Search

We present a fast and accurate, semi-analytical method for predicting hydrogen adsorption in nanoporous materials. For any temperature and pressure, the adsorbed amount is calculated as an integral over the energy density of adsorption sites (guest-host interactions) plus an average guest-guest term. The guest-host interaction energy is calculated using a classical force field with hydrogen modelled as a single-site probe. The guest-guest interaction energy is approximated using an average coordination number, which is regressed using Gaussian Process Regression (GPR). Local adsorption at each site is then modelled using a Langmuir isotherm, which when weighted with its probability density gives an accurate description of hydrogen adsorption. The method is tested on 933 metal-organic frameworks (MOFs) from the Computation-Ready Experimental (CoRE) MOF database at 77 K from 10 − 5 to 100 bar, and the results are compared against GCMC predictions. To demonstrate the utility of the method, we calculated hydrogen adsorption isotherms for 12,914 existing MOF structures, at two different temperatures at a speed about 100 times that of GCMC simulations and analyzed the results. We found 13 MOFs with predicted deliverable capacities exceeding the DOE target of 50 g/L for adsorption at 100 bar, 77 K and desorption at 5 bar, 160 K. [ABSTRACT FROM AUTHOR]

Hydrogen is an appealing energy storage solution for electric vehicles due to its low environmental impact and faster recharge times compared to batteries. However, there are many engineering challenges involved in safely storing a sufficient amount of hydrogen onboard a vehicle with a reasonable volumetric density. Nanoporous materials such as metal–organic frameworks (MOFs) have the potential to store hydrogen at high density and only moderate pressure. Considerable research has been devoted to finding new MOFs for hydrogen storage in recent years; however, a MOF that provides sufficient hydrogen density and is suitable to commercial applications has not yet been found. Much of this research makes use of molecular modelling to screen thousands of materials in a high-throughput way. Computational screening can be an effective tool for gaining insight into structure-performance relationships as well as finding specific candidates for an application. Recently, some research groups have also used machine learning to analyze data more effectively and accelerate the screening process. In this review, we discuss some recent advances in using molecular modelling and machine learning to find materials for hydrogen storage. We also discuss and compare some popular models for the hydrogen molecule and the accuracy of different equations of state, which are important considerations for accurate molecular simulations. [ABSTRACT FROM AUTHOR]

A new software package, RASPA, for simulating adsorption and diffusion of molecules in flexible nanoporous materials is presented. The code implements the latest state-of-the-art algorithms for molecular dynamics and Monte Carlo (MC) in various ensembles including symplectic/measure-preserving integrators, Ewald summation, configurational-bias MC, continuous fractional component MC, reactive MC and Baker's minimisation. We show example applications of RASPA in computing coexistence properties, adsorption isotherms for single and multiple components, self- and collective diffusivities, reaction systems and visualisation. The software is released under the GNU General Public License. [ABSTRACT FROM PUBLISHER]

Computational high-throughput screening was carried out to assess a large number of experimentally reported metal-organic frameworks (MOFs) and zeolites for their utility in hexane isomer separation. Through the work, we identified many MOFs and zeolites with high selectivity (SL+M > 10) for the group of n-hexane, 2-methylpentane, and 3-methylpentane (linear and monobranched isomers) versus 2,2-dimethylbutane and 2,3-dimethylbutane (dibranched isomers). This group of selective sorbents includes VICDOC (Fe2(BDP)3), a MOF with triangular pores that is known to exhibit high isomer selectivity and capacity. For three of these structures, the adsorption isotherms for a 10-component mixture of hexane and heptane isomers were calculated. Subsequent simulations of column breakthrough curves showed that the DEYVUA MOF exhibits a longer process cycle time than VICDOC MOF or MRE zeolite, which are previously reported, high-performing materials, illustrating the importance of capacity in designing MOFs for practical applications. Among the identified candidates, we synthesized and characterized a MOF in a new copper form with high predicted adsorbent capacity (qL+M > 1.2 mol/L) and moderately high selectivity (SL+M ≈ 10). Finally, we examined the role of pore shape in hexane isomer separations, especially of triangular-shaped pores. We show through the potential energy surface and three-dimensional siting analyses that linear alkanes do not populate the corners of narrow triangular channels and that structures with nontriangular pores can efficiently separate hexane isomers. Detailed thermodynamic analysis illustrates how differences in the free energy of adsorption contribute to shape-selective separation in nanoporous materials. [ABSTRACT FROM AUTHOR]

