Structures and Reaction Mechanisms of Cumene Formation via

Thana Maihom , Supawadee Namuangruk , Tanin Nanok and Jumras Limtrakul ... of pore system construction in zeolite-catalyzed chemical industry processe...
9 downloads 0 Views 286KB Size
Download PDF
12626

J. Phys. Chem. B 2006, 110, 12626-12631

Structures and Reaction Mechanisms of Cumene Formation via Benzene Alkylation with Propylene in a Newly Synthesized ITQ-24 Zeolite: An Embedded ONIOM Study Bavornpon Jansang,†,‡ Tanin Nanok,†,‡ and Jumras Limtrakul*,†,‡ Laboratory for Computational and Applied Chemistry, Department of Chemistry, Faculty of Science, Kasetsart UniVersity, Bangkok 10900, Thailand, and Center of Nanotechnology, Kasetsart UniVersity Research and DeVelopment Institute, Kasetsart UniVersity, Bangkok 10900, Thailand ReceiVed: March 16, 2006; In Final Form: May 7, 2006

The cumene formation via benzene alkylation with propylene on the new three-dimensional nanoporous catalyst, ITQ-24 zeolite, has been investigated by using the ONIOM2(B3LYP/6-31G(d,p):UFF) method. Both consecutive and associative reaction pathways are examined. The contributions of the short-range van der Waals interactions, which are explicitly included in the ONIOM2 model, and an additional long-range electrostatic potential from the extended zeolite framework to the energy profile are taken into consideration. It is found that benzene alkylation with propylene in the ITQ-24 zeolite prefers to occur through the consecutive reaction mechanism. The benzene alkylation step is the reaction rate-determining step with an estimated activation energy of 35.70 kcal/mol, comparable with an experimental report in β-zeolite of 34.9 kcal/mol. The electrostatic potential from the extended zeolite framework shows a much more significant contribution to the transition state selectivity than the van der Waals interactions.

1. Introduction Zeolites are fascinating nanostructured materials with a wide range of industrial applications such as the separation of organic molecules, catalysis, and ion-exchange.1-3 A fundamental understanding of the catalytic reaction of molecules in zeolites provides insight into developing new nanoporous catalysts, thus making the study of the reaction mechanism in such materials interesting. Since the product distribution is correlated with the size and shape of the molecules relative to the zeolite micropores, the concept of size and shape selectivity has been very constructive in designing new, improved zeolite catalysts, especially in reactions concerning bulky aromatic reactant molecules, where the size and shape selection becomes considerably important.4 Alkylation of aromatics is a good example of reactions where the transition state selectivity plays a dominant role in controlling the selectivity of zeolite catalysts. The alkylation of benzene with propylene is one of the most commercially important reactions that have been widely used to study the catalytic performance of zeolites.5,6 This reaction is an elementary process in cumene or isopropylbenzene production, an intermediate for phenol production.7 The process can be operated either in the vapor or in liquid phase. A medium-pore zeolite like ZSM-5 is commonly applied as a catalyst to operate in the vapor phase, which usually requires high temperatures and leads to troublesome competing reactions.8 The attempt to tackle this problem has been extended to operate in the liquid phase with the exploitation of the novel large-pore acid zeolites, including modenite (MOR), β-zeolite, and faujasite (Y).5 Under liquidphase conditions, the process is able to be operated at moderate temperatures and consequently acquires better thermal control. * To whom correspondence should be addressed. E-mail: [email protected]. † Department of Chemistry, Kasetsart University. ‡ Kasetsart University of Research and Development Institute, Kasetsart University.

Recently, a new large-pore three-dimensional zeolite named ITQ-24 has successfully been synthesized by Castanˇeda et al.9 The acid form of ITQ-24 has been found to be highly reactive and selective for cumene formation with small amounts of propylene oligomers and n-propylbenzene.9 Although cumene formation has been extensively studied in various conditions, full details of the reaction mechanism are still unclear. Rozanska et al.10 have theoretically investigated the formation of cumene from benzene and propylene with the periodic density functional theory in proton-exchange mordenite (H-MOR). They proposed that the reaction occurs preferentially via the direct rather than the step-by-step mechanism. Similar results have been reported by Nameungrak et al. for the study of benzene alkylation with ethylene in H-FAU.11 However, in the conditions of an excess amount of benzene with respect to propylene, the mechanism was experimentally proposed to occur through two competitive pathways, consecutive and associative mechanisms, depending on the acidity and topology of zeolites.6,12 Therefore, the study of this reaction in different zeolite topologies at molecular level theories may provide a good explanation of the transition state selectivity. To our knowledge, ITQ-24 has never been studied theoretically for catalytic activity in any reactions. In the present work, we thus employ the two-layered our Own N-layered Integrated molecular Orbital + molecular Mechanics (ONIOM) method13 to study the formation of cumene from benzene and propylene to understand details of the reaction such as the activation energy, rate-determining step, and reaction mechanism in this zeolite. 2. Models and Methods The crystal lattice structure of ITQ-24 was taken from the work of Castanˇeda et al.9 The topology is an orthorhombic space group Cmcm with unit cell parameters a ) 21.254 Å, b ) 13.521 Å, and c ) 12.609 Å, respectively. The framework

10.1021/jp061644h CCC: $33.50 © 2006 American Chemical Society Published on Web 06/07/2006

Cumene Formation via Benzene Alkylation

J. Phys. Chem. B, Vol. 110, No. 25, 2006 12627

Figure 1. Illustration of the three types of channels of ITQ-24 zeolite viewed along the c-axis.

system consists of three types of cavities: two 12- and one 10membered ring (MR) channels (see Figure 1). First, the 12MR straight channel runs perpendicularly to the ab plane with pore diameters of about 7.7 × 5.6 Å, and the latter is the sinusoidal channel placed on the a-axis with a pore diameter of about 7.2 × 6.2 Å and one 10MR channel that intersects perpendicularly to both 12MR channels with a pore diameter of about 5.7 × 4.8 Å. In the framework system of ITQ-24, there are four inequivalent tetrahedral sites (T1,...,T4) and 10 oxygen sites (O1,...,O10). The study of incorporating germanium (Ge) in the zeolitic framework found that Ge can be incorporated in the T1 site more than the other sites.9 This result indicates that the T1 site of ITQ-24 is the most preferential local site of aluminum (Al) in the zeolite framework. The proton (Hz) sitting on O10, which is connected with the T1 site and located at the intersection of the three channels, was selected to be the Brønsted active site. The 28T nanocluster model cut from the crystal lattice of ITQ-24 was used to represent the local active site of the ITQ-24 zeolite. The two-layer ONIOM (ONIOM2) scheme was employed to study the reaction mechanisms. For computational efficiency, only the small active region is treated quantum mechanically with the density functional theory method, while the contribution of interactions from the rest of the model is approximated by a less computationally expensive method. The B3LYP/6-31G(d,p) level of theory was applied for the 5T tetrahedral quantum cluster, which is considered to represent the active site. The remainder of the 28T extended framework connected to the 5T active site was treated with the universal force field (UFF). This force field has been found to provide a good description of the short-range van der Waals (vdW) interactions.11,14-15 All calculations were performed using the Gaussian03 code.16 During optimization, only the 5T bare cluster was permitted to relax while the remainders were fixed along the crystallographic coordinates. The frequency calculations were performed at the same level of theory to ensure that an obtained transition state structure has only one imaginary frequency that corresponds to a saddle point of the required reaction coordinate. To include the effect of the electrostatic Madelung potential from the extended zeolite framework on the reaction energy surface, the 5T quantum cluster enclosed with the UFF force field was embedded in the two sets of point charges placed at the atomic zeolite positions. The first one was represented by the fixed partial atomic charges (-1.0 e for O

