Low-Temperature Continuous-Flow Dehydration of Xylose Over Water-Tolerant Niobia–Titania Heterogeneous Catalysts

Article abstract of DOI:10.1002/cssc.201801414

The sustainable conversion of vegetable biomass-derived feeds to useful chemicals requires innovative routes meeting environmental and economical criteria. The approach herein pursued is the synthesis of water-tolerant, unconventional solid acid monolithic catalysts based on a mixed niobia–titania skeleton building up a hierarchical open-cell network of meso- and macropores, and tailored for use under continuous-flow conditions. The materials were characterized by spectroscopic, microscopy, and diffraction techniques, showing a reproducible isotropic structure and an increasing Lewis/Br?nsted acid sites ratio with increasing Nb content. The catalytic dehydration reaction of xylose to furfural was investigated as a representative application. The efficiency of the catalyst was found to be dramatically affected by the niobia content in the titania lattice. The presence of as low as 2 wt % niobium resulted in the highest furfural yield at 140 °C under continuous-flow conditions, by using H₂O/γ-valerolactone as a safe monophasic solvent system. The interception of a transient 2,5-anhydroxylose species suggested the dehydration process occurs via a cyclic intermediates mechanism. The catalytic activity and the formation of the anhydro intermediate were related to the Lewis acid sites (LAS)/Br?nsted acid sites (BAS) ratio and indicated a significant contribution of xylose–xylulose isomerization. No significant catalyst deactivation was observed over 4 days usage.

Full text of DOI:10.1002/cssc.201801414

Accepted Article
Title: Low temperature continuous flow dehydration of xylose over
water-tolerant niobia-titania heterogeneous catalysts
Authors: Carmen Moreno-Marrodan, Pierluigi Barbaro, Stefano
Caporali, and Filippo Bossola
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201801414
Link to VoR: http://dx.doi.org/10.1002/cssc.201801414
A Journal of
www.chemsuschem.org

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
Low temperature continuous flow dehydration of xylose over
water-tolerant niobia-titania heterogeneous catalysts
Carmen Moreno-Marrodan,[a] Pierluigi Barbaro,*[a] Stefano Caporali,[b] and Filippo Bossola[c]
The sustainable conversion of vegetable biomass-derived feeds
to useful chemicals requires innovative routes matching
environmental and economical criteria. The approach herein
pursued is the synthesis of water-tolerant, unconventional solid
dramatically affected by the niobia content in the titania lattice.
The presence of as low as 2% wt niobium resulted in the highest
furfural yield at 140 °C reaction temperature under continuous
flow conditions, using H2O / g-valerolactone as safe monophasic
acid monolithic catalysts based on
a
mixed niobia-titania
solvent system. The interception of
a
transient 2,5-
skeleton building up a hierarchical open-cell network of meso
and macropores, and tailored for use under continuous flow
conditions. The materials were characterized by spectroscopic,
microscopy and diffraction techniques showing a reproducible,
isotropic structure and an increasing Lewis / Brønsted acid sites
ratio with increasing Nb content. The catalytic dehydration
reaction of xylose to furfural was investigated as representative
application. The efficiency of the catalyst showed to be
anhydroxylose species suggested the dehydration process to
occur via a cyclic intermediates mechanism. The catalytic
activity and the formation of the anhydro intermediate were
related to the LAS/BAS ratio and indicated
a significant
contribution of xylose-xylulose isomerization. No significant
catalyst deactivation was observed over 4 days usage.
transfer.[10,11] When exposed to a continuous flow of a liquid
phase, the pressure drop per unit reactor length (DP/L)
generated by a hierarchically porous monolith is ca. one
order magnitude lower compared to the corresponding
mesoporous material under the same flow conditions.[12]
Based on the Darcy's law, this is attributed to the high
permeability coefficient of hierarchically porous monoliths (k >
0.25 µm2),[9] which is proportional to the macropore size (DP/L
= µv/k, µ viscosity; v linear velocity).[13]
The development of efficient, continuous flow reactor
systems is critical in catalysis, due to the several advantages
compared to batch operations including increased safety and
reactor volume productivity, better heat transfer, reduced
development time, improved process control, simpler
downstream processing and smaller hold-up volumes
(solvents, reagents).[14,15] Continuous refresh of catalyst
surface may also significantly reduce active site inhibition due
to adsorbed co-products.[16] Monolithic catalysts have shown
to be particularly useful to this regard, as they overcome
most issues typical of packed-bed systems, such as high
backpressure evolution, low contacting efficiency, broad
distribution of residence times, formation of hot-spots or
stagnation zones, which result in uncontrolled fluid dynamics,
hence in unsatisfactory performance.[17]
Introduction
A monolith is “a shaped, fabricated intractable article with a
homogeneous microstructure that does not exhibit any
structural
components
distinguishable
by
optical
microscopy”.[1] According to this definition, conventional
foams or honeycomb-type materials with millimetre size
cavities,[2] commonly reported in the chemical engineering
literature for gas-phase unselective thermal processes,[3,4] do
not belong to this category. By contrast, unconventional
monoliths with sub-millimetre scale porosity have been
described in various flow-through applications for the fine
chemistry, including chromatography and catalysis, thanks to
the enhanced contact with molecular substrates because of
the large surface area.[5] Particularly, monoliths featuring an
isotropic, hierarchically porous network of narrowly size
distributed, open-cell macropores (1-20 µm) and mesopores
(2-50 nm) within the struts (1-5 µm)[6,7] feature a peculiar
hydrodynamic behaviour joining the advantages of efficient
processing, typical of mesoporous materials, and fast
diffusion, due to macropores.[8,9] Compared to mesoporous-
only materials, these monoliths provide considerable benefits
in heterogeneous catalysis in terms of reduced pore clogging,
better active sites accessibility and improved mass
Polymeric materials were the first to prove the benefits of
monoliths in catalysis.[18,19] However, they show several
drawbacks, including limited thermal, mechanical and
chemical stability, shrinking phenomena, porosity changes
with swelling, which may result in non-uniform radial
permeability and in significant pressure drop at high flow
rates.[20,21] In order to circumvent these issues, different types
of inorganic monoliths have been synthesized by spinodal
decomposition showing improved resistance and rigidity.[22,23]
Examples include silica,[24,25,26,27] alumina,[28,29,30] titania,[31,32]
zirconium phosphate[33] and AlPO4.[34] Despite of their
advantages, use of hierarchically porous inorganic monoliths
in flow catalysis has been scarcely investigated so far, being
essentially explored for HPLC applications.[35,36] One reason
[a] Dr. C. Moreno-Marrodan, Dr. P. Barbaro
Consiglio Nazionale delle Ricerche, Istituto di Chimica dei Composti
Organo Metallici,
Via Madonna del Piano 10, 50019 Sesto Fiorentino, Firenze (Italy) E-
mail: pierluigi.barbaro@iccom.cnr.it
[b] Dr. S. Caporali
Consorzio Interuniversitario Nazionale per la Scienza e Tecnologia dei
Materiali, Via Giusti 9, 50121, Firenze
and Consiglio Nazionale delle Ricerche, Istituto dei Sistemi Complessi,
Via Madonna del Piano 10, 50019 Sesto Fiorentino, Firenze, Italy.
[c] Dr. F. Bossola
Consiglio Nazionale delle Ricerche, Istituto di Scienze e Tecnologie
Molecolari, Via Golgi 19, 20133 Milano (Italy)
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/cctc.xxxx
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
for this is the hardly reproducible preparation of monoliths
with 2 - 15 mm diameter to perform microsynthesis,[37,38]
owing to the extreme structure sensitivity from the phase
system composition.[39] Specific functionalization of the
monolithic material may also be required for catalytic
applications.[40]
batch (e.g. propane dehydrogenation)[84] and in gas-phase
continuous flow conditions (butan-1-ol[85,86] and toluene total
oxidation[87]).
In the present work, the mixed niobia-titania monoliths
synthesized were carefully characterized and their catalytic
activity was demonstrated under a continuous flow of a liquid
phase, in a reaction of great relevance to the development of
sustainable bio-economy i.e. the acid-catalyzed dehydration
reaction of xylose to furfural. With the diminishing and
dwindling supply of fossil resources, and the climatic
problems associated with the use of these, end-users started
looking for alternative raw materials for their productions.
