COMBINED PROCESS AND DEVICE SIMULATIONS TO SUPPORT BBR3 TUBE FURNACE DIFFUSION PROCESS OPTIMIZATION FOR INDUSTRIAL HIGH-EFFIICENCY N-TYPE CRYSTALLINE SILICON SOLAR CELLS

Abstract

This thesis focuses on the development of a predictive simulation framework which supports process optimization for BBr3 diffusion and improves the efficiency of silicon solar cells. Five main topics are investigated: a. A Sentaurus TCAD process simulation program is developed and calibrated by experiments. b. The passivation of heavily diffused p+ silicon surfaces is characterized. c. A systematic method to quantify the metallization-induced recombination losses on heavily diffused p+ silicon surfaces is developed, which involves intensity-dependent photoluminescence imaging and image processing with Griddler-AI. d. The experimentally observed microscopic metal-Si interface is studied by 2D unit-cell simulations. e. Combining the developed process and device simulations, cell efficiency improvements due to processing changes are quantified as a function of the process parameters. Experimentally, the efficiency of n-type bifacial front and back contacted silicon solar cells with screen-printed contacts is improved from 18.9 ± 0.5% to 20.8 ± 0.5%.