Conversion of 5-hydroxymethylfurfural to a cyclopentanone derivative by ring rearrangement over supported Au nanoparticles

Article abstract of DOI:10.1039/c3cc49591d

Supported Au nanoparticles showed efficient catalytic performance for the ring rearrangement of 5-hydroxymethylfurfural (HMF) to a cyclopentanone derivative, 3-hydroxymethylcyclopentanone (HCPN), by taking advantage of the selective hydrogenation on Au nanoparticles and the Lewis acid catalysis of metal oxide supports. Among various metal oxide supported Au catalysts, the highest yield of HCPN was obtained by using Au/Nb₂O₅ (86% yield). the Partner Organisations 2014.

Full text of DOI:10.1039/c3cc49591d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Conversion of 5-hydroxymethylfurfural to a
cyclopentanone derivative by ring rearrangement
over supported Au nanoparticles†
Junya Ohyama,*ab Ryusuke Kanao,a Akihiko Esakia and Atsushi Satsuma*ab
Cite this: Chem. Commun., 2014,
50, 5633
Received 18th December 2013,
Accepted 26th March 2014
DOI: 10.1039/c3cc49591d
www.rsc.org/chemcomm
Supported Au nanoparticles showed efficient catalytic performance
for the ring rearrangement of 5-hydroxymethylfurfural (HMF) to a
cyclopentanone derivative, 3-hydroxymethylcyclopentanone (HCPN),
by taking advantage of the selective hydrogenation on Au nano-
particles and the Lewis acid catalysis of metal oxide supports. Among
various metal oxide supported Au catalysts, the highest yield of HCPN
was obtained by using Au/Nb2O5 (86% yield).
Biorefining is being intensely investigated as a key technology
for a sustainable future. In the biorefinery process, 5-hydroxy-
methylfurfural (HMF), which can be synthesized by dehydration of
hexoses, is regarded as one of the most promising precursors to
produce valuable chemicals and fuels.1–3 For example, by an
oxidation reaction, HMF is converted to 2,5-furandicarboxylic acid
which is a potential alternative to terephthalic acid.4 Hydrogenation
Scheme 1 Hydrogenation and ring rearrangement of HMF to HCPN.
and/or hydrogenolysis of HMF give 2,5-bis(hydroxymethyl)furan considered to proceed via selective hydrogenation of an aldehyde to
(BHF), 2,5-bis(hydroxymethyl)tetrahydrofuran (BHTF), 1,6-hexandiol, an alcohol and the following ring opening and closing reactions
and 2,5-dimethylfuran (DMF), etc., which can be used for synthesizing (Scheme 1). It is well known that platinum group metal catalysts
polymers, fine chemicals, and fuels.5–9 Aldol condensation of show high hydrogenation activity but low selectivity towards hydro-
HMF with acetone on an acid–base catalyst and the following genation of a multi-functional compound.17,18 The low selectivity
hydrogenation and hydrogenolysis provide jet fuel.10 In a similar would decrease the preferential synthesis of cyclopentanone from
fashion, furfural synthesized by dehydration of pentoses can also furfural due to the hydrogenation of the furan ring. Yang et al.
be converted to valuable compounds. Very recently, the ring reported that the Ni–Cu bimetallic catalyst, which is known for the
rearrangement of furfural has attracted increasing attention as a new selective hydrogenation of furfural to furfuryl alcohol in gas-phase,
transformation of biomass derived materials to useful chemicals.11–14 exhibited the ring rearrangement reaction of furfural to give high
Hronec et al. reported the rearrangement of the furan ring of furfural yield of cyclopentanone (62%) in aqueous solution; however, the
to cyclopentanone over supported platinum group metal catalysts Ni–Cu catalyst did not convert HMF to a cyclopentanone derivative
(Pt, Pd, and Ru), and carbon supported Pt (Pt/C) provided the highest at all.12 The Au catalyst is also known for the selective hydrogenation
yield of cyclopentanone.15,16 The rearrangement of the furan ring is of unsaturated aldehydes.18–21 In addition to the selective hydro-
genation reaction, acid–base properties of supports can promote the
ring opening and closing reactions.5,6 Thus, Au supported on metal
a Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan.
oxides with acid–base properties is expected to show efficient
performance for the ring rearrangement reaction. However, to our
knowledge, there is no report on the ring rearrangement reaction
over supported Au catalysts as well as the effect of supports on the
ring rearrangement reaction.
E-mail: ohyama@apchem.nagoya-u.ac.jp
b Elements Strategy Initiative for Catalysts and Batteries (ESICB), Kyoto University,
Katsura, Kyoto 615-8520, Japan
† Electronic supplementary information (ESI) available: Experimental proce-
dures; ICP analysis of Au catalysts; NMR data of HCPN, HCPEN, HHD, and
BHF; TEM images of Au/Nb2O5 before and after the reaction; Au L3-edge EXAFS
and XANES spectra of Au catalysts; and results of curve fitting analysis of the
EXAFS spectra. See DOI: 10.1039/c3cc49591d
Previously, we demonstrated that Au/Al2O3 showed high
catalytic activity and selectivity for the hydrogenation of HMF
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 5633--5636 | 5633

View Article Online
Communication
ChemComm
to BHF (496% yield).5 The basic supports did not facilitate the
ring opening reaction to provide BHF with high selectivity. On
the other hand, the acidic supports promoted the ring-opening
reaction to produce 1-hydroxyl-2,5-hexanedione (HHD), which has a
potential to form ring compounds through intramolecular aldol
condensation. Herein, we investigated the ring rearrangement of
HMF over supported Au catalysts by taking advantage of the
selective hydrogenation over Au and the acid–base properties of
supports. The present study first demonstrates the efficient ring
rearrangement of HMF to a cyclopentanone derivative, 3-hydroxy-
methylcyclopentanone (HCPN), which can be an intermediate
chemical for production of fragrances, pesticides, and polymers.
Supported Au catalysts were prepared by the deposition
precipitation method using HAuCl4 as a precursor containing
10 mg of Au and 0.99 g of various metal oxide supports (Nb2O5,
Al2O3, ZrO2, TiO2, SiO2–Al2O3 (SA), La2O3, CeO2, and hydrotalcite
(HT)).22 The catalysts were calcined at 573 K under air for 3 h, and
treated with H2 for 1 h and then O2 for 0.5 h at 473 K prior to the
reaction. This pretreatment is different from that for the prepara-
tion of Au subnanoparticles which provided a high BHF yield in
the previous study. The pretreatment will increase rather than
decrease the particle size; however, such small Au particles were
not required for the ring rearrangement reaction as described
below. The catalytic reaction was performed under the following
conditions: 10 mg of supported gold catalysts; 3 mL of 0.067 M
HMF aqueous solution; 8 MPa of H2; 413 K. In this study, we
identified 4-hydroxy-4-hydroxymethyl-2-cyclopentenone (HCPEN)
as an intermediate for production of HCPN.
Table 1 Results for conversion of HMF into HCPN in the presence of
various metal oxides supported Au catalysts and Nb2O5 supported Pt, Pd,
and Rua catalysts
Yieldb (%)
Conversion
Entry Catalyst
(%)
HCPN HHD BHF HCPEN
1
2
3
4
5
6
7
8
Au/Nb2O5
Au/Al2O3
Au/ZrO2
Au/TiO2
Au/SA
Au/La2O3
Au/CeO2
Au/HT
499
499
499
499
92
89
92
86
72
62
54
53
1
1
0
84
1
2
0
0
0
28
43
66
82
7
3
1
4
12
3
2
1
10
3
5
0
0
0
0
7
2
5
0
0
0
0
1
0
0
16
0
5
3
0
0
0
0
0
0
0
0
0
0
0
0
13
9
13
0
8
9
0
0
0
0
0
0
0
98
9c
10d
11e
12
13
14 f
15
16
17
18g
Recycled Au/Nb2O5 499
Au/Nb2O5
Recycled Au/Nb2O5
Blank
37
40
10
11
24
499
499
499
Nb2O5
Au/Nb2O5
Pt/Nb2O5
Pd/Nb2O5
Ru/Nb2O5
Au/Nb2O5 without 499
O2 treatment
Pt/Nb2O5 without
O2 treatment
19g
20h
a
499
4
1
27
60
0
0
0
6
Au/Nb2O5
+
81
8.5 mM H3PO4
Reaction conditions: catalyst 10 mg, 0.067 M HMF aqueous solution 3 mL,
H2 8 MPa, 413 K, 12 h. b Yield: (mol product)/(mol initial HMF) Â 100,
HMF: 5-hydroxymethylfurfural, HCPN: 3-hydroxymethylcyclopentanone,
HHD: 1-hydroxyl-2,5-hexanedione, BHF: 2,5-bishydroxymethylfuran,
c
HCPEN: 4-hydroxy-4-hydroxymethyl-2-cyclopentenone. Au/Nb2O5 was
d
recycled by washing with acetone and water, and drying at 353 K. Reaction
time: 1 h. e Recycle test. Reaction time: 1 h. f The catalytic test was
performed under N2 without H2. g The catalysts were pretreated with H2
but without O2. h The catalytic test was carried out in the presence of
8.5 mM H3PO4.
Table 1 lists the results of the ring rearrangement reaction
over the various metal oxide supported Au catalysts (entries 1–8).
The highest yield of HCPN was obtained on Au/Nb2O5 among the
Au catalysts (86%). Au/Nb2O5 was reusable without any decrease in
the initial conversion of HMF and the yield of HCPN (Table 1
The time course of the conversion of HMF to the product
entries 1 and 9–11). No decrease in the initial conversion of HMF yields on Au/Nb2O5 is shown in Fig. 1. As HMF was consumed,
(Table 1 entries 10 and 11) suggests that the gold nanoparticles of the yields of BHF and HCPEN increased for ca. 1 h, and HHD
Au/Nb2O5 do not grow during the reaction, and their hydrogenation for ca. 6 h. The yields of these intermediates decreased and the
activity does not decrease. Actually, particle size analysis by TEM yield of HCPN increased up to 86% for 12 h. On the basis of
showed that the size of gold nanoparticles did not change during these results, we proposed that HCPN is produced by a stepwise
the reaction for 12 h (Fig. S1, ESI†). The ICP analysis (Table S1, ESI†) reaction as shown in Scheme 1. The selective hydrogenation
indicated that the Au loading of Au/Nb2O5 did not change after the over Au gives BHF. The furan ring of BHF opens on acid sites of
reaction (0.56 wt% of Au loading). The results indicate no leaching the supports to provide 1-hydroxy-3-hexene-2,5-dione (HHED)
of gold during the reaction.
(not detected). Hydrogenation of HHED produces HHD, and
The Au supported on Al2O3, ZrO2, TiO2, and SA also gave intramolecular aldol condensation of HHD over supports and
HCPN with relatively high yield. These metal oxide supports the following hydrogenation generate HCPN. On the other
including Nb2O5 are conventionally known to have acidic hand, intramolecular aldol reaction of HHED also proceeds to
properties. On the other hand, the Au supported on basic give HCPEN. The resulting HCPEN is dehydrated and hydro-
supports, CeO2, La2O3, and HT, provided ring rearranged pro- genated to produce HCPN.12,23 In this scheme, the acidity of the
ducts (HCPN and HCPEN) in much lower yields in comparison supports first works for the ring opening reaction, and then the
with the acid metal oxide supported Au catalysts. The results aldol and dehydration reactions. As shown above, the basic
suggest that the acid on metal oxide supports promotes the ring- metal oxides supported on Au catalysts gave low yields of the
opening and closing reaction (hydration and dehydration). It ring rearrangement products (HCPEN and HCPN). This could
should also be mentioned that, in our previous study, Au/Al2O3 be due to low catalytic activity of basic metal oxides for the ring
provided BHF, but, in the present study, gave HCPN with high opening of BHF. In fact, it has been reported that the ring
yield. This is because the longer reaction time in this study opening of BHF proceeds on acidic supports, but does not on
caused the ring opening and closing on Al2O3 which has not only basic supports.5,6 Therefore, the acidic properties of supports
base sites but also acid sites.
are necessary for the ring rearrangement and generation of
5634 | Chem. Commun., 2014, 50, 5633--5636
This journal is ©The Royal Society of Chemistry 2014

View Article Online
ChemComm
Communication
Fig. 1 Time course of the yields of HCPN (K), HCPEN (.), HHD (’), BHF
(m), and HMF ( ) for the ring rearrangement reaction over Au/Nb2O5. The
~
catalyst test at each time was carried out at least three times, and the error
bars represent the minimum to maximum values.
Fig. 2 Fourier transformed Au L3-edge EXAFS spectra of Au supported on
(a) Nb2O5, (b) Al2O3, (c) ZrO2, (d) TiO2 and (e) SA, together with (f) the Au
foil and (g) Au2O3 as references.
