Search for: All records

With the growing emphasis on sustainability, criticality, and availability in materials research, providing actionable information about mineral commodities is crucial for informed decision-making and strategic planning by researchers, policy makers, and industry stakeholders. While the United States Geological Survey (USGS) offers valuable information on mineral-commodity summaries, their unstructured nature makes analysis challenging. To address this, we present a comprehensive data-analytics application () that processes the past 10 years of USGS mineral-commodity summaries into actionable insights. The application offers country-specific insights into global elemental production and reserves, along with quantitative metrics such as the Herfindahl-Hirschman index (HHI) to evaluate market concentration, identifying risks and opportunities in resource availability. It also features an artificial-intelligence assistant powered by a large language model (LLM) and a retrieval–augmented generation (RAG) system, enabling users to query various aspects of raw materials, including reserves, production, market share, usage, price, substitutes, recycling, and more. We evaluated multiple open-source LLMs for the RAG task and selected the best-performing model, , to implement in the system. This application provides valuable support for material scientists in assessing sustainability, criticality, and market risks, thereby aiding in the development of new materials. We demonstrate its application in energy materials, and by describing the application architecture and providing open access to the code, we aim to enable data-driven advancements in materials research. 

Advancements in materials discovery tend to rely disproportionately on happenstance and luck rather than employing a systematic approach. Recently, advances in computational power have allowed researchers to build computer models to predict the material properties of any chemical formula. From energy minimization techniques to machine learning-based models, these algorithms have unique strengths and weaknesses. However, a computational model is only as good as its accuracy when compared to real-world measurements. In this work, we take two recommendations from a thermoelectric machine learning model, TaVO[Formula: see text] and GdTaO[Formula: see text], and measure their thermoelectric properties of Seebeck coefficient, thermal conductivity, and electrical conductivity. We see that the predictions are mixed; thermal conductivities are correctly predicted, while electrical conductivities and Seebeck coefficients are not. Furthermore, we explore TaVO[Formula: see text]’s unusually low thermal conductivity of 1.2 Wm[Formula: see text]K[Formula: see text], and we discover a possible new avenue of research of a low thermal conductivity oxide family. 

Prompted by limited available data, we explore data-aggregation strategies for material datasets, aiming to boost machine learning performance. Our findings suggest that intuitive aggregation schemes are ineffective in enhancing predictive accuracy. 

Diffusion Models outperform Generative Adversarial Networks (GANs) and Wasserstein GANs in material discovery.