Lignocellulose-derived materials has thus become of
increasing importance as renewable source of carbon for the
large-scale manufacture of chemicals.[88] Furfural is
considered one of the "top" platform molecules for the
production of polymers, solvents and fine-chemicals,
obtainable either from hemicellulose or xylose, which is its
main component.[89,90]
Few functionalized inorganic monoliths with hierarchical
porosity have been described, either bearing supported metal
nanoparticles (Ag,[41] Pd[42,43]), for metal-type catalysis (e.g.
hydrogenations), or grafted acidic/basic groups (-SO3H, -NH2,
Al2O3, HPO42-), for organic-type catalysis (e.g. dehydrations,
condensations).[44,45] The catalytic performance of these
monoliths has been compared with both packed-bed systems
under continuous flow and with batch setups. The better
productivity was clearly demonstrated for the monolithic
systems in the Pd-catalyzed hydrogenation reaction of
cyclohexene,[46,47]
Knoevenagel,[48,49]
for
metal-free
trans-esterification,
and Friedländer
Diels-Alder[50]
reactions.[51] In terms of selectivity, Pd@SiO2 monolith also
performed better than Pd onto conventional mesoporous
silica in the partial hydrogenation reaction of 3-hexyn-1-ol
and 3-halogeno-nitrobenzenes, both under flow and batch
conditions.[52] Irrespective of the supported catalyst and
reaction, monolithic systems have shown remarkable stability
for prolonged reaction times, with no significant activity decay
observed up to 3 days time-on-stream.[46,50] All the above
findings show the advantages of unconventional inorganic
monolithic catalysts compared to the classical heterogeneous
ones.[53]
Solid acid catalysis is of paramount importance in both
the hydrocarbon[54,55] and the biorefinery industry, wherein
conversion of real feedstocks (e.g. aqueous solution of
carbohydrates),[56,57] usually requires tolerance to water and
to medium-high temperatures.[58] Prompted by the low
efficiency and the unsatisfactory thermal stability of the
previously reported monolithic acid catalysts (i.e. zirconium
phosphate[45] and sulfonated silica[59]), we sought to develop
alternative materials with improved performance for the
conversion of biomass-derived substrates under continuous
flow conditions. Herein, we report the first example of 3D
isotropic, niobia-titania monoliths featuring a well-defined,
hierarchically porous structure with interconnected meso and
macropores. Choice of mixed titanium-niobium oxide
materials has multiple motivations: the reproducible synthesis
of titania monoliths (see above), the thermal and chemical
stability of TiO2,[60,61] the strong solid acid properties,[62,63]
tolerance to water[64] and catalytic activity of Nb2O5,[65,66] the
fast deactivation of pure niobia catalysts.[67] Other potentially
useful properties for further catalytic applications include the
Figure 1. Optical image of dual porosity niobium-titanium oxide monolith.
Table 1. Mesoporosity features of monolithic materials.[a]
D
V
S
Monolith
[nm]
[cm3 g-1]
[m2 g-1]
TiO2-MNL
3.8 ± 0.2
3.9 ± 0.6
4.0 ± 0.4
0.20 ± 0.02
0.20 ± 0.03
0.20 ± 0.01
154 ± 9
146 ± 5
145 ± 4
NbTiO-MNL1
NbTiO-MNL2
[a] Pore diameter (D), porous volume (V), BET surface area (S).
Results and discussion
68
[
strong metal-support interaction (SMSI) of both TiO2 ] and
Nb2O5,[69] the strong metal-promoter interaction (SMPI) of
niobia,[70,71] the photocatalytic activity of titanium and niobium
oxides.[72,73]
Synthesis and characterization of the hierarchically
porous niobium-titanium oxide monoliths
Bimodal macro-mesoporous, mixed oxides monoliths
(NbTiO-MNL, typically 5 mm diameter, 30 mm length, ca.
480 mg, Figure 1) were synthesized by a combination of
spinodal decomposition and sol-gel transition method.[91]
Thus, polymerization-induced phase separation and
gelation of a composition-modulated NbCl5-TPO–NMFA–
PEO water-phase system led to the reproducible formation
of monoliths of desired size. A labelling scheme was
adopted according to the niobium molar content % of the
starting reaction mixture (Table 1). As previously reported
Mixed niobia-titania materials have some precedents in
the literature with a Nb loading up to 40% atomic percentage.
These include thin films, ceramic pellets, mesoporous and
hierarchically micro(meso)macroporous powders. Most
materials are based on niobium oxide layered onto
preformed TiO2 supports,[74,75,76] although few mixed oxides
have also been reported.[77,78,79,80] Applications focused on
sensing[81,82] and photocatalysis.[83] Studies in chemocatalysis
were restricted to unselective oxidative processes under
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
for other inorganic oxides, the procedure allowed to
reinforce the monolith skeleton, resulting in the formation of
uniform and crack-free materials featuring a hierarchically
porous structure.[39,46] During thermal treatment in the aging
stage, temperature was cautiously kept below 350 °C, since
the high acid strength of hydrated niobium pentoxide
(Hammett acidity H0 » - 5.6, comparable to that of 70%
H2SO4) is known to be drastically reduced above 500
°C.[92,93]
Textural properties. Porosity and morphology of the
monoliths were investigated by N2 porosimetry, mercury
intrusion, X-ray microtomography, scanning and transmission
electron microscopy.
Nitrogen isotherms showed the presence of narrow
distributions of mesopores with average pore diameter 3-4 nm,
connected through windows of ca. 2 nm, corresponding to 0.1-
0.2 cm3 g-1 cumulative pore volumes and 136-154 m² g-1 BET
surface areas. Selected porosity data are reported in Table 1.
The mesopores size distribution of NbTiO-MNL2 is reported in
Figure 2 left, as representative example. No significant changes
of mesopore width, volume or surface area were detected up to
2% niobium content.
The as-prepared monoliths were thoroughly characterized
by a combination of spectroscopic, microscopy and analytical
techniques.
SEM analysis showed the monoliths to have a well-defined,
isotropic co-continuous microstructure consisting in a skeleton of
ca. 3 µm thickness, forming an open-cell network of macropores
of average size of ca. 4 μm.[94] Representative SEM images are
reported in Figure 3 for NbTiO-MNL2. Neither aggregates of
globular particles within the struts, nor significant morphologic
differences were detected by SEM up to 2% Nb loading.[95]
The macroporosity and mechanical stability of the monoliths
were examined by Hg intrusion measurements, showing narrow
distributions of macropore diameters with mean values of 3 μm,
in fair agreement with SEM evidences. The pore size diagram of
NbTiO-MNL2 is reported in Figure 2, right, as example. In that
case, the total macroporous volume and surface area were 0.56
cm3 g-1 and 2.09 m2 g-1, respectively, corresponding to an overall
porosity of 53.7%.
TEM analysis showed the monoliths skeleton to be
formed by spherical crystallites of 4-7 nm size building up
the mesopores cavities. The nanocrystals are attached with
no preferred orientation. A TEM image of NbTiO-MNL2 is
reported in Figure 4. No significant difference of crystallite
size was displayed by TEM analysis of monoliths with
different Nb loadings.
Figure 2. Pore size distribution in NbTiO-MNL2: mesopores size calculated
from the N2 desorption branch using the BJH method (left) and macropores
size from mercury porosimetry (right).
The overall morphology of the monoliths was confirmed
by X-ray tomography, which showed an isotropic, uniform
radial distribution of voids and struts, and a very narrow size
distribution of windows and diffusion pores. The data
calculated were in line with those obtained from SEM and
Hg porosimtery. In the case of NbTiO-MNL2 an average 5
μm skeleton thickness and 5 μm macropores diameter were
obtained, corresponding to a total porosity of 49%.[46]
A
tomography reconstruction image of NbTiO-MNL2 is
reported in Figure 5.
Figure 3. Typical SEM images of NbTiO-MNL2 (secondary electrons, 5 keV,
high vacuum). Left: 11000 magnifications. Right: 29000 magnifications.
Figure 5. X-ray tomography 3D-reconstruction image of NbTiO-MNL2.
Figure 4. Typical TEM image of NbTiO-MNL2 (200 keV).
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
other than Nb, Ti and O were detected above the sensitivity
limit of the instrument (0.1% wt).
Table 2. Actual bulk composition of monolithic materials.[a]
The XRD patterns of all monolithic materials showed the
characteristic peaks of anatase-type crystalline TiO2 due to
(1 0 1), (0 0 4) and (2 0 0) reflections at 2θ of 25.39º, 38.01º
and 48.00º, respectively (Figure 6). This indicates that the
anatase phase is retained upon incorporation of up 2% wt
Nb, as previously reported for similar niobium
Nb
Ti
Ti / Nb
Monolith
[wt%]
[wt%]
molar ratio
TiO2-MNL
-
59.8 ± 0.4
83.5 ± 0.4
84.6 ± 0.4
-
NbTiO-MNL1
NbTiO-MNL2
1.0 ± 0.1
2.0 ± 0.3
162.1 ± 0.2
82.1 ± 0.4
concentrations in other Nb-doped TiO2 materials.[85,87]
A
[a] Data from ICP-OES analysis. Mean values over three samples.
weak peak at 2θ = 30.8º was detected, attributable to the (1
2 1) diffraction plane of brookite phase. The calculated
percentage of brookite in the monoliths was less than 1 wt%
in any case. No other phases, such as Nb2O5, were
observed in the XRD profiles. The peak at ca. 55° could be
deconvoluted into two components, corresponding to the
(105) and (211) diffraction planes, which showed small
shifts to lower 2θ values with increasing niobium content
(Table 3). According to the Bragg's law, this can be
explained in terms of replacement of Ti4+ (0.61 Å) by larger
Nb5+ (0.64 Å) ions in the titania lattice.[81,96] The size of
crystallites in the monolithic materials, calculated using the
Debye-Scherrer formula, showed the crystallite size to
slightly increase with niobium loading (Table 3), as
previously reported.[97]
Figure 6. XRD patterns of niobia-titania monoliths. The inset is the
deconvoluted peak in the (105) and (211) region for NbTiO-MNL2.
