Intramolecular cyclization of (+)-citronellal using supported 12-tungstophosphoric acid on MCM-41

Article abstract of DOI:10.1016/j.molcata.2012.03.002

Supported heteropolyacids (HPAs) have been used as catalysts by fine-tuning their heterogenous and Br?nsted acidic properties. In this work, supported H₃PW₁₂O₄₀ (HPW) on MCM-41 was prepared (2-40 wt%) and applied in the intramolecular cyclization of (+)-citronellal. The synthesized materials were characterized by ³¹P MAS NMR, XRD, FTIR, pyridine gaseous adsorption, low temperature nitrogen adsorption and thermal analysis. The impregnation of 20% HPW on the surface of MCM-41 decreased the characteristic crystallographic reflections of the support, suggesting that HPW modifies the long-range order. The results show that the HPW was deposited preferentially inside the mesopores of the support and affected the acid strength of the supported catalyst. FTIR studies of pyridine adsorption confirmed the presence of Br?nsted and hydrogen bonded acid sites. All synthesized materials were active in the cyclization reaction, but the 20%HPW/MCM-41 sample was the most active. This catalyst showed about 96% conversion and 65% selectivity to the most important stereoisomer (-)-isopulegol, under 1 h reaction. This catalyst was reused four times with a fair degree of deactivation. The decreased activity was attributed to the formation of small agglomerates in the channels of MCM-41.

Full text of DOI:10.1016/j.molcata.2012.03.002

Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Intramolecular cyclization of (+)-citronellal using supported
ଝ
1
2-tungstophosphoric acid on MCM-41
Patricia R.S. Braga , Andréia A. Costa , Elon F. de Freitas , Rafael O. Rochac,1, Julio L. de Macedob,∗
a
a
b
,
d,2
b
b,∗
Antonio S. Araujo , José A. Dias , Sílvia C.L. Dias
a
Universidade de Brasília – Faculdade UnB- Gama – Engenharia de Energia, Brasília-DF, 72405-610, Brazil
b
c
Universidade de Brasília – Campus Darcy Ribeiro – Asa Norte – Instituto de Química - Laboratório de Catálise (A1-62/21), caixa postal 4478, Brasília-DF, 70904-970, Brazil
Universidade de Brasília – Campus Darcy Ribeiro – Asa Norte – Instituto de Química - Laboratório de Isolamento e Transforma c¸ ão de Moléculas Orgânicas (LITMO), caixa postal
478, Brasília-DF, 70904-970, Brazil
4
d
Universidade Federal do Rio Grande do Norte – Instituto de Química, Natal-RN, 59078-970, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Supported heteropolyacids (HPAs) have been used as catalysts by fine-tuning their heterogenous and
Brønsted acidic properties. In this work, supported H3PW12O40 (HPW) on MCM-41 was prepared
Received 21 September 2011
Received in revised form 24 February 2012
Accepted 4 March 2012
(
2–40 wt%) and applied in the intramolecular cyclization of (+)-citronellal. The synthesized materials
31
were characterized by P MAS NMR, XRD, FTIR, pyridine gaseous adsorption, low temperature nitrogen
adsorption and thermal analysis. The impregnation of 20% HPW on the surface of MCM-41 decreased the
characteristic crystallographic reflections of the support, suggesting that HPW modifies the long-range
order. The results show that the HPW was deposited preferentially inside the mesopores of the support
and affected the acid strength of the supported catalyst. FTIR studies of pyridine adsorption confirmed
the presence of Brønsted and hydrogen bonded acid sites. All synthesized materials were active in the
cyclization reaction, but the 20%HPW/MCM-41 sample was the most active. This catalyst showed about
96% conversion and 65% selectivity to the most important stereoisomer (−)-isopulegol, under 1 h reac-
tion. This catalyst was reused four times with a fair degree of deactivation. The decreased activity was
attributed to the formation of small agglomerates in the channels of MCM-41.
Available online 12 March 2012
Keywords:
Cyclization of (+)-citronellal
Supported heteropolyacid on MCM-41
1
2-Tungstophosphoric acid
Brønsted acidity
Molecular sieve MCM-41
©
2012 Elsevier B.V. All rights reserved.
1
. Introduction
converted to isopulegols on catalysts containing exclusively Lewis
or strong Brønsted acid sites but can also be obtained with more
selectivity using solids containing Brønsted /Lewis acidity. Mertens
et al. [1] reported the use of the Pt/H-BEA catalyst, which also
contains both acid sites. Therefore, the literature is controversial,
and a wide range of materials can fit the challenge to obtain (−)-
isopulegol with high selectivity [3,4].
The reported heterogenous acid catalysts generally present a
low stereoselectivity to the desired isomer. Neatu et al. [2] com-
pared the behavior of three zeolites (mordenite ZM510, mordenite
CBV-20A and beta) and showