Mapping the Porous and Chemical Structure–Function Relationships of Trace CH₃I Capture by Metal–Organic Frameworks using Machine Learning

Abstract

Large-scale computational screening has become an indispensable tool for functional materials discovery. It, however, remains a challenge to adequately interrogate the large amount of data generated by a screening study. Here, we computationally screened 1087 metal−organic frameworks (MOFs), from the CoRE MOF 2014 database, for capturing trace amounts (300 ppmv) of methyl iodide (CH 3 I); as a primary representative of organic iodides, CH 3 129 I is one of the most difficult radioactive contaminants to separate. Furthermore, we demonstrate a simple and general approach for mapping and interrogating the high-dimensional structure−function data obtained by high-throughput screening; this involves learning two-dimensional embeddings of the highdimensional data by applying unsupervised learning to encoded structural and chemical features of MOFs. The resulting various porous and chemical structure−function maps are human-interpretable, revealing not only top-performing MOFs but also complex structure−function correlations that are hidden when inspecting individual MOF features. These maps also alleviate the need of laborious visual inspection of a large number of MOFs by clustering similar MOFs, per the encoding features, into defined regions on the map. We also show that these structure−function maps are amenable to supervised classification of the performances of MOFs for trace CH 3 I capture. We further show that the machine-learning models trained on the 1087 CoRE MOFs can be used to predict an unseen set of 250 MOFs randomly selected from a different MOF database, achieving high prediction accuracies.

Computational screening and functional tuning of chemically stable metal organic frameworks for I2/CH3I capture in humid environments

“…The aforementioned results were well consistent with the previous researches. 47 We picked the several top-performing materials and marked their names in Figure 1 . Zn(1,3-BDP), MIL-53-Al, NOTT-300, Al-PMOF, and JUC-110 presented the well-performing for I 2 adsorption, exhibiting the large uptake amounts of 66.07, 49.85, 47.83, 29.94, and 26.02 cm 3 /g, respectively.…”

Optimal Surrogate Models for Predicting the Elastic Moduli of Metal–Organic Frameworks via Multiscale Features

“…Due to the proficiency of SOAP descriptors in accurately representing the atomic environment, including the type of neighboring atoms and short-range intermolecular interactions, they are well suited for not only predicting static data but also projecting kinetic data. Thus, although SOAP descriptors might be less straightforward than RAC descriptors, , they were selected as local features due to these strengths. The software, parameters, and notation used to extract SOAP descriptors are presented in Section S4.2.…”

“…Similarly, SOAP descriptors provide a numerical representation of the local atomic environment surrounding each atom within a specified distance. By encoding information about chemical bonds and short-range intermolecular interactions (for example, π–π stacking) using spherical harmonics and radial basis functions, SOAP descriptors can effectively capture the local chemical and physical information on an MOF. , Additionally, as the prediction of material performance degradation over timeknown as kinetic databecomes increasingly vital, , it is also essential in the field of MOFs to characterize their time-dependent behaviors . Due to the proficiency of SOAP descriptors in accurately representing the atomic environment, including the type of neighboring atoms and short-range intermolecular interactions, they are well suited for not only predicting static data but also projecting kinetic data.…”

Metal–organic framework clustering through the lens of transfer learning

“…Escobar-Hernandez and coworkers used k -means clustering to evaluate MOF models that deal with thermal stability, 56 Rosen et al employed UMAP to determine quantum properties in MOFs, 57 and Wu et al used UMAP and k -means to condense MOF features into an accessible representation of the space. 58…”

“…Escobar-Hernandez and coworkers used k-means clustering to evaluate MOF models that deal with thermal stability, 56 Rosen et al employed UMAP to determine quantum properties in MOFs, 57 and Wu et al used UMAP and k-means to condense MOF features into an accessible representation of the space. 58 Regarding an accessible representation of the space to use in clustering, we utilize principal component analysis (PCA) to effectively represent and understand the MOF textural space. PCA is a statistical method used to reduce the dimensionality of a dataset while retaining the most important information or patterns present in the data.…”