H2 Adsorption in Metal‐Organic Frameworks: Dispersion or Electrostatic Interactions?

Abstract

Materials to store molecular hydrogen for mobile applications have been intensively studied over the past years. In summary, two storage mechanisms have been proposed: chemisorption (e.g. metal hydrides, aminoboranes), and physisorption in nanoporous materials. In contrast to most hydride storage media, materials physisorbing H2 offer reversible (un)loading processes without intensive external heating or cooling. As H2 is a nonpolar molecule, the two principal contributions to the adsorption energy are weak London (dispersion) interactions (LDI) and interactions due to the electrostatic potential of the host material. LDI depend on the polarisability of the host material and on the distance between H2 and the host surface. Therefore, systems designed for H2 storage should be highly polarisable and have a large specific surface area with favourable pore sizes of ~0.6 nm. Graphitic (sp) carbon structures (graphene slit pores, carbon nanotubes, fullerenes and more advanced materials (C60 intercalated graphite, [11] honeycomb graphite etc.)) belong to this group. However, with none of them the 2010 goal of the US Department of Energy (6 wt.% of stored H2 and 45 gL 1 volumetric density) could be reached for moderate pressure and ambient temperature. Higher H2 adsorption capacities might be possible if attractive electrostatic interactions are introduced by a non-negligible charge separation in the host. One of the most promising materials with these properties are metal-organic frameworks (MOFs, see Figure 1a), a family of nanoporous materials that are built of well-defined building blocks, polar metal oxide centers (connectors) and nonpolar organic linkers containing aromatic carbons. As it is possible to tailor their chemical composition and pore size distribution, many potential applications have been proposed for MOFs, among them H2 storage. It has been shown experimentally that some MOFs show indeed excellent storage capacities for H2. [21,22] It is, however, unclear, which underlying mechanism is responsible for this property. To tune the capability of MOFs to store H2 the fundamental interactions leading to the adsorption have to be well understood. So far, it is not clear which interaction (LDI or electrostatics, for certain connectors possibly even chemisorption) is responsible for the H2 adsorption in MOFs. Experimental evidence emphasizes that the strongest H2 adsorption sites are close to the metal oxide connectors, which is interpreted such that M O (M=Zn, Cu, Mg, etc.) dipoles are most effective in polarizing the gas molecules and lead to strong interactions. There is no consensus in the interpretation of the adsorption mechanism; the quantification of the adsorption energy depends on various variables and is matter of discussion, but lowenergy adsorption sites have been identified in agreement between experiment and theory. It is important to obtain the host–guest potential theoretically, as it cannot be accessed experimentally due to the complex nature of the interaction. Also, the theoretical determination is not straightforward: So far, severe approximations had to be made in all theoretical approaches, and no final conclusion on the interaction mechanism could be drawn from their results: Either, an extended model for the MOF/H2 system was made, but the interaction energy has been calculated using density-functional theory (DFT) which is well-known to fail to describe LDI. The second approach is to reduce the MOF structure to model clusters (MOF connectors and linkers); however, the host–guest interaction is treated at higher computational level, most commonly using MP2 theory, as it is the compu[a] A. Kuc, Prof. G. Seifert Physikalische Chemie, Technische UniversitAt Dresden 01062 Dresden (Germany) [b] A. Kuc, Prof. T. Heine School of Engineering and Science, Jacobs University Bremen 28759 Bremen (Germany) Fax: (+49)421200493223 E-mail : t.heine@jacobs-university.de [c] Prof. H. A. Duarte Departamento de QuGmica–ICEx Universidade Federal de Minas Gerais 31.270-901 Belo Horizonte, MG (Brazil) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.200800878.