Chemisorption of CO and mechanism of CO oxidation on supported platinum nanoclusters

Kinetic, isotopic, and infrared studies on well-defined dispersed Pt clusters are combined here with first-principle theoretical methods on model cluster surfaces to probe the mechanism and structural requirements for CO oxidation catalysis at conditions typical of its industrial practice. CO oxidation turnover rates and the dynamics and thermodynamics of adsorption-desorption processes on cluster surfaces saturated with chemisorbed CO were measured on 1-20 nm Pt clusters under conditions of strict kinetic control. Turnover rates are proportional to O(2) pressure and inversely proportional to CO pressure, consistent with kinetically relevant irreversible O(2) activation steps on vacant sites present within saturated CO monolayers. These conclusions are consistent with the lack of isotopic scrambling in C(16)O-(18)O(2)-(16)O(2) reactions, and with infrared bands for chemisorbed CO that did not change within a CO pressure range that strongly influenced CO oxidation turnover rates. Density functional theory estimates of rate and equilibrium constants show that the kinetically relevant O(2) activation steps involve direct O(2)* (or O(2)) reactions with CO* to form reactive O*-O-C*=O intermediates that decompose to form CO(2) and chemisorbed O*, instead of unassisted activation steps involving molecular adsorption and subsequent dissociation of O(2). These CO-assisted O(2) dissociation pathways avoid the higher barriers imposed by the spin-forbidden transitions required for unassisted O(2) dissociation on surfaces saturated with chemisorbed CO. Measured rate parameters for CO oxidation were independent of Pt cluster size; these parameters depend on the ratio of rate constants for O(2) reactions with CO* and CO adsorption equilibrium constants, which reflect the respective activation barriers and reaction enthalpies for these two steps. Infrared spectra during isotopic displacement and thermal desorption with (12)CO-(13)CO mixtures showed that the binding, dynamics, and thermodynamics of CO chemisorbed at saturation coverages do not depend on Pt cluster size in a range that strongly affects the coordination of Pt atoms exposed at cluster surfaces. These data and their theoretical and mechanistic interpretations indicate that the remarkable structure insensitivity observed for CO oxidation reactions reflects average CO binding properties that are essentially independent of cluster size. Theoretical estimates of rate and equilibrium constants for surface reactions and CO adsorption show that both parameters increase as the coordination of exposed Pt atoms decreases in Pt(201) cluster surfaces; such compensation dampens but does not eliminate coordination and cluster size effects on measured rate constants. The structural features and intrinsic non-uniformity of cluster surfaces weaken when CO forms saturated monolayers on such surfaces, apparently because surfaces and adsorbates restructure to balance CO surface binding and CO-CO interaction energies.     

Estimation of Surface Structure and Carbon Monoxide Oxidation Site of Shape‐Controlled Pt Nanoparticles

Surface structures of shape‐controlled Pt nanoparticles have been estimated using cyclic voltammetry (CV) and infrared reflection absorption spectroscopy (IRAS). Cubic and cuboctahedral Pt nanoparticles are prepared using a capping polymer. These nanoparticles give CVs similar to those of single crystal electrodes of Pt in sulfuric acid solution. The CV of cubic nanoparticles is similar to that of the Pt(510) [=5(100)–(110)] electrode, while the CV of cuboctahedral nanoparticles is reproduced well with the convolution of Pt(766) [=13(111)–(100)] and Pt(17 1 1) [=9(100)–(111)] electrodes. These results suggest that the planes of the cubic and cuboctahedral nanoparticles are composed of step‐terrace and atomically flat terraces, respectively. Adsorbed carbon monoxide (CO) on the shape‐controlled nanoparticles gives the IR bands that are assigned to on‐top and bridged CO. The band of on‐top CO is deconvoluted to two bands: the higher and the lower frequency bands are assigned to the CO on the plane and the edges of the nanoparticles, respectively. On‐top CO adsorbed on the edges is oxidized at more negative potential than that on the planes. Edge sites of the nanoparticles promote CO oxidation.     

