Multi-objective optimization of performance metrics of all-metallic vacuum field effect transistors

Abstract

Vacuum field effect transistors (VacFETs) are a promising alternative to traditional semiconductor-based transistors, offering a superior frequency response, radiation hardness, and high-temperature operation. In this work, we illustrate the use of a multi-objective optimization (MOO) technique to systematically explore the design space of all-metallic VacFETs. The approach is based on a MATLAB implementation of an elitist genetic algorithm to identify the Pareto front of optimal configurations resulting from trade-offs between key VacFET performance metrics, such as a large ON–OFF field emission (FE) current ratio and a low leakage current. The analysis is illustrated for the case of an all-metallic VacFET with either one or two in-plane side gates based on a two-dimensional analysis of the effects of electrostatics. A comparison of the FE characteristics demonstrates that the optimized dual-gate VacFET achieves an approximately three orders of magnitude improvement in both the ON/OFF current ratio and the ON/Leakage current ratio compared to the similarly optimized single-gate configuration. Based on these results, a first-order small signal high-frequency model is developed, which predicts a unity current gain cutoff frequency of several hundred GHz for the optimized dual-gate VacFET. The versatile MOO technique proposed here can be readily extended to three-dimensional simulations incorporating various gate configurations (top, bottom, or lateral gates) and material combinations with increasing number of constraints and an implementation of their corresponding cost functions. The insights gained from MOO techniques offer valuable guidelines for designing next-generation vacuum nanoscale devices with substantially improved figures of merit.