Mechanistic insights into the production of methyl lactate by catalytic conversion of carbohydrates on mesoporous Zr-SBA-15

Article abstract of DOI:10.1016/j.jcat.2015.10.013

The as-synthesized Zr-SBA-15 catalysts with tunable mesoporous structures showed excellent catalytic performance for the conversion of carbohydrates to methyl lactate in a "one-pot" process using near-critical methanol or methanol-water mixture as the solvents. The effects of reaction conditions, including temperature, reaction time, and catalyst loading amount, on the conversions of carbohydrates and the yields of methyl lactate were investigated. The high yields of methyl lactate, up to 41% and 44%, were produced from pentose and hexose, respectively, in the near-critical methanol at 240 °C. Moreover, the Si/Zr ratio of the Zr-SBA-15 catalysts profoundly affected the Lewis acidity and therefore the catalytic activity and selectivity to methyl lactate in the conversion of carbohydrates. The pore size of the Zr-SBA-15 catalysts, tuned by the synthesis temperature, strongly affected the formation of solid residues. The key intermediates such as glyceraldehyde, glycolaldehyde, and pyruvaldehyde were used as probe reactants to understand the mechanism. The role of the Zr-SBA-15 catalyst in the aldol- and retro-aldol condensation, isomerization, and Cannizzaro reactions of carbohydrates and their derivatives was discussed. Furthermore, 28% and 27% yields of methyl lactate were obtained from cellulose and starch, respectively, in methanol-water mixture (5 wt% water and 95 wt% methanol) at 240 °C. The Zr-SBA-15 catalyst was relatively stable in short term without regeneration.

Full text of DOI:10.1016/j.jcat.2015.10.013

Journal of Catalysis 333 (2016) 207–216
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Mechanistic insights into the production of methyl lactate by catalytic
conversion of carbohydrates on mesoporous Zr-SBA-15
Lisha Yang a,1, Xiaokun Yang a,1, Elli Tian , Vivek Vattipalli , Wei Fan , Hongfei Lin
b
c
c
⇑
a, ,1
a
Department of Chemical and Materials Engineering, University of Nevada, Reno, 1664 N. Virginia St., Reno, NV 89557, USA
Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD 21205, USA
Department of Chemical Engineering, University of Massachusetts Amherst, Amherst, MA 01003, USA
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
The as-synthesized Zr-SBA-15 catalysts with tunable mesoporous structures showed excellent catalytic
performance for the conversion of carbohydrates to methyl lactate in a ‘‘one-pot” process using near-
critical methanol or methanol–water mixture as the solvents. The effects of reaction conditions, including
temperature, reaction time, and catalyst loading amount, on the conversions of carbohydrates and the
yields of methyl lactate were investigated. The high yields of methyl lactate, up to 41% and 44%, were pro-
duced from pentose and hexose, respectively, in the near-critical methanol at 240 °C. Moreover, the Si/Zr
ratio of the Zr-SBA-15 catalysts profoundly affected the Lewis acidity and therefore the catalytic activity
and selectivity to methyl lactate in the conversion of carbohydrates. The pore size of the Zr-SBA-15 cat-
alysts, tuned by the synthesis temperature, strongly affected the formation of solid residues. The key
intermediates such as glyceraldehyde, glycolaldehyde, and pyruvaldehyde were used as probe reactants
to understand the mechanism. The role of the Zr-SBA-15 catalyst in the aldol- and retro-aldol condensa-
tion, isomerization, and Cannizzaro reactions of carbohydrates and their derivatives was discussed.
Furthermore, 28% and 27% yields of methyl lactate were obtained from cellulose and starch, respectively,
in methanol–water mixture (5 wt% water and 95 wt% methanol) at 240 °C. The Zr-SBA-15 catalyst was
relatively stable in short term without regeneration.
Received 30 May 2015
Revised 17 September 2015
Accepted 19 October 2015
Keywords:
Methyl lactate
Carbohydrate biomass
Mesoporous silica
Zr-SBA-15
Lewis acid
Ó 2015 Elsevier Inc. All rights reserved.
1
. Introduction
The development of efficient methods to make fuels and chem-
Shape selectivity enables excellent adjustment of catalytic trans-
formation exclusivity, and also acts on the activity and stability
of the catalyst by either protecting the acid sites from potential
contaminants (in particular, coke precursors) which are contained
in the feeds, or by inhibiting the formation of coke precursors in
the pores. However, few reports exist on biomass conversion with
mesoporous silicate materials. This scarcity of literature is due to
the fact that a pure mesoporous silica material possesses a neutral
framework, as well as its propensity to exhibit inherently negative
traits such as poor hydrothermal stability and low catalytic activ-
ity, which limit its application as a catalyst medium [14,15]. In
order to enable their application in catalysis, different structured
mesoporous silicate materials have been investigated and various
metal ions have been incorporated into the host mesoporous silica
materials [16–20]. In particular, the isomorphic substitution of sil-
icon with transition metals has proven to be an excellent strategy
to generate catalytically active sites in mesoporous silicate materi-
als [14,21–23].
icals from lignocellulosic biomass is vital to reduce the dependence
on fossil fuel feedstocks [1–3]. The production of value-added
chemicals from carbohydrates using non-toxic heterogeneous cat-
alysts is an appealing environmentally benign process. Various
endeavors have been made over the past few years to develop
stable, recyclable solid catalysts for the conversion of biomass by
catalytic methods [4–7]. Microporous zeolites are efficient cata-
lysts in the transformation of small molecules; however, they
become unsuitable when the essential reactants involved in the
catalytic process are sized comparably with the zeolite pore
dimensions. The rational approach to overcoming mass-transfer
limitations would be to increase the diameter of the pores, thus
bringing them into the mesoporous range with a key factor here
being the shape selectivity properties of these catalysts [8–13].
Among the mesoporous materials, the SBA-15 of 2D hexagonal-
ordered structure with tunable pores in 4–10 nm has received
much attention in the past decades due to its relatively large pore
⇑
Corresponding author.
E-mail address: HongfeiL@unr.edu (H. Lin).
Fax: +1 775 327 5059.
1
http://dx.doi.org/10.1016/j.jcat.2015.10.013
021-9517/Ó 2015 Elsevier Inc. All rights reserved.
0

2
08
L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
size and high hydrothermal stability in comparison with other
mesoporous silica materials, such as MCM-41, its analog in M41S
family [24]. This large pore channel network provides a distinctive
open space, with easy and direct access for both guest and host
species, thus facilitating inclusion and/or diffusion throughout
the pore channels without pore blockage. Such properties spur
the prospective utilization as catalysts and adsorbents [21,25,26].
On the other hand, zirconia-based materials have been widely
applied in catalyzing various types of reactions such as oxidation,
dehydration, hydrogenation, and hydroxylation [27–30]. Meso-
porous silicate materials containing zirconium have high special
surface areas and potential Lewis acid properties [14,31]. The basic
structural unit of mesoporous silicate frameworks consists of a sil-
icon atom that is coordinated with four oxygen atoms. Zirconium
atoms have a coordination number of 7 or 8 in zirconia materials.
can act not only as a solvent but also as a reactant, which can serve
as a hydrogen donor agent to remove oxygen from biomass and a
radical quenching agent to retard repolymerization and formation
of humins [62–65]. Methanol, a small and highly polar molecule,
still exhibits weak hydrogen bonding even at the critical tempera-
c
ture (T = 239.4 °C) [66,67], which facilitates the methanolysis of
large biomass molecules.
Herein we report our findings on the catalytic conversion of car-
bohydrates to methyl lactate using a mesoporous Zr-SBA-15 cata-
lyst in near-critical methanol solvents (T < 240 °C), which
combines the lactic acid production and esterification into a
‘‘one-pot” reaction system. A possible reaction mechanism and
structure–activity relationship are proposed to explain the perfor-
mance of the Zr-SBA-15 as a heterogeneous Lewis acid catalyst in
the production of ML from various carbohydrates including pen-
tose, hexose, starch and cellulose.
4
+
4+
When replacing Si with Zr , a zirconium atom has only 4 coordi-
nated oxygen atoms, resulting in empty zirconium d-orbitals
which can act as electron acceptors, i.e. Lewis acid sites [32]. Hane-
feld and co-workers found that Zr-TUD-1 with zirconium ion incor-
porated in three-dimensional mesoporous TUD-1 silicate possesses
predominately Lewis acidity [33,34].
2
. Experimental section
2.1. Materials
Carbohydrates constitute the largest portion of lignocellulosic
biomass, and various strategies for their efficient use as a commer-
cial chemical feedstock as a petroleum supplement are being
established to synthesize value-added chemicals. The synthesis of
lactate acid esters in related alcohols with renewable carbohydrate
biomass as the feedstock invokes lots of interests in that lactate
acid esters are ‘‘green” solvents which have numerous applications
in the chemical, food, pharmaceutical, and cosmetic industries
The following reagents and products were used as received
without further purification.
