases the reaction rate compared to the hetaeraphase catalytic
process. Carrying out a chemical process at a temperature below the equi-
librium allows one to synthesize compounds which are unstable at higher
temperatures or the selectivity of their synthesis at higher temperatures is
low. These peculiarities of the behavior of chemical processes under
plasma action have shown to be a quite promising in the use of large-scale
chemical production. A high speed of the reaction which is required for
industrial technologies is achieved by using of branched chain processes.
But their major drawback is associated with the occurrence of an explo-
sive process, which significantly increases the risk of production. This
disadvantage is eliminated by initiating the chain process under external
action out of the self-ignition region.

4.1 Classification of chain processes

Chain processes are the chemical and nuclear transformation, where

a large number of elementary reactions (chain) occur due to an active par-
ticle (free radical, an atom with an unpaired electron, an excited particle -
an ion, a molecule - in chemical reactions, a neutron - in nuclear trans-
formations). The active particle is regenerated as a result of the elemen-
tary process.
A chain process consists of three stages: the initiation, the continua-
tion and the termination of the chain. The initiation involves the for-
mation of radicals through the decay of a molecule of the intial product. It
requires a high activation energy (~ 3.5 eV/molecule for hydrocarbons),
i.e. the reaction can occur in the equilibrium with an appreciable rate only

105
at temperatures above 1200 K. The continuation of chain consists of the
interaction of free radicals with the initial molecules resulting in the for-
mation of a stable molecule of the reaction product and a new radical. The
new radical, in turn, interacts with the initial molecules. This stage of the
chain process requires much lower activation energy (0.87 eV/molecule
for CH4), therefore, its implementation requires a lower temperature. It is
important to note that the major part of chemical reactions occur at the
stage of the continuation. An active radical that is formed at the initiation
stage can initiate 103 - 105 of the reactions at the stage of the chain pro-
cess.
Fig. 39 shows the change in energy consumption for conducting the
chemical transformation at different stages of the chain process of radia-
tion-thermal cracking of methane.

Fig.39. The chart illustrating the power inputs at stages of chain process
of methane conversion

At temperatures below 700 K, when the thermal initiation of reaction

does not occur, under the influence of the electron beam the active centers
- free radicals, ions or excited molecules are formed, which can start a
new chain reaction. Such a reaction will take place at a temperature of
150-200 degrees below than the temperature of a normal thermal process,
but with the same speed because the impact of plasma facilitates the most
energy-intensive phase –the phase of the reaction initiation.
Since 1913 when M. Bodenstein first discovered chain reactions
many other different types of chain reactions have been found. There are
unbranched chain reactions, chain reactions with quadratic branching,

106
chain reactions with degenerate branching, chain reactions with energy
branching and others.

4.1.1 Unbranched chain reactions

An example of unbranched chain reaction is a reaction of synthesis

of hydrogen chloride from a mixture of chlorine and hydrogen molecules.
In 1916, W. Nernst (Nobel Prize in 1920) proposed a mechanism of chain
reactions which involves Cl atoms.

Initiation of chain process by electron impact

In the plasma formed though electric discharge in the gas mixture of
Cl2 + H2, the following processes may occur:
1 Dissociative ionization: Cl2 + е + 11.6 эВ → (Cl2) + + 2е →
2е + Сl + Cl+
2 Dissociative electron attachment Cl2 + е + 2.4 эВ → (Cl2)¯ → Cl +
Сl¯

3 Dissociation though inelastic Cl2 + е + 2.4 эВ → Cl + Сl + е

collision
Development of chain process
At the stage of chain development the following reactions take place:
Сl + Н2 → НСl + Н + 0.05 eV at 100 °С k1 = 5·107 l/ mole ·с (145)
(slow reaction)
H + Cl2 → HCl + Cl + 1.97 eV at 100 °С k2 = 3·1010 l/ mole ·с (146)
(fast reaction)
Cl atom, formed in the reaction (146) is identical to the Cl atom, re-
acted in the reaction (145) therefore, the sequence of reactions could be
repeated many times (basicly 105-106 times), the same as the chain links
are repeated.
4 Chain termination
At formation of hydrogen chloride from a mixture of H2 + Cl2, the
active particles, which continue the chain, are the atoms of H and Cl. The
rate of reaction (2) is much higher than the rate of reaction (1), so the
concentration of hydrogen atoms is low and the reaction of chain termina-
tion is most likely to occur
Cl + Cl + М = Сl2 + М

107
 4.1.2 Chain reactions with quadratic branching

The experimental studies have shown that there exists a class of

chain reactions, which are characterized by the fact that at the interaction
of the active particles with the initial product along with the reaction
product a new active particle is formed. In the course of the reaction the
number of active particles increases continuously which leads to a sharp
increase in its velocity. This reaction can end up by the explosion. Thus,
in the case of the reactions of H2 with O2 the cycle of chain consists from
the following reactions [44]:
H* + O2 → OH* + O*,
OH* + H2 → H2O + H*,
O* + H2 → OH* + H*,
This cycle of reactions is called the cycle of chain with quadratic
branching. Each of the H* and OH* particles formed begins a new cycle
of the chain. As a result, their concentration and thus the reaction rate
avalanche-like increases, and the entire reaction is completed in a fraction
of a second at the initiation of combustion. It is important to note that the
heat of the reaction of 2H2 + O2 = 2H2O, which is equal to 483 kJ/mole, is
not immediately converted into kinetic energy of the reaction products,
and partially transformed into chemical energy of the active chemical par-
ticles of H* and OH*.
It is obvious that branching reactions can occur only as exothermic
reactions such as the following: the oxidation of hydrogen, phosphorus
and sulfur vapors as well as methane, other hydrocarbons, CS2 carbon di-
sulfide, carbon monoxide, SiH4 silane and the decomposition of NCl3. All
of these reactions are characterized by the existence of the so-called igni-
tion peninsula – a region with a pressure and temperature at which the re-
action proceeds.

4.1.3 Chain reactions with degenerate branching

This name was given to the numerous radical-chain reactions, which

are characterized by the formation of intermediate metastable products.
After reaching a maximum speed the reaction is slowed down. The char-
acteristic time of their occurrence is not a fraction of a second, but tens of
minutes or hours. The most studied chain reaction with degenerate
branching is the oxidation of hydrocarbons. These reactions will be dis-
cussed in the more details in Section 4.4.

108
 4.1.4 Electron-induced chain reactions

When an external electric field is applied to a chemically reactive

plasma the free electrons may act as active particles of chain process. In
this case, during the time between the release of an electron (ionization of
the molecule) and its collision with a neutral molecule the electron gains
an additional kinetic energy in the external electric field. The scheme of
this process was first proposed by Yu. N. Novoselov. It is shown that in
the case of non-self-sustained discharge initiated by an electron beam, the
energy consumption for the formation of a molecule of methanol can be
4-5 times less than at the self-sustained discharge, which is several elec-
tron volts. Much attention was given to the ion-molecular processes that
may lead to the emergence of the radicals. In particular, the chain of reac-
tion of

yields OH radicals, HO2 and the electron, which can repeatedly par-
ticipate in their formation, obtaining the necessary energy from the elec-
tric field.

4.2 Chain chemical processes under external action

The initiation and development of chain processes in external ac-

tion have been only researched for the last 10-15 years. The develop-
ment of chain processes such as oxidation of hydrogen, cracking of
hydrocarbon gases, partial oxidation of methane, and others under
equilibrium conditions has been well studied. Changing the conditions
of the chain process is interesting from the standpoint of the synthesis
of new products or elimination of defects such as explosibility.

4.2.1 Investigation of the induction period of ignition of the oxy-

gen - hydrogen mixture

A significant drawback of the chain reaction of hydrogen oxidation

from the standpoint of technological application is a significant delay (in-
duction period) in ignition. Fig.40 shows the change in pressure in the re-
actor during the development of the chain process of oxidation of hydro-
gen in equilibrium, reflecting the presence of the induction period [44].

109
 Fig. 40. The curves of the oxidation kinetics of the mixture of 2H2 + O2 at
485°C and at various initial pressures of: 1 - 8.2 Torr; 2 - 7.8 Torr; 3 - 7.4
Torr; 4 - 7.1 Torr; 5 - 6.8 Torr; 6 - 6.4 Torr; 7 - 6.1 Torr; and 8 - 5.8 Torr

The dependence of the initial induction period on the pressure of a

stoichiometric mixture of hydrogen and oxygen for the equilibrium condi-
tions is shown in Fig. 41.
To reduce the induction period of ignition of the mixture of 2H2 + O2
in [45] were proposed using a pulsed laser excitation of the gas mixture. It
is shown that the excitation of O2 molecules by resonance laser radiation
with λ = 762 nm leads to the formation of new channels of O, H and OH,
and acceleration of the ignition.

Fig. 41. Relationship between the induction period and the pressure of
mixture of 2H2 + O2 at different temperatures: 1 - 435 °C; 2 - 445 °C; 3 - 458
°C. Curves are calculation

110
 Fig. 42 shows the calculated dependences of the induction on the ini-
tial temperature of the stoichiometric oxygen-hydrogen mixture at a pres-
sure of 1 kPa in the case of a 1 ms laser irradiation and different values of
radiation density.

Fig. 42. The induction period versus temperature with the laser radiation
energy of 0 (1), 0.017 (2), 0.082 (3) and 0.15 (4) eV /molecule. Curve 5 - The
diffusion time of hydrogen from the excitation region versus temperature

At laser energy of 0.082 eV/molecule and a temperature of 300 K,

the estimated induction period was 0.2 sec. At a temperature of 735K and
pressure of 1 kPa the duration of the induction period of ignition of the
oxygen-hydrogen mixture reduced from 0.15 to 0.005 sec, trough the la-
ser radiation with energy of 0.082 eV/molecule. Under the equilibrium
conditions for the ignition of the oxygen-hydrogen mixture to occur at the
given pressure it is necessary to heat the gas mixture to a temperature
above 758K (see Fig. 42), which corresponds to the energy input of over
0.13 eV/molecule for oxygen or hydrogen. Fig. 43 shows the scheme of
formation of chain processes.

111
 Fig. 43. Formation of the chain mechanism by ignition of H2 + O2 + H2O
mixture without the excitation of H2O molecules (a) and with the excitation of
asymmetric vibrations of H2O caused by the radiation with λ = 2.66 µm (b)

In [46, 47,] the experimental studies on the dynamics of the induc-

tion period of ignition of hydrogen and oxygen stoichiometric mixture by
a pulsed electron beam were presented. The studies were performed on a
TEU-500 high-current electron accelerator [48]. The parameters of the
electron beam were as follows: the kinetic energy was 450-500 keV, the
half-height pulse duration was 60 ns, the pulse energy was up to 200 J,
the beam diameter of 5 cm. The electron beam was injected into a closed
reactor which is a stainless steel cylinder with an inner diameter of 9 cm
and a volume of 1.6 liters. The reactor pressure was controlled by a fast-
response pressure sensor [49, 50]. The initial temperature of the gas mix-
ture was 300K.
When the pressure in the reactor is 110 Torr (curve 1) the electron
beam initiates a chain reaction of oxidation, which is accompanied with
the energy release and the gas heating, which causes an increase in pres-
sure in the reactor. The pulse radiolysis leads to a significant reduction in
the induction period. Fig. 44 shows the waveform of pressure in the
closed reactor which is filled with the stoichiometric mixture of oxygen
and hydrogen, in the initial period after the injection of an electron beam.
The time (t = 0) when the electron beam was injected corresponds to a
spike in the waveform. This spike is due to an electrical signal from the
recording equipment when the accelerator launches. Fig. 44 shows that
the time delay between the electron beam injection and the beginning of
the raise in pressure in the reactor (induction period) is a few millisec-
onds. This is significantly lower than the calculated values of the induc-
tion period in similar conditions. For example, for laser energy of 0.082

112
eV/molecule (absorbed dose is 260 kGy) at a temperature of 300 K the
induction period was estimated to be 0.2 according to the data in [2].

Fig. 44. The change in pressure (initial part) in the reactor filled with a
mixture of 2H2 + O2 at different initial pressures of the mixture. The initial
temperature is 300 K

Similarly to the processes of self-ignition of a stoichiometric mixture

of oxygen and hydrogen under equilibrium conditions [1], it was revealed
that the induction period changes when the initial pressure of the gas mix-
ture is changed though the impact of a pulsed electron beam. Fig. 45
shows the dependence of the induction period of ignition of 2H2 + O2
mixture on the initial pressure of the reagent mixture.

Fig. 45. The dependence of the induction period on the initial pressure of
the gas mixture for different volumes of the reactors: 1.6 liters (1) and 3.2 liters
(2)

113
 The dependence of the induction period on the pressure has a shape
which is peculiar to chain processes – an increase in the induction period
near the combustion region (see Figure 40.). Increasing the reactor vol-
ume reduces an inhibitory effect of the reactor walls on the development
of the chain process, which increases the speed of the chain process and
thus causes a decrease in the induction period (curve 2 in Fig. 45). This
confirms that under the influence of a pulsed electron beam on the stoi-
chiometric oxygen-hydrogen mixture the hydrogen oxidation occurs as a
chain process.
To compare the effectiveness of different types of impact on the ini-
tiation of ignition of oxygen-hydrogen mixture the values of the induction
period for similar conditions are shown in Table 6. At high temperatures,
when the ignition of oxygen-hydrogen mixture is possible even without
an external action, a laser, or ionizing radiation causes a sharp accelera-
tion of the ignition and reduction in the induction period. The initiation of
the ignition occurs at low temperatures, under the external action, with a
low-energy input, which does not exceed a few percent of the dissociation
energy of the original molecules.

Table 6. The values of the induction period under different external effects
Energy
Tem- Induc-
Type of exter- Mixture of gas- Pres- input,
pera- tion
nal effect es sure eV/moluc
ture period
ule
No effect 2H2 + O2 7.6 Torr 730К 0 0.15 s
No effect 2H2 + O2 25 Torr 730К 0 0.2 s
Laser, λ = 762
2H2 + O2 7.6 Torr 300 К 0.082 0.2 s
nm
Laser, λ = 762
2H2 + O2 7.6 Torr 600К 0.082 0.01 s
nm
Laser, λ = 762 100
2H2 + O2 400К 0.082 0.2 s
nm Torr
Laser, λ = 2H2+O2 + 5%
7.6 Torr 300К
2.66 µm H2O
Pulsed
5%H2+air 76 Torr 770К 0 103 s
discharge
Pulsed 5∙10-4
5%H2+ air 76 Torr 770К 0.23
discharge s
Pulsed 5%H2+ air 7.6 Torr 400К 2.3 0.1 s

114
discharge
380
Continuous
Hydrogen- Torr 0.003
ionizing 400К 2s
aerial mixture 0.05 (20 kGy)
radiation
Mpa
Pulsed elec- 100 0.013 (114 0.003
2H2 + O2 300К
tron beam Torr kGy) s

The analysis of theoretical and experimental data shows that, under

the comparable conditions, ionizing radiation is more effective for the ini-
tiation of ignition than the laser radiation. At a temperature of 400K with
the laser radiation (λ = 762 nm), the energy input of more than 0.08
eV/molecule is needed to initiate the ignition. With a continuous or
pulsed radiolysis the ignition occurs when the energy input is below 0.01
eV/molecule. At the same time the pulse radiolysis by the electron beam
leads to a significant reduction in the induction period compared to the
continuous ionizing radiation. At a lower temperature and lower initial
pressure the induction period, in the case of the electron beam impact,
was 3 ms, which is nearly 1,000 times smaller than for the continuous ra-
diation.

4.2.2 Investigation of the displacement of the ignition limits of a

stoichiometric oxygen-hydrogen mixture under external action

A significant drawback (for technological applications) of chain re-

actions of the oxidation of hydrogen is also a high ignition temperature,
which exceeds 700K. Fig. 46 shows the limits of self-ignition of oxygen-
hydrogen mixture under equilibrium conditions [1].

115
 Fig. 46. The limits of ignition of a stoichiometric of hydrogen-oxygen mix-
ture in a spherical vessel with a diameter of 7.4 cm, covered with KCl. The first
and third limits are extrapolated (dashed line)

In [2] it is shown that the excitation of O2 molecules though the res-

onant laser radiation leads to the formation of new channels of O, H and
OH, which also leads to the reduction in self-ignition temperature. The
theoretical analysis has shown that with the laser excitation (λ = 762 nm,
τ = 1 ms) of oxygen molecules with an absorbed dose of 260 kGy (0.082
eV/molecule) at a pressure of 100 Torr, the temperature of self-ignition of
the mixture should decrease to 300K. The ignition temperature in the first
approximation was determined by the authors from the equation of the in-
duction period and the diffusion time of hydrogen atoms.
In [51] the authors studied the influence of the initial concentration
of free radicals and the rate of radiolysis of the molecular hydrogen and
oxygen on the displacement of ignition limits of the stoichiometric hy-
drogen-oxygen gas mixture. The values of the minimum initial concentra-
tion (pulse radiolysis) and minimum rate of generation (continuous radi-
olysis) of hydrogen and oxygen, at which the reduction of the ignition
temperature of the gas mixture begins with, were obtained. It was shown
that the ignition temperature of the first and second limits is the most sen-
sitive to the rate of radiolysis of molecular hydrogen and oxygen. The re-
sults of numerical simulations [13] showed that a significant shift of the
limits of ignition for the pulse radiolysis begins with the magnitude of the
initial mole fraction of hydrogen atoms of [Н]0 > 10-4. The results of the
calculations of ignition limits at a value of [H]0 = 0.002 showed that the

116
temperature of ignition at the first limit should reduce to 100-140 degrees
(see Figure 47).

