Inverse Modeling and Synthesis of Ion Implementation Process

摘要

Process synthesis is a top-down design methodology and can effectively reduce the process design time. This paper focuses on the synthesis of one of the key process steps: Ion implementation. First neural networks are used to build a basic inverse model for ion implantation, which is based on the implanted profiles with one type of substrate, one implantation, and using the Gaussian or single Pearson model in simulating procedure. Then the synthetic model is expanded to deal with more complex cases such as profiles generated by implanting into multilayer substrates, by using the dual-Pearson model in simulating procedure when ion type is B or BF2. In addition, synthesis of profiles generated by two implantations is discussed.