Efficient separation of xylene isomers by a guest-responsive metal–organic framework with rotational anionic sites

The separation of xylene isomers (para-, meta-, orth-) remains a great challenge in the petrochemical industry due to their similar molecular structure and physical properties. Porous materials with sensitive nanospace and selective binding sites for discriminating the subtle structural difference of isomers are urgently needed. Here, we demonstrate the adaptively molecular discrimination of xylene isomers by employing a NbOF52−-pillared metal–organic framework (NbOFFIVE-bpy-Ni, also referred to as ZU-61) with rotational anionic sites. Single crystal X-ray diffraction studies indicate that ZU-61 with guest-responsive nanospace/sites can adapt the shape of specific isomers through geometric deformation and/or the rotation of fluorine atoms in anionic sites, thereby enabling ZU-61 to effectively differentiate xylene isomers through multiple C–H···F interactions. ZU-61 exhibited both high meta-xylene uptake capacity (3.4 mmol g−1) and meta-xylene/para-xylene separation selectivity (2.9, obtained from breakthrough curves), as well as a favorable separation sequence as confirmed by breakthrough experiments: para-xylene elute first with high-purity (≥99.9%), then meta-xylene, and orth-xylene. Such a remarkable performance of ZU-61 can be attributed to the type anionic binding sites together with its guest-response properties.


Powder X-ray diffraction structure analysis
Powder X-ray diffraction (PXRD) was carried out at room temperature on a Bruker D8 Advance diffractometer using Cu-Kα radiation (λ=1.5418 Å).

C8 aromatics vapor adsorption
ZU-61, SIFSIX-1-Cu, SIFSIX-2-Cu-i, and zeolite NaY were degassed at certain temperature until the pressure dropped below 10μm Hg. Nitrogen adsorption-desorption isotherms at 77 K were collected using ASAP 2020 Analyzer (Micromeritics). The single-component vapor adsorption isotherm of the activated samples were collected using ASAP 2020 Analyzer equipped with a vapor dosing tube. Each xylene isomers and ethylbeneze was purified by being degassed on ASAP 2020 through freeze-pump-thaw cycles.

Vapor-phase breakthrough tests
The vapor-phase multi-component breakthrough tests were carried out in a dynamic vapor breakthrough equipment. All experiments were conducted using a stainless-steel column (4.6 mm inner diameter × 50 mm). The column packed with adsorbent was firstly purged with He flow at room temperature. The mixed gases of pX, mX, oX (1:1:1) and pX, mX, oX, EB (1:1:1:1) in nitrogen were produced by nitrogen-blow bubble method. Nitrogen was passed through the container of C8 aromatics liquid mixture at desired rate. After the concentration of each isomer were tuned to the desired value, introduce the mixed gas in nitrogen at 25 mL/min. Outlet gas from the column was monitored using gas chromatography (GC-2010, SHIMADZU). The vapor mixture was separated by a capillary column (Agilent). It should be note that the time caused by the void volume of the pipeline and the column have been deducted when processing the breakthrough data.
To control and calculate the component ratios of the mixture, specific procedures were performed: 1) Before performing the breakthrough experiments, we firstly measured and got the Relative Quality Correction Factor of pX, mX, oX, and EB (Supplementary Table 5) on our gas chromatography (GC) with Flame Ionization Detector (FID). 2) According to the Antoine equation, the saturated pressure of each xylene isomer and EB can be calculated (Supplementary Table 6). Then, the vapor of xylene mixtures with known liquid compositions were injected and measured on the GC. Therefore, the corresponding relationship between each xylene isomer concentration and the response value of FID detector were obtained (Supplementary Table 7). Based on the above information, we can test the concentration/pressure of respective component. 3) For the vapor-phase breakthrough tests, the mixed vapors of pX, mX, oX (1:1:1) in nitrogen were produced by nitrogen-blow bubble method. Nitrogen was passed through the container of C8 aromatics liquid mixture at desired rate and injected into GC to measure the concentration. According to the tested concentrations, adjust the liquid mixtures until the gas concentration reached approximately 1:1:1.

The Antoine equation
The saturated pressure of pX, mX, oX and ethylbenzene Here, P is the pressure expressed in mmHg. A, B and C is a constant and the physical property data can be found in various manuals. T is the temperature expressed in ℃. The A, B, C values of pX, mX, oX and ethylbenzene are provided in the Supplementary Table 6.

Liquid-phase breakthrough tests
The liquid-phase multi-component breakthrough tests were carried out in a liquid breakthrough equipment. All experiments were conducted using a stainless-steel column (4.6 mm inner diameter × 100 mm). The liquid mixture of pX, mX, and oX (1:1:1) was diluted with heptane or hexane, the concentration of each xylene isomer is 0.01 mmol/mL. The fluid is pumped by the HPLC pump with flow rate of 0.2 ml/min. The concentration of effluent was detected using GC with FID detector. Furthermore, the column of ZU-61 was washed by para-diethylbenzene (diluted with heptane), and then on the next cycle the liquid mixture of pX, mX, and oX (1:1:1, diluted with heptane) was pumped into the column. In addition, we evaluated the separation performance of ZU-61 for binary pX/mX mixtures using the liquid-phase breakthrough equipment.

Differential Scanning Calorimetry (DSC)
Enthalpy of adsorption for xylene isomers was measured using the PE TGA and PE DSC 7. ZU-61 and zeolite NaY were activated on the PE TGA at certain temperatures under dry N2 flow until the weights remained stable. The activated samples were transferred to the PE DSC 7 to measure the adsorption enthalpy, the baseline was obtained under dry N2 flow at 25°C, then the N2 was changed to xylene isomers and the DSC signal were monitored to the obtain the heat of adsorption.

Density-functional theory calculations
The density-functional theory (DFT) calculations were performed to calculate the static binding energy using the CASTEP code. A semi-empirical addition of dispersive forces to conventional DFT was included in the calculation to account for van der Waals interactions. We used Vanderbilt-type ultrasoft pseudopotentials and generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) exchange correlation. A cutoff energy of 544 ev and a 1×1×2 kpoint mesh (generated using the Monkhosrt-Pack scheme) were found to be enough for the total energy to converge within 0.01 meV/atom. The structure of ZU-61 was first optimized. The optimized structures are good matches for the experimentally determined crystal structures of the coordination networks. The PX molecules were then introduced into the optimized structure, followed by a full structural relaxation. The initial location of pX molecules were obtained from the experiment XRD data. To obtain the gas binding energy, an isolated gas molecule placed in a supercell (with the same cell dimensions as the MOF crystal) was also relaxed as a reference. The static binding energy (at T = 0 K) was then calculated using: EB = E(MOF) + E(gas) -E(MOF+gas).
In order to confirm the adsorption configuration of oX and mX in ZU-61, the first-principle density functional theory (DFT) and plane-wave ultrasoft pseudopotentil implemented in the CASTEP code were performed. Firstly, six possible adsorption configurations of oX/mX in the structure of ZU-61 were constructed based on the experimental single-crystal X-ray diffraction data. Then, the single point energy calculations were performed to calculate the energy of these six possible configurations. All of the calculations were performed under the generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) exchange correlation. The cutoff energy of 544 eV and 1×1×2 k-point mesh with smearing 0.1 ev were adopted in the calculation. Finally, the energy of these configurations was compared to confirm the stable oX/mX configuration in ZU-61 structure ( Table 4). The higher negative value of single point energy corresponds to a more stable adsorption configuration.  t3 t1