Automatic stress analysis method for semiconductor materials based on dual-wavelength infrared photoelasticity

Abstract

In this paper, an automatic method of stress analysis was proposed, based on dual-wavelength infrared photoelasticity, which was prospectively applicable to quality inspection in semiconductor manufacturing engineering. This method employed a strategy and corresponding algorithm to locate reference points by analyzing the rate of intensity change between two photoelastic images captured at dual wavelengths, relative to the material fringe constant, guiding the unwrapping process of the isoclinic value and isochromatic value maps. It enabled the full-field stress analysis without any manual intervention or any priori knowledge or information about the stress to analyze. An optic instrument was developed to realize the measurement of dual-wavelength infrared photoelasticity. Using this device, the feasibility of the proposed method was verified by quantifying the internal stress fields of different samples made in monocrystalline silicon wafers. The experimental results illustrated the full automation of the proposed method, and its high accuracy as well. Additionally, an optimized scanning scheme was further discussed to balance the efficiency and the accuracy based on the above method and device of dual-wavelength infrared photoelasticity.