Full Papers
for example as active sites for methanol synthesis,[33] and has
recently been proposed to be a pivotal factor in ethanol con-
version to yield acetaldehyde and/or ethylene.[11] With respect
to the surface hydroxyl groups on polar surfaces, there is also
an open discussion of their role in catalysis. This is because, in
many catalytic reactions, one or more reaction steps involve
hydrogenation/dehydrogenation of the reaction species, likely
to be performed by these surface moieties.[34] Furthermore, it is
known that the presence of surface hydroxyl groups could
modify the acid–base surface properties of the material.
Indeed, it has been shown that in many solid bases, such as
rare earth oxides[35] or hydrotalcites,[36] surface hydroxylation
partially controls their reactivity.
Scheme 1. Crystallographic ZnO wurtzite structure. Preferential surface ori-
entations are displayed with colored planes and direction vectors for the
hexagonal (left image) and the orthorhombic (right image) unit cells.
Orange and blue spheres denote O and Zn atoms, respectively.[33]
dipole moments.[21] As ZnO exhibits a variety of morphologies,
which include single-crystal surfaces,[22] thin films,[23] nanostruc-
tures,[24] and well-faceted nanoparticles,[25] with distinct polar/
nonpolar facet ratios, many experimental studies have ad-
dressed this point by relating the catalytic behavior to structur-
al factors. However, the discussion remains open to a large
extent. On one hand, some research groups assign the highest
catalytic activities to polar surfaces[24–26] and point out that
such surfaces are the most unstable, that is, exhibit the highest
surface energy, and consequently are prone to react more
easily. On the other hand, others have shown that nonpolar
surfaces[22] are responsible for the catalytic performance.
In the present work, we explore the relationship between
the ZnO morphology and the surface reactivity during the de-
hydrogenation of ethanol for a series of seven nanometric
samples, six synthesized in the lab and one available commer-
cially. The structural and chemical differences encountered for
the different nanostructures are analyzed as a function of the
specific surface planes exposed to the medium. The main char-
acterization tools applied are surface area (SBET) determination,
XRD, and SEM for structural and morphological details. Diffuse
reflectance infrared Fourier transform spectroscopy (DRIFTS)
and catalytic tests of isopropanol dehydrogenation/dehydra-
tion are used to determine the main acid–base surface proper-
ties. The information obtained from these characterization
methods is used to discuss the reactivity of the exposed facets
of these polycrystalline ZnO samples in the ethanol decompo-
sition reaction.
Surface crystalline structures are well known to be a funda-
mental aspect in the activity and selectivity in heterogeneous
catalysis using metal oxides.[27] The surface structure sensitivity
phenomenon, which implies that active sites are different from
one crystalline face to another, was suggested by Boudart for
metallic nanoparticles.[28] Structure sensitivity on metal oxides
for oxidation reactions was demonstrated for the first time by
Volta et al.[29] for propene partial oxidation to acrolein, who
used a new method to prepare MoO3 crystals with specific ori-
entations. In the particular case of ZnO, apart from photocatal-
ysis, there is very little data that deals with the influence of the
morphology or surface properties of ZnO powders in their be-
havior towards chemical reactions in catalysis. Indeed, there is
a lack of fundamental understanding on how these ZnO pow-
ders work under real catalytic reactions, and the aspects on
which the catalytic activity and selectivity depend on surface
structures are still matters of debate. In some studies, the cata-
lytic activity of the polar surfaces was found to be higher than
that of the nonpolar surfaces, such as in methanol synthesis,[30]
in the decomposition of terminal alkynes, and in the decompo-
sition of acetic and propionic acids.[31] However, most of these
studies concern single crystals and/or high-vacuum conditions.
It is known that the surfaces of polycrystalline ZnO particles
comprise a large number of defects such as steps, edges, cor-
ners, kinks, and vacancies, which are not present on perfect
single-crystal surfaces.
Results and Discussion
Some characterization results that concern the textural, struc-
tural, and morphological properties of the samples are sum-
marized in Table 1. Although samples ZnO-E3, ZnO-E4, ZnO-ox,
and ZnO-E5 have a relatively high BET surface area (20–
40 m2gÀ1), the other three zinc oxides, ZnO-hc, ZnO-A, and
ZnO-h, exhibit lower specific surface areas, less than 7 m2 gÀ1
,
¯
Table 1. SBET, intensity ratio of XRD (1010)/(0002) peaks, morphological
observations, range of particle sizes measured from the SEM micrographs,
and average particle size calculated from the principal XRD peak of the
ZnO samples.
Sample SBET
[m2 gÀ1
(1010)/(0002) Morphological
issues[a]
Particle size[a] [nm]
SEM
¯
]
XRD[b]
ZnO-E3 38
ZnO-E4 35
ZnO-ox 23
ZnO-E5 23
0.97
1.06
1.31
1.44
1.27
1.27
1.29
brick[c]
hexagonal[c]
hexagonal disk
needle[c]
hexagonal disk 200–1000 >100
hexagonal disk
20–200
32
37
45
43
10–50
30–80
20–120
As mentioned above, polar surfaces on ZnO are intrinsically
unstable, and different stabilization processes can occur. Exam-
ples of stabilizing processes are surface reconstruction,[32] satu-
ration of the surface with hydroxyl groups through the adsorp-
tion of hydrogen or water,[21] and the formation of oxygen va-
ZnO-h
ZnO-hc
ZnO-A
0.8
6.7
2.9
80–200
86
>100
hexagonal prism 100–400
[a] Observed by SEM. [b] Particle sizes calculated by the application of the
Debye–Scherrer equation to the principal ZnO XRD peaks. [c] Studies by
TEM also support these morphological observations, as reported in
Refs. [20,26]
¯
cancies on the polar O-terminated (0001) surface. The latter
was proposed to play an important role in catalysis on ZnO,
ChemSusChem 2015, 8, 2223 – 2230
2224
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim