Hydrogenolysis of tetrahydrofuran-2-carboxylic acid over tungsten-modified rhodium catalyst

Article abstract of DOI:10.1016/j.apcata.2020.117723

Catalysts for reduction of tetrahydrofuran-2-carboxylic acid (THFCA), which can be synthesized from furfural via oxidation and hydrogenation, were explored among the combinations of noble metal and reducible metal oxide supported on SiO₂. Rh-WO_x/SiO₂ catalysts showed activity in C-O hydrogenolysis at 2-position of THFCA (to δ-valerolactone and 5-hydroxyvaleric acid) and higher yield ratio of these C-O hydrogenolysis products to carboxylic acid hydrogenation products than other bimetallic catalysts. The activity of Rh-WO_x/SiO₂ catalysts was highest at W/Rh = 0.25 mol/mol. XRD, TPR, CO adsorption and XAFS characterizations showed that the Rh-WO_x/SiO₂ (W/Rh = 0.25) catalyst contained Rh metal particles with surface modification with isolated W²⁺ oxide species. The mechanism that hydride-like species formed on Rh atom attacks the C atom at the α-position (2-position) of adsorbed carboxylate on W atom is proposed based on the similar kinetics and similar catalyst structure to Rh-MO_x/SiO₂ (M = Re, Mo) which is known to be active in THFA hydrogenolysis to 1,5-pentanediol.

Full text of DOI:10.1016/j.apcata.2020.117723

Journal Pre-proof
Hydrogenolysis of tetrahydrofuran-2-carboxylic acid over
tungsten-modified rhodium catalyst
Takehiro Asano (Conceptualization) (Methodology) (Investigation)
(Writing - original draft), Yoshinao Nakagawa (Conceptualization)
(Methodology) (Writing - review and editing), Masazumi Tamura
(Formal analysis), Keiichi Tomishige (Supervision) (Writing - review
and editing)
PII:
S0926-860X(20)30316-1
DOI:
https://doi.org/10.1016/j.apcata.2020.117723
APCATA 117723
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
2 May 2020
22 June 2020
24 June 2020
Please cite this article as: Asano T, Nakagawa Y, Tamura M, Tomishige K, Hydrogenolysis of
tetrahydrofuran-2-carboxylic acid over tungsten-modified rhodium catalyst, Applied Catalysis
A, General (2020), doi: https://doi.org/10.1016/j.apcata.2020.117723
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Hydrogenolysis of tetrahydrofuran-2-carboxylic acid over tungsten-
modified rhodium catalyst
Takehiro Asano,a Yoshinao Nakagawa,a,b,* Masazumi Tamura,c Keiichi Tomishigea,b,*
a
Department of Applied Chemistry, School of Engineering, Tohoku University, 6-6-07 Aoba,
Aramaki, Aoba-ku, Sendai, 980-8579, Japan
b Research Center for Rare Metal and Green Innovation, Tohoku University, 468-1, Aoba, Aramaki,
Aoba-ku, Sendai 980-0845, Japan
c
Research Center for Artificial Photosynthesis, The Advanced Research Institute for Natural
Science and Technology, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi, Osaka, 558-8585,
Japan
* Corresponding authors.
yoshinao@erec.che.tohoku.ac.jp, Tel/Fax: +81-22-795-7215 (Y. Nakagawa)
tomi@erec.che.tohoku.ac.jp, Tel&Fax: +81-22-795-7214 (K. Tomishige)
Graphical abstract
1

Highlights




Tetrahydrofuran-2-carboxylic acid (THFCA) was reduced with M-M’Ox/SiO2 catalysts.
Rate ratio of C-O hydrogenolysis/carboxyl group hydrogenation is high over Rh-WOx.
Rh-WOx/SiO2 has Rh metal particles partially covered with monomeric W2+ species.
Similar hydrogenolysis mechanism of THFCA and 2-functionalized alcohol is proposed.
Abstract
Catalysts for reduction of tetrahydrofuran-2-carboxylic acid (THFCA), which can be synthesized
from furfural via oxidation and hydrogenation, were explored among the combinations of noble
metal and reducible metal oxide supported on SiO2. Rh-WOx/SiO2 catalysts showed activity in C-
O hydrogenolysis at 2-position of THFCA (to -valerolactone and 5-hydroxyvaleric acid) and
higher yield ratio of these C-O hydrogenolysis products to carboxylic acid hydrogenation products
than other bimetallic catalysts. The activity of Rh-WOx/SiO2 catalysts was highest at W/Rh=0.25
mol/mol. XRD, TPR, CO adsorption and XAFS characterizations showed that the Rh-WOx/SiO2
(W/Rh=0.25) catalyst contained Rh metal particles with surface modification with isolated W2+
oxide species. The mechanism that hydride-like species formed on Rh atom attacks the C atom at
the -position (2-position) of adsorbed carboxylate on W atom is proposed based on the similar
kinetics and similar catalyst structure to Rh-MOx/SiO2 (M=Re, Mo) which is known to be active
in THFA hydrogenolysis to 1,5-pentanediol.
Keywords: Hydrogenolysis; Rhodium; Tungsten; Ring opening; Carboxylic acid
2

1. Introduction
Reduction with hydrogen, especially selective C-O hydrogenolysis, is an important reaction in
biomass utilization, because most bulk chemicals have lower oxygen content and smaller number
of C-O bonds than raw biomass or biomass-derived platform chemicals which can be synthesized
and isolated from biomass [1-4]. Major platform chemicals are sugar alcohols, furans and
carboxylic acids [5-7]. Furfural is produced in large scale by depolymerization and dehydration of
hemicellulose in abundant lignocellulose [8,9]. Various compounds such as furfuryl alcohol,
tetrahydrofurfuryl
alcohol
(THFA),
2-methylfuran,
pentanediols
and
cyclopentanone/cyclopentanol can be catalytically synthesized by furfural reduction with
hydrogen [10,11]. These compounds can be further converted into useful chemicals and polymers.
On the other hand, oxidation of furfural can produce 2-furancarboxylic acid in good yield [12-17];
however, the demand of 2-furancarboxylic acid is not large and the conversion of 2-
furancarboxylic acid to useful chemicals has been less investigated than that of furfural or furfuryl
alcohol. Reductive conversion of 2-furancarboxylic acid has been investigated by limited groups
including ours. Reduction of carboxyl group has been reported by using homogeneous complex
catalysts [18-20] and heterogeneous metal catalysts [21]; however, these reactions are less
attractive because the products (furfuryl alcohol or THFA) can be more easily synthesized by
furfural hydrogenation. The products with retention of the carboxyl group are more attractive. We
reported the hydrodeoxygenation of 2-furancarboxylic acid to valeric acid over Pt-MoOx/TiO2
catalyst [22,23]. We also reported the reductive conversion of 2-furancarboxylic acid to 5-
hydroxyvaleric acid and its derivatives (-valerolactone and ester with the alcohol solvent) with
H2 over Pt/Al2O3 catalyst [24]. At almost the same time, Haichao Liu et al. individually reported
Pt/SiO2-catalyzed reduction of 2-furancarboxylic acid with H2 to 5-hydroxyvaleric acid [25]. All
3

