Drift-Diffusion-Reaction and Machine Learning Modeling of Cu Diffusion in CdTe Solar Cells

In this paper we introduce the PVRD-FASP solver for studying carrier and defect transport in CdTe solar cells on an equal footing by solving 1D and 2D drift-diffusion-reaction model equations. The diffusion constants and activation energies of the defect and the defect chemical reactions require reaction rate constants that are calculated using density functional theory ( DFT). The PVRD-FASP solver can propose solutions that can reduce the development cost of thinfilm photovoltaics ( TFPV) because up- and down-stream process optimization, required due to complex interactions, is replaced by predictive modeling. An in-house implementation of a machine-learning approach for modeling of Cu diffusion in the CdTe absorber layer of the CdTe solar cell is also discussed.