Significant Transient Mobility of Platinum Clusters via a Hot Precursor

Nov 7, 2016 - Atsushi Beniya†, Hirohito Hirata‡, and Yoshihide Watanabe†. † Toyota Central R&D Labs., Inc., 41-1 Yokomichi, Nagakute, Aichi 48...
0 downloads 0 Views 3MB Size
Letter pubs.acs.org/JPCL

Significant Transient Mobility of Platinum Clusters via a Hot Precursor State on the Alumina Surface Atsushi Beniya,† Hirohito Hirata,‡ and Yoshihide Watanabe*,† †

Toyota Central R&D Labs., Inc., 41-1 Yokomichi, Nagakute, Aichi 480-1192, Japan Toyota Motor Corporation, 1200 Mishuku, Susono, Shizuoka 410-1193, Japan



S Supporting Information *

ABSTRACT: Relaxation dynamics of hot metal clusters on oxide surfaces play a crucial role in a variety of physical and chemical processes. However, their transient mobility has not been investigated as much as other systems such as atoms and molecules on metal surfaces due to experimental difficulties. To study the role of the transient mobility of clusters on the oxide surface, we investigated the initial adsorption process of size-selected Pt clusters on a thin Al2O3 film. Soft-landing the size-selected clusters while suppressing the thermal migration resulted in the transient migration controlling the initial adsorption states as an isolated and aggregated cluster, as revealed using scanning tunneling microscopy. We demonstrate that transient migration significantly contributes to the initial cluster adsorption process; the cross section for aggregation is seven times larger than the expected value from geometrical considerations, indicating that metal clusters are highly mobile during a energy dissipation process on the oxide surface.

E

strated,10 indicating that the energy released by the chemical reaction on the cluster surface was stored in the cluster and induced transient migration. Therefore, transient migration of clusters should be considered to elucidate catalytic reaction dynamics. It is also very helpful to evaluate the transient migration length quantitatively for fabricating nanomaterials such as cluster-monodispersed and -assembled materials by controlling the cluster size and intercluster distance.4,5,19−21 Progress toward the fundamental understanding of adsorption processes, catalytic reactions, and nanomaterial fabrications necessitates precise investigation of the transient migration of clusters. Oxide-supported metal clusters are of particular importance for catalysis including electro- and photocatalytic reactions.20−26 However, few experimental studies have reported details of the transient migration of metal clusters on oxide surfaces due to experimental limitations.27,28 First, atomic vapor deposition, employed as a cluster growth method in earlier studies,29 precludes size selection and hence does not allow direct determination of the cluster size and its transient mobility. Second, experimental study of the oxide surface is often hampered because it is difficult to use methods involving electrons and ions on almost perfect insulators.30 As elaborated below, the above shortcomings are overcome in the present study that focuses on size-selected Pt clusters soft-landed on a thin Al2O3 film formed by oxidation of a NiAl(110) substrate.

nergy dissipation processes of atoms, molecules, and clusters on the surface play an important role in various processes such as adsorption, catalytic reaction, and material fabrication.1−17 During the adsorption process, adsorbing species trapped by the surface (hot precursor) gradually dissipate their translational and adsorption energy via an electron−hole pair (EHP) and phonon excitations until they become thermalized on the surface.12 Through the energy dissipation process, the hot precursors are mobile on the surface. Such a lateral motion of hot precursors is called transient migration.1 Transient migration is also induced by surface chemical reactions.15 In dynamic catalytic reaction environments where bond-breaking and -making processes occur on the catalyst surface, the reaction energy excites the product species and induces transient migration until the reaction energy dissipates to the substrate. When the substrate consists of more than hundreds of atoms, the substrate has various phonon modes and a continuum electronic structure; hence, the reaction energy immediately distributes to various phonon modes in the substrate via EHP excitations.15,16 The situation changes significantly by decreasing the size of the catalyst to several atoms; according to gas-phase cluster studies, chemical reactions occurring on the cluster surface lead to cluster fragmentation.18 Because small atom clusters have limited numbers of vibrational modes and discrete electronic structures, a chemical reaction induces highly excited vibrational states via electronic excitations in clusters and vibrational relaxation is accomplished by cluster fragmentation. Recently, reaction-induced ripening of deposited clusters was demon© XXXX American Chemical Society

