Bowker M., Davies P.R. (Eds.) Scanning Tunneling Microscopy in Surface Science, Nanoscience and Catalysis
Подождите немного. Документ загружается.
two consecutive STM images of the coadsorbed NO/CO structure. The images were
taken 1 min apart, during which time a small number of rearrangement events
occurred (two are marked by arrows). In the analysis of 10 images with approximately
400 top sites per image, 64 events were recorded. Fifty-six of these events were a
single event of a top site NO adsorbate moving to an adjacent top site. The other
events were adsorbates that hopped two top sites or the appearance or disappearance
of a top site. Figure 7.25 includes a schematic for vacancy-mediated diffusion of
a high-coverage overlayer. The study also shows that a phase transition occurred
Figure 7.24 (a) 80 Å 80 Å STM image of Pt
in the presence of 1.2 10
5
Torr benzene at
300 K. Species is molecular benzene. Lines in
the [1 1 0]-type directions of the underlying
platinum lattice and a unit cell of the adsorbate
structure have been drawn. (b) Diagram of
proposed (2H3 2H3)R30.0
benzene
structure. Inset shows bonding structure.
214
j
7 Surface Mobility of Atoms and Molecules Studied with High-Pressure STM
through a phase boundary that had previously been observed during Torr pressure
STM studies of NO adsorption [42]. The high-pressure structure has not yet been
conclusively assigned [43–45]. The study of CO and NO conversion on Rh(1 1 1)
shows that high-pressure STM can be used not only to study the adsorbate structure
and its mobility but also to study the reaction kinetics of single-crystal model catalysis
under realistic conditions.
Figure 7.25 STM image: sequential (55 s delay) 200 Å 115 Å
images of Rh(1 1 1) with 0.50 Torr CO and 0.92 Torr NO
(I ¼260 pA, V ¼50 mV). The top arrow shows one top site
occupied by NO in the top image and by CO in the bottom one.
The bottom arrow shows the opposite.
7.5 Adsorbate Mobility During Catalytic Reactions
j
215
7.6
Summary
The possible pressure gap of adsorbates of reactant molecules and catalyst surface
structure between low-pressure and realistic high-pressure industrial catalysis con-
ditions drives the study of adsorption and surface structure at high pressures of
reactant gases. One of the most valuable techniques is high-pressure high-temper-
ature STM, with the capability of working at the entire pressure range, from UHV to
several bars of reactant gases, and at variable temperature. This technique and
instrumentation was presented on the basis of the recently developed high-pressure
high-temperature STM reactor/UHV system at Berkeley. Using high-pressure STM,
we revealed the pressure-dependent surface structure of several systems, including
hex-Pt(1 0 0) under an environment of CO at 300 K. The flexibility of the surface
structure and the mobility of the metal atoms of the catalyst surface were demon-
strated at high pressure. Our high-pressure high-temperature STM studies of
catalytic systems also revealed the mobility of reactant species at high pressure. We
also demonstrated that high-pressure STM can be used to study reaction kinetics for
single-crystal model catalysts.
Acknowledgments
This work was supported by the Director, Office of Science, Office of Basic Energy
Sciences, Division of Materials Sciences and Engineering of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231.
References
1 Binnig, G. and Rohrer, H. (1983)
Scanning Electron Microsc., 3, 1079–1082.
2 Mcintyre, B.J., Salmeron, M.B., and
Somorjai, G.A. (1992) Catal. Lett., 14,
263–269.
3 Somorjai, G.A., Tao, F., and Park, J.Y.
(2008) Top. Catal., 47,1–14.
4 Westerstrom, R., Wang, J.G., Ackermann,
M.D., Gustafson, J., Resta, A., Mikkelsen,
A., Andersen, J.N., Lundgren, E., Balmes,
O., Torrelles, X., Frenken, J.W.M., and
Hammer, B. (2008) J. Phys.: Condens.
Matter, 20, 184018.
5 Laegsgaard, E., Osterlund,L., Thostrup, P.,
Rasmussen, P.B., Stensgaard, I., and
Besenbacher, F. (2001) Rev. Sci. Instrum.,
72, 3537–3542.
6 Kolmakov, A. and Goodman, D.W. (2003)
Rev. Sci. Instrum., 74, 2444–2450.
7 Rasmussen, P.B., Hendriksen, B.L.M.,
Zeijlemaker, H., Ficke,H.G., and Frenken,
J.W.M. (1998) Rev. Sci. Instrum., 69,
3879–3884.