and +2.0 e for Si) locally positioned at 13-17 Å from the quantum region center, accounting for the medium effect of electrostatic potential. Another one was described by the point charges extended up to 38 Å from the quantum center. These point charges were optimized to reproduce the long-range Madelung potential from the rest of the infinite zeolite lattice. This embedded scheme has been performed in our previous calculation for pyridine adsorption in H-FAU.17 3. Results and Discussion The formation of cumene from benzene alkylation with propylene was considered to follow two possible reaction pathways, consecutive or stepwise, and associative or concerted mechanisms. For the consecutive route, the classical concept of carbocation formation was used to propose the intermediate species generated along the reaction process.18 Due to the fact that the secondary carbocation is more stable than the primary carbocation, in this study, only the isopropoxide, which is the product in the protonation step of propylene, was regarded as the reactive intermediate. This species was developed through the pseudo-2-propyl cation-like transition state. Reaction of this reactive species with an electron-rich benzene molecule finally leads to the cumene product. Unlike in the consecutive reaction mechanism where the propylene protonation and benzene alkylation occur in step-by-step sequence, the protonation and alkylation take place simultaneously and thus there is no intermediate species created through the associative route. Because of the different number and size of molecular species involved in each reaction step, it might be expected that the confinement effect by the zeolite framework becomes the most important contribution to determine the preferential reaction mechanism. 3.1. Consecutive Reaction Pathway. For the consecutive reaction pathway of cumene formation, we were guided by the experimental study of benzene alkylation with propylene in HZSM-11.6 It has been proposed that propylene is readily adsorbed onto the Brønsted acid site in an excess amount of benzene and converted immediately into chemisorbed isopropoxide at the zeolite wall before alkylating benzene. The subsequent reaction leads eventually to the formation of the cumene product. This reaction mechanism has been theoretically modeled in various types of zeolite.10,11 The reaction pathway

12628 J. Phys. Chem. B, Vol. 110, No. 25, 2006

Jansang et al.

Figure 3. Illustration of the optimized geometry of the transition state, Ts_asso, in the concerted reaction pathway. Optimized structures of the adsorption complex, transition state complex, and product are tabulated in Table 3 of the Supporting Information.

Figure 2. Illustration of the optimized geometry of transition states: (a) Ts_pr and (b) Ts_alk, in the consecutive reaction pathway. Optimized structures of isolated reactants, adsorption complexes, transition state complexes, and products are tabulated in Tables 1 and 2 of the Supporting Information.

is initiated with the adsorption of propylene on the zeolite acid site via the hydrogen bonding complex of a propylene CdC bond with the Brønsted proton. This results in the lengthening of the O1-Hz and C1dC2 bonds and hence becomes ready to begin the protonation process. In our calculations, it is found that propylene prefers to adsorb at the intersecting channel between the 12MR and the 10MR channels of the ITQ-24 micropore with the C1‚‚‚Hz and C2‚‚‚Hz distances of 2.115 and 2.333 Å, respectively (see atom labels and Ads_pr in Figures 2a and 4a, respectively). The adsorption energy is predicted to be -12.45 kcal/mol. In the protonation step, propylene is attacked by the acidic proton at the primary carbon atom C1 to generate the more stable secondary carbenium ion. However, it has been shown that only bulky carbocations can be observed experimentally within the zeolite pore.19,20 In contrast, small carbenium ions were theoretically found to be unstable species and usually expressed as the transition state.18,21-22 In this transition state, Ts_pr, the acidic proton, is transferred toward the C1 of propylene (Hz‚‚‚C1 ) 1.211 Å) at the same time that C2 forms a covalent bond with the neighboring zeolitic oxygen framework O2 (C2‚‚‚O2 ) 2.539 Å) to form the isopropoxide species (see Figure 2a). The geometry of the organic fragment at this state is similar to the classical 2-propyl carbocation. The activation energy with respect to the adsorption complex is calculated to be 26.14 kcal/ mol, and the reaction energy for isopropoxide formation with respect to the adsorption complex is endothermic by 3.31 kcal/ mol. Once propylene becomes protonated and converted to isopropoxide, P_alk, it can instantaneously interact with the proximity adsorbing benzene to form the cumene product. Benzene is first physisorbed nearby the isopropoxide and located close to the 12MR window connected to the 10MR channel

