AI for Better and Faster Photovoltaic Materials

The quest for sustainable energy solutions has been a major focus of scientific research for decades. Solar energy, a clean and renewable source, has emerged as a promising alternative to traditional fossil fuels. In particular, perovskite solar cells have gained significant attention due to their flexibility and sustainability.

A Collaborative Approach

A recent breakthrough in materials science has accelerated the discovery of novel perovskite materials. By leveraging the power of artificial intelligence (AI) and high-throughput synthesis, researchers have been able to identify promising molecules that can significantly enhance the efficiency of perovskite solar cells.

An international team inclusive of scientists from the Helmholtz Institute Erlangen-Nuremberg for Renewable Energy (HI ERN), a branch of Forschungszentrum Jülich, and researchers from the Karlsruhe Institute of Technology (KIT) pioneered this innovative approach. They combined the computational prowess of AI with fully automated high-throughput synthesis that can be transferred to other areas of materials research, such as the search for new battery materials. The outcome of the research was published in the renowned journal Science.

The Power of AI-Driven Material Design

The researchers began by creating a virtual library of one million molecules that make perovskite solar cells particularly efficient as conductors of positive charge. These million molecules are produced and tested by tenure-track professor Pascal Friederich from the KIT Institute of Nanotechnology and Professor Christoph Brabec from HI ERN. “With only 150 targeted experiments, a breakthrough was achieved that would otherwise have required hundreds of thousands of tests. The developed workflow opens up new possibilities for the fast and cost-efficient discovery of high-performance materials in a wide range of application fields,” explains Brabec. They increased the efficiency of a reference solar cell by around 2 percent to 26.2 percent using one of the materials discovered.

A database with the structural formulas of about one million virtual molecules that could be produced from commercially available substances was the origin at HI ERN. The KIT researchers examined energy levels, polarity, geometry, and other characteristics using established quantum mechanical methods from 13,000 randomly read virtual molecules.

AI training with data from only 101 molecules

To make the research as authentic as possible, the researchers selected 101 molecules that had distinct characteristics. With the help of a robotic system, these molecules were automatically produced at HI ERN to measure their efficiency. Christoph Brabec, who led the work at HI ERN stated, “”It was crucial for the success of our strategy that we were able to generate truly comparable samples thanks to our highly automated synthesis platform and thus determine reliable values for efficiency.”

The KIT researchers trained an AI model based on the efficiencies and the characteristics of the associated molecules. The model further proposed an additional 48 molecules for synthesis based on two criteria: expected high efficiency and unpredictable properties. “If the machine learning model is unsure about predicting the efficiency, it is worthwhile to produce the molecule in order to investigate it in more detail,” explains Pascal Friederich, while mentioning the second criterion, “It could surprise with a high degree of efficiency.”

In fact, the molecules proposed by AI could be used to build exceptional solar cells that surpass others. “We can’t be sure that we have really found the best among a million molecules, but we are certainly close to the optimum,” states Friederich.

KI versus chemische Intuition

Due to AI’s intuitive capabilities, the researchers were able to understand its molecular proposals to a certain limit. In addition, the AI proposals were based on features such as the presence of certain chemical groups such as amines, which the chemists had previously overlooked.

Christoph Brabec and Pascal Friederich are confident of their strategy booming in materials research and by extension, the optimization of entire components.

The transformative power of AI in Materials Science

The integration of AI and high-throughput synthesis represents a major leap forward in materials science. By automating the laborious process of material discovery, researchers can now explore vast chemical spaces and identify groundbreaking materials at an unprecedented rate.

The research results, developed in collaboration with researchers from the University of Erlangen-Nuremberg, the South Korean Ulsan National Institute of Science, the Chinese Xiamen University, and the University of Electronic Science and Technology in Chengdu, China, were recently published in the renowned journal Science.

Original Source: https://www.fz-juelich.de/de/aktuelles/news/pressemitteilungen/2024/mit-ki-schneller-zu-besseren-photovoltaik-materialien

Original Publication
Jianchang Wu, Luca Torresi, ManMan Hu, Patrick Reiser, Jiyun Zhang, Juan S. Rocha-Ortiz, Luyao Wang, Zhiqiang Xie, Kaicheng Zhang, Byung-wook Park, Anastasia Barabash, Yicheng Zhao, Junsheng Luo, Yunuo Wang, Larry Lüer, Lin-Long Deng, Jens A. Hauch, Dirk M. Guldi, M. Eugenia Pérez-Ojeda, Sang Il Seok, Pascal Friederich, Christoph J. Brabec
Journal: Science
Article Title: Inverse design workflow discovers hole-transport materials tailored for perovskite solar cells
Article Publication Date: 12-12-2024
DOI: 10.1126/science.ads0901
Website: https://www.fz-juelich.de/

Contacts
Prof. Christoph Brabec
Institute of Energy Technologies, Helmholtz Institute Erlangen-Nuremberg for Renewable Energies (IET-2 / HI ERN)
Telephone: +49 9131/85-25462
E-Mail: c.brabec@fz-juelich.de

Jessica Pölloth
Institute of Energy Technologies, Helmholtz Institute Erlangen-Nuremberg for Renewable Energies (IET-2 / HI ERN)
Telephone: +49 9131-12538204
E-Mail: j.poelloth@fz-juelich.de

Press Contact
Dr. Regine Panknin
Press Officer Forschungszentrum Jülich
Telephone: +49 2461/61-9054
E-Mail: r.panknin@fz-juelich.de