New heterogeneous catalysts for greener routes in the synthesis of fine chemicals

Article abstract of DOI:10.1016/j.jcat.2007.08.001

New strong Lewis acid SnTf-MCM-41 and SnTf-UVM-7 catalysts with unimodal and bimodal pore systems were prepared in a two-step synthesis in which the triflic acid (Tf) was incorporated to previously synthesized mesoporous tin-containing silicas. The Sn incorporation inside the pore walls was carried out through the Atrane method. The SnTf-UVM-7 catalysts were prepared by aggregating nanometric mesoporous particles defining a hierarchic textural-type additional pore system. Following these procedures, catalysts with different Si/Sn ratios-21.8 to 50.8 for SnTf-MCM-41 and 18.4 for SnTf-UVM-7-were prepared. These new materials were tested in the acylation of aromatic sulfonamides using acetic acid as the acylating agent and in the synthesis of (dl)-[α]-tocopherol through the condensation of 2,3,6-trimethylhydroquinone (TMHQ) with isophytol (IP). The activity data indicate that these heterogeneous catalysts are very active, corresponding to high yields in acylated compounds as 65.5% and very high selectivity to (dl)-[α]-tocopherol (94%, for a conversion of 98%).

Full text of DOI:10.1016/j.jcat.2007.08.001

Journal of Catalysis 251 (2007) 388–399
www.elsevier.com/locate/jcat
New heterogeneous catalysts for greener routes in the synthesis
of fine chemicals
Simona M. Coman a,∗, Georgeta Pop a, Cristina Stere a, Vasile I. Parvulescu a, Jamal El Haskouri b,
Daniel Beltrán b, Pedro Amorós b
a
Department of Chemical Technology and Catalysis, Faculty of Chemistry, University of Bucharest, Bdul Regina Elisabeta, 4-12, Bucharest 030016, Romania
b
Institut de Ciencia dels Material, Universitat de Valencia, Valencia E46071, Spain
Received 2 May 2007; revised 30 July 2007; accepted 1 August 2007
Available online 14 September 2007
Abstract
New strong Lewis acid SnTf-MCM-41 and SnTf-UVM-7 catalysts with unimodal and bimodal pore systems were prepared in a two-step
synthesis in which the triflic acid (Tf) was incorporated to previously synthesized mesoporous tin-containing silicas. The Sn incorporation inside
the pore walls was carried out through the Atrane method. The SnTf-UVM-7 catalysts were prepared by aggregating nanometric mesoporous
particles defining a hierarchic textural-type additional pore system. Following these procedures, catalysts with different Si/Sn ratios—21.8 to
50.8 for SnTf-MCM-41 and 18.4 for SnTf-UVM-7—were prepared. These new materials were tested in the acylation of aromatic sulfonamides
using acetic acid as the acylating agent and in the synthesis of (dl)-[α]-tocopherol through the condensation of 2,3,6-trimethylhydroquinone
(TMHQ) with isophytol (IP). The activity data indicate that these heterogeneous catalysts are very active, corresponding to high yields in acylated
compounds as 65.5% and very high selectivity to (dl)-[α]-tocopherol (94%, for a conversion of 98%).
© 2007 Elsevier Inc. All rights reserved.
Keywords: Heterogeneous catalysis; Mesoporous structures; Tin triflate; Acylated sulfonamides; (dl)-[α]-Tocopherol
1. Introduction
antibacterial agents [2], and prostaglandin F1α sulfonamides for
the potential treatment of osteoporosis, incorporate the N-acyl
sulfonamide moiety [3]. N-acyl sulfonamide synthesis usually
starts with the coupling of the parent sulfonamides with acid
chlorides or carboxylic anhydrides using trialkylamines, pyri-
dine [2–5], or alkali hydroxides [6–8]. Unfortunately, with no
exception, all of these methods lead to substantial waste prod-
ucts; for example, acid chlorides generate hydrochloric acid as
byproduct, whereas the carboxylic anhydrides use only 50%
of the molecule, achieving low atom economy. Less common
reports mentioning this transformation under acidic conditions
(Brönsted or Lewis acids) have not systematically examining
the purpose and limitations of the reaction [9].
The use of carboxylic acids as acylating agents on alcohol or
aromatics is rare [10–15] and usually associated with the pres-
ence of hypernucleophilic species (DCC, EDC, or 4-DMAP)
(see Aldrich Chemical Company Technical Bulletin AL114).
Otherwise, in the case of sulfonamide acylation, besides the
above-described synthesis methodologies, routes using carbox-
Today, there are significant demands by the pharmaceuti-
cal industry for new and efficient catalysts. Moreover, together
with traditional criteria, such as good activity and selectivity,
these new catalysts must include additional requisites, such as
low cost and environmental friendliness. Whereas the cost of
the process (including catalysts) is not a determinant constraint
for specific high-added-value drugs, produced at low scale, it
constitutes a first-order parameter when considering generic
drugs for diseases affecting a broad population sector or a para-
pharmaceutical product, such as food complement and/or ingre-
dient or a cosmetic, due to the high production costs involved.
Some newly developed drugs, including therapeutic agents
for Alzheimer’s disease [1], inhibitors of tRNA synthetases as
*
Corresponding author.
E-mail addresses: s.coman@chem.unibuc.ro, simona_cmn@yahoo.com
(S.M. Coman).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.08.001

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
389
ylic acids as a source of carbocation in direct-coupling ho-
mogeneous and heterogeneous catalytic processes have been
described [16–19]. These latter approximations, which have in
common the presence of nucleophilic activators, overcome the
undesirable wastes (because water is the main byproduct), and
the entire carboxylate amount is reactive.
hexadecyltrimethylammonium bromide [CTABr], SnCl2, triflic
acid, NaOH, HCl, and ethanol) were analytically pure and were
used as received from Aldrich. The unimodal (SnTf-MCM-41)
and bimodal (SnTf-UVM-7) porous catalysts were prepared
in a two-step synthesis in which the triflic acid was incorpo-
rated into previously synthesized mesoporous tin-containing
silicas. The Sn incorporation inside the pore walls was carried
out using the Atrane method [27], which enables preparation
of a unimodal or bimodal porous mixed oxide working under
strong (pH 11 to 12) or moderate (pH 8 to 9) basic conditions,
respectively. The resulting tin-doped silicas are designated Sn-
MCM-41 and Sn-UVM-7, respectively.
On the other hand, the synthesis of (dl)-[α]-tocopherol
is of great importance for the pharmaceutical industry and
also in the area of functional foods. In classical synthesis,
this is done through the acid-catalyzed condensation of 2,3,6-
trimethylhydroquinone (TMHQ) with isophytol (IP) using both
Brønsted and Lewis acid catalysts. Corrosion caused by the
acidic media, contamination of the wastewater with acids and
zinc ions, and the difficult purification of (dl)-[α]-tocopherol
by distillation under high vacuum at 200 ◦C are the main prob-
lems hindering industrial-scale synthesis [20]. Recently, data
concerning the use of heterogeneous catalysts, the main merits
of which are the simple separation of the solid catalyst from
the reaction mass, the absence of washwater containing the
dead catalyst, and a high purity of α-tocopherol, were published
[21–23]. Unfortunately, high reaction temperatures and long re-
action times are required, leading in many cases to diminished
yields as a result of byproduct formation. Recently, some metal
derivatives of triflic acid have been suggested as a cleaner al-
ternative to using metal chlorides as Lewis acid and versatile
catalysts in organic synthesis [24–26], avoiding such environ-
mental problems as large amounts of waste mineral acids and
zinc or aluminum residues, as well as the use of harmful organic
solvents in industrial-scale processes.
Keeping this background in mind, and with the aim of de-
signing a greener approach to the above set of reactions, we
have prepared new surfactant-assisted mesoporous heteroge-
neous Sn triflate–silica catalysts. These new family of catalysts
combine the high hydrophilic surface and accessible pore sys-
tem typical of mesoporous silicas with the presence of well-
dispersed and anchored tin triflate species able to act as strong
acid Lewis catalytic sites. The system architecture has been
built up, taking into account the described activity of metal
triflate complexes and the compatibility of Sn(IV) and Si(IV)
centers in silica networks. Thus we anticipate that our mate-
rials will combine the strong Lewis acidity of anchored tin
triflate with the water-sink capability of the hydrophilic silica
network, becoming efficient coupling-dehydrating agents or ac-
tive Friedel–Crafts solid catalysts.
2.1.1. Synthesis of Sn-MCM-41
In a typical synthesis, the molar ratio of the reagents in the
mother liquor was adjusted to 2 − x Si:x Sn:7 TEAH3:0.5
NaOH:0.52 CTABr:180 H2O. For example, the Si/Sn = 37
mesoporous material (x = 0.05) (working pH = 11) was ob-
tained as follows: 0.5 g (0.0125 mol) of NaOH was dissolved at
60 ◦C in 23 mL (0.172 mol) of TEAH3, and after a few minutes,
10.7 mL (0.0478 mol) of TEOS and 0.36 g (0.0016 mol) of
SnCl2 were added while stirring, and the mixture was heated at
130 ◦C for 5 min. The resulting solution was cooled to 110 ◦C,
and 4.68 g (0.0128 mol) of CTABr was added under stirring.
Then 80 mL (4.44 mol) of water was added under vigorous
stirring at a mixing temperature of 60 ◦C; shortly, a white sus-
pension appeared. This mixture was aged at room temperature
for 24 h. The resulting mesostructured powder was filtered
off, washed with water and ethanol, and air-dried. Finally, to
open the mesopore system, the surfactant was extracted from
the as-synthesized solid using an acetic acid/ethanol solution
(CTMA+/H+ exchange). Here ca. 1 g of mesostructured pow-
der was suspended in a solution containing 8 mL of CH3COOH
(80%) and 120 mL of ethanol (99%), and this mixture was
heated at reflux (60 ◦C) for 2 h under stirring. Later, after re-
newing the CH3COOH/ethanol solution, and to complete the
extraction process, the suspension was reheated at 60 ◦C for
16 h under stirring. The resulting (mesoporous) powder was
collected by filtration, washed with ethanol, and air-dried.
2.1.2. Synthesis of Sn-UVM-7
This mesophase was prepared almost exactly as for Sn-
MCM-41, starting from silatrane and stannatrane complexes
and working in a TEAH3-rich medium (i.e., using the Atrane
route) and in the absence of NaOH (with an apparent work-
ing pH of ca. 9.3). Thus the molar ratio of the reagents in the
mother liquor was adjusted to 2 − x Si:x Sn:7 TEAH3:0.52
CTABr:180 H2O (x = 0.1). In a typical synthesis to obtain
the Si/Sn = 26 bimodal porous material, a mixture of TEOS
(10.5 mL; 0.047 mol), SnCl2 (0.56 g, 0.0024 mol), and TEAH3
(23 mL, 0.172 mol) was heated at 150 ◦C for 10 min to prepare
Atrane complexes. The resulting solution was cooled to 110 ◦C,
and 4.68 g of CTABr (0.0128 mol) was added. Then 80 mL of
water was slowly added under vigorous stirring at 80 ◦C. After
a few minutes, the resulting suspension was aged at room tem-
perature for 4 h. The resulting mesostructured solid was then
separated by centrifugation, washed with water and ethanol,
Here we report that these new materials are effectively able
to facilitate the green acylation of aromatic sulfonamides us-
ing acetic acid as an acylating agent. These catalysts are also
highly active in the synthesis of (dl)-[α]-tocopherol through the
condensation of 2,3,6-trimethylhydroquinone (TMHQ) with is-
ophytol (IP).
2. Experimental
2.1. Catalyst preparation
All of the synthesis reagents (tetraethyl orthosilicate [TEOS],
triethanolamine [N(CH2–CH2–OH)3, hereinafter TEAH3],

390
S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
and air-dried. To obtain the final catalyst, the surfactant was
chemically extracted as described earlier for Sn-MCM-41. The
second step corresponds to the formation of Sn-triflate com-
plexes at the Sn-MCM-41 or the Sn-UVM-7 surface.
H2O = 40:60; flow rate, 0.5 mL/min; wavelength, 235 nm; vol-
ume sample, 15 µL. The products were characterized by HPLC-
MS and 1H NMR techniques.
For comparison, the aforementioned sulfonamides were acy-
lated in the presence of pure Sn(OTf)2 (10 mg). After reaction,
10 mL of saturated aqueous NaHCO3 solution was added, and
the products were extracted with Et2O (20 mL × 2). The or-
ganic extract was dried over anhydrous Na2SO4 a, after which
the solvent was separated from reaction product under vacuum
at 80 ◦C.
For synthesis of (dl)-[α]-tocopherol, in a typical procedure,
152 mg of TMHQ was dissolved in 10 mL of solvent (ace-
tonitrile, hexane, toluene, DMC, or hexane:DMC [50:50]) in
a two-necked flask. Then 50 mg of catalyst was added to this
mixture. Once the temperature reached 100 ◦C, 0.4 mL of IP
(TMHQ:IP molar ratio = 1:1) was added dropwise under stir-
ring. The reaction was carried out under reflux conditions for
1 h. After the reaction was stopped, the catalyst was filtered
off and the product separated from solvent under vacuum at
80 ◦C, as light-yellow–brown oil. Yields and purity were deter-
mined by HPLC analysis under following conditions: column,
EC 125/4.6 NUCLEOSIL 120-5 C18; eluent, acetonitrile; flow
rate, 0.8 mL/min; wavelength, 280 nm; volume sample, 15 µL.
The compounds were identified with the aid of pure samples
(standards).
For comparison, the synthesis of (dl)-[α]-tocopherol was
done in the presence of pure Sn(OTf)2 (10 mg), using acetoni-
trile as the solvent. After reaction, saturated aqueous NaHCO3
(10 mL) solution was added and the products were extracted
with hexane (20 mL × 2). The organic extract was dried over
anhydrous Na2SO4, after which the solvent was separated
from reaction product under vacuum, at 80 ◦C, as light-yellow–
brown oil.
2.1.3. Preparation of the SnTf-MCM-41 and SnTf-UVM-7
catalysts
The catalysts were prepared by heating the tin-doped meso-
porous silicas in methanolic solutions of triflic acid by the fol-
lowing procedure. A suspension of Sn-MCM-41 or Sn-UVM-7
chemically extracted porous silicas (2 g) was stirred under re-
flux in a solution of triflic acid (5 g) in methanol (50 mL) for
15 h. The resulting porous samples were collected by filtration,
washed with cold methanol to eliminate any excess triflic acid
on the silica surface, and air-dried at 80 ◦C. The obtained sam-
ples were designated SnTf-MCM-41 and SnTf-UVM-7.
2.2. Catalyst characterization
All solids were analysed for Sn, Si, and S by electron probe
microanalysis (EPMA) using a Philips SEM-515 instrument. X-
ray powder diffraction (XRD) data were recorded on a Seifert
3000TT θ–θ diffractometer using CuKα radiation. Patterns
were collected in steps of 0.02◦ (2θ) over the angular range
1◦–10◦ (2θ) for 25 s per step. To detect the presence of some
crystalline bulk phase, additional patterns were recorded with a
larger scanning step [0.05◦ (2θ)] over the angular range 10◦–
60◦ (2θ) for 10 s per step. Transmission electron microscopy
(TEM) was carried out with a Philips CM-10 electron micro-
scope operating at 100 kV. Samples were ground gently in do-
decane, and microparticles were deposited on a holey carbon
film supported on a Cu grid. Surface area and pore size values
were calculated from nitrogen adsorption–desorption isotherms
(−196 ◦C) recorded on a Micromeritics ASAP-2010 automated
instrument. Calcined samples were degassed for 15 h at 130 ◦C
and 10−6 Torr before analysis. Surface areas were estimated
according to the BET model, and pore size dimensions were
calculated using the BJH method. Room-temperature diffuse
reflectance spectra were recorded (200–800 nm) using a Shi-
madzu UV–vis 2501PC spectrophotometer. FTIR spectra were
collected on a Nicolet 4700 spectrometer (200 scans with a res-
olution of 4 cm−1) using self disks of 1% sample in KBr.
Recycling tests were carried out as well. After each reac-
tion ended, the catalyst was filtered off, rinsed with the solvent,
and dried under ambient conditions for 2 h. Each catalyst was
reused in at least four catalytic cycles. Washed catalysts were
analyzed by ICP-AES and FTIR.
3. Results and discussion
3.1. Catalyst characterization
2.3. Catalytic tests: Acylation of sulfonamides
As shown in Section 2, the catalysts were prepared in a two-
step synthesis in which the triflic acid was incorporated into
previously synthesized mesoporous tin-containing silicas. In-
corporation of Sn inside the pore walls was done using the
Atrane method [27]. The triflate ligands were included follow-
ing a protocol adapted from the synthesis method described
in the bibliography for tin–triflate complexes in solution. The
selection of the pH conditions of the first step allowed us to
prepare two different catalyst types with a unimodal or bimodal
pore system. At high pH values, unimodal Sn-MCM-41, in
the form of large micrometric particles, was obtained; work-
ing under moderately basic conditions, Sn-UVM-7 materials
(a nanometric version of the typical M41 solids) were isolated.
These latter catalysts were constructed by aggregating nano-
In a typical procedure, 1.25 mmol of sulfonamide (benzene-
sulfonamide, p-nitrobenzenesulfonamide, p-methoxybenzene-
sulfonamide) were dissolved in 4 mL of acetonitrile in a 10-mL
glass vessel. To this mixture, 3.75 mmol of acylating agent
(acetic acid) and 15 mg of catalyst were added (acylating
agent:substrate ratio = 3:1). All reactions were carried out at
80 ◦C for 18 h under vigorous stirring. After the reaction was
stopped, the catalyst was filtered off and the product separated
from solvent under vacuum at 80 ◦C as a white crystalline solid.
The product was redissolved in the high-pressure liquid chro-
matography (HPLC) eluent and analyzed under the following
conditions: column, GROM-SIL 80 ODS-2 FE; eluent, MeOH,

