«by Jun-Chieh Wang A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Electrical Engineering) ...»
MODELING STUDIES OF ATMOSPHERIC PRESSURE MICROPLASMAS:
PLASMA DYNAMICS, SURFACE INTERACTION AND APPLICATIONS
A dissertation submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
in the University of Michigan
Professor Mark J. Kushner, Chair
Professor John E. Foster Professor Brian E. Gilchrist Professor Yogesh B. Gianchandani Professor Euisik Yoon
DEDICATIONTo those who always stand behind me, Thanks for your continued support and encouragement.
To those who stick by me through tough times, I cannot accomplish any of these on my own.
ACKNOWLEDGEMENTSFive years has gone so fast. It feels like I just came here as a newbie yesterday, felt excited about everything I saw and everyone I met, and all of a sudden, I am standing here bragging about my research and begging for a PhD degree. It seems crazy that one thing just leads to another like it was planned way before I was born. In fact, none of these can be done without a great deal of assistance and support from all of my family, professors, collaborators and friends.
In some particular order, I would like to thank my parents. It’s impossible to thank them adequately for everything they have done and sacrificed for me from raising me in a stable household to unconditionally loving me.
A great deal of credit is due to my dissertation committee: Professor John E. Foster, Professor Yogesh B. Gianchandani, Professor Euisik Yoon, and Professor Brian Gilchrist for serving on my committee and all the insightful comments they provide me throughout the process;
this work will not be possible without their help.
To my current and former labmates: Juline Shoeb, Mingmei Wang, Michael Logue, Seth Norberg, Sang-Heon Song, Pen Tian, Wei Tian, Yiting Zhang; your companionship is invaluable.
Special thanks go to Dr. Zhongmin Xiong and Dr. Natalia Babaeva for all the guidance and useful conversation we had; thanks to Yan Yang for serving as a good example for me; and Julia for helping me through all the processes and paper work. I would also like to acknowledge my collaborators at HP Lab: Daihua Zhang, Henryk Birecki, Michael Lee, Napoleon Leoni, Omer
Gila, Seongsik Chang, and Tom Anthony; and graduate students at Prof. Gianchandani’s group:
iii Christine Eun and Xin Luo, who have contributed greatly to the development, fabrication, testing and validation of the devices that we are interested. I appreciate this great opportunity to work with them on such challenging and interesting topics.
I would like to extend my gratitude to my teachers at the National Cheng Kung University:
Prof. Kuan-Ren Chen, Prof. Sunny Wing-Yee Tam, Prof. Ker-Chung Shaing, Prof. C. Z. Frank Cheng and Prof. Jang-Yu Hsu. Your training and lectures not only broadened my horizons, but also equipped me with the necessary knowledge and skills to pursue my research career.
I also wish to thank all my wonderful friends for being always right behind me. It was at times a long, difficult road, but I am glad it was long and difficult because if I hadn’t gone through it to get there, the memories and lessons might not be so clear. I will cherish all the great moments we shared: stupid drinking game at parties, BBQs at Gallup Park, studying at library, every Christmas Eve and sleepy Sunday afternoon-I will carry that with me.
I have saved the best for last: I would like to express my deepest gratitude to my research supervisor, Prof. Mark Kushner who has supported and encouraged me for all these years. Mark is always actively involved in the work of all his students. Time after time, his easy grasp of physics at its most fundamental level helped us in the struggle for understanding. As a successful and world-leading scientist, Mark is still the most diligent and enthusiastic researcher than everyone I know. Day and night, weekday and weekend, he is always passionate, hardworking and eager about making a positive contribution to this society. Just when I thought that comfort and luxury were the chief requirements of life, his passion for science and education reminds me that all we need to make us really happy is something to be enthusiastic about. I am fortunate to have him as my mentor, and I could not ask for a better supervisor and role-model.
LIST OF FIGURES
1.1 Plasmas: The Fourth State of Matter
1.2 Plasmas: The Definition
1.3 Plasmas: 99% of the Entire Visible Universe
1.4 Townsend Breakdown
1.5 Spark Breakdown
1.6 Dielectric Barrier Discharge (DBDs)
1.7 Microdischarges and Electrophotography
1.8 Microdischarge-based Pressure Sensor
1.9 Issues to Be Discussed
DESCRIPTION OF THE MODEL
2.2 Geometry, Mesh Generation, and Discretization
2.3 Interpolation of E/N
2.4 Plasma Dynamics
2.5 Transport Coefficients, Rate Coefficients, and Electron Temperature
2.6 Radiation Transport
2.7 Secondary Electrons and Field Emission
2.8 Monte Carlo Simulation
2.9 Parallel Implementation of EMCS
ELECTRON CURRENT EXTRACTION FROM RADIO FREQUENCYEXCITED MICRO-DIELECTRIC BARRIER DISCHARGES
3.2 Description of the Model and Reaction Mechanism
3.3 Plasma Dynamics and Current Extraction from the mDBD
3.4 Dielectric Charging Characteristics and Ionization Processes
3.5 Total Charge Collection
3.6 Concluding Remarks
CHARACTERISTIC OF A RF MICRO-DIELECTRIC BARRIERDISCHARGE ARRAY
4.2 Description of the Model and Reaction Mechanism
4.3 mDBD Plasma Properties
4.4 Small Arrays of mDBDs
4.5 Concluding Remarks
CHARGING OF MOVING SURFACES BY CORONA DISCHARGESSUSTAINED IN AIR
5.2 Description of the Model and Reaction Mechanism
5.3 Corona Properties and Charging of a Stationary Surface
5.4 Charging of a Moving Surface
5.5 Concluding Remarks
ATMOSPHERIC PRESSURE MICRODISCHARGES PRODUCED BYCONDUCTIVE CHARGE ROLLER
6.2 Description of the Model and Reaction Mechanism
6.3 Characteristics of Microplasmas and Charging of a Stationary PC Surface............ 163
6.4 Charging of a Moving Surface
6.5 Concluding Remarks
SIMULATION OF MICRODISCHARGE-BASED PRESSURESENSORS………
7.2 Description of the Model and Reaction Mechanism
7.3 Plasma Dynamics in the Microdischarge Pressure Sensor
7.4 Periodic IW and Current Collection
7.5 Differential Current and External Pressure
7.6 Concluding Remarks
SUMMARY AND FUTURE WORK
8.1 Summary and Future Work
APPENDIX: AUTHOR’S BIOGRAPHY
Figure 1.1 Plasma can be characterized by electron density and temperature.
Electron Debye length λD (constant along the dashed lines) and the number of charged particles in a Debye cube ND (constant along solid lines) are also indicated in the figure...... 18 Figure 1.2 Illustration of avalanche and Townsend breakdown.[12-13]
Figure 1.3 The charge distribution and electric field in the gap (a) before and (b) after the avalanche reaches the anode.
E0, E’, and E represent the applied electric field, the field produced by space charges, and the total field (E = E0 + E’), respectively.[6,10Figure 1.4 Schematics of a (a) cathode-directed (or positive) streamer and (b) anode-directed (or negative) streamer at t1 and t2 (t2 t1). The anode-directed (or negative) streamer occurs in a longer gap.[6,10-11]
Figure 1.5 A time evolution of DBDs between a pair of parallel electrodes biased with AC voltage.
The DBD prevents the formation of an arc.
Figure 1.6 Elements of electrophotography (EP).
(a) Six steps of EP printing. (b) Surface charging by a charge roller (CR) and a corona discharge.
Figure 1.7 Schematic of ionography.
(a) A typical printing head. (a) Printing process of iconography.…………………………………….
Figure 2.1 Block diagram of nonPDPSIM.
Figure 2.2 Model geometry and unstructured mesh used for the simulation of microdischargebased pressure sensor.
The refinement zones are highlighted by red rectangles... 50 Figure 2.3 Control volume (CV) mesh and actual mesh. (a) The vertex-centered control volume is constructed by identifying the intersections of the perpendicular bisectors between a node and it nearest neighbors; the “cell corners” (A, B, C, D, E, or F) are defined as the intersections of the perpendicular bisectors. Actual mesh consists of vertices, faces, and cells. Nodes (no. 1–7) and solid lines which connect the nodes are output from the mesh generator. (b) The CV is centered around the vertex.
Figure 2.5 A schematic of the Newton-Raphson iterative method.
Figure 2.6 Schematic of radiation transport.
Seeding of electrons far away from the avalanche is produced by photons from excited species in the ionization. Both photoionization and photoemission (ϕS) from the surface result from photons emitted by species j.55 Figure 2.7 An electron Monte Carlo simulation is used to follow trajectories of secondary electrons from surfaces. The structured Cartesian mesh (CM) is overlaid onto the unstructured mesh (UM). The CM overlays only the portion of the UM in which beam electron transport is expected to be important. The resolution of the CM is chosen to be fine enough to capture the small scale feature of the UM.................. 56 Figure 2.8 A schematic of original and optimized particle trajectory algorithms in the EMCS.
In the original algorithm, particles are (a) first launched and from node 1. After all particles from node 1 are tracked and removed from EMCS domain, new particles are then released from (b) node 2 and then (c) node 3. The process repeats until particles are launched and recorded from all emission nodes. In the optimized algorithm (d), particles are simultaneously released from all emission nodes........ 57 Figure 2.9 The speedup and efficiency for (a) Np = 400; and (b) Np = 40,000 from each emission node.
Figure 3.1 Schematic of the cylindrical symmetric mDBD device.
(a) Entire device and full computational domain. (b) Enlargement of mDBD cavity and location of sites A, B, C and D that are used to provide surface properties and ionization characteristics. 80 Time evolution of electron density (log scale, cm-3) and E/N (at the center of contour Figure 3.2 labels, Td) in the mDBD cavity at different phases of the rf driving voltage of 1.4 kV during a 40 ns (25 MHz) cycle. The cycle begins with 0 V on the buried rf electrode.
The top electrode is biased with +2 kV. At high electron density, the electric field is shielded and E/N is reduced. …………………
Electron density in the plume (log scale, cm-3) in the gap at different phases of the rf Figure 3.3 driving voltage of 1.4 kV during a 40 ns (25 MHz) cycle.
Figure 3.4 Electron temperature (eV), electron impact ionization sources from bulk electrons (Se) and ionization source by sheath accelerated secondary electrons (Ssec) in the mDBD cavity during a 40 ns (25 MHz) cycle for Vrf = 1.
4 kV and extraction voltage of 2 kV. The ionization sources are plotted on a log scale.
The surface charge density σS (dash-dotted, cm-3), surface voltage on the dielectric Figure 3.7 (solid, V) and voltage on the buried rf electrodes (dashed, V) are shown for frequencies of a) 25, b) 5 and c) 2.5 MHz at site A (see Fig. 3.1) for Vrf = 1.4 kV and extraction voltage of 2 kV. Oscillations in charging of the dielectric appear at low frequencies.
Figure 3.8 mDBD characteristic are shown at site B as a function of time (see Fig.
3.1) for Vrf =
1.4 kV and extraction voltage of 2 kV for a) 25 MHz and b) 2.5 MHz: rf voltage (dashed, V), local potential (dash-dotted, V) and E/N (solid, Td), electron density (dash-dotted, cm-3) and electron ionizations source Se and Ssec (solid and dotted, cms ).
Figure 3.9 Current collection on the top electrode.
a) Experimentally observed triple current pulsed obtained at 2.5 MHz and b) simulation results in a similar sandwich mDBD device for equivalent biasing
Figure 3.10 Charge density at site D and current collection on the top electrode at 25 MHz.
a) Electron (solid, cm-3) and positive ion density (dash, cm-3) at site D for top bias voltage Vtop = 2, 1.5 and 1 kV. b) Electron current collected by the top biased electrode for Vtop = 2, 1.5 and 1 kV. A pulsed modulated dc current is collected at Vtop = 2 kV.
Figure 3.11 Charge collection as a function of time for 25 MHz to 2.