Switching Adsorbent Layered Material that Enables Stepwise Capture of C8 Aromatics via Single-Crystal-to-Single-Crystal Transformations

Separation of the C8 aromatic isomers, xylenes (PX, MX, and OX) and ethylbenzene (EB), is important to the petrochemical industry. Whereas physisorptive separation is an energy-efficient alternative to current processes, such as distillation, physisorbents do not generally exhibit strong C8 selectivity. Herein, we report the mixed-linker square lattice (sql) coordination network [Zn2(sba)2(bis)]n·mDMF (sql-4,5-Zn, H2sba or 4 = 4,4′-sulfonyldibenzoic acid, bis or 5 = trans-4,4′-bis(1-imidazolyl)stilbene) and its C8 sorption properties. sql-4,5-Zn was found to exhibit high uptake capacity for liquid C8 aromatics (∼20.2 wt %), and to the best of our knowledge, it is the first sorbent to exhibit selectivity for PX, EB, and MX over OX for binary, ternary, and quaternary mixtures from gas chromatography. Single-crystal structures of narrow-pore, intermediate-pore, and large-pore phases provided insight into the phase transformations, which were enabled by flexibility of the linker ligands and changes in the square grid geometry and interlayer distances. This work adds to the library of two-dimensional coordination networks that exhibit high uptake, thanks to clay-like expansion, and strong selectivity, thanks to shape-selective binding sites, for C8 isomers.

Quest diffractometer equipped with a CMOS detector and IμS microfocus X-ray source Cu Kα (λ =1.54178 Å).APEX4 was used for collecting, indexing, integrating and scaling the data. 1 Absorption corrections were performed by multi-scan method. 2 Space groups were determined using XPREP 3 as implemented in APEX4.All the scaled data were solved using intrinsic phasing method (XT) 4 and refined on F 2 using SHELXL 5 inbuilt in OLEX2 v1. 5 (2020) program. 6All nonhydrogen atoms present in the frameworks were refined anisotropically.Hydrogen atoms were located at idealized positions from the molecular geometry and refined isotropically with thermal parameters based on the equivalent displacement parameters of their carriers.Crystallographic data reported in this paper are summarized in Tables S3 and S4.These crystal structures have been deposited to the Cambridge Crystallographic Data Centre.

Powder X-ray diffraction (PXRD).
Diffractograms were recorded using a PANalytical Empyrean™ diffractometer equipped with a PIXcel 3D detector operating in scanning line detector mode with an active length of 4 utilizing 255 channels.The diffractometer is outfitted with an Empyrean Cu LFF (long fine-focus) HR (9430 033 7310x) tube operated at 40 kV and 40 mA and Cu Kα radiation (λα = 1.540598Å) was used for diffraction experiments.Continuous scanning mode with the goniometer in the theta-theta orientation was used to collect the data.Incident beam optics included the Fixed Divergences slit with anti-scatter slit PreFIX module, with a 1/8° divergence slit and a 1/4° anti-scatter slit, as well as a 10 mm fixed incident beam mask and a Soller slit (0.04 rad).Divergent beam optics included a P7.5 anti-scatter slit, a Soller slit (0.04 rad), and a Ni-β filter.In a typical experiment, 25 mg of sample was dried, ground into a fine powder and was loaded on a zero background silicon disks.
The data was collected from 3°-50° (2θ) with a step-size of 0.0131303° and a scan time of 30 seconds per step.Crude data were analyzed using the X'Pert HighScore Plus™ software V 4.1 (PANalytical, The Netherlands).

Thermogravimetric analysis (TGA)
Thermograms were recorded under N2 atmosphere using TGA instrument TA Q50 V20.13 Build 39.Aluminium pans and a flow rate of 60 cm 3 min -1 for the nitrogen gas were used for the experiments.The data was collected in the High Resolution Dynamic mode with a sensitivity of 1.0, a resolution of 4.0, and a temperature ramp of 10°C min -1 up to 550°C.with a 1/4° divergence slit and a Soller slit (0.04 rad).Divergent beam optics included a P7.5 antiscatter slit, a Soller slit (0.04 rad), and a Ni-β filter.In a typical experiment, 20 mg of sample was loaded on a zero background sample holder made for Anton Paar TTK 450 chamber.The data was collected from 4°-40° (2θ).Crude data were analyzed using the X'Pert HighScore Plus™ software V 4.1 (PANalytical, The Netherlands).The sample was heated up to 350°C under N2 atmosphere.