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
391
Fig. 2. Low-angle XRD patterns of (a) Sn-UVM-7 and (b) SnTf-UVM-7 cata-
lysts.
Fig. 1. Low-angle XRD patterns of (a) Sn-MCM-41 and (b) SnTf-MCM-41
catalysts.
nal cell in the case of the Sn-UVM-7 derivative (first step) that
practically disappeared after reaction with triflic acid (second
step). In both series (MCM-41 and UVM-7 like), the peak po-
sitions remained practically unchanged. This finding confirms
that the mesostructures were not altered after reaction with tri-
flic acid/methanol and thus supports our selection of solvent. In
fact, a significant loss of order (and consequently mesostructure
degradation) was observed in preliminary syntheses carried out
in acetonitrile as solvent. On the other hand, the second reac-
tion step affected the signal intensities only slightly. The small
loss of X-ray intensity observed in the XRD corresponding to
SnTf-MCM-41 and SnTf-UVM-7 (compared with Sn-MCM-
41 and Sn-UVM-7, respectively) must be attributed to a certain
phase-cancellation phenomena associated to the introduction
of scattering material (triflate anions) into the pores, although
the possibility of a slight decrease in symmetry cannot be dis-
carded. In all samples, the absence of XRD peaks at high angle
values indicates that no phase segregation leading to SnO or
SnO2 bulk oxides occurred.
The TEM images (Figs. 3a, 3b) were completely corre-
lated with XRD observations. The observation in both cases
of a dominant (single-type) particle morphology supports the
monophasic nature of the catalysts. Typical morphologies—
hexagonal ordered large particles and aggregates of disor-
dered nanoparticles—were observed for the SnTf-MCM-41 and
SnTf-UVM-7 catalysts. Moreover, TEM images of both cata-
lysts also showed small dark spots (average size ca. 2 nm) that
could be due to Sn-rich domains.
metric mesoporous particles defining a hierarchic textural-type
additional pore system (UVM-7-like materials). The resulted
catalysts are designated SnTf(22)-MCM-41 (Si/Sn = 21.81),
SnTf(47)-MCM-41 (Si/Sn = 47.32), SnTf(51)-MCM-41 (Si/Sn
= 50.84), and SnTf(18)-UVM-7 (Si/Sn = 18.40).
The symmetry and order degree of the catalysts before and
after reaction with triflic acid were studied by XRD and TEM.
XRD provides information only on the intraparticle mesopore
relative organization (i.e., the surfactant templating pore sys-
tem). Fig. 1 shows the XRD patterns corresponding to Sn-
MCM-41 and SnTf-MCM-41 samples. Both materials display
XRD patterns in the low-angle region with one strong diffrac-
tion peak, usually associated with the (100) reflection when a
hexagonal cell (similar to that described in the case of MCM-
41) is assumed, along with two other resolved small signals that
can be indexed to the (110) and (200) reflections. These features
indicate that the ordered hexagonal mesopore symmetry typical
of pure MCM-41 silica was maintained after the Sn incorpo-
ration (first step) and even after the reaction with triflic acid
(second step).
In a similar way, the disordered intraparticle porosity typi-
cal of UVM-7 materials was preserved even after incorporation
of Sn as a triflate (Fig. 2). Both solids showed XRD patterns
with an intense peak in the low-angle domains, ascribed to the
(100) reflection. Apart from the intense peak at low 2θ values,
a broad signal of relatively low intensity can be indexed to the
overlapped (110) and (200) reflections of the typical hexago-

392
S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
Fig. 5. N adsorption–desorption isotherms of (a) Sn-UVM-7 and (b) SnTf-
2
UVM-7 catalysts.
templated mesopores. Moreover, the isotherms of UVM-7-like
catalysts showed a second adsorption step at high relative pres-
sure values (P/P0 > 0.8), associated with filling of the large
interparticle cage-like pores. In both series, the unimodal or
bimodal porous topology together with a high BET surface
area was maintained before and after reaction with triflic acid
(Table 1). The pore size after this second reaction step was prac-
tically unaltered, with only a slight decrease of both parameters
observed. This evolution supports the insertion of triflate anions
inside the pores.
Fig. 6 shows more evidence on the formation of the Sn
triflate species. FTIR spectra indicate that the MCM-41 and
UVM-7 supports did not retain even a trace of triflate. In con-
trast, on doped Sn-silica samples, triflic acid was incorporated
in amounts that increased with increasing Sn content. No an-
choring occurred on the silica wall, but anchoring was obvious
on the Sn sites and on the Sn silica. The results of chemical
analysis and TGA of these samples were consistent with FTIR
results and with the reported S/Si ratios.
Fig. 3. TEM images of (a) SnTf-MCM-41 [the insets show enlarged TEM im-
ages showing the ordered pore array (×5) and the rich Sn domains (×3)] and
(b) SnTf-UVM-7 catalysts [the inset shows an enlarged (×3) TEM image].
A summary of the textural and chemical characteristics of
the prepared catalysts is given in Table 1.
3.2. Catalytic behavior in acylation of various sulfonamides
with acetic acid
The aforementioned catalysts were tested in the acylation of
benzenesulfonamide (1a, R = H), p-nitrobenzenesulfonamide
(1b, R = NO2), and p-methoxybenzenesulfonamide (1c, R =
MeO) with acetic acid (Scheme 1). The only byproduct in this
case was water. Data presented in Table 2 confirm that in the
presence of these catalysts, the acylation of the above substrates
with acetic acid, which usually requires high temperatures (nec-
essary to activate the acylating agent), occurred with yields
from moderate to good even at temperatures as low as 80 ◦C.
Of note, the selectivity was 100% irrespective of the structure
of the substrate or the nature of the catalyst. On the other hand,
the fact that for Sn-MCM and Sn-UVM, neither triflic acid-
impregnated support had any activity in this reaction provides
additional evidence that the active site was associated with tin
triflate species.
Fig. 4. N adsorption–desorption isotherms of (a) Sn-MCM-41 and (b) SnTf-
2
MCM-41 catalysts.
The unimodal and bimodal porous character of MCM-
41 and UVM-7-like catalysts was further confirmed by N2
adsorption–desorption isotherms (Figs. 4 and 5). All samples
presented a clear and sharp adsorption step at intermediate rel-
ative pressure values (0.15 < P/P0 < 0.4) due to the capillary
condensation of N2 inside the intra-nanoparticle surfactant-

