NO and NO2 production in pulsed low pressure dc discharge

Article abstract of DOI:10.1063/1.1935046

Simultaneous measurements of both NO and NO2 are performed downstream a pulsed low pressure dc discharge in flowing air using tunable diode laser absorption spectroscopy in the infrared region. Pulse duration and repetition rate range from 20 μs to 5 ms and from 50 to 1000 Hz, respectively. The gas pressure is 4 mbar and the peak current is 80 mA. Experimental results show that NO and NO2 production depends only on the duty cycle ratio, that is, on the average power at a given current. A numerical computation of a simplified kinetics agrees well with experiment results.

Full text of DOI:10.1063/1.1935046

NO and NO 2 production in pulsed low pressure dc discharge
Antoine Rousseau, Lina V. Gatilova, Jürgen Röpcke, Alexander V. Meshchanov, and Yury Z. Ionikh
Citation: Applied Physics Letters 86, 211501 (2005); doi: 10.1063/1.1935046
View online: http://dx.doi.org/10.1063/1.1935046
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/86/21?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Optical emission spectroscopic diagnostics of a non-thermal atmospheric pressure helium-oxygen plasma jet for
biomedical applications
J. Appl. Phys. 113, 233302 (2013); 10.1063/1.4811339
Atmospheric pressure plasma jet—Living tissue interface: Electrical, optical, and spectral characterization
J. Appl. Phys. 113, 183302 (2013); 10.1063/1.4804319
Atmospheric pressure resistive barrier air plasma jet induced bacterial inactivation in aqueous environment
J. Appl. Phys. 113, 093302 (2013); 10.1063/1.4794333
Photocatalyst activation in a pulsed low pressure discharge
Appl. Phys. Lett. 87, 221501 (2005); 10.1063/1.2136415
Numerical simulation of de- N O x performance by repetitive pulsed discharge when added with hydrocarbons
such as ethylene
J. Appl. Phys. 97, 023301 (2005); 10.1063/1.1828242
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.22.67.107 On: Sat, 22 Nov 2014 18:15:58