Efforts to computationally characterize large numbers of nanoporous materials often rely on databases of experimentally resolved crystal structures. The accuracy of experimental crystal structures used in such calculations has a significant impact on the reliability of the results. In this work, we report structures optimized using periodic density functional theory (DFT) for more than 800 experimentally synthesized metal-organic frameworks (MOFs). Many MOFs changed significantly upon structural optimization, particularly materials that were crystallographically resolved in their solvated form. For each MOF, we simulated the adsorption of CH4 and CO2 using grand canonical Monte Carlo both before and after DFT optimization. The DFT optimization has a large impact on simulated gas adsorption in some cases. For example, CO2 loading at 1 bar changed by more than 25% in over 25% of the MOFs we considered. [ABSTRACT FROM AUTHOR]

The methane deliverable capacity of adsorbent materials is a critical performance metric that will determine the viability of using adsorbed natural gas (ANG) technology in vehicular applications. ARPA-E recently set a target deliverable capacity of 315 cc(STP)/cc that a viable adsorbent material should achieve to yield a driving range competitive with incumbent fuels. However, recent computational screening of hundreds of thousands of materials suggests that the target is unattainable. In this work, we aim to determine whether the observed limits in deliverable capacity (~200 cc(STP)/cc) are fundamental limits arising from thermodynamic or material design constraints. Our efforts focus on simulating methane adsorption isotherms in a large number of systems, resulting in a broad exploration of different combinations of spatial distributions and energetics of adsorption sites. All systems were classified into five adsorption scenarios with varying degrees of realism in the manner that adsorption sites are created and endowed with energetics. The scenarios range from methane adsorption on discrete idealized lattice sites to adsorption in metal–organic frameworks with coordinatively unsaturated sites (CUS) provided by metalated catechol groups. Our findings strongly suggest that the ARPA-E target is unattainable, although not due to thermodynamic constraints but due to material design constraints. On the other hand, we also find that the currently observed deliverable capacity limits may be moderately surpassed. For instance, incorporation of CUS in IRMOF-10 is predicted to yield a 217 cc(STP)/cc deliverable capacity. The modified material has a ~0.85 void fraction and a heat of adsorption of ~15 kJ/mol. This suggests that similar, moderate improvements over existing materials could be achieved as long as CUS incorporation still maintains a relatively large void fraction. Nonetheless, we conclude that more significant improvements in deliverable capacity will require changes in the currently proposed operation conditions. [ABSTRACT FROM AUTHOR]

Experimentally refinedcrystal structures for metal–organicframeworks (MOFs) often include solvent molecules and partially occupiedor disordered atoms. This creates a major impediment to applying high-throughputcomputational screening to MOFs. To address this problem, we haveconstructed a database of MOF structures that are derived from experimentaldata but are immediately suitable for molecular simulations. The computation-ready,experimental (CoRE) MOF database contains over 4700 porous structureswith publically available atomic coordinates. Important physical andchemical properties including the surface area and pore dimensionsare reported for these structures. To demonstrate the utility of thedatabase, we performed grand canonical Monte Carlo simulations ofmethane adsorption on all structures in the CoRE MOF database. Weinvestigated the structural properties of the CoRE MOFs that governmethane storage capacity and found that these relationships agreewell with those derived recently from a large database of hypotheticalMOFs. [ABSTRACT FROM AUTHOR]