Lithium-doped metal-organic frameworks for reversible H2 storage at ambient temperature.

The use of hydrogen as a future energy carrier is an essential part of future energy strategies; however, application as an energy carrier in vehicles and portable electronics is limited by problems with hydrogen storage.1 Two major strategies being pursued toward hydrogen storage for fuel cell applications are dissociative adsorption (DA) and associated adsorption (AA) of hydrogen. DA is the mode for metal alloys, which dissolve the H atoms separately in the matrix breaking the H-H bond. These systems tend to suffer from large barriers in dissociating the H-H bond in storing the hydrogen and large barriers in reassociating the H atoms to desorb H2 for input to the fuel cell. In contrast AA binds the H2 as a molecule, reducing the rate problems with adsorbing or desorbing. But here the challenge has been to obtain sufficiently strong bonding to molecular H2 to achieve the target of 6.0 wt % H2 near room temperature with pressures e100 bar.2 A recent major advance in AA systems is the development of metal-organic frameworks (MOFs), which are crystalline materials composed of metal oxide and organic units.3 At 60 bar and 77 K, IRMOF-1 stores 5.0 wt % of H2 and MOF-177 stores 7.5 wt %.3h However, their H2 uptake capability decreases dramatically near room temperature to ∼0.5 wt %, far too low for practical use. Although the H2 storage capability at room temperature can be increased to 1.8 wt % at 298 K and 100 bar by hydrogen spillover techniques,3i the current materials do not meet the 2010 DOE (Department of Energy) criteria for use in transportation (6.0 wt % in the temperature range of -30 to 80 °C).2 We report here that Li-doped MOFs significantly improve H2 uptake capacity near ambient conditions. Indeed, we predict that at -30 °C and 100 bar the Li-MOF-C30 leads gravimetric H2 uptake of 6.0 wt %, reaching (barely) the 2010 DOE target. Figure 1 illustrates the metal (Zn) oxide secondary building unit and the organic carboxylate links of the MOFs investigated here. We considered systems based on a cubic lattice with optimized lattice parameters of 26.025 for MOF-C6, 30.252 for MOF-C10, 34.374 for MOF-C16, 38.652 for MOF-C22, and 42.824 Å for MOF-C30, which are in good agreement (within 0.7%) of available experimental lattice parameters.3h Starting with these structures, we used quantum mechanics (QM) calculations (X3LYP flavor of DFT)4 to predict the structure for Li atoms bound to aromatic organic linkers with up to nine fused rings. We find that Li atoms prefer to bind at the centers of the hexagonal aromatic rings, but Li atoms on adjacent aromatic rings are on opposite sides. To predict the strength of binding H2 to these structures we used the results for QM calculations [high quality second-order Møller-Plesset (MP2) at the quadruple-ú QZVPP and triple-ú TZVPP basis sets]5 to calculate the van der Waals interaction between H2 and the metal-oxide clusters and between H2 and the organic linkers. These QM results were used to determine the nonbond H-C, H-O, and H-Zn interactions in the final force field (FF). The FF leads to structures and energies in very good agreement with the QM data. The FF parameters and comparison to QM are in the Supporting Information (SI). With this first-principles-derived FF, we used grand canonical ensemble Monte Carlo (GCMC) simulations6 to calculate the H2 uptake behavior of the Li-MOFs. This determines the equilibrium loading of H2 as a function of pressures and temperature, as shown in Figure 2. To eliminate boundary effects, we use an infinite threedimensionally periodic cell containing four independent sheets each with 32 Zn atoms. Additional calculation details are included in the SI. Our simulations show that the H2 uptake of the MOF-C6 at 77 K and 1 bar is 1.28 wt % which compares well with the experimental results of 1.32 wt %, while for MOF-C10 at 77 K and 1 bar we calculate 1.62 wt % which compares well with experiment, 1.50 wt %.3e We calculate that MOF-C6 has 4.17 wt % at a pressure of 20 bar and 77 K (experimental value of ∼4.6)3h and 4.89 wt % at a pressure of 50 bar and 77 K (experimental result of ∼5.0 wt %).3h In addition, at 300 K our simulation predicts that MOF-C6 has 0.35 wt % at 60 bar (experimental value of 0.45 wt %)3j and MOF-C10 has 0.3 wt % at 30 bar (experimental value Figure 1. Li-doped MOFs. In each case the Zn4O(CO2)6 connector couples to six aromatic linkers through the O-C-O common to each linker. The MOFs are named according to the number of aromatic carbon atoms. The large violet atoms in the linkers represent Li atoms above the linkers while small violet Li atoms lie below the linkers. The CxLi ratio considers only aromatic carbon atoms.