DOE Plasma Science Center 3rd Annual Meeting

Controlling Vibrationally Excited Nitrogen and Overall Plasma Chemistry with Surface Micro-discharge in Ambient Air

Yukinori Sakiyama and David B. Graves Department of Chemical and Biomolecular Engineering University of California at Berkeley

University of California, Berkeley

Outline

1. Surface micro-discharge (SMD) 2. Distribution of RONS at low-power • RONS in discharge layer and afterglow • Comparison with FTIR 3. Mode transition in afterglow • UV O3 measurement • Fitting model with N2 vibrational mode • Correlation with bactericidal effect 4. Concluding remarks

Plasma Science Center Predictive Control of Plasma Kinetics

SMD: discharge characteristics G. Morfill et al., New J. Phys. 11 (2009) 115019, T. Shimizu et al., New J. Phys. 13 (2011) 023026

SMD = surface micro-discharge (surface dielectric barrier discharge in ambient air at room temperature)

Cu (powered)

HV

image of emission

dielectric

~ 10 cm SS mesh (grounded)

RNS ROS

treated surface

Frequency: Voltage: Power: Distance to sample: Exposure time:

1-10 kHz 1-10 kVpp 2 0.01-1 W/cm 1-10 mm 1-1000 s

Plasma Science Center Predictive Control of Plasma Kinetics

SMD: anti-microbial effect Various microbes on agar plate G. Morfill ., New J. Phys. 11 (2009) 115019

untreated

E. coli on pig skin M. Pavlovich,, Plasma. Process. Polm. (submitted)

30 s

gram positive/negative bacteria, spores, viruses, etc… Plasma Science Center Predictive Control of Plasma Kinetics

SMD: multi-scale phenomena 1 ns

pulse excitation electron impact reactions

1 s

1 ms

1s

1000 s

charge transfer, recombination

Ozonizer Pollution control

neutral reactions applied voltage period gas diffusion

SMD for biomaterial treatment

exposure time

Plasma Science Center Predictive Control of Plasma Kinetics

Modeling: domain and equations • •

Humid air at 1 atm: 79% N2, 20% O2, 1% H2O (30% relative humidity) Gas temperature: 300 K

dp = 0.1 mm

discharge layer e

ROS, RNS

n p t

  k j  nr , j 

 pg 

afterglow

dg = 10 mm

ng t

dp

j

ion

ROS, RNS

 pg

D ( n p  ng ) (d p  d g ) / 2

  k j  nr , j  j

 pg dg

treated surface Plasma Science Center Predictive Control of Plasma Kinetics

Modeling: simulation procedure (solver: MATLAB) pls(=1 ns)

pulse-like electric field

E  E0 exp{(t /  pls ) 2 / 2} discharge layer (e, ion, neutral)

afterglow (neutral)

100 s for single cycle rep=100 s

Cycle-averaged reaction rates discharge layer (neutral)

afterglow (neutral)

for gas=1 s

Plasma Science Center Predictive Control of Plasma Kinetics

Modeling: humid air plasma chemistry •

53 species: electrons, 16 positive ions, 10 negative ions, and 26 neutrals

•

624 reactions 23 electron impact excitation/ionization 84 electron recombination/attachment/detachment 169 charge transfer and ion conversion 231 ion-ion recombination 116 neutral-neutral reactions

References • H.Matzing, Adv. Chem. Phys. 80 (1991) 315 • I. A. Kossyi, Plasma Sources Sci. Technol. 1 (1992) 207 • R. Atkinson, Atmos. Chem. Phys., 4 (2004) 1461 • M.Capitelli, “Plasma kinetics in atmospheric gases” (Springer, Berlin 2000) etc., etc., etc…. Plasma Science Center Predictive Control of Plasma Kinetics

Modeling: discharge layer at low power Power density: 0.05 W/cm2 Peak density: ~1019 m−3 Negative ions at 1000 s

Positive ions at 1000 s 8

8

total

-3

density [10 m ]

6

18

+

4

NxOy

+

Nx

2 0

+

Ox

+ OxHy

10

-9

10

-8

6 -

NOx

18

-3

density [10 m ]

total

4

electrons

2

-

Ox

OxHy -7

10 10 time [s]

-6

10

-5

10

-4

0 10

-9

10

-8

-7

10 10 time [s]

-6

10

-5

10

-4

Plasma Science Center Predictive Control of Plasma Kinetics

Modeling: neutrals still in transient after 1000 s Power density: 0.05 W/cm2 Peak density: > 1019 m−3

10

24

10

23

density

-3

[m ]

Cycle-averaged neutral density at 1000 s

10

22

10

21

10

20

10

19

Ox NxOy HxNyOz OxHy

1

2

3

4

5 6 7 89

10

2

3

time

4

[s]

