A washer gun plasma system for

microwave—plasma interaction
experiments
Cite as: Rev. Sci. Instrum. 90, 013502 (2019); https://doi.org/10.1063/1.5051802
Submitted: 13 August 2018 • Accepted: 13 December 2018 • Published Online: 07 January 2019

Anitha V. P., Priyavandana J. Rathod, Jayesh Raval, et al.

ARTICLES YOU MAY BE INTERESTED IN

Characterization of a medium-sized washer-gun for an axisymmetric mirror

Review of Scientific Instruments 89, 043503 (2018); https://doi.org/10.1063/1.5013118

Development of a low-energy and high-current pulsed neutral beam injector with a

washer-gun plasma source for high-beta plasma experiments
Review of Scientific Instruments 83, 083504 (2012); https://doi.org/10.1063/1.4744960

The upgraded Large Plasma Device, a machine for studying frontier basic plasma physics
Review of Scientific Instruments 87, 025105 (2016); https://doi.org/10.1063/1.4941079

Rev. Sci. Instrum. 90, 013502 (2019); https://doi.org/10.1063/1.5051802 90, 013502

© 2019 Author(s).
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

A washer gun plasma system for

microwave—plasma interaction experiments
Cite as: Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802
Submitted: 13 August 2018 • Accepted: 13 December 2018 •
Published Online: 7 January 2019

Anitha V. P.,a) Priyavandana J. Rathod, Jayesh Raval, Renu Bahl, and Y. C. Saxena

AFFILIATIONS
Institute for Plasma Research, HBNI, Bhat, Gandhinagar 382428, India

a)
Electronic mail: anitha@ipr.res.in

ABSTRACT
A washer-gun based plasma system has been developed to enable high power microwave (HPM)—plasma interaction in a system
for microwave plasma experiments. The critical pre-requisites of the plasma are density, ne ∼ (1–10) × 1017 m−3 , uniformity over a
radial extent ≈10 cm and axial extent ≈20–30 cm, an axial density gradient of scale-length Ln ≈ wavelength of HPM, and ambient
pressure low enough to maintain electron-neutral collision frequency much less than plasma frequency. The system developed
deploys a ten stage pulse forming network, discharged to the washer-gun to produce pulsed (τ pulse ∼ 100 µs) discharges that get
ejected into an experimental chamber. The system is capable of generating ne ∼ 1018 m−3 and Te ∼ 10 eV. Temporal and spatial
regimes are identified to obtain the required extents of radial and axial ne uniformity of 10 cm and 20 cm, respectively, and a steep
axial gradient Ln ≈ 10 cm. Based on the desired frequency of the interacting HPM (in the range 3-5 GHz) planned for a particular
experimental campaign, the density and spatial density profiles of the plasma can be tailored. The present paper presents an
account of the plasma source and characterization of the plasma.
Published under license by AIP Publishing. https://doi.org/10.1063/1.5051802

I. INTRODUCTION permittivity, and k is the Boltzmann constant, the wave

An experimental system, SYstem for Microwave PLasma
Experiments (“SYMPLE”)1 has been developed at Institute for
Plasma Research in order to investigate the physics of inter-
action of intense electromagnetic (e.m.) waves with an over-
dense, i.e., f p > f µ , plasma, where f p is the plasma frequency
and f µ is the frequency of the interacting wave. It is well
known that an over-dense, unmagnetised plasma do not sup-
port propagation of e.m. waves of conventional power levels,
say a few to a few tens of Watts, where the quiver velocity
attained by the interacting plasma electrons vq = eEµ /mω µ
< vth where Eµ is the wave field, ω µ = 2πf µ , and vth is the
plasma electron thermal velocity. As the intensity of the inter-
acting wave goes up, a variety of linear and nonlinear inter-
actions occur between the wave and the plasma at the criti-
cal layer where f p ≈ f µ ,2–17 resulting in the absorption of the
wave in plasma. When the intensity of the wave under con-
sideration is moderate, with the condition (1/2)o E2µ /ne kTe ∼ 1
satisfied, where ne and Te are the plasma electron density
and electron temperature, respectively,  o is the free space
absorption in plasma is by parametric effects.2–5,11 On the
other extreme, if the wave is highly intense with vq ∼ c (speed
of light), relativistic effects play a governing role in wave
absorption and associated instabilities.6–8,10,16
Although substantial theoretical and simulation studies
exist in this field of e.m. wave absorption in an over-dense
plasma,6,7,10–16 experimental studies are limited, owing to the
limited availability of required resources. On the other hand,
these studies have implications not only in understanding
the underlying physics but also in relating to various issues
of inertial fusion involving laser—plasma interactions.6,7,17
Experiments on laser and plasma interaction8,9,17 have been
undertaken to understand the wave absorption and associ-
ated instabilities. One of the major challenges in carrying out
experiments with lasers is to have appropriate diagnostics
for the phenomena of interest. The typical length scales of
the phenomena are very short (∼skin depth ∼ microns), and
the time scales of interest are extremely rapid (∼a few femto
seconds) causing difficulties in diagnosing the interaction.

Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802 90, 013502-1
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

Furthermore, the plasma needs to have near solid densities, the over-dense regime should be higher, in the range (2–10)
to make it over-dense to the incoming laser beam. × 1017 m−3 , depending on the exact microwave frequency and
gradient scale length. The axial extent of density uniformity
The above diagnostic constraints on resolving appropri-
beyond the steep gradient should be such that the fast elec-
ate length and time scales in case of laser–plasma experi-
trons should remain within the plasma over time scales asso-
ments can be made less stringent by carrying out a frequency
ciated with saturation of the instability. In other words, the
scaled down investigation on microwave–plasma interaction.
distance traversed by the fast electrons in a time duration
Another advantage of such experiments is that plasma with
of the order of the instability growth should be less than the
much lower range of density, ∼(1–10) × 1017 m−3 , yet remain-
axial extent of the uniform plasma. In the radial direction, the
ing over-dense to microwave of frequency of a few GHz,
density uniformity requirement is determined by two con-
can be chosen. Earlier experiments in this direction2–5 have
siderations. The radial extent should accommodate the wave
been limited to interaction with plasma with microwaves of
focusing area ≈ λ2µ . Furthermore, the radial extent should also
power of a few hundred watts to a few hundred kilowatts,
cover a few perpendicular wavelengths of instability, λ⊥ .
where parametric effects have been explored in consider-
able detail although there were difficulties with respect to The growth rate and λ⊥ of instability, as estimated for
time resolved measurements of instabilities on account of parameters valid for SYMPLE, with the help of a 2-D dispersion
the limitations in the data acquisition techniques compati- relation discussed in Ref. 6, show that the plasma should have
ble with high (a few gigahertz) frequencies. Furthermore, the an axial extent of density uniformity in the range 20–50 cm
scale length of plasma density inhomogeneity was greater than considering extreme frequency values (3-5 GHz) and vq /c
the wavelength of the wave, causing limitation in the spec- ∼ 0.2. Furthermore, λ⊥ is ∼1 cm, requiring a radial extent of
trum of phenomena that could be investigated. Experiments ∼10 cm of uniform plasma column. Thus the major parametric
with higher power (a few hundered megawatts) microwaves requisites of the plasma are ne ∼ (1–10) × 1017 m−3 , Te ∼ 10 eV,
are needed to explore many nonlinear and relativistic phe- ambient pressure ∼10−3 mbar to maintain electron neutral col-
nomena governing the wave absorption in plasma. There is a lision ν e−n  f p , radial ne uniformity extent ∼10 cm, axial ne
dearth of experimental investigation in this direction due to uniformity extent ∼20-50 cm, and a steep axial density gra-
non availability of such high power microwave (HPM) sources dient having scale length, Ln = n/(∂n/∂z) ∼ λµ . A schematic
for routine laboratory experiments and due to the technolog- representation of the characteristics required for the plasma
ical complexity, necessitating non-conventional approach of column is given in Fig. 1.
coupling these HPM power to the plasma. The recent advent
A washer-gun based plasma source has been chosen for
in the field of high power microwaves,18–20 with power rang-
our experimental studies as such a source has a potential to
ing from a few hundered megawatts to a few gigawatts, opens
generate high density plasmas.21–25 Furthermore, the mov-
up a new arena of experimental investigations. The proposed
ing front of the plasma ejected into the experimental cham-
experiments in SYMPLE include systematic investigation of
ber would have steep density gradients as required for this
various wave plasma interaction phenomena covering linear,
study. The basic system parameters such as the gun geometry,
nonlinear, and relativistic effects.
parameters of the energizer (power, time duration τ pulse ), and
Generation of plasma that satisfies the critical paramet- the differential pressure scheme needed to satisfy require-
ric requirements has been one of the major experimental ments of the value and profiles of the density, discharge pulse
challenges. The exact requirement of the plasma parameters duration, and low ambient pressure are determined by car-
depends on the particular experimental study under consid- rying out various test experiments and simple theoretical
eration. We briefly discuss the specific case of studying one
of the prominent beam instabilities, the Weibel instability,6,7,17
for the electron beam generated at the critical density layer by
the highly intense microwave (power ∼ 1 GW). This is because
the prerequisites of the plasma estimated for this case of rel-
ativistic wave plasma interaction cover the requirements for
a broad spectrum of investigations covering linear, nonlin-
ear, and relativistic effects. One of the prime requirements is
that the plasma should have a steep axial gradient, Ln ∼ λµ ,
at the wave interaction regime, where λµ is the wavelength
of HPM. Considering the range of frequency, 3–5 GHz, of
the HPM source, this gradient scale length should be around
6–10 cm. This steep gradient (upward as seen in the direc-
tion of the incoming microwave) should be followed by a
regime having uniformity in density. In order to have the con- FIG. 1. A schematic representation of the required plasma characteristics, showing
dition f p = f µ satisfied, the plasma density within the density axial (a) and radial (b) density profiles. The interaction of the incident microwave
gradient regime should be in the range 1–3 × 1017 m−3 ; the occurs in the density gradient regime. The density gradient is followed by a mini-
exact value to be maintained depending on the frequency of mum axial extent of uniformity ∼20-40 cm (depending on f µ ). The radial uniformity
the interacting wave. This essentially means the density in extent should be ∼10 cm.

Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802 90, 013502-2
Published under license by AIP Publishing
 Review of
Scientific Instruments
estimates. The system with these estimated parameters incor-
porated is further subjected to detailed characterization of
the plasma produced. The discussions presented in this paper
include the design considerations of major sub-components
as well as the characterization of the plasma. Emphasis is also
given to optimization of operational parameters such as the
discharge voltage, V d , and ambient pressure, to identify tem-
poral and spatial regimes suitable for wave-plasma interaction
experiments.
The basic experimental setup is presented in Sec. II fol-
lowed by a discussion on the major design considerations in
Sec. III. The results and discussions are given in Sec. IV, and
the conclusions are presented in Sec. V.
II. EXPERIMENTAL SETUP
A washer gun based plasma source has been chosen for
the present study because the moving plasma ejected out of
the gun into the experimental volume can satisfy the require-
ment of steep density gradient. The plasma would however
appear quasistatic to the incoming wave as the ne scale-
length/plasma velocity is τ µ ∼ 100 ns. Furthermore, the
washer gun is a potential source to produce high density
(1018 –1019 m−3 ) plasma,21–23 and therefore even after expan-
sion to the experimental chamber, the plasma is expected
to meet the requirement of the density in the experimental
volume.
The basic configuration of the plasma system is shown in
Fig. 2. The system consists of a vacuum chamber, 1 m long and
30 cm diameter, evacuated to 10−6 mbar pressure, using a dif-
fusion pump and a rotary pump (No. 13). The washer gun is
installed at one end of the vacuum chamber (No. 2).
Argon gas is let into the gun through an opening on the
backside plenum of the washer-gun, with the flow control
achieved by a needle valve. The working pressure in the cham-
ber is controlled by throttling of various valves: the diaphragm
FIG. 2. Schematic of the plasma system (1) chamber, (2) washer-gun, (3) gas inlet,
(4) power supply, (5) charging resistor, (6) matching resistor, (7) bleeding circuit,
(8) gun energizer (single capacitor or PFN), (9) ignitron switch, (10) Pirani guage
G1, (11) Pirani gauge G2, (12) Langmuir probes, (13) diffusion–rotary pumps, and
(14) interspace pumping.
Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802
Published under license by AIP Publishing
ARTICLE scitation.org/journal/rsi
valve installed in the interspace pumping line (No. 14), the
roughing valve installed in the main pumping line (No. 13)
between the rotary pump and the chamber and the baffle
valve separating the diffusion pump and the chamber. Energy
(≈1-6 kJ) stored either in a capacitor or in a pulse forming net-
work (PFN)26 (No. 8) is discharged to the washer-gun upon
closure of the ignitron switch (No. 9).
The discharge voltage V d is measured using a Tektronix
voltage probe (75 MHz, Model P6015) and the discharge cur-
rent Id using a Pearson make current transformer (Pearson,
Model 301X). The differential pressure between the regions
close to the gun (with pressure P1 ) and the experimental cham-
ber (with pressure P2 ) is measured by monitoring pressures
using gauges G1 (Pirani Gauge, No. 10) and G2 (Pirani or Pen-
ning, depending on the pressure range, No. 11), as shown in
Fig. 2. The plasma density, electron temperature, and float-
ing potential are measured using radially and axially movable
Langmuir probes (No. 12).
III. MAJOR DESIGN CONSIDERATIONS
Keeping in view of the critical requisites of the plasma,
certain estimations and test experiments were carried out
to start with, to determine the system components. A brief
discussion on the considerations behind choosing the type
and parameters of major components such as the gun ener-
gizer, gun geometry, and the differential pumping scheme is
presented here.
A. Washer-gun energizer
Experiments are carried out both on capacitive (10 µF, 20
kV) discharges and on discharges using a PFN. The washer-
gun electrode close to the gas plenum is used as the cathode
and the electrode connected to the chamber as the anode.
Experiments to understand the electrical characteristics of
the circuit showed a ringing in current of ∼10 kHz, with-
out connecting an appropriate series resistor with the load
(washer-gun), indicating a line inductance of ∼10 µH, causing
the capacitors to undergo change of polarity. To keep the cir-
cuit in critically damped condition, a resistance of ≈0.9 Ω is
connected in series with the gun, which removed the ringing
effects. Typical plot of V d measured across the series com-
bination of the resistor and washer-gun (curve “a”) and across
the washer-gun alone (curve “b”) are shown in Fig. 3 along with
the discharge current (Id ) (curve “c”) in case of a PFN discharge.
The gun impedance is about 0.05 Ω. The energy dissipation in
the series resistor as well as the gun is measured by graphi-
cal integration of I2 (t)R. It is seen that about 90% of the total
energy is dissipated across the series resistor, and the rest is
coupled to the washer gun.
In order to attain the required axial density uniformity
of the moving plasma at least in some temporal regime, it is
necessary to have a flat top in the discharge pulse, equiva-
lent to the ratio of the required uniformity extent (20-50 cm)
to the flow velocity. Experiments with capacitive discharges,
shown as insets in Fig. 4 although not discussed in detail
90, 013502-3
 Review of
Scientific Instruments
FIG. 3. Temporal profiles of the discharge voltage (V d ) measured across the series
combination of the resistor and washer-gun (a) and across the washer-gun alone
(b) and discharge current (Id ) (c). The inset shows zoomed version of the curve