of these systems using Pt as active metal involved opening of furan ring before hydrogenation to
tetrahydrofuran ring, and tetrahydrofuran-2-carboxylic acid (THFCA), ring-hydrogenation
product of 2-furancarboxylic acid, was not reactive and left as by-product. Synthesis of THFCA
from 2-furancarboxylic acid by hydrogenation in high yield is possible, by using Pd catalyst [24-
27]. Therefore, conversion of THFCA to useful chemical has importance in biomass refinery.
Possible reduction products of THFCA are summarized in Figure 1. Reduction of carboxyl
group of THFCA gives tetrahydrofurfuryl alcohol (THFA). THFA can be further reduced to 1,5-
pentanediol (1,5-PeD), 1,2-pentanediol (1,2-PeD) and 2-methyltetrahydrofuran via selective C-O
hydrogenolysis. Production of THFA or these THFA-derived compounds has been already
intensively investigated using furfural as the reactant, and a number of systems with high
selectivity to one of these compounds have been reported [10,11]. This is a similar situation to the
reduction of 2-furancarboxylic acid: the reduction of carboxyl group does not give attractive
products. On the other hand, hydrogenolysis of C-O bond in the tetrahydrofuran ring gives 5-
hydroxyvaleric acid (C-O hydrogenolysis at 2-position (-position of carboxyl group)) or 2-
hydroxyvaleric acid (C-O hydrogenolysis at 5-position). In particular, 5-hydroxyvaleric acid (5-
HVA) is an attractive target because it is an ,-difunctionalized compound that might be used as
a monomer of resin. The lactone form of 5-HVA, -valerolactone (DVL), can be reversibly formed
from 5-HVA, and it can be also regarded as a target product.
Selective C-O hydrogenolysis of an -functionalized carboxylic acid at -position is a difficult
reaction. Hydrogenolysis of THFCA to 5-HVA as the main target has not been investigated in the
literature. As a related reaction, lactic acid deoxygenation to propanoic acid has been reported in
a few papers. Xinli Li et al. reported hydrodeoxygenation of lactic acid to propanoic acid over iron
oxide catalyst at very high temperature around 663 K with moderate yield (maximum 46.7%) [28].
4

Deoxygenation of lactic acid to propanoic acid in H2-free conditions over molybdenum-based
catalysts has been reported at similar high temperature with ≤61% yield [29,30]. There are other
several biomass-derived compounds with an -functionalized carboxylic acid moiety such as
malic acid, citric acid and sugar acids. The development of selective C-O hydrogenolysis at -
position of carboxylic acid is meaningful, but not achieved.
On the other hand, selective C-O hydrogenolysis of -position of -CH2OH group, for example
THFA and glycerol hydrogenolysis to 1,5-pentanediol and 1,3-propanediol, respectively, has been
well investigated typically by using metal-oxide(M’Ox)-modified noble metal(M) catalysts [31-
38]. In particular, the combinations of Rh or Ir as M and Re or Mo as M’ have been reported to be
effective [39-53]. Rh metal has higher intrinsic catalytic activity in C-O hydrogenolysis [54] than
Pt and Pd which were used in the conversion of 2-furancarboxylic acid to THFCA as main product
or by-product. The activity and regioselectivity of Rh in C-O hydrogenolysis is much enhanced by
the combination with M’ (Re or Mo) [35]. The C-O hydrogenolysis activity of Ir, which is located
just below Rh in the periodic table, is also much enhanced by the combination with M’[35]. These
Rh- and Ir-based bimetallic catalysts have the common structure that noble metal particles (M; Rh
or Ir) are partially covered with reduced M’Ox species via direct M-M’ bond, and the interface is
proposed to be the active site. This type of catalysts tends to show high activity in carboxylic acid
hydrogenation, for example, the Rh-MoOx catalysts are an effective one for the hydrogenation of
amino acids to the corresponding amino alcohols [55,56]. The Ru-MoOx catalysts are effective for
the hydrogenation of lactic acid to 1,2-propanediol [57] and levulinic acid to 1,4-pentanediol [58].
Levulinic acid hydrogenation to 1,4-pentanediol has been also reported with other Mo-modified
noble metal catalysts [59,60].
In this study, we explored the catalysts for the conversion of THFCA to 5-HVA or its derivatives
5

among M-M’Ox/SiO2 catalysts. In this target reaction, the carboxyl groups should be maintained
in the product; however, this type of catalysts tend to show high activity in the hydrogenation of
carboxyl groups. Therefore, it is important to develop the catalysts with high activity in the ring-
opening C-O hydrogenolysis and low activity in the hydrogenation of carboxyl groups. We found
that Rh-WOx/SiO2 catalyst showed relatively good selectivity. Although the promoting effect of
W addition to Rh catalyst in C-O hydrogenolysis has been known for more than 15 years [61],
recent studies of Rh-M bimetallic catalysts more focused on Rh-Mo and Rh-Re combinations
because the activity of Rh-W catalysts in most hydrogenolysis reactions is lower than that of Rh-
Mo and Rh-Re combinations [40,41,62-65]. Because of the lower activity, the characterization of
supported Rh-W bimetallic catalysts was hardly carried out in the literature. Here we characterized
Rh-WOx/SiO2 catalyst with various techniques to elucidate the active site structure.
2. Experimental section
2.1.Catalyst preparation
M-M’Ox/SiO2 (4 wt% M, M’/M=0.25, M=Rh, Pd, Pt, Ru, Ir, M’=none, Mo, Re, W) catalysts
were prepared by the same sequential impregnation method as the previous reports [41,62]: SiO2
(CARiACT G6, 475 m2/g, Fuji Silysia, calcined at 973 K for 1 h) was used for support, and the
following metal precursors were used: RhCl3·3H2O (Wako), Pd(NO3)2 (NE chemcat),
Pt(NO2)2(NH3)2 in HNO3aq (Tanaka Kikinzoku Kogyo), Ru(NO)(NO3)3-x(OH)x (Sigma-Aldrich),
H2IrCl6 in HClaq (Furuya Metals), (NH4)10W12O40·xH2O (Sigma-Aldrich), (NH4)6Mo7O24·4H2O
(Wako), and NH4ReO4 (Mitsuwa Chemicals). The catalysts after impregnation and drying were
calcined at 773 K for 3 h.
6

2.2.Activity tests
An autoclave with 190 mL volume was used as reactor. A glass vessel was inserted into the
reactor in order to avoid the contact of the reaction mixture with the stainless-steel wall of the
reactor. In the cases except Ru and Ir-based catalysts, the catalyst, solvent, and substrate (as
received) as well as a magnetic spinner were put into the glass vessel. In the case of typical catalyst
(Rh-WOx/SiO2, W/Rh=0.25), the mixture of water (19 g) + catalyst (0.1 g) had pH of 4.3, and the
addition of THFCA substrate (10 mmol) decreased the pH value to 2.0. The autoclave was purged
three times with 1 MPa H2 and then heated to 413 K. The H2 pressure became 1.5 MPa when the
autoclave was heated to 413 K (standard reaction temperature; about 30 min heating). The pressure
was set to the desired value by adding or drawing out H2 if necessary. The spinning rate was 500
rpm. The catalysts were reduced during the heating.
In the cases of Ru and Ir catalysts, the reaction temperature (413 K) may not be enough to totally
reduce the active metal [44]. Therefore, the catalyst, solvent and spinner without substrate were
first loaded to the autoclave and they were heated at 473 K under 3 MPa H2 (2 MPa at r.t.) for 1 h
to reduce the catalyst. After the reduction treatment, the mixture was cooled, and the substrate was
quickly added to the mixture under air at room temperature. The autoclave was purged with H2
and then heated to the reaction temperature (413 K) in a similar manner to the normal runs as
described above.
After the reaction time (0-48 h; typically 4 h), the autoclave was cooled with water and the gases
were collected into a gas bag. The glass vessel and its contents were washed with 1,4-dioxane. The
internal standard substance (2-methoxyethanol, 10 mmol) was added, and mixture was transferred
to a vial. The products in the liquid phase after filtering out the catalyst were analyzed by FID-GC
with an HP-FFAP capillary column, and those in the gas phase were analyzed by two FID-GCs
7

