ARTICLE pubs.acs.org/JPCA
Theoretical Study of 1,2-Hydride Shift Associated with the Isomerization of Glyceraldehyde to Dihydroxy Acetone by Lewis Acid Active Site Models Rajeev S. Assary*,†,‡ and Larry A. Curtiss*,†,§ †
Materials Science Division, Argonne National Laboratory, Argonne, Illinois 60439, United States Chemical & Biological Engineering, Northwestern University, Evanston, Illinois 60208, United States § Centers for Nanoscale Materials, Argonne National Laboratory, Argonne, Illinois 60439, United States ‡
ABSTRACT: The isomerization of glyceraldehyde to dihydroxy acetone catalyzed by the active site of Sn-beta zeolite is investigated using the B3LYP density functional and MP2 levels of theory. Structural studies were aimed to understanding the binding modes of glyceraldehyde with the active site, and the detailed free energy landscape was computed for the isomerization process. The rate-limiting step for the isomerization is the 1,2-hydride shift, which is enhanced by the active participation of the hydroxyl group in the hydrolyzed Sn-beta active site analogues to the one seen in the xylose isomerase. On the basis of the assessment of the activation barriers for isomerization by the Sn, Zr, Ti, and Si zeolite models, the activity of the catalysts are in the order of Sn > Zr > Ti > Si in aqueous dielectric media.
1. INTRODUCTION Conversion of glucose to hydroxy-methyl-furfural (HMF), levulinic acid, and lactic acid is one of the key steps for utilizing biomass as chemical feedstock and producing a platform chemical for alternative fuel precursors.1 8 Both experimental and theoretical studies have a lot to offer in this area of biomass conversion to platform chemicals and the molecular level understanding of reaction pathways.2 4,9 11 The isomerization of glucose to fructose is very significant, due to its high selectivity to produce HMF and in the industrial production of highfructose corn syrup.12 Recent studies of such isomerization of aldoses to ketoses in nonaqueous solution catalyzed by zeolites for the hexoses and trioses also deserves special attention due to their significance in producing alternative platform chemicals, such as alkyl lactates from sugar molecules.5,13 These isomerization processes typically can be catalyzed by enzymes and inorganic bases and acids, where industrial level enzymatic catalysis is limited by processing costs, pH condition, narrow temperature ranges, etc.14 Base-catalyzed reactions are limited by their low yields and stability of monosaccharides in basic solution.15,16 Similarly, removal of water through mineral acid-catalyzed reaction of glucose leads to undesired products or “humins” in addition to the desired product HMF, which limits the use of mineral acids. Heterogeneous porous catalysts with the desired Lewis acidity in aqueous solution and with an adequate pore size that allows diffusion of hexoses are increasingly envisioned as the ideal candidates for biomass conversion. Recently, Moliner et al. have reported on a tin-containing zeolite (Sn-beta) that acts as a highly efficient catalyst for the glucose fructose isomerization in aqueous solution through a hydride shift.17,18 r 2011 American Chemical Society
The Sn-beta zeolite was shown to be active over a wide range of temperatures and acidic and organic solutions and to possess strong interactions with hydroxyl/keto compounds.19,20 Corma and co-workers have reported the detailed nature of the Sn-beta catalytic center and the catalysis of reduction of keto compounds.19,21,22 These zeolite frameworks can also perform the reduction of keto groups upon substitution of tin with Zr, Ta, and Nb.23,24 It has been suggested that the actual active site of Sn-beta is the (SiO)3Sn(OH) center, and previous density functional studies were reported using this model to understand the energetics and mechanism of the Meerwein Ponndroff Verley Oppenauer (MPVO) reactions between cyclohexanone and 2-butanol.25 In this work, the objective is to understand the aldose ketose conversion process catalyzed by the Lewis acids for the glyceraldehyde to dihydroxy acetone system (Scheme 1), which was chosen due to its significance in alkyl lactate or lactic acid production by zeolites and its similarity to the glucose-to-fructose isomerization. Our investigation aims to understand the following: (1) the initial binding of glyceraldehyde to the zeolite active site model of Sn-beta, (2) the detailed free energy profile of the reaction, and (3) the effect of substitution (by Zr, Ti, Si) on isomerization. The basic mechanism of isomerization can occur through acidcatalyzed (hydride shift), base-catalyzed (proton shift), or at neutral (proton-coupled hydride shift) and is shown for the isomerization of glyceraldehyde to dihydroxy acetone in Scheme 2, which is a similar mechanism to that of the glucose. Received: May 10, 2011 Revised: June 25, 2011 Published: June 27, 2011 8754
dx.doi.org/10.1021/jp204371g | J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A
ARTICLE
Computational details are given in the next section, and the detailed nature of binding structures and the energetics of isomerization are presented in the Results and Discussion section.
