APPLIED PHYSICS LETTERS 90, 051504 共2007兲
Atmospheric-pressure argon/oxygen plasma-discharge source
with a stepped electrode
Jin-Pyo Lim and Han S. Uhma兲
Department of Molecular Science and Technology, Ajou University, San 5 Wonchon-Dong,
Youngtong-Gu, Suwon 443-749, Korea
Shou-Zhe Li
State Key Laboratory of Materials Modification by Laser, Ion and Electron Beams
and Department of Physics, Dalian University of Technology, Dalian 116024, China
共Received 4 December 2006; accepted 6 January 2007; published online 2 February 2007兲
The nonequilibrium glow discharge in argon mixed with oxygen at atmospheric pressure was
generated in a parallel plate reactor with a stepped electrode powered by a 13.56 MHz
radio-frequency power supplier. The stepped-electrode reactor consists of a narrow and wide gap
structure. A strong electric field occurred at the narrow gap region preionizes Ar/ O2 gas and assists
to generate a large volumetric plasma in the wide gap region. Therefore, the stepped-electrode
reactor makes it easy to operate Ar/ O2 glow discharge, providing a stable, uniform, and broad
plasma jet at atmospheric pressure. © 2007 American Institute of Physics.
关DOI: 10.1063/1.2450654兴
In recent years, atmospheric-pressure glow discharge
共APGD兲 sources have been developed to overcome the difficulties posed by low-pressure glow discharges which require expensive and complicated vacuum systems.1 These
APGD sources are the dielectric barrier discharge,2–4 the
one-atmospheric uniform glow discharge plasma,5 the cold
plasma torch,6,7 and the atmospheric-pressure plasma jet.8
Although some of the present APGDs operated in helium gas
have a relatively low breakdown voltage and a broad operating window in the glow discharge region, these sources
have problems related to long treatment time, restricted treatment space, and large consumption of working gas. There
have been several attempts to replace helium by argon as
working gas in APGDs.9,10 However, improvement of stable
argon plasma source is still required for the generation of
large volumetric and uniform glow discharge plasmas, when
electronegative gases are added, because O2, CF4, and SF6
are important in material processing such as ashing, etching,
cleaning, and sterilization. The atmospheric-pressure glow
discharges in Ar/electronegative gas are known to be difficult
to sustain and to operate in ␣ mode, even if the gap distance
between parallel electrodes is reduced to increase the electric
field in typical capacitively coupled rf plasma, because of a
high breakdown voltage due to a small mean free path of
argon and a narrow operating window.
In this study, we introduce an atmospheric-pressure
Ar/ O2 plasma-discharge source with a stepped-electrode,
generating a stable Ar/ O2 glow discharge and explain a
steady-state Ar/ O2 discharge in 13.56 MHz rf field in terms
of the electron number balance due to the attachment.11
A schematic of the experiment setup and the steppedelectrode configuration is shown in Figs. 1共a兲 and 1共b兲. The
stepped-electrode reactor consists of two parallel stainlesssteel plates of 3 ⫻ 7 cm2. The powered electrode was fixed in
the upper part of an insulating support and the grounded
electrode can be up and down along the z axis with a micrometer. Both bare metal electrodes were cooled by water
a兲
Electronic mail: [email protected] and [email protected]
0003-6951/2007/90共5兲/051504/3/$23.00
circulation and a quartz window was placed in the left lateral
side to observe the argon plasma discharge. Figure 1共b兲
shows the lateral side view of the parallel electrodes configured as stepped-electrode structure and experimental setup.
The rf generator operating at 13.56 MHz was connected to
the upper powered electrode through the matching network.
The grounded electrode has a step of 2 cm length and
1.3 mm height in the region labeled by A. Thus, the steppedelectrode reactor consists of two parts; a narrow gap region
labeled by A and a wide gap region labeled by B. The working gases, argon and oxygen, flowed into the plasma reactor
through the mass flow controller, and the argon gas flow was
10 l ” min and the oxygen gas flow was operated from
0 to 100 cc/ min. The gases pass through the narrow region
A and then flow into the wide region B in the x direction, as
shown in Fig. 1共b兲. In order to reduce the high breakdown
voltage of Ar/ O2 discharge, oxygen gas was added after generating pure argon plasma discharge. The voltage applied
between bare electrodes was measured by a high-voltage
probe 共Tektronix P6015A with bandwidth of 75 MHz兲, and
the results were recorded on an oscilloscope 共Tektronix
TDS320 with bandwidth of 100 MHz兲. Photo images of
Ar/ O2 discharges were recorded by a digital camera 共Cannon, power shot 1S3兲 in the y direction through a quartz
window and at front view in the x direction.
