Theoretical Study of the Aluminum Distribution Effects on the Double

Article pubs.acs.org/JPCA

Theoretical Study of the Aluminum Distribution Effects on the Double Proton Transfer Mechanisms upon Adsorption of 4,4′-Bipyridine on H-ZSM-5 Yamina Akacem,† Martine Castellà-Ventura,‡ and Emile Kassab*,‡ †

Laboratoire de Modélisation Moléculaire, Institut de Chimie, Université des Sciences et de Technologie, LTMM-USTHB, BP 32 Dar El Beida, Alger, Algérie ‡ Laboratoire de Chimie Théorique, CNRS-UMR7616, Université Pierre et Marie Curie, 4 Place Jussieu, 75252 Paris Cedex 05, France S Supporting Information *

ABSTRACT: The aluminum distribution effects on the adsorption of 4,4′-bipyridine (44BPY) in the straight channel of H-ZSM-5 simulated by two ten-membered ring clusters (2−10T) have been investigated by DFT methods. The energetic and structural properties of the complexes formed upon interaction of 44BPY with the zeolite Brønsted acid sites for six different aluminum distributions were determined by B3LYP/6-31+G* calculations. Dispersion energies were estimated by performing single point calculations at the MP2 and M06-2X levels. Interaction energies were corrected for basis set superposition error (BSSE). The minimum energy pathways of the double proton transfer from H-ZSM-5 to 44BPY were characterized. Two mechanisms are proposed: a concerted mechanism in which both protons are simultaneously transferred giving the bidentate ion pair complex (44BPYH22+/2−10T2−) and a consecutive mechanism in which one proton is transferred directly leading to the monodentate ion pair complex (44BPYH+/2−10T−), whereas the second proton can be transferred according to Al distribution. The formation of monodentate or bidentate complexes strongly depends on the Al distribution.

1. INTRODUCTION Zeolites in their protonated form are widely used as solid acid catalysts for many reactions in petrochemical and fine chemical industries.1−3 The catalytic activity of these materials is attributed to the presence of Brønsted acid sites associated with bridging hydroxyl groups Si−OH−Al. These sites are strong enough to adsorb host molecules by hydrogen bonding interactions. A proton transfer can often occur between the acid site and the adsorbed molecule and can thus initiate a series of reactions that can lead to the formation of new species. The characterization of such acid sites and how they interact with the adsorbed molecules will certainly contribute to further understand the adsorption properties and catalytic activity of zeolite and also allow elucidating the mechanism of the catalytic reaction occurring at these sites. To better understand the role of acidity in catalytic processes, it is important to know the elementary steps of the interaction of basic molecules with Brønsted acid sites and the role of hydrogen bonding precursors. Several experimental studies with spectroscopic methods such as NMR,4−9 Raman,10,11 and IR5,12−25 have been carried out to characterize the structure and interactions at molecular level, with sometimes inconclusive results. In addition, these methods do not generally allow us to provide information on intermediate species having extremely short lifetimes involved in catalytic reactions. Theoretical © 2012 American Chemical Society

calculations can then be done to obtain more precise information such as the structure of these intermediate species, and to suggest interpretation of the experimental results by providing a direct link between structure and spectra. In addition, theoretical methods offer a powerful tool that sheds light on the reaction mechanisms complementing experimental investigations. Theoretical studies6,13,26−41 conducted in recent years by different approaches on the adsorption of aromatic molecules on zeolites have been devoted to energetic properties. Many previous studies by ab initio and DFT calculations have been reported in the past on small clusters as models of zeolite.42−46 These clusters do not correspond to a particular zeolite, but rather to a subunit containing the tetrahedral generic Brønsted acid site in an environment without constraint. These small clusters can give a good description of the adsorption of small molecules. However, for larger molecules such as aromatic species, with dimensions close to the opening pores of zeolites such as H-ZSM-5, there are multiple electrostatic interactions between hydrogen atoms of adsorbed molecules and the oxygen atoms of the zeolite Received: September 30, 2011 Revised: January 5, 2012 Published: January 5, 2012 1261

dx.doi.org/10.1021/jp209445s | J. Phys. Chem. A 2012, 116, 1261−1271

The Journal of Physical Chemistry A

Article

framework. Thus, the cluster model must be large enough to account for these interactions. It has been well-known that the aluminum distribution in the zeolite frameworks controls the location of the strongly acidic protons countering the negative charge of the AlO4− framework in the protonic zeolites.47 However, in addition to the location of protons in the framework of silicon-rich zeolites such as H-ZSM-5, the distances between the protons might be also expected to affect their function and the reaction pathway in acid-catalyzed reactions. Several works have reported that the strong Brønsted acidity occurs only on acid sites of which the Al atom has no Al atoms at its Next-Nearest-Neighbor (NNN).48−51 In addition, it has been reported that the NextNext-Nearest-Neighbor (NNNN) distribution is more acidic than the NNN one when the two Al atoms are located in the same ring.52 Several efforts have been undertaken to predict the distribution of Al atoms in silicon-rich zeolites,53−58 and a few experimental studies have been devoted to its effect on the catalytic activity.8,9,24 It has been shown by 27Al MAS NMR studies that the Al distribution is not random, but substantially depends on the conditions of zeolite synthesis.55,57,58 However, the Al distribution still remains a challenge in heterogeneous catalysis.59 We recently reported DFT calculations31 at the B3LYP level on the adsorption of 4,4′-bipyridine (44BPY) inside the straight channel of H-ZSM-5 simulated by a cluster composed of two ten-membered rings (2−10T). We investigated the double proton transfer reaction between 44BPY ligand and the Brønsted acid sites of the zeolite with two particular Al distributions adapted to concerted mechanism, leading to the formation of the bidentate ion pair complex (44BPYH22+/ 2−10T2−). Indeed, the diprotonated 44BPYH22+ was observed by Raman spectrometry upon adsorption of 44BPY inside H-ZSM-5.11 In the present study, which is the extension of our previous work,31 other Al distributions have been considered. Extensive cluster B3LYP calculations have been carried out in order to gain an understanding at the molecular level of the influence of the Al distribution on the different mechanisms of double proton transfer from Brønsted acid sites of H-ZSM-5 to the nitrogen atoms of 44BPY. In particular, this work presents the energetic and structural analysis of the adsorbed species implicated upon interaction of 44BPY with 2−10T cluster according to the Al distribution. Although we are only interested in the relative trends of the proton transfer reaction as a function of the Al distribution, the dispersion energy has been taken into account by performing single-point calculations using the MP2 method and the new M06-2X density functional.60

Figure 1. Schematic representation of 44BPY in the straight channel of H-ZSM-5 simulated by a 2−10T cluster. The position and the orientation of 44BPY are defined by α and β angles. The Brønsted active regions are represented in balls and sticks.

