J. Phys. Chem. B 2006, 110, 10461-10466
10461
Mechanism and Regioselectivity for the Reactions of Propylene Oxide with X(100)-2×1 Surfaces (X ) C, Si, Ge): A Density Functional Cluster Model Investigation Zheng Guo and Xin Lu* State Key Laboratory of Physical Chemistry of Solid Surface & Center for Theoretical Chemistry, Department of Chemistry, College of Chemistry and Chemical Engineering, Xiamen UniVersity, Xiamen 361005, People’s Republic of China ReceiVed: February 7, 2006; In Final Form: March 17, 2006
We have performed density functional cluster model calculations to explore the mechanism and regioselectivity for the reactions of propylene oxide with X(100)-2×1 surfaces (X ) C, Si, and Ge). The computations reveal the following: (i) the reactions on Si(100) and Ge(100) are barrierless and highly exothermic; (ii) the reactions on X(100) (X ) Si and Ge) are initiated by the formation of a dative-bonded precursor state followed by regioselective cleavage of the C2-O bond (C2 directly connected to the methyl-substituent) in propylene oxide, giving rise to a five-membered ring surface species; and (iii) the reaction on C(100), although highly exothermic, requires a large activation energy and would be kinetically forbidden at room temperature.
1. Introduction Enhanced attention has recently been paid to the organic functionalization of semiconductor surfaces because of potential applications in new and emerging technologies such as molecular electronics,1 nanotechnology,2 and biotechnology sensors.3 To obtain the fundamental aspects of the related chemistry, a large number of experimental and theoretical investigations have been performed concerning the attachments of various organic compounds onto the well-defined single-crystalline semiconductor surfaces such as the reconstructed X(100)-2×1 (X ) C, Si, Ge) and Si(111)-7×7 surfaces.4 Thermally reconstructed X(100)-2×1 (X ) C, Si, Ge) surfaces have ordered arrays of dimeric >XdX< unit, which has a σ-bond and a weak π-bond between the two X atoms, similar to the CdC double bond in simple alkenes.4 The π-bond strength within the dimeric XdX unit was estimated to be ∼28 kcal/mol for C(100)5a and ∼5-10 kcal/mol for Si(100)5b and Ge(100),5c much smaller than that of ethylene (56 kcal/mol).5d As a result, these surface dimers also display substantial diradical character, and more specifically zwitterionic character on Si(100) and Ge(100) surfaces.6 With such fascinating bonding characters, the surface dimers appear to be highly reactive toward various unsaturated organic compounds, for example, readily undergoing [4+2] cycloaddition (Diels-Alder reaction) with conjugated dienes and related compounds,7,8 1,3-dipolar cycloaddition,9,10 and [2+2] cycloaddition with alkenes11 and alkynes.12 In addition, due to their zwitterionic nature, the surface dimers on Si(100) and Ge(100) are subject to nucleophilic or electrophilic reactions with such molecules as OH-,13 NH-,14,15 and BHcontaining16 molecules. For several years, our laboratory has been involved in elucidating theoretically the reaction mechanisms for the attachments of various organic compounds onto semiconductor surfaces, including alkenes11c-e and alkynes,12d,8c aromatic hydrocarbons and allied compounds (e.g., thiophene and furan),8,17 acetonitrile,17a methanol, formaldehyde, formic acid,18 and 1,3-dipolar compounds,10 etc. In this manuscript, * Corresponding author. Fax: +86-592-2183047. E-mail: xinlu@ xmu.edu.cn.
