Insight into the hydrogenation of pure and crude HMF to furan diols using Ru/C as catalyst

Article abstract of DOI:10.1016/j.apcata.2019.04.007

5-hydroxymethylfurfural (HMF) is one of the most important renewable platform-chemicals, a very valuable precursor for the synthesis of bio-fuels and bio-products. In this work, the hydrogenation of HMF to two furan diols, 2,5-bis(hydroxymethyl)furan (BHMF) and 2,5-bis(hydroxymethyl)tetrahydrofuran (BHMTHF), both promising renewable monomers, was investigated. Three commercial catalysts, Ru/C, Pd/C and Pt/C, were tested in the hydrogenation of aqueous HMF solutions (2–3 wt%), using a metal loading of 1 wt% respect to HMF content. By appropriate tuning of the process conditions, either BHMF or BHMTHF were obtained in good yields, and Ru/C resulted the best catalyst for this purpose, allowing us to obtain BHMF or BHMTHF yields up to 93.0 and 95.3 mol%, respectively. This catalyst was also tested for in the hydrogenation of a crude HMF-rich hydrolyzate, obtained by one-pot the dehydration of fructose. The influence of each component of this hydrolyzate on the hydrogenation efficiency was investigated, including unconverted fructose, rehydration acids and humins, in order to improve the yields towards each furan diol. Moreover, ICP-OES and TEM analysis showed that the catalyst was not subjected to important leaching and sintering phenomena, as further confirmed by catalyst recycling study.

Full text of DOI:10.1016/j.apcata.2019.04.007

Accepted Manuscript
Title: Insight into the hydrogenation of pure and crude HMF
to furan diols using Ru/C as catalyst
Authors: Sara Fulignati, Claudia Antonetti, Domenico Licursi,
Matteo Pieraccioni, Erwin Wilbers, Hero Jan Heeres, Anna
Maria Raspolli Galletti
PII:
S0926-860X(19)30159-0
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2019.04.007
APCATA 17039
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
1 November 2018
18 March 2019
8 April 2019
Please cite this article as: Fulignati S, Antonetti C, Licursi D, Pieraccioni M, Wilbers
E, Heeres HJ, Galletti AMR, Insight into the hydrogenation of pure and crude
HMF to furan diols using Ru/C as catalyst, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.04.007
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. 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.

Insight into the hydrogenation of pure and crude HMF to furan diols
using Ru/C as catalyst
Sara Fulignatia, Claudia Antonettia, Domenico Licursia, Matteo Pieraccionia, Erwin
Wilbersb, Hero Jan Heeresb, Anna Maria Raspolli Gallettia,*
a Department of Chemistry and Industrial Chemistry, University of Pisa, Via G. Moruzzi
13, 56124, Pisa, Italy
b Green Chemical Reaction Engineering, ENTEG, University of Groningen, Nijenborgh
4, 9747 AG Groningen, The Netherlands
*Corresponding author, e-mail: anna.maria.raspolli.galletti@unipi.it
1

Graphical abstract
Highlights
 The tunable hydrogenation of both aqueous HMF solutions and crude fructose
hydrolyzate to renewable diols BHMF and BHMTHF, was studied.
 5 wt% Ru/C resulted an active and robust catalyst, showing negligible sintering and
leaching processes.
 Yields of BHMF up to 93 mol% and of BHMTHF up to 95 mol% were ascertained
starting from aqueous HMF.
 This study proves the feasibility of renewable monomers synthesis from the crude
fructose hydrolyzate, opening the way for their production directly from
lignocellulosic biomass through a cascade process.
Abstract
5-hydroxymethylfurfural (HMF) is one of the most important renewable platform-
chemicals, a very valuable precursor for the synthesis of bio-fuels and bio-products. In
this work, the hydrogenation of HMF to two furan diols, 2,5-bis(hydroxymethyl)furan
(BHMF) and 2,5-bis(hydroxymethyl)tetrahydrofuran (BHMTHF), both promising
renewable monomers, was investigated. Three commercial catalysts, Ru/C, Pd/C and
Pt/C, were tested in the hydrogenation of aqueous HMF solutions (2-3 wt%), using a
metal loading of 1 wt% respect to HMF content. By appropriate tuning of the process
2

conditions, either BHMF or BHMTHF were obtained in good yields, and Ru/C resulted
the best catalyst for this purpose, allowing us to obtain BHMF or BHMTHF yields up to
93.0 and 95.3 mol%, respectively. This catalyst was also tested for in the hydrogenation
of a crude HMF-rich hydrolyzate, obtained by one-pot the dehydration of fructose. The
influence of each component of this hydrolyzate on the hydrogenation efficiency was
investigated, including unconverted fructose, rehydration acids and humins, in order to
improve the yields towards each furan diol. Moreover, ICP-OES and TEM analysis
showed that the catalyst was not subjected to important leaching and sintering
phenomena, as further confirmed by catalyst recycling study.
Keywords
5-hydroxymethylfurfural; aqueous-phase hydrogenation; 2,5-bis(hydroxymethyl)furan;
2,5-bis(hydroxymethyl)tetrahydrofuran; crude hydrolyzate.
3

1. Introduction
Nowadays, the dwindling supplies of worldwide fossil resources and the growth of carbon
dioxide emission make the production of chemicals and fuels from renewable resources
a key topic of the industrial chemistry [1-3]. Biomass is a very promising alternative
feedstock, being abundant, cheap, widespread and precursor for the production of several
valuable products [4-6]. In particular, 5-hydroxymethylfurfural (HMF) is considered as
one of the most important bio-based compounds [7]. HMF may be obtained by the
dehydration of model compounds, including monosaccharides, such as glucose [8-11]
and fructose [11-15], polysaccharides, such as inulin [12,13,16], starch [11,15] and
cellulose [11,15,17,18] and, more advantageously, real lignocellulosic biomasses, such
as corn stover, pinewood, switchgrass and poplar [14,19]. The presence of different
reactive groups (an aldehyde group, a hydroxyl group and a furan ring) makes HMF a
very important platform-chemical, precursor of bio-fuels, such as 2,5-dimethylfuran
(DMF) [20,21], 5-ethoxymethylfurfural (EMF) [22] and long chain alkanes [23]. In
addition, it is possible to convert HMF into interesting monomers, such as 2,5-
furandicarboxylic acid (FDCA) [24,25], 2,5-bis(hydroxymethyl)furan (BHMF) [26-28],
2,5-bis(hydroxymethyl)tetrahydrofuran (BHMTHF) [29-31] and caprolactone [32] and
many more valuable products [33-35].
In this work, the selective synthesis of two important furan diols, BHMF and BHMTHF,
is investigated. The first one derives from the hydrogenation of the aldehyde group of
HMF, whereas the second one stems from the hydrogenation of both aldehyde group and
furan ring. Their application for the synthesis of resins, fibres, foams and polymers has
been recently proven, underlining their high potential as monomers for the synthesis of
alternative and renewable materials [36-40]. Regarding their possible synthesis, the
majority of the literature investigations employs molecular hydrogen as reducing agent,
in particular for the synthesis of BHMTHF, whereas the hydrogenation of HMF to BHMF
4