equipped with an Rtx-1 PONA capillary column (analysis of organic carbons such as small
alkanes) or a Porapak N packed column in combination with a methanator (analysis of CO and
CO2).
Conversion, yield, and selectivity were calculated by the following equations.
[
]
푎푚표푢푛푡 표푓 푑푒푡푒푐푡푒푑 푠푢푏푠푡푟푎푡푒 푚푚표푙
[ ]
conversion % = (1 −
) × 100
[
]
푎푚표푢푛푡 표푓 푙표푎푑푒푑 푠푢푏푠푡푟푎푡푒 푚푚표푙
[
]
푎푚표푢푛푡 표푓 푝푟표푑푢푐푡 푚푚표푙 ×푛푢푚푏푒푟 표푓 푐푎푟푏표푛 푎푡표푚푠 ꢀ푛 푡ℎ푒 푝푟표푑푢푐푡 푚표푙푒푐푢푙푒
yield [%] =
×
[
]
푎푚표푢푛푡 표푓 푙표푎푑푒푑 푠푢푏푠푡푟푎푡푒 푚푚표푙 ×푛푢푚푏푒푟 표푓 푐푎푟푏표푛 푎푡표푚푠 ꢀ푛 푎 푠푢푏푠푡푟푎푡푒 푚표푙푒푐푢푙푒
100
selectivity [%] = 푐표푛푣푒푟푠ꢀ표푛 [%] × 100
푦ꢀ푒푙푑 [%]
Carbon balance was calculated by the comparison of the sum of carbon atoms in the detected
substrate-derived molecules before and after reaction, defined as follows:
carbon balance [%]
∑
[
]
(푎푚표푢ꢂ푡 표푓 푑푒푡푒푐푡푒푑 푚표푙푒푐푢푙푒 푚푚표푙 × ꢂ푢푚푏푒푟 표푓 푐푎푟푏표ꢂ 푎푡표푚푠 ꢁꢂ 푡ℎ푒 푚표푙푒푐푢푙푒)
=
[
]
푎푚표푢ꢂ푡 표푓 푙표푎푑푒푑 푠푢푏푠푡푟푎푡푒 푚푚표푙 × ꢂ푢푚푏푒푟 표푓 푐푎푟푏표ꢂ 푎푡표푚푠 ꢁꢂ 푎 푠푢푏푠푡푟푎푡푒 푚표푙푒푐푢푙푒
× 100
2.3.Catalyst characterization
X-ray diffraction (XRD) patterns were recorded by Rigaku MiniFlex60 diffractometer (Cu Kα
(λ = 0.154 nm), 45 kV, 40 mA) under air. The particle (crystallite) size (d) of Rh metal was
calculated by the Scherrer’s equation using the (111) peak of fcc Rh at around 41°.
The CO adsorption amount was measured with MicrotracBEL BELCAT II instrument. About
50 mg of the calcined catalyst sample was pre-reduced with H2 at 473 K for 30 min. The flow gas
was changed to He, and the sample tube was purged at 473 K for 30 min. Subsequently the sample
tube was cooled to 308 K, and adsorption was performed by introducing 10% CO/He pulses (0.888
cm3/pulse). The pulse introductions continued until the TCD signals of last three pulses became
8

constant (difference < 1%).
Temperature-programmed reduction with H2 (H2-TPR) was measured with a home-made
apparatus composed of flow controllers, glass tube reactor, frozen acetone trap and TCD. The
sample amount was about 50 mg. The sample was heated under 5% H2/Ar with 30 mL/min flow
rate from 293 K to 1123 K at heating rate of 10 K/min.
The XAFS spectra were measured at the BL14B2 station at SPring-8, JASRI. The proposal
numbers for the measurement were 2018B1805 and 2019A1716. The catalyst samples after
reduction or reaction were put into UV/vis cells with 10 mm optical path under inert atmosphere
(glove bag). Detectors (ion chambers) for I0 were filled with 100% N2 and 100% Ar for W L3-edge
and Rh K-edge measurement, respectively. Detectors for I1 were filled with 70% N2 + 30% Ar and
70% Ar + 30% Kr for W and Rh, respectively. The data were mainly collected in a transmission
mode (edge jumps: 0.3-1.0) except W L3 edge measurement of the sample with very low W amount
(Rh/W=0.13) where a fluorescence mode with 19-SSD was used. The data processing and curve
fitting were performed using REX2000 ver 2.6.0 program (Rigaku). For X-ray absorption near
edge structure (XANES) analysis, the raw data were first processed by subtraction of a modified
Victoreen equation and normalization with the edge height. The intensity of white line, which is
the first absorption peak in XANES, was determined by fitting with one arctangent and one
Gaussian functions, which represent baseline and white line, respectively. For extended X-ray
absorption fine structure (EXAFS) analysis, the oscillation was extracted from the raw spectrum
with a spline smoothing and then processed by k3-weighting and Fourier transformation. In the
curve fitting, the curves and the empirical phase shifts for W-O and Rh-Rh shells (bonds) were
extracted from the data of Na2WO4 and Rh foil, respectively. Curves for the W-Rh, Rh-W and W-
W shells (bonds) were theoretically obtained with FEFF 8.20 program.
9

ICP-AES measurement for determination of leached metal amount into the reaction media was
carried out with ThermoFisher iCAP6500 instrument.
3. Results and discussion
3.1.Hydrogenolysis of THFCA over various catalysts
SiO2 was used as support for catalysts because the modification effect of additive metal species
occurs easily on this inert support. Table 1 lists the results of the catalyst survey in the reaction of
THFCA over noble metal catalysts (M/SiO2, M=Rh, Pd, Pt, Ru, Ir) without and with the
modification with M’Ox (M’=W, Re, Mo) at the fixed loading amount of M (4 wt%) and the fixed
molar ratio (M’/M=0.25). Full yield data are shown in Table S1 (Supplementary Material). Among
monometallic catalysts (Entries 1, 5, 9, 13, 17), Rh, Ru, and Ir catalysts gave the formation of
DVL; however, the activity was rather low. The addition of M’ to Rh/SiO2 increased significantly
the conversion and the yield of DVL (+5-HVA). In most runs showing significant activity, the loss
of carbon balance (100% – carbon balance [%]) was not negligible in comparison with the total
yield of detected compounds, and it is difficult to calculate exact selectivity values. Therefore, we
discuss the data based on the yield values of each compound. Considering the yield ratio of DVL
(+ 5-HVA) to THFA+1,5-PeD+1-PeOH, which are the products by the reactions containing the
hydrogenation of carboxyl group, the ratio of Rh-WOx/SiO2 (Entry 2, 0.55) is clearly higher than
those of Rh-ReOx/SiO2 (Entry 3, 0.37) and Rh-MoOx/SiO2 (Entry 4, 0.34). At the same time, other
combinations such as Pt-WOx/SiO2 (Entry 10) and Pt-MoOx/SiO2 (Entry 12) gave relatively high
yield of DVL and the yield ratio of DVL (+ 5-HVA) to THFA+1,5-PeD+1-PeOH on Pt-WOx/SiO2
(Entry 10) and Pt-MoOx/SiO2 (Entry 12) was 0.28 and 0.18, respectively, which were clearly lower
than that on Rh-WOx/SiO2 (Entry 2, 0.55). Based on the comparison from the viewpoint of the
10

