10 Physics and chemistry of non-thermal plasma at

atmospheric pressure relevant to surface treatment

10.1 INTRODUCTION
The atmospheric pressure non-thermal plasma interacting with a surface is an extremely
intricate object for detailed modelling. Complexity of plasma-surface interaction at conditions
typical for polymer treatment consists in a wide variety of plasma and polymer surface parameters.
In particular, the sort of a plasma-forming gas varies according to a purpose of polymer surface
modification. Both, surface atomic profile and chemical composition of a monomer are different for
different polymers. Nowadays we have rudimentary knowledge about processes taking place on a
boundary between plasma and polymer bulk.
Generally, the plasma surface processing can be broken into following stages: 1) plasma
production; 2) transport of plasma activated gas to the polymer surface; 3) interaction of plasma
products with polymer surface. Respectively, a theoretical model of a whole technological process
can be reduced to a combination of specific models for each of the stages.
Recently, the plasma physics scientific community has paid much attention to the development
of atmospheric-pressure non-thermal plasma sources. Progress in this field is very rapid. A number
of overviews about non-thermal plasma production at atmospheric pressures and some applications
appeared recently [1-7].
Approaches to develop atmospheric pressures discharges operated at strongly non-equilibrium
conditions vary, and the spectrum of discharge devices already existing is rather broad. Starting
from the Dielectric Barrier Discharge (DBD, also called silent discharge) and corona discharges
known for more than hundred years, nowadays a number of different discharge techniques
succeeded in the development of devices, which are sources of non-thermal plasma at atmospheric
pressure. The interest in this topic is dictated by a potential economic benefit from numerous non-
thermal plasma technologies. It is anticipated that many of these (for example, protective coating
deposition, toxic and harmful gas decomposition, electromagnetic wave shielding, polymer and
fabric surface modifications, bio-medical applications, and so on) may gain a strong impetus and
rapid progress. Due to the wide variety of potential applications, requirements for the relevant
plasma parameters are also rather diverse.
Along with the progress in the discharge techniques the theory of gas discharge formation and
of plasma properties was developed. The specific feature of non-thermal plasma produced by the
DBD, corona, and glow discharge is high sensitivity of its characteristics to the sort of gas. In
contrast to thermal plasma, which can be completely characterized by a few parameters (gas
pressure, temperature, dissipated power density), the number of parameters required for a full
description of plasma properties is rather large. Typically, electron energy distribution is strongly
non-equilibrium, while ions have nearly the same average energy as atoms and molecules. Exactly,
free electrons effectively gain energy from the electric field, which further is dissipated in numerous
reactive collisions (excitation of molecular vibrations, dissociation, ionization and excitation of
electronic states of atoms and molecules). For correct description of non-thermal plasma it is
mandatory to treat properly electron kinetics.
Chemical kinetics of plasma involves many species produced in plasma including radicals and
excited atoms and molecules. Above-equilibrium concentrations of such species change completely
reactivity of gas in comparison with thermal kinetics and lead to appearance of a number of
transient compounds capable to react at nearly room temperature. Theoretical description of these
processes is possible provided the extensive knowledge about numerous rate coefficients is
accumulated. Hence, electron kinetics and plasma chemistry have a fundamental value for theory.

10.2 DISCHARGE TECHNIQUES

Plasma processing has been widely used in various engineering applications.
Dry plasma technology offers an environmentally friendly alternative to the conventional wet
chemical methods to improve the surface properties. Most of industrially used plasma technologies
are based on use of so-called non-thermal reactive plasmas consisting of gaseous mixture of
charged particles (electrons, ions) and neutral activated species including gas molecules, free
radicals, metastables and UV-photons. Modification of the surface properties of heat-
sensitive materials like fabrics and polymer films, for example, has received widespread
attention for their practical importance. It is important that plasma modification
influences only the outermost molecular surface layers without changing the bulk properties.
Another advantage of plasma method is that resulting treatment effect depends on easily adjustable
discharge parameters like electric power, treatment time, gas composition and gas pressure. This
circumstance allows the controlling the reaction process and predefines non-thermal plasma
treatment as hugely versatile and multifunctional process that can be applied to different
technologies.
In general non-thermal reactive plasma can be created with different kinds of gas discharges
under low and atmospheric pressure. At low pressure, it is more or less easy to produce
homogeneous plasma with a volume of up to 10-30 liters. However generation of non-thermal
reactive plasma at low-pressure requires using of large-size and massive metallic vacuum chambers
and expensive vacuum equipment. Besides, plasma processing at low pressure is not continuous – it
is necessary to load and unload the samples in batches in and out of the vacuum chamber.
Our main concern in this chapter is physics and chemistry of atmospheric pressure non-thermal
plasma. This is because high-pressure non-thermal plasma allows us to eliminate expensive vacuum
equipment and to perform continuous plasma processing that elucidate considerably its
incorporation into industrial production lines. As mentioned above many kinds of gas
discharges at atmospheric pressure have become more deeply involved in non-
thermal plasma processing. The most of atmospheric pressure discharges
generate non-equilibrium plasma due to the formation in gas gap of large
numbers of non-stationary micro-discharges generating numerous current
filaments (so-called streamers) of nano- or micro-second duration. As a result,
highly reactive non-equilibrium conditions in cold gas can be provided at
atmospheric pressure.
At atmospheric pressure, the life-time of most active species generated by plasma is very short.
It is therefore necessary either to produce active species immediately at the surface or to provide a
fast transfer of active species from the plasma region to the surface. Both ideas are used in all
known approaches for the generation of non-thermal plasma at atmospheric pressure. The first idea
is used in the case of so-called “corona treaters” based on dielectric barrier discharge at typical
frequencies of 10-40 kHz. However, in connection with surface treatment, there is serious problem
inherent to the barrier discharge, in which the electrical current passes through the surface to be
treated. In some cases, this circumstance may lead to the destruction of the fabric.
The second idea is realized in so-called “remote surface treatment”, in which non-thermal
plasma in a form of “plasma torches”, “plasma pencils” or “plasma jet” enriched with active species
is blown out of the plasma source and directed onto the treated surface. Several novel types of
plasma sources providing remote treatment at atmospheric pressure are described in [8]. Remote
treatment works well for many applications, but it does not provide a 100% use of the active species
that are produced in the plasma source, due to inevitable losses during their transport by gas flow.
However, in contrast to corona treaters, the remote plasma sources can treat the surface of materials
of any thickness.
Two non-thermal plasma sources working at atmospheric pressure which were used in our
experiments for remote surface treatment are described below. Side view of these sources marked
as No1 and No2 are presented below in Figures 10.1 and 10.2.
Plasma source No1 (see details in [9]) creates plasma jet in ambient air blowing out with
average velocity about of 30-50 m/s. This atmospheric pressure plasma source can work in two
modes - diffusive and filamentary or streamer mode. Namely streamer mode of the discharge in
airflow is shown in Figure 10.1a. The effect of blowing off of streamers down the gas flow is seen.
The radius of streamers was evaluated from the experimental observations, and was taken as 0.5
mm. Typical waveforms of discharge voltage and current are shown in Figure 10.1b. Strong
fluctuations in the amplitudes of current pulses are recorded. Typical repetition frequency of the
current pulses is about 20 kHz. In our experiments it was revealed that homogeneous discharge in
air is about 6 times less effective in polymer surface treatment than the discharge with numerous
non-stationary streamers at the same average energy loading. For this reason, our main concern with
discharge in airflow was a reproducible realization of its non-stationary streamer mode.

a
b

Figure 10.1a–b Atmospheric pressure plasma source No1 used for remote surface treatment outside
of discharge zone. a Side view of plasma source No1 in the streamer mode; ambient air, gas flow
velocity is 40m/s; average electric power is P=35-70 W; maximum length of streamer plasma jet in
air is about 3 cm. b The waveforms of discharge voltage and current of plasma source No1 in the
non-stationary streamer mode. Scales: voltage is 5 kV per unit, current is 200 mA per unit, time is
50 s per unit.

b
Figure 10.2a–b Information on steady-state atmospheric pressure plasma source No2. a Side view
of diffusive plasma jet generated by source No2 in pure (99.999%) nitrogen. b Longitudinal
distribution of gas temperature along plasma jet (x=0 is the nozzle of plasma source. Average
electric power is P=70 W; N2 flow velocity is 30m/s; diameter of plasma jet at outlet is about 8mm;
maximum length of diffusive plasma jet in pure nitrogen can reach 10-15 cm.

Plasma source No2 creates plasma jet in pure (99.999% of purity) nitrogen blowing out with
average velocity about of 30-50 m/s. In the experiments devoted to remote treatment this steady-
state atmospheric pressure plasma source was used only in diffusive mode. The length of plasma jet
depends on sort of ambient gas, and in the case of nitrogen this length can reach 15 cm.
Long-length plasma jet in pure nitrogen is an effective instrument for surface treatment. It
should be noted, however, our measurements revealed that composition of active species in nitrogen
plasma jet includes not only nitrogen excited molecules but other species as well generated in
plasma due to existence of trace admixtures like oxygen and hydrocarbons.

