Publications

Ethanol and Water Adsorption in Conventional and Hierarchical All-Silica MFI Zeolites

S.Pahari, M. D. de Mello, M. S. Shah, T. R. Josephson, L. Ren, H.G. T. Nguyen, R. D. Van Zee, M. Tsapatsis, and J. I. Siepmann

ACS Phys. Chem Au 2, 79–88 (2022)

DOI: 10.1021/acsphyschemau.1c00026

Hierarchical zeolites containing both micro- (less than 2 nm) and mesopores (2−50 nm) have gained increasing attention in recent years because they combine the intrinsic properties of conventional zeolites with enhanced mass transport rates due to the presence of mesopores. The structure of the hierarchical self-pillared pentasil (SPP) zeolite is of interest because all-silica SPP consists of orthogonally intergrown single-unit-cell MFI nanosheets and contains hydrophilic surface silanol groups on the mesopore surface while its micropores are nominally hydrophobic. Therefore, the distribution of adsorbed polar molecules, like water and ethanol, in the meso- and micropores is of fundamental interest. Here, molecular simulation and experiment are used to investigate the adsorption of water and ethanol on SPP. Vapor-phase single-component adsorption shows that water occupies preferentially the mesopore corner and surface regions of the SPP material at lower pressures (P/P₀ less than 0.5) while loading in the mesopore interior dominates adsorption at higher pressures. In contrast, ethanol does not exhibit a marked preference for micro- or mesopores at low pressures. Liquid-phase adsorption from binary water−ethanol mixtures demonstrates a 2 orders of magnitude lower ethanol/water selectivity for the SPP material compared to bulk MFI. For very dilute aqueous solutions of ethanol, the ethanol molecules are mostly adsorbed inside the SPP micropore region due to stronger dispersion interactions and the competition from water for the surface silanols. At high ethanol concentrations (C_EtOH greater than 700 g L⁻¹), the SPP material becomes selective for water over ethanol.

Machine learning the quantum-chemical properties of metal–organic frameworks for accelerated materials discovery

A. S. Rosen, S. M. Iyer, D. Ray, Z. Yao, A. Aspuru-Guzik, L. Gagliardi, J. M. Notestein, R. Q.Snurr

Matter 4, 1578-1597 (2021)

DOI: 10.1016/j.matt.2021.02.015

The modular nature of metal–organic frameworks (MOFs) enables synthetic control over their physical and chemical properties, but it can be difficult to know which MOFs would be optimal for a given application. High-throughput computational screening and machine learning are promising routes to efficiently navigate the vast chemical space of MOFs but have rarely been used for the prediction of properties that need to be calculated by quantum mechanical methods. Here, we introduce the Quantum MOF (QMOF) database, a publicly available database of computed quantum-chemical properties for more than 14,000 experimentally synthesized MOFs. Throughout this study, we demonstrate how machine learning models trained on the QMOF database can be used to rapidly discover MOFs with targeted electronic structure properties, using the prediction of theoretically computed band gaps as a representative example. We conclude by highlighting several MOFs predicted to have low band gaps, a challenging task given the electronically insulating nature of most MOFs.

M11plus: A Range-Separated Hybrid Meta Functional with Both Local and Rung-3.5 Correlation Terms and High Across-the-Board Accuracy for Chemical Applications

P. Verma, B. G. Janesko, Y. Wang, X. He, G. Scalmani, M. J. Frisch, and D. G. Truhlar

J. Chem. Theory Comput. 15, 4804-4815 (2019)

DOI: 10.1021/acs.jctc.9b00411

The way to improve Kohn–Sham density functional theory is to improve the exchange–correlation functionals, and functionals have been successively improved by adding new ingredients, especially local spin density gradients, nonlocal Hartree–Fock exchange, and local meta terms based on kinetic energy density. Here, we present a new kind of functional obtained by adding rung-3.5 terms to a functional including local gradients, local meta terms, and range-separated Hartree–Fock exchange. A rung-3.5 term has short-range nonlocality designed to account for nondynamic correlation; we add two kinds of rung-3.5 terms, one kind modeled on position-dependent Hartree–Fock exchange and another modeled on the spin density at a point interacting with the opposite-spin exchange hole at the same point. Optimization of the functional yields broad accuracy for both ground states and excited states with especially significant improvement for systems with strong correlation.

