ARTICLE IN PRESS JID:PLA
AID:15985 /SCO
Doctopic: Plasma and fluid physics
[m5+; v 1.65; Prn:10/08/2006; 14:32] P.1 (1-5)
1
58
2
59
3
60
Physics Letters A ••• (••••) •••–•••
4
www.elsevier.com/locate/pla
61
5
62
6
63
7
64
Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column
8
10
13
a
a
C. Brault , A. Escarguel , Th. Pierre
14 15
a,∗
b
17
21
Abstract
24
30
Keywords: Plasma sources; Recombining plasma
31 32 33
35
1. Introduction
36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
UN CO RR E
PACS: 52.75.-d; 52.50.Dg; 52.25.Xz; 52.75.Xx
It has been recently recognized that removing the heat in front of the plasma facing components inside the divertor of tokamaks is a very important problem. The high neutral pressure inside the divertor leads to the spreading of the heat. The “detached divertor” regime is then obtained where plasma recombination is the dominant process when the electron temperature is very low. In order to get a better understanding of the different processes involved, several linear divertor simulators have been recently built. In this Letter, we report results on the recombination of a very cold plasma obtained in a way different from the existing divertor simulators for instance PISCES-A, Nagdis-II or Magnum-PSI [1–3]. In these devices, a plasma gun is injecting a dense hot plasma into a target chamber where the high pressure induces a rapid cooling of the plasma and leads to the
52 53 54 55 56 57
CT
In a new divertor simulator, a very cold recombining plasma is produced after transverse electric extraction from a dense magnetized plasma column. The plasma is characterized using probes, spectroscopic measurements, and ultra-fast imaging of spontaneous emission. This new technique is shown to be very useful for the investigation of the recombination processes. © 2006 Published by Elsevier B.V.
29
34
a
ED
22
28
a
Communicated by F. Porcelli
20
27
68
a
Received 7 June 2006; accepted 3 August 2006
19
26
b
67
, R. Redon , A. Bois , M. Koubiti , F. Rosmej , R. Stamm , K. Quotb a
18
25
66
a LPIIM, UMR6633 du CNRS, Université de Provence, Case 321, 13397 Marseille cedex 20, France b Laboratoire PROTEE, Université du Sud Toulon Var, 20132 La Garde cedex, France
16
23
PR OO
11 12
65
F
9
* Corresponding author.
E-mail address:
[email protected] (Th. Pierre).
69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
recombination. The evolution of the various parameters, e.g., electron density, electron temperature, is studied along the axis of the high pressure chamber. On the contrary, the recombining plasma is created in our device in a completely different way. It is not obtained after injection of an arc plasma in a dense neutral gas cell, but rather the plasma is created in situ by the injection of energetic electrons and local ionization. An end-plate collector is used in order to drive an axial current in the range 5–50 A. However, this simple technique would only produce a dense plasma with a rather high electron temperature due to the local ionization by the energetic electrons along the magnetized plasma column. Our original design to get a low-temperature recombining plasma consists in involving two additional settings: the plasma chamber is filled at a rather high pressure of working gas in order to get the maximum cooling of the electrons, and a transverse electric field is externally applied across the central magnetized plasma column. A cold plasma jet is obtained after that electric extraction. More exactly, the energetic electrons are collected axially leading to a rapid cooling of the plasma during the early stage of the E × B extraction. A low-temperature drifting
0375-9601/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.physleta.2006.08.018 Please cite this article as: C. Brault et al., Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column, Physics Letters A (2006), doi:10.1016/j.physleta.2006.08.018.
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
ARTICLE IN PRESS AID:15985 /SCO
Doctopic: Plasma and fluid physics
2 3 4 5 6
plasma is formed whose particle flux is directly related to the central density column. At low axial magnetic field, a very fast ion beam is obtained because the drift velocity scales as 1/B. The velocity of the drifting plasma along the extracted plasma jet is measured using various electrical and optical diagnostic techniques.