APPLIED PHYSICS LETTERS 86, 211501 ͑2005͒
NO and NO2 production in pulsed low pressure dc discharge
Antoine Rousseaua͒ and Lina V. Gatilova
Laboratoire de Physique et Technologie des Plasmas GdR Cataplasme, Ecole Polytechnique-UMR 7648,
91128 Palaiseau Cedex, France
Jürgen Röpcke
INP-Greifswald, F.-L.-Jahn-Strasse 19, 17489 Greifswald, Germany
Alexander V. Meshchanov and Yury Z. Ionikh
St. Petersburg State University, Faculty of Physics, Ulianovskaya 1, 198904 St. Petersburg, Russia
͑Received 22 December 2004; accepted 11 April 2005; published online 16 May 2005͒
Simultaneous measurements of both NO and NO2 are performed downstream a pulsed low pressure
dc discharge in flowing air using tunable diode laser absorption spectroscopy in the infrared region.
Pulse duration and repetition rate range from 20 µs to 5 ms and from 50 to 1000 Hz, respectively.
The gas pressure is 4 mbar and the peak current is 80 mA. Experimental results show that NO and
NO2 production depends only on the duty cycle ratio, that is, on the average power at a given
current. A numerical computation of a simplified kinetics agrees well with experiment results.
© 2005 American Institute of Physics. ͓DOI: 10.1063/1.1935046͔
Due to recent development of plasma based technologies
for environmental applications, extensive literature exists on
the NOx ͑NO+NO2͒ kinetics in atmospheric pressure
discharges.1–5 In a recent study concerning the potentials of a
pulsed microwave discharge at atmospheric pressure for gas
treatment and environmental applications, it was shown that
the NOx ͑NO+NO2͒ formation depends only on the average
power injected in the plasma, and not on the pulse duration
itself, i.e., a short pulse ͑30 µs͒ with a high repetition rate
will lead to the same amount of NOx as a longer pulse ͑e.g.,
1 ms͒ with low repetition rate, provided that the averaged
power is the same.5 The use of pulses even as short as 30 µs
could not prevent significantly the NOx formation. However,
no straightforward analysis could be performed since the
pulsed microwave discharge had strong space gradients. On
the contrary, positive columns of dc low pressure discharges
are spatially homogeneous and have been extensively stud-
ied, both theoretically and experimentally by Gordiets et al.,6
and Rybkin et al.,7 for a wide range of parameters. In this
letter we report, experimental results concerning the NOx
formation in a low pressure pulsed dc discharge. The simul-
taneous measurement of NO and NO2 concentration is very
useful to validate reaction schemes occurring in the dis-
charge and in the post-discharge for different pulse duration,
repetition rate, and discharge current. NO and NO2 are mea-
sured downstream the plasma region using tunable diode la-
ser absorption spectroscopy ͑TDLAS͒ in the infrared region.
The discharge is generated in a 50 cm long and 1.9 cm
inside diameter Pyrex tube ͑Fig. 1͒. The dc generator is
pulsed by a master pulse generator. Pulse duration and rep-
etition rate are adjustable from 20 µs to 5 ms and from 50 to
1000 Hz, respectively; the peak current is set to 80 mA. The
gas flow is 20 sccm of dry air-like mixture ͑N2/O2͒; the gas
pressure inside the discharge tube is 4 mbar. The exhaust gas
is pumped out of the discharge tube through a multipath cell
͑optical path length is 100 m͒, where NO and NO2 species
are simultaneously detected using a two-laser beam infrared
system ͑TOBI͒ ͑Fig. 1͒.8 The pressure inside the cell was
adjusted from 0.1 to 2 mbar, in order to optimize the signal
to noise ratio. Essential basics of absorption spectroscopy are
the relation between the incident ͑I0͒ and the transmitted ͑I͒
intensity of the laser radiation:
L
I0
ln
=
k n͑z͒dz,
␯
ͩ ͪ
͵
I
0
where k is the absorption cross section at the wave number
␯
; L is the absorption length, and n is the density of the
v
absorbing species. NO2 and NO lines, which have been used,
are located in the 1626 and 1880 cm−1 spectral region, re-
spectively. Absorption lines with different line strengths were
used in order to increase the dynamic of the detection sensi-
tivity, from 1 to 10 000 ppm typically.
In Figs. 2 and 3 the influence of the duty cycle ratio
͑DCR͒ on the NO and NO2 production is presented for dif-
ferent pulse repetition rates, respectively. DCR is the ratio of
the pulse duration to the pulse period. NO increases monoto-
nously from 40 to 15 000 ppm as the duty cycle ratio in-
creases from 0.1% to 50% ͑Fig. 2͒. On the contrary, NO2
decreases continuously from 160 to 20 ppm ͑Fig. 3͒. It is
a͒
Author to whom correspondence should be addressed; electronic mail:
FIG. 1. Experimental set up: pulsed dc discharge, the infrared laser TOBI
and the multipath cell.
antoine.rousseau@lptp.polytechnique.fr
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
0003-6951/2005/86͑21͒/211501/3/$22.50 86, 211501-1 © 2005 American Institute of Physics
129.22.67.107 On: Sat, 22 Nov 2014 18:15:58

211501-2
Rousseau et al.
Appl. Phys. Lett. 86, 211501 ͑2005͒
FIG. 2. Influence of the duty cycle ratio, DCR, on the formation of NO for
different pulse repetition rates. Peak current 80 mA, 4 mbar, dry air, 20
sccm. The solid line is the computed results from numerical code.
FIG. 3. Influence of the duty cycle ratio, DCR, on the formation of NO2 for
different pulse repetition rates. Labels identical to Fig. 2.
worthwhile to notice that NO and NO2 depend only on the
DCR: the experimental conditions of 4 ms pulse duration at
50 Hz are equivalent, in terms of NOx production, to 200 µs,
1 kHz ͑DCR=20%͒. The scaling of NO2 concentration with
the DCR ͑Fig. 3͒ is less remarkable than that of NO; but the
discrepancies in NO2 density are in the range of 10 ppm,
which is much less than NO density fluctuations. It is worth
noting that DCR is proportional to the time-averaged dis-
charge power for the given current. This shows that NO and
NO2 density scales as a function of the time-averaged depos-
ited power. The peak discharge power P may be found from
P= j·E, where j=I/S=2.8ϫ10−2 A cm−2 is the current den-
sity for I=80 mA. E=63 V cm−1 is calculated from the re-
duced electric field for our discharge conditions E/N
=65 Td.9 This corresponds to the peak discharge power P
=1.7 W cm−3 which gives the average discharge power in-
creasing from 0.72 to 360 kJ/I for DCR increasing from
0.1% to 50%, respectively. Recently a similar observation
was published for a pulsed microwave discharge near atmo-
spheric pressure where both NO and NO2 production only
depend on the average power.5
An extensive analysis of the NO and NO2 kinetics in a
low pressure pulsed dc discharge will be published in a forth-
coming paper. However, let us show that a simplified kinetics
of NO gives a very good agreement with experimental re-
sults. Table I presents reactions taken into account in our
model. The related kinetic equations are numerically solved
with the following assumptions: ͑i͒ the electron energy dis-
tribution function ͑eedf͒ is computed using a code developed
10
in the Kurchatov Institute ͑Moscow͒ with Tg=500 K, Tv
=4000 K, and E/N=65 Td. NO density being less than 1.2%
is not taken into account for the computation of eedf; ͑ii͒ the
electron density and E/N are supposed to be constant during
the ON phase of the pulse and equal to zero during the OFF
phase of the pulse; ͑iii͒ no ion kinetics are included in the
model; ͑iv͒ the computation procedure stands as follows: a
TABLE I. List of reactions taken into account in the numerical model.
Rate constant
͑cm3 s−1
No.
Reactions
͒
References
+
1
N2 +e→N*2͑A3⌺u ͒+e
5ϫ10−11
7.7ϫ10−11
10
2
N2 +e→N*2͑B3⌸g͒+e
10
1
−
3
N +e→N*͑a
⌺
͒+e
u
4.4ϫ10−12
10
Ј
2
2
4
N2 +e→N*͑2D͒+N+e
8.1ϫ10−13
10 and 11
5
O2 +e→O+O+e
4.3ϫ10−11
10
6
O2 +e→O+O*͑1D͒+e
NO+e→N+O+e
2.0ϫ10−10
11
7
3.0ϫ10−11
11
N*2͑A3⌺u ͒+O2 →N2 +O2
2.5ϫ10−12ϫ͑T/300͒
0.65ϫk8
12
+
0.55
8
+
9
N*2͑A3⌺u ͒O2 →N2 +O+O
12
+
10
11
12
13
14
15
16
17
18
19
20
21
N*2͑A3⌺u ͒+O→N2 +O
2.8ϫ10−11
12
N*2͑A3⌺u ͒+O→NO+N*͑2D͒
0.4ϫ10−11
13
+
N*2͑B3⌸g͒+N2 →N*2͑A3⌺u ͒+N2
5.0ϫ10−11
13
+
N*2͑B3⌸g͒+O2 →N2 +O+O
3.0ϫ10−10
13
1
1
−
−
N*͑a
⌺
⌺
͒+O2 →N2 +O+O
͒+NO→N2 +N+O
2.8ϫ10−11
13
Ј
Ј
2
u
N*͑a
3.6ϫ10−10
13
2
u
N+NO→N2 +O
3.1ϫ10−11
14
N*͑2D͒+O2 →NO+O
O*͑1D͒+N2 →O+N2
O*͑1D͒+O2 →O+O*2
O*͑1D͒+NO→N+O2
O+Oads→O2
9.7ϫ10−12 e−185/T
1.8ϫ10−11 e107/T
3.2ϫ10−11 e67/T
1.7ϫ10−10
12
13
13
13
25.4
For ␥=0.5ϫ10−3
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.22.67.107 On: Sat, 22 Nov 2014 18:15:58

