The energies of silver (Ag) atoms in Ag nanoparticles supported on different cerium and magnesium oxide surfaces, determined from previous calorimetric measurements of metal adsorption energies, were analyzed with respect to particle size. Their stability was found to increase with particle size below 5000 atoms per particle. Silver nanoparticles of any given size below 1000 atoms had much higher stability (30 to 70 kilojoules per mole of silver atoms) on reduced CeO2(111) than on MgO(100). This effect is the result of the very large adhesion energy (approximately 2.3 joules per square meter) of Ag nanoparticles to reduced CeO2(111), which we found to be a result of strong bonding to both defects and CeO2(111) terraces, apparently localized by lattice strain. These results explain the unusual sinter resistance of late transition metal catalysts when supported on ceria.

The adsorption energies and growth morphology of silver on reduced CeO 2 (111) thin films at 300 K have been studied using adsorption microcalorimetry in combination with low energy ion scattering (ISS), Auger electron spectroscopy (AES), X-ray photoelectron spectroscopy (XPS), electron energy loss spectroscopy (EELS), sticking probability measurements, low energy electron diffraction (LEED), and scanning tunneling microscopy (STM). Thin films (4 nm thick) of CeO 2-x (111), with x ) ∼0.1, were grown on a Pt(111) singlecrystal. The AES and ISS signals varied with Ag coverage in a manner indicative of 3D Ag particle growth, with a density of ∼4 × 10 12 particles/cm 2 . The initial heat of adsorption of Ag vapor was ∼200 kJ/mol, which is much lower than the heat of sublimation of Ag (285 kJ/mol). Further reduction of the surface by heating in vacuum (to reach x ) ∼0.2) resulted in a slight increase in the initial heat of adsorption to ∼220 kJ/mol; subsequently, higher heats of adsorption were observed until the heat of sublimation of bulk Ag was reached on both films. This increase was attributed to stronger bonding of Ag particles to oxygen vacancies. The initial adsorption energy of Ag on a thinner film (1 nm thick) of CeO 1.9 (111) was ∼250 kJ/mol, ∼50 kJ/mol higher than the 2, 3, and 4 nm films. This is probably due to an interaction of the Ag particles with the underlying Pt(111) surface. The sticking probability of Ag was measured to be near unity (>0.98) on all these reduced CeO 2 (111) films between 1-4 nm thick at 300 K and at all Ag coverages.

We present calorimetric measurements of the effect of cluster size on the adsorption enthalpy of carbon monoxide on Pd nanoclusters sized from 120 to 4900 Pd atoms per particle, which were grown in situ on a well-ordered Fe 3 O 4 / Pt͑111͒ film. A substantial decrease in the initial heat of adsorption amounting to about 20-40 kJ mol −1 was observed on the smallest Pd nanoparticles as compared to the larger Pd clusters and the extended Pd͑111͒ single-crystal surface. We discuss this effect in terms of the size-dependent properties of the Pd nanoparticles. DOI: 10.1103/PhysRevB.81.241416 PACS number͑s͒: 68.43.Ϫh, 82.65.ϩr This Rapid Communication addresses the question: how does the heat of chemisorption of a molecule change when comparing a metal single-crystal surface with a supported metal nanoparticle, and how does it depend on particle size? This is an exceptionally important question that lies at the very heart of understanding particle size effects in catalysis. 1 The energetics of interaction of gaseous molecules, particularly carbon monoxide, with well-defined metal nanoparticles were previously addressed indirectly in nonisothermal temperature-programed desorption ͑TPD͒ experiments 2 and in isothermal modulated molecular-beam studies, 3 where the adsorption energies were obtained by modeling the desorption process and analyzing the lifetimes of the adsorbate on the surface. However, these indirect methods did not provide a clear trend in the changes in the adsorption energy with the particles size: whereas the TPD studies found a decrease in the adsorption energy by about 10 kJ mol −1 on the 2.5 nm sized Pd particles as compared to the extended single-crystal surfaces, the kinetic model used for analysis of the molecular-beam experiments predicted a pronounced increase in the adsorption energy by about 35 kJ mol −1 on the particles smaller than 1.5 nm.A strategy to overcome the shortcomings of such indirect methods based on the model assumptions is direct calorimetric measurement of adsorption enthalpies. Recently, two types of calorimeters were developed that allow direct adsorption energy measurements on single-crystal surfaces. 4,5 However, many phenomena inherent to dispersed supported catalysts and technically relevant nanoparticle-based materials cannot be addressed on such simplified model systems since they do not reproduce some properties of realistic surfaces such as different particles sizes or interactions between nanoparticles and their support material. Only a limited amount of calorimetric data is available today on dispersed supported metal powder catalysts, 6,7 suffering, however, from a high degree of inhomogeneity in metal particle size distribution and low support homogeneity. A strategy to surmount this shortcoming was the development of welldefined, single-crystal-based model surfaces, consisting of metal nanoparticles deposited on flat thin oxide films. 8 The structural properties of these model systems such as particle size and shape can be controlled and characterized in grea...

The adsorption of Ca on poly(3-hexylthiophene) (P3HT) has been studied by adsorption microcalorimetry, atomic beam/surface scattering, X-ray photoelectron spectroscopy (XPS), low-energy He(+) ion scattering spectroscopy (LEIS), and first-principles calculations. The sticking probability of Ca on P3HT is initially 0.35 and increases to almost unity by 5 ML. A very high initial heat of adsorption in the first 0.02 ML (625-500 kJ/mol) is attributed to the reaction of Ca with defect sites or residual contamination. Between 0.1 and 0.5 ML, there is a high and nearly constant heat of adsorption of 405 kJ/mol, which we ascribe to Ca reacting with subsurface sulfur atoms from the thiophene rings of the polymer. This is supported by the absence of LEIS signal for Ca and the shift of the S 2p XPS binding energy by -2.8 eV for reacted S atoms. The heat of adsorption decreases above 0.6 ML coverage, reaching the sublimation enthalpy of Ca, 178 kJ/mol, by 4 ML. This is attributed to the formation of Ca nanoparticles and eventually a continuous solid Ca film, on top of the polymer. LEIS and XPS measurements, which show only a slow increase of the signals related to solid Ca, support this model. Incoming Ca atoms are subject to a kinetic competition between diffusing into the polymer to react with subsurface thiophene units versus forming or adding to three-dimensional Ca clusters on the surface. At approximately 1.6 ML Ca coverage, Ca atoms have similar probabilities for either process, with the former dominating at lower coverage. Ultimately about 1.6 ML of Ca (1.2 x 10(15) atoms/cm(2)) reacts with S atoms, corresponding to a reacted depth of approximately 3 nm, or nearly five monomer-unit layers. Density-functional theory calculations confirm that the heat of reaction and the shift of the S 2p signal are consistent with Ca abstracting S from the thiophene rings to form small CaS clusters.

The energy of adsorbed hydroxyl on Pt(111) was measured by dosing D2O gas onto oxygen precovered Pt(111) at 150 K while following the heat of reaction with single-crystal adsorption calorimetry. The adsorption of D2O on oxygen precovered Pt(111) is known to produce surface OD (hydroxyl) coadsorbed with molecular D2O in a well-defined structure. The heat of reaction and sticking probability of D2O on Pt(111) were measured as a function of oxygen precoverage and D2O dose. With 0.25 monolayers (ML) of oxygen atoms, the differential heat of adsorption is nearly constant at 61.3 kJ/mol for the first ∼1/3 ML but drops to 57.9 kJ/mol by 0.50 ML and 50.5 kJ/mol by saturation (0.62 ML). Similar experiments with Oad precoverages of 0.18 and 0.07 ML gave lower saturation D2O coverages (0.55 and 0.22 ML, respectively) and lower heats of reaction by ∼3.4 and ∼4.6 kJ/mol, respectively, except at very low D2O coverage where step sites may play a role. From the integral heat of D2O adsorption, the standard enthalpy of formation of the (D2O···OD)ad complex is estimated to be −527 kJ/mol. Assuming that the adsorbed D2O molecules in this (D2O···OD)ad complex have the same enthalpy as when adsorbed as a pure D2O film on Pt(111), the enthalpy of formation of adsorbed OD is found to be −226 kJ/mol, and the Pt–OD bond enthalpy to be 263–274 kJ/mol. From these values, we estimate that the reaction D2Oad → ODad + Dad is endothermic by 29–39 kJ/mol. At D2O coverages below 0.4 ML, the sticking probability of D2O on Pt(111) at 150 K is ∼0.9 for oxygen precoverages between 0.14 and 0.25 ML.

Abstract:The adsorption of Ca on the MgO(100) surface at 300 K has been studied using microcalorimetry, in combination with LEED, AES, ISS, work function, sticking probability measurements, and density functional theory (DFT) calculations. The MgO(100) thin films (∼4 nm thick) were grown epitaxially on a 1 µm thick Mo(100) single-crystal. The sticking probability of Ca on MgO(100) at 300 K is unity. On the basis of AES and ISS measurements, it was determined that Ca grows mainly as 3D particles on the MgO(100) surface with a density of ∼1 × 10 12 islands/cm 2 . Ca adsorbs initially at defect sites with a very high heat of adsorption (∼410 kJ/mol). DFT calculations attribute this high initial heat to Ca binding to kink sites (376 kJ/mol), step sites (205 kJ/mol), and lower concentrations of stronger binding sites. The heat of adsorption decreases rapidly with coverage, reaching a minimum of 162 kJ/mol at ∼0.3 ML, where Ca is mainly adding to small 3D Ca clusters. Afterward, it increases to the value of bulk Ca heat of sublimation (178 kJ/mol) at ∼1.2 ML, attributed to the increase in stability with increasing Ca particle size. A 1.0 eV decrease of the work function with Ca coverage from 0 to 0.3 ML indicates that Ca adsorbed at defects is cationic, in agreement with calculations showing that Ca donates electron density to the MgO. Light ion sputtering of the MgO-(100) surface generates point defects, but these do not change the heat of adsorption versus coverage, implying that they do not nucleate Ca particles. Oxygen vacancies are a likely candidate; DFT calculations show that F and F+ center vacancies bind Ca more weakly than terrace sites. More extensive sputtering creates extended defects (such as steps and kinks) that adsorb Ca with heats of adsorption up to ∼400 kJ/mol, similar to that at the intrinsic defect sites.

A new ultrahigh vacuum microcalorimeter for measuring heats of adsorption and adsorption-induced surface reactions on complex single crystal-based model surfaces is described. It has been specifically designed to study the interaction of gaseous molecules with well-defined model catalysts consisting of metal nanoparticles supported on single crystal surfaces or epitaxial thin oxide films grown on single crystals. The detection principle is based on the previously described measurement of the temperature rise upon adsorption of gaseous molecules by use of a pyroelectric polymer ribbon, which is brought into mechanical∕thermal contact with the back side of the thin single crystal. The instrument includes (i) a preparation chamber providing the required equipment to prepare supported model catalysts involving well-defined nanoparticles on clean single crystal surfaces and to characterize them using surface analysis techniques and in situ reflectivity measurements and (ii) the adsorption∕reaction chamber containing a molecular beam, a pyroelectric heat detector, and calibration tools for determining the absolute reactant fluxes and adsorption heats. The molecular beam is produced by a differentially pumped source based on a multichannel array capable of providing variable fluxes of both high and low vapor pressure gaseous molecules in the range of 0.005-1.5 × 10(15) molecules cm(-2) s(-1) and is modulated by means of the computer-controlled chopper with the shortest pulse length of 150 ms. The calorimetric measurements of adsorption and reaction heats can be performed in a broad temperature range from 100 to 300 K. A novel vibrational isolation method for the pyroelectric detector is introduced for the reduction of acoustic noise. The detector shows a pulse-to-pulse standard deviation ≤15 nJ when heat pulses in the range of 190-3600 nJ are applied to the sample surface with a chopped laser. Particularly for CO adsorption on Pt(111), the energy input of 15 nJ (or 120 nJ cm(-2)) corresponds to the detection limit for adsorption of less than 1.5 × 10(12) CO molecules cm(-2) or less than 0.1% of the monolayer coverage (with respect to the 1.5 × 10(15) surface Pt atoms cm(-2)). The absolute accuracy in energy is within ∼7%-9%. As a test of the new calorimeter, the adsorption heats of CO on Pt(111) at different temperatures were measured and compared to previously obtained calorimetric data at 300 K.

The adsorption of atoms and molecules on single crystal surfaces allows one to produce well-characterized atomic, molecular, or dissociated adsorbates. Microcalorimetric measurement of the resulting adsorption energies, i.e., single crystal adsorption calorimetry, allows determination of the standard enthalpies of formation of these adsorbates. Methods are described for making an improved heat detector for such measurements, which greatly improves the signal-to-noise ratio, particularly at low temperatures (down to 100 K). The heat detector is an adaptation of a previously introduced design, based on a metallized pyroelectric polymer (beta-polyvinylidene fluoride), which is pressed against the back of a single crystal during measurement but removed during sample preparation and annealing. The improvement is achieved by selectively etching the metal coating of the polymer, thus reducing the pyro- and piezoelectric noise from all nonessential regions of the polymer. We, furthermore, describe how to achieve a better thermal contact between the sample and the pyroelectric polymer, without increasing the thermal mass of the detector, resulting in significantly improved sensitivities for both 1 and 127 microm thick samples. The result is a detector which, using 1 microm samples, is approximately 40 times more sensitive at 100 K than the traditional polymer-based detector, showing a pulse-to-pulse standard deviation in the heat of adsorption of just 1.3 kJ/mol with gas pulses containing only 1.1% of a monolayer onto Pt(111), for which 1 ML (monolayer) is 1.5x10(15) species/cm(2). For measurements at 300 K, where especially pyroelectric noise is likely of less concern, the new design improves the sensitivity 3.6-fold compared to the traditional detector. These improvements are furthermore used to propose a new detector design that is able to measure heats of adsorption on samples as thick as 127 microm with reasonable sensitivity.