Predicting the decomposition temperatures of metal-organic frameworks based on the pair distribution functions using a deep learning model

Nhut Huynh,Riley Wagner,Xiang He,Kim-Doang Nguyen

DOI: https://doi.org/10.1016/j.micromeso.2024.113042

IF: 5.876

2024-02-09

Microporous and Mesoporous Materials

Abstract:The thermal stability is one of the factors that challenge the widespread applications of metal–organic frameworks (MOFs). In order to leverage the high surface areas and remarkable porosities of MOFs, it's of critical importance to understand the dependence of the thermal stability of MOFs on the atomic structures, achieve rapid and accurate predictions on the decomposition temperatures of MOFs, and discover MOF structures with excellent thermal robustness. Herein, a deep learning framework was built to predict the thermal stability of MOFs using the pair distribution functions (PDFs) and certain physical properties of MOFs, where feature engineering was carried out with the random forest algorithm to extract the importance of each feature and various approaches, such as transformer and dense layers, were used to unravel the complicated relationships among features. As a result, the deep learning model was demonstrated highly effective in predicting the decomposition temperatures of MOFs with a dominant accuracy of ±10 °C. The outstanding predictive performance of the model primarily stems from the PDF features as confirmed by their overwhelming importance compared to other features, which highlights the significant role of the local structures of MOFs in determining their structural robustness at elevated temperatures.

materials science, multidisciplinary,chemistry, physical,nanoscience & nanotechnology, applied