Can I read Machine Learning-Assisted Defect Analysis and Optimization for P-I-N-Structured Perovskite Solar Cells on EtoBox?

Machine Learning-Assisted Defect Analysis and Optimization for P-I-N-Structured Perovskite Solar Cells by Seongtak Kim; Younghun Jeong; Dong-Woon Han; Chan Bin Mo is a Engineering article available to read on EtoBox.

What is Machine Learning-Assisted Defect Analysis and Optimization for P-I-N-Structured Perovskite Solar Cells about?

A solar cell capacitance simulator (SCAPS-1D) was used to prepare 3611 cell data with different defect densities in the bulk and interface of p-i-n-structured perovskite solar cells. The training was conducted using four machine learning algorithm models. The random forest algorithm had an accuracy and root mean square error of 0.999 and 0.00306, respectively. As per the explanatory Shapley additive explanations (SHAP) analysis, the bulk defects of the perovskite and the hole transfer layer/ perovskite and perovskite/electron transfer layer interface defects greatly affected the power conversion efficiency of the solar cells. It was also confirmed that when the perovskite defect density was low, the cell performance was more sensitive to the interface defect densities. Based on the predictive analysis of machine learning, a strategy to improve the solar cell efficiency of the p-i-n structure was presented, and the efficiency was improved from 17.97% to 24.66% in the poly(triarylamine)/ perovskite/phenyl-C61-butyric acid methyl ester structure by optimizing the defect density and resistance. It is expected that this methodology will not only help in identifying the factors affecting

Who reads Machine Learning-Assisted Defect Analysis and Optimization for P-I-N-Structured Perovskite Solar Cells?

Author

Seongtak Kim; Younghun Jeong; Dong-Woon Han; Chan Bin Mo

Publisher

Springer Science and Business Media LLC

Published

2023

Language

EN

Field

Engineering (Physical Sciences)