Received: October 11, 2016 Accepted: November 7, 2016 Published: November 7, 2016 4710

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715

Letter

The Journal of Physical Chemistry Letters Depositing size-selected clusters, where the cluster size was exactly controlled atom-by-atom, resolves the complexity of the cluster size distribution.6 Furthermore, because the severalatomic-layer thin Al2O3 film on the NiAl(110) substrate could avoid charge accumulation,30,31 analysis methods such as scanning tunneling microscopy (STM), which is the technique allowing for elucidation of transient migration phenomena at the atomic level, can be applied.1 Size-selected Pt15 clusters were soft-landed on the Al2O3/ NiAl(110) surface at 300 K where thermal migration of Pt15 clusters was suppressed.28Figure 1 shows STM topographic

Figure 2. Histograms of apparent cluster height as a function of coverage. Pt clusters adsorbed (a) on the Al2O3 terraces and (b) on the DBs. STM measurements were performed with Vs = 3.5 V and It = 0.1 nA at 78 K. Apparent heights of 81−779 and 27−177 clusters were measured for (a) and (b), respectively.

Figure 1. STM topographic images of Pt15/Al2O3/NiAl(110) as a function of deposited Pt15 coverage (Vs = 3.5 V, It = 0.1 nA, 50 × 50 nm2): (a) 0.007 Pt15, (b) 0.021 Pt15, (c) 0.043 Pt15, and (d) 0.071 Pt15 clusters/nm2. Pt15 clusters were deposited at 300 K, followed by STM measurements at 78 K.

tunneling current), the apparent cluster height was overestimated by ∼0.26 nm compared with the real cluster height due to the band gap of the Al2O3 film.28 With increasing coverage, clusters start to aggregate; above 0.021 Pt15 clusters/ nm2, three-dimensional (3D) two-layer clusters (Pt30 clusters formed by aggregation of two Pt15 clusters) were observed at ∼0.65 nm. Upon increasing the coverage to 0.071 Pt15 clusters/ nm2, 3D three-layer clusters (Pt45 clusters formed by aggregation of three Pt15 clusters) were observed at an apparent cluster height of ∼0.85 nm. This result clearly indicates that cluster aggregation starts from a low coverage of ∼0.021 Pt15 clusters/nm2. Interestingly, similar coverage dependence is observed in Figure 2b, indicating that cluster aggregation occurs on both the Al2O3 terraces and DBs. The height histograms (Figure 2) also suggest that the morphology and binding energy of Ptn clusters on the Al2O3 terrace are almost the same as those on DBs.28 We previously reported an analytical model to calculate the fraction of isolated and aggregated clusters as a function of coverage.38 In this model, the clusters were randomly distributed on the surface, and clusters adsorbed below certain intercluster distances were counted as aggregates. This intercluster distance defines the cross section (σ) of cluster aggregation. Assuming that the lattice constant of the 2D Pt15 clusters is the same as that of Pt(111), the amount of isolated clusters (θm; monomer coverage) becomes θm = θT exp(−σθT),

images of Pt15 deposited on an Al2O3/NiAl(110) surface at various coverages. The Al2O3 surface is terminated by oxygen atoms.31,32 The Al2O3 film grows in two reflection domains tilted by ±24° with respect to the [1−10] direction of NiAl(110).31 The bright stripes are domain boundaries (DBs), which appear as line defects of the Al2O3 film.33−37 The topographic contrast of the DBs was not from the height difference but from the electronic effect; the structure is atomically flat across the DBs.33−37 The protrusions observed in Figure 1a were assigned to size-selected Pt15 clusters.28 After Pt15 cluster deposition, we did not find any increase in surface defects or fragmented Pt monomers, which confirms that Pt15 clusters were soft-landed intact on the Al2O3 surface.28 At 0.021 Pt15 clusters/nm2 (Figure 1b), the density of the protrusions increases and larger clusters start to appear. With increasing coverage (Figure 1c,d), the number of larger clusters increased; thus, they were assigned to aggregated clusters. Figure 2 shows histograms of apparent cluster height as a function of coverage. Figure 2a,b shows histograms for clusters adsorbed on the Al2O3 terraces and on the DBs, respectively. In Figure 2a, below 0.007 Pt15 clusters/nm2, the histogram is centered at ∼0.4 nm which can be ascribed to the twodimensional (2D) planar structure.28 Note that at the STM conditions of Vs = 3.5 V and It = 0.1 nA (Vs: sample bias; It: 4711

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715

Letter

The Journal of Physical Chemistry Letters where θT is the total coverage of deposited clusters (see the Supporting Information). On the basis of this model, the coverage dependence of the isolated Pt15 clusters would provide the cross section for cluster aggregation. In this study, θm was obtained in two ways. First, it was estimated from the histogram of the apparent cluster height. As shown in Figure 3a at θT = 0.071 Pt15 clusters/nm2, fractions of

monomer is 56%. Therefore, θm is 0.040 Pt15 clusters/nm2 at θT = 0.071 Pt15 clusters/nm2. θm values estimated from the height histogram and the density uptake are plotted as black and gray points in Figure 3d, respectively. The θm values agree well with each other. Here, θm estimated from the height histogram is slightly lower than that from the density uptake due to Pt15 trimer formation. Solid curves represent fitted results using the formula θT exp(−σθT); the parameter σ obtained was 9.4 nm2 (black curve) and 8.1 nm2 (gray curve). To confirm the obtained σ value, we further analyzed the STM images. The intercluster distance provides the cross section in a straightforward way; thus, we analyzed nearestneighbor distances (NNDs) between the observed clusters (Figure 4a,b). Figure 4a shows the histogram of the NND as a function of coverage. With increasing coverage, the distributions become sharper and shift to shorter NNDs. Solid curves represent the calculated results based on a random distribution model without aggregation.39 Below 0.007 Pt15 clusters/nm2, the histograms show broad distribution peaks at ∼5 nm, which nearly follow the random distribution model. However, above 0.021 Pt15 clusters/nm2, the histograms strongly deviate from the calculated results; the experimental value is lower than the calculated one below NND of ∼3 nm and vice versa above this distance. This result means that clusters adsorbed at short NNDs aggregated, and aggregated clusters would be counted at long NNDs. Clusters with a NND below 1.57 nm were not observed (see the inset of Figure 4a at 0.071 Pt15 clusters/nm2), indicating that Pt15 clusters aggregate when the intercluster distance is below this distance. Thus, σ was estimated to be 7.7 nm2 (= π × 1.572) (Figure 4c), which is in reasonable agreement with the above results. The cross section of cluster aggregation (σ) can be determined by a cluster sectional area and mobility on the surface. The cluster sectional area provides the geometrical cross section (σg). σg can be defined from the sectional area of two clusters separated infinitely. Because Pt15 clusters are adsorbed as 2D planar structures and incoming Pt15+ cluster ions are 3D structures,40,41 σg between 2D planar (adsorbed) and 3D (gas-phase) Pt15 clusters should be considered (Figure 4c). Using the lattice constant of the Pt(111) surface, the geometrical area of the 2D Pt15 cluster is calculated to be 0.57 nm2. The sectional area of the 3D Pt15 cluster would be ∼0.13 nm2.40 Thus, the radius (rg) of σg becomes (0.57/π)0.5 + (0.13/ π)0.5, and then, σg is 1.3 nm2. The experimentally obtained σ was 7.7−9.4 nm2. Therefore, σ is about seven times larger than the expected value from geometrical considerations, indicating that mobility significantly contributes to σ. Note that considering the σg between 2D Pt15 clusters, σg becomes 2.3 nm2, and the obtained σ is about four times larger than the σg. Ptn clusters were thermally immobile at 300 K on Al2O3/ NiAl(110); transient migration should be included in order to explain the obtained cross section.27,28 Pt15+ cluster ions approaching the Al2O3/NiAl(110) surface would be gradually accelerated by the image potential, which shifts energy levels of the Pt15+ and reduces the electron-tunneling barrier.42 In this experiment, the average impact energy of clusters to the substrate was 7 eV/cluster; the average speed of incoming Pt15+ clusters was 680 m/s. As a result, it takes 1.5 ps to travel 1 nm. This time is sufficiently long compared with the time scale of charge transfer.43 Therefore, neutralization of the clusters would occur via electron tunneling at a certain distance from the surface. Electron tunneling would occur from the NiAl

