Carbonyl Bond Cleavage by Complementary Active Sites - The

Article pubs.acs.org/JPCC

Carbonyl Bond Cleavage by Complementary Active Sites W. Hunter Woodward,† A. C. Reber,‡ Jordan C. Smith,† S. N. Khanna,*,‡ and A. W. Castleman, Jr.*,† †

Departments of Chemistry and Physics, The Pennsylvania State University, University Park, Pennsylvania 16802, United States Department of Physics, Virginia Commonwealth University, Richmond, Virginia 23284, United States

‡

S Supporting Information *

ABSTRACT: We have studied the size-selective reactivity of Aln− clusters with formaldehyde to determine if carbonyl bonds may be broken by complementary active sites. Gas phase experiments reveal that Aln−, where n = 8−12, react with formaldehyde to form Aln−2CH2−, which demonstrates that the carbonyl bond is broken in the reaction with the cluster, while Al13− is found to be resistant to reaction. The most likely leaving group is determined to be Al2O. We also found n = 15−19 to be reactive with the products being a mix of Aln−2CH2− and Aln(OCH2)m−. Theoretical investigations find that the adjacent Lewis acid and Lewis base sites stabilize the resonance structure in which the carbonyl is reduced to a single bond which encourages carbonyl cleavage. A transition state analysis of the cleavage of the carbonyl bond confirms the size selective cleavage of the carbonyl bond and supports the importance of complementary Lewis acid−Lewis base active sites in governing the reactivity.

1. INTRODUCTION The reactivity of metal clusters and nanoparticles with small molecules may be controlled through their electronic structure.1−10 The quantum confinement of electrons by the spherical potential in spherical clusters leads to the bunching of electronic levels into shells which exhibit a uniform surface distribution of charge. On the other hand, metal clusters with distorted geometries can have uneven charge distributions that may promote complementary active sites that can activate polar covalent bonds.11−16 Herein, we present experimental and theoretical findings which reveal that the strong CO carbonyl bond of formaldehyde can be split by complementary active sites on size-selective aluminum cluster anions. The adjacent Lewis acid and Lewis base sites bind with O and C to stabilize the resonance structure in which the carbonyl is reduced to a single bond which promotes carbonyl cleavage. Previous work has addressed the role of complementary sites in breaking the hydroxyl polar bond. In particular, the sizeselective reactivity of aluminum cluster anions with water has been previously identified as being due to the existence of adjacent complementary active sites where one site serves as a Lewis acid and the other site serves as a Lewis base.11−14 A Lewis acid site accepts electrons from the oxygen atom and a neighboring Lewis base site donates electrons to the hydrogen, which stabilizes the transition state for bond cleavage. The reactivity is not governed by the electronic shell closuresthe usual explanation for aluminum cluster reactivity with oxygen17−20but rather is due to the existence of these complementary active sites. Surprising control of the reactivity is possible, and specific aluminum clusters with two or more adjacent active sites have been shown to selectively release hydrogen gas, via a cooperative mechanism.12 This raises the question whether such complementary active sites may cleave stronger bonds such as carbonyls. © 2013 American Chemical Society

The carbonyl bond serves as a reaction center in countless organic reactions, making it one of the most important bonds in all of chemistry.21 Carbonyl compounds also play a role in the decomposition of alcohols on metal surfaces22,23 and are important intermediates in various reactions such as the reduction of CO2 to methane.24 The carbonyl’s importance motivates our effort to identify clusters and surfaces that possess well-patterned Lewis acid/Lewis base sites optimal to attacking carbonyls25 and extend their use to metal oxides, bimetallic interfaces, defects, or cluster-assembled materials.25−30 The dissociation of the carbonyl group in formaldehyde is particularly difficult to achieve with a bond dissociation energy of 751.5 kJ/mol versus only 498.8 kJ/mol for the O−H bond in water, which suggests that the CO bond will be harder to cleave using the same mechanism.31 On the other hand, formaldehyde has a resonance hybrid structure, with positive and negative charges on the C and O atoms, respectively. We hypothesize that the canonical C−O singlebond structure may be stabilized by the complementary active sites and subsequently promote the weakening of the CO double bond, thereby allowing the splitting of such a strong interaction, as proposed in some theoretical studies involving surfaces.32,33 In the present study, reactions between formaldehyde and anionic aluminum clusters were observed and investigated under multicollisional conditions in a fast flow reactor.11,12,34,35 Experiments find that Aln−, where n = 8−12 and n = 15−19, react with formaldehyde. We observed the formation of Aln−2CH2− peaks, which are explained by the loss of Al2O after the cleavage of the carbonyl bond of formaldehyde, Received: April 16, 2012 Revised: March 11, 2013 Published: March 18, 2013 7445

