Understanding Micro Plasmas
J.Winter, J. Benedikt, M. Böke, D. Ellerweg, T. Hemke, N. Knake,T. Mussenbrock, B. Niermann, D.
Schröder, V. Schulz-von der Gathen, A. von Keudell
Institute for Experimental Physics II, Ruhr- Universität Bochum, Germany
Keywords: Micro plasmas, jets, diagnostics, modeling
Abstract
Micro plasmas are operated around atmospheric pressure exhibiting pronounced non-equilibrium
characteristics, i.e. they possess energetic electrons while ions and neutrals remain cold. They have
gained significant interest due to their enormous application potential e.g. in the biomedical, surface
modification and light source areas, just to name a few. Many different configurations are in use. Their
understanding and quantification is mandatory for further progress in applications. We report on recent
progress in the diagnostics and simulation of the entire micro plasma system from gas introduction,
via the plasma discharge up to the samples at the example of a plasma jet operated in He/O2 in an
ambient air environment.
Introduction
not sufficient, though indispensible, to
characterize the source as such, but the whole
set up from gas introduction via plasma
discharge to the target has to be carefully
considered. Without quantitative data further
progress is very much hampered and
interpretation of the complex phenomena
involving micro plasma applications will
remain in the realm of tell stories.
Micro plasmas are characterized by high
electron densities in small volumes. The
characteristic length scale for the charged
particle dynamics, the Debye length, contracts
below a value of 1 µm. The mean free path of
particles and photons in these plasmas
becomes rather small. Energetic electrons,
however, may have mean free paths of the
order of the discharge vessel, leading to non
local plasma heating. The surface-to-volume
ratio is extreme, yielding very high power
densities.
There are many different micro plasma
configurations in use from micro jets, micro
hollow cathode arrangements to arrays of
integrated structures with excitations from the
microwave range, RF range, pulsed high
voltage, to DC. Being limited in time and
space, this paper focuses on measurements
made
with
the
coplanar
microscale
atmospheric pressure plasma jet (µ-APPJ) and
addresses selected problems. Still we will be
able to draw some general conclusions for
other systems. We will present quantitative
data on the kinetics of oxygen species in a RF
excited jet operated in He/O2 and their
interaction with surfaces, discuss the problem
of energy flow in the jet by metastable He and
Ar atoms which is associated with the stability
of the micro jet and the problem of impurity
intrusion into jets when operated in ambient
atmosphere.
Research on microscopic plasmas nowadays is
either focused on the development and
optimization of plasma for light emission or
for the production of reactive species from a
molecular precursor gas. Light emission is
exploited in display panels or in short arcs as
used for extreme UV lithography. Reactive
microplasmas have been developed for surface
treatment of materials ranging from polymers
up to living tissue in the emerging field of
plasma medicine. Also depositing gases are
used for the deposition of functional organic or
inorganic coatings and thin films.
Absolute
measurements
of
discharge
parameters and fluxes are needed to perform
ultimate benchmarks and tests for the
falsification or verification of current
hypotheses and theories. As will be shown it is
1
Fig. 2 shows the absolute oxygen densities
along the axis of the jet, with the measuring
spot always in the middle between electrodes.
In agreement with modelling, the maximum O
density in the jet is achieved for an admixture
of 0.6% O2 to He.
The µAPPJ
The micro plasma jet (μ-APPJ) used for the
experiments is a capacitively coupled radio
frequency discharge (13.56 MHz, ∼15W RF
power) designed for optimized optical
diagnostic access (Fig.1). The coplanar
geometry and an electrode width of about 1mm
provide a discharge profile of 1mm2 for
localized surface treatment at a low gas
consumption. It is operated in a homogeneous
glow mode with a noble gas flow (typically 1.4
slmHe) containing a small admixture of
molecular oxygen (typically ∼0.5%). For some
investigations the jet was prolonged by a non
powered section made from isolating materials
to avoid immediate contact with ambient air
and guide the gas flow.
Complementary measurements of reactive
oxygen species were measured by calibrated
molecular beam mass spectroscopy in the
effluent. The sensitivity of this diagnostics was
improved significantly by introducing a
rotating shutter. In addition to oxygen, the
method also detects reactive species like O3
and others. O densities in the effluent agree
very well regarding the trend. Absolute
numbers differ by about a factor of 0.27 which
is likely to be due to the presence of the
surface of the mass spectrometer (see below).
It is also seen that significant amounts of O3
are produced by reactions of O with O2
whereby one source of O remote from the
nozzle may be due VUV radiation propagating
along the He beam leaving the nozzle.
