Interaction of sulfate groups with the surface of zirconia: An HRTEM characterization study

Article abstract of DOI:10.1006/jcat.1996.0231

High-resolution transmission electron microscopy (HRTEM) has been used to characterize the morphology and the surface structure at an atomic level of sulfated zirconia and sulfate-free zirconia. Our study shows that HRTEM can be used to directly observe sulfate layers adsorbed on the surface of zirconia crystallites. The results indicate that the presence of sulfate groups not only stabilizes the tetragonal zirconia phase, but also induces the formation of well-faceted small zirconia crystallites. In particular, it is observed that the presence of sulfate groups induces the preferential formation of relatively long-flat (110) plane of tetragonal zirconia. HRTEM images of this plane revealed the presence of an adsorbed sulfate layer. It is proposed that the geometry of the (110) plane is such that it can accommodate sulfate groups in a two-or threefold coordination. Although, zirconia crystallites with relatively long-flat {110} planes containing adsorbed sulfate groups were clearly predominant in the sulfated zirconia catalyst; its role in the formation of the highly acidic sites is not clear. Our study also reveals the presence of few zirconia crystallites containing rough surfaces, crystallographically speaking high-Miller-index surfaces, which if they were to contain sulfate groups, they could be the locus of the highly acidic sites. Thus, it is suggested that caution must be taken when performing spectroscopic studies using techniques such as IR, NMR, XPS, and Raman to distinguish between spectator sulfate groups, which could be a majority if they were the ones observed on the (110) plane, and those participating in the formation of the highly acidic site, which could be associated to the few high-Miller-index-containing zirconia crystallites.

Full text of DOI:10.1006/jcat.1996.0231

JOURNAL OF CATALYSIS 161, 694–703 (1996)
ARTICLE NO. 0231
Interaction of Sulfate Groups with the Surface of Zirconia: An HRTEM
Characterization Study
M. Bena ¨ı ssa, J. G. Santiesteban,�,1
�
G. D ı´ az, C. D. Chang, and M. Jos e´ -Yacam a´ n
�
Instituto de F ´ı sica, Universidad Nacional Aut o´ noma de M e´ xico, A.P. 20-364, M e´ xico D.F. 01000, Mexico; and Mobil Research and Development Corp.,
Central Research Laboratory, Princeton, New Jersey 08540
Received October 24, 1995; revised March 12, 1996; accepted March 25, 1996
superacid activity (1, 2), facilitates paraffin isomerization
High-resolution transmission electron microscopy (HRTEM) has reactions under conditions much milder, near room temper-
been used to characterize the morphology and the surface structure ature, then those used with conventional solid acid catalysts,
at an atomic level of sulfated zirconia and sulfate-free zirconia. Our
study shows that HRTEM can be used to directly observe sulfate
layers adsorbed on the surface of zirconia crystallites. The results
indicate that the presence of sulfate groups not only stabilizes the
tetragonal zirconia phase, but also induces the formation of well-
faceted small zirconia crystallites. In particular, it is observed that
the presence of sulfate groups induces the preferential formation
of relatively long-flat (110) plane of tetragonal zirconia. HRTEM
images of this plane revealed the presence of an adsorbed sulfate
for example, zeolites. Despite extensive characterization
studies using a great variety of spectroscopic techniques
(
3–12), the nature and structure of the highly acidic sites
2�
on SO /ZrO2 are still a matter of controversy. An under-
4
standing of the nature of the highly acidic sites could be
reached by a good knowledge of the factors which control
the surface structure of the catalyst. High-resolution trans-
mission electron microscopy (HRTEM) is a powerful tool
layer. It is proposed that the geometry of the (110) plane is such that which can provide direct images of a projected structure of
it can accommodate sulfate groups in a two- or threefold coordina- individual metallic particles and/or oxide surfaces (13–16).
tion. Although, zirconia crystallites with relatively long-flat {110} In the present study, information at an atomic level on local
planes containing adsorbed sulfate groups were clearly predomi-
nant in the sulfated zirconia catalyst; its role in the formation of the
highly acidic sites is not clear. Our study also reveals the presence of
few zirconia crystallites containing rough surfaces, crystallograph-
est parallel to the electron beam. This allows one to record
ically speaking high-Miller-index surfaces, which if they were to
atomic detail of both the bulk and the one-dimensional pro-
contain sulfate groups, they could be the locus of the highly acidic
sites. Thus, it is suggested that caution must be taken when perform-
ing spectroscopic studies using techniques such as IR, NMR, XPS,
surfaces of sulfate-free zirconia and sulfated zirconia par-
ticles has been obtained using HRTEM imaging. Our ap-
proach consisted of aligning the crystallite surface of inter-
jection of the structure of the crystallite edge. Elucidation
of the structure of the crystallite edge, which is interpreted
in terms of the crystallite surface, may yield atomic-scale
and Raman to distinguish between spectator sulfate groups, which
could be a majority if they were the ones observed on the (110) plane, information on the surface structure. HRTEM images have
and those participating in the formation of the highly acidic site, been recorded around Scherzer defocus, so that every dark
which could be associated to the few high-Miller-index-containing spot corresponds to a single atomic column and/or assem-
zirconia crystallites. �c 1996 Academic Press, Inc.
bly of atomic columns. This has the advantage of avoiding
Fresnel-like oscillations obtained when working at white
contrast defocus, which generally confuses surface details.
The HRTEM images presented in this study are devoted
to help in understanding the correlation between surface
structure, which governs the nature of electronic surface
states, and the high acid activity exhibited by the sulfated
zirconia catalyst.
1
. INTRODUCTION
In the past few years, sulfated zirconia catalyst has been
reported to be the most promising solid acid candidate
to replace liquid acids and halogen-containing solid acids,
which are highly corrosive and are subject to stringent en-
vironmental regulations. The high acid activity of the sul-
fated zirconia catalyst, referred to by some researchers as
2. EXPERIMENTAL PROCEDURE
2
.1. Materials
Sulfated zirconia, SO /ZrO2, was prepared by using
1
To whom correspondence should be addressed at current address: Air
2
�
Products and Chemicals, Inc., 7201 Hamilton Boulevard, Allentown, PA
4
1
8195-1501. Fax: (610) 418-7719.
a two-step procedure. First, Zr(OH)4 was synthesized by
694
0021-9517/96 $18.00
Copyright �c 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

HRTEM STUDY OF SULFATE GROUPS INTERACTIONS
695
precipitating a zirconyl nitrate (ZrO(NO3)2) solution with
an aqueous NH4OH solution (28% NH3). The precipitated
gel, Zr(OH)4, was washed thoroughly with deionized wa-
ter, filtered, and air dried at 383 K for 14 h. In the second
step, the dried Zr(OH)4 was impregnated with a 1 N aque-
ous solution of (NH4)2SO4. The resulting slurry was dried
at 383 K for 14 h, followed by dry air calcination at 998 K
for 1 h. Sulfur content, as determined by chemical analysis,
2
�
was 1.33 wt%, which corresponds to 4.0 wt% SO . The
4
2
BET surface area was 90 m /g. The estimated surface con-
centration of sulfate groups, assuming a diameter for the
2�
2�
2
SO4 anion of 0.58 nm, is 2.9 SO /nm , or approximately
4
half a monolayer in terms of overall surface coverage. It
has been reported that at this surface density, SO2 groups
�
4
exist as a mixture of isolated sulfates and complex polynu-
clear sulfates (5).
Two sulfate-free ZrO2 samples were also prepared for
comparison purposes. They were prepared by calcining
Zr(OH)4, which was prepared according to the above pro-
cedure, at two different temperatures, 773 and 998 K. These
samples are designated in the text and figures as ZrO2(773)
and ZrO2(998), and their BET surface areas were 90 and
2
2
5 m /g, respectively.
2
.2. High-Resolution Transmission Electron Microscopy
HRTEM was carried out using a JEM4000EX electron
microscope operated at 400 keV. Electron micrographs
were recorded at magnification of 600,000� . Samples were
supported on holey carbon-coated copper grids by simply
grinding the specimen between two glass plates and bring-
ing the powder into contact with the grid. All images were
recorded on particles which protruded over openings in the
carbon support of the copper grids.
Powder X-ray diffraction analyses were also carried out
for phase identification of the sample.
FIG. 1. X-ray diffraction patterns of (a) sulfated zirconia calcined at
998 K, (b) sulfate-free zirconia calcined at 773 K, ZrO2(773), and (c) sulfate-
free zirconia calcined at 998 K, ZrO2(998).
3
. RESULTS AND DISCUSSION
3
.1. Particle Size, Morphology, and Crystal Structure
2
�
XRD pattern of the SO /ZrO2 sample, Fig. 1a, indicates
It is apparent from these XRD results, and the results
4
that the crystalline form of ZrO2 is mainly tetragonal. An of authors (2, 17), that the presence of SO2 stabilizes the
�
4
�
aditional small peak that appears at the angle 2� = 28.2
metastable tetragonal ZrO2 and retards the sintering pro-
may be interpreted as a presence of monoclinic zirconia cess. The stabilization of the metastable tetragonal ZrO2
traces. has also been reported to take place with the incorporation
Figures 1b and 1c show the XRD patterns of sulfate- of many different metal oxides such as MgO, CaO, Sc2O3,
free ZrO2 samples calcined at 773 and 998 K, respectively. Y2O3, La2O3, and CeO2, and the mechanism governing the
These XRD patterns reveal that a mixture of tetragonal process of stabilization is still not fully understood (18).
and monoclinic phases is present in the ZrO2(773) material, Although it is generally accepted that the tetragonal phase
2
�
Fig. 1b, whiletheZrO2(998) materialcontainssolelythemon- is present in the highly acidic SO /ZrO2 catalysts, it is not
4
oclinic phase. In addition, the sharp XRD peaks exhibited clear if it plays any direct role in the formation of the highly
bytheZrO2(998) materialindicatethatasignificantsintering, acidic sites present in this material.
which gives rise to large crystallite sizes, took place during
the high-temperature calcination.
Now, let us discuss the morphology and particle size of
the ZrO2 samples. Figure 2a shows a typical region of the

¨
BENAISSA ET AL.
6
96
FIG. 2. Typical low-magnification electron micrographs for (a) sulfated zirconia calcined at 998 K, (b) sulfate-free zirconia calcined at 773 K,
ZrO2(773), and (c) sulfate-free zirconia calcined at 998 K, ZrO2(998).
sulfated ZrO2. The average particle size of this material
Typical low-magnification electron micrographs of the
remains relatively small, <10 nm, despite the high calcina- sulfate-free ZrO2 materials, ZrO2(773) and ZrO2(998), are
tion temperature, 998 K, used in its preparation. The most presented in Figs. 2b and 2c. The particle size in the case
striking feature of the sulfated ZrO2 is the abundance of of ZrO2(773), Fig. 2b, is in the 10 nm range, and their shape is
faceted particles. This result indicates that the SO2 groups more or less rounded, indicating defective (or stepped) ter-
�
4
not only retard the sintering process of the tetragonal ZrO2 minations. Considerablesintering, asalsorevealedbyXRD,
crystallites but also decrease to a great extent their defec- is observed in the high-calcination temperature sulfate-free
tive terminations.
ZrO2(998) sample, Fig. 2c. Its average particle size is >20 nm.

HRTEM STUDY OF SULFATE GROUPS INTERACTIONS
697
FIG. 2—Continued
3
.2. Surface Atomic Structure
Surface defects are of great interest to catalysis since
they contain low-coordination number atoms with coordi-
natively unsaturated bonds ready to serve as adsorption
sites. Therefore, one could argue that the defects observed
on the sulfate-free zirconia crystallites are seats of acid ac-
tivity; however, the acid strength of these sites must be too
weak since pure ZrO2 does not catalyze reactions that de-
In this section, we compare the surface atomic structure
of the sulfated ZrO2, material calcined at 998 K, with that
of the sulfate-free ZrO2(773), calcined at 773 K. Only the
tetragonal ZrO2 particles of the former are discussed, as
this is the phase present in the sulfated ZrO2 catalyst.
3.2.a. ZrO2(773). Typical high-resolution TEM micro- mand relatively strong acid sties, such as butane isomeri-
graphs of crystallites observed in the ZrO2(773) material are zation.
shown in Figs. 3 and 4. Two crystallographic orientations
were frequently observed; namely h111i (Fig. 3) and h010i
3.2.b. Sulfated zirconia. As in the case of the sulfate-
Fig. 4). The zirconia crystallite surfaces are free of con- free zirconia, the h010i and h111i orientations were also
(
tamination so that their surface morphology can be char- clearly predominant in the sulfate-modified material. A
acterized at the atomic level. Along the h111i orientation, HRTEM image of a sulfated zirconia crystallite along the
Fig. 3, long {101} plane terminations are clearly visible, to- h010i orientation is presented in Fig. 5. The contrast exhib-
gether with some short {011} surface planes that give rise ited in this image is similar to that of Fig. 4, although here
to surface atomic steps. Similarly, many monoatomic steps a slight disorientation with respect to the electron beam
can be identified from the HRTEM zirconia image along axis means that the fringes run preferably in one direction.
the h010i orientation in Fig. 4. Along this orientation, the The crystallite of Fig. 5 shows a faceted shape exhibiting
{101} faces do not show long plane terminations. This can two kind of faces; namely regular {101} and {001} faces and
probably be related to the crystallite size. The energetically slightly rough {100} face. This surface roughness can also
stable configuration of the zirconia crystallite and the sub- be interpreted in terms of surface atomic steps. Crystal-
stantial rearrangement of the crystallite profile are likely to lographically speaking, the combination of these surface
be sensitive either to the crystallite size or to the adsorp- atomic steps gives rise to surfaces of high Miller index,
tion of foreign species, or both. Since the images show very which from an energetic point of view produce very favor-
clean zirconia surfaces, the crystallite surface configuration able adsorption sites (19). It would be appealing to propose
in Fig. 4 is then due to the size effect. This result leads us to that the presence of sulfate groups near or on these high-
conclude that there is a critical crystallite size (� 5 nm) in Miller-index surfaces could give rise to the highly acidic
the sulfate-free ZrO2 under which the appearance of many sites of sulfated zirconia. The presence of sulfate groups
one-atom-height steps is favored. Above this crystallite size on those particles is very likely, as the proven effect of the
the number of surface steps is considerably reduced.
presence of sulfate is the stabilization of tetragonal zirconia

¨
BENAISSA ET AL.
6
98
FIG. 3. HRTEM micrograph for sulfate-free zirconia calcined at 773 K, ZrO2(773), along the h111i crystallographic orientation. Note the presence
of short surface atomic steps.
particles at the high calcination temperature used in this mentioned above, the sulfated zirconia particles have a
study (998 K). However, the presence of sulfate groups markedly faceted shape; with the {110} plane terminations
on these high-Miller-index short surfaces was not appar- being clearly predominant. The {110} faces do not exhibit
ent from our study due to the limitations of the HRTEM apparent monoatomic steps, as was the case on the {110}
technique. A rough surface is a 3D object, while the edge of faces present on the sulfate-free zirconia particles (see
the particle provided by a micrograph is only 1D, so details Fig. 3). It is proposed that the faceted shape of the relatively
on the rough surfaces are lost during transmission electron small zirconia crystallites in the sulfate-containing material
microscopy studies. It is also important to mention that the results from both the presence of sulfate and the thermal
number of ZrO2 crystallites that have “rough surfaces” or effect. The latter is indeed enhancing the formation of the
high Miller indices is very small compared to the particles most stable faces, while the presence of sulfate is favoring
that have flat or regular surfaces that are discussed below. the formation of small particles of the tetragonal zirconia
These results imply that the number of highly acidic sites phase.
on the sulfated zirconia would be low if they were present
on these small highly defective particles.
While the HRTEM image along the h010i orientation,
Fig. 5, does not show any extra surface features, the
HRTEM images of sulfated zirconia along the h111i ori- HRTEM images in Figs. 6a and 6b show surface details
entation are shown in Figs. 6a and 6b. In these images, in the form of quasi-continuous line of repeated dark
we can recognize the periodicity already observed in the spots along the {110} top atomic layer. Careful examination
HRTEM images of sulfate-free zirconia (Fig. 3). As it was shows that the top layer contrast is quite different from the

HRTEM STUDY OF SULFATE GROUPS INTERACTIONS
699
FIG. 4. HRTEM micrograph for sulfate-free zirconia calcined at 773 K, ZrO2(773), along the h010i crystallographic orientation. Note the abundance
of defective terminations.
immediate underneath layers. These surface features were The presence of sulfate groups on this sample is evidenced
visible over a limited range of defoci (around Scherzer) by the stabilization of the metastable tetragonal zirconia.
recorded experimentally (the focus difference between im- As discussed in Section 3.1, sulfate-free zirconia calcined
ages in Figs. 6a and 6b, respectively, is on the order of 10 nm). at 998 K, which is the same calcination temperature used
This slight variation in defocus makes the detection of the for the preparation of the sulfated zirconia sample, contains
variable atomic occupancy on the surface unmistakable. solely monoclinic zirconia phase. In addition, it was clear
The variation of the defocus clearly shows surface vacan- from our observations that the presence of sulfate groups
cies located on the top layer of the particle. Moreover, the induces (110) faceting of the tetragonal zirconia crystallites.
image of Fig. 6b exhibits white fringes having a periodicity The question regarding the image of Fig. 6 is whether a
of 0.515 nm, equal to the double of the ZrO2 (110) plane dis- sulfate complex layer could be evidenced by the phase con-
tance. If we follow the white-fringe periodicity perpendic- trast, and the answer is yes. Image simulations have recently
ularly from the bulk toward the edge, the white line termi- demonstrated (20, 21) that an adsorbed sulfur overlayer on
nates exactly at the top layer. This contrast periodicity was a Pt particle produces sufficient contrast to be observed
not visible in the sulfate-free zirconia samples and would be by HRTEM.
difficult to explain based on surface roughness or anion va-
Let us go back to the image of Fig. 6, and try to ex-
cancies. The observed surface features in Figs. 6a and 6b can tract information about the possible sulfate groups loca-
be explained by the formation of a sulfate complex layer. tion upon the zirconia surface. The spacing between two

¨
BENAISSA ET AL.
7
00
FIG. 5. HRTEM micrograph for sulfated zirconia calcined at 998 K along the h010i crystallographic orientation. A faceted particle showing two
kind of surface terminations namely regular {101} and {001} faces and a slightly rough {100} face.
dark spots along the (110) plane is equal to 0.368 nm. Com- proposed to exist on the ZrO2 surface. For example, Riemer
pared to the extra-top-layer periodicity, we notice that the et al. (11) carried out Raman and MAS NMR studies, and
dark spots along the adsorbed layer have exactly the same proposed a sulfate structure, suggested earlier by Bensitel
periodicity, but the spots are shifted by 0.184 nm, which et al. (5), that consists of a single S == O double bond with
is half the spacing along the (110) plane. This result indi- the sulfur atom anchored to the oxide surface through three
cates that the sulfate groups are located between Zr atoms. S–O–Zr linkages, as shown in Fig. 7a. In contrast, several
The side view image of the (110) plane cannot distinguish researchers (10) support Tanabe’s earlier bidenatate struc-
whether the sulfate group is bridging across two or three ture (2), which consisted on two S == O double bonds with
zirconium atoms, as depicted by the top view of the (110) the sulfur atom anchored to the oxide surface through two
plane in Figs. 7a and 7b. Both sulfate structures have been S–O–Zr linkages, as shown in Fig. 7b.

