T h e o p e n – a c c e s s j o u r n a l f o r p h y s i c s

New Journal of Physics

Complex plasma laboratory PK-3 Plus onthe International Space Station

H M Thomas 1, G E Morfill 1,5, V E Fortov 2, A V Ivlev 1,V I Molotkov 2, A M Lipaev 2, T Hagl1, H Rothermel 1,S A Khrapak 1, R K Suetterlin 1, M Rubin-Zuzic 1, O F Petrov 2,V I Tokarev 3 and S K Krikalev 4

1 Max-Planck-Institut für extraterrestrische Physik, 85741 Garching, Germany2 Institute for High Energy Densities, Russian Academy of Sciences,Moscow, Russia3 Yuri Gagarin Cosmonaut Training Centre, Star City, Russia4 RSC-Energia, 141070 Korolev, RussiaE-mail: gem@mpe.mpg.de

New Journal of Physics 10 (2008) 033036 (14pp)Received 18 January 2008Published 27 March 2008Online athttp://www.njp.org/doi:10.1088/1367-2630/10/3/033036

Abstract. PK-3 Plus is the second-generation laboratory for the investigationof complex plasmas under microgravity conditions on the International SpaceStation (ISS). It has more advanced hardware, software and diagnostics than itsprecursor PKE-Nefedov (Nefedovet al2003New J. Phys.5 33). The first experi-ments with PK-3 Plus show the perfect functioning of the apparatus and providemuch better insights into the properties of complex plasmas. In particular, the‘void’ in the center of the complex plasma cloud can now be easily closed,thus providing a much better homogeneity of the complex plasma—a featurewhich was hardly achievable before—but which is essential for many precisionstudies. Moreover, the use of the function generator at frequencies above thedust plasma frequency provides many possibilities for future experiments. Otherinteresting phenomena are related to high densities of the microparticles in thecomplex plasma. These so-called heartbeat and filamentary mode instabilitiescan be investigated in detail, by comparing particle motion with the dischargeglow characteristics.

5 Author to whom any correspondence should be addressed.

New Journal of Physics 10 (2008) 0330361367-2630/08/033036+14$30.00 © IOP Publishing Ltd and Deutsche Physikalische Gesellschaft

2

Contents

1. Introduction: complex plasmas under microgravity conditions 22. Technical description of the PK-3 Plus laboratory 2

2.1. The physical apparatus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2. The Telescience apparatus. . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.3. Estimates of the plasma parameters. . . . . . . . . . . . . . . . . . . . . . . . 7

3. First results from the ISS 73.1. Basic experiments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

4. Conclusion and outlook 12Acknowledgments 13References 14

1. Introduction: complex plasmas under microgravity conditions

The research in complex plasmas is a fast growing field with many applications in differentareas, e.g. in soft matter physics, solid-state physics and fluid physics [1]–[3]. The reason isthat various phenomena, like melting, self-organization, wave propagation, transport, etc can bestudied at the most fundamental, the kinetic level [4]. Complex plasmas are conventional lowtemperature plasmas consisting of electrons, ions and neutrals containing small solid particlestypically in the range of a few micrometres, so-called microparticles. The particles gain chargethrough the interaction with the charged component of the plasma, they are shielded by electronsand ions and interact with each other via the resulting electrostatic potential. This interaction canbe so strong that the particle system can even show crystalline behavior [4]. The microparticlesare typically separated by about 100–500µm, hence using laser light scattering, the dynamicscan be observed at the kinetic level at all relevant time scales. This can provide interesting newinsights into the physics of condensed matter.

Microparticles are strongly affected by gravity, contrary to the other plasma particles, theelectrons and ions [5]. On the ground, a dc electric field is usually employed to compensategravity in a small region of the plasma chamber, in the sheath. This allows measurementsof the complex plasma under two-dimensional (2D) (or 2 1/2-dimensional) conditions. Toperform certain precision measurements, especially of large 3D isotropic systems, microgravityconditions are necessary. Such experiments allow the study of systems and processes notattainable on the ground [6]–[8].

In the next sections, we will describe the new laboratory PK-3 Plus for the research ofcomplex plasmas in microgravity in detail and show the first experimental results.

