Self-Assembly of Chiral Porous Metal–Organic Polyhedra from Trianglsalen Macrocycles

Metal–organic polyhedra (MOPs) can exhibit tunable porosity and functionality, suggesting potential for applications such as molecular separations. MOPs are typically constructed by the bottom-up multicomponent self-assembly of organic ligands and metal ions, and the final functionality can be hard to program. Here, we used trianglsalen macrocycles as preorganized building blocks to assemble octahedral-shaped MOPs. The resultant MOPs inherit most of the preorganized properties of the macrocyclic ligands, including their well-defined cavities and chirality. As a result, the porosity in the MOPs could be tuned by modifying the structure of the macrocycle building blocks. Using this strategy, we could systematically enlarge the size of the MOPs from 26.3 to 32.1 Å by increasing the macrocycle size. The family of MOPs shows experimental surface areas of up to 820 m2/g, and they are stable in water. One of these MOPs can efficiently separate the rare gases Xe from Kr because the prefabricated macrocyclic windows of MOPs can be modified to sit at the Xe/Kr size cutoff range.


NMR Spectra
NMR spectra were recorded at ambient probe temperature on a Bruker 400 NMR spectrometer at 400 MHz ( 1 H) and 100 MHz ( 13 C) and referenced against the residual 1 H or 13 C signal of the solvent.

Infra-red (IR) Spectra
IR spectra were recorded on a Bruker Tensor 27 FT-IR using ATR measurements for solids as neat samples, or using transmission mode on a 96-well silica wafer deposited as a thin film as part of the high-throughput analysis.

Elemental analysis
CHN analysis was performed on a Thermo EA1112 Flash CHNS-O Analyser using standard microanalytical procedures.

High resolution mass spectrometry (HRMS)
HRMS was carried out using an Agilent Technologies 6530B accurate-mass QTOF Dual ESI mass spectrometer (capillary voltage 4000 V, fragmentor 225 V) in positive-ion detection mode.The mobile phase was methanol (MeOH) + 0.1% formic acid at a flow rate of 0.25 mL/min.

Powder X-ray diffraction
Laboratory powder X-ray diffraction (PXRD) data patterns were collected in transmission mode on samples held on thin Mylar film in aluminium well plates on a Panalytical Empyrean diffractometer, equipped with a high throughput screening XYZ stage, X-ray focusing mirror, and PIXcel detector, using Cu-Kα (λ = 1.541Å) radiation.For indexing, samples were loaded into borosilicate glass capillaries, and PXRD patterns were recorded in transmission mode on a Panalytical Empyrean diffractometer, equipped with a sample spinner, X-ray focusing mirror, and PIXcel detector, using Cu-Kα (λ = 1.541Å) radiation.

Gas sorption
Nitrogen isotherms of M1 and M2 were collected at 77 K using an ASAP2020 volumetric adsorption analyzer (Micrometrics Instrument Corporation).M1 and M2 were degassed at 100 °C for 15 hours under a dynamic vacuum prior to gas analysis.Isotherm measurements of MOPs were performed using a Micromeritics 3flex surface characterization analyzer, equipped with a Cold-Edge technologies liquid helium cryostat chiller unit for temperature control.N 2 isotherms for MOPs were collected at 77 K. Kr and Xe isotherms for MOPs were collected at 273 and 298 K. MOPs activated by solvent exchange method using acetonitrile (MeCN) were degassed at 100 °C for 15 hours under dynamic vacuum prior to gas analysis.Zn 6 (M2-(R, R)) 4 activated by the supercritical CO 2 drying was degassed at room temperature for 15 hours under dynamic vacuum prior to gas analysis.

Thermogravimetric analysis (TGA)
TGA was carried out using a Q5000IR analyzer (TA instruments) with an automated vertical overhead thermobalance.The samples were heated at the rate of 10 °C /min using dry N 2 as the protective gas.(R, R)) 4 after further dilution were 0.60 mg/mL and 0.35 mg/mL, respectively.Then, the CD (mdeg) signal of each sample was determined by a circular dichroism spectrometer (model: JASCO J-1700, testing optical path: 1 mm).The mdeg was converted to molar ellipticity by calculation.The formula is [θ](deg•cm 2 •dmol -2 ) = mdeg•10 3 /(l•c) (l (mm) represents the optical path, and c (mM) represents the molar concentration of the sample).

Inductively coupled plasma optical emission spectrometry (ICP-OES)
ICP-OES measurements were conducted on an ICP-OES Agilent 5110.Before the measurements, the ICP-OES instrument was calibrated using four standards at three different wavelengths (202.548nm, 213.857 nm, and 472.215 nm), and the relative standard deviations (% RSDs) were determined from the calibration measurements.The filtered MOPs immersing solutions were diluted using distilled water before analysis.