(b).
FIG. 4. Typical temporal profiles of (a) V d , (b) Id , (c) and ne in case of discharges
using PFN (ten-stage, Ctotal = 50 µF, Ltotal = 50 µH), showing a flat top of ≈100 µs.
Shown in insets are the corresponding plots obtained for capacitive (10 µF, 20 kV)
discharges, having ≈20 µs FWHM.
Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802
Published under license by AIP Publishing
ARTICLE scitation.org/journal/rsi
in this paper, showed plasma temperature ∼10 eV, pointing
to a flow velocity assumed to be ion accoustic, ∼5000 m/s.
Thus discharge pulses of ≈100 µs flat top are desired. Fur-
thermore, with short duration (≈20-40 µs) discharges, it is
not possible to have a duration where the discharge is devoid
of energetic electrons, as discussed in Sec. IV. Simple energy
balance calculations taking into consideration argon ioniza-
tion energy, required density of the plasma, and confinement
time (average axial/radial extent upon ion accoustic velocity)
show that generation of argon plasma of density ∼1019 m−3
requires coupling of ≈3 MW to the washer-gun. Consider-
ing 10% efficiency of coupling of power from the energizer
to the washer-gun, the energy stored in the driver circuitry
should be ≈3 kJ. A ten stage PFN (≈6 kJ) having a capacitance
of 5 µF and an inductance of 5 µH in each stage is developed,
yielding trapezoidal electrical pulses of 15 kV (maximum) with
flat top ≈100 µs and rise time ≈10 µs. Details of design and
development of this PFN is given in Ref. 26. Typical temporal
profiles of V d , Id , and ne are shown in Fig. 4. Attainment of a
plateau of ≈100 µs in V d and Id is clearly visible, in comparison
with the corresponding plots for capacitive discharges (inset).
The density can be seen to mimic the temporal behavior
of Id .
B. The washer-gun configuration
The washer-gun has a set of alternate metal (stainless
steel, 100 mm OD, 5 mm thickness) and insulator washers
(perspex, 120 mm OD, 12 mm thickness) stacked to a length
15 cm. Figure 5(a) shows typical schematic of a washer-gun.
Shown in Figs. 5(b)–5(d) are photographs of three types of
FIG. 5. Sectional view of the washer-gun and the pictures of different washer-guns.
(a) Schematic sectional view. (b) WG30. (c) WG50. (d) Four-gun.
90, 013502-4
 Review of
Scientific Instruments
FIG. 6. Radial density profiles, (a) with WG30, (b) with WG50, and (c) with the
four–gun.
washer-guns used in the present study: WG30 having a single
central hole of 30 mm diameter, WG50 with a single central
hole of 50 mm diameter, and a “four-gun” configuration with
four symmetrically placed holes each of diameter 25 mm,
respectively.
Figure 6 gives a comparison of a typical radial density
dependence, with the use of WG30, WG50, and the four–gun
at z = 20 cm, for V d = 8 kV and ambient pressure in the cham-
ber (P2 ) ∼ 0.1 mbar. Here, z = 0 corresponds to the region
where the gun opens to the chamber. Keeping the discharge
parameters and the gas throughput same, it is observed that
the density of the plasma produced by WG50 is higher as
compared to the plasma produced by WG30. This could be
attributed to the increased gun volume and the associated
reduction in the argon pressure within the gun, estimated to
be in the range of a few tens of Torrs. For the same voltage
8 kV applied between the electrodes, the variation in the E/P
(where E is the electric field, V d /electrode gap) is found to lie
in the range where the increase in E/P results in an enhanced
ionization. As is evident from the figure, the profile for WG50
(curve “b”) has radial extent of ≈10 cm where the density is
reasonably uniform (within about 10%) in comparison with
the profile (curve “a”) of WG30 where the radial profile has a
peak. Thus, WG50 performs better in comparison with WG30,
in terms of density as well as the radial profile. In case of
the four-gun, the density is much lower in comparison with
WG30 and WG50. However, the radial uniformity (curve “c”)
in case of the four-gun is better in comparison with the other
two gun configurations, due to the configuration adopted for
this gun. Here the E/P of the individual cylindrical gun con-
stituent remains the same as that of WG50, and thus, the
density of plasma generated in each of the four paths would
be approximately the same as that of the density produced by
the gun WG50. However, as discussed, the radial uniformity
is enhanced resulting in the averaged density being less. As is
seen in the figure, there is a double folded enhancement in
the spatial extent over which density uniformity is obtained
in case of four-gun in comparison with plasma generated
by WG50, and the density corresponding to the four-gun
Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802
Published under license by AIP Publishing
ARTICLE scitation.org/journal/rsi
FIG. 7. Axial density profile of the density plateau, for WG50, at V d = 8 kV and P2
∼ 0.1 mbar, at radial centre. Shown in the inset is the extrapolation of the observed
density dependence to z values close to the gun opening.
plasma is approximately one half of the density of the WG50
plasma, consistent with the attribution based on density
averaging.
Figure 7 is the typical axial density profile of the den-
sity plateau at V d ≈ 8 kV and P2 ∼ 0.1 mbar, with the use
of WG50, plotted at the radial centre. The density values at
axial locations close to the gun are estimated approximately
by extrapolating the experimentally observed density values,
as shown in the inset, which is needed for the estimation of
differential pumping requirement as discussed below. As is
clear, at these parameters, although there is a steep density
gradient, no axial extent is obtained with density uniformity,
as desired in the present study. All these test experiments
