Jackson S.D., Hargreaves J.S.J. Metal Oxide Catalysis
Подождите немного. Документ загружается.
284 6 Photoelectron Spectroscopy of Catalytic Oxide Materials
further increased the Te concentration (and possibly increased V and Nb as well),
while in the follow - up experiment in propene oxidation without water, Te partly
segregated back into the bulk. A post - reaction sample, after quenching the reac-
tion, taken from an atmospheric reactor with typical feed composition gave approx-
imately similar results to in situ experiments with water. The characteristic
enrichment/depletion profi les of M1 was further validated by LEIS experiments
[219, 220] . The surface composition thus cast doubts on the validity of the naive
view that the surface could be terminated by an ideal basal plane. On the contrary,
high - resolution TEM indicated that the crystals were encapsulated by a 1 nm thick
layer without long - range ordering [217] . It was concluded that the short - range
ordered terminating layer was necessary to accommodate the essential transport
of lattice oxygen and electrons for the selective catalytic function.
Table 6.4 shows some variation in the actual binding energies, which can be
partly related to different materials and partly to the lack of universality applied to
aligning the E
B
scale. The reference point is typically the C 1s line at somewhat
arbitrary energies (284.4 – 285 eV) and some publications do not indicate their ref-
erencing procedure at all. However, there are clearly some obvious tendencies to
observe in Table 6.4 . For example, most materials contain vanadium in mainly the
4+ formal oxidation state; molybdenum is usually fi tted to obtain two (6+ and 5+)
components, that is, its peak maximum is observed between the position of these
two oxidation states; the E
B
values of niobium are somewhat lower, and those of
Figure 6.12 Bulk sensitive Elemental
composition of an M1 material (M1#1),
(according to EDX) in comparison with its
surface composition at different conditions
as measured by XPS. XPS experiments were
carried out in situ . Conditions were from left
to right: 0.25 mbar C
3
H
8
+
O
2
mixture at 623 K;
0.25 mbar C
3
H
8
+
O
2
+
H
2
O at 623 K; 0.25 mbar
C
3
H
6
+
O
2
at 623 K; 0.3 mbar O
2
at room
temperature after catalytic reaction at 1 bar
(C
3
H
8
+
O
2
+
H
2
O + N
2
).
6.3 Case Studies 285
tellurium higher, than the corresponding binary (Nb
2
O
5
, TeO
2
) materials. These
observations become even clearer in Table 6.5 , in which the binding energies for
two M1 materials recorded in situ during propane oxidation are listed in compari-
son to reference materials measured, also in situ , under oxygen at elevated tem-
perature. This procedure of investigating the reference materials under in situ
conditions ensures that only clean materials are compared, which becomes impor-
tant owing to the small sampling depth probed by the low - KE electrons. The
insignifi cant difference in O 1s binding energy clearly indicates that the “ ionicity ”
of lattice oxygen of the reference binary systems is very similar, and that O 1s will
be an insensitive tool for obtaining information on oxygen species. On the other
hand, the more sensitive valence band might contain information that could facili-
tate the understanding of the electronic structure of binary and complex mixed -
metal oxides. Since the M1 phase contains mainly molybdenum (6+) oxide, we
shall fi rst examine the valence band of MoO
3
. Figure 6.13 depicts the measured
valence band using synchrotron radiation with different photon energies com-
pared with DFT cluster calculation. As stated previously, the formal oxidation
states would dictate an ionic picture with fi lled O 2p and empty Mo 4d orbitals.
From this picture, no signifi cant modifi cation of the valence band envelope is
expected, since the various symmetry components of oxygen 2p orbitals would not
be preferentially enhanced or weakened. The observed results are, however, truly
different. Intensity in the low - lying region of 6 – 9.5 eV is strongly attenuated at
lower excitation energies, that is, for lower KE electrons. Conversely, orbitals
Table 6.5 XPS binding energies of the M1 phase MoVTeNb
materials under reaction conditions (see Figure 6.12 ) in
comparison with reference binary oxides measured under
oxygen at elevated temperature (623 – 673 K). At these
conditions charging was practically not observed. Experiments
used low kinetic energy electrons, and hence were sensitive
to the topmost layers.
Sample O1s
(eV)
V2p (3/2)
(eV)
Mo3d (5/2)
(eV)
Nb3d (5/2)
(eV)
Te3d (5/2)
(eV)
V
2
O
5
530.1 517.2
MoO
3
530.5 232.6
Nb
2
O
5
530.1
a)
207.2
a)
TeO
2
530.1 – 530.4
b)
575.9
b)
M1#1 C
3
H
8
+ O
2
530.1 516 232.3
( ∼ 231)
206.4 576.1
M1#1 C
3
H
8
+ O
2
+ H
2
O 530.1 – 530.2 516 232.2
( ∼ 231)
206.4 576.2
M1#2 C
3
H
8
+ O
2
+ H
2
O 530.2 516.1 232.3
c)
206.6 576.2
a XPS Handbook: 530.6 eV; 207.5 eV.
b XPS Handbook: 530.2 eV; 575.7 eV.
c Measured in C
3
H
8
+ O
2
(without addition of water).
286 6 Photoelectron Spectroscopy of Catalytic Oxide Materials
around 4.3 eV are strongly enhanced. The changes can be easily understood con-
sidering the variations in the cross - section ( CS ) of corresponding orbitals. Typi-
cally all CS values increase towards lower energies, but their slope can vary. The
ratio of O - 2p/Mo - 4d CS strongly increases from 0.07 at 1041 eV to almost 1.6 at
151.4 eV (Table 6.6 ). Therefore the spectrum at low - energy excitation corresponds
mainly to the various oxygen orbitals, while the enhanced intensities at higher
energy are due to molybdenum 4d orbitals. All this is in perfect agreement with
the DFT calculations [68] shown for comparison. As a consequence, MoO
3
is not
fully ionic and there are appreciable covalent contributions. The calculated Mul-
liken populations show that the molybdenum centers can be described as Mo
2.2+
and the negative oxygen charges scale with coordination (from − 0.5 to − 1) [68,
221] . Therefore the bridging oxygens exhibit more ionic character and are more
prone to vacancy formation than the terminal oxygens, which are more covalent.
Moving to further reference materials listed in Table 6.5 , all of them, except
V
2
O
5
, exhibit similar CS variations (Table 6.6 ) giving the opportunity to investigate
their possible covalency by XPS. The excitation - energy - dependent valence bands
of Nb
2
O
5
and TeO
2
(not shown) indicate a qualitatively similar picture to that
detailed for MoO
3
, suggesting a similar mixed covalent/ionic structure. Although
the ratio of O 2p/V 3d CS values remains unchanged in the 150 – 1040 eV region,
Figure 6.13 Synchrotron based valence band
spectra of MoO
3
at different excitation
energies. MoO
3
was measured at 623 K in
0.5 mbar O
2
to remove carbon contamination
and minimize reduced molybdenum sites.
As comparison, the partial density of states
according to DFT cluster calculation [68] of
O 2p and Mo 4d is shown. DFT result is
reproduced with kind permission of K.
Hermann. The theoretical curves are aligned
to the experiment to coincide with the onset
of high - lying valence orbitals at ∼ 2.5 eV. (The
peaks above 10 eV in experiments correspond
to gas - phase O
2
signal, while those of the
calculation are due to bond saturation of
peripheral oxygen by hydrogen. Therefore this
latter has to be considered as a cluster
artefact).
6.3 Case Studies 287
DFT calculations, as discussed in Section 6.3.1 , yield a similar picture. Further-
more, modifi cation of the valence band envelope as a function of h ν was charac-
teristic for all M1 and reference catalysts. Therefore it should be stressed that
although we adopt the formalism of oxidation states it is essentially always only a
formal oxidation state, which is just a chemical tool but does not represent any
serious physical meaning, since the electron occupations are considerably
different.
