Long Time CO2 Storage Under Ambient Conditions in Isolated Voids of a Porous Coordination Network Facilitated by the “Magic Door” Mechanism

Abstract A coordination network containing isolated pores without interconnecting channels is prepared from a tetrahedral ligand and copper(I) iodide. Despite the lack of accessibility, CO2 is selectively adsorbed into these pores at 298 K and then retained for more than one week while exposed to the atmosphere. The CO2 adsorption energy and diffusion mechanism throughout the network are simulated using Matlantis, which helps to rationalize the experimental results. CO2 enters the isolated voids through transient channels, termed “magic doors”, which can momentarily appear within the structure. Once inside the voids, CO2 remains locked in limiting its escape. This mechanism is facilitated by the flexibility of organic ligands and the pivot motion of cluster units. In situ powder X‐ray diffraction revealed that the crystal structure change is negligible before and after CO2 capture, unlike gate‐opening coordination networks. The uncovered CO2 sorption and retention ability paves the way for the design of sorbents based on isolated voids.


Table of Contents
1 H (400 MHz) and 13 C NMR spectra were measured using JEOL JNM-ECA400 II spectrometer.The samples were dissolved in deuterated chloroform (CDCl3) containing tetramethylsilane (TMS), and the chemical shifts were referenced against the TMS peak (δ 0.0).
Single crystal X-ray diffraction data for the ligand (L) and the as-synthesized networks 1 and 2 containing MeCN in the pores were measured using Rigaku VariMax X-ray diffractometer with Saturn.The crystals were cooled under nitrogen flow to 123 K using the Rigaku GNNP cryogenic cooler.Graphite-monochromated Mo K (λ = 0.71075 Å) was used as an X-ray source, and the detector was a two-dimensional CCD detector.The collected diffraction data were analyzed by Rigaku CrysAlisPro software.
Single crystal X-ray diffraction data for 2@activated and 2@CO2 at 90 K were measured at the BL-5A beamline, Photon Factory in the Institute of Materials Structure Science (IMSS), High Energy Accelerator Research Organization, KEK.Single crystals were cooled to 90 K by a Rigaku CryoCooler and irradiated with synchrotron radiation (λ = 0.7500 Å).Diffraction patterns were recorded by a Dectris Pilatus3 S6M detector.The measured diffraction data were integrated by XDS.For the measurements, the crystals of 2@CO2 were soaked in oil immediately after taking them out of the CO2 atmosphere.
Single crystal X-ray diffraction data for 2@CO2 at 298 K was collected on a Rigaku Synergy-R/DWTI APEX II instrument with a Hypix-6000HE detector equipped with Rigaku GNNP low temperature device using PhotonJet-R X-ray source with MicroMax™-007 rotating anode Cu Kα radiation (λ = 1.54184Å).
Adsorption/desorption isotherms of network 1 for N2 (298 K) and CO2 (298 K), and network 2 for N2 (298 K) and CO2 (273, 283, 298, 313 and 333 K) were measured using Micromeritics 3Flex Adsorption Analyzer.The powdered samples (ca.70 mg) were ground in an agate mortar prior to being placed into glass sample tubes.The solvent in the pores of 1 and 2 was removed by heating the networks at 553 K and 473 K, respectively, under dynamic vacuum for 12 h using Micromeritics VacPrep 061 attached to a rotary pump.The 3Flex Version 5.03 program was used to analyze the experimental results.
The adsorption isotherms for N2 at 77 K and CO2 at 195 K were measured by MicrotracBEL BELSORP MAX.The adsorption volume was measured using the constant volume gas adsorption method without removing the sample cell from the apparatus.The measurement temperature was controlled by a bath of liquid nitrogen for 77 K and acetone with dry ice for 195 K.The experimental results were analyzed using the BEL MASTERTM analysis program.
Solid-state UV-vis absorption spectra were collected using a UV-Vis-NIR Spectrophotometer V-770 from JASCO Corporation (Japan Spectroscopy Co., Ltd.).The samples were ground in an agate mortar and then transferred to the sample holder.The measurements were performed in the diffuse reflectance mode using a φ = 60 mm integrating sphere unit.BaSO4 white plate was used as a background.The final spectra were converted using the Kubelka-Munk (KM) transform.
Infrared absorption spectra were collected using a Nicolet™ iS™ 50 Fourier transform infrared spectrophotometer from Thermo Fisher Scientific.A liquid nitrogen-cooled MCT-A detector was used for diffuse reflectance measurements.Solid samples were diluted with potassium bromide (KBr) while pure KBr was used as the background.
TGA-DSC data was collected by Simultaneous Thermogravimetry STA449 F3 Jupiter analyzer from NETZCH.Approximately 5 mg of each sample was placed in ceramic pans and heated from room temperature to 700 °C under nitrogen flow.Before measurements, the inside of the furnace was purged with nitrogen for 10 min.
Elemental analysis (CHN) was performed using Elementar's vario MICRO cube macro-organic elemental analyzer.Powder X-ray diffraction (PXRD) measurements were performed on a SmartLab fully automated multi-purpose X-ray diffractometer from Rigaku using Cu K (λ = 1.5418Å) as an X-ray source.The powdered samples were placed between two Mylar sheets and the patterns were collected in the transmission mode on a D/teX Ultra (1D) detector.
In situ PXRD experiments under a CO2 atmosphere were performed at the RIKEN Materials Science beamline BL44B2 in SPring-8.The measurements were carried out using a diffractometer equipped with an OHGI system[1] in transmission mode, with a wavelength of λ = 0.5201 Å.The measurement apparatus was connected to a gas cylinder and a vacuum pump via BELSORP-G, which allowed for gas introduction with controllable pressure.Powdered samples of 2@activated were packed into glass capillaries and assembled to allow gas flow within the capillary (Figure S1).It was mounted in the apparatus, heated under vacuum at 493 K, followed by introduction of CO2.The system was placed under approximately 100 kPa of CO2 and the time dependent PXRD patterns were collected for 12 h at 298 K.The calculations were carried out by PreFerred Potential with dispersion force correction (PFP+D3) version 3.0.0 on Matlantis TM , a high-speed versatile atomic-scale simulator powered by a neural network potential (NNP).The initial model was a unit cell obtained from the single crystal structure and contained about 3000 atoms.By specifying periodic boundary conditions for each face of the unit cell, a complete network structure was reproduced.Using this model, the enthalpies of adsorption (ΔHads) and the diffusion path of CO2 through the network were calculated.