revealed requirement of further improvizations in the system,
to attain required density values and gradients.
C. Differential pressure
In order to have ν e−n  f p , the ambient pressure in the
chamber should be maintained ∼10−3 mbar. Furthermore, as
is clear from the discussions above, the values of density at
chamber pressure P2 ∼ 0.05–0.1 mbar remained generally low,
irrespective of the type of the washer-gun used. One way of
enhancing the density is by lowering of pressure as is evi-
dent from the dependence of density on pressure, as shown
in Fig. 8(a). It was not possible to sustain the plasma below a
certain pressure, called Pmin . For WG30 (curve 1), Pmin was 0.05
mbar, whereas for WG50 (curve 2) and the four–gun, the Pmin
was still higher, ≈0.1 mbar.
The above observations indicated that if, with the use
of WG50 which has a potential to give high density and
acceptable radial profile, the sustainable chamber pressure is
brought down by about two orders of magnitude, the sys-
tem performance can be enhanced in terms of collisional-
ity and attainable density. This necessitated installation of
90, 013502-5
 Review of
Scientific Instruments
FIG. 8. Density variation with pressure prior to upgradation in differential pumping
(a) and after installing appropriate differential pumping (b). Curve (1) shows the
behavior with use of WG30, whereas curves (2) and (3) show the dependence
when WG50 was used.
an appropriate differential pumping arrangement at the port
connecting the gun to the chamber. In order to achieve this
with optimum cost, the required pumping speed (S1 ) of the
differential pumping arrangement is estimated keeping in view
of the flow rate requirements to attain desired plasma density.
This estimate is performed in the following manner. Using the
observed axial dependence of ne , the value of density at the
mouth of the gun (nwg ) is estimated approximately by extrap-
olation, as shown in the inset of Fig. 7. Argon throughput (Q1 )
at the mouth of the gun is estimated from the particle deple-
tion rate given by the product of nwg , velocity, and Awg . Here
Awg is the area of the washer-gun opening and velocity is
taken as ≈ion acoustic. For different values of S1 , the value
of P1 (=Q1 /S1 ) is estimated, as plotted in Fig. 9. Similarly, the
pressure in the chamber (P2 ) is estimated by first calculat-
ing the gas flow rate Q2 to the chamber which is the product
of differential pressure (P1 − P2 ≈ P1 ) and the conductance
of the port connecting the gun to the chamber. P2 = Q2 /S2 ,
where S2 is the speed of the diffusion pump. The value of P2 is
also plotted in the figure. These approximate estimates gave a
good agreement with the P1 and P2 experimentally monitored,
FIG. 9. Theoretically estimated values of P1 and P2 against differential pumping
speed. The curves are cross checked with experimental values marked “◽” for
speed of rotary pump and “•” for speed of Roots pump.
Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802
Published under license by AIP Publishing
ARTICLE scitation.org/journal/rsi
with use of a rotary pump (1000 LPM) for differential pump-
ing. The experimentally observed values are shown in the fig-
ure as points marked “◽.” Having got the theoretical estimate
validated by experiments, the optimum value of S1 needed to
reduce the working pressure (P2 ) by about two orders of mag-
nitude is estimated from the graphs and found to be ≈300 L/s.
A Roots pump (300 L/s) based differential pumping scheme is
therefore installed in the system. The experimentally moni-
tored values of P1 and P2 in this case, for gas throughput as
same as maintained in the earlier experiments (where P2 = 0.1
mbar), are shown in the figure as points marked “•.” Clearly,
the chamber pressure is brought down by about two orders
of magnitude, with this arrangement. ne vs pressure depen-
dence of WG50, after incorporation of this differential pump-
ing arrangement, is shown in Fig. 8(b) (curve 3). Compared to
the curve 2 [Fig. 9(a)] showing Pmin as 0.1 mbar for WG50,
where the dependence is plotted prior to installation of the
Roots pump, the Pmin with the use of Roots pump is reduced to
5 × 10−3 mbar.
The test experiments and theoretical estimates discussed
above lead to finalizing the major system components: the
washer-gun configuration chosen as that corresponding to
WG50 driven by the ten stage PFN and interspace pumping
between the gun mouth and chamber attained with the help
of an appropriate Roots pump.
IV. PLASMA CHARACTERIZATION
Experiments are carried out to optimize operational
parameters and to characterise the plasma. The parameters
generally maintained are listed in Table I.
A. Electron temperature
The electron temperature Te is measured at typical spa-
tial locations with the help of I-V characteristics of the Lang-
muir probes obtained by shot-to-shot measurements. Here,
for any fixed location of the probes, keeping the bias volt-
age fixed at one value, current measurements were carried
out on four (minimum) numbers of plasma shots (each of
100 µs time duration.). Starting from −70 V, the bias voltage
was increased in appropriate steps (of 2 V or 5 V depending
on the regime in the I-V curve) covering the range of bias
voltage −70 V- +20 V. I-V plots corresponding to different
temporal points of the plasma pulse are generated by not-
ing current values averaged over 10 µs, centred around the
temporal point chosen and further averaging over four plasma
shots.
TABLE I. Operational parameters.
Type of Source Washer-gun WG50
Gun energizer unit PFN, 10 stages, 10 µF, 10 µH, 100 µs flat top
Differential pumping Roots pump with speed 300 L/s
Ambient pressure 5 × 10− 3 mbar
Discharge voltage V d 10 kV
Discharge current Id 10 kA
90, 013502-6
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