HCPN. The reaction for more than 12 h decreased HCPN yield
(45% of HCPN yield for 72 h). Although the products were not
currently identified, 3-hydroxymethylcyclopentanol might be and Au/SA, which gave HCPN (Table 1), together with Au foil and
formed by hydrogenation of HCPN. Au2O3 as references. All of the samples showed a peak around 2.6 Å
The reaction without the supported Au catalysts showed no derived from Au–Au scattering of the Au metal. The peak intensities
yield of HCPN (Table 1 entry 12). In addition, the reaction with of the samples were smaller than that of the Au foil, indicating that
only Nb2O5 or without H2 (under N2) gave no HCPN (Table 1 the samples are composed of Au nanoparticles supported on metal
entries 13 and 14). The results indicate that the hydrogenation oxides. The peak intensity of the Au–Au scattering and the coordina-
of HMF on Au is required to generate HCPN. In these control tion number (Au–Au pair) determined by curve fitting analysis
experiments, the color of the reaction solutions changed from (Table S2, ESI†) increased in the following order: Au/SA o
yellow to dark yellow, and some conversion of HMF was Au/Nb2O5 o Au/TiO2 o Au/ZrO2 o Au/Al2O3, which represents
observed. It is possible that HMF was polymerized by heat the order of the size of Au nanoparticles. The sizes of Au nano-
and the catalysts. The reaction was also carried out using Nb2O5 particles of Au/SA, Au/Nb2O5, and Au/TiO2 were estimated to be
supported Pt, Pd, and Ru catalysts, which are known to have high ca. 2–3 nm, and those for the other catalysts to be larger than 3 nm,
catalytic activity for the hydrogenation reaction. The Nb2O5 sup- assuming that the Au nanoparticles have a cuboctahedral struc-
ported Pt, Pd, and Ru catalysts were prepared by the conventional ture.24 It was noted that the Au particle size had no relation to HCPN
impregnation method or the deposition precipitation method. yield (Table 1), although our previous study showed that small Au
Prior to the reaction, the catalysts were treated with H2 for 1 h nanoparticles have high catalytic performance for the selective
and then O2 for 0.5 h at 473 K. The Pt, Pd, and Ru catalysts hydrogenation of HMF to BHF. It is reasonable to conclude that
provided 28, 43, and 66% yield of HCPN, respectively (Table 1 the selective hydrogenation on Au nanoparticles triggers the for-
entries 15–17). Compared with them, higher HCPN yield was mation of HCPN, but the hydrogenation rate varied with Au particle
obtained on Au/Nb2O5. The selective hydrogenation of HMF over size does not significantly affect the yield of HCPN.
Au is considered to promote the efficient transformation of HMF to
Based on the above results, we concluded that the acidic
HCPN. Interestingly, although the O2 treatment of Au/Nb2O5 prior to properties of catalysts mainly control the HCPN yield. For
the reaction almost did not affect the HCPN yield (Table 1 entries 1 further investigation of the effect of acidity on the HCPN
and 18), the pretreatment of Pt/Nb2O5 improves the HCPN yield formation, we carried out the ring rearrangement reaction on
(Table 1 entries 15 and 19). It is likely that (partially) oxidized Pt Au/Nb2O5 in the presence of a liquid Brønsted acid, H3PO4. As
enhances the selective hydrogenation. The variation of catalysis a result, the addition of H3PO4 reduced the yield of HCPN
under pretreatment conditions is a subject for future study.
(Table 1 entry 20). It is suggested that Brønsted acidity does not
Our previous study demonstrated that the catalytic activity for the contribute to the HCPN formation. Fig. 3 shows the Fourier
selective hydrogenation of HMF to BHF over Au/Al2O3 significantly transformed infrared (FT-IR) spectra of pyridine adsorbed on
depends on the Au particle size.5 Thus, we investigated the effect of the supported Au catalysts which provided HCPN. Although
the Au particle size on the HCPN yield. The Au particle size was Au/SA showed the small band assignable to the ring stretching
evaluated by the extended X-ray absorption fine structure (EXAFS) n19b of pyridine adsorbed on the Brønsted acid site at 1545 cmÀ1
,
spectral analysis (Fig. S1, ESI†). Fig. 2 shows the Fourier transformed the other supported Au catalysts presented almost no band at
Au L3-edge EXAFS spectra of Au/Nb2O5, Au/Al2O3, Au/ZrO2, Au/TiO2, around 1545 cmÀ1. On the other hand, the strong band at
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 5633--5636 | 5635

View Article Online
Communication
ChemComm
(MEXT), Japan – the ‘‘Elements Strategy Initiative to Form Core
Research Center’’ program (since 2012) and Young Scientists
(B) (No. 25820393). XAFS measurements were carried out on the
approval of the Photon Factory Program Advisory Committee
(2012G763).
Notes and references
1 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411.
2 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem., Int. Ed.,
2007, 46, 7164.
3 R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra,
H. J. Heeres and J. G. de Vries, Chem. Rev., 2013, 113, 1499.
4 N. K. Gupta, S. Nishimura, A. Takagaki and K. Ebitani, Green Chem.,
2011, 13, 824.
