New insights into the encapsulation and stabilization of heteropolyacids inside the pore walls of mesostructured silica materials

Article abstract of DOI:10.1039/b816172k

A method of synthesizing stable and reactive modified mesoporous hybrid materials with intact heteropolyacid (HPA) Keggin units is reported. The materials were synthesized by a one-pot method using a mixture of structure directing agents (CTAB and Pluronic 123). The key to obtaining stable HPA containing materials was the sequential calcination and extraction steps, and the authors present a working hypothesis for this phenomenon. Studies of several other key factors concerning synthetic parameters are also reported. All materials were characterized by a wide variety of physical and spectroscopic methods at each step of the synthesis. The kinetics of the esterification of acetic acid by butanol mediated by several of the materials and relevant homogeneous acid catalysts showed that the HPA containing hybrid material was a highly effective catalyst.

Full text of DOI:10.1039/b816172k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
New insights into the encapsulation and stabilization of heteropolyacids inside
the pore walls of mesostructured silica materials†
a
b
Veronique Dufaud, Frederic Lefebvre, Gerald P. Niccolaic and Mimoun Aouined
*
*
ꢀ
ꢀ ꢀ
Received 18th September 2008, Accepted 24th November 2008
First published as an Advance Article on the web 20th January 2009
DOI: 10.1039/b816172k
A method of synthesizing stable and reactive modified mesoporous hybrid materials with intact
heteropolyacid (HPA) Keggin units is reported. The materials were synthesized by a one-pot method
using a mixture of structure directing agents (CTAB and Pluronic 123). The key to obtaining stable
HPA containing materials was the sequential calcination and extraction steps, and the authors present
a working hypothesis for this phenomenon. Studies of several other key factors concerning synthetic
parameters are also reported. All materials were characterized by a wide variety of physical and
spectroscopic methods at each step of the synthesis. The kinetics of the esterification of acetic acid by
butanol mediated by several of the materials and relevant homogeneous acid catalysts showed that the
HPA containing hybrid material was a highly effective catalyst.
precipitation of HPAs on the outer surface of silica support,5d
1. Introduction
leaching of the active species and/or loss of mesoporous ordering
in the material.6a
The diversity of known polyoxometalates (POMs) provides
a broad spectrum of chemical properties in terms of morphology,
polarity, oxidation potentials, acidity and solubility.1 Thus, one
encounters polyoxometalates in very diverse settings (medicine,
catalysis, material science).2 In catalysis, polyoxometalate
compounds have been extensively used as acid and oxidation
catalysts in many reactions since their acid–base and redox
properties can be tuned easily by simply changing polyanion
chemical composition.2a,3 Their utility for industrial application
has been limited, notably due to their tendency toward low
surface area (1–10 m2/g). Considerable effort has been directed
towards their heterogenization onto large surface area supports.
In particular, periodic mesoporous silica materials have been
used to take advantage of their uniform and narrow pore size
distributions, highly ordered structure and high thermal
stability.4
Polyoxometalates immobilization by stronger chemical
bonding was also reported using organically modified meso-
porous silica.7 For example, an HPA was immobilized inside the
channels of MCM-41 which had been modified by amine func-
tional groups: the HPA was held through an ionic bond (anion,
ammonium) and catalytic activity was shown to be equivalent to
free HPA in solution. The strong ionic bond leads to higher
stability for the catalyst in polar solvent media.7a Transition-
metal
substituted
polyoxometalates
of
the
type
[MII(H2O)PW11O39]5ꢀ (M ¼ Co, Zn), have also been chemically
anchored to aminopropyl modified silica surfaces through dative
bonding.7b The open coordination site available to these metal
clusters allows them to bind various functional ligands leading to
the creation of solids containing highly dispersed clusters. More
recently, polyoxometalate compounds were introduced onto
periodic ordered mesoporous silica by covalent linkages.8 This
interesting approach involves the co-condensation of TEOS
(tetraethoxysilane) with Keggin-type monovacant poly-
oxometalate of the type SiW11O398ꢀ, in the presence of block
copolymers. Covalent Si–O–W bonds were readily formed from
the reaction of TEOS with the lacunary SiW11 leading to the
attachment of the POM to the mesopore walls as a ‘‘SiW11Si2’’
intermediate. The obtained hybrid materials exhibited higher
stability in water-leaching experiments compared to impregnated
samples suggesting that the nature of the link but also the
strategy of synthesizing the material is important.