2-1. General methods
All reagents were purchased from NARD Institute, Kanto Chemical, Tokyo Chemical Industry and Wako Chemical and used without further purification.
2-2.Synthesis of 3,3′-5,5′-tetrakis(5-pyrimidyl) bimesityl (L) To a 100 mL Schlenk flask, tetraiodobimesityl (0.52 g, 0.70 mmol), 5-pyrimidyl boronic acid (0.41 g, 3.3 mmol), tetrakis(triphenylphosphine)palladium (0.24 g, 0.22 mmol), and potassium carbonate (2.84 g, 21 mmol) were added.Toluene (30 mL), ethanol (20 mL), and distilled water (10 mL) were degassed by bubbling N2 for 30 min, and then transferred to the reaction flask.The reaction mixture was refluxed at 80 °C under N2 atmosphere for 2 days.The solvents were removed by evaporation, and the residue was re-dissolved in chloroform (200 mL).The organic phase was washed with water and brine, dried over anhydrous MgSO4, and evaporated to dryness to give a brown solid.The crude compound was dissolved in a small amount of chloroform and purified by silica gel column chromatography with ethyl acetate/triethylamine (100:1) as an eluent.The final product was obtained as a white powder (0.27 g, 69%).2-3.Synthesis of network 1 L (11 mg, 0.020 mmol), copper(I) iodide (19 mg, 0.10 mmol), and potassium iodide (0.83 g, 5.0 mmol) were placed inside a Teflonlined stainless-steel autoclave.A mixture of acetonitrile (5.4 mL), distilled water (3.6 mL) and N,N′-dimethylformamide (DMF) (1 mL) was added and the reaction vessel was heated in an oven at 120 °C for 64 h.After that, the oven was slowly cooled to r. t. for half a day.The precipitate was removed by vacuum filtration, washed with DMF, water and acetonitrile, and dried to give a mixture of yellow plate-like (network 1) and prism-like (network 2) crystals.To separate the two networks, the mixed powder was immersed into a dibromomethane/dichloromethane (3:4) mixture (7 mL) inside a glass tube.After 5 min, the network 1 crystals floated to the top, whereas the network 2 crystals sunk to the bottom.The former were removed from the glass tube using a pipette and then isolated by vacuum filtration.Pure network 1 was obtained with the yield of 74 % based on L. The phase purity was confirmed by PXRD.Elemental analysis.Calcd for {Cu2I2[C34H30N8(L)]•(CH3CN)0.73•(H2O)1.12}:C,43.39;H,3.53;N,12.46. Found: C,43.39;H,3.30;N,.Synthesis of network 2 L (11 mg, 0.020 mmol), copper(I) iodide (30 mg, 0.16 mmol), potassium iodide (0.83 g, 5.0 mmol), and triphenylphosphine (5.2 mg, 0.020 mmol) were placed inside a Teflon-lined stainless-steel autoclave.A mixture of acetonitrile (5.4 mL), distilled water (3.6 mL) and ethanol (1 mL) was added and the reaction vessel was heated in an oven at 140 °C for 64 h.After that, the oven was slowly cooled to r. t. for half a day.The precipitate was removed by vacuum filtration, washed with DMF, water and acetonitrile, and dried to give a mixture of yellow plate-like (network 1) and prism-like (network 2) crystals.The same density separation method as described for the network 1 was used to purify the network 2. After the separation was achieved, the sunken crystals of 2 were removed from glass tube using a pipette and then isolated by vacuum filtration.Pure network 2 was obtained with the yield of 58 % based on L. The phase purity was confirmed by PXRD.Elemental analysis.Calcd for｛Cu4I4[C34H30N8(L)]•(CH3CN)2•(H2O)0.71｝:C, 32.43; H, 2.68; N, 9.95.Found: C, 32.44; H, 2.68; N, 9.76.