Now, by comparing the valence - band region of catalytic materials with the refer-
ence oxides (Figure 6.14 ) we may hope to get a better understanding of the surface
electronic structure of our M1 phase. The comparison is made at 180 eV excitation
that probes mainly oxygen orbitals (see above). Oxygen 2p in the binary oxides is
clearly distinguishable by its unique features. Without going into detail we would
like to stress just two points. First, the valence bands of the reference oxides do
not show any intensity in the band gap ( ∼ 1 eV) that would be an indication of a
considerable number of oxygen vacancies reducing the corresponding metal
cations. Second, tellurium dioxide, because of its tellurium 5p electrons, shows a
measurable intensity at 2 eV, which is typically lower than the HOMO of the other
oxides. The valence - band spectra of the catalytic materials recorded under reaction
conditions are unexpectedly very similar, indicating that they are likely to share
the same structural unit at the surface. This becomes important considering the
widely different selectivity of the samples: MoV is unable to produce acrylic acid
( AA ) while the two samples with M1 phase reach a selectivity to AA of 70% or
more. All samples show spectral intensity in the band gap ( E
B
: 1 eV), which fi ts
well with the vanadium state being in its formal 4+ oxidation state. That the inten-
sity in the band gap is small should be no surprise considering the low vanadium
content at the top layers (Table 6.4 , Figure 6.14 ). Moreover, it shows clearly that
molybdenum cannot be reduced in signifi cant amounts, otherwise the band - gap
feature would be more intense [68] . It is also obvious that no V
2
O
5
- like
Table 6.6 Theoretical atomic subshell photoionization cross -
sections for MoVTeNb - related valence band orbitals at
different photon energies, according to the calculations of Yeh
and Lindau [177].
151.4 eV 200 eV 600 eV 1041 eV
O 2p 0.404 0.184 0.0061 0.000 89
V 3d; 4s 0.645; 0.028 0.311; 0.019 0.01; 0.0032 0.0015; 0.001
Mo 4d; 5s 0.254; 0.009 0.257; 0.0054 0.047; 0.0011 0.012; 0.0004
Nb 4d; 5s 0.191; 0.0086 0.184; 0.0052 0.03; 0.0011 0.0079; 0.0004
Te 5s; 5p 0.064; 0.06 0.044; 0.047 0.0083; 0.011 0.0033; 0.0049
O2p/V3d 0.63 0.59 0.61 0.59
O2p/Mo4d 1.59 0.72 0.13 0.07
O2p/Nb4d 2.12 1 0.2 0.11
O2p/Te5p 6.31 4.18 0.73 0.27
288 6 Photoelectron Spectroscopy of Catalytic Oxide Materials
micro - domain could exist in any of the catalysts. Further, the lack of intensity at
2 eV demonstrates that the high excess of tellurium cannot be in the form of binary
TeO
2
at the surface. Hence tellurium should be either in a local environment as
for the other cations, or in a formal oxidation state of 6+. The fi nal observation we
emphasize is the shift of both the O 2p maximum (3.7 eV) and the onset of 2p at
the top of valence band (2 eV) relative to the main constituent MoO
3
. This is in
line with the slight shift of the O1s peak shown in Table 6.5 .