Plasma is sustained only when the ionization process
dominates electron loss. The main mechanism of removing
electrons in Ar/ O2 discharge is the electron attachment of
oxygen. Therefore, ionization frequency must be equal to
attachment frequency for steady-state discharge, i.e., the ionization coefficient ␣ = ␯i / ␷d, is equal to the attachment coefficient ␩ = ␯a / ␷d, where ␯i and ␯a are ionization and attachment frequencies, respectively, and ␷d is the electron drift
velocity. Therefore, the required electric field E for sustaining a steady-state discharge can be obtained from ␣eff = 0,
where ␣eff is the effective coefficient defined by ␣eff = ␣ − ␩.11
The ionization coefficient ␣ and the attachment coefficient ␩
are generally expressed as11–13
90, 051504-1
© 2007 American Institute of Physics
051504-2
Appl. Phys. Lett. 90, 051504 共2007兲
Lim, Uhm, and Li
to the instantaneous value of E共t兲, while the instantaneous
frequency of ionization is equal to the ionization frequency
␯i共E兲. Hence, the rf ionization frequency ␯i,rf is simplified to
be an average of ␯i关E共t兲兴 in time,11 i.e.,
␯i,rf =
␻
2␲
冕
2␲/␻
␯i共Ea兩sin ␻t兩兲dt =
0
2
␲
冕
␲/2
␯i共Ea sin ␸兲d␸ . 共4兲
0
Since ␯i共E兲 is a very steep function, ionization occurs as a
short surge when the field reaches its maximum value, while
the electrons disappear predominately during the rest of the
period. If the ionization coefficient is given in the Townsend
form of Eq. 共1兲, the expansion of 1 / sin ␸ in the neighborhood of ␸ = ␲ / 2 yields a formula useful for calculations,11
␯i,rf = ␯i共Ea兲冑2Ea/␲g j p.
共5兲
The square-root term of the right hand side in Eq. 共5兲 indicates the effective fraction of time during which ionization
occurs. Similar to Eq. 共5兲, the attachment frequency ␯a,rf in rf
field can be evaluated as11,13
␯a,rf = ␯a共Ea兲冑2Ea/␲s j p.
共6兲
The steady state of Ar/ O2 plasma discharge must satisfy
␯i,rf − ␯a,rf = 0. Therefore, the steady-state condition is obtained from Eqs. 共5兲 and 共6兲, and is given by
␯i共Ea兲 −
FIG. 1. Schematic presentation of the atmospheric-pressure argon/oxygen
plasma discharge source: 共a兲 the front view of the stepped-electrode reactor
and 共b兲 the side view of the reactor and experimental setup.
冉 冊
冉 冊
p
␣ j共E/p兲 = h j p exp − g j ,
E
共1兲
p
,
E
共2兲
␩ j共E/p兲 = u j p exp − s j
where j indicates gas species, and the constants h j and g j in
Eq. 共1兲 and the constant u j and s j in Eq. 共2兲 can be found
from tables or determined experimentally. In our case, the
oscillating field E can be presented by E = Ea sin ␻t, where
Ea is the amplitude of the oscillating rf field with ␻ = 2␲ f
= 0.85⫻ 108 s−1. The ionization frequency ␯i,rf共Ea兲 in rf field
is determined by an electron energy spectrum depending on
the field frequency ␻. The ionization rate is characterized by
the energy loss frequency ␯m␦ of electrons, where ␯m is an
effective collisional frequency and ␦ represents the energy
loss coefficient between electrons and atoms 共or
molecules兲.11 For ␻ Ⰷ ␯m␦, the conditions for sustaining nonequilibrium rf plasma do not differ, in principle, from those
of the positive column of a dc glow discharge, because the
electron energy spectrum responds very weakly to the oscillations of the rf field. Therefore, to sustain a steady-state
discharge in this case, the effective coefficient ␣eff in a mixture of argon and oxygen is expressed as
␣eff共Ea/p兲 = 共1 − ␹兲␣Ar + ␹共␣O2 − ␩O2兲 = 0,
共3兲
where ␹ is the oxygen mole fraction. Note from Eq. 共3兲 that
the attachment coefficient ␩Ar of noble gas is neglected. In
the opposite limit of ␻ Ⰶ ␯m␦, the electron energy spectrum
depends on the rf field oscillation, constantly adjusting itself
冉 冊
gAr/O2
1/2
␯a共Ea兲 = 0,
sAr/O2
共7兲
where
gAr/O2
sAr/O2
=
共1 − ␹兲gAr + ␹gO2
␹ s O2
共8兲
.
