Pt

---

Subscriber access provided by University of Sussex Library

Article

Interaction of CO with Structurally Well Defined Monolayer PtAu/Pt(111) Surface Alloys Menhild Eyrich, Thomas Diemant, Heinrich Hartmann, Joachim Bansmann, and R. Jürgen Behm J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/jp302469c • Publication Date (Web): 03 May 2012 Downloaded from http://pubs.acs.org on May 4, 2012

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Interaction of CO with structurally well defined monolayer PtAu/Pt(111) surface alloys M. Eyrich, T. Diemant, H. Hartmann, J. Bansmann, and R.J. Behm* Institute of Surface Chemistry and Catalysis, Ulm University, D-89069 Ulm, Germany

The adsorption properties of structurally well defined bimetallic PtAu/Pt(111) monolayer surface alloys, with known lateral distribution of the respective surface species and varied Au surface contents, were studied by temperature programmed desorption (TPD) and infrared reflection absorption spectroscopy (IRAS), using CO as probe molecule. The surface composition and the lateral distribution of surface atoms in the PtAu/Pt(111) surface alloys were previously determined by high-resolution scanning tunneling microscopy (STM) [A. Bergbreiter et al., ChemPhysChem 11 (2010) 1505], showing a tendency for lateral segregation of both metals in the surface layer. CO adsorption on these surfaces is dominated by adsorption on Pt on-top sites, the fraction of bridge bonded COad decays rapidly with increasing Au surface content and is completely absent for surfaces with 35% and more surface Au content. Adsorption on Au sites is negligible at 100 K. The roles of electronic ligand and strain effects and of geometric ensemble effects on the CO adsorption properties, including energetics, maximum COad coverage, vibrational properties, and adsorption kinetics, are mapped out and discussed. The results are compared with CO adsorption on comparable surface alloys such as PtAg/Pt(111), PdAu/Pd(111) and PdAg/Pd(111) as well as on the inverse system PtAu/Au(111). Keywords: Bimetallic Surfaces; Surface Alloys; Site specific adsorption, Temperature programmed desorption; Infrared absorption spectroscopy; Gold; Platinum; Carbon monoxide Resubmitted to J. Phys.Chem. C: 02.05.2012

*

Author, to whom correspondence should be addressed, e-mail: [email protected]

ACS Paragon Plus Environment

The Journal of Physical Chemistry

Page 2 of 43

2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 Introduction Bimetallic supported catalysts have attracted enormous interest due to their potential for enhanced reactivity and selectivity compared to monometallic catalysts.

1-5

Among other

effects, dilution of the active metal by an inactive species was supposed to increase the selectivity for specific reaction products and often also the activity, mainly by reducing the self-poisoning of the catalyst by reactants.6;7 In a more general picture, the modified reaction behavior of bimetallic catalysts was explained by electronic ligand effects

8-10

and geometric

ensemble effects.8;10 In the first case, modifications in the reaction characteristics were attributed to changes in the electronic properties of the reaction site / reaction ensemble due to variations in the neighboring shell of surface atoms; in the second model, they were related to variations in the size and composition of the reaction ensemble. Later, electronic strain effects, reflecting modifications in the chemical properties of the surface species compared with the pure bulk surface due to variations in the lattice spacing, were introduced as an additional contribution.11;12 Supported PtAu catalysts, which can be considered as a case where the active metal Pt is diluted by the inert component Au, were found to be promising candidates for hydrocarbon conversion,6;13 or, more recently, as core-shell catalysts for O2 reduction 14 or in NO sensors.15 They were also among the first ones where the modified reaction behavior was studied systematically using planar model systems.6;16 In these earlier studies, however, the twodimensional (2D) distribution of the surface species was not accessible and hence unknown. Therefore, any interpretation using an assumed distribution of the surface species was somewhat speculative. In the meantime, the distribution of surface atoms in bimetallic surfaces has become accessible by high resolution scanning tunneling microscopy (STM) with chemical contrast,1724

allowing systematic studies of the correlation between surface structure and chemical ACS Paragon Plus Environment

Page 3 of 43

The Journal of Physical Chemistry

3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

properties. These studies aim to explain the chemisorption and chemical properties in terms of local reactivities of specific surface ensembles.22;25-30 In the present study, we will use this approach to investigate the chemisorption behavior of structurally well defined monolayer PtAu/Pt(111) surface alloys using adsorbed CO as probe molecule. A previous quantitative high resolution STM study on the distribution of surface atoms in PtAu monolayer surface alloys of varying compositions on Pt(111), which was performed in our laboratory, had shown a tendency towards two-dimensional (2D) clustering of similar type surface atoms in the surface layer, i.e., segregation to homo-atomic islands, reflecting an energetic preference for alike surface neighborhoods compared to bonding between unlike surface atoms.30 A more detailed summary of the results of this study is given at the beginning of Section 3. These finding were confirmed by a combined density functional theory (DFT) / Monte Carlo (MC) study by Stephens et al..31 This distribution, which qualitatively resembles previous findings for a number of other monolayer surface alloys such as PdRu/Ru(0001),27, PtAg/Pt(111),32 and PdAg/Pd(111),33 has significant consequences on the adsorption properties of the bimetallic surface. The chemical properties of PtAu/Pt(111) surface alloys, prepared by deposition of multilayer Au amounts on the surface and subsequent heating to temperatures above 1000 K, were studied before by Sachtler and Somorjai.6 Due to the preparation method, these alloy systems were not confined to the topmost layer, but also contained sizeable Au amounts in deeper layers. Therefore, the Pt surface atoms were not necessarily located on a Pt subsurface layer, but may be residing on Au containing PtxAu3-x sites in the second layer. The CO adsorption capacity was found to decrease with increasing Au concentration. Additionally, a shift of the CO desorption temperature from the Pt ensembles to higher temperature was found at higher COad coverage, which was tentatively explained by a reduction of the repulsive interactions between neighboring CO molecules due to formation of diluted Pt ensembles with increasing ACS Paragon Plus Environment

The Journal of Physical Chemistry

4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Au concentration. Sizeable ligand effects of adjacent Au surface atoms on the adsorption properties of the Pt ensembles were excluded.6 In this publication, we present results from Temperature Programmed Desorption (TPD) and Infrared Reflection Absorption Spectroscopy (IRAS) measurements, using CO as probe molecule, and correlate the chemical properties of the PtAu/Pt(111) surface alloys with their atomic structure. After a brief summary of the previous STM results on the surface atom distribution in the PtAu/Pt(111) monolayer surface alloys,30 we present and discuss CO TPD data, which aim at elucidating Au induced modifications of the adsorption strength of COad on ensembles of Pt surface atoms (section 3.1). We will also discuss the Au induced modification of the CO adsorption kinetics on these surfaces. Subsequently, information on the adsorption site occupation and on the variations in the C-O bond strength is derived from series of IRAS measurements (section 3.2). Finally, the results are discussed in comparison with findings reported for chemically comparable surface alloy systems PdAu/Pd(111),22;25;34;35 PtAg/Pt(111),36 and PdAg/Pd(111) 37;38 as well as on the inverse system PtAu/Au(111).39

2. Experimental The measurements were performed in an ultrahigh vacuum (UHV) system equipped with standard facilities for surface preparation and characterization such as an Ar+ sputter gun, a metal evaporator, and a cylindrical mirror analyzer for Auger Electron Spectroscopy (AES). The interaction of CO with the bimetallic surfaces was investigated using a shielded mass spectrometer (Pfeiffer Vacuum, QMS 200) for TPD measurements and an IR spectrometer (Bruker, Tensor 27) with a mercury-cadmium-telluride (MCT) detector for IRAS measurements. Additionally, the UHV system is equipped with facilities for X-ray Photoelectron Spectroscopy (XPS) measurements (XR3E2 X-ray source, CLAM 2 electron analyzer, VG Scientific).

