Hydrogenolysis of CO bond over Re-modified Ir catalyst in alkane solvent

Article abstract of DOI:10.1016/j.apcata.2013.09.021

Hydrogenolysis of alcohols was carried out using n-heptane solvent and IrReO_x/SiO₂ catalyst, which has been known to be active in water solvent. Hydrogenolysis of trans-1,2-cyclohexanediol proceeded more smoothly in n-heptane than in water. The maximum yield of cyclohexanol was 74%, and at longer reaction time cyclohexane was selectively formed (>80% yield). Stronger adsorption of substrate on catalyst surface in n-heptane than in water is one of factors in obtaining the good yields. Alkane solvent was also advantageous to water solvent in hydrogenolysis of mono-alcohols. The reaction route via acid-catalyzed dehydration and subsequent hydrogenation is enhanced in alkane solvent. On the other hand, the "direct" hydrogenolysis driven by the hydride-like species is suppressed in alkane solvent, leading lower activity in n-heptane for hydrogenolysis of tetrahydrofurfuryl alcohol or 1,2-hexanediol, which smoothly react over IrReO_x/SiO₂ catalyst in water.

Full text of DOI:10.1016/j.apcata.2013.09.021

Applied Catalysis A: General 468 (2013) 418–425
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Hydrogenolysis of C O bond over Re-modified Ir catalyst
in alkane solvent
∗
Yoshinao Nakagawa , Kazuma Mori, Kaiyou Chen, Yasushi Amada,
∗∗
Masazumi Tamura, Keiichi Tomishige
Department of Applied Chemistry, School of Engineering, Tohoku University, 6-6-07 Aoba, Aramaki, Aoba-ku, Sendai, 980-8579, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 June 2013
Received in revised form 2 August 2013
Accepted 11 September 2013
Available online 21 September 2013
Hydrogenolysis of alcohols was carried out using n-heptane solvent and Ir ReO /SiO catalyst, which has
x
2
been known to be active in water solvent. Hydrogenolysis of trans-1,2-cyclohexanediol proceeded more
smoothly in n-heptane than in water. The maximum yield of cyclohexanol was 74%, and at longer reac-
tion time cyclohexane was selectively formed (>80% yield). Stronger adsorption of substrate on catalyst
surface in n-heptane than in water is one of factors in obtaining the good yields. Alkane solvent was also
advantageous to water solvent in hydrogenolysis of mono-alcohols. The reaction route via acid-catalyzed
dehydration and subsequent hydrogenation is enhanced in alkane solvent. On the other hand, the “direct”
hydrogenolysis driven by the hydride-like species is suppressed in alkane solvent, leading lower activity
in n-heptane for hydrogenolysis of tetrahydrofurfuryl alcohol or 1,2-hexanediol, which smoothly react
over Ir ReOx/SiO2 catalyst in water.
Keywords:
Hydrogenolysis
Alcohol
Solvent effect
Iridium
Rhenium
© 2013 Elsevier B.V. All rights reserved.
1
. Introduction
The importance of catalytic C O hydrogenolysis (dissociation
In a practical view, water has a large heat of vaporization. The use
of nonpolar solvent such as alkanes can overcome these problems.
In addition, it has been reported that adsorption of secondary alco-
hols on Ir ReOx/SiO2 is weak in water, which can be explained
by the competitive adsorption of water [26]. The use of alkanes
with weak adsorption ability may promote the adsorption of sub-
strates. In this paper, we report the C O hydrogenolysis catalysis
of C O bond plus addition of H ) is rapidly growing, because con-
2
version of oxygen-rich biomass to valuable fuels and chemicals has
gained much attention in recent years [1–7]. Recently various het-
erogeneous catalysts have been reported for C O hydrogenolysis
of various substrates, such as Cu-based catalysts for conversion
of glycerol to 1,2-propanediol [8–10], Pt W catalysts for con-
version of glycerol to 1,3-propanediol [11–14], Ru catalysts for
conversion of furfuryl alcohol to 1,2-pentanediol [15], and mod-
ified Rh catalysts for conversion of tetrahydrofurfuryl alcohol
to 1,5-pentanediol [16–21]. We have very recently discovered
that bimetallic Ir ReOx/SiO2 catalyst can promote various C O
hydrogenolysis reactions with unique selectivities such as glyc-
erol to 1,3-propanediol [22–24], erythritol to butanediols [25],
cyclic ethers to ring-opening reaction products [26], and sugars or
sugar alcohols to n-alkanes [27]. The very low activity in the C C
hydrogenolysis (cracking) is one important feature of the catalysis
of Ir ReOx/SiO in alkane solvent, and discuss the difference from
2
that in water solvent.
2. Experimental
2.1. Materials
Ir ReOx/SiO2 catalyst was prepared by sequential impregna-
tion method using Fuji Silysia G-6 SiO2 support with size of less
than 100 mesh and aqueous H2IrCl6 (Furuya Metals) and NH4ReO4
(Soekawa Chemical) precursors, as reported earlier [22,23]. The
loading amount of Ir was 4 wt%. The molar ratio of Re to Ir was
typically 2. Ir/SiO2 (4 wt% Ir) and ReOx/SiO2 (4 wt% Re) catalysts
were prepared by impregnation method using the same support
and precursors as Ir ReOx/SiO2. Rh-ReOx/SiO2 catalyst was pre-
pared by sequential impregnation method using Fuji Silysia G-6
silica and aqueous solutions of RhCl ·3H O (Soekawa Chemical)
of Ir ReOx/SiO . However, the reactions catalyzed by Ir ReOx/SiO2
2
have been always conducted in water solvent. Water is a green
solvent and can dissolve many multi-functionalized molecules;
however, substrates with longer chains are less soluble in water.
3
2
and NH ReO [16,19]. These catalysts were calcined in air at 773 K
4
4
for 3 h after drying at 383 K for 12 h. Commercial Rh/C, Pd/C, Pt/C
and Ru/C catalysts with 5 wt% metal loading were purchased from
Wako. All organic substrates and solvent (n-heptane) were com-
mercially available and used as received.
∗
Corresponding author. Tel.: +81 22 795 7215; fax: +81 22 795 7215.
Corresponding author. Tel.: +81 22 795 7214; fax: +81 22 795 7214.
E-mail addresses: yoshinao@erec.che.tohoku.ac.jp (Y. Nakagawa),
∗
∗
tomi@erec.che.tohoku.ac.jp, tomi@tulip.sannet.ne.jp (K. Tomishige).
0
926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.09.021

Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
419
2.2. Activity tests
Activity tests as well as reduction of catalysts were performed in
a 190 ml stainless-steel autoclave with an inserted glass vessel. The
catalyst (25–100 mg), a spinner and reaction solvent (n-heptane
or water; typically 10 ml) were put into an autoclave and heated
at 473 K under 8 MPa H2 (Nippon Peroxide Co., Ltd.) for 1 h for
the reduction pretreatment. After the pretreatment, the autoclave
was cooled and H2 was removed. The substrate was quickly put
into the autoclave. After sealing the reactor, the air was purged
by flushing three times with 1 MPa N2 (Tanuma Sanso). The auto-
clave was heated to the reaction temperature (393 K) under N2.
The temperature was monitored by using a thermocouple inserted
into the autoclave. After reaching the reaction temperature, H was
2
introduced to start the reaction. Typical H2 partial pressure was
6
.8 MPa with the total pressure of 8 MPa. Commercial catalysts and
Rh ReOx/SiO catalyst were used without reduction pretreatment
2
since they can be readily reduced with H below the reaction tem-
2
perature (393 K). After an appropriate reaction time, the reactor
was cooled in water, and the gases were collected in a gas bag.
The products in both gas and liquid phases were analyzed by GC
(
Shimadzu GC-2014) equipped with an InertCap 5MS/Sil or TC-
WAX capillary column and FID. Products were also identified using
GC–MS (QP5050, Shimadzu).
2.3. Catalyst characterization
The Ir ReOx/SiO catalysts after reduction in n-heptane or after
2
use for hydrogenolysis reaction were characterized. The condi-
tions of the catalytic reaction were catalyst 100 mg, n-heptane
1
0 ml, trans-1,2-cyclohexanediol 0.5 g, H partial pressure 6.8 MPa,
2
reaction temperature 393 K and reaction time 4 h. The catalyst
samples for XRD and TEM were separated from the solvent by
centrifugation, washed with n-heptane, and dried. XRD patterns
were recorded with a Rigaku Ultima IV diffractometer. The cata-
lyst sample for extended X-ray absorption fine structure (EXAFS)
was transferred from the reactor together with a small amount of
solvent to a measurement cell in a glove bag filled with nitrogen.
The EXAFS spectra were measured at the BL01B1 station at SPring-
Fig. 1. Hydrogenolysis of trans-1,2-cyclohexanediol over Ir ReOx/SiO2 in n-
heptane (A) or water (B). Reaction conditions: Ir ReOx/SiO2 (Ir 4 wt%, Re/Ir = 2)
100 mg, trans-1,2-cyclohexanediol 0.5 g, n-heptane or water 10 ml, H2 6.8 MPa,
8
with the approval of the Japan Synchrotron Radiation Research
Institute (JASRI; Proposal No. 2012A1089). The thickness of the cell
filled with the powder was 2 mm to give an edge jump of 0.9 for Re
L3-edge measurement. The EXAFS data were collected in a trans-
mission mode. The oscillation was first extracted from the EXAFS
data using a spline smoothing method [28]. Fourier transforma-
393 K.
and water are shown in Fig. 1. The main products were cyclohex-
anol and cyclohexane. Small amount of cis-1,2-cyclohexanediol,
which can be produced via dehydrogenation and hydrogenation
3
tion of the k -weighted EXAFS oscillation from the k space to the
r space was performed to obtain a radial distribution function. The
inversely Fourier filtered data were analyzed using a usual curve fit-
ting method [29,30]. For curve fitting analysis, the empirical phase
shift and amplitude function for the Re O bond were extracted
from data for NH ReO . Theoretical functions for the Re Ir bond
[
26], was observed at shorter reaction time in both solvents. In
n-heptane (Fig. 1A), the substrate was smoothly converted, and
almost complete conversion was achieved at 24 h. At that time
cis-1,2-cyclohexanediol was also converted almost completely.