Figure 3. (a) Histogram of the apparent cluster height at 0.071 Pt15 clusters/nm2. Fractions of Pt15 monomer, dimer, and trimer were estimated by fitting, as shown by the red, blue, and green curves, respectively. (b) Schematic illustrations of the Pt15 monomer, dimer, and trimer. (c) Density of observed clusters as a function of coverage. The solid curve represents the fitted result using a second-order polynomial. The dashed line represents the density uptake without aggregation. (d) θm as a function of coverage. The estimated θm from the apparent cluster height histogram and density uptake is plotted as black and gray points, respectively. Solid curves are fitted results using the equation θm = θT exp(−σθT).

the isolated Pt15 clusters (Pt15 monomer) and its aggregates (Pt15 dimer and trimer) were estimated by fitting the apparent height histogram. Figure 3b shows schematic illustrations of the structures of Pt15 from the monomer to its trimer. The fraction of Pt15 monomer was 69% of the observed clusters. Fractions of the Pt15 dimer and trimer were 26 and 5% of the observed clusters, respectively. We could then estimate the fraction of the Pt15 monomer to be 0.69/(0.69 + 2 × 0.26 + 3 × 0.05) = 50% of the deposited Pt15 clusters. As a result, θm was obtained as 0.036 Pt15 clusters/nm2 at θT = 0.071 Pt15 clusters/nm2. The coverage dependence of θm was obtained, as shown in Figure S1. Second, θm was estimated from the density of observed clusters as a function of θT, as shown in Figure 3c. Initially, the density grew linearly, indicating that Pt15 clusters adsorb without aggregation. Above θT ≈ 0.04 Pt15 clusters/nm2, the experimental results clearly deviated from the dashed line (the density without aggregation). At θT = 0.071 Pt15 clusters/nm2, the observed density was 78% of θT due to cluster aggregation. Because the experimental data can be fitted reasonably well by a second-order polynomial (solid curve), it is deduced that dimerization of Pt15 clusters dominates the aggregation. Thus, it was estimated that 44% [= 2 × (1−0.78)] of the deposited Pt15 clusters become Pt15 dimers. Hence, the fraction of Pt15 4712

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715

Letter

The Journal of Physical Chemistry Letters

Figure 4. (a) Histograms of NNDs (bars) and calculated random distribution without aggregation (solid curves) as a function of coverage. The inset at 0.071 Pt15 clusters/nm2 shows the leading edge region; the blue line is the linear fitted result at a NND of 1.5−2.0 nm. (b) Schematic illustration of NNDs of the clusters. (d) Schematic illustration of the geometrical and estimated cross sections between gas-phase and adsorbed Pt15 clusters.

atoms, the potential energy surface becomes flatter and more isotropic as the cluster size increases. This is because the cluster inner bonds screen the cluster−surface interactions and bond directionality, which would also support the observed large cross section.17 Energy dissipation via EHP excitations on DBs could be more efficient than that on the Al2O3 terrace because of the smaller band gap of the DBs.37 We have reported that ∼20% of the deposited Ptn clusters adsorbed on the DBs, although the area fraction of the DBs was ∼10% of the Al2O3 surface, indicating that transiently migrating Ptn clusters would be thermalized more efficiently on DBs than on the Al2O3 terrace.28 We have also reported that Ptn clusters adsorbed on the TiO2(110) surface without preferential adsorption at step edges.48 Because the band gap of the TiO2 substrate is much smaller than that of Al2O3, transient migration of Ptn clusters on TiO2 would not contribute significantly to the initial adsorption sites. Note that atomic corrugation of the TiO2(110) surface is larger than that of the Al2O3 surface.35 Atomic corrugation of the surface would also contribute to the transient migration length: the larger the surface corrugation, the more corrugated the potential energy surface; thus, hot adsorbates would have more of a chance of inelastic collision with the surface. Recently, Bonanni et al. reported that ripening of the sizeselected Ptn clusters deposited on TiO2(110) was enhanced by CO oxidation.10 They concluded that the CO oxidation energy excites the Ptn/TiO2(110) system, which induces cluster ripening via transient migration. As demonstrated in the present study, a slow energy dissipation process would significantly contribute to reaction-induced ripening. Note that dimerization is evident from the low coverage at ∼0.02 Pt15 clusters/nm2. This was consistent with a previous study for size-selected Au clusters on SiO2 and Al2O3 surfaces49 that indicates that transient migration of the Au clusters should be considered. To fabricate cluster-monodisperse surfaces, we infer that the coverage of deposited clusters should be below

substrate to the cluster because the ionization potential of the gas-phase Pt15 cluster (∼6.5 eV) is smaller than the valence band maximum of the Al2O3 film (9.1 eV from the vacuum level).44,45 By this neutralization, the NiAl substrate would obtain an energy of ∼1.9 eV [the difference between the work function of Al2O3/NiAl(110) (∼4.6 eV) and the ionization potential of the Pt15 cluster (∼6.5 eV)].46 The structure of the gas-phase Pt15+ cluster ion would be maintained after neutralization.41 Neutralized Pt15 clusters collide with the surface inelastically and dissipate excess translational and adsorption energy to the heat bath of the substrate. Using the excess energy, various electronic and vibrational modes are excited in the cluster. Pt15 clusters efficiently store the excess energy to their vibrational degrees of freedom because they have various vibrational modes (15 × 3 = 45 modes including frustrated translation and rotational modes). Here, we did not find any evidence of cluster fragmentation after cluster deposition. This suggests that the excess energy does not localize in a specific vibrational mode relating to a reaction coordinate of cluster fragmentation but distributes over all of the accessible vibrational modes, although the excess energy would be larger than the binding energy of Pt atoms in the Pt15 cluster.40 During the energy dissipation process, EHP excitations inside of the substrate would efficiently transfer the excess energy.47 However, because the Al2O3 film has a band gap of ∼6.7 eV, EHP excitations below the band gap energy are impossible. Furthermore, the large mass difference between the Pt and surface oxygen atom means that the excess energy would not be transferred efficiently to the substrate via the rigid-body collision process. Therefore, the vibrationally excited clusters would slowly become thermalized on the Al2O3 surface. Using the vibrational energy, the clusters could change their atomic configuration by means of structural isomerization. In this process, the atom-by-atom rearrangement would result in transient migration on the surface. Thus, the slow vibrational energy dissipation process would result in a long transient migration length and a large cross section for aggregation. It is reported that for clusters composed of several 4713

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715

Letter

The Journal of Physical Chemistry Letters ∼0.05 ML (1 ML = 1.5 × 1015 atoms/cm2), at which ≤30% of deposited clusters would aggregate. Our results demonstrated for the first time that the transient migration of clusters significantly contributes to the initial aggregation process on the oxide surface. The cross section for aggregation was seven times larger than the expected value from geometrical considerations. The large number of vibrational modes in the cluster and a lack of EHP excitations in the oxide would decrease the energy transfer rate from the cluster to the substrate. Hence, the clusters showed significant transient mobility. Fundamental understanding of transient migration is crucial to make progress in various fields ranging from nanomaterial fabrication to catalytic reaction dynamics.



