Investigation of nitrogen fixation in low-

pressure microwave plasma via rotational–

vibrational NO and N2 kinetics
Cite as: J. Appl. Phys. 133, 113303 (2023); https://doi.org/10.1063/5.0138298
Submitted: 10 December 2022 • Accepted: 12 February 2023 • Published Online: 15 March 2023

Omid Samadi Bahnamiri, Filippo Manaigo, Abhyuday Chatterjee, et al.

COLLECTIONS

This paper was selected as Featured

J. Appl. Phys. 133, 113303 (2023); https://doi.org/10.1063/5.0138298 133, 113303

© 2023 Author(s).
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

Investigation of nitrogen fixation in low-pressure

microwave plasma via rotational–vibrational NO
and N2 kinetics
Cite as: J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298
Submitted: 10 December 2022 · Accepted: 12 February 2023 · View Online Export Citation CrossMark
Published Online: 15 March 2023

Omid Samadi Bahnamiri,1,a) Filippo Manaigo,1,2 Abhyuday Chatterjee,1 Rony Snyders,1,3

4,b)
Federico Antonio D’Isa, and Nikolay Britun5

AFFILIATIONS
1
Chimie des Interactions Plasma-Surface (ChIPS), CIRMAP, Université de Mons, 23 Place du Parc, 7000 Mons, Belgium
2
Research group PLASMANT, Department of Chemistry, University of Antwerp, Universiteitsplein 1, BE-2610 Antwerp, Belgium
3
Materia Nova Research Center, Parc Initialis, 7000 Mons, Belgium
4
Max Planck Institute for Plasma Physics, Boltzmannstr. 2, 85748 Garching, Germany
5
Center for Low-temperature Plasma Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8603, Japan

a)
Author to whom correspondence should be addressed: [email protected]
b)
Present address: Consorzio RFX, Corso Stati Uniti 4, 35127 Padova (PD), Italy

ABSTRACT
A pulsed microwave surfaguide-type discharge used for nitrogen fixation in N2 –O2 gas mixtures is characterized by optical emission spec-
troscopy. Results show that both rotational and vibrational temperatures are elevated in the active zone near the waveguide, decaying along
the discharge tube in both upstream and downstream. The characteristic length of optical emission from NO(A-X) transition gets contracted
when pressure increases, specifically at P
2 Torr. The degree of vibrational non-equilibrium (defined as the ratio between vibrational and
rotational temperatures) is decreased by a factor of two when pressure changes from 0.6 to 10 Torr. Non-equilibrium likely disappears as the
discharge pressure rises, resulting in a gas temperature elevation. A correlation between gas residence time, pulse duration, and characteristic
times for different energy transfer channels is discussed. The rotational–vibrational dynamics differs for NO and N2 during the pulse. Both
species lose vibrational excitation at the beginning of the pulse, whereas N2 gets re-excited again during the second half of the pulse, which
may occur as a result of an efficient pumping-up effect through the vibrational–vibrational energy transfer. At the same time, vibrational
relaxation of NO takes place primarily due to a strong vibrational–translational exchange via NO–O2 and NO–O collisions.

Published under an exclusive license by AIP Publishing. https://doi.org/10.1063/5.0138298

I. INTRODUCTION cost-effective approach when considering the utilization of inter-

mittent green energy sources. Low-temperature plasmas provide an
Reactive nitrogen species, such as NH3 or NOx , are used as a
feedstock to produce fertilizers. The process of creating these opportunity for N-compounds production due to the high chemi-
species from chemically inert N2 , known as nitrogen fixation, has cal reactivity occurring as a result of a non-equilibrium between
attracted much interest.1 Today, the industrial nitrogen fixation for the electrons and heavy particles.3,4 Owing to the stability of the N2
NH3 production via Haber–Bosch (H–B) process dominating artifi- triple bond, interrogation of efficient routes for N2 dissociation
cial fertilizer manufacturing is associated with high energy costs remains a key hurdle to improving the nitrogen fixation process.
and high CO2 emissions causing a dramatic environmental Among the plasma technologies under consideration, micro-
impact.2 Therefore, the development and integration of alternative wave (MW) discharges5,6 offer an efficient dissociation pathway
processes are of key importance. due to a high degree of non-equilibrium at low pressure. This is an
In the quest for alternative nitrogen fixation technologies, inherent feature of low-pressure MW discharges where typically
plasma-assisted nitrogen fixation is a highly promising more than 90% of applied electric power is absorbed by electrons

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-1

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

and channelized toward electronic and rotational–vibrational following reaction:

(ro-vibrational) excitation of heavy particles.7 This energy is
further transmitted to highly excited vibrational states, whereas the N2 (X, v
13) þ O(3 P) ! NO(X) þ N(4 S), (R3)
gas heating remains relatively low.3 In addition, the use of dis-
charge power modulation, i.e., working in the pulsed regime with a whereas the destruction of NO(X) occurs via the reverse Zeldovich
certain repetition rate, presents an interesting strategy for gas mechanism:
heating limitation in MW plasmas.4,8,9 This effect has also proven
to be very efficient for molecular dissociation.10 In pulsed MW dis- NO(X) þ N(4 S) ! N2 (X, v ≃ 3) þ O(3 P) (R4)
charges, the gas temperature decreases rather rapidly between
pulses while the vibrational temperature reduces more slowly, as corresponding to high rate coefficients: R3 ¼ 1013 cm3 s1 ;
reported by Jivotov11 Pulsed MW plasma in pure N2 at low pres- R4 ¼ 3:3  1011 cm3 s1 , at Tgas ¼ 1000 K. In addition, the vibra-
sure (5 mbar) has been previously studied by Baeva et al.12 con- tionally excited NO has a fast V–T self-relaxation rate of about 5 
firming the high level of vibrational excitation present. An 1013 cm3 s1 at 900 K which is 5 orders of magnitude faster than
extensive coverage of the fundamental kinetics of N2 and N2 –O2 that of N2 and CO molecules, as shown elsewhere.3,16,17
plasmas by Capitelli et al.13 emphasizes the profound role of vibra- On the other hand, the NO(A) state is predominantly pro-
tional and electronic excitation on discharge properties. duced by collisions of N2 (A) metastables with ground state NO(X)
The importance of vibrational excitation is motivated by three molecules. The population of this state is reduced by quenching
main factors leading to higher non-equilibrium processes:3 (i) elec- with N2 , O2 , and NO.14
tron impact vibrational excitation (e-V energy transfer) in which To date, numerous efforts have been focused on a deeper
plasma electrons excite the lowest vibrational levels of N2 in understanding of plasma-assisted nitrogen fixation in various low-
ground electronic state at relatively low mean electron energies temperature non-equilibrium discharges, including particular atten-
(e.g., in the range of 1–3 eV); (ii) redistributing these vibrational tion on vibrational kinetics. From the experimental point of view,
quanta via vibrational–vibrational (V–V) transfer, sometimes called the characterization of the plasma-assisted nitrogen fixation process
as “vibrational ladder climbing”; and (iii) a relatively slow vibra- is of much interest nowadays, since an in-depth insight into the
tional–translational (V–T) energy transfer which is responsible for related mechanisms could be achieved, including both space- and
gas heating through the quenching of vibrationally excited N2 (and time-resolved information. In spite of the numerous modeling and
other molecules). The latter occurs over a longer time (especially at experimental studies devoted to nitrogen fixation in MW
low gas temperature and pressure), while the e-V process followed plasmas,18–20 dielectric barrier discharges (DBDs),21 propeller-arc
by V–V transfer is fast. Thus, by limiting the V–T transfer, one discharges,22 glow discharges,23 spark discharges,24 gliding arc
may expect boosting N2 dissociation as well.8 In addition, due to plasmas (GAPs),25 the behavior of NO and N2 ro-vibrational
the anharmonicity of the potential energy curve of N2 (X), the V–V dynamics have not yet been well understood, requiring special
energy exchanges are not exactly resonant having larger rate coeffi- attention in the microwave discharge case. In particular, the role of
cients for the exchanges that lead to a gain of quanta on the top of pulsed power and its influence on V–V and V–T energy transfers,
the vibrational ladder and a loss at the bottom. This gives rise to a specifically for low-pressure MW plasmas is still an open question.
pumping-up effect in the ladder, which is balanced by the loss of The aim of this paper is to give an insight into the kinetics of
vibrational quanta by V–T processes.14 NO formation in a surfaguide-type microwave discharge working
As mentioned earlier, the non-equilibrium kinetics of low- in N2 –O2 gas mixtures. For this purpose, the temporal dynamics of
temperature plasmas in N2 –O2 mixtures is of high importance. NO NO, N2 , and O atoms, as well as the space-resolved measurements
production in the N2 –O2 plasma is often associated with so-called have been investigated using optical emission spectroscopy (OES).
Zeldovich mechanism promoted by vibrational excitation,3 An attempt to answer the question of how the degree of vibrational
non-equilibrium changes when discharge parameters such as
N2 (X, v) þ O(3 P) ! NO(X) þ N(4 S), (R1) power, pressure, and pulse duration are varied, is undertaken.

II. EXPERIMENTAL SETUP AND METHODOLOGY

N(4 S) þ O2 (3 Σ, v) ! NO(X) þ O(3 P), (R2)
A. The MW plasma source
where X stands for the ground state, whereas v indicates vibration- The study has been conducted in a surfaguide-type MW
ally excited state. This mechanism can, thus, be exploited by over- 2.45 GHz discharge working in the 0.5–12 Torr pressure range in
populating the vibrational levels via V–V energy transfer. N2 :O2 gas mixtures. An experimental setup used in this study is
Depopulation of N2 (X, v), however, takes place through V–T shown in Fig. 1(a). The wave was launched by the surfaguide situ-
exchanges associated with N2 –O2 , N2 –N, and N2 –O collisions. The ated at the middle of a quartz tube. The waveguide was connected
influence of these V–T channels becomes progressively smaller to a fine (three-stub) tuner and a movable reflector (matching) to
with the increase of rate coefficient for the reaction (R1), which is tune the electrical field to optimal conditions for an electrical
the predominant mechanism for the formation of NO and N.15 breakdown in order to minimize the reflected power. The discharge
The kinetics of nitric oxide states, such as NO(X) and NO(A) was sustained in the quartz tube with a 14 mm inner diameter and
in the pressure range of 0.1–10 Torr, have been investigated by about 31 cm length, surrounded by a Plexiglass tube. Silicone oil
modeling.14 In this case, the NO(X) creation mainly occurs via the circulating between two tubes was served for cooling of the quartz

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-2

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

TABLE I. Main parameters related to the MW plasma source and optical emission
spectroscopy used in this study.

Parameter Value
Plasma source
MW discharge type Surfaguide
Base (microwave) frequency 2.45 GHz
Pulse repetition rate, f 0.1–2.0 kHz
Mean applied power, p 0.5–0.9 kW
Gases used N2, O2, CO2
Gas flow rate range, F 0.02–5 slm
Pressure range, P 0.5–12 Torr
Quartz tube inner diameter 14 mm
Quartz tube length 31 cm
Plexiglass tube inner diameter 32 mm
Optical emission diagnostics
Monochromator Andor SR750
Grating 1800 g/mm, 3600 g/mm
N2
Resolution for Trot 20 pm
NO
Resolution for Trot 32 pm
Resolution for Tvibr 50–70 pm
Optical detector Andor iStar740 ICCDa
ICCD triggering mode External
ICCD accumulation number 20-200
ICCD exposure time 0.05 s (typically)
Trigger frequency 0.5 kHz
Number of averaged pulsesb 500–5000
a
Intensified charge coupled device.
b
Number of averaged pulses = (trigger frequency, Hz) × (ICCD exposure
time, s) × (ICCD accumulation number).

spectrometer with a 20 μm entrance slit and 3600 g/mm grating,

yielding a high resolution of about 20 pm for Trot determination.
FIG. 1. 3D (a) and 2D (b) schematic view of the MW plasma source, as well as The emission lines were recorded by an Andor iStar740 intensified
the monochromator and ICCD detector as spectroscopic tools used in this work. charge coupled device (ICCD) camera working in accumulative
The N2 :O2 (1:1) discharge appearance (c) in the quartz tube taken at
mode. The details related to the OES diagnostics are also listed in
p ¼ 0:7 kW, f ¼ 0:5 kHz with duty ratio of 50%.
Table I. The kinetic acquisition mode of the ICCD camera with a
time step of 50 μs was used during the time-resolved measure-
ments. The discharge and ICCD camera were synchronized using
tube down to about 10 C. This geometry allows us to carry out the the delay generator. For the space-resolved profile, the fiber was
OES-based measurements either along the discharge tube or per- moving along the discharge tube [X axis defined as D in Fig. 1(c)]
pendicular to the discharge column aiming to assess the space- with a step equal to about 0.5 cm. The spatial interval between 25
resolved evolution [see Fig. 1(b)]. The main parameters of the and 25 mm was inaccessible for optical emission measurements
described MW plasma source are summarized in Table I, while due to the presence of the waveguide.
more details can be found elsewhere.26–28 The discharge appear-
ance of the N2 :O2 (1 : 1) gas mixtures in the quartz tube is depicted
in Fig. 1(c). The discharge has been modulated in the pulse regime C. Ro-vibrational analysis
triggered externally using a TGP110 delay generator.
1. N2 C3 Πu !B3 Πg
A typical wide emission spectrum from the discharge active
B. Spectroscopy tools
zone for N2 –O2 gas mixtures is shown in Fig. 2(a). Spectral transi-
For discharge characterization, an Andor SR750 monochro- tions of a second positive system (SPS) of N2 C3 Πu !B3 Πg are
mator with either 1800 g/mm or 3600 g/mm diffraction gratings used to determine the rotational (Trot ) and the vibrational (Tvibr )
has been utilized, respectively, for the determination of vibrational temperatures of N2 measured in the discharge area. The rotational
(Tvibr ) and rotational (Trot ) temperatures in the discharge area. An spectra were fitted using a residual minimizing code based on sim-
optical fiber relays the light emitted by plasma emission into the ulated datasets providing Trot of N2 as a result of the fitting process

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-3

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

FIG. 2. (a) The typical wide emission spectrum taken in the pulsed MW discharge for N2 –O2 gas mixtures at an average applied power of 0.7 kW and a gas pressure of
3 Torr. (b) The vibrational band sequences from the N2 (C-B) used for Tvibr determination in the N2 (C) state. (c) Enlarged view of the N2 (C,0 - B,0) rotational band at
337.2 nm used for Trot determination with 20 pm resolution.