ACS Paragon Plus Environment

Page 4 of 43

Page 5 of 43

The Journal of Physical Chemistry

5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Pt(111) crystal was initially cleaned by Ar+ ion sputtering at room temperature (1.2 kV, ion current 2.5 µA), treatment with O2 (800 K, 1 × 10-7 mbar, 10 min) and flash annealing to 1200 K. Repeated cycles of this procedure resulted in a surface with an impurity level below the detection limit of XPS and AES (< 1% of the surface).30 Before each experiment, surface contaminations were removed by brief Ar+ ion sputtering and heating to 1200 K. Subsequently, the sample was exposed 3 times to a dose of 7.5 L O2 (1 L = 10-6 Torr·s) and heated to 1000 K to remove oxidizable contaminations. The purity of the Pt(111) surface was controlled before Au deposition by a TPD spectrum of a saturated CO adlayer. This TPD spectrum was used also as reference for the COad coverage on the PtAu/Pt(111) surface alloys, using the well-known saturation coverage of 0.68 monolayer (ML) at 120 K

40-42

on

the unmodified Pt(111) surface as standard. PtAu/Pt(111) surface alloys were prepared by deposition of sub-monolayer amounts of Au with a home-made resistively heated evaporator on the Pt(111) surface (deposition rate 0.16 ML min-1, deposition at room temperature) and subsequent annealing at 1000 K for 10 sec. (annealing / cooling rate 4 K s-1 / 2 K s-1), similar to the procedure used in the previous STM study.30 For the TPD measurements, CO was dosed at 100 K and the heating ramp was controlled at 4 K s-1. The mass spectrometer was shielded against desorption from the sample holder by a cup with a central aperture of 4 mm width. The distance between cup and sample could be adjusted reproducibly via an electrical contact.43 The IR spectra were recorded at a detection angle of 7° with respect to the surface plane (resolution 4 cm-1), co-adding 1024 spectra. Reference spectra of the COad-free surfaces were recorded at a sample temperature of 600 K. XPS measurements were performed using non-monochromatic Al Kα radiation and a hemispherical sector analyzer operated in the fixed transmission mode at pass energies of 20.0 eV (detail scans) or 50.0 eV (survey scans). After subtraction of a Shirley background, ACS Paragon Plus Environment

The Journal of Physical Chemistry

Page 6 of 43

6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the intensity ratio of the Pt(4f) and Au(4f) peaks was determined. The precision of the Au surface contents is ±3 % (absolute). The XPS intensity ratio was calibrated using CO TPD spectra of a surface, where the minimum COad uptake was reached and where increasing the amount of Au caused no further decrease in the COad desorption. At this point, ~2% of the COad saturation coverage on a clean Pt(111) surface was observed after surface alloy formation. The remaining small COad coverage on the PtAu/Pt(111) surface alloy results from the presence of isolated small Pt aggregates and atoms in the surface, which remain independent of further Au evaporation before annealing. Thus, the maximum Au content in the AuPt/Pt(111) surface alloy is 98% Au. STM/AES measurements had shown that the loss of Au surface atoms during surface alloy formation did not exceed 10% − 15% of the initially deposited Au amount after annealing at 1000 K for 10 s.30 XPS measurements performed in this work on these bimetallic surfaces before and after the annealing step agree with these results. In agreement with previous literature reports,44 the Au loss is mainly assigned to bulk diffusion.

3. Results Before presenting the TPD and IR data, we will briefly summarize the most relevant results of the structural characterization of similarly prepared PtAu/Pt(111) surface alloys by high resolution STM, which were published recently.30 The procedure for preparing the PtAu/Pt(111) surface alloys leads to incorporation of the Au atoms in flat terraces separated by monolayer steps.30 Quantitative evaluation of atomically resolved STM images with chemical contrast of these surfaces, which could distinguish between ‘dark’ Au and ‘bright’ Pt atoms, allowed for a statistical analysis of the two-dimensional (2D) distribution of the two surface

species,30

using

the

Warren-Cowley

short-range-order

parameter

for

quantification.24;45 Unambiguous identification of the respective species was possible by the

ACS Paragon Plus Environment

Page 7 of 43

The Journal of Physical Chemistry

7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

trends in surface concentration with increasing Au deposition. The analysis revealed a pronounced clustering of like atoms, i.e., a tendency for 2D phase separation. This is illustrated in the three STM images depicted in Fig. 1, which were taken from ref. 30. Images and results of the quantitative analysis fit well to the low miscibility of Pt and Au in bulk alloys.31;46-48 Although the energetics favor the confinement of alloy formation to the outermost layer, entropy effects will nevertheless favor bulk dissolution of surface Au at the high temperatures, where kinetic limitations for two-dimensional (2D) intermixing are overcome.30 Therefore, care had to be taken to not exceed the annealing conditions described above. The STM images did not show any indications of surface dislocations or of the formation of a Moiré pattern, which would be indicative of a relaxed, non-pseudomorphic structure of the surface layer.49 Accordingly, the incorporation of the larger Au surface atoms in the Pt layer results in (local) compressive strain. Without going into details, the structural data allow two important conclusions relevant for the understanding of the adsorption data. First, due to the limitation of alloy formation to the outermost layer, the composition of the second atom layer of the substrate can be considered as homogeneous and therefore differences in vertical ligand effects arising from an inhomogeneous composition of that layer can be neglected. Second, the pronounced clustering of like atoms – with large homo-atomic ensembles – reduces the density of mixed ensembles and of Pt surface atoms with mixed neighborhoods. The latter reduces the impact of lateral electronic ligand and geometric ensemble effects on the adsorption properties of these surface alloys.

3.1 Interaction of CO with PtAu/Pt(111) surface alloys - temperature programmed desorption

ACS Paragon Plus Environment

The Journal of Physical Chemistry

8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Temperature programmed desorption (TPD) data sets were recorded on the pure Pt(111) surface and on PtAu/Pt(111) surface alloys with different surface Au contents (5, 20, 37, 42, 63, 90, and 98%) after CO adsorption at 100 K. Figure 2 displays sets of CO desorption spectra recorded on the different substrates with increasing COad exposure. The spectra recorded for the pure Pt(111) surface (Fig. 2a) agree well with results in previous studies.40;42;50-52 After small CO exposures, they show a symmetric desorption peak with a maximum at 466 K. With increasing coverage, this shifts to lower temperatures, e.g., 451 K after 0.6 L CO exposure, and becomes broader and less symmetric. At saturation coverage, the maximum is located at 382 K. The shift of the TPD peak is due to increasing repulsive interactions between the COad molecules on the surface. The desorption energy of 122 kJ mol-1, which was obtained by extrapolating the data of several CO coverages to a coverage of zero and assuming a first order mechanism for CO desorption with a value of 1013 s-1 for the pre-exponential factor,53 fits well to the values between 112 and 134 kJ mol-1 reported previously for the desorption barrier of CO from Pt(111).6;40;54-57 Sets of CO desorption spectra recorded on monolayer PtAu/Pt(111) surface alloys with Au contents between 5 and 98%, where the latter corresponds to a nearly complete Au monolayer with about 2% Pt atoms in the topmost layer on top of the Pt(111) single crystal, are shown in panels 2b – 2h. For comparison, a desorption spectrum recorded on the pure Pt(111) surface after COad saturation is included in each panel as dashed line. Despite of the wide variation in surface composition, the general appearance of the spectra does not change significantly and CO desorption takes place in the temperature range typical for desorption from Pt(111) surfaces, between ~300 K and 500 K (if the high temperature tail is not considered). Also the position of the main peak remains roughly the same, at about 460 K for all surface alloys (cf. Fig. 3a), in the limits of low COad coverages. This is slightly down-shifted (~6 K) with respect ACS Paragon Plus Environment

Page 8 of 43

Page 9 of 43

The Journal of Physical Chemistry

9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

to pure Pt(111), cf. Fig. 3b. Consequently, the desorption energy Edes (extrapolated to zero COad coverage) is more or less independent of the Au content (for θAu ≥ 5 %). For all surface alloys, the peak grows to lower temperatures with increasing COad coverage, indicative of either repulsive interactions between COad molecules or occupation of lower energy adsorption sites. Major difference between the different surfaces is the pronounced decay of the COad saturation coverage with increasing Au surface content (cf. Fig. 4b). A closer look, however, reveals a number of subtle differences between desorption from Pt(111) and from the PtAu/Pt(111) surface alloys. With increasing Au coverage, the trailing edge of the CO desorption peak shifts to slightly lower temperatures, which gets particularly obvious at the highest Au concentrations (≥ 90%). For surface alloys with higher Au contents (beyond 40%), a distinct shoulder forms at the low-temperature side of the CO desorption peak, in addition to a mere peak broadening with increasing COad coverage. For Au rich surfaces (>90% Au), this shoulder dominates the TPD spectra even after CO saturation. A possible explanation for this behavior is that the lower temperature TPD feature is related to desorption of the first of two COad molecules adsorbed on an isolated Pt dimer surrounded by Au surface atoms, while the second COad desorbs in the high temperature peak. A similar TPD feature has been reported before for PdAu/Pd(111)

25

and PdAg/Pd(111)

37;38

surface

alloys (with a TPD peak located at around ~250 K and 325 K for PdAu/Pd(111) and PdAg/Pd(111) surface alloys, respectivley). Adsorption of two CO molecules on an isolated Pd (or Pt) dimer is possible, but results in a lower effective binding energy for both COad molecules, due to repulsive adsorbate-adsorbate interaction. Accordingly, desorption of the first COad molecule occurs at significantly lower temperature compared to desorption of the second CO molecule from the Pt dimer. CO adsorption on Au sites can be excluded at these temperatures, since no corresponding COad species were observed by IR spectroscopy at 300 K.

