Shape effect of ZnO crystals as cocatalyst in combined reforming- hydrogenolysis of glycerol

Article abstract of DOI:10.1021/cs400526s

Disk- and rod-shaped hexagonal ZnO crystals with various length-to-diameter aspect ratios were controllably synthesized via a facile solution route by adjusting the precursor concentration. The shape and the dimension of the synthesized ZnO crystals were observed by scanning electron microscopy (SEM). The wurtzite structure and the growth habit were determined by powder X-ray diffraction (XRD) and transmission electron microscopy (TEM) coupled with selected-area electron diffraction (SAED). It is found that with the lowering of the precursor concentrations, the ZnO crystals were elongated along the c-axis, and the diameter of the {001} planes was reduced, leading to shape evolution from hexagonal disk to prismatic rod. As a result, the ZnO crystals were different from each other in the proportion of the {100} nonpolar planes and the {001} polar planes. The well-defined ZnO crystals were used as the cocatalyst with skeletal Ni₄₀Mo₁₀ in the combined reforming-hydrogenolysis (CRH) of glycerol in the absence of adventitious H ₂. A remarkable shape-dependent effect on the selectivity to the C3 hydrogenolysis products and the production rate of 1,2-propandiol (1,2-PDO) was identified. ZnO with a larger proportion of the nonpolar planes was more effective in the CRH of glycerol to the C3 products. An excellent linear relationship between the surface area of the {100} nonpolar planes and the production rate of 1,2-PDO was identified. This is attributed to the in situ enhancement of the Lewis acidity of the nonpolar planes of ZnO by chemisorbed CO₂ from the reforming of glycerol, which greatly accelerates the dehydration of glycerol to acetol, the intermediate to 1,2-PDO.

Full text of DOI:10.1021/cs400526s

Research Article
pubs.acs.org/acscatalysis
Shape Effect of ZnO Crystals as Cocatalyst in Combined Reforming−
Hydrogenolysis of Glycerol
Jiye Hu, Yiqiu Fan, Yan Pei, Minghua Qiao,* Kangnian Fan, Xiaoxin Zhang, and Baoning Zong*
†
†
†
,†
†
‡
,‡
†Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Fudan University, Shanghai
00433, People’s Republic of China
2
‡
State Key Laboratory of Catalytic Materials and Reaction Engineering, Research Institute of Petroleum Processing, Beijing 100083,
People’s Republic of China
ABSTRACT: Disk- and rod-shaped hexagonal ZnO crystals with various length-to-
diameter aspect ratios were controllably synthesized via a facile solution route by adjusting
the precursor concentration. The shape and the dimension of the synthesized ZnO crystals
were observed by scanning electron microscopy (SEM). The wurtzite structure and the
growth habit were determined by powder X-ray diffraction (XRD) and transmission
electron microscopy (TEM) coupled with selected-area electron diffraction (SAED). It is
found that with the lowering of the precursor concentrations, the ZnO crystals were
elongated along the c-axis, and the diameter of the {001} planes was reduced, leading to
shape evolution from hexagonal disk to prismatic rod. As a result, the ZnO crystals were
different from each other in the proportion of the {100} nonpolar planes and the {001}
polar planes. The well-defined ZnO crystals were used as the cocatalyst with skeletal Ni Mo in the combined reforming−
40
10
hydrogenolysis (CRH) of glycerol in the absence of adventitious H . A remarkable shape-dependent effect on the selectivity to
2
the C3 hydrogenolysis products and the production rate of 1,2-propandiol (1,2-PDO) was identified. ZnO with a larger
proportion of the nonpolar planes was more effective in the CRH of glycerol to the C3 products. An excellent linear relationship
between the surface area of the {100} nonpolar planes and the production rate of 1,2-PDO was identified. This is attributed to
the in situ enhancement of the Lewis acidity of the nonpolar planes of ZnO by chemisorbed CO from the reforming of glycerol,
2
which greatly accelerates the dehydration of glycerol to acetol, the intermediate to 1,2-PDO.
KEYWORDS: shape, zinc oxide, glycerol, hydrogenolysis, promotion effect
1
. INTRODUCTION
Al O and Pt/Al O showed a synergetic effect between them,
2 3 2 3
leading to glycerol conversion of 50.1% with 1,2-PDO
Glycerol, as the main byproduct of biodiesel derived from
biomass, is currently widely available. Because it is highly
12
1
,2
selectivity of 47.2% at 493 K under 1.4 MPa N for 6 h.
2
They found that, as compared with N atmosphere, H
2
2
functionalized, glycerol is considered a versatile platform3
atmosphere adversely affected 1,2-PDO selectivity on their
catalyst because of excessive methanation. Gandarias et al.
reported that the amorphous silicoalumina-supported Pt
catalyst (Pt/ASA) gave glycerol conversion of 26.6% with
chemical for the production of a set of valuable chemicals.
Among various utilization routes of glycerol, its reaction with
H to produce 1,2-PDO is of special interest
PDO is a desirable chemical that has wide applications in the
cosmetic, pharmaceutical, and polymer industries.
4
−8
because 1,2-
2
1
,2,9
1,2-PDO selectivity of 35.3% at 493 K under 4.5 MPa N for 24
2
Tradi-
13
h. Wawrzetz et al. investigated the glycerol transformation
over the Pt/Al O3 catalyst under N atmosphere in a
continuous fixed-bed reactor. By kinetically controlled
experiments, they disclosed that 1,2-PDO was formed through
glycerol dehydration to acetol on the support, followed by
acetol hydrogenation on the metal.
tionally, 1,2-PDO is fabricated in the petrochemical route1
2
2
through the hydration of hazardous propylene oxide.
1
4
Production of 1,2-PDO from biomass-derived glycerol is
obviously more sustainable and environmentally more friendly.
In the hydrogenolysis of glycerol to 1,2-PDO, H can be fed
2
into the reactor externally or generated on-site by converting a
portion of glycerol over catalysts with reforming ability. The
concept of on-site H generation can produce 1,2-PDO from
10
Recently, we found that when physically mixed with the
rapidly quenched skeletal NiMo catalyst (RQ Ni Mo ), ZnO
40
10
2
was far superior to other oxides in directing the product to 1,2-
glycerol with high atomic economy and operational safety, thus
attracting much attention. Many works have been focused on
the conversion of glycerol to 1,2-PDO without adventitious H2.
D’Hondt et al. achieved the production of 1,2-PDO from
glycerol under an inert atmosphere over the noble Pt/NaY
catalyst. This leading work triggered successive research in
pursuing other effective catalyst systems. Roy et al. reported
that an admixture catalyst comprising equal weights of Ru/
15
PDO in the CRH of glycerol. In light of the study by Wang et
16,17
al. on CO adsorption on (100) nonpolar plane of ZnO,
we
2
tentatively attributed the promotion effect to CO -enhanced
2
1
1
Received: July 8, 2013
Revised: August 28, 2013
Published: August 29, 2013
©
2013 American Chemical Society
2280
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
Lewis acidity of ZnO, which accelerates the dehydration of
glycerol to acetol, the reaction intermediate to 1,2-PDO.
However, the commercial ZnO has ill-defined morphology,
which makes it difficult to quantify the effect of the nonpolar
plane of ZnO on the promotion effect. Such knowledge may
offer new opportunity for maximizing the beneficial effect of
ZnO in the CRH of glycerol through rational shape control.
Regular crystal with well-defined morphology and limited
surface sites is conductive to a better understanding of the
The Raman spectra were recorded on a Jobin-Yvon LabRam-
−1
1B spectrometer with spectral resolution of ∼1.2 cm . The
laser line at 632 nm of a He−Ne laser with an output of 4.3
mW was used as the excitation source.
Room temperature PL spectra were acquired on an
Edinburgh Photonics FLS 920 photoluminescence instrument
with an excitation wavelength of 325 nm. Prior to the
measurements, the ZnO samples were dispersed in deionized
water by ultrasonication to prepare homogeneous suspension
18
−1
microscopic origin of its function in heterogeneous catalysis.
with the same concentration of 1.0 mg mL .
It is documented that ZnO crystal tends to take on a prismatic
shape with two hexagonal {001} bottom planes and six {100}
The XPS spectra were recorded on a Perkin-Elmer
PHI5000C spectrometer using Mg Kα radiation (hν = 1253.6
eV) as the excitation source. The samples were pressed into a
self-supported disc and loaded onto the sample plate, then the
samples were degassed in the pretreatment chamber in vacuo at
room temperature for 4 h before being transferred to the
analyzing chamber with a background pressure >2 × 10− Torr.
All binding energy (BE) values were referenced to the C 1s
peak of contaminant carbon at 284.6 eV with an uncertainty of
1
9
side planes. The two {001} planes are terminated with either
2
+
2−
Zn or O ions and are, thus, polar, whereas the six {100}
20
planes are terminated with both ions and are, thus, nonpolar.
The solution route is versatile in synthesizing ZnO crystals with
adjustable growth habits.
21,22
9
With these shape-controlled ZnO
crystals, one may find a direct correlation between the surface
area of the nonpolar planes and the promotion effect of ZnO in
the CRH of glycerol, provided that our previous argument
based on commercial ZnO is justified.
In the present work, we synthesized hexagonal ZnO crystals
with different exposure degrees of the nonpolar planes by the
±
0.2 eV.
The multipoint Brunauer−Emmett−Teller (BET) surface
areas of the ZnO samples were measured by N physisorption
2
at 77 K on a Micromeritics TriStar3000 apparatus. Prior to the
solution method and physically mixed them with RQ Ni Mo
40
10
measurements, the samples were purged by N at 423 K for 2 h.
2
for the CRH of glycerol. The physical properties of the ZnO
crystals were characterized by XRD, SEM, and TEM coupled
with SAED, Raman spectroscopy, photoluminescence (PL)
2
.3. Activity Testing and Product Analysis. The CRH of
glycerol was conducted on a continuous-flow fixed-bed tubular
reactor (6 mm i.d.). The reaction temperature was 508 K, and
the system pressure was 3.1 MPa regulated by Ar. An aqueous
solution containing 10 wt % of glycerol was pumped from the
bottom of the reactor at varying flow rates to obtain the desired
weight hourly space velocity (WHSV, weight of glycerol per
weight of RQ Ni Mo per hour). The weight ratio between
spectroscopy, and N physisorption. The possible electronic
2
interaction between ZnO and RQ Ni Mo was discriminated
4
0
10
by X-ray photoelectron spectroscopy (XPS). Correlations of
the structural peculiarities of the ZnO crystals with their
promotion effects in the CRH of glycerol were discussed on the
basis of detailed catalytic results.
40
10
ZnO and RQ Ni Mo in the physically mixed ZnO−RQ
40
10
Ni Mo catalysts was kept at 1:1. The reaction course was
40
10
2
. EXPERIMENTAL SECTION
.1. Preparation of ZnO. The preparation procedure for
hexagonal ZnO crystals with disk to rod shapes was a
monitored by sampling the gas and liquid products at intervals,
followed by gas chromatographic analysis. In the gas products,
H and CO were separated by a 5 Å molecular sieve-packed
2
2
2
21
modification of the method developed by Li et al. with
some adjustments in the precursor concentration and the
crystallization temperature and time. First, 3.00 g of Zn-
column and examined by a thermal conductivity detector
(TCD) on a GC102 M gas chromatograph. Light alkanes were
separated by a Porapak R packed column and examined by a
flame ionization detector (FID) on the same gas chromato-
graph. The liquid products were condensed and separated by a
CP-Wax 52 CB capillary column (30 m × 0.25 mm × 0.25 μm)
and examined by an FID on a GC9560 gas chromatograph. The
liquid products were also qualified by GC/MS (Finnigan
Voyager) with an HP-5 capillary column. The carbon balance
was within ±5% for all catalytic runs.
(
CH COO) ·2H O (ZnAc ) and 1.92 g of hexamethylenetetr-
3 2 2 2
amine (HMT) were dissolved in 24, 120, 240, and 360 mL of
deionized water. After ultrasonication for 10 min, transparent
solutions with decreasing precursor concentrations were
obtained. These vessels were covered by preservative films
and kept at 363 K for 12 h. After being cooled to room
temperature, the white precipitates were collected by
centrifugation and washed with ethanol, followed by deionized
water to neutrality. The products were dried at 333 K overnight
and calcined at 473 K for 2 h in air for the purpose of
characterization and catalytic testing.
3
. RESULTS
3.1. Crystal Phase and Morphology. The crystal phase
2
.2. Characterization. The XRD patterns were acquired on
and phase purity of the ZnO samples were examined by XRD.
As presented in Figure 1, all the diffraction peaks can be
indexed to ZnO with hexagonal wurtzite structure (space group
a Bruker AXS D8 Advance X-ray diffractometer using Ni-
filtered Cu Kα radiation (λ = 0.15418 nm). The tube voltage
was 40 kV, and the current was 40 mA.
23
P6 mc, JCPDS 36-1451). The intense and sharp diffraction
3
The SEM images were recorded on a Philips XL30 electron
microscope. The accelerating voltage was 20−30 kV. The
samples were attached to a double-sided carbon conductive
tape on an aluminum slab and coated with gold. TEM
observations were conducted on a JEOL JEM2011 microscope
operated at 200 kV. The ZnO samples were dispersed in
ethanol, dropped onto a carbon-film-coated copper grid, and
dried in air.
peaks indicate that these ZnO crystals are in general of high
crystallinity. No impurity peaks were found. It is noted that the
relative intensities between the (100) and (002) diffraction
peaks are quite different among these samples. As implemented
in the brackets in Figure 1, the I(100)/I(002) intensity ratio
increases in the order of 0.80 < 1.05 < 1.77 < 2.61 with the
decrease in the precursor concentration, indicating the
21
preferential growth of the ZnO crystals along the c-axis.
2
281
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
basis of the shape, we labeled these ZnO crystals as ZnO-disk
(
Figure 2a), ZnO-rod1 (Figure 2b), ZnO-rod2 (Figure 2c), and
ZnO-rod3 (Figure 2d). As summarized in Table 1, from ZnO-
Table 1. Physical Characteristics of the As-Synthesized ZnO
Samples
b
c
length
(μm)
diameter
(μm)
S
S
BET,n
2
BET
a
(m g−1)
2
(m g−1)
sample
R
ZnO-disk
1.4 ± 0.2
4.7 ± 1.5
5.6 ± 1.4
4.0 ± 1.4
0.3
1.2
4.7
2.9
1.6
2.0
ZnO-
rod1
ZnO-
rod2
4.9 ± 1.7
5.7 ± 1.8
1.5 ± 0.6
1.3 ± 0.4
3.3
3.4
3.9
2.9
3.5
ZnO-
rod3
4.4
a
b
Length-to-diameter aspect ratio. Total specific surface area
c
determined by N physisorption. Specific surface area of the nonpolar
2
surfaces calculated from the total specific surface area and the
Figure 1. XRD patterns of the ZnO-disk, ZnO-rod1, ZnO-rod2, ZnO-
dimension of the ZnO measured by SEM.
rod3 samples.
disk to ZnO-rod3, the average length increased monotonically
from 1.4 to 5.7 μm, and the average diameter of the hexagonal
plane decreased monotonically from 5.6 to 1.3 μm, resulting in
a drastic increase in the length-to-diameter aspect ratio from 0.3
to 4.4.
The microstructures of the ZnO crystals were elucidated by
TEM using ZnO-disk and ZnO-rod3 as examples for the disk-
and rod-shaped ZnO samples, respectively. Figure 3a is the plan
view of ZnO-disk, which clearly evidenced the regular
hexagonal shape fully consistent with the SEM observation.
The SAED pattern of ZnO-disk as the inset in Figure 3a can be
SEM unambiguously confirmed the shape change of ZnO
from disk to rod with the lowering of the precursor
concentration. According to Figure 2a, relatively uniform and
perfect hexagonal ZnO disks were harvested from the highest
precursor concentration. By reducing the precursor concen-
tration, the diameter of the hexagonal surface was reduced and
the length was increased, giving rise to hexagonal ZnO
prismatic rods with six smooth planes throughout their length
(
Figure 2b, c, d). This concentration-dependent shape change
is in accordance with the observation by Vayssieres. On the
24
Figure 2. SEM images of (a) ZnO-disk, (b) ZnO-rod1, (c) ZnO-rod2, and (d) ZnO-rod3.
2
282
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
length of the hexagonal ZnO crystals with the lowering of the
precursor concentration.
3
.2. Optical Properties. Figure 4 shows the Raman spectra
of the ZnO crystals. There is a prominent feature peaked at
Figure 3. (a) TEM image and (b) HRTEM image of ZnO-disk. Inset
in part a shows the corresponding SAED pattern. (c) TEM image and
Figure 4. Raman spectra of the ZnO-disk, ZnO-rod1, ZnO-rod2, and
ZnO-rod3 samples.
(d) HRTEM image of ZnO-rod3. Inset in part c shows the
corresponding SAED pattern.
437 cm− in all spectra, which is characteristic of the Raman
1
∼
35
active optical-phonon E mode of hexagonal wurtzite ZnO.
Some weak features at around 333 cm were also observed,
2
−1
25,26
indexed to the diffraction along the [001] zone axis,
which
28,35
which are assigned to the multiphonon E −E modes.
2H
2L
reflects the single-crystalline nature of the ZnO disk with
hexagonal {001} polar surfaces. The high-resolution TEM
−1
The virtual absence of the scattering peaks at 550−600 cm
due to a superimposition of the A and E (LO) modes
1
1
(
HRTEM) image of this sample is presented in Figure 3b. The
28,36
evidences the good phase purity of the ZnO crystals.
It is
ordered lattice fringes with the interplanar spacing of 0.282 nm
is clearly seen at the crystal edge, which can be ascribed to the
{
also observed that the E mode became stronger with the
2
increase in the aspect ratio, implying that the crystallinity of
2
7,28
100} nonpolar planes of wurtzite ZnO.
Figure 3c presents
37
ZnO became better from ZnO-disk to ZnO-rod3.
the side view of ZnO-rod3. The corresponding SAED pattern is
Room-temperature PL spectra of the ZnO samples are
shown in Figure 5. In all cases, the luminescence band at
around 385 nm and the broad band centered at 609 nm are
observed. The former resulted from the near-band edge
29
indexed as the [100] zone spots, which also evidence the high
crystal quality of the ZnO rods. The HRTEM image of ZnO-
rod3 (Figure 3d) shows ordered lattice fringes with the
interplanar spacing of ∼0.263 nm, which coincides with the
38
emission of the wide band gap ZnO, and the latter in the
3
0,31
{
002} planes of wurtzite ZnO
and confirms the [001]
orientation of the ZnO rods.
It is acknowledged that many factors integrate to affect the
nucleation and growth of a crystal, thus leading to different
3
2
shapes. In the present synthesis process, the precursors and
preparation conditions (temperature and time) were identical.
The only difference was the precursor concentration used.
Therefore, the precursor concentration must be largely
responsible for the shape difference of the ZnO crystals. In
−
the preparation, ZnAc is the zinc source and also supplies Ac
2
ions, and HMT serves as the pH buffer and provides a slow and
−
32,33
22
continuous supply of OH ions.
According to Peng et al.,
−
at higher precursor concentration, the Ac ions are more
inclined to adsorb on the (001) polar plane of ZnO terminated
with Zn cations, which inhibits the growth along the c-axis and,
consequently, giving rise to ZnO disks, although at lower
−
precursor concentration, the OH ions that functions as the
3
4
−
precipitant tend to substitute Ac and adsorb on the (001)
2
1
−
polar plane. The adsorption of the OH ions will accelerate
the growth of the (001) plane, resulting in the formation of
ZnO rods elongated at the c-axis. The above argument justifies
the observed decrease in the diameter and increase in the
Figure 5. Room-temperature PL spectra of the ZnO-disk, ZnO-rod1,
ZnO-rod2, and ZnO-rod3 samples.
2
283
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
visible-light region is assigned as an overall reflection of surface
40,41
RQ Ni Mo gave 42% selectivity to the hydrogenolysis
products, with 32% selectivity to 1,2-PDO. In the presence of
40
10
2
1,39
detects,
such as oxygen vacancy and interstitial defects.
Judging from the intensities of the visible-light region bands, we
found that the peak intensity follows the order of ZnO-disk >
ZnO-rod1 > ZnO-rod2 > ZnO-rod3, which is consistent with
the observation by Li et al. when comparing the PL spectra of
ZnO, all four ZnO−RQ Ni Mo admixture catalysts exhibited
40
10
improved selectivities to the hydrogenolysis products inclusive
of 1,2-PDO. It is particularly noteworthy that from ZnO-disk to
ZnO-rod3, the selectivity to the hydrogenolysis products
improved from 51% to 72% with the concomitant improvement
of the selectivity to 1,2-PDO from 38% to 51%. Meanwhile, the
selectivities to the gas and cracking products dropped from 31%
to 17% and from 18% to 11%, respectively. Interestingly, the
order of the selectivity to the hydrogenolysis products as well as
1,2-PDO is consistent with the order of the aspect ratio of the
ZnO crystals, which infers an inherent connection between the
product distributions and the shapes of the ZnO crystals in the
CRH of glycerol.
21
their ZnO disk with the rod. It indicates that the amount of
the surface defects on these ZnO samples decreased in that
42
order. As claimed by Strunk et al., the {100} nonpolar planes,
which is more abundant on the ZnO sample with the higher
aspect ratio, have low surface energies and, thus, are more
stable than the high-energy {001} polar planes, so to minimize
the total surface energy, the dissolution of the high-energy polar
planes, although invisible by SEM or TEM, is faster than that of
the nonpolar planes, which introduces more surface defects on
21
the ZnO sample abundant in the {001} polar planes.
In an endeavor to relate the production rate of 1,2-PDO to
the shapes of the ZnO crystals, kinetically controlled experi-
ments were carried out by adjusting the WHSV to restrict the
glycerol conversion within the range of 10%−15% to exclude
3.3. Catalytic Results in the CRH of Glycerol. The
synthesis of the ZnO crystals exposing different proportions of
the polar and nonpolar planes allows us to differentiate their
respective influence in the CRH of glycerol. The detailed
physicochemical properties of RQ Ni Mo have been
46
the transport limitations. Figure 6 presents the production
4
0
10
15
described elsewhere. The ZnO−RQ Ni Mo catalysts
were homogeneous mixtures of ZnO with different crystal
4
0
10
shapes and RQ Ni Mo with a weight ratio of 1:1. Previous
4
0
10
work using commercial ZnO as the cocatalyst has shown that
15
this weight ratio could present a good promotion effect.
Table 2 compiles the catalytic results over bare RQ Ni Mo
10
40
and the ZnO−RQ Ni Mo admixture catalysts at high
4
0
10
Table 2. Catalytic Results of the ZnO−RQ Ni Mo
4
0
10
Admixture Catalysts in the CRH of Glycerol with ZnO
a
Crystals of Different Shapes
carbon-based product selectivity (%)
b
c
d
ZnO
conversion (%)
gas
hydrogenolysis
cracking
none
disk
99
98
99
97
98
41
31
27
21
17
42 (32)
51 (38)
56 (41)
67 (49)
72 (51)
17
18
16
13
11
Figure 6. The production rate of 1,2-PDO over the ZnO−RQ
Ni Mo admixture catalysts with hexagonal ZnO crystals of different
rod1
rod2
rod3
40
10
shapes. Reaction conditions: 0.5 g of RQ Ni Mo , ZnO/RQ
40
10
Ni Mo weight ratio = 1.0, T = 508 K, pAr = 3.1 MPa, and the
40
10
conversion of glycerol within the range of 10−15%.
a
Reaction conditions: T = 508 K, p = 3.1 MPa, ZnO/RQ Ni Mo
10
Ar
40
−1
weight ratio = 1.0, WHSV = 1.2 h (relative to RQ Ni Mo ), and t =
4
0
10
b
c
8
h. Gas products (including CO and light alkanes). Hydrogenolysis
rate of 1,2-PDO expressed as turnover frequency (TOF) with
respect to the metallic Ni sites. In compliance with our previous
2
products (including acetone, 1-propanol, 2-propanol, acetol, propionic
acid, and 1,2-PDO). In the parentheses is the selectivity to 1,2-PDO.
Cracking products in liquid phase (including methanol, ethanol,
15
work using commercial ZnO as the cocatalyst, the figure
clearly illustrates that the physically mixed ZnO markedly
improved the TOF of 1,2-PDO. The TOF of 1,2-PDO was 0.72
d
acetaldehyde, acetic acid, and ethylene glycol).
−
1
min on bare RQ Ni Mo and markedly increased to 1.1
40
10
−1
conversion levels of glycerol. Aqueous-phase transformation
of glycerol over heterogeneous catalysts involves complicated
reaction networks;
min on the ZnO-disk−RQ Ni40Mo10 catalyst. The TOF of
1,2-PDO was increased steadily with the elongation of the ZnO
4
3,44
−1
however, this process can be approx-
rod, achieving a value of 1.5 min on the ZnO-rod3−RQ
Ni40Mo10 catalyst.
imately categorized into two kinds of pathways in parallel:
namely, reforming and hydrogenolysis, involving C−C bond
45
4. DISCUSSION
cleavage and C−O bond cleavage, respectively. Thus, to ease
the interpretation of the catalytic results, we classified the
products into three groups in Table 2: (i) gas products
resulting from the aqueous-phase reforming (APR) of glycerol
and successive gas phase reactions, (ii) hydrogenolysis products
resulting from the dehydration/hydrogenation reactions, and
When employed as catalyst or cocatalyst, there are four factors
that are often considered to account for the role of metal oxide:
47
21,48
49
surface area, crystal defect,
electronic effect, and crystal
50
habit. A larger surface area usually means more surface sites
on which the reactants can adsorb and react and, thus, a higher
activity. For example, Deng et al. found that the ZnO hollow
sphere was much more active in photodegradation of dyes than
commercial ZnO, which was attributed to the larger BET
(
iii) cracking products in the liquid phase resulting from C−C
bond cleavage of glycerol or C intermediates. According to
3
Table 2, the product distributions over these catalysts were
14
47
qualitatively similar to that over the Pt/Al O catalyst. Bare
surface area of the former. As listed in Table 1, the total BET
2
3
2
284
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
surface areas of these ZnO crystals ranked as ZnO-disk > ZnO-
rod2 > ZnO-rod3 > ZnO-rod1; however, according to Table 2
and Figure 6, ZnO-disk with the highest total BET surface area
was the least effective in promoting the production of the
hydrogenolysis products, whereas ZnO-rod1 with the lowest
BET surface area performed in-between ZnO-disk and ZnO-
rod2. It is evident that the total BET surface area did not
dominate the promotion effect of ZnO in the CRH of glycerol.
Li et al. reported that the oxygen vacancy was more
important for the catalytic activity than the BET surface area
when investigating the morphology effect of ZnO in N-
2
1
formylation reactions. The {001} polar planes, which
intrinsically possess more defects, were more active than the
{
100} nonpolar planes. Our Raman and PL spectra reveal that
ZnO-disk is the most defective, but this sample displayed the
least distinct promotion effect. On the contrary, the promotion
effect of ZnO-rod3 with the highest crystal perfection was the
most remarkable. These results do not favor the possibility that
the promotion effect of ZnO is determined by surface defects.
A recent work showed that, when ZnO was employed as the
cocatalyst with Cu nanoparticles (NPs) for methanol synthesis,
the morphology of ZnO affected the selectivity to methanol by
4
9
∼
30%. They observed by XPS, the electron transfer from
platelike ZnO to Cu NPs, which was then related to the
synergetic effect between physically mixed ZnO and Cu NPs.
Following that work, we collected the Ni 2p and Zn 2p spectra
of RQ Ni Mo , ZnO-rod3, and ZnO-rod3−RQ Ni Mo , as
40
10
40
10
presented in Figure 7. For the ZnO-rod3−RQ Ni Mo
4
0
10
catalyst, the Ni 2p BE is 852.7 eV (Figure 7b), which is
3
/2
identical to that of bare RQ Ni Mo (Figure 7a) with its Ni in
4
0
10
46
the metallic state. The Zn 2p3/2 BE of the ZnO-rod3−RQ
Ni Mo catalyst is 1021.8 eV (Figure 7d), which is the same
40
10
as that of bare ZnO-rod3 (Figure 7c) with its Zn oxidation state
49
consistent with that of bulk ZnO, so the electronic interaction
between ZnO and RQ Ni Mo could be excluded in the
4
0
10
ZnO−RQ Ni Mo admixture catalysts.
4
0
10
The crystal habits have been demonstrated to exert a
Figure 7. (Top) Ni 2p spectra of (a) bare RQ Ni Mo and (b) the
40
10
significant impact on the reaction due to different atomic
ZnO-rod3−RQ Ni Mo admixture catalyst. (Bottom) Zn 2p spectra
40
10
5
1−56
arrangements.
Si and Flytzani-Stephanopoulos found that
of (c) bare ZnO-rod3 and (d) the ZnO-rod3−RQ Ni40Mo10 admixture
the CeO rod enclosed by the {110} and {100} planes was
catalyst.
2
more active for Au stabilization/activation than the CeO cube
2
5
1
and polyhedra in the water−gas shift reaction. Co O4
3
nanorods predominantly exposing the {110} planes not only
catalyzed CO oxidation at as low as 196 K, but also remained
stable in a moist stream, which was attributed to the
be chemisorbed and formed a stable tridentate carbonate
species on the (100) nonpolar plane of ZnO, with the C atom
of CO bonding to one surface O anion, and both O atoms of
2
3
+
52
preferential presence of Co on these planes. Among the
CO forming equivalent bonds with neighboring Zn cations.
2
Co O4 cube, truncated octahedron, and octahedron, the
Note that CO cannot adopt such an adsorption mode on the
3
2
2
+
2−
octahedron with the {111} planes had the highest reversible
capacity and the best rate capability as anode material in lithium
{001} polar planes of ZnO terminated solely with Zn or O
ions. CO was adsorbed on every second Zn cation of the
20
2
5
3
ion battery. The {010} planes of TiO had both a favorable
(100) nonpolar plane, resulting in alternating occupied and
unoccupied surface Zn sites. The strong charge-accepting
2
surface atomic structure and a surface electronic structure,
which gave rise to a faster H evolution rate in photocatalytic
property of the adsorbed CO made the unoccupied Zn cation
2
2
5
4
splitting of water. In photodegradation of Rhodamine B on α-
Fe O with different morphologies, Zhou et al. found that α-
more electron-deficient, which greatly enhanced the Lewis
16
acidity of the nonpolar plane. On the other hand, according
2
3
8,57,58
Fe O nanorod exposing more low-coordinated surface Fe
to the dehydration mechanism of glycerol hydrogenolysis,
2
3
55
cations was much more active than other α-Fe O . Jang et al.
glycerol initially undergoes dehydration to acetol on the acid
sites, which is subsequently hydrogenated to 1,2-PDO on the
2
3
reported that ZnO nanoplates with a high population of the Zn
0001) polar plane had the highest photocatalytic activity for
H O generation, which was understood as a result of the facile
(
metallic sites. In the CRH of glycerol, CO is one of the main
2
gas products; such a CO -enhanced Lewis acidity of ZnO is
2
2
2
56
adsorption of O on that plane.
In our previous work, we tentatively attributed the
thus expected. The addition of Lewis acidic components to
accelerate the dehydration of glycerol have been widely
2
15
7
,59−61
promotion effect of commercial ZnO to the CO -enhanced
practiced.
Therefore, if our argument on commercial
2
1
6,17
Lewis acidity of ZnO. According to Wang et al.,
CO could
ZnO is justified, one may anticipate a positive correlation
2
2
285
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
between the surface area of the nonpolar planes and the
promotion effect of ZnO.
the total surface area, crystal defects, and electronic interaction
with RQ Ni Mo were excluded by linking these properties
with the catalytic results. The excellent linear relationship
between the surface area of the {100} nonpolar planes of ZnO
and the TOF of 1,2-PDO directly evidenced the assumption
40
10
To validate this deduction, the relationship between the
surface areas of the ZnO {100} nonpolar planes and the TOFs
of 1,2-PDO was quantified. On the basis of the well-defined
shape and dimension of these ZnO crystals, the BET surface
areas of the {100} planes were calculated to be 1.6, 2.0, 2.9, and
that the CO -enhanced Lewis acidity of the nonpolar planes of
2
ZnO is responsible for the remarkable shape-dependent
promotion effect of ZnO. The shape-function relationship
demonstrated by this work may provide guidance for designing
more effective catalyst system for the CRH of glycerol.
2
−1
3
.5 m g , respectively (Table 1). It is encouraging that an
excellent linear relationship emerges between the BET surface
area of the ZnO {100} planes and the TOF of 1,2-PDO, as
plotted in Figure 8. This linearity convincingly evidences our
AUTHOR INFORMATION
Corresponding Author
E-mail: mhqiao@fudan.edu.cn (M.H.Q.); zongbn.ripp@
sinopec.com (B.N.Z.).
■
*
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the National Basic Research
Program of China (2012CB224804), the NSF of China
■
(
21073043), the Science and Technology Commission of
Shanghai Municipality (10JC1401800, 08DZ2270500), and the
Program of New Century Excellent Talents (NCET-08-0126).
REFERENCES
■
(
1) Pagliaro, M.; Ciriminna, R.; Kimura, H.; Rossi, M.; Pina, C. D.
Angew. Chem., Int. Ed. 2007, 46, 4434−4440.
(
2
(
2) Bienholz, A.; Schwab, F.; Claus, P. Green Chem. 2010, 12, 290−
95.
3) Brandner, A.; Lehnert, K.; Bienholz, A.; Lucas, M.; Claus, P. Top.
Figure 8. A plot of the TOF of 1,2-PDO over the ZnO−RQ Ni Mo
4
0
10
admixture catalysts against the SBET of the {100} nonpolar planes of
Catal. 2009, 52, 278−287.
ZnO. Reaction conditions: 0.5 g of RQ Ni Mo , ZnO/RQ Ni Mo
weight ratio = 1.0, T = 508 K, pAr = 3.1 MPa, and the conversion of
glycerol within the range of 10−15%.
40
10
40
10
(
(
4) Lahr, D. G.; Shanks, B. H. J. Catal. 2005, 232, 386−394.
