Characterization of a cell process from IV cell-test data

Several studies have been presented relating large data sets at cell test to the process parameters from the production line using statistics, regression, and machine learning (1,2). In these previous examples, laser-marks were used with tracking to associate each wafer’s cell-test parameters to each step in the process, which require additional laser marking tools and barcode readers and may impact the solar cell efficiency, as well as the cost of manufacturing. In the current paper, another approach is presented that can accomplish similar goals in many cases. An advanced production cell tester was used to measure all of the standard IV parameters, but also measured substrate resistivity, series resistance power loss at the maximum power point and the recombination parameters. This enables a more sophisticated assessment of each wafer as well as detailed statistical conclusions for large numbers of wafers. Examples of results from these techniques for production n-type industrial TOPCon cells and p-type PERC cells will be discussed. In each case, fundamental properties of key physical parameters are isolated by using the large data-set to evaluate dependencies with other parameters held constant. The goal is to have information from the IV-test data that identifies specific process parameters to optimize. This is especially critical for the production of both n- or p-type high-efficiency solar cells.