Component Content of Active Particles in Plasma-Chemical Reactor Based on a Volume Barrier Discharge
I.A.Soloshenko1, V.V.Tsiolko1, S.S.Pogulay1, A.G.Terent’eva1, V.Yu.Bazhenov1,
A.I.Shchedrin1, A.V.Ryabtsev1, A.I.Kuzmichev2
1Institute of Physics of National Academy of Sciences of Ukraine, 46 Nauki Ave., 03028, Kiev, Ukraine,
2National Technical University “KPI”, 37 Peremogy Ave., KPI-2230, 03056, Kiev, Ukraine
Abstract. In this proceeding the results of theoretical and experimental studies of the component content of active
particles formed in plasma-chemical reactor composed of multiple-cell generator of active particles based on volume
barrier discharge, and working chamber are presented. For calculation of the content of uncharged plasma components an
approach is proposed which is based on averaging of introduced power over the entire volume. Advantages of such
approach consist in absence of fitting parameters, such as dimensions of microdischarges, their surface density, and rate
of breakdowns. The calculation and the experiment were accomplished with the use of dry air (20% relative humidity) as
plasma generating medium. Concentrations of O3, HNO3, HNO2, N2O5, and NO3 were measured experimentally in the
discharge volume and working chamber for transient time of the particles on the discharge of 0.3 s and more, and the
discharge specific power of 1.5 W/cm3. It has been determined that the best agreement between the calculation and the
experiment occurs at calculated gas medium temperature in the discharge plasma of about 400÷425 K, which corresponds
to experimentally measured rotational temperature of nitrogen. In the most cases calculated concentrations of O3, HNO3,
HNO2, N2O5, and NO3 for barrier discharge and working chamber are in a good agreement with respective measured
Keywords: barrier discharge, microdischarge, component content.
In the last decade non-thermal discharges at one atmosphere pressure attain more and more applications in
technology. Particularly, possibilities of the use of corona and barrier discharges were demonstrated for efficient
removal of nitrogen oxides NOx and NxOy from industrial gases, sterilization of medical articles, modification of
surface features of polymer materials and cleaning the surfaces contaminated in result of chemical weapon action.
For solving the last three tasks, the most optimum is the use of plasma-chemical reactor composed of two parts –
1) generator of active particles with one ore several gas discharge cells; 2) working chamber having relatively large
volume for placement of processed articles or materials. Such design allows minimum influence of processed
articles on the discharge and, consequently, on the component content of active particles in it.
Purpose of the present work consists in theoretical and experimental study of the component content of active
particles and its dependence on different factors of the concentrations, both in volume barrier discharge, and in
working chamber of the reactor developed by us, with the use of ambient air with ≈ 20% relative humidity as
EXPERIMENTAL SETUP AND METHODS OF MEASUREMENTS
The plasma-chemical reactor consisted from working chamber made of polymethylmethacrylate having 80 liters
volume (430x430x430 mm) and sixteen volume barrier discharge cells which were evenly located at top wall of the
chamber. Active particles formed in the discharges were supplied to the chamber, and after that left the chamber by
passing to deactivation system via opening located at the chamber bottom. Air was supplied to the discharge cells
through wetting/desiccation system which enabled relative humidity (RH) variations in range of 20-90% at 20-220C
temperature. Volume pumping rate through each of the discharge gaps was varied from 1 cm3/s up to 8 cm3/s (at that
volume rates of pumping through the chamber comprised 1÷8 liters per minute). These volume rates corresponded to
mean transient time values in the discharge gap τ = 2.4÷0.3 s (Mean transient time of the particles in the discharge is
τ = V/2υ, where V is the discharge volume). Alternating current source with voltage up to 15 kV and 400 Hz
frequency was used for powering of the discharges.
NO3 concentration was calculated on a basis of the lamp radiation absorption at wavelength λ = 662 nm and
623 nm. Densities of O3, HNO3, HNO2, N2O5, H2O2 were calculated with the use of curve of integral absorption by
these particles in 200÷300 nm wavelength range by means of automated fitting routine. The routine performed
selection of concentration values for the particles until reaching coincidence of experimental and calculated curves
with pre-determined precision. Possibility of use of such method is based on fact that in 200÷300 nm wavelength
range spectrum dependencies of absorption cross sections for these particles are essentially different. At calculations
of concentrations of the particles, cross section values taken from  were used.
RESULTS OF NUMERICAL MODELING OF COMPONENT CONTENT OF THE
PARTICLES IN BARRIER DISCHARGE VOLUME AND IN WORKING CHAMBER
As it is known, barrier discharge represents an assembly of filamentary microdischarges each having
≈ 10÷100 ns duration and ~ 0.1 mm diameter which are stochastically spread in time and in the discharge volume.
