Surface science experiments

The strategy of the surface science approach is to collect structural and kinetic data on well-defined model surfaces, develop structure-activity relationships, and compare the kinetic properties with those of technical surfaces containing the same components. In principle, a model surface is comparable to a technical surface if both surfaces exhibit the same kinetic properties. When properties are comparable, the operation of the model surface may provide insight into how the technical surface works and information to guide the development of new and/or improved catalysts. 

Simple metal single crystals prepared under well-controlled conditions are extensively characterized with respect to their geometric and electronic properties [28, 29]. With techniques such as MBS, one can describe details of reaction mechanisms and obtain reaction probabilities and rate constants for elementary reaction steps, activation barriers for surface processes, and adsorption quantities [30-33], Kinetic data from single crystal experiments may be vastly different from that of technical 

In recent years, studies moved beyond single crystals and a number of groups have fabricated model catalysts with greater complexity to more closely represent the important features of technical materials. For example, metal-support interactions are modified by depositing metal atoms or nanoparticles on metal oxide single crystals [35, 38, 39]. 

In an MBS experiment a pulsed or chopped beam of atoms or molecules is directed at a target surface (e.g., a specific plane of metal single crystal) typically mounted on a substrate heater in a 

From MBS experiments one can determine the reaction probability, Rp, which is expressed as follows  

---

Figure Al.7.11. Schematic diagram of a generic surface science experiment. Particles, such as photons, electrons, or ions, are mcident onto a solid surface, while the particles emitted from the surface are collected and measured by the detector.

The catalysts with the simplest compositions are pure metals, and the metals that have the simplest and most uniform surface stmctures are single crystals. Researchers have done many experiments with metal single crystals in ultrahigh vacuum chambers so that unimpeded beams of particles and radiation can be used to probe them. These surface science experiments have led to fundamental understanding of the stmctures of simple adsorbed species, such as CO, H, and small hydrocarbons, and the mechanisms of their reactions (42) they indicate that catalytic activity is often sensitive to small changes in surface stmcture. For example, paraffin hydrogenolysis reactions take place rapidly on steps and kinks of platinum surfaces but only very slowly on flat planes however, hydrogenation of olefins takes place at approximately the same rate on each kind of surface site. 

CO oxidation catalysis is understood in depth because potential surface contaminants such as carbon or sulfur are burned off under reaction conditions and because the rate of CO oxidation is almost independent of pressure over a wide range. Thus ultrahigh vacuum surface science experiments could be done in conjunction with measurements of reaction kinetics (71). The results show that at very low surface coverages, both reactants are adsorbed randomly on the surface CO is adsorbed intact and O2 is dissociated and adsorbed atomically. When the coverage by CO is more than 1/3 of a monolayer, chemisorption of oxygen is blocked. When CO is adsorbed at somewhat less than a monolayer, oxygen is adsorbed, and the two are present in separate domains. The reaction that forms CO2 on the surface then takes place at the domain boundaries. 

This reaction is catalyzed by iron, and extensive research, including surface science experiments, has led to an understanding of many of the details (72). The adsorption of H2 on iron is fast, and the adsorption of N2 is slow and characterized by a substantial activation energy. N2 and H2 are both dis so datively adsorbed. Adsorption of N2 leads to reconstmction of the iron surface and formation of stmctures called iron nitrides that have depths of several atomic layers with compositions of approximately Fe N. There is a bulk compound Fe N, but it is thermodynamically unstable when the surface stmcture is stable. Adsorbed species such as the intermediates NH and NH2 have been identified spectroscopically. 

It is now assumed that each active site consists of four Pt atoms and the reactivity of 1 g of catalyst is tested under conditions where the rate is first order in oxygen concentration. The flow over the reactor is set to 100 mL min with 21% oxygen, the temperature 500 K, the pressure to 1 bar, and the TOE (turnover frequency per site) per Pt site under the chosen conditions is known from surface science experiments to be 0.001 s . The amount of oxygen converted is considered negligible. 

Figure 3.3. Schematic representation of the adsorption, surface diffusion, and surface reaction steps identified by surface-science experiments on model supported-palladium catalysts [28]. Important conclusions from this work include the preferential dissociation of NO at the edges and defects of the Pd particles, the limited mobility of the resulting Nads and Oads species at low temperatures, and the enhancement in NO dissociation promoted by strongly-bonded nitrogen atoms in the vicinity of edge and defect sites at high adsorbate coverages. (Figure provided by Professor Libuda and reproduced with permission from the American Chemical Society, Copyright 2004).

Surface science experiments and DFT have often been teammates in very successful projects. DFT has been used along with ultra-high-vacuum surface science experiments such as scanning tunneling microscopy (STM), temperature-programmed desorption, X-ray diffraction, and X-ray photoelectron spectroscopy... 

The piezoelectric stepper described in the previous section, the louse, is a somewhat complicated device, which requires substantial effort to make it work. In many surface-science experiments, the actual location on the sample surface does not matter. A one-dimensional stepper is sufficient. In its simplest form, a micrometer, or a fine-pitch lead screw, can make controlled steps of a few micrometers. However, it is extremely difficult for STM, where the range of the z-piezo is typically of the order of O.ljL/m. 

The link between the microscopic description of the reaction dynamics and the macroscopic kinetics that can be measured in a catalytic reactor is a micro-kinetic model. Such a model will start from binding energies and reaction rate constants deduced from surface science experiments on well defined single crystal surfaces and relate this to the macroscopic kinetics of the reaction. 

Electronic structure theory has developed to a point where realistic bond energies and activation barriers can be calculated. Typically the model catalysts used in such calculations are even more idealized than in the surface science experiments (perfect surfaces, ordered overlayers etc.), but the insight into the details of the potential energy surface of the reaction is much greater. 

To conduct ultra-high-vacuum surface science experiments, often a clean surface is... 

Surface science experiments are, for the most part, conducted under ultra-high vacuum conditions (UHV), generally with base pressures in the range of 10 8—10 11 Torr. This is done in order to maintain... 

Unfortunately, no quantitative data on any of these reaction steps exist, which are obtained under strict control of the experimental variables as in usual surface science experiments with metallic surfaces. Micro-kinetic modeling which could support in a quantitative way the picture derived so far has to await such well-defined experiments. 

In typical surface science experiments as presented previously, oxide-supported metal nanoparticles are deposited onto a clean oxide surface by physical vapor deposition. The precursor in this process is metal atoms in the gas phase, which impinge on the surface, diffuse until they eventually get trapped (either at surface defects or by dimer formation), and then act as nuclei for the growth of larger particles. These processes are well understood for ideal model systems under ultrahigh vacuum (UHV) conditions [56, 57]. In contrast, most realistic supported metal catalyst... 

Realistic operating pressures in catalytic reactions are orders of magnitude higher than those used in most surface-science experiments, and the chemical potential of the gas, usually neglected in the UHV experiments, becomes a significant contribution to the free energy of the surface layer. This pressure difference implies that the structures monitored under... 

The innate complexity of practical catalytic systems has lead to trial and error procedures as the common approach for the design of new and more proficient catalysts. Unfortunately, this approach is far from being efficient and does not permit to reach a deep insight into the chemical nature of the catalytic processes. The consequence of this difficulty is a rather limited knowledge about the molecular mechanisms of heterogeneous catalysis. To provide information about catalysis on a molecular scale, surface science experiments on extremely well controlled conditions have been designed and resulted in a new research field in its own. However, even under these extremely controlled conditions it is still very difficult, almost impossible, to obtain precise information about the molecular mechanisms that underlie catalytic processes without an unbiased theoretical guide. The development of new and sophisticated experimental techniques that enable resolution at... 

Nonlinear optical infrared-visible sum frequency generation (IR-vis SFG) is a versatile surface-specific vibrational spectroscopy that meets the requirements mentioned above. SFG provides vibrational spectra of molecules adsorbed on a surface, while the molecules in the gas phase do not produce a signal. Consequently, SFG can be operated in a pressure range from UFIV to ambient conditions and still detects only the adsorbed species. A direct comparison of adsorbate structures under UFIV and elevated pressure is therefore feasible. Furthermore, SFG can be applied to molecules adsorbed on single crystals, thin films, metal foils, and supported nanoparticles (46,116-121) and is thus a promising tool to extend surface science experiments to more realistic conditions. 

Although a surface in an ambient environment may appear clean, in reality the surface is covered with a layer or layers of adsorbed species. These species may be either physically or chemically adsorbed to the surface, but in both instances prevent the study of a truly clean surface. Under a vacuum environment, such surface contaminants can be removed into the gas phase by sputtering the surface with energetic ions. Once the surface contaminants have been removed, ultrahigh vacuum conditions are required to keep the sample clean. For example, at a pressure of 1 X 10 Torr, a sample receives a flux of approximately 5 X 10 " molecules s cm . Assuming a surface density of 10 atoms cm and a sticking probability of 1 for the adsorption of the gas phase species, the sample would be covered by one monolayer of adsorbed species in seconds. Conversely, by working at pressures below 1 x 10 Torr, a sample can be easily maintained free of contaminants for times on the order of hours, the time frame required for most surface science experiments. 

These results suggest that acetylene should also adsorb onto palladium in the presence of a vinylidene monolayer at sufficiently high pressures. As noted above, surface science experiments conclude that benzene is formed from acetylene on Pd(lll) under UHV conditions by an initial reaction between two acetylene molecules to form a C H intermediate [42,43], followed by reaction with a third acety-... 

Assuming hydrogen dissociatively adsorbs on the surface and that the reaction rate is proportional to the hydrogen atom coverage as suggested by surface science experiments, the hydrogen pressure dependence of the acetylene hydrogenation rate can be expressed as ... 

Linic S, Barteau MA (2003) Construction of a reaction coordinate and a microkinetic model for ethylene epoxidation on silver from DFT calculations and surface science experiments. J Catal 214 200... 

Surface science experiments have shown that such intermediates can be formed rather easily [131]. It has also been demonstrated that acetic acid can readily absorb to a Pd surface by reacting with preadsorbed oxygen [132] (Scheme 5.8). 

---