2. COMPUTATIONAL DETAILS The B3LYP26 functional is used to study the geometry of all of the species involved in this investigation. The 6-31+G(d) (BS1), 6-31+G(2df,p) (BS2), 6-311+G(3df,2p) (BS3), and 6-311++G(3df,3pd) (BS4) basis sets were used for first and second row elements. Comparison of energetics for isomerization evaluated from the density functional theory calculations were made with the Scheme 1. Isomerization Processes Investigated in This Worka
MP2/BS4, G4,27 and CCSD(T)/BS3//B3LYP/BS2 levels of theory. For geometry optimization, the Lewis acid central atom is treated using CC-pVDZ-PP,28 LANL2TZ(f),29 Ahlrich’s VTZ,30 and 6-31+G(d) for Sn, Zr, Ti, and Si at the B3LYP level of theory. To address the dispersion energies and basis set superposition errors, a single-point energy evaluation is performed at the MP2 level of theory with aug-CC-pVQZ (BS5) for Sn,28Ti,31 and Zr32 and BS4 for the rest of the atoms using the geometries obtained from the density functional studies. Effective core potentials (ECPs) for Sn and Zr, including relativistic effects, were used in DFT and MP2 calculations.33 Frequency calculations were performed to verify the nature of all the stationary points as either minima or transition states (TSs) and to provide zeropoint energy (ZPE) corrections. Free energy and enthalpy corrections at 298 K were also evaluated using the B3LYP level of theory. To account for the effects of the aqueous environment, calculations were also performed in water dielectric using the recently developed SMD solvation model34 at the B3LYP level of theory with the same basis sets used for the geometry evaluations. The calculations for this investigation were done using Gaussian 09.35
3. RESULTS AND DISCUSSION a
The circled hydrogen undergoes hydride shift during the isomerization.
3.1. Glyceraldehyde to Dihydroxy Acetone. 3.1.1. CatalystFree Isomerization. Initially, we investigated the energetics of
acid-catalyzed, base-catalyzed, and uncatalyzed isomerization of glyceraldehyde to dihydroxy acetone, as shown in Scheme 2. The key step in the acid- and base-catalyzed isomerizations are the hydrogen shift and proton transfer, respectively, whereas in a neutral medium, isomerization proceeds through a concerted proton-coupled hydrogen shift (PCHS). The computed enthalpy of the reaction (ΔHrxn), enthalpy of activation (ΔH†), Gibbs free energy of reaction (ΔGrxn), and the Gibbs free energy of activation (ΔG†) are tabulated in Table 1 at various levels of theory. Compared to accurate G4 or CCSD(T)/BS3 levels of theory, density functional calculations underestimates the computed barriers by 1 3 kcal/mol. The computed activation
Scheme 2. Schematic Representation of Isomerization of Glyceraldehyde to Dihydroxy Acetone in Acid, Base, and Neutral Media
Scheme 3. Active Site Model of the T-9 Site of Sn-beta
Table 1. Computed Enthalpy of Reaction (ΔHrxn), Activation Enthalpy Barrier (ΔHrxn), Free Energy of the Reaction (ΔGrxn), and Activation Gibbs Free Energy Barrier (ΔG†) for Tautomerization of Glyceraldehyde to Dihydroxy Acetone at 298 K Using the CCSD(T), G4, and the B3LYP Levels of Theorya acid catalyzed method
ΔHrxn
ΔH
†
ΔGrxn
base catalyzed †
ΔG
ΔHrxn
†
ΔH
ΔGrxn
neutral ΔG
†
ΔHrxn
ΔH
†
ΔGrxn
ΔG†
CCSD(T)/BS3
5.7
14.0
5.9
15.0
1.9
6.4
1.8
7.2
3.2
42.2
4.0
43.1
G4
6.1
14.4
6.2
15.4
1.9
5.6
1.8
6.5
3.3
41.6
4.0
42.5
B3LYP/BS2
7.1
11.6
7.2
12.6
2.4
6.1
2.3
7.0
4.4
38.0
5.2
38.8
B3LYP/BS1
4.9
12.9
4.7
13.5
0.3
7.8
0.0
11.0
4.4
39.6
3.8
40.5
MP2/BS4
6.4
13.6
6.2
14.2
1.2
4.3
0.9
4.9
3.5