Fig. 47. The displacement of the ignition limits at the initial mole fraction
of hydrogen atoms: solid curve is [Н]0 = 0, points is [Н]0 = 2 10-3

With the radiation output of hydrogen atoms equal to 2.14 per 100
eV, for the production of 1.2∙1021 of hydrogen atoms in 1 mole, the ener-
gy of electron beam of 1350 J is required. For the mass of gas of 2.12 g in
one mole, the absorbed dose is 110 kGy. This value of the absorbed dose
of electron beam energy is consumed only for the decomposition of hy-
drogen molecules. In a mixture of oxygen and hydrogen the injected elec-
tron beam is scattered not only by the hydrogen molecules but also the
oxygen.
To estimate the total absorbed dose let us take into account that the
energy loss of an electron with a kinetic energy of 0.5 MeV in hydrogen
is 4.2 4.2 МeV·cm2/g, and in oxygen is МeV·cm2/g [52]. Then, for a stoi-
chiometric mixture of oxygen and hydrogen the electron with an energy
of 0.5 MeV 3.4 more loses energy on 1 cm path due to the scattering by
oxygen molecules than due to the scattering by hydrogen molecules.
Therefore, the electron beam absorbed dose required for the production
the molar fraction of hydrogen atoms of 2·10-3 in the stoichiometric mix-
ture of hydrogen and oxygen is 480 kGy. In [3, 4], the experimental stud-
ies on the initiation of ignition of the stoichiometric oxygen-hydrogen
mixtures by a pulsed electron beam were presented. For this investigation
a TEU-500 high-current electron accelerator was used. The studies of the
ignition of 2H2 + O2 mixture thought the pulsed electron beam irradiation
showed that with the injection of electrons (the absorbed dose is not more

117
than 114 kGy) the ignition of the gas mixture occurs in a room tempera-
ture. The temperature bias, as a result of the pulse radiolysis, for the first
limit of ignition is much greater than the calculated values given in [13]
where it was estimated by a numerical simulation method. When the ini-
tial mole fraction of hydrogen atoms equals to 0.002 (absorbed dose of
electrons is 480 kGy), the lower temperature limit of ignition should de-
crease to 100-140 degrees and be equal to 600K.
To compare the effectiveness of different types of impact on the dis-
placement of the ignition limits both the experimental and theoretical data
for similar conditions are shown in Table 7.

Table 7. The effect of different types of external action on the displace-

ment of the ignition limits
Energy
Type of exter- Mixture of gas- Tempera- input,
Pressure
nal effect es ture eV/molu
cule
No effect 2H2 + O2 7.6 0 710
No effect 2H2 + O2 25 0 755
Laser, λ = 762
2H2 + O2 7.6 0.017 500
nm
Laser, λ = 762
2H2 + O2 7.6 0.082 300
nm
Laser, λ = 762
2H2 + O2 100 0.082 400
nm
Laser, λ = 2.66 2H2+O2+5%
7.6 2.5 J/cm2 300
µm H2O
Pulsed
5%H2+air 76 0.23 770
discharge
Pulsed
5%H2+ air 7.6 2.3 400
discharge
Continuous Stoichiometric
380 0.003
ionizing hydrogen-aerial 400
(20 kGy)
radiation mixture
0.05
Pulse radiolysis 2H2 + O2 100 680
(460 kGy)
Pulsed electron 0.013 (114
2H2 + O2 100 300
beam kGy)

118
 The ionizing radiation helps reducing the limits of ignition more ef-
fectively than the laser radiation. With an initial pressure 100 Torr in the
mixture of 2H2 + O2 under the influence of resonant laser radiation (λ =
762 nm) with an energy of 0.08 eV/molecule the ignition temperature
limit is shifted to 400K. With the same initial pressure, in the case of the
pulsed electron beam radiation with an energy of 0.013 eV/molecule the
ignition of 2H2 + O2 mixture at temperatures below 300K occur.
The use of the electron beam to initiate the ignition of oxygen - hy-
drogen mixture creates a unique environment for a study of the originali-
ty, development and termination of t combustion processes. This is be-
cause the duration of the electron beam effect is much smaller than the
induction period. The change in pressure in the mixture does not affect
the absorbed dose (at a low-pressure) provided that the density of electron
beam is constant. Furthermore, the change in the conditions of chain is
possible by introducing the atomic hydrogen into the reactor volume. The
initiation of ignition of oxygen - hydrogen mixture in gas discharge pro-
vides a high uniformity of the initial concentration of the atoms and radi-
cals in the reactor volume (ideal mixing).

4.3 Radiation-thermal cracking of methane

A great number of studies on the decomposition of hydrocarbons

during heating show that the process is implemented as a chain (thermal
cracking). The reactions of thermal cracking of normal alkanes occur at
high temperatures, for example it is about 800-900K for heptane –vapor
[53]. According to the chain theory of cracking, this process consists of
three stages:
1 The initiation of the reaction though radicals, due to the thermal
decomposition of molecules of the original product;
2 The continuation of the chain, which is due to the interaction of the
free radicals with the molecules. This leads to the formation a stable mol-
ecule of the reaction product and a new radical;
3 The chain termination.
The stage of initiation requires the activation energy of 3.5
eV/molecule (80 kcal/mol), i.e., the reaction can proceed at a noticeable
rate, but only after the temperature exceeds 900K. The stage of the chain
continuation requires activation energy of about 0.87 eV/molecule (20
kcal/mol), and therefore for its implementation a lower temperature is re-
quired.
At low temperatures, the thermal initiation of the reaction does not
occur due to the absence of the radiation. In this case, with the radiation

119
exposure the active centers which are free radicals, may occur, which can
start a chain reaction of hydrocarbon cracking. The interaction of the rad-
icals, which contain a small number of carbon atoms, with the initial hy-
drocarbon is a distinct chain process, which is initiated by the radicals
formed due to the radiation.
In a nook of V.V.Sarajeva [54] there are four temperature regions of
the radiation-thermal cracking, they are shown in Fig. 48.

Fig. 48. The influence of the temperature on the radiation decomposition

of hydrocarbons. The temperature ranges are: 1 – radiation, 2 - radiation-
thermal, 3 – Thermal – radiation 4 – Thermal [39]

The first region is radiation, at which the radicals are generated only
by the radiation. In this region, the process is not a chain, and the activa-
tion energy is only 0.04 eV/molecule, which characterizes the radiolysis
process.
The second region is the thermal radiation. Within this temperature
range the process becomes a chain with activation energy of 0.87
eV/molecule.
RH + e (3.5 эВ) → R* + H*
R* + RH + 0.87 эВ→ stable product + R1 *
The third region is the thermal radiation. It occurs when the R1 radi-
cal is thermally unstable and can decompose into two active radical due to
the reaction of
R1* → R2* + R3*
In this temperature range it is branched chain process (the activation
energy is of the order of hundreds of kilojoules per mole).
The fourth region is thermal. Here the generation rate for the active
radicals is determined by the reaction

120
 RH + tº (3.5 e) → R* + H*
The generation rate increases due to the radiation. The activation en-
ergy in this temperature range is determined only by the thermal process
and is equal to 3.5 eV. The speed of the chain process depends on the rate
of radical generation, which also depends on the dose. The influence of
the temperature and dose on the development of the chain process in de-
scribed in more details in [16] as an example of radiation oxidation of hy-
drocarbons.
Based on the analysis of experimental data on the radiolysis of liq-
uid hydrocarbons of the paraffin series it was obtained that:
1 The reaction yield of radical formation in γ-radiolysis, and a tem-
perature of 300 K is 7-10 radicals. Assuming that the average energy is
3.5 eV, it is easy to see that the chemical reaction consumes no more than
30% of the absorbed energy of ionizing radiation, and the rest of the en-
ergy is spent on heating (by collisions or energy transfer).
2 The yield of the radiolysis products at 300K contains about 80%
of the products of C - H bond breaking. It results in formation of molecu-
lar hydrogen and hydrocarbon, and about 20% are the products of C - C
bond breaking.
3 With an increase of the temperature above 600 K the radiation-
chemical yields rise dramatically, reaching 1.5 104 molecules per 100 eV
at 600 K, and the product composition approaches the composition of the
products peculiar to the thermal cracking.
Chain process of the decomposition of methane (radiation-thermal
cracking) can be described by the following basic equations [55]:
CH4 + 3.5 эВ = (CH3)* +
Chain initiation due to the thermal in-
H*
fluence
Chain initiation due to the radiation in-
CH4 + e = (CH3)* + H*
fluence
Continuation of the chain CH4 + H* = (CH3)* + H2
(CH3)* + CH4 = C2H6 + H*
(CH3)* + H* = CH4
Termination of the chain
(CH3)* + (CH3)* = C2H6
H* + H* = H2
In [56, 57] the experimental data on plasma catalysis in the conver-
sion of methane into carbon and hydrogen were presented. Preheated to a
temperature of 700 - 1100K methane was treated for a short time pulse
microwave discharge (frequency 9 GHz, pulse power up to 100 kW, pulse
length 1 ms, repetition rate 1 kHz), which caused a sharp increase in its
degree of conversion. It was shown that this effect can not be explained

121
by the thermal influence of discharge and the role of plasma is to generate
the active particles, which accelerates the conversion process. Fig. 49
shows the experimental dependence of the degree of conversion of me-
thane in the microwave discharge plasma on the temperature.

Fig. 49. The dependence of the degree of methane conversion on the tem-
perature (a): 1 - decomposition of methane without the influence of plasma, 2,
3, 4 - influence of plasma on gas, preheated to 950, 850 and 750K, respectively.
(b)The dependence of energy consumption for the formation of hydrogen mole-
cule on the plasma energy

It is important to note that the products of the plasma pyrolysis (H2

and nanosized graphite) were significantly different from the products of
the pyrolysis of CH4 in equilibrium conditions (acetylene, ethylene and
ethane).

4.4 Chain oxidation of methane under external action

It is known that the oxidation of the hydrocarbon is a branched de-

generate chain process [58, 59]. The most energy-consuming element is
the formation of primary radicals, i.e., the initiation of the chain. Under
the equilibrium conditions, the initiation of the chain process is a result of
thermal decomposition of the valence-saturated molecules or the interac-
tion of two valence-saturated molecules. The remaining links are initiated
by the free radicals, which significantly reduces their activation energy
compared with the molecular reactions. Thus, the low-temperature
boundary of the oxidation region should be considered as a boundary with
the minimum temperature, which, in the given conditions of pressure,
mixture composition, surface of the reactor, could still initiate chain oxi-
dation. Further reduction in temperature will lead to their elimination of

122
the initiation reaction and, for this reason, the disappearance of the rest of
the chain. Thus, if it were possible to carry out the initiation of free radi-
cals at a lower temperature than the temperature of the low boundary of a
slow oxidation region, then the oxidation and further conversion of hy-
drocarbons could be continued even in these conditions.
With this oxidative conversion of hydrocarbons at low temperatures,
the synthesis of new compounds may occur. In a complex chain process
the radicals can enter not into one reaction but also in two or more reac-
tions. The temperature influences on the outcome of the radical reactions.
The lower the temperature of the process is, the smaller activation ener-
gies is required for radical to determine the preferred direction of the re-
action. For example, at 400K the difference between the activation ener-
gies of two reactions of 0.04 eV/molecule leads to the fact that 93% of the
response will take the path with a lower activation energy (provided that
cross the sections for interaction are equal).
The oxidation at the temperatures lower than the normal tempera-
tures in oxidation, should result in a significant decrease in the quantities
of hydrocarbons which decompose and oxidize. This follows from the
fact that out of two possible reactions of alkyl radicals - the decomposi-
tion and the oxidation
1 The decomposition of R → products of decomposition, the activa-
tion energy of 1.1-1.7 eV/molecule.
2 The oxidation of R + O2 → RO2 the activation energy of 0.02-
0.09 eV/molecule at temperature of the oxidation of hydrocarbons of 450
K. We can assume that the cracking does not occur.
The second consequence of lowering the temperature of the reaction
is the formation a significant amount of peroxides as a result of conver-
sion of peroxide radicals RO2 .Here there are two ways: the disintegration
of RO2 radicals with the formation of aldehydes and alcohols, and its in-
teraction the initial hydrocarbon with the formation of the alkylhydroper-
oxides. The activation energy of decomposition is about 0.8 - 0.9
eV/molecule. The energy of activation of the bimolecular interaction of
RO2 with hydrocarbon does not exceed 0.2-0.3 eV/molecule. For a mix-
ture of 150 Torr RH + 150 Torr O2 at 450 K the ratio of the rate of the de-
composition reaction of RO2 to the reaction of alkylhydroperoxides for-
mation is 0.53. In the case of the oxidation at 400K, this ratio is equal to
0.043. Thus, the reduction in the oxidation temperature should facilitate
the formation of the alkylhydroperoxides and cause a decrease in the yield
of aldehydes and alcohols, which are, according to the mechanism, the
decomposition products of the RO2 radicals.

123
 One can also assume that at a low temperature of oxidation the al-
kylhydroperoxides will have an isometric structure rather than the normal
structure. The difference in the energy of a bond breakage of the primary
and the secondary C-H bonds is about 0.17 eV/molecule. This difference
will lead to the fact that at low temperatures the separation of the hydro-
gen atom from CH2group (rather than CH3) will occur in hydrocarbon and
the peroxide formed will have an isomeric structure.
In general, it is expected that a significant reduction in the reaction
temperature causes a change in the direction of oxidative conversion. The
alkylhydroperoxides will be mainly formed. The reduction in the tem-
perature of oxidation can be achieved by facilitating the generation by
primary free radicals. This can be done either by supplying the active rad-
icals from the outside, or addition of certain substances that are able to
decompose to radicals at low temperatures (i.e. below the boundary of the
normal region slow oxidation of hydrocarbons).

4.4.1 Oxidation of methane in the equilibrium conditions at a low

pressure
The oxidation of methane under equilibrium conditions is well stud-
ied and described in the books written by V.J. Stern [20], and V.S.
Arutyunov [21]. Experimental studies have shown that at a low pressure
(below 10 atm.) in the initial mixture CH4 + O2 the oxidation process is
very slow. Fig. 50 shows the dependence of the oxidation of methane on
the pressure and area of ignition of the methane-air mixture under equilib-
rium conditions. You can see that the chain process of methane oxidation
of occurs at temperatures above 800 K.

Fig. 50. The dependence of the oxidation time of methane on the mixture
pressure at temperature of 650 K and the region of ignition of hydrocarbon-air
mixtures. 1 – 13% CH4, 2 – 10 % C2H6

Fig. 51 and Fig. 52 show the kinetic curves of the accumulation of

the products of methane oxidation.

124
 Fig. 51. The kinetics of methane oxidation obtained by: changing the total
pressure, consumption of the initial products and the accumulation of the end
products of the reaction.

A mixture of CH4 + 2O2; Rinitial = 235 Torr, T = 723 K. 1 – CH4; 2 – O2;

3 – H2O; 4 – CO; 5 – CO2; 6 – ∆Р

Fig. 52. The accumulation kinetics of formaldehyde and hydrogen perox-

ide at the oxidation of methane. A mixture of СН4+2O2; Рinit = 235 Тоrr;
Т = 723 К. 1 – peroxide, 2 – formaldehyde

The experiments were conducted under static conditions in a quartz

vessel within the temperature interval of 723-813 K with a mixture of
CH4 + 2O2. The main products of oxidation of methane under equilibrium
conditions were CO2, CO and H2O. The formaldehyde was oxidized at
high temperatures to carbon monoxide and water, so in the reaction prod-
ucts the content of it is negligible. The slow oxidation of methane under

125
equilibrium at low temperature is characterized by a low degree of con-
version. The oxidation process is terminated long before the formation of
the original products (see Fig. 51, lines 1 and 2).

4.4.2 The initiation of methane oxidation by an external influ-

ence

The mercury and bromine sensitized photochemical oxidation of hy-

drocarbons is reviewed in a book of V.J. Stern [20]. The photochemical
oxidation of propane at room temperature was the most studied. The main
product of the oxidation was hydroperoxide of isopropyl (see Fig.53).