Low Pressure Gas Adsorption Studies
For gas sorption experiments, ultrahigh-purity gases were used as received from BOC Gases Ireland: research-grade CO2 (99.995%), and N2 (99.998%).Adsorption experiments (up to 1 bar) were performed on Micromeritics Tristar II 3030 instrument.Before sorption measurements, activation of sql-4,5-Zn- was achieved by degassing the air-dried samples on a SmartVacPrep™ using dynamic vacuum overnight.About 100 mg of activated samples were used for the measurements.A Julabo temperature controller was used to maintain a constant temperature in the bath throughout the experiment.The low temperature at 77 K and 195 K were controlled by a 4 L Dewar filled with liquid N2 and a mixture of dry ice/acetone, respectively.At every interval of two independent isotherms recorded for any sorbent, samples were regenerated by degassing under high vacuum, before commencing the next sorption experiment.

Dynamic Vacuum Vapor Sorption
Dynamic vapor sorption measurements were conducted using a Surface Measurement Systems DVS Vacuum at 298 K. Activated samples of sql-4,5-Zn-β were further degassed under high vacuum (1x10 -4 Torr) in-situ and stepwise increase in relative pressure were controlled by equilibrated weight changes of the sample (dM/dT = 0.01%/min) from 0 to 95%.Vacuum pressure transducers were used with the ability to measure from 1x10 -6 to 760 Torr with a resolution of 0.01%.Approximately 10 mg of sample was used for each experiment.The mass of the sample was determined by comparison to an empty reference pan and recorded by a high resolution microbalance with a precision of 0.1 μg.

C8 aromatics selectivity studies using Nuclear Magnetic Resonance (NMR)
~30 mg samples of sql-4,5-Zn-β were separately immersed in equimolar (1 g each) binary/ternary/quaternary liquid of C8 aromatics at room temperature for one day.PXRD patterns (Figure S2) and TGA curves (Figure 2) reveal that sql-4,5-Zn-β completely adsorbed the C8 aromatics.Then the saturated samples were filtered and air-dried (ca. 10 min) under ambient conditions (ca. 25 º C) to remove xylenes adhering to the surface of samples.After that, the samples were soaked in 1 mL DMSO-d6 for two days.
The supernatant of each sample after soaking in DMSO-d6 were filtered and collected to measure 1 H NMR spectra (JEOL ECX400 NMR spectrometer).The reliability of NMR has been verified in our recent paper. 7The selectivity is defined as: where S is the selectivity of component i relative to component j, xi and xj are the mole fractions of components i and j in the adsorbed phase, and yi and yj are the mole fractions of components i and j in the liquid phase.For equimolar binary phase, the selectivity can be simplified as The ratio of xi/xj can be derived from the integrated area ratio of corresponding methyl groups or methylene group of C8 aromatics in NMR spectra.When component i and j are both xylene isomers, the selectivity is defined as: Where qi and qj are the relatively integrated area of corresponding methyl groups of xylene isomers.
When component i is one of xylene isomers while j is ethylbenzene, the selectivity is defined as: Where qi is the relatively integrated area of corresponding methyl groups (including 6 H) of xylene isomers, while qj is the relatively integrated area of corresponding methylene group (including 2 H) of ethylbenzene.

C8 aromatics selectivity studies using Gas Chromatography (GC)
Vapour phase: 20 mg samples of sql-4,5-Zn-β were separately kept in small vials which stand inside bigger capped vials containing the mixtures of C8 aromatics at 303 K for three days.Then the saturated samples were air-dried under ambient conditions (ca. 25 º C) to remove xylenes adhering to the surface of samples until the sample is free to move.After that, they were soaked in 1.5 mL CH2Cl2 for about three days, allowing C8 aromatics to be completely extracted by with a ramp rate of 10°C/min to a maximum temperature of 180°C.The injector and detector were kept at 220°C and nitrogen was used as carrier gas with a flow rate of 1 ml/min. 1 μl of each liquid sample was injected through the GC inlet with a split ratio of 100:1 and a split flow rate of 142.45 ml/min.Dichloromethane HPLC/GC grade 99.9% (Sigma-Aldrich) was used as eluent and solvent for the standard solutions of 500 ppm of all the C8 aromatic isomers.With the method described above, we could obtain separate retention times for PX, MX, OX and EB.

