ion flows and temperatures in a helicon plasma source

Transcription

1 Time-resolved, laser-inducedfluorescence measurements of ion flows and temperatures in a helicon plasma source Earl E. Scime* June, 2010 International Conference on Spectral Line Shapes * Ioana Biloiu, Jerry Carr, Saikat Chakraborty Thakur, Sam Cohen, Matt Galante, Alex Hansen, Amy Keesee, Dustin McCarren, Stephanie Sears, and Xuan Sun

2 HIGHLIGHTS LIF in multiple species demonstrated in low density plasmas with a single 10 mw diode laser. Absorption line width, shift, and amplitude each provide quantitative measures of key plasma source parameters. LIF in expanding helicon plasmas used to demonstrate existence of spontaneous double layer Detailed LIF measurements of ion beam (tomographic, timeresolved, and polarization resolved) address key aspects of double layer formation and stability.

3 OUTLINE LIF schemes employed for Ar II, Ar I, He I Examples of science issues examined through LIF Source optimization (line widths) Three dimensional flows (line shifts) Velocity distribution anisotropy (multiplexed LIF) Absolute He density (optical depth) LIF measurements of ion beams accelerated by double layers Asymmetric optical pumping (Zeeman cooling due to gradients) Tomographic and time-resolved LIF Evidence for beam driven instability suppression of double layer in pulsed plasmas

4 Argon Ion Emission TYPICALLY WE BEGIN OUR THREE-LEVEL LIF SCHEMES FOR LOW-TEMPERATURE PLASMAS WITH EXCITATION FROM A LOW-LYING METASTABLE STATE AR II velocity Distribution Ar II diode laser nm4p4 D 5/ nm T Ar II =.22 ev 3d 4 F 7/ s 4 P 3/2 0.1 Ar II dye laser 461 nm 4s 2 D 5/2 4p 2 F 7/ nm 3d 2 G 9/ Frequency ( ) (GHz) o Stark broadening and natural linewidth are ignorable. Zeeman splitting ignorable for perpendicular injection. For parallel measurements, single circular polarization used.

5 Helium Neutral Emission Argon Neutral Emission OCCASIONALLY NON-METASTABLE AND 4-LEVEL SCHEMES ARE ALSO EMPLOYED FOR LIF ON HE I AND AR I 0.5 Ar I diode laser nm 4p ( 2 P 0 1/2) nm T Ar I =.03 ev 4s ( 2 P 0 1/2) s ( 2 P 0 1/2) 1 4s( 2 P 0 3/2) 1 4s( 2 P 0 3/2) 2 metastable Frequency ( ) (GHz) o 0.35 He I diode laser nm 3 1 P 3 1 D excitation transfer nm T He I =.03 ev S 2 1 P Frequency ( ) (GHz) o

6 COMPACT, PORTABLE, DIODE-LASER BASED SYSTEM FOR LIF LIF Signal (a.u.) Iodine Reference Cell Signal (a. u.) 1: scope 11: PMT 2: powermeter 12: Toptica laser 3: wavemeter 13: chopper 4: ¼ wave retarder / polarizer 5: vacuum chamber 14: laser controller 6: plasma 15: lock-in amp. 7: LIF volume 16: DAQ card 8: laser dump 17: laptop PC 9: collection optics 18: signal 10 interference filter LIF measurements performed with low power (~ 10 mw) diode laser, new (~ 400 mw) diode laser, and 1.2 W dye laser. Shift relative to iodine line or wavemeter standard yields overall flow speed Laser Frequency (GHz)

7 AR II LINE WIDTH (TEMPERATURE) MEASUREMENTS IDENTIFY OPTIMAL ( ~ LH) OPERATING CONDITIONS AND ANTENNA GEOMETRY FOR HELICON PLASMA SOURCE m =+1 helical Nagoya III

8 AR II ION HEATING PEAKS AT SAME MAGNETIC FIELDS AND RF FREQUENCIES AS PREDICTED BY 1 THEORY 1 1 = lh ce ci pi ci (k / ) v thi ~ 1 -> ion Landau Damping Strong ion heating correlated with the lower hybrid frequency in the edge (frequency downshifts because of plasma frequency term)

9 2D AR LINE SHIFT (FLOW) MEASUREMENTS INDICATE STRONG ROTATION. 2D LIF INTENSITY INDICATES CORE ~ 3 CM X 3 CM collection optics Velmex Unislide Motor injection optics

10 MULTIPLEXED AR LIF ENABLES SIMULTANEOUS PARALLEL AND PERPENDICULAR ION VELOCITY DISTRIBUTION MEASUREMENTS Isotropic ion distribution for these source parameters

11 Normalized Intensity LIF Amplitude (a.u.) HE LIF OPTICAL DEPTH MEASUREMENTS FOR ABSOLUTE NEUTRAL DENSITIES Location along beam Z = 55 cm Z = 85 cm Z = 115 cm The optical depth 0 I / I e 0 e z Ng A / 2 8g1 v th L 0.6 Exponential fit: τ = cm -1, corresponds to a metastable density of 1.47 x cm Frequency Shift (GHz) 0.6 The metastable density is related to ground state neutral density (n z-1 (1) where z is the degree of ionization) through a steady-sate collisional-radiative model for an excited state p 0.4 n z p R p n 1 0 e z 1 e z n R p n n Axial Distance from rf Antenna (cm) Using measured plasma parameters, the predicted ground state neutral density is 5.9 x cm -3, ~ 10% ionized.

12 ELECTROSTATIC DOUBLE LAYER (DL) BASICS Scale length ~ 10 s of Debye lengths (shielding distances) quasi-neutrality is violated! accelerated ions Fluid theory for collisionless plasma L.P. Block, Astrophysics and Space Science (1978)

13 Plasma potential (V) SIGNATURE OF SPONTANEOUS DL IN HELICON IS A DOWNSTREAM AR ION BEAM [CHARLES AND BOSWELL, 2003; COHEN ET AL., 2003] Beam V p double layer from RFEA probe

14 IN MC-PIC, ION BEAM FORMS AT EXPANSION POINT, TWO ION POPULATIONS DOWNSTREAM OF DL beam ~ 5 km/s sheath double layer Joint ANU-France Monte Carlo, particle-in-cell simulation

15 FULL DL STRUCTURE, INCLUDING PRE-SHEATH REGION, MEASURED IN HELIX. EXCELLENT AGREEMENT WITH MC-PIC MODEL [Sun et al., PRL 2005] The plasma potential measurements are consistent with the LIF ion energy measurements. The plasma potential tracks the magnetic field strength - decreasing along z. The pre-sheath and sheath are clearly visible and large enough for detailed study. pre-sheath sheath

16 f(v) IONS ACCELERATED TO ~ 10 KM/S DOWNSTREAM OF DL weak double layer as E Beam ~ 3kT e Ion Velocity (m/s) Position (cm)

17 Ion beam energy (ev) ASYMMETRIC OPTICAL PUMPING: AT DL LOCATION IN MNX HELICON SOURCE THERE IS A MAGNETIC FIELD GRADIENT AND AN AR FLOW SPEED GRADIENT Beyond Aperture Plasma Source Region z (cm)

18 ASYMMETRIC AR II LIF ZEEMAN SIGNAL IN MNX SOURCE nm 4 4s P 5/ pd 5/ 2 Ar II nm 4 3d F 7/ 2 m 5/ 2 : 5/ 2 m 7 / 2 : 7 / 2 Each ion group in the metastable state with different m should be equally populated in the source (T ~ 8 ev)?

19 LIF Amplitude (arb) n e 2 Te 0.5 (10 23 cm -6 ev 0.5 ) WHICH COMPONENT IS ABNORMAL (TOO MUCH + OR TOO LITTLE - )? since Clearly, - components are suppressed but how? Radius (cm)

20 ASYMMETRIC OPTICAL PUMPING (ZEEMAN COOLING) Absorption out of the i th state of free ions:* d B N ij i( z ) N i( z ) d L i( ) I ( z,, t ) dt 4 0 * 2 i ( ) i D exp D L M T T * [ I ib( z)][1 V ( z) / c] I( z,, t) laser instensity I( z,, t) I ( ) B ij o Einstein Coefficient o During ions moving from a to b, laser is pumping ions the whole time: N i b a d Ni( z) dz V ( z) dt Distance between a and b is ion mean-free-path The LIF signal is proportional to the pumping intensity of the remaining ions: Bij Ai ( b) ( Ni Ni) d Li ( ) I( b,, t) 4 0 Summing over all Zeeman sublevels: A( b) A ( b) i 6 1 i *M. J. Geockner and J. Goree, J. Vac. Sci. Technol. (1989)

21 R MODEL PREDICTION BASED ON MEASURED COLLISIONALITY CONSISTENT WITH OBSERVATIONS 5 [ ( ( ))(1 ( ) / )] 6 6 I ib zo V zo c Ni( zo) DT o i o i 0 o o N i 1 i 1 i [ o ( I i B ( z ))(1 V ( z ) / c )] i() B i z 1 T e D dz 4 z z i DT V ( z) A ( z ) A ( z ) (1 ) M e I ( ) d N z M N R A A /10 i 10 i Nozzle Magnetic Field (kg) i Potential Significance: Remote measurements of ion collision frequency and ion density for highly ionized plasma And.

22 LIF TOMOGRAPHIC STUDIES IDENTIFY MIRROR RATIO THRESHOLD FOR DL FORMATION

23 THE DL TYPICALLY REQUIRES 10 S OF MS NO SHORT PULSE ROCKETS? More detailed study with 1 ms time resolution: the LIF-determined argon ion velocity distribution function during a 100 ms plasma pulse surface plot showing fast (~ 7.1 km/s) and a slow (~ 0.4 km/s) ion populations.

24 BEAM DELAY DEPENDS ON DEAD TIME, I.E., PERSISTENCE

25 PULSED AND ANTENNA FREQUENCY STUDIES TO EXPLORE BEAM FORMATION PHASE DO BEAMS ATTEMPT TO FORM, BUT CANNOT? Background Beam Moderate mirror ratio of 30 case persistent ion beam Wave amplitude

26 CLEAR CORRELATION WITH FASTER & MORE INTENSE BEAM AND APPEARANCE OF INSTABILITY ~7.2km/s 8 km/s Faster, Mirror Large mirror ratio more of ratio, intense 30, no large waves beam waves before waves appear

27 LARGE MIRROR RATIO CASE SMALL MIRROR RATIO CASE beam-wave anti -correlation

28 HIGHLIGHTS LIF in multiple species demonstrated in low density plasmas with a single 10 mw diode laser. Line shape, width, and amplitude each provide quantitative measures of key plasma source parameters. LIF in expanding helicon plasmas used to demonstrate existence of spontaneous double layer Detailed LIF measurements of ion beam (tomographic, timeresolved, and polarization resolved) address key aspects of double layer formation and stability. Multi-specie LIF in double layers demonstrates that the ions accelerate to a common, bulk, sound speed in the presheath.

29 TIME RESOLVED LIF DEVELOPED TO INVESTIGATE DL FORMATION PHASE ANU experiments indicate some DL formation within 100 s (RFEA measurements difficult to quantify). Stenzel experiments in supersonically expanding plasmas indicate DL forms within a few ms (just an ambipolar field effect?). In HELIX, the DL forms within a few ms, but ion beam energy continues to increase until ~ 100 ms into discharge pulse. This measurement is in the DL and once the DL forms, the background ions are unable to reach the measurement location - so only one population is observed. ion beam Ion Velocity (arb)

30 MONTE-CARLO -PIC SIMULATION SUGGESTS DENSITY GRADIENT TRIGGERS DL FORMATION Common geometry for all helicon source experiments reporting DLs A spatially dependent loss rate models the divergent magnetic field [Meige et al. Phys. Plasmas (2005)]. DL spontaneously forms when the loss rate exceeds a critical value.

31 Amplitude Amplitude A PUZZLE IN THE SUN S MAGNETIC FIELD MEASUREMENTS Asymmetric Stokes V profile could yield erroneous stellar field measurements How astrophysicists measure the Sun s magnetic field? Wavelength (nm) Why asymmetric? Sigwarth: Our results indeed show that unresolved magnetic and velocity fields are responsible for various observed profiles! Stokes V profile ----M. Sigwarth, Astrophysical Journal, 2001 Wavelength (nm)