Implementation of a Core–Shell Design Approach for Constructing MOFs for CO2 Capture

Adsorption-based capture of CO2 from flue gas and from air requires materials that have a high affinity for CO2 and can resist water molecules that competitively bind to adsorption sites. Here, we present a core–shell metal–organic framework (MOF) design strategy where the core MOF is designed to selectively adsorb CO2, and the shell MOF is designed to block H2O diffusion into the core. To implement and test this strategy, we used the zirconium (Zr)-based UiO MOF platform because of its relative structural rigidity and chemical stability. Previously reported computational screening results were used to select optimal core and shell MOF compositions from a basis set of possible building blocks, and the target core–shell MOFs were prepared. Their compositions and structures were characterized using scanning electron microscopy, transmission electron microscopy, and powder X-ray diffraction. Multigas (CO2, N2, and H2O) sorption data were collected both for the core–shell MOFs and for the core and shell MOFs individually. These data were compared to determine whether the core–shell MOF architecture improved the CO2 capture performance under humid conditions. The combination of experimental and computational results demonstrated that adding a shell layer with high CO2/H2O diffusion selectivity can significantly reduce the effect of water on CO2 uptake.


General Methods
All reagents and solvents were commercially available and used as received.
Powder X-ray diffraction (PXRD) patterns were collected using a Bruker AXS D8 Discover powder diffractometer at 40 kV, 40 mA for Cu Kα, (λ = 1.5406 Å) with a scan speed of 0.20 s/step from 5 to 30° at a step size of 0.02°. The data were analyzed using the EVA program from the Bruker powder analysis software package. The simulated powder patterns were calculated using Mercury 3.8 based on MOF crystal structures.
Scanning electron microscopy (SEM) data were collected using a ZEISS Sigma 500 VP scanning electron microscope. Samples were dispersed in ethanol and drop cast on TEM grids (Ted Pella Inc., 200 mesh carbon film copper grids, catalog No. NC0733370). The TEM grids were dried under ambient conditions before SEM studies. A STEM sample holder was used to mount the TEM grids.
Transmission electron microscopy (TEM) images used to determine size distributions of MOF crystallites were collected on an FEI Morgagni 268 operated at 80 kV with an AMT side mount CCD camera system. Scanning transmission electron microscopy -energy dispersive X-ray spectroscopy (STEM) imaging and STEM-EDS studies were conducted on a JEOL JEM-2100F equipped with a Gatan Orius camera operated at 200 kV. Samples were dispersed in acetonitrile and drop cast on TEM grids [Ted Pella Inc., 200 mesh carbon film copper grids, catalog no. NC0733370]. The TEM grids were dried under ambient conditions before TEM and STEM-EDS analyses. EDS data were acquired using 1024 channels from 0 to 20 keV. Elemental maps were collected for 10-15 min with a pixel dwell time of 100 μs and a pixel resolution of 1024 × 1024. EDS maps and line-scans for zirconium and palladium were generated using the Zr Kα1 line intensity at 15.7 keV and the Pd Lα1 line intensity at 2.8 keV.
Gas adsorption isotherms were collected on a Quantachrome Autosorb-1 instrument or on a Micromeritics 3-flex gas adsorption analyzer. Approximately 40-60 mg of each sample was exchanged with dichloromethane 3 times a day for 1 day at 65 o C to remove N,Ndimethylformamide. After that, samples were degassed at 100 o C for 24 h on a Micromeritics SmartVacPrep under vacuum. A liquid N2 bath was used for the N2 adsorption experiments at 77 K. A water bath was used for N2 and CO2 adsorption experiment at 298 K. Ultra-high purity grade N2 and CO2 (99.999 %) was used. S-5

2,2'-dicyclohexylamino-[1,1'-biphenyl]-4,4'-dicarboxylic acid (4')
Compound 3' (474 mg, 1.74 mmol) was dissolved in 2 mL of DMF in a 25 mL 3-neck flask under N2 flow. NaH (192 mg, 8 mmol) was washed with hexane (2 mL, 3x) and dispersed S-8 in 2 mL of DMF. The NaH dispersion was then added to 3-neck round-bottomed flask at 273 K. After stirring the mixture for 1 h, iodocyclohexane (840 mg, 4 mmol) was added dropwise. The reaction temperature was allowed to rise to room temperature, and the mixture was allowed to stir for 48 h under N2 atmosphere. After completion of the reaction, the mixture was quenched with saturated NaHCO3 solution. The resulting solution was washed with ethyl acetate (50 mL, 3x) to remove the unreacted iodocyclohexane. 1 M HCl was then added to the aqueous solution until reaching acidic pH to give a light brown solid (741 mg, 97.5%). 1

Multi-gas tests results
After solvent exchange using the same method as gas adsorption isotherm tests, the MOF samples were activated at 120 o C for 24 hours in a vacuum oven before multi-gas test in order to remove any solvent molecules potentially remaining in the pore. For multi-gas measurement, ~100 mg of sample was sealed in a gas tight module with silicon o-rings. The feed gas was purged through the module for 12 h (Scheme S1). The feed contained a mixture of N 2 and CO 2 (85/15) with 0%, 15% and 30% relative humidity with a total feed flow rate of 200 mL/min; feed composition was adjusted using mass flow controller. The MOF sample was then re-activated at 120 o C. The compositions of desorbed gas were analyzed by gas chromatography (Agilent-GC8860). He gas was introduced with volumetric flow rate of 50 mL/min as a sweep gas. The plot of velocity-time was used to calculate the uptake of each gas. Tests at each condition were performed twice to confirm the reliability of the results and the average gas uptake of two trials was used for analysis. After multigas test at each condition, PXRD patterns and N 2 adsorption isotherms at 77 K were collected to assess the integrity of the MOFs. Table S1. Multi-gas tests results at 0% RH.  S-24 Figure S28. N2 adsorption isotherms at 77 K of NH2-UiO-67 before multi-gas tests (black square), after 15% RH tests (red circle) and after 30% RH tests (blue triangle). The calculated BET surface areas were 2280 m 2 g -1 (before multi-gas tests), 2150 m 2 g -1 (after 15% RH tests) and 1810 m 2 g -1 (after 30% RH tests). Figure S29. N2 adsorption isotherms at 77 K of (CyNH)2-UiO-67 before multi-gas tests (black square), after 15% RH tests (red circle) and after 30% RH tests (blue triangle). The calculated BET surface areas were 1780 m 2 g -1 (before multi-gas tests), 1650 m 2 g -1 (after 15% RH tests) and 1390 m 2 g -1 (after 30% RH tests). . N2 adsorption isotherms at 77 K of cs-MOF-1 before multi-gas tests (black square), after 15% RH tests (red circle) and after 30% RH tests (blue triangle). The calculated BET surface areas were 1830 m 2 g -1 (before multi-gas tests), 1820 m 2 g -1 (after 15% RH tests) and 1610 m 2 g -1 (after 30% RH tests). Figure S31. N2 adsorption isotherms at 77 K of cs-MOF-2 before multi-gas tests (black square), after 15% RH tests (red circle) and after 30% RH tests (blue triangle). The calculated BET surface areas were 1810 m 2 g -1 (before multi-gas tests), 1730 m 2 g -1 (after 15% RH tests) and 1480 m 2 g -1 (after 30% RH tests).