Transcript Document
THE LONG TERM EFFECTS OF RANDOM DBD STREAMERS ON THIN LIQUID LAYERS OVER TISSUES* Wei Tiana) and Mark J. Kushnerb) University of Michigan, Ann Arbor, MI USA 48109 a)Department of Nuclear Engineering and Radiological Science, [email protected] b)Department of Electrical Engineering and Computer Science, [email protected] October 2014 * Work supported by US Department of Energy Office of Fusion Energy Science and the National Science Foundation. MODELING OF PLASMA INTERACTION WITH TISSUE Plasma treatment of tissue involves Plasma kinetics, dynamics and gas phase plasma chemistry Plasma-liquid interactions and liquid phase chemistry Biological functionality Plasma treatment of tissue usually operates from seconds to minutes, consisting of 100’s to 10,000’s pulses. To investigate the plasma treatment of tissue, modelling of multiple pulses followed by long-term afterglow is necessary. MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. MODELING IN PLASMA MEDICINE: DBD TREATMENT Powered Electrode Dielectric Air Plasma Evaporation Ionization Wave (ns - ms) hn E-Field Plasma-liquid Interaction (ms - s) Tissue Multi-scale, multiIntratissue Transport phenomena, multi(mins – days, mm - cm) disciplinary Many desired outcomes. Membrane MIPSE_2014 Kinetics Cellular response AGENDA In this poster, we present results from a computational investigation of DBDs interacting with liquid covering tissue for up to 100 pulses. Radical production with time (or pulses), terminal species and transit species. Stationary scheme: locally repeated streamers “Memory-effect” scheme: streamers are repeated at selected locations Random scheme: streamers are repeated randomly MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. MODELING PLATFORM: nonPDPSIM Poisson’s equation: ( q j N j s ) j Transport of charged and neutral species: Surface Charge: s N j t j S j q j j S j t j material Electron Temperature (transport coefficient obtained from Boltzmann’s equation) ne 5 j E ne i K i N i e Te Te t 2 i Radiation transport and photoionization: S m (ri ) N m (ri ) 3 mk Ak N k r j 'Gk r j ' , ri d r j ' k MIPSE_2014 ri exp lk N l rj 'drj ' l r ' j G rj ' , ri 2 4 rj 'ri University of Michigan Institute for Plasma Science & Engr. TREATMENT OF LIQUID Liquid plasma is treated identically to gas as a partially ionized substance. Higher density Specified susceptibility/atom to provide known permittivity Surface tension is addressed by specifying species able to pass through vapor/liquid interface. Diffusion into water is limited by Henry’s law equilibrium at the surface layer. Liquid evaporates into gas with source given by its vapor pressure. MIPSE_2014 Evaporation Ions Dissolved gases Solvated Liquid University of Michigan Institute for Plasma Science & Engr. WATER REACTION MECHANISM Plasma Photons, M* e H2O+ (H2O)e M- R Water H2 M+ H,OH,O,O3, O2,NO,NO2 H RH H OH OH R RH O2- OH- H,OH,O,O3, O2,NO,NO2 O,O- H3O+ O3,O3H 3 O+ H 2 O2 R H O2 RH N2O3,N2O4 R RH HO2 H3O+ NO2- NO3- = reaction with H2O MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. GEOMETRY DBD, 1.5 mm gap 200 mm thick liquid layer with “plasma water”. Humid air (N2/O2/H2O=79.0/20.9/0.1) at 1 atm with H2O evaporating from the liquid surface. Dissolved O2aq in liquid – 8 ppm MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. DBD DISCHARGE DYNAMICS DBDs in contact with water act like traditional DBDs with a lossy floating electrode underneath. After the discharge channel establishes, the ne, Te and Se are concentrated on the top. -18 kV, 5 ns pulse, 200 mm water, O2aq 8 ppm in water MIN MIPSE_2014 MAX Log scale University of Michigan Institute for Plasma Science & Engr. SCHEME FOR MULTI-PULSES 10 ms Water evaporation 10 ns Plasma Time The source functions are recorded and used for following pulses. 