Fig. 53. The dependence of the yield of photochemical propane oxidation

on the temperature [60].Contact time 30 sec. 1 – peroxide, 2 – acetaldehyde, 3
– formaldehyde, 4 – acetone, 5 – CO

The quantum yield of the reaction was determined. It was calculated

as the ratio of an amount of the molecules of isopropyl hydroperoxide
which are formed per unit time to the number of photons absorbed by the
system. It turned out that the lower limit of the quantum yield equals two,
i.e. the length of the chain oxidation of propane in the studied conditions
is very small. In the all experiments carried out at room temperature no
more than 0.5 % of propane entered into the oxidation reaction.
The photochemical oxidation of methane and ethane was studied by
A.B. Nalbandian and N.V. Fok, both at room and elevated temperatures.
It was found that at the room temperature the main products are the hy-
droperoxides. As a temperature increases, together with the peroxides, the
aldehydes start formimg. Starting from 500K, the yield of peroxides de-
creases, while the yield of aldehydes continues growing and at 600 K the
main products are the aldehydes CO and CO2. The kinetic curves of the

126
peroxides and aldehydes formation are similar to the curves obtained by
the oxidation of propane.
The study of photochemical oxidation of hydrocarbons at low pres-
sures showed that:
1 The facilitation of the initiation enables the oxidation at low tem-
peratures
2 The main product of the low-temperature oxidation is the corre-
sponding alkylhydroperoxides
3 If the conversion is very low, the oxidation process is slow as well.
A small chain length in these conditions makes, however, the low-
temperature oxidation of hydrocarbons ineffective with respect to the
yields of products and the degree of conversion.
In 1958, the oxidation of methane (a mixture of 4SN4 + O2 and CH4
+ O2) under the action of a fast electron beam at room temperature was
studied by B.M. Mikhailov and others [61]. The dose rate (at Pinit = 760
Torr) was 2,7·1016 eV/cm3·s (for the mixture of CH4 + O2 the dose was 4
kGy/s). The source of fast electrons was an electron tube with an acceler-
ating voltage of 120 kV and an output current of 100 µA. The experi-
ments were conducted with a steel vessel, the irradiation time was 30 min.
The reaction yield for the mixture CH4 + O2 was equal to 7 molecules of
O2 and 6.2 molecules of CH4 at 100 eV. Fig. 54 shows the change in pres-
sure in the irradiated methane-oxygen mixture during the reaction.

Fig. 54. The change in pressure of nitrogen and methane-oxygen mixtures

in the initial period of the irradiation by the fast electrons at room tempera-
ture[31]. 1 – CH4:O2 = 4:1, 2 – CH4:O2 =1:2, 3 – CH4:O2 = 1:1, 4 – N2

127
 A sharp increase in pressure for 20-30 seconds in the beginning of
the curve in Fig. 16 is associated with heating. Later the pressure curve
dpops during 3-4 min, then straight-line section can be obsereved, which
lasts until one of the initial components of the mixture is used up. After
that, the pressure drop slows down. It was found that the reaction rate in-
creases with an increase of the partial pressure of methane and the total
pressure. The consumption of methane and oxygen is proportional to the
time of irradiation. It is important to note that the rate of oxidation of me-
thane in the case of the irradiation by the electron beam is much higher
than under equilibrium conditions, although the temperature of the gas
mixture was significantly lower (300 K at the irradiation, 723K under
equilibrium conditions).
At room temperature, and initial pressure of 760 Torr the methane
mixture was subjected to transformation by 32 % (a mixture of CH4 +
O2). This greatly exceeds the degree of conversion in the mercury sensi-
tized photochemical initiation (0.5 % for propane). The products of oxida-
tion found are CO and CO2, H2O, H2, HCOOH and a small amount of
CH3OH and peroxides. The formaldehyde was found only in trace
amounts (see Fig. 55).

Fig. 55. The dependence of the oxidation products of methane on the du-
ration of electron beam irradiation. CH4 + O2 mixture: Pinit = 760 Torr., T =
300 K. 1 – H2O, 2 – H2, and 3 – HCOOH, 4 – CO and CO2, 5 – peroxides and
alcohols

About 50 % of methane consumed goes into the liquid phase, the

main product of which is HCOOH.
These data suggest that during the methane oxidation initiated by ra-
diation at room temperature, in contrast to the thermal and photochemical

128
oxidation, the degree of conversion is higher (32% with the irradiation
and 0.5 % under equilibrium). The authors believe that when bombarded
with the fast electrons 50% of the exposed methane is converted to the
excited molecules, and 50 % to the ions.
In [62, 63]] the partial oxidation of methane by the two types of mi-
crowave discharge was studied. These discharges are: the pulse-periodic
discharge (streamer pseudo-corona discharge, the wavelength of 3 cm,
pulse power of 300 kW, average power of 300 W, pulse duration of 1 ms,
repetition rate of 1 kHz) and the continuous discharge (coaxial jet dis-
charge, the frequency of 2.45 GHz, the power of 1-5 kW). The initial re-
actants were heated up to 800-1200 K and supplied into the discharge
chamber, which were combined with the methane combustion zone. The
authors note that the use of the microwave discharge acts on the system in
two ways. Firstly, it effectively introduces an additional thermal energy
even in strongly heated reagents due to the high temperature of plasma.
Secondly, the plasma generates the active particles which contribute to
the oxidation of methane in the chain reactions, and play a role of the ini-
tiator of the combustion. For the reaction of partial oxidation of methane
the energy consumption of the microwave discharge were 0.25
eV/molecule when the degree of conversion methane was 70 %, and in-
creased to 0.5 eV/ molecule with the degree of conversion about 100 %.
In this case the energy input due to thermal heating (according to the de-
gree of conversion) was 2.6 -2.8 eV/molecule. The heating of the mixture
of gases has greatly increased the length of the chain reaction of oxidation
and with the inexternal initiation of the chain process significantly in-
creased the degree of methane conversion at the given temperature.

4.5 The conversion of carbon disulfide CS2 in the atmospheric

air

Chain mechanism of the reaction was discovered by a group headed

by N. Novoselov while they studied the decomposition of carbon disul-
fide under the electron beam and in the semi-self-maintained discharge
initiated by an electron beam. In [64, 65] the results of experimental and
theoretical studies of the action of the pulsed processes of nano - and mi-
crosecond duration on the nitrogen-oxygen mixture at atmospheric pres-
sure with an addition of carbon disulfide CS2 were presented. The two re-
gimes of the conversion of impurities were found. These regimes depend
on the type of beam and are characterized by the mechanism and the end
products. The effect an external electric field on the conversion of impuri-
ties was investigated.

129
 For the experiments the model gas mixtures of nitrogen and oxygen
were used. The content of nitrogen N2 was ranged from 89 to 99 % and
oxygen from 0.1 to 10 %. The content of the investigated impurities CS2
varied from 0.005 to 1 %. For the irradiation of the gas mixtures two elec-
tron accelerators with different parameters were used. A compact “Ra-
dan”accelerator was used in a nanosecond time scale. It generated the
electron beam with the following parameters: the electron energy was 180
kV, the half-hight pulse duration of 3 ns, the current density of 800
A/cm2, the pulse repetition rate of 10 Hz. In some experiments the non-
self-sustained discharge, initiated by a nanosecond electron beam was ob-
served. The discharge was ignited in a gap with a length of 1 cm when a
12 nF capacitor, charged to 1-25 kV was connected to the gap. In the mi-
crosecond time scale an accelerator with a plasma-cathode was used. This
accelerator formes a radially divergent electron beam with the following
parameters: the area of ~ 1.5 m2, the electron energy of 280-300 keV, the
current density raged from 0.1 to 10 mA/cm2, the pulse duration of 48 ms.
The volume of 170 liters was subjected to the irradiated. In the same vol-
ume the nonself-sustained discharge maintained by an electron beam was
ignited in the gap of 10 cm in length. The maximum field strength in the
volume discharge plasma was 5 kV/ cm.
In a mixture irradiated by a nanosecond electron beam the initial
concentration of CS2 has a little effect on the angle of the dependence of
concentration of the carbon disulfide molecules on the number of irradia-
tion pulses. The minimum value of energy consumption was equal to 0.8
eV/molecule, which is ignificantly less than the energy of dissociation of
carbon disulfide, equal to 7.6 eV. The conversion process of CS2 in the
nonself-sustained discharge initiated by a nanosecond beam was also in-
vestigated. Fig. 56 shows the dependence of the energy consumption on
the electric field strength in the volume discharge. It was found that in a
range of the electric field strength from 0 to 1.5 kV/cm, the value of ener-
gy varies slightly and is ranged from 0.6 to 0.9 eV/molecule. This is much
lower than the dissociation energy of carbon disulfide and oxygen.

130
 Fig. 56. The dependence of the energy required for the removal of one
molecule of the impurity on the field strength in the column of the nonself-
sustained discharge initiated by a nanosecond electron beam

The experiments showed that with the comparable values of energy

introduced to a gas by the electron beam the irradiation regimes are dif-
ferent and characterized by the end-products of CS2 conversion. As a re-
sult of the gas irradiation by high-current electron beam of nanosecond
duration the main products of the reaction are the sulfur and carbon ox-
ides. After irradiation of the initial mixture by an electron beam of low
current density and microsecond duration in a plasma-chemical reactor a
solid phase as well as viscous liquid were found. The analysis showed
that these substances are composed of (-CS-)n, the presence of SO2 with
trace amounts was detected.
According to the analysis the ionization by a high-current electron
beam of nanosecond duration leads to a chain mechanism of oxidation of
CS2, in which a triggering agent is atomic oxygen:
CS2 + O2 → CS + SO (147)
SO + CS → CO + S2 (148)
O + S2 → SO + S (149)
SO + O2 → SO2 + O (150)
S + O2 → SO + O (151)
The recovery of atomic oxygen occurs under the influence of the
beam through the electronic states of O2, and in the reactions with elec-
tronically excited nitrogen molecules N2*:
e + O2 → 2O (152)
*
N2 + O2 → N2 + 2O (153)
With an electron beam of a sufficient intensity in the reactions (153)
the quantity of O radicals is enough to trigger a chain mechanism (147) -

131
(152). By reducing the concentration of CS2 below a certain critical value
the chain mechanism is interrupted, the atomic oxygen ceases to turn out
in the reactions (152), (153) and the oxidation of carbon disulfide is ter-
minated. According to the calculations, the main products of decomposi-
tion of CS2 are sulfur oxides SO2 and carbon oxides CO. In small quanti-
ties NO, SO and S2 were produced, which was observed in the experi-
ments. The presence of nitrogen in a mixture leads to the loss of radicals
O in the reactions of nitrogen oxide synthesis:
N2(ν) + O → NО +N (154)
N + O2 → NО + О
Here N2(ν) is the vibrationally excited nitrogen molecules.
Another mechanism for the conversion of carbon disulfide is imple-
mented with the excitation of N2-O2 -CS2 mixture by a low-current elec-
tron beam of microsecond duration. In this case, the beam does not have
enough intensity and the density of plasma electrons is two orders of
magnitude lower. The concentration of charged particles generated by a
low-current beam is low and comparable with the concentration of free
radicals O. In this case the chain mechanism does not occur, although the
oxidation takes place: the estimated concentration of SO2 is three orders
of magnitude smaller than in the high-current irradiation regime. In the
case of the low-current electron beam the ion losses due to electron-ion
recombination are sharply reduced and the plasma conditions are more
favorable for the conversion of CS2 via the formation of the cluster ions.
The concentration of these ions is [A±] ~ 103 cm-3. The reduction in the
concentration of CS2 occurs in the fast reactions of clusters formation of
A ± (CS2), followed by the polymerization of carbon disulfide and the
formation of (-CS-)n molecules. The rate of these processes is much high-
er than the rate of the reactions involving free radicals O. It was found
that the cluster ions in these experiments are the negative ions O2-, the rate
of which is almost entirely controlled by the process of three-body at-
tachment of the electrons to molecular oxygen.
When the ignition of a volume discharge is initiated by an electron-
beam of nanosecond duration, the influence of the electric field is mani-
fested in a monotonic decrease in the number of converted molecules of Δ
[CS2], which leads to an increased energy consumption. The main reason
for the decrease in Δ [CS2] is the loss of atomic oxygen in the reaction
(12), becausee with an increase in electric field strenth in the discharge
the rate of vibrational excitation of nitrogen increases as well
е + N2 → е + N2(ν)
A more complex nature of the influence of an electric field is ob-
sereved in the discharge, maintained by a weak electron beam. [66]. Since

132
the loss of molecules of carbon disulfide in this regime is associated with
the formation of clusters which are the ions of O2-, the behavior of Δ
[CS2] should be determined by the rate of the cluster ions formation in the
reaction of the three-body attachment
е + М + О2 → М + O2–
Fig. 57 shows the scheme of basic processes of the decomposition of
CS2 caused by a pulsed electron beam

Fig. 57. The scheme of processes occurring in mixture of N2 + O2 +

CS2 initiated by a pulsed electron beam of 3 ns (1) and 48 ms (2) in dura-
tion
To analyze the mechanisms of conversion of CS2 a model of plasma-
chemical processes initiated by a pulsed electron beam in nitrogen-
oxygen mixture with an addition of carbon disulfide, has been designed.
[30].

133
 Chapter 5. Applied plasma-chemistry

Gas discharges and other sources of low-temperature plasmas are

widely used in chemical manufacture, namely in the following: synthesis
and decomposition of inorganic compounds, plasma-chemical synthesis
of nanosized materials, catalytic plasma-chemical processes and others.
Low-temperature plasma is applied in organic chemistry as well: conver-
sion of hydrocarbon gases (methane, propane, landfill gas) into high mo-
lecular weight compounds, synthesis of hydrogen and carbon clusters
(fullerenes), processing of polymer coatings, leather, fur, etc.

5.1 Nonequilibrium synthesis of nanosized oxide powders

Titanium dioxide (TiO2) is one of the most important inorganic ma-

terials. It is used as a white pigment, a photocatalyst, a semiconductor in
solar cells, etc. The world consumption of TiO2 in 2001 exceeded 4 mil-
lion tons. Titanium dioxide nanoparticles have been studied well and are
prepared by various processes; therefore, the basic features of different
manufacturing processes can be compared by considering the characteris-
tics of synthetic TiO2. The use of nonequilibrium plasma-assisted pro-
cesses is a promising way of oxide synthesis. The nonequilibrium charac-
ter of the synthesis process makes it possible to considerably lower power
consumption for the reaction and to alter its conditions.
The use of nonequilibrium plasma-assisted processes is a promising
way of oxide synthesis. The non-equilibrium character of the synthesis
process makes it possible to considerably lower power consumption for
the reaction and to alter its conditions. Nanosized silicon dioxide (SiO2)
was prepared in a nonequilibrium plasma-chemical process induced by a
pulsed electron beam [67, 68]. The energy consumption in an electro-
physical unit for the conversion of silicon tetrachloride (SiCl4) did not ex-
ceed 0.02 eV/molecule.

5.1.1 Comparison of the available processes for production of

nanosized TiO2 particles

1 The chlorine process is the most widespread vapor phase proce-

dure for the preparation of TiO2 from titanium tetrachloride (TiCl4). This
process makes it possible to synthesize crystalline TiO2 with a given ratio
between the rutile and anatase crystalline phases. In this process, titanium
tetrachloride is oxidized by oxygen
TiCl4 (gas) + O2 (gas) = TiO2 (solid) + 2Cl2 (gas)

134
 However, the high synthesis temperature (1400–1700°С) compli-
cates the manufacturing process [69, 70] and requires a considerable
power input. To reduce the temperature and to decrease the powder parti-
cle size, Bin et al.[69] investigated the gas phase hydrolysis of TiCl4:
TiCl4 (gas) + 2H2O (gas) = TiO2 (solid) + 4HCl (gas)
They showed that amorphous TiO2 was produced at a synthesis tem-
perature of 260-400°С and a crystalline material (anatase) was formed at
400-600°С. The average particle size decreased with an increase in the
synthesis temperature from 120 nm (at 260°C) to 18 nm (at 525°C). The
results of investigation of the preparation of TiO2 nanoparticles by ther-
mal vapor-phase decomposition of tetrabutoxytitanium (C4H9O)4Ti are
reported in [71, 72, 73]. The degradation temperature (Td) of (C4H9O)4Ti
is below 500°C, but the onset of formation of particles with the crystalline
phase (anatase) was detected only at Td > 500°C and all particles pos-
sessed this structure at Td > 600°C. The formation of the rutile crystal
structure began at Td > 1200°С [71]. Ayllón et al. [72] carried out the de-
composition of (C4H9O)4Ti in a mixture with oxygen in an RF discharge
plasma. Despite of a high temperature in the discharge zone, the degrada-
tion of(C4H9O)4Ti yielded amorphous TiO2 particles with an average size
of 25 nm and the crystal structure was formed during subsequent anneal-
ing at 600–800°C. The rutile lattice in noticeable amounts was formed on-
ly at T > 800°C. Li et al. [73] also detected the formation of TiO2 nano-
particles with the anatase lattice at Td of tetrabutoxytitanium of 600°C.
During the annealing of these particles (>800°С), rearrangement into the
rutile lattice took place.
The investigation of the synthesis of TiO2 nanoparticles from abla-
tion plasma produced by a pulse laser showed that it is the formation con-
ditions, not the initial temperature of particles, that are important for the
formation of the crystal lattice. Upon laser ablation of titanium or titania
(rutile or anatase) rods in an He + O2 atmosphere, the formation of TiO2
(anatase) particles of 10–50 nm in size was revealed [74]. The laser abla-
tion of a TiO2 rod in an inert atmosphere generated amorphous TiO2 na-
noparticles. Seto et al. [75] reported the results of study of nanoparticles
prepared by pulsed laser ablation of a TiO2 (rutile) substrate in a helium
atmosphere. It was shown that the particles had a metallic titanium core
and a shell composed of crystalline TiO2 (rutile and anatase).
In the synthesis of TiO2 nanoparticles, the appearance of the initial
titanium-containing material has no determining effect on the crystal
structure of the particles produced. Oh et al. [76], reported the results of
investigation of TiO2 preparation by oxidation of titanium nitride (TiN)
particles in a microwave plasma. A TiN powder was carried to the dis-

135
charge zone by a stream of oxygen mixed with an inert carrier gas. At a
low oxygen concentration, the particles had a TiN core and a TiO2 (rutile)
shell. When oxygen was in excess, the particles were composed com-
pletely of TiO2 but had the anatase structure.
Unlike vapor-phase processes for the manufacture of TiO2 nanopar-
ticles, liquid-phase synthesis occurs at a lower temperature; for example,
at T < 100°C in the sulfate process [69]. However, the particles produced
are amorphous and the formation of a crystal structure requires subse-
quent annealing at T > 600°C. In addition, thermogravimetric studies [69,
70] have shown that the removal of hydroxyl groups and residual organic
components take place at T > 500°С. Heating amorphous or anatase TiO2
to this temperature leads to the formation of the rutile lattice, thus compli-
cating the synthesis of anatase TiO2 by the liquid-phase procedure.
The explosion of conducting wires, providing a non-equilibrium na-
ture of the formation of nanoparticles [77, 78], allows one to synthesize
particles with new properties, in particular with a high excess energy. In
this method, a mixture micrometer-sized and nanosized particles is
formed. Be means of special filters it is posiible to collect the synthesized
nanoparticles with a specific area of up to 40-80 m2/g, which make up 30
% of the total mass of the particles. The capacity of the plant is 100-200
g/h. The explosion of metal wires in oxygen or air leads to the formation
of nanosized metal oxides. It was found that with the oxygen concentra-
tion in the mixture of an inert gas over 20 % (volume persent) the degree
of oxidation of metal particles synthesized by the wire explosion is up to
100 %. Synthesized nanooxides have a spherical shape and relatively
smooth surface.
The data from the comparison of parameters of titanium dioxide syn-
thesized via different processes are summarized in Table 8.

