Parallel 1d3v Particle in Cell/Monte Carlo Collision (PIC/MCC)
Simulation of a Glow Discharge Millimeter Wave Detector
Cemre Kusoglu-Sarikaya, Hakan Altan and Demiral Akbar
Department of Physics, Middle East Technical University, 06800, Ankara, Turkey
Keywords:
Terahertz (THz), Glow Discharge Detector.
Abstract:
Glow discharge detectors can be a good alternative to existing Schottky diodes, Golay cells and pyroelectric
detectors because they are inexpensive and can detect mm-wave and sub-mm radiation successfully. This
detection occurs as a result of the interaction of the radiation with the electrons in the plasma. It is required
to understand this interaction mechanism to obtain optimum detection parameters. Previous methods have
focused on understanding the interaction using analytical models, where the radiation is generally thought to
increase the collision frequency of electrons in the plasma, however these theories were not tested against
real discharge parameters. For that reason, in this study, the plasma formed inside the detector is simulated
by using parallel 1d3v PIC/MCC code, which was previously developed (Kusoglu-Sarikaya et al., 2016) to
better understand how the glow discharge forms under different pressure and gas concentrations. The effec-
tiveness of the simulation is compared with mm-wave experiments performed on both commercially obtained
and home-built glow discharge detectors. Initial results show that the 1d3v PIC/MCC code can simulate the
discharge parameters that are observed in the measurements. Using this platform future studies will focus on
understanding the effect of the sub-THz radiation on the collision frequency and observed parameters of the
discharge.
1 INTRODUCTION
Technologies and applications that are based on mm
waves (30-100GHz) have gained much attention re-
cently. Historically, once a difficult region in the elec-
tromagnetic spectrum for the development of sources
and detectors, their potential for use in civil and de-
fense applications technologies have been driving re-
search rapidly in the laboratory. One of the best com-
mercial examples of their use has been in the devel-
opment of imaging systems and their use in airport
security screening areas. The use of these systems
especially in the USA has made air travel even safer.
One of the main disadvantages of these systems is that
the sources and detectors used are still expensive to
produce. Since mm-wave photon energies are on the
order of meV and the background levels (blackbody)
are typically high the methods used in the manufac-
ture of the sources and detectors require advanced ma-
terials science development as well as cryogenic cool-
ing which tends to increase the overall cost of these
systems.
Current commercial direct detectors used to de-
tect THz waves are Schottky diodes, Golay cells and
pyroelectric detectors. However, these detectors are
expensive and have limitations in terms of speed and
responsivity (Hou et al., 2011). One research area,
first started in the 1970s, which was to use glow dis-
charge lamps to detect THz and millimeter waves has
become relevant again in the international arena due
to the high interest in developing THz technologies.
These recent studies show that glow discharge detec-
tors (GDDs) have no such limitations as with other
room temperature commercial detectors (Abramovich
et al., 2007). Moreover, they are very cheap, which
gives these detectors an extra advantage.
In order to understand the detection mechanism
of GDDs, it is necessary to understand what is in-
volved in the mm wave-plasma interaction. First,
it is known that THz waves can penetrate through
the plasma because the plasma frequency in the dis-
charge tube (sub-GHz to GHz range) is smaller than
the THz frequency (Kopeika, 1978). In addition, the
mm wave radiation and the plasma are thought to in-
teract in two ways: Cascade ionization and diffusion
current (Kopeika, 1975). While this gives a broad un-
derstanding of the method of interaction, the physi-
cal phenomena resulting from this interaction which
are observed in experiments have not been studied
in detail. Thus how the radiation interacts with the
110
Kusoglu-Sarikaya, C., Altan, H. and Akbar, D.
Parallel 1d3v Particle in Cell/Monte Carlo Collision (PIC/MCC) Simulation of a Glow Discharge Millimeter Wave Detector.
DOI: 10.5220/0006716001100115
In Proceedings of the 6th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2018), pages 110-115
ISBN: 978-989-758-286-8
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

plasma is still not known exactly. Experiments per-
formed by our group has shown that the mm wave ra-
diation is detected by a change in the plasma current
for commercially available discharge lamps, however
the same experiments when performed with a home-
built discharge chamber were inconclusive suggesting
that gas composition, pressure as well as electrode
composition all play an important role in the detec-
tion of the radiation (Alasgarzade et al., 2016). To
better understand the role these parameters play the
GDD simulated using the parallel 1d3v PIC / MCC
code previously developed (Kusoglu-Sarikaya et al.,
2016) and the results compared with controlled ex-
periments performed in the laboratory. These results
will aid in understanding the plasma parameters that
dominate the detection observed with commercial de-
tectors. This will be done later by adding the energy
associated with the radiation to the plasma allowing
to better understand the physical reactions of the par-
ticles inside.
