Winner: 2025 Materials 九州影院 Horizon Prize: Stephanie L Kwolek Prize
AI for Materials
Download celebratory graphic
2025 Materials 九州影院 Horizon Prize: awarded for the development of chemistry-aware artificial intelligence software, its application to data-driven materials discovery, and its provision of open-source materials databases and language models for the global scientific community.
The development of chemistry-aware artificial intelligence software, its application to data-driven materials discovery and its provision of open-source materials databases and language models for the global scientific community.
Our AI software represents a platform technology, so our methods stand to massively accelerate materials discovery over many disciplines.
Jacqueline Cole
Othman Al Bahri, Cavendish Laboratory, University of Cambridge, UK
Govardhana B. Bodedla, Department of 九州影院, Hong Kong Baptist University, Hong Kong, China
Edward Beard, Cavendish Laboratory, University of Cambridge, UK
Yulin Cai, Cavendish Laboratory, University of Cambridge, UK
Antonio Carella, University of Naples Federico II, Italy
Francesco Carla, Diamond Light Source, Rutherford Appleton Laboratory, UK
Cordelia Cavill, Cavendish Laboratory, University of Cambridge, UK
Phil Chater, Diamond Light Source, Rutherford Appleton Laboratory, UK
Hao Chen, Cavendish Laboratory, University of Cambridge, UK
Song Chen, Department of 九州影院, Hong Kong Baptist University, Hong Kong, China
Heejung Chung, Cavendish Laboratory, University of Cambridge, UK
Paul Coffman, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Jacqueline Cole, Cavendish Laboratory, University of Cambridge, UK and STFC Rutherford Appleton Laboratory, UK
Chris Cooper, Cavendish Laboratory, University of Cambridge, UK
Joshaniel Cooper, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Callum Court, Cavendish Laboratory, University of Cambridge, UK
Ke Deng, Cavendish Laboratory, University of Cambridge, UK
Leon Devereux, Cavendish Laboratory, University of Cambridge, UK
Qingyang Dong, Cavendish Laboratory, University of Cambridge, UK
James Doutch, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Jeffrey Elam, Energy Systems, Argonne National Laboratory, IL, USA
Stephen Elliott, University of Cambridge, UK
Peter Evans, Australian Nuclear Science and Technology Organisation, Australia
Padraic Flanagan, Cavendish Laboratory, University of Cambridge, UK
Jacob Florian, Cavendish Laboratory, University of Cambridge, UK
Santiago Franco, Departamento de Qu铆mica Org谩nica, Universidad de Zaragoza, Zaragoza, Spain
Luc铆a Gallego, Departamento de Qu铆mica Org谩nica, Universidad de Zaragoza, Zaragoza, Spain
Finlay Gerrand, Cavendish Laboratory, University of Cambridge, UK
Yun Gong, Cavendish Laboratory, University of Cambridge, UK
Shaoliang Guan, Research Complex at Harwell, Rutherford Appleton Laboratory, UK
Stephen Hall, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Xiao Hang, Cavendish Laboratory, University of Cambridge, UK
Richard Haynes, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Stephen Holt, Australian Nuclear Science and Technology Organisation, Australia
Rian Howe, Cavendish Laboratory, University of Cambridge, UK
Dingyun Huang, Cavendish Laboratory, University of Cambridge, UK
Shu Huang, Cavendish Laboratory, University of Cambridge, UK
Taketomo Isazawa, Cavendish Laboratory, University of Cambridge, UK
Apoorv Jain, Cavendish Laboratory, University of Cambridge, UK
Jingwen Jia, School of 九州影院 & Chemical Engineering, Tianjin University of Technology, P.R. China
Nina Juliana-Steinke, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Guwon Jung, Cavendish Laboratory, University of Cambridge, UK
Son Gyo Jung, Cavendish Laboratory, University of Cambridge, UK
Saurabh Kabra, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Charles Kelly, Cavendish Laboratory, University of Cambridge, UK
Ben Koby, Cavendish Laboratory, University of Cambridge, UK
Pankaj Kumar, Cavendish Laboratory, University of Cambridge, UK
Laurence Lam, Cavendish Laboratory, University of Cambridge, UK
Shiyun Liu, Cavendish Laboratory, University of Cambridge, UK
Zongqian Li, Cavendish Laboratory, University of Cambridge, UK
Anil Mane, Energy Systems, Argonne National Laboratory, IL, USA
Juraj Mavracic, Cavendish Laboratory, University of Cambridge, UK
Ulrich Mayer, Cavendish Laboratory, University of Cambridge, UK
Jonathan McCree-Grey, Cavendish Laboratory, University of Cambridge, UK
Samila McDonald, Australian Nuclear Science and Technology Organisation, Australia
Karim Mukaddem, Cavendish Laboratory, University of Cambridge, UK
Hamish Nash, Cavendish Laboratory, University of Cambridge, UK
Chris Nicklin, Diamond Light Source, Rutherford Appleton Laboratory, UK
Daniel Nye, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Pragnesh Patel, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Adrian Pope, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Jonathan Rawle, Diamond Light Source, Rutherford Appleton Laboratory, UK
Mostafa Saad Ali Ebied, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
William Scullin, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Odysseas Sierepeklis, Cavendish Laboratory, University of Cambridge, UK