SCHEME 1: Reaction of Organic Epoxide with a SidSi Dimer on Si(100) and Possible Products
we report the results of a density functional cluster model investigation on the attachment of a model epoxide, propylene oxide (PO), to the X(100)-2×1 surfaces (X ) C, Si, Ge). An organic epoxide has a three-membered C2O ring that is subject to either nucleophilic or electrophilic reactions. Recently, Linford et al.19 investigated experimentally the reactions of organic epoxides with Si(100)-2×1 surface and found that the surface reaction on a surface dimer produces a five-membered ring surface species by cleaving a C-O bond of the epoxide (Scheme 1), giving rise to a uniform organic adlayer on the surface. Obviously, the regioselectivity of this surface reaction is a key factor that controls the uniformity of the thus-formed organic monolayer. Unfortunately, no clear information is available regarding the regioselectivity of this surface reaction, that is, whether the final surface species is 3, 3′, or even a mixture of both (Scheme 1). This invoked our interest to explore the mechanism and regioselectivity of this newly disclosed surface reaction by means of density functional cluster model calculations. For the sake of simplicity, propylene oxide is taken as an example of organic epoxides. In addition, we have also investigated the reactions of propylene oxide with the Ge(100)2×1 and C(100)-2×1 surfaces to see if analogous reactions can be employed to chemically functionalize these surfaces. 2. Computational Details All calculations were performed with the Gaussian 98 package.20 The hybrid density functional method, including Beck’s three parameters nonlocal-exchange functional21 with the correlation functional of Lee-Yang-Parr (B3LYP),22 was employed with the standard all-electron split-valence basis set
10.1021/jp0607972 CCC: $33.50 © 2006 American Chemical Society Published on Web 05/05/2006
10462 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Figure 1. B3LYP/6-31G*-optimized geometries of single-dimer cluster models (key bond length in angstroms) to represent the surface XdX dimer of X(100)-2×1 surfaces (X ) C, Si, Ge): (a) C9H12, (b) Si9H12, and (c) Ge2Si7H12.
6-31G*. For stationary points that may be singlet diradicals, the spin-unrestricted UB3LYP method was used. Geometry optimizations were performed with no constrained degrees of freedom. The computed transition states are verified by harmonic vibrational analyses to have only one imaginary frequency and have been further confirmed by intrinsic reaction coordinate (IRC) calculations23 at the B3LYP/6-31G* level. For each stationary point, the computed frequencies are used to derive its zero-point energies (ZPE) without scaling. The energies reported have been corrected by inclusion of zero-point energies (ZPE), unless otherwise specified explicitly. Spin-projected correction24 has been employed to derive the energies of the stationary points that may be singlet diradicals. The XdX dimeric site on the X(100)-2×1 surface (X ) C, Si) is modeled by the X9H12 single-dimer cluster (Figure 1a and b). For the Ge(100)-2×1 surface, the cluster model Ge2Si7H12 is employed (Figure 1c), which is derived from the Si single-dimer cluster model by substituting the two topmost Si atoms with Ge atoms. This simplified model was previously employed by Wang et al. to investigate some surface reactions on the Ge(100) surface with reduced computational cost.25 The B3LYP-optimized structures of these clusters are given in Figure 1. Note that the topmost CdC dimer in the C9H12 cluster is symmetric, whereas the SidSi and GedGe dimers are asymmetric and buckled, consistent with the experimental observations.4b,26 Despite the simplicity that they neglect interdimer
Guo and Lu interactions, these cluster models have been proven to give reasonable energetic and geometric predictions for the chemisorptions and reactions on the X(100)-2×1 (X ) C, Si, Ge) surfaces.25,27 Our B3LYP/6-31G* calculation predicts a Ge-Ge bond length of ∼2.38 Å for the Ge2Si7H12 cluster model, slightly shorter than the previous B3LYP/6-311+G** prediction (2.41 Å)27b for the same Ge2Si7H12 cluster and the previous B3LYP/ 6-31+G* prediction (2.39 Å)10d for a Ge9H12 cluster model. It appears that the smaller basis set tends to underestimate the bond length of the topmost GedGe dimer. Furthermore, to explore the basis set effect on the reaction energetics, we also have performed single-point B3LYP/6-311+G** calculations for the reactions on the Si and Ge surfaces. 3. Results and Discussion 3.1. Propylene Oxide/Si(100). On the Si(100) surface, the surface dimers are known to be zwitterionic, in which the buckled-down and buckled-up Si atoms are positively and negatively charged, respectively. It is likely that PO can undergo nucleophilic attack with the lone pair of its oxygen atom to the electron-deficient buckled-down Si atom, forming a POfSi dative bond. Figure 2 depicts the computed dative-bonded states (LMa-LMe) of PO/Si9H12. The formations of these dativebonded states are essentially barrierless. The strength of the asformed dative bond ranges within 12.3-17.2 kcal/mol at the B3LYP/6-31G* level of theory. The optimal lengths for the PO-Si dative bonds (∼1.97 Å) are much longer than that of a typical Si-O covalent bond (∼1.7 Å). Upon coordination of PO, the dimeric bond (Si-Si) is elongated by ∼0.13 Å (from 2.22 to ∼2.35 Å), indicating the breakage of its weak π-bond. Notably, the two C-O bonds in PO are also elongated with the C2-O bond becoming longer than the C1-O bond. Such a phenomenon suggests the C2-O bond is more severely activated than the C1-O bond upon datively bonding. These dativebonded states should be precursors for further reaction of PO with the surface dimer.
