Generative Machine Learning Models for the Deconvolution of Charge Carrier Dynamics in Organic Photovoltaic Cells

TL;DR

This paper introduces ta-LLODE, a machine learning framework that models charge carrier dynamics in organic solar cells, enabling detailed analysis, simulation, and prediction of device behavior from experimental data.

Contribution

The paper presents ta-LLODE, a novel machine learning approach that disentangles and reconstructs charge carrier dynamics, providing interpretability and predictive capabilities for organic photovoltaic research.

Findings

Charge carrier decay is well described by a compressed exponential model.

The latent space enables accurate simulation of experimental conditions.

The framework supports device optimization through predictive modeling.

Abstract

Charge carrier dynamics critically affect the efficiency and stability of organic photovoltaic devices, but they are challenging to model with traditional analytical methods. We introduce \b{eta}-Linearly Decoded Latent Ordinary Differential Equations (\b{eta}-LLODE), a machine learning framework that disentangles and reconstructs extraction dynamics from time-resolved charge extraction measurements of P3HT:PCBM cells. This model enables the isolated analysis of the underlying charge carrier behaviour, which was found to be well described by a compressed exponential decay. Furthermore, the learnt interpretable latent space enables simulation, including both interpolation and extrapolation of experimental measurement conditions, offering a predictive tool for solar cell research to support device study and optimisation.