Calculation of the Langmuir surface area
The well-known Langmuir isotherm model can be expressed by the following equation: Where Q/(cm 3 g -1 ) is the amount adsorbed; Q0/(cm 3 g -1 ) is the saturated amount adsorbed; P/mmHg is the equilibrium pressure; and b/mmHg -1 is the adsorption affinity.
A line expression for the Langmuir equation can be written as following: A least-squares fit is performed on the ( Using the results of the above calculations, the Langmuir surface area can be calculated as following: Where Sg is the Langmuir surface area (m 2 /g); Am = molecular cross-sectional area (nm 2 ) of adsorbate i.e. 0.1700 nm 2 for CO2, and NA = 6.02 × 10 23 .

The Antoine equation
For each batch of experiments, the saturated pressure of each xylene isomers was calculated by the following equation:

Survey of crystallographic and topological databases
The list of MOMs having sql net topology was obtained from the TTO TOPOS database 8

Interactions between Framework and C8 Aromatics.
Table S13.A summary of interaction between C8 aromatics and framework.

1. 4
In-situ Variable Temperature Powder X-ray Diffraction (VT-PXRD) Diffractograms at different temperature were recorded using a PANalytical X'Pert Pro-MPD diffractometer equipped with a PIXcel3D detector operating in scanning line detector mode with an active length of 4 utilizing 255 channels.Anton Paar TTK 450 stage coupled with the Anton Paar TCU 110 Temperature Control Unit was used to record the variable temperature diffractograms.The diffractometer is outfitted with an Empyrean Cu LFF (long fine-focus) HR (9430 033 7300x) tube operated at 40 kV and 40 mA and CuKα radiation (λα = 1.54056Å) was used for diffraction experiments.Continuous scanning mode with the goniometer in the theta-theta orientation was used to collect the data.Incident beam optics included the Fixed Divergences slit,

CH2Cl2.
The liquid was collected for GC measurements to calculate the ratios and selectivity coefficients.The samples were also submitted to second extraction in CH2Cl2 for GC measurements to make sure the extraction experiments in the first time are complete.We calculated the saturated vapor pressure (SVP) of C8 aromatics based on Antoine equation under 303K (the equation is presented in supporting information).The values of SVP of C8 aromatics and the ratios of liquid phase are listed below and provided equimolar gaseous phases of C8 aromatics.Saturated vapor pressure (SVP) of C8 aromatics at 303K: SVP(OX) = 8.85841 SVP(MX) = 11.04598SVP(PX) = 11.62993SVP(EB) = 12.61894The ratios of liquid phase to get equimolar gaseous phases of C8 aromatics: Ratio (MX/OX) = 1:1.2470Ratio (PX/OX) = 1:1.3129Ratio (EB/OX) = 1:1.4245Ratio (PX/MX) = 1:1.0529Ratio (EB/MX) = 1:1.1424Ratio (EB/PX) = 1:1.0850Ratio (EB/MX/OX) = 1:1.1424:1.4245Ratio (PX/MX/OX) = 1:1.0529:1.3129Ratio (EB/PX/OX) = 1:1.0850:1:1.4245Ratio (EB/PX/MX) = 1:1.0850:1.1424Ratio (EB/PX/MX/OX) = 1:1.0850:1.1424:1.4245Liquid phase: 20 mg samples of sql-4,5-Zn-β were separately immersed in equimolar (1 g each) binary/ternary/quaternary liquid of C8 aromatics at room temperature for one day.Then the saturated samples were filtered and air-dried (ca. 10 min) under ambient conditions (ca. 25 º C) to remove xylenes adhering to the surface of samples.After that, the samples were soaked in 1.5 mL CH2Cl2 for two days, allowing C8 aromatics to be completely extracted by CH2Cl2.The liquid was collected for GC measurements to calculate the ratios and selectivity coefficients.The samples were also extracted by CH2Cl2 for the second time for GC measurements to make sure the extraction experiments in the first time are complete.The peak areas of individual C8 aromatics shown in chromatograms were used to calculate the selectivity coefficients by the following method.S10 The selectivity coefficient of component i relative to component j is defined as: where xi and xj are the mole fractions of components i and j in the adsorbed phase, and yi and yj are the mole fractions of components i and j in the liquid phase.The following equation was used to obtain the adsorption selectivity for quaternary mixture adsorption (SPX/OME) GC instrument settings: The analyses were carried out on an Agilent 6890A gas chromatograph fitted with a 7683B ALS (automated liquid sampler) equipped with a flame ionization detector (FID).The column used was an Agilent DB-Wax (Length: 30 m, Inner diameter: 320 μm, Film thickness: 0.25 μm).An initial temperature of 40°C and initial hold time of two minutes were used

