ZrO2-SBA-15 catalysts for the one-pot cascade synthesis of GVL from furfural

Article abstract of DOI:10.1039/c8cy01121d

Controlling the thickness of zirconia monolayers coated over SBA-15 offers an effective way to tune catalytic performance for the acid-mediated and hydrogen transfer (Meerwein Ponndorf Verley, MPV) cascade transformation of furfural to γ-valerolactone. Co

Full text of DOI:10.1039/c8cy01121d

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Iglesias, J. A.
Melero, G. Morales, M. Paniagua, B. Hernandez, A. Osatiashtiani, A. F. Lee and K. Wilson, Catal. Sci.
Technol., 2018, DOI: 10.1039/C8CY01121D.
This is an Accepted Manuscript, which has been through the
Volume 6 Number 1 7 January 2016 Pages 1–308
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
Accepted Manuscripts are published online shortly after
www.rsc.org/catalysis
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
ISSN 2044-4753
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
PAPER
Qingzhu Zhang et al.
Catalytic mechanism of C–F bond cleavage: insights from QM/MM
analysis of fluoroacetate dehalogenase
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/catalysis

Please do not adjust margins
Catalysis Science & Technology
Page 1 of 11
View Article Online
DOI: 10.1039/C8CY01121D
Journal Name
ARTICLE
ZrO2-SBA-15 catalysts for the one-pot cascade synthesis of GVL
from furfural
J. Iglesias,*a J. A. Melero,a G. Morales,a M. Paniagua,a B. Hernández,a A. Osatiashtiani,b A. F. Leec
and Karen Wilsonc
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
Controlling the thickness of zirconia monolayers coated over SBA-15 offers an effective way to tune catalytic performance
for the acid-mediated and hydrogen transfer (Meerwein Ponndorf Verley, MPV) cascade transformation of furfural to γ-
valerolactone. Complementary mechanistic and kinetic modelling establishes the existence of the two distinct zirconium
active species (weak and strong acid sites), whose balancing enables optimisation of the cascade and hence maximal γ-
www.rsc.org/
valerolactone
(GVL)
production.
collection of bioproducts, ranging from green fuels to bio-
polymers.2 GVL can be synthesized from furfural through a
catalytic cascade of reactions (Scheme 1),3 notably through the
combination of acid and hydrogenation catalysts.4-5 However,
the use of molecular H2 necessitates high operating pressures
(typically >30 bar) and precious metal catalysts. Catalytic
transfer hydrogenation by a Meerwein Ponndorf Verley (MPV)
reaction, therefore represents a promising alternative route to
reduce levulinic acid in GVL production.6-8 Roman-Leshkov and
co-workers reported GVL production from furfural by the
combination of discrete Lewis and Brønsted catalysts (Zr-beta
and Al-MFI zeolites) in the presence of isobutanol as a
hydrogen donor.9 Bifunctional Sn-Al and Zr-Al β-zeolite
catalysts possessing both Lewis and Brønsted sites10-14 were
subsequently demonstrated to be active for the same reaction.
Despite these promising results, and the simple post-synthesis
protocol used to prepare such bifunctional zeolites,15 their
high manufacturing costs currently prohibits commercial
exploitation for GVL production.
Introduction
Within the context of a bio-based economy, lignocellulosic
biomass is an interesting and sustainable source of raw
materials that provide solutions for the production of many
different products conventionally obtained from crude oil,
including transportation fuels and chemicals. This type of
biomass can be transformed into sugar monosaccharides
(glucose and xylose) that can be subsequently used to produce
bio-fuels and bio-products. In general, such processes require
several discrete chemical transformations, each usually driven
by a specific class of catalyst to achieve the desired bio-
product. Syntheses involving multiple intermediates and
competing side reactions are difficult to commercialise at
industrial scale, and hence there is growing interest in one-pot
catalytic cascades to valorize sugars into platform chemicals.1
This approach unlocks process advantages through the
elimination of multiple separation and product work-up
stages, but is also dependent on the discovery of new
multifunctional catalysts able to promote each step along the
reaction pathway.
R
OH
O
OH
O
R
O
R
O
O
O
O
Furfural is an abundant and low cost bio-derived product of
xylose dehydration (the latter obtained from hemicellulose),
O
H2O
R
OH
R
O
LEV
FE
FOL
FAL
R
OH
and therefore an attractive building block to synthesise γ-
R
O
valerolactone (GVL). GVL is an appealing compound due to its
attractive physicochemical properties (low toxicity and
biodegradability) and unique fuel characteristics. It can be
directly used as food additive, green solvent, liquid
R
OH
O
Acid-driven reactions
MPV reactions
R
O
OH
O
O
GVL
fuel/additive for gasoline, and transformed into
a wide
Scheme 1. Reaction scheme for the production of GVL from furfural by
the combination of Lewis and Brønsted acid catalyst
.
Zirconia is an abundant and low cost catalyst with relatively
high acid and basic loadings,16 properties that make it
attractive for MPV reductions,17,18 a key step in the GVL
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 2 of 11
View Article Online
DOI: 10.1039/C8CY01121D
ARTICLE
Journal Name
production from furfural. However, its naturally occurring neutralizer and Mg Kα X-ray source (hν = 1253.6 eV). Spectra
dense form, zirconia exhibits poor textural properties which were recorded at normal emission with an analyser pass
limit catalytic utility. Nevertheless, ZrO2 coating of porous energy of 20 eV and X-ray power of 225 W. ZrO2 film thickness
silicas (such as mesoporous SBA-15)19 offers a simple way to was calculated from attenuation of the Si 2p signal as
enhance surface area and active site accessibility. In addition, previously described (see data in Table ESI-1). Acid loadings
ZrO2-SBA-15 materials have been reported to be excellent were determined by NH3 temperature-programmed
catalysts for the MPV reduction of alkyl levulinates to GVL.20,21 desorption
Here we take a further step, and demonstrate the application (micromeritics). TPD profiles were fitted to calculate
(TPD)
using
a
Autochem
instrument
a
of ZrO2-SBA-15 materials for the one-pot catalytic maximum NH3 desorption temperature and hence quantify
transformation of furfural into GVL. The combination of acid strength. Base site loading was quantified by means of
excellent textural properties, and ability to tune the solid CO2 pulse titration using
a Quantachrome ChemBET
acidity and H-transfer capability by altering the thickness of instrument. Brønsted/Lewis acid character was calculated
conformal ZrO2 monlayers, enable optimisation of the overall from diffuse reflectance FTIR spectroscopy (DRIFTS) of pyridine
cascade and GVL yields >30 %.
saturated samples under gas and liquid phase conditions.
Infrared spectra were recorded on a Thermo Nicolet iS50 FT-IR
spectrometer equipped with a LN2 cooled MCT detector, and
either Praying Mantis (gas phase) or ATR (liquid) attachments.
DRIFT spectra were recorded on dry catalyst powders diluted
in KBr onto which pyridine was dropped, and then dried in-
vacuo overnight prior to analysis. The ATR liquid flow-through
cell featured a trapezoidal, 10 bounce ZnSe crystal (80 mm×10
mmx4 mm). Spectra were the average of 120 scans recorded
at room temperature and 4 cm−1 resolution, using the
appropriate dried catalyst film for spectral background
subtraction. Slurries of catalyst powders were prepared by
adding 20 mg catalyst to 2 mL isopropanol and subsequent
sonication for 15 min. 1 mL of slurry was then added dropwise
to the ZnSe internal reflection element, and evaporated to
obtain a uniform thin layer of catalyst over the ATR crystal. In-
situ spectra of chemisorbed pyridine in the liquid phase were
recorded by first passing a flow of isopropanol through the cell
to saturate the catalyst surface for 1 h. A 0.1 vol% solution of
pyridine in isopropanol was then introduced at 0.2 ml.min−1
using a syringe pump for 1.5 h, and spectra of adsorbed
pyridine then recorded.
Experimental
Catalyst preparation
ZrO2-SBA-15 was synthesized following our literature
procedure:19 10 g pure SBA-15 silica was dried at 300 °C for 4 h
before refluxing with 58.5g Zr(OPr)4 (70 wt% in isopropanol,
Aldrich) in 300 mL of anhydrous hexane. This method favors
the reaction between surface hydroxyls (silanols) and
zirconium propoxide resulting in a conformal ZrO2 monolayer
encapsulating the SBA-15 surface. Refluxing was performed
overnight, and the resulting solid obtained by filtration then
washed three times with hexane to remove any unreacted
Zr(OPr)4. Samples were subsequently rehydrated in 300 mL of
deionised water under stirring for 4 h to fully hydrolyse
residual propoxide groups, and provide new anchoring points
for additional ZrO2 monolayers. Samples were filtered and
dried at 80 °C overnight, and the coating procedure repeated
to obtain bilayer and trilayer zirconia coated SBA-15. Finally,
samples were calcined at 550 °C in static air for 3 h, and
denoted ZrO2-SBA-15(1), ZrO2-SBA-15(2), and ZrO2-SBA-15(3)
to reflect the number of ZrO2 coatings. A Zr-SBA-15 reference,
containing isolated zirconium sites within the mesostructured
silica matrix, was also prepared through a direct synthesis
procedure,22 using a zirconocene dichloride (ABCR, 99 wt%)
precursor.
Catalytic tests
The one-pot cascade transformation of furfural to GVL was
performed in a 200 mL stainless steel, stirred autoclave fitted
with a temperature controller and a pressure gauge. Reactions
used 3.75 g furfural in a 1:50 molar ratio of furfural (Aldrich,
99 wt%) to isopropanol (Scharlab, 98 wt%), and 2.5:1 mass
ratio of furfural to catalyst. Decane (Acros, 99 wt%) was added
as an internal standard for reaction analysis at a concentration
of 10 g·L-1. Catalysts were thermally treated in air at 450 °C for
5 h under static conditions prior to use. The reaction mixture
was stirred at 1000 rpm, and heated to the desired
temperature (130-170 °C) over a 1 h period. Aliquots of the
reaction mixture were periodically withdrawn during 7 h for
analysis.
