Synthesis and characterization of triflic acid-functionalized mesoporous Zr-TMS catalysts: Heterogenization of CF3SO3H over Zr-TMS and its catalytic activity

Article abstract of DOI:10.1016/S0021-9517(03)00282-3

Triflic acid-functionalized Zr-TMS (zirconium oxide with a mesostructured framework; TMS, transition metal oxide mesoporous molecular sieves) catalysts have been synthesized by functionalizing triflic acid onto the walls of Zr-TMS via post synthesis method. The synthesized materials were characterized by powder XRD, N₂-sorption, FT-IR spectroscopy, elemental analysis, solid-state ¹³C CP and DD/MAS NMR spectroscopy, FT-Raman analysis, NH₃-TPD, SEM, TEM, TG-DTA, and DTG techniques. All these results revealed that an ordered Zr-TMS material was synthesized and triflic acid was anchored on the walls of the Zr-TMS. Typical XRD patterns of the Zr-TMS and functionalized Zr-TMS (f-Zr-TMS) showed ordered structures. Synthesized materials showed type IV isotherms. The chemical shift observed (≈119 ppm) and ¹³C-¹⁹F coupling (J_C-F≈310 Hz) by ¹³C DD/MAS NMR showed that the triflic acid was intact on the catalyst framework. According to Raman spectral analysis, triflate was adsorbed on the zirconia surface at all loadings as a tridentate ligand through three equivalent S-O bonds (local C_3v symmetry). Ammonia TPD measurements revealed an increase in number of acid sites with an increase in loading of triflic acid. Functionalized amorphous ≡Zr-O-SO₂-CF ₃ catalysts were also synthesized by an in situ method and SO ₄^2-/ZrO₂ was obtained for comparison. The catalytic activity of the materials was tested in the acetalization of ethylacetoacetate and in the benzoylation of biphenyl in a batch reactor at 100 and 150°C, respectively. Recycling was performed in the acetalization of ethylacetoacetate using f-Zr-TMS-30 three times and no major deactivation of the catalyst was observed.

Full text of DOI:10.1016/S0021-9517(03)00282-3

Journal of Catalysis 220 (2003) 442–456
www.elsevier.com/locate/jcat
Synthesis and characterization of triflic acid-functionalized mesoporous
Zr-TMS catalysts: heterogenization of CF SO H over Zr-TMS
3
3
and its catalytic activity
a
b
a,∗
b
M. Chidambaram, D. Curulla-Ferre, A.P. Singh, and B.G. Anderson
a
Catalysis Division, National Chemical Laboratory, Pune 411008, India
b
Schuit Institute of Catalysis, Eindhoven University of Technology, PO Box 513, 5600 MB Eindhoven, The Netherlands
Received 1 April 2003; revised 18 June 2003; accepted 18 June 2003
Abstract
Triflic acid-functionalized Zr-TMS (zirconium oxide with a mesostructured framework; TMS, transition metal oxide mesoporous molecu-
lar sieves) catalysts have been synthesized by functionalizing triflic acid onto the walls of Zr-TMS via post synthesis method. The synthesized
1
3
materials were characterized by powder XRD, N -sorption, FT-IR spectroscopy, elemental analysis, solid-state C CP and DD/MAS NMR
2
spectroscopy, FT-Raman analysis, NH -TPD, SEM, TEM, TG-DTA, and DTG techniques. All these results revealed that an ordered Zr-TMS
3
material was synthesized and triflic acid was anchored on the walls of the Zr-TMS. Typical XRD patterns of the Zr-TMS and functionalized
Zr-TMS (f-Zr-TMS) showed ordered structures. Synthesized materials showed type IV isotherms. The chemical shift observed (≈ 119 ppm)
1
3
19
13
and C– F coupling (J
≈ 310 Hz) by C DD/MAS NMR showed that the triflic acid was intact on the catalyst framework. According
C–F
to Raman spectral analysis, triflate was adsorbed on the zirconia surface at all loadings as a tridentate ligand through three equivalent S–O
bonds (local C symmetry). Ammonia TPD measurements revealed an increase in number of acid sites with an increase in loading of triflic
3
v
2
−
acid. Functionalized amorphous ≡Zr–O–SO –CF catalysts were also synthesized by an in situ method and SO
/ZrO was obtained for
2
3
4
2
comparison. The catalytic activity of the materials was tested in the acetalization of ethylacetoacetate and in the benzoylation of biphenyl in
◦
a batch reactor at 100 and 150 C, respectively. Recycling was performed in the acetalization of ethylacetoacetate using f-Zr-TMS-30 three
times and no major deactivation of the catalyst was observed.

2003 Elsevier Inc. All rights reserved.
Keywords: f-Zr-TMS; Triflic acid; Functionalization; Acetalization of ethylacetoacetate; Benzoylation of biphenyl
1
. Introduction
by Sayari and Liu [9], although a large number of elements
are able to form such materials, only few of these exhibit
ordered porous structures.
Synthesis of highly acidic inorganic materials by heterog-
enization of homogeneous systems is currently the subject of
a great deal of research in green chemistry that aims to facil-
itate the recovery of the catalyst and to minimize the pollu-
tion. Many acidic catalysts have been developed using silica,
alumina, metal oxides, and microporous zeolites as supports
Transition metal oxides are widely used as industrial cat-
alysts and as catalyst supports. Unfortunately they usually
have poorly defined pore structures. The synthesis of meso-
porous silica partially substituted by titanium and zirconium
has been attempted to circumvent the difficulty of preparing
stable mesoporous titania and zirconia. Zirconium oxide is
of particular interest because it contains both acidic and ba-
sic surface sites. These are critical for some reactions that
need bifunctional (amphoteric) catalysts. The search for zir-
conium oxide as supports for various catalysts has received
[
1–5]. However, there are still many problems because of
limited acidity and diffusion. Soon after the discovery of the
MCM-41 materials, it was demonstrated that it is also possi-
ble to synthesize non-silica-based mesostructured materials.
Niobium-, hafnium-, or cerium-based materials are now fre-
quently cited [6–8]. However, as noted in a recent review
2−
keen interest in the past decade. In recent years SO4 /ZrO2
has attracted attention as it catalyzes various industrially im-
portant reactions such as isomerization, condensation, and
Friedel–Crafts acylation reactions [10]. However, its nonuni-
*
Corresponding author.
E-mail address: apsingh@cata.ncl.res.in (A.P. Singh).
0021-9517/$ – see front matter  2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0021-9517(03)00282-3

