An efficient approach to explore the adsorption of benzene and phenol on nanostructured catalysts: a DFT analysis

Abstract

Adsorption of benzene and phenol on the 8T cluster model of ZSM-5 and Al-ZSM-5 catalysts, defined as ((H)₃SiO)₃–Si–O–Si–(OSi(H)₃)₃ and ((H)₃SiO)₃–Si–O(H)–Al–(OSi(H)₃)₃ structures, respectively, has been investigated comparatively using B3LYP, M06-2X, and wB97XD functionals employing the 6-311++G** standard basis set. Geometric parameters predict one and two types of hydrogen bonding in the guest-ZSM-5 and guest-Al-ZSM-5 complexes, respectively. Variations of adsorption energy, isotropic chemical shifts, δ_iso, of ¹H, ¹⁷O, ²⁷Al, and ²⁹Si atoms contributing in the hydrogen bonding as well as quadrupole coupling constant, C_Q, and asymmetry parameter, η_Q, of ²H, ¹⁷O, and ²⁷Al atoms have been well correlated with the strength of the hydrogen bonds. Atom in molecules (AIM) calculations showed a covalent nature for the hydrogen bonds in the phenol⋯Al-ZSM-5 adsorption complex. Furthermore, based on AIM, NQR and NMR, the C–H⋯O and O–H⋯π hydrogen bonds have been confirmed in the benzene adsorbate zeolite, which may highlight a crucial feature of the adsorption of benzene molecules inside the pores of zeolites. The differences in the adsorption behavior between benzene and phenol on the ZSM-5 and Al-ZSM-5 are attributed to the differences in the strength of the hydrogen bonding interactions. Finally, Al-ZSM-5 appears to be an efficient adsorbent for phenol and benzene.

---

Introduction

Zeolites, which are employed in a wide range of important petroleum and petrochemical processes as adsorbents and are the most important heterogeneous catalysts for separation of hydrocarbons and other processes, are crystalline, nano porous molecular sieves. They owe their extensive applicability to their unique characteristics such as tunable acidity, high surface area, uniform pore size, and good thermal/mechanical stability.¹ ZSM-5 catalysts are one of the most valuable catalysts with MFI-type topology. They are crystalline aluminosilicates with high silica to alumina ratio, based on frameworks with periodic arrangements of channels or pores, widely used in chemical industries including petroleum refining. The ZSM-5 network consists of two sets of intersecting 10-ring channel systems, running perpendicular to each other: straight circular (0.53 × 0.56 nm in diameter) and sinusoidal elliptical (0.51 × 0.55 nm in diameter) channels. Trivalent Al substitutions in the tetravalent Si framework cause the acidity of this framework structure, leading to a charge imbalance. The charge imbalance, which is compensated with a counter ion, is centered on a bridging oxygen atom located between aluminum and silicon atoms. When the counter ion is a proton, a Brönsted acid site, available for catalytic reactions is formed.^2,3

Probing the type of interaction between guest molecule and ZSM-5 is of great importance because a lot of industrially important processes begin with the adsorption of guest molecules inside the pores of the ZSM-5. Therefore, adsorption of guest molecules, especially hydrocarbons in the ZSM-5, is widely considered experimentally and theoretically.^4–10 Similar dimensions of zeolite channels to the diameter of benzene molecule (0.58 nm) as well as the acid sites in MFI framework makes the adsorption of aromatic hydrocarbons in MFI type zeolites particularly interesting. Adsorption of aromatic hydrocarbons such as benzene and phenol is significant. Benzene is directly oxidized to phenol using ZSM-5 type zeolite and nitrous oxide as the catalyst and oxidizing agent, respectively, as an alternative to the traditional cumene process. Good performance of [Al] MFI in selective oxidation of benzene to phenol has been reported by Hensen et al.^6–9,11 The adsorption of basic probe molecules is mostly based on the formation of hydrogen bonding between the ZSM-5 and guest molecule and formation of A⋯B adducts, where A and B are ZSM-5 and weak base, respectively. C–H⋯O and O–H⋯π bonds are well considered as hydrogen bonds which are associated with unconventional donor/acceptor functional group.^12–14 C–H could be a hydrogen bond donor concerning C–H⋯O hydrogen bonds according to Huggins's suggestion in 1936.¹⁵ In addition, given a potential relevance to heterogeneous catalysis, O–H⋯π hydrogen bond has attracted the interest of chemists. Electrons could function as hydrogen bond acceptors under some circumstances. The structural influence and functional relevance of hydrogen bonds by the application of a variety of experimental and theoretical means are still subject to hot debates among material scientists, biochemists, and others interested in fundamental structure/property issues.¹⁶ Thus, valuable information regarding adsorption process of guest molecules can be provided by the determination of hydrogen bonds and comparison of their strength. In order to explore the electronic structure of compounds, several spectroscopic methods such as solid state nuclear magnetic resonance (NMR) and nuclear quadrupole resonance (NQR) spectroscopy have been employed. NQR spectroscopy is one of the very sensitive of these techniques to characterize the electronic environment and reveal the details of charge distribution around the nuclei with spin angular momentum of over one half (I > 1/2).^17–19 Another powerful, high potential tool for the exploration of the local electronic structure of framework atoms in the solid catalysts is based on solid state NMR spectroscopy, which offers invaluable evidence about the electrostatic interactions.^20,21 Hydrogen bondings, at the other extreme, can be described as far as delocalization (or charge transferring) effects are concerned. These suggest a covalent nature for these interactions.²²

Despite several studies on adsorption of benzene and phenol on ZSM-5 zeolite,^23–27 few detailed computational works have considered the type of interaction between benzene and phenol with ZSM-5 host lattice through NMR, NQR, and AIM analysis. Our understanding of adsorption process of aromatic hydrocarbons may be facilitated by the investigation of interactions and comparison of strength of these complexes. Importantly, we are interested in the explanation of the nature and consequences of hydrogen bonds. The type of interactions and relationship between alternation of spectroscopic parameters and strength of hydrogen bondings are revealed by quantum chemical calculations. We have systematically implemented a quantum mechanical investigation of charge distribution on the benzene and phenol adsorbate complex on the solid acids of different strengths including ZSM-5 and Al-ZSM-5 zeolites by the application of NMR and NQR parameters as well as Natural Bond Orbital (NBO) and quantum theory of atoms in molecules (QTAIM) analysis to reach this purpose. Combination of nuclear magnetic resonance (NMR) and nuclear quadrupole resonance (NQR) techniques thus presents a powerful tool for the investigation of the type of interactions in the adsorption complexes and successful interpretation of the observed results. NBO and QTAIM analyses finally complete the structural analysis.