yield of DVL (+ 5-HVA) and the yield ratio of DVL (+ 5-HVA) to THFA+1,5-PeD+1-PeOH, we
used Rh-WOx/SiO2 in the following studies. Ru- and Ir-based catalysts showed very low activity
(low conversion) even with the modification of W, Re or Mo (Entries 13-20), except in the case of
Ru-ReOx/SiO2 (Etnry 15) which showed some activity in only carboxylic acid hydrogenation. The
activity of Ru-Re catalyst and lower activity of Ir-Re catalyst than other M-Re catalysts (M: noble
metal) in carboxylic acid hydrogenation have been reported in the literature [66].
Table 2 (details: Table S2 in Supplementary Material) lists the effect of W/Rh in Rh-WOx/SiO2
in the reaction of THFCA. The addition of W to Rh/SiO2 clearly enhanced the conversion and the
yield of DVL (+ 5-HVA). The yield of DVL (+ 5-HVA) showed a volcano-curve dependence on
W amount and the maximum yield was obtained at W/Rh=0.25 (Entry 4). The conversion and total
yield values showed also near-volcano-type-dependence with the maximum at W/Rh=0.25,
although there were some fluctuations. Similar volcano-curve dependence of activity on W amount
has been reported for hydrogenolysis of glycerol [62] and 1,2-propanediol [63] over the same type
of catalysts, although the optimized W amount was slightly different. The decrease in the
conversion and the yield of DVL (+ 5-HVA) can be explained by the decrease of the number of
the active sites due to the covering of Rh metal surface with WOx species. Such phenomena of
activity decrease due to the covering of active sites have been reported for C-O hydrogenolysis by
Rh-MoOx/SiO2 catalysts [41,67]. An important point is that the yield ratio of DVL (+ 5-HVA) to
THFA+1,5-PeD+1-PeOH increased monotonously in the range of W/Rh from zero to 0.5, and it
was also saturated at above W/Rh=0.5. It is also demonstrated that the catalytic performance of
Rh-WOx/SiO2 (W/Rh=0.25) was higher than that of monometallic Rh/SiO2 (Entry 1) and
WOx/SiO2 catalysts, suggesting that the synergy between Rh and WOx can be connected to the
formation of DVL (+ 5-HVA). In order to make the relation between the synergetic effect and the
11

distance of Rh-WOx, physical mixture of Rh/SiO2 + WOx/SiO2 was also evaluated (Table 3;
details: Table S3 in Supplementary Material), where the loading amount of WOx in WOx/SiO2 was
adjusted to the same amount of Rh-WOx/SiO2 (W/Rh=0.25). The addition of WOx/SiO2 to Rh/SiO2
(Entry 4) also enhanced the conversion and the yield of DVL (+ 5-HVA) in comparison with those
on Rh/SiO2 (Entry 2); however, the promoting effect by physical mixing is smaller than the case
of the direct addition of WOx to Rh/SiO2 (Rh-WOx/SiO2). This behavior can be explained by the
partial leaching of WOx species from WOx/SiO2 and re-deposition on Rh/SiO2. In particular,
tungsten oxide species with high oxidation state has solubility in acidic water [68], and W species
tends to be leached before the reduction. In fact, the ICP-AES analysis of the mixture of unreduced
Rh-WOx/SiO2 (W/Rh=0.25) catalyst and aqueous THFCA solution after heating at 413 K under
Ar atmosphere for 0.5 h showed significant leaching: 29% of total W and 5% of total Rh were
dissolved. On the other hand, the ICP-AES analysis of the solution using Rh-WOx/SiO2
(W/Rh=0.25) catalyst after C-O hydrogenolysis reaction of THFCA for 4 h showed slight leaching
of W (2% of total W) and negligible leaching of Rh (<0.1% of total Rh). These elemental analysis
results indicate the leaching and re-deposition of some WOx species at the initial stage of the
catalytic reaction. As a result, although there are some mobile WOx species, the direct addition of
WOx to Rh/SiO2 is more suitable than that by physical mixture, and the distance between Rh metal
particles and WOx species is important and the synergy is suggested to be caused by the direct
interaction of WOx species with Rh metal surface.
3.2.Effect of reaction conditions in THFCA hydrogenolysis over Rh-WOx/SiO2 (W/Rh=0.25)
The reaction time dependence of the THFCA hydrogenolysis over Rh-WOx/SiO2 is shown in
Table 4 (details: Table S4 in Supplementary Material). The conversion and the yield of DVL (+ 5-
12

HVA) increased monotonously with the reaction time; however, the increase became slower at
longer reaction time. The conversion-reaction time curve can be fitted by second-order reaction
(Fig. S1 in Supplementary Material), but such reaction order is unlikely. In fact, as described later,
the reaction order with respect to THFCA concentration was much lower than 1. The discrepancy
suggests the gradual deactivation of the catalyst. The selectivity to DVL (+ 5-HVA) was 18% at
the initial stage, and it was decreased a little to be 14% at 48 h. The yield ratio of DVL (+ 5-HVA)
to THFA+1,5-PeD+1-PeOH decreased gradually in the time range from 4 h to 48 h. In addition,
the carbon balance decreased significantly at longer reaction time. The decreases of the selectivity
to DVL (+ 5-HVA) and the carbon balance suggest the consecutive reaction of DVL to the
compounds that cannot be detected by GC such as polymers. As a result, longer reaction time is
not suitable in the reaction of THFCA over Rh-WOx/SiO2 (W/Rh=0.25).
Table 5 (details: Table S5 in Supplementary Material) lists the effect of reaction temperature
and time of the reaction of THFCA over Rh-WOx/SiO2 (W/Rh=0.25). The yield ratio decreased
with increasing the conversion at each reaction temperature. When the yield is compared at similar
conversion level (Entries 2 and 6; Entries 4, 7 and 9), lower reaction temperature is more favorable
in terms of the yield ratio of DVL (+ 5-HVA) to THFA+1,5-PeD+1-PeOH.
Table 6 (details: Table S6 in Supplementary Material) lists the effect of H2 pressure on the
reaction of THFCA over Rh-WOx/SiO2 (W/Rh=0.25). The conversion increased with increase of
H2 pressure almost proportionally (first-order reaction). The comparison of the yield ratio of DVL
(+ 5-HVA) to THFA+1,5-PeD+1-PeOH at similar conversion level (Entries 2 and 5; Entries 4 and
6; Entries 7 and 8) indicates that the yield ratio of DVL (+ 5-HVA) to THFA+1,5-PeD+1-PeOH is
decreased (0.68 to 0.55; 0.50 to 0.40; 0.36 to 0.28, respectively) under higher H2 pressure. The
balance between the conversion rate and the yield ratio is thought to be satisfied at the medium H2
13

pressure range such as 1.5 MPa.
Table 7 (details: Table S7 in Supplementary Material) lists the effect of THFCA concentration
in water solvent on the reaction of THFCA over Rh-WOx/SiO2 (W/Rh=0.25). Here, the
concentration of THFCA was adjusted by changing the amount of THFCA and H2O (THFCA
concentration: 1.6 wt%-100 wt%). The conversion amount of THFCA (THFCA amount / mmol ×
conversion / %) was in the range of 1.17 to 2.33 mmol and the effect of the concentration was not
strong. At the same time, the formation amount of DVL (+ 5-HVA) was also in the range of 0.16
to 0.33 mmol and the effect of the concentration was also not strong. These results suggest the
almost zero reaction order with respect to the concentration of THFCA in the target and side
reactions of THFCA in the standard reaction conditions, although at lower concentrations there
was some dependence on the concentration (reaction order of 0.36 from conversion data in Entries
1-4). This may be explained by the strong interaction of THFCA with the catalyst surface. These
dependency trends on H2 pressure and substrate concentration were similar to the cases of C-O
hydrogenolysis at 2-position of functionalized primary alcohols such as THFAover Rh-ReOx/SiO2
and Rh-MoOx/SiO2 catalysts [62,63,67,69]. Regarding the yield ratio of DVL (+ 5-HVA) to
THFA+1,5-PeD+1-PeOH, the ratio increased with decrease of THFCA concentration below 11
wt% (Entries 1-4). The ratio seemed to increase at very high THFCA concentration (Entries 5 and
6); however, there were large amounts of unidentified products (“Others”) in these runs, and the
ratio might be overestimated.
3.3.Reaction of related substrates over Rh-WOx/SiO2 (W/Rh=0.25) catalyst
Table 8 lists the results of the reaction of THFCA and the related substrates. In the reaction of
THFCA using methanol as a solvent (Entry 2), the main reaction was esterification, and the
14