ACS Paragon Plus Environment

The Journal of Physical Chemistry

10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Different from the trailing edge, which shows only a subtle shift to lower temperatures, the position of the main peak after COad saturation (peak maximum) increases from 382 K on the pure Pt(111) surface to 455 K on the PtAu/Pt(111) surface alloy with 90% Au (see Fig. 4a). The up-shift is explained by a lowering of the repulsive COad – COad interactions at COad saturation with increasing Au content, due to increasing adsorption on separated individual Pt surface atoms and small ensembles. This is reflected also by the smaller shift of the C-O band in the IR spectra (Fig. 6) with increasing Au surface content. An analogous up-shift of the TPD maxima at CO saturation was also observed by Sachtler and Somorjai,6 who investigated the CO desorption behavior on both epitaxial Au films on Pt(111) and on (multilayer) PtAu/Pt(111) surface alloys. For epitaxial Au films, the TPD peaks for COad saturation hardly changed in temperature with increasing Au content; while on the surface alloys, the peaks (peak maxima) exhibited a shift to higher temperatures with increasing Au content. From the negligible difference between the positions of the CO desorption peak for very low COad coverages on Pt(111) and for COad saturation on PtAu/Pt(111) surface alloy with very low Pt surface contents, hence in the absence of COad – COad repulsions, those authors concluded that ligand effects do not have a significant influence on the adsorption energy of CO on PtAu/Pt(111) surface alloys. The COad saturation coverage decays continuously with increasing Au surface content, to almost zero at the highest Au surface content (cf. Fig. 4b). Hence, under present experimental conditions, CO adsorption on / desorption from Au sites of the surface alloys at 100 K can essentially be excluded, with the exception from a small number of under-coordinated Au atoms (see IR measurements). This interpretation is in complete agreement with results of previous DFT studies on CO adsorption on the flat Au(111) surface, which determined a binding energy of only ~30 kJ mol-1.58;59 Stable CO adsorption becomes possible, however, at 100 K on under-coordinated atoms in the Au layer, e.g., on defect-rich (sputtered) Au(111)

ACS Paragon Plus Environment

Page 10 of 43

Page 11 of 43

The Journal of Physical Chemistry

11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

surfaces

60

and on stepped Au surfaces, e.g., on Au(110)-(1×2),61 Au(332),62;63 or Au(211).64

For the Au(110)-(1×2) surface, Gottfried et al. found a decrease of the isosteric heat of CO adsorption from 59 kJ mol-1 at very low coverage to 35 kJ mol-1 at θ ≈ 0.45 ML. In analogy, our IR measurements on the surface alloys demonstrate the adsorption of CO molecules on a small number of Au sites (see below). Because of the low coverage of these species (40%, the decay in COad coverage is less steep. We tentatively explain this behavior by edge effects, i.e., by high COad densities along the perimeter of the Pt areas, which apparently overcompensate the effects being dominant at lower Au surface contents. This is most easily illustrated for very small Pt ensembles. For

ACS Paragon Plus Environment

The Journal of Physical Chemistry

12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

example, we expect that similar to findings for PdAu/Pd(111),25 Pt monomers and dimers can take up 1 or 2 COad molecules, respectively, which would correspond to a normalized coverage of 1 COad per Pt surface atom. If adsorption on very small Pt ensembles were dominant, one would expect a slope of 1 (loss of 1 COad per surface Pt atom). The fact that this is not evident from the present data reflects that even at very high Au surface contents Pt monomers are not dominant. Even at 98% surface Au, the CO TPD spectra do not consist of a single peak, as expected for desorption from monomers, but exhibit a considerable low temperature shoulder, as would be expected for desorption of the first COad molecules out of Pt dimers/trimers.25;38 In fact, for PdAg/Pd(111) with its highly disperse distribution of Pd and Ag surface species, a slope of -1 is observed for very low surface Pd contents (60%), this behavior of the sticking coefficient becomes less pronounced (63% Au) and is finally no more visible (93% Au). Up to a Au amount of around

ACS Paragon Plus Environment

Page 12 of 43

Page 13 of 43

The Journal of Physical Chemistry

13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

42%, the development of the sticking coefficient can be described via the Kisliuk adsorption model,68 which was derived for adsorption via an extrinsic precursor state: θ   s ( θ) = s 0  1 + K ⋅  1 − θ 

−1

(1)

where θ denotes the relative COad coverage. For each surface, the respective saturation coverage was set to θ = 1, and for the initial sticking coefficient s0, the experimentally determined value was used. For the parameter K, a value of 0.08±0.01 was obtained for all surface fitted, the results are displayed as dashed lines in Fig. 5a. The development of the initial sticking coefficient with increasing surface Au content is plotted in Fig. 5b. First, there is a pronounced, almost linear decrease of the initial sticking coefficient with increasing Au surface content, from 0.9 for the pure Pt(111) surface to about 0.56 for a surface alloy with 42% surface Au content. For higher Au surface contents, the decrease in the initial sticking coefficient deviates slightly from the linear behavior, with the experimentally determined values of s0 being higher than anticipated for the linear relation. For the two samples with the highest Au surface contents, with 90 and 98% Au, we determined initial sticking coefficients of 0.2 and 0.15, which is much higher than expected from the linear correlation indicated (dashed line). This development of the initial sticking coefficient with increasing Au content fits to an adsorption model where (temporary) CO adsorption on Au sites acts as precursor state in the adsorption process. In general, we can define two limiting situations for CO adsorption on the PtAu/Pt(111) surface alloys. In the one case, CO adsorption is only possible if the adsorbing molecule impinges on a Pt atom or adsorption sites. In that case, the initial sticking coefficient should decrease linearly with the Pt surface content until zero for a closed Au layer on the Pt(111) substrate. In the extreme case of an infinite lifetime of the precursor state, it would not matter whether CO impinges on Au or Pt surface sites, it would always reach neighbored ACS Paragon Plus Environment

The Journal of Physical Chemistry

Page 14 of 43

14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pt sites. This would lead to a constant initial sticking coefficient, independent of Au surface content. The deviation of the measured initial sticking coefficient with decreasing Pt content clearly indicates that COad molecules impinging on Au surface atoms or adsorption sites can migrate a limited distance on the surface before they either find a Pt adsorption site or desorb.

3.2 Interaction of CO with PtAu/Pt(111) surface alloys - IR Spectroscopy We continue with the results of IR measurements at 100 K on unmodified Pt(111) and on PtAu/Pt(111) surface alloys with varying Au surface contents. The spectra recorded for increasing CO exposure/coverage on the pure Pt(111) surface (Fig. 6) are in excellent agreement with previous reports50;69-75 Up to θCO = 0.33 ML, a single band is detected, which shifts with increasing COad coverage from 2088 cm-1 to 2096 cm-1 and which is assigned to CO adsorbed on on-top sites. For higher COad coverages, the occupation of bridge sites is indicated by the appearance of a second IR band at ~1850 cm-1. At θCO = 0.5 ML, a c(4×2) structure is known to be formed

40;50

with an equal occupation of COad in on-top (2100 cm-1)

and bridge sites (1845 cm-1). A further increase of the COad coverage, up to saturation (θCO = 0.68 ML), leads to a shift of the on-top related band to 2105 cm-1 and to a slight decrease of the intensity of the bridge related feature, which results either from a successive backshift of some COad molecules from bridge to on-top sites with increasing CO coverage or from a change in the cross section of the respective features.41;73 For all surface alloys, only a single absorption band was detected for on-top COad on Pt. Apparently,

differences between

the

electronic

properties arising form

different

neighborhoods of the respective Pt atoms (ligand effects) are too small to be resolved. Instead, the IR spectra at COad saturation coverage (Fig. 7) show a change in full width at half maximum (FWHM), which increases from 6.4 cm-1 for the Pt(111) surface to 10.2 cm-1 for a surface alloy with 59% Au. For a surface alloy with 93% Au, the FWHM has a value of

