Anda di halaman 1dari 9


discussions, stats, and author profiles for this publication at:

Observation of dust torus with poloidal rotation

in direct current glow discharge plasma

Article in Physics of Plasmas · March 2015

DOI: 10.1063/1.4916065


2 201

6 authors, including:

Manjit Kaur Sayak Bose

Swarthmore College Institute for Plasma Research


Joydeep Ghosh Yogesh Saxena

Institute for Plasma Research Institute for Plasma Research


Some of the authors of this publication are also working on these related projects:

Non-neutral plasma in a small aspect ratio "C" shaped toroidal trap View project

Physics of Taylor state plasmas. View project

All content following this page was uploaded by Yogesh Saxena on 25 March 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Observation of dust torus with poloidal rotation in direct current glow discharge
Manjit Kaur, Sayak Bose, P. K. Chattopadhyay, Devendra Sharma, J. Ghosh, and Y. C. Saxena

Citation: Physics of Plasmas (1994-present) 22, 033703 (2015); doi: 10.1063/1.4916065

View online:
View Table of Contents:
Published by the AIP Publishing

Articles you may be interested in

Flush-mounted probe diagnostics for argon glow discharge plasma
Rev. Sci. Instrum. 85, 093505 (2014); 10.1063/1.4895644

Direct-current glow discharges in atmospheric pressure air plasmas

J. Appl. Phys. 91, 2678 (2002); 10.1063/1.1435421

Observation of submicron dust particles trapped in a diffused region of a low pressure radio frequency plasma
Phys. Plasmas 8, 3051 (2001); 10.1063/1.1371237

Initial experiments in the Idaho Dusty Plasma Device

Phys. Plasmas 8, 1746 (2001); 10.1063/1.1365407

Mechanism of dust-acoustic instability in a direct current glow discharge plasma

Phys. Plasmas 7, 1374 (2000); 10.1063/1.873954

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
PHYSICS OF PLASMAS 22, 033703 (2015)

Observation of dust torus with poloidal rotation in direct current glow

discharge plasma
Manjit Kaur,a) Sayak Bose, P. K. Chattopadhyay,b) Devendra Sharma, J. Ghosh,
and Y. C. Saxena
Institute for Plasma Research, Gandhinagar 382428, India
(Received 1 August 2014; accepted 10 March 2015; published online 24 March 2015)
Observation of dust cloud rotation in parallel-plate DC glow discharge plasma is reported here.
The experiments are carried out at high pressures (130 Pa) with a metallic ring placed on the
lower electrode (cathode). The dust cloud rotates poloidally in the vertical plane near the cathode
surface. This structure is continuous toroidally. Absence of magnetic field rules out the possibility
of E  B induced ion flow as the cause of dust rotation. The dust rotational structures exist even
with water cooled cathode. Therefore, temperature gradient driven mechanisms, such as thermo-
phoretic force, thermal creep flow, and free convection cannot be causing the observed dust rota-
tion. Langmuir probe measurement reveals the existence of a sharp density gradient near the
location of the rotating dust cloud. The gradient in the density, giving rise to a gradient in the ion
drag force, has been identified as the principal cause behind the rotation of dust particles. VC 2015

AIP Publishing LLC. []

