Achieving Molecular Sieving of CO2 from CH4 by Controlled Dynamical Movement and Host–Guest Interactions in Ultramicroporous VOFFIVE-1-Ni by Pillar Substitution

Engineering the building blocks in metal–organic materials is an effective strategy for tuning their dynamical properties and can affect their response to external guest molecules. Tailoring the interaction and diffusion of molecules into these structures is highly important, particularly for applications related to gas separation. Herein, we report a vanadium-based hybrid ultramicroporous material, VOFFIVE-1-Ni, with temperature-dependent dynamical properties and a strong affinity to effectively capture and separate carbon dioxide (CO2) from methane (CH4). VOFFIVE-1-Ni exhibits a CO2 uptake of 12.08 wt % (2.75 mmol g–1), a negligible CH4 uptake at 293 K (0.5 bar), and an excellent CO2-over-CH4 uptake ratio of 2280, far exceeding that of similar materials. The material also exhibits a favorable CO2 enthalpy of adsorption below −50 kJ mol–1, as well as fast CO2 adsorption rates (90% uptake reached within 20 s) that render the hydrolytically stable VOFFIVE-1-Ni a promising sorbent for applications such as biogas upgrading.


S3. Characterization
Powder X-ray diffraction patterns of all synthesized HUMs were collected on a Bruker D8 powder diffractometer (Karlsruhe, Germany) using Cu-Kα1,2 radiation (λ1 = 1.5406Å, λ2 = 1.5444Å) and operated at 40 kV and 40 mA.A capillary holder was employed for all measurements and patterns were collected in a 2θ-range of 5 to 100° using a step size of 0.01° at ambient conditions.Scanning electron microscopy (SEM) images were taken using a Zeiss Merlin Field Emission Scanning Electron Microscope (Oberkochen, Germany) using an acceleration voltage of 2 kV.Volumetric CO2, N2, and H2O sorption isotherms were recorded on HUM samples activated at 423 K for 3 h under dynamic vacuum (1×10 −4 Pa) using a Micromeritics SmartVacPrep (Norcross, USA).After activation, equilibrium N2 sorption isotherms were recorded at 77 K as well as CO2, N2, and H2O sorption isotherms at 293 K which were collected using a temperature-controlled water bath at 293 K on a Micromeritics ASAP2020 surface area analyzer (Norcross, USA).Heat of adsorption was calculated using Clausius-Clapeyron equation (Equation 7) on the CO2 adsorption recorded at 423 K and 433 K. CO2/CH4 selectivity in this study is reported as the uptake ratio between CO2 and CH4 and the use of ideal adsorption solution theory (IAST) was omitted due to that molecular sieving nature of the samples, which falls outside one of assumptions of IAST where each sorbent should have equal accessibility to the sorbate area.Thermogravimetric profiles were recorded on a Mettler Toledo Thermal Analysis System TGA 2 and Mettler Toledo Thermal Analysis System TGA/DSC 3+ (Columbus, United States) using air, N2, and/or CO2 at a flow-rate of 50 ml min −1 .

S.3.1.1. Transmission electron microscopic (TEM) analysis
Specimens were loaded onto copper grids coated with a layer of carbon.The investigative procedures were conducted using a JEOL JEM2100 microscope (Akishima, Japan) operating at 200 kV.The TEM images were recorded utilizing a Gatan Orius 833 CCD camera with a resolution of 2048 x 2048 pixels and an individual pixel dimension of 7.4 µm.Furthermore, the collection of electron diffraction patterns was facilitated by a Timepix pixel detector QTPX-262k, featuring a pixel count of 512 x 512 and a pixel size of 55 µm.This pixel detector was innovatively manufactured by Amsterdam Scientific Instruments (Amsterdam, Netherlands).

