Modeling X-ray proximity lithography

Abstract

Advanced semiconductor circuits, such as DRAMs, are based on very complex fabrication processes. Because of the cost and complexity involved, it is rapidly becoming impossible to adopt a “trial-and-error” approach in the development stage of a new process. Fortunately, the advances in computer power spurred by the new semiconductor devices have made it possible to compute the response of complex systems in a reasonable time on workstations. Thus, the study of a virtual representation of the process (that is, a model) can represent a solution to the high cost of process development—of course, after verification of the model accuracy through controlled experiments. A correct physical interpretation of the process under study is necessary in order to implement a model that is both accurate and extendible. This is particularly true for new approaches, such as those involved in X-ray lithography. We have studied the process of image formation in X-ray lithography and have implemented several models to predict the intensity distribution at the wafer plane. The models can be applied to the definition of an optimal exposure system that will provide the maximum exposure latitude, and to the study of new types of X-ray masks.