Luminescence Image Analysis Using Finite-Element Models: Finished Solar Cell Analysis

A general computational routine which seeks cell parameters that best explain a set of luminescence imaging data is presented in this article. The routine combines Griddler, a finite-element simulator for solar cells, and tailored multivariate regression techniques. It is applied to a dataset of luminescence images from about 80 monocrystalline silicon Al back surface field solar cells with varying front metal grid contact resistance and recombination. The fitted cell parameters including contact resistance spatial distribution, saturation current densities of the wafer's passivated regions, and under the metal contacts lead to simulated I-V characteristics that are in agreement with experimental data. The extracted contact resistance also correlates well with transmission line method measurements. In a separate study of luminescence imaging data from a multicrystalline passivated emitter rear local contacted cell (PERC) cell, this analysis routine is also compared with an alternative high spatial resolution method, yielding comparable results.