Selective hydrogenolysis of 2-furancarboxylic acid to 5-hydroxyvaleric acid derivatives over supported platinum catalysts

Article abstract of DOI:10.1039/c9gc03315g

The conversion of 2-furancarboxylic acid (FCA), which is produced by oxidation of furfural, to 5-hydroxyvaleric acid (5-HVA) and its ester/lactone derivatives with H₂ was investigated. Monometallic Pt catalysts were effective, and other noble metals were not effective due to the formation of ring-hydrogenation products. Supports and solvents had a small effect on the performance; however, Pt/Al₂O₃ was the best catalyst and short chain alcohols such as methanol were better solvents. The optimum reaction temperature was about 373 K, and at higher temperature the catalyst was drastically deactivated by deposition of organic materials on the catalyst. The highest yield of target products (5-HVA, δ-valerolactone (DVL), and methyl 5-hydroxyvalerate) was 62%, mainly obtained as methyl 5-hydroxyvalerate (55% yield). The byproducts were mainly ring-hydrogenation compounds (tetrahydrofuran-2-carboxylic acid and its ester) and undetected ones (loss of carbon balance). The catalyst was gradually deactivated during reuses even at a reaction temperature of 373 K; however, the catalytic activity was recovered by calcination at 573 K. The reactions of various related substrates were carried out, and it was found that the O-C bond in the O-CC structure (1,2,3-position of the furan ring) is dissociated before CC hydrogenation while the presence and position of the carboxyl group (or methoxy carbonyl group) much affect the reactivity.

Full text of DOI:10.1039/c9gc03315g

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Selective hydrogenolysis of 2-furancarboxylic acid
to 5-hydroxyvaleric acid derivatives over
supported platinum catalysts†
Cite this: Green Chem., 2019, 21,
6133
Takehiro Asano,a Hiroshi Takagi,a Yoshinao Nakagawa,
*
a,b Masazumi Tamuraa,b
a,b
and Keiichi Tomishige
*
The conversion of 2-furancarboxylic acid (FCA), which is produced by oxidation of furfural, to 5-hydroxy-
valeric acid (5-HVA) and its ester/lactone derivatives with H2 was investigated. Monometallic Pt catalysts
were effective, and other noble metals were not effective due to the formation of ring-hydrogenation
products. Supports and solvents had a small effect on the performance; however, Pt/Al2O3 was the best
catalyst and short chain alcohols such as methanol were better solvents. The optimum reaction tempera-
ture was about 373 K, and at higher temperature the catalyst was drastically deactivated by deposition of
organic materials on the catalyst. The highest yield of target products (5-HVA, δ-valerolactone (DVL), and
methyl 5-hydroxyvalerate) was 62%, mainly obtained as methyl 5-hydroxyvalerate (55% yield). The bypro-
ducts were mainly ring-hydrogenation compounds (tetrahydrofuran-2-carboxylic acid and its ester) and
undetected ones (loss of carbon balance). The catalyst was gradually deactivated during reuses even at a
reaction temperature of 373 K; however, the catalytic activity was recovered by calcination at 573 K. The
reactions of various related substrates were carried out, and it was found that the O–C bond in the
O–CvC structure (1,2,3-position of the furan ring) is dissociated before CvC hydrogenation while the
presence and position of the carboxyl group (or methoxy carbonyl group) much affect the reactivity.
Received 21st September 2019,
Accepted 9th October 2019
DOI: 10.1039/c9gc03315g
rsc.li/greenchem
as intermediate compounds.18–22 For example, γ-butyrolactone
and γ-valerolactone can be synthesized by hydrogenation of
1. Introduction
Utilization of biomass is one of the approaches to escape from succinic acid and levulinic acid, respectively, which are plat-
dependence on fossil resources.1–8 Remarkably, replacement form chemicals produced by fermentation and catalytic con-
of petroleum-derived materials in the fiber and resin indus- version of hexoses, respectively.23–28 Furfural is one of the
tries with biomass-derived compounds has attracted most important platform chemicals because it can be readily
attention.9–11 Production of fiber and resin typically uses produced from hemicellulose by hydrolysis and dehydration.29
molecules functionalized at two positions, especially at Conversions of furfural to various compounds have been inten-
α,ω-positions, as monomers.12–16 These monomers include sively investigated with various reactions such as reduction,
diols, dicarboxylic acids, and hydroxy acids (or their lactones). oxidation, condensation, and so on.30–34 Reduction of furfural
It is difficult to directly produce monomers from biomass to pentanediols with H2 is well studied.35,36 1,2-Pentanediol
because of the very complex structure of biomass.17 A typical can be synthesized from furfural or furfuryl alcohol, which is
approach in the biomass refinery is the use of “platform the CvO hydrogenation product of furfural, using various cat-
chemicals”, which are relatively easily obtained from biomass, alysts such as Pt/hydrotalcite,37 Pt/CeO2,38–40 Pt/Co2AlO4,41 Ru/
44
MnOx,42 Cu-Al2O3,43 and Cu-Mg3AlO4.5
.
1,5-Pentanediol was
also produced simultaneously in some systems for 1,2-penta-
nediol production.37,43,44 On the other hand, 1,5-pentanediol
can be obtained as the main product from furfural over Rh–Ir-
ReOx/SiO2 or Pd–Ir-ReOx/SiO2 catalysts.45,46 In these two
systems, tetrahydrofurfuryl alcohol which is the total hydro-
genation product of furfural is the key intermediate,45–49 while
in the systems for 1,2-pentanediol production tetrahydrofur-
furyl alcohol is not an intermediate but a by-product of parallel
reactions.37,39,41–44
aDepartment of Applied Chemistry, School of Engineering, Tohoku University,
6-6-07 Aoba, Aramaki, Aoba-ku, Sendai, 980-8579, Japan.
E-mail: yoshinao@erec.che.tohoku.ac.jp; Fax: +81-22-795-7215;
Tel: +81-22-795-7215
bResearch Center for Rare Metal and Green Innovation, Tohoku University, 468-1,
Aoba, Aramaki, Aoba-ku, Sendai 980-0845, Japan.
E-mail: tomi@erec.che.tohoku.ac.jp; Fax: +81-22-795-7214; Tel: +81-22-795-7214
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9gc03315g
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6133

View Article Online
Paper
Green Chemistry
2-Furancarboxylic acid (FCA) can be easily produced by oxi- the yields were low (total yield <20%).53,54 Our previous studies
dation of furfural.50–52 However, studies of utilization of FCA did not aim at 5-HVA production but for valeric acid pro-
as a platform chemical are very limited. The literature data of duction, and thus catalyst and reaction conditions were not
catalytic FCA reduction with H2 are summarized in Table S1,† the best ones for the production of 5-HVA derivatives. An old
and many of them are the ring-hydrogenation to tetrahydro- report in 1949 mentioned the formation of DVL from FCA over
furan-2-carboxylic acid or its derivatives and the hydrogenation the Adams catalyst (Pt oxide);72 however, a detailed procedure
of the carboxyl group to furfuryl alcohol. We have very recently and yield/selectivity values were not provided, and there was
reported hydrodeoxygenation of FCA to valeric acid over a Pt- another old report describing that the Adams catalyst pro-
MoOx/TiO2 catalyst with ∼60% yield.53,54 In this system, the duced tetrahydrofuran-2-carboxylic acid from FCA instead of
combination of these metals (Pt and Mo) and water solvent is 5-HVA derivatives.73
essential to obtain good yield of valeric acid; Pt/TiO2 and MoOx/
In this study, we investigated the catalytic reaction for selec-
TiO2 catalysts showed much lower activity and selectivity to tive production of 5-HVA derivatives from FCA. A simple Pt/
valeric acid than the Pt-MoOx/TiO2 catalyst. As a related Al2O3 catalyst gave good yield (>60%) of 5-HVA derivatives
reaction, conversion of 2,5-furandicarboxylic acid (FDCA) to when the reaction conditions were appropriately chosen, and
adipic acid has been investigated by a few research groups.55–58
In the context of fiber and resin, as described above,
ω-hydroxycarboxylic acids are useful chemicals, and the pro-
duction of 5-hydroxyvaleric acid (5-HVA) from FCA will be
important, although there is no study which aims for this con-
version, to the best of our knowledge. Reported production
it was reusable after regeneration by calcination.