ACS Paragon Plus Environment

Page 15 of 43

The Journal of Physical Chemistry

15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

6.2 cm-1 which is close to that of the Pt(111) surface. A similar peak broadening of the on-top COad band was observed by Mott et al. for Au rich (> 65%) bimetallic PtAu nanoparticles (NPs, size between 2 and 4 nm) on SiO2, in CO adsorption measurements on a series of PtAu NPs with different composition, although the positions of the IR bands are slightly different.76 The peak broadening may result, e.g., from different magnitudes of ligand and strain effects depending on the number of neighboring Au atoms, which in turn may also lead to modifications in the local CO adlayer arrangement. Both effects disappear at high surface Au contents, where Pt monomers with a single COad species adsorbed on top become increasingly dominant, leading to the observed decrease of FWHM on a Au-rich alloy. The band related to CO adsorbed on Pt-Pt bridge sites (at ~1850 cm-1) looses rapidly intensity upon alloy formation and is only hardly detectable for a surface alloy with 35% Au. Since CO adsorption at low coverages occurs, both for Pt(111) and for the PtAu/Pt(111) surface alloys, exclusively on on-top sites this must be related to the rapid loss of larger Pt areas which would allow to form densely packed CO adlayers similar to those appearing on Pt(111) at θCO = 0.5 and higher. In addition to the on-top and bridge bands of CO on Pt, a further band appears for high Au contents (above 59%) at 2115 cm-1, which based on its frequency is typical for CO adsorbed on Au surface atoms. For surface alloys with 59%Au, the band is only visible for high COad coverage; for a surface alloy with 93% Au, however, it is already detected at a low COad coverage. IR measurements at variable temperature show that this band disappears after heating above 200 K. Mott et al. detected a similar feature for CO adsorption on Au-rich (>65% Au) PtAu nanoparticles on SiO2;76 and a feature in this wave number region was observed before also in IR studies of CO adsorption on defect rich

60

or stepped

62;77;78

Au

single crystal surfaces. This band is attributed to CO adsorption on under-coordinated Au sites on the surface alloy, which are present, e.g., at step edges and defects. A band related to COad ACS Paragon Plus Environment

The Journal of Physical Chemistry

16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

on Au was also found in vibrational spectra of COad on Au island covered Pd(111) surfaces with a high number of Au step atoms

79

and on 10 ML thick PdAu alloy layers on

Mo(110).34;35 Similar to our findings, Ruggiero et al. also noted in experiments on stepped Au(332) that this band disappeared after heating to above 200 K, and concluded to an upper limit of ~50 kJ mol-1 for the CO binding energy on these sites.62 A more precise assessment of the binding energy of COad on these sites was not possible in the present experiments, as we could not resolve them in the TPD spectra. The simultaneous occupation of Pt and Au sites, which is observed on Au-rich alloys (93% Au, Fig. 6) already at low COad coverages, is not caused by a comparable adsorption energy of CO on Pt monomers and Au sites (COad bonds are much stronger to Pt than to Au sites), but by kinetically hindered surface diffusion of COad over the flat Au surface areas, after this has adsorbed and is trapped at a more stable (undercoordinated Au or Pt) adsorption site. Hence, at low COad coverages the occupation of undercoordinated Au sites represents a metastable configuration. Finally, we concentrate on frequency shifts of the bands related to CO adsorbed on-top on Pt surface atoms with increasing COad coverage and with increasing Au surface content. The IR spectra recorded on the surface alloys (Figs. 6b-f) after COad saturation show a continuous shift of this band to lower wave numbers with increasing Au surface content, from 2105 cm-1 for the unmodified Pt(111) to 2100 cm-1 and 2073 cm-1 for surface alloys with 7 and 93% Au, respectively (cf. spectra and inset (squares) in Fig. 7). This shift results from a combination of different effects; including dynamic dipole-dipole interactions between adjacent COad, a chemical shift due to interaction between COad molecules, and electronic effects induced by surface alloy formation. To eliminate the first two contributions, which both result from COadCOad interactions in the (saturated) CO adlayer, we compare IR spectra obtained after the smallest CO exposure (εCO = 0.08 L); the band positions are also plotted in the inset in Fig. 7 (open triangles). Also in the low COad coverage regime, the spectra show a steady red-shift of

ACS Paragon Plus Environment

Page 16 of 43

Page 17 of 43

The Journal of Physical Chemistry

17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

this IR band, at least up to a Au surface content of ~20%. From these data (inset in Fig 7 and spectra in Fig. 6), it is obvious that this band exhibits a blue-shift with increasing COad coverage on all surface alloys, similar to the finding for unmodified Pt(111). The shift of the band position from the lowest to the highest COad coverage, however, becomes smaller with increasing Au content in the surface alloys. After slightly increasing from 17 cm-1 for unmodified Pt(111) to 21 cm-1 for a Au coverage of 14%, it decreases in the following to only 16 or 11 cm-1 for surface alloys with 35% and 59% Au, respectively. The decrease of the COad coverage induced shift can directly be explained by the reduction of COad - COad interactions with increasing Au content due to the increasing separation of COad species on the respective Pt atoms/ensembles. (The absence of a detected shift at 93% surface Au is also due to the fact that already very low exposures lead to substantial local COad coverages, normalized to the number of Pt surface atoms.) A shift of the on-top COad band as a function of particle composition was observed also for bimetallic PtAu nanoparticles (size between 2 and 4 nm) on SiO2 surfaces by Mott and coworkers,76 they reported a shift from 2068 cm-1 for Pt-rich nanoparticles (65% Pt) to about 2050 cm-1 for Pt-poor particles (4% Pt). This result and its implications for the CO adsorption on the surface alloys will be discussed in more detail in the following section.

4. Discussion In this section, we will address a number of points which have not been discussed in detail in the preceding section and which are of general relevance also for the systematic understanding of adsorption on bimetallic surfaces. These include i) the physical origin of the observed modification of the adsorption properties of the PtAu/Pt(111) surface alloys as compared to Pt(111), and ii) characteristic differences in the adsorption properties compared

ACS Paragon Plus Environment

The Journal of Physical Chemistry

Page 18 of 43

18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

to the closely related PtAg/Pt(111),36 PdAu/Pd(111),25;34;35 and PdAg/Pd(111)

37;38

surface

alloys, including also the inverse systems AuPt/Au(111) 39;80-82 and PdAu/Au(111). 22

Physical origin of the modifications in adsorption energy compared to Pt(111). The discussion of the chemical properties of bimetallic surfaces is mostly carried out in the context of electronic ligand 6;8-10 and geometric ensemble effects 6;8;10 as well as modifications in the local electronic structure based on strain effects, caused by the insertion of a foreign metal atom in the surface layer of another metal with a different lattice constant.11;12 In the following, we will restrict ourselves to the modifications of the chemical properties of the Pt surface atoms/ensemble, since CO adsorption on Au sites was either negligible (35% Au) under present adsorption conditions. Even in the latter case it could not be quantified. Focusing on CO adsorption on PtAu/Pt(111) surface alloys, geometric ensemble effects can be neglected, at least for medium and higher Au surface contents (≥ 30%), since on these surfaces CO adsorbs exclusively in on-top sites, and also for low Au surface contents bridge bonded COad is only observed at higher COad coverage, stabilized by COad - COad interactions. In the case of pseudomorphic PtAu surface layers on the Pt(111) surface, the mismatch of the two fcc-lattice constants (Pt: 392 pm, Au: 408 pm) leads to increasing compressive strain in the surface layer with increasing Au surface content and thus, according to the d-band model,11;12;83 to a down-shift of the d-band center and consequently, to a lowering of the adsorption strength with increasing Au surface content. On the other hand, electronic ligand effects, which account for the modification of the electronic properties of the central atom or central ensemble carrying the adsorbed species due to interaction with neighboring surface atoms and thus for the resulting modification of the adsorption properties, can be assumed to strengthen the bond between adsorbate and Pt

ACS Paragon Plus Environment

Page 19 of 43

The Journal of Physical Chemistry

19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

surface atom with increasing number of neighboring Au surface atoms, due to the weaker PtAu bond as compared to Pt-Pt bonds.82 Vertical ligand effects, accounting for modifications in the second layer of the substrate, are absent due to the homogeneous composition of that layer (no Au atoms in the second layer). The role of electronic ligand effects on CO adsorption on PtAu/Pt(111) surface alloys has already been investigated by Sachtler and Somorjai.6 Evaluating the temperatures of the peak maxima (TM) in the CO TPD spectra recorded on pure Pt(111) and on various PtAu/Pt(111) surface alloys in the absence of COad – COad repulsions (in the limits of zero COad coverage on Pt(111), for COad saturation in the limits of very low Pt surface contents for PtAu/Pt(111) surface alloys), they found, within the experimental error of ±5 K, essentially no difference between the peak temperatures for CO desorption from Pt(111) or from PtAu/Pt(111) surface alloys.6 In both cases, COad – COad interactions can be neglected and desorption is solely determined by the substrate-adsorbate interaction. From these results they concluded that electronic ligand effects are negligible for CO adsorption on PtAu/Pt(111) surface alloys, and hence most likely for adsorption on these surfaces in general.84 (Note that those authors did not distinguish between electronic ligand and strain effects, and that therefore the above statement refers to the sum of both of these electronic effects.) A similar independence of the CO – Pt interaction from the number of neighboring Au surface atoms was also derived by Groß et al. in DFT calculations.85 Using Pt(111), Pt0.67Au0.33/Pt(111) and Pt0.33Au0.67/Pt(111) surfaces, they found that going from 0 to 3 to 6 Au neighbors, the CO adsorption energy increased very little, by ~0.1 eV. This agrees also with results reported by Stephens et al.,31 who calculated a small up-shift of 0.09 eV for the dband center for Pt monomers in a PtAu/Pt(111) surface alloy compared to Pt(111). It is important to note that the rather small modification of the CO-Pt bond energy is not in contrast to the strain induced adsorbate bond weakening expected for PtAu/Pt(111) surface ACS Paragon Plus Environment