10.3 DISCHARGE MODELLING

10.3.1 Introduction
Complexity of plasma description, in particular at high pressures is associated with two points:
(1) a vast amount of reactions taking place in plasma, which are very specific for given gas
composition; (2) non-uniformity of plasma in majority of discharge types. Evidently, taking into
account both of these points presents a formidable problem. Traditionally, an approach to modeling
of realistic plasma devices consists of two phases. At a first step, plasma is considered to be
uniform, and all transport processes are ignored. At this step, a major attention is paid to correct
description of electron kinetics and of initiated by electrons reactions of excited particles and
chemical radicals. Because of a strong difference in masses of electrons and neutral
molecules/atoms, even in a low electric field electrons are strongly heated, and their kinetics is
described by dynamics of an electron energy distribution function (EEDF). Molecular vibrations
can be easily excited by electron collisions. Provided relaxation of vibrational energy is slow,
vibrational degrees of freedom are non-equilibrium, too. This requires description of molecular
vibrations to be done in terms of a vibrational distribution function. Usually, rates of ion-molecular
reactions are remarkably higher than rates of reactions of neutral species. It means that despite
rather low number densities of charged particles, their role in plasma evolution can be significant.
The full kinetic model comprises self-consistent combination of electron and vibration kinetics in
terms of respective distribution functions with chemical and ion-molecular kinetics in terms of
reactive species concentrations. There are a few groups in the world doing full kinetic models:
Centro de Fizica dos Plasmas, Lisbon, Portugal; University of Bari, Italy; Centre de Physique
Atomique de Toulouse, France; Institute of Physics, Belgrade, Serbia, and the SRC TRINITI,
Troitsk, Russia.
Actually, for any specific system far not all processes are equally important. The full kinetic
model can be a good starting point for analyzing relative roles played by different species. An
additional analysis of a specific system at specified conditions (gas temperature, pressure and
composition, typical electric current density) allows one to make a significant simplification of a
model. Such a reduced model containing a reasonably small amount of reactions can be employed
on a second step for further analysis of effects of spatial non-uniformities and/or time variations.
For example, it was shown in [10], that for modeling of atmospheric pressure glow discharge in fast
dry airflow one may neglect plasma density gradients and describe evolution of plasma as it is
transported with the gas flow (plug-flow model). Such a plug-flow model was developed in [10-14]
for atmospheric air plasma. It was demonstrated a good agreement between theoretical predictions
and experimental measurements for fast airflow glow discharge.
The model incorporates kinetic equations for charged and neutral species including electrons,
positive and negative ions, molecules, excited species, and radicals. An efficiency of production of
the chemically active species is determined by an electron energy distribution function (EEDF)
formed by electric field and collisions with atoms and molecules. The EEDF is found numerically
by solving the electron Boltzmann equation. The influence of physical parameters such as gas
temperature, pressure, and applied electric field on the EEDF shape can be described adequately by
one combined parameter controlling practically all important discharge characteristics. This is the
value of the reduced electric field strength E/N, where E is the electric field strength and N is the
total gas density. Besides, gas chemical composition is an important factor, as well. For typical
conditions of plasma treatment, E/N parameter and gas composition change in time. Therefore, the
electron Boltzmann equation should be recalculated when necessary.
Generally, discharges producing non-thermal plasma at atmospheric pressure are inherently non-
uniform and proceed in a form of a manifold of micro-discharges (DBD) or as fast propagating
streamers (pulse coronas). However, location of micro-discharges in DBD is not fixed and varies
from pulse to pulse covering in average all electrode surface uniformly. Then it is reasonable to
employ a zero-dimensional model for simulation of evolution of a pulse repetitive single micro-
discharge with following averaging over whole discharge space. Such an approach allows one to
model plasma chemical processes in realistic gas mixtures. This approach was implemented
successfully for nitrogen oxides generation in pulse-periodic DBD [15].
 In general, evolution of micro-discharges (MD) and streamers proceeds through formation of
self-organized transient structures controlled by correlations between electric field and spatial
charge distributions. All these factors make actual multi-dimensional modeling of discharges. Due
to complexity of space-time modeling, only simplest kinetic models can be included. Micro-
discharge dynamics in the DBDs of various geometries in oxygen has been numerically simulated
by 3D fluid models in collaborative work of [16]. Earlier theoretical studies are reviewed in [17].
2D approach to modeling the MDs is reported in [18]. Formation and evolution of streamers in
pulsed coronas were analyzed with help of 3D axial-symmetry fluid models by groups [19-24].
Typical diameters of an MD or streamers are 10-100 m, and typical times of their existence are
10-1000 ns. At such conditions, one has to include into consideration effects of gas dynamics and
dissipation processes like diffusion and thermal diffusion. M. Kushner [25] has made such
modeling for the MD in a metal-dielectric-metal cylindrical device in Ar atmosphere. It was found
that properties of the MDs are nearly the same as of classical glow discharge resulting in sensitivity
of plasma parameters to variations of secondary emission coefficients of the treated surface.
Actually, classic negative and positive corona discharges operate in a self-pulsation mode in a
wide range of conditions. In the negative corona these are widely known regular Trichel pulses [26]
and in positive coronas these are current oscillations in a number of gases.
In air at atmospheric pressure, effects of non-locality of electron and ion distributions and
diffusion processes, as well, could play some role only in the very vicinity of the point electrode. It
is known that the cathode layer parameters are insensitive to the kinetic model used. Therefore, the
simplest fluid model can be taken where all the transport and kinetic coefficients are functions of
the local value of E/N. To describe the pulse mode of the negative and positive point-to-plane
coronas it is sufficient to solve the well-known continuity equations for:
electrons ne/ t + div newe = ne + d nn, (1)
positive ions np/ t + div np wp= ne, (2)
negative ions nn/ t + div nn wn = ne - d nn,, (3)
and Poisson equation div E = e(np - ne - nn)/0, (4)
where the indexes e, p and n refer to electrons, positive and negative ions, respectively, np , ne
and nn are the positive ion, electron and negative ion number densities, wp, we and wn their drift
velocities, I, a, and d are the ionization, attachment, and detachment frequencies, e is the
electronic charge, 0 is the permittivity of free space. The electron drift velocity and kinetic
coefficients are to be determined from solving the electron Boltzmann equation, the ion drift
velocities are calculated using the known ion mobilities. The current in the external circuit, I, is
determined from the equation
V = U0 - RI, (5)
 where V and U0 are the discharge and power supply voltage, R is the ballast resistor.
This relatively simple system of differential equations with transport and kinetic coefficients
calculated by an offline Boltzmann solver is rather hard for numerical simulations for typical
geometries of electrodes and/or plasma objects. The challenge to numerical methods is a very
strong variation in a size of the current channel as a function of time or distance. In a rigorous
approach one has to solve 3D space and time dependent equations. Such the approach was realized
for axial symmetry of discharges in [16, 19-24, 27]. Having knowledge from experiments about
structure of discharge, it is possible to reduce dimensionality of the mathematical problem. Such an
approach to modeling Trichel pulses in negative corona and self-oscillations in positive corona was
realized by authors [28-30]. Similar approach was employed in [31] to describe theoretically
corona-to-glow transition taking place in fast flow atmospheric pressure discharge with current
increase. In all these cases, using the shape of the current channel a rather good agreement was
achieved between predicted and observed discharge characteristics and parametric dependences of
pulse repetition frequencies and average currents.

10.3.2 Electron kinetics

The detailed knowledge of electron kinetics in non-thermal plasma is required for plasma
chemistry modeling in technological gases (air, N2-, Ar- or He-based mixtures and others). For wide
variety of discharge types (glow, silent, corona, pulse-periodic, cw) the EEDF plays a key role in
control of gas plasma parameters. In particular, a mean electron energy determines the excitation,
dissociation and ionization rates and the partition of discharge energy input into different channels
for the given gas mixture. Experimental study of the EEDF in atmospheric pressure plasma meets
challenges associated with numerous fast transients in plasma. Therefore, theoretical approaches are
of particular importance. The most common theoretical approach to EEDF determination in gas
plasma is Boltzmann equation (BE) solving in the two-term approximation. This approach reveals
to be sufficiently simple, reliable and had been proved by comparison with the more accurate model
methods, for example Monte-Carlo calculations.
Electron motion in the gas under applied electric field is determined by collisions with
molecules and other electrons. Electron-electron collisions begin to play some role at a relatively
high gas ionization degree, usually starting from values of 10-4 - 10-3 for molecular gases. Present
review is aimed upon description of the modern state of the discharge modelling for gases under
normal conditions with total molecule concentration more than 1019 cm-3. Typical electron densities
in corona and glow discharge are much lower than 1015 cm-3, hence electron-electron interactions
may be neglected with a good accuracy.
 The influence of physical parameters such as gas temperature, pressure, applied electric field on
the EEDF shape can be described adequately by one combined parameter determining practically
all important discharge characteristics. This is the value of the reduced electric field strength E/N,
where E is the electric field strength and N is the total gas density. The most of calculated plasma
characteristics for the fixed gas composition are the functions of E/N only without detectable
dependence on other physical parameters in wide ranges of variation.
The most important components of air are O2, N2, CO2, and H2O. Their concentrations depend
on many conditions and may vary in a wide range. Furthermore, numerous minor additives may
present in gas, which do not exert strong influence on the shape of the EEDF. These admixtures
may be strongly electronegative or chemically aggressive. To describe their role in discharges it is
sufficient to average cross sections of interest with the EEDF found neglecting these contaminants.
A lot of data exist about electron collision cross sections corresponding to different processes in
the above listed gases and some rare gases (Ar, He).
An important problem does exist with formation of an electron scattering cross section set for a
given molecule. Actually, for many practical simulations of plasma devices existing information
about electron scattering cross sections is excessive. For example, the cross sections for electron
scattering from N2 molecules are measured for excitation of 17 electronic states [32] and the
recommended in [33] set of cross sections includes excitation of 11 electronic states. Reactions for
molecules in many of these excited states are not known, in particular for gas mixtures containing
H2O, CO2, and O2. In such a situation, one has two options for kinetic modelling. The first is to
introduce voluntarily lacking quenching processes; the second is to reduce the quantity of molecular
states incorporated into the model. The second way is preferable when modelling electric discharge
evolution in time and space. However, one has to take care of consistency of BE analysis for the
reduced number of collision processes accounted for in the model. Generally, it is necessary to
create a reduced set of electron collision cross sections for every species included into the kinetic
model.
In order to guarantee that this reduced set of cross sections is appropriate for kinetic modeling,
the following procedure should be realized. After formulation of the reduced kinetic model, the
cross sections for proper electron inelastic processes for every species are constructed from the
known full set of cross sections for this species. For example, excitation rates for a number of
electronic levels of a molecule can be summed, and instead of a manifold of levels one effective
lump level is introduced. Then the procedure of adjustment of effective cross sections should be
applied similar to that developed earlier by A. V. Phelps [34]. Namely, transport coefficients
(electron drift velocity, transverse and longitudinal diffusion coefficients) and kinetic coefficients
(for ionization, attachment, excitation) are calculated by solving the BE. Comparing calculated
transport and kinetic coefficients as functions of E/N with experimental ones, the differences
between them are minimized by fitting amplitudes of effective cross sections. The effective cross
sections found in this way differ from the cross sections entering into the full set of cross sections.
The set of cross sections generated with such the methodology is called traditionally as a self-
consistent set. Change of any one of such cross sections, made f. e. after appearance of new data in
direct measurements, may result in a wrong description of the electron kinetics. Any change of
cross sections for processes included into the model should include as an obligatory step the
procedure of fitting calculated transport and kinetic coefficients to the experimental data. The sets
of electron collision processes accounted for in such reduced models of technological plasma for
N2, O2, H2O, CO2 species are listed in Table 10.1.
Table 10.1. Reactions of electrons with molecules.
No Process Threshold, Comments References
eV
1 O2 + e  O2(vibr) + e 0.193 v=1,2,3,4 [35, 50]
2 O2 + e  O2(0.98) + e 0.977 O2(a1) [35, 50]
3 O2 + e  O2(1.63) + e 1.627 O2(b1) [35, 50]
4 O2 + e  O- + O 4.2 dissociative attachment [36, 50]
5 O2 + e  O2(4.5) + e 4.48 O2(c1, C3, A3) [50]
6 O2 + e  O + O + e 5.58 [50]
7 O2 + e  O + O(1D) + e 8.4 [35, 50]
8 O2 + e  O2+ + e + e 12.1 [35, 50]
9 O2 + e  O + O+ + e + e 19.5 [35, 50]
10 O2+e+(O2)O2- + (O2) 0.051 three body attachment [35]
11 N2 + e  N2(vibr) + e 0.29 v=1,2,...,8 [11]
12 N2 + e  N2(E=1) + e 6.17 N2(A3) [11]
13 N2 + e  N2(esum) + e 7.35 N2(B3), N2(C3) [11]
14 N2 + e  N + N + e 9.0 [11]
15 N2 + e  N2+ + e + e 15.6 [11]
16 H2O + e  H2O(rot) + e 0.040 [37]
17 H2O + e  H2O(vibr1) + e 0.198 [37]
18 H2O + e  H2O(vibr2) + e 0.453 [37]
19 H2O + e  H + OH + e 7.1 [37]
20 H2O + e  H2O+ + e + e 12.61 [37]
21 H2O + e  OH + H- 5.7 [37]
22 H2O + e  H2 + O- 4.9 [37]
23 CO2 + e  CO2(vibr) + e 0.083 [37]
24 CO2 + e  CO + O + e 6.0 [37]
25 CO2 + e  CO2+ + e + e 13.3 [37]
26 CO2 + e  CO + O- 3.3 [37]

Figure 10.3a–b The self-consistent reduced set of electron scattering cross sections: a for N2 [11]. 1
– momentum transfer including inelastic scattering processes; 2 – resonant vibrational excitation, 2’
near threshold vibrational excitation (process 11 in Table 2.5.1); 3 – A3+u electronic level
excitation (12); 4 – total electronic levels excitation with threshold higher than 7 eV (13); 5 –
dissociation (14); 6 – ionization (15); b for CO2 [37]. 1 – momentum transfer including inelastic
scattering processes; 2 – vibrational excitation (process 23 in Table 10.1); 3 - dissociative
attachment (26); 4 –dissociation (24); 5 – ionization (25).