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
443
form pore size, low porosity, and small surface area limit its
potential application for catalyzing reactions of bulky mole-
cules. Despite these limitations zirconia has a high melting
point, low thermal conductivity, high corrosion resistance,
and amphoteric behavior, all of which can be useful proper-
ties for a support material. The possibility of obtaining such
material with a mesoporous texture has made this oxide even
more interesting [11].
The recent discoveries of ordered, high surface area,
porous zirconia in the presence of cationic surfactants by
Knowles and Hudson [12,13], mesoporous zirconia from
anionic and neutral surfactants by Pacheco et al. [14], nano-
structured mesoporous zirconia by Blin et al. [15], meso-
porous sulfated zirconium dioxide by Huang et al. [16],
surfactant-assisted synthesis of mesoporous zirconia by
Larsen et al. [17], and aluminum-doped mesoporous zirco-
nia by Zhao et al. [18] are attracting increased interest in the
use of zirconia as a catalyst support. Great effort has been
made to prepare mesoporous zirconium hydroxide with high
surface area via a surfactant templating route.
mined by different techniques. Since the catalysts are highly
acidic, acetalization of ethylacetoacetate (EAA) to ethyl
3,3-ethylenedioxybutyrate (fructone) and benzoylation of
biphenyl (BP) to 4-phenylbenzophenone (4-PBP) were car-
ried out to check the catalytic activity of the synthesized
catalysts. Both reactions were also carried out by using func-
tionalized amorphous ≡Zr–O–SO2–CF3 catalysts, Zr-TMS,
2
−
SO4 /ZrO2, and triflic acid for comparison. Fructone, an
apple scent has been used as fragrance, in detergents, in
pharmaceutical industries, as food and beverage additives,
and also in cosmetics [29]. The 4-PBP is an important inter-
mediate in the pharmaceutical industry to synthesize an an-
tifungal agent (bifonazole) and as a photo-initiator [30,31].
2. Experimental
2.1. Materials
The syntheses of f-Zr-TMS catalysts were carried out us-
ing zirconium (IV) butoxide (80 wt% solution in 1-butanol,
Aldrich, USA), a 25 wt% aqueous solution of tetramethyl-
ammonium hydroxide (TMAOH, Loba Chemie, India),
a 25 wt% aqueous solution of N-cetyl-N,N,N trimethyl-
ammonium bromide (CTMABr, Loba Chemie, India), and
trifluoromethanesulfonic acid (CF3SO3H, Lancaster, UK).
“
Triflic acid (trifluoromethanesulfonic acid, CF3SO3H)
is known to be a ‘strong’ acid.” It is widely used in many
organic reactions such as Koch carbonylations [19], Friedel–
Crafts reactions [20], polymerizations [20], etc. [21,22].
However, the recovery of the triflic acid from the reaction
mixture results in the formation of large amounts of waste,
which is environmentally unacceptable. The desire to per-
form acid-catalyzed reactions over solids has resulted in
new research in this area and supported triflic acid is now
becoming available to replace homogeneous acid solutions
2.2. Synthesis of Zr-TMS
The synthesis of Zr-TMS was carried out using the fol-
lowing gel composition and procedure:0.07 Zr(OC4H9)4:1.4
BuOH:0.02 CTMABr:0.014 TMAOH:1.7 H2O.
[
23–25].
Trifluoromethanesulfonic acid and its conjugate base
have extremely high thermal stability and resistance to both
reductive and oxidative cleavage. They do not provide a
source of fluoride ions even in the presence of strong nu-
cleophiles. In addition, the use of triflates or triflic acid as
homogeneous acid catalysts has received attention because
of the electron-withdrawing effect exerted by the trifluo-
romethanesulfonyl group. Recently, a silica-supported poly-
trifluoromethanesulfosiloxane solid superacid [26], a silica-
embedded triflate [27], and metal-triflate [28] were reported
to be heterogeneous catalysts for reactions of bulkier mole-
cules. Hence, we thought that the immobilization of triflic
acid over Zr-TMS would develop a novel class of heteroge-
neous solid acid catalysts with enhanced acidity. Moreover,
the use of a heterogeneous CF3SO3H system would of-
fer ease of catalyst recovery and reuse, and minimize the
production of waste currently formed during triflic acid re-
covery.
A mixture of zirconium (IV) butoxide (80 wt% solution
in 1-butanol) and 1-butanol was stirred for 10 min. Then
the required amount of water was added dropwise into this
mixture under stirring to hydrolyze the zirconium (IV) bu-
toxide to Zr(OH)4. The precipitated Zr(OH)4 mixture was
added to an aqueous solution of CTMABr (25 wt% aq so-
lution) and TMAOH (25 wt% aq solution) under continuous
stirring. After further stirring for 2 h, the resulting synthe-
sis gel (pH = 10.5–11.0) was transferred to a round bottom
◦
flask, sealed, and refluxed at 90 C for 48 h under stirring.
The solid product was recovered by filtration, washed with
◦
deionized water and acetone, and dried at 100 C for 2 h.
The surfactant was removed from the synthesized material
by extraction with a mixture containing 100 g of ethanol and
2.5 g of HCl (36 wt%) per gram of the solid material [20,32]
under reflux for 48 h. The Zr-TMS was washed with water
◦
and acetone and dried at 100 C for 6 h.
In this paper, we disclose the first preparation of tri-
flic acid functionalized Zr-TMS (f-Zr-TMS, Zr-TMS–O–
SO2–CF3, mesoporous ≡Zr–O–SO2–CF3) materials using
Zr-TMS as solid support and triflic acid as an acidic com-
ponent (added by post synthesis). Sol–gel method was
employed for the synthesis of Zr-TMS. The environment
of the triflic acid on the walls of Zr-TMS was deter-
2.3. Synthesis of f-Zr-TMS catalysts
The resulting solid mesoporous material, Zr-TMS, was
functionalized with triflic acid by post synthesis proce-
dure using the molar composition 0.07 Zr-TMS:0.7 dry
toluene:0.03 CF3SO3H.

444
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
1
1
3
3
C CP/MAS (cross-polarization/magic-angle spinning) and
C DD/MAS (dipolar decoupled/magic angle spinning)
NMR for the confirmation of functional group (–CF3) in
the catalyst, FT-Raman analysis for the binding mode of
the triflate, temperature-programmed desorption (TPD) of
NH3 measurement for total acidity, scanning electron mi-
croscopy (SEM) for the particle size and morphology, trans-
mission electron microscopy (TEM) to view the crystalline
structure, and thermogravimetric-differential thermal analy-
sis (TG-DTA and DTG) to study the decomposition and ther-
mal stability of the catalysts.
Scheme 1.
Triflic acid (0.03 mol) was added dropwise into the mix-
ture of toluene and Zr-TMS at 90 C under nitrogen at-
mosphere and further refluxed for 2 h. Then the mixture was
cooled, filtered, washed with acetone, and dried at 100 C
for 6 h. Soxhlet extraction of the material (to remove the un-
reacted triflic acid) was carried out at 75 C for 24 h using a
mixture of dichloromethane (100 g) and diethyl ether (100 g)
per gram of catalyst. The sample was then dried at 200 C
for 10 h. The synthesis of the catalyst is shown in Scheme 1.
The syntheses of different loadings of triflic acid (5 to 25
wt%) on Zr-TMS were carried out by varying the molar ra-
tios of zirconium (IV) butoxide, CTMABr, TMAOH, water,
and CF3SO3H.
◦
◦
◦
The powder X-ray diffraction patterns of synthesized Zr-
TMS, f-Zr-TMS, and ≡Zr–O–SO2–CF3 (A) were recorded
on a Rigaku D MAX III VC (Ni-filtered Cu-Kα radiation,
◦
◦
λ = 1.5404 Å) between 1.5 and 60 (2θ) with a scan rate
◦
of 4 /min. The BET surface area, total pore volume, and
average pore diameter were measured by N2 adsorption–
desorption method by NOVA 1200 (Quantachrome) at
◦
−
196 C. For this particular measurement, the samples were
◦
activated at 180 C for 3 h under vacuum and then the
adsorption–desorption was conducted by passing nitrogen
onto the sample, which was kept under liquid nitrogen. The
2
≡
.4. Synthesis of functionalized amorphous
Zr–O–SO2–CF3
−1
FT-IR spectra were obtained in a range of 400 to 4000 cm
Synthesis of functionalized amorphous ≡Zr–O–SO2–
on a Shimadzu FTIR 8201 PC using a diffuse reflectance
scanning disk technique. Elemental analysis for C and S
were done by an EA1108 elemental analyser (Carlo Erba
CF3 catalysts was carried out via the following procedure.
A mixture of zirconium (IV) butoxide (0.07 mol) and 1-
butanol (1.4 mol) was placed in a three-necked 250-ml
round-bottom flask equipped with a magnetic stirrer and a
Instruments).
13C CP/MAS and 13C DD/MAS solid-state NMR studies
◦
septum. The flask was held at 90 C and stirred for 10 min.
were carried out on a Bruker DRX-500 NMR spectrome-
ter. The resonance frequency was 125.757 MHz. The finely
powdered sample f-Zr-TMS-30 was placed in 4.0-mm zir-
conia rotors and spun at 810 kHz. The CP/MAS spectrum
was recorded under Hartmann-Hahn match conditions using
a contact time of 1 m/s. A relaxation delay of 10 and 4 s was
used for DD/MAS and CP/MAS spectra, respectively. The
chemical shift was referred to an external adamantane CH2
peak reference taken as 28.7 ppm with respect to TMS. The
raw data obtained for DD/MAS (2000 scans) and CP/MAS
Then 0.28 mol of water was added dropwise into this mix-
ture under stirring to hydrolyze the zirconium (IV) butoxide
to Zr(OH)4. Further, after 30 min of stirring, triflic acid
0.03 mol) was slowly added by syringe into the above mix-
ture and stirring was continued for 2 h. The mixture was
cooled, filtered, washed with deionized water and acetone,
and then dried at 100 C for 6 h. Soxhlet extraction was car-
ried out for 24 h. Then the sample was dried at 200 C for
0 h. The syntheses of different loadings of triflic acid over
(
◦
◦
1
amorphous Zr(OH)4 (5 to 25 wt%) were carried out by vary-
ing the molar ratios of zirconium (IV) butoxide, water, and
CF3SO3H. The synthesized materials were white in color.
Functionalized amorphous materials are designated as ≡Zr–
O–SO2–CF3 (A).
(
3000 scans) were processed with an exponential function
of line broadening of 20 and 60 Hz, respectively, prior to
Fourier transformation for sensitivity enhancement.
Powder samples (ca. 10 mg) were compressed into a
small stainless-steel holder. Raman spectra of these samples
were measured at room temperature, exposed to atmosphere,
using a Bruker RFS/100S FT-Raman spectrometer. A Nd-
The SO42 /ZrO2 was obtained from MEL Chemicals
−
◦
(
Manchester, UK) and activated at 500 C for 10 h in sta-
−1
tic air prior to the catalytic run.
YAG laser (1064 cm ; 50 mW) was used as an excitation
source. Scattered radiation was collected at 180 C and de-
◦
2
.5. Catalyst characterization
tected using an InGaAs detector. Spectra were recorded in
−1
−1
the Stokes region between 2000 and 25 cm at 4 cm res-
olutions. One hundred scans were coadded. For comparison,
the Raman spectrum of liquid triflic acid (Aldrich, synthesis
grade; in a standard 4-mm NMR tube) was collected under
similar conditions. Only 50 scans needed to be coadded.
Temperature-programmed desorption was carried out to
determine the total acidity and strength of acid sites on the
catalysts using ammonia as an adsorbate [33]. In a typi-
The synthesized materials were mainly characterized by
powder X-ray diffraction (XRD) for phase purity and crys-
tallinity, N2 adsorption–desorption techniques for specific
surface area, total pore volume and average pore diameter,
Fourier transform infrared (FTIR) spectroscopy for func-
tional group confirmation, elemental analysis for C and S
to measure the triflic acid loading in the material, solid-state

