Simulation of Z-Shaped Graphene Geometric Diodes Using Particle-in-Cell Monte Carlo Method in the Quasi-Ballistic Regime

Geometric diodes are planar conductors patterned asymmetrically to provide electrical asymmetry, and they have exhibited high-frequency rectification in infrared rectennas. These devices function by ballistic or quasi-ballistic transport in which the transport characteristics are sensitive to the device geometry. Common methods for predicting device performance rely on the assumption of totally ballistic transport and neglect the effects of electron momentum relaxation. We present a particle-in-cell Monte Carlo simulation method that allows the prediction of the current-voltage characteristics of geometric diodes operating quasi-ballistically, with the mean-free-path length shorter than the critical device dimensions. With this simulation method, we analyze a new diode geometry made from graphene that shows an improvement in rectification capability over previous geometries. We find that the current rectification capability of a given geometry is optimized for a specific mean-free-path length, such that arbitrarily large mean-free-path lengths are not desirable. These results present a new avenue for understanding geometric effects in the quasi-ballistic regime and show that the relationship between device dimensions and the carrier mean-free-path length can be adjusted to optimize device performance.