``But what does it all mean, Basil?''

Modern semiconductor manufacturing uses plasma assisted processes for a number of critical fabrication steps. Understanding how the plasma interacts with the walls (wafers, materials targets, etc) is paramount. (Actually, this is true in any plasma device but I mostly concentrate on temperatures and densities appropriate for manufacturing.) My thesis looks at high frequency resonances and waves at the walls in low temperature plasmas.

By the way, my thesis won the David J. Sakrison Memorial Prize for outstanding thesis research in the Berkeley EECS department and the American Physical Society Outstanding Doctoral Thesis in Plasma Physics Award.

An excerpt from the thesis abstract:
Chapter 1, Chapter 2 and Chapter 3 conduct simulation studies of electron series resonance sustained discharges with comparisons to theory and experiment. These plasmas have many desirable characteristics (resistive V-I phase, frequency tunable density, low-temperature, low-pressure).
Surface wave plasmas are the natural extension to resonant plasmas and are promising for use in large-area plasma sources. Appropriate for large-area device modeling, an electromagnetic theory of surface wave propagation in a warm non-uniform plasma is developed and compared to previous theoretical work (Chapter 4 and Chapter 5). In Chapter 6, several PIC simulations are conducted to validate the electromagnetic theory. In Chapter 7, numerical techniques suitable for computing the wave dispersion and impedance in a large-area low-temperature plasma are developed.
Utilizing much of the research conducted here, Chapter 8 demonstrates a novel application of surface waves. Through a resonant wave-particle interaction (``Landau resonant heating''), the electron velocity distribution function is controllably modified by a standing surface wave excited with a distributed periodic electrode. Simulation results indicate this Landau resonant heating can be used to dramatically enhance important reactions in low-temperature low-pressure plasmas including electron-impact excitation and electron-impact ionization.
In conducting this research, an algorithm to effectively eliminate cache thrashing in a particle-in-cell simulation was developed, resulting in a 40 to 70 percent performance gain on typical workstations. The algorithm is described in Chapter 9. Also several implicit methods for solving the Maxwell equations were developed with superior noise and stability properties compared to the standard explicit leap-frog finite-difference-time-domain algorithm (Chapter 10).
The listings below are not exhaustive. It only covers journal articles for which I have electronic copies. Several conference papers were also given for this research. These other publications are cited in the thesis.

Note: Both PDF and gzipped PostScript versions of these documents are available. The PDF versions occasionally have some font conversion problems in the figures so if you want a typeset quality document for printing, use the PostScript version. In the true sprit of UNIX, if you can't read gzipped PostScript files, go buy yourself a real computer.

Everything

Electron Series Resonant Discharge Investigation

All these articles were submitted to Plasma Sources Science and Technology in December 2000.

Plasma Surface Wave Investigation

All these articles were submitted to Physics of Plasmas in February 2001.

Landau Resonant Heating Investigation

This article has been accepted for publication as an invited paper in the Physics of Plasmas May 2002 Special Issue.

Related Algorithm Development

These articles were submitted to Journal of Computational Physics in January 2001. If you have any questions or wish to comment on any of the above documents, feel free to contact me at:

kevin dot j dot bowers at ieee dot org.

(Do the obvious to get my correct email address; I prefer correspondence from people---not spam-bots.)