(see alk_bz in Figure 4b). The adsorption energy with respect to the isopropoxide intermediate is calculated to be -13.31 kcal/ mol. In the alkylation step, the transition state is associated with the C-C bond formation between benzene and propoxide (see Figure 2b) with the activation energy, Ts_alk, of 43.17 kcal/ mol with respect to the isopropoxide intermediate. The propylene C2 atom forms a bond with the C3 atom of benzene with a distance of 2.141 Å. The bond between the organic fragment and the zeolitic wall is already broken down, and the 2-propyl cation is distorted from the classical carbocations in which the C2 hybridization is more tetrahedral than the one in the protonation step (see also Figure 2a). Due to the steric hindrance of the bulky transition state structure, therefore, using a small quantum cluster without the zeolite framework constraints may not be enough to give reliable results.23 It is to be noted that the transition state structure is not concerned with proton backdonation to the zeolite framework but only to the C-C bond formation. Therefore, the existence of the alkylated carbocation intermediate becomes questionable. It has been found that the carbocation product can be observed theoretically and is slightly more stable than the transition state.24 However, there has been no experimental data reported on the stability of this intermediate species. The reaction is completed with the formation of cumene by proton back-donation from the alkylated benzene to the zeolitic framework (see P_cumene in Figure 4b). The reaction energy is exothermic by 19.44 kcal/mol. 3.2. Associative Reaction Pathway. Alternatively, for the associative or concerted mechanism, the reaction is started with the coadsorption complex, Co_ads, between propylene and benzene on the Brønsted acid site (see Co_ads in Figure 5). The propylene is adsorbed on the acidic proton via the hydrogen bond complex, similar to that in the associative mechanism (C1‚ ‚‚Hz ) 2.083 Å and C2‚‚‚Hz ) 2.426 Å) with the neighboring adsorption of the benzene molecule. The adsorption energy with respect to the isolated reactants is calculated to be -25.02 kcal/ mol. After the adsorption of both propylene and benzene, the protonation and alkylation processes occur simultaneously through the transition state shown in Figure 3. In this transition state, Ts_asso, the proton is transferred to protonate propylene at the C1 atom (C1‚‚‚Hz ) 1.335 Å) at the same time that C2 forms a bond with the C3 atom of benzene (C2‚‚‚C3 ) 2.321

Cumene Formation via Benzene Alkylation

Figure 4. Calculated energy profiles for the consecutive reaction pathway: (a) protonation step and (b) alkylation step. The energies in parentheses were derived from embedded ONIOM2 calculations.

Å). No alkoxide intermediates are generated along this reaction pathway. The activation energy with respect to the coadsorption complex for this process is estimated to be 45.24 kcal/mol. The overall geometry of the associative transition state structure is similar to the transition state in the alkylation step of the consecutive mechanism. The cumene product, P_cumene, is formed after the back-donation of the proton to the zeolite framework. The reaction energy for the cumene formation with respect to the coadsorption complex is exothermic by 9.98 kcal/ mol. The reaction energy profiles of both the consecutive and the associative reaction mechanisms of cumene formation are shown in Figures 4 and 5. It can be seen that the consecutive reaction pathway is dominant in the cumene formation. The alkylation step, rather than the propylene protonation, is found to be the rate-limiting step. This result is consistent with experimental reports that propylene is easily protonated in acidic zeolite and that the consecutive mechanism becomes dominant

