Temperature‐Dependent Separation of CO2 from Light Hydrocarbons in a Porous Self‐Assembly of Vertexes Sharing Octahedra

Abstract Design of flexible porous materials where the diffusion of guest molecules is regulated by the dynamics of contracted pore aperture is challenging. Here, a flexible porous self‐assembly consisting of 1D channels with dynamic bottleneck gates is reported. The dynamic pendant naphthimidazolylmethyl moieties at the channel necks provide kinetic gate function, that enables unusual adsorption for light hydrocarbons. The adsorption for CO2 is mainly dominated by thermodynamics with the uptakes decreasing with increasing temperature, whereas the adsorptions for larger hydrocarbons are controlled by both thermodynamics and kinetics resulting in an uptake maximum at a temperature threshold. Such an unusual adsorption enables temperature‐dependent separation of CO2 from the corresponding hydrocarbons.


S1.1 Starting materials
S1] All other reagents were purchased from commercial sources and used without purification.

S1.2 Physical measurements
Thermogravimetric analyses (TGA) were performed using a TG/DTA6300 system at a rate of 10 °C/min.Powder X-ray diffraction (PXRD) patterns were obtained on a Rigaku 2100 diffractometer using Cu-Kα radiation with flat plate geometry.High-resolution Electrospray Ionization Time of Flight Mass Spectrometry (ESI-TOF-MS) measurements were performed on a DECAX-30000 LCQ Deca XP system.Scanning electron microscopy (SEM) was performed on a Hitachi SU1510 scanning electron microscope.The sample of 1 was desolvated at 100 °C under high vacuum for 5 h to remove the guest molecules before gas sorption measurements.Gas sorption isotherms of activated-1 were measured on a Micromeritics ASAP 2020 surface area analyzer or on a 3Flex instrument.Both the instruments are produced by Micromeritics Instrument Corporation for conventional gases sorption measurements.The maximum vacuum degree of 3Flex is as low as 1.3×10 -9 bar, while thar for ASAP 2020 is only about 1.0×10 -5 bar.In addition, the equilibrium time for isotherm measurement of 3Flex is automatically set to longer than that of ASAP 2020.

S1.3 Crystallography
S2] The hydrogen atoms are geometrically generated and refined using a riding model.The PLATON/SQUEEZE procedures [S3] were used to treat the highly disordered solvents in the voids of the structure of 1.The X-ray crystallographic coordinates for structures reported in this article have been deposited at the Cambridge Crystallographic Data Centre (CCDC), under deposition numbers CCDC 2300877−2300882.These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif. Detail crystallographic data are listed in Table S1.

S 1.4 Breakthrough measurements.
Breakthrough experiments were performed on a fixed bed.A stainless-steel column with a length of 180 mm and an internal diameter of 3 mm was used for sample packing.The flow rates of all gases are regulated by mass flow controllers, and the effluent gas stream from the column is monitored by gas chromatography (GC) detector.All measurements were performed following a protocol established by literatures.The crystals of 1 (0.896 g) were packed into a stainless-steel column tightly.After the column was activated with the He flow under 100 °C for 5 h, a mixture flow was dosed into the column.Breakpoints were determined when the first peak was detected.

S1.5 Calculation of the isosteric heat of adsorption (Qst).
A virial-type expression of the above form was used to fit the combined isotherm data of activated-1 at 273, 298, and 308 K, where P is the pressure described in mmHg, N is the adsorbed amount in mg/g, T is the temperature in K, ai and bi are virial coefficients, and m and n are the number of coefficients used to describe the isotherms.Qst is the coverage-dependent heat of adsorption and R is the universal gas constant.

S1.6 Calculation of the selectivities by IAST.
Single-component gas equilibrium adsorption isotherms were fitted with the Langmuir-Freundlich model, given by the following equation: where N is the amount of adsorbed gas (mmol g -1 ), p is the bulk gas phase pressure (atm), A1 is the adsorption saturation capacities for site 1 (mmol g -1 ), b1 is the affinity coefficient of site 1 (1/kPa), c1 is the Langmuir-Freundlich exponent (dimensionless) for the adsorption sites A.
The parameters of A1, b1, and c1 were used to predict the adsorption selectivities based on IAST, which is finally defined as: where S is the ideal selectivity of component 1 over component 2, xi and yi are the mole fractions of component i (i = 1, 2) in the adsorbed and bulk phases, respectively.

S1.7 Evalution of the diffusion rates of C2H2 in activated-1
The CO2-adsorbed and C2H2-adsorbed structures of activated-1 were simulated by Grand Canonical Monte Carlo (GCMC), and the Molecular dynamics (MD) method has been used to analyze the mean square displacement (MSD).The energy minimization and geometry optimization process of these models have been performed in the forcite module.And then, the annealing process of these molecular models has been performed to obtain stable conformations.Finally, MD simulations of these molecular models have been performed with a 10 ps dynamic simulation under equilibrium run with constant volume and temperature (NVT) and a 5 ps dynamic simulation under the constant volume and energy (NVE) ensemble.After MD simulation, the diffusion coefficient have been analyzed in the forcite analysis module.All the above simulation results have been completed by Materials Studio software.Table S1.Crystallographic data of 1 and activated-1.

Figure S1 .
Figure S1.Molecular structures of L and btcn.

Figure S3 .
Figure S3.(a) Side view of the channel wall formed by hexagonally aligned π-stacked helical columns, the different helical columns are shown in different colors.(b) One of the π-stacked helical columns possessing a pitch of 61.2 Å.

Figure S4 .
Figure S4.TGA for 1. TG analyses show a weight loss below 100 °C which corresponds to the loss solvent molecules.

Figure S5 .
Figure S5.PXRD patterns of 1 showing the exceptional thermal and chemical stabilities and the exceptional regeneration property.The desolvated sample was activated at 100 °C under vacuum for 10 h.

Figure S6 .
Figure S6.Scheme showing the regenerating process of 1.

Figure S7 .
Figure S7.The N2 sorption data at 77 K, Ar2 sorption data at 87 K, and CO2 sorption data at 195 K of desolvated-1, inset): pore size distribution of activated-1.The pore volume was calculated based on the CO2 uptake at P/P° = 0.96, and the pore size distribution was extracted with the Non-Local Density Functional Theory (NLDFT) method[S4]  .

Figure S8 .
Figure S8.(a) The triangular gate of the channels in 1.(b) The gate size increasing with temperature.(The size is described using the shortest H•••H distance betweem adjacent naphthimidazolylmethyl arms).

Figure S10 .
Figure S10.(a) The virial fitting of the adsorption of activated-1 towards CO2.(d) The isosteric heats of adsorption (Qst) for CO2 on activated-1 calculated by the virial method.

Figure S12 .
Figure S12.The propylene (a) and propyne (b) adsorption isotherms accompanying with elapsed times at 298 K.

Figure S20 .
Figure S20.Comparison of the sorption isotherms of CO2 with N2 (a) and methane (b) at 298 K. (c) Comparison of the sorption isotherms of ethylene and acetylene at 298 K.(d) Comparison of the sorption isotherms of propylene and propyne at 298 K measured on 3Flex.