Natural-gas sweetening (separation of CO2 from CH4) is an industrially significant separation process as CO2 represents a substantial (up to 70 %) impurity in natural — gas wells [80]. The presence of CO2 reduces the energy content of the natural gas, and its acidity in the presence of water can result in the corrosion of natural-gas lines. Physical solvent-based processes for CO2 removal from natural-gas are abundant, however, the large amount of water recycling needed makes solvent- based processes highly limited in this application due to solvent degradation and loss during operation [81]. Porous solids present a more efficient and environ­mentally friendly way to capture CO2 from natural-gas wells. In this case, sepa­ration largely proceeds based on quadrupole moment, due to the similar properties of the two gases in other respects (kinetic diameter, polarizability, dipole moment). Additionally, the flexible structure of some MOFs upon adsorption-desorption (in contrast with “rigid” adsorbents such as carbons and zeolites), may result in dynamic and stepwise adsorption at different pressures. This is generally known as a “gate opening” phenomenon, and arises mainly from the flexibility of the net­works and their affinity for particular guests [82]. In MIL-53 [Cr(OH)(bdc)], for example, the selective adsorption of CO2 over CH4 is strongly affected by the presence of water which causes dramatic changes in the pore structure [27].

Neutron scattering has also been used extensively to study CH4 confined in porous materials, in particular to study methane confined in MOFs, commensurate

with the increasing work investigating these hosts for application in CO2/CH4 separations.

The metal sites, including open-metal sites, in many MOFs also interact with CH4. NPD studies of the Mg2(dobdc) material show the binding of one CD4 molecule per open-metal site, resulting in the large CH4 storage capacity of 160-174 cm3(at standard temperature and pressure, STP)/cm3, approaching the DOE target of 180 cm3(STP)/cm3 for solid-based CH4 storage at room temperature [83]. Direct determination of CD4 sorption sites in Zn(mIm)2 and Zn4O(bdc)3 were gained using NPD (Fig. 3.8) [84]. The primary CD4 adsorption sites are associated with the organic linkers in Zn(mIm)2 and the metal oxide clusters in Zn4O(bdc)3. In Zn4O(bdc)3 the first binding sites (“cup” sites) were not found to alter the Fm 3m symmetry of the host-guest system. CD4 at these primary sites possesses well — defined orientations, implying relatively-strong binding with the framework. With higher CD4 loading, additional CD4 molecules populate secondary sites and are confined in the framework. The confined CD4 at these secondary sites is orienta­tionally disordered and stabilized by the intermolecular interactions. The CD4 guest is a high symmetry guest whose ordered location (at the primary sites) significantly alters the symmetry of system. The “hex” and “ZnO2” CD4 sites caused a symmetry lowering of the system to I4/mmm as a result of the symmetry incompatibility of the tetrahedral CD4 molecules with the local geometry. At higher CD4 loadings a P4 mm structure was found, where CD4 sites aligned themselves along the c axis and further lowered the symmetry.

Using a similar approach, a comprehensive mechanistic study of CD4 was performed in Cu3(btc)2, Cu2(sbtc) where sbtc = frans-stilbene-3, 3′, 5, 5′-tetracar — boxylate, and Cu2(adip) where adip = 5, 5′-(9, 10-anthracenediyl)di-isophthalate and also known as PCN-14, allowing a comparison of structures that consist of the same dicopper-paddlewheel secondary-building units (the well-known dicopper acetate unit), but contain different organic linkers, leading to cage-like pores with various sizes and geometries (Fig. 3.9). This work revealed that CD4 uptake takes place primarily at two types of strong adsorption site: (1) the open Cu sites which exhibit enhanced coulombic attraction toward CD4, and (2) the van der Waals potential pocket sites in which the total dispersive interactions are enhanced due to the molecule being in contact with multiple “surfaces”. Interestingly, the enhanced van der Waals sites are present exclusively in small cages and at the windows to these cages, whereas large cages with relatively flat pore surfaces bind very little CD4 [85].

The self-diffusion of CH4 was measured directly using QENS in the isostructural Cr(OH)(bdc) and V(O)(bdc) materials [86]. The hydroxyl groups in Cr(OH)(bdc) were expected to hinder CH4 mobility, although this work revealed a global one­dimensional diffusion mechanism of CH4 in both materials, echoing the single-file diffusion regime found for CO2 [49]. An interesting result of this work was that CH4 diffusivities are significantly higher in V(O)(bdc) than in Cr(OH)(bdc) over the whole range of investigated CH4 loadings.

There have been several neutron-scattering studies targeting the separation mechanism of CO2 from CH4. The polymer/selective-flake nanocomposite mem­branes exhibiting selectivity for O2 over N2 (discussed in Sect. 3.3.3) also shows substantial selectivity of CO2 over CH4. Again, SANS results revealed that this occurs within only 10 wt% of the AlPO layers. The Zr6O4(OH)(bdc)6 material, also known as UiO-66(Zr), is a MOF with encouraging properties for CO2/CH4 gas separation, achieved by combining good selectivity with a high working capacity

and the ability for regeneration under relatively-mild conditions [87]. Zr6O4(OH) (bdc)6 is built from Zr6O4(OH)4 octahedra that extend into three-dimensions via bdc ligands, resulting in two types of microporous cages. The dynamics of CO2 and CH4 within Zr6O4(OH)(bdc)6 was measured using QENS and matched with results from MD simulations [87] (Fig. 3.10). Importantly, this work established the

concentration dependence of the diffusivities of CH4 and CO2 (self and transport, respectively) within the material. The flexibility of the framework was found to influence significantly the diffusivity of the two species, and CH4 was found to diffuse faster than CO2 over a broad concentration range, a result that is in contrast to zeolites with narrow windows, for which opposite trends were observed. Further analysis of the MD trajectories for CH4 provided insights into the global micro­scopic diffusion-mechanism, proposed to occur by a combination of intracage motions and jump sequences between the material’s tetrahedral and octahedral cages. The coadsorption of CO2 and CH4 in the material, from both a thermody­namic and a kinetic perspective, was also studied using this approach. It was shown that each type of guest adsorbs preferentially in the two different pores, where CO2 occupies the tetrahedral cages and CH4 the octahedral cages. Further, a very unu­sual dynamic behaviour was also noted in the study of CH4/CO2 mixtures in Zr6O4(OH)(bdc)6 in that the slower CO2 molecule was found to enhance the mobility of the faster CH4, again contrasting with the usual observation for CO2/ CH4 mixtures in narrow-window zeolites, where the molecules diffuse indepen­dently or slow the partner species.

The self-diffusion properties of pure CH4 and its binary mixture with CO2 within the NaY zeolite have also been investigated by the combined QENS/MD approach. This material combines several favourable features including a good selectivity, high working capacity, and potential easy regenerability that make it a good can­didate for the selective adsorption of CO2 over CH4 [88]. The QENS measurements at 200 K led to an unexpected self-diffusivity profile for pure CH4 with the presence of a maximum for a loading of 32 CH4/unit cell, which was previously unobserved for the diffusion of an apolar species in a zeolite with large windows. The QENS measurements report only a slight decrease of the self-diffusivity of CH4 in the presence of CO2 when the CO2 loading increases. MD calculations successfully reproduce this experimental trend and suggest a microscopic diffusion-mechanism in the case of this binary mixture [89].