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450

The Journal of Physical Chemistry C

Article

although at larger sizes, 12 and 15−19 Aln(OCH2)− peaks are also observed. Al13− was found to be resistant to reaction, and the reactivity of Al14− appears to be reduced as compared to other sizes but a possible secondary product makes this determination uncertain. Theoretical studies on the reactivity of formaldehyde with Aln−, n = 9, 11, 12, and 13, were performed, and accessible transition states were found for n = 9, 11, and 12, and a much higher barrier for the cleavage of the carbonyl bond on Al13− was found. Furthermore, the role of complementary active sites in the breaking of the carbonyl bond on aluminum cluster anions was confirmed by analysis of the electronic structure at the transition state.

2. EXPERIMENTAL/THEORETICAL SECTION Clusters were produced in a laser vaporization source consisting of an aluminum rod (99.999%, Puratronic) which was rotated and translated as it was ablated by a focused 532 nm Nd:YAG laser. The clusters were then carried out of the source through an expansion nozzle via a helium carrier gas maintained at ∼12 000 standard cubic centimeters per second. Clusters were cooled to room temperature in a 90 cm laminar flow tube, which was maintained at ∼0.75 Torr. Formaldehyde was introduced ∼60 cm downstream from the source through a reactant gas inlet. The formaldehyde was created by heating paraformaldehyde (97%, Alfa Aesar),36 which was previously dried in a nitrogen glove box, and was then passed through a low-flow needle valve (SS-SS4-VH, Swagelok) to control reactant gas concentration. Products were sampled through a differentially pumped vacuum apparatus before being mass selected using a quadrupole mass spectrometer (Extrel CMS) and detected via a channeltron electron multiplier. The theoretical studies used a first-principles molecular orbital approach within a gradient corrected density functional framework. The molecular orbitals are expressed as a linear combination of atomic orbitals that were, in turn, formed via a linear combination of Gaussian functions located at the atomic sites. The exchange correlation contributions are included within the GGA-PBE gradient corrected density functional formalism.37 The calculations were carried out, at an all electron level, using the deMon2K.38 The DZVP basis set was used for Al, and TZVP was used for oxygen, carbon, and hydrogen. Transition states were found using a hierarchical transition state search algorithm.39 The fragment analysis was performed using the ADF code40 with a TZ2P basis set with the GGA-PBE formalism.37 A fragment analysis is used in which system studied is broken into “fragments” which serve as a basis set for the calculation of the full system. Here, we chose the aluminum cluster and reactant to serve as fragments. This enables an analysis of the change in charge density in the complex versus the charge density of the individual molecule and cluster and allows us to determine which molecular orbitals are involved in the bonding.

Figure 1. Aluminum cluster anion distribution (a) before and (b) after reaction with formaldehyde. Aluminum clusters (Aln−) are labeled with blue numbers, formaldehyde additions [Aln(OCH2)−] are labeled with red numbers, and oxygen losses [Aln(CH2)−] are labeled with green numbers. Anomalies are labeled with black numbers. Intensities are arbitrary.

The intensity of Al12− shows the largest decrease for clusters below Al13− upon exposure to formaldehyde, while Al13− reveals virtually no reactivity. Both of these observations are similar to the previously observed reactivity with water.11,12 By contrast, Al11− diminishes in intensity when exposed to formaldehyde yet is resistant to etching with water. Al14− shows some resistance to formaldehyde etching; its intensity decreases, but it maintains a large pure aluminum peak, suggesting a relatively slow reaction. Because of the tendencies of formaldehyde to polymerize on surfaces,40 it was not possible to determine the exact amount of formaldehyde introduced to the aluminum cluster distribution, so rate constants are not reported here. Previously, the products of Aln− reactions with water produced peaks corresponding to Aln(H2O)− in which the water dissociatively coupled to the cluster, while etching experiments with molecular oxygen produced smaller Aln− clusters corresponding to the loss of two neutral Al2O species. In reactivity studies with water and alcohols, n = 11, 13, and 20 were resistant to reaction, while in the oxygen etching experiments n = 13, 23, and 37 were resistant to reaction.11−13,17−20 In the present studies with formaldehyde, peaks were observed in the product spectrum corresponding to Aln(CH2)− complexes (+14 m/z) for clusters (n = 6−10, 13− 17; green labels in Figure 1b) and Aln(OCH2)− complexes (+30 m/z) for n = 12, 17, 19, and other larger clusters (red labels in Figure 1b). Methylene complexes are due to the adsorbed formaldehyde molecule losing an oxygen atom through reaction with each aluminum cluster. Al2O is the most energetically favorable leaving group of an aluminum cluster−oxygen atom complex, being >1.5 eV more exothermic than the loss of O, AlO, or Al3O.41−43 This suggests that each Aln−2(CH2)− product peak observed corresponds to the loss of a single Al2O from an Aln+2(OCH2)− complex as shown in eq 1.

3. RESULTS AND DISCUSSION Initial and final spectra are shown in Figure 1. Although a convenient reference, this figure is deceptive as peak heights are easily discernible while integrated peak areas are not. It is important to use the latter rather than the former due to occasional signal spikes and peak broadening at higher masses, which is common and difficult to avoid when using a quadrupole mass spectrometer. Therefore, integrated peak areas are shown in Figure 2.

Al n− + H 2CO → Al n − 2(CH 2)− + Al 2O 7446

(1)

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450

The Journal of Physical Chemistry C

Article

from Al2O loss in the Aln− clusters where n = 17−19 so the carbonyl is being cleaved in a significant percentage of these reactions. In summary, the experimental results presented here suggest that the carbonyl bond in formaldehyde is being size selectively cleaved by the aluminum cluster anions. To support the proposed mechanism, we have studied the reactions with first principles density functional theory.37−39 We first examined the reaction pathway of Al9− with formaldehyde (Figure 3A). The charge density plotted in red

Figure 2. (a) Ratio of summed integrated product intensities (Aln−, Aln−2(CH2)−, Aln(OCH2)−, and Aln(OCH2)2−) versus initial integrated Aln− product intensities. Deviations from unity are due to the added baseline noise from each of the three additional product peaks. This error is exacerbated at higher masses, as is common with quadrupole mass spectrometers. (b) Percent additions for each of the products (e.g., Aln−2(CH2)−) divided by summed integrated product intensities from (a) for each n. Additions of CH2, OCH2, and (OCH2)2 are represented by green, red, and black, respectively. It can be seen how the smaller aluminum cluster anions (7 < n < 14) and n = 18, 19 favor OCH2 addition and therefore Al2O loss. It can also be seen how the larger clusters (n > 13) begin to show OCH2 activation, while Al19− shows the strongest (OCH2)2 addition.

The largest methylene peak is Al10(CH2)− and the most diminished pure aluminum peak is Al12−, confirming the loss of Al2O. Also, the absence of an Al11CH2− peak suggests the inactivity of Al13−. This is similar to the O−H bond cleavage of past studies that are explained by complementary active sites.11−14 The presence of AlnCH2− peaks of n = 6−10 and 13−17 tells us that Aln− with n = 8−12 and n = 15−19 are reacting with formaldehyde and cleaving the carbonyl bond. Additionally, the nearly complete absence of Al12(CH2)− suggests that Al14− has reduced reactivity, although secondary products of Al13− with the loss of [AlOCH2]due to the stability of Al13−may be a possible explanation for the weak Al12CH2− signal. Theoretical studies find that the loss of Al2O from Al14− and formaldehyde is exothermic by only 0.43 eV, as compared to 0.98−1.35 eV for n = 9−11, reactions which predominantly produce Al2O loss. Second, the loss of AlOCH2 is found to be endothermic by 0.26 eV, which makes it a less stable but plausible product. Aln− (n ≤ 6) clusters did not react, which was also observed in the reactivity of these clusters with water and methanol and is likely due to the distribution of the excess charge protecting the smaller clusters from nucleophilic attack.11−13 Formaldehyde additions without oxygen loss at increasingly larger cluster sizes are most likely due to the larger clusters containing a greater number of vibrational degrees of freedom over which the energy released by the aluminum−oxygen bond formation may be redistributed. We observe peaks resulting