The Journal of Physical Chemistry

20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

alloys (see above), since the calculated adsorption energy includes both strain effects (adsorbate bond weakening) and (lateral) ligand effects (adsorbate bond strengthening). The same is true also for the experimental finding of a negligible shift of the CO desorption peak in the absence of COad – COad interactions. Hence, for CO adsorption on Pt sites / ensembles in PtAu/Pt(111) surface alloys, the counteracting (lateral) ligand effect and the strain effect almost cancel out. These ideas fully agree with the trends observed in the present TPD measurements on different PtAu/Pt(111) surface alloys. This is different for the IR data, however, which show a significant shift of the C-O vibration (on-top adsorbed CO) to lower wave numbers with increasing Au surface content, both for low COad coverages as well as for COad saturation (see Fig. 7), which seems to be contradictory to the TPD results and also to the calculated adsorption energies. Most plausibly, this apparent discrepancy is related to the fact that the CO adsorption bond on late transition metals and hence also the CO adsorption energy is related to both electron donation from the CO 5σ orbital, which is predominately located at the carbon atom, to the sp states of Pt (in this case) and back donation of electron density from the Pt 5d-bands into the empty 2π* orbital of CO, where the latter is bonding for the Pt-C bond, but antibonding for the C-O bond.86 In the d-band model, which uses the d-band center as essential parameter,83 the main focus is on the interaction between the Pt d-band and the CO 2π∗ orbital. Based on DFT calculations, Tsuda and Kawai

87

recently demonstrated that the C-O bond frequency

shows little correlation with the CO adsorption energy on different Pt surface alloys, and that a much better correlation is obtained for the Pt-C bond frequency, which shifts to lower wave numbers with increasing repulsive strain. Hence, in their picture, the shift in C-O wave number is not necessarily a good measure of the change in adsorption energy, since it does not account for changes in the Pt-C bond strength due to charge donation from the CO 5σ ACS Paragon Plus Environment

Page 20 of 43

Page 21 of 43

The Journal of Physical Chemistry

21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Orbital. An influence of strain on metal-C stretch frequency is also known on other metals, e.g., for CO on Cu, where the stretch frequency decreases with increasing strain.88 It should be noted that the situation is significantly different when looking at the (much weaker) adsorption on Au sites and its modification by Pt substrate and surface atoms, compared to CO adsorption on Au(111). Considering the preference of CO for on-top adsorption on Au surfaces, ensemble effects should be negligible. Based on the d-band model,12 electronic ligand and strain effects should both result in a weakening of the Au-CO bond, considering the stronger Au-Pt bond compared to Au-Au and the smaller lattice constant of Pt(111) compared to Au(111). Interestingly, Gao et al.

89

reported a strongly

enhanced binding energy of about 69 kJ mol-1 for CO adsorption on the Au sites of a bulk alloy PdAu(100) surface and interpreted this finding in terms of a ligand effect.

Comparison with similar type surface alloys. Next we compare our results with findings for the closely related cases of CO adsorption on PdAu/Pd(111),25 PtAg/Pt(111),36 and PdAg/Pd(111)

37;38

surface alloys and thin PdAu films (10 ML) on Mo(110).34;35 These

surface alloys differ substantially in the distribution of surface atoms, with a more or less pronounced phase separation for PtAu/Pt(111),30 and PtAg/Pt(111),90 a disperse distribution for PdAu/Pd(111),25;91 and a distribution changing from disperse at low Ag contents to phase separated clustering at high Ag contents for PdAg/Pd(111).33 In spite of these differences, comparison with the present adsorption system is instructive. First of all, the Pt(111)-based surface alloys differ distinctly from the Pd(111)-based surface alloys by the considerable ensemble effects on the latter surfaces, which originates from the tendency for CO to adsorb on Pd(111) on threefold hollow sites at low COad coverages, and to occupy bridge bonded and on-top sites only at higher coverages.92;93 Accordingly, adsorption of a CO molecule on a Pd dimer surrounded by Au surface atoms was calculated to be much more stable on a bridge site (Ead = 0.87 eV) than in an on-top configuration (Ead = 0.76 eV), ACS Paragon Plus Environment

The Journal of Physical Chemistry

22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

and was also supported experimentally by high resolution electron energy loss spectroscopy (HREELS).25 Equally, CO adsorption on a Pd trimer surrounded by Au surface atoms was calculated to be more stable on a threefold hollow site (Ead = 1.11 eV) than in an on-top configuration (Ead = 0.49 eV). From these findings, Ruff et al. concluded that ensemble effects are dominant for CO interaction with the Pd sites of PdAu/Pd(111), while the contribution of (lateral) ligand plus strain effects is almost negligible.25 Similar trends were concluded also for CO adsorption on PdAg/Pd(111).38 On Pt(111), in contrast, adsorption at low COad coverages is most stable in a linear configuration, on on-top sites. Similarly, on-top adsorption is preferred for PtAu/Pt(111) surface alloys and is the only adsorption geometry detected for CO adsorption after small CO exposures. On Au-rich PtAu/Pt(111) surface alloys with isolated Pt monomers or small Pt ensembles (dimers / trimers), a second on-top site will be occupied upon further CO adsorption, which is energetically less favorable because of COad – COad repulsions, due to direct and indirect (electronic) interactions, but does not involve a simultaneous change in adsorption site of the first COad already present before adsorption of the second CO molecule, as observed, e.g., for CO adsorption on Pd dimers on PdAu/Pd(111) surface alloys. These differences in adsorption behavior between Pd-based and Pt based surface alloys are schematically indicated in Fig. 8. In total, for CO adsorption on Pd (Pt) sites/ensembles, ensemble effects have a pronounced effect on the Pd-based surface alloys, while they are absent for the Pt-based systems. The combined impact of lateral ligand plus strain effects is very small in all cases.

Comparison with inverse surface alloy system PtAu/Au(111). Finally, we want to discuss the present results with respect to the CO adsorption on the inverse system PtAu/Au(111), which has been studied experimentally in more detail by Pedersen et al..39 Besides a much higher tendency of the deposited metal (Pt) atoms to diffuse into the bulk of the host metal ACS Paragon Plus Environment

Page 22 of 43

Page 23 of 43

The Journal of Physical Chemistry

23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(Au), a stabilization of the CO adsorption on Pt atoms in a PtAu overlayer formed on the Au(111) substrate in comparison to CO adsorption on pure Pt(111) was evident. The highest temperature of the TPD peak (25 K higher than for the pure Pt(111) surface) is reached for a Pt content of 1.3 ML, afterwards, the temperature of the TPD maximum decreases and, for a Pt coverage of 2.5 ML, the desorption occurs at a similar temperature as on a pure Pt(111) surface. The authors attributed this behavior to the upward shift of the d-band center caused by the expansive strain resulting from insertion of Pt atoms in the Au lattice. The finding of an increased binding strength of CO on a Pt overlayer on a Au(111) substrate is also in accordance with recent DFT calculations.80-82

Comparison with PtAu nanoparticle catalysts. Tenney et al. analyzed the adsorption of CO on PtAu NPs on TiO2(110), which were prepared by sequential evaporation, using TPD and low energy ion scattering (LEIS).94 Similar to our findings, these authors observed that the CO adsorption on these bimetallic NPs closely resembles that on pure Pt NPs. Adsorption of CO on Au sites was found to be possible at 100 K, desorption occurred immediately at the beginning of the TPD spectra at 110-120 K. The CO oxidation rate of these bimetallic NPs was directly related to the number of Pt sites on the surface. It has to be kept in mind that the as-deposited NPs consist of a Pt core, with a more or less complete Au cover layer. Only for subsequent annealing at higher temperatures, intermixing sets in, but is accompanied by significant particle growth. Upon CO exposure, Pt atoms were found to segregate to the surface, stabilized by the strong Pt-CO bond. Auten et al. studied the adsorption of CO on pure Pt NPs and on bimetallic PtAu NPs (Au:Pt= 1:1), which were supported on different dispersed metal oxide substrates, by IR spectroscopy.95 Comparing the IR band positions at COad saturation coverages, they observed a red-shift of about 10 cm-1 going from pure Pt NPs to the bimetallic NP. This agrees excellently with our findings on planar PtAu/Pt(111) surface alloys (cf. inset of Fig. 7). In ACS Paragon Plus Environment