2. Technical description of the PK-3 Plus laboratory

The PK-3 Plus laboratory is based on the long-term microgravity laboratory PKE-Nefedov(a comparison is shown in table1) and the experience gained over more than four years ofoperation on the International Space Station (ISS) [9]. It consists of two parts, the experimentalblock, containing the physical apparatus, electronics and the experiment computer housed ina hermetically sealed container, and the so-called Telescience apparatus, which is used for the

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

3

Insulator

Plasma & particles

Ground

rf-electrode

Field ofviews

Glass

Particle dispenser

rf electrode

6 cm

9 cm

10 cm

3 cm 5.4 cm

Gas inlet

Gas outlet

Dispenser

rf-electrode

Guard ring

Electronic box

Vacuum flange

Window

Figure 1. The sketches show the 2D (left) and 3D view of the plasma chamber(right).

control of the experiments via a ruggedized Dolch laptop and the storage of the experimentaldata.

2.1. The physical apparatus

The heart of the physical apparatus is a symmetrically driven rf-plasma chamber. A cross-sectional and perspective schematic is shown in figure1. The vacuum chamber consists of aglass cuvette in the form of a cuboid with a quadratic cross-section. Top and bottom flangesare aluminium plates. They include the rf-electrodes, electrical feedthroughs and the vacuumconnections. The electrodes are circular plates of aluminium with a diameter of 6 cm. Thedistance between electrodes is 3 cm. The electrodes are surrounded by a 1.5 cm wide groundshield including three microparticle dispensers on each side. The dispensers are magneticallydriven pistons with a storage volume at their ends. The storage volumes are filled withmicroparticles and 1 mm diameter metal balls and covered with a sieve with an adapted meshsize. The monodisperse particle sizes, their distribution and materials are: 1.55± 0.04µm silica,2.55± 0.04, 3.42± 0.06, 6.81± 0.1, 9.19± 0.09 and 14.9± 0.26µm melamine–formaldehyde(http://www.microparticles.de). The microparticles are dispersed through the sieve into theplasma chamber by electromagnetically driven strokes of the piston. The metallic balls helpto destroy/break up bigger agglomerates and keep the sieve clean.

The electrodes are thermally coupled to the metal flanges. An insulator ring of high ther-mal conductivity (alumina) prevents a temperature gradient between the electrodes and thestructure. The metallic parts of the chamber, the electrodes, flanges and connectors betweenthe upper and lower flange are manufactured from highly conducting aluminium to additionallyprevent a temperature gradient between the upper and lower part of the chamber. This isimportant, because for micrometre particles a temperature gradient of only 1 K cm−1 would giverise to a thermophoretic force of≈10−1 g, which obviously would destroy the microgravityconditions.

The gas–vacuum system, consisting of a turbomolecular pump, a gas reservoir of pureargon and neon and a flow and pressure control, provides high vacuum conditions down

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

4

Table 1. Differences between PKE-Nefedov and PK-3 Plus.

Item PKE-Nefedov PK-3 Plus

Electrodes 42 mm diameter (V2A) 60 mm diameter (Al6060)Width of guard ring 1.5 mm (V2A) 15 mm (Al)Dispensor position Center of rf-electrodes Evenly distributed in grounded guard ringsNumber of dispensors 2 6Particles 3.4 and 6.8µm 1.55, 2.55, 3.42, 6.81, 9.19 and 14.92µmMetal structure of chamber V2A Al6060Connections V2A AlInsulator Macor Alumina of high thermal conductivityGas flow No YesChoice of gases Argon Argon, neon and mixtureTurbo pump Outside of experiment container Inside, close to plasma chamberType of cameras Interlaced PAL CCD camera Progressive scan PAL CCD camera

(25 Hz) (50 Hz full frame rate)Number of cameras 2 4Video recording format High8 video recorder Digital video recorderNumber of recorders 2 4Function generator (FG) Sinusoidal Programmable static voltages on both

0.1–10 Hz; 1–100 Hz Sinus, square wave, triangle wave, randomon both electrodes 1–255 Hz

