Etherification of biomass-derived furanyl alcohols with aliphatic alcohols over silica-supported nickel phosphide catalysts: Effect of surplus P species on the acidity

Article abstract of DOI:10.1016/j.apcata.2020.117763

The acidity of nickel phosphide (Ni₂P) catalysts plays a crucial role in producing a desired hydrodeoxygenation molecule from biomass-derived substrates; yet, it has never been explored in acid-catalyzed reactions. Herein, we demonstrated the activity of silica-supported Ni₂P catalyst prepared with the nominal P/Ni ratio of 2 (Ni₂P/SiO₂-2P) in the etherification of furanyl alcohols (particularly, 5-(hydroxymethyl)furfural) with aliphatic alcohols including ethanol. By comparing the characteristics of Ni/SiO₂, P_xO_y/SiO₂, and Ni₂P/SiO₂-xP (x = 0.5 and 1), Ni₂P/SiO₂-2P was revealed to contain the Br?nsted and Lewis acid sites of which both contributed to the etherification reaction. Notably, the Br?nsted acidity was associated with the surplus P species added to produce the Ni₂P phase. Consequently, supported Ni₂P catalysts can work in acid-catalyzed reactions if an adequate ratio of Br?nsted to Lewis acid sites is provided by the amount of the surplus P species determined by adjusting the P/Ni ratio.

Full text of DOI:10.1016/j.apcata.2020.117763

Journal Pre-proof
Etherification of biomass-derived furanyl alcohols with aliphatic alcohols
over silica-supported nickel phosphide catalysts: Effect of surplus P
species on the acidity
Mi Shin (Data curation) (Formal analysis) (Investigation)
(Methodology) (Validation)<ce:contributor-role>Writing -original
draft) (Visualization), Jinsung Kim (Formal analysis) (Investigation),
Young-Woong Suh (Conceptualization) (Funding acquisition)
(Project administration) (Supervision) (Writing - review and editing)
PII:
S0926-860X(20)30356-2
DOI:
https://doi.org/10.1016/j.apcata.2020.117763
APCATA 117763
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
1 June 2020
10 July 2020
23 July 2020
Please cite this article as: { doi: https://doi.org/
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.

Etherification of biomass-derived furanyl alcohols with aliphatic
alcohols over silica-supported nickel phosphide catalysts: Effect
of surplus P species on the acidity
Mi Shin,a Jinsung Kim,a and Young-Woong Suha,b,*
aDepartment of Chemical Engineering, Hanyang University, Seoul 04763, Republic of Korea
bResearch Institute of Industrial Science, Hanyang University, Seoul 04763, Republic of Korea
*Corresponding author. E-mail address: ywsuh@hanyang.ac.kr (Y.-W. Suh)
Graphical Abstract
Highlights
 Ni2P/SiO2 catalysts are capable of etherifying HMF with aliphatic alcohols.
 Furanyl alcohols can be etherified with ethanol over Ni2P/SiO2.
 The P–OH group could be located on the surface of Ni2P particles and SiO2
support.
 The Brønsted acidity of Ni2P/SiO2 is associated with the surplus P species.
 Both Brønsted and Lewis acid sites contribute to the activity of Ni2P/SiO2.
1

Abstract
The acidity of nickel phosphide (Ni2P) catalysts plays a crucial role in producing a desired
hydrodeoxygenation molecule from biomass-derived substrates; yet, it has never been
explored in acid-catalyzed reactions. Herein, we demonstrated the activity of silica-supported
Ni2P catalyst prepared with the nominal P/Ni ratio of 2 (Ni2P/SiO2-2P) in the etherification of
furanyl alcohols (particularly, 5-(hydroxymethyl)furfural) with aliphatic alcohols including
ethanol. By comparing the characteristics of Ni/SiO2, PxOy/SiO2, and Ni2P/SiO2-xP (x = 0.5
and 1), Ni2P/SiO2-2P was revealed to contain the Brønsted and Lewis acid sites of which both
contributed to the etherification reaction. Notably, the Brønsted acidity was associated with
the surplus P species added to produce the Ni2P phase. Consequently, supported Ni2P
catalysts can work in acid-catalyzed reactions if an adequate ratio of Brønsted to Lewis acid
sites is provided by the amount of the surplus P species determined by adjusting the P/Ni
ratio.
Keywords: Etherification; Furanyl alcohol; Aliphatic alcohol; Nickel phosphide; Acidity
1. Introduction
Since the first report of Sweeny et al. in 1958 that nickel phosphide (Ni2P) could catalyze
the reduction of nitrobenzene to aniline [1], Ni2P catalysts have been used for a great number
of hydroprocessing reactions. Their representative applications are hydrodenitrogenation
(HDN), hydrodesulfurization (HDS), and hydrodeoxygenation (HDO) that are essential in the
production of ultra clean motor fuels by removing heteroatom-containing molecules from
conventional fossil fuel and renewable biomass feedstock [2–4]. By virtue of unique
performance of Ni2P catalysts, extensive effort has been exerted to synthesize Ni2P
nanoparticles that are unsupported or often loaded onto various support materials [5].
Particularly, numerous studies have revealed a bifunctional character of supported Ni2P
affecting the product selectivity. The HDN and HDS reactions undergo hydrogenation on
metallic Ni site and protonation on acid site of Ni2P catalyst. Imporatnatly, the activity is
correlated with a density of the Brønsted acid site originating from the P–OH group present
2

in the surface of Ni2P [6–8]. The bifunctionality of Ni2P is more vivid in the HDO reaction of
biomass-derived chemicals because the removal of O is absolutely associated with acid site.
In the case of guaiacol that is a phenolic model compound for lignin, benzene was
preferentially formed over Ni2P/SiO2 at higher space times because the Brønsted acid site
adjacent to Ni site could facilitate the dehydroxylation of phenol by protonating the oxygen in
phenol [9,10]. Moreover, supported Ni2P catalysts enabled selective transformation of
cellulose into sorbitol: the acidity and metallic site were responsible for the hydrolysis of
cellulose to glucose and subsequent hydrogenation of glucose to sorbitol, respectively [11,12].
In recent years, the HDO reaction of furfural, which is the dehydration product from xylose,
was studied over Ni2P catalysts. Jiménez-Gómez et al. reported high selectivity towards 2-
methylfuran over Ni2P/SiO2, which is caused by the hydrogenation capacity of Ni site in the
conversion of furfural to furfuryl alcohol that is coupled with the hydrogenolysis function of
acid site in the subsequent transformation to 2-methylfuran [13]. Therefore, the acidity of
Ni2P catalysts clearly affects their HDO performance.
While understanding the catalytic actions of Ni2P, we realized that the performance was
determined only under H2-rich conditions in order to assess superior hydrotreating ability. If
it is true that Ni2P contains the acidity, then it should display an activity in acid-catalyzed
reactions taking place under inert atmosphere. Yet, this attempt has never been explored so
far as we know. In literature we found a report that the conversion of furfural to 2-
methylfuran over unsupported Ni2P involves a step of the condensation of furfuryl alcohol
into difurfuryl ether [14]. Although the authors employed HDO conditions for this
transformation, their results prompted us to certify the capability of Ni2P to catalyze an
acidity-demanding reaction. In this context, the target reaction is etherification of furfuryl
alcohols with aliphatic alcohols. Typically, 5-(hydroxymethyl)furfural (HMF), formed by
dehydration of hexose sugars [15,16], can be etherified to 5-(alkoxymethyl)furfural that can
be used as additive and precursor of drop-in fuel; for instance, the etherification product from
HMF and ethanol, that is, 5-(ethoxymethyl)furfural (EMF) has a comparable energy density
(8.7 kWh L–1) to gasoline (8.8 kWh L–1) and diesel (9.7 kWh L–1) [17].
The catalyst employed for the etherification reaction is SiO2-supported Ni2P (Ni2P/SiO2)
with the nominal P/Ni molar ratio of 2. Note that as the support acidity may initiate the
reaction [18], SiO2 is selected as a support due to no apparent activity. We have examined the
catalytic behavior of Ni2P/SiO2 under H2 for the conversion of HMF in the solvent of
3