Table 8. Comparison of parameters of TiO2 particles synthesized by different

technologies

Preparation meth- Reactants Grain size Tempera- Lattice

ods ture of syn- structure
thesis
Chlorine process TiCl4 + O2 100-400 1400-1500
rutile
nm ºС +anatase
Oxidation in diffuse TiCl4 + O2 15–30 nm 1700 ºС anatase +
flame rutile
Gas-phase hydroly- TiCl4 + H2O 18-120 nm lower than amor-
sis 600 ºС phous+

136
 anatase
Droplet-to-particle TTIP. + propa- 20 – 40 500–1200 anatase
method nol nm ºС
Decomposition in TTIP. + О2 Average amor-
RF-discharge plas- size 25 nm phous
ma
Laser ablation Ti (or TiO2) + 10 - 50 nm anatase
O2
Ablation by pulsed Ti + O2 4 - 45 nm rutile
ion-beam 0.3-2 µm +anatase
Oxidation in ther- TiN + O2 average 50 anatase
mal microwave- nm
discharge plasma
Sulfate process (wet TiOSO4 + H2O 100-400 hydrolysis rutile
method) nm up to 80ºС, +anatase
annealing -
above
700ºС
Non-equilibrium TiCl4 + H2 + 30 - 200 lower 600 rutile
plasmochemical O2 nm ºС +anatase
synthesis

The review of different processes for the synthesis of nanosized tita-

nium dioxide has shown that crystalline particles are formed at the tem-
peratures above 500 ºС (anatase) and 800 ºС (rutile) only.

5.1.2 An overview of the techniques to synthesize nanosized

(TiO2)x(SiO2)1-x

Along with pure TiO2 nanoparticles, a composite material containing

TiO2 and SiO2 is of great interest. Many useful properties of TiO2 can be
substantially enhanced by varying its structure in the presence of amor-
phous SiO2. In addition, the use of an inexpensive support for TiO2
makes it possible to reduce considerably the cost of the material with re-
taining its useful properties. The composite material (TiO2)x(SiO2)1 − x is
also interesting from the viewpoint of investigation of change in the struc-
ture of crystalline TiO2 upon insertion in the amorphous SiO2 matrix.
For the synthesis of a composite (TiO2)x(SiO2)1-x material, use is
largely made of the sol-gel method. The resulting amorphous material is

137
further subjected to thermal treatment at the temperature above 500 °C to
remove the hydroxyl group and the precursor material left.
In [79], amorphous (TiO2)x(SiO2)1-x (x = 0.08, 0.18 and 0.41) was
investigated using X-ray phase analysis, neutron diffraction, infra-red
spectrometry, thermal gravitometry, and other analytical techniques. As
precursor materials use was made of tetraethylorthosilicate (TEOS) and
titanium tetraisopropoxide (C4H9O)4Ti. It was found out that at a compar-
atively low concentration (х = 0.08) titanium is embedded into the lattice
of titanium dioxide to form the Ti-O-Si bond. This concentration of tita-
nium is below its limit of solubility in silicon dioxide. Given a high con-
centration (х = 0.41), the phases of titanium dioxide and silicon dioxide
are separated already in the initial gel, and the general structure of the re-
sulting material is amorphous. When the composite (TiO2)x(SiO2)1-x (х =
0.41) material is heated to the temperature above 500 ºC, titanium dioxide
forms a crystal lattice with the structure of anatase. It was noted in [5] that
the presence of amorphous silicon dioxide retards rearrangement of the
crystal structure of titanium dioxide of the anatase type into that of the ru-
tile type. No rutile-type lattice was found out when the composite materi-
al (х = 0.41) was heated to 800 ºC. Initially, the composite material
(TiO2)x(SiO2)1-x with the average concentration of titanium (х = 0.18) was
characterized by a Ti-O-Ti bonding. Upon heating the gel to 500 ºC, all
the titanium atoms formed a Ti-O-Si structure. When the gel was further
heated (750 ºC and higher), this structure was decomposed to form sepa-
rate phases of silicon dioxide and titanium dioxide (anatase).
In [80] the results of investigation of the composite (TiO2)x(SiO2)1-x
(x = 0.1, 0.3 and 0.5) are presented, which was formed by the sol-gel syn-
thesis tetraethylorthosilicate and titanium tetraisopropoxide. The initial
gel was a composite of TiO2 and SiO2 with a conspicuous number of Si-
O-Ti structures. As the temperature was increased, there was an increase
in the number of Si-O-Ti bonds that decomposed at higher temperatures.
The structure of the composite after thermal treatment was a matrix made
up by pure silicon dioxide with particles of crystalline titanium dioxide
having either an anatase or rutile lattice, depending on the treatment tem-
perature. A concurrent research in [80] was made into the changes in the
structure of titanium dioxide synthesized by the same method from titani-
um tetraisopropoxide. Initially in the course of heating to 420 ºC, the
amorphous gel TiO2 acquired crystal structure with the anatase-type lat-
tice. When heated to above 800 ºC, it exhibited a rearrangement of its
crystal lattice into the rutile form. The composite material rearranged
from the amorphous structure into crystalline (anatase) at the temperature
above 600 ˚C and maintained the anatase lattice in heating to 1000 ºC.

138
 The results of investigation into the formation of nanosized compo-
site material from a mixture of titanium dioxide (anatase) gel and silicon
dioxide gel with different mixture ratios are given in [81]. The mixture of
these gels was dried in air at 110 ˚C and was further annealed at 800 ºC
for 1 hour. The structure of pure TiO2 during the high-temperature anneal-
ing treatment was partially rearranged into the rutile-type lattice. An addi-
tion of silicon dioxide inhibited the rearrangement of the crystal lattice of
the former during pyrolysis. The authors conclude that an annealing
treatment for 1 hour at 800 ºC did not form any solid solution of Ti-O-Si,
and the initial oxides were present in the composite material as separate
phases.
A study of the physical characteristics of silicon dioxide aerogel with
5 % titanium dioxide was made in [82]. The composite material was syn-
thesized from nanosized SiO2 gel in ethanol with an addition of nanosized
TiO2 (anatase). The emulsion was dried in air and subjected to thermal
treatment. An IR-spectrometry analysis did not reveal any formation of
the Ti-O-Si bonding either in the initial composite material or after an-
nealing up to 1000 ºC, the initial oxides did not form any solid solution.
With the use of TEM it was found out that TiO2 particles are built into the
structure of silicon dioxide particles 5 nm in size. As in the previous pa-
pers, it was noted that the presence of silicon dioxide inhibits rearrange-
ment of the initial structure of titanium dioxide from anatase to rutile even
at 1000 ºC.
In [83], the results of investigation of a composite (TiO2)x(SiO2)1-x
material synthesized by a thermal decomposition of the metalloorganic
precursor containing silicon and titanium atoms in one molecule. In that
work, three compounds were studied [Si(OBut)2OTi(acac)2O]2,
[(ButO)3SiO]2Ti(OPri)2 and [(ButO)3SiO]3Ti(OPri), which differed by the
number of titanium atoms in the molecule of titansiloxanes. The compo-
site material was formed as a nanostructured film with the grain size from
4 to 200 nm (depending on the pyrolysis temperature). An examination of
the adsorption IR-spectra demonstrated that with increase in temperature
the intensity of the peak from the Si-O-Ti bond is decreased, while the in-
tensity of the adsorption peaks from the Si-O-Si and Ti-O-Ti bonds is in-
creased. This is indicative of the formation of separate phases of titanium
dioxide and silicon dioxide at a high-temperature pyrolysis. At high tem-
perature of decomposition of the initial metalloorganic compound, titani-
um dioxide formed a crystal anatase-type lattice. It is worth noting that
the minimum temperature required for titanium dioxide to crystallize in
the composite material under study was 600-800 ˚C for different precur-
sors. With increase in the number of titanium atoms in the precursor mol-

139
ecule, the minimum pyrolysis temperature at which the formation of crys-
tal structure in TiO2 is decreased. A crystal phase of titanium dioxide of
the rutile type was formed in minor quantities at the pyrolysis temperature
1100 ˚C only.

5.1.3 Non-equilibrium plasmochemical synthesis of nanosized

particles of metal oxides

Investigation into the synthesis of oxide nanoparticles was per-

formed with a TEU-500 pulsed electron accelerator [84]. The electron
beam parameters were as follows: the kinetic electron energy of 500 keV,
the pulse duration (full width at half maximum) of 60 ns, the energy per
pulse of 100 J, the pulse repetition frequency of 0.5 Hz, and the beam di-
ameter of 5 cm. The experimental system is presented in Fig. 58.

Fig.58. Experimental setup: 1 – vacuum pump; 2 – pressure gauge; 3 –

plasmochemical reactor, 4 – electron beam; 5 – anode foil

The reactor was a fused-silica cylinder of 14 cm inner diameter and

6 l volume. Titania nanoparticles were produced by injection of a pulsed
electron beam in a gas-phase mixture of TiCl4, oxygen, and hydrogen.
The complete conversion of TiCl4 was achieved by passing one electron
beam pulse; the energy consumption for conversion was 2.2 kJ/mol. The
reactor used was a cylinder made of quartz glass with an internal diameter
of 14 cm and a volume of 6 liters. Nanosized titanium dioxide was syn-
thesized by injecting a pulsed electron beam into a gas-phase mixture of
titanium tetrachloride, oxygen and hydrogen.

140
 Temperature Regime of Preparation of Oxide Nanoparticles
The use of a closed plasma reactor and the pulsed character of the
process make it possible to calculate a change in temperature during the
synthesis from pressure changes in the reactor. Fig. 59 shows the time de-
pendence of the reactor temperature. The temperatures were calculated by
the ideal gas law from experimentally measured pressures.

Fig. 59 (1) Change in temperature in the plasma reactor during the manu-
facture of oxide nanoparticles and (2) the temperature change due to heating a
gas mixture by electron beam only (without ignition)

The measurements showed that the temperature of the gas-phase

mixture during synthesis did not exceed 600°C and the reaction time was
less than 0.1 s. The electron beam-initiated process of TiCl4 degradation
in a mixture with hydrogen and oxygen is explosive in nature. Along with
the existence of the lower explosion limit (by pressure) and a low energy
consumption, this indicates the branched chain character of the process of
TiO2 synthesis by the conversion of TiCl4 in a mixture with oxygen and
hydrogen.
Geometric Size of TiO2 Nanoparticles
The particle size of a TiO2 powder was determined with a JEM-
100CXII electron transmission microscope. Fig. 60 and 61 shows a mi-
crophotograph of the particles and grain-size distribution (sample of over
1000 measurements). The grain size varied from 90 to 250 nm.

141
 Fig. 60.TEM image and bar chart of grain-size distribution of the na-
nosized titanium dioxide powder. Initial mixture composition in mmol: H2 + O2
+ TiCl4 (50 : 25 : 10)

Fig. 61. (a) TEM photograph and (b) the histogram of particle size distri-
bution of a nanosized titanium dioxide powder

The investigations also showed that the geometrical size of the syn-
thesized powder depends on the operating mode of the synthesis. A de-
crease in the titanium tetrachloride concentration in the initial mixture or
an addition of a buffer gas resulted in a grain size decreasing. This is in-
dicative of a bulk character of the process of synthesis. It is noteworthy
that the nanosized particles did not have any inner voids.
The shape and structure of nanosized particles of synthesized titani-
um dioxide were also controlled by the conditions of synthesis. At a low
concentration of titanium tetrachloride in the initial reactive mixture, TiO2
particles had a hexagonal nucleus with a characteristic tubular shell (Fig.
62a). With increased concentration of titanium tetrachloride in the initial

142
mixture, TiO2 particles mostly round, with the surface covered by smaller-
sized round particles (Fig. 62b). A further increase in the concentration of
TiCl4 in the initial mixture tended to yield particles with hexadral and cu-
bic cut and clear surface (Fig. 62c).

a b c
Fig. 62. Shape of particles of synthesized titanium dioxide for different
concentration of titanium tetrachloride in the initial mixture: 5 mmol (a), 10
mmol (b), and 15 mmol (c) Size of photo 250·250 nm (a), 1000·1000 nm (b) and
2000·2000 nm (c)

The particle size ranged from 90 to 200 nm and depended on synthe-

sis conditions. A decrease in TiCl4 concentration in the initial reactant
mixture led to a decrease in the numberaverage particle size to 30–40 nm,
thus indicating the occurrence of the process in the bulk. It is also note-
worthy that the prepared particles did not have internal voids.
X-ray Diffraction Analysis of TiO2
A crystalline TiO2 structure was formed under the given experi-
mental conditions, with the crystal lattice type (anatase or rutile) depend-
ing on the synthesis conditions (reactant mixture composition, total pres-
sure, etc.). Fig. 63 depicts X-ray diffraction patterns of a nanosized TiO2
powder, and Table 1 presents data on the ratio between the rutile and ana-
tase phases for different TiO2 samples. Table 1 also gives values for the
goodness of fit with the relevant tabulated data (21-1275 for anatase and
21-1276 for rutile).

143
 Fig. 63. X-ray diffraction patterns of a nanosized titanium dioxide powder
(pattern numbers 1, 3, 4 correspond to sample numbering in Table 1): (0) ana-
tase and (×) rutile

The fifth column in Table 9 gives the values of К, the percentage of

coincidence of the X-ray pattern with the tabular data (21-1272 for ana-
tase and 21-1276 for rutile).

Table 9. Results of X-ray diffraction analysis of nanosized titanium diox-

ide

Specimen Concentra- Rutile, Anatase, K, Average sd

tion of TiCl4, % % % size, nm
mmol
1 5 93 7 92
84-
2 5 90 10 138 44
86
3 10 86 14 85 148 54
4 15 33 67 80
5 10 + Ar 39 61 85 77 24

The initial gas-phase mixture had the following molar composition:

H2 + O2 + TiCl4 = 50:25:n. A high degree of coincidence of X-ray diffrac-
tion peaks with data of the standard spectra of crystalline titanium dioxide

144
indicate a low amount of defects in the crystal structure and the amor-
phous phase of the prepared TiO2 samples.
Despite that the temperature of the gas-phase mixture during the syn-
thesis did not reach 800°C, crystalline nanoparticles with the rutile crystal
lattice were formed, unlike the case of other TiO2 preparation methods.
This formation is due to the nonequilibrium character of the process. An
increase in TiCl4 concentration in the reactant mixture or the addition of
an inert gas (argon) primarily led to the formation of an anatase type lat-
tice.
IR Spectrometry of Nanosized TiO2
The prepared nanosized particles were studied with a Nicolet 5700
IR spectrometer over the range of 400-4000 cm–1 (resolution 4 cm–1).
Typical IR spectra of nanosized TiO2 is shown is Fig. 64.

