Aqueous-sensitive reaction sites in sulfonic acid-functionalized mesoporous silicas

Article abstract of DOI:10.1016/j.jcat.2007.12.011

Local differences in surface hydrophilicities/hydrophobicities of propyl- and arene-sulfonic-acid modified mesoporous silica and organosilica catalysts have been compared and correlated with their bulk catalytic properties for aqueous-sensitive organic reactions. Syntheses of propyl- and arene-SO₃H-modified mesoporous silicas and organosilicas yield materials with different hydrophilicities, especially when ethylsiloxane moieties are incorporated into the silica frameworks. Solid-state two-dimensional (2D) ¹³C{¹H} and ²⁹Si{¹H} heteronuclear correlation (HETCOR) NMR spectra prove that the incorporation of hydrophobic ethylsiloxane groups into functionalized mesoporous silica frameworks result in reduced interactions of adsorbed water with the silica framework in general and, importantly, in the immediate vicinities of the SO₃H active sites. The hydrophilic/hydrophobic character of the surface, as well as the active site properties depend on the functional species attached. Propyl-sulfonic acid moieties are less acidic but more hydrophobic than arene-SO₃H species, leading to superior overall activities for water-mediated acid-catalyzed organic reactions. The etherification of vanillyl alcohol (4-hydroxy-3-methoxybenzylalcohol) with 1-hexanol to yield 4-hydroxy-3-methoxybenzyl-1-hexyl ether is shown to proceed significantly more effectively on SO₃H-modified mesoporous organosilicas, compared to wholly siliceous mesoporous supports. The correlation of macroscopic adsorption and reaction results with 2D NMR measurements allows the hydrophilic/hydrophobic surface properties of the mesoporous support to be optimized with respect to water-retention capacities and activities for water-sensitive organic reactions.

Full text of DOI:10.1016/j.jcat.2007.12.011

Journal of Catalysis 254 (2008) 205–217
www.elsevier.com/locate/jcat
Aqueous-sensitive reaction sites in sulfonic acid-functionalized
mesoporous silicas
Gabriel Morales a, George Athens b, Bradley F. Chmelka b, Rafael van Grieken a, Juan A. Melero a,∗
a
Department of Chemical and Environmental Technology, ESCET, Rey Juan Carlos University, C/Tulipán s/n, 28933 Móstoles (Madrid), Spain
b
Department of Chemical Engineering, University of California, Santa Barbara, CA 93106, USA
Received 7 November 2007; revised 17 December 2007; accepted 18 December 2007
Abstract
Local differences in surface hydrophilicities/hydrophobicities of propyl- and arene-sulfonic-acid modified mesoporous silica and organosilica
catalysts have been compared and correlated with their bulk catalytic properties for aqueous-sensitive organic reactions. Syntheses of propyl-
and arene-SO H-modified mesoporous silicas and organosilicas yield materials with different hydrophilicities, especially when ethylsiloxane
3
13
1
29
1
moieties are incorporated into the silica frameworks. Solid-state two-dimensional (2D) C{ H} and Si{ H} heteronuclear correlation (HET-
COR) NMR spectra prove that the incorporation of hydrophobic ethylsiloxane groups into functionalized mesoporous silica frameworks result
in reduced interactions of adsorbed water with the silica framework in general and, importantly, in the immediate vicinities of the SO H active
3
sites. The hydrophilic/hydrophobic character of the surface, as well as the active site properties depend on the functional species attached. Propyl-
sulfonic acid moieties are less acidic but more hydrophobic than arene-SO H species, leading to superior overall activities for water-mediated
3
acid-catalyzed organic reactions. The etherification of vanillyl alcohol (4-hydroxy-3-methoxybenzylalcohol) with 1-hexanol to yield 4-hydroxy-
3-methoxybenzyl-1-hexyl ether is shown to proceed significantly more effectively on SO H-modified mesoporous organosilicas, compared to
3
wholly siliceous mesoporous supports. The correlation of macroscopic adsorption and reaction results with 2D NMR measurements allows the
hydrophilic/hydrophobic surface properties of the mesoporous support to be optimized with respect to water-retention capacities and activities for
water-sensitive organic reactions.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Mesostructured silica; Mesoporous organosilica; Surface hydrophobicity; Etherification; Benzyl alcohol; Sulfonic acid; 2D NMR
1. Introduction
portunity for new solid-acid catalysts to replace strong liquid
acids [3,4].
However, to achieve the combination or even a subset of
the process advantages that heterogeneous catalysts may per-
mit, numerous materials challenges must be overcome. These
include mitigating or exploiting the effects of reduced active
site mobility, lower entropy, confinement, diffusion limitations
to mass transport, and competitive adsorption at nearby surface
sites. Heterogeneous acid catalysts for fine chemical syntheses
face all of these issues and, in particular the latter, because of
the difficulty in managing the inherent amphiphilicity of the
organo-acid functional species. The relative incompatibilities
of the hydrophilic acid sites, their accompanying hydrophobic
organic links to the support, and their interactions with the or-
ganic reactants and products of interest tend to be exacerbated
when attached to solid surfaces.
The development of heterogeneous catalysts to replace ho-
mogeneous systems for the production of fine chemicals is
motivated by several potential process improvements [1,2].
These include easier separation of the catalyst from the reaction
medium, greater catalyst stability, improved regenerability, and
enhanced product selectivity, when the active sites are attached
to a solid surface, typically within high-surface-area porous
materials. Likewise, tightening environmental regulation of tra-
ditional acid-catalyzed chemical processes (alkylation, acyla-
tion, isomerization, oligomerization, etc.) has provided an op-
*
Corresponding author.
E-mail address: juan.melero@urjc.es (J.A. Melero).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.12.011