[see Fig. 2(c)]. The simulated spectra were synthesized using using the freely available LIFBASE (v 2.1.1) software.30 The rota-
MassiveOES software.29 The rotational spectra were acquired with tional structures of the measured NO(A2 Σþ , v0 ¼ 0  X2 Π, v00 ¼ 2)
about 20 pm of spectral resolution. The vibrational band sequences band (shortened to NO(A,0  X,2) below) located at 247 nm are
from N2 (C,v0  B,v00 ) with v0 -v00 ¼ 1,  2,  3, and 4 were well fitted and accurately matched with the simulation one, as
used for the determination of Tvibr of N2 in the N2 (C) excited state shown in Fig. 3(b). The fitting range for rotational spectra of NO
[see Fig. 2(b)] using the Boltzmann plot approach, similar to previ- has been reduced to 245–249 nm to avoid the influence of the NO
ous works.26,27 The non-uniform spectral transmission of the (A,1  X,3) emission band. In this case, after shrinking the range,
Plexiglass tube and the monochromator spectral response were Trot of NO, given in Sec. III, is overestimated by about 5%. The
taken into account. temperature error is defined by a simulation step, which is equal to
50–100 K in the rotational temperature case, whereas for vibrational
2. NO A2 Σþ !X2 Π temperature, this step is 300 K.
The OES imaging for NO(A,0  X,2) band at 247 nm and O
A typical time-averaged emission spectrum of the NO
triplet at 844.6 nm has been performed through a 247 nm/10 nm
A2 Σþ !X2 Π transitions in the 205–275 nm region, a part of the
and 840 nm/10 nm optical bandpass filters attached to the ICCD
γ-system, is shown in Fig. 3(a). The NO spectra were collected
camera.
using a UV–VIS optical fiber along the discharge tube. The Trot
and Tvibr of NO were also determined by the custom-made residual
minimizing code mentioned above. For this purpose, the analytical III. RESULTS AND DISCUSSIONS
spectra synthesized in the range of both rotational and vibrational First, the ro-vibrational dynamics of N2 and NO species and
temperature followed by fitting these two parameters during the the axial distribution of Tvibr and Trot in the pulsed MW discharge
same run. The calculation of the rotational spectra was performed are described, since they are necessary for understanding of the

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-4

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

FIG. 3. The time-averaged emission spectra taken in N2 –O2 gas mixtures for NO A2 Σþ  ! X2 Π transition (a part of the γ-system) in the range of 205–275 nm (a); mag-
nified view of the NO (A,0  X,2) rotational band used for Trot and Tvibr determinations (b).

effects of power, pressure, and plasma pulse duration, which are Based on the estimated total quenching time (τ est , as a result of
discussed later. radiative decay and quenching), the excited states of N2 (C) and CO
(B) may not be thermalized with the reference gas in our case. The
thermalization times, τ RT , for N2 (C) and CO(B) are about 180
A. Gas temperature
and 125 ns at 10 Torr and at 1000 K, respectively.33,34 This suggests
The emission corresponding to the N2 (C  B) and CO(B  A) that the N2 (C) and CO(B) excited states are not in thermal equilib-
transitions has been chosen for Tgas determination in CO2 :N2 (4:1) rium with the gas during their lifetime, since τ RT is longer than
gas mixtures [the details related to the Boltzmann fit of CO(B) are τ est in our case.
described in Appendix A]. The first consideration to make is On the other hand, as long as the main population mecha-
whether Trot measured by N2 (C) and CO(B) can be expected to be nism for the excited state is direct electron impact from the ground
in equilibrium with the translational (gas) temperature. Estimates state, the Trot is a good estimation for the gas temperature. In this
show that the radiative lifetime of N2 (C) is about 36 ns,31 while the case, the rotational distribution in the excited state correlates with
corresponding lifetime for CO(B) is about 23 ns32 (see Table II). the one of the ground state, which, in turn, is equilibrated with gas
temperature, since electron excitation happens much faster than
the R–T energy relaxation for either ground or exited atomic states.
TABLE II. Radiative lifetime (τ0), rate coefficient (k), total quenching rate (Qtot) for Silva et al.28 reported that the CO(B) state is mainly produced by
electronic energy transfer, total quenching time (τest), and thermalization time (τR−T)
estimated at 1000 K. The calculations are taken at CO2:N2 (4:1) gas mixture at
the direct electron excitation from CO(X) in a similar MW dis-
10 Torr. charge at low pressure, allowing Trot (B1 Σþ )  Trot (X 1 Σþ g ).
Therefore, the following correlation between the rotational temper-
N2(C) CO(B) Reference [N2]/[CO] atures of CO(B) and CO(X) state takes place:37
τ0 (ns) 36 23 31/32
Bv (X, v ¼ 0)
kN2 (10−11 cm3 s) 1.24 16.5 35/32 Tgas (K) ¼ Trot (B)   Trot (B), (1)
kO2 (10−11 cm3 s) 29 55.4 35/32 Bv (B, v ¼ 0)
kCO2 (10−11 cm3 s) 32 81 35/32
Qtota (107 s−1) 8.7 22 where Bv (X, v ¼ 0) and Bv (B, v ¼ 0) are the rotational constants
τestb (ns) 8.8 3.8 for CO(X, v ¼ 0) and CO(B, v ¼ 0) state, respectively.38 As a
τR−T (ns) 180 125 34, 36, 33 result, Trot of CO(B) in the discharge area can be considered as a
Tgas in our case.
a
The Qtot is estimated based on the CO2:N2 (4:1) gas mixtures. Trot of N2 (C) yields a higher value than that for CO(B) at the
b
τest = (1/τ0 + Qtot)−1. same discharge (see Fig. 12 in Appendix B). Since both N2 (C) and

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-5

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

CO(B) rotational temperatures have a similar trend during the m) and vgas is the gas velocity (in m s1 ). vgas in the tube was esti-
plasma pulse ON and OFF time, we can only conclude that there is mated based on the known discharge tube cross section, measured
an additional excitation mechanism for N2 (C) electronic state, as gas flow rate and pressure, as well as the gas temperature deter-
discussed further below. mined along the discharge tube. The vgas can be expressed as

F 760 Tgas
B. NO and N2 ro-vibrational dynamics vgas (m s1 )  0:166    , (3)
πr 2 Pgas 298
Temporal behavior of the emission intensity and characteristic
temperatures in the active discharge zone are shown in Fig. 4. The where F is the gas flow rate (in standard liter per minute, slm), r is
pulse repetition rate was set at 0.5 kHz with a total gas flow rate of the radius of the discharge tube (in cm), and Pgas is the gas pressure
2 slm, corresponding to a discharge pressure of 5 Torr. (in Torr). vgas in our MW system was estimated in the range of 43–
Gas residence time (τ res , defined as the time during which the 130 m s1 (depending on the gas flow rate). Based on the estimated
gas is located in the active discharge zone) can be estimated using vgas , τ res of about 1.4 ms at a total gas flow rate of 2 slm was found.
the expression This certifies the τ res in the plasma region is comparable to the
inter-pulse interval, thus ensuring that any gas molecule will be
1
τ res (s)  Lplasma  vgas , (2) exposed to plasma pulse when passing through the discharge.
Trot of both NO and N2 starts growing at the beginning of the
where Lplasma is the plasma length measured from plasma emission plasma pulse showing peaks of about 2200 and 1200 K at about
and captured by ICCD camera without optical bandpass filters (in 0.7 ms, respectively [see Figs. 4(a) and 4(b)]. These undergo a

FIG. 4. Time evolution of the Trot of NO(A) and N2 (C) determined based on the corresponding rotational bandheads (a) and (b); Tvibr of NO(A) and N2 (C) excited state (c)
and (d). More information is available in the legend.

