Classification of crystal defects in multicrystalline silicon solar cells and wafer using spectrally and spatially resolved photoluminescence

In this contribution, spectral photoluminescence (SPL) imaging detecting both the spectral distribution and the lateral position is applied on recombination active defects in multicrystalline silicon solar cells and wafers. The result is analysed by a Multivariate Curve Resolution (MCR) algorithm using the spectral photoluminescence response and their positions. (i) Without any pre-assumptions made, the algorithm distinguishes four different recombination active defect types. Looking at the spatial distribution, it is shown that two of these defect types coincide with two defect types that have been distinguished on solar cell level using an analysis of forward and reverse biased electroluminescence (denoted as Type-A and -B) previously. (ii) Using SPL, all previously classified defects can also be distinguished at the wafer level. Therefore, the defects limiting the solar cell efficiency are already present in the wafer material and not introduced by the solar cell process. This is of particular interest for the question of how to predict the solar cell efficiency based on the PL measurements at the wafer level. The SPL is able to distinguish between the recombination activity of the dominant Type-A and -B defects that cannot be distinguished by classical PL measurements of the band-to-band recombination at the wafer level. The technique also highlights the changes in recombination activity of the given defects throughout the fabrication process. (iii) Additionally, it is shown that the spectral peak positions of Type-A defects coincide with the known D3 and D4 lines and of Type-B defects with the D1 line on both solar cell and wafer level. Two further defects are captured by the MCR algorithm denoted as Type-VID3 and Type-D07 defects occurring as spot-like defects in isolated positions. Their spectral PL response is analysed as well. (C) 2016 AIP Publishing LLC.