7 8
2. The experimental setup
9
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
ED
12
CT
11
The new device [4] is depicted in Fig. 1. It consists in a high pressure cylindrical interaction chamber (40 cm in diameter, 1 meter in length) containing the target plasma where recombination processes are studied. The plasma is created in a large chamber (1.4 m diameter, 1 m length) where a low pressure is maintained. An electric discharge inside the source chamber creates a homogeneous plasma with a typical density in the range 1010 e/cm3 . The anode consists in an insulated stainless steel cylinder (1.2 m in diameter, 80 cm length) and the cathode is made of 32 tungsten filaments (0.2 mm diameter, 15 cm length) heated to a temperature of about 2200 K. The discharge voltage is typically 60 volts and the discharge current is lower than 10 amperes. The ionizing electrons are focused on the axis of the target chamber using a conical magnetic structure inside the source. This multipolar magnetic structure is made of a special arrangement of ferrite permanent magnets. A large part of the ionizing electrons produced by the cathode are focused at the entrance of the interaction chamber. The system is equivalent to a compact high power plasma source without the drawbacks of that system especially the overheating of the plasma source. The source chamber is connected to the interaction chamber composed of a metallic tube (40 cm in diameter, 1.2 meter in length) through an injection hole (3 cm in diameter). The interaction chamber is surrounded by 25 equally spaced water-cooled coils (48 cm internal diameter) producing a magnetic field with intensity up to 0.04 tesla. The working gas (argon or helium) is injected in the interaction chamber. Only the source chamber is evacuated through a water-cooled diffusion pump (2000 l/s pumping rate). The pressure in the target chamber ranges from base pressure to 20 Pa. The end-plate collector located at the end of the interaction chamber is biased to a positive voltage in order to drive an axial current along the target chamber. The typical voltage is in the range 40–60 V. An axial current in the range 1 to 50 A is
UN CO RR E
10
48 49 50 51 52 53 54 55 56 57
recorded. This leads to rather high heat deposition on the collector. The measurement of the plasma parameters are obtained using electric probes, high frequency resonance probes, spectroscopic analysis and ultra-fast imaging of spontaneous emission. The Langmuir probe characteristics allows to determine the radial and axial evolution of the electron temperature. Using a high-frequency network analyzer, a high-frequency double probe with 1 cm radial spacing is used for detection of the lower hybrid resonance cone [5]. This technique gives a good estimate of the electron density inside the central plasma column. Spectroscopic measurements are performed with two different experimental devices: the first one is composed of a standard visible spectrometer device (2400, 1200 and 600 mm−1 grooves, 50 cm focal length) coupled to an intensified detector array. It allows to record neutral helium spectra centered at 355 nm with integration times of several seconds. The second one (CERCO) is dedicated to high dispersion Doppler shift measurements. This spectrometer is based on a Bowen chamber with a large numerical aperture (#1.7), and a 3600 grooves/mm holographic grating (dispersion: 13.2 Å/mm) [6]. A 1 mm width entrance slit is used with no significant degradation of the apparatus function, allowing high signal to noise ratio. A high quality objective (magnification factor: 10) is coupled to the output of the CERCO to obtain the very high dispersion (1.32 Å/mm) needed to detect Doppler shift of emission lines. An optical multichannel analyser and an intensified photodiode array, with integration times of 10 s, is used to acquire the spectra.
F
1
[m5+; v 1.65; Prn:10/08/2006; 14:32] P.2 (1-5)
C. Brault et al. / Physics Letters A ••• (••••) •••–•••
2
PR OO
JID:PLA
Fig. 1. Schematics of the Mistral-B device.
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
3. The E × B extraction
88 89 90
In order to apply a transverse electric field leading to the extraction of the plasma from the central plasma column, the end plate collector is segmented into two plates. The plasma potential across the central plasma column is directly influenced by the polarisation of the collecting plates. The potential difference between the plates is chosen typically between 5 and 10 volts. This is sufficient to create a transverse electric field that is established along the whole axial extent of the central column. The ejection of a plasma jet in the limiter shadow due to the E × B drift of both electrons and ions is recorded. The volume of the plasma jet is close to 104 cm−3 leading to a large volume of recombining plasma compared to the volume obtained in other divertor simulators. This technique of electric extraction has very seldom been used. To the best of our knowledge, a comparable proposal has been reported in a preliminary numerical study of the efficiency of isotope separation of uranium ions after selective photo-ionization [7]. It is also possible to use the set-up in pulsed mode applying potential variations to the end plates with very short duration, typically a few tens microseconds. In this situation, the spatio-temporal evolution of the extracted plasma cloud is tracked using a very sensitive ultra-fast camera developed recently (64 pixels spatial resolution, 16 bits encoding). The characteristic convection time and decay lengths of the recom-
Please cite this article as: C. Brault et al., Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column, Physics Letters A (2006), doi:10.1016/j.physleta.2006.08.018.
