Engineering topics
Snurr, Randall Q.
Publications and source records attributed to Snurr, Randall Q..
Two-Dimensional Energy Histograms as Features for Machine Learning to Predict Adsorption in Diverse Nanoporous Materials
A major obstacle for machine learning (ML) in chemical science is the lack of physically informed feature representations that provide both accurate prediction and easy interpretability of the ML model. In this work, we describe adsorption systems using novel two-dimensional energy histogram (2D-EH) features, which are obtained from the probe-adsorbent energies and energy gradients at grid points located throughout the adsorbent. The 2D-EH features encode both energetic and structural information of the material and lead to highly accurate ML models (coefficient of determination R2 ~ 0.94–0.99) for predicting single-component adsorption capacity in metal–organic frameworks (MOFs). Here, we consider the adsorption of spherical molecules (Kr and Xe), linear alkanes with a wide range of aspect ratios (ethane, propane, n-butane, and n-hexane), and a branched alkane (2,2-dimethylbutane) over a wide range of temperatures and pressures. The interpretable 2D-EH features enable the ML model to learn the basic physics of adsorption in pores from the training data. We show that these MOF-data-trained ML models are transferrable to different families of amorphous nanoporous materials. We also identify several adsorption systems where capillary condensation occurs, and ML predictions are more challenging. Nevertheless, our 2D-EH features still outperform structural features including those derived from persistent homology. The novel 2D-EH features may help accelerate the discovery and design of advanced nanoporous materials using ML for gas storage and separation in the future.
MOFX-DB: An Online Database of Computational Adsorption Data for Nanoporous Materials
Machine learning and data mining coupled with molecular modeling have become powerful tools for materials discovery. Metal-organic frameworks (MOFs) are a rich area for this due to their modular construction and numerous applications. Here, we make data from several previous large-scale studies in MOFs and zeolites from our groups (and new data for N 2 and Ar adsorption in MOFs) easily accessible in one place. The database includes over 3 million simulated adsorption data points for H 2 , CH 4 , CO 2 , Xe, Kr, Ar, and N 2 in over 160 000 MOFs and zeolites, textural properties like pore sizes and surface areas, and the structure file for each material. We include metadata about the Monte Carlo simulations to enable reproducibility. The database is searchable by MOF properties, and the data are stored in a standardized JSON format that that is interoperable with the NIST adsorption database. We also identify several MOFs that meet high performance targets for multiple applications, such as high storage capacity for both hydrogen and methane or high CO 2 capacity plus good Xe/Kr selectivity. Here, by providing this data publicly, we hope to facilitate machine learning studies on these materials, leading to new insights on adsorption in MOFs and zeolites.
Insights and Heuristics for Predicting Diffusion Rates of Chemical Warfare Agents in Zirconium Metal–Organic Frameworks
Not provided.
Effect of Composition and Local Environment on CO 2 Adsorption on Nickel and Magnesium Oxide Solid Solutions
CO 2 adsorption energies on the (100) surfaces of a nickel oxide doped with Mg, magnesium oxide doped with Ni, and their 50:50 solid solution were calculated using density functional theory. The composition and atomic arrangement of the adsorption site was varied to understand how the local environment affects CO 2 adsorption and the basicity of the surfaces at the atomic level. The dispersive and electronic contributions to the adsorption energies were quantified, and the results indicate that the variation of the adsorption energy with adsorption site configuration and metal composition is dominated by electronic interactions. Interestingly, for magnesium oxide doped with nickel, a single substitution can create stronger CO 2 binding sites, which implies stronger basic sites, even though nickel oxide is less basic than magnesium oxide. The effect of double substitution at the binding site can be reasonably approximated by summing the effects of single substitutions. Furthermore, this work provides guidance for the preparation of metal oxides with tailored Lewis basicity.
Computational Investigation of Metal Oxides as Candidate Hydrogen Storage Materials
Not provided.
MOFX-DB: An online database of computational adsorption data for nanoporous materials.
Abstract not provided.
Interfacial Unit-Dependent Catalytic Activity for CO Oxidation over Cerium Oxysulfate Cluster Assemblies
Atomically precise cerium oxo clusters offer a platform to investigate structure–property relationships that are much more complex in the ill-defined bulk material cerium dioxide. We investigated the activity of the MCe 70 torus family (M = Cd, Ce, Co, Cu, Fe, Ni, and Zn), a family of discrete oxysulfate-based Ce 70 rings linked by monomeric cation units, for CO oxidation. CuCe 70 emerged as the best performing MCe 70 catalyst among those tested, prompting our exploration of the role of the interfacial unit on catalytic activity. Temperature-programmed reduction (TPR) studies of the catalysts indicated a lower temperature reduction in CuCe 70 as compared to CeCe 70 . In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) indicated that CuCe 70 exhibited a faster formation of Ce 3+ and contained CO bridging sites absent in CeCe 70 . Isothermal CO adsorption measurements demonstrated a greater uptake of CO by CuCe 70 as compared to CeCe 70 . The calculated energies for the formation of a single oxygen defect in the structure significantly decreased with the presence of Cu at the linkage site as opposed to Ce. Furthermore, this study revealed that atomic-level changes in the interfacial unit can change the reducibility, CO binding/uptake, and oxygen vacancy defect formation energetics in the MCe 70 family to thus tune their catalytic activity.
How Reproducible are Surface Areas Calculated from the BET Equation?