I. INTRODUCTION probable cause behind the formation of vortices. However,

simulations by Akdim et al.15 showed that the dust vortices
Dusty plasmas are low temperature plasmas, which in
reported by Morfill et al.3 were generated by a combination
addition to electrons, ions, and neutrals, contain micron-
of the ion drag force, electric force, and screened Coulomb
sized dielectric particles. Dust particles acquire large charges
force acting on the dust particles, not by thermophoretic
depending upon their size and background plasma parame-
force. Vaulina et al.4,16–18 and Samarian et al.19 attributed
ters. They increase the complexity of the whole system. The
that dust charge gradient in an inhomogeneous plasma in the
presence of charged dust grains not only modifies the exist-
presence of a non-electrostatic force orthogonal to it causes
ing low-frequency waves but also supports the formation of
an instability which leads to the formation of the dust vorti-
crystalline structures,1,2 voids and self-organized struc-
ces. Mittic et al.20 reported the formation of dust vortices
tures,3,4 waves5,6 and rotation of dust particles, etc. Dust
due to thermal creep flow (TCF) in a low frequency dis-
rotation is an interesting fundamental issue, which has been
charge produced in an inhomogeneously heated vertical
observed in many laboratory experiments. In some of these
glass tube at low pressures. Later, Flanagan et al.21 designed
experiments, magnetic field was applied.7–13 In presence of
an experiment to isolate TCF from other temperature gradi-
magnetic field, the electric field inside the plasma induces an
ent driven effects and verified its role on vorticity formation
azimuthal (E  B) rotation of ions. It causes the levitated
at low pressures. Uchida et al.22 reported 2D dust vortex
dust particles to rotate in azimuthal direction at low pres-
flow in dc plasma near the edge of a metal plate. They attrib-
sures. An exception to this E  B induced dust rotation para-
digm was reported by Carstensen et al.12 They suggested uted the dynamics as the effect of an asymmetry in ion drag
that the dust rotation can also arise due to the ion-induced force near the metal plate. Schwabe et al.23 reported the pres-
neutral flow. Rotation of dust particles and convective vortex ence of two types of gas convection in the presence of tem-
motion in the absence of magnetic field in laboratory plas- perature gradient; (a) free (Rayleigh-Benard) convection at
mas as well as under microgravity conditions has also been high (almost atmospheric) gas pressures and (b) convection
reported. Dust rotation in crystalline structures induced by a induced by thermal creep at low pressures.
biased probe immersed in plasma was observed by Law In a nutshell, there are different mechanisms that can
et al.14 It was suggested that ion wake fields generated due cause dust rotation. In this paper, we present the observation
to the biased probe gave rise to local space charge accumula- of poloidal rotation of dust particles formed between
tion. The variation in space charge leading to the formation parallel-plate dc glow discharge plasma. The structure is also
of a non-uniform electric field was proposed to be the origin continuous toroidally. The novelty of our experiment is that
of the circulation. Morfill et al.3 reported the observation of the dust particles are placed on the lower electrode (not
void (dust-free region) in the central region of a discharge, sprinkled from top). A metallic ring surrounds these dust par-
around which dust vortices get formed under microgravity ticles. The metallic ring causes inhomogeneity in the plasma
conditions. They speculated thermophoretic force as the and the presence of dielectric dust particles enhances it fur-
ther. We find that the observed rotational structures cannot
be explained with the mechanisms proposed by previous
Electronic mail: works mentioned above. We suggest that the rotation arises
Electronic mail: due to a gradient in the ion drag force in the presence of an

1070-664X/2015/22(3)/033703/7/$30.00 22, 033703-1 C 2015 AIP Publishing LLC


This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-2 Kaur et al. Phys. Plasmas 22, 033703 (2015)

electric field. Additionally, the dust particle velocity is The whole body of the cylindrical system along with the
observed to be very high (4–5 cms1) and has a radial gra- lower electrode is acting as a grounded cathode. The inter-
dient in it. The observed rotation/vortices in this kind of electrode separation can be varied from 1 to 10 cm. For the
structures may have interesting relation to various processes present experiment, it is kept fixed at 4 cm, so that the dis-
involving rotational flow of matter, ranging from those at the charge occurs mainly between the two electrodes, the vessel
astronomical scales, such as galactic star formation,24 to does not take part. Vessel side walls are at least 7 cm away
those at the micro- or nano-scales, like mixing of heteroge- from the cathode. Thus, the loss of charged particles diffus-
neous fluids.25 Vortex flow is fundamental to fluid dynamics ing to side walls is small.
also as the vortices are analytic building blocks in a scale- The discharge voltage is varied from 270 V to 470 V
free dynamical formulation of the fluid turbulence.26 with discharge current ranging from 1 mA to 50 mA. A ro-
The vortices forming in laboratory dusty plasmas provide the tary pump allows a base pressure of 1 Pa. The pressure of the
opportunity to study these dynamic building blocks at the filling gas, Argon, is varied between 20 Pa and 150 Pa using
microscopic level of the system which is impossible in any a needle valve. No external magnetic field is used in these
other conventional fluid medium.27,28 experiments. Mono-dispersed Melamine Formaldehyde
The rest of the paper is organised in the following way. particles of diameter 6.48 6 0.08 lm and mass density of
A description of the experimental setup is given in Sec. II. 1.51 g cm3 are placed on the lower electrode inside the ring.
Experimental results are described in Sec. III. Section IV The levitated dust particles are illuminated by a vertical laser
provides estimates of different forces acting on the dust par- sheet of thickness 2 mm from a 100 mWatt, 532 nm laser
ticles. The dust velocity profile analysed with the help of beam and a cylindrical lens. The vertical laser sheet is intro-
Particle Image Velocimetry (PIV) technique is discussed in duced in the system through one of the horizontal radial
Sec. V. A brief discussion is provided in Sec. VI. Finally, the ports. The illuminated dust particles are viewed with a
summary is provided in Sec. VII. sCMOS camera through one of the axial ports of the system.
Coordinate representation of the electrode geometry along
II. EXPERIMENTAL SET UP with dust vortex is shown in Fig. 2. The direction of gravity
is in the negative z direction; r, h, and u represent the usual
The present experiment is carried out in a cylindrical toroidal coordinates. The laser sheet illuminates the dust
vessel of 31 cm diameter and 50 cm length as shown in torus in the middle section. Thus, diametrically opposite
Fig. 1. The discharge is produced between two horizontal dust-torus cross–sections, which look like two vertical dust
parallel-plate stainless steel electrodes. The upper electrode, vortices, represented by two circular disks are seen (as
a disk of diameter 10 cm, is the anode. Its edges are covered shown in Fig. 2). The dust cloud (poloidal cross-section of
by boron nitride (BN) ceramic, an insulator which allows the torus) is considered circular. When viewed from perpen-
only the front part of the anode to take part in plasma pro- dicular to the plane of the laser sheet, the vertical motion of
duction. The exposed diameter of the anode is 9 cm. The the dust particles can be recorded. The sCMOS camera has a
lower electrode, a disk of 13 cm diameter, is a grounded resolution of 2560  2160 pixels with a frame grab rate of
cathode. The cathode has provision for water cooling and 100 frames per second. Higher frame grab rates are possible
thermocouple based temperature measurement. A metallic for lower resolution. The camera is used with a macro-lens
ring, having an inner diameter of 63 mm and outer diameter of focal length 100 mm. A narrow band laser-line filter is
of 82 mm, is kept concentrically on the cathode surface. Its placed between the camera and lens. This arrangement
height above the cathode surface is 3 mm. The ring is electri- allows the laser light scattered from the dust particles to pass
cally and thermally in good contact with the lower electrode. through the filter while blocking the background plasma
light. A DSLR camera is used through another axial port for
viewing as well as capturing pictures of the dust-torus.
A single Langmuir probe, whose tip has a diameter
0.125 mm and length 5 mm, is used to measure the plasma

