sandia national laboratories albuquerque,...
TRANSCRIPT
DOE-OFS Plasma Science Center Edward V. Barnat
Greg Hebner, Kraig Frederickson, Paul Miller
Sandia National LaboratoriesAlbuquerque, N.M.
Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
Diagnostic tools for multi-dimensional plasmas
Outline of the talkIntroduction to Sandia
• What we’ve done, what we are doing, where we are goingPlasma density and temperature diagnostics
• Laser-collision induced fluorescence (LCIF)Electric field diagnostics
• Laser-induced fluorescence-dip (LIF-Dip)Surface diagnostics
• Picosecond sum frequency generation ps-SFGFuture directions and concluding thoughts
• Where can connections be made?
LCIF
LIF-Dip
ps-SFG
Low temperature plasma studies at Sandia Sandia has rich history in LTP
• Broad range of customersLong-term investment by DOE
• Responsive to broad range of needsPast
• Sematech, Applied Materials• DOE-BES• Various internal customers
Present• WFO (NASA and others)• Internal customers• DOE-OFS PLSC
Future• Maintain open door and stay responsive
AMAT WFO
DOE-BES
NASA WFO
Internal SNL
Time-resolved, two dimensional electron density and temperature measurements
Laser-collision induced fluorescence (LCIF)• Pump population to an intermediate state• Plasma electrons redistribute portion of population• Monitor fluorescence from neighboring states
Electroncollisions
Degree of redistribution depends on ne, Te
A "good" model is required to predict transfer between levels
Employ a collisional-radiative model (CRM):
5
∑ ∑∑∑∑∑∑ ⎥⎦
⎤⎢⎣
⎡−+⎥
⎦
⎤⎢⎣
⎡−+⎥
⎦
⎤⎢⎣
⎡−=
≠≠<>≠≠ kk
jij
ajki
jii
aikj
jij
jji
jiiije
jij
eji
jii
eij
j NNKNKNANAnNKNKdt
dN
∫∞
⎟⎟⎠
⎞⎜⎜⎝
⎛ −⎟⎟⎠
⎞⎜⎜⎝
⎛==
0
22
42
exp)(2
23
dvvTkvmv
kTmvK
eB
eij
e
eeij
eij πσ
πσ
Cross sections: Yu. Ralchenko, R. K. Janev, T. Kato, D. V. Fursa, I. Bray, F. J. De Heer, Atomic Data and Nuclear Data Tables 94, 603 (2008)
Temperature dependence is introduced via collisional rates Keij
"Electron mixing" "Photon mixing" "Neutral mixing"
Computed and measured excitation rates in Helium
Success is dependant on knowledge of rates
Key transitions in Helium
Density and temperature dependent trends derived from CMR results
Computed temporal evolution of LCIF • Focus on key transitions
6
Key transitions
These transitions have yielded best signals
0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250
ne=1010 e/cm3
ne=1011 e/cm3
ne=1012 e/cm3
“Integration window”Time (ns)
Nor
mal
ized
LC
IF 33P -> 23S 33D -> 23P(389 nm) (588 nm)
33S -> 23P(707 nm)
43D -> 23P(447 nm)
λ=706 nm
Rat
io to
[λ] t
o 38
9 nm
Rat
io 0
f 447
nm
to 5
88 n
mElectron density (cm-3)
kTe=0.5 eV
kTe=1 eV
kTe=2 eV
kTe=4 eV
kTe=6 eV
λ=389 nm
kTe=2 eV
Electron density (cm-3)
Normalized time dependent LCIF trends
Time integrated, ratio trends
Demonstration of LCIF technique:Time modulated plasma
Proof of principle: Examine time evolution of transient plasma • Afterglow of RF modulated plasma,• Pulse bias generated to planar electrode
Setup Data
Captures decay of the plasma• Fast kTe, slower ne,very slow
metastableReasonable agreement between LCIF and probe
• Probe problematic, uncertainties in rates
Analysis
Proof of principle demonstrated
Examine spatial structure around biased electrode• Representative LCIF data used for analysis
8
Key transitions
Temperature measurements become tricky in the sheath….
33P LIF (@389 nm) 33D LCIF (@588 nm)
Electron Density Electron Temperature1010electrons/cm
3
eV
x (mm)
y (m
m)
x (mm)
y (m
m)
x (mm)
y (m
m) C
ounts/s
Counts/s
Counts/s
43D LCIF (@447 nm)
Demonstration of LCIF technique:Static structure of a sheath
Demonstration of LCIF technique:Formation of ion sheath
Proof of principle: Two-dimensional maps of electron density• 50 μs after positive pulse applied to electrode• 20 ns snapshots of LCIF, 30 ns steps
Setup Data
Voltage applied
"Matrix" sheath
Electrons bunching?
Sheath evolving
Initial structure
Final sheath structure
Decent temporal and spatial resolution demonstrated
Current plasmas of interest
We are looking at various plasma systems • Plasma double layers and anode glows• Expanding and flowing plasmas• ECR and magnetized plasmas (B. Weatherford & J. Foster, U. Mich)
What other systems can this technique be applied to?
Transient double layers and anode glows Expanding and flowing plasmas
Time-resolved, two dimensional electric fields
Laser-induced fluorescence-dip spectroscopy (LIF-dip)• Pump population into intermediate state• Redistribute portion of this population to Rydberg state with probe laser• Monitor fluorescence from intermediate state
Rydberg levels offer varying sensitivity to electric fields
Typical 2D LCIF-dip experimental arrangement
Firing of lasers synched to rf phase.13 MHz rf, 20 Hz lasersTime resolved rf voltages
Gate ICCD after firing of the lasers2D snapshot of LIFAccumulate for ~ 100's of laser shots
Repeat as probe laser is incrementally stepped
Typically 30 discrete steps
Post process to determine electric fields
Plot LIF vs. wavelength for each pixelAssign electric field, create 2D map
Scenario: Metal-dielectric interface
How will electric field distribution influence energy deposited into the plasma?
Test case: We know that the electric fields will be non-uniform
Structured boundaries introduce non-uniformities
400 mTorr Argon; 320 Vpp @ 13.56 MHz; 1.5 mm Teflon (κ ~ 2.1)1 M. A. Lieberman, IEEE Trans. Plasma Sci. V17, 338 (1989)
2D measurements demonstrate techniqueElectric fields impacted by boundaries
Analysis above both surfacesSheath potentials are not equal Potential drop across dielectric
Compare measured electric fields to modelsValidate predictions
Analysis of electric fieldsComparison to model
Measured distribution of electric fields
Time resolved studies provide more insightSingle "snapshot" is indicative - at risk of missing whole picture
• Time resolved measurements yield more complete picturePhase-resolved measurements
• Phase-locked LIF-Dip• Phase resolved optical emission (PROES)
Phase-resolved optical emissionPhase-resolved LIF-dip
Capture transient phenomena with time resolved measurements
Incorporate additional diagnostics to help complete the picture
Lower fields mean less energy deposited into the plasma above the dielectric
Less excitation, less ionization
Gradients in potential across the sheath are maintained by a horizontal component of the electric field
Influence angular distribution of ions hitting surface
Electrode topology impacts plasma and surface interactions
Scenario: Different conductorsLess obvious - both are conductors
No induced voltage across either one
Is there any noticeable difference introduced by the dissimilar metals?
0-15 15-10 -5 5 10Electrode Position (mm)
Aluminum
Hei
ght (
mm
)
Fiel
d (V
/cm
)
0
2000
0
5
Stainless Steel
Different metals do influence discharge structure….Under certain circumstances
0-15 15-10 -5 5 10Electrode Position (mm)
Aluminum Stainless Steel
Hei
ght (
mm
)
Fiel
d (V
/cm
)
0
800
0
5
0-15 15-10 -5 5 10
0
8
4
Electrode Position (mm)
Aluminum Stainless Steel
Hei
ght (
mm
)
LIF
(kC
ount
s)
0
2
0-15 15-10 -5 5 10
0
8
4
Electrode Position (mm)
Aluminum Stainless SteelH
eigh
t (m
m)
LIF
(kC
ount
s)
0
2
1s4 Distribution (LIF)Electric Fields
100 mTorr
400 mTorr
320 Vpp @ 13.56 MHz
At a pressure of 100 mTorr, very little difference is observed
Electric Fields 1s4 Distribution (LIF)
At a pressure of 400 mTorr, there is a difference
The choice of metal and choice of operating conditions influences the plasma
Secondary electrons become more important in plasma heating at higher pressures
λγγ x
enn Δ
≈0
Plasma is coupled to emission from the surface,surface is coupled to the plasma……
Species hitting the surface can liberate electrons (Secondary emission γ)Electrons accelerated by sheath into the plasma
Cascading collisions arise if mean-free path (λ) is short compared to sheath thickness
Fraction of γ electrons
Scenario: Electric fields around a charged "dust grain"
Initially funded through BES - Dusty plasmas"Invested" in electric field diagnostics
Thrust II area of interest
Ion wake
Dust particle
E. V. Barnat and G. A. Hebner. J. Appl. Phys. 101, 013306 (2007).
"Far" - Langmuir probe
"Near" - Dusty plasma
Structure is symmetric in bulk plasma, independent of neutral density
Field profiles around the probe are symmetric, and extend ~ 1 mm from probe
200 mTorr
Sheath structure changes with bias• Higher fields, slightly thicker sheath
Sheath structure is independent of pressure• Not dependent on collisions??
Interaction between the probe and plasma extends beyond the sheath
Unexpected changes in excitation observed around the probe:
Differences extend beyond measured sheath
• Sheath thickness ~ 1 mm• Change extends > 5 mm
Differences are not symmetric• Dependent on location w.r.t. powered
electrode?What is the source of the loss?
Float
FloatBias
LIFLIFLIFDifference −
≡
The presence of this “small particle” impacts “large region”
Relative changes in 1s4 LIF
200 mTorr, 25 Watts
Ion dynamics depend on properties of the grain
Above the sheath• VProbe < VLocal
• Some "bending"• Depletion of space charge by
the probe
At the sheath edge• Big q/m grain• VProbe ~ VLocal• Hard to detect focusing
In the sheath• Small q/m grain• VProbe > VLocal
• Ions repelled from probe
200 mTorr, 25 Watts
Sheath bending
Smaller dE/dx
dE/dx = 0
Smaller dE/dx
Floating Probe: VProbe ~ 5 V
SNL is implementing cw laser diagnostics for ion velocity measurements
Developing and implementing cw-laser based diagnostic for measuring ion velocities
• K. FredericksonOur “deliverable” for FY10
Ion wake
Dust particle
This is going to be a challenge…
Argon ion LIF
Velocity
LIF
Pump 3d4F[7/2] @ 811.29 nmMonitor LIF @ 480.6 nm
Future directions: Ultrafast laser diagnostics to probe plasma-surface interface
“Holy Grail” of LTP• Synergistic effects• Peer into mechanisims
Diagnostics are “challenging”• LTP community actively engaged
(Ultra) fast spectroscopy offers access to surface/interface• Intense photon flux to drive non-linear processes• Faster than molecular and electronic relaxation timescales
Sum-frequency generation Time-resolved pump-probe
ωpump
ωProbe
ΔtωVis
ωIR
ωSFG
|0>|1>
ωIR
ωVis
|v>ωSFG
Future directions: Sum frequency generation“Sum-frequency generation (SFH)
• Second order, non-linear Raman technique• ISFG ~ |χ(2)EVisEIR|2
Powerful diagnostic for surface studies• Sensitive to chemical composition• Molecular orientation• Broad range of pressuresHas been implemented in our lab• Darcy Farrow
Conversation is underway with Task III PI's to identify paths towards implementation
ResultsSetup
Concept
ωVis
ωIR
ωSFG
|0>|1>
ωIR
ωVis
|v>ωSFG
Future directions: Time-resolved pump-probeTime-resolved pump-probe spectroscopy
• Femtosecond laser pulses, temporally delayedProbe electronic structure of various materials
• Relaxation times depend on electronic structure of materialPotential PLSC applications
• Poisoned catalysts• Metal hydrides, nitrides, oxide.• Film thickness
Bench-top setup looking for the right mission…
Concept
Untreated
Plasma treated
Samples provided by S. G. Walton, NRL
Setup
Guiding data
Hohlfeld et.at., Chemical Physics 251 (2000) 237-258.
Results
Concluding remarks
28
Thank you!
Emphasis placed on more recent experiments performed mostly by EVBMore powerful, "unique capabilities" at our disposal"Tip of the iceberg"
It is our goal to capitalize on the unique opportunities offered to us through funding of PLSC
Great chance to open the doors and invite community in…..
There are many opportunities to host members of the PLSC (Professors, post-docs and students …)
Is there interest from members of the PLSC at large to pursue these opportunities?
Contact information
29
Greg HebnerManagerSandia National LaboratoriesDept 1128 - Lasers, Plasmas, and Remote Sensing1515 Eubank Blvd. SEAlbuquerque, NM 87185-1423Phone: (505) 844-6831Fax: (505) 844-5459Email: [email protected]
Ed BarnatTechnical StaffSandia National LaboratoriesDept 1128 - Lasers, Plasmas, and Remote Sensing1515 Eubank Blvd. SEAlbuquerque, NM 87185-1423Phone: (505) 284-9828Fax: (505) 844-5459Email: [email protected]
Appendix: Microwave diagnostics
30
Large array of microwave hardware and expertise accessible for plasma characterization
(P. Miller and G. Hebner)
Appendix: Probes
31
Specialization in probes and probe arrays for characterizing plasmaP. Miller, G. Hebner, E. Barnat
Ion Flow "B-dot"
Hairpin resonator
Appendix: Multi-frequency rf plasmas
32
WFO with Applied Materials - High frequency and multi frequency scaling studies
Electromagnetic effectsRadial current density Radial sheath voltage
Frequency scaling Dual frequency
13 MHz 60 MHz 160 MHz
300 mm chamber
Radial current density
Appendix: Magnetized rf plasmas
33
WFO with Applied Materials - Model validation and predictive development
Modified GEC ref cell I & V trends Electric fields
2D Excitation profiles
Magnetized Hydrogen plasma