Adsorptive Molecular Sieving of Styrene over Ethylbenzene by Trianglimine Crystals

The separation of styrene (ST) and ethylbenzene (EB) mixtures is of great importance in the petrochemical and plastics industries. Current technology employs multiple cycles of energy-intensive distillation due to the very close boiling points of ST and EB. Here, we show that the molecular sieving properties of easily scalable and stable trianglimine crystals offer ultrahigh selectivity (99%) for styrene separation. The unique molecular sieving properties of trianglimine crystals are corroborated by DFT calculations, suggesting that the incorporation of the nonplanar EB requires a significant deformation of the macrocyclic cavity whereas the planar ST can be easily accommodated in the cavity.

was added to make system mixed solvent with different polarity. Further, 1 ml EB and 1 ml of ST was added to that solution and kept for crystallization. After 3 days block prism shaped single crystals were found and suitable for single crystal X-ray diffraction. Bulk purities was verified by powder X-ray diffraction (PXRD).               Single Crystal X-ray Diffraction. Single crystals of the macrocycle trianglimine, were mounted in a Hampton cryoloop with Paratone® N oil cryoprotectant. In each case, a suitable crystal of appropriate size was selected from the mother liquor and immersed in Paratone® N oil and then it was mounted on the tip of a glass fiber and cemented using epoxy resin. Single crystal X-ray diffraction (SCXRD) was performed using a Bruker D8-Venture single crystal X-ray diffractometer equipped with a digital camera diffractometer using graphite-monochromated Mo-Kα radiation (0.71073 Å) at 120 K temperature. The linear absorption coefficients, scattering factors for the atoms and the anomalous dispersion corrections were taken from International Tables for X-ray Crystallography. Data integration and reduction were performed using SaintPlus 6.01 2 software. Absorption correction was performed by multi-scan method implemented in SADABS. 3 Space group was determined using XPREP implemented in APEX-III. 4 Structure was solved using Direct Methods (SHELXS-97) 5 and refined using SHELXL-2014 6 program package (full-matrix least squares on F 2 ) contained in WinGX. 7 For all the cases nonhydrogen atoms were refined anisotropically. All other hydrogen atoms are geometrically fixed using riding atom model. Attempts to identify the highly disordered solvent molecules have failed in some cases. Instead, a new set of F 2 (hkl) values with the contribution from the solvent molecules withdrawn was obtained by the SQUEEZE procedure implemented in PLATON. 8 The crystal data and refinement conditions for all the macrocycles trianglimine were collected in Table S1-S3. Table S1. Crystal data and structure refinements for EB@1.

Identification code EB@1
Empirical formula C128H130N12 # Two EB solvent molecules are found to be disordered and could not be refined. The final refinement was performed using the PLATON SQUEEZE by removing the solvents. 6,8 The presence of solvent molecules could easily be seen by the residual peaks located in the open channels. The estimated electron count of two ethylbenzene (EB) solvent molecules (116 electrons; found 116.1) was considered per unit cell for EB@1. Taking into account the number of solvent molecules which were squeezed we determined host to guest ratio as 1:1.

Computational Details
All the geometries were optimized with Gaussian 16 program packages, 9 using hybrid generalized gradient approximation (h-GGA) DFT functional PBE0. 10 To account for dispersion effects, empirical dispersion-corrected model at GD3 level were applied. 11 The electronic configuration of all atoms were described with the Ahlrichs split-valance polarization basis function Def2-SVP 12 The geometries were optimized without any symmetry constraints. Harmonic force constants were computed at the optimized geometries to characterize the stationary points as minima or saddle points. For further validation of energetics, single-point calculations were performed on the PBE0-D3/Def2-SVP optimized geometries using the long-range corrected hybrid density functional, ωB97xD, with damped atom−atom dispersion corrections 13 employing a valence triple-ζtype of basis set Def2-TZVP 14 . The solvent effects (ethyl benzene, ɛ = 2.4339) were evaluated implicitly by a self-consistent reaction field (SCRF) approach using the SMD continuum solvation model. 15 The rigid-rotor harmonic-oscillator approximation was applied for evaluating the thermal and entropic contributions that are needed to derive the enthalpies and Gibbs free energies. The The Gibbs free energy in the solution phase is a rough approximation, which is not appropriate for association and dissociation processes because there is a significant degree of denial of the translational degrees of freedom upon moving from the gas to solution phase. As a result, the Sackur-Tetrode equation, which is generally used to S26 calculate the gas-phase translational entropy, cannot be applied directly in the solution phase. Therefore, the solvation entropy (S298 Sol ) has been estimated as two-thirds of the gas-phase value. 16 Finally, the ∆G298 Sol was calculated as: ∆G298 Sol = ∆H298 Sol − T∆S298 Sol . Here, ∆S298 Sol represents the solvation entropy, which was estimated as 2/3 of the gas phase value.

Activation strain-distortion/interaction analysis
In this method the bond dissociation energy De of a complex AB is divided into the instantaneous interaction energy (∆Eint) and the distortion energy (∆Edis) according to eq 1.

∆E ( = −De) = ∆Eint + ∆Edis (1)
The ∆Edis is the energy that required to promote fragments A and B from their equilibrium geometries in the electronic ground state to the geometries in the corresponding intermediates/complexes, whereas ∆Eint is the actual interaction energy between the prepared fragments in the respective intermediates/complexes. The ∆Eint can be further divided into different components as explain in the following ETS-NOCV analysis.

Energy Decomposition Analysis (EDA)
We have performed the energy decomposition analysis (EDA) using the program package ADF 2019.301. 17 All analyses were performed by employing the BLYP 18 functional with triple-ξ-quality basis set using uncontracted Slater-type orbitals (STOs) augmented by two sets of polarization functions (TZ2P) for all atoms, with no frozen-core approximation for inner core electrons. 19 To compute the dispersion effects, we have utilized Grimme's D3BJ 20 empirical correction. The EDA method, developed independently by Morokuma 21 and Ziegler and Rauk, 22 gives a quantitative description of the chemical bonds in molecules.
In this method the ∆Eint can be divided into three main components: In eq 2, ΔEelstat is the quasiclassical electrostatic interaction energy between the fragments. ΔEPauli refers to the repulsive interactions between the fragments, which are caused by the fact that two electrons with the same spin cannot occupy the same region in space and can be calculated by enforcing the Kohn−Sham determinant on the superimposed fragments to obey the Pauli principle by antisymmetrization and renormalization. The stabilizing orbital interaction term ΔEorb is calculated in the final step of the energy partitioning analysis when the Kohn−Sham orbitals relax to their optimal form. Explanation of crystallographic alerts of ST@1:

PLAT331_ALERT_2_A Small Aver Phenyl C-C Dist C078 --C07K. 1.34 Ang.
Response: This is because of C078 atom is not ideally shaped, however, this does not indicate an incorrect atom-type assignment. We observed deviations of expected thermal parameters which are in agreement with a slight rotational disorder of this particular carbon atom.