Selective hydrogenolysis of glycerol to 1,3-propanediol over a Pt/WO 3/TiO2/SiO2 catalyst in aqueous media

Article abstract of DOI:10.1016/j.apcata.2010.10.002

SiO₂-supported Pt/WO₃/TiO₂ catalysts were prepared; they were found to be more active and selective than the Pt/WO ₃/TiO₂ catalyst for glycerol hydrogenolysis to 1,3-propanediol in a slurry batch reactor. The influences of catalyst component, reaction medium, reaction temperature, H₂ pressure and reaction time on glycerol hydrogenolysis over the Pt/WO₃/TiO₂/SiO ₂ catalyst were investigated. XRD, TEM, NH₃-TPD and Py-IR characterization were employed to reveal the roles of WO₃ and TiO₂ in the performance of the Pt based-catalysts. XRD patterns and TEM images showed that the presence of TiO₂ species in the catalyst favored the dispersion of platinum. The weak Br?nsted acid sites formed by addition of WO₃ to the catalyst were concluded to play a key role in selective formation of 1,3-propanediol, based on the results of NH ₃-TPD and Py-IR characterization.

Full text of DOI:10.1016/j.apcata.2010.10.002

Applied Catalysis A: General 390 (2010) 119–126
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Selective hydrogenolysis of glycerol to 1,3-propanediol over a Pt/WO /TiO /SiO
2
3
2
catalyst in aqueous media
a,c
a
a,b,∗
, Ronghe Lina,c, Jingwei Lia,c, Wenda Donga,c
,
Leifeng Gong , Yuan Lu , Yunjie Ding
a
a
Tao Wang , Weimiao Chen
Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
Graduate School of Chinese Academy of Sciences, Beijing 100049, PR China
a
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 June 2010
Received in revised form
2 3 2
SiO -supported Pt/WO /TiO catalysts were prepared; they were found to be more active and selective
than the Pt/WO3/TiO2 catalyst for glycerol hydrogenolysis to 1,3-propanediol in a slurry batch reactor.
The influences of catalyst component, reaction medium, reaction temperature, H2 pressure and reaction
time on glycerol hydrogenolysis over the Pt/WO3/TiO2/SiO2 catalyst were investigated. XRD, TEM, NH3-
TPD and Py-IR characterization were employed to reveal the roles of WO3 and TiO2 in the performance
of the Pt based-catalysts. XRD patterns and TEM images showed that the presence of TiO2 species in the
catalyst favored the dispersion of platinum. The weak Brønsted acid sites formed by addition of WO3 to
the catalyst were concluded to play a key role in selective formation of 1,3-propanediol, based on the
results of NH3-TPD and Py-IR characterization.
3
0 September 2010
Accepted 1 October 2010
Available online 27 October 2010
Keywords:
Glycerol
Hydrogenolysis
1
,3-Propanediol
© 2010 Elsevier B.V. All rights reserved.
Acidity
1
. Introduction
,3-Propanediol (1,3-PD) is an important chemical intermediate
vent containing tungstic acid. After reaction for 24 h at 473 K under
3
2 MPa pressure, 1,3-PD was produced with 45% selectivity [7].
1
Shell patented an approach converting glycerol in water-sulfolane
mixed solvent at 413 K and 6 MPa pressure in presence of homoge-
nous palladium complex and methane sulfonic acid. After a 10 h
reaction, 31% selectivity to 1,3-PD was obtained [8]. The direct
hydrogenolysis of glycerol over heterogeneous catalysts has been
studied in order to overcome the common problem of catalyst sep-
aration in homogeneous processes. Chaminand et al. first reported
that the addition of tungstic acid to Ru/C catalytic system improved
the conversion of glycerol to 1,3-PD in water, and 1,3-PD selectiv-
ity of 12% was obtained in sulfolane [2]. Kurosaka et al. carried
used mostly in the manufacture of polytrimethylene terephtha-
late (PTT) [1]. Based on petrochemicals, the industrial production
of 1,3-PD begins with hydroformylation of ethylene oxide (Shell
route) or hydration of acrolein (Degussa–DuPont route) [2]. With
the decrease of petroleum resource, it is increasingly more impera-
tive to develop an alternative route for the sustainable production
of 1,3-PD.
Glycerol is one of the top 12 building block chemicals derived
from plant sources [3]. The rapid development of bio-diesel produc-
tion makes large quantities of glycerol available as a byproduct. As
a result, glycerol could be the ideal renewable feedstock to produce
out the transformation of glycerol over a Pt/WO /ZrO2 catalyst
3
in 1,3-dimethyl-2-imidazolidinone solvent; the yields to 1,3-PD,
1,2-PD and 1-propanol were 24.2%, 12.5% and 27.5%, respectively
[9]. Huang et al. converted solvent-free glycerol into 1,3-PD over
1
,3-PD. Research has been reported on the conversion of glycerol
into 1,3-PD through bio-catalytic, homogeneous or heterogeneous
processes [4–6]. Glycerol fermentation produced 1,3-PD in good
yield (>70%), but with a low metabolic efficiency [4]. Che developed
a process of glycerol hydrogenolysis using a homogeneous rhodium
Cu-H SiW12O40/SiO2 in vapor phase; the selectivities to 1,3-PD
4
and to 1,2-PD were detected to be 32.1% and 22.2%, respectively
[10]. Tomishige and co-workers have done much work on glycerol
complex catalyst (Rh(CO) (acac)) in 1-methyl-2-pyrrolidinone sol-
hydrogenolysis. They have found that modification of Rh/SiO cata-
2
2
lyst with Re was in favor of 1,3-PD formation from glycerol [11,12].
Recently, they developed the rhenium-modified iridium catalyst
with H SO4 as additive, over which the yield of 1,3-PD reached
2
∗ Corresponding author at: Dalian National Laboratory for Clean Energy, Dalian
Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China.
Tel.: +86 411 84379143; fax: +86 411 84379143.
3
8% at 81% conversion of glycerol [13]. Although the progress men-
tioned above is substantial, to put this reaction into industrial use,
one must solve many remaining problems. Most solid catalysts
E-mail address: dyj@dicp.ac.cn (Y. Ding).
0
926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.10.002

1
20
L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
Table 1
a
Performance of hydrogenolysis of glycerol over different catalysts .
Catalysts
Conv. (%)
Selectivity (mol%)b
,3-PD
1,3-PD/1,2-PD
1
1,2-PD
1-PO
2-PO
Othersc
Pt/WO3/TiO2
Pt/WO3/TiO2/SiO2
Pt/WO3/TiO2/AC
Pt/WO3/TiO2/Al2O3
Pt/WO3/TiO2/HZSM-5
7.5
15.3
5.5
5.1
3.8
43.7
50.5
30.8
11.3
41.1
11.7
9.2
29.0
59.5
7.6
28.2
25.1
14.5
6.7
7.6
8.8
6.1
2.1
1.6
8.8
6.4
19.6
20.4
28.5
3.74
5.49
1.06
0.19
5.41
21.2
a
Reaction conditions: 10 wt.% glycerol aqueous solution, 40 ml; initial H2 pressure, 5.5 MPa; reaction temperature, 453 K; reaction time, 12 h; catalyst loading, 2 ml.
b
c
1
,3-PD: 1,3-propanediol, 1,2-PD: 1,2-propanediol, 1-PO: 1-propanol, 2-PO: 2-propanol.
Others include ethanol, methanol, methane, ethane, propane and carbon dioxide.
without liquid acids as additive are not selective in the formation of
tion for 12 h, the reactor was cooled down to room temperature.
Gas phase products were collected in a pocket sampler and ana-
lyzed using an Agilent 3000 A Micro GC. The sample of liquid phase
products was esterified by acetic anhydride and analyzed using an
Agilent 7890 equipped with a flame ionization detector and a HP-5
capillary column. The conversion of glycerol and the selectivity to
products were calculated by the following equations:
1
1
,3-PD from glycerol. 1,3-PD selectivity is no more than 35%. Much
,2-PD is produced as byproduct (1,3-PD/1,2-PD ≤ 2). Another dis-
advantage of existing heterogeneous processes arises from use of
organic solvents, which violates environment protection. It is well
known that glycerol is strongly hygroscopic and that crude glyc-
erol from bio-diesel production contains water unavoidably, which
makes water the desired solvent for glycerol conversion in view of
environmental and economical viability. Accordingly, it is mean-
ingful to study glycerol hydrogenolysis in aqueous media. As far
as supported metal catalysts reported are concerned, coexistence
of metal components with acidic species in reaction system was
supposed to be essential for the selective formation of 1,3-PD from
glycerol. In this research, we loaded metal–acid bi-functional cat-
conversion of glycerol (%)
moles of glycerol initially added − moles of glycerol that remained
=
× 100
moles of glycerol initially added
C mole of specific product
sum of C mole of glycerol consumed
selectivity (%) =
× 100
alyst Pt/WO /TiO2 on silica and hydrogenolysed glycerol in water
3
in an attempt to develop an environmentally benign and economi-
cally feasible route of glycerol conversion to 1,3-PD, which holds a
much higher commercial value than 1,2-PD.
2.3. Catalyst characterization
BET surface area and average pore volume of supports were
measured on a Quantachrome AUTOSORB-1 adsorption apparatus
using nitrogen as adsorbate.
2
. Experimental
Powder X-ray diffraction (XRD) analysis was performed on an
X’pert PRO/PANalytical Diffractometer using Cu-K␣ radiation and
operated at 40 KV and 40 mA. The X-ray patterns were recorded in
2.1. Catalyst preparation
◦
◦
◦
2
ꢀ values ranging from 10 to 70 at the scanning rate of 10 /min.
Commercial supports such as SiO2 (Qingdao Haiyang Chemi-
cal), activated carbon (AC for short, Beijing Guanghua Activated
TEM measurements were carried out using a HITACHIH600
microscope operated at an accelerating voltage of 120 kV. The cat-
alysts (<100 mesh) were suspended in ethanol and placed onto a
carbon film supported over a copper grid.
Carbon Co., Ltd), Al O3 (Chalco) and HZSM-5 (Nankai Catalyst)
2
were used as received. TiO2 as support used in Pt/WO /TiO2 cat-
3
alyst was obtained by hydrolysis of tetrabutyl titanate by adding
drop-wise an ammonium hydroxide solution, filtering and washing
the precipitate obtained with deionized water to pH = 7, followed
by drying at 373 K for 24 h and then calcining at 873 K for 6 h
in air. All the catalysts used in this research were prepared by
stepwise impregnation. Pt/WO /TiO /SiO sample was prepared by
Temperature programmed desorption of ammonia (NH -TPD)
3
was carried out in a Micromeritics Autochem 2910. A 200 mg sam-
ple was used. Prior to each measurement, a sample was pretreated
in a helium flow at 873 K for 0.5 h. After cooling to 423 K under
continuous flow of helium, the sample was adsorbed with NH3
using pulse model until saturation. Then the sample was heated
from 373 K to 823 K with a ramp of 10 K/min in a helium flow of
3
2
2
impregnating silica with an ethanol solution of tetrabutyl titanate,
an aqueous (NH ) (H W12O40) solution and an aqueous H PtCl6
4
6
2
2
3
0 ml/min and the desorbed NH3 was monitored simultaneously
with TCD.
Fourier transform infrared spectra of adsorbed pyridine were
solution in sequence; each impregnating step was followed by
drying at 393 K overnight and calcination in air at 873 K for 6 h.
Pt/WO /TiO /Al O , Pt/WO /TiO /AC, Pt/WO /TiO /HZSM-5 and
3
2
2
3
3
2
3
2
obtained using a BRUKER EQUINOX 55 spectrometer equipped with
a DTGS detector. The self- supported sample (20 mg) was first pre-
treated at 623 K for 0.5 h. Subsequently, a FT-IR spectrum (denoted
as spectrum A) was recorded. After adsorption of pyridine at room
Pt/WO /TiO catalysts were prepared in ways similar to that used
3
2
for Pt/WO /TiO /SiO catalyst. Loadings of Pt, WO and TiO in sup-
3
2
2
3
2
ported Pt/WO /TiO /SiO catalysts were 2 wt.%, 5 wt.% and 10 wt.%,
3
2
2
respectively, unless otherwise specified. The weight ratio of Pt,
temperature, the sample was evacuated at 423 or 473 K for 0.5 h
WO3 and TiO2 in Pt/WO /TiO2 was 2:5:100.
−2
3
(
10 Pa), and then spectrum B was recorded. All spectra were
−
1
recorded at 4 cm resolution with 16 co-added scans. Subtracting
spectrum A from spectrum B gave a final spectrum.
2.2. Catalyst evaluation
Glycerol hydrogenolysis reactions were carried out in a 100 ml
3. Results and discussion
autoclave. 40 ml of glycerol aqueous solution (10 wt.%) and 2 ml
of catalyst were introduced into the autoclave. After being purged
3.1. Influence of support
six times with pure H , the reactor was pressurized with H2 up to
2
5
.5 MPa, and then heated to the reaction temperature of 453 K at a
Table 1 summarizes the experimental results of glycerol
hydrogenolysis over different catalysts in aqueous media. The
ramp of 5 K/min. The agitation rate was set at 700 rpm. After reac-

L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
121
Table 2
A
Textural properties of different supports.
5
4
0
0
Conversion
1,3-PD selectivity
SBETa
Vpb
Sample
TiO2
SiO2
AC
Al2O3
HZSM-5
5
355
1276
191
0.04
1.10
0.99
0.50
0.21
30
289
a
2
Multipoint BET surface area (m /g).
Total pore volume (cm /g).
2
1
0
0
0
b
3
Pt/WO /TiO catalyzed glycerol to 1,3-PD with a high selectiv-
3
2
ity of 43.7%, but glycerol conversion was as low as 7.5%, because
2
the support TiO2 was bulk-like (BET surface area of 5 m /g, as
shown in Table 2). The low surface area of Pt/WO /TiO2 led
0
2
4
6
8
10
3
us to consider the potential of improving glycerol conversion
by supporting catalytic components on materials with high sur-
face areas, such as SiO , Al O , HZSM-5 and AC. As shown in
WO loading (wt%)
3
6
0
B
2
2
3
Conversion
1,3-PD selectivity
Table 1, glycerol conversion over the Pt/WO /TiO /SiO cata-
3
2
2
50
lyst was up to 15.3%, almost twice as large as that obtained
over the Pt/WO /TiO2 catalyst, more importantly, higher 1,3-PD
3
selectivity and higher 1,3-PD/1,2-PD ratio were also obtained.
40
The Pt/WO /TiO /SiO catalyst was obviously superior to the
3
2
2
Pt/WO /TiO catalyst. However, glycerol conversion and 1,3-PD
3
2
3
2
0
0
selectivity over the Pt/WO /TiO /AC, Pt/WO /TiO /HZSM-5 and
3
2
3
2
Pt/WO /TiO /Al O catalysts were lower than those obtained over
3
2
2
3
the Pt/WO /TiO catalyst. Rather than a promotional effect of SiO2
3
2
on catalytic performance, some negative effects were observed in
the case of AC, HZSM-5 and Al O .
10
0
2
3
Textural properties of different supports are listed in Table 2. The
order of surface area of supports is as follows: AC > SiO > HZSM-
2
5
> Al O > TiO , which is not in good coincidence with the catalytic
2 3 2
0
2
4
6
8
10
12
14
16
performances of their corresponding catalysts. This indicates that
the support nature affected catalytic activity more dramatically
than the textural property. Because of its unique properties, sil-
ica is widely used as an inorganic support for organic reactions
TiO loading (wt%)
2
Fig. 1. Influence of WO3 (A) and TiO2 (B) loadings on Pt/WO3/TiO2/SiO2 catalysts.
in water [14]. The better performance over the Pt/WO /TiO /SiO
catalyst was ascribed to not only the high surface area of SiO2,
but also to the synergistic effect supposed to exist between the
3
2
2
Fig. 1. The increase of TiO2 loading resulted in the increase of glyc-
erol conversion and 1,3-PD selectivity, but further increase led to
the decrease of glycerol conversion and 1,3-PD selectivity. A similar
Pt/WO /TiO component and silica. Strong affinity for glycerol of
3
2
tendency was observed in the case of WO . The optimal loadings
3
silica was found in the etherification and esterification of glycerol
15,16]. The stronger interaction between silica and glycerol than
of TiO2 and WO3 were 10% and 5%, respectively, in which cases
glycerol conversion and 1,3-PD selectivity attained maxima: 15.3%
[
between silica and other supports, such as HZSM-5, Al O3 and AC,
2
resulted in more glycerol absorbing on the catalyst surface and get-
ting access to catalytic sites, and thus higher glycerol conversion
was achieved over the Pt/WO /TiO /SiO catalyst. Enhancement of
3
2
2
: Pt
1
,3-PD selectivity may be due to the hydrophilic nature of silica
which allows a rapid desorption of the product from the catalyst
surface; this inhibits further hydrogenolysis reaction, as reported
in silica-based acid catalysts for the esterification of glycerol [16].
2
Pt/SiO
Pt/5WO /SiO
Pt/10TiO /SiO
Pt/5WO /10TiO /SiO
Pt/5WO /5TiO /SiO
2Pt/5WO /15TiO /SiO
Pt/2WO /10TiO /SiO
2
2
3
.2. Influences of TiO2 and WO3 on Pt/WO /TiO /SiO catalysts
3 2 2
2
As shown in Table 3, Pt/SiO sample showed poor catalytic activ-
2
2
ity. Only 1.1% of glycerol was converted; even worse, no 1,3-PD
was produced from glycerol, while 1,2-PD was the main product.
The presence of TiO2 in Pt/TiO /SiO2 catalyst enhanced glycerol
2
2
conversion up to 4%, however, 1,3-PD selectivity was still negli-
2Pt/10WO /10TiO /SiO
gible (6.3%). The existence of WO3 in Pt/WO /SiO2 catalyst made
3
1
,3-PD selectivity increase greatly from 0.0% to 35.6%. The coexis-
tence of TiO2 and WO3 highlighted Pt/WO /TiO /SiO catalyst in
terms of the highest activity (15.3%) and 1,3-PD selectivity (50.5%),
which indicates that TiO2 and WO3 were essential components in
3
2
2
1
0
20
30
40
50
60
70
2
θ /degree
Pt/WO /TiO /SiO catalyst for selective hydrogenolysis of glycerol
3
2
2
Fig. 2. XRD patterns of Pt/WO3/TiO2/SiO2 catalysts with different WO3 and TiO2
loadings. The numbers before Pt, WO3 and TiO2 stand for their weight percentages
in the catalysts.
to 1,3-PD. The detailed effects of TiO2 and WO3 loadings on the
catalytic performance of Pt/WO /TiO /SiO catalysts are shown in
3
2
2

1
22
L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
Table 3
Influences of WO3 and TiO2 components on the performance of Pt/WO3/TiO2/SiO2 catalysts.
a
Catalysts
Conv. (%)
Selectivity (mol%)
1,3-PD/1,2-PD
1,3-PD
1,2-PD
1-PO
2-PO
Others
Pt/SiO2
Pt/WO3/SiO2
Pt/TiO2/SiO2
Pt/WO3/TiO2/SiO2
1.1
2.8
4.0
0.0
35.6
6.3
41.6
14.7
66.1
9.2
4.5
19.1
4.5
0.0
7.6
0.8
8.8
53.4
23.0
22.3
6.4
0.00
2.42
0.10
5.49
15.3
50.5
25.1
a
Reaction conditions: 10 wt.% glycerol aqueous solution, 40 ml; initial H2 pressure, 5.5 MPa; reaction temperature, 453 K; reaction time, 12 h; catalyst loading, 2 ml.
and 50.5%, respectively. Glycerol conversion increased linearly with
increase of TiO2 loading from 0.0% to its optimal value, and a lin-
ear correlation was observed between 1,3-PD selectivity and WO3
loading with WO3 loading increasing from 0.0% to 5%. Accordingly,
peaks are observed on only two TiO -free catalysts (i.e. Pt/SiO2
2
and Pt/WO /SiO ), which implies that the introduction of TiO
3
2
2
component improved the dispersion of platinum or inhibited the
crystallization of platinum on the catalysts. On the other hand, The
TiO influenced the reaction rate more greatly, while WO affected
XRD signals of platinum crystal peaks over the Pt/WO /SiO2 sam-
2
3
3
1
,3-PD selectivity more drastically. XRD, TEM, NH -TPD and Py-IR
ple are much weaker than those on the Pt/SiO2 sample, indicating
the WO3 contributed to the enhancement of platinum dispersion
to some extent.
3
characterizations were carried out to reveal the function of TiO2
and WO3 species in the catalyst.
XRD patterns of Pt/WO /TiO /SiO catalysts with different WO
and TiO2 loadings are presented in Fig. 2. No crystalline phase
related to tungstic and titanic species can be observed, indicat-
The TEM images give direct evidence of the platinum parti-
cle sizes. As shown in Fig. 3, the platinum particle sizes varied
with catalyst components. Big platinum particles with an aver-
age size of up to 28.1 nm were present in TEM images of the
Pt/SiO2 catalyst, moreover, platinum particles obviously congre-
3
2
2
3
ing good dispersions of WO and TiO2 on the surface of SiO2.
3
◦
◦
◦
The peaks at 2ꢀ = 39.8 , 46.2 and 67.5 represent the reflection
of Pt (1 1 1), Pt (2 0 0) and Pt (2 2 0), respectively. Although plat-
inum content was identical (i.e. 2%) in all catalysts, platinum metal
gated. The average particle sizes in the Pt/WO /SiO , Pt/TiO /SiO
3
2
2
2
and Pt/WO /TiO /SiO samples were 19.5 nm, 11.0 nm, 10.3 nm,
3
2
2
Fig. 3. TEM images of (a) Pt/SiO2; (b) Pt/WO3/SiO2; (c) Pt/TiO2/SiO2 and (d) Pt/WO3/TiO2/SiO2.

L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
123
o
2
00 C, 5WO /10TiO /SiO
3
2
2
1540
1447
o
2
00 C, 10TiO /SiO
2
2
o
1
50 C, 5WO /10TiO /SiO
3 2 2
o
1
50 C, 10TiO /SiO
2
2
100
150
200
250
300
350
400
450
500
550
o
Temperature ( C)
1800
1700
1600
1500
1400
1300
1200
-
1
Wavenumber (cm )
B
Fig. 5. FT-IR spectra of adsorbed pyridine of the 10TiO2/SiO2 and 5WO3/10TiO2/SiO2
catalysts at 150 C and 200 C.
◦
◦
of TiO2 loading resulted in a slight drop, which is consistent with
the reaction result shown in Fig. 1(A). The increase of WO3 load-
ing from 2% to 5% led to a distinct increase of the amount of acid
sites from 182 to 231 ␮mol/g, and the ammonia desorption peak
◦
◦
temperature increased slightly from 172 C to 177 C. The addition
of WO3 enhanced the acidity of catalyst in terms of number and
strength. WO3 loading of 5 wt.% seemed to provide suitable acidity
in the catalyst to improve 1,3-PD selectivity to a high level of 50.5%,
and further increase of WO loading resulted in a negative influence
3
1
00 150 200 250 300 350 400 450 500 550
on the catalytic performance of the catalyst, as shown in Fig. 1(B).
One possible reason is that, for the Pt/WO3/TiO2/SiO2 sample, a bi-
functional catalyst, the number of acid sites should match with the
number of metal sites.
o
Temperature ( C)
Fig. 4. NH3-TPD profiles of WO3/TiO2/SiO2 with different TiO2 (A) and WO3 (B)
loadings.
To get further information about the type of acid sites in
the WO /TiO /SiO sample, we recorded the infrared spectra of
3
2
2
adsorbed pyridine of the 10TiO /SiO and 5WO /10TiO /SiO cat-
2
2
3
−1
2
2
respectively, much smaller than that in the Pt/SiO2 catalyst. No
◦
◦
alysts at 150 C and 200 C. The band at 1450 cm corresponds to
the coordinated pyridine adsorbed on Lewis acid sites, and the band
congregation of platinum particles was found in TiO -containing
2
catalysts. Due to weak interaction between platinum and silica,
platinum particles on Pt/SiO2 catalyst are apt to grow up and crys-
tallize in calcination at high temperature. Based on XRD and TEM
characterization, introduction titania to the catalyst enhanced plat-
inum dispersion. TiO2 interacts strongly with Pt, which play a role
in preventing the crystallization of platinum in calcination at ele-
vated temperature [17]. It has been reported that no Pt diffraction
peaks were detected by TEM in Pt-based catalysts used in glycerol
hydrogenolysis with Pt particles of about 10 nm, to be specific, the
Pt particle size was 12.9 nm as measured by TEM, but no XRD peak
of platinum was observed [18].
−1
at 1540 cm is attributed to pyridine ions formed by interaction
with BrØnsted acid sites [23]. As shown in Fig. 5, the TiO /SiO2 and
2
WO /TiO /SiO samples exhibited the characteristic band of Lewis
3
2
2
−1
acid sites at 1447 cm , whose integrated areas are almost iden-
tical in Table 4, indicating the same number of Lewis acid sites in
both samples. The band of BrØnsted acid sites at 1540 cm was
−1
observed remarkably in the WO /TiO /SiO sample, while it was
3
2
2
indiscernible in the TiO /SiO2 catalyst. This is confirmed by their
2
integrated areas in Table 4. WO3 plays a key role in generating
BrØnsted acid sites, which is consistent with the literature [24].
BrØnsted acid sites are supposed to be beneficial for dehydration
of glycerol to 3-hydroxypropionaldehyde, which can be hydro-
genated to 1,3-PD, so WO3 affected 1,3-PD selectivity, as shown
Glycerol hydrogenolysis is considered as a bi-functional reaction
[
19–22]. 1,3-PD cannot be obtained selectively over those sup-
ported metal catalysts such as Ru/C and Rh/C without the presence
of an acidic component. Acidity of the catalyst plays a crucial role in
Table 4
selective hydrogenolysis of glycerol to 1,3-PD. The NH -TPD tech-
3
Integrated areas of L and B acid sites in FT-IR spectra of adsorbed pyridine on the
nique is usually used to measure the acidic properties of supports.
The strength of acidity is characterized by the temperature of the
desorptionpeak, while the numberof acidic sitesis measured bythe
amount of NH3 desorbed. NH -TPD profiles of the WO /TiO /SiO
10TiO2/SiO2 and 5WO3/10TiO2/SiO2 catalysts.
◦
Integrated areaa
Samples
Temperature ( C)
3
3
2
2
L
B
sample with different TiO and WO loadings are shown in Fig. 4. All
2
3
1
0TiO2/SiO2
150
171
68
177
56
4
1
28
10
the samples exhibit a single ammonia desorption peak at around
2
00
150
00
◦
1
75 C, indicating that all weak acid sites on the WO /TiO /SiO
5WO /10TiO /SiO
3
2
2
3
2
2
2
samples were observed. The amount of acid sites increased gradu-
ally with increase of TiO loading. The maximum (231 ␮mol/g) was
a
Integrated areas of L and B acid sites were based on band at 1447 cm−1 and
2
−
1
obtained at TiO2 loading of 10%. It was found that further increase
1540 cm , respectively.

1
24
L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
5
5
4
4
3
3
2
2
1
1
5
0
5
0
5
0
5
0
5
0
55
10
8
50
45
40
35
30
25
Conversion
,2-PD
1,3-PD/1,2-PD
1,3-PD
1-PO
glycerol conversion
1
1
,3-PD selectivity
,2-PD selectivity
1
6
4
2
1
1
0
5
0
2
5
0
5
0
160
170
180
190
Temperature (°C)
water
ethanol
NMI
sulfane
Fig. 7. Influence of reaction temperature on the hydrogenolysis of glycerol over the
Pt/WO3/TiO2/SiO2 catalyst.
Fig. 6. Influence of solvent on hydrogenolysis of glycerol over the Pt/WO3/TiO2/SiO2
catalyst. NMI: N-methylimidazole.
case of water as solvent. Hydrogenolysis of glycerol to 1,3-PD can
be taken as a substitution reaction in which H atom substitutes
for the central hydroxyl group in glycerol. It is well known that
hydroxyl groups of alcohols can be activated by Lucas reagent,
a kind of acid catalyst, and that secondary alcohols are a little
more active than primary ones for chlorination reaction in Lucas
reagent. Nimlos et al. reported that proton affinity for the central
in Fig. 1(B). The integrated area of the bands drastically decreased
when the samples were evacuated from 150 C to 200 C, which
indicated that the acid sites were relatively weak ones, as confirmed
◦
◦
by NH -TPD profiles in Fig. 1.
3
On the basis of the above observations, the presence of WO3
and TiO2 on the Pt/WO /TiO /SiO catalyst impact greatly on the
3
2
2
glycerol hydrogenolysis reaction. Suitable loading of TiO increases
2
−1
hydroxyl in glycerol was 195.4 kcal mol , while for the termi-
the activity and selectivity to 1,3-PD by improvement of platinum
dispersion, while the main role of WO3 is to regulate acidity of the
catalyst by introducing BrØnsted acid sites, which were shown to
be essential for 1,3-PD formation.
−
1
nal ones it was 194.8 kcal mol
[25], which means the central
hydroxyl group in glycerol is slightly more easily activated by pro-
tons. With the central hydroxyl group protonized, glycerol can
more easily be dehydrated towards 3-hydroxypropanal, which was
detected in products by GC-MS, but in trace amount, because with
water in large excess, the formation of 3-hydroxypropional by
acid-catalyzed dehydration of glycerol is probably limited by unfa-
vorable equilibrium. 3-Hydroxypropanal could be hydrogenated by
the active hydrogen atom produced over the Pt metal sites of the
catalysts to 1,3-PD [2]. We suppose that the central hydroxyl group
in glycerol might be less accessible to the active acidic sites over
solid catalyst than the two terminal ones due to the steric hin-
drance, which would result in the poor activity and selectivity to
3.3. Influence of Pt loading
Besides acid functionality, metal functionality is important for
a bi-functional catalyst. Table 5 shows the influence of Pt load-
ing on the catalytic performance of glycerol hydrogenolysis over
the Pt/WO /TiO /SiO catalyst. WO /TiO /SiO performed poorly
3
2
2
3
2
2
in glycerol hydrogenolysis. Only 0.7% of glycerol was converted.
,3-PD selectivity was 15.5%, which is much lower than 1,2-PD
selectivity. When 0.5% platinum was loaded on the WO /TiO /SiO
1
3
2
2
1
,3-PD in those organic solvents tested in this paper. Water as sol-
sample, glycerol conversion was 7.4%, and 1,3-PD selectivity was
enhanced to 39.3%. Further increase of Pt loading to 2.0% on the
Pt/WO /TiO /SiO catalyst leads to the linear increase of glycerol
vent might facilitate the transfer of the proton produced on the
BrØnsted acidic sites of WO3 species over the Pt/WO /TiO /SiO
3
2
2
3
2
2
catalyst to react with the central hydroxyl group through the liq-
uid phase, overcoming the steric hindrance of the central hydroxyl
group in glycerol and leading to the high activity as well as the high
selectivity to 1,3-PD.
conversion, but Pt loading of higher than 1.0 wt.% shows a tempered
influence on 1,3-PD selectivity. We can conclude that Pt possessed
hydrogenation activity, and took part in the hydrogenating reaction
intermediate to 1,3-PD. A proper number of Pt metal sites available
was required in the hydrogenolysis of glycerol without varying the
acidity in the Pt/WO /TiO /SiO catalyst.
3
2
2
3.5. Influence of reaction temperature
3
.4. Influence of solvent
The influence of reaction temperature on glycerol conversion
and product selectivity plotted in Fig. 7 shows that glycerol con-
version increased monotonically from 6.2% up to 17.8% along with
Previous research by Chaminand et al. and Ritter showed that
organic solvents favored the formation of 1,3-PD from glycerol
2,3], in our research, however, this tendency was not found to be
the case over the Pt/WO /TiO /SiO catalyst. As shown in Fig. 6,
◦
◦
the increase of the temperature from 160 C to 190 C. The maximal
1,3-PD selectivity of 52.7% was obtained at reaction temperature
[
◦
of 170 C. The highest 1,3-PD to 1,2-PD ratio (7.3) occurred at
3
2
2
◦
compared to water, organic solvents such as ethanol, sulfane and
N-methylimidazole made the Pt/WO /TiO /SiO catalyst perform
160 C. Product distribution showed tiny changes as temperature
◦
◦
increased from 170 C to 180 C, but further increase of tempera-
ture caused a decrease of 1,3-PD selectivity accompanied with an
increase of 1,2-PD selectivity. We suggest that a reaction tempera-
3
2
2
far more poorly. When the above organic solvents were used, both
glycerol conversion and 1,3-PD selectivity were much lower than
those in the case of water as solvent. For example, in the case
of N-methylimidazole as solvent, glycerol conversion and 1,3-PD
selectivity were 2.6% and 15.3%, respectively, less than those in
◦
ture higher than 180 C might allow more terminal hydroxyl groups
of glycerol to be activated and make the hydrogenolysis reaction
less selective to 1,3-PD.

L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
125
Table 5
a
Influence of platinum loading on the hydrogenolysis of glycerol over the Pt/WO3/TiO2/SiO2 catalyst .
Pt loading
Conv. (%)
Selectivity
,3-PD
1,3-PD/1,2-PD
1
1,2-PD
1-PO
2-PO
Others
2.0
1.0
0.5
0.0
15.3
10.1
7.4
50.5
46.7
39.3
15.5
9.2
11.6
12.6
21.8
25.1
25.6
22.0
6.8
8.8
8.7
8.5
0.0
6.4
7.4
17.6
55.9
5.49
4.03
3.12
0.71
0.7
a
Reaction conditions: 10 wt.% glycerol aqueous solution, 40 ml; initial H2 pressure, 5.5 MPa; reaction temperature, 453 K; reaction time, 12 h; catalyst loading, 2 ml.
6
5
4
3
2
1
0
0
0
0
0
0
8
6
4
2
0
the whole, increase of hydrogen pressure led to 1,3-PD selectivity
growth, which was consistent with previous reports [10].
Conversion
,2-PD
,3-PD/1,2-PD
1,3-PD
1-PO
1
1
3.7. Influence of reaction time
The influence of reaction time on glycerol hydrogenolysis was
investigated. As illustrated in Fig. 9, glycerol conversion reached
0.9% with 47.0% selectivity for 1,3-PD after a 6 h reaction. Glyc-
1
erol conversion gradually increased with reaction time. When
the reaction time was up to 24 h, a high glycerol conversion
(33.9%) with a slightly decreased selectivity of 1,3-PD (41.3%) was
achieved, suggesting that the prolonged reaction time led to the
further hydrogenolysis of 1,3-PD. Unlike 1,3-PD, 1,2-PD selectivity
increased with time, because it is easier for 1,3-PD to be further
hydrogenolysed than 1,2-PD [20,21]. The decrease of 1-propanol
selectivity with time was attributed to the hydrogenolysis of 1-
propanol to propane, which was detected in large amounts in gas
products when the reaction time was extended.
4
.5
5.0
5.5
6.0
Pressure (MPa)
Fig. 8. Influence of initial hydrogen pressure on the hydrogenolysis of glycerol over
the Pt/WO3/TiO2/SiO2 catalyst.