Abstract Porosity and surface area analysis play a prominent role in modern materials science. At the heart of this sits the Brunauer–Emmett–Teller (BET) theory, which has been a remarkably successful contribution to the field of materials science. The BET method was developed in the 1930s for open surfaces but is now the most widely used metric for the estimation of surface areas of micro‐ and mesoporous materials. Despite its widespread use, the calculation of BET surface areas causes a spread in reported areas, resulting in reproducibility problems in both academia and industry. To prove this, for this analysis, 18 already‐measured raw adsorption isotherms were provided to sixty‐one labs, who were asked to calculate the corresponding BET areas. This round‐robin exercise resulted in a wide range of values. Here, the reproducibility of BET area determination from identical isotherms is demonstrated to be a largely ignored issue, raising critical concerns over the reliability of reported BET areas. To solve this major issue, a new computational approach to accurately and systematically determine the BET area of nanoporous materials is developed. The software, called “BET surface identification” (BETSI), expands on the well‐known Rouquerol criteria and makes an unambiguous BET area assignment possible.
Exploring mechanistic routes for light alkane oxidation with an iron–triazolate metal–organic framework
In this work, we computationally explore the formation and subsequent reactivity of various iron-oxo species in the iron–triazolate framework Fe 2 (m-OH) 2 (bbta) (H2bbta = 1H,5H-benzo(1,2-d:4,5- d0)bistriazole) for the catalytic activation of strong C–H bonds. With the direct conversion of methane to methanol as the probe reaction of interest, we use density functional theory (DFT) calculations to evaluate multiple mechanistic pathways in the presence of either N 2 O or H 2 O 2 oxidants. These calculations reveal that a wide range of transition metal-oxo sites – both terminal and bridging – are plausible in this family of metal–organic frameworks, making it a unique platform for comparing the electronic structure and reactivity of different proposed active site motifs. Based on the DFT calculations, we predict that Fe 2 (m-OH)2(bbta) would exhibit a relatively low barrier for N 2 O activation and energetically favorable formation of an [Fe(O)] 2+ species that is capable of oxidizing C–H bonds. In contrast, the use of H 2 O 2 as the oxidant is predicted to yield an assortment of bridging iron-oxo sites that are less reactive. We also find that abstracting oxo ligands can exhibit a complex mixture of both positive and negative spin density, which may have broader implications for relating the degree of radical character to catalytic activity. In general, we consider the coordinatively unsaturated iron sites to be promising for oxidation catalysis, and we provide several recommendations on how to further tune the catalytic properties of this family of metal–triazolate frameworks.
In silico design of microporous polymers for chemical separations and storage
Polymers of intrinsic microporosity (PIMs) are a family of materials with potential to be effective and scalable solutions for challenging adsorbent and membrane applications. The broad range of repeat unit chemistry, microporous structural features, and polymer processing makes exploration of the expansive PIM design space inefficient via chemical and materials intuition alone. Computational techniques such as molecular simulations and machine learning can provide a leap in capabilities to address this polymer design challenge and will be central to the future development of PIMs. In this work, we highlight recent microporous material studies that arrived at key results by employing computational techniques and provide our perspective on the prospects for in silico design and development of PIMs.
Realizing the data-driven, computational discovery of metal-organic framework catalysts
Metal-organic frameworks (MOFs) have been widely investigated for challenging catalytic transformations due to their well-defined structures and high degree of synthetic tunability. These features, at least in principle, make MOFs ideally suited for a computational approach towards catalyst design and discovery. Nonetheless, the widespread use of data science and machine learning to accelerate the discovery of MOF catalysts has yet to be substantially realized. In this review, we provide an overview of recent work that sets the stage for future high-throughput computational screening and machine learning studies involving MOF catalysts. This is followed by a discussion of several challenges currently facing the broad adoption of data-centric approaches in MOF computational catalysis, and we share possible solutions that can help propel the field forward.
Applying design principles to improve hydrogen storage capacity in nanoporous materials
Hydrogen is an attractive option for energy storage because it can be produced from renewable sources and produces environmentally benign byproducts. However, the volumetric energy density of molecular hydrogen at ambient conditions is low compared to other storage methods like batteries, so it must be compressed to attain a viable energy density for applications such as transportation. Nanoporous materials have attracted significant interest for gas storage because they can attain high storage density at lower pressure than conventional compression. Here, we examine how to improve the cryogenic hydrogen storage capacity of a series of porous aromatic frameworks (PAFs) by controlling the pore size and increasing the surface area by adding functional groups. We also explore tradeoffs in gravimetric and volumetric measures of the hydrogen storage capacity and the effects of temperature swings using grand canonical Monte Carlo simulations. We also consider the effects of adding functional groups to the metal–organic framework NU-1000 to improve its hydrogen storage capacity. We find that highly flexible alkane chains do not improve the hydrogen storage capacity in NU-1000 because they do not extend into the pores; however, rigid chains containing alkyne groups do increase the surface area and hydrogen storage capacity. Finally, we demonstrate that the deliverable capacity of hydrogen in NU-1000 can be increased from 40.0 to 45.3 g/L (at storage conditions of 100 bar and 77 K and desorption conditions of 5 bar and 160 K) by adding long, rigid alkyne chains into the pores.
Nanoconfinement and mass transport in metal–organic frameworks
The ubiquity of metal–organic frameworks in recent scientific literature underscores their highly versatile nature. MOFs have been developed for use in a wide array of applications, including: sensors, catalysis, separations, drug delivery, and electrochemical processes. Often overlooked in the discussion of MOF-based materials is the mass transport of guest molecules within the pores and channels. Given the wide distribution of pore sizes, linker functionalization, and crystal sizes, molecular diffusion within MOFs can be highly dependent on the MOF-guest system. In this paper, we discuss the major factors that govern the mass transport of molecules through MOFs at both the intracrystalline and intercrystalline scale; provide an overview of the experimental and computational methods used to measure guest diffusivity within MOFs; and highlight the relevance of mass transfer in the applications of MOFs in electrochemical systems, separations, and heterogeneous catalysis.