Figure 2. B3LYP/6-31G*-computed geometries (key bond length in angstroms, and bond angle in degrees) and energies (relative to isolated reactants) for the dative-bonded states of propylene oxide on Si9H12. The optimized geometry of propylene oxide is also given.
Propylene Oxide on X(100) Surfaces
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10463
Figure 4. UB3LYP/6-31G*-computed geometries (key bond length in angstroms, and bond angle in degrees) for open-shell singlet transition states leading to C-O bond cleavage of PO on Si9H12.
TABLE 1: UB3LYP/6-31G*-Predicted Energies (RE in kcal/mol, Relative to Isolated Reactants), Net Spin Densities, and the Values of 〈S2〉 for the Open-Shell Singlet Transition States of C-O Bond Breakage of Propylene Oxide Adsorbed on the Si9H12 Cluster Model net spin TS1-OS TS2-OS a
Figure 3. B3LYP/6-31G*-optimized geometries (key bond length in angstroms, and bond angle in degrees) and energies (RE in kcal/mol, relative to isolated reactants) for transition states and products of PO reaction with the Si9H12 cluster model starting from the precursor LMa.
Among all of the dative-bonded states, LMa is the most stable with the longest C-O bonds. Meanwhile, the C-O bonds in this dative-bonded state are the most severely activated. Accordingly, we choose LMa as the starting point of further reaction. As shown in Figure 3, two possible pathways starting from LMa have been disclosed. TS1 and TS2 are the transition states leading to breakages of the C1-O and C2-O bonds in PO, respectively. They are higher by 11.6 and 9.1 kcal/mol in energy than the precursor LMa, respectively. The formation energies of the products LM1 and LM2 are quite large (>80.0 kcal/mol with respect to isolated propylene oxide and Si9H12), showing the surface reaction is highly exothermic. Both products have a five-membered ring surface species, differing in the position of the methyl-substituent. More interestingly, the transition states TS1 and TS2 are lower by 5.6 and 8.1 kcal/ mol in energy than the isolated state, respectively. The overall surface reaction should be barrierless by following either pathway. Because TS2 is by 2.5 kcal/mol lower than TS1 in energy, cleavage of the C2-O bond is kinetically favorable over breakage of the C1-O bond, leading dominantly to the product LM2. It should be noted that in transition states TS1 and TS2, the cleaving C-O bond length is around 1.93 Å, whereas the forming C-Si2 bond length is too large (∼3.61 Å). According to the Hammond postulate,28 they are structurally reactant-like and, thus, belong to early transition states, in accordance with the rather low activation barriers (with respect to LMa) and the high exothermicity of the surface reaction. However, due to the diradical character of the surface dimer, it is also possible that these transition states may be diradicals, as was found for the reactions of alkenes with the silicon surface dimer.7f,11c To
REa
〈S2〉
C
Si2
-7.7 -9.3
0.16 0.09
-0.27(Χ1) -0.18(C2)
0.23 0.16
Relative energy after spin-projected correction.
