Automated Optimization of Light Dose Distribution for Moving-Mask Lithography

In the present paper the automated optimization of the exposure dose for UV moving-mask lithography is presented. Moving-mask lithography is a promising new technique for micro patterning thick positive photo resists. The design task is to design the mask and the mask motion such that the requested micro pattern is achieved as good as possible. As an intermediate step towards a computational engineering tool for moving mask lithography, the present paper puts focus on the automated optimization of the required light dose distribution. The method relies on a recent model which links the spatial distribution of the dissolution rate to the light dose distribution at the top surface of the photo resist. This model is combined with a search-free and, thus, cost efficient quality measure. For the latter, adjoint sensitivities can be calculated using a backward integration scheme. The fact that adjoint sensitivities can be retrieved easily, implies that the light dose can be obtained with an high spatial resolution. The proposed method will be demonstrated on the basis of a number of test problems.