Prediction of exciton binding energies in two dimensional hybrid organic-inorganic perovskites using machine learning

In recent years, there have been a great number of studies focused on two dimensional hybrid organic-inorganic perovskites (2D HOIPs), owing to their unique optoelectronic properties that render them promising candidates for various photovoltaic technologies. A subject of particular importance is the relationship between the structure of a material and its charge carrier dynamics, an understanding of which is necessary for the design of finely-tuned photovoltaic devices. Building upon previous efforts to model this relationship, such as those by Hansen et al.¹, we have worked to develop a neural network to predict the excitonic behavior within a material given a set of structural parameters. Through the use of this model, we hope to be able to efficiently obtain accurate values for the exciton binding energy and radius without the need for costly DFT calculations or time-consuming sample fabrication and experimentation. This project aims to help improve the tunability of the properties of such 2D HOIPs, and has the potential to contribute to the development of more efficient photovoltaic technologies.

[1] Kameron R. Hansen, et al. Mechanistic origins of excitonic properties in 2D perovskites: Implications for exciton engineering, Matter, Volume 6, Issue 10, 2023, Pages 3463-3482, ISSN 2590-2385, https://doi.org/10.1016/j.matt.2023.07.004.