FIG. 2. Co-ordinate representation: z represents the direction opposite to

FIG. 1. Schematic drawing of the experimental set-up (side view). The cam- gravity, r represents radial direction, h represents the toroidal/azimuthal
era has been aligned in the x direction; laser sheet is pointed in the y direc- direction and u represents the poloidal direction (direction of rotation of the
tion and z represents the vertical direction. dust particles). BN ceramic has been used to cover the edges of anode.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-3 Kaur et al. Phys. Plasmas 22, 033703 (2015)

parameters. The temperature is determined from the slope of

the linear region of lnðIelectron Þ vs VBias plot near the floating
potential,29,30 where Ielectron represents the electronic part of
the probe current and VBias represents the probe bias voltage.
The plasma density is determined from the ion current using
the results of the modified TALBOT and CHOU model.31,32
In order to prevent the dust particle deposition on the
probe,33,34 the probe tip is maintained at a potential much
less than the floating potential by using a DC power supply.
A single cycle triangular waveform is superposed on the DC
bias at a frequency of 200 Hz to acquire the I-V characteris-
tics. This is done in order to ensure that, the time duration,
for which the probe potential is near the plasma potential, is
small (in terms of frequency, it is around 2.5 kHz). As the
dust plasma frequency is <25 Hz, the dust particles will not
be able to respond to this positive probe potential within
such a small duration of time. Hence, the presence of dust
particles is not expected to alter the probe characteristics.