(11) Matsushima, T. The Mechanism of the CO2 Formation on Pt(111) and Polycrystalline Surfaces at Low-Temperatures. Surf. Sci. 1983, 127, 403−423. (12) Tully, J. C. The Dynamics of Adsorption and Desorption. Surf. Sci. 1994, 299-300, 667−677. (13) Yoshinobu, J.; Tsukahara, N.; Yasui, F.; Mukai, K.; Yamashita, Y. Lateral Displacement by Transient Mobility in Chemisorption of CO on Pt(997). Phys. Rev. Lett. 2003, 90, 248301. (14) Novko, D.; Blanco-Rey, M.; Juaristi, J. I.; Alducin, M. Ab Initio Molecular Dynamics with Simultaneous Electron and Phonon Excitations: Application to the Relaxation of Hot Atoms and Molecules on Metal Surfaces. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 92, 201411. (15) Meyer, J.; Reuter, K. Modeling Heat Dissipation at the Nanoscale: An Embedding Approach for Chemical Reaction Dynamics on Metal Surfaces. Angew. Chem., Int. Ed. 2014, 53, 4721−4724. (16) Bunermann, O.; Jiang, H. Y.; Dorenkamp, Y.; Kandratsenka, A.; Janke, S. M.; Auerbach, D. J.; Wodtke, A. M. Electron-Hole Pair Excitation Determines the Mechanism of Hydrogen Atom Adsorption. Science 2015, 350, 1346−1349. (17) Vasco, E.; Sacedon, J. L. Role of Cluster Transient Mobility in Pulsed Laser Deposition-Type Growth Kinetics. Phys. Rev. Lett. 2007, 98, 036104. (18) Johnston, R. L. Atomic and Molecular Clusters; Taylor & Francis: London, U.K., 2002. (19) Claridge, S. A.; Castleman, A. W.; Khanna, S. N.; Murray, C. B.; Sen, A.; Weiss, P. S. Cluster-Assembled Materials. ACS Nano 2009, 3, 244−255. (20) Tyo, E. C.; Vajda, S. Catalysis by Clusters with Precise Numbers of Atoms. Nat. Nanotechnol. 2015, 10, 577−588. (21) Watanabe, Y. Atomically Precise Cluster Catalysis Towards Quantum Controlled Catalysts. Sci. Technol. Adv. Mater. 2014, 15, 063501. (22) Takao, S.; Sekizawa, O.; Samjeske, G.; Nagamatsu, S.; Kaneko, T.; Yamamoto, T.; Higashi, K.; Nagasawa, K.; Uruga, T.; Iwasawa, Y. Same-View Nano-XAFS/STEM-EDS Imagings of Pt Chemical Species in Pt/C Cathode Catalyst Layers of a Polymer Electrolyte Fuel Cell. J. Phys. Chem. Lett. 2015, 6, 2121−2126. (23) Kwon, G.; Ferguson, G. A.; Heard, C. J.; Tyo, E. C.; Yin, C. R.; DeBartolo, J.; Seifert, S.; Winans, R. E.; Kropf, A. J.; Greeley, J.; Johnston, R. L.; Curtiss, L. A.; Pellin, M. J.; Vajda, S. Size-Dependent Subnanometer Pd Cluster (Pd4, Pd6, and Pd17) Water Oxidation Electrocatalysis. ACS Nano 2013, 7, 5808−5817. (24) Proch, S.; Wirth, M.; White, H. S.; Anderson, S. L. Strong Effects of Cluster Size and Air Exposure on Oxygen Reduction and Carbon Oxidation Electrocatalysis by Size-Selected Ptn (n ≤ 11) on Glassy Carbon Electrodes. J. Am. Chem. Soc. 2013, 135, 3073−3086. (25) Nesselberger, M.; Roefzaad, M.; Faycal Hamou, R.; Ulrich Biedermann, P.; Schweinberger, F. F.; Kunz, S.; Schloegl, K.; Wiberg, G. K. H.; Ashton, S.; Heiz, U.; Mayrhofer, K. J. J.; Arenz, M. The Effect of Particle Proximity on the Oxygen Reduction Rate of Size-Selected Platinum Clusters. Nat. Mater. 2013, 12, 919−924. (26) Berr, M. J.; Schweinberger, F. F.; Doblinger, M.; Sanwald, K. E.; Wolff, C.; Breimeier, J.; Crampton, A. S.; Ridge, C. J.; Tschurl, M.; Heiz, U.; Jackel, F.; Feldmann, J. Size-Selected Subnanometer Cluster Catalysts on Semiconductor Nanocrystal Films for Atomic Scale Insight into Photocatalysis. Nano Lett. 2012, 12, 5903−5906. (27) Beniya, A.; Isomura, N.; Hirata, H.; Watanabe, Y. Lateral Displacement in Soft-Landing Process and Electronic Properties of Size-Selected Pt7 Clusters on the Aluminum Oxide Film on NiAl(110). Chem. Phys. Lett. 2013, 576, 49−54. (28) Beniya, A.; Isomura, N.; Hirata, H.; Watanabe, Y. Morphology and Chemical States of Size-Selected Ptn Clusters on an Aluminium Oxide Film on NiAl(110). Phys. Chem. Chem. Phys. 2014, 16, 26485− 26492. (29) Heiz, U.; Landman, U. Nanocatalysis; Springer: Berlin, Germany; 2007.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.6b02362. Experimental methods, analytical model to estimate the cross section for cluster aggregation, coverage estimation of the Pt15 monomer and its aggregates from an apparent height histogram, and kinetic energy distribution of Pt15+ clusters (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Yoshihide Watanabe: 0000-0001-9999-0486 Notes

The authors declare no competing financial interest.



