State-of-the-art Boltzmann solvers for modeling and simulation of GaN HEMT technology

March 3, 2021


The increased demand for power, efficiency and linearity in next-generation T/R MMIC technology presents new challenges for designers and process engineers. A critical component of the GaN design cycle is modeling and simulation. Accurate physics-based models are paramount for closing the feedback loop between design and process. This talk will focus on modeling and simulation of GaN HEMT technology using Boltzmann solvers. After discussing the requirements for accurate simulations of GaN HEMTs, I will present a comprehensive review of seminal works in this area. This literature review will feature semi-classical deterministic solvers (drift-diffusion, hydrodynamics and energy transport methods), semi-classical stochastic solvers (ensemble Monte Carlo, cellular Monte Carlo, etc.), and quantum mechanical solvers (non-equilibrium Green’s function methods). I will thoroughly examine the strengths and weaknesses of each method. Finally, I will discuss AFRL’s semi-classical deterministic Boltzmann solver called Fermi kinetics transport and will show several simulation results compared to measurements.


Dr. Nicholas C. Miller is a research electronics engineer at the Air Force Research Laboratory Sensors Directorate near Dayton, OH. Before starting this position in 2017, he was a Ph.D. student at Michigan State University. He received his B.S., M.S., and Ph.D. in electrical and computer engineering from Michigan State University in 2013, 2015, and 2017, respectively.Nicholas’ research interests span a broad range of fundamental, applied, and computational physics, applied mathematics, electrical engineering, linear and nonlinear RF characterization, and linear and nonlinear RF modeling and simulation. He is actively pioneering a physics-based electronic device modeling paradigm at AFRL to advance highly accurate first principles semiconductor modeling and simulations, discrete transistor simulations with a full-wave EM electronic device solver, and physics-based compact modeling solutions in tandem in academic and industrial modeling experts.