8:30 AM-9:15 AM
New Hampshire Ballroom
Chair: Linda R. Petzold, University of California, Santa Barbara, USA
The need to understand the relationship between processing and device performance for chemical vapor deposition vapor (CVD) processes is driving the coupling of molecular level simulations with traditional macroscopic transport phenomena descriptions. The development of predictive, efficient models that bridge across multiple length and time scales raises challenges in terms of simulation strategies, numerical algorithms, and experimental validation. These are exemplified through studies linking quantum chemistry, kinetic Monte Carlo (MC), and macroscopic finite element simulations. Process examples are drawn from of organometallic CVD of metals and compound semiconductors. Experimental observations and quantum chemistry predictions of elementary surface reactions are incorporated into MC simulations to provide new understanding of microstructure evolution. By flux balances and level set methods, the results of these computations are subsequently incorporated into self-consistent feature and reactor scale models. Comparisons with experimental data are given at each length scale.
Klavs F. Jensen