Catalytic conversion of levulinic acid and its esters to γ-valerolactone over silica-supported zirconia catalysts

Article abstract of DOI:10.1246/bcsj.20140205

Levulinic acid and its esters, important intermediate chemicals produced from biomass resources, were converted to γ-valerolactone via a catalytic transfer hydrogenation process over silica-supported zirconia catalysts. Silica support with a higher surface area provided highly dispersed active zirconium oxide species, which exhibited superior catalytic activity to the conventional ZrO₂ catalyst.

Full text of DOI:10.1246/bcsj.20140205

Short Articles
Table 1. Conversion
of
Methyl
Levulinate
to
γ-Valerolactone with 2-Propanola)
Catalytic Conversion of
Levulinic Acid and Its Esters
to γ-Valerolactone over Silica-
Supported Zirconia Catalysts
b)
ZrO2 SBET
T
/°C
Conv.c) Yieldc)
Entry Catalyst
/wt % /m2 g
/%
/%
¹1
1
2
3
4
5
6
7
8
9
ZrO2/SiO2
ZrO2/MCM-41
ZrO2/SBA-15
9.6
9.7
9.8
9.8
9.8
19.4
49.7
10.0
10.0
100
®
250
783
863
863
863
753
597
211
91
150
150
90
93
98
6.9
59
98
93
54
96
61
45
79
88
5.5
49
88
84
46
83
48
42
Yasutaka Kuwahara,*1,2 Wako Kaburagi,1
and Tadahiro Fujitani1
120
150
150
150
150
150
150
150
1Research Institute for Innovation in Sustainable Chemistry,
National Institute of Advanced Industrial Science
and Technology (AIST), 1-1-1 Higashi,
ZrO2/γ-Al2O3
ZrO2/TiO2
Tsukuba, Ibaraki 305-8565
10 ZrO2
11 SBA-15
112
1040
2Division of Materials and Manufacturing Science,
Graduate School of Engineering, Osaka University,
2-1 Yamada-oka, Suita, Osaka 565-0871
no reaction
a) Reaction conditions: catalyst (40 mg as ZrO2), methyl
levulinate (2 mmol), 2-PrOH (10 mL), Ar (1.0 MPa), 2 h.
b) Determined by BET method from N2 adsorption isotherms.
c) Determined by GC.
E-mail: kuwahara@mat.eng.osaka-u.ac.jp
Received: July 16, 2014; Accepted: July 30, 2014;
Web Released: November 15, 2014
improved dispersibility of active Zr sites and thus afford
increased catalytic activities.10,11 In particular, mesoporous
materials having large surface areas are expected to be suitable
supports for creating highly-dispersed active Zr sites.12
In this work, we report a CTH process for the production of
GVL from LA and its esters using several alcohols as H-donors
over silica-supported Zr oxide catalysts, which are cheap and
readily prepared. Optimization of support type, catalyst com-
position, and reaction conditions was performed, and the scope
of substrates and alcohols was also investigated.
Levulinic acid and its esters, important intermediate
chemicals produced from biomass resources, were converted
to γ-valerolactone via a catalytic transfer hydrogenation
process over silica-supported zirconia catalysts. Silica sup-
port with a higher surface area provided highly dispersed
active zirconium oxide species, which exhibited superior
catalytic activity to the conventional ZrO2 catalyst.