15
1.4x10
15
1.2x10
15
1.0x10
15
8.0x10
14
6.0x10
14
4.0x10
14
2.0x10
14
O (MBMS)
O3 (MBMS)
O (TALIF)
-3
density [cm ]
Fig.1 The µAPPJ micro plasma jet
1.6x10
Absolute measurements of reactive
O species
x 0.27
0.0
Ground state atomic oxygen densities were
measured by two-photon absorption laserinduced fluorescence spectroscopy (TALIF)
providing space resolved density maps. The
quantitative calibration of the TALIF setup is
performed by comparative measurements with
xenon.
Fig. 3: Comparison of TALIF and MBMS
measurements in the effluent of the µAPPJ
Fig. 2: Absolute O densities in the discharge
region and in the effluent of the µAPPJ
Fig. 4: Model calculations of neutral reactive
species
0
5
10
15
20
25
30
35
40
45
50
z - axis [mm]
2
Modelling using nonPDPSIM is in good
agreement with these data but also suggests a
prime contribution of O- and O2- in the jet
which still has to be assessed experimentally.
Fig. 7: O profiles in front of targets
surface. A simple model allowing for a fully
absorbing wall (particle loss probabilty ß =1),
a fully reflecting surface (ß=0), and an O
source at the surface (ß<0) predicts the local O
concentration close to the stagnation point.
Profile measurements with a polyehtylene
(PET) target show full absorption of O. O
etches PET and the reaction products CO and
CO2 are probably released. Etching was
confirmed by measuring the surface
topography. These data also show in addition
to a localized “deep” etching pattern from O
and UV a very broad affected zone which is
not associated with the narrow beam particles
and radiation but stems most likely from O3
which is simultaneously formed in reactions of
O with O2 from air. Experiments with a gold
surface show O production at the surface
(ß<0). This is explained by a catalytic
decomposition of O3 forming O. Ozone
measurements using absorption spectroscopy
confirm this picture.
Fig. 5: Model calculations of charged species
O in front of a surface
Various materials were exposed to the effluent
of the jet at different distances and the local O
density was measured by TALIF up to about
0.1 mm before the targets. This distance was
limited by spatial access. The flow has a
stagnation point at the target surface, but
diverts from perpendicular to parallel flow
direction before the target, defining thus the
diffusion zone across which species reach the
surface. Flow modeling indicates the diffusion
zone to start at about 0.5 mm in front of the
He metastables, the role of impurity
intrusion
The absolute concentrations of He* and Ar*
metastables were measured using tunable laser
diode absorption spectroscopy (TDLAS).
Profiles across and along the jet axis are shown
in Fig. 8. He* is located in the sheath regions
where fast electrons lead to their excitation.
They constitute an important reservoir of
potential energy, which is converted into
ionization by pooling reactions.
Fig.6: Set up for measuring the O impinging
on a surface
3
Conclusion
Our analysis of the entire integrated system of
a micro-plasma jet operated in He/O2 and our
attempts to provide quantitative data, have led
to significant progress in understanding the
behaviour of the complex system from gas
introduction to the treated target. We have
been able to benchmark model predictions. It
became obvious that it is not enough just to
quantify the discharge as such (which may be
difficult enough) but that it is important to
consider carefully all intervening external
factors. Nevertheless, although a significant
effort has gone into characterizing the system
quantitatively, some problems still await
resolution.
1.8
Vertical Position [mm]
1.6
High Density
6.3E+10
1.4
5.5E+10
1.2
4.4E+10
1.0
3.3E+10
0.8
2.1E+10
0.6
1.0E+10
Acknowledgement
0.4
-1.2E+09
0.2
Low Density
0.0
0
5
10
15
20
25
30
35
The work was supported by DFG within the
Research Unit FOR 1123
40
Horizontal Position [mm]
Fig 8: He* and Ar* profiles across the µAPPJ
(top) and map of He* along the jet axis
(bottom).
The lifetime of He* was measured by pulsing
the plasma. Metastables are rapidly quenched
by three body collisions and, in particular, by
impurities. This may be one of the causes
limiting the admixture of molecular O2 to the
jet. Since the lifetime is a very sensitive probe
for the impurity concentration in the jet, a
calibration curve of lifetime versus added N2
flow was established. This allowed scanning
quantitatively the impurity concentration (as
N2 equivalent) for different operation modes.
In order to investigate the role of air ingress
when the jet is operated at ambient air,
measurements were performed with the jet
located in a vacuum chamber which was
backfilled either with He or air, respectively.
Analysis showed, that a significant backflow
of air occurs from the nozzle into the jet even
with large He flow. Under realistic application
conditions, this has to be considered carefully.
4