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
393
Table 1
The textural and chemical characteristics of the SnTf-MCM-41 and SnTf-UVM-7 samples
a
a
b
2
c
c
Sample
Si/Sn
Sn/S
d
100
(nm)
S
BET
(m /g)
Small pore (nm)
Large pore (nm)
1
2
3
4
5
6
Sn-MCM-41
38.0
21.8
47.3
50.8
19.0
18.4
–
4.0
4.1
3.7
3.7
4.8
4.5
937
725
878
812
820
646
2.5
2.5
2.6
3.1
2.4
2.1
–
–
–
–
32.3
36.6
SnTf(22)-MCM-41
SnTf(47)-MCM-41
SnTf(51)-MCM-41
Sn-UVM-7
2.3
1.4
1.3
–
SnTf(18)-UVM-7
1.1
a
b
c
Si/Sn and Sn/S molar ratio estimated from electron probe microanalysis (EPMA).
Cell parameter calculated assuming a hexagonal cell (XRD).
Textural characteristics from the adsorption–desorption isotherms of liquid N at 77 K.
2
Table 2
The N-acyl sulfonamides yield and TOF as a function of the substrate and the catalyst nature
−1
)
Catalyst
Substrate
Yield (%)
TOF (h
1
2
3
4
5
6
7
8
9
SnTf(22)-MCM-41
SnTf(22)-MCM-41
SnTf(47)-MCM-41
SnTf(47)-MCM-41
SnTf(51)-MCM-41
SnTf(51)-MCM-41
SnTf(18)-UVM-7
SnTf(18)-UVM-7
Benzenesulfonamide
p-Nitrobenzenesulfonamide
Benzenesulfonamide
p-Nitrobenzenesulfonamide
Benzenesulfonamide
p-Nitrobenzenesulfonamide
Benzenesulfonamide
p-Nitrobenzenesulfonamide
Benzenesulfonamide
p-Nitrobenzenesulfonamide
30.0
17.4
45.2
11.4
48.6
11.7
65.5
24.2
76.0
26.1
8.9
2.8
9.1
2.9
9.3
2.3
4.6
1.7
9.9
3.4
a
a
Sn(OTf)
Sn(OTf)
2
2
10
◦
Note. Conditions: substrate (1.25 mmol), acetic acid (3.75 mmol), catalyst: (15 mg), solvent (4 mL of acetonitrile), reaction temperature (80 C), reaction time
(18 h).
a
Catalyst: (10 mg), reaction time (4 h).
Fig. 6. Comparative FTIR spectra of UVM-7, Sn(18)-UVM-7 and SnTf(18)-UVM-7 samples.
Scheme 1. The acylation of aromatic sulfonamides.

394
S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
The nucleophilic substrates reaction rate with acylium
cations would be expected to increase as the induced electronic
density on C4 and thus the N basicity increase. Nevertheless,
the observed rate of the sulfonamide acylation followed the
sequence benzenesulfonamide > p-nitrobenzenesulfonamide
> p-methoxybenzenesulfonamide and was very sensitive to
the nature of the aromatic hydrogen substituent. The selectiv-
ity in acylated p-methoxybenzenesulfonamide did not exceed
7% irrespective of the catalyst nature. This corresponds to
an approximate relative yield of benzenesulfonamide:p-nitro-
benzenesulfonamide:p-methoxybenzenesulfonamide of 10:4:1
(Table 2). This result contrasts with that obtained from Brön-
sted acid (i.e., sulfuric)-catalyzed homogeneous acylation of the
same substrate set using acetic anhydride in the same solvent
and conditions, where yield (98%, 95%, 91% for R = NO2, H,
OMe, respectively) did not significantly depend on the nucle-
ophile’s substituent nature [18]. The catalysts nature is obvi-
ously responsible for the behavior observed in our system.
The best yield (65.5%; entry 7, Table 2) to acylated sul-
fonamides 2 was obtained in the presence of the SnTf(18)-
UVM-7 catalyst (Si/Sn ratio = 18.40). In heterogeneous
acidic catalysis, isolated species usually are much more ac-
tive than oligomeric and polymeric species. In conventional
co-hydrolysis methods, good dispersion of monomeric Sn in the
silica network is not easily achieved. The hydrolysis rate of tin
(IV) is much higher than that of silicium, with subsequent low
chemical homogeneity and even phase segregation. Therefore,
the Atrane route to Sn-MCM-41 synthesis provides an attrac-
tive solution, favoring the harmonic heterometallic hydrolysis–
condensation processes. With this preparation method, high
concentrations of well-dispersed Sn atoms into the framework
of the mesoporous silica can be attained. Regardless of the
support type (MCM-41 or UVM-7), UV–vis characterization
suggested the presence of highly dispersed tetrahedral isolated
Sn4+ with an absorption band centered at 250 nm, ascribed to
charge transition from O2− to Sn4+ in a tetrahedral coordina-
tion environment [28]. In the SnTf(18)-UVM-7 sample, these
Sn sites seemed to be the dominant Sn4+ species. However,
in the SnTf-MCM-41 catalysts, the additional observed band
(centered at ca. 289 nm, assigned to charge transitions from
O2− to Sn4+ in octahedral environments with a certain degree
of polymerization [29]) could be associated with the existence
of small casiterite-like nanodomains (Fig. 7) with a subsequent
decrease in active sites from ca. 30 to 50% for low and high
Sn-containing MCM-41 samples, respectively. This tendency
to form Sn–O–Sn bonds is due to the higher pH values used
for preparing MCM-41 materials compared with that used for
preparing UVM-7. Moreover, chemical analysis of the reaction
products clearly indicated no leaching of the active species.
The presence of Sn4+ sites was also suggested by XPS mea-
surements. The values measured for the Sn 3d5/2 species were
closer to those of Sn4+ than those of Sn2+ [30,31].
Fig. 7. Diffuse reflectance UV–vis spectra of (a) SnTf-MCM-41 and (b) SnTf-
UVM-7 catalysts.
Scheme 2. The model of the active site in incorporated Sn-triflate catalysts.
the Si/Sn ratio. The increased concentration of the Sny(O,OH)z
species exhibited a negative effect on the yield. These data also
demonstrated that the network in which Sn is incorporated is
very important, making it more or less accessible to complexa-
tion by triflate. Although the concentration of Sn increased from
SnTf(51) to SnTf(22), neither the evolution of TOF nor the con-
centration of Sn(OTf)x followed the same trend, demonstrating
the contribution of isolated Sn(OTf)x sites in this reaction.
The additional experiments carried out in homogeneous con-
ditions using pure Sn(OTf)2 (entries 9 and 10, Table 2) con-
firmed the foregoing conclusion about the contribution of the
isolated Sn(OTf)x sites. The TOFs measured for this catalyst
were slightly higher than those found for heterogeneous cata-
lysts; however, the differences were enough small to indicate
that the homogeneous system is more active.
The observed tendency toward formation of Sn–O–Sn bonds
and the small amount of triflate ligands in the SnTf(22)-MCM-
41 dropped the proportion of active Sn(OTf)x sites compared
with SnTf(51)-MCM-41, in accordance with the catalytic re-
sults. Thus, the pH value at the first reaction step can be
considered the key to controlling the concentration of active
sites, avoiding basic high-coordinated Sny(O,OH)z species or
SnO2 segregation. Under moderate basic conditions, the cross-
condensation among Si and Sn species is favored, and SnTf-
UVM-7 catalysts must have significantly higher amounts of
active sites than SnTf(22)-MCM-41 with similar Si/Sn ratios.
Moreover, the low Sn/S = 1.1 molar ratio in UVM-7 catalysts
warrants that all tin surface sites exist as tin-triflate centers (Ta-
Scheme 2 proposes a model of the active sites in the inves-
tigated catalysts. TOF values given in Table 2 and their cor-
relation with the molar concentration of the highly dispersed
tin triflate (Sn(OTf)x) are in line with this model. Fig. 8 shows
the variation of TOF with the content of Sn(OTf)x species and

