High entropy alloys for thermal, electronic and optical applications

New semiconductor materials are needed in many areas within renewable energy and energy saving as well as in electronics. One group of materials has so far evaded attention as potential semiconductors: high entropy alloys (HEA). They consist of at least five different metallic elements in similar proportions, adopt simple structures, and can show unusual and promising electronic properties. However, until now, the emphasis has been placed on structural applications due to the strength, corrosion resistance, and stability of HEA. The goal of HEATER was to identify novel semiconductors based on HEA. The methodology employed was based on a systematic high-throughput search of a large number of existing and hypothetical HEA compositions for properties relevant to semiconductor technologies. This space is huge; not only is the number of combinations of elements large, but the freedom to vary the amount of each element around the equimolar amount (e.g. around 20% for five elements) makes the number of possibilities explode. This called for a systematic, high-throughput approach where the properties of numerous compositions were investigated simultaneously. In HEATER this was done both experimentally and theoretically. Multi-source magnetron sputtering generated graded films with thousands of compositions represented in a single disk. A series of high-throughput characterization techniques were employed to map out the physical properties along the composition ranges. The screening was taken further with a machine learning concept, where intelligent data analysis was used to predict the crystal symmetry of new compounds. Vast composition spaces were probed including oxides, nitrides, silicides, and antimonides. The major challenge of the screening process was to ensure phase stability with most of the compositions corresponding to multiphase microstructures. Through stabilization strategies, several new semiconductor phases completely unknown to science have been produced and characterized in terms of transport properties. The new screening strategies and machine learning algorithms have been implemented at the partners. These tools are extremely important to future research and are expected to impact the whole scientific field.

HEATER explored high-entropy alloys and high-entropy compounds, such as oxides, nitrides, silicides and antimonides. Fundamental knowledge on complex alloys and compounds was acquired by the consortium as documented by the many communications and publications that resulted from the project. The work developed in the context of HEATER on the educated exploration of these materials and on high-throughput synthesis and characterization led to important research spin-offs such as the projects ANSWER (RCN 280545), Magnificent (RCN 287979), Allotherm (RCN 314778) and START (EU 101058632). Significant scientific outcomes of HEATER include: • Demonstration of entropy stabilization of several compounds. • New and unexpected crystal structures adopted by high-entropy compounds. • (Meta)stabilization of multiphase compositions through amorphization. • Identification of new semiconductors such as Fe10Co8Ni8Cu14Ge11O50 and Zn25Fe14Co12Ni8Mn8Sb33. • Some high entropy compounds showed exotic fundamental behavior, such as inversion from p-type to n-type conduction with temperature and exchange bias at room temperature. The HEATER project allowed the groups involved at SINTEF, UiO and NTNU to strengthen their knowledge in solid-state physics of high entropy alloys and compounds. In addition, a series of theoretical tools and experimental protocols have been developed, serving as the basis for future research. The research concept and its results were extensively disseminated during the course of the project as attested by the scientific communications and publications added to the Cristin repository. Notably, several popular science dissemination activities and one master thesis were also carried out in the framework of the project. The consortium expects to continue publishing results that stem from HEATER. In addition, a workshop with industrial stakeholders is planned for February 2024 from which new applied research is expected to arise. This workshop will be held jointly with other projects on high-entropy materials that are also led by SINTEF, such as ANSWER (RCN 280545), Magnificent (RCN 287979) and Allotherm (RCN 314778).

New seminconductor materials are required in many areas within renewable energy and energy saving as well as in electronics. One group of materials has so far avoided attention as potential semiconductors: high entropy alloys (HEA). They typically consist of at least five different metallic elements in approximately equimolar amounts and adopt simple crystal structures. HEA can exhibit unusual and promising properties for many applications, but up to now the main emphasis has been to use them as structural materials due to their strength, corrosion resistance and/or stability. The main idea of the present project is to initiate a systematic search for properties that are relevant for semiconductor-based technologies within the space of hypothetical HEA compositions. This space is enormous: not only is the number of combinatorial possibilities huge, but the freedom to vary the amount of each element around the equimolar quantity (e.g. 20% for five elements) explodes the number of potential compositions. This calls for a systematic, high-throughput approach where the properties of a large number of compositions can be investigated simultaneously. The HEATER project will do this both experimentally and theoretically. Multi-source magnetron sputtering will generate graded films with thousands of compositions represented in a single disk. A range of high-throughput characterization techniques will be used to map out several properties for composition ranges. Simultaneously, electronic scale modelling will be used to complement the experimental work using a host of state-of-the-art techniques. The optimization will be taken further with a machine learning concept, where intelligent data analysis is used to parameterize models that can predict compositions completely different from the usual suspects. If successful, this project will give a tremendous contribution to materials technology as well as to fundamental materials science.