The reduced sets of electron scattering cross sections appropriate for technological plasma
modeling are shown in Figures 10.3, 10.4. Usage of these reduced sets of cross sections allows us to
reproduce with a good accuracy (within a few percent) swarm data on electron drift velocity,
transverse and longitudinal diffusion coefficients and ionization and attachment coefficients, when
applicable. Comparison between calculated ionization and attachment coefficients for O 2 and H2O is
made in Figures 10.5, 10.6.
Figure 10.4 The self-consistent reduced set of electron scattering cross sections for H2O [37]. 1 –
momentum transfer including inelastic scattering processes; 2, - excitation rotational levels (process
16 in Table 2.5.1); 3, 4 – vibrational excitation (17, 18); 5, 6 - dissociative attachment with creative
O– and H– respectively (21, 22); 7 –dissociation (19); 8 – ionization (20).

a b

Figure 10.5a–b a Ionization coefficient as a function of E/N in H2O. Markers are experimental
data, solid line calculation with the reduced set of cross sections. + - [38]; ○ - [39]; ◊ - [40]. Here
the reduced electric field E/N is measured in the Td, 1 Td = 10-17 Vcm2. b Dissociative electron
attachment coefficient as a function of E/N in H2O. Markers are experimental data, solid line
calculation with the reduced set of cross sections. + - [38]; ○ - [39]; ◊ - [40], Δ - [41]; × - [42]. Here
the reduced electric field E/N is measured in the Td, 1 Td = 10-17 Vcm2.

Numerical solving of the BE for the EEDF allows one to calculate rate coefficients of
elementary processes involving electrons (see Table 10.1) as a function of E/N.
Figure 10.6 Dissociative electron attachment (η/N) and ionization (α/N) coefficients as a function
of E/N in O2. Markers are experimental data, solid line calculation with the reduced set of cross
sections. ○- compilation: [39]; □ - [43]; Δ - [44]; ◊ - [45];  - [46];  - [47]. Here the reduced
electric field E/N is measured in the Td, 1 Td = 10-17 Vcm2.

10.3.3 Plasma chemistry

For description of any particular plasma chemical system one needs a lot of information about
mechanisms of chemical reactions and about rate coefficients for various conditions, which are
characterized by gas pressure and temperature, excitation degree and so on. An amount of different
chemical species in a typical system is rather large. For reliable modeling of any particular system
requires great efforts to be put forth for accumulation of data on processes in this system. Rather
extensive data sets for modeling plasma chemistry in dry air can be found in [4, 11, 48, 49]. Most
detailed information about plasma chemistry in oxygen plasma can be found in [50]. Data for
plasma chemistry at room temperature in polluted gases treated by beams of high-energy electrons
are collected in [51].
Table 10.2 contain information about chemical reactions of neutral gas components appropriate
for modeling of plasma polymer surface treatment in order to enrich the upper layers with oxygen.
The enrichment of the polymer with oxygen results in improvement of its surface wettability.
Table 10.2. Neutral species reactions
No Process Rate Coefficient, cm3/s Reference

O2(4.5) + M ® O2 + M 2.3×10-14 [52]

O2(4.5) + M ® O2(a1D) + M 1.86×10-13 [52]

O2(4.5) + M ® O2(b1S) + M 8.1×10-14 [52]

O + O + M ® O2+ M 1.9×10-30×T-1×exp(-170/T) [52]

O + O + M ® O2(a1D) + M 1.3×10-30×T-1×exp(-170/T) [52]

O + O + M ® O2(b1S) + M 6×10-31×T-1×exp(-170/T) [52]

O + O + M ® O2(4.5) + M 1.2×10-34 [52]

O + O2 + O2 ® O3 + O2 8.6×10-31×T-1.25 [53]

O + O2 + N2 ® O3 + N2 5.6×10-29×T-2 [53]

O + O3 ® O2 + O2 9.5×10-12×exp(-2300/T) [54]
[52]

6.3×10-12×exp(-2300/T) [54]
O + O3 ® O2(a1D) + O2
[52]

3.2×10-12×exp(-2300/T) [54]
O + O3 ® O2(b1S) + O2
[52]
O + NO + M ® NO2 + M 9.1×10-28×T-1.6 [55]

O + NO2 ® NO + O2 3.26×10-12×T0.18 [48]

O + NO2 + M® NO3 + M 8.1×10-27×T-2 [53]

1.7×10-11 [51]

O + NO3 ® NO2 + O2
10-11 [54]
[56]

O(1D) + N2O ® N2 + O2 4.4×10-11 [55]

O(1D) + N2O ® NO + NO 7.2×10-11 [55]

O(1D) + N2O ® O + N2O 10-12 [53]

O(1D) + NO ® O2 + N 8.5×10-11

O(1D) + NO2 ® O2 + NO 1.4×10-10 [51]

O(1D) + O ® O + O 7.5×10-11 [57]

O(1D) + O2 ® O + O2(b1S) 2.56×10-11×exp(67/T) [57]

O(1D) + O2 ® O + O2 6.4×10-12×exp(67/T) [57]

O(1D) + N2 ® O + N2 1.8×10-11×exp(107/T) [53]

O(1D) + O3 ® O2 + O + O 1.2×10-10 [52]

O(1D) + O3 ® O2 + O2 2.3×10-11 [52]

O(1D) + O3 ® O2(a1D) + O2 1.5×10-11 [52]

O(1D) + O3 ® O2(b1S) + O2 7.7×10-12 [52]

O(1D) + O3 ® O2(4.5) + O2 7.4×10-11 [52]

O2(a1D) + O ® O2 + O 1.3×10-16 [54]

O2(a1D) + N2 ® O2 + N2 1.4×10-19 [53]

O2(a1D) + O3 ® O2 + O2 + O 5×10-11×exp(-2830/T) [54]

O2(a1D) + O2(a1D) ® O2(b1S) + O2 2.2×10-17 [54]

O2(b1S) + O ® O2(a1D) + O 8×10-14 [57]

4×10-17 [55]
O2(b1S) + O2 ® O2 + O2
1.5×10-16 [57]

O2(b1S) + N2 ® O2 + N2 2×10-15 [54]

O2(b1S) + O2 ® O2(a1D) + O2 4×10-17 [57]

O2(b1S) + N2 ® O2 + N2 2×10-15 [54]

O2(b1S) + O2 ® O2(a1D) + O2 4×10-17 [57]

O2(b1S) + O3 ® O2 + O2 + O 1.5×10-11x [57]

O2(b1S) + O3 ® O2(a1D) + O2 + O 7×10-12 [52]

N2(A3S) + N2(A3S) ® N2(esum) + N2 2×10-10

N2(A3S) + O2 ® N2 + O + O 2.29×10-12 [58]

N2(A3S) + O2 ® N2O + O 4.6×10-15 [58]

N2(A3S) + N2 ® N2 + N2 3.7×10-16 [56]

2.8×10-11 [56]
N2(A3S) + NO ® N2 + NO [51]
1.5×10-10

1.4×10-11 [56]
N2(A3S) + N2O ® N2 + N2 + O
8×10-11 [51]

N2(A3S) + NO2 ® NO + O + N2 10-12 [56]

[51]

N2(esum) + M ® N2(A3S) + M 10-13

 N2(esum) + O2 ® N2 + O + O 2×10-12 [59]

N + NO ® N2 + O 3.1×10-11 [60]

N + NO2 ® N2O + O 1.4×10-12 [53]

N + NO2 ® NO + NO 6×10-13 [56]

N + O2 ® NO + O 1.1×10-14×T×exp(-3150/T) [54]

N + O2(a1D) ® NO + O 10-16 [53]

10-15 [53]
N + O3 ® NO + O2
5.7×10-13 [54]

N + N + M ® N2 + M 8.3×10-34×exp(500/T) [54]

N + O + M ® NO + M 1.8×10-31×T-0.5 [54]

N2 + M ® N + N + M 6.1×10-3×T1.6×exp(-113200/T) [54]

NO + M ® N + O + M 6.6×10-4×T-1.5×exp(-75500/T) [54]

NO + O3 ® NO2 + O2 9×10-13×exp(-1200/T) [54]

NO + NO3 ® NO2 + NO2 2×10-11 [53]

NO + NO + O2 ® NO2 + NO2 3.3×10-39×exp(526/T) [53]

NO + NO3 ® NO + NO + O2 2.71×10-11×T-0.23×exp(-947/T) [51]

NO2 + M ® O + NO + M 1.8×10-8×exp(-33000/T) [54]

NO2 + NO2 ® NO + NO + O2 3.3×10-12×exp(-13540/T) [54]

NO2 + NO3 + M ® N2O5 +M 5.3×10-20×T-4.1 [53]

NO2 +O3 ® NO3 + O2 1.2×10-13×exp(-2450/T) [51]

NO2 + O3 ® NO + O2 + O2 10-18 [60]

NO2 + NO3 ® NO2 + NO + O2 2.3×10-13×exp(-1600/T) [51]

NO3 +NO3 ® NO2 + NO2 + O2 7.5×10-12×exp(-3000/T) [51]

N2O5 +M ® NO2 + NO3 + M 1.7×108×T-4.4×exp(-11080/T) [53]

10.3.4 EXPERIMENTAL EMISSION SPECTRA

 Composition of active species generated by specific plasma-jet source and distribution of these
species along plasma jet can be determined from spectroscopic measurements. Optical spectra of
light emitted by plasma jets of sources No1 and No2 were recorded with spectroscope. Figure 10.7a
shows spectral lines of different species formed and excited in plasma jet in ambient air. One can
see that N- and O-atoms, NO- and OH-radicals, and especially molecular nitrogen excited in N 2(C)
and N2(B) states give major input in the radiation from the air plasma jet. Decay in the light
emission intensity of different active species along air plasma jet is shown in Figure 10.7b.

a b

Figure 10.7a–b a Total optical spectrum in UV and visible region of a radiation emitted from
plasma jet in airflow (plasma source No1). Gas flow velocity is 50 m/s. Discharge power is 35 W.
b Longitudinal emission intensity distribution of active species in the air plasma jet. Coordinate x=0
is the outlet of the source. 1 - N2(C3u, v`=0  B3g, v``=0) transition, λ=337.1 nm, N2 second
positive system; 2 - О(5S05P) transition, λ=777.2nm; 3 - OH(A2Σ+, v`=0X2П, v``=0) transition,
λ=306nm.

Spectroscopic information on UV and visible light emission from the middle of discharge zone
of N2-plasma source is shown in Figure 10.8a. Main input to emission from this region corresponds
to 1+ and 2+ systems of nitrogen. Low intensity emission of CN is recorded only from the region
close to the outlet of discharge zone (see Figure 10.8b). Afterglow UV and visible spectrum of a
radiation emitted from N2-plasma jet at different distances away from the outlet of the plasma
source nozzle are presented in Figure 10.9. We would like to attract serious attention to a surprising
fact in afterglow: despite high purity of nitrogen (99.999%) used as plasma forming gas the light
emission in spectral lines does not correlate with emission from the excited states of nitrogen itself
as it shown in Figure 10.8a. The very small impurities of radicals CN and NH give main
contribution to the light emission from active N2 plasma jet afterglow at atmospheric pressure. It
means that small impurities in pure gas can influence strongly on the efficiency of plasma
treatment.
 a b

Figure 10.8a–b a UV and visible spectrum of a radiation emitted from the middle of discharge
zone (plasma source No2). Discharge power is 70 W. Gas flow rate of nitrogen is 440 cm3/s.
b Change in intensity of CN spectrum with location of the emitting discharge region inside plasma
source No2. 1 – region close to the inlet; 2 - region close to the outlet.

a b

Figure 10.9a–b Afterglow UV and visible spectrum of a radiation emitted from N2-plasma jet
(plasma source No2) at different distances away from the outlet of the plasma source nozzle.
Discharge power is 70 W. Gas flow rate of nitrogen is 440 cm3/s. a 1.5 mm away from nozzle. b 10
mm away from nozzle.

As one can see in Figure 10.9a, the light emission at short distances from the nozzle is caused,
mainly, by radicals CN and NH while at the longer distances (10 mm or more) the essential
contribution to the plasma emission is made by NO (γ-band) (see Figure 10.9b).
Under plasma processing, narrow N2-plasma jet strikes the sample, expands over the surface and
occupies large area. It is interesting therefore to know a radial distribution of active species and gas
temperature in plasma over the treated zone. Such kind of experimental information is presented in
Figure 10.10.

a b

Figure 10.10a–b Radial distribution: a of light intensity of CN(B2Σ+ → X2Σ+) (1) and
NH(A3П→ X2Σ+) (2) emitted from N2-plasma jet expanded over the surface to be treated. b of gas
temperature in plasma jet expanded over the surface to be treated. The surface is placed at a
distance of 3 mm away from nozzle of plasma source. Discharge power is 70 W. Gas flow rate of
nitrogen is 440 cm3/s.

So, due to an existence in pure nitrogen of trace gaseous admixtures like CH 4, H2O, O2, plasma
composition can be changed drastically. Our experimental results presented above prove
unambiguously that small gas additives play crucial role in generation by plasma of active species,
and their interaction with polymer surface has to be taken therefore into account in theory and
practice.
Another subject of our investigations is CF 4 that is widely used as plasma forming gas in the
DBD-systems for hydrophobic processing of the fabrics and polymers. Purity of CF4 used in the
experiments is 99.5%, and air is the dominant impurity in CF 4 (about 0.5%). Survey spectrum of
light emitted from DBD-plasma in gas mixture CF 4 (99.5% CF4 + 0.5% air) for wavelength range
from 240 nm to 800 nm is shown in Figure 10.11.
Figure 10.11 Survey UV-visible emission spectra from DBD. Pressure of working gas P=1.2atm.
Gas mixture is 99.5% CF4 and 0.5% air.

Emission spectra with higher resolution in two wavelength ranges 240 nm – 450 nm and 450 nm
– 800 nm for two gas mixtures (99.5% CF4 + 0.5% air) and (10% CF4 and 90% nitrogen) are
presented in Figure 10.12. One can see that CF3-radicals give the major input in the light emission
from DBD-plasma in gas mixtures containing CF4. It should be noted that spectral lines of N2-
second positive system emitted from gas mixture (99.5% CF4 + 0.5% air) appear due to air
admixture.
For gas mixture (99.5% CF4 + 0.5% air), broad band emission of CF3-radical is seen clearly in
Figure 10.12b. Atomic spectral lines of the ionized oxygen observed in the spectrum appear
probably due to admixture of air in the working gas mixture. For comparison, emission spectrum of
pure nitrogen is shown in Figure 10.12b as well.

a b

Figure 10.12a–b The emission spectrum from DBD-plasma in gas mixtures containing CF4. Total
gas pressure P=1.2atm. 1 – 99.5% CF4 and 0.5% air; 2 – 10% CF4 and 90% nitrogen; 3- pure
nitrogen. a in the range 250-450 nm; b in the range 450-800 nm.
10.3.5 INFLUENCE OF IMPURITIES ON COMPOSITION OF GAS
ACTIVATED BY NON-THERMAL PLASMA
One of important parameters, which can serve for optimization of the plasma surface treatment
is a buffer (transportation) gas. A key criterion for buffer gas choice is a feasibility to realize high
reduced energy input in non-thermal discharge. Besides, the energy consumed by the buffer gas
should effectively used for production of chemically active species and UV radiation. In
experiments, there are a lot of factors making it difficult to identify mechanisms for observed
variations of discharge treatment efficiency of a specific surface. Theoretical modelling can
elucidate significantly analyses of experimental observations.
Here an example is considered when experimentally observed [61] high efficiency of PP and
PET surfaces treatment by a glow discharge in “pure” nitrogen can be explained by existence in this
nitrogen of traces of other species at ppm level. It was revealed experimentally (Figures 10.8, 10.9)
that plasma emission spectrum of this discharge exhibits molecular systems of species, which can
appear only from impurities.
Specification of gas composition supplied by manufacturers in high-pressure vessels is usually
given by the supplier. Typical level of small admixture is on the level of a few ppm. In particular, in
so-called “pure” nitrogen there exist different admixtures like O2, H2O, CH4 and other
hydrocarbons. To evaluate the role of these admixtures in discharge products formation, simulations
were performed for conditions close to our experiments with micro-streamer discharge for additions
of O2, H2O and mixture of all three components.
It was found that at the afterglow phase atomic oxygen decays rather fast due to conversion to
ozone. As a result, at time 69 s atomic oxygen concentration in polluted (10 ppm O2) nitrogen is
higher than in air. It is interesting to find a maximum in oxygen atom concentration in N 2 with
variable concentration of O2. Actually, this maximum depends on the moment when the
measurement was done. Figure 10.13a shows the dependence of oxygen atom number density on
concentration O2 in the mixture O2:N2 at the moment 100 s after electric discharge. There is the
pronounced maximum of oxygen atom concentration 3.6·1016 cm-3 at the concentration O2 equal to
0.5%. For comparison, the oxygen atom concentration is equal to 2.4·1014 cm-3 and 1.7·1014 cm-3 for
initial concentration O2 5 ppm and 20% (synthetic air), correspondingly. Using mixture 0.5%O2-N2,
which is typical for so-called “commercial purity nitrogen”, the oxygen atom density increases 200
times in comparison with dry air. In our calculation we used the electrical circuit with the fixed
parameters: discharge capacity, charging voltage on the discharge capacity and others. During
variation gas mixture in our simulation the input discharge energy was changed in the range 600-
750 J/liter.
 a b

Figure 10.13a–b Dependence of oxygen atom (a), radicals and active molecules (b) number
density on concentration of O2 in the mixture O2:N2 at two moments after electric discharge. Solid
line – 100 s, dashed line – 1s. P0 = 1 atm, T0 = 300 K, Ein = 600 J/liter.

Figure 10.13b shows dependence of radical and active molecule number densities on
concentration O2 in the mixture O2:N2 at the time 1 s and 100 s after electric discharge. In
comparison with Figure 10.13a, Figure 10.13b uses the logarithmical scale. The first moment (1 s)
corresponds to the end of discharge. The second moment (100 s) approximately corresponds to the
time of transport plasma to the surface of the processed film.

a b

Figure 10.14a–b a Radical and active molecule number density as a function of H2O in the
mixture H2O with dry air at two moments after electric discharge. Dashed line – 1s, solid line –
100 s. Conditions are as for Figure 10.13. b Atom and molecule number density as a function of
time for the mixture of nitrogen and 5 ppm CH4, 5 ppm O2, 5 ppm H2O.
 Figure 10.14a shows dependence of radical and active molecule number densities on
concentration of H2O in the mixture H2O with dry air at the moments 1 s and 100 s after electric
discharge. Presence of a few percent of H2O in the air decreases the oxygen atom density 10 times
at the moment 100 s. The number density of OH radicals increases with increasing concentration
of H2O. At the initial concentration of H2O 5%, the OH radical number density at the moment 100
s is of the same order as the oxygen atom number density.
Simulations were performed also for nitrogen with additions of CH4, O2, and H2O with equal
concentrations 5 ppm. Respectively, our data base on electron scattering cross sections was
extended by inclusion of a self-consistent set of cross sections for CH4 molecules. It is seen from
Figure 10.14b that admixture molecules are destroyed in afterglow phase in more degree than in the
discharge (about 0.5 s). This is the typical situation for impurities, which are removed in reactions
with active species produced in the discharge from the main components. Time behavior of a
number of radicals produced in the discharge and afterglow is illustrated in Figure 2.5.24.

Figure 10.15 Radical number density as a function of time for the mixture of nitrogen and 5 ppm
CH4, 5 ppm O2, 5 ppm H2O. Conditions are as for Figure 10.13.

To explain discharge emission spectrum observed experimentally, information about how

electronically excited molecules are formed is needed. Such information can be found in Table
10.3.
Table 10.3. Reactions leading to electronically excited molecules and radicals
# Reaction Rate constant, Reference
cm3/s, cm6/s
1 N2(X, ,v>12) + CN(X2) -> N2(X, ) + CN(B2+) 1.2·10-11 [62]

2 N2(A3) + CN(X2) -> N2(X,) + CN(B2+) 1.2·10-10 [63]

3 C + N + M -> CN(B2+) + M 9.4·10-33 [63]

4 N2(A3) + OH(X2) -> N2(X,) + OH(A2) 1·10-10 [64]

5 N2(A3) + NO(X2) -> N2(X,) + NO(A2) 6.9·10-11 [65]

The amount and sort of radicals approaching the surface can vary strongly with change of gas
composition. In particular, small admixtures (O2, H2O, CxHy) to pure gases, such as Ar or N2, can
significantly influence on the activation doze. Our numerical data demonstrate an opportunity to
increase processing efficiency by a proper choice of the components of work gas mixture.

10.4 Kinetic model for chemical reactions on PP surface in

atmospheric pressure air plasma

10.4.1 Description of kinetic model

Modeling of plasma surface interactions for conditions of air non-thermal plasma at atmospheric
pressure is the most challenging part of the theoretical approach to simulations of the material
processing by plasma. The model is based on estimations, which show that the surface processing
by plasma products at atmospheric pressure is controlled by neutral chemistry reactions. The
surface reaction mechanism for PP was taken from references [66, 67-70]. Here we describe
important features of this mechanism.
PP is a saturated hydrocarbon polymer with carbon backbone containing hydrogen and methyl
(CH3) groups (see Figure 10.16). Gaseous particles approaching the surface of the perfect PP
material meet only hydrogen atoms exposed outside. In general, the reactivity of the hydrogen
atoms depends on the nature of the C atom to which they are attached. Carbon atoms in the PP
monomer unit can be divided into three groups depending on the number of neighbouring C atoms:
primary, secondary and tertiary C atoms (see Figure 10.16). According to the paper of R. Dorai and
M. Kushner [66], the reactivity of hydrogen bound to these C atoms scales as: Htert > Hsec > Hpri.

Figure 10.16 Three types of C atoms in polypropylene monomers.

 Further we will use the term “in chain” to designate radicals and groups formed in the polymer
chain (not at the end of the chain) and the term “end free” to designate radicals and groups formed
at the end of polymer chain. “End free” alkyl radicals are appeared due to the bond scission
processes. Some of radicals considered in the model are listed in Table 10.4.

Table 10.4. Examples (symbol * designates chemical vacant site) of surface radicals structure
“In chain” tertiary alkyl radical “End free” primary alkyl radical
H H H H
*

-
-
-

- - C C ~

*
~ C C C ~

-
-
-
-

H CH3
H CH3 H

“In chain” secondary alkyl radical “End free” secondary alkyl radical

H H H H
- *

- -

- -
- -
- -
- -

C C C ~ ~ C C

*
~

H CH3 H H CH3

“In chain” tertiary alkoxy radical “End free” secondary alkoxy radical
*

H O H
H H
-
-
-

~ C - C - C ~
- -
-

-
~ C C O

*
-
-
-

H CH3 H
-

-
H CH3

In description of reactions “in chain” radicals and groups are designated using term PP. For
example: PP* is in chain alkyl radical, PPO* is in chain alkoxy radical and so on. For designation
of “end free” groups and radicals the term R is used. For example: R* is end free alkyl radical,
ROO* is end free peroxy radical and so on. If it will be necessary to indicate explicitly the type of C
atom (primary, secondary or tertiary) with a given group or radical, then the terms 1PP, 2PP, 3PP,
1R and 2R will be used. For example, 2PPO* is secondary in chain alkoxy radical. Let us note that
the type of end free C atoms can be only primary or secondary (see Table 10.4). The full list of
processes taken into account in the model is presented in Table 10.5.

Table 10.5. Processes included in the kinetic model for the plasma treatment of PP film in
atmospheric pressure discharge in air
Rate constants (cm2/s), Comments,
Processes frequencies (s-1) or ref.
probability
1 3PP + O => 3PP* + OH 0.001 [66]
2 2PP + O => 2PP* + OH 10-4 [66]
3 1PP + O => 1PP* + OH 10-5 [66]
4 3PP + OH => 3PP* + H2O 0.25 [66]
5 2PP + OH => 2PP* + H2O 0.05 [66]
6 1PP + OH => 1PP* + H2O 0.0025 [66]
7 3PP* + O => 3PPO* 0.1 [66]
8 2PP* + O => 2PPO* 0.01 [66]
9 1PP* + O => 1PPO* 0.01 [66]
10 3PP* + O3 => 31PPO* + O2 1 [66]
11 2PP* + O3 => 2PPO* + O2 0.5 [66]
12 1PP* + O3 => 1PPO* + O2 0.5 [66]
13 3PP* + O2 => 3PPOO* 0.001 [66]
14 2PP* + O2 => 2PPOO* 2.3 10-4 [66]
15 1PP* + O2 => 1PPOO* 5. 10-4 [66]
16 3PP* + H => 3PP 0.2 [66]
17 2PP* + H => 2PP 0.2 [66]
18 1PP* + H => 1PP 0.2 [66]
19 3PP* + OH => 3PPOH 0.2 [66]
20 2PP* + OH => 2PPOH 0.2 [66]
21 1PP* + OH => 1PPOH 0.2 [66]
22 3PPOH + O => 3PPO* + OH 7.5 10-4 [66]
23 2PPOH + O => 2PPO* + OH 7.5 10-4 [66]
24 1PPOH + O => 1PPO* + OH 7.5 10-4 [66]
25 3PPOH + OH => 3PPO* + H2O 8.2 10-3 [66]
26 2PPOH + OH => 2PPO* + H2O 8.2 10-3 [66]
27 1PPOH + OH => 1PPO* + H2O 8.2 10-3 [66]
28 3PPOO* + 3PP => 3PPOOH + 3PP* 5.5 10-16 cm2/s [66]
29 2PPOO* + 2PP => 2PPOOH + 2PP* 5.5 10-16 cm2/s [66]
30 1PPOO* + 1PP => 1PPOOH + 1PP* 5.5 10-16 cm2/s [66]
31 3PPO* + 3PP => 3PPOH + 3PP* 8. 10-14 cm2/s [66]
32 2PPO* + 2PP => 2PPOH + 2PP* 8. 10-14 cm2/s [66]
33 1PPO* + 1PP => 1PPOH + 1PP* 8. 10-14 cm2/s [66]
34 1PPOO* + NO => 1PPO* +NO2 0.02 [67]
35 2PPOO* + NO => 2PPO* +NO2 0.02 [67]
36 3PPOO* + NO => 3PPO* +NO2 0.02 [67]
37 3PPO* => 3PPO + 1R* 500 s-1 [66]
(3PPO – Ketones)
38 2PPO* => 2R* + 2PPO 10 s-1 [66]
(2PPO – Aldehydes)
39 2PPO + O => *2PPO + OH 0.04 [66]
(*2PPO – carbonyl radicals)
40 2PPO + OH => *2PPO + H2O 0.4 [66]
41 *2PPO + O => A* + CO2 0.4 [66]
42 *2PPO + OH => OH2PPO 0.12 [66]
(OH2PPO – acids)
43 1PPOO* + HO2 => 1PPOOH + O2 0.01 See the text
44 2PPOO* + HO2 => 2PPOOH + O2 0.01 See the text
45 3PPOO* + HO2 => 3PPOOH + O2 0.01 See the text
46 1PPO* + HO2 => 1PPOH + O2 0.01 See the text
47 2PPO* + HO2 => 2PPOH + O2 0.01 See the text
48 3PPO* + HO2 => 3PPOH + O2 0.01 See the text
49 1PPOO* + OH => 1PPOH + O2 0.01 Estimate
50 2PPOO* + OH => 2PPOH + O2 0.01 Estimate
51 3PPOO* + OH => 3PPOH + O2 0.01 Estimate
52 1PPO* + OH => 1PPOOH 0.001 Estimate
53 2PPO* + OH => 2PPOOH 0.001 Estimate
54 3PPO* + OH => 3PPOOH 0.001 Estimate
55 1PPOO* + 1PPOO* => 1PPOO1PP + O2 10-15 cm2/s Estimate
56 1PPOO* + 2PPOO* => 1PPOO2PP + O2 10-15 cm2/s Estimate
57 1PPOO* + 3PPOO* => 1PPOO3PP + O2 10-15 cm2/s Estimate
58 2PPOO* + 2PPOO* => 2PPOO2PP + O2 10-15 cm2/s Estimate
59 2PPOO* + 3PPOO* => 2PPOO3PP + O2 10-15 cm2/s Estimate
60 3PPOO* + 3PPOO* => 3PPOO3PP + O2 10-15 cm2/s Estimate
61 1PPO* + 1PPO* => 1PPOO1PP 10-15 cm2/s Estimate
62 1PPO* + 2PPO* => 1PPOO2PP 10-15 cm2/s Estimate
63 1PPO* + 3PPO* => 1PPOO3PP 10-15 cm2/s Estimate
64 2PPO* + 2PPO* => 2PPOO2PP 10-15 cm2/s Estimate
65 2PPO* + 3PPO* => 2PPOO3PP 10-15 cm2/s Estimate
66 3PPO* + 3PPO* => 3PPOO3PP 10-15 cm2/s Estimate
67 1PPOO* + 1PPO* => 1PPO1PP + O2 10-15 cm2/s Estimate
68 1PPOO* + 2PPO* => 1PPO2PP + O2 10-15 cm2/s Estimate
69 1PPOO* + 3PPO* => 1PPO3PP + O2 10-15 cm2/s Estimate
70 2PPOO* + 1PPO* => 2PPO1PP + O2 10-15 cm2/s Estimate
71 2PPOO* + 2PPO* => 2PPO2PP + O2 10-15 cm2/s Estimate
72 2PPOO* + 3PPO* => 2PPO3PP + O2 10-15 cm2/s Estimate
73 3PPOO* + 1PPO* => 3PPO1PP + O2 10-15 cm2/s Estimate
74 3PPOO* + 2PPO* => 3PPO2PP + O2 10-15 cm2/s Estimate
75 3PPOO* + 3PPO* => 3PPO3PP + O2 10-15 cm2/s Estimate
76 1R* + O => 1RO* 0.1 [68]
77 1R* + O2 => 1ROO* 0.01 [68]
78 1R* + O3 => 1RO* + O2 0.001 Estimate
79 2R* + O => 2RO* 0.1 [68]
80 2R* + O2 => 2ROO* 0.01 [68]
81 2R* + O3 => 2RO* + O2 0.001 Estimate
82 1R* + OH => 1ROH 10-5 [68]
83 2R* + OH => 2ROH 10-5 [68]
84 1RO* + HO2 => 1ROH + O2 0.01 Estimate
85 1RO* + OH => 1ROOH 0.001 Estimate
86 2RO* + HO2 => 2ROH + O2 0.01 Estimate
87 2RO* + OH => 2ROOH 0.001 Estimate
88 1ROO* + HO2 => 1ROOH + O2 0.01 Estimate
89 1ROO* + OH => 1ROH + O2 0.01 Estimate
90 2ROO* + HO2 => 2ROOH + O2 0.01 Estimate
91 2ROO* + OH => 2ROH + O2 0.01 Estimate
92 2ROO* + NO => 2RO* + NO2 0.02 Estimate
93 1ROO* + NO => 1RO* + NO2 0.02 Estimate
94 1RO* + 1PP => 1ROH + 1PP* 10-16 cm2/s Estimate
95 1RO* + 2PP => 1ROH + 2PP* 10-15 cm2/s Estimate
96 1RO* + 3PP => 1ROH + 3PP* 10-15 cm2/s Estimate
97 2RO* + 1PP => 2ROH + 1PP* 10-16 cm2/s Estimate
98 2RO* + 2PP => 2ROH + 2PP* 10-15 cm2/s Estimate
99 2RO* + 3PP => 2ROH + 3PP* 10-15 cm2/s Estimate
100 1ROO* + 1PP => 1ROOH + 1PP* 10-16 cm2/s Estimate
101 1ROO* + 2PP => 1ROOH + 2PP* 10-15 cm2/s Estimate
102 1ROO* + 3PP => 1ROOH + 3PP* 10-15 cm2/s Estimate
103 2ROO* + 1PP => 2ROOH + 1PP* 10-16 cm2/s Estimate
104 2ROO* + 2PP => 2ROOH + 2PP* 10-15 cm2/s Estimate
105 2ROO* + 3PP => 2ROOH + 3PP* 10-15 cm2/s Estimate
106 1ROO* + 1PPOO* => 1ROO1PP + O2 10-15 cm2/s Estimate
107 1ROO* + 2PPOO* => 1ROO2PP + O2 10-15 cm2/s Estimate
108 1ROO* + 3PPOO* => 1ROO3PP + O2 10-15 cm2/s Estimate
109 2ROO* + 1PPOO* => 2ROO1PP + O2 10-15 cm2/s Estimate
110 2ROO* + 2PPOO* => 2ROO2PP + O2 10-15 cm2/s Estimate
111 2ROO* + 3PPOO* => 2ROO3PP + O2 10-15 cm2/s Estimate
112 1RO* + 1PPOO* => 1RO1PP + O2 10-15 cm2/s Estimate
113 1RO* + 2PPOO* => 1RO2PP + O2 10-15 cm2/s Estimate
114 1RO* + 3PPOO* => 1RO3PP + O2 10-15 cm2/s Estimate
115 2RO* + 1PPOO* => 2RO1PP + O2 10-15 cm2/s Estimate
115 2RO* + 2PPOO* => 2RO2PP + O2 10-15 cm2/s Estimate
117 2RO* + 3PPOO* => 2RO3PP + O2 10-15 cm2/s Estimate
118 1ROO* + 1PPO* => 1RO1PP + O2 10-15 cm2/s Estimate
119 1ROO* + 2PPO* => 1RO2PP + O2 10-15 cm2/s Estimate
120 1ROO* + 3PPO* => 1RO3PP + O2 10-15 cm2/s Estimate
121 2ROO* + 1PPO* => 2RO1PP + O2 10-15 cm2/s Estimate
122 2ROO* + 2PPO* => 2RO2PP + O2 10-15 cm2/s Estimate
123 2ROO* + 3PPO* => 2RO3PP + O2 10-15 cm2/s Estimate
124 1ROO* + 1ROO* => 1RO* + 1RO* + O2 10-15 cm2/s Estimate
125 1ROO* + 2ROO* => 1RO* 2RO* + O2 10-15 cm2/s Estimate
126 2ROO* + 2ROO* => 2RO* + 2RO*+ O2 10-15 cm2/s Estimate
127 1R* + 1R* => 1R1R 10-15 cm2/s [68]
128 2R* + 2R* => 2R2R 10-15 cm2/s [68]
129 1R* + 2R* => 1R2R 10-15 cm2/s [68]

The brief description of the reaction mechanism for the PP treatment in air plasma proposed in
papers [66, 67] is the following. The modification of PP surface is started with the abstraction of H
atoms by O atoms and OH radicals and, respectively, production of alkyl radicals PP* (nos 1-6 in
Table 10.5):
PP + O => PP* + OH
PP + OH => PP* + H2O
The alkyl radicals react with O and O3 to form alkoxy radicals PPO*, with O2 to produce peroxy
radicals PPOO*, with OH to form alcohols PPOH and with H to regenerate PP monomer (nos 7-21
in Table 10.5):
PP* + O => PPO*
PP* + O3 => PPO* + O2
PP* + O2 => PPOO*
PP* + OH => PPOH
PP* + H => PP
Further reactions of the alcohol groups with O and OH radicals result in the abstraction of
hydrogen atoms and formation of alkoxy radicals (nos 22-27 in Table 10.5):
PPOH + O => PP* + OH
PPOH + OH => PPO* + H2O
The peroxy radicals can abstract hydrogen atoms from neighboring sites to produce
hydroperoxide groups PPOOH and alkyl radicals. The similar processes take place for the alkoxy
radicals, i.e. alkoxy radicals abstract hydrogen atoms from neighboring sites to form, respectively,
alcohol groups and alkyl radicals (nos 28-33 in Table 10.5):
PPOO* + PP => PPOOH + PP*
PPO* + PP => PPOH + PP*
Besides, peroxy radicals can react with NO molecules to produce alkoxy radicals (nos 34-36 in
Table 10.5):
PPOO* + NO => PPO* + NO2
There are also backbone scission processes due to the reactions of alkoxy radicals with PP
backbone, scissions on tertiary carbon atoms lead to the formation of ketones ~(CH3)C=O (no 37
in Table 10.5)
- *