In the initial studies, several ZrO2 catalysts supported on
various oxide supports were prepared and tested in the CTH
reaction of methyl levulinate with 2-PrOH (Table 1). Supported
ZrO2 catalysts were prepared by in situ hydrolysis of zirconium
n-butoxide (Zr(OnBu)4, 70% in n-butanol) on oxide support
in an organic solvent, followed by filtration, washing, drying,
and calcination. The reaction was carried out using a stainless
autoclave reactor under 1.0 MPa of Ar. As shown in Table 1,
GVL was synthesized from methyl levulinate over various
ZrO2 catalysts. Transesterification compounds (i.e., isopropyl
levulinate ester and 4-hydroxypentanoic acid isopropyl ester)
were produced as main by-products, and acetone and methanol
were also detected, suggesting that the reaction proceeds via a
CTH reaction and the following dealcoholation steps. Sup-
ported ZrO2 catalysts showed higher catalytic activities than
bulk ZrO2, although the selectivity to GVL slightly decreased
(Entries 1 and 8-10). ZrO2 supported on γ-Al2O3 (SBET = 211
Increasing concern over the depletion of fossil fuel resources
has necessitated the use of biomass as renewable, natural re-
sources for the sustainable production of fuels and chemicals.1
Among various chemicals synthesized from lignocellulosic
biomass, γ-valerolactone (GVL) is considered one of the most
appealing molecules which can be used as an ideal precursor
for the production of olefins, fuels, polymers, and other
valuable chemicals.2 GVL can conventionally be produced by
hydrogenation of cellulosic biomass-derived levulinic acid
(LA) and its esters over precious metal catalysts such as Ru,3
Pd,4 and Ir5 under the presence of high-pressure H2.6
Catalytic transfer hydrogenation (CTH) is another viable
option for selective transformation of LA and its esters to GVL,7
because the reaction requires alcohols as H-donors instead
of high-pressure and flammable H2 gas and can be catalyzed
by nonprecious metal catalysts. Chia and Dumesic reported a
CTH process to convert LA and its esters into GVL using
bulk ZrO2 and 2-propanol/2-methyl-1-propanol as H-donors at
150 °C under pressurized inert gas conditions (300 psig He,
1 Pa = 1.45037738007 © 10¹4 psig).8 However, such a system
still requires high-temperature and -pressure conditions and the
catalyst loses its catalytic activity during the reaction.9 Previous
literature regarding the CTH process using Zr-based hetero-
geneous catalysts have claimed that oxide supports provide
m2 g¹1) showed as high an activity as that on SiO2 (SBET
=
250 m2 g¹1), whereas ZrO2 supported on TiO2 (SBET = 91
m2 g¹1) gave a modest catalytic activity. A further improve-
ment of catalytic activity was achieved by enlarging surface
area of silica support; ZrO2 supported on typical hexagonal
mesoporous silica, such as MCM-41 and SBA-15, both
afforded 88% yield of GVL with 98% conversion at 150 °C
for 2 h (Entries 2 and 5), while silica support itself is inactive
1252 | Bull. Chem. Soc. Jpn. 2014, 87, 1252–1254 | doi:10.1246/bcsj.20140205
© 2014 The Chemical Society of Japan

Table 2. EXAFS Curve-Fitting Parameters of ZrO2,
ZrO2/SiO2, and ZrO2/SBA-15a)
(A)
(B)
Monoclinic ZrO2
Monoclinic ZrO2
Sample
Phaseb)
Shell
C.N.c)
R/¡d)
ZrO2
monoclinic
Zr-O
Zr-Zr
Zr-O
Zr-Zr
Zr-O
Zr-Zr
Zr-O
Zr-Zr
Zr-O
Zr-Zr
7
7
2.160
3.460
2.122
3.431
2.121
3.401
2.121
3.404
2.128
3.693
ZrO2/SiO2 (9.6 wt %)
ZrO2/SiO2
(9.6 wt %)
ZrO2/SiO2
(9.6 wt %)
amorphous
amorphous
amorphous
5.5
1.6
4.5
0.9
5.1
1.0
7.1
3.2
ZrO2/SBA-15 (9.8 wt %)
ZrO2/SBA-15
(9.8 wt %)
ZrO2/SBA-15
(9.8 wt %)
ZrO2/SBA-15
(19.4 wt %)
ZrO2/SBA-15
(49.7 wt %)
ZrO2/SBA-15 (19.4 wt %)
ZrO2/SBA-15 (49.7 wt %)
ZrO2/SBA-15
(19.4 wt %)
ZrO2/SBA-15
(49.7 wt %)
17950
18000
18050
18100
0
1
2
3
4
5
amorphous
+ tetragonal
r / Å
Energy / eV
a) k3-weighted EXAFS, 3.0 ¯ k(¡¹1) ¯ 13.0. b) Determined
by XRD. c) Average coordination number. d) Average inter-
atomic distance.
