Kinetics of nanosecond discharges at high specific energy ...mipse.umich.edu/files/Starikovskaia_presentation.pdfCapillary nanosecond discharge Low voltage Electrode (not grounded)

Kinetics of nanosecond discharges at high

specific energy release

Laboratoire de Physique des Plasmas

Svetlana Starikovskaia

Laboratory of Plasma Physics, Ecole Polytechnique, Paris, France

Co-authors and collaborators for this work

LPP PostDocs@PhD students: Sergey Shcherbanev, Yifei Zhu,Nikita Lepikhin, Sergey Stepanyan, Andrey Klochko

LPP fellowship PostDocs@students: Ilya Kosarev, Brian Baron,Andrey Khomenko, James Williams, Georgy Pokrovskiy,

2

Kinetic calculations and theory: Dr. Nikolay Popov (MSU)

Andrey Khomenko, James Williams, Georgy Pokrovskiy,Mhedine Alicherif

PAC/PAI: Mohamed A. Boumehdi, Guillaume Vanhove,Pascale Desgroux (Lille University)

2D streamer (early) calculations: Dr. Vladimir Soloviev (MIPT)

Plan of the presentation

I. Introduction

II. Problem formulation

III. Experimental tools III. Experimental tools

IV. Results and discussion: -capillary nanosecond discharge at moderate pressure -surface dielectric barrier discharge (nSDBD) at 1 atm-surface dielectric barrier discharge at high pressure, filamentation in nanosecond single-shot discharge

V. Conclusion

3

Low temperature – high temperature plasma: where we are?

High T plasma

6

7

8

9

10

Z-pinch

Gas lasers

Sun corona

Nuclear synthesis

Sun core

[K])

4

Low T plasma

-2 0 2 4 6 8 10 12 14 16 18 20 22 240

1

2

3

4

5Gas lasers

Streamer

PLasma torchLightning, spark

Cathode spot

Ionosphere plasma

Intetstellar plasmaPLasma of metals

lg(T

e[K])

lg(N e[cm -3])

Motivation: plasma-assisted combustion

• Fast flows:

150 ms x 2.5 m/ms (M=8) L= 375 m

5

• To estimate the length:TGV Eurostar: TransManche Super Train LTGV=395 m (20 cars)

Need to have shorter ignition delay

Decrease of ignition delay time:moderate gas densities; uniform ns discharge

103

104

105

CH4, auto, 0.4-0.7 atm

CH , auto, 2 atm

Ig

nitio

n de

lay

time,

µµ µµ s

Auto Exp, C2H6 Auto Calc, C2H6 PAI Exp, C2H6 PAI Calc, C2H6, , , C3H8, , , C4H10, , , C5H12

6

0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85

100

101

102

10

CH4-C

5H

12,

PAI, 0.2-0.7 atm

C2H

6-C

5H

12,

auto, 0.2-0.7 atm

CH4, auto, 2 atm

Igni

tion

dela

y tim

e,

1000/T, K-1

N L Aleksandrov, S V Kindysheva, I N Kosarev, S M Starikovskaia, A Yu Starikovskii, Proc. of Combustion Institute, 32 (2009) 205-212

High pressure ignition of combustibe mixtures

- Ignition of lean mixtures- Multi-point ignition- Ignition at lower temperatures- Ignition of bio-fuels

7

- Ignition of bio-fuels

Different potential applications

Flow speed up to 100 m/s

85 min after plasma action

Pressures up to 15 bar

What is peculiar with nanosecond discharges? E-field range

1,0

O2 vibr

Ene

rgy

bran

chin

g N2 vibr

RF ns

9

0,1 1 100,0

0,5

O2 electr

N2 electr

O2, N

2 ion

elastic

O2, N

2 rot

Ene

rgy

bran

chin

g

E/N, 10 -16 V cm 2 N.L. Aleksandrov, E.E. Son, 1980, in: Plasma Chemistry, B.M.Smirnov, ed., Atomizdat Publ., Moscow, V.7, pp. 35-75

Main pathway for initiation [any] chemistry by ns discharges

100 ps- high electric fields

- excitation by electron impact

- dominant population of electronically

10

1 ns

100 ns

- dominant population of electronicallyexcited levels

- dissociation

- relaxation with the rest of the energygoing to heat

Aim of present work

To underline specific features of pulsednanosecond discharges under conditions of

- high specific energy density;

- high reduced electric fields.- high reduced electric fields.

To specify the most important changes inplasma chemical kinetics at the mentionedconditions

11

Experimental setup: discharge parameters and spectroscopy

12

DT - Discharge tube; BCS - Back Current Shunt; TG - Triggering generator; HVG - High Voltage Generator; CP - Capacitive Probe; ICCD - Intensified CCD Camera; OSC - Oscilloscope; MC - Monochromator

Typical discharges and discharge cells

FIW, 1-10 Torr, L=20cm

CD, 20 Torr, 8 cm

13

nSDBD, 8 bar 10cm/0.5 mm

nSDBD, 15 bar 10cm/0.5 mm (T<700 K)

Experimental tools

U(t),I(t): calibrated custom made backcurrent shunts;

E/N(t,x): calibrated custom made capacitivedetectors; emission spectroscopy;

14

detectors; emission spectroscopy;

v(t,x), morphology: ICCD (filtered) imaging

ne(t,x): electrical current measurements,Stark broadening

[O](t): actinometry, TALIF

Specific deposited energy, eV/mol

15

FIW,

10-3 eV/mol

1 atm SDBD

streamer,

0.1 eV/mol

Capillary

discharge,

>1 eV/mol

Scientific elegance: “diluted” chemistry in high fields at nanoseconds

103

FIW front structure

ne, cm -3

E/N, Td

1012

1013

16

0 5 10 15 20 25

102

[N2

+(B

2ΣΣΣΣ

+

u)], cm-3

[N2(C3ΠΠΠΠ

u)], cm -3

Time, ns

1010

1011

S.V.Pancheshnyi et al J. Phys. D: Appl. Phys, 1999, 32 2219

Example of application: N2(C3Πu)

quenching rate measurements

17S.V.Pancheshnyi et al Chem. Phys., 2000, 262, 349-357

Fast ionization waves (U=+10 kV in the cable, FWHM=30 ns)

Discharge in air at 1-10 Torr (N=1017 cm-3):

I(t): 60 A (3 A/cm2);

v(t,x): 50 mm/ns;

18

v(t,x): 50 mm/ns;

E/N(t,x): 1-5 kTd in the front and 100-300 Tdin the region of energy deposition after thefront;

ne(t,x): 1012 cm-3

Scale of specific deposited energy

19

FIW,

10-3 eV/mol

1 atm SDBD

streamer,

0.1 eV/mol

Capillary

discharge,

>1 eV/mol

Capillary nanosecond discharge

Low voltageElectrode

(not grounded)

Quartz tubeInner diameter:1.5 4 mmor

50-70 mm

HV electrode

50 mm

Flow in 10-50 sccm

(gas renewedevery 10 ms)

Gas flow out4-50 mbar

N :O mixture2 2

Grounded screen

Plasma

electrode

20U(t):

Electric field and current waveforms

500

600

700

E fi

eld,

Td 40

50

60

21

-10 0 10 20 30 40 50 60

0

100

200

300

400

E fi

eld,

Td

I, A

Time, ns

0

10

20

30

40

Specific energy and gas heating

3

4 First pulse Second pulse W1+W2 Third pulse W1+W2+W3

Dep

osite

d en

ergy

, eV

/mol

ecul

e

22

10 20 30 40 50

0

1

2

(b)

Dep

osite

d en

ergy

, eV

/mol

ecul

e

Pressure, mbar

Energy: a few eV/mol Temperature: a few kK/µs>1 eV/mol, >3000 K

0.1

1 Experimental Theoretical

PM

T s

igna

l, V

N2(C3Πu) emission decay in nitrogen

33 ns [1]

0 40 80 120 160

0.01

0.1

PM

T s

igna

l, V

Time, ns 23[1] Pancheshnyi, S. V., Starikovskaia, S. M., and Starikovskii, A. Y., Chemical Physics, 262 (2000)

0.1

1 Experimental Theoretical

PM

T s

igna

l, V

N2(C3Πu) emission decay in nitrogen

9 ns

33 ns [1]

0 40 80 120 160

0.01

0.1

PM

T s

igna

l, V

Time, ns 24[1] Pancheshnyi, S. V., Starikovskaia, S. M., and Starikovskii, A. Yu., Chemical Physics, 262 (2000)

9 ns

17 ns

N2(C3Πu) reverse decay vs pressure

0.08

0.10

0.12

0.91 eV/mol.

0.93 eV/mol.

0.90 eV/mol.

ns-1

1.5mm tube 1.5mm tube 1.5mm tube theoretical 4mm tube 4mm tube 4mm tube

Specific depositedenergy decreasing

25

0 10 20 30 40 50 600.02

0.04

0.06

0.08

0.58 eV/mol.

0.72 eV/mol.

1/ττ ττ,

ns

P, mbar

4mm tube 4mm tube

Analysis of the experimental results

Possible causes:

– Stimulated emission

– Variation of quenching rate with temperature

– Quenching by atoms

– Quenching by excited species (up to 10% of N2 are – Quenching by excited species (up to 10% of N2 are electronically excited)

– Quenching by electrons: ne=1015 cm-3

– The electron density at t>10 ns is sustained by reactions of associative ionization:

N2(A3Σ+

u) + N2(a’1Σ-u) � N4

+ + e

N2(a’1Σ-u) + N2(a’1Σ-

u) � N4+ + e

26

Calculated ne and excited species densities

1016

27 mbar

N (A)

N2(a)

N2(B)

Den

sity

, cm

-3

27● electron density slightly changes after discharge pulse

101 1021014

1015

N2(C)

N2(A)

Ne

Den

sity

, cm

Time, ns

Calculated and measured electron density in the afterglow in nitrogen

10

associative ionization

ne afterglow experimental

ne main pulse experimental

den

sity

, 10

14cm

-3

28

40 80 120 160 200 240

1

ionization

Ne

dens

ity, 1

0

Time, ns

no associative ionization

101 102

1015

1016

keC

=0 cm3/s

5x10-8

Experiment

keC

=10-7 cm3/s

N2(C

3 ΠΠ ΠΠu)

dens

ity, c

m-3

Time, ns

“Non-diluted” plasma chemistry: how this will influence the diagnostics?

N2(A3Σ+

u) + N2(a’1Σ-u) � Prod

N2(C3Πu) + e � Prod

A* + e � Prod?

A* + M* � Prod?

29

Actinometry in nanosecond capillary discharge in air, 20 Torr

1

7 ns

Inte

nsit

y, a

.u. 22 ns 8

30Actinometry: 5.3% Ar in air; OES of Ar(2p1), 750 nm // O(3p3P), 845 nm

0 10 20 30 40 50 60

0.1

7 ns

Inte

nsit

y, a

.u.

Time, ns

N2(C3)-theory

N2(C3)-experiment

0 10 20 30 40 50 600

2

4

6

O(3p3P) Ar(2p1)

Inte

nsit

y, a

.u.

Time, ns

?

Temporal behavior of Ar* and N2(C)

31

Temporal behavior of O*, Ar*/O* ratio and rates of O* production/decay

32

O-density: 3e16 cm-3; am expression for the discharge phase

33

O-TALIF: scheme of the levels

34

Niemi, K., Schulz-von der Gathen, V., and Döbele, H. F., Plasma Source Science and Technology, Vol. 14, No. 2, 2005, pp. 375-386

TALIF: O-atoms density in the afterglow

35

Long time trend: actinometry and TALIF

1016

1017

O(3P), TALIF: w/ Q(T) O(3P), TALIF: no Q(T)

2.