PQ 1 Q
, P) designated pairs where P Q is the independent variable and  is the dependent variable.The following are calculated: a) Slope ( 0 , g/cm 3 STP) b) Y-intercept ( 0 1 bQ , g• mmHg/cm 3 STP) c) Error of the slope (g/cm 3 STP) d) Error of the y-intercept (g• mmHg/cm 3 STP)

12 .
Figure S15.Rotation of the rings in the ditopic N-donor ligand for guest free phase of sql-4,5-Zn- and the porous

Figure S22 . 4 , 5 -S32Figure S23 . 4 , 5 -
Figure S22.Magnified 1 H NMR spectrum recorded using the DMSO-d6 extract of C8 aromatics obtained from sql-4,5-Zn that was prior subjected to the equimolar binary liquid of PX/MX until saturated.Since the peaks of MX and PX at 2.25 and 2.26 are combined and it is hard to get a good discrimination of these two compounds, so the peaks at 7.05 (four H atoms in benzene ring for PX) and 7.11, 7.13, 7.14 (one H atom in benzene ring for MX) were chosen to calculate the selectivity.The selectivity of PX/MX is 1.5.

Figure S24 . 4 , 5 -
Figure S24.Magnified 1 H NMR spectrum recorded using the DMSO-d6 extract of C8 aromatics obtained from sql-4,5-Zn that was prior subjected to the equimolar binary liquid of EB/PX until saturated.

Figure S26 .
Figure S26.Langmuir fit corresponding to the second step from the 195 K CO2 data for sql-4,5-Zn-.Squares and line represent the experimental data and linear fitted data, respectively.

Figure S29 .
Figure S29.Gas chromatograms (GC) used to quantify the selectivity coefficients of sql-4,5-Zn- after being soaked in the equimolar liquid binary mixtures.

Figure S30 .
Figure S30.GC used to quantify the selectivity coefficients of sql-4,5-Zn- after being soaked in the equimolar liquid ternary or quaternary mixtures.

Figure S31 .
Figure S31.GC used to quantify the selectivity coefficients of sql-4,5-Zn- that was prior subjected to the equimolar vapor binary mixtures until saturated.

Figure S32 .
Figure S32.GC used to quantify the selectivity coefficients of sql-4,5-Zn- that was prior subjected to the equimolar vapor ternary or quaternary mixtures until saturated.

Table S1 .
Comparison of physicochemical properties for C8 aromatics.

Table S2 .
A summary of reported bent dicarboxylate ligand based sql networks assembled from parallel layers with paddlewheel units found in the database survey.
(version: Dec 2021); valence-bonded MOFs in standard representation were used.The listed MOM crystal structures from the TTO database were analyzed using to the Cambridge Structural Database (CSD version 5.43, March 2022) through ConQuest software. 9

Table S3 .
Crystallographic data and refinement parameters of sql

Table S5 .
Crystallographic data and refinement parameters of sql-

Table S9 .
Selectivities of C8 aromatics on some representative metal organic materials (MOMs).Note: NG. refers to not given; a Measured by HPLC or GC; b Liquid phase binary mixture (equimolar) separation from 1 H NMR; c Vapor phase binary mixture (equimolar) separation from GC; d Liquid phase binary mixture (equimolar) separation from GC.

Table S10 .
Flexible or switching adsorbents for the separation of C8 aromatics.

Table S11 .
Some representative OX selective MOMs, arranged in descending order of year.Note: NG. refers to not given; a 303 K; b no uptake; c 7.1 mbar, 333 K; d 398 K.
Note: NG. refers to not given.

Table S14 .
Separation performance of binary, ternary and quaternary C8 aromatics (liquid phase, equimolar) from GC Result.

Table S15 .
Separation performance of binary C8 aromatics (vapor phase, equimolar) from GC Result at 303K.

Table S16 .
Separation performance of binary C8 aromatics (liquid phase) from 1 H NMR.

Table S17 .
The values of A, B and C for C8 aromatics from the Antoine equation.