FIG. 6. HRTEM micrograph for sulfated zirconia calcined at 998 K along the h111i crystallographic orientation. (a) and (b) are images of the same
particle but with a slight different defocus to show surface features. Note in (b) the appearance of some vacancies (arrows) at the top-layer and white
fringes having a periodicity equal to the double of that of (110) planes.

¨
BENAISSA ET AL.
7
02
imaging contrast (21), even if low-energy electron diffrac-
tion (LEED) data were available in the literature. In this
context, it should be mentioned that the adsorption sites of
sulfate on any surface of zirconia are not precisely known.
While extensive image simulations have been obtained, the
agreement with our experimental images is still uncertain.
Although the periodicity of surface features observed in the
HRTEM images will not change by inverting the structures,
agreement with experiments is not easily obtained mainly
because of two factors; (i) the difference in the scattering
factors of Zr, O, and S, and (ii) the lack of the exact distribu-
tion of the interatomic distances and coordination numbers
of the different atoms involved along the surface and within
the three to four zirconia top layers.
4
. CONCLUSIONS
Our HRTEM study provided information at an atomic
level on the sulfated zirconia surface configuration. We
demonstrated the potential of HRTEM to directly observe
sulfate surface complexes on sulfated zirconia catalysts.
This capability combined with other characterization tech-
niques and catalytic measurements should prove useful in
elucidating the structure and/or nature of the highly acidic
sites present on the sulfated zirconia material.
Specifically, the following conclusions can be drawn from
our study:
1) The presence of SO2 groups not only stabilizes small
�
(
4
tetragonal ZrO2 crystallites, but also induces the formation
of well-faceted particles, i.e., decreases the number of de-
fective terminations.
(
2) The shape of the sulfated zirconia crystallites exhibits
two types of surfaces: long-smooth and a relatively few
short-rough planes. With the long-smooth (110) plane being
clearly predominant.
FIG. 7. Top view of proposed structures of adsorbed sulfate groups
on the (110) plane of zirconia. (a) Threefold coordination and (b) twofold
coordination.
(
3) Comparison between sulfate-free and sulfated zir-
2�
conia materials indicates that the presence of SO4 in-
duces the (110) faceting of the zirconia particles. A layer
of isolated sulfate groups has been directly observed on the
To construct a model of adsorbate sulfates which ac- long-flat (110) surface of the zirconia. The geometry of this
counts for the observed HRTEM image contrast, we must plane is such that it can accommodate the sulfate groups in a
at least take into account a compatible arrangement with two- or threefold coordination, i.e., bidentate or tridentate
existing surface studies on model crystals. Unfortunately, structure. The HRTEM image contrast cannot distinguish
the lack of studies regarding zirconia single crystals loaded between these two structures. It is not clear whether the ob-
with sulfates makes the prediction of a reasonable model served sulfate groups on the presumably defect-free (110)
not straightforward. An example is given by early stud- plane play any role in the formation of the highly acidic
ies on the morphological changes of platinum particles oc- site or they are just “spectators.” This could mask informa-
curring upon sulfiding (15). While extensive modeling and tion on the actual acid sites, which could be present on the
image calculations have led to a refined model of a sul- particles described below. Thus, it is suggested that caution
fur monolayer on a cubic platinum particle which match must be taken when performing spectroscopic studies us-
very well with HRTEM surface images (20), other investi- ing techniques such as IR, NMR, XPS, and Raman, since
gations using the same sample and the same model showed the information derived could originate from “spectator”
many difficulties in reconciling using the same sample and sulfate groups.
the same model showed many difficulties in reconciling the
(4) Our study revealed the presence of a small number of
transmitted contrast of simulated images with the surface zirconia crystallites containing high-Miller-index surfaces.

HRTEM STUDY OF SULFATE GROUPS INTERACTIONS
703
It was proposed that the presence of sulfate groups near or
on these particles could give rise to the highly acidic sites of
sulfated zirconia. The number of such catalytic sites would
therefore be small.
9. Nascimiento, P., Akratopoulou, C., Oszagyan, H., Coudurier, G.,
Travers, C., Joly, J. F., and Vedrine, J. C., in “Proceedings, 10th Inter-
national Congress on Catalysis, Budapest, 1992.” (L. Guczi, F. Soly-
mosi, and P. Tetenyi, Eds.), p. 1185. Akad e´ miai Kiad o´ , Budapest,
1993.
1
1
1
0. Lunsford, J. H., Song, H., Campbell, S. M., Liang, C. H., and Anthony,
R. G., Catal. Lett. 27, 305 (1994).
ACKNOWLEDGMENTS
1. Riemer, T., Spielbauer, D., Hunger, M., Mekhemer, H. A. H., and
Knozinger, H., J. Chem. Soc. Chem. Commun. 1181 (1994).
2. Adeeva, V., de Haan, J. W., Lei, G. D., Schunemann, V., van de Ven,
L. J. M., Sachtler, W. M. H., and van Santen, R. A., J. Catal. 151, 364
(1995).
The authors are indebted to Mr. Luis Rendon for technical help and to
Dr. John Higgins for helpful discussions.
REFERENCES
13. Marks, L. D., and Smith, D. J., Nature 303, 316 (1983).
1
2
. Hino, M., and Arata, K., J. Chem. Soc. Chem. Commun. 1148 (1979).
. Tanabe, K., Misono, M., Yoshio, O., and Hatton, H., in “New
14. Smith, D. J., Bursill, L. A., and Jefferson, D. A., Surf. Sci. 175, 673
(1986).
Solid Acids and Bases: Their Catalytic Properties,” p. 199. Elsevier, 15. Jefferson, D. A., and Harris, P. J. F., Nature 332, 617 (1988).
Amsterdam/Kodansha, Tokyo, 1989.
16. Yacam a´ n, M. J., Herrera, R., G o´ mez, A., Tehuacanero, S., and Schabes,
P. , Surf. Sci. 237, 248 (1990).
17. Parera, J. M., Catal. Today 15, 481 (1992).
18. Mercera, P. D. L., Ph.D. thesis, University of Twente, 1991.
19. Somorjai, G. A., in “Chemistry in Two Dimensions: Surfaces,” p. 157.
Cornell Univ. Press, Cornell, NY, 1981.
3
4
5
. Yamaguchi, T., Jin, T., and Tanabe, K., J.Phys. Chem. 90, 3148 (1986).
. Hino, M., Arata, K., and Yabe, K., Shokubai 22, 232 (1980).
. Bensitel, M., Saur, O., and Lavalley, J. C., Mater. Chem. Phys. 19, 147
(
1988).
6
7
. Babou, F., Coudurier, G., and Vedrine, J., J. Catal. 152, 341 (1995).
. Kustov, L. M., Kazansky, V. B., Figueras, F., and Tichit, D., J. Catal. 20. Uppenbrink, J., Kirkland, A. I., Tang, D., and Jefferson, D. A., Surf.
50, 143 (1994). Sci. 238, 132 (1990).
. Morterra, C., Cerrato, G., Emanuel, C., and Bolis, V., J. Catal. 142, 349 21. Gribelyuk, M. A., Harris, P. J. F., and Hutchison, J. L., Phil. Mag. B 69,
1994). 655 (1994).
1
8
(