1.NO

N(4S)O(3P)

O a

tom

s, c

m-3

1E16

1E17

1E18

Den

sity

, cm

-3

e O

36

100 101 102 103 104

1014

1015

N(4S)

N2(B+C)/5

O( P)

O a

tom

s, c

m

Time, ns1 10

1E14

1E15

Den

sity

, cm

Time, ns

O2 N2 Ar N2(C3)

Capillary discharge (U=+10 kV in the cable, T=30 ns)

Discharge in air at 27 mbar (N=7x1017 cm-3):

I(t): 60 A (2.5 kA/cm2);v(t,x): 15 mm/ns;

37

v(t,x): 15 mm/ns;

E/N(t,x): 1-3 kTd in the front and 200-400 Tdin the region of energy deposition after thefront;ne(t,x): 1015 cm-3;[O](t): 50-100% dissociation of O2 at 2 µs


38

FIW,

10-3 eV/mol

1 atm SDBD

streamer,

0.1 eV/mol

Capillary

discharge,

>1 eV/mol

Atmospheric pressure nanosecond SDBD

39

2014 S.A.Stepanyan, V.R.Soloviev and S.M.Starikovskaia J.Phys.D.:Appl. Phys, 47, 485201 (13pp)

Propagation by ICCD imaging (ICCD gate is 1 ns, N2(C

3Πu) emission)

40

2003 First time: D.V. Roupassov, A.Yu.Starikovskii, MIPT(GE Project //PhD Thesis)

nSDBD modeling; difference between–U and +U

-U

41

+U

100 µµµµm thickness


Distribution of parameters near the surface streamer head

Q

Te

422016 N.Yu.Babaeva, D.V.Tereshonok and G.V.NaidisPlasma Sources Sci. Technol. 25 044008

E

ne

nSDBD ICCD imaging: 2+ system of N2

No diaphragm With diaphragm

0.5 ns gate, (7x7) mm per pixel

43

nSDBD development: 1 atm air, |U|=24 kV

-

44

+


-

45

+


-

46

+


-

47

+


-

48

+


-

49

+


-

50

+


x-t diagram: streamer head vs time

25

30

35 HV pulse shape calculations, + 24 kV calculations, -24 kV

Dis

tanc

e fr

om H

V e

lect

rode

, mm

+U, 25 mm span -U, 25 mm span

51

0 5 10 15 20 25 30 35 400

5

10

15

20

Dis

tanc

e fr

om H

V e

lect

rode

, mm

Time, ns

-U, 25 mm span +U, 50 mm span -U, 50 mm span -U, cylindrical, 63 mm

perimeter of the electrode -U, 80 mm span

A B

N2(C3Πu) emission 3 mm away from

HV electrode and current in the streamer

100

Inte

nsity

of N

2(C

) em

issi

on, a

.u.

Positive Negative

0.8

1.0

1.2

1.4

Cur

rent

, A/c

hann

el

U<0

HV

A B A B

52

0 5 10 15 20 25 30 35

1

10

Inte

nsity

of N

2(C

) em

issi

on, a

.u.

Distance, mm

Noise level

0 5 10 15 20 25 30 35-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

Cur

rent

, A/c

hann

el

Time, nsSegment |AB|

Streamer (U<0) stops after ≈10 ns. Emission intensity and electrical current on |AB| are ≈const

What we can estimate from the experimental data

) = 0

Ech = 2U/Lmax

0 5 10 15 20 25

0

5

10

15

20

25

30

35

Dis

tanc

e fr

om H

V e

lect

rode

, mm

Time, ns

Positive Negative

53

) = 0

0 10 20 30 40-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Cur

rent

, A/c

hann

el

Time, ns

0 5 10 15 20 25 30 35

1

10

100

Inte

nsity

of e

mis

sion

, a.u

.Distance, mm

Positive Negative

Noise level

Logics of data treatment

Electron density ne on the segment |AB|

Absolute value of N2(C) on the segment |AB|

Current and radius measurements I(t), r

E/N on the segment |AB|, E=U/Lmax

Emission intensity measurements

54

Temporal profile of N2(C) concentration

f(E/N)=Eµ/kexc=j(t)/kexcne

Temporal profile of E/NElectron density ne(t)

0 5 10 15 20 25 30 35

1

10

100

Inte

nsity

of e

mis

sion

, a.u

.

