STM and reactivity studies on nanostructures

Get Complete Project Material File(s) Now! »

STM and reactivity studies on nanostructures

In the past, studying reactions on model systems, such as single crystals under UHV conditions, has helped to reveal many fundamental processes, adsorption sites and reaction mechanisms even at the atomic scale [38, 39]. The complexity of research increases drastically with the dimension of the surface reducing from the large domain for a uniform surface (single crystal) to the nanometer domain (small clusters). Many research groups dealing with heterogeneous catalysis are interested in the reactivity of supported nanoparticles. Traditionally, they would study single crystals of transition metals in UHV as model systems or try to bridge the so-called material gap between single crystals and real industry catalysts by implementing metal nanoparticles deposited on high surface area oxide supports. Yet such systems show a broad distribution in both particle size and distance between neighboring particles. This is a problem, since the catalytic activity and selectivity is known to depend on the structure of the catalytic active site [40–43]. Therefore such nanoparticles do not represent a good model for the study of this structural influence. Few catalysis groups have therefore chosen to study reactivity of model uniform nanostructured surfaces prepared by nanolithography [44–49]. However these studies were mainly kinetic studies and did not focus on the mechanism of the reaction. Bobrov and Guillemot [50, 51] have studied the reaction of the Cu(110)-(2×1)O-nanostructured surface with water. Their STM study at low temperature (200 K) revealed stages of the reaction mechanism, yet they did not try to modify the surface and study the influence of the dimensions of the oxidized stripes on the reactivity.
The main technique employed within the present thesis is the scanning tunneling microscopy described in detail in section 2.2.2. STM is a very powerful tool for studying local phenomena in reactivity, since we can image steps and defects and also obtain atomic resolution. In surface science, reactions are classically studied by the means of e.g. x-ray photoelectron spectroscopy (XPS), low-energy electron diffraction (LEED), surface x-ray diffraction (XRD) or secondary ion mass spectrometry (SIMS). However STM stands out from these techniques, since it provides the possibility to study a reaction locally. In spite of the obvious advantage of STM in imaging processes at the atomic level, STM image acquisition during a reaction deals with an obstacle due to surface mobility. Most atoms and many molecules are highly mobile at room temperature, therefore STM studies of reactions often require cooling down to temperatures around 150 K or lower.
In the present thesis, one of few STM studies, in which STM measurement has been performed during a reaction, was carried out. We performed STM measurements during exposure to a reactive gas at a pressure up to 10−6 mbar. The study of the reactivity of a nanostructure at room temperature and in real time conditions has revealed reaction mechanisms at the atomic scale.

Cu(110)-(2×1)O nanostructure

This thesis is dedicated to the study of a self-organized nanostructured surface, the Cu(110)-(2×1)O. This surface serves as a model example for a self-organized template with nanodimensions. The following section summarizes the known information about the system from the literature and finally focuses on the objectives of the present study.

Cu(110)

Copper is a transition metal with a face-centered cubic (fcc) lattice and a lattice constant a = 3.61 Å. In Fig. 1.3, a ball model of the arrangement of atoms on the (110) plane is shown. The primitive surface unit cell has the dimensions a = 3.61 Å and b = 2.55 Å. The (110) surface is the most open and therefore the most reactive of all the low-index surfaces. The clean Cu(110) is thermodynamically stable in this unreconstructed state. The LEED pattern of a clean Cu(110) surface is shown in Fig. 1.4.

Interaction of Cu(110) with oxygen

Oxygen on Cu(110) has been studied extensively since the pioneering work of Ertl [52], revealing that molecular oxygen chemisorbs dissociatively on Cu(110), and that the LEED pattern shows a (2×1) structure at an O coverage of 0.5 monolayers (ML). A LEED pattern of an oxygen-saturated Cu(110) surface is shown in Fig. 1.5. The half- and integer-order spots of the LEED pattern of comparable intensity indicate a surface superstructure. The adsorbed oxygen is located at the long-bridge sites along the [001] rows [53–56], as seen in Fig. 1.6, which shows the rearrangement of the atoms on the (110) face of Cu after the (2×1) oxygen-induced reconstruction.
The reconstruction, previously believed to be of “missing-row” [57–61], or “buckled-row” [62–66] type, was studied and correctly identified for the first time in 1990 in the parallel work of Coulman, Jensen and Kuk [67–69]. The real mechanism is described as a (2×1) “added-row” reconstruction. A detailed STM analysis was needed to identify the mechanism, since the “added-row” model and “missing-row” model are equivalent at saturation coverage (θO = 0.5 ML), but the two differ significantly with respect to the mass transport. While for the “added-row” model, Cu atoms are supplied from step edges, the “missing-row” model would lead to mass transport from terraces to step edges.

The Marchenko-Vanderbilt model

The pioneering theoretical work aimed to understand self-organization has been done by Marchenko [87], Alerhand and Vanderbilt [88]. They proposed elastic continuum models which describe the interplay between two opposing forces: the long-range elastic relaxation energy yielded by the underlying crystal and the domain boundary energy. The former is minimum when two phases A and B are as far apart as possible on the surface and the latter is minimum when they are present as a single domain. The result will be an equilibrium between these two states leading to the formation of a pattern.
A sketch for the stripe phase in elastic continuum model is shown in Fig. 1.10. ω1 and ω2 are the widths of the stripes and P is the periodicity, given by P = ω1 + ω2. The two coexisting phases on a surface have different values of intrinsic surface stress tensor σi, which will induce effective force monopoles Fi, applied to the domain boundaries. These forces will give rise to atomic displacements uj in the whole crystal due to elastic relaxations. The gain in the elastic energy of the surface S is calculated by integrating the product of forces by displacements: 1 Z X FiujdS.

READ  Acquiring academic skills: building blocks and determinants 

Interaction of clean and oxidized Cu(110) with sulfur

Chapter 4 deals with results obtained from sulfidation experiments on the nanostruc-tured Cu(110)-(2×1)O surface. The present section is a summary of studies known from the literature on the interaction of copper and oxidized copper surfaces with sulfur. Examples of the reaction with sulfur on other clean and oxidized metals are discussed as well. Lastly, we focus on the objectives of the present study.
The reaction between sulfur and oxygen on metal surfaces is important for many catalytic processes. For example, it has been observed that the selectivity of carbon-free steam reforming of methane can be obtained by the presence of sulfur on the nickel catalyst [90]. On the other hand, sulfur is also a catalyst poison for the industrially important water-gas shift reaction [91, 92] and a poison for corrosion protection [93, 94]. Therefore studying its reaction on metal surfaces is of high importance.

Sulfur adsorption on clean copper

Different structures in which sulfur adsorbs on the (110) face of copper, depending on the concentration of sulfur, have been established. They are, starting from the lowest concentration: c(2×2), p(5×2) and p(3×2) [95–99]. Other high concentration structures, such as c(8×2), have also been reported [100]. The structures of adsorbed sulfur have been studied by LEED [100, 101], STM [100, 102] and SEXAFS [103]. While the SEXAFS study of Atrei et al. showed that, independently of the sulfur coverage, the adatoms occupy the two-fold hollow site on Cu(110) with the same bond length of 2.37 ± 0.03 Å, the STM data of Parker et al. favored the model of sulfur adsorption in both hollow and bridge sites. A reconciliation came with the work of Carley [99] in the year 2000. In his STM study, Carley has noticed how sulfur present at low concentrations only became visible on STM images after oxygen co-adsorption. The previously highly mobile sulfur adatoms were hindered by the oxygen chains. The mobility of sulfur is especially enhanced along the [1¯10] direction parallel to the copper rows, which provides the lowest diffusion barrier. The inability to image sulfur at low concentrations explains the discrepancy between the previous SEXAFS and STM studies.
In the present study, the two low concentration structures of sulfur on Cu(110) have been observed, the c(2×2) and the p(5×2), and therefore are discussed below in more detail.

Table of contents :

Introduction
Outline of the thesis
1 State of the art and objectives 
1.1 Self-organization
1.2 STM and reactivity studies on nanostructures
1.3 Cu(110)-(2×1)O nanostructure
1.3.1 Cu(110)
1.3.2 Interaction of Cu(110) with oxygen
1.3.3 The Marchenko-Vanderbilt model
1.3.4 Objectives
1.4 Interaction of clean and oxidized Cu(110) with sulfur
1.4.1 Sulfur adsorption on clean copper
1.4.1.1 S-c(2×2) phase
1.4.1.2 S-p(5×2) phase
1.4.2 Sulfur adsorption on oxidized copper
1.4.3 Objectives
1.5 Surface diffusion of large clusters
1.5.1 Objectives
2 Experimental 
2.1 Experimental setup
2.2 Methods
2.2.1 Ultra high vacuum
2.2.2 Scanning tunneling microscopy
2.2.3 Auger electron spectroscopy
2.3 Sample preparation
2.4 Tip preparation
2.5 STM in praxis
2.6 Experimental conditions
2.6.1 Preparation of the Cu(110)-(2×1)O nanostructure: Classical and S co-adsorption method
2.6.2 Sulfidation of the Cu(110)-(2×1)O nanostructure
2.6.2.1 Kinetics by AES
2.6.2.2 Mechanism by STM
3 Tuning the Cu(110)-(2×1)O nanostructure 
3.1 STM of the Cu(110)-(2×1)O surface
3.2 Nanostructures prepared by S co-adsorption
3.3 Influence of step bunching
3.4 Modified Marchenko-Vanderbilt model for the S co-adsorption method
3.5 Summary and discussion
4 Sulfidation of the Cu(110)-(2×1)O nanostructure 
4.1 STM during sulfidation
4.2 Reaction kinetics: sulfidation of the clean and oxidized Cu(110)
4.3 Sulfidation of narrow oxidized stripes
4.4 Sulfidation of wide oxidized stripes
4.4.1 Initial stages of the sulfidation
4.4.2 Influence of the exposure conditions on the mechanism
4.4.2.1 Exposure at high pressures
4.4.2.2 Exposure at low pressures
4.5 Summary and discussion
5 Dynamics of the S-c(2×2) islands 
5.1 Behavior of S islands at sub-saturation sulfur coverages
5.2 Behavior of S islands on a sulfur-saturated surface
5.3 Summary and discussion
Conclusion and perspectives
List of figures
List of tables
References

GET THE COMPLETE PROJECT

Related Posts