Solid acid co-catalyst for the hydrogenolysis of glycerol to 1,3-propanediol over Ir-ReOx/SiO2

Article abstract of DOI:10.1016/j.apcata.2012.05.009

Hydrogenolysis of aqueous glycerol was conducted with Ir-ReO _x/SiO₂ catalyst and solid acid co-catalyst. Considering the reusability and activity, H-ZSM-5 is the most suitable solid co-catalyst. The property of Ir-ReO_x/SiO₂ + H-ZSM-5 system including kinetics and selectivity trends in various reaction conditions is similar to the case of Ir-ReO_x/SiO₂ + H₂SO₄. The catalyst stability, activity, and the maximum yield of 1,3-PrD of Ir-ReO _x/SiO₂ + H-ZSM-5 were slightly lower than Ir-ReO _x/SiO₂ + H₂SO₄. The role of added acid may be to protonate the surface of ReO_x cluster to increase the number of hydroxorhenium site, which activates glycerol by the formation of glyceride species.

Full text of DOI:10.1016/j.apcata.2012.05.009

Applied Catalysis A: General 433–434 (2012) 128–134
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Solid acid co-catalyst for the hydrogenolysis of glycerol to 1,3-propanediol over
Ir-ReOx/SiO2
Yoshinao Nakagawa, Xuanhe Ning, Yasushi Amada, 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:
Hydrogenolysis of aqueous glycerol was conducted with Ir-ReOx/SiO2 catalyst and solid acid co-catalyst.
Considering the reusability and activity, H-ZSM-5 is the most suitable solid co-catalyst. The property of
Ir-ReOx/SiO2 + H-ZSM-5 system including kinetics and selectivity trends in various reaction conditions
is similar to the case of Ir-ReOx/SiO2 + H2SO4. The catalyst stability, activity, and the maximum yield of
1,3-PrD of Ir-ReOx/SiO2 + H-ZSM-5 were slightly lower than Ir-ReOx/SiO2 + H2SO4. The role of added acid
may be to protonate the surface of ReOx cluster to increase the number of hydroxorhenium site, which
activates glycerol by the formation of glyceride species.
Received 25 February 2012
Received in revised form 31 March 2012
Accepted 10 May 2012
Available online 17 May 2012
Keywords:
Hydrogenolysis
Glycerol
© 2012 Elsevier B.V. All rights reserved.
Iridium
Rhenium
Zeolite
1. Introduction
1,3-PrD [19]. We have very recently reported that silica-supported
rhenium-modified iridium catalyst (Ir-ReOx/SiO2) is very effec-
Utilization of biomass is essential in the sustainable develop-
ment of our society [1–7]. Glycerol is one of the most important
compounds in the chemical conversion of biomass, since vast
amount of glycerol is co-produced in the manufacture of biodiesel
fuel from vegetable oil and lower alcohols [8,9]. Hydrogenolysis of
glycerol is an attractive route in the chemical conversion of glycerol
because of the large demand of products. The primary products are
1,2-propanediol (1,2-PrD) and 1,3-propanediol (1,3-PrD). Overhy-
drogenolysis reactions produce 1-propanol (1-PrOH), 2-propanol
(2-PrOH) and propane (Scheme 1). Other side-reactions include
degradation which produces C2 compounds such as ethylene glycol
and C1 compounds such as methane. In these compounds, 1,3-PrD
is most valuable and can be used in large scale for polyester pro-
duction. Despite the recent extensive studies in the developments
of catalytic systems for glycerol hydrogenolysis [10], selective for-
mation of 1,3-PrD is very difficult and the reported systems have
been limited [11–23]. In the literature, rhodium- or platinum-based
catalysts combined with tungsten-based co-catalyst or modifier
are relatively effective in the glycerol hydrogenolysis to 1,3-PrD.
Our group reported that rhenium is the best modifier for rhodium
catalysts in terms of hydrogenolysis activity and selectivity to
tive in the selective hydrogenolysis of glycerol to 1,3-PrD [20,21].
Almost no activity is observed for monometallic Ir/SiO2 and
ReOx/SiO2 catalysts, indicating that the synergy between iridium
and rhenium enables the hydrogenolysis. The maximum 1,3-PrD
yield is 38% [20], which value is still the highest in the litera-
ture of the catalytic hydrogenolysis of aqueous glycerol. However,
this system requires the addition of sulfuric acid to the reaction
mixture. The presence of corrosive sulfuric acid makes it diffi-
cult to design the reaction system including reactor and product
separation system. Therefore, substituting solid acids for sulfuric
acid is advantageous to the practical use of the Ir-ReOx/SiO2-based
catalytic system. In this study, various solid acids including an ion-
exchange resin, silica–alumina and zeolites were applied to the
co-catalyst for the glycerol hydrogenolysis over Ir-ReOx/SiO2. In
terms of the activity and reusability, H-ZSM-5 is the best additive
among the solid acids tested. The role of acid in this reaction is also
discussed.
2. Experimental
2.1. Catalysts
Ir-ReOx/SiO2 catalyst was prepared by sequential impregnation
method using Fuji Silysia G-6 silica, H2IrCl6 aq and NH4ReO4
aq followed by drying and calcination at 773 K, as described in
the previous reports [20,21]. Iridium nitrate was also used as a
∗
Corresponding author. Tel.: +81 22 795 7214; fax: +81 22 795 7214.
E-mail addresses: tomi@erec.che.tohoku.ac.jp, tomi@tulip.sannet.ne.jp
(K. Tomishige).
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2012.05.009

Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
129
HO
OH
HO
1,3-Propanediol
1-Propanol
(1-PrOH)
(1,3-PrD)
HO
OH
Propane
OH
HO
Glycerol
OH
OH
1,2-Propanediol
(1,2-PrD)
2-Propanol
(2-PrOH)
Scheme 1. Reaction routes in the hydrogenolysis of glycerol.
precursor instead of H2IrCl6, and the catalytic performances were
same. The loading amount of Ir was 4 wt.%, and the ratio of Re to
Ir was 1 in molar basis (3.9 wt.% Re). MgO (500A, Ube Material
Industries), CeO2 (HS, Daiichi Kigenso), ZrO2 (RC-100P, Daiichi
Kigenso), TiO2 (P25, Nippon Aerosil Co., Ltd.), silica–alumina (JRC-
SAL-3, Nikki Shokubai Kasei and Catalysis Society of Japan), H-beta
(HSZ-930HOA, Tosoh, Si/Al2 = 27), H-mordenite (HSZ-620HOD,
Tosoh, Si/Al2 = 15.7), and H-ZSM-5 (JRC-Z5-90H(1), Süd-Chemie
Catalysts and Catalysis Society of Japan, Si/Al2 = 90) were used
as received. The dry-based compositions of the silica–alumina
were SiO2 86%, Al2O3 13 wt.%, Na2O 0.02 wt.%, SO4 0.6 wt.%, and Fe
0.01 wt.%. Removal of Fe from the silica–alumina was conducted by
washing JRC-SAL-3 with 0.5 M HNO3 aq at room temperature for
4 h. Ir-ReOx/silica–alumina was prepared by the similar procedure
as Ir-ReOx/SiO2 using washed silica–alumina instead of SiO2.
2.3. Characterization
Temperature-programmed reduction (TPR) was carried out in
a fixed-bed reactor equipped with a thermal conductivity detector
using 5% H2 diluted with Ar (30 ml/min). The amount of catalyst was
0.05 g, and temperature was increased from room temperature to
1123 K at a heating rate of 10 K/min. X-ray diffraction (XRD) pat-
terns were recordedbyadiffractometer (Rigaku UltimaIV). Average
metal particle size was estimated using the Scherrer equation [24].
TEM images were taken with HITACHI HF 2000EDX.
3. Results and discussion
3.1. Addition of acid in Ir-ReOx/SiO2-catalyzed glycerol
hydrogenolysis
First, the effect of sulfuric acid in the Ir-ReOx/SiO2-catalyzed
hydrogenolysis was investigated in detail. Table 1 shows the
hydrogenolysis of glycerol and the related compounds over Ir-
ReOx/SiO2 with or without addition of H2SO4. In the hydrogenolysis
of glycerol, 1,3-PrD was more favorably produced than 1,2-PrD. It
should be noted that 1-PrOH and 2-PrOH were significantly pro-
duced even in the initial stage, indicating the presence of direct
formation routes for PrOHs in addition to the overhydrogenolysis of
free PrDs. While the initial selectivity to 1,3-PrD is almost the same
(50–60%) for both cases with and without H2SO4, the selectivities
to the other products were different. The H2SO4 addition signifi-
cantly decreased the formation of 1,2-PrD (25% select. → 14%). At
the same time, direct formation of 1-PrOH was promoted (14%
select. → 21%). The initial rates of hydrogenolysis of all substrates
(glycerol and PrDs) became almost twice by the addition of H2SO4.
At longer reaction time, the selectivities to PrDs were decreased
and that to 1-PrOH was increased because of the overhydrogenoly-
sis. The selectivity to 2-PrOH was hardly changed with time, which
can be explained by the low selectivity to 2-PrOH in the overhy-
drogenolysis reactions. At a similar conversion level, the selectivity
to 1,3-PrD was similar for both cases with and without the addition
of H2SO4. The selectivity to 1,2-PrD in the case with the addition of
H2SO4 was lower and that to 1-PrOH was higher than in the case
without the addition. These differences were mainly caused by the
difference of the initial selectivity patterns of glycerol hydrogenol-
ysis, considering that the rates of both glycerol hydrogenolysis and
overhydrogenolysis of PrDs were increased at a constant ratio by
acid addition.