REFERENCES

(1) Barth, J. V. Transport of Adsorbates at Metal Surfaces: From Thermal Migration to Hot Precursors. Surf. Sci. Rep. 2000, 40, 75− 149. (2) Weiss, P. S.; Eigler, D. M. Adsorption and Accommodation of Xe on Pt(111). Phys. Rev. Lett. 1992, 69, 2240−2243. (3) Barth, J. V.; Zambelli, T.; Wintterlin, J.; Ertl, G. Hot Precursors in the Adsorption of Molecular Oxygen on Ag(110). Chem. Phys. Lett. 1997, 270, 152−156. (4) Johnson, G. E.; Priest, T.; Laskin, J. Charge Retention by Gold Clusters on Surfaces Prepared Using Soft Landing of Mass Selected Ions. ACS Nano 2012, 6, 573−582. (5) Mitsui, M.; Nagaoka, S.; Matsumoto, T.; Nakajima, A. SoftLanding Isolation of Vanadium-Benzene Sandwich Clusters on a Room-Temperature Substrate Using n-Alkanethiolate Self-Assembled Monolayer Matrixes. J. Phys. Chem. B 2006, 110, 2968−2971. (6) Popok, V. N.; Barke, I.; Campbell, E. E. B.; Meiwes-Broer, K.-H. Cluster−Surface Interaction: From Soft Landing to Implantation. Surf. Sci. Rep. 2011, 66, 347−377. (7) Gao, D. Z.; Watkins, M. B.; Shluger, A. L. Transient Mobility Mechanisms of Deposited Metal Atoms on Insulating Surfaces: Pd on MgO (100). J. Phys. Chem. C 2012, 116, 14471−14479. (8) Moseler, M.; Häkkinen, H.; Landman, U. Supported Magnetic Nanoclusters: Soft Landing of Pd Clusters on a MgO Surface. Phys. Rev. Lett. 2002, 89, 176103. (9) Winkler, A.; Tumbek, L. Nucleation of Organic Molecules via a Hot Precursor State: Pentacene on Amorphous Mica. J. Phys. Chem. Lett. 2013, 4, 4080−4084. (10) Bonanni, S.; Aït-Mansour, K.; Harbich, W.; Brune, H. ReactionInduced Cluster Ripening and Initial Size-Dependent Reaction Rates for CO Oxidation on Ptn/TiO2(110)-(1 × 1). J. Am. Chem. Soc. 2014, 136, 8702−8707. 4714