42.3
2.9
43.3
a
The labels are BS1 = 6-31+G(d), BS2 = 6-31G(2df,p), BS3 = 6-311+G(3df,2p), and BS4 = 6-3111++G(3df,3pd). The enthalpy and free energy corrections from the B3LYP/BS2 and B3LYP/BS1 levels of theory were added to the CCSD(T) and the MP2 methods. All values are reported in kcal/ mol. 8755
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A
ARTICLE
Figure 1. Computed gas-phase Gibbs free energy profile (298 K) for the isomerization of glyceraldehyde to dihydroxy acetone using the Sn(OSiH3)3 OH model at the MP2 level of theory. The free energy corrections were computed at the B3LYP level of theory. All energies are reported in kcal/mol.
Figure 2. Computed gas-phase Gibbs free energy profile (298 K) for the isomerization of glyceraldehyde to dihydroxy acetone using the Sn(OSiH3)3 OH (type 2) model at the MP2 level of theory. The free energy corrections were computed at the B3LYP level of theory. All energies are reported in kcal/mol.
barriers at the MP2/BS4//B3LYP/BS1 level of theory are within 1 kcal/mol of the G4 level of theory. Isomerization of glyceraldehyde to dihydroxy acetone is thermodynamically downhill (by 4.0 kcal/mol) and requires an activation energy of 41.6 kcal/mol in the gas phase (computed at the G4 level of theory), whereas the acid-catalyzed hydrogen shift requires an activation of 14.4 kcal/mol. Therefore, it is evident that deprotonation of the hydroxyl group and the activation of the keto group significantly reduces the activation barrier for the hydride shift. 3.1.2. Isomerization Catalyzed by the Sn-beta Active Site Models. Similar to the previous density functional study by Corma et al.,25 an active site model for the T-9 site of Sn-beta is employed in this study (shown in Scheme 3). We note that binding of molecules, such as glyceraldehyde, in the zeolite pore will be affected by the dispersive and van der Waals interactions from the atoms in the cavity. Therefore, the magnitude of the binding energy may
depend on the pore size, dispersion, solvation energy of the substrate, and the cavity. An accurate calculation would require an extensive computational study beyond the scope of the current work, which is focused on relative energies. The energetics of all binding and subsequent reaction pathways were computed at the MP2 levels of theory with sufficiently large basis sets (combination of BS4 and BS5) to address the dispersion interactions for the small cluster model system employed here. It is believed that the partially hydrolyzed tin active site (Si O )3 Sn OH center is responsible for the isomerization of aldose to ketose by protonation and deprotonation of the carbohydrate with Sn OH center analogues to that one seen in xylose isomerase.36 Here, we have considered both the hydrolyzed and the unhydrolyzed active site models to understand the effect of active participation of the hydroxyl group in the isomerization energetics. The unhydrolyzed model can be represented by the substitution of hydroxyl group (Scheme 3) OSiH3 group. 8756
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A
ARTICLE