A first-principles investigation of the effect of Pt cluster size on CO and NO oxidation intermediates and energetics

As catalysis research strives toward designing structurally and functionally well-defined catalytic centers containing as few active metal atoms as possible, the importance of understanding the reactivity of small metal clusters, and in particular of systematic comparisons of reaction types and cluster sizes, has grown concomitantly. Here we report density functional theory calculations (GGA-PW91) that probe the relationship between particle size, intermediate structures, and energetics of CO and NO oxidation by molecular and atomic oxygen on Pt(x) clusters (x = 1-5 and 10). The preferred structures, charge distributions, vibrational spectra, and energetics are systematically examined for oxygen (O(2), 2O, and O), CO, CO(2), NO, and NO(2), for CO/NO co-adsorbed with O(2), 2O, and O, and for CO(2)/NO(2) co-adsorbed with O. The binding energies of oxygen, CO, NO, and of the oxidation products CO(2) and NO(2) are all markedly enhanced on Pt(x) compared to Pt(111), and they trend toward the Pt(111) levels as cluster size increases. Because of the strong interaction of both the reactants and products with the Pt(x) clusters, deep energy sinks develop on the potential energy surfaces of the respective oxidation processes, indicating worse reaction energetics than on Pt(111). Thus the smallest Pt clusters are less effective for catalyzing CO and NO oxidation in their original state than bulk Pt. Our results further suggests that oxidation by molecular O(2) is thermodynamically more favourable than by atomic O on Pt(x). Conditions and applications in which the Pt(x) clusters may be effective catalysts are discussed.     

Structures of small Pd–Pt bimetallic clusters by Monte Carlo simulation

Segregation phenomena of Pd–Pt bimetallic clusters with icosahedral and decahedral structures are investigated by using Monte Carlo method based on the second-moment approximation of the tight-binding (TB-SMA) potentials. The simulation results indicate that the Pd atoms generally lie on the surface of the smaller clusters. The three-shell onion-like structures are observed in 55-atom Pd–Pt bimetallic clusters, in which a single Pd atom is located in the center, and the Pt atoms are in the middle shell, while the Pd atoms are enriched on the surface. With the increase of Pd mole fraction in 55-atom Pd–Pt bimetallic clusters, the Pd atoms occupy the vertices of clusters first, then edge and center sites, and finally the interior shell. It is noticed that some decahedral structures can be transformed into the icosahedron-like structure at 300 and 500 K. Comparisons are made with previous experiments and theoretical studies of Pd–Pt bimetallic clusters.     

Binding energy of large icosahedral and cuboctahedral Lennard-Jones clusters

It is widely believed that the lowest energy configurations for small rare gas clusters have icosahedral symmetry. This contrasts with the bulk crystal structures which have cuboctahedral fcc symmetry. It is of interest to understand the transition between this finite and bulk behavior. To model this transition in rare gas clusters we have undertaken optimization studies within the Lennard-Jones pair potential model. Using a combination of Monte Carlo and Partan Search optimization methods, the lowest energy relaxed structures of Lennard-Jones clusters having icosahedral and cuboctahedral symmetry were found. Studies were performed for complete shell clusters ranging in size from one shell having 13 atoms to 14 shells having 10,179 atoms. It was found that the icosahedral structures are lower in energy than the cuboctahedral structures for cluster sizes having 13 shells or fewer. Additional studies were performed using the more accurate Aziz-Chen [HFD-C] pair potential parameterized for argon. The conclusions appear to be relatively insensitive to the form of the potential.     

CO oxidation catalyzed by silver nanoclusters: mechanism and effects of charge

The mechanism of the CO oxidation promoted by a neutral Ag(55) cluster was investigated extensively, using density functional theory calculations. The CO oxidation process catalyzed by anionic and cationic Ag(55) clusters was also studied, to clarify the effects of the charge state. The Eley-Rideal (ER) and Langmuir-Hinshelwood (LH) mechanisms were discussed in detail. Six reaction pathways were found for the Ag(55)-mediated CO oxidation. It was found that the ER mechanism competed with the LH mechanism. The rate-limiting step of the CO oxidation was the reaction of CO with the Ag(55)O species. All of the anionic, neutral, and cationic Ag(55) clusters were able to promote CO oxidation at low temperatures. The present results enrich our understanding of the catalytic oxidation of CO by nano-sized Ag-based catalysts.     

Evolution of the structural stability of large Cu,Ni, Pd,and Ag clusters with size: An analysis within the embedded atom method

A study of the structural stability of clusters made up of a single component has been carried out within the Embedded Atom Method. Perfect icosahedral and cuboctahedral Cu, Ni, Pd, and Ag clusters with up to 5083 atoms have been compared. The icosahedron is found to be the stable structure for small clusters, and a change of structure from icosahedral to cuboctahedral is found as the cluster size increases. A contraction of the interatomic distances results when the cluster size decreases.     

