Furfural production in a biphasic system using a carbonaceous solid acid catalyst

Article abstract of DOI:10.1016/j.apcata.2019.117180

The formation of furfural from xylose was investigated under heterogeneously catalyzed conditions with Starbon450-SO₃H as a catalyst in a biphasic system. Experiments were performed based on a statistical experimental design. The variables considered were time and temperature. Starbon450-SO₃H was characterized by scanning electron microscopy, N₂-physisorption, thermogravimetric analysis, diffuse reflectance infrared Fourier transform, Raman spectroscopy, pyridine titration and X-ray photoelectron spectroscopy. The results indicate that sulfonated Starbon450-SO₃H can be an effective solid acid catalyst for furfural formation. A maximum furfural yield and selectivity of 70 mol% was achieved at complete xylose conversion under optimum experimental conditions. The present paper suggests that functionalized Starbon450-SO₃H can be employed as an efficient solid acid catalyst that has significant hydrothermal stability and can be reused for several cycles to produce furfural from xylose.

Full text of DOI:10.1016/j.apcata.2019.117180

Accepted Manuscript
Title: Furfural production in a biphasic system using a
carbonaceous solid acid catalyst
Authors: Gerardo G o´ mez Mill a´ n, Josphat Phir, Mikko M a¨ kel a¨ ,
Thad Maloney, Alina M. Balu, Antonio Pineda, Jordi Llorca,
Herbert Sixta
PII:
DOI:
Article Number:
S0926-860X(19)30335-7
https://doi.org/10.1016/j.apcata.2019.117180
117180
Reference:
APCATA 117180
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
21 March 2019
23 July 2019
26 July 2019
Please cite this article as: Mill a´ n GG, Phir J, M a¨ kel a¨ M, Maloney T, Balu
AM, Pineda A, Llorca J, Sixta H, Furfural production in a biphasic system
using a carbonaceous solid acid catalyst, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.117180
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.

Furfural production in a biphasic system using a
carbonaceous solid acid catalyst
[
a,b]
[a]
[a,c]
[a]
[d]
Gerardo Gómez Millán , Josphat Phiri , Mikko Mäkelä , Thad Maloney , Alina M. Balu ,
[d]
[b]
[a]
Antonio Pineda , Jordi Llorca , Herbert Sixta *
G. Gómez Millán, J. Phiri, DSc. M. Mäkelä, Prof. Thad Maloney, Prof. H. Sixta
Department of Bioproducts and Biosystems School of Chemical Engineering, Aalto University Vuorimiehentie 1, 02150 Espoo (Finland)
E-mail: herbert.sixta@aalto.fi
G. Gómez Millán, Prof. J. Llorca
Department of Chemical Engineering, Institute of Energy Technologies and Barcelona Research Center in Multiscale Science and Engineering
Universitat Politècnica de CatalunyaEduard Maristany 10-14, 08019 Barcelona (Spain)
DSc. M. Mäkelä
Department of Forest Biomaterials and Technology Swedish University of Agricultural Sciences Skogsmarksgränd, 90183 Umeå (Sweden)
Prof. A. M. Balu, Dr. A. Pineda
Departmento de Química Orgánica Universidad de Cordoba Campus Rabanales, Edificio Marie Curie (C-3) Ctra Nnal IV-A, km 396 Cordoba
(Spain)
Graphical abstract
Highlights



Furfural formation using a carbonaceous solid acid catalyst in a biphasic system
including CPME was studied.
At complete xylose conversion, a high furfural yield (70%) at 175 °C in 18 h was
obtained in the biphasic system.
The carbonaceous solid acid catalyst demonstrated hydrothermal stability and
similar catalytic activity was reached after three reusability cycles.

Abstract
The formation of furfural from xylose was investigated under heterogeneously catalyzed
conditions with sulfonated Starbon®450-SO H as catalyst in a biphasic system. Experiments
3
were performed based on a statistical experimental design. The variables considered were time
and temperature. Starbon®450-SO H was characterised by scanning electron microscopy, N -
3
2
physisorption, thermogravimetric analysis, diffuse reflectance infrared Fourier transform, Raman
spectroscopy, pyridine titration and X-ray photoelectron spectroscopy. The results indicate that
sulfonated Starbon®450-SO H can be an effective solid acid catalyst for furfural formation. A
3
maximum furfural yield and selectivity of 70 mol% was achieved at complete xylose conversion
under optimum experimental conditions. The present paper suggests that functionalized
Starbon®450-SO H can be employed as an efficient solid acid catalyst that has significant
3
hydrothermal stability and can be reused for several cycles to produce furfural from xylose.
Keywords: xylose, furfural, carbonized starch, cyclopentyl methyl ether, heterogeneous
catalysis, biorefinery
1. Introduction
Furfural (FUR) has been highlighted as one of the top ten most rewarding bio-based building
blocks by the United States Department of Energy. FUR can be employed directly as a chemical
solvent and selective extractant, fungicide and as a component of disinfectors, rust removers
and pesticides [1, 2]. Furthermore, FUR withholds the potential to be further transformed directly
or indirectly into more than 80 valuable chemicals [2, 3]. FUR can also be hydrogenated to
furfuryl alcohol, which has applications in the biofuel and food industries (furfuryl alcohol
represents around 60% of the FUR market [4]). It can also be used in the manufacturing of

chemical resistant furanic resins. Other attractive chemicals that can be obtained from FUR are
levulinic acid, 2-methyltetrahydrofuran, furan and furoic acid [2, 5].
FUR is typically produced by dehydration of C -sugars (arabinose and xylose) contained in the
5
hemicellulose of lignocellulosic biomass. The production of FUR at industrial scale is associated
with high reaction temperatures (approximately 200 °C) and mineral acids (usually sulfuric and
hydrochloric acids) that have various process drawbacks, such as the production of toxic
effluents, equipment corrosion and consumption of high stripping-steam-to-FUR ratios.
Furthermore, the number of side reactions under these conditions limits FUR yields to
approximately 50% [6]. Recent research in this field has focused on increasing the FUR yield
with reusable solid acid catalysts to replace typically used homogenous acid catalyzed
conditions. A wide range of solid acid catalysts have been developed for this purpose to
produce FUR from xylose; such as zirconia [7-10], alumina [10, 11], zeolites [10, 12-17],
aluminosilicates supported with metals [18], modified silica [19-26], sulfonated graphenes [27],
heteropolyacids [8, 28, 29], coated activated carbon [30], and resins [22, 31, 32]. However, one
of the main challenges of heterogeneous catalysis is the hydrothermal stability of the solid
catalysts and the blocking of active sites by humins (insoluble polymeric products formed via
condensation and resinification of furanic compounds) [7, 16, 33].
Carbon-based catalysts offer high hydrothermal stability [34]. Carbon materials, such as
functionalized activated carbon [30], provide exciting opportunities for the catalytic conversion of
biomass into value-added chemicals due to their hydrothermal stability, their potential to be
produced from biomass and ability to be functionalized by various methods. Sairanen et al [30]
impregnated activated carbon with H SO , HNO and a combination of both in order to form
2
4
3
FUR from xylose in aqueous media. Even though xylose conversion is complete under the
reported experimental conditions, FUR yield is not shown in the paper and the reusability of the
catalyst is not mentioned. In a similar way, Termvidchakorn et al [35] functionalized multi-wall

