Designing complex concentrated alloys with quantum machine learning and language modeling

Designing novel complex concentrated alloys (CCAs) is an essential topic in materials science. However, due to the complicated high-dimensional component-property relationship, tuning material properties by researchers’ experience is challenging, even when guided by physical or empirical rules. Here, we adopt quantum computing (QC) technology and machine learning models to provide a proof-of-concept application of QC in physical metallurgy. We propose a quantum support vector machine (QSVM) model to predict single-phase CCAs. We show that fine-tuned quantum kernels with entanglement deliver promising performance, with a maximum accuracy of 89.4%. The QSVM model is then used to identify 1,741 lightweight CCAs jointly with a new text-mining-based method. Meanwhile, we devise a controllable approach to study the effect of noise on model performance and find that the noise level needs to be minimized for high-performance QSVM models. This study provides a practical and general approach to designing CCAs based on quantum technologies. [Display omitted] •We designed about 1,700 complex concentrated alloys using two machine learning methods•We provide a proof-of-concept application of quantum machine learning in alloy design•Our fine-tuned quantum machine learning model achieved an accuracy of 89.4%•We devise an approach to study the effect of noise on quantum machine learning models Complex concentrated alloys are an important class of materials with excellent mechanical, functional, and other properties. Due to the huge compositional space, designing complex concentrated solid solutions is challenging. Here, we combine quantum machine learning and language modeling to provide a proof-of-concept example for designing such alloys. In addition, noise in hardware is well known to affect the quality of quantum computing and quantum machine learning, but the study of data noise is insufficient. We devise an approach to evaluating the influence of data noise in quantum machine learning. We combine two different types of machine learning models to design complex concentrated alloys. Our target is to predict single-phase solid solutions. The quantum machine learning model predicts the likeliness of an alloy as a solid solution. Then, the language model provides input to calculate the “context similarity” of chemical elements for further screening. Eventually, our work provides a list of alloy candidates based on a few criteria.