Selective guest docking in metal-organic framework materials.

The 1990s saw a quest towards synthesizing modular porous materials from metal-containing linker groups and organic molecule struts. The idea, perhaps in analogy to the spools and sticks of the popular “Tinkertoy” sets, was to use molecular linkers with multiple chelating sites and molecular struts of different lengths to generate porous solid materials with custom-designed pore sizes and shapes. In 1998 Omar M. Yaghi, currently at UCLA, succeeded in developing a systematic route towards synthesizing metal–organic framework (MOF) materials. Since then, MOFs with many metal ion/metal oxide based linkers and a wide variety of bidentate (or multidentate) organic struts have been synthesized. The MOFs are structurally robust, withstanding relatively high temperatures and pressures. They have found many applications in gas-phase separations and gas storage, prominently in H2 storage and CO2 separation from flue gases. MOF-5 (or IRMOF-1) is the prototypical MOF phase with a cubic unit cell, Zn4O metal oxide linkers (with a formal +6 charge) bound to p-benzenedicarboyxlate organic struts in three perpendicular directions. Most MOFs show some degree of selectivity towards guest adsorption from the gas or solution phase. The simplest mechanism of guest separation in MOF materials is based on pore aperture size where large gas or solute molecules are excluded from the solid phase and smaller gas molecules diffuse through the pores to get to the other side of the MOF barrier, much like the action of a sieve. This type of separation is based on considerations of the van der Waals radii of the guests and MOF molecules lining the pore apertures. Geometric selectivity operates in the so-called “sorting domain” and is effective when the pore aperture diameter is close to the molecular diameters of the molecules in the mixture. MOFs with large pore apertures cannot separate components of a mixture based on geometric factors alone. However, many MOFs have organic struts with electron-rich functional groups such as aromatic rings. In these cases the MOFs can separate mixture components based on the different van der Waals interaction strengths between guests and the framework. Guests with greater numbers of electrons overall and stronger total van der Waals interactions, such as CO2, are selectively attracted to electron-rich parts of the framework over smaller guests like CO, N2, CH4, or H2. [8] A similar mechanism can apply to cases where the metal ion linker in the MOF has unsaturated (open) coordinate sites. The representative example of this class of MOF is HKUST-1 where the unsaturated Cu coordination site can act as a centre of H2 [9] or acetylene guest binding. The mechanism of guest adsorption and separation based on different guest–host interaction strengths is called the “coverage domain”. The selectivity in these cases is based on the electron density of the guest molecules and will have limited chemical specificity when gas-phase molecules of the same type are used. The coverage domain mechanism is operative at low and medium external guest pressures. An example of the preferential binding of CO2 molecules to specific electron-rich sites in the ZIF 68 structure at low pressure is shown in Figure 1. Achieving better chemical specificity for guest molecule binding requires custom-designed sites which geometrically and chemically bind to specific adsorbates but at the same time do not disrupt the MOF framework topology. Designing such a system is the important advance achieved by Li, Yahgi, Stoddart and coworkers, who have incorporated large macrocyclic polyether moieties into a cubic isoreticular MOF lattice. Li et al. use a modified form of the electron-rich bis-p-phenylene-34-crown10 (BPP34C10) molecule shown in Figure 2. A 1,5-naphtho-p-phenylene-36crown-10 analogue was also synthesized but is not further discussed. The electron-rich cyclic BPP34C10 forms a pseudorotaxane complex with the electrondeficient N,N’-dimethyl-4,4’-bipyridinium dication (paraquat, PQT , Figure 2) where the two pyridinium rings of PQT are aligned parallel to the two aromatic rings in BPP34C10 by p–p stacking interactions. Hydrogen bonding by the macrocycle ether oxygen atoms with the a-CH atoms of paraquat and electrostatic interactions between the hydroquinone ring oxygen atoms and the paraFigure 1. The selective binding of CO2 guest molecules to specific sites in ZIF-68 pores.