211501-3
Rousseau et al.
Appl. Phys. Lett. 86, 211501 ͑2005͒
sequence of n pulse periods is computed, starting from pure
N2/O2 air like mixture, followed by t=2 min. n is defined by
the ratio of the experimental residence time of the gas inside
the discharge tube 1.2 s to the period; t is the gas averaged
residence time between the discharge tube and the multipath
cell where infrared measurements are performed; ͑v͒ NO2
kinetics have not been included.
In the present study, tunable diode laser absorption spec-
troscopy has been performed downstream a pulsed low pres-
sure dc discharge in air, in order to measure simultaneously
the formation of NO and NO2 for various pulse durations and
repetition rates. It is shown that the NOx production depends
only on the pulse duty cycle ratio, which is proportional to
the pulse power. This is similar to recent measurements per-
formed downstream a pulsed microwave discharge near at-
mospheric pressure. NO density increases with DCR up to
about 1.2% for an averaged specific energy equal to 360 kJ/l.
A numerical computation, solving a set of 21 kinetic equa-
tions including some excited states of N2 and N, gives a very
good agreement with experimental results. NO formation is
Figure 2 also shows the good agreement between the
results of the numerical computation of the model presented
above and the experimental results for NO. This agreement
covers a wide range of pulse frequency and duration as well
as they NO density. NO molecules are formed and destroyed
mainly during the plasma pulse. The major process leading
to NO formation is the reaction between N*2͑A3⌺u+͒ molecules
and atomic oxygen O ͑reaction 11͒. In fact in this reaction
two NO molecules are created, because the excited N*͑2D͒
atoms, that are produced in reaction ͑11͒, collide with O2 to
produce another NO molecule in reaction ͑17͒. The destruc-
+
driven by the time averaged density of N*2͑A3⌺u ͒, which is
an increasing function of the DCR.
1B. Penetrante, M. Hsiao, B. Meritt, G. Vogtlin, and P. Wallman, IEEE
Trans. Plasma Sci. 23, 67 ͑1995͒.
1
−
tion of NO occurs by the collision with electrons, N ͑a ⌺ ͒
Ј
2
u
2F. Fresnet, G. Baravian, L. Magne, S. Pasquiers, C. Postel, V. Puech, and
A. Rousseau, Appl. Phys. Lett. 77, 4118 ͑2000͒.
molecules and O*͑1D͒ atoms ͑reactions 7, 15, and 20͒.
The physical reason for the dependency of the NOx on
the DCR may be summarized as follows. The rate of
3R. Dorai and M. Kushner, J. Phys. D 34, 574 ͑2001͒.
4Th. Hammer, Plasma Sources Sci. Technol. 11, 196 ͑2002͒.
5A. Rousseau, A. Dantier, L. Gatilova, Y. Ionikh, J. Röpcke, and Y. Tol-
machev, Plasma Sources Sci. Technol. 14, 70 ͑2005͒.
6B. F. Gordiets, C. M. Ferreira, V. L. Guerra, J. M. A. H. Loureiro, J.
Nahorny, D. Pagnon, M. Touzeau, and M. Vialle, IEEE Trans. Plasma Sci.
23, 750 ͑1995͒.
+
N*2͑A3⌺u ͒ molecules depends mostly on the electron density
͑process 1͒, which is constant during the pulse and equal to
zero after the pulse; thus the time average value of this rate is
proportional to the duty cycle ratio. The same is valid for NO
molecule destruction ͑processes 7, 15, and 20͒. The destruc-
7S. A. Smirnov, V. V. Rybkin, I. V. Kholodov, and V. A. Titov, High Temp.
40, 323 ͑2002͒.
+
tion of the N*2͑A3⌺u ͒ excited state goes via the collision
with O2 and O ͑processes 10 and 11͒, where the time aver-
aged O atom density depends only on the DCR for a given
pulse amplitude; as a result NO concentration depends only
on the DCR and shows some saturation effect for high DCR
corresponding to a high O atom density.
NO2 kinetics were not included in the kinetic modeling;
its density is much lower than the NO density for DCR
Ͼ1%. NO2 is formed downstream the plasma region via oxi-
dation of NO by ozone:
8J. B. McManus, D. Nelson, M. Zahniser, L. Mechold, M. Osiac, J.
Röpcke, and A. Rousseau, Rev. Sci. Instrum. 74, 2709 ͑2003͒.
9Y. P. Raizer, Gas Discharge Physics ͑Springer, Berlin, 1991͒.
10N. A. Dyatko, I. V. Kochetov, and A. P. Napartovich, J. Phys. D 26, 418
͑1993͒.
11J. Nahorny, C. M. Ferreira, B. Gordiets, D. Pagnon, M. Touzeau, and M.
Vialle, J. Phys. D 28, 738 ͑1995͒.
12J. T. Herron, J. Phys. Chem. Ref. Data 28, 1453 ͑1999͒.
13I. A. Kossyi, A. Yu. Kostonsky, A. A. Matveev, and V. P. Silakov, Plasma
Sources Sci. Technol. 1, 207 ͑1992͒.
O3 + NO → NO2 + O2,
where the O3 density is strongly reduced with increasing gas
temperature above 300 K. This would explain why the NO2
production decreases with increasing DCR.
14R. Atkinson, D. L. Baulch, R. A. Cox, R. F. Hampson, J. A. Kerr, and J.
Troe, J. Phys. Chem. Ref. Data 18, 881 ͑1989͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.22.67.107 On: Sat, 22 Nov 2014 18:15:58