2-6. Adsorption of CO2 into network 2
The crystals of activated network 2 without interstitial solvent molecules were placed into a 5 mL glass vial, and a CO2 balloon was attached through a needle.The crystals were kept in a pure CO2 atmosphere overnight allowing a near complete saturation of the network pores.After that, the network sample was removed from the vial and further experiments were carried out under ambient atmosphere.The presence of CO2 was confirmed by single crystal X-ray diffraction analysis and infrared spectroscopy.

Figure S1 .
Figure S1.Schematic diagram of the sample holder used for the in situ powder X-ray diffraction experiment.

Figure S4 .
Figure S4.Single crystal structure of the as-synthesized network 1 showing a) the voids outlined in dark yellow, b) a single 2D layer and c) multiple layers represented by different colors.C -grey, N -blue, and Cl -light green; hydrogen atoms were omitted for clarity.

Figure S5 .
Figure S5.Single crystal structure of the as-synthesized network 2. a) connectivity of the ligand with CuI cubane clusters and disordered acetonitrile in the pore and (b) packed structure view down the b axis, (acetonitrile molecules are shown using spacefilling model) C -grey, N -blue, Cu -orange, and I -purple, hydrogen atoms were omitted for clarity.

Figure S18 .
Figure S18.Time-dependent PXRD patterns of 2@activated collected in situ as the network was being exposed to CO2 atmosphere at 100 kPa.

Figure S19 .
Figure S19.Optimized structure of network 2 containing CO2 simulated by Matlantis TM .(a) short contacts between network 2 and the CO2 guest, (b) CO2 orientation inside the pore (dark yellow), C -grey, N -blue, O -red, Cu -orange, and I -purple, hydrogen atoms were omitted for clarity.