Fig. 64. IR absorption spectra of nanoparticles samples TiO2 (1), rutile

TiO2 (2), and anatase TiO2. K is the reflectance

For TiO2 nanoparticles, the superposition of absorption bands due to

vibrations of the Ti-O-Ti bond in the TiO4 (650 cm–1, anatase lattice) and
TiO6 (500 cm–1, rutile lattice) groups is observed [85, 86].
Chemical composition of synthesized powder
In order to determine the chemical composition of synthesized pow-
ders, we performed an X-ray luminescence analysis in an Olympic
DC2000. Its results are given in Table 10.

145
 Table 10. The results of X-ray luminescence analysis

Element Concentration, Wt Element Concentration, Wt

% %
Ti 99.09 ± 0.08 Fe 0.48 ± 0.04
Si 0.24 ± 0.04 Cu 0.03
Cr 0.10 ± 0.01 Zn 0.03

This spectrometer is capable of registering the elements with the or-

der number more than 10 (Na and higher), so there is no data on oxygen
and chlorine. Taking into account the content of oxygen in the synthe-
sized titanium dioxide powder, the impurity density is as low as 0.6 %.

5.1.4 Research of composite nanosized oxides (TiO2)x(SiO2)1-x

synthesized using a non-equilibrium plasmochemical process

Nanosized composite (TiO2)x(SiO2)1-x was synthesized by injecting a

pulsed electron beam into a gas-phase mixture of titanium tetrachloride,
silicon tetrachloride, oxygen and hydrogen. The experimental system is
presented in fig. 65.

Fig. 65. Experimental setup: 1 – vacuum pump; 2 – pressure gauge; 3 –

plasmochemical reactor, 4 – electron beam; 5 – anode foil

146
 Decomposition of mixture of halides (TiCl4+ SiCl4 or SiCl4+CCl4)
was completed within a single electron beam pulse. With the electron
beam energy 100 J, the energy consumption for decomposition of mixture
of halides was 5 kJ/mol. This is by far smaller than the TiCl4 dissociation
energy that is equal to 804 kJ/mol. The process of destruction of TiCl4+
SiCl4 with hydrogen and oxygen with an electron beam injected into this
mixture had an explosive character. This circumstance, the presence of a
lower limit (on pressure) of the reactive mixture ignition, and the low en-
ergy consumption are indicative of a cross-linked chain character of the
process of composite (TiO2)x(SiO2)1-x synthesis through decomposition of
TiCl4 + SiCl4 in a mixture with oxygen and hydrogen.
In order to understand whether it is possible to simultaneously syn-
thesize nanosized oxides of different materials, a number of experiments
have been conducted to excite a mixture of oxygen, hydrogen, titanium
tetrachloride and tetrachlorsilane by a pulsed electron beam. For the syn-
thesis of composite oxides use was made of technological tetra-
chlorsilane. The reactor was heated to 90 ºС and, before filling in the gas
mixture, evacuated to ~1 Pa. Upon injecting the electron beam into the
following mixture (mmol): H2 + O2 + SiCl4 + TiCl4 (50 : 25 : 17 : 10), na-
nosized powder was formed in the reaction chamber. Fig. 66 shows a
SEM image of the resulting powder.

Fig. 66. Scanning electron image of the synthesized composite

(TiO2)x(SiO2)1-x

Presented in fig. 67 are a TEM image of the powder and a bar chart
of grain-size distribution. Note that the mean size of the composite pow-
der particles is smaller compared to that of pure nanosized titanium diox-
ide synthesized under similar experimental. This might be attributed to a
change in the conditions of coagulation of the particles formed upon addi-
tion of a new material.

147
 Fig. 67 TEM image and a bar chart of grain-size distribution from
(TiO2)x(SiO2)1-x powder. The mean size is 29 nm. The initial mixture in mmol:
H2 + O2 + SiCl4 + TiCl4 (50 : 25 : 17 : 10)

The resulting composite nanosized oxides may be present as a mix-

ture of individual particles of silicon dioxide and titanium dioxide proper.
In order to determine their lattice structure and mixture ratio, the synthe-
sized powder was analyzed in a high-resolution transmission electron mi-
croscope. In addition to TEM images, an analysis was made of the dark-
field images of the particles. Brighter images corresponded to the region
of coherent scattering. All the particles of the composite oxide under ex-
amination exhibited reflections in the diffraction pattern. Smaller particles
in the dark-field image were fully bright (fig. 68).

148
 Fig. 68. Dark-field TEM-image of a small composite powder particle and
its diffraction pattern

Large particles had individual regions of coherent scattering in their

structure (Fig. 50).

Fig.69. Dark-field TEM-image of a large composite powder particle and

its diffraction pattern

The high-resolution electron microscopy examination of the synthe-

sized composite oxide demonstrated that in our case the mixture of SiO2
and TiO2 in one particle are formed.

149
 The chemical composition of the synthesized composite powder was
determined in an Oxford ED2000 X-ray fluorescence spectrometer, and
the results are listed in Table 11. Taking into account the content of oxy-
gen in the synthesized composite powder, the impurity concentration is
less than 0.4 %.

Table 11. Chemical composition of the powder (TiO2)x(SiO2)1-x

El- Content, El- Content,

ement mass % ement mass %
Si 55.90±0,08 Mn 0.049±0,01
Ti 43.58±0,01 Cu 0.040±0,01
Fe 0.225±0,01 Zn 0.040±0,01
Cr 0.100±0,01

The method of plasmochemical synthesis developed here ensures

that the chemical composition is maintained for different-size nanoparti-
cles. Presented in Fig. 70 are the images of large (over 100 nm) and small
(20…40 nm) particles of (TiO2)x(SiO2)1-x and the total EDX-spectrum.
The EDX-spectrum was also measured separately for large and small par-
ticles of the synthesized powder. The area of the respective peaks is given
in Table 12 (in percent).

Fig. 70. TEM-image and an EDX-spectrum from the (TiO2)x(SiO2)1-x pow-

der. Large and small nanoparticles together

150
 Table 12. Percentage of oxygen, silicon and titanium in the composite
powder (TiO2)x(SiO2)1-x according to the EDX-spectra

Element Full spectrum, % Large particles, % Small particles, %

O 43 41.4 44.7
Si 37.5 39.1 43
Ti 19.4 19.5 12.2
An X-ray diffraction analysis of the synthesized composite na-
nosized powder demonstrated that both titanium dioxide proper and com-
posite nanopowder (TiO2)x(SiO2)1-x have crystal structure. It is well de-
scribed by a composition of two types of lattices characteristic of titanium
dioxide proper (rutile or anatase). Fig. 71 shows the X-ray diffraction pat-
terns from two specimens of the synthesized oxides (rad. Co, λ =1.7901
A), and table 13 presents the data on the proportion of rutile and anatase
phases of (TiO2)x(SiO2)1-x for different specimens.

Table 13. The proportion of rutile and anatase phases of (TiO2)x(SiO2)1-x

for different specimens
Spec- Ru- Ana-
Regime of synthesis
imen tile tase
1 46 mmol H2+ 23 mmol O2 +25 mmol SiCl4+ 9 85
15%
mmol TiCl4 %
2 46 mmol H2+ 23 mmol O2 +25 mmol SiCl4+ 18 47
53%
mmol TiCl4 %

Fig. 71. X-ray diffraction pattern from nanosized (TiO2)x(SiO2)1-x powder

(curves 1 and 2 correspond to table 3)

For comparison, Fig. 72 depicts the X-ray diffraction patterns from

silicon dioxide, titanium dioxide, and composite dioxide (TiO2)x(SiO2)1-x.

151
 Fig. 72. X-ray diffraction patterns from the TiO2, SiO2 and
(TiO2)x(SiO2)1-x specimens

All the powders were synthesized by exposing the respective halide

with a stoichiometric mixture of oxygen and hydrogen to a pulsed elec-
tron beam. A characteristic feature of the composite powder is the pres-
ence of a few maximums within the low-angle region (2Θ = 20-45°) and
in the region from 75 to 90 degrees, which are due to the presence of an
amorphous phase of silicon dioxide.
The synthesized nanoparticles were examined in an IR-spectrometer
Nicolet 5700 within the range from 400 to 4000 cm-1, the resolution being
4 cm-1. The powder under study was preliminary mixed with KrB and
compacted into a tablet. The reflectance spectrum of KrB proper was de-
ducted from that of the mixture. Fig. 10 presents a characteristic absorp-
tion spectrum from composite (TiO2)x(SiO2)1-x nanopowder. For compari-
son, shown in the figure are also the IR absorption spectra from compo-
site powder and silicon and titanium oxides proper. The absorption peaks
with the wave numbers 1190, 1080, and 790 cm-1 correspond to tetrahe-
dral structure of silicon dioxide (νas(SiO4) LO, νas(SiO4) TO and νs(SiO4),
respectively) [79]. In [87], absorption at 960 cm-1 is attributed to the pres-
ence of a Ti-O-Si bond. Absorption in this region, however, is observed
not only for the composite oxide (TiO2)x(SiO2)1-x but also for silicon diox-
ide proper (Fig. 73). It is, therefore, very likely that absorption of IR light
at 960 cm-1 corresponds to vibrational excitation of a Si-O-H bond, whose
absorption peak corresponds to this wave number [87].

152
 Fig. 73. IR spectra from the specimens made of (TiO2)x(SiO2)1-x (1), SiO2
(2), and TiO2 (3)

The proposed method of synthesizing nanosized titanium dioxide

and composite nanosized powder (TiO2)x(SiO2)1-x exhibits very high out-
put efficiency (1-1.1 kg/h calculated per final product) and low e-beam
energy consumption (0.1-0.15 (kW∙h)/kg). The initial bulk density of the
nanopowder was (6-10) g/l.

5.1.5 Synthesis of composite oxides Si-C-Ox

One of the large-capacity businesses intensively developing nowa-

days is the production of technological carbon. The major consumers of
carbon – tire manufactures – need high-quality technological carbon that
will ensure higher service safety of vehicles. Moreover, an improved
quality of technical carbon also reduces fuel consumption and increases
tire service life. Composite nanosized materials containing carbon and sil-
icon dioxide have been recently introduced into tire production, which al-
lows the manufacturers to improve a number of service characteristics of
the tires. Currently, there is an evident interest to the research into alterna-
tive techniques for carbon production, including plasmochemical meth-
ods. Our experimental investigations have demonstrated that a non-
equilibrium plasmochemical process initiated by a pulsed electron beam
is an effective way to synthesize composite nanosized oxides Si-C-Ox.

153
Shown in Fig. 74 are the TEM-image of the synthesized powder and the
bar chart of grain-size distribution.

Fig. 74. TEM-image and a bar chart of grain-size distribution in Si-С-Ox.

Initial mixture: 23 mmol O2 +46 mmol H2 + 25 mmol SiCl4 + 20 mmol СCl4

The average grain size of the resulting composite Si-C-Ox powder

(46.5 nm) is smaller than that of nanosized silicon dioxide proper. An X-
ray diffraction analysis of the synthesized powder reveals that it is
amourphous. Fig. 75 shows a high-resolution TEM- image of a particle of
the synthesized Si-C-Ox powder.

Fig. 75. TEM-image of the synthesized nanosized Si-C-Ox powder

The synthesized nanoparticles were examined in a Nicolet 5700 IR

spectrometer within the range from 400 to 4000 cm-1. Fig. 76 presents a
characteristic absorption spectrum from the nanosized composite Si-C-Ox
powder.

154
 Fig. 76. IR-spectrum from silicon dioxide proper (1) and composite Si-C-
Ox oxide (2)

The experimental research carried out in this work has demonstrated

that a plasmochemical process initiated by a pulsed electron beam is an
effective way to synthesize nanosized composite oxides from a gas-phase
mixture of oxygen, hydrogen, and a mixture of halides. The resulting na-
nooxides exhibit homogeneous composition and their particles are spheri-
cal in shape with facetting and without voids. Crystalline TiO2 nanoparti-
cles were prepared from a gaseous mixture of oxygen, hydrogen, and
TiCl4 in a nonequilibrium plasma-chemical process initiated by pulsed
electron-beam radiation. The use of a mixture of chlorides TiCl4 and SiCl4
makes it possible to prepare nanoparticles of mixed oxides. The obtained
oxides are homogeneous in composition, and their particles have a char-
acteristic faceted shape without inner voids. A change in the composition
of the reactant mixture allows the crystal structure, as well as the shape
and size, of synthetic TiO2 to be varied. A distinctive feature of the pro-
cess developed in this study is a considerable reduction in the synthesis
temperature of particles with a crystalline (rutile and anatase) structure, a
result which is due to the nonequilibrium nature of the processes in-
volved.

155
 5.2. Plasmochemical conversion of methane

Processing of natural gas and associated hydrocarbon gas is one of

the most important tasks of contemporary gas chemistry. Geological stud-
ies have shown that not only biogenic methane formation processes, but
also the continuing degassing of the Earth play dominating role in the
formation of natural gas reserves. As a result, up to 2 trillion cubic meters
of gas (its world production level) a year get into the Earth crust. This al-
lows us to treat natural gas as a partially renewable material and energy
source.
A wealth of studies on methane conversion in various electric dis-
charges and under continuous and pulsed electron-beam irradiation have
been performed to date, thus making it possible to reveal the specifics of
methane degradation reactions in a low-temperature plasma. The aim of
this review is to analyze the influence of the nonequilibrium conditions in
plasma on the energy consumption, selectivity, and the product composi-
tion. The studies of the pyrolysis, partial oxidation, and carbon dioxide
and steam reforming of methane in a low-temperature plasma are consid-
ered.

5.2.1 Plasma pyrolysis of methane

The pyrolysis of methane (conversion without addition of oxygenat-

ed compounds) is the most attractive process of natural gas reforming in
which compounds with a higher molecular mass of the motor fuel range
can be synthesized. The plasma synthesis of acetylene by methane pyrol-
ysis in arc discharge has been studied and worked out on the engineering
scale to the greatest extent. The degree of conversion in this process
reaches 95-98 % at an acetylene selectivity of 90-95 % and an energy
consumption of 2.5-3 eV/molecule [88]. Under equilibrium conditions, an
energy consumption of 1.8 eV per methane molecule is required for the
reaction 2CH4 = C2H2 + 3H2, regardless of the energy source.
Numerous studies of hydrocarbon decomposition upon heating have
shown that the process is chain in character (thermal cracking). At low
temperatures when thermal initiation of cracking is impossible, an electric
discharge generates active centers-free radicals, ions, or excited mole-
cules-which can initiate chain hydrocarbon cracking processes. The chain
mechanism of methane pyrolysis in pulsed microwave-discharge plasma
(9 GHz, 100 kW) was experimentally studied in [89, 90]. In the case of
methane preheating to 700-1100 K, the expenditure of plasma energy for

156
hydrogen synthesis does not exceed 1 eV/molecule. Similar results were
obtained in a study of methane pyrolysis in a gliding alternating-current
(50 Hz, 3 kW) arc discharge [91]. At a gas flow rate of 2 m3/h, a pressure
of 6 atm, and a temperature of 1400 K, 34 % of CH4 was converted into
hydrogen and acetylene. The arc-discharge energy consumption for me-
thane pyrolysis was 0.82 eV/molecule.
The pyrolysis of CH4 in atmospheric-pressure spark, pulsed stream-
er, and dielectric-barrier (dc and ac) discharges at room temperature re-
quires quite a considerable consumption of energy [92]. For methane deg-
radation, the energy costs (eV/molecule) are 14-25 in spark discharge, 17-
21 in streamer discharge, 28-57 in dc dielectric-barrier discharge, and
116-175 in ac barrier discharge. Pulsed spark discharge results in a high
yield of acetylene (54%) and hydrogen (51 %) at a CH4 conversion of
69 %. In the case of barrier discharge, ethane prevails in the conversion
products.
A high efficiency and selectivity of thermocatalytic processes of
conversion of methane and other hydrocarbons stimulated studies on the
plasma pyrolysis of CH4 in the presence of catalysts. Li et al. [93] report-
ed the results of study of the two-stage pyrolysis process in pulsed spark
discharge (first step) and further in a reactor with a catalyst (second step),
which was Fe-, Co-, Ni-, Cu-, or Zn-coated grains of the HZSM-5 catalyst
(SiO2/Al2O3 = 35). The catalyst temperature was 623 K. In the absence of
the catalyst (Fig. 77), the methane conversion was 41 % at an energy con-
sumption of 17 eV/molecule and the product composition was 84 % acet-
ylene and 75 % hydrogen (percent of the maximal theoretical yield).

157
 Fig. 77. Methane conversion and product composition (1) in the absence
of catalyst and in the presence of the (2) HZSM-5, (3) Fe+HZSM-5, (4)
Co+HZSM-5, (5) Ni+HZSM-5, (6) Cu+HZSM-5, and (7) Zn+HZSM-5 cata-
lysts
The addition of the uncoated HZSM-5 catalyst increases the yield of
aromatic hydrocarbons (primarily, benzene, toluene, and xylene) to 15 %
with rise in the energy costs to 19 eV/molecule. The Ni coating of grains
provides for the highest yield of aromatic hydrocarbons (47 %) and a sta-
ble catalytic activity for 3 h. Methane pyrolysis in plasma is comprehen-
sively surveyed in [94, 95], and the data are collated in Table 14.

Table 14. Plasma pyrolysis of methane

Discharge type Reactant Main prod- Conversion Energy con-

mixture ucts % sumption,
eV/molecule
Pulsed microwave, 700 CH4 C2H2, H2 90 - 100 1
– 1100 K
Gliding arc, 50 Hz, CH4 C2H2, H2 34 0.82
1400 K
Gliding, 50 Hz CH4 + H2 (75 - 84)% 18 1.6 - 4.4
+ Ar C2
Direct-current arc, 60 CH4 + H2 C2H2 95 4.2
kW
Direct-current arc, 160 CH4 + N2 33% H2 80 8 – 10
V, 75 A
Pulsed, 8 kHz CH4 - - 8.5
Pulsed glow, 50 – 200 CH4 - - 9.1 - 10.3
Hz
Pulsed double-pointed, CH4 85% C2H2 23.5 3.8
2 – 20 kHz
Corona, 0.2 – 10 kHz CH4 - 25 52
Spark, 50 Hz, 5 kV CH4 C2H2 70 6.1
Streamer, 4.6 W CH4 C2* 19 - 41 17 - 21
Dielectric-barrier, 75 CH4 C2H6 6 - 13 38 - 57
kHz, 17 W
Pulsed electron beam, CH4 C2* 5-7 18
450 kV, 10 kA
Gliding radiofrequen- CH4 + Ar C2H2, C2H4 75 12
cy, 13.56 MHz, 60 torr
Dielectric-barrier with CH4 C2H6 + H2 45 52
SiO2 catalyst

158
 Spark with Fe, Co, or CH4 46% (C6H6, 40 19
Ni catalyst on HZSM-5 C7H8,
C8H10)
Gliding (50 Hz) with CH4 + H2 (83 - 100)% 18 2.4 – 7.3
Al2O3 catalyst + Ar C2*
* C2 includes ethane, ethylene and acetylene.

Analysis of studies on the plasma pyrolysis of methane has shown

that the minimal energy consumption and the maximal degree of conver-
sion were obtained under equilibrium conditions, in dc arc discharge
(3 eV/ molecule, 95 %). The use of different types of pulsed discharge for
CH4 pyrolysis neither reduces the specific energy consumption nor in-
creases the conversion. The best results were obtained in pulsed spark
discharge (3.8 eV/molecule, 28.5 %), but they are worse than the results
of methane pyrolysis in arc discharge plasma. The main degradation
products are C2 hydrocarbons with the highest proportion of acetylene.
The nonequilbrium character of plasma interaction was revealed on-
ly in the case of microwave-discharge or gliding arc discharge treatment
of methane preheated to the thermal degradation onset temperature. In
this case, the plasma energy consumption for decomposition is lower than
that under the equilibrium conditions and does not exceed 1 eV/molecule.

5.2.2 Partial oxidation of methane

At present, the overwhelming bulk of natural gas is processed into

synthetic liquid fuel via the Fischer–Tropsch process designed in 1923.
For manufacture of liquid fuel, carbon compounds (natural gas, black and
brown coals, heavy petroleum fractions, etc.) are converted into synthesis
gas (mixture of CO and H2) and, then, into synthetic crude oil. The liquid
fuel manufacture technology requires high capital and operation costs at
the syngas production step.
The methane oxidation process is a branched-chain process; it has
been thoroughly studied under equilibrium conditions and is described in
[96, 97]. The use of low-temperature plasma in the partial oxidation of
CH4 holds promise primarily for initiation of the chain process. Data on
the partial oxidation of methane via the chain process in microwave dis-
charges of two types are reported in [98]. The plasma energy consump-

159
tion was 0.25 eV/molecule at a methane conversion of 70 % and in-
creased to 0.5 eV/molecule when the conversion increased to 100 %.
Data that are interesting from the viewpoint of industrial implemen-
tation were obtained in a study of the partial oxidation of methane in arc
discharges of different types. In a two-electrode arc plasma reactor, the
partial oxidation of CH4 in a stoichiometric mixture with oxygen is char-
acterized by an energy consumption of at most 1 eV/molecule [99].
Methane conversion in a mixture with oxygen (CH4 : O2 = 1 : 2) in a
rotating-arc plasma reactor also showed a low energy consumption [100].
As the discharge energy input to the gas increased from 2.6 to 4.8 kJ/l at a
gas flow rate of 15-30 l/s, the consumption for methane conversion in-
creased from 1.6 to 2.2 eV/molecule. Fig. 78 shows the results of a study
of the conversion process.

Fig. 78. The degree of conversion of methane α and hydrogen content η in

the products conversion versus the input energy of discharge

Similar results were obtained during the partial oxidation of CH4 in

its mixture with air in vortex gliding arc discharge (10 kV, 15–20 kHz)
[101]. As the initial methane concentration in the reactant mixture was
increased from 20 to 44 %, the energy consumption for conversion in-
creased from 1.4 to 5 eV/molecule. Fig.79 shows the plasma-chemical re-
actor.

160
 Fig. 79. Plasma-chemical reactor with a surface arc discharge

In the case of pulsed electron-beam irradiation (350–450 keV,

10 kA, 60 ns), partial methane oxidation in a mixture with oxygen and
hydrogen occurred in the nonequilibrium mode [102]. The process of in-
complete oxidation of CH4 was effective at a high initial concentration of
oxygen (>30 %) in the reactant mixture. As the initial pressure of methane
in the H2 + O2 + CH4 mixture was increased at a constant oxygen pres-
sure, the proportion of CO and H2 in the products increased. The energy
consumption for CH4 conversion did not exceed 0.05 eV/molecule.
It was experimentally shown that the joint action of low-temperature
plasma and catalyst substantially alters the composition of the products of
methane partial oxidation. Indarto et al. [103] studied the conversion of
methane in a 4:1 mixture with oxygen into methanol in a dielectric-barrier
discharge plasma (10 kV, 50 W). To increase the conversion efficiency,
the CZA (copper–zinc–aluminum) catalyst placed at the end of the plas-
ma reactor was used without additional heating. In the absence of the cat-
alyst, the main products of partial methane oxidation in the barrier dis-
charge were CO, CO2, and H2, and the methanol content did not exceed
11 %. In the presence of the catalyst, the yield of methanol increased by a
factor of 2. The process comprised two steps; the production of synthesis
gas in low-tem perature plasma and the subsequent formation of methanol
on the catalyst. At CH4 conversion of 28-33 % and a feed-mixture flow
rate of 30 ml/min, the energy consumption for methane conversion was
88-100 eV/molecule. Promoting the catalyst with different metals made it
possible to increase the selectivity for methanol (Fig. 80).

161
 Fig. 80. Concentration of methanol in the products of plasma reforming of
the CH4 + O2 mixture in the presence of different catalysts.

The data on the partial oxidation of CH4 in different discharges are

summarized in Table 15, and a more detailed review is presented in [94,
95].
Table 15. Partial oxidation of methane in plasma

Discharge type Reactant Main prod- Conversion Energy con-

mixture ucts % sumption,
eV/molecule
Direct-current arc, 10 CH4 + O2 CO + H2 - 1
kW + Ar
Rotating arc, 2 kW, 5 - CH4 + O2 50-60 % H2 80 1.6-2.2
20 kHz
Vortex rotating arc CH4 + Air H2 / CO = 30 1.4-5
1.1-2.7
Microwave, 2.45 GHz, CH4 + O2 CO + H2 100 0.5
1-5 kW
Dielectric-barrier, 75 CH4 + O2 CO, CH3OH 14-27 5-13
kHz
Dielectric-barrier, 20 CH4 + O2 22 % 24 11
kHz CH3OH
Dielectric-barrier with CH4 + O2 30- 35 % 28-33 88-100
CZA catalyst CH3OH
Dielectric-barrier with CH4 + O2 20-23 % 50 88
Yt + Zr catalyst CH3OH

Thus, the highest yield of the synthesis gas at a low energy con-
sumption was obtained in various arc discharges, which provide a high

162
methane conversion capacity per unit volume of the reactor and do not
require a high-voltage power source. The use of combined (plasma +
catalyst) reactors makes it possible to control the composition of conver-
sion products, in particular, to increase the selectivity for CH3OH from 3
to 35 %.

Carbon dioxide reforming of methane

Methane conversion in the presence of carbon dioxide is considered
by many investigators to be an alternative CO2 utilization process aimed
to mitigate the greenhouse effect and to simultaneously synthesize useful
products. Under equilibrium conditions, the reaction:
CH4 + CO2 + 2.5 eV = 2CO + 2H2
is run over a Ni catalyst at a temperature of 850 ºC and a pressure of
2-3 MPa [97].
Plasma methane conversion in a controlled arc dis charge was stud-
ied in [99]. In the case of conversion of the CH4 + CO2 mixture at atmos-
pheric pressure, the energy consumption for decomposition was
15 eV per methane molecule. The use of an improved reactor design with
gliding arc discharge (480 V, 20 A) made it possible to reduce the energy
consumption to 1.5-2.3 eV/molecule [99]. At a gas flow rate of
1.8-4.7 m3/h, atmospheric pressure, and ambient temperature, the main
products were H2, CO, and H2O.
When the reactant gas mixture was heated to 1200 K, the energy
consumption of a pulsed corona discharge for carbon dioxide methane re-
forming did not exceed 0.7 eV/molecule, thereby indicating the chain
mechanism of the conversion [104].
The conversion of methane in a mixture with carbon dioxide
(CH4/CO2 = 1) in dielectric-barrier discharge plasma requires high energy
consumption [105, 106]. The study of the degree of conversion depending
on the voltage pulse amplitude, pulse repetition frequency, and the form
(uni- or bipolar pulses, 4-12 kV, 0.4-2 kHz) shows that the conversion
depends only on the specific energy input to the discharge. Variations in
the energy input by changing the voltage (pulsed power) and pulse repeti-
tion frequency (average power) have the same effect on the conversion.
As the energy input was varied from 16 to 70 J/cm3, the degree of CH4
conversion in the case of the bipolar pulse form increased from 9.8 to
22.8 %, and the energy consumption increased from 38 to 70
eV/molecule. The data on carbon dioxide reforming of methane in differ-
ent discharges are collated in Table 16.

163
Table16. Carbon dioxide reforming of methane in plasma

Discharge Reactant mix- Main products Conversion, Energy

type ture % Consumption,
eV/molecule
Deirect- CH4+CO CO+H2 _ 1.5-3.3
current glid-
ing arc
Pulsed coro- CH4+CO2 CH2+CO+H2 16 0.7
na, 1200K
Alternating- CH4+C2H6+CO2 H2+C2 8 20
current
glididng arc
Dielectric- CH4+CO2 CO+H2 9.8-22.8 38-70
barrier, 0.4-2
kHz
Dielectric- CH4+CO2 CO+H2 25-65 73-100
barrier, 2-40
kHz
Dielectric- CH4+CO2 CO+H2+C2H6 35-75 52-85
barrier, 30
kHz
Dielectric- CH4+CO2 CO+H2+C2 30-50 _
barrier with
(Ir,Pd,Re)/γ-
Al2O3 cata-
lyst
Dielectric- CH4+CO2 HCOH,CH3OH 48 16
barrier with
catalyst

The results that hold promise for industrial implementation were ob-
tained only for methane conversion in a gliding discharge. In discharges
of other types, the separate conversion of CO2 into CO and CH4 into hy-
drocarbons (C2H6, C2H4, C2H2, etc.) takes place. The electrode material
and the catalyst have a great effect on the composition of the conversion
products of the CH4 + CO2 mixture.

Steam reforming of methane

Steam methane reforming provides the maximal yield of hydrogen

and, therefore, holds promise for hydrogen economy and is being inten-

164
sively studied. Under equilibrium conditions, the steam reforming of me-
thane proceeds in accordance with the reaction:
CH4 + H2O +2.1 eV = CO + 3H2
at a temperature of 1300–1900 K and a pressure above l5 MPa [97].
The conversion of methane in a mixture with steam in gliding dis-
charge plasma is characterized in that the energy consumption for CH4
decomposition does not exceed 0.92 eV/molecule [107]. A specific fea-
ture of the reactor design is the rotation of the plasma string by an exter-
nal magnetic field (Fig. 81). At a discharge power of 1 kW, a feed mixture
flow rate of 30 l/min, a temperature of 150 ºC, a pressure of 0.1 MPa, and
a CH4/H2O ratio of 0.67, the degree of conversion reaches 50 %. The
main products are hydrogen (55 vol %) and CO (10 vol %); the amount of
CO2 and C2H2 does not exceed 1 vol %.

Fig.81: A scematic of a new reactor

Energy consumptions close to the enthalpy of the process were

achieved in the case of steam methane reforming in a rotating gliding arc
plasma reactor [108]. For the CH4/H2O = 0.2 mixture, a 44 % conversion
was attained at an energy cost of 2.6 eV per methane molecule.

165
 The results of investigation of CH4 conversion in a mixture with
steam via the chain mechanism are reported in [109]. The experiments
were conducted in a microwave plasma generator (915 MHz; pulse pow-
er, 200 kW) with preheating the reactant mixture to 500-570 ºC; the
CH4/H2O ratio was varied from 0.5 to 1. The gaseous products of the pro-
cess were syngas and CO2. Azizov et al. [109] note that the total energy
consumption of the process is reduced by 30-60 % when the discharge is
switched on, although the discharge energy does not exceed 5 % of the
thermal energy. At a cumulative energy input of 2 J/cm3, the concentra-
tion of the product hydrogen was 11 %, a value that corresponds to the
total energy consumption for H2 synthesis of 4.2 eV/molecule.
Sekine et al. [110] examined the dependence of the conversion and
the product composition on the spark-discharge pulse repetition frequency
(20-300 Hz) and the proportion of CH4 in a mixture with water vapor. In
all experiments a quartz tube of 4.0 mm inner diameter was used as a flow
type reactor as shown in Fig. 82.

Fig. 82. Experimental setup and configuration

At a methane conversion of 55 %, a discharge power of 2.1 W, and a

feed mixture flow rate of 10 cm3/min, the energy consumption for conver-
sion was 5.2 eV/molecule. A distinguishing feature was a high H2/CO ra-
tio, 4-4.5. This value is noticeably higher than the ratio characteristic of
steam reforming under equilibrium conditions, H2/CO = 3.
In addition to steam reforming of methane, authors [110] investigat-
ed about steam reforming of other hydrocarbons and alcohols using non-
equilibrium pulsed discharge. Propane served as a model for LPG and n-
hexane and cyclohexane served as a model for gasoline. The results are
shown in Fig. 83.

166
 Fig. 83. Steam reforming of various fuels using non-equilibrium pulsed
discharge.
The cited authors [110] noted that the energy utilization efficiency of
the pulse discharge for the synthesis of hydrogen and CO reached 60 %.
Zhdanok et al. [111] reported the results of study of steam and
steam–air methane reforming in a combined reactor employing a high-
voltage atmospheric-pressure discharge and catalysts (Ni or Fe2O3). Fig. 1
gives a diagram of experimental setup to make studies on partial methane
oxidation in APHVD plasma. The major components of the setup are a
plasmachemical reactor and a gas chromatograph. A cooler and a hotwell
are installed at the reactor output (Fig. 84).
The given experimental study used two types of catalysts: hydro-
genation nickel catalyst Synetix PRICAT 400 (RP) 2.5 mm (15.7 % Ni on
Al2O3) and carbonic oxide conversion catalyst СТК - TZC 3/1 (Fe2O3)
with chrome additives.

167
 Fig.84. Experimental set-up

The use of nickel made it possible to reduce the discharge energy

consumption for H2 production by a factor of 2.5. Data on steam reform-
ing in discharges of different types are collated in Table 17.

Table 17. Steam methane reforming in plasma

Discharge type Main prod- Conversion, % Energy

ucts
consuption
eV/molecule
Direct-carrent gliding H2+CO 50 0.92
arc H2+CO 44 2.6
Direct-current roraiting
gliding arc H2+CO+CO2 - 2.1
Microwave discharge, 55
915 MHz, 500-700 0C H2+CO 5.3
Spark discharge, 20- 77

168
 300 Hz H2+CO+CO2 1.6
Spark discharge with
Ni catalyst

From the data presented in the table, it is seen that steam methane re-
forming in most low-temperature discharge plasmas requires a low energy
consumption, equal to or lower than the enthalpy of the process. This
mode of plasma-assisted conversion of methane is the most promising for
industrial implementation; however, it has been little studied.

