The variability in the printing of small contacts with electron-beam lithography in a fast high resolution resist was characterized using automated SEM image analysis of an identical array of contact holes. The goal of this study was to evaluate the sources and severity of pattern noise in an e-beam system. A matrix of 391 contacts, 17x23 (dictated by the dimensions of the SEM display 768x1024), was printed at 100KeV on the LBNL nano-writer in KRS-XE2 photoresist. The doses ranged from 28uC/cm∧2 to 851uC/cm∧2 with 8nm, 16nm, 24nm, 32nm, and 40nm contact holes. Printed contacts were counted by image processing of SEM images using NIH's ImageJ program. The amount of pattern noise was found to be 14X larger than the noise that would be predicted by the traditional Poisson shot noise of 5500 electrons per contact. Surprisingly, the pattern noise was independent of PEB time and resist thickness. The main source of noise was found to be associated with the surface of the resist, most likely outgassing of acid in the e-beam vacuum chamber. The contact hole experiment provides a practical method for quantifying random effects in evaluating resists, processes, and treatments.
© (2006) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
