Investigating the impact of parameter and process variations on multicrystalline PERC cell efficiency

In this work, we thoroughly investigate the impact of parameter and process variations in an industrially feasible multicrystalline silicon (mc-Si) passivated emitter and rear cell (PERC) process. In an experiment, where in total more than 800 mc-Si PERC cells are fabricated, we distribute 51 neighboring highperformance mc-Si wafers and characterize them in depth to garner only the process-induced variations within a larger experiment. We use numerical three-dimensional (3-D) device simulations to model the width of the resulting distribution of solar cell efficiencies of this group. The main responsible parameters leading to a broadening of the distribution are extracted and ranked by their impact, highlighting the parameters that need to be tuned and controlled most accurately. In our case, from the variations in cell efficiencies amounting to 0.9 %abs., we identify the rear side passivation and a wrap-around during the emitter etch back process to be in charge of almost 60 % of the total fluctuations. With the presented approach and its findings, a better understanding of the underlying dependencies can be developed and subsequently applied to improve the ramp-up and the development of mc PERC cell production. In addition, the relevant parameters and their range obtained and denoted in this work are a fundamental input for modelling a mc-Si PERC solar cell, drawing a realistic illustration of parameters achieved in today's research facilities and during ramp-up of PERC production.