J. Phys. Chem. B, Vol. 110, No. 25, 2006 12629 for cumene formation where the amount of benzene is relatively large as compared with propene.6 It is to be noted that, by using the ONIOM2 method, only the small active site (5T cluster) and the adsorbing molecules are treated quantum mechanically, whereas the rest is described by the UFF force field which does not account for electrostatic interactions. Therefore, the energy of the system is mainly contributed by the van der Waals interactions which are rather constant through the reaction pathway and are very important to the adsorption process of aromatic molecules in zeolite pores.14,25 However, when dealing with the ionic species such as the transition states in this study, the electrostatic interaction from the extended zeolite framework cannot be neglected.17,24 To take into account such electrostatic potential on the reaction energy surface, single-point energy calculations at the same level of theory were carried out by using the embedded scheme. The same trend as computed by the ONIOM2 method is obtained: the consecutive mechanism is preferred over the associative route. Depending on the structural geometry, the intermediate complexes are stabilized by at most 2.3 kcal/mol with respect to the ONIOM2 energies (see Figures 4 and 5). Much difference is found in the stabilization energy of the transition states. In the alkylation step, the activation energy is lowered by 7.46 kcal/mol relative to the ONIOM2 method for the consecutive reaction, while it is lowered only by 1.55 kcal/mol for the associative reaction. This is in accordance with the higher ionicity of the transition state organic fragment of the consecutive reaction Ts_alk, C9H13+ (+0.899 e), as compared with that of the associative reaction Ts_asso (+0.762 e). The zeolite proton has already transferred to the organic fragment in the former case, while it is between the zeolite framework and the organic fragment in the latter case. The activation energies of the alkylation step are corrected to be 35.70 and 43.69 kcal/ mol for the consecutive and associative mechanisms, respectively. The former is comparable with the alkylation activation energy of benzene alkylation with propylene in β-zeolite of 34.9 kcal/mol.26 Concerning the size of the quantum clusters, we should mention that a former adsorption study of small molecules (NO, CO, and CH3OH)27 indicated that good convergence of adsorption properties could already be reached by the 5T cluster. Similar conclusions also hold up well for larger molecules, such as pyridine.17 To test this aspect, single-point energy calculations of full quantum clusters of 28T at the B3LYP/6-31G(d,p) level of theory were used to study the transition state stability. The rate-determining steps of both concerted and stepwise reactions were carried out. The same trend as computed by the successfully calibrated 5T/28T ONIOM2(B3LYP:UFF) is obtained: the apparent activation energies for the concerted mechanism are 16.97 and 17.42 kcal/mol for the 5T/28T ONIOM2(B3LYP: UFF) and the 28T QM(B3LYP), respectively, while the corresponding apparent activation energies for the alkylation step of the stepwise mechanisms are 12.81 vs 11.69 kcal/mol. Comparison of the results contained in this study with the same reaction in other similar large-pore zeolites may lead to the conceptual idea of the shape-selective property of different zeolites. In contrast to our present study, the associative reaction was drastically preferred for the cumene formation in H-MOR.10 The activation energies of the alkylation step were virtually the same in both the consecutive and associative reactions, whereas the protonation step of propylene in the consecutive reaction required a much higher activation energy and was referred to as the rate-determining step. From these results, one can see

12630 J. Phys. Chem. B, Vol. 110, No. 25, 2006

Jansang et al.

Figure 5. Calculated energy profiles for the concerted reaction pathway. The energies in parentheses were derived from embedded ONIOM2 calculations.

that the favored reaction pathway is highly selective to the topology of the local active site of zeolites. 4. Conclusions The combined QM/MM calculations at the B3LYP/6-31G(d,p):UFF method utilizing an ONIOM2 scheme have been performed to study the cumene formation through benzene alkylation with propylene in the nanoporous ITQ-24 zeolite. To understand the zeolite confinement effects on the transition state selectivity, both consecutive and associative reaction mechanisms have been examined for comparison purposes. In contrast to previous theoretical studies of the alkylation of benzene with small olefins in large-pore zeolites, the consecutive reaction pathway is more preferred over the associative mechanism in ITQ-24 zeolite. The alkylation step of benzene with the reactive isopropoxide intermediate is the reaction ratelimiting step with the corrected activation barrier of 35.70 kcal/ mol, which compares well with an experimental activation energy of 34.9 kcal/mol for benzene alkylation in β-zeolite. By use of the ONIOM2 method in combination with the newly developed electronic embedding scheme, it can be seen that the electrostatic potential from the extended zeolite framework plays a major role in the transition state selectivity with minor support from the van der Waals contribution. Compared with other similar large-pore dimensional zeolites, one can also obtain from this study that the transition state selectivity is highly sensitive to the zeolite topology. Acknowledgment. This work was supported in part by grants from the Thailand Research Fund (TRF Senior Research Scholar to J.L.) and the Kasetsart University Research and Development Institute (KURDI), as well as the Ministry of University Affairs under the Science and Technology Higher Education Development Project (MUA-ADB funds). The support from the National Nanotechnology Center (NANOTEC, Thailand) is also acknowledged.