FG amplitude 13.6 V, 180◦ (push pull) +55 to−55 V on both electrodes, arbitrary phaseRF power range 0–4 Watt (programmable) SameRF modulation None Sinusoidal 1–255 Hz modulation depth 1–100%

to 10−6 mbar and working conditions between 5 and 255 Pa with and without gas flow. Thelatter produces a gas curtain symmetrically around the electrode system flowing from the lowerto the upper side (see 2D sketch in figure1). The prevacuum is provided by a connection of thevacuum system to a valve opening to space. The gas flow is produced by a specially designedsystem, allowing us to operate with the lowest flow rates necessary to provide clean conditionsfor the plasma experiments. For each gas there exists a gas bottle with a volume of 1 litre at≈4.5 bar. The gas bottles are connected to the vacuum system by three-way-cocks. One rotationof this valve releases fresh gas into the vacuum system. During the rotation, the gas bottle is firstconnected to a small reservoir volume of a few mm3 and then this reservoir is emptied into anauxiliary expansion volume of about 100 cm3 from where it continues with laminar flow througha capillary to the inlets around the ground shield of the lower electrode system into the chamber,producing the above-mentioned gas curtain. However, when the valve is energized (typically fora period of 10–100 s), a minor increase of a few percent of the nominal working pressure reachesthe chamber volume.

The plasma is excited by an rf-generator of 4 W maximum at 13.56 MHz with asymmetrical output to drive both electrodes in a push–pull mode. This symmetrical couplingprovides a symmetric, homogeneous and isotropic distribution of the plasma between the twoelectrodes, a mandatory condition for the investigation of microgravity effects in complexplasmas. The following rf-parameters are measured directly at the electrodes through special

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

5

Figure 2. The particle detection system is shown in this figure by means of thefields of views of the three different magnifications and camera positions. Thehigh-resolution camera can be moved along the central axis.

circuits inside of the electronic boxes above and below the plasma chamber (see figure1):effective rf-voltage and rf-current, current in the harmonics and the plasma current to theelectrodes. The current measurements are good indicators for the existence of a plasma andcan be used to control the plasma-on condition at very low power level via an automatic servoloop. Furthermore, they indicate global plasma conductivity and degree of ionization.

The optical particle detection system consists of a laser illumination system and videorecording. For the sake of redundancy two diode lasers (λ = 686 nm) have been implemented.The continuous wave optical power is 40 mW. Depending on the particle size the laser power isadjusted by a pulse width modulation where both period and duty cycle can be adjusted such thatthe illumination is optimal. Both lasers are collimated by a system of several lenses, one of thema cylinder, producing a laser sheet perpendicular to the electrode surface with different openingangles and focal points. Three progressive scan CCD-cameras observe the reflected light at 90◦

with three different magnifications and fields of view. An overview camera has a field of viewof 58.6 mm× 43.1 mm. It shows the full field between the electrodes. A second camera has afield of view of 35.7 mm× 26.0 mm. It shows the left part of the interelectrode system (abouthalf of the full system). A third camera with the highest resolution can be moved along thecentral axis and has a field of view of 8.1 mm× 5.9 mm. The latter is used for high precisionposition measurements of the microparticles. The fields of view of the three cameras are shownin figure2. The cameras follow the PAL standard, have a resolution of 768× 576 pixels and eachprovide two composite interlaced video channels with 25 Hz. The latter can be mixed off-line toa 50 Hz progressive scan video. Interference filters matched to the laser wavelength are used tofilter out the plasma glow. The cameras and lasers are mounted on a horizontal translation stageallowing a depth scan through and, therefore, a 3D view of the complex plasma.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

6

A fourth CCD-camera of the same type as above is added to observe the glowcharacteristics of the plasma. Its field of view is identical to the overview camera. It providesintegrated wavelength information on the local plasma conditions.

The whole physical apparatus is mounted on a base plate. Below this plate, the electronicsand experiment computer are placed. The experiment computer has real time software allowingto measure and process the housekeeping values from the experimental apparatus and tocontrol the experiments in an automatic mode. Pre-programmed software procedures runningadditionally on the experiment computer are used to perform the detailed experiments accordingto scientific needs.

