Sub-half-micron contact window design with 3D photolithography simulator

Using advanced simulation to aid microlithography development

“…If the topology of the surface changed, even greater problems arose. Fortunately, considerable advances involving the mathematical and numerical treatment of surface evolution has occurred since the start of this work, as discussed in [27], [30]- [40], and in the level-set work in [41] and [42].…”

“…Consequently, a number of parameters that are only indirectly related to the detailed physics have been explored by researchers [79]- [86] to determine how well they can be utilized to gain some of the predictability that exists in more fundamental simulation approaches [14], [17], [27], [30]- [40]. A large number of choices exist, including: 1) information in the optical image space, sampled from in and out of focus [81], [92], [85]; 2) information in the focus plane, but along the direction of the normal to the contour predicted from the threshold model [79], [80]; 3) as a "diffused " aerial image [79], [83]- [86], to attempt to account for baking and other "diffusion related" effects; 4) information such as the slope of the aerial image and the maximum peak intensity along a normal to the threshold [82]; or, even 5) information like the curvature of the shape predicted by the threshold model.…”

“…The extent to how well these models work, and for which resist conditions, will undoubtedly continue to be investigated as tests are made, and as new resist processing is developed. Clearly, none of these approaches can hope to reproduce the more detailed physical models conventionally used [14], [17], [27], [30]- [40], but that is not really the aim; rather, the intent is to find methods that work reasonably well for well-controlled processes, and to use these models, with reasonably rapid calibration methods, to find the necessary fitting parameters for the resist process in question.…”