Abstract

Laser light scattering is commonly used for surface-defect inspections in nanofabrication processes such as the manufacture of semiconductors and hard disks. In this paper, convex/concave discrimination of nano-scale defects on surfaces utilizing laser light scattering is described. Angular distribution of scatter intensity is analyzed for common silicon wafer surface defects (particles, crystal originated particles, and microscratches) using Discrete Dipole Approximation simulation. We show an example of a new method for convex/concave defect discrimination that utilizes angular distribution in both the azimuthal and elevational directions on the basis of calculated defect scatter characteristics. The results of our evaluation show that discrimination accuracy improves compared with the conventional method that utilizes angular intensity distribution only in the elevational direction.