10 ms 10 ms 2nd pulse 3rd pulse ••• ••• 1 s afterglow 100th pulse Time Before the first pulse, water is evaporated into the gap for 10 ms. The plasma is computed for the 1st pulse. The source functions at the end of plasma pulse are recorded and used as initial conditions for later pulses. Poisson’s equation is not solved and quasi-neutrality is assumed after the pulse. MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. THREE MULTI-PULSES SCHEMES Stationary: repeated streamers “Memory-effect”: 5 selected locations Dielectric Evaporation Repeated Streamer Plasma Air E-Field hn E-Field Random Streamers Plasma Air Random Locations E-Field MIPSE_2014 Plasma Air Selected hn Locations Different multi-pulses schemes Repeated single streamers “Memory-effect” streamers Random streamers hn University of Michigan Institute for Plasma Science & Engr. ELECTRON DENSITY (1015 cm-3) “Memory-Effect” Stationary Random Electron density is shown for each multi-pulses scheme. In stationary scheme, plasma is concentrated at the center. In “memory-effect” scheme, 5 striking locations are shown. In random scheme, plasma appears randomly. MIN MIPSE_2014 MAX Log scale University of Michigan Institute for Plasma Science & Engr. TERMINAL SPECIES IN LIQUID “Memory-effect” Stationary Random Density: the average density in liquid layer MIPSE_2014 O3aq accumulates during 100 pulses, up to 1016 cm-3. In stationary scheme, NOaq and H2O2aq both increase with time. In “memory-effect” scheme, NOaq and H2O2aq still increase but with lower densities. In random scheme, NOaq is almost consumed up and H2O2aq starts to fall after 20 pulses. TRANSIT SPECIES IN LIQUID Density: the average density in liquid layer MIPSE_2014 Different from terminal species, OHaq , HO2aq and Haq can hardly accumulate during multi-pulses. OHaq reacts with NO2aq to form HNO3aq and HOONOaq. Haq forms HO2aq, which then reacts with NOaq to form HNO3aq and HOONOaq. University of Michigan Institute for Plasma Science & Engr. HYDROGEN PEROXIDE EVOLUTION In stationary scheme, H2O2aq narrows to the width of streamer after 5 pulses. In “memory-effect” scheme, H2O2aq shows two significant high density region. MIN MAX Log scale MIPSE_2014 In random scheme, H2O2aq is more uniform. University of Michigan Institute for Plasma Science & Engr. NITROGEN OXIDE EVOLUTION In stationary scheme, NOaq is consumed at the center but diffuse deep aside, since NOaq reacts with H2O2aq. MIN MAX Log scale MIPSE_2014 In “memory-effect” scheme, NOaq and H2O2aq are mixed at selected locations leaving NOaq only remaining at selected locations. In random scheme, H2O2aq is “wellstirred” with NOaq, which is no longer able to diffuse to tissues. FLUENCES TO UNDERLYING TISSUE In stationary scheme, the fluences profiles show non-uniform distributions. In “memory-effect” scheme, fluences profiles peak at selected locations. In random scheme, fluences profiles become uniform. MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. FLUENCES TO UNDERLYING TISSUE The fluences profiles of charged species basically have the similar characteristics as that of neutral species. In “memory-effect” and random schemes, “well-stirred” radicals produce more ONOO-aq. MIPSE_2014 University of Michigan Institute for Plasma Science & Engr. CONCLUDING REMARKS Results from modeling of multiple DBD pulses incident onto thin water layers followed by long-term afterglow are discussed. The accumulation of reactivity, such as O3aq and H2O2aq, depend significantly on the multi-pulses schemes. Locally stationary streamers result in non-uniform fluences. NOaq can reach tissue by avoiding reacting with H2O2aq. “Memory-effect” scheme concentrates the reactivity at selected locations. Randomly placed streamers result in more uniform fluences. Random streamers provide a well-stirred environment of precursors in gas phase, resulting in NOaq being consumed by H2O2aq before reaching the tissue. MIPSE_2014 University of Michigan Institute for Plasma Science & Engr.