Physics-informed machine learning to accelerate stability research on perovskite solar cells

Información del proyecto

ACCELERATE-PER

Identificador del acuerdo de subvención: 101059891

DOI

10.3030/101059891

Proyecto cerrado

Fecha de la firma de la CE 19 Julio 2022

Fecha de inicio 1 Julio 2023

Fecha de finalización 30 Junio 2025

Financiado con arreglo a

• Marie Skłodowska-Curie Actions (MSCA)

Coste total

Sin datos

Aportación de la UE

€ 199 694,40

Coordinado por

AALTO KORKEAKOULUSAATIO SR
Finland

Periodic Reporting for period 1 - ACCELERATE-PER (Physics-informed machine learning to accelerate stability research on perovskite solar cells)

Período documentado: 2023-07-01 hasta 2025-06-30

Resumen del contexto y de los objetivos generales del proyecto

The development of sustainable energy production technologies, such as new types of solar cells, has traditionally taken decades. These research cycles need to be accelerated in order to respond to the urgent climate change crisis. Perovskite solar cells (PSCs) are a recent technology that boast high electricity production efficiencies, however they suffer from insufficient lifetimes. Since there are thousands of material candidates even for a single layer of a PSC, it is challenging to search for stable materials and infer device degradation mechanisms.

This project aimed to implement machine learning (ML) as a part of the solution. The objectives were to create an accelerated, ML-assisted research cycle for improving the lifetime of PSCs; with focus on optimizing the composition of the light absorbing layer – perovskite – for stability. This included showing the resulting stability improvements in full PSCs and develop faster inference of the remaining PSC degradation mechanisms. The acceleration in the cycle arises from i) using ML to save experimental bandwidth, ii) designing experiments that are directly compatible with ML (opposed to first collecting data and then figuring out how to use it with ML), and iii) maximizing effectiveness by teaching physics to the ML algorithm. For example, at the materials level, the algorithm needs to know that certain perovskite compositions do not exist as mixed materials according to density functional theory. The physics integration aligns ML algorithms with physical reality and alleviates data scarcity that has traditionally hindered the use of ML in experimental science.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

The scientific activities in this project were focused on three topics: development of physics-informed active learning for finding optimally stable perovskite materials for solar cells, understanding perovskite solar cell degradation under environmental stress, and developing further machine learning methodology for materials optimization.

The main achievements during the project were i) the development of hyperspectral camera -assisted accelerated characterizations of perovskite semiconductor materials for the faster development of new types of solar cells, ii) mapping the stability of a range of hybrid perovskite alloys against phase separation using computational and machine learning methods, iii) and developing more robust and reliable machine learning methods to be used for experimental materials optimization (including, human-in-the-loop Bayesian optimization and Bayesian optimization that learns the relevant contextual variables automatically during the optimization process).

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

The main results of the project are i) a novel hyperspectral camera -assisted accelerated characterization pipeline of perovskite semiconductor materials for the faster development of new types of solar cells, ii) a map of the stability of a range of hybrid perovskite alloys against phase separation, iii) and sharing of novel, more robust and reliable machine learning methods to be used for experimental materials optimization (i.e. human-in-the-loop Bayesian optimization and Bayesian optimization that learns the relevant contextual variables automatically during the optimization process).

These new methods and equipment will allow faster materials optimization for solar cells and other use cases from now on - by the authoring team and also the broader public because the methods are shared in open software repositories. Now, these approaches need to be tested for a wider set of materials to wet them for real use, which the researcher (Tiihonen) will complete in her own research laboratory from October 2025 onward. There is potential for commercializing these technologies as a service, or as novel laboratory equipment.