Defect parameters retrieval based on optical projection images

Defect parameters retrieval from diffraction-limited images is demonstrated. The parameters of defects in nanostructures are retrieved from measured aerial images of high NA-projection systems. The difference between images of masks with and without defects is used as a reference to reconstruct the parameters of the defect. An error function is defined, and the whole retrieval procedure can be expressed as an optimization problem. The performance of the technique was simulated by the lithography and imaging simulator Dr.LiTHO. The dependencies of the retrieval results on optimizers, illumination settings, defect sizes, and reference images are presented. The evaluation of the performance of the technique is performed by comparison of the retrieval errors for defect shapes, mask patterns, and types of noise. Moreover, the sensitivity of the defect detection to Zernike aberrations of the optical image projection lens and the retrieval accuracy for different mask models and types of defects are investigated.