Computational procedure

Zeolite cluster models

The cluster models are presumed based on the crystallographic coordinates of ZSM-5 zeolites reported by van Koningsveld et al.²⁸ It is compulsory to model unit cell of ZSM-5 zeolite with 288 atoms (96 silicon and 192 oxygen atoms) to obtain reliable results. Modeling of a large unit cell of ZSM-5 zeolite is, nevertheless, computationally expensive. Thus, a fragment of appropriate size from the zeolite framework must be chosen. Clusters containing eight tetrahedral centers are reliable models compared to experimental results, as indicated by previous works.^29–31 This fragment of zeolite framework includes the intersection of the straight and sinusoidal channels which is accessible to the interaction between bridging hydroxyl group and guest molecules as a catalytic active site. Therefore, in the present study, the clusters containing eight tetrahedral centers (8T), defined as ((H)₃SiO)₃–Si–O–Si–(OSi(H)₃)₃ and ((H)₃SiO)₃–Si–O(H)–Al–(OSi(H)₃)₃ structures, in which one of the Si sites had been substituted by an Al atom in the latter, were chosen. In order to obtain a neutral cluster, the boundary Si atoms were saturated with hydrogen atoms. Cluster models are referred to as 8T-ZSM-5 and 8T-Al-ZSM-5, respectively, from here on for the sake of brevity.

Theoretical background and methods of analysis

GAMESS program package has been used to carry out all quantum mechanical calculations in this work.³² B3LYP, M06-2X, and wB97XD functionals in conjugation with 6-311++G** basis set that contains polarization and diffuse functions on all atoms has been applied to perform geometry optimization of benzene and phenol adsorbate complexes. B3LYP has been selected as a conventional functional has been widely used in literature. Additionally, M06 and wB97XD have been used as specifically parameterized functionals for systems having dominant weak dispersive interactions in order to improve the description of noncovalent and π-stacking interactions. Specially, M06-2X has been employed to explore the zeolite-catalyzed reactions.^33–36 Coordinates of all atoms in the cluster, excluding boundary Si and H atoms, kept fixed in their crystallographic positions, were allowed to relax throughout the geometry optimization procedure. The optimized cluster models have been subjected to chemical shielding (CS) and electric field gradient (EFG) tensors at the as well as NBO and QTAIM analyses.

A traceless, second-rank tensor, whose principal components are defined as |q_zz| ≥ |q_yy| ≥ |q_xx|, can be used to describe the EFG. Two other experimentally measurable NQR parameters include the nuclear quadrupole coupling constant, C_Q, and its associated asymmetry parameter, η_Q. The amount of interaction energy between the electric quadrupole moment, e_Q, of the nucleus and the EFG at the quadrupole nucleus site due to the anisotropic charge distribution in the system can be shown by C_Q. This gives information regarding electron distribution in the molecule and is defined as:^17,37

C_Q(MHz) = e²Qq_zz/h (1)

where e is the electron charge and h is the Planck constant. The standard values of Q were employed as Q(²H) = 2.86 mb, Q(¹⁷O) = 25.58, and Q(²⁷Al) = 146.6 mb, as reported by Pyykkö.³⁸ The deviation of EFG tensors from cylindrical symmetry at the site of quadrupolar nucleus is measured by asymmetry parameter, η_Q, using eqn (2):

η_Q = |q_yy − q_xx|/|q_zz|, 0 ≤ η_Q ≤ 1 (2)

To carry out chemical shielding calculations, the gauge independent atomic orbital (GIAO)³⁹ method was employed. The chemical shielding tensor is one of the important NMR parameters, which shows the magnetic shielding effect of electrons, especially valence electrons around atomic nucleus. σ₃₃ > σ₂₂ > σ₁₁ are defined as the three principal components of the corresponding CS tensor. The chemical shielding isotropy, σ_iso, is used in addition to the three principal components, σ_iso = (σ₃₃ + σ₂₂ + σ₁₁)/3 to describe a CS tensor. The predicted ¹H and ²⁹Si, ¹⁷O, and ²⁷Al chemical shifts (in ppm) are derived from equation δ_iso = σ_iso,ref − σ_iso,cal, in which σ_iso,cal is the absolute shielding and σ_iso,ref refers to the absolute chemical shielding of tetramethylsilane (TMS), liquid water and chabasite, respectively.^20,40–42 Natural Bond Orbital (NBO) analysis⁴³ used to calculate natural charges for cluster models were performed on wave functions calculated at the M06-2X/6-311++G** level. AIM2000 package was used to implement the AIM analysis for cluster models considered at the M06-2X/6-311++G** computational level.⁴⁴

Finally, for each cluster, the adsorption energy ΔE_ads of benzene and phenol on the zeolites has been calculated which is defined as the energy difference between the complex system and the sum of the separated fragments.

Results and discussion

For the purpose of clarity, this part of the work is divided into three different subdivisions including geometry, NMR and NQR, NBO and AIM parameters.

Geometrical aspects of adsorption

The geometries and hydrogen bonds of the optimized adsorption complexes at M06-2X/6-311++G** are depicted in Fig. 1. The calculated bond lengths and angles between the most important atoms of these cluster models, as well as full details of geometrical parameters of hydrogen bonds are summarized in Table 1. Throughout this work, Z, P and B subscripts denote zeolite, phenol and benzene, respectively.

Fig. 1 Representation of hydrogen bonding interactions in: (a) 8T-ZSM-5⋯benzene, (b) 8T-ZSM-5⋯phenol, (c) 8T-Al-ZSM-5⋯benzene, and (d) 8T-Al-ZSM-5⋯phenol cluster models at M06-2X/6-311++G** level.

