Modification of the development parameter for a chemically amplified resist simulator

Abstract

It is necessary to have more appropriate resist parameters in order for a lithography simulator to predict real photoresist profiles. These process parameters are usually obtained by flood exposure experiments without pattern masks. However, real processes are performed with pattern masks. Since the intensity on the wafer is different with and without a pattern, the development parameters must be modified in order to predict real processes. Especially, the development parameters, one example of the process parameters, are crucial to mimic real processes. It has been reported that the development parameters of a photoresist with or without underlying patterns are different. In this paper, we modified the flood exposure development parameters of a 248-nm chemically amplified resist (CAR) to get patterned development parameters and compared them with the simulation results. First, we obtained the development parameters by using a flood exposure experiment and applied them to our lithography simulator LUV. The simulated resist profiles were then compared to SEM microphotographs. Second, we modified the development parameters for the simulated resist profile to match the SEM photographs. We also determined the relationship between the changes of the parameters and the pattern profile. We could see the effect of the modification in different line widths and sidewall angle.