Efficient one-pot production of γ-valerolactone from xylose over Zr-Al-Beta zeolite: Rational optimization of catalyst synthesis and reaction conditions

Article abstract of DOI:10.1039/c7gc01969f

The one-pot conversion of xylose into γ-gammavalerolactone in 2-propanol over bifunctional Zr-Al-Beta zeolites, prepared via a post-synthetic route, was optimized in terms of both catalyst synthesis and reaction conditions. In the catalyst preparation, the use of Zr(NO₃)₄ as zirconium source as well as the tuning of the amount of water used during the impregnation had a strong impact on the activity of the Zr species due to an improved dispersion of Zr species. As for the aluminium to zirconium exchange, an optimal Al/Zr ratio of 0.20 was identified to provide a catalyst with better activity. The modelization of the catalytic system through experimental design methodology allowed to identify the optimal values of the most influential reaction conditions: temperature 190 °C, catalyst loading 15 g L^-1, and starting xylose concentration 30.5 g L^-1. Under these optimized reaction conditions, Zr-Al-Beta catalyst provides a GVL yield from xylose (ca. 34%) after only 10 h. The catalysts are stable and reusable after thermal regeneration at 550 °C.

Full text of DOI:10.1039/c7gc01969f

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. A. Melero, G.
Morales, J. Iglesias, M. Paniagua, C. López-Aguado, K. Wilson and A. Osatiashtiani, Green Chem., 2017,
DOI: 10.1039/C7GC01969F.
Volume 18 Number
7
7
April 2016 Pages 1821–2242
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
Accepted Manuscripts are published online shortly after
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
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
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Please do not adjust margins
Green Chemistry
Page 1 of 9
View Article Online
DOI: 10.1039/C7GC01969F
Journal Name
ARTICLE
Efficient one-pot production of γ-valerolactone from xylose over Zr-
Al-Beta zeolite: rational optimization of catalyst synthesis and
reaction conditions
Received 00th January 20xx,
Accepted 00th January 20xx
a,
a
a
a
a
Juan A. Melero *, Gabriel Morales , Jose Iglesias , Marta Paniagua , Clara López-Aguado , Karen
DOI: 10.1039/x0xx00000x
b
b
Wilson , Amin Osatiashtiani
www.rsc.org/
The one-pot conversion of xylose into γ-gammavalerolactone in 2-propanol over bifunctional Zr-Al-Beta zeolites, prepared
via a post-synthetic route, was optimized in terms of both catalyst synthesis and reaction conditions. In the catalyst
3 4
preparation, the use of Zr(NO ) as zirconium source as well as the tuning of the amount of water used during the
impregnation had a strong impact on the activity of the Zr species due to an improved dispersion of Zr species. As for the
aluminium to zirconium exchange, an optimal Al/Zr ratio of 0.20 was identified to provide a catalyst with better activity.
The modelization of the catalytic system through experimental design methodology allowed to identify the optimal values
-1
of the most influential reaction conditions: temperature 190 ºC, catalyst loading 15 g·L , and starting xylose concentration
-
1
3
0.5 g·L . Under these optimized reaction conditions, Zr-Al-Beta catalyst provides a GVL yield from xylose (ca. 34%) after
only 10 h. The catalysts are stable and reusable after thermal regeneration at 550ºC.
noble-metals, the reduction of LA and its esters by catalytic transfer
Introduction
hydrogenation (CTH) via Meerwein-Ponndorf-Verley (MPV) reaction
in the presence of an H-donor alcohol has been widely described in
The interest of ɤ-valerolactone (GVL) as a versatile biomass-derived
platform chemical has grown fast in the last decade due to its good
properties as green solvent and fuel additive, as well as precursor
11-12
literature
. This approach offers a high chemo-selectivity for the
reduction of carbonyl groups under mild conditions and without the
requirement of expensive metals as active phase or high hydrogen
pressures (Fig. ESI-1). Zirconium-based catalysts of different nature
1
-10
for the production of liquid alkanes and high-value biopolymers
.
The currently proposed strategy for GVL production is a multi-step
process starting by the selective fractionation and depolymerization
of the lignocellulosic biomass. This is followed by the acid catalyzed
transformation of the released monosaccharides into levulinic acid
(
2 2 3 2 2 4
ZrO , Al O -ZrO , Zr-SBA-15, ZrO /SBA-15, Zr(OH) , ZrFeO, etc.)
have been tested for the production of GVL from LA and its esters in
the presence of different alcohols (mainly 2-propanol because of its
high reactivity in MPV reduction) yielding in most of the catalytic
(
LA), which is then hydrogenated to the desired GVL, using
13-19
assays GVL yields over 80 %
. On the other hand, pure-silica
hydrogen gas, and catalytic bed reactors bearing supported noble-
metals (Ru, Pt and Ir). Although this approach has been reported to
provide high yields towards GVL, it also has important drawbacks:
the use of mineral acids for the acid-catalyzed steps; only hexoses
are transformed into GVL; the use of expensive catalysts in the
hydrogenation step; and the requirement of high hydrogen
pressure (> 30 bar).
zeolites doped with Zr and Sn provide the presence of acid Lewis
sites, which efficiently promote the CTH reactions. In this way,
Roman-Leshkov et al have recently studied the catalytic
performance of different Lewis metal-modified zeolites (Zr-Beta, Al-
Beta, Sn-Beta and Ti-Beta) in the production of GVL from methyl-
2
0
levulinate in 2-butanol . Among these zeolites, Zr-beta showed the
best catalytic results with a GVL selectivity close to 100 % after 5
As an alternative to the conversion of LA to GVL using H
2
over hours of reaction and complete levulinate conversion. Later, Wang
et al. demonstrated that Zr-Beta zeolite is also a robust and active
2
1
catalyst for the MPV reduction of levulinic acid to GVL . The high
activity of this catalyst was attributed to the presence of Lewis
acidic sites with moderate strength coming from the incorporation
of zirconium atoms within lattice positions of the zeolite
framework.
