Predicting very rare stochastic defects in EUVL processes for full-chip correction and verification

The quality of any lithographic pattern is dependent on the photon absorption events that a photoresist pattern receives during exposure and subsequent molecular events. The failure probability is not governed by the stochastics of any one process and is conditional on multiple stochastic and conventional variations (such as resist thickness). These processes jointly result in the observed distribution with multiple pathways resulting in the same rare event. This paper will examine how probabilistic processes combine to form rare defects so that these defects can be predicted and corrected. Rigorous simulations will be used to study the distributions and will be applied to existing stochastic compact model forms.