2. Experimental section
2.1. Catalyst preparation
methods of 5-HVA and its derivatives such as δ-valerolactone Pt/support catalysts (support = Al2O3, CeO2, ZrO2, SiO2, and
(DVL) include Baeyer–Villiger oxidation of cyclopentanone TiO2) were prepared by an impregnation method. Used sup-
which is one of the reduction products of furfural,59–61 partial ports were Al2O3 (γ-type, AEROXIDE Alu C, 100 m2 g−1
,
,
oxidation of 1,5-pentanediol which is also one of the reduction AEROSIL, calcined at 973 K for 1 h), CeO2 (HS, 84 m2 g−1
products of furfural,62–69 and reduction of glutaric acid.70 Daiichi Kigenso Kagaku Kogyo, calcined at 973 K for 1 h), ZrO2
Additionally, DVL is produced by dehydrogenation of 1,5-pen- (RC-100P, 100 m2 g−1, Daiichi Kigenso Kagaku Kogyo, calcined
tanediol which is produced by hydrogenation of glutaric acid at 973 K for 1 h), SiO2 (CARiACT G6, 475 m2 g−1, Fuji Silysia,
in industry.71 These production methods of 5-HVA derivatives calcined at 973 K for 1 h), and TiO2 (AEROXIDE P25, 30 m2
from furfural involve oxidation of reduced functional groups, g−1, AEROSIL, calcined at 973 K for 1 h). The Pt precursor was
requiring expensive oxidants or a larger amount of H2 as the H2PtCl6 aq. (Pt 4.56 wt%; Sigma-Aldrich) except for Pt/SiO2
reducing agent than the production via FCA (Scheme 1). On where Pt(NO2)2(NH3)2 solution in HNO3 (Tanaka Kikinzoku
the other hand, our Pt-MoOx/TiO2 catalyst system actually pro- Kogyo) was used instead. The loading amount of the noble
duces 5-HVA, 5-oxovaleric acid, and DVL from FCA although metal was 4 wt%. After impregnation, the catalysts were dried
Scheme 1 Production of 5-hydroxyvaleric acid derivatives from furfural or petroleum.
6134 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
at 383 K overnight and then they were calcined at 773 K for 3 h carbon balance values were significantly lower than 100%
under air. As another type of Pt/Al2O3 catalyst, the Pt/Al2O3 (typically 70–90%). The phrase “derived from substrate” in the
catalyst with lower dispersion was also prepared by changing following equations means that the alkoxy group in the ester
the calcination temperature to 873 K (Pt/Al2O3-sintered). The was excluded from the carbon amount.
ꢀ
ꢁ
amount of detected substrate ½mmolꢀ
amount of loaded substrate ½mmolꢀ
Conversion½%ꢀ ¼ 1 ꢂ
ꢁ 100
amount of C atoms derived from substrate in the product ½mmolꢀ
amount of C atoms in loaded substrate ½mmolꢀ
Yield½%ꢀ ¼
ꢁ 100
prepared Pt catalysts were used for catalytic runs as the cal-
cined form because they can be easily reduced with H2.
Commercial Pt/C, Rh/C, Ru/C, and Pd/C catalysts (loading
amount of noble metals: 5 wt%) were purchased from Wako
Yield½%ꢀ
Selectivity½%ꢀ ¼
ꢁ 100
Conversion½%ꢀ
P
ðamount of C atoms derived from substrate in the detected substrate and product ½mmolꢀÞ
amount of C atoms in loaded substrate ½mmolꢀ
Carbon balance ½%ꢀ ¼
ꢁ 100
Pure Chemical Industries, Ltd. The characterization data (CO
adsorption amount) of these commercial catalysts were
described in our previous work.74
As discussed in the Results and discussion section, the loss of
carbon balance was regarded as the formation of undetected by-
products. The loss of carbon balance (100% – carbon balance) is
connected to the sum of selectivities by the following equation.
2.2. Activity test
Loss of carbon balance½%ꢀ ¼
Activity tests of various catalysts were carried out with a 190 ml
stainless-steel autoclave. A glass vessel was inserted into the
autoclave to avoid contamination from the autoclave.
Substrates are commercially available and were used as
received (2-furancarboxylic acid and 3-furancarboxylic acid:
Wako Pure Chemical Industries, Ltd; tetrahydrofuran-2-car-
boxylic acid, furan, 2,3-dihydrofuran, 2,5-dihydrofuran, and
n-butyl vinyl ether: Tokyo Chemical Industry Co., Ltd).
Typically, the catalyst (0.050 g), solvent (29 g), and substrate
(10 mmol; 1.12 g in the case of FCA) were loaded in the
inserted glass vessel. The glass vessel was set to the autoclave
with a spinner, and the reactor was purged by flushing three
times with 1 MPa of H2. After that, the autoclave was filled
with 4 MPa hydrogen at room temperature. The autoclave was
P
100 ꢂ ðSelectivity½%ꢀÞ
ðConversion½%ꢀÞ
100
The used catalysts were recycled in two different recovery
and treatment methods; (i) separation by centrifugation (6000
rpm, 5 min) under air, washing with methanol three times,
drying at 383 K overnight, and (ii) calcination in air at 573 K
for 1 h after separation and washing. We carried out multiple
runs at the same time under the same conditions in order to
collect sufficient amount of used catalyst. The loss of the cata-
lyst was compensated by reducing the number of runs.
2.3. Catalyst characterization
then heated to 373 K. The reaction time was defined as 0 h X-ray diffraction (XRD) patterns were recorded with
a
when the temperature of the reactor reached 373 K. The MiniFlex600 (Rigaku) instrument under air after the samples
heating to 373 K took about 20 min. The reaction mixture in were used or reduced in methanol under 5 MPa H2 at 373 K.
the reactor was always magnetically stirred at 500 rpm. After Cu Kα (λ = 0.154 nm, 45 kV, 40 mA) radiation was used as an
scheduled reaction time, the autoclave was cooled to room X-ray source. The Pt particle size (d) of the catalysts was calcu-
temperature with a water bath. The gas was collected in a gas lated by Scherrer’s equation using the Pt (111) peak at around
bag, and the reaction mixture was collected with a washing 40°. The support signals overlapping with the Pt peak were
solvent (methanol 15 g, Kanto Chemical Industries, Ltd). After removed by subtracting the XRD pattern of the support from
the catalyst was separated from the reaction mixture by fil- the pattern of the sample.
tration, the reaction solution was analyzed by using a FID-GC
The amount of CO chemisorption was measured by a pulse
(Shimadzu GC-2025) equipped with an HP-FFAP capillary adsorption method with a BELCAT II (MicrotracBEL). The cal-
column, and the gas was analyzed by using a FID-GC cined catalyst (about 50 mg) was loaded to a sample tube, and
(Shimadzu GC-2014) equipped with a Porapak N packed pre-reduced with H2 at 373 K for 30 min. After reduction, the
column and a methanator. GC-MS (QP-2010, Shimadzu) was sample tube was purged with He at 473 K for 30 min. Adsorption
also used for product identification.