Two types of binding modes were considered for the Sn(OSiH3)3 OH model. In the first mode (Figure 1), the hydroxyl group is involved directly during the entire pathway of the reaction, and in the second mode (Figure 2), the hydroxyl group acts more like a spectator and is not involved in any reaction sequence. The computed Gibbs free energy (gas phase, 298 K) profile at the MP2 level of theory for the first mode of binding is shown in Figure 1. All the energies are given in kcal/mol and are with respect to isolated glyceraldehyde and the Sn active site model. The detailed energetics, including electronic energy, enthalpy, and free energy, at the B3LYP and MP2 levels of theory are shown in Table 2. Because of the accuracy of MP2 methods for predicting the barriers, the energetics based on the MP2 level of theory is discussed throughout this paper. The binding of the glyceraldehyde molecule on the Sn active site center occurs by Table 2. Computed Gas-Phase Relative Electronic Energy (ΔEe), Enthalpy (ΔH298 K), and Free Energy (ΔG298 K) of Intermediates (II to VIII wrt I) at the B3LYP and the MP2 Levels of Theorya B3LYP ΔEe
a
ΔH298 K
MP2 ΔG298 K
ΔEe
ΔH298 K
ΔG298 K
I
0.0
0.0
0.0
0.0
0.0
II
1.2
2.5
13.4
8.1
9.4
0.0 6.6
III
17.3
15.6
0.1
22.5
20.8
5.1
IV V
3.2 20.8
2.4 19.4
19.2 4.1
2.1 24.5
3.0 23.0
13.8 7.8
VI
6.7
8.2
7.8
12.0
13.4
2.5
VII
14.4
12.7
1.3
18.2
16.5
2.5
VIII
3.7
4.4
3.8
2.8
3.5
2.9
See Figure 1 for the molecular structures of intermediates I VIII. The geometries and enthalpy and free energy corrections were computed at the B3LYP/BS1 level of theory, where the CC-pVDZ-PP basis set is employed for the Sn atom. All values are reported in kcal/mol.
condensation of a water molecule from the hydroxyl group of the tin active site and the proton from the hydroxyl group of the glyceraldehyde hydroxyl group. This process is thermodynamically downhill by 5.1 kcal/mol and requires an activation free energy of 6.6 kcal/mol. Upon formation of a water molecule, the glyceraldehyde anion binds strongly to the Sn center and, hence, forms a stable intermediate (III). The hydride shift from the C2 carbon to C1 carbon occurs through the transition state (IV), which requires an apparent activation free energy barrier of 13.8 kcal/mol. Because the formation of intermediate III is thermodynamically downhill (5.1 kcal/mol), the barrier height is significant here (18.9 kcal/ mol), and this could be affected by the free energy of formation of the initial complex formation when considering all free energy contributions of binding in an active site pore of Sn-beta zeolite. Overall, from our calculations, the hydride shift is thermodynamically downhill from intermediate III to V (by 1.6 kcal/mol). Notably, the computed activation barrier for the hydride shift (13.8 kcal/mol) here is significantly smaller than the activation free energy barriers for the uncatalyzed (42.5 kcal/mol) or acidcatalyzed isomerization (15.4 kcal/mol, Table 1). The next step is the protonation of the oxygen atom of the C1 carbon of the complex (V) by the deprotonation process of the water ligand attached to the Sn center. This requires a relatively small activation free energy barrier of 2.5 kcal/mol (VI) and leads to the formation of dihydroxy acetone complexed with the active site through the coordinative bond between the carbonyl group and the tin center. Desorption of the dihydroxy acetone from this complex is thermodynamically downhill (by 0.5 kcal/mol), as is the overall process of conversion of glyceraldehyde to dihydroxy acetone (by 2.9 kcal/mol). In the second mode of binding for the Sn(OSiH3) OH center with the glyceraldehyde substrate, we have considered the hydroxy group as a spectator, and the detailed free energy landscape for the isomerization of glyceraldehyde to dihydroxy acetone is shown in Figure 2. The free energy analysis shows that the formation of the initial complex between the Sn center and glyceraldehyde (II) is thermodynamically uphill by 4.1 kcal/mol.