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
ARTICLE IN PRESS JID:PLA
AID:15985 /SCO
Doctopic: Plasma and fluid physics
[m5+; v 1.65; Prn:10/08/2006; 14:32] P.3 (1-5)
C. Brault et al. / Physics Letters A ••• (••••) •••–•••
bining plasma can be obtained after analysis of the video record (64 000 frames, 200 000 frames per second).
3 4 5
4. Experimental results
6
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
ED
9
The working gas used in preliminary experiments was argon in order to test the efficiency of the new experimental set-up. The pressure is chosen at about 1 Pa inside the target chamber and a lower pressure, i.e., ten times lower, is maintained inside the source chamber. An axial current larger than 20 A is collected by the end plates when an axial magnetic field of 0.02 T is applied. In that typical situation, the electron density reaches 2.1018 m−3 on the axis of the magnetized plasma column, as inferred from the extrapolation of the results of the high frequency resonance probe. The peak electron temperature on the axis of the target chamber is measured by conventional Langmuir probes. Its typical value is about 3 eV. The evolution of the collected current when the voltage of the collector is increased is seen on Fig. 2. The Child–Langmuir law (I ∝ V 3/2 ) is verified in the case of our plasma source: the experimental points are very well distributed along the theoretical curve. The extraction of the plasma is easily obtained when half of the segmented collecting plate is kept floating. In that case, a radial transverse electric field is induced inside the central magnetized plasma column. As a consequence, a spiral plasma jet is extracted from the central column. Fig. 3 displays the spontaneous light emission of the steady-state plasma jet. The central plasma column is not seen because it is hidden behind the end plates. The spatial extension of the jet is related to the recombination processes and to the axial collection of the particles.
CT
8
UN CO RR E
7
The measurement of the velocity of the ejected plasma in argon is obtained from Doppler-shift experiments performed with the spectroscopic device CERCO described in Section 2. A new subpixel method is used to extract the Doppler shift from the ArII (434.8 nm) experimental spectra. Fig. 4 shows the velocity of the argon ions deduced from the Doppler shift, versus the biasing potential between the two parts of the segmented end-collector. The ion velocity is controlled by the E × B drift ejection energy proportional to the applied transverse potential, and by the local plasma potential profile along the line of sight. Local electric field measurements with a Langmuir probe give ion velocities in the range of a few km/s, in agreement with the spectroscopic results. A radially movable Langmuir probe located axially at the middle of the plasma chamber gives the radial profile of the electron temperature inside the extracted plasma jet. A very fast decrease of the electron temperature is seen inside the jet. The temperature is below 1 eV at a radial position distant of 10 mm from the edge of the central magnetized plasma column. As previously reported in the literature, such a very low
F
2
PR OO
1
3
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
Fig. 3. Plasma jet induced by E × B drift of both electrons and ions.
96 97 98 99 100 101 102 103 104
48
105
49
106
50
107
51
108
52
109
53
110
54
111
55 56 57
Fig. 2. Evolution of the current collected by the end-plate with increasing the biasing. The experimental data is in very good agreement with the Child–Langmuir law (V 3/2 ) is recorded (full line).
112
Fig. 4. Ion velocity vi of the ejected argon plasma versus the biasing potential Vap between the two collectors.
Please cite this article as: C. Brault et al., Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column, Physics Letters A (2006), doi:10.1016/j.physleta.2006.08.018.
113 114
ARTICLE IN PRESS
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
58 59 60 61 62 63 64 65 66
Fig. 5. Typical helium spectrum of the recombining ultra-cold ejected plasma. Recombination lines 1s2s 3 P–1sxsd 3 D, with x = 8 to 19, can be observed in the 345–359 nm spectral region.