POM-functionalized ordered mesoporous materials have been
obtained either by impregnation5 and/or by direct synthesis using
sol–gel techniques.6 In such cases, the HPAs are not covalently
connected to the silica surface, but are presumably held by weak
interactions between surface silanols and the HPA acidic
protons, and so some degradation of the modified materials by
polar solvents may occur, including clustering by dissolution/
a
ꢀ
Laboratoire de Chimie, UMR 5182 CNRS-ENS Lyon, 46 Allee d’Italie,
F-69364 Lyon cedex 07, France. E-mail: vdufaud@ens-lyon.fr
bLaboratoire de Chimie, Catalyse, Polymeꢁres et Proceꢀdeꢀs UMR 5265
CNRS-CPE Lyon, 43 bd du 11 Novembre 1918, 69622 Villeurbanne
In this paper we wish to report our efforts to obtain stable and
reactive modified mesoporous hybrid materials with intact het-
eropolyacid (HPA) Keggin units which are resistant to extrac-
tion. Different methods are reported, satisfactory results
depending on a three step synthetic sequence: co-condensation of
HPA and silica precursors, followed by calcination then extrac-
tion. The state of the solid structure and the integrity of the
Keggin units were characterized at each major step of
cedex, France. E-mail: lefebvre@cpe.fr
c
ꢀ
ICAR, UMR 5191 ENS LSH, 15 parvis Rene-Descartes, 69342 Lyon
cedex 07, France
dInstitut de Recherches sur la Catalyse et l’Environnement de Lyon, 2
avenue Albert Einstein, 69626 Villeurbanne cedex, France
† Electronic supplementary information (ESI) available: TGA, IR
spectra, 29Si CP MAS NMR, XRD patterns, TEM micrographs,
TEM/EDX analysis, BET, pore size distribution, DR UV-Vis and
Raman spectra. See DOI: 10.1039/b816172k
1142 | J. Mater. Chem., 2009, 19, 1142–1150
This journal is ª The Royal Society of Chemistry 2009

View Article Online
elaboration by several methods among the following: X-ray
powder diffraction at small angles, elemental analysis, ther-
mogravimetric analysis, infrared spectroscopy, diffuse reflec-
tance UV-Visible spectroscopy, 31P NMR spectroscopy, nitrogen
sorptions, and transmission electron microscopy (with EDX).
The accessibility of the heteropolyacid groups was shown
through a model catalytic esterification reaction.
409PC was used for simultaneous thermal analysis combining
thermogravimetric (TGA) and differential thermoanalysis
(DTA) at a heating rate of 10 C minꢀ1 in air from 25–1000 ꢁC.
ꢁ
31P MAS NMR spectra were recorded on a Bruker DSX-300
spectrometer operating at 121.51 MHz with a classical 4 mm
probehead allowing spinning rates up to 10 kHz. The spectra
were obtained by direct irradiation of phosphorus and proton
decoupling. In all cases it was checked that there was a sufficient
delay between the scans allowing a full relaxation of the nuclei.
Depending of the amount of phosphorus in the samples and the
nature of the samples between 100 and 10 000 scans were accu-
mulated. The chemical shifts are given relative to external 85%
H3PO4. Liquid NMR spectra were recorded on a Bruker AC-300
spectrometer and referenced as following: 1H (300 MHz),
internal SiMe4 at d ¼ 0.00 ppm, 13C (75 MHz), internal CDCl3 at
d ¼ 77.2 ppm, and 31P (121 MHz), external 85% H3PO4 at d ¼
0.00 ppm. Metal determinations were performed by ICP-OES
(Activa Jobin Yvon) spectroscopy from a solution obtained by
treatment of the solid catalyst with a mixture of HF, HNO3 and
H2SO4 in a Teflon reactor at 150 ꢁC. High resolution trans-
mission electron microscopy (HRTEM) was performed with
a 200 kV JEOL 2010 microscope, with a point to point resolution
of 0.195 nm (spherical aberration coefficient Cs ¼ 0.5 mm). This
instrument was equipped with an EDS X-ray analyzer LINK-
ISIS. The preparation of the sample was performed by dispersing
the catalyst powder within an EPON resin. Very thin slices (20–
40 nm) were obtained by ultramicrotomy of the sample. The
Ultramicrotome is a Leica Ultra cut UCT equipped with a dia-
mond knife. The slices are then recovered on a microscopy
copper grid (3.05 mm; 200 mesh) previously coated with a holey-
carbon film. The Raman spectra were achieved on powders at
room temperature with a LabRAM HR spectrometer (Jobin
Yvon) equipped with a CCD detector. The emission line at
514.53 nm from an Ar+ ion laser (Spectra Physics) was focused
on the sample under the microscope, the analyzing spot being ꢂ1
mm. Time of acquisition was varied according to the intensity of
2. Experimental
2.1 Materials
Tetraethoxysilane (TEOS), poly- (ethylene oxide)–poly-
(propylene oxide)–poly(ethylene oxide) block copolymer
(Pluronic 123, MW: 5000) and phosphomolybdic acid,
H3PMo12O40$xH2O, were purchased from Aldrich Chemicals
and used without further purification. Cetyltrimethylammonium
bromide and tungstophosphoric acid, H3PW12O40$xH2O, were
obtained from Acros.
