The Open Polymers 2026 (OPoly26) Dataset and Evaluations

TL;DR

The paper introduces the OPoly26 dataset, comprising over 6.57 million DFT calculations on polymeric systems, to enhance machine learning models for polymer property prediction and facilitate broader development of atomistic models.

Contribution

The creation and public release of the large-scale OPoly26 dataset, filling a gap in polymer data for machine learning applications and improving predictive performance.

Findings

Augmenting ML training with OPoly26 improves polymer prediction accuracy.

OPoly26 captures diverse polymer chemistries, including monomer types and chain architectures.

Dataset supports development of universal atomistic models for polymers.

Abstract

Polymers-macromolecular systems composed of repeating chemical units-constitute the molecular foundation of living organisms, while their synthetic counterparts drive transformative advances across medicine, consumer products, and energy technologies. While machine learning (ML) models have been trained on millions of quantum chemical atomistic simulations for materials and/or small molecular structures to enable efficient, accurate, and transferable predictions of chemical properties, polymers have largely not been included in prior datasets due to the computational expense of high quality electronic structure calculations on representative polymeric structures. Here, we address this shortcoming with the creation of the Open Polymers 2026 (OPoly26) dataset, which contains more than 6.57 million density functional theory (DFT) calculations on up to 360 atom clusters derived from polymeric systems, comprising over 1.2 billion total atoms. OPoly26 captures the chemical diversity that makes polymers intrinsically tunable and versatile materials, encompassing variations in monomer composition, degree of polymerization, chain architectures, and solvation environments. We show that augmenting ML model training with the OPoly26 dataset improves model performance for polymer prediction tasks. We also publicly release the OPoly26 dataset to help further the development of ML models for polymers, and more broadly, strive towards universal atomistic models.