Upload
dinhhanh
View
220
Download
1
Embed Size (px)
Citation preview
Ultra-thin silicate films on metal
single crystals vorgelegt von
Master - Naturwissenschaften
Xin Yu
aus Hefei
Von der Fakultät II - Mathematik und Naturwissenschaften
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
Doktor der Naturwissenschaften
Dr. rer. nat.
genehmigte Dissertation
Promotionsausschuss:
Vorsitzender: Prof. Dr. rer. nat. Martin Oestreich
Gutachter: Prof. Dr. rer. nat. Hans-Joachim Freund
Gutachter: Prof. Dr. rer. nat. Reinhard Schomäcker
Tag der wissenschaftlichen Aussprache: 06. September 2013
Berlin 2013
D 83
Diese Arbeit wurde im Zeitraum von September 2009 bis Juli 2013 am
Fritz-Haber-Institut der Max-Planck-Gesellschaft, Berlin, Deutschland,
Abteilung Chemische Physik, unter der Leitung von Herrn Prof. Hans-
Joachim Freund angefertigt.
I
ACKNOWLEDGEMENTS
First and foremost, I would like to express my deeply-felt thanks to my advisor,
Prof. Dr. Hans-Joachim Freund, for giving me the opportunity to perform my PhD
work in his fascinating research group and for his thoughtful guidance and
encouragement.
I would like to thank Prof. Dr. Reinhard Schomäcker (the Technical University
of Berlin) for being my co-supervisor and reviewing my thesis.
I gratefully acknowledge Dr. Shamil Shaikhutdinov for his supervision, advice,
and guidance. He has given me constant encouragement and support in various
ways. Without his wise ideas, this thesis could not reach the present form.
I am also greatly indebted to Dr. Daniel Löffler, who has instructed and helped
me a lot in the beginning of my PhD studies.
Many thanks go in particular to Dr. Jorge Anibal Boscoboinik and Dr. Bing
Yang for their valuable advices in science discussions.
I would like to thank my group members, Dr. Yingna Sun, Yulia Martynova, Bo
Hong Liu, Dr. Irene Groot, Emre Emmez, Martin McBriarty, for their supports in my
work.
I would like to express my appreciations to Uwe Härtel, Matthias Naschitzki,
Klaus-Peter Vogelgesang, Walter Wachsmann and Burkhard Kell, for their technical
assistance.
I am grateful to Manuela Misch, Gabriele Mehnert and Daniela Nikolaus for
their tireless help in administrative works.
II
I am thankful to Bettina Menzel for her cares and supports in IMPRS and
thank Dr. Niklas Nilius, Dr. Tobias Kampfrath and all other people in IMPRS. I have
had a very good time with all of you through seminars, block courses and workshops.
I would like to thank my Chinese friends in our department. They are Dr.
Xiang Shao, Dr. Huifeng Wang, Dr. Yi Pan, Dr. Hengshan Qiu, Dr. Yi Cui, Dr. Yu Lei,
Dr. Xiao Lin, Dr. Kai Huang, Xin Song and Qiushi Pan. Thank you all for helping
with my life in Berlin. I also want to thank all other Chinese friends, especially
Guoxu Zhang, at the Fritz Haber Institute.
I would also like to thank all other colleagues at the department, past and
present, for providing helps and supports with all kinds of things I could not handle
on my own.
Finally my deepest gratitude goes to my beloved parents and sister for their
endless love, persistent encouragement and great confidence in me. They have been
a constant source of inspiration and this work is especially dedicated to them.
III
ABSTRACT
Silica (SiO2) is one of the most widely used and important materials in
heterogeneous catalysis and other technological applications. A thin silicate film
grown on metal single crystal is a suitable model system to study the electronic
properties and surface chemistry of silica based materials at the atomic level using a
variety of surface sensitive techniques.
In this work, we studied the preparation and atomic structure of silica films on
a Ru(0001) substrate. The results show that silica can be grown as either monolayer,
which shares the structure virtually identical with SiO2.5/Mo(112), or a bilayer film.
The bilayer silica film, which is weakly bound to the substrate, can be grown either
as crystalline or amorphous (vitreous) depending on the preparation conditions. In
addition, the film may contain interface oxygen, thus exhibiting “o-rich” and “o-poor”
states.
To understand the role of a metal support in the structure of the silica films,
the growth of silica films on Pt(111) has been studied for comparison. The results
show that the oxyfilicity (oxygen affinity to a metal) of the substrates plays a decisive
role for film structure. The lattice mismatch between a film and a substrate may
affect the film morphology, but to a lesser extent.
The bilayer silica film on Ru(0001) has been employed as a template for
preparation of metal-containing silica films, namely aluminosilicate and iron-silicate
films. It is shown that the aluminosilicate films expose strongly acidic OH-species
upon water adsorption. Iron-containing silica films show a close similarity to ideal
nontronite (Fe-rich smectites). These films represent well-suited model systems for
studying surface chemistry of zeolites and clay minerals at the atomic level.
IV
V
ZUSAMMENFASSUNG
Silika (SiO2) ist eines der meistgenutzten und wichtigsten Materialien, sowohl
auf dem Gebiet der heterogenen Katalyse als auch bei anderen technologischen
Anwendungen. Ein dünner Silikat-Film, aufgebracht auf Metall-Einkristallen, ist ein
geeignetes Modellsystem, um die elektronischen Eigenschaften und die
Oberflächenchemie von silikabasierten Materialien auf der atomaren Ebene mit Hilfe
einer Vielzahl von oberflächensensitiven Techniken zu untersuchen.
In dieser Arbeit untersuchen wir die Präparation und die atomare Struktur
eines Silikafilms auf einem Ru(0001)-Substrat. Die Ergebnisse zeigen, dass Silka
einerseits als Monolage wachsen kann, welches die gleiche Struktur wie
SiO2.5/Mo(112) besitzt, oder andererseits als Doppellagenfilm. Der Doppellagenfilm,
welcher schwach an das Substrat gebunden ist, kann entweder kristallin oder
amorph aufgebaut sein, abhängig von den Präparationsbedingungen. Des Weiteren
kann das System Sauerstoff an der Grenzfläche vom Film zum Metall enthalten, und
daher im O-reichen oder O-armen Zustand vorliegen.
Um den Einfluss des Metallsubstrates auf die Struktur der Silikafilme zu
verstehen, wurde das Wachstum von Silikafilmen auf Pt(111) vergleichend
untersucht. Die Sauerstoff-Affinität des Substrates spielt eine entscheidende Rolle
für die Filmstruktur. Unterschiede der Gitterkoonstanten von Film und Substrat
können ebenfalls die Film-Morphologie geringfügig beeinflussen.
Der Doppellagen-Silikafilm auf Ru(0001) wurde als Grundlage für die
Präparation von, Aluminosilikate und Eisensilikaten verwendet. Aluminosilikat-Filme
weisen nach H2O-Adsorption stark saure OH-Spezies auf. Eisen-dotierte Silikafilme
besitzen Gemeinsamkeiten mit idealem Nontronit (Eisen-reiches Smektit). Diese
Filme stellen geeignete Modellsysteme für die Untersuchung der Chemie von Zeoliten
und Tonmineralien auf atomarer Ebene dar.
VI
Contents
ACKNOWLEDGEMENTS ...................................................................................................... I
ABSTRACT .............................................................................................................................. III
ZUSAMMENFASSUNG ........................................................................................................V
Chapter 1 Introduction ....................................................................................................... 1
Chapter 2 Experimental setup ......................................................................................... 9
2.1 Photoelectron Spectroscopy ................................................................................... 9
2.1.1 X-ray photoelectron spectroscopy .............................................................. 11
2.1.2 Ultraviolet photoelectron spectroscopy ..................................................... 17
2.2 Low Energy Electron Diffraction (LEED) .......................................................... 18
2.3 Scanning Tunneling Microscopy (STM) ............................................................ 21
2.4 Infrared Reflection Adsorption Spectroscopy (IRAS) .................................... 24
Chapter 3 Silica films on Ru(0001) ............................................................................... 31
3.1 Preparation ................................................................................................................ 31
3.2 Monolayer silica film on Ru(0001) ..................................................................... 36
3.3 Bilayer silica film on Ru(0001) ............................................................................ 42
3.3.1 Crystalline and vitreous films ...................................................................... 50
3.3.2 O-rich and O-poor films ................................................................................. 54
3.4 “Thick” silica films on Ru(0001) .......................................................................... 57
3.5 Summary .................................................................................................................... 59
Chapter 4 Silica films on Pt(111) ................................................................................... 61
4.1 Silica films on Pt(111) ............................................................................................ 61
4.2 Support effects ......................................................................................................... 70
4.3 Summary .................................................................................................................... 73
Chapter 5 Metal-containing silicate films on Ru(0001) .......................................... 75
5.1 Aluminosilicate films .............................................................................................. 76
5.2 Iron-silicate films ..................................................................................................... 85
5.3 Summary .................................................................................................................... 99
Chapter 6 Summary and outlook ................................................................................ 101
BIBLIOGRAPHY ................................................................................................................. 105
LIST OF FIGURES ............................................................................................................. 121
ABBREVIATIONS .............................................................................................................. 127
List of Publications: .......................................................................................................... 129
Conference Contributions: ............................................................................................. 131
Curriculum Vitae .............................................................................................................. 133
Erklärung ............................................................................................................................ 135
1
Chapter 1 Introduction
A catalyst is a substance which increases the rate of chemical reaction without
being produced or consumed. It is involved in most chemical processes in the
industry. If the catalyst and the species involved in the reaction (reactants and
products) are in different phases, such as in the case of a reaction happening at the
solid-gas interface, then it is called a heterogeneous catalyst. From now on, when I
talk about catalysis, I will refer to heterogeneous catalysis.
The main steps of a catalytic reaction include: 1) adsorption of the reactants
onto the active site of a catalyst, 2) formation and conversion of various adsorbed
intermediates, 3) desorption of the products from catalyst [1]. Mostly, the process of
a heterogeneous catalytic reaction is rather complicated, bringing difficulties to the
investigation of reaction mechanisms. One approach to study the properties of the
catalyst and the reaction process is building up simplified systems, known as model
catalysts [2-7]. For example, real catalysts expose several types of active site, and it
is therefore hard to study their activity. However, in model systems the nature of
exposed sites can be controlled using different preparation methods, thus allowing
the study of each of them independently [8]. Since the reactions happen on the
surface of the catalyst, it is therefore important to investigate the properties of the
surface and the species adsorbed on it. The field dedicated to these studies is
Surface Science and a large number of tools have been developed over the last
century to study the structure of surfaces and processes happening on it.
The simplest model catalyst is a single crystal surface of an active metal.
However, the catalysts used in the industry are much more complicated than model
systems. To bridge this “material gap”, one approach is to prepare small particles
supported on a flat surface. The most frequently used support materials in industry
2
are highly porous materials, such as SiO2, Al2O3, zeolites and so on, to get large
active surface area to volume ratio [9, 10]. Consequently, a single crystal oxide
surface can be used as the flat support. However, the difficulty of applying surface
science techniques to such a system is caused by the poor electrical and thermal
conductivity of the bulk oxides. Alternatively, a thin oxide layer grown on a metal
substrate can be used as a suitable support for the catalyst particles since, in this
manner, techniques that require conductive samples can be used.
The growth of high quality oxide films can be a complex problem in itself. It is
then necessary to study the growth of these films on metals. Physical vapor
deposition (PVD) is usually employed to grow contaminant-free thin films on metal
substrates in ultra-high vacuum (UHV) conditions.
Silicon dioxide (silica, SiO2) has been intensively studied due to its importance
in many modern technological applications. It is not only an important material as a
support in catalysis, but also plays a critical role in the metal-oxide-semiconductor
field effect transistor (MOSFET), which dominates contemporary integrated circuit
(IC) technology [11]. Therefore, it is necessary to build up a model system to study
the structure and property relationships of silica related materials. Thin silica films
grown on a metal support are suitable model systems to understand the structure
and properties of silica related materials.
Much effort has been made on attempting to grow well-defined silica films on
single crystals in the last two decades. Recently, S. Shaikhutdinov and H. J. Freund
have written a comprehensive review about silica film growth [12]. Below, a brief
survey of these studies will be presented.
A straightforward way of growing ultra-thin silica films is to do it by oxidation
of a Si single crystal surface. The knowledge about the properties of SiO2/Si interface
is crucial for semiconductor device. Numerous studies of SiO2/Si system have been
3
carried out using a variety of surface science techniques [13-28]. These studies
indicate that silica films grown on Si crystals have an amorphous structure and a
transition layer with a thickness varying from half to several nanometers exists at
the SiO2/Si interface. However, the SiO2/Si system is not an ideal model system for
silica related materials due to lack of structural knowledge at the atomic level.
The firstly reported silica film grown on a metal was prepared on Mo(110)
substrate by Xu and Goodman [29, 30]. Later on, the same group reported the
preparation of ultrathin silica films on a Mo(100) surface [31]. In the following years,
silica films have been grown on many metal substrates, Ni(111) [32], Pd(100) [33],
Pd(111) [34], Pt(111) [35, 36]. However, the atomic structure of these films is not
clear.
Silica films grown on Mo(112) substrate have been extensively studied, not
only experimentally but also theoretically, by several groups [37-73]. For the first
time, a well-ordered thin silica film grown on Mo(112) was reported by Schroeder et
al [56-59]. During many years of study, several models were proposed for the
crystalline single layer silica film on Mo(112). Among these, the most convincing one
was given by Freund’s group, after thorough studies using numerous techniques.
The experimental results indicate the formation of a [SiO4] unit [58, 59]; two oxygen
species were found, which can be assigned to Si-O-Si and Si-O-Mo linkages [43]; the
film consists a hexagonal network with a high degree of crystallinity [58, 68]. All the
evidences lead to one conclusion, the thin film consists of a monolayer network of
corner-sharing [SiO4] tetrahedra connected to the substrate via Si-O-Mo linkages [67].
This hypothesis was further supported by density functional theory (DFT)
calculations by Sauer and co-workers [69]. Among several models, the most stable
structure is a monolayer silica film with SiO2.5 stoichiometry, formed via a network of
corner-sharing [SiO4] tetrahedra, where one corner of each tetrahedron downward to
substrate forming a Si-O-Mo bond. (see in Fig 1.1 (a)).
4
For this SiO2.5/Mo(112) system, in addition to the oxygen in the film structure
described before, there is another oxygen species chemisorbed on the Mo(112)
substrate. “O-rich” and “O-poor” states (see in Fig 1.1) can be obtained by tuning the
amount of chemisorbed oxygen [65]. The chemisorbed oxygen atoms are shown in
blue in figure 1.1.
Fig. 1.1 Top and cross views of “O-poor” (a) [69] and “O-rich” (b) [65] SiO2.5/Mo(112)
films.(O: small red, small blue; Si: big orange)
The sub-monolayer silica film grown on Mo(112) forms stripe-like structures
[49]. For the thicker silica film grown on Mo(112), the structure ends up with a
three-dimensional amorphous phase [66].
Other widely used materials in heterogeneous catalysis are zeolites. The latter
are not only important as a supports for active particles but also as catalysts in their
5
own right. Zeolites are natural or synthetic microporous crystalline aluminosilicates,
composed of corner-sharing [TO4] tetrahedra (T = Si, Al). An extreme case is that in
which all T atoms are Si, resulting in neutral framework. However, after introducing
Al into the silica framework, the +3 charge on the Al makes the framework negatively
charged. Therefore, the presence of extraframework cations, such as H+ or alkali
metal cations, is required to keep the overall framework neutral. When the
compensating cation is a proton, as shown in Fig. 1.2, the materials behave as solid
acids [74]. The properties of surface acidity and shape selectivity make zeolites
valuable materials in several areas of chemistry.
Fig. 1. 2 Hydrogen bonded to an O atom bridging Si and Al, in zeolite mordenite. [74]
Mostly, the current understanding of the relationship of structure and surface
chemistry of silicates and related materials come from studies using bulk-sensitive
techniques and from theoretical calculations[75]. Due to their poor electrical and
thermal conductivities, these materials are barely studied using surface-sensitive
techniques. It is therefore useful to prepare suitable model systems for zeolite, using
thin aluminosilicate films on metal single crystals.
6
To the best of our knowledge, the first attempts to prepare ‘surface science’
models of zeolites were reported by Somorjai and co-workers [76]. Many techniques
were used to characterize the films structure, and results indicated that the films are
homogeneous but amorphous.
Well-defined crystalline ultrathin aluminosilicate films were prepared using the
previously mentioned monolayer silica film on Mo(112) as a starting point [77]. The
results show that the film structure is a highly-ordered honeycomb-like network,
basically same as pure silica films. The Al atoms, which substituted the Si atoms,
are randomly distributed in the film. Two possible models were proposed based on
experimental results. One possibility was that the aluminosilicate film on Mo(112)
consists of a monolayer of corner-sharing [TO4]. In this AlO4 model, the Al3+ ions are
each coordinated to four O2- ions forming same structure as [SiO4] in pure silica film.
The extra charge in the film induced by Al3+ ions is compensated by the metal
support. Another possibility is that the Al3+ ions are only coordinated to three O2-
ions in the top-most layer resulting in a AlO3 model. The stability of two models were
basically the same. However scanning tunneling microscopy (STM) image
simulations and X-ray photoelectron spectroscopy (XPS) studies using synchrotron
radiation clearly favored the AlO3 model. Due to the lack of negative framework
charge present in zeolites, this film has no acidic Si-OH-Al species.
The aim of this thesis is to study the properties of silica related materials using
a surface science approach, e.g. investigation of suitable model systems by applying
surface science techniques. As a strategy to study the structure and properties of
silica, which is an insulator, a well-defined thin silica film grown on a metal
substrate is needed. In this work, we have chosen to use a Ru(0001) single crystal as
a substrate for the growth of silica films [78, 79]. To understand the role of metal
supports in the atomic structure of the framework, we have compared the growth of
silica films on Pt(111) with the growth on Mo(112) and Ru(0001) [80]. Furthermore,
7
we have explored the substitution of Si atoms in the framework by other elements,
such as Al and Fe, in order to model other materials of interest, such as zeolites and
clays [81].
The thesis is organized as follows:
a. In Chapter 2, experimental tools and their theoretical backgrounds are
described.
b. Chapter 3 presents the preparation and characterization of silica films
with different thickness on Ru(0001).
c. Chapter 4 addresses the growth of silica films on Pt(111). With
comparison silica/Pt(111), silica/Ru(0001) and silica/Mo(112), the
substrate effect on silica film growth is discussed.
d. Chapter 5 discusses the growth of metal-containing silicate thin films on
Ru(0001), i.e. aluminosilicate and iron-silicate.
e. Summary and outlook are presented in Chapter 6.
8
9
Chapter 2 Experimental setup
The work was mainly done in the UHV system shown in Fig. 2.1, which is
equipped with X-ray photoelectron spectroscopy (XPS), Ultraviolet photoelectron
spectroscopy (UPS), Low energy electron diffraction (LEED), Scanning tunneling
microscopy (STM) and Infrared reflection adsorption spectroscopy (IRAS) for
characterization. In addition, some of the experiments presented in this work were
performed at BESSY II. This chapter briefly describes the techniques used in this
study.
Fig. 2.1 Schematic illustration of the UHV system (MEGA machine).
2.1 Photoelectron Spectroscopy
Due to its unique ability to provide information on the elemental composition,
oxidation state of the elements and especially the dispersion of the different phase
10
within the sample surface up to a depth of several nanometers, photoelectron
spectroscopy (PES) has become one of the most widely used techniques in many
fields, such as the study of heterogeneous catalysis, new materials development and
semiconductor technology.
PES is based on the photoelectric effect which was developed around 1905.
After absorbing a photon of energy, , a core or valence electron with binding energy
is ejected from an atom, and then escapes from the surface(see in Fig 2.2). The
measured kinetic energy, , of the ejected electron is:
2.1
where φ is the work function of the spectrometer. By collecting the intensity of
ejected electrons as a function of energy, one can generate a photoelectron spectrum.
Depending on the energy of the photons used to excite electrons in the sample,
photoelectron spectroscopy can be divided into two main branches, named X-ray
photoelectron spectroscopy (XPS), or so called electron spectroscopy for chemical
analysis (ESCA), and Ultraviolet photoelectron spectroscopy (UPS).
Fig. 2.2 Schematic illustration of the process of for the creation of photoelectron. The
Auger process as one possible decay is included.
11
To interpret the photoemission process, several models have been proposed,
such as three step model and one step model. Three step model is the most widely
used model. As shown in Fig. 2.3, the three steps are: 1) photoexcitation of the
electrons; 2) Travel to the surface with concomitant production of secondaries
(shaded); 3) penetration through the surface (barrier) and escape into the vacuum.
[82]
Fig. 2.3 PES as a stree-step process. [82]
2.1.1 X-ray photoelectron spectroscopy
XPS was first developed by Siegbahn’s group during the 1950s. In 1981
Siegbahn received the Nobel Prize in Physics for his efforts in the field. The
experimental setup for X-ray photoelectron spectroscopy contains two main parts,
12
those being an X-ray source and an electron energy analyzer (see Fig. 2.4). The most
commonly used commercial laboratory X-ray source is a dual anode X-ray gun with
Al and Mg anodes, which provides mildly different radiation energies ( Mg Kα =
1253.6 eV and Al Kα = 1486.3 eV). The availability of two different photos energies
enables the user to unambiguously distinguish Auger peaks from core-lines in
photoelectron spectra. As an alternative to laboratory sources, synchrotron radiation
is also occasionally used. This type of source provides a beam of light with high
intensity and variable energy. Hemispherical analyzers are widely used electron
energy analyzer in photoelectron spectrometers. During measurement a high
vacuum environment is necessary to 1) avoid electron scattering on gas molecules in
the path from sample to analyzer, and 2) avoid attenuation and distortion of the
spectra by surface contamination. [83]
Fig. 2.4 Principle of X-ray photoelectron spectroscopy.
13
Binding energy and chemical shifts
Binding energies are not only element-specific, but also contain information
about the chemical-state of the atoms, which is probed as small shifts in the energy
levels of the core electrons. Such shifts are typically in the range of 0 to 3 eV [84].
The photoemission process begins with an atom containing N electrons, and a
total initial-state energy Ei. The atom is then ionized after absorbing a photon of
energy hν. An electron with kinetic energy escapes from the atom and creates a
hole in the final-state configuration of the atom. The energy for the final-state is Ef.
Since energy must be conserved, we write the energy balance for this process as:
2.2
Such that the measured can be expressed as:
2.3
where is the work function of the spectrometer. Typically this process is
interpreted using the sudden approximation, which assumes that electrons in the
final-state do not have time to redistribute within the time scale of the photoemission
process. However, in reality, such rearrangements often occur upon. Some of the
photoelectrons lose part of their kinetic energy to excitations of other electrons,
leaving the latter electrons in unoccupied orbitals (shake up) or unbound states
(shake off). The resultant decreases photoelectron kinetic energy give rise to satellite
peaks on the higher binding energy side of the main features. Another effect is
multiplet splitting, which occurs when the initial state atom contains unpaired
electron that later interacts with the unpaired core-level electron created by the the
photoelectron process. [83]
14
In photoelectron spectroscopy, Auger peaks are frequently observed because
the Auger process (see in Fig. 2.2) provides one of the most abundant pathways for
final-state decay. In this process, an electron from a higher level fills the vacancy
created by the initially ejected photoelectron. The excess energy may result in the
ejection of a secondary electron, the Auger electron. This process gives rise to an
important technique called Auger electron spectroscopy.
The ability to observe core-level binding energy shifts with chemical-state
variation makes XPS a powerful technique to study the chemical environment of the
surface. For example, when one atom is bonded to another with higher
electronegativity, a charge transfer to the latter atom occurs and the effective charge
of the former becomes positive, thus increasing the binding energy. Therefore, the
binding energy of the metal element in metal oxides is typically higher than in metal.
The physics behind such binding energy shifts can be explained using a charge
potential model, by means of the formular:
∑ 2.4
where E is the biding energy of an electron from an atom i; q is the charge on the
atom; k is a constant; q is the charge on a neighboring atom j; r is the distance
between atom i and atom j; E is a suitable energy reference. In this equation, the
first term (kq ) indicate that with increasing the positive charge on the atom, from
which the photoelectron originates, the binding energy increases. In ionic solids, the
second term counteracts the first due to the charge on a neighboring atom will have
the opposite sign. [85]
In addition to initial state effects (i.e. changes in the core potential due to the
alteration of the local environment), final state effects (for example, polarization of
the core and valence electrons as well as of the surrounding atoms by the core hole
15
created during photoionisation) can also significantly contribute to core level
chemical-shifts. [83]
Intensity
Because a set of binding energies are characteristic for an element, XPS can be
used to analyze the composition of samples. In a homogenous sample, the intensity
of each element can be expressed as:
2.4
where is the intensity of the XPS peak (area); is the x-ray flux on the sample;
is the transmission function of the spectrometer; is the cross-section for
photoemission; is the concentration in number of atoms per unit volume; z is
the depth below the surface; is the inelastic mean free path of the photoelectron;
and is the take-off angle of the photoelectron. Therefore, by comparing the
intensity of different peaks, on can determine the relative atomic concentration of
different components within mixed systems. [85]
Surface sensitive
As a surface sensitive technique, XPS can provide the information about the
top-most layers of a sample. The intensity of a photoelectron at take-off angle with
respect to the surface normal can be expressed by the Beer-Lambert law,
2.5
where the inelastic mean free path is determined by the kinetic energy of the
photoelectron and the nature of the material. Therefore, one application of
synchrotron radiation is doing energy dependent analysis, which allows one to vary
the surface-sensitivity of the technique from several nanometers to single atomic
layers in some cases. As can be seen in the “universal curve” plotted in Fig. 2.5,
16
optimal surface sensitivity is usually obtained for photoelections with kinetic energy
in 25-100 eV range. Another important analysis method is angle dependent XPS, in
which the angle between analyzer and the surface normal is varied. Clearly, the
depth sensitivity can be tuned by changing take-off angle , the larger is, the more
surface information is observed. Therefore, by taking spectra at different θ,
information on the relative depth distribution of the different component can be
obtained.
Fig. 2.5 The mean free path of an electron depends on its kinetic energy, and
determines how much surface information it carries. Optimum surface sensitivity is
obtained with electrons in the 25 to 200 eV range. [86]
XPS Line Shapes
The line shape of photoelectron spectra can be affected by many contributions,
the most important being the natural lifetime and Gaussian broadening. Due to
17
Heisenberg’s uncertainty relationship, the natural line-width is determined by the
life-time of the core-ionized atom. The lifetime is generally on the order of femto-
seconds, resulting in natural widths of 100-200 meV.
There are three factors that contribute to Gaussian broadening: experimental
energy resolution, vibrational and inhomogeneous broadening. The experimental
energy resolution is determined by the analyzer broadening and line width of the X-
ray source. Therefore, it can be improved by lowering the pass energy of the analyzer
and replacing the Al or Mg Kα source, whose line width is about 1 eV, with a
monochromatic X-ray, whose line width is better than 0.3 eV. In crystalline materials,
atomic vibrations impose a small Gaussian broadening, which is temperature
dependent. [83]
2.1.2 Ultraviolet photoelectron spectroscopy
UPS differs from XPS in that UV light is used instead of X-rays to excite the
sample. The most frequently used sources are helium discharge lamps, which
generate light at 21.2 eV (He I) and 40.8 eV (He II). In this case, photoemission is
limited to valence electrons due to the low excitation energies. Therefore, UPS is
particularly well suited to probe bonding in metals, molecules and adsorbed species.
It also provides a quick measure of the macroscopic work function. The high binding
energy cut-off in the spectrum originates from photoelectrons with zero kinetic
energy, = 0. Meanwhile, the electrons form the Fermi-level have the highest kinetic
energy, . Therefore the work function can be expressed as:
2.6
where Δ is the kinetic energy difference between the lowest and highest , i.e. the
width of the spectrum. [85]
18
2.2 Low Energy Electron Diffraction (LEED)
Low energy electron diffraction (LEED) is the principal technique used to
determine the surface structure of single crystal surfaces and ordered adsorbate
overlayers. In 1924 Louis de Broglie introduced wave mechanics and proposed the
wavelike nature of all particles. The de Broglie hypothesis laid the theoretical
foundation of electron diffraction techniques. According to the de Broglie formula:
2.7
where is the wavelength of the electron; = 6.626 × 10-34 , is Plank’s constant;
= 9.11×10-31 , is the mass of an electron; and is the energy of the electron. For
electrons with energy in the range of 20 eV -200 eV, their wavelengths will be in the
range of 2.7 Å -0.87 Å. These wavelengths are comparable with typical atomic
spacings, which is why interaction of such electrons with ordered surfaces should
result in diffraction. Due to the strong interaction between low energy electrons and
the atoms in crystals, the elastic mean free path of the electrons is rather low.
Therefore, LEED gives information about top1~3 layers of a crystal, making it a
surface sensitive technique.
A standard LEED setup consists of a hemispherical fluorescent screen and an
electron gun aligned along the central axis of the screen (see Fig. 2.6). A beam of
monoenergetic electrons emitted from the electron gun impinges up on a grounded
sample and then scatters elastically from the surface. The back-scattered electrons
are accelerated onto the screen, which is biased at 3-6 kV. The resulting diffraction
pattern contains information about the surface structure of the sample. The energy-
filtering grids before screen remove the inelastically scattered electrons. [87]
19
Fig. 2.6 The principle of LEED.
The electrons scattered from a surface give an interference pattern with
constructive interference in the directions for which the difference path lengths of
the electrons are integral multiples of the electron wavelength. Hence, each spot in
the pattern observed on the fluorescent screen corresponds to a direction in which
constructive interference takes place. The diffraction process is shown in Fig. 2.7.
Fig. 2.7 Schematic representation of the diffraction process.
20
The LEED pattern reveals the translational symmetry and the point group of
the surface in reciprocal space [88, 89]. Therefore, the separation between two LEED
spots is larger when interatomic distance is smaller, and vice-versa. The relation
between the real surface lattice and the reciprocal lattice of the sample is
· 2.8
where are the base vectors of the real lattice (i=1,2); are the base vectors of the
reciprocal lattice (j=1,2); is the Kronecker delta; 1 , 0.
The clean non-reconstructed surface gives a rather simple diffraction pattern.
However, an overlayer structure with a rather large unit mesh and various domains
may be more difficult to interpret. Fig. 2.8 gives an example of a c(2×2) overlayer on
an fcc(100) surface.
Fig. 2.8 The diagram of a real space c( 2 × 2 ) structure (left) and the corresponding
diffraction pattern (right).
it can be seen that in real space, b1=b2=√2a1=√2a2, b1*=b2*= a1*= a2*, the vectors
for the adsorbate overlayer are rotated with respect to those of the substrate by 45°.
21
2.3 Scanning Tunneling Microscopy (STM)
Scanning tunneling microscopy (STM) was first developed by Binnig and
Rohrer in 1981, and the 1986 Nobel Prize in physics was awarded to the inventors
for their contributions to this work. Within a few years, STM was quickly established
as one of the most powerful tools for the detailed investigation of surface structure.
STM is based on quantum tunneling, which refers to the quantum mechanical
phenomenon where a particle, described in terms of a wave-function, with a given
energy E tunnels through a potential barrier that it is classically forbidden to cross.
In a one-dimensional model, the quantum mechanical description of an electron with
energy E moving in a potential U(z) is a wave function Ψ(z), That satisfies
Schrödinger’s equation,
ħ
ψ ψ ψ 2.9
with a solution,
ψ ψ 0 2.10
where ψ 0 is the wave function of the surface state of the sample at z=0, and
ħ is the decay constant. The wave function ψ decays exponentially with
distance z.
In an STM experiment, the tunneling current is
∑ |ψ 0 | 2.11
where V is the bias voltage applied between two electrodes. If V is small enough, Eq.
2.11 can be written in terms of the local density of states (LDOS) at the Fermi level,
, ∑ |ψ | 2.12
22
Therefore, the tunneling current can be written as
0, 0, . 2.13
where is the barrier height which has a typical value ~4 eV, and z is in unit on Å.
The tunneling current decays by an order of magnitude when z is changed by ~1 Å.
Electron tunneling can occur from the sample surface to the tip, or vice versa,
depending on the polarity bias. When the tip is biased with a positive voltage and the
sample is at earth potential, electrons from occupied states of the sample tunnel into
the empty states of the tip surface as shown in Fig. 2.9. In this case, the STM images
the occupied states of the surface. Conversely, if the potential is reversed the current
flows in the other direction, and the STM images the unoccupied states. Therefore,
STM does not necessary image atoms, but rather a portion of the density of states
near the Fermi level.
Fig. 2.9 The principle of tunneling between the surface and the metal tip for positive tip
bias V. Tunneling of electrons from occupied surface states into the metal tip. [85]
23
However, the one-dimension model shown above does not adequately describe
the physical reality of the STM tunneling junction. Among several models, the
approach described by Bardeen, which considers the physical nature of the tunnel
junction, is widely used [90]. This model assumes that the wave functions of the
surface states of the tip ψt and sample ψs overlap in order to carry out perturbation
theory. The transition probability from ψt to ψs is called the tunneling matrix
element Mts:
Mћ
dS ψ ψ ψ ψ 2.14
Fermi’s Golden Rule gives the rate for electron transfer across the barrier, and is
written as
ћ
|M | 2.15
where and are the energy of states ψt and ψs, respectively. The implies
tunneling occurs only between electron levels with the same energy. Since electrons
can tunnel in both directions, the tunneling current can be calculate by summing
over all states, as
ћ
∑ , 1 1 · |M | 2.16
where 1 / , and are the energies of the unperturbed wave
functions of tip and sample, respectively. The Fermi functions enter here because
only tunneling from an occupied to an unoccupied state is allowed.
24
Fig. 2.10 Schematic illustration of the working principle of an STM instrument [91].
The typical STM instrumental apparatus is shown in Fig. 2.10. The two
essential electrodes are a sample and a sharp tip mounted on a piezoceramic
scanner, which is the heart of STM. The lower part of the piezoceramic tube can
change the length by applying a voltage Uz thereby tuning the tip in the z direction;
the upper part is surrounded by four equivalent electrodes, which can bend the tube
in different directions by applying suitable voltage. Therefore, the tip can move in all
possible spatial directions. During scanning, the tip is controlled by the piezoceramic
tube using either constant height or constant current mode.
2.4 Infrared Reflection Adsorption Spectroscopy (IRAS)
Infrared (IR) spectroscopy is one of the most frequently used spectroscopic
techniques in catalysis [92-94]. The technique is commonly applied to identify the
adsorbed species and to study the adsorption manner of these species on the surface
of the catalyst. In addition, the infrared spectra of adsorbed probe molecules can
provide valuable information with regards to the chemical properties of the surface.
25
In the following, a simple model will be discussed to understand the molecular
vibrations responsible for the characteristic bands observed in infrared. The
potential energy for the harmonic oscillation of a diatomic molecule in gas phase is
2.17
where k is the force constant of the vibrating bon; r is the distance between the
vibrating atoms; req is the equilibrium distance between the atoms. The vibrational
frequency is
µ 2.18
where µ is the reduced mass, which is obtained from µ
, and are the
mass of the vibrating atoms. It is obvious that when the bond strength increases
and/or the mass of the atoms decreases, the vibrational frequency will decrease.
From quantum mechanical point of view, the molecules can only exist in quantized
energy states, as shown in Fig. 2.11, thus, molecules possess discrete levels of
vibrational energy. The energy of the n-th vibrational level En is
2.19
. The transitions are allowed only when the change of the vibrational quantum
number is Δn = ± 1. [85, 95]
26
Fig. 2.11 Potential energy versus internuclear distance for a diatomic harmonic
oscillator.
The excitation of molecular vibration by light can be described using
perturbation theory. The first order perturbation operator, , is given as the scalar
product of the molecular dipole moment and the electric field component of the
irradiated light:
2.20
using Fermi’s Golden rule, the transition probability is given by:
Ψ ′ | |Ψ ′′ 2.21
where Ψ ′′ and Ψ ′ are the wave function of the initial and final state, respectively,
and is the dipole operator. The expression
′′ ′ Ψ ′ | |Ψ ′′ 2.22
27
then describes the transition dipole moment. Only vibration where the dynamic
dipole moment ( ) changes along the normal coordinate may be excited.
Fig. 2.12 Dipoles oriented parallel and perpendicular to the surface create image dipoles
within the metal surface, which leads to effective screening of dipole moment changes
parallel to the surface but reinforcement of dipole moment changes normal to the
surface. [96]
Infrared reflection absorption spectroscopy (IRAS) is used to study the
vibrations of both molecules and thin films on a metal substrate. An addition rule,
metal surface selection rule (MSSR) applies in IRAS [93, 96, 97]. One reason for this
selection rule is that dipoles located in the proximity of a metal surface create image
dipoles within the metal surface. As shown in Fig. 2.12, when the dipole oriented
parallel to the surface, the image dipole moment is in the opposite orientation
resulting in effective screening of dipole moment changes. Whereas, the dipole
oriented normal to the surface cause a image dipole moment in the same orientation,
thus reinforcement of the dipole moment changes. Another reason is the part of the
IR light with a polarization parallel to the surface (s-polarization) is reflected with a
phase shift of 180°, leading to a complete loss of the electric field parallel to the
surface due to destructive interference. Therefore, vibrations aligned along the
28
surface plane will not give rise to any signal in an IRA spectrum. The phase shift for
p-light is strongly dependent on the angle of incidence φ (see Fig. 2.13) [97], and it
has been demonstrated that a maximal absorption can be achieved with an incident
angle of 83° [98].
Fig. 2.13 Reflection of light at a surface. The electric field components are denoted by
Ep and Es. The component parallel to the surface for p-light (Ep//) is not shown. [97]
The IRA-spectra of adsorbed molecules provide plenty of useful information,
such as molecular symmetry, metal-adsorbate bond, dynamic dipole moment and so
on. Carbon monoxide (CO) is one of the most frequently used probe molecules due to
its simple vibrational and electronic structure. [97]
The vibrational frequency of CO in gas phase is 2143 cm-1. A polarization of the
CO molecule is caused by the electric field which is created by a positive charge,
such as an alkaline metal cation, resulting in the increase of the force constant of C-
O bond and the decrease of the bond length, therefore, the frequency shifts to higher
wavenumers, blue-shift. An upwards shift is also observed when CO interacts with a
Lewis acidic center, which is an acceptor of electron pairs. In this case, the value of
the blue shift increases with the increase of the Lewis acidic strength. [99]
29
The decrease of the CO stretching frequency, red-shift, can occur when CO
adsorbs on metal substrate, and the value of the shift from the vibrational frequency
of CO in gas (2143 cm-1) provides geometric information about the adsorption site.
For instance, the observed frequency in the range of 1800-1900 cm-1 is for the
threefold site, 1900-2000 cm-1 is for the twofold site and 2000-2100 cm-1 is for the
on-top site. [97]
In the case of CO adsorbing on transition cations possessing partially filled d
orbitals, both blue- and red- shift with respect to gas phase can be observed. The
transition cations possess the partially occupied d orbitals, when the empty orbital is
in a suitable energy, the electron can transfer from CO to this empty orbital resulting
in a blue-shift; in another case, back-donation of d-electron from the metal atom into
anti-bonding CO obitals, which stabilizes the metal-molecule bond and weaken s the
C-O bond, leads to a red-shift.
A typical IRAS setup is shown in Fig. 2.14. Fourier-transform (FT)
spectrometers with Michelson interferometers, which measure all of the infrared
frequencies simultaneously by dividing a beam of radiation into two paths and then
recombining the two beams after a path difference has been introduced, result in
interference between the beams, and are used to generate spectra. The components
of a Michelson interferometer include a beam splitter, and a fixed and moving mirror.