Deep neural network learning of complex binary sorption equilibria from molecular simulation data

Y. Sun, R. F. DeJaco, and J. I. Siepmann

Chem. Sci. 10, 4377-4388 (2019)

DOI: 10.1039/C8SC05340E

We employed deep neural networks (NNs) as an efficient and intelligent surrogate of molecular simulations for complex sorption equilibria using probabilistic modeling. Canonical (N1N2VT) Gibbs ensemble Monte Carlo simulations were performed to model a single-stage equilibrium desorptive drying process for (1,4-butanediol or 1,5-pentanediol)/water and 1,5-pentanediol/ethanol from all-silica MFI zeolite and 1,5-pentanediol/water from all-silica LTA zeolite. A multi-task deep NN was trained on the simulation data to predict equilibrium loadings as a function of thermodynamic state variables.

Multiconfigurational Self-Consistent Field Theory with Density Matrix Embedding: The Localized Active Space Self-Consistent Field Method

M. R. Hermes and L. Gagliardi

J. Chem. Theory Comput. 15, 972–986 (2019)

DOI: 10.1021/acs.jctc.8b01009

We recently encountered evidence that the approximate single-determinantal bath picture inherent to DMET is sometimes problematic when the complete active space self-consistent field (CASSCF) is used as a solver and the method is applied to realistic models of strongly correlated molecules. Here, we show this problem can be defeated by generalizing DMET to use a multiconfigurational wave function as a bath without sacrificing practically attractive features of DMET, such as a second-quantization form of the embedded subsystem Hamiltonian, by dividing the active space into unentangled active subspaces each localized to one fragment. We introduce the term localized active space (LAS) to refer to this kind of wave function.

Energy-based descriptors to rapidly predict hydrogen storage in metal–organic frameworks

B. J. Bucior, N. S. Bobbitt, T. Islamoglu, S. Goswami, A. Gopalan, T. Yildirim, O. K. Farha, N. Bagheri, and R. Q. Snurr

Mol. Syst. Des. Eng. 4, 162-174 (2019)

DOI: 10.1039/C8ME00050F

Since MOFs can be made from many combinations of metal nodes, organic linkers, and functional groups, the design space of possible MOFs is enormous. Experimental characterization of thousands of MOFs is infeasible, and even conventional molecular simulations can be prohibitively expensive for large databases. In this work, we have developed a data-driven approach to accelerate materials screening and learn structure–property relationships. We report new descriptors for gas adsorption in MOFs derived from the energetics of MOF–guest interactions. Using the bins of an energy histogram as features, we trained a sparse regression model to predict gas uptake in multiple MOF databases to an accuracy within 3 g L⁻¹.

First principles Monte Carlo simulations of unary and binary adsorption: CO₂, N₂, and H₂O in Mg-MOF-74

E. O. Fetisov, M. Shah, J. R. Long, M. Tsapatsis, and J. I. Siepmann

Chem. Commun. 54, 10816-108192 (2018)

DOI: 10.1039/C8ME00050F

Dative bonding of adsorbate molecules onto coordinatively-unsaturated metal sites in metal–organic frameworks can lead to unique adsorption selectivities. However, the distortion of the electron density during dative bonding poses a challenge for force-field-based simulations. Here, we report first principles Monte Carlo simulations with the PBE-D3 functional for the adsorption of CO₂, N₂, and H₂O in Mg-MOF-74, and obtain accurate predictions of the unary isotherms without any of the adjustments or fitting often required for systems with strong adsorption sites. Simulations of binary CO₂/N₂ and H₂O/CO₂ mixtures yield selectivities of 200 and 160, respectively, and indicate that predictions from ideal adsorbed solution theory need to be viewed with caution.

Air Separation by Catechol-Ligated Transition Metals: A Quantum Chemical Screening

S. J. Stoneburner and L. Gagliardi

J. Phys. Chem. C 122, 22345–22351 (2018)

DOI: 10.1039/C8ME00050F

The separation of O₂ and N₂ from air is of great importance in a variety of industrial contexts, but the primary means of accomplishing the separation is cryogenic distillation, an energy-intensive process. A material that could enable air separation to occur at conventional temperatures would be of great economic and environmental benefit. Metalated catecholates within metal–organic frameworks have been considered for other gas separations and are shown here to have significant potential for air separation. Calculations of interaction energies between catecholates with first-row transition metals and guests O₂ and N₂ were performed using density functional theory and multireference complete active space self-consistent field followed by second-order perturbation theory. A general recipe is offered for active space selection for metalated catecholate systems. The multireference results are used to rationalize O₂ binding in terms of redox activity with the metalated catecholate. O₂ is predicted to bind more strongly than N₂ for all cases except Cu²⁺, with general agreement in the binding trends among all methods.

