22nd International Symposium on Plasma Chemistry
July 5-10, 2015; Antwerp, Belgium
2D time-resolved measurement and modeling of electric fields associated with
atmospheric pressure plasma streams propagation in dielectric capillaries
T. Darny1, E. Robert1, F. Pechereau2, S. Dozias1, A. Bourdon3 and J.-M. Pouvesle1
1
GREMI UMR 7344 CNRS / Université d’Orléans, Orléans, FR-45067, France
2
CERFACS, 42 Av Coriolis, FR-31057 Toulouse, France
3
LPP UMR7648 CNRS / Ecole Polytechnique, Palaiseau / Université Pierre et Marie Curie, Paris, France
Abstract: This work reports on time-resolved measurement of longitudinal and radial
electric fields (EF) associated with plasma propagation in dielectric capillaries. Plasma
propagation occurs in a region where longitudinal EF exists ahead the ionization front
position usually revealed from plasma emission with ICCD measurement. The ionization
front propagation induces the sudden rise of a radial EF component. Both of these EF
components have a few kV/cm in amplitude for helium or neon plasmas. Their amplitude
is kept almost constant along a few tens of cm long capillary. All these experimental
measurements are in excellent agreement with electrostatic and 2D fluid model calculations
which are used to infer EF data on capillary axis.
Keywords: plasma jet, electric field, ionization wave
1. Introduction
While cold atmospheric pressure plasma jets have
shown their great potential for many biomedical
applications, there still exists a strong need for their
diagnostics. Excited and reactive species quantification
[1,2], interplay between plasma jets and the carrying gas
flow [3,4],interaction of plasma jets with targets of
various nature [5], electric field measurement [6,7] are
some of the most exciting but challenging topics under
study targeting a deeper characterization of plasma
properties and potential optimization for relevant
applications. This work presents a preliminary non
intrusive experimental and modelling analysis of electric
fields associated with the generation and propagation of
atmospheric pressure plasma streams in dielectric tubes
and ambient air. Dealing with electric field diagnostics,
several techniques have been already proposed. Shashurin
et al. [6] proposed a method based on stopping ionization
front propagation with DC potential applied on external
metallic ring surrounding plasma plume. Very intense
electric field amplitude as high as 100 kV/cm were
deduced in [6]. The method may suffer severe limitation
for some plasma jets, where the use of metal ring may
strongly modify the plasma plume features.
Other
technique
based
onStark
polarization
spectroscopy has been proposed by Ivkovic et al. [7],
based on He line and its forbidden counterpart in a helium
plane to plane DBD. This optical method is non intrusive
and give access to spatial resolution of electric field
amplitude in the cathode sheath region. However,
homogenous discharge is required, in appropriate pressure
range (200 and 800 mbar pressure), and the technique is
limited with lower detection rate of 3kV/cm and the only
use of helium gases or helium and hydrogen admixture
[8]. Authors indicate a good agreement between their
O-2-2
experimental value and the usual local electrical field
calculates by numerical models in this condition (around
10 kV/cm). A large effort has also been addressed on the
development of adequate models likely to give insight for
some hard to reach experimental parameters [8-12]. In the
present work, we report and analyse first electric field
measurements with a new device (bi component EOP
Kapteos probe) for the diagnostics of pulsed atmosphericpressure plasma streams inside long dielectric tube,
produced by the Plasma Gun (PG). PG is a coaxial
dielectric barrier discharge reactor with a quartz capillary,
flushed with rare gas and powered, in this work by μs
duration voltage pulses in the kHz regime.
2. Experimental Setup
Figure 1 shows the experimental setup. A 42 cm long
dielectric quartz capillary with a 4 mm inner diameter and
6 mm outer diameter is used. The inner electrode, 2cm
long, is set inside the capillary. The rare gas (helium or
neon) buffer (1 L/min) is injected through in the inner
hollowed electrode (0,8 mm inner diameter). Grounded
ring electrode is set on the outer surface of the quartz
capillary at the tip of the inner electrode.
Some experiments have also been performed with the
use of a single HV cylindrical electrode set on the outer
surface of the quartz capillary, the discharge reactor being
positioned on the axis of a grounded cylinder, 20 cm in
diameter, in order to get the best agreement with the
model discharge configuration. Data discussed in this
work have been measured with a 16 kV peak voltage
amplitude applied across the PG powered at a constant
1 kHz repetition rate.
1
usually detected from ICCD measurement or plasma
emission optical detection, expanding over a distance of
about 5 cm. This was previously revealed from model
calculation [9] where the existence of the higher energy
electrons was reported ahead the electron density and
ionization source function peak position during ionization
wave propagation.
Fig. 1. PG with the probe set 10 cm downstream from the
grounded electrode, 2 mm from the capillary wall. The
probe is represented by it space orientation, here allowing
Ex and Er measurements.