Table 1 Selected optimized geometrical parameters (bond lengths in pm and angles in degrees) of the cluster models before and after adsorption calculated at the M06-2X/6-311++G** level of theory. The values in parentheses represent the data before adsorption

Cluster models
8T-ZSM-5⋯benzene		8T-ZSM-5⋯phenol		8T-Al-ZSM-5⋯benzene		8T-Al-ZSM-5⋯phenol
d(CB–HB)	108.4 (108.4)	d(CP–OP)	135.3 (136.3)	d(CB–HB)	108.5 (108.4)	d(CP–OP)	139.0 (136.3)
d(Si–OZ1)	160.2 (159.7)	d(OP–HP)	96.7 (96.1)	d(Al–OZ1)	181.2 (182.4)	d(OP–HP)	98.1 (96.1)
d(Si–OZ2)	159.5 (159.9)	d(Si–OZ1)	161.5 (159.7)	d(Al–OZ2)	169.0 (168.9)	d(Al–OZ1)	180.2 (182.4)
∠(OZ1–M–OZ2)	104.1 (104.9)	d(Si–OZ2)	159.2 (159.9)	d(OZ1–HZ)	97.7 (96.9)	d(Al–OZ2)	170.9 (168.9)
d(CB–HB⋯OZ1)	254.3	∠(OZ1–M–OZ2)	103.8 (104.9)	∠(OZ1–M–OZ2)	91.7 (90.5)	d(OZ1–HZ)	102.5 (96.9)
∠(CB–HB⋯OZ1)	117.5	d(OP–HP⋯OZ1)	198.6	d(CB–HB⋯OZ2)	246.7	∠(OZ1–M–OZ2)	92.8 (90.5)
		∠(OP–HP⋯OZ1)	161.0	∠(CB–HB⋯OZ2)	116.5	d(OZ1–HZ⋯O)	153.3
				d(CB–HB⋯OZ2)	259.9	∠(OZ1–HZ⋯O)	161.8
				∠(CB–HB⋯OZ2)	111.8	d(OP–HP⋯OZ2)	180.3
				d(OZ1–HZ⋯π)	218.0	∠(OP–HP⋯OZ2)	140.9
				∠(OZ1–HZ⋯π)	155.8

---

It can be concluded that bond distances such as benzene C–H and that of the acidic hydroxyl group of the isolated Al-ZSM-5 cluster model are in excellent agreement with the experimental observation of 108.4 and 97.5 pm, respectively.^45,46

It should be pointed out that for the analysis of the interaction of benzene and phenol with the zeolite, initial geometry was started where benzene and phenol were assumed at the center, with benzene ring oriented approximately perpendicular to the plane of the zeolite. These orientations are approximately preserved for 8T-ZSM-5 adsorption complex after optimization whereas for 8T-Al-ZSM-5 adsorption complex, the aromatic ring of benzene and phenol were moved approximately parallel to the plane of the zeolite. It's noteworthy that optimizations of different initial orientations of complexes resulted in conformations as reported. These conformations correspond to the best orientation, based on creating most interactions resulting in most stable configuration, and are in good agreement with similar previous studies.^29,31

Fig. 1 and the results of Table 1 show a single molecule of phenol interacts with the site of 8T-ZSM-5 and 8T-Al-ZSM-5 zeolite frameworks via one and two type of hydrogen bonds, respectively. In the latter, the oxygen atom of phenol (O_P) interacts with the hydrogen of Brönsted acid (H_Z) with an O_P⋯H_Z bond length of 153.3 pm, and another hydrogen bond is formed between the hydrogen atom of phenol (H_P) and the oxygen atom of 8T-Al-ZSM-5 framework (O_Z2) with an O_Z⋯H_P bond length of 180.3 pm. Consequently, a six membered ring is formed (see Fig. 1). Based on Jeffrey's classification,⁴⁷ hydrogen bonds are classified into strong, moderate and weak types, with bond lengths 120–150, 150–220 and 220–320 pm, respectively. Consequently, based on the results presented in Table 1, O_P⋯H_Z and O_Z⋯H_P hydrogens are classified as moderate types. Furthermore, based on the results of Table 1 the hydrogen bond formed in the phenol⋯ZSM-5 adsorption complex is moderate, too. It is worth mentioning that benzene molecule interacts with the zeolite framework via one and three hydrogen bonds in the benzene⋯ZSM-5 and benzene⋯Al-ZSM-5, respectively (see Fig. 1). In the benzene⋯ZSM-5 adsorption complex, hydrogen atom of benzene interacts with oxygen atom of ZSM-5 with a H_B⋯O_Z bond length of 254.3 pm. In the latter, the hydrogen of Brönsted acid (H_Z) interacts with the π electrons of benzene with a π⋯H_Z–O_Z bond length of 218.0 pm and the other hydrogen bond is formed between two hydrogen atoms of benzene (H_B) and the oxygen atom of 8T-Al-ZSM-5 framework (O_Z2) with an O_Z⋯H_B bond length of 253.3 pm, on average, resulting in a five membered ring. From the geometrical results, it is concluded that all the hydrogen bonds, excluding π⋯H_Z–O_Z, in the benzene adsorption complex are categorized as classified weak and complex formation takes place, associated with weak hydrogen bonds. These results are in good agreement with the study of Koch and Popelier, and Panigrahi et al.^48,49 on the basis of weak hydrogen bonds.

Generally, Table 1 shows that geometry parameters including bonds and angles of hydrogen bonds of phenol⋯ZSM-5 and phenol⋯Al-ZSM-5 adsorption complexes are more proper than those of benzene⋯ZSM-5 and benzene⋯Al-ZSM-5 adsorption complexes. Comparison of the calculated bond lengths in Table 1 shows that the O_Z1–H_Z distance of the 8T-Al-ZSM-5 is elongated upon complexation. Increasing of O_Z1–H_Z bond length becomes higher when the guest molecule is changed from benzene to phenol (97.7 and 102.5 vs. 96.9 pm due to formation of hydrogen bonding with the benzene and phenol, respectively) confirming stronger hydrogen bonds in the phenol adsorption complex compared to benzene. It is worth mentioning that this case is well correlated with the long-range interactions. In addition the Al–O_Z1 bond lengths decrease upon complexation with benzene and phenol by about 1.2 and 2.2 pm, respectively, whereas Si–O_Z1 bond lengths increase. Variation of Si–O_Z1 and Al–O_Z1 bond lengths becomes greater in going from benzene⋯zeolite to phenol⋯zeolite. Table 1 shows that the complexation can also affect O_Z1–M–O_Z2 bond angles. O_Z1–Al–O_Z2 bond angles increase whereas O_Z1–Si–O_Z2 bond angles decrease. O_Z1–Al–O_Z2 bond angle, acting as adsorption sites in the 8T-Al-ZSM-5 adsorption complexes, is observed to open up due to benzene and phenol adsorption. O_P–H_P bond length increase upon complex formation. In addition, benzene adsorption causes a slight elongation of C_B–H_B bond compared to the isolated molecule.

Also, B3LYP and wB97XD functionals in combination with 6-311++G** basis sets were used to optimize adsorption complexes. The geometrical parameters of hydrogen bonds of the optimized adsorption complexes at B3LYP/6-311++G** and wB97XD/6-311++G** are listed in Table S1 of ESI.† As seen from Table S1,† the hydrogen bond lengths of adsorption complexes, especially the O_Z1–H_Z⋯π bond distance, are sensitive to the level of theory used. This indicates that dispersion may play an important role in host/guest interactions.