show such a possibility, we have conducted UB3LYP/6-31G* calculations (with the broken-symmetry algorithm) to reoptimize the structures of TS1 and TS2. Figure 4 depicts the computed open-shell singlet transition states TS1-OS and TS2-OS that lead to C1-O and C2-O cleavages, respectively. Their relative energies, 〈S2〉 values, and net spin densities on the critical C and Si2 atoms are listed in Table 1. Despite that these openshell singlet transition states are lower by ∼2 kcal/mol in energy than the corresponding closed-shell singlet transition states, their 〈S2〉 values and the atomic spin densities on C1/C2 and Si2 atoms are small, indicating neither transition state is typical singlet diradical. We have also computed reaction pathways starting from other dative-bonded states given in Figure 2 and found that they are kinetically less favorable than the pathways depicted in Figure 3. Thus, we do not present those unfavorable reaction pathways herein. 3.2. Propylene Oxide/Ge(100). Similar to the PO/Si(100) case, several dative-bonded states can be formed barrierlessly for the attachment of PO to the positively charged buckleddown Ge1 atom of the Ge2Si7H12 cluster; the two most stable of them are depicted in Figure 5. The predicted bonding strength for the most stable state GLMa is 17.0 kcal/mol, nearly equivalent to the dative bond strength in the PO-Si9H12 (LMa). Accompanying the formation of the POfGe dative bond, the two C-O bonds in PO are substantially activated and elongated by ∼0.04 Å. By taking the most stable dative-bonded states GLMa as a precursor, two pathways have been found and depicted in Figure 6. It can be seen that the PO/Ge(100) system displays reaction pathways and regioselectivity similar to those of the PO/Si(100) reaction. That is, after formation of the dative-bonded precursor state, both C-O bonds in PO can be cleaved via the transition states GTS1/GTS2, giving rise to the final five-membered ring products GLM1/GLM2; the C2-O bond cleavage is kinetically
10464 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Guo and Lu
Figure 5. B3LYP/6-31G*-computed structures (key bond length in angstroms, and bond angle in degrees) and energies (RE in kcal/mol, relative to isolated reactants) for the two most stable dative-bonded states of PO on the Ge2Si7H12 cluster.
Figure 7. B3LYP/6-31G*-optimized geometries (key bond length in angstroms, and bond angle in degrees) and relative energies (RE in kcal/mol) for the transition states and products of PO reaction with the C9H12 cluster model.
Figure 6. B3LYP/6-31G*-optimized geometries (key bond length in angstroms, and bond angle in degrees) and energies (RE in kcal/mol, relative to isolated reactants) for transition states and products of PO reaction with the Ge2Si7H12 cluster model starting from the most stable dative-bonded state GLMa.
favorable over the cleavage of the C1-O bond in PO. However, the thermodynamics of the PO/Ge(100) system differs substantially from that of the PO/Si(100) system. Because the covalent Ge-O and Ge-C bondings are generally weaker than the Si-O and Si-C bondings, the predicted exothermicity for the PO/ Ge(100) reaction is 61.4 kcal/mol for the favored product GLM2, ∼20 kcal/mol lower than that of the PO/Si(100) reaction (81.4 kcal/mol for the favored product LM2). In addition, in the PO/Ge(100) case, the transition state GTS1 that leads to C1-O breakage is by 2.0 kcal/mol higher than the isolated
reactants, whereas the transition state GTS2 is by only -0.8 kcal/mol lower than the isolated reactants. So the PO/Ge(100) reaction would preferentially pass through the transition state GTS2 and is barrierless, too. It should be noted that the activation barrier with respect to the most stable precursor is 16.2 kcal/mol at GTS2 for the PO/Ge(100) reaction, but only 9.1 kcal/mol at TS2 for the PO/Si(100) case. As a consequence, the kinetic stability of the metastable dative-bonded precursor state is much higher in the PO/Ge(100) system than in the PO/ Si(100). That is, the dative-bonded state GLMa in the PO/ Ge(100) system should have a much longer lifetime than the one (LMa) in the PO/Si(100) system and is likely detectable at low temperatures. 3.3. Propylene Oxide/C(100). The computed pathways for the reaction of PO with the C9H12 cluster are shown in Figure 7. Different from the mechanisms predicted for the PO/Si(100) and PO/Ge(100) systems, no dative-bonded states can be formed between PO and the CdC dimer of C(100). The predicted exothermicity of the PO/C(100) reaction (>90 kcal/mol) is much higher than those of the PO/Si(100) and PO/Ge(100) reactions, showing the covalent C-O and C-C bondings are much stronger than the corresponding X-O and X-C bondings (X ) Si, Ge). Unfortunately, the PO/C(100) reaction requires a much high activation energy (29.5 kcal/mol at the favored transition state CTS2) and can hardly occur at room temperature. The much lower reactivity of the CdC site on C(100) with respect to the SidSi and GedGe sites on Si(100) and Ge(100) can be related to the fact that the CdC dimer on C(100) is not
Propylene Oxide on X(100) Surfaces
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10465
TABLE 2: Relative Energies (kcal/mol)a of the Stationary Pointsb Predicted for the Reaction of Propylene Oxide with the Si9H12 Cluster Model theory
reactants
LMa
TS1
LM1
TS2
LM2
B3LYP/6-31G* B3LYP/ 6-311++G**
0.0 0.0
-17.2 -13.5
-5.6 -3.0
-85.8 -82.8
-8.1 -5.3
-81.4 -78.2
a The energies are ZPE-corrected, whereas the ZPE for each stationary point was computed at the B3LYP/6-31G* level. b See Figure 3 for the definition and geometry of the stationary points.