Since Mo 3d is the most often reported and depicted core level in the MoVTeNb
literature, it is worth considering it here as well. In many publications, partially
reduced molybdenum (Mo
5+
) was reported (see Table 6.4 ) and its importance for
catalysis was considered. Unfortunately, in situ experiments (Figure 6.15 A) cannot
validate the sometimes relatively high concentration of reduced Mo species. Apart
from a 0.4 eV binding - energy shift of the main Mo 3d peak relative to the position
of MoO
3
, the concentration of Mo
5+
( ∼ 231 eV) in M1 is below 5% under propane
oxidation condition. This is in good agreement with the in situ XAS experiments
by Safanova and coworkers [222] , although the Mo K - edge is intrinsically bulk
sensitive. The main component of Mo 3d can be well represented by one relatively
narrow and symmetric Gauss – Lorenz function, indicating the presence of only
Figure 6.14 Synchrotron based valence band
spectra (h ν = 180 eV) of clean reference binary
oxides (A) and catalytic materials under
reaction conditions (B). The reference oxides
were recorded at elevated temperature (623 –
673 K) in 0.2 – 1 mbar oxygen to remove carbon
contamination and minimize reduction. The
M1 catalysts (1: M1#1, 2: M1#2) are under
propane oxidation conditions with addition of
steam to the feed. 3: Mo
5
O
14
- type MoVW -
oxide under propene oxidation with addition
of steam. 4: Mo
5
O
14
- type MoV - oxide under
propene oxidation. Reaction pressure: 0.25 –
0.6 mbar; temperature 623 K. (The small peaks
above 10 eV and around 6 eV correspond to
gas phase signals. Their positions relative to
the surface signal depend on the potential
fi eld in front of the sample and are thus not
constant on the conventional E
B
scale
referenced to the work function).
6.3 Case Studies 289
one species. That the situation may be rather more complex is illustrated in Figure
6.15 B showing in situ Mo 3d spectra of the MoV oxide. The bottom curve obtained
at room temperature obviously contains reduced Mo sites. In parallel, a signifi cant
amount of carbon was observed on the surface. However, as the sample was heated
up in the reaction mixture (here propene + O
2
), carbon and Mo
5+
disappeared (or
decreased strongly) thus underlining the correlation of the two species. During
reaction, no surface carbon could be distinguished and only the gas - phase signal
of the feed was observed. The Mo 3d signal became essentially symmetric and
could have been approximated by one broad symmetric function. The reason we
did not do this is shown in relation to the top curve observed after introduction of
water into the reaction feed. The peak is now clearly asymmetric and can be well
fi tted with two narrower functions. The position of the higher energy component
is close to the position of “ classical 6+ ” in the binary system and the low E
B
com-
ponent is situated at ∼ 232.0 eV. This latter peak can clearly not be interpreted
within the initial state approximation with a formal ionic picture (Mo
5+
: ∼ 231 eV).
If we disregard the ionic formalism, however, the binding - energy shift could rep-
resent a small partial reduction of the localized positive charge of the cation. This
can be the case if removal of an insignifi cant amount of lattice oxygen is accom-
panied by reorganization of the lattice metal polyhedra (from corner to edge
sharing) and the corresponding reduction could be delocalized throughout a whole
micro - domain. Another explanation could be that the E
B
is also a function of the
Figure 6.15 Synchrotron based Mo 3d spectra
of (A), 1: clean, insignifi cantly reduced MoO
3
,
2: M1#1, 3: M1#2. The spectra of M1
catalysts (2, 3) were recorded during propane
oxidation with steam (as shown previously).
Spectra in (A) were collected with high
surface sensitivity (h ν : 410 eV). (B) shows Mo
3d of Mo
5
O
14
- type MoV - oxide in 1: propene
oxidation feed at RT, 2 (and 3): propene
oxidation feed at 623 K without (with steam).
Spectra in (B) were recorded with 710 eV
excitation energy. The shaded component
corresponds to Mo
5+
.
290 6 Photoelectron Spectroscopy of Catalytic Oxide Materials
fi nal excited state. If the screening of the core hole can be modifi ed, this could
give rise to non - uniform binding energies. As pointed out by Wertheim [223] , this
can indeed happen with Mo 3d of, for example, K
0.3
MoO
3
, in which he identifi ed
two alternative screening channels, owing to the appearance of a conduction band
of Mo 4d character below the Fermi level. Since M1 is an n - type semiconductor
under reaction conditions, with high - lying electrons as the main charge carriers
(Figure 6.14 B), there is a possibility of substituting the classical (O 2p – Mo 3d)
charge - transfer screening process by a d – d transition. Owing to the low concentra-
tion of Mo
5+
species this would, however, indicate an “ interatomic ” V 3d – Mo 3d
screening process.
The somewhat low E
B
of Nb 3d in M1 as compared to Nb
2
O
5
and the asymmetry
in its envelope observed at each and every condition (not shown, but also noted
in ref. [207] ), would indicate similar fi nal - state effects being operative in defi ning
the niobium binding energy. It is interesting in this context that tellurium 3d has
negligible asymmetry in its peak form. The slightly higher E
B
, relative to TeO
2
,
may suggest a partially more ionic cation, as suggested by the lack of TeO
2
- related
intensity at 2 eV in the valence band, which could make oxygen insertion facile
from TeO
x
.
Turning now to the reaction side, it is generally accepted that the rate - limiting
step of propane selective (amm)oxidation is activation of the fi rst C
−
H bond.
Vanadium 5+ was proposed by several authors as being responsible for this ele-
mentary step [148, 204 – 206] . However, XPS does not seem to support this pro-
posal, since vanadium was observed in its formal 4+ oxidation state on the surface
of M1 (Tables 6.4 and 6.5 ). The defi ciency of the surface in vanadium and the
minority of its V
5+
fraction (if any) relative to V
4+
cast serious doubts that vanadium
in it highest oxidation state is responsible for the most demanding elementary
step of C
−
H bond activation of propane. Ueda and coworkers [212, 224] have
shown that the Te - and Nb - free MoV with M1 structure can activate propane as
effectively as MoVTeNb, although with poor selectivity towards AA. Taking into
account the mechanism proposed for butane activation on VPOs (Section 6.3.3.1 ),
it is more reasonable to consider the importance of Lewis basic V
4+
sites for the
fi rst C
−
H activation step. The suspiciously high tellurium content at the M1
surface indicates that Te is an essential catalytic component, promoting the forma-
tion of AA, possibly by the oxidation of a propene - like surface intermediate. As Te
is effective in allylic oxidation of olefi ns, the higher activity of Te - containing cata-
lysts for propene conversion over the Te - free M1 catalysts caused Ueda and
coworkers [212, 224] to conclude that the rapid allylic - hydrogen abstraction from
propene on the Te site to produce acrolein is the origin of the effective nature of
M1 - (Te). It is also well known [225 – 228] that MoV - based mixed oxides can convert
acrolein to AA with very high selectivity at almost full conversion, but at much
lower temperature. By using in situ diffuse refl ectance Fourier transform spectros-
copy, Vogel and coworkers [229] concluded that this reaction was facilitated by
Lewis acid sites on which acrolein adsorbs with its carbonyl group. On the other
hand, Tichy [226] stressed the importance of vanadium 4+ sites and the presence
of a high steam concentration, which will block Lewis acid sites and increase
Br ø nsted acidity, enabling desorption of acids from the surface. The high surface
Te concentration might be further benefi cial by diluting possible neighboring
acrolein oxidation sites, thus decreasing the temperature margin between acrolein
oxidation (500 K for M1 - phase MoV - oxide) and propane oxidation to avoid over -
oxidation. To conclude, the uniqueness of the MoVTeNb M1 phase materials
seems to be related to their ability to host a surface structure that facilitates the
elementary steps of propane (amm)oxidation and is able to supply oxygen from
the sub - surface hexagonal/heptagonal channel [208, 222] and transfer electrons
through its subsurface interlayer [217] . The dynamic character of the surface
(reversible modifi cation of surface composition) with the high similarity of the O
2p - related valence bands of M1 and model MoV oxide catalysts indicate that the
top few layers, although being high - resolution TEM amorphous, should posses a
certain structural motif that allows high specifi city during catalytic turnover if the
correct cationic centers are placed next to each other. The combination of the right
surface structures, diluted Lewis basicity and a large excess of stable TeO
x
(or
analogous) sites seems to hold the key to successful propane oxidation
chemistry.
6.4
Summary
The rational design of heterogeneous catalysts requires an understanding of
surface processes on a molecular level that inherently involves knowledge of the
active state of the solid surface. Since catalysis is a dynamic event, requiring a
fl exible and metastable site for catalytic turnover, the only plausible choice for
spectroscopic experiments is the application of a reaction - like environment during
observation. For experimentalists, PES provides a tool to investigate the surface
electronic structure of catalytically relevant materials. Recent developments in in
situ experimentation have now made it possible to investigate the active surface
state during catalytic turnover. The examples detailed in this chapter clearly indi-
cate the need for rapidly expanding the number of in situ studies in many catalytic
systems, as these will contribute to a greater and deeper understanding of the
working principles of heterogeneous catalysis. Further experimental development
(photon sources with higher brilliance, smaller aperture sizes, more effi cient lens
systems and better electron detection) will allow us to reach even more realistic
pressure conditions, which will broaden the spectrum of those catalytic reactions
that will be successfully implemented in in situ XPS investigations. Since the
valence and conduction band of the solid defi nes the “ playground ” of the surface
processes, it is very desirable that in situ studies progressively concentrate on these
spectral regions. However, as their interpretation requires massive support from
theoreticians, the new age of exploration of surface reactions requires substantial
joint efforts from experimentalists and theoreticians.
6.4 Summary 291
292 6 Photoelectron Spectroscopy of Catalytic Oxide Materials
Acknowledgments
The authors would like to thank their immediate coworkers who participated in
the BESSY experiments: M. H ä vecker, A. Knop - Gericke, P. Schn ö rch, S. Zafeiratos
and E. Kleimenov for their invaluable contribution to the experiments detailed in
this text. Support from the BESSY synchrotron in Berlin is gratefully acknowl-
edged. Finally thanks to the SfB Germany and Athena Project (jointly funded by
Johnson Matthey Catalysts and EPSRC of the UK) for fi nancial support.
References
1 Fadley , C.S. ( 1978 ) Basic Concept of
X - ray Photoelectron Spectroscopy , in
Electron Spectroscopy: Theory, Techniques
and Applications II (eds C.R. Brundle
and A.D. Baker ), Academic , New York ,
p. 2 .
2 H ü fner , S. ( 2003 ) Photoelectron
Spectroscopy; Principles and Applications ,
Springer , Berlin Heidelberg New York .
3 Siegbahn , K. , Nordling , C. , Johansson ,
G. , Hedman , J. , Heden , P.F. , Hamrin ,
K. , Gelius , U. , Bergmark , T. , Werme ,
L.O. , Manne , R. and Baer , Y. ( 1969 )
ESCA Applied to Free Molecules , North -
Holland Publishing Company .
4 Joyner , R.W. , Roberts , M.W. and
Yates , K. ( 1979 ) Surface Science , 87 ,
501 .
5 Ruppender , H.J. , Grunze , M. , Kong ,
C.W. and Wilmers , M. ( 1990 ) Surface
and Interface Analysis , 15 , 245 .
6 Ogletree , D.F. , Bluhm , H. , Lebedev , G. ,
Fadley , C.S. , Hussain , Z. and Salmeron ,
M. ( 2002 ) Review of Scientifi c
Instruments , 73 , 3872 .
7 Zhu , J.F. , Kinne , M. , Fuhrmann , T. ,
Denecke , R. and Steinr ü ck , H. - P. ( 2003 )
Surface Science , 529 , 384 .
8 St ö hr , J. ( 1992 ) NEXAFS Spectroscopy ,
Springer - Verlag , Berlin - Heidelberg,
New York .
9 Grant , J.T. ( 2003 ) Software for Processing
AES and XPS Data , IM Publications
and SurfaceSpectra Limited , Charton
Manchester .
10 Tougaard , S. ( 2003 ) Quantifi cation of
Nano - structures by Electron Spectroscopy ,
IM Publications and SurfaceSpectra
Limited , Charton Manchester .
11 Clark , D.T. and Dilks , A. ( 1976 ) Journal of
Polymer Science Part A – Polymer
Chemistry , 14 , 533 .
12 Bancroft , G.M. ( 1975 ) Chemical Physics
Letters , 32 , 173 .
13 Birkenstock , U. , Holm , R. , Reinfandt , B.
and Storp , S. ( 1985 ) Journal of Catalysis ,
93 , 55 .
14 Fuggle , J.C. and Martensson , N. ( 1980 )
Journal of Electron Spectroscopy and Related
Phenomena , 21 , 275 .
15 Gelius , U. ( 1974 ) Physica Scripta , 9 , 133 .
16 Wertheim , G.K. ( 1975 ) Journal of Electron
Spectroscopy and Related Phenomena , 6 , 239 .
17 Ivanovsky , A.L. , Gubanov , V.A. , Shveikin ,
G.P. and Kurmaev , E.Z. ( 1981 ) Journal of
the Less - Common Metals , 78 , 1 .
18 Larsson , S. ( 1976 ) Journal of Electron
Spectroscopy and Related Phenomena , 8 , 171 .
19 Jadoul , M.A. ( 1978 ) Etude des Spectres
E.S.C.A. des Halogenures de Fer(II) et
(III) , Ms Thesis, Facultes Universitaires
Notre - Dame de la Paix , Namur .
20 Morgan , W.E. , Van Wazer , J.R. and Stec ,
W. ( 1973 ) Journal of American Chemical
Society , 95 , 751 .
21 van der Laan , G. , Westra , C. , Haas , C.
and Sawatzky , G.A. ( 1981 ) Physical Review
B , 23 , 4369 .
22 Scrocco , M. ( 1981 ) Physical Review B , 23 ,
4381 .
23 Richter , B. , Kuhlenbeck , H. , Freund , H.J.
and Bagus , P.S. ( 2004 ) Physical Review
Letters , 93 , 26805 .
24 Wertheim , G.K. , DiCenzo , S.B. ,
Buchanan , D.N. and Bennett , P.A. ( 1985 )
Solid State Communications , 53 , 377 .
25 Mavrikakis , M. , Stoltze , P. and Norskov ,
J.K. ( 1999 ) Catalysis Letters , 64 , 101 .
References 293
26 Joseph , Y. , Besnard , I. , Rosenberger ,
M. , Guse , B. , Nothofer , H.G. , Wessels ,
J.M. , Wild , U. , Knop - Gericke , A. , Su ,
D.S. , Schl ö gl , R. , Yasuda , A. and
Vossmeyer , T. ( 2003 ) Journal of Physical
Chemistry B , 107 , 7406 .
27 Martensson , N. , Berndtsson , A. and
Nyholm , R. ( 1980 ) Journal of Electron
Spectroscopy and Related Phenomena , 19 ,
299 .
28 Martensson , N. and Johansson , B.
( 1979 ) Solid State Communications , 32 ,
791 .
29 Johansson , G. , Hedman , J. , Berndtsson ,
A. , Klasson , M. and Nilsson , R. ( 1973 )
Journal of Electron Spectroscopy and
Related Phenomena , 2 , 295 .
30 Jentoft , F.C. , Weinberg , G. , Wild , U. ,
Schl ö gl , R. and Wakai , F. ( 1999 )
Photoelectron Spectroscopy as a Tool
for Studying Ceramic Interfaces: A
Tutorial , Wiley - VCH Verlag GmbH ,
Weinheim .
31 Barr , T.L. ( 1989 ) Journal of Vacuum
Science and Technology A , 7 , 1677 .
32 Wark , M. , Koch , M. , Bruckner , A. and
Gr ü nert , W. ( 1998 ) Journal of the
Chemical Society – Faraday Transactions ,
2033 , 94 .
33 Hupfer , B. , Schupp , H. , Andrade , J.D.
and Ringsdorf , H. ( 1981 ) Journal of
Electron Spectroscopy and Related
Phenomena , 23 , 103 .
34 Storp , S. and Holm , R. ( 1977 ) Surface
Science , 68 , 10 .
35 Holm , R. and Storp , S. ( 1976 ) Applied
Physics , 9 , 217 .
36 Abu Haija , M. , Guimond , S. ,
Romanyshyn , Y. , Uhl , A. , Kuhlenbeck ,
H. , Todorova , T.K. , Ganduglia - Pirovano ,
M.V. , D ö bler , J. , Sauer , J. and Freund ,
H. - J. ( 2006 ) Surface Science , 600 ,
1497 .
37 Freund , H.J. ( 1995 )
Physica Status Solidi
B , 192 , 407 .
38 Freund , H.H. , Kuhlenbeck , H. ,
Staemmler , V. , Twerenbold , D. and
Webb , R.H. ( 1996 ) Reports on Progress
in Physics , 59 , 283 .
39 Shaikhutdinov , S. and Weiss , W. ( 2000 )
Journal of Molecular Catalysis A:
Chemical , 158 , 129 .
40 Che , M. and Bond , G.C. ( 1985 )
Adsorption and Catalysis on Oxide
Surfaces , Elsevier Science Publishers ,
Amsterdam .
41 Ketteler , G. , Weiss , W. , Ranke , W. and
Schl ö gl , R. ( 2001 ) Physical Chemistry
Chemical Physics , 3 , 1114 .
42 Boehm , H.P. ( 1966 ) Angewandte Chemie ,
78 , 617 .
43 Heber , M. and Gr ü nert , W. ( 2000 )
Journal of Physical Chemistry B , 104 , 5288 .
44 Chen , J.G. , Crowell , J.E. and Yates , J.T.
( 1986 ) Journal of Chemical Physics , 84 , 5906 .
45 Bukhtiyarov , V.I. , H ä vecker , M. , Kaichev ,
V.V. , Knop - Gericke , A. , Mayer , R.W. and
Schl ö gl , R. ( 2003 ) Physical Review B , 6723 ,
5422 .
46 Knop - Gericke , A. , H ä vecker , M. , Schedel -
Niedrig , T. and Schl ö gl , R. ( 2001 ) Topics
in Catalysis , 15 , 27 .
47 Che , M. and Tench , A.J. ( 1983 ) Advances
in Catalysis , 32 , 1 .
48 Xu , C. , Hassel , M. , Kuhlenbeck , H. and
Freund , H.J. ( 1991 ) Surface Science , 258 ,
23 .
49
Freund , H.J. , Dillmann , B. , Seiferth , O. ,
Klivenyi , G. , Bender , M. , Ehrlich , D. ,
Hemmerich , I. and Cappus , D. ( 1996 )
Catalysis Today , 32 , 1 .
50 Hermann , K. , Witko , M. , Druzinic , R. ,
Chakrabarti , A. , Tepper , B. , Elsner , M. ,
Gorschl ü ter , A. , Kuhlenbeck , H. and
Freund , H.J. ( 1999 ) Journal of Electron
Spectroscopy and Related Phenomena , 98 – 9 ,
245 .
51 Weiss , W. , Boffa , A.B. , Dunphy , J.C. ,
Galloway , H.C. , Salmeron , M.B. ,
Somorjai , G.A. , Freund , J. and Umbach ,
E. ( 1993 ) The Structure and Reactivity of
Oxide Thin Films Grown on Transition
Metal Surfaces , Springer Verlag .
52 Dupuis , A. - C. , Haija , M.A. , Richter , B. ,
Kuhlenbeck , H. and Freund , H. - J. ( 2003 )
Surface Science , 539 , 99 .
53 Rainer , D.R. and Goodman , D.W. ( 1998 )
Journal of Molecular Catalysis A: Chemical ,
131 , 259 .
54 Corneille , J.S. , He , J.W. and Goodmann ,
D.W. ( 1995 ) Surface Science , 338 , 221 .
55 Tessier , L. , Bordes , E. and Gubelmann -
Bonneau , M. ( 1995 ) Catalysis Today , 24 ,
335 .
56 Engeser , M. , Schlangen , M. , Schr ö der ,
D. , Schwarz , H. , Yumura , T. and
Yoshizawa , K. ( 2003 ) Organometallics , 22 ,
3933 .