Using a new database of electronic structure properties derived from highthroughput density functional theory calculations for thousands of metal-organic frameworks (MOFs), we benchmark a variety of machine learning models to accurately and rapidly predict MOF band gaps. Unsupervised dimensionality reduction techniques are also used to map the MOF feature space and identify otherwise subtle structure-property relationships. We anticipate that machine learning models derived from this new database will accelerate the discovery of promising MOFs with targeted quantum-chemical properties.

In this work, we leverage advances in computational screening based on periodic density functional theory (DFT) to study a diverse set of experimentally derived metal−organic frameworks (MOFs) with accessible metal sites for the oxidative activation of methane. We find that the thermodynamic favorability of forming the metal-oxo active site has a strong, inverse correlation with the reactivity toward C−H bond activation for a wide range of MOFs. This scaling relationship is found to hold over MOFs with varying coordination environments and metal compositions, provided the bonds of the framework atoms are conserved. The need to conserve bonds is an important constraint on the correlations but also demonstrates a route to intentionally break the scaling relationship to generate novel catalytic reactivity. Periodic trends are also observed across the data set of screened MOFs, with later transition metals forming less stable but more reactive metal-oxo active sites. Collectively, the results in this work provide robust rules-ofthumb for choosing MOFs to investigate for the activation of methane at moderate reaction conditions.

The modular nature of metal−organic frameworks (MOFs) leads to a very large number of possible structures. Highthroughput computational screening has led to a rapid increase in property data that has enabled several potential applications for MOFs, including gas storage, separations, catalysis, and other fields. Despite their rich chemistry, MOFs are typically named using an ad hoc approach, which can impede their searchability and the discovery of broad insights. In this article, we develop two systematic MOF identifiers, coined MOFid and MOFkey, and algorithms for deconstructing MOFs into their building blocks and underlying topological network. We review existing cheminformatics formats for small molecules and address the challenges of adapting them to periodic crystal structures. Our algorithms are distributed as open-source software, and we apply them here to extract insights from several MOF databases. Through the process of designing MOFid and MOFkey, we provide a perspective on opportunities for the community to facilitate data reuse, improve searchability, and rapidly apply cheminformatics analyses.

Metal–organic frameworks (MOFs) are a class of nanoporous materials with highly tunable structures in terms of both chemical composition and topology. Due to their tunable nature, high‐throughput computational screening is a particularly appealing method to reduce the time‐to‐discovery of MOFs with desirable physical and chemical properties. In this work, a fully automated, high‐throughput periodic density functional theory (DFT) workflow for screening promising MOF candidates was developed and benchmarked, with a specific focus on applications in catalysis. As a proof‐of‐concept, we use the high‐throughput workflow to screen MOFs containing open metal sites (OMSs) from the Computation‐Ready, Experimental MOF database for the oxidative C—H bond activation of methane. The results from the screening process suggest that, despite the strong C—H bond strength of methane, the main challenge from a screening standpoint is the identification of MOFs with OMSs that can be readily oxidized at moderate reaction conditions. © 2019 Wiley Periodicals, Inc.

Metal–organic frameworks (MOFs) with coordinatively unsaturated metal sites are appealing as adsorbent materials due to their tunable functionality and ability to selectively bind small molecules. Through the use of computational screening methods based on periodic density functional theory, we investigate O2 and N2 adsorption at the coordinatively unsaturated metal sites of several MOF families. A variety of design handles are identified that can be used to modify the redox activity of the metal centers, including changing the functionalization of the linkers (replacing oxido donors with sulfido donors), anion exchange of bridging ligands (considering μ-Br–, μ-Cl–, μ-F–, μ-SH–, or μ-OH– groups), and altering the formal oxidation state of the metal. As a result, we show that it is possible to tune the O2 affinity at the open metal sites of MOFs for applications involving the strong and/or selective binding of O2. In contrast with O2 adsorption, N2 adsorption at open metal sites is predicted to be relatively weak across the MOF dataset, with the exception of MOFs containing synthetically elusive V2+ open metal sites. As one example from the screening study, we predicted that exchanging the μ-Cl– ligands of M2Cl2(BBTA) (H2BBTA = 1H,5H-benzo­(1,2-d:4,5-d′)­bistriazole) with μ-OH– groups would significantly enhance the strength of O2 adsorption at the open metal sites without a corresponding increase in the N2 affinity. Experimental investigation of Co2Cl2(BBTA) and Co2(OH)2(BBTA) confirms that the former exhibits weak physisorption of both N2 and O2, whereas the latter is capable of chemisorbing O2 at room temperature in a highly selective manner. The O2 chemisorption behavior is attributed to the greater electron-donating character of the μ-OH– ligands and the presence of H-bonding interactions between the μ-OH– bridging ligands and the reduced O2 adsorbate.