Investigation on resist development rate model for synchrotron radiation X-ray lithography

Since the synchrotron radiation X-ray lithography (SRXRL) was put forward, it hascaught many people’s attention day after day. It has much benefit, such as high-structuralresolution, large process window, high throughput. It is generally thought a very goodlithography technique when dimensions shrink to 0.25μm and below. Because of the im-

Since the synchrotron radiation X-ray lithography (SRXRL) was put forward, it has caught many people's attention day after day. It has much benefit, such as high-structural resolution, large process window, high throughput. It is generally thought a very good lithography technique when dimensions shrink to 0.25 mu m and below([1]). Because of the importance of SRXRL in the future, in 1990 the lithographers in China established the first SRXRL station in 3B1A beamline which is located in Beijing Synchrotron Radiation Facility (BSRF), and they finished the first SRXRL experiment successfully in June, 1990. In the SRXRL experiments, it is very important to select the product of exposure time and exposure current accurately (thin product is expressed by XK below), because this selection will influence the resist development rate directly, and thus can influence the quality of the image transferring later. In this note, we use the nonlinear regression analysis to establish several resist development rate models for SRXRL; all these models can pass hypothesis testing. Using these models to control XX in exposures, good results can be obtained.