has been also carried out using formic acid [41] or isopropanol [42,43] as hydrogen donor,
or through electrochemical processes [28]. Concerning the catalyst selection, mainly
heterogeneous catalysts have been used and only a few papers have described the use of
homogeneous ones [41, 44]. Advantages of the former are ease of separation from the
reaction medium and recyclability. The most largely adopted heterogeneous catalysts are
represented by metals, such as Ru, Pt, Pd, Au, Ir, Ni, Cu supported on several oxides,
polymers or carbon species [3,30,36,45-53]. Ru-, Pd-, and Pt-catalysts are particularly
attractive for this purpose, because of their high intrinsic activity, and it is usually easy to
have them dispersed as nanoparticles on an appropriate support [53]. In most of the
published works, the HMF hydrogenation has been carried out working with toxic,
expensive, and non-renewable solvents, such as ionic liquids and organic solvents
(mainly tetrahydrofuran, 1,4-dioxane and alcohols). However, from a green chemistry
point of view, the use of water is certainly preferred. In addition, the HMF upgrading in
water is of more practical importance, because HMF is expected to be directly supplied
as an aqueous solution in a biorefinery process, thus minimizing or, even better, avoiding
expensive and unnecessary purification steps. Unfortunately, lower diols selectivities are
reported when the reaction is performed in water, rather than in organic or biphasic
solvent systems [30,54,55]. In fact, water-based HMF hydrogenation may lead to
different products, because this reaction environment could enable the hydrolytic ring
opening, the hydrodeoxygenation and rearrangements of the furan ring. On this basis, the
selectivity of the water-based hydrogenation of HMF to BHMF and BHMTHF is
markedly determined by the hydrogenation activity and acid–base properties of the
chosen catalytic system. Functional sites determine the selectivity of products in a
catalytic reaction system for the hydrogenation of HMF in water, using supported metal
catalysts, in particular the metal surface for hydrogenation/hydrogenolysis, the support
surface for acid–base catalysis, the metal-support interface for the unique adsorption of
5

reactants, and the acid–base catalysis, determined by the presence of water and other
compounds of the reaction mixture. Focusing the attention on the catalysts of interest in
this work, the best results in aqueous medium have been obtained using non-commercial,
ad hoc synthesized catalysts. In this regard, Chen et al. [46] carried out the hydrogenation
of diluted aqueous HMF solution employing Ru clusters immobilized on nanosized
mesoporous zirconium silica (Ru/MSN-Zr) as catalyst, reaching the maximum BHMF
yield of 90 mol%. They have also investigated the synthesis of BHMTHF from HMF in
water [48] adopting Pd catalyst supported on amine-functionalized metal-organic
frameworks (Pd/MIL-101(Al)-NH2), obtaining the BHMTHF yield of 96 mol%. Despite
the promising catalytic performances of many ad hoc synthetised catalysts, some
problems still limit their application on a larger scale, such as the reproducibility of
catalyst formulation (and therefore of its properties), the cost of recovery of the precious
metal from the spent catalyst after its use, the cost of the support and catalyst production.
Commercial catalysts still remain the preferred choice for hydrogenation reactions, in
particular those carbon-based, because of their lower cost, high surface area, chemical
inertness, thermal stability in non-oxidizing atmospheres, and ease of metal recovery,
allowed by simple calcination. The use of commercial catalysts for the synthesis of the
diols was reported by Schiavo et al. [57]. A wide range of noble metals supported on
carbon, such as 10 wt% Pd/C, 10 wt% Pt/C and 10 wt% Ru/C, was tested together with
Raney Ni, platinum oxide and copper chromite. All catalysts were active towards the
HMF hydrogenation in water to BHMF and BHMTHF and, particularly, 10 wt% Ru/C
showed promising results, by properly tuning the reaction time. In particular, the BHMF
yield of 95 mol% was obtained after 30 min, whereas prolonging the reaction to 4 h gave
the BHMTHF yield of 92 mol%. These results were obtained adopting a low HMF
concentration (1.3 wt%), which could be responsible of the ascertained good selectivity.
Moreover, the authors did not perform a systematic investigation and they did not study
6

the influence of the main reaction parameters to give the target diols. Regarding the
possible role of the catalytic support, Alamillo et al. [30] proved that acidic supports, such
as SiO2, have a detrimental role on the HMF hydrogenation, favouring the selective
formation of ring opening triols and tetrol, such as 1,2,5-hexanetriol, 1,2,5,6-hexanetetrol
and 1,2,6-hexanetirol, formed by the hydrogenation of acid-catalyzed degradation
products of BHMF. In this context, also the presence of other homogeneous acids, which
are typical of the hydrolyzate solutions, such as H2SO4 and levulinic acid, also caused a
significant reduction of furan diols selectivity. These statements suggest that the pH of
the aqueous reaction solution has certainly a strong influence on the furan diols selectivity
and, in particular, a low pH causes undesired ring opening and also degradation of HMF
and reaction intermediates, leading to the undesired formation of humins. This issue could
be partially solved on laboratory scale by using a biphasic system, which allows HMF
extraction from the aqueous phase, minimizing its degradation to acids and humins. This
application is still academically interesting, even if not very practical from an industrial
point of view. In this context, Alamillo et al. [30] studied the activity of Ru black in
different solvents and the highest BHMTHF selectivity (67 mol%) at complete HMF
conversion was obtained using tetrahydrofurfuryl alcohol, whereas it markedly decreased
to 46 mol% employing the biphasic water/1-butanol (1/2 v/v) system and even more in
water (22 mol%). The decrease of BHMTHF yield was attributed to the formation of
polyols, such as 1,2,6-hexanetriol and 1,2,5-hexanetriol, deriving from the hydration of
the intermediate BHMF, together with additional degradation pathways occurring in
water. In addition to the acid–base properties of a reaction system, also the hydrogenation
rate has an influence on the yield or selectivity of BHMF and BHMTHF, because BHMF
is relatively unstable under hydrothermal reaction conditions. In this sense, the maximum
yield of BHMF can be obtained by increasing the hydrogenation rate of HMF beyond that
of BHMF ring opening. On the basis of the above statements, HMF hydrogenation in
7

water is certainly challenging for industrial applications, but very difficult to tune,
depending on the contribute of many different components, which simultaneously act
within the reaction environment, and which must be individually and experimentally
considered, for a better understanding of the reaction.
Starting from the work of Schiavo et al. [30], in this work, the hydrogenation of more
concentrated HMF aqueous solution (2-3 wt%) has been carried out in the presence of
commercial noble metals supported on carbon, Ru/C, Pd/C and Pt/C. The choice of
carbon support for this purpose is appropriate, thanks to its relative inertness, which
prevents the occurrence of unwanted reactions catalyzed by the support surface acidity,
thus allowing us to focus the attention on the sole effect of the reaction mixture. The
adopted HMF concentration was similar to that reached in water for crude HMF synthesis
from fructose hydrolysis, previously optimized by us [13]. The BHMF and BHMTHF
yields have been optimized by properly tuning the process conditions, in order to
minimize the ring opening issue. These optimized reaction conditions have been
subsequently employed in the cascade process for the direct hydrogenation of the HMF-
rich crude hydrolyzate obtained from the dehydration of fructose to HMF, without any
intermediate separation step. With this approach, the intermediate separation and
purification steps to obtain pure HMF are avoided, because unnecessary, with a positive
impact on the techno-economic viability of the overall process from fructose to renewable
furan diols.
2. Methods
2.1 Materials
5-hydroxymethylfurfural (95%) was supplied by AVA Biochem. 2,5-
bis(hydroxymethyl)furan (95%) and 2,5-bis(hydroxymethyl)tetrahydrofuran (95%) were
provided by AKos GmbH (Germany). Ru/C (5 wt%), Pd/C (5 wt%), formic acid (99.8%),
8