The cathode is covered with the dust particles from the
centre to the inner diameter of the metallic ring. Vacuum
vessel is first purged with Argon gas at high pressure
( 300 Pa) for at least 10 min and then pumped down to low
FIG. 3. Cross sections of the dust torus in the vertical r  z plane at different
pressures. This process is repeated several times before pro- pressures; (a) 70 Pa, (b) 76 Pa, (c) 100 Pa, and (d) 110 Pa, respectively. The
ducing the discharge. The discharge is produced at low pres- laser sheet is passed along the diameter of the cathode through the side radial
sure (p ¼ 20 Pa) with low discharge current (few mA) port and the camera is placed perpendicular through the axial port.
between the two electrodes. Initial observations of the dusty (Multimedia view) [URL:]
plasma illuminated by laser sheet are made by direct sight-
ing. Cathode sheath more than 1 cm thick is seen and the get successively closer and merge into one cloud virtually.
levitated dust particles oscillate with very high amplitudes in These two rotating dust clouds are cross-sections of the torus
the central region inside the metallic ring. Cathode sheath illuminated by the laser at different angles. A top view of
thickness decreases as the working pressure increases. As the the schematic of the scanning of the laser sheet is shown in
pressure is increased to 50 Pa, a stationary three dimensional Fig. 4(a) and the merging of the images of dust clouds taken
crystalline structure is visible at the centre and thick dust with the help of a DSLR camera are shown in Fig. 4(b).
clouds of very fine particles are visible above the metallic When the laser sheet is passed along the diameter of the elec-
ring surface. It is not possible to resolve individual dust par- trode, two structures of rotating dust clouds are well sepa-
ticles in these dust clouds formed above the metallic ring rated, as seen in Fig. 4(b) (i). As the laser sheet is scanned
surface as the particles are too small. The presence of these horizontally away from the diameter, structures get succes-
fine particles might be due to the breaking of mono- sively closer as shown in Fig. 4(b) (ii)–(v). Fig. 4(b) (vi)
dispersed micro-particles by ion bombardment. As we fur- shows the edge of the dust torus. From these images, it is
ther increase the pressure up to 69 Pa, the cathode sheath clear that the dust structures are toroidally continuous. The
thickness decreases further. The mono-dispersed dust par- toroidal nature arises because of the circular symmetry of the
ticles form slanted horizontal sheets of dust particles above electrode and the metallic ring.
the sheath. In addition, a stationary three dimensional crys- In the working range of pressure ðp > 100 PaÞ and dis-
talline structure is observed at the centre. Beyond 70 Pa charge current ðIdis > 20 mAÞ of the present experiment, the
pressure, the dust particles start showing mild rotation in the cathode sheath (thickness  6 mm) covers only a small frac-
vertical plane as is shown in Fig. 3(a). At 76 Pa pressure, the tion of the inter-electrode separation. This kind of rotation is
dust number density as well as the extent of vertical rotation generally observed at low pressures in a horizontal plane in
in the cloud increases, the central three-dimensional crystal- the presence of external magnetic field. The radial electric
line structure disappears and the size of the dust cloud also field induced due to ambipolar diffusion and the vertical
decreases with increase in pressure as shown in Fig. 3(b). At magnetic field produce E  B drift motion of plasma par-
100 Pa, we see a clear vertical rotation of the dust particles ticles resulting in solid body like azimuthal motion of the
in the clouds as shown in Fig. 3(c). Two such fully grown dust particles. However, in the present experiment, there is
rotating clouds at a pressure of 110 Pa are visible near the no external magnetic field. There have been many experi-
diametrically opposite sides of the metallic ring as shown in ments where the metallic ring of finite thickness35 has been
Fig. 3(d) in r  z plane. When the laser light is scanned in used to create a confining potential. The purpose of the ring
the horizontal plane, the images of two rotating dust clouds in those experiments was to confine the levitated dust

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-4 Kaur et al. Phys. Plasmas 22, 033703 (2015)

FIG. 5. (a) Radial profile of the ion density in the absence of dust particles,
(b) radial ion density profiles, and (c) radial plasma potential profiles at dif-
ferent heights from the cathode surface in the presence of the dust.

of the cathode. With the dust particles, it is found that the ra-
dial profile of the plasma density peaked slightly outside the
edge. In addition, the density gradient is steeper as compared
to that without dust particles. The density has a vertical varia-
tion too. As one moves from anode to cathode, the plasma
density is found to increase (ni at z ¼ 9 mm is more than that
at z ¼ 11 mm) as shown in Fig. 5(b). It attains a maximum
value at a height above cathode surface (away from sheath)
and then again decreases in the cathode sheath region.
Presence of dielectric dust particles makes the discharge even
more concentrated near the metal ring. This results in steep
density gradient near the metallic ring. Thus, it would be
interesting to know whether plasma density gradient plays
any role in the formation of rotating dust structures.
FIG. 4. (a) Top view of the assembly of camera, laser and dust torus for
imaging the intersection of the dust torus with the horizontal scan of the ver-
tical laser sheet and (b) different cross-sections of the dust torus viewed
from the axial port using a DSLR camera when the vertical laser sheet is
As discussed in the previous literature, the dust rotation
moved horizontally. The sub parts (i) to (vi) correspond to the position of
the laser sheet as it is moved away from the diameter of the torus, i.e., from can occur as a result of several forces acting on the dust par-
1 to 2. ticles in plasma. Now, we elaborate the different forces act-
ing on the dust particles in our experiments.
particles above the electrode in the region inside the ring. In One of the main forces is due to gravity which arises
our experiment, we have kept a metallic ring on the cathode due to the finite mass of the dust particles. The force due to
and dust particles are permanently placed within this ring gravity ðFg ¼ md gÞ on the micro-particles is estimated to be
before evacuating the vacuum vessel. While viewing through 2.15  1012 N and it acts downward. Using the approach
the naked eye, it is observed that the discharge conditions, described by Khrapak and Morfill36 and taking into consider-
with or without dust particles, are quite different. ation the ion-neutral collisionality the surface potential of
We have measured plasma density at different horizontal the dust particle comes out to be 1:158Te =e. In the ambient
as well as vertical locations using a Langmuir probe at a pres- plasma with plasma screening length kD , the capacitance37
sure of 130 Pa. The results of the measurements are plotted in of the dust particle can be written as
Fig. 5. Without dust particles, it is observed that plasma den-  
sity peaks inside the ring (in between the cathode centre and Cd ¼ 4p0 rd 1 þ :
the ring) and falls off rapidly outside the ring towards the end kD