Figure 1. (A) Zr K-edge XANES spectra and (B) Zr K-edge
radial distribution functions of monoclinic ZrO2, ZrO2/
SiO2, and ZrO2/SBA-15 with varied ZrO2 content.
100
for this reaction (Entry 11). The pore size of mesoporous silica
had little effect on the catalytic activity (For reaction kinetics
data, see Figure S1). These results clearly indicate that surface
area of oxide support alters the local structure of Zr species and
consequently affects the catalytic activity.
Optimization of catalyst composition and reaction conditions
was performed on ZrO2/SBA-15 catalyst (Table 1). Reaction
rate in the CTH process increased as the reaction temperature
increased (Entries 3-5). On the other hand, increasing ZrO2
content from 9.8 to 49.7 wt % led to an appreciable reduction of
catalytic activity (Entries 5-7), despite retaining most of its
surface area. This result also suggests a strong link between the
dispersion state of Zr species and the catalytic activity.
ZrO2/SBA-15 (9.8 wt %)
80
60
40
20
0
bulk ZrO2
1st
2nd
3rd
4th
5th
Number of cycles
Figure 2. Catalyst reusability of ZrO2/SBA-15 and bulk
ZrO2. Reaction conditions are the same as those shown in
Table 1.
To obtain insights to the local structure of Zr species sup-
ported on silica, X-ray diffraction (XRD) and X-ray absorption
fine structure (XAFS) measurements were performed. In the
XRD patterns of the catalysts, no crystalline zirconia phase
was identified below 20 wt % of ZrO2 loading levels, but a
tetragonal ZrO2 phase was observed at 49.7 wt % of ZrO2
loading (Figure S2). The Zr K-edge X-ray absorption near edge
structure (XANES) spectra of the supported ZrO2 samples
resembled to that of ZrO2, suggesting that Zr species exist on
silica surface as zirconium(IV) oxides (Figure 1A). In the radial
distribution functions, the peaks corresponding to Zr-O and Zr-
Zr bonds were observed at around r = 1.6 and 3.0 ¡ (the phase
shift was uncorrected), respectively (Figure 1B). Curve-fitting
analysis of the signals showed that the average coordination
number (C.N.) of the first Zr-O shell was 4.5-5.5 at low ZrO2
content, but was 7.1 at 49.7 wt % ZrO2 content, which is similar
to that of bulk ZrO2 (Table 2), suggesting a different coordi-
nation state of Zr.13 The FT magnitude of the Zr-Zr shell signal
for the supported ZrO2 samples was considerably damped
compared to that for bulk ZrO2. As for 9.8 wt % ZrO2 loaded
SBA-15, the interatomic distance between two zirconium atoms
was calculated to be 3.40 ¡ and the C.N. was 0.9, suggesting
the presence of highly dispersed Zr oxide species. As the
ZrO2 content increased from 9.8 to 49.7 wt %, the C.N. for the
Zr-Zr shell became higher (0.9 ¼ 3.2), clearly evidencing the
formation of agglomerated ZrO2 species at higher ZrO2 level.
It should be also noted that ZrO2/SBA-15 (49.7 wt %) sam-
ple exhibited the Zr-Zr shell signal at r = 3.3 ¡ (R = 3.69 ¡)
which corresponds to that of tetragonal ZrO2 (C.N. = 12, R =
3.63 ¡),14 being consistent with the XRD result. The estimated
C.N., rather than the surface area, provided a more distinct
relationship toward the catalytic activity (Figure S3). These
combined analyses suggest that high-surface-area silica support
provides a surface environment suitable for generating highly-
dispersed zirconium oxide species with a low-coordination state
which is highly active for the CTH reaction.
To evaluate catalyst reusability, ZrO2/SBA-15 (9.8 wt %)
and bulk ZrO2 catalysts were used 5 times repeatedly for
the CTH reaction (Figure 2).15 The supported ZrO2 catalyst
retained more than 70% of its initial activity even after 4
repeated cycles, while unsupported ZrO2 lost most of its
activity during the same number of cycles. The observed loss of
catalytic activity is due to strong adsorption of organic residues
on catalyst surface, which inhibits the access of organic sub-
strate to the Zr active sites (Figure S4).