D-(+)-xylose (99%), D-(+)-glyceraldehyde
(
98%), Fructose (99%), Sucrose (99%), glycolaldehyde dimer, pyruvalde-
hyde (40 wt% solution in water), furfural (99%), 5-(hydroxymethyl) fur-
fural (99%), hydrochloric acid (36.5–38.0%, BioReagent), triblock
copolymer Pluronic P123, tetraethyl orthosilicate (>99.0%), n-butanol
(
>99.0%), and zirconyl chloride octahydrate (98%) were purchased from
Sigma Aldrich. (+)-Glucose (Reagent ACS Grade) were purchased from
D
[
35–39]. Various strategies to produce lactic acid esters have been
Acros Organics. Methyl lactate (97%), erythrose syrup (70% w/v),
methyl levulinate (99%), methyl glycolate (98%) and glycolaldehyde
dimethylacetal (98%) were purchased from Alfa Aesar. Microcrystalline
established including both fermentative processes [40] and cat-
alytic transformations [41,42]. Catalytic methods have many
advantages over that involving fermentation method especially
considering the latter approach’s unavoidable large amounts of
salts by-products, which impose a high environmental remediation
cost. Studies on the conversion of carbohydrates to lactate esters
over heterogeneous catalysts have emerged recently. For instance,
Holm et al. [43,44] reported that Lewis acidic zeotype materials,
such as Sn-Beta, catalyzed the conversion of mono- and disaccha-
rides to methyl lactate (ML) in methanol at 160 °C with a 16-h
reaction time. With sucrose as the substrate, the ML yield reached
cellulose (average particle size 50 lm) and cellobiose (98%) were
purchased from Acros Organics. Starch (powder, certified ACS, soluble)
and sucrose (crystalline, certified ACS) were purchased from Fisher
Scientific. Galactose, mannose, and arabinose were purchased from
Carbosynth.
2
.2. Catalyst preparation
The Zr-SBA-15 materials were synthesized following the proce-
dure described by Ref. [24]. Briefly, 2 g of Pluronic P123 was added
2 3
a record high of 68%, while when the alkali ion of K CO was added
to the solvent mixture, the ML yield was further promoted to 75%
at 170 °C [45]. Carlos reported that with the Sn-MCM-41 which has
an atomic ratio of Si/Sn = 55, a 43% yield of ML was produced from
the conversion of glucose after 20 h at 160 °C [46]. The carbon–
silica composite grafted Sn(IV) showed well-balanced Lewis/
Brønsted acidity and yielded 45% ML from the conversion of
sucrose in methanol [47]. Sn-MWW zeolite was also demonstrated
to be an effective and selective catalyst for the direct conversion of
mono- and disaccharides to ML [48]. However, the main draw-
backs of tin-based materials are the typically complex and lengthy
synthesis process and the toxicity of tin precursors, which may
hamper their industrial applications [49]. Solid base catalysts,
including hydrotalcites [50] and magnesium oxide [51], and sup-
ported noble metal catalysts [41] were also used for catalyzing
the formation of lactic acid or ML from glucose at rather low yields.
Recently, alcohols have been utilized as an alternative solvent in
the liquefaction of various types of biomass, including cellulose,
lignin, sewage sludge, and microalgae, due to their advantages
with better solubility of organic intermediates, hydrogen donor
properties, and easier separation due to their low boiling points
to 75 mL of 1.6 M HCl solution. The mixture was stirred at 40 °C for
3
h until all P123 was dissolved. Next, 4.25 g of TEOS and an appro-
priate amount of zirconyl chloride octahydrate were added into the
solution and the mixture was stirred for another 24 h at 40 °C. The
resulting gel was placed in the Teflon-lined autoclave and heated
at a range of temperatures of 80–150 °C for 24 h. The solid product
was filtered with mild washing, dried at 100 °C overnight, and cal-
cined in flowing air at 550 °C for 6 h. The x (in Zr-SBA-15-x-y°C)
represents the mole ratio of Si/Zr, and the y represents the
hydrothermal temperature. Zr-SBA-15-y°C without x means that
the mole ratio of Si/Zr is 20. Zr-SBA-15-x without y means that
the catalyst was synthesized at 100 °C. Zr-SBA-15 without x and
y means that the catalyst was synthesized at 100 °C with Si/Zr = 20.
2.3. Catalyst characterization
Small-angle X-ray scattering (SAXS) was performed using a
sample-to-detector distance of 172.1 cm, which provided a two-
theta range of approximately 0.3–2.0°. Data were typically col-
lected over 30 s at a temperature of 20 °C. The X-ray source was
[
52–61]. Compared to water, alcohols such as methanol and etha-
nol have much lower critical temperatures and pressures. Thus at
relatively mild conditions, near-critical and supercritical alcohols
Cu Ka radiation with a wavelength of 1.54 Å, which was generated
by a Rigaku Ru-200BVH rotating anode. Measurements were made

L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
209
on a Siemens HI-STAR multi-wire area detector and were corrected
for background and nonlinearities in the detector. Integration of
the 2D measurement provided a 1D plot of intensity (arbitrary
units) versus the two-theta scattering angle, which was peak-
fitted using Material Data Incorporated’s JADE.
In the catalyst stability tests, the spent catalysts were collected
and dried overnight in an oven at 105 °C, and then were reused
without further regeneration under the identical reaction
conditions.
Transmission electron microscope (TEM) micrographs were
captured using JEOL-JEM 2100F operating at 200 kV. The samples
were dispersed in 1-butanol, and a drop of the suspension was
placed on lacey carbon supported on 300 mesh copper grids.
2
.5. Product analysis
After the reaction, the resultant liquid phase product samples
were prepared for analysis with a gas chromatography coupling
with flame ionization detector (GC–FID), a high performance liquid
chromatography (HPLC) and a gas chromatography coupling with
mass spectrometer (GC–MS).
The liquid products (e.g. furfural, HMF) were quantified by HPLC
analysis using a Shimadzu HPLC system equipped a UV–VIS Detec-
tor (Shimadzu SPD 10-AV) and Refractive Index Detector (Shi-
madzu RID-6A). The liquid phase after reaction was filtered
through a 0.45
DI water. The samples were separated in an Aminex 87-H column
from Bio-Rad, using 5 mM H SO as the mobile phase (0.7 mL/min
flow rate) at a column temperature of 55 °C. The UV–VIS detector
was utilized at 208 nm and 290 nm.
2
N physisorption isotherms were measured on an Autosorb-iQ
system (Quantachrome) at 77 K. Outgassing was done at 523 K
until pressure rise in the test cell was less than 25 mTorr/min. Pore
size distribution and cumulative adsorbed volume were calculated
by using nonlocal density functional theory (NLDFT) adsorption
2
model which describes N adsorbed onto silica at 77 K in cylindri-
cal pores (AsiQwin 1.02, Quantachrome). NLDFT model considers
the configuration of adsorbates in pores on a molecular level and
is widely used to characterize ordered porous materials with dif-
ferent pore geometries. With adequate fluid–fluid and fluid–solid
interaction parameters, it has been used to quantitatively predict
the capillary condensation and evaporation transitions of adsor-
bates in mesoporous materials.
lm syringe filter, and then diluted 10 times with
2
4
The liquid products identified in the methanol were qualified
and quantified by GC–MS and GC–FID analysis, respectively. The
3
In a typical NH temperature programmed desorption (TPD)
experiment using a Micromeritics AutoChem II 2920 Chemisorp-
tion Analyzer, the catalyst was first degassed in helium at 250 °C
for 1 h, and then the temperature was cooled till 100 °C in helium
flow. After that, 10% ammonia in helium was adsorbed on the cat-
alyst at 100 °C for 60 min and then the helium flow at 50 mL/min
was used to remove physically adsorbed ammonia. Finally, the
NH TPD was carried out from 100 °C to 550 °C with a temperature
3
ramp of 10 °C/min.
Drift-IR study was performed on EQUINOX 55 (Bruker)
equipped with a MCT detector. The samples were degassed at
liquid phase after reaction was filtered through a 0.45 lm syringe
filter before being diluted 10 times with methanol. The sample was
injected in an Agilent 6890 series GC–MS equipped with an Agilent
DB5-MS column (30 m ꢁ 0.25 mm ID, 0.25
lm film thickness) and
an Agilent 5973 Mass Selective Detector. The same prepared sam-
ple was also injected in a Shimadzu GC-2010 equipped with an
SHRXI-5MS column (30 m ꢁ 0.25 mm ID, 0.25
lm film thickness)
and an FID detector.
A Euro EA3000 CHNS-O analyzer (Eurovector) was used to mea-
sure the carbon content of the solid residue samples.
5
50 °C for 1 h under helium in a high temperature reaction cham-
TM
ber containing a Praying Mantis diffuse reflection attachment
3
. Results and discussion
(
Harrick). Small aliquots of pyridine were carried by helium and
exposed to the sample at room temperature for 15 min. Prior to
the characterization, the physically adsorbed pyridine was
removed by flowing helium at 250 °C under helium for 1 h. All
spectra were collected at 120 °C.