206
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
Such heterogenized catalysts are often based on porous sil-
yielding mesostructured organically-modified silica [11,13].
The incorporation of the acidic functional species occurs in a
single process step, after which the surfactant species are re-
moved to produce the porosity needed to allow access to the
acid sites in the mesopore channels. High concentrations of the
acid precursor species tend to disrupt mesostructural ordering
and make it difficult to control bulk morphologies (e.g., particle,
film, fiber, or monolith). For powders used in catalytic appli-
cations, however, these challenges are less troublesome, and
moreover, higher and more uniform surface concentrations of
acid-centers have been achieved using co-condensation strate-
gies, which, when suitable, are often more convenient than post-
synthetic grafting approaches.
In particular, a co-condensed, ‘one-pot’ synthesis protocol
of mesoporous solid-acid catalysts with hybrid organosilica
frameworks allows for substantial control of local hydropho-
bic/hydrophilic environments near the active site. This strategy
relies on the co-condensation of hydrolyzed trialkoxyorganosi-
lanes of the type (RꢀO)3Si–R–Si(ORꢀ)3, where the bridging
moiety, –R–, can be composed of various organic groups (e.g.,
ethane, benzene, thiophene, or biphenylene) [31–34], in what
have been referred to as ‘periodic mesoporous organosilicas’
(PMOs). These materials combine several advantageous as-
pects of both organic and inorganic species into mesoporous
solids with mechanical, adsorption, and reaction properties
that are different from either of the wholly organic or inor-
ganic components. Furthermore, recent work has shown im-
proved catalytic performance of sulfonic-acid active sites sup-
ported on mesoporous organosilica in water-sensitive reactions
[20,35–37]. The higher catalytic activity of these materials has
been attributed to increased hydrophobicity near the sulfonic-
acid moieties and enhanced diffusion of reactant and products
within the hydrophobic mesopores.
Thus, interactions among the –SO3H groups, adjacent moi-
eties (including framework and grafted species) and water
molecules are important to elucidate and control. A number of
different techniques have been employed to investigate struc-
tural properties, acid capacities, organic incorporation, oxida-
tion efficiencies and acid strength of these sulfonic acid modi-
fied mesostructured materials. Furthermore, advanced 2D solid-
state NMR spectroscopy techniques have been shown to be
powerful tools to characterize and identify sites, species and in-
teractions in mesostructured materials [38]. Recently, Trebosc
et al. [39] reported the utilization of high resolution solid-state
2D heteronuclear correlation (HETCOR) NMR spectroscopy
under fast magic angle spinning to provide highly resolved
spectra between protons and low-gamma nuclei (13C and 29Si)
in allyl-functionalized MCM-41 samples. These methods pro-
vide detailed structural characterization of surfaces of meso-
porous solid and herein they are applied for the first time to sul-
fonic acid-modified silica and organosilica materials. Specifi-
cally, interactions among the different 29Si, 13C, and 1H species
have been examined to obtain surface-structure correlations
with catalytic reaction properties. These are expected to be
important especially when highly polar molecules (e.g., water
molecules) are present in the reaction medium. We demon-
strate how control of molecular surface hydrophilicity or hy-
ica supports, primarily because of their high surface areas and
porosities, excellent stabilities (chemical and thermal), and
facile functionalization with catalytically active organic groups
with catalytic activity that can be robustly anchored to the
support surfaces. Mesostructured silica materials are partic-
ularly promising as supports for active species, due to their
very high surface areas (>800 m2/g), large pores (2–50 nm)
for low diffusional resistances to mass transport, and adjustable
and narrow pore size distributions with high active-site ac-
cessibilities [5–7]. A number of recent studies, including our
own, have reported the incorporation of sulfonic-acid groups
into mesostructured pore channels of MCM-41, HMS, or
SBA-type materials [8–15] to introduce acid-catalysis or ion-
conducting properties. Such properties, can furthermore be
modified by different moieties adjacent to the sulfonic-acid
site, such as arene-sulfonic [13] or perfluorosulfonic groups
[15–17], whose electron-withdrawing characters lead to en-
hanced acidity. Efforts to maximize the hydrophobicity [18]
or hydrophilicity [15] in the near vicinities of the sulfonic-
acid sites are enabling new applications for these materials.
For example, sulfonic-acid functionalized mesoporous sili-
cas have been shown to have improved reaction properties
over conventional homogeneous or heterogeneous catalysts
for a wide range of acid-catalyzed reactions, including es-
terification [12,19–21], condensation and addition [9,22,23],
etherification [24], rearrangement [25–27], Friedel-Crafts acy-
lation [28], alkylation [27] and conversion of biorenewable
molecules [29,30]. Recently they have also shown enhanced
proton-conductivities [15].
The anchoring of organic acid moieties to mesoporous silica
surfaces by covalent bonds has generally followed two proto-
cols, either post-synthetic grafting of the acid species to accessi-
ble pore surfaces after the silica framework has been formed or
co-condensation of the acid functional species simultaneously
as the silica framework cross-links. Post-synthesis grafting
methods are based primarily on the reaction of organosilanes
((RꢀO)3SiR) or chlorosilanes (Cl3SiR) with silanol groups on
the interior mesopore channels of previously synthesized (e.g.,
self-assembled and then calcined or extracted) mesoporous sil-
ica. This functionalization approach allows for a wide range of
organic species to be anchored to the silica surface without af-
fecting the mesostructural ordering of the separately prepared
mesoporous silica support. Furthermore, the versatile process-
ability of mesostructured block-copolymer-directed silica per-
mits facile control over particle or bulk morphologies and sep-
arate surface incorporation of multiple functional species into
the mesoporous silica. However, grafting techniques often lead
to rather low loading of functional groups and/or require the use
of multiple separate processing steps. Hence, there is interest in
the incorporation of relatively high functional groups loadings
by single-step co-assembly processes.
Co-condensation (frequently referred to also as “one-pot
syntheses”) involves the simultaneous condensation of tetra-
alkoxysilanes ((RO)4Si) with terminal tri-alkoxyorganosilanes
(RO)3SiRꢀ, where Rꢀ can be a sulfonic-acid-containing organic
moiety in the presence of a structure-directing surfactant agent,

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
207
drophobicity in propyl- and arene-sulfonic acid-modified PMO
materials can influence the etherification of vanillyl alcohol,
a bulky benzyl alcohol, in the presence of 1-hexanol, which
react to yield 4-hydroxy-3-methoxybenzyl-1-hexyl ether, a rep-
resentative fine chemical compounds used in perfumes and fla-
vorings.
The data were analysed using the BJH model on the adsorption
branch of the isotherm, and the pore volume (Vp) was taken at
point where P/P0 = 0.975 single point. Transmission electron
micrographs (TEM) were obtained using a JEOL 2000 electron
microscope operating at 200 kV.
Content and thermal stability of the organic and sulfonic-
acid species within the mesoporous and organosilica materials
were evaluated by elemental analysis (HCN) in a Vario EL III
apparatus, and also indirectly by means of thermogravimetry
analysis (TGA) (SDT 2960 Simultaneous DSC-TGA, from TA
Instruments). Samples were heated in a Mettler TGA/sDTA-
851e ThermoGravimetric Analyzer from 25 to 800 ◦C at a rate
of 10 ◦C/min under constant nitrogen gas flow of 50 ml/min.
The concentrations of surface sulfonic acid groups were deter-
mined as cation-exchange capacities by titration with 0.01 M
NaOH [11]. Table 1 summarizes data related to incorporation
of the organic species and the resulting acidic properties.
Temperature programmed desorption (TPD) experiments
were carried out in a Micromeritics AutoChem 2910 equipment
provided with thermal conductivity detector (TCD) using he-
lium as carrier gas. Prior to the TPD experiment, the sample
was outgassed by thermal treatment in He flow. After cooling,
the material was contacted with gas-phase water until saturation
by successive injections of pulses, in amounts ranging from 0.1
to 2 µl. Reversibly adsorbed water was then removed in He
flow for 2 h at 40 ◦C. TPD tests were carried out by heating
the sample from 40 to 400 ◦C at 15 ◦C/min with constant He
flow, keeping the final temperature for an additional period of
15 min. The effluent was continuously screened for water detec-
tion by means of a TCD detector. From TPD curves, the amount
2. Experimental
2.1. Catalyst preparation
Propyl-sulfonic-acid functionalized mesostructured silica
and organosilica were synthesized following a one-pot co-
condensation procedure previously employed for the prepara-
tion of propyl-sulfonic-acid functionalized silica SBA-15 [11].
For the case of the organosilica, 1,2-bis-(triethoxysilyl)ethane
(BTSE, Aldrich) was additionally incorporated. In a typi-
cal synthesis, 4 g of poly(alkylene oxide) triblock copoly-
mer (EO20PO70EO20) (Aldrich) were dissolved in 125 ml
of 1.9 M HCl. The solution was heated to 40 ◦C and then
tetraethoxysilane (TEOS, Aldrich), or a mixture of TEOS and
BTSE in the desired ratio, was added, after which the mix-
ture was allowed to prehydrolyze for 45 min. Then mercapto-
propyltrimethoxysilane (MPTMS, Aldrich) and an aqueous so-
lution of H2O2 (30 wt%, Merck) were added at once. The mix-
ture was stirred vigorously at 40 ◦C for 20 h and then aged at
100 ◦C for 24 h under static conditions. Solid products were re-
covered by filtration and air-dried overnight. The block copoly-
mer species were subsequently removed by ethanol washing
under reflux for 24 h (2 g of as-synthesized material per 200 ml
of ethanol). Molar composition of the mixtures, relative to 4 g
of triblock copolymer, were: (0.0369 − x)TEOS:(x/2)BTSE:
0.0041MPTMS:0.0369H2O2:0.24HCl: ≈ 6.67H2O, with x be-
ing either 0 or 0.0184. The amount of MPTMS precursor
species was fixed, so that approximately 10% of total silicon
atoms were functionalized with propyl-sulfonic acid groups.
Arene-sulfonic-acid functionalized mesostructured silica
and organosilica were synthesized similarly to that described
for the propyl-sulfonic-acid functionalized mesoporous solids
above [13]. In this case, following the prehydrolysis step of
the alkoxysilane precursor species (TEOS or TEOS + BTSE),
2-(4-chloro-sulfonylphenyl)-ethyl-trimethoxysilane (CSPTMS,
Gelest) was added. Afterwards, the synthesis proceeded as de-
scribed above. Molar composition of the mixtures, relative
to 4 g of triblock copolymer, were: (0.0369 − x)TEOS:(x/2)
BTSE:0.0041CSPTMS:0.24HCl: ≈ 6.67H2O, with x being 0
or 0.0184. Similarly, the amount of CSPTMS precursor species
was fixed, so that approximately 10% of total silicon atoms
were functionalized with arene-sulfonic-acid groups.
of water retained and desorbed (WH O) and the temperature cor-
2
responding to the maximum of the desorption peak (TH O) are
2
obtained.
Solid-state two-dimensional (2D) 13C{1H} and 29Si{1H}
heteronuclear chemical shift correlation (HETCOR) NMR ex-
periments were performed on a Bruker AVANCE-500 NMR
spectrometer operating at 500.13, 125.76, and 99.35 MHz
for 1H, 13C, and 29Si, respectively. Experiments were con-
ducted at room temperature under magic-angle spinning (MAS)
conditions at 10 kHz by using a Bruker 1H/X double-resonance
MAS probehead with 4.0-mm zirconia rotors. The HETCOR
technique allows different dipole-dipole-coupled moieties to be
differentiated and identified by spreading their chemical shifts
into a 2D frequency map. The HETCOR experiment is similar
to standard cross-polarization magic-angle spinning (CP-MAS)
1
NMR [40], with the key exception that the H magnetization
is allowed to evolve for an incremented evolution time pe-
riod t1 prior to magnetization transfer to heteronuclei, such
as 13C or 29Si, whose responses are measured directly dur-
ing the detection period t2 [41]. Double Fourier transformation
converts the time domain signal S(t1, t2) into the frequency
domain F(ω1, ω2), which is typically presented as a 2D con-
tour plot spectrum. For the 13C{1H} HETCOR experiments,
a 4.2 µs 90◦ pulse, followed by a 3 ms contact time, were
used for cross-polarization. 864 acquisitions with a 2 s recy-
cle delay were collected for 96t1 increments. For the 29Si{1H}
HETCOR experiments, a 6.8 µs 90◦ pulse, followed by a 2 ms
2.2. Catalyst characterization
X-ray powder diffraction (XRD) patterns were acquired on
a PHILIPS X’PERT diffractometer using CuKα radiation. The
data were recorded from 0.6◦ to 5◦ (2θ) with a resolution
of 0.02◦. Nitrogen adsorption and desorption isotherms at 77 K
were measured using a Micromeritics TRISTAR 3000 system.