TABLE 3: Relative Energies (kcal/mol)a of the Stationary Pointsb Predicted for the Reaction of Propylene Oxide with the Ge2Si7H12 Cluster Model theory B3LYP/6-31G* B3LYP/ 6-311++G**
reactants GLMa GTS1 GLM1 GTS2 GLM2 0.0 0.0
-17.0 -12.0
2.0 6.6
-65.7 -59.0
-0.8 3.9
-61.4 -54.0
a The energies are ZPE-corrected, whereas the ZPE for each stationary point was computed at the B3LYP/6-31G* level. b See Figure 6 for the definition and geometry of the stationary points.
zwitterionic at all. On the contrary, the zwitterionic character of the SidSi and GedGe sites facilitates the nucleophilic attack of the PO molecule as well as the subsequent C-O bond activation. A similar mechanism holds for the barrierless O-H bond cleavage involved in the dissociative chemisorptions of alcohols and amines on the Si(100) and Ge(100) surfaces.18,27 3.4. Basis Set Effects on the Reaction Energies. The energies for the reactions of PO on the Si and Ge surfaces have been further computed at the single points by using a larger basis set of 6-311++G**. For comparison, the relative energies computed by the B3LYP/6-311++G** and B3LYP/6-31G* calculations are listed in Tables 1 and 2 for the reactions on the Si and Ge surfaces, respectively. It is clear that when the more rigorous 6-311++G** basis set is employed, the relative energies of the stationary points (precursors, transition states, and final products) are moved upward by ∼3 kcal/mol for the PO/Si9H12 system (Table 2) and by 5-7 kcal/mol for the PO/ Ge2Si7H12 system (Table 3). Nevertheless, the regioselectivity predicted for either surface reaction is the same at both levels of theory. 4. Concluding Remarks By means of density functional cluster model calculations, we have shown that the reactions of propylene oxide with X(100)-2×1 surfaces (X ) Si and Ge) are barrierless and highly exothermic, whereas the reaction of propylene oxide with the C(100)-2×1 surface should be kinetically prohibited at room temperature. The surface reactions on X(100) (X ) Si and Ge) are initiated by the formation of a dative-bonded precursor state followed by regioselective cleavage of the C2-O bond (C2 directly connected to the methyl-substituent) in propylene oxide, leading finally to the formation of a five-membered ring surface species. The barrierless C2-O bond breakage of propylene oxide on the X(100) (X ) Si, Ge) is related to the zwitterionic character of the surface XdX dimer. Further investigation is underway to explore the effect of different substituents of organic epoxides on the regioselectivity of the surface reactions. Acknowledgment. This work was sponsored by the NSFC (Grants No. 20021002, 20425312, 20203013, 20423002, 90206038), MOE of China (Grant No. 20010384005), Fok YingTung Education Foundation, and NSF of Fujian Province