H O H H O H
=
-

~ C - C - C ~ => ~ C - C + C ~
*
-

H CH3 H H CH3 H

and scissions on secondary carbon atoms result in aldehydes ~HC=O (no 38 in Table 10.5)
*

H O H H H O H H
=

- - - - -
-

~ C C C C ~ => ~ C C + C C ~
*
-

CH3 H CH3 H CH3 H CH3 H

 In the kinetic model [66] there are no processes for the ketones degradation. As for the aldehydes,
there are processes for the production of carbonyl radicals (~C=O) by the abstraction of H atoms
by O and OH radicals (nos 39-40 in Table 10.5).
H O H O

- =

* =
- -

- -

- -
~ C C + O => ~ C C + OH

CH3 H CH3

H O H O

- =

* =
- -
- -

- -
~ C C + OH => ~ C C + H2 O

CH3 H CH3

These carbonyl radicals further react with O atoms to produce CO2 molecules and respective
remainder radicals, i.e. the model includes the processes resulting in etching of PP surface (no 41
in Table 10.5).
H O H
* =

-
-

-
~ C C + O => ~ C + CO2

*
-

CH3
-
CH3

Carbonyl radicals can also react with OH to form acids ~(OH)C=O (no 42 in Table 10.5).
H O H O
* =

- - - =
- -

- -

~ C C + OH => ~ C C

CH3 CH3 OH

Along with formation of ketones and aldehydes, backbone scission reactions lead also to the
appearance of remainder radicals 1R* and R2*. Processes with these radicals are not considered in
[66].
Rate constants, frequencies and probabilities for the reactions discussed above were taken the
same as in [66, 67]. It should be also noted that in our previous paper [70] the only these reactions
were taken into account in the kinetic model.
In present paper a number of processes are included in the model in addition to reaction
mechanism presented in [66, 67, 70]. According to results of discharge modeling, in humid air
plasma a noticeable amount of HO2 radicals is formed and these radicals can react with peroxy and
alkoxy radicals to form hydroperoxides and alcohols, respectively [69], (nos 43-48 in Table 10.5):
PPOO* + HO2 => PPOOH + O2 (6)
PPO* + HO2 => PPOH + O2 (7)
To estimate the probabilities of these reactions we have looked for analogous gas phase chemical
reactions. It appeared that the rate constants for the gas phase reactions like
HO2 + CH3C(O)CH2OO* => CH3C(O)CH2OOH + O2
HO2 + (CH3)3CCH2OO* => (CH3)3CCH2OOH + O2
are about 10-11 cm3/s [71], so the probability for the reaction (1) is expected to be high. The value of
this probability was estimated in the following manner [66]. It was assumed that a unit surface
reaction probability corresponds to a gas phase kinetic rate constant ( 10-10 cm3s-1). Then the
surface reaction probability for the reaction (6) can be as high as 0.1. In our model correspondent
probability was taken as 0.01. Unfortunately, we have not found analogous gas phase chemical
reactions for the process (7), so the correspondent probability was taken the same as for the process
(6). It should be noted that in paper [68] the probability for the reaction (6) is taken as 10 -5 and
probably is underestimated. The authors did not explain the reason of such a choice.
Termination reactions also occur between OH gas radical and alkoxy and peroxy surface
radicals [69] (nos 49-54 in Table 10.5):
PPOO* + OH => PPOH + O2 (8)
PPO* + OH => PPOH (9)
The probabilities of these processes were taken as 0.01 and 0.001 respectively.
As it is pointed out in paper [69], the termination of alkoxy and peroxy radicals can occur by
bimolecular recombination (nos 55-75 in Table 10.5):
PPO* + PPO* => PPOOPP (10)
PPO* + PPOO* => PPOPP + O2 (11)
PPOO* + PPOO* => PPOOPP + O2 (or PPO* + PPO* + O2 ) (12)
As for the reaction (12), for the “in chain” peroxy radicals the first reaction channel is taken into
account and for the “end free” peroxy radicals the second one is assumed (see below). The rate
constants for the processes (10)-(12) was taken as 10-15 cm2/s.
If the polymer peroxy and/or alkoxy radicals are in neighbouring positions, they recombine to
form stable cyclic peroxides or expoxides [69]. For example:
*

O O
-

H O O H H O - O H
-

~ C - C - C - C ~ => ~ C - C - C - C ~ + O2
-

H CH3 H CH3 H CH3 H CH3

*

O
*
-

H O O H H O H
-

~ C - C - C - C ~ => ~ C - C - C - C ~ + O2
-

H CH3 H CH3 H CH3 H CH3

Reactions between peroxy and/or alkoxy radicals in different polymer chains cause cross
linking. Let us also note that reactions between two peroxy radicals and between peroxy radicals
and HO2 gas radicals can also lead to formation of C=O groups [69]. At present these processes are
not included in the kinetic model.
Processes with the participation of “end free” radicals are also included in the model. These
reactions are similar to that for “in chain” radicals. The following processes are taken into account.
Formation of alkoxy and peroxy radicals in reactions with active plasma species (nos 76-81 in
Table 10.5):
R* + O => RO* (13)
R* + O2 => ROO* (14)
R* + O3 => RO* + O2 (15)
Formation of alcohols and hydroperoxides in reactions with active plasma species (nos 82-91 in
Table 10.5):
R* + OH => ROH (16)
RO* + OH => ROOH (17)
RO* + HO2 => ROH + O2 (18)
ROO* + OH => ROH + O2 (19)
ROO* + HO2 => ROOH + O2 (20)
Decay of “end free” peroxy radicals in reactions with NO molecules (nos 92-93 in Table 10.5):
ROO* + NO => RO* + NO2 (21)
Abstraction of H atoms from polymer surface by “end free” alkoxy and peroxy radicals (nos 94-105
in Table 10.5):
RO* + PP => ROH + PP* (22)
ROO* + PP => ROOH + PP* (23)
Recombination reactions between “end free” and “in chain” alkoxy and peroxy radicals, these
reactions lead to cross linking (nos 106-123 in Table 10.5):
ROO* + PPOO* => ROOPP + O2 (24)
ROO* + PPO* => ROPP + O2 (25)
RO* + PPOO* => ROPP + O2 (26)
RO* + PPO* => ROOPP + O2 (27)
Recombination reactions between “end free” peroxy radicals (nos 124-126 in Table 10.5):
ROO* + ROO* => RO* + RO* + O2 (28)
Corresponding probabilities and rate constants were partially taken from [69] and partially were
assigned by ourselves by analogy with “in chain” radical reactions. It was assumed that the typical
value of rate constant is 10-15 cm2/s.
Alkyl radicals can react with each other to crosslink. We included in the model crosslink processes
for the “end free” alkyl radicals (nos 127-129 in Table 10.5) with an estimated rate of 10-15 cm2/s
[68].
R* + R* => RR (29)
It should be noted that the role of these processes under considered conditions (atmospheric
pressure air discharge) is minor and the density of cross-linked alkyl radicals is small, since the rate
of the alkyl radicals decay due to reactions with active plasma species is essentially higher than the
rate of cross linking.
Evolution of surface densities of different functional groups (including H-sites) with treatment
time was calculated by numerical solving of system of balance equations [70]:

, i = 1, 2, 3 …M, (30)

where Ni [cm-2] is the surface concentration of i-th functional group and M is the number of
functional groups considered in the model. The terms in the right hand side of equation (30)
describe the total rates of production and loss of the i-th functional group.
The kinetic model described includes reactions of three different types: (I) reactions between
plasma species and surface species (including etching); (II) reactions between different surface
species and (III) chain backbone scission reactions due to interaction of surface species with
polymer backbone. The explicit expressions for the corresponding terms on the right hand side of
(30) depend on the type of reaction. Let us illustrate the procedure by particular examples.
(I). Abstraction of H atoms from H-sites by OH radicals:

H-site (surface) + OH (gas) = alkyl radical (surface) + H2O (gas). (31)

Partial production rate of alkyl radicals (equal to the decay rate of H-sites) in reaction (26) is
expressed as
Qprod(alkyl rad) = Qloss(H-sites) = JOH NH /K, (32)

where NH is the surface concentration of H-sites, K is the total surface density of H-sites on a virgin
polymer surface, is the reaction probability per unit area and JOH is the flux of OH radicals to the
polymer surface. The flux of OH radicals is estimated as
JOH = nOHOH/4, (33)
where nOHis the concentration of OH radicals in plasma near the surface and vOH is the thermal
velocity of OH radicals.
(II) Abstraction of H atoms from H-sites by alkoxy radicals:

Alkoxy radical(surface) + H-site(surface) = Alcohol group (surface) + Alkyl radical (surface). (34)
 Partial production rates of alcohol groups and alkyl radicals (and the decay rates of H-sites and
alkoxy radicals) for this process are calculated as
Qprod(alcohol)=Qprod(alkyl rad)=Qloss(alkoxy rad)=Qloss(H-sites) = Nalkoxy kNH , (35)

where Nalkoxy is the surface concentration of alkoxy radicals and k is the rate constant of reaction
(34).

(III) Chain backbone scission due to interaction of alkoxy radicals with polymer backbone:

Alkoxy radical (surface)+Backbone = Aldehyde (surface) + Backbone scission radical. (36)

Partial production rates of aldehydes and backbone scission radicals (and decay rate of alkoxy
radicals) for this process are calculated as
Qprod (aldehyde)= Qprod (scission)= Qloss(alkoxy rad)=NAlkoxy -1, (37)

where  is the characteristic time of reaction.

10.4.2 Results of modeling and comparison with experimental data

Conditions under consideration are similar to that in [70]. The initial surface densities of all the
functional groups (except H-sites) were taken to be zero. To specify the initial surface concentration
of H-sites the surface density of C atoms (the number of C atoms in PP monolayer per cm2) was
first estimated. It was done by using known density of PP polymer ( 0.91 g/cm) and assuming that
a thickness of PP monolayer is 0.5 nm. According to our estimations the surface density of C atoms
is about 2. 1015 cm-2. Respectively, the initial surface density of H-sites was chosen as 2. 1015 cm-2,
the surface densities of different kinds of sites were taken in proportion
[primary]/[secondary]/[tertiary] = 0.25/0.5/0.25 similar to that in [66].
The degree of the surface modification depends on the concentrations of active species in the
gas phase and on the time of processing. In our calculations concentrations of active species (O,
OH, HO2, H, O2 and O3) in the gas phase were considered as time-independent parameters.
Gas phase species concentrations were taken from glow discharge modeling within frames of
the model described in [10, 91]: [O] = 8 1012 cm-3, [O2] = 5 1018 cm-3, [O3] = 1.8 1014 cm-3, [OH] =
1.8 1012 cm-3, [H] = 4 108 cm-3, [NO] = 1.3 1013 cm-3, [HO2] = 1.1013 cm-3. Gas temperature and
pressure were 300 K and 1 atmosphere, respectively.
It simulations it was assumed that PP film is treated during 3 s and then is stored at ambient air.
Calculations were performed from 0 to 10 s to include post discharge period. Results of calculations
are shown in Figures 10.17-10.20.
Let us briefly discuss the general features of the time variation of surface concentrations of
different groups and radicals during treatment procedure and post treatment period. The decay of H-
sites densities under considered conditions is mainly due to the abstraction of H atoms by OH
radicals. The characteristic decay time for 3PP and for 2PP during treatment time (3 s) is shorter
than that for 1PP, and concentrations of these sites at the end of the time interval become very small
with respect to their initial values (see Figure 10.17a). The density of 1PP is decreased after the end
of treatment, this is due to the abstraction of H atoms by alkoxy and peroxy radicals formed during
treatment.

a b

Figure 10.17a–b Calculated surface concentrations:a of H-sites as functions of time; b of alkyl

radicals and cross links as functions of time. Treatment time is 3 s.