2.2 One pot synthesis of H3PW12O40 modified SBA-15 type
silica using a mixed surfactant system
The general approach for the synthesis of HPA modified hybrid
materials involves the polymerization of a silica precursor in the
presence of heteropolyacid using a mixture of non-ionic (Plur-
onic 123) and ionic (CTAB) structure directing agents. Compo-
sition for 4 g Pluronic 123: 125 g HCl (1.9 M); 8.16 g TEOS
(0.0392 mol); 2.3 g H3PW12O40 (8 10ꢀ4 mol); 0.2 g CTAB (5.5
10ꢀ4 mol). In a typical procedure, tetraethoxysilane (TEOS),
aqueous acid (HCl) and templates were combined and allowed to
react at 40 ꢁC for 30 minutes (prehydrolysis reaction). The
phosphotungstic acid, solubilized in a minimum of water, was
then introduced dropwise. The temperature was maintained at 40
ꢁ
ꢁC for 20 hours followed by ageing during 24 hours at 100 C.
The solid was recovered by filtration, washed with water and
ꢁ
dried at 70 C (as-made solid, 2). The templates were removed
the Raman scattering. The spectral resolution was about 2 cmꢀ1
.
quantitatively from the as-made material by calcination (2-C)
followed by extraction with methanol (2-CE). For comparison,
blank materials (without H3PW12O40) were prepared according
to the same procedure and are referred to as 1 and 1-CE for the
as-made and calcined/extracted solids, respectively.
Homogeneity of the samples was checked focusing the laser
beam on various points.
2.4 Catalytic tests
Protocol. The liquid phase esterification of acetic acid with
n-butanol was chosen as a model reaction to evaluate the
performance of our hybrid materials. The catalytic reaction was
carried out in a two-necked flask under nitrogen. The reaction
mixture was composed of acetic acid (5.24 g, 0.087 mol),
n-butanol (7.29 g, 0.098 mol), nonane (0.80 g, 6.23 10ꢀ3 mol) and
toluene (4.9 ml). When a strong acid catalyst was added the
amount was adjusted to give 7.85 ꢃ 10ꢀ6 mol based on
H3PW12O40 (100 mg for 2-CE and 22.6 mg for molecular
H3PW12O40). In the case of the silica blank, 100 mg of 1-CE was
2.3 Characterization
Low-angle X-ray powder diffraction (XRD) data were acquired
on a Bruker D5005 diffractometer using Cu Ka monochromatic
˚
radiation (l ¼ 1.054184 A). Nitrogen adsorption–desorption
isotherms at 77 K were measured using a Micromeritics ASAP
2010M physisorption analyzer. The samples were evacuated at
ꢁ
180 C for 24 h before the measurements. Specific surface areas
were calculated following the BET procedure. Pore size distri-
bution was obtained by using the BJH pore analysis applied to
the desorption branch of the nitrogen adsorption/–desorption
isotherm. Infrared spectra were recorded from KBr pellets using
a Mattson 3000 IRFT spectrometer. Liquid UV-vis spectra were
recorded using a Vector 550 Bruker spectrometer. Solid UV-vis
and near infrared spectra were recorded from aluminium cells
with Suprasil 300 quartz windows, using a PerkinElmer Lambda
950 and PE Winlab software. A Netzsch thermoanalyser STA
ꢁ
used. The reactor was placed in a preheated oil bath at 80 C.
Analysis. The reaction was monitored by taking aliquots (0.2
ml) periodically (30 minutes), which were diluted with toluene
(ꢂ1 ml) and analyzed on a Varian CP-3800 gas chromatograph
equipped with a flame ionization detector, a Varian CP-8400
autosampler and a CP-Sil5CB capillary column (30 m, 0.32 mm
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1142–1150 | 1143

View Article Online
internal diameter, 0.25 mm film thickness). Nitrogen was used as
gas carrier. Conversion and yield were determined by GC based
on relative area of the GC-signals referred to an internal stan-
dard (nonane) calibrated to the corresponding pure compounds.
Thus, the polycondensation of TEOS in the presence of
a mixture of non-ionic (Pluronic 123) and ionic (CTAB) structure
directing agents was carried out in aqueous HCl at 40 ꢁC for 30
minutes before dropwise addition of phosphotungsticꢁ acid,
H3PW12O40 (aq). The temperature was maintained at 40 C for
ꢁ
20 hours followed by ageing at 100 C. The solid was recovered
by filtration, washed with water and dried at 70 ꢁC. The two-
template method was performed without addition of HPA
(blank material 1) and was reproduced with the addition of
different HPAs (based on tungsten and molybdenum) at various
theoretical loadings. One example of a stable material in which
H3PW12O40 was incorporated by direct synthesis, 2, serves as the
basis for most of the other materials in this report.
3. Results and discussion
The overall objective of the project was the production of an
HPA containing catalytic material with the following criteria:
high surface area materials in which stable, intact Keggin units
have been integrated without significant changes in reactivity
compared to the HPA in solution. Thus, one must balance the
need for a robust link to the surface with the avoidance of
profound changes in the chemical nature of the HPA. In order to
better understand each step of the synthesis, a high level of
characterisation was required throughout involving both
molecular level techniques (variously NMR, Raman, UV-Vis)
and bulk properties (XRD, TGA, BET). Finally, the materials
were tested as catalysts, with the sole objective of demonstrating
that the active HPA species was fully accessible.
