Robustness of Parameter Estimation Procedures for Bulk-Heterojunction Organic Solar Cells

TL;DR

This paper investigates the reliability of parameter estimation methods for bulk-heterojunction organic solar cells using a 2D model, revealing limitations of 1D fitting and emphasizing the importance of multi-characterization approaches for accurate interpretation.

Contribution

It provides a detailed analysis of parameter estimation robustness in BHJ solar cells and offers practical recommendations to improve reliability and interpretation of estimated parameters.

Findings

Fitting only current-voltage data with 1D models can produce nonsensical parameters.

Simultaneous fitting of multiple characterization results enhances robustness.

The paper offers guidelines to avoid pitfalls in parameter estimation.

Abstract

Parameter estimation procedures provide valuable guidance in the understanding and improvement of organic solar cells and other devices. They often rely on one-dimensional models, but in the case of bulk-heterojunction (BHJ) designs, it is not straightforward that these models' parameters have a consistent physical interpretation. Indeed, contrarily to two- or three-dimensional models, the BHJ morphology is not explicitly described in one-dimensional models and must be implicitly expressed through effective parameters. In order to inform experimental decisions, a helpful parameter estimation method must establish that one can correctly interpret the provided parameters. However, only a few works have been undertaken to reach that objective in the context of BHJ organic solar cells. In this work, a realistic two-dimensional model of BHJ solar cells is used to investigate the behavior of state-of-the-art parameter estimation procedures in situations that emulate experimental conditions. We demonstrate that fitting solely current-voltage characteristics by an effective medium one-dimensional model can yield nonsensical results, which may lead to counter-productive decisions about future design choices. In agreement with previously published literature, we explicitly demonstrate that fitting several characterization results together can drastically improve the robustness of the parameter estimation. Based on a detailed analysis of parameter estimation results, a set of recommendations is formulated to avoid the most problematic pitfalls and increase awareness about the limitations that cannot be circumvented.