Table 3. Selected XRD properties of monolithic materials.
2 Theta [°]
(105)
Crystallite size
[nm]
Monolith
(211)
55.42
55.25
55.16
TiO2-MNL
54.28
54.09
54.03
6.8
7.2
7.4
NbTiO-MNL1
NbTiO-MNL2
Chemical composition. The composition of the
monoliths was analyzed by ICP-OES, EDS, XRD and XPS.
Irrespective of the Nb loading, bulk ICP-OES analyses
revealed the solids to have a metal content in line with that
of the starting reaction mixture, thus indicating the
quantitative incorporation of Nb and Ti into the monoliths.
Representative data are reported in Table 2. The values of
the Ti / Nb ratio (82-162) indicated the niobium content to
be well below the reported solubility limit of Nb into anatase
materials (ca. 4) in any case,[83,98] which is fair agreement
with the absence of different phases detected in the XRD
spectra (vide infra). EDS analyses, either multipoint or 2D
maps, were in agreement with ICP-OES data within the
experimental errors, showing homogeneous radial and
longitudinal compositions within the monoliths. No elements
Figure 7. XPS spectrum and curve fit of NbTiO-MNL2 in the Ti 2p (top) and
Nb 3d (bottom) regions.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
g-1), respectively (Table 5).[104] The overall acid density
showed to decrease with increasing Nb content, i.e. TiO2-
MNL > NbTiO-MNL1 > NbTiO-MNL2, while the LAS / BAS
ratio (11-160) significantly increased in the same order.
Anatase TiO2 has been reported to have weak acidic
properties, significantly dependent from its structure and
morphology, thus to be strongly affected by the preparation
method.[105,106] The acidity of Nb2O5·nH2O has been examined
in greater detail. The structure of niobium pentoxide is mainly
composed of distorted NbO6 octahedra and NbO4 tetrahedra.
While NbO4 tetrahedra function as Lewis acid sites, the highly
polarized Nb-O bonds in octahedra cause part of the surface
OH groups to act as Brønsted acid sites.[107,108] The solid acid
activity of both sites in water have been demonstrated in
several instances.[107,109,110] The acidic properties of niobic
acid depend on its thermal treatment. Brønsted and Lewis
acid densities are highest at 100 °C and 300 °C, respectively,
while decrease until being almost negligible at 500 °C,[93]
which is in fair agreement with our results.
The ratio between the amount of Lewis and Brønsted
sites have been previously quantified in other porous mixed
niobium-titanium oxide materials, showing a trend analogous
to that observed in the monoliths herein described, in the
same Nb content range.[85] Although not exactly comparable
with our materials, previous studies on titania-supported
niobium oxide revealed the absence of BAS and a non linear
decrease of LAS with niobia loading, that was explained in
terms of partial replacement of existing sites.[111]
Table 4. Selected XPS binding energy data [eV] for niobia-titania materials.[a]
Nb
Ti
Ti / Nb
ratio [b]
ref.
Material
3d5/2
2p3/2
NbTiO-MNL1 206.9
NbTiO-MNL2 206.9
(2.7)
(2.7)
(2.8)
458.5
458.4
(5.7)
(5.7)
145
72
this work
this work
[95]
Nb2O5
TiO2
207.3
459.3
(6.2)
[96]
[a] Spin-orbit coupling values in parenthesis. [b] Estimated molar ratio.
Table 5. Lewis (LAS) and Brønsted (BAS) acid sites concentration on
monolithic materials.[a]
LAS
[μmol g-1]
BAS
[μmol g-1]
LAS / BAS
ratio
Monolith
TiO2-MNL
244
241
160
22
8
11
NbTiO-MNL1
NbTiO-MNL2
30
1
160
[a] From FT-IR pyridine-desorption at 150 °C. LAS 1448 cm-1, BAS 1540 cm-1.
XPS measurements were carried out on all monolithic
samples in order to define the oxidation state of the species
involved. Representative XPS spectra and the curve fittings
in the Ti 2p and Nb 3d regions, where the usual doublets are
observed, are reported in Figure
7 for NbTiO-MNL2.
Regardless of the niobium content, all spectra could be
deconvoluted in one component only, whose binding energy
values are in agreement with those previously reported for
other Nb-doped TiO2 anatase materials for Nb(V) (Nb 3d5/2
206.9 ± 0.1 eV, spin-orbit coupling 2.7 eV) and Ti(IV) (Ti 2p3/2
458.5 ± 0.1 eV spin-orbit coupling 5.7 eV).[77,85] Significant
data are summarized in Table 4. No peaks attributable to
other oxidation states, namely Nb(IV) and Ti(III), were
detected above the detection limit of the technique. The
observed binding energies were somewhat lower than those
of pure Nb2O5[98] and anatase TiO2[99] (Table 4), which can be
attributed to a minor lowering of local symmetry due to the
substitution of slightly larger Nb5+ ions for Ti4+ in the anatase
network.[83] It was previously suggested that, when Nb5+
replace Ti4+ into a TiO2 lattice, the excess of positive charge
may be compensated either by the creation of one Ti cation
vacancy for every four Nb5+ ions, by the reduction of one Ti
atom from (IV) to (III) oxidation state for every Nb5+ ion, or by
incorporation of one interstitial oxygen per two Nb5+ ions, with
the formation of vacancies being the preferred mechanism
for up to 10 at.% Nb content.[100,101] In all monolithic samples,
the Ti / Nb molar ratios evaluated from XPS data was in fair
agreement with those obtained from ICP-OES analysis within
the experimental errors, which indicates a uniform surface /
bulk composition.[102]
Scheme 1. Pathways for the acid-catalyzed dehydration reaction of xylose to furfural.
Catalytic dehydration of xylose to furfural
In order to evaluate the catalytic performance of the niobium-
titanium oxide monoliths under continuous flow conditions, as
well as the effect of niobia presence in the titania lattice, a
representative reaction of great relevance to the valorisation
of vegetable biomass was used,[112,113] namely the
dehydration of xylose to furfural.
Furfural is an important platform chemical obtainable from
hemicellulose-derived xylose via challenging threefold acid-
catalyzed dehydration at high temperatures.[114,115] Current
production methods rely on aqueous phase batch operations
and use of homogeneous mineral acids (H2SO4).[116,117]
Several oil-alternative, key industrial intermediates and
consumer chemicals are accessible through further furfural
processing.[118,119] Despite the significance of the reaction and
the number of in-depth studies performed, the mechanism of
xylose dehydration is not yet definitely agreed and several
Acid capacity. The concentration of Lewis (LAS) and
Brønsted (BAS) solid acid sites in the NbTiO monoliths was
investigated by pyridine desorption FT-IR spectroscopy
experiments at 150 °C.[103] Integration of peaks at 1540 and
1448 cm-1 revealed a moderate and a high concentration of
Brønsted (22-1 µmol g-1) and Lewis acid sites (244-160 µmol
possibilities have been hypothesized.[120]
A
generally
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
accepted pathway is sketched in Scheme 1.[121] It was
suggested that xylose undergoes a Brønsted acid-catalyzed
conversion to various (anhydro) intermediates and other
decomposition products,[122] followed by dehydration to
furfural. It is known that the presence of Lewis acids
Effect of the reaction temperature. The xylose
dehydration reaction was investigated under fixed residence
times in the temperature range 100 - 150 °C. The xylose
conversion and the selectivity to furfural observed at t 85 s
are reported in graphical format in Figure 8, left, for NbTiO-
MNL2 catalyst, as representative example, whereas the
reaction solution composition is reported in Figure 8, right.
Furfural selectivity was calculated as:
increases
isomerization to xylulose.[123,124] The selectivity of the process
is limited by number of side-reactions, including
the
xylose-furfural
conversion rate
via
a
condensation of xylose and furfural with intermediates,
decompositions, oligomerisations and resinifications.[125,126]
The formation of significant amounts of unidentified by-
products, such as solid humins or polymers, is responsible
for the difference usually observed between the amount of
xylose consumed and products detected in solution.
(
)
Co furfural
(
)
Sel. furfural =
·100
(
)
(
)
Ci xylose -Co xylose
where Co and Ci are the observed and initial concentrations,
respectively. In addition to xylose, the only compounds
detected were furfural and the mono-dehydration product
2,5-anhydroxylose (GC-MS, Figure S7). The formation of
insoluble humins was indicated by a brownish colouring of
the solid catalyst.[133] In line with reported batch studies, the
overall amount of dehydrated products (furfural and 2,5-
anhydroxylose) corresponded to ca. a 50% yield, consistently
with a significant contribution of side reactions.[134] Both
conversion (up to 100%) and selectivity to furfural (up to
35%) increased with increasing reaction temperature, with a
maximum observed at 140 °C. Selectivity slightly decreased
above that temperature. If these data confirm that the xylose
conversion process is favoured by high temperatures, two
justifications can be found for the observed selectivity trend.
