Learning from the Harvard Clean Energy Project: The Use of Neural Networks to Accelerate Materials Discovery

TLDR

The study proposes using multilayer perceptron neural networks within a computational funneling framework to accelerate high‑throughput design of organic materials. This approach employs multilayer perceptrons as surrogate models to predict material properties, thereby reducing the need for exhaustive quantum‑chemical calculations. The neural‑network surrogate accurately reproduces quantum‑chemical results, dramatically cuts the fraction of the screening library that must be explicitly computed, and enables large‑scale molecular screening that expands the size and diversity of explored libraries.

Abstract

Here, the employment of multilayer perceptrons, a type of artificial neural network, is proposed as part of a computational funneling procedure for high‐throughput organic materials design. Through the use of state of the art algorithms and a large amount of data extracted from the Harvard Clean Energy Project, it is demonstrated that these methods allow a great reduction in the fraction of the screening library that is actually calculated. Neural networks can reproduce the results of quantum‐chemical calculations with a large level of accuracy. The proposed approach allows to carry out large‐scale molecular screening projects with less computational time. This, in turn, allows for the exploration of increasingly large and diverse libraries.