An Open Access Digital Twin to Support High-Throughput Experiments on New Materials

Databases of computed material properties have become essential tools for materials research. On the other hand, databases of experimental properties (and databases of experimental materials research results) have not reached the same level of maturity.

In our group, we can now manage, analyze, and share our experimental high-throughput material development data on the public, open-access NOMAD platform (https://nomad-lab.eu/). Synthesis data comes from combinatorial thin-film growth of a novel family of thin-film compounds (phosphosulfides) by reactive sputter deposition in a unique setup. We have implemented:

- a synthesis data/metadata structure based on the Chemical Methods Ontology (CHMO) for interoperability, but with additional features to accommodate our specific type of growth process

- logfile readers for easily sharing and visualizing details of every growth process

- a tracking system for deposited samples and deposition sources based on QR codes attached to objects in the NOMAD platform

On the characterization side, data is geared towards optoelectronic applications. We have developed custom NOMAD parsers and data analyzers for high-throughput energy-dispersive x-ray spectroscopy (elemental composition), scanning electron microscopy (morphology), x-ray diffraction (structure, phase mix), spectroscopic ellipsometry (optical constants), photoluminescence spectroscopy (minority carrier lifetimes, radiative efficiency). We have implemented:

- linking of raw characterization data and the material properties extracted by high-throughput analysis of such data to each combinatorial sample on NOMAD

- linking our own experimental characterization data, our own first-principles calculation data, and the existing calculation data already present on NOMAD

- various artificial intelligence tools to be used both for decision-making in the lab and for understanding complex composition-structure-process-property-performance relationships.

I will describe how this digital infrastructure can now enable cross-laboratory collaboration, meaningful sharing of both “positive” and “negative” results, and progress towards autonomous experimental materials research.