(Grants No. E0210001 and 2002F010). We thank the reviewers for their insightful comments. Supporting Information Available: Cartesian coordinates, total electronic energies, zero-point energies, and number of imaginary frequencies for stationary points predicted at the B3LYP/6-31G* level of theory. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Lopinski, G. P.; Wayner, D. D. M.; Wolkow, R. A. Nature 2000, 406, 48. (2) Hersam, M. C.; Guisinger, N. P.; Lyding, J. W. Nanotechnology 2000, 11, 70. (3) Strother, T.; Hamers, R. J.; Smith, L. M. Nucleic Acids Res. 2000, 28, 3535. (4) (a) Hamers, R. J.; Wang, Y. J. Chem. ReV. 1996, 96, 1261. (b) Hamers, R. J.; Coulter, S. K.; Ellison, M. D.; Hovis, J. S.; Padowtiz, D. F.; Schwartz, M. P.; Greenlief, C. M.; Russell, J. N. Acc. Chem. Res. 2000, 33, 617. (c) Lu, X.; Lin, M. C. Int. ReV. Phys. Chem. 2002, 21, 137. (5) (a) Hukka, T. I.; Pakkanen, T. A.; D’Evelyn, M. P. J. Phys. Chem. 1994, 98, 12420. (b) Doren, D. J. AdV. Chem. Phys. 1996, 95, 1. (c) D’Evelyn, M. P.; Cohen, S. M.; Rouchouze, E.; Yang, Y. L. J. Chem. Phys. 1993, 98, 3560. (d) McMurry, J. Organic Chemistry, 4th ed.; Brooks/Cole Publishing Co.: Pacific Grove, CA, 1996; pp 23-24. (6) Buriak, J. M. Chem. ReV. 2002, 102, 1271. (7) (a) Hovis, J. S.; Liu, H. B.; Hamers, R. J. J. Phys. Chem. B 1998, 102, 6873. (b) Lee, S. W.; Nelen, L. N.; Ihm, H.; Scoggins, T.; Greenlief, C. M. Surf. Sci. 1998, 410, L773. (c) Wang, G. T.; Bent, S. F.; Russell, J. N., Jr.; Butler, J. E.; D’Evelyn, M. P. J. Am. Chem. Soc. 2000, 122, 744. (d) Fitzgerald, D. R.; Doren, D. J. J. Am. Chem. Soc. 2000, 122, 12334. (e) Fink, A.; Menzel, D.; Widdra, W. J. Phys. Chem. B 2001, 105, 3828. (f) Choi, C. H.; Gordon, M. S. J. Am. Chem. Soc. 1999, 121, 11311. (g) Jung, Y.; Gordon, M. S. J. Am. Chem. Soc. 2005, 127, 3131. (h) Lee, H. S.; Choi, C. H.; Gordon, M. S. J. Am. Chem. Soc. 2005, 127, 8485. (8) (a) Lu, X.; Xu, X.; Wang, N.; Zhang, Q.; Lin, M. C. J. Phys. Chem. B 2001, 105, 10069. (b) Lu, X.; Lin, M. C.; Xu, X.; Wang, N.; Zhang, Q.; PhysChemPhys 2001, 13. (c) Lu, X.; Zhu, M.; Wang, X.; Zhang, Q. J. Phys. Chem. B 2004, 108, 4478. (9) (a) Barriocanal, J. A.; Doren, D. J. J. Vac. Sci. Technol., A 2000, 18, 1959. (b) Barriocanal, J. A.; Doren, D. J. J. Phys. Chem. B 2000, 104, 12269. (10) (a) Lu, X.; Fu, G.; Wang, N.; Zhang, Q.; Lin, M. C. Chem. Phys. Lett. 2001, 343, 212. (b) Lu, X.; Xu, X.; Wang, N.; Zhang, Q. J. Phys. Chem. B 2002, 106, 5972. (c) Lu, X.; Xu, X.; Wang, N.; Zhang, Q. J. Org. Chem. 2002, 67, 515. (d) Lu, X.; Su, Z.; Xu, X.; Wang, N.; Zhang, Q. Chem. Phys. Lett. 2003, 371, 172. (11) For examples of recent investigations on the [2+2] cycloadditions of alkenes, see: (a) Liu, H.; Hamers, R. J. J. Am. Chem. Soc. 1997, 119, 7593. (b) Lopinski, G. P.; Moffatt, D. J.; Wayner, D. D. M.; Wolkow, R. A. J. Am. Chem. Soc. 2000, 122, 3548. (c) Lu, X. J. Am. Chem. Soc. 2003, 125, 6384. (d) Lu, X.; Zhu, M.; Zhang, Q. J. Phys. Chem. B 2004, 108, 7359. (e) Lu, X.; Zhu, M. Chem. Phys. Lett. 2004, 