5 6 7 89

100

2

3

4

5 6 7 89

1000

Plasma Science Center Predictive Control of Plasma Kinetics

10

21

10

19

10

17

N2O

Simulation at 100 [s] H2

H2O2

O3

HNO3 NO3

HNO2 NO2

N2O5

10

4

10

2

10

0

10

-2

FTIR measurement (qualitative comparison) 0.12

IR beam

200 scans for 60-120 [s]

transmittance [-]

density

-3

10

23

density [ppm]

[m ]

Modeling: distributions of neutrals

O3

0.08 0.04

N2O

0.00 400035003000 2500

HNO3 HNO3 N2O5 N2O5

2000 1500 wave number

HNO3

1000 [cm ] -1

500

Plasma Science Center Predictive Control of Plasma Kinetics

Outline

1. Surface micro-discharge (SMD) 2. Distribution of RONS at low-power • RONS in discharge layer and afterglow • Comparison with FTIR 3. Mode transition in afterglow • UV O3 measurement • Fitting model with N2 vibrational mode • Correlation with bactericidal effect 4. Concluding remarks

Plasma Science Center Predictive Control of Plasma Kinetics

Gas-phase ozone: modulation by power density UV 254 nm

• power density: 0.03-3 W/cm2 • gas gap: ~10 mm • temporal resolution: 1 s

concentration [ppm]

5000 4000 3000 2000 1000

Intermediate power (0.25 W/cm2)

Low power: “ozone-mode” (0.025 W/cm2)

ozone-mode in SMD (our study) High mode” Highpower: power:“nitrogen “nitrogenoxides oxide mode” (2.64 (2.64W/cm W/cm22))

0

U. Kogelschatz, Ozone Sci. 367 0 20 et al.,40 60 Eng. 10 80(1988)100

120

time [s] Plasma Science Center Predictive Control of Plasma Kinetics

Mode transition: N2 vibrational state Ref: M. Capitelli, “Plasma kinetics in atmospheric gases” (Springer, Berlin 2000)

e

O3

O

N2 N2(v)

NO2

NO

v > 12 Energy transfer efficiency 1.0

10 10

-2

10

-4

0.2

10

-6

0.0

10

-8

F (v >12)

Qv / Qttl

0.8

0

0.6 0.4

0

2 4 6 8 mean electron energy [eV]

10

Cumulative distribution function (v >12)

2000

4000 Tv

6000 [K]

8000

10000

Plasma Science Center Predictive Control of Plasma Kinetics

Mode transition: simplified fitted model discharge layer

mixed layer

afterglow

R1: O + O 2  M  O3  M R2: N 2 (v)  O  NO  N R3: O3  NO  NO 2  O 2

diffusion loss

R4: O  NO + M  NO 2  M

Governing equations dnO3 nO3  k1nM nO nO2  k3nNO nO3   dif dt dnNO n  k2 nN 2 (v) nO  k3nNO nO3  k4 nO nNO nM  NO dt  dif nN 2 (v)  nN 2 Fv 12

2 unknown variables • nO3 and nNO 3 fitting parameters • nO, Tvmax, v

12 v  nN 2 exp( ) kTv

Tv  Tg  Tvmax {1  exp(t /  v )} Plasma Science Center Predictive Control of Plasma Kinetics

Mode transition: O3 model and NxOy mode fitted parameters: nO = 8×1017 m−3, Tvmax = 5000 K, v = 2 s concentration [ppm]

8000

O3

800

v

600

6000

Tv

4000 400

Tv

nO

NO

2000

200 0

0

5

10

ozone mode

15 time [s]

20

25

0 30

Vibrational temperature [K]

1000

nitrogen oxides mode

Plasma Science Center Predictive Control of Plasma Kinetics

[-]

3

1500

2

1000

1

500

0 0.001

0

[ppm]

2000

Intermediate Low power 0.1 High power (< (>power 1.0 Wcm Wcm22)) Wcm2) O33:: low increase ••(0.1-1.0 O O3: quenched L-R: increase •• L-R: low • L-R: high Avg O3 for 30s

4

Log reduction

Gas-phase ozone: bactericidal activity on agar plate

0.01 0.1 1 2 power density [W/cm ]

• •

ozone is not responsible for inactivation? Nonlinear reaction on lipid membrane?

Plasma Science Center Predictive Control of Plasma Kinetics

Concluding Remarks 1. We developed multi-scale model of SMD. Our simulation results at low power (0.05 W/cm2) shows good agreement with our FTIR measurement. 2. We presented one example of modulating plasma chemistry in SMD. The modulation is achieved by controlling distribution functions of electrons and neutrals through pulsing of electric field.

Plasma Science Center Predictive Control of Plasma Kinetics

Acknowledgements Prof. G. Morfill (Max-Planck Institute) Dr. T. Shimizu (Max-Planck Institute) Prof. D. Clark (UC Berkeley) M. Pavlovich (Ph.D. candidate, UC Berkeley) H.-W. Chang (Ph.D. candidate, National Taiwan University)

Plasma Science Center Predictive Control of Plasma Kinetics