Obviously, there are significantly more possible Al distributions in the framework of H-ZSM-5 than those investigated in this study. The adsorption of 44BPY in H-ZSM-5 was modelized by placing the molecule between two 10T rings and perpendicularly to their planes, so that, when the molecule bends during the geometry optimization, at least one nitrogen atom may be directed to the proton of the bridging hydroxyl Si−OH−Al group. For the sake of simplicity, co-planarity of both pyridyl rings of 44BPY will be assumed. In order to reduce the repulsion and spatial hindrance between pyridyl rings and the framework wall oxygen atoms in more realistic zeolite system, the molecular center of 44BPY was assumed to be fixed on the cluster straight channel axis during the geometry optimization. The quantum chemical calculations were carried out with the DFT method using the B3LYP functional and the 6-31+G* basis set, having five d functions. The geometries were optimized by using the standard techniques implemented in the Gaussian03 package.63 Since the B3LYP functional is not able to properly describe the van der Waals effects, the dispersion energy contributions were estimated by single point calculations at the MP2 and M06-2X levels in the Gaussian09 package.64 The adsorption energy Eads was computed as the difference between the energy of the optimized adsorption complex, and the sum of the energies of optimized 44BPY isolated molecule and 2−10T bare cluster. The minimum energy pathways for proton transfer from zeolite Brønsted acid sites to 44BPY were computed at the B3LYP level in a series of calculations using fixed values of O...H distance and fully optimizing all other geometrical parameters. Single point calculations at the MP2 and M06-2X

2. MODELS AND METHODS The straight channel of H-ZSM-559,61,62 is simulated by two ten-membered ring clusters (2−10T), arranged in parallel way, that we have used in our previous work31 as indicated in Figure 1. Each ring is composed of 9 silicon atoms and 1 aluminum atom. The dangling bonds of the two 10T rings were terminated with hydrogen atoms. The distribution of the two Al atoms in two parallel ten-membered ring clusters generates one hundred combinations, giving only six nonequivalent Al distributions.31 We note that our cluster is a particular view of H-ZSM-5 and corresponds to a restrictive situation, because the Al atoms are constrained to be in the two ten-membered rings, and the wall atoms of the channel are not taken into account. 1262



Article

Figure 2. Schematic representation (top view) of the different configurations according to the position of acidic protons associated with two Al atoms in the 2−10T cluster (example of the (Al1, Al3′) distribution). a The (HAl, HAl) and (AlH, AlH) configurations are identical.

and (AlH, AlH), as shown in Figure 2. However, our model is two parallel ten-membered ring structure, thus for the six Al distributions, (HAl, HAl) and (AlH, AlH) isomers are uniform (Figure 2). Moreover, for the two symmetrical distributions (Al1, Al1′) and (Al1, Al6′), the configurations (HAl, AlH) and (AlH, HAl) are also uniform. Consequently, there are only 16 different configurations. In order to reduce the computational work, our analysis has been limited to eight configurations: all (HAl, AlH) configurations in which the two protons are positioned outside of the two Al atoms for the six Al distributions, the (HAl, HAl) configurations for the (Al1, Al5′) and (Al1, Al6′) distributions. 3.1.2. Proton Transfer Reaction Pathways. Initially, for each Al distribution, 44BPY is inserted into the two tenmembered ring channel so that its CC inter-ring bond is along to the channel axis. The OH bonds of 2−10T cluster are oriented within the channel (Figure 1). Starting from this position by requiring that the plane of 44BPY remains perpendicular to the planes of 2−10T cluster during the geometry optimization processes, there is no proton transfer from Brønsted acid sites to the adsorbed molecule because the zeolite acidic protons are far from the nitrogen atoms of 44BPY (N...H > 3 Å), and on the other hand the nonbonding orbital of N is not directed toward the OH bond. When we remove the perpendicularity constraint by allowing the α tilt angle to relax, adsorbed 44BPY bends easily during the geometry optimization pointing one or two nitrogen atoms to the acidic protons Si−OH−Al; at the same time, these protons begin to move toward the N atoms of 44BPY. Whatever the Al distribution, it has been stressed that, because of the strong basicity of 44BPY, no neutral hydrogen bonded complex either as mono- or bidentate forms was found to be stable. In the present work, two double proton transfer mechanisms have been proposed depending on the Al distribution: a concerted mechanism in which both protons are simultaneously transferred, and a consecutive mechanism in which the transfer of the two protons takes place in a step-bystep sequence. 3.1.2.1. Concerted Mechanism. As discussed in our previous work,31 only two Al configurations are potential candidates for the study of the concerted double proton transfer reaction leading to the formation of the diprotonated 44BPYH22+: the (HAl, HAl) configuration for the symmetrical distribution (Al1, Al6′), and the (HAl, AlH) configuration for the (Al1, Al5′)

levels were also performed on the B3LYP minimum energy pathways. The B3LYP, MP2, and M06-2X calculations of the energetics of the proton transfer reaction were systematically corrected for the basis set superposition error (BSSE) estimated by the counterpoise method.65

3. RESULTS AND DISCUSSION 3.1. Energetics. 3.1.1. Localization of two Al Atoms and Associated Protons. For all calculations, the proton transfer reaction pathway is initiated with the location of the 44BPY molecule within the straight channel of H-ZSM-5 such as the long molecular axis of 44BPY is superimposed on the straight channel axis. The location and orientation of 44BPY can be described by two intermolecular angles: the α tilt angle between the long molecular axis of 44BPY and the straight channel axis, and the β rotational angle around the long molecular axis of 44BPY, defined as the angle between the Brønsted OH bond and the vector normal to the 44BPY plane (Figure 1). The positioning of the two Al atoms, each located in a 10T ring, gives six different distributions. Starting from the symmetrical distribution of the two Al atoms, in which the two parallel 10T rings can be perfectly superimposed (Cs symmetry) denoted as (Al1, Al1′) distribution (Figure 1), five other Al distributions can be generated by letting the first Al atom of the first 10T ring stay at its position (Al1), whereas the second Al atom is moved around the second 10T ring from left to right: four unsymmetrical distributions denoted as (Al1, Al2′), (Al1, Al3′), (Al1, Al4′), and (Al1, Al5′) and one symmetrical distribution (Ci symmetry) in which the two Al atoms are placed in opposite position, denoted as (Al1, Al6′) (Figure 1). The isomorphous substitution of Si atom by Al atom in the zeolite framework introduces a negative charge (AlO4−) that must be neutralized by a proton. The proton can be bounded to one of four bridging oxygen atoms adjacent to Al. In our 2− 10T cluster model, the proton positions are restricted to the only two ring oxygen atoms (Figure 1). Thus, for every Al position there are two OH positions designated by HAl and AlH, as shown in Figure 2. Taking into account the positioning of the protons associated with two Al atoms, 24 (Al, H) configurations can be generated by six Al distributions: (HAl, HAl), (HAl, AlH), (AlH, HAl), 1263



