Morphology Tailoring of Sulfonic Acid Functionalized Organosilica Nanohybrids for the Synthesis of Biomass-Derived Alkyl Levulinates

Article abstract of DOI:10.1002/chem.201501219

Morphology evolution of sulfonic acid functionalized organosilica nanohybrids (Si(Et)Si-Pr/ArSO₃H) with a 1D tubular structure (inner diameter of ca. 5nm), a 2D hexagonal mesostructure (pore diameter of ca. 5nm), and a 3D hollow spherical structure (shell thickness of 2-3nm and inner diameter of ca. 15nm) was successfully realized through P123lated sol-gelcocondensationstrategies and fine-tuning of the acidity followed by aging or a hydrothermal treatment. The Si(Et)Si-Pr/ArSO₃H nanohybrids were applied in synthesis of alkyl levulinates from the esterification of levulinic acid and ethanolysis of furfural alcohol. Hollow spherical Si(Et)Si-Pr/ArSO₃H and hexagonal mesoporous analogues exhibited the highest and lowest catalytic activity, respectively, among three types of nanohybrids; additionally, the activity was influenced by the ￡SO₃H loading. The activity differences are explained in terms of different Bronsted acid and textural properties, reactant/product diffusion, and mass transfer rate, as well as accessibility of ￡SO₃H sites to the reactant molecules. The reusability of the nanohybrids was also evaluated.

Full text of DOI:10.1002/chem.201501219

DOI: 10.1002/chem.201501219
Full Paper
&
Heterogeneous Catalysis
Morphology Tailoring of Sulfonic Acid Functionalized Organosilica
Nanohybrids for the Synthesis of Biomass-Derived Alkyl
Levulinates
[b]
[b]
[b]
[a]
[a]
Sai An, Daiyu Song, Bo Lu, Xia Yang, and Yi-Hang Guo*
Abstract: Morphology evolution of sulfonic acid functional-
tion of levulinic acid and ethanolysis of furfural alcohol.
ized organosilica nanohybrids (Si(Et)Si-Pr/ArSO H) with a 1D
Hollow spherical Si(Et)Si-Pr/ArSO H and hexagonal mesopo-
3
3
tubular structure (inner diameter of ca. 5 nm), a 2D hexago-
nal mesostructure (pore diameter of ca. 5 nm), and a 3D
hollow spherical structure (shell thickness of 2–3 nm and
inner diameter of ca. 15 nm) was successfully realized
through P123-templated sol–gel cocondensation strategies
and fine-tuning of the acidity followed by aging or a hydro-
rous analogues exhibited the highest and lowest catalytic
activity, respectively, among three types of nanohybrids; ad-
ditionally, the activity was influenced by the ÀSO H loading.
3
The activity differences are explained in terms of different
Brønsted acid and textural properties, reactant/product dif-
fusion, and mass transfer rate, as well as accessibility of À
thermal treatment. The Si(Et)Si-Pr/ArSO H nanohybrids were
SO H sites to the reactant molecules. The reusability of the
3
3
applied in synthesis of alkyl levulinates from the esterifica-
nanohybrids was also evaluated.
Introduction
important to reduce the dependence on non-renewable fossil
reserves, such as esterification, transesterification, hydrolysis,
dehydration, and polycondensation, which lead to biofuels and
Silica chemistry is one of the most active research subjects be-
cause of its extensive applications. Silica-based materials with
various compositions, structures, and morphologies can be
prepared by using abundant silicate species in the presence of
ionic or nonionic surfactants, and diverse chemical interactions
result in the assembly, dissolution, and reassembly of silicate/
[
11–14]
value-added chemicals.
The morphological design of catalysts is an important
aspect for improving their catalytic activity and selectivity. To
a great extent, the catalytic activity of catalysts is determined
by their morphological characteristics and textural properties,
in addition to the structure and composition, which can influ-
ence the number of active sites, reactant/product diffusion to/
from active sites, and accessibility of the active sites to reac-
tants. Most of the reported sulfonic acid functionalized organo-
silica materials possess a periodic mesostructure, leading to ex-
cellent textural properties, including large surface area and
high porosity. Unfortunately, they suffer correspondingly from
slow in-pore diffusion because of their isolated and long 2D
pore channels, which limit their catalytic activity. The design of
sulfonic acid functionalized organosilicas with new morpholo-
gies is therefore an effective strategy for the further improve-
ment of their catalytic performance.
[
1–6]
surfactant species on the molecular scale.
Compared with
inorganic silica-based materials with a pure siliceous frame-
work, organosilica-based materials with organic/inorganic
(
O SiÀRÀSiO ) units completely and homogeneously distrib-
1.5
1.5
uted over the whole framework at the molecular level have
found extensive applications in materials science, separation
[
7]
science, pharmaceuticals, and catalysis. For example, due to
advantages such as strong Brønsted acid properties, a plentiful
number of acid sites, tunable textural and morphological prop-
erties, and high hydrothermal stability, sulfonic acid functional-
ized organosilica materials represent attractive alternatives to
hazardous liquid acids, and they have been widely applied in
[
6,8–10]
diversified acid-catalyzed reactions.
More recently, they
In the search for novel sulfonic acid functionalized organosil-
ica materials, herein, morphology tailoring of sulfonic acid
functionalized organosilica nanohybrids has been carried out
by one-step sol–gel cocondensation of bridged silsesquioxane
and 3-mercaptopropyltrimethoxysilane or 2-(4-chlorosulfonyl-
phenyl)ethyl trimethoxysilane by using nonionic surfactant
P123 (EO PO EO ) as a template. By carefully tailoring the
have also been successfully employed in environmentally
friendly biomass transformation processes that are especially
[
a] Dr. X. Yang, Prof. Y.-H. Guo
School of Environment, Northeast Normal University
20
70
20
2
555 Jingyue Street, Changchun 130117 (P.R. China)
preparation conditions, morphology evolution of Si(Et)Si-Pr/
Fax: (+86)431-89165626
E-mail: guoyh@nenu.edu.cn
ArSO H nanohybrids with a 1D tubular nanostructure (Si(Et)Si-
3
PrSO H-NT), a 2D hexagonal periodic mesostructure (Si(Et)Si-
3
[
b] S. An, D. Song, B. Lu
PrSO H-2Dhex), and 3D hollow spherical nanostructure (Si(Et)Si-
School of Chemistry, Northeast Normal University
3
5
268 Renmin Street, Changchun 130024 (P.R. China)
Pr/ArSO H-HNS) were successfully realized.
3
Chem. Eur. J. 2015, 21, 10786 – 10798
10786
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
To reveal the relationship between the heterogeneous acid
catalytic activity and morphology, as well as evaluate the reus-
ability of the prepared three types of sulfonic acid functional-
ized organosilica nanohybrids, the synthesis of alkyl levulinates
from the esterification of levulinic acid (Scheme 1a) and etha-
Results and Discussion
Preparation and characterization of Si(Et)Si-Pr/ArSO H nano-
hybrids
3
Sulfonic acid functionalized organosilica nanohybrids were pre-
pared by a P123 template directed sol–gel route. The process
included cocondensation of both organic building blocks,
namely, the bis(silylated) organic unit 1,2-bis(trimethoxysilyl)-
ethane (BTMSE; main organic building block to construct the
silica/carbon framework of the nanohybrids) with 3-mercapto-
propyltrimethoxysilane (MPTMS; secondary organic building
block to provide PrSO H sites after in situ oxidation with 30%
3
Scheme 1. Synthesis of alkyl levulinates from the esterification of levulinic
H O ) or 2-(4-chlorosulfonylphenyl)ethyl trimethoxysilane
2
2
acid and ethanolysis of furfural alcohol, as catalyzed by Si(Et)Si-Pr/ArSO
nanohybrids.
3
H
(
CSPTMS; secondary organic building block to provide ArSO H
3
sites directly) under acidic conditions. After formation of the
Pr/ArSO H-functionalized silica/carbon (i.e., ethane-bridged or-
3
nolysis of furfural alcohol (Scheme 1b) were selected as the
model reactions. Alkyl levulinates have found extensive appli-
cations in the flavoring and fragrance industry or as the pre-
cursors to produce chemicals (e.g., g-valerolactone) that can
ganosilica) framework, they nanohybrids underwent hydrother-
mal or aging treatments to further fasten the linkages of the
silica/carbon framework with organosulfonic acid groups, and
hence, cleavage of the SiÀC bond in subsequent catalytic pro-
[
15–17]
be converted into liquid alkanes and transportation fuels.
cesses was expected to be avoided. Finally, Si(Et)Si-Pr/ArSO H
3
The threat of a shortage of fossil fuel oil is stimulating the
search for more sustainable resources to fuels and chemicals.