On one hand, the concentration of furfural increased with
The Nb2O5-TiO2 monoliths were scrutinized under a
variety of flow conditions by modifying the reaction
temperature, the residence time t (i.e. the amount of time
that the feed spends inside the reactor depending on the
substrate solution flow rate),[127] and the combinations
thereof. Thus, after cladding the monolith with a heat
shrinkable Teflon tube and connecting to a continuous flow
system, a stream of xylose solution in water : GVL = 1 : 9
(v/v) was allowed to flow through the monoliths under
controlled conditions (Figure S6). Under our experimental
conditions, highly reproducible flow patterns were observed
within very short equilibration times. The reaction mixture
collected at the outlet of the reactor was periodically analyzed
via HPLC, GC-Ms and ICP-OES. Choice of the solvent
mixture was motivated by the use of water as real xylose
feed and by the fact that the addition of organic solvents is
known to improve the selectivity of the acid-mediated xylose
dehydration.[128,129] GVL (g-valerolactone) is a water miscible,
safe solvent obtainable from lignocellulosic biomass via
catalytic methods.[130,131] Compared to pure water, benefits of
water-GVL mixtures in terms of furfural selectivity have been
attributed to the decrease in the activation energy of xylose
dehydration and the increase of that of furfural degradation,
on the basis of kinetic studies using both homogeneous and
heterogeneous, Brønsted and Lewis catalysts.[132]
increasing reaction temperature at the expenses of
a
decreasing concentration of 2,5-anhydroxylose (Figure 8,
right). On the other hand, it is known that xylose dehydration
is faster than furfural loss reactions (e.g. resinification) at high
temperatures, which is called "entropy effect".[135,136] As a
matter of fact, the collective yield of dehydration products
(2,5-anhydroxylose plus furfural) reached a maximum at ca.
130 °C. Above 140 °C, the yield of furfural decreased, likely
due to decomposition.[137] The detection of the 2,5-
anhydroxylose intermediate indicates that the first
dehydration step of xylose is faster than the further
dehydrations (Scheme 2). Earlier kinetic studies of H2SO4-
catalyzed xylose dehydration in water confirms this
hypothesis.[138] Analogous findings have been reported for
the two-fold dehydration reaction of sorbitol to isosorbide via
1,4-sorbitan.[139,140]
Figure 8. Selected data for the dehydration reaction of xylose over NbTiO-
MNL2 catalyst under continuous flow conditions. Effect of temperature at fixed
residence time (t 85 s, xylose 0.02 M in H2O : GVL = 1:9 v/v, average values over
three experiments). Left: conversion and selectivity to furfural. Right: xylose, 2,5-
anhydroxylose, furfural and collective yield of dehydration products.
Scheme 2. Acid-catalyzed xylose dehydration mechanism via cyclic
intermediates. Adapted from Ref [120] - Published by The Royal Society of
Chemistry.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
Effect of the residence time. The continuous flow
dehydration reaction of xylose using monolithic catalysts
was examined in the residence time range 71-224 s (0.4-
0.1 mL min-1), under fixed reaction temperature. No
backpressure was generated by the monolithic reactors at
these solution flow rates. Selected data at 130 °C for the
NbTiO-MNL2 catalyst are shown in Figure 9, as
representative example. The effect of residence time
change was similar to that above described for the reaction
temperature. An increase of the residence time, i.e. a longer
contact time between catalyst and substrate, resulted in
increasing xylose conversion and selectivity to furfural. Full
conversion was achieved for t 141 s. No selectivity
decrease was observed up to t 224 s (0.1 mL min-1). At this
residence time value, the furfural and the dehydration
product collective yields were 39% and 54%, respectively.
The yield of furfural increased with increasing residence
time, at the expenses of the concentration of the transient
2,5-anhydroxylose specie.
Scheme 3. Dehydration of xylose via isomerization to xylulose.
Overall, the combination of reaction conditions resulting
in the best, non-optimized (i.e. within the temperature and
t ranges examined) furfural yield of 42% was 140 °C and
141 s residence time (Figure S10).
Figure 9. Selected data for the dehydration reaction of xylose over NbTiO-
MNL2 catalyst under continuous flow conditions. Effect of residence time at
fixed reaction temperature (130 °C, xylose 0.02 M in H2O : GVL = 1:9 v/v,
average values over three experiments). Left: conversion and selectivity to furfural.
Right: xylose, 2,5-anhydroxylose, furfural and collective yield of dehydration
products.
Interception of the 2,5-anhydroxylose supports that, at
least under our experimental conditions, the dehydration of
xylose occurs via a “cyclic intermediates rearrangement"
mechanism (Scheme 2),[141,142] that was suggested to be
energetically more favourable compared to a "dehydration
of acyclic intermediates” one.[143,144] Nonetheless, we
cannot rule out a contribution to the mechanism based on
xylose-xylulose isomerisation followed by dehydration, due
to the combined action of Lewis and Brønsted acid sites
(Scheme 3).[145,146,147] Actually, minor amounts of xylulose (<
2%) were only detected using the monolithic catalyst with
the highest LAS / BAS ratio, i.e. NbTiO-MNL2. To the best
of our knowledge, the formation of the 2,5-anhydroxylose
intermediate, despite being proposed in various
studies,[118,119] has never been demonstrated by direct
observation during acid-catalyzed xylose dehydration. 2,5-
Anhydroxylose can be isolated by Swern oxidation of
anhydroarabitol in two steps,[148] or in 15% yield from
sugarcane bagasse after extraction with aqueous
ethanol.[149] The fact that in our system the rate of formation
of 2,5-anhydroxylose is higher than the rate of its further
dehydration can be tentatively attributed to the unusually
high LAS/BAS ratio (160), which enables fast 2,5-
anhydroxylose formation via Lewis acid-catalyzed xylose-
xylulose isomerisation (Scheme 3).
Figure 10. Selected data for the continuous flow dehydration reaction of
xylose over monolithic catalysts at different residence times (130 °C, xylose
0.02 M in H2O : GVL = 1:9 v/v, average values over three experiments). Left:
conversion. Right: selectivity to furfural.
Comparison among monolithic catalysts. The
monolithic
materials
showed
significantly
different
performance in the catalytic dehydration reaction of xylose.
Representative data are reported in graphical format in
Figure 10. As above described for NbTiO-MNL2, xylose
conversion increased with increasing residence time using
both NbTiO-MNL1 and TiO2-MNL, however with remarkably
lower values in the same t range (20-45%) (Figure 10, left).
By contrast, while selectivities to furfural were in the same
range observed using NbTiO-MNL2, their values decreased
with increasing t (Figure 10, right) using NbTiO-MNL1 and
TiO2-MNL, due to the formation of higher amounts of by-
products (see Supporting Information). Compared to
NbTiO-MNL2, selectivities at the same conversion level
were therefore much lower using the NbTiO-MNL1 and
TiO2-MNL catalysts. A similar behaviour was observed upon
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
Table 6. Continuous flow dehydration reaction of xylose to furfural using monolithic catalysts under identical conditions.[a]
Conversion [b]
[%]
Dehydration yield [c]
[%]
Furfural yield
[%]
Selectivity [d]
[%]
Productivity [e]
[mol g-1 h-1]∙103
STY [f]
[kg L-1 h-1]∙103
Catalyst
1
2
3
TiO2-MNL h
31
32
98
22
22
50
11
12
29
34
37
30
0.1
0.1
0.2
7.2
NbTiO-MNL1 i
NbTiO-MNL2 l
7.6
19.0
[a] Reaction conditions: xylose 0.02 M in H2O : GVL = 1:9 v/v, 0.1-0.4 mL min-1, residence time t = 106 s, 130 °C. Start time: attainment of steady state
conditions ca. 0.5 h. Data from HPLC analysis. [b] Conversion average value over three experiments ± 2 %. [c] Collective yield of dehydration products 2,5-
anhydroxylose and furfural. [d] Selectivity to furfural. [e] Furfural productivity calculated at the yield indicated and per gram of catalyst. [f] Furfural space-time-
yield calculated for the yield indicated and monolithic reactor volume.
change of the reaction temperature, indicating that NbTiO-
MNL2 was significantly the most efficient among the three
monolithic catalysts. It turns out that, under identical flow
conditions, use of NbTiO-MNL2 resulted in a much higher
dehydration yield and furfural productivity. For the sake of
comparison, selected data are shown in Table 6 for xylose
conversions below 100%, wherein space-time-yields (STY,
-1
kgfurfural Lreactor h-1) and furfural productivities are reported.
Test reactions in batch conditions confirmed this trend. The
above results indicate that the addition of as low as 2% wt
niobia into the titania lattice had a dramatic effect on the
catalytic performance of the monoliths.
A
possible
explanation can be found in the different acidic properties of
the materials. Inspection of data reveals that the
productivity of furfural is closely related to the LAS/BAS
ratio in the monolithic catalysts. STY results are shown in
Figure 11 as illustrative example. The higher the ratio, the
higher was the reactor productivity. By contrast, no direct
connection was highlighted between productivity and total
number of acidic sites. These findings confirm that the
dehydration reaction of xylose over heterogeneous
catalysts is ruled by subtle combinations of their acidic
properties.[150] Similar conclusions were previously reported
using batch setups, wherein higher LAS/BAS ratios resulted
Figure 12. Continuous flow catalytic dehydration reaction of xylose over
NbTiO-MNL2 monolith (t 72 s, 130 °C, xylose 0.02 M in H2O : GVL = 1:9 v/v,
0.3 mL min-1, 276 mg catalyst). Effect of time-on-stream. Drop lines represent
overnight switch-offs.
Catalyst resistance. The monolithic catalysts were
typically tested for 8 hours continuous flow reaction time.
Negligible catalytic activity and selectivity decreases were
observed up to 4 days usage. Selected data are reported in
Figure 12 for NbTiO-MNL2 catalyst under 130 °C and 72 s
residence time, as representative example. The system was
switched off overnight (due to safety reasons) and restart
the day after under the same conditions. After four days and
29 hours overall time-on-stream, conversion and selectivity
observed were 95 ± 2% and 20 ± 1 %, respectively, with >
98 % retention of the starting catalytic efficiency. In no
circumstance niobium or titanium above the detection limit
of ICP-OES were detected in the reaction solution
recovered. These results indicate the stability of the
monolithic catalysts under the reaction conditions adopted,
as well as their resistance to pore blocking by coke and
humins, as usually observed in the dehydration reaction of
carbohydrates.[152] Despite no catalyst regeneration was
required within the above reaction time span, the starting
white colour of the catalysts could be restored by
calcination at 350 °C for 2 hours, that indicates the
possibility to remove the adsorbed humins.
in higher catalytic activity and in
a lower selectivity,
respectively, however using inhomogeneous acid series
(i.e. Zr or Nb phosphate, sulfated zirconia, SiO2-Al2O3,
WOX/ZrO2, g-Al2O3, HY or Hb zeolites).[151]
Several heterogeneous catalysts have been reported for
the dehydration reaction of xylose to furfural under batch
conditions, including conventional TiO2,[153] Nb2O5,[154] solid-
Figure 11. Relationship between LAS / BAS ratio and space-time-yield for the
continuous flow dehydration reaction of xylose to furfural using monolithic
catalysts under the same flow conditions (t = 106 s, 130 °C). Data from Table
5 and 6.
supported
zirconia)[155,156,157] and titanoniobate nanosheets.[158]
detailed discussion of these catalysts is outside the scope
Nb2O5
powders
(silica,
alumina,
A
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
of the present manuscript and comprehensive reviews can
be found in recent papers.[119,134,159,160] Nonetheless, it must
be noticed that the large majority of these systems use
biphasic solvent conditions (e.g. water-toluene / methyl
isobutyl ketone / butanol) or harmful organic solvents (DMF,
DMSO). Monophasic aqueous systems (e.g. water, water-
GVL) require mild (110-150 °C) to harsh conditions (170-
250 °C) to achieve poor (10-14 %) and moderate to good
(29-62%) yields of furfural, respectively. Best results were
obtained using zeolite beta nanocrystals (77% @ 170 °C,
water)[161] or H-mordenite (81% @ 175 °C, 10% wt H2O in
GVL) catalysts.[162]
parameter in determining the catalytic performance. A non-
optimized furfural yield of 42% under 140 °C reaction
temperature and 141 s residence time was obtained using
the catalyst with the higher LAS/BAS ratio without
separation / purification steps.
Unprecedented interception of the 2,5-anhydroxylose
intermediate suggests the reaction to occur via a cyclic
dehydration mechanism with a significant contribution of
xylose-xylulose isomerisation. No catalyst deactivation was
observed over four days usage, demonstrating
a
remarkable stability of the catalyst in the aqueous phase
and its resistance to poisoning by side products.
To the best of our knowledge, only one continuous flow
system has been previously reported, based on modified
tantalum oxide, for the catalytic conversion of xylose to
furfural.[163] The conversion and selectivity trends with
reaction temperatures and residence times were similar to
those herein described. An optimized furfural yield of 59%
at 96% xylose conversion was obtained under 180 °C and
3600 s residence time, using a H2O-butanol 1:1.5 (v/v)
biphasic solvent mixture and 20 bar N2 backpressure. No
details of the by-products observed were given.
The performance of the catalyst herein described
compares favourably with that of the heterogeneous
systems reported in the literature, whereas the dehydration
reaction is usually carried out in batch using biphasic / toxic
solvents (toluene, DMF, butanol), higher temperatures, or in
the homogenous phase using mineral acids.
These results may help in the design and in the
development of more sustainable catalytic processes for the
production of useful platform chemicals and fuels from
renewable biomass.
The monolithic reactor herein described compares
favourably with the performances reported for batch
catalysts in the low temperature range, with the benefits of
continuous flow operations, use of monophasic, safe
reaction solvent and short reaction times.
Experimental
General information
Conclusions
Hierarchical meso / macroporous titania monoliths (TiO2-MNL) were
prepared as previously described.[46] Details of experimental
methods are reported in the Supporting Information.
The design of sustainable processes for the conversion of
biomass to useful chemicals has been identified as a
priority in the biorefinery industry.[114,164] To this aim, the
development of heterogeneous acid catalysts featuring high
resistance to water and efficiency under mild reaction
conditions is deemed critical.
Synthesis of hierarchically porous niobia-titania
monoliths
We have reported the synthesis and characterization of
unconventional, mixed niobia-titania monolithic catalysts
and their application to the dehydration reaction of xylose to
furfural under continuous flow conditions. Materials with up
to 2% wt niobium content feature a well-defined and
isotropic, hierarchically porous network of interconnected
mesopores and macropores. The results indicate that, at
this Nb loading, Ti4+ ions are substituted by Nb5+ ions in the
TiO2 anatase lattice. The Lewis / Brønsted acid sites ratio
shows to increase with increasing Nb content. The textural
properties of the monoliths, while granting effective
interaction between reactants and active sites, allow to
minimize mass transfer limitations and pore blocking typical
of mesoporous materials, thus resulting in no significant
back pressure evolution under a continuous flow of a fluidic
phase.
The catalytic dehydration reaction of xylose to furfural
was conveniently achieved under mild reaction conditions,
using H2O:GVL as safe, single mobile phase, wherein
adoption of continuous flow arrangement represents a
significant process intensification. Addition of as low as 2%
wt niobia to the titania lattice showed to dramatically
improve the efficiency of the catalyst. Use of a uniform
series of materials sharing the same textural properties,
allowed to identify the LAS/BAS ratio as the critical
Meso-/macroporous mixed niobia-titania monoliths (NbTiO-MNL)
were prepared as follows. All the reagents were cooled down to 0
°C prior of use and maintained at that temperature using an ice bath
during the whole procedure. In a typical synthesis, two separate
solutions were prepared: A) a selected amount of NbCl5 was
dissolved in 1.14
g of 37% HCl (11.6 mmol); B) 0.38g of
polyethylene oxide (PEO, 0.039 mmol) were dissolved in 1.72 g of
distilled water (135.5 mmol). The solutions A and B were added in
the order at 0 °C to Ti(OPr)4 (TPO, 6.96 g, 24.0 mmol) under
vigorous stirring. The mixture was stirred for 2-3 minutes. N-
Methylformamide (NMFA, 1.09 g, 18.3 mmol) was added dropwise
with stirring until a clear solution was obtained. The molar ratio
composition of the mixture was: 1 Ti / x Nb /0.48 HCl / 5.64 H2O /
0.76 NMFA / 0.0016 PEO, with x = 0.0 - 0.01 - 0.02. The mixture
was then poured into a polypropylene tube (inner diameter 6-7 mm,
length 4 cm), which was sealed and kept at 40 ºC during 24 h for
gelation and phase separation. The resulting gel was aged at 60 ºC
for 24 h (heating ramp = 1°C/min). The wet monolith obtained was
removed from the mould and washed with 1-propanol (5 x 25 mL,
room temperature), before being dried at room temperature for 24 h
and at 40 ºC for 1 week. The final white monolith was obtained after
calcination of the solid at 350 ºC for 5 h (heating ramp = 0.5
ºC/min).
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
[5]
[6]
[7]
[8]
K. Nakanishi, in Hierarchically Structured Porous Materials, B. Su, C.
Sanchez, X., Yang, Eds.; Wiley-VCH, Weinheim, 2012, Chapter 8.
C. M. A. Parlett, K.Wilson A. F. Lee, Chem. Soc. Rev. 2013, 42, 3876-
3893.
Catalytic reactions in continuous flow mode
Catalytic reactions under continuous flow conditions were carried
out in air atmosphere using a home-made, metal-free continuous-
flow reactor system (Scheme S1). The reaction temperature was
regulated using an oven comprising a Kapton® flexible heating
element and a temperature controller accurate ± 0.1 °C. In a typical
experiment, a sample of monolithic catalyst (5 mm diameter, 7 mm
A. Feinle, M. S. Elsaesser, N. Hüsing, Chem. Soc. Rev. 2016, 45,
3377-3399.
A. Galarneau, A. Sachse, B. Said, C. H. Pelisson, P. Boscaro, N. Brun,
L. Courtheoux, N. Olivi-Tran, B. Coasne, F. Fajula, C. R. Chimie 2016,
19, 231-247.
[9]
A. Galarneau, Z. Abid, B. Said, Y. Didi, K. Szymanska, A. Jarzebski, F.
Tancret, H. Hamaizi, A. Bengueddach, F. Di Renzo, F. Fajula
Inorganics 2016, 4, 9-34.
length, 114 mg, volume 137
µL), was cladded with a heat-
shrinkable PTFE tube and connected to the continuous flow system
via 1/16” PEEK tubing. A solution of xylose in H2O : GVL = 1 : 9
(v/v) was allowed to flow through the monolith. Residence times
were in the range 70 - 224 s, corresponding to 0.1 - 0.4 mL min-1
substrate solution flow rate. The attainment of steady state
conditions (ca. 0.5 h) was taken as the reaction start time. At the
outlet of the reactor, the reaction solution was cooled down to room
temperature and periodically analyzed for conversion and selectivity
by GC, GC-MS and HPLC. Aliquots were sampled at 1 h intervals
for Nb and Ti leaching determination by ICP-OES. The reaction
products were unequivocally identified by comparison of the GC
and HPLC retention times and mass spectra with those of authentic
specimens. Quantitative analyses were carried out via HPLC and
GC based on calibration curves of the pure compounds.
[10] M. Guisnet, M. Guidotti, Microporous and Mesoporous Solid Catalysts,
in Catalysts for Fine Chemical Synthesis, Vol. 4, E. G. Derouane, Ed.;
John Wiley & Sons, Weinheim, 2006, Chapter 2.
[11] B. Louis, G. Laugel, P. Pale, M. Maciel Pereira ChemCatChem 2011, 3,
1263-1272
[12] H. Minakuchi, K. Nakanishi, N. Soga, N. Ishizuka, N.Tanaka, J.
Chromatogr. A 1997, 762, 135-146.
[13] S. Whitaker, Transport Porous Med. 1986, 1, 3-25.
[14] C. Wiles, P. Watts, Green Chem. 2014, 16, 55-62.
[15] R. M. Myers, D. E. Fitzpatrick, R. M. Turner, S. V. Ley, Chem. Eur. J.
2014, 20, 12348-12366.
[16] N. G. Anderson, Org. Process Res. Dev. 2012, 16, 852-869.
[17] A. Stankiewicz, Chem. Eng. Sci. 2001, 56, 359-364.
[18] M. R. Buchmeiser, Polymer 2007, 48, 2187-2198.
[19] M. Baumann, I. R. Baxendale, S. V. Ley, N. Nikbin, C.D. Smith, Org.
Biomol. Chem. 2008, 6, 1587-1593.
[20] F. Svec, T.B. Tennikova, Z. Deyl, Monolithic Materials: Preparation,
Properties, and Applications, Elsevier, Amsterdam, 2003.
[21] S. Xie, R.W. Allington, J.M. Fréchet, F. Svec in Advances in
Biochemical Engineering/Biotechnology, R. Freitag, Ed., Springer,
London, 2002.
Acknowledgements
Thanks are due to: Dr. Samuele Ciattini (Centro di
Cristallografia University of Firenze, Italy) for X-ray
tomography
measurements,
Centro
Microscopie
[22] K. Nakanishi, H. Shikata, N. Ishizuka, N. Koheiya, N. Soga, J. High
Resol. Chromatogr. 2000, 23, 106-110.
Elettroniche (CNR, Firenze, Italy) for technical assistance,
Dr. Andreana Piancastelli (ISTEC - CNR, Faenza, Italy) for
Hg intrusion measurements, Dr. A. Ienco (ICCOM-CNR) for
XRD discussion, C. Almansa (SSTTI Universidad de
Alicante) for TEM analysis. F.B. thank financial support from
EIT Raw Materials through project FREECATS (Project no.
15054).
[23] B. Gaweł, K.Gaweł, G. Øye, Materials 2010, 3, 2815-2833.
[24] J. Babin, J. Iapichella, B. Lefèvre, C. Biolley, J. P. Bellat, F. Fajula, A.
Galarneau, New J. Chem. 2007, 31, 1907-1917.
[25] T. Amatani, K. Nakanishi, K. Hirao, T. Kodaira, Chem. Mater. 2005, 17,
2114-2119.
[26] X. Guo, R. Wang, H. Yu, Y. Zhu, K. Nakanishi, K. Kanamori, H. Yang,
Dalton Trans. 2015, 44, 13592-13601.
[27] S. Hartmann, M. S. Elsaesser, N. Hüsing, Z. Anorg. Allg. Chem. 2014,
640, 641-648.
[28] S. Hartmann, A. Sachse, A. Galarneau, Materials 2012, 5, 336-349.
[29] B. Gaweł, G. Øye, Mater. Lett. 2013, 95, 86-88.
[30] B. Gaweł, K. Gaweł, T. C. Hobæk, M. Yasuda, G. Øye, Mater. Chem.
Phys. 2012, 137, 414-420.
Conflicts of interest
The authors declare no conflict of interest.
[31] G. Hasegawa, K. Kanamori, K. Nakanishi, T. Hanada, J. Am. Ceram.
Soc. 2010, 93 3110-3115.
Keywords: sustainable chemistry • heterogeneous catalysis •
[32] J. Konishi, K. Fujita, K.Nakanishi, K. Hirao, Chem. Mater. 2006, 18,
6069-6074.
biorenewable resources • continuous flow • xylose
[33] Y. Zhu, T. Shimizu, T. Kitajima, K. Morisato, N.Moitra, N.Brun, T.
Kiyomura, K. Kanamori, K. Takeda, H. Kurata, M.Tafu, K. Nakanishi,
New J. Chem. 2015, 39, 2444-2450.
Abbreviations: HPLC, high-performance liquid chromatography, ICP-OES,
Inductively Coupled Plasma Atomic Emission Spectroscopy; BET, Brunauer–
Emmett–Teller model; ESEM, Environmental Scanning Electron Microscopy;
EDS, Energy Dispersive X-ray Spectrometry; TEM, Transmission Electron
Microscopy; XRD, X-ray diffraction; XPS, X-ray Photoelectron spectroscopy;
[34] X. Cai, W. Zhu, H. Yang, C. Xu, K. Nakanishi, K. Kanamori, X. Guo,
New J. Chem. 2015, 39, 6238-6243.
[35] N. Moitra, K. Kanamori, T. Shimada, K. Takeda, Y. H. Ikuhara, X. Gao,
K. Nakanishi, Adv. Funct. Mater. 2013, 23, 2714-2722.
[36] N. Moitra, S. Fukumoto, J. Reboul, K. Sumida, Y. Zhu, K. Nakanishi, S.
Furukawa, S. Kitagawa, K. Kanamori, Chem. Commun. 2015, 51, 3511-
3514.
GVL,
g-valerolactone;
GC, gas
chromatography; GC-MS,
gas
chromatography-mass spectrometry; STY, space-time-yield.
[1]
[2]
R. J. Jones, Definitions of Terms Relating to the Structural and
Processing of Sols, Gels, Networks and Inorganic–Organic Hybrid
Materials. IUPAC Recommendations. Pure Appl. Chem. 2007, 79,
1801-1829.
[37] J. Wegner, S. Ceylan, A. Kirschning, Chem. Commun. 2011, 47, 4583-
4592.
[38] R. Ricciardi, J. Huskens, W. Verboom, ChemSusChem 2015, 8, 2586-
2605.
A. Cybulski, J. A. Moulijn, Structured Catalysts and Reactors, Marcel
Dekker Ltd.: New York, 2005, p 19.
[39] K. Nakanishi, J. Porous Mat. 1997, 4, 67-112.
[40] A. Sachse, A. Galarneau, B. Coq, F. Fajula, New J. Chem. 2011, 35,
259-264.
[3]
[4]
R. M. Heck, S. Gulati, R. J.Farrauto, Chem. Eng. J. 2001, 82, 149-156.
T. Vesna, J. Franjo, Appl. Catal., A 2006, 311, 112-121.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
[41] H. Yu, Y. Zhu, H. Yang, K. Nakanishi, K. Kanamori, X. Guo, Dalton
Trans. 2014, 43, 12648-12656.
[82] S. Hasegawa, H. Aritani, M. Kudo, Catal. Today 1993, 16, 371-377.
[83] A. Kubacka, G. Colón, M. Fernández-García Catal. Today 2009, 143,
286-292.
[42] P. He, S. J. Haswell, P. D. I. Fletcher, S. M. Kelly, A. Mansfield,
Beilstein J. Org. Chem. 2011, 7, 1150-1157.
[84] P. Viparelli, P. Ciambelli, J. C. Volta, J. M. Herrmann, Appl. Catal. A:
Gen. 1999, 182, 165-173.
[43] C. H. Pélisson, T. Nakanishi, Y. Zhu, K. Morisato, T. Kamei, A. Maeno,
H. Kaji, S. Muroyama, M. Tafu, K. Kanamori, T. Shimada, K. Nakanishi,
ACS Appl. Mater. Interfaces 2017, 9, 406-412.
[85] M. Franco Finol, J. Rooke, S. Siffert, R. Cousin, P. Carniti, A. Gervasini,
J. M. Giraudon, B. L. Su, J. F. Lamonier, New J. Chem. 2014, 38, 1988-
1995.
[44] A. Sachse, V. Hulea, A. Finiels, B. Coq, F. Fajula, A. Galarneau J.
Catal. 2012, 287, 62-67.
[86] T. Onfroy, O. V. Manoilova, S. B. Bukallah, D. M. Hercules, G. Clet, M.
Houalla, Appl. Catal. A: Gen. 2007, 316, 184-190.
[45] Y. Zhu, K. Kanamori, N. Brun, C. H. Pélisson, N. Moitra, F. Fajula, V.
Hulea, A. Galarneau, K. Takeda, K. Nakanishi, Catal. Commun. 2016,
87, 112-115
[87] J. C. Rooke, T. Barakat, M. Franco Finol, P. Billemont, G. De Weireld,
Y. Lie, R. Cousin, J. M. Giraudon, S. Siffert, J. F. Lamonier, B.L. Su
Appl. Catal. B: Environ. 2013, 142-143, 149-160.
[46] N. Linares, S. Hartmann, A. Galarneau, P. Barbaro, ACS Catal. 2012,
2, 2194-2198.
[88] B. M. Matsagar, S. A. Hossain, T. Islam, H. R. Alamri, Z. A. Alothman,
Y. Yamauchi, P. L. Dhepe, K. C. W. Wu, SciRep 2017, 7, 13508-13513.
[89] B. M. Matsagar, P. L. Dhepe, New J. Chem. 2017, 41, 6137-6144.
[90] J. J. Bozell, G. R. Petersen, Green Chem. 2010, 12, 539-554.
[91] J. Konishi, K. Fujita, K. Nakanishi, K. Hirao, K. Morisato, S. Miyazaki,
M. Ohira, J. Chromatogr. A 2009, 1216, 7375-7383.
[47] A. Sachse, N. Linares, P. Barbaro, F. Fajula, A. Galarneau, Dalton
Trans. 2013, 42, 1378-1384.
[48] A. El Kadib, R. Chimenton, A. Sachse, F. Fajula, A.Galarneau, B. Coq
Angew. Chem. Int. Ed. 2009, 48, 4969-4972.
[49] C. P. Haas, T. Müllner, R. Kohns, D. Enkeb, U. Tallarek, React. Chem.
Eng. 2017, 2, 498-511.
[92] K. Tanabe, Catal. Today 1990, 8, 1-11.
[50] A. Sachse, A. Galarneau, F. Fajula, F. Di Renzo, P. Creux, B. Coq,
Micropor. Mesopor. Mater. 2011, 140, 58-68.
[93] T. Iizuka, K. Ogasawara, K. Tanabe, Bull. Chem. Soc. Jpn. 1983, 56,
2927-2931.
[51] A. Sachse, R. Ameloot, B. Coq, F. Fajula, B. Coasne, D. De Vos, A.
Galarneau, Chem. Commun. 2012, 48, 4749-4751.
[94] Y. Zhu, Y. Morimoto, T. Shimizu, K. Morisato, K. Takeda, K. Kanamori,
K. Nakanishi, J. Ceram. Soc. Jpn. 2015, 123, 770-778.
[95] J. Konishi, K. Fujita, K. Nakanishi, K. Hirao, Chem. Mater. 2006, 18, 864-
866.
[52] F. Liguori, P. Barbaro, B. Said, A. Galarneau, V. Dal Santo, E.
Passaglia, A. Feis, ChemCatChem 2017, 9, 3245-3258.
[53] A. J. Grano, F. M. Sayler, J. H. Smått, M. G.Bakker, J. Porous Mater.
2014, 21, 1113-1122.
[96] Y. Liu, J. M. Szeifert, J. M. Feckl, B. Mandlmeier, J. Rathousky, O.
Hayden, D. Fattakhova-Rohlfing, T. Bein, ACS Nano 2010, 4, 5373-
5381.
[54] A. Corma, Chem. Rev. 1995, 95, 559-614.
[55] G. Busca, Chem. Rev. 2007, 107, 5366-5410.
[56] P. Gupta, S. Paul, Catal. Today 2014, 236, 153-170.
[57] K. Shimizu, A. Satsuma, Energy Environ. Sci. 2011, 4, 3140-3153.
[58] R. A. Sheldon, Green Chem. 2014, 16, 950-963.
[59] A. Koreniuk, K. Maresz, K. Odrozek, A. B. Jarzębski, J.Mrowiec-Białoń,
Appl. Catal. A: Gen. 2015, 489, 203-208.
[97] X. Lu, X. Mou, J. Wu, D. Zhang, L. Zhang, F. Huang, F. Xu, S. Huang,
Adv. Funct. Mater. 2010, 20, 509-515.
[98] N. Özer, D. G. Chen, C. M. Lampert, Thin Solid Films 1996, 277, 162-
168.
[99] G. M. Ingo, S. Dirè, F. Babonneau, Appl. Surf. Sci. 1993, 70, 230-234.
[100] L. Sheppard, T. Bak, J. Nowotny, C.C. Sorrell, S. Kumar, A. R. Gerson,
M. C. Barnes, C. Ball, Thin Solid Films 2006, 510, 119-124.
[101] T. Bak, J. Nowotny, M. Rekas, C.C. Sorrell, J. Phys. Chem. Solids
2003, 64, 1057-1067.
[60] X. Chen, S. S. Mao, Chem. Rev. 2007, 107, 2891-2959.
[61] U. Diebold, Surf. Sci. Rep. 2003, 48, 53-229.
[62] K. Tanabe, Catal. Today 2003, 78, 65-77.
[63] I. Nowak, M. Ziolek, Chem. Rev. 1999, 99, 3603-3624.
[64] T. Okuhara, Chem. Rev. 2002, 102, 3641-3666.
[65] M. Ziolek, Catal. Today 2003, 78, 47-64.
[102] V. León, Surf. Sci. 1995, 339, L931–L934.
[103] M. Tamura, K. Shimizu, A. Satsuma, Appl. Catal. A: Gen. 2012, 433,
135-145.
[66] Y. Zhao, X. Zhou, L. Ye, S. C. E. Tsang, Nano Reviews 2012, 3, 17631.
[67] P. Carniti, A. Gervasini, S. Biella, A. Auroux, Catal. Today 2006, 118, 373-
378.
[104] C. A. Emeis, J. Catal. 1993, 141, 347-354.
[105] A. Chareonlimkun, V. Champreda, A. Shotipruk, N. Laosiripojana,
Bioresour. Technol. 2010, 101, 4179-4186.
[68] A. Lewera, L. Timperman, A. Roguska, N. Alonso-Vante, J. Phys.
Chem. C 2011, 115, 20153-20159
[106] S. Dutta, S. De, A. K. Patra, M. Sasidharan, A. Bhaumik, B. Saha, Appl.
Catal. A 2011, 409-410, 133-139.
[69] E. I. Ko, R. Bafrali, N. T. Nuhfer, N. J. Wagner, J. Catal. 1985, 95, 260-
270.
[107] K. Nakajima, Y. Baba, R. Noma, M. Kitano, J. N. Kondo, S. Hayashi, M.
Hara, J. Am. Chem. Soc. 2011, 133, 4224-4227.
[70] T. Beutel, V. Siborov, B. Tesche, H. Knözinger, J. Catal. 1997, 167,
379-390.
[108] J. M. Jehng, I. E. Wachs, J. Phys. Chem. 1991, 95, 7373-7379.
[109] A. Gervasini, P. Carniti, M. Marzo, A. Auroux, J. Catal. 2012, 296, 143-
155.
[71] T. Hoffer, S. Dobos, L. Guczi, Catal. Today 1993, 16, 435-446.
[72] Materials for Energy Conversion Devices, C. C. Sorrell, J. Nowotny, S.
Sugihara, Eds., CRC Press, Boca Raton, 2005.
[110] K. Nakajima, J. Hirata, M. Kim, N. K. Gupta, T. Murayama, A. Yoshida, N.
Hiyoshi, A. Fukuoka, W. Ueda, ACS Catal. 2018, 8, 283-290.
[111] J. Datka, A. M. Turek, J. M. Jehng, I.E. Wachs, J. Catal. 1992, 135,
186-199.
[73] A. L. Linsebigler, G. Lu, J. T. Jr. Yates, Chem. Rev. 1995, 95, 735-758.
[74] R. M. Pittman, A. T. Bell, Catal. Lett. 1994, 24, 1-13.
[75] M. Franco Finol, J. Rooke, B. L. Su, M. Trentesaux, J. M. Giraudon, J.
F. Lamonier, Catal. Today 2012, 192, 154-159.
[112] R. Luque, Curr. Opin. Green Sustain. Chem. 2016, 2, 6-9.
[113] M. J. Climent, A. Corma, S. Iborra, Green Chem. 2014, 16, 516-547.
[114] B. M. Matsagar, M. K. Munshi, A. A. Kelkar, P. L. Dhepe, Catal. Sci.
Technol. 2015, 5, 5086-5090.
[76] Niobium-doped titanates have also been described, see e.g.: E. Wada,
M. Kitano, K. Yamamoto, K. Nakajima, S. Hayashi, M. Hara, Catal. Sci.
Technol. 2016, 6, 4832–4839.
[115] A.S. Mamman, J.M. Lee, Y.C. Kim, I.T. Hwang, N.J. Park, Y.K. Hwang,
J.S. Chang, J. S. Hwang, Biofuels Bioprod. Biorefin. 2008, 2, 438-454.
[116] K J. Zeitsch, The chemistry and technology of furfural and its many by-
products. Volume 13. Amsterdam, Elsevier, 2000.
[77] M.Z. Atashbar, H.T. Sun, B. Gong, W. Wlodarski, R. Lamb Thin Solid
Films 1998, 326, 238-244.
[78] A. Mattsson, M. Leideborg, K. Larsson, G. Westin, L. Österlund, J.
Phys. Chem. B 2006, 110, 1210-1220.
[117] C. M. Cai, T. Zhang, R. Kumar, C. E. Wyman, J. Chem. Technol.
Biotechnol. 2014, 89, 2-10.
[79] D. Morris, Y. Dou, J. Rebane, C.E.J. Mitchell, R.G. Egdell, D.S.L. Law,
A. Vitadini, M. Casarin, Phys. Rev. B 2000, 61, 13445-13457.
[80] Y. Wang, B.M. Smarsly, I. Djerdj, Chem. Mater. 2010, 22, 6624-6631.
[81] W. Zeng, T. Liu, Z. Wang, Sensor. Actuat. B: Chem. 2012, 166-167,
141-149.
[118] R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sádaba, M. López
Granados Energy Environ. Sci. 2016, 9, 1144-1189.
[119] K. Yan, G. Wu, T. Lafleur, C. Jarvis, Renew. Sust. Energ. Rev. 2014,
38, 663-676.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801414
ChemSusChem
FULL PAPER
[120] B. Danon, G. Marcotullio, W. de Jong Green Chem. 2014, 16, 39-54.
[121] K. R. Enslow, A. T. Bell, ChemCatChem 2015, 7, 479-489.
[122] M. B. Fusaro, V. Chagnault, D. Postel, Carbohydr. Res. 2015, 409, 9-
19.
[156] C. García-Sancho, I. Agirrezabal-Telleria, M. B. Güemez, P. Maireles-
Torres, Appl. Catal. B: Environ. 2014, 152-153, 1-10
[157] I. Agirrezabal-Telleria, C. García-Sancho, P. Maireles-Torres, P. L.
Arias, Chinese J. Catal. 2013, 34, 1402-1406.
[123] V. Choudhary, S. I. Sandler, D. G. Vlachos, ACS Catal. 2012, 2, 2022-
2028.
[158] A.S. Dias, S. Lima, D. Carriazo, V. Rives, M. Pillinger, A. A. Valente, J.
Catal. 2006, 244, 230-237.
[124] M. S. Holm, Y. J. Pagán-Torres, S. Saravanamurugan, A. Riisager, J.
A. Dumesic, E. Taarning, Green Chem. 2012, 14, 702-706.
[125] K. Lamminpää, J. Ahola, J. Tanskanen, RSC Adv. 2014, 4, 60243-
60248.
[159] X. Li, P. Jia, T. Wang, ACS Catal. 2016, 6, 7621-7640.
[160] I. Agirrezabal-Telleria, I. Gandarias, P.L. Arias, Catal. Today 2014, 234,
42-58.
[161] L. R. Ferreira, S. Lima, P. Neves, M. M. Antunes, S. M. Rocha, M.
Pillinger, I. Portugal, A. A. Valente, Chem. Eng. J. 2013, 215-216, 772-
783.
[126] L. Filiciotto, A. M. Balu, J. C. Van der Waal, R. Luque, Catal. Today
2018, 302, 2-15.
[127] S. M. Walas, Chemical Process Equipment. Selection and Design.
Butterworth-Heinemann, Washington, 1990, Chapter 17.
[128] J. N. Chheda, Y. Román-Leshkov, J. A. Dumesic, Green Chem. 2007,
9, 342-350.
[162] E. I. Gürbüz, J. M. R. Gallo, D. M. Alonso, S. G. Wettstein, W. Y. Lim, J.
A. Dumesic, Angew. Chem. Int. Ed. 2013, 52, 1270-1274.
[163] X. Li, L. T. Pan, J. Deng, Y. Fu, H. J. Xu, RSC Adv. 2015, 5, 70139-
70146.
[129] A.S. Dias, M. Pillinger, A. A. Valente, Appl. Catal. A: Gen. 2005, 285,
126-131.
[164] A Roadmap for moving to a competitive low carbon economy in 2050,
European Commission, COM(2011) 112 final.
[130] F. Liguori, C. Moreno-Marrodan, P. Barbaro, ACS Catal. 2015, 5, 1882-
1894.
[131] I. T. Horváth, H. Mehdi, V. Fabos, L. Boda, L. T. Mika, Green Chem.
2008, 10, 238-242.
[132] M. A. Mellmer, C. Sener, J. M. R. Gallo, J. S. Luterbacher, D. M.
Alonso, J. A. Dumesic, Angew. Chem. Int. Ed. 2014, 53, 11872-11875.
[133] X. Hu, C. Lievens, C. Z. Li, ChemSusChem 2012, 5, 1427-1434.
[134] X. Zhang, K. Wilson, A. F. Lee, Chem. Rev. 2016, 116, 12328-12368.
[135] A.S. Dias, S. Lima, M. Pillinger, A. A. Valente, in Ideas in Chemistry
and MolecularSciences: Advances in Synthetic Chemistry, B.
Pignataro, Ed., Wiley-VCH, Weinheim, 2010, Chapter 8.
[136] K. Dussan, B. Girisuta, M. Lopes, J. J. Leahy, M. H. B. Hayes,
ChemSusChem 2015, 8, 1411-1428.
[137] W. Zhang, Y. Zhu, S. Niu, Y. Li, J. Mol. Cat. A: Chemical 2011, 335, 71-
81.
[138] M. J. Jr. Antal, T. Leesomboon, W. S. Mok, G. N. Richards, Carbohyd.
Res. 1991, 217, 71-85.
[139] P. Barbaro, F. Liguori, C. Moreno-Marrodan, Green Chem. 2016, 18,
2935-2940.
[140] J. U. Oltmanns, S. Palkovits, R. Palkovits, Appl. Catal. A: Gen. 2013,
456, 168-173.
[141] S. Verma, R. B. N. Baig, M. N. Nadagouda, C. Len, R. S. Varma, Green
Chem. 2017, 19, 164-168.
[142] L. Hu, G. Zhao, W. Hao, X. Tang, Y. Sun, L. Lin, S. Liu, RSC Adv.
2012, 2, 11184-11206.
[143] M. R. Nimlos, X. Qian, M. Davis, M. E. Himmel, D. K. Johnson, J. Phys.
Chem. A 2006, 110, 11824-11838.
[144] T. Ahmad, L. Kenne, K. Olsson, O. Theander, Carbohydr. Res. 1995,
276, 309-320.
[145] Y. Yang, C. W. Hu, M. M. Abu-Omar, ChemSusChem 2012, 5, 405-
410.
[146] V. Choudhary, A. B. Pinar, S. I. Sandler, D. G. Vlachos, R. F. Lobo,
ACS Catal. 2011, 1, 1724-1728.
[147] O. Ershova, J. Kanervo, S. Hellstena, H. Sixta, RSC Adv. 2015, 5, 66727-
66737.
[148] K. Stensrud, (Archer Daniels Midland Company). Patent No.
WO2017/30684 A1, 2017.
[149] M. H. L. Silveira, B. A. Vanelli, M. L. Corazza, L. Pereira Ramos,
Bioresource Technol. 2015, 192, 389-396.
[150] R. Weingarten, G. A. Tompsett, W. C. Jr. Conner, G. W Huber, J. Catal.
2011, 279, 174-182.
[151] B. Pholjaroen, N. Li, Z. Wang, A. Wang, T. Zhang, J. Energy Chem.
2013, 22, 826-832.
[152] S. Lima, M. Pillinger, A. A. Valente, Catal. Commun. 2008, 9, 2144-
2148.
[153] N. K. Gupta, A. Fukuoka, K. Nakajima, ACS Catal. 2017, 7, 2430-2436.
[154] C. García-Sancho, J. M. Rubio-Caballero, J. M. Mérida-Robles, R.
Moreno-Tost, J. Santamaría-González, P. Maireles-Torres, Catal.
Today 2014, 234, 119-124.
[155] M. J. Campos Molina, M. López Granados, A. Gervasini, P. Carniti,
Catal. Today 2015, 254, 90-98.
This article is protected by copyright. All rights reserved.