Table 3 shows the scope of substrates and alcohols for the
production of GVL over ZrO2/SBA-15 (9.8 wt %) catalyst.
Among various alcohols, secondary alcohols such as 2-butanol
and cyclohexanol worked as efficient H-donors (Entries 4-6).
Primary alcohols such as methanol, ethanol, and 1-propanol
gave quite low GVL yields (Entries 1-3), mainly affording
transesterification products. When 2-propanol was used as a
Bull. Chem. Soc. Jpn. 2014, 87, 1252–1254 | doi:10.1246/bcsj.20140205
© 2014 The Chemical Society of Japan | 1253

Table 3. Scope of Substrates and Alcohols for the Produc-
tion of γ-Valerolactone over ZrO2/SBA-15 (9.8 wt %)
Catalysta)
References
1
For reviews, see: a) D. M. Alonso, J. Q. Bond, J. A.
Dumesic, Green Chem. 2010, 12, 1493. b) J. J. Bozell, G. R.
Petersen, Green Chem. 2010, 12, 539. c) J. N. Chheda, G. W.
Huber, J. A. Dumesic, Angew. Chem., Int. Ed. 2007, 46, 7164.
d) A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411.
e) J. C. Serrano-Ruiz, R. Luque, A. Sepúlveda-Escribano, Chem.
Soc. Rev. 2011, 40, 5266.
2
For example, see: a) J. Q. Bond, D. M. Alonso, D. Wang,
Conv.d)
/%
Yieldd)
/%
Entry
Substratec)
Alcohol
R. M. West, J. A. Dumesic, Science 2010, 327, 1110. b) D. M.
Alonso, S. G. Wettstein, J. A. Dumesic, Green Chem. 2013, 15,
584. c) M. J. Climent, A. Corma, S. Iborra, Green Chem. 2014, 16,
516. d) J. C. Serrano-Ruiz, D. Wang, J. A. Dumesic, Green Chem.
2010, 12, 574.
1b)
2
3
4
5
6
7
8
9
ML
ML
ML
ML
ML
ML
EL
EL
BL
LA
MeOH
EtOH
1-PrOH
2-PrOH
2-BuOH
CyOH
4.7
87
88
98
95
97
10.4
86
68
99
1.2
19
31
88
79
88
9.7
81
58
83
3
a) M. G. Al-Shaal, W. R. H. Wright, R. Palkovits, Green
Chem. 2012, 14, 1260. b) D. J. Braden, C. A. Henao, J. Heltzel,
C. C. Maravelias, J. A. Dumesic, Green Chem. 2011, 13, 1755.
c) L. Deng, J. Li, D.-M. Lai, Y. Fu, Q.-X. Guo, Angew. Chem., Int.
Ed. 2009, 48, 6529. d) A. M. R. Galletti, C. Antonetti, V. De Luise,
M. Martinelli, Green Chem. 2012, 14, 688. e) J. M. Nadgeri, N.
Hiyoshi, A. Yamaguchi, O. Sato, M. Shirai, Appl. Catal., A 2014,
470, 215.
EtOH
2-PrOH
2-PrOH
2-PrOH
10
a) Reaction conditions: catalyst (40 mg as ZrO2), substrate
(2 mmol), alcohol (10 mL), Ar (1.0 MPa), 150 °C, 2 h. b) At
90 °C. c) ML: methyl levulinate, EL: ethyl levulinate, BL:
n-butyl levulinate, LA: levulinic acid. d) Determined by GC.
4
K. Yan, T. Lafleur, G. Wu, J. Liao, C. Ceng, X. Xie, Appl.
Catal., A 2013, 468, 52.
5
J. Deng, Y. Wang, T. Pan, Q. Xu, Q.-X. Guo, Y. Fu,
ChemSusChem 2013, 6, 1163.
6
1657.
7
W. R. H. Wright, R. Palkovits, ChemSusChem 2012, 5,
H-donor, GVL could be synthesized from ethyl levulinate (EL),
butyl levulinate (BL), and levulinic acid (LA) (Entries 8-10).