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715

Letter

The Journal of Physical Chemistry Letters (30) Nilius, N. Properties of Oxide Thin Films and their Adsorption Behavior Studied by Scanning Tunneling Microscopy and Conductance Spectroscopy. Surf. Sci. Rep. 2009, 64, 595−659. (31) Jaeger, R. M.; Kuhlenbeck, H.; Freund, H. J.; Wuttig, M.; Hoffmann, W.; Franchy, R.; Ibach, H. Formation of a Well-Ordered Aluminium Oxide Overlayer by Oxidation of NiAl(110). Surf. Sci. 1991, 259, 235−252. (32) Kresse, G.; Schmid, M.; Napetschnig, E.; Shishkin, M.; Kohler, L.; Varga, P. Structure of the Ultrathin Aluminum Oxide Film on NiAl(110). Science 2005, 308, 1440−1442. (33) Kulawik, M.; Nilius, N.; Rust, H. P.; Freund, H. J. Atomic Structure of Antiphase Domain Boundaries of a Thin Al2O3 Film on NiAl(110). Phys. Rev. Lett. 2003, 91, 256101. (34) Simon, G. H.; Konig, T.; Heinke, L.; Lichtenstein, L.; Heyde, M.; Freund, H. J. Atomic Structure of Surface Defects in Alumina Studied by Dynamic Force Microscopy: Strain-Relief-, Translationand Reflection-Related Boundaries, Including their Junctions. New J. Phys. 2011, 13, 123028. (35) Simon, G. H.; Konig, T.; Rust, H. P.; Heyde, M.; Freund, H. J. Atomic Structure of the Ultrathin Alumina on NiAl(110) and its Antiphase Domain Boundaries As Seen by Frequency Modulation Dynamic Force Microscopy. New J. Phys. 2009, 11, 093009. (36) Nilius, N.; Kulawik, M.; Rust, H. P.; Freund, H. J. DefectInduced Gap States in Al2O3 Thin Films on NiAl(110). Phys. Rev. B: Condens. Matter Mater. Phys. 2004, 69, 121401. (37) Schmid, M.; Shishkin, M.; Kresse, G.; Napetschnig, E.; Varga, P.; Kulawik, M.; Nilius, N.; Rust, H. P.; Freund, H. J. Oxygen-Deficient Line Defects in an Ultrathin Aluminum Oxide Film. Phys. Rev. Lett. 2006, 97, 046101. (38) Beniya, A.; Hirata, H.; Watanabe, Y. Uniform Deposition of Size-Selected Clusters Using Lissajous Scanning. J. Vac. Sci. Technol., A 2016, 34, 031601. (39) Knorr, N.; Brune, H.; Epple, M.; Hirstein, A.; Schneider, M. A.; Kern, K. Long-Range Adsorbate Interactions Mediated by a TwoDimensional Electron Gas. Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 65, 115420. (40) Kumar, V.; Kawazoe, Y. Evolution of Atomic and Electronic Structure of Pt Clusters: Planar, Layered, Pyramidal, Cage, Cubic, and Octahedral Growth. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 77, 205418. (41) Chaves, A. S.; Rondina, G. G.; Piotrowski, M. J.; Tereshchuk, P.; Da Silva, J. L. F. The Role of Charge States in the Atomic Structure of Cun and Ptn (n = 2−14 atoms) Clusters: A DFT Investigation. J. Phys. Chem. A 2014, 118, 10813−10821. (42) Kasi, S. R.; Kang, H.; Sass, C. S.; Rabalais, J. W. Inelastic Processes in Low-Energy Ion-Surface Collisions. Surf. Sci. Rep. 1989, 10, 1−104. (43) Arnolds, H.; Bonn, M. Ultrafast Surface Vibrational Dynamics. Surf. Sci. Rep. 2010, 65, 45−66. (44) Andersson, S.; Brühwiler, P. A.; Sandell, A.; Frank, M.; Libuda, J.; Giertz, A.; Brena, B.; Maxwell, A. J.; Bäumer, M.; Freund, H. J.; Mårtensson, N. Metal−Oxide Interaction for Metal Clusters on a Metal-Supported Thin Alumina Film. Surf. Sci. 1999, 442, L964−L970. (45) Nie, A. H.; Wu, J. P.; Zhou, C. G.; Yao, S. J.; Luo, C.; Forrey, R. C.; Cheng, H. S. Structural Evolution of Subnano Platinum Clusters. Int. J. Quantum Chem. 2007, 107, 219−224. (46) Song, W.; Yoshitake, M. A Work Function Study of Ultra-Thin Alumina Formation on NiAl(1 1 0) Surface. Appl. Surf. Sci. 2005, 251, 14−18. (47) Nienhaus, H. Electronic Excitations by Chemical Reactions on Metal Surfaces. Surf. Sci. Rep. 2002, 45, 1−78. (48) Isomura, N.; Wu, X.; Watanabe, Y. Atomic-Resolution Imaging of Size-Selected Platinum Clusters on TiO2(110) Surfaces. J. Chem. Phys. 2009, 131, 164707. (49) Vajda, S.; Winans, R. E.; Elam, J. W.; Lee, B. D.; Pellin, M. J.; Seifert, S.; Tikhonov, G. Y.; Tomczyk, N. A. Supported Gold Clusters and Cluster-Based Nanomaterials: Characterization, Stability and Growth Studies by in situ GISAXS Under Vacuum Conditions and in the Presence of Hydrogen. Top. Catal. 2006, 39, 161−166. 4715

DOI: 10.1021/acs.jpclett.6b02362 J. Phys. Chem. Lett. 2016, 7, 4710−4715