Graph Convolutional Neural Networks for Polymers Property Prediction

TL;DR

This paper demonstrates that graph convolutional neural networks can accurately predict polymer dielectric constants and bandgaps using only morphological data, outperforming other machine learning methods and eliminating the need for handcrafted features.

Contribution

The study introduces GCNN as a fast, accurate, and descriptor-free approach for polymer property prediction, validated against DFT calculations.

Findings

GCNN achieves high agreement with DFT results

Outperforms other machine learning algorithms

Requires only morphological data, no handcrafted descriptors

Abstract

A fast and accurate predictive tool for polymer properties is demanding and will pave the way to iterative inverse design. In this work, we apply graph convolutional neural networks (GCNN) to predict the dielectric constant and energy bandgap of polymers. Using density functional theory (DFT) calculated properties as the ground truth, GCNN can achieve remarkable agreement with DFT results. Moreover, we show that GCNN outperforms other machine learning algorithms. Our work proves that GCNN relies only on morphological data of polymers and removes the requirement for complicated hand-crafted descriptors, while still offering accuracy in fast predictions.