Feature selection method suitable for optical scattering measurement

Abstract

The invention belongs to the related technical field of optical scattering measurement, and discloses a feature selection method suitable for optical scattering measurement, which comprises the following steps of: (1) determining the nanostructure morphology, the key size to be measured and the material optical constant of a structure to be measured according to a semiconductor process; (2) constructing a nanostructure forward optical characteristic model of the to-be-measured structure; (3) dividing a sparse grid XS, and calculating an all-optical feature combination fa corresponding to all discrete points in the sparse grid so as to form a sparse optical feature library omega S; (4) obtaining candidate feature combinations fc from the all-optical feature combinations fa based on a filtering type feature selection algorithm; and (5) further extracting the feature combination used by the evaluation function in the library matching method from the candidate feature combination fc basedon a wrapped feature selection algorithm to obtain a final optimized feature combination f*. According to the method, the offline library building time in the library matching method is shortened, andthe parameter extraction accuracy is improved.

Description

technical field [0001] The invention belongs to the related technical field of optical scatterometry, and more specifically relates to a feature selection method suitable for optical scatterometry. Background technique [0002] In the semiconductor process, the manufacturing process of integrated circuits includes multiple processes such as material preparation, photolithography, cleaning, etching, doping, and chemical mechanical polishing, among which the photolithography process is the most critical. The main indicators of the lithography process are key dimensions, resolution, depth of focus, alignment and overlay accuracy, etc. In the photolithography process, in order to evaluate and control the processing accuracy of the photolithographic pattern, a special line pattern is designed to reflect the characteristic line width of the integrated circuit. The characteristic size of the line pattern is called the critical dimension (CD). Therefore, in the semiconductor manufa...

AI Technical Summary

Problems solved by technology

The literature "Scatterometry-based metrology with feature region signatures matching" and CN104807398B both propose to calculate the evaluation function based on the optical features with higher sensitivity to improve the accuracy of library matching, but they do not consider the redundancy between highly sensitive features

Images