C2 Adsorption in Zeolites: In Silico Screening and Sensitivity to Molecular Models

M. S. Shah, E. O. Fetisov, M. Tsapatsis, and J. I. Siepmann

Mol. Syst. Des. Eng. 3, 619-626 (2018)

DOI: 10.1039/C8ME00004B

Efficient separation of ethane and ethylene has been a long-standing challenge for the chemical industry. In this study, we use molecular modeling to identify zeolite and zeotype frameworks that have the potential to be the next-generation solution for the separation of these C2 compounds. Using two different united-atom versions of the transferable potentials for phase equilibria (TraPPE) force field, the zeolitic structures in the database of the International Zeolite Association are screened for the separation of ethane and ethylene. A detailed analysis, with regards to accessibility of favorable sites and sensitivity to molecular models (also considering the explicit-hydrogen TraPPE model for ethane), is carried out on the top-performing structures. This study provides insights on the performance and limitations of molecular models for predicting mixture adsorption in zeolites.

Catechol-Ligated Transition Metals: A Quantum Chemical Study on a Promising System for Gas Separation

S. J. Stoneburner, V. Livermore, M. E. McGreal, D. C. Yu, K. D. Vogiatzis, R. Q. Snurr, and L. Gagliardi

J. Phys. Chem. C 121, 10463-10469 (2017)

DOI: 10.1021/acs.jpcc.7b02685

Metal–organic frameworks (MOFs) have received a great deal of attention for their potential in atmospheric filtering, and recent work has shown that catecholate linkers can bind metals, creating MOFs with monocatecholate metal centers and abundant open coordination sites. In this study, M–catecholate systems (with M = Mg²⁺, Sc²⁺, Ti²⁺, V²⁺, Cr²⁺, Mn²⁺, Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, and Zn²⁺) were used as computational models of metalated catecholate linkers in MOFs. Nitric oxide (NO) is a radical molecule that is considered an environmental pollutant and is toxic if inhaled in large quantities. Binding NO is of interest in creating atmospheric filters, at both the industrial and personal scale. The binding energies of NO to the metal–catecholate systems were calculated using density functional theory (DFT) and complete active space self-consistent field (CASSCF) followed by second-order perturbation theory (CASPT2). Selectivity was studied by calculating the binding energies of additional guests (CO, NH₃, H₂O, N₂, and CO₂). The toxic guests have stronger binding than the benign guests for all metals studied, and NO has significantly stronger binding than other guests for most of the metals studied, suggesting that metal–catecholates are worthy of further study for NO filtration. Certain metal–catecholates also show potential for separation of N₂ and CO₂ via N₂ activation, which could be relevant for carbon capture or ammonia synthesis.

Impact of the Strength and Spatial Distribution of Adsorption Sites on Methane Deliverable Capacity in Nanoporous Materials

D. A. Gómez-Gualdrón, C. M. Simon, W. Lassman, D. Chen, R. L. Martin, M. Haranczyk, O. K. Farha, B. Smit, and R. Q. Snurr

Chem. Eng. Sci. 159, 18-30 (2016)

DOI: 10.1016/j.ces.2016.02.030i

In this work, we address the question of which thermodynamic factors determine the deliverable capacity of methane in nanoporous materials. The deliverable capacity is one of the key factors that determines the performance of a material for methane storage in automotive fuel tanks. To obtain insights into how the molecular characteristics of a material are related to the deliverable capacity, we developed several statistical thermodynamic models. The predictions of these models are compared with the classical thermodynamics approach of Bhatia and Myers [Bhatia and Myers, Langmuir, 2005, 22, 1688] and with the results of molecular simulations in which we screen the International Zeolite Association (IZA) structure database and a hypothetical zeolite database of over 100,000 structures. Both the simulations and our models do not support the rule of thumb that, for methane storage, one should aim for an optimal heat of adsorption of 18.8 kJ mol⁻¹. Instead, our models show that one can identify an optimal heat of adsorption, but that this optimal heat of adsorption depends on the structure of the material and can range from 8 to 23 kJ mol⁻¹. The different models we have developed are aimed to determine how this optimal heat of adsorption is related to the molecular structure of the material.

Identifying Optimal Zeolitic Sorbents for Sweetening of Highly Sour Natural Gas

M. S. Shah, M. Tsapatsis, and J. I. Siepmann

Angew. Chem. Int. Ed. 55, 5938-5942 (2016)

DOI: 10.1002/anie.201600612

Raw natural gas is a complex mixture comprising methane, ethane, other hydrocarbons, hydrogen sulfide, carbon dioxide, nitrogen, and water. For sour gas fields, selective and energy‐efficient removal of H₂S is one of the crucial challenges facing the natural‐gas industry. Separation using nanoporous materials, such as zeolites, can be an alternative to energy‐intensive amine‐based absorption processes. Herein, the adsorption of binary H₂S/CH₄ and H₂S/C₂H₆ mixtures in the all‐silica forms of 386 zeolitic frameworks is investigated using Monte Carlo simulations. Adsorption of a five‐component mixture is utilized to evaluate the performance of the 16 most promising materials under close‐to‐real conditions. It is found that depending on the fractions of CH₄, C₂H₆, and CO₂, different sorbents allow for optimal H₂S removal and hydrocarbon recovery.

The Materials Genome in Action: Identifying the Performance Limits for Methane Storage

C. Simon, J. Kim, D. A. Gomez-Gualdron, J. S. Camp, Y. G. Chung, R. Martin, R. Mercado, M. W. Deem, D. Gunter, M. Haranczyk, D. S. Sholl, R. Q. Snurr, and B. Smit

Energ. Environ. Sci. 8, 1190-1199 (2015)

DOI: 10.1002/anie.201600612

Natural gas, mostly methane, is an attractive replacement for petroleum fuels for automotive vehicles because of its economic and environmental advantages. However, it suffers from a low volumetric energy density, necessitating densification to yield reasonable driving ranges from a full fuel tank. Densification strategies in the market today, liquefaction and compression, require expensive and cumbersome vehicular fuel tanks and refill station infrastructure. If we are able to develop a nanoporous adsorbent material to store natural gas at ambient temperature and moderate pressure, one could envision a simple fuel tank that can be refilled at home. Modern, advanced nanoporous materials are highly tunable, inundating researchers with practically infinite possibilities of materials to synthesize and test for methane storage. The current research is focused on finding among these millions of possible materials one that can be used to store natural gas without using liquefaction or compression processes. In this Perspective, we adopt a computational approach to screen databases of over 650,000 nanoporous material structures. Using this nanoporous materials genome approach, we reveal relationships between structural characteristics and performance, and suggest that it may be difficult, if not impossible, to reach the current Advanced Research Project Agency-Energy (ARPA-E) target for natural gas storage using nanoporous materials.

In Silico Prediction of MOFs with High Deliverable Capacity or Internal Surface Area

Y. Bao, R. L. Martin, M. Haranczyk, and M. W. Deem

Phys. Chem. Chem. Phys. 17, 11962-11973 (2015)

DOI: 10.1039/C5CP00002E

Metal–organic frameworks (MOFs) offer unprecedented atom-scale design and structural tunability, largely due to the vast number of possible organic linkers which can be utilized in their assembly. Exploration of this space of linkers allows identification of ranges of achievable material properties as well as discovery of optimal materials for a given application. Experimental exploration of the linker space has to date been quite limited due to the cost and complexity of synthesis, while high-throughput computational studies have mainly explored MOF materials based on known or readily available linkers. Here an evolutionary algorithm for de novo design of organic linkers for metal–organic frameworks is used to predict MOFs with either high methane deliverable capacity or methane accessible surface area. Known chemical reactions are applied in silico to a population of linkers to discover these MOFs. Through this design strategy, MOF candidates are found in the ten symmetric networks acs, cds, dia, hxg, lvt, nbo, pcu, rhr, sod, and tbo. The correlation between deliverable capacities and surface area is network dependent.

