Background and Objective In the recent years, a rapid emergence of a new class of solar cell based on mixed organic and inorganic halide perovskites. Incorporating perovskites into semiconductor devices such as solar cells has shown good performance with a confirmed power conversion efficiencies reaching to 20%. However, many potential tunable parameters impact the properties of interest remains unknown. The objective of this project is to develop models that find relationships between the structures of hybrid organic perovskites (HOIPs) and its properties that are important for materials design. This project will incorporate a wide range of data, including published experimental data, simulated data from the Li research group, data from collaborators, as well as data from the Materials Project database. Additionally, the development of Python software framework will allow the integration of many types of data to be used for statistical inference and design, including electronic structure calculations, spectroscopic analysis, and molecular dynamics. The goal of our team is to enable the rational design of perovskite solar cells by optimizing tunable parameters to yield desired properties and behaviors.
Sarah Floris, Yongquan Xie, Hongbin Liu
We have used the data science techniques to identify the optimal combination of elements in the HOIPs for the energy conversion in the solar cells. The lead based HOIP is so far still the best candidate for the solar cell application. A prediction in the full range of FA/MA, I/Br ratios has been made based on the neural networks.