Materials informatics has the potential to help material scientists explore a much larger part of composition space than what was possible with empirical or computational material science. In this talk, two materials informatics applications are demonstrated: First, based on a traditional Calphad databases for salts (FTsalt) to screen composition space for low-temperature eutectics as phase-change materials. Second, the combination of the Calphad database aiMP (ab initio materialsproject.org) with a traditional Calphad database for pure substances, SGTE Pure Substances, is used to screen for promising coating materials for SiC/SiC tubes which are a promising future cladding material for accident-tolerant fuels in nuclear reactors, but suffer from severe steam oxidation at very high temperature [1]. aiMP is a Calphad database built on ab initio calculated 0 K formation enthalpies from Materials Project [2] and machine learning models to a) make formation enthalpies consistent with existing, experimentally based Calphad databases, and b) describe temperature dependent entropy and heat capacity [3]. It is shown how filters can be applied to analyze hundreds of thousands of equilibrium calculations.

References:

[1] M. Steinbrueck et al., “Oxidation of Silicon Carbide Composites for Nuclear Applications at Very High Temperatures in Steam” Coatings 2022, 12, 875. https://doi.org/10.3390/coatings12070875.

[2] A. Jain et al., “A high-throughput infrastructure for density functional theory

calculations,” Comput. Mater. Sci., vol. 50, no. 8, pp. 2295–2310, Jun. 2011, doi:

10.1016/j.commatsci.2011.02.023.

[3] https://gtt-technologies.de/wp-content/uploads/2022/08/AIMP_AIOQ_v5_Docu.pdf