Figure 3. Computed Gibbs free energy profile (gas phase, 298 K) for the isomerization of glyceraldehyde to dihydroxy acetone using the Sn(OSiH3)4 model complex at the MP2 level of theory. The free energy corrections were computed at the B3LYP level of theory. All energies are reported in kcal/mol. 8757
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A
ARTICLE
The isomerization to dihydroxy acetone proceeds through a stepwise mechanism, via proton transfer (III) and hydride shift (V) transition states. These processes require apparent activation free energy barriers of 19.5 and 23.3 kcal/mol, respectively, and both processes are thermodynamically downhill from the initial complex (II). From Figures 1 and 2, the hydride shift is computed as the ratelimiting step and the involvement of the hydroxyl group is vital to the isomerization through a relatively lower barrier, 13.8 kcal/mol, compared to 23.3 kcal/mol when the zeolite active site has no hydroxyl group. We further assessed the energetics of the isomerization on the tin site when lacking the hydroxyl functional groups or the unhydrolyzed sites. The computed free energy profile is shown in Figure 3. Similar to the glyceraldehyde binding arrangement with the tin center, where the hydroxyl group acts Table 3. Computed Relative Free Energies (Gas Phase, 298 K) for the Tautomerization of Glyceraldehyde to Dihydroxy Acetone by the 9-T Cluster of Tin-beta Zeolite (Si (SiO)3(OH)) at the MP2/BS4 and BS5 Levels of Theorya
intermediates
Ti(OSiH3)3 OH
Zr(OSiH3)3 OH
Si(OSiH3)3 OH
ΔG (298 K)
ΔG (298 K)
ΔG (298 K)
ΔG (298 K)
I
0.0
0.0
0.0
0.0
II (TS)
6.6
8.4
3.5
36.6
5.1
1.1
12.7
3.8
IV (TS)
13.8
19.6
7.4
40.9
V VI (TS)
7.8 2.5
4.5 4.6
15.8 7.3
2.1 34.8
III
a
Sn(OSiH3)3 OH
VII
2.4
1.3
6.5
1.2
VIII
2.9
2.9
2.9
2.9
The geometry and free energy corrections were obtained using the B3LYP/BS1 level of theory (see the Computational Details section for the basis set details of metal atoms). All energies are reported in kcal/ mol.
as a spectator, the computed activation free energy for the hydride shift in this model is also significant (23.9 kcal/mol) compared to the binding mode when the hydroxyl group is actively participating (Figure 1). Additionally here, the initial binding of glyceraldehyde with the tin center is computed to be thermodynamically uphill (6.4 kcal/mol). Thus, from the free energy landscapes of various binding modes of glyceraldehyde in the tin active site models, we can conclude that the Sn(OSiH3)3 OH active site with the active participation of the hydroxyl group in the reaction catalyzes the isomerization of glyceraldehyde to dihydroxy acetone effectively by providing a lower free energy barrier for the rate-limiting hydride transfer compared to the rest of the models that we have considered. Thus, the presence of the hydroxyl group is a key factor in the catalytic activity of the Sn-beta zeolite for the isomerization of glyceraldehyde to dihydroxy acetone. 3.1.3. Comparison of Sn/Ti/Si/Zr Activities. To understand how the substitution of the central metal affects the catalytic property of the Sn-beta active site, we have performed computations to evaluate the free energy landscape of the isomerization of glyceraldehyde by substituting Sn with Si, Ti, and Zr. All reaction profiles were similar to that of the glyceraldehyde/Sn OH active site with protonation and deprotonation mediated by the hydroxyl group (Figure 1). The computed Gibbs free energies (gas phase, 298 K in kcal/mol) of all intermediates and transition states at the MP2 level of theory are shown in Table 3. Comparing the free energy profiles of all the molecular species, it is obvious that the rate-limiting step for all metal active sites is the hydride shift. The optimized structures of the transition state for the rate-limiting hydride shift considered in this investigation are shown in Figure 4. Figure 5 compares the activation free barriers of the hydride shift for all metal active sites relative to the reactants at 298 K in the gas phase and in the water dielectric medium. Experimentally, the conversion efficiency and activity are in the order of Sn > Ti > Zr > Si, where efficiency of the Ti and Zr-beta catalysts were found to be very similar.13 The computed activation free energy barrier of Zr is 7.4 kcal/mol, which is much lower than the
Figure 4. Optimized geometries of the transition states for the rate-limiting hydride shift at the B3LYP level of theory. 8758
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A
ARTICLE
Figure 5. Relative Gibbs free energy of activation (298 K) at gas phase (a) and in the water dielectric (b) of the rate-limiting step for the isomerization of glyceraldehyde to dihydroxyl acetone at the MP2 level of theory using the M(OSiH3)3 OH model complex, where M = Sn, Zr, Ti, and Si. The solvation free energy is computed using the SMD solvation model at the B3LYP/BS1 level of theory. All energies are reported in kcal/mol.