Article

unc c Table 1. BSSE Uncorrected (Eads ) and BSSE Corrected (Eads ) Adsorption Energies, and Estimated Dispersion Energiesa (Edisp) + − for Monodentate Ion Pair Complexes 44BPYH /2−10T of Different Aluminum Distributions, Computed at Different Levels of Theory with the 6-31+G* Basis Setb

B3LYP//B3LYP distributions (Al1, (Al1, (Al1, (Al1, (Al1, (Al1, a

Al1′) Al2′) Al3′) Al4′) Al5′) Al6′)

unc Eads

−25.2 −25.4 −25.9 −27.6 −26.4 −27.4

M06-2X//B3LYP

BSSE

c Eads

unc Eads

+4.2 +4.1 +4.2 +4.5 +4.7 +4.8

−21.0 −21.3 −21.7 −23.1 −21.7 −22.6

−30.0 −30.1 −31.0 −33.4 −32.0 −34.1

BSSE

c Eads

+3.7 +3.8 +3.9 +3.7 +4.1 +4.4

−26.3 −26.3 −27.1 −29.7 −27.9 −29.7

MP2//B3LYP Edisp

unc Eads

BSSE

c Eads

Edisp

−5.3 −5.0 −5.4 −6.6 −6.2 −7.1

−40.2 −40.3 −40.9 −44.2 −44.7 −46.5

+13.8 +13.7 +13.8 +14.9 +16.1 +16.6

−26.4 −26.6 −27.1 −29.3 −28.6 −29.9

−5.4 −5.3 −5.4 −6.2 −6.9 −7.3

c c Edisp = Eads (M06-2X or MP2) − Eads (B3LYP). bEnergies are given in kcal/mol.

capture dispersion interactions, but our results clearly show that the strong interaction energies are heavily contaminated by the BSSE errors. As expected,66 the BSSE is much larger for MP2 than for B3LYP. Whereas the B3LYP computed BSSE values are fairly constant and amount to ∼4.4 kcal/mol (∼17% of unc Eads ), the MP2 computed BSSE values vary from 13.7 for (Al1, Al2′) to 16.6 kcal/mol for (Al1, Al6′) representing ∼35% of the unc Eads . This points out that BSSE correction is necessary for calculating adsorption energy of the complexes using the MP2 method. The values of BSSE in M06-2X calculations were slightly lower than those in the B3LYP calculations. With the inclusion of the BSSE correction, the corrected c adsorption energy (Eads ) values obtained by MP2 and M06-2X methods are almost identical for all monodentate complexes, and the largest difference between two levels is found to be 0.7 kcal/mol for (Al1, Al5′) complex. The MP2 or M06-2X adsorption energy values that are larger than those obtained from the B3LYP calculations are mainly attributed to the van der Waals “non-local” interactions. In our study, the dispersion energy (Edisp) was approximated as the difference between the corrected adsorption energy calculated at the MP2 or M06-2X level and that calculated at the B3LYP level. Interestingly, we found that the Edisp values estimated by M06-2X and MP2 calculations are quite identical for all monodentate complexes. The Edisp values vary from 5.3 kcal/mol for (Al1, Al2′) to 7.3 kcal/mol for (Al1, Al6′) distribution. They introduce an increase of the adsorption energies (20−28%) of the monodentate complexes making their relative energies more pronounced, without altering their order of stability established by the B3LYP calculations. It should be noted that in the B3LYP calculations, the decrease of adsorption energies introduced by BSSE corrections (4.1−4.8 kcal/mol) is well compensated with the increase of the attractive dispersion energies (5.3−7.3 kcal/mol) not accounted for by B3LYP calculations. Thus, it is very important to recognize that at least with the 6-31+G* basis set, the B3LYP uncorrected Eads values and the MP2 corrected Eads values are almost identical for all complexes and the largest difference between two levels is found to be 2.5 kcal/mol for (Al1, Al6′) complex. Since the position and orientation of 44BPYH+relative to the first 10T ring are similar in all optimized structures of monodentate complexes (as discussed in Structures section), thus, these complexes differ only by the positions of the Al atom and its proton in the second 10T ring. So we might expect that the slightly different stability of these complexes is due to this second 10T ring. The contribution of the second neutral 10T ring to the adsorption energy for these complexes was partly estimated in the B3LYP calculations, because the

distribution obtained simply by exchanging the Si and Al atoms of the bridging active region Si−OH−Al in the (Al1, Al6′) distribution. For these two Al configurations, the double proton transfer reaction occurs in one step by a concerted mechanism in which the two N atoms of 44BPY can be synchronously directed toward the two bridging hydroxyl groups Si−OH−Al. The large distance between the two acidic sites situated in positions opposite to each other in these distributions strongly reduces the repulsion and spatial hindrance between pyridyl rings and the framework wall atoms. For both distributions, the calculated potential energy surface (PES) for the concerted double proton transfer shows only one minimum corresponding to the formation of the bidentate complex (44BPYH22+/2−10T2−), in agreement with the Raman results in which the diprotonated dication 44BPYH22+ is observed upon adsorption of 44BPY in H-ZSM-5.11 3.1.2.2. Consecutive Mechanism. The other six Al configurations are not suitable for studying the concerted double proton transfer, since the two N atoms of 44BPY cannot interact with Brønsted acid sites simultaneously: when one N atom can be in front of the bridge proton Si−OH−Al, the other N is in front of the bridge oxygen atom (SiOSi or SiOAl); thus, the second bridging Si−OH−Al proton is located at a distance farther away from the second N atom to be transferred. Thus, we expect that the double proton transfer reaction likely occurs via a consecutive mechanism rather than via the concerted one. For the consecutive mechanism, the first step reaction is started when the adsorbed molecule begins to lean itself in the channel toward a Brønsted acid site region. Initializing the geometry optimization with a position in which 44BPY is placed perpendicular to two rings of zeolite cluster, adsorbed 44BPY bends easily during the geometry optimization process and only one proton begins to be transferred without energy barrier from the zeolite acid site to the basic nitrogen atom of the pyridyl ring leading to the formation of the monodentate ion pair complex (44BPYH+/2−10T−). Thus, for these six Al configurations, the potential energy surfaces show only one minimum corresponding to these monodentate complexes. The adsorption energies with and without BSSE corrections c unc (Eads and Eads , respectively) for monodentate ion pair complexes (44BPYH+/2−10T−) for different Al distributions were calculated at different levels of calculation and listed in unc Table 1. For all monodentate complexes, the Eads values calculated at the MP2 level (−40.2 to −46.5 kcal/mol) are much larger in absolute value by 58−70% and 32−40% than those estimated by B3LYP (−25.2 to −27.6 kcal/mol) and M06-2X (−30.0 to −34.1 kcal/mol) calculations, respectively. It is widely appreciated that MP2 is more affordable choice among the traditional correlation methods that are able to 1264