In the next sections, the importance of dispersive functionals has been evaluated while only the M06-2X has been selected to study the adsorption complexes as the most reliable functional to explain the host/guest interactions.

The predicted hydrogen bonds can be well elucidated with NMR, NQR, NBO and AIM parameters, which are better criteria for investigation of hydrogen bonding than the geometrical parameters. Significantly, the changes of geometrical parameters upon the adsorption of benzene, including a lengthening of the O_Z1–H_Z, O_P–H_P and C_B–H_B, inspired us to confirm this lengthening with the parameters mentioned. The following sections support these findings.

NMR and NQR analysis

Gauge Independent Atomic Orbital (GIAO) method employing DZVP2 basis set was implemented to confirm the predicted hydrogen bonding interactions based on NMR and NQR parameters of absorption complexes. We examined the ability of three density functionals including B3LYP, M06-2X, and wP97XD to evaluate the NMR parameters. Previous studies have demonstrated that DZVP2 basis set successfully predicts the spectroscopic parameters of probe molecules adsorbed on zeolites.^29,50,51 We are going to mainly discuss the NMR and NQR parameters of the oxygen and hydrogen atoms of adsorption complex, which contribute to the hydrogen bonding interactions as well as Al and Si atoms at the neighboring Brönsted acid site. The calculated ¹⁷O, ¹H, ²⁷Al and ²⁹Si chemical shielding tensors and isotropic chemical shift as well as the EFG principal components and the corresponding parameters of ¹⁷O, ²H, and ²⁷Al for bare and adsorption cluster models are summarized in Table 2, respectively. Also, Tables S2 and S3,† in the ESI, present the NMR and NQR parameters of mentioned atoms at the B3LYP/DZVP2 and wB97XD/DZVP2 levels, respectively.

Table 2 Selected calculated ¹⁷O, ¹H, and ²⁷Al NMR chemical shifts (in ppm), quadrupole coupling constant (in MHz for ¹⁷O and ²⁷Al nuclei, and kHz for ²H nuclei), and asymmetry parameter for ¹⁷O, ²H, and ²⁷Al nuclei before and after adsorption in the cluster models at the M06-2X/DZVP2 level of theory. The values in parentheses represent the calculated data before the adsorption

Cluster models	Nucleus	δiso	CQ	ηQ
8T-ZSM-5⋯benzene	HB	8.77(8.05)	215.82(220.17)	0.06(0.06)
	OZ1	68.34(56.21)	6.36(6.59)	0.25(0.21)
	OZ2	5.19(1.23)	6.57(6.68)	0.29(0.24)
	Si	−80.67(−81.98)
8T-Al-ZSM-5⋯benzene	HB	8.72(8.05)	212.86(220.17)	0.05(0.06)
	OZ1	66.09(52.05)	8.34(8.89)	0.88(0.80)
	HZ	4.31(3.96)	290.46(317.15)	0.11(0.09)
	OZ2	−1.17(−10.03)	4.05(4.10)	0.75(0.74)
	Al	72.43(73.48)	19.39(21.19)	0.25(0.21)
	Si	−80.47(−80.68)
8T-ZSM-5⋯phenol	OP	102.55(92.31)	11.01(11.43)	0.86(0.91)
	HP	6.42(3.74)	312.13(341.92)	0.13(0.13)
	OZ1	76.04(56.21)	5.92(6.59)	0.43(0.21)
	OZ2	9.69(1.23)	6.48(6.68)	0.25(0.24)
	Si	−80.06(−81.98)
8T-Al-ZSM-5⋯phenol	OP	92.60(92.31)	10.11(11.43)	0.92(0.91)
	HP	8.29(3.74)	266.60(341.92)	0.10(0.12)
	OZ1	55.32(52.05)	6.34(8.89)	0.93(0.80)
	HZ	12.52(3.96)	168.36(317.15)	0.18(0.09)
	OZ2	10.99(−10.03)	4.03(4.10)	0.76(0.74)
	Al	73.11(73.48)	15.82(21.19)	1.00(0.21)
	Si	−80.52(−80.68)

---

Fig. 1 shows that there are two distinct oxygen sites in the ZSM-5, and two distinct oxygen sites and one hydrogen site in the Al-ZSM-5. Based on the calculated results, ¹⁷O, ¹H, ²⁷Al and ²⁹Si chemical shift isotropy, δ_iso, are observed to change from bare to the one in the adsorption cluster models due to the participation in the intermolecular hydrogen bonding interactions. Alteration in charge density around a particular nucleus could be manifested in its NMR chemical shift. The magnitude of these variations depends on its contribution to the interactions. The ¹⁷O chemical shift value of 8T-ZSM-5 (O_Z1) of bare cluster model is 56.21 ppm, increasing with a Δδ_iso of 12.13 and 19.83 ppm in the benzene⋯ZSM-5 and phenol⋯ZSM-5 cluster models, respectively. The ¹⁷O chemical shift value of 8T-Al-ZSM-5 (O_Z1) of bare cluster model is 52.05 ppm and it increases about 14.04 and 3.27 ppm in the benzene⋯Al-ZSM-5 and phenol⋯Al-ZSM-5 cluster models, respectively. In spite of the stronger hydrogen bonding of phenol⋯Al-ZSM-5, its δ_iso (¹⁷O_Z1) alternation is smaller than benzene⋯Al-ZSM-5 as predicted through the geometry parameters. Thus, what is the origin of the observed violation in the trend of the δ_iso (¹⁷O_Z1) of phenol⋯Al-ZSM-5? It seems that hydroxyl group of phenol in the phenol⋯Al-ZSM-5 contributes in two hydrogen bondings through O_P and H_P (Fig. 1) resulting in formation of a six membered ring. Since the phenol molecule is a stronger base probe compared to benzene, its adsorption on the Brönsted acid site will result in a much stronger proton transfer from zeolite to phenol. In fact, one could conclude that O_Z1 acts like O_Z2 and vice versa. Formation of six membered ring has been confirmed by AIM analysis in the next section which is in good agreement with similar previous theoretical works.^36,52,53

Similarity, ¹⁷O_Z2 chemical shift values increase in the benzene⋯ZSM-5, phenol⋯ZSM-5, benzene⋯Al-ZSM-5, and phenol⋯Al-ZSM-5 adsorption complexes compared to bare cluster model confirming the contribution of ¹⁷O in the hydrogen bonding in four adsorption complexes (Table 1). More alternation of chemical shift value in the phenol adsorption complexes compared to benzene adsorption complexes is attributed to the stronger interaction in the phenol adsorption complexes.

