Hundreds of features used in candidate metal-organic frameworks are ranked and evaluated.
Metal-organic frameworks (MOFs) are a promising candidate for the efficient transport of carbon dioxide (CO) due to their high porosity and surface area.2) adsorption. Because of its importance, it is highly desirable to be able to quickly screen different MOFs for their potential efficiency. Teng and Shan developed a machine learning model to determine the key properties that should be used to screen MOFs.
The duo analyzed the effect of 23 structural and molecular features and 765 calculated features on the model and ranked the importance of different features at different pressures. They found that, regardless of pressure, the molecular structure and pore size of the MOFs were important in improving the accuracy of the predictive models by up to 20 percent.
“High-throughput screening using our machine learning model can reduce the demand for computational resources and minimize the need for expert intervention,” said author Guangcun Shan. “Furthermore, the machine learning model in this study can be used to provide theoretical support for other prediction results and achieve higher prediction accuracy.”
Understand how individual characteristics affect CO2To study adsorption, the authors added sequentially structural, molecular, and computational features to their model. Then, to understand why some features were ranked higher than others, the team applied their expertise in intermolecular forces, secondary bonds, and electrical potentials.
The research team plans to continue investigating how MOFs adsorb CO.2at the atomic level using similar machine learning techniques.
“We aim to utilize other advanced deep learning models, such as graph neural networks, to further assist in screening MOFs with high adsorption capacity,” Shan said.
sauce: “Interpretable Machine Learning for Materials Discovery: Predicting CO2 Adsorption Properties of Metal-Organic Frameworks,” by Yukun Teng and Guangcun Shan, APL Materials (2024). You can find this article here: https://doi.org/10.1063/5.0222154 .
This paper is part of the Emerging Leaders in Materials Science collection. Read more here .