carbon nanotubes with mineral and organic acids. They employed the functionalized catalysts to
form FUR from xylose and achieved the highest xylose conversion (95%) when adding Co
(
Co(NO ) ·6H O was used as precursor) in 3 h at 170 °C. Nevertheless, FUR yield was not
3
2
2
reported. Moreover, the carbon-based catalysts employed were not investigated for their
hydrothermal stability or reusability potential. Unlike previous carbon-based catalysts, Lam et al
[
27] developed a sulfonated graphene oxide that yielded 62% FUR in 35 min at 200 °C in water.
Nevertheless, the production of graphene oxide includes several steps and various chemicals.
Jalili et al [36] reported in their recent paper that graphene derivatives contain silicon, which has
a significant impact on their performance. Among these carbon-based catalysts, a mesoporous
material derived from renewable bio-resources (potato and corn starches) known as
Starbon®450-SO H has demonstrated superior selectivities and activities in various acid
3
catalyzed aqueous phase reactions, such as the esterification of succinic acid [37-40]. In
addition, Starbon®450-SO H functionalities, such as hydrophilicity, can be tuned which makes it
3
possible to dehydrate xylose in aqueous phase.
The aim of the present paper was to employ Starbon®450-SO H as a solid acid catalyst for the
3
dehydration of xylose to produce FUR. Cyclopentyl methyl ether (CPME) was added to the
aqueous xylose solution to extract formed FUR into the organic phase as part of a biphasic
system [41]. CPME has demonstrated to be an efficient green solvent, due to its lower toxicity in
comparison to other ethers, in the extraction of FUR [42, 43]. Furthermore, the hydrothermal
stability of the solid acid catalyst and its reusability were thoroughly investigated. Under the
experimental conditions provided in this paper, it is demonstrated that Starbon®40-SO H can
3
produce high FUR yields in a biphasic system and its catalytic activity remains similar after 3
reusability cycles. Besides, the characteristics of the solid catalyst were studied in detail by
SEM, EDX, N -physorption, Py-titration, TGA, DRIFT, Raman spectroscopy and X-ray
2
photoelectron spectroscopy.

2. Experimental
2
.1.Materials
D-Xylose powder (99%, Sigma Aldrich), CPME (99.9%, Sigma Aldrich), furfural (99%, Sigma
Aldrich), potato starch (Sigma Aldrich), sulfuric acid (49–51%, HPLC grade, Sigma Aldrich) were
used in the experiments without further purification. Formic acid (98%, Sigma Aldrich), levulinic
acid (99%, Sigma Aldrich) and acetic acid (99%, HPLC grade, Sigma Aldrich) were used for the
preparation of calibration standards for HPLC analysis. Iso-butanol (99.9%, Sigma Aldrich) was
used as internal standard (IS) for GC analysis. Millipore-grade water was used for preparing the
solutions.
2
.2.Methods
.2.1. Determination of FUR and by-products
2
From the biphasic system, samples for analysis were drawn from both the top (organic phase)
and the bottom part (aqueous phase) after hydrothermal reaction. Xylose, carboxylic acids
(formic, acetic and levulinic acids) and FUR from aqueous phase were analyzed separately by
High Performance Liquid Chromatography (HPLC) operating a Dionex UltiMate 3000 HPLC
(
Dionex, Sunnyvale, CA, USA) device equipped with refractive index (RI) and ultraviolet (UV)
+
diode array detectors. Product separation was achieved on a Rezex ROA-Organic Acid H (8%)
-1
LC column (7.8 mm × 300 mm, Phenomenex, USA). Aqueous sulfuric acid (0.0025 mol l ) was
-1
used as eluent with a flow rate of 0.5 ml min . A temperature of 55 °C was set for the column
temperature and the RI-detector. The FUR concentration was determined by the UV-detector at
a wavelength of 280 nm. Xylose concentration was simultaneously analyzed by the RI-detector
and the UV-detector at 210 nm [44]. The samples were filtered through a 0.45 m syringe filter
before the analysis. For calibration of the HPLC, a series of calibration standards was prepared
from the following chemicals: xylose, furfural, formic acid, acetic acid, levulinic acid. From a

parent standard solution (0.1 g diluted in 100 ml of Milli-Q water) calibration standards in four
concentrations (0.1 ml, 0.5 ml, 1 ml and 1.5 ml in 10 ml of Milli- Q water) were prepared.
FUR from the organic phase was analyzed by gas chromatography with a flame ionization
detector (GC-FID) relative to iso-butanol as internal standard (IS). The column used was a DB-
WAXetr (30 m, 0.32 mm i.d., 1 µm film thickness) from Agilent Technologies Inc. The volume of
injected samples was 0.5 µL and they were subjected to a splitless ratio of 20:1 in the inlet,
which was maintained at 250 °C and a pressure of 13 psi. Helium was used as the carrier gas.
The oven was initially maintained at 80 °C for 1 min, after which the temperature was increased
-1
to 250 °C at 30 °C min . The FID was operated at 250 °C with hydrogen, air, and helium
-1
-1
-1
delivered at 30 mL min , 380 ml min , and 29 ml min , respectively.
In this study conversion was defined in terms of moles of reactant converted per unit volume of
reactor (Eq. 1). Selectivity, at an instant, was the generated number of moles of desired product
referred to the moles of reactant converted (Eq. 2). Yield was calculated as the amount in moles
of desired product (FUR) produced related to the amount of xylose converted (Eq. 3) [45]. The
following equations were used for deriving these parameters:
ꢅꢆ
ꢈ
ꢄ
ꢇꢄ
ꢀ
ꢁꢂ ꢀꢁꢂ
ꢃ
ꢃ
ꢃ
ꢉꢊꢋꢋꢌ[ꢍ]
(Eq. 1),
(Eq. 2)
(Eq. 3),
ꢀ
ꢁꢂ
ꢅꢆ
ꢄ
ꢀ
ꢁꢂ
ꢄ
ꢄ
ꢈꢏꢐ
ꢎꢈꢏꢐ
ꢉꢊꢋꢋꢌ[ꢍ]
ꢅꢆ
ꢀ
ꢁꢂ
ꢈꢏꢐ
ꢄꢈꢏꢐ
ꢅꢆ
ꢑ
ꢉꢊꢋꢋꢌ[ꢍ]
ꢀꢁꢂ
ꢈ
ꢀꢁꢂ
ꢄ
ꢇꢄ
ꢀ
ꢁꢂ
where X, S, Y are the– conversion of xylose, selectivity to FUR and FUR yield, respectively; c is
-1
the– concentration in mmol l (the subscripts xyl, fur, in, f refer to xylose, FUR, initial, final).

Once the concentrations of FUR and xylose had been determined in each sample, individual
prediction models were built for xylose conversion, furfural yield and selectivity separately by
solving the general linear regression equation (Eq. 4) [46].
ꢒꢃꢓꢔꢕꢖ
(Eq. 4)
by minimizing the sum of squares of model residuals through the least-squares estimate (Eq. 5):
ꢘ
ꢚꢊ ꢘ
ꢔꢃꢗꢓ ꢓꢙ ꢓ ꢒ
(Eq. 5)
where y denoted a vector of response values, Z the mean-centered and coded design matrix
including interaction and second-order terms, b a vector of model coefficients and e the model
residuals. Statistically insignificant model terms (p > 0.10) were excluded based on an F-test
that compared the effects with the respective model residuals. The performance of the models
2
was expressed through the R value, which indicated the proportion of data variation explained
by each individual model.
2.2.2.
Catalyst preparation
Starbon®450-SO H catalyst was synthesized according to known literature procedure with
3
minor modifications [47]. First, the starting material (starch from potato, Sigma-Aldrich) was
heated up in water to 140 °C for 2 h (150 g starch in 3 L deionized water). Upon cooling the
warm solution was poured into a vial at room temperature, and it was further cooled down to 5
°
C for 48 h until formation of a porous gel in water. To avoid the structure to collapse while
drying, several solvent exchange steps were conducted until water was fully replaced by ethanol
5 times), and finally by acetone (2 times) to stabilize the porous network. The resulting
(
materials were then filtered off and dried overnight at 50 °C under vacuum, rendering the
mesoporous starch structure, subsequently calcined at 450 °C under inert atmosphere (N2,
-1
5
0 mL min ) by using the following heating conditions: from RT to 450 °C, heating rate 1 °C
-1
min ; temperature maintained for 1 h. A purge with nitrogen prior to carbonization was
conducted to ensure the absence of oxygen in the first steps of carbonization.

For sulfonation, the calcined Starbon®450 material was suspended in H SO of 95-97% purity
2
4
(
10 mL acid per gram of material and 4 h at 80 °C). After sulfonation, samples were thoroughly
washed with distilled water until neutral pH value, and finally oven dried at 100 °C overnight.
The resulting functionalized mesoporous acid material is denoted as Starbon®450-SO H
3
(
STARch carBONized at 450 °C with sulfonic acid groups).
.2.3. Catalyst characterization
2
Scanning electron microscopy (SEM) images were recorded at 5 kV using a JEOL JSM-7800F
PRIME Schottky Field Emission Scanning Electron Microscope equipped with a high resolution
Gentle Beam (GBSH). Samples were deposited on conductive carbon tabs. The instrument has
a field emission gun and it is also equipped with an energy dispersive X-ray (EDX) detector for
chemical analysis.
Thermogravimetric analysis (TGA) was carried out in a Setaram Setsys 12 using air as carrier
-1
gas (50 mL min ). The sample was loaded in ceramic crucibles with-Al O used as reference
2
3
-
1
compound and a Pt/Pt-Rh (10%) thermocouple. The heating rate employed was 10 K min in all
cases.
Infrared studies were done using Diffuse Reflectance Infrared Fourier Transform (DRIFT).
Spectra were recorded on an ABB BOMEM MB 3000 Instrument equipped with an
environmental chamber (Spectra Tech, P/N, 0030-100) placed in the diffuse reflectance
-1
attachment. The resolution was 8 cm and 256 scans were averaged to obtain the spectra in
-1
the 4000-400 cm range. Spectra were recorded by using KBr as a reference. The samples for
DRIFTS studies were prepared by mechanically grinding all reactants to a fine powder
(
sample/KBr 1:5.7 ratio).
A Micromeritics Tristar II-Physisorption Analyzer was utilized to record the nitrogen sorption
isotherms for fresh and spent catalysts. All samples were dried at 105 °C and exposed to

nitrogen gas for 12 h before measurement and the isotherms were taken at 196.15 °C. The
samples were exposed to ~20% humid room air for about 1 minute during the transfer to the
holders. The specific surface area (ABET) was determined by the Brunauer-Emmett-Teller (BET)
model [48] at relative pressures between 5 and 35% where the data points were observed to
arrange linearly. The specific pore volume (V ) was estimated from N uptake at a p/p value of
p
2
0
0.99 while recording approximately 150 equilibrium data points. The pore width distribution (dp)
was deduced from the desorption branch using the Barrett-Joyner-Halenda (BJH) method [49].
Xylose adsorption tests were done by stirring 3 mL of an aqueous solution of 186 mmol l
-1
3
-1
xylose using a borosilicate glass reactor (V = 10 cm ) with magnetic stirring (600 min ) and 50
mg of Starbon®450-SO H. Agitation of the suspension for 24 h at room temperature (25 °C).
3
Determination of xylose adsorption was performed by HPLC analysis.
Pyridine (PY) titration experiments were conducted similarly to the method found in the literature
with few modifications [50]. The experiments were performed at 200 °C via gas phase
adsorption of the basic probe molecules utilising a pulse chromatographic titration methodology.
Briefly, probe molecules (typically 2-5 μL) were injected into a gas chromatograph (GC) through
a microreactor in which the solid acid catalyst was previously placed. Basic compounds were
adsorbed to full saturation, from where the peaks of the probe molecules in the gas phase were
detected in the GC. The quantity of probe molecule adsorbed by the solid acid catalyst could
subsequently be easily quantified.
Raman spectra were measured using a WITec alpha300 R Raman microscope (alpha 300,
WITec, Ulm, Germany) equipped with a piezoelectric scanner using a 532 nm linear polarized
excitation laser. The measurement was conducted directly on the powder catalyst after washing
and drying.
The surface characterization was done with X-ray photoelectron spectroscopy (XPS) on a
SPECS system equipped with an Al anode XR50 source operating at 150 mW and a Phoibos

-
7
1
50 MCD-9 detector. The pressure in the analysis chamber was always below 10 Pa. The area
analyzed was about 2 mm × 2 mm. The pass energy of the hemispherical analyzer was set at
5 eV and the energy step was set at 0.1 eV.
Peak fitting and quantification analysis were performed using the software package CasaXPS
Casa Software Ltd., UK). Binding energy (BE) values for Starbon®450-SO H were referred to
2
(
3
the adventitious C 1s signal at 284.8 eV. Atomic surface ratios were obtained by using peak
areas normalized on the basis of acquisition parameters after background subtraction,
experimental sensitivity factors and transmission factors provided by the manufacturer.
2.2.4.
Catalytic activity tests
3
In a typical experiment, the glass reactor (V = 8 cm ) was loaded with 0.75 ml of an aqueous D-
-1
xylose solution in a concentration typical for biomass hydrolysate (186 mmol l ) [43], 2.25 ml of
-
1
CPME and 21 mg of the catalyst. The glass reactor included magnetic stirring (600 min )
(
Figure S1 in the Supplementary Information). The vials were heated up in a silicone oil bath
until the desired temperature was reached. Time zero was set when the vials were immersed
into the oil bath. Towards the reaction end time, vials were rapidly pulled up from the silicone oil
bath and cooled in an ice bath. The prepared solutions were used for determining FUR yield,
selectivity to FUR and xylose conversion in different reaction conditions. The individual
experiments were organized according to a central composite design on two variables. This
design enabled to determine the effects and statistical significance of the reaction conditions on
sample properties. The variables, reaction temperature and time, were varied on three different
levels. The design thus consisted of 11 experiments, including 3 replicated experiments at the
design center (175 °C in 12.5 h) (Table 2).

Results and Discussion
In order to know more about the unique characteristics of the carbonaceous acid catalyst,
various analytical techniques were employed on Starbon®450-SO H and how its characteristics
3
affect the production of FUR from xylose. The results of its analyses are described below.
2
.3.Scanning electron microscopy (SEM) and energy-dispersive X-ray (EDX)
For the purpose of demonstrating various properties related to surface topography and chemical
composition, scanning electron microscopy (SEM) and energy-dispersive X-ray (EDX) deliver
simple, non-destructive analyses. Fig. 1 corresponds to a representative image of Starbon®450-
SO H catalyst powder before hydrothermal reaction revealing the characteristic morphology of
3
particles with sharp edges similar to that reported in the literature [51]. Particles are compact
and their size is in the range of 50-100 μm.
a
b
100 µm
10 µm
Figure 1. SEM images of the Starbon®450-SO H catalyst powder obtained at different
3
magnifications, (a) 180X and (b) 650X.

EDX analyses showed that the catalyst had a very homogeneous composition. Table 1
compiles the mean values obtained in three different regions of the sample. They showed small
variations in the content of the elements detected (0.2%).
Table 1. Energy-dispersive X-ray spectroscopy (EDX) analysis of Starbon®450-SO H, textural
3
properties (i.e., BET (ABET), Pore volume (V ) and Pore diameter (d )) and acid properties of
p
p
Starbon®450-SO H.
3
Energy-dispersive X-ray
spectroscopy (EDX)
Textural properties
Xylose
Acid site
density
a
b
adsorption
-
1
Element Wt. %
Atomic %
2
-1
3
-1
(mmol g )
(mmol Py
ABET (m g ) V (cm g ) dp (nm)
-1
p
g )
C
O
S
68.9 ± 0.2
29.3 ± 0.2
1.8 ± 0.1
100.0
75.2 ± 0.2
24.0 ± 0.2
0.8 ± 0.0
100.0
264.5 ±
62.9
0.04 ±
0.004
3.5 ±
0.6
0
.32
0.29
Total:
a
-1
5
0 mg of Starbon®450-SO H in 3 mL of a xylose solution (186 mmol l ). Agitation of the
3
suspension for 24 h at room temperature (25 °C). Determination of xylose adsorption by HPLC
analysis. Adsorbed amount of xylose per gram of catalyst.
b
Pyridine (PY) titration value at 200 °C.
2
.4.N -physisorption – Py TPD
2
As published in literature, a large surface area is preferable due to a high amount of available
acid sites and to facilitate the accessibility of xylose [52]. Another important textural
characteristic of the catalyst, the pore width, affects the diffusivity. Fast diffusion in large pores
can prevent FUR decomposition, hence increased selectivity can be achieved [52]. Brunauer-
Emmett-Teller (BET) specific surface area, pore volume, xylose adsorption capacity and acid
site density of Starbon®450-SO H are compiled in Table 1. Starbon®450-SO H shows a BET of
3
3
2
-1
3
-1
2
20 m g , a pore volume of 0.4 cm g and a pore width of 3.9 nm; which are similar to
previously reported literature on similar materials [40]. The pore width distribution shown in Fig.
S2 (in the Supplementary Information) reveals a narrow pore width of approximately 10 nm.

Xylose adsorption was investigated to determine the availability of the reaction starting material
xylose at the surface. Adsorption of sugar solutions (xylose [30, 53] and fructose [54]) in
aqueous phase onto solid materials has been investigated by previous researchers. According
to Sairanen et al, [30] if the amount of xylose adsorbed on the catalyst surface is higher than the
amount of acid sites, which indicates that xylose is adsorbed at sites other than the Brönsted
sites as determined in our studies by pyridine titration. Brönsted acid sites are associated with
direct dehydration of xylose into FUR, while Lewis acid sites are known to shift the equilibrium
towards the isomers (especially to xylulose) [10, 55, 56]. Our results showed that the
concentration of xylose adsorbed on the surface was slightly higher than the catalyst’s acid site
density, suggesting that acid sites other than Brönsted are present on the surface of the
catalyst. Nevertheless, even if Lewis acid sites were present on the surface of Starbon®450-
SO H, they did not play a significant role in the isomerization of xylose into xylulose, since
3
xylulose was not detected by HPLC. Other literature reported materials including mesoporous
-1
silica SBA-15 have been reported to have lower acid site densities (e.g. 0.16 mmol g ) [57].
.5.Infrared spectroscopy (IR) and thermal gravimetric analysis (TGA)
The infrared (IR) spectrum of Starbon®450-SO H is shown in Figure 2. It exhibits two maxima
2
3
-1
-1
at 1735 cm and 1612 cm that could be assigned to asymmetric stretching vibrations of –
-
COOH carboxyl and –COO carbonyl and/or –CꢃO ketone units and to the stretching vibrations
-1
of CꢃC bonds in aromatic carbon rings, respectively [58]. There is also a peak at 1227 cm ,
which can be assigned to asymmetric stretch of –C-C-C bridges in ketonic groups and/or to
deformation vibrations of O-H vibration in the carboxylic acid and alcoholic groups from sugars
present in the starch-based material, which have been formed during the carbonization of the
-1
starch precursor [59]. There is also a small peak at 1057 cm , which could be assigned to
symmetric SꢃO stretching of sulfonic groups attached to the material [59-63]. Bands observed
-1
in low region at 865 cm could be related to symmetrical C-O-C stretching [59]. In the low

