Exposure Optimization in Scanning Laser Lithography

Scanning laser and probe-based exposure offers an attractive alternative to standard lithographic methods for prototyping and low volume production. As the speed of nanopositioning systems increases, these methods will become increasingly competitive. Simple devices can already be exposed in less than a second and current research aims to create millions of features in a similar time frame. This article focuses on the problem of finding an exposure pattern which optimizes the geometrical fidelity of the developed features. The solution is based on a nonlinear programming approach that can be solved with a gradient based method. By changing the beam profile function, this method is applicable to all forms of serial lithography including e-beam, probe-based, and scanning laser. Current research includes adapting the algorithm to handle images with a massive number of features and/or ultra-high resolution. It is also necessary to consider uncertainty in the optical and photoresist models, for example, variations in film thickness and photoresist constants, among others. Although technical challenges still exist, the development of this technology will dramatically improve access to a low-cost, ultrahigh resolution lithographic process. We hope that this will stimulate the development of new fabrication processes and myriad new technologies and devices that rely on them.