The X-ray diffraction (XRD) powder pattern of the ‘‘as-made’’
material, 2, showed three clear peaks in the 2q-range of 0.6 to 3ꢁ,
characteristic of hexagonally ordered mesophases, with a d(100)
˚
spacing of 98 A (see Fig. 1, left, and Table 1). Analogous results
were obtained for the blank material 1 (Table 1). In the case of 2,
the presence of H3PW12O40 was indicated by elemental analysis
(Table 1, W 12.8 ꢄ 0.05%wt, P 0.2 ꢄ 0.05%wt, W/P ¼ 11 ꢄ 2.5)
and the structural integrity of the HPA Keggin unit was
confirmed by the presence of a single, very sharp peak at ꢀ15.3
ppm in the MAS 31P NMR spectrum (Fig. 1, right).
3.1 Mixed surfactant templating approach for the one pot
synthesis of H3PW12O40 modified SBA-15 type silica
3.2 Template removal methods and HPA Keggin unit stability
Given the failure of post-synthetic methods to provide stable
modified materials, we chose to pursue direct synthetic or one-
pot-approaches to our objective. The basic SBA-15 protocol
chosen involves the polycondensation of a silica precursor,
typically tetraethoxysilane (TEOS), in the presence of a non-
ionic structure directing agent (SDA, Pluronic 123).9,4c As has
been shown elsewhere, superior results, in terms of mesoscopic
order, for organically modified SBA-15 silica were obtained by
first allowing polycondensation of TEOS for a fixed period
before introduction of the organic moiety.10 It has also been
reported that the use of an ionic SDA, cetyltrimethylammonium
bromide (CTAB) favors stronger HPA/support interactions, but
attempts to use pure CTAB leads to HPA precipitation from the
synthetic mixture.11 We postulated that addition of smaller
quantities of CTAB to a synthesis based on Pluronic 123 as SDA
might ameliorate HPA inclusion without leading to precipitation
and without disrupting the ordering of the micelles.
The removal of the SDA from the pores while maintaining the
ordering of the mesostructure is a key issue to address particu-
larly for direct synthesis. In this study, both of the generally cited
methods, extraction of the template and calcination led to
quantitative removal of the SDA resulting in highly ordered,
SDA free materials (2-E and 2-C, respectively). The extraction
was performed by the soxhlet method using methanol (16 hours).
The calcination was performed overnight under a flow of air at
500 ꢁC. Extraction of the templates was confirmed by TGA and
by infrared spectroscopy. In the case of the as-made material 2,
ꢁ
a 22.7% weight loss was observed between 80 to 160 C in the
TGA, whereas in the 2-E and 2-C materials, no notable weight
loss was observed above 80 ꢁC (see ESI Fig. S1†). Furthermore,
the n(CH) bands in the infrared spectrum of 2 were not at all
presentintheinfraredspectraof2-Eand2-C(seeESIFig.S2†).The
X-ray diffractograms of 2-E and 2-C (see Fig. 2, left, (a) and (b)
Fig. 1 XRD pattern (left) and MAS 31P NMR (right) of 2.
1144 | J. Mater. Chem., 2009, 19, 1142–1150
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Table 1 Physical and textural properties of SBA-15 type silica materials containing H3PW12O40
Structural and textural properties
a
d100 /A
b
Wall thickness /A
e
f
Dp /A
Smc/m2gꢀ1
Vmd/cm3gꢀ1
Vp /cm3gꢀ1
SBET/m2gꢀ1
˚
˚
˚
Sample
Si(%wt
)
W(%wt
)
1
—
—
14
—
—
12.8
98
94
98
—
53.9
—
—
64.7
—
—
0.022
—
—
1.00
—
—
54.4
—
—
831
—
1-CE
2
2-E
2-C
2-CE
35.2
26.1
25.8
0.14
23.6
17.3
102
92
96
44.6
39.2
43.1
231.5
133.5
152.5
0.094
0.057
0.068
0.94
0.452
0.53
73.5
67
67.5
740
305
428
a
b
d(100) spacing (the error is estimated as 3 A). Calculated by a0 ꢀ pore size with a0 ¼ 2d(100)/O3. c Micropore surface determined using the t-plot
method. d Micropore volume determined using the t-plot method. e Total pore volume at P/P0 ¼ 0.980. f Pore size from desorption branch applying the
BJH pore analysis.
˚
Fig. 2 XRD pattern (left) and MAS 31P NMR (right) of (a) 2-E; (b) 2-C; and (c) 2-CE.
respectively, and Table 1) indicate that the long range mesostruc-
tural order is maintained in all of the materials.
HPA modified material, 2-CE. Evidence for the Keggin unit
reconstruction is found in the 31P MAS NMR of 2-CE (Fig. 2,
right, spectrum c) where the only resonance observed was the fine
peak at ꢀ15 ppm corresponding to the hydrated HPA molecular
moiety.13 Microanalysis showed a small decrease of the tungsten
loading (17.3%wt for 2-CE vs. 23.6%wt for 2-C) which could be
due to the loss of HPA and/or to the partial filling of the meso-
porous material by the alcohol, but certainly not the total elimi-
nation of the HPA moiety as was observed for extraction without
calcination, 2-E. The X-ray diffractograms of 2-CE showed long
range mesostructural ordering, although some loss of order with
respect to 2-C and 2-E was indicated by the diminished peak
intensity (Fig. 2, left, diffractogram c). Furthermore, the charac-
teristic peaks of the HPA in the 3ꢁ to 80ꢁ (2q scale) region of the
diffractograms were absent, indicating that the HPA was homo-
geneously dispersed over the solid (see ESI, Fig. S4†).
The effect of SDA removal by these methods on the HPA
Keggin unit, as determined by 31P MAS NMR, appeared cata-
strophic, at least at first glance. When the SDA was removed by
soxhlet extraction using methanol, 2-E, no resonance peak
between ꢀ200 and +200 ppm appeared (Fig. 2, right, spectrum
a), suggesting that all of the HPA had been extracted along with
the templates. This observation was confirmed by microanalysis
where only traces of tungsten were found (Table 1, 0.14%wt for
2-E vs. 12.8%wt for 2).
The results for SDA removal by calcination, 2-C, were less
absolute. The tungsten microanalysis (Table 1) showed no
significant loss of tungsten in the material.12 However, in the 31
P
MAS NMR spectrum, a very broad and intense series of reso-
nances between 10 and ꢀ30 ppm surrounded the central, fine
peak at ꢀ15 ppm of 2-C (Fig. 2, right, spectrum b). Other
analogous materials prepared in this series (differing HPA,
different loading) showed only the broad resonances with no
such fine central peak whatsoever (see ESI Fig. S3†). Clearly,
calcination leads to significant, even quantitative breakdown of
the highly symmetric HPA Keggin unit of 2.
The TEM/EDX study of the surface of 2-CE showed well
ordered long range ordering of the material and, most impor-
tantly, that tungsten was uniformly present over most of the
ordered surface. Some ‘‘dark’’ pores showed high tungsten
content, presumably plugged by the aggregation of many HPA
units (see ESI Fig. S5†).
In the context of a leaching study, 2-C was submitted to
soxhlet extraction with methanol. To our initial surprise, this
treatment led to the reformation of the Keggin unit and a stable
The textural data derived from the BET analysis of nitrogen
adsorption and desorption experiments are presented in Table 1.
The materials exhibit type IV isotherms characteristic of
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1142–1150 | 1145

View Article Online
˚
mesoporous solids, with large pore diameters (55–75 A, see ESI
behind this approach was based on previous authors’ hypotheses
that such mixed micelles with ionic surfactants is an easy way to
change the surface charge distribution (at the surface of the
micelle) and may lead to stronger interactions between the
template and the inorganic precursor.14 We hypothesized that we
could pre-organize the highly polar HPA at the surface of the
micelle in an analogous manner.
Fig. S6–S8† for graphical data). It is interesting to note that while
the material which was synthesized in the absence of HPA, 1-CE,
had little microporosity (specific microporous volume of 0.022
cm3 gꢀ1 and microporous surface area of 64.7 m2 gꢀ1), the
materials prepared in the presence of HPA, 2-E and 2-CE, had
significant microporous nature (0.094 cm3 gꢀ1 and 231.5 m2 gꢀ1
;
0.068 cm3 gꢀ1 and 152.5 m2 gꢀ1 respectively). Also note that 2-E,
the solid in which all of the HPA has been removed, has signif-
icantly more microporous volume and surface than the HPA
containing 2-CE.
To confirm the importance of this approach, materials were
also prepared using more CTAB (the same mass of Pluronic, 1 :
1.6 molar Pluronic/CTAB ratio) and using only Pluronic. The
XRD study of these two materials (Fig. 3, bottom) shows that
mesoscopic ordering was present throughout (as-made, calcined,
and calcined-extracted stages) for these two templating methods.
The 31P MAS NMR spectra of the calcined-extracted materials
(Fig. 3, top, spectra c) show that the Keggin unit was not fully
restored (compared to Fig. 2c, right). In the absence of CTAB,
the central Keggin unit peak at 15.3 ppm was surrounded by
a very broad and intense resonance band, and in the case of the
higher CTAB concentration, no central peak at all was observed.
The very clean results obtained for the 1 : 0.8 molar ratio mixed
template appear to be near optimal.
Thus, material 2-CE appears to conform to the structural
criteria of the synthetic project. The 31P MAS NMR and the
diffuse reflectance UV-Vis and infrared spectroscopies (ESI
Fig. S9 and Fig. S10† respectively) would seem to indicate that
chemical properties of the HPA inclusions may be similar to the
free acid. Before confirming the accessibility and chemical
activity of the modified material by catalytic benchmarking,
a brief discussion of several side studies is presented.
3.3 Investigation of synthesis conditions
Influence of the concentration of CTAB. As mentioned above,
the materials 2 are based on a synthetic procedure in which the
SDA mixture is Pluronic/CTAB in a 1 : 0.8 molar ratio. The logic
Influence of the prehydrolysis reaction time. As mentioned
above, previous authors have indicated that the incorporation of
HPA to the synthesis is facilitated by allowing some
Fig. 3 XRD pattern (bottom) and MAS 31P NMR (top) of H3PW12O40 modified SBA-15 type silica in the absence (left) and in the presence (right) of
CTAB: (a) as-made; (b) a after calcination; and (c) b after extraction.
1146 | J. Mater. Chem., 2009, 19, 1142–1150
This journal is ª The Royal Society of Chemistry 2009

View Article Online
Fig. 4 MAS 31P NMR (left) and XRD pattern (right) of H3PW12O40 modified SBA-15 type silica after simultaneous introduction of TEOS and
tungstophosphoric acid: (a) as-made; (b) a after calcination; and (c) b after extraction.
polymerisation of the silica precursor (TEOS) before introduc-
tion of the functionalised silica precursors. The present study
confirms a similar effect in the case of HPA inclusion in the
synthetic protocol. When TEOS and HPA were simultaneously
added to the aqueous solution of the mixed SDA templates, an
ordered solid was obtained which includes the HPA in its intact
Keggin unit. (Fig 4, bottom left). Mesoscopic order was main-
tained throughout the calcination and extraction, but again the
restoral of the Keggin unit in the extracted-calcined product was
far from complete (Fig. 4, top left).
3.4 Demonstration of the accessibility of the HPA sites
To establish that the stabilization of the HPA in the walls of the
material does not inhibit its catalytic properties, a series of
catalytic experiments were undertaken using a catalytic bench-
mark reaction, the esterification of acetic acid by n-butanol. The
basic protocol and kinetic analysis of Sepulveda and coworkers
were used to determine the observed second order rate constant
for the reaction.16 If the heteropolyacid sites are accessible, the
catalytic reaction rate should be superior to a blank system (no
catalyst, unmodified SBA-15.).
Rate constants were determined by fitting experimental data to
the kinetic model described by Sepulveda. The basic second order
rate law was tested:
Applicability of the synthetic method to various HPAs. Initial
feasibility studies were performed on other heteropolyacids,
namely H3PMo12O40 and Na3PW12O40. Encouraging results in
terms of mesoscopic order and Keggin unit restoration were
obtained (see Fig. 5 and 6) and further studies into the optimi-
sation of the synthesis and the applications of these materials are
underway. It is also interesting to note the failure of the mixed
addenda phosphovanadomolybdic acid H4PVMo11O40 in this
protocol. Again, little restructuring of the Keggin unit was
observed in the calcined-extracted material, but significantly,
elemental analyses indicated the total absence of vanadium after
calcination. Clearly, the vanadium was irreversibly extracted
from the polyanion framework upon thermal treatment.15
d½Aꢅ
ꢀ
¼ k½Aꢅ½Bꢅ
dt
where [A] is the concentration of acetic acid, [B] is the concen-
tration of butanol and t is time of reaction. The units are chosen
such that the rate constant k is expressed in ml molꢀ1 hꢀ1. The
integration of this equation leads to the expression:
ꢀ
ꢁ
 Ã
1
Mð1 ꢀ cAÞ
M ꢀ cA
A kt ¼
ln
1 ꢀ M
Fig. 5 XRD pattern (right) and MAS 31P NMR (left) of H3PMo12O40 modified SBA-15 type silica: (a) as-made; (b) a after calcination; and (c) b after
extraction (Mo 6.73%wt).
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1142–1150 | 1147

View Article Online
Fig. 6 XRD pattern (right) and MAS 31P NMR (left) of Na3PW12O40 modified SBA-15 type silica: (a) as-made; (b) a after calcination; and (c) b after
extraction (W 6.75%wt; Na 0.57%wt).
where M is the ratio of the initial concentrations of acetic acid
and butanol:
constant. Given that the acid catalysts were added such that the
ratio of substrate to strong acid was equal for those runs, and
that other relevant parameters were constant (reactant ratios,
temperature, solvent, etc.) these rate constants can be compared
in a meaningful manner.
½A0ꢅ
M ¼
½B0ꢅ
Four systems were tested in which all parameters except the
presence and identity of the catalytic adduct were kept constant:
2-CE, molecular H3PW12O40, an SBA-15 having no added HPA,
and no acid adduct. In the first two cases, where strong acid
catalysts were used, the substrate/acid molar ratio was main-
tained at ꢂ11 000.17 In all cases, the observed data
and cA is the conversion of acetic acid:
½Aꢅ ꢀ½Aꢅ
0
cA
½Aꢅ
The experimental data is plotted in the form of ‘‘kt’’ vs. time,
¼
0
and thus the slope of the plot is equal to the observed rate
Fig. 7 Esterification of acetic acid by n-butanol: conversion of acetic acid versus time (left) and rate constant determination (right).
1148 | J. Mater. Chem., 2009, 19, 1142–1150 This journal is ª The Royal Society of Chemistry 2009

View Article Online
Table 2 Esterification of acetic acid by n-butanol
Second order rate constantb/
ml molꢀ1 hꢀ1
Catalyst
Conversiona at t ¼ 6 h
Homogeneous catalyst
Heterogeneous catalysts
Blank
H3PW12O40
24%
49%
13%
14%
16
49
6.6
7.7
HPA@SBA-15-CE
Silica blank
No catalyst
a
Conversions based on unreacted acetic acid were determined by GC. b Determined at 80 ꢁC by fitting experimental data to the kinetic model described
by Sepulveda et al. Applied Catalysis A: General 2005, 288, 18.
Scheme 1 Sketch of proposed interpretation of data.
(concentration of acetic acid followed by GC) exhibit an excel-
lent fit to the second order kinetic model and thus the second
order rate constants could be determined (see Fig. 7 and Table 2).
possibility would be that calcination has reduced the size of the
entrance of the micropore host of the, now thermally degraded,
HPA moiety (Scheme 1, top center). When the methanol
extraction of the calcined material is performed, that part of the
degraded HPA which was on the surface is eliminated. However,
the surface restructuration inhibits the extraction of the HPA in
the micropores, and methanol-mediated reformation of the
Keggin structure takes place in the solid.18 Thus, the final
material contains trapped, structurally intact and accessible HPA
moieties at the surface of the channels of the mesoporous silica
(Scheme 1, top right).
ꢁ
The second order rate constant determined at 80 C for 2-CE
(49 ml molꢀ1 hꢀ1) was more than three times higher than that
determined for the molecular H3PW12O40 (16 ml molꢀ1 hꢀ1) and
6–7 times higher than the SBA-15 only and the no catalyst blank
reactions (6.6, and 7.4 ml molꢀ1 hꢀ1 respectively).
Clearly, the HPA moiety of 2-CE was accessible to reactants
and active, the primary objective of this series of experiments.
The factors leading to this enhancement of the activity of the
solid supported catalyst with respect to the homogeneous system
(enhanced acidity, confinement .) as well as the stability of the
solid with respect to leaching are under further study.
4. Conclusion
In summary, new, stable and reactive HPA modified mesoporous
hybrid materials have been obtained by one-pot synthesis in the
presence of HPA and two template molecules, one ionic and one
non-ionic, together. Template removal by calcination–extraction
sequence was essential to yield stable materials with intact
Keggin unit. The catalytic results showed that not only the sites
were accessible, but also that the catalytic activity of the new
material was three times that of the homogeneous HPA system.
Further work is currently underway to demonstrate the struc-
tural factors leading to the stabilization of the HPA unit during
template removal and to extend the scope of this new synthetic
method to multifunctional catalyst design.
3.5 Proposed rationalisation of observations
At this stage, the reasons behind the reconstruction of the Keggin
unit into a stably fixed and reactive form have not been
demonstrated, but we believe that our working hypothesis merits
mention in this report. The proposed explanation, represented in
Scheme 1, is based on two key hypotheses, each of which
warrants further study.
Firstly, we propose that when the HPA is introduced, it takes
up a position at the periphery of the micelle creating surface
microporocity by molecular imprinting. Thus, after polymeri-
sation of TEOS, some of the HPA is at the center of micropores
in the interior walls of the as-made material (Scheme 1, upper
left). The extraction protocol, if performed at this step, leads to
total elimination of the HPA from these micropores (Scheme 1,
lower left).
References
1 M. T. Pope, Heteropoly and isopoly oxometalates, Springer-Verlag,
Berlin, 1983.
2 (a) N. Mizuno and M. Misono, Chem. Rev., 1998, 98, 199–217; (b)
D. E. Katsoulis, Chem. Rev., 1998, 98, 359–387; (c) J. T.
Rhule, G. L. Hill and D. A. Judd, Chem. Rev., 1998, 98, 327–357;
The second key hypothesis is that calcination leads to some
restructuring of the surface which inhibits extraction. One
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 1142–1150 | 1149

View Article Online
(d) I. V. Kozhevnikov, Catalysis by Polyoxometalates, Wiley,
Chichester, England, 2002.
9 D. Zhao, Q. Huo, J. Feng, B. F. Chmelka and G. D. Stucky, J. Am.
Chem. Soc., 1998, 120, 6024–6036.
3 R. D. Hetterley, E. F. Kozhevnikov and I. V. Kozhevnikov, Chem.
Rev., 2006, 782.
10 D. Margolese, J. A. Melero, S. C. Christiansen, B. F. Chmelka and
G. D. Stucky, Chem. Mater., 2000, 12, 2448–2459.
4 (a) J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz,
C. T. Kresge, K. D. Schmitt, C. T.-W. Chu, D. H. Olson,
E. W. Sheppard, S. B. McCullen, J. B. Higgins and J. L. Schlenker,
J. Am. Chem. Soc., 1992, 114, 10834–10843; (b) C. T. Kresge,
M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature,
1992, 359, 710–712; (c) D. Zhao, J. Feng, Q. Huo, N. Melosh,
G. H. Fredrickson, B. F. Chmelka and G. D. Stucky, Science, 1998,
279, 548–552.
11 J. Toufaily, M. Soulard and J.-L. Guth, C. R., Acad. Sci. Paris,
Chimie/Chemistry, 2001, 4, 675–681.
12 Recall that calcination removed the template molecules, and so some
increase in the tungsten analysis (W 23.6 ꢄ 0.05%wt) is normal.
13 The same phenomenon was observed when a batch extraction with
water was performed.
14 L.-R. Dai, T.-W. Wang, L.-T. Bu and G. Chen, Colloids and Surfaces
A: Physicochem. Eng. Aspects, 2001, 181, 151–157.
ꢀ ꢀ
ꢀ
15 (a) R. Bayer, C. Marchal, F. X. Liu, A. Teze and G. Herve, J. Mol.
Catal. A: Chem., 1996, 110, 65; (b) R. Bayer, C. Marchal-Roch,
5 (a) A. Lapkin, B. Bozkaya, T. Mays, L. Borello, K. Edler and
B. Crittenden, Catalysis Today, 2003, 81, 611–621; (b) A. Ghanbari-
Siahkali, A. Philippou, J. Dwyer and M. W. Anderson, Applied
Catalysis A: General, 2000, 192, 57–69; (c) T. Blasco, A. Corma,
A. Martinez and P. M. Escolano, J. Catal., 1998, 177, 306–313; (d)
M. J. Verhoef, P. J. Kooyman, J. A. Peters and H. Van Bekkum,
Microporous and Mesoporous Materials, 1999, 27, 365;
I. V. Kozhevnikov, K. R. Kloestra, A. Sinnema, H. W. Zandbergen
and H. van Bekkum, Journal of Molecular Catalysis A: Chemical,
1996, 114, 287–298.
ꢀ ꢀ
F. X. Liu, A. Teze and G. Herve, J. Mol. Catal. A: Chem., 1996,
114, 277; (c) T. Ilkenhans, B. Herzog, T. Braun and R. Schlogl,
J. Catal., 1995, 153, 275; (d) E. Cadot, C. Marchal, M. Fournier,
ꢀ ꢀ
A. Teze and G. Herve, inPolyoxometalates: From Platonic Solids to
Anti-Retroviral Activity, ed. M. T. Pope and A. Muller, Kluwer,
ꢀ
€
ꢀ
€
Dordrecht, 1994, p. 315.
ꢁ
16 J. H. Sepulveda, J. C. Yori and C. R. Ver, Applied Catalysis A:
General, 2005, 288, 18–24.
6 (a) K. Nowinska, R. Formaniak, W. Kaleta and A. Waclaw, Applied
Catalysis A: General, 2003, 256, 115–123; (b) H. S. Yun,
M. Kuwabara, H. S. Zhou and I. Honma, Journal of Materials
Science, 2004, 39, 2341–2347; (c) L. Yang, y. Qi, X. Yuan, J. Shen
and J. Kim, Journal of Molecular Catalysis A: Chemical, 2005, 229,
199–205; (d) J. Toufaily, M. Soulard, J.-L. Guth, J. Patarin,
L. Delmote, T. Hamieh, M. Kodeih, D. Naoufal and H. Hamad,
Colloids and Surfaces A: Physicochemical and Engineering Aspects,
2008, 316, 285–291.
7 (a) W. Kaleta and K. Nowinska, Chem. Commun., 2001, 535–536; (b)
B. J. S. Johnson and A. Stein, Inorganic Chemistry, 2001, 40, 801–808;
(c) N. V. Maksmchuk, M. S. Melgunov, Y. A. Chesalov, J. Mrowiec-
Bialon, A. B. Jarzebski and O. A. Kholdeeva, J. Catal., 2007, 246,
241–248.
17 In the case of 2-CE, for the purposes of the calculation of the quantity
of catalyst added and the subsequent calculation of rate constants, it is
assumed that all HPA content is catalytically active. Thus, the results
represent the ‘‘best case’’ scenario. It actually may be the case that
a fraction of sites are active, and thus the substrate to catalyst ratio
is much higher and the effective second order rate constant reported
below is underestimated. These considerations do not affect the
central purpose of the catalytic study, the test of whether some
HPA contained in the mesopore walls is available for catalytic
reaction.
18 (a) Similar reconstruction has been demonstrated for thermally
decomposed HPA when exposed to water vapour. See
€
K. Bruckman, M. J. CheHaber and J. M. Tatibouet, Catal. Lett.,
1994, 25, 225, and (b) Y. Konishi, K. Sakata, M. Misono and
Y. Yoneda, J. Catal., 1982, 7, 169.
8 R. Zhang and C. Yang, J. Mater. Chem., 2008, 18, 2691–2703.
1150 | J. Mater. Chem., 2009, 19, 1142–1150
This journal is ª The Royal Society of Chemistry 2009