OPC verification on cell level using fully rigorous mask topography simulation

Starting with the 45nm node, the minimum feature size on the mask has reached sub-wavelength dimension. In this regime the electromagnetic field induced in the mask is significantly impacted by the mask topography. These so called mask topography effects play an important role in the image formation process and need to be compensated for in the optical proximity correction (OPC) model. Looking ahead to the 32nm process node, mask topography effects will become even more pronounced. So, including these effects into the OPC model has become a must for advanced process nodes. Modern OPC engines start to apply electromagnetic field (EMF) compensation techniques to take these effects into account. Of course, due to the severe run time constrains for OPC models most EMF aware OPC models need to rely on approximate methods. A reliable OPC verification process needs to include a fully rigorous treatment of the mask topography effects with taking into account oblique light incidence and polarization of light. In this paper we investigate the impact of rigorous mask topography simulation on the reliability of OPC verification and determine the influence of EMF aware OPC models on OPC quality. We use lithography simulations on OPCed layout cells where we apply a fully rigorous parallelized EMF solver to the mask model. Two different OPC models are used in this study; one based on the conventional approach and another one using EMF compensation techniques. The results of the rigorous lithography simulations are used to verify both OPC models. The impact of the EMF simulation on OPC verification quality is illustrated by direct comparison with the corresponding Kirchhoff simulations for both OPC models.