Fig. 1 shows the yields of the main products using xylose as the
feedstock with and without adding catalysts. A relatively low yield
of ML (ꢂ8%) was produced at 240 °C without adding a catalyst.
With only the pure SBA-15, the ML yield was ꢂ7.0%, which is close
to that without catalyst. Both the as-synthesized Zr-SBA-15 and
2.4. Catalytic reactions
the commercial ZrO
2
resulted in noticeably higher yields of ML at
2
40 °C for 1 h. The 35.9% yield of ML was achieved with the
Reactions were carried out in a 100 mL stirred Parr micro reac-
tor, whereby the catalyst was suspended in a solution of biomass
substrate in methanol (20 mL) and the reactor was charged with
4
2
00 psi N initially and then heated at a ramp rate of 10 °C/min
until the desired set temperature was reached. During the reaction,
mixing was achieved through an internal propeller operating at
7
00 RPM. Once the set temperature was attained, the reactor
was held for the set reaction time, and then quenched quickly in
an ice bath to stop the reaction. The reactor was cooled to
approximately 25 °C before being vented after the gas pressure
was recorded. The reactor was then immediately broken down
and the solid residue remaining in the reactor was recovered and
dried. The aqueous and solid fractions were separated using a
centrifuge.
The solid residue (S.R.) yield and the product carbon yield are
calculated using the following equations:
moles of carbon in the dried solid residue
S:R: Yield ¼
moles of carbon in the biomass feed
Product Carbon Yield
Fig. 1. Comparison of different catalysts toward the conversion of xylose in
methanol. Reaction conditions: 240 °C, 1 h, 400 psi N , 0.2 g xylose loading, and
0.1 g catalyst. ML: methyl lactate, GADMA: glycolaldehyde dimethyl acetal, MG:
2
moles of carbon in product ꢀ moles of carbon in methyl groups
¼
moles of carbon in the biomass feed
methyl glycolate, MLE: methyl levulinate.

2
10
L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
Zr-SBA-15 catalyst, which was more than twice of the 16.4% yield
with the ZrO catalyst. Notably, without a catalyst, there were neg-
ligible amount of furfural yields, 0.7%. In contrast, when SBA-15,
ZrO and Zr-SBA-15 were employed, the furfural yields reached
Zr-SBA-15 catalysts with various Si/Zr molar ratios under
otherwise identical reaction conditions. Initially, with increasing
the Zr loading, the yields of increased steadily. However, as the zir-
conia loading further increased till the Si/Zr mole ratio of 10, the
yields of all three products decreased implying the abrupt change
of the catalyst properties. The maximum yield of ML was found to
be in the range of the Si/Zr mole ratios of 40–20, e.g., with the Zr-
SBA-15-40 or Zr-SBA-15-20 catalyst, and approximately 19% ML
and 11% GADMA were produced at 180 °C. However over-loading
of Zr ions onto SBA-15 inhibited the catalyst’s performance. There-
fore, to confirm the structures of the Zr-SBA-15 silicate materials at
different Si/Zr ratios, the small-angle X-ray scattering (SAXS) char-
acterization was performed. As depicted in Fig. 4, the SAXS spectra
exhibited the strong (100), (110) and (200) diffraction peaks at 2h
angles between 0.5° and 2° for the samples with the Si/Zr ratios
from 100 to 20, which indicated the structural ordering with the
symmetry of the 2D-hexagonal space group p6mm [9,14,31]. The
powder XRD spectra (Fig. S1) displayed the broad diffraction peak
at 2h of 20–30° for all the Zr-SBA-15 samples which corresponds to
the amorphous structure of the silicate materials. However, a fur-
ther increase in Zr loading (Si/Zr = 10) drastically lowered the peak
intensity suggesting that incorporating excess Zr heteroatoms was
detrimental on the structure of the mesoporous SBA-15 frame-
work, which was coincident with its inferior catalytic performance.
The high-resolution TEM images (Fig. S2) confirmed that highly
ordered pore structure was evident for the Zr-SBA-15 materials
at the Si/Zr mole ratio of 20, while the ordered mesoporous struc-
ture was partially collapsed when the Si/Zr mole ratio reached 10,
which were consistent with the SAXS data. Although the total acid-
2
2
to 5.1%, 6.4% and 7.9%, respectively. The mesoporous structure of
the silica framework with the presence of zirconium ions confers
strong Lewis acidity as well as weak Brønsted acidity [68]. It is well
accepted that Lewis acid catalysis can facilitate the retro-aldol con-
densation of a sugar molecule, which is the initial step in the con-
version of sugars to lactic acid [43], while a Brønsted acid catalyzes
the dehydration of xylose to form furfural. The co-production of ML
and furfural from xylose with the Zr-SBA-15 catalyst suggests that
both the Lewis and Brønsted acidic properties may coexist on the
catalyst surface.
To further examine the catalytic effect of zirconium loading of
the Zr-SBA-15 catalyst, the catalysts with different Si/Zr molar
ratios of 100, 60, 40, 20 and 10 were prepared. As shown in
Fig. 2, the total acid strength analyzed by the NH
with increasing the zirconium loading on the SBA-15 silicate.
Fig. shows the yields of the three major products, ML,
glycolaldehyde dimethyl acetal (GADMA), and furfural, on the
3
TPD increased
3
3
ity of the Zr-SBA-15-10 sample (0.74 mmol NH /g) was the highest,
Zr-SBA-15-10 had the inferior performance compared to the Zr-
SBA-15-20 sample. Note that the total acidities of both samples
were very close even the Zr content of Zr-SBA-15-10 was doubled.
Therefore, the collapse of the mesoporous structure of the Zr-SBA-
1
5-10 sample implied the overloading of Zr content, which may
lead to the formation of ZrO nanoclusters that have multi-
2
functional surface properties, and thus the selectivity to ML
became lower on Zr-SBA-15-10.
To maximize ML yield, process conditions were optimized for
the xylose conversion reactions. As shown in Fig. 5, the varying
temperature has a pronounced effect on the production of ML,
yielding a steadily increasing amount of ML up to 35.9% with an
increase in temperature from 160 °C to 240 °C for a 1-h reaction.
The yields of other products such as furfural also increased steadily
3
Fig. 2. Temperature-programmed desorption of ammonia (NH -TPD) for Zr-SBA-15
materials with different Si/Zr molar ratios. All materials were synthesized at the
hydrothermal temperature of 100 °C.
Fig. 3. Comparison of different Zr4+ ion loading on SBA-15 toward the conversion of
xylose in methanol. Reaction conditions: 180 °C, 60 min, 400 psi initial N
.2 g xylose, and 0.1 g catalyst. ML: methyl lactate, GADMA: glycolaldehyde
dimethylacetal.
2
pressure,
Fig. 4. Small-angle X-ray scattering patterns of Zr-SBA-15 materials with different
Si/Zr molar ratios. All materials were synthesized at the hydrothermal temperature
of 100 °C.
0

L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
211
Fig. 5. Effect of temperature on the yields of liquid-phase products of xylose
conversion with the Zr-SBA-15 catalyst in methanol. Reaction conditions: (a) 1 h,
Fig. 7. Effect of catalyst loading on the yields of liquid-phase products of xylose
conversion with the Zr-SBA-15 catalyst in methanol. Reaction conditions: 240 °C,
4
2 2
00 psi initial N pressure, 0.2 g xylose, and 0.1 g catalyst; (b) 10 h, 400 psi N , 0.2 g
1
2
h, 400 psi initial N pressure, 0.2 g xylose. ML: methyl lactate, GADMA:
xylose, and 0.1 g catalyst. ML: methyl lactate, GADMA: glycolaldehyde dimethy-
lacetal, MG: methyl glycolate, MLE: methyl levulinate.
glycolaldehyde dimethylacetal, MG: methyl glycolate, MLE: methyl levulinate.
adding any catalyst. Thus the Lewis acid property of the Zr-SBA-
1
5 catalyst appeared to vanish and be transformed to Brønsted acid
in water.
For chemical reactions in confined channel space in mesoporous
silicate materials, interesting pore size effects can be expected [69–
1]. The pore size of the Zr-SBA-15 catalyst can be tuned by opti-
7
mizing the hydrothermal treatment conditions during the synthe-
sis [72]. Fig. S3 shows the pore size distribution determined by
NLDFT model for the Zr-SBA-15 materials with different synthesis
temperatures. The pore sizes of Zr-SBA-15-80 °C, Zr-SBA-15-
1
00 °C, Zr-SBA-15-120 °C and Zr-SBA-15-150 °C were 7.6, 9.1, 9.8
and 10.6 nm, respectively (Table 2). With increasing the hydrother-
mal temperature, the pore diameter of the SBA-15 materials
increased while the BET surface area decreased. The ML yield var-
ied slightly with increasing the pore size except the catalyst syn-
thesized at 80 °C in the smallest pore size, as shown in Table 3.
