A combined ML and DFT strategy for the prediction of dye candidates for indoor DSSCs

Abstract

The excellent ability of dye-sensitized solar cells (DSSCs) to capture ambient light and convert it into electric current makes them attractive power sources for indoor applications, including powering Internet of Things (IoT) devices. In this context, substantial research efforts have been devoted to the discovery of novel organic dyes able to harvest energy from a wide range of indoor light sources at different intensities. However, such activities are often based on trial-and-error procedures which are frequently expensive and time-consuming. Here, Machine Learning (ML) techniques and Density Functional Theory (DFT) methods have been combined in a two-stage approach, with the aim to accelerate the design of new, synthetically accessible organic dyes for indoor DSSC applications. By predicting the power conversion efficiency (PCE) under different indoor light sources and intensities, potentially high-performance organic dyes have been identified.