393, 124. (12) For examples of recent investigations on the [2+2] cycloadditions of alkynes, see: (a) Liu, Q.; Hoffmann, R. J. Am. Chem. Soc. 1995, 117, 4082. (b) Mezhenny, S.; Lyubinetsky, I.; Choyke, W. J.; Wolkow, R. A.; Yates, J. T. Chem. Phys. Lett. 2001, 344, 7. (c) Xu, S. H.; Keeffe, M.; Yang, Y.; Chen, C.; Yu, M.; Lapeyre, G. J.; Rotenberg, E.; Denlinger, J.; Yates, J. T. Phys. ReV. Lett. 2000, 84, 939. (d) Lu, X.; Lin, M. C. Phys. Chem. Chem. Phys. 2000, 2, 4213. (13) (a) Foraker, A.; Doren, D. J. J. Phys. Chem. B 2003, 107, 8507. (b) Lu, X.; Zhang, Q.; Lin, M. C. Phys. Chem. Chem. Phys. 2001, 3, 2156. (14) (a) Nelen, L. M.; Fuller, K.; Greenlief, C. M. Appl. Surf. Sci. 1999, 150, 65. (b) Xu, X.; Kang, S. Y.; Yamabe, T. Phys. ReV. Lett. 2002, 8, 5351. (15) (a) Avoyan, A.; Tindall, C.; Hemminger, J. C. J. Phys. Chem. B 1998, 102, 5069. (b) Davies, B. M.; Craig, J. H. Surf. Sci. 2003, 529, 231. (16) (a) Konecny, R.; Doren, D. J. J. Phys. Chem. B 1997, 101, 10983. (b) Long, L.; Lu, X.; Tian, F.; Zhang, Q. J. Org. Chem. 2003, 68, 4495. (17) (a) Lu, X.; Xu, X.; Wang, N.; Zhang, Q. New J. Chem. 2002, 26, 160. (b) Lu, X.; Wang, X.; Yuan, Q.; Zhang, Q. J. Am. Chem. Soc. 2003, 125, 7923. (18) Lu, X.; Zhang, Q.; Lin, M. C. Phys. Chem. Chem. Phys. 2001, 3, 2156. (19) Lua, Y. Y.; Lee, M. V.; Fillmore, W. J. J.; Matheson, R.; Sathyapalan, A.; Asplund, M. C.; Fleming, S. A.; Linford, M. R. Angew. Chem., Int. Ed. 2003, 42, 4046. (20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
10466 J. Phys. Chem. B, Vol. 110, No. 21, 2006 D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998. (21) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (22) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (23) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154. (24) (a) Yamaguchi, K.; Jensen, F.; Dorigo, A.; Houk, K. N. Chem. Phys. Lett. 1988, 149, 537. (b) Yamanaka, S.; Kawakami, T.; Nagao, H.; Yamaguchi, K. Chem. Phys. Lett. 1994, 231, 25.
Guo and Lu (25) (a) Wang, G. T.; Mui, C.; Musgrave, C. B.; Bent, S. F. J. Am. Chem. Soc. 2002, 124, 8990. (b) Wang, G. T.; Mui, C.; Tannaci, J. F.; Filler, M. A.; Musgrave, C. B.; Bent, S. F. J. Phys. Chem. B 2003, 107, 4982. (26) Kang, H. C. Surf. Interface Anal. 1999, 28, 92. (27) (a) Fattal, E.; Radeke, M. R.; Reynolds, G.; Carter, E. A. J. Phys. Chem. B 1997, 101, 8658. (b) Mui, C.; Bent, S. F.; Musgrave, C. B. J. Phys. Chem. A 2000, 104, 2457. (c) Widjaja, Y.; Mysinger, M. M.; Musgrave, C. B. J. Phys. Chem. B 2000, 104, 2527. (d) Widjaja, Y.; Musgrave, C. B. Surf. Sci. 2000, 469, 9. (e) Loh, Z. H.; Kang, H. C. J. Chem. Phys. 2000, 112, 2444. (f) Bacalzo-Gladden, F.; Lu, X.; Lin, M. C. J. Phys. Chem. A 2001, 105, 4368. (g) Xu, X.; Kang, S. Y.; Yamabe, T. Bull. Chem. Soc. Jpn. 2001, 14, 817. (h) Kato, T.; Kang, S. Y.; Xu, X.; Yamabe, T. J. Phys. Chem. B 2001, 105, 10340. (28) Hammond, G. S. J. Am. Chem. Soc. 1954, 77, 334.