levulinic acid (98%), ethanol (96%), dichloromethane (99.9%), sodium bicarbonate and
water for HPLC were purchased from Sigma-Aldrich. Pt/C (5 wt%) was supplied by
Strem Chemicals. Fructose was food grade. Amberlyst-70 was provided by Rohm and
Haas. All catalysts and chemicals were employed as received.
2.2 Hydrogenation of HMF
Hydrogenation reactions were carried out in a 300 mL stainless steel Parr 4560 autoclave
equipped with a P.I.D. controller (4843). In a typical experiment, a weight ratio metal to
HMF of 1 wt% was used. In this regard, the catalyst employed as received, was weighted
and introduced into the autoclave, which was subsequently closed and evacuated to 65 Pa
with a mechanical vacuum pump. 50 mL of a HMF aqueous solution was introduced into
the autoclave by suction, and the reaction mixture was stirred using a mechanical
overhead stirrer. Then, the reactor was pressurized with hydrogen till the desired pressure
was reached at the pre-set temperature, always under mechanical stirring. The pressure in
the reactor was manually held constant at the pre-determined value by repeated hydrogen
addition, when necessary. The reaction progress was monitored by sampling periodically
the liquid through a dip tube. The liquid samples were analysed using HPLC. All the
experiments were carried out in triplicate and the reproducibility of the techniques was
within 3%. For the recycling tests, the employed catalyst was recovered by filtration and
re-used within two subsequent runs. At the end of the third cycle, the recovered catalyst
was washed with acetone, dried and re-used for an additional recycling test.
2.3 Synthesis of a HMF-rich hydrolyzate from fructose
The hydrolyzate was prepared using a microwave reactor CEM Discover S-class System,
according to the procedure reported by Antonetti et al. at optimum reaction conditions
[13]. At the end of the hydrolysis reaction, the heterogeneous catalyst Amberlyst-70 was
separated by the liquid fraction through centrifugation, and the isolated liquid fraction
was employed as raw feedstock of the subsequent hydrogenation.
9

2.4 Hydrogenation of hydrolyzate with Ru/C
The hydrogenation of the HMF-rich hydrolyzate from fructose was conducted
analogously to that of pure HMF and, also in this case, a weight ratio metal to HMF of 1
wt% was adopted. The progress of the reaction was monitored by sampling periodically
the liquid through a dip tube. The liquid samples were analyzed by HPLC.
2.5 Analytical equipment
2.5.1 High-Pressure Liquid Chromatography
High-pressure liquid chromatography (HPLC) analysis of the liquid samples deriving
from HMF hydrogenation runs was carried out with Perkin Elmer Flexer Isocratic
Platform equipped with a column Benson 2000-0 BP-OA (300 mm x 7.8 mm). A 0.005
M H2SO4 aqueous solution was adopted as mobile phase, maintaining the column at 60
°C with a flow-rate of 0.6 mL/min. The concentrations of products were determined from
calibration curves obtained with standard solutions. Conversion, products yield and
products selectivity were expressed in mol%. The carbon balance was evaluated as the
sum of the moles of products and unconverted HMF respect to the initial moles of HMF
and it was expressed in mol%.
2.5.2 Gas Chromatography coupled with Mass Spectrometer
The by-products formed during the hydrogenation of HMF were qualitatively identified
by gas chromatography coupled with mass spectrometer (GC-MS). Before the analysis,
the aqueous solution was extracted with dichlorometane. A GC-MS (Agilent 7890B-
5977A) equipped with HP-5MS capillary column (30 m × 0.25 mm × 0.25 µm) (5%-
phenyl)-methylpolysiloxane was employed for the analysis. The carrier gas was helium
with a flow of 1 mL/min. The injector and detector temperatures were 250 °C and 280
°C, respectively. The following temperature program was adopted for the
chromatographic run: 70 °C isothermal for 2 min; 12 °C/min up to 250 °C; 250 °C
isothermal for 2 min.
10

2.5.3 Transmission Electron Microscopy
Transmission Electron Microscopy (TEM) measurements in bright field mode were
conducted with a CM12 microscope (Philips), operating at 120 keV. The catalysts were
suspended in ethanol by ultra-sonication, and the obtained sample was dropped onto
carbon coated 400 mesh copper grids. Images were taken on a slow scanning CCD
camera. The ruthenium particle size distribution was evaluated by measuring at least 100
particles with the software Nano Measurer 1.2.
2.5.4 Nitrogen physisorption
Nitrogen physisorption experiments were carried out in a Micromeritics ASAP 2020 at
−196.2 °C. Before the measurement, the samples were degassed under vacuum at 150 °C
for 6 h. The surface area was estimated using the standard BET method. The single point
desorption total pore volume (VT) was calculated from the amount of gas adsorbed at a
relative pressure of 0.98 in the desorption branch.
2.5.5 Thermogravimetric analysis
Thermogravimetric analysis (TGA) of the fresh and used catalysts was determined using
a TGA Q50 system (TA Instrument). The samples were heated in a nitrogen atmosphere,
employing a temperature range between 20 and 650 °C, and a heating rate of 10 °C/min.
2.5.6 Inductively coupled plasma-optical emission spectrometry
Inductively coupled plasma-optical emission spectrometry (ICP-OES) was employed to
determine the metal content in the catalyst after reaction using an Optima 7000 DV
(PerkinElmer) analyser equipped with a CCD array detector. Sample digestion was
carried out in a microwave oven (CEM MARS 5). 20 mg of catalyst was weighted and
introduced in the vessel together with a mixture of HNO3 (7 mL), HCl (1 mL) and HF (2
mL). The vessel was closed and heated at 200 °C for 2 h. Subsequently, the vessel was
cooled to room temperature and diluted to 50 mL with double-distilled water, prior to the
ICP-OES analysis.
11

2.5.7 Gas-phase analysis
The Micro-GC Agilent 3000 equipped with a thermal conductivity detector was
employed for the CO identification. The channel used for CO analysis was the molecular
sieve column Molsieve 5A (10 m x 0.32 mm x 12 μm), adopting argon as carrier gas.
3. Results and discussion
3.1 Metal species screening
Starting from the work of Schiavo et al. [30], a preliminary screening of the catalytic
performances of different commercial catalysts (Ru/C, Pd/C and Pt/C, 5 wt%) was
performed at 140 °C, 70 bar H2, with the initial HMF concentration of 2 wt% and the
metal to HMF ratio of 1 wt%. The results are reported in Table 1.
Table 1, near here
Pt/C resulted the least active system and the HMF conversion was only 64.5 mol% after
1 h of reaction. Both Pd/C and Ru/C were more active and complete HMF conversion
was reached after 1 h. Regarding products distribution, Pt/C gave a very low selectivity
to BHMF (< 16 mol%). GC-MS analysis of the reaction mixture (Figure S1) showed the
presence of several by-products deriving from the hydrodeoxygenation of HMF, such as
2,5-hexanedione, 5-methyl-2-furaldehyde and 5-hydroxy-2-hexanone (Scheme 1A), in
agreement with the literature [58].
Scheme 1, near here
Under the adopted reaction conditions, Pt/C mainly promotes the hydrodeoxygenation
and ring opening of HMF, resulting in a poor selectivity towards the desired furan diols.
On the other hand, both Pd/C and Ru/C favour hydrogenation reactions, leading to
improved BHMTHF yields (55.8 and 88.6 mol% for Pd/C and Ru/C, respectively).
Among the three catalysts, Ru/C is surely the most promising, combining high HMF
conversion with high BHMTHF selectivity, as also confirmed by the best carbon balance,
the closest to 90 mol%. In this case, only minor amounts of by-products were detected,
12

such as tetrahydrofurfuryl alcohol, 5-methyl-tetrahydrofurfuryl alcohol, 2,5-
dimethyltetrahydrofuran and tetrahydro-2H-pyran-2-methanol. These are known by-
products, deriving from BHMTHF degradation reactions which are promoted at elevated
temperatures [59] (Figure S2 and Scheme 1B). For Pd/C catalyst, not only the
hydrogenation of both aldehyde group and furan ring of HMF occurs, but also
hydrodeoxygenation and ring opening reactions (Scheme 1C), as ascertained by the
presence of typical by-products, such as 5-methyl-tetrahydrofurfuryl alcohol, 2,5-
hexanedione, 5-hydroxy-2-hexanone and 1,2,6-hexanetriol (Figure S3), in agreement
with the literature [55,60,61].
On the basis of this exploratory screening, Ru/C was identified as the most promising
commercial catalyst for the aqueous hydrogenation of HMF to the target diols, and
therefore it was adopted for subsequent optimization studies.
3.2 Optimization of BHMF and BHMTHF yields in the presence of Ru/C
The above reported preliminary screening has been performed under harsh reaction
conditions (140 °C and 70 bar H2). Subsequently, in order to improve the selectivity of
the reaction, milder reaction conditions have been adopted, in terms of temperature and
H2 pressure. In particular, the influence of temperature (100-140 °C) on the catalytic
performances at 70 bar H2 was investigated, and the obtained results are reported in
Figure 1.
Figure 1, near here
The conversion of HMF was almost complete already at short reaction time (30 min), for
all the adopted temperatures. On the other hand, temperature strongly influenced the
products distribution. In fact, at 140 °C (Figure 1A), the amount of BHMF was negligible
during the whole reaction due to the extensive hydrogenation of the furan ring and
BHMTHF yield of 79.3 mol% was ascertained after 30 min. BHMTHF yield reached the
maximum value of 88.6 mol% after 60 min, and then decreased, due to the formation of
13

by-products, as evidenced by the corresponding trend of carbon balance (run 4, Table
S1). Working at 120 °C (Figure 1B), the hydrogenation of the furan ring was slower and,
after 30 min, the BHMF yield of 16.0 mol% was obtained. The maximum BHMTHF yield
shifted from 60 to 180 min (Figures 1A and 1B), when it resulted higher than that
ascertained at 140 °C, reaching 92.8 mol%. These results underline that the by-products
formation is favoured at high temperature, as confirmed by the corresponding trend of
carbon balance at the different temperatures reported in Table S1, and by the results
obtained decreasing the reaction temperature up to 100 °C (Figure 1C). In fact, in this last
case, the BHMTHF yield continuously increased with the time, reaching the highest value
of 95.0 mol% after 240 min.
On this basis, the effect of the decrease of H2 pressure to 50 bar was further investigated
working at 100 °C, and the obtained results are reported in Table 2.
Table 2, near here
The decrease of H2 pressure did not influence the HMF conversion, that resulted almost
complete (compare runs 7 and 8, Table 2) during the whole reaction. On the other hand,
at short reaction time (30 min), the lower pressure led to a higher BHMF yield at the
expense of BHMTHF yield, due to the reduced hydrogenation of the furan ring.
The concentration of the starting feedstock is another very important parameter because,
usually, high substrate concentrations promote side-reactions, and the yields and/or
selectivities towards target products fall down. The initial HMF concentration was
increased from 2 to 3 wt% (runs 8 and 9, Table 2) at 50 bar H2, on the basis of the HMF
concentration obtained from fructose in a previous study adopting Amberlyst A-70 as
acid catalyst in water [13], in the perspective of a feasible cascade approach.
The presence of a higher amount of the initial substrate did not limit the hydrogenation
of the aldehyde group of HMF, whose conversion resulted unchanged, but caused a
slowdown of the furan ring hydrogenation, as evidenced in particular at short reaction
14

times. However, at the end of the reaction, analogous BHMTHF yields were ascertained
starting from 2 and 3 wt% HMF solutions. Moreover, in order to prove the key role of
Ru/C towards the activation of the HMF hydrogenation, a blank run without the catalyst
was performed under the same reaction conditions (100 °C, 50 bar H2, 3 wt% HMF
solution). In this case, after 240 min, the conversion of HMF resulted 10.6 mol% and only
BHMF in trace was detected, confirming the necessity of employing a suitable catalyst
for the hydrogenation of HMF. In conclusion, the highest BHMTHF yield of 95.3 mol%
was reached starting from 3 wt% HMF aqueous solution at 100 °C, 50 bar H2 after 240
min. This represents a very promising result, considering that BHMTHF yields over 90
mol% have been reported working only on less concentrated water solution, adopting
higher H2 pressures [56,57], and/or higher temperatures [57], longer reaction times
[48,57] and, in the presence of ad hoc synthetized catalysts, which are not still really
interesting for industrial applications in the immediate future.
Once having optimized the synthesis of BHMTHF, this study was focused on the
optimization of BHMF synthesis, where the sole hydrogenation of the aldehyde group of
HMF is required. The above results suggested that it was necessary to adopt milder
reaction conditions and thus, the H2 pressure was reduced to 30 bar (run 10, Table 2). The
hydrogenation of HMF was slowed down and, for the first time, its conversion was not
complete within the first hour of reaction. As a consequence, also the hydrogenation of
the furan ring was limited, causing the increase of BHMF yield, which resulted 79.5
mol% after 30 min. However, under the H2 pressure of 30 bar, BHMF underwent other
side-reactions, as evidenced by the worsening of the carbon balance in run 10, reaching
the value of 29.6 mol%, after 240 min. The pressure of 30 bar H2 was not sufficient to
promote the hydrogenation of the furan ring, which requires high temperatures and high
pressures to occur [53]. However, as reported in the literature [59,62], at high
temperatures and low pressures, the ring opening of BHMF prevailed, resulting faster
15