Moreover, GVL could be produced from EL by using EtOH
as a H-donor, although the reaction rate is considerably low
(Entry 7). In an ideal biomass conversion process, levulinate
esters are produced by alcoholysis of sugars, which results in a
mixture of alcohols and the corresponding levulinate esters.16
Expanding the scope of substrates and alcohols would open
up a simple and low-cost way to produce GVL, because, if
the alcohol media could directly be used as H-donors in CTH
process, several energy-consuming steps associated with iso-
lation and purification can be skipped.
In summary, GVL, an important feedstock chemical, was
produced from biomass-derived levulinic acid and its esters
using alcohols as H-donors via a CTH process over supported
ZrO2 catalysts, which are cheap and readily prepared. In partic-
ular, ZrO2 supported on high-surface-area silica was demon-
strated to be a more active and reusable catalyst compared to
the conventional bulk ZrO2.
R. S. Assary, L. A. Curtiss, J. A. Dumesic, ACS Catal.
2013, 3, 2694.
8
9
M. Chia, J. A. Dumesic, Chem. Commun. 2011, 47, 12233.
a) X. Tang, L. Hu, Y. Sun, G. Zhao, W. Hao, L. Lin, RSC
Adv. 2013, 3, 10277. b) X. Tang, H. Chen, L. Hu, W. Hao, Y. Sun,
X. Zeng, L. Lin, S. Liu, Appl. Catal., B 2014, 147, 827.
10 a) K. Inada, M. Shibagaki, Y. Nakanishi, H. Matsushita,
Chem. Lett. 1993, 1795. b) S. Nishiyama, M. Yamamoto, H.
Izumida, S. Tsuruya, J. Chem. Eng. Jpn. 2004, 37, 310.
11 a) S. H. Liu, S. Jaenicke, G. K. Chuah, J. Catal. 2002, 206,
321. b) Y. Zhu, S. Jaenicke, G. K. Chuah, J. Catal. 2003, 218, 396.
c) Y. Zhu, G. Chuah, S. Jaenicke, J. Catal. 2004, 227, 1. d) J.
Wang, S. Jaenicke, G.-K. Chuah, RSC Adv. 2014, 4, 13481.
12 For reviews, see: a) K. Ariga, A. Vinu, Y. Yamauchi, Q. Ji,
J. P. Hill, Bull. Chem. Soc. Jpn. 2012, 85, 1. b) W. Chaikittisilp,
K. Ariga, Y. Yamauchi, J. Mater. Chem. A 2013, 1, 14. c) X. Qian,
K. Fuku, Y. Kuwahara, T. Kamegawa, K. Mori, H. Yamashita,
ChemSusChem 2014, 7, 1528.
13 For the curve-fitting analysis, the phase shift and amplitude
functions for Zr-O and Zr-Zr were extracted from the data for
monoclinic ZrO2.
14 F. Meneghetti, E. Wendel, S. Mascotto, B. M. Smarsly,
E. Tondello, H. Bertagnolli, S. Gross, CrystEngComm 2010, 12,
1639.
This work was financially supported by Research Institute
for Innovation in Sustainable Chemistry in the National Insti-
tute of Advanced Industrial Science and Technology (AIST).
The synchrotron radiation experiments were performed at the
BL01B1 beam line at SPring-8 with the approval of JASRI
(No. 2014A1045).
15 After each catalytic run, catalysts were retrieved from the
reaction mixture by filtration, washed with acetone, dried at
100 °C, and then subjected to next catalytic run.
16 a) Y. Kuwahara, W. Kaburagi, K. Nemoto, T. Fujitani, Appl.
Catal., A 2014, 476, 186. b) Y. Kuwahara, T. Fujitani, H.
Yamashita, Catal. Today 2014, 237, 18. c) K. Tominaga, A. Mori,
Y. Fukushima, S. Shimada, K. Sato, Green Chem. 2011, 13, 810.
Supporting Information
Reaction kinetics data, XRD patterns, activity-structure
relationships, and thermogravimetric measurement data. These
materials are available electronically on J-STAGE.
1254 | Bull. Chem. Soc. Jpn. 2014, 87, 1252–1254 | doi:10.1246/bcsj.20140205
© 2014 The Chemical Society of Japan