High entropy / muliti-component alloys

High Entropy Alloys (HEAs) are a new class of material,  concocted to have no single base element, instead having many elemental components mixed in near equal atomic proportions. Competing thermodynamic contributions, including a high entropy of mixing, can favour solid solutions with simple crystalline phases over the formation of brittle intermetallic compounds.

This complex thermodynamic balance between the many constituent elements results in a myriad of mixing reactions, all accessible by subtle changes in composition. As such, HEAs can be engineered to form single solid solutions or multiple solid solutions, as well as undergoing spinodal decomposition and nanoprecipitation, which results in changes to the alloy microstructure and functional properties. Control of the alloy microstructure is of particular importance for the fabrication of new soft magnetic alloys, critical for applications within generators and transformers for power generation. This can be used to reduce losses in these machines, crucial for improving the efficiency of the grid alongside enabling green technologies. 

Just as the possibilities of HEAs are endless, so are the ways in which the constituent elements can be combined. This presents an enormous problem, in that the composition parameter space is too big to be explored with traditional fabrication and characterisation techniques. To resolve this, the group has invested in developing high throughput alloy fabrication and characterisation techniques, which enables rapid screening of HEAs. These techniques allow the fabrication and magnetic characterisation of over 180 unique alloy compositions in a single measurement. In this way, promising HEAs can be identified for detailed characterisation, allowing one to home in the regions of the phase diagram that are of interest and understand the chemical contributions that give rise to this.   

In addition, computational data-driven approaches to HEA discovery are being developed. Here, the group is exploring using Machine Learning to predict HEAs with optimal magnetic properties. These algorithms are trained on large datasets of known magnetic alloys and use this to ‘learn’ relationships between alloy compositions and  magnetic properties. With sufficient training data, this could be used to predict the magnetic properties of HEAs, reducing the need for time consuming fabrication and characterisation.