Chemical shift values are in excellent agreement with the experimental observations which are 3.6–4.3, 6–8.5, and 4 ppm for ²H_Z, ²H_B, and ²H_P, respectively.^{54–56 2}H_Z chemical shift value of 8T-Al-ZSM-5 is 3.96 ppm increasing about 0.35 and a remarkable Δδ_iso 8.56 ppm relative to the one in the benzene and phenol adsorption complex, respectively. Chemical shift is further alternated due to the stronger contribution of H_z in the hydrogen bonds in the phenol adsorption complex compared with benzene adsorption complex. These results confirm high sensitivity of chemical shielding tensors to local bonding environments.

²⁹Si and ²⁷Al chemical shift values are in good agreement with the experimental results and previous studies; −112.8 and 60 ppm, respectively.^42,57,58 It is noteworthy that alternations of ²⁹Si and ²⁷Al chemical shift values are insignificant because of indirect contributing in hydrogen bondings.

The influence of the hydrogen bonding interactions on the ²H, ¹⁷O, and ²⁷Al C_Q and η_Q for the bare and adsorption cluster models is now focused on. NQR parameters of these atoms are also changed due to hydrogen bonding interactions. According to Table 2, the C_Q (²H_Z) value decreases from 317.15 kHz for Al-ZSM-5 to 290.46 and 168.360 kHz in benzene⋯Al-ZSM-5 and phenol⋯Al-ZSM-5, respectively, whereas the corresponding values of η_Q increase. Similar to H_z, the C_Q (O_Z1) and C_Q (O_Z2) values decrease in benzene⋯Al-ZSM-5 and phenol⋯Al-ZSM-5 complexes whereas the corresponding values of η_Q increase.

Two factors control the value of q_zz for a quadrupolar nucleus: the charge density at the nucleus and the symmetry of EFG around the nucleus. The hydrogen bonding interactions increase the charge density at acceptor atoms. In addition, EFG is more asymmetric in atoms due to hydrogen bonding. On the other hand, if the asymmetry of EFG increases, then q_zz would consequently decrease. As a result, the competing effects of charge density and EFG asymmetry on C_Q offset each other, leading to only a small decrease in the C_Q values of these compounds.

It should be pointed out that Al atom do not directly contribute to the hydrogen bonding, but alternation of C_Q parameter is considerable. It can be concluded ²⁷Al NQR parameters are in good agreement with the experimental observation of C_Q = 16 MHz and η_Q = 0.1.⁵⁸ According to Table 2, the C_Q (²⁷Al) value decreases about 1.80 MHz and remarkable ΔC_Q = 4.63 MHz in the benzene⋯Al-ZSM-5 and phenol⋯Al-ZSM-5, respectively. All the results are in good agreement with geometry and NMR parameters discussed in the previous sections. In the next section, we intend to confirm our results with NBO and AIM calculations.

QTAIM and NBO analysis

To gain insight into the nature of interaction between benzene and zeolite as well as phenol and zeolite, the Bader's quantum theory⁵⁹ of atoms in molecules (QTAIM) has been applied at M06-2X/6-311++G** level. The total electronic density, ρ(r_c), the Laplacian of electron density, ∇²ρ(r_c), and total energy density (H) at the Bond Critical Points (BCPs) and Ring Critical Point (RCP), are summarized in Table 3. The hydrogen bond energies are listed in Table 4.

Table 3 Selected calculated BCP and RCP data (au) before and after adsorption in the cluster models at the M06-2X/6-311++G** level of theory. The values in parentheses represent the data before adsorption

Bond	ρ(r)	∇^2ρ(r)	H(r)
8T-ZSM-5⋯Benzene
CB–HB	0.2851(0.2838)	−1.0486(−1.0175)	−0.3009(−0.2961)
Si–OZ1	0.1407(0.1422)	1.1604(1.1848)	−0.0269(−0.0273)
Si–OZ2	0.1436(0.1418)	1.2000(1.1800)	−0.0285(−0.0275)
OZ1⋯HB	0.0093	0.0345	0.0010
8T-Al-ZSM-5⋯Benzene
CB–HB	0.2859(0.2838)	−1.0443(−1.0175)	−0.2998(−0.2961)
	0.2859(0.2838)	−1.0416(−1.0175)	−0.2996(−0.2961)
OZ1–HZ	0.3304(0.3425)	−2.0714(−2.1590)	−0.5801(−0.6025)
Al–OZ1	0.0709(0.0684)	0.5850(0.5598)	0.0123(0.0121)
Al–OZ2	0.1019(0.1027)	0.9420(0.9494)	0.0108(0.0104)
HZ⋯π	0.0155	0.0428	0.0009
OZ2⋯HB	0.0116	0.0409	0.0009
	0.0098	0.0352	0.0011
8T-ZSM-5⋯Phenol
OP–HP	0.3552(0.3675)	−2.1600(−2.150)	−0.6094(−0.6132)
Si–OZ1	0.1366(0.1422)	1.1031(1.1848)	−0.0257(−0.0273)
Si–OZ2	0.1448(0.1418)	1.2120(1.1800)	−0.0291(−0.0275)
OZ1⋯HP	0.0223	0.0697	−0.0007
8T-Al-ZSM-5⋯phenol
OP–HP	0.3370(0.3675)	−2.0700(−2.150)	0.5832(−0.6132)
OZ1–HZ	0.2797(0.3425)	−1.4844(−2.1590)	0.4443(−0.6025)
Al–OZ1	0.074(0.0684)	0.6084(0.5598)	0.0116(0.0121)
Al–OZ2	0.0971(0.1027)	0.8776(0.9494)	0.0111(0.0104)
HZ⋯OP	0.0700	0.1419	−0.0157
OZ2⋯HP	0.0357	0.1156	−0.0008
RCP	0.0147	0.0758	0.0021

---

Table 4 NBO analysis of donor–acceptor interactions in adsorption complexes showing stabilization energy E(2) values (kJ mol⁻¹) at the M06-2X/DZVP2 level of theory, hydrogen bonding energies (E_HB) values (kJ mol⁻¹) obtained from AIM analysis and main contributions to the adsorption energy (kJ mol⁻¹) at the M06-2X/6-311++G** level of theory

Adsorption complexes	Donor⋯acceptor	EHB	E(2)	ΔEint	ΔE^def	ΔE′ads	ΔEads
8T-ZSM-5⋯benzene	CB–HB⋯OZ1	−8.64	0.84	−26.44	0.95	−25.49	−34.18
8T-Al-ZSM⋯5-benzene	OZ1–HZ⋯π	−11.70	0.59	−39.37	2.77	−36.60	−47.79
	CB–HB⋯OZ2	−8.72/−11	1.25
8T-ZSM-5⋯phenol	OP–HP ⋯OZ1	−24.7	2.84	−48.24	4.63	−43.61	−59.09
8T-Al-ZSM-5⋯phenol	OZ1–HZ⋯OP	−87.90	9.62	−91.25	20.38	−70.87	−105.02
	OP–HP⋯OZ2	−40.10	1.97

---

A shared interaction such as lone pairs and covalent bonds is shown by a negative Laplacian whereas positive Laplacian indicates where the electron density is declining as in ionic, hydrogen bonds, and van der Waals interactions.