Nanoclusters Au_(19)Pd and Au_(19)Pt Catalyzing CO Oxidation: a Density Functional Study

The gold atoms on the Au20 cluster had been substituted by the palladium and platinum atoms to obtain the doped clusters with more stable geometries as a function of the bind energy and interaction energy in the previous study. Therefore, we investigated the catalytic activities of the Au_(19)Pd and Au_(19)Pt clusters for CO oxidation along the Langmuir-Hinshelwood mechanism. It is found that the coadsorption of CO and O2 on the doped clusters is obviously stronger than on the Au20 cluster, especially on the doped atom, which makes potential energy of the transition state lower than the total energy of the reactants so that it can promote CO oxidation. The reaction on these doped clusters with the heteroatom on the vertex is more difficult. However, the Au_(19)Pd(S) is more prone to catalyzing the CO oxidation, in which the rate-limiting step has the lower energy barrier of 38.84 kJ/mol for this study. Therefore, the single atom can be modified to change the catalytic activity of the cluster for the CO oxidation. Meanwhile, the different sites on the clusters have different strengths of activity for the reaction.     

Studies of surface processes of electrocatalytic reduction of CO_2 on Pt(210), Pt(310) and Pt(510)

Surface processes of CO2 reduction on Pt(210), Pt(310), and Pt(510) electrodes were studied by cyclic voltammetry. Different surface structures of these platinum single crystal electrodes were obtained by various treatment conditions. The experimental results illustrated that the electrocatalytic activity of Pt single crystal electrodes towards CO2 reduction is decreased in an order of Pt(210)>Pt(310)>Pt(510), i.e., with the decrease of (110) step density on well-defined surfaces. When the surfaces were reconstructed due to oxygen adsorption, the catalytic activity of all the three electrodes has been enhanced to a cer- tain extent. Although the activity order remains unchanged, the electrocatalytic activity has been en- hanced more significantly as the density of (110) step sites is more intensive on the Pt single crystal surface. It has revealed that the more open the surface structure is, the more active the Pt single crystal electrode will be, and the easier for the electrode to be transformed into a surface structure that exhib- its higher activity under external inductions. However, the relatively ordered surfaces of Pt single crystal electrode are comparatively stable under the same external inductions. The present study has gained knowledge on the interaction between CO2 and Pt single crystal electrode surfaces at a micro- scopic level, and thrown new insight into understanding the surface processes of electrocatalytic re- duction of CO2.     

Size and charge effects on the binding of CO to late transition metal clusters

We report on the size and charge dependence of the C-O stretching frequency, nu(CO), in complexes of CO with gas phase anionic, neutral, and cationic cobalt clusters (Co(n)CO(-0+)), anionic, neutral, and cationic rhodium clusters (Rh(n)CO(-0+)), and cationic nickel clusters (Ni(n)CO(+)) for n up to 37. We develop models, based on the established vibrational spectroscopy of organometallic carbonyl compounds, to understand how cluster size and charge relate to nu(CO) in these complexes. The dominating factor is the available electron density for backdonation from the metal to the CO pi* orbital. Electrostatic effects play a significant but minor role. For the charged clusters, the size trends are related to the dilution of the charge density at the binding site on the cluster as n increases. At large n, nu(CO) approaches asymptotes that are not the same as found for nu(CO) on the single crystal metal surfaces, reflecting differences between binding sites on medium sized clusters and the more highly coordinated metal surface sites.     

Adsorption energetics of CO on supported Pd nanoparticles as a function of particle size by single crystal microcalorimetry

The heat of adsorption and sticking probability of CO on well-defined Pd nanoparticles were measured as a function of particle size using single crystal adsorption microcalorimetry. Pd particles of different average sizes ranging from 120 to 4900 atoms per particle (or from 1.8 to 8 nm) and Pd(111) were used that were supported on a model in situ grown Fe(3)O(4)/Pt(111) oxide film. To precisely quantify the adsorption energies, the reflectivities of the investigated model surfaces were measured as a function of the thickness of the Fe(3)O(4) oxide layer and the amount of deposited Pd. A substantial decrease of the binding energy of CO was found with decreasing particle size. Initial heat of adsorption obtained on the virtually adsorbate-free surface was observed to be reduced by about 20-40 kJ mol(-1) on the smallest 1.8 nm sized Pd particles as compared to the larger Pd clusters and the extended Pd(111) single crystal surface. This effect is discussed in terms of the size-dependent properties of the Pd nanoparticles. The CO adsorption kinetics indicates a strong enhancement of the adsorbate flux onto the metal particles due to a capture zone effect, which involves trapping of adsorbates on the support and diffusion to metal clusters. The CO adsorption rate was found to be enhanced by a factor of ～8 for the smallest 1.8 nm sized particles and by ～1.4 for the particles of 7-8 nm size.     

Theoretical study of the adsorption and dissociation of oxygen on Pt(111) in the presence of homogeneous electric fields

The effect of homogeneous electric fields on the adsorption energies of atomic and molecular oxygen and the dissociation activation energy of molecular oxygen on Pt(111) were studied by density functional theory (DFT). Positive electric fields, corresponding to positively charged surfaces, reduce the adsorption energies of the oxygen species on Pt(111), whereas negative fields increase the adsorption energies. The magnitude of the energy change for a given field is primarily determined by the static surface dipole moment induced by adsorption. On 10-atom Pt(111) clusters, the adsorption energy of atomic oxygen decreased by ca. 0.25 eV in the presence of a 0.51 V/A (0.01 au) electric field. This energy change, however, is heavily dependent on the number of atoms in the Pt(111) cluster, as the static dipole moment decreases with cluster size. Similar calculations with periodic slab models revealed a change in energy smaller by roughly an order of magnitude relative to the 10-atom cluster results. Calculations with adsorbed molecular oxygen and its transition state for dissociation showed similar behavior. Additionally, substrate relaxation in periodic slab models lowers the static dipole moment and, therefore, the effect of electric field on binding energy. The results presented in this paper indicate that the electrostatic effect of electric fields at fuel cell cathodes may be sufficiently large to influence the oxygen reduction reaction kinetics by increasing the activation energy for dissociation.     

Chemical identification of icosahedral structure for cobalt and nickel clusters

Reactions of bare and hydrogenated cobalt and nickel clusters with ammonia and with water are used to determine cluster geometrical structure. Saturation measurements determine the total number of ammonia binding sites on cluster surfaces. A pattern of minima in the number of such sites is found to correlate with the sequence of closed shells and subshells expected for icosahedral packing in the 50- to 120-atom size range (50- to 200-atom range for hydrogenated clusters). In many cases there are 12 sites at the minima, the number that would be expected for preferred ammonia binding sites on closed (sub)shells of icosahedral clusters. The equilibrium adsorption of a single water molecule provides a sensitive measure of changes in cluster-water binding energy. A pattern of binding energy maxima is found, once again correlating with icosahedral structure, but for clusters having one metal atom more than the closed (sub)shells. In general, hydrogenation enhances the patterns of minima and maxima. These observations are explained in terms of the expected nature of ammonia and water binding to icosahedral clusters.     

CO oxidation over supported Pt clusters at different CO coverage

CO adsorption and oxidation over supported Pt₁₄ with different CO coverage on TiO₂(110) surface were investigated using density functional theory (DFT) calculations and thermodynamic analysis. According to the phase diagram, Pt₁₄/TiO₂(110) and 11CO@Pt₁₄/TiO₂(110) were chosen to represent the low and high CO coverage of Pt clusters, respectively. Our study shows that the high coverage of CO can induce the structural change of supported Pt clusters and weaken the interaction between Pt clusters and TiO₂ support. The CO adsorption and oxidation mechanism depends on the CO coverage, which is determined by the experimental reactant composition, pressure, and temperature. At low CO coverage, the dissociated oxygen is active specie to form CO₂ by reacting with CO. At high coverage, the molecular oxygen can directly react with CO via the formation of OOCO intermediate. Our proposed mechanisms provide useful information for understanding the CO oxidation over Pt clusters with different CO coverage. © 2016 Wiley Periodicals, Inc.     

Adsorbate binding energies for ammonia on cobalt and nickel clusters

Equilibrium reactions of ammonia with cobalt and nickel clusters are analyzed to determine cluster-adsorbate binding energies. The temperature dependence of reaction equilibrium constantsK _eq are measured and ?ΔH ⁰ values obtained from plots of lnK _eq vs 1/T. We find that binding energies generally decrease with increasing ammonia coverage, and that for a given number of NH₃ molecules binding energies increase with increasing cluster size. The pattern of binding energies is found to be consistent with proposed geometrical structures for Co₁₉ and for clusters in the 55-atom size range. Cluster-ammonia binding energies are generally somewhat higher than for bulk metal surfaces, an expected result considering the character of the cluster surface and the nature of the NH₃-metal interaction.     

Cluster chemistry: size-dependent reactivity induced by reverse spill-over

In the present work, the CO oxidation rate on size-selected Pd clusters supported on thin MgO films is investigated in pulsed molecular beam experiments. By varying the cluster coverage independent of the cluster size, we were able to change the ratio of direct and diffusion flux (reverse spill-over) of CO onto the cluster catalyst and thus probe the influence of reverse spill-over on the reaction rate for different cluster sizes (Pd(8) and Pd(30)). The experimental results show that the change in reaction rate as a function of cluster coverage is different for Pd(8) and Pd(30). In order to explain these findings, the CO flux onto the clusters is modeled utilizing the collection zone model for the given experimental conditions. The results indicate that, for small clusters (Pd(8)), the reaction probability of an impinging CO molecule is independent of whether it is supplied by diffusion or direct flux. By contrast, for larger clusters (Pd(30)) a reduced reaction probability is found for CO supplied by reverse spill-over compared to CO supplied by direct flux.     

In situ infrared reflection absorption spectroscopy of carbon monoxide adsorbed on Pt(S)-[n(100)x(110)] electrodes

Structural effects on the adsorption of CO have been studied using infrared reflection absorption spectroscopy (IRAS) on Pt(S)-[n(100)x(110)] surfaces (n = 2, 5, 9) that have densely packed kink atoms in the step. Coverage and potential dependence of the IRAS spectra are scrutinized. On-top and bridge-bonded CO are found on all of the surfaces examined. CO is adsorbed on only kink at low coverage (thetaCO < or = 0.2). Adsorbed CO on kink gives an IR band at lower frequency than that on step. CO is adsorbed on both kink and terrace at 0.3 < or = thetaCO. Water is adsorbed on the terrace of Pt(510) n = 5 and Pt(910) n = 9 at low CO coverage, but water is not found on Pt(210) n = 2 of which the first layer is composed of only kink atoms. It is suggested that coadsorbed water on the terrace enhances the activity for the oxidation of adsorbed CO on the kink remarkably.     

Structure sensitivity in the oxidation of CO on Ir surfaces

We report results on the catalytic oxidation of carbon monoxide (CO) over clean Ir surfaces that are prepared reversibly from the same crystal in situ with different surface morphologies, from planar to nanometer-scale facets of specific crystal orientations and various sizes. Our temperature-programmed desorption (TPD) data show that both planar Ir(210) and faceted Ir(210) are very active for CO oxidation to form CO2. Preadsorbed oxygen promotes the oxidation of CO, whereas high coverages of preadsorbed CO poison the reaction by blocking the surface sites for oxygen adsorption. At low coverages of preadsorbed oxygen (< or = 0.3 ML of O), the temperature Ti for the onset of CO2 desorption decreases with increasing CO coverage. At high coverages of preadsorbed oxygen (> 0.5 ML of O), T(i) is < 330 K and is independent of CO coverage. Moreover, we find clear evidence for structure sensitivity in CO oxidation over clean planar Ir(210) versus that over clean faceted Ir(210): the CO2 desorption rate is sensitive to the surface morphological differences. However, no evidence has been found for size effects in CO oxidation over faceted Ir(210) for average facet size ranging from 5 to 14 nm. Energetically favorable binding sites for O/Ir(210) are characterized using density functional theory (DFT) calculations.     

Nanosized (mu12-Pt)Pd164-xPtx(CO)72(PPh3)20 (x approximately 7) containing Pt-centered four-shell 165-atom Pd-Pt core with unprecedented intershell bridging carbonyl ligands: comparative analysis of icosahedral shell-growth patterns with geometrically related Pd145(CO)x(PEt3)30 (x approximately 60) containing capped three-shell Pd145 core

Presented herein are the preparation and crystallographic/microanalytical/magnetic/spectroscopic characterization of the Pt-centered four-shell 165-atom Pd-Pt cluster, (mu(12)-Pt)Pd(164-x)Pt(x)(CO)(72)(PPh(3))(20) (x approximately 7), 1, that replaces the geometrically related capped three-shell icosahedral Pd(145) cluster, Pd(145)(CO)(x)(PEt(3))(30) (x approximately 60), 2, as the largest crystallographically determined discrete transition metal cluster with direct metal-metal bonding. A detailed comparison of their shell-growth patterns gives rise to important stereochemical implications concerning completely unexpected structural dissimilarities as well as similarities and provides new insight concerning possible synthetic approaches for generation of multi-shell metal clusters. 1 was reproducibly prepared in small yields (<10%) from the reaction of Pd(10)(CO)(12)(PPh(3))(6) with Pt(CO)(2)(PPh(3))(2). Its 165-atom metal-core geometry and 20 PPh(3) and 72 CO ligands were established from a low-temperature (100 K) CCD X-ray diffraction study. The well-determined crystal structure is attributed largely to 1 possessing cubic T(h) (2/m3) site symmetry, which is the highest crystallographic subgroup of the noncrystallographic pseudo-icosahedral I(h) (2/m35) symmetry. The "full" four-shell Pd-Pt anatomy of 1 consists of: (a) shell 1 with the centered (mu(12)-Pt) atom encapsulated by the 12-atom icosahedral Pt(x)Pd(12-x) cage, x = 1.2(3); (b) shell 2 with the 42-atom nu(2) icosahedral Pt(x)Pd(42-x) cage, x = 3.5(5); (c) shell 3 with the anti-Mackay 60-atom semi-regular rhombicosidodecahedral Pt(x)Pd(60-x) cage, x = 2.2(6); (d) shell 4 with the 50-atom nu(2) pentagonal dodecahedral Pd(50) cage. The total number of crystallographically estimated Pt atoms, 8 +/- 3, which was obtained from least-squares (Pt(x)/Pd(1-x))-occupancy analysis of the X-ray data that conclusively revealed the central atom to be pure Pt (occupancy factor, x = 1.00(3)), is fortuitously in agreement with that of 7.6(7) found from an X-ray Pt/Pd microanalysis (WDS spectrometer) on three crystals of 1. Our utilization of this site-occupancy (Pt(x)Pd(1-x))-analysis for shells 1-3 originated from the microanalytical results; otherwise, the presumed metal-core composition would have been (mu(12)-Pt)Pd(164). [Alternatively, the (mu(12)-Pt)M(164) core-geometry of 1 may be viewed as a pseudo-Ih Pt-centered six-shell successive nu(1) polyhedral system, each with radially equivalent vertex atoms: Pt@M(12)(icosahedron)@M(30)(icosidodecahedron)@M(12)(icosahedron)@M(60)(rhombicosidodecahedron)@M(30)(icosidodecahedron)@M(20)(pentagonal dodecahedron)]. Completely surprising structural dissimilarities between 1 and 2 are: (1) to date 1 is only reproducibly isolated as a heterometallic Pd-Pt cluster with a central Pt instead of Pd atom; (2) the 50 atoms comprising the outer fourth nu(2) pentagonal dodecahedral shell in 1 are less than the 60 atoms of the inner third shell in 1, in contradistinction to shell-by-shell growth processes in all other known shell-based structures; (3) the 10 fewer PR3 ligands in 1 necessitate larger bulky PPh(3) ligands to protect the Pd-Pt core-geometry; (4) the 72 CO ligands consist of six bridging COs within each of the 12 pentagons in shell 4 that are coordinated to intershell metal atoms. SQUID magnetometry measurements showed a single-crystal sample of 1 to be diamagnetic over the entire temperature range of 10-300 K.     

Oxygen adsorption on small PtNi nanoalloys

The adsorption of an oxygen molecule on nanoclusters of Pt and PtNi, in the size range between 13 and 55 atoms, has been studied using first-principle simulations. The structures have been obtained as a function of size and chemical composition of the clusters by means of the parallel excitable-walkers basin hopping method. O(2) preferentially adsorbs along the edge between two (111) facets due to a massive distortion of the Pt-Pt bond length. This bond elongation favours the adsorption in such a way that the binding energy of oxygen on a pure 55-atom cluster is still twice the value on the clean Pt(111). On the other hand, on 55-Pt(shell)Ni(core) nanoparticles, the O(2) binding energy is slightly lower than on Pt(111), because nickel core inhibits the stretching of the Pt-bond because of their size mismatch. However, as soon as its concentration is increased, Ni appears at the surface and its oxyphilic nature contributes to bind the oxygen molecule stronger.