Scientists at the NOMAD Laboratory of the Fritz Haber Institute of the Max Planck Society have recently proposed a workflow that will facilitate the search for novel materials with better properties. They demonstrate the power of the method by identifying more than 50 strong thermally insulating materials. This will help alleviate the ongoing energy crisis, by allowing for more efficient thermoelectric elements, ie, devices capable of converting otherwise wasted heat into useful electrical voltage.
The discovery of new and reliable thermoelectric materials is of utmost importance for the use of more than 40% of the energy provided as waste heat worldwide and helps to alleviate the growing challenges of renewable energy. in the climate. One way to increase the thermoelectric efficiency of a material is to reduce its thermal conductivity, κ, and thereby maintain the temperature gradient required to generate electricity.
However, the cost associated with studying these properties has limited computational and experimental investigations of κ to only a minute subset of all possible materials. A team at NOMAD Laboratory recently made efforts to reduce these costs by creating an AI-guided workflow that hierarchically screens out materials to efficiently find new and better thermal insulators. .
The work was recently published in npj Computational Materials proposes a new way to use Artificial Intelligence (AI) to guide high-throughput searches for new materials. Instead of using physical/chemical intuition to screen materials based on general, known or suspected trends, the new method learns the conditions that lead to desired results in advanced method of AI. This work has the potential to quantify the search for new energy materials and increase the efficiency of these searches.
The first step in designing these workflows is to use advanced statistical and AI methods to estimate the target property of interest, κ in this case. For this purpose, the sure-independence screening and sparsifying operator (SISSO) method is used. SISSO is a machine learning method that reveals fundamental dependencies between different material properties from a set of billions of possible expressions.
Compared to other “black-box” AI models, this approach is equally accurate, but additionally provides analytic relationships between different material properties. This allows us to use modern feature importance criteria to determine which material properties are most important. In the case of κ, it is the molar volume, Vm; the limit of high temperature Debye Temperature, θD, ∞; and the anharmonicity metric factor, σA.
In addition, the described statistical analysis allows the elimination of rules for individual parts that enable a priori estimation of the potential of the material to be a thermal insulator. Working with the three most important main features therefore allows the creation of AI-guided computational pathways for the discovery of new thermal insulators.
These workflows use state-of-the-art electronic structure programs to calculate each of the selected features. During each step the materials are evaluated as poor insulators based on their V valuesmiD, ∞and σA. With this, it is possible to reduce the number of calculations required to find thermally insulating materials by more than two orders of magnitude.
In this work, it is shown by identifying 96 thermal insulators (κ < 10 Wm-1K-1) of the initial set of 732 materials. The reliability of this method was further verified by calculating κ for 4 of these predictions with the highest possible accuracy.
Besides facilitating the active search for new thermoelectric materials, the formalisms proposed by the NOMAD team can also be used to solve other pressing problems in material science.
Thomas AR Purcell et al, Accelerating materials-space exploration for thermal insulators by mapping materials properties through artificial intelligence, npj Computational Materials (2023). DOI: 10.1038/s41524-023-01063-y
Provided by the Max Planck Society
Citation: Efficient discovery of energy-enhanced materials through a new AI-guided workflow (2023, July 18) retrieved 18 July 2023 from https://phys.org/news/2023-07-efficient-discovery -energy-materials-ai-guided .html
This document is subject to copyright. Except for any fair dealing for the purpose of private study or research, no part may be reproduced without written permission. Content is provided for informational purposes only.