208
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
Table 1
Sulfonic acid and organic contents of the sulfonic-acid-functionalized mesoporous silica and organosilica catalysts
a
29
Catalyst
Elemental analysis
Sulfur content
Si MAS NMR
Titration
m
Carbon content
(mmol C/g SiO )
T
sites
b
Acid capacity
(mmol H /g SiO )
+
(mmol S/g SiO )
(mol%)
2
2
2
Propyl-SO H silica
1.52 ( 0.12)
1.53 ( 0.18)
1.66 ( 0.08)
1.79 ( 0.21)
6.15 ( 0.35)
15.44 ( 0.65)
16.10 ( 0.53)
25.20 ( 1.08)
11
57
12
56
1.58
1.45
1.71
1.83
3
Propyl-SO H organosilica
3
Arene-SO H silica
3
Arene-SO H organosilica
3
a
Mean values and experimental error obtained from three analysis replications.
Molar percentage of T sites obtained from Si MAS NMR spectra as follows = [integrated area of T signals/(area T signals + area Q signals)]×100.
b
m
29
m
m
n
contact time, were used for cross-polarization. 768 acquisitions
modified PMOs. Likewise, regardless of the type of sulfonic
acid moiety, a gradual increase in the final nitrogen adsorbed
volume at P/P0 > 0.8 is observed in the hybrid PMO materi-
als. This unexpected increase in the adsorbed volume at high
relative pressures might be attributed either to macropores or
to interparticulate porosity, which takes place when the particle
diameters are very small and closely packed. Thus, this effect
might be related to a morphology change in the particles when
introducing BTSE in the synthesis of the sulfonic-acid modified
mesostructured materials (see TEM images in Supplementary
material, Figure SI.1).
with a 2 s recycle delay were collected for 128t1 increments.
1
The H and 29Si chemical shifts were referenced to tetrakis-
(trimethylsilyl)silane [(CH3)3Si]4Si, and the 13C chemical shift
was referenced to adamantane (C10H16). All of the 2D HET-
COR spectra are presented with contour levels shown to 20%
of full intensity.
2.3. Catalytic reaction tests
Etherification experiments were carried out in liquid phase
at 50 ◦C in a stirred Teflon-lined stainless-steel autoclave under
autogeneous pressure (with an initial nitrogen pressure of 4 bar
to ensure that a liquid phase is maintained). Reaction temper-
ature was controlled using a thermocouple immersed into the
reaction mixture. Samples were periodically withdrawn at time
intervals ranging from 0 to 2 h. Typically, weight composition
of the reaction mixture was: 30 g of 1-hexanol, 3 g of vanillyl
alcohol and constant catalyst loading of 0.03 g (0.1 wt%). Re-
action samples were analyzed by GC (Varian 3900 chromato-
graph) using a CP-SIL 8 CB column (30 m × 0.25 mm, film
thickness = 0.25) and a flame ionization detector. Benzyl ether
was the only detected reaction product, and a control blank re-
action gave no conversion of vanillyl alcohol in absence of acid
catalyst. Catalytic results are shown either in terms of absolute
conversion of vanillyl alcohol or in terms of initial rate values
(mmol of reacted alcohol per g of catalyst and minute), calcu-
lated at the beginning of the reaction (15 min). To allow the
comparison of the catalytic activity between different types of
sulfonic acid groups (in different surface micro-environments),
a specific initial rate per acid site has also been utilized (mmol
of reacted alcohol per minute and mmol of acid site after 15 min
of reaction).
Powder X-ray diffraction analyses of propyl- and arene-
sulfonic acid functionalized samples give diffractograms that
are typical signals for hexagonally-ordered mesopores [42]
(Fig. 1). Samples without ethylsiloxane functionalities in the
composition of their pore walls, present the typical small-angle
X-ray scattering (SAXS) reflections associated with p6mm
symmetry. The organically-modified samples (PMOs) have in
common a single prominent SAXS reflection centered at 2θ
below 1◦, assigned to the crystallographic direction [100]. The
signals [110] and [200], indicative of long-range hexagonal or-
dering, are less pronounced than in the silica materials. This
kind of diffraction pattern has been previously associated with
“worm-hole” structures [43,44], and similar to those described
in literature for related organically modified solids [45,46].
Elemental analyses and acid titration studies establish the
incorporation of organic species within the silica framework
and the accessibilities of the sulfonic-acid sites at the meso-
pore surfaces. Comparison of the sulfur concentrations deter-
mined by elemental analysis and the proton concentrations de-
termined by cationic-exchange of the acid sites by sodium ions
(Table 1) reveal that >95% of the acid catalyst sites are ac-
cessible. In addition, the presence of ethylsiloxane moieties in
the silica framework does not appear to affect the incorpora-
tion of sulfonic-acid species into the material. Further analy-
sis of the incorporated sulfonic-acid and organic functionalities
by single-pulse 29Si MAS NMR (Table 1) establishes that the
propyl- and arene-sulfonic-acid species are inserted into the sil-
ica framework to comparable extents. For example, single-pulse
29Si MAS NMR analyses provide quantitative information on
the relative populations of 29Si sites resolved for different silica
and organosilica moieties associated with the mesopore frame-
works. Linewidths of approximately 9 ppm are observed for
29Si signals associated with these materials, consistent with
a large distribution of local 29Si sites in the disordered sil-
3. Results and discussion
3.1. Catalyst structures, compositions, and stabilities
Nitrogen adsorption isotherms measured for the sulfonic-
acid functionalized materials were of Type IV according to
IUPAC classification (Fig. 1). For materials synthesized with-
out BTSE, the isotherms displayed a pronounced hysteresis
loop at relative pressures in the range 0.6–0.8, consistent with
data previously reported [11,13]. The increase in adsorption
in the mesoporous range is less defined in the sulfonic acid-

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
209
Fig. 1. (a) Nitrogen adsorption isotherms at 77 K and (b) small-angle X-ray scattering patterns for sulfonic-acid modified mesostructured silicas. See Supplementary
material, Table SI.1, which summarizes the textural properties of these materials.
ica frameworks. Nevertheless, single-pulse 29Si MAS NMR
spectra reveal resolved signals from 29Si sites with different
extents of condensation and numbers of organic ligands. Quan-
titative intensities are observed and attributed to the following
silica and organosilica 29Si species: Qn = Si(OSi)n(OH)4−n
,
where n = 2–4 (Q2 at ca. −90 ppm, Q3 at ca. −100 ppm, and
Q4 at ca. −110 ppm), and T m = RSi(OSi)m(OH)3−m, where
m = 2–3 (T 3 at ca. −65 ppm and T 2 at ca. −57 ppm). Both the
propyl- and arene-sulfonic-acid functionalized materials pro-
vide ratios of T m signals that are consistent with bulk elemen-
tal analyses (Table 1). Furthermore, the 29Si signal intensities
assigned to T m silicon atoms account not only for ethoxysi-
lane functionalities, but also for the sulfonated species, propyl-
SO3H and arene-SO3H. Combining the number of T m sites
from 29Si MAS NMR (i.e., the number of Si–R moieties) with
carbon elemental analyses (Table 1), the differences in car-
bon concentrations in the propyl- and arene-sulfonic acid func-
tionalized materials are consistent with the different molecular
weights of the functional groups and not differences in their re-
spective loadings.
Scheme 1. Production of 4-hydroxy-3-methoxybenzyl-1-hexyl ether via the
etherification of vanillyl alcohol and 1-hexanol.
species, and (3) above 350 ◦C, from the thermal decomposi-
tion of sulfonic-acid groups and ethylsiloxane moieties within
the silica framework. Higher mass losses are observed in re-
gion 3 for the organosilica materials compared to the silica
materials (19.5 vs 13.2% for the propyl-sulfonic acid materi-
als and 25.5 vs 18.4% for the arene-sulfonic acid materials), in
spite of having a constant sulfonic-acid functionality loading.
The increased mass loss above 350 ◦C is quantitatively consis-
tent with simultaneous decomposition of the ethylsiloxane and
sulfonic-acid moieties in the organosilica materials. Progres-
sive decomposition of the ethylsiloxane moieties over a wide
temperature range is evidence of a distribution of local envi-
ronments around the framework ethylsiloxane functionalities.
Notably, the mass loss assigned to water desorption is lower
by ∼2% in the organosilicas compared to the silica materi-
als. The difference, although small, is indicative of the more
hydrophobic nature in the ethylsiloxane-modified organosilica
materials, even in the presence of highly hydrophilic sulfonic-
acid species.
High thermal stabilities of incorporated sulfonic-acid and or-
ganic framework species are important in acid-functionalized
mesoporous silica and organosilica catalysts. The main ad-
vantage of using the co-condensation method for inclusion
of functional species in mesoporous silicas is the introduc-
tion of thermally-stable covalent Si–C anchoring bonds with
high incorporation yields, in contrast to low incorporation typ-
ically obtained by post-synthetic grafting methods. Thermo-
gravimetric analyses (see Supplementary material, Figure SI.2)
of the acid-functionalized mesoporous catalyst materials typi-
cally show weight loss profiles that can be divided into three
regions: (1) from room temperature to 150 ◦C attributed to
water desorption, (2) between 150 and 350 ◦C, attributed to
remaining surfactant molecules and surface-adsorbed ethoxy
groups from the ethanol-extraction to remove the surfactant
3.2. Catalyst performance
Propyl- and arene-sulfonic acid modified mesostructured
SBA-15 silicas have previously been demonstrated as active
catalysts in the etherification of benzyl alcohols, and for which
the hydrophobicity of the local environment surrounding the
acid sites plays a key role in the catalytic performance [47].

210
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
Table 2
Catalytic activity of propyl- and arene-sulfonic acid modified mesoporous silica
and organosilica catalysts in the etherification of vanillyl alcohol with 1-hexanol
a
b
ꢀ c
Catalyst
Propyl-SO H silica
X
r
r
vanillyl
0
0
35
54
33
51
14.4
22.2
13.6
21.0
11.7
21.2
11.1
16.9
3
Propyl-SO H organosilica
3
Arene-SO H silica
3
Arene-SO H organosilica
3
a
Conversion of vanillyl alcohol at 15 min.
−1
b
−1
Initial reaction rate values [(mmol)(g )(min )] calculated from mmols
cat
of reacted vanillyl alcohol within the first 15 min of reaction per gram of cata-
lyst.
c
Initial specific reaction rate values per acid site [(mmol)(mmol acid-
−1
−1
center )(min )] calculated from mmols of reacted vanillyl alcohol within
the first 15 min of reaction and referred to mmol of acid sites per gram of each
catalyst.
Fig. 2. Etherification of vanillyl alcohol with 1-hexanol over propyl- and arene-
sulfonic-acid modified mesoporous silica and organosilica catalysts. Reaction
◦
conditions: batch reactor at 50 C and 4 bar; 30 g 1-hexanol, 3 g vanillyl alco-
hol, 0.03 g catalyst (0.1 wt%)
.
when using the organosilica catalysts. Normalizing the initial
reaction rates per acid site (r0ꢀ ), the enhancement of catalytic
activity in the organosilicas compared to the silica materials
is even higher (1.8 times higher in the case of propyl-SO3H
organosilica). Incorporation of hydrophobic ethylsiloxane moi-
eties into the silica support framework has a beneficial effect
on the catalyst performance, presumably due to reduced water
poisoning of the acid active sites.
The sulfonic-acid catalyst sites are readily poisoned by water
as reaction by-product. Balancing the hydrophobic-hydrophilic
surface properties of the catalyst allows one to modulate the
reaction conversion rates. Here, the etherification of vanillyl
alcohol (4-hydroxy-3-methoxybenzylalcohol) with 1-hexanol
to yield 4-hydroxy-3-methoxybenzyl-1-hexyl ether (Scheme 1)
has been chosen as a technologically-relevant, acid-catalyzed
reaction to examine the effects of increased hydrophobicity
of the mesoporous organosilica support material on the acid-
catalyst performance.
3.3. Water retention capacities
The comparison of the bulk hydrophilicities of acid-function-
alized mesoporous silica and organosilica catalysts confirms
that incorporation of ethylsiloxane moieties within the silica
frameworks reduces the water retention capacities of the ma-
terial. Temperature-programmed-desorption (TPD) were per-
formed to study the interaction of adsorbed water with the acid-
functionalized surfaces of the mesoporous silica and organosil-
ica materials. The TPD curves for water over sulfonic-acid
modified mesoporous silica and organosilica samples show two
desorption peaks, the first one (100–200 ◦C) corresponding to
the desorption of water interacting strongly with acid sites.
Desorption signal over 200 ◦C would be attributed to thermal
decomposition of SO3H and Si–CH2–CH2–Si moieties, and to
dehydration of surface Si–OH groups. This assumption is valid
since decomposition of acid and ethylene moieties occurs at
higher temperature ranges (observed in TGA data [14]). Fur-
thermore, Shimizu et al. [48] also employed this approximation
for TPD results over sulfonic-acid modified silica samples. It
should be noted that the adsorbed-desorbed amounts of water
shown in Table 3 correspond to dynamic experiments in contin-
uous helium flow with a previous physi-desorption step at 40 ◦C
to remove weakly retained water. Accordingly, the amounts
shown account only for water that is more strongly bounded
to the functionalized surfaces and are significantly smaller than
the adsorption capacity corresponding to the total pore filling
of the mesopores. Thus, the relative amounts of desorbed water
in TPD experiments, shown in Table 3, can be used to compare
the hydrophilicities of the samples.
Comparisons of vanillyl alcohol conversion in batch reac-
tions over propyl- and arene-sulfonic-acid mesoporous silica
and organosilica catalysts establish the beneficial performance
effects of incorporating hydrophobic ethylsiloxane moieties in
the mesoporous silica support. As shown in Fig. 2, both propyl-
and arene-sulfonic-acid species on mesoporous silica display
high conversion rates, reaching conversions of over 77% (arene-
SO3H) and over 95% (propyl-SO3H) after only 2 h of reaction.
Despite their lower acid strength, the propyl-sulfonic acid sites
exhibit higher etherification activity than the arene-sulfonic-
acid groups. This is attributed to lower hydrophilicity of the
alkyl chain, compared to the aromatic ring, which yields lo-
cal environments near the sulfonic-acid active sites that are less
sensitive to the poisoning effect of co-adsorbed water mole-
cules. This is further supported by the activities measured for
the propyl-SO3H and arene-SO3H groups anchored on more
hydrophobic mesoporous organosilica supports.
For both the propyl- and arene-sulfonic acid functionalized
mesoporous organosilicas, clear improvements in their respec-
tive etherification activities are observed compared to otherwise
identical mesoporous silica-supported SO3H catalysts. This
improvement is especially apparent at initial reaction times,
where deactivation effects are minimal and the conditions are
far away from chemical equilibrium. The acid-functionalized
mesoporous organosilica catalysts show higher initial conver-
sions of vanillyl alcohol after 15 min reaction time, compared
to the corresponding acid-functionalized mesoporous silica cat-
alysts (Table 2). Furthermore, assuming linear behavior during
the first 15 min, initial reaction rates are also about 1.5 higher
The sulfonic-acid modified mesoporous organosilicas are
more hydrophobic than the corresponding sulfonic-acid mod-

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
211
Table 3
compositions, structures, and dynamics of framework, func-
tional, and adsorbed species in the mesoporous organosilica cat-
alysts. In particular, 2D 13C{1H} and 29Si{1H} HETCOR mea-
surements provide significantly enhanced spectral resolution
and clear identification of otherwise overlapping 1D 1H signals
associated with the various dipole–dipole-coupled molecular
moieties in acid-functionalized mesoporous materials. For the
arene- or propyl-SO3H-modified catalysts under consideration
here, these include grafted sulfonic acid species, framework
ethylsiloxane moieties, and weakly surface-adsorbed ethoxy
groups from either ethanol washing or noncondensed ethoxysi-
lanes (Figs. 3a, 4a, 5a, 6a). In a 2D HETCOR spectrum, cor-
Temperature-programmed-desorption (TPD) of water in propyl- and arene-
sulfonic-acid modified mesoporous silica and organosilica catalysts
a
W
H O
2
b
Mesoporous catalysts
T
H O
2
◦
+
( C)
(mg/g) (mmol/g) (mmol H O/mmol H
)
2
Silica
Propyl-SO H silica
3
Propyl-SO H organosilica 117
Arene-SO H silica
83
119
0.74 0.04
13.81 0.77
6.77 0.38
14.55 0.81
7.83 0.49
–
0.62
0.36
0.65
0.40
3
139
3
Arene-SO H organosilica 137
3
a
Temperature corresponding to the maximum of the water desorption peak.
Water amount retained and desorbed by the solid in the temperature range
b
◦
of 50–200 C.
1
related H and 13C or 29Si signal intensities can be observed
for directly bonded or spatially proximate (ca. 1 nm) species.
ified mesoporous silicas, although much less compared to
mesoporous silica. As shown in Table 3, the sulfonic-acid
organosilicas adsorb about half the amount of water (expressed
as mg g−1) than the sulfonic-acid silicas, and approximately
10 times that corresponding to pure silica. That is, the water-
retention capacity is clearly increased by incorporating acid
functional species, but reduced when incorporating alkyl moi-
eties within the framework of the mesostructure. Additionally,
the temperature corresponding to the peak of maximum des-
orption remains practically constant when considering catalyst
supports functionalized by the same acid moiety (about 117–
119 ◦C for propyl-sulfonic-acid materials, and 137–139 ◦C for
arene-sulfonic-acid materials), which suggests that the inter-
actions between the –SO3H groups and water molecules does
not depend strongly on the surface hydrophobicity. Compar-
ing the effect of neighboring molecular groups adjacent to
the –SO3H sites, arene-sulfonic-acid materials provide slightly
higher water loadings and retain water to much higher reten-
tion temperatures relative to the propyl-sulfonic acid materials,
in agreement with enhanced acid strengths and resulting hy-
drophilicities.
A consequence of this is that the acid centers may be
subjected to H2O-poisoning in reactions involving non-polar
organic molecules. Hence, the higher catalytic performance
displayed by the SO3H-modified PMO samples might be
attributed to weaker interactions between the sulfonic-acid
groups in these catalysts and water molecules generated dur-
ing reaction (Scheme 1). This is expected to lead to better
accessibilities of alcohol reactant molecules to the sulfonic-acid
active sites. Since the –SO3H moieties are highly hydrophilic,
the hydrophobicity of the nearby acid site environments ex-
ert substantial influence on their activities in water-sensitive
organic reactions. Macroscopically, the water-retention proper-
ties of these materials, as determined by TGA and H2O-TPD
measurements (Table 3), and the catalytic results obtained in
the etherification of vanillyl alcohol with 1-hexanol (Fig. 2) are
consistent with this hypothesis. However, insights and under-
standing of the molecular origins of these macroscopic adsorp-
tion and reaction behaviours remain unknown.
By adjusting the H → 13C or H → 29Si cross-polarization
contact times, it is furthermore possible to distinguish between
strongly and weakly interacting species. By this method, unam-
biguous information is obtained on molecular-level structures
and local compositions in acid-functionalized mesoporous cat-
alysts that account for differences in their macroscopic catalytic
properties.
1
1
3.4.1. Arene-SO3H-functionalized mesoporous silica and
organosilica catalysts
Specifically, 2D HETCOR spectra allow different grafted,
surface-adsorbed, and framework moieties to be identified
and their mutual proximities established. Moreover, differ-
ences in the extents to which arene- and propyl-sulfonic acid-
functionalized mesoporous silica or mesoporous organosilica
materials adsorb water can be understood at a molecular-level
and correlated with their efficacies for catalyzing heteroge-
neous etherification reactions, such as the etherification of
vanillyl alcohol with 1-hexanol. For example, Fig. 3 shows 2D
13C{1H} HETCOR spectra obtained from arene-SO3H acid-
functionalized mesoporous silica (Fig. 3b) and mesoporous
organosilica, (Fig. 3c) with the various proton moieties of the
arene-sulfonic acid groups, surface-adsorbed ethoxy groups,
framework ethylsiloxane groups, and adsorbed water all yield-
ing well resolved signals. The 2D 13C{1H} HETCOR spectrum
in Fig. 3b shows strong correlated signal intensities at 1.2 ppm
in the 1H dimension and at 16 and 28 ppm in the 13C dimension,
which are associated with the alkyl –CH2– groups (labeled ‘1’
and ‘2’ in Fig. 3a) of the arene-SO3H species and the –CH3
groups (labeled ‘5’ in Fig. 3a) of surface ethoxy species, respec-
tively. The –OCH2– (4) groups of the surface ethoxy species
1
give rise to 2D signal intensity at 3.8 ppm in the H dimen-
sion and 58 ppm in the 13C dimension. The aromatic protons of
the arene-SO3H species yield correlated 2D signal intensities at
7.6–8.0 ppm in the 1H dimension and between 124–146 ppm in
the aromatic region of the 13C dimension. In addition to the sig-
1
nal intensities arising from directly bonded H–13C pairs, two
clearly resolved signals in the 2D 13C{1H} HETCOR spec-
1
trum in Fig. 3b at 6.4 ppm in the H dimension and 124 and
126 ppm in the 13C dimension (dotted red lines) are attributed
to adsorbed water molecules that are in close spatial proximity
to the aromatic carbon atoms and –SO3H groups of the arene-
sulfonic acid species.
3.4. Local site compositions and structures
Powerful two-dimensional (2D) nuclear magnetic resonance
(NMR) spectroscopy techniques are especially sensitive to local

212
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
13
1
Fig. 3. (a) Schematic representations of the organic moieties present and identified in room-temperature 2D C{ H} HETCOR spectra of (b) mesoporous silica
13
1
and (c) mesoporous ethylsiloxane functionalized with arene-SO H moieties. Separate C CP-MAS and H MAS spectra are plotted along their corresponding
3
13
1
13
1
axes. The 2D C{ H} HETCOR spectra show strongly correlated C and H signals that permit their unambiguous assignments to grafted arene-sulfonic acid
moieties, surface ethoxy groups, framework ethylsiloxane moieties, and adsorbed water. Note the strong 2D intensity correlation (dotted red lines) between protons
of adsorbed water and the aromatic C nuclei in (b) and its absence in (c), which manifests the different hydrophilicities/hydrophobicities near the acid sites of the
13
silica and organosilica supports.
By comparison, the 2D 13C{1H} HETCOR spectrum for
the arene-sulfonic acid-functionalized organosilica catalyst
(Fig. 3c) shows the same correlations as for the wholly siliceous
support with two notable exceptions. First, the additional cor-
catalyst. It establishes that upon incorporation of ethylsiloxane
species into the mesoporous silica framework, adsorbed water
no longer resides in the local environments near (ca. 1 nm) the
–SO3H catalytic acid sites.
1
related 2D signal intensity at 1.2 ppm in the H dimension
The 2D 29Si{1H} HETCOR spectra acquired for the same
arene-sulfonic acid-functionalized silica and organosilica cat-
alysts confirm the covalent grafting of the acid moieties and
corroborate the increased hydrophobicity of the mesoporous
organosilica support. As shown in Fig. 4 for both samples, 2D
29Si{1H} intensity correlations are clearly resolved between the
alkyl protons of the arene-SO3H species and the protons of ad-
sorbed water molecules with the various 29Si sites in the mate-
rials. For example, in Fig. 4b, the 29Si{1H} HETCOR spectrum
acquired for the arene-sulfonic acid functionalized silica (with-
out ethylsiloxane incorporated) shows 2D signal intensity at
and 6 ppm in the 13C dimension arises, due to the presence of
–CH2– groups (‘3’ in Fig. 3a) associated with the ethylsiloxane
species incorporated into the organosilica framework. Further-
1
more, while the single-pulse H NMR spectrum shown along
1
the H dimension of the 2D contour plot contains an intense
signal at 6.9 ppm corresponding to the adsorbed water, there
is no 2D signal correlated to any of the 13C species, including
the aromatic carbon atoms adjacent to the –SO3H acid sites of
the arene-sulfonic acid species. This is an important difference
from the arene-sulfonic acid functionalized mesoporous silica

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
213
Fig. 4. (a) Schematic representations of the various organic moieties present in arene-SO H-functionalized silica catalyst materials, with room-temperature 2D
3
29
1
29
1
Si{ H} HETCOR spectra acquired for (b) mesoporous silica and (c) mesoporous organosilica (ethylsiloxane) supports. Separate one-pulse Si and H MAS
spectra are plotted along their corresponding axes. All measurements were conducted on the same samples as used in Fig. 3. The Si{ H} HETCOR spectra show
strongly correlated signal intensities between Si Q and Q framework sites and H nuclei of the grafted sulfonic acid moieties and/or adsorbed water.
29
1
29
3
4
1
1.2 ppm in the 1H dimension and −65 ppm in the 29Si dimen-
sion. Such correlated intensity arises from dipole–dipole cou-
plings between the –CH2– (1, 2) protons of the arene-sulfonic
acid moieties and cross-linked silica T 3 29Si sites associated
with the silica framework (T n refers to four-coordinate silica
units with n (an integer ꢀ3) silicon next-nearest neighbors,
one covalent Si–C bond associated with an organic moiety, and
(3−n) uncondensed Si–O bonds, e.g., hydroxyl species). In ad-
dition, strong 2D signal intensity is observed at 6.4 ppm in the
1H dimension and −102 ppm in the 29Si dimension, reflecting
strong interactions between protons of adsorbed water mole-
cules and partially polymerized Q3 29Si sites. (Qn refers to
four-coordinate silica units with n (an integer ꢀ 4) silicon next-
nearest neighbors and (4 − n) uncondensed Si–O bonds, e.g.,
hydroxyl species). The T 3 29Si signal at −65 ppm in Fig. 4b is
also correlated with that of the 1H nuclei of the adsorbed water
molecules in the wholly siliceous silica support (6.4 ppm).
Similarly, for the arene-sulfonic acid functionalized organo-
silica, 2D 29Si{1H} HETCOR results confirm acid-group graft-
ing and increased hydrophobicity of the mesoporous support.
The 2D 29Si{1H} HETCOR spectrum in Fig. 4c shows cor-
1
related signal intensity at 6.8 ppm in the H dimension and
−100 and −110 ppm in the 29Si dimension (dotted red lines),
corresponding to interactions between adsorbed water protons
and both the Q3 and spatially proximate, and presumably sur-
face, Q4 29Si sites in the framework. In addition, correlated
intensity at 1.2 ppm in the 1H dimension and −65 ppm in
the 29Si dimension is also observed, corresponding to interac-
tions between the –CH2– (1, 2) protons of the arene-sulfonic
acid moieties and the T 3 29Si framework sites; a weak inten-
sity correlation is also observed at −55 ppm corresponding to
incompletely cross-linked T 2 29Si sites. As expected for the
organosilica support, higher populations of T 2 and T 3 29Si sites
are present, relative to the Q3 and Q4 species and compared to

214
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
13
1
Fig. 5. (a) Schematic representations of the organic moieties present and identified in room-temperature 2D C{ H} HETCOR MAS spectra of (b) mesoporous
13
1
silica and (c) mesoporous organosilica functionalized with propyl-SO H acid moieties. Separate C CP-MAS and H MAS spectra are plotted along their corre-
3
13
1
13
1
sponding axes. The 2D C{ H} HETCOR spectra show similarly strongly correlated C and H signals from grafted propyl-sulfonic acid moieties, surface and
framework organic groups, and adsorbed water. The strong 2D intensity correlation (dotted red lines) observed in (b) between protons of adsorbed water and the
propyl C nuclei is absent in (c), manifesting the very different hydrophilic/hydrophobic environments near the acid sites of the silica and organosilica supports.
13
the mesoporous silica catalyst, as established by the different
integrated signal intensities in the one-pulse 29Si MAS spec-
tra in Figs. 4b and 4c. The notable absence of any 2D signal
eties of the remaining surfactant yield weak 2D signal intensity
1
at 3.8 ppm in the H dimension and 71 ppm in the 13C di-
mension. Notably, weak 2D correlated signal intensity is also
1
1
intensity at 6.8 ppm in the H dimension and −65 ppm in the
observed at 6.5 ppm in the H dimension and 56 ppm in the
29Si dimension establishes that adsorbed water molecules do
not reside near (ca. 1 nm) a significant fraction of the arene-
SO3H moieties in the functionalized organosilica catalyst.
13C dimension, corresponding to interactions between the pro-
tons of adsorbed water molecules near the propyl-SO3H groups
and 13C nuclei of the nearest –CH2– (1) moieties.
By comparison, the 2D 13C{1H} HETCOR spectrum in
Fig. 5c for the propyl-SO3H-functionalized organosilica sup-
port shows signal intensities corresponding to the same cor-
relations as in Fig. 5b for the propyl-SO3H-functionalized sil-
ica material, but with two important differences. An additional
2D signal is present in Fig. 5c at 1.1 ppm in the 1H dimension
and 6 ppm in the 13C dimension, corresponding to the –CH2–
(4) groups of the incorporated framework ethylsiloxane species.
Importantly, in addition, there is no 2D signal intensity in the
3.4.2. Propyl-SO3H-functionalized mesoporous silica and
organosilica catalysts
Similar corroborative results and conclusions are obtained
from the propyl-sulfonic acid-functionalized silica and organo-
silica catalysts. The 2D 13C{1H} HETCOR spectra shown in
Figs. 5b and 5c allow the various overlapping signals in the 1D
1
single-pulse H spectra from the propyl-sulfonic acid, surface
ethoxy, and ethylsiloxane species (Fig. 5a) to be clearly re-
solved. For example in Fig. 5b, 2D signal intensities from each
of the different –CH2– (1, 2, 3) moieties of the propyl-sulfonic
acid species at 1.1 ppm in the 1H dimension are clearly resolved
at 56, 18, and 4 ppm in the 13C dimension. The –OCH2– moi-
1
spectrum near 7.0 ppm in the H dimension and 56 ppm in
the 13C dimension, establishing that no detectable interactions
are measured between adsorbed water and the 13C nuclei ad-
jacent to the propyl-sulfonic acid sites. This corroborates at a

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
215
Fig. 6. (a) Schematic representations of the various organic moieties present in propyl-SO H-functionalized silica catalyst materials, with room-temperature 2D
3
29
1
29
1
Si{ H} HETCOR spectra for (b) mesoporous silica and (c) mesoporous organosilica (ethylsiloxane) supports. Separate one-pulse Si and H MAS spectra are
plotted along their corresponding axes. All measurements were conducted on the same samples as used in Fig. 5. The Si{ H} HETCOR spectra show strongly
correlated signal intensities between Si Q and Q framework sites and H nuclei of the grafted SO H moieties and/or adsorbed water.
29
1
29
3
4
1
3
molecular level the greater hydrophobicity of the organosilica-
supported propyl-SO3H catalysts.
alysts confirm the covalent grafting of the acid moieties and
corroborate the increased hydrophobicity of the mesoporous
organosilica support. As shown in Fig. 6 for both the silica
and organosilica supports, 2D 29Si{1H} HETCOR spectra show
correlated signal intensity between protons associated with the
alkyl –CH2– (1, 2, 3) groups or adsorbed water and various
29Si sites in the materials. For example, in Fig. 6b, the 2D
29Si{1H} HETCOR spectrum acquired for the wholly siliceous
mesoporous support shows strong correlated signal intensities
at 6.5 ppm in the 1H dimension and −100 and −110 ppm in the
29Si dimension (dotted red lines), corresponding to strong inter-
actions between adsorbed water molecules and the Q3 and Q4
29Si sites, respectively, in the silica framework. In addition,
the T 3 29Si sites share strongly correlated 2D signal intensity
at −67 ppm with the alkyl protons of the propyl-sulfonic acid
species at 1.1 ppm.
Likewise, comparing spectra in Fig. 5 (propyl-SO3H cata-
lysts) with those in Fig. 3 (arene-SO3H catalysts), higher inten-
sity in the correlation between water protons (signal in the spec-
1
tra near 7.0 ppm in the H dimension) and the corresponding
13C nuclei next to the sulfonic acid sites (aromatic 13C nuclei
at 124–146 ppm for arene-SO3H catalysts and alkyl 13C nuclei
at 54 ppm for propyl-SO3H catalysts in the 13C dimension) is
clearly evidenced, especially for the silica materials (spectra de-
noted as b). This confirms the macroscopic information above
presented from catalytic and TPD data, where the environment
in propyl-SO3H sites is more hydrophobic than in arene-SO3H
moieties.
The 2D 29Si{1H} HETCOR spectra acquired for the same
propyl-sulfonic acid-functionalized silica and organosilica cat-

216
G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
Similarly, for the propyl-sulfonic acid-functionalized organo-
vironments near the propyl-sulfonic acid sites result in reduced
poisoning by adsorbed water. The relationships between the
macroscopic catalytic behaviors and molecular surface proper-
ties of the materials are expected to be generally applicable to
other water-sensitive reactions, where hydrophilic/hydrophobic
interactions among reacting species and active/adsorption sites
must be balanced.
silica, 2D 29Si{1H} HETCOR results confirm acid-group graft-
ing and increased hydrophobicity of the mesoporous support.
The 2D 29Si{1H} HETCOR spectrum in Fig. 6c shows the
same correlated signals associated with water-Q3/Q4 interac-
tions (dotted red lines) as for the wholly siliceous mesoporous
support, although with weaker intensities. This is consistent
with the more hydrophobic character of the ethylsiloxane-
incorporated organosilica support. In addition, a weak corre-
lation is also observed between the –CH2– (1, 2, 3) protons
(1.2 ppm) of the propyl-sulfonic acid moieties with the in-
completely cross-linked T 2 29Si sites (−55 ppm). Finally, as in
Fig. 6b, there is no 2D signal intensity arising from interactions
between the acid grafting sites (T n 29Si species) and adsorbed
water protons. These results confirm the increased hydropho-
bicity of mesopore surfaces in the propyl-SO3H-functionalized
organosilica mesoporous catalyst.
Acknowledgments
Spanish co-authors thank the research project CICYT (CTQ-
2005-02375) for financial support. The UCSB co-authors’ con-
tributions were supported in part by the U.S. Department of
Energy, Office of Basic Energy Sciences, Catalysis Sciences
Grant No. DE-FG02-03ER15467 and by MURI program of
the USARO under grant No. DAAD 19-03-1-0121. The 2D
NMR measurements made use of the Central Facilities of the
UCSB Materials Research Laboratory supported by the MR-
SEC Program of the NSF under award No. DMR 05-20415.
B.F.C. thanks the Spanish Ministerio de Educación y Cultura
for fellowship support during a recent sabbatical visit.
As discussed before for the 13C{1H} HETCOR spectra,
comparing 29Si{1H} HETCOR spectra in Fig. 4b (propyl-
SO3H silica) and Fig. 6b (arene-SO3H silica), a clear difference
in the correlation between water protons (signal in the spec-
tra near 7.0 ppm in the 1H dimension) and the T 3 29Si species
(at −67 ppm in the 29Si dimension) is deduced. While in the
case of arene-SO3H silica such a correlation is evident, for
the propyl-SO3H no interaction can be observed. Again, this
confirms that the environment in propyl-SO3H sites is more hy-
drophobic than in arene-SO3H moieties.
Supplementary material
The online version of this article contains additional supple-
mentary material.
Please visit DOI: 10.1016/j.jcat.2007.12.011.
4. Conclusions
References
The incorporation of hydrophobic ethylsiloxane units within
the frameworks of propyl- and arene-SO3H-modified meso-
porous silica materials have been shown to provide enhanced
catalytic activities for the etherification of vanillyl alcohol
and 1-hexanol. Water produced as a by-product of the re-
action coadsorbs near the sulfonic acid centers, resulting in
their partial deactivation due to competition with the alco-
hol reactant species. The 2D 13C{1H} and 29Si{1H} HET-
COR NMR results establish unambiguously that the incor-
poration of ethylsiloxane moieties into the silica framework
of mesoporous acid catalysts introduces important hydropho-
bicity locally around the sulfonic acid catalyst sites. In both
the arene- and propyl-SO3H-functionalized catalysts, the 2D
13C{1H} HETCOR spectra prove that the hydrophobic eth-
ylsiloxane groups in the organosilica framework lead to re-
duced interactions of adsorbed water with the sulfonic acid
catalyst sites. The 2D 29Si{1H} HETCOR results, furthermore,
establish that interactions between adsorbed water molecules
and 29Si sites on the interior mesopore surfaces are also di-
minished upon the incorporation of ethylsiloxane groups in
the silica framework. These results are in agreement with the
lower water-retention capacities, as determined by means of
TPD and TGA, and higher etherification activities of SO3H-
functionalized mesoporous organosilicas, compared to wholly
siliceous mesoporous supports. Despite their lower intrinsic
acid strengths, propyl-SO3H acid sites yield higher catalytic
activities than arene-SO3H groups; more hydrophobic local en-
[1] T. Okuhara, Chem. Rev. 102 (2002) 3641.
[2] J.H. Clark, D.J. Macquarrie, Handbook of Green Chemistry and Technol-
ogy, Blackwell, Oxford, 2002.
[3] M.A. Harmer, W.E. Farneth, Q. Sun, Adv. Mater. 10 (1998) 1255.
[4] K. Wilson, J.H. Clark, Pure Appl. Chem. 72 (2000) 1313–1319.
[5] A. Vinu, K.Z. Hossain, K. Ariga, J. Nanosci. Nanotechnol. 5 (2005) 347.
[6] G. Oye, J. Sjoblom, M. Stocker, Adv. Colloid Interface Sci. 89–90 (2001)
439.
[7] A. Corma, Chem. Rev. 97 (1997) 2373.
[8] W. Van Rhijn, D. De Vos, W. Bossaert, J. Bullen, B. Wouters, P. Grobet,
P.A. Jacobs, Stud. Surf. Sci. Catal. 117 (1998) 183.
[9] W.M. Van Rhijn, D.E. De Vos, B.F. Sels, W.D. Bossaert, P.A. Jacobs,
Chem. Commun. (1998) 317.
[10] M.H. Lim, C.F. Blanford, A. Stein, Chem. Mater. 10 (1998) 467.
[11] D. Margolese, J.A. Melero, S.C. Christiansen, B.F. Chmelka, G.D. Stucky,
Chem. Mater. 12 (2000) 2448.
[12] I. Díaz, F. Mohino, J. Pérez-Pariente, E. Sastre, Appl. Catal. A Gen. 205
(2001) 19.
[13] J.A. Melero, G.D. Stucky, R. van Grieken, G. Morales, J. Mater. Chem. 12
(2002) 1664.
[14] J.A. Melero, R. van Grieken, G. Morales, Chem. Rev. 106 (2006) 3790.
[15] G.L. Athens, Y. Ein-Eli, B.F. Chmelka, Adv. Mater. 19 (18) (2007) 2580.
[16] M. Alvaro, A. Corma, D. Das, V. Fornés, H. García, Chem. Commun.
(2004) 956.
[17] D.J. Macquarrie, S.J. Tavener, M.A. Harmer, Chem. Commun. (2005)
2363.
[18] I. Díaz, C. Márquez-Alvárez, F. Mohino, J. Pérez-Pariente, E. Sastre,
J. Catal. 193 (2000) 283.
[19] W.D. Bossaert, D.E. De Vos, W.M. Van Rhijn, J. Bullen, P.J. Grobet, P.A.
Jacobs, J. Catal. 182 (1999) 156.
[20] I. Díaz, C. Márquez-Alvárez, F. Mohino, J. Pérez-Pariente, E. Sastre,
J. Catal. 193 (2000) 295.

G. Morales et al. / Journal of Catalysis 254 (2008) 205–217
217
[21] J.A. Melero, R. van Grieken, G. Morales, M. Paniagua, Energy Fuels 21
(2007) 1782.
[36] B. Show, S. Hamoudi, M.H. Zahedi-Niaki, S. Kaliaguine, Micropor.
Mesopor. Mater. 79 (2005) 129.
[22] D. Das, J.F. Lee, S. Cheng, J. Catal. 223 (2004) 152.
[23] K. Wilson, A.F. Fee, D.J. Macquarrie, J.H. Clark, Appl. Catal. A Gen. 228
(2002) 127.
[24] J.G.C. Shen, R.G. Herman, K. Klier, J. Phys. Chem. 106 (2002) 9975.
[25] R. van Grieken, J.A. Melero, G. Morales, Appl. Catal. A Gen. 289 (2005)
143.
[37] P.L. Dhepe, M. Ohashi, S. Inagaki, M. Ichikawa, A. Fukuoka, Catal. Lett.
102 (3–4) (2005) 163.
[38] J.D. Epping, B.F. Chmelka, Curr. Opin. Colloid Interface 11 (2006)
81.
[39] J. Trebosc, J.W. Wiench, S. Huh, V.S.Y. Lin, M. Pruski, J. Am. Chem. Soc.
127 (2005) 7587.
[26] X. Wang, C.C. Che, S.Y. Chen, Y. Mou, S. Cheng, Appl. Catal. 281 (2005)
47.
[40] M.J. Duer, Introduction to Solid-State NMR Spectroscopy, Blackwell
Publishing, Oxford, 2004.
[27] B. Rác, A. Molnár, P. Forgo, M. Mohai, I. Bertoti, J. Mol. Catal. A Chem.
244 (2006) 46.
[28] J.A. Melero, R. van Grieken, G. Morales, V. Nuño, Catal. Commun. 5
(2004) 131.
[29] I.K. Mbaraka, B.H. Shanks, J. Am. Oil Chem. Soc. 83 (2006) 79.
[30] J.A. Bootsma, B.H. Shanks, Appl. Catal. A Gen. 327 (2007) 44.
[31] Q. Yang, M.P. Kapoor, S. Inagaki, J. Am. Chem. Soc. 124 (2002) 9694.
[32] M.P. Kapoor, Q. Yang, Y. Goto, S. Inagaki, Chem. Lett. 32 (2003) 914.
[33] S. Hamoudi, S. Royer, S. Kaliaguine, Micropor. Mesopor. Mater. 71
(2004) 17.
[41] R.R. Ernst, G. Bodenhausen, A. Wokaun, Principles of Nuclear Magnetic
Resonance in One and Two Dimensions, Oxford Univ. Press, New York,
1997.
[42] D. Zhao, Q. Huo, J. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc.
120 (1998) 6024.
[43] P.T. Tanev, T.J. Pinnavaia, Science 267 (1999) 865.
[44] S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science 269 (1999) 1242.
[45] O. Muth, C. Schellbach, M. Fröba, Chem. Commun. (2001) 2032.
[46] H. Zhu, D.J. Jones, J. Zajac, J. Roziere, R. Dutartre, Chem. Commun.
(2001) 2568.
[34] X. Yuan, H.I. Lee, J.W. Kim, J.E. Yie, J.M. Kim, Chem. Lett. 32 (2003)
650.
[47] R. van Grieken, J.A. Melero, G. Morales, J. Mol. Catal. A Chem. 256
(2006) 29.
[35] Q. Yang, M.P. Kapoor, N. Shirokura, M. Ohashi, S. Inagaki, J.N. Kondo,
K.J. Domen, J. Mater. Chem. 15 (2005) 666.
[48] K. Shimizu, E. Hayashi, T. Hatamachi, T. Kodama, T. Higuchi, A. Sat-
suma, Y. Kitayama, J. Catal. 231 (2005) 131.