Catalyst characterization
Textural properties were assessed by N2 physisorption at 77 K
using a Micromeritics TRISTAR 3000 porosimeter. Total pore
volumes were recorded at P/P0=0.985, and pore size
distributions calculated by the BJH-KJS method. Mesoscopic
ordering was evaluated by low-angle powder X-ray diffraction
(XRD) on a PHILIPS X-PERT diffractometer using the Cu Kα line
in the range 2θ = 0.6-5.0° with a step size of 0.02°. Phase
identification was evaluated by wide-angle XRD in the range
2θ=10−90° with a step size of 0.04°. Bulk zirconium content
was determined by ICP-OES. High-resolution transmission
electron microscopy (HRTEM) was performed on a Philips
TECNAI-20T microscope operated at 200 kV. X-ray
photoelectron spectroscopy (XPS) was performed on a Kratos
Axis HSi photoelectron spectrometer equipped with a charge
Analysis of reaction products
The reaction mixture was analysed by gas chromatography
using a Varian 3900 GC fitted with a CP-WAX column (30
mx0.25 mmx0.25 µm) and FID detector. Calibration curves
were obtained from stock solutions, using decane as an
internal standard, for furfuryl alcohol (FOL, Aldrich 98 wt%)
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 3 of 11
View Article Online
DOI: 10.1039/C8CY01121D
Journal Name
ARTICLE
and γ-valerolactone (GVL, Aldrich 99 wt%). Isopropyl levulinate coated samples (Fig. ESI-1A), although the loss of (110) and
(i-LEV) was prepared according to methods described (200) reflections suggests that surface modification by zirconia
elsewhere,23 and its purity assessed by 1H and 13C NMR. degrades the local ordering of mesopores in hexagonal arrays.
Catalyst performance was evaluated according to the following Surface area and porosity progressively decreased with
equations:
zirconia film thickness, resulting in a reduction in mesopore
diameter and corresponding rise in wall thickness (Fig. ESI-2).
TEM imaging confirmed that zirconia coated materials retained
the hexagonal ordering of parallel mesopore channels present
in the parent SBA-15 (Fig. ESI-3) which SEM imaging shows
extend over hundreds of microns (Fig. ESI-4). No evidence of
bulk ZrO2 domains was observed for any ZrO2-SBA-15
materials, although wide angle XRD reveals the emergence of
ꢄ
ꢃ
ꢂꢈ
ꢂꢃꢄ
Xꢀ ꢁ ꢂ ꢅꢂ ꢆ 100;
Yꢇ ꢁ ꢆ 100
ꢃ
ꢂꢃꢄ
where Xi and Yj are the mean conversion of i and yield towards
j, respectively. Subscripts i and j denote reactants and products
respectively, and superscript 0 refers to the initial conditions.
Results and discussion
small crystallites of t-ZrO2
increase with film thickness.
(Fig. 1) whose intensity slowly
Catalyst characterization and preliminary screening
Table 1 summarises the main physicochemical properties of
the ZrO2-SBA-15 materials, alongside commercial ZrSiO4
(Zircon, Aldrich, 99 wt%), tetragonal ZrO2 (Aldrich, 99 wt%),
and the parent SBA-15. Layer-by-layer conformal growth of
zirconia overlayers was previously evidenced by XPS, which
revealed quantitative agreement between the predicted and
experimentally determined Si 2p XP signal attenuation
following the growth of the different ZrO2 monolayers.24 The
observation of complete ZrO2 monolayers suggests that
adsorption of molecular Zr species is accompanied by the loss
of multiple isopropoxide groups, possibly arising from partial
hydrolysis by neighbouring silanols, which would increase the
Zr concentration within each adlayer. Based on the unit cell
dimensions of t-ZrO2 two layers of ZrOx octahedra are required
to generate a structure with ‘bulk-like’ termination.
Table 1. Physicochemical properties of catalysts.
Fig. 1 Wide angle XRD patterns of ZrO2-SBA-15, and pure SBA-15 and tetragonal ZrO2
Textural properties
Zr loading
Catalyst
SBET
(m2g-1)b
Vp
Dp
(nm)d
(wt%)a
(cm3g-1)c
The zirconium chemical environment within ZrO2-SBA-15
materials was subsequently explored by XPS (Fig. 2). Zr 3d XP
spectra exhibit two distinct spin-orbit split components
associated with the presence of interfacial Zr-O-Si and Zr-O-Zr
species with 3d3/2 binding energies of 183.8 eV and 183.1 eV
respectively, as previously reported.19 The low binding Zr-O-Zr
feature grows with film thickness, consistent with the
formation of an ordered t-ZrO2 phase observed by XRD, and
the appearance of a low binding energy state in the O 1s
spectra at 530.5 eV. Note that it is not possible to identify the
presence of Zr-OH species, which would appear ~532 eV,28 due
to overlap by the intense, broad signal from the underlying
silica substrate at 532.6 eV
t-ZrO2
74.0
49.8
8.4
-
<5
0.01
0.01
1.26
1.18
0.75
0.57
0.43
-
ZrSiO4
<5
-
Zr-SBA-15
SBA-15
533
451
367
349
296
12.3
11.1
8.3
6.7
6.5
ZrO2-SBA-15(1)
ZrO2-SBA-15(2)
ZrO2-SBA-15(3)
17.0
24.4
30.0
c
aICP-OES; bN2 porosimetry; Total pore volume for p/p0=0.98;d Mean pore size
from BJH-KJS method.