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
Green Chemistry
Page 2 of 9
View Article Online
DOI: 10.1039/C7GC01969F
ARTICLE
Journal Name
These studies are focused on the use of LA and/or its esters, but Catalysts preparation
this approach needs a source of high-purity LA. Hence, a great
The procedure for the synthesis of zirconium-modified Beta zeolites
interest has arisen to obtain GVL not from levulinic acid but directly
from other available biomass-derived molecules in a one-pot
approach through the combined action of acid (Brønsted) and
reduction (Lewis) sites, thus avoiding the need of isolating LA. In
this way, the group of Roman-Leshkov firstly reported an integrated
catalytic process for the efficient production of GVL from furfural in
a one-pot process using a physical mixture of Lewis and Brønsted
(
Zr-Al-Beta) was slightly modified from that reported in
2
1,24
literature
. A commercially available H-Beta zeolite (Si/Al=22)
was purchased from Zeolyst International and used as parent
support. Partial dealumination was carried out by treatment in
3
nitric acid solution (60% aq. HNO , Scharlau), using varying acid
-1
concentrations at room temperature (1h, 20 mL·g ). Higher
20
dealumination degrees were achieved using harsher conditions
catalysts . In the cascade of reactions, furfural is firstly converted
into furfuryl alcohol through a transfer hydrogenation reaction
promoted by a Lewis acid catalyst (Zr-Beta zeolite) using 2-butanol
as the hydrogen donor. Next, a Brønsted acid catalyst (Al-ZSM-5
zeolite) converts furfuryl alcohol into a mixture of levulinic acid and
butyl levulinate through hydrolytic ring-opening reactions. Finally,
both levulinic acid and butyl levulinate undergo a second transfer-
hydrogenation step, catalyzed again by Zr-Beta, to produce GVL
with a yield over 70%.
-
1
(
3
10M HNO , 100 ºC, 20 mL·g ) and increasing times. After washing
with deionized water and centrifugation to recover the solid
products, the resulting materials were dried overnight (110 ºC).
Thereafter, zirconium incorporation was accomplished by incipient
wetness impregnation of the dealuminated zeolites with an
aqueous solution of the Zr precursor (zirconium oxychloride,
Aldrich, or zirconium (IV) nitrate, Chemical Point). The amount of
zirconium precursor was adjusted to match that of extracted
aluminium. Solvent was removed under vacuum from the resultant
A step forward has been recently described in several works, aiming slurries, and the solids were dried overnight and then calcined in air
-1
to combine Lewis and Brønsted acid sites within the same zeolite at 200 ºC for 6h (3 ºC·min heating ramp) and thereafter at 550 ºC
2
2
framework (bifunctional catalysts). Thus, Winoto et al
have for another 6h (same heating rate). The different so-synthesised Zr-
reported a bifunctional Sn-Al-Beta zeolite for the production of GVL modified Beta zeolites were consecutively named as Zr-Al-Beta-n
from furfural. Thereby, Lewis acid Sn-Beta zeolites with Brønsted (Table 1).
acid Al sites in the framework were synthesized via a post-synthesis
Catalysts characterization
procedure based on the partial dealumination of the parent Al-Beta
followed by the incorporation of tin species. An optimized Sn-Al-
Zirconium and aluminium contents were determined by Inductively
Beta zeolite with a Si/Sn ratio of 63 and a Si/Al ratio of 473
Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) with a
exhibited the highest GVL yield (60%) at 180 ºC in 2-butanol. More
23
Varian Vista AX apparatus. Textural properties of the tested
catalysts were calculated from argon adsorption-desorption
isotherms, which were recorded at 87K using an AutoSorb
equipment (Quantachrome Instruments). Average pore size was
estimated using the NLDFT method, and total pore volume was
taken at P/Po = 0.98. X-ray powder diffraction (XRD) patterns were
collected in a Philips X-pert diffractometer using the Cu Kα line in
the 2θ angle range from 5º to 65º (step size 0.04º). Transmission
electron microscopy (TEM) images were obtained on a PHILIPS
TECNAI-20 electronic microscope operating at 200kV. Acid
properties of the materials were determined by means of
recently, Li et al have successfully prepared meso-Zr-Al-beta
zeolites via a multiple-step post-synthesis strategy with the
combination of controlled dealumination, desilication and metal
ions incorporation. Under optimized conditions (catalysts and
reaction parameters), these materials achieved an outstanding GVL
yield of 95% after 24 h of reaction at 120 ºC in the presence of 2-
propanol.
Our research group, going a step further, has firstly reported the
direct transformation of the parent monosaccharide, i.e. xylose,
2
4
into GVL through a single reaction step over Zr-Al-beta zeolites
.
temperature-programmed desorption of NH
3
in a Micromeritics
We demonstrated that a proper tuning of the Brønsted/Lewis acid
sites ratio is essential to efficiently promote the alternant MPV and
2
910 (TPD/TPR) equipment fitted with a TCD detector. Diffuse
acid-driven reaction steps. In this way, a zirconium-modified Beta reflectance infra-red Fourier transform (DRIFT) spectra were
obtained using a Thermo Scientific Nicolet iS50 FT-IR spectrometer
with Smart Collector accessory, mid/near infrared source and
mercury cadmium telluride (MCT-A) photon detector at -196 °C
zeolite with an Al/Zr molar ratio of 0.22 attained a 35 % GVL yield
starting from xylose after 48 h at 190 ºC using 2-propanol as H-
donor. In this contribution, we have optimized the synthesis
conditions of Zr-Al-Beta in order to increase the Zr species
dispersion with the purpose of improving the production of GVL
from xylose. Likewise, the influence of reaction conditions,
including temperature, and catalyst and xylose concentration, has
been investigated through an experimental design methodology.
(
liquid N2). All the spectra are average of 64 scans measured at a
−1
resolution of 4 cm . Samples were diluted homogeneously to 10
wt. % with KBr. For ex situ pyridine adsorption, diluted samples
were impregnated with neat pyridine, and excess physisorbed
pyridine was removed in vacuo at 40 °C for 18 h prior to room
temperature measurement in vacuo. X-ray Photoelectron
Spectroscopy (XPS) measurements were obtained using a Specs
Phoibos 150 9MCD photoelectron spectrometer.
Experimental
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 9
Pleas Ge rdeo e nn o Ct ah de j mu si ts mt r ya rgins
View Article Online
DOI: 10.1039/C7GC01969F
Journal Name
ARTICLE
Catalytic tests
(XXYL) or in terms of yields towards the different products (Y ). The
i
definitions of these parameters are as follows:
Catalytic tests were typically performed with the reaction
-
1
ꢄꢅꢆꢇꢈꢅꢉꢊꢋꢌꢍꢊꢌꢎꢊꢏꢐꢍꢌꢑꢅ
composition: catalyst loading 0.01 g·mL , xylose:2-propanol
ꢀꢁꢂꢃ
ꢕꢔ =
=
× 100
(1)
(2)
ꢒꢓꢔꢈꢔꢆꢍꢊꢋꢌꢍꢊꢌꢎꢊꢏꢐꢍꢌꢑꢅ
(
Scharlau, 98%) molar ratio 1:50, xylose (XYL) (Sigma Aldrich, 99%)
-1
loading 0.026 mol·gCAT . Catalytic runs were carried out in a
stainless steel stirred autoclave (200 mL) with temperature control
and pressure gauge. After sealing the reactor, stirring was fixed at
ꢖꢌꢗꢋꢅꢉꢊꢋꢌꢍꢊꢌꢎꢊꢔ
× 100
ꢒꢓꢔꢈꢔꢆꢍꢊꢋꢌꢍꢊꢌꢎꢊꢏꢐꢍꢌꢑꢅ
1
000 rpm in order to avoid mass transfer hindrances, and a heating
-1
rate of 6 ºC·min was established to reach the desired temperature
within 1 h. This point is taken as the beginning of the reaction,
ignoring the time needed for heating up. Decane (Acros Organics Optimization of the reaction conditions
-
1
>
99%) was used as internal standard (0.01 g·mL ). Samples were
withdrawn periodically and filtered into a vial before analysis.
Response surfaces were calculated to optimize the operation
conditions using Al-Zr-Beta zeolite to maximize GVL yield. Three
independent operation variables (factors) were selected: the
reaction temperature (X ), the catalyst loading (X ), and the starting
Products analysis
1
2
Conversion of xylose was evaluated by HPLC with an Evaporative
Light Scattering Detector (ELSD) (Agilent Technologies 1260
3
xylose concentration (X ). Table 1 lists the experimental range for
the selected operation variables, which are based on preliminary
+
Infinity), using a Hi-PLEX H column (300 mm-long, 7.7 mm-i.d., 8
24
experimental results , as well as the corresponding levels of the
µm particle size). Reaction samples were also analysed by gas
chromatography (GC), using a Varian 3900 gas chromatograph
fitted with a ZB-WAX Plus column (30 m x 0.25 mm, DF=0.25 µm)
and a FID detector. Compounds detected by GC were furfural
factors. A face-centered central composite design consisting of 18
experiments was performed, comprising 8 factorials points, 6 star
points, and 4 cube center replicas.
(Sigma Aldrich, 99%), furfuryl alcohol (Sigma Aldrich, 98%), Table 1. Full factorial face-centered composite design matrix of three
variables and the observed responses in the optimization of GVL production
in presence of Al-Zr-Beta.
isopropyl furfuryl ether (Manchester Organics, 97%), isopropyl
levulinate (iL) (98%, non-commercial, see below for its preparation),
levulinic acid (LA) (Sigma Aldrich 98%), α/β-angelica lactones (LACT)
Coded levels
Factor
(Sigma Aldrich, 98%), γ-valerolactone (GVL) (Sigma Aldrich, 99%)
-1
150
5
0
+1
190
15
and methyl-β-D-xylopyranoside (Sigma Aldrich, 99%) as
a
X
X
X
1
2
3
Temperature (ºC)
Catalyst loading (g·L )
Starting xylose concentration (g·L )
170
10
-
1
representative of the family of isopropyl xylosides formed by
etherification of xylose with 2-propanol. Isopropyl levulinate was
synthetized by esterification of levulinic acid in 2-propanol with
sulfuric acid as catalyst. Purification of the resultant levulinate was
-
1
10
40
70
2
0
All experiments were carried out in a randomized order to minimize
the effect of unexplained variability in the observed response.
Experimental data were analyzed by response surface methodology
using a second-order polynomial equation such as the following
one:
accomplished following a previously described procedure . After
removal of 2-propanol by distillation under reduced pressure, the
product mixture was dissolved in MTBE and extracted 3 times with
saturated aq. KHCO solution and 3 times with distilled water. The
3
solution was dried with anhydrous Na
2 4
SO and then filtered. Finally,
the solvent was removed by distillation under reduced pressure.
3
3
3
1
13
Purification grade was assessed by H NMR and C NMR
2
n
Y = β0
+
βn ⋅ X +
βnn ⋅ X +
βnm ⋅ X ⋅ Xm
(3)
∑
n
∑
∑∑
n
1
spectroscopy. H NMR (400 MHz, CDCl
3
): δ 4.97 (sep, J = 6.3 Hz, 1H),
n=1
n=1
n<m
2
.72 (t, J = 6.7 Hz, 2H), 2.52 (dt, J = 0.4, 6.7 Hz, 2H), 2.17 (s, 3H), 1.2
1
3
(
(
)
d, J = 6.3 Hz, 6H) ppm; C NMR (100 MHz, CDCl ): δ 206.98 where Y is the response (GVL yield, mol%) and β
0
, β
C(O)C- are the coefficients of the model, representing the intercept, linear,
) ppm. quadratic, and binary interactions, respectively. X and X are the
n
, βnn and βnm
3
CH
3
C(O)-); 172.45 (-C(O)OiPr), 68.18 (-C(CH
2
)CH
2
), 38.19 (CH
)CH
3
, 30.07 (CH
3
C(O)CH C-), 28.58 (CH C(O)-), 21.96 (-C(CH
2
3
2
2
n
m
The identification of others bio-products, such as 2-propoxy glycol studied independent factors. An analysis of variance (ANOVA) was
and isopropyl lactate coming from the retro-aldol condensation of performed to calculate the regression coefficients. The analysis of
xylose (Fig. ESI-2) was checked by GC-MS using a Bruker 320-MS GC statistical significance was based on the total error criteria with a
Quadrupole Mass Spectrometer equipped with a capillary column confidence level of 95.0%.
(BR-SWax, 30 m x 0.25 mm, DF=0.25 µm), using He as carrier gas.
Product quantification was based on a previous calibration of the
analysis unit with standard stock solutions of pure commercially-
Results and discussion
available chemicals using decane as internal standard. Catalytic Within this work, we have optimized the synthesis of Zr-modified
results are shown either in terms of absolute conversion of xylose Beta catalysts in order to maximize the production of GVL in the
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
Green Chemistry
Page 4 of 9
View Article Online
DOI: 10.1039/C7GC01969F
ARTICLE
Journal Name
cascade reaction from xylose. The starting point for this species. On the contrary, the sample prepared from ZrOCl displays
2
optimization was the catalyst that has so far provided the best some minor small dark contrast areas, indicating some degree of
2
4
performance , i.e. Zr-Al-Beta prepared via
a post-synthetic inhomogeneity probably coming from the accumulation of poorly
dealumination procedure, using zirconium oxychloride as zirconium dispersed Zr species. The catalytic evaluation of both samples is
precursor, with 0.22 Al/Zr ratio, and 0.17 Brønsted/Lewis acid sites depicted in Figure 2. The better zirconium dispersion achieved for
ratio. This catalyst was able to provide 22.6 % yield of GVL after 48h Zr-Al-Beta-2 results in an enhancement of the cascade reaction
at 170ºC, reaching 35 % when increasing the temperature up to rate, as it can be inferred from the higher final GLV yield (30 % vs 24
190ºC. The proposed strategies for the optimization of the %, after 48h). Analysing the evolution of the different species with
synthesis have followed two different routes: i) increasing the time, some remarkable aspects highlighting the acceleration of the
surface dispersion of Zr species in the dealuminated zeolite, and ii) process towards GVL are the lower 2-propyl xylosides
exploring a further reduction of the Al/Zr ratio, by means of concentration, the better catalytic performance in dehydration
reducing the amount of aluminium. For the first effect, a different reactions, as well as lower contents on furfural and iso-propyl
zirconium precursor, Zr(NO
displays higher water solubility than the previously used ZrOCl
thus a better dispersion of Zr species during the impregnation can (MPV transformations) because of a higher activity of the better-
3
)
4
, has been tested. This compound levulinate obtained over Zr-Al-Beta-2. This improvement can be
2
and ascribed to an enhancement of both carbonyl groups reductions
2
5
be obtained . In the same line, an increase of water amount during dispersed zirconium. As a side effect, the enhancement of Zr sites
4
+
the impregnation step, in order to favour Zr ions mobility, has also activity also contributes to increase the extent of the retro-aldol
been evaluated. Finally, lower Al/Zr ratios have been obtained by condensation of xylose, providing products such as 2-propyl lactate
increasing the harshness of the dealumination procedure (longer and 2-propoxyethanol (see Fig. ESI-2), which can be ascribed to a
refluxing times in concentrated nitric acid solution) while keeping higher Lewis acidity in these materials.
constant the Zr loading.
Regarding the structural characterization of the materials included
Considering all these factors, a series of modified Zr-Al-Beta zeolites in Table 2, as a general trend, all the samples prepared using
were prepared and Table 2 summarises their physicochemical zirconium nitrate preserved the typical XRD pattern of the parent
properties as well as the GVL yield achieved after 48h at 170ºC. The Beta zeolite (Fig.ESI-3a). Additionally, typical XRD reflections of
3 4 2 2
benefits of using Zr(NO ) vs. ZrOCl as zirconium precursor are crystalline ZrO could not be observed in these materials, indicating
highlighted when comparing samples Zr-Al-Beta-1 and Zr-Al-Beta-2. the absence of large domains of this metal oxide, and a proper
Both materials come from the same dealuminated Beta zeolite and incorporation of Zr into lattice positions within the zeolite structure.
have very similar composition and structure, except for the NH
TPD analysis that shows a slight decrease in acidity from 0.35 to porosimetry (Fig.ESI-3b) confirm the absence of significant structure
.28 mmolH+/g. TEM analysis (Fig. 1) shows high homogeneity of damage during the synthesis procedure.
material Zr-Al-Beta-2, without any evidence of ZrO bulk
3
-
Likewise, textural properties measured by argon manometric
0
2
agglomerates, in consonance with a lattice incorporation of the Zr
Table 2. Physicochemical properties of the Zr-Al-Beta zeolites and catalytic performance in the production of GVL from xylose.
b
Composition
Acidity
a
BET
Total pore
Yield to
GVL
(%)
Zirconium
precursor
H
2
O
c
d
g
Catalyst
-1
area
volume
(mL·g )
2
-1
3 -1
+
-1
e
f
(
m g )
(cm g )
%
Al
% Zr
Si/Al
Si/Zr
Al/Zr
(mmol H g )
B/L ratio
Beta (parent)
Zr-Al-Beta-1
Zr-Al-Beta-2
-
-
2.0
0.0
5.0
4.5
22
-
-
623
582
561
0.36
0.37
0.36
0.41
0.35
0.28
0.56
0.17
0.03
0
ZrOCl
2
5
5
0.28
0.27
149
156
28
32
0.19
0.20
24
30
Zr(NO
3
)
4
4
Zr-Al-Beta-3
Zr-Al-Beta-4
Zr-Al-Beta-5
Zr-Al-Beta-6
Zr-Al-Beta-7
Zr(NO
Zr(NO
Zr(NO
Zr(NO
Zr(NO
3
)
)
)
)
)
5
5
0.12
0.06
0.04
0.27
0.23
4.0
4.5
6.4
4.5
354
704
1027
156
183
36
32
22
32
31
0.10
0.05
0.02
0.20
0.17
601
567
559
685
681
0.39
0.33
0.36
0.38
0.38
0.21
0.18
0.15
0.29
0.04
0.05
0.07
0.05
0.10
22
14
10
37
32
3
3
3
3
4
4
4
4
5
10
15
4.6
0.27
a
b
c
Amount of water used during the impregnation of Zr precursor. % Al, % Zr (w/w); Si/Al, Si/Zr, Al/Zr (atomic ratios) measured by ICP-OES. Surface area calculated by the
d
e
f
B.E.T. method. Total pore volume recorded at P/P
molecular probe. Yield to γ-valerolactone from xylose; reaction conditions: 170°C, reaction time 48h, xylose:2-propanol molar ratio 1:50, catalyst loading 10 g·L .
0
= 0.975. Acid sites loading analysed by NH
3
-TPD. Brønsted/Lewis acid sites ratio determined by FTIR using pyridine as
g
-1
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
Green Chemistry
Page 5 of 9
View Article Online
DOI: 10.1039/C7GC01969F
Journal Name
ARTICLE
materials, the use of Zr(NO
3
)
2
as zirconium precursor seems to lead
2
to B/L ratios lower than when ZrOCl is used, even at similar Al y Zr
contents (see, e.g., Zr-Al-Beta-1 vs. Zr-Al-Beta-6). As a hypothesis,
this could be explained by the better dispersion of Zr species when
using the more soluble zirconium nitrate. Thus, a better dispersion
of Zr species would cover larger portion of the zeolite surface,
affecting in a larger extent the Al environments and possibly
decreasing part of the Brønsted acidity.
Figure 1. TEM images of (a) Zr-Al-Beta-1 and (b) Zr-Al-Beta-2 zeolites.
Aluminum content was then decreased in order to check the effect
of using lower Al/Zr ratios. In this way, materials Zr-Al-Beta-3 to 5
were prepared in the same way than Zr-Al-Beta-2, except for using
increasingly harsher dealumination processes. As a result, Al/Zr
could be tuned from 0.20 to as low as 0.02 (Table 2). As expected,
the higher extent of Al removal from the network, together with an
approximately constant Zr loading, resulted in a progressive
3
decrease of the NH -TPD acidity of the materials. The assessment of
the catalytic activity of these materials resulted in lower final GVL
yields, showing a progressive reduction of the production of GVL
insofar as the Al content decreases. This result evidences the
necessity of a minimum amount of Al sites to effectively catalyze
Brønsted acidity reaction steps in the cascade reaction (sugar
dehydration, furfuryl alcohol hydration and isomerisation…). On the
other hand, samples Zr-Al-Beta-6 and Zr-Al-Beta-7 were prepared
using higher water concentrations during the impregnation with
Figure 2. Xylose conversion and yields to products in 2-propanol over a) Zr-
Al-Beta-1, b) Zr-Al-Beta-2. Reaction conditions: 170°C, xylose:2-propanol
-1
-1
-1
molar ratio 1:50, catalyst loading 10 g·L .
zirconium nitrate, double (10 mL·g ) and triple (15 mL·g ),
respectively, than in Zr-Al-Beta-2. Despite the composition, textural
and structural properties do not vary significantly among the three
samples (indeed, they keep an almost constant Al/Zr ratio), there is
a remarkable effect on the catalytic activity. Thus, doubling the
amount of water led to 37 % GVL yield, while tripling it had a lower
impact on the catalytic performance, leading to 32 % GVL yield.
Considering all the above descriptions, it can be concluded that
there is an optimal Al/Zr ratio around 0.20 for the production of
GVL from xylose over zirconium-modified Beta zeolites. However,
this is not the only influential parameter. The zirconium precursor
used in the post-synthetic preparation method also results critical,
being the zirconium nitrate, a more water-soluble zirconium
precursor, more appropriate than the oxychloride counterpart to
achieve higher metal dispersions. Furthermore, the amount of
water used during the impregnation has an impact on the
The relative presence of Brønsted and Lewis acid sites was assessed
by means of DRIFT spectroscopy for adsorbed pyridine on each
sample (Fig. ESI-4). Table
2
incorporates the obtained
Brønsted/Lewis acid sites ratios. Among the different synthesized
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
Green Chemistry
Page 6 of 9
View Article Online
DOI: 10.1039/C7GC01969F
ARTICLE
Journal Name
dispersion of Zr species and, consequently, on their distributed obtained in four replicas of the central point (Runs #1-4). Figure 3
incorporation within the dealuminated Beta zeolite. Sample Zr-Al- depicts the evolution of the products distribution with time,
Beta-6 led to the higher yield to γ-valerolactone (Fig. ESI-5) and obtained under the reaction conditions corresponding to the center
therefore it was selected to continue with the optimization of the of the experimental field. Repeatability of the catalytic assays is
reaction conditions.
quite high, as it is evident from the comparison of the four different
reaction runs depicted in Fig. 3. Actually, reproducibility is better for
longer reaction times, as the first stages of the reaction cascade
proposed for the transformation of xylose into GVL (etherification,
dehydration, H-transfer,…) are quite fast, and slight differences in
operational variables like temperature, the catalyst loading or
starting xylose concentration, may lead to the larger divergences
observed during the early moments of the reaction. Nevertheless,
long reaction times, such as 24 h, provide quite similar results in
products distribution. In this way, the curvature of the model at 24
h has resulted to be significant for the selected response, the GVL
yield.
Optimization of the reaction conditions
Enhancing the yield towards the desired GVL product requires
understanding the influence of the different reaction key factors
affecting the progress of the described reaction cascade from xylose
to GVL. For this purpose, three operation factors have been varied
according to
a factorial design of experiments to build a
mathematical model that can predict the GVL yield as a function of
the reaction conditions. This study not only allows getting a
predictive mathematical model to infer the GVL yield, but also
studying and get a deep knowledge on the influence of the studied
variables on this response factor. Thus, reaction temperature,
catalyst loading, and xylose concentration in the reaction medium
have been selected as key variables affecting the final GVL yield. For
this optimization study, evolution of the reaction media
composition was monitored up to 24h. Table
3 lists the
experimental face-centered cube design, including the yields to the
main products achieved in these catalytic tests.
Table 3. Full factorial face-centered composite design matrix of three
variables and the observed responses after 24 h of reaction in the
optimization of GVL production over Zr-Al-Beta-6 zeolite.
d
Experimental factors
Main products yields
Run
a
b
2
c
3
X
1
X
X
Y
iL
Y
LA
Y
LACT
Y
GVL
1
2
3
4
5
6
7
8
9
0
0
0
0
0
0
0
0
0
0
0
0
1.5
1.7
1.9
1.4
0.7
2.4
3.2
5.5
0.0
0.0
3.1
0.4
5.7
0.2
4.9
0.9
0.2
4.6
1.6
1.6
1.8
1.1
0.8
5.0
6.6
9.6
0.0
0.0
3.4
0.6
11.0
0.1
6.3
0.6
0.4
1.3
1.6
26.5
26.4
24.8
25.2
19.5
12.7
1.4
1.8
1.2
-1
-1
-1
-1
+1
+1
+1
+1
-1
+1
0
+1
-1
-1
+1
+1
-1
-1
+1
0
-1
-1
+1
+1
-1
-1
+1
+1
0
3.4
Figure 3. Repeatability of the catalytic test corresponding to the operation
conditions at the center of the experimental field: 170ºC, xylose:2-propanol
12.5
11.3
11.1
0.0
-
1
molar ratio 1:50, catalyst loading 0.1 g·mL .
5.7
30.4
27.5
20.2
29.6
10.2
28.8
17.3
30.8
27.3
12.2
1
1
1
1
1
1
1
1
1
0
1
2
3
4
5
6
7
8
0.0
A quadratic model was fitted to predict GVL yield as a function of
the operation conditions. The mathematical expression, including
only the regression coefficients corresponding to significant factors
and interactions (at 95% confidence level), is as follows:
1.8
0.3
13.5
0.3
0
0
2
1
2
3
(4)
-1
+1
0
0
6.0
Y
GVL
(%)
= 24.5+17.4⋅X
1
+ 7.4⋅X
2
−9.7⋅X
3
−7.5⋅X
−6.9⋅X
+ 4.3⋅X
1
⋅X
3
0
0
1.5
In this equation, YGVL represents the GVL yield in mol% units. X
i
0
-1
+1
0.5
refers to the independent coded factors, which adopt values
between -1 and +1. This mathematical model is valid within the
explored experimental region. The value of the regression
0
0
6.5
5.6
a
b
c
Coded temperature; Coded catalyst loading; Coded xylose starting
d
concentration. Yields calculated as (mol produced/mol starting xylose)x100,
obtained after 24h of reaction. Products: iL, isopropyl levulinate; LA, levulinic
acid; LACT, α/β-angelica lactones; GVL, γ-valerolactone.
2
coefficient, r = 0.9447, indicated a high degree of correlation
between the experimentally observed and the predicted values for
GVL yield (Fig. ESI-6). Figure 4 depicts the contour plots showing the
predicted influence of the operation conditions (temperature,
catalyst loading and starting xylose concentration) on the obtained
The experimental catalytic results were used to test, in a first step,
the variability of the analysis procedure by comparing the results
6
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 9
Pleas Ge rdeo e nn o Ct ah de j mu si ts mt r ya rgins
View Article Online
DOI: 10.1039/C7GC01969F
Journal Name
ARTICLE
yield towards GVL, based on the generated model. The influence of As for the influence of the reaction temperature (X
the distinct operation parameters has been evaluated through the starting xylose concentration (X ), both factors play a strong
analysis of the response surfaces depicted in Fig. 4. Thus, the influence, though opposite, on the yield towards GVL. Thus, the
catalyst loading (X ), though significant, seems to exert a limited reaction temperature exerts a positive influence on the response,
1
) and the
3
2
influence on the GVL yield. This influence is positive, so that the because an increase in temperature, accordingly increases the
higher catalyst loading, the higher the GVL yield. However, the reaction rates of each one of the transformations involved in the
influence is linear, which is clearly evident from the equidistant proposed reaction cascade (Fig. ESI-7). Thus, elevating the reaction
separation of the response surfaces generated at three different temperature leads to a fast transformation of the starting xylose
catalyst loadings (Fig. 4A). This result is probably a consequence of and its 2-propyl ethers (xylosides), and the rest of the chemical
the studied range of catalyst loadings in quite dilute conditions, intermediates. However, raising the reaction temperature also
thus avoiding the presence of an excessive concentration of solid leads to an enhanced humins formation, which explains the
within the reactor and preventing mass transfer hindrances.
negative influence of the quadratic term of temperature (Eq. (4)).
This is evident from the dark colored solutions obtained when
operating at higher temperatures (Fig. ESI-7). Surprisingly, the
extension of the retro-aldol condensation reactions seems to be
independent on the reaction temperature, as no great differences
were observed in the production of 2-propyl lactate and 2-propyl
glycol when varying the reaction temperature. Nevertheless, the
overall influence of the temperature over the yield towards GVL,
under the tested experimental conditions, is positive, as it is clearly
visible in Fig.4.
As for the influence of the starting xylose concentration on GVL
production, this factor exerts a negative influence on the predicted
GVL yield (in both linear and quadratic terms). A detailed analysis of
the evolution of the reaction products distribution as a function of
the starting xylose concentration (Fig. ESI-8), revealed interesting
facts. Thus, lower xylose loadings conducted to higher yields
towards the products coming from the retro-aldol condensation of
2
-propyl xylosides (2-propyl glycol and 2-propyl levulinate). This
suggests that the proposed cascade of reactions from xylose to GVL,
through alternate H-transfer reactions and acid driven
transformations, is more sensitive to the concentration of the
starting sugar than the competing reaction pathway involving the
rupture of xylose. On the contrary, the formation of humins seems
to be favored when higher concentrations of the starting sugar are
used.
Analyzing the complete mathematical model, the maximal
predicted value for the yield towards GVL at 24 h within the
experimental region is 34.4 mol%. This optimal value corresponds
-1
to a reaction temperature of 190ºC, a catalyst loading of 15 g·L ,
-
1
and a starting xylose concentration of 30.5 g·L (X
1 2 3
= +1, X = +1, X
=
-0.31, in coded values). A catalytic experiment carried out under
these reaction conditions over the optimized catalyst (Zr-Al-Beta-6)
provided 34.3% YGVL, in fairly good agreement with the model
prediction. Noticeably, under these optimized reaction conditions,
the catalyst reaches its maximum value of GVL yield (ca. 34%)
already after only 10 h (Fig. ESI-9).
Figure 4. Surface and contour plots obtained for the predicted influence of
the operation conditions on the yield to GVL at 24h. A) Surface plots as a
-1
-1
-1
function of the catalyst loading (red: 5 g·L ; green: 10 g·L ; blue: 15 g·L ). B)
Contour plots representing GVL yield as a function of the starting xylose
Stability and reusability of the catalyst was evaluated under the
optimized reaction conditions. In this way, a hot filtration test
discarded the leaching of active metal species to the reaction
medium, which could lead to a misleading homogeneous catalysis
-1
concentration and the temperature (catalyst loading 15 g·L ).
(
Fig. ESI-10). Additionally, for the reutilization experiments, Zr-Al-
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
Green Chemistry
Page 8 of 9
View Article Online
DOI: 10.1039/C7GC01969F
ARTICLE
Journal Name
Beta-6 was recovered by filtration after a first reaction cycle of 24 h. Financial support from the Spanish Ministry of Economy and
The sample, once dried, presented dark color, which can be Competitiveness through the projects CTQ2014-52907-R and
attributed to the adsorption of reaction products and non-identified CTQ2015-68844-REDT, as well as from the Regional Government of
compounds, i.e. humins. In order to remove the organic deposits Madrid through the project S2013-MAE-2882, is gratefully
from the catalysts surface, a thermal treatment in air (5 h at 550ºC) acknowledged. C. López-Aguado thanks the Spanish Ministry of
was applied. Thereafter, the catalyst was used again in a second Economy and Competitiveness for a FPI grant (BES-2015-072709).
identical reaction cycle. The procedure was repeated for a third
cycle. Figure 5 includes the results in terms of yields to the
References
identified products. As shown, Zr-Al-Beta-6 displayed a very good
reusability in the transformation from xylose to GVL, evidencing
that the active sites keep their activity essentially intact even after
an accumulated reaction time of 72h.
1
J. Q. Bond JQ, D. M. Alonso, D. Wang, R. M. West, J. A.
Dumesic, Science, 2010, 327, 1110-1114.
J. J. Bozell, Science, 2010, 329, 522-523.
J.P. Lange, R. Price, P. M. Ayoub, J. Louis, L. Petrus, L. Clarke,
H. Gosselink, Angewandte Chemie International Edition,
2
3
4
4
3
3
2
2
1
1
5
0
5
0
5
0
5
0
5
0
2
010, 49, 4479-4483.
GVL
4
5
I. T. Horvath, H. Mehdi, V. Fabos, L. Boda, L. T. Mika, Green
Chemistry, 2008, 10, 238-242.
S. Van de Vyver, Y. Roman-Leshkov, Catalysis Science &
Technology, 2013, 3, 1465-1479.
K. Yao, C. Tang, Macromolecules, 2013, 46, 1689-1712.
D. Fegyverneki, L. Orha, G. Láng, I. T. Horváth, Tetrahedron,
2-propyl-lactate
2-propoxy-ethanol
6
7
2
010, 66, 1078-1081.
D. M. Alonso, J. Q. Bond, J. A. Dumesic, Green Chemistry,
010, 12,1493-1513.
R. Palkovits, Angewandte Chemie International Edition, 2010,
, 4336-4338.
8
9
1
2
4
9
0 K. Yan, Y. Yang, J. Chai, Y. Lu, Applied Catalysis B:
Environmental, 2015, 179, 292-304.
1 M. J. Gilkey, B. Xu, ACS Catalysis, 2016, 6, 1420-1436.
2 A. Osatiashtiani, A. F. Lee, K. Wilson, Journal of Chemical
Technology and Biotechnology, 2017, 92, 1125-1135.
3 M. Chia, J. A. Dumesic, Chemical Communications, 2011, 47,
1
1
1
1
1
1
1
2
3
12233-12235.
4 X. Tang, H. Chen, L. Hu, W. Hao, Y. Sun, X. Zeng, L. Lin, S. Liu,
Applied Catalysis B: Environmental, 2014, 147, 827-834.
5 J. He, H. Li, Y. M. Lu, Y. X. Liu, Z. B. Wu, D. Y. Hu, S. Yang,
Applied Catalysis A: General, 2016, 510, 11-19.
Use
Figure 5. Reutilization of Zr-Al-Beta-6. Yields to products starting from xylose
in three consecutive uses with intermediate thermal activation of the
catalyst. Reaction conditions: 190ºC, 24h, catalyst loading 15 g·L , xylose
6 J. Iglesias, J. A. Melero, G. Morales, J. Moreno, Y. Segura, M.
-1
Paniagua, A. Cambra, B. Hernández, Catalysts, 2015,
927.
7 Y. Kuwahara, W. Kaburagi, Y. Osada, T. Fujitani, H.
5, 1911-
-1
concentration 30.5 g·L .
1
1
Yamashita, Catalysis Today, 2017, 281, 418-428.
Conclusions
18 S. S. Enumula, V.R.B. Gurram, M. Kondeboina, D. R. Burri,
S.R.R. Kamaraju, RSC Advances, 2016, , 20230-20239.
9 H. Li, Z. Fang, S. Yang, ACS Sustainable Chemistry &
Engineering, 2016, , 236-246.
6
Within this work we have shown that for the synthesis of Zr-Al-Beta
zeolite, via dealumination and post-synthetic grafting of Zr, there is
1
4
an optimal Al/Zr ratio around 0.20. Likewise, the use of Zr(NO
3
)
4
as 20 L. Bui, H. Luo, W. R. Gunther, Y. Román-Leshkov,
Angewandte Chemie International Edition, 2013, 52, 8022-
zirconium source instead of ZrClO , combined with a proper
2
8
1 J. Wang, S. Jaenicke and G. K. Chuah, RSC Advances, 2014, 4,
13481-13489.
2 H. P. Winoto, B. S. Ahn, J. Jae, Journal of Industrial and
Engineering Chemistry, 2016, 40, 62-71.
025.
adjustment of water concentration during the impregnation, allows
for an improved catalytic activity in the cascade of reactions from
xylose to GVL, which is attributed to a better dispersion of Zr
species on the zeolite. Modelling of the catalytic system identified
2
2
2
3 G. Wu, W. Dai, N. Guan, L. Li, Applied Catalysis B:
Environmental, 2017, 205, 393-403.
-
the optimal values of temperature (190 ºC), catalyst loading (15 g·L
1
-1
)
, and xylose concentration (30.5 g·L ), in the analyzed 24 B. Hernández, J. Iglesias, G. Morales, M. Paniagua, C. López-
Aguado, J. L. García-Fierro, P. Wolf, I. Hermans, J. A. Melero,
Green Chemistry, 2016, 18, 5777-5781.
5 J. Wang, K. Okumura, S. Jaenicke, G.-K. Chuah, Applied
Catalysis A: General, 2015, 493, 112-120.
experimental range. Under these optimized reaction conditions, the
optimized Zr-Al-Beta catalyst reaches its maximum value of GVL
yield from xylose (ca. 34%) after only 10 h of reaction. The catalyst
showed a good reusability after thermal regeneration in the
reaction from xylose, confirming its stability.
2
Acknowledgements
8
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 9
Green Chemistry
View Article Online
DOI: 10.1039/C7GC01969F
Efficient one-pot production of γ-valerolactone from xylose over Zr-Al-
Beta zeolite: rational optimization of catalyst synthesis and reaction
conditions
a,
a
a
a
a
Juan A. Melero *, Gabriel Morales , Jose Iglesias , Marta Paniagua , Clara López-Aguado , Karen
b
b
Wilson , Amin Osatiashtiani
Green Chemistry
Graphical Abstract