5) Maris, E. P.; Ketchie, W. C.; Murayama, M.; Davis, R. J. J. Catal.
2
(
007, 251, 281−294.
6) Zhou, J. X.; Guo, L. Y.; Guo, X. W.; Mao, J. B.; Zhang, S. G.
Green Chem. 2010, 12, 1835−1843.
7) Balaraju, M.; Rekha, V.; Prasad, P. S. S.; Devi, B. L. A. P.; Prasad,
R. B. N.; Lingaiah, N. Appl. Catal., A 2009, 354, 82−87.
8) Miyazawa, T.; Kusunoki, Y.; Kunimori, K.; Tomishige, K. J. Catal.
15
hypothesis on commercial ZnO and clearly verifies that the
exposure degree of the {100} nonpolar planes is responsible for
the shape-dependent promotion effect of ZnO in the CRH of
glycerol.
(
(
2
(
(
006, 240, 213−221.
Interestingly, when the straight line in Figure 8 was extended
to the zero BET surface area, the extrapolated TOF of 1,2-PDO
was about 11% higher than that on bare RQ Ni Mo ,
9) Perosa, A.; Tundo, P. Ind. Eng. Chem. Res. 2005, 44, 8535−8537.
10) Huber, G. W.; Cortright, R. D.; Dumesic, J. A. Angew. Chem., Int.
40
10
Ed. 2004, 43, 1549−1551.
suggesting the occurrence of the synergistic effect between
(11) D’Hondt, E.; Van de Vyver, S.; Sels, B. F.; Jacobs, P. A. Chem.
Commun. 2008, 45, 6011−6012.
ZnO and RQ Ni Mo . It seems that the CRH of glycerol over
40
10
(12) Roy, D.; Subramaniam, B.; Chaudhari, R. V. Catal. Today 2010,
bare RQ Ni Mo is limited by its dehydration activity, rather
40
10
1
(
56, 31−37.
than the hydrogenation ability. By first principles calculation of
the thermodynamic profile of glycerol hydrogenolysis on
transition metal catalysts including Ni, Coll et al. concluded
13) Gandarias, I.; Arias, P. L.; Requies, J.; Guemez, M. B.; Fierro, J.
̈
L. G. Appl. Catal., B 2010, 97, 248−256.
(14) Wawrzetz, A.; Peng, B.; Hrabar, A.; Jentys, A.; Lemonidou, A.
44
that the metal catalyst has three roles. Aside from catalyzing
the hydrogenation of acetol to 1,2-PDO, it also stabilizes acetol
and renders the dehydration more exothermic. Moreover, the
adsorption on the surface modifies the relative stability of
acetol, with implications on the reaction selectivity.
A.; Lercher, J. A. J. Catal. 2010, 269, 411−420.
(15) Hu, J. Y.; Liu, X. Y.; Fan, Y. Q.; Xie, S. H.; Pei, Y.; Qiao, M. H.;
Fan, K. N.; Zhang, X. X.; Zong, B. N. Chin. J. Catal. 2013, 34, 1020−
1026.
(16) Wang, Y. M.; Kovac
Staemmler, V.; Qiu, H. S.; Traeger, F.; Langenberg, D.; Muhler, M.;
Woll, C. Angew. Chem., Int. Ed. 2007, 46, 5624−5627.
17) Wang, Y. M.; Xia, X.; Urban, A.; Qiu, H. S.; Strunk, J.; Meyer,
B.; Muhler, M.; Woll, C. Angew. Chem., Int. Ed. 2007, 46, 7315−7318.
18) Somorjai, G. A. Introduction to Surface Chemistry and Catalysis;
John Wiley & Sons: New York, 1994, Chapter 7.
́ ̌
ik, R.; Meyer, B.; Kotsis, K.; Stodt, D.;
5. CONCLUSIONS
̈
(
Hexagonal ZnO crystals with different aspect ratios were
̈
synthesized and employed as cocatalysts with RQ Ni Mo in
40
10
(
the CRH of glycerol. The promotion effect of these ZnO
crystals was highly dependent on their crystal habits; that is, the
more abundant the nonpolar planes, the more prominent the
improvement in the selectivity to the hydrogenolysis products
and the production rate of 1,2-PDO. The possible influences of
(
19) Garcia, S. P.; Semancik, S. Chem. Mater. 2007, 19, 4016−4022.
(20) Wang, Z. L. J. Phys.: Condens. Matter 2004, 16, R829−R858.
(21) Li, G. R.; Hu, T.; Pan, G. L.; Yan, T. Y.; Gao, X. P.; Zhu, H. Y. J.
Phys. Chem. C 2008, 112, 11859−11864.
2
286
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287

ACS Catalysis
Research Article
(
22) Peng, Y.; Xu, A. W.; Deng, B.; Antonietti, M.; Co
̈
lfen, H. J. Phys.
(57) Miyazawa, T.; Koso, S.; Kunimori, K.; Tomishige, K. Appl.
Catal., A 2007, 318, 244−251.
Chem. B 2006, 110, 2988−2993.
(
(
(
23) Cheng, B.; Samulski, E. T. Chem. Commun. 2004, 8, 986−987.
(58) Miyazawa, T.; Koso, S.; Kunimori, K.; Tomishige, K. Appl.
Catal., A 2007, 329, 30−35.
24) Vayssieres, L. Adv. Mater. 2003, 15, 464−466.
25) Hu, J. Q.; Li, Q.; Wong, N. B.; Lee, C. S.; Lee, S. T. Chem.
(59) Chai, S. H.; Wang, H. P.; Liang, Y.; Xu, B. Q. Green Chem. 2007,
9, 1130−1136.
Mater. 2002, 14, 1216−1219.
26) Shen, L.; Bao, N.; Yanagisawa, K.; Domen, K.; Grimes, C. A.;
Gupta, A. J. Phys. Chem. C 2007, 111, 7280−7287.
27) Yoshida, Y.; Mitani, Y.; Itoi, T.; Izumi, Y. J. Catal. 2012, 287,
90−202.
28) Bai, S.; Hu, J. W.; Li, D.; Luo, R.; Chen, A.; Liu, C. C. J. Mater.
Chem. 2011, 21, 12288−12294.
29) Guo, L.; Ji, Y. L.; Xu, H. B. J. Am. Chem. Soc. 2002, 124, 14864−
4865.
30) Wang, G.; Deng, Y.; Guo, L. Chem.Eur. J. 2010, 16, 10220−
0225.
31) Xu, F.; Zhang, P.; Navrotsky, A.; Yuan, Z. Y.; Ren, T. Z.; Halasa,
M.; Su, B. L. Chem. Mater. 2007, 19, 5680−5686.
32) Kuo, C. L.; Kuo, T. J.; Huang, M. H. J. Phys. Chem. B 2005, 109,
0115−20121.
33) Tong, Y.; Liu, Y.; Dong, L.; Zhao, D.; Zhang, J.; Lu, Y.; Shen,
D.; Fan, X. J. Phys. Chem. B 2006, 110, 20263−20267.
34) Li, P.; Liu, H.; Lu, B.; Wei, Y. J. Phys. Chem. C 2010, 114,
1132−21137.
35) Umar, A.; Hahn, Y. B. Appl. Phys. Lett. 2006, 88, 173120−
73122.
(
(60) Alhanash, A.; Kozhevnikova, E. F.; Kozhevnikov, I. V. Appl.
Catal., A 2010, 378, 11−18.
(
1
(
(61) Katryniok, B.; Paul, S.; Bellier
Green Chem. 2010, 12, 2079−2098.
̀
e-Baca, V.; Reye, P.; Dumeignil, F.
(
1
(
1
(
(
2
(
(
2
(
1
(
(
36) Wu, J. J.; Liu, S. C. J. Phys. Chem. B 2002, 106, 9546−9551.
37) Cong, G. W.; Wei, H. Y.; Zhang, P. F.; Peng, W. Q.; Wu, J. J.;
Liu, X. L.; Jiao, C. M.; Hu, W. G.; Zhu, Q. S.; Wang, Z. G. Appl. Phys.
Lett. 2005, 87, 231903−231906.
(
(
38) Wu, J. J.; Liu, S. C. Adv. Mater. 2002, 14, 215−218.
39) Beane, G. A.; Morfa, A. J.; Funston, A. M.; Mulvaney, P. J. Phys.
Chem. C 2012, 116, 3305−3310.
40) Wu, X. L.; Siu, G. G.; Fu, C. L.; Ong, H. C. Appl. Phys. Lett.
001, 78, 2285−2287.
41) Andelman, T.; Gong, Y. Y.; Polking, M.; Yin, M.; Kuskovsky, I.;
Neumark, G.; O’Brien, S. J. Phys. Chem. B 2005, 109, 14314−14318.
42) Strunk, J.; Kahler, K.; Xia, X.; Muhler, M. Surf. Sci. 2009, 603,
776−1783.
43) Auneau, F.; Michel, C.; Delbecq, F.; Pinel, C.; Sautet, P.
Chem.Eur. J. 2011, 17, 14288−14299.
44) Coll, D.; Delbecq, F.; Aray, Y.; Sautet, P. Phys. Chem. Chem.
Phys. 2011, 13, 1448−1456.
45) Cortright, R. D.; Davda, R. R.; Dumesic, J. A. Nature 2002, 418,
64−967.
46) Xie, F. Z.; Chu, X. W.; Hu, H. R.; Qiao, M. H.; Yan, S. R.; Zhu,
(
2
(
(
1
(
(
(
9
(
Y. L.; He, H. Y.; Fan, K. N.; Li, H. X.; Zong, B. N.; Zhang, X. X. J.
Catal. 2006, 241, 211−220.
(
47) Deng, Z.; Chen, M.; Gu, G.; Wu, L. J. Phys. Chem. B 2008, 112,
1
6−22.
48) Xu, L.; Hu, Y. L.; Pelligra, C.; Chen, C. H.; Jin, L.; Huang, H.;
Sithambaram, S.; Aindow, M.; Joesten, R.; Suib, S. L. Chem. Mater.
009, 21, 2875−2885.
49) Liao, F.; Huang, Y.; Ge, J.; Zheng, W.; Tedsree, K.; Collier, P.;
Hong, X.; Tsang, S. C. Angew. Chem., Int. Ed. 2011, 50, 2162−2165.
50) Xia, S.; Yuan, Z.; Wang, L.; Chen, P.; Hou, Z. Appl. Catal., A
011, 403, 173−182.
51) Si, R.; Flytzani-Stephanopoulos, M. Angew. Chem., Int. Ed. 2008,
7, 2884−2887.
52) Xie, X. W.; Li, Y.; Liu, Z. Q.; Haruta, M.; Shen, W. J. Nature
009, 458, 746−749.
53) Xiao, X. L.; Liu, X. F.; Zhao, H.; Chen, D. F.; Liu, F. Z.; Xiang, J.
H.; Hu, Z. B.; Li, Y. D. Adv. Mater. 2012, 24, 5762−5766.
54) Pan, J.; Liu, G.; Lu, G. Q.; Cheng, H. M. Angew. Chem., Int. Ed.
011, 50, 2133−2137.
55) Zhou, X.; Lan, J.; Liu, G.; Deng, K.; Yang, Y.; Nie, G.; Yu, J.;
Zhi, L. Angew. Chem., Int. Ed. 2012, 51, 178−182.
56) Jang, E. S.; Won, J. H.; Hwang, S. J.; Choy, J. H. Adv. Mater.
006, 18, 3309−3312.
(
2
(
(
2
(
4
(
2
(
(
2
(
(
2
2
287
dx.doi.org/10.1021/cs400526s | ACS Catal. 2013, 3, 2280−2287