For equilibrium molecular geometry, the molecular graphs (including the critical points and bond paths of interacting atoms) for the adsorbate complexes are presented in Fig. 2. Inspection of molecular graphs indicates that for O_Z1⋯H_B, O_Z2⋯H_B, H_Z⋯π, O_Z1⋯H_P, O_Z2⋯H_P and H_Z⋯O_P, there are corresponding bond paths and critical points (CPs) within the equilibrium structures, confirming the interaction between benzene, phenol and zeolite. Another useful characterization of hydrogen bonding is the Ring Critical Point (RCP) in the resonance assisted hydrogen bonds as an important criterion to describe the hydrogen bonding. Molecular graphs of phenol⋯Al-ZSM-5 indicate the existence of RCP assisted O_Z2⋯H_P and H_Z⋯O_P hydrogen bonds which confirms the formation of six membered ring. According to the third column in Table 3, for all adsorption complexes, Laplacian of total electronic densities at BCPs of O_Z1⋯H_B, O_Z2⋯H_B, H_Z⋯π, O_Z1⋯H_P, O_Z2⋯H_P and H_Z⋯O_P, are positive at the critical points, confirming predicted hydrogen bonds. The greater ρ_A⋯B values correspond to stronger hydrogen bond. Thus, phenol⋯Al-ZSM-5 forms the strongest interaction among adsorption complexes. Weak hydrogen bonds (E_HB < 50 kJ mol⁻¹) show both ∇²ρ(r_c) and H_BCP > 0, and medium hydrogen bonds (50 < E_HB < 100 kJ mol⁻¹) show ∇²ρ(r_c) > 0 and H_BCP < 0, as reported by Rozas et al.,⁶⁰ whereas strong hydrogen bonds (E_HB > 100 kJ mol⁻¹) show both ∇²ρ(r_c) and H_BCP < 0. Espinosa–Molins–Lecomte⁶¹ formula was applied to evaluate hydrogen bond energies for all of the adsorption complexes being investigated based on the electron density distribution at the BCPs: E_HB = 0.5V(r). The energy of O_Z2⋯H_P and H_Z⋯O_P hydrogen bonding interactions are 40.10 and 87.90 kJ mol⁻¹, respectively, and are classified as nearly medium and medium hydrogen bonds of partially covalent nature. The energy of C–H ⋯O and C–H⋯π hydrogen bonds are less than 50 kJ mol⁻¹ (8.72 and 11.00 kJ mol⁻¹ for C–H ⋯O and 11.70 kJ mol⁻¹ for O–H⋯π in the 8T-Al-ZSM-5⋯Benzene and 8.64 kJ mol⁻¹ for C–H ⋯O in the 8T-ZSM-5⋯Benzene) and are categorized as electrostatic weak hydrogen bonds. The energy of H_Z⋯O_P is 24.7 kJ mol⁻¹, which may be treated as weak hydrogen bonding.

Fig. 2 Molecular graph of adsorption complexes: (a) 8T-ZSM-5⋯benzene, (b) 8T-ZSM-5⋯phenol, (c) 8T-Al-ZSM-5⋯benzene and (d) 8T-Al-ZSM-5⋯phenol. Nuclei and bond critical points are represented by big and small spheres, respectively.

It was reported that bonds with positive value of ρ(r) and small negative value of H(r) at BCP are termed as partially covalent in nature. For O_z–H_Z, O_P–H_P and C_B–H_B BCP, ρ(r) and H(r) are negative, indicating a covalent character. The results in Table 3 show that the electron density and its Laplacian at O_Z1–H_Z BCP of Al-ZSM-5 decrease upon complexation with benzene and phenol. In addition, the values of ρ(r) and ∇²ρ(r) at O_P–H_P BCP of phenol and C_B–H_B BCP of benzene decrease upon complexation. Therefore, complexation causes a decrease in the covalent nature in all complexes. This decrease is greater in 8T-Al-ZSM-5 adsorption complex than in 8T-ZSM-5 adsorption complexes. The decrease of BCP data for phenol adsorbate complex is greater than that of benzene adsorbate complex due to stronger hydrogen bondings. According to these results, it can be concluded that the interaction between phenol and zeolite in both model complexes is stronger than that between benzene and zeolite. In addition, the decrease in BCP data for O_Z1–H_Z bond is greater than that in O_P–H_P. Based on these results, it can be concluded that the interaction between oxygen of phenol and hydrogen of zeolite is stronger than that between hydrogen of phenol and zeolite oxygen. This result well correlate with hydrogen bonding length reported in Table 1. All the results are in agreement with NMR, NQR and NBO, reported in the previous sections.

Other method for prediction of hydrogen bonding is charge analysis based on Weinhold's NBO calculation.⁴³ The deviation of the molecule from the Lewis structure is caused by delocalization of electron density from filled (bonding or lone pair) Lewis type NBOs to other neighboring electron deficient orbitals (non-Lewis type NBOs, such as anti-bonding or Rydberg) properly oriented in a NBO representation. This can be applied as a measure of hyper conjugation resulting in a stabilizing effect. Second-order perturbation theory can be used to describe the stabilization energies of these interactions, E(2), associated with delocalization i → j and is estimated as E(2) = −q_jF(i,j)²/(ε_i − ε_j), where q_j is the donor orbital occupancy, ε_i, ε_j are diagonal elements (orbital energies) and F(i,j) is the off-diagonal NBO Fock matrix elements.

The E(2) values are summarized in Table 4. Table 4 shows that E(2) in the case of the phenol⋯ZSM-5 and phenol⋯Al-ZSM-5 adsorption complexes is larger than that of benzene⋯ZSM-5 and benzene⋯Al-ZSM-5 adsorption complexes. Inspection of these energies indicates that E(2) values well correlate with the strength of hydrogen bonding. More values of E(2) were obtained in the phenol adsorption complexes compared to benzene complexes.