Article

much larger than those for the first group as shown in Table 2. When comparing the B3LYP contribution values (∼2.7 kcal/ mol) with those obtained by MP2 calculations (∼6.8 kcal/ mol), we can find that only ∼60% of these contributions are due to dispersion interactions, while the remaining ∼40% results from electrostatic interactions between acidic active region Si−OH−Al and the basic nitrogen atom of adsorbed molecule. The possibility for transferring the second acidic proton depends of its distance from the basic nitrogen atom of 44BPYH+. The longer the N...H distances are, the more difficult the transfer of the second proton becomes. Consequently, we expect that the energy barriers for transferring the second proton from Brønsted acid site of neutral 10T ring to adsorbed monoprotonated 44BPYH+ leading to the formation of the bidentate ion pair complexes (44BPYH22+/2−10T2−) increase with increasing N...H distances in the following order: (Al1, Al6′) < (Al1, Al5′) < (Al1, Al4′) < (Al1, Al3′) < (Al1, Al2′) < (Al1, Al1′). Unlike the case of monodentate complexes, the relative energetic stabilities of the bidentate complexes with respect to the most stable one corresponding to the (Al1, Al6′) distribution are clearly very significant (Table 3), indicating the important effects of the Al distribution. The adsorption energies with and without BSSE corrections c unc 2− and Eads , respectively) for (44BPYH2+ (Eads 2 /2−10T ) of different Al distributions were calculated at different levels of calculations and listed in Table 3. Whatever the method of calculation, the adsorption energies in absolute value strongly increase in the following order: (Al1, Al1′) < (Al1, Al2′) < (Al1, Al3′) < (Al1, Al4′) < (Al1, Al5′) < (Al1, Al6′). Except for the (Al1, Al5′) complex which has nearly the same adsorption energy as the (Al1, Al6′) complex, the differences between the calculated adsorption energy values are significantly large, indicating the strong effect of the Al distribution. This is reasonable because framework oxygen atoms bonded to Al atoms are more basic than those bonded to Si atoms,47 and thus the basicity of the bridging oxygen atom depends on its position in the 10T ring and decreases with the increasing of the number of Si atoms between it and the Al tetrahedron. As can be shown from Table 3, the B3LYP uncorrected Eads values for these complexes vary from −12.3 for (Al1, Al1′) to −47.2 kcal/mol for (Al1, Al6′) distribution. Even though the dispersion energy has a strong effect on Eads of monodentate complexes, particularly for the first group where Edisp represents 32−51% compared to 16−22% estimated for the second group, it is found that the inclusion of BSSE corrections (from 5.8 to

B3LYP functional does not describe properly the van der Waals interactions. In order to estimate the contribution of the second neutral 10T ring, we have also calculated the adsorption energies for a complex formed with only one 10T ring (44BPYH+/10T−) cut from the optimized structures of two ten-membered ring complexes (44BPYH+/2−10T−) of different Al distributions. According to the Al distribution, the results can be sorted into two groups. The first group includes the three monodentate complexes formed with (Al1, Al1′), (Al1, Al2′) and (Al1, Al3′) distributions which have nearly the same c values estimated at the adsorption energies. The average Eads B3LYP, M06-2X, and MP2 levels for these complexes are ∼−21.3, −26.6, and −26.7 kcal/mol, respectively (Table 1). In these complexes, the second acidic proton is very far from the basic nitrogen atom of 44BPYH+ (OH...N ≥ 4.7 Å) and the nearest bridging oxygen atom SiOSi interacting with N is also far from Al atom. Consequently, for these complexes, the c contributions of the second neutral 10T ring to Eads are essentially due to dispersion interactions. Indeed, as shown in Table 2, the B3LYP contribution values (∼0.5 kcal/mol) are Table 2. Contributions of the Second Neutral 10T Ring to the Corrected Adsorption Energies for Monodentate Ion Pair Complexes 44BPYH+/2−10T− of Different Aluminum Distributions Computed at Different Levels of Theory with the 6-31+G* Basis Seta distributions (Al1, (Al1, (Al1, (Al1, (Al1, (Al1, a

Al1′) Al2′) Al3′) Al4′) Al5′) Al6′)

B3LYP//B3LYP

M06-2X//B3LYP

−0.5 (2.4) −0.3 (1.4) −0.6 (2.8) −2.3 (10.0) −3.1 (14.3) −2.8 (12.4)

−2.5 (9.5) −2.1 (8.0) −2.3 (8.5) −4.7 (15.8) −6.2 (22.2) −5.6 (18.9)

MP2//B3LYP −3.5 −3.1 −3.4 −5.9 −7.4 −7.1

(13.3) (11.7) (12.5) (20.1) (25.9) (23.7)

Values are given in kcal/mol, and in parentheses in %.

c , whereas those obtained by negligible, less than 2.8% of Eads M06-2X (∼2.3 kcal/mol) and MP2 (∼3.3 kcal/mol) c , respectively. calculations represent ∼8.7% and ∼12.5% of Eads The second group with the best stability includes the other three monodentate complexes formed with (Al1, Al4′), (Al1, c Al5′), and (Al1, Al6′) distributions with average Eads values of ∼−22.5 kcal/mol computed at the B3LYP level and of ∼ −29.2 kcal/mol estimated at the M06-2X and MP2 levels (Table 1). In these complexes, the second acidic proton is at a distance less than 3.2 Å from the nitrogen atom of the neutral pyridyl ring of 44BPYH+, thus the contributions of the second neutral 10T ring to the adsorption energies for these complexes are

unc c Table 3. BSSE Uncorrected (Eads ) and BSSE Corrected (Eads ) Adsorption Energies, and Estimated Dispersion Energiesa (Edisp) 2+ 2− for Bidentate Ion Pair Complexes 44BPYH2 /2−10T of Different Aluminum Distributions, Computed at Different Levels of Theory with the 6-31+G* Basis Setb

B3LYP//B3LYP distributions (Al1, (Al1, (Al1, (Al1, (Al1, (Al1, a

Al1′) Al2′) Al3′) Al4′) Al5′) Al6′)

unc Eads

−12.3 −16.6 −24.1 −34.0 −45.4 −47.2

M06-2X//B3LYP

BSSE

c Eads

unc Eads

+5.8 +5.9 +6.1 +6.5 +6.4 +6.4

−6.5 −10.7 −18.0 −27.5 −39.0 −40.8

−16.1 −22.6 −31.1 −41.4 −52.6 −55.5

BSSE

c Eads

+5.1 +5.2 +5.6 +6.1 +6.0 +5.9

−11.0 −17.4 −25.5 −35.3 −46.6 −49.6

MP2//B3LYP Edisp

unc Eads

BSSE

c Eads

Edisp

−4.5 −6.7 −7.6 −7.8 −7.6 −8.8

−32.2 −37.8 −46.3 −55.8 −68.4 −70.0

+18.9 +19.0 +19.7 +20.2 +20.4 +20.8

−13.3 −18.8 −26.6 −35.6 −48.0 −49.2

−6.8 −8.1 −8.7 −8.1 −9.0 −8.4

c c Edisp = Eads (M06-2X or MP2) − Eads (B3LYP). bEnergies are given in kcal/mol.

