Automatic etch pit density analysis in multicrystalline silicon

Abstract

This publication contains a description and is published in combination with the source code for a tool capable of determining the etch pit density (EPD) on multicrystalline silicon image data. The algorithm is capable of classifying grain boundaries and polishing scratches and removes these structures from the analysis result. Included with the analysis code are methods for plotting EPD maps as well as relative EPD frequency. This is combined with a brief description of the experimental steps of wafer preparation and defect etching as well as a discussion of the analysis limitations.

Design of automatic detection algorithm for dislocation contrasts in birefringence images of SiC wafers

“…The result of the automatic detection for the birefringence image was comparable to that for the etch pit counting of semiconductor wafers. 29,30) The present results indicate that threading dislocation density and the position of commercial SiC wafers are possible to be detected by the birefringence imaging as well as the current automatic detection algorithm using variance filter. This implies that the birefringence imaging is the candidate for a non-destructive inspection method for the threading dislocations in SiC wafers.…”

Scalable morphological accessibility of complex microstructures

“…Due to the complexity of some interconnected networks, named microstructures when real media are considered, their structural characterization often requires combination of various operators, named descriptors, focusing either on local -morphological-or global -topological-features [9,10,11]. Among topological measures, connectivity notion is considered in various descriptors, as connectivity number and Euler number, and concepts too, as percolation and accessibility [1,12,3]. This latter turned out to have a central interest on various experiments considering hindrance phenomenon, making use of different flowing particles and measurement means [13,14,15,16].…”