FEATURE CONTRAST IN DOSE-EQUALIZATION SCHEMES USED FOR ELECTRON-BEAM PROXIMITY CONTROL

Dose equalization by the GHOST technique fails to provide homogeneous contrast when the two-Gaussian model does not adequately portray the b eam-energy dose distribution. This leads to an undesirable pattern dep endence to the contrast obtained. The degree to which this breakdown o ccurs is shown in computer simulation. The most severe breakdown occur s in enclosed (annular) structures where the effect of ''higher order' ' Gaussian deposition peaks are concentrated. Unanticipated background fluctuations of about 10% of the nominal exposure dose are obtained. As a result of this study, test structures are proposed which can exte nd the utility of dose equalization.