4. Conclusion
3.6. Influence of initial hydrogen pressure
Glycerol can be more effectively and selectively converted
to 1,3-PD in water medium over the Pt/WO /TiO /SiO catalyst
3
2
2
Fig. 8 shows the influence of initial hydrogen pressure on glyc-
than over the Pt/WO /TiO2 catalyst. The existence of TiO2 species
3
erol hydrogenolysis over the Pt/WO /TiO /SiO catalyst. Glycerol
3
2
2
improves the dispersion of Pt metal in the metal–acid bi-functional
conversion increases with the increase of hydrogen pressure in the
range of from 4.5 MPa to 5.5 MPa, accompanied with the increase
of selectivity to 1,3-PD and the decrease of selectivity to 1,2-PD.
But further increase of hydrogen pressure leads to a increase of
selectivity to 1,2-PD. Glycerol conversion and 1,3-PD/1,2-PD ratio
attain their maxima of 15.3% and 5.49, respectively, at a hydro-
gen pressure of 5.5 MPa. A similar tendency of glycerol conversion
with hydrogen pressure has been reported over a Ru/CsPW catalyst;
the researchers suggested that reduction of W(VI) may be respon-
sible for the decrease of catalyst activity above 5.5 MPa [26]. On
Pt/WO /TiO /SiO catalyst based on the results from XRD, TEM
3
2
2
characterization. NH -TPD and IR results show that WO3 species
3
regulates the acidity of the Pt/WO /TiO /SiO catalyst by producing
3
2
2
BrØnsted acid sites, which play a key role during 1,3-PD formation.
The nature of the solvent also shows distinct influence on glycerol
conversion; water is more suitable than ethanol, sulfane and NMI
to be used as solvent of the reaction.
Acknowledgement
This work was financially supported by National the Natural
Science Foundation of China (contract no. 20503032).
55
50
45
40
35
30
25
20
15
10
5
10
References
Conversion
,2-PD
1,3-PD/1,2-PD
1,3-PD
1-PO
1
8
[
[
1] G.A. Kraus, Clean: Soil, Air, Water 36 (2008) 648–651.
2] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green
Chem. 6 (2004) 359–361.
[
[
3] S.K. Ritter, Chem. Eng. News 82 (2004) 31–39.
4] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D. Pina, Angew. Chem. Int. Ed.
6
4
6 (2007) 4434–4440.
[5] A. Behr, J. Eilting, K. Irawadi, J.L. Leschinski, F. Lindner, Green Chem. 10 (2008)
3–30.
1
[6] C. Zhou, J.N. Beltramini, Y. Fan, G. Lu, Chem. Soc. Rev. 37 (2008) 527–549.
[7] T.M. Che, US Patent 4,642,394 (1987).
4
[
[
8] E. Drent, W. Jager, US Patent 6,080,898, (2000).
9] T. Kurosaka, H. Maruyama, I. Naribayashi, Y. Sasaki, Catal. Commun. 9 (2008)
1360–1363.
[
[
10] L. Huang, Y. Zhu, H. Zheng, G. Ding, Y. Li, Catal. Lett. 131 (2009) 312–320.
11] A. Shimao, S. Koso, N. Ueda, Y. Shinmi, I. Furikado, K. Tomishige, Chem. Lett. 38
2
5
10
15
20
25
(
2009) 540–541.
12] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B: Environ.
4 (2010) 318–326.
Reaction time (h)
[
9
Fig. 9. Influence of reaction time on the hydrogenolysis of glycerol over the
Pt/WO3/TiO2/SiO2 catalyst.
[13] Y. Nakagawa, Y. Shinmi, S. Koso, K. Tomishige, J. Catal. 272 (2010) 191–194.
[14] S. Minakata, M. Komatsu, Chem. Rev. 109 (2009) 711–724.

1
26
L. Gong et al. / Applied Catalysis A: General 390 (2010) 119–126
[
[
15] Y. Gu, A. Ahmed, Y. Pouilloux, F. JérÔme, J. Barrault, Green Chem. 10 (2008)
64–167.
16] A. Karam, Y. Gu, F. JérÔme, J.P. Douliez, J. Barrault, Chem. Commun. 2223 (2007)
222–2224.
[21] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 240 (2006)
213–221.
[22] M. Balaraju, V. Rekha, P.S. Sai Prasad, B.L.A. Prabhavathi Devi, R.B.N. Prasad, N.
Lingaiah, Appl. Catal. A: Gen. 354 (2009) 82–87.
1
2
[
[
[
17] M.Y. Kim, Y.S. You, H.S. Han, G. Seo, Catal. Lett. 120 (2008) 40–47.
18] Z. Yuan, P. Wu, J. Gao, X. Lu, Z. Hou, X. Zheng, Catal. Lett. 130 (2009) 261–265.
19] Y. Kusunoki, T. Miyazawa, K. Kunimori, K. Tomishige, Catal. Commun. 6 (2005)
[23] E.P. Parry, J. Catal. 109 (1963) 371–379.
[24] M. Kobayashi, A. Morita, M. Ikeda, Appl. Catal. B: Environ. 71 (2007) 94–100.
[25] M.R. Nimlos, S.J. Blanksby, X. Qian, M.E. Himmel, D.K. Johnson, J. Phys. Chem. A
110 (2006) 6145–6156.
[26] A. Alhanash, E.F. Kozhevnikova, I.V. Kozhevnikov, Catal. Lett. 120 (2008)
307–311.
6
45.
20] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal. A: Gen. 318 (2007)
44–251.
[
2