2.2. Reaction
Catalytic reactions were performed in a 190-ml stainless steel
autoclave with an inserted glass vessel. Ir-ReOx/SiO2 catalyst was
put into an autoclave together with a spinner and an appropriate
amount of water and heated at 473 K with 8 MPa H2 for 1 h for
the reduction pretreatments. After the pretreatment, the autoclave
was cooled down, and hydrogen was removed. Glycerol (Wako
Pure Chemical Industries, Ltd., >99%) and solid acid co-catalyst was
quickly put into the autoclave. After sealing the reactor, the air con-
tent was purged by flushing thrice with 1 MPa hydrogen (99.99%;
Takachiho Trading Co., Ltd.). The autoclave was then heated to
393 K, and the temperature was monitored using a thermocouple
inserted in the autoclave. After the temperature reached 393 K, the
H2 pressure was increased to 8 MPa. During the experiment, the
stirring rate was fixed at 500 rpm (magnetic stirring), while lower-
ing the stirring rate to 250 rpm did not change the results ( 0.5% in
both conversion and selectivities) in the standard reaction condi-
tions (20 wt.% aqueous glycerol solution 20 g, Ir-ReOx/SiO2 150 mg,
H-ZSM-5 60 mg, H2 8 MPa, 393 K, 24 h). Products in both liquid and
gas phases were analyzed using a gas chromatograph (Shimadzu
GC-2014) equipped with FID and TC-WAX capillary column. Prod-
ucts were also identified with GC–MS (Shimadzu QP5050). The
mass balance was also confirmed in each result and the difference
in mass balance was always in the range of the experimental error,
suggesting that polymeric byproducts were not formed. The used
catalyst was separated and collected by centrifugation. When nec-
essary, the recovered catalyst was calcined at 773 K for 3 h. In reuse
experiments, the loss of the catalyst in recovery process was com-
pensated by addition of small amount of fresh catalyst (ca. 10%)
and the pretreatment (under 8 MPa H2 at 473 K in water) was car-
ried out again regardless of whether calcined or not. The amount of
eluted metal into the reaction solution was analyzed by inductively
coupled plasma atomic emission spectrometry (ICP-AES, Thermo
Scientific iCAP6500).
Next, the addition effect of various metal oxides or solid acids
instead of sulfuric acid was investigated (Table 2). Dehydration
products such as acetol [24–27] and 3-hydroxypropanal were not
detected at all in all cases. The addition of strongly basic MgO and
weakly basic CeO2 led much decreases in both activity and 1,3-
PrD selectivity (Entries 3 and 4). The loss of 1,3-PrD selectivity was
related to the decrease of the reactivity ratio of secondary/primary
OH groups. The addition of weakly acidic ZrO2 or TiO2 showed

130
Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
Table 1
Hydrogenolysis of glycerol and related compounds over Ir-ReOx/SiO2.
Additive
Substrate
Time (h)
Conv. (%)
Selectivity (%)
1,3-PrD
1,2-PrD
1-PrOH
2-PrOH
Propane
None
None
H2SO4
H2SO4
None
H2SO4
None
H2SO4
Glycerol
Glycerol
Glycerol
Glycerol
1,3-PrD
1,3-PrD
1,2-PrD
1,2-PrD
4
40
4
24
4
4
4
4
8.4
61.3
17.1
61.1
2.1
4.0
22.9
38.9
51.0
43.3
57.6
43.2
–
–
–
–
25.2
14.5
14.1
9.2
–
–
–
–
14.1
31.1
20.9
36.6
99.1
99.1
85.5
89.7
9.3
10.3
7.3
9.7
–
–
13.8
10.3
0.4
0.8
<0.1
1.3
0.9
0.6
0.7
<0.1
PrD, propanediol; PrOH, propanol.
Reaction conditions: substrate (43 mmol; glycerol: 4 g, propanediol: 3.3 g), solvent: water (total 20 g), Ir-ReOx/SiO2 (150 mg), H2SO4 (0 or 1.5 mg; H+/Ir = 1), H2 (8 MPa), 393 K.
little effects on the activity and selectivities (Entries 5 and 6).
The addition of small amount of acidic zeolites showed effects in
both activity and selectivities similar to those of H2SO4 (Entries
7, 9 and 11). The activities of the systems with small amount of
zeolites were similar to one another while the pH values of the
reaction mixtures of these systems were different. However, larger
amount of H-mordenite and H-beta decreased the activity (Entries
8 and 10). In contrast, the addition effect of H-ZSM-5 was almost
unchanged by excess amount of H-ZSM-5 (Entry 12). The differ-
ence of the effect of larger addition amount may be caused by the
difference in the stability of zeolites in hydrothermal conditions
and/or the difference of the pore size of the zeolites. The addi-
tion of as-received silica–alumina decreased the activity (Entry 13).
We thought that the Fe impurity may poison the catalyst since
we found that the activity of Ir-ReOx/SiO2 is much lowered when
steel reactor without glass inner tubes is used. Then we washed
the silica–alumina with dilute nitric acid to remove Fe. The addi-
tion of washed silica–alumina enhanced the catalytic performance,
although the effects were slightly smaller than those of H2SO4 or
zeolites (Entries 14 and 15). We also prepared Ir-ReOx catalyst sup-
ported on washed silica–alumina. However, the activity was much
lower than the case where silica–alumina was externally added
or the case even without acid addition (Entry 16). The interaction
between aluminum site and iridium (and/or rhenium) may disturb
the formation of active iridium-rhenium interface. The addition of
acidic ion-exchange resin Amberlyst 70 is very effective for activity
and the effect was even larger than that of H2SO4 although the pH
value of the reaction mixture is higher (Entries 17 and 18). It should
be noted that Amberlyst is better acid than H2SO4 in Ru/C + acid sys-
tem for the hydrogenolysis of glycerol to 1,2-PrD [28–31]. Based on
the above results, we chose Amberlyst 70 and H-ZSM-5 as additives.