Computational details
An isolated molecule, extracted from the experimental crystal structure of M1-(R, R) and Zn 6 (M1-(R, R)) 4 , was optimized by the GFN2-XTB method with D4 dispersion 1 model in the gas phase with defaults for convergence.3][4] Based on that, MD simulation was performed for 200 ps, in which 100 ps for equilibration and 100 ps for production with timestep of 2 fs, along with SHAKE restraints on all bonds, 5 in the NVT ensemble using the Berendsen thermostat 6 to maintain the temperature of 298 K. Structures were dumped every 2 ps.
Pywindow 7 was used to calculate the widows diameter of the molecular dynamics trajectories including 50 geometries obtained from xTB calculations.

Synthetic procedures
All enantiomers were synthesized using the same reaction procedures by using the appropriate chiral start materials.M1-(R, R) was synthesized as described previously 8 with the slight adjustments to the method.A solution of trans-(1R,2R)-diaminocyclohexane (560 mg, 5 mmol) in 100 ml was added into 150 ml solution of 2,5-dihydroxyterephthalaldehyde (815 mg, 5 mmol) in CHCl 3 .The mixture was stirred under nitrogen atmosphere at room temperature for 7 days.After that time, to the mixture was added ethanol (100 mL).The product crystallized as yellow-orange solid in 80% yield.3,3'-Dihydroxy[1,1'-biphenyl]-4,4'-dicarboxaldehyde (121 mg 0.5 mmol) was dissolved in tetrahydrofuran (25 mL).Then, 25 mL (1R,2R)-1,2-diaminocyclohexane (57 mg, 0.5 mmol) dissolved in ethanol added to the reaction slowly.The mixture was stirred at room temperature for 12 h, then the solvent was evaporated at room temperature.The yellow solid product was obtained by filtering and washing with ethanol.Yield: 107.6 mg (67.3%).M2 (50 mg) was dissolved in tetrahydrofuran (10 mL).After filtering the mixture through a syringe filter (0.2 µm PTFE membrane), the clear solution was divided evenly into five vials (2 mL per vial).Ethanol was allowed to vapour-diffuse slowly into the vials over a period of 3 to 7 days at room temperature.Yellow needle crystals of M3-(R, R) were obtained by filtering and washing with ethanol.Yield: 23 mg (46.0%).-    6 .6 0 6 .6 5 6 .7 0 6 .7 5 6 .0 6 .8 5 6 .9 0 6 .9     [a] Due to the severe disorder of the solvent molecules in the large crystal pores, a solvent mask implemented in Olex2 [9] using the BYPASS [10] method mask was used during the final refinement cycles.b] Due to the severe disorder of the solvent molecules in the large crystal pores, a solvent mask implemented in Olex2 [9] using the BYPASS [10] method mask was used during the final refinement cycles.The solvent mask found 6233 electrons in an interconnected, 23718 Å 3 sized pore.S-28

ICP data
Table S2.ICP-OES analysis of water and 0.1 M HCl (aq) solutions that had been used to immerse the solid samples of Zn 6 (M1-(R, R)) 4 and Zn 6 (M2-(R, R)) 4 for 10 days at room temperature.We used the ICP values as markers to determine if the solid samples of Zn 6 (M1-(R, R)) 4 and Zn 6 (M2-(R, R)) 4 decomposed during these experiments.The analysis was conducted after filtering the solutions to remove the remaining solids, and the resulting filtrates were diluted with water before the measurement.Each measurement was performed at three different wavelengths (202.548nm, 213.857 nm, and 472.215 nm).Each wavelength was calibrated beforehand using four standards, and the relative standard deviations (% RSDs) were determined from these calibration measurements.

Figure S29 .Figure S30 .
Figure S29.TGA plot of (a) M1-(R, R) (black line), and as-crystallized Zn 6 (M1-(R, R)) 4 from DEF after exchanging the DEF crystallization solvent with MeCN solvent and drying at 100 °C under vacuum for 12 h (red line).TGA plot of (b) M2-(R, R) (black line), as-crystallized Zn 6 (M2-(R, R)) 4 from DEF after exchanging the DEF crystallization solvent with MeCN solvent and drying at 100 °C in vacuum for 12 h (red line), and as-crystallized Zn 6 (M2-(R, R)) 4 from DEF after exchanging the DEF crystallization solvent with EtOH solvent followed by supercritical CO 2 (SCO 2 ) drying (blue line).All the samples were loaded onto the TGA pans in air, and we attribute the lower temperature weight losses (<100 °C) to water adsorbed by the samples before the measurements than MeCN from the solvent exchanges.The higher temperature weight losses are likely due to trace DEF solvent and water in the crystal pores.In the TGA plots, we observed far lower higher temperature weight losses for the supercritical CO 2activated Zn 6 (M2-(R, R)) 4 sample than the MeCN-exchanged Zn 6 (M2-(R, R)) 4 sample and subsequently activated CO 2 -activated Zn 6 (M2-(R, R)) 4 at room temperature under vacuum before performing gas sorption analysis.