5 J. Ohyama, A. Esaki, Y. Yamamoto, S. Arai and A. Satsuma, RSC Adv.,
2013, 3, 1033.
6 R. Alamillo, M. Tucker, M. Chia, Y. Pagan-Torres and J. Dumesic,
Green Chem., 2012, 14, 1413.
7 Y. Nakagawa and K. Tomishige, Catal. Commun., 2010, 12, 154.
8 M. Chia, Y. J. Pagan-Torres, D. Hibbitts, Q. Tan, H. N. Pham,
A. K. Datye, M. Neurock, R. J. Davis and J. A. Dumesic, J. Am. Chem.
Soc., 2011, 133, 12675.
9 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic, Nature,
2007, 447, 982.
Fig. 3 FT-IR spectra of pyridine adsorbed on (a) Nb2O5, (b) Al2O3, (c) ZrO2,
(d) TiO2, and (e) SA supported Au catalysts.
10 J. N. Chheda and J. A. Dumesic, Catal. Today, 2007, 123, 59.
´
11 L. Bui, H. Luo, W. R. Gunther and Y. Roman-Leshkov, Angew. Chem.,
1440–1455 cmÀ1 assignable to the n19b mode of pyridine
adsorbed on the Lewis acid site was observed for all of the
catalysts. It is noteworthy that Au/Nb2O5, which gave the highest
Int. Ed., 2013, 52, 8022.
12 Y. Yang, Z. Du, Y. Huang, F. Lu, F. Wang, J. Gao and J. Xu, Green
Chem., 2013, 15, 1932.
13 S. Sato, J. Igarashi and Y. Yamada, Appl. Catal., A, 2013, 453, 213.
´
HCPN yield, presented a strong band for pyridine on the Lewis 14 T. Buntara, S. Noel, P. H. Phua, I. Melian-Cabrera, J. G. de Vries and
H. J. Heeres, Angew. Chem., Int. Ed., 2011, 50, 7083.
15 M. Hronec and K. Fulajtarova, Catal. Commun., 2012, 24, 100.
16 M. Hronec, K. Fulajtarova and T. Liptaj, Appl. Catal., A, 2012,
acid site, but almost no band for pyridine on the Brønsted acid.
These results indicate that the ring rearrangement reaction
proceeds on Lewis acid sites. Recently, Nb2O5 has been reported
to have unique Lewis acid sites working in water.25,26 Such Lewis
acid sites might contribute to the high yield of HCPN.
437–438, 104.
17 A. Corma and P. Serna, Science, 2006, 313, 332.
18 E. Bus, R. Prins and J. A. van Bokhoven, Catal. Commun., 2007, 8, 1397.
19 J. E. Bailie and G. J. Hutchings, Chem. Commun., 1999, 2151.
In summary, we first demonstrated the ring rearrangement 20 J. E. Bailie, H. A. Abdullah, J. A. Anderson, C. H. Rochester,
N. V. Richardson, N. Hodge, J.-G. Zhang, A. Burrows, C. J. Kiely
of HMF to HCPN over supported Au catalysts by taking advantage
and G. J. Hutchings, Phys. Chem. Chem. Phys., 2001, 3, 4113.
of the selective hydrogenation on Au and the Lewis acid catalysis of
21 C. Milone, M. L. Tropeano, G. Gulino, G. Neri, R. Ingoglia and
metal oxide supports. Among the Au catalysts, Au/Nb2O5 exhibited
the highest catalytic performance for the ring rearrangement, and
gave 86% yield of HCPN. The ring rearrangement of HMF over Au
S. Galvagno, Chem. Commun., 2002, 868.
22 M. Haruta, Catal. Today, 1997, 36, 153.
23 Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal., 2013,
3, 2655.
catalysts proposed in this study can contribute to the development 24 J. Ohyama, K. Teramura, S.-i. Okuoka, S. Yamazoe, K. Kato,
T. Shishido and T. Tanaka, Langmuir, 2010, 26, 13907.
25 K. Nakajima, Y. Baba, R. Noma, M. Kitano, J. N. Kondo, S. Hayashi
of the biorefinery for a sustainable future.
This work was supported by a Grant-in-Aids from the
and M. Hara, J. Am. Chem. Soc., 2011, 133, 4224.
Ministry of Education, Culture, Sports, Science and Technology 26 Y. Koito, K. Nakajima, M. Kitano and M. Hara, Chem. Lett., 2013, 42, 873.
5636 | Chem. Commun., 2014, 50, 5633--5636
This journal is ©The Royal Society of Chemistry 2014