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
395
Fig. 8. The variation of the yield to acyl-benzenesulfonamide (a) and TOF (b) with the content of Sn(OTf) species and Si/Sn ratio in the acylation of benzenesul-
x
fonamide with acetic acid (Si/Sn = 22.0, 47.0 and 51.0 for MCM structure, Si/Sn = 18.0 for UVM structure).
ble 1) with a Sn/triflate proportion of between 1 and 2. The
superacid active site of the catalyst would be the anchored
Sn(OTf)x species, and the silica network would be the dehy-
drating agent. Recycling experiments revealed no deactivation
in this reaction.
ca. 60%. This can possibly explained by the carbocation stabi-
lization in polar solvents by strong ion dipole interactions.
We observe that every tested material exhibited excellent
conversion rates regardless of the Si/Sn or Sn/S ratio. Nev-
ertheless, when referring to reaction selectivity, the behavior
differed; as it occurred in the case of sulfonamide acylation,
well-dispersed Sn(OTf)x species seemed to affect the catalytic
performance. The selectivity to (dl)-[α]-tocopherol was high
for catalysts with a low oligomeric tin species content (Fig. 9a).
Thus, optimum results were obtained for SnTf(51)-MCM-41,
whereas SnTf(22)-MCM-41 exhibited very poor selectivity. (In
fact, this latter material would be considered a selective catalyst
for the benzofurane derivative.) Again, SnTf(18)-UVM-7 was
more selective to (dl)-[α]-tocopherol than SnTf(22)-MCM-41
irrespective of the nature of the solvent and demonstrated sim-
ilar (but somewhat poorer) performance than SnTf(51)-MCM-
41 (see Table 4). The tin-triflated UVM-7 silicas were not op-
timized; SnTf(18)-UVM-7 gave reasonable results, but a lower
Si/Sn ratio would be expected to give this material better selec-
tivity in tocopherol synthesis.
3.3. Catalytic behavior in the condensation of the
2,3,6-trimethylhydroquinone with isophytol
To extend the scope of application of these catalysts, the
condensation of the 2,3,6-trimethylhydroquinone (TMHQ, 3)
with isophytol (IP, 4) for the synthesis of (dl)-[α]-tocopherol (5)
was attempted (Scheme 3). This synthesis occurs in two steps:
(i) chemoselective acid-catalyzed Friedel–Crafts alkylation of
trimethylhydroquinone (TMHQ) with isophytol, and (ii) chro-
mane ring closure. The nature of the solvent is very important
for this reaction (Table 3). The best solvent seems to be a non-
polar solvent (e.g., hexane, ε = 2.02). As the polarity of the
solvent increased, the catalytic performances became poorer;
thus in acetonitrile (ε = 36.6), the conversion of TMHQ was

396
S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
Scheme 3. The synthesis of (dl)-[α]-tocopherol 5 by the condensation of TMHQ 3 and IP 4.
Table 3
The TMHQ conversion, TOF and (dl)-[α]-tocopherol selectivity as a function of the solvent nature
Catalyst/
solvent
SnTf(22)-MCM-41
SnTf(18)-UVM-7
Conversion
(%)
TOF
(h
Selectivity
(%)
Conversion
(%)
TOF
(h
Selectivity
(%)
−1
−1
)
)
Hexane
Toluene
DMC
99.8
93.1
90.3
58.5
69.8
65.1
63.1
40.9
13.0
19.2
15.0
10.3
100
30.6
29.8
28.3
18.9
76.5
46.7
20.5
67.9
97.3
92.7
61.8
Acetonitrile
◦
Note. Conditions: TMHQ (152 mg), IP (0.4 mL), solvent (10 mL), catalyst (50 mg), reaction temperature (100 C), time (1 h); Conversion and selectivity are based
on TMHQ by HPLC-analysis of isolated material.
Fig. 9 shows the way in which the selectivity (Fig. 9a) and
TOF (Fig. 9b) correlate with the content of Sny(O,OH)z and
Sn(OTf)x species and the Si/Sn ratio. As for the acylation of
sulfonamides for MCM-41-type catalysts, even though the Sn
concentration increased from SnTf(51) to SnTf(22), neither
the evolution of TOF nor the concentration of Sn(OTf)x fol-
lowed the same trend, thus demonstrating the effect of highly
dispersed Sn(OTf)x sites in this reaction. On the other hand,
Fig. 9a shows a negative effect of low-coordinated Sny(O,OH)z
species in this reaction. As expected, the increased concen-
tration of these species caused a decrease in the selectivity to
(dl)-[α]-tocopherol.
Table 4 also presents data collected for the Sn(OTf)2 cata-
lyst (entry 5). As for the acylation of sulfonamides, TOF was
in the same range as that of the most active MCM-41-type cat-
alysts, providing another argument for the activity of dispersed
tin-triflate species. The selectivity to (dl)-[α]-tocopherol serves
as more evidence in this sense.
[32] seems to be the key step in successful synthesis. Specially,
the benzofuran derivative/(dl)-[α]-tocopherol ratio is a function
of the relative stability of i-phytil and n-phytil cations. It is
very well known that the stability of the carbenium ions fol-
lows the sequence primary < secondary < tertiary, whereas
their reactivity follows the sequence primary > secondary >
tertiary. As a result, the benzofurane derivative is the principal
byproduct due to i-phytil–n-phytil cation isomerization. More-
over, phytadienes readily undergo polymerization and condense
with TMHQ to form mixtures of “isomeric” tocopherols that
are heterocyclic in nature. Then the benzofurane derivatives
are not the only byproduct in this reaction. It is well known
that alcohols, preferably tertiary allylic alcohols like 4, readily
dehydrate in the presence of acids [33]. Others have reported
that formation of some other intermediates and byproducts is
also possible [34,35]. The solvent itself also can be a source of
byproducts.
The ground-state activated electrophile (isophyl species)
forms relatively tight ion pairs with the Sn(OTf)x species and
may show a considerable degree of covalent bonding. How-
ever, basic Sny(O,OH)z oligomers present on the silica surface
will compete with monomers as carbocation-bonding centers.
The i-phytil and n-phytil cations, which have different charges
(and acidities), interact with all tin centers on the surface. The
strongest interaction, between n-phytil and basic Sny(O,OH)z
The mechanistic details of the alkylation step involves the
interaction of an alkylating agent (isophytol) with the protons
generated through the interaction of the acid catalyst with the
solvent to form an activated electrophile, which adds to the aro-
matic ring of TMHQ acting as a nucleophile, followed by pro-
ton elimination. The availability and proportion of the different
carbocations (i-phytil, n-phytil, and all isomeric phytadienes)

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
397
Fig. 9. The variation of the selectivity to tocopherol (a) and TOF (b) with the content of Sn(OTf) species and Si/Sn ratio in the synthesis of tocopherol (Si/Sn = 22.0,
x
47.0 and 51.0 for MCM structure, Si/Sn = 18.0 for UVM structure); after 50 min.
Table 4
The TMHQ conversion, TOF and (dl)-[α]-tocopherol selectivity as a function of the Si/Sn ratio
a
a
b
−1
b
Sample
Si/Sn
Sn/S
Conversion (%)
TOF (h
)
Selectivity (%)
1
2
3
4
5
SnTf(22)-MCM-41
SnTf(47)-MCM-41
SnTf(51)-MCM-41
SnTf(18)-UVM-7
21.8
47.3
50.8
18.0
–
2.3
1.4
1.3
1.1
–
99.8
99.0
98.0
100
69.8
86.1
84.5
30.6
83.4
13.0
25.2
94.0
76.5
95.2
c
Sn(OTf)
100
2
a
b
c
Si/Sn and Sn/S molar ratio estimated from electron probe microanalysis (EPMA).
Conditions: TMHQ (152 mg), IP (0.4 mL), solvent (hexane, 10 mL), catalyst (50 mg), reaction temperature (100 C), time (1 h).
Solvent (acetonitrile, 10 mL), catalyst (10 mg), reaction time (30 min). Conversion and selectivity are based on TMHQ by HPLC-analysis of isolated material.
◦
domains, will prevent the desired TMHQ alkylation. The tin-
rich catalysts (obtained at high pH values) will give the poorest
results.
Table 5 presents the results of recycling the SnTf(51)-MCM-
41 catalysts in the synthesis of (dl)-[α]-tocopherol. After four
cycles, the conversion of TMHQ was preserved at >80%. In
addition, the decrease in selectivity to (dl)-[α]-tocopherol was
<10% after four cycles. These data demonstrate that the investi-
gated catalysts are quite robust in a reaction for which previous
reports claimed rapid catalyst deactivation [36,37].
Fig. 10 shows typical FTIR spectra collected for the Sn-
Tf(51)-MCM catalyst tested in this reaction after its separation

398
S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
Fig. 10. The FTIR spectra collected for the SnTf(51)-MCM-41 sample (a) in time and (b) after the first recycling.
Table 5
with the solvent and recycled; the results were very similar to
those presented in Table 5 for the second recycle. In this way,
we also can also explain the decrease of the TOF in the second
cycle and the slight decrease of the selectivity (Table 5). Spectra
presented in Fig. 10b confirm this behavior.
Therefore, this new catalyst family is characterized by a
triple functionality: (i) the Lewis Sn(OTf)x activating cen-
ters, which generate carbocations; (ii) the basic blocking
Sny(O,OH)z oligomers, which account for the yield variations
observed in the sulfonamide acylation and, with respect to the
Friedel–Crafts alkylation, are able to modulate the isophyl equi-
librium affecting then the selectivity; and (iii) the high-surface
area silica as the necessary dehydrating support. This latter as-
pect is most important, because keeping water away allows
working at temperatures below the boiling point of water.
Synthesis of (dl)-[α]-tocopherol with reused, non-regenerated SnTf(51)-MCM-
41 catalyst
−1
Catalytic cycle
Conversion (%)
Selectivity (%)
TOF (h
)
1 (fresh catalyst)
98.0
90.0
81.4
80.0
94.0
93.6
90.1
86.0
84.5
77.6
70.2
70.0
2
3
4
Note. Conditions: TMHQ (152 mg), IP (0.4 mL), solvent (hexane, 10 mL),
catalyst (50 mg), reaction temperature (100 C), time (1 h); conversion and
selectivity are based on TMHQ by HPLC-analysis of isolated material.
◦
at different reaction times. Initially, the spectrum of the fresh
catalyst showed only bands assigned to support and triflate
species; however, after 30 min, new bands at 2900–2800 cm−1
,
assigned to –OCH2 and –CH2 groups, were detected. The in-
tensity of these bands reached a plateau after 1 h of reaction.
These bands were assigned to strongly chemisorbed IP and
byproduct molecules on the most active triflate species. Thus,
the decreased catalyst activity is likely related to blockage of
the very active sites by these compounds. To confirm this sup-
position, the catalyst recovered after 1 h of reaction was rinsed
4. Conclusion
Our data demonstrate that in the investigated SnTf-MCM-
41 and SnTf-UVM-7 catalysts, triflate is associated with well-
dispersed tin species. The synthesis of these catalysts was
achieved in two steps, with triflic acid incorporated into previ-