hydrogenolysis of C-O bond and hydrogenation of carboxyl group were suppressed. In the reaction
of THFA (Entry 3), the main product was 1,5-pentanediol, as reported in our previous study [41].
The selectivity of Rh-WOx/SiO2 to 1,5-pentanediol was not so remarkable in comparison with Rh-
ReOx/SiO2 and Rh-MoOx/SiO2-catalyzed reactions in the literature [41]. THFA has a similar
structure to THFCA, and the only difference is in the side chain: -CH2OH (THFA) and -C(=O)OH
(THFCA). The production of 1,5-pentanediol from THFA corresponds to the target reaction of
THFCA to 5-HVA. The reactivity of THFA is comparable to that of THFCA judging from the
conversion of the substrates (Entries 1 and 3). According to the studies on the reaction mechanism
of THFAhydrogenolysis on Rh-ReOx/SiO2 [31,35,69], the strong interaction between the -CH2OH
group of THFAand Rh-ReOx/SiO2 to give the alkoxide of Re species has been suggested by almost
zero reaction order with respect to the concentration of THFA. The similar reactivity of THFA and
THFCA over Rh-WOx/SiO2 and similar dependence (almost zero reaction order) on substrate
concentration in THFA and THFCA reactions suggest that adsorbed THFA and THFCA species
had similar reactivity. The reactivity of tetrahydrofuran over Rh-WOx/SiO2 (W/Rh=0.25) was
much lower (Entry 4), and the similar tendency was also observed in the case of Rh-ReOx/SiO2
[39], which indicates that high reactivity of THFA is caused by the strong interaction between -
CH2OH group and ReOx or WOx. The similarity between THFA and THFCA in both the structure
and reactivity suggests that THFCA is activated by the strong interaction between the -OH of the
carboxyl group and WOx. In the reaction of DVL, the conversion was clearly higher than that of
THFCA (Entry 5), which can explain the relatively low yield of DVL in the reaction of THFCA.
An important point is that the carbon balance of this test was very low (Entry 5, only 55%), and
the main reaction of DVL gave the products which cannot be detected by GC. One possible
candidate of the reaction of DVL is the ring-opening polymerization. The similar low carbon
15

balance in the reaction of THFCA at longer reaction time (Table 5) is also interpreted by the
consecutive ring-opening polymerization. 2-Methoxyacetic acid has similar partial structure of
THFCA, and the reaction of 2-methoxyacetic acid was also tested (Entry 6). The reactivity of 2-
methoxyacetic acid toward C-O hydrogenolysis of -position of carboxyl group (acetic acid
product) is much lower than that of THFCA. The reactivity of 2-methoxyacetic acid (Entry 6) and
acetic acid (Entry 7) toward reduction of the carboxyl group was also lower than that of THFCA;
however, the difference was smaller than that for C-O hydrogenolysis. Overall, the yield ratio of
the carboxylic compounds (acetic acid) to the hydrogenated compounds (2-methoxyethanol and
ethanol) in 2-methoxyacetic acid reduction is lower than the case of THFCA reduction.
3.4.Characterization of Rh-WOx/SiO2
In comparison with Rh-ReOx/SiO2 and Rh-MoOx/SiO2 catalysts, Rh-WOx/SiO2 catalyst has
been much less investigated and characterized. The Rh-ReOx/SiO2 and Rh-MoOx/SiO2 catalysts
with optimized Re or Mo amount (Re/Rh=0.5 or Mo/Rh=0.13) have the structure of Rh metal
particles partially covered with two-dimensional ReOx clusters or isolated MoOx species, as
proposed by the metal particle size (XRD and TEM), CO adsorption amount, TPR and XAFS
analysis [67,69]. On the other hand, in our previous studies Rh-WOx/SiO2 catalyst was only
characterized by CO adsorption, and the results were similar to Rh-MoOx/SiO2 catalysts [62]. In
the literature, Miessner et al. measured IR spectrum of adsorbed CO on Rh-WOx/SiO2 and Rh-
MoOx/SiO2 [70]. The vibration band was shifted from the position in Rh/SiO2 catalyst to higher
wavenumbers for both modified catalysts, suggesting the change of electronic state of Rh to
positively charged one caused by the WOx or MoOx species formed on the Rh particle surface.
Here, we characterized Rh-WOx/SiO2 catalyst to confirm the similar structure to Rh-MoOx/SiO2.
16

Figure 2 shows the H2-TPR profiles of Rh-WOx/SiO2 catalysts with various W/Rh ratios. The
most H2 consumption on Rh/SiO2 and Rh-WOx/SiO2 was observed in the temperature range of
300 K to 423 K. Considering the pressure difference between the TPR experiments and the reaction
test (1.5 MPa), it is thought that the reaction of THFCA proceeds on the reduced Rh-WOx/SiO2
catalysts at the reaction temperature of 413 K. Because some of Rh species was reduced during
the stabilization process at room temperature, as indicated by the smaller H2 consumption amount
than that required for reduction of Rh2O3 (0.39 mmol × 1.5 = 0.59 mmol) in most samples, it is
difficult to determine the reduction degree of W by these TPR profiles.
Figure 3 shows XRD patterns of Rh-WOx/SiO2 catalysts reduced in water at 413 K which is the
same temperature as the reaction temperature. The XRD pattern of catalyst after the reaction is
also shown. The particle sizes of Rh metal on Rh/SiO2 and Rh-WOx/SiO2 were determined to be
about 3 nm (2.9 to 3.3 nm), which are summarized in Figure 4. The additive effect of WOx on Rh
metal particle size is not strong. At the same time, the CO adsorption amount was measured on
Rh/SiO2 and Rh-WOx/SiO2, and the results are also plotted as a function of W/Rh ratio (Figure 4).
The amount of CO adsorption on Rh-WOx/SiO2 decreased with increase of the additive amount of
WOx, as reported in our previous paper [62]. The dispersion from the CO adsorption was clearly
lower than that from the XRD patterns, and the dispersion difference increased with increasing the
molar ratio of W/Rh. This behavior was also observed in the case of Rh-ReOx and Rh-MoOx
catalysts [67]. This is interpreted by the partial covering of the surface of Rh metal particles with
WOx species. Based on these results, in order to analyze the oxidation state of WOx and the
interaction between WOx and Rh metal surface, the XAFS experiment was carried out for Rh-
WOx/SiO2. Figure 5 shows W L3-edge XANES of Rh-WOx/SiO2. The edge energy of Rh-
WOx/SiO2 after reduction or reaction was lower than that of WO2, suggesting that the W species
17

in Rh-WOx/SiO2 was reduced to the oxidation state lower than 4. The oxidation state of W in Rh-
WOx/SiO2 was further estimated by the correlation between the oxidation state and the white line
area of L3-edge XANES. Here, only W metal and WO2 were used as reference compounds because
hexavalent W compounds show complex white line patterns depending on the coordination state
[71]. The oxidation state of W species on Rh-WOx/SiO2 after the reduction treatment was
estimated to be in the range of +1.2 to +2.2 (Figure 6). The XANES spectra and the estimated
oxidation state of W species in Rh-WOx/SiO2 (W/Rh=0.25) were almost the same (average valence
1.9) between the samples after the reduction and after reaction, indicating that the oxidation state
was maintained during the reaction. Figure 7 shows the W L3-edge EXAFS spectra and Table 9
lists the results of the curve fitting analysis. In all the Rh-WOx/SiO2 catalysts after the reduction
and the reaction, the contribution of W-O and W-Rh bonds was assigned. The bond length of W-
O was clearly shorter than that of W=O in Na2WO4, indicating the presence of W-O single bond.
The coordination number of the W-O bond corresponds to the oxidation state of W determined
from XANES (Figures 5 and 6). An important point is that the contribution of W-Rh bond was
present. The bond length of W-Rh bond was a little shorter than that of Rh-Rh in Rh metal (0.268
nm), indicating that this bond was the direct bond between W and Rh without the presence of oxide
ions. Similar structures have been reported on Rh-ReOx/SiO2 and Rh-MoOx/SiO2 [67]. In the case
of Rh-WOx/SiO2 catalysts, the contribution of W-W bond was not detected in the EXAFS analysis,
suggesting that monomeric W oxide species on Rh metal surface. The coordination number of W-
Rh in W/Rh ≥ 0.25 catalysts was around 3-4, which suggests that the W species can be located at
three-fold hollow site on Rh (111) or four-fold hollow site on Rh (100). On the other hand, the
catalyst with very small amount of W (W/Rh=0.13) had large coordination number of W-Rh (5.8)
and lower average W valence than 2 (1.2). Some W atoms in Rh-WOx/SiO2 (W/Rh=0.13) can be
18

