Achieving sub-10-nm resolution using scanning electron beam lithography

Abstract

Achieving the highest possible resolution using scanning-electron-beam lithography (SEBL) has become an increasingly urgent problem in recent years, as advances in various nanotechnology applications have driven demand for feature sizes well into the sub-10-nm domain. While SEBL has the highest resolution of nearly any conventional patterning technique available, reliably defining features at these length scales has been a challenge, as well as an interesting scientific problem. In this work I have investigated, both theoretically and experimentally, many of the factors that limit SEBL resolution and attempted to understand and minimize their influence on the process. This includes resist development, where we have thoroughly characterized the temperature dependence of poly(methylmethacrylate) (PMMA) resist contrast and used the results to create transferable patterns smaller than nearly any published results to date with this resist chemistry. We have also examined the process of electron-beam exposure and attempted to characterize the various factors that affect the way energy is distributed in the resist by the beam, using theoretical arguments, Monte Carlo simulations, and experimental data. We have used the results of these investigations to create some of the smallest structures reported to date, using hydrogen silsesquioxane (HSQ) resist. Finally, we have applied some of the previously-gained knowledge to the design of a unique bilayer process for patterning high-resolution metal structures using evaporation and liftoff, while simultaneously developing a broadly-useful new model for the kinetics of resist development.