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-6

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

smooth decrease until the end of the plama pulse (1 ms), thereafter, TABLE III. Degree of vibrational non-equilibrium (defined in this work as Tvibr/Trot)
get saturated in the OFF time. A similar trend for the temporal evo- estimated at different pulse delay. The pulse repetition rate and pressure were kept
lution of the gas temperature was found in an air pulsed discharge at 0.5 kHz and 5 Torr, respectively.
with a pulse duration of 5 ms working at 1 Torr, as reported by
Species Tvibr/Trot at pulse ON time
Guerra et al.39 They showed that gas heating occurred during the
plasma pulse is mainly a result of collisional quenching of the elec- 0.25 ms 0.65 ms 0.95 ms
tronically excited states of N2 by oxygen molecules and electronic N2 8.4 7.0 8.3
impact processes causing energy to be released into the transla- NO 2.7 1.8 2.0
tional degrees of freedom. This is in line with other modeling Tvibr/Trot at pulse OFF time
studies on gas heating.40 In addition, this leads to a continuous 1.25 ms 1.55 ms 1.95 ms
increase in V–T N2 –O relaxation (as discussed in detail below) to N2 10.7 9.2 8.0
gas heating until the end of the plasma pulse. The other important NO 2.1 2.3 2.6
gas heating mechanism could be the highly exothermic reverse
Zeldovich reaction (R4) transferring 2.45 eV to gas heating. As a
result, the reaction (R4) and V–T N2 –O energy transfer are of par-
ticular importance in gas heating, since the corresponding rate energy transfer mechanisms, such as e-V, V–V, V–T relaxation pro-
coefficients increase with the gas temperature.39 During the plasma cesses in N2 –O2 discharge, are given in Table IV. As an example,
pulse OFF time, the observed significant decrease in gas heating is for N2 molecule, the characteristic times for aforementioned pro-
likely related to the fast decay on the electron density and electronic cesses can be expressed as62
excited states.39
The discrepancy in rotational temperature values between NO 1
τ eV (s) ¼ (ne  keV
1!0 ) , (4)
(A) and N2 (C) states reflects the absence of equilibrium between
them. Similar conclusions had also been reported by Bruggeman
1
et al.37 showing a larger Trot of NO(A) in comparison with Trot of τ VV (s) ¼ (nN2  k1!0
0!1 ) , (5)
N2 (C). Another observation was found in an atmospheric spark
discharge, reported by Britun et al.24 They showed that Trot of NO
(A) was reached to 4500 K, whereas Tgas was about 1400 K. Van τ VT (s) ¼ (nN2  k1!0 )1 , (6)
Gessel and Bruggeman41 found that Trot of NO(A) (1830 K)
might be higher than the Trot of NO(X) (860 K) in an atmo- where keV 1!0
1!0 , k0!1 , and k1!0 are the rate coefficients for e-V, V–V,
spheric pressure He-air microwave plasma jet. Staack and and V–T relaxation reactions in cm3 s1 , respectively. ne is the elec-
Fridman42 also reported different rotational temperatures measured tron density and nN2 is the N2 number density. kVT can be deter-
in the same atmospheric pressure glow discharge, resulting in Trot mined using a semi-empirical expression from3 involving the V–T
of NO(A) and N2 (C) equal to 2200 and 1200 K, respectively. rate coefficient for N2 ,
In our case, we suggest that the non-equilibrium between NO 1=3
μ T0
0:68 0:3
(A) and N2 (C) might be a result of two effects: (i) a double kVT  3  106 ε2:66 μ2:06 e(0:49ε )
, (7)
Boltzmann slope for NO rotational distribution giving higher tem-
perature (not verified yet), whereas N2 rotational distribution at where ε is the vibrational energy for quantum oscillator (in K), μ is
337.2 nm reveals only one slope (as proved by additional fitting); the reduced mass of colliding particles (in a.m.u.), and T0 is the gas
(ii) a single slope rotational distribution of NO may be excited by temperature (in K). The expressions related to the keV and kVV
additional rotational transfer, being not in equilibrium with the gas can be found elsewhere.3
temperature. Based on our present understanding, we incline The e-V energy transfer stimulating the higher vibrational
toward the second hypothesis, since the double slope of NO(A) levels exhibits significantly higher reaction rates than V–T relaxa-
rotational distribution was not reported in the literature. tion (see Table IV). The noticeable difference in the rate coefficient
A degree of vibrational non-equilibrium defined as Tvibr /Trot between V–V N2 –N2 and V–T N2 –N2 energy transfers is responsi-
(as an indicator of the extent of thermal non-equilibrium between ble for the excessive vibrational population to higher vibrational
translational and vibrational temperatures) is determined (see energy levels, limiting the gas heating in the plasma.
Table III). Rigorously speaking, it should be for Tvibr =Tgas , but we The vibrational deactivation of N2 (X 1 Σþ
g , v) molecules in mix-
presented it as Tvibr =Trot in order to show the degree of vibrational tures containing O2 usually takes place through the mechanisms
non-equilibrium for different molecules at the same time. The ratio involving oxygen atoms (V–T N2 –O relaxation), as well as through
of Tvibr =Trot of N2 is about 8 at the end of the pulse (t  0:95 ms) the Zeldovich mechanism (R1).39 The rate coefficient for V–T N2 –
resulting in a significant vibrational non-equilibrium between the O relaxation was reported both experimentally and theoretically
gas and vibrational temperatures. At the same time, Tvibr =Trot of elsewhere.39,52,63 Based on available data, it exhibits a large rate of
NO goes down by a factor of about 4, which is probably a result of about 6  1014 cm3 s1 at 1000 K. Such a high rate makes the V–T
additional rotational pumping of NO(A) state, as suggested above. N2 –O process one of the dominant deactivation mechanisms of
To better understand the obtained results during the plasma vibrationally excited N2 . This competes with V–T N2 –N and V–V
pulse, the characteristic times for different energy exchanges are N2 –N2 processes39 becoming the main deactivation mechanism (V–
estimated in this work. The mean characteristic times for the main T N2 –O) for larger O2 content (.5%). Less relevant mechanisms

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-7

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

TABLE IV. Rates coefficients and the corresponding characteristic times estimated for different energy transfer channels in N2:O2 (1:1) discharges.

Characteristic time (τ)

3 −1
Reaction type Rate coefficient (cm s ) 1 Torr 10 Torr Comment Reference
−9 −3
e-V a
4.4 × 10 110 μs … Te = 1.47 eV, ne = 2 × 10 cm 12
43, 44
e-Va 2.1 × 10−9 … 9.3 μs Te = 0.75 eV, ne = 5 × 1013 cm−3 43, 44
V–V N2–N2 5.4 × 10−14, 4.9 × 10−14, 9 × 10−14 3.3 ms 0.33 ms Tgas = 1000 K 39, 45, 46b
V–V N2–NOc 1.6 × 10−14, 1.1 × 10−15 7 ms 0.7 ms Tgas ≈ 1000 K 17, 47b
V–V N2–O2 1.9 × 10−15, 5 × 10−15, 6 × 10−16 57 ms 5.7 ms Tgas = 1000 K 48, 49b
V–T O2–O 20 μs 2 μs Tgas = 1000 K 50
V–T NO–O 3 × 10−11 80 μs 8 μs Tgas ≈ 800 K 50
V–T N2–N 2 × 10−13 1 ms 0.1 ms Tgas = 1000 K 51
V–T N2–O 6.5 × 10−14, 6.1 × 10−14 2 ms 0.2 ms Tgas = 1000 K 39, 52b
V–T NO–O2 2.4 × 10−14, 3 × 10−14, 3.4 × 10−14 2 ms 0.2 ms Tgas = 300 K 53, 54, 55, 56b
V–T NO–NO 2.5 × 10−13, 4 × 10−13, 4.5 × 10−13 14 ms 2.2 ms Tgas = 1000 K 3, 57, 58, 59b
V–T O2–O2 60 ms 6 ms Tgas = 1000 K 50, 60
V–T O2–N2 150 ms 15 ms Tgas = 1000 K 50
V–T N2–O2 1.3 × 10−17 8s 0.8 s Tgas = 1000 K 61
−17
V–T N2–N2 10 , 1.1 × 10−17, 1.3 × 10−17 18 s 1.8 s Tgas = 1000 K 45, 46, 61b
R–T N2(C) 1.8 μs 0.18 μs Tgas = 1000 K 34, 36
R–T CO(B) 1.25 μs 0.125 μs Tgas = 1000 K 33, 34, 36
a
Electron impact vibrational excitation.
b
Averaged characteristic time.
c
For reaction: N2(X, v = 1) + NO(X, v = 0) → N2(X, v = 0) + NO(X, v = 1).