than the hydrogenation of the furan ring. In fact, the low hydrogen pressure is
disadvantageous for hydrogen solubilisation in water, inhibiting the conversion of
BHMF to BHMTHF, and promoting the formation of partially hydrogenated products,
which are intermediates for the synthesis of polyols, such as 1,2,6-hexanetriol, 1,2-
hexanediol, 1,6-hexanediol, 1-hydroxyl-2,5-hexanedione and 1,2,5-hexanetriol [30, 59].
In fact, as reported in the literature, 1,2,6-hexanetriol derives from the ring opening of
BHMF and hydrogenation of the intermediate, whereas 1,2-hexanediol and 1,6-
hexanediol originate from the break of the C-O bond of C6 or C2 of 1,2,6-hexanetriol,
respectively [59]. Regarding 1-hydroxyl-2,5-hexanedione and 1,2,5-hexanetriol, the first
one derives from the rearrangement of BHMF, favoured in water, followed by ring
opening, whereas 1,2,5-hexanetriol is the product of 1-hydroxyl-2,5-hexanedione
complete hydrogenation [30]. In order to optimize the BHMF production, the
investigation of temperature within the range 50-120 °C was carried out at 30 bar H2, and
the results are reported in Figure 2.
Figure 2, near here
As expected, HMF conversion increased with temperature, which strongly influences the
distribution of products. In fact, when the reaction was performed at 50 °C, the BHMF
yield continuously increased, reaching the highest value of 93.0 mol%, after 240 min.
When the temperature was raised to 70 °C, the maximum of the BHMF yield (90.0 mol%)
shifted to shorter reaction time (120 min), and then it strongly decreased by prolonging
the reaction. The further increase of the reaction temperature, first to 100 °C and then to
120 °C, promoted the BHMF decomposition, as confirmed also by the progressive
decrease of carbon balance (runs 13 and 14, Table S2).
Therefore, the highest BHMF yield (93.0 mol%) was reached on 3 wt% HMF aqueous
solution at 50 °C, 30 bar H2, and after 240 min, with a Ru/HMF ratio of 1 wt% (run 11).
16

In the literature, analogous BHMF yields are reported starting from aqueous HMF
solutions only employing significantly less sustainable reaction conditions [27,45,46,57].
3.3 Hydrogenation of crude HMF-rich hydrolyzate obtained from fructose dehydration
The synthesis of BHMF and BHMTHF starting from pure HMF is scarcely attractive in
an industrial perspective due to the high cost of HMF, caused by its low yield in both
production and purification steps. On this basis, the hydrogenation of a crude HMF-rich
hydrolyzate was also investigated, thus evaluating the effect of other compounds, which
are typical of a real HMF-rich hydrolyzate, on the catalytic performances towards the
next HMF hydrogenation step. The hydrolyzate was obtained from the dehydration of
fructose, according to our previous work, in the presence of the commercial resin
Amberlyst-70 as acid catalyst, and the best HMF yield of 45.6 mol% was reached [13].
At the end of the hydrolysis reaction, the catalyst was separated by filtration and the
hydrolyzate was composed of 3 wt% of HMF, 2 wt% of unreacted fructose, 0.08 wt% of
formic acid and 0.15 wt% of levulinic acid, showing a pH=2.6, due to the significant
presence of the organic acids. This real hydrolyzate was subjected to hydrogenation at
100 °C and 50 bar H2 (Figure 3, run 15).
Figure 3, near here
Comparing the above reaction profile with that of the hydrogenation of pure HMF, which
was carried out under the same reaction conditions (run 9, Table 2), it is evident that,
starting from the real hydrolyzate, the HMF conversion and the yields of the diols were
significantly lower than those achieved starting from pure HMF. This is due to the
significant formation of by-products, as confirmed by the very low carbon balance (run
15, Table S3). These include the unconverted fructose, rehydration acids, formic and
levulinic ones, and soluble humins. In order to verify the influence of these compounds
on the hydrogenation performances, some model mixtures, having the typical
concentrations of the raw hydrolyzate, were prepared, thus separately investigating the
17

effect of the addition of these components on the HMF hydrogenation. In this regard, four
model mixtures were prepared and hydrogenated: 1) HMF (3 wt%) with fructose (2 wt%)
(Figure 4A, run 16); 2) HMF (3 wt%) with formic acid (0.08 wt%) and levulinic acid
(0.15 wt%) (Figure 4B, run 17); 3) HMF (3 wt%) with formic acid (0.08 wt%) (Figure
4C, run 18); 4) HMF (3 wt%) with levulinic acid (0.15 wt%) (Figure 4D, run 19).
Figure 4, near here
The HMF hydrogenation in the presence of fructose (Figure 4A) proceeded similarly to
that of pure HMF (run 10, Table 2), showing that the presence of the unreacted
monosaccharide had no influence on the cascade reaction. On the contrary, formic and
levulinic acids had a detrimental effect on the hydrogenation of HMF, causing a
significant decrease of the reaction rate and a marked drop of BHMF and BHMTHF
yields, which respectively reached only 10 and 5 mol% after 30 and 120 min, respectively
(Figure 4B). As before evidenced, this result is in agreement with the literature. In fact,
not only the acid conditions promote the decomposition of the furan diols [30,57], but it
is known the strong deactivating adsorption of formic acid, which remained in the
reaction mixture because its decomposition to CO/CO2 was not significant under the
adopted mild conditions [63]. This peculiar behaviour of formic acid was also confirmed
comparing the catalyst performances in the hydrogenation of the HMF model mixtures
with formic (Figure 4C) or levulinic acids (Figure 4D). In fact, in the presence of formic
acid, the conversion of HMF was slower than that found in the HMF hydrogenation with
levulinic acid. This is in agreement with the literature results, already reported for the
hydrogenation of levulinic acid, where it is underlined that formic acid can be easily and
strongly adsorbed on Ru particles in its formate form, limiting the availability of the
active sites for the substrate [63-65]. The deactivation of the catalyst, due to the presence
of formic acid, was also evidenced by the products formation. In fact, in Figure 4C the
conversion of HMF did not lead to diols but rather to other by-products, indicating that
18

the hydrogenation of HMF was strongly limited. On the other hand, in the presence of
levulinic acid (Figure 4D), considerable amount of BHMF was obtained at short reaction
time, proving that the HMF hydrogenation occurred, but the acidity of the mixture had a
detrimental effect with prolonging the reaction, causing the decreasing of the furan diols
yields. Moreover, the formation of humins, deriving from HMF acid condensation
[12,13,66], contributed to the catalyst surface passivation [67]. Their formation was
confirmed by the very low carbon balance ascertained during the whole reaction in the
presence of rehydration acids (runs 17, 18 and 19, Table S3). However, the conversion of
HMF reached in the raw hydrolyzate was even lower than that starting from the model
mixtures of HMF with rehydration acids, due to the presence of soluble humins already
present in the raw hydrolyzate.
In order to overcome this drawback, the raw hydrolyzate was neutralized with NaHCO3
until pH = 7, and then subjected to hydrogenation at 100 °C and 50 bar H2 (Figure 5).
Figure 5, near here
The neutralization gave an improvement of the catalytic performances, and the BHMF
yield markedly improved, reaching the value of 73.2 mol% respect to the starting amount
of HMF. This value corresponds to a BHMF yield of 33.4 mol% respect to the starting
fructose employed in this cascade approach, being the yields of HMF from fructose in the
hydrolysis step equal to 45.6 mol% [13]. However, comparing this run with the
hydrogenation of pure HMF (run 9, Table 2), HMF conversion (Figure 5) and the carbon
balance (run 20, Table S3) for the neutralized hydrolyzate resulted still lower, and the
major product was BHMF, instead of BHMTHF, underlining that the hydrogenation
reaction remained almost limited. This evidence can be justified taking into account that
the neutralizing step counteracted the acid conditions, responsible for the ring opening
by-products and further humins formation in the hydrogenation step, but the passivation
19

effect of soluble humins already present in the hydrolyzate remained, thus limiting the
hydrogenation reaction [67].
Regarding the reaction mechanism of HMF hydrogenation, it is well-known in the
literature that Ru-based catalysts favor the hydrogenation of C=O to give 2,5-
bis(hydroxymethyl)furan at relatively low temperatures, which would further be
converted to 2,5-dimethylfuran via hydrogenolysis, occurring at relatively high
temperatures, with 5-methylfurfuryl alcohol and 2,5-hexanedione as intermediate and by-
product, respectively [68]. Moreover, regarding the reactivity of HMF, literature studies
on aldehydes have shown that decarbonylation path takes place on metals of groups 8, 9,
and 10, including ruthenium, especially at high temperatures, leading to the formation of
furfuryl alcohol and CO [60]. On the basis of our data, in order to experimentally confirm
the HMF hydrogenation mechanism as the main one responsible for the production of
BHMF and BHMTHF performed under mild reaction conditions, 50 °C, 30 bar H2 and
100 °C, 50 bar H2 respectively, the reaction mixtures obtained under these conditions
starting from pure HMF were analysed by GC-MS and the gas-phase reaction products
by GC analysis. Only trace amounts of products deriving from hydrogenolysis or
decarbonylation reactions of HMF and/or of BHMF, and/or from subsequent
hydrogenation/hydrogenolysis reactions on the obtained hydrogenolysis or
decarbonylation products were detected (Figure S4 and S5). These products can include
5-methylfurfural, furfuryl alcohol, 5-methylfurfuryl alcohol, 5-methyltetrahydrofurfuryl
alcohol, 2,5-dimethyltetrahydrofuran, tetrahydrofurfuryl alcohol, some of which were
present in low amounts (not negligible) when the reaction was performed at 140 °C, 70
bar H2, as already reported in Figure S2, confirming that hydrogenolysis and
decarbonylation pathways become more important at high temperatures. Also in the gas-
phase, only trace amounts of CO were detected, in agreement with the literature,
highlighting as the hydrogenation mechanism is the main one for ruthenium catalysts in
20

the production of BHMF and BHMTHF from HMF [68]. In this regard, it is reasonable
that, when the C=O hydrogenation is the main reaction pathway, the preferential HMF
adsorption mode on the active metal occurs in the η2(C,O)-aldehyde configuration. By
this way, BHMF could be selectively formed from this η2(C,O) species,. Once BHMF
was obtained in the reaction mixture, this molecule may be adsorbed in two different
modes for the subsequent hydrogenation step: parallel and tilted. The parallel mode may
lead to complete hydrogenation, forming BHMTHF, whereas the tilted one may cause
the ring opening, through the C-O bond cleavage, with the final formation of 1,2,6-
hexanetriol, after hydrogenation step. This proposed mechanism is reported in the
Scheme 2, and it is in agreement with the literature data [59,60].
Scheme 2, near here
In order to better evaluate the amount of carbonaceous material on the catalyst surface at
the end of the reaction and how it affects the physical properties of the employed catalyst,
TGA (Figure 6) and N2 physisorption (Figure S6 and Table S4) analyses were carried out
on fresh and spent Ru/C catalysts at the end of hydrogenation reactions performed
adopting different starting materials: solutions of pure HMF (run 9, Table 2), the raw
hydrolyzate (run 15, Figure 3) and the neutralized one (run 20, Figure 5).
Figure 6, near here
Figure 6 shows that the amount of carbonaceous material (humins) on the spent catalysts
is strongly influenced by the type of the starting substrate. In fact, when the raw
hydrolyzate was employed as starting material, the lowest residual weight was acquired
at the end of the analysis, confirming that, in this case, the highest amount of humins was
deposited on the catalyst, originating from both the crude hydrolyzate and the HMF
condensation that took place during the hydrogenation reaction. The neutralizing step
allowed the reduction of humins formation, thus the residual weight recorded at the end
of the analysis was higher than that obtained for the crude hydrolyzate, but lower than
21

that for the catalyst employed in the hydrogenation of pure HMF. This explains the trend
found for HMF conversion and it is in agreement with the N2 physisorption experiments
reported in Figure S6 and Table S4. In fact, the isothermal curves and the specific surface
area values show that the surface area of the spent catalysts depend on the adopted
substrate, following this order: pure HMF (153 m2/g) > neutralized hydrolyzate (62 m2/g)
> raw hydrolyzate (6 m2/g), being equal to 770 m2/g that of fresh Ru/C system. The
catalyst support plays a significant influence on the catalytic activity in the selective
hydrogenation of HMF. The obtained catalytic trend is in agreement with the literature:
supports with high surface area favor the dispersion of active metal particles on their
surfaces, providing more active catalytic sites for the hydrogenation reactions [36].
Moreover, it is evident that the surface area of the catalyst recovered after the
hydrogenation of pure HMF was lower than that of the fresh Ru/C, indicating that, also
in this case, some organic material could be adsorbed on the catalyst surface, as
previously observed by the comparison of the thermogravimetric curves of these two
catalysts reported in Figure 6.
In order to improve the yields towards BHMTHF starting from the crude HMF, harsher
reaction conditions (140 °C, 70 bar H2) were adopted, and the results are shown in Figure
7 (run 21).
Figure 7, near here
Both HMF conversion and carbon balance were similar to those obtained working at 100
°C and 50 bar H2, but the product distribution significantly changed. In fact, in this case,
the prevailing furan diol resulted BHMTHF, which after 240 min reached the yield of
81.1 mol% respect to the amount of initial HMF present in the hydrolyzate, which
corresponds to the value of 37.0 mol% respect to the starting fructose, taking into account
that in the first step the yield of HMF starting from fructose was 45.6 mol% [13].
22

Up to now, only few papers report the synthesis of these diols directly from fructose
[31,50,69,70]. However, to the best of our knowledge, in this work, for the first time, both
fructose dehydration and hydrolyzate hydrogenation were carried out in water instead of
organic solvents, ionic liquids or organic-water mixtures. In particular, the first step of
this cascade approach is the most critical, due to the possible decomposition of HMF to
humins and rehydration acids, which causes the lowering of the HMF yield respect to
those obtained with different solvent systems, which should allow an almost quantitative
HMF yield [7]. Therefore, in the second step (HMF hydrogenation), the literature
investigations performed in organic solvent are based on hydrolyzates which don’t
include the presence of rehydration acids and humins, thus allowing the maximization of
the furan diols yields. However, the employment of organic media or ionic liquid makes
the literature processes significantly less sustainable under economic, environmental and
safety points of view.
3.4 Catalyst stability
When a heterogeneous catalyst is employed, the evaluation of its stability is an essential
issue. For this purpose, the fresh and spent Ru/C catalysts recovered at the end of the
optimized reactions for the synthesis of both BHMTHF (run 9, Table 2) and BHMF (run
11, Figure 2), both starting from pure HMF were analysed through ICP-OES and TEM
techniques. The first one proved that the leaching of ruthenium in the solution was
negligible when it was employed for the synthesis of BHMF and BHMTHF. The TEM
pictures and the distributions of the ruthenium particles size for the fresh and the spent
Ru/C catalysts are reported in Figure 8.
Figure 8, near here
The TEM image of fresh Ru/C catalyst shows that this system is characterized by
ruthenium particles with very small average size, 1.5 nm, in agreement with the results
reported in the literature [71]. On the other hand, the ruthenium particles sizes in the spent
23

catalysts were 2.5 and 2.3 nm, for those employed for the synthesis of BHMTHF and
BHMF, respectively. In order to investigate the recyclability of the catalyst, the catalytic
system employed in run 9 (Table 2) was recovered at the end of the reaction by filtration,
and reused in two subsequent tests, using the same reaction conditions adopted in run 9.
The obtained results are reported in Figure 9.
Figure 9, near here
During these three cycles (1, 2 and 3), a slight decrease of the catalytic activity was
observed. In fact, the HMF conversion was not complete in the recycling runs and, after
the third one, a decrease of 13.8 mol% was obtained. Moreover, a modest increase of
BHMF yield (4.2 mol% in the third cycle) was observed, due to the passivation of catalyst
surface. At the end of the third cycle, the recovered catalyst was washed with acetone,
dried and reused again in another subsequent recycling test. After the washing treatment,
the catalyst performances were almost entirely restored, proving that the increase of
ruthenium particle sizes did not influence the catalytic activity and confirming that the
adopted washing treatment represents an efficient and simple reactivation method, able
to remove humins from catalyst surface.
These results underline that catalytic performances of the Ru/C catalyst can be restored,
in agreement with our previous research on hydrogenation of raw biomass-derived
levulinic acid to γ-valerolactone (GVL) [65], or to 2-methyltetrahydrofuran, and to 2-
butanol [72]. The prevailing deactivation of the catalyst can be related only to humin
deposition on the surface, which could be removed through washing and/or thermal
treatments.
4. Conclusion
Ru/C, Pd/C and Pt/C catalysts were studied in the hydrogenation of pure HMF aqueous
solutions to obtain the furan diols 2,5-bis(hydroxymethyl)furan (BHMF) and 2,5-
bis(hydroxymethyl)tetrahydrofuran (BHMTHF). Under the same reaction conditions,
24

Pt/C and Pd/C promoted the HMF hydrodeoxygenation and ring opening, whereas Ru/C
mainly activated the HMF hydrogenation, thus resulting as the best catalyst, in terms of
conversion and selectivity, towards the desired products. The investigation on Ru/C
catalyst revealed that mild reaction conditions were appropriate for obtaining high BHMF
yield, whereas higher temperature and H2 pressure were necessary to hydrogenate also
the furan ring, thus selectively obtaining BHMTHF. From the composition of the reaction
mixtures, in terms of ascertained by-products, informations on the reaction mechanism
were inferred. The hydrogenation of HMF-rich hydrolyzate obtained from the
dehydration of fructose aqueous solution was subsequently studied. The investigation
evidenced the detrimental role of formic and levulinic acids, which promote the formation
of ring opening by-products and humins, which can passivate the catalyst surface.
However, the neutralization of the hydrolyzate allowed the improvement of the catalyst
performances, preventing the humins formation, as confirmed by N2 physisorption and
TGA analyses of spent catalysts. These results evidence, for the first time, the feasibility
of the BHMF and BHMTHF synthesis with good yields, starting from aqueous crude
HMF and commercial Ru/C catalyst. Moreover, the recycling data obtained in batch
reactor are promising and experiments in continuous set-up are now in progress in order
to investigate the catalyst performances for long time on stream.
References
[1] I.K.M. Yu, D.C.W. Tsang, Biores. Technol. 238 (2017) 716-732.
[2] C. Antonetti, D. Licursi, S. Fulignati, G. Valentini, A.M. Raspolli Galletti, Catalysts
6 (2016) 196-224.
[3] L. Hu, L. Lin Z. Wu, S. Zhou, S. Liu, Renew. Sust. Energ. Rev. 74 (2017) 230-257.
[4] D.P. Serrano, J.A. Melero, G. Morales, J. Iglesias, P. Pizarro, Catal. Rev. 60 (2018)
1-70.
25

[5] C. Antonetti, E. Bonari, D. Licursi, N. Nassi o Di Nasso, A.M. Raspolli Galletti,
Molecules 20 (2015) 21232-21253.
[6] A. Galia, B. Schiavo, C. Antonetti, A.M. Raspolli Galletti, L. Interrante, M. Lessi, O.
Scialdone, M.G. Valenti, Biotechnol. Biofuels 8 (2015) 218-236.
[7] S.P. Teong, G. Yi, Y. Zhang, Green Chem. 16 (2014) 2015-2026.
[8] M. Li, W. Li, Y. Lu, H. Jameel, H. Chang, L. Ma, RSC Adv. 7 (2017) 14330-14336.
[9] B. Agarwal, K. Kailasam, R.S. Sangwan, S., Elumalai, Renew. Sust. Ener. Rev., 82
(2018) 2408-2425.
[10] A.A. Marianou, C.M. Michailof, A. Pineda, E.F. Iliopoulou, K.S. Triantafyllidis,
A.A. Lappas, Appl. Catal. A: Gen. 555 (2018) 75-87.
[11] L. Atanda, A. Shrotri, S. Mukundan, Q. Ma, M. Konarova, J. Beltramini,
ChemSusChem 8 (2015) 2907-2916.
[12] C. Antonetti, M. Melloni, D. Licursi, S. Fulignati, E. Ribechini, S. Rivas, J.C. Parajó,
F. Cavani, A.M. Raspolli Galletti, Appl. Catal. B: Env. 206 (2017) 364-377.
[13] C. Antonetti, A.M. Raspolli Galletti, S. Fulignati, D. Licursi, Catal. Commun. 97
(2017) 146-150.
[14] A.N. Chermahini, H. Hafizi, N. Andisheh, M. Saraji, A. Shahvar, Res. Chem.
Intermed. 43 (2017) 5507-5521.
[15] I.A. Masiutin, A.A. Novikov, A.A. Litvin, D.S. Kopitsyn, D.A. Beskorovaynaya,
E.V. Ivanov, Starch/Stärke 68 (2016) 637-643.
[16] X. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr., Green Chem. 12 (2010) 1855-1860.
[17] H. Shirai, S. Ikeda, E.W. Qian, Fuel Process. Technol. 159 (2017) 280-286.
[18] R. Weingarten, A. Rodriguez-Beuerman, F. Cao, J.S. Lutherbacher, D.M. Alonso,
J.A. Dumesic, G.W. Huber, ChemCatChem 6 (2014) 2229-2234.
[19] Y. Yang, C.W. Hu, M.M. Abu-Omar, ChemSusChem 5 (2012) 405-410.
[20] M.J. Gilkey, D.G. Vlachos, B. Xu, Appl. Catal. A: Gen. 542 (2017) 327-335.
26

[21] J. Li, J. Liu, H. Liu, G. Xu, J. Zhang, J. Liu, G. Zhou, Q. Li, Z. Xu, Y. Fu,
ChemSusChem 10 (2017) 1436-1447.
[22] G. Raveendra, A. Rajasekhar, M. Srinivas, P.S. Sai Prasad, N. Lingaiah, Appl. Catal.
A: Gen. 520 (2016) 105-113.
[23] B.O. De Beeck, M. Dusselier, J. Geboers, J. Holsbeek, E. Morré, S. Oswald, L.
Giebeler, B.F. Sels, Energy Environ. Sci. 8 (2015) 230-240.
[24] G. Yi, S.P. Teong, Y. Zhang, Green Chem. 18 (2016) 979-983.
[25] D.X. Martínez-Vargas, J.R. De La Rosa, L. Sandoval-Rangel, J.L. Guzmán-Mar,
M.A. Garza-Navarro, C.J. Lucio-Ortiz, D.A. De Haro-Del Río, Appl. Catal. A:
Gen. 547 (2017) 132-145.
[26] G. Bottari, A.J. Kumalaputri, K.K. Krawczyk, B.L. Feringa, H.J. Heeres, K. Barta,
ChemSusChem 8 (2015) 1323-1327.
[27] M. Chatterjee, T. Ishizaka, H. Kawanami, Green Chem. 16 (2014) 4734-4739.
[28] J.J. Roylance, T.W. Kim, K. Choi, ACS Catal. 6 (2016) 1840-1847.
[29] N. Perret, A. Grigoropoulos, M. Zanella, T.D. Manning, J.B. Claridge, M.J.
Rosseinsky, ChemSusChem 9 (2016) 521-531.
[30] R. Alamillo, M. Tucker, M. Chia, Y. Pagán-Torres, J. Dumesic, Green Chem. 14
(2012) 1413-1419.
[31] H. Cai, C. Li, A. Wang, T. Zhang, Catal. Today 234 (2014) 59-65.
[32] 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.
[33] T. Buntara, I. Melián-Cabrera, Q. Tan, J.L.G. Fierro, M. Neurock, J.G. De Vries,
H.J. Heeres, Catal. Today 210 (2013) 106-116.
[34] K. Gupta, R.K. Rai, S.K. Singh, ChemCatChem 10 (2018) 2326-2349
[35] B. Ma, Y. Wang, X. Guo, X. Tong, C. Liu, Y. Wang, X. Guo, Appl. Catal. A: Gen.
552 (2018) 70-76.
27

[36] X. Tang, J. Wei, N. Ding, Y. Sun, X. Zeng, L. Hu, S. Liu, T. Lei, L. Lin, Renew.
Sust. Ener. Rev. 77 (2017) 287-296.
[37] M. Gomes, A. Gandini, A.J.D. Silvestre, B. Reis, J. Polym. Sci. Part A: Polym.
Chem. 49 (2011) 3759-3768.
[38] D. Zhang, M. Dumont, J. Polym. Sci., Part A: Polym. Chem. 55 (2017) 1478-1492.
[39] L. Hu, J. Xu, S. Zhou, A. He, X. Tang, L. Lin, J. Xu, Y. Zhao, ACS Catal. 8 (2018)
2959-2980.
[40] A. Gandini, T.M. Lacerda, A.J.F. Carvalho, E. Trovatti, Chem. Rev. 116 (2016)
1637-1669.
[41] T. Thananatthanachon, T.B. Rauchfuss, ChemSusChem 3 (2010) 1139-1141.
[42] T. Wang, J. Zhang, W. Xie, Y. Tang, D. Guo, Y. Ni, Catalysts 7 (2017) 92-99.
[43] L. Hu, M. Yang, N. Xu, J. Xu, S. Zhou, X. Chu, Y. Zhao, Korean J. Chem. Eng. 35
(2018) 99-109.
[44] T. Pasini, G. Solinas, V. Zanotti, S. Albonetti, F. Cavani, A. Vaccari, A. Mazzanti,
S. Ranieri, R. Mazzoni, Dalton Trans. 43 (2014) 10224-10234.
[45] J. Ohyama, A. Esaki, Y. Yamamoto, S. Arai, A. Satsuma, RSC Adv. 3 (2013) 1033-
1036.
[46] J. Chen, F. Lu, J. Zhang, W. Yu, F. Wang, J. Gao, J. Xu, ChemCatChem 5 (2013)
2822-2826.
[47] J. Han, Y.H. Kim, H.S. Jang, S.Y. Hwang, J. Jegal, J.W. Kim, Y.S. Lee, RSC Adv.
6 (2016) 93394-93397.
[48] J. Chen, R. Liu, Y. Guo, L. Chen, H. Gao, ACS Catal. 5 (2015) 722-733.
[49] Y. Nakagawa, K. Takada, M. Tamura, K. Tomishige, ACS Catal. 4 (2014) 2718-
2726.
[50] P.P. Upare, Y.K. Hwang, D.W. Hwang, Green Chem. 20 (2018) 879-885.
28

[51] Y. Zhu, X. Kong, H. Zheng, G. Ding, Y. Zhu, Y.W. Li, Catal. Sci, Technol. 5 (2015)
4208-4217.
[52] X. Kong, Y. Zhu, H. Zheng, F. Dong, Y. Zhu, Y.W. Li, RSC Adv. 4 (2014) 60467-
60472.
[53] S. Chen, R. Wojcieszak, F. Dumeignil, E. Marceau, S. Royer, Chem. Rev. 18 (2018)
11023–11117.
[54] F. Liu, M. Audemar, K. De Oliveira Vigier, J.M. Clacens, F. De Campo, F. Jérôme,
Green Chem. 16 (2014) 4110-4114.
[55] X. Hu, S. Kadarwati, Y. Song, C.Z. Li, RSC Adv. 6 (2016) 4647-4656.
[56] Y. Nakagawa, K. Tomishige, Catal. Commun. 12 (2010) 154-156.
[57] V. Schiavo, G. Descotes, J. Mentech, Bull. Soc. Chim. Fr. 128 (1991) 704-711.
[58] J. Luo, L. Arroyo-Ramirez, R.J. Gorte, AIChE 61 (2015) 590-597.
[59] S. Yao, X. Wang, Y. Jiang, F. Wu, X. Chen, X. Mu, ACS Sustainable Chem. Eng. 2
(2014) 173-180.
[60] D.P. Duarte, R. Martínez, L.J. Hoyos, Ind. Eng. Chem. Res. 55 (2016) 54-63.
[61] J. Mitra, X. Zhou, T. Rauchfuss, Green Chem. 17 (2015) 307-313.
[62] B. Zhang, Y. Zhu, G. Ding, H. Zheng, Y. Li, Green Chem. 14 (2012), 3402-3409.
[63] H. Heeres, R. Handana, D. Chunai, C.B. Rasrendra, B. Girisuta, H.J. Heeres, Green
Chem. 11 (2009) 1247-1255.
[64] A.M. Ruppert, M. Jędrzejczyk, O. Sneka-Płatek, N. Keller, A.S. Dumon, C. Michel,
P. Sautet, J. Grams, Green Chem. 18 (2016) 2014–2028.
[65] S. Rivas, A.M. Raspolli Galletti, C. Antonetti, D. Licursi, V. Santos, J.C. Parajó,
Catalysts 8 (2018) 169-184.
[66] G. Tsilomelekis, M.J. Orella, Z. Lin, Z. Cheng, W. Zheng, V. Nikolakis, D.G.
Vlachos, Green Chem. 18 (2016) 1983-1993.
29