Combined Processes of methane conversion in plasma

The combination of the endoergonic reactions of carbon dioxide and

steam reforming with the exothermic reactions of partial oxidation of me-
thane makes it possible to substantially reduce the energy consumption
for CH4 conversion. The carbon dioxide reforming of natural gas in glid-
ing ac arc discharge was studied by Chavadej et al. [112]. The addition of
O2 to the CH4 + C2H6 + C3H8 + CO2 mixture at a 75 % hydrocarbon con-
tent of the initial reactant mixture reduced the energy consumption for the
decomposition of reactant molecules from 19.4 to 14.3 eV/molecule, and
the energy consumption for H2 production decreased from
32 to 19.4 eV/molecule.
In a study of the plasma reforming of methane in a mixture of steam
and air in a direct-current plasmatron, a 40 % hydrogen yield (on a hy-
drogen content in methane basis) and an energy consumption of
2.1 eV/molecule were achieved [113]. The use of the Ni/Al2O3 catalyst in
combined methane reforming in the plasmatron made it possible to in-
crease the yield of H2 to 100 % and to reduce the energy consumption to
0.9 eV/molecule at a 70 % methane conversion.
To reduce the energy consumption for CH4 conversion in a mixture
with air and steam in an arc plasmatron, the Ni/Al2O3 catalyst (United
Catalyst C-11) was used [114]. At a 70 % methane conversion, the energy
consumption for hydrogen production was reduced from 1 eV/molecule
(in the absence of catalyst) to 0.35 eV/molecule.
Steam methane conversion with a an admixture of 5 % oxygen in a
continuous microwave discharge (915 MHz, 200 kW) is described in
[90]. At a gas flow rate of 200 m3/h and a pressure of 0.1-1 atm, a 90 %
conversion of CH4 was achieved with an energy consumption for H2 pro-

169
duction of 0.9-1.0 eV/molecule. The data on combined methane reform-
ing in plasma are presented in Table 18.

Table 18. Combined reforming of methane in plasma

Discharge type Reactant mix- Main products Conver- Energy

ture sion, % consuption
eV/molecul
e
Gliding arc, CH4+C2H6+C3H H2,CH4,C2H2,C 30-70 14.3
17.5 kV,300 Hz 8+CO2+O2 O
Direct-current
arc, 3.5 kW CH4+H2O+air H2 80 4.2
Direct-current
arc and CH4+H2O+air H2, CO, CO2 70 0.35
NiO/Al2O3 cata-
lyst ____
Roraiting, 10
kV, 50 Hz CH4+H2O+air ____ 2-4 6
Microwave dis-
charge, 915 CH4+H2O+O2 90 0.9-1
MHz, 200kW

Thus, the most important parameter from the view point of commer-
cialization of the plasma reforming of methane is the energy consumption
for its decomposition. All types of discharge can be divided into two
groups in accordance with the energy consumption, degree of conversion,
and selectivity for products, namely, inhomogeneous (arc, spark, gliding)
and space (dielectric-barrier, and corona) discharges.
In inhomogeneous discharges, the methane conversion efficiency is
higher, the energy onsumption for decomposition is below
10 eV/molecule; the conversion in arc discharges exceeds 90 % and the
selectivity reaches 90 % for certain products (C2H2 in plasma pyrolysis,
H2 and CO in steam reforming). In space discharges, a high degree of
conversion (>50 %) is reached only at high energy consumptions for me-
thane decomposition (>40-50 eV/molecule) and a broad range of products
with a low selectivity are synthesized.
A substantial reduction in the discharge energy consumption for me-
thane decomposition is attained if the conversion is conducted in the
chain mode; in this case, the energy consumption does not exceed 1
eV/molecule. A promising line in the plasma reforming of methane is its

170
conversion in a mixture with steam. In this case, the energy consumption
of gliding discharge for CH4 decomposition is lower than the enthalpy of
the steam methane reforming under the equilibrium conditions. The yield
of H2 considerably exceeds the equilibrium values, and the H2/CO ratio is
5-10.
The review of the experimental works on gas-phase plasmochemical
processes has demonstrated that the conditions developed under the action
of a pulsed electron beam on gas are favorable for initiation of chain
chemical processes. In contrast to other plasma formation techniques, a
high-current pulsed electron beam provides a considerable reduction in
the energy of electrophysical installation for completing the chemical re-
action, which is a critical factor in the case of limited power supply. Un-
der non-equilibrium conditions generated by a pulsed electron beam, no
inhibiting effect of high oxygen concentration on partial methane oxida-
tion is manifested. Unlike other techniques, a pulsed electron beam irradi-
ation helps synthesize nanosized titanium dioxide particles with crystal
structure at low temperature. A considerable reduction in the temperature
threshold of particle crystal structure formation has also been achieved for
composite nanosized particles of (TiO2)x(SiO2)1-x oxides.

171
 Conclusion
In conclusion we would like emphysise some other promising areas
of practical application of plasma chemistry. The studies of liquid-phase
decomposition of hydrocarbons during heating showed that the process is
implemented as a chain (thermal cracking). At low temperatures, in the
absence of radiation the thermal initiation of the reaction does not occur;
as a result of the radiation exposure the active centres are formed, which
are free radicals that can start a chain reaction of hydrocarbon cracking.
The interaction of radicals containing a small number of carbon atoms
with the initial hydrocarbon is a distinct chain process, it is initiated by
the radicals formed though the radiation. Such a chain reaction of radia-
tion-thermal cracking will take place at a temperature of 150-200 degrees
below the temperature of a normal thermal process, but with the same
specific rate, because the radiation facilitates the most energy-intensive
stage - the thermal initiation of reaction. In addition, with a sufficient
length of chain the total energy consumption for the chemical process can
be reduced. The main source of energy in this case is the thermal energy
of the original material or the exothermic energy of elementary chemical
chain reactions (e.g., oxidation or polymerization). This can significantly
reduce the energy consumed by the chemical process. Conducting a
chemical process at a low temperature allows one to synthesize com-
pounds which are unstable at higher temperatures or selectivity of synthe-
sis of which at high temperatures is low. The above features from the
chain of chemical processes which occur in plasma are particularly rele-
vant for the petrochemical industry.
But at the interaction of ionizing radiation with matter in the liquid
phase the major part of the secondary electrons produced though ioniza-
tion of the medium, have a small path, due to its relatively low energy.
Therefore the electrons produce ionization and excitation in a local area,
close to the place of their formation. This effect observed in the radiolysis
of liquid media is called "Franck-Rabinowitch effect". This effect is to
maintain a high concentration of radicals and long-lived excited products
formed in the spurs, after the completion of the primary excitation act.
The influence of this effect is particularly noticeable in the general output
of the radiolysis products, which are less formed in the liquid phase than
in the gas phase. This leads to the fact that at temperatures below 600 K,
the radiation-chemical yield does not exceed a few molecules per 100 eV.
Pulse radiolysis of liquid-phase hydrocarbons at low temperature un-
der the action of an electron beam with a high current density (200
A/cm2), is of a great interest. At the same time the conditions favorable to
multiple collisions of particles with excess energy stored in the internal

172
degrees of freedom are formed. These conditions are similar to nonequi-
librium excitation of gas-phase environment and contribute to a chain re-
action. A requirement for creating of such conditions of radiolysis is a
high absorbed dose of the radiation.The tracks are formed during the en-
tire duration of the pulse and during the radiation the relaxation of excited
molecules and radicals recombination should be negligible. The lifetime
of active radicals formed in the track during the radiolysis does not ex-
ceed 10-7-10-8 sec. Therefore, the duration of radiation exposure should
not exceed 100 ns. Radiolysis of liquid hydrocarbons in these conditions
has not been studied and is of scientific and practical interest for studing
the mechanism of radiation liquid phase cracking of hydrocarbons at low
temperatures.
Another promising area of application of the gas discharge and
pulsed electron beams is the initiation of soot formation in non-
equilibrium conditions. It is known that the conversion of carbon-
hydrogen fuel during the combustion passes through the stage of for-
mation of polycyclic aromatic hydrocarbons. They are the centers of for-
mation of the soot particles (nuclei). For the soot formation the radical-
precursors requires; therefore the process under equilibrium conditions
does not occur at low temperatures. In addition, the soot precursors are
pyrolyzed and oxidized at high temperatures, so that the process of soot
formation under equilibrium conditions (1000 - 2000 K) completes the
synthesis of pyrolytic carbon. The formation of radicals thought the action
of a pulsed electron beam on a mixture of hydrocarbon gases and oxygen
can initiate the formation of polycyclic aromatic hydrocarbons at a tem-
perature below the equilibrium sooting. Under these conditions, the pro-
cess of soot formation can be stopped at the stage of the synthesis of aro-
matic hydrocarbons, if the duration of external exposure will not exceed
the duration of this stage. This may allow the development of a new tech-
nology of the synthetic liquid fuels or the complex hydrocarbons with
isomeric structure from a natural gas.

173
 Questions
1. Please give the definitions to the following concepts in
chemical kinetics: rate of chemical reaction, law of mass action, reaction
rate constant.
2. Please explain the following: Arrhenius equation, kinetic
equation of the reaction, kinetic scheme of the chemical process
3. Please give the definitions to the following terms: plasma,
Debye radius, degree of ionization, electron energy distribution function.
4. Explain the following: elementary processes in plasmas:
classification, rate elementary processes, collision cross section.
5. Direct and stepwise ionization in the plasma.
6. The relaxation of excited particles in the plasma. The practi-
cal application of the relaxation processes.
7. Treanor mechanism.
8. Mechanisms of electrical discharge. Tawsend discharge
9. Paschen curve. Limitations of the Paschen law.
10. Types of gas discharge
11. The streamer and spark forms of gas discharges.
12. Gas breakdown in an alternating voltage. Microwave
discharge.
13. Typical designs of gas-discharge sources of low-temperature
plasma.
14. The specific characteristics of plasma-chemical reactions.
Quasi-equilibrium plasma-chemical processes.
15. Nonequilibrium plasma-chemical processes. Arrhenius equa-
tion for nonequilibrium plasma chemical reactions.
16. Principles of organization of plasma chemical processes.
17. The main types of reactions in plasma chemistry.
18. The methods of nonequilibrium excitation of the molecules.
19. Chain gas-phase processes. The kinetic scheme of chain pro-
cess.
20. Classification of chain processes. The induction period and
limits of ignition.
21. Chain oxidation of hydrogen.
22. Chain-chemical processes under external action. Methods of
initiation of chain processes.
23. Conversion of methane in low-temperature plasma. Classifi-
cation of conversion methods. Conversion products.
24. The kinetic scheme and the main plasma chemical reaction of
the methane pyrolysis

174
 25. The steam conversion of methane in low-temperature plas-
ma. The kinetic scheme and conversion products.
26. Combined methods of plasma-chemical conversion of me-
thane
27. Chain plasma-chemical conversion of methane. The kinetic
scheme and the energy balance of the chain process.
28. Plasma-chemical synthesis of nanosized particles. Classifica-
tion, advantages and disadvantages of synthesis methods.
29. Non-equilibrium plasma-chemical synthesis of nanosized
metal oxides.
30. Non-equilibrium plasma-chemical synthesis of nanosized
composite oxides. The application of titanium nanosized composite ox-
ides.
31. Plasma-chemical methods for the formation of carbon
nanostructures. Allotropic forms of carbon.
32. Beam-plasma technology of hardening and surface modifica-
tion of metal products
33. Applications of low-temperature plasma in chemical produc-
tion
34. Plasma-chemical processing of medical polymers.
35. Plasma-chemical methods of waste treatment.

175
 References

1. Pushkarev A., Novoselov Yu., Remnev G. Chain processes in low

temperature plasma. – Novosibirsk.: Nauka, 2006. – 226 p.
2. Pushkarev A.I., Remnev G.E. Application of Pulsed Electron
Beams in Plasma Chemistry // Известия вузов. Физика. – 2006. – T. 49.
– № 11 (Приложение). – C. 462–466.
3. Goldston R.,Rutherford P. Introduction to Plasma Physics. – Phil-
adelphia.: IOP Rublishing Ltd, 1995. – 479 p.
4. Fitzpatrick R. Lectures on Plasma Physics. – Cambridge UK:
Cambridge University Press, 1994. – 247 p.
5. Cecchi J. L. Handbook of Plasma Processing Technology Park
Ridge, NJ: Noyes Publications, 1990. – 149 p.
6. Thornton J. A, Penfold A. S. Thin Film Processes. – New York:
Academic Press, 1978. – 75 p.
7. Chen F. F. Introduction to Plasma Physics and Controlled Fusion,
2nd ed. – New York: Plenum Press, 1984. – 126 p.
8. Venugopalan M. Reactions Under Plasma Conditions. – New
York: Wiley-Interscience, 1971. – Vol. 1. – 608 p.
9. Bell A. T. Techniques and Applications of Plasma Chemistry. –
New York: John Wiley & Sons, 1974. – 403 p.
10. Cordey J. G., Goldston R. J., Parker R. R. Progress Toward a
Tokamak Fusion Reactor // Phys. Today. – 1992 – Vol. 45. – P. 22–28.
11. Raizer Yu.P. Gas discharge physics. – New York: Springer,
1997. – 463 p.
12. Cooray G. V. The Lightning Flash. – London: The Institution of
Engineering and Technology ,2003. – 574 p.
13.Weissler G.L. Introduction to Plasma Physics // Encyclopedia of
Physics. – 1956 – Vol. 16. – P. 445–454.
14. Richadson O.W. The emission of electricity from hot bodies. –
London: LongmansGreen, 1921. – 314 p.
15. Schottky W.Z. Schottky Effect // J. Phys. – 1923. – № 14. –
P. 63–74.
16. Fowler R., Nordheim L. Field emission // Proc. R. Soc. – 1928. –
Vol. 119. – P. 173–181.
17. Loeb L.B. Basic processes of gaseous electronics. – Los angeles:
University of California Press, 1955. – 1029 p.
18.Raether H. Electrical Coronas // J. Angew. Phys. – 1955. – №7. –
P.50–57.
19. Loeb L.B., Meek J.M. The mechanism of the electric spark.
Stanford: StanfordUniversity Press, 1940. – 211 p.

176
 20. Loeb L.B. Electrical coronas: Their basic physical mechanism. –
Los angeles: Universityof California Press, 1965. – 417 p.
21. Raether H. Avalanches and streamers // J. Phys. – 1939. –
Vol. 112. – P.464–471.
22. Meek J.M. Formation of a negative streamer // J. Phys. Rev. –
1940. – Vol. 57. – P. 722–762.
23. Dawson G.A., Winn W.P. A spherical space charge // J. Phys. –
1965. – Vol. 183. – P. 159 –163.
24. Gallimberti I.J., Bacchiege G.L., Gazzani A., Bernadi M., and
Bondiou A. // Proceedings of the 1994 international aerospace and
groundconference on ‘ Lightning and static electricity’. – 1994. – P. 23–
29.
25. Gallimberti I. The mechanism of the long spark formation // J.
Phys. – 1979. – № 40. – Р. 193–250.
26. Towsend J.S. Electricity in gases. – Oxford: Oxford Press, 1914.
– 493 p.
27. Nasser E. Fundamentals of gaseous ionization and plasma elec-
tronics. – New York:Wiley-Interscience, 1971. – 386 p.
28. Dakin T., Luxa G., Oppermann G., Vigreux J., Wind G. Paschen
curve for air // Electra – 1974. – Vol. 32. – P. 61–69.
29. House J. E. Principles of chemical kinetics // James E. House. –
2007. – P. 963–975.
30. Frinman A. Plasma Chemistry. – New York: Cambridge Univer-
sity Press, 2008. – 725 p.
31. Polak L. Elementary chemical processes and kinetics in a
nonequilibrium and quasi-equilibrium plasma // J. Applied Chem. – 1970.
–
P. 307–342.
32. Полак Л.С. Кинетика неравновесной плазмы// Первый все-
союзный симпозиум по плазмохимии. – 1971. C. 1–15.
33. Полак Л.С. Применение вычислтельной математики в физи-
ческой и химической кинетике. – М.: Наука, 1966. – 421 c.
34. Von Engel A. Ionized Gases. – Clarendon: Oxford, 1955. –
421 p.
35. Spitzer L. Physics of Fully Ionized Gases. – New York.: Inter-
science: New York, 1962. – 358 p.
36. Shih-J-Pay. Magneto Gas Dynamics and Plasma Dynamics. –
Springer: Vienna, 1962. – 412 p.
37. Wall F. T., Hitler L. A., Mazur J. Chemical reactions under low-
temperature plasma conditions // J. Chem. Phys. – 1961. – Vol. 35. –

177
P. 1284–1295.
38. Blais N. C., Bunker D. L. Chemical reactions // J. Che. Phys. –
1963. – Vol. 39. – P. 315–319.
39. Keck J. C. Chemical reactions // J. Chem. Phys. – 1958. – Vol.
29. – P. 410–426.
40. Schmeltekopf A. L., Ferguson E. E. Fesenfeld F. C. Chemical
reactions under low-temperature plasma conditions // J. Chem. Phys. –
1968. – Vol. 48. – P. 2966–2974.
41. Месяц Г.А. Импульсная энергетика и электроника. – М.:
Наука, 2004. – 704 с.
42. Рыжов В.В., Ястремский А.Г. Распределение энергии элек-
тронного пучка в плазме азота // Физика плазмы. – 1978. – Т. 4,
вып.6. – С. 1262–1266.
43. Норман Г.Э., Полак Л.С., Сопин П.И., Сорокин Г.А.
Сильноточные релятивистские электронные пучки в плазмохимии //
Синтез соединений в плазме, содержащей углеводороды. – М.:
ИНХС АН СССР, 1985. – 166 с.
44. Налбандян А.Б., Воеводский В.В. Механизм окисления и
горения водорода. – М.: Изд-во АН СССР, 1949. – 179 с.
45. Старик А.М., Титова Н.С. О кинетических механизмах
инициирования горения водородно-кислородных смесей при
возбуждении электронных степеней свободы молекулярного
кислорода лазерным излучением // Журнал технической физики. –
2003. – Т. 73. – № 3. – С. 59–68.
46. Пушкарев А.И., Ремнев Г.Е. Инициирование окисления
водорода импульсным электронным пучком // Физика горения и
взрыва. – 2005. – № 3. – С. 46–51.
47. Пушкарев А.И., Ремнев Г.Е. Колебательный характер
процесса окисления водорода при инициировании импульсным
электронным пучком // Физика горения и взрыва. – 2005. – № 4. –
C. 18–21.
48. Ремнев Г.Е., Фурман Э.Г., Пушкарев А.И. и др. Импульсный
сильноточный ускоритель с согласующим трансформатором //
Приборы и техника эксперимента. – 2004. – № 3. – С. 130–134.
49. Пушкарев А.И., Пушкарев М.А., Ремнев Г.Е. Исследование
звуковых волн, генерируемых при поглощении импульсного
электронного пучка в газе // Акустический журнал. – 2002. – Т. 48,
№2. – С. 260–265.

178
 50. Pushkarev A., Isakova J., Kholodnaya G., Sazonov R. Sound
Waves Generated Due to the Absorption of a Pulsed Electron Beam
//Advances in Sound localization. – Vienna: INTECH, 2011. – 672 p.
51. Селезенев А.А., Алейников А.Ю., Ярошенко В.В. Влияние
радиолиза на смещение пределов воспламенения водород-
кислородной газовой смеси. // Химическая физика. – 1999. – Т. 18. –
№ 5. – С. 65–71.
52. Комар А.П., Круглов С.П., Лопатин И.В. Измерение полной
энергии пучков тормозного излучения от электронных ускорителей.
– Л.: Наука. – 1972. –172 с.
53. Радиолиз углеводородов. Некоторые физико-химические
проблемы / Под ред. А.В. Топчиева, Л.С. Полака.– М.: Изд-во АН
СССР. – 1962. – 208 с.
54. Сараева В.В. Радиолиз углеводородов в жидкой фазе.
Современное состояние вопроса. – М.: Изд-во Моск. ун-та. – 1986. –
256 с.
55. Химия высоких энергий / Под ред. Л.Т. Бугаенко, М.Г.
Кузьмин, Л.С. Полак. – М.: Химия, 1988. – 368 с.
56. Бабарицкий А.И., Деминский М.А., Демкин С.А., Животов
В.К. Эффект плазменного катализа при разложении метана // Химия
высоких энергий. – 1999. – Т. 33, №1. – С. 49–56.
57. Бабарицкий А.И., Баранов И.Б., Дёмкин С. А. и др.
Плазменный катализ процессов конверсии углеводородов // Химия
высоких энергий. – 1999. – Т. 33. – № 6. – С. 458–462.
58. Штерн В.Я. Механизм окисления углеводородов в газовой
фазе. – М.: Изд-во АН СССР, 1960. – 496 с.
59. Арутюнов B.C., Крылов О.В. Окислительные превращения
метана. – М.: Наука, 1998. – 361 с.
60. Кармилова Л.В., Ениколопян Н.С., Налбандян А.Б. К
вопросу о вырожденном разветвлении. Роль формальдегида при
окислении метана // Журнал физической химии. – 1957. – Т. 31. –
С. 851–864.
61. Михайлов Б.М., Куимова М.Е., Богданов В.С. Действие
ионизирующих излучений на неорганические и органические
системы. – М.: Изд. АН СССР, 1958. – 223 с.
62. Русанов В.Д., Бабарицкий А.И., Герасимов Е.Н. и др.
Стимулирование процесса парциального окисления метана в
микроволновом разряде // Доклады Академии наук. – 2003. – Т. 389.
– № 3. – С. 324–327.

179
 63. Русанов В.Д., Бабарицкий А.И., Баранов И.Е. и др.
Неравновесное воздействие плазмы микроволнового разряда
атмосферного давления на процесс конверсии метана и керосина в
синтез-газ. // Доклады Академии наук. – 2004. – Т. 395. – № 5. – С.
637–640.
64. Денисов Г.В., Новоселов Ю.Н., Суслов А.И., Устер А.М.
Продукты окисления сероуглерода в ионизованном воздухе // Журн.
техн. физики. – 2001. – Т. 71. – № 1. – С. 136–138.
65. Новоселов Ю.Н., Суслов А.И., Кузнецов Д.Л. Воздействие
импульсных пучков электронов на примесь сероуглерода в воздухе //
Журн. техн. физики. – 2003. – Т. 73. – № 6. – С. 123–129.
66. Новоселов Ю.Н., Рыжов В.В., Суслов А.И. Эффект
электрического поля при конверсии сероуглерода в ионизованном
воздухе. // Письма в ЖТФ. – 2002. – Т. 28. – № 6. – С. 35–41.
67. Remnev G.E. , Pushkarev A.I., Khim. Plasma-chemical synthesis
of SiO2 nanopowders // High Energy Chem. – 2004. – Vol. 38. – №. 5. –
P. 348–354.
68. Remnev G.E., Pushkarev A.I. Plasma-chemical synthesis of TiO2
nanopowders // IEEJ Trans. Fundam. Mater. – 2004. – Vol. 124. – № 6. –
P. 483–489.
69. Bin Xia, Li. W., Zhang B. et al. Low temperature vapor-phase
preparation of TiO2 nanopowders // Journal of Materials Science. – 1999.
– Vol. 34. – P. 3505 – 3511.
70. Jang H. D., Kim S.-K. and Kim S.-J. Effect of particle size and
phase composition of titanium dioxide nanoparticles on the photocatalytic
properties // J. of Nanoparticle Res. – 2001. – Vol.3. – № 2 – 3. P. 141–
146.
71. Ahonen P.P., Moisala A., et al. Gas-phase crystallization of tita-
nium dioxide nanoparticles // J. of Nanoparticle Res. – 2002. – Vol . 4. –
№ 1– 2. – P. 43–46.
72. Ayllon J. A., Figueras A., Garelik S., et al. Preparation of TiO2
powder using titanium tetraisopropoxide decomposition in a plasma en-
hanced chemical vapor deposition (PECVD) reactor // J. of Mater. Sci.
Lett. – 1999. – Vol. 16. – № 18. – P. 13– 9.
73. Li W., Ni C., Lin H. et al. Size dependence of thermal stability of
TiO2 nanoparticles // J. of Appl. Phys. – 2004. – Vol. 96. – №. 11. –
P. 6663–6667.
74. Harano A., Shimada K. et al. Crystal phases of TiO2 ultrafine
particles prepared by laser ablation of solid rods // J. of Nanoparticle Res.
– 2002. – Vol. 4. – № 3. – P. 215–217.

180
 75. Seto T., Kawakami Y. et al. Evaluation of morphology and size
distribution of silicon and titanium oxide nanoparticles generated by laser
ablation // J. of Nanoparticle Res. – 2001. – Vol. 3. – № 2–3. – P. 185–
189.
76. Oh S.-M., Park D.-W. and Ishigaki T. Plasma Synthesis of
Spherical Titanium Dioxide from Titanium Nitride // Proc. of 16th Intern.
Symp. on Plasma Chemistry. Taormina. – 2003. – P. 45–52.
77. Kotov Yu A. Electric explosion of wires as a method for prepara-
tion of nanopowders // Journal of Nanoparticle Research. – 2003. – № 5.
– P. 539–550.
78. Назаренко О.Б. Электровзрывные порошки. Получение,
свойства, применение / Под ред. А.П. Ильина. – Томск: Изд-во
Томского университета, 2005. – 148 с.
79. Wallidge G. W., Anderson R., Mountjoy G., Pickup D. M., Gun-
awidjaja P., Newport R. J., Smith M. E. Advanced physical characteriza-
tion of the structural evolution of amorphous (TiO2)x(SiO2)1-x sol-gel ma-
terials // Journal of materials science. – 2004. – Vol. 39. – P. 6743–6755.
80. Ingo G. M., Riccucci C., Bultrini G., Dire S. and Chiozzini G.
Thermal and microchemical characterization of sol-gel SiO2, TiO2 and
xSiO2-(1-x)TiO2 ceramic materials // Journal of Thermal Analysis and
Calorimetry. – 2001. – Vol. 66. – P. 37–46.
81. Machida M., Norimoto K., Watanabe T., Hashimoto K., Fuji-
shima A. (1999) The effect of SiO2 addition in super-hydrophilic property
of TiO2 photocatalyst // Journal of materials science. – 1999. – Vol. 4. –
P. 2569–2574.
82. Young-Geun Kwon, Se-Young Choi, Eul-Son Kang, Seung-Su
Baek Ambient-dried silica aerogel doped with TiO2 powder for thermal
insulation // Journal of Materials Science. – 2000. – Vol. 35. – № 24. –
P. 6075–6079.
83. Takahiro G., Takayuki K., Yoshimoto A. Crystallization Behav-
ior of SiO2-TiO2 ceramics derived from titanosiloxanes on pyrolysis //
Journal of Sol-Gel Science and Technology. – 1998. – Vol. 13. – №1–3. –
P. 975–979.
84. Remnev G.E., Isakov I.F., Pushkarev A.I. A pulsed ion beam ac-
celerator for material modification // Surf. Coat. Technol. – 1999. – Vol.
114. – P. 206–217.
85. Nakamoto K. Infrared and Raman Spectra of Inorganic and Co-
ordination Compounds. – New York: Wiley, 1986. – 326 p.
86. Kuptsov A.Kh., Zhizhin, G.N. Fourier-Transform Raman and In-
frared Absorption Spectra of Polymers. – M.: Fizmatlit, 2001 – 126 p.

181
 87. Ahn W.S., Kang K.K., Kim K.Y. Synthesis of TS-1 by micro-
wave heating of template-impregnated SiO2–TiO2 xerogels // Catalysis
Letters. – 2001. – Vol. 72. – № 3–4. – P. 229–232.
88. Fincke J. R., Anderson R. P., Hyde T. et al. Plasma thermal con-
version of methane to acetylene // Plasma Chemistry and Plasma Pro-
cessing. – 2002. – Vol. 22. – № 1. – P. 107–138.
89. Babaritskii A.I., Deminskii M.A., Demkin S.A., Zhivotov, V.K.,
A low temperature thermal initiation of cracking // High Energy Chem. –
1999. – Vol. 33. – № 1. – Р. 45–49.
90. Zhivotov V.K., Potapkin B.V., Rusanov V.D. Encyclopedia of
Low-Temperature Plasma. – M.: Fizmatlit. – 2005. – 41 p.
91. Czernichowski A., Glid arc assisted production of H2 from CH4,
CO or H2S // Proc. Int. Symp. Hydrogen Power, Theoretical and Engi-
neering Solutions. – 1999. – P. 98–104.
92. Li X.S., Zhu A.M., Wang K.J., Xu Y., Song Z.M. Methane con-
version to C2 hydrocarbons and hydrogen in atmospheric non-thermal
plasmas generated by different electric discharge techniques // J. Catal. –
2004. – Vol. 98. – P. 617–625.
93. Li X.S., Shi C., Xu Y., Wang K.J., Zhu, A.M. A process for a
high yield of aromatics from the oxygen-free conversion of methane:
combining plasma with Ni/HZSM-5 catalysts // Green Chem. – 2007. –
Vol. 9. – P. 647–689.
94. Pushkarev A.I., Novoselov Yu.N., Remnev G.E., Tsepnye
protsessy v nizkotemperaturnoi plazme. – Novosibirsk: Nauka, 2006. –
123 p.
95. Pushkarev A.I, Remnev G.E., and Ponomarev D.V., Fundamen-
tal’nye problemy prilozhenii fiziki nizkotemperaturnoi plazmy: Materialy
lektsii Vserossiiskogo simpoziuma molodykh uchenykh, studentov i as-
pirantov // Proc. All-Russia Symp. of Young Scientists, Students, and As-
pirants on Fundamental Problems of Application of Low-Temperature
Plasma Physics. – 2005. – Vol. 1. – P. 26–29.
96. Shtern V.Ya., Mekhanizm okisleniya uglevodorodov v gazovoi
faze. – Moscow: AN SSSR, 1960. – 123 p.
97. Arutyunov V.S. and Krylov O.V. Okislitel’nye prevrashcheniya
metana. – M: Nauka, 1998. – 136 p.
98. Rusanov V.D., Babaritskii A.I., Gerasimov E.N., Deminskii
M.A., Demkin S.A., Zhivotov V.K, Moskovskii A.S., Potapkin B.V.,
Smirnov R.V. and Strelkova M.I. Dokl. Akad. Nauk. – 2003. – Vol. 389.
– № 3. – Р. 324–327.

182
 99. Lesueur H., Czernichowski A., Chapelle A. Electrically assisted
partial oxidation of methane //. J. Hydrogen Energy. – 1994. – Vol. 19. –
№ 2. – P. 139–142.
100. Lee D.H., Kim K.T., Song Y.-H and Cha M.S. Characteristic of
Methane Processing by Rotating Arc Reactor Proc. // 18th Int. Symp. on
Plasma Chemistry. – 2007. – P. 153–159.
101. Heo J., Choi J.-W., Lee H., Sekiguchi H. and Song H.K. Syn-
thesis Gas Production from Methane and Air Mixture with a Vortex Glid-
ing Arc Reactor Proc. // 18th Int. Symp. on Plasma Chemistry. – 2007. –
P. 554–559.
102. Remnev G.E., Pushkarev A.I., Ezhov V.V., Partial methane and
nitrogen oxidation initiated by pulsed electron beam Proc // 13th Int.
Symp. on High Current Electronics. – 2004. – P. 447–451.
103. Indarto A., Choi J.W., Lee H., Song H.K. Plasma methane oxi-
dation for methanol synthesis with Cu-Zn-Al catalyst Proc. // 18th Int.
Symp. on Plasma Chemistry. – 2007. – P. 194–198.
104. Mutaf-Yardimci O., Saveliev A., Fridman A., Kennedy L. Em-
ploying plasma as catalyst in hydrogen prodaction Int. // J. Hydrogen En-
ergy. – 1998. – Vol. 23. – № 12. – Р. 1109–1114.
105. Lee H., Choi J.-W., Song H.K., Lee C.-H. The effect of the
electric pulse polarity on CO2 reforming of CH4 using dielectric-barrier
discharges Proc. // 4th Int. Symp. on Pulsed Power and Plasma Applica-
tions. – 2003. – P. 146–157.
106. Song H., Lee H., Choi J.,Na B. Effect of electrical pulse forms
on the CO2 reforming of methane using atmospheric dielectric barrier dis-
charge // Plasma Chem. Plasma Process. – 2004. – Vol. 24. – №. 1. – P.
57–61.
107. Rusu I. and Cormier J.-M. On a possible mechanism of the me-
thane steam reforming in a gliding arc reactor // Chem. Eng. J. – 2003. –
Vol. 91. – № 1. – Р. 23–29.
108. Cormier J., Rusu I. Syngas production via methane steam re-
forming with oxygen: plasma reactors versus chemical reactors // J. Phys.
D: Appl. Phys. – 2001. – Vol. 34. – P. 2798–2801.
109. Azizov R.I., Babaritskii A.I., Demkin S.A. In Gazokhimiya v
XXI veke. Problemy i perspektivy // Gas Chemistry in XXI Century:
Problems and Prospects. – 2003. – P. 45–53.
110. Sekine Y., Urasaki K., Kado S., Asai S., Matsukata M., Kikuchi
Steam reforming of hydrocarbons and alcohols using non-equilibrium
pulsed discharge E // Proc. 16th Int. Symp. on Plasma Chemistry,
Taormina. – 2003. – P. 213–216.

183
 111. Zhdanok S.A., Krauklis A.V., Bouyakov I.F. Studying methane
conversion in atmospheric pressure high-voltage discharge with different
oxidizers in presence of catalysts // Proc. IV Intern. School-Seminar
Modern Problems of Combustion and its Application. – 2001. – P. 66–72.
112. Chavadej S., Rueangjitt N., Sreethawong T. Reforming of CO-
containing natural gas with partial oxidation using an AC gliding arc sys-
tem // Proc. 18th Int. Symp. on Plasma Chemistry. – 2007. – P. 552–561.
113. Bromberg L., Cohn D., Rabinovich A., O’Brien C. and
Hochgreb S. Plasma catalytic reforming of methane Energy Fuels // Proc.
18th Int. Symp. on Plasma Chemistry – 1998. – Vol. 1. – №. 1. – Р. 11–
16.
114. Bromberg L., Cohn D., Rabinovich A., Alexeev N. Partial me-
thane and nitrogen oxidation initiated by pulsed electron beam // Int. J.
Hydrogen Energy. – 1999. – Vol. 24. – P. 1131–1136.

184