3.2. Stability of Ir-ReOx/SiO2 + Amberlyst 70 or H-ZSM-5
The life of the catalyst was investigated with reuse experi-
ments. The results were summarized in Table 3. When the mixture
of Ir-ReOx/SiO2 + Amberlyst 70 or H-ZSM-5 was reused as recov-
ered, the activity monotonously dropped to about half in three
uses. In the reported Ir-ReOx/SiO2 + H2SO4 system, the recovered
catalyst can be reused without loss of activity or leaching after
calcination at 773 K. However, calcination of the mixture of Ir-
ReOx/SiO2 + Amberlyst 70 burns the ion-exchange resin. Even when
Amberlyst 70 was added again after calcination of used cata-
lyst mixture, the activity was slightly decreased. In the case of
Ir-ReOx/SiO2 + H-ZSM-5, the decrease of activity in the repeated
uses was much slower than the cases of Ir-ReOx/SiO2 only or
Ir-ReOx/SiO2 + Amberlyst 70 when the recovered catalysts were
calcined before the next use. Selectivities were almost unchanged
during the reuses.
The leaching of Ir or Re from the Ir-ReOx/SiO2 + H-ZSM-5 was
investigated with ICP-AES. The leached amount in the first use of
Ir-ReOx/SiO2 + H-ZSM-5 was 0.5% of Ir and 0.4% of Re. The amount
Table 2
Hydrogenolysis of glycerol over Ir-ReOx/SiO2 in the presence of acid or metal oxide.
Entry
Additive (amount/mg)
Conv. (%)
Selectivity (%)
pHa
1,3-PrD
1,2-PrD
1-PrOH
2-PrOH
Propane
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
None
33.7
61.1
2.8
46.2
43.2
24.4
26.4
41.0
47.7
42.8
46.7
49.4
55.9
43.4
44.7
62.3
53.0
49.4
59.8
44.4
39.0
25.8
9.2
18.4
36.6
9.0
9.6
9.7
7.2
<0.1
1.3
0.8
0.8
0.7
<0.1
0.8
1.6
1.0
0.8
1.3
1.3
5.9
1.4
1.5
0.5
0.2
1.4
5.7
2.8
10.6
7.2
6.6
6.3
6.4
6.4
5.5
4.8
4.8
4.4
6.1
5.2
4.6
3.4
4.5
3.9
H2SO4 (1.5; H+/Ir = 1)
MgO (10)
58.5
57.0
19.2
19.0
10.7
14.9
9.9
12.4
12.1
9.3
21.0
13.4
12.8
13.3
8.0
CeO2 (10)
ZrO2 (10)
TiO2 (10)
H-mordenite (10)
H-mordenite (50)
H-beta (10)
H-beta (50)
H-ZSM-5 (10)
2.4
7.8
8.0
38.9
39.2
55.8
39.2
54.7
36.9
54.6
58.8
18.6
53.7
56.4
31.5
69.7
69.3
25.9
21.9
34.7
26.1
30.9
22.2
32.0
35.3
6.9
24.6
27.5
17.7
39.4
43.6
13.1
11.4
11.0
10.8
8.8
8.6
11.1
9.4
3.9
7.6
8.8
8.7
H-ZSM-5 (60)
Silica–alumina (as-received, 150)
Silica–alumina (washed, 50)
Silica–alumina (washed, 150)
Ir-ReOx/silica–aluminab
Amberlyst 70 (10)
Amberlyst 70 (50)
8.0
9.0
7.0
PrD, propanediol; PrOH, propanol.
Reaction conditions: glycerol (4 g), water (16 g), Ir-ReOx/SiO2 (150 mg), H2 (8 MPa), 393 K, 24 h.
a
Room temperature pH value of stirred aqueous glycerol solution (20 wt.%, 20 g) added with the metal oxide or acid.
Ir-ReOx catalyst supported on washed silica–alumina (150 mg) was used instead of Ir-ReOx/SiO2.
b

Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
131
Table 3
Reusability of Ir-ReOx/SiO2.
System
Usage times
Conv. (%)
Selectivity (%)
1,3-PrD
1,2-PrD
1-PrOH
2-PrOH
Propane
1
2
3
33.7
11.9
6.6
46.2
45.3
44.0
25.8
25.4
28.2
18.4
17.3
17.8
9.6
9.4
8.8
<0.1
2.7
1.2
Without co-catalyst, reused as recovered
1
2
3
33.7
25.3
21.0
46.2
47.1
45.9
25.8
21.1
24.0
18.4
20.3
18.1
9.6
11.0
11.2
<0.1
0.5
0.9
Without co-catalyst, calcined before reuse
+Amberlyst 70 reused as recovered
1
2
3
69.7
49.9
37.5
44.4
44.7
48.6
8.0
9.4
9.6
39.4
36.1
32.8
8.0
8.5
8.0
0.2
1.4
0.9
1
2
3
69.7
62.9
51.1
44.4
41.4
50.1
8.0
5.4
6.3
39.4
43.0
34.5
8.0
8.7
8.3
0.2
1.4
0.9
+Amberlyst 70 calcined and re-added before
reuse
1
2
3
58.8
36.1
24.9
44.7
50.3
51.2
9.3
12.8
13.9
35.3
27.3
26.1
9.4
8.8
8.8
1.3
0.8
<0.1
+H-ZSM-5 reused as recovered
+H-ZSM-5 calcined before reuse
PrD, propanediol; PrOH, propanol.