Supporting Information Available: Tables giving the optimized structures of isolated reactants, adsorption complexes, intermediates, transition states, and products for the consecutive and associative reaction pathways. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Ackley, M. W.; Rege, S. U.; Saxena, H. Microporous Mesoporous Mater. 2003, 61, 25-42. (2) Hattori, H. Chem. ReV. 1995, 95, 537. (3) Clearfield, A. Chem. ReV. 1988, 88, 125. (4) Corma, A. Chem. ReV. 1995, 95, 559-614. (5) Perego, C.; Amarilli, S.; Millini, R.; Bellussi, G.; Girotti, G.; Terzoni, G. Microporous Mater. 1996, 6, 395-404. (6) Derouane, E. G.; He, H.; Hamid, S. B. D.-A.; Ivanova, I. I. Catal. Lett. 1999, 58, 1-19. (7) Pujado, P. R.; Salazar, J. R.; Berger, C. V. Hydrocarbon Process. 1976, 55, 91. (8) Kaeding, W. W.; Holland, R. E. J. Catal. 1988, 109, 212-216. (9) Castanˇeda, R.; Corma, A.; Forne´s, V.; Rey, F.; Rius, J. J. Am. Chem. Soc. 2003, 125, 7820-7821. (10) Rozanska, X.; Barbosa, L. A. M. M.; van Santen, R. A. J. Phys. Chem. B 2005, 109, 2203-2211. (11) Namuangruk, S.; Pantu, P.; Limtrakul, J. J. Catal. 2004, 225, 523530. (12) Fu, J.; Ding, C. Catal. Commun. 2005, 6, 770-776. (13) Maseras, F.; Morokuma, K. J. Comput. Chem. 1995, 16, 11701179. (14) Namuangruk, S.; Pantu, P.; Limtrakul, J. ChemPhysChem 2005, 6, 1333-1339. (15) Pantu, P.; Pabchanda, S.; Limtrakul, J. ChemPhysChem 2004, 5, 1901-1906. (16) 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.

Cumene Formation via Benzene Alkylation 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 B.05; Gaussian, Inc.: Wallingford, CT, 2004. (17) Injan, N.; Pannorad, N.; Probst, M.; Limtrakul, J. Int. J. Quantum Chem. 2005, 105, 898-905. (18) Rozanska, X.; Demuth, T. H.; Hutschka, F.; Hafner, J.; van Santen, R. A. J. Phys. Chem. B 2002, 106, 3248-3254. (19) Haw, J. F.; Nicholas, J. B.; Xu, T.; Beck, L. W.; Ferguson, D. B. Acc. Chem. Res. 1996, 29, 259-267. (20) Xu, T.; Haw, J. F. J. Am. Chem. Soc. 1994, 116, 7753-7759. (21) Boronat, M.; Zicovich-Wilson, C. M.; Viruela, P.; Corma, A. J. Phys. Chem. B 2001, 105, 11169-11177.

J. Phys. Chem. B, Vol. 110, No. 25, 2006 12631 (22) Bhan, A.; Joshi, Y. V.; Delgass, W. N.; Thomson, K. T. J. Phys. Chem. B 2003, 107, 10476-10487. (23) Vos, A. M.; Schoonheydt, R. A.; Proft, F. D.; Geerling, P. J. Phys. Chem. B 2003, 107, 2001-2008. (24) Vos, A. M.; Rozanska, V.; Schoonheydt, R. A.; van Santen, R. A.; Hutschka, F.; Hafner, J. J. Am. Chem. Soc. 2001, 123, 2799-2809. (25) Rozanska, X.; van Santen, R. A.; Hutschka, F.; Hafner, J. J. Am. Chem. Soc. 2001, 123, 7655-7667. (26) Sridevi, U.; Rao, B. K. B.; Pradhan, N. C.; Tambe, S. S.; Satyanarayana, C. V.; Rao, B. S. Ind. Eng. Chem. Res. 2001, 40, 31333138. (27) Treesukol, P.; Limtrakul, J.; Truong, T. J. Phys. Chem. B 2001, 105, 2421-2428.