Figure 3. Theoretically determined reaction coordinate diagrams of Aln− + OCH2 for (A) n = 9, (B) n = 12, and (C) n = 13. For each initial structure, the HOMO and LUMO (or LUMO+1) are shown in red and blue, respectively. The results reveal that Al9− and Al12− will react readily at the complementary active sites and subsequently lose an Al2O, while Al13− does not have active sites and has both a barrier to carbonyl cleavage and Al2O release is endothermic.

and blue on the Al9− cluster indicates the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) of the clusters which serve as Lewis base and Lewis acid active sites, respectively. The energies of reaction pathways studied are given in Table S1. Formaldehyde binds most favorably when the O atom binds to the Lewis acid site, and the C atom binds to the Lewis base site. Once the oxygen has donated charge to the Lewis acid site, the single bond resonance structure is favored, leading to a more weakly bonded C−O intermediate. The large binding energy between the formaldehyde molecule and the aluminum cluster (1.76 eV) indicates that covalent bonding has occurred. The C−O bond has stretched from 1.22 to1.46 Å (which compares well with the 1.44 Å bond length of methanol), and the O−C−H bond angle changes from 121.8° in the formaldehyde molecule to 7447

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450

The Journal of Physical Chemistry C

Article

green area of overlap which confirms the role of the Lewis acid site. An analysis of the symmetrized fragment orbitals confirms that the LUMO of the Al9− fragment has an occupation of 0.47. The purple isosurface indicates the vacated charge and shows that charge has been removed from a π-like orbital on formaldehyde, confirming that the hybridization of carbon has shifted from sp2 to sp3. Second, the overlap between the red Lewis Base site and the yellow bond result in an orange area of overlap, confirming the role of the Lewis base site. In Figure 4b, a similar analysis is performed at the transition state, and while the C−O bond distance has stretched significantly, the green overlap area between the yellow Al−O bond and the orange overlap between the Lewis base site and the Al−C bond are even more pronounced. Following the breaking of the C−O bond in Figure 4c, the O and CH2 are incorporated more fully into the cluster, and the bonding is now more complex as each moiety is bound to adjacent faces of the cluster. The fragment analysis finds that the complementary active sites remain important in stabilizing the Al−O and Al−C bonds at the transition state. Interestingly, active sites on Al11− which cleave formaldehyde do not cleave water, making formaldehyde more sensitive than water to the strength of the Al11− active sites. Experimentally this cluster reacts with formaldehyde despite being resistant to water etching, and theory suggests two energetically favorable pathways for Al2O release (Figure 5). In the first pathway, the O atom binds in a bridging site adjacent with the Lewis acid site of the Al11− cluster, and the C interacts with the HOMO in a manner similar to Al9− and Al12−. There is enough energy to cleave the C−O bond and then release an Al2O molecule, forming the observed Al9CH2− peak. The second pathway, in