activation free energy barrier catalyzed by the tin active site at the gas phase. The computed activation free energy barriers for the hydride shift when Sn is substituted by Ti and Si are 19.7 and 40.9 kcal/mol, respectively. A previous theoretical calculation for the MPVO reduction of cyclohexanone with 2-butanol reported a lower activation energy for the Zr active site model compared to the Sn active site for the hydride shift in the gas phase.25 This is in agreement with our calculation. Considering the effects from the aqueous dielectric significantly changed the order of the computed apparent activation barriers for the hydride shift catalyzed by various zeolite models (Figure 5b).The computed apparent activation free energy for the isomerization is 15.4, 23.0, 28.2, and 40.7 kcal/mol, respectively, for Sn, Zr, Ti, and Si-beta model complexes in the water dielectric medium. The computed apparent activation barrier for the rate-limiting step catalyzed by the Sn(OSiH3)4 model complex in aqueous dielectric is 25.1 kcal/mol, and this indicates the lower catalytic activity for the unhydrolyzed tin active site compared to the hydrolyzed site. The effect of an aqueous dielectric solvent on the Zr active site model complex is found to be more significant than for the Ti model complex. Introduction of an aqueous dielectric increased the apparent activation barriers for the isomerization by 15.6 and 8.5 kcal/mol for Zr and Ti active site models, respectively. Experimentally, Sn-beta, Ti-beta, and Zr-beta were found to promote the hydride shift for the conversion of pyruvaldehyde hemiacetal to methyl lactate or pyruvaldehyde hydrate to lactic acid; however, no activity was observed for pure Si-beta zeolite.13 This is in excellent agreement with our computations, where the apparent activation free barrier for the rate-limiting step catalyzed by the Si-model complex is about 41 kcal/mol both in gas phase and in an aqueous dielectric medium.
4. CONCLUSIONS The isomerization of glyceraldehyde to dihydroxy acetone was investigated using density functional theory and the MP2 level of theory to gain insight into the catalytic activity of the tin-beta
active site using a simple cluster model of the active site. On the basis of the detailed structural and free energy profiles investigated here, the following conclusions can be drawn from this study: (1) The rate-limiting step of the acid-catalyzed isomerization of the glyceraldehyde to dihydroxy acetone is the 1, 2-hydride shift. (2) The isomerization requires an active participation of the hydroxyl group coordinated to the Sn-beta active site. In gas phase, the activation free energy barrier for the ratelimiting hydride transfer is 13.8 kcal/mol for the hydrolyzed model, whereas the barrier required is 23.3 kcal/mol for an unhydrolyzed active site. (3) On the basis of the assessment of the activation barriers for isomerization by the Sn, Zr, Ti, and Si zeolite models, the activity of the catalysts is in the order of Sn > Zr > Ti > Si in aqueous dielectric media. Future studies are essential to understanding the effect of the zeolite cavity, such as dispersion, diffusion, and mass transfer, during the catalysis. This investigation provides an ideal platform to investigate the catalytic activity, kinetics, and structure of isomerization for the complex hexoses, such as glucopyranose, by the Sn-beta zeolites.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected] (R.S.A),
[email protected] (L.A.C.). Phone: 630-252-7020 (R.S.A), 630-252-7380 (L.A.C.). Fax: 630-252-9555 (R.S.A), 630-252-9555 (L.A.C.).
’ ACKNOWLEDGMENT This work was supported by the U.S. Department of Energy under Contract DE-AC0206CH11357. This material is based upon work supported as part of the Institute for Atom-efficient Chemical Transformations (IACT), an Energy Frontier 8759
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
The Journal of Physical Chemistry A Research Center funded by the U.S. Department of Energy, Office of Science, and Office of Basic Energy Sciences. We gratefully acknowledge grants of computer time from EMSL, a national scientific user facility located at Pacific Northwest National Laboratory, the ANL Laboratory Computing Resource Center (LCRC), and the ANL Center for Nanoscale Materials.
ARTICLE
(35) 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.; et al. Gaussian 09, rev. A.02; Gaussian, Inc.: Wallingford, CT, 2009. (36) Kovalevsky, A. Y.; Hanson, L.; Fisher, S. Z.; Mustyakimov, M.; Mason, S. A.; Trevor Forsyth, V.; Blakeley, M. P.; Keen, D. A.; Wagner, T.; Carrell, H. L.; et al. Structure 2010, 18, 688.