In contrast, the furfural yield and the corresponding solid residue
yield decreased significantly with increasing the pore size. It is well
known that aldehyde compounds such as furfural are prone to
polymerize and form humins at elevated temperatures. Appar-
ently, large pores in a Zr-SBA-15 catalyst provide more open pore
space where guest and host species can easily access, thus facilitat-
ing the diffusion in and out the stoma channels without pore
restriction. It turns out that in order to minimize the undesirable
by-products such as furfural and humins, it is important to lift
the pore diffusion limit of the Zr-SBA-15 catalyst for this particular
ML production reaction.
Fig. 6. Effect of reaction time on the yields of liquid-phase products of xylose
conversion with the Zr-SBA-15 catalyst in methanol. Reaction conditions: 240 °C,
4
2
00 psi initial N pressure, 0.2 g xylose, and 0.1 g catalyst. ML: methyl lactate,
GADMA: glycolaldehyde dimethylacetal, MG: methyl glycolate, MLE: methyl
levulinate.
with increasing temperature. The yield of GADMA, however,
decreased from 10.9% at 180 °C to 1.6% at 240 °C. The highest ML
yield of 40.8% was achieved when the reaction time extended to
6
h at 240 °C as shown in Fig. 6. However, the ML yield varied little
as the reaction time was longer than 3 h. Conversely, the yields of
GADMA and furfural consistently decreased with extending the
reaction time. Of particular interest is the fact that GADMA almost
completely vanished after 3 h at 240 °C. These results suggested
that longer reaction time leads to the decomposition of GADMA
and furfural. It is also interesting to find that the methyl levulinate
yield increased with reaction time at 240 °C. The effects of different
catalyst loading amounts on the conversion of xylose were also
examined. As depicted in Fig. 7, the yields of both ML and furfural
showed a similar uptrend with constantly increasing catalyst load-
ing, while those of GADMA and methyl glycolate decreased stea-
dily. However, as the mass ratio of catalyst to xylose was larger
than 0.2, the yield of ML was almost unchanged and reached to a
plateau. By substituting methanol with water as the solvent
Recyclability and reusability are of importance for the heteroge-
neous Zr-SBA-15 catalysts. During the typical reaction, the yield of
solid residue on the catalyst was relatively low, ꢂ10.8%. However,
the activity of the catalyst could decrease if solid residue continues
to build up on the catalyst surface in the subsequent repeating
reactions. Fig. 8 shows that the yield of ML from xylose only
decreased slightly over five consecutive runs with the reused cat-
alyst without regeneration, suggesting that the Zr-SBA-15 catalyst
was relatively stable, although a regeneration process to remove
the solid residues would be desirable for the catalyst reuse. It is
very interesting to find that the yield of solid residue (10.7%) on
the catalyst after the 5th run almost did not change compared with
that after the first run (10.8%), implying that coke did not accumu-
late inside the pores after reuse of the spent catalysts without any
regeneration in multiple repeated reactions. The high-resolution
TEM images of the Zr-SBA-15 samples before and after reaction
showed that the highly ordered pore structure was maintained
(
Table 1 Entry 4), the lactic acid yield was ꢂ5.9% while the furfural
yield was ꢂ42.3% with Zr-SBA-15. In contrast, the yields of furfural
and lactic acid were only 21.2% and 3.0%, respectively, without

2
12
L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
Table 1
Comparison of the conversion of different carbohydrate feedstocks with the Zr-SBA-15 catalyst in methanol and/or water solvents.
Entry
Feedstock
Catalyst
Time (h)
Solvent
Carbon yield (%)
ML
GADMA
MG
MLE
Furfural
HMF
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
Xylose
Xylose
Xylose
Xylose
Glyceraldehyde
Glyceraldehyde
Dihydroxy acetone
Dihydroxy acetone
Pyruvaldehyde
Pyruvaldehyde
Glycoaldehyde
Glycoaldehyde
GADMA
Xylose
Glucose
Sucrose
Fructose
Galactose
Mannose
Arabinose
Cellobiose
Cellobiose
Cellulose
/
1
1
1
1
1
1
1
1
1
1
1
1
1
6
6
6
6
6
6
6
10
10
10
10
10
10
Methanol
Methanol
Water
8.1
35.9
3.0 (Lactic acid)
5.9 (Lactic acid)
19.9
78.8
29.5
84.5
64.5
98.7
3.9
61.3
24.0
40.8
37.3
39.5
44.1
14.8
31.3
33.4
24.3
26.4
16.7
28.1
24.1
7.9
1.6
/
6.3
0.2
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
0.2
2.2
1.7
2.1
1.6
0.1
2.1
1.3
0.7
1.0
0.2
1.5
Zr-SBA-15
/
Zr-SBA-15
/
Zr-SBA-15
/
Zr-SBA-15
/
21.2
42.3
/
/
/
/
/
/
/
Water
/
/
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
95% Methanol
Methanol
95% Methanol
Methanol
95% Methanol
2.3
2.5
1.9
1.9
0.7
0.8
32.6
8.4
/
0.6
0.6
0.4
0.7
0.8
1.2
1.3
0.9
1.6
0.1
0.3
1.4
0.7
2.9
0.8
3.3
0.9
0.2
0.2
0.7
0.6
0.2
0.5
1.0
0.3
0.6
0.1
0.3
0.5
0.3
0.1
0.1
0.4
0.1
0.2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
Zr-SBA-15
/
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
Zr-SBA-15
/
/
/
2.7
1.3
2.5
3.1
7.2
8.9
5.4
4.8
3.1
0.7
2.0
0.9
1.1
3.2
0.9
1.6
1.7
0.9
1.3
3.7
1.9
2.3
0.2
1.8
0.1
0.6
Cellulose
Starch
Starch
26.8
2
Reaction conditions: 240 °C, 400 psi initial N pressure, 0.2 g biomass substrate, and 0.1 g Zr-SBA-15. ML: methyl lactate, GADMA: glycolaldehyde dimethyl acetal, MG:
methyl glycolate, MLE: methyl levulinate.
Table 2
The physicochemical properties of Zr-SBA-15 silicates.
Catalysta
BET surface
Pore volume determined by
Pore size determined by
NLDFT model (nm)
Acid strength NH
3
quantity (mmol/g)
b
2
ꢀ1
NLDFT model (cm3
c
ꢀ1
c
area (m
g
)
g
)
SBA-15-100 °C
876
988
841
618
393
0.998
1.275
1.357
1.384
1.215
7.0
7.6
9.1
9.8
10.6
0.02
0.65
0.72
0.69
0.51
Zr-SBA-15-80 °C
Zr-SBA-15-100 °C
Zr-SBA-15-120 °C
Zr-SBA-15-150 °C
a
b
c
The mole ratio of Si/Zr is 20 for the catalysts.
BET surface area calculated using the BET (Brunauer–Emmett–Teller) equation at relative pressures between 0.05 and 0.25 using the adsorption branch.
Cylindrical pore model applied for N adsorption on silica at 77 K for the adsorption branch.
2
Table 3
Comparison of the effect of different pore sizes of Zr-SBA-15 on the catalytic conversion of xylose to methyl lactate in methanol solvent.
Entry
Feedstock
Catalyst
Conversion (%)
Solid residue (%)
Carbon yield (%)
ML
GDA
MG
MLE
Furfural
1
2
3
4
Xylose
Xylose
Xylose
Xylose
Zr-SBA-15-80 °C
Zr-SBA-15-100 °C
Zr-SBA-15-120 °C
Zr-SBA-15-150 °C
>99
>99
>99
>99
17.4
10.8
9.2
30.9
35.9
34.1
34.3
1.0
1.6
3.0
8.7
0.1
0.2
1.0
1.2
0.6
0.3
0.2
0.2
9.2
7.9
7.3
6.4
6.9
2
Reaction conditions: 240 °C, 1 h, 400 psi N pressure, 0.2 g feedstock loading, and 0.1 g catalyst. ML: methyl lactate, GADMA: glycolaldehyde dimethylacetal, MG: methyl
glycolate, MLE: methyl levulinate.
without any noticeable pore size shrinkage or blockage, as shown
in Fig. S4. Furthermore, ML was used as reactant to test its stability.
It was found that only 3% of ML was decomposed after a one-hour
reaction at 240 °C with the Zr-SBA-15 catalyst in methanol
xylose to ML over the Zr-SBA-15 catalyst begins at the retro-aldol
condensation forming glycolaldehyde and glyceraldehyde, as
shown in Scheme 1. Fig. S6 illustrates the total acidity of the Zr-
SBA-15 samples calculated from the NH
icate did not show any appreciable ammonia adsorption (only
0.02 mmol NH /g), while the Zr-SBA-15 samples presented high
/g. The DRIFT-IR characterization was
3
TPD. The pure SBA-15 sil-
(Fig. S5), thus indicating that ML, the final product of xylose con-
version, was very stable in this reaction system.
3
It is well accepted that Lewis acids can promote retro-aldol con-
densation as the initial step in the conversion of sugars to lactic
acid [43]. We thus propose that the reaction pathway of converting
acidity of ꢂ0.7 mmol NH
3
performed to distinguish the Lewis acid and Brønsted acid sites
on the Zr-SBA-20 sample. As shown in Fig. S7, FTIR spectra of

L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
213
reversibly converted to each other. Then the aldol-condensation
reactions occurred in the presence of Zr-SBA-15 to form C4 and
C6 aldehydes. And subsequently ML was produced via retro-
aldol condensation reactions and became the major product.