The Journal of Physical Chemistry

24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

addition, they observed a small peak related to CO adsorbed on Au at room temperature (~2113-2119 cm-1, atmospheric pressure), which disappeared upon removing CO from the gas phase or warming up to 373 K, which equally fits well to our observations. Mott and co-workers investigated the interaction of CO with PtAu/SiO2 supported catalysts with different metal loadings, which were prepared by deposition of preformed PtAu NPs on a SiO2 support followed by controlled annealing (particle size between 2.2 nm and 4.8 nm for pure Au and Pt NPs, 1.8 nm for the alloy particles).76 CO adsorption on PtAu NPs with more than 65% Au content showed IR bands at 2115 cm-1 (Au related ) and at 2060-2096 cm-1 (Ptrelated) in 4% CO in N2. The Pt band was found to shift to higher wave numbers with increasing Pt content, from ~2060 cm-1 for Au-rich to 2096 cm-1 for pure Pt NPs. Also these findings agree fully with the results for the present study.

4. Conclusions The interaction of CO with PtAu/Pt(111) monolayer surface alloys of varying composition and with well defined lateral distribution of the two types of surface atoms in the 2D alloy was investigated by a combination of TPD and IR measurements. These lead to the following conclusions: 1. Desorption of CO adsorbed on-top of Pt monomers, which dominates the spectra at lower COad coverage for low Pt surface contents (98% Au), occurs in a peak centered at 460 K. The absence of significant differences in the peak temperature between CO adsorption on Pt monomers on Au-rich surfaces and CO adsorption on Pt(111), at low coverages, illustrates that the influence of electronic ligand effects (including (local) strain effects) due to adjacent Au atoms is negligible for CO adsorption on PtAu/Pt(111) surface alloys. This is explained by an almost complete compensation of these two effects, the electronic ligand effect (adsorption bond strengthening) and the electronic strain effect (adsorption bond

ACS Paragon Plus Environment

Page 24 of 43

Page 25 of 43

The Journal of Physical Chemistry

25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

weakening). Vertical ligand effects can be neglected due to the low concentration of Au atoms in the subsurface layer. 2. The distinct clustering of Pt surface atoms is reflected by the fact that even for the highest Au surface content (98%) the CO TPD peak covers a broad temperature range with a distinct low-temperature shoulder in addition to the peak at 460 K. While the latter peak is due to desorption of CO molecules adsorbed on small Pt ensembles (monomers, dimers etc.) without COad – COad repulsions (1 COad on Pt monomers or dimers), the low temperature shoulder is related to desorption of COad molecules with COad – COad repulsions (second COad molecule on a Pt dimer etc.) This is very different from Pd-based surface alloys (PdAg/Pd(111) and PdAu/Pd(111)), where already at 90% Ag or Au surface content, surface Pd atoms are exclusively present as isolated monomers, with a single CO desorption peak even at COad saturation. 3. Under our experimental conditions, CO adsorption on PtAu/Pt(111) surface alloys is dominated by on-top adsorption on the much stronger binding Pt sites, in particular for medium and higher surface Au contents. Accordingly, ensemble effects play a negligible role for CO adsorption on these surface alloys, in contrast to the corresponding Pd based surface alloys, where ensemble effects dominate the adsorption energetics. 4. CO adsorption on the Pt-based PtAu surface alloy (PtAu/Pt(111)) differs from the Au based analogon (PtAu/Au(111)) by the much higher tendency for intermixing in the latter system, which causes intermixing also in deeper layers and a much less defined structural situation. Furthermore, on PtAu/Au(111) surface alloys, CO adsorption on Pt surface atoms is stabilized as compared to adsorption on Pt(111) due to the coinciding effect of (both vertical and lateral) ligand effects and strain effects, which all lead to stronger adsorption compared to Pt(111) or PtAu/Pt(111).

ACS Paragon Plus Environment

The Journal of Physical Chemistry

26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

5. Different from the negligible contribution of electronic ligand effects (including strain effects) to the CO adsorption energy, they do affect the C-O vibrational frequency, which shows a distinct down-shift with increasing surface Au content, both for low COad coverages (in the absence of COad – COad interactions) and for COad saturation (including COad – COad interactions). This is explained in the way that the CO adsorption energy (Pt – CO interaction) and the C-O vibrational frequency are not strictly correlated. While changes in the C-O vibrational frequency are dominated by back donation of electron charge from the Pt d-band into the empty CO 2π* orbital, the Pt – CO interaction is affected by both donation from the CO 5σ orbital into the Pt sp-band and the back donation as described above. It was proposed, that the Pt-C vibration is a much better measure of the Pt-CO adsorption energy. 6. At 100 K, CO does not adsorb on Au atoms in the surface, only a few defect Au sites are occupied at this temperature (less than 3%). At lower Au surface contents, the COad saturation coverage (on the Pt sites) decreases nearly linearly with increasing Au surface content, steeper than expected for constant COad density per Pt surface atom, which is explained by a hindered formation of high density COad phases on smaller Pt surface aggregates. This is supported also by the preferential loss of low temperature intensity in the TPD spectra. At Au surface contents >40%, the decay is less steep, which we explain by the increasing influence of additional CO uptake at edges of Pt aggregates. The tendency for Pt surface clustering finally is reflected by the fact that even at 98% surface Pt the COad saturation TPD spectra exhibit an extended low temperature desorption range, characteristic for desorption of ‘high coverage COad’ from small Pt aggregates such as dimers, trimers etc., rather than showing the single peak shape expected for desorption from Pt surface monomers, as it is found for PdAu/Pd(111) and PdAg/Pd(111) at very low surface Pd contents.

ACS Paragon Plus Environment

Page 26 of 43

Page 27 of 43

The Journal of Physical Chemistry

27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

7. Quantitative evaluation of the TPD spectra revealed high sticking coefficients for Au-rich surfaces, which provide clear evidence that adsorption takes place by either direct adsorption on Pt-sites or temporary adsorption on Au sites and subsequent spill-over to empty Pt sites. In total, adsorption of CO has demonstrated that on PtAu/Pt(111) surface alloys Au surface atoms act as an almost ideal diluent, without significantly affecting the chemical properties of the Pt surface atoms.

Acknowledgement This work was supported by the ‘Baden-Württemberg Stiftung’ within the ‘Kompetenznetz Funktionale Nanostrukturen‘ and by the Deutsche Forschungsgemeinschaft via Research Group 1613 (Be-1201/19-1). We gratefully acknowledge discussions with A. Bergbreiter (Ulm University, now Zeiss AG, Oberkochen) and H.E. Hoster (Ulm University, now TUM CREATE Centre for Electromobility, Singapore), as well as with A. Gross (Ulm University, Inst. of Theoretical Chemistry).

ACS Paragon Plus Environment

The Journal of Physical Chemistry

28 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Reference List

1.

Sinfelt, J. H. Acc. Chem. Res. 1977, 10, 15.

2.

Sinfelt, J. H. Bimetallic Catalysts: Discoveries, Concepts and Applications, Wiley: NewYork, 1983.

3.

Ponec, V. J. Mol. Catal. 1998, 133, 221.

4.

Ponec, V. Appl. Catal. A 2001, 222, 31.

5.

Alexeev, O. S.; Gates, B. C. Ind. Eng. Chem. Res. 2003, 42, 1571.

6.

Sachtler, W. M. H.; Somorjai, G. A. J. Catal. 1983, 81, 77.

7.

Macleod, N.; Keel, J. M.; Lambert, R. M. Appl. Catal. A 2004, 261, 37.

8.

Ponec, V.; Sachtler, W. M. H. J. Catal. 1972, 24, 250.

9.

Sinfelt, J. H.; Carter, J. J.; Yates, D. J. C. J. Catal. 1972, 24, 283.

10. Sachtler W. M. H., in Handbook of Heterogeneous Catalysis; Ertl, G., Knözinger, H, and Weitkamp, J, Eds. (VCH-Wiley: Weinheim, 1997) pp. 1077-1084. 11. Mavrikakis, M.; Hammer, B.; Nørskov, J. K. Phys. Rev. Lett. 1998, 81, 2819. 12. Hammer, B.; Nørskov, J. K. Adv. Catal. 2000, 45, 71. 13. O´Cinneide, A.; Gault, F. G. J. Catal. 1975, 37, 311. 14. Kulp, C.; Chen, X.; Puschof, A.; Schwamborn, S.; Schuhmann, W.; Bron, M.

