Photovoltaics (PV) are a renewable energy technology experiencing rapidly increasing commercial adoption today. Nonetheless, many proposed PV applications
still require higher efficiencies, lower costs and comparable reliability to currently
available in commercial devices (typically made from silicon). To enable the rigorous study of a much wider range of materials and novel design concepts, particularly
those based on compound thin films, Concentrated Photovoltaics (CPV), cells with
bifaciality, a comprehensive modeling framework is developed to couple photon absorption, carrier transport, photon recycling, and thermal transport in PV devices.
The universality of this framework manifest itself in approaching various PV related
problems as follows: 1) exploring the novel design of wide-Eg GaInP solar cells as
an intermediate step to enhance the efficiency of multijunction PV devices; 2) characterizing the open-circuit voltage (VOC) degradation in thin-film vapor liquid solid
(TF-VLS) grown InP solar cell through combined device and circuit model for interpreting photoluminescence (PL) image; 3) establishing optic-electric-thermal coupled
framework to assess and compare the passive cooling effect for Silicon CPV devices
that employ porous soda-lime glass radiative cooler and conventional copper cooler
respectively; 4) Investigating and formulating the analytic solution of the optimal
design that minimizes combined optical shadowing loss and electrical resistive loss
for two types of bifacial PV devices: a) interdigitated back contact (IBC) Silicon
heterojunction (SHJ) solar cells and b) Copper Indium Gallium DiSelenide (CIGSe)
solar cell with Al2O3 passivation; and 5) Constructing an Neural Network Autoen- coder (NNA) that compresses and reconstructs the J-V characteristics obtained from
TCAD simulation and literature for rapid screening and automated classification.