The overhydrogenolysis of cyclohexanol to cyclohexane rapidly
proceeded after 16 h. The highest yield of cyclohexanol was 74%,
which was obtained at 16 h. High yield of cyclohexane (>80%) was
obtained at >24 h. On the other hand, in water (Fig. 1B) the reac-
tion rate was much decreased when the conversion approached
4
4
were calculated using the FEFF8.2 program [31]. The Re Ir bonds
are represented by the Re Ir (or Re) in the curve fitting results.
This is because it is very 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.).
8
0%. The selectivity to cyclohexanol was no more than 75% and
gradually decreased after the conversion reached 70% (16 h). The
maximum yield of cyclohexanol was 54% (20 h), which is much
lower than that obtained in n-heptane.
3
. Results and discussion
3
.1. Hydrogenolysis of trans-1,2-cyclohexanediol in alkane
The lower conversion in water, especially at longer reaction
time, is connected to the low adsorption ability of trans-1,2-
cyclohexanediol on the catalyst surface. In the previous paper,
we reported that the reaction order with respect to trans-1,2-
cyclohexanediol concentration in water was 0.7 [26]. The much
shorter reaction time for complete conversion in n-heptane can be
solvent
First, we carried out the hydrogenolysis of trans-1,2-
cyclohexanediol with two secondary OH groups. The time courses
of the reaction over Ir ReOx/SiO2 (Re/Ir = 2) catalyst in n-heptane

4
20
Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
Fig. 3. Model structure of reduced Ir ReOx/SiO2 catalyst.
metal catalysts without addition of Re. These data showed that
Fig. 2. Effect of substrate concentration on hydrogenolysis of trans-1,2-
cyclohexanediol over Ir ReOx/SiO2 in n-heptane. Reaction conditions:
Ir ReOx/SiO2 (Ir 4 wt%, Re/Ir = 2) 50 mg, trans-1,2-cyclohexanediol 0.5 g, n-heptane
Ir ReO /SiO2 (Re/Ir = 2) is a good catalyst for the hydrogenolysis
of trans-1,2-cyclohexanediol to cyclohexanol or cyclohexane.
x
We have reported that addition of acid (sulfuric acid or solid acid
such as H-ZSM-5) to Ir ReOx/SiO2 enhances the activity and the
resistance to leaching in hydrogenolysis of glycerol in water [23,24].
However, when addition of H-ZSM-5 was applied to hydrogenolysis
of trans-1,2-cyclohexanediol in n-heptane (Entry 12), the conver-
sion was very low and significant amount of by-products was
formed including smaller alkanes. Addition of strong acid is unsuit-
able for this reaction in alkane solvent. Water-soluble metal species
is generally insoluble to alkanes, and thus leaching would not be a
significant problem in alkane solvent.
5
–15 ml, H2 6.8 MPa, 393 K, 2 h.
explained by the difference of adsorption ability. The dependence
of the reaction on trans-1,2-cyclohexanediol concentration in n-
heptane is shown in Fig. 2. The reaction rate was almost constant
in different trans-1,2-cyclohexanediol concentration. The reaction
order was calculated to be −0.25, suggesting the strong adsorption.
The strong adsorption of trans-1,2-cyclohexanediol is also shown
by the drastic change of the rate of overhydrogenolysis of cyclohex-
anol to cyclohexane at 20 h in Fig. 1A. The fast overhydrogenolysis
after 20 h means that the intrinsic hydrogenolysis rate of cyclohex-
anol to cyclohexane is larger than that of trans-1,2-cyclohexanediol.
The presence of trans-1,2-cyclohexanediol strongly suppressed the
hydrogenolysis of cyclohexanol, which indicates the strong adsorp-
tion of trans-1,2-cyclohexanediol.
3.2. Catalyst structure in n-heptane
In previous papers, we have conducted the characterization
of Ir ReOx/SiO2 catalyst in dry conditions using various tech-
niques such as TPR, XRD, TEM, CO adsorption, XANES and EXAFS
[22,23,32]. The catalyst after reduction contains Ir metal particles
modified with partially-oxidized ReOx clusters, as shown in Fig. 3.
We also showed that the structures after reduction or catalytic
use in water solvent are the same as that formed by reduction in
dry conditions [23,32]. We checked the structure of Ir ReOx/SiO2
(Re/Ir = 2) catalyst reduced or used in n-heptane. From XRD patterns
(Fig. 4) and TEM images (Fig. 5), Ir metal particles with the size of
around 2 nm were observed in both samples after reduction and
Various catalysts were tested for hydrogenolysis of trans-1,2-
cyclohexanediol, and the results are summarized in Table 1.
Monometallic Ir/SiO2 (Entry 1) and ReOx/SiO2 (Entry 6) cata-
lysts showed no hydrogenolysis activity. The activity increased
with increasing Re loading in Ir ReOx/SiO catalysts (Entries 2–5).
2
The effect of Re is more evident at low Re/Ir ratio. When Re/Ir
ratio is greater than 1, further increase of Re/Ir has no evident
effect. These trends are the same as reported for the same cata-
lysts in the hydrogenolysis of glycerol [23] and tetrahydrofurfuryl
alcohol [26] in water solvent. Commercial noble metal catalysts
and Rh-ReOx/SiO2 catalyst [16–21] showed lower hydrogenolysis
activity than Ir ReOx/SiO2 (Re/Ir = 2) (Entries 7–11). Isomeriza-
tion to cis-1,2-cyclohexanediol was the main reaction over noble
after catalytic use. Re L -edge EXAFS analysis showed that Re O
3
and Re Re (or Ir) bonds are present in both samples (Fig. 6 and
Table 2). These patterns were essentially the same as those of sam-
ples prepared in water [23], indicating the same structure for all
samples.
Table 1
a
Hydrogenolysis of trans-1,2-cyclohexanediol over various catalysts in n-heptane.
Entry
Catalyst
Conv./%
Selectivity/%
Cyclohexanol
cis-1,2-Cyclohexanediol
Cyclohexane
1
2
3
4
5
6
7
8
9
Ir/SiO2
1.3
20
23
35
39
0.0
5.9
18
4.3
0.7
2.3
4.2
<5
76
78
75
79
–
67
2
13
<5
42
20
>95
18
15
17
14
–
33
98
87
>95
58
19
<5
6
7
9
7
Ir ReOx/SiO2 (Re/Ir = 0.25)
Ir ReOx/SiO2 (Re/Ir = 0.5)
Ir ReOx/SiO2 (Re/Ir = 1)
Ir ReOx/SiO2 (Re/Ir = 2)
ReOx/SiO2
Rh-ReOx/SiO2 (Re/Rh = 0.5)
Ru/C
Rh/C
–
<1
<1
<1
<5
<1
22
1
1
1
0
1
2
Pd/C
Pt/C
Ir ReOx/SiO2 (Re/Ir = 2) + H-ZSM-5b
a
Reaction conditions: catalyst 100 mg (active metal 4 and 5 wt% for SiO2- and C-supported catalysts, respectively), trans-1,2-cyclohexanediol 0.5 g, n-heptane 10 ml, H2
6
.8 MPa, 393 K, 4 h.
b
H-ZSM-5 (100 mg). Other products include cyclohexanone and smaller alkanes.

Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
421
substrates, and the results were summarized in Table 3. Mono-
alcohols were reacted to afford the corresponding alkanes without
C-C hydrogenolysis (cracking) or isomerization of alkane chain
(
Entries 1–6). The reaction rates for these substrates were higher
−
1
−1
than 5 mmol g-cat
h in n-heptane, and also higher than those
in water solvent for the same substrates. The improvement of
activity by changing the solvent from water to n-heptane (>3
times higher reaction rate) is larger than that by addition of acid in
water solvent (up to 2 times) [24]. On the other hand, in the cases
of cis-1,2-cyclohexanediol, 1,2-hexanediol and tetrahydrofurfuryl
alcohol, the reaction rates in n-heptane were low (<2.5 mmol g-
−
1
−1
cat
h ) (Entries 9, 11, 13). In addition, regio-selectivities in the
hydrogenolysis of 1,2-hexanediol and tetrahydrofurfuryl alcohol
in n-heptane were significantly lower than those in water (Entries
1
1–14). Therefore, in the catalysis of Ir ReOx/SiO , alkane solvent
2
is only good at the hydrogenolysis of mono-alcohols and trans-1,2-
cyclohexanediol. Most multi-functionalized substrates react better
in water. The higher solubility of multi-functionalized substrates in
water than in n-heptane also calls for the use of water solvent, while
alkanes are good solvents for hydrophobic higher mono-alcohols
in terms of the solubility as well as the hydrogenolysis rate.
The difference of catalytic performance between in these sol-
vents was further investigated in view of reaction mechanism. We
have proposed the following “direct” mechanism of Ir ReOx/SiO2
catalyst in water for most multi-functionalized substrates such as
glycerol, 1,2-propanediol, erythritol, 2-alkoxyethanol, and tetrahy-
drofurfuryl alcohol (Fig. 7A) [22–26]: substrate molecules are
adsorbed on ReOx site of the catalyst with one of the OH groups
Fig. 4. XRD patterns of Ir ReOx/SiO2 (Re/Ir = 2) catalyst after (a) reduction in n-
heptane at 473 K and (b) catalytic use. Conditions for catalytic use were the same as
used in Table 1.
3.3. Hydrogenolysis of various alcohols over Ir ReOx/SiO2
(
Re/Ir = 2)
The difference of the performance of Ir ReOx/SiO2 catalyst
between in n-heptane and water was investigated for various
Fig. 5. TEM images of Ir ReOx/SiO2 (Re/Ir = 2) catalyst after (a) reduction in n-heptane at 473 K and (b) catalytic use. Conditions for catalytic use were the same as used in
Table 1.
3
3
Fig. 6. Results of Re L3-edge EXAFS analysis of model compounds and Ir ReOx/SiO2 (Re/Ir = 2). (I) k -Weighted EXAFS oscillations. (II) Fourier transform of k -weighted Ir
−1
L3-edge EXAFS, FT range: 30–120 nm . (III) Fourier filtered EXAFS data (solid line) and calculated data (dotted line), Fourier filtering range: 0.129–0.316 nm. (a) Re powder,
b) NH4ReO4, (c) Ir ReOx/SiO2 (Re/Ir = 2) after the catalytic use. The reaction conditions were the same as used in Table 1.
(

4
22
Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
Table 2
Curve fitting of Re L3-edge EXAFS analysis of Ir ReOx/SiO2 (Re/Ir = 2) catalysts.
−
1
ꢀ/10 nma
−1
ꢁE0/eVb
Rf/%c
Sample
Shells
Re
Coordination number
Bond distance/10 nm
Used in n-heptane
O
1.6
6.2
2.04
2.69
0.081
0.082
−2.6
2.4
Re Ir (or Re)
9.6
Used in waterd
Re
O
1.7
6.1
2.03
2.68
0.080
0.082
0.6
8.1
2.5
Re Ir (or Re)
a
Debye–Waller factor.
b
Difference in the origin of photoelectron energy between the reference and the sample.
Residual factor.
Ref. [23]. After the catalytic use for glycerol hydrogenolysis.
c
d
of the substrate molecule to form ORe group. Hydride-like hydro-
gen species is formed on the Ir metal site by the activation of H2
on the catalyst. The adsorbed substrate is attacked by the hydride-
like species from the backside of the breaking C O bond (SN2-type
attack). This mechanism leads to the selective dissociation of
Even in the “direct” mechanism (Fig. 7A), when glycerol is adsorbed
with the secondary OH group, 1,2-propanediol is produced via the
attack of terminal OH group neighboring the secondary one [24,37].
The reaction kinetics, especially the reaction order with respect
to H2 pressure, is valuable information in determination of the
reaction mechanism. Here, we assume that the substrate and H2
is not competitively adsorbed on the catalyst surface. The “direct”
mechanism shows first-order kinetics with respect to H2 pres-
sure. The two-step indirect mechanism typically shows zero-order
kinetics. This is supported by the fact that dehydration product
is usually not observed during the reaction, which means that
the dehydration is the rate-determining step. A negative reac-
tion order means that the reaction involves a dehydrogenation
step, such as the first step in the three-step indirect mechanism.
The reaction orders of the hydrogenolysis of various substrates in
water over Ir ReOx/SiO2 are shown in Table 4 and the detailed
data is shown in Table S1 (Supplementary Information). In pre-
vious papers we reported that the reaction order with respect to
C O bond neighboring the terminal CH OH group such as 1,2-
2
propanediol to 1-propanol and tetrahydrofurfuryl alcohol to 1,5-
pentanediol. However, this “direct” mechanism cannot be applied
to 1,2-cyclohexanediols, whose OH groups cannot be attacked
from the backside because of the fixed configuration of adsorbed
substrate [26]. Mono-alcohols are also unsuitable substrates for this
mechanism. There are two types of well-known mechanisms for the
C O hydrogenolysis: One is the two-step indirect mechanism that
composed of acid-catalyzed dehydration and subsequent hydro-
genation (Fig. 7B) [7]. This two-step mechanism can be applied
to any alcohols, and the systems for which this mechanism has
been proposed include glycerol hydrogenolysis to 1,2-propanediol
in acidic conditions [8,33,34]. We have already proposed that this
two-step indirect mechanism works in the hydrogenolysis of sec-
H2 pressure in Ir ReOx/SiO -catalyzed hydrogenolysis in water
2
ondary mono-alcohols to alkanes over Ir ReOx/SiO in water based
is first for glycerol [22,23], erythritol [25], and tetrahydrofurfuryl
alcohol [26], supporting the “direct” mechanism. In the cases of
cis- and trans-1,2-cyclohexanediol, the reaction order is −0.5 and
−0.2, respectively [26]. Hydrogenolysis of cis-1,2-cyclohexanediol
in water clearly involves a dehydrogenation step. One explanation
is the three-step indirect mechanism. However, this mechanism
does not work over Ir ReOx/SiO2 for glycerol, which is the most
typical substrate for this mechanism, according to the data that the
yield of 1,2-propanediol (minor by-product) is slightly increased
2
on the data that the reaction rate is much enhanced by the addi-
tion of an acid [27]. The other is the three-step indirect mechanism
that composed of dehydrogenation, dehydration and hydrogena-
tion (Fig. 7C) [35]. This three-step mechanism can be applied to only
limited substrates, and the typical system for which this three-step
mechanism has been proposed is the hydrogenolysis of glycerol to
1
,2-propanediol over Cu-based catalysts [36]. It should be noted
that 1,2-propanediol can be produced in any mechanism in Fig. 7:
Table 3
Hydrogenolysis of various alcohols over Ir ReOx/SiO2 (Re/Ir = 2) in n-heptane or water.a
Entry
Substrate
Solvent
Catalyst
amount/mg
Time/h
Conv./%
Product (selectivity/%)
Reaction rate/mmol
g-cat
−
1
−1 b
h
1
2
3
4
5
6
7
1-Hexanol
1-Propanol
2-Hexanol
2-Propanol
Cyclohexanol
Cyclohexanol
trans-1,2-Cyclohexanediol
n-Heptane
Water
n-Heptane
Water
n-Heptane
Water
n-Heptane
100
100
100
100
25
1
2
1
2
1
1
2
14
3.3
19
7.1
16
5.6
12
n-Hexane (>99)
n-Propane (>99)
n-Hexane (>99)
Propane (>99)
Cyclohexane (>99)
Cyclohexane (>98)
Cyclohexanol (62), cyclohexane (7),
cis-1,2-cyclohexanediol (31)
Cyclohexanol (63), cyclohexane (21),
cis-1,2-cyclohexanediol (16)
Cyclohexanol (73), cyclohexane (8),
trans-1,2-cyclohexanediol (19)
Cyclohexanol (52), cyclohexane (7),
trans-1,2-cyclohexanediol (42)
1-Hexanol (57), 2-hexanol (35),
n-hexane (8)
6.8
1.4
9.1
3.0
31
100
50
2.8
5.0
8
9
trans-1,2-Cyclohexanediol
cis-1,2-Cyclohexanediol
cis-1,2-Cyclohexanediol
1,2-Hexanediol
Water
100
100
100
100
100
100
2
8
1
4
2
2
20
7.2
22
10
28
9.6
4.4
0.4
9.6
1.1
5.9
2.4
n-Heptane
Water
1
1
1
1
0
1
2
3
n-Heptane
Water
1,2-Hexanediol
1-Hexanol (81), 2-hexanol (12),
n-hexane (7)
1,5-Pentanediol (50), 1-pentanol (30),
Tetrahydrofurfuryl alcohol
n-Heptane
2-methyltetrahydrofuran (12),
n-pentane (8)
1
4
Tetrahydrofurfuryl alcohol
Water
100
0.1
5.6
1,5-Pentanediol (97), 1-pentanol (3)
28
a
Reaction conditions: Ir ReOx/SiO2 (Ir 4 wt%, Re/Ir = 2) 25–100 mg, substrate 0.5 g, solvent 10 ml, H2 6.8 MPa, 393 K.
Based on total products.
b

Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
423
Fig. 7. Possible hydrogenolysis mechanisms.
with increasing H2 pressure [23]. Formation of 1,2-propanediol
for cis-1,2-cyclohexanediol [26]; however, the reaction order is
more close to zero and other mechanisms such as the two-step
indirect hydrogenolysis may be involved. Hydrogenolysis of cyclo-
hexanol and 2-propanol showed zero-order kinetics, which can
be explained by the two-step indirect mechanism. Hydrogenol-
ysis of 1-propanol showed the reaction order of 0.83, suggesting
the involvement of the same hydride-like species as present in the
hydrogenolysis of multi-functionalized substrates such as glycerol,
from glycerol over Ir ReOx/SiO is probably via the “direct” mech-
2
anism and/or the two-step indirect mechanism. We have not ruled
out the possibility of other mechanisms than three-step indi-
rect mechanism for the hydrogenolysis of cis-1,2-cyclohexanediol,
such as dehydrogenation followed by direct hydrogenolysis [26].
For the hydrogenolysis of trans-1,2-cyclohexanediol in water, we
have tentatively supposed that the mechanism is the same as that
Table 4
Reaction orders of hydrogenolysis with respect to H2 pressure over Ir ReOx/SiO2 in water.
Substrate
Catalyst
Reaction ordera
Reference
Glycerol
Glycerol
Erythritol
Tetrahydrofurfuryl alcohol
Ir ReOx/SiO2 (Re/Ir = 1) + H2SO4
Ir ReOx/SiO2 (Re/Ir = 1) + H-ZSM-5
Ir ReOx/SiO2 (Re/Ir = 1) + H2SO4
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
Ir ReOx/SiO2 (Re/Ir = 2)
1
[23]
[24]
[25]
[26]
0.97
0.85
1.0
0.93
−0.5
−0.2
−0.1
0.0
b
1
,2-Hexanediol
–
cis-1,2-Cyclohexanediol
trans-1,2-Cyclohexanediol
Cyclohexanol
[26]
[26]
b
–
–
–
b
2
1
-Propanol
-Propanol
b
0.83
a
With respect to H2 pressure; based on the total formation rate of all products.
This work; reaction conditions were the same as shown in Table 3 with variation in H2 pressure (0.8–6.8 MPa). The detailed data is shown in Supplementary Information
b
(
Table S1).

4
24
Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
Table 5
Reaction orders of hydrogenolysis over Ir ReOx/SiO2 (Re/Ir = 2) in n-heptane.
site on Ir ReOx/SiO2 is sensitive to small change in the structure
of substrate for both “direct” hydrogenolysis and dehydration. It
should be noted that the dependence of the catalytic activity of
Ir ReOx/SiO2 on Re loading amount is almost the same for glyc-
erol or tetrahydrofurfuryl alcohol hydrogenolysis in water (“direct”
mechanism”) [23,26] and trans-1,2-cyclohexanediol hydrogenoly-
sis in n-heptane (Table 1; two-step indirect mechanism). Therefore
the active site for the “direct” mechanism, which we proposed
is the interface between the Ir metal and ReOx cluster [22–26],
may be also the acid site for the two-step indirect mechanism,
although the very catalytic center should be different: The Ir
atom on the Ir Re interface activates hydrogen to give active
species for “direct” mechanism, and Re (or ReO H [21]) on the
same Ir Re interface is the acidic center for the two-step mech-
anism (Fig. 7A and B). From Fig. 3, the catalytically active Ir Re
interface site is surrounded by large ReOx clusters. The steric
crowding by large ReOx clusters may induce the sensitivity of the
activity to the substrate structure; in addition, adsorption of the
substrate on the outer surface of ReOx clusters may block the
adsorption on the catalytically active Ir Re interface site. How-
ever, further investigation is necessary in the state of adsorbed
substrates.
