Applying an interferometric exposure model to analyze the influences of process parameters on the linewidth

We utilize a modified interferometric exposure model, enhanced with the Beer-Lambert law, to study how some process parameters influence the structural dimensions within the whole exposure area. An experimental apparatus is built to verify the accuracy of this model. The simulation results indicate that when the incident angle is larger than 15 degrees, the effect of the beam deformation cannot be neglected. One cannot readily obtain periodic structures with the same dimensions during static exposure because of the Gaussian distribution of the light intensity. The theoretical results match the experimental ones quite well. The variation of Dill's parameter A has a greater influence on the transmittance and the linewidth when A is decreasing. If a poor contrast fringe is exposed in the photoresist, it will not only cause a greater nonuniformity of the structural dimensions but also a decreased aspect ratio in the structure after the development process.