109.5° after binding to the cluster, all of which confirms that the C−O single bond is present and that the C atom has moved from sp2 to sp3 hybridization. The formaldehyde molecule bonds in a manner similar to an η2-H2CO conformation with a bulk surface.32 The barrier to cleave the C−O bond is 1.22 eV (117.7 kJ/mol); however, there is sufficient energy from the binding to affect this transformation. There is some distortion of the cluster at the transition state, although the large binding energy provides the cluster with enough energy for reconstruction. Finally, once the carbonyl bond is cleaved, 3.04 eV of energy is released, which is enough to eject an Al2O molecule from the surface of the cluster, consistent with the proposed origin for the observed Al7CH2− peak. Additionally, we examined the reaction pathway between Al12− and formaldehyde, with similar results (Figure 3B). The barrier for splitting the carbonyl is even lower than the barrier on Al9−, which is consistent with Al12− reacting more readily, and the cleavage of the C−O bond releases 3.03 eV of energy, which is sufficient to release an Al2O molecule. In the case of Al13−, the activation energy required to break the CO bond designated by the peak in the reaction coordinate diagram (Figure 3C) is greater than the binding energy gained from the initial interaction. Thus, the reaction does not proceed, agreeing with our experimental results. Al13− is an icosahedral cluster, with a closed electronic shell, whose electron density is uniformly distributed on all of the surface atoms, which means that no atom serves as a superior Lewis acid or base, so the cluster is uniquely deficient of active sites. The absence of active sites on Al13−, the presence of such active sites on the other clusters discussed herein, and the lack of observable Al13− reactivity suggest that active sites are necessary to promote carbonyl cleavage of formaldehyde on aluminum cluster anions. To support the importance of the complementary active sites in the breaking of the carbonyl bond, we have performed a fragment analysis at selected points along the reaction pathway. Figure 4 shows a fragment analysis in which the charge density of the isolated fragments, Al9− and OCH2, are subtracted from the total charge density of the complex. Figure 4a shows the analysis and frontier orbitals at the local minimum for binding of formaldehyde. Yellow indicates increased charge density and bond formation, and the Lewis acid active site is marked by blue. The overlap between the bond and LUMO results in a

Figure 5. Theoretically determined reaction pathways for Al11− + OCH2. The LUMO and LUMO+1 are shown in blue and light blue, respectively, while the HOMO and HOMO−1 are shown in red and yellow, respectively. In the upper diagram, the oxygen attacks the LUMO and the carbon bonds to the HOMO, while in the lower pathway the O binds to a second site with significant LUMO density, and the C binds to the HOMO−1. In both cases, there is sufficient energy available for the reaction to proceed and an Al2O is lost.

Figure 4. A plot of the difference in charge density between the OCH2−Al9− complex and the fragments (Al9− and OCH2) with yellow indicating increased charge density and purple indicating decreased charge density. The HOMO and LUMO of the of the Al9− fragment are plotted in red and blue, respectively. (a) Initial binding of OCH2, (b) the transition state for breaking the carbonyl bond, and (c) the final state with the carbonyl bond broken. 7448

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450

The Journal of Physical Chemistry C

Article

that there is no hidden reactivity wherein the C or CO favorably leaves the cluster once the bond is broken. Additionally, since the nucleophilic attack proposed here involves the donation of electrons to an anion, we argue that the lack of a charge-dipole interaction due to the nonpolar C− O bonds of CO2 is a negligible factor. This reactivity is interesting, as it suggests that formaldehyde and acetone must be more reactive because the methylene and 2-propylene, respectively, allow a greater tendency to the canonical resonance structure of fully charged C and O atoms. In the case of carbon dioxide, the highly electronegative oxygen atom is likely to resist the localization of electrons on the neighboring carbon atom, preventing the weakening of the C−O bond. In the case of carbon monoxide, a canonical double-bonded resonance structure is feasible; however, even with such a system there still exists a double bond that is likely too strong to break.

which the carbonyl is split along an edge site, differs in that the O binds to a site with LUMO density serving as a Lewis acid site, and the C binds to a site with HOMO−1 charge density serving as a Lewis base. The barrier of this second pathway is only 0.04 eV higher in energy than the first, and both are energetically plausible. In comparison, the transition state for CO cleavage on Al12− is 0.4 eV higher in energy at the edge site of Al12− than the pathway shown in Figure 3B, so not all clusters have multiple sites which react with formaldehyde. To test the limitations of this activity, experiments were carried out in which aluminum cluster anions were reacted with three carbonyl-containing species of differing bond strength: acetone (771.4 kJ/mol), carbon dioxide (532.2 kJ/mol), and carbon monoxide (1076.4 kJ/mol).21 In these experiments, similar reactivity to formaldehyde was observed with acetone, while carbon dioxide and carbon monoxide showed no reactivity, even though carbon dioxide has a lower bond dissociation enthalpy than formaldehyde and acetone (Figure 6). The absence of any noticeable change in the intensity of the cluster distribution for the reactivity with CO and CO2 suggests