S.3.1.2. Continuous rotation electron diffraction (cRED) collection
The data acquisition was carried out employing Instamatic software, 3 utilizing a single-tilt tomography holder with a range of −70° to +70° for tilting within the TEM.The cRED data collection involved the use of an approximately 0.6 μm diameter aperture.Goniometer tilt occurred at a speed of 1.1° s −1 , and each frame was exposed for 0.3 s.This strategic approach ensured that each data set was assembled S-4 within a span of 2 min, thereby minimizing beam-related degradation and enhancing data quality.The integration of data sets for structure determination harnessed the capabilities of XDS packages. 4

S.3.1.3. cRED data processing
Due to the small crystal size of NbOFFIVE-2-Ni (< 2 μm), 3DED, namely continuous rotation electron diffraction (cRED), 5,6 was applied for crystal structure analysis.Reconstructed 3D reciprocal lattice from the cRED data indicated the unit cell parameters to be a = 10.The dataset exhibited a high resolution of 0.80 Å with data completeness 86.0% for NbOFFIVE-2-Ni. 7,8 Du to the high 3D electron diffraction data quality, the structure could be solved via ab initio by direct methods.The final refinement converged to a R1 value of 0.1408.Details of data collection and structure refinement are shown in Table S1.
The refinement of NbOFFIVE-2-Ni's unit cell parameters (Figure S2) was carried out through Pawley fitting by using TOPAS 4.1.Background correction was performed using a 16th-order Chebychev S-5 polynomial.Refinement was executed utilizing a Pearson VII-type peak profile function, followed by the refinement of unit cell and zero-shift parameters.Finally, the unit cell parameters were refined to a = b = 9.901(5) Å and c = 8.355(8) Å.The corresponding R-values were converged to: Rp = 0.0433, Rwp = 0.0582, and Rexp = 0.0514 (Table S2).Unit cell parameters obtained from cRED data are typically larger than those obtained from X-ray diffraction.This is generally assumed to be due to lens distortions and changes in sample height during data collection. 9

S3.2. Rietveld and Pawley refinements
Pawley and Rietveld refinements were carried out using the structure analysis software TOPAS (v.6).

S-26
Table S6.Unit cell parameters and interatomic distances of as-synthesized HUM compounds.S-27

S3.3.2. Hydrolytic stability
The hydrolytic stability of the as-synthesized HUMs was investigated by stirring the materials (approx. 0.01 g) in 15 ml of deionized water for 3 months, after which the crystallinity of the samples was evaluated by powder X-ray diffraction.
All CO2 isotherms were fitted using the Langmuir-Freundlich model: Where qeq [mmol g −1 ] is the equilibrium CO2 uptake at pressure p [bar], qsat [mmol g −1 ] the saturation uptake capacity of an adsorption site, b the affinity parameter for the specific adsorption site.Time-resolved CO2 uptake data were obtained from time-dependent IR spectra.The IR spectra were collected by rapidly introducing CO2 gas (>99.999%,provided by Linde gas company), into the manifold and subsequently collecting IR spectra every 0.25 s for up to 60 min.

S-34
The single beam absorption spectra of the sorbent without CO2 (I0) and with CO2 (I) were used to calculate the absorption spectra for CO2 adsorbed on the sorbent using the relationship A = log10(I0/I).
This calculation was carried out for every time point.

Figure S1 .
Figure S1.a) Reconstructed 3D reciprocal lattice of NbOFFIVE-2-Ni (inset is the image of the crystal on which the cRED data was collected) and 2D slice cuts of 3D reciprocal lattice of NbOFFIVE-2-Ni show the (b) hk0, (c) 0kl, and (d) hkl planes.

Figure S25 .Figure S29 .
Figure S25.(a) N2 sorption isotherm of VOFFIVE-1-Ni recorded at 77 K (filled and open circles represent the adsorption and desorption branches, respectively), showing a typical type II non-porous material (b) BET plot showing the linear region wherein the BET model would be valid for porous materials, and (c) Rouquerol plot (the linear region is highlighted in red).BET model was applied to the data here for consistency and quantitative comparison with other samples.

4 Where 5 Figure S38 .
Figure S38.CO2 profiles derived from in situ infrared data fitted using the short-time expression of the transient diffusion equation.The infrared measurements were carried out at 303 K and 1000 ppm CO2.

Table S8 .
Summary of fitting parameters.