2 GLOW DISCHARGE CHAMBER
Commercially available neon indicator lamps, namely
glow discharge detectors (GDDs) have been our
primary research interest since they are proven to
be cost-effective, low noise, fast response mm-
wave/Terahertz (THz) detectors (Rozban et al., 2008;
Abramovich et al., 2007). Previously, we have in-
vestigated various GDDs in terms of their speed, fre-
quency response and polarization dependence based
on their orientation with respect to the incident light
(Alasgarzade et al., 2016). In order to investigate con-
tributing effects to mm-wave/THz detection mecha-
nism such as the effects of discharge breakdown and
glow scenarios for various inert gasses as well as var-
ious Penning mixtures, a small plasma vacuum cham-
ber is designed and built, shown in Fig. 1 and Fig. 2.
A breakdown is achieved in gas mixture by apply-
ing a bias DC voltage to the electrodes. Under suf-
ficient conditions the electric field of the modulated
incident radiation increases the total electric field and
can generate variations in the plasma current. The ef-
fect of incident radiation on the plasma current de-
pends on several parameters such as plasma region,
type of the gas mixture, electrode geometry and po-
larization of radiation. Based on the measured dis-
charge current values, we expect to operate within the
abnormal regime of the glow, before the arc region
(Braithwaite, 2000). In this region the electric field of
the incoming mm wave radiation, when aligned with
the applied DC field is expected to increase the rate of
excitation collision (Rozban et al., 2008).
Figure 1: The gas mixer system designed and built to mix
the inert gases inside the home-built chamber allows up to 3
different gases to mixed under normal (partial atmosphere)
to low vacuum (0.01 Torr) conditions. The gases are mixed
inside a housing before being sent into the chamber.
Figure 2: Plasma Discharge Chamber. The electrode sepa-
ration is controlled using a micrometer.
The vacuum chamber having a dimension of
roughly 10x10x10 cm
3
, has two quartz windows with
40 mm diameter that allow the transmission of inci-
dent THz radiation through the DC glow discharge
between the electrodes. Electrodes with different ge-
ometries can be used. The electrode separation can be
controlled with 10 µm resolution and can be extended
up to 2 cm. Also the chamber admits a floating probe
allowing the measurement of changes in plasma cur-
rent and plasma voltage. There are two feedthroughs
on the top of plasma chamber. One of them is con-
nected to the gas distribution system and the other is
connected to the Multi-Gauge Controller (Varian) for
measuring pressure inside chamber instantaneously.
With the rotary vane pump system, the pressure in-
side chamber can be reduced to 10
−2
torr.
After sealing the pump line, routine operations of
plasma glow are achieved with a backfill pressure of
25 torr for pure Neon Gas (99.999%). The breakdown
is achieved at around 350 V for a 1 mm separation be-
tween anode and cathode. Compared to commercially
available Neon lamps these values are thought to be
much higher. In commercially available Neon lamps
typically discharges are obtained for breakdown volt-
ages 80-150 V for similar electrode spacing and pres-
sure. Thus the commercially available Neon lamps
Parallel 1d3v Particle in Cell/Monte Carlo Collision (PIC/MCC) Simulation of a Glow Discharge Millimeter Wave Detector
111

are thought to contain a mixture of gases. In order to
better understand the interaction of the mm wave ra-
diation with the glow simulations were carried out to
see if one could simulate the observed experiments.
3 MODEL
Parallel 1d3v PIC/MCC code, which was developed
previously (Kusoglu-Sarikaya et al., 2016), is used to
simulate the glow discharge detector by using neon
gas and Ne-Ar mixture separately. This code includes
elastic scattering, excitation and direct ionization pro-
cesses between electrons and neutrals as in Fig. 3.
Since it is well known that Ne(
3
P
2
) is responsible
for Penning ionization of argon gas (Kopeika, 1978),
only the excitation which causes the formation of this
metastable atom has been taken into account. Fluid
approximation,
∂n
∂t
− D
∂
2
n
∂x
2
= S, (1)
is incorporated with the PIC/MCC numerical
model to analyze the density distribution of Ne(
3
P
2
)
metastable atoms. Here, D is the diffusion coefficient
and S is the excitation source term. Metastable neon
atoms are assumed to be absorbed at the boundaries.
The diffusion coefficient was taken (Phelps and Mol-
nar, 1953; Phelps, 1959) as 150 cm
2
s
−1
torr.
In addition, it is assumed that mainly isotropic
scattering and charge transfer collisions occur be-
tween ions and neutrals and cross-section values are
taken (Cramer, 1958) as 2 × 10
−19
and 3 × 10
−19
m
2
respectively.
10
-25
10
-24
10
-23
10
-22
10
-21
10
-20
10
-19
10
-18
0 20 40 60 80 100
Cross section (m
2
)
Electron energy (eV)
Elastic
Excitation (Ne(
3
P
2
))
Direct ionization
Figure 3: Electron cross-sections for elastic, excitation
and direct ionization collisions in neon, used in the
model. Cross-sections were taken from www.lxcat.net (Bi-
agi (Magboltz versions 8.9 and higher)).
Simulation of the Ne-Ar mixture also includes
Penning ionization reaction with the cross-section
(Fridman and Kennedy, 2004; Franz, 2009) of 1 ×
10
−19
m
2
,
Ne(
3
P
2
) + Ar → Ar
+
+ Ne + e. (2)
Here, the ionization energy of argon is 15.76 eV
and the energy of metastable neon atom is 16.62 eV.
Thus, it is expected that electrons with an energy of
0.86 eV will form as a result of this reaction.
It is also known that the reactions occurring at the
boundaries have an important role in the formation of
plasma. Therefore, ion induced secondary electron
emission and electron reflection are considered and
they were assumed to have coefficient value of 0.2.
Remaining parameters used in the GDD simulation
are summarized in Table 1.
Table 1: The Parameters Used in the PIC/MCC simulation
of GDD.
Species e, Ne
+
, Ne(
3
P
2
)
Weighting (10
8
) 1
Grid Number 600
Time step (s) 1 × 10
−12
Distance (mm) 1
Neutral Temperature (K) 300
Pressure (Torr) 25
Voltage (Volt) 110
4 SIMULATION RESULTS (PURE
NEON GAS)
Electric field and potential profiles can be seen in
Fig. 4 and Fig. 5. According to these figures, the
quasineutral region is between 0.1 and 0.5 mm. This
quasineutrality can be seen more clearly in the den-
sity distribution profiles (see Fig. 6). In addition, the
Cathode region appears near 1 mm with strong elec-
tric field.
Fig. 7 and Fig. 8 illustrate the mean energy dis-
tribution of electrons and ions between the elec-
trodes. As expected, highly energetic particles are
seen near the cathode. However, in quasineutral re-
gion, low energetic particles constitute the majority
as the frequency of the collisions increases in this re-
gion. These low and high energetic particles cause the
energy distribution function profiles to consist of two
different regions (Fig. 9 and Fig. 10).
High electron current density and low ion current
density are seen in Fig. 11 and Fig. 12. However, it
is noteworthy that the electron current density is very
noisy. This noise can be reduced by decreasing the
weighting value of the super particle number.
PHOTOPTICS 2018 - 6th International Conference on Photonics, Optics and Laser Technology
112

0
20
40
60
80
100
120
0 0.2 0.4 0.6 0.8 1
Potential (V)
x (mm)
Figure 4: Potential profile obtained by using pure neon gas.
-100
0
100
200
300
400
500
0 0.2 0.4 0.6 0.8 1
Electric field (kV/m)
x (mm)
Figure 5: Electric field profile obtained by using pure neon
gas.
0
5
10
15
20
25
0 0.2 0.4 0.6 0.8 1
Density (10
16
/m
3
)
x (mm)
Electrons
Ions
Ne(
3
P
2
)
Figure 6: Electron, ion and metastable neon density profiles
obtained by using pure neon gas.
5 SIMULATION RESULTS
(NEON-ARGON MIXTURE)
Simulation results were re-examined by adding 1 per-
cent argon gas to the neon gas. For the Ne-Ar gas
mixture (1% Argon, 99% Neon), a plasma could be
0
10
20
30
40
50
60
0 0.2 0.4 0.6 0.8 1
Electron mean energy (eV)
x (mm)
Figure 7: Electron mean energy distribution obtained by us-
ing pure neon gas.
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1
Ion mean energy (eV)
x (mm)
Figure 8: Ion mean energy distribution obtained by using
pure neon gas.
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
0 5 10 15 20
EEDF (eV
-3/2
)
E (eV)
Figure 9: Electron energy distribution function as a function
of electron kinetic energy obtained by using pure neon gas.
obtained at 110 Volts (as in pure neon gas). The in-
crease in electron and ion density with the effect of
Penning ionization is clearly visible in Fig. 13 and
Fig. 14. Given this increase, it turns out that when
a pure neon gas is used, a higher voltage is needed
to achieve the same profile. This clearly indicates the
Parallel 1d3v Particle in Cell/Monte Carlo Collision (PIC/MCC) Simulation of a Glow Discharge Millimeter Wave Detector
113

10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
0 1 2 3 4 5 6
IEDF (eV
-3/2
)
E (eV)
Figure 10: Ion energy distribution function as a function of
electron kinetic energy obtained by using pure neon gas.
-400
-200
0
200
400
0 0.2 0.4 0.6 0.8 1
Electron current density (A/m
2
)
x (mm)
Figure 11: Electron current density obtained by using pure
neon gas.
-2
0
2
4
6
8
10
12
14
0 0.2 0.4 0.6 0.8 1
Ion current density (A/m
2
)
x (mm)
Figure 12: Ion current density obtained by using pure neon
gas.
reason for the use of the Ne-Ar mixture.
0
5
10
15
20
25
0 0.2 0.4 0.6 0.8 1
Density (10
16
/m
3
)
x (mm)
Electrons
Ions
Ne(
3
P
2
)
Figure 13: Electron, ion and metastable neon density pro-
files obtained by using Ne-Ar mixture.
0
5
10
15
20
25
0 0.2 0.4 0.6 0.8 1
Electron density (10
16
/m
3
)
x (mm)
Pure Ne
Ne-Ar mixture
Figure 14: Comparison of the electron density profiles ob-
tained by using pure Ne and Ne-Ar mixture, separately.
6 CONCLUSIONS
Parallel 1d3v PIC/MCC (Kusoglu-Sarikaya et al.,
2016) simulation of GDD filled with neon gas and
Ne-Ar mixture is performed separately and compared
with experimental results. It is seen that the simula-
tion describes the observed experimental parameters
adequately for the home-built glow chamber. Mix-
tures of gases reduce the required breakdown volt-
age due to Penning ionization and the obtained elec-
tron densities agree well with expected values. Fu-
ture work will target the effect of mm wave radia-
tion on the energy distributions of electrons inside the
plasma. By obtaining an understanding of the param-
eters associated with the glow, better GDD structures
can be designed and implemented for mm wave ap-
plications.
PHOTOPTICS 2018 - 6th International Conference on Photonics, Optics and Laser Technology
114

ACKNOWLEDGEMENTS
The work was supported by the Scientific and
Technical Research Council of Turkey (TUBITAK)
115F226. The simulations were performed us-
ing High Performance and Grid Computing Center
(TRUBA Resources) at TUBITAK ULAKBIM.
REFERENCES
Abramovich, A., Kopeika, N. S., Rozban, D., and Farber, E.
(2007). Appl. Opt., 46:7207.
Alasgarzade, D. N., Takan, T., Mansuroglu, D., Sahin,
A. B., Uzun-Kaymak, I. U., and Altan, H. (2016).
Interaction of a narrow gap glow discharge plasma
with far infrared radiation. In 41st International Con-
ference on Infrared, Millimeter, and Terahertz waves
(IRMMW-THz).
Braithwaite, N. S. J. (2000). Plasma Sources Sci. Technol.,
9:517.
Cramer, W. H. (1958). J. Chem. Phys., 28:688.
Franz, G. (2009). Low Pressure Plasmas and Microstruc-
turing Technology. Springer.
Fridman, A. and Kennedy, L. A. (2004). Plasma Physics
and Engineering. CRC Press.
Hou, L., Park, H., and Zhang, X. C. (2011). IEEE J. Sel.
Top. Quantum Electron, 17:177.
Kopeika, N. S. (1975). Proc. IEEE, 63:981.
Kopeika, N. S. (1978). IEEE T. Plasma Sci., 6:139.
Kusoglu-Sarikaya, C., Rafatov, I., and Kudryavtsev, A. A.
(2016). Phys. Plasmas, 23:063524–1.
Phelps, A. V. (1959). Phys. Rev., 114:1011.
Phelps, A. V. and Molnar, J. P. (1953). Phys. Rev., 89:1202.
Rozban, A. D., Kopeika, N. S., Abramovich, A., and Farber,
E. (2008). J. Appl. Phys., 103:093306.
Parallel 1d3v Particle in Cell/Monte Carlo Collision (PIC/MCC) Simulation of a Glow Discharge Millimeter Wave Detector
115