Bowker M., Davies P.R. (Eds.) Scanning Tunneling Microscopy in Surface Science, Nanoscience and Catalysis

Подождите немного. Документ загружается.

Due to the increased complexity at high pressure (e.g., the impinging reactant

gases are no longer under molecular ﬂow, causing the gas environment near the

surface to be markedly different from the ambient gases), much more work is

required before establishing a ﬁrm correlation between the surface structure and the

surface reactivity. The difﬁculty in using STM to measure the reaction rate directly

Figure 3.22 QMS signals and STM images

simultaneously measured during CO oxidation

on the Pt(1 1 0) surface at a temperatureof 425 K

in a 3.0 ml min

1

flow of mixtures of CO and/or

at 0.5 bar. (Upper panel) QMS signals of

, CO, and CO

, measured directly from the

reactor cell. Labels (A)–(H) correspond to the

STM images in the lower panel. R

low

and R

high

denote the low and high CO

production rate

branches. P

indicates the threshold value

of the CO pressure at which the rate switched

from R

low

to R

high

and the surface changed

simultaneously from smooth to rough. (Lower

panel) STM images (210 nm 210 nm) from an

STMmovie(65s/image,I

¼0.2 nA,V

¼80 mV).

The images were differentiated to enhance the

contrast. Images (A), (B), (E), (F), and (H) show

flat terraces separated by steps of the Pt lattice.

This corresponds to the metallic, CO-covered

surface. Image (C) shows the change in the

surface accompanying the step in activity at

t ¼2109s (the image was built up from bottom

to top). Images (D) and (G) show the rough

surface consisting of protrusions,with heights of

0.2–0.4 nm, and pits (see inset). (Reprinted with

The American Physical Society.)

3 In Situ STM Studies of Model Catalysts

can lead to uncertainties in correlating local surface structure with the measured

global reaction rate.

STM measurements during high-pressure reaction present challenges for resolv-

ing the surface structures at high resolution. Rapid diffusion of surface adsorbates

further prevents precise characterization of surface-active species. In this case, the

application and development of the fast-scanning technique is essential and could

provide more precise surface information during the high-pressure reaction. None-

theless, the above study demonstrates the necessity of expanding high-pressure STM

studies in model catalysis. The effect of high pressure intuitively is equivalent to the

effect of lowering the temperature. Low-temperature UHV experiments might

reproduce the same adsorbate structures as do high pressures. However, due to

the restricted diffusion of metal atoms at low temperature, the dynamic change of

surface structures under catalytically realistic conditions could not be observed in the

UHV experiments.

3.5

In Situ Studies of Supported Model Catalysts

In an effort to bridge the material gap, in situ STM studies of supported model

catalysts have emerged in recent years. Among supported model catalysts, Au

clusters supported on TiO

(1 1 0) are the most investigated model system by in situ

STM. The tremendous interest in studying this model system originates from the

remarkable catalytic reactivity of supported Au clusters discovered a few decades

ago [73–80]. It was generally believed that Au, the noblest metal, is inert as a catalyst.

The TiO

support alone also exhibits limited catalytic reactivity. In contrast, Au

clusters in the 2–4 nm size range supported on TiO

exhibit a remarkable catalytic

reactivity for low-temperature CO oxidation and selective hydrogenation reactions.

Although the cause of the extraordinary activity of Au clusters is still under debate,

supported model Au catalysts are playing a key role in the development of our

understanding of the relative importance of cluster morphology and the cluster

support. In this section, we highlight a few recent in situ STM studies in monitoring

the synthesis of supported model catalysts, revealing the mechanism of the Strong

metal support interaction (SMSI) effect and measuring the sintering kinetics of

supported Au catalysts.

3.5.1

Monitoring the Growth Kinetics of Supported Metal Catalysts

Supported model catalysts are frequently prepared by thermally evaporating metal

atoms onto a planar oxide surface in UHV. The morphology and growth of supported

metal clusters depend on a number of factors such as substrate morphology, the

deposition rate, and the surface temperature. For a controlled synthesis of supported

model catalysts, it is necessary to monitor the growth kinetics of supported metal

3.5 In Situ Studies of Supported Model Catalysts

clusters. Figure 3.23 shows a series of STM images as a function of increased Au

coverage obtained on the same area of a TiO

(1 1 0) surface at room temperature [81].

At the very early stages of growth, Au clusters preferentially decorate the step edges

(Figure 3.23a). With an increase in Au coverage, the prevailing role of the step edges

as major nucleation sites decreases, whereas new clusters begin to nucleate on

terraces (Figure 3.23b–f). Au clusters grown at step edges grow much faster than

those at terraces. A quantitative analysis of Au cluster distribution suggests a bimodal

size distribution for the growth of Au clusters at step edges and terraces. In situ STM

studies allow the growth mode of Au clusters to be followed on a cluster-to-cluster

basis. The results demonstrate that step edges play a dominant role in the growth of

Au clusters on TiO

(1 1 0) at room temperature.

Using in situ STM also makes it possible to monitor the growth of supported alloy

model catalysts. The simplest way to synthesize supported alloy model catalysts is to

Figure 3.23 STM images of the same area of a TiO

(1 1 0) surface

after the deposition of (a) 0.17 ML Au; (b) 0.34 ML Au; (c) 0.51 ML

Au; (d) 0.69ML Au; (e) 0.86 ML Au; and (f) 1.3 ML Au. All images

have the same dimensions of 100 nm 100 nm. The circle

highlighted in white in each image indicates the identical area.

The Japan Society of Applied Physics.)

3 In Situ STM Studies of Model Catalysts

evaporate both metal atoms in UHVdirectly onto an oxide support surface. However,

this method does not give information on whether the two metals mix and form alloy

clusters or separate to form two types of monometallic clusters. In situ STM allows

this problem to be addressed by investigating the growth kinetics of supported alloy

catalysts.

In our study, Ag–Au alloy clusters of varying ratios were synthesized on TiO

(1 1 0)

and the growth kinetics for the clusters determined [82]. To understand the growth of

Au clusters in the presence of Ag clusters, sequential in situ STM measurements were

carried out as a function of Au coverage on a Ag precovered TiO

(1 1 0) surface.

A particular precoverage of 0.08 ML Ag (Figure 3.24a) was chosen to essentially

saturate the step edge sites with Ag clusters, allowing the investigation of Au growth

in the absence of their preferred nucleation sites. The series of in situ STM images

Figure 3.24 STM images (100 nm 100 nm) during Au

deposition on an Ag precovered (0.08 ML) TiO

(1 1 0) surface (a)

after deposition of (b) 0.17 ML Au, (c) 0.85 ML Au; or on an Ag

precovered (0.033 ML) TiO

(1 1 0) surface (d) after deposition of

(e) 0.17 ML Au, (f) 0.51 ML Au. The white circles in image (d) and

(e) show the appearance of new Au clusters. (Reprinted with

3.5 In Situ Studies of Supported Model Catalysts

(Figure 3.24a–c) show a preferential growth of Au on the existing Ag clusters. As the

Au coverage is increased, new Au clusters begin to appear on the surface. In

another set of experiments, the TiO

(1 1 0) surface was precovered with less Ag

so that a fraction of empty step edge sites remain open for the nucleation of new

Au clusters (Figure 3.24d). The series of in situ STM images (Figure 3.24d–f) show

even a small amount of Au deposition (0.17 ML) leads to the nucleation of new Au

clusters.

These systematic experiments of Au growth on TiO

(1 1 0) precovered with Ag

clusters show the versatility of in situ STMexperiments in controlling the synthesis of

supported alloy model catalysts. Simply by blocking step edge sites to a varying extent,

either monometallic clusters or bimetallic clusters could be selectively synthesized

and monitored. By changing the sequence of metal deposition, the surface structure

of alloy clusters can also be altered. Adopting the same in situ approach, the

synthesis of supported alloy model catalysts has been extended to Pd/Au clusters

on TiO

(1 1 0) [83] and Pt/Rh clusters on TiO

(1 1 0) [84]. It is expected that this

method will be generally adopted for the preparation and characterization of

supported model catalysts.

3.5.2

Studies of the SMSI Effect

In the late 1970s, Tauster et al. [85–87] discovered the unusual properties of group

VIII metal, for example, Pt a nd Ir,when supported on TiO

and reduced at relatively

high temperatures. The Pt and Ir clu sters supported on TiO

show a suppressed CO

and H

chemisorption and an increased metha nation catalytic activity. Strong

metal support interaction has since be en introduced to describe the unusual

properties of group VIII metal s upported on reducible oxides such as TiO

.Afew

reasons for the SMSI effect have been postulated, including alloy formation,

altered electronic effects at the interface, or encapsulation of supported metal

clusters.

Bowker et al. [88, 89] have investigated the SMSI effect using in situ STM. As shown

in Figure 3.25, Pd clusters with a mean size of approximately 4 nm were supported

on TiO

(1 1 0) and heated to 673 K. The surface was then exposed to 5 10

8

mbar O

and monitored by STM as a function of time. In the presence of oxygen, the

reoxidation of the TiO

(1 1 0) surface proceeds. Since the dissociation probability

of O

on Pd is about three magnitudes higher than on TiO

(1 1 0), O

dissociation

mainly occurs on supported Pd clusters. The dissociated oxygen atoms subsequently

diffuse onto the adjacent surface, that is, spillover, and coalesce with Ti interstitials

from the bulk, causing the enrichment of TiO

species surrounding the Pd clusters.

The regrowth of TiO

layers thus accelerates at the periphery of Pd clusters.

Eventually, the growth of TiO

layers surpass the height of the Pd clusters and

fully encapsulates them within the TiO

overlayers, that is, in situ STM study

veriﬁes that encapsulation is a major contributor to the SMSI effect. This is also

the ﬁrst study to have directly observed in situ the spillover, an elementary step in

catalytic process.

3 In Situ STM Studies of Model Catalysts

3.5.3

Sintering Kinetics of Supported Au Clusters

Although supported Au catalysts have great potential in technological applications,

such as removing CO selectively from hydrogen fuel for fuel cells, a major problem

for the commercialization of supported Au catalysts is that these catalysts deactivate

rapidly [78, 90, 91]. Similar deactivation has been observed for model catalysts

consisting of Au clusters deposited on a reduced TiO

(1 1 0) where rapid deactivation

has been shown to correlate with sintering of the Au clusters [91, 92]. To understand

the sintering of supported Au catalysts, we have studied the sintering kinetics of Au/

TiO

(1 1 0) in UHV and in the presence of reactant gases.

The ﬁrst issue in sintering kinetics is to determine the primary mass transport

process. The sintering of supported nanoclusters can occur in either one or two

modes: (a) the migration and coalescence of whole clusters or (b) the migration of

monomers (single metal atoms or metal complexes). The ﬁrst mode, known as

cluster migration, entails the migration of whole clusters on the surface that then

coalesce with neighboring clusters. In the second mode, known as Ostwald ripening,

monomers dissociate from small clusters, diffuse to, and coalesce with large

clusters [93, 94].

Figure 3.25 A sequence of STM images taken

at 673 K of oxygen spillover from Pd clusters in

an ambient pressure of 5 10

8

mbar of O

(doubled before image (f)). The total exposures

to oxygen from image (a)–(f) were (in units

of L): 114, 178, 237, 282, 344, and 531. The

spillover begins with the formation of one layer

around the Pd clusters, which then becomes

multilayer growth, eventually burying the Pd

cluster with TiO

. In image (f), a total of seven

layers of TiO

have grown around the Pd on top

of the original TiO

surface. (Reprinted with

Elsevier)

3.5 In Situ Studies of Supported Model Catalysts

In situ STM allows one to monitor individual clusters under realistic conditions

and to directly determine the mass transport mode controlling the kinetics. Mitchell

et al. studied Au clusters supported on TiO

(1 1 0) at approximately 750 K as a

function of time [95]. Their STM results show Au clusters are mostly static while

some small clusters with size below 3 nm diffuse along the surface. The diffusion of

small clusters does not behave in a continuous manner. Instead, the clusters stick for

a relatively long time before taking a sudden jump. This unconventional diffusion of

clusters, termed as Levy ﬂight, leads to the speculation that Au clusters sinter in the

mode of cluster migration. Such results are in agreement with the in situ STM study

by Kolmakov et al. [12], who examined the stability of Au/TiO

(1 1 0) after thermal

treatment at 950 K. Dramatic morphological changes were observed. Accompanying

the decomposition of surface steps, the density of Au clusters decreased by 20%,

whereas the volume of the Au clusters increased. Most Au clusters changed their

positions on the TiO

(1 1 0) surface. Clearly, cluster migration has played a role in the

thermally induced sintering of supported Au clusters under UHV. Interestingly, the

above studies were both conducted at the temperature under which the TiO

(1 1 0)

substrate becomes unstable and starts to decompose. To fully explore the sintering

mode induced by thermal treatment, it is necessary to conduct such in situ experi-

ments at intermediate temperatures where the TiO

(1 1 0) lattice remains stable.

The sintering of Au/TiO

(1 1 0) was also studied by our group in the presence of

reactant gases [96]. Figure 3.26 shows the morphological changes of 0.5 ML Au

clusters supported on TiO

(1 1 0) in the presence of CO/O

at 300 K. Under UHV

conditions, the Au clusters remain immobile and show no detectable change of either

cluster size or shape for more than 4 h. Sintering of Au clusters began immediately

upon the introduction of a 0.1 Torr CO/O

gas mixture (1 : 1 ratio). Figure 3.26 shows

no apparent change of cluster positions before and after gas exposure; however, the

small Au clusters gradually decay whereas the large Au clusters grow in size. The

gradual change in the cluster size and static cluster positions are consistent with an

Ostwald ripening process. Most Au clusters with a height around 4.6 Å or less

(assuming the height of Au clusters of two-layer thickness is 4.6 Å [97]) disappear

within 2 h of CO oxidation reaction. Gold clusters supported on TiO

(1 1 0) were also

exposed to O

and CO separately to investigate the inﬂuence of the chemisorbed

reactants. Minimal changes were observed for 0.5 ML Au clusters supported on

TiO

(1 1 0) after an exposure of 0.1 Torr O

or 3 Torr CO for more than 1.5 h at 300 K.

Therefore, a synergetic effect of CO/O

was found to induce and accelerate the

sintering of Au clusters on TiO

(1 1 0).

The synergetic effect of a CO and O

mixture can be explained by a reaction-

induced mechanism, which might involve hot electrons generated by CO oxida-

tion [98–100], inducing the detachment of Au monomers from supported Au clusters

and initiating the sintering process. On the basis of the sintering model of Ostwald

ripening, the activation energy for Au clusters with diameters around 3 nm is

estimated to be approximately 10 kJ mol

1

, a value that closely tracks the activation

energy of CO oxidation on supported Au clusters. The large difference in comparing

this number with the activation energy of sintering for supported Au clusters in UHV,

approximately 280 kJ mol

1

[101, 102], demonstrates that CO oxidation itself does

3 In Situ STM Studies of Model Catalysts

indeed inﬂuence cluster sintering. This effect may very well be general for CO

oxidation over other supported metal catalysts.

3.6

Outlook

This chapter reviews the recent progress in in situ STM studies of model catalysts.

From revealing reaction pathways to delineating active sites, in situ STM studies in

UHVand on extended surfaces have demonstrated their power to solve fundamental

questions in catalysis and enhance our understanding of the elementary steps of

Figure 3.26 75 nm 75 nm in situ STM images

of0.5 ML Auclusterssupportedon the TiO

(11 0)

surface in the presence of 0.1 Torr CO and O

mixture at 300 K. (a–f ) are consequently taken

at the same surface area. The time intervals are

(a) 0 min; (b) 28 min; (c) 42 min; (d) 63 min; (e)

120 min; and (f) 280 min. Tunneling parameters

are V

¼2 V and I

¼0.1 nA. (Reprinted with

American Chemical Society.)

3.6 Outlook

surface catalytic processes. Recent efforts in expanding in situ STM studies to high

pressures and to supported model catalysts illustrate the need to study model

catalysts under realistic conditions.

The studies discussed in this chapter have shown surface structures are extremely

dynamic under reaction conditions, consistent with the hypothesis promoted by

Somorjai [103–107]. The hypothesis, termed as the ﬂexible surface, supposes that

metal surfaces with low coordinated metal atoms or curved cluster surfaces are highly

ﬂuxional during reaction. Those atoms with high diffusivity are responsible for

high activity of metal surfaces. To this end, it is critical to compare the timescale and

energetics of metal atom diffusion with their catalytic reactivity. With the develop-

ment of fast scan techniques, in situ STM now provides a great opportunity to test this

hypothesis. The studies to date have shown that small can be quite different and

less can sometimes mean more. In the coming years, there is no doubt that in situ

STM studies on well-deﬁned nanostructures will exponentially increase and

contribute invaluably to an atomic-level understanding of reactions catalyzed by

solid surfaces.

Acknowledgments

We gratefully acknowledge the support for this work by the Department of Energy

(DOE), Ofﬁce of Basic Energy Sciences, Division of Chemical Sciences, and the

Robert A. Welch Foundation.

References

1 Langmuir, I. (1915) Phys. Rev., 6, 79.

2 Bartels, L., Wang, F., Moller, D., Knoesel,

E.,andHeinz,T.F.(2004)Science,305 , 648.

3 Che, M. and Bennett, C.O. (1989) Adv.

Catal., 36, 55.

4 Wintterlin, J., Trost, J., Renisch, S.,

Schuster, R., Zambelli, T., and Ertl, G.

(1997) Surf. Sci., 394, 159.

5 Kuipers, L., Hoogeman, M.S., and

Frenken, J.W.M. (1993) Phys. Rev. Lett., 71,

3517.

6 Ludwig, C., Gompf, B., Glatz, W.,

Petersen, J., Eisenmenger, W., Mobus,

M., Zimmermann, U., and Karl, N. (1992)

Z. Phys. B, 86, 397.

7 Rost, M.J., Crama, L., Schakel, P., Van Tol,

E., Van Velzen-Williams, G.B.E.M.,

Overgauw, C.F., Horst, H., Dekker, H.,

Okhuijsen, B., Seynen, M., Vijftigschild,

A., Han, P., Katan, A.J., Schoots, K.,

Schumm, R., Van Loo, W., Oosterkamp,

T.H., and Frenken, J.W.M. (2005) Rev. Sci.

Instrum., 76, 053710.

8 Rossler, M., Geng, P., and Wintterlin, J.

(2005) Rev. Sci. Instrum., 76, 023705.

9 Laegsgaard, E., Osterlund, L., Thostrup,

P., Rasmussen, P.B., Stensgaard, I., and

Besenbacher, F. (2001) Rev. Sci. Instrum.,

72, 3537.

10 Rasmussen, P.B., Hendriksen, B.L.M.,

Zeijlemaker, H., Ficke, H.G., and

Frenken, J.W.M. (1998) Rev. Sci. Instrum.,

69, 3879.

11 Mcintyre, B.J., Salmeron, M., and

Somorjai, G.A. (1993) Rev. Sci. Instrum.,

64, 687.

12 Kolmakov, A. and Goodman, D.W. (2002)

Chem. Rec., 2, 446.

13 Kolmakov, A. and Goodman, D.W. (2003)

Rev. Sci. Instrum., 74, 2444.

3 In Situ STM Studies of Model Catalysts

14 Melmed, A.J. (1991) J. Vac. Sci. Technol. B,

9, 601.

15 Oliva, A.I., Romero, A., Pena, J.L.,

Anguiano, E., and Aguilar, M. (1996) Rev.

Sci. Instrum., 67, 1917.

16 Weinstein, V., Slutzky, M., Arenshtam, A.,

and Benjacob, E. (1995) Rev. Sci. Instrum.,

66, 3075.

17 Ren, B., Picardi, G., and Pettinger, B.

(2004) Rev. Sci. Instrum., 75, 837.

18 Fried, G.A., Wang, X.D., and Hipps, K.W.

(1993) Rev. Sci. Instrum., 64, 1495.

19 Iwami, M., Uehara, Y., and Ushioda, S.

(1998) Rev. Sci. Instrum., 69, 4010.

20 Schroder, U., Mcintyre, B.J., Salmeron,

M., and Somorjai, G.A. (1995) Surf. Sci.,

333, 337.

21 Reichelt, R., Gunther, S., Rossler, M.,

Wintterlin, J., Kubias, B., Jakobi, B., and

Schlogl, R. (2007) Phys. Chem. Chem.

Phys., 9, 3590.

22 Hendriksen, B.L.M., Bobaru, S.C.,

and Frenken, J.W.M. (2005) Top. Catal.,

36, 43.

23 Hendriksen, B.L.M. and Frenken, J.W.M.

(2002) Phys. Rev. Lett., 89, 046101.

24 Wintterlin, J. (2000) Adv. Catal., 45, 131.

25 Hahn, J.R. and Ho, W. (2001) Phys. Rev.

Lett., 87, 166102.

26 Hla, S.W. and Rieder, K.H. (2003) Annu.

Rev. Phys. Chem., 54, 307.

27 Rieder, K.H., Meyer, G., Hla, S.W.,

Moresco, F., Braun, K.F., Morgenstern,

K., Repp, J., Foelsch, S., and Bartels, L.

(2004) Philos. Trans. Roy. Soc., 362,

1207.

28 Wong, K.L., Rao, B.V., Pawin, G.,

Ulin-Avila, E., and Bartels, L. (2005)

J. Chem. Phys., 123, 201102.

29 Longwitz, S.R., Schnadt, J., Vestergaard,

E.K., Vang, R.T., Laegsgaard, E., Stens-

gaard, I., Brune, H., and Besenbacher, F.

(2004) J. Phys. Chem. B, 108, 14497.

30 Vestergaard, E.K., Thostrup, P., An, T.,

Laegsgaard, E., Stensgaard, I., Hammer,

B., and Besenbacher, F. (2002) Phys. Rev.

Lett., 88, 259601.

31 Osterlund, L., Rasmussen, P.B.,

Thostrup, P., Laegsgaard, E., Stensgaard,

I., and Besenbacher, F. (2001) Phys. Rev.

Lett., 86, 460.

32 Thostrup, P., Vestergaard, E.K., An, T.,

Laegsgaard, E., and Besenbacher, F.

(2003) J. Chem. Phys., 118, 3724.

33 Vang, R.T., Wang, J.G., Knudsen, J.,

Schnadt, J., Laegsgaard, E., Stensgaard, I.,

and Besenbacher, F. (2005) J. Phys. Chem.

B, 109, 14262.

34 Mitsui, T., Rose, M.K., Fomin, E.,

Ogletree, D.F., and Salmeron, M. (2003)

Nature, 422, 705.

35 Salmeron, M. (2005) Top. Catal., 36, 55.

36 Mitsui, T., Rose, M.K., Fomin, E.,

Ogletree, D.F., and Salmeron, M. (2003)

Surf. Sci., 540,5.

37 Diebold, U. (2003) Surf. Sci. Rep.,

48, 53.

38 Diebold, U., Lehman, J., Mahmoud, T.,

Kuhn, M., Leonardelli, G., Hebenstreit,

W., Schmid, M., and Varga, P. (1998)

Surf. Sci., 411, 137.

39 Schaub, R., Thostrup, R., Lopez, N.,

Laegsgaard, E., Stensgaard, I., Norskov,

J.K., and Besenbacher, F. (2001) Phys. Rev.

Lett., 87, 266104.

40 Wendt, S., Schaub, R., Matthiesen, J.,

Vestergaard, E.K., Wahlstrom, E.,

Rasmussen, M.D., Thostrup, P., Molina,

L.M., Laegsgaard, E., Stensgaard, I.,

Hammer, B., and Besenbacher, F. (2005)

Surf. Sci., 598, 226.

41 Bikondoa, O., Pang, C.L., Ithnin, R.,

Muryn, C.A., Onishi, H., and Thornton,

G. (2006)

Nat. Mater., 5, 189.

42 Wendt, S., Matthiesen, J., Schaub, R.,

Vestergaard, E.K., Laegsgaard, E.,

Besenbacher, F., and Hammer, B. (2006)

Phys. Rev. Lett., 96, 066107.

43 Zhang, Z., Bondarchuk, O., Kay, B.D.,

White, J.M., and Dohnalek, Z. (2006)

J. Phys. Chem. B, 110, 21840.

44 Li, S.C., Zhang, Z., Sheppard, D., Kay,

B.D., White, J.M., Du, Y., Lyubinetsky, I.,

Henkelman, G., and Dohnalek, Z. (2008)

J. Am. Chem. Soc., 130, 9080.

45 Zhang, Z.R., Bondarchuk, O., White,

J.M., Kay, B.D., and Dohnalek, Z. (2006)

J. Am. Chem. Soc., 128, 4198.

References