1265



Article

6.5 kcal/mol) and MP2 dispersion energies (from 6.8 to 9.0 kcal/mol) does not improve significantly the uncorrected B3LYP adsorption energies. Unlike the case of monodentate complexes, the E disp estimated by M06-2X and MP2 calculations are somewhat different for all bidentate complexes. Thus, the corrected Eads values obtained by MP2 (−13.3 to −49.2 kcal/mol) and M06-2X (−11.0 to −49.6 kcal/mol) calculations differ in the same way for all bidentate complexes, and the largest difference between two levels is found to be 2.3 kcal/mol for (Al1, Al1′) distribution. It should be noted that the hazardous similarity between the B3LYP uncorrected Eads and the MP2 corrected Eads values for all bidentate complexes is due to the cancellation of opposite effects of dispersion energies not accounted for by B3LYP calculations and BSSE errors. The largest difference between the two levels is found to be 2.6 kcal/mol for (Al1, Al5′) complex. The reaction energy (Ereac) for the bidentate complex formation with respect to monodentate complex (44BPYH+/ 2−10T− → 44BPYH 22+/2−10T 2− ) depends on the Al distribution. For the first group (Al1, Al1′), (Al1, Al2′), and (Al1, Al3′) distributions, the proton transfer reaction does not occur (Table 4), the B3LYP uncorrected Eads in absolute value

Figure 3. B3LYP/6-31+G* minimum energy pathways of the double proton transfer via the consecutive mechanism from 2−10T cluster to 44BPY for (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions. The reaction coordinate is the O1...H distance in the second 10T ring. The energies are defined with respect to the isolated 44BPY and 2−10T cluster.

Table 4. Reaction Energies (Ereac) of the Proton Transfer Reaction 44BPYH+/2−10T- → 44BPYH22+/2−10T2− for Different Aluminum Distributions, Computed at Different Levels of Theory with the 6-31+G* Basis Seta,b distributions (Al1, (Al1, (Al1, (Al1, (Al1, (Al1,

Al1′) Al2′) Al3′) Al4′) Al5′) Al6′)

B3LYP//B3LYP

M06-2X//B3LYP

MP2//B3LYP

+12.9 (+14.4) +8.8 (+10.6) +1.8 (+3.8) −6.4 (−4.4) −19.0 (−17.3) −19.8 (−18.2)

+13.9 (+15.3) +7.5 (+8.9) −0.1 (+1.6) −8.0 (−5.6) −20.6 (−18.7) −21.4 (−19.9)

+8.0 (+13.1) +2.5 (+7.8) −5.4 (+0.5) −11.6 (−6.3) −23.7 (−19.4) −23.5 (−19.3)

3.2 Å and the proton transfer reactions are exothermic (Table 5). For these distributions, the B3LYP minimum energy pathways Table 5. Activation Energies (Eact) of the Proton Transfer Reaction 44BPYH+/2−10T− → 44BPYH22+/2−10T2− for Different Aluminum Distributions, Computed at Different Levels of Theory with the 6-31+G* Basis Seta,b

a The BSSE corrected energies are given in parentheses. bValues are given in kcal/mol.

distributions

B3LYP//B3LYP

M06-2X//B3LYP

MP2//B3LYP

(Al1, Al4′) (Al1, Al5′) (Al1, Al6′)

+27.5 (+29.7) +16.4 (+17.5) +17.2 (+19.0)

+24.5 (+27.0) +14.5 (+15.9) +14.4 (+16.2)

+19.2 (+26.9) +12.9 (+17.2) +11.5 (+17.3)

a

The BSSE corrected energies are given in parentheses. bValues are given in kcal/mol.

for the bidentate complexes are less than those for the corresponding monodentate complexes by 12.9, 8.8, and 1.8 kcal/mol, respectively. This is due to the strong repulsive interactions between the two negatively charged Al tetrahedra of the two 10T rings in these bidentate complexes, wherein their positions do not allow the second pyridinium ring of dication 44BPYH22+ to intercalate between them to reduce their repulsion. It should be noted that the attractive interactions between this second pyridinium ring and the bridging oxygen atoms SiOSi (O5, O4, and O3) are not strong enough to stabilize the bidentate complexes. On the other hand, for the second group (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions, the proton transfer reaction is exothermic, the B3LYP uncorrected Eads in absolute value for the bidentate complexes are much larger than those for the corresponding monodentate complexes by 6.4, 19.0, and 19.8 kcal/mol, respectively (Table 4 and Figure3). It is essentially due to the strong attractive interactions of the second pyridinium ring of 44BPYH22+ with the bridging oxygen atoms of great basicity bonded to Al tetrahedron or its next Si atom (O10Al, and O2Si, respectively) in the second 10T ring. The activation energies (Eact) of the transfer of the second acidic proton Si−OH−Al to the nitrogen atom of 44BPYH+ were only calculated for the (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions in which the OH...N distances are less than

connecting the two minima corresponding to the mono- and bidentate complexes are shown in Figures 3 and 4. The potential energy surfaces determined by single-point calculations with MP2 and M06-2X methods are also reported in Figure 4. The uncorrected B3LYP energy barrier values calculated with respect to the monodentate complexes are 27.5, 16.4, and 17.2 kcal/mol for (Al1, Al4′), (Al1, Al5′), (Al1, Al6′) distributions, respectively (Table 5), and the corresponding reverse barriers are 33.9, 35.4, and 37.0 kcal/mol, respectively (Figure 3). The effect of BSSE on the potential energy surfaces is more significant in the MP2 calculations than in B3LYP or M06-2X calculations. As a result, the corrected PESs are upward shifted compared to the uncorrected profiles. More importantly, this shift is not the same for reactants, transition states and products, particularly in MP2 calculations. The BSSE introduces an increase of Eact by 1.1−2.5 kcal/mol in the DFT calculations, while this increase is much larger by 4.3−7.7 kcal/mol in the MP2 calculations. The dispersion energy has an opposite effect to that of BSSE on the energetics of adsorption and proton transfer reaction. If the dispersion interactions are crucial for calculating absolute adsorption energies of aromatic molecules in the zeolites, these interactions contribute less significantly to describe the PES of 1266