However, it is unclear how the etherification and esterification
reactions with methanol evolved in parallel to the aldol conden-
sation and retro-aldol condensation reactions. Since most
glycolaldehyde intermediates could be further converted to ML
with the Zr-SBA-15 catalyst in near critical methanol, the yields
of ML from pentoses (xylose, arabinose) and hexoses (glucose,
mannose) had no distinct difference, unlike the results of convert-
ing sugars to ML at 160–170 °C using the Sn-Beta catalyst
reported by Taarning and co-workers [44].
Furfural was the undesirable by-product from the dehydration
of xylose, usually catalyzed by Brønsted acid. The furfural yield
decreased sharply at high temperatures or for a long reaction time,
while methyl levulinate yield increased at the same time during
the xylose conversion with the Zr-SBA-15 catalyst. Using furfural
as the probe reactant, the main product was methyl levulinate in
the yields of 3.6% and 33.2% after reactions for 1 h and 10 h, respec-
tively, at 240 °C, as shown in Fig. S10. Alkyl levulinate is known to
be produced from furfuryl alcohol with solid acid catalyst [75]. A
hydrogenation step is needed in the conversion of furfural to
methyl levulinate, while methanol was the only hydrogen source
in the probe reaction. We thus contemplate that furfural was con-
verted into furfuryl alcohol via Meerwein–Ponndorf–Verley (MPV)
transfer hydrogenation with methanol as the hydrogen donor,
which was probably promoted by the Lewis acid, followed by the
conversion of furfuryl alcohol into methyl levulinate through
ring-opening reactions with the aid of the weak Brønsted acid sites
on the Zr-SBA-15 catalyst. van Bekkum and co-workers demon-
strated that the zeolite Ti-Beta exhibited Lewis-acidic and moder-
ate weak Brønsted acidic properties and were selective and water
stable catalyst in the Meerwein–Ponndorf–Verley reduction
[76,77]. Corma’s group also demonstrated the intermolecular
transfer hydrogenation between alcohols and ketones in organic
solvents using zeolite Sn-Beta and Zr-Beta as solid Lewis acid cat-
alysts [78–80].
Fig. 8. Reuse of the Zr-SBA-15 catalyst in the conversion of xylose. The spent
catalysts were not regenerated. Reaction conditions: 240 °C, 1 h, 400 psi initial N
pressure, 0.2 g xylose, 0.1 g catalyst. ML: methyl lactate, GADMA: glycolaldehyde
2
dimethylacetal, MG: methyl glycolate, MLE: methyl levulinate.
adsorbed pyridine on Zr-SBA-15-20 revealed the presence of pre-
dominantly Lewis acid sites. Prominent adsorption band at
ꢀ
1
1
446 cm
is assigned to the presence of Lewis acid (L) sites,
ꢀ1
whereas the weak adsorption band at 1481 cm is attributed to
a combination of Brønsted and Lewis (B + L) acid sites. A very weak
band at 1541 cm typically corresponding to the Brønsted acid (B)
ꢀ1
4+
sites, was almost negligible. The Zr ions in the SBA-15 silica
framework as the Lewis acid sites first interacted with the carbonyl
group of xylose, and then broke the C5 xylose molecule down to C3
glyceraldehyde and C2 glycolaldehyde. Glyceraldehyde underwent
dehydration to form 2-hydroxypropenal, then to pyruvaldehyde
via keto-enol tautomerization, and finally to ML in methanol sol-
vent by possible intramolecular Cannizzaro and esterification reac-
tions. Although trioses were able to be converted to ML via
Meerwein–Ponndorf–Verley reduction with methanol [73], the
direct evidence of converting pyruvaldehyde to ML through
intramolecular Cannizzaro reaction was confirmed by isotopic
The proposed reaction pathway for the conversion of xylose to
ML can be generalized for the conversion of other cellulosic bio-
mass. The retro-aldol condensation of pentoses and hexoses, how-
ever, may form different aldehyde or ketone products with solid
Lewis acid catalysts [43,44]. The conversion of pentoses to ML
can be divided as the following steps: retro-aldol condensation of
an aldopentose leads to glyceraldehyde and glycolaldehyde, while
retro-aldol condensation of a ketopentose forms dihydroxyacetone
and glycolaldehyde. The outcome of both reaction pathways is the
formation of a triose and a glycolaldehyde [44]. In contrast, the
fragments formed from a hexose depend on the ketose or aldose
form. Disintegration of an aldohexose leads to the fragments of a
C4 aldotetrose and a C2 glycolaldehyde, while the ketohexose frag-
mentation results in two C3 fragments, dihydroxyacetone and
glyceraldehyde [43]. With a Lewis acid catalyst, however, glucose
and fructose can be inter-converted through isomerization before
experiencing subsequent reactions [46,81]. Table 1 shows that
the disparate yields of ML were produced from glucose and fruc-
tose, respectively, while the comparable yields are seen for fruc-
tose and sucrose, the dimer of glucose and fructose (products
were identified by GC/MS as shown in Fig. S11). Glucose yielded
a lower amount of ML relative to fructose (37.3% versus 44.1%).
The difference in the ML yields from different monosaccharides
reflects the dynamic equilibrium between the isomerization,
retro-aldol condensation and degradation reactions. In all cases,
the conversions of both monosaccharides and disaccharides were
greater than 99% after 6 h of reaction at 240 °C and the consider-
able yields of ML were formed using the Zr-SBA-15 catalyst
3
studies in CD OD solvent [74]. On the other hand, the by-
product, GADMA, was formed by the acetalization of glycolalde-
hyde with methanol.
To validate our hypothesis of the reaction pathway, the possi-
ble key intermediates, glyceraldehyde, dihydroxyacetone, pyru-
valdehyde, glycolaldehyde, and GADMA were used as the probe
reactants. It was found that much higher ML yields (78.8% and
8
4.5%) were obtained from glyceraldehyde and dihydroxyacetone,
respectively (Table 1 Entries 6 and 8), than from xylose. The yield
of ML was even close to 100% from pyruvaldehyde (Table 1 Entry
1
0). Yet in the absence of the Zr-SBA-15, much lower amounts of
ML were produced from all three probe reactants. More interest-
ingly, using glycolaldehyde as the probe, a high amount of
GADMA was produced without a catalyst while ꢂ61.3% ML was
obtained with the Zr-SBA-15 catalyst. It appeared that C–C bond
forming aldol condensation reactions occurred in the presence
of Zr-SBA-15 in that the C4 acid ester products, e.g., methyl
vinylglycolate and methyl 2-hydroxybutanoate, were also
observed as the products from the C2 reactant, glycolaldehyde
(Fig. S8). Very interestingly, when using GADMA as the probe
reactant, it was converted to ML in a 24% yield with the
Zr-SBA-15 catalyst and the product distribution was similar to
that using glycolaldehyde as the reactant. Methoxyacetaldehyde
dimethylacetal was another major product obtained through
etherification while other products in the much lower yields were
methyl vinylglycolate and methyl 2-hydroxybutanoate, as shown
in Fig. S9. It revealed that GADMA and glycoaldehyde were

2
14
L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
Scheme 1. Proposed reaction mechanism for the conversion of xylose to methyl lactate and other intermediate and final products in methanol solvent with the Zr-SBA-15
catalyst.
(Table 1). Using disaccharides or polysaccharide including cel-
4. Conclusion
lobiose, starch and cellulose as the reactants, substantially lower
conversions were observed (products were identified by GC/MS
as shown in Fig. S12). Cellobiose and starch, which are soluble in
methanol, reached the similar yields of ML, 24.3% and 24.1%,
respectively. However, only 16.7% ML was produced from cellulose
at 240 °C for a 10 h reaction, suggesting that depolymerization is
the bottleneck for cellulose conversion. In order to enhance the
yield of ML from cellulose, a small amount of water (5 wt%) was
added into methanol to facilitate the hydrolysis of cellulose. As
the result, as high as 28.1% yield of ML was obtained directly from
cellulose.
In summary, we demonstrate that methyl lactate was yielded
from various carbohydrates in methanol solutions at near critical
conditions with the Zr-SBA-15 catalysts. Under the optimum reac-
tion conditions, the highest methyl lactate yields were 42% and
44% from pentoses and hexoses, respectively. The Zr-SBA-15 cata-
lyst was relatively stable to produce methyl lactate from xylose
after five consecutive catalytic reaction cycles without regenera-
tion. The Lewis acid sites on the Zr-SBA-15 facilitated the retro-
aldol condensation of carbohydrates, which was the initial step
of the conversion of carbohydrates to methyl lactate. Increasing

L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
215
zirconium loading on the SBA-15 silica framework increased the
total acid strength of the Zr-SBA-15 catalyst. However, over-load
of zirconium could destroy the mesoporous structure of the SBA-
[11] R. Koster, B. Van Der Linden, E. Poels, A. Bliek, The mechanism of the gas-phase
esterification of acetic acid and ethanol over MCM-41, J. Catal. 204 (2001)
333–338, http://dx.doi.org/10.1006/jcat.2001.3356.