4. CONCLUSIONS We have demonstrated that the reactivity of aluminum cluster anions with formaldehyde is size-selective and entails complementary active sites that can cleave the strong CO bond of formaldehyde. We suggest that the primary products are the Aln−2CH2− species after an Al2O molecule is released from the cluster. It was discovered that Al11− was reactive with formaldehyde but not water, which is consistent with the strong binding due to the cluster stabilizing the singly bonded C−O resonance structure. Further exploration of gas-phase metal cluster chemistry is necessary to screen additional metals and combinations to fully understand the usefulness of these active sites for their bond activation properties.

■

ASSOCIATED CONTENT

S Supporting Information *

Calculated binding energies, transition state energies for carbonyl cleavage, and dissociative binding energy. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (A.W.C.); [email protected] (S.N.K.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Air Force Office of Science Research under AFOSR Awards FA9550-10-1-0071 and FA9550-09-1-0371 (PSU and VCU, respectively).

■

REFERENCES

(1) Castleman, A. W., Jr.; Bowen, K. H., Jr. Clusters: Structure, Energetic, and Dynamics of Intermediate States of Matter. J. Phys. Chem. 1996, 100, 12911−12944. (2) Castleman, A. W., Jr.; Khanna, S. N. Clusters, Superatoms, and Building Blocks of New Materials. J. Phys. Chem. C 2009, 113, 2664− 2675. (3) Claridge, S. A.; Castleman, A. W., Jr.; Khanna, S. N.; Murray, C. B.; Sen, A.; Weiss, P. S. Cluster-Assembled Materials. ACS Nano 2009, 3, 244−255. (4) Bohme, D. K.; Schwarz, H. Gas-Phase Catalysis by Atomic and Cluster Metal Ions: The Ultimate Single-Site Catalysts. Angew. Chem., Int. Ed. 2005, 44, 2336−2354.

Figure 6. Aluminum cluster anion distribution after reaction with (a) acetone, (b) carbon dioxide, and (c) carbon monoxide. Initial distributions were all similar to that shown in Figure 1a. Acetone additions [Aln(OC(CH3)2)−] are labeled with red numbers, and oxygen losses [AlnC(CH3)2)−] are labeled with green numbers. Note that several of the larger peaks are cut off. Additional reactivity with acetone is present due to the methyl leaving group. This reactivity is not pertinent here and will be discussed in a future publication. In (b) and (c), no reactivity is observed; two contaminant peaks in (c) can be attributed to (†) Fe(CO)4− and (‡) Fe(CO)5−, which are two common contaminants in bottled carbon monoxide. 7449

