ABSTRACTThin-film devices pose unique challenges to device simulators. Relaxation approximations to the Boltzmann Transport Equation may be vitiated as the width and length of the thin-film components approach the mean free path lengths of electrons and phonons. Concomitant with these reduced lengths, surface and other interface scattering mechanisms exert greater influences on the transport processes. Consequently, it may not be effective to use transport models that depend upon analytic or other pre-ordained representations of electron, phonon, and photon distribution functions. Finally, thin-film simulations must account for tunneling processes, local defects, nanoscale dopant gradients, nanoscale roughness, and nanoscale variations in local geometries. Consequently, it is desirable to develop a nanoscale simulation technique that (1) can arbitrarily vary material properties in real space while (2) tracking reciprocal space scattering for arbitrary and volatile distributions of electrons, phonons, and photons. The Discrete State Simulation (DSS) forms the backbone of such a simulation. Written with the run-time flexibility of object-oriented code, the Discrete State Simulation (DSS) is a coupled cellular automata (CA) simulator that builds upon the objects and rules of quantum mechanics. The DSS represents global non-equilibrium processes as patterns that emerge through an ensemble of scattering events that are localized at vibronic nodes. The run-time flexibility of the DSS enables dynamic rule construction - computational algorithms that evolve in response to the conditions that are being simulated. The reported DSS effort simulated nanoscale transport processes at dopant-defined junctions by coupling electron-phonon and electron-photon scattering mechanisms. Using electronic band structures, phonon band structures, and deformation potentials analogous to silicon <100> material properties, the DSS generated data depicted the dynamic band bending induced by femtosecond variations in applied electric fields.
Read full abstract