Substrate
Reaction ordersa
H2 pressure
Substrate concentration
Tetrahydrofurfuryl alcohol
0.94
0.78
0.06
0.10
1
,2-Hexanediol
trans-1,2-Cyclohexanediol
Cyclohexanol
−0.04
−0.25
0.02
0.12
2
1
-Hexanol
-Hexanol
−0.03
0.34
0.10
−0.13
a
Based on the total formation rate of all products. Reaction conditions were the
same as shown in Table 3 with variation in H2 pressure (0.8–6.8 MPa) or n-heptane
amount (5–15 ml). The detailed data is shown in Supplementary Information (Tables
S2 and S3).
although the reaction rate was much smaller than in the typical
“
direct” mechanism.
The reaction orders in n-heptane solvent are shown in Table 5
and the detailed data is shown in Tables S2 and S3 (Supplemen-
tary Information). All substrates showed zero-order kinetics with
respect to substrate concentration. These data indicate the high
coverage of substrate on the catalyst surface in n-heptane, and
also indicate that the adsorption of substrate does not prevent the
adsorption and activation of H . On the H pressure, hydrogenolysis
2
2
4. Conclusions
of tetrahydrofurfuryl alcohol and 1,2-hexanediol showed a reaction
order of around one, suggesting the “direct” mechanism similarly
to the cases of the same substrates in water. The lower reaction
rates for these substrates in n-heptane than in water indicate that
alkanes are less efficient solvents for the “direct” mechanism. The
reason of low activity may include the difficulty in the forma-
tion of anionic hydrogen species in non-polar alkane solvent. The
lower regioselectivity in the hydrogenolysis of tetrahydrofurfuryl
alcohol or 1,2-hexanediol in n-heptane than in water may be due
to different character of hydrogen species or different structure
of adsorption species: in water the substrate is adsorbed prefer-
ably with the terminal OH group [22–26], while in n-heptane the
substrate may be adsorbed with any oxygen-containing groups
in it. Hydrogenolysis of cyclohexanol and 2-hexanol in n-heptane
showed zero-order kinetics with respect to H2 pressure similarly
to in water, suggesting the two-step indirect mechanism. The
high reactivity of cyclohexanol and 2-hexanol can be explained
by higher acid strength of the catalyst surface in alkane solvent
than in water, since water molecule can be coordinated to the acid
site. Hydrogenolysis of 1-hexanol showed a small positive reaction
1
. Ir ReOx/SiO2 catalyst can be used in alkane solvent for C O
hydrogenolysis of alcohols without dissociation of C C bonds
as well as in water solvent. The use of alkane solvent widens the
substrate scope to higher alcohols.
2
. Alkane solvent is advantageous to water solvent in the
hydrogenolysis of secondary mono-alcohols and trans-1,2-
cyclohexanediol. This is due to the stronger adsorption of
substrate on the catalyst surface and the higher activity for
the reaction route that composed of acid-catalyzed dehydration
and subsequent hydrogenation (the two-step indirect mech-
anism). In the hydrogenolysis of trans-1,2-cyclohexanediol,
Ir ReOx/SiO2 catalyst shows the highest cyclohexanol yield of
7
1
4%. At longer reaction time, the total hydrodeoxygenation of
,2-cyclohexanediol to cyclohexane is also possible.
3
. Hydrogenolysis of 1,2-hexanediol and tetrahydrofurfuryl alco-
hol, which can be readily converted in water to 1-hexanol and
1
,5-pentanediol, respectively, is less efficient in alkane solvent.
The “direct” mechanism which is driven by the hydride-like
active hydrogen species works better in water solvent.
. Alkane solvent also suppresses the reaction route involving a
dehydrogenation step, which is observed in hydrogenolysis of
cis-1,2-cyclohexanediol in water solvent.
order (0.34) with respect to H pressure. Both the two-step indirect
2
4
5
mechanism and the mechanism involving the hydride-like species
may be involved in the hydrogenolysis of 1-hexanol in n-heptane.
Hydrogenolysis of trans-1,2-cyclohexanediol in n-heptane showed
the reaction order of almost zero. Considering that hydrogenolysis
of cis-1,2-cyclohexanediol showed much lower rate in n-heptane
than in water, the mechanism for cis-1,2-cyclohexanediol substrate
in water (mechanism involving a dehydrogenation step) does not
work in n-heptane. Hydrogenolysis of trans-1,2-cyclohexanediol
in n-heptane may well proceed via the two-step indirect mecha-
nism.
. The structure of Ir ReOx/SiO catalyst in both alkane and water
2
solvents is the same. The active site for the “direct” mechanism,
the interface between the Ir metal and ReOx cluster, may be also
the acid site for the two-step indirect mechanism.
Acknowledgements
From above, hydrogenolysis in n-heptane proceeds via the
two-step indirect mechanism for mono-alcohols and trans-1,2-
cyclohexanediol, while this mechanism does not, strangely, work
well for cis-1,2-cyclohexanediol and 1,2-hexanediol. This suggests
that the acid site required for the two-step indirect mecha-
nism is hard for cis-1,2-cyclohexanediol and 1,2-hexanediol to
access. In our previous paper, we have reported that that 1,2,3-
butanetriol and threitol (stereoisomer of erythritol) hardly react
over Ir ReOx/SiO2 catalyst, while glycerol and erythritol react
smoothly [25]. These tendencies are interesting, but difficult to
explain at present. Overall, the catalytic activity of the active
This work is supported by the Cabinet Office, Government of
Japan through its “Funding Program for Next Generation World-
Leading Researchers”. Authors appreciate Prof. Tokushi Kizuka
(Univ. of Tsukuba) for TEM observation of the catalyst.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.apcata.
2013.09.021.

Y. Nakagawa et al. / Applied Catalysis A: General 468 (2013) 418–425
425
References
[21] M. Chia, Y.J. Pagán-Torres, D. Hibbitts, Q. Tan, H.N. Pham, A.K. Datye, M. Neurock,
R.J. Davis, J.A. Dumesic, J. Am. Chem. Soc. 133 (2011) 12675–12689.
[
[
[
[
[
[
22] Y. Nakagawa, Y. Shinmi, S. Koso, K. Tomishige, J. Catal. 272 (2010)
91–194.
23] Y. Amada, Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal.
B: Environ. 105 (2011) 117–127.
[
[
[
1] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044–4098.
2] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
3] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007)
1
7
164–7183.
24] Y. Nakagawa, X. Ning, Y. Amada, K. Tomishige, Appl. Catal. A: Gen. 433–434
[
[
4] J.J. Bozell, G.R. Petersen, Green Chem. 12 (2010) 539–554.
5] 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) 484–489.
(
2012) 128–134.
25] Y. Amada, H. Watanabe, Y. Hirai, Y. Kajikawa, Y. Nakagawa, K. Tomishige, Chem-
SusChem 5 (2012) 1991–1999.
26] K. Chen, K. Mori, H. Watanabe, Y. Nakagawa, K. Tomishige, J. Catal. 294 (2012)
[
[
[
[
6] A.-L. Marshall, P.J. Alaimo, Chem. Eur. J. 16 (2010) 4970–4980.
7] M. Schlaf, Dalton Trans. (2006) 4645–4653.
8] Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 1 (2011) 179–190.
9] S. Zhu, X. Gao, Y. Zhu, Y. Zhu, H. Zheng, Y. Li, J. Catal. 303 (2013) 70–79.
1
71–183.
27] K. Chen, M. Tamura, Z. Yuan, Y. Nakagawa, K. Tomishige, ChemSusChem 6
2013) 613–621.
(
[
[
28] J.W. Cook, D.E. Sayers, J. Appl. Phys. 52 (1981) 5024–5031.
29] K. Okumura, J. Amano, N. Yasunobu, M. Niwa, J. Phys. Chem. B 104 (2000)
[
[
10] S. Xia, R. Nie, X. Lu, L. Wang, P. Chen, Z. Hou, J. Catal. 296 (2012) 1–11.
11] J. ten Dam, K. Djanashvili, F. Kapteijn, U. Hanesfeld, ChemCatChem 5 (2013)
1
050–1057.
30] K. Okumura, S. Matsumoto, N. Nishiaki, M. Niwa, Appl. Catal. B: Environ. 40
2003) 151–159.
31] A.L. Ankudinov, B. Ravel, J.J. Rehr, S.D. Conradson, Phys. Rev. B 58 (1998)
565–7576.
4
97–505.
[
[
[
[
[
[
[
12] T. Mizugaki, T. Yamakawa, R. Arundhathi, T. Mitsudome, K. Jitsukawa, K.
(
Kaneda, Chem. Lett. 41 (2012) 1720–1722.
[
[
13] L.-Z. Qin, M.-J. Song, C.-L. Chen, Green Chem. 12 (2010) 1466–1472.
14] T. Kurosaka, H. Maruyama, I. Naribayashi, Y. Sasaki, Catal. Commun. 9 (2008)
7
32] Y. Amada, H. Watanabe, M. Tamura, Y. Nakagawa, K. Okumura, K. Tomishige, J.
Phys. Chem. C 116 (2012) 23503–23514.
33] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 240 (2006)
1
360–1363.
[
[
15] B. Zhang, Y. Zhu, G. Ding, H. Zheng, Y. Li, Green Chem. 14 (2012) 3402–3409.
16] S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori, K. Tomishige, Chem.
Commun. (2009) 2035–2037.
2
13–221.
34] M.A. Dasari, P.-P. Kiatsimkul, W.R. Sutterlin, G.J. Suppes, Appl. Catal. A: Gen.
81 (2005) 225–231.
35] C. Montassier, J.C. Ménézo, L.C. Hoang, C. Renaud, J. Barbier, J. Mol. Catal. 70
1991) 99–110.
[
[
17] S. Koso, N. Ueda, Y. Shinmi, K. Okumura, T. Kizuka, K. Tomishige, J. Catal. 267
2
(
2009) 89–92.
18] K. Chen, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, ChemCatChem 2 (2010)
47–555.
19] S. Koso, Y. Nakagawa, K. Tomishige, J. Catal. 280 (2011) 221–229.
20] S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, Appl. Catal. B:
Environ. 111–112 (2012) 27–37.
(
5
[
[
36] S. Wang, Y. Zhang, H. Liu, Chem. Asian J. 5 (2010) 1100–1111.
37] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B: Environ.
[
[
94 (2010) 318–326.