Distance, mm

Positive Negative

Noise level

Comparison with results of numerical calculations

150

200

250

300 Calculations Experiment

E/N

, Td

E/N

|AB|

Electric field Electron density

[2]

55

0 5 10 15 20 25 30

0

50

100

150

Time, ns

E/N

, Td

|AB| [2]

[1][1] Results of Babaeva et al.[2] Results of Soloviev et al.

|AB|

Lines: 1D kinetic modeling, N.A.Popov (MSU)

Streamer discharge (U=+10 kV in the cable, T=30 ns)

nSDBD in air at 1 atm:

I(t): 0.2 A/streamer (600 A/cm2);

56

I(t): 0.2 A/streamer (600 A/cm );

v(t,x): 2-5 mm/ns;

E/N(t,x): 100 Td after the front for U<0,<<100 Td for U>0;

ne(t,x): 1014 cm-3

Flow control: 2D modeling and experiments

57


High pressure filamentary SDBD?

58

HV

High pressure nSDBD: streamer-to-filament transition

Intensity (filament)=50 x Intensity (streamer)

59

-35 kV, 2 bar, air -46 kV, 4 bar, air

Analysis of intensities (streamer vs filamentary mode)

60

Radial distribution of the frontal discharge emission; P=3 bar, U=-46 kV

Optical emission in the filaments

ICCD images and radial intensity profiles of single filament: without filter and with (480±5) nm and (340±5) nm narrowband filte rs

0

20000

40000

60000

80000

no filter

N2(C3ΠΠΠΠu)CW spectrum

61

ICCD camera gate is 10 ns-120 -80 -40 0 40 80 120 160

500

1000

1500

2000

2500

500

1000

1500

2000

Distance, x(03BC)m

340 nm filter

Inte

nsity

, a.u

.

480 nm filter

Filamentation curves for nitrogen,air and N2/CH4 mixtures

38

40

42

44

N2/CH4 (1%)

Vol

tage

, kV

U<0

35

40

45

50

N2/CH4 (6%)

Vol

tage

, kV

U>0

62

2 4 6 8 10 12

28

30

32

34

36

N2/CH4 (1%)

N2

Vol

tage

, kV

Pressure, bar

Air

4 6 8 10 1210

15

20

25

30

35

AirN2

Vol

tage

, kV

Pressure, bar

Continuous spectra in filamentary discharge

63

Electron density in the filaments: H-atom emission at 656 nm

1

2

3

4

Hig

h vo

ltage

, a.u

.

20

25

30

35

FW

HM

, nm

1019

64

0 20 40 60

-4

-3

-2

-1

0

1

Hig

h vo

ltage

, a.u

.

Discharge time, ns

0

5

10

15

20

FW

HM

, nm

1017

1018

ne (

cm-3

)

Group of support: for high electron densities!

ne=5x1018 cm -3

ne=9x1018 cm -3

65

Van der Horst R., Verreycken T, Bruggeman P. et al J. Phys.D. (2012) 345201

Lo A., Cessou A, Lacour C. et al PSST (2017) 045012

ne=5x10 cm

D. Pai, GEC 2016

ne=1.5x1018 cm -3

Correlation in time: U-pulse, FWHM of O-line, intensity of CW emission, O* –atom density

4

5

6

Hig

h vo

ltage

, a.u

.

8

10

Line

777

nm

Lor

entz

FW

HM

, nm

1000

1500

Are

a lin

e 77

7 nm

by

hand

, a.u

.

80 90 100 110 120 130

Delay, ns

80

100

120

Con

tinuo

us s

pecr

um in

tens

ity a

t 777

nm

, a.u

.

66

0 10 20 30 40 50

0

1

2

3

Hig

h vo

ltage

, a.u

.

Discharge time, ns

0

2

4

6

Line

777

nm

Lor

entz

FW

HM

, nm

0

500

1000

Are

a lin

e 77

7 nm

by

hand

, a.u

.

0

20

40

60

80

Con

tinuo

us s

pecr

um in

tens

ity a

t 777

nm

, a.u

.