1
2
3
58.8
54.2
51.6
44.7
46.4
49.3
9.3
8.4
8.0
35.3
34.0
31.8
9.4
10.0
10.0
1.3
1.1
0.9
Reaction conditions: glycerol (4 g), water (16 g), Ir-ReOx/SiO2 (150 mg), acid (Amberlyst 70: 10 mg; H-ZSM-5: 60 mg), H2 (8 MPa), 393 K, 24 h.
is larger than that in the case in the presence of H2SO4 (<0.3% for
both Ir and Re) [20] and smaller than that in the case without any
acid (2% for Ir and 0.9% for Re) [20]. Therefore, H-ZSM-5 has the
enhancing effect on the stabilization of Ir-ReOx/SiO2 like H2SO4,
although the effect is weaker.
The XRD pattern of used Ir-ReOx/SiO2 + H-ZSM-5 showed the
peaks assignable to SiO2, Ir metal and the MFI-type structure
of H-ZSM-5 (Fig. 1). No peak assignable to IrO2 or rhenium
oxide was observed. The pattern except the peaks of zeolite
structure was almost the same as that of used Ir-ReOx/SiO2 or
Ir-ReOx/SiO2 + H2SO4 [21]. We could not observe the structure
difference in XRD among these catalysts. The TPR profile of Ir-
ReOx/SiO2 after use and subsequent calcination was also similar
regardless of whether acids were added or not in the catalytic test
(Fig. 2). TEM image of Ir-ReOx/SiO2 + H-ZSM-5 after use showed
two types of support particles (Fig. 3). Support particles with small
(2–3 nm) metal particles can be assigned to Ir-ReOx/SiO2 and those
without metal particles can be assigned to H-ZSM-5. Therefore,
the bulk phases of Ir-ReOx/SiO2 catalysts where iridium particles
with 2–3 nm size are covered with partially reduced ReOx cluster
[20,21] are not affected by the presence of an acid. The particle
surface of active components and/or the interface between iridium
and rhenium species is probably affected.
3.3. Effect of reaction conditions
Table 4 shows the H2 pressure dependence of Ir-ReOx/SiO2-
catalyzed glycerol hydrogenolysis combined with H-ZSM-5. The
log(H2 pressure [MPa]) vs. log(conversion) plot gave a line with
a slope of 0.97, indicating the first-order kinetics. The first-order
kinetics has been also observed for Ir-ReOx/SiO2 + H2SO4 sys-
tem [21]. The selectivities to 1,3-PrD and 1,2-PrD were gradually
decreased with increasing H2 pressure and the trends can be
explained by the progress of overhydrogenolysis accompanied by
higher conversion.
Fig. 2. TPR profiles of catalysts after use and subsequent calcination (773 K, 3 h). (a)
Ir-ReOx/SiO2, (b) Ir-ReOx/SiO2 + H2SO4, (c) Ir-ReOx/SiO2 + H-ZSM-5 (150 mg:60 mg).
The reaction conditions were the same as shown in Table 2. H2/Ir: molar ratio of
amount of H2 consumption to Ir content.
Fig. 1. XRD patterns of used catalysts. (a) Ir-ReOx/SiO2, (b) Ir-ReOx/SiO2 + H-ZSM-5
(150 mg:60 mg). The reaction conditions were the same as shown in Table 2.

132
Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
Table 4
Hydrogenolysis of glycerol under H2 of various pressures over Ir-ReOx/SiO2 + H-ZSM-5.
H2 pressure (MPa)
Conv. (%)
Selectivity (%)
1,3-PrD
1,2-PrD
1-PrOH
2-PrOH
Propane
2
3
5
6
7
8
14.6
23.5
35.2
42.7
51.2
58.8
52.2
53.4
49.6
48.2
45.9
44.7
14.0
13.3
12.1
9.9
9.6
9.3
26.0
26.0
29.4
32.8
33.8
35.3
6.0
6.4
8.1
9.1
9.3
9.4
1.7
1.0
0.7
0.1
1.4
1.3
PrD, propanediol; PrOH, propanol.
Reaction conditions: glycerol (4 g), water (16 g), Ir-ReOx/SiO2 (150 mg), H-ZSM-5 (60 mg), H2 (2–8 MPa), 393 K, 24 h.
Table 5
Hydrogenolysis of glycerol with various concentrations over Ir-ReOx/SiO2 + H-ZSM-5.
Entry
Glycerol conc. (wt.%)
Reaction time (h)
Conv. (%)
Selectivity (%)
1,3-PrD
1,2-PrD
1-PrOH
2-PrOH
Propane
1
2
3
4
5
6
7
8
5a
10b
20
40
67
80
80
80
24
24
24
24
24
24
36
48
31.0
46.6
58.8
58.3
57.1
51.6
75.2
79.6
52.1
48.5
44.7
46.7
47.3
53.3
43.9
39.6
14.6
11.9
9.3
7.8
6.3
6.3
4.1
4.3
23.7
29.8
35.3
35.7
36.3
33.5
44.0
47.5
9.1
8.8
9.4
8.7
8.2
6.4
7.2
7.6
0.5
1.1
1.3
1.0
1.9
0.5
0.7
1.0
PrD, propanediol; PrOH, propanol.
