Thickness Optimization Simulation to Minimize Effective Reflectance in Broadband for SiO2/TiO2 Anti-Reflection Coating on Silicon Solar Cells

Jong-Chol Kima, Yun Ria, Un-Song Pakb, Kyong-Chol Kimc, Hyok-Chang Kwond aInstitute of Silicate Engineering, National Academy of Science, Pyongyang, Democratic People's Republic of Korea bPyongyang Transportation University, Pyongyang, Democratic People's Republic of Korea cFaculty of Applied Chemical Engineering, Kim Chaek University of Technology, Pyongyang, Democratic People's Republic of Korea dInstitute of Intelligent Information Technology, National Academy of Science, Pyongyang, Democratic People's Republic of Korea

Full Text PDF

By the finite-difference time-domain method, we modeled the reflection coefficient with the ratios of the layer thicknesses of SiO2/TiO2. SiO2/TiO2 thin film thickness ratios that minimize the effective reflectance in the broadband of 400-800 nm were simulated using an artificial neural network. For the Si/SiO2/TiO2 system, the results were obtained with the following layer parameters: n(Si)=3.7-5.6, n(TiO2)=2.3-2.7, d(TiO2)=25 nm, and n(SiO2)=1.5, d(SiO2)=24 nm. The average effective reflectance coefficient in the broadband of 400-800 nm was about 6.6%. Accordingly, the optimization of the thickness parameters of the anti-reflection film has shown that it is possible to significantly reduce the total effective reflectance in the visible range, thereby increasing the efficiency of the solar cells.

DOI:10.12693/APhysPolA.141.557 topics: silicon solar cell, FDTD, neural network, effective reflectance