Abstract
In this paper we present a novel method to describe the quality of multi-crystalline as-cut wafer based on photoluminescence imaging (PL). PL has a high potential to detect efficiency relevant defects already on as-cut wafers. Defects that can be detected are for instance crystal dislocations and contaminations from iron precipitates or the crystallization crucible. We present reliable image processing algorithms to detect and quantify quality features related to specific defects. For an interpretable presentation the quality features are combined in a histogram. We show that the histogram contains a large fraction of the physically relevant information to predict the open circuit voltage of the finished solar cells by an artificial neural network. This proves that the features can be used to establish a meaningful rating of the wafer quality.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