[
12] I. Agirrezabal-Telleria, J. Requies, M.B. Güemez, P.L. Arias, Dehydration of D-
1
5 silica. The large-pore Zr-SBA-15 catalysts inhibited the yields
xylose to furfural using selective and hydrothermally stable arenesulfonic
SBA-15 catalysts, Appl. Catal. B: Environ. 145 (2014) 34–42, http://dx.doi.org/
of undesirable by-products such as humins and furfural. Through
the probe reaction studies, we conclude that C3 aldehydes/ketones
including glyceraldehyde, dihydroxyacetone, and pyruvaldehyde
were the key intermediates to form methyl lactate from a variety
of carbohydrate biomass. Zr-SBA-15 also catalyzed aldol condensa-
tion of C2 aldehydes to form C4 and C6 sugars, followed by subse-
quent retro-aldol condensation, and finally to produce methyl
lactate in high yields. The conversions of polysaccharides such as
starch and cellulose are much more challenging than those of
monosaccharides and thus longer reaction time, as well as adding
water co-solvent to facilitate the hydrolysis, was beneficial. Over-
all, this ‘‘one-pot” process using the Zr-SBA-15 catalyst and near
critical methanol solvent is efficient and environmentally-
friendly to produce ML from cellulosic biomass feedstocks.
10.1016/j.apcatb.2012.11.010.
[
[
[
13] C.R. Marcilly, Where and how shape selectivity of molecular sieves operates in
refining and petrochemistry catalytic processes, Top. Catal. 13 (2000) 357–
366, http://dx.doi.org/10.1023/A:1009007021975.
14] S.-Y. Chen, J.-F. Lee, S. Cheng, Pinacol-type rearrangement catalyzed by Zr-
incorporated SBA-15, J. Catal. 270 (2010) 196–205, http://dx.doi.org/10.1016/j.
jcat.2009.12.020.
15] T.S. Jiang, X.P. Zhou, Y.H. Li, Q. Zhao, H.B. Yin, Characterization of Zr-based
MCM-41 mesoporous molecular sieves obtained by microwave irradiation or
hydrothermal method, Inorg. Mater. 47 (2011) 296–302, http://dx.doi.org/
10.1134/S0020168511030113.
16] J. Zhu, Q. Gao, Z. Chen, Preparation of mesoporous copper cerium bimetal
oxides with high performance for catalytic oxidation of carbon monoxide,
Appl. Catal. B: Environ. 81 (2008) 236–243, http://dx.doi.org/10.1016/j.
apcatb.2007.12.017.
[17] S. Lima, M.M. Antunes, A. Fernandes, M. Pillinger, M.F. Ribeiro, A.A. Valente,
Acid-catalysed conversion of saccharides into furanic aldehydes in the
presence of three-dimensional mesoporous Al-TUD-1, Molecules 15 (2010)
[
3863–3877, http://dx.doi.org/10.3390/molecules15063863.
[
18] M.S. Morey, S.O. Brien, S. Schwarz, G.D. Stucky, Hydrothermal and
postsynthesis surface modification of cubic, MCM-48, and ultralarge pore
SBA-15 mesoporous silica with titanium, Chem. Mater. 12 (2000) 898–911.
Acknowledgments
This material is based upon work supported by the National
Science Foundation under Grant No. CBET 1337017. Parts of this
work were carried out in the Characterization Facility, University
of Minnesota, which receives partial support from NSF through
the MRSEC program. The authors would like to thank Dr. Jacob
Warner of the University of Minnesota Characterization Facility
for the SAXS measurements, Dr. Steven Spain for the support of
the analytical instruments in the Shared Instrumentation Labora-
tory (SIL) in Chemistry department of UNR, and Dr. Mojtaba Ahma-
diantehrani in the Chemical and Materials Engineering Department
for the TEM test.
[19] Q. Pan, Intrinsic Kinetics of Lower Alcohols: C2, C3 Dehydration Over Lewis
Acidic Ordered Mesoporous Silicate: Zr-KIT-6, University of Kansas, 2013.
[
[
[
20] T.W. Kim, F. Kleitz, B. Paul, R. Ryoo, MCM-48-like large mesoporous silicas with
tailored pore structure: facile synthesis domain in ternary triblock
a
copolymer–butanol–water system, J. Am. Chem. Soc. 127 (2005) 7601–7610,
http://dx.doi.org/10.1021/ja042601m.
21] Q. Pan, A. Ramanathan, W.K. Snavely, R.V. Chaudhari, B. Subramaniam,
Synthesis and dehydration activity of novel Lewis Acidic ordered
mesoporous silicate: Zr-KIT-6, Ind. Eng. Chem. Res. 52 (2013) 15481–15487,
http://dx.doi.org/10.1021/ie4019484.
22] F. Bérubé, B. Nohair, F. Kleitz, S. Kaliaguine, Controlled postgrafting of titanium
chelates for improved synthesis of Ti-SBA-15 epoxidation catalysts, Chem.
Mater. 22 (2010) 1988–2000, http://dx.doi.org/10.1021/cm9030667.
[23] T. Tsoncheva, L. Ivanova, J. Rosenholm, M. Linden, Cobalt oxide species
supported on SBA-15, KIT-5 and KIT-6 mesoporous silicas for ethyl acetate
total oxidation, Appl. Catal. B: Environ. 89 (2009) 365–374, http://dx.doi.org/
10.1016/j.apcatb.2008.12.015.
Appendix A. Supplementary material
[
24] S.Y. Chen, L.Y. Jang, S. Cheng, Synthesis of Zr-incorporated SBA-15 mesoporous
materials in a self-generated acidic environment, Chem. Mater. 16 (2004)
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2015.10.013.
4174–4180, http://dx.doi.org/10.1021/cm049247b.
[25] A. Prabhu, L. Kumaresan, M. Palanichamy, V. Murugesan, Synthesis and
characterization of aluminium incorporated mesoporous KIT-6: efficient
catalyst for acylation of phenol, Appl. Catal. A: Gen. 360 (2009) 59–65,
http://dx.doi.org/10.1016/j.apcata.2009.03.004.
References
[
26] A. Vinu, N. Gokulakrishnan, V.V. Balasubramanian, S. Alam, M.P. Kapoor, K.
Ariga, et al., Three-dimensional ultralarge-pore ia3d mesoporous silica with
various pore diameters and their application in biomolecule immobilization,
Chem. A: Eur. J. 14 (2008) 11529–11538, http://dx.doi.org/10.1002/
chem.200801304.
[
1] D.J. Braden, C.A. Henao, J. Heltzel, C.C. Maravelias, J.A. Dumesic, Production of
liquid hydrocarbon fuels by catalytic conversion of biomass-derived levulinic
acid, Green Chem. 13 (2011) 1755–1765, http://dx.doi.org/10.1039/
c1gc15047b.
[
[
[
[
2] B. Girisuta, Levulinic Acid from Lignocellulosic Biomass, University of
Groningen, 2007. <http://dissertations.ub.rug.nl/FILES/faculties/science/2007/
b.girisuta/thesis.pdf> (accessed 21.11.13).
3] J.A. Geboers, S. Van de Vyver, R. Ooms, B. Op de Beeck, P.A. Jacobs, B.F. Sels,
Chemocatalytic conversion of cellulose: opportunities, advances and pitfalls,
Catal. Sci. Technol. 1 (2011) 714–726, http://dx.doi.org/10.1039/c1cy00093d.
4] H. Deng, L. Lin, S. Liu, Catalysis of Cu-doped Co-based perovskite-type oxide in
wet oxidation of lignin to produce aromatic aldehydes, Energy Fuels 24 (2010)
[27] L. Yang, J. Su, S. Carl, J.G. Lynam, X. Yang, H. Lin, Catalytic conversion of
hemicellulosic biomass to lactic acid in pH neutral aqueous phase media, Appl.
Catal. B: Environ. 162 (2015) 149–157, http://dx.doi.org/10.1016/j.
apcatb.2014.06.025.
[28] Y. Zhu, S. Jaenicke, G.K. Chuah, Supported zirconium propoxide – a versatile
heterogeneous catalyst for the Meerwein–Ponndorf–Verley reduction, J. Catal.
218 (2003) 396–404, http://dx.doi.org/10.1016/S0021-9517(03)00160-X.
[29] H. Lin, J. Strull, Y. Liu, Z. Karmiol, K. Plank, G. Miller, et al., High yield
production of levulinic acid by catalytic partial oxidation of cellulose in
aqueous media, Energy Environ. Sci. 5 (2012) 9773–9777, http://dx.doi.org/
10.1039/c2ee23225a.
4
797–4802, http://dx.doi.org/10.1021/ef100768e.
5] W. Deng, Y. Wang, Q. Zhang, Y. Wang, Development of bifunctional catalysts
for the conversions of cellulose or cellobiose into polyols and organic acids in
water, Catal. Surv. Asia 16 (2012) 91–105, http://dx.doi.org/10.1007/s10563-
[30] Y. Zhu, S. Liu, S. Jaenicke, G. Chuah, Zirconia catalysts in Meerwein–Ponndorf–
Verley reduction of citral, Catal. Today 97 (2004) 249–255, http://dx.doi.org/
10.1016/j.cattod.2004.07.002.
0
12-9136-1.
[
[
[
6] J. Hegner, K.C. Pereira, B. DeBoef, B.L. Lucht, Conversion of cellulose to glucose
and levulinic acid via solid-supported acid catalysis, Tetrahedron Lett. 51
[31] L. Fuxiang, Y. Feng, L. Yongli, L. Ruifeng, X. Kechang, Direct synthesis of Zr-SBA-
15 mesoporous molecular sieves with high zirconium loading: characterization
and catalytic performance after sulfated, Microporous Mesoporous Mater. 101
(2007) 250–255, http://dx.doi.org/10.1016/j.micromeso.2006.12.018.
[32] E. Rodr ´ı guez-Castellón, A. Jiménez-López, P. Maireles-Torres, D. Jones, J.
Rozière, M. Trombetta, et al., Textural and structural properties and surface
acidity characterization of mesoporous silica–zirconia molecular sieves, J.
Solid State Chem. 175 (2003) 159–169, http://dx.doi.org/10.1016/S0022-4596
(03)00218-4.
(
2010) 2356–2358, http://dx.doi.org/10.1016/j.tetlet.2010.02.148.
7] J. Jow, G.L. Rorrer, M.C. Hawley, D.T.A. Lamport, Dehydration of
D-fructose to
levulinic acid over LZY zeolite catalyst, Biomass 14 (1987) 185–194, http://dx.
doi.org/10.1016/0144-4565(87)90046-1.
8] R. Mellaerts, C.A. Aerts, J. Van Humbeeck, P. Augustijns, G. Van den Mooter, J.A.
Martens, Enhanced release of itraconazole from ordered mesoporous SBA-15
silica materials, Chem. Commun. (2007) 1375–1377, http://dx.doi.org/
1
0.1039/b616746b.
[
9] A. Katiyar, S. Yadav, P.G. Smirniotis, N.G. Pinto, Synthesis of ordered large pore
SBA-15 spherical particles for adsorption of biomolecules, J. Chromatogr. A
[33] A. Ramanathan, M.C. Castro Villalobos, C. Kwakernaak, S. Telalovic, U.
Hanefeld, Zr-TUD-1:
a
Lewis acidic, three-dimensional, mesoporous,
1
122 (2006) 13–20, http://dx.doi.org/10.1016/j.chroma.2006.04.055.
zirconium-containing catalyst, Chem. – Eur. J. 14 (2008) 961–972, http://dx.
doi.org/10.1002/chem.200700725.
[
10] P. Valle-vigón, M. Sevilla, A.B. Fuertes, Sulfonated mesoporous silica–carbon
composites and their use as solid acid catalysts, Appl. Surf. Sci. 261 (2012)
[34] A. Ramanathan, D. Klomp, J.a. Peters, U. Hanefeld, Zr-TUD-1:
a novel
5
74–583, http://dx.doi.org/10.1016/j.apsusc.2012.08.059.
heterogeneous catalyst for the Meerwein–Ponndorf–Verley reaction, J. Mol.

2
16
L. Yang et al. / Journal of Catalysis 333 (2016) 207–216
Catal. A: Chem. 260 (2006) 62–69, http://dx.doi.org/10.1016/j.molcata.2006.
6.057.
35] H. Kishida, F. Jin, X. Yan, T. Moriya, H. Enomoto, Formation of lactic acid from
glycolaldehyde by alkaline hydrothermal reaction, Carbohydr. Res. 341 (2006)
[59] X. Huang, T.I. Korányi, M.D. Boot, E.J.M. Hensen, Catalytic depolymerization of
0
lignin in supercritical ethanol, ChemSusChem 7 (2014) 2276–2288, http://dx.
doi.org/10.1002/cssc.201402094.
[60] S. Mukai, T. Koyama, K. Tsujii, S. Deguchi, Anomalous long-range repulsion
between silica surfaces induced by density inhomogeneities in supercritical
ethanol, Soft Matter 10 (2014) 6645–6650, http://dx.doi.org/10.1039/
c4sm00787e.
[
[
[
2
619–2623, http://dx.doi.org/10.1016/j.carres.2006.06.013.
36] R.H.W. Maas, R.R. Bakker, G. Eggink, R.A. Weusthuis, Lactic acid production
from xylose by the fungus Rhizopus oryzae, Appl. Microbiol. Biotechnol. 72
(
2006) 861–868, http://dx.doi.org/10.1007/s00253-006-0379-5.
[61] Q. Zhang, Y. Xu, Y. Li, T. Wang, Q. Zhang, L. Ma, et al., Investigation on the
esterification by using supercritical ethanol for bio-oil upgrading, Appl. Energy
(2015) in press. doi: http://dx.doi.org/10.1016/j.apenergy.2014.12.063.
[62] S. Brand, R.F. Susanti, S.K. Kim, H.S. Lee, J. Kim, B.I. Sang, Supercritical ethanol
as an enhanced medium for lignocellulosic biomass liquefaction: influence of
physical process parameters, Energy 59 (2013) 173–182, http://dx.doi.org/
10.1016/j.energy.2013.06.049.
[63] N.P. Vasilakos, D.M. Austgen, Hydrogen-donor solvents in biomass
liquefaction, Ind. Eng. Chem. Process Des. Dev. 24 (1985) 304–311.
[64] T. Nakagawa, H. Ozaki, T. Kamitanaka, H. Takagi, T. Matsuda, T. Kitamura, et al.,
Reactions of supercritical alcohols with unsaturated hydrocarbons, J. Supercrit.
Fluids 27 (2003) 255–261, http://dx.doi.org/10.1016/S0896-8446(02)00269-3.
[65] Z. Yuan, S. Cheng, M. Leitch, C. (Charles) Xu, Hydrolytic degradation of alkaline
lignin in hot-compressed water and ethanol, Bioresour. Technol. 101 (2010)
9308–9313, http://dx.doi.org/10.1016/j.biortech.2010.06.140.
[66] T. Ebukuro, A. Takami, Y. Oshima, S. Koda, Raman spectroscopic studies on
hydrogen bonding in methanol and methanol/water mixtures under high
temperature and pressure, J. Supercrit. Fluids 15 (1999) 73–78, http://dx.doi.
org/10.1016/S0896-8446(98)00126-0.
[67] A.R. Bazaev, I.M. Abdulagatov, E.A. Bazaev, A.A. Abdurashidova, A.E.
Ramazanova, PVT measurements for pure methanol in the near-critical and
supercritical regions, J. Supercrit. Fluids 41 (2007) 217–226, http://dx.doi.org/
10.1007/s10765-007-0164-4.
37] L. Wang, B. Zhao, B. Liu, B. Yu, C. Ma, F. Su, et al., Efficient production of
L
-lactic
acid from corncob molasses, a waste by-product in xylitol production, by a
newly isolated xylose utilizing Bacillus sp. strain, Bioresour. Technol. 101
(
2010) 7908–7915, http://dx.doi.org/10.1016/j.biortech.2010.05.031.
[
[
[
38] X. Yan, F. Jin, K. Tohji, A. Kishita, H. Enomoto, Hydrothermal conversion of
carbohydrate biomass to lactic acid, AIChE J. 56 (2010) 2727–2733, http://dx.
doi.org/10.1002/aic.
39] Y. Fan, C. Zhou, X. Zhu, Selective catalysis of lactic acid to produce commodity
chemicals, Catal. Rev. 51 (2009) 293–324, http://dx.doi.org/10.1080/01614940
9
03048513.
40] M.E. Himmel, S.-Y. Ding, D.K. Johnson, W.S. Adney, M.R. Nimlos, J.W. Brady,
et al., Biomass recalcitrance: engineering plants and enzymes for biofuels
production, Science 315 (2007) 804–807, http://dx.doi.org/10.1126/science.
1
137016.
41] A. Onda, T. Ochi, K. Kajiyoshi, K. Yanagisawa, A new chemical process for
catalytic conversion of -glucose into lactic acid and gluconic acid, Appl. Catal.
[
[
[
[
[
[
D
A: Gen. 343 (2008) 49–54, http://dx.doi.org/10.1016/j.apcata.2008.03.017.
42] L. Kong, G. Li, H. Wang, W. He, F. Ling, Hydrothermal catalytic conversion of
biomass for lactic acid production, J. Chem. Technol. Biotechnol. 388 (2008)
3
83–388, http://dx.doi.org/10.1002/jctb.
43] M.S. Holm, S. Saravanamurugan, E. Taarning, Conversion of sugars to lactic acid
derivatives using heterogeneous zeotype catalysts, Science 328 (2010) 602–
6
05, http://dx.doi.org/10.1126/science.1183990.
[68] D.M. Do, S. Jaenicke, G.K. Chuah, Mesoporous Zr-SBA-15 as a green solid acid
catalyst for the Prins reaction, Catal. Sci. Technol. 2 (2012) 1417–1424, http://
dx.doi.org/10.1039/c2cy20084h.
[69] Y. Konishi, M. Okazaki, K. Toriyama, T. Kasai, Nanotube effect on a liquid-phase
photoreaction in mesoporous silica, J. Phys. Chem. B 105 (2001) 9101–9106,
http://dx.doi.org/10.1021/jp010899m.
[70] D. Wei, W. Te Chueh, G.L. Haller, Catalytic behavior of vanadium substituted
mesoporous molecular sieves, Catal. Today 51 (1999) 501–511, http://dx.doi.
org/10.1016/S0920-5861(99)00036-X.
44] M.S. Holm, Y.J. Pagán-Torres, S. Saravanamurugan, A. Riisager, J.a. Dumesic, E.
Taarning, Sn-Beta catalysed conversion of hemicellulosic sugars, Green Chem.
1
4 (2012) 702–706, http://dx.doi.org/10.1039/c2gc16202d.
45] S. Tolborg, I. Sádaba, C.M. Osmundsen, P. Fristrup, M.S. Holm, E. Taarning, Tin-
containing silicates: alkali salts improve methyl lactate yield from sugars,
ChemSusChem 8 (2015) 613–617, http://dx.doi.org/10.1002/cssc.201403057.
46] B. Murillo, A. Sánchez, V. Sebastián, C. Casado-Coterillo, O. de la Iglesia, M.P.
López-Ram- de-Viu, et al., Conversion of glucose to lactic acid derivatives with
mesoporous Sn-MCM-41 and microporous titanosilicates, J. Chem. Technol.
Biotechnol. 89 (2014) 1344–1350, http://dx.doi.org/10.1002/jctb.4210.
[71] X. Hu, M.L. Foo, G.K. Chuah, S. Jaenicke, Pore size engineering on MCM-41:
3
selectivity tuning of heterogenized AlCl for the synthesis of linear alkyl
[
47] F. de Clippel, M. Dusselier, R. Van Rompaey, P. Vanelderen, J. Dijkmans, E.
Makshina, et al., Fast and selective sugar conversion to alkyl lactate and lactic
acid with bifunctional carbon–silica catalysts, J. Am. Chem. Soc. 134 (2012)
benzenes, J. Catal. 195 (2000) 412–415, http://dx.doi.org/10.1006/jcat.
2000.2965.
[72] C. Pirez, J. Caderon, J. Dacquin, A.F. Lee, K. Wilson, Tunable KIT-6 mesoporous
sulfonic acid catalysts for fatty acid esterification, ACS Catal. 2 (2012) 1607–
1614.
[73] J. Wang, Y. Masui, M. Onaka, Conversion of triose sugars with alcohols to alkyl
lactates catalyzed by Brønsted acid tin ion-exchanged montmorillonite, Appl.
Catal. B: Environ. 107 (2011) 135–139, http://dx.doi.org/10.1016/j.apcatb.
2011.07.006.
1
0089–10101, http://dx.doi.org/10.1021/ja301678w.
[
[
[
[
[
[
48] Q. Guo, F. Fan, E.A. Pidko, W.N.P. van der Graaff, Z. Feng, Li Can, et al., Highly
active and recyclable Sn-MWW zeolite catalyst for sugar conversion to methyl
lactate and lactic acid, ChemSusChem 6 (2013) 1352–1356.
49] P.Y. Dapsens, C. Mondelli, J. Pérez-Ramírez, Highly selective Lewis acid sites in
desilicated MFI zeolites for dihydroxyacetone isomerization to lactic acid,
ChemSusChem 6 (2013) 831–839, http://dx.doi.org/10.1002/cssc.201200703.
50] A. Onda, T. Ochi, K. Kajiyoshi, K. Yanagisawa, Lactic acid production from glucose
over activated hydrotalcites as solid base catalysts in water, Catal. Commun. 9
[74] P.P. Pescarmona, K.P.F. Janssen, C. Delaet, C. Stroobants, K. Houthoofd, A.
Philippaerts, et al., Zeolite-catalysed conversion of
3
C sugars to alkyl
lactates, Green Chem. 12 (2010) 1083–1089, http://dx.doi.org/10.1039/
b921284a.
(
2008) 1050–1053, http://dx.doi.org/10.1016/j.catcom.2007.10.005.
51] Z. Liu, W. Li, C. Pan, P. Chen, H. Lou, X. Zheng, Conversion of biomass-derived
carbohydrates to methyl lactate using solid base catalysts, Catal. Commun. 15
[75] J.P. Lange, W.D. van de Graaf, R.J. Haan, Conversion of furfuryl alcohol into
ethyl levulinate using solid acid catalysts, ChemSusChem 2 (2009) 437–441,
http://dx.doi.org/10.1002/cssc.200800216.
[76] J.C. Van Der Waal, E.J. Creyghton, P.J. Kunkeler, K. Tan, H. Van Bekkum, Beta-
type zeolites as selective and regenerable catalysts in the Meerwein–
Ponndorf–Verley reduction of carbonyl compounds, Top. Catal. 4 (1997)
261–268.
[77] J.C. van der Waal, K. Tan, H. van Bekkum, Zeolite titanium beta: a selective and
water resistant catalyst in Meerwein–Ponndorf–Verley–Oppenauer reactions,
Catal. Lett. 41 (1996) 63–67, http://dx.doi.org/10.1007/BF00811714.
[78] M. Boronat, A. Corma, M. Renz, Mechanism of the Meerwein–Ponndorf–
Verley–Oppenauer (MPVO) redox equilibrium on Sn- and Zr-beta zeolite
(
2011) 82–87, http://dx.doi.org/10.1016/j.catcom.2011.08.019.
52] S. Brand, F. Hardi, J. Kim, D.J. Suh, Effect of heating rate on biomass
liquefaction: differences between subcritical water and supercritical ethanol,
Energy 68 (2014) 420–427, http://dx.doi.org/10.1016/j.energy.2014.02.086.
53] P. Lalanne, J.M. Andanson, J.-C. Soetens, T. Tassaing, Y. Danten, M. Besnard,
Hydrogen bonding in supercritical ethanol assessed by infrared and Raman
spectroscopies, J. Phys. Chem. A 108 (2004) 3902–3909, http://dx.doi.org/
1
0.1021/jp0309466.
[
54] J. Peng, P. Chen, H. Lou, X. Zheng, Upgrading of bio-oil over aluminum silicate
in supercritical ethanol, Energy Fuels 22 (2008) 3489–3492.
[
55] B. Jin, P. Duan, Y. Xu, B. Wang, F. Wang, L. Zhang, Lewis acid-catalyzed in situ
transesterification/esterification of microalgae in supercritical ethanol,
Bioresour. Technol. 162 (2014) 341–349, http://dx.doi.org/10.1016/j.
biortech.2014.03.157.
56] S. Brand, J. Kim, Liquefaction of major lignocellulosic biomass constituents in
supercritical ethanol, Energy 80 (2015) 64–74, http://dx.doi.org/10.1016/j.
energy.2014.11.043.
57] W. Chen, Z. Luo, C. Yu, G. Li, Y. Yang, H. Zhang, Upgrading of bio-oil in
supercritical ethanol: catalysts screening, solvent recovery and catalyst
stability study, J. Supercrit. Fluids 95 (2014) 387–393, http://dx.doi.org/
catalysts, J. Phys. Chem.
10.1021/jp063249x.
[79] A. Corma, M.E. Domine, L. Nemeth, S. Valencia, Al-free Sn-Beta zeolite as a
catalyst for the selective reduction of carbonyl compounds (Meerwein–
Ponndorf–Verley reaction), J. Am. Chem. Soc. 124 (2002) 3194–3195, http://
dx.doi.org/10.1021/ja012297m.
[80] A. Corma, M.E. Domine, S. Valencia, Water-resistant solid Lewis acid catalysts:
Meerwein–Ponndorf–Verley and Oppenauer reactions catalyzed by tin-beta
zeolite, J. Catal. 215 (2003) 294–304, http://dx.doi.org/10.1016/S0021-9517
(03)00014-9.
B 110 (2006) 21168–21174, http://dx.doi.org/
[
[
1
0.1016/j.supflu.2014.09.041.
[81] G. Li, E.A. Pidko, E.J.M. Hensen, Synergy between Lewis acid sites and hydroxyl
groups for the isomerization of glucose to fructose over Sn-containing zeolites:
a theoretical perspective, Catal. Sci. Technol. 4 (2014) 2241–2250, http://dx.
doi.org/10.1039/c4cy00186a.
[
58] C. Tang, H. Tao, X. Zhan, X. Xie, Mechanism of esters formation during
cellulose liquefaction in sub- and supercritical ethanol, BioResources 9 (2014)
4
946–4957.