Title:	Leveraging Machine Learning for Understanding Charge Carrier Mobility in PEDOT Polymer
DNr:	NAISS 2024/22-501
Project Type:	NAISS Small Compute
Principal Investigator:	Ali Beikmohammadi <beikmohammadi@dsv.su.se>
Affiliation:	Stockholms universitet
Duration:	2024-04-03 – 2025-05-01
Classification:	10406
Keywords:

Abstract

We propose to develop a multiscale framework empowered by machine-learned charge transfer integrals to predict charge carrier mobilities in PEDOT polymer. Our approach entails exploring various molecular representations and employing kernel-based algorithms to accurately predict transfer integrals. By systematically comparing different representations and algorithms, we aim to identify the optimal combination for achieving high prediction accuracies. The ultimate goal is to significantly improve our understanding of structure-property relationships in organic electronics while reducing computational costs. With access to computational resources, we anticipate substantial advancements in the field and welcome the opportunity to contribute to this exciting research area.