tetrahydrofuran (THF) or ethanol. The etherification performance of Ni2P/SiO2 is confirmed
under inert N2 in the reaction of HMF with ethanol, and furthermore in the reactions
involving other furanyl alcohols and aliphatic alcohols. The characteristics of the prepared
Ni2P/SiO2 are investigated and compared with the samples prepared with different P/Ni ratios
of 0.5 and 1 in order to find out the origin of acidity. Finally, catalyst recyclability and
regeneration are examined for the etherification reaction.
2. Experimental
2.1. Catalyst preparation
The Ni2P/SiO2 catalyst with the nominal P/Ni ratio of 2, which is named Ni2P/SiO2-2P,
was prepared by temperature-programmed reduction (TPR). Typically, an aqueous solution
of Ni(NO3)2·6H2O (1.25 g; 97%, Daejung Chemical) was first precipitated by adding a
solution of (NH4)2HPO4 (1.14 g; 98%, Sigma-Aldrich). After dissolving the precipitate by 10
wt% nitric acid, a SiO2 support with the surface area of 350 m2 g–1 (2 g; Alfa Aesar) was
added, followed by aging for 2 h under vigorous stirring. The catalyst precursor was dried at
105 °C, calcined at 500 °C for 3 h, and finally reduced under H2 (99.9%) at a flow rate of 200
mL min−1 in a quartz tube according to the following program: heating from 25 °C to 350 °C
at a rate of 3.5 °C min−1, further to 700 °C at 1 °C min−1, and holding at 700 °C for 1 h. After
cooling to 25 °C under N2, the sample was finally obtained after passivation at 25 °C under
N2 gas flow for 1 h. In addition, Ni2P/SiO2-0.5P and Ni2P/SiO2-1P were prepared using the
above recipe except the nominal P/Ni ratio being 0.5 and 1, respectively.
For comparison, several reference catalysts were prepared. Bulk (or unsupported) Ni2P
was prepared by reduction of commercial Ni2P (Sigma-Aldrich; ~100 mesh) under a H2 flow
of 200 cm3 min−1 using the temperature program identical to that employed for Ni2P/SiO2. A
SiO2-supported Ni catalyst (Ni/SiO2) was prepared by wet impregnation, where the nominal
Ni loading was 10 wt% which is identical to that in Ni2P/SiO2. SiO2 (2 g) was added to an
aqueous solution of Ni(NO3)2·6H2O (1.10 g) in distilled water (30 mL). Then, the solution
was aged for 2 h under vigorous stirring, followed by water evaporation. The sample was
dried at 105 °C for 12 h and subsequently calcined at 500 °C for 3 h in an air flow of 100 mL
min−1. Finally, H2 reduction was carried out using the same procedure as that employed for
Ni2P/SiO2. On the other hand, a SiO2-supported phosphate catalyst (PxOy/SiO2) with the
nominal P loading of 10 wt% was prepared by the method analogous to Ni/SiO2 except the
use of (NH4)2HPO4 instead Ni source without a step of H2 reduction.
4

As ion-exchange resin catalyst, Amberlyst 15 (4.7 meq g−1) was kindly provided by Dow
Chemical in Korea, while Nafion NR50 (0.8 meq g−1) was purchased from Sigma-Aldrich.
2.2. Catalyst characterization
Powder X-ray diffraction (XRD) analysis was conducted using a Rigaku MiniFlex600
with a GADDS diffractometer using a Cu Kα radiation of 40 kV and 40 mA with a scan
speed of 10° min−1 and a step size of 0.02°. The crystallite size was calculated using the
Scherrer equation Dp = Kλ/βcosθ, where Dp is the particle size in a unit of Å, K is the shape
factor (0.94 in this study), λ is the wavelength of the X-rays, β is the full width at half
maximum, and θ is the Bragg angle. The transmission electron microscopy (TEM) images
were taken with a JEOL JEM 2100F microscope operating at 200 kV. The specific surface
area and pore volume were measured using a Micromeritics 3Flex instrument at −196 °C
with N2 after degassing at 105 °C for 1 h under vacuum. X-ray photoelectron spectroscopy
(XPS) was performed using a Thermo Fisher Scientific Messtechnik K-Alpha+ spectrometer
with an Al Kα radiation of 1486.6 eV. The actual elemental compositions were obtained by
an inductively coupled plasma atomic emission spectroscopy (ICP-AES) using a PerkinElmer
Optima 8300 spectrometer operated with an RF generator power of 13000 W, RF frequency
of 40 MHz, plasma gas flow of 12 L min−1, auxiliary gas flow of 0.2 L min−1, and nebulizer
gas flow of 0.55 L min−1. Prior to ICP-AES measurement, a sample of 0.03 g was mixed with
a mixture of nitric acid (4 mL), hydrofluoric acid (4 mL), and perchloric acid (1 mL) over a
hot plate, followed by dilution with water of 20 mL.
Pyridine-chemisorbed FT-IR analysis was performed to quantitatively measure the
Brønsted and Lewis acid sites. The probe molecule was adsorbed in a Specac HTHP cell
coupled with a Thermo Scientific Nicolet 6700 spectrometer with an MCT-A detector. The
self-supporting disc of a sample (20 mg) was treated at 300 °C for 1 h in a vacuum of 1×10−3
mbar and then pyridine of 5 μL was directly injected at 100 °C. After degassing at 100 °C for
30 min, IR spectra were recorded with a scan number of 32 and a resolution of 4 cm−1.
2.3. Activity test
The furanyl alcohols employed in this work are HMF, furfuryl alcohol (98%), and 5-
methylfurfuryl alcohol (97%) purchased from Sigma-Aldrich, TCI Chemical, and Acros
Organics, respectively. Also, a solution of 2,5-bis(hydroxymethyl)furan in ethanol was
prepared by the hydrogenation of HMF over Cu/Al2O3 catalyst at 70 °C for 3 h, resulting in a
5