Special electronic components to be mentioned are a video mixer and a low-frequencyfunction generator. The four cameras deliver eight interlaced PAL signal channels, which thevideo mixer distributes on demand on the four video recorders which are part of the Telescienceapparatus (see below). Additionally, the mixer fades a time code signal into the video framesfor synchronization between the different cameras and the housekeeping data and the mixerprovides housekeeping information as teletext in the overview and glow camera signals forquick look purposes as shown in figure2. If both channels from a single camera are recorded, afull frame rate 50 Hz progressive scan video is accomplished by combining the correspondinghalf frames afterwards.

The function generator provides amplitudes at low frequencies to the electrodes, overlaidon the rf-signal. Frequencies between 1 and 255 Hz with amplitudes up to±55 V can be set withdifferent waveforms, e.g. sinusoidal, triangle, rectangle etc and even different phases betweenthe electrodes. This is a very important tool for manipulation and excitation of the complexplasmas, which provides information about e.g. the dispersion relation (DR), wave and shockpropagation.

An alternative possibility to excite the complex plasma by the function generator is amodulation of the rf-amplitude by a sinusoidal function with a frequency between 1 and255 Hz and a modulation depth of 1–100%. Instead of changing the dc-configuration of theplasma system, the latter provides changes in the plasma parameters similar to a pulsedplasma process.

2.2. The Telescience apparatus

The Telescience apparatus contains the control computer, a ruggedized Dolch laptop and anexpansion unit, which includes four digital video recorders. It is connected to the experimentalphysical apparatus via electric cables. The laptop runs on Linux and has a special Panelapplication software for the control of the physical apparatus and the video recorders. Withthis software, the cosmonaut can start automatic preprogrammed experiments, can control itscorrect progress and can manually interfere if errors or off-nominals occur. Additionally, thesoftware allows the full manual control of the experiment. The digital video recorders storethe data in mov-file format with a compression of 4 on 80GB hard drives, which are mountedin flight exchangeable frames. Used hard drives are transported back to Earth for the full datadownload. Additionally video data are transferred live during the experiments from the ISS tothe ground stations in Russia using S-band and are available in the Mission Control Centre inKorolev for online evaluation by engineers and scientists.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

7

2.3. Estimates of the plasma parameters

Siglo-2D is a commercial 2D code [10] which allows the simulation of rf-discharges related toits geometry using a fluid model. The simulation results show the most important parametersof the discharge, e.g. electron temperature, ion/electron density, ionization rate and the plasmapotential, in a 2D plot with isolines for the chosen parameters. These simulations give us arough estimate of the parameters of the PK-3 Plus discharge without microparticles. (A fullcharacterization of the plasma, using Langmuir probes for example, causes a lot of problems,due to the special geometry of this set-up.) In figure3, four simulation results for argon and neonat a pressure of 60 Pa and a rf-amplitude of 20 V are shown. The simulated electric potential,plasma production rate (ionization rate), as well as the measured plasma glow are shown forargon (on the left) and neon (on the right). For argon the ion density is peaked at the center of thedischarge, while the maximum of the electron temperature is close to the electrodes, where themain acceleration of electrons occurs. For neon, due to the longer mean free path of the electronsat a given gas pressure, the ionization is more homogeneously distributed over the full plasmavolume, as can be seen in the plasma glow distribution. The central ion density dependence onthe rf-amplitude and on the neutral gas pressure is shown in figure4 for argon and neon. Fortypical experiment parameters which are used on the ISS the ion density is below 109 cm−3.The electron temperature is for all simulations around 4 eV for argon and 5–6 eV for neon. It isreasonable to expect that the introduction of a large number of microparticles will change theplasma parameters locally, as is known from experiments with nanometre particles, injected orgrown in plasma devices [11]. This has not been measured experimentally for a homogeneousdistribution of particles with sizes larger than one micrometre so far under controlled conditions(we need microgravity to do this). The combination with the plasma glow camera provides thefirst opportunity for such local studies.

3. First results from the ISS

The first experiments performed on the ISS with the new PK-3 Plus equipment were conductedin January 2006 by cosmonaut Valery Tokarev. The goal of these experiments was to comparethe new hardware with PKE-Nefedov [9], which was operational on the ISS from March2001 until July 2005. Therefore, the same particle sizes as those for PKE-Nefedov (3.4 and6.8µm in diameter) and the same discharge parameters were chosen. Additionally, we usednew parameters, e.g. neon as a neutral gas, and different complex plasma manipulation, e.g.the function generator, to extend our knowledge of the behavior of the system. These so-called‘basic experiments’ help us to understand the special features of the new setup and provideinformation for dedicated experiments planned for the future.