The beam from an external source (a SiC globar) incident on an ideal beam splitter
will be divided into two equal intensity beams where 50% is reflected to the fixed
mirror and the other 50% is transmitted to the moving mirror. The light is then
reflected off both mirrors back to the beam splitter where 50% is sent to the detector
and the other half is lost to the source. As the moving mirror scans a defined
distance (x), the path difference between the two beams is varied, and is two times of
the distance traveled by the moving mirror. The interferometer records the
interferogram caused by phase-dependent interference of light with varying optical
30
retardation. Then the spectrometer acquires and digitizes the interferogram and
performs a mathematical function known as a Fourier transform. This function
deconvolutes all individual cosine waves that contribute to the interferogram. The
result of the transformation is a plot of intensity versus frequency or wavelength. [95]
Fig. 2.14 Schematic of the Infrared Reflection-Absorption Spectroscopy setup:
Michelson interferometer, ultrahigh vacuum chamber with sample, mercury cadmium
telluride detector.
31
Chapter 3 Silica films on Ru(0001)
As mentioned in chapter 1, a crystalline ultra thin silica film can be grown on
Mo(112). The film consists of a single layer of corner-sharing [SiO4] tetrahedra
forming a honeycomb-like network, and is strongly bonded to the substrate via Si-O-
Mo linkage resulting a large crosstalk of film and the substrate. Attempt to grow
thicker silica film results in an amorphous silica. In this work, Ru(0001) was chosen
as the substrate based on several considerations. Ru(0001) shares the same
hexagonal symmetry of the monolayer silica film on Mo(112). It has been reported
that a structurally similar hexagonal graphene overlayer grows on Ru(0001) [100]. In
addition, Ru-O bond is much weaker than Mo-O, which is considered as the
limitation of growing a thicker crystalline silica film on Mo(112) [66], therefore it may
give the possibility of growing silica films with different thickness on Ru(0001).
3.1 Preparation
We have tested several preparation recipes of silica films grown on Ru(0001).
Here we describe the method by which the best film quality is achieved. The clean
Ru(0001) surface is obtained in a standard way by repeated cycles of Ar+ sputtering
and high temperature (1300K), annealing in UHV.
Before silicon deposition, the substrate surface is precovered by a 3O(2×2)
overlayer. It was prepared by exposing the clean Ru(0001) surface to 3 × 10-6 mbar
O2 at 1200 K for 5 min and then cooling to 500 K prior to evacuating the oxygen
from the chamber. The XPS result (see fig 3.1) shows that the O1s peak is at ~530 eV,
which is assigned to the chemisorbed oxygen. LEED (see Fig. 3.2 (b)) gives a 2×2
pattern and STM image (see Fig. 3.3) with atomic resolution reveals the surface has
hexagonal structure, which may have a template effect for silica film growth.
32
Fig. 3.1 O1s region in XP-spectrum of the of 3O(2×2) precovered Ru(0001)
Fig. 3.2 LEED pattern of clean Ru(0001) (a); and 3O(2×2) precovered Ru(0001) (b).
33
Fig. 3.3 STM images of 3O(2×2) precovered Ru(0001), V=1.2 V, I=0.12 nA (proved by B.
Yang).
Silicon was evaporated in 2×10-7 mbar O2 by an e-beam assisted evaporator.
During deposition the sample is biased at the same potential as the Si rod to avoid
the uncontrolled defects caused by acceleration of the ions toward the sample. The
sample was held at low temperature, ~100K. The idea of low temperature deposition
is inspired by the growth of CeO2(111) films on Ru(0001): where even the lattice
mismatch between CeO2 film and the Ru substrate is large, ~40%, the two
dimensional structure is favored at low temperature because the adhesion is
improved [101, 102]. It can be explained by the suppression to the diffusivity of the
atoms on the surface at this temperature.
34
Fig. 3. 4 O1s region in XP-spectrum of as deposited film (top), Si2p region in XP-spectra
of as deposited film with different photon energy (bottom).
The (2×2) LEED pattern disappeared after silicon deposition on to the surface.
In Fig 3.4, XPS results show that the 3O(2×2) overlayer is consumed by silicon, and
35
a new broad O1s peak centered at 532.6 eV is formed. The Si is partially oxidized
resulting in a complex of silicon with different oxidation states (see Fig. 3.4).
Applying different incident X-ray energy, 200 eV, 400eV and 1253.6 eV, the Si 2p
peak at high binding energy side attenuates when the photon energy increases,
indicating that the metallic Si is close to the substrate and the Si4+ is on top.
Fig. 3.5 IRA-spectrum of as deposited film.
IRAS (in Fig 3.5) spectra show a broad peak centered at 1249 cm-1. This peak
has been observed by Goodman for the amorphous silica films grown on metals and
assigned to the asymmetric longitudinal-optical phonon mode [103].
The final oxidation step is performed in 3×10-6 O2 mbar at ~1200K for 10
minutes. The oxygen is pumped out after the sample temperature cooled down to
500K.
36
We will present the characterization results of the films following the order of
the film thickness.
3.2 Monolayer silica film on Ru(0001)
In Fig. 3.6 we can see that the Si2p peak is centered at 102.3 eV, which is in
the range for Si4+, indicating the chemical state of Si in the film is 4+ [14, 104, 105].
The Si binding energy in this film is lower than the one in bulk SiO2 and can be
explained by the screening effect from metal substrate to the thin oxide film. It
excludes the formation of Ru silicide which has binding energy ~99.5 eV. This is
further proved by the Ru3d spectrum, where the Ru3d peaks attenuated but no new
features are observed.
In fig 3.6, we can see the O1s contains at least two components, 531.3eV and
529.8eV, and the intensity ratio of these two peaks is about 3:2, which is the same
as for monolayer silica film on Mo(112) where two oxygen species are assigned to the
oxygen atoms bonding to two Si atoms and the oxygen atom in Si-O-Mo linkage [43].
Therefore, similarly, the peak of binding energy at 531.3 eV is assigned to the oxygen
in the Si-O-Si bond and the oxygen atoms with binding energy at 529.8 eV are in the
Si-O-Ru linkage.
37
Fig. 3.6 Si2p region in XP-spectrum of monolayer film (top); O1s region in XP-spectrum
of monolayer film, deconvoluted into two peaks as indicated (bottom).
38
Fig. 3.7 IRA-spectrum of monolayer film.
As shown in Fig. 3.7, the IRAS of this silica film on Ru(0001) shows a sharp
phonon at 1134 cm-1 and less pronounced phonons at 1074 cm-1, 790 cm-1, 687 cm-1.
A similar observation has been found for monolayer silica on Mo(112), which has a
sharp phonon at 1060 cm-1 and two weaker signals at 770 cm-1 and 675 cm-1 [69].
Previously, the 1060 cm-1 phonon is assigned to the stretching vibrations of the Si-O-
Mo linkage. Therefore we suggest the 1134 cm-1 phonon was assigned to the
stretching vibrations of the Si-O-Ru linkage. The frequency shift for the principal
band (i.e., from 1060 to 1134 cm-1) can be explained in the first approximation by
increasing the reduced mass of a Si-O-Ru oscillator as compared to Si-O-Mo.
39
Fig. 3.8 STM images of the monolayer silica film on Ru(0001). V=8 V, I=0.1 nA (a); V=2.0
V, I=0.1 nA (b); The respective LEED pattern (at 60 eV) is shown as an inset in (a).
A large scale STM image of this silica film is present in Fig. 3.8 (a). It shows a
flat film with randomly distributed small holes and particles covering the whole
substrate. The STM images with atomic resolution (see Fig. 3.8 (b)) indicates that the
film consists of multiple domains, all of them having a honeycomb-like network
structure with a 5.4 Å periodicity. In agreement with the LEED result, it shows a
(2×2) pattern with respect to the Ru(0001) substrate. The domains are shifted by
half of the lattice with respect to each other and separated by domain boundaries
imaged as protruding lines. The depth of the randomly distributed holes is 1.4 Å.
(see in fig 3.9).
40
Fig. 3.9 STM image of the monolayer silica film on Ru(0001), V=1.2 V, I=0.1 nA (top);
The profile line is measured along the line indicated in STM (bottom).
Therefore, the combined STM, XPS and IRAS results clearly indicate the
formation of a monolayer silica film on Ru(0001) at low Si coverage. In order to
validate the model, density functional theory (DFT) calculations of the monolayer
silica film on Ru(0001) have been performed by Sauer and co-workers.
41
Fig. 3.10 Top and side views of the most stable structure of Si4O10·2O/Ru(0001) found
by DFT, Si yellow, O red, Ru gray[79].
Fig. 3.11 Caption IRA-spectrum (a) and XP-spectrum (b) simulated for the monolayer,
Si4O10·2O/Ru(0001) structure model (see Fig. 3.10). The frequency in computed IRA-
spectrum is scaled by a factor of 1.0341.15 [78]. The bar height in the XP-spectrum is
proportional to the number of the respective O atoms in the structure. BE shifts are
given with respect to the lowest O1s state.
42
DFT results show that among the several monolayer models studied, only the film
containing two O(Ru) atoms per unit cell (see Fig. 3.10), was found to be stable [79].
Fig. 3.11 shows the harmonic IRA spectrum simulated for this model,
Si4O10·2O/Ru(0001). The most intensive mode at about 1160 cm-1 originates from
the in-phase combination of asymmetric stretching vibrations of the Si–O–Ru
linkages. The mode at 1076 cm-1 with very weak intensity involves combinations of
symmetric O–Si–O stretching vibrations. The bands at 820 and 677 cm-1 are the
combinations of asymmetric stretching of Si–O–Ru linkages and O–Si–O bending
modes. The positions and relative intensities of these calculated bands are in good
agreement with the experimental data provided in Fig. 3.7.
The simulated XP spectrum of the O1s core levels is shown in Fig. 3.11 (b).
The spectrum displays two groups of signals. The signal at higher BEs originates
from oxygen atoms forming Si–O–Si bridges, and the three signals at lower BEs arise
from oxygen atoms bound to the metal substrate, i.e., O(Ru) and those of the Si–O–
Ru linkages in the different (top and hollow) sites with respect to Ru(0001). This
result is in good agreement with the experimental result (Fig. 3.6).
Therefore, the film with silica amount is 1MLE grows as monolayer film,
SiO2.5/Ru(0001). This film consists of a two-dimensional honeycomb-like network of
corner-sharing [SiO4] tetrahedra, with one oxygen atom from each tetrahedron
bonding downward to Ru substrate. The other three oxygen atoms form Si-O-Si
bonds with the neighboring tetrahedra.
3.3 Bilayer silica film on Ru(0001)
In this section, we doubled silicon amount to two monolayer equivalent (2
MLE). After high temperature oxidation, the XPS Si2p spectrum (Fig. 3.12) shows a
peak at 102.5 eV (the full width at half maximum, FWHM, is ~1.3 eV), which
43
indicates that the Si presents only one species and is in the 4+ oxidation state. The
intensity of this peak is two times of the one for silica monolayer on Ru(0001) as
expected. As presented in Fig. 3.12, the XPS O1s spectrum shows an intensive peak
at 531.7 eV with a small shoulder at about 529.9 eV, which has the same binding
energy as chemisorbed 3O(2×2) on Ru(0001). When performing the measurement at
grazing emission angle (80° with respect to the surface normal), the shoulder at
529.9 eV is attenuated more than the peak at 531.7 eV, which suggests that the O
species is underneath the silica film. Therefore the oxygen species at 529.9 eV is in
between the silica film and the substrate, and chemisorbed on Ru substrate (i.e.
interfacial oxygen). Therefore the film has two oxygen components, which can be
assigned to oxygen bonded to silicon atoms (O1s(Si)) and the one bond to Ru
(O1s(Ru)), respectively. The ratio of the integrated areas of these two peaks is about
12:1, which is different from the monolayer silica film on Ru(0001), where the ratio of
the two oxygen peaks is 3:2. Therefore, the oxygen atoms in this film are primarily in
Si-O-Si linkages. In addition, the separation of Si2p and O1s(Si) peaks is ~429.2 eV,
which is closer to the reported value for SiO2 films on Si, e.g. 429.9 eV, rather than
the ones for Si2O3 (432.6 eV) and SiO (432.8 eV) bulk compounds [104]. Therefore,
the stoichiometry of this 2MLE silica film on Ru(0001) is SiO2.
44
Fig. 3.12 Si2p (hν = 200 eV) region in XP-spectrum of bilayer film (top). Normalized XP-
spectra of O1s (hν = 630 eV) region of 2 MLE film at different emission angles (bottom).
The vibrational properties of this film are shown in Fig. 3.16. Two sharp peaks
are observed in IRAS, ~1300 cm-1 and ~690 cm-1, respectively. This differs from the
45
IRAS spectra for SiO2.5/Ru(0001) where the most intensive peak is at 1134 cm-1. To
the best of our knowledge, the 1300 cm-1 peak has not been reported before. To
investigate the structure property given by these phonon vibrations, 18O-labeled films
have been prepared. The IRAS show that the main peak for this film is at 1247 cm-1,
i.e. 53 cm-1 lower than the film oxidized by 16O, and the less intensive band is at 664
cm-1, 28 cm-1 red shift from the 692 cm-1 peak. This results is in good agreement
with the reduced masses of a Si-O-Si oscillator. Since only vibrations with the
dynamic dipole moment normal to the metal surface contribute to an IRAS signal[97],
the very high intensity of the ~1300 cm-1 band (~10%) suggests this band to be
associated with the Si-O-Si bonds oriented perpendicular to the surface.
Fig. 3.13 IRA-spectrum of bilayer films prepared using 32O2 (black), 36O2 (red). The blue
bars show the position and relative intensity of the IR active vibrations calculated by
SFT for the structure shown in Fig. 3.15 (a).
46
Depending on different preparation conditions, LEED shows different patterns,
either (2×2) (with respect to Ru(0001) substrate) pattern or diffuse ring, or both
coexist. In most cases, the LEED give a (2×2) pattern with a weak ring, which infers
the surface consists of hexagonal network with a lattice constant is 5.42 Å (two times
the periodicity of Ru(0001)). More morphological information about the film is
presented by STM results. Large scale images reveal that the substrate is covered by
a flat silica film with a height of ~5 Å. In the atomically resolved STM images, see Fig.
3.14, we can see that the film is composed of a hexagonal network, which shares the
same in plane structure as monolayer silica on Ru(0001).
Fig. 3.14 STM images of the bilayer silica film on Ru(0001). V=1.3 V, I=0.07 nA.
(provided by B. Yang)
47
Fig. 3.15 Structural models of bilayer silica films, “o-poor” (a) and “o-rich” (b).
All evidence leads to one conclusion: the 2MLE silica film contains two silica
layers which are connected together via Si-O-Si, i.e. bilayer silica film. This bilayer
film has no dangling bonds on either side, resulting in a weak interaction with the
Ru(0001) substrate.
This model was confirmed by DFT calculations performed by Sauer and co-
workers [78]. First, the double layer silica sheet is placed over the clean Ru(0001)
surface. The most stable structure is shown in Fig. 3.15. In this model, all O atoms
in the lowest layer are located above the Ru atoms. The results show that the
adhesion energy of the silica film to the Ru(0001) support is 3.1 KJ·mol-1·Å-2.
The calculated phonon spectrum, see in Fig. 3.13, reveals only two IR active
modes above 600 cm-1 measured. The most intense mode at 1296 cm-1 is given by an
48
in-phase combination of asymmetric Si-O-Si stretching vibrations of the Si-O-Si
linkage between the two layers. Another mode at 642 cm-1 is a combination of
symmetric Si-O-Si stretching vibrations of Si-O-Si bonds nearly parallel to the
surface. This result is in good agreement with our experimental result in both the
positions, 1300 vs 1296 cm-1 and 692 vs 642 cm-1, and the relative intensities of the
vibrations [78].
The bilayer silica film can also be obtained with layer by layer growth. First,
one makes a monolayer silica film on Ru(0001), then 1MLE silicon is deposited onto
this monolayer film, followed by high temperature oxidation. In IRA-spectrum, the
1134 cm-1 peak, the feature for monolayer silica film, was not ovserved and the 1300
cm-1 phonon, which is the infrared feature for bilayer silica film, appears. We
suppose that during this process the Si-O-Ru bond breaks to form Si-O-Si bond,
which is thermodynamically unfavorable for monolayer silica on Mo(112). Therefore,
under our condition, the structure of the silica films on Ru(0001) is governed by Si
coverage. This conclusion can be further proved by the growth of silica film with
intermediate silicon coverage, between 1 MLE and 2 MLE.
When the amount of silicon deposited is about 1.5 MLE, the film contains both
monolayer and bilayer structure. In Fig. 3.16, we can see the infrared spectrum has
both 1137 cm-1 band, which is given by Si-O-Ru bond corresponding to a monolayer
silica film, and 1300 cm-1 band, which infers the formation of bilayer silica film. As
shown in Fig. 3.17, LEED shows a (2×2) pattern with a weak ring, and large scale
STM image reveals flat morphology where wide terraces of Ru(0001) can still be
recognized. Large domains within the same terrace are separated by steps with
apparent heights of ~ 1.5 and ~ 5 Å when measured with respect to small holes
exposing the underlying O–Ru surface. The heights of ~ 1.5 and ~ 5 Å can be
assigned to monolayer silica film and a bilayer silica film, respectively.
49
Fig. 3.16 IRA-spectrum of 1.5 MLE film.
Fig. 3.17 STM image of the 1.5 MLE silica film on Ru(0001), V= 8.0 V, I=0.1 nA; The
respective LEED pattern (at 60 eV) is shown as an inset (provided by B. Yang).
50
3.3.1 Crystalline and vitreous films
Depending on different preparation conditions, the LEED pattern of the bilayer
silica films can be either (2×2) spots or a (2×2) diffuse ring, or coexisting of (2×2)
spots and the ring. First, we give an example of a bilayer silica film which has LEED
pattern with coexisting (2×2) spots and ring. The corresponding LEED and STM
images are shown in Fig. 3.18. In the atomically resolved STM image, we can clearly
see the domains of ordered structures, which are identified as crystalline silica film,
coexisting with disordered structures, which could be identified as two-dimensional
vitreous silica.
Fig. 3. 18 STM image of the bilayer silica film prepared by two sequential deposition and
oxidation steps of 1 MLE Si each, V= 2 V, I=0.1 nA The respective LEED pattern (at 60
eV) is shown as an inset (provided by B. Yang).
51
To study the effect of the preparation conditions on the morphology of bilayer
silica film, several films have been prepared under different conditions by varying
annealing temperature, time, oxygen pressure, etc. The results indicate that the rate
of sample cooling after the high-temperature oxidation step is an important factor in
controlling film crystallinity. Fig. 3.19 shows LEED and STM results of two films
prepared using same condition except the cooling rate after oxidation. The film
prepared with slow cooling rate (below 1 K s-1) gives sharp (2×2) LEED pattern
indicating a highly ordered surface. The respective STM image shows a regular
honeycomb-like network formed by six membered rings, which contains six [SiO4]
units. On the contrary, the film obtained by fast cooling rate (~ 5 K s-1), shows a
diffraction ring in addition to the weak (2×2) LEED pattern. The STM image indicates
that the film consists of rings with different size, composed of 4 to 9 [SiO4] units. The
number of the [SiO4] units in a ring is determined by the number of the nearest
neighboring rings. Therefore, slow cooling rate leads to the formation of a crystalline
silica bilayer film and a fast cooling rate results in a vitreous bilayer film. The XPS
and IRAS of crystalline and vitreous films are nearly identical which infers that the
electronic and vibrational properties for both films are the same. This bilayer silica
film opens a playground for studying glass and glass transition in atomic level [106,
107].
52
Fig. 3. 19 (top) LEED patterns (at 60 eV) and STM images (bottom) of the bilayer silica
films prepared by slow (on the left) and fast (on the right) cooling after the high
temperature oxidation step in the film preparation. (Tunneling parameters: 2 V, 0.1 nA
(left); 3.3 V, 0.1 nA (right).)
The crystallinity of the bilayer film is irreversible: the crystalline film, which is
obtained by slow cooling, cannot be transformed into the vitreous state by re-
oxidation of the same sample followed by fast cooling, and vice versa. The film
decomposes when applying higher temperature for re-oxidation. These findings agree
53
well with the very high barrier for crystal-to-glassy transformation predicted by the
DFT calculations [106].
The film quality can also be affected by other preparation factors. In particular,
we found that lower deposition temperatures (ca. 100K) appear to favor the
formation of more closed films. Fig. 3.20 gives the STM images of two films prepared
using same silicon dosage, 1.8 MLE, but different deposition temperature, 300 K and
100 K. XPS results show the same Si and O amount in the oxidized films. Infrared
spectra suggest two films have the same vibrational property. However, an obvious
difference is observed by STM. The film obtained by depositing silicon at low
temperature, 100 K, has less open patches than the one prepared by depositing
silicon at 300 K. This likely implies the formation of more two-dimensional SiOx
overlayers at lower temperatures, which may form as a result of limited surface
diffusivity. When such films are then oxidized at high temperatures, the increased
wetting of the “precursor state” is then presumably reflected in the final state of the
prepared film.
Fig. 3.20 STM images of bilayer silica film with different deposition temperature 300 K
(a), V=10V, I=0.1 nA; and 100 K (b), V=5.4 V, I=0.14 nA.
54
3.3.2 O-rich and O-poor films
As previously mentioned, there may be oxygen species located between the
bilayer silica film and Ru substrate, which are chemisorbed only on Ru(0001). This
chemisorbed interfacial O on Ru(0001) can desorb and re-adsorb by annealing the
sample in UHV and re-oxidation, causing binding energy shifts of Si2p and O1s[108].
This phenomenon is observed by a thermal treatment experiment, annealing the
prepared sample, o-rich film, step by step from 1030 K to 1150 K in UHV, forming
“O-poor” film, and finally re-oxidation at 1140 K. All the spectra were taken below
400 K after cooling down. (This work was carried out in Bessy II, where we used a
different sample holder caused a different temperature reading than the work
described in previous sections.) As shown in Fig. 3.21, the binding energy of Si2p
increases gradually with increasing annealing temperature, from 102.5 eV as
prepared to 103.3 eV after annealing in UHV at 1150K. The Si2p binding energy is
restored to 102.5 eV after re-oxidation. Similarly, the main peak in O1s spectrum,
which is assigned to the oxygen connected to Si (marked as O1s(Si)), shifts from
531.8 eV to 532.3 eV with increasing the annealing temperature and restored by re-
oxidation. The intensity and FWHM of Si2p and O1s(Si) are not changed indicating
the silica film maintains its chemical composition and stoichiometry. Whereas, the
shoulder at 529.9 eV, which is assigned to interface oxygen (marked as O1s(Ru)),
does not change the binding energy but is attenuated during the thermal treatment
and is finally restored after re-oxidation. Fig. 3.22 gives the binding energy as a
function of annealing temperature, where we can clearly obverse the trends. Both the
Si2p and O1s(Si) peaks shift by ~0.8 eV from o-rich film to o-poor film, whereas
O1s(Ru) does not change.
55
Fig. 3.21 The Si2p (hν =200 eV) (top) and O1s (hν = 630 eV) (bottom) regions in XP
spectra of a bilayer film as a function of annealing temperature in UHV as indicated and
after sample reoxidation at 1140 K for 10 min in 2 × 10−6 mbar O2
56
Fig. 3.22 BE shifts measured from the spectra shown in Fig. 3.21.
UPS results are shown in Fig. 3.23. The broad band from 5 to 9 eV is assigned
to the nonbonding O2p in silica, and the band region from 10 to 12.5 eV is
associated with the hybridized O2p-Si3s,sp bonding orbitals [58]. The work function
difference between o-rich (~ 7.2 eV) and o-poor (~ 6.2 eV) states is ~ 1eV. The same
trend is observed for silica free system (not shown), the work function of O(2×1) pre-
covered Ru(0001), e.g. 6.2 eV is larger the one of clean Ru(0001), e.g. 5.45 eV.
In this thermal treatment experiment, the LEED pattern, STM images and
infrared spectra remain the same for different film states, therefore the film structure
was not affected. These findings open the possibility of tuning the electronic property
of oxide/metal systems without changing the thickness and structure of the oxide
overlayer.
57
Fig. 3. 23 UP spectra of o-rich film and o-poor film, light source: He I.
3.4 “Thick” silica films on Ru(0001)
To see whether the bilayer silica film can be further grown in a layer-by-layer
manner, we have analyzed silica films with 4 MLE silicon deposited. The results for
the preparation in one step and the preparation in two sequential 2 MLE deposition–
oxidation steps were similar. XPS results (not shown) indicate that higher
temperature (about 25K higher) is needed to fully oxidize the film.
The large scale STM image of the 4 MLE film (Fig.3.24) reveals the smooth
surface, however the surface is not atomically flat. Atomic resolution was
unsuccessful, because the STM became unstable at low biases due to the strong
attenuation of the tunneling probability for the thick insulating films. It is only
possible to get atomic resolution in the depressed areas, and it shows a pore
structure similar to the vitreous bilayer silica film.
58
Fig. 3.24 STM images of the 4 MLE silica film on Ru(0001), V= 8.6 V, I=0.12 nA; The
respective LEED pattern (at 60 eV) is shown as an inset. (provided by B. Yang)
Fig. 3.25 IRA-spectrum of 4 MLE film.
59
Fig. 3.25 shows the IRAS results for 4 MLE film. It can be seen that the 1300
cm-1 and 694 cm-1 phonons, which are the characteristic phonons for a bilayer silica
film, are less intense while a phonon at 1257cm-1 with a shoulder at 1164cm-1
appears. The 1300 cm-1 and 694 cm-1 phonons can be assigned to the areas with
depressed contrast which show the pore structure, attenuate by a factor of three.
The 1257 cm-1 phonon corresponds to the rest of the film, which has been reported
for several nanometers thick silica overlayer on silicon single crystal [25]. The
phonon can be assigned to the asymmetric longitudinal optical vibration modes as in
quarts-like compounds [18]. Base on the results given above, we can conclude that
no monolayer or bilayer but a silica film with three-dimensional network of [SiO4]
tetrahedra grows on the bilayer silica film when the silicon thickness is greater than
2 MLE.
3.5 Summary
In this chapter, we have discussed the growth of silica films on a Ru(0001)
substrate. The results show that the atomic structure of silica films on Ru(0001) is
determined by the thickness of the film. For the silicon amount no more than 1MLE,
the film grows as monolayer film chemisorbed on Ru(0001) via Si-O-Ru bond, and
the characteristic phonon vibration is at ~1135 cm-1. The film which contains 2MLE
silicon grows as a bilayer film, two silica layers are connected via Si-O-Si linkage
which has a sharp phonon at ~1300 cm-1. The influence of the metal substrate
should be minimized in a good model system, therefore, this weakly bound bilayer
silica film is a suitable model system for investigation of silica related materials.
When increasing the film thickness more than 2 MLE, a IRAS feature at ~1257 cm-1
develops, which can be assigned to bulk-like silica, therefore the film grows as three-
dimensional.
60
The bilayer silica film can be grown as either crystalline or vitreous, or both
coexists, depending on different preparation conditions, respectively, slow cooling or
fast cooling after high temperature oxidation. It can be used as a suitable model
system for studying glass and glass transition in atomic level.
The electronic property of the bilayer silica film grown on Ru(0001) can be
tuned by the interface oxygen amount. The maximum of the work function difference
between o-rich film and o-poor film is about 1eV.
61
Chapter 4 Silica films on Pt(111)
As previously mentioned, the silica films grow as monolayer silica film on
Mo(112), however, on Ru(0001) both monolayer and bilayer silica films can be
obtained. In addition, the bilayer silica film on Ru(0001) can be grown as either
crystalline or vitreous depending on the preparation conditions. To understand the
role of the metal substrate on the structure of the silica film, we employed Pt(111) as
another substrate for a comparative study. Pt has a weaker interaction with oxygen
as compared to Mo and Ru. Pt(111) shares the same hexagonal symmetry with
Ru(0001), but the lattice constant, 2.77Å, is slightly larger than Ru(0001), (2.71 Å).
The Pt(111) surface has been widely used as a substrate for growing transition metal
oxide thin films, e.g., titania, iron oxides, ceria, etc [109-111].
4.1 Silica films on Pt(111)
Similar to silica films grown on Ru(0001), we started with making an oxygen
overlayer on Pt(111). We prepare the oxygen precovered Pt(111) surface by exposing
to 3×10-6 mbar O2 at ~1200 K for 5 min and cooling to room temperature before
pumping out oxygen.
Depositing silicon onto Pt(111) at 373K in 2×10-7 mbar oxygen results in the
XPS Pt4f peaks shifting to higher binding energy, while Si2p at ~100e V refers to
silicon in silicide. To avoid the formation of silicide, silicon was deposited at low
temperature, 100 K, in higher oxygen pressure, 1.5×10-6 mbar.
Finally, the film was oxidized at 1200 K in 1×10-5 mbar oxygen ambient. Due to
the low adhesion of oxygen on Pt, high oxygen pressure was necessary.
We begin by discussing the structure of 2 MLE silica film on Pt(111).
62
Fig. 4.1 Si2p region (top) and O1s region (bottom) in XP-spectrum of 2 MLE film.
XPS results presented in Fig. 4.1 show only one chemical state in the Si 2p
region with a binding energy of 102.8 eV corresponding to Si4+, indicating the
formation of [SiO4] tetrahedra. The O1s region has one peak located at 531.8 eV.
These results are similar to bilayer film on Ru(0001), except that there is no obvious
63
shoulder at lower BE, which was previously assigned to interfacial oxygen for the
SiO2/Ru(0001) system [78].
IRAS is a unique technique to identify the principal structure of silica films. In
Fig. 4.2 we can see a sharp phonon at 1294 cm-1 and a less intense peak at 690 cm-1,
which is virtually identical to those observed for the bilayer films on Ru(0001) (~1300
and 692 cm-1). The phonon at 1294 cm-1 can be assigned to an in-phase combination
of asymmetric Si-O-Si stretching vibrations of the Si-O-Si linkage between two silica
layers. The high intensity of the 1294 cm-1 peak informs that the Si-O-Si bonds are
oriented normal to the surface plane. These results indicate the formation of silica
bilayer film on Pt(111).
Fig. 4.2 IRA-spectrum of 2 MLE film on Pt(111).
As shown in Fig. 4.3, large scale STM reveals that a flat film with randomly
distributed holes covers the whole substrate. The wide terraces are separated by
monoatomic steps of Pt(111). The depth of the distributed holes in this film is ~2 Å,
64
which is considerably lower than the bilayer silica film on Ru(0001), where the
measured height of the film is ~5 Å[78].
Fig. 4.3 STM images of the 2 MLE silica film on Pt(111). V =4.4 V; I =0.1 nA (a); V=1.3 V,
I=0.13 nA (b).
LEED shows a Pt(111) (1×1) pattern accompanied by a diffuse (2×2) ring
pattern. Therefore, this silica film lacks long range ordering, which is also proven by
atomically resolved STM. Fig. 4.3 shows that the film is composed of corner-sharing
N-membered rings, with N varying between 4 and 9. This film structure has been
65
observed for the vitreous bilayer silica film on Ru(0001) [79, 106]. It was previously
found for the SiO2/Ru(0001) film that lower cooling rates favor the formation of
crystalline films. In the case of the films grown on Pt(111), no crystallinity was
observed even for very low cooling rates. So it can be concluded that the 2 MLE silica
on Pt(111) grows as a vitreous silica film.
It was reported that monolayer silica films can be grown on both Mo(112) and
Ru(0001) due to the formation of stable O-Mo and O-Ru bonds [43, 56, 69, 79]. In
order to study whether the O-metal bond is crucial for growing monolayer silica films,
a 1MLE silica film was prepared on Pt(111). If the silica film grows as a monolayer,
the whole substrate should be covered by the film. However, after annealing the
sample to 1100 K in O2 ambient, STM image (Fig. 4.4) shows that only half of the
substrate surface is covered by the film. The film is composed of two-dimensional
islands linked by straight stripes. In the small scale STM image, we can see that
these two-dimensional islands share the same surface structure as the vitreous
bilayer silica films. The apparent height of the islands is ~2 Å, same as for 2 MLE
film. The stripes bridging the islands have the same height, ~2 Å, and run along the
principal crystallographic directions of Pt(111). Most of these ‘bridges’ are ~7 Å in
width, although ~4 Å and ~10 Å wide lines are observed as well. These stripes
resemble the one-dimensional silica rows formed on Mo(112) at low coverage [49].
66
Fig. 4.4 (a) STM image of the 1 MLE film on Pt(111). Height profile along the A-B line,
indicated in (a), is shown in (b). The inset shows a close-up image demonstrating the
vitreous state of silica islands (cf., Fig. 4.4 (b)). The arrow indicates silica stripes
bridging the islands (tunneling parameters 0.8 V, 0.06 nA).
67
Fig. 4.5 XP-spectra of Si2p region (top) and O1s region (bottom) of 1 MLE film (red), and
2 MLE film (black).
XPS results presented in Fig. 4.5, show only one Si2p peak at 102.8 eV. It
should be noticed that the spectrum for the O1s core level shows only one species at
68
531.8 eV , which is obviously different from 1 MLE silica grown on Mo(112) and
Ru(0001), where the O1s peak can be deconvoluted into two components,
representing the O from Si-O-Si and the O from Si-O-Ru. It can be seen the O is half
of the intensity but keeps the same binding energy as 2 MLE silica film, therefore the
O ions are in the same chemical environment as in bilayer silica on Pt(111).
Fig. 4.6 IRA-spectra of 1MLE film (red), 2 MLE film (black).
The monolayer films grown on Mo(112) and Ru(0001) show characteristic
phonon vibrations located at 1060 cm-1 for Mo(112) [69] and 1136cm-1 for Ru(0001)
[79]. If a monolayer film were grown on Pt(111), phonon vibrations in the same
regions would also be found in the IRA spectra of such films. However, as shown in
Fig. 4.6, the 1 MLE silica film has only two sharp phonons at the same frequencies
found for the 2 MLE silica on Pt(111), with half of the intensity. The most intense
peak is located at 1294 cm-1 and the less intense peak is at 690 cm-1, indicating that
the film contains Si-O-Si stretching oscillators oriented perpendicular to the surface
plane. In addition, the absence of a phonon peak in the 1000-1200 cm-1 range
69
indicates that there is no Si-O-Pt bond formation. It is clear then that the 1MLE film
is a bilayer in nature.
Therefore, we can conclude that silica films on Pt(111) only grow as bilayer film
regardless of the silicon thickness.
Fig. 4.7 the BE shift of Si2p (Black) and O1s (red) XP spectra of SiO2/Pt(111) as a
function of annealing temperature in UHV as indicated and after reoxidation.
In addition, the experiment of transformation between O-rich and O-poor
states was carried out. In this experiment, all the XP spectra were taken at 100 K.
As shown in Fig. 4.7, both Si2p and O1s peaks shift to higher binding energy after
anneal the sample at 423 K in UHV. The binding energy change is ~0.2 eV when the
annealing temperature reaches 573 K. Further increase the annealing temperature
does not cause obvious change. Both peak of Si2p and O1s restored after re-
oxidation at 1200 K. These results indicate this SiO2/Pt(111) has reversible O-rich
and O-poor states, which is similar as SiO2/Ru(0001). However, the maximum
70
values of the binding energy change are quite different, ~0.2 eV for SiO2/Pt(111) and
~1.0 eV for SiO2/Ru(0001) [108]. This can be explained by the interface oxygen
amount in SiO2/Pt(111) is less than SiO2/Ru(0001). On the other hand, bond of the
interface oxygen on Pt(111) is weaker than on Ru(0001), therefore, lower annealing
temperature, 573 K (1150 K for SiO2/Ru(0001)), is required to desorb all of the
interface oxygen.
4.2 Support effects
Here, we summarize the silica structures on Mo(112), Ru(0001) and Pt(111).
On all substrates, the films are formed by a network of corner sharing [SiO4]
tetrahedra. On Mo(112), the film can only grow as monolayer silica film, which is
strongly bonded to the substrate via Si-O-Mo bond, which is schematically shown in
Fig. 4.8 (a). On Ru(0001), the silica film grows as a monolayer silica film, which is
bonded to the substrate via Si-O-Ru linkage, when the thickness is below 1MLE.
When the thickness is 2MLE, the film grows as a bilayer film weakly bonded to the
Ru(0001) substrate. The bilayer films can be grown as crystalline structure, which
consists of hexagonal rings composed of six [SiO4] units forming a network with long
range order(see in Fig. 4.8 (b)), or vitreous structure, which is composed of rings with
different ring size, where the number (N) of [SiO4] units, forming the ring, varies
between 4 and 9. On Pt(111), the silica film can only grow as a bilayer regardless of
the thickness and only the vitreous state has been observed.
71
Fig. 4.8 Schematic representation of a monolayer (a) and a bilayer (b) film on a metal
support. Top and cross views are shown. Si ion positions are in [SiO4] tetrahedra is
indicated by dots.
The principal structure of silica films grown on different metals correlates with
the oxygen affinity to the metal support. The reported heats of dissociative
adsorption of oxygen on Mo, Ru, and Pt are -544, -220 and -133 kJ/mole
respectively [112], which correlates with standard heats of formation for MoO2 (-588
kJ/mole), RuO2 (-153 kJ/mole) and PtO2(-71 kJ/mole) [113]. Therefore, the strong
O-Mo bonds leads to the formation of only monolayer silica films on Mo(112). Due to
the weakness of the O-Pt bond, the silica film on Pt(111) does not grow as monolayer
but bilayer, where two sheets of [SiO4] tetrahedra are interconnected via O ions
forming a ‘closed shell’ structure and the film terminated by the fully saturated
oxygen layer on either side. Ruthenium exhibits the intermediate properties, silica
film can grow as monolayer at low thickness. When thickness is more than 1MLE,
72
the O-Ru bond, which is stable but much weaker than O-Mo bond, in the Si-O-Ru
linkage can break, then the O connects with Si in another silica layer leading to the
formation of Si-O-Si linkage ultimately resulting in a bilayer silica film.
Fig. 4.9 STM images of the bilayer silica films: crystalline bilayer on Ru(0001) (a),
vitreous bilayer on Ru(0001) (b), and bilayer silica film on Pt(111) (c).
It is known, in heteroepitaxy, that another important factor influencing the
growth of the film is the lattice mismatch between an oxide film and a metal
substrate, which is obviously different for the metals studied. The lattice constant of
the unsupported, free-standing silica bilayer is computed to be 5.32 Å [50]. When the
films are grown on Ru(0001) (lattice constant 2.71 Å) and Pt(111) (lattice constant
2.77 Å), the most natural structure for this porous silica film to accommodate these
metal supports is the (2×2) structure. From this point of view, the Ru(0001), where
the periodicity of (2×2) structure is 5.42 Å, is better suited than Pt(111), where the
periodicity of (2×2) structure is 5.54 Å. This may explain why the silica bilayer film
can form a crystalline structure on Ru(0001) under certain conditions [78, 79],
whereas the bilayer film favors the formation of vitreous structure on Pt(111) in
order to release the tension caused by lattice mismatch.
73
Mo(112) has a rectangular unit cell (5.46 Å× 8.92 Å), which hardly fit a
hexagonal silica ovelayer. As a result, the monolayer silica films on Mo(112) form a
honeycomb-like structure with shortened periodicity of 5.2 Å in the Mo[-3 1 1]
direction and extended periodicity of 5.46 Å in the Mo[-1 -1 1] direction [43, 69].
However, in spite of the distortion, the monolayer silica form a perfectly ordered film
on the larger scale on the Mo(112) substrate, whereas the monolayer silica film on
Ru(0001) contains a multi-domain structure with a high density of domain
boundaries [79]. Therefore, we can conclude that the strong Si-O-Mo bond stabilizes
the well-ordered monolayer structure despite the energy cost of the silica lattice
distortion.
Fig. 4.10 STM images of the monolayer silica films on Mo(112) (a) [43], and on Ru(0001)
(b).
4.3 Summary
In this chapter, we have discussed the preparation and characterization of
ultrathin silica films on Pt(111). It is shown that the silica film only forms vitreous
bilayer structure on Pt(111). Based on the comparison of the structure of silica films
74
grown on Mo(112), Ru(0001) and Pt(111), we suggest that the metal-oxygen bond
strength plays a decisive role on the atomic structure of silica overlayers on metal
substrates. Lattice mismatch also influences the morphology of the film in a certain
way.
75
Chapter 5 Metal-containing silicate films on
Ru(0001)
Silicates are one of the most widely existing minerals in nature. These minerals
and synthetic silicate materials are extensively used in our society. Although there
are several compounds known to contain [SiO6], such as the high pressure phases
K[(AlSi3)[6]O8 with the hollandite-type structure, in most silicate, each silicon atom is
strongly bonded to four oxygen atoms forming a [SiO4] tetrahedron [114]. When
silicate contains only Si and O, it is known as silica.
Among silicate materials, zeolites received a lot of attention due to their
importance in heterogeneous catalysis [115]. However, the relation between
structure and reactivity of these complex and highly porous materials is mostly
studied using bulk-sensitive techniques and theoretical calculations [116]. To better
understand the catalytic reactions, which happens on the surface of catalyst, it is
necessary to employ the surface sensitive techniques. Therefore, a zeolite model
system that is accessible to surface science techniques is very valuable.
As mentioned in chapter one, a thin aluminosilicate film has been prepared on
Mo(112) substrate using monolayer silica as the template [77]. This thin film is
strongly bound to the substrate, lacking of negative charge resulting in unability to
be protonated by water adsorbing to form bridging OH groups, which exist in zeolite
playing an important role in many reaction due to the acidity of this OH group. In
this work, we prepared the aluminosilicate film using the bilayer silica film on
Ru(0001) as template, because the weakly bound film may be less affected by the
substrate.
76
Metal promoted zeolites are commonly used as catalysts. For example, Fe-
containing zeolites and Fe-silicalites efficiently catalyze several industrially important
oxidation reactions. [117] Basically, such materials are formed by substitution of a
small fraction of Si4+ with Fe3+ in the framework. However, the nature of active
species in these materials remains debating owing to a huge variety of Fe
coordination in and out of the crystalline framework [118]. In addition, iron-silicates
are found in many minerals, such as chamosite, greenalite, cronstedtite, nontronite,
glauconite, etc. Currently, the structure and property of iron-silicate is poorly
understood, especially, compared to the iron-oxide system. In this work, we studied
the growth of thin iron-silicate films on Ru(0001) using bilayer silica film as a
starting point.
Therefore, in this work, two metal-containing silicate systems, e.g.
aluminosilicate and iron-silicate, were studied.
5.1 Aluminosilicate films
To substitute Si in SiO2/Ru by Al, different methods were performed. Among
these methods, the one for the best film quality achieved is using co-deposition of
silicon and aluminium. Firstly, using the same procedure as described in chapter 3,
a 3O(2×2) overlayer was prepared on the clean Ru(0001) substrate. Then, Si and Al
were sequentially deposited by e-beam evaporation on to the 3O(2×2)-Ru(0001)
surface at 100 K in 2×10-7 mbar oxygen. Finally, the deposited mixture is annealed
at 1200K, in 3×10-6 mbar O2 for 10 minutes [81, 119].
Before presenting the characterization results, it is necessary to mention that
the composition of the aluminosilicate is written as AlxSi(1-x)O2, where x is the Al
molar fraction. Using the pure bilayer silica film on Ru(0001) as a reference, where
x=0, several samples with different Al/Si ratio were studied.
77
The XPS results of the prepared aluminosilicate films reveal that aluminum is
in the highest oxidation state (see Fig. 5.1). As shown in Fig. 5.2, the Si2p has the
same binding energy as in pure silica bilyer film inferring the Si4+ oxidation state. In
the O1s region, shown in Fig. 5.3, in addition to the two components observed in the
bilayer silica film, the one at 531.7 eV assigned to Si-O-Si and 529.9 eV assigned to
O(Ru), a new feature at ~530.7 eV is observed. This new feature gets more
pronounced with increasing Al/Si ratio and has previously been assigned to the Si-
O-Al linkage [77].
Fig. 5.1 Al2s region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red).
78
Fig. 5.2 Si2p region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red).
Fig. 5.3 O1s region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red).
79
Comparing with the bilayer silica film, which gives a sharp and strong phonon
at ~1300 cm-1 together with a phonon at ~690 cm-1, the Al-substituted films do not
give any new feature but these phonons are shifted in the infrared spectrum. While
increasing the Al content, the ~1300 cm-1 phonon vibration gradually red-shifts, but
the intensity and width of this phonon remains the same (see in Fig. 5.4). This
finding indicates that the Al atoms incorporate into the bilayer structure without
forming a new phase.
Fig. 5.4 IRA-spectrum of bilayer silica film on Ru(0001) (black), and Al0.2O0.8O2 film on
Ru(0001) (red).
At low Al/Si ratio, LEED shows a (2×2) pattern, with respect to Ru(0001),
indicating the formation of a crystalline structure. STM images reveal that the
surface is nearly atomically flat. STM images of an Al0.12Si0.88O2 film are presented in
Fig. 5.5. Some irregularly shaped areas (labeled A) with a slightly different contrast
are separated by protruding domain boundaries from other areas (labeled B). The
areas labeled B show the honeycomb-like network which is typical structure of the
80
crystalline bilayer silica films. In the domains A, the protrusions assigned to the
positions of oxygen in the topmost layer are brighter than those in areas B. The
coverage of areas A increases with Al/Si ratio indicating the Al-containing nature of
these areas. Therefore, the Al atoms are not randomly distributed across the surface,
but segregate into small domains. This finding seems to contradict Dempsey’s
statement, based on electrostatic considerations, that Al atoms arrange in zeolitic
structures as far as possible from each other [120]. Considerting lattice strain
induced by the Al incorporation into the silicate frame, the Al atoms prefer locating
near each other to minimize the strain.
Fig. 5.5 Large scale STM image of Al0.12Si0.88O2 film (a), high resolution STM images in
domain A (b), in domains B (c), V=0.15 V, I=0.07 nA.
81
Fig. 5.6 Large scale STM image of Al0.35Si0.65O2 film, V= 8.3 V, I=0.11 nA.
The large scale STM image of Al0.36Si0.64O2 film, Fig. 5. shows a flat film with
wide terraces separated by the monoatomic steps of the Ru(0001) substrate further
proving that Al atoms incorporates into the framework rather than forms an alumina
phase. For the films with Al/Si ratio close to 1, the integrity of the film is not
conserved. This is in accordance with Lowenstein’s rule, which states that Al-O-Al
linkages in zeolitic frameworks are forbidden [121].
Based on the results discussed above, one conclusion can be achieved that Al
incorporates into silica frame by substituting Si in [SiO4] tetrahedra. Therefore the
resulted films are bilayer aluminosilicate films weakly bound to Ru(0001) substrate.
To study where is the favorite location, in top layer or in bottom layer, for Al
atoms, angle resolved XPS experiment has been down. For the Al0.25Si0.75O2 film (see
in Fig. 5.7), the normalized signal ratio of Al to Si decreases while increasing the
82
angle between the analyzer and surface normal indicating the Al locates in the
bottom layer. This finding is predictable due to the charge imbalance caused by
substitution must be compensated. Since neither of extra cations and water, which
is used to protonate the oxygen bridging Al and Si, were introduced during
preparation, Al atom would prefer the bottom layer to get closer to the metal
substrate to let the metal substrate compensate the extra charge. As a result, Al
atom will not present in the top layer until the bottom layer is saturated at x=0.25.
Fig. 5.7 Angle resolved XP-spectra of Al0.25Si0.75O2 film, Al/Si normalized signals as a
function of the angle between analyzer and surface normal, α.
83
Fig. 5.8 IRA spectra of the Al0.4Si0.6O2 films, hydroxylated with H2O (black) and D2O (red),
recorded in 2×10-5 mbar CO atmosphere. The spectra were divided by reference spectra
taken before CO exposure.
To form surface OH groups, the sample was exposed to water at ~ 100 K,
resulting in a solid water layer, followed by heating to 300 K to desorb the weakly
bound water molecules. It has to be mentioned that on pure silica film, IRAS shows a
peak at 3750 (2763) cm-1 which can be assigned to the stretching vibration of OH
(OD) bonds in silanol (Si-OH (Si-OD)) groups associated with defect sites [122]. For
the aluminosilicate films, at low Al content (x < 0.25) no other OH related features
were observed. However, for the films with higher Al content a peak appears at 3594
cm-1, which falls into the vibration range of hydroxyl groups in the bridging Si-OH-Al
positions in zeolites [122-124]. Therefore, differing from the repoted aluminosilicate
films on Mo(112), the aluminosilitate films, with high Al content, weakly bound to
Ru(0001) substrate can be protonated forming Si-OH-Al species, or as OHbr. In
84
addition, the films were protonated using D2O resulting in a peak at 2652 cm-1,
which can be assigned to OD in Si-OD-Al species.
The OH groups in Si-OH-Al can be replaced by OD. When D2O was adsorbed
on the OHbr-containing surface at 100K forming a solid D2O layer and then followed
by heating to 300 K, the OHbr signal at 3594 cm-1 disappears and the peak at 2652
cm-1, which is assigned to ODbr, was detected, thus indicating the H/D exchange,
which is a well-known phenomenon in zeolite chemistry [125].
Fig. 5.9 IRA spectra of the OH-terminated surface of the Al0.4Si0.6O2 films upon
subsequent hydroxylation with D2O.
The acidity of the Si-OH-Al (Si-OD-Al) groups can be quantified by adsorption
of CO as a weak base, which binds to the acidic proton through the C atom to form a
CO···HO adduct. A red-shift in ν(OH) together with a blue-shift in ν(CO) (compared to
CO in gas phase 2143 cm-1) are induced as a consequence, and the magnitude of the
85
shift is proportional to the degree of acidity [122, 126, 127]. Fig. 5.9 shows the
examples of an Al0.4Si0.6O2 film, After exposing the protonated surface in CO ambient
(2×10-5 mbar), the red-shift of OHbr was observed, the peak shifts to 3215 cm-1 by
379 cm-1 from 3594 cm-1. Similarly, the ODbr signal red-shift to 2409 cm-1 from 2652
cm-1 by 243 cm-1. Meanwhile, the CO stretching vibration peak is at 2183 cm-1 is
blue-shifted from 2143 cm-1, which is the frequency of CO in gas phase. Similarly,
the ODbr signal red-shift to 2409 cm-1 from 2652 cm-1 by 243 cm-1. Therefore, the
results indicate the formation of acidic OH(OD) groups on the aluminosilicate film
with high Al/ratio.
To summarize, the aluminosilicates film grown on Ru(0001) has structure
composed of [SiO4] and [AlO4] building blocks, thus structurally similar as zeolites.
The weakly bound films with higher Al content, x>0.25, can be protonated forming
acidic OHbr (ODbr) groups which is an important feature of zeolites in catalysis. In
addition, the H/D exchange occurs on this system, which is a well-known
phenomenon in zeolite chemistry. Therefore, the bilayer aluminosilicate films weakly
bound to Ru(0001) can be used as a suitable model system for zeolites [81].
5.2 Iron-silicate films
The preparation of iron silicate film was similar to aluminosilicate, which starts
with making a 3O(2×2) overlayer on clean Ru(0001) surface, followed by sequential
deposition of Si and Fe in oxygen ambient. Finally the deposited mixture was
annealed at ~1150 K in 5×10-6 mbar O2. We define the Fe molar fraction in the
resulted films as Fe/(Fe+Si), determined by XPS, the sum of the molar amounts of Si
and Fe was equal to the silicon in a bilayer silica film.
We begin with the IRAS results since IRAS turned out to be the most sensitive
to the principal structure of the films. As previously mentioned, the most
86
characteristic features of the pure silica bilayer film are two strong phonon
vibrations that can be easily seen by IRAS at 1300 cm-1, corresponding to a
combination of in-phase asymmetric Si-O-Si stretching vibrations of the Si-O-Si
linkage between the two layers and a peak at 692 cm-1 associated with a combination
of symmetric Si-O-Si stretching vibrations of Si-O-Si bonds nearly parallel to the
surface [78]. Fig. 5.10 gives a series of IR spectra of the films with different Fe
content. A sharp new peak at 1005 cm-1 together with a less intensity peak at ~674
cm-1 appear with the introduction of Fe into the films. These two new features gain
intensity with increasing Fe content, meanwhile the features of silica bilayer film,
~1300 cm-1 and 692 cm-1, gradually attenuate.
Fig. 5.10 IRA-spectra of FexSi1- xO2 films on Ru(0001) as a function of the Fe content (x)
as indicated.
87
It is interesting to compare this system with aluminosilicate bilyer films on
Ru(0001). IRAS results show that the 1300 cm-1 phonon shifts to lower frequency
after introducing Al into the silica bilayer framework, where the Al atoms do not form
any separate phase, but are incorporated into the bilayer structure forming Al-O-Si-
O-Al linkage in plane. With increasing the Al/Si ratio, the phonon is further red
shifted, for instance the phonon vibration is ~1270 cm-1 when the Al content is 40%
[81, 119]. Apparently, iron doped silica films behave differently from aluminosilicate
films. There is no red shift of the 1300 cm-1 phonon indicating that the film has
regions of pure silica, and the new phonon at 1005 cm-1 indicates the formation of
forming a new phase related to the presence of iron. Therefore, we suggest that the
iron doped silica films have two different phases, a pure bilayer silica region and an
iron-contained silica region. Although Al ions exhibited some islanding at low x, as
shown by STM in Fig. 5.5, in the case of an Fe-silicate film, changes in the reduced
masses of respective oscillators are much greater than in an Al-containing silicate
film, and vibrational coupling of the two phases is hardly possible.
When the Fe content reaches 50%, the 1300 cm-1 peak is almost gone, and the
intensity of the 1005 cm-1 peak is comparable to the 1300 cm-1 peak for silica bilayer
film. It is obvious that when the Fe/Si ratio equals 1, the vertical Si-O-Si oscillator
between layers is no longer present. On the other hand the 1005 cm-1 vibration is
close to the phonon vibration of silica monolayer grown on metal substrates (i.e.
1134 cm-1 for Si-O-Ru on Ru(0001) and 1060 cm-1 for Si-O-Mo on Mo(112)).
Therefore the new phonon frequency of 1005 cm-1 can also be assigned to the Si-O-
Fe oscillator.
This hypothesis is supported by the water absorption experiments (see Fig.
5.11). After making a D2O ice layer on the films by dosing water at 100 K, the 1005
cm-1 phonon vibration shifts to ~1022 cm-1. This is similar to what was observed for
ice layers deposited on a monolayer silica film on Mo(112). A 9 cm-1 blue shift of the
88
Si-O-Mo stretching vibration is observed after adsorption H2O (D2O) at 100K on a
monolayer silica film on Mo(112), which may be caused by polarization of the
water/silica interface [44]. Similarly, the shift seen in our case can be caused by the
polarization of the water/iron-containing film interface. Note that, after desorbing the
ice layer by heating the sample to 300K, no feature which can be assigned to Si-OH-
Fe groups is observed for the films with any Fe content (from 0.07 up to 0.5), which
is different from the situation for aluminosilicate with Al content higher than 0.25.
Therefore, we can conclude that no negative charge is present in the top layer for all
the films.
Fig. 5.11 IRA-spectra of Fe0.2Si0.8O2 films on Ru(0001) before D2O adsorption (black), and
with D2O ice layer (red).
For comparison, an iron oxide film grown on Ru(0001) using the same
preparation conditions, which resulted in a FeO(111) film is IRAS-silent in the
89
spectral region above 600 cm-1 (see Fig. 5.10). This finding rules out the formation of
the alone-standing FeO(111) structure in the Fe-silicate films.
Fig. 5.12 LEED patterns (60 eV) of the FexSi1- xO2 films on Ru(0001) at x = 0 (a), 0.2 (b),
0.5 (c), and 1 (d). The unit cells are indicated.
Fig. 5.12 shows the LEED images of films with different Fe content, all images
were taken at a beam energy of 60 eV. The spots in image (a) corresponds to a pure
silica bilayer showing a (2×2) structure, with respect to the Ru(0001) underlying
lattice, typical of the ordered crystalline film and a (2×2) ring characteristic of
amorphous regions. After introducing Fe into the film (b), the LEED gives the same
(2×2) pattern as seen for the silica bilayer film coexisting with a 30°–rotated pattern
with a Moiré structure, indicateing the formation of a new phase with long range
order different than the pure silica film. It should be noted that the 30°–rotated
90
pattern has a slightly larger distance between spots than Ru(0001)-(2×2), so the
lattice constant of the film is shortened to about 5.25 Å (cf 5.42 Å in the original film).
As the Fe/Si ratio is increased, the moiré pattern gets more pronounced, following
the same trend found with the 1005 cm-1 phonon in IRAS. For the film with Fe
content x ~ 0.5, the LEED shows only the rotated 2×2 spots with the moiré pattern.
The LEED pattern for iron oxide grown on Ru(0001) using the same preparation
condition as iron-silicate films is shown in Fig4.(d), which is the pattern for FeO(111)
[128]. This LEED pattern is not observed in Fig 4.(b) and (c), which gives evidence
that there is no formation of the reported FeO(111) species when iron oxide is grown
on Ru(0001).
Fig. 5.13 Large-scale (a) and high-resolution (b) STM images of the Fe0.2Si0.8O2 film.
To get more details about the morphology of the films, STM measurements
were performed. In the large scale STM image shown in Fig. 5.13 (a), two different
areas, with and without moiré structure, are found. The measured periodicity of the
moiré structure is ~ 26 Å, which is in good agreement with LEED results. With
comparison of STM images of films with different Fe content, a trend is revealed that
the area with moiré superstructure grows with increasing the Fe content. Therefore,
91
the area showing a moiré pattern can be assigned to the iron-containing phase,
associated with the 1005 cm-1 phonon in IRAS. The step height between terraces
with identical structure is about 2.5 Å, which corresponds to a monoatomic step on
Ru(0001) underneath the film. Within the same terrace, the moiré area is slightly
higher than the silica area. The atomically resolved STM image, Fig. 5.13 (b), shows
that the film with moiré superstructure rotates 30° with respect to the Ru(0001)
substrate. The area without moiré structure has a honeycomb-like structure with a
periodicity of 5.5 Å, which corresponds to the pure silica bilayer film. In addition, the
unrotated crystalline and vitreous silica structure was observed in some small areas
close to the boundary of the iron-containing film and the pure silica film. These
results indicate that Fe is not distributed uniformly over the whole surface but
segregates forming new phases which separate from the pure silica bilayer. In
addition, the area covered by the moiré pattern is two times larger than the Fe molar
fraction, e.g. when the Fe content is 35%, the total area of the film with moiré
pattern is about 70-80%. This is different from the aluminosilicate film for which the
Al containing phase covers the whole surface when the Al content reaches 25% [81],
whereas for the films described here the Fe-containing phase (moiré pattern) covers
the whole substrate only when the Fe content reaches 50%.
92
Fig. 5.14 Si2p region in XP-specta of silica bilayer (black), Fe0.5Si0.5O2 (red) and Fe(111)
(blue).
Fig. 5.15 Fe2p region in XP-specta of silica bilayer (black), Fe0.5Si0.5O2 (red) and Fe(111)
(blue).
93
Fig. 5.16 Top: XP-spectra of the O1s core level in pure SiO2, FeO(111) and Fe0.5Si0.5O2
films on Ru(0001) as indicated. Bottom: Spectral deconvolution for the spectrum of the
Fe0.5Si0.5O2 film.
For all of the films, as shown in Fig. 5.14, the XPS spectra show that the Si2p
binding energy has the same value, 102.6eV, as pure silica bilayer film, which refers
to Si4+. Therefore Si is in the same chemical environment, and each Si atom is
surrounded by four O atoms forming [SiO4] tetrahedra, which is the same as in
mono- and bi- layer silica and aluminosilicate on Ru(0001). In fig 5.15, we can see
that the intensity of the Fe2p signal scales with the Fe amounts, while the signal
shape is virtually identical to that observed for a FeO(111)/Ru(0001) film. The only
differences are observed for the O1s level. At low Fe content, the O1s spectra shows
94
a main peak at 531.8 eV with shoulder at low BE, which is related to the oxygen
atoms bonded to metal (Ru or Fe). With increasing Fe/Si ratio, the main peak
attenuates and the shoulder becomes more pronounced. When the Fe content
reaches ~0.5, a broad O1s peak forms. Fig. 5.16 displays the XP-spectra in the O1s
region for a pure SiO2 film, which shows a main peak centered at 531.8 eV and a
shoulder at ~530 eV, a pure FeO(111) film, which gives a peak with binding energy at
530eV, and for a film with Fe content of ~0.5, i.e. Fe0.5Si0.5O2. Spectral
deconvolution of the O1s region of the ~0.5 Fe film reveals three oxygen species,
which could tentatively be assigned to the O atoms in Si-O-Si (at 532 eV), Si-O-Fe
(531.0 eV) and Fe-O-Fe (530 eV) coordination environments.
To summarize the results of the film with Fe content of ~ 0.5, the surface is
uniformly covered with a well-ordered film that is rotated by 30° with respect to
Ru(0001) and shows the Moiré structure and the honeycomb-like framework. The
IRAS spectrum for this film shows a sharp and intense phonon at 1005 cm-1,
whereas the ~1300 cm-1 phono no longer exists indicating that the Si-O-Si between
layer is not present in this film. Based on the reduced mass analysis, the band at
1005 cm-1 can reasonably be assigned to Si-O-Fe bonds. In addition, differing from
the aluminosilicate film grown on Ru(0001), this film does show acidic bridging OH
group upon water adsorption. XPS results show that there are three oxygen species
in the film, which can be attributed to the oxygen in Si-O-Si, Si-O-Fe and Fe-O-Fe
bonds.
Now we are in the situation of raising a hypothesis of this iron-containing
silicate film structure. As a starting point, one can consider a bilayer structure where
Si is substituted by Fe like in the previously studied aluminosilicate films.
Accordingly, the Fe ions could be placed in two possible positions, in the top or the
bottom cation layer, or both. Based on the high-resolution STM images of Fe-
95
containing areas, it can be suggested that formation of only one structure rather
than a mixture of several ones due to the uniform contrast observed. As shown in Fig.
5.13, it is clear that the film segregates into two phases, the Fe-containing one and
pure silica film, therefore, it is not consistent with the random distribution of Fe in
cation layers. Moreover, the XPS results show that there is a substantial amount of
oxygen coordinated only to Fe, like in FeO(111) (Fig. 5.16) indicating the the Fe
atoms are not randomly distributed in two layers like Al atom in aluminosilicate; and
for the film with Si : Fe ratio equal to 1, O1s XP-spectrum is consistent with the
assumption that Fe ions substitute all Si in one layer, either the top one or the
bottom one. Since the Fe-O bond length (~2.0 Å in iron oxides) is much larger than
the Si-O bond (~ 1.65 Å in silica), a severe lattice distortion will be raised by the Si
substitution by Fe. To minimize the effect of this substitution, one possible solution
is that Fe atoms located in the bottom layer interact with the Ru(0001) support,
which must be much stronger than the van der Waals interaction observed for the
pure silica bilayer films. In addition, if Fe substitutes Si in the top layer, it would be
in 4+ oxidation state which is rather unusual for the iron. In the case of
aluminosilicate films, charge imbalance for Al3+ in the top layer was compensated by
protons which formed bridging (Si-OH-Al) hydroxyls clearly identified by IRAS. [81]
However, the water adsorption experiments on Fe-silicate films show that there is no
formation of Si-OH-Fe species. In conclution, the experimental results suggest that
the Fe-containing layer must be closer to the metal support than the Si-layer. This
would be also consistent with the STM results showing the same honeycomb-like
surface structure as in a pure silicate film.
DFT calculations for this system were done by R. Włodarczyk, and J. Sauer.
Among several models, the most stable model, where Fe content is 0.5, is shown in
Fig. 5.17. The film is composed of a monolayer silica film grown on an iron oxide
layer. The calculated IRAS spectrum of this structure is shown in Fig. 5.18, it gives a
96
sharp phonon at 1002 cm-1 together with a less intense phonon at 652 cm-1, which is
in excellent agreement with our experimental results (1005 cm-1 and 674 cm-1).
Therefore, the Fe-silicate films are structurally very different from the similarly
prepared aluminosilicate films, where Al is rather uniformly distributed in the bilayer
silicate framework, following the Lowenstein rule stating that the Al-O-Al linkages
are forbidden. Such a layered structure, observed here for the Fe-silicate film, is
characteristic for clay minerals. Fig. 5.19 depicts the atomic structure of an idealized
nontronite Fe2Si4O10(OH)2, representative of the Fe-rich smectites [129]. It consists of
two tetrahedral silicate sheets that sandwich an octahedral Fe-hydroxide sheet, thus
classified as 2:1 to differentiate from 1:1 clay minerals formed by alternating tetra-
and octahedral sheets. There is a clear similarity between the Fe-silicate and
nontronite structures as shown in Figs. 5.17 and 5.19. In essence, the Fe-silicate
film can be formed by a cleavage of an idealized nontronite along the mirror plane.
Note that the presence of OH groups almost parallel to the surface within the iron
oxide sheet is hard to identify by IRAS due to the well-known metal selection rules
[92].
97
Fig. 5.17 Top and side views of the most stable structure of Fe4Si4O162O/Ru(0001) found
by DFT, Si yellow, Fe dark violet, O red, Ru gray. (provided by R. Włodarczyk)
Fig. 5.18 IRA-spectrum simulated for the structural models depicted in Fig. 5.16. The
spectra are scaled by a factor of 1.0341 (see ref. [78]). (provided by R. Włodarczyk)
98
Fig. 5.19 Top and side views of an idealized nontronite (in polyhedral representation)
formed by tetrahedral silicate sheet (in blue) and octahedral Fe hydroxide sheet (in
yellow). Structural protons are seen in the top view as small balls. [129]
In conclusion, this iron-silicate film grown in Ru(0001) substrate can be used
as a model system to study the iron-containing clay materials, which contain redox-
active Fe3+ species participating in electron transfer reactions and play an important
role in biogeochemical processes [129, 130].
99
5.3 Summary
We have prepared aluminosilicate films and iron-silicate films on Ru(0001).
The aluminosilicate thin films consist of corner-sharing [TO4] tetrahedra, where T is
Si or Al. When the Al content is higher than 0.25, the film exposes the acidic OHbr
after being protonated by water and H/D exchange has been observed on this
protonated surface. Therefore, this thin film is an ideal model system for studying
zeolite materials not only structurally but also chemically.
The iron-silicate film is composed of a silica layer which consists of corner-
sharing [SiO4] tetrahedra forming a honeycomb-like network on top of an iron oxide
layer with hexagonal structure. This film has a similar structure to some clay
materials, therefore this thin iron-silicate film can be used as a model system to
study iron-containing clay.
100
101
Chapter 6 Summary and outlook
Due to the complexity of real catalytic reactions, it is a challenge to extract the
information about reaction mechanisms. One approach to study the mechanism of
catalytic reactions is to use systems with reduced complexity, i.e. model catalysts. It
is necessary to get the knowledge of chemical and physical properties of the model
system to ensure its suitability to mimic the real catalyst. Based on work carried out
over many decades, surface science techniques have been shown to be powerful tools
to get fundamental insight into the properties of these model systems.
One driving force for performing research on the growth of silica films on
Ru(0001) substrates is to develop a suitable model systems to study the widely used
silica related materials, especially the usage as support material in catalysis. The
principal structure of the silica films grown on Ru(0001) is controlled by the silicon
amount in the film. For lower silicon contents, the film grows as monolayer silica,
SiO2.5/Ru(0001), which shares a similar structure with the well-studied
SiO2.5/Mo(112) system. Most importantly, a bilayer silica film weakly bound to
Ru(0001) was obtained. This film consists of two silica layers connected together via
Si-O-Si linkages, resulting in a SiO2 stoichiometry. Interestingly, using different
preparation conditions, the film can be grown either as a crystalline or a vitreous
structure. In addition, the electronic properties of this bilayer film can be tuned by
varying the amount of interfacial oxygen. An important feature of this film is that it
is not chemically bound to the substrate and the interaction is mainly due to Van
der Waals forces. All in all, this SiO2/Ru(0001) film represents a good model system
for studying the structure and properties relationship of silica-based materials.
We have extended our studies to silica films grown on a Pt(111) substrate to
get an insight into the support effects, by comparing these results with silica films
102
grown on Ru(0001) and Mo(112). It turned out that the principal structure of the
silica film is strongly affected by the oxygen affinity of the metal supports, which
influences the stability of the Si-O-Metal bonds. Only monolayer silica films can be
grown on Mo(112) and only bilayer silica films are observed on Pt(111), whereas both
mono- and bi-layer films can be prepared on Ru(0001). In addition, the lattice
mismatch between silica films and the metal supports affects the crystallinity of the
films, such that a large lattice mismatch tends to favors vitreous structures for the
bilayer case. Therefore, to prepare a suitable model system of catalyst, it is important
to choose the proper metal support.
Two model systems of silicates, aluminosilicates and iron-silicates, were
established using the bilayer silica structure on Ru(0001) as a starting point. The
resulting aluminosilicate films with higher Al content ( 0.25 ) show similar
properties as bulk zeolite materials. For example, these films contain acidic bridging
hydroxyls (OHbr), which readily exhibit H/D exchange. The similarities of these films
with their three-dimensional counterparts, i.e.: zeolites, make them suitable models
to study the catalytic properties of active sites of zeolites using surface techniques.
The iron-silicate films with Fe content 0.5 consist of a monolayer silica film on
top of a monolayer iron oxide (also with honeycomb structure), linked together via Si-
O-Fe bonds. This finding provides an approach to study iron-containing clay
materials. In fact, the film grown in this work has the same structure part of
nontronite layer, which is an important iron-containing clay.
In summary, the work presented in this thesis opens a new playground for
studying silica related materials. The bilayer silica film on Ru(0001) can be used as a
good support for catalytically active particles due to the weak interaction of the film
and the metal substrate, which is important for a good model system. Since the
structure of the films can be affected by the metal substrates, it is important to
103
choose the proper support when preparing the model systems. The prepared
aluminosilicate films on Ru(0001) can be further used to study the catalytic
properties of zeolites due to the similarity of chemical properties between the films
and real zeolites, e.g. both of them contain acidic bridging OH groups. The iron-
silicate films present a suitable model system for studying iron-containing clay
materials. Therefore, we believe these findings open the possibility of using a surface
science approach to properly model silica-supported catalysts and other silicate
materials.
104
105
BIBLIOGRAPHY
1. G. Ertl, H. Knoezinger, J. Weitkamp, Handbook of heterogeneous catalysis.
Vol. 1. 2008, Weinheim: Wiley-VCH.
2. H. J. Freund, Model studies in heterogeneous catalysis. Chem. Eur. J., 2010.
16(31): p. 9384-9397.
3. D.W. Goodman, Model catalysts: From extended single crystals to supported
particles. Surface Review and Letters, 1995. 02(01): p. 9-24.
4. H.J. Freund, Model studies on heterogeneous catalysts at the atomic level.
Catalysis Today, 2005. 100: p. 3-9.
5. N. Nilius, T. Risse, S. Schauermann, S. Shaikhutdinov, M. Sterrer, H.J.
Freund, Model Studies in Catalysis. Topics in Catalysis. 54(1-4): p. 4-12.
6. S. Shaikhutdinov, H.J. Freund, Ultrathin Oxide Films on Metal Supports:
Structure-Reactivity Relations. 2012. 63.
7. H.J. Freund, G. Pacchioni, Oxide ultra-thin films on metals: new materials
for the design of supported metal catalysts. Chemical Society Reviews, 2008.
37(10): p. 2224-2242.
8. H.J. Freund, J. Libuda, M. Bäumer, T. Risse, A. Carlsson, Cluster, facets,
and edges: Site-dependent selective chemistry on model catalysts. The
Chemical Record, 2003. 3(3): p. 181-201.
9. G. Ertl, H. Knoezinger, J. Weitkamp, Handbook of heterogeneous catalysis.
Vol. 3. 2008, Weinheim: Wiley-VCH.
106
10. J.A. Moulijn, P.W.N.M.v. Leeuwen., R.A.v. Santen, Catalysis : an integrated
approach to homogeneous, heterogeneous and industrial catalysis. 1993,
Amsterdam, New York: Elsevier Science.
11. D.A. Muller, T. Sorsch, S. Moccio, F.H. Baumann, K. Evans-Lutterodt, G.
Timp, The electronic structure at the atomic scale of ultrathin gate oxides.
Nature, 1999. 399(6738): p. 758-761.
12. S. Shaikhutdinov, H.J. Freund, Ultrathin Silica Films on Metals: The Long
and Winding Road to Understanding the Atomic Structure. 2013. 25: p. 49.
13. D.W. Berreman, Infrared Absorption at Longitudinal Optic Frequency in
Cubic Crystal Films. Physical Review, 1963. 130(6): p. 2193-2198.
14. W. Braun, H. Kuhlenbeck, Chemical structure of ultrathin thermally grown
oxides on a Si(100)-wafer using core level photoemission. Surface Science,
1987. 180(1): p. 279-288.
15. J. Derrie, M. Commandre, SiO2 ultra thin film growth kinetics as investigated
by surface techniques. Surface Science, 1982. 118: p. 32-46.
16. G. Feliciano, B. Angelo, P. Alfredo, Atomistic models of the Si(100)-SiO2
interface: structural, electronic and dielectric properties. Journal of Physics:
Condensed Matter, 2005. 17(21): p. S2065.
17. J. Finster, D. Schulze, F. Bechstedt, A. Meisel, Interpretation of XPS core
level shifts and structure of thin silicon oxide layers. Surface Science, 1985.
152-153(0): p. 1063-1070.
18. F. Giustino, A. Pasquarello, Infrared Spectra at Surfaces and Interfaces from
First Principles: Evolution of the Spectra across the Si(100)-SiO2 Interface.
Physical Review Letters, 2005. 95(18): p. 187402.
107
19. P.J. Grunthaner, M.H. Hecht, F.J. Grunthaner, N.M. Johnson, The
localization and crystallographic dependence of Si suboxide species at the
SiO2/Si interface. Journal of Applied Physics, 1987. 61(2): p. 629-638.
20. N. Herbots, J.M Shaw, Q.B. Hurst, M.P. Grams, R.J. Culbertson, D.J Smith,
V. Atluri, P. Zimmerman, K.T. Queeney, The formation of ordered, ultrathin
SiO2/Si(100) interfaces grown on (1 × 1) Si(100). Materials Science and
Engineering: B, 2001. 87(3): p. 303-316.
21. K. Hirose, H. Nohira, K. Azuma, T. Hattori, Photoelectron spectroscopy
studies of SiO2/Si interfaces. Progress in Surface Science, 2007. 82(1): p. 3-
54.
22. S. Ichimura, A. Kurokawa, K. Nakamura, H. Itoh, H. Nonaka, K. Koike,
Ultrathin SiO2 film growth on Si by highly concentrated ozone. Thin Solid
Films, 2000. 377-378(0): p. 518-524.
23. C.T. Kirk, Quantitative analysis of the effect of disorder-induced mode
coupling on infrared absorption in silica. Physical Review B, 1988. 38(2): p.
1255-1273.
24. K.T. Queeney, Y.J. Chabal, M.K. Weldon, K. Raghavachari, Silicon Oxidation
and Ultra-Thin Oxide Formation on Silicon Studied by Infrared Absorption
Spectroscopy. physica status solidi (a), 1999. 175(1): p. 77-88.
25. K.T. Queeney, M.K. Weldon, J.P. Chang, Y.J. Chabal, A.B. Gurevich, J.
Sapjeta, R. L. Opila, Infrared spectroscopic analysis of the Si/SiO2 interface
structure of thermally oxidized silicon. Journal of Applied Physics, 2000.
87(3): p. 1322-1330.
108
26. J.A. Schaefer, W. Göpel, Initial stages of oxidation of Si{100}(2 × 1): A
combined vibrational (EELS) and electron binding energy (XPS) study.
Surface Science, 1985. 155: p. 535-552.
27. J.M. Shaw, N. Herbots, Q.B. Hurst, D. Bradley, R.J. Culbertson, V. Atluri, K.
T. Queeney, Atomic displacement free interfaces and atomic registry in
SiO2/(1 x 1) Si(100). Journal of Applied Physics, 2006. 100(10): p. 104109.
28. P.A. Thiry, M. Liehr, J. J. Pireaux, R. Sporken, R. Caudano, J. P. Vigneron, A.
A. Lucas, Vibrational study of the SiO2/Si interface by high resolution
electron energy loss spectroscopy. Journal of Vacuum Science & Technology
B: Microelectronics and Nanometer Structures, 1985. 3(4): p. 1118-1121.
29. X. Xu, D.W. Goodman, New approach to the preparation of ultrathin silicon
dioxide films at low temperatures. Applied Physics Letters, 1992. 61(7): p.
774-776.
30. X. Xu, D.W. Goodman, The preparation and characterization of ultra-thin
silicon dioxide films on a Mo(110) surface. Surface Science, 1993. 282(3): p.
323-332.
31. J.W. He, X. Xu, J.S. Corneille, D.W. Goodman, X-ray photoelectron
spectroscopic characterization of ultra-thin silicon oxide films on a Mo(100)
surface. Surface Science, 1992. 279: p. 119-126.
32. M. Kundu, Y. Murata, Growth of single-crystal SiO2 film on Ni(111) surface.
Applied Physics Letters, 2002. 80(11): p. 1921-1923.
33. Z. Zhang, Z. Jiang, Y. Yao, D. Tan, Q. Fu, X. Bao, Preparation and
characterization of atomically flat and ordered silica films on a Pd(100)
surface. Thin Solid Films, 2008. 516(12): p. 3741-3746.
109
34. D.C. Kershner, J.W. Medlin, Adsorption and reaction of SiH4 and oxygen on
Pd(111). Surface Science, 2008. 602(3): p. 786-794.
35. D.C. Kershner, M.P. Hyman, J.W. Medlin, Density functional theory studies
of submonolayer oxidized silicon structures on Pd(111) and Pt(111). Surface
Science, 2008. 602(23): p. 3603-3610.
36. D.C. Kershner, W. Zhang, J. Will Medlin, Investigation of submonolayer SiOX
species formed from oxidation of silane on Pt(111). Surface Science, 2008.
602(21): p. 3225-3231.
37. M. Chen, D.W. Goodman, The structure of monolayer SiO2 on Mo(112): A 2-D
[Si-O-Si] network or isolated [SiO4] units? Surface Science, 2006. 600(19): p.
L255-L259.
38. M. Chen, D.W. Goodman, Reply to comment on "The structure of monolayer
SiO2 on Mo(112): A 2-D [Si-O-Si] network or isolated [SiO4] units?". Surface
Science, 2007. 601(2): p. 591-593.
39. M.S. Chen, A.K. Santra, D.W. Goodman, Structure of thin SiO2 films grown
on Mo(112). Physical Review B, 2004. 69(15): p. 155404.
40. C. Freysoldt, P. Rinke, M. Scheffler, Ultrathin Oxides: Bulk-Oxide-Like Model
Surfaces or Unique Films? Physical Review Letters, 2007. 99(8): p. 086101.
41. L. Giordano, A. Pasquarello, Structure and vibrational spectra of crystalline
SiO2 ultra-thin films on Mo(112). Surface Science, 2005. 584: p. 225-236.
42. L. Giordano, D. Ricci, G. Pacchioni, P. Ugliengo, Comment on "The structure
of monolayer SiO2 on Mo(112): A 2-D [Si-O-Si] network or isolated [SiO4]
units?". Surface Science, 2007. 601(2): p. 588-590.
110
43. S. Kaya, M. Baron, D. Stacchiola, J. Weissenrieder, S. Shaikhutdinov, T.K.
Todorova, M. Sierka, J. Sauer, H.J. Freund, On the geometrical and
electronic structure of an ultra-thin crystalline silica film grown on Mo(112).
Surface Science, 2007. 601(21): p. 4849-4861.
44. S. Kaya, J. Weissenrieder, D. Stacchiola, S. Shaikhutdinov, H.J. Freund,
Formation of an Ordered Ice Layer on a Thin Silica Film. The Journal of
Physical Chemistry C, 2006. 111(2): p. 759-764.
45. S. Kaya, J. Weissenrieder, D. Stacchiola, T.K. Todorova, M. Sierka, J. Sauer,
S. Shaikhutdinov, H.J. Freund, Formation of one-dimensional molybdenum
oxide on Mo(112). Surface Science, 2008. 602(21): p. 3338-3342.
46. Y.D. Kim, J. Stultz, T. Wei, D.W. Goodman, Surface Characterization Using
Metastable Impact Electron Spectroscopy of Adsorbed Xenon. The Journal of
Physical Chemistry B, 2002. 107(2): p. 592-596.
47. Y.D. Kim, T. Wei, D.W. Goodman, Identification of Defect Sites on SiO2 Thin
Films Grown on Mo(112). Langmuir, 2002. 19(2): p. 354-357.
48. T. Kinoshita, S. Mizuno, Surface structure determination of silica single layer
on Mo(112) by LEED. Surface Science, 2011. 605: p. 1209-1213.
49. J.L. Lu, S. Kaya, J. Weissenrieder, T.K. Todorova, M. Sierka, J. Sauer, H.J.
Gao, S. Shaikhutdinov, H.J. Freund, Formation of one-dimensional
crystalline silica on a metal substrate. Surface Science, 2006. 600(13): p.
L164-L168.
50. U. Martinez, L. Giordano, G. Pacchioni, Nature of Point Defects on
SiO2/Mo(112) Thin Films and Their Interaction with Au Atoms. The Journal
of Physical Chemistry B, 2006. 110(34): p. 17015-17023.
111
51. B.K. Min, W.T. Wallace, D.W. Goodman, Synthesis of a Sinter-Resistant,
Mixed-Oxide Support for Au Nanoclusters. The Journal of Physical Chemistry
B, 2004. 108(38): p. 14609-14615.
52. E. Ozensoy, B.K. Min, A.K. Santra, D.W. Goodman, CO Dissociation at
Elevated Pressures on Supported Pd Nanoclusters. The Journal of Physical
Chemistry B, 2004. 108(14): p. 4351-4357.
53. D. Ricci, G. Pacchioni, Structure of ultrathin crystalline SiO2 films on
Mo(112). Physical Review B, 2004. 69(16): p. 161307.
54. G.M. Rignanese, A. De Vita, J.C. Charlier, X. Gonze, R. Car, First-principles
molecular-dynamics study of the (0001) α-quartz surface. Physical Review B,
2000. 61(19): p. 13250-13255.
55. A.K. Santra, B.K. Min, D.W. Goodman, Ag clusters on ultra-thin, ordered
SiO2 films. Surface Science, 2002. 515(1): p. L475-L479.
56. T. Schroeder, M. Adelt, B. Richter, M. Naschitki, M. Bäumer, H.J. Freund,
Growth of well-ordered silicon dioxide films on Mo(112). Microelectronics
Reliability, 2000. 40: p. 841-844.
57. T. Schroeder, M. Adelt, B. Richter, M. Naschitki, M. Bäumer, H.J. Freund,
Epitaxial growth of SiO2 on Mo(112). Surface Review and Letters, 2000. 7(1-
2): p. 7-14.
58. T. Schroeder, J.B. Giorgi, M. Bäumer, H.J. Freund, Morphological and
electronic properties of ultrathin crystalline silica epilayers on a Mo(112)
substrate. Physical Review B, 2002. 66(16): p. 165422.
112
59. T. Schroeder, A. Hammoudeh, M. Pykavy, N. Magg, M. Adelt, M. Bäumer, H.J.
Freund, Single crystalline silicon dioxide films on Mo(112). Solid-State
Electronics, 2001. 45(8): p. 1471-1478.
60. J. Seifert, D. Blauth, H. Winter, Evidence for 2D-Network Structure of
Monolayer Silica Film on Mo(112). Physical Review Letters, 2009. 103(1): p.
017601.
61. J. Seifert, M. Busch, A. Schüller, D. Blauth, S. Wethekam, H. Winter,
Structure of ultrathin silica films on Mo(112) studied via Classical and
Quantum Mechanical Rainbow Scattering of Fast Atoms. Surface and
Interface Analysis, 2010. 42(10-11): p. 1575-1580.
62. J. Seifert, A. Schüller, H. Winter, J. Sauer, M. Sierka, Diffraction of fast
atoms during grazing scattering from the surface of an ultrathin silica film on
Mo(112). Physical Review B, 2010. 82(3): p. 035436.
63. J. Seifert, H. Winter, Structure of monolayer silica on Mo(112) investigated by
rainbow scattering under axial surface channeling. Surface Science, 2009.
603(17): p. L109-L112.
64. M. Sierka, Synergy between theory and experiment in structure resolution of
low-dimensional oxides. Progress in Surface Science, 2010. 85: p. 398-434.
65. M. Sierka, T.K. Todorova, S. Kaya, D. Stacchiola, D. Stacchiola, J.
Weissenrieder, J.L. Lu, H.J. Gao, S. Shaikhutdinov, H.J. Freund, J. Sauer,
Interplay between theory and experiment in the quest for silica with reduced
dimensionality grown on a Mo(112) surface. Chemical Physics Letters, 2006.
424: p. 115-119.
113
66. D. Stacchiola, M. Baron, S. Kaya, J. Weissenrieder, S. Shaikhutdinov, H.J.
Freund, Growth of stoichiometric subnanometer silica films. Applied Physics
Letters, 2008. 92(1): p. 011911.
67. T.K. Todorova, M. Sierka, J. Sauer, S. Kaya, J. Weissenrieder, J.L. Lu, H.J.
Gao, S. Shaikhutdinov, H.J. Freund, Atomic structure of a thin silica film on
a Mo(112) substrate: A combined experimental and theoretical study.
Physical Review B, 2006. 73(16): p. 165414.
68. T.K. Todorova, M. Sierka, J. Sauer, S. Kaya, J. Weissenrieder, J.L. Lu, H.J.
Gao, S. Shaikhutdinov, H.J. Freund, Erratum: Atomic structure of a thin
silica film on a Mo(112) substrate: A combined experimental and theoretical
study [Phys. Rev. B 73, 165414 (2006)]. Physical Review B, 2006. 73(24): p.
249902.
69. J. Weissenrieder, S. Kaya, J.L. Lu, H.J. Gao, S. Shaikhutdinov, H.J. Freund,
M. Sierka, T.K. Todorova, J. Sauer, Atomic Structure of a Thin Silica Film on
a Mo(112) Substrate: A Two-Dimensional Network of SiO4 Tetrahedra.
Physical Review Letters, 2005. 95(7): p. 076103.
70. S. Wendt, M. Frerichs, T. Wei, M.S. Chen, V. Kempter, D.W. Goodman, The
interaction of water with silica thin films grown on Mo(112). Surface Science,
2004. 565: p. 107-120.
71. S. Wendt, Y.D. Kim, D.W. Goodman, Identification of defect sites on oxide
surfaces by metastable impact electron spectroscopy. Progress in Surface
Science, 2003. 74: p. 141-159.
72. S. Wendt, E. Ozensoy, T. Wei, M. Frerichs, Z. Cai, M.S. Chen, D.W. Goodman,
Electronic and vibrational properties of ultrathin SiO2 films grown on Mo(112).
Physical Review B, 2005. 72(11): p. 115409.
114
73. J. Yang, E.G. Wang, Water adsorption on hydroxylated α-quartz (0001)
surfaces: From monomer to flat bilayer. Physical Review B, 2006. 73(3): p.
035406.
74. S.M. Auerbach, K.A. Carrado, P.K. Dutta, Handbook of zeolite science and
technology. 2003: CRC Press.
75. J. Cejka, A.Corma, S. Zones, Zeolites and catalysis : synthesis, reactions and
applications. 2010, Wiley-VCH: Weinheim.
76. I. Böszörményi, T. Nakayama, B. Mclntyre, G. Somorjai, Model silica-alumina
acid catalysts for surface science and catalysis studies prepared by argon ion
beam sputter deposition using HY-zeolite targets. Catalysis Letters, 1991.
10(5-6): p. 343-355.
77. D. Stacchiola, S. Kaya, J. Weissenrieder, H. Kuhlenbeck, S. Shaikhutdinov,
H.J. Freund, M. Sierka, T.K. Todorova, J. Sauer, Synthesis and Structure of
Ultrathin Aluminosilicate Films. Angewandte Chemie International Edition,
2006. 45(45): p. 7636-7639.
78. D. Löffler, J.J. Uhlrich, M. Baron, B. Yang, X. Yu, L. Lichtenstein, L. Heinke,
C. Büchner, M. Heyde, S. Shaikhutdinov, H.-J. Freund, R. Wlodarczyk, M.
Sierka, J. Sauer, Growth and Structure of Crystalline Silica Sheet on
Ru(0001). Physical Review Letters. 105(14): p. 146104.
79. B. Yang, W.E. Kaden, X. Yu, J.A. Boscoboinik, Y. Martynova, L. Lichtenstein,
M. Heyde, M. Sterrer, R. Wlodarczyk, M. Sierka, J. Sauer, S. Shaikhutdinov,
H.-J. Freund, Thin silica films on Ru(0001): monolayer, bilayer and three-
dimensional networks of [SiO4] tetrahedra. Physical Chemistry Chemical
Physics. 14(32): p. 11344-11351.
115
80. X. Yu, B. Yang, J.A. Boscoboinik, S. Shaikhutdinov, H.-J. Freund, Support
effects on the atomic structure of ultrathin silica films on metals. Applied
Physics Letters, 2012. 100(15): p. 151608-4.
81. J.A. Boscoboinik, X. Yu, B. Yang, F.D. Fischer, R. Wlodarczyk, M. Sierka, J.
Sauer, S. Shaikhutdinov, H.-J. Freund, Modeling Zeolites with Metal-
Supported Two-Dimensional Aluminosilicate Films. Angew. Chem. Int. Ed.,
2012. 51: p. 6005-6008.
82. S. Hüfner, Photoelectron spectroscopy : principles and applications. 2003,
Berlin: Springer.
83. D. Briggs, J.T.Grant, Surface analysis by Auger and X-ray photoelectron
spectroscopy. 2003, Chichester,: IM Publications.
84. J. Chastain, J.F.Moulder, W.F. Stickle, Handbook of x-ray photoelectron
spectroscopy: a reference book of standard spectra for identification and
interpretation of XPS data. 1992, Eden Prairie, Minn.: Perkin-Elmer Corp.
85. J.W. Niemantsverdriet, Spectrosopy in catalysis. 2007, Weinheim,: WILEY.
86. G.A. Somorjai, Chemistry in Two Dimensions, Surfaces. 1981, Ithaca. :
Cornell University Press
87. L.J. Clarke, Surface Crystallography. 1985: John Wiley & Sons Ltd.
88. L.J. Clarke, Principles of adsorption and reaction on solid surface. 1996:
Wiley.
89. D.P. Woodruff, T.A. Delchar, Modern techniques of surface science. 1994:
Cambridge university Press.
116
90. J. Bardeen, Tunnelling from a Many-Particle Point of View. Physical Review
Letters, 1961. 6(2): p. 57-59.
91. M. Wilms, New and versatile ultrahigh vacuum scanning tunneling
microscope for film growth experiments Review of Scientific Instruments 1998.
69.
92. R.G. Greenler, Infrared Study of Adsorbed Molecules on Metal Surfaces by
Reflection Techniques. The Journal of Chemical Physics, 1966. 44(1): p. 310-
315.
93. R.G. Greenler, Reflection Method for Obtaining the Infrared Spectrum of a
Thin Layer on a Metal Surface. The Journal of Chemical Physics, 1969. 50(5):
p. 1963-1968.
94. Y.J. Chabal, Surface infrared spectroscopy. Surface Science Reports, 1988.
8(5–7): p. 211-357.
95. P.J. Larkin, Infrared and Raman spectroscopy : principles and spectral
interpretation. 2011, Amsterdam: Elsevier.
96. K. Wandelt, Surface and interface science. 2012, Weinheim: Wiley.
97. F.M. Hoffmann, Infrared reflection-absorption spectroscopy of adsorbed
molecules. Surface Science Reports, 1983. 3: p. 107-192.
98. F.M. Hoffmann, Infrared reflection-absorption spectroscopy of adsorbed
molecules. Surface Science Reports, 1983. 3(3): p. 107-192.
99. C. Lamberti, A. Zecchina, E. Groppo, S. Bordiga, Probing the surfaces of
heterogeneous catalysts by in situ IR spectroscopy. Chem. Soc. Rev., 2010.
39: p. 4951-5001.
117
100. P.W. Sutter, J.I. Flege, E.A. Sutter, Epitaxial graphene on ruthenium. Nat
Mater, 2008. 7(5): p. 406-411.
101. J.L. Lu, H.J. Gao, S. Shaikhutdinov, H.J. Freund, Morphology and defect
structure of the CeO2(111) films grown on Ru(0001) as studied by scanning
tunneling microscopy. Surface Science, 2006. 600(22): p. 5004-5010.
102. D.R. Mullins, P.V. Radulovic, S.H. Overbury, Ordered cerium oxide thin films
grown on Ru(0001) and Ni(111). Surface Science, 1999. 429: p. 186-198.
103. X. Xu D. W. Goodman, Appl. Phys. Lett. 61,774 (1992)
104. F.J Grunthaner, R.P. Vasquez, B.F. Lewis, J. Masergian, A. Madhukar, Local
atomic and electronic structure of oxide/GaAs and SiO2/Si interfaces using
high-resolution XPS. Journal of Vacuum Science and Technology, 1979. 16(5):
p. 1443-1453.
105. F. Rochet, C. Poncey, G. Dufour, H. Roulet, C. Gruillot, F. Sirotti, Suboxides
at the Si/SiO2 interface: a Si2p core level study with synchrotron radiation.
Journal of Non-Crystalline Solids, 1997. 216(0): p. 148-155.
106. L. Lichtenstein, C. Büchner, B. Yang, S. Shaikhutdinov, M. Heyde, M. Sierka,
R. Wlodarczyk, J. Sauer, H.-J. Freund, The Atomic Structure of a Metal-
Supported Vitreous Thin Silica Film. Angewandte Chemie International
Edition. 51(2): p. 404-407.
107. L. Lichtenstein, M. Heyde, H.J. Freund, Crystalline-Vitreous Interface in Two
Dimensional Silica. Physical Review Letters. 109(10): p. 106101.
108. R. Włodarczyk, M. Sierka, J. Sauer, D. Löffler, J. Uhlrich, X. Yu, B. Yang,
I.M.N. Groot, S. Shaikhutdinov, H.-J. Freund, Tuning the electronic structure
118
of ultrathin crystalline silica films on Ru(0001). Physical Review B, 2012.
85(8): p. 085403.
109. P. Luches, F. Pagliuca, S. Valeri, Morphology, Stoichiometry, and Interface
Structure of CeO2 Ultrathin Films on Pt(111). The Journal of Physical
Chemistry C, 2011. 115(21): p. 10718-10726.
110. W. Weiss, M. Ritter, Metal oxide heteroepitaxy: Stranski-Krastanov growth
for iron oxides on Pt(111). Physical Review B, 1999. 59(7): p. 5201-5213.
111. U. Berner, K.-D. Schierbaum, Cerium oxides and cerium-platinum surface
alloys on Pt(111) single-crystal surfaces studied by scanning tunneling
microscopy. Physical Review B, 2002. 65(23): p. 235404.
112. C.N.R. Rao, P.V. Kamath, S. Yashonath, Molecularly adsorbed oxygen on
metals: electron spectroscopic studies. Chemical Physics Letters, 1982. 88(1):
p. 13-16.
113. B.A. Staskiewicz, J.R. Tucker, P.E. Snyder, The Heat of Formation of
Molybdenum Dioxide and Molybdenum Trioxide1. Journal of the American
Chemical Society, 1955. 77(11): p. 2987-2989.
114. F. Liebau, Structural chemistry of silicates : structure, bonding and
classification. 1971, Berlin Springer.
115. C. Martínez, A. Corma, Inorganic molecular sieves: Preparation, modification
and industrial application in catalytic processes. Coordination Chemistry
Reviews, 2011. 255(13–14): p. 1558-1580.
116. S.P. Davis, E.V.R. Borgstedt, S.L. Suib, Growth of zeolite crystallites and
coatings on metal surfaces. Chemistry of Materials, 1990. 2(6): p. 712-719.
119
117. G.I. Panov, Advances in Oxidation Catalysis; Oxidation of Benzene to Phenol
by Nutrous Oxide. CATTECH, 2000. 4(1): p. 18-31.
118. A. Zecchina, M. Rivallan, G. Berlier, D. Lamberti, G. Ricchiardi, Structure
and nuclearity of active sites in Fe-zeolites: comparison with iron sites in
enzymes and homogeneous catalysts. Physical Chemistry Chemical Physics,
2007. 9(27): p. 3483-3499.
119. J.A. Boscoboinik, X. Yu, B. Yang, S. Shaikhutdinov, H.J. Freund, Building
blocks of zeolites on an aluminosilicate ultra-thin film. Microporous
Mesoporous Mater., 2013. 165: p. 158-162.
120. E. Dempsey, Acid strength and aluminum site reactivity of Y zeolites. Journal
of Catalysis, 1974. 33(3): p. 497-499.
121. W. Loewenstein, The distribution of aluminum in the tetrahedra of silicates
and aluminates. American Mineralogist, 1954. 39(1-2): p. 92-96.
122. S. Bordiga, L. Regli, D. Cocina, C. Lamberti, M. Bjorgen, k.P. Lillerud,
Assessing the acidity of high silica chabazite H-SSZ-13 by FTIR using CO as
molecular probe: Comparison with H-SAPO-34. J. Phys. Chem. B, 2005. 109:
p. 2779-84.
123. L.J. Smith, A. Davidson, A.K. Cheetham, A neutron diffraction and infrared
spectroscopy study of the acid form of the aluminosilicate zeolite, chabazite
(H- SSZ-13). Catal. Lett., 1997. 49(3-4): p. 143-146 LA - English.
124. M. Sierka, J. Sauer, Proton Mobility in Chabazite, Faujasite, and ZSM-5
Zeolite Catalysts. Comparison Based on ab Initio Calculations. J. Phys. Chem.
B, 2001. 105(8): p. 1603-1613.
120
125. C. Tuma, J. Sauer, A hybrid MP2/planewave-DFT scheme for large chemical
systems: proton jumps in zeolites. Chemical Physics Letters, 2004. 387(4–6):
p. 388-394.
126. A. Zecchina, S. Bordiga, G. Spoto, D. Scarano, G. Petrini, G. Leofanti, M.
Padovan, C.O. Arean, Low-temperature Fourier-transform infrared
investigation of the interaction of CO with nanosized ZSM5 and silicalite. J.
Chem. Soc., Faraday Trans., 1992. 88(19): p. 2959-2959.
127. G. Busca, The surface acidity of solid oxides and its characterization by IR
spectroscopic methods. An attempt at systematization. Phys. Chem. Chem.
Phys., 1999. 1(5): p. 723-736.
128. G. Ketteler, W. Ranke, Heteroepitaxial Growth and Nucleation of Iron Oxide
Films on Ru(0001). The Journal of Physical Chemistry B, 2003. 107(18): p.
4320-4333.
129. V. Alexandrov, A. Neumann, M. Scherer, K.M. Rosso, Electron Exchange and
Conduction in Nontronite from First-Principles. The Journal of Physical
Chemistry C, 2013. 117(5): p. 2032-2040.
130. N. Anke, S. Michael, B.H. Thomas, Redox Properties of Structural Fe in
Smectite Clay Minerals, in Aquatic Redox Chemistry. 2011, American
Chemical Society. p. 361-379.
121
LIST OF FIGURES
Fig. 1.1 Top and cross views of “O-poor” (a) [68] and “O-rich” (b) [64] SiO2.5/Mo(112)
films.(O: small red, small blue; Si: big orange) ................................................. 4
Fig. 1. 2 Hydrogen bonded to an O atom bridging Si and Al, in zeolite mordenite. [73]
........................................................................................................................ 5
Fig. 2.1 Schematic illustration of the UHV system (MEGA machine). ......................... 9
Fig. 2.2 Schematic illustration of the process of for the creation of photoelectron. The
Auger process as one possible decay is included. ........................................... 10
Fig. 2.3 PES as a stree-step process. [82] ................................................................ 11
Fig. 2.4 Principle of X-ray photoelectron spectroscopy. ........................................... 12
Fig. 2.5 The mean free path of an electron depends on its kinetic energy, and
determines how much surface information it carries. Optimum surface
sensitivity is obtained with electrons in the 25 to 200 eV range. [86] .............. 16
Fig. 2.6 The principle of LEED. ............................................................................... 19
Fig. 2.7 Schematic representation of the diffraction process. ................................... 19
Fig. 2.8 The diagram of a real space c( 2 × 2 ) structure (left) and the corresponding
diffraction pattern (right). .............................................................................. 20
Fig. 2.9 The principle of tunneling between the surface and the metal tip for positive
tip bias V. Tunneling of electrons from occupied surface states into the metal
tip. [85] .......................................................................................................... 22
Fig. 2.10 Schematic illustration of the working principle of an STM instrument [91].
...................................................................................................................... 24
Fig. 2.11 Potential energy versus internuclear distance for a diatomic harmonic
oscillator. ....................................................................................................... 26
Fig. 2.12 Dipoles oriented parallel and perpendicular to the surface create image
dipoles within the metal surface, which leads to effective screening of dipole
122
moment changes parallel to the surface but reinforcement of dipole moment
changes normal to the surface. [96] ............................................................... 27
Fig. 2.13 Reflection of light at a surface. The electric field components are denoted by
Ep and Es. The component parallel to the surface for p-light (Ep//) is not
shown. [97] .................................................................................................... 28
Fig. 2.14 Schematic of the Infrared Reflection-Absorption Spectroscopy setup:
Michelson interferometer, ultrahigh vacuum chamber with sample, mercury
cadmium telluride detector. ........................................................................... 30
Fig. 3.1 O1s region in XP-spectrum of the of 3O(2×2) precovered Ru(0001) ............... 32
Fig. 3.2 LEED pattern of clean Ru(0001) (a); and 3O(2×2) precovered Ru(0001) (b). . 32
Fig. 3.3 STM images of 3O(2×2) precovered Ru(0001), V=1.2 V, I=0.12 nA (proved by
B. Yang). ........................................................................................................................ 33
Fig. 3. 4 O1s region in XP-spectrum of as deposited film (top), Si2p region in XP-
spectra of as deposited film with different photon energy (bottom). ................... 34
Fig. 3.5 IRA-spectrum of as deposited film. ....................................................................... 35
Fig. 3.6 Si2p region in XP-spectrum of monolayer film (top); O1s region in XP-
spectrum of monolayer film, deconvoluted into two peaks as indicated (bottom).
........................................................................................................................................ 37
Fig. 3.7 IRA-spectrum of monolayer film. .......................................................................... 38
Fig. 3.8 STM images of the monolayer silica film on Ru(0001). V=8 V, I=0.1 nA (a);
V=2.0 V, I=0.1 nA (b); The respective LEED pattern (at 60 eV) is shown as an
inset in (a). .................................................................................................................... 39
Fig. 3.9 STM image of the monolayer silica film on Ru(0001), V=1.2 V, I=0.1 nA (top);
The profile line is measured along the line indicated in STM (bottom). ............. 40
Fig. 3.10 Top and side views of the most stable structure of Si4O10·2O/Ru(0001)
found by DFT, Si yellow, O red, Ru gray[78]. .......................................................... 41
123
Fig. 3.11 Caption IRA-spectrum (a) and XP-spectrum (b) simulated for the monolayer,
Si4O10·2O/Ru(0001) structure model (see Fig. 3.10). The frequency in
computed IRA-spectrum is scaled by a factor of 1.0341.15 [77]. The bar height
in the XP-spectrum is proportional to the number of the respective O atoms in
the structure. BE shifts are given with respect to the lowest O1s state. ............ 41
Fig. 3.12 Si2p (hν = 200 eV) region in XP-spectrum of bilayer film (top). Normalized
XP-spectra of O1s (hν = 630 eV) region of 2 MLE film at different emission
angles (bottom). ............................................................................................................ 44
Fig. 3.13 IRA-spectrum of bilayer films prepared using 32O2 (black), 36O2 (red). The
blue bars show the position and relative intensity of the IR active vibrations
calculated by SFT for the structure shown in Fig. 3.15 (a). ................................. 45
Fig. 3.14 STM images of the bilayer silica film on Ru(0001). V=1.3 V, I=0.07 nA.
(provided by B. Yang) .................................................................................................. 46
Fig. 3.15 Structural models of bilayer silica films, “o-poor” (a) and “o-rich” (b). ......... 47
Fig. 3.16 IRA-spectrum of 1.5 MLE film. ............................................................................ 49
Fig. 3.17 STM image of the 1.5 MLE silica film on Ru(0001), V= 8.0 V, I=0.1 nA; The
respective LEED pattern (at 60 eV) is shown as an inset (provided by B. Yang).
........................................................................................................................................ 49
Fig. 3. 18 STM image of the bilayer silica film prepared by two sequential deposition
and oxidation steps of 1 MLE Si each, V= 2 V, I=0.1 nA The respective LEED
pattern (at 60 eV) is shown as an inset (provided by B. Yang). ............................ 50
Fig. 3. 19 (top) LEED patterns (at 60 eV) and STM images (bottom) of the bilayer
silica films prepared by slow (on the left) and fast (on the right) cooling after the
high temperature oxidation step in the film preparation. (Tunneling parameters:
2 V, 0.1 nA (left); 3.3 V, 0.1 nA (right).) .................................................................... 52
124
Fig. 3.20 STM images of bilayer silica film with different deposition temperature 300
K (a), V=10V, I=0.1 nA; and 100 K (b), V=5.4 V, I=0.14 nA. ................................. 53
Fig. 3.21 The Si2p (hν =200 eV) (top) and O1s (hν = 630 eV) (bottom) regions in XP
spectra of a bilayer film as a function of annealing temperature in UHV as
indicated and after sample reoxidation at 1140 K for 10 min in 2 × 10−6 mbar
O2 ................................................................................................................................... 55
Fig. 3.22 BE shifts measured from the spectra shown in Fig. 3.21. ............................. 56
Fig. 3. 23 UP spectra of o-rich film and o-poor film, light source: He I. ....................... 57
Fig. 3.24 STM images of the 4 MLE silica film on Ru(0001), V= 8.6 V, I=0.12 nA; The
respective LEED pattern (at 60 eV) is shown as an inset. (provided by B. Yang)
........................................................................................................................................ 58
Fig. 3.25 IRA-spectrum of 4 MLE film. ............................................................................... 58
Fig. 4.1 Si2p region (top) and O1s region (bottom) in XP-spectrum of 2 MLE film. ... 62
Fig. 4.2 IRA-spectrum of 2 MLE film on Pt(111). ..................................................... 63
Fig. 4.3 STM images of the 2 MLE silica film on Pt(111). V =4.4 V; I =0.1 nA (a);
V=1.3 V, I=0.13 nA (b). .................................................................................. 64
Fig. 4.4 (a) STM image of the 1 MLE film on Pt(111). Height profile along the A-B line,
indicated in (a), is shown in (b). The inset shows a close-up image
demonstrating the vitreous state of silica islands (cf., Fig. 4.4 (b)). The arrow
indicates silica stripes bridging the islands (tunneling parameters 0.8 V, 0.06
nA). ............................................................................................................... 66
Fig. 4.5 XP-spectra of Si2p region (top) and O1s region (bottom) of 1 MLE film (red),
and 2 MLE film (black)................................................................................... 67
Fig. 4.6 IRA-spectra of 1MLE film (red), 2 MLE film (black). .................................... 68
Fig. 4.7 the BE shift of Si2p (Black) and O1s (red) XP spectra of SiO2/Pt(111) as a
function of annealing temperature in UHV as indicated and after reoxidation.
..................................................................................................................... 69
125
Fig. 4.8 Schematic representation of a monolayer (a) and a bilayer (b) film on a metal
support. Top and cross views are shown. Si ion positions are in [SiO4]
tetrahedra is indicated by dots. ...................................................................... 71
Fig. 4.9 STM images of the bilayer silica films: crystalline bilayer on Ru(0001) (a),
vitreous bilayer on Ru(0001) (b), and bilayer silica film on Pt(111) (c). ............ 72
Fig. 4.10 STM images of the monolayer silica films on Mo(112) (a) [42], and on
Ru(0001) (b). .................................................................................................. 73
Fig. 5.1 Al2s region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red). ................................................................ 77
Fig. 5.2 Si2p region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red). ................................................................ 78
Fig. 5.3 O1s region in XP-spectra of bilayer silica film on Ru(0001) (black), and
Al0.35O0.65O2 film on Ru(0001) (red). ................................................................ 78
Fig. 5.4 IRA-spectrum of bilayer silica film on Ru(0001) (black), and Al0.2O0.8O2 film
on Ru(0001) (red). .......................................................................................... 79
Fig. 5.5 Large scale STM image of Al0.12Si0.88O2 film (a), high resolution STM images
in domain A (b), in domains B (c), V=0.15 V, I=0.07 nA. ................................. 80
Fig. 5.6 Large scale STM image of Al0.35Si0.65O2 film, V= 8.3 V, I=0.11 nA. ............... 81
Fig. 5.7 Angle resolved XP-spectra of Al0.25Si0.75O2 film, Al/Si normalized signals as a
function of the angle between analyzer and surface normal, α. ...................... 82
Fig. 5.8 IRA spectra of the Al0.4Si0.6O2 films, hydroxylated with H2O (black) and D2O
(red), recorded in 2×10-5 mbar CO atmosphere. The spectra were divided by
reference spectra taken before CO exposure. .................................................. 83
Fig. 5.9 IRA spectra of the OH-terminated surface of the Al0.4Si0.6O2 films upon
subsequent hydroxylation with D2O. .............................................................. 84
Fig. 5.10 IRA-spectra of FexSi1- xO2 films on Ru(0001) as a function of the Fe content
(x) as indicated. ............................................................................................. 86
126
Fig. 5.11 IRA-spectra of Fe0.2Si0.8O2 films on Ru(0001) before D2O adsorption (black),
and with D2O ice layer (red). .......................................................................... 88
Fig. 5.12 LEED patterns (60 eV) of the FexSi1- xO2 films on Ru(0001) at x = 0 (a), 0.2
(b), 0.5 (c), and 1 (d). The unit cells are indicated. .......................................... 89
Fig. 5.13 Large-scale (a) and high-resolution (b) STM images of the Fe0.2Si0.8O2 film.
..................................................................................................................... 90
Fig. 5.14 Si2p region in XP-specta of silica bilayer (black), Fe0.5Si0.5O2 (red) and
Fe(111) (blue). ............................................................................................... 92
Fig. 5.15 Fe2p region in XP-specta of silica bilayer (black), Fe0.5Si0.5O2 (red) and
Fe(111) (blue). ............................................................................................... 92
Fig. 5.16 Top: XP-spectra of the O1s core level in pure SiO2, FeO(111) and
Fe0.5Si0.5O2 films on Ru(0001) as indicated. Bottom: Spectral deconvolution for
the spectrum of the Fe0.5Si0.5O2 film. .............................................................. 93
Fig. 5.17 Top and side views of the most stable structure of Fe4Si4O162O/Ru(0001)
found by DFT, Si yellow, Fe dark violet, O red, Ru gray. (provided by R.
Włodarczyk) ................................................................................................... 97
Fig. 5.18 IRA-spectrum simulated for the structural models depicted in Fig. 5.16.
The spectra are scaled by a factor of 1.0341 (see ref. [77]). (provided by R.
Włodarczyk) ................................................................................................... 97
Fig. 5.19 Top and side views of an idealized nontronite (in polyhedral representation)
formed by tetrahedral silicate sheet (in blue) and octahedral Fe hydroxide sheet
(in yellow). Structural protons are seen in the top view as small balls. [128] .. 98
127
ABBREVIATIONS
Å Angstrom
BE Binding energy
BL Bilayer
CO Carbon monoxide
DFT Density functional theory
ESCA Electron spectroscopy for chemical analysis
FT Fourier-transform
IC Integrated circuit
IR Infrared
IRAS Infrared reflection absorption spectroscopy
LEED Low energy diffraction
LDOS Local density of states
MLE Monolayer equivalent
MOSFET Metal-oxide-semiconductor field effect transistor
MSSR Metal surface selection rule
nA Nano Ampere
PES Photoelectron spectroscopy
PVD Physical vapor deposition
STM Scanning tunneling microscopy
UPS Ultraviolet photoelectron spectroscopy
XPS X-ray photoelectron spectroscopy
128
129
List of Publications:
10. The atomic structure of a Fe-silicate ultrathin film grown on Ru(0001)
Xin Yu, R. Włodarczyk, J. A. Boscoboinik, B. Yang, F. Fischer, J. Sauer, S.
Shaikhutdinov, H.-J. Freund
In preparation
9. Probe Molecules on the Bridging Hydroxyls of Zeolites: a Surface Science
Approach
J. A. Boscoboinik, Xin Yu, E. Emmez, B. Yang, F. D. Fischer, R. Wlodarczyk, S.
Shaikhutdinov, Joachim Sauer, Hans-Joachim Freund
Journal: J. Phys. Chem. C, Volume: accepted.
8. Hydroxylation of metal supported sheet-like silica films
B. Yang, E. Emmez, W. E. Kaden, Xin.Yu, J. A. Boscoboinik, M. Sterrer, S.
Shaikhutdinov, H.-J. Freund
Journal: J. Phys. Chem. C, Volume: 117, Page: 8336–8344, Year: 2013
7. Building Blocks of Zeolites on an Aluminosilicate Ultra-Thin Film
J. A. Boscoboinik, Xin Yu, B. Yang, S. Shaikhutdinov and H.-J. Freund.
Journal: Microporous Mesoporous Mater., Volume: 165, Page: 158-162, Year:
2013
6. Thin Silica Films on Ru(0001): Monolayer, Bilayer, and Three-dimensional
Networks of [SiO4] Tetrahedra
B. Yang, W. Kaden, Xin Yu, J. Boscoboinik, Y. Martynova, L. Lichtenstein, M.
Heyde, M. Sterrer, R. Wlodarczyk, M. Sierka, J. Sauer, S. Shaikhutdinov, H.-J.
Freund
Journal: Phys. Chem. Chem. Phys., Volume: 14, Page: 11344-11351, Year: 2012
5. Modeling Zeolites with Metal-supported Two-dimensional Aluminosilicate Films
J. A. Boscoboinik, Xin Yu, B. Yang, F. D. Fischer, R. Wlodarczyk, M. Sierka, S.
Shaikhutdinov, J. Sauer, H.-J. Freund
130
Journal: Angew. Chem. Int. Ed., Volume: 51, Page: 6005-6008, Year: 2012
Modellierung von Zeolithen durch zweidimensionale Aluminosilicatfilme auf
Metallunterlagen
J. A. Boscoboinik, Xin Yu, B. Yang, F. D. Fischer, R. Wlodarczyk, M. Sierka, S.
Shaikhutdinov, J. Sauer, H.-J. Freund
Journal: Angew. Chem., Volume: 124, Page: 6107-6111, Year: 2012
4. Support Effects on the Atomic Structure of Ultrathin Silica Films on Metals
Xin Yu, B. Yang, J.A. Boscoboinik, S. Shaikhutdinov, H.-J. Freund
Journal: Appl. Phys. Lett., Volume: 100, Page: 151608-1-4, Year: 2012
3. Low Temperature CO Oxidation on Ruthenium Oxide Thin Films at Near-
atmospheric Pressures
Y. Martynova, B. Yang, Xin Yu, J.A. Boscoboinik, S. Shaikhutdinov, H.-J.
Freund
Journal: Catal. Lett., Volume: 142, Page: 657-663, Year: 2012
2. Tuning the Electronic Structure of Ultrathin Crystalline Silica Films on Ru(0001)
R. Wlodarczyk, M. Sierka, J. Sauer, D. Löffler, J.J. Uhlrich, Xin Yu, B. Yang,
I.M.N. Groot, S. Shaikhutdinov, H.-J. Freund
Journal: Phys. Rev. B, Volume: 85, Page: 085403-1-8, Year: 2012
1. Growth and Structure of Crystalline Silica Sheet on Ru(0001)
D. Löffler, J.J. Uhlrich, M. Baron, B. Yang, Xin Yu, L. Lichtenstein, L. Heinke, C.
Büchner, M. Heyde, S. Shaikhutdinov, H.-J. Freund, R. Wlodarczyk, M. Sierka, J.
Sauer
Journal: Phys. Rev. Lett., Volume: 105, Page: 146194-1-4, Year: 2010
131
Conference Contributions:
Xin. Yu, B. Yang, J.A. Boscoboinik, J.J. Uhlrich, D. Löffler, S. Shaikhutdinov, H.-J.
Freund: Support Effect on The Atomic Structure of Ultrathin Silica Films on Metal.
16th edition of the International Conference on Solid Films and Surfaces, July 2012,
Genoa (Italy)
Xin. Yu, J.A. Boscoboinik, B. Yang, S. Shaikhutdinov, H.-J. Freund: The atomic
structure of a Fe-silicate ultrathin film grown on Ru(0001). Deutsche Physicalische
Gesellschaft spring meeting, March 2013, Regensburg (Germany)
132
133
Curriculum Vitae
Xin Yu
Geboren am 29.08.1983
In Nei Mongol, China
Schulausbildung
1989-1995 Xilinhot 3rd primary school, Nei Mongol, China
1995-1998 Xilinhot 3rd middle school, Nei Mongol, China
1998-2001 Xinlingol Mongolian school, Nei Mongol, China
Studium
2001-2005 Bachelor of Natural Science, in Material Chemistry
China University of Geosciences, Beijing, China
2006-2009 Master of Natural Science, in Physical Chemistry
University of Science and Technology of China, Hefei, China
Supervisor: Prof. Weixin Huang
Thema: Photoelectron Spectroscopy Study of Au/TiO2(110)
model System
Promotion
2009-2013 Doktorandin an der Technischen Universität Berlin und
Forschungstätigkeit unter der Leitung von Prof. Hans-Joachim
Freund am Fritz-Haber-Institut der Max-Planck-Gesellschaft,
Abteilung Chemische Physik, Berlin, Deutschland
Thema: Ultra-thin silicate films on metal single crystals
134
135
Erklärung an Eides statt
Ich versichere an Eides statt, dass ich die vorliegende Arbeit selbständig
verfasst und nur die angegebenen Hilfsmittel und Quellen benutzt habe.
Berlin, 01. Juli 2013