purity of over 99% (confirmed by gas chromatography (GC) measurement). The alcohol
solvents are methanol (99.9%, J.T. Baker), ethanol (99.9%, Daejung Chemical), 1-propanol
(99%, Samchun Chemicals), 2-propanol (99.5%, Daejung Chemical), tert-butanol (99.5%,
Daejung Chemical), and 1-hexanol (98%, TCI Chemical).
For catalytic experiments, a stainless-steel batch reactor of 100 mL (Parr Instrument) was
charged with a catalyst (100 mg), furanyl substrate, and solvent (19 mL) in which the
substrate concentration was set at 0.21 M. The reactor was then purged with inert N2 at 25 °C
for 1 h and heated to a desired temperature (usually, 150 °C). The reaction started with
stirring at 600 rpm and was performed for 3 h. After the reaction was complete, the heating
and stirring were stopped and cooled to 25 °C. From the collected etherification product
mixture, a sample (1.5 mL) was taken and then mixed with 0.08 g of toluene (99.5%) that is
an external standard for GC analysis using an Agilent 7890A GC with a flame ionization
detector (FID) and a INNOWAX capillary column (30 m × 0.25 mm × 0.25 μm). While the
injector and detector were set at 250 and 300 °C, respectively, the oven temperature was
programmed as follows: 40 °C for 2 min, ramping to 160 °C at 20 °C min−1, holding at 160
°C for 2 min, followed by ramping to 250 °C at 10 °C min−1, and finally holding at 250 °C
for 5 min. The product was confirmed by GC-MS analysis (Agilent INNOWAX column). For
quantitative measurement, the experiments were repeated at least three times and the average
product compositions were reported in this context. Finally, the conversion of furanyl alcohol
and the selectivity and yield of each product were calculated as follows:
(
)
mole of furanyl alcohol converted
(
)
Conversion of furanyl alcohol mol% =
×100
(
)
mole of initial furanyl alcohol
(mole of the etherification product formed)
(mole of the total products formed)
(
)
Product selectivity mol% =
×100
(mole of the etherification product formed)
(mole of initial furanyl alcohol)
(
)
Product yield mol% =
×100
3. Results and discussion
3.1. Etherification activity of Ni2P/SiO2-2P catalyst
6

Scheme 1. Reaction pathway for conversion of HMF via hydrodeoxygenation (Pathway A)
and etherification (Pathway B).
We examined the catalytic action of Ni2P/SiO2-2P in THF at 50 bar H2 and 210 °C. As
displayed in Fig. 1a, the products formed are 5-methylfurfural (5-MF) and 2,5-dimethylfuran
(DMF) produced by the hydrodeoxygenation of HMF (Pathway A in Scheme 1) [13,19,20].
These HDO products were also found in the experiment at 240 °C showing the HMF
conversion of 74.5% that is higher than 47.7% at 210 °C (Fig. 1b). However, they were not
observed in the majority when the solvent was changed to ethanol. Instead, different GC
peaks (marked by the asterisk in Fig. 1a) were detected from the products obtained by the
tests at 210 and 240 °C. Based on GC-MS analysis of several authentic samples, the major
products could be assigned to EMF and its acetalization product, 5-(ethoxymethyl)furfural
diethyl acetal (EMFDEA), along with ethyl levulinate (EL) and ethyl formate (EF) in minor.
These originate from the etherification of HMF with the solvent ethanol [21–24], as depicted
in the Pathway B in Scheme 1. Quantitative results indicate that the conversion of HMF is
96.7% and 99.8% with the yield of all etherification products corresponding to 88.3% and
85.2% at 210 and 240 °C, respectively. The larger HMF consumption in ethanol than in THF
is caused by the preferred etherification reaction over Ni2P/SiO2-2P. Besides, intermolecular
dehydration of ethanol to diethyl ether (DEE) could occur.
7

Figure 1. Conversion of HMF over Ni2P/SiO2-2P. a Gas chromatograms of the product
mixtures obtained with a solvent of tetrahydrofuran (THF) and ethanol (EtOH) under the
reaction conditions: catalyst 100 mg, HMF 4 mmol, solvent 19 mL, 210 and 240 °C, 50 bar
H2, and 3 h. b HMF conversion and product yields by the HDO and etherification pathways,
calculated on the basis of the results in panel a. c Gas chromatograms of the product mixtures
obtained using ethanol at 1 bar N2 with different catalysts such as Ni2P/SiO2-2P (100 mg) at
150 and 180 °C, Amberlyst 15 (A-15, 0.6 mg) at 100 °C, and Nafion NR50 (NR50, 3.6 mg)
at 150 and 180 °C under the reaction conditions: HMF 4 mmol, ethanol 19 mL, and 3 h. d
HMF conversion and yields of etherification products (EMF, EMFDEA, and EL+EF),
calculated on the basis of the results in panel c.
Additional tests were carried out at low temperatures (e.g., below 200 °C) under N2 due to
the aforementioned reaction conditions suitable for the HDO reaction. The catalyst loading
used was determined based on the molar density of Brønsted acid site (BAS) relative to the
initial mole of HMF; the ratio of BAS to HMF was 0.07%. Figure 1c exhibits that Ni2P/SiO2-
2P catalyzes only the etherification reaction between HMF and ethanol at 150 and 180 °C.
The yield of EMF was higher at 180 °C owing to the nearly complete conversion of HMF
while the formation of EMF was more selective as 63.3% at 150 °C (Fig. 1d). When the
BAS/HMF ratio increased to 0.14% and 0.50%, the HMF conversion at 150 °C was enhanced
to 73.0% and 86.0%, respectively (Entries 6 and 7 in Table 1). However, the selectivity of
EMF was reduced to 56.9% and 48.0% at 0.14% and 0.50% BAS/HMF, respectively. This
suggests that the conversion of EMF into EL via hydration and further into EF via
esterification can occur over Ni2P/SiO2-2P under the employed reaction conditions.
8

