Machine learning frontier orbital energies of nanodiamonds

TL;DR

This paper introduces ND5k, a new dataset of nanodiamond structures with computed frontier orbital energies, and evaluates machine learning models, especially equivariant graph neural networks, for predicting these energies.

Contribution

The creation of ND5k dataset and the benchmarking of ML models, notably PaiNN, for predicting nanodiamond orbital energies, including extrapolation to larger structures.

Findings

PaiNN achieves best interpolation and extrapolation performance.

ML models can predict orbital energies with high accuracy.

Extrapolation to larger structures is feasible with advanced GNNs.

Abstract

Nanodiamonds have a wide range of applications including catalysis, sensing, tribology and biomedicine. To leverage nanodiamond design via machine learning, we introduce the new dataset ND5k, consisting of 5,089 diamondoid and nanodiamond structures and their frontier orbital energies. ND5k structures are optimized via tight-binding density functional theory (DFTB) and their frontier orbital energies are computed using density functional theory (DFT) with the PBE0 hybrid functional. We also compare recent machine learning models for predicting frontier orbital energies for similar structures as they have been trained on (interpolation on ND5k), and we test their abilities to extrapolate predictions to larger structures. For both the interpolation and extrapolation task, we find best performance using the equivariant graph neural network PaiNN. The second best results are achieved with a message passing neural network using a tailored set of atomic descriptors proposed here.