Article

kcal/mol for (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions, respectively. However, due to the opposite effects of the dispersion energy and the BSSE correction on the energetics of the proton transfer reaction, it is not surprising to find, as shown in Figure 4, that the PES determined by uncorrected B3LYP calculations are almost completely superimposed to those obtained by MP2 and M06-2X calculations. The largest differences between uncorrected B3LYP and corrected MP2 are 1.0 kcal/mol for the activation energy and 0.8 kcal/mol for reaction energy. 3.2. Structures. The main optimized geometrical parameters, determined at the B3LYP level, of 44BPY free and within the mono- and bidentate complexes formed with 2−10T clusters of all Al distributions are compared in Tables S1 and S2 of the Supporting Information, and illustrated in Figures 5 and 6. Selected structural parameters of 2−10T cluster bare and within the complexes are also given in Tables S1 and S2 of the Supporting Information. For bare 2−10T cluster, the structural parameters do not depend on Al distribution, due to the large Al...Al distances. 3.2.1. Monodentate Ion Pair Complexes. The geometrical parameters of the occluded monoprotonated 44BPYH+ do not practically depend on the Al distribution. The structural parameters of the neutral pyridyl ring are quite similar to the pyridyl moieties of neutral 44BPY. However, those of pyridinium ring present the same geometric modifications as pyridinium rings of free 44BPYH22+,31,67 resulting from quaternization of the ring upon protonation of the nitrogen atom. The C4C4′ bond is slightly smaller than the value calculated in neutral 44BPY. The orientation of 44BPYH+, defined by the two angles α and β, depends slightly on the Al distribution (∼ 43° and ∼137°, respectively) (Table S1 of the Supporting Information). The NH bond of the pyridinium ring (1.066−1.078 Å) has a similar value as in the ion pair complexes formed between pyridine and zeolite clusters.30 On the pyridinium ring side, the N...O1 distance is about 2.63 Å, whatever the Al distribution, it is indicative of a rather strong hydrogen bond, as yet evidenced in ion pair complexes with pyridine.30 However, the NHO1 angle is not far from linearity (∼165°) for the (Al1, Al5′) and (Al1, Al6′) distributions, and nearly linear (∼176°) for the other Al distributions. On the neutral pyridyl ring side, only for the two (Al1, Al5′) and (Al1, Al6′) distributions, the O1H...N distances are less than 2.5 Å, and the NHO1 angles are about 160°. However, for the other four Al distributions, O1H...N distances are very large, they increase in the following order: (Al1, Al4′) < (Al1, Al3′) < (Al1, Al2′) < (Al1, Al1′) from 3.2 up to 6.4 Å, and the NHO1 angles range from 112° to 135° (Figure 5, Table S1 of the Supporting Information). Likewise, the distances On...N between the nitrogen atom of the neutral pyridyl ring and the On oxygen atom of the neutral 10T ring located in front of it varies from ∼3.2 Å for the (Al1, Al6′) distribution to ∼3.6 Å for the (Al1, Al2′) distribution. Whatever the Al distribution, the distances between the ortho and especially meta H atoms of 44BPYH+ with the cluster oxygen atoms are larger than 2.5, and 3.0 Å, respectively, indicating insignificant attractive interactions (Figure 5). For the zeolite, the most important geometrical modifications occur around the Brønsted acid site of the anionic 10T cluster for all Al distributions, and of the neutral 10T cluster for the (Al1, Al5′) and (Al1, Al6′) distributions (Table S1 of the Supporting Information).

Figure 4. Potential energy surfaces of the double proton transfer via the consecutive mechanism from 2−10T cluster to 44BPY for (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions computed at different levels of theory with the 6-31+G* basis set. The reaction coordinate is the O1...H distance in the second 10T ring. The energies are defined with respect to the isolated 44BPY and 2−10T cluster.

the considered proton transfer reactions, as shown in Figure 4. Indeed, for calculating energy differences between nearby configurations of adsorbed species, for example reaction energy or energy barrier, the dispersion effect essentially cancels, leaving an energy difference controlled by local electronic interactions. Our results show that MP2 dispersion energies reduce the B3LYP corrected energy barrier values by only 2.8, 0.3, and 1.7 1267



Article

Figure 5. Monodentate ion pair complexes (44BPYH+/2−10T−) of different aluminum distributions.

3.2.2. Bidentate Ion Pair Complexes. Whatever the Al distribution, the geometrical parameters of the first pyridinium ring are similar to those of the pyridinium ring in the monodentate complexes. On moving from the (Al1, Al6′) to the (Al1, Al1′) distribution, the geometrical structure of the second pyridinium ring is getting closer and closer to the structure of a free pyridinium ring.30 Contrarily to the free 44BPYH22+,31 the inter-ring bond C4C4′ length is slightly shorter than in neutral 44BPY. For all Al distributions, the α tilt angle is of the same order of magnitude as in the monodentate complexes (∼37°). The β rotation angle ranges from 50° to 67° (Table S2 of the Supporting Information). The NH length of the first pyridinium ring is close to that in the monodentate complexes (1.066−1.078 Å). In the second pyridinium ring, for (Al1, Al5′) and (Al1, Al6′) distributions, the NH bond length is quite the same as in the first pyridinium ring. However, for the other Al distributions, the NH bond has nearly the same value as in the free pyridinium ring (∼1.025− 1.036 Å).30 For both (Al1, Al5′) and (Al1, Al6′) distributions, the NH...O1 (SiO1Al) and NH...O10 (SiO10Al) hydrogen bonds in

the first and second pyridinium rings, are rather strong, as indicated by the nearly linear NHO angles (163°−169°), and the smallest O...H (∼1.57 Å) and N...O distances (∼2.62 Å) (Figure 6). For the four other distributions, the NH...O hydrogen bonds in both pyridinium rings are less strong: the NHO angles decrease (up to 153°), the O...H and the N...O distances lengthen (up to 1.87 and 2.83 Å, respectively). For all Al distributions, the hydrogen bonds between the zeolite oxygen atoms and the hydrogen atoms of 44BPYH22+ are weak in the ortho position with CH...O distances ranging from ∼2.1 to 2.7 Å, and negligible in meta position (Figure 6, Table S2 of the Supporting Information). For the 2−10T cluster, the geometric parameters modified upon complexation are principally located in the Brønsted active regions (Table S2 of the Supporting Information).

4. CONCLUSIONS The proton transfer reactions upon adsorption of 4,4′bipyridine on H-ZSM-5 zeolite were investigated by DFT calculations at the B3LYP/6-31+G* level using a two tenmembered ring cluster. The energetic and structural properties of the species formed upon interaction of 44BPY with the 1268



Article

Figure 6. Bidentate ion-pair complexes (44BPYH22+/2−10T2−) of different aluminum distributions.

