"Point-to-point nanosecond pulse "diffuse filament" discharges
for studies of energy transfer and nonequilibrium chemical
reaction mechanisms in molecular plasmas
D. Burnette, S. Bowman , K. Frederickson, B. Goldberg, S. Lanier, A. Montello,
M. Nishihara, A. Roettgen, Z. Yin, I.V. Adamovich, and W.R. Lempert
Objective / Outline
Main thrust: isolating molecular energy transfer processes and reaction pathways in
nonequilibrium, high-pressure, highly transient molecular plasmas
• Previous results in plane-to-plane nsec pulse discharges
• Point-to-point nsec pulse discharges to achieve high specific energy loadings
• Vibrational level populations and temperature measurements by psec CARS:
results for nitrogen and air
• CARS and spontaneous Raman spectroscopy results analysis, comparison with
kinetic modeling: do we understand kinetics mechanisms involved?
• Overview of absolute species concentrations measurements in reacting air plasmas:
NO LIF, O and N atom TALIF
• Path to measuring both reactant and product species in vibrationally enhanced
plasma chemical reactions
• Electric field measurements of in nsec pulse discharges by 4-wave mixing
• Work in progress: development of electron density measurements by Thomson
scattering
Previous results: repetitive nanosecond pulse discharge
in plane-to-plane geometry
(6.3 cm x 2.2 xm x 1 cm, ~25 kV, ~10 nsec pulses, pulse rep rate 10 kHz)
Air, T0=300 K, P=50 torr, Pulse #10
Air
Pulse #1
Pulse #10
Pulse #100
T0=300 K
P=60 torr
T0=300 K
P=120 torr
T0=500 K
P=200 torr
Side view (0.5 nsec gate)
End view (50 nsec gate)
Preheating enhances diffusion and thermal conductivity, greatly improves plasma uniformity
Instability onset is not controlled by reduced electric field, E/N
Kinetic modeling: coupled master equation /
Boltzmann equation model of nonequilibrium air plasma
dnv (t )
 ( El . imp ) v  (VT ) v  (VV ) v  (VE ) v  (V  Chem ) v
dt
RHS terms represent vibrational quantum state change by the following processes:
El. Imp.: inelastic electron impact processes by free electrons
VT: vibration-to-translation/rotation energy relaxation
VV: vibration-to-vibration energy exchange
VE: electronic-vibration energy transfer during collisional quenching
V-Chem: vibrational – chemistry coupling for vibrationally enhanced reactions such as
N2(v) + O → NO + N,
O2(v) + N → NO + O
•
Rotational and translational modes are in equilibrium at a gas kinetic temperature
•
Single vibrational quantum change processes dominate at low temperatures involved
•
Significant body of theory and experimental validation data for the rates used
•
Master equation coupled to Boltzmann equation for electron energy distribution, species
concentrations equations, and quasi-1-D compressible flow equation
•
Nonequilibrium air plasma chemistry, excited electronic states kinetics are included
•
0-D, with correction for diffusion
•
E/N, ne waveforms predicted by separate nsec pulse, plane-to-plane discharge model
Excitation of air by nsec pulse discharge burst:
Psec CARS measurements, master equation modeling
• Dry air, P=100 torr, T0=300 K, burst of 100-150 pulses,
repetition rate 10 kHz
• Trot, Tv(N2) measurements by psec CARS
• ~40% of coupled energy is loaded into N2 vibrational mode
• Model predicts N2(X,v) populations, electron impact /
vibrationally stimulated reaction products
• Vibrational relaxation mainly by O atoms, ozone
• Specific energy loading fairly low, ~1 meV/molecule/pulse
• Temperature rise ~ 1 K/pulse, Tv(N2) remains fairly low
New Test Bed: Diffuse Filament Nsec Pulse Discharge
between two bare metal spherical electrodes
Use of small (a few mm diameter), bare spherical electrodes increases power loading
(up to ~1 eV/molecule/pulse at P=100 Torr, coupled pulse energy up to ~10-20 mJ)
AND creates plasma large enough to be easily probed by CARS, spontaneous
Raman spectroscopy, and two-photon absorption LIF (TALIF)
4 mm
2 mm
10 mm
Air, single-pulse nsec pulse discharge:
100 ns gate, during the pulse (left) and
1 μsec after the pulse (right)
Air, single-pulse nsec pulse discharge in point-to-point
geometry: schlieren images and waveforms
Compression waves formed by “rapid” heating, on subacoustic time scale
From known initial temperature & pressure, voltage,
current, and filament diameter
→
Reduced electric field (E/N) and electron density (ne) for
kinetic modeling
Psec BOX-CARS using broadband dye mixture:
good spatial and time resolution, access to multiple vibrational levels
Ekspla Nd:YAG laser
- ~150 psec pulses, 125 mJ per pulse max @ 532 nm
Modeless Psec Dye Laser
-Broadband ~592-610 nm FWHM, ~7-10% conversion
-Broadband Pyrromethene Dye Mixture
Folded Box-CARS CARS signal beam
Interrogation
volume
(*S. Tedder, et al, 2011)
Spectral Resolution ~0.4 cm-1
95% of signal generated over ~0.5 mm
95% region
Data
Gaussian
Fit
Integrated Signal Intensity [au]
3
2.5
2
1.5
1
0.5
0
-0.4
-0.2
0
0.2
Distance [mm]
0.4
Typical psec CARS Spectra, 100 torr N2
(Normalized to v=0, corrected for dye laser spectral profile)
100 laser “shot” averaged spectra
vs. time after rising edge of current pulse
Vibrational level populations inference:
least squares fitting to Voigt line shape
v=0
↓
t = 200 nsec
v=3
↓
v=6
v=9
↓
↓
t = 200 nsec
(zoomed)
v=0-9 are detected
Direct evidence of additional N2 vibrational excitation
after discharge pulse
t = 200 nsec
t = 100 μsec
Square root of integrated band CARS signal
proportional to difference in vibrational level populations, nv –nv+1
Spatial resolution: “1st level” Tv(N2) measurements,
100 torr N2 at a lower pulse energy (5 mJ/pulse)
Tv 01 
v
lnn0 n1 
2 mm
10 mm
N2, single-pulse nsec pulse
discharge in point-to-point
geometry, 100 ns gate
N2 “first level” vibrational
temperature distribution 100 μs
after the pulse
Rotational Temperature Measurements
Data
Tfit=300K
25
Tfit
 =302 K
20
150
Counts
Sqrt(Int.) [au]
200
100
15
10
50
0
5
2320
2325
2330
-1
Raman Shift [cm ]
100-shot accumulation spectrum in
“cold” 100 torr air, with Sandia
CARSFIT best fit synthetic spectrum.
0
280
290
300
Tfit [K]
310
320
Histogram plot from measuring
and fitting 80 such spectra.
95% confidence interval ~ ± 9 K.
CARS results summary:
time-resolved N2 vibrational populations in nitrogen
Significant vibrational excitation after
the pulse (~100 nsec long), followed by
eventual relaxation
Energy appears to come to N2(v=1-8)
from an “internal storage”, not from
electron impact during the discharge
Increase in both “first level” N2
vibrational temperature, Tv01, and total
number of quanta per molecule, Q
Tv 01 
v
lnn0 n1 
9
Q   vfv
v 0
CARS results summary:
time-resolved N2 vibrational populations in air
Significant vibrational excitation after
the pulse (~100 nsec long), followed by
eventual relaxation
Energy appears to come to N2(v=1-8)
from an “internal storage”, not from
electron impact during the discharge
Increase in both “first level” N2
vibrational temperature, Tv01, and total
number of quanta per molecule, Q
Tv 01 
v
lnn0 n1 
9
Q   vfv
v 0
Air, experiment vs. model:
number of vibrational quanta per molecule
Average number of vibrational
quanta per molecule (Nquanta):
9
N quanta   vfv
v 0
• Significant increase of number of vib.
quanta per molecule ~1-10 μs after the pulse
(by ~60%)
• At variance with the model, which predicts
Nquanta=const after the pulse (V-V exchange
conserves quanta)
Air, experiment vs. model:
N2(X,v) vibrational level populations
The VDF evolution can be divided into 3 phases:
(1) Initial appearance and growth of all vibrational levels observed (Δt ~ 100 nsec – 1 μsec)
(2) Steady growth of vibrational levels detected (Δt ~ 1 μsec – 100 μsec)
(3) Vibrational energy decay: V-T relaxation (by O atoms) and diffusion (Δt ~ 100 μsec – 10 msec)
As time evolves, v=0, 1 level populations are well predicted by model; higher level populations
are significantly underestimated. Tv01(N2) is not a good metric.
N2, experiment vs. model:
number of vibrational quanta per molecule
Average number of vibrational
quanta per molecule (Nquanta):
9
N quanta   vfv
v 0
• Significant increase of number of vib.
quanta per molecule ~1-10 μs after the pulse
(by more than a factor of 2)
• At variance with the model, which predicts
Nquanta=const after the pulse (V-V exchange
conserves quanta)
N2, experiment vs. model:
N2(X,v) vibrational level populations
N2, experiment vs. model: Tv01(N2) and T
Tv 01 
v
lnn0 n1 
Tv01(N2) rise is primarily due to N2–N2 V-V exchange during relaxation: v=0, w → v=1, w-1
CARS results summary: time-resolved
rotational/ translational temperature in nitrogen vs. air
• Both in N2 and air, model overpredicts “rapid” heating ,
likely N2(A,B,C,a) + M → N2(X,v) + M (E-V processes)
• In air, also model underpredicts “slow” heating (absent in N2),
likely V-T relaxation by O: N2(X,v) + O → N2(X,v-1) + O
and O atom recombination: O + O + M → O2 + M
Where is the energy coming from?
Effect of possible electronic-to-vibrational energy coupling in N2
30% energy into N2(X,v) during N2*(A,B,C,a) quenching, e.g.
N 2 (C )  N 2 ( X )  N 2 ( B)  N 2 ( X , v)
N 2 ( A)  N 2 ( A)  N 2 ( B, C )  N 2 ( X , v)
k ( v)  k (T )1  exp( )exp v 
 vib 
  (v ) k ( v )
v
k (T )
α – adjustable parameter controlling energy into vibrational mode,
 vib
Better agreement of N2 vibrational and rotational temperature with the data
Where is the energy coming from?
Effect of possible electronic-to-vibrational energy coupling in N2
N2(v=1-8) rise at t ~ 0.1-1 μsec overpredicted, but better agreement at quasi steady state
Spontaneous Raman spectroscopy: consistent with
psec CARS results (nitrogen, N2(v=0-12), P=100 torr)
Signal collection
region (2.75 mm)
5 μs delay
Integrated band Raman signal proportional to
vibrational level population, nv
More vibrational levels detected
Effect of vibrational quanta rise after discharge
pulse observed again
Spontaneous Raman results in N2 vs. “baseline” kinetic
model predictions
Again, N2(v=2-5) rise at t ~ 1-10 μsec is not reproduced by the model
Model clearly missing electron impact excitation processes of N2(v>8)
Can electron impact excitation processes, N2(v) + e → N2(w) + e (v≥0, w>8) , be the key?
Spontaneous Raman results in N2 vs. modified model predictions
(30% of energy defect to N2(X,v) during E-V transfer)
Better agreement with data at t ~ 1 μsec – 1 msec
Previous work: NO LIF
Previous and current work: calibrated N and O TALIF
Air, 60 torr, plane-to-plane, ~0.3 meV/molecule
Mole
fractions
Mole Mole
fractions
fractions
1.0E-4
1.0E-4
1.0E-41.0E-4
Air, 40 torr, plane-to-plane, ~6 meV/molecule (21 pulses)
O, NO
O
O
O
NO
O
NO
ONO
O
1.0E-5
1.0E-5
1.0E-5
1.0E-5
NO
1.0E-6
1.0E-6
N
NNO
O3
ON3
N2(A)
NO23(A)
O2(b)
1.0E-6
1.0E-6
O
NO
NO2
(A)
ON22(b)
(b)
NO (X,v)
N2(X,v) 22
- (X,v)
eN
2
1.0E-7
1.0E-7
1.0E-7
1.0E-7
1E-6 1E-5 1E-4 1E-3 1E-2 1E-1 1E+0
NO2
Time, seconds
1E-6
1E-6 1E-5
1E-5 1E-4
1E-4 1E-3
1E-3 1E-2
1E-2 1E-1
1E-1 1E+0
1E+0
N2, 124 torr, point-to-point, ~60 meV/molecule
[NO] prediction
including reaction
Time,
Time, seconds
seconds
N2(X,v≥12) + O(3P)  NO + N
Psec CARS (spontaneous Raman), NO LIF, and
calibrated TALIF diagnostics:
Reactants, products, and temperature can be
measured for the same experimental conditions,
state-specific reactant rates can be inferred.
N
Sub-nsec resolution electric field measurement
in hydrogen by CARS-like 4-wave mixing
E-Field Results (kV/cm)
10
Hydrogen, P=100 torr, L=10 mm gap
Line: high voltage probe
8
Symbols: CARS
6
4
2
-0.2
0
0.2
seconds
0.4
0.6
Field in the range of the 1-10 kV/cm has been measured (averaging over 128 discharge pulses)
Thomson Scattering for electron density measurements
Triple Grating Spectrometer Schematic*,**
Mask
↓
* Patterned after Y. Noguchi, et al., Jpn. J. Appl. Phys 40, 2001
** Acknowledgement: U. Czarnetzki, Ruhr-University Bochum
Summary
• Time-resolved, spatially resolved T, Tv(N2), and N2 (X,v=0-12) population
measurements (psec CARS, spontaneous Raman) in high energy loading nsec pulse
discharges in air and nitrogen
• Results suggest coupling between electronic and vibrational mode energies in N2
• Time-resolved, spatially resolved measurements of absolute species concentrations
in reacting air plasmas: NO (LIF), O and N atoms (TALIF)
• Results demonstrate feasibility of measuring both reactant and product species in
vibrationally enhanced plasma chemical reactions, N2(X,v≥12) + O(3P)  NO + N,
inference of state-specific reaction rates
• Electric field measurements (4-wave mixing / CARS) in nsec pulse discharges in H2
• Good progress
measurements
on
Thomson
scattering
diagnostics
for electron
density
• Comprehensive set of optical diagnostics for characterization of nsec pulse
discharges in molecular gases