Xenon excimer emission from pulsed high-pressure capillary
microdischarges
Byung-Joon Lee,
a
Hasibur Rahaman, Isfried Petzenhauser, and Klaus Frank
Physics Department I, F.A.-University of Erlangen-Nuremberg, D-91058 Erlangen, Germany
Konstantinos P. Giapis
Division of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, California 91125
Received 18 April 2007; accepted 20 May 2007; published online 14 June 2007
Intense xenon vacuum ultraviolet
VUV
emission is observed from a high-pressure capillary
cathode microdischarge in direct current operation, by superimposing a high-voltage pulse of 50 ns
duration. Under stagnant gas conditions, the total VUV light intensity increases linearly with
pressure from 400 to 1013 mbar for a fixed voltage pulse. At fixed pressure, however, the VUV light
intensity increases superlinearly with voltage pulse height ranging from 08 to 2.8 kV. Gains in
emission intensity are obtained by inducing gas flow through the capillary cathode, presumably
because of excimer dimer survival due to gas cooling. ©
2007 American Institute of Physics
.
DOI:
10.1063/1.2748314
Several topical reviews in recent years emphasized the
scientific interest and the important applications of microdis-
charges operated in the hollow cathode mode.
1
–
3
These mi-
crodischarges form stably at high pressures
1 atm or higher
through the presence of a sufficient concentration of high-
energy electrons undergoing a Pendel oscillation. Operation
under these conditions
high pressure, high-energy electrons
favors the formation of excimer dimers, which renders these
microdicharges useful as sources of intense vacuum ultravio-
let
VUV
emission from rare gases
4
–
7
and UV emission
from rare gas halides.
8
,
9
However, most of the reported ex-
periments were performed in direct current
dc
microdis-
charges, where gas heating can destroy excimer dimers thus
limiting the radiant power of the VUV emission. Thermal
heating can be reduced by pulsing the discharge. Recently,
an increase in the radiant power of xenon
10
and argon
11
ex-
cimer emission, from planar dc microdischarges was ob-
served after a high-voltage pulse
sub-kilovolts 20 ns
was
superimposed. However, the enhancement was attributed to
the increase in the plasma area over the cathode surface re-
sulting from high voltage and current. This effect is incom-
patible with an excimer microlaser, which requires an on-
axis increase in plasma volume.
7
In this letter, we report enhanced VUV emission from
the application of a high-voltage
1kV
pulse to a dc cap-
illary microdischarge. The enhancement is believed to origi-
nate from the increase in on-axis plasma volume in the me-
tallic capillary cathode. Applying a high-voltage pulse to the
capillary microdischarge may result in current densities
above the threshold for the glow-to-arc transition
GAT
.
This condition was avoided by selecting the pulse duration to
be shorter than the time constant of the dominant instability
that causes the GAT transition.
12
For improved VUV charac-
teristics, it is critically important to avoid emission caused by
secondary current pulses as the pulsed voltage relaxes to
zero.
13
The latter effect was avoided by using a self-matched
transmission line pulsen generator,
14
consisting of a
10-m-long 50
coaxial cable, which fixed the pulse dura-
tion to 50 ns. The matching impedance minimized the reflec-
tion of the current pulse that arises from the time-variable
impedance of the gas discharge. A fast rise time of the pulsed
voltage was obtained by using a spark gap.
The experimental setup for the pulsed capillary dis-
charge is shown in Fig.
1
. The cathode consists of a stainless
steel capillary tube with inside diameter of 180
m and a
length of 5 mm. A stainless steel mesh served as the anode.
A 250-
m-thick mica sheet with a centered hole of 200
m
prevented the plasma from expanding on the cathode outer
surface. The microdischarge was generated by two dc power
supplies, one for establishing a cw microdischarge and the
other for providing the 50 ns voltage pulses through the self-
matched transmission line. Experiments were conducted in
pure Xe with and without forced gas flow through the micro-
discharge.