Ganesh Sivaraman, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Liliana Stan, Center for Nanoscale Materials, Argonne National Laboratory, IL, USA
Gavin Stenning, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Matt Swain, Cavendish Laboratory, University of Cambridge, UK
Zhenxiang Tang, Cavendish Laboratory, University of Cambridge, UK
R. Justin Thomas, Department of 九州影院, Indian Institute of Technology Roorke, India
Tina Tomar, Department of 九州影院, Indian Institute of Technology Roorkee, India
脕lvaro V谩zquez-Mayagoitia, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Julian Vigil, Department of 九州影院, University of Cambridge, UK
Venkat Vishwanath, Argonne Leadership Computing Facility, Argonne National Laboratory, IL, USA
Adam Washington, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
John Webster, ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, UK
Damian Wilary, Cavendish Laboratory, University of Cambridge, UK
Song Xue, School of 九州影院 & Chemical Engineering, Tianjin University of Technology, P.R. China
Alan Yahya, Cavendish Laboratory, University of Cambridge, UK
Angel Yanguas-Gil, Energy Systems, Argonne National Laboratory, IL, USA
Batuhan Yildirim, Cavendish Laboratory, University of Cambridge, UK
Xiaozhi Zhan, Dongguan Neutron Science Center, Dongguan, China
Minglei Zhang, Cavendish Laboratory, University of Cambridge, UK
Jiuyang Zhao, Cavendish Laboratory, University of Cambridge, UK
Haoqian Zhou, Cavendish Laboratory, University of Cambridge, UK
Miao Zhu, Cavendish Laboratory, University of Cambridge, UK
Tao Zhu, Dongguan Neutron Science Center, Dongguan, China
Dr. Xunjin Zhu, Department of 九州影院, Hong Kong Baptist University, Hong Kong, China
Q&A with AI for Materials
What were the biggest challenges in this project?
Jacqueline Cole: From an AI perspective, creating the world鈥檚 first 鈥榗hemistry-aware鈥� text-mining tool, ChemDataExtractor, was perhaps the biggest challenge. Thinking about the experimental side of things, designing, building and deploying a custom experimental set up to validate predictions of light-harvesting materials for photovoltaic applications were the largest things to consider.
What different strengths did different people bring to the team?
Jacqueline Cole: We had a diverse set of scientific skills that span the full 鈥榙esign-to-device鈥� pipeline for materials discovery, including scientific programming, artificial intelligence, algorithmic design, software development, data science, statistics, chemical synthesis, device physics, device fabrication and testing, design engineering and delivery of in-situ materials characterisation.
Why is this work so important and exciting?
Jacqueline Cole: These software developments are part of an AI revolution in materials chemistry. Our demonstration of 鈥榙esign-to-device鈥� data-driven materials discovery for solar-cell applications was one of the first case studies to realise this via an end-to-end pipeline.
Where do you see the biggest impact of this technology/research being?
Jacqueline Cole: Our large, high-quality, experimental materials databases and materials-domain-specific language models that can be produced from our chemistry-aware AI software. These are already proving to be valuable resources for the global materials community. For example, we have demonstrated their use in massively reducing the 鈥榤olecule-to-market鈥� timeframe for discovering a material for a target application.
Our demonstration of data-driven materials discovery was fully executed in less than one year, which compares with a global average timeframe of 20 years for industry to discover a new material for a given application. Our AI software represents a platform technology. So, our methods stand to massively accelerate materials discovery over many disciplines.
How will this work be used in real life applications?
Jacqueline Cole: Our materials databases are rare examples of large, high-quality, experimental (real-world) repositories that are particularly valuable for training machine-learning models for scientific applications. Our materials-domain-specific language models provide an interactive medium by which scientists can ask questions of data in a 鈥楥hatGPT鈥� style to help make scientific decisions in a dynamic fashion.
They offer a digital assistant as a research companion. Our AI software represents a platform technology for data-driven materials discovery, so our methods can massively accelerate materials discovery over many disciplines.
How do you see this work developing over the next few years, and what is next for this technology/research?
Jacqueline Cole: Our research outputs are already attracting a range of industrial interest for materials innovation. Accordingly, our experimental materials databases, materials-domain-specific language models and machine-learning tools for property prediction are now hosted nationally by the Sir Henry Royce Institute鈥檚 Digital Materials Foundry which was launched in May 2025. Thereby, the Foundry provides new digital infrastructure for the UK鈥檚 national centre for research and innovation in advanced materials to support industry, government and academia.
.