was performed after cooling to 308 K. Adsorption gas was 10 vol%
Conversion, yield, selectivity and carbon balance were calcu- CO/He and 0.888 cm3 per pulse. Adsorption was measured until
lated by the following equations. In almost all cases, the the difference of last three pulses after adsorption became lower
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6135

View Article Online
Paper
Green Chemistry
than 1%. Field emission scanning transmission electron catalysts. All the tested monometallic Pt catalysts showed mod-
microscopy (FE-STEM) images were obtained with a HD-2700 erate selectivity to 5-HVA derivatives, indicating that Pt rather
(Hitachi). The samples dispersed in methanol were dropped on than the support is the essential element for the production of
Cu microgrids under air. The average Pt particle size (d) was cal- 5-HVA derivatives. Among the Pt/support catalysts, Pt/Al2O3
culated by the following equation: d = ∑(nidi3)/∑(nidi2) (di: size of showed relatively high activity and highest selectivity to target
each particle, ni: number of particles with size of di). The TG-DTA 5-HVA derivatives. The effect of supports of Pt/support catalysts
profile was measured by using a Rigaku Thermo plus EVO II was not strong in the production of 5-HVA derivatives from
instrument. The sample amount was about 10 mg, and the FCA, and this tendency was similar to the case of VA pro-
sample was heated up to 1073 K at a rate of 10 K min−1 under air. duction over Pt-MoOx catalysts, although the best support was
different (Al2O3 in this case; TiO2 for Pt-MoOx catalysts for VA
production).53,54 The dispersion values determined by CO
adsorption are also shown in Table 1. The XRD patterns of the
3. Results and discussion
Pt catalyst are shown in Fig. S1,† and the Pt crystallite size
First, we tested various supported noble metal catalysts in determined from the XRD pattern fairly agreed with the dis-
reduction of FCA with H2 in methanol solvent, and the results persion values in Table 1. Among Pt catalysts except Pt/Al2O3
are shown in Table 1 (details: Table S2†). Potential products of and Pt/CeO2, the activity (conversion) fairly increased with
reduction of FCA in alcohol solvent are summarized in increasing dispersion. The Pt/Al2O3 catalyst showed higher
Scheme 2. Platinum catalysts were mainly tested because our activity than this trend, suggesting that Al2O3 plays an
previous study for FCA reduction53,54 and many literature additional role in the catalysis such as adsorption of the sub-
studies for furfural reduction to pentanediols used platinum strate.75 We also tested the sintered Pt/Al2O3 catalyst with
as the main active metal.37–41 The Pt/Al2O3 catalyst gave good lower dispersion by calcination at higher temperature during
selectivity to methyl 5-hydroxyvalerate (Me-5-HV), which is one preparation (entry 8). In spite of the very low dispersion, the
of the target products (5-hydroxyvaleric acid (5-HVA) deriva- activity of the sintered Pt/Al2O3 catalyst showed comparable
tives) (entry 1). At longer reaction time (entry 2), the total yield activity to other Pt catalysts on inert supports such as Pt/C,
of target products (Me-5-HV, 5-HVA, and δ-valerolactone (DVL)) indicating the higher activity of Pt on the Al2O3 support. Other
reached 62% over the Pt/Al2O3 catalyst. Main byproducts were active metals (Rh, Ru, and Pd) were also tested as carbon-sup-
ring-hydrogenation products, namely tetrahydrofuran-2-car- ported catalysts (entries 9–11). All these catalysts showed high
boxylic acid (THFCA) and its methyl ester (MeTHFC). Other selectivity to ring-hydrogenation products (THFCA and
C–O hydrogenolysis products were hardly produced, such as MeTHFC) and the selectivity to target products (5-HVA deriva-
methyl 2-hydroxyvalerate, 2-hydroxyvaleric acid, valeric acid tives) was much lower than that to Pt catalysts. According to
(VA), methyl valerate, and the hydrogenation products of the the literature results for FCA hydrogenation (Table S1†),
carboxyl group in C–O hydrogenolysis products such as 1,5- Rh,53,54,76 Pd,53,54,76–79 and Ru76,80 catalysts have been actually
pentanediol and 1,2-pentanediol. Furthermore, the ester of the reported as active ones in ring-hydrogenation of FCA or its
substrate, methyl 2-furancarboxylate, was also hardly observed ester. The effect of Mo addition to Pt was investigated (entries 5
in all entries. Entries 3–7 show the effect of the support of Pt and 12) because we have reported that the Pt-MoOx/TiO2 catalyst
Table 1 Reduction of 2-furancarboxylic acid (FCA) over various catalysts in methanol solvent
Sel./%
Entry
Catalyst
Conv./%
Me-5-HV
5-HVA
DVL
MeTHFC
THFCA
Others
C.B./%
Dispersion (CO/M)
1
Pt/Al2O3
Pt/Al2O3
Pt/ZrO2
Pt/CeO2
Pt/TiO2
Pt/SiO2
Pt/C
91
>99
58
91
25
34
16
17
51
55
46
37
38
34
24
32
6
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
6
7
4
7
5
6
6
2
<1
<1
<0.1
2
12
20
13
10
8
5
5
10
2
6
24
8
23
25
11
60
61
89
11
<1
1
<1
<1
3
<1
1
2
<1
<1
<1
3
82
86
84
80
90
90
94
93
93
92
97
93
0.29
0.29
0.42
—
0.22
2a
3
b
4
5
6
0.30
7c
8
0.15d
0.02
Pt/Al2O3-sinterede
Rh/C
9
9c
10c
11c
12
>99
>99
93
26
23
8
0.24d
0.35d
0.08d
0.33 f
Ru/C
Pd/C
Pt-MoOx/TiO2
6
<0.1
33
f
16
10
Reaction conditions: 4 wt% M, Wcat = 0.050 g, WFCA = 1.12 g, WMeOH = 29 g, P(H2) = 4 MPa at r.t., T = 373 K, t = 1 h. a t = 4 h. b Not measured
because of the adsorption of CO on CeO2. c Wcat = 0.040 g commercial M/C (5 wt% M) catalyst. d Reported in ref. 72. e Catalyst calcined at 873 K to
decrease Pt dispersion. f Pt 4 wt%, Mo 0.5 wt% catalyst reported in ref. 53 and 54. Conv.: conversion, Sel.: selectivity, C.B.: carbon balance, FCA:
2-furancarboxylic acid, Me-5-HV: methyl 5-hydroxyvalerate, 5-HVA: 5-hydroxyvaleric acid, DVL: δ-valerolactone, MeTHFC: methyl tetrahydrofuran-
2-carboxylate, THFCA: tetrahydrofuran-2-carboxylic acid, Others: methyl valerate, valeric acid, 1,5-pentanediol, methyl 2-hydroxyvalerate, 2-hydro-
xyvaleric acid, and methyl 2-furancarboxylate. Detailed data are shown in Table S1.†
6136 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
Scheme 2 Possible products in reduction of FCA in alcohol solvent.
is much more effective in FCA reduction to VA than Pt/TiO2 in due to the difference in solvent (this study: methanol; previous
view of both activity and selectivity. Mo addition did not work: water), as also discussed next. The difference in other
increase the selectivity to 5-HVA derivatives, and the activity was reaction conditions was not large (this study: FCA concentration
rather significantly decreased. Therefore, Mo addition has a ∼3 wt%, 4 MPa H2 at r.t., 373 K; previous studies: FCA concen-
rather negative effect on the production of 5-HVA derivatives. tration ∼5 wt%, 1.5 MPa H2 at 413 K, 413 K).53,54
The different trends for the Mo addition effect in FCA conver-
Table 2 (details: Table S3†) shows the effect of the solvent
sion between this study and our previous studies53,54 can be for selective hydrogenolysis of FCA to 5-HVA derivatives over
Table 2 Solvent effect on hydrogenolysis of FCA over the Pt/Al2O3 (4 wt% Pt) catalyst
Sel./%
Entry
Solvent
Conv./%
R-5-HV
5-HVA
DVL
AVA
RTHFC
THFCA
Others
C.B./%
1
2
3
4
5
6
7
8
9
Methanol
Ethanol
1-Propanol
2-Propanol
tert-Butanol
1,4-Dioxane
THF
91
90
96
97
81
88
64
55
99
51
44
27
4
<0.1
—
—
—
—
<0.1
<1
3
5
7
5
4
<0.1
4
6
—
—
—
—
—
—
—
1
12
3
1
<1
<0.1
—
—
—
—
10
19
23
27
28
35
34
27
30
<1
<1
<1
<1
<1
<1
<1
<1
6
82
81
80
79
77
78
83
97
73
13
24
42
37
34
34
66
33
Acetic acid
Water
—
Reaction conditions: Pt/Al2O3(4 wt% Pt), Wcat = 0.050 g, WFCA = 1.12 g, Wsolvent = 29 g, P(H2) = 4 MPa at r.t., T = 373 K, t = 1 h. Conv.: conversion,
Sel.: selectivity, C.B.: carbon balance, <0.1: not detected, FCA: 2-furancarboxylic acid, R-5-HV: alkyl 5-hydroxyvalerate, 5-HVA: 5-hydroxyvaleric
acid, DVL: δ-valerolactone, AVA: 5-acetoxyvaleric acid, RTHFC: alkyl tetrahydrofuran-2-carboxylate, THFCA: tetrahdyrofuran-2-carboxylic acid,
Others: alkyl valerate, valeric acid, 1,5-pentanediol, alkyl 2-hydroxyvalerate, 2-hydroxyvaleric acid, and alkyl 2-furancarboxylate. Detailed data are
shown in Table S2.†
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6137

View Article Online
Paper
Green Chemistry
the Pt/Al2O3 catalyst. Generally, the effect of the solvent on the
total selectivity to 5-HVA derivatives was not large.
Nevertheless, among alcohol solvents (entries 1–5), the selecti-
vity to 5-HVA derivatives decreased and that to ring-hydrogen-
ation products increased with the increase of the carbon
number or the change from primary alcohol to secondary or
tertiary ones. The distribution in the target products was
changed from the ester-rich one to DVL-rich one with the
increase of the carbon number of solvent or the change of the
solvent from primary alcohol. This trend in the ratio of target
products to ring-hydrogenation can be connected to the
polarity of the solvent: a highly polar solvent tends to give
higher selectivities to target products. In fact, ether solvents
with low polarity showed relatively low selectivity to target pro-
ducts (∼40%; entries 6 and 7), and acetic acid solvent showed
good selectivity to target products, mainly as DVL (66%; entry
8). A small amount of 5-acetoxyvaleric acid (AVA) was also
detected, which is the ester of acetic acid and 5-HVA. AVA can
also be one of the target products. However, the activity in
acetic acid was significantly lower than that in methanol, prob-
ably because of the competitive adsorption of the solvent and
substrate molecules on the catalyst surface.53,54 Considering
also the handling issue of acetic acid solvent, we selected
methanol solvent in the following studies. Water solvent
showed a different result from other solvents: in spite of the
high polarity, the selectivity was rather similar to that in ether
solvent, and water solvent is not suitable for 5-HVA derivative
Fig. 1 Effect of reaction temperature on hydrogenolysis of FCA over
the Pt/Al2O3 catalyst. Reaction conditions: Pt/Al2O3 (4 wt% Pt), Wcat
=
0.050 g, WFCA = 1.12 g, WMeOH = 29 g, P(H2) = 4 MPa at r.t., t = 1 h.
Circle: conversion; triangle: carbon balance; red: methyl 5-hydroxyvale-
rate (Me-5-HV); green: δ-valerolactone (DVL); dark blue: methyl tetra-
hydrofuran-2-carboxylate (MeTHFC); orange: tetrahydrofuran-2-car-
boxylic acid (THFCA); gray: others (methyl valerate, valeric acid, 1,5-pen-
tanediol, methyl 2-hydroxyvalerate, 2-hydroxyvaleric acid, and methyl
2-furancarboxylate). Detailed data are shown in Table S3.†
production. The activity of the Pt/Al2O3 catalyst in water negligible at 373 K. On the other hand, at 413 K, the reaction
solvent was rather high (activity order: Pt/Al2O3 in water stopped at around 50% conversion level. This suggests that de-
(Table 2, entry 9) >Pt/Al2O3 in methanol (Table 1 or 2, entry 1) activation of the catalyst occurred at higher reaction tempera-
≫Pt/TiO2 in methanol (Table 1, entry 5) >Pt-MoOx/TiO2 in ture. The TG-DTA profile (under air) of the Pt/Al2O3 catalyst
methanol (Table 1, entry 11)), which was in contrast to the after reaction at 413 K for 4 h is shown in Fig. S2.† An exother-
reported low activity of Pt/TiO2 for FCA reduction in water mic signal with a weight loss of ca. 2% was observed at
solvent (activity order: Pt-MoOx/TiO2 in water ≫Pt-MoOx/TiO2 470–590 K, which suggests the deposition of the organic
in methanol >Pt/TiO2 in water).54 It is well known that the material on the catalyst surface. Although the amount of
γ-Al2O3 support is transformed into boehmite (AlOOH) in hot deposit (weight loss in TG: ca. 1 mg per 50 mgcat) was much
water solvent.81 The different trends from other solvents or the lower than the loss of carbon balance (0.17 g) in the reaction,
Pt/TiO2 catalyst may be due to the change of the support it is large enough to cover the surface (1 mg-C per 50 mgcat
structure.
0.08 mmol-C per 0.01 mmol-Pt). The loss of carbon balance
Fig. 1 (details: Table S4†) shows the effect of reaction temp- was probably mainly due to soluble polymeric byproducts.
erature of FCA hydrogenolysis over the Pt/Al2O3 catalyst. From
Fig. 3 (details: Table S7†) shows the effect of H2 pressure.
=
343 K to 373 K, the conversion increased (30% → 91%) and Up to 4 MPa (at r.t.), the conversion and yield of 5-HVA deriva-
the selectivity to 5-HVA derivatives also increased (38% → tives increased with increasing H2 pressure. However, the con-
57%), while the selectivity to ring-hydrogenation products version decreased above 4 MPa. In order to check whether the
(THFCA and MeTHFC) was almost unchanged (∼22%). Higher decrease of conversion at high H2 pressure was due to the reac-
reaction temperature is better in this range (≤373 K). However, tion rate itself or the catalyst deactivation, we measured the
from 373 K to 393 K, the reaction results were almost initial reaction rates at different H2 pressures (at r.t.) with a
unchanged, and above 393 K, the conversion as well as selecti- smaller catalyst amount (Fig. 4; details: Table S8 and Fig. S3,
vity to 5-HVA derivatives was even decreased. The optimum S4†). The initial reaction rate actually decreased at higher H2
reaction temperature of this reaction is about 373 K. The time pressure than 4 MPa. This volcano-type curve of reaction rate
courses of FCA hydrogenolysis at 373 K and 413 K are shown versus H2 pressure can be explained by competitive adsorption
in Fig. 2 (details: Tables S5 and S6†). At 373 K, the conversion of FCA and H2 on the same active site (Pt). There are other pos-
was increased from 91% at 1 h to >99% at 4 h, and the selecti- sibilities that caused this volcano-curve kinetics. One typical
vity was hardly changed during this period. This indicates that example is the change of the catalyst structure by reduction of
the reaction proceeded until the substrate (FCA) was comple- the catalyst; however, it is unlikely that the Pt/Al2O3 catalyst
tely consumed while the overreaction of 5-HVA derivatives is undergoes structural change by different H2 pressures because
6138 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
Fig. 3 Effect of hydrogen pressure on hydrogenolysis of FCA over the
Pt/Al2O3 (4 wt% Pt) catalyst. Reaction conditions: Pt/Al2O3 (4 wt% Pt),
Wcat = 0.050 g, WFCA = 1.12 g, WMeOH = 29 g, T = 373 K, t = 1 h. Circle:
conversion; triangle: C.B. (carbon balance); red: methyl 5-hydroxyvale-
rate (Me-5-HV); green: δ-valerolactone (DVL); dark blue: methyl tetra-
hydrofuran-2-carboxylate (MeTHFC); orange: tetrahydrofuran-2-car-
boxylic acid (THFCA); gray: others (methyl valerate, valeric acid, 1,5-pen-
tanediol, methyl 2-hydroxyvalerate, 2-hydroxyvaleric acid, and methyl
2-furancarboxylate). Detailed data are shown in Table S6.†
Fig. 2 Time-course of hydrogenolysis of FCA over Pt/Al2O3 (4 wt% Pt)
catalyst at (I) 373 K and (II) 413 K. Reaction conditions: Pt/Al2O3 (4 wt%
Pt), Wcat = 0.050 g, WFCA = 1.12 g, WMeOH = 29 g, P(H2) = 4 MPa at r.t.