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-5 Kaur et al. Phys. Plasmas 22, 033703 (2015)

Thus, the charge on the dust particles (Z ¼ Zd e) is esti- 5 rd
mated to be 8.1  103 e for Argon plasma with Te  3 eV Fth ¼  pffiffiffi kB rTn
4 2 rLJ
and ni  3  109 cm–3 . These micron-sized particles after
getting charged experience a repulsive force upward due to where rLJ
3:4Å for Argon gas.
the sheath electric field and get levitated. Besides the above, In this way, the estimated thermophoretic force experi-
ion drag force is also a dominant force. In the case of dc enced by the dust particles for a temperature gradient of
glow discharge, the directed ion flow towards the cathode rT
1 Kcm1 in our experiments is about 1.0  1013 N,
surface drags the dust particles along their way. The ion ve- which is much less compared to the other forces mentioned
locity has mainly two components: thermal and drift. In above. Therefore, the thermophoretic force cannot cause
quasi-neutral bulk plasma at high pressures ðp  100 PaÞ, dust rotation. Additionally, being a conservative force, it
the ion drift velocity is much less than the thermal velocity cannot give rise to rotation of dust particles as its curl will
vi  vTi . For sub-thermal ion flows, the ion drag force on always be zero.
dust particles can be determined using the expression given Apart from thermophoretic force, two other phenomena
by Ivlev et al.38 For frequent collisions case39 with can occur in presence of temperature gradient. One of them,
li  bT kD , the kinetic effects disappear completely and the thermal creep flow, arises because of non-uniformly heated
expression for ion drag force is given by, gas being in contact with a solid surface. In presence of a
   non-uniformly heated body in a rarefied gas, the gas starts
1 Zd e 2 kD flowing along the boundary in the direction of the tempera-
Fi ffi MT ;
6 kD li ture gradient. This phenomenon requires a hot boundary and
occurs only at low pressures (less than 90 Pa)20,21,28 near the
here, MT ¼ vvTii , is the ion thermal Mach number, bT ¼ jZ d je
kD Ti is
boundary. In our case, thermal creep flow has no role, primar-
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ily as the working gas pressure is high. Other phenomenon,
the ion-dust interaction parameter and kD ffi kDi ¼ Ti =4pe2 ni
free convection flow is driven by buoyancy. Free convection
is the effective plasma Debye length. Zd , ni , Ti, vi and vTi are is a phenomenon of heat transport in the fluids generated as a
dust particle charge number, plasma (ion) density, ion tem- result of the density difference in the fluid due to temperature
perature, ion drift velocity and ion thermal velocity, respec- gradient. For free convection to be the reason behind the rota-
tively. Electron temperature is found to be nearly constant tion of the dust clouds, the Rayleigh number should be of the
radially. From the above expression of ion drag force, it is order of few thousand. Rayleigh number for our experiments
observed that there is a strong gradient in the ion drag force is less than 1. Thus, the working pressure range in our experi-
(  1  1010 Nm1) at the region of strong radial density ments is too high for the existence of thermal creep flow and
gradient. This strong radial gradient in ion drag force exerts too low for the existence of free convection.
a torque on the dust cloud present in the vicinity of the Dust particle dynamics can be affected by directed neu-
gradient. tral flow. To mitigate this, Argon gas is fed through the bot-
Cathode heating takes place due to the ion bombardment tom port. This port is connected to the rotary pump via a
on cathode surface and it increases with the discharge cur- gate valve for evacuating the system as shown in Fig. 1. So,
rent. Presence of finite temperature gradient can give rise to only diffused gas is allowed to enter into the system. During
thermo-phoretic force. This may affect dust dynamics the experiment, we closed both the gate valve located at the
depending on its relative magnitude compared to other mouth of the rotary pump and the needle valve (being used
forces. In our experiments, the cathode can attain a tempera- for feeding the gas) simultaneously. By doing so, we do not
ture from 40 C to 65 C without water cooling. We have observe any significant effect on the dust cloud formation
observed that it takes more than an hour for the cathode tem- and rotation. This observation rules out the presence of any
perature to increase up to the maximum temperature depend- directed gas flow and its effect on the dynamics of the dust
ing upon the discharge current. For example, it takes more cloud. Thus, the neutral gas can be assumed to be at rest and
than two hours to reach 65 C with a discharge current of acting as frictional background on the dust particles, resisting
25 mA at a pressure of 120 Pa. However, our experiments are their motion. The value of the neutral frictional force (which
conducted for short periods. Besides, to exclude the effect of is proportional to the instantaneous velocity of the dust parti-
temperature gradient, experiments are carried out with cle) can be estimated40 as
actively cooled cathode. The cathode surface is maintained
at room temperature. In this scenario also, we observe dust Fnd ¼ md dn vd ;
rotation. Although, with active water cooling, temperature
where md , dn , and vd are dust mass, dust neutral collision
gradient is not expected, a temperature gradient of rT
frequency and dust velocity, respectively. For the observed
Kcm1 is considered to estimate the thermophoretic force.
dust velocity of 4 cm/sec and neutral gas density of
The neutral pressure is so high in our experiments that the
3  1016 cm3 at a pressure of 130 Pa, the value of the neu-
distance between the dust cloud and the cathode surface
tral frictional force is 1.4  1012 N.
(cathode-fall thickness) is much larger than the neutral mean
free path (few microns). Thus, the thermophoretic force can
be assumed to be independent of the neutral pressure and
mass of the neutral gas. The expression for thermophoretic A snapshot of the dust cloud captured with the help of
force under such conditions can be written40 as digital camera is shown in Fig. 6(a). Under the same

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-6 Kaur et al. Phys. Plasmas 22, 033703 (2015)

openPIV software, is shown in Fig. 6(c). The particle motion

is recorded at a frame rate of 200 frames/sec (exposure time
is set at 5 ms) and a total of 800 still images are recorded.
Here, the interrogation window size is taken as 32  32 with
a 50% overlapping. PIV analysis provides a total of 799 flow
fields for averaging. The average dust flow velocity profile is
a representative for the majority of the flow fields. Flow ve-
locity is seen to be rotational as indicated by the velocity
vectors. It is observed that the dust velocity is not uniform
throughout the cloud poloidally. As the dust particles move
towards the bottom their velocity increases and attain a peak
value at the bottom. Since the dust particles move very fast
at the bottom of the cloud, so they look like lines (as they
travel a number of pixels in a single exposure time) as they
reach near the bottom of the cloud and become faint in com-
parison to rest of the cloud. From the still images, it is seen
that the particles appear as faint lines near the bottom of the
cloud and as well-defined points at the top. For this reason,
PIV analysis gives errors at bottom locations of the cloud
and reasons attributed are:
(a) The dust density at locations, like (900, 600), is much
higher than that at other locations.
(b) The particle (due to its appearance as a line) occupies
large fraction of the interrogation box and moves a
considerably large distance in the interrogation box in
a single exposure time (in two consecutive frames).
The PIV velocity profiles are best obtained when the
particle moves a small distance41 (less than 30%)
across the interrogation box.
(c) Due to the faintness of the particles (i.e., the lines) at
bottom location i.e., low signal-to-noise ratio, in some
of the trials we do not obtain a velocity vector and it
results into lesser number of velocity vectors at those
locations, for example, from (1200, 600) to (1000,
650), in the ensemble average.
FIG. 6. (a) Image of the poloidal cross-section of the dust torus captured by
a DSLR camera, while the laser sheet is passed along the diameter of torus, For bottom locations, the dust particle velocity can be
(b) image captured by a sCMOS camera at an exposure time of 0.005 s, and
(c) ensemble average of 799 flow fields (800 frames) in PIV analysis with an
estimated by determining its track length in single exposure
interrogation window size of 32  32 with a 50% overlap. Velocity vectors time. Using this method, the dust velocity at (1150, 650) is
show the direction of rotation of the dust particles; the colour bar represents estimated as  7 cm/s. For the rest of the cloud, PIV analysis
the value of dust velocity in cms1. provides very accurate results. In the velocity profile ana-
lysed with the help of PIV, the dust particles are moving in
conditions, a number of sequences of still images are stored
the vertical plane with a velocity of around 4–5 cm/s towards
with the help of sCMOS camera and one of these is shown in
the bottom electrode as shown in Fig. 6(c). The dust particle
Fig. 6(b). PIV analysis is performed on these images to mea-
velocity has been found to be varying radially too in the
sure the particle streaming velocity. The velocity profile can
cloud. At the centre of the cloud (which is close to bottom)
be reconstructed by performing cross-correlation between
dust velocity is minimal and increases as one move away
two consecutive images.41 On each image, identical regions
from the centre and reaches its maximum value at the edges
with the illuminated grains are decomposed into a rectangu-
of the cloud. This is a velocity pattern similar to that of rotat-
lar grid of interrogation boxes. Each interrogation box is a
ing solid bodies.
square of m  m pixels in size, for example, 32  32 pixels,
64  64 pixels, etc. A cross-correlation of two images of a
single interrogation box is used to construct a displacement
vector for all particles within the interrogation box. Since the The main forces, responsible for dust rotation in the sys-
time interval between the two images is known, a velocity tem under consideration, are downward directed sheared ion
vector for each interrogation spot can be computed. For car- drag force, neutral drag force, and the forces due to sheath
rying out PIV analysis, MATLAB based free software called electric field and gravity. The dust particles in the cloud are
openPIV42 is used. A two-dimensional velocity profile of the observed to be flowing towards the cathode surface against
rotating dust cloud, in the r  z plane measured using the sheath electric field in a region, where the ion density

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
033703-7 Kaur et al. Phys. Plasmas 22, 033703 (2015)

and hence the ion drag force is maximum. The dust particles G. E. Morfill, H. M. Thomas, U. Konopka, H. Rothermel, M. Zuzic, A.
start moving upward near the sheath edge, where the ion Ivlev, and J. Goree, Phys. Rev. Lett. 83, 1598 (1999).
O. S. Vaulina, O. F. Petrov, V. E. Petrov, G. E. Morfill, H. M. Thomas,
drag force becomes less. Yu. P. Semenov, A. I. Ivanov, S. K. Krikalev, and Yu. P. Gidzenko, Phys.
In the simplest terms, the above observations of dust Scr. T107, 224 (2004).
cloud rotation can be interpreted as follows. The levitated N. N. Rao, P. K. Shukla, and M. Y. Yu, Planet. Space Sci. 38, 543
dust cloud is confined effectively by a balance of the non- 6
J. Pramanik, G. Prasad, A. Sen, and P. K. Kaw, Phys. Rev. Lett. 88,
uniform electrostatic space potential present above the bottom 175001 (2002).
electrode and the gravitational force acting downwards. This 7
U. Konopka, D. Samsanov, A. V. Ivlev, J. Goree, V. Steinberg, and G. E.
is subjected to additional non-conservative forces, e.g., the Morfill, Phys. Rev. E 61, 1890 (2000).
N. Sato, G. Uchida, T. Kaneko, S. Shimiju, and S. Lizuka, Phys. Plasmas
ion-drag force, neutral friction, etc. For conditions with a spa-
8, 1786 (2001).
tial gradient of the ion drag force is present across the cloud 9
K. Matyash, M. Frohlich, H. Kersten, G. Thieme, R. Schneider, M.
dimensions, the dust cloud experiences a torque. The value of Hannemann, and R. Hiller, J. Phys. D: Appl. Phys. 37, 2703 (2004).
coupling parameter (C) for these clouds in our experiments is F. Cheung, A. Samarian, and B. james, New J. Phys. 5, 75.1 (2003).
I. Pilch, T. Reichstein, and A. Piel, Phys. Plasmas 15, 103706 (2008).
typically 2500 indicating, that the dust particles in these 12
J. Carstensen, F. Greiner, L.-J. Hou, H. Maurer, and A. Piel, Phys. Plasmas
clouds are in strongly coupled state and the cloud as a whole 16, 013702 (2009).
is behaving like a solid body. If two unequal forces act on ei- Y. Saitou and O. Ishihara, Phys. Rev. Lett. 111, 185003 (2013).
ther end of a solid body, it can cause rotation. In an alternate D. A. Law, W. H. Steel, B. M. Annaratone, and J. E. Allen, Phys. Rev.
Lett. 80, 4189 (1998).
regime where the dust cloud behaves like a fluid, the dust fluid 15
M. R. Akdim and W. J. Geodheer, Phys. Rev. E 67, 056405 (2003).
interacts with the vorticity of the dragging fluid43 (ions, neu- 16
O. S. Vaulina, A. A. Samarian, A. P. Nefedov, and V. E. Fortov, Phys.
trals in the present set up). Both these mechanisms can pro- Lett. A 289, 240 (2001).
duce an effective rotation of the dust cloud similar to what is O. S. Vaulina, A. A. Samarian, O. F. Petrov, B. James, and F. Melandso,
Plasma Phys. Rep. 30, 918 (2004).
observed in the present experiment. In a very recent publica- 18
O. S. Vaulina, A. A. Samarian, O. F. Petrov, B. W. James, and V. E.
tion,44 it has been shown that the presence of a gradient in the Fortov, New J. Phys. 5, 82.1 (2003).
ion drag force in a direction not parallel to the electric field A. Samarian, O. S. Vaulina, W. Tsang, and B. W. James, Phys. Scr. 2002,
can cause the formation of dust rotation/vortices under micro- 123.
S. Mittic, R. Sutterlin, A. V. Ivlev, H. Hofner, H. M. Thoma, S. Zhdanov,
gravity conditions. Using this approach, the curl of neutral and G. E. Morfill, Phys. Rev. Lett. 101, 235001 (2008).
frictional force in our experiment comes out to be  21
T. M. Flanagan and J. Goree, Phys. Rev. E 80, 046402 (2009).
3  1010 Nm1 for a measured vorticity value of 12 s1, 22
G. Uchida, S. Lizuka, T. Kamimura, and N. Sato, Phys. Plasmas 16,
which is of the order of gradient in the ion drag force. This 053707 (2009).