compared to the V–T N2 –O relaxation are the V–T N2 –O2 and V– for NO, and at the same time, an increase in Tvibr for N2 could be a
T O2 –N2 relaxations, as given in Table IV. result of a V–V energy transfer between N2 and NO. In this case, N2
A rather different behavior for the vibrational excitation of is being pumping vibrationally due to the energy exchange with NO
NO and N2 is found in our case, as shown in Figs. 4(c) and 4(d). (and with N2 as well). Note that the short V–T NO–O2 relaxation
Both species loose vibrational excitation at the beginning of the time (in our case) is estimated at room temperature, since there are
plasma pulse. This is likely due to the V–T processes for both NO no corresponding rates reported at high temperatures. Thus, we
and N2 where Tvibr is decreasing. N2 , however, gets quickly expect that this time might be even shorter at a higher temperature,
re-excited in the middle of the pulse also evolving much faster showing a decisive role of the NO vibrational relaxation by O2 .
having the growth time of about 1 ms. The characteristic times esti- The vibrationally excited NO molecules give a fast V–T self-
mated for V–V N2 –N2 energy transfer (giving 1.6 and 0.7 ms at 2 relaxation rate of about 4:5  1013 cm3 s1 , which is 5 orders of
and 5 Torr, respectively) are comparable to the observed growth magnitude faster than that of N2 .16 Although the corresponding
time of Tvibr of N2 . Based on Table IV, the V–V N2 –N2 energy rate constant is high, the NO density is not high enough in our
transfer gives a higher rate compared to the V–V N2 –O2 and V–V case (compared to either N2 or O2 density) to provide a significant
N2 –NO exchanges. This means that the V–V energy transfer for rate of this process. To the best of our knowledge, no experimental
N2 almost entirely occurs during the pulse as a result of an efficient data are available for NO relaxation by N2 yet.
pumping-up effect originated by V–V energy exchange of N2 . This The emission from the NO (A,0  X,2) band at 247 nm and
effect in N2 has been primarily pointed out by Capitelli et al.64 the O triplet at 844.6 nm are shown by OES imaging data in Fig. 5.
Unfortunately, due to the lack of literature data on rates for V–V This figure is composed by three sets of images captured by the
NO–NO collisions, the corresponding characteristic time was ICCD camera filtered using bandpass optical filters corresponding
not estimated in our case. On the other hand, the V–T N2 –N2 and to NO and O emissions. These data also give rough information on
V–T N2 –O2 energy transfers do not play a significant role, whereas the electron cloud evolution in the active zone during the discharge
the V–T N2 –O and V–T N2 –N exchanges occur predominantly. pulse. When looking at the NO emission at 2 Torr, after a delay of
More importantly, the high V–T NO–O2 and V–T NO–O relaxa- about 0.6 ms, the emitters occupy all the visible discharge area fol-
tion rates could be the main mechanisms responsible for Tvibr lowing the electron density. In addition, at any moment of time,
decrease of NO, probably signifying a V–T transfer during the the emission from NO at 2 Torr is relatively symmetric to the tube
pulse ON time. These might also point out on a stronger V–T center, whereas this behaves asymmetrically at 5 Torr implying the
transfer for NO, leading to gas heating. main excitation of NO and O occurs in the downstream. Further
Let us note a clear vibrational transfer taking place between NO discussion with respect to the plasma emission along the discharge
and N2 during the plasma ON time. As we can see, a decrease in Tvibr tube is provided below devoted to the pressure effect.

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-8

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

C. Space-resolved measurements
A space-resolved experimental investigation of the emission
intensity as well as Tvibr and Trot in a N2 :O2 (1:1) discharge has
been undertaken (see Fig. 6). These experiments were performed at
three different pressures: 0.5, 2, and 5 Torr. The applied power,
pulse frequency, and duty ratio were kept constant at 0.75 kW,
0.5 kHz, and 50%, respectively.
The emission intensities of excited species, such as N2
(337.2 nm) and O (844.6 nm), reveal maximum near the waveguide,
whereas they decrease at a higher distance from the waveguide, as
evident in Fig. 6(a). This observation is a result of the electron
impact excitation near the waveguide.12 Based on Moisan et al.,6
the plasma sustained by a traveling electromagnetic wave is axially
and radially inhomogeneous and the plasma density decreases
going away from the wave launcher.
The axial distribution of Tvibr of N2 in N2 –O2 plasma
columns is shown in Fig. 6(b). We can see that vibrational tempera-
ture is likely having a maximum in the center at 5 Torr, whereas it
is clearly shifted upstream at low pressure (at 2 Torr). The observed
earlier vibrational excitation of N2 in the upstream at lower

FIG. 6. Space-resolved profiles measured in a pulsed MW discharge perpendic-

FIG. 5. OES imaging data representing the time evolution of the NO(A,0 - X,2) ularly to the discharge tube. Relative intensity of N2 at 337.2 nm and O at
rotational band (a) and (b); and O triplet at 844.6 nm (c), measured in N2 :O2 844.6 nm (a); Tvibr of the N2 (C) state (b); and Trot obtained by N2 (C,0  B,0)
(1:1) discharge at 2 and 5 Torr. Pulse duration = 1 ms, f = 0.5 kHz, p = 0.7 kW. rotational band (c). The measurements were performed at the middle of the
The central part is screened by the waveguide. The emission intensity from NO plasma pulse (0.5 ms). More information about the discharge is available in the
and O is separately normalized in each series. legend.

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-9

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

pressure is likely a result of the more efficient e-V transfer happen-

ing earlier in the upstream due to higher electron density.
Spatial profiles of Trot have the temperature maximum close to
the waveguide center (inaccessible region) at different discharge
pressures, as shown in Fig. 6(c). In particular, the shape of the Trot
distribution is more symmetrical at 0.5 Torr as the plasma fully
occupies the discharge tube. At 2 and 5 Torr, however, the fast Trot
increases in the upstream, and the slow Trot decreases in the down-
stream take place, likely due to a shorter plasma length and a dif-
ferent electron distribution (i.e., at higher pressure electrons are
closer to waveguide mainly). Interestingly, after the waveguide (in
the downstream), we observed almost similar temperature trends at
different discharge pressures.
In addition, the Trot elevation at the tube center is likely due
to the electron-translational (e-T) transfer because the electron
density reveals maximum near the waveguide vicinity.65 In addi-
tion, the higher V–T energy transfer at the tube center may also
cause an increase in gas heating. A similar conclusion had been
carried out by Silva et al.28 in experiments in a CO2 discharge. As a
result, a combination of these two processes (i.e., e-T + V–T energy
transfers) is the most probable outcome in this case.10
Overall, the found Trot behavior seem not being correlating
with the one of Tvibr : we can see that very different Tvibr behavior for
2 and 5 Torr result in the same Trot growth (and decay as well) both
before and after the waveguide. The higher Trot values obtained after
the waveguide at 2 and 5 Torr suggest that a stronger V–T energy
transfer occurs mainly in the downstream by increasing the pressure
and gas starts heating further away from the waveguide.

