Prediction of Inorganic Solar Cell Materials with Double Perovskite Structure Using Machine Learning Approach
Project Description

Perovskite based solar cell materials attracts great attention in the clean energy technology. Within a decade, the power conversion efficiency of hybrid organic-inorganic perovskites (HOIPs) material has grown from 3.8% to 22.1%. However, HOIPs materials are unstable under heat and especially moisture, where the moisture could hydrate and destroy the structure.To overcome this challenge, researchers have been exploring stable inorganic materials with double perovskite structures as alternative. For instance, Kuang et al.2 showed that Cs2PdBr6 could withstand 1 sun illumination in the air for more than 1000 hours without significant structure damage, in comparison to 100 hours for HOIPs.3 In the double perovskite framework, thousands of plausible double perovskite structures could be proposed. But for now, only 24 double perovskites materials were synthesized and characterized.

Supervisor
SU Haibin
Co-Supervisor
LIN Zhenyang
Quota
1
Course type
UROP1000
UROP1100
UROP2100
UROP3100
UROP4100
Applicant's Roles

employ a set of such algorithms including Kernel Ridge Regression, Gaussian Progressor Regression, Least Absolute Shrinkage and Selection Operator, and Neural Network to develop machine learning models for inorganic materials with double perovskite structures

Applicant's Learning Objectives

training in machine learning algorithm; research experiences in solar cell materials

Complexity of the project
Challenging