Finally, the adsorption energy, ΔE_ads, of benzene and phenol on the zeolites was considered as the energy difference between the absorbed complex system and the total energy of separated fragments as follows:

ΔE_ads = E_complex − (E_{benzene/phenol} + E_zeolite) (3)

where E_complex represents the single-point energy of the optimized adsorption complex while E_zeolite and E_{benzene/phenol} are the single-point energies of the optimized bare zeolite, separate benzene or phenol, respectively.

The total adsorption energy (ΔE′_ads) is assumed as follows:

ΔE′_ads = ΔE_int + ΔE_def (4)

In which ΔE_int represents the total energy consisting the hydrogen bonds and dispersive interactions which is currently calculated as follows:

ΔE_int = E_complex − (E^S_{benzene/phenol} + E^S_zeolite) (5)

where E^S_zeolite and E^S_{benzene/phenol} are the single-point energies of the zeolite and adsorbate, respectively, at the configurations of host/guest system having no further structural optimization. The ΔE_def term represents the deformations of both zeolite and adsorbate molecules which is often separately calculated as follows:

ΔE^def_zeolite = E^S_zeolite − E_zeolite (6)

ΔE^def_{benzene/phenol} = E^S_{benzene/phenol} − E_{benzene/phenol} (7)

Counterpoise calculations were carried out at M06-2X/6-311++G** level to evaluate how calculated adsorption energies are affected by the basis set superposition error (BSSE).⁵⁹ The various contributions of adsorption energy obtained at the M06-2X/6-311++G** level of theory are summarized in Table 4. According to obtained data, the deformation energy is more prominent for Al-ZSM-5⋯phenol adsorption complex compared to other adsorption complexes that can be due to the stronger hydrogen bonding in Al-ZSM-5⋯phenol.

Comparing ΔE′_ads and ΔE_ads values reported in Table 4, the difference between them is attributed to basis set superposition error (BSSE) which is neglected in ΔE′_ads.

It is worth mentioning that the energy calculation with different analysis obey the similar trend from hydrogen bonding strength standpoint.

Comparing the performance of applied functionals, their obtained data across the adsorption complexes are summarized in Fig. 3 and 4, including the absolute adsorption energy and absolute chemical shift of atoms, the atoms with available experimental chemical shift, respectively, obtained through B3LYP, wB97XD, M06-2X functionals. M06-2X and wB97XD noticeably improve the accuracy of dispersion energy estimation for systems containing hydrogen-bonding. From Fig. 3 and 4, the M06-2X functional has a higher capability to show and distinguish the dispersion phenomenon in noncovalent interactions compared to wB97XD and B3LYP. The obtained findings are in agreement with previous works describing noncovalent interactions in host/guest systems.^33,34,36,62

Fig. 3 Calculated and experimental absolute adsorption energies of applied functionals of benzene⋯ZSM-5, benzene⋯Al-ZSM-5, phenol⋯ZSM-5, and phenol⋯Al-ZSM-5 complexes.

Fig. 4 Calculated and experimental ²⁹Si, ²⁷Al, ¹H_Z, ¹H_B, ¹H_P absolute chemical shift of applied functionals of benzene, phenol, ZSM-5, and Al-ZSM-5.

Conclusion

DFT study was carried out to shed some light on the intermolecular hydrogen bonds in the adsorption complex of the 8T cluster model of ZSM-5 and Al-ZSM-5 catalysts with benzene and phenol comparatively through B3LYP, M06-2X, and wB97XD functionals employing 6-311++G** standard basis set. The computed geometry parameters display one and two types of intermolecular hydrogen bonding possibilities in the guest-ZSM-5 and guest-Al-ZSM-5 complexes, respectively. The C–H⋯O and O–H⋯π hydrogen bonds are recognized in the complex formation of benzene adsorbate zeolites. Compared to popular B3LYP functional, M06-2X offered a highly improved performance to concern the noncovalent interactions. All the predicted hydrogen bonds are confirmed via NMR, NQR, NBO, and QTAIM techniques. Variations of adsorption energy, isotropic chemical shifts, δ_iso, of ¹H, ¹⁷O, ²⁷Al, and ²⁹Si atoms contribute to the hydrogen bonding as well as quadrupole coupling constant, C_Q, and asymmetry parameter, η_Q. ²H, ¹⁷O, and ²⁷Al atoms are well correlated with the strength of hydrogen bonding. The formation of a hydrogen bond between a surface hydroxyl group and an adsorbed molecule leads to an enhanced isotropic ¹H and ¹⁷O NMR chemical shift whereas C_Q values of corresponding atoms decrease. Although Al and Si do not participate directly in the interactions, they are affected by these interactions resulting in alternations in NMR and NQR parameters. The total electronic density, ρ(r_c), the Laplacian of electron density, ∇²ρ(r_c) and energy density (H) estimated by AIM calculations, show that hydrogen bonds in the phenol⋯Al-ZSM-5 adsorption complex are partially covalent in nature. The differences in the adsorption behavior between benzene and phenol on the ZSM-5 and Al-ZSM-5 are attributed to the differences in types of interactions. Finally, Al-ZSM-5 appears to be an efficient adsorbent for phenol and benzene.