D. Power, pressure, and pulse duration effects

1. Power effect
The Lplasma variation and the evolution of the Tvibr and Trot are
examined as a function of the applied power taken at 3 and 4.6 Torr.
The Lplasma was estimated using OES imaging taken by an ICCD
camera. A continuous rise in Lplasma is found when power increases
[see Fig. 7(a)]. We believe that a higher applied power could result in
an increase in both electron temperature and density, facilitating a
large power coupling followed by a plasma length expansion.
A higher energy input results in the enhanced production of N-
and O-reactive species, primarily as a result of the electron impact
excitation as well as the vibrational excitation of N2 and O2 [see Fig. 7
(b)]. Those reactive species could then combine through the Zeldovich
mechanisms: (R1) and (R2) reactions, resulting in a higher NO con-
centration, as explained thoroughly in our previous work.26 A higher
applied power, however, facilitates the V–T relaxation processes when
the vibrational energy is transferred to gas heating [see Fig. 7(c)].

2. Pressure effect
Tvibr and Trot of N2 show opposite trends as the working pres-
sure increases in the reactor, as shown in Fig. 8. In particular, in
the pressure range of 0.6–10 Torr, Trot grows by about 270 K,
whereas Tvibr (of N2 (C) state) decreases by about 2000 K. Note FIG. 7. Lplasma (a); Tvibr (b); and Trot (c), determined as a function of the
that an increase in gas flow rate is associated with the increase of applied power. The measurements were taken at f ¼ 0:5 kHz with a duty ratio
of 50% at different discharge pressures. The plasma length was estimated using
the discharge pressure in our case. The measurements were per- OES imaging taken by an ICCD camera without optical bandpass filters.
formed in N2 :O2 (1:1) gas mixtures at a mean applied power of

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-10

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

TABLE V. Total gas flow (F), corresponding discharge pressures (P), gas velocities
(vgas) in the quartz tube, gas residence time (τres), as well as degree of vibrational
non-equilibrium (Tvibr/Trot) estimated in this study. The pulse repetition rate and
power were kept at 0.5 kHz and 0.9 kW, respectively.

F (slm) P (Torr) vgas (m s−1) τres (ms) Tvibr/Trota (for N2)

0.12 0.6 43 4.7 10
1 3 78 2.1 8.6
1.75 4.6 90 1.6 8.1
2 5 96 1.4 7.9
4.5 10 130 0.6 5.6
a
Based on Fig. 7.

higher pressure. This non-uniformity might be a result of the gas

flow direction, but the excitation areas reduce almost symmetrically
from 1 to 10 Torr in the downstream.

FIG. 8. Tvibr and Trot measured in the pulse MW discharge using N2 :O2 (1:1) 3. Pulse duration effect
gas mixtures as a function of the discharge pressure. The measurements were
taken at a p ¼ 0:9 kW, f ¼ 0:5 kHz with duty ratio of 50%. To study the pulse duration effect on ro-vibrational kinetics,
the plasma pulse ON time was varied where the mean power

0.9 kW and at a pulse repetition rate of 0.5 kHz with duty ratio of
50%. Lplasma contraction appears as pressure increases, leading to
a significant change in the plasma characteristics. At low pressure
in which electron temperature is high,27 the electron energy is
channelized to vibrational excitation of various plasma molecules.
However, when the pressure increases, the collisional energy
exchange from vibrational to translational (V–T) degrees of
freedom increases66 leading to the gradual increase of gas temper-
ature, which is visible in Fig. 8. Since the characteristic time of V–
T relaxation for different collisions (such as N2 –N2 , N2 –O2 , N2 –
N, and N2 –O) is smaller at higher pressures (see Table IV), we
could expect a higher gas temperature occurs as a result of a more
efficient V–T energy transfer.
According to recent studies27,67 dedicated to CO2 conversion
in the same discharge geometry, a noticeable decrease in the degree
of discharge non-equilibrium was found as pressure increases. They
showed that the electron temperature decreases at high pressures.
In the present work, we, therefore, assume that the electron tem-
perature is most likely reduced at high pressures in N2 –O2 dis-
charge. Another indication of electron temperature decrease might
be the drop of vibrational excitation of N2 found in this case.
The degree of vibrational non-equilibrium for N2 drops from
10 to about 5.6 when the pressure increases from 0.6 to 10 Torr in
our case (as given in Table V). This table also shows an increase in
vgas as well as a drop in τ res with increasing the pressure as esti-
mated in the discharge tube. The short τ res in the plasma region
limits the population of higher vibrational levels. The higher gas
temperature, however, could depopulate the higher vibrational
levels and increase the recombination.68
The emission from NO(A,0  X,2) at 247 nm captured by
ICCD camera as a function of discharge pressure is shown in FIG. 9. OES imaging of the NO(A,0  X,2) rotational band located at 247 nm as
a function of discharge pressure measured in N2 :O2 (1:1) MW discharge. The
Fig. 9. This figure reveals that NO is more excited upstream at a
measurements are taken at p ¼ 0:9 kW, f ¼ 0:5 kHz with duty ratio of 50%.
lower pressure (at 1 Torr), whereas this displaces to downstream at