The constants gAr/O2 and sAr/O2 in Eq. 共7兲 are the constants in
the exponential term of Eqs. 共1兲 and 共2兲 for mixing gases of
argon and oxygen. As shown in Eq. 共8兲, the constants can be
expressed in terms of the mole fraction of mixing gases. The
ionization and attachment frequencies, ␯i共Ea兲 and ␯a共Ea兲 in
Eq. 共7兲, in a mixture of argon and oxygen are expressed as
␯i共Ea兲 = ␷d关共1 − ␹兲␣Ar + ␹␣O2兴,
共9兲
and
␯a共Ea兲 = ␷d关␹␩O2兴.
共10兲
Substituting Eqs. 共1兲, 共2兲, and 共8兲–共10兲 into Eq. 共7兲 gives the
solution for ␻ Ⰶ ␯m␦,
冉
共1 − ␹兲hAr exp − gAr
=
冋
冊
共1 − ␹兲gAr + ␹gO2
␹ s O2
冉
p
p
+ ␹hO2 exp − gO2
Ea
Ea
册
1/2
冉
␹uO2 exp − sO2
冊
冊
p
,
Ea
共11兲
where the constants h, g, u, and s are given by
hAr = 8.74⫻ 103 cm−1 atm−1, hO2 = 4.94⫻ 103 cm−1 atm−1, gAr
= 1.34⫻ 105 V cm−1 atm−1, gO2 = 1.44⫻ 105 V cm−1 atm−1,
uO2 = 1.88⫻ 102 cm−1 atm−1, and sO2 = 3.66⫻ 104 V cm−1
atm−1.11,14 Therefore, the required electric field in the steadystate Ar/ O2 discharge is determined by Eqs. 共3兲 and 共11兲 in
each case of ␻ Ⰷ ␯m␦ and ␻ Ⰶ ␯m␦.
Figure 2 presents the plots of rf field amplitude versus
the oxygen mole fraction, obtained from Eq. 共3兲 共solid line兲
051504-3
Appl. Phys. Lett. 90, 051504 共2007兲
Lim, Uhm, and Li
FIG. 2. Plot of the rf field amplitude Ea vs O2 mole fraction ␹ in steady-state
Ar/ O2 discharge at atmospheric pressure. Solid line is calculated from Eq.
共3兲 for ␻ Ⰷ ␯m␦ and dashed line is obtained from Eq. 共11兲 for ␻ Ⰶ ␯m␦. The
circular marks 共쎲兲 are measured values obtained from E = Vcutoff / dA.
for ␻ Ⰷ ␯m␦ and Eq. 共11兲 共dashed line兲 for ␻ Ⰶ ␯m␦. The circular marks in Fig. 2 are the measured electric fields, E
= Vcutoff / dA, which are obtained at the gap spacing dA in region A and the minimum voltage Vcutoff for discharge for
each oxygen mole fraction ␹. Energy loss coefficient ␦ depends on the oxygen mole fraction because electrons in molecular gases such as oxygen molecules dissipate energy
mostly by exciting vibrational and rotational energy levels
and the inelastic loss coefficient of electrons in oxygen molecules is about a few hundred times more than the elastic
loss coefficient of electrons in argon atoms.11 Therefore the
coefficient ␦ may increase as the oxygen mole fraction increases. As shown in Fig. 2, the measured electric fields 共쎲兲
in region A of the stepped-electrode reactor are close to the
dashed line for the limit of ␻ Ⰶ ␯m␦, as the oxygen mole
fraction increases. The rf field amplitude Ea for the steadystate discharge calculated from Eqs. 共3兲 and 共11兲 is compared
with the measured electric field in narrow region A, demonstrating that the electric field in narrow region plays a pivotal
role in sustaining the Ar/ O2 discharge. In the stepped reactor
operated at 10 l ” min of argon gas flow, the maximum concentration of O2 was about 0.1 vol % 共␹max ⬇ 0.001兲 for dA
= 0.3 mm, ␹max ⬇ 0.007 for dA = 0.25 mm, and ␹max ⬇ 0.01
for dA = 0.23 mm. The stable Ar/ O2 plasma discharge is sustained only for the stepped-electrode structure with the working gas entering from narrow region A to plasma region B. In
this context, energetic electrons produced by the strong field
in the narrow region may make the Ar/ O2 gas to be preionized and to be excited. Atoms and molecules excited by electron impacts in the narrow region can easily be ionized by
subsequent collisions in the wide region. Remember that the
ionization cross sections of excited atoms and molecules are
rather high,11 thereby being easily ionized by a relatively low
electric field in the wide region and generating plasmas.
Figure 3 shows photographs of Ar/ O2 plasma discharge
taken by a digital camera in 1 / 600 s of exposure time for
dA = 0.25 mm and dB = 1.55 mm at 10 l ” min of Ar gas flow,
and 50 cc/ min of O2 gas. The labels 共a1兲–共a3兲 in Fig. 3 are
photographs of the front views 共x direction in Fig. 1兲 and the
labels 共b1兲–共b3兲 are photographs of the lateral side views 共y
direction in Fig. 1兲 of the plasma reactor. As shown in Fig.
FIG. 3. 共Color online兲 Photographs of Ar/ O2 plasma discharges taken
by a digital camera in 1 / 600 s of exposure time. 关共a1兲 and 共b1兲兴 Pure Ar
plasma discharge at 100 W. 关共a2兲 and 共b2兲兴 Ar/ O2 plasma discharge at
100 W and 0.5 vol % of O2. 关共a3兲 and 共b3兲兴 Ar/ O2 plasma discharge at
150 W and 0.5 vol % of O2, which were operated at 10 l ” min of Ar gas
flow and dA = 0.25 mm and dB = 1.55 mm gap distances. The photos labeled
by 共a1兲–共a3兲 are the front views and the photos labeled by 共b1兲–共b3兲 are the
lateral side views of the reactor.
3共a2兲, the Ar/ O2 plasma was unstable due to an insufficient
power and reduced its volume, when 0.5 vol % of oxygen
was added into pure argon plasma. In Figs. 3共a3兲 and 3共b3兲,
the plasma was stable again and fully filled without streamer
and arcing as the input power increased from 100 to 150 W.
The plasma-flame temperatures exiting the reactor measured
by a thermocouple were about 90– 120 ° C. As shown in
Figs. 3共b1兲–3共b3兲, the bulk plasmas exist only in region B,
when the Ar/ O2 plasma discharge is generated.
In summary, we developed the atmospheric-pressure
Ar/ O2 plasma-discharge source with a stepped electrode,
which has many advantages, generating a stable Ar/ O2 glow
discharge, providing a wide operating window in glow region, and forming a relatively large volumetric plasma.
Moreover, argon plasma with high electron density, about
two times higher than helium plasma, is expected to reduce
treatment time compared with other APGD sources using
helium gas.15 In this context, the stable, uniform, and broad
plasma jet produced by the atmospheric-pressure Ar/ O2 discharge in the stepped-electrode reactor has a great potential
in industrial applications.
1
A. Schütze, J. Y. Jeong, S. E. Babayan, J. Park, G. S. Selwyn, and R. F.
Hicks, IEEE Trans. Plasma Sci. 26, 1685 共1998兲.
2
B. Eliasson and U. Kogelschatz, IEEE Trans. Plasma Sci. 19, 1063
共1991兲.
3
B. Eliasson and U. Kogelschatz, IEEE Trans. Plasma Sci. 19, 309 共1991兲.
4
F. Massines, A. Rabehi, P. Decomps, R. B. Gadri, P. Segur, and C. Mayoux, J. Appl. Phys. 83, 2950 共1998兲.
5
J. R. Roth, Phys. Plasmas 12, 057103 共2005兲.
6
H. Koinuma, H. Ohkubo, T. Hashimoto, K. Inomata, T. Shiraishi, A. Miyanaga, and S. Hayashi, Appl. Phys. Lett. 60, 816 共1992兲.
7
K. Inomata, H. Koinuma, Y. Oikawa, and T. Shiraishi, Appl. Phys. Lett.
66, 2188 共1995兲.
8
J. Park, I. Henins, H. W. Herrmann, G. S. Selwyn, J. Y. Jeong, R. F. Hicks,
D. Shim, and C. S. Chang, Appl. Phys. Lett. 76, 288 共2000兲.
9
S. Wang, V. Schulz-von der Gathen, and H. F. Döbele, Appl. Phys. Lett.
83, 3272 共2003兲.
10
H. Wang, W. Sun, H. Li, C. Bao, X. Gao, and H. Luo, Appl. Phys. Lett.
89, 161504 共2006兲.
11
Y. P. Raizer, Gas Discharge 共Springer, Berlin, 1997兲, Vol. 1, Chaps. 2, 4,
9, and 13, pp. 16–18, 52–66, 222–228, and 378–414.
12
E. Nasser, Fundamentals of Gaseous Ionization and Plasma Electronics
共Wiley, New York, 1971兲, Vol. 1, Chaps. 7 and 8, pp. 188–226.
13
H. S. Uhm and W. M. Lee, Phys. Plasmas 4, 3217 共1997兲.
14
J. B. Freely and L. H. Fisher, Phys. Rev. 133, A304 共1964兲.
15
S. Z. Li, J. P. Lim, J. G. Kang, and H. S. Uhm, Phys. Plasmas 13, 093503
共2006兲.