COST 18234

Modern society in Europe needs a source of energy that is generated without harming the environment.

The efficiency of renewable energy converting devices such as water splitting with electrochemical cells based on nano-scaled oxides relies on a sensible choice of material components. However, larger scale material and device properties such as interface segregation, grain boundary movement, ionic diffusion through porous materials, and mechanical loading also strongly impact performance, making the theoretical simulation of realistic devices a challenging multi-scale problem. Although the scientific community has developed expertise in the individual modelling fields, much less effort has been devoted to integrating and combining the scales toward a multi-scale approach. The ultimate central challenge will be to generate a multiscale modelling platform that will be used world-wide for conducting state-of-the-art multi-scale property prediction of materials.

CIG 18234

Modelling nanoscaled catalysts is an important goal for better utilizing energy and chemicals, but it is a challenging task involving combining several methodologies able to capture the structural and functional complexity. Each methodology can provide a significant amount of valuable data, including material composition, geometrical features, structural architecture, and electronic and chemical capabilities.

By combining the data accumulated by European research experts during our Cost Action with available databases we will develop and train machine learning algorithms that are capable of predicting catalytic performance. As a result, we will provide the scientific and industrial community with chatGPT-type and machine learning tools for identifying the composition and shape of efficient nanocatalysts.