Commonly  at determining concentrations of particles in the discharge, plasma kinetics in separate current
channels of the microdischarges is calculated at first, and after that in time of an order of diffusion one (~ 10-3 s)
averaging of concentrations of all components over the entire discharge volume is performed. With such approach
there exists a set of parameters that are poorly
known and essentially depend on design of the
discharge cell and kind of gas, such as rates of
occurrence of current channels, their dimensions
and surface density at the discharge electrodes.
Those parameters are commonly fitting ones.
Calculation presented in the present paper is
based on another approach, at which power
introduced into the discharge is immediately
averaged over the discharge volume. With such
approach, correct description is provided for
processes with linear dependence on electron
density, as well as for nonlinear ones with typical
time of reaction being longer than the diffusion
time (t > 10-3 s). Since typical duration of
chemical reactions between dissociation products
in current channels does not exceed 10-2 s,
approach used by us is valid. Besides, an
advantage of this approach consists in the absence
of fitting parameters.
Calculations of component content of the particles in the discharge gap and working chamber were performed at
specific power in the discharge Wd=1.5 W/cm3, transient time of gas mixture in the discharge gap up to 4.8 s, and
gas mixture temperature in the discharge gap Тd = 300, 375, 400, 425 and 500 K. (In case of necessity, known
experimental temperature dependencies of reaction rates in range of 200÷450 K were extrapolated to 500 K).
Relative humidity of air supplied to the discharge gap comprised 20% at 200C. In Fig.1 calculated dependencies of
concentrations of neutral components of barrier discharge plasma on transient time of gas mixture in the discharge
gap τ are presented for gas mixture temperature Тd = 425 K. (It was assumed in the calculations that the particles are
permanently present in the discharge gap for time duration τ, and only after that are ejected outside it).
Concentrations of H, HNO and O(d) components are not shown in the figure due to their smallness (Ni ≤ 108 cm-3).
One can see from the figure that at initial stages of the discharge development (before ~ 10-4÷10-1 s) concentrations
of all components exhibit practically linear growth in time, and after that either reach saturation (H2O2, NO2, NO, N,
O, OH, HO2), or start to decrease (O3, NO3, N2O5), or continue to grow with smaller rate (N2O4, N2O, HN03, HNO2).
One can also see from the figure that: 1) maximum quasi-stationary concentration values ≈ (1÷5)⋅1016 cm-3 are
exhibited by O3, N2O, N2O4 and HNO3 components; 2) HNO2, NO and NO2 concentrations reach values of
1014÷1015 cm-3; 3) at τ > 10-2 s concentrations of majority of the components possess weak dependence on transient
FIGURE 1. Calculated dependencies of concentrations of
different neutral particles in barrier discharge volume on their
transient time in the discharge gap τ. Wd = 1.5 W/cm3.
Gas temperature Td = 425 K, air RH 20%.
time of gas mixture in the discharge; 4) only N2O4, HNO3 and N2O concentrations essentially depend on transient
time τ – at its increase from 0.3 to 2.4 s their concentrations grow up by practically an order of magnitude.
In Fig.2 calculated dependencies of concentrations of the particles in working chamber on accumulation time of
the particles tc are presented for transient time of gas
mixture in the discharge gap τ = 0.3 s, and temperature
of gas mixture in the discharge and working chamber
Td = 425 K and Tc = 300 K, respectively. First of all, it
should be noted that unlike the case of the discharge
volume, concentrations of N, O and OH components in
working chamber do not exceed 108 cm-3 (respective
curves are not shown in Fig.2). It is due to fact that the
reactions leading to formation of these particles are
practically absent in working chamber volume, and the
particles coming from barrier discharge volume
quickly “burn out” forming more stable molecules. In
time interval tc ≈ (10-8÷10) minutes, linear growth of
concentration with time is inherent for majority of the
particles, thus giving evidence to uniform in time
coming of these particles, that is, their coming to the
chamber is determined by processes of their carrying
out of the discharge volume. NO3 concentration
linearly grows for tc up to ~ 10-3 s; after that it starts its
decrease, and than after passing the minimum starts the
increase again reaching its quasi-stationary value in a
time of about 10 minutes. Temporal dependence of HO2 concentration exhibits more complex behavior – initially
linear dependence is replaced by oscillating one at t ~ 5⋅10-3 minutes. Concentrations of NO, NO2 and HNO4 reach
their maximum values in time from 0.01 to 0.5 minutes, and after that they start their decrease due to influence of
death processes. Concentrations of O3, HNO3, HNO2 and N2O5 components linearly increase up to quasi-stationary
values for 10÷20 minutes.