44BPY, leading to the formation of the monodentate complex. In the second step, the possibility for transferring the second acidic proton leading to the formation of the bidentate complex depends on the Al distribution. Our BSSE uncorrected B3LYP/ 6-31+G* results show that for the (Al1, Al1′), (Al1, Al2′) and (Al1, Al3′) distributions, the bidentate complexes are less stable than the corresponding monodentate complexes by 12.9, 8.8, and 1.8 kcal/mol, respectively, whereas for (Al1, Al4′), (Al1, Al5′) and (Al1, Al6′) distributions, the bidentate complexes are much more stable than the corresponding monodentate complexes by 6.4, 19.0, and 19.8 kcal/mol, respectively. The energy barrier values of the minimum energy pathways connecting two minima corresponding to mono- and bidentate complexes are 27.5, 16.4, and 17.2 kcal/mol for (Al1, Al4′), (Al1, Al5′), and (Al1, Al6′) distributions, respectively. The formation of the monodentate or bidentate complexes strongly depends on the Al distribution. From a methodological point of view, the comparison of DFT with MP2 calculations indicates that for energetics of the proton transfer reaction, the M06-2X approach gives similar results to those obtained by MP2 calculations. However, it should be noted that the similarity of the energetic results

Brønsted acid sites of the zeolite with different Al distributions were characterized. The dispersion interactions were estimated using MP2 and M06-2X methods. Two mechanisms have been proposed according to the Al distribution: a concerted mechanism in which both protons are simultaneously transferred, and a consecutive mechanism in which the transfer of the two protons takes place successively. For the concerted mechanism, only two particular (Al1, Al5′) and (Al1, Al6′) distributions in which the two acidic protons are farthest from each other can favor the simultaneous transferring of these two protons to the bidentate 44BPY ligand. The calculated potential energy surface of this concerted double proton transfer reaction has only one minimum corresponding to the formation of the bidentate ion pair complex (44BPYH22+/2−10T2−). For the other Al distributions, the double proton transfer reaction generally occurs via a consecutive mechanism. The calculated potential energy surface is characterized by two minima corresponding to the formation of a monodentate (44BPYH+/2−10T−) and a bidentate (44BPYH22+/2−10T2−) ion pair complexes. In the first step, only one proton is directly transferred without any barrier from the Brønsted acid site to 1269



Article

(15) Buzzoni, R.; Bordiga, S.; Ricchiardi, G.; Lamberti, C.; Zecchina, A.; Bellussi, G. Langmuir 1996, 12, 930−940. (16) Trombetta, M.; Alejandre, A. G.; Solis, J. R.; Busca, G. Appl. Catal., A 2000, 198, 81−93. (17) Armaroli, T.; Bevilacqua, M.; Trombetta, M.; Alejandre, A. G.; Ramirez, J.; Busca, G. Appl. Catal., A 2001, 220, 181−190. (18) Daniell, W.; Topsoe, N.-Y.; Knözinger, H. Langmuir 2001, 17, 6233−6239. (19) Busch, O. M.; Brijoux, W.; Thomson, S.; Schüth, F. J. Catal. 2004, 222, 174−179. (20) Karge, H. G. C. R. Chimie 2005, 8, 303−319. (21) Zheng, S.; Jentys, A.; Lercher, J. A. J. Catal. 2006, 241, 304−311. (22) Bevilacqua, M.; Montanari, T.; Finocchio, E.; Busca, G. Catal. Today 2006, 116, 132−142. (23) Bregolato, M.; Bolis, V.; Busco, C.; Ugliengo, P.; Bordiga, S.; Cavani, F.; Ballarini, N.; Maselli, L.; Passeri, S.; Rossetti, I.; Forni, L. J. Catal. 2007, 245, 285−300. (24) Márquez-Alvarez, C.; Pinar, A. B.; García, R.; Grande-Casas, M.; Pérez-Pariente, J. Top Catal. 2009, 52, 1281−1291. (25) Jin, F.; Li, Y. Catal. Today 2009, 145, 101−107. (26) Vos, A. M.; Rozanska, X.; Schonnheydt, R. A.; van Santen, R. A.; Hutschka, F.; Hafner, J. J. Am. Chem. Soc. 2001, 123, 2799−2809. (27) Arstad, B.; Nicholas, J. B.; Haw, J. F. J. Am. Chem. Soc. 2004, 126, 2991−3001. (28) Yuan, S.; Shi, W.; Li, B.; Wang, J.; Jiao, H.; Li, Y.-W. J. Phys. Chem. A 2005, 109, 2594−2601. (29) Llopis, F. J.; Sastre, G.; Corma, A. J. Catal. 2006, 242, 195−206. (30) Castellà-Ventura, M.; Akacem, Y.; Kassab, E. J. Phys. Chem. C 2008, 112, 19045−19054. (31) Kassab, E.; Castellà-Ventura, M.; Akacem, Y. J. Phys. Chem. C 2009, 113, 20388−20395. (32) Rozanska, X.; Saintigny, X.; van Santen, R. A.; Hutschka, F. J. Catal. 2001, 202, 141−155. (33) Rozanska, X.; van Santen, R. A.; Hutschka, F.; Hafner, J. J. Am. Chem. Soc. 2001, 123, 7655−7667. (34) Rozanska, X.; van Santen, R. A.; Hutschka, F. J. Phys. Chem. B 2002, 106, 4652−4657. (35) Rozanska, X.; Barbosa, L. A. M. M.; van Santen, R. A. J. Phys. Chem. B 2005, 109, 2203−2211. (36) Zhou, D.; Bao, Y.; Yang, M.; He, N.; Yang, G. J. Mol. Catal. A 2006, 244, 11−19. (37) Zheng, A.; Zhang, H.; Chen, L.; Yue, Y.; Ye, C.; Deng, F. J. Phys. Chem. B 2007, 111, 3085−3089. (38) Lomratsiri, J.; Probst, M.; Limtrakul, J. J. Mol. Graphics Modelling 2006, 25, 219−225. (39) Boekfa, B.; Choomwattana, S.; Khongpracha, P.; Limtrakul, J. Langmuir 2009, 25, 12990−12999. (40) Buurmans, I. L. C.; Pidko, E. A.; de Groot, J. M.; Stavitski, E.; van Santen, R. A.; Weckhuysen, B. M. Phys. Chem. Chem. Phys. 2010, 12, 7032−7040. (41) Hansen, N.; Kerber, T.; Sauer, J.; Bell, A. T.; Keil, F. J. J. Am. Chem. Soc. 2010, 132, 11525−11538. (42) Brand, H. V.; Curtiss, L. A.; Iton, L. E. J. Phys. Chem. 1993, 97, 12773−12782. (43) Sauer, J.; Ugliengo, P.; Garrone, E.; Saunders, V. R. Chem. Rev. 1994, 94, 2095−2160. (44) van Santen, R. A.; Kramer, G. J. Chem. Rev. 1995, 95, 637−660. (45) Corma, A. Chem. Rev. 1995, 95, 559−614. (46) Kassab, E.; Jessri, H.; Allavena, M.; White, D. J. Phys. Chem. A 1999, 103, 2766−2774. (47) Sung, C.-Y.; Broadbelt, L. J.; Snurr, R. Q. J. Phys. Chem. 2009, 113, 15643−15651. (48) Barthomeuf, D. J. Phys. Chem. 1979, 83, 249−256. (49) Janssens, G. O. A.; Toufar, H.; Baekelandt, B. G.; Mortier, W. J.; Schoonheydt, R. A. J. Phys. Chem. 1996, 100, 14443−14450. (50) Hunger, B.; Heuchel, M.; Clark, L. A.; Snurr, R. Q. J. Phys. Chem. B 2002, 106, 3882−3889. (51) Xu, B.; Bordiga, S.; Prins, R.; van Bokhoven, J. A. Appl. Catal., A 2007, 333, 245−253.