Reaction conditions: glycerol (4 g), water (1–16 g), Ir-ReOx/SiO2 (150 mg), H-ZSM-5 (60 mg), H2 (8 MPa), 393 K.
a
Glycerol (1 g), water (19 g), Ir-ReOx/SiO2 (37.5 mg), H-ZSM-5 (15 mg).
Glycerol (2 g), water (18 g), Ir-ReOx/SiO2 (75 mg), H-ZSM-5 (30 mg).
b
Table 5 shows the effect of glycerol concentration on the Ir-
ReOx/SiO2 + H-ZSM-5 system. In this experiment, the total amount
of glycerol was fixed and the concentration of glycerol was adjusted
by the amount of H2O as a solvent. In the cases of very low
concentration (≤10%) of glycerol, all of the amounts of glycerol,
Ir-ReOx/SiO2 catalyst and H-ZSM-5 co-catalyst were lowered to fix
the total amount of the reaction solution. When the range of glyc-
erol concentration was less than 20% (Entries 1–3), the conversion
of glycerol increased with increasing glycerol concentration. The
conversion was almost constant when the glycerol concentration
was above 20% (Entries 3–6). Therefore, the reaction order with
respect to glycerol concentration was zero and the adsorption of
glycerol molecule at the active site was saturated in this concen-
tration range. The selectivity to 1,3-PrD was slightly increased and
that to 1,2-PrD was decreased with increasing glycerol concentra-
tion, indicating that high concentration of glycerol is advantageous
to obtain high 1,3-PrD yield. These trends are the same as those
observed for the Ir-ReOx/SiO2 + H2SO4 system [21]. Longer reaction
time in high glycerol concentration gave maximum 1,3-PrD yield
of 33% (Entry 7; Eq. (1)). Although this value of 1,3-PrD yield is
slightly lower than that obtained in Ir-ReOx/SiO2 + H2SO4 system
(38%) [20], it is still higher than or comparable to the values for
other catalysts reported in the literature.
H2 (8 MPa)
HO
OH
HO
OH
OH
Ir-ReOx/SiO2 150 mg,
H-ZSM-560 mg, 393 K,
36 h
33% (75% conv.)
4 g (80%aq.)
(1)
3.4. Mechanism of the promoting effect of acid addition on the
catalysis of Ir-ReOx/SiO2
As shown above, the addition effect of solid acid to Ir-ReOx/SiO2
is similar to the effect of H2SO4 on the activity, selectivity, and
stability of the catalyst. We have proposed the mechanism of Ir-
ReOx/SiO2-catalyzed glycerol hydrogenolysis [20,21] as follows:
first, substrate (glycerol) is adsorbed on the surface of ReOx clus-
ter at the CH2OH group to form terminal alkoxide. Next, hydride
activated on the Ir metal attacks the 2-position of the alkoxide
to break the C O bond. The hydrolysis of the reduced alkoxide
releases the product. We have also proposed a similar mechanism
for Rh-MOx (M = Mo, Re) catalysts supported on silica or carbon
in the hydrogenolysis of ethers and poly-ols including glycerol
[19,32–37]. Based on this mechanism, changes in the surface struc-
ture of ReOx cluster can change the coordination mode of adsorbed
glycerol, leading different selectivities and activities. Here we pro-
pose that hydroxorhenium site (Re OH) is the adsorption site of
substrates (Eq. (2)). The presence of hydroxo group on the surface
Fig. 3. TEM image of Ir-ReOx/SiO2 + H-ZSM-5 (150 mg:60 mg) after catalytic use.

Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
133
of ReOx cluster has been suggested by characterization results of
Ir-ReOx/SiO2 after the catalytic use with Re L3-edge EXAFS [21].
acid addition. Considering the high selectivity to 1-PrOH in the
hydrogenolysis of 1,2-PrD (Table 1), the adsorption of 1,2-PrD at 2-
position is very minor regardless of whether acid is present or not.
Therefore, the OH group at 3-position and oxorhenium site may
direct the adsorption of glycerol at 2-position as described below.
OH
HO
HO
O
HO
OH
+ H2O
+
Re
Re
OH
OH
(2)
HO
OH
OH
O
O
>
O
O
O
In neutral or basic conditions, hydroxorhenium site can
reversibly release proton to form oxorhenium site.
Re
Re
Re
Re
Re
HO
O
The rise in selectivity to 1-PrOH by acid addition is difficult to
explain. As discussed in Section 3.1, direct route for 1-PrOH for-
mation from glycerol is present and promoted by acid addition.
Overhydrogenolysis of free PrDs also produces 1-PrOH, however,
acid addition does not affect the selectivity to 1-PrOH from this
route because acid addition increases the rates of both glycerol
hydrogenolysis and overhydrogenolysis of PrDs at a constant ratio.
Significant formation of 1-PrOH and 2-PrOH from glycerol at low
conversion has been frequently observed in various hydrogenol-
ysis systems that can give 1,3-PrD as one of the main products
[13,14,16–18,23]. One explanation for the formation of 1-PrOH
is based on the dehydration–hydrogenation mechanism where
acrolein is formed by acid-catalyzed dehydration of glycerol and
subsequently hydrogenated to 1-PrOH (Eq. (5)). However, direct
formation of 2-PrOH cannot be explained by this mechanism since
the precursor of 2-PrOH (acetone) cannot be produced by simple
dehydration of glycerol (Eq. (6)) [27].