Optimizing Nanoporous Materials for Gas Storage

C. M. Simon, J. Kim, L.-C. Lin, R. L. Martin, M. Haranczyk, and B. Smit

Phys. Chem. Chem. Phys. 16, 5499-5513 (2014)

DOI: 10.1039/C3CP55039G

In this work, we address the question of which thermodynamic factors determine the deliverable capacity of methane in nanoporous materials. The deliverable capacity is one of the key factors that determines the performance of a material for methane storage in automotive fuel tanks. To obtain insights into how the molecular characteristics of a material are related to the deliverable capacity, we developed several statistical thermodynamic models. The predictions of these models are compared with the classical thermodynamics approach of Bhatia and Myers [Bhatia and Myers, Langmuir, 2005, 22, 1688] and with the results of molecular simulations in which we screen the International Zeolite Association (IZA) structure database and a hypothetical zeolite database of over 100 000 structures. Both the simulations and our models do not support the rule of thumb that, for methane storage, one should aim for an optimal heat of adsorption of 18.8 kJ mol−1. Instead, our models show that one can identify an optimal heat of adsorption, but that this optimal heat of adsorption depends on the structure of the material and can range from 8 to 23 kJ mol⁻¹. The different models we have developed are aimed to determine how this optimal heat of adsorption is related to the molecular structure of the material.

Comprehensive Study of Carbon Dioxide Adsorption in the Metal-Organic Frameworks, M₂(dobdc) (M=Mg, Mn, Fe, Co, Ni, Cu, and Zn)

W. L. Queen, M. R. Hudson, E. D. Bloch, J. A. Mason, M. Gonzalez, J. Lee, D. Gygi, J. D. Howe, K. Lee, T. A. Darwish, M. James, V. K. Peterson, S. J. Teat, B. Smit, J. B. Neaton, J. R. Long, and C. M. Brown

Chem. Sci. 5, 4569-4581 (2014)

DOI: 10.1039/C4SC02064B

Analysis of the CO2 adsorption properties of a well-known series of metal–organic frameworks M₂(dobdc) (dobdc4− = 2,5-dioxido-1,4-benzenedicarboxylate; M = Mg, Mn, Fe, Co, Ni, Cu, and Zn) is carried out in tandem with in situ structural studies to identify the host–guest interactions that lead to significant differences in isosteric heats of CO₂ adsorption. Neutron and X-ray powder diffraction and single crystal X-ray diffraction experiments are used to unveil the site-specific binding properties of CO₂ within many of these materials while systematically varying both the amount of CO₂ and the temperature. Unlike previous studies, we show that CO₂ adsorbed at the metal cations exhibits intramolecular angles with minimal deviations from 180°, a finding that indicates a strongly electrostatic and physisorptive interaction with the framework surface and sheds more light on the ongoing discussion regarding whether CO₂ adsorbs in a linear or nonlinear geometry. This has important implications for proposals that have been made to utilize these materials for the activation and chemical conversion of CO₂. For the weaker CO₂ adsorbents, significant elongation of the metal–O(CO₂) distances are observed and diffraction experiments additionally reveal that secondary CO₂ adsorption sites, while likely stabilized by the population of the primary adsorption sites, significantly contribute to adsorption behavior at ambient temperature. Density functional theory calculations including van der Waals dispersion quantitatively corroborate and rationalize observations regarding intramolecular CO₂ angles and trends in relative geometric properties and heats of adsorption in the M₂(dobdc)–CO₂ adducts.

Mail-Order Metal-Organic Frameworks (MOFs): Designing Isoreticular MOF-5 Analogues Comprising Commercially Available Organic Molecules

R. L. Martin, L.-C. Lin, K. Jariwala, B. Smit, and M. Haranczyk

J. Phys. Chem. C 117, 12159-12167 (2013)

DOI: 10.1021/jp401920y

Metal–organic frameworks (MOFs), a class of porous materials, are of particular interest in gas storage and separation applications due largely to their high internal surface areas and tunable structures. MOF-5 is perhaps the archetypal MOF; in particular, many isoreticular analogues of MOF-5 have been synthesized, comprising alternative dicarboxylic acid ligands. In this contribution we introduce a new set of hypothesized MOF-5 analogues, constructed from commercially available organic molecules. We describe our automated procedure for hypothetical MOF design, comprising selection of appropriate ligands, construction of 3D structure models, and structure relaxation methods. 116 MOF-5 analogues were designed and characterized in terms of geometric properties and simulated methane uptake at conditions relevant to vehicular storage applications. A strength of the presented approach is that all of the hypothesized MOFs are designed to be synthesizable utilizing ligands purchasable online.