in metallic state and alloyed with Rh.
Based on the coordination number of W-Rh bond in W L3-edge EXAFS analysis, the
coordination number of Rh-W bond in Rh K-edge EXAFS analysis can be calculated using
CN(Rh-W)/CN(W-Rh) = W/Rh. Figure 8 shows the Rh K-edge EXAFS spectra on Rh-WOx/SiO2
catalysts, and Table 10 lists the results of the curve fitting analysis. Here, two kinds of curve fitting
results are listed: one-wave fitting with Rh-Rh and two-wave fitting with Rh-Rh and Rh-W. In
both cases, the good fitting results are obtained and this indicates that Rh-W bond is not necessary
in the Rh K-edge EXAFS analysis of Rh-WOx/SiO2 catalysts. This can be interpreted by much
smaller contribution of surface Rh-W bond than that of Rh-Rh bond in the bulk of Rh metal
particles.
The characterization results shown above suggests that the optimized Rh-WOx/SiO2 catalyst
(W/Rh=0.25) has the Rh metal particles with ca. 3 nm size partially covered with monomeric WOx
species with low oxidation state of W (ca. +2). The structures are similar to those of Rh-MoOx/SiO2
and Rh-ReOx/SiO2 catalysts where Mo~4+ or Re2~3+ oxide species partially cover the Rh particles
with similar size (~3 nm) [67].
3.5.Mechanism of THFCA hydrogenolysis
Relatively high reaction order with respect to H2 pressure suggests that the rate determining step
is the reaction of the adsorbed substrate with the active hydrogen species. On the basis of the
previous reports on Rh-ReOx/SiO2 and Rh-MoOx/SiO2 catalysts [67,69], it is suggested that the
active hydrogen species can be hydride formed at the interface between Rh metal and metal oxide
species by the heterolytic dissociation of H2. In the case of hydrogenolysis of 2-functionalized
primary alcohols such as THFA, the terminal OH group is proposed to be coordinated with Re or
19

Mo atoms, and hydride species on Rh attacks the carbon atom at 2-position [67,69]. Rather low
reaction order with respect to the THFCA concentration suggests the strong adsorption of THFCA
on Rh-WOx/SiO2 catalysts. Carboxylic acid groups can interact with both metal and metal oxide
surfaces [35]. Therefore, it is thought that carboxylic acid groups can interact with Rh metal
surface and WOx species. On the other hand, the adsorption of the carboxylic acid group does not
seem to block the hydrogen activation site such as the interface between Rh metal and WOx. If it
blocks the hydrogen activation site, strongly negative reaction order like -1 should be obtained.
Therefore, the carboxylic acid group in THFCA probably mainly interacted with WOx species, not
Rh species which are the activation sites of hydrogen. Here we propose essentially the same
mechanism of C-O hydrogenolysis of THFCA to that of THFA: the carboxyl group is coordinated
with the W atom, and hydride species on Rh attacks the carbon atom at 2-position (Figure 9).
Based on the results of the reaction tests, hydrogenation of carboxyl groups ((b) in Figure 9) as
well as C-O hydrogenolysis ((a) in Figure 9) is catalyzed by Rh-WOx/SiO2, and the selectivity to
specific products by two reactions is not so high. This suggests that the active hydrogen species
such as hydride can attack two carbon atoms in THFCA: one is the carbon atom binding the
carboxyl groups in the tetrahydrofuran ring to give DVL and the other is the carbon atom in the
carboxyl group to give THFA. In addition, it is not easy to suppress the consecutive reaction of
DVL and the yield of DVL formation is not high at present. Highly selective attack to the carbon
atom binding the carboxyl groups in the tetrahydrofuran ring to give DVL and the suppression of
the consecutive reaction of DVL may enhance the yield of DVL. Nevertheless, the selectivity of
Rh-WOx/SiO2 to C-O hydrogenolysis products was higher than those of Rh-MoOx/SiO2 and Rh-
ReOx/SiO2 with similar direct interaction between Rh and added metal oxide species. Re species
is known to be active in hydrogenation of carboxylic acid [66,72,73], and the combination with
20

Rh much increases the activity of Re [66,72]. Mo species is known to work as a promoter of noble
metal catalyst for carboxylic acid hydrogenation, especially in combination with Pt or Rh [55-60].
The higher activity or promotion ability of Re or Mo species than W species may lower the
selectivity to C-O hydrogenolysis products by promoting the carboxylic acid hydrogenation as the
side reaction. While the combination of Ir and Re shows excellent activity in C-O hydrogenolysis
of 2-functionalized primary alcohols, both the activity in C-O hydrogenolysis of THFCA in our
study and the reported activity in hydrogenation of carboxylic acid of Ir-based catalysts are lower
than those of Rh-based catalysts [60,66]. The adsorption of carboxylic acid on Ir site might be
stronger than that on Rh, which decreases the activity of hydrogen activation. Pt-, Pd- and Ru-
based catalysts have lower activity in C-O hydrogenolysis in both THFCA and 2-functionalized
primary alcohols because of the different nature of activated hydrogen species: neutral hydrogen
species formed by homolytic dissociation of H2 which are suitable to C=C hydrogenation.
4. Conclusions
C-O hydrogenolysis at 2-position of tetrahydrofuran-2-carboxylic acid (THFCA) was
investigated, which has not been the target reaction in the literature. Even as a related reaction (C-
O hydrogenolysis at -position of functionalized carboxylic acid), lactic acid hydrogenolysis to
propionic acid has been rarely investigated; only carried out at very high temperature. This is the
first comprehensive study of oxide-modified noble metal catalysts for C-O hydrogenolysis at -
position of functionalized carboxylic acid. Among the tested combinations (M-M’Ox/SiO2, M =
Rh, Pt, Ru, Ir; M’ = None, W, Re, Mo), Rh-WOx/SiO2 catalyst showed relatively high yield of C-
O hydrogenolysis products (-valerolactone (DVL) and 5-hydroxyvaleric acid) and yield ratio of
C-O hydrogenolysis products to carboxylic acid hydrogenation products (tetrahydrofurfuryl
21

alcohol, 1,5-pentanediol and 1-pentanol). Although the effects of W/Rh ratio and various
parameters of reaction conditions were investigated in detail, the selectivity to C-O hydrogenolysis
products is not so high at present; it is not easy to suppress the consecutive reaction of DVL. The
promoting effect of W is more remarkable when W is directly added to the catalyst than the case
when mixture of Rh/SiO2 and WOx/SiO2 is used, suggesting that the direct interaction of WOx
species with Rh metal surface is important in the catalysis. The direct interaction between WOx
species and Rh metal surface is confirmed by the characterizations: reduced W oxide species
(average oxidation state ~+2) partially cover the surface of Rh metal particles with ca. 3 nm size.
Strong adsorption of carboxylic acid groups on the interface of Rh and WOx species can promote
the C-O hydrogenolysis at the -position of carboxylic acid groups. The structure and promotion
mechanism are similar to the case of Rh-ReOx/SiO2 and Rh-MoOx/SiO2 catalysts which are active
in C-O hydrogenolysis at 2-position of 2-functionalized primary alcohols. The lower activity or
promotion ability of WOx species in carboxylic acid hydrogenation than ReOx or MoOx species
may cause relatively good selectivity to C-O hydrogenolysis products.
CRediT author statement
Takehiro Asano: Conceptualization, Methodology, Investigation, Writing - Original Draft;
Yoshinao Nakagawa: Conceptualization, Methodology, Writing - Review & Editing; Masazumi
Tamura: Formal analysis; Keiichi Tomishige: Supervision, Writing - Review & Editing.
Declaration of interests
The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
22

