Efficient, Interpretable Graph Neural Network Representation for Angle-dependent Properties and its Application to Optical Spectroscopy

TL;DR

This paper introduces ALIGNN-d, an extension of graph neural network encoding that includes bond and dihedral angles, enabling accurate and interpretable predictions of optical properties in disordered atomic systems.

Contribution

The paper presents a memory-efficient graph neural network encoding that captures complete atomic geometry, including angles, for improved property prediction and interpretability.

Findings

Bond and dihedral angles significantly influence optical absorption spectra.

ALIGNN-d effectively predicts infrared responses of disordered Cu(II) complexes.

Structural distortions correlate with absorption intensity changes.

Abstract

Graph neural networks are attractive for learning properties of atomic structures thanks to their intuitive graph encoding of atoms and bonds. However, conventional encoding does not include angular information, which is critical for describing atomic arrangements in disordered systems. In this work, we extend the recently proposed ALIGNN encoding, which incorporates bond angles, to also include dihedral angles (ALIGNN-d). This simple extension leads to a memory-efficient graph representation that captures the complete geometry of atomic structures. ALIGNN-d is applied to predict the infrared optical response of dynamically disordered Cu(II) aqua complexes, leveraging the intrinsic interpretability to elucidate the relative contributions of individual structural components. Bond and dihedral angles are found to be critical contributors to the fine structure of the absorption response, with distortions representing transitions between more common geometries exhibiting the strongest absorption intensity. Future directions for further development of ALIGNN-d are discussed.