ED
3
value means that recombination processes are dominant in the extracted plasma jet. The detailed mechanisms leading to the rapid cooling of the plasma jet have been carefully investigated. Systematic measurements of the electron distribution function at different radial positions indicate that the fast ionizing and the fast thermal electrons are lost very quickly by axial collection by the end plates. In fact, the E × B extraction is a very slow process compared to the very short transit time of the fast thermal electrons along the plasma column. After the electrons with a high parallel velocity are collected, the collisional thermalisation of the extracted plasma jet leads to the recorded very low electron temperature. On the other hand, no ionization process occurs behind the limiter and this leads to the depicted situation with a very cold recombining plasma. It is worth noting that similar mechanisms are present in the scrape-off-layer of tokamaks and lead to the formation of the electron temperature pedestal. This process has been extensively studied and numerical studies have investigated the detailed mechanism leading to the existence of the fast decrease of the electron temperature in the scrape-off-layer of tokamaks [8,9]. It is also very important to keep in mind that in the case of a turbulent central plasma column, transverse turbulent electric fields are self excited inside the column. This leads to the expulsion of plasma bursts around the column in the limiter shadow. Inside the bursts, the same cooling mechanism occurs and leads to the observed recombining turbulent halo around the column. This has been also observed on the Nagdis-II device recently [10] in accordance with our recent results on Mistral [11]. Using helium as working gas, similar results are obtained with a lower value of the collected current compared to the case of the argon working gas. The cold ejected plasma has been analyzed by emission spectroscopy (first experimental device described in part 2). The plasma light is coupled to the entrance slit of the spectrometer with an 0.94 mm core diameter optical fiber and a collimating optics. The line of sight is located along the high pressure target chamber, in the shadow of the limiter. Fig. 5 shows a typical spectrum observed in the 335– 360 nm spectral region. Emission lines due to highly excited levels (xd → 1s2p, with x = 9–19) can be observed, indicating important radiative recombination processes typical of very low temperature plasmas. For comparison, these helium recombining lines cannot be observed when the line of sight looks at the warmer central plasma. Langmuir probe characteristic inside the ejected plasma, 1 cm away from the central one, give electron temperatures Te between 0.5 and 1.4 eV, depending on the experimental configuration. The high frequency double probe at the same location gives electron density comprised between 109 and 3 × 1010 cm−3 . It should be noted that in several divertor simulators, Langmuir probe give electron temperature higher than the one obtained by spectroscopic methods. The influence of primary energetic electrons is given to understand this fact. However, this must not happen in our case, because primary ionizing electrons are located only in the central plasma. Next step in the investigation will need the use of a monochromator coupled to a photomultiplicator device in order to
CT
2
[m5+; v 1.65; Prn:10/08/2006; 14:32] P.4 (1-5)
C. Brault et al. / Physics Letters A ••• (••••) •••–•••
UN CO RR E
1
Doctopic: Plasma and fluid physics
F
4
AID:15985 /SCO
PR OO
JID:PLA
study the dynamical evolution of the recombining plasma. The results will be correlated with the series of frames given by the ultra-fast camera. During the analysis of the spatio-temporal evolution of the recombination, the ultra-fast camera is located at a distance of 3 meters on the axis of the magnetized plasma column. It is imaging a squared section (12 × 12 cm2 ) of the target chamber with 1.5 cm spatial resolution. The high sensitivity of the camera is obtained using a squared array of 8 × 8 amplified phototransistors and 16 bits A/D parallel conversion of the signals. Each record is up to 64 000 frames duration at a record rate of 200 000 frames per second. When the central plasma column is turbulent, bursts of plasma are convected around the column by E × B drift. In this situation, the record exhibits luminescent propagating spots. The trajectory of each spot is clearly detected by the camera and the spatio-temporal evolution is analyzed. Fig. 6 depicts a series of pictures corresponding to one event with 5 µs time lag between each frame. The convection velocity is inferred from the frames. The burst moves 6 cm within 30 ms, giving 4 km/s as E × B drift. This leads to a radial electric field of 0.8 V/cm at B = 0.02 T. Most of the measurements give a transverse electric field in the range 50 to 100 V/m. The local measurement of the density inside the jet gives an estimate for the ion flux across the B-field in the range 1019 –1020 m−2 /s. A more precise analysis would give the precise trajectory of the plasma jet across the magnetic field is directly determined by the evolution of the amplitude and orientation of the transverse electric field along the trajectory. At the ejection location, the electric field is induced by the biasing of the segmented endplate. In our set-up, the extraction electric field is horizontal, giving a vertical initial jet. At larger radius, the dominant electric field is the radial electric field established naturally between the confinement cylindrical vessel at ground potential and the
Please cite this article as: C. Brault et al., Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column, Physics Letters A (2006), doi:10.1016/j.physleta.2006.08.018.