Zircon and tetragonal zirconia (t-ZrO2) were highly
crystalline (Fig. ESI-1) reflecting their correspondingly low
surface area and porosity, and negligible acid loading (NH3
TPD). In contrast, Zr-SBA-15 materials displayed a high surface
area and large mesopore size, a moderate acid loading, and
XRD reflections characteristic of only zircon domains.22
Brønsted acidity in Zr-SBA-15 has previously been attributed to
surface silanols, whereas Zr provides Lewis acidity25-27
ZrO2-SBA-15 exhibited a sharp (100) reflection indicative of
ordered mesopores within the parent SBA-15 for all ZrO2
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 4 of 11
View Article Online
DOI: 10.1039/C8CY01121D
ARTICLE
Journal Name
dry non-polar solvent in their synthesis (chosen to favour
growth of ZrO2 monolayers) compared with that of Kuwahara
et al20-21 who employed a polar solvent (which would likely
favour ZrO2 cluster formation).
Preliminary reactions were undertaken to benchmark the
performance of ZrO2-SBA-15 for the cascade transformation of
furfural in isopropanol at 130 °C. The resulting furfural
conversion and product yields are presented in Table 2
.
Table 2. Furfural reaction over Zr catalysts.
Minimal furfural reduction occurred over commercial t-
Product yield
Furfural
(mol %)b
Catalyst
conversion
(mol %)
FOL
i-FE
i-LEV GVL
Total
t-ZrO2
12
9
3
-
-
-
3
ZrSiO4
<1
1
-
-
-
<1
56
82
73
79
Zr-SBA-15
86
40
62
60
60
14
7
6
4
1
2
3
3
Fig. 2. (A) O 1s, and (B) Zr 3d XP spectra of ZrO2-coated SBA-15.
ZrO2-SBA-15(1) 99
ZrO2-SBA-15(2) 93
ZrO2-SBA-15(3) 90
11
5
Surface acidity of ZrO2-SBA-15 materials was determined
13
by NH3 TPD (Fig. ESI-5) and pyridine adsorption (Figure ESI-6-
aReaction conditions: 130 °C, 7 h; furfural : 2-propanol=1:50 (by mols); furfural :
catalyst=2.5:1 (by mass, equivalent to 10 g·L-1 catalyst). b Furfuryl alcohol (FOL);
2-propyl furfuryl ether (i-FE); 2-propyl levulinate ( i-LEV); γ-valerolactone (GVL).
7). Total acid loadings from ammonia titration increased from
0.227 to 0.357 mmoleqH+·g-1 with ZrO2 film thickness,
accompanied by a concomitant decrease in acid strength
(ammonia desorption temperature) summarised in Table ESI-
ZrO2 and ZrSiO4, presumably a consequence of their low
surface area, in turn reflecting the stability of zirconium within
the octahedral coordination geometry in these crystal
structures; saturating the Zr coordination sphere limits
opportunities for furfural and isopropanol binding, critical to
the MPV mechanism.31-33
1. The stronger (Lewis) acidity of zirconium cations at the
interface with silica than within t-ZrO2 is attributed to the
greater electronegativity of silicon versus zirconium, and
consequent stronger polarisation of interfacial Zr4+ cations.29
The Brønsted/Lewis character of ZrO2-SBA-15 materials was
probed by both ex-situ (DRIFTS, Fig. ESI-6) and in-situ (ATR-IR,
Fig. ESI-7) pyridine adsorption. Both measurements were in
good quantitative agreement, with the Brønsted:Lewis ratio
decreasing from ~0.3-0.4 to zero with increasing ZrO2 film
In contrast, Zr-SBA-15 and all ZrO2-SBA-15 catalyst exhibit
furfural conversions
≥86 %, attributable to the presence of
undercoordinated Zr species either in the framework, at the
interface with silica,34 or within highly dispersed t-ZrO2
crystallites.32 Zr-SBA-15 material displayed a much lower
catalytic performance in MPV reactions than ZrO2-
functionalized samples, as consequence of the lower
concentration of Lewis acid sites in this material. Despite
efficient furfural activation, neither Zr-SBA-15 nor any ZrO2-
SBA-15 produce high GVL yields, with furfuryl isopropyl ether
(i-FE) the dominant product, formed by the condensation of
furfuryl alcohol and the sacrificial isopropanol. The latter
transformation requires acid sites, and is the dominant
reaction over Zr-SBA-15 at 130 °C.34 Isomerisation of furfuryl
isopropyl ether to i-propyl levulinate (i-LEV) is disfavoured
under these conditions, consistent with reports that this
transformation requires Brønsted acid sites.9 The existence of
an alternative (albeit slow) Lewis acid catalysed ring opening
is, however, suggested by the increase in i-LEV yield with Lewis
acid strength across the Zr- and ZrO2-SBA-15 series. In any
event, i-FE isomerisation to i-LEV appears rate-limiting. The
final step in the reaction pathway from furfural to GVL is the
reduction of the ketone group in i-LEV to yield 4-hydroxy-
isopropyl valerate, and the latter’s subsequent intramolecular
thickness (Table ESI-1 and
2). This observation is readily
understood since Brønsted acidity is solely attributed to
surface silanols,25–27 which are capped by consecutive zirconia
monolayers, the latter imparting Lewis acidity arising from
undercoordinated Zr4+ sites. It is important to note that
significant Lewis acid sites remain at the ZrO2-SBA-15 surface
even during immersion in isopropanol (Fig. ESI-7), the same
solvent employed for our subsequent cascade reactions, and
hence are catalytically relevant. This is an important finding,
since the interconversion of Lewis to Brønsted acid sites has
been previously proposed following alcohol adsorption over
different metal oxides.30
Surface basicity assessed by CO2 pulse chemisorption
(
Table ESI-4) revealed base site loadings orders of magnitude
(0.002-0.011 mmol/g, Table ESI-4) lower than the
corresponding acid site loadings (>0.205 mm/g), and hence is
unlikely to contribute to the observed catalysis. It is also
important to note that levulinic acid would rapidly neutralise
such a low concentration of base sites during reaction. The
extremely low basicity of our materials may reflect the use of a
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 5 of 11
View Article Online
DOI: 10.1039/C8CY01121D
Journal Name
ARTICLE
transesterification (lactonisation) to GVL. Negligible 4-hydroxy-
isopropyl valerate was detected, suggesting that reduction of
the levulinate intermediate through H-transfer is the limiting
stage, and thus, the GVL yield is controlled by the rate of the
hydrogen transfer reaction. Despite the low GVL yields, ZrO2-
functionalized SBA-15 appears more active than Zr-SBA-15 for
the MPV reaction, being favoured with each layer of ZrO2
which provide increasing density of (weak) Lewis acid sites.
Temperature dependence and kinetic modelling
The influence of reaction temperature on the cascade
transformation of furfural to GVL was further investigated over
ZrO2-SBA-15 catalysts between 130 to 170 °C in order to both
optimise the process and provide extended datasets for kinetic
modelling (Fig. 3). Complete furfural conversion was achieved
in all cases in under 4 h, with the rate proportional to
temperature. As anticipated, furfuryl alcohol (FOL) was the
primary reaction product, and reached a maximum yield
between 1 and 3 h, before decreasing at the expense of
subsequent products in the cascade. The maximum FOL yield,
and time at which this was attained, decreased with reaction
temperature, suggesting that its rate of removal through
etherification exhibits
a strong temperature sensitivity.
Furfuryl isopropyl ether (i-FE) is the next product observed,
whose rates of formation and removal are again proportional
to
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 6 of 11
View Article Online
DOI: 10.1039/C8CY01121D
ARTICLE
Journal Name
Fig. 3. Influence of temperature on the reaction of furfural over ZrO2-SBA-15 catalysts. Reaction conditions: furfural:2-propanol=1:50 (by mols); furfural:catalyst=2.5:1 (by mass).
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 7 of 11
View Article Online
DOI: 10.1039/C8CY01121D
Journal Name
ARTICLE
according to the following model equations for the reaction
network:
FOL
FE
i-LEV
GVL
100
90
80
70
60
50
40
30
20
10
0
ꢉꢊꢋꢌꢍ
ꢁ −ꢏꢐ ꢆ ꢊꢋꢌꢍ − ꢏꢑ ꢆ ꢊꢋꢌꢍ
ꢉꢎ
ꢉꢊꢋꢒꢍ
ꢁ ꢏꢐ ꢆ ꢊꢋꢌꢍ − ꢏꢓ ꢆ ꢊꢋꢒꢍ − ꢏꢔ ꢆ ꢊꢋꢒꢍ
ꢉꢎ
ꢉꢊꢋꢕ
ꢁ ꢏꢓ ꢆ ꢊꢋꢒꢍ − ꢏꢖ ꢆ ꢊꢋꢕ − ꢏꢗ ꢆ ꢊꢋꢕ
ꢉꢎ
ꢉꢊꢘꢅꢍꢕꢙ
ꢁ ꢏꢖ ꢆ ꢊꢋꢕ − ꢏꢚ ꢆ ꢊꢘꢅꢍꢕꢙ
ꢉꢎ
ꢉꢊꢛꢙꢍ
ꢁ ꢏꢚ ꢆ ꢊꢘꢅꢍꢕꢙ
ꢉꢎ
ꢉꢊꢜ
ꢁ ꢏꢔ ꢆ ꢊꢋꢒꢍ + ꢏꢗ ꢆ ꢊꢋꢕ + ꢏꢑ ꢆ ꢊꢋꢌꢍ
ꢉꢎ
ZrO -SBA-15 (1)
ZrO -SBA-15 (2) ZrO -SBA-15 (3)
2
2
2
where C is the molar concentration of each component, and ki
is the apparent rate constant of step i. and H designates
humins. The Nelder-Mead method was used to optimise the
rate constants through minimising the following objective
Fig. 4. Product distributions for the reaction of furfural over ZrO2-SBA-15 catalysts.
Reaction conditions: 170 °C;
furfural:catalyst=2.5:1 (by mass).
7
h; furfural:2-propanol=1:50 (by mols);
temperature, although the latter remained relatively slow
even at 170 °C, consistent with our hypothesis that
isomerisation of the ether into isopropyl levulinate (i-LEV) is
rate-limiting for the overall cascade. Since the maximum i-FE
yield reached ~70 %, overcoming this bottleneck in the
cascade could unlock efficient GVL production. i-LEV (slowly)
accumulated at all temperatures, confirming that MPV
reduction to form GVL was inefficient. Considering the impact
of ZrO2 film thickness, the mono- and bilayer functionalised
SBA-15 catalysts showed increasing catalytic performance,
being ZrO2-SBA-15(2) slightly more efficient for the final MPV
function:
ꢥꢭ
ꢝꢞꢟꢠ ꢁ ꢡ ꢢꢡꢣꢊꢤ,ꢥ,ꢦꢧꢨꢦ − ꢊꢤ,ꢥ,ꢩꢪꢫꢬꢓꢯ
ꢤ
ꢥꢮꢐ
were Cm,n,calc and Cm,n,exp are the calculated and experimental
concentrations of component m at a given reaction time n. The
resulting apparent first-order rate constants for the three
ZrO2-SBA-15 catalysts are reported in Table ESI-5
. The
proposed pseudo-first order kinetic model provides a good fit
to the experimental datasets for all the catalysts (Fig. ESI-8).
step at 170 °C, achieving the largest GVL yield for any system
of 37 %. In contrast, the trilayer ZrO2-SBA-15(3) catalyst proved
ineffective for the isomerisation of i-FE to i-LEV even at 170 °C,
consistent with its lack of Brønsted acid sites. The product
distributions of the three ZrO2-SBA-15 catalysts at 170 °C are
summarised in Fig 4, which highlight that ZrO2-SBA-15(2) offers
good activity for furfural conversion to furfuryl alcohol, good
rates of i-FE isomerisation, and modest rates of i-LEV reduction
to GVL; we attribute this to its optimal balance of acid loading,
strength, and trace Brønsted acidity. Undesired side reactions
and resultant humin formation accounted for less than 15 % of
the reactively-formed products over the more weakly acidic
ZrO2-SBA-15(2) and ZrO2-SBA-15(3) catalysts, whereas stronger
acid sites in ZrO2-SBA-15(1) led to a poorer mass balance
deficit of 33 % due to humin production.
A
plausible reaction pathway for the cascade
transformation of furfural into GVL in alcohol media over ZrO2-
grafted SBA-15 is proposed in Scheme ESI-1. Rate constants for
each step can be derived by pseudo-homogeneous kinetic
modelling, with the exception of 4-hydroxy isopropyl valerate
lactonisation into GVL which is lumped together with the rate
of i-LEV reduction, due to the rapid nature of the former
intramolecular transesterification. Each step is fitted to a first-
order equation, assuming stirred batch reactor operation,
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 8 of 11
View Article Online
DOI: 10.1039/C8CY01121D
ARTICLE
Journal Name
Brønsted:Lewis character provides a simple approach to
optimise reaction steps in the cascade transformation of
furfural to GVL. Future work will address the challenge of
preparing high area, high acid loading, ZrO2-SBA-15 analogues
that possess enhanced Brønsted acidity to further accelerate
the (rate-limiting) isomerisation of furfuryl isopropyl ether to
isopropyl levulinate.
Catalysts reusability
Catalyst reuse was assessed over four consecutive reactions for
the ZrO2-SBA-15(2) sample which showed highest activity for
GVL production. The first three reaction cycles (170 °C, 7 h)
were performed by a direct reuse of the catalyst, which was
recovered by filtration after each cycle and simply air dried
overnight without any re-activation treatment. The originally
white sample progressively darkened to a brown colour on re-
use, indicating the adsorption of reactively-formed organics
(desired products and humins), as confirmed by TGA of the
used catalyst after three reaction cycles (Fig. ESI-9). This colour
change was accompanied by partial deactivation and an
inability to convert furfuryl isopropyl ether into isopropyl
levulinate (the most challenging reaction in the overall
cascade); this step is catalysed by strong acid sites which are
themselves most susceptible to poisoning. A subsequent
thermal reactivation (550 °C calcination in air for 5 h) restored
the original white colour of the catalyst, and recovered the
Fig. 5. Apparent first-order pseudo-homogeneous kinetic constants (ki) at 170 °C for
original catalyst activity in a subsequent fourth reaction cycle
steps 1, 2, 3, and 4 in the reaction of furfural over ZrO2-SBA-15 catalysts.
(
Fig. 6). Catalyst deactivation appears associated with
adsorption of organic residues at strong acid sites.
Furfural MPV reduction to furfuryl alcohol (k1) and its
etherification with 2-propanol (k2) are the fastest steps over
all catalysts, with furfuryl isopropyl ether isomerisation to
isopropyl levulinate (k3) confirmed as rate-limiting in all cases.
These findings are in accordance with the observed acid
properties of our three ZrO2-SBA-15 catalysts; the first and the
last steps (MPV transformations) in the cascade sequence are
Lewis acid, and base,18,21 catalysed, while the intermediate
etherification and isomerisation steps are favoured by
Brønsted and strong Lewis acids.9
The impact of zirconia film thickness on GVL production
can now be rationalised. ZrO2-SBA-15(1) possesses the highest
rate for furfuryl alcohol etherification (k2) and i-LEV
isomerisation (k3), but it does not effectively promote MPV
reductions of furfural (k1) and isopropyl levulinate (k4). In
contrast, ZrO2-SBA-15(3) is the best catalyst for MPV
reduction, but is extremely inefficient for the intermediate
steps in the cascade (etherification and isomerisation). Humin
production (Scheme ESI-1, reactions 5-7) is favoured at higher
reaction temperature, but remains a slow process over all
catalysts. Under these conditions, the bilayer ZrO2-SBA-15(2)
catalyst occupies the ‘Goldilocks zone’, offering an activity that
falls between the thinner and thicker zirconia coatings for
every step in the cascade. Nevertheless, the rate of humin
formation seems to correlate with the acid strength, being all
the humin yielding transformations faster in presence of ZrO2-
SBA-15(1). Tailoring the physicochemical properties of zirconia
monolayers, notably their acid loading, acid strength, and
Fig. 6. Product distributions obtained with ZrO2-SBA-15(2) in 4 consecutive reutilization
cycles in the cascade transformation of furfural to GVL. Catalyst regeneration was
accomplished by calcination in air after the 3rd reutilization cycle. Reaction conditions:
170 °C; 7 h; furfural:2-propanol=1:50 (by mols); furfural:catalyst=2.5:1 (by mass).
Conclusions
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 9 of 11
View Article Online
DOI: 10.1039/C8CY01121D
Journal Name
ARTICLE
Conformal zirconia monolayers grown by a wet chemical route gratefully acknowledge M.Eng. P. Juárez for assistance in
over mesoporous SBA-15 silica offer a flexible catalytic solution catalysts recycling study.
to drive the one-pot cascade transformation of furfural into
GVL in isopropanol. The zirconia coatings contain two Zr
References
species, those at the interface with silica and those within
highly dispersed t-ZrO2 crystallites prevalent in bi- and
trilayers, associated with strong and weak acid sites
respectively. Pyridine titration in vacuo and during immersion
in isopropanol evidences predominantly Lewis acidity for all
zirconia monolayer thicknesses; a small concentration of
Brønsted acid sites observed for a single zirconia monolayer
are probably associated with residual surface silanols. Furfural
reduction and etherification with isopropanol is quite efficient
at 130 °C, however a reaction temperature of 170 °C is
required to drive the subsequent reaction steps.
1
I. Delidovich, K. Leonhard, R. Palkovits, Energy Environ. Sci.,
2014, , 2803.
K. Yan, Y. Yang, J. Chai, Y. Lu, Appl. Catal. B Environ., 2015, 179
7
2
,
292–304.
3
4
R. Xing, W. Qi, G.W. Huber, Energy Environ. Sci., 2011, 4, 2193.
D. Ding, J. Wang, J. Xi, X. Liu, G. Lu, Y. Wang, Green Chem., 2014,
16, 3846.
J. Cui, J. Tan, T. Deng, X. Cui, H. Zheng, Y. Zhu, Y. Li, Green
Chem., 2015, 17, 3084–3089.
5
6
7
8
M.J. Gilkey, B.J. Xu, ACS catalysis, 2016,
6, 1420-1436.
M. Chia, J.A. Dumesic, Chem. Commun., 2011, 47, 12233.
R.S. Assary, L.A. Curtiss, J.A. Dumesic, ACS Catal., 2013, 3, 2694–
2704.
L. Bui, H. Luo, W.R. Gunther, Y. Roman-Leshkov, Angew. Chemie
- Int. Ed., 2013, 52, 8022–8025.
Kinetic modelling of furfural reactivity over ZrO2-SBA-15
catalysts identifies two distinct regimes associated with the
strong and weak acid Zr species. Interfacial, strong Lewis acid
Zr species within the first monolayer favour etherification and
isomerisation of reactively-formed furfuryl alcohol but show
poor activity for hydrogen transfer (MPV) chemistry. In
contrast, more abundant, but weaker Lewis acid sites in t-ZrO2
present in the bi- and trilayers promote furfural reduction and
the MPV reduction of isopropyl levulinate to GVL. A zirconia
bilayer coating, ZrO2-SBA-15(2), offers the optimum balance of
acidity and MPV hydrogen transfer activity to drive the
cascade transformation of furfural to GVL with 37 % overall
yield. Isomerisation of isopropyl furfuryl ether (formed by
etherification of furfuryl alcohol) to isopropyl levulinate is the
rate-limiting step in GVL production, suggesting that increasing
the GVL yield will require either a higher density of strong
Lewis sites or the stabilisation/introduction of additional
Brønsted acidity. our catalytic tests were (deliberately)
performed under dilute conditions with a view to minimize
9
10 M.M. Antunes, S. Lima, P. Neves, A.L. Magalhães, E. Fazio, A.
Fernandes, F. Neri, C.M. Silva, S.M. Rocha, M.F. Ribeiro, M.
Pillinger, A. Urakawa, A. A. Valente, J. Catal., 2015, 329, 522–
537.
11 M.M. Antunes, S. Lima, P. Neves, A.L. Magalhães, E. Fazio, F.
Neri, M.T. Pereira, A.F. Silva, C.M. Silva, S.M. Rocha, M. Pillinger,
A. Urakawa, A.A. Valente, Appl. Catal. B Environ., 2016, 182
485–503.
,
12 J.A. Melero, G. Morales, J. Iglesias, M. Paniagua, C. López-
Aguado, K. Wilson, A. Osatiashtiani, Green Chem., 2017, 19
,
5114-5121.
13 H.P. Winoto, B.S. Ahn, J. Jae, J. Ind. Eng. Chem., 2016, 40, 62–
71.
14 S. Song, L. Di, G. Wu, W. Dai, N. Guan, L. Li, Appl. Catal. B
Environ., 2017, 205, 393–403.
15 C. Hammond, S. Conrad, I. Hermans, Angew. Chemie Int. Ed.,
2012, 51, 11736–11739.
16 T. Viinikainen, H. Rönkkönen, H. Bradshaw, H. Stephenson, S.
Airaksinen, M. Reinikainen, P. Simell, O. Kraus, App. Catal. A:
General, 2009, 362, 169-177.
side reactions. Humin by product formation remains
a
challenge, and further work is required to optimise ZrO2-SBA-
15 catalysts for the selective transformation from furfural to
GVL. Although other catalytic systems (e.g. Zr-Al-β zeolite)12
display superior GVL selectivity, these much more expensive
17 C. Battilocchio, J.M. Hawkins, S.V. Ley, Org. Lett., 2013, 15
,
2278-2281.
and difficult materials to synthesise. Finally, it is important to 18 T. Komanoya, K. Nakajima, M. Kitano, M. Hara, J. Phys. Chem. C,
2015, 119, 26540-26546.
note that our experiments were conducted in the presence of
excess solvent, and that such operation would decrease the
profitability of any corresponding industrial process, since a
large fraction of the feedstream would be constituted by the
isopropanol solvent (although much of this could be recovered
and recycled). Ongoing research is directed at reducing the
amount of solvent through process optimisation.
19 G. Morales, A. Osatiashtiani, B. Hernández, J. Iglesias, J. A.
Melero, M. Paniagua, R. Brown, M. Granollers, A.F. Lee, K.
Wilson, Chem. Commun., 2014, 50, 11742–11745.
20 Y. Kuwahara, W. Kaburagi, T. Fujitani, Bull. Chem. Soc. Jpn.,
2014, 87, 1252–1254.
21 Y. Kuwahara, W. Kaburagi, Y. Osada, T. Fujitani, H. Yamashita,
Catal. Today, 2017, 281, 418-428.
22 J.A. Melero, L.F. Bautista, J. Iglesias, G. Morales, R. Sánchez-
Vázquez, Cat. Today., 2012, 195, 44-53.
23 H.J.M. Bosman, E.C. Kruissink, J. Vanderspoel, F. Vandenbrink, J.
Catal., 1994, 148, 660-672.
Acknowledgements
Financial support from the Spanish Ministry of Economy and
Competitiveness through the project CTQ2014-52907-R, as
well as from the Regional Government of Madrid through the
project S2013-MAE-2882 is gratefully acknowledged. B.
Hernández thanks the Spanish Ministry of Economy and
Competitiveness for a FPI grant (BES-2012-061144). Authors
24 A. Osatiashtiani, A.F. Lee, M. Granollers, D.R. Brown, L. Olivi, G.
Morales, J.A. Melero, K. Wilson, ACS Catalysis, 2015, 5, 4345-
4352.
25 M. M. Antunes, P. Neves, A. Fernandes, S. Lima, A. F. Silva, M. F.
Ribeiro, C. M. Silva, M. Pillinger and A. A. Valente, Catal. Sci.
Technol., 2016, 6, 7812–7829.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 10 of 11
View Article Online
DOI: 10.1039/C8CY01121D
ARTICLE
Journal Name
26 J.A. Melero, L.F. Bautista, J. Iglesias, G. Morales, R. Sánchez-
Vázquez, Catal. Today, 2012, 195, 44–53.
27 J. Geboers, X. Wang, A.B. de Carvalho, R. Rinaldi, J. Mol. Catal. A
Chem., 2014, 388, 106–115.
28 Y. Mi, J. Wang, Z. Yang, Z. Wang, H. Wang, S. Yang, RSC Adv.
2014, 4, 6060-6067.
29 J.A. Anderson, C. Fergusson, I. Rodríguez-Ramos, A. Guerrero-
Ruiz, J. Catal., 2000, 192, 344-354.
30 E. V. Beletskiy, X. Hou, Z. Shen, J. R. Gallagher, J. T. Miller, Y.
Wu, T. Li, M. C. Kung,H. H. Kung, J. Am. Chem. Soc., 2016, 138
,
4294–4297.
31 A. Corma, M. E. Domine, S. Valencia, J. Catal., 2003, 215, 294–
304.
32 J. C. van der Waal, K. Tan, H. van Bekkum, Catal. Letters, 1996,
41, 63–67.
33 F. Gonell, M. Boronat, A. Corma, Catal. Sci. Technol., 2017, 7,
2865, 2873.
34 J. Iglesias, J. A. Melero, G. Morales, J. Moreno, Y. Segura, M.
Paniagua, A. Cambra, B. Hernández, Catalysts, 2015,
5
.
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 11 of 11
Catalysis Science & Technology
View Article Online
DOI: 10.1039/C8CY01121D
190x275mm (96 x 96 DPI)