Catalysis − the reduction of activation energy in a chemical reaction by a catalyst − plays a key role in the chemical industry, as well as in the development of sustainable technologies essential for achieving a low-carbon economy. However, the search for high-performance and sustainable catalysts is often costly and time-consuming. It can be accelerated through data-driven catalysis research. Yet experimental data are often not available in machine-readable and standardised formats. This is where the Catalysis App comes in: “With the Catalysis App, we have created a solution that enables researchers to store and share their data in a unified format. This facilitates the comparability of results and lays the foundation for future AI-supported catalyst development,” says Dr Annette Trunschke, Principal Investigator of FAIRmat.
The Catalysis App was developed by the FAIRmat consortium within the National Research Data Infrastructure (NFDI). An article describing this work has now been published in the journal Nature Catalysis. It is the result of a collaboration between researchers from FAIRmat at the Centre for the Science of Materials Berlin (CSMB) at Humboldt-Universität zu Berlin (HU), the Fritz Haber Institute of the Max Planck Society, and Helmholtz-Zentrum Berlin. A key contributor was Dr Julia Schumann, FAIRmat Catalysis Expert at HU, while the scientific lead was Dr Annette Trunschke, Principal Investigator at FAIRmat.
Added value increases with available data
The Catalysis App allows researchers to work with structured, comparable and machine-readable data − either via an intuitive graphical user interface (GUI) or programmatically via an application programming interface (API). These interfaces are designed to enable exploration of data from different perspectives. For example, suitable catalysts for specific reactions can be identified, or reaction products for given starting materials can be analysed. In addition, numerous parameters − such as synthesis methods, catalyst forms or reaction conditions − can be filtered, for instance to specifically search for high-pressure or low-temperature reactions.
A particular advantage of the tool is its integrated data visualisation. Data can be uploaded either manually (via the GUI) or programmatically (via the API). Special data structures and templates have been developed to enable simple and standardised data entry.
Ultimately, the scientific value of the app depends significantly on the scope, diversity and quality of the available data. “As community participation grows, the benefits of the platform will continue to increase. Researchers worldwide are therefore invited to use the application, contribute data and provide feedback in order to actively help shape its further development,” says Dr Annette Trunschke.
FAIRmat: Handling large volumes of research data
Modern research institutions require effective data management infrastructure and tools to ensure the quality, traceability and long-term usability of data. This is where the FAIR principles come in: Findability, Accessibility, Interoperability and Reusability of digital data. These principles are implemented by the FAIRmat consortium within the National Research Data Infrastructure (NFDI), which is coordinated at Humboldt-Universität zu Berlin. The goal of research data management is to promote transparency and reproducibility in research while avoiding redundant work through the reuse of existing data. As such, it is a central component of good scientific practice. Within FAIRmat, these principles are implemented in the NOMAD data infrastructure.
NOMAD: Publicly accessible data infrastructure for the Catalysis App
NOMAD (Novel Materials Discovery) is a data infrastructure for research data management in materials science. It provides the necessary flexibility for the development of domain-specific tools for managing standardised experimental catalysis data. Within NOMAD, the Catalysis App was developed with the aim of establishing FAIR-compliant data practices in catalysis. Since its launch in 2014 as a repository for computed data, NOMAD has been continuously developed and expanded.