Acknowledgement
This work is supported by JSPS KAKENHI grant numbers of 18H05247, 20H02519 and
20H02520. We thank Mr. Lujie Liu and Ms. Ayaka Kuwata for their help with experiments.
References
[1] A.M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem. Int. Ed. 51 (2012) 2564-2601.
https://doi.org/10.1002/anie.201105125
[2] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827-1870.
https://doi.org/10.1021/cr4002269
[3] S. De, B. Saha, R. Luque, Bioresour. Technol. 178 (2015) 108-118.
https://doi.org/10.1016/j.biortech.2014.09.065
[4] S. Kim, E.E. Kwon, Y.T. Kim, S. Jung, H.J. Kim, G.W. Huber, J. Lee, Green Chem. 21 (2019)
3715-3743. https://doi.org/10.1039/C9GC01210A
[5] J.J. Bozell, G.R. Petersen, Green Chem. 12 (2010) 539-554. https://doi.org/10.1039/B922014C
[6] L.T. Mika, E. Cséfalvay, Á. Németh, Chem. Rev. 118 (2018) 505-613.
https://doi.org/10.1021/acs.chemrev.7b00395
[7] G.G. Millán, S. Hellsten, J. Llorca, R. Luque, H. Sixta, A.M. Balu, ChemCatChem 11 (2019)
2022-2042. https://doi.org/10.1002/cctc.201801843
[8] R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sádaba, M.L. Granados, Energy Environ. Sci. 9
(2016) 1144-1189. https://doi.org/10.1039/C5EE02666K
[9]
X.
Li,
P.
Jia,
T.
Wang,
ACS
Catal.
6
(2016)
7621-7640.
23

https://doi.org/10.1021/acscatal.6b01838
[10] Y. Nakagawa, M. Tamura, K. Tomishige, ACS Catal. 3 (2013) 2655-2668.
https://doi.org/10.1021/cs400616p
[11] S. Chen, R. Wojcieszak, F. Dumeignil, E. Marceau, S. Royer, Chem. Rev. 118 (2018) 11023-
11117. https://doi.org/10.1021/acs.chemrev.8b00134
[12] M. Douthwaite, X. Huang, S. Iqbal, P.J. Miedziak, G.L. Brett, S.A. Kondrat, J.K. Edwards,
M. Sankar, D.W. Knight, D. Bethell, G.J. Hutchings, Catal. Sci. Technol. 7 (2017) 5284-5293.
https://doi.org/10.1039/C7CY01025G
[13] F. Menegazzo, M. Signoretto, T. Fantinel, M. Manzoli, J. Chem. Technol. Biotechnol. 92
(2017) 2196-2205. https://doi.org/10.1002/jctb.5240
[14] N.K. Gupta, A. Fukuoka, K. Nakajima, ACS Sustainable Chem. Eng. 6 (2018) 3434-3442.
https://doi.org/10.1021/acssuschemeng.7b03681
[15] C.P. Ferraz, M. Zieliński, M. Pietrowski, S. Heyte, F. Dumeignil, L.M. Rossi, R. Wojcieszak,
ACS
Sustainable
Chem.
Eng.
6
(2018)
16332-16340.
https://doi.org/10.1021/acssuschemeng.8b03330
[16] B. Peng, C.-L. Ma, P.-Q. Zhang, C.-Q. Wu, Z.-W. Wang, A.-T. Li, Y.-C. He, B. Yang, Green
Chem. 21 (2019) 5914-5923. https://doi.org/10.1039/C9GC02499A
[17] A. Roselli, Y. Carvalho, F. Dumeignil, F. Cavani, S. Paul, R. Wojcieszak, Catalysts 2020, 10,
73. https://doi.org/10.3390/catal10010073
[18] S. Werkmeister, K. Junge, B. Wendt, E. Alberico, H. Jiao, W. Baumann, H. Junge, F. Gallou,
M. Beller, Angew. Chem. Int. Ed. 53 (2014) 8722-8726. https://doi.org/10.1002/anie.201402542
[19] M. Naruto, S. Agrawal, K. Toda, S. Saito, Sci. Rep. 7 (2017) 3425.
https://doi.org/10.1038/s41598-017-03436-y
24

[20] B.M. Stadler, P. Puylaert, J. Diekamp, R. van Heck, Y. Fan, A. Spannenberg, S. Hinze, J.G.
de Vries, Adv. Synth. Catal. 360 (2018) 1151-1158. https://doi.org/10.1002/adsc.201701607
[21] J. Ullrich, B. Breit, ACS Catal. 8 (2018) 785-789. https://doi.org/10.1021/acscatal.7b03484
[22] T. Asano, M. Tamura, Y. Nakagawa, K. Tomishige, ACS Sustainable Chem. Eng. 4 (2016)
6253-6257. https://doi.org/10.1021/acssuschemeng.6b01640
[23] T. Asano, Y. Nakagawa, M. Tamura, K. Tomishige, ACS Sustainable Chem. Eng. 7 (2019)
9601-9612. https://doi.org/10.1021/acssuschemeng.9b01104
[24] T. Asano, H. Takagi, Y. Nakagawa, M. Tamura, K. Tomishige, Green Chem. 21 (2019) 6133-
6145. https://doi.org/10.1039/C9GC03315G
[25] Q. Sun, S. Wang, H. Liu, ACS Catal.
https://doi.org/10.1021/acscatal.9b04074
9
(2019) 11413-11425.
[26] M. Maris, W.-R. Huck, T. Mallat, A. Baiker, J. Catal. 219 (2003) 52-58.
https://doi.org/10.1016/S0021-9517(03)00184-2
[27] X. Zhang, L. Chen, Y. Li, H. Li, Z. Xie, Q. Kuang, L. Zheng, Dalton Trans. 48 (2019) 5515-
5519. https://doi.org/10.1039/C9DT00348G
[28] X. Li, Z. Zhai, C. Tang, L. Sun, Y. Zhang, W. Bai, RSC Adv. 6 (2016) 62252-62262.
https://doi.org/10.1039/C6RA10096A
[29] T.J. Korstanje, H. Kleijn, J.T.B.H. Jastrzebski, R.J.M.K. Gebbink, Green Chem. 15 (2013)
982-988. https://doi.org/10.1039/C3GC36874B
[30] X. Li, J. Pang, J. Zhang, C. Yin, W. Zou, C. Tang, L. Dong, Ind. Eng. Chem. Res. 58 (2019)
101-109. https://doi.org/10.1021/acs.iecr.8b04713
[31]
Y.
Nakagawa,
K.
Tomishige,
Catal.
Today
195
(2012)
136-143.
http://doi.org/10.1016/j.cattod.2012.04.048
25

