Evaluation of developers based on automatic classification of line-and-space pattern SEM images

Novel resists are being developed to meet the demands for lower dose consumption and reduced pattern defects in Extreme Ultraviolet (EUV) lithography. Consequently, standard developers are reaching their limits as the chemical properties of the resists evolve. New developers are needed to improve pattern quality. This study evaluated various developers during the development process. The resulting line-and-space (L/S) patterns, developed with different developers, were assessed using the Hough transform method. The evaluated feature values were then used to classify the SEM images into different classes using clustering methods. The class maps illustrated the pattern qualities obtained with different developers. The results were further explained through the simulation of Quartz Crystal Microbalance (QCM) chart results, which included frequency change (corresponding to the dissolution rate) and impedance change (corresponding to the viscosity change). This study provides deeper insights into how the chemical properties of developers impact the development process.