S.M. Coman et al. / Journal of Catalysis 251 (2007) 388–399
399
ously synthesized mesoporous tin-containing silicas for Sn con-
tent up to Si/Sn > 18. Both SnTf-MCM-41 and SnTf-UVM-7
are highly active green catalysts for acylation with acetic acid
and synthesis of (dl)-[α]-tocopherol.
[15] J.M. Escola, M.E. Davis, Appl. Catal. A Gen. 214 (2001) 111.
[16] P. Gomes, J.R.B. Gomes, M. Rodrigues, R. Moreira, Tetrahedron 59
(2003) 7473.
[17] A.R. Katrizky, S. Hoffmann, K. Suzuki, ARKIVOC 12 (2004) 14.
[18] M.T. Martin, F. Roschangar, F.F. Eaddy, Tetrahedron Lett. 44 (2003) 5461.
[19] I.E. Pop, B.P. Déprez, A.L. Tartar, J. Org. Chem. 62 (1997) 2594.
[20] T. Netscher, Chimica 50 (1996) 563.
Acknowledgments
[21] F. Schager, W. Bonrath, J. Catal. 1 (1999) 282.
[22] R.F. Childs, A. Mika-Gibala, J. Org. Chem. 47 (1982) 4207.
[23] A. Heidekum, M.A. Harmer, W.F. Hölderich, J. Catal. 176 (1998) 260.
[24] S. Luo, L. Zhu, A. Talukdar, G. Zhang, X. Mi, J.-P. Cheng, P.G. Wang,
Mini Rev. Org. Chem. 2 (2005) 546.
The authors thank the MEC (projects MAT2003-08568-
C03-01 and CTQ2006-15456-C04-03) for support. J.E.H. also
thanks the MEC for a Ramón & Cajal contract.
[25] S. Kobayashi, M. Sugira, H. Kitagawa, W.W.-L. Lam, Chem. Rev. 102
(2002) 2227.
References
[26] A. Corma, H. Garcia, Chem. Rev. 103 (2003) 4307.
[27] S. Cabrera, J. El Haskouri, C. Guillem, J. Latorre, A. Beltrán, D. Beltrán,
M.D. Marcos, P. Amorós, Solid State Sci. 2 (2000) 405.
[28] S. Samanta, N.K. Mal, A. Manna, A. Bhaumik, Appl. Catal. A Gen. 273
(2004) 157.
[29] Z.C. Liu, H.R. Chen, W.M. Huang, J.L. Gu, W.B. Bu, Z.L. Hua, J.L. Shi,
Microporous Mesoporous Mater. 89 (2006) 270.
[30] C.D. Wagner, J.F. Moulder, L.E. Davis, W.M. Riggs, Handbook of X-Ray
Photoelectron Spectroscopy, Perking–Elmer Corporation, Physical Elec-
tronics Division, 1993.
[31] D. Briggs, M.P. Seah, Practical Surface Analysis, vol. 1, Auger and X-
Ray Photoelectron Spectroscopy, second ed., Wiley/Salle & Sauerländer,
Chichester, New York, Brisbane, Toronto, Singapore/Aarau, Frankfurt/M,
Salzburg, 1990.
[32] J. March, Advanced Organic Chemistry, third ed., Wiley, New York, 1985,
p. 902.
[33] M.B. Tarabrin, G.G. Shestakov, R.P. Evstigneeva, E.Yu. Bulychev, Pharm.
Chem. J. 18 (1984) 572.
[34] S. Wang, W. Bonrath, H. Pauling, F. Kienzle, J. Supercrit. Fluids 17 (2000)
135.
[1] T. Hasegawa, H. Yamamoto, Bull. Chem. Soc. Jpn. 73 (2000) 423.
[2] M.G. Banwell, C.F. Crasto, C.J. Easton, A.K. Forrest, T. Karoli, D.R.
March, L. Mensah, M.R. Nairn, P.J. O’Hanlon, M.D. Oldham, W. Yue,
Bioorg. Med. Chem. Lett. 10 (2000) 2263.
[3] Y. Wang, D.L. Soper, M.J. Dirr, M.A. DeLong, B. De, J.A. Wos, Chem.
Pharm. Bull. 48 (2000) 1332.
[4] K. Kondo, E. Sekimoto, J. Nakao, Y. Murakami, Tetrahedron 56 (2000)
5843.
[5] K. Kondo, E. Sekimoto, K. Miki, Y. Murakami, J. Chem. Soc. Perkin
Trans. 1 (1998) 2973.
[6] N. Ishizuka, K.-I. Matsumura, K. Hayashi, K. Sakai, T. Yamamori, Syn-
thesis 6 (2000) 784.
[7] N. Ishizuka, K. Matsumura, Japanese Patent JP-10045705, 1998.
[8] T. Inoe, O. Myahara, A. Takahashi, Y. Nakamura, Japanese Patent JP-
08198840, 1996.
[9] Y. Morisawa, M. Kataoka, H. Negahori, T. Sakamoto, N. Kitano, K. Ku-
sano, J. Med. Chem. 23 (1980) 1376.
[10] M. Akçay, Appl. Catal. A Gen. 269 (2004) 157.
[11] C.L. Padró, C.R. Apesteguía, J. Catal. 226 (2004) 308.
[12] A.K. Pandey, A.P. Singh, Catal. Lett. 44 (1997) 129.
[13] I. Neves, F. Jayat, P. Magnoux, G. Pérot, F.R. Ribeiro, M. Gubelmann, M.
Guisnet, J. Mol. Catal. 93 (1994) 169.
[35] W. Bonrath, C. Dittel, L. Giraudi, T. Netscher, T. Pabst, Catal. Today 121
(2007) 65.
[36] M.C. Laufer, W. Bonrath, W.F. Hölderich, Catal. Lett. 100 (2005) 101.
[37] H. Wang, B.-Q. Xu, Appl. Catal. A Gen. 275 (2004) 247.
[14] V.D. Kumari, G. Saroja, A. Ratnamala, M. Noorjahan, M. Subrahmanyam,
React. Kinet. Catal. Lett. 79 (2003) 43.