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450

The Journal of Physical Chemistry C

Article

Transition States, and Electron Detachment Energies. J. Chem. Phys. 2005, 122, 214309. (27) Bonačić-Koutecký, V.; Burda, J.; Mitrić, R.; Ge, M. F.; Zampella, G.; Fantucci, P. Density Functional Study of Structural and Electronic Properties of Bimetallic Silver-Gold Clusters: Comparison with Pure Gold and Silver Clusters. J. Chem. Phys. 2002, 117, 3120−3131. (28) Kaden, W. E.; Wu, T.; Kunkel, W. A.; Anderson, S. L. Electronic Structure Controls Reactivity of Size-Selected Pd Clusters Adsorbed on TiO2 Surfaces. Science 2009, 326, 826−829. (29) Supported Metals in Catalysis; Anderson, J. A., García, M. F., Eds.; Catalytic Science Series Vol. 5; Imperial College Press: London, 2005. (30) Surface Chemistry and Catalysis; Carley, A. F., Davies, P. R., Hutchings, G. J., Spencer, M. S., Eds.; Fundamental and Applied Catalysis; Kluwer Adademic/Plenum Publishers: New York, 2002. (31) Bond dissociation energies calculated using ΔfH° values from: Chase, M. W., Jr. NIST-JANAF Thermochemical Tables, 4th ed.; J. Phys. Chem. Ref. Data Monogr. 1998, 9, 1−1951. (32) Delbecq, F.; Sautet, P. Low-Temperature Adsorption of Formaldehyde on a Pt(111) Surface: A Theoretical Study. Langmuir 1993, 9, 197−207. (33) Gomes, J. R. B.; Gomes, J. A. N. F.; Illas, F. First-Principles Study of the Adsorption of Formaldehyde on the Clean and Atomic Oxygen Covered Cu(111) Surface. J. Mol. Catal. A: Chem. 2001, 170, 187−193. (34) Castleman, A. W., Jr.; Weil, K. G.; Sigsworth, S. W.; Leuchtner, R. E.; Keesee, R. G., Jr. Considerations of the Rates and Lifetimes of Intermediate Complexes for the Association of Various Ligands to Metal-Ions: Ag+ And Cu+. J. Chem. Phys. 1987, 86, 3829−3835. (35) Guo, B. C.; Wei, S.; Chen, Z.; Kerns, K. P.; Purnell, J.; Buzza, S.; Castleman, A. W., Jr. Generation of Metal-Carbon and Metal-Nitrogen Clusters with a Laser-Induced Plasma Technique. J. Chem. Phys. 1992, 97, 5243−5245. (36) Helander, K. G. Formaldehyde Prepared from Paraformaldehyde is Stable. Biotech. Histochem. 2000, 75, 19−22. (37) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (38) Köster, A. M.; Calaminici, P.; Casida, M. E.; Dominguez, V. D.; Flores-Moreno, R.; Geudtner, G.; Goursot, A.; Heine, T.; Ipatov, A.; Janetzko, F.; et al. deMon2k, Ver. 2.3.6, The deMon Developers, Cinvestav, Mexico, 2010. Available at http://www.deMon-software. com. (39) del Campo, J. M.; Köster, A. M. A Heirarchical Transition State Search Algorithm. J. Chem. Phys. 2008, 129, 024107. (40) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. Chemistry with ADF. J. Comput. Chem. 2001, 22, 931−967. (41) Ruatta, S. A.; Anderson, S. L. Reaction of Aluminum Clsuter Ions with Oxygen and Nitrous-Oxide: Energetics and Dynamics of Cluster Oxidation. J. Chem. Phys. 1988, 89, 273−286. (42) Cooper, B. T.; Parent, D.; Buckner, S. Oxidation Reactions and Photochemistry of Aluminum Cluster Anions (Al3− to Al23−). Chem. Phys. Lett. 1998, 284, 401−406. (43) Burgert, R.; Schnockel, H.; Grubisic, A.; Li, X.; Stokes, S. T.; Bowen, K. H.; Gantefor, G. F.; Kiran, B.; Jena, P. Spin Conservation Accounts for Aluminum Cluster Reactivity Pattern with O2. Science 2008, 319, 438−442.