A Pockels effect based fiber-like sensor equipped with
an isotropic crystal probed by a laser beam give
simultaneous access to two orthogonal components of the
electric field. A specially designed 1.75 mm in diameter,
1 mm long crystal embedded in an alumina tube, set at
one end of an optical fiber was used as a sensor. The
sensor was move along the quartz capillary, with a
constant 2 mm gap from the quartz outer surface. This
corresponds to a distance between the capillary axis and
the crystal center of 6 mm. This 2 mm gap was checked,
through ICCD measurement, to induce no detectable
modification of the plasma front propagation emission
pattern and velocity. Thus, the probe allow for the nonintrusive, non perturbative measurement of the
longitudinal (Ex) and radial (Er) EF components. The EF
field amplitude necessarily reflects space averaged (1.75
mm3) value while the full detection system affords
nanosecond temporal resolution.
3. Results
Figure 2 presents the voltage pulse and the temporal
evolutions of longitudinal and radial EF for the probe set
10 cm downstream from the inner electrode tip. For this
probe position, the peak longitudinal EF amplitude, 5 kV/cm, is measured with an 1800 ns delay from the
voltage onset. This delay should be assigned to the
combination of the discharge production delay around 4
kV applied voltage and the consecutive plasma
propagation along the first 10 cm path. For the helium
buffer and a 16 kV amplitude, the mean velocity of the
plasma is around 107cm/s, inducing a 1 µs delay for 10
cm long plasma propagation. One can note that the EF
probe detects longitudinal EF increase about 500 ns
before the 1800 ns peak. This indicates that there exists a
significant longitudinal EF ahead the ionization front
2
Fig. 2. Ex and Er time evolutions for the probe set at the
10 cm position. The time evolution of the voltage pulse is
also superimposed on the graphic. Helium buffer is used.
Figure 2 also indicates the sudden rise of an intense
radial EF component, presenting a very sharp rising front
and appearing almost synchronously with the peak of the
longitudinal EF component, i.e. around 1800 ns for the
present experimental conditions. The rise of such radial
component following the ionization wave propagation
was also reported in [9] where EF was shown to
essentially consists in a radial component all along the
ionization channel following the ionization front while a
rather extended region where longitudinal EF component
predominates ahead the ionization front.
At longer delays, the probe detects the EF associated
with the combination of the electric field imposed across
the PG electrodes, and the EF induced through the plasma
column following the ionization front. The voltage
applied to this plasma column is controlled by the voltage
applied across the PG. Equipotential line calculation have
been performed using Comsol multiphysics® for two
situations corresponding first, to the start of the ionization
wave propagation and second, after an 8 cm long
ionization front propagation. In this rough calculation,
constant voltage of 10 kV in amplitude is applied on an
on axis 2 mm wide HV electrode. Ground potential is
imposed in a cylindrical region 1 cm high, 2 mm wide,
mimicking the PG ground electrode. The plasma column
is considered as an equipotential downstream extension of
the HV electrode. As awaited, the electric field is
basically longitudinal ahead the ionization front, while
following plasma propagation, the plasma column induces
an almost radial EF pattern, in agreement with results in
fig.2. This confirms the role in high potential extension of
O-2-2
the plasma column in between the ionization front and the
discharge reactor.
Fig. 3. Equipotential line patterns at the beginning of
plasma propagation (top) and after an 8 cm path (bottom).
Color scale ranges from 0.1 to 10 kV (applied voltage).
Both the longitudinal and radial EF measured 6 mm
from the capillary axis, exhibit amplitudes of a few
kV/cm with no direct relationship with the electric field
imposed across the PG electrodes, this latter being
essential intense in the electrode zone and quite
undetectable a few cm away from this region without
plasma propagation. This confirms that transient intense
electric field could be delivered a few tens of cm away
from the DBD reactor and may play a critical role for
biomedical applications.
Figure 4 presents the longitudinal and radial EF for
different downstream distances from the HV electrode tip
for helium and neon buffer gases. At a first glance, the EF
components behaviour and amplitude are close for the
two rare gases. The faster propagation of neon plasma is
confirmed by the sooner appearance of the EF
components for the same downstream positions (10 or 25
cm). Sudden rise of radial EF holds true for neon plasma
which exhibits slightly higher EF amplitudes and shorter
duration and rising front partly associated with the faster
propagation velocity in front of the EF probe. The
measurements in long tubes for neon plasma indicate that
the EF amplitudes are almost constant along a few tens of
cm propagation, in agreement with previous calculation
[9]. This confirms the specific nature and peculiar interest
of plasma streams generated in confined dielectric tubes,
having the ability to preserved most of plasma parameters
along very long distances.
O-2-2
Fig. 4. Ex and Er time evolutions for different
downstream probe positions with helium (a) and neon (b)
buffer gas.