Circle: conversion; triangle: carbon balance; closed circle: yield; red:
methyl 5-hydroxyvalerate (Me-5-HV); green: δ-valerolactone (DVL); dark
blue: methyl tetrahydrofuran-2-carboxylate (MeTHFC); orange: tetra-
hydrofuran-2-carboxylic acid (THFCA); gray: others (methyl valerate,
valeric acid, 1,5-pentanediol, methyl 2-hydroxyvalerate, 2-hydroxyvale-
ric acid, and methyl 2-furancarboxylate). Detailed data are shown in
Tables S4 and S5.†
Fig. 4 Dependence of the initial reaction rate of FCA hydrogenolysis on
H2 pressure. Circle: conversion rate; square: 5-HVA derivatives (methyl
5-hydroxyvalerate (Me-5-HV), 5-hydroxyvaleric acid (5-HVA), and
δ-valerolactone (DVL)) production rate. Reaction conditions: Pt/Al2O3
(4 wt% Pt), Wcat = 0.010 g, WFCA = 1.12 g, WMeOH = 29 g, T = 373 K, t =
0–1.5 h. Detailed data are shown in Table S7.†
Al2O3 is not reducible and Pt is easily fully reduced to the
metallic state with H2 under very mild conditions.
Table 3 (details: Table S9†) shows the effect of substrate
concentration at the constant amount of FCA and different
amounts of solvent. The conversion slightly increased with the
increase of concentration (decrease of the solvent amount),
which agreed with the unsaturated adsorption of the substrate
on the catalyst surface by the presence of H2. However, the
selectivity to 5-HVA derivatives (5-HVA and DVL) slightly
decreased with the increase of concentration. We selected
lower concentration of FCA to obtain the highest 5-HVA deriva-
tive yield.
After all, the optimized conditions are methanol solvent,
373 K temperature, 4 MPa H2 (at r.t.), 1.1 g FCA/29 g methanol,
which are the same conditions as those in Table 1. The
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6139

View Article Online
Paper
Green Chemistry
Table 3 Effect of FCA concentration (methanol solvent amount) on hydrogenolysis of FCA over the Pt/Al2O3 catalyst
Sel./%
Entry
WMeOH/g
Conv./%
Me-5-HV
5-HVA
DVL
MeTHFC
THFCA
Others
C.B./%
1
29
29
19
9
91
>99
95
>99
>99
51
55
54
55
53
< 0.1
< 0.1
< 0.1
< 0.1
< 1
6
7
4
3
4
12
20
18
18
16
10
2
5
5
7
< 1
1
1
1
2
82
86
83
84
82
2a
3
4
5
4
Reaction conditions: Pt/Al2O3 (4 wt% Pt), Wcatalyst = 0.050 g, WFCA = 1.12 g, P(H2) = 4 MPa at r.t., T = 373 K, t = 1 h. a t = 4 h. Conv.: conversion,
Sel.: selectivity, C.B.: carbon balance, FCA: 2-furancarboxylic acid, Me-5-HV: methyl 5-hydroxyvalerate, 5-HVA: 5-hydroxyvaleric acid, DVL:
δ-valerolactone, MeTHFC: methyl tetrahydrofuran-2-carboxylate, THFCA: tetrahydrofuran-2-carboxylic acid, Others: methyl valerate, valeric acid,
1,5-pentanediol, methyl 2-hydroxyvalerate, 2-hydroxyvaleric acid, and methyl 2-furancarboxylate. Detailed data are shown in Table S8.†
highest yield of 5-HVA derivatives was 62% (Me-5-HV 55%,
DVL 7%; Table 1, entry 2).
Reusability of Pt/Al2O3 was investigated. First, the used cata-
lyst was collected by centrifugation and used for the next run
without any treatment. In this case, the activity rapidly
decreased during reuses (95% conversion in the first run →
29% conversion in the fourth run; Table S10†). Considering
the rapid deactivation of Pt/Al2O3 in the reaction tests at
higher temperature (Fig. 1 and 2), the catalyst could be also
gradually deactivated by deposition of the organic material
even at this reaction temperature (373 K). Second, another re-
usability test was carried out with regeneration for the recov-
ered catalyst: the recovered catalyst was calcined in air at 573 K
for 1 h before the next run. The calcination temperature was
selected on the basis of the TG-DTA profile of the deactivated
Pt/Al2O3 catalyst (Fig. S2;† exothermic peak top temperature:
539 K). In this case (Fig. 5; details: Table S11†), the conversion
was slightly decreased in the first reuse, and it was almost con-
Fig. 5 Reusability test of Pt/Al2O3 with regeneration by calcination at
573 K for hydrogenolysis of FCA. Reaction conditions: Pt/Al2O3 (4 wt%
stant in the subsequent reuses (95% → 77% → 76% → 78%;
yield of HVA derivatives: 57% → 38% → 36% → 37%). The
XRD patterns of the Pt/Al2O3 catalyst after reduction, after the
first run, and after the 4th run are shown in Fig. 6. The particle
sizes of Pt were slightly increased during the first run (2.1 nm
(b) → 2.5 nm (c); XRD), which can explain the conversion
decrease of the second run after calcination. The particle size
was not changed during the 2–4th run when the recovered cata-
lyst was directly used without calcination (Fig. 6(a)), and it was
Pt), Wcat = 0.050 g, WFCA = 1.12 g, WMeOH = 29 g, P(H2) = 4 MPa at r.t.,
T = 373 K, t = 2 h. Circle: conversion; triangle: carbon balance; red:
methyl 5-hydroxyvalerate (Me-5-HV); green: δ-valerolactone (DVL); dark
blue: methyl tetrahydrofuran-2-carboxylate (MeTHFC); orange: tetra-
hydrofuran-2-carboxylic acid (THFCA); gray: others (methyl valerate,
valeric acid, 1,5-pentanediol, methyl 2-hydroxyvalerate, 2-hydroxyvale-
ric acid, and methyl 2-furancarboxylate). Detailed data are shown in
Table S10.†
very slightly increased during the 2–4th run when reused after 2), no hydrogenolysis product was detected, and the main reac-
regeneration (Fig. 6(e)). TEM images of Pt/Al2O3 catalysts after tion is the esterification with methanol solvent. This indicates
reduction and after reactions are shown in Fig. 7. The Pt par- that THFCA is not an intermediate in the formation of 5-HVA
ticle size estimated from TEM images agreed well with that cal- derivatives from FCA on Pt/Al2O3. This also indicates that the
culated from XRD: 2.4 nm after reduction, 2.9 nm after the ring-opening reaction can proceed before the total hydrogen-
first run, and 3.1 nm after the 4th run with regeneration. ation of the furan ring to tetrahydrofuran ring. In the case of
These changes of the Pt particle size agreed with the change of methyl 2-furancarboxylate (entry 3), the target products, 5-HVA
activity (conversion) of used catalysts after regeneration (Fig. 5 derivatives, were obtained in similar yield (total 46%) to that
and Table S11†). The deactivation in the case without any in the case of the FCA substrate (53%, entry 1). As shown in
treatment was not due to the change of the Pt particle size, but Fig. 2(I), the selectivity to methyl 2-furancarboxylate in the
probably due to the deposition on the catalyst surface as reaction of FCA was very low even at the initial stage. These
described above.