Accordingly, the efficient conversion of further downstream
acid hydrolysis products of biomass-derived C or C carbohy-
nanohybrids were formed after the removal of P123 by wash-
ing with boiling ethanol. Compared with the complex hard
template method, which involves a multiple-step process, the
current one-step soft template strategy is simple and saves
time. Importantly, the direct preparation route can result in
5
[
6
18]
drates, that is, levulinic acid or furfural alcohol, to form alkyl
levulinates has attracted particular interest. Homogenous
acids, such as HF, H SO , and H PO , can catalyze the above
a uniform distribution and tunable loading of Pr/ArSO H sites;
3
moreover, the bonding strength of the Pr/ArSO H sites to the
2
4
3
4
3
processes effectively; nevertheless, the implementation of
these acids requires significant costs in terms of separation,
reuse, and treatment of the mixture after the reactions. The
development of solid acid catalysts is one of the key technolo-
gies to establish environmentally friendly catalytic processes
for the production of alkyl levulinates. Some solid acids, includ-
ing macroporous ion-exchange resins (e.g., Amberlyst-15 or
silica/carbon framework is high enough through covalently at-
tachment to prevent sulfur from leaching during the catalytic
process. However, compared with the hard template method,
it is more difficult to construct the nanohybrids with well-de-
fined morphologies. Herein, by fine-tuning of the preparation
conditions, morphology evolution of Si(Et)Si-Pr/ArSO H nano-
3
hybrids with hollow tubular and spherical structures, as well as
a 2D hexagonal periodic mesostructure, was realized.
[
19,20]
[21]
-
70),
mesoporous aluminosilicates (e.g., Al-TUD-1),
mi-
[22,23]
croporous zeolites (e.g., HZSM-5),
sulfated metal oxides
Amphiphilic copolymer surfactants (e.g., P123 or F127) con-
taining both hydrophilic (e.g., poly(ethylene oxide (PEO)) and
hydrophobic (e.g., poly(propylene oxide) (PPO)) blocks can
readily self-assemble into micelles through hydrogen bonding
and hydrophobic/-philic interactions under acidic conditions.
The micelles can further aggregate to form lyotropic liquid-
crystal structures with various morphologies, such as spherical,
tubular, and cylindrical shapes, through fine-tuning of the
preparation conditions, including micelle concentration, silica
precursor concentration, and acidity of the preparation system,
2
À
2À
[20]
(
(
e.g., SO4 /TiO and SO /ZrO2), sulfonic acid grafted silica
2
4
[
22,24]
e.g., SO H–SBA-15),
and mesoporous heteropolyacid
3
[25,26]
hybrid catalysts (e.g., H PW O /ZrO ÀSi(Ph)Si)
have been
3
12 40
2
applied in the above two reactions. However, moderate yields
of alkyl levulinates were obtained; additionally, catalyst deacti-
vation was observed during recycling of the solid acids. By
combining strong Brønsted acidity, well-defined morphology
(
especially for two hollow Si(Et)Si-Pr/ArSO H analogues), and
3
low water affinity, the sulfonic acid functionalized organosilica
nanohybrids are expected to exhibit excellent heterogeneous
acid catalytic activity and stability in the synthesis of alkyl levu-
linates from biomass-derived platform molecules. Special at-
tention is paid to the enhanced acid catalytic activity of two
[
27,28]
as well as hydrothermal or aging treatments.
Herein,
Si(Et)Si-Pr/ArSO H nanohybrids with a 1D tubular nanostruc-
3
ture, a 2D hexagonal periodic mesostructure, and a 3D hollow
spherical nanostructure were successfully fabricated by adjust-
ing the acidity of the preparation system and aging or hydro-
thermal treatments. As illustrated in Scheme 2 and Figure 1a–
e, for a total Si/P123/HCl molar ratio of 52:0.86:282, with 1,3,5-
trimethylbenzene (TMB) as the micelle-expanding agent, as
hollow Si(Et)Si-Pr/ArSO H nanohybrids with respect to their
3
mesoporous counterparts.
well as aging treatment, Si(Et)Si-Pr/ArSO H hollow nanospheres
3
were formed, in which the total silicon molar number was cal-
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10787
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 2. Illustration of the P123-templated, one-step preparation of propylsulfonic acid functionalized, ethane-bridged, organosilica hollow nanospheres
with a total Si/P123/HCl molar ratio of 52:0.86:282.
culated by 2nBTMSE +nMPTMS or 2nBTMSE +nCSPTMS. In the absence
of TMB, Si(Et)Si-PrSO H nanotubes were created under the
3
same conditions (Scheme 3 and Figure 1 f). For the same total
amount of Si and P123, but with a significantly decreased
amount of HCl, that is, a Si/P123/HCl molar ratio of 52:0.86:36,
the morphology of Si(Et)Si-PrSO H transformed into a 2D hex-
3
agonal periodic mesostructure after hydrothermal treatment
(
Scheme 3 and Figure 1g).
On the basis of the above results, it can be inferred that, at
a suitable concentration of P123 micelles and silica precursors
BTMSE and MPTMS/CSPTMS), the acidity of the preparation
(
system played a key role in the construction of Si(Et)Si-Pr/
ArSO H nanohybrids with different morphologies. At higher
3
acidity (e.g., Si/P123/HCl molar ratio=52:0.86:282), tubular or
spherical P123 lyotropic liquid-crystal phases were formed by
self-assembly. In the liquid-crystal structures, hydrophobic PPO
blocks formed the core of the micelles, whereas the hydrophil-
ic PEO blocks formed a hydrated corona around the core
(
Scheme 2). On the other hand, higher acidity can facilitate co-
hydrolysis and -condensation of BTMSE with MPTMS/CSPTMS
at a faster rate; this is favorable to form individual tubular or
spherical nanoparticles instead of consolidated porous struc-
tures during the self-assembly of hydrolyzed Ar/PrSO H-modi-
3
fied organosilica species around the corona of P123 micelles
through supramolecular interactions (e.g., hydrogen bonding,
electrostatic, and/or van der Waals forces). In the presence of
TMB, its hydrophobic molecules can dissolve in the core of
tubular P123 micelles as a swelling agent to expand micelles
to obtain spherical micelles, and the charge density on the sur-
face of micelles further decreases with the penetration of TMB
into the micelles, which finally leads to well-dispersed hollow
nanospheres instead of nanotubes. In the absence of TMB, tub-
ular Si(Et)Si-PrSO H was formed at a molar ratio of Si/P123/HCl
3
of 52:0.86:282. After aging treatment and subsequent extrac-
tion with boiling ethanol, tubular or spherical HSO CH CH CH -
Figure 1. TEM images of various Si(Et)Si-Pr/ArSO
ArSO H-HNS3.1, b) Si(Et)Si-ArSO H-HNS7.6, c) Si(Et)Si-ArSO
H-HNS6.8, e) Si(Et)Si-HNS, f) Si(Et)Si-PrSO H-NT8.6, and
H-2Dhex6.7. Scale bars 50 nm, except in b) in which it is
3
H nanohybrids: a) Si(Et)Si-
3
2
2
2
3
3
3
H-HNS10.4,
Si(OSi) -CH CH -Si(OSi) frameworks formed and constructed
3
2
2
3
d) Si(Et)Si-PrSO
g) Si(Et)Si-PrSO
3
3
3
the wall or shell structure of the nanohybrids (Schemes 2 and
3).
1
00 nm.
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10788
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 3. Illustration of the P123-templated, one-step preparation of propylsulfonic acid functionalized, ethane-bridged, organosilica nanohybrids with a 1D
tubular structure and a 2D hexagonal periodic mesostructure.
At lower acidity, that is, at a Si/P123/HCl molar ratio of
2:0.86:36, P123 micelles can aggregate linearly to form a cylin-
Si(Et)Si-PrSO H nanohybrids exhibit 1D hollow tubular mor-
3
5
phology with several tubes twisting together, and the estimat-
ed inner diameter of the tubes is about 5 nm (Figure 1 f). By
changing the total Si/P123/HCl molar ratio to 52:0.86:36 fol-
drical lyotropic liquid-crystal structure; meanwhile, lower acidi-
ty can slow down the cohydrolysis and -condensation rates of
BTMSE with MPTMS. Both factors lead to the self-assembly of
lowed by hydrothermal treatment, the Si(Et)Si-PrSO H nanohy-
3
hydrolyzed PrSO H-modified organosilica species with P123 mi-
brids exhibit a 2D hexagonal periodic mesostructure with pore
diameters of about 5 nm (Figure 1g).
3
celles to form a periodic mesostructure instead of individual
hollow nanoparticles. After hydrothermal treatment and subse-
quent extraction with boiling ethanol, a 2D hexagonal meso-
Nitrogen gas porosimetry measurements were applied to
characterize the textural properties of various Si(Et)Si-Pr/
structured Si(Et)Si-PrSO H was formed with the same wall com-
ArSO H nanohybrids, and the corresponding nitrogen gas ad-
3
3
position as that of its tubular or spherical analogue
sorption–desorption isotherms (Figure 2a and c) and BJH pore
size distribution curves (Figure 2b and d) were obtained. The
results show that all tested nanohybrids exhibit type IV iso-
therms, which implies that they are mesoporous. Interestingly,
(
Scheme 3).
Pore morphologies and textural properties
the Si(Et)Si-Pr/ArSO H nanohybrids with different morphologies
3
High-resolution TEM was applied to reveal the morphology
possess different hysteresis loop shapes, which reflect the in-
fluence of the morphologies on the porosity properties. For
transformation of the Si(Et)Si-Pr/ArSO H nanohybrids with the
3
preparation conditions (Figure 1). From the images in Fig-
the mesoporous Si(Et)Si-PrSO H-2Dhex6.7 nanohybrid, it has an
3
ure 1a–e, it can be observed that the Si(Et)Si-Pr/ArSO H nano-
H1-type hysteresis loop and one capillary condensation step at
3
hybrids, prepared from
a
Si/P123/HCl molar ratio of
a medium relative pressure (P/P ) of 0.45–0.70 (Figure 2a). The
0
5
2:0.86:282 with TMB as the mi-
celle-expanding agent followed
by aging treatment, are com-
posed of uniform spheres with
hollow interiors, regardless of
the structure of the organic
building blocks and loading of
organosulfonic acid groups.
Table 1. Textural parameters and acid capacity of various Si(Et)Si-Pr/ArSO3H nanohybrids.
Catalysts
S
BET
2
D
[nm]
p
V
p
D
[nm]
I
S
A
À1 [a]
[b]
3
À1 [c]
[d]
[d]
+
À1 [e]
[m g
]
[cm g
]
[nm]
[meq(H )g ]
Si(Et)Si-HNS
Si(Et)Si-ArSO
Si(Et)Si-ArSO H-HNS7.6
830
539
529
468
667
715
709
14.9
12.6
12.6
11.8
14.6
1.90
1.27
1.02
0.71
1.50
1.17
0.70
16
15
15
14
16
5
20
21
19
18
21
–
–
387
944
1290
836
1059
823
3
H-HNS3.1
3
Moreover, the Si(Et)Si-Pr/ArSO H
Si(Et)Si-ArSO
3
H-HNS10.4
3
Si(Et)Si-PrSO
Si(Et)Si-PrSO
Si(Et)Si-PrSO
3
H-HNS6.8
H-NT8.6
H-2Dhex6.7
nanospheres possess extremely
small particle sizes (18–21 nm;
Table 1) and thin shells (2–3 nm;
Table 1), and the estimated inner
diameter of the nanospheres is
about 15 nm. At the same total
Si/P123/HCl molar ratio, but in
the absence of TMB, the formed
3
3
4.0/23.0
4.8
–
–
[
(
a] Surface area (SBET) was calculated by using the Brunauer–Emmett–Teller (BET) equation. [b] Pore diameter
) was estimated from the Barrett–Joyner–Halenda (BJH) adsorption determination. [c] Pore volume (V ) was
isotherm curve at P/
) and particle size (S) of the hollow nanospheres were estimat-
D
p
p
estimated from the pore volume determination by using the adsorption branch of the N
=0.99 single point. [d] The inner diameter (D
2
P
0
I
ed from TEM images. [e] The acid density (A) was determined by the inductively coupled plasma optical emis-
sion spectroscopy (ICP-OES) method.
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10789
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
a lower ArSO H loading (i.e.,
3
3
.1%), they possess the same
isotherm shape as that of
Si(Et)Si-ArSO H-HNS7.6. Upon in-
3
creasing the ArSO H loading to
3
1
0.4%, only one capillary con-
densation at P/P =0.45–0.85
0
was found. This result may
be because the void space
of Si(Et)Si-ArSO H-HNS10.4 is
3
blocked severely by the anchor-
ing ArSO H group, which leads
3
to a loss of the secondary pores.
The aforementioned interest-
ing porosity properties are re-
flected in the corresponding BJH
pore size distribution curves; ad-
ditionally, the results obtained
are in agreement with the TEM
observations
displayed
in
Figure 1. From the results shown
in Figure 2b, it is found that
Si(Et)Si-PrSO H-2Dhex6.7 exhibits
3
a sharp pore size distribution
curve centered at 4.8 nm, which
further confirms the well-distrib-
uted mesopores of the nanohy-
Figure 2. Nitrogen gas adsorption–desorption isotherms (a, c) and BJH pore size distribution profiles (b, d) of vari-
ous Si(Et)Si-Pr/ArSO H nanohybrids.
3
brids. For tubular Si(Et)Si-PrSO H-
3
result signifies that the nanohybrid has periodic pores without
interconnecting channels. In the cases of two types of hollow
NT8.6, its BJH pore size distribution curve exhibits double
peaks, which implies a bimodal porous structure. The smaller
(primary) pore centered at 4.0 nm is well distributed and fea-
tures a narrower curve, which corresponds to uniform nano-
tube channels of the nanohybrids. The larger (secondary) pore
centered at 23.0 nm is unevenly distributed, featuring wider
curve. This type of pore originates from the void space be-
tween the loosely packed tubular particles. For two nano-
sphere analogues, Si(Et)Si-PrSO H-HNS6.8 and Si(Et)Si-ArSO H-
nanohybrids, tubular Si(Et)Si-PrSO H-NT8.6 and spherical
3
Si(Et)Si-PrSO H-HNS6.8 or Si(Et)Si-ArSO H-HNS7.6, they both ex-
3
3
hibit two capillary condensation steps (Figure 2a). For tubular
Si(Et)Si-PrSO H-NT8.6, its capillary condensation occurs at P/
3
P0 =0.4–0.80 and 0.8–0.99, respectively. Compared with tubular
Si(Et)Si-PrSO H-NT8.6, spherical Si(Et)Si-PrSO H-HNS6.8 or
3
3
Si(Et)Si-ArSO H-HNS7.6 possesses much larger and smaller hys-
3
3
3
teresis loops at lower (0.45–0.90) and higher P/P (0.90–0.99),
HNS7.6, only the smaller pore centered at 12.6 and 14.6 nm, re-
spectively, is found, which is attributed to the uniform hollow
interior of the nanospheres. The secondary pore, which origi-
nates from the void space between the aggregated spherical
nanoparticles, is not found under the current testing condi-
tions, probably due to the unsuitable testing range. Similar re-
0
respectively; moreover, the hollow nanospherical analogues
have a much higher nitrogen gas adsorbed capacity (Fig-
ure 2a). From a combination of the porosity and morphologi-
cal characteristics of hollow nanohybrids, it is speculated that
the formation of the first hysteresis loop (which appeared at
lower P/P ) is due to the hollow interior of the tubular or
sults are also obtained for Si(Et)Si-ArSO H-HNS3.1 and Si(Et)Si-
0
3
spherical nanoparticles, whereas the second hysteresis loop
ArSO H-HNS10.4 (Figure 2d).
3
(
which appeared at higher P/P ) originates from the void space
From the calculated textual parameters based on nitrogen
gas adsorption–desorption isotherms (Table 1), three types of
0
between the loosely packed tubular or spherical nanoparticles.
The increased gas adsorption capacity of spherical Si(Et)Si-
PrSO H-HNS6.8 or Si(Et)Si-ArSO H-HNS7.6 is due to a sufficient
Si(Et)Si-Pr/ArSO H nanohybrids all have a large BET surface area
3
2
À1
3
À1
(468–715 m g ) and high pore volume (0.70–1.50 cm g );
3
3
hollow interior space and much more unplugged mesostruc-
ture.
meanwhile, the ÀSO H loading and morphology of the nano-
3
hybrids influence the BET surface area and pore volume. For in-
stance, the BET surface area and pore volume of Si(Et)Si-
To further study the influence of ArSO H loading on the tex-
3
tural properties of the hollow spherical nanohybrids, Si(Et)Si-
ArSO H-HNS nanohybrids gradually decrease with increasing À
3
HNS, Si(Et)Si-ArSO H-HNS3.1, and Si(Et)Si-ArSO H-HNS10.4 were
SO H group loading from 3.1 and 7.6% to 10.4%. The ÀSO H-
3
3
3
3
also tested, and the obtained nitrogen gas adsorption–desorp-
tion isotherms are presented in Figure 2c. The results show
that, for ÀSO H-free Si(Et)Si-HNS and Si(Et)Si-ArSO H-HNS with
free Si(Et)Si-HNS nanohybrid possesses the largest BET surface
2
À1
3
À1
area (830 m g ) and highest pore volume (1.90 cm g ). This
result indicates that most of the ArSO H groups are confined
3
3
3
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10790
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
inside the interior of hollow nanospheres because Si(Et)Si-HNS
possesses a large fraction of voids in the interior to accommo-
date the guest components. However, the incorporation of
a large amount of ArSO H groups could occupy most of the
3
space of Si(Et)Si-HNS, which in turn leads to a decreased BET
surface area and pore volume. At a similar ArSO H loading,
3
three types of Si(Et)Si-PrSO H nanohybrids (e.g., Si(Et)Si-
3
PrSO H-2Dhex6.7, Si(Et)Si-PrSO H-NT8.6, and Si(Et)Si-PrSO H-
3
3
3
HNS6.8) with 2D hexagonal periodic mesostuctures, hollow
tubular nanostructures, and spherical nanostructures possess
different BET surface areas and pore volumes: Si(Et)Si-PrSO H-
3
2
À1
NT8.6 has the largest BET surface area (715 m g ), whereas
Si(Et)Si-PrSO H-HNS6.8 has the highest pore volume
3
3
À1
(
1.50 cm g ).
Structural information
The formation of the silica/carbon framework and successful
incorporation of ArSO H or PrSO H groups into the prepared
3
3
nanohybrids by the current one-step cocondensation prepara-
2
9
tion route were confirmed by Si MAS (Figures 3a and 4a) and
13
C CP-MAS NMR spectroscopy (Figures 3b and 4b), and
Si(Et)Si-ArSO H-HNS10.4 and Si(Et)Si-PrSO H-HNS6.8 were
3
3
chosen as representative samples.
2
9
13
Figure 4. a) Si MAS and b) C-CP MAS NMR spectra of the Si(Et)Si-PrSO
HNS6.8 nanohybrid.
3
H-
29
13
Figure 3a and b displays Si MAS and C CP-MAS NMR
spectra, respectively, of the Si(Et)Si-ArSO H-HNS10.4 nanohy-
3
29
brids. In the Si MAS NMR spectrum, the characteristic reso-
nance at d=À66.6 ppm is assigned to organosiloxane species
3
of T (ÀCH CH ÀSi(OSi) ) within the ethane-bridged organosilica
2
2
3
framework; this confirms that the ethane-bridged organosilica
framework (-Si-O -Si-CH -CH -Si-O -) has been successfully
1.5
2
2
1.5
fabricated after sol–gel cocondensation of BTMSE followed by
aging treatment. Additionally, the resonance signals corre-
3
sponding to inorganic SiO species, such as Q (Si(OSi) (OH);
4
3
4
d=À90 ppm) and Q (Si(OSi) ; d=À120 ppm), are not found,
4
which implies that cleavage of the SiÀC bond in the silica/
13
carbon framework of the sample has been avoided. In the
C
CP-MAS NMR spectrum, five signals at d=16.3, 28.3, 127.4,
40.1, and 148.0 ppm belong to five types of carbon species of
1
1
5
arenesulfonic acid groups of Si(Et)Si-ArSO H-HNS10.4 (C ÀC in
3
Figure 3b), in which the first two signals originate from the
1
2
methylene carbon atoms (C and C in Figure 3b) of the
ArSO H group, whereas the last three signals correspond to ar-
3
3
4
5
omatic carbon atoms (C , C , and C in Figure 3b) of the
[
29]
ArSO H group. The most intense signal at d=5.4 ppm is as-
3
6
signed to carbon species (C in Figure 3b) of the bridging
[
11]
29
13
ethyl group of the silica/carbon framework. Through a combi-
Figure 3. a) Si MAS and b) C CP-MAS NMR spectra of the Si(Et)Si-ArSO
HNS10.4 nanohybrid.
3
H-
29
13
nation of Si MAS and C CP-MAS NMR spectra, it was con-
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10791 ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
firmed that arenesulfonic acid groups were bonded on the
ethane-bridged organosilica framework through a -Si-O -Si-
1.5
CH -CH -Si-O -CH CH -C H -SO H covalent linkage during the
2
2
1.5
2
2
6
4
3
preparation process, and the formed organic–inorganic frame-
work constructed the shell structure of the Si(Et)Si-ArSO H-
3
HNS10.4 nanohybrid.
2
9
Similar results can be found in the Si MAS (Figure 4a) and
13
C CP-MAS (Figure 4b) NMR spectra of the Si(Et)Si-PrSO H-
3
2
9
HNS6.8 nanohybrids. In the Si MAS NMR spectrum, two char-
acteristic signals at d=À59.8 and À66.2 ppm are found; these
2
correspond to siloxane species of ÀCH CH Si(OSi) (OH) (T ) and
2
2
2
3
ÀCH CH Si(OSi) (T ) within the ethane-bridged organosilica
2
2
3
framework, respectively; additionally, the signals corresponding
1
3
to inorganic SiO species are not detected. In the C CP-MAS
4
1
2
3
NMR spectrum, signals at d=5.3 (C ), 53.8 (C ), 18.4 (C ), and
4
1
1.9 ppm (C ) are found; these are attributed to carbon atoms
from the bridging ethyl group in the silica/carbon framework
and the incorporated propylsulfonic acid groups, respectively.
Additionally, the signals (d=29.0 and 22.4 ppm) corresponding
[
30]
to carbon atoms on the -Si-(CH ) -SH units are not found,
2
3
which suggests that thiol groups from MPTMS are oxidized
completely during the preparation procedure. Signals (d=36.5
and 39.0 ppm) originating from carbon species (-CH -S-S-CH2
2
and -CH -S-S(O )-CH ) of incomplete oxidation of mercapto-
2
2
2
[30]
propyl groups are also not identified. The above results fur-
ther confirm that mercaptopropyl groups from MPTMS are
converted completely into propylsulfonic acid groups by in
2
9
13
situ H O oxidation. From a combination of Si MAS and
C
2
2
Figure 5. TGA plots of a) Si(Et)Si-ArSO
HNS6.8 nanohybrids.
3
3
H-HNS10.4 and b) Si(Et)Si-PrSO H-
CP-MAS NMR spectra of Si(Et)Si-PrSO H-HNS6.8, it can be con-
3
cluded that the -Si-O -Si-CH -CH -Si-O -CH CH CH -SO H
1
.5
2
2
1.5
2
2
2
3
framework is formed during the nanohybrid preparation pro-
cess.
Heterogeneous acid catalytic activity of the Si(Et)Si-Pr/
ArSO H nanohybrids
3
Esterification of levulinic acid
The heterogeneous acid catalytic activity of the Si(Et)Si-Pr/
ArSO H nanohybrids was first evaluated by esterification of lev-
3
Thermal properties
ulinic acid with methanol to yield methyl levulinate under the
conditions of a levulinic acid/methanol molar ratio of 1:7,
2 wt% catalyst, 658C, and atmospheric pressure.
The thermal stability of the Si(Et)Si-Pr/ArSO H-HNS nanohybrids
3
was studied by thermogravimetric analysis (TGA) in the range
of 25 to 6008C, and the TGA plots of representative samples
First, the influence of morphology on the heterogeneous
(
Si(Et)Si-ArSO H-HNS10.4 and Si(Et)Si-PrSO H-HNS 6.8) are pre-
acid catalytic activity of the Si(Et)Si-Pr/ArSO H nanohybrids was
3
3
3
sented in Figure 5. The TGA plots show three weight loss
steps: the first weight loss (ca. 5%), which occurred below
studied (Figure 6a). At similar PrSO H loadings, three types of
3
Si(Et)Si-PrSO H nanohybrids with different morphologies exhib-
3
1
008C, is related to the loss of physisorbed water at the sur-
ited different acid catalytic activities, and their activity followed
face of the nanohybrids; the second weight loss, which hap-
the order of Si(Et)Si-PrSO H-2D 6.7<Si(Et)Si-PrSO H-NT8.6<
3
hex
3
pened in the range of 200–4008C (6.4% for Si(Et)Si-ArSO H-
Si(Et)Si-PrSO H-HNS6.8. For instance, for the Si(Et)Si-PrSO H-
3 3
3
HNS10.4 and 7.5% for Si(Et)Si-PrSO H-HNS6.8), is mainly attrib-
2Dhex6.7-, Si(Et)Si-PrSO H-NT8.6-, and Si(Et)Si-PrSO H-HNS6.8-cat-
3 3
3
uted to the decomposition of residual P123; and the last
alyzed esterification reaction, the yields of methyl levulinate
reached 51.2, 72.1, and 89.4%, respectively, after the reaction
proceeded for 90 min. For arenesulfonic and propylsulfonic
acid functionalized ethyl-bridged organosilica hollow nano-
spheres, Si(Et)Si-ArSO H-HNS7.6 and Si(Et)Si-PrSO H-HNS6.8, the
weight loss, which was found in the range of 400–6008C
(
13.7% for Si(Et)Si-ArSO H-HNS10.4 and 17.0% for Si(Et)Si-
3
PrSO H-HNS6.8), is related to the decomposition of organic
3
components in the nanohybrids, including ArSO H or PrSO H
3
3
3
3
groups and the -Si-O -Si-CH -CH -Si-O - framework. There-
former was somewhat more active than the latter. Under the
same conditions, the yield of methyl levulinate reached 98.2%
1
.5
2
2
1.5
fore, the Si(Et)Si-Pr/ArSO H nanohybrids are thermally stable
3
below 4008C.
for the Si(Et)Si-ArSO H-HNS7.6-catalyzed esterification reaction.
3
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10792
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The esterification activity of the most active nanohybrid,
Si(Et)Si-ArSO H-HNS7.6, was further compared with those of
3
conventional homo- (e.g., HCl) and heterogeneous (e.g., H-
ZSM-5 and Amberlyst-15) Brønsted acid catalysts in terms of
TOF (Figure 6c) and the yield of methyl levulinate (Figure 6c,
inset). The TOF of Si(Et)Si-ArSO H-HNS7.6 outperformed the
3
abovementioned reference Brønsted acid catalysts. However,
when considering the yield of methyl levulinate, the four
tested acid catalysts followed the activity order HCl>Si(Et)Si-
ArSO H-HNS7.6>Amberlyst-15>H-ZSM-5, and after the reac-
3
tion proceeded for 90 min, the yield of methyl levulinate
reached 100.0, 98.2, 86.1, and 10.0%, respectively, for the HCl-,
Si(Et)Si-ArSO H-HNS7.6-, Amberlyst-15-, and H-ZSM-5-catalyzed
3
esterification reactions.
Ethanolysis of furfural alcohol
The synthesis of alkyl levulinates by direct esterification of lev-
ulinic acid with the corresponding alkyl alcohols suffers from
drawbacks such as the high cost of raw material (i.e., levulinic
acid) and the production of the byproduct water. An atom-
economic and convenient method for the synthesis of alkyl
levulinates is expected by ethanolysis of furfural alcohol in the
presence of acid catalysts. Herein, the Si(Et)Si-Pr/ArSO H nano-
3
hybrids with various morphologies were applied in the etha-
nolysis of furfural alcohol to synthesize ethyl levulinate at
1
208C and a furfural alcohol/ethanol molar ratio of 1:60
(
Figure 7). The results show that spherical Si(Et)Si-PrSO H-
3
HNS6.8 or Si(Et)Si-ArSO H-HNS7.6 exhibited higher ethanolysis
3
activity than the tubular (Si(Et)Si-PrSO H-NT8.6) or periodic
3
mesoporous (Si(Et)Si-PrSO H-2Dhex6.7) analogues; in addition,
3
Si(Et)Si-ArSO H-HNS7.6 is more active than Si(Et)Si-PrArSO H-
3
3
HNS6.8. For example, for the Si(Et)Si-PrSO H-2Dhex6.7-, Si(Et)Si-
3
PrSO H-NT8.6-, Si(Et)Si-PrSO H-HNS6.8-, and Si(Et)Si-ArSO H-
3
3
3
HNS7.6-catalyzed ethanolysis reactions, the yields of ethyl levu-
linate reached 53.0, 54.7, 56.8, and 59.4%, respectively, over
a period of 120 min. The above activity order is the same as
that of the esterification reaction.
The ethanolysis activity of Si(Et)Si-ArSO H-HNS7.6 was also
3
compared with the reference Brønsted acid catalysts HCl, H-
ZSM-5, and Amberlyst-15 in terms of TOF (Figure 7b) and the
yield of ethyl levulinate (Figure 7b, inset). The TOF of Si(Et)Si-
Figure 6. Influence of a) morphology and b) ArSO
catalytic activity of Si(Et)Si-Pr/ArSO H-HNS nanohybrids towards esterification
of levulinic acid with methanol to produce methyl levulinate. c) Esterification
activity comparison of Si(Et)Si-ArSO H-HNS7.6 with respect to H-ZSM-5, Am-
berlyst-15, or HCl in terms of turnover frequency (TOF) and the yield of
methyl levulinate (inset). Molar ratio of levulinic acid/methanol=1:7; 2 wt%
catalyst; 658C.
3
H group loading on the
3
3
ArSO H-HNS7.6 is somewhat higher than that of H-ZSM-5, but
3
much higher than those of HCl or Amberlyst-15. However,
when considering the yield of ethyl levulinate, the four tested
acid catalysts followed the activity order HCl=Amberlyst-15>
Si(Et)Si-ArSO H-HNS7.6>H-ZSM-5, and after the reaction pro-
3
Subsequently, the influence of ÀSO H loading on the esterifi-
ceeded for 120 min, the yield of ethyl levulinate reached 86.6,
87.4, 59.4, and 38.4%, respectively, for the HCl-, Amberlyst-15-,
3
cation activity was evaluated by selecting hollow spherical
Si(Et)Si-ArSO H-HNS nanohybrids as representative catalysts.
Si(Et)Si-ArSO H-HNS7.6-, and H-ZSM-5-catalyzed ethanolysis re-
3
3
From the results shown in Figure 6b, the three tested nanohy-
actions.
brids follow the activity order Si(Et)Si-ArSO H-HNS3.1<Si(Et)Si-
3
ArSO H-HNS10.4<Si(Et)Si-ArSO H-HNS7.6. For instance, after
3
3
Discussion
the reaction was carried out for 90 min, the yields of methyl
levulinate reached 37.7 (Si(Et)Si-ArSO H-HNS3.1), 98.2 (Si(Et)Si-
On the basis of the above catalytic testing results, the esterifi-
3
ArSO H-HNS7.6), and 88.8% (Si(Et)Si-ArSO H-HNS10.4), respec-
cation and ethanolysis activity of the Si(Et)Si-Pr/ArSO H nano-
3
3
3
tively.
hybrids is mainly dominated by their Brønsted acid properties,
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10793
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
this leads to a stronger Brønsted acid strength of Si(Et)Si-
ArSO H-HNS7.6, and thus, a stronger proton-releasing ability.
3
As a consequence, Si(Et)Si-ArSO H-HNS7.6 is somewhat more
3
catalytically active than Si(Et)Si-PrSO H-HNS6.8 in both esterifi-
3
cation and ethanolysis reactions.
On the other hand, the heterogeneous acid catalytic activity
of the Si(Et)Si-Pr/ArSO H nanohybrids is determined remarkably
3
by their morphological characteristics and textural properties
in addition to Brønsted acid properties. The three types of
nanohybrids prepared possess unique morphological charac-
teristics, including a 1D tubular nanostructure, a 2D hexagonal
periodic mesostructure, and a 3D hollow spherical nanostruc-
ture, which can improve the accessibility of Pr/ArSO H sites to
3
levulinic acid or furfural alcohol molecules and facilitate mass
transfer of levulinic acid/methyl levulinate or furfural alcohol/
ethyl levulinate molecules. Additionally, unique morphological
characteristics lead to the excellent textural properties of the
nanohybrids, including a large BET surface area and high pore
volume, which can provide better dispersion of the Pr/ArSO H
3
sites throughout the nanohybrid catalysts and increase the
population of Pr/ArSO H sites. Both factors can significantly in-
3
crease the esterification and ethanolysis reaction rates. In com-
parison with periodic mesoporous Si(Et)Si-PrSO H-2Dhex6.7,
3
with a longer diffusion path owing to its longer pore channels,
the PrSO H site confined hollow interiors of tubular Si(Et)Si-
3
PrSO H-NT8.6 or spherical Si(Et)Si-PrSO H-HNS6.8 can serve as
3
3
a nanoreactor for the synthesis of alkyl levulinates, and the
shortened diffusion path ensures efficient reactant/product dif-
fusion and acid site accessibility during the reaction. Addition-
ally, extremely thin walls or shells (several nanometers) of the
hollow nanohybrids can significantly facilitate mass transfer of
the reactant/product through the successful reduction in the
mass transfer resistance from the outside surface of the nano-
spheres/-tubes to the interior. As a consequence, Si(Et)Si-
PrSO H-NT8.6 or Si(Et)Si-PrSO H-HNS6.8 show higher catalytic
Figure 7. a) Catalytic activity of Si(Et)Si-Pr/ArSO
nolysis of furfural alcohol to produce ethyl levulinate. b) Comparison of the
ethanolysis activity Si(Et)Si-ArSO H-HNS7.6 with respect to H-ZSM-5, Amber-
lyst-15, or HCl in terms of TOF and the yield of ethyl levulinate (inset). Molar
ratio of furfural alcohol/ethanol=1:60; 1.5 wt% catalyst; 1208C.
3
H nanohybrids towards etha-
3
including acid site density and acid strength. For the Si(Et)Si-
Pr/ArSO H-catalyzed esterification reaction of levulinic acid, it
3
3
3
started from protonation of the carbonyl groups of levulinic
activity to both of the target reactions than Si(Et)Si-PrSO H-
3
acid molecules by ÀSO H groups, whereas protonation of the
2Dhex6.7, although the three counterparts have comparable
Brønsted acid densities and BET surface areas. In addition, the
higher pore volume and larger pore diameter of Si(Et)Si-
PrSO H-NT8.6 or Si(Et)Si-PrSO H-HNS6.8 than those of Si(Et)Si-
3
hydroxyl groups of furfural alcohol molecules by ÀSO H
3
groups was the first step of the Si(Et)Si-Pr/ArSO H-catalyzed
3
[31]
ethanolysis reaction of furfural alcohol. Strong Brønsted acid
3
3
strengths of the Si(Et)Si-Pr/ArSO H nanohybrids can ensure
PrSO H-2Dhex6.7 also positively influence the activity. For two
3
3
that both of the reactions proceed at a considerably fast rate.
Additionally, the Brønsted acid site density significantly influen-
ces the catalytic activity. For example, hollow spherical Si(Et)Si-
ArSO H-HNS3.1 and Si(Et)Si-ArSO H-HNS7.6 possess similar tex-
hollow nanostructured analogues, spherical Si(Et)Si-PrSO H-
3
HNS6.8 and tubular Si(Et)Si-PrSO H-NT8.6, the higher acid cata-
3
lytic activity of the former is due to a shorter diffusion path
than that in the latter, as well as its three-dimensional nano-
structure, which can accelerate reactant/product diffusion to/
3
3
tural properties (Table 1); however, the Brønsted acid site den-
+
À1
sity of the former (387 meq(H )g ) is clearly lower than that of
from PrSO H sites and improve the acid site accessibility. Final-
3
+
À1
the latter (944 meq(H )g ), which finally leads to a lower cata-
ly, for hollow spherical Si(Et)Si-ArSO H-HNS10.4, although it has
3
+
À1
lytic activity of the former. For Si(Et)Si-ArSO H-HNS7.6 and
the highest Brønsted acid site density (1290 meq(H )g ) of
3
Si(Et)Si-PrSO H-HNS6.8, they both possess the same hollow
the three tested hollow spherical Si(Et)Si-ArSO H-HNS nanohy-
3
3
spherical morphology, as well as similar textural properties and
acid site densities; their activity difference is mainly deter-
brids, it exhibits lower catalytic activity than that of Si(Et)Si-
ArSO H-HNS7.6. The decreased catalytic activity is due to inferi-
3
mined by the microenvironment of the ÀSO H sites in the
or porosity properties, including the smaller BET surface area
3
[
32,33]
nanocatalysts.
For the Si(Et)Si-ArSO H-HNS7.6 nanohybrid,
and pore volume of Si(Et)Si-ArSO H-HNS10.4. This result further
3
3
phenyl in the arenesulfonic acid group can provide stronger
electron-withdrawing environments than propyl in the propyl-
confirms the influence of the porosity properties on the cata-
lytic activity.
sulfonic acid group of the Si(Et)Si-PrSO H-HNS6.8 nanohybrid;
3
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10794
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Regeneration and reusability
56.0%, respectively, for the first, second, and third cycles.
These results confirm that the catalyst also shows excellent
catalytic stability under the current solvothermal conditions.
Additionally, to evaluate the catalytic stability of the Si(Et)Si-
Evaluation of the reusability of hybrid acid catalysts is an im-
portant issue, especially under hydrothermal or solvothermal
conditions because of leaching of the anchored acid sites
under harsh conditions. Additionally, the adsorption of carbo-
naceous byproducts on the catalyst surface may severely dete-
Pr/ArSO H nanohybrids in the presence of water in the reac-
3
tion media, the representative Si(Et)Si-ArSO H-HNS7.6 nanohy-
3
brid was further tested in the hydrolysis of esters, including
ethyl acetate, ethyl benzoate, and butyl benzoate (Table 2). For
riorate the catalytic activity. Herein, by taking Si(Et)Si-ArSO H-
3
HNS10.4 and Si(Et)Si-ArSO H-HNS7.6 as the representative cata-
3
lysts, the catalytic stability of the sulfonic acid functionalized
organosilica nanohybrids was evaluated by reusing them in
the esterification and ethanolysis reactions three times. After
each catalytic cycle, the catalysts were recovered by centrifuga-
tion, and then they were washed with dichloromethane three
times. After being dried at 1008C overnight, the catalysts were
weighed and applied to the second and third cycles.
Table 2. Conversion of the ester in hydrolysis reactions catalyzed by the
Si(Et)Si-ArSO H-HNS7.6 nanohybrid.
3
[a]
[a]
Substrate
1st cycle
2nd cycle
53.5
3rd cycle
60.3
52.1
The catalytic stability of Si(Et)Si-ArSO H-HNS10.4 in the
3
esterification of levulinic acid in methanol at reflux (658C) for
8
7
5.8
90.0
68.2
83.4
66.2
2
h is presented in Figure 8a, and the yields of methyl levuli-
nate were 88.1, 97.3, and 99.9%, respectively, for the first,
second, and third cycles. The results indicate that the catalyst
is stable, and it can maintain a similar level of activity for at
4.1
least three catalytic cycles. For the Si(Et)Si-ArSO H-HNS7.6-cata-
3
lyzed ethanolysis reaction of furfural alcohol carried out at
1
208C for 2 h, the yield of ethyl levulinate is 59.4, 58.7, and
[
a] Conversion (%)=(c
of the ester (determined by GC) initially and at time t, respectively.
b] The reaction was performed in a 5 wt% aqueous solution (10 mL, in
0
Àc
t 0 0 t
)/c , in which c and c were the concentrations
[
which 5.84 mmol of ethyl acetate and 520 mmol of water were included)
at 608C, 67 mg of catalyst, 2 h reaction time. [c] The reaction was per-
formed in a 2 wt% aqueous solution (10 mL, in which 1.29 mmol of ethyl
benzoate, 450 mmol of water, and 40 mmol of acetonitrile were included)
at 808C, 67 mg of catalyst, 2 h reaction time. [d] The reaction was per-
formed in a 2 wt% aqueous solution (10 mL, in which 1.08 mmol of butyl
benzoate, 450 mmol of water, and 40 mmol of acetonitrile were included)
at 808C, 67 mg of catalyst, 2 h reaction time.
the three model reactions, no clear activity loss was found
after three consecutive catalytic cycles (Table 2); meanwhile,
elution of the SO H groups into the reaction media was barely
3
detected by ICP-OES. Both results confirm the excellent catalyt-
ic stability of the Si(Et)Si-Pr/ArSO H nanohybrids in ester hy-
3
drolysis reactions, except in esterification or ethanolysis reac-
tions.
To support the above results, the composition, chemical
structure, and morphology of Si(Et)Si-ArSO H-HNS7.6 after the
3
13
third cycle were characterized by ICP-OES, C CP-MAS NMR
spectroscopy (Figure 9a), and TEM (Figure 9b). The ÀSO H
3
loading of Si(Et)Si-ArSO H-HNS7.6 after the third cycle is 7.4%,
3
which is close to its initial value (7.6%); leaching of ÀSO H
3
groups into the reaction medium is minimal. Additionally, all
characteristic signals concerning carbon species of arenesulfon-
1
5
ic acid groups (C ÀC in Figure 9a; d=16.0, 28.3, 127.7, 140.6,
and 147.5 ppm) and ethane-bridged organosilica units (-Si-
6
6
CH - CH -Si- in Figure 9a; d=5.2 ppm) are found in the
2
2
Figure 8. Reusability of a) Si(Et)Si-ArSO
levulinic acid with methanol to produce methyl levulinate (molar ratio of
levulinic acid/methanol=1:7; 2 wt% catalyst; 658C) and b) Si(Et)Si-ArSO H-
HNS7.6 towards ethanolysis of furfuryl alcohol to produce ethyl levulinate
molar ratio of furfural alcohol/ethanol=1:60; 1.5 wt% catalyst; 1208C).
3
H-HNS10.4 towards esterification of
Si(Et)Si-ArSO H-HNS7.6 catalyst after the third cycle. The hollow
3
nanospherical morphology of Si(Et)Si-ArSO H-HNS7.6 after the
3
3
third cycle remained intact (Figure 9b) without further aggre-
gation. These results suggest that the physicochemical proper-
(
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10795
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
areas and high pore volumes), except for strong Brønsted acid
properties; this resulted in excellent acid catalytic activity to-
wards the synthesis of alkyl levulinates from the esterification
of levulinic acid and ethanolysis of furfural alcohol. Moreover,
hollow tubular or spherical Si(Et)Si-PrSO H with extremely thin
3
walls or shells and accessible mesochannels showed higher
catalytic activity than the periodic mesoporous counterpart;
this was attributed to the shortened diffusion path and more
efficient mass transfer. As a consequence, improved accessibili-
ty of ÀSO H sites to the reactant molecules was obtained. Ad-
3
ditionally, the higher acid catalytic activity of spherical Si(Et)Si-
PrSO H than its tubular analogue was due to a much shorter
3
diffusion path and three-dimensional nanostructure, which fur-
ther accelerated reactant/product diffusion to/from PrSO H
3
sites and improved the acid site accessibility. Importantly,
Si(Et)Si-Pr/ArSO H nanohybrids exhibited excellent reusability
3
in both target reactions, which showed their potential applica-
tions in a variety of environmentally friendly biomass transfor-
mation processes.
Experimental Section
Materials
BTMSE (97%) and Pluronic P123 (EO PO EO , in which EO=
20
70
20
CH CH O and PO=CH (CH )CHO) were purchased from Sigma-Al-
2
2
2
3
drich. MPTMS (97%) was purchased from J&K CHEMICA (P.R.
China). CSPTMS (50% in dichloromethane) was purchased from
Gelest, Inc. TMB was purchased from Tianjin Guangfu Fine Chemi-
cal Research Institute (P.R. China). Furfural alcohol was purchased
from Sinopharm Chemical Reagent Co. Ltd. (P.R. China). All other
chemicals were of analytical grade, and they were purchased from
Beijing Fine Chemical Co. (P.R. China).
1
3
Figure 9. a) C CP-MAS NMR spectrum and b) TEM image of the Si(Et)Si-
ArSO H-HNS7.6 nanohybrid after the third cycle.
3
ties of the Si(Et)Si-ArSO H-HNS7.6 catalyst are retained after the
3
third cycle; this reveals excellent catalytic stability. The excel-
lent reusability of the organosulfonic acid functionalized orga-
nosilica nanohybrids is due to strong chemical interactions be-
tween organosulfonic acid groups and the silica/carbon frame-
work; meanwhile, the hydrophobic character of alkyl-contain-
ing organosilica can reduce acid site deactivation associated
with adsorption of polar byproducts. Therefore, these hybrid
materials work effectively as recyclable solid acid catalysts in
both esterification and ethanolysis reactions.
Catalyst preparation
Arenesulfonic or propylsulfonic acid functionalized ethane-
bridged organosilica hollow nanospheres Si(Et)Si-Pr/ArSO H-
HNS):
3
Si(Et)Si-ArSO H-HNS: Typically, P123 (0.5 g) was dissolved in a mix-
3
À1
ture of water (12.7 mL), HCl (12 molL , 2.4 mL), and TMB (1.3 mL)
under stirring at room temperature for 1 h. Subsequently, BTMSE
(
0.6 mL) and CSPTMS (0.11, 0.22 and 0.33 mL, respectively, for are-
nesulfonic or propylsulfonic acid to adjust the loading of ÀSO H
3
group in the products) were added successively to the above clear
acidic solution of P123/TMB at an interval of 45 min. The molar
composition of the starting materials in the synthetic gel was
P123/water/HCl/TMB/BTMSE/CSPTMS=0.86:7056:282:93:23.4:(2.6,
Conclusion
Sulfonic acid functionalized organosilica nanohybrids with dif-
ferent morphologies were facilely fabricated by adjusting the
preparation conditions of a sol–gel cocondensation process:
5
.2 or 7.8). The resulting suspension was stirred at 408C for 24 h
and then aged at 1008C for an additional 48 h. Subsequently, the
product was air-dried at 60, 80, and 1008C for 12, 12, and 2 h, re-
spectively. P123 in the product was removed by washing with boil-
ing ethanol three times; each time lasted 12 h. The surfactant-free
product was air-dried at 1008C overnight, and was denoted as
Si(Et)Si-ArSO H-HNSx, in which x represented the loading of ÀSO H
1
5
) at higher acidity, or with a Si/P123/HCl molar ratio of
2:0.86:282 and TMB as the micelle-expanding agent followed
by aging treatment, Si(Et)Si-Pr/ArSO H hollow nanospheres
3
formed; 2) in the absence of TMB in the above preparation
3
3
system, Si(Et)Si-PrSO H nanotubes were developed; and 3) at
3
groups in the product (herein, x=3.1, 7.6, and 10.4%, respectively,
lower acidity, or with a Si/P123/HCl molar ratio of 52:0.86:36
followed by hydrothermal treatment, 2D hexagonal periodic
as determined by ICP-OES).
mesostructured Si(Et)Si-PrSO H was created. The three types of
3
Si(Et)Si-PrSO H-HNS: The preparation process was the same as
3
nanohybrids possessed unique morphological characteristics
and excellent textural properties (including large BET surface
that described above, but CSPTMS was replaced with MPTMS
(0.1 mL or 0.52 mmol) to provide PrSO H groups after in situ H O
2
3
2
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10796
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
À1
À1
oxidation. The molar composition of the starting materials in the
spectively; TOF (h )=[M /(Am)]t , in which m (g) is the mass of
D
synthetic gel was P123/water/HCl/TMB/BTMSE/MPTMS/H O =
the catalyst used in the reaction, t (h) is the reaction time (herein,
t=1 h), and A is the acid density of the tested catalyst. The cata-
lysts were dried at 1208C in vacuum for 2 h before the tests.
2
2
0
.86:7056:282:93:23.4:5.2:46.6. The ÀSO H loading in Si(Et)Si-
3
PrSO H-HNS was determined to be 6.8%.
3
Si(Et)Si-HNS: The preparation process was the same as that de-
For the esterification reaction, it was performed with stirring in
a three-necked round-bottomed glass flask under reflux, with a lev-
ulinic acid (1.474 g)/methanol (3.7 mL) molar ratio of 1:7 and
scribed above in the absence of MPTMS and H O .
2
2
2
wt% (with respect to the amount of reactants) catalyst. At peri-
Propylsulfonic acid functionalized ethane-bridged organosilica
odical intervals, aliquots (0.1 mL) of the reaction mixture were
withdrawn, ethyl laurate was added as the internal standard, and
then it was diluted with acetone to 5 mL. The diluted suspension
was centrifuged to obtain a clear solution, which was analyzed by
means of a Shimadzu 2014C gas chromatograph (GC) to determine
the concentration of methyl levulinate. The GC was equipped with
a HP-INNOWAX capillary column (film thickness, 0.5 mm; inner di-
ameter (i.d.), 0.32 mm; length, 30 m) and a flame ionization detec-
tor. The operating temperature was 2208C, and the flow rate of ni-
nanotubes (Si(Et)Si-PrSO H-NT): The preparation route was the
3
same as that used for Si(Et)Si-PrSO H-HNS without the addition of
3
TMB. The ÀSO H loading in Si(Et)Si-PrSO H-NT was determined to
3
3
be 8.6%.
2
D hexagonal periodic mesoporous propylsulfonic acid func-
tionalized ethane-bridged organosilica (Si(Et)Si-PrSO H-2Dhex):
3
P123 (0.5 g) was dissolved in a mixture of water (12.7 mL) and HCl
À1
(
12 molL , 0.3 mL) under stirring at room temperature for 1 h.
À1
trogen gas was 1.0 mLmin .
Subsequently, BTMSE (0.6 mL) was added dropwise to the above
solution at 408C. After stirring the mixture for 45 min, MPTMS
For the ethanolysis reaction, it was performed at 1208C in a Teflon-
lined autoclave reactor with vigorous stirring and in an oil batch to
hold isothermicity. The molar ratio of furfural alcohol (0.1 mL) to
ethanol (4 mL) was 1:60, and 1.5 wt% (with respect to the amount
of reactants) catalyst was used. After the reaction, all suspension in
the autoclave was decanted into a 10 mL centrifuge tube, and the
clear reaction liquid was separated by centrifugation. After adding
ethyl laurate (internal standard) to the above reaction liquid, the
resulting mixture was diluted with acetone to 10 mL. The concen-
tration of ethyl levulinate was analyzed by means of a Shimadzu
(
0.1 mL) and H O (0.5 mL) were added successively. The molar
2 2
composition of the starting materials in the synthetic gel was
P123/water/HCl/BTMSE/MPTMS/H O =0.86:7056:36:23.4:5.2:46.6.
2
2
The resultant white suspension was stirred at 408C for 24 h, and
then it was transferred to an autoclave, followed by heating at
À1
1
008C with a heating rate of 28Cmin for an additional 24 h. The
white product obtained was recovered by filtration and then air-
dried at 608C overnight. P123 in the product was removed by
washing with boiling ethanol for 12 h (repeated three times). The
final product was obtained by air drying at 1008C overnight. The À
2
014C GC under the conditions described below. The injection
SO H loading determined for the product was 6.7%.
3
port temperature was 2508C, the oven temperature was main-
tained at 608C for 5 min and then raised to 1208C for 30 min at
À1
Catalyst characterization
a heating rate of 88Cmin , and the flow rate of nitrogen gas was
À1
1
.0 mLmin .
Nitrogen gas porosimetry measurements were performed on a Mi-
cromeritics ASAP 2020M surface area and porosity analyzer after
the samples were outgassed under vacuum at 363 K for 1 h and
Acknowledgements
3
73 K for another 12 h. TEM observations were performed on
a JEM-2100F high-resolution transmission electron microscope at
This work was supported by the Natural Science Fund Council
of China (21173036, 51278092, and 51478095) and New Centu-
ry Excellent Talents in University of Ministry of Education of
China (NCET-13-0723).
1
3
29
an accelerating voltage of 200 kV. C CP-MAS and Si MAS NMR
spectra were recorded on a Bruker AVANCE III 400 WB spectrome-
ter equipped with a 4 mm standard bore CP MAS probe head. The
dried and finely powdered samples were packed in the ZrO rotor
2
closed with a Ke-F cap, which were spun at a rate of 12 kHz. Chem-
1
3
29
ical shifts for C CP-MAS and Si MAS NMR spectra were refer-
Keywords: biomass
·
heterogeneous
catalysis
·
·
enced to the signals of the C H (dCH =38.5) and 3-(trimethyl-
silyl)-1-propanesulfonic acid sodium salt (d=0.0) standards, respec-
1
0
16
2
nanostructures
silicates
· organic–inorganic hybrid composites
tively. The sulfur content or loading of ÀSO H in the prepared
3
nanohybrids were determined by using an ICAP6300-Thermoscien-
[
[
1] R. Ciriminna, A. Fidalgo, V. Pandarus, F. Beland, L. M. Ilharco, M. Pagliaro,
Chem. Rev. 2013, 113, 6592.
2] C. Urata, H. Yamada, R. Wakabayashi, Y. Aoyama, S. Hirosawa, S. Arai, S.
Takeoka, Y. Yamauchi, K. Kuroda, J. Am. Chem. Soc. 2011, 133, 8102–
tific ICP-OES instrument, and therefore, the Brønsted acid site den-
+
À1
sity (A, meq(H )g ) of the product was calculated. Before determi-
À1
nation, the samples were dissolved in a 2 molL solution of NaOH
and then diluted with deionized water to suitable volumes.
8105.
[
[
3] T. Yokoi, K. Ogawa, D. Lu, J. N. Kondo, Y. Kubota, T. Tatsumi, Chem.
Mater. 2011, 23, 2014–2016.
4] Y. Chen, P. Xu, H. Chen, Y. Li, W. Bu, Z. Shu, Y. Li, J. Zhang, L. Zhang, L.
Pan, X. Cui, Z. Hua, J. Wang, L. Zhang, J. Shi, Adv. Mater. 2013, 25,
Catalytic tests
Esterification of levulinic acid with methanol to produce methyl
levulinate and ethanolysis of biomass-derived furfural alcohol to
produce ethyl levulinate were selected as the model reactions. The
catalytic activity of the catalysts was evaluated quantitatively from
3
100–3105.
5] J. H. Kim, B. Fang, M. Y. Song, J. S. Yu, Chem. Mater. 2012, 24, 2256–
264.
6] X. H. Zhang, F. Su, D. Y. Song, S. An, B. Lu, Y. H. Guo, Appl. Catal. B 2015,
63, 50–62.
[
[
2
À1
the yields of methyl or ethyl levulinate (Y, %) and TOF (h ). Herein,
1
Y (%)=(M /M )”100, in which M and M are the number of moles
of methyl levulinate or ethyl levulinate produced and expected, re-
[7] F. Hoffmann, M. Cornelius, J. Morell, M. Froba, Angew. Chem. Int. Ed.
2006, 45, 3216–3251; Angew. Chem. 2006, 118, 3290–3328.
D
T
D
T
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org 10797
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
8] Q. H. Yang, J. Liu, J. Yang, M. P. Kapoor, S. Inagaki, C. Li, J. Catal. 2004,
28, 265–272.
9] M. Rat, M. H. Zahedi-Niaki, S. Kaliaguine, T. O. Do, Microporous Mesopo-
rous Mater. 2008, 112, 26–31.
[22] P. Demma Carà, R. Ciriminna, N. R. Shiju, G. Rothenberg, M. Pagliaro,
ChemSusChem 2014, 7, 835–840.
[23] K. Y. Nandiwale, P. S. Niphadkar, S. S. Deshpande, V. V. Bokade, J. Chem.
Technol. Biotechnol. 2014, 89, 1507–1515.
2
[
[
10] J. A. Melero, R. Grieken, G. Morales, Chem. Rev. 2006, 106, 3790–3812.
11] M. H. Tucker, A. J. Crisci, B. N. Wigington, N. Phadke, R. Alamillo, J. P.
Zhang, S. L. Scott, J. A. Dumesic, ACS Catal. 2012, 2, 1865–1876.
[24] J. A. Melero, G. Morales, J. Iglesias, M. Paniagua, B. Hernµndez, S.
Penedo, Appl. Catal. A 2013, 466, 116–122.
[25] F. Su, L. Ma, D. Y. Song, X. H. Zhang, Y. H. Guo, Green Chem. 2013, 15,
885–890.
[26] F. Su, Q. Y. Wu, D. Y. Song, X. H. Zhang, M. Wang, Y. H. Guo, J. Mater.
Chem. A 2013, 1, 13209–13221.
[27] H. Lee, K. Char, ACS Appl. Mater. Interfaces 2009, 1, 913–920.
[28] M. Mandal, M. Kruk, Chem. Mater. 2012, 24, 123–132.
[29] J. A. Melero, G. D. Stucky, R. van Grieken, G. Morales, J. Mater. Chem.
2002, 12, 1664–1670.
[30] X. G. Wang, S. Cheng, J. C. C. Chan, J. Phys. Chem. C 2007, 111, 2156–
2164.
[31] G. F. Wang, Z. Q. Zhang, L. H. Song, Green Chem. 2014, 16, 1436–1443.
[32] D. Zuo, J. Lane, D. Culy, M. Schultz, A. Pullar, M. Waxman, Appl. Catal. B
2013, 129, 342–350.
[
[
[
[
[
[
[
12] J. Z. Zhang, X. Liu, M. Sun, X. H. Ma, Y. Han, ACS Catal. 2012, 2, 1698–
1
702.
13] D. M. Alonso, J. Q. Bond, J. A. Dumesic, Green Chem. 2010, 12, 1493–
513.
14] J. C. Serrano-Ruiz, R. Luque, A. Sepulveda-Escribanoa, Chem. Soc. Rev.
011, 40, 5266–5281.
15] R. Le Van Mao, Q. Zhao, G. Dima, D. Petraccone, Catal. Lett. 2011, 141,
71–276.
16] Z. H. Zhang, K. Dong, Z. B. (Kent) Zhao, ChemSusChem 2011, 4, 112–
18.
17] A. DØmolis, N. Essayem, F. Rataboul, ACS Sustainable Chem. Eng. 2014,
, 1338–1352.
18] L. Hu, G. Zhao, W. Hao, X. Tang, Y. Sun, L. Lin, S. Liu, RSC Adv. 2012, 2,
1184–11206.
1
2
2
1
2
[33] K. Nakajima, I. Tomita, M. Hara, S. Hayashi, K. Domen, J. N. Kondo, Adv.
Mater. 2005, 17, 1839–1842.
1
[
[
19] J. P. Lange, W. D. Graaf, R. J. Haan, ChemSusChem 2009, 2, 437–441.
20] D. R. Fernandes, A. S. Rocha, E. F. Mai, C. J. A. Mota, V. T. Silva, Appl.
Catal. A 2012, 425–426, 199–204.
[
21] P. Neves, M. M. Antunes, P. A. Russo, J. P. Abrantes, S. Lima, A. Fer-
nandes, M. Pillinger, S. M. Rocha, M. F. Ribeiro, A. A. Valente, Green
Chem. 2013, 15, 3367–3376.
Received: March 28, 2015
Published online on June 18, 2015
Chem. Eur. J. 2015, 21, 10786 – 10798
www.chemeurj.org
10798
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim