Direct hydrogenolysis of glycerol into 1,3-propanediol over rhenium-modified iridium catalyst

Article abstract of DOI:10.1016/j.jcat.2010.04.009

Rhenium-oxide-modified supported iridium nanoparticles on silica catalyzes direct hydrogenolysis of glycerol to 1,3-propanediol in an aqueous media. The selectivity to 1,3-propanediol at an initial stage reaches 67 ± 3%. The yield of 1,3-propanediol reaches 38% at 81% conversion of glycerol. The characterization of catalyst and the reactivity of alcohols suggest that 1,3-propanediol is produced by the attack of active hydrogen species on iridium metal to 1-glyceride species formed on the oxidized rhenium cluster.

Full text of DOI:10.1016/j.jcat.2010.04.009

Journal of Catalysis 272 (2010) 191–194
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Priority Communication
Direct hydrogenolysis of glycerol into 1,3-propanediol over rhenium-modified
iridium catalyst
Yoshinao Nakagawa a,b,1, Yasunori Shinmi a, Shuichi Koso a, Keiichi Tomishige a,b,
*
a Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1, Tennodai, Tsukuba, Ibaraki 305-8573, Japan
b International Center for Materials Nanoarchitectonics Satellite (MANA), National Institute of Materials Science (NIMS) and University of Tsukuba, 1-1-1, Tennodai, Tsukuba,
Ibaraki 305-8573, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Rhenium-oxide-modified supported iridium nanoparticles on silica catalyzes direct hydrogenolysis of glyc-
erol to 1,3-propanediol in an aqueous media. The selectivity to 1,3-propanediol at an initial stage reaches
67 3%. The yield of 1,3-propanediol reaches 38% at 81% conversion of glycerol. The characterization of cat-
alyst and the reactivity of alcohols suggest that 1,3-propanediol is produced by the attack of active hydrogen
species on iridium metal to 1-glyceride species formed on the oxidized rhenium cluster.
Ó 2010 Elsevier Inc. All rights reserved.
Received 23 February 2010
Revised 29 March 2010
Accepted 14 April 2010
Available online 14 May 2010
Keywords:
Glycerol
Hydrogenolysis
Biorefinery
Iridium
Rhenium oxide
1. Introduction
low as 26.8% [11]. One example of co-catalyst is ReOx for Rh/
SiO2. We have very recently reported that the co-catalyst much en-
For the sustainable development of our society, much attention
has been paid to biorefinery, where renewable biomass-related
raw material is converted to valuable chemicals and fuel [1,2].
Glycerol is one of the most important building blocks in the
biorefinery feedstock, since enormous amount of glycerol is
expected to be produced by the biodiesel production process as
well as traditional routes such as soap manufacture [3,4]. Among
the chemicals that can be derived from glycerol, 1,3-propanediol
is a very promising target because of the high cost of conventional
processes of 1,3-propanediol production and the large-scale pro-
duction of polyester and polyurethane resin from 1,3-propanediol
[4,5]. Therefore, the development of an efficient conversion process
of glycerol into 1,3-propanediol will make the biodiesel process
more profitable. However, selective hydrogenolysis of glycerol into
1,3-propanediol is not easy. Although glycerol hydrogenolysis has
been heavily studied in recent years, the main products are less
valuable such as 1,2-propanediol and propanols in most cases
[4]. Significant amount of 1,3-propanediol is formed only in the
cases where oxide of group 6 or 7 metal is added as a co-catalyst
[6–12], and the maximum reported yield of 1,3-propanediol is as
hances the glycerol conversion (22 times) and 1,3-propanediol
selectivity [12]. However, the selectivity to 1,3-propanediol is still
lower than 30% even in the initial stage, and its maximum yield
was 11%. Iridium lies just under rhodium in the periodic table. It
has been known that iridium complexes, as well as rhodium coun-
terparts, can activate H2 to give hydride complexes [13]. However,
metallic iridium has been received less attention for hydrogenation
or hydrogenolysis catalysis because of the lower activity than other
noble metals, and little is known of the modification effect of irid-
ium catalyst with metal oxide [14,15]. In this study, we report that
the ReOx-modified iridium nanoparticles catalyzes the selective
hydrogenolysis of glycerol into 1,3-propanediol in water. The mod-
ification effect of Ir with ReOx is much more remarkable than the
case of Rh in terms of the initial selectivity to 1,3-propanediol
(68%) and maximum yield (38%).