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-11

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

understanding of a surfaguide pulsed microwave discharge used for

nitrogen fixation. We have shown that the characteristic length of
optical emission from NO(A-X) transition gets contracted when
pressure increases, specifically at P
2 Torr gets shorter as the dis-
charge pressure increases from 1 to 10 Torr. The degree of vibra-
tional non-equilibrium (a ratio between Tvibr and Trot ) for N2
drops only by a factor of 2 when the pressure increases about 17
times in our case. The gas heating is not consistent along the dis-
charge tube due to a discharge non-uniformity, especially at higher
pressure (at P
2 Torr). The measurements demonstrate that
working at a higher pressure results in a significant loss in vibra-
tional excitation. This has a negative effect on NO production due
to the gas temperature elevation and recombination processes, as
discussed in our recent work.26
This work also demonstrates that the power modulation is
beneficial to reduce the gas heating as it can effectively suppress the
unfavorable V–T N2 –N, V–T N2 –O, as well as V–T NO–O2 energy
transfers. The gas residence time plays a decisive role in an efficient
NO formation. The plasma pulse duration itself is of great impor-
tance as a higher vibrational excitation could be achieved when
pulse duration, gas residence time, and the V–V energy transfer
FIG. 10. Evolutions of the Tvibr and Trot as a function of duty ratio measured in times are comparable. The residence time estimated in our work
N2 :O2 gas mixtures. Further information regarding the discharge is depicted in
the legend. The error bars indicated for Tvibr correspond to the Boltzmann plot
(ranges from 0.6 to 4.7 ms) is comparable to the V–V N2 –N2
fitting error. energy transfer time (ranges from 0.33 to 3.3 ms), meaning that the
V–V N2 –N2 relaxation time is affected by gas residence time. This
likely offers a direct link between the dominant N2 dissociation
pathway and vibrational excitation, known to be an efficient means
injected per pulse was kept constant (see Fig. 10). The pulse dura- for nitrogen fixation. The V–T N2 –N and V–T N2 –O relaxation
tion is defined as the amount of power applied during the plasma times are affected by gas residence time as well, having the time
pulse ON time. The experiments were carried out along the dis- about 0.1–2 ms.
charge tube at a pulse repetition rate of 0.5 kHz, at a mean power Remarkably, the time-evolutions of N2 and NO vibrational
of 0.7 kW, and a discharge pressure of 2 Torr. excitation appear different. In particular, the slow decrease in vibra-
As it can be seen from Fig. 10, a longer pulse results in a tional excitation of NO appears during the pulse ON time, whereas
slightly higher Tvibr , whereas Trot is barely sensitive to the pulse N2 undergoes a fast decrease up to the middle of the pulse ON
duration. In addition, the degree of vibrational non-equilibrium time, then followed by a slow increase. This points out a stronger
slightly increases from 9 to 11 when the pulse duration rises from V–T energy transfer for NO due to the high V–T NO–O2 and V–T
0.4 to 1.4 ms (see inset in Fig. 10). The trend can be explained by NO–O relaxation rates.
the τ res effect, since this time (1.4 ms estimated at 2 Torr) is compa- A correlation between the spatial drop in Tvibr of N2 and slow
rable to the pulse duration in our case. This leads to a vibrational gas temperature decay in the downstream is found. Based on the
excitation elevation of N2 at a pulse duration equal to τ res provok- OES imaging data, the emission from NO tends to be symmetric at
ing a V–V exchange growth. Although the longer pulses imply a a lower pressure (up to 2 Torr), whereas it displaces to downstream
slightly higher vibrational excitation for vibrational ladder climbing, as pressure increases. This results in gas heating elevation occurred
a pulsed MW plasma with a longer pulse duration is, thus, reflected mainly at downstream, especially at higher pressure where the
less interest from an energy cost point of view. This can be due to Tvibr =Trot ratio is low. The non-equilibrium gradually disappears as
more power consumption, but not displaying a significantly higher the pressure increases.
vibrational excitation. All in all, the results of this work show the importance of the
In our recent work,26 we have shown that a proper-tuning of discharge power interruption as well as the correlation between gas
the pulse repetition rate could play an important role in improving residence time, pulse duration, and characteristic times estimated
NO production. The pulse duration may additionally affect vibra- for different energy transfer channels in N2 –O2 discharge for the
tional excitation, especially if the plasma pulse duration is compa- optimum vibrational excitation.
rable with the V–T energy relaxation constant.

IV. CONCLUSIONS ACKNOWLEDGMENTS

In this work, we made an attempt to clarify the behavior of This research was supported by the FNRS-FWO project
the discharge parameters, such as the rotational and vibrational “NITROPLASM,” EOS O005118F. The authors thank Dr. Tiago
temperatures for N2 and NO species aiming at a deeper Silva (Lisbon University, Instituto Superior Instituto Superior

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-12

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

Tecnico) for providing the data for ro-vibrational exchange rate APPENDIX B. N2(C) AND CO(B) ROTATIONAL
coefficients. TEMPERATURES
Trot of N2 (C) and CO(B) taken at N2 -CO2 gas mixtures
AUTHOR DECLARATIONS during the plasma pulse is shown in Fig. 12. The differences in
Conflict of Interest rotational temperature between N2 (C) and CO(B) are already dis-
cussed above. The determination of Trot of CO(B) cannot be
The authors have no conflicts to disclose. achieved at a lower CO2 admixture since N2 band overlaps
with CO Angstrom band. Results show that an admixture of 80%
Author Contributions CO2 in N2 for Tgas determination had negligible effect on the Trot
Omid Samadi Bahnamiri: Conceptualization (equal); Investigation of N2 , giving almost the same value (for Trot of N2 ) in N2 –O2 and
(lead); Methodology (lead); Writing – original draft (lead). Filippo CO2 –N2 plasmas. CO2 admixture will certainly change other prop-
Manaigo: Conceptualization (supporting); Investigation (support- erties of the plasma, e.g., plasma chemistry. However, in this paper,
ing); Methodology (supporting); Writing – review & editing (sup- we only used CO2 gas to determine Tgas through CO(B) band, as
porting). Abhyuday Chatterjee: Conceptualization (equal); well as to study the fundamental plasma physics of gas heating
Methodology (supporting); Validation (supporting); Writing – during a plasma pulse.
review & editing (supporting). Rony Snyders: Conceptualization
(equal); Supervision (equal); Validation (equal); Writing – review
& editing (equal). Federico Antonio D’Isa: Conceptualization
(supporting); Software (supporting); Writing – review & editing
(supporting). Nikolay Britun: Conceptualization (equal);
Supervision (equal); Validation (equal); Writing – review & editing
(equal).

DATA AVAILABILITY
The data that support the findings of this study are available
within the article.

APPENDIX A. BOLTZMANN PLOT OF CO(B)

The emission of CO(B,0  A,3) rotational band at 561 nm,
acquired along the discharge, is used for Tgas determination in a
CO2 :N2 (4:1) gas mixture. A high-resolution structure of this band
where the emission peaks from the Q-branch used for the
Boltzmann fit is shown in Fig. 11. The corresponding Boltzmann
plot used for Trot determination is also depicted. A good agreement
with the Boltzmann distribution was found for the rotational Q
branch at J0 ¼ 8–22.

FIG. 12. Evolution of the Trot determined based on the CO(B,0  A,3) and N2
(C,0  B,0) rotational bands as a function of time delay measured in CO2 :N2
(4:1) gas mixtures.

REFERENCES
1
B. S. Patil, Q. Wang, V. Hessel, and J. Lang, Catal. Today 256, 49 (2015).
2
N. Cherkasov, A. O. Ibhadon, and P. Fitzpatrick, Chem. Eng. Process. Process
Intensif. 90, 24 (2015).
3
A. Fridman, Plasma Chemistry (Cambridge University Press, Cambridge, 2008).
4
V. D. Rusanov, A. A. Fridman, and G. V. Sholin, Sov. Phys. Usp. 24, 447
(1981).
5
M. Moisan, M. Chaker, Z. Zakrzewski, and J. Paraszczak, J. Phys. E 20, 1356
(1987).
FIG. 11. Structure of the CO(B,0  A,3) rotational band from the Angstrom CO 6
M. Moisan and H. Nowakowska, Plasma Sources Sci. Technol. 27, 073001
system taken in CO2 :N2 (4:1) gas mixtures. A Boltzmann plot (see inset) corre-
(2018).
sponding to the marked rotational peaks utilized for Trot determination is also 7
depicted. C. Boisse-Laporte, C. Chave-Normand, and J. Marec, Plasma Sources Sci.
Technol. 6, 70 (1997).

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-13

Published under an exclusive license by AIP Publishing
 Journal of ARTICLE scitation.org/journal/jap
Applied Physics

8 35
S. van Alphen, V. Vermeiren, T. Butterworth, D. C. van Den Bekerom, G. J. van G. Dilecce, Plasma Sources Sci. Technol. 23, 015011 (2014).
36
Rooij, and A. Bogaerts, J. Phys. Chem. C 124, 1765 (2020). J. G. Parker, Phys. Fluids 2, 449 (1959).
9 37
V. Vermeiren and A. Bogaerts, J. Phys. Chem. C 123, 17650 (2019). P. J. Bruggeman, N. Sadeghi, D. C. Schram, and V. Linss, Plasma Sources Sci.
10
N. Britun, T. Silva, G. Chen, T. Godfroid, J. Van Der Mullen, and R. Snyders, Technol. 23, 023001 (2014).
38
J. Phys. D: Appl. Phys. 51, 144002 (2018). V. N. Ochkin, Spectroscopy of Low Temperature Plasma (Wiley-VCH Verlag
11
V. Jivotov, J. Phys. IV 8, 401–410 (1998). GmbH & Co. KGaA, Weinheim, 2009).
12 39
E. Benova, T. Petrova, A. Blagoev, and I. Zhelyazkov, J. Appl. Phys. 84, 147 V. Guerra, A. Tejero-del Caz, C. D. Pintassilgo, and L. L. Alves, Plasma
(1998). Sources Sci. Technol. 28, 073001 (2019).
13 40
M. Capitelli, R. Celiberto, G. Colonna, F. Esposito, C. Gorse, K. Hassouni, N. A. Popov, J. Phys. D: Appl. Phys. 44, 285201 (2011).
A. Laricchiuta, and S. Longo, “Fundamental aspects of plasma chemical physics: 41
A. F. Van Gessel and P. J. Bruggeman, J. Chem. Phys. 138, 204306 (2013).
42
Kinetics,” in Springer Series on Atomic, Optical, and Plasma Physics (Springer, D. Staack, B. Farouk, A. F. Gutsol, and A. A. Fridman, Plasma Sources Sci.
New York, NY, 2015). Technol. 15, 818 (2006).
14 43
J. Loureiro, V. Guerra, P. A. Sá, C. D. Pintassilgo, and M. L. D. Silva, Plasma The LXCat open-access, LXCat.
44
Sources Sci. Technol. 20, 024007 (2011). L. L. Alves, J. Phys. Conf. Ser. 565, 12007 (2014).
15 45
V. Guerra and J. Loureiro, J. Phys. D: Appl. Phys. 28, 1903 (1995). L. Terraz, T. Silva, A. Morillo-Candas, O. Guaitella, A. Tejero-del Caz,
16
I. V. Adamovich, S. Saupe, M. J. Grassi, S. O. Macheret, and J. William Rich, L. L. Alves, and V. Guerra, J. Phys. D: Appl. Phys. 53, 094002 (2020).
46
in AIAA 24th Plasma Dynamics, and Lasers Conference 1993 (AIAA, 1993), G. D. Billing and E. R. Fisher, Chem. Phys. 43, 395 (1979).
47
Vol. 174, pp. 3199. R. L. Taylor, M. Camac, and R. M. Feinberg, Symp. Combust. 11, 49 (1967).
17 48
M. E. Whitson and R. J. McNeal, J. Chem. Phys. 66, 2696 (1976). F. R. Gilmore, E. Bauer, and J. W. McGowan, J. Quant. Spectrosc. Radiat.
18
R. I. Asisov, V. K. Givotov, V. D. Rusanov, and A. Fridman, Sov. Phys. 14, 366 Transf. 9, 157 (1969).
49
(1980). R. L. Taylor and S. Bitterman, Rev. Mod. Phys. 41, 26 (1969).
19 50
S. Kelly and A. Bogaerts, Joule 5, 3006 (2021). G. P. Oblapenko, J. Phys. Chem. A 122, 9615 (2018).
20 51
L. S. Polak, A. A. Ovsiannikov, D. I. Slovetsky, and F. B. Vurzel, M. Nauk. 46, F. Esposito, M. Capitelli, and C. Gorse, Chem. Phys. 257, 193 (2000).
52
295 (1975). A. Starikovskiy and N. Aleksandrov, Prog. Energy Combust. Sci. 39, 61
21
B. S. Patil, N. Cherkasov, J. Lang, A. O. Ibhadon, V. Hessel, and Q. Wang, (2013).
53
Appl. Catal. B 194, 123 (2016). B. Bohn, A. Doughty, G. Hancock, E. L. Moore, and C. Morrell, Phys. Chem.
22
X. Pei, D. Gidon, Y. J. Yang, Z. Xiong, and D. B. Graves, Chem. Eng. J. 362, Chem. Phys. 1, 1833 (1999).
54
217 (2019). R. P. Fernando and I. W. Smith, J. Chem. Soc. Faraday Trans. Mol. Chem.
23
X. Pei, D. Gidon, and D. B. Graves, J. Phys. D: Appl. Phys. 53, 044002 (2020). Phys. 77, 459 (1981).
24 55
N. Britun, V. Gamaleev, and M. Hori, Plasma Sources Sci. Technol. 30, B. D. Green, G. E. Caledonia, R. E. Murphy, and F. X. Robert, J. Chem. Phys.
08LT02 (2021). 76, 2441 (1982).
25 56
F. Jardali, S. Van Alphen, J. Creel, H. Ahmadi Eshtehardi, M. Axelsson, G. Hancock, M. Morrison, and M. Saunders, Chem. Phys. Lett. 425, 216
R. Ingels, R. Snyders, and A. Bogaerts, Green Chem. 23, 1748 (2021). (2006).
26 57
O. S. Bahnamiri, C. Verheyen, R. Snyders, A. Bogaerts, and N. Britun, Plasma K. L. Wray, J. Chem. Phys. 36, 2597 (1962).
58
Sources Sci. Technol. 30, 065007 (2021). H. E. Bass and G. L. Hill, J. Chem. Phys. 5179, 5179 (1973).
27 59
N. Britun, T. Godfroid, and R. Snyders, J. CO2 Util. 41, 101239 (2020). G. Kamimoto and H. Matsui, J. Chem. Phys. 53, 3987 (1970).
28 60
T. Silva, N. Britun, T. Godfroid, and R. Snyders, Plasma Sources Sci. Technol. R. W. Lutz and J. H. Kiefer, Phys. Fluids 9, 1638 (1966).
61
23, 025009 (2014). V. A. Sal’nikov and A. M. Starik, High Temp. 33, 121 (1995).
29 62
J. Voráč, L. Kusýn, and P. Synek, Rev. Sci. Instrum. 90, 123102 (2019). M. Capitelli, G. Colonna, G. D’Ammando, A. Laricchiuta, and L. D. Pietanza,
30
D. Luque and J. Crosley, “LIFBASE: database and spectral simulation Plasma Sources Sci. Technol. 26, 034004 (2017).
63
program,” SRI Int. Rep. 9, 99 (1999); available at https://www.sri.com/platform/ R. J. McNeal, M. E. Whitson, and G. R. Cook, J. Geophys. Res. 79, https://doi.
lifbase-spectroscopy-tool/. org/10.1029/JA079i010p01527, 1527 (1974).
31 64
K. H. Becker, H. Engels, and T. Tatarczyk, Chem. Phys. Lett. 5, 111 (1977). M. Capitelli, C. Gorse, and G. D. Billing, Chem. Phys. 52, 299 (1980).
32 65
F. Di Teodoro, J. E. Rehm, R. L. Farrow, and P. H. Paul, J. Chem. Phys. 113, M. Moisan and Z. Zakrzewski, J. Phys. D: Appl. Phys. 24, 1025 (1991).
66
3046 (2000). A. Berthelot and A. Bogaerts, J. Phys. Chem. C 121, 8236 (2017).
33 67
C. A. Brau and R. M. Jonkman, J. Chem. Phys. 52, 477 (1970). G. Chen, R. Snyders, and N. Britun, J. CO2 Util. 49, 101557 (2021).
34 68
F. J. Uribe, E. A. Mason, and J. Kestin, J. Phys. Chem. Ref. Data 19, 1123 S. Kelly, A. van de Steeg, A. Hughes, G. van Rooij, and A. Bogaerts, Plasma
(1990). Sources Sci. Technol. 30, 055005 (2021).

J. Appl. Phys. 133, 113303 (2023); doi: 10.1063/5.0138298 133, 113303-14

Published under an exclusive license by AIP Publishing