(5) Shimojo, F.; Ohmura, S.; Kalia, R. K.; Nakano, A.; Vashishta, P. Molecular Dynamics Simulations of Rapid Hydrogen Production from water Using Aluminum Clusters as Catalyzers. Phys. Rev. Lett. 2010, 104, 126102. (6) Parent, D. C.; Anderson, S. L. Chemistry of Metal and Semimetal Cluster Ions. Chem. Rev. 1992, 92, 1541−1565. (7) Romero, C. P.; Avila, J. I.; Trabol, R. A.; Wang, H.; Vantomme, A.; Van Bael, M. J.; Lievens, P.; Cabrera, A. L. Pd as a Promoter to Reduce Co Cluster Films at Room Temperature. Int. J. Hydrogen Energy 2010, 35, 2262−2267. (8) Mills, G.; Gordon, M. S.; Metiu, H. Oxygen Adsorption on Au Clusters and a Rough Au(111) Surface: The Role of Surface Flatness, Electron Confinement, Excess Electrons, and Band Gap. J. Chem. Phys. 2003, 118, 4198−4205. (9) Lei, Y.; Mehmood, F.; Lee, S.; Greeley, J.; Lee, B.; Seifert, S.; Winans, R. E.; Elam, J. W.; Meyer, R. J.; Redfern, P. C.; et al. Increased Silver Activity for Direct Propylene Epoxidation via Subnanometer Size Effects. Science 2010, 328, 224−228. (10) Das, U.; Raghavachari, K.; Jarrold, C. C. Addition of Water to Al5O4− Determined by Anion Photoelectron Spectroscopy and Quantum Chemical Calculations. J. Chem. Phys. 2005, 122, 014313. (11) Reber, A. C.; Khanna, S. N.; Roach, P. J.; Woodward, W. H.; Castleman, A. W., Jr. Reactivity of Aluminum Cluster Anions with Water: Origins of Reactivity and Mechanisms for H2 Release. J. Phys. Chem. A 2010, 114, 6071−6081. (12) Roach, P. J.; Woodward, W. H.; Castleman, A. W., Jr.; Reber, A. C.; Khanna, S. N. Complementary Active Sites Cause Size-Selective Reactivity of Aluminum Cluster Anions with Water. Science 2009, 323, 492−495. (13) Reber, A. C.; Roach, P. J.; Woodward, W. H.; Khanna, S. N.; Castleman, A. W., Jr. Edge-Induced Active Sites Enhance the Reactivity of Large Aluminum Cluster Anions with Alcohols. J. Phys. Chem. A 2012, 116, 8085−8091. (14) Abreu, M. B.; Powell, C.; Reber, A. C.; Khanna. Ligand-Induced Active Sites: Reactivity of Iodine-Protected Aluminum Superatoms with Methanol. J. Am. Chem. Soc. 2012, 134, 20507−20512. (15) Zemski, K. A.; Justes, D. R.; Castleman, A. W., Jr. Studies of Metal Oxide Clusters: Elucidating Reactive Sites Responsible for the Activity of Transition Metal Oxide Catalysts. J. Phys. Chem. B 2002, 106, 6136−6148. (16) Johnson, G. E.; Mitrić, R.; Bonačić-Koutecký, V.; Castleman, A. W., Jr. Clusters as Model Systems for Investigating Nanoscale Oxidation Catalysis. Chem. Phys. Lett. 2009, 475, 1−9. (17) Reber, A. C.; Khanna, S. N.; Roach, P. J.; Woodward, W. H.; Castleman, A. W., Jr. Spin Accommodation and Reactivity of Aluminum Based Clusters with O2. J. Am. Chem. Soc. 2007, 129, 16098−16101. (18) Roach, P. J.; Woodward, W. H.; Reber, A. C.; Khanna, S. N.; Castleman, A. W., Jr. Crystal Field Effects on the Reactivity of Aluminum-Copper Cluster Anions. Phys. Rev. B 2010, 81, 195404. (19) Leuchtner, R. E.; Harms, A. C.; Castleman, A. W., Jr. Thermal Metal Cluster Anion Reactions: Behavior of Aluminum Clusters with Oxygen. J. Chem. Phys. 1989, 91, 2753−2754. (20) Leuchtner, R. E.; Harms, A. C.; Castleman, A. W., Jr. Aluminum Cluster Reactions. J. Chem. Phys. 1991, 94, 1093−1101. (21) Solomons, T. W. G.; Fryhle, C. B. Organic Chemistry, 9th ed.; Wiley: Hoboken, NJ, 2008. (22) Somorjai, G. A.; Li, Y. Introduction to Surface Chemistry and Catalysis; Wiley: Hoboken, NJ, 2010. (23) Gates, B. C. Supported Metal-Clusters: Synthesis, Structure, and Catalysis. Chem. Rev. 1995, 95, 511−522. (24) Peterson, A. A.; Norskov, J. K. Activity Descriptors for CO2 Electroreduction to Methane on Transition-Metal Catalysis. J. Phys. Chem. Lett. 2012, 3, 251−258. (25) Iglesia, E.; Barton, D. G.; Biscardi, J. A.; Gines, M. J. L.; Soled, S. L. Bifunctional Pathways in Catalysis by Solid Acids and Bases. Catal. Today 1997, 38, 339−360. (26) Guevara-García, A.; Martínez, A.; Ortiz, J. V. Addition of Water, Methanol, and Ammonia to Al3O3− Clusters: Reaction Products, 7450

dx.doi.org/10.1021/jp303668b | J. Phys. Chem. C 2013, 117, 7445−7450