Continuous spectrum base level

Possible reasons of CW spectrum-1

Bremsstrahlung and recombination radiation

eionebr TNNQ ⋅⋅∝

1x10-24

1x10-23

Inte

nsity

of e

mis

sion

, a.u

.

bremsstrahlung recombination

67

e

ionerec

T

NNQ

⋅∝

Intensity of bremsstrahlung and recombination radiation is

a strong function of density of charged species

and weakly depends on the electron temperature

0 10 20 30 40 50

1x10-25Inte

nsity

of e

mis

sion

, a.u

.Mean electron energy, eV

Possible reasons of CW spectrum-2

SPS (337.1 nm) is a reference radiation

N2(C) = kC(E/N)·ne·N2/ νq

QC = N2(C)·FFK·A00

QC = 3·1022quantum/cm3/s.

68

Electron density from recombination emission:

220 102)( ⋅=

⋅⋅⋅= ω

ωω d

T

nnCQ

e

ioneCW η

At λ = 337 nm and ∆λ = 3 nm, dω = 5·1013 s-1

In the discharge Te~2-3 eV → ne~ 6·1018 cm-3

In the afterglow Te~0.5 eV → ne~ 4·1018 cm-3

quantum/cm3/s

QC(t=0 ns)/QCW(t=5 ns)~(1.5-2)

1

1022

1023

Ne = 4 1018 cm -3

Ne = 6 1018 cm -3

Q, q

uant

um /

cm3 / s

Te, eV

Continuous spectra in filamentary discharge: Planck emission so gas heating?

+U -U

69

T=7200KTrot=7200K

The most typical features of the filament

-the electron density is extremely high, 1018-1019 cm-3

at 6 bar, but the decay is slow, tens of ns;

- no molecular emission is observed from thefilament, only cw and atomic lines (no material of

70

filament, only cw and atomic lines (no material ofelectrodes, no energetic ion lines);-- observed cw spectrum is due to recombination andPlanck emission;

- there is strong (kK) and fast (units of ns) gasheating within the filament

Each feature needs additional analysis: molecular emission

-no molecular emission is observed from thefilament, only cw and atomic lines

No molecules, Quenching by

71

No molecules, 100% dissociation

Quenching by electrons?

ne=1019 cm-3,kq=10-7 cm3/s N2

++e=N+N,

kne2t>[M] during 1 ns

Each feature needs additional analysis: decay time for electrons

-the electron density is extremely high,1018-1019 cm-3 at 6 bar, and the decay is low, 10s of ns

Recombination decay,

1019 calc, T

e = 1 eV

cacl, Te = 2 eV

72

Thermal ionization?

decay, N++e+e

is too fast(parts of

nanoseconds)20 30 40 50

1017

1018

Ne, c

m-3

Time, ns

Each feature needs additional analysis: decay time = gas cooling for electrons?

1018

1019

15 ns

25 ns

, cm

-3

73

Thermal ionization?

6 8 10 12 14 161017

N2 : O

2 = 4 : 1

40 ns50 ns

Ne, c

m

Temperature, kKTrot=7200K

And finally, how this transition starts?

Ionization-heating

instability?

Increase of T by kinetic FGH

Decrease of gas density within a few µm distance

T ⇑

74

1 3p nsτ = −

Decrease of gas density within a few µm distance

N ⇓

Increase of E/N

E

N⇑

Increase of ionization

Conclusions

Specific deposited energy in nanosecond microplasmas can be ashigh as electronvolts per molecule. At high reduced electric fieldsand moderate gas pressures, “non-diluted chemistry” in thedischarge and early afterglow is important

75

N2(C3Πu) emission in P=1 atm surface streamer discharge,

electrical current and morphology provide important informationabout the plasma parameters

High pressure filamentary nSDBD is characterized by abrupt cwemission and by high electron density. Appearance and physicsof streamer-to-filamentary transition are studied

Acknowledgments

This work was partially supported by the following

projects and organizations:

–ANR ASPEN

–LIA KaPPA (France-Russia)

–LabEx Plas@Par

76

Thank you for your attention

77

Documents

Kinetics of nanosecond discharges at high specific energy ...mipse.umich.edu/files/Starikovskaia_presentation.pdfCapillary nanosecond discharge Low voltage Electrode (not grounded)