For comparison, we tested the two well-known ion-exchange resins, Amberlyst 15 (4.7
meq H+ per gram) and Nafion NR50 (0.8 meq H+ per gram), as solid catalyst because the
etherification reaction proceeds by the protons tethered to the resin. Note that the relative
concentration of proton to HMF ([H+]/[HMF]), which is equivalent to the BAS/HMF ratio of
Ni2P/SiO2-2P, was set at 0.07% for activity tests. First, when Amberlyst 15 was tested under
N2 at 100 °C, the HMF conversion was measured to be 53.8% even though the employed
reaction temperature was as low as 100 °C due to low thermal stability (Table 1). However,
EL and EF were produced in larger amounts due to improved conversion of EMF over
Amberlyst 15 that has higher affinity to water [21]. On the other hand, Nafion NR50, which
is a perfluorinated acid resin and thermally stable compared to Amberlyst 15, was tested at
150 °C. The HMF conversion was pretty high as 80.5%, but the yield of EMF (34.2%) was
lower than that obtained with Ni2P/SiO2-2P along with the formation of DEE, because of the
well-known superacidity of Nafion NR50 [25]. The experiment at 180 °C confirmed the
stronger acidity of Nafion NR50: the conversion of HMF was 95.7% with the EMF yield of
51.5% and DEE yield of 0.13%. However, since the yields of EMF at 150 and 180 °C were
lower than those with Ni2P/SiO2-2P, it can be suggested that Ni2P/SiO2-2P is a little superior
to ion-exchange resin catalysts in terms of the selective formation of EMF. Also, Ni2P/SiO2-
2P appears to be comparable to the reported Zr-based catalysts [26,27], although the
employed reaction conditions are different (Table S1).
Table 1. Activity results in the etherification of HMF [a]
Entry Catalyst
Alcohol BAS/HMF Temperature HMF conversion Product selectivity (mol%)
(ROH) (mol%)
(°C)
150
180
100
150
180
150
150
(mol%)
71.2
96.5
53.8
80.5
95.7
73.0
86.0
EMF EMFDEA EL+EF
1
2
3
4
5
6
7
Ni2P/SiO2-2P EtOH 0.07
Ni2P/SiO2-2P EtOH 0.07
Amberlyst 15 EtOH 0.07
Nafion NR50 EtOH 0.07
Nafion NR50 EtOH 0.07
Ni2P/SiO2-2P EtOH 0.14
Ni2P/SiO2-2P EtOH 0.50
67.3 14.7
58.2 23.0
29.4 25.4
42.4 19.1
53.8 18.5
56.9 24.1
48.0 28.3
22.0
18.8
45.2
38.5
27.7
19.0
23.8
[a] Reaction conditions: HMF 4 mmol, alcohol 19 mL, 1 bar N2, and 3 h.
We further tried verifying the etherification capability of Ni2P/SiO2-2P by replacing either
HMF or ethanol with different alcohol substrates. The first approach is to react HMF with
different primary alcohols such as methanol, 1-propanol, and 1-hexanol. As the reaction
pathway in Fig. 2 represents, possible products are 5-(alkoxymethyl)furan (denoted as RMF,
where R is Me, 1-Pr, and 1-Hex representing methyl, propyl, and hexyl, respectively), its
acetalization product (RMFDRA), alkyl levulinate (RL), and alkyl formate (RF) [26]. In case
9

of the etherification with methanol at 150 °C and 0.07% BAS/HMF, the conversion of HMF
was 90.9% while the selectivity to MeMF plus MeMFDMeA was 58.1%. Among the
products, MeF was determined to be the major product (36.8% selectivity), meaning facile
transformation of MeMF into MeF via MeL due to the short methyl group. As the alkyl group
of alcohols was bulkier, the conversion of HMF decreased to 57.4% for 1-propanol and to
49.4% for 1-hexanol under the identical reaction conditions. However, the selectivity to the
etherification products corresponding to RMF and RMFDRA was 63.0% and 67.7% for 1-
propanol and 1-hexanol. Therefore, Ni2P/SiO2-2P can produce the etherification product in
the reaction of HMF with various primary alcohols.
Figure 2. Activities of Ni2P/SiO2-2P in the reaction of HMF with various aliphatic
alcohols. Reaction conditions: HMF 4 mmol, 0.07% BAS/HMF, alcohol 19 mL, 150 °C, 1
bar N2, and 3 h.
Furthermore, we tried using branched alcohols as the substrate for the etherification of
HMF over Ni2P/SiO2-2P (Fig. 2), where the used alcohols are 2-propanol (secondary alcohol)
and tert-butanol (tertiary alcohol). In case of the etherification of 2-propanol (R = 2-Pr) at
150 °C and 0.07% BAS/HMF, the conversion of HMF was 28.9% with the yield of RMF plus
RMFDRA at 20.8%. However, only (2-Pr)MFD(2-Pr)A was detected (Fig. S1), because the
formation of (2-Pr)MF is favorable at high reaction temperatures. Thus, the increase in
reaction temperature to 180 °C resulted in the enhancement in the HMF conversion (83.8%)
and (2-Pr)MF yield (26.0%). Additionally, Ni2P/SiO2 was tested in the reaction of HMF with
tert-butanol (R = t-Bu) at 150 °C. As a result, the HMF conversion was 24.6% and the yield
10

of RMFDRA was 13.4% with no formation of RMF. Surprisingly, isobutene was produced at
the yield of 3.6% (Fig. S1), indicating that Ni2P/SiO2 can catalyze the dehydration of tert-
butanol. The formation of isobutene was also observed using Nafion NR50. Therefore,
Ni2P/SiO2-2P possibly catalyzes the etherification of HMF with secondary and tertiary
alcohols, even if its performance is reduced compared to when primary alcohols are used.
Figure 3. Etherification of furanyl alcohols with ethanol over Ni2P/SiO2-2P. a
Conversion of furfuryl alcohol (FA) into furfuryl ethyl ether (FEE), 2-(diethoxymethyl)furan
(DOF), and EL. b Conversion of 5-methylfurfuryl alcohol (MFA) into 2-(ethoxymethyl)-5-
methylfuran (EMMF). c Conversion of 2,5-bis(hydroxymethyl)furan (BHMF) into 5-
(ethoxymethyl)furfuryl alcohol (EMFA) and 2,5-bis(ethoxymethyl)-furan (BEMFA).
Reaction conditions: furanyl alcohol 4 mmol, 0.07% BAS/HMF, ethanol 19 mL, 150 °C, 1
bar N2, and 3 h. To denote the products in the reaction scheme, the gas chromatograms of the
product mixtures and substrate are presented in the right-handed side.
The other approach we attempted is to replace HMF with different furanyl alcohols such
as furfuryl alcohol (FA), 5-methylfurfuryl alcohol (MFA), and 2,5-bis(hydroxymethyl)furan
(BHMF) for the etherification with ethanol. At 150 °C and 0.07% BAS/HMF, Ni2P/SiO2-2P
transformed FA into furfuryl ethyl ether (FEE) with the selectivity of 63.7% (Fig. 3a). When
a high reaction temperature of 180 °C was employed, the conversion of FA increased to 96.5%
while the selectivity to FEE was reduced to 55.7% accompanied by the increase in the
selectivity to EL (36.6%). The next example is MFA that can react with ethanol to produce 2-
(ethoxymethyl)-5-methylfuran (EMMF) [28]. After the etherification over Ni2P/SiO2-2P at
150 °C for 3 h, 78.0% of MFA was converted into EMMF with negligible side products (Fig.
11