The abstraction of H atoms from H-sites leads to the formation of “in chain” alkyl radicals,
surface concentrations of which get maximum (~ 1011 cm-2) at times about 10-3 s (Figure 10.17b). At
later times these radicals are consumed by the reactions with O2 and O3 to form peroxy and alcoxy
radicals. As the free sites 2PP and 3PP on the PP surface are depleted, the densities of the alkyl
radicals decrease. The maximum surface density of the “end free” alkyl radicals (appeared due to
the bond scission processes) is about ~ 1010 cm-2, these radicals are also consumed by the reactions
with O2 and O3 to form “end free” peroxy and alcoxy radicals. After the end of treatment, the main
source of alkyl radical production is diminished (the abstraction of H atoms by OH radicals and
formation of “in chain” peroxy radicals, which are responsible for the bond scission processes) and
surface concentration of these radicals is decreased. At later times “in chain” alkyl radicals are
formed due to the abstraction of H atoms by alkoxy and peroxy radicals stored during treatment.
“End free” alkyl radicals are also formed, since there is some amount of “in chain” alkoxy radicals,
which produce bond scissions.
Surface concentrations of alkoxy and peroxy radicals (see Figure 10.18a) are not very high (in
total ~3. 104 cm-2 at t = 3s). Under discharge conditions the decay rates of these radicals in reactions
with OH and HO2 plasma radicals are rather high. After the treatment the decay is provided by
recombination reactions. Concentrations of aldehydes, acids and carbonyl radicals are small and
concentration of ketones is about ~ 3.1014 cm-2 (see Figure 10.18b).

a b

Figure 10.18a–b Calculated surface concentrations: a of alkoxy and peroxy radicals as functions
of time; b of different groups and radicals with double bonded (=O) oxygen atoms as functions of
time. Treatment time is 3 s. t

It can be seen from Figure 10.19a that the major amount of O atoms is introduced onto the PP
surface in the form of hydroperoxides and alcohols, the concentration of “in chain” groups is higher
than the concentration of “end free” groups. The total concentration of the products of
recombination (see Figure 10.19b) is about 2.3 1014 cm-2. ROOPP and ROPP groups (in total ~ 1.2
1014 cm-2 ) as well as some part of PPOOPP and PPOPP groups are the cross links. Let us note that
the concentration of cross linked “end free” alkyl radicals is very small and is not shown in the
figures.

a b
Figure 10.19a–b Calculated surface concentrations: a of alcohols and hydroperoxides as functions
of time; b of products of alkoxy and peroxy radicals recombination as functions of time. Treatment
time is 3 s.

The total concentration of O atoms incorporated onto PP surface is 3. 1015 cm-2, the
concentration of “in chain” O atoms is 1.6 times higher then the concentration of “end free” atoms
(see Figure 10.20a).

a b

Figure 10.20a–b Calculated surface concentrations: a of O atoms as functions of time; b of bond

scissions and number of etched C atoms as functions of time. Treatment time is 3 s.

According to our calculations (see Figure 10.20b) at t = 3 s the number of C atoms etched from
1 cm2 is about 2.5 1014, i.e. about 13% of the upper PP layer is etched. Besides, there appeared a
large number of bond scissions. From the standpoint of experiments it means that the upper layer of
PP polymer contains a lot of low molecular weight moieties. In fact, such a situation responds to
over-treatment, since these moieties can be easily removed from the surface, for example, by a
mechanical means. In general, the treatment time should be long enough to provide necessary
degree of oxidation but it is restricted by noticeable destruction of the upper layer of polymer. It
follows from our studies that under considered conditions acceptable treatment time is about 1 s.
Now let us compare results of calculations with available experimental data [70]. It follows from
Figure 10.20a that the characteristic time for the surface concentration of O atoms introduced into
PP surface is about 12 s, this value is in a reasonable agreement with that estimated from the
results of contact angle measurements depending on treatment time [70].
We can also compare calculated and measured O/C ratios. According to the XPS analysis under
experimental conditions [70], O/C ratio (at t = 3 s) is about 0.25. The characteristic depth where
from the signal was collected was about L0 ~7.5 nm, this depth corresponds to 15 PP atomic layers.
 It is possible to estimate the value of O/C ratio, which is found from XPS analyses, provided
that concentration of O atoms introduced in the surface of PP polymer is known [70]. To do this, we
have assumed that XPS signal intensity from C atoms located at different depths exponentially
decreases with depth with a characteristic length . Then the signal from i-th layer (i = 0, 1, 2, …; i
= 0 denotes the upper layer) of C atoms can be calculated as
SCi = kXkC[C1]exp(-i/), (38)
where [C1] =2. 1015 cm-2 is the surface density of C atoms in a layer,  is the thickness of PP
monolayer (0.5 nm), kX is a numerical factor proportional to the power of the X-ray source and kC is
a numerical factor proportional to the probability of the electron emission from C atoms.
The total signal from C atoms is calculated as

. (39)

In our model O atoms are introduced only in the upper layer of PP polymer. For this reason, the
XPS signal intensity from O atoms can be estimated as
SO = kXkO[O], (40)
where [O] is the surface concentration of O atoms and kO is a numerical factor proportional to the
probability of the electron emission from O atoms.
The O/C ratio measured in experiments is the ratio of signals from O and C atoms
normalized to corresponding probabilities of the electron emission. Respectively, O/C ratio is
calculated as

(41)

To estimate the value of  let us note that in experiments the depth of analysis was about L0=7.5
nm [66], i.e. signal was collected from 15 PP atomic layers. The contribution from deeper layers is
neglected. Therefore it is reasonably to assume that  = L0/3 = 2.5 nm, since in this case 15 PP
atomic layers provide about 95% of the total signal calculated from equation (39).
For given values of  and  (and taking into account only 15 PP atomic layers) equation (41)
can be rewritten as

, (42)

According to our calculations, at t = 3 s the total surface concentration of O atoms is about 3.

1015 cm-2. Corresponding O/C ratio estimated by formula (42) is about 0.29, it is in agreement with
our experimental data 0.25 reported in [70].
CONCLUSIONS
Non-thermal plasma produced by gas discharges at atmospheric pressure is an attractive tool for
polymer surface treatments for both increasing and decreasing wettability of the surface. Despite
numerous successful demonstrations of polymer surface improvement by this technology the
quantitative description of processes taking place on plasma-surface boundary is still far from being
satisfactory. To achieve better understanding of these processes one has to develop an extended
kinetic model of non-thermal gas discharges in various environments and a kinetic model for
interface chemical processes taking into account fluxes of radicals and excited species on the
surface. In this chapter the state-of-the-art situation with theoretical modeling is described and the
whole problem is illustrated by example of polypropylene surface treatment by non-thermal plasma
at atmospheric pressure. A satisfactory agreement is demonstrated with respect to experimental data
on O/C ratio measured by the XPS method.