ChemPhysChem 2010, 11, 2854. 15. Skelton, D. C.; Wang, H.; Tobin, R. G.; Lambert, D. K.; DiMaggio, C. L.; Fisher, G. B.

J. Phys. Chem. B 1999, 105, 204. 16. Stephan, J. J.; Ponec, V.; Sachtler, W. M. H. Surf. Sci. 1975, 47, 403. 17. Schmid, M.; Biedermann, A.; Stadler, H.; Slama, C.; Varga, P. Appl. Phys. A 1992, 55, 468. 18. Schmid, M.; Stadler, H.; Varga, P. Phys. Rev. Lett. 1993, 70, 1441. 19. Wouda, P. T.; Schmid, M.; Nieuwenhuys, B. E.; Varga, P. Surf. Sci. 1998, 417, 292. 20. Hofer, W. A.; Ritz, G.; Hebenstreit, W.; Schmid, M.; Varga, P.; Redinger, J.; Podloucky, R. Surf. Sci. 1998, 405, L514-L519.

ACS Paragon Plus Environment

Page 28 of 43

Page 29 of 43

The Journal of Physical Chemistry

29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

21. Buatier de Mongeot, F.; Scherer, M.; Gleich, B.; Kopatzki, E.; Behm, R. J. Surf. Sci.

1998, 411, 249. 22. Maroun, F.; Ozanam, F.; Magnussen, O. M.; Behm, R. J. Science 2001, 293, 1811. 23. Schmid M.; Varga, P., in Alloy surfaces and surface alloys; Woodruf, D.P. Ed. (Elsevier; Amsterdam, 2002) Chapter 4, pp. 118-151. 24. Hoster, H. E.; Bergbreiter, A.; Erne, P.; Hager, T.; Rauscher, H.; Behm, R. J. Phys.

Chem. Chem. Phys. 2008, 10, 3812. 25. Ruff, M.; Takehiro, N.; Liu, P.; Nørskov, J. K.; Behm, R. J. ChemPhysChem 2007, 8, 2068. 26. Rauscher, H.; Hager, T.; Diemant, T.; Hoster, H.; Buatier de Mongeot, F.; Behm, R. J.

Surf. Sci. 2007, 601, 4608. 27. Hartmann, H.; Diemant, T.; Bergbreiter, A.; Bansmann, J.; Hoster, H.; Behm, R. J. Surf.

Sci. 2009, 603, 1439. 28. Hartmann, H.; Diemant, T.; Bansmann, J.; Behm, R. J. Surf. Sci. 2009, 603, 1456. 29. Diemant, T.; Bergbreiter, A.; Bansmann, J.; Hoster, H. E.; Behm, R. J. ChemPhysChem

2010, 11, 3123. 30. Bergbreiter, A.; Alves, O. B.; Hoster, H. E. ChemPhysChem 2010, 11, 1505. 31. Stephens, J. A.; Ham, H. C.; Hwang, G. S. J. Phys. Chem. C 2010, 114, 21516. 32. Röder, H.; Schuster, R.; Brune, H.; Kern, K. Phys. Rev. Lett. 1993, 71, 2086. 33. Engstfeld, A. K., Hoster, H. E., and Behm, R. J., submitted for publication. 34. Yi, C.-W.; Luo, K.; Wei, T.; Goodman, D. W. J. Phys. Chem. B 2005, 109, 18535. 35. Wei, T.; Wang, J.; Goodman, D. W. J. Phys. Chem. C 2007, 111, 8781. 36. Paffet, M. T.; Campbell, C. T.; Taylor, T. N. Langmuir 1985, 1, 741. 37. Ma, Y.; Bansmann, J.; Diemant, T.; Behm, R. J. Surf. Sci. 2009, 603, 1046. 38. Ma, Y.; Diemant, T.; Bansmann, J.; Behm, R. J. Phys. Chem. Chem. Phys 2011, 13, 10741. 39. Pedersen, M. O.; Ruban, A.; Stensgard, I.; Nørskov, J. K.; Besenbacher, F. Surf. Sci.

1999, 426, 395. 40. Ertl, G.; Neumann, M.; Streit, K. M. Surf. Sci. 1977, 64, 393. ACS Paragon Plus Environment

The Journal of Physical Chemistry

30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

41. Avery, N. J. Chem. Phys. 1981, 74, 4202. 42. Paffett, M. T.; Gebhardt, S. C.; Windham, R. G.; Koel, B. E. J. Phys. Chem. 1990, 94, 6831. 43. Smentkowski, V. S.; Yates Jr., J. T. J. Vac. Sci. Technol. A 1989, 7, 3325. 44. Davies, P. W.; Quinlan, M. A.; Somorjai, G. A. Surf. Sci. 1982, 121, 290. 45. Cowley, J. M. J. Appl. Phys. 1950, 21, 24. 46. Darling, A. S.; Mintern, R. A.; Chaston, J. C. J. Inst. Metals 1952, 53, 125. 47. Münster, A.; Sagel, K. Z. Phys. Chem. 1960, 23, 415. 48. Chou, T. S.; Perlman, M. L.; Watson, R. E. Phys. Rev. B 1976, 14, 3248. 49. Günther, C.; Vrijmoeth, J.; Hwang, R. Q.; Behm, R. J. Phys. Rev. Lett. 1995, 74, 754. 50. Hayden, B. E.; Bradshaw, A. M. Surf. Sci. 1983, 125, 787. 51. Gland, J. L.; Kollin, E. B. J. Chem. Phys. 1983, 78, 963. 52. McEwen, J.-S.; Payne, S. H.; Kreuzer, H. J.; Kinne, M.; Denecke, R.; Steinrück, H.-P.

Surf. Sci. 2003, 545, 47. 53. Redhead, P. A. Vacuum 1962, 12, 203. 54. Collins, D. W.; Spicer, W. E. Surf. Sci. 1977, 69, 85. 55. McCabe, R. W.; Schmidt, L. D. Surf. Sci. 1977, 65, 189. 56. Lin, T. H.; Somorjai, G. A. Surf. Sci. 1981, 107, 573. 57. Steininger, H.; Lehwald, S.; Ibach, H. Surf. Sci. 1982, 123, 264. 58. Mavrikakis, M.; Stoltze, P.; Nørskov, J. K. Catal. Lett. 2000, 64, 101. 59. Piccolo, L.; Loffreda, D.; Cadete Santos Aires, F. J.; Deranlot, C.; Jugnet, Y.; Sautet, P.; Bertolini, J. C. Surf. Sci. 2004, 566-568, 995. 60. Yim, W.-L.; Nowitzki, T.; Necke, M.; Schnars, H.; Nickut, P.; Biener, J.; Biener, M. M.; Zielasek, V.; Al Shamery, K.; Klüner, T.; Bäumer, M. J. Phys. Chem. C 2007 , 111, 445. 61. Gottfried, J. M.; Schmidt, K. J.; Schroeder, S. L. M.; Christmann, K. Surf. Sci. 2003,

536, 206. 62. Ruggiero, C.; Hollins, P. J. Chem. Soc. Faraday Trans. I 1996, 92, 4829. 63. Ruggiero, C.; Hollins, P. Surf. Sci. 1997, 377-379, 583.

ACS Paragon Plus Environment

Page 30 of 43

Page 31 of 43

The Journal of Physical Chemistry

31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

64. Kim, J.; Samano, E.; Koel, B. E. J. Phys. Chem. B 2006, 110, 17512. 65. Jiang, L. Q.; Koel, B. E.; Falconer, J. L. Surf. Sci. 1992, 273 , 273. 66. Tränkenschuh, B.; Fritsche, N.; Fuhrmann, T.; Papp, C.; Zhu, J. F.; Denecke, R.; Steinruck, H. P. J. Chem. Phys. 2006, 124, 074712-9. 67. Poelsema, B.; Verheij, L.; Comsa, G. Surf. Sci. 1985, 152-153, 496. 68. Kisliuk, P. J. J. Phys. Chem. Solids 1957, 3, 95. 69. Hollins, P.; Pritchard, J. Progr. Surf. Sci. 1985, 19, 275. 70. Olsen, C. W.; Masel, R. I. Surf. Sci. 1988, 201, 444. 71. Schweizer, E.; Persson, B. N. J.; Tüshaus, M.; Hoge, D.; Bradshaw, A. M. Surf. Sci.