behaviors indicate that the rate of esterification of FCA to
Table 4 shows the results of hydrogenolysis of various methyl 2-furancarboxylate is much slower than the hydrogeno-
related substrates. In the case of THFCA as the substrate (entry lysis of FCA to 5-HVA derivatives. Therefore, the hydrogenolysis
6140 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
the reactivity. In the case of the Al2O3 support where 2-furan-
carboxylic acid and its ester have similar reactivity, the sub-
strates are probably adsorbed with the carbonyl group on
Al2O3. Considering that the reaction media are protic (alcohol
solvent and carboxylic acid substrate) and not strictly anhy-
drous (water is actually produced by the formation of ester or
lactone), the Al–OH site rather than the Lewis acid site is
present and can be the adsorption site for the carbonyl group.
The reactivity of 3-furancarboxylic acid (entry 5) was much
lower than that of FCA, although the C–O hydrogenolysis pro-
ducts, 2-methyl-4-hydroxybutyric acid derivatives, were actually
observed as the lactone form, 2-methyl-γ-butyrolactone. This
result indicates that the position of the carboxyl group in FCA
is essential to the high reactivity. Entries 6–8 show the results
of hydrogenolysis of furan or dihydrofuran. Under the stan-
dard reaction conditions, furan or dihydrofuran was almost
totally converted to THF which is the hydrogenation product
(data not shown). This reaction was very fast and probably
completed during heating to the reaction temperature (373 K).
As shown in Fig. 1, selectivity to hydrogenolysis products can
be improved at higher reaction temperature. Then, we carried
out reaction tests of furan and dihydrofuran by supplying H2
after heating the reaction solution to 373 K under Ar. In the
case of furan (entry 5), 1-butanol (1-BuOH) which is the hydro-
genolysis product was produced with moderate yield (41%),
and THF was still obtained with 40% yield. The yield of
1-BuOH from 2,3- and 2,5-dihydrofurans (entries 7 and 8,
respectively) was much lower than that from furan, suggesting
that the C–O hydrogenolysis of furan mainly occurs before the
partial hydrogenation to dihydrofurans. The main product
from 2,3-dihydrofuran (entry 7) was the acetal (2-methoxytetra-
hydrofuran) produced by the addition of methanol. The high
yield of acetal suggests that acetals are stable under these reac-
tion conditions. A small amount of the C–O hydrogenolysis
product (1-BuOH) was also observed. From 2,5-dihydrofuran,
although the main product was THF, the C–O hydrogenolysis
product (1-BuOH) was formed in a significant amount. One
explanation for the formation of 1-BuOH from 2,5-dihydro-
furan is that 2,3,5-trihydrofuran species which is formed by
addition of one hydrogen atom to either dihydrofuran works
as an intermediate. Another model molecule with an acyclic
structure, n-butyl vinyl ether, was also tested. The vinyl oxy
group can be regarded as a model structure of 1,2,3-positions
of the furan ring. Similarly to furan, hydrogenation of the vinyl
Fig. 6 XRD patterns of Pt/Al2O3 catalysts. (I) Raw patterns, (II) differ-
ence patterns. (a): γ-Al2O3 support without Pt, (b): after reduction with 4
MPa H2 (at r.t.) in methanol at 373 K, (c): after reaction, (d): after the 4th
use (without any treatment, Table S9†), (e): after the 4th use (with regen-
eration: calcination at 573 K for 1 h, Fig. 5 and Table S10†).
of methyl 2-furancarboxylate to 5-HVA derivatives does not group was very fast at low temperature and the main product
involve hydrolysis of the methyl ester. The similar reactivity of was n-butyl ethyl ether under the standard reaction conditions
FCA and methyl 2-furancarboxylate suggests that the acidic (data not shown). We carried out the reaction test of n-butyl
proton in FCA does not play a critical role in the hydrogenoly- vinyl ether by supplying H2 after heating, similarly to the cases
sis of FCA to 5-HVA derivatives over Pt/Al2O3. However, the of furan and dihydrofurans. In this case (entry 9), the C–O
reactivity of methyl 2-furancarboxylate was much lower in the hydrogenolysis product 1-BuOH was actually formed. However,
case of the Pt/CeO2 catalyst (entry 4: conversion = 22%) than the corresponding C2 fragment was hardly detected in both
that of FCA (Table 1, entry 4: conversion = 91%). The CeO2 liquid and gas phases, suggesting that the fragment was poly-
support is a basic material, and thus FCA is much more merized over the catalyst surface.
strongly adsorbed on the CeO2 support at the carboxyl group
Based on these data and the discussion in our previous
than the methyl ester. These reaction results suggest that the papers for the Pt-MoOx/TiO2 catalyst,53,54 we propose the reac-
adsorption of the substrate on the support actually increases tion mechanism of FCA hydrogenolysis to 5-HVA derivatives
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6141

View Article Online
Paper
Green Chemistry
Fig. 7 TEM images of the Pt/Al2O3 catalyst (a) and (d) after reduction with 4 MPa H2 (at r.t.) in methanol at 373 K, (b) and (e) after reaction under
standard reaction conditions, and (c) and (f) after the 4th use with regeneration by calcination at 573 K for 1 h.
over the Pt/Al2O3 catalyst (Fig. 8). The 1,2,3-position of the reactivity trends of FCA and 3-furancarboxylic acid. The con-
furan ring (O–CvC) is the key structure, while the presence sideration of the “direction” effect is also a reason for the
and position of the carboxyl group (or methoxy carbonyl adsorption mode in Fig. 8 where the carboxyl group interacts
group) much affect the reactivity. FCA molecules are adsorbed with the Pt surface. Next, the C–O bond at the 1,2-position is
on Pt with the CvO group (step (i)), and the adsorption is dissociated, and the dienolate species is produced (step (iii)).
assisted by the Al–OH site of the Al2O3 support. This adsorp- Polar solvents can more stabilize the anionic dienolate species
tion mode is proposed based on the following reasons: both than aprotic solvents, and thus polar solvents might promote
carboxyl (–COOH) and methoxy carbonyl (–COOMe) groups in the formation of the dienolate species (step (iii)). The dienolate
the substrate similarly promote the hydrogenolysis reactions; species is then protonated (step (iv)), hydrogenated and des-
the Al2O3 support has a positive effect on the activity although orbed (step (v)) to give 5-HVA. The overall mechanism is
Pt catalysts on inert supports have also activity in the hydroge- similar to the case of the Pt-MoOx/TiO2 catalyst in water
nolysis to some extent; too high H2 pressure decreases the solvent, especially from step (i) to step (iii).54 The most
reaction rate, which can be explained by the competitive different point is the reaction of the dienolate species formed
adsorption of H2 and substrate on the Pt surface. The in step (iii): while the dienolate species is simply protonated
reduction starts with addition of H atoms to the 2-position of and hydrogenated over the Pt/Al2O3 catalyst, over the Pt-MoOx/
the furan ring (step (ii)). Although the addition breaks the con- TiO2 catalyst it is trapped by the Mo4+ site which works as a
jugation between the furan ring and the carboxyl group, the 2-electron reducing agent to further deoxygenate the dienolate
carboxyl group directs the position of the addition of H atoms species, giving the diene species (pentadienoic acid) which is
(the carbon atom which is bonded to the carboxyl group is then hydrogenated to valeric acid.