3b). The final test was done for BHMF in which the hydroxyl groups on both sides of the
furan ring are changed step by step into ether groups; namely, the first etherification product
5-(ethoxymethyl)furfuryl alcohol (EMFA) followed by 2,5-bis(ethoxymethyl)furan (BEMFA)
[28]. Surprisingly, 99.7% of BHMF could be converted at 150 °C into EMFA (15.4% yield)
and BEMFA (80.3% yield) with several unknown products (Fig. 3c). The superior catalytic
performance in the etherification of BHMF is ascribed to the intrinsic ability of electron-
donating hydroxyl groups that are symmetrically placed in BHMF. Consequently, Ni2P/SiO2-
2P can catalyze the etherification of various furanyl alcohols with ethanol, where the reaction
efficiency depends on chemical vulnerability of furan structure.
3.2. The acidity of Ni2P/SiO2-2P catalyst
X-ray photoelectron spectroscopy (XPS) was performed to distinguish the oxidation states
of surface P and Ni species in the Ni2P/SiO2-2P catalyst (Fig. 4a). In the Ni 2p core level, the
peaks representing divalent Ni (Ni2+) at 857 and 873 eV are detected along with shakeup
satellite peaks [29,30]. Another peaks at 853 and 870 eV indicate the presence of Niδ+ species
(0 < δ < 1), which results from the electron density transfer from Ni to P atom in Ni2P
[13,29,31]. All these peaks were confirmed in the spectrum of bulk Ni2P. In the P 2p region
of 140 to 128 ppm, Ni2P/SiO2-2P exhibited two peaks at 135.1 and 129.6 eV. The former
peak can be assigned as pentavalent P (P5+) since PxOy/SiO2 exhibits only the peak at 135.4
eV. This P5+ species is associated with the surplus of the phosphorus added for the
preparation of Ni2P/SiO2-2P, which will be discussed later. The peak at 129.6 eV represents
the presence of partially reduced P species corresponding to Pδ−. Therefore, the presence of
both Niδ+ and Pδ− species in the Ni2P/SiO2-2P suggests the strong covalent bonding between
Ni and P atoms in the supported Ni2P particles.
The acidity of Ni2P/SiO2-2P was examined by pyridine-chemisorbed FT-IR using
PxOy/SiO2 and Ni/SiO2 as a reference material (Fig. 4b). Ni/SiO2 shows only the absorption
bands at 1448 and 1596 cm–1, indicating the presence of Lewis acid sites (LAS). In contrast,
only the bands corresponding to the BAS are clearly noticed in the sample PxOy/SiO2. The
LAS and BAS were quantitatively calculated using the band areas at 1448 and 1546 cm–1,
respectively [32]. As a result, the acid site density of Lewis acidic Ni/SiO2 is 14.1 µmol g–1
and that of Brønsted acidic PxOy/SiO2 is 62.4 µmol g–1. For Ni2P/SiO2-2P, the bands for the
BAS and LAS are all seen, which is consistent to the results of previous reports [6,33]. The
calculated density of BAS (28.0 µmol g–1) is over double that of LAS (11.4 µmol g–1).
12

Moreover, another footprint by pyridine chemisorption was found in the hydroxyl region of
4000 to 3400 cm–1. Ni2P/SiO2-2P shows the negative peaks at 3746 and 3661 cm–1
corresponding to the groups of Si–OH and P–OH, respectively. This is supported by the
findings on the spectra of Ni/SiO2 (only Si–OH) and PxOy/SiO2 (both Si–OH and P–OH).
Therefore, the group of P–OH in Ni2P/SiO2-2P can play the role of Brønsted acidity in the
etherification reaction.
Figure 4. Characterization results of Ni2P/SiO2-2P. a XP spectra in the regions of Ni 2p
and P 2p, compared to those of PxOy/SiO2 (for P 2p) and bulk Ni2P (for Ni 2p). b Pyridine-
chemisorbed FT-IR spectrum, compared to those of PxOy/SiO2 and Ni/SiO2.
The phase of Ni2P in the prepared Ni2P/SiO2-2P was confirmed by the standard reflections
of Ni2P (PDF #03-0953) and X-ray diffraction (XRD) peaks of unsupported (or called bulk)
Ni2P particles (Fig. 5a). Using the Ni2P (111) reflection at 2θ of 40.8°, the size of Ni2P
particles in Ni2P/SiO2 was calculated to be 23.1 nm that is much smaller compared to bulk
Ni2P (49.2 nm). The major planes of Ni2P phase such as (111) and (201) were identified in the
selected area electron diffraction (SAED) patterns of TEM image (Fig. 5b). This is the reason
why the Ni2P/SiO2-2P with the nominal P/Ni molar ratio of 2 was prepared in this work.
However, elemental analysis revealed that the actual P/Ni molar ratio was 1 (11.0 wt% for Ni
and 5.8 wt% for P), in which the difference between the nominal and actual values is caused
by loss of volatile phosphorus in the TPR process to form the Ni2P phase. This explains that
the surplus phosphorus species can exist on the surface of Ni2P/SiO2-2P, when considering
the theoretical P/Ni ratio of 0.5 in Ni2P.
To understand the acidity of Ni2P/SiO2-2P, we prepared Ni2P/SiO2-0.5P and Ni2P/SiO2-1P
of which the actual P/Ni molar ratio was measured to be 0.5 and 0.9, respectively (Table 2).
The XRD patterns of these samples show that the peaks corresponding to Ni12P5 phase are
more intense with the decrease in the P/Ni ratio along with the peaks of Ni2P phase (Fig. 5a).
13

This is consistent to the previous reports in which Ni12P5 is transformed into Ni2P as the P/Ni
ratio increases [8,34]. Therefore, when the P/Ni molar ratio is close to the theoretical value of
Ni2P, the major phase of supported nickel phosphide is Ni12P5 and not Ni2P.
Figure 5. Characterization results of Ni2P/SiO2-xP (x = 0.5, 1, and 2). a XRD patterns. b
TEM image of Ni2P/SiO2-2P in the scale bar of 5 nm and SAED patterns of Ni2P phase (1–3)
and amorphous SiO2 (4). c XP spectra in the region of P 2p, compared to those of PxOy/SiO2.
d Pyridine-chemisorbed FT-IR spectra. For comparison, the results of Ni2P/SiO2-2P are
reproduced herein.
Table 2. Compositions of Ni and P in Ni2P/SiO2-xP, Ni/SiO2, and PxOy/SiO2
14

Catalyst
Ni/SiO2
Ni2P/SiO2-0.5P 12.1
Ni2P/SiO2-1P 10.6
Ni2P/SiO2-2P 11.0
Actual composition Actual
Surface composition Surface Area percentage of
P/Ni ratio P species [c]
(wt%) [a]
Ni
P/Ni ratio (at%) [b]
P
Ni
P
P5+
-
Pδ–
-
10.3
-
-
0.27
0.51
0.52
0.56
-
-
-
3.1
5.2
5.8
9.3
0.5
0.9
1.0
-
0.22
1.14
1.84
6.95
0.4
2.2
3.3
-
19
58
88
100
81
42
12
0
PxOy/SiO2
-
[a] Measured by ICP-AES analysis. [b] Measured by XPS analysis. [c] Calculated by peak area in Fig. 5c.
XP spectra in the region of P 2p were compared to notice the difference in the
environment of phosphorus (Fig. 5c). As the P/Ni ratio decreases from 2 to 0.5, the peak of
P5+ becomes weak and is gradually shifted to 134.3 eV, meaning the disappearance of
phosphate group. On the contrary, the peak of Pδ– grows up with the P/Ni ratio decreasing:
the area percentage of Pδ– increases from 12% for Ni2P/SiO2-2P to 81% for Ni2P/SiO2-0.5P
(Table 2). This possibly suggests that 81% of the P species present in Ni2P/SiO2-0.5P takes
parts in the formation of Ni–P bond whereas 88% of the P species in Ni2P/SiO2-2P exists in
the form of phosphate group. In other words, the surplus phosphorus is required to produce
the phase of Ni2P in Ni2P/SiO2-2P but the P species left after forming the Ni–P bond is used
for the phosphate group.
The surface compositions measured by XPS are listed in Table 2. For Ni2P/SiO2-xP, the
surface composition of Ni is similar at around 0.5 at% whereas that of P increases from 0.22
to 1.84 at% with the x value increasing. Hence, the surface P/Ni molar ratio increases in the
following order: 0.4 (Ni2P/SiO2-0.5P) < 2.2 (Ni2P/SiO2-1P) < 3.3 (Ni2P/SiO2-2P). More
importantly, the surface P/Ni molar ratios in the latter two catalysts are higher than the actual
values. This implies that the surplus P species is more located on the surface for Ni2P/SiO2-1P
and Ni2P/SiO2-2P. According to the literature [7,35,36], the surplus P species, including the
4−
−
unreduced phosphate (such as HnPO4(3−n)−, P2O7 , and (PO3 )n) and the element P (such as P2
and P4), may locate on the surfaces of Ni2P particles and/or SiO2 support. These species as
well as the P species on the Ni2P particle surface can be combined with the water formed
during the TPR process employed for catalyst preparation, leading to the formation of the P–
OH group [35–37].
The densities of the BAS and LAS in Ni2P/SiO2-xP and PxOy/SiO2 were measured by
pyridine-chemisorbed FT-IR analysis (Fig. 5d and Table 3). The density of BAS increases in
the following order: 0.0 (Ni/SiO2) < 1.9 (Ni2P/SiO2-0.5P) < 22.3 (Ni2P/SiO2-1P) < 28.0
(Ni2P/SiO2-2P) < 62.4 μmol g–1 (PxOy/SiO2). The BAS densities of Ni2P/SiO2-1P and
15

Ni2P/SiO2-2P can be explained by the surplus P species. In contrast, Ni2P/SiO2-0.5P contains
the very little density of BAS compared to that of LAS (14.0 μmol g–1), while it shows the
very small peak of P–OH group ascribed to 3661 cm–1. This suggests that as the actual P/Ni
ratio is close to the theoretical value, an ideal Ni2P/SiO2 would be Lewis acidic though a very
little BAS can still exist on the surface of Ni2P particles. However, it should be noted here
that Ni2P/SiO2-0.5P exhibits the phases of both Ni2P and Ni12P5 with the latter being in major
(according to its XRD pattern). In this respect, the higher LAS density in Ni2P/SiO2-1P
compared to Ni2P/SiO2-0.5P can be rationalized by the fact that the latter sample has
abundant Ni2P phase. Because of this feature, the surplus P species added for Ni2P/SiO2-1P
may exist in the form of the unreduced phosphate (i.e., the P–OH group corresponding to the
BAS) that mainly locates on the surface of SiO2 support. This explanation is also reasonable
for the higher BAS density of Ni2P/SiO2-2P. Furthermore, the surplus P species would be
located as the form of the element P on the surface of Ni2P particles, as suggested in Chen et
al. [35]. This is corroborated by the considerably reduced density of LAS is in Ni2P/SiO2-2P.
Besides, the molar ratio of BAS to LAS increases with the P/Ni ratio in the order: 0.14
(Ni2P/SiO2-0.5P) < 0.83 (Ni2P/SiO2-1P) < 2.46 (Ni2P/SiO2-2P). Such a ratio is consistent to
the fraction of surface P5+ species. Consequently, the P–OH group could be located on the
surface of Ni2P particles and SiO2 support, as suggested in the previous report [35].
Table 3. Activity results in the etherification of HMF with ethanol over Ni/SiO2,
PxOy/SiO2, and Ni2P/SiO2-xP [a] and catalyst acidity
Catalyst
HMF conversion EMF yield Product selectivity (mol%) Acidity (μmol g–1) [e]
(mol%)
25.6
(mol%)
5.4
17.7
38.0
46.4
54.0
EMF EMFDEA EL+EF BAS
LAS
14.1
14.0
26.8
11.4
–
[b]
Ni/SiO2
20.9 10.8
41.1 21.2
67.3 12.2
67.3 11.7
72.5 8.1
68.3
37.7
16.7
21.0
19.4
–
1.9
22.3
28.0
62.4
Ni2P/SiO2-0.5P [c] 43.0
Ni2P/SiO2-1P [c] 56.5
Ni2P/SiO2-2P [c,d] 68.9
[d]
PxOy/SiO2
74.5
[a] Reaction conditions: HMF 4 mmol, ethanol 19 mL, 150 °C, 1 bar N2, 600 rpm, and 3 h. [b] 0.07 mol%
LAS/HMF. [c] Catalyst 0.1 g. [d] 0.07 mol% BAS/HMF. [e] Calculated by the spectra shown in Fig. 5d.
Ni2P/SiO2-xP, Ni/SiO2, and PxOy/SiO2 were tested in the etherification of HMF with
ethanol. Table 3 lists the detailed activity results and Fig. 6 presents the HMF conversion and
EMF yield with the densities of the BAS and LAS. The blank test at 150 °C showed the HMF
conversion of ca. 12% with the EMF yield of 2.6%. In the case of Ni/SiO2 possessing only
the LAS, the HMF conversion and EMF yield are estimated to be 25.6% and 5.4%,
respectively. In contrast, Brønsted acidic PxOy/SiO2 presents the high HMF conversion
(74.5%) and EMF yield (54.0%), which is similar to the activity result (HMF conversion =
16

77.4% and EMF yield = 47.3%) of SiO2-supprted nickel phosphate corresponding to the
precursor of Ni2P/SiO2-2P. The HMF conversion of Ni2P/SiO2-xP increases with the P/Ni
ratio in the following order: 43.0% (Ni2P/SiO2-0.5P) < 56.5% (Ni2P/SiO2-1P) < 68.9%
(Ni2P/SiO2-2P).
First, we suggest that the HMF etherification can proceed not only over the BAS but also
over the LAS, because Ni/SiO2 and Ni2P/SiO2-0.5P show a little activity even if the BAS is
absent in the former and very low in the latter. Next, the contribution of the BAS to the
activity is more significant than that of the LAS, when considering the catalytic performance
of PxOy/SiO2 and Ni2P/SiO2-2P. The last finding is associated with the result that the HMF
conversion and EMF yield are relatively similar with PxOy/SiO2 and Ni2P/SiO2-2P though the
BAS density of the former is 2.2-fold higher than that of the latter. Lee and Oyama reported
that the P–OH group ascribed to the BAS was stronger for PxOy/SiO2 compared with
Ni2P/SiO2 [6]. Li et al. found from NH3-TPD profiles that the Lewis acidity on Ni2P/SiO2
was stronger than the Brønsted acidity on PxOy/SiO2 [38] and bulk Ni2P [39]. Form these
reports, we can presume that the comparable activity between PxOy/SiO2 and Ni2P/SiO2-2P is
attributed to the stronger Lewis acidity of Ni2P/SiO2-2P. Consequently, it is believed that both
Brønsted and Lewis acidities of Ni2P/SiO2-2P are responsible for the etherification activity.
Figure 6. Relation between the etherification activity and acidity of Ni/SiO2, PxOy/SiO2,
and Ni2P/SiO2-xP.
3.3. Recyclability of Ni2P/SiO2-2P catalyst in the etherification of HMF with ethanol
The recyclability of Ni2P/SiO2-2P was confirmed by repeated use in the etherification of
HMF with ethanol, where the reaction temperature of 180 °C was selected so as to clearly
identify the transformation of the active Ni2P phase. In Fig. 7a the catalytic first-run activity
17

of Ni2P/SiO2-2P is displayed as a single bar representing the yields of all the formed products
(using the same values given in Fig. 1d). The spent Ni2P/SiO2-2P was obtained after washing
with the ethanol solvent without post heat-treatment. As a result, the HMF conversion
decreased a little to 89.2% and 86.8% in the second and third runs, respectively. Meanwhile,
the EMF yield was reduced to 57.5% and 52.4% in the recycle runs. Since the observed
activity loss may be related with the transformation of Ni2P phase, the spent Ni2P/SiO2
obtained after the first run was regenerated by H2 reduction at 700 °C for 1 h. As a result,
both the HMF conversion and EMF yield could be recovered to 95.7% and 63.3%,
respectively, which are comparable to the performance of the fresh Ni2P/SiO2-2P.
The spent Ni2P/SiO2-2P catalysts were characterized by the aforementioned techniques.
Their XRD patterns show that the intensities of Ni2P reflections in the spent samples are not
changed considerably (Fig. 7b). This suggests that the Ni2P phase in the bulk is possibly
preserved in the course of the etherification reaction. Nonetheless, it should be noted that the
peak at 2θ of 26.6° corresponding to the elemental carbon (PDF #03-0953) was detected,
which indicates the formation of coke species during the reaction. This coke formation can
result from the high reactivity of HMF at high temperatures producing humin [40].
We further characterized the spent and regenerated Ni2P/SiO2-2P samples (obtained after
the first activity run and after H2 reduction of the spent catalyst at 700 °C, respectively) to
compare with the fresh one. The XP spectrum of the spent Ni2P/SiO2-2P shows that in the
region of Ni 2p, the peaks corresponding to Niδ+ species are reduced and, in turn, the peaks of
Ni2+ are enlarged, while the peak of Pδ− species also decreases in the P 2p region (Fig. 7c).
This suggests that the oxidation of Ni2P particles during the reaction. Since the instability of
the Ni2P phase in high-temperature water was reported previously [12], another spent catalyst
sample was prepared by exposing Ni2P/SiO2-2P into pure ethanol under the reaction
conditions. The XP spectrum of this sample exhibits that the species of Niδ+ and Pδ− are
maintained (Fig. S2). Therefore, water molecules produced by the etherification reaction can
induce the oxidation of Ni2P surface and formation of P–OH group, assuming the increase in
the Brønsted acidity. Nevertheless, the BAS density in the spent Ni2P/SiO2-2P decreased to
18.0 µmol g–1, although the compositions of Ni and P in the spent catalysts were not largely
different from those in the fresh one (Table S2). This indicates that the formed coke
(examined by XRD analysis) might be responsible for the activity loss in repeated runs.
However, the characteristics of the regenerated Ni2P/SiO2-2P were similar to those of the
18

fresh one even though a little discrepancy was observed in the XPS results. This suggests that
the regeneration we employed can fully recover the catalytic performance of Ni2P/SiO2-2P.
Figure 7. Recyclability of Ni2P/SiO2-2P catalyst in the etherification of HMF with
ethanol. a Activity results over the fresh catalyst (Ni2P/SiO2-F), the spent catalysts
(Ni2P/SiO2-S1 and Ni2P/SiO2-S2 obtained after the first and second reaction runs,
respectively), and the regenerated catalyst (Ni2P/SiO2-R obtained after H2 reduction of
Ni2P/SiO2-S1 at 700 °C). Reaction conditions: catalyst 100 mg, HMF 4 mmol, ethanol 19
mL, 180 °C, 1 bar N2, and 3 h. b XRD patterns of Ni2P/SiO2-F, Ni2P/SiO2-S1, Ni2P/SiO2-S2,
Ni2P/SiO2-S3 (obtained after the third reaction run), and Ni2P/SiO2-R. c XP spectra of
Ni2P/SiO2-F, Ni2P/SiO2-S1, and Ni2P/SiO2-R in the regions of Ni 2p and P 2p.
19

4. Conclusion
We herein demonstrated the capability of a SiO2-supported Ni2P catalyst in the
etherification of furanyl alcohols with aliphatic alcohols under inert atmosphere. All the
activity tests disclosed that the Ni2P/SiO2 prepared with the nominal P/Ni ratio of 2 (namely,
Ni2P/SiO2-2P) could produce the etherification products (e.g., EMF and EMFDEA in the
reaction of HMF with ethanol) at a fairly good yield. The examined etherification activity
was attributed to both Brønsted and Lewis acidity of Ni2P/SiO2-2P that was much better than
the Lewis acidic Ni/SiO2 and comparable to the Brønsted acidic PxOy/SiO2. In Ni2P/SiO2-2P,
the Lewis acid sites was related with Ni species while the Brønsted acid site ascribed to the
P–OH group could be located on the surface of Ni2P particles and SiO2 support. The catalyst
recyclability tests revealed the activity loss due to coke formation but the catalytic
performance was fully recovered after regeneration by high-temperature H2 reduction.
Consequently, we suggest that Ni2P/SiO2 can catalyze the etherification reaction demanding
the acidity only if the actual P/Ni ratio is above the theoretical value so that the Brønsted acid
site is sufficiently created. This study will greatly contribute to understanding the acidity of
supported Ni2P catalysts examined in hydrotreating reactions.
Credit author statement
Mi Shin: Methodology, Validation, Formal Analysis, Investigation, Data Curation, Writing –
Original Draft, Visualization. Jinsung Kim: Formal Analysis, Investigation. Young-Woong
Suh: Conceptualization, Supervision, Project Administration, Writing – Review and Editing,
Funding Acquisition.
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.
Acknowledgements
This work was financially supported by the Next Generation Carbon Upcycling Program
20

through the National Research Foundation of Korea (NRF) funded by the Ministry of Science
and ICT, Republic of Korea (2017M1A2A2043148). We acknowledge technical supports
from the Hanyang LINC+ Analytical Equipment Center in Seoul.
Notes and references
[1] N. P. Sweeny, C. S. Rohrer, O. W. Brown, J. Am. Chem. Soc. 80 (1958) 799–800.
[2] J. L. Wagner, E. Jones, A. Sartbaeva, S. A. Davis, L. Torrente-Murciano, C. J. Chuck, V. P.
Ting, Dalton Trans. 47 (2018) 1189–1201.
[3] M. V. Landau, M. Herskowitz, T. Hoffman, D. Fuks, E. Liverts, D. Vingurt, N. Froumin, Ind.
Eng. Chem. Res. 48 (2009) 5239–5249.
[4] J. A. Cecilia, A. Infantes-Molina, E. Rodríguez-Castellón, A. Jiménez-López, S. T. Oyama,
Appl. Catal. B 136–137 (2013) 140–149.
[5] M. E. Bussell, React. Chem. Eng. 2 (2017) 628–635.
[6] Y.-K. Lee, S. T. Oyama, J. Catal. 239 (2006) 376–389.
[7] S. T. Oyama, Y.-K. Lee, J. Catal. 258 (2008) 393–400.
[8] S. T. Oyama, X. Wang, Y.-K. Lee, K. Bando, F. G. Requejo, J. Catal. 210 (2002) 207–217.
[9] S. Wu, P. Lai, Y. Lin, H. Wan, H. Lee, Y. Chang, ACS Sustainable Chem. Eng. 1 (2013)
349−358.
[10] X. Lan, E. J. M. Hensen, Th. Weber, Appl. Catal. A 550 (2018) 57–66.
[11] L. Ding, A. Wang, M. Zheng, T. Zhang, ChemSusChem 3 (2010) 818–821.
[12] P. Yang, H. Kobayashi, K. Hara, A. Fukuoka, ChemSusChem 5 (2012) 920–926.
[13] C. P. Jiménez-Gómez, J. A. Cecilia, R. Moreno-Tost, P. Maireles-Torres, ChemCatChem 9
(2017) 2881–2889.
[14] Y. Wang, F. Liu, H. Han, L. Xiao, W. Wu, ChemistrySelect 3 (2018) 7926–7933.
[15] J. Jeong, C. A. Antonyraj, S. Shin, S. Kim, B. Kim, K.-Y. Lee, J. K. Cho, J. Ind. Eng. Chem. 19
(2013) 1106–1111.
[16] I. K. Yu, D. C. W. Tsang, Bioresource Technol. 238 (2017) 716–732.
[17] S. Alipour, H. Omidvarborna, D.-S. Kim, Renew. Sust. Energ. Rev. 71 (2017) 908–926.
[18] H. Shi, J. Chen, Y. Yang, S. Tian, Fuel Process. Technol. 118 (2014) 161–170.
[19] Y.-B. Huang, M.-Y. Chen, L. Yan, Q.-X. Guo, Y. Fu, ChemSusChem 7 (2014) 1068–1070.
[20] D. P. Duarte, R. Martínez, L. J. Hoyos, Ind. Eng. Chem. Res. 55 (2016) 54–63.
[21] M. Balakrishnan, E. R. Sacia, A. T. Bell, Green Chem. 14 (2012) 1626–1634.
[22] H. Wang, T. Deng, Y. Wang, X. Cui, Y. Qi, X. Mu, X. Hou, Y. Zhu, Green Chem. 15 (2013)
2379–2383.
[23] M. M. Antunes, P. A. Russo, P. V. Wiper, J. M. Veiga, M. Pillinger, L. Mafra, D. V. Evtuguin,
N. Pinna, A. A. Valente, ChemSusChem 7 (2014) 804–812.
[24] M. Musolino, M. J. Ginés-Molina, R. Moreno-Tost, F. Arico, ACS Sustainable Chem. Eng. 7
(2019) 10221–10226.
[25] S. Pariente, N. Tanchoux, F. Fajula, Green Chem. 11 (2008) 1256–1261.
21

[26] P. Lanzafame, D. M. Temi, S. Perathoner, G. Centi, A. Macario, A. Aloise, G. Giordano, Catal.
Today 175 (2011) 435–441.
[27] M. M. Antunes, S. Lima, P. Neves, A. L. Magalhães, E. Fazio, F. Neri, M. T. Pereira, A. F.
Silva, C. M. Silva, S. M. Rocha, M. Pillinger, A. Urakawa, A. A. Valente, Appl. Catal. B 182
(2016) 485–503.
[28] E. R. Sacia, M. Balakrishnan, A. T. Bell, J. Catal. 313 (2014) 70–79.
[29] P. Bui, J. A. Cecilia, S. T. Oyama, A. Takagaki, A. Infantes-Molina, H. Zhao, E. Rodríguez-
Castellón, A. J. López, J. Catal. 294 (2012) 184–198.
[30] L. Gua, Y. Zhao, Z. Yao, Dalton Trans. 45 (2016) 1225–1232.
[31] S. J. Sawhill, D. C. Phillips, M. E. Bussell, J. Catal., 215 (2003) 208–219.
[32] C. A. Emeis, J. Catal. 141 (1993) 347–354.
[33] S. T. Oyama, T. Gott, K. Asakura, S. Takakusagi, K. Miyazaki, Y. Koike, K. K. Bando, J. Catal.
268 (2009) 209–222.
[34] V. O. O. Goncalves, P. M. Souza, T. de Cabioc’h, V. T. da Silva, F. B. Noronha, F. Richard,
Catal. Comm. 119 (2019) 33–38.
[35] J. Chen, T. Guo, K. Li, L. Sun, Catal. Sci. Technol. 5 (2015) 2670–2680.
[36] C. Stinner, Z. Tang, M. Haouas, Th. Weber, R. Prins, J. Catal. 208 (2002) 456–466.
[37] S. T. Oyama, T. Gott, H. Zhao, Y. Lee, Catal. Today 143 (2009) 94–107.
[38] K. Li, R. Wang, J. Chen, Energy. Fuels 25 (2011) 854–863.
[39] Y. Wang, X. Feng, S. Yang, L. Xiao, W. Wu, J. Nanopart. Res. 22 (2020) 67.
[40] M. Schlaf, Z.C. Zhang, Reaction pathways and mechanisms in thermocatalytic biomass
conversion I, first ed., Springer, Singapore, 2016.
22