-1
frequency range, the line at 679 cm indicates the SꢃO stretching mode of –SO H [59, 64]. The
3
-1
-1
peak at 594 cm is assigned to the S-O stretching mode and 532 cm was assigned to the C-S
stretching mode, suggesting the existence of covalent sulfonic acid groups [64, 65].
Starbon®450 before sulfonation was also analysed with IR (Fig. S3 in the Supplementary
Information). Significant changes can be seen after Starbon®450 was sulfonated. Mena [59]
reported Starbon® carbonized at 300 °C. However, the parent Starbon® displays bands
corresponding to residues of p-toluenesulfonic acid, which she used in its synthesis.
In order to study the thermal stability and the decomposition rate of Starbon®450-SO H,
3
DTA/TGA curves were recorded (Figure S4 in the Supplementary Information). The DTA curve
for Starbon®450-SO H displays two endothermic peaks at 100 °C and 540 °C due to moisture in
3
the sample and to the combustion of the carbon matrix, respectively [41, 59]. The TGA curve of
Starbon®450-SO H showed a steep weight loss of nearly 70% between 350 °C and 600 °C,
3
which could be associated to the combustion of the carbonaceous material.
1057
1735
1612
679
1227
865
5
5
94
32
CO2
2500
2000
1500
1000
500
-
1
Wavenumber (cm )

Figure 2. Infrared spectra of Starbon®450-SO H
3
2
.6.Catalytic activity tests in monophasic system
Xylose dehydration into FUR in aqueous phase leads to rapid decomposition of FUR and
provides low product yields [7]. As Tables S1, S2 and Figure S5 displays (in the Supplementary
-
1
Information) the auto-catalyzed system of xylose dehydration (3 ml of a 186 mmol l xylose
solution) at 170 °C in various reaction times (1 – 6 h), the highest FUR yield was 38% at a
xylose conversion of 58% after 6 h (Fig. S5a). A selectivity to FUR (74%) was reached after 5 h,
which decreased to 66% after 6 h (Fig. S5c). With the addition of Starbon®450-SO H (50 mg) to
3
-1
the aqueous xylose solution (3 ml of 186 mmol l ) at 170 °C in various reaction times (1-6 h),
the highest FUR yield was 42% at a xylose conversion of 73% (after 6 h, Figure S5b). When
adding Starbon®450-SO H, FUR yield and xylose conversion increase in comparison to the
3
auto-catalyzed system. This is due to the addition of acid sites into the system. In a similar
published system, a high selectivity to FUR (67%) was reached after 2 h at 170 °C, which
gradually decreased with increasing reaction time. Under similar conditions (2 h at 170 °C),
alumina on cordierite reached a selectivity to FUR of 30%; whereas polymeric resins, such as
Nafion NR40 and Amberlyst DT showed a selectivity to FUR of 48% and 27%, respectively [7].
In order to avoid FUR decomposition, and to increase its yield, a biphasic system was
developed adding an organic solvent that would protect FUR formed.
2
.7.Tests in biphasic system
.7.1. Partition coefficient
2
The partitioning of FUR was determined by performing hydrothermal reactions employing a
solution of 5 wt% FUR in water, which was heated with CPME for 60 min at 170 °C at five
different ratios of CPME to aqueous: 5:1, 2:1, 1:1, 1:2 and 1:5 (v/v) and 25 mg of Starbon®450-

SO H. Figure S6 displays the FUR partition coefficients (P) obtained with CPME, where P was
3
determined by using Eq. 6 [66].
[
ꢜꢝꢞ]ꢟꢐꢠ
ꢛꢃ
(Eq. 6)
[
ꢜꢝꢞ]ꢡꢢ
At an aqueous to CPME fraction ratio of 5:1, a FUR partition coefficient of 3.4 was obtained.
This partition coefficient value declined to 3.3, 3.2, 3.0 and 2.8 as the aqueous to CPME fraction
ratio experienced an increment to 1:2, 1:1, 2:1 and 5:1, respectively. Similar values have been
recently reported when studying the partition coefficient of FUR in CPME-aqueous phases at
190 °C for 30 min under auto-catalyzed conditions [43]. The selection of the aqueous to organic
phase ratio in a range of 1:5 to 5:1 has a rather low influence on the partition coefficient as
highlighted in Fig. S6 (in the Supplementary Information).
2.7.2.
Effect of the Ratio Water-CPME on Furfural Yields
In order to evaluate the effect of the ratio water-CPME on catalyzed xylose conversion and
furfural yields, eight ratios of aqueous to organic phase (5:0, 5:1, 3:1, 2:1, 1:1, 1:2, 1:3 and 1:5;
v/v) were studied. When the reaction was performed in pure water (ratio 5:0), the yield of the
produced FUR did not exceed 10%, and the selectivity was approximately 37% (Figure 3).
Furthermore, the conversion of xylose was around 27%. In the biphasic system, xylose
conversion remained around 20% when aqueous to CPME phase ratio decreased from 5:1 to
2:1 (v/v). However, when adding CPME further to decrease the aqueous to organic phase ratio
from 1:1 to 1:5 xylose conversion increased from 22% to 35%. This could be due to xylose
fragmentation into carboxylic acids [30]. Generated carboxylic acids (especially formic and
acetic acid) during hydrothermal reaction concentrate in the aqueous phase with increasing
CPME proportion, thus enhancing the acidity in the aqueous phase. Therefore xylose
conversion increases as lower aqueous to CPME phase ratios are used. A similar effect has
been reported in the literature [67]. By increasing the aqueous to organic phase ratio, the

selectivity to FUR increased with values of 60%, 56%, 55% and 51% for water-CPME
volumetric ratios of 5:1, 3:1, 2:1 and 1:1, respectively. Additionally, when the water-to-CPME
volumetric ratios increased even more to 1:3 and 1:5, the FUR yield increased to 18%, due to a
concomitant increase of the conversion rate. Experiments employing pure CPME were excluded
from this study, since xylose has been proven to be almost insoluble in this organic solvent,
hence FUR yields are generally minimal [67]. Even though the selectivity and xylose conversion
are higher at water to CPME volumetric ratios of 1:5 than at 1:3, it is not practical for industrial
applications due to high organic solvent requirements. Thus, the ratio of 1:3 was selected for
further experiments. This is in accordance with published literature, since biphasic systems
benefit from higher organic to aqueous phase ratios [68].
FUR yield
Xylose conversion
FUR selectivity
60
50
40
30
20
10
0
5
:0 5:1 3:1 2:1 1:1 1:2 1:3 1:5
Aqueous - CPME ratio (v/v)
Figure 3. Effect of aqueous to organic ratio on FUR yield when using CPME. The effect was
-
1
determined for a solution of xylose (186 mmol l ) heated for 60 min at 170 °C with 25 mg of
Starbon®450-SO H (and then cooled down to 4 °C) at eight different ratios of aqueous to
3
organic solvent: 5:0, 5:1, 3:1, 2:1, 1:1, 1:2, 1:3 and 1:5 (v/v).

A design of experiments was developed using a 1:3 water-to-CPME phase ratio (v/v) and 21 mg
of Starbon®450-SO H at various reaction temperatures and times. The original experimental
3
design and the calculated results are illustrated in Table 2.
Table 2. Variables and the calculated response values based on the experiments. Each
experiment consisted of 21 mg of Starbon®450-SO H using an aqueous to CPME phase ratio
3
of 1:3 (v/v) at various reaction temperatures (150, 175 and 200 °C) and times (1, 12.5 and 24 h).
Exp.
No
1
Furfural Yield
(%)
Xylose Conversion
Selectivity to Furfural
T (°C)
t (h)
1
(%)
1.3
(%)
53.1
72.5
65.6
21.0
69.2
50.3
64.8
60.2
61.8
63.2
64.6
150
200
150
200
150
200
175
175
175
175
175
0.7
2
1
69.5
52.2
21
95.9
79.5
100
78.5
100
30.8
100
100
100
100
3
24
4
24
5
12.5
12.5
1
54.4
50.3
20
6
7
8
24
60.2
61.8
63.2
64.6
9
12.5
12.5
12.5
1
0
1
1
Based on the results, practically all xylose was converted at 200 °C in 1 h. At 150 °C only 1.3%
of the xylose was dehydrated (Figure 4a). Yield of FUR varied from 0.7% to 69.5%. Based on
the results, FUR yield increased when reaction temperature and reaction time increased, but
decreased when reaction times longer than 17 h at 200 °C were used (Figure 4b). Selectivity to
FUR ranged from 1.5% and 72.5%. Figure 2c shows that the highest selectivities were achieved

at low reaction temperatures (150 °C–170 °C) and long reaction times (>10 h) or short reaction
times (1–10 h) and high temperatures (170 °C–200 °C).
Prediction models for xylose conversion, furfural yield and selectivity were also successfully
2
determined based on the results. R values indicated that the models explained 92–99% of
variation in the sample properties. The obtained models were then used to predict xylose
conversion and FUR yield of the samples within the experimental design range. As illustrated in
Figure 2, both reaction temperature and time were statistically significant for xylose conversion,
FUR yield and selectivity. Interaction effects between reaction temperature and time were also
significant based on the determined models. As an example, longer reaction times increased
xylose conversion at lower temperatures, but the effect of time decreased at higher
temperatures (Fig. 4a). Longer reaction times also decreased significantly FUR yield at higher
temperatures (Fig. 4b). The highest selectivities were thus obtained by combining high reaction
temperature with low reaction time, or low to medium reaction temperature with medium to high
reaction time (Fig. 4c). The overlay contour plot in Fig. 4d also suggested that a local optimum,
where both xylose conversion and FUR yield were maximized, existed within the experimental
design range. Even though another optimum could be found at reaction temperatures above
200 °C and reaction times below 6 h (Fig. 4d), these conditions were not possible to perform in
the present set-up. Furthermore, those experimental conditions could present a challenge for
the hydrothermal stability of the catalyst. A verification experiment was thus performed by
combining a reaction temperature of 175 °C with a reaction time of 18 hours. The obtained
results indicated that 100% of xylose was converted to 69.3% FUR during the first cycle. The
verification results were in good agreement with the model predictions, which suggested that
101 ± 25.0% of xylose would be converted to 66.9 ± 10.5% (α=0.10) FUR. The analyses of
variance are summarised in Tables S4, S5 and S6 (in the Supplementary Information), for FUR
yield, xylose conversion and selectivity to FUR, respectively.

2
Figure 4. Contour plots based on model prediction for (a) conversion of xylose (%, R = 0.92);
2
2
(
b) furfural yield (%, R = 0.99); (c) selectivity (%, R = 0.92), and; (d) an overplay plot of xylose
conversion and furfural yield. The yellow patterned area in (d) indicates 100% xylose conversion
and a furfural yield of 60% based on the model predictions. Each experiment consisted of 21
mg of Starbon®450-SO H using an aqueous to CPME phase ratio of 1:3 (v/v) at various
3
reaction temperatures (150, 175 and 200 °C) and times (1, 12.5 and 24 h).

2.7.3.
Reusability
The hydrothermal stability of Starbon®450-SO H under the investigated reaction conditions was
3
tested by employing the same catalyst in a series of xylose dehydration reactions. Prior to each
cycle, the sample was washed with deionized water and dried at 105 °C. Figure 5 shows three
consecutive reaction runs of Starbon®450-SO H (at 175 °C in 18 h using 21 mg of
3
-1
Starbon®450-SO H in 0.75 ml of xylose concentration (186 mmol l ) and 2.25 ml of CPME).
3
The notation for Starbon®450-SO H after the reusability test includes a hyphen and the
3
reusability cycle number, e.g. Starbon®450-SO H .
3
—1
After 3 cycles, the catalytic activity of the reused catalyst stayed stable, yielding 70% FUR at
complete xylose conversion. Under similar conditions (175 °C, 18 h and 1:3 aqueous to CPME
phase ratio), the auto-catalyzed system reaches 100% xylose conversion and 59% FUR yield.
As a non-destructive method, Raman spectroscopy is commonly used to characterize the
structure of carbon-based materials. As shown in Figure 6, all the samples showed two
-1
-1
pronounced bands: D and G bands at around 1347 cm and 1593 cm , respectively, which are
a typical characteristic for graphitic carbon and show the presence of aromatic carbon sheets
3
[
69]. The D-band is associated with the breathing modes of sp atoms and is activated only in
the presence of defects and disorder in the carbon structure, whilst the G-band is attributed to
2
the vibrations of sp bonded carbon atoms in the hexagonal lattice. The higher frequency
position of the G-band and the broad and intense D-band indicate the presence of amorphous
phase [69].

100
Conversion
Selectivity
Furfural yield
80
60
40
20
0
1
2
3
Catalyst cycle
Figure 5. Reusability of Starbon®450-SO Hfor the dehydration of xylose to FUR using 21 mg of
3
catalyst using an aqueous to CPME phase ratio of 1:3 (v/v) at 175 °C for 18 h (xylose
conversion (white bar), FUR yield (blue bar) and selectivity to FUR (striped bar)).
As shown in Figure 6, the Raman spectra of the catalyst before and after the cycles were very
similar. The intensity ratio of the G and D band, I /I can be used to estimate the defect level.
D
G
The I /I ratios were virtually identical: 0.984, 0.961, 0.962 and 0.964 for Starbon®450-SO H
—
D
G
3
fresh, Starbon®450-SO H , Starbon®450-SO H
and Starbon®450-SO H , respectively.
3 —3
3
—1
3
—2
Therefore, the catalyst was stable and reusable, in agreement with the reusability tests which
also did not show any loss of yield and selectivity between the three cycles.

D
G
Cycle 3
Cycle 2
Cycle 1
Starbon450-SO42-
500
1000
1500
2000
2500
3000
-
1
Raman shift (cm )
Figure 6. Raman spectra of the catalyst before and after reusability cycles. Each reusability
cycle consisted of adding 21 mg of catalyst to biphasic system employing an aqueous to CPME
phase ratio of 1:3 (v/v) at 175 °C for 18 h. After each cycle, the catalyst was filtered, washed
with deionized water and dried at 105 °C.
Apart from Raman spectra studies, a detailed X-ray photoelectron spectroscopy (XPS) analysis
was performed to get a deeper insight about the surface composition of the materials. In Fig. S7
and Table S3 (In the Supplementary Information) the binding energy values and surface atomic
composition of Starbon®450-SO H before and after hydrothermal reaction are shown. As it can
3

be demonstrated, there were no significant changes in the chemical composition of the surface
of Starbon®450-SO H before and after hydrothermal reaction. Different approaches to measure
3
stability and reusability of solid catalysts have been under discussion in recent years. In order to
design stable catalysts with practical applications, Christopher W. Jones [70] highlighted the
need of understanding deactivation mechanisms of solid catalysts. In this way, catalysis as a
kinetic phenomenon should be used to assess recyclability, stability and deactivation. We agree
with his approach. However, in the present set-up it was not possible to withdraw samples and
analyze them periodically from the reactor as it was done in the referred literature.
Nevertheless, we do not only report number of reusability cycles and yield, but we also included
experimental conditions, such as reaction time and temperature. This adds consistency to the
continuous catalytic activity of Starbon®450-SO H after various reusability cycles. We also
3
agree with Jones that deactivation plays a key role in the improvement of solid catalysts.
Nevertheless, deactivation studies of Starbon®450-SO H in the present paper are not included,
3
but we certainly think they should be addressed in future work.
Starbon®450-SO H is a very attractive solid acid catalyst to form FUR from xylose due to its
3
high BET surface area, excellent hydrothermal stability, and high acid site density. It is
interesting to compare the performance of Starbon®450-SO H with those of other carbon-based
3
catalysts employed in similar set-ups. Wang et al [71] developed a Miscanthus-based catalyst
with sulfonic groups for the FUR formation from xylose and xylan in a CPME/H O 3:1 phase
2
ratio (v/v). Under optimized conditions, they reported a FUR yield of 60% and 42% from xylose
and xylan, respectively (at 190 °C in 1 h). A sulfated lignin-based catalyst was developed by
Antonyraj and Haridas [72] to form FUR and hydroxymethylfurfural (HMF) from xylose and
fructose, respectively, in a methyl isobutyl ketone (MIBK)/H O 7:3 phase ratio (v/v) system.
2
They reported yields of up to 65% FUR at 175 °C in 3 h from xylose and 27% HMF at 150 °C in
3
h from fructose.

In this work, we have shown how higher FUR yields and complete xylose conversion can be
achieved when using Starbon®450-SO H in comparison to similar systems using carbonaceous
3
catalysts. Moreover, Starbon®450-SO H can be funtionalized with sulfuric acid and has shown
3
hydrothermal stability under the experimental conditions presented in this paper. Moreover, it
can be easily separated from the reaction media and further reused without losing its catalytic
activity. Additionaly, statistical methods have been rarely employed to optimize reaction
conditions in converting xylose to FUR. Among the few who have used design of experiments,
Lamminpää [73] supported her research employing experimental design to determine the
interactions of lignin in FUR formation from xylose using formic and sulphuric acid.
A possible alternative to reduce the reaction time presented in this study, could be to increase
the reaction temperature around 200 °C. As Fig. 4 shows, an area with FUR yields above 60%
can be reached at approximately 200 °C under 6 h. Naturally, hydrothermal stability of
Starbon®450-SO H and its feasible reuse under similar experimental conditions have to be
3
investigated. An interesting option to avoid further FUR decomposition would be to modify the
batch system for a plug-flow reactor with an optimized residence time.
3. Conclusions
Furfural formation from xylose in the presence of Starbon®450-SO H was studied in a biphasic
3
system using CPME as organic solvent. The major product of the catalyzed dehydration
reaction of xylose was FUR when adding Starbon®450-SO H. Minor products detected in the
3
system were decomposition products, such as humins, but they were not quantified. The
maximum mole fraction yield obtained was 70% in 18 h at 175 °C at complete xylose
conversion. Whereas the auto-catalyzed system at same conditions yielded 59% FUR.
Starbon®450-SO H showed a high selectivity to FUR in both the monophasic and biphasic
3
systems. Under these experimental conditions, it is demonstrated that the reusability potential of
the carbon-based mesoporous material is possible without decreasing its catalytic activity.

4. Acknowledgements
This research has been done in collaboration with Stora Enso and funded through Erasmus
Mundus Joint Doctoral Programme SELECT+, the support of which is gratefully acknowledged.
GGM acknowledges the support of COST Action FP1306 to embark upon a Short-Term
Scientific Mission at Universidad de Cordoba. The authors are also grateful for the support of
the staff at the Department of Bioproducts and Biosystems at Aalto University, especially to
Carlo Bertinetto and Hans Orassaari; and at Universidad de Cordoba in the Nanoscale
Chemistry and Biomass/Waste Valorisation group. This work made use of Aalto University
Bioeconomy Facilities. JL is a Serra Húnter Fellow and is grateful to ICREA Academia program
and grant GC 2017 SGR 128.
References
[
1] S. Peleteiro, S. Rivas, J.L. Alonso, V. Santos and J.C. Parajó, Bioresour. Technol. 202
2016) 181-191.
(
[
2] K.J. Zeitsch, in: K.J. Zeitsch (Ed.), The chemistry and technology of furfural and its many by-
products, Elsevier B.V., 2000, pp. 98-103.
[
3] G. Gómez Millán, S. Hellsten, J. Llorca, R. Luque, H. Sixta and A.M. Balu, ChemCatChem
1 (2019) 2022-2042.
1
[
4] S.J. Canhaci, R.F. Perez, L.E.P. Borges and M.A. Fraga, Appl. Catal. , B 207 (2017) 279-
85.
2
[
5] Y. Luo, Z. Li, X. Li, X. Liu, J. Fan, J.H. Clark and C. Hu, Catal. Today 319 (2019) 14-24.

[
6] A.S. Mamman, J. Lee, Y. Kim, I.T. Hwang, N. Park, Y.K. Hwang, J. Chang and J. Hwang,
Biofuels, Bioprod. Biorefin. 2 (2008) 438-454.
[
7] G. Gómez Millán, Z. El Assal, K. Nieminen, S. Hellsten, J. Llorca and H. Sixta, Fuel Process
Technol 182 (2018) 56-67.
[
[
8] A.S. Dias, S. Lima, M. Pillinger and A.A. Valente, Catal. Lett. 114 (2007) 151-160.
9] H. Li, A. Deng, J. Ren, C. Liu, W. Wang, F. Peng and R. Sun, Catal. Today 234 (2014) 251-
256.
[
10] R. Weingarten, G.A. Tompsett, W.C. Conner Jr. and G.W. Huber, J. Catal. 279 (2011) 174-
82.
1
[
[
11] S.J. You, Y.T. Kim and E.D. Park, React. Kinet. , Mech. Catal. 111 (2014) 521-534.
12] C. Moreau, R. Durand, D. Peyron, J. Duhamet and P. Rivalier, Ind. Crops Prod. 7 (1998)
95-99.
[
13] R. O'Neill, M.N. Ahmad, L. Vanoye and F. Aiouache, Ind. Eng. Chem. Res. 48 (2009) 4300-
306.
4
[
[
14] R. Sahu and P.L. Dhepe, ChemSusChem 5 (2012) 751-761.
15] S. Lima, A. Fernandes, M.M. Antunes, M. Pillinger, F. Ribeiro and A.A. Valente, Catal. Lett.
135 (2010) 41-47.
[
16] E.I. Gürbüz, J.M.R. Gallo, D.M. Alonso, S.G. Wettstein, W.Y. Lim and J.A. Dumesic,
Angew. Chem. , Int. Ed. 52 (2013) 1270-1274.

[
17] L.R. Ferreira, S. Lima, P. Neves, M.M. Antunes, S.M. Rocha, M. Pillinger, I. Portugal and
A.A. Valente, Chem. Eng. J. 215–216 (2013) 772-783.
[
18] A. Yepez, A. Garcia, M.S. Climent, A.A. Romero and R. Luque, Catal. Sci. Technol. 4
2014) 428-434.
(
[
[
19] M.J.C. Molina, M.L. Granados, A. Gervasini and P. Carniti, Catal. Today 254 (2015) 90-98.
20] G.H. Jeong, E.G. Kim, S.B. Kim, E.D. Park and S.W. Kim, Microporous Mesoporous Mater.
144 (2011) 134-139.
[
21] I. Agirrezabal-Telleria, J. Requies, M.B. Güemez and P.L. Arias, Appl. Catal. , B 115–116
2012) 169-178.
(
[
22] I. Agirrezabal-Telleria, C. García-Sancho, P. Maireles-Torres and P.L. Arias, Chin. J. Catal.
4 (2013) 1402-1406.
3
[
23] I. Agirrezabal-Telleria, J. Requies, M.B. Güemez and P.L. Arias, Appl. Catal. , B 145 (2014)
4-42.
3
[
[
[
24] P. Bhaumik, T. Kane and P.L. Dhepe, Catal. Sci. Technol. 4 (2014) 2904-2907.
25] A.S. Dias, M. Pillinger and A.A. Valente, J. Catal. 229 (2005) 414-423.
26] S. Kaiprommarat, S. Kongparakul, P. Reubroycharoen, G. Guan and C. Samart, Fuel 174
(
2016) 189-196.
[
27] E. Lam, J.H. Chong, E. Majid, Y. Liu, S. Hrapovic, A.C.W. Leung and J.H.T. Luong, Carbon
0 (2012) 1033-1043.
5
[
28] A.S. Dias, M. Pillinger and A.A. Valente, Appl. Catal. , A 285 (2005) 126-131.

[
29] A.S. Dias, M. Pillinger and A.A. Valente, Microporous Mesoporous Mater. 94 (2006) 214-
25.
2
[
30] E. Sairanen, K. Vilonen, R. Karinen and J. Lehtonen, Top. Catal. 56 (2013) 512-521.
[
31] E. Lam, E. Majid, A.C.W. Leung, J.H. Chong, K.A. Mahmoud and J.H.T. Luong,
ChemSusChem 4 (2011) 535-541.
[
32] I. Agirrezabal-Telleria, A. Larreategui, J. Requies, M.B. Güemez and P.L. Arias, Bioresour.
Technol. 102 (2011) 7478-7485.
[
33] C. Sener, A.H. Motagamwala, D.M. Alonso and J.A. Dumesic, ChemSusChem 11 (2018)
321.
2
[
[
34] H. Xiong, H.N. Pham and A.K. Datye, Green Chem. 16 (2014) 4627-4643.
35] C. Termvidchakorn, V. Itthibenchapong, S. Songtawee, B. Chamnankid, S. Namuangruk, K.
Faungnawakij, T. Charinpanitkul, R. Khunchit, N. Hansupaluk, N. Sano and H. Hinode, Adv.
Nat. Sci. : Nanosci. Nanotechnol. 8 (2017) 035006.
[
36] R. Jalili, D. Esrafilzadeh, S.H. Aboutalebi, Y.M. Sabri, A.E. Kandjani, S.K. Bhargava, E.
Della Gaspera, T.R. Gengenbach, A. Walker, Y. Chao, C. Wang, H. Alimadadi, D.R.G. Mitchell,
D.L. Officer, D.R. MacFarlane and G.G. Wallace, Nat. Commun. 9 (2018) 5070.
[
37] V.L. Budarin, J.H. Clark, R. Luque, D.J. Macquarrie, A. Koutinas and C. Webb, Green
Chem. 9 (2007) 992-995.
[
[
38] V. Budarin, R. Luque, D.J. Macquarrie and J. Clark, Chem. - Eur. J. 13 (2007) 6914-6919.
39] V.L. Budarin, J.H. Clark, R. Luque and D.J. Macquarrie, Chem. Commun. (2007) 634-636.