Reactive ion etching of 193-nm resist candidates: current platforms and future requirements

Abstract

We have examined the reactive ion etch (RIE) resistance of two families of 193 nm photoresist candidates, poly(methacrylates) and vinyl-polymerized poly(cyclic olefins), in three RIE processes. Correlation of these measurements to polymer structure and composition using known methods (Ohnishi and Ring Parameter fits) was moderately successful in demonstrating global trends but proved generally inadequate for providing quantitative predictions. To address this shortcoming, we have developed a new empirical structural parameter which provides a much more precise model for predicting RIE rates within a given family of polymers. The model is applicable across polymer platforms, with two caveats: (1) The methacrylate and cyclic olefin families examined to date fall on essentially parallel, offset curves when examined with the new model, (2) The offset between polymer family curves is RIE tool- and process-dependent. While these caveats imply a setback to the idea of a truly `universal' model, they may in fact represent a powerful and unanticipated feature; the model appears to separate chemical RIE processes which affect individual functionalities within a polymer from predominantly polymer-family dependent processes such as global backbone degradation. In the course of conducting these studies, we have encountered several potential pitfalls in the measurement of etch rates. These illustrate the complex nature of plasma: resist interactions and highlight the careful experimental design and controls that are required if meaningful RIE rate comparisons between polymer and resist families are desired.