References

1. J. Kärger and D. M. Ruthven, Diffusion in zeolites and other microporous solids, Wiley, 1992.
2. S. Kulprathipanja, Zeolites in industrial separation and catalysis, John Wiley & Sons, 2010.
3. R. Xu, W. Pang, J. Yu, Q. Huo and J. Chen, Chemistry of zeolites and related porous materials: synthesis and structure, John Wiley & Sons, 2009.
4. B. Smit and T. L. M. Maesen, Chem. Rev., 2008, 108, 4125–4184.
5. J. O'Brien-Abraham, M. Kanezashi and Y. S. Lin, J. Membr. Sci., 2008, 320, 505–513.
6. M. Petryk, S. Leclerc, D. Canet and J. Fraissard, Catal. Today, 2008, 139, 234–240.
7. M. F. Fellah, I. Onal and R. A. van Santen, J. Phys. Chem. C, 2010, 114, 12580–12589.
8. E. J. M. Hensen, Q. Zhu, R. A. J. Janssen, P. C. M. M. Magusin, P. J. Kooyman and R. A. van Santen, J. Catal., 2005, 233, 123–135.
9. A. Reitzmann, E. Klemm and G. Emig, Chem. Eng. J., 2002, 90, 149–164.
10. G. Yang and L. Zhou, Catal. Sci. Technol., 2014, 4, 2490–2493.
11. N. Hansen, T. Kerber, J. Sauer, A. T. Bell and F. J. Keil, J. Am. Chem. Soc., 2010, 132, 11525–11538.
12. R. H. Crabtree, O. Eisenstein, G. Sini and E. Peris, J. Organomet. Chem., 1998, 567, 7–11.
13. F. Sánchez -Viesca, F. Cortés, R. Gómez and M. Berros, Am. J. Chem., 2012, 2, 343–346.
14. I. Alkorta, I. Rozas and J. Elguero, Chem. Soc. Rev., 1998, 27, 163–170.
15. M. L. Huggins, J. Org. Chem., 1936, 1, 407–456.
16. G. R. Desiraju and T. Steiner, The weak hydrogen bond: in structural chemistry and biology, Oxford university press, 2001.
17. T. P. Das and E. L. Hahn, Nuclear quadrupole resonance spectroscopy, Academic Press, New York, 1958.
18. N. Hadipour and S. Javadian, J. Mol. Struct., 2000, 525, 129–134.
19. S. Javadian and R. Araghi, J. Mol. Graphics Modell., 2009, 27, 620–627.
20. M. J. Duer, Solid state NMR spectroscopy: principles and applications, John Wiley & Sons, 2008.
21. F. Ektefa, M. Anafche and N. L. Hadipour, Comput. Theor. Chem., 2011, 977, 1–8.
22. N. V. Sidwich, The Electronic Theory of Valency, Clarendon, Oxford, 1972.
23. C. Raksakoon and J. Limtrakul, J. Mol. Struct.: THEOCHEM, 2003, 631, 147–156.
24. R. Rungsirisakun, B. Jansang, P. Pantu and J. Limtrakul, J. Mol. Struct., 2005, 733, 239–246.
25. P. M. Esteves and B. Louis, J. Phys. Chem. B, 2006, 110, 16793–16800.
26. Z. Yang, G. Yang, X. Liu and X. Han, Catal. Lett., 2013, 143, 260–266.
27. E. A. P. G. Li, R. A. van Santen, Z. Feng, C. Li and E. J. M. Hensen, J. Catal., 2011, 284, 194–206.
28. H. van Koningsveld, H. van Bekkum and J. C. Jansen, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1987, 43, 127–132.
29. A. Zheng, H. Zhang, L. Chen, Y. Yue, C. Ye and F. Deng, J. Phys. Chem. B, 2007, 111, 3085–3089.
30. D. Yi, H. Zhang and Z. Deng, J. Mol. Catal. A: Chem., 2010, 326, 88–93.
31. S. Yuan, J. Wang, Y. Li and H. Jiao, J. Nat. Gas Chem., 2003, 12, 93–97.
32. M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis and J. A. Montgomery, J. Comput. Chem., 1993, 14, 1347–1363.
33. S. Grimme, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2011, 1, 211–228.
34. Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241.
35. I. N. Levine and P. Learning, Quantum chemistry, Pearson Prentice Hall, Upper Saddle River, NJ, 2009.
36. J. Van der Mynsbrugge, K. Hemelsoet, M. Vandichel, M. Waroquier and V. Van Speybroeck, J. Phys. Chem. C, 2012, 116, 5499–5508.
37. E. A. C. Lucken, Nuclear Quadrupole Coupling Constants, Academic Press, London, 1969.
38. P. Pyykkö, Mol. Phys., 2001, 99, 1617–1629.
39. K. Wolinski, J. F. Hilton and P. Pulay, J. Am. Chem. Soc., 1990, 112, 8251–8260.
40. S. Sternhell, Org. Magn. Reson., 1983, 21, 770.
41. J. K. M. Cambridge, Magn. Reson. Chem., 1987, 25, 280.
42. S. Sklenak, J. Dědeček, C. Li, B. Wichterlová, V. Gábová, M. Sierka and J. Sauer, Angew. Chem., Int. Ed., 2007, 46, 7286–7289.
43. A. E. Reed, L. A. Curtiss and F. Weinhold, Chem. Rev., 1988, 88, 899–926.
44. F. Biegler-König, J. Schönbohm and D. Bayles, J. Comput. Chem., 2001, 22, 545–559.
45. H. Huo, L. Peng and C. P. Grey, J. Phys. Chem. C, 2011, 115, 2030–2037.
46. E. G. Cox, D. W. J. Cruickshank and J. A. S. Smith, Proc. R. Soc. A, 1958, 247, 1–21.
47. G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, Oxford, 1997.
48. U. Koch and P. L. A. Popelier, J. Phys. Chem., 1995, 99, 9747–9754.
49. S. K. Panigrahi and G. R. Desiraju, Proteins: Struct., Funct., Bioinf., 2007, 67, 128–141.
50. A. Zheng, L. Chen, J. Yang, Y. Yue, C. Ye, X. Lu and F. Deng, Chem. Commun., 2005, 2474–2476.
51. A. Zheng, L. Chen, J. Yang, M. Zhang, Y. Su, Y. Yue, C. Ye and F. Deng, J. Phys. Chem. B, 2005, 109, 24273–24279.
52. C. Tuma and J. Sauer, Chem. Phys. Lett., 2004, 387, 388–394.
53. M. M. Mestdagh, W. E. Stone and J. J. Fripiat, J. Phys. Chem., 1972, 76, 1220–1226.
54. D. Pavia, G. Lampman, G. Kriz and J. Vyvyan, Introduction to Spectroscopy, Cengage Learning, 2008.
55. Y. Jiang, J. Huang, W. Dai and M. Hunger, Solid State Nucl. Magn. Reson., 2011, 39, 116–141.
56. J. Dědeček, S. Sklenak, C. Li, B. Wichterlová, V. Gábová, J. Brus, M. Sierka and J. Sauer, Zeolites and Related Materials: Trends, Targets and Challenges, proceedings of the 4th International FEZA, Paris, 2008.
57. W. C. Conner, R. Vincent, P. Man and J. Fraissard, Catal. Lett., 1990, 4, 75–83.
58. D. Freude, H. Ernst and I. Wolf, Solid State Nucl. Magn. Reson., 1994, 3, 271–286.
59. R. F. Bader, Atoms in molecules, Wiley Online Library, 1990.
60. I. Rozas, I. Alkorta and J. Elguero, J. Am. Chem. Soc., 2000, 122, 11154–11161.
61. E. Espinosa, E. Molins and C. Lecomte, Chem. Phys. Lett., 1998, 285, 170–173.
62. B. Boekfa, P. Pantu, M. Probst and J. Limtrakul, J. Phys. Chem. C, 2010, 114, 15061–15067.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra20657j

---