M. Schwabe, L. H. J. Hou, S. Zhdanov, A. V. Ivlev, H. M. Thomas, and G.
further supports our claim of gradient in ion drag force as the E. Morfill, New J. Phys. 13, 083034 (2011).
principal cause behind the formation of dust vortices. 24
M. A. Dopita, C. Heisler, S. Lumsden, and J. Bailey, Astrophys. J. 498,
570–578 (1998).
VII. SUMMARY J. P. Gleeson, Phys. Fluids 17, 100614 (2005).
A. J. Chorin, Vortivity and Turbulence (Springer-Verlag, New York, 1998).
In a parallel-plate dc glow discharge plasma, the poloi- H. M. Thomas, U. Konopka, and Victor Steinberg, Phys. Rev. Lett. 92,
175004 (2004).
dal dust rotation in the vertical plane in absence of any exter- 28
M. Schwabe, S. Zhdanov, C. Rath, D. B. Graves, H. M. Thomas, and G. E.
nal magnetic field has been observed and studied in detail. A Morfill, Phys. Rev. Lett. 112, 115002 (2014).
concentric metallic ring is placed on the cathode. Dust par- V. A. Lisovskiy and S. D. Yakovin, Plasma Phys. Rep. 26, 1066–1075 (2000).
ticles are placed on the cathode surface unlike other experi- V. A. Lisovskiy, N. D. Kharchenko, and V. D. Yegorenkov, Phys. D:
Appl. Phys. 41, 125207 (2008).
ments where dust is sprinkled from the top using a dispenser. 31
M. Tichy, M. Sicha, P. David, and T. David, Contrib. Plasma Phys. 34(1),
We have observed a stable dust cloud rotation in the poloidal 59–68 (1994).
(vertical) plane and it is also toroidally continuous. Plasma M. Tichy, P. Kudrna, J. F. Behnke, C. Csambal, and S. Klagge, J. Phys. IV
behaviour is analysed with 2D profile measurements of France 07, C4-397–C4-411 (1997).
M. Klindworth, O. Arp, and A. Piel, Rev. Sci. Intrum. 78, 033502 (2007).
plasma density, electron temperature, and floating potential 34
M. Klindworth, see for Fundamentals and
obtained with the help of electrical probes. Probe measure- Applications of Langmuir Probe Diagnostics in Complex Plasmas (2004).
ments reveal a strong radial density gradient at the location K. Takahashi, T. Oishi, K.-i. Shimomai, Y. Hayashi, and S. Nishino, Phys.
of the dust cloud. The dust cloud, with a high Coulomb cou- Rev. E 58, 7805 (1998).
S. A. Khrapak and G. E. Morfill, Phys. Plasmas 15, 114503 (2008).
pling parameter, is behaving as a solid body. The radial den- 37
A. Melzer, see
sity gradient causes a shear/gradient in the ion drag force. ag_melzer/skript_public.pdf for Introduction to Colloidal (Dusty) Plasmas
The gradient in the ion drag force exerts a torque on the dust 38
particle cloud. The dust particles move downwards against V. Ivlev, S. A. Khrapak, S. K. Zhdanov, and G. E. Morfill, Phys. Rev.
Lett. 92, 205007 (2004).
the sheath electric field in the regions of high ion drag force 39
V. E. Fortov and G. E. Morfill, Complex and Dusty Plasmas: From
and upward in the regions with low values of ion drag force. Laboratory to Space (CRC Press, 2010).
PIV analysis is also performed to estimate the velocity field P. K. Shukla and A. A. Mamun, Introduction to Dusty Plasma Physics
(IOP, Bristol, 2002).
profile of these vortex rotations. This analysis reveals the ve- 41
E. Thomas, Jr., Phys. Plasmas 6, 2672 (1999).
locity profile is similar to that of a rotating solid body. 42
A. Liberzon, R. Gurka, and Z. Taylor, see
piv-matlab for OpenPIV Home Page (2009).
1 43
H. Thomas, G. E. Morfill, V. Demmel, J. Goree, B. Feuer-bacher, and D. M. Laishram, D. Sharma, and P. K. Kaw, Phys. Plasmas 21, 073703 (2014).
Mohlmann, Phys. Rev. Lett. 73, 652 (1994). T. Bockwoldt, O. Arp, K. Ole Menzel, and A. Piel, Phys. Plasmas 21,
J. H. Chu and I. Lin, Phys. Rev. Lett. 72, 4009 (1994). 103703 (2014).

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: Downloaded to IP: On: Tue, 24 Mar 2015 15:38:22
View publication stats