Even the first live video that was transferred to ground during the experiment showed theadvanced features of PK-3 Plus, especially concerning the homogeneity and isotropy of thecomplex plasma.

3.1. Basic experiments

The basic experiments are designed to cover a broad range of parameters in particle size anddensity, gas type (argon and neon) and pressure, rf-power and manipulation by low-frequencyexcitation. The first experiment was a pure plasma experiment, to test the system reliability after

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

8

Figure 3. Simulation results for argon and neon and corresponding plasma glowmeasurements.

the stresses during the launch. The experiment was compared to a reference measurement takenshortly before the launch in Baikonur, because for a pure plasma gravity or microgravity hasno influence. This experiment showed the perfect functioning of the laboratory, so the complexplasma experiments could start. The following experiments were dedicated to basic complexplasma conditions.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

9

Figure 4. Simulated ion densities in the center of the electrode system dependingon rf-amplitude and the neutral gas pressure for argon and neon, respectively.

These basic complex plasma experiments always start with a pure plasma. Here, already,a clear difference between argon and neon plasma is visible for similar conditions (shown infigure 3) in the plasma glow, which is in agreement with siglo-2D simulations for the sameparameters. The argon plasma shows the typical features of brighter glow, corresponding to ahigher ionization rate, close to the electrodes due to the wave riding effect at higher pressures.The neon plasma at the same neutral gas pressure shows a homogeneous distribution of the glowbetween the two electrodes due to the larger mean free path of the electrons. Although one couldimagine that a homogeneous plasma distribution would cause a homogeneous and isotropiccomplex plasma distribution, the experiments show the opposite—the argon distribution givesthe best conditions for a homogeneous, void-free complex plasma, while the neon glowdistribution—unexpectedly—results in a big void! The physical reason for this difference isnot yet clear. There are a number of important differences between complex plasma parametersin neon and argon, e.g. different ion mass and mean free path, different plasma density, differentelectron temperature and hence different microparticle charge, different distribution of theambipolar electric field, etc. At the moment we cannot give simple arguments explaining theobserved behavior. Deeper investigations and numerical simulations are needed, which will bepart of our future work. The results will be reported elsewhere.

The injection of the microparticles (in figure5, we show 3.4µm particles) changes theplasma glow tremendously. During the injection the glow is first ‘pushed’ out of the centerby the moving particles, before it relaxes back when the microparticles find their equilibriumpositions and settle centrally. Clear qualitative evidence also exists that the glow changeswith the presence of the microparticles. Comparing the glow before and after the injectionin figures5(a) and (c)—it is seen that the glow increases and spreads towards the center ofthe discharge. In contrast, particle injection into a neon plasma does not change the dischargeconfiguration substantially for the same parameters (figures5(d)–(f)). The resulting distribution(same number of injected microparticles) is dominated by a huge void and a very high densityof the complex plasma cloud. So, neon produces a totally different structure, this is an importantresult for future experiments. Further on in this work, we will concentrate on experimentsin argon.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

10

Figure 5. The combined measurement of the glow and the particle positionsbefore ((a) and (d)), during ((b) and (e)) and after the injection ((c) and (f)) of3.4µm microparticles into the particle free plasma for argon at 30 Pa (a)–(c)and neon at 60 Pa (d)–(f). The glow light intensity is additionally visualizedby contour lines. The microparticles are flowing into the field of view fromthe left side. The ‘bulge’ visible just below the center is a reflection of theplasma glow at the interference filter of the overview camera mounted on theopposite side of the plasma chamber (so it scales with the glow intensity). Moviesof the injection process showing the particle and the plasma glow motion areavailable fromstacks.iop.org/NJP/10/033036/mmedia: (i) movie-fig5a-c.mov:representing (a)–(c) for argon, (ii) movie-fig5d-f.mov: representing (d)–(f) forneon.