[32] Y. Nakagawa, M. Tamura, K. Tomishige, J. Mater. Chem. A 2 (2014) 6688-6702.
https://doi.org/10.1039/C3TA15384C
[33] Y. Nakagawa, M. Tamura, K. Tomishige, Res. Chem. Intermed. 44 (2018) 3879-3903.
https://doi.org/10.1007/s11164-018-3481-2
[34] K. Tomishige, Y. Nakagawa, M. Tamura, Curr. Opin. Green Sustainable Chem. 22 (2020) 13-
21. https://doi.org/10.1016/j.cogsc.2019.11.003
[35] K. Tomishige, Y. Nakagawa, M. Tamura, Green Chem. 19 (2017) 2876-2924.
https://doi.org/10.1039/C7GC00620A
[36] S. Feng, A. Nagao, T. Aihara, H. Miura, T. Shishido, Catal. Today 303 (2018) 207-212.
https://doi.org/10.1016/j.cattod.2017.08.058
[37] C. Wang, J.D. Lee, Y. Ji, T.M. Onn, J. Luo, C.B. Murray, R.J. Gorte, Catal. Lett. 148 (2018)
1047-1054. https://doi.org/10.1007/s10562-018-2323-6
[38] E. Soghrati, C.K. Poh, Y. Du, F. Gao, S. Kawi, A. Borgna, ChemCatChem 10 (2018) 4652-
4664. https://doi.org/10.1002/cctc.201800783
[39] S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori, K. Tomishige, Chem. Commun.
(2009) 2035-2037. https://doi.org/10.1039/B822942B
[40] A. Shimao, S. Koso, N. Ueda, Y. Shinmi, I. Furikado, K. Tomishige, Chem. Lett. 38 (2009)
540-541. https://doi.org/10.1246/cl.2009.540
[41] S. Koso, N. Ueda, Y. Shinmi, K. Okumura, T. Kizuka, K. Tomishige, J. Catal. 267 (2009) 89-
92. https://doi.org/10.1016/j.jcat.2009.07.010
[42] K. Chen, S. Koso, T. Kubora, Y. Nakagawa, K. Tomishige, ChemCatChem 2 (2010) 547-555.
https://doi.org/10.1002/cctc.201000018
[43] Y. Nakagawa, Y. Shinmi, S. Koso, K. Tomishige, J. Catal. 272 (2010) 191-194.
26

https://doi.org/10.1016/j.jcat.2010.04.009
[44] Y. Amada, Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B 105
(2011) 117-127. https://doi.org/10.1016/j.apcatb.2011.04.001
[45] L. Liu, S. Kawakami, Y. Nakagawa, M. Tamura, K. Tomishige, Appl. Catal. B 256 (2019)
117775. https://doi.org/10.1016/j.apcatb.2019.117775
[46] L. Liu, T. Asano, Y. Nakagawa, M. Tamura, K. Okumura, K. Tomishige, ACS Catal. 9 (2019)
10913-10930. https://doi.org/10.1021/acscatal.9b03824
[47] M. Chia, Y.J. Pagán-Torres, D. Hibbitts, Q. Tan, H.N. Pham, A.K. Datye, M. Neurock, R.J.
Davis,
J.A.
Dumesic,
J.
Am.
Chem.
Soc.
133
(2011)
12675-12689.
https://doi.org/10.1021/ja2038358
[48] T. Buntara, S. Noel, P.H. Phua, I. Melián-Cabrera, J.G. de Vries, H.J. Heeres, Angew. Chem.
Int. Ed. 50 (2011) 7083-7087. https://doi.org/10.1002/anie.201102156
[49] T. Buntara, S. Noel, P.H. Phua, I. Melián-Cabrera, J.G. de Vries, H.J. Heeres, Top. Catal. 55
(2012) 612-619. https://doi.org/10.1007/s11244-012-9839-6
[50] M. Chia, B.J. O’Neill, R. Alamillo, P.J. Dietrich, F.H. Ribeiro, J.T. Miller, J.A. Dumesic, J.
Catal. 308 (2013) 226-236. https://doi.org/10.1016/j.jcat.2013.08.008
[51] S.H. Hakim, C. Sener, A.C. Alba-Rubio, T.M. Gostanian, B.J. O’Neill, F.H. Ribeiro, J.T.
Miller, J.A. Dumesic, J. Catal. 328 (2015) 75-90. https://doi.org/10.1016/j.jcat.2014.12.015
[52] P.U. Karanjkar, S.P. Burt, X. Chen, K.J. Barnett, M.R. Ball, M.D. Kumbhalkar, X. Wang, J.B.
Miller, I. Hermans, J.A. Dumesic, G.W. Huber, Catal. Sci. Technol. 6 (2016) 7841-7851.
https://doi.org/10.1039/C6CY01763K
[53] A. Said, D. D. S. Perez, N. Perret, C. Pinel, M. Besson, ChemCatChem 9 (2017) 2768-2783.
https://doi.org/10.1002/cctc.201700260
27

[54] I. Furikado, T. Miyazawa, S. Koso, A. Shimao, K. Kunimori, K. Tomishige, Green Chem. 9
(2007) 582-588. https://doi.org/10.1039/B614253B
[55] M. Tamura, R. Tamura, Y. Takeda, Y. Nakagawa, K. Tomishige, Chem. Commun. 50 (2014)
6656-6659. https://doi.org/10.1039/C4CC02675F
[56] M. Tamura, R. Tamura, Y. Takeda, Y. Nakagawa, K. Tomishige, Chem. Eur. J. 21 (2015)
3097-3107. https://doi.org/10.1002/chem.201405769
[57] Y. Takeda, T. Shoji, H. Watanabe, M. Tamura, Y. Nakagawa, K. Okumura, K. Tomishige,
ChemSusChem 8 (2015) 1170-1178. https://doi.org/10.1002/cssc.201403011
[58] J. Cui, J. Tan, Y. Zhu, F. Cheng, ChemSusChem 11 (2018) 1316-1320.
https://doi.org/10.1002/cssc.201800038
[59] M. Li, G. Li, N. Li, A. Wang, W. Dong, X. Wang, Y. Cong, Chem. Commun. 50 (2014) 1414-
1416. https://doi.org/10.1039/C3CC48236G
[60] T. Mizugaki, Y. Nagatsu, K. Togo, Z. Maeno, T. Mitsudome, K. Jitsukawa, K. Kaneda, Green
Chem. 17 (2015) 5136-5139. https://doi.org/10.1039/C5GC01878A
[61] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green Chem. 6
(2004) 359-361. https://doi.org/10.1039/B407378A
[62] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl. Catal. B 94 (2010) 318-
326. https://doi.org/10.1016/j.apcatb.2009.11.021
[63] Y. Amada, S. Koso, Y. Nakagawa, K. Tomishige, ChemSusChem 3 (2010) 728-736.
https://doi.org/10.1002/cssc.201000040
[64] Y. Nakagawa, R. Tamura, M. Tamura, K. Tomishige, Sci. Technol. Adv. Mater. 16 (2015)
014901. https://doi.org/10.1088/1468-6996/16/1/014901
[65] T. Arai, M. Tamura, Y. Nakagawa, K. Tomishige, ChemSusChem 9 (2016) 1680-1688.
28

https://doi.org/10.1002/cssc.201600295
[66] Y. Takeda, Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 2 (2012) 2221-2223.
https://doi.org/10.1039/c2cy20302b
[67] S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, Appl. Catal. B 111-112
(2012) 27-37. https://doi.org/10.1016/j.apcatb.2011.09.015
[68] V.C. Nguyen, A. Dandach, T.T.H. Vu, P. Fongarland, N. Essayem, Mol. Catal. 476 (2019)
110518. https://doi.org/10.1016/j.mcat.2019.110518
[69] S. Koso, Y. Nakagawa, K. Tomishige, J. Catal. 280 (2011) 221-229.
https://doi.org/10.1016/j.jcat.2011.03.018
[70] A. Trunschke, H. Ewald, D. Gutschick, H. Miessner, M. Skupin, J. Mol. Catal. 56 (1989) 95-
106. https://doi.org/10.1016/0304-5102(89)80175-0
[71] S. Yamazoe, Y. Hitomi, T. Shishido. T. Tanaka, J. Phys. Chem. C 112 (2008) 6869-6879.
https://doi.org/10.1021/jp711250f
[72] Y. Takeda, M. Tamura, Y. Nakagawa, K. Okumura, K. Tomishige, ACS Catal. 5 (2015) 7034-
7047. https://doi.org/10.1021/acscatal.5b01054
[73] M. Tamura, Y. Nakagawa, K. Tomishige, Asian J. Org. Chem. 9 (2020) 126-143.
https://doi.org/10.1002/ajoc.201900667
29