+
H+
Re
Re
(3)
Formation of oxorhenium site decreases the number of adsorbed
substrate molecules, leading lower catalytic activity for all sub-
strates. The addition of acid regenerates the hydroxorhenium site
to enhance the activity. Excess amount of acid does not enhance
the activity when all of the oxorhenium sites are converted to
hydroxorhenium sites. On the other hand, if the step of the for-
mation of active hydrogen species is affected by the presence of
acid, the activity should be monotonously increased with increas-
ing acid amount since the coverage of active hydrogen species is
well below saturation as indicated by first-order kinetics on H2
pressure. If the step of the transfer of active hydrogen species to
adsorbed substrates, which is the rate-determining step, is affected
by acid, the activity should be also monotonously increased with
increasing acid amount. The observed dependency of activity on
acid amount in this study and the previous report [21] disagrees
with these two cases. In addition, the dehydration–hydrogenation
mechanism, which is frequently proposed for the hydrogenoly-
sis of glycerol to 1,2-PrD in acidic conditions, also contradicts the
experimental data: The first-order kinetics on H2 pressure and the
zero-order kinetics on acid amount mean that the latter hydro-
genation step is rate-determining. However, in fact, hydrogenation
of keto compounds such as acetol was very fast over Ir-ReOx/SiO2.
It should be noted that dehydration products were not detected in
the reaction mixture.
+2H2
-2H2O
HO
OH
HO
O
OH
(5)
-xH2O
+H2
HO
OH
X
OH
OH
O
The change of selectivity in addition of acid is rather com-
plex. 1,2-PrD can be produced when glycerol is adsorbed at the
2-position (Eq. (4)) [19,21]. The attack of hydrogen species at the
(6)
C
O bond neighboring to the C ORe site gives 1,2-PrD.
Another explanation for the production of 1-PrOH is the overhy-
drogenolysis of 1,3-PrD just after the hydrogenolysis of adsorbed
glycerol and before desorption. Our proposed mechanism for the
production of 1,3-PrD from glycerol involves the formation of 3-
hydroxypropoxyrhenium species as a rate-determining step. Most
3-hydroxypropoxyrhenium species is released from the rhenium
site as 1,3-PrD. If the nascent 3-hydroxypropoxyrhenium species is
reacted with another hydrogen species or acid, the remaining OH
group is lost and 1-PrOH is formed (Eq. (7)). The direct formation of
2-PrOH from glycerol is explained by the overhydrogenolysis of 1-
hydroxyisopropoxyrhenium species just after the hydrogenolysis
of glycerol adsorbed at the 2-position (Eq. (8)).
OH
HO
HO
O
HO
OH
+ H2O
+
Re
Re
OH
(4)
The decrease of 1,2-PrD selectivity by acid addition means that
adsorption of glycerol at 2-position significantly occurs without
+H2, -H2O
O
+H2, -H2O
HO
hot intermediate
O
HO
O
Re
Re
Re
OH
(7)
(8)
OH
O
HO
+H2, -H2O
HO
O
+H2, -H2O
O
Re
Re
hot
Re
intermediate

134
Y. Nakagawa et al. / Applied Catalysis A: General 433–434 (2012) 128–134
The ratio of initial 1,3-PrD selectivity to 1-PrOH selectivities can be
determined by the rate ratio of the ligand exchange to the over-
hydrogenolysis of activated alkoxide. Experimental data in Table 4
show that the ratio of 1,3-PrD/1-PrOH selectivity is almost unaf-
fected by H2 pressure when the effect of different conversion levels
is considered, suggesting that the active hydrogen species for the
overhydrogenolysis of activated alkoxide is different from that of
the main reaction which shows first-order dependence on H2 pres-
sure. Experimental data in Table 2 also show that acid addition
promotes the 1-PrOH formation, however, the effect is saturated
with excess amount of acid, as similar to the effect on the activity
of main hydrogenolysis. The involvement of the hydroxorhenium
sites in the overhydrogenolysis is one explanation: The acidic
hydroxorhenium site can catalyze the indirect route that consists
of acid-catalyzed dehydration and metal-catalyzed hydrogenation
(Eq. (9)).
of New Energy and Industrial Technology Development Organiza-
tion (NEDO) of Japan, and the Cabinet Office, Government of Japan
through its “Funding Program for Next Generation World-Leading
Researchers”.
References
[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)
7164–7183.
[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.
[6] A.-L. Marshall, P.J. Alaimo, Chem. Eur. J. 16 (2010) 4970–4980.
[7] M. Schlaf, Dalton Trans. (2006) 4645–4653.
[8] C.-H. Zhou, J.N. Beltramini, Y.-X. Fan, G.Q. Lu, Chem. Soc. Rev. 37 (2008)
527–549.
[9] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D. Pina, Angew. Chem. Int. Ed.
46 (2007) 4434–4440.
[10] Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 1 (2011) 179–190.
[11] T.M. Che, US Patent, 4,642,394 (1987).
[12] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green
Chem. 6 (2004) 359–361.
hot intermediate
+H2
O
-H2O
O
HO
O
Re
Re
HO
O
Re
HO
Re
Re
Re
(9)
[13] N. Suzuki, Y. Yoshikawa, M. Takahashi, M. Tamura, WO Patent, 2007129560
(2007).