8 Rossler, M., Geng, P., and Wintterlin, J.
(2005) Rev. Sci. Instrum., 76, 023705.
9 Vestergaard, E.K., Vang, R.T., Knudsen, J.,
Pedersen, T.M., An, T., Laegsgaard, E.,
Stensgaard, I., Hammer, B., and
Besenbacher, F. (2005) Phys. Rev. Lett., 95,
126101.
10 Hendriksen, B.L.M., Bobaru, S.C., and
Frenken, J.W.M. (2005) Catal. Today, 105,
234–243.
11 Westerstrom, R., Gustafson, J., Resta, A.,
Mikkelsen, A., Andersen, J.N.,
Lundgren, E., Seriani, N., Mittendorfer,
F.,Schmid,M.,Klikovits,J.,Varga,P.,
Ackermann, M.D., Frenken, J.W.M.,
216
j
7 Surface Mobility of Atoms and Molecules Studied with High-Pressure STM
Kasper, N., and Stierle, A. (2007) Phys.
Rev. B, 76, 155410.
12 Tao, F., Tang, D., Salmeron, M., and
Somorjai, G.A. (2008) Rev. Sci. Instrum.,
79, 084101.
13 Somorjai, G.A. (1999) Introduction to
Surface Chemistry and Catalysis,
Wiley-VCH, Weinheim.
14 Somorjai, G.A. and McCrea, K.R. (2000)
Adv. Catal., 45, 385–438.
15 Cremer, P.S., Su, X.C., Shen, Y.R., and
Somorjai, G.A. (1996) J. Am. Chem. Soc.,
118, 2942–2949.
16 Shen, Y.R. (1989) Nature, 337, 519–525.
17 Zhu, X.D., Suhr, H., and Shen, Y.R. (1987)
Phys. Rev. B, 35, 3047–3050.
18 Ogletree, D.F., Bluhm, H., Lebedev, G.,
Fadley, C.S., Hussain, Z., and Salmeron,
M. (2002) Rev. Sci. Instrum., 73, 3872–3877.
19 Salmeron, M. and Schlogl, R. (2008) Surf.
Sci. Rep., 63, 169–199.
20 Tao, F., Grass, M.E., Zhang, Y.W., Butcher,
D.R., Renzas, J.R., Liu, Z., Chung, J.Y.,
Mun, B.S., Salmeron, M., and Somorjai,
G.A. (2008) Science, 322, 932–934.
21 Somorjai, G.A. (1981) Chemistry in Two
Dimensions: Surfaces, Cornell University,
Ithaca, NY.
22 Somorjai, G.A. (1991) Langmuir, 7,
3176–3182.
23 Mcintyre, B.J., Salmeron, M., and
Somorjai, G.A. (1993) J. Vac. Sci. Technol.
A, 11, 1964–1968.
24 Mortensen, K., Klink, C., Jensen, F.,
Besenbacher, F., and Stensgaard, I. (1989)
Surf. Sci., 220, L701–L708.
25 Batteas, J.D., Dunphy, J.C., Somorjai, G.A.,
and Salmeron, M. (1996) Phys. Rev. Lett.,
77, 534–537.
26 Blakely, D.W. and Somorjai, G.A. (1977)
Surf. Sci., 65, 419–442.
27 Vang, R.T., Laegsgaard, E., and Besen-
bacher, F. (2007) Phys. Chem. Chem. Phys.,
9, 3460–3469.
28 Tang, D.C., Hwang, K.S., Salmeron, M.,
and Somorjai, G.A. (2004) J. Phys. Chem. B,
108, 13300–13306.
29 Dunphy, J.C., Sa utet, P., Ogletree, D.F.,
Dabbousi, O., and Salmeron, M.B.
(1993) Phys.Rev.B, 47, 2320–2328.
30 Koeleman, B.J.J., Dezwart, S.T., Boers,
A.L., Poelsema, B., and Verheij, L.K. (1986)
Phys. Rev. Lett., 56, 1152–1155.
31 Steininger, H., Ibach, H., and Lehwald, S.
(1982) Surf. Sci., 117, 685–698.
32 Longwitz, S.R., Schnadt, J., Vestergaard,
E.K., Vang, R.T., Laegsgaard, E.,
Stensgaard, I., Brune, H., and Besen-
bacher, F. (2004) J. Phys. Chem. B, 108,
14497–14502.
33 Montano, M., Bratlie, K., Salmeron, M.,
and Somorjai, G.A. (2006) J. Am. Chem.
Soc., 128, 13229–13234.
34 Chen, P., Westerberg, S., Kung, K.Y.,
Zhu, J., Grunes, J., and Somorjai, G.A.
(2002) Appl. Catal. A – Gen., 229,
147–154.
35 Montano, M., Salmeron, M., and
Somorjai, G.A. (2006) Surf. Sci., 600,
1809–1816.
36 Bratlie, K.M., Montano, M.O., Flores,
L.D., Paajanen, M., and Somorjai, G.A.
(2006) J. Am. Chem. Soc., 128,
12810–12816.
37 Yang, M.C., Chou, K.C., and Somorjai,
G.A. (2004) J. Phys. Chem. B, 108,
14766–14779.
38 Koel, B.E., Blank, D.A., and Carter, E.A.
(1998) J. Mol. Catal. A – Chem., 131,
39–53.
39 Saeys, M., Reyniers, M.F., Marin, G.B.,
and Neurock, M. (2002) Surf. Sci., 513,
315–327.
40 Ertl, G., Neumann, M., and Streit, K.M.
(1977) Surf. Sci., 64, 393–410.
41 Rider, K.B., Hwang, K.S., Salmeron, M.,
and Somorjai, G.A. (2002) J. Am. Chem.
Soc., 124, 5588–5593.
42 Rider, K.B., Hwang, K.S., Salmeron, M.,
and Somorjai, G.A. (2001) Phys. Rev. Lett.,
86, 4330–4333.
43 Wallace, W.T., Cai, Y., Chen, M.S., and
Goodman, D.W. (2006) J. Phys. Chem. B,
110, 6245–6249.
44 Popa, C., Flipse, C.F.J., Jansen, A.P.J.,
van Santen, R.A., and Sautet, P. (2006)
Phys. Rev. B, 73, 245408.
45 Requejo, F.G., Hebenstreit, E.L.D.,
Ogletree, D.F., and Salmeron, M. (2004)
J. Catal., 226,83–87.
References
j
217
8
Point Defects on Rutile TiO
2
(1 1 0):
Reactivity, Dynamics, and Tunability
Chi L. Pang and Geoff Thornton
8.1
Introduction
Metal oxides are exploited in a number of technologies including gas sensing,
microelectronics, and catalysis [1, 2]. A number of oxides directly act as catalysts
while manyothers are employed as supports on which an active metal is dispersed. As
there is evidence that the oxide support influences the reactivity of the dispersed
metal [3, 4], understanding the behavior of oxides becomes important both when they
are used as supports and when they are the active component in a catalyst.
To date, rutile TiO
2
(1 1 0) is the most studied single-crystal TiO
2
surface and its
1 1 termination is considered the prototypical metal oxide surface for fundamental
studies [5]. There are a number of reasons why this surface has been studied so
extensively, such as the ready availability of high-quality crystals, its thermodynamic
stability, its significant electrical conductivity following vacuum annealing, and the
relative simplicity of its geometric structure. The vast majority of the literature
concerns work carried out in ultrahigh vacuum (UHV) that ensures clean, well-
defined surfaces, although there are efforts directed toward working both in ambient
pressures and in liquids [6–8].
The chemistry of both TiO
2
(1 1 0) and metal oxide surfaces in general has long
been thought to be dominated by reactions involving defects [9, 10]. Despite this, even
on the unreconstructed 1 1 phase of the model oxide – TiO
2
(1 1 0) – point defects in
scanning probe microscopy (SPM) images have not been well understood until very
recently. In this chapter, we explore how these point defects were unambiguously
identified in SPM images, then review a series of studies made possible by the
positive identification of point defects.
All the work we describe in this chapter was carried out in UHV on the rutile
TiO
2
(1 1 0)1 1 surface. Exposures to vapors and gases are given in Langmuirs (L)
where 1 L ¼1.333 10
6
mbar s. Coverages of defects or molecules adsorbed at the
surface will be given in monolayers (ML), where 1 ML corresponds to the density of
primitive surface unit cells.
j
219
8.2
Methods
We use commercial TiO
2
crystals (Pi-Kem) cut and polished to within 0.3
of the
(1 1 0) face and we prepare them further with cycles of Ar
þ
bombardment and UHV
annealing to approximately 950–1100 K, typically 5–10 min for each cycle. The
samples are mounted onto tantalum back-plates via strips of tantalum spot-welded
to the back-plate. Annealing is performed by high-energy electron bombardment of
the back-plate from a hot filament. Temperatures are measured from optical
pyrometers (Minolta) focused on the back-plate. The temperatures are not measured
directly from the samples because they are translucent and get darker with each
sputter/anneal cycle.
Our laboratories are currently equipped with three UHV Omicron microscopes, a
variable-temperature scanning tunneling microscope (STM), a room-temperature
atomic force microscope (AFM)/STM, and a low-temperature liquid helium bath
cryostat STM, all of which are currently driven by Omicron Scala software and
electronics.
All the STM results from ourgroup presented in this chapter employed the variable
temperature STM, with tips made by electrochemical etching of tungsten wire. For
noncontact AFM (NC-AFM), we employ commercial conducting silicon cantilevers
with force constants of approximately 2–14 N m
1
and resonant frequencies of
approximately 60–350 kHz (Nanosensors and Mikromasch). The NC-AFM images we
present here were recorded in collaboration with Professor Onishi at Kobe University
and employed a UHV JEOL (JSPM-4500A) microscope.
To reduce the exposure to residual gas from the vacuum, our samples are
sputtered/annealed in separate preparation chambers attached via valves to the SPM
chambers. After the final anneal cycle of the cleaning procedure, the sample is
transferred to the SPM chamber within a minute or so where the pressure is lower.
Whenever possible, we perform experiments in the same part of the surface before
and after some operation. The operation may involve exposure of the surface to a
reactant such as water or O
2
, for example; or it may be some switch in the
functionality of the microscope, for instance, imaging first in AFM mode and then
in STM mode. For reactions, we simply stop the scan and retract the tip by up to
1000 nm, which ensures it does not crash when we admit gases or vapors to the
chamber via the high-precision leak valves.
We have also exploited multichannel imaging. As described in Section 8.3,
we simultaneously recorded STM and NC-AFM images. Briefly, NC-AFM operates
by oscillating a cantilever (which carries the tip) at its resonant frequency. As the
tip a pproaches the surface, a force is felt by the cantilever and the resonant
frequency changes, that is, there is a frequency shift ( Df ). Typically, a topographic
map is then constructed by scanning the surface at a constant Df. Simultaneous
measurement of S TM can be ac hieved simply by applying a bias between the tip
and the sample during normal NC-AFM imaging and recording the tunneling
current.
220
j
8 Point Defects on Rutile TiO
2
(1 1 0): Reactivity, Dynamics, and Tunability
8.3
Water Dissociation at Oxygen Vacancies and the Identification of Point Defects
As discussed in the introduction, point defects are important on this surface as they
are thought to dominate surface reactivity. As for water, TiO
2
photocatalyzes its
dissociation and it is this photoactivity that underpins the use of TiO
2
in Green
Technology such as self-cleaning surfaces, air/water purification devices, and novel
solar cells. Apart from the intrinsic importance of water on TiO
2
, understanding
water chemistry is also a crucial issue because water is present in the environment in
almost all conceivable applications of TiO
2
.
A model of TiO
2
(1 1 0) is shown in Figure 8.1. The key features are the fivefold
coordinated Ti (Ti
5c
) rows that run in the [0 0 1] direction and alternate with bridging
O(O
b
) rows. In almost all published STM images of TiO
2
(1 1 0), tunneling is into
empty substrate states (positive bias) because tunneling from the substrate tends to
Figure 8.1 Schematic depiction of water dissociation on
TiO
2
(1 1 0). Blue and red spheres denote lattice O and Ti,
respectively. The light blue spheres are O
b
atoms, which lie in the
[0 0 1] azimuth of the substrate. Parallel Ti rows that lie between
the O
b
rows are Ti
5c
atoms. Green spheres indicate O atoms
bonded to H atoms that are shaded pink.
8.3 Water Dissociation at Oxygen Vacancies and the Identification of Point Defects
j
221
be unstable. Of these images, most show alternating bright and dark rows with
bright spots (type-A defects) appearing between the bright rows, as shown in
Figure 8.2b [5, 11–20]. Images that appear differently, such as those with dark
depressions on bright rows [12, 21] are poorly understood and will not be discussed
further here.
Density functional theory (DFT) simulations of STM images indicate that the Ti
5c
atoms should appear brighter than the O
b
atoms in STM images [11] despite the Ti
5c
atoms lying roughly 1.5 Å below the O
b
atoms [22, 23]. This is due to the dominance of
the Ti 3d states in the conduction band of TiO
2
. Although not explicitly simulated,
Diebold et al. [11] arguethat bridging oxygen vacancies (O
b
-vacs) should appear bright
and between bright Ti
5c
rows because the vacancies expose Ti sites. This was further
supported by STM simulations of a missing row of O
b
atoms that appear as a bright
row. As such, the bright spots seen between bright Ti
5c
rows in STM images were
assigned to O
b
-vacs [11, 12]. In contrast, based both on STM images and on electron
stimulated desorption experiments, Suzuki et al. [13] suggested that type-A defects
originate from surface hydroxyl (OH
b
). The hydrogen was thought to originate from
residual water vapor or dihydrogen in the vacuum chamber.
Brookes et al. [14] gave support to Suzuki et al.s interpretation. STM images were
recorded at low temperature (200 K), then warmed to room temperature (290 K) to
follow the dissociation of water. At low temperature, bright spots are seen on bright
Figure 8.2 175 Å 90 Å STM images of TiO
2
(1 1 0) exposed to
3.5 L H
2
O at 150 K, and consecutively heated to the measurement
temperatures of (a) 195 K and (b) 290 K. (Modified with
permission from Ref. [14].)
222
j
8 Point Defects on Rutile TiO
2
(1 1 0): Reactivity, Dynamics, and Tunability
Ti
5c
rows whereas at room temperature, bright spots appear between bright Ti
5c
rows
(see Figure 8.2). At low temperature, the bright spots were assigned to molecular
water adsorbed on Ti
5c
sites. This was corroborated in subsequent scanned energy
mode photoelectron diffraction (PhD) measurements, which show that water bonds
to Ti
5c
sites via its oxygen atom [24]. As for the bright spots seen in STM images at
room temperature, Brookes et al. [14] assigned these to OH
b
resulting from water
dissociation, on the basis that they have a similar concentration to those observed by
Suzuki et al. [13] on sputtered/annealed surfaces.
Although Brookes et al. show that OH
b
do indeed show up as bright spots between
the bright Ti
5c
rows, no indication was given as to the appearance of O
b
-vacs in STM
images. Schaub et al. [15] resolved this issue, in part, by showing that two kinds of
type-A defects coexist (see Figure 8.3). The brighter (A
b
) defects were assigned to O
b
-
vacs and the darker (A
d
) defects to OH
b
. Schaub et al. gave two reasons for their
assignment: (i) their DFTsimulations of STM images suggest that O
b
-vacs should be
brighter than OH
b
and (ii) the number of type-A defects after exposing TiO
2
(1 1 0) to
water was counted at twice that of the A
b
defects before exposure. This ratio is in line
with the expectation from other experimental work [14, 25–27], which suggested that
water dissociates in O
b
-vacs forming two OH
b
species.
STM experiments carried out over the last few years, however, show that the
assignment by Schaub et al. [15] is incorrect. One of the keys to correctly identifying
these defects was hinted at in studies in which the STM tip was scanned across the
TiO
2
(1 1 0) surface at a raised bias of þ3 V. Type-A defects were removed during
these high bias scans, the process being interpreted as tip-induced healing of O
b
-vacs
by Diebold et al. [12] and as tip-induced removal of hydrogen from OH
b
by
Suzuki et al. [13].
Bikondoa et al. [16, 17] applied electrical pulses (approximately þ3 V) to individual
type-A defects to remove them one by one. Although this is almost always
successful for A
b
defects (an example is shown in Figure 8.3), A
d
defects could
never be removed in this way. As such, this was used as a way of distinguishing the
two types of defects.
Figure 8.3 Sequential (150 Å)
2
STM images of
sputter/annealed TiO
2
(1 1 0). The densities of
A
b
and A
d
defects are approximately 1.5 and
3.0% ML, respectively. (a) Before the voltage
pulse and (b) after the voltage pulse. The voltage
pulse ( þ3 V, 0.35 nA, 300 ms) was applied to
the A
b
defect circled in (a). In (b), A
b
, and A
d
defects are marked with red and blue crosses,
respectively, and the pulsed A
b
defect is missing.
(Modified with permission from Ref. [16].)
8.3 Water Dissociation at Oxygen Vacancies and the Identification of Point Defects
j
223