REFERENCES
1. Kunhardt, E. E. (2000) Generation of Large-Volume, Atmospheric-Pressure, Nonequilibrium
Plasmas. IEEE Transactions on Plasma Science, 28 (1), 189-200.
2. Chang, J.-S., Lawless, P. A. and Yamamoto, T. (1991) Corona discharge processes. IEEE Trans.
Plasma Science, 19 (6), 1152-1166.
3. Napartovich, A. (2001) Overview of atmospheric pressure discharges producing non- thermal
plasma. Plasmas and Polymers, 6 (1), 1-14.
4. Becker, K., Barker,R., Kogelschatz, U. and Schoenbach, K.(eds) (2004) Non-Equilibrium Air
Plasmas at Atmospheric Pressure, IOP Publ., Bristol.
5. Conrads, H. and Schmidt, M. (2000) Plasma generation and plasma sources. Plasma Sources Sci.
Technol., 9 (4), 441–454.
6. Kogelschatz, U. (2002) Industrial innovation based on fundamental physics. Plasma Sources Sci.
Technol., 11 (3A), A1–A6.
7. Fridman, A., Chirokov, A. and Gutsol, A. (2005) Non-thermal atmospheric pressure
discharges. J. Phys. D: Appl. Phys., 38 (2), R1–R24.
8. Akishev, Yu. S., Grushin, M. E., Napartovich, A. P. and Trushkin, N. I. (2002) Novel AC and
DC Non-Thermal Plasma Sources for Cold Surface Treatment of Polymer Films and Fabrics at
Atmospheric Pressure. Plasmas and Polymers, 7 (3), 261-289.
9. Akishev, Yu. S., Aponin, G. I., Grushin, M. E., Karal’nik, V. B., Pan’kin, M. V., Petryakov, A.
V. and Trushkin, N. I., (2008) Alternating Nonsteady Gas-Discharge Modes in an Atmospheric-
Pressure Air Flow Blown through a Point-Plane Gap. Plasma Physics Reports, 34 (4), 312–324.
10. Akishev, Yu. S., Deryugin, A. A., Elkin, N. N., Kochetov, I. V., Napartovich, A. P., and
Trushkin, N. I. (1994) Calculation of Air Glow Discharge Spatial Structure. Plasma Physics
Reports, 20 (5), 437-441.
11. Akishev, Yu. S., Deryugin, A. A., Karal’nik, V. B., Kochetov, I. V., Napartovich, A. P. and
Trushkin, N. I. (1994) Numerical Simulation and Experimental Study of a Atmospheric-Pressure
Direct-Current Glow Discharge. Plasma Physics Reports, 20 (6), 511-524.
12. Akishev, Yu. S., Deryugin, A. A., Kochetov, I. V., Napartovich, A. P. and Trushkin, N. I.
(1994) Generation Efficiency of Chemically -Active Particles in Self-Maintained Discharge.
Plasma Physics Report, 20 (6), 525-532.
13. Napartovich, A. P., Akishev, Yu. S., Deryugin, A. A., Kochetov, I. V. and Trushkin, N. I.
(1993) Glow Discharge with fast Gas Flow for Flue Gas Processing, in Non-Thermal Plasma
Techniques for Pollution Control, (eds B. M. Penetrante and S. E. Schultheis), Springer-Verlag,
NATO ASI Series G 34, Part B, pp. 355-370.
14. Akishev, Yu. S., Deryugin, A. A., Kochetov, I. V., Napartovich, A. P. and Trushkin, N. I.
(1993) DC glow discharge in air flow at atmospheric pressure in connection with waste gases
treatment. J. Phys. D: Appl. Phys., 26 (10), 1630-1637.
15. Stefanović, I., Bibinov, N. K., Deryugin, A. A., Vinogradov, I. P., Napartovich, A. P. and
Wiesemann, K. (2001) Kinetics of ozone and nitric oxides in dielectric barrier discharges in O2/NOx
and N2/O2/NOx mixtures. Plasma Sources Sci. Technol., 10 (3), 406-416.
16. Gibalov, V. I. and Pietsch, G. J. (2000) The development of dielectric barrier discharges in gas
gaps and on surfaces. J. Phys. D: Appl. Phys., 33 (20), 2618-2636.
17. Eliasson, B. and Kogelschatz, U. (1991) Modeling and applications of silent discharge plasmas.
IEEE Trans. Plasma Sci., 19 (2), 309-323.
18. Steinle, G., Neundorf, D., Hiller, W. and Petralla, M. (1999) Two-dimensional simulation of
filaments in barrier discharges. J. Phys. D: Appl. Phys., 32 (12), 1350-1356.
19. Babaeva, N. Yu. and Naidis, G. V. (2000) Modeling of streamer propagation, in Electrical
Discharges for Environmental Purposes: Fundamentals and Applications, (ed E. M. van
Veldhuizen), Nova Science Publishers, Inc., New York, pp. 21-48.
20. Kulikovsky, A. A. (1994) The structure of streamers in N2. I. fast method of space-charge
dominated plasma simulation. J. Phys. D: Appl. Phys., 27 (12), 2556-2563.
21. Kulikovsky, A. A. (1994) The structure of streamers in N2. II. Two- dimensional simulation. J.
Phys. D: Appl. Phys., 27 (12), 2564-2569.
22. Kulikovsky, A. A. (1997) Positive streamer between parallel plate electrodes in atmospheric
pressure air. J. Phys. D: Appl. Phys., 30 (3), 441-450.
23. Kulikovsky, A. A. (1997) The mechanism of positive streamer acceleration and expansion in air
in a strong external field. J. Phys. D: Appl. Phys., 30 (10), 1515-1522.
24. Pancheshnyi, S. V., Starikovskaya, S. M. and Starikovskii, A. Yu. (2001) Role of
photoionization processes in propagation of cathode-directed streamer. J. Phys. D: Appl. Phys., 34
(1), 105-115.
25. Kushner, M. J. (2005) Modelling of microdischarge devices: plasma and gas dynamics. J.
Phys. D: Appl. Phys., 38 (11), 1633-1643.
26. Trichel, G. W. (1938) The mechanism of the negative Point to Plane Corona near Onset. Phys.
Rev., 54 (12), 1078-1084.
27. Akishev, Yu. S., Kochetov, I. V., Loboiko, A. I. and Napartovich, A. P. (2002) Numerical
Simulations of Trichel Pulses in a Negative Corona in Air. Plasma Physics Reports, 28 (12), 1049-
1059.
28. Napartovich, A. P., Akishev, Yu. S., Deryugin, A. A., Kochetov, I. V., Pan’kin, M. V. and
Trushkin, N. I. (1997) A numerical simulation of Trichel-pulse formation in a negative corona. J.
Phys. D: Appl. Phys., 30 (19), 2726-2736.
29. Akishev, Yu. S., Grushin, M. E., Kochetov, I. V., Napartovich, A. P. and Trushkin, N I. (1999)
An Establishment of Regular Trichel Pulses in a Negative Corona in Air. Plasma Physics Reports,
25 (11), 922-927.
30. Akishev, Yu. S, Grushin, M. E, Deryugin, A. A, Napartovich, A. P, Pan’kin, M. V. and
Trushkin, N. I. (1999) Self-oscillations of a positive corona in nitrogen. J. Phys. D: Appl. Phys. 32
(18), 2399-2409.
31. Akishev, Yu. S., Grushin, M. E., Kochetov, I. V., Napartovich, A. P., Pan’kin, M. V. and
Trushkin, N. I. (2000) Transition of a Multipin Negative Corona in Atmospheric Air to a Glow
Discharge. Plasma Physics Reports, 26 (2), 157-163.
32. Itikava, Y., Hayashi, M., Ichimura, A., Onda, K., Sakimoto, K., Takanayagi, K., Nakamura, M.,
Nishimura, H. and Takanayanagim, T. (1986) Cross Sections for Collisions of Electrons and
Photons with Nitrogen Molecules. Journal of Physical and Chemical Reference Data, 15 (3), 985-
1010.
33. Phelps, A. V. and Pitchford, L. C. (1985), 26 th report. JILA Information Center Report,
Boulder.
34. Pack, J. L., Voshall, R. E., and Phelps, A. V. 1962, Drift Velocities of Slow Electrons in
Krypton, Xenon, Deuterium, Carbon Monoxide, Carbon Dioxide, Water Vapor, Nitrous Oxide, and
Ammonia. Phys. Rev., 127 (6), 2084-2089.
35. Eliasson, B. and Kogelschatz, U. (1986) Basic Data for Modelling of Electric Discharge in
Gases: Oxygen. Asea Brown Boweri Forschungszntrum CH-5405, Baden, KLR 86-11C.
36. Rapp, D. and Englander-Golden, P. (1965) Total cross sections for ionization and attachment in
gases by electron impact. I. Positive ionization. J. Chem. Phys., 43 (12), 1464-1479.
37. Yousfi, M., Azzi, N., Segur, P., Gallimberti I. and Stangherlin I. (1987) Electron-molecule
collision cross sections and electron swarm parameters in some atmospheric gases (N 2, O2, CO2 and
H2O), Centre de Physique Atomique de Toulouse & Istituto di Elettrotecnica ed Elettronical
Universita di Padova, Toulouse Padova.
38. Ryzko, H. (1966) Ionization, attachment and drift velocity of electrons in water vapor and dry
air. Arkiv foer Fysik, 32 (1), 1-18.
39. Prasad, A. N. and Craggs, J. D. (1960) Measurement of Ionization and Attachment Coefficients
in Humid Air in Uniform Fields and the Mechanism of Breakdown. Proc. Phys. Soc. London, 76
(2), 223-232.
40. Risbud, A. V. and Naidu, M. S. (1979) Ionization and attachment in water-vapor and ammonia.
Journal de Physique, 40 (7), 77-78.
41. Crompton, R. W., Rees, J. A. and Jory, R. L. (1965) The Diffusion and Attachment of Electrons
in Water Vapour. Aust. J. Phys., 18 (6), 541-551.
42. Parr, J. E. and Moruzzi, J. L. (1972) Electron attachment in water vapour and ammonia. J. Phys.
D: Appl. Phys., 5 (3), 514-524.
43. Chanin, L. M., Phelps, A. V. and Biondi, M. A. (1962) Measurements of the Attachment of
Low-Energy Electrons to Oxygen Molecules. Phys. Rev., 128 (1), 219–230.
44. Rees, J. A. (1965) Behavior of free and attached electrons in oxygen. Aust. J. Phys., 18 (1), 41-
57.
45. Huxley, L. G. H., Crompton, R. W. and Bagot, C. H. (1959) A New Method for Measuring the
Attachment of Slow Electrons in Gases. Australian J. Phys., 12 (3), 303-308.
46. Chatterton, P. A. and Craggs, J. D. (1961) J. Electron. Control, 11, 425-437.
47. Doehring, A. (1952) Measurement of the probability of addition of electrons to oxygen by a
time-of-transit method. Zeitschrift fuer Naturforschung, 7A, 253-270.
48. Kossyi, I. A., Kostinskii, A. Y., Matveev, A. A. and Silakov, V. P. (1992) Kinetic scheme of the
non-equilibrium discharge in nitrogen-oxygen mixtures. Plasma Sources Sci. Techn., 1 (3), 207-
220.
49. Capitelli, M., Ferreira, C. M., Gordiets, B. F. and Osipov A. I. (2000) Plasma kinetics in
atmospheric gases, Springer-Verlag, Berlin.
50. Ionin, A. A., Kochetov, I. V., Napartovich, A. P. and Yuryshev, N. N. (2007) Physics and
engineering of singlet delta oxygen production in low-temperature plasma. J. Phys. D: Appl. Phys.,
40 (2), R25-R61.
51. Maetzing, H. (1991) Chemical Kinetics of Flue Gas Cleaning by Irradiation with Electrons, in
Adv. in Chemical Phys., 80, (eds I. Prigogine and S.A. Rice), John Wiley& Sons, pp.315-402.
52. Zakharov, A. I., Klopovskii, K. S., Osipov, A. P., Popov, A. M., Popovicheva, O. B.,
Rakhimova, T. V., Samorodov, V. A. and Sokolov, A. P. (1988) Kinetics of Processes Excited by a
Self-Sustaining Volume Discharge in Oxygen. Plasma Physics Reports, 14 (3), 191-195.
53. Mukkavilli, S., Lee, C.K., Varghese, K. and Tavlarides, L.L. (1988) Modeling of the
electrostatic corona discharge reactor. IEEE Trans. Plasma Sci., 16 (6), 652-660.
54. Mc Ewan, M. J. and Phillips, L. F. (1975) Chemistry of the Atmosphere, Edward Arnold Ltd.,
London.
55. Atkinson,R, Baulch, D. L., Cox, R.A., Hampson, R.F., Kerr, J.A. and Troe, J. (1989) Evaluated
kinetic and photochemical data for atmospheric chemistry: Supplement III - IUPAC Subcommittee
on Gas Kinetic Data Evaluation for Atmospheric Chemistry. Journal of Physical and Chemical
Reference Data, 18 (2), pp.881-1097.
56. Person, J.C. and Ham, D.O. (1988) Removal of SO2 and NOx from stack gases by electron beam
irradiation. Radiat. Phys. Chem., 31 (1-3), 1-8.
57. Feoktistov, V. A., Lopaev, D. V., Klopovsky, K. S., Popov, A. M., Popovicheva, O. B.,
Rakhimov, A. T. and Rakhimova, T. V. (1993) Low pressure RF discharge in electronegative gases
for plasma processing. J. Nucl. Materials, 200 (3), p. 309-314.
58. Fraser, M.E. and Piper, L.G. (1989) Product branching ratios from the N2(A3Σu+) + O2
interaction. J. Phys. Chem., 93 (3), p.1107-1111.
59. Slovetskii, D. I. (1980) Mechanisms of chemical reactions in non-equilibrium plasma, Nauka,
Moscow.
60. Gentile, A. C. and Kushner, M. J. (1995) Reaction chemistry and optimization of plasma
remediation of NxOy from gas streams. J. Appl. Phys., 78 (3), 2074-2085.
61. De Geyter, N., Morent, R., Leys, C., Gengembre, L. and Payen, E. (2007) Treatment of polymer
films with a dielectric barrier discharge in air, helium and argon at medium pressure. Surface &
Coatings Technology, 201 (16-17), 7066-7075.
62. Hubenák, J. and Krcma, F. (2000) Determination of a hydrocarbon concentration in the N2 dc
flowing afterglow and its application. J. Phys. D: Appl. Phys., 33 (23), 3121-3128.
63. Pintassilgo, C. D., Cernogora, G. and Loureiro, J. (2001) Spectroscopy study and modelling of
an afterglow created by a low-pressure pulsed discharge in N2-CH4. Plasma Sources Sci. Technol.,
10 (2), 147-161.
64. Dilecce, G., Simek, M. and De Benedictis, S. (2001) The N2(A3Σu+) energy transfer to OH(A2Σ+)
in low-pressure pulsed RF discharges. J. Phys. D: Appl. Phys., 34 (12), 1799-1806.
65. Fresnet, F., Baravian, G., Magne, L., Pasquiers, S., Postel, C., Puech, V. and Rousseau, A.
(2002) Influence of water on NO removal by pulsed discharge in N2/H2O/NO mixtures. Plasma
Sources Sci. Technol., 11 (2), 152-160.
66. Dorai, R. and Kushner, M. J. (2003) A model for plasma modification of polypropylene using
atmospheric pressure discharges. J. Phys. D: Appl. Phys., 36 (6), 666–685.
67. Dorai, R. and Kushner, M. J. http://www.ee.ualberta.ca/icops2002 (30 January 2009).
68. Bhoj, A. N. and Kushner, M. J. (2007) Continuous processing of polymers in repetitively pulsed
atmospheric pressure discharges with moving surfaces and gas flow. J. Phys. D: Appl. Phys., 40
(22), 6953-6968.
69. Rabek, J. F. (1995) Polymer Photodegradation: Mechanisms and Experimental Methods, 1st
edn, Chapman and Hall, London.
70. Akishev, Yu., Grushin, M., Dyatko, N., Kochetov, I., Napartovich, A.,
Trushkin, N., Tran, Minh Duc and Descours, S. (2008) Studies on cold plasma-polymer
surface interaction by example of PP- and PET-films. J. Phys. D: Appl. Phys., 41 (23), 1-13.
71. ANSI 17-2Q98. (1998) Standard Reference Database, American National Standards Institute,
New York. http://kinetics.nist.gov/kinetics/ (9 February 2009).