EUV and non-EUV inspection of reticle defect repair sites

Abstract

We report the actinic (EUV wavelength) and non-actinic inspection of a multilayer-coated mask blank containing an array of open-field defect repair sites created in different ways. The comparison of actinic brightfield and darkfield measurements shows the importance of having both local reflectivity and scattering measurements. Although effective mask blank repair capabilities have not been adequately demonstrated, the data acquired in this experiment have been very instructive. Correlation with non-actinic inspection methods shows the difficulty of establishing a successful predictive model of the EUV response without EUV cross-comparison. The defect repair sites were also evaluated with SEM, AFM, and 488-nm-wavelength confocal microscopy. The data raise important questions about mask quality specifications and the requirements of future commercial actinic inspection tools.