obtained by BSSE uncorrected B3LYP and BSSE corrected MP2 calculations is fortuitous, it is obviously due to the cancellation of opposite effects of BSSE error and dispersion correction not accounted for by B3LYP calculations. The largest differences between two levels are 0.8 kcal/mol for the activation energy and 0.5 kcal/mol for the reaction energies. On the basis of a consideration of both the computational expense and accuracy of the MP2 method, one must conclude that the DFT still appears as a practical and useful method with a small BSSE error to explore the PESs of proton transfer reactions in zeolite, and the M06 functional family seems to be particularly attractive in this regard.60

■

ASSOCIATED CONTENT

S Supporting Information *

The main optimized geometrical parameters, determined at the B3LYP/6-31+G* level, of 44BPY free and within the monoand bidentate complexes formed with 2−10T clusters of all Al distributions are listed in Tables S1 and S2. Selected structural parameters of 2−10T cluster bare and within the complexes are also given in Tables S1 and S2. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +33(0)1 44 27 75 08; fax: +33(0)1 44 27 41 17; e-mail: [email protected].

■

ACKNOWLEDGMENTS We regret to announce the death of our colleague, Prof. Yamina Akacem, on October 18th, 2011. We dedicate this paper to her memory.

■

REFERENCES

(1) Chen, N. Y.; Garwood, W. E.; Dwyer, F. G. Shape Selective Catalysis and Industrial Applications, second ed.; Dekker: New York, U.S., 1996. (2) Guisnet, M. Zeolites of Cleaner Technologies; Gilson, J. P., Ed.; Imperial College Press: London, U.K., 2002. (3) Bhattacharyya, K. G.; Talukdar, A. K. Catalysis in Petroleum and Petrochemica Industries; Narosa Publishing House Pvt Ltd.: Dehli, In, 2005. (4) Lercher, J. A.; Gründling, C.; Eder-Mirth, G. Catal. Today 1996, 27, 353−376. (5) Boréave, A.; Auroux, A.; Guimon, C. Microporous Mater. 1997, 11, 275−291. (6) Ehresmann, J. O.; Wang, W.; Herreros, B.; Luigi, D.-P.; Venkatraman, T. N.; Song, W.; Nicholas, J. B.; Haw, J. F. J. Am. Chem. Soc. 2002, 124, 10868−10874. (7) Hunger, M. Catal. Today 2004, 97, 3−12. (8) Sazama, P.; Dĕdeček, J.; Gábová, V.; Wichterlová, B.; Spoto, G.; Bordiga, S. J. Catal. 2008, 254, 180−189. (9) Román-Leshkov, Y.; Moliner, M.; Davis, M. E. J. Phys. Chem. C 2011, 115, 1096−1102. (10) Hendra, P. J.; Passingham, C.; Warnes, G. M.; Burch, R.; Rawlence, D. J. Chem. Phys. Lett. 1989, 164, 178−184. (11) Moissette, A.; Batonneau, Y.; Brémard, C. J. Am. Chem. Soc. 2001, 123, 12325−12334. (12) Parker, L. M.; Bibby, D. M.; Burns, G. R. J. Chem. Soc., Faraday Trans. 1991, 87, 3319−3323. (13) Kubelkova, L.; Kotrla, J.; Florian, J. J. Phys. Chem. 1995, 99, 10285−10293. (14) Barzetti, T.; Selli, E.; Moscotti, D.; Forni, L. J. Chem. Soc., Fraraday Trans. 1996, 92, 1401−1407. 1270



Article

(52) Zhou, D.; He, N.; Wang, Y.; Yang, G.; Liu, X.; Bao, X. J. Mol. Struct. (Theochem) 2005, 756, 39−46. (53) Rice, M. J.; Chakraborty, A. K.; Bell, A. T. J. Catal. 1999, 186, 222−227. (54) Rice, M. J.; Chakraborty, A. K.; Bell, A. T. J. Catal. 2000, 194, 278−285. (55) Sklenak, S.; Dĕdeček, J.; Li, C.; Wichterlová, B.; Gábová, V.; Sierka, M.; Sauer, J. Angew. Chem., Int. Ed. 2007, 46, 7286−7289. (56) van Bokhoven, J. A.; Lee, T.-L.; Drakopoulos, M.; Lamberti, C.; Thieß, S.; Zegenhagen, J. Nat. Mater. 2008, 7, 551−555. (57) Sklenak, S.; Dĕdeček, J.; Li, C.; Wichterlová, B.; Gábová, V.; Sierka, M.; Sauer, J. Phys. Chem. Chem. Phys. 2009, 11, 1237−1247. (58) Dĕdeček, J.; Lucero, M. J.; Li, C.; Gao, F.; Klein, P.; Urbanova, M.; Tvaruzkova, Z.; Sazama, P.; Sklenak, S. J. Phys. Chem. C 2011, 115, 11056−11064. (59) Nishi, K.; Kamiya, N.; Yokomori, Y. Microporous Mesoporous Mater. 2007, 101, 83−89. (60) Zhao, Y.; Truhlar, D. J. Phys. Chem. C 2008, 112, 6860−6868. (61) van Koningsveld, H. Acta Crystallogr. 1990, B46, 731−735. (62) Olson, D. H.; Kokotailo, G. T.; Lawton, S. L.; Meier, W. M. J. Phys. Chem. 1981, 85, 2238−2243. (63) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision D.02; Gaussian, Inc.: Wallingford CT, 2004. (64) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.02; Gaussian, Inc.: Wallingford CT, 2009. (65) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553−566. (66) Tuma, C.; Boese, A. D.; Handy, N. C. Chem. Phys. Phys. Chem. 1999, 1, 3939−3947. (67) Castellà-Ventura, M.; Kassab, E. J. Raman Spectrosc. 1998, 29, 511−536. (68) Almenningen, A.; Bastiansen, O. K. Nor. Vidensk. Selsk. Skr. 1958, 4, 1−16.

1271


Theoretical Study of the Aluminum Distribution Effects on the Double

Recommend Documents