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
445
Scheme 2.
Scheme 3.
cal run, 1.0 g of activated sample was placed in a silica
also identified by GC-MS (Shimadzu 2000 A) analysis and
◦
1
13
tubular reactor and heated at 200 C under nitrogen flow of
confirmed by H and C NMR spectra (Bruker AC-200).
Recycling of the catalyst was done during the acetaliza-
tion reaction by using f-Zr-TMS-30. After workup of the re-
action, the catalyst was separated by filtration, washed with
◦
5
0 ml/min for 6 h. The reactor was then cooled at 30 C and
adsorption conducted at that temperature by exposing the
sample to ammonia (10 ml/min) for 30 min. Physically ad-
sorbed ammonia was removed by purging the sample with a
◦
acetone, and dried at 200 C for 10 h in the presence of air.
◦
nitrogen stream flowing at 50 ml/min for 15 h at 30 C. The
The recovered catalyst after each cycle was characterized for
its crystallinity by XRD and its chemical composition by el-
emental analysis.
acid strength distribution was obtained by raising the cata-
◦
lyst temperature (10 K/min) from 30 to 300 C in a number
of steps with the flow of nitrogen (50 ml/min). The NH3
evolved was trapped in the HCl solution and titrated with a
standard NaOH solution.
3
. Results and discussion
The SEM micrographs of Zr-TMS and f-Zr-TMS-30 ma-
terials were taken by JEOL-JSM-5200 scanning microscopy.
TEM was performed on a JEOL JEM-1200EX instrument
with 100 kV of acceleration voltage to probe the mesoporos-
ity of the materials. The TG-DTA and DTG analysis of the
Zr-TMS and f-Zr-TMS-30 catalysts were carried out with a
3.1. Powder X-ray diffraction study
The powder X-ray diffraction (XRD) patterns of all the
synthesized catalysts are shown in Fig. 1. The template-
extracted Zr-TMS (Fig. 1A, Zr-TMS) and f-Zr-TMS (Fig. 1A,
e
Mettler Toledo 851 equipment using an alumina pan under
◦
a nitrogen (80 ml/min) atmosphere from ambient to 1000 C
with an increasing rate of 20 C/min.
5
–30 wt%) catalysts exhibited a single, broad reflection at
◦
◦
low 2θ (2.5–4.0 ). This reflection corresponds to the XRD
pattern of ordered mesoporous ZrO2 [11,13–18] and Si–
Zr mesoporous materials [34]. However, in these studies
no higher order reflections were observed. The presence of
these in our spectra may indicate that the pore walls are
rather amorphous or that there is a lack of correspondence
between the structures of adjacent pores. Other reflections
were observed at about 31 (broad) and at about 50 (small)
in all the samples. They are attributed to the tetragonal, mon-
oclinic, and cubic phases of ZrO2. They are readily formed
after calcination of the sample at higher temperature [15].
2
.6. Catalytic activity
The catalytic activity of all f-Zr-TMS was examined in
the acetalization of ethylacetoacetate with ethylene glycol
EG) using toluene as a solvent (Scheme 2) and benzoylation
(
◦
◦
of biphenyl with benzoyl chloride (BOC) using nitroben-
zene as a solvent (Scheme 3). The reactions were carried out
in a batch reactor equipped with two-necked 50-ml round-
bottom flask with septum, an oil bath, and condenser. AR
grade chemicals were used without further purification. In
a typical run, EAA and EG were added in a required mo-
lar ratio to the activated catalyst (0.1 g). In another run, BP
and BOC were added to the activated catalyst (0.5 g) in
the required molar ratio. The reaction mixtures were mag-
netically stirred and heated to the required temperature at
atmospheric pressure. The products were analyzed in a gas
chromatograph (HP 6890) equipped with a flame ionization
detector (FID) and a capillary column (5-µm-thick cross-
linked methyl silicone gum, 0.2 × 50 m). The products were
◦
The broad reflection at 2θ = 31 may also be partially due
to the presence of amorphous materials as was reported
in siliceous MCM-41 [35]. The crystallinity of the triflic
acid containing Zr-TMS materials decreased as the triflic
acid loading increased. Moreover, low intensity and the ab-
sence of higher order reflections show that the order and
mesostructure were entirely different from that measured for
the mesoporous silica [11]. ≡Zr–O–SO2–CF3 (A) catalysts
(Fig. 1B, 5–30 wt%) were indeed purely amorphous show-
◦
ing low intensity reflections at about 5, 31, and 50 (2θ).

446
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
Fig. 1. XRD patterns of (A) Zr-TMS and f-Zr-TMS (5–30 wt%) and (B)
Zr–O–SO –CF (A) (5–30 wt%).
≡
2
3
Fig. 2. N adsorption isotherms of (A) Zr-TMS and (B) f-Zr-TMS-30.
2
3
.2. Nitrogen adsorption–desorption study
triflic acid. The isotherms and pore size distributions of Zr-
TMS confirm the presence of mesopores. These results are
comparable with those previously reported for a mesoporous
ZrO2 [13].
The incorporation or anchoring of any medium (acid
or base) or metal in the framework positions and/or into
the walls of the supporting medium leads to a progres-
sive decrease in surface area [27,34,36]. The BET surface
area of the Zr-TMS, f-Zr-TMS, ≡Zr–O–SO2–CF3 (A), and
Figs. 2A and 2B show the nitrogen adsorption isotherm of
Zr-TMS and f-Zr-TMS-30, respectively. Both the isotherms
are indicative of type IV behavior characteristic of meso-
porous materials [11,13]. The hysteresis in the desorption
branch clearly shows the existence of mesopores. The des-
orption isotherm of Zr-TMS is similar to the calcined ZrO2
aerogels previously reported [13]. Before the functionaliza-
tion (Fig. 2A) the position of the inflection in the P/P0 = 0.4
to 0.83 region depends on the diameter of the mesopores
(broad pore size distribution) and its sharpness indicates the
uniformity of the pore size distributions. After the function-
alization (Fig. 2B), the position of the inflection is changed
in the P/P0 = 0.35 to 0.90 region. This is indicative of a
slight structural collapse in the material. However, pore size
distribution analysis still indicates that mesoporosity was not
lost.
2
−
SO4 /ZrO2 are given in Table 1. The surface area for the
2
−1
Zr-TMS was 371 m g , which is comparable to that mea-
sured previously for a Zr-TMS synthesized using a surfac-
tant with C16 carbon chain [11,13]. Surface area gradually
decreased with increasing triflic acid loading. The surface
area decreased to 284 m g for the maximum loading of
triflic acid (30 wt%) over Zr-TMS. In the same way, the
2
−1
2
−1
surface area of amorphous Zr(OH)4 was 80 m g and de-
2
−1
creased to 52 m g for ≡Zr–O–SO2–CF3 (A), 30. The
surface area of SO42 /ZrO2 was found to be 101 m g
−1
.
−
2
Zr-TMS and f-Zr-TMS-30 were analyzed by detailed N2
sorption studies. The BET surface area, average pore diam-
eter, and total pore volume of Zr-TMS and f-Zr-TMS-30
2
−1
3
−1
2
−1
were 371 m g , 44.3 Å, 0.31 cm g , and 284 m g ,
3
−1
3
4.0 Å, 0.25 cm g , respectively (Table 1). The decrease in
surface area, pore diameter, and pore volume of f-Zr-TMS-
0 may be attributed to the functionalization of Zr-TMS by
Figs. 3A and 3B show the BJH pore size distributions of
Zr-TMS and f-Zr-TMS-30, respectively. Zr-TMS possesses
a fairly narrow distribution of mesopores centered at circa 40
3

Table 1
Physico-chemical properties of synthesized catalysts
b
−1
◦
c
d
Catalyst
Elemental analysis
output) (wt%)
Loading of
BET surface
Crystal size
NH desorbed (mmol g ) between temperature ( C) (%)
NH chemisorbed
3
3
a
2
−1
◦
−1
(
CF SO H (wt%)
area (m g
)
(µm)
30–70
70–110
110–150
150–200
200–300
at 30 C (mmol g
)
3
3
C
S
Input
Output
Amorphous
Zr(OH)
–
–
–
5.0
–
4.7
9.6
12.2
15.2
24.2
27.7
7.8
80
72
68
63
59
56
52
101
0.35
0.40
0.35
0.38
0.37
0.40
0.39
0.45
12
11
10
17
14
13
13
–
27
33
33
29
30
29
29
–
33
37
38
35
37
38
37
–
20
12
15
15
15
17
17
–
8
6
3
3
3
3
4
–
0.49
0.81
1.05
1.27
1.36
1.53
1.67
1.45
4
e
f
≡
≡
≡
≡
≡
≡
SO
Zr–O–SO –CF (A) -5
0.7
1.5
1.5
1.7
2.0
2.7
–
1.00
2.04
2.6
3.19
5.16
5.91
3.08
2
3
Zr–O–SO –CF (A)-10
10.0
15.0
20.0
25.0
30.0
10.0
2
3
Zr–O–SO –CF (A)-15
2
3
Zr–O–SO –CF (A)-20
2
3
Zr–O–SO –CF (A)-25
2
3
Zr–O–SO –CF (A)-30
2
3
2−
g
/ZrO
4
2
Mesoporous
h
Zr-TMS
–
–
–
5.0
–
4.7
7.1
13.0
16.6
18.0
22.8
371
355
344
320
309
292
284
0.33
0.37
0.34
0.38
0.40
0.36
0.40
14
16
17
17
13
12
12
26
26
28
29
27
27
28
32
34
33
33
37
41
41
20
17
16
15
17
16
16
8
8
6
6
6
4
3
0.50
0.77
0.87
1.05
1.31
1.47
1.55
f
f-Zr-TMS-5
0.98
0.97
1.52
1.62
1.82
2.00
1.01
1.52
2.78
3.54
3.85
4.86
f-Zr-TMS-10
f-Zr-TMS-15
f-Zr-TMS-20
f-Zr-TMS-25
f-Zr-TMS-30
10.0
15.0
20.0
25.0
30.0
i
a
◦
Measured by N adsorption–desorption at −196 C.
2
b
c
d
e
f
Measured by JEOL SEM (JSM-5200).
◦
Percentage of acid strength distribution calculated based on NH desorbed from 30 to 300 C.
Total acid sites determined in the solid catalyst by NH adsorption–desorption.
3
3
A denotes amorphous.
Numbers denote wt% (input) of triflic acid loading over Zr-TMS.
g
h
i
◦
NH desorbed from 30 to 600 C.
3
3
−1
Total pore volume (cm g ) = 0.31 and average pore diameter (Å) = 44.3 measured by N adsorption isotherm.
2
3
−1
Total pore volume (cm g ) = 0.25 and average pore diameter (Å) = 34.0 measured by N adsorption isotherm.
2

448
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
Fig. 3. BJH pore size distributions of (A) Zr-TMS and (B) f-Zr-TMS-30.
to 50 Å with some as large as 100 Å. After functionalization
the distribution is clearly broadened. The material also con-
tains smaller pores of 10 to 20 Å and the amount of larger
pores (between 75 and 150 Å) is greatly increased. It is ap-
parent that a pore formation mechanism may be in effect that
differs from that of M41S materials. Pore size uniformity
does not necessarily imply bi- or three-dimensional X-ray
detectable ordering [11,13].
Fig. 4. IR spectrum of (A) Zr-TMS and f-Zr-TMS (5–30 wt%) and (B)
Zr–O–SO –CF (A) (5–30 wt%).
≡
3
.3. Elemental analysis study
2
3
The carbon and sulfur contents of the samples are shown
acid [37,38]. The C–S link in f-Zr-TMS also gives a medium
−1
in Table 1. To a first approximation, the sulfur content was
assigned to the loading of triflic acid over Zr-TMS and it was
observed that acid loading over Zr-TMS indeed increased
with the increase in the amount of CF3SO3H introduced. Af-
ter removal of surfactant no nitrogen was detected in the el-
emental analysis as expected. These results were confirmed
by infrared and NMR spectroscopies (see below).
band at 601 cm . This band is assigned to the SO defor-
2
−1
mation mode and a sharp band at 1043 cm is assigned to
C–F band [37,38]. Moreover, the spectra of f-Zr-TMS and
≡Zr–O–SO –CF (A) are similar to the reported silver tri-
2
3
flate spectrum [39]. Further, from Fig. 4A, Zr-TMS, it can
−
1
be seen that the stretching (3200–2800 cm ) and bending
−1
(1500–1300 cm ) modes of the methyl groups of CTMABr
completely disappear after 48 h of extraction (13, 15). Thus,
all these results indicate that the final material was free from
surfactant and that triflic acid was functionalized onto the
walls of Zr-TMS.
3
.4. Infrared spectroscopy study
The infrared spectra of Zr-TMS, f-Zr-TMS and ≡Zr–O–
SO2–CF3 (A) catalysts are shown in Figs. 4A and B, re-
spectively. The strong and broad band between 3600 and
3.5. Solid-state 13C-CP/MAS and 13C-DD/MAS NMR
studies
−1
3
200 cm corresponds to the stretching mode of hydroxyl
groups present on the surface (as Zr(OH)4). The weak un-
−
1
resolved band between 850 and 700 cm is attributed to
Zr–O stretching modes. The sharp band in the region 1650–
1
In order to confirm that the –CF group exists in the
3
material, the solid-state 13C-CP/MAS and DD/MAS NMR
−1
600 cm is due to the bending mode of associated water
spectra were recorded at 125.757 MHz at a rotational speed
of 8 kHz for f-Zr-TMS-30. The DD/MAS NMR spectrum
(Fig. 5) was found to be more sensitive than the CP/MAS
NMR spectra, indicating the absence of a proton environ-
ment in the proximity of the carbon under observation. The
molecules. The f-Zr-TMS (Fig. 4A, 5–30 wt%) and ≡Zr–
O–SO2–CF3 (A) (Fig. 4B, 5–30 wt%) spectra show addi-
−1
tional bands (at 1296, 1184, 1043, and 601 cm ) that are
absent in Zr-TMS (Fig. 4A, Zr-TMS). The broad and in-
tense band at 1296 cm and medium band at 1184 cm
are due to S=O stretching mode of the incorporated triflic
−1
−1
13
13
quartet nature of the C spectrum, arising due to the C–
1
9
F scalar coupling (JC–F ≈ 310 Hz), unambiguously shows

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
449
13
13
Fig. 5. Solid-state C CP/MAS and C DD/MAS NMR spectra of
f-Zr-TMS-30.
Fig. 7. FT-Raman spectra of adsorbed triflate at different acid loadings (A)
on amorphous Zr(OH) and (B) on Zr-TMS. The spectra of the solid ma-
4
−1
terial were subtracted in each case. (4 cm , 100 scans, 50 mW). The
spectrum of liquid triflic acid is also shown in each case for comparison.
sensitive to the presence or absence of mesopores. Fig. 7A
shows the Raman spectra of f-Zr-TMS with triflic acid load-
ings between 5 and 30 wt%. The spectrum of the Zr-TMS
support was subtracted from each. Also shown is the spec-
trum of liquid triflic acid. Clearly, the spectra of all the f-
Zr-TMS samples are very similar to each other and to that
of liquid trifluoromethanesulfonic acid. No bands appear in
the spectra of the f-Zr-TMS samples that do not appear in
the spectrum of the acid itself. Only small peak shifts exist
for some bands on the supported samples relative to the liq-
uid acid. These will be discussed in detail later. Comparing
Fig. 7A, 5–30 wt%, one observes that the intensity of each
band increases with increased loading of triflate. Fig. 7A,
20 wt%, is somewhat anomalous and may indicate that the
sample actually contains more than 20 wt% triflate. Elemen-
tal analysis in fact revealed that the sulfur content was lower
than 20 wt% (16.6 wt%). According to Table 1 the actual
change in S content between spectra 7A, 20–30 wt%, was
smaller than theoretically calculated (13–22.8 wt% rather
than 15–30 wt%). Fig. 7B shows the spectra for the ≡Zr–O–
SO2–CF3 (A) samples. Again the spectrum of the support
was subtracted. These spectra are again very similar to that
of the liquid acid.
−
1
Fig. 6. FT-Raman spectra of amorphous Zr(OH) and Zr-TMS, 4 cm
,
4
100 scans, 50 mW.
the presence of CF3 groups. Moreover, the chemical shift
observed at ≈ 119 ppm is very close to that reported for Na–
O–SO2–CF3 (≈ 120 ppm) [40]. These features indicate that
the –CF3 group is intact in the sample. Further, no peak cor-
responding to the surfactant was observed, which would nor-
mally appear in the range 10–100 ppm in the 13C CP/MAS
NMR spectrum.
3
.6. FT-Raman analysis
Fig. 6 shows the Raman spectra of Zr-TMS and amor-
phous Zr(OH)4. The spectra of the two zirconia materials
are identical, with bands at 120 cm (w), 84 cm (m), and
9 cm (w). Apparently these low-frequency modes are not
−1
−1
−1
3

450
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
Table 2
−
1
Compilation of Raman bands measured (cm ) for f-Zr-TMS, ≡Zr–O–SO –F (A) and for liquid CF SO H
f-Zr-TMS
2
3
3
3
a
≡Zr–O–SO –CF (A)
CF SO H (liq.)
Assignment [41]
2
3
3
3
a
This study
Literature [41]
ꢀ
–
–
2999
1410
1248
∼ 1228
1151
1130
1030
929
2997 [2]
∼ 1400
1244
∼ 1232
∼ 1152
1122
υ(OH) A [42]
ꢀꢀ
1
1
1
412
256
229
–
–
038
1396
1266
1228
1185
–
1040
924
769
υas(SO ) A
υs(SO ) A
υ (CF ) E
υs(CF ) A
2
ꢀ
2
d
3
3
1
ꢀ
δ (SOH) A
−
1
1034 [42],1038 [43]
υs(S–O ) [42,43]
3
ꢀ
9
7
29
68
930
775
υ(SO(H)) A
ꢀ
773
δs(CF )A [41],
3
ꢀ
υ(CS) A [42]
ꢀ
∼
620
76
13
480
50
19
619
578
518
492
351
321
619
568
499
478 (sh)
339
622
570
499
478 (sh)
340
δ(SO bend) A
2
5
5
δ (CF ) E
d 3
ꢀ
δ(SO(H) wag) A
ꢀꢀ
δ(SO rock) A
2
∼
ꢀ
3
3
δ(SCF wag) A
3
ꢀ
309
312
υ(CS) A [41],
ꢀꢀ
CF wag A [42]
3
∼
223
213
202
201
ρ(CF ) E
3
The band assignments were made previously [41,42] and apply to the liquid acid molecule only.
a
Assignments were made assuming that the CF SO H molecule belongs to the point group Cs, however, with local C symmetry (CF group).
3
3
3v
3
The peak data from Figs. 7A and B are compiled in Ta-
ble 2. Also listed are previously published data regarding
the Raman spectrum of liquid trifluoromethanesulfonic acid
and the band assignments previously made for this molecule
The C3v point group has 18 normal vibrational modes that
ꢀ
ꢀꢀ
can be classified by the symmetry types: Γvib = 5A + A +
ꢀꢀ
6E. All modes, excepting the A mode, are both IR and Ra-
man active. Hall and Hansma [43] studied the adsorption
and orientation of sulfonic acids on alumina using inelas-
tic tunneling spectroscopy and found that sulfonate groups
also bind through three equivalent oxygen atoms in a triden-
tate form on this surface. The four possible binding modes
[
41,42]. As shown in Table 2, excellent agreement exists
between the measured and the previously reported Raman
spectra of liquid triflic acid. The assignments are from the
work of Katsuhara et al. [41] who were the first to measure
−
−1
of CF SO3 (a) monodentate, (b) bidentate, (c) tridentate,
3
and assign the complete spectrum between 1500 and 0 cm
.
and (d) tridentate (distorted) can be distinguished by their
spectra in the following manner. The lowered symmetry of
monodentate sulfonates (bound via only one oxygen) leads
They assumed that the CF3SO3H molecule belonged to the
point group Cs with, however, local C3v symmetry (the CF3
group). They also assumed that this group was free to ro-
tate around the C–S bond. The most relevant bands for this
−
to splitting of the three degenerate S–O3 stretching modes
−1
and produces a spectrum containing three distinct modes
study are the SOH deformation mode at 1122 cm and
the symmetric and asymmetric stretching modes of SO2 (at
−1
−1
spanning about 500 cm , centered around 1100 cm .
Bidentate ligands (bound via two oxygen atoms) also pro-
duce three different S–O stretching modes centred around
−
1
1
248 and 1410 cm , respectively). The symmetric stretch-
−
ing band of the triflate anion (CF3SO3 ) is also included in
− −1
1
1
150 to 1250 cm and separated by at least 300 cm . The
−1
−1
Table 2 (1034 cm [42], 1038 cm [43]). This peak was
not observed in very pure acid since there is very little dis-
totally symmetric tridentate sulfonate group (C3v) has only
two different S–O stretching modes, the symmetric stretch
υs and the doubly degenerate asymmetric stretch υa. Slight
distortion from C3v symmetry results in loss of degener-
acy of these two modes resulting in two asymmetric S–O
−1
sociation. We observed a small band at 1030 cm . Since no
attempt was made to purify the acid or to keep it totally dry,
trace water obviously led to some dissociation in our case.
Finally, the OH stretching frequency, previously reported at
−1
bands slightly split (ꢀ 32 cm ) about their degenerate po-
−1
−1
2
997 cm [42], was observed at 2999 cm in this study.
−
sition. The vibrations involving the S–O3 group for the
CF3SO3 anion and their measured frequencies (using IR
Previous work on bulk samples of metal sulfonate salts [43]
showed that the sulfonate oxygen atoms were in equivalent,
symmetric positions. Normal coordinate analyses of IR and
−
and Raman) are shown in Table 3. In a recent previous pub-
lication [44] some of us reported the transmission infrared
spectrum of f-Zr-TMS. The spectrum contained four bands
not present in the IR spectrum of the Zr-TMS support. These
−
Raman spectra indicated that the CX3SO3 anions (X = H,
D, F, Cl) had C3v symmetry in which the three oxygen atoms
coordinate in an equivalent manner to a metal ion with three-
fold rotational symmetry about the C–S bond. Essentially the
sulfonate group acts as a tripod.
−
1
−1
bands appeared at 1296 cm (vs), 1184 cm (m), 1043
−1
−1
cm (vs), and 601 cm (m, br). The first two bands were
previously assigned to υa and υs of the CF3S–O3 group
−

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
451
Table 3
−
−
−1
Vibrational modes of S–O group (assuming C symmetry) for CF SO and their measured IR and Raman frequencies (cm
)
3
3v
3
3
Mode
IR [43]
Raman [43]
Raman [this study]
On amorphous Zr(OH)
On Zr-TMS
4
−
ꢀ
υs(S–O ) A
1019–1049 (s–vs)
1152–1194 (s–vs)
630–676 (s–vs, br)
515–531 (w–s)
1038 (s–vs) p
1188 (vw) dp
635 (s) p
520 (vw) dp
353–359 (m) dp
1038
–
1041
1185
619
518
351
3
−
υa(S–O ) E
3
−
ꢀ
δs(S–O ) A
∼ 620
3
−
δa(S–O ) E
513
3
−
ρa(S–O ) E
350–356 (w–s)
350
3
bound to the surface. However, a more careful examination
bonded to the solid support, it could not be treated above
300 C (above this temperature triflic acid is lost from the
−1
◦
of the spectrum shows an unresolved shoulder at 1200 cm
.
As shown in Table 3, this band should be assigned to υa
solid support [20,26,27,36]). Table 1 shows the total number
of acid sites (determined via NH3 chemisorbed at 30 C) of
Zr-TMS, f-Zr-TMS and ≡Zr–O–SO2–CF3 (A). The quanti-
tative distributions of the acid sites (in the 5 regions) are also
shown.
The total number of acid sites on the catalysts was found
to increase proportionally with increased loading of triflic
acid supported on Zr-TMS. The same trend of increasing the
total number of acid sites with increasing triflic acid loading
was observed in the ≡Zr–O–SO2–CF3 (A) catalysts (see Ta-
ble 1). A larger number of acid sites were observed for the
amorphous catalyst (≡Zr–O–SO2–CF3 (A)-30) than for the
mesoporous catalyst (f-Zr-TMS-30). This is consistent with
the results of elemental analysis (see Table 1).
−
◦
(
CF3S–O3 ) E. This very weak mode was observed at 1038
−1
cm on the ≡Zr–O–SO2–CF3 (A) samples (see Table 2).
−1
The band at 1296 cm should be assigned to υa (C–F3)
E. According to [43], this band has been observed in the IR
−
1
−1
between 1259 and 1280 cm (vs, br) and at 1285 cm
vw, dp) in some Raman spectra. Very weak bands were de-
(
−
1
tected at 1256 and 1266 cm in the spectra of the f-Zr-TMS
and ≡Zr–O–SO2–CF3 (A) materials, respectively. In addi-
−1
tion, the band previously reported at 601 cm in the IR
spectrum is more probably two unresolved bands. Table 3
shows that the symmetric deformation mode of triflate an-
−
ꢀ
ion, δs (S–O3 ) A appears at slightly higher frequencies
− −1
1
(
630–676 cm (s–vs, br)) in the IR, and at 635 cm (s, p)
−1
in the Raman. Bands were observed at ∼ 620 cm on both
the amorphous and the mesoporous samples. The second
component of the band is probably due to the asymmetric
deformation mode of the CF3 group [δd (C–F3) E]. As re-
ported in [43], this mode appears between 577 and 591 cm
vw–s) in the IR, and at 580 cm in the Raman (vw, dp).
The total number of acid sites on the Zr-TMS was found
−1
to be 0.50 mmol g ; the Brønsted acid sites are formed
by the hydroxyl groups that exist in the Zr-TMS mater-
◦
ial. The total amount of NH3 chemisorbed at 30 C was
0.77 mmolg
−1
−1
−1
for
for f-Zr-TMS-5, and 1.55 mmolg
−
1
2−
(
f-Zr-TMS-30. The total number of acid sites on SO4 /ZrO2
−
1
◦
Bands were observed at 576 cm in this study.
(measured between 30 and 600 C) was found to be 1.45
−1
mmol g . Silica-supported triflic acid showed superacid
character (i.e., the acidity was greater than that of 100%
H2SO4 [26]. Triflic acid alone is also known to be a su-
peracid [20].
3
.7. Ammonia adsorption–desorption study
Experimental data obtained using the Hammett acidity
function have shown that triflic acid and perfluorinated acid
Further, the acid strength distribution (in %) of Zr-TMS,
f-Zr-TMS, amorphous Zr(OH) , and ≡Zr–O–SO –CF (A)
resins are superacids with Ho values between −12 and
4
2
3
−
14 [45,46]. However discrepancies and uncertainties in
are given in Table 1. The results of ammonia TPD measure-
−1
◦
the practical applications of Hammett acidity made us look
for other methods to determine the acid strength of our
functionalized materials. Marziano et al. [36] and Corma
et al. [47] used potentiometric titration with standardized
NaOH and ammonia TPD measurements to obtain the total
acidity and the distribution of their acid strengths of triflate
ments reveal that the total acidity (mmol g at 30 C) in-
creases with the increase in loading of triflic acid over amor-
phous Zr(OH) and Zr-TMS. However, the acid strength dis-
4
tribution (measured in five steps) remains nearly unchanged
◦
except between 110 and 150 C. It is observed that the per-
◦
centage of the acid sites (between 110 and 150 C) increases
and SO42 /ZrO2 materials, respectively. Hence, the ammo-
nia TPD technique was used to measure the acidity of the
synthesized catalysts.
−
from 33 to about 38%, and 32 to 41% when the triflic acid
loading was increased from 0 to 30 wt% over amorphous
Zr(OH) and Zr-TMS, respectively.
In addition to the total number of acid sites, a quantitative
distribution of acid strengths of the sites on the function-
alized materials was obtained by measuring the amounts
in five arbitrarily defined temperature regions during the
3.8. Scanning electron microscopic study
The particle size and morphology of Zr-TMS and f-Zr-
TMS-30 were studied by SEM. SEM micrographs of Zr-
TMS (Fig. 8A) and f-Zr-TMS-30 (Fig. 8B) are typical of or-
dered ZrO2- [48] and MCM-41- [49] type materials exhibit-
◦
ammonia TPD experiment (between 30 and 70 C, 70 and
◦
◦
◦
1
3
10 C, 110 and 150 C, 150 and 200 C, and 200 and
◦
00 C). Since the functionalized materials are covalently

452
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
Fig. 8. Scanning electron micrographs of (A) Zr-TMS and (B) f-Zr-
TMS-30.
Fig. 9. Transmission electron micrographs of (A) Zr-TMS and (B) f-Zr-
TMS-30.
ing the winding-worm type. It is noteworthy that previous
authors have reported mesoporous zirconia to be very disor-
dered [11]. Zr-TMS consisted of ordered particles (winding-
worm type) of approximately 0.33 µm in size. The parti-
cle size of f-Zr-TMS-30 was slightly larger (approximately
3
.10. Thermal analysis (TG, DTA, and DTG)
Thermal analysis of all synthesized catalysts showed sim-
ilar results. Figs. 10A and 10B show typical TG, DTA, and
DTG profiles measured for Zr-TMS and f-Zr-TMS-30 sam-
ples, respectively. The TG curve of Zr-TMS shows three
stages of weight loss. The weight loss (endothermic) be-
tween 70 and 200 C corresponds to the loss of loosely
bound water (desorption of adsorbed water) [52] and resid-
ual butanol (the latter was used as solvent in the synthesis
0
.4 µm). The particles appeared to be of a distorted winding-
worm type. It appears that functionalization alters the par-
ticle size to some extent. The increase in the particle size
with the increase of CF3SO3H loading is attributed to an-
choring of triflic acid on Zr-TMS. This decrease in ordering
and crystallinity was also observed by XRD.
◦
◦
mixture). Further, the weight loss between 200 and 400 C
corresponds to the decomposition of Zr(OH)4 into ZrO2
◦
[
11,13]. A slight decrease at about 450 C in the TG curve
3
.9. Transmission electron microscopic study
and a corresponding sharp exothermic peak at the same
temperature in the DTA curve are indicative of an addi-
tional phase coexisting with zirconia. This may be a quasi-
amorphous tetragonal phase produced in the decomposition
process of Zr(OH) [53]. No further weight losses were ob-
served above 650 C. At this point the residue is anhydrous
The samples were prepared for TEM analysis by sonifica-
tion in isopropanol followed by deposition on carbon-coated
copper grids. Transmission electron micrographs of Zr-TMS
4
◦
(
Fig. 9A) and f-Zr-TMS-30 (Fig. 9B) reveal that these ma-
terials have a highly porous nature and contain either disor-
dered hexagonal phases or pseudo-hexagonal arrays. These
results are in good agreement with the earlier studies of Zr-
TMS [50,51].
ZrO2.
The TG curve of f-Zr-TMS-30 (Fig. 10B) shows four
stages of weight loss as described above. The decrease be-
tween 70 and 200 C corresponds to the loss of loosely
◦

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
453
also carried out using SO42 /ZrO2 and triflic acid for com-
−
parison.
3
.12. Acetalization of ethylacetoacetate
The results of acetalization of ethylacetoacetate to fruc-
tone are given in Table 4. Under similar reaction conditions,
the f-Zr-TMS catalysts were more active than Zr-TMS and
comparable to SO42 /ZrO2 and ≡Zr–O–SO2–CF3 (A). The
Zr-TMS and amorphous Zr(OH)4 showed 21 and 18 wt%
conversion of EAA, respectively, due to the limited num-
ber of acid sites. However, in terms of selectivity amorphous
Zr(OH)4 gave 97 wt% to fructone whereas Zr-TMS was
−
1
00 wt% selectivity to fructone. This higher selectivity is
attributed to the presence of well-defined pores. Since the
acidity of the functionalized mesoporous and the amorphous
catalysts are similar with respect to the triflic acid loading
(
Table 1), both series of catalysts showed more or less sim-
ilar conversion of EAA. The EAA conversion was 82 wt%
for f-Zr-TMS-5 and 93 wt% for f-Zr-TMS-30. Further, the
EAA conversion over ≡Zr–O–SO2–CF3 (A)-5 and ≡Zr–O–
SO2–CF3 (A)-30 was 83 and 92 wt%, respectively. All the
f-Zr-TMS catalysts showed the conversion of EAA between
8
2 and 93 wt% with 100 wt% selectivity to fructone (Ta-
Fig. 10. Thermogravimetric differential thermal analysis (TG, DTA, and
DTG) profiles of (A) Zr-TMS and (B) f-Zr-TMS-30.
ble 4). The selectivity of the ≡Zr–O–SO2 CF3 (A) catalysts
to fructone was much poorer (between 64 and 82 wt%) than
over the f-Zr-TMS catalysts. SO42 /ZrO2 showed 93 wt%
conversion of EAA and 90 wt% selectivity to fructone. The
lower selectivity of SO42 /ZrO2 may be correlated due to its
nonuniform pore size, low porosity, and smaller surface area.
Pure CF3SO3H showed higher conversion (95 wt%) than all
the catalysts, with a selectivity of 74 wt%. The homoge-
neous medium of strong acid and the water formed during
this reaction can cause the hydrolysis of the ester, producing
the corresponding EDBA (3,3-ethylenedioxybutanoic acid).
The formation of this product not only reduces the yield of
fructone, but also it can alter the organoleptic characteris-
tics of the final product. The turnover frequency (TOF) was
calculated as number of moles of EAA converted per moles
of sulfur per hour. Since the conversions were almost the
same in all the cases with respect to the loading of triflic
acid, TOFs was found to be approximately similar for both
−
bound water and residual butanol. The large weight losses
−
◦
◦
between 320 and 400 C [28] and between 400 and 440 C
correspond to the loss of the trifluoromethane sulfonate
(
O–SO2–CF3) group from the zirconia surface and to the
complete phase change of Zr(OH)4 to ZrO2, respectively.
Hence these functionalized materials cannot be treated
◦
above 300 C.
Differential thermal gravimetric analysis of the catalysts
agreed well with the results obtained by TG-DTA. Fig. 10A
shows the DTG profile of the Zr-TMS catalyst. Two en-
◦
dothermic peaks were observed (at about 110 and 450 C)
for Zr-TMS whereas three endothermic peaks were observed
◦
at about 100, 370, and 440 C) for f-Zr-TMS-30 (Fig. 10B).
(
◦
The endothermic peak at 360 C corresponds to the loss of
trifluoromethanesulfonate. The sharp endothermic peak at
◦
−1
−1
4
40 C is attributed to the additional phase and the phase
types of catalysts (Table 4). The TOFs (in mol S h ) for
f-Zr-TMS-5, f-Zr-TMS-30, and ≡Zr–O–SO2–CF3 (A)-5 and
≡Zr–O–SO2–CF3 (A)-30 were 259 and 61.2 and 265.6 and
change of zirconia. In addition, a sharp exothermic peak at
◦
about 900 C in the DTA curve corresponds to the crystal-
2
−
lization of material and the phase change from tetragonal to
monoclinic and it is also observed in Zr-TMS (Fig. 10A).
49.8, respectively. The TOFs of triflic acid and SO4 /ZrO2
were 54.0 and 96.9, respectively.
3
.11. Catalytic activity
3.13. Benzoylation of biphenyl
The catalytic activity of the synthesized materials was
The results for the benzoylation reaction of BP to 4-PBP
are given in Table 5. As in the acetalization reaction, due to
the limited acidity, both Zr-TMS and amorphous Zr(OH)4
showed comparable conversion of BP, 3.2 and 3.1 wt%,
respectively. Further, the biphenyl conversion dramatically
examined in the acetalization of ethylacetoacetate with eth-
ylene glycol to fructone, and in the benzoylation of biphenyl
◦
with benzoyl chloride to 4-PBP in batch reactors at 100 C
◦
for 1 h and 170 C for 6 h, respectively. Both reactions were

454
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
Table 4
a
Acetalization of ethylacetoacetate
b
c
Catalyst
Conversion of EAA
TOF
Product distribution (wt%)
−
1
−1
(wt%)
(h mol S)
EDBA
Fructone
Amorphous
Amorphous-Zr(OH)
18.0
83.0
83.0
89.0
94.0
89.0
92.0
95.0
93.0
–
3
36
18
27
21
22
35
26
10
97
64
82
73
79
78
65
74
90
4
≡
≡
≡
≡
≡
≡
Zr–O–SO –CF (A)-5
265.6
130.2
105.5
94.1
55.2
49.8
54.0
96.6
2
3
Zr–O–SO –CF (A)-10
2
3
Zr–O–SO –CF (A)-15
2
3
Zr–O–SO –CF (A)-20
2
3
Zr–O–SO –CF (A)-25
2
3
Zr–O–SO –CF (A)-30
2
3
d
CF SO H
3
3
2−
SO
/ZrO
4
2
Mesoporous
Zr-TMS
f-Zr-TMS-5
f-Zr-TMS-10
f-Zr-TMS-15
f-Zr-TMS-20
f-Zr-TMS-25
f-Zr-TMS-30
21.0
82.0
87.0
88.0
88.0
91.0
93.0
–
0
0
0
0
0
0
0
100
100
100
100
100
100
100
259.0
183.0
101.3
79.5
75.6
61.2
a
◦
Reaction conditions: catalyst (g) = 0.1; ethylacetoacetate (mmol) = 10; ethylene glycol (mmol) = 10; reaction time (h) = 1; reaction temperature ( C) =
00; toluene (g) = 25.
1
b
TOF is given as moles of EAA transformed per hour per mole of sulfur.
c
EDBA, 3,3-ethylenedioxybutanoic acid; Fructone, ethyl 3,3-ethylenedioxybutyrate.
Catalyst (g) = 0.125; reaction time (h) = 1.
d
Table 5
Benzoylation of biphenyl
a
b
c
Catalyst
Conversion of biphenyl
TOF
Product distribution (wt%)
−1
−1
−1
(wt%)
(10
h
mol S)
2-PBP
4-PBP
Amorphous
Amorphous-Zr(OH)
3.1
5.0
–
0
3
2
2
4
3
2
7
0
100
97
98
98
96
97
98
93
100
4
≡
≡
≡
≡
≡
≡
Zr–O–SO –CF (A)-5
5.34
5.55
6.36
5.34
4.2
3.7
54.3
0.98
2
3
Zr–O–SO –CF (A)-10
10.2
15.5
16.0
19.1
20.5
63.7
7.2
2
3
Zr–O–SO –CF (A)-15
2
3
Zr–O–SO –CF (A)-20
2
3
Zr–O–SO –CF (A)-25
2
3
Zr–O–SO –CF (A)-30
2
3
d
CF SO H
3
3
2−
SO
/ZrO
4
2
Mesoporous
Zr-TMS
f-Zr-TMS-5
f-Zr-TMS-10
f-Zr-TMS-15
f-Zr-TMS-20
f-Zr-TMS-25
f-Zr-TMS-30
3.2
22.0
27.0
31.5
35.0
37.0
41.0
–
23.2
0
0
0
0
0
0
0
100
100
100
100
100
100
100
21.9
12.1
10.5
10.2
9.0
a
◦
Reaction conditions: catalyst (g) = 0.5; biphenyl (mmol) = 10; benzoyl chloride (mmol) = 10; reaction time (h) = 6; reaction temperature ( C) = 170;
nitrobenzene (ml) = 20.
b
TOF is given as moles of BP transformed per hour per mole of sulfur.
c
2-PBP, 2-phenylbenzophenone; 4-PBP, 4-phenylbenzophenone.
d
Catalyst (g) = 0.125; reaction time (h) = 3.
increased after the functionalization of triflic acid over Zr-
TMS. All the f-Zr-TMS catalysts showed higher conversion
of BP with 100 wt% selectivity to 4-PBP. The conversion
of BP over the f-Zr-TMS catalysts was between 22 and 41
wt% when the triflic acid loading was increased from 5 to
30 nominal wt%. In the same manner, the conversion of BP
over the ≡Zr–O–SO2–CF3 (A) catalysts was between 5 and
20 wt% for nominal loadings of triflic acid increasing from
5 to 30 wt%. In terms of selectivity, the f-Zr-TMS catalysts
showed 100 wt% to 4-PBP whereas the ≡Zr–O–SO2–CF3

M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
455
Table 6
a
Recycling of f-Zr-TMS-30 in acetalization reaction
b
c
Cycle
Elemental analysis (wt%)
Conversion
TOF
1
Selectivity to fructone
(wt%)
−
−1
C
S
of EAA (wt%)
(h mol S)
Fresh
First recycle
Second recycle
2.0
1.96
1.92
4.86
4.79
4.71
93.0
86.0
86.0
61.2
61.8
68.7
100
100
100
a
Reaction conditions: catalyst (g) = 0.1; ethylacetoacetate (mmol) = 10; ethylene glycol (mmol) = 10; reaction time (h) = 1; toluene (g) = 25; reaction
◦
temperature ( C) = 100.
b
Elemental analysis by EA1108 elemental analyzer (Carlo Erba).
TOF is given as moles of EAA transformed per hour per mole of sulfur.
c
showed that both sulfur content and catalyst crystallinity de-
creased after each cycle. The conversion of EAA decreased
to some extent on recycling (from 93 to 86 wt%), whereas
the selectivity to fructone did not change. The loss of sulfur
and the decrease in the crystallinity of the catalyst were the
likely causes of the decrease in catalytic activity.
4
. Conclusions
In conclusion, Zr-TMS catalysts have been synthesized
based on the sol-gel method by using CTMABr and TMAOH.
The template was extracted from the synthesized materi-
als by using ethanol and HCl. The extracted Zr-TMS was
successfully functionalized with triflic acid by postsynthesis
treatment to obtain covalently bonded f-Zr-TMS catalysts.
Various loadings of triflic acid over Zr-TMS were pre-
pared by varying the molar ratios of Zr(OC4H9)4, CTMABr,
TMAOH, H O, and CF SO H. Functionalized amorphous
Fig. 11. XRD patterns of fresh and recycled catalysts (a) f-Zr-TMS-30
fresh), (b) after one recycle, and (c) after two recycles.
2
3
3
(
≡
Zr–O–SO2–CF3 catalysts were also synthesized and char-
acterized for comparison. The catalysts were characterized
by various physico-chemical techniques such as XRD, N2
adsorption–desorption, FTIR, FT-Raman, elemental analy-
(
A) catalysts produced a small amount of 2-PBP due to its
2
−
nonporous structure. Triflic acid and SO4 /ZrO2 showed
3.7 and 7.2 wt% conversions of BP, respectively. The se-
1
3
6
sis, C DD/MAS NMR, TPD of NH , SEM, TEM, and
3
lectivity to 4-PBP over CF3SO3H was only 93 wt% due to
its nonshape-selective nature. These results demonstrate that
the BP benzoylation reaction essentially needs a porous cat-
alyst with uniform pore size and higher acidity. The TOFs
were calculated as the number of moles of BP converted per
thermal analysis. BET surface area and pore size distribution
results were in general agreement with previous values re-
ported for mesoporous ZrO . The NH -TPD measurements
2
3
1
3
showed that the catalysts were highly acid. C DD/MAS
NMR revealed that the –CF group remained intact in the
3
−1
−1
−1
mole of sulfur per hour. TOFs (in 10 mol S h ) of BP
conversion over f-Zr-TMS-5, f-Zr-TMS-30, ≡Zr–O–SO2–
CF3 (A)-5, and ≡Zr–O–SO2–CF3 (A)-30 were 23.2, 9.0,
material. FT-Raman analysis demonstrated that the triflic
acid was bonded in an identical fashion on both amorphous
Zr(OH) and Zr-TMS at all loadings. Triflate ligands bound
4
−
1
−1
−1
5
.34, and 3.7, respectively. The TOFs (in 10 mol S h )
via 3 equivalent oxygen atoms to zirconium atoms form-
2
−
of CF3SO3H and SO4 /ZrO2 were 54.3 and 0.98, respec-
tively.
ing tripod structures. TEM studies teach that the material
contained disordered channels, unlike MCM-41 mesoporous
molecular sieves. Acetalization of EAA to fructone and ben-
zoylation of BP to 4-PBP reactions were performed on f-Zr-
3
.14. Catalyst recycling study
2
−
TMS, ≡Zr–O–SO2–CF3 (A), triflic acid, and SO4 /ZrO2.
f-Zr-TMS catalysts were found to be the most active and se-
lective in both reactions due to their mesoporosity and to
an increase in the number of acid sites with the “right” acid
strength. The stability and recycle effect of the catalysts were
checked in the acetalization reaction by using f-Zr-TMS-30.
No major loss of activity was observed after two recycles,
Recycle of the synthesized catalysts was studied in the ac-
etalization reaction using f-Zr-TMS-30 in order to check the
stability and activity of recycled catalysts (Table 6). Three
reaction cycles (fresh and two recycles) were carried out
under similar reaction conditions, using the same catalyst.
Elemental analysis (Table 6) and XRD analysis (Fig. 11)

456
M. Chidambaram et al. / Journal of Catalysis 220 (2003) 442–456
but a decrease of the sulfur content and catalyst crystallinity
was observed.
[22] D.G. Russel, J.B. Senior, Can. J. Chem. 52 (1974) 2975.
[
[
[
[
23] J.W. Brockington, R.H. Bennett, US patent 3,970,721, 1976.
24] J.F. Joly, C. Marcilly, E. Benazzi, US patent 5,336,833, 1994.
25] S.I. Hommeltoft, Eur. patent 0,987,237A, 2000.
26] D.Q. Zhou, J.H. Yang, G.M. Dong, M.Y. Huang, Y.Y. Jiang, J. Mol.
Catal. A 159 (2000) 85.
Acknowledgments
[27] A.N. Parvulescu, B.C. Gagea, M. Alifanti, V. Parvulescu, V.I. Parvu-
lescu, S. Nae, A. Razus, G. Poncelet, P. Grange, J. Catal. 202 (2001)
The authors are thankful to Drs. C. Gopinathan, S.S. De-
sphande, P.R. Rajamohanan, H.S. Soni, N.E. Jacob, A.A.
Belhekar, R. Parischa, S. Violet, R. Marimuthu, and C. Ven-
katesan for their valuable contribution and cooperation in
characterizing the catalysts.
319.
[
[
28] K. Wilson, A. Renson, J.H. Clark, Catal. Lett. 61 (1999) 51.
29] K. Bauer, D. Garbe, H. Surburg, Common Fragrances and Flavors Ma-
terials, second ed., VCH, New York, 1990.
[30] A. Walczak, J. Rzasa, S. Labus, Pol PL. 170632, 1997.
[
31] J.I. Kroschwitz, M. Howe-Grant (Eds.), Kirk–Othmer Encyclopedia of
Chemical Technology, 4th ed., Wiley, New York, 1995, p. 454.
32] S. Hitz, R. Prins, J. Catal. 168 (1997) 194.
33] M. Chamumi, D. Brunel, F. Fajula, P. Geneste, P. Moreau, J. Solof,
Zeolites 14 (1994) 283.
[
[
References
[
[
[
[
[
[
1] Y.D. Xia, W.M. Hua, Y. Tang, Z. Goa, Chem. Commun. (1999) 1899.
2] T. Jin, T. Yamaguchi, K. Tanabe, J. Phys. Chem. 90 (1986) 4797.
3] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1979) 1148.
4] A. Corma, Chem. Rev. 95 (1995) 559.
5] Z. Liu, W. Ji, L. Dong, Y. Chen, Mater. Chem. Phys. 56 (1998) 134.
6] D.M. Antonelli, J.Y. Ying, Angew. Chem., Int. Ed. Engl. 35 (1996)
[34] A.O. Bianchi, M. Champanati, P.M. Torres, E.R. Castellon, A.J. Lo-
pez, A. Vaccari, Appl. Catal. A 220 (2001) 105.
[35] U. Ciesla, S. Schacht, G.D. Stucky, K.K. Unger, F. Schuth, Angew.
Chem. Int. Ed. Engl. 35 (1996) 541.
[36] N.C. Marziano, L.D. Ronchin, C. Tortato, A. Zingales, A.A. Sheikh-
Osman, J. Mol. Catal. A 174 (2001) 265.
4
26.
[37] G. Herzberg, Infrared and Raman Spectra of Polyatomic Molecules,
Van Nostrand, New York, 1945, p. 285.
[38] L.J. Bellamy, The Infrared Spectra of Complex Molecules, Wiley, New
York, 1960, p. 328.
[
[
7] D.M. Antonelli, J.Y. Ying, Chem. Mater. 8 (1996) 874.
8] D. Terrible, A. Trovarelli, J. Llorca, C. de Leitenburg, G. Dolcettin,
J. Catal. 178 (1998) 299.
[
9] A. Sayari, P. Liu, Micropor. Mater. 12 (1997) 149.
[39] C.J. Pouchert, The Aldrich Library of IR Spectra, 3rd ed., 1981, p. 533.
[40] C.J. Pouchert, J. Behnke, The Aldrich Library of C and H FT NMR
Spectra, Vol. 1, p. 1431.
[41] Y. Katsuhara, R.M. Hammaker, D.D. DesMarteau, Inorg. Chem. 19
(1980) 607.
[42] H.G.M. Edwards, Spectrochim. Acta 45A (1989) 715.
[43] T. Hall, P.K. Hansma, Surf. Sci. 71 (1978) 1, and references therein.
[44] M. Chidambaram, C. Venkatesan, P.R. Rajamohanan, A.P. Singh,
Appl. Catal. A: Gen. 244 (2003) 27.
[45] C.H. Rochester, Acidity Functions, Academic Press, London, 1970.
[46] G.A. Olah, G.K. Surya Prakash, J. Sommer, Science 206 (1973) 13.
[47] A. Corma, M.I. Fornes, J.M. Juan-Rajadell, L. Neito, Appl. Catal.
A 116 (1994) 151.
[48] X. Ju, P. Huang, N. Xu, J. Shi, J. Membr. Sci. 166 (2000) 41.
[49] S.C. Laha, P. Mukherjee, S.R. Sainkar, R. Kumar, J. Catal. 207 (2002)
213.
13
1
[
[
10] G.D. Yadav, J.J. Nair, Micropor. Mesopor. Mater. 33 (1999) 1.
11] V.I. Parvulescu, H. Bonnemann, V. Parvulescu, B. Endruschat,
A.Ch.W. Rufinska, B. Tesche, G. Poncelet, Appl. Catal. A 214 (2001)
73.
12] J.A. Knowles, M.J. Hudson, J. Chem. Soc., Chem. Commun. (1995)
083.
13] M.J. Hudson, J.A. Knowles, J. Mater. Chem. 6 (1) (1996) 89.
14] G. Pacheco, E. Zhao, E.D. Valdes, A. Garcia, J.J. Fripiat, Micropor.
Mesopor. Mater. 32 (1999) 175.
15] J.L. Blin, R. Flamant, B.L. Su, Int. J. Inorg. Mater. 3 (2001) 959.
16] Y.Y. Huang, T.J. Mccarth, W.M.H. Sachtler, Appl. Catal. A 148 (1996)
35.
17] G. Larsen, E. Lotero, M. Nabity, L.M. Petkovic, D.S. Shobe, J. Catal.
64 (1996) 246.
18] E. Zhao, S.E. Hardcastle, G. Pacheco, A. Garcia, A.L. Blumenfeld, J.J.
Fripiat, Micropor. Mesopor. Mater. 31 (1999) 9.
2
[
2
[
[
[
[
1
[
[
[
1
[50] M.S. Wong, J.Y. Ying, Chem. Mater. 10 (1998) 2067.
[
[
[
51] D.M. Antonelli, Adv. Mater. 11 (1999) 487.
19] D.S. Sood, S.C. Sherman, A.V. Iretskii, J.C. Kenvin, D.A. Schiraldi,
M.G. White, J. Catal. 199 (2001) 149.
52] M.S. Wong, J.Y. Ying, Chem. Mater. 10 (1998) 2067.
53] J.A. Wang, M.A. Valenzuela, J. Salmones, A. Vazquez, A.G. Ruiz, X.
Bokhimi, Catal. Today 68 (2001) 21.
[20] R.D. Howells, J.D. Mc Cown, Chem. Rev. 77 (1977) 69.
21] A. Senning, Chem. Rev. 65 (1965) 385.
[