FIG. 11. Density variation with V d at different axial locations (r = 0).

FIG. 10. Typical Langmuir probe I-V plots at z = 20 cm corresponding to t = 40

µs (a) and t = 90 µs (b). Curve “a” reveals the presence of energetic electrons,
estimated to be ≈30–50 eV, in addition to plasma electrons. The discharge at a
later period (curve “b”) is devoid of energetic electrons.

It is observed that, for the initial period of the flat top,

in addition to the plasma electrons, the discharge contains
energetic electrons (≈30–50 eV). The contribution of the ener-
getic electrons is seen to disappear gradually as time pro-
gresses, and during the discharge period t = 90 µs, the plasma
is nearly devoid of energetic electrons. Typical dependence
of electron current Ie , obtained by subtracting the ion sat-
uration current (Is ) from the total current collected by the
probe, on the bias voltage, corresponding to two typical tem-
poral points, i.e., at t = 40 µs and t = 90 µs, is shown in FIG. 12. Temporal profile of the floating potential (a) at z = 20 cm is compared
with the density profile shown in curve (b). A negative dip in the floating potential
Fig. 10, for the location of the probe z = 20 cm. Here t = 0 during the initial discharge period is indicative of the presence of energetic electron
coincides with the initiation of discharge current. The obser- population.
vations indicate that there is a beam at the gun exit which dis-
perses and thermalizes as the plasma travels from the source.
At any particular location, this reflects in a non-Maxwellian floating probe and the chamber on a 1 MΩ oscilloscope. The
I-V characteristic owing to the contribution of the beam at values lying in the range 10–15 V depending on the location
the initial period and a Maxwellian like plasma at a later of measurement were also cross checked with the estimate
period. obtained from the I-V plots. The curve (a) of Fig. 12 shows the
typical temporal profile of floating potential V f . For compar-
The measurements were carried out at various axial and
ison, the temporal profile of Is (b) is also shown. For about
redial locations. Leaving the axial locations close to the gun (z
initial 40 µs, the potential exhibits a dip in the negative direc-
< 15 cm), Te is found to lie in the range 10–12 eV, with an error
tion followed by a positive flat top of ≈12 V. The negative dip
bar of ±1 eV.
in V f also indicates the presence of energetic electron popu-
lation during the initial discharge period, as also revealed by
B. Density dependence on V d the Langmuir probe I-V characteristics during this discharge
Figure 11 shows the dependence of density on the dis- duration, as discussed above.
charge voltage V d at three axial locations (r = 0). For the volt-
age range studied, the density increases with V d , consistent D. Velocity of the plasma
with energy balance estimate.
Figure 13(a) shows the temporal profile of Is for various
axial locations, at r = 0. It is observed that the front of the dis-
C. Floating potential charge pulse propagates with a velocity of about 1.5 × 104 m/s,
Floating potential (V f ) at various axial and radial locations which is thrice the ion acoustic velocity, consistent with
are measured by directly noting the voltage drop between a the observations in an earlier Ref. 2. The trailing end has a

Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802 90, 013502-7
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

FIG. 13. Temporal profiles of Is at different axial (a) and radial (b) locations. The
Is = 0 lines are shifted in order to make the time shifts easily visible. The front of
the discharge pulse is seen to move about 3 times faster as compared to the ion
acoustic velocity. The trailing edge moves with ion acoustic velocity. In the radial
direction, no propagation velocity could be detected.

propagation velocity of ≈5 × 103 m/s which matches with

the ion acoustic speed. In the radial direction, no propa-
gation velocity could be detected. Typical temporal profiles
obtained at z = 30 cm, at various radial locations is shown in
Fig. 13(b).
E. The density profiles
Typical density profiles have been shown as 3-D plots
in Fig. 14. Profiles at lower discharge voltages (V d = 8 kV or
less) are shown in “a” and those at higher voltages (V d ≥ 10
kV), in “b.” All the profiles correspond to t = 90 µs. In gen-
eral, irrespective of the discharge voltage, there is a steep fall
in density (Ln ≈ 10–14 cm) close to the gun, up to a distance
about 20-30 cm. For lower discharge voltages, the density fall
continues for z > 30 cm (Ln ≈ 20 cm) although with a reduced
slope.
A reasonable axial extent of density uniformity could be
obtained at higher discharge voltages (V d ≥ 10 kV). For the
case shown in Fig. 14(b), it is observed that the steep gradi-
ent at location axially close to the gun is followed by a region
FIG. 14. The r-z profiles, at t = 90 µs, of plasma density at V d = 8 kV (a) and V d =
10 kV (b). A steep axial density gradient exists close to the gun in both cases. In
case of “a,” the density continues to decrease throughout the length of the plasma
column. In (b), an axial extent, ≈20 cm, having uniform density is seen, followed
by a gradient with Ln ≈ 10 cm. The regime where a 3 GHz HPM pulse is to be
focused for wave-plasma interaction studies is also shown in (b).
of 20 cm where the density is uniform (within 10%). This den-
sity plateau with value ≈2 × 1017 m−3 is followed by a steep fall,
having scale length ≈10 cm. For normalized growth rate ≈0.4
(Fig. 1) and considering about five growth times, the distance
traversed by fast electrons (v/c ∼ 0.2) is close to 20 cm for
f µ ∼ 3 GHz. Furthermore, the critical density for a 3 GHz
e.m. wave which is 1.1 × 1017 m−3 coincides with the mid-
dle of the density gradient. This parametric regime therefore
appears suitable for the study of interaction of 3 GHz wave

Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802 90, 013502-8
Published under license by AIP Publishing
 Review of ARTICLE scitation.org/journal/rsi
Scientific Instruments

with plasma. The microwave pulse, as shown in Fig. 14(b), REFERENCES

should thus be focused to about 100 cm2 area at the cen- 1
V. P. Anitha, P. J. Rathod, R. Singh, and D. V. Giri, IEEE Trans. Plasma Sci.
ter of the cylindrical plasma column at the axial location, 44, 2226–2231 (2016).
z = 45 cm, at t = 90 µs. 2
I. R. Gekker and O. V. Sizukhin, JETP Lett. 9, 243–246 (1969) [Pis’ma Zh.
Eksp. Teor. Fiz. 9, 408 (1969) (in Russian)].
While complete axial and radial profiles at higher dis- 3
A. Y. Wong and R. L. Stenzel, Phys. Rev. Lett. 34, 727–730 (1975).
charge voltages are presently under way, observations at typi- 4
R. L. Stenzel, A. Y. Wong, and H. C. Kim, Phys. Rev. Lett. 32, 654–657 (1974).
cal locations indicate that the uniform axial extent can further 5
R. Stenzel and A. Y. Wong, Phys. Rev. Lett. 28, 274–277 (1972).
be enhanced to more than 30 cm. However, keeping the other 6
L. O. Silva, R. A. Fonseca, J. W. Tonge, W. B. Mori, and J. M. Dawson, Phys.
parameters same, the density plateau and the value of ne at the Plasmas 9, 2458 (2002).
mid-gradient are higher, pointing to suitability of these higher 7
Y. Sentoku, K. Mima, Z. M. Sheng, P. Kaw, K. Nishihara, and K. Nishikawa,
power discharges to incident e.m. waves of higher frequency, Phys. Rev. E 65, 046408-1–046408-7 (2002).
up to 5 GHz. These high voltage discharges can however be 8
A. S. Sandhu, G. R. Kumar, S. Sengupta, A. Das, and P. K. Kaw, Phys. Rev.
possibly made suitable for low frequency waves (3 GHz) as well Lett. 95, 025005-1–025005-4 (2005).
9
as by tailoring the pressure parameters. A. S. Sandhu, A. K. Dharmadhikari, P. P. Rajeev, G. R. Kumar, S. Sengupta,
A. Das, and P. K. Kaw, Phys. Rev. Lett. 89, 225002-1–225002-4 (2002).
10
P. K. Kaw and J. M. Dawson, Phys. Fluids 12, 2586–2591 (1969).
V. CONCLUSIONS 11
P. Kaw, E. Valeo, and J. M. Dawson, Phys. Rev. Lett. 25, 430–433 (1970).
12
A washer-gun based, pulsed (100 µs) plasma system has W. L. Kruer and J. M. Dawson, Phys. Fluids 15, 446–453 (1972).
13
W. L. Kruer, Radio Sci. 25, 1351–1357, https://doi.org/10.1029/rs025i006p01351
been developed for the purpose of experimental investiga-
(1990).
tion of high power microwaves with plasma in SYMPLE. The 14
W. L. Kruer, Phys. Plasmas 7, 2270–2277 (2000).
pre-requisites for plasma such as ne ∼ (1–10) × 1017 m−3 , 15
W. L. Kruer, Phys. Plasmas 10, 2087–2092 (2003).
radial uniform extent of ≈10 cm, axial uniformity extent of 16
P. Kaw and J. Dawson, Phys. Fluids 13, 472–481 (1970).
20-50 cm, and a steep axial density gradient scale length ∼λµ 17
F. N. Beg, A. R. Bell, A. E. Dangor, C. N. Danson, A. P. Fews, M. E. Glinsky,
are arrived at from considerations of wave absorption and B. A. Hammel, P. Lee, P. A. Norreys, and M. Tatarakis, Phys. Plasmas 4, 447–
associated instability growth rate and scale lengths. The sys- 457 (1997).
18
tem components and parameters were designed to enable S. Burkhart, J. Appl. Phys. 62, 75–78 (1987).
19
attainment of these requisites of plasma. These components S. H. Gold and G. S. Nusinovich, Rev. Sci. Instrum. 68, 3945–3974 (1997).
include a ten-stage PFN (50 µF Ctotal , 50 µH Ltotal , 100 µs 20
S. K. Chatterjee and Rajeswari Chatterjee, Def. Sci. J. 40, 25–38 (1990).
21
flat-top) that energizes the washer-gun (of 50 cm diame- K. K. Jain, P. I. John, A. M. Punithavelu, and P. P. Rao, J. Phys. E: Sci. Instrum.
13, 928–930 (1980).
ter) and a differential pressure scheme to sustain low enough 22
M. J. McCarrick, R. F. Ellis, J. H. Booske, and M. Koepke, J. Appl. Phys. 61,
pressure in the chamber. The system is capable of generat-
1747–1752 (1987).
ing pulsed plasma with a flat-top density of ∼1018 m−3 and 23
G. Fiksel, A. F. Almagri, D. Craig, M. Iida, S. C. Prager, and J. S. Sarff, Plasma
Te ≈ 10 eV. Operational parameters are also optimized to iden- Sources Sci. Technol. 5, 78–83 (1996).
tify temporal and spatial regimes suitable for HPM absorp- 24
D. J. D. Hartog, D. J. Craig, G. Fiksel, and J. S. Sarff, Plasma Sources Sci.
tion in an over-dense plasma, covering various linear and Technol. 6, 492–498 (1997).
nonlinear phenomena such as resonant wave–plasma inter- 25
T. Ii, K. Gi, T. Umezawa, T. Asai, M. Inomoto, and Y. Ono, Rev. Sci. Instrum.
actions and non-resonant processes involving excitation of 83, 083504-1–083504-5 (2012).
various instabilities, enhanced anomalous collisions, and wave 26
P. J. Rathod, V. P. Anitha, Z. H. Sholapurwala, and Y. C. Saxena, Rev. Sci.
breaking. Instrum. 85, 063503-1–063503-7 (2014).

Rev. Sci. Instrum. 90, 013502 (2019); doi: 10.1063/1.5051802 90, 013502-9
Published under license by AIP Publishing