In order to obtain high yield of 1,3-PrD, it is necessary to improve
the selectivity to 1,3-PrD. The initial selectivity pattern shows that
the formation of 1,2-PrD (ca. 10% selectivity) and direct formations
of 1-PrOH (ca. 20%) and 2-PrOH (ca. 10%) decrease the selectiv-
ity to 1,3-PrD. In the case of Eq. (1) where the highest 1,3-PrD
yield is obtained, overhydrogenolysis of desorbed 1,3-PrD further
decreases the selectivity to 1,3-PrD by ca. 15%. Therefore, the sup-
pression of direct formation of PrOHs is essential in order to obtain
high yield of 1,3-PrD, and the acidity of additive oxide (ReOx) may
be the key. Further investigation in the mechanism of the direct
formation of PrOHs is necessary.
[14] T. Kurosaka, H. Maruyama, I. Naribayashi, Y. Sasaki, Catal. Commun. 9 (2008)
1360–1363.
[15] L. Huang, Y. Zhu, H. Zheng, G. Dong, Y. Li, Catal. Lett. 131 (2009)
312–320.
[16] L.-Z. Qin, M.-J. Song, C.-L. Chen, Green Chem. 12 (2010)
1466–1472.
[17] O.M. Daniel, A. DeLaRiva, E.L. Kunkes, A.K. Datye, J.A. Dumesic, R.J. Davis, Chem-
CatChem 2 (2010) 1107–1114.
[18] L. Gong, Y. Lu, Y. Ding, R. Lin, J. Li, W. Dong, T. Wang, W. Chen, Appl. Catal. A:
Gen. 390 (2010) 119–126.
[19] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B: Environ.
94 (2010) 318–326.
[20] Y. Nakagawa, Y. Shinmi, S. Koso, K. Tomishige, J. Catal. 272 (2010)
191–194.
[21] Y. Amada, Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal.
B: Environ. 105 (2011) 117–127.
4. Conclusions
[22] J. Oh, S. Dash, H. Lee, Green Chem. 13 (2011) 2004–2007.
[23] S. Zhu, Y. Zhu, S. Hao, L. Chen, B. Zhang, Y. Li, Catal. Lett. 142 (2012)
267–274.
[24] F. Locatelli, B. Didillon, D. Uzio, G. Niccolai, J.P. Candy, J.M. Basset, J. Catal. 193
(2000) 154–160.
Hydrogenolysis of glycerol over Ir-ReOx/SiO2 is promoted by the
addition of various types of acids such as zeolites, silica–alumina,
ion-exchange resin and sulfuric acid. Among the solid acids tested,
Amberlyst 70 is most effective to enhance the activity although
Amberlyst 70 is burned during calcination in the recycling process.
Considering the reusability, H-ZSM-5 is a more suitable co-catalyst.
The property of Ir-ReOx/SiO2 + H-ZSM-5 system including kinetics
and selectivity trends in various reaction conditions is similar to the
case of Ir-ReOx/SiO2 + H2SO4. The maximum 1,3-PrD yield is 33%,
which is slightly lower than that obtained with H2SO4 co-catalyst
which has been reported earlier. Supporting iridium and rhenium
on silica–alumina decreased the activity, while silica–alumina can
be used as a co-catalyst with Ir-ReOx/SiO2. The addition of acid does
not affect the bulk phases of Ir-ReOx/SiO2 catalyst. The proposed
role of added acid is to protonate the surface of ReOx cluster to
increase the number of hydroxorhenium site, which is responsible
for the activation of glycerol by the formation of glyceride.
[25] C.-W.
Chiu,
M.A.
Dasari,
G.J.
Suppes,
AIChe
J.
52
(2006)
3543–3548.
[26] S. Sato, M. Akiyama, R. Takahashi, T. Hara, K. Inui, M. Yokota, Appl. Catal. A: Gen.
347 (2008) 186–191.
[27] Y.T. Kim, K.-D. Jung, E.D. Park, Appl. Catal. A: Gen. 393 (2011)
275–287.
[28] Y. Kusunoki, T. Miyazawa, K. Kunimori, K. Tomishige, Catal. Commun. 6 (2005)
645–649.
[29] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 240 (2006)
213–221.
[30] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal. A: Gen. 318 (2007)
244–251.
[31] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal. A: Gen. 329 (2007)
30–35.
[32] S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori, K. Tomishige, Chem.
Commun. (2009) 2035–2037.
[33] S. Koso, N. Ueda, Y. Shinmi, K. Okumura, T. Kizuka, K. Tomishige, J. Catal. 267
(2009) 89–92.
[34] K. Chen, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, ChemCatChem 2 (2010)
547–555.
[35] Y. Amada, S. Koso, Y. Nakagawa, K. Tomishige, ChemSusChem
728–736.
3 (2010)
Acknowledgment
[36] S. Koso, Y. Nakagawa, K. Tomishige, J. Catal. 280 (2011) 221–229.
[37] S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, Appl. Catal. B:
Environ. 111/112 (2012) 27–37.
This research is supported by the JSPS KAKENHI (23760737), the
Industrial Technology Research Grant Program (No. 08B40001c)