attacked by hydrogen). The “direction” effect of the carboxyl
The reaction mechanism may be also related to that of fur-
group has been proposed in our previous study of FCA furyl alcohol hydrogenolysis to 1,2- or 1,5-pentanediols over Pt
reduction over the Pt-MoOx/TiO2 catalyst.54 The addition of catalysts combined with basic supports.37–41 Although there is
hydrogen atoms to the 2-position of the furan ring of 3-furan- no widely accepted mechanism, Pt catalysts have been used for
carboxylic acid, which is the first step of the ring-opening of the hydrogenolysis of furfuryl alcohol and the C–O bonds in
3-furancarboxylic acid, is intrinsically much easier than that at O–CvC groups in furfuryl alcohol are dissociated. The Pt-
any position of the furan ring of FCA. On the other hand, the based systems used alcohol solvents, and the systems cannot
actual reactivity of 3-furancarboxylic acid in both hydrogen- open the ring of tetrahydrofurfuryl alcohol. These features are
ation and ring-opening is much lower than that of FCA. The similar to our case of Pt catalysts for FCA hydrogenolysis to
idea that the carbon atom which is bonded to the carboxyl 5-HVA derivatives. The use of a base in furfuryl alcohol hydro-
group is attacked by the hydrogen atom on Pt can explain the genolysis can be explained by the necessity of suppression of
6142 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
Table 4 Hydrogenolysis of various substrates over the Pt/Al2O3 catalyst
Entry
1
Substrate
Conv./%
91
Product (yield/%)
C.B./%
82
2
3
82
95
91
84
4a
22
22
96
90
5
6b
7b
8b
9b
93
90
96
89
>99
92
90
C2H6
(3)
102
Reaction conditions: Pt/Al2O3(4 wt% Pt), Wcat = 0.050 g, substrate = 10 mmol, WMeOH = 29 g, P(H2) = 4 MPa at r.t., T = 373 K, t = 1 h. a Pt/CeO2
(4 wt% Pt) was used as catalyst. b 1 MPa Ar + 3 MPa H2 (H2 was introduced at 373 K). Conv.: conversion, C.B.: carbon balance.
acid-catalyzed polymerization of furfuryl alcohol,37,38 while in
FCA hydrogenolysis similar acid-catalyzed polymerization does
not occur and is not necessary to suppress. Further mechanis-
tic studies on FCA hydrogenolysis and furfural/furfuryl alcohol
hydrogenolysis will uncover the total picture of C–O hydroge-
nolysis of O–CvC groups.
4. Conclusions
2-Furancarboxylic acid (FCA) can be reduced with Pt catalysts
and H2 to give 5-hydroxyvaleric acid (5-HVA) and its derivatives
(ester and δ-valerolactone (DVL)). Solvents and supports have a
Fig. 8 Proposed reaction mechanism of the conversion of FCA to
5-HVA derivatives over the Pt catalyst.
limited effect on the catalysis; nevertheless, polar solvents
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6143

View Article Online
Paper
Green Chemistry
such as methanol and Al2O3 support give better results. The 19 J. C. Serrano-Ruiz, R. Luque and A. Sepúlvega-Escribano,
highest yield of 5-HVA derivatives (55% methyl 5-hydroxyvale- Chem. Soc. Rev., 2011, 40, 5266–5281.
rate and 7% DVL) is obtained under the conditions of metha- 20 M. Besson, P. Gallezot and C. Pinel, Chem. Rev., 2014, 114,
nol solvent, low substrate concentration (10 mmol per29 g 1827–1870.
methanol), 4 MPa (at r.t.) H2 and 373 K. The catalyst is de- 21 X. Zhang, K. Wilson and A. F. Lee, Chem. Rev., 2016, 116,
activated during the reaction probably due to the deposition of 12328–12368.
the organic material on the surface of Pt metal, especially at 22 L. T. Mika, E. Cséfalvay and Á. Németh, Chem. Rev., 2018,
higher temperature. The catalyst can be regenerated by
calcination.
118, 505–613.
23 C. Delhomme, D. Weuster-Botz and F. E. Kühn, Green
Chem., 2009, 11, 13–26.
24 I. Bechthold, K. Bretz, S. Kabasci, R. Kopitzky and
A. Springer, Chem. Eng. Technol., 2008, 31, 647–654.
25 D. M. Alonso, S. G. Wettstein and J. A. Dumesic, Green
Chem., 2013, 15, 584–595.
Conflicts of interest
There are no conflicts of interest to declare.
26 K. Yan, Y. Yang, J. Chai and Y. Lu, Appl. Catal., B, 2015,
179, 292–304.
27 F. D. Pileidis and M.-M. Titirici, ChemSusChem, 2016, 9,
562–582.
28 S. Dutta, I. K. M. Yu, D. C. W. Tsang, Y. H. Ng, Y. S. Ok,
J. Sherwood and J. H. Clark, Chem. Eng. J., 2019, 372, 992–
1006.
Acknowledgements
This work was supported by JSPS KAKENHI Grant Number
18H05247.
29 P. Zhou and Z. Zhang, Catal. Sci. Technol., 2016, 6, 3694–
3712.
References
1 J. H. Clark, V. Budarin, F. E. I. Deswarte, J. J. E. Hardy, 30 Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal.,
F. M. Kerton, A. J. Hunt, R. Luque, D. J. Macquarrie, 2013, 3, 2655–2668.
K. Milkowski, A. Rodriguez, O. Samuel, S. J. Tavener, 31 R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sádaba and
R. J. White and A. J. Wilson, Green Chem., 2006, 8, 853–860. M. L. Granados, Energy Environ. Sci., 2016, 9, 1144–1189.
2 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 32 X. Li, P. Jia and T. Wang, ACS Catal., 2016, 6, 7621–7640.
2411–2502.
33 K. Gupta, R. K. Rai and S. K. Singh, ChemCatChem, 2018,
10, 2326–2349.
34 S. Chen, R. Wojcieszak, F. Dumeignil, E. Marceau and
S. Royer, Chem. Rev., 2018, 118, 11023–11117.
3 F. Xie, E. Polletb, P. J. Halley and L. Avérous, Prog. Polym.
Sci., 2013, 38, 1590–1628.
4 R. A. Sheldon, Green Chem., 2014, 16, 950–963.
5 C. Chatterjee, F. Pong and A. Sen, Green Chem., 2015, 17, 35 Y. Nakagawa, M. Tamura and K. Tomishige, Catal. Surv.
40–71.
Asia, 2015, 19, 249–256.
6 R. A. Sheldon, ACS Sustainable Chem. Eng., 2018, 6, 32–48.
7 H. Li, A. Riisager, S. Saravanamurugan, A. Pandey,
R. S. Sangwan, S. Yang and R. Luque, ACS Catal., 2018, 8,
148–187.
8 G. D. Bhowmick, A. K. Sarmah and R. Sen, Bioresour.
Technol., 2018, 247, 1144–1154.
36 F. Liu, Q. Liu, J. Xu, L. Li, Y.-T. Cui, R. Lang, L. Li, Y. Su,
S. Miao, H. Sun, B. Qiao, A. Wang, F. Jérome and T. Zhang,
Green Chem., 2018, 20, 1770–1776.
37 T. Mizugaki, T. Yamakawa, Y. Nagatsu, Z. Maeno,
T. Mitsudome, K. Jitsukawa and K. Kaneda, ACS Sustainable
Chem. Eng., 2014, 2, 2243–2247.
9 T. Iwata, Angew. Chem., Int. Ed., 2015, 54, 3210–3215.
10 X. Zhang, M. Fevre, G. O. Jones and R. M. Waymouth,
Chem. Rev., 2018, 118, 839–885.
11 M. Wang, J. Ma, H. Liu, N. Luo, Z. Zhao and F. Wang, ACS
Catal., 2018, 8, 2129–2165.