Comparison of Figures 1 and 2 shows that: 1) quasi-stationary values of O3, HNO3, HNO2, N2O, N2O4, H2O2 and
N2O5 concentrations in the chamber exceed the densities of these components in the discharge thus giving evidence
to their accumulation in working chamber. However, while in case of O3, HNO3, HNO2, N2O, N2O4, H2O2
components this increase comprises a factor of 1.5÷2, N2O5 density in the chamber increases by factor of ≈ 15. This
effect is, first of all, due to strong decrease of decomposition rate for this component at injection of the mixture from
the discharge with temperature Td = 425 K to the chamber with temperature Tc = 300 K; 2) on the contrary to
mentioned above components, NO3, HNO4, NO2, HO2, and NO concentrations in the chamber are less than those in
the discharge by factor of (5÷106), first of all, due to their ”burn-out”.
RESULTS OF EXPERIMENTAL STUDIES OF CONCENTRATIONS OF PARTICLES
IN DISCHARGE VOLUME AND WORKING CHAMBER
The rates of many reactions essentially depend on temperature of the particles. Thus, for correct comparison of
the results of calculations and experiments it is necessary to know gas mixture temperature. It has been shown in 
that rotational temperature of nitrogen molecules Trot is close to the temperature of translational movement. For
determining rotational temperature, 0-0 transition of SPS of nitrogen was used in the present work. Measurements of
dependency of rotational temperature on the discharge glowing time for τ = 0.3 s and Wd = 1.5 W/cm3 have shown
that Trot reaches its quasi-stationary value of 400 ± 12 K approximately at 15÷20 minute of the discharge glowing.
It should be noted that we did not succeed in measurement of N2O5 and H2O2 concentrations in the discharge gap
and H2O2 in the working chamber. It gives evidence to fact that concentrations of these particles in the discharge are
lower than sensitivity thresholds of used method of measurements which comprise ≈ 1014 and 1015 cm-3,
In Fig.3 experimentally obtained dependencies of O3, HNO3, HNO2, NO3 concentrations in the discharge gap on
the discharge glowing time td are presented for τ = 0.3 s and Wd = 1.5 W/cm3 (discharge glowing time td is one
measured from the point of turning on the discharge powering voltage till the point of its turning off). Each point at
the figure represents data averaged for 3÷4 measurements
FIGURE 2. Calculated dependencies of concentrations of
the particles in working chamber on accumulation time tc at
τ = 0.3 s, Wd = 1.5 W/cm3. Td = 425 K, Tc = 300 K; air
Comparison of Figs.1 and 3 shows that calculated values of O3 and NO3 concentrations exceed respective
experimental values by factor of about 1.5÷2. In case of HNO3 this ratio is somewhat higher and comprises about
2.5. The largest discrepancy between the
experiment and the calculation is observed for
HNO2 – calculated concentration values are less
than experimentally measured ones by more than
one order of magnitude.
The measurements of concentration of O3,
HNO3, N2O5, HNO2, H2O2 and NO3 components
in the chamber were performed at transient time
τ = 0.3 s (it corresponded to 8 liters per minute
volume rate of blowing the mixture through the
chamber), Wd = 1.5 W/cm3 at distances 65, 215
and 365 mm from top wall of the chamber. Spread
of measured densities of the particles along the
chamber height is minimum one for О3 comprising
≈ 10%, and maximum one for HNO2 being about
In Fig.4 the dependencies of O3, NO3, N2O5,
HNO3, HNO2 concentrations averaged over the
chamber height on accumulation time in the
chamber tc are presented (each point in the figure
represents data averaged for 6÷9 measurements).
Comparison of measured concentration values for
particles in the chamber with the calculations (Fig.2) shows that: 1) calculated values of O3, HNO3 and N2O5
concentrations exceed their measured values by factor of about two; 2) in cases of NO3 and HNO2 the contrary
situation is observed – calculated values of the concentrations are about twice less than those measured in the
One can see from figures 1-4 that in cases of
O3, HNO3 and N2O5 behavior of changes of their
concentrations at gas mixture injection from the
discharge to the chamber is the same both in the
calculations and the experiments, at that: 1) O3
concentrations in the discharge and the chamber
are practically equal; 2) HNO3 and N2O5
concentrations increase by factors of about (2÷3)
and (10÷15), respectively.
As to NO3 and HNO2 components, the
following mismatch between the calculation and
the experiment is observed: 1) theoretically
calculated NO3 concentration in the discharge
exceeds that in the chamber by factor of about
five, whereas measured concentrations of these
particles are practically equal; 2) calculated values
of HNO2 concentrations in the discharge and the
chamber are practically the same, whereas in the
experiment concentration of these particles in the
chamber is less by almost one order of magnitude
than that in the discharge volume.
1. Atkinson R, Baulch D L, Cox R A et al, Atmos. Phys. 4, 1461-1738 (2004).
2. Stefanovic I, Bibinov N K, Deryugin A A et al, Plasma Sources Sci. Technol., 10, 406-412 (2001).
3. Sakharov A. D., Izvestiya AN SSSR. Seriya Fiz. 12(4), 372-375 (1948) (in Russian).
Опубликовано в рубрике Documents