Leaving the Valley: Charting the Energy Landscape of Metal/Organic Interfaces via Machine Learning

TL;DR

This paper introduces a machine learning approach to efficiently explore the energy landscape of metal/organic interfaces, enabling prediction of polymorphs and defects with limited DFT data, demonstrated on tetracyanoethylene on Ag(100).

Contribution

The study presents a novel machine learning method that predicts formation energies of interface polymorphs using minimal DFT calculations, significantly reducing computational effort.

Findings

Successfully predicted polymorph energies for tetracyanoethylene on Ag(100)

Explained experimentally observed anisotropic ordering

Reduced DFT calculations needed for interface energy landscape exploration

Abstract

The rich polymorphism exhibited by inorganic/organic interfaces is a major challenge for materials design. In this work we present a method to efficiently explore the potential energy surface and predict the formation energies of polymorphs and defects. This is achieved by training a machine learning model on a list of only 100 candidate structures that are evaluated via dispersion-corrected Density Functional Theory (DFT) calculations. We demonstrate the power of this approach for tetracyanoethylene on Ag(100) and explain the anisotropic ordering that is observed experimentally.