1989, 213, 49. 72. Persson, B. N. J. J. Chem. Phys. 1990, 92, 5034. 73. Tüshaus, M.; Berndt, W.; Conrad, H.; Bradshaw, A. M.; Persson, B. N. J. Appl. Phys. A

1990, 51, 91. 74. Kizhakevariam, N.; Jiang, X.; Weaver, M. J. J. Chem. Phys. 1994, 100, 6750. 75. Andersen, M.; Johansson, M.; Chorkendorff, I. J. Phys. Chem. B 2005, 109, 10285. 76. Mott, D.; Luo, J.; Njoki, P. N.; Lin, Y.; Wang, Li.; Zhong, C. J. Catal. Today 2007, 122, 378. 77. Ruggerone, P.; Kley, A.; Scheffler, M. Progr. Surf. Sci. 1997, 54, 331. 78. Meier, D. C.; Bukhtiyarov, V.; Goodman, D. W. J. Phys. Chem. B 2003, 107, 12668. 79. Ruff, M.; Frey, S.; Gleich, B.; Behm, R. J. Appl. Phys. A 1998, 66, S513. 80. Gohda, Y.; Gross, A. J. Electroanal. Chem. 2007, 607, 47. 81. Gohda, Y.; Groß, A. Surf. Sci. 2007, 601, 3702. 82. Groß, A. J. Phys. : Condens. Matter 2009, 21, 084205. 83. Hammer, B.; Morikawa, Y.; Nørskov, J. K. Phys. Rev. Lett. 1996, 76, 2141. 84. Hammer B.; Nørskov, J. K., in Chemisorption and Reactivity on Supported Clusters and

Thin Films, Lambert R.M. and Pacchiomni, G., Eds. (Kluwer Academic Publ., Dordrecht, 1997); pp. 285-351. 85. Gohda, Y. and Groß, A., unpublished work

ACS Paragon Plus Environment

The Journal of Physical Chemistry

32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

86. Blyholder, G. J. Phys. Chem. 1964, 68, 2772. 87. Tsuda, M.; Kasai, H. Phys. Rev. B 2006, 73, 155405. 88. Kampshoff, E.; Hahn, E.; Kern, K. Phys. Rev. Lett. 1994, 73, 704. 89. Gao, F.; Wang, Y.; Goodman, D. W. J. Am. Chem. Soc. 2009, 131, 3734. 90. Rötter, R., Diploma thesis, Ulm University 2009 91. Behm, R. J. Z. Phys. Chem. 2009, 223, 9. 92. Surnev, S.; Sock, M.; Ramsey, M. G.; Netzer, F. P.; Wiklund, M.; Borg, M.; Andersen, J. N. Surf. Sci. 2000, 470, 171. 93. Morkel, M.; Unterhalt, H.; Klüner, T.; Rupprechter, G.; Freund, H.-J. Surf. Sci. 2005,

586, 146. 94. Tenney, S. A.; Ratliff, J. S.; Roberts, C. C.; He, W.; Ammal, S. C.; Heyden, A.; Chen, D. A. J. Phys. Chem. C 2010, 114, 21652. 95. Auten, B. J.; Lang, H.; Chandler, B. D. Appl. Catal. B 2008, 81 , 225.

ACS Paragon Plus Environment

Page 32 of 43

Page 33 of 43

The Journal of Physical Chemistry

33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure captions Fig. 1

Representative atomically resolved STM images (10 nm x 10 nm) of the PtAu/Pt(111) surface alloys with chemical contrast (a) Au30Pt70; (b) Au40Pt60, (c) Au79Pt21 (keeping 1000 K for 30s) (from ref. 30, with permission).

Fig. 2

Sets of CO TPD spectra recorded on (a) pure Pt(111) and (b-h) PtAu/Pt(111) surface alloys with increasing Au content, see values given in each panel (CO exposures at 100 K: 0.16, 0.24, 0.42, 0.60, 0.76, 0.94, 1.72, 3.28, 16.66 L). Heating rate: 4 K s-1. The CO TPD spectrum of pure Pt(111) (dashed line) is included for comparison.

Fig. 3

(a) CO TPD spectra after the same low CO exposure (0.16 L) recorded on PtAu/Pt(111) surface alloys with varying Au fractions. (b) TPD peak temperature at low COad coverage for AuPt/Pt(111) surface alloy with varying Au fraction.

Fig. 4

(a) Peak temperature at COad saturation coverage for different PtAu/Pt(111) surface alloys, (b) COad saturation coverage on different PtAu/Pt(111) surface alloys. The dashed line indicates a linear decrease of the COad saturation coverage with increasing Au surface content.

Fig. 5

(a) CO sticking coefficient vs. CO coverage on Pt(111) and PtAu/Pt(111) surface alloys with different Au surface contents, (b) Initial sticking coefficient s0 for CO adsorption at 100 K on different PtAu/Pt(111) surface alloys. The dashed line indicates a linear decrease of s0 with increasing Au content.

Fig. 6

Infrared spectra from Pt(111) and PtAu/Pt(111) surface alloys with varying Au surface contents (Au content given in each panel); CO exposures at 100 K: 0.60, 0.94, 1.80, 3.28, 9.0, 16.66 L.

ACS Paragon Plus Environment

The Journal of Physical Chemistry

34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fig. 7

IR spectra from COad saturated Pt(111) and PtAu/Pt(111) surface alloys with 6, 14, 35, 59 and 93% surface Au. The inset shows the CO band positions at COad saturation coverage (squares) and at low coverage (triangles) vs. the surface Au content.

Fig. 8

Schematic representation of the differences in CO adsorption behavior for CO adsorption on small Pt ( a), c), d), g), h) ) or Pd ( b), e), f), i), j) ) ensembles in Ptbased and Pd-based MeAu/Me(111) and MeAg/Me(111) surface alloys (Me = Pt or Pd), illustrating the influence of ensemble effects for these systems (Pt: absence of ensemble effects, Pd: presence of ensemble effects). The combined impact of lateral ligand plus strain effects is very small in all cases.

ACS Paragon Plus Environment

Page 34 of 43

Page 35 of 43

The Journal of Physical Chemistry

35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fig. 1 ACS Paragon Plus Environment

M. Eyrich et al.

The Journal of Physical Chemistry

Page 36 of 43

PCO / a.u.

36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a)

Pt(111)

(b)

5 % Au

(c)

20 % Au

(d)

37 % Au

(e)

42 % Au

(f)

63 % Au

(g)

90 % Au x20

(h)

98 % Au x20

300

400

500

300

400

500

600

Temperature / K

Fig. 2 ACS Paragon Plus Environment

M. Eyrich et al.

Page 37 of 43

The Journal of Physical Chemistry

37

PCO / a.u.

Pt 5% Au 20% Au 42% Au 63% Au 90% Au 98% Au

(a)

300

400

500

Temperature / K 480

(b)

475

Temperature / K

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

470 465 460 455 450 0

20

40

60

80

100

Au surface content / %

Fig. 3 ACS Paragon Plus Environment

M. Eyrich et al.

The Journal of Physical Chemistry

Page 38 of 43

38

500

Peak temperature / K

(a)

CO saturation coverage / ML

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

450

400

(b) 0.6

0.4

0.2

0.0 0

20

40

60

80

100

Au surface content / %

Fig. 4 ACS Paragon Plus Environment

M. Eyrich et al.

Page 39 of 43

The Journal of Physical Chemistry

39

1.2

(a)

Pt(111) 20 % Au 42 % Au 63 % Au 90 % Au

s / s0, Pt

1.0 0.8 0.6 0.4 0.2 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

CO coverage / ML

Initial sticking coefficient s0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1.0

(b)

0.8 0.6 0.4 0.2 0.0 0

20

40

60

80

100

Au surface content / %

Fig. 5 ACS Paragon Plus Environment

M. Eyrich et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Intensity / a.u.

The Journal of Physical Chemistry

2100

2000

Page 40 of 43

Pt(111)

7 % Au

14 % Au

35 % Au

59 % Au

93 % Au

1900

2100

2000

1900

Wave number / cm

2100

2000

1900

-1

Fig. 6

ACS Paragon Plus Environment

M. Eyrich et al.

Page 41 of 43

41

wave number / cm

-1

0.20

Pt(111) 0.15

7 % Au Intensity a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

The Journal of Physical Chemistry

0.10

2100 2090 2080 2070

14 % Au

0

20

40

60

80

100

Au surface content / % 35 % Au 0.05

59 % Au 93 % Au 0.00 2140

2120

2100

2080

2060

2040

2020

2000

-1

Wave number / cm

Fig. 7

ACS Paragon Plus Environment

M. Eyrich et al.

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fig. 8

ACS Paragon Plus Environment

Page 42 of 43

M. Eyrich et al.

Page 43 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

38x17mm (300 x 300 DPI)

ACS Paragon Plus Environment