’ REFERENCES (1) Chheda, J. N.; Huber, G. W.; Dumesic, J. A. Angew. Chem., Int. Ed. 2007, 46, 7164. (2) Roman-Leshkov, Y.; Barrett, C. J.; Liu, Z. Y.; Dumesic, J. A. Nature 2007, 447, 982. (3) Huber, G. W.; Chheda, J. N.; Barrett, C. J.; Dumesic, J. A. Science 2005, 308, 1446. (4) Huber, G. W.; Iborra, S.; Corma, A. Chem. Rev. 2006, 106, 4044. (5) Holm, M. S.; Saravanamurugan, S.; Taarning, E. Science 2010, 328, 602. (6) Regalbuto, J. R. Science 2009, 325, 822. (7) Zhao, H.; Holladay, J. E.; Brown, H.; Zhang, Z. C. Science 2007, 316, 1597. (8) Corma, A.; Iborra, S.; Velty, A. Chem. Rev. 2007, 107, 2411. (9) Assary, R. S.; Redfern, P. C.; Hammond, J. R.; Greeley, J.; Curtiss, L. A. Chem. Phys. Lett. 2010, 497, 123. (10) Assary, R. S.; Redfern, P. C.; Greeley, J.; Curtiss, L. A. J. Phys. Chem. B 2011, 115, 4341. (11) Assary, R. S.; Redfern, P. C.; Hammond, J. R.; Greeley, J.; Curtiss, L. A. J. Phys. Chem. B 2010, 114, 9002. (12) Nikolla, E.; Roman-Leshkov, Y.; Moliner, M.; Davis, M. E. ACS Catal. 2011, 1, 408. (13) Taarning, E.; Saravanamurugan, S.; Holm, M. S.; Xiong, J.; West, R. M.; Christensen, C. H. ChemSusChem 2009, 2, 625. (14) Bhosale, S. H.; Rao, M. B.; Deshpande, V. V. Microbiol. Rev. 1996, 60, 280. (15) De Wit, G.; Kieboom, A. P. G.; van Bekkum, H. Carbohydr. Res. 1979, 74, 157. (16) Yang, B. Y.; Montgomery, R. Carbohydr. Res. 1996, 280, 27. (17) Moliner, M.; Roman-Leshkov, Y.; Davis, M. E. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 6164. (18) Roman-Leshkov, Y.; Moliner, M.; Labinger, J. A.; Davis, M. E. Angew. Chem., Int. Ed. 2010, 49, 8954. (19) Corma, A.; Garcia, H. Chem. Rev. 2002, 102, 3837. (20) Corma, A.; Nemeth, L. T.; Renz, M.; Valencia, S. Nature 2001, 412, 423. (21) Boronat, M.; Concepcion, P.; Corma, A.; Renz, M. Catal. Today 2007, 121, 39. (22) Boronat, M.; Concepcion, P.; Corma, A.; Renz, M.; Valencia, S. J. Catal. 2005, 234, 111. (23) Corma, A.; Domine, M. E.; Valencia, S. J. Catal. 2003, 215, 294. (24) Corma, A.; Llabres i Xamena, F. X.; Prestipino, C.; Renz, M.; Valencia, S. J. Phys. Chem. B 2009, 113, 11306. (25) Boronat, M.; Corma, A.; Renz, M. J. Phys. Chem. B 2006, 110, 21168. (26) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (27) Curtiss, L. A.; Redfern, P. C.; Raghavachari, K. J. Chem. Phys. 2007, 126, 084108. (28) Peterson, K. A. J. Chem. Phys. 2003, 119, 11099. (29) Roy, L. E.; Hay, P. J.; Martin, R. L. J. Chem. Theory Comput. 2008, 4, 1029. (30) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571. (31) Balabanov, N. B.; Peterson, K. A. J. Chem. Phys. 2005, 123, 064107. (32) Peterson, K. A.; Figgen, D.; Dolg, M.; Stoll, H. J. Chem. Phys. 2007, 126, 124101. (33) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. (34) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378. 8760
dx.doi.org/10.1021/jp204371g |J. Phys. Chem. A 2011, 115, 8754–8760