3.1.1. Void closure following a gas cycle.The gas cycle which produces continuous gas flowthrough the plasma chamber at first leads to a short disturbance of the particle cloud. Sincethe gas inlets are circularly distributed around the guard ring (see figure1—2D sketch) the gaspressure symmetrically compresses the cloud, leading to void closure for a short time as can beseen in figure6. Depending on the gas pressure and therefore neutral gas damping, this voidclosure lasts from a few seconds to about a minute. Slowly, the void reappears after that time.Sometimes a few particles remain trapped inside the void, without leaving it. This is a hint thatthe potential in the central area is very flat.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

11

Electrode

Electrode

Centre

1 cm

1 cm

Sheath

Sheath

Figure 6. Trajectories of the microparticles (3.4µm) for an exposure time of 4 safter a gas cycle while the void was closed in argon at 60 Pa and a rf-amplitudeof ≈15 V.

3.1.2. Void closure by power decrease.Under special conditions, it is possible to close the voidby reducing the rf-amplitude/power if the number of microparticles is sufficiently high, but nottoo high. In the latter case instabilities occur, which will be described below. The critical numberof microparticles for the onset of instabilities depends on their size, the neutral gas, neutral gaspressure and the rf-power (plasma density). The detailed analysis of this dependency is part ofa future work.

As can be seen from the siglo-2D simulations presented above, decreasing the rf-amplitudedecreases the ion density and, additionally, the plasma potential gets flatter, while the electrontemperature remains nearly constant. A reduced ion density leads to an increased shieldinglength, so that the interaction between neighboring microparticles becomes stronger. This is dueto the fact that the charge remains practically constant (because it mostly depends on the electrontemperature). At the same time, the main bulk forces on the microparticles, the electrostatic(inward) force and the (outward) ion drag force, are decreased in such a way [12] that the ratioof both can be shifted in favor of void closure.

3.1.3. Void closure by low-frequency excitation.The last possibility to close the void describedhere uses the low-frequency function generator with frequencies above the dust plasmafrequency. As mentioned above, the function generator acts on both electrodes with amplitudesup to 55 V and frequencies up to 255 Hz. In principle, we use the function generator to measurethe DR of low-frequency dust modes. This takes place at low pressure (argon: 8 and 15 Pa,neon: 15 Pa) to achieve low damping rates. To measure the DR we start with a low frequencyat about 1 Hz and low amplitude to reach a linear response of the complex plasma [13, 14].Frequency and amplitude are increased stepwise in intervals of a few seconds to measure thefrequency dependence. Since the sheath acts like a diode, the increasing amplitude can be seenas an additional dc-bias on the electrodes which squeezes the particle cloud and finally leads tothe closing of the void (see figure7). This is a method for a very stable void closure, far awayfrom plasma off conditions. It can be used for so-called string formation in electrorheologicalplasmas [15].

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

12

(a)(a) (b)(b)

Figure 7. At frequencies above the dust-plasma frequency and high excitationamplitudes the void is closed. The figures show the stable condition for argon at15 Pa with (a) 3.4 and (b) 6.8µm microparticles, respectively.

3.1.4. Heartbeat and other instabilities.One necessary condition to observe instabilities incomplex plasmas is a high density of the microparticle cloud. There are two differentappearances of the instabilities, depending on the motion of the particles and the plasma. Thefirst one is the so-called heartbeat instability [11], a continuous contraction and expansion ofthe void which the microparticles follow. Ahead of the contraction the plasma glow shows anenhanced intensity in the central void area—the particles at the void boundary fall into thecenter, are stopped by the enhanced plasma drag and again depleted from the void area untilthey reach their equilibrium position (see figures8(a)–(c)). Then, the sequence repeats.

The second instability is the so-called filamentary mode [11, 16]. It appears whenthe void is closed and the glow moves, most probably, rotationally symmetric aboutthe chamber axis. The change in the glow intensity, associated with a plasma densitychange, affects the microparticles. These follow the motion of the glow-like surface waves(see figures8(d) and (e)).