New calculations and experiments have been performed
with a specific attention for plasma propagation analysis
and associated electric fields properties. Figure 5 presents
the experimental EF measurements obtained with the
single outer HV electrode PG configuration and their
comparison with the results of a 2D fluid model [12] in
the same setup. An excellent agreement is observed
concerning the time evolution of the EF components, the
calculation also confirms the sudden rise of the radial EF
following ionization wave propagation. A good
agreement is also found concerning the EF amplitudes,
confirming that radial peak amplitude is higher than the
longitudinal one. Considering the already achieved good
simulation of the experimental ionization wave
propagation and EF nature, model has been used to infer
the on axis EF. Figure 6 presents such on axis
longitudinal EF evolution along the axis of the quartz
capillary. Besides the electric field induced in the
electrode region, an intense (9 kV/cm in amplitude)
longitudinal component is generated presenting some
ahead extension over distance of a few cm.
3
Fig. 5. Experimental (top) and 2D fluid model calculation
(bottom) of Ex and Er in the single electrode PG
configuration. The probe was set 3 cm downstream the
HV electrode tip, voltage amplitude was 16 kV, helium
buffer was used.
0
-1
0
2
4
6
8
10
12
14
16
18
20
Ex (kV/cm)
-2
-3
-4
-5
-6
5. Acknowlegments
This work is supported by ANR BLAN 093003
PAMPA, APR PLASMEDNORM, TD is supported by
MENSR. The authors are grateful to N. Semmar for
Comsol calculation implementation, and L. Duvillaret
(Kapteos) for EOS probe design and validation.
-7
-8
-9
-10
Distance from electrode (cm)
Fig. 6. 2D fluid model calculation of on capillary axis Ex
profile along a 10 cm long capillary. Discharge
propagation from left to right.
Work is in progress to extend EF characterization for
the plasma plume delivered in ambient air but also over
targets relevant for biomedical applications. While the
non perturbative setting of the optical probe has not yet
been fully demonstrated, preliminary measurements
indicate first that EF having peak amplitudes in the range
from 10 to 20 kV/cm are detected in the plasma plume,
and second, that EF amplitudes around 1kV/cm persists
below 3 mm thick tissue layers.
4. Conclusion
Time-resolved experimental non intrusive and non
perturbative measurement of longitudinal and radial
electric field components associated with helium and
neon atmospheric pressure plasma propagation in long
4
dielectric tubes has been achieved using a new probe
based on Pockels effect. Peak voltage amplitudes of a few
kV/cm have been measured for both components a few
mm apart from capillary axis, for both rare gas buffers.
The experimental measurements reveal that plasma
propagates in region where an intense longitudinal
component exists a few cm ahead the ionization front
usually revealed by strong plasma emission in optical
diagnostics. Correlated with the ionization front
propagation, the extension of a plasma tail connecting this
latter with the powered electrode of the plasma jet device,
induces the sudden generation of an intense radial electric
field component. These observations are in good
agreement with electrostatic calculations and confirm that
electric field amplitudes are almost constant along the full
plasma propagation over distances of a few tens of cm.
Calculations with a 2D fluid model, in a configuration
very close to the experimental setup and including the
temporal evolution of the voltage pulse delivered across
the discharge reactor, indicate very good correlation of
the electric field temporal profiles, electric field amplitude
and allow inferring electric field on capillary axis. On axis
longitudinal and radial electric field amplitude around
10 kV/cm have been obtained inside helium fed
capillaries while electric field amplitude in the plume
region ranges from 10 to 20 kV/cm, in our experimental
conditions.
6. References
[1] S. Iseni et al New Journal of Physics16 123011 (2014)
[2] S.Hubner et al J.Phys.D :Appl Phys. 47 432001 (2014)
[3] E. Robert et al. Plasma Sources Sci. Technol. 23
012003 (2014)
[4] S Zhang et al J. Phys. D: Appl. Phys. 48 015203
(2015)
[5] S. Bornholdt et al Eur. Phys. J. D10.1140/epjd/e201000245-x (2010)
[6] A.Shashurin et al Plasma Sources Sci. Technol. 21
034006 (2012)
[7] B M. Obradovic et al AppliedPhysicsletters92, 191501
(2008)
[8] S S. Ivkovic et al J.Phys.D :Appl Phys. 45 275204
(2012)
[9] Z. Xiong and M J. Kushner Plasma Sources Sci.
Technol. 21 034001 (2012)
[10] J. Jansky et alJ. Phys. D: Appl. Phys. 44335201
(2011)
[11] J-P Boeuf et al J. Phys. D: Appl. Phys.46
015201(2013)
[12] J. Jansky et al Plasma Sources Sci. Technol23025001
(2014)
O-2-2