2. Experimental
2.1. Preparation of catalyst
An Ir/SiO2 catalyst was prepared by impregnating SiO2 with an
aqueous solution of H2IrCl6 (Furuya Metal Co., Ltd.). The SiO2 (G-6,
BET surface area 535 m2/g, <100 mesh powder) was supplied by
Fuji Silysia Chemical Ltd. After impregnation, the catalyst was
dried at 383 K for 12 h. The loading amount of Ir was 4 wt.%.
Ir–ReOx/SiO2 catalysts were prepared by impregnating Ir/SiO2 after
* Corresponding author. Present address: School of Engineering, Tohoku Univer-
sity, 6-6-07 Aoba, Aramaki, Aoba-ku, Sendai, Miyagi 980-8579, Japan. Fax: + 81 22
795 7214.
E-mail address: tomi@tulip.sannet.ne.jp (K. Tomishige).
Present address: School of Engineering, Tohoku University, 6-6-07 Aoba, Aramaki,
1
Aoba-ku, Sendai, Miyagi 980-8579, Japan.
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.04.009

192
Y. Nakagawa et al. / Journal of Catalysis 272 (2010) 191–194
the drying procedure with aqueous solution of NH4ReO4 (Soekawa
Chemical Co., Ltd.). ReOx/SiO2 was prepared by impregnation of
SiO2 with the same precursor as that for Ir–ReOx/SiO2. All catalysts
were calcined in air at 773 K for 3 h after drying (at 383 K for 12 h).
The ratio of Re to Ir was optimized to maximize the selectivity to
1,3-propanediol in the glycerol hydrogenolysis and was deter-
mined to be Re/Ir = 1 (molar basis).
search Institute (JASRI; Proposal No. 2008B1235). The storage ring
was operated at 8 GeV. A Si (1 1 1) single crystal was used to obtain
a monochromatic X-ray beam. Two ion chambers for I0 and I were
filled with 85% N2 + 15% Ar and 100% Ar, respectively, for Re and Ir
L3-edge measurement. The data were collected in a transmission
mode. The calcined catalyst was reduced with 8 MPa H2 in water
at 473 K for 1 h. After cooling, the wet catalyst powder was trans-
ferred to a measurement cell (thickness 2 mm) using a glove box.
For curve fitting analysis, the empirical phase shift and amplitude
functions for the Re–O, Re–Re and Ir–Ir bonds were extracted from
data for NH4ReO4, Re powder and Ir powder, respectively. Theoret-
ical functions for the Re–Ir bond were calculated using the FEFF8.2
program [16], and this bond is denoted as Re–Ir (or -Re) since it is
difficult to distinguish between Ir and Re as a scattering atom.
Analyses of EXAFS data were performed using a computer program
(REX2000, ver. 2.5.9; Rigaku Corp.).
2.2. Catalytic test
Catalytic tests were performed in a 190-ml stainless steel auto-
clave with an inserted glass vessel. Water, a spinner and an appro-
priate amount of catalyst were placed into the autoclave. After
sealing the reactor, the air content was purged by flushing thrice
with 1 MPa hydrogen. The reactor was then heated to 473 K and
pressurized to 8 MPa. After 1 h, the reactor was cooled down and
the H2 inside the reactor was removed. This procedure is needed
because the reduction temperature is higher than the reaction
temperature. When the pretreatment was conducted at 393 K,
the reduction in Ir was incomplete (confirmed by EXAFS) and the
catalytic performance was lower. An appropriate amount of glyc-
erol and sulfuric acid was quickly added to the reaction mixture.
In this case, it can be possible to avoid direct contact of the reduced
catalyst with air because the catalyst is covered with water. After
sealing the reactor, the air content was purged by flushing thrice
with 1 MPa hydrogen. The reactor was then heated to the reaction
temperature (393 K) and pressurized to 8 MPa. After an appropri-
ate reaction time, the reactor was cooled down and the gas inside
the reactor was collected in a gas bag. The products in both liquid
and gas phases were analyzed using GC (Shimadzu GC-2014)
equipped with FID and GC–MS (Shimadzu QP5050). The used
catalyst was separated from the reaction mixture by centrifuga-
tion, washed with excess water, dried in air and calcined at
773 K for 3 h. A slight loss (<10% in weight) was observed during
the recovery process and was compensated with fresh catalyst in
each reuse experiment. The yield and selectivity were calculated
on the carbon basis and defined as follows:
3. Results and discussion
3.1. Catalytic performance
The time course of the glycerol hydrogenolysis catalyzed by
Ir–ReOx/SiO2 (Re/Ir = 1) is shown in Fig. 1. The initial selectivity
to 1,3-propanediol was 67 3%. The maximum yield of 1,3-pro-
panediol was 38% at 36 h, which value was much higher than those
reported for the other catalytic glycerol hydrogenolysis systems.
The ICP analysis of the reaction solution after the filtration of the
catalyst showed no appreciable leaching of Ir or Re (<0.3%). The
Ir–ReOx/SiO2 catalyst could be reused at least thrice without the
change of the reaction rate and selectivities. In contrast, monome-
tallic Ir/SiO2 showed very low catalytic activity (ca. 2 orders lower
than Ir–ReOx/SiO2) and selectivity to 1,3-propanediol (<10% at 0.3%
conversion; the main product is 1,2-propanediol). It has been
known that the activity of ReOx/SiO2 was very low [12]. Therefore,
it is concluded that the synergy between Ir and Re enabled the
hydrogenolysis of glycerol to 1,3-propanediol. It should be noted
that the additive effect of Re over Ir catalysts on the selectivity to
ðmol of the productÞ ꢂ ðnumber of carbon atoms in the productÞ
ðmol of glycerol chargedÞ ꢂ 3
Yield ð%Þ ¼
ꢂ 100
ðmol of the productÞ ꢂ ðnumber of carbon atoms in the productÞ
ðmol of glycerol consumedÞ ꢂ 3
Selectivity ð%Þ ¼
ꢂ 100
2.3. Catalyst characterization
1,3-propanediol and its maximum yield was much more drastic
than that over Rh catalysts.
X-ray diffraction (XRD) patterns were recorded by a diffractom-
eter (Philips X’pert). Transmission electron microscope (TEM)
images were taken with JEOL JEM 2010. The samples were dis-
persed in ethanol with supersonic waves and placed on Cu grids
under air atmosphere. Average particle size was calculated by
The present hydrogenolysis system was applied to other alco-
hols (Table 1). The reactivity decreased in the following order:
1,2-propanediol P glycerol ꢀ 1,3-propanediol ꢁ 2,3-butanediol ꢁ
propanols. In the hydrogenolysis of 1,2-propanediol, the secondary
OH group was selectively removed to give 1-propanol as a main
product. These trends show that the OH group adjacent to a pri-
mary OH group has much higher reactivity than the other OH
groups.
P
P
nid3= nid2 (di: particle size, ni: number of particles with di).
i
i
The amount of CO chemisorption was measured in a high-vacuum
system using a volumetric method under about 1.1 kPa CO at room
temperature after the pretreatment in H2 at 473 K for 1 h. Temper-
ature-programmed reduction (TPR) was carried out in a fix bed
reactor equipped with a thermal conductivity detector using 5%
H2/Ar in the temperature range from room temperature to
1123 K (heating rate 10 K/min).
3.2. Characterization and reaction mechanism
In the temperature-programmed reduction (TPR) profile (Fig. S1
in Supplementary Information) of Ir/SiO2, H2 consumption signal
with a peak at 510 K was observed between 380 and 600 K. The to-
tal amount of H2 consumption was 2.0 equiv. to Ir, which agreed
The EXAFS spectra were measured at the BL01B1 station at
SPring-8 with the approval of the Japan Synchrotron Radiation Re-

Y. Nakagawa et al. / Journal of Catalysis 272 (2010) 191–194
193
Fig. 1. Time course of the hydrogenolysis of glycerol over Ir–ReOx/SiO2 (Re/Ir = 1).
Conditions: catalyst (150 mg, 31 lmol Ir), glycerol (4 g), water (1 g), sulfuric acid
(H+/Ir = 1) and H2 (8 MPa) at 393 K.
Scheme 1. Possible reaction mechanism of glycerol hydrogenolysis catalyzed by Ir–
ReOx/SiO2.
Table 1
Hydrogenolysis of alcohols over Ir–ReOx/SiO2 (Re/Ir = 1).
ide. The larger formation constants of metal alkoxides of primary
alcohols than of secondary alcohols have been reported for nio-
bium- or vanadium-containing tungsten oxide clusters and ex-
plained by the steric crowding between the cluster framework
and the groups bonded with the C–OH carbon of secondary alco-
hols [19,20]. Next, hydrogen activated on the Ir metal attacks the
2-position of the 2,3-dihydroxypropoxide to produce 3-hydroxy-
propoxide. The hydrolysis of 3-hydroxypropoxide releases 1,3-pro-
panediol. The growth of oxidized metal cluster large enough to
induce steric effect for the preferential terminal alkoxide formation
may be the key to high 1,3-propanediol selectivity. In this mecha-
nism, only substrates with OH groups at both 1- and 2-positions
such as 1,2-propanediol and glycerol are reactive. The possible
mechanisms of the hydrogenolysis of the other positions include
the hydrogenolysis of alkoxide, the formation of 1,3-dihydroxyiso-
propoxide intermediate [12] and the indirect hydrogenolysis com-
posed of acid-catalyzed dehydration and successive hydrogenation
[21]. These mechanisms can lead the side-reactions of glycerol
hydrogenolysis such as overhydrogenolysis and 1,2-propanediol
formation.
Entry Substrate
(mmol)
Converted Products (selectivity (%))
(mmol)
1
Glycerol (43)
27
1,3-Propanediol (49), 1,2-
propanediol (10), 1-propanol
(33), 2-propanol (8)
1-Propanol (85), 2-propanol
(10), propane (5)
2
1,2-Propanediol (53) 38
1,3-Propanediol (53) 12
2,3-Butanediol (44)
1-Propanol (67)
2-Propanol (67)
3
4
5
6
1-Propanol (>99)
8
10
14
2-Butanol (81), n-butane (19)
Propane (>99)
Propane (>99)
Reaction conditions: catalyst (150 mg, 31
lmol Ir), alcohol (4 g), water (16 g), sul-
furic acid (H+/Ir = 1) and H2 (8 MPa) at 393 K for 24 h.
with the stoichiometry of the reduction of IrIV to Ir0. In the case of
ReOx/SiO2, where the initial valence of Re was 7, H2 consumption
occurred between 500 and 800 K. The total amount of H2 consump-
tion was 2.2 equiv. to Re, indicating that low-valent rhenium oxide
was formed. In the case of Ir–ReOx/SiO2, intense H2 consumption
signal with a peak at 450 K was observed between 380 and
480 K. The consumption amount was 3.8 equiv. to Ir, indicating
that Ir and Re were simultaneously reduced. The XRD pattern (2h
range: 10–70°; Fig. S2) of Ir–ReOx/SiO2 after the catalytic use
showed only one broad peak at 2h = 41° assignable to Ir metal in
addition to the peak of SiO2 (2h = 22°). The crystal size of Ir was
estimated to be 2.3 nm from the linewidth of the peak. The TEM
observation also detected particles of 2.0 0.2 nm size (Fig. S3).
The curve fitting of Ir L3-edge EXAFS of Ir–ReOx/SiO2 reduced at
473 K detected Ir–Ir (or -Re) bond with length (R) of 2.68 Å and
coordination number (CN) of 9.3 (Table S1 and Fig. S4). The Re
L3-edge EXAFS detected Re–O bond (R = 2.03 Å, CN = 1.2) as well
as Re–Ir (or -Re) bond (R = 2.69 Å, CN = 5.5) (Table S1 and
Fig. S5). These data showed that iridium and rhenium in the
Ir–ReOx/SiO2 catalyst were reduced to the metallic state and
low-valent metal oxide, respectively. The amount of CO adsorbed
on Ir–ReOx/SiO2 after the reduction at 473 K was 0.07 equiv. to Ir,
while 48% of Ir is calculated to be located on the surface of
2.3-nm Ir crystal [17]. This discrepancy and the single-peak TPR
profile suggest that the oxidized Re clusters were attached to the
Ir metal particles [12,18].
Acknowledgments
This work was in part supported by the Industrial Technology
Research Grant Program (No. 08B40001c) of New Energy and
Industrial Technology Development Organization (NEDO) of Japan
and World Premier International Research Center (WPI) Initiative
on Materials Nanoarchitectonics, MEXT. We appreciate Dr. Tokushi
Kizuka for TEM observation of the catalyst.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcat.2010.04.009.
References
[1] A.J. Ragauskas, C.K. Williams, B.H. Davison, G. Britovsek, J. Cairney, C.A. Eckert,
W.J. Frederick Jr., J.P. Hallett, D.J. Leak, C.L. Liotta, J.R. Mielenz, R. Murphy, R.
Templer, T. Tschaplinski, Science 311 (2006) 484.
[2] S. Fernando, S. Adhikari, C. Chandrapal, N. Murani, Energy Fuels 20 (2006)
1727.
On the basis of all results, we propose a possible reaction mech-
anism in Scheme 1. First, glycerol was adsorbed on the surface of
ReOx cluster at the terminal position to form 2,3-dihydroxypropox-
[3] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D. Pina, Angew. Chem. Int. Ed.
4434 (2007) 4434.

194
Y. Nakagawa et al. / Journal of Catalysis 272 (2010) 191–194
[4] C.-H. Zhou, J.N. Beltramini, Y.-X. Fan, G.Q. Lu, Chem. Soc. Rev. 37 (2008) 527.
[13] D. Milstein, Acc. Chem. Res. 17 (1984) 221.
[5] M. Schlaf, Dalton Trans. (2009) 4645.
[6] T. Kurosaka, H. Maruyama, I. Naribayashi, Y. Sasaki, Catal. Commun. 9 (2008)
1360.
[7] Y. Yoshikawa, M. Tamura, M. Takahashi, JP Patent 2007-326849, 2007.
[8] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green
Chem. 6 (2004) 359.
[14] P. Reyes, M.C. Aguirre, I. Melian-Cabrera, M. Lopez Granados, J.L.G. Fierro, J.
Catal. 208 (2002) 229.
[15] S. Subramanian, J.A. Schwarz, J. Catal. 127 (1991) 201.
[16] A.L. Ankudinov, B. Ravel, J.J. Rehr, S.D. Conradson, Phys. Rev. B 58 (1998) 7565.
[17] F. Locatelli, B. Didillon, D. Uzio, G. Niccolai, J.P. Candy, J.M. Basset, J. Catal. 193
(2000) 154.
[9] T.M. Che, US Patent 4 642 394, 1987.
[10] L. Ma, D. He, Catal. Today 149 (2010) 148.
[18] S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori, K. Tomishige, Chem.
Commun. (2009) 2035.
[11] L. Huang, Y. Zhu, H. Zheng, G. Ding, Y. Li, Catal. Lett. 131 (2009) 312.
[12] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B:
Environ. 318 (2010) 318.
[19] Y. Nakagawa, K. Uehara, N. Mizuno, Inorg. Chem. 44 (2005) 14.
[20] V.W. Day, W.G. Klemperer, C. Schwartz, J. Am. Chem. Soc. 109 (1987) 6030.
[21] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 240 (2006) 213.