4. Conclusion and outlook

This paper describes the new laboratory PK-3 Plus that was launched to the ISS in December2005. The operational phase started in January 2006. So-called basic experiments have beenperformed to test the apparatus and to explore the parameter space available for the complexplasmas. A homogeneous and isotropic distribution of the complex plasma can be obtainedunder certain conditions. In particular, the use of the function generator at frequencies abovethe dust plasma frequency with amplitudes above 10 V provides a very stable void closure andenables new manipulation possibilities for future experiments. Instabilities like the heartbeator the filamentary mode appear at high microparticle densities and are strongly related tochanges in the plasma glow. In contradiction to expectations—based on the plasma glowcharacteristics—it was found that neon forms a much larger void compared to that of argonplasmas for identical gas and electrical parameters.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

13

Figure 8. (a)–(c): Consecutive images of combined particle and plasma glowmeasurements for a heartbeat instability at 30 Pa in argon with 6.8µm particles.rf-amplitude is 18.6 V. (d) and (e): Consecutive images of combined particleand plasma glow measurements for a filamentary mode instability at 30 Pain argon with 6.8µm particles, appearing at lower rf-amplitude (15.8 V). Theglow light intensity is additionally visualized by contour lines. Movies of theoriginal particle motion and combined with the plasma glow are available fromstacks.iop.org/NJP/10/033036/mmedia: movie-fig8a-c-orig.mov (movie-fig8d-e-orig.mov) shows the original movie of the microparticle motion between theelectrodes for the heartbeat instability (filamentary mode) representing (a)–(c)((d) and (e)); movie-fig8a-c-comb.mov (movie-fig8d-e-comb.mov) shows thesame heartbeat instability (filamentary mode) as before including the overlaidplasma glow intensity and contour lines.

Acknowledgments

This work was supported by DLR/BMWi grant no 50WP0203 and by RFBR grant no06–02–08100. We thank the firm Kayser-Threde for a constructive collaboration duringdevelopment, fabrication, test and delivery of the lab, RSC-Energia for all its support over thefull project, the Mission Control Centre in Korolev for help in planning and performing theexperiments, and, finally, the Yuri Gagarin Cosmonaut Training Centre and the cosmonauts fortheir perfect work.

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)

14

References

[1] Morfill G E, Annaratone B M, Bryant P, Ivlev A V, Thomas H M, Zuzic M and Fortov V E 2002PlasmaPhys. Control. FusionB 44263

[2] Rubin-Zuzic M, Morfill G E, Ivlev A V, Pompl R, Klumov B A, Bunk W, Thomas H M, Rothermel H, Havnes Oand Fouquet A 2006Nat. Phys.2 181–5

[3] Fortov V E, Ivlev A V, Khrapak S A, Khrapak A G and Morfill G E 2005Phys. Rep.4211–103[4] Thomas H M and Morfill G E 1996Nature379806–9[5] Thomas H M, Morfill G E, Demmel V, Goree J, Feuerbacher B and Möhlmann D 1994Phys. Rev. Lett.73

652–5[6] Morfill G E, Thomas H M, Konopka U, Rothermel H, Zuzic M, Ivlev A V and Goree J 1999Phys. Rev. Lett.

831598–601[7] Ivlev A V et al2003Phys. Rev. Lett.90055003[8] Thomas H M, Morfill G E, Ivlev A V, Nefedov A P, Fortov V E, Rothermel H, Rubin-Zuzic M, Lipaev A M,

Molotkov V I and Petrov O F 2005Micrograv. Sci. Technol.16317–21[9] Nefedov A Pet al2003New J. Phys.5 33

[10] Boeuf J P and Pitchford L C 1995Phys. Rev.E 511376[11] Mikikian M, Coudel L, Cavarroc M, Tessier Y and Boufendi L 2007New J. Phys.9 268[12] Lipaev A M et al2007Phys. Rev. Lett.98265006[13] Annibaldi S V, Ivlev A V, Konopka U, Ratynskaia S, Thomas H M, Morfill G E, Lipaev A M, Molotkov V I,

Petrov O F and Fortov V E 2007New J. Phys.9 327[14] Schwabe M, Zhdanov S K, Thomas H M, Ivlev A V, Rubin-Zuzic M, Morfill G E, Molotkov V I, Lipaev

A M, Fortov V E and Reiter T 2008New J. Phys.10033037[15] Ivlev A V et al2008Phys. Rev. Lett.100095003[16] Samsonov D and Goree J 1999Phys. Rev.E 591047

New Journal of Physics 10 (2008) 033036 (http://www.njp.org/)