Comparison of writing strategies subject to resist heating

During electron beam (e-beam) exposure, the temperature of the resist varies locally, causing critical dimension distortion. An experimental comparative study of resist heating at 50 kV for a number of writing strategies was made. Exposure, materials, and processing were equal; the same e-beam column was used, but beam deflection was modified so that raster scanning could be compared to vector scan beam (VSB) exposure at high throughput. These strategies were compared for single-pass and four-pass exposures. Simulations of temperature rise were done using the TEMPTATION (Temperature simulation) software tool. The maximum temperature rise was 38 degrees C for 614-degrees m wide raster exposure, while over 220 degrees C for VSB exposure. Good agreement of simulation and experiment was found. A 16% effective dose change due to resist heating was found at full coverage, single-pass exposure, while in VSB, this value was over 200%. These results indicate that raster writing has an advantage over VSB exposures with regard to resist heating.