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
ARTICLE IN PRESS JID:PLA
AID:15985 /SCO
Doctopic: Plasma and fluid physics
[m5+; v 1.65; Prn:10/08/2006; 14:32] P.5 (1-5)
C. Brault et al. / Physics Letters A ••• (••••) •••–•••
5
1
58
2
59
3
60
4
61
5
62
6
63
7
64
8
65
10
Fig. 6. Ultra-fast record of the expulsion of one the plasma bursts (200 000 frames/sec, 12 cm × 12 cm field) allowing to determine the velocity of the plasma burst. The measurement of the velocity gives the local (vertical) electric field.
F
9
12 13 14 15 16 17 18 19
central magnetized plasma column. In typical experiments, the plasma potential at the edge of the plasma column is about 20 volts. This radial electric field follows a 1/r radial evolution giving approximately an electric field of 1 V/cm across the plasma jet at mid-radius of the containing vessel. This evaluation is in agreement with the velocity measured from the record of the ultra-fast camera.
20
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
5. Conclusions
We have performed experiments in a new divertor simulator where a very cold recombining plasma is obtained by E × B extraction. This new technique leads to the production of a large volume of drifting recombining plasma. The ejection ion velocity of the plasma is measured by Doppler high dispersion spectroscopy. Inside the cold plasma jet, the electron temperature and the electron density are measured using simple and double electric probes. Helium recombination spectra are obtained in helium and exhibit typical features of recombining plasma. The spatio-temporal evolution of the recombining plasma is analyzed during the E × B drift using an ultra-fast camera in agreement with the calculated spatio-temporal evolution of the ejected plasma.
CT
23
UN CO RR E
22
Acknowledgements The authors thanks Dr. G. Leclert for fruitful discussions about the basic physics of the electric extraction. References
[1] C.L. Rettig, W.A. Peebles, N.C. Luhmann Jr., P. Gohil, Rev. Sci. Instrum. 59 (1998) 1536. [2] N. Ezumi, N. Ohno, K. Aoki, D. Nishijima, S. Takamura, Contrib. Plasma Phys. 38 (1998) 31. [3] V. Veremiyenko, R.P. Dahiya, Z. Ahmad, B. de Groot, W.J. Goedheer, R. Engeln, N.J. Lopes Cardozo, in: Proc. 29th EPS-Conference on Plasma Physics and Controled Fusion Montreux, 2002, P2-2.05. [4] M. Matsukuma, Th. Pierre, A. Escarguel, D. Guyomarc’h, G. Leclert, F. Brochard, E. Gravier, Y. Kawai, Phys. Lett. A (2003) 163. [5] Th. Pierre, V. Rohde, A. Piel, Phys. Fluids 4 (1992) 2661. [6] A. Escarguel, R. Redon, A. Bois, Th. Pierre, P. Martinez, in: E. Dalimier (Ed.), Proc. 17th Int. Conf. Spectral Lines Shapes, Frontier Group, 2004, p. 478. [7] N.S. Demidova, V.A. Mishin, Plasma Phys. Rep. 26 (2000) 1045. [8] B. LaBombard, et al., Phys. Plasmas 8 (2001) 2107. [9] S.I. Krasheninnikov, A.Yu. Pigarov, D.A. Knoll, B. LaBombard, B. Lipschultz, D.J. Sigmar, T.K. Soboleva, J.L. Terry, F. Wising, Phys. Plasmas 4 (1997) 1638. [10] D. Nishijima, U. Wenzel, M. Motoyoma, N. Ohno, S. Takamura, S.I. Krasheninnikov, J. Nucl. Mater. 290 (2001) 688. [11] Th. Pierre, A. Escarguel, D. Guyomarc’h, R. Barni, C. Riccardi, Phys. Rev. Lett. 92 (2004) 065004.
ED
21
PR OO
11
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
48
105
49
106
50
107
51
108
52
109
53
110
54
111
55
112
56
113
57
114
Please cite this article as: C. Brault et al., Experimental study of a drifting low temperature plasma extracted from a magnetized plasma column, Physics Letters A (2006), doi:10.1016/j.physleta.2006.08.018.