38 R. Ma, X.-P. Wu, T. Tong, Z.-J. Shao, Y. Wang, X. Liu, Q. Xia
and X.-Q. Gong, ACS Catal., 2017, 7, 333–337.
39 T. Tong, Q. Xia, X. Liu and Y. Wang, Catal. Commun., 2017,
101, 129–133.
40 T. Tong, X. Liu, Y. Guo, M. N. Banis, Y. Hu and Y. Wang,
J. Catal., 2018, 365, 420–428.
12 C. K. Williams, Chem. Soc. Rev., 2007, 36, 1573–1580.
13 F. H. Isikgor and C. R. Becer, Polym. Chem., 2015, 6, 4497– 41 W. Xu, H. Wang, X. Liu, J. Ren, Y. Wang and G. Lu, Chem.
4559.
Commun., 2011, 47, 3924–3926.
14 M. Hong and E. Y.-X. Chen, Nat. Chem., 2016, 8, 42–49.
15 G. Z. Papageorgiou, D. G. Papageorgiou, Z. Terzopoulou
and D. N. Bikiaris, Eur. Polym. J., 2016, 83, 202–229.
16 Y. Ren, Z. Wei, T. Wu, Y. Bian, X. Leng, C. Zhou and Y. Li,
RSC Adv., 2016, 6, 45791–45801.
42 B. Zhang, Y. Zhu, G. Ding, H. Zheng and Y. Li, Green
Chem., 2012, 14, 3402–3409.
43 H. Liu, Z. Huang, H. Kang, C. Xia and J. Chen,
Chin. J. Catal., 2016, 37, 700–710.
44 H. Liu, Z. Huang, F. Zhao, F. Cui, X. Li, C. Xia and J. Chen,
Catal. Sci. Technol., 2016, 6, 668–671.
17 P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538–1558.
18 J. J. Bozell and G. R. Petersen, Green Chem., 2010, 12, 539– 45 S. Liu, Y. Amada, M. Tamura, Y. Nakagawa and
554.
K. Tomishige, Catal. Sci. Technol., 2014, 4, 2535–2549.
6144 | Green Chem., 2019, 21, 6133–6145
This journal is © The Royal Society of Chemistry 2019

View Article Online
Green Chemistry
Paper
46 S. Liu, Y. Amada, M. Tamura, Y. Nakagawa and 64 G. Chen, S. Wu, H. Liu, H. Jiang and Y. Li, Green Chem.,
K. Tomishige, Green Chem., 2014, 16, 617–626. 2013, 15, 230–235.
47 K. Tomishige, Y. Nakagawa and M. Tamura, Green Chem., 65 W. Zhong, H. Liu, C. Bai, S. Liao and Y. Li, ACS Catal.,
2017, 19, 2876–2924. 2015, 5, 1850–1856.
48 S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori 66 A. S. Touchya and K. Shimizu, RSC Adv., 2015, 5, 29072–29075.
and K. Tomishige, Chem. Commun., 2009, 2035–2037.
49 S. Koso, N. Ueda, Y. Shinmi, K. Okumura, T. Kizuka and
K. Tomishige, J. Catal., 2009, 267, 89–92.
50 M. Sajid, X. Zhao and D. Liu, Green Chem., 2018, 20, 5427–
5453.
67 H. Miura, K. Nakahara, T. Kitajima and T. Shishido, ACS
Omega, 2017, 2, 6167–6173.
68 Z.-z. Zhou, M. Liu, L. Lv and C.-J. Li, Angew. Chem., Int. Ed.,
2018, 57, 2616–2620.
69 D. Nandan, G. Zoppellaro, I. Medřík, C. Aparicio, P. Kumar,
M. Petr, O. Tomanec, M. B. Gawande, R. S. Varma and
R. Zbořil, Green Chem., 2018, 20, 3542–3556.
51 Z. Zhang and G. W. Huber, Chem. Soc. Rev., 2018, 47, 1351–
1390.
52 M. Douthwaite, X. Huang, S. Iqbal, P. J. Miedziak, 70 Y. Takeda, M. Tamura, Y. Nakagawa, K. Okumura and
G. L. Brett, S. A. Kondrat, J. K. Edwards, M. Sankar, K. Tomishige, Catal. Sci. Technol., 2016, 6, 5668–5683.
D. W. Knight, D. Bethell and G. J. Hutchings, Catal. Sci. 71 R. Pinkos, C. Bauduin, A. Paul, G. Fritz and H. Wagner, US
Technol., 2017, 7, 5284–5293. patent US2011/0237806A1, 2011.
53 T. Asano, M. Tamura, Y. Nakagawa and K. Tomishige, 72 H. A. Smith and J. F. Fuzek, J. Am. Chem. Soc., 1949, 71,
ACS Sustainable Chem. Eng., 2016, 4, 6253–6257. 415–419.
54 T. Asano, Y. Nakagawa, M. Tamura and K. Tomishige, ACS 73 W. E. Kaufmann and R. Adams, J. Am. Chem. Soc., 1923, 45,
Sustainable Chem. Eng., 2019, 7, 9601–9612. 3029–3044.
55 T. R. Boussie, E. L. Dias, Z. M. Fresco and V. J. Murphy, US 74 A. Miyagawa, Y. Nakagawa, M. Tamura and K. Tomishige,
patent US2010/0317822A1, 2010. Mol. Catal., 2019, 471, 60–70.
56 M. J. Gilkey, A. V. Mironenko, D. G. Vlachos and B. Xu, ACS 75 T. Mizugaki, T. Yamakawa, R. Arundhathi, T. Mitsudome,
Catal., 2017, 7, 6619–6634. K. Jitsukawa and K. Kaneda, Chem. Lett., 2012, 41, 1720–1722.
57 M. J. Gilkey, R. Balakumar, D. G. Vlachos and B. Xu, Catal. 76 M. Sebek, J. Holz, A. Börner and K. Jähnisch, Synlett, 2009,
Sci. Technol., 2018, 8, 2661–2671. 461–465.
58 L. Wei, J. Zhang, W. Deng, S. Xie, Q. Zhang and Y. Wang, 77 M. Maris, W.-R. Huck, T. Mallat and A. Baiker, J. Catal.,
Chem. Commun., 2019, 55, 8013–8016. 2003, 219, 52–58.
59 G.-J. ten Brink, I. W. C. E. Arends and R. A. Sheldon, Chem. 78 M. Maris, T. Mallat, E. Orglmeister and A. Baiker, J. Mol.
Rev., 2004, 104, 4105–4124. Catal. A: Chem., 2004, 219, 371–376.
60 T. Hara, M. Hatakeyama, A. Kim, N. Ichikuni and 79 X. Zhang, L. Chen, Y. Li, H. Li, Z. Xie, Q. Kuang and
S. Shimazu, Green Chem., 2012, 14, 771–777. L. Zheng, Dalton Trans., 2019, 48, 5515–5519.
61 P. K. Saikia, P. P. Sarmah, B. J. Borah, L. Saikia, K. Saikia 80 S. Noël, D. Bourbiaux, N. Tabary, A. Ponchel, B. Martel,
and D. K. Dutta, Green Chem., 2016, 18, 2843–2850.
62 T. Mitsudome, A. Noujima, T. Mizugaki, K. Jitsukawa and
K. Kaneda, Green Chem., 2009, 11, 793–797.
E. Monflier and B. Léger, Catal. Sci. Technol., 2017, 7, 5982–
5992.
81 R. S. Gates, M. Hsu and E. E. Klaus, Tribol. Trans., 1989, 32,
357–363.
63 M. S. Ide and R. J. Davi, J. Catal., 2013, 308, 50–59.
This journal is © The Royal Society of Chemistry 2019
Green Chem., 2019, 21, 6133–6145 | 6145