Figure
2
shows the time resolved measurement of the
electrical characteristics for the pulsed capillary microdis-
charge in a pressure of 1013 mbar. Initially, the dc microdis-
charge was formed at a sustaining voltage of 200 V with a
current of 3 mA. Then, a 1.5 kV pulse with a rise time
10–
90%
of 7 ns was superimposed for 50 ns
full width at half
maximum
FWHM
to this dc discharge. The voltage across
the discharge was measured by a voltage probe
Tektronix
P5100
and the current was measured by a current monitor
Pearson 2877
. The voltage pulse coincides with an abrupt
increase in discharge current of equal duration. The peak
a
Electronic mail: bjlee@physik.uni-erlangen.de
FIG. 1. Schematic diagram of the experimental setup.
APPLIED PHYSICS LETTERS
90
, 241502
2007
0003-6951/2007/90
24
/241502/3/$23.00
© 2007 American Institute of Physics
90
, 241502-1
Downloaded 25 Jul 2007 to 131.215.225.175. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
current value was approximately 28 A. During the decay of
the voltage pulse, the discharge voltage drops to zero but
recovers to its dc value
200 V
after 20
s. The temporal
development of the Xe excimer emission at 172 nm during
pulsing is also shown in Fig.
2
. The emission was collected
through the mesh anode and was focused on the entrance slit
of the VUV monochromator
Acton Research VM 505,
1200 G/mm grating blazed at 190 nm
using a magnesium
fluoride
MgF
2
lens. The temporal development of the VUV
emission was detected by means of a fast VUV photomulti-
plier
Valvo AVP 56 B
integrating over a bandwidth of 5 nm
centered at 172 nm. The internal delay time of the photomul-
tiplier was compensated with respect to the voltage signal
and the current signal. Very low intensity of the VUV emis-
sion was observed during dc operation, consistent with the
low discharge current. However, when the high-voltage pulse
was applied, the emission intensity increased rapidly in step
with the discharge current, reaching a value almost two or-
ders of magnitude larger than that for the dc operation. Al-
though the peak VUV intensity coincides with the discharge
current maximum, the emission decays exponentially to its
dc value with a time scale one order of the magnitude longer
than the duration of the applied voltage pulse.
The temporal behavior of the VUV emission was also
investigated as a function of the pressure from
400 to 1013 mbar
see Fig.
3
. In this pressure range, the
VUV intensity during dc operation is small and changes very
little. Superposition of the pulsed voltage caused the peak
VUV intensity to increase roughly proportional to the pres-
sure. For instance, the peak VUV intensity at 1013 mbar was
about 2.5 times larger than that at 400 mbar. No time delay
for the appearance of the VUV emission pulses was observed
with increasing pressure. The width
FWHM
of the VUV
emission pulse was about 70 ns regardless of pressure. The
exponential decay of the VUV emission occurred slightly
faster at the higher pressures. This suggests an increased col-
lisional rate for the Xe
*
conversion into Xe
2
*
, that is, an in-
creased collisional quenching effect.
11
The decay of the sec-
ond continuum Xe
2
*
results from the loss rate of the Xe
*
1
S
5
.
15
The total VUV intensity, obtained by integrating the
VUV emission wave forms, increased linearly with the gas
pressure
not shown
.
The VUV emission intensity also depends on the mag-
nitude of the applied voltage pulse. Figure
4
illustrates the
variation in total integrated intensity when the applied volt-
age pulse height is raised from 0.8 to 2.8 kV at 1013 mbar.
The observed superlinear dependence is attributed to non-
equilibrium effects in the pulsed discharge.
10
The VUV emis-
sion efficiency, defined as the ratio of the total VUV intensity
over the electrically consumed energy, first increased slightly
with pulsed voltage up to 0.9 kV, but then it decreased lin-
early with pulsed voltage. Thus, to increase VUV emission,
FIG. 5. Temporal development of the VUV emission for different gas flow
rates.
FIG. 2. Temporal development of the discharge current, voltage, and the
VUV emission at 172 nm.
FIG. 3. Temporal development of the VUV emission for the different pres-
sures
400–1013 mbar
.
FIG. 4. VUV efficiency and total VUV intensity for the different amplitudes
of voltage pulses.
241502-2 Lee
etal.
Appl. Phys. Lett.
90
, 241502
2007
Downloaded 25 Jul 2007 to 131.215.225.175. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
a disproportionately larger power must be delivered to the
microdischarge.
The influence of forced gas flow through the capillary
cathode on VUV emission was also investigated for pulsed
discharge operation. Figure
5
shows the temporal evolution
of the VUV emission as a function of the flow rate of ambi-
ent Xe through the capillary tube. The gas flow rate was
controlled by means of a flow meter, while maintaining an
ambient pressure of 1013 mbar. For this particular experi-
ment, the applied voltage on the VUV photomultiplier was
reduced in order to avoid any possible saturation in the mea-
surement. When the pulsed voltage was applied to the micro-
discharge under forced flow, the temporal evolution of the
VUV emission exhibited differences from that observed in
the stagnant gas
see Fig.
3
. The peak VUV intensity im-
proved with flow rate, possibly due to increased neutral gas
density and reduced decomposition of the excimer dimers as
a result of the cooling effect of the gas flow. The improve-
ment in VUV emission with gas flow for pulsed microdis-
charge operation is very small when compared with that re-
ported for dc operation at much higher discharge currents,
where heating effects are a lot more substantial.
7
In conclusion, superimposing a high-voltage pulse on a
dc capillary cathode microdischarge operated in Xe gas am-
bient permits substantial gains in VUV emission intensity,
especially with increasing voltage pulse height, operating
pressure, and forced flow rate through the microdischarge.
Thus, pulsed operation may increase the chances of con-
structing an excimer microlaser using stacked capillary cath-
ode microdischarges to provide sufficient active gain length.
7
The authors gratefully acknowledge the support of this
work by the “Deutscher Akademischer Austauschdienst”
Code No. A/02/08987
.
1
K. H. Becker, K. H. Schoenbach, and J. G. Eden, J. Phys. D
39
,R55
2006
.
2
J. G. Eden, S.-J. Park, and K.-S. Kim, Plasma Sources Sci. Technol.
15
,
S67
2006
.
3
R. M. Sankaran and K. P. Giapis, J. Phys. D
36
, 2914
2003
.
4
A. El-Habachi and K. H. Schoenbach, Appl. Phys. Lett.
72
,22
1998
.
5
A. El-Habachi and K. H. Schoenbach, Appl. Phys. Lett.
73
,885
1998
.
6
T. I. Lee, K. W. Park, H. S. Hwang, J. P. Jegal, and H. G. Baik, Appl.
Phys. Lett.
88
, 211502
2006
.
7
R. M. Sankaran, K. P. Giapis, M. Moselhy, and K. H. Schoenbach, Appl.
Phys. Lett.
83
, 4728
2003
.
8
K. H. Schoenbach, A. El-Habachi, M. Moselhy, and W. Shi, Phys.
Plasmas
7
, 2186
2000
.
9
A. El-Habachi, W. Shi, M. Moselhy, R. H. Stark, and K. H. Schoenbach,
J. Appl. Phys.
88
, 3220
2000
.
10
M. Moselhy, W. Shi, R. H. Stark, and K. H. Schoenbach, Appl. Phys. Lett.
79
, 1240
2001
.
11
M. Moselhy, I. Petzenhauser, K. Frank, and K. H. Schoenbach, J. Phys. D
36
, 2922
2003
.
12
E. E. Kunhardt, IEEE Trans. Plasma Sci.
28
, 189
2000
.
13
R. J. Carman, R. P. Mildren, B. K. Ward, and D. M. Kane, J. Phys. D
37
,
2399
2004
.
14
F. A. Tuema, S. J. MacGregor, and R. A. Fouracre, Meas. Sci. Technol.
9
,
1989
1998
.
15
I. Petzenhauser, L. D. Biborosch, U. Ernst, K. Frank, and K. H.
Schoenbach, Appl. Phys. Lett.
83
, 4297
2003
.
241502-3 Lee
etal.
Appl. Phys. Lett.
90
, 241502
2007
Downloaded 25 Jul 2007 to 131.215.225.175. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp