Transcript PPTX - DOE Plasma Science Center

x 10
10
Photos placed in horizontal position
with even amount of white space
between photos and header
100
mm
120
140
10
5
0
2-D Profiles of Electron and Metastable Densities
in Helium Fast Ionization Wave Discharges
B. R. Weatherford and E. V. Barnat
Sandia National Laboratories
Z. Xiong, B. T. Yee, M. J. Kushner, and J. E. Foster
University of Michigan
Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin
Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000. SAND NO. 2011-XXXXP
Overview
 Background on Fast Ionization Waves
 Basic Description of FIW Propagation
 Current Understanding of FIW Discharges




Propagation Velocity
Electron Density Profiles
Metastable Density Profiles
Calculated Electric Fields
(w/ Xiong & Kushner, U. Mich.)
Radial Position, mm
 Experimental & Simulation Setup
 Discussion of Results
16 Torr
x 10
10
10
5
0
5
-5
-10
20
40
60
80
Axial Position, mm
 Influence of Radial E-Fields on FIW Profiles
 Laser Absorption Spectroscopy Study (w/ Yee & Foster, U. Mich.)
 Summary
100
120
140
0
Fast Ionization Waves (FIWs)
 Nanosecond-duration, overvoltage (> breakdown) E-fields
 Diffuse volume discharge at elevated pressures
 Large volume, uniform, high pressure production of:
Photons, charged particles, and excited species
 Interesting Science:
 High voltage + short timescales + fast wave speeds = Hard to capture!
 Large values of E/N  Efficiently drives ionization/excitation processes
 Interesting Applications:




Pulsed UV light sources / laser pumping
High-pressure plasma chemistry
Plasma-assisted combustion
Runaway electron generation
3
FIW Propagation – Positive Polarity
Pre-Pulse Conditions
• High voltage anode, grounded cathode; coaxial geometry
• Grounded outer conductor in contact w/ cathode
• FIW always starts at powered electrode
• Positive polarity FIWs: trace background ionization needed
4
FIW Propagation – Positive Polarity
Application of +HV Pulse
• Applied voltage accelerates electrons toward anode
• Reverse-directed avalanche  electron multiplication
• Electrons move to shield potential @ anode
• Region vacated by electrons = positive space charge
• Photons move ahead of wavefront, add to preionization
5
FIW Propagation – Positive Polarity
Ionization Wavefront Propagation
•
•
•
•
Process continues along length of the tube
Potential gradient moves away from anode
FIW wavefront = moving region of positive space charge + ionization
Residual plasma remains behind wavefront
• Weak fields  relatively little excitation
6
Current Understanding of FIWs
 Capacitive probes  Average E-fields, e- density[1]
 Optical emission  2-D profiles, wave speeds[2-3]
 Radial variations important, but still unclear
 Varying E-field? Higher density or Te? Photons?
 Applications may require volume uniformity
 What process causes the FIW shape to change?
What do profiles tell us about the physics?
Increasing Pressure
 Axial FIW propagation studied extensively
Takashima (2011)
Helium FIW, 20 Torr, 11 kV
Positive Polarity
Negative Polarity
Vasilyak (1994)
7
Experimental Setup - Chamber
 Discharge Tube: 3.3 cm ID x
25.4 cm long
 Coaxial layout: low inductance
 HV electrode inside Teflon
sleeve, grounded shield
 Imaged area: 20-140 mm from
ground electrode





Helium feed gas
Pressure 1-20 Torr
~14 kV (open load) +HV pulses
20 ns duration, 3 ns rise time
1 kHz pulse rep rate
8
2-D LCIF Diagnostic Scheme
 2-D maps of electron densities acquired from
helium line intensity ratios
Data Set A: A
Eff
= ANom
Ratio to [l] to 389 nm
 Pump 23S metastables to 33P with 389 nm laser
l=389 nm
10
3
3
 Electron collisions transfer from 3 P  3 D
l=707 nm
3
3
10
 Image LIF @ 389 nm (3 P-2 S) and LCIF @ 588
nm
(33D-23P) after the laser pulse
10
 Ratio depends linearly on e- density
0
-1
-2
-3
to 587 nm Ratio
to [l]
Ratio
nm
587nm
nmtoto389
0f 447
10
9
10
10
-1
10
kTe=2 eV
Data SetBarnat
A: AEff(2009)
= ANom
10
11
10
12
10
l=389 nm
0
10
kTl=707
e=6 eV nm
-1
10
kTe=4 eV
kTe=2 eV
-2
-2
10
-3
10
kTe=1 eV
9
10
9
10
-1
10
kTe=2 eV
kTe=0.5 eV
1010
11
10 11
Electron density (cm-3)
kTe=6 eV
12
12
10
10
9
Timing & High Voltage Waveform
 Reflected energy: ~85% (V²)
 Long HV cable (15.4 m)
used to separate incident
& reflected pulses
 ICCD optically
synchronized to FIW
Interrogation
Region
 t0 = initial detection of 389
emission without laser
 Reflected pulse: t = 170 ns
 Interrogation: 100-120 ns
 Timing jitter ~ 2-3 ns
 Laser duration ~ 5 ns
 Images accumulated from
repeatable pulses
Forward Pulse
Reflected Pulse
10
2-D Simulation Setup - nonPDPSIM
 2-D fluid model
 Radiation-photon transport
 Includes stepwise ionization
 EEDF from two-term expansion
of Boltzmann equation
 Same pulse shape as experiment
 Approx. open load pre-pulse:
14 kV peak applied at anode
 Assumptions:
 0.1% O2 concentration
(photoionization)
 Initial ne = 108 cm-3
(Xiong and Kushner)
11
Results - FIW Velocities
 FIW speed estimated
from optical emission
intensities vs. time
 389 nm & 588 nm emission




Wavefront Motion
= 389 nm
= 588 nm
FIW Speed: 5 – 20 mm/ns
Peaks @ moderate Pgas
1 Torr: stalls @ x = 40 mm
Decay along tube length
 Due to residual E-field
behind wavefront
Distance to Cathode, mm
12
FIW Speed and ne: Comparison
 Model & experiment agree in several ways:
 Comparable wave velocities:
 Experiment: 0.5 – 2.0 cm/ns, peaks at 2-4 Torr
 Simulation: 0.5 – 1.5 cm/ns, peaks at 8 Torr
 Trend in e- density and ionization rate:
 Peaks at intermediate pressure, then decreases
 Transition from center-heavy to wall-heavy “hollow” profile
 Weak ionization behind wavefront, conserves spatial profile
 Absolute densities are sometimes different
 Experiment: 5x1010 – 3x1011 cm-3
 Simulation: 5x1010 – 8x1012 cm-3
 Model  e- and He* profiles always identical
13
Density Profiles – 1 Torr
 Electrons, metastables
are center-peaked
 Production stops
@ x = 40 mm
Radial Position, mm
 In both LCIF and Model
LCIF Measurements
1 Torr
x 10
3
10
5
2
0
1
-5
-10
20
40
60
80
100
140
Ne,
cm-3
0
Axial Position, mm
NHe*,
arb.
 Corresponds with
decay of FIW speed
 Weak ionization in
residual plasma
 Electrons, metastables
track one another in
simulation
120
11
Simulation Results
Ne,
cm-3
Se,
cm-3-s-1
14
Density Profiles – 4 Torr
Radial Position, mm
 More volume-filling
than @ 1 Torr
 Maximum electron,
metastable densities
 Corresponds to
maximum FIW speeds
in experiment
 Wave traverses the
entire gap
LCIF Measurements
4 Torr
x 10
3
10
5
2
0
1
-5
-10
20
40
60
80
100
120
140
11
Ne,
cm-3
0
Axial Position, mm
NHe*,
arb.
Simulation Results
Ne,
cm-3
Se,
cm-3-s-1
15
Density Profiles – 8 Torr
 Asymmetry due to
slight offset in outer
conductor?
 Radial shift predicted
by simulation
 Metastable profile still
volume-filling
 Residual ionization
follows ne profile
LCIF Measurements
Radial Position, mm
 Electron densities shift
to off-center peak
8 Torr
x 10
2
10
5
1.5
0
1
-5
0.5
-10
20
40
60
80
100
120
140
11
Ne,
cm-3
0
Axial Position, mm
NHe*,
arb.
Simulation Results
Ne,
cm-3
Se,
cm-3-s-1
16
Density Profiles – 16 Torr
Radial Position, mm
 Electron densities
strongly wall-peaked
 Simulation shows
excellent agreement in
electron profile shape
 Metastable densities
volume-filling, with
slight shoulder off-axis
LCIF Measurements
16 Torr
x 10
10
10
5
0
5
-5
-10
20
40
60
80
100
120
140
10
Ne,
cm-3
0
Axial Position, mm
NHe*,
arb.
Simulation Results
Ne,
cm-3
Se,
cm-3-s-1
17
Radial Profiles – Experimental
1 Torr
1 Torr
LIF (kCounts)
Metastable
distribution
Helium Metastable
4
2.0
3
1.5
2
1.0
1
0.5
0
-10
LIF (k Counts)
1.0
75High
mmPressure:
Concave ne
140 mm
profile
0.2
0.0
5
Radial position (mm)
10
High Pressure:
Broad He*
profile
10 mm
0.4
X = 10 mm
0
10
0.6
0.5
-5
Electron densities
-5
0
5
Radial position (mm)
0.8
75 mm
-10
-10
10
140 mm
1.5
Low Pressure:
Center-peaked
ne and He*
0.0
Radial position (mm)
2.0
20 Torr
20 Torr
Helium Metastable
-5
0
5
Electron densities
(x1011 e/cm3)
0.0
-10
-5
0
5
Radial position (mm)
10
18
 Experiment: Different ne, NHe*
radial profiles @ high pressure
 Metastables shifted to center
 Model: ne, NHe* track each other
 Model results rule out:
Radial Position, mm
Electrons vs. Metastables
16 Torr
x 10
10
10
5
ne
0
-5
5
-10
20
40
60
80
100
120
0
140
Axial Position, mm
NHe*
ne
 Volume photoionization
 Photoelectrons from wall
He* Profiles - Experiment
NHe*
16 Torr Profiles - Simulation
Top: Experiment
Bottom: Simulation
(Behind wavefront)
10
Key Questions:
• Why are these
profiles different?
• What does this say
about FIW physics?
19
Energy Deposition in Wavefront
 Simulations  Strong radial E near wall
 Radial E exceeds axial E just behind FIW front
Radial E
Radial E-field
fills much of
the volume
Axial E
Radial E-field
drops rapidly
away from wall
1 Torr
Electric field exceeds
runaway e- threshold
(~210 Td in He)
16 Torr
20
Cross Sections & Path Lengths
• σiz peaks near 150 eV, σHe* near 25 eV
• Path lengths drastically diverge above 30 eV
Cross-sections
Ground Ionization, Metastable
Mean Free Paths vs. Pressure
Ionization & Metastable
Ionization
Metastable
Tube Diameter
1 Torr
4 Torr
16 Torr
21
1-D Fast Electron Model
 Simple 1-D model assumptions:
 Discretize radial position & electron energies @ each position
 Initial fast e- flux, radially inward, originating @ wall:
 Electrons lose energy via collisions:
 Elastic, Excitation from ground state, Ionization from ground state
 Flux Conservation:
 Gain & Loss Terms:
 Solve (iteratively) for
to find:
22
Model Results: Fixed Initial Energy
 Sample of results from band of 55-65
eV initial energy e- flux
 All curves normalized to 1 @ wall
 Competition of effects:
Varied Pressure: 55-65 eV e-
 Low Pressure: 1/r focusing of flux
 High Pressure: Attenuation & e- Cooling
 0.5 Torr:
 Se center-peaked
 SHe* also center-peaked
 2 Torr:
 Se more uniform
 SHe* center-peaked, but broad
 8 Torr:
 Se wall-peaked
 SHe* peaks off-axis, near wall
23
Model Results: Fixed Pressure
Varied Energy, 8 Torr
 Pressure fixed @ 8 Torr
 Solved for various energy
“bands” at wall, 10 eV wide
 Little divergence for 20 eV
electrons
 Higher energies:
 Short Ionization Pathlength
 Long Metastable Pathlength
 Divergence in Profiles
Spatial separation is influenced by:
• Neutral Pressure
• Fast e- Energy Distribution
24
1-D Electron Model Limitations
 Not included:




Stepwise ionization
Randomization of electron motion (2-D)
Electrons which pass through origin
Axial component of fast electron flux
 But still qualitatively captures…
 Center-to-wall transition in radial profiles
 Low pressure:
 Center-peaked e- density and metastables
 Metastables slightly broader, as seen in LCIF
 High pressure:
 Wall-peaked e- densities
 Broadened metastable profiles, closer to axis than e- profiles
 What does a more sophisticated model say?
25
Laser-Absorption Spectroscopy
- B. Yee & J. Foster
●
●
Parallel effort to study FIW processes – Ph.D. Thesis for B. Yee
Advantages
–
–
–
–
●
Non-perturbing
Excellent time resolution (~ns)
Absolute measurement
Simple calibration
Difficulties
–
–
–
Pathlength-integrated measurement
Optimizing detector response sensitivity
Electrical noise
The Goal: Use absolute metastable densities + plasma induced emission
to clarify e- energies (and E/N?) during FIW propagation.
Laser Absorption Setup
●
DFB laser swept (in wavelength)
across transition from He 23S
●
●
●
1083 nm (23S  23P)
Laser absorbed by metastables in
plasma – quantified with PDs
Absorption curve fit gives metastable:
–
–
–
Densities
Temperatures
Drifts
DFB:
FI:
ND:
AP:
PD:
Distributed Feedback Laser
Faraday Isolator
Neutral Density Filter
Aperture
Photodiode
27
Initial Measurements
Sample Location
During Density Buildup
Negligible Gas Heating
Reflected Pulses
(Long Decay)
Deposited
Energy
Interpretation of LAS Data
 Immediate temporal evolution of NHe* 
 Ground-Metastable excitation rates (from pulse duration & initial ne)
  Effective e- Temperature & E/N (if electrons are local)
  Some kind of info on e- energies in wavefront (even if nonlocal)
 Plasma induced emission (PIE) measured w/ monochromator
 more constraints on energy distribution
 Global Model/CRM under development (Yee & Foster)
1.
2.
3.
4.
5.
Energy balance to calculate e- energies from applied pulse
Collisional radiative model  PIE from energies & densities
Measure metastable & electron densities
Input densities into CRM, calculate emissions of various transitions
Compare to measured PIE  does it match what’s expected from
E/N? If not, why?  validity of E/N in FIW wavefronts.
29
First-order Example
During Pulse Buildup
If valid?
E/N
30
Summary
 2-D maps of electron and 23S metastable densities in a positive
polarity He FIW measured using LCIF/LIF
 Center-peaked ne at low pressure, wall-peaked at high pressure
 Metastable profiles shift from center-peaked to volume-filling
 2-D fluid simulations capture similar trends in ne
 Center-to-wall transition, trends in FIW velocity, ne profiles
 Predicts metastable distributions which track e- densities
 Radial E-fields yielding runaway e- may explain the difference




Strong E-field @ wall (~kTd) = source of fast “runaway” electrons
Dropoff in E at high pressure  e- from walls lose energy
High energy  ionization; Lower energy  metastable production
Energy decay along radius causes spatial separation in profiles
 Laser absorption measurements of He* + CRM may yield more
information on electron energetics in the FIW
31
Thank you!
 Questions?
 Comments?
This work was supported by the Department of Energy Office of
Fusion Energy Science Contract DE-SC0001939.
References:
1. S. M. Starikovskaia, N. B. Anikin, S. V. Pancheshnyi, D. V. Zatsepin, and A. Yu. Starikovskii.
Plasma Sources Sci. Tech., 10:344–355, 2001.
2. K. Takashima, I. V. Adamovich, Z. Xiong, M. J. Kushner, S. Starikovskaia, U. Czarnetzki, and D.
Luggenholscher. Phys. Plasmas, 18:083505, 2011
3. L. M. Vasilyak, S. V. Kostyuchenko, N. N. Kudryavtsev, and I. V. Filyugin. Phys. Uspekhi, 37:247269, 1994.
32
Key Questions:
What causes the transition in e- densities?
Can we explain this with a model?
Radial Position, mm
Radial Position, mm
Radial Position, mm
Radial Position, mm
Increasing Pressure
 Density maps @ fixed
rate & voltage, 1-16 Torr
 Peak densities on scale
of 1011 cm-3 for all
pressures
 Low P  center-peaked
 High P  wall-peaked
 Volume-filling, max. ne
at intermediate pressure
Radial Position, mm
Electron Densities vs. Pressure
1 Torr
x 10
3
10
5
11
2
0
1
-5
-10
20
40
60
80
100
120
0
140
Axial Position,
2 Torr mm
x 10
3
10
5
11
2
0
1
-5
-10
20
40
60
80
100
120
0
140
Axial Position,
4 Torr mm
x 10
3
10
5
11
2
0
1
-5
-10
20
40
60
80
100
120
0
140
Axial Position,
8 Torr mm
x 10
2
10
5
1.5
0
1
-5
11
0.5
-10
20
40
60
80
100
120
0
140
Axial16
Position,
Torr mm
x 10
10
10
5
0
5
-5
-10
20
40
60
80
100
120
0
140
Axial Position, mm
Wavefront Motion
33
10
 Helium 23S metastable
profiles, 1-16 Torr
 Relative densities from
LIF intensities
 Laser absorption
measurements for
absolute values (B. Yee)
Increasing Pressure
Metastable Densities vs. Pressure
 Similar trends, but less
drastic than ne
 Center-peaked to volumefilling / uniform
 Similar FIW decay lengths
Wavefront Motion
34
Electron Profiles vs. Pressure
1 Torr
8 Torr
4 Torr
t = 100 ns
16 Torr
35
Time Dependence After FIW
Electron Density
23S Metastable Density
• Shape of radial profile established by 40 ns
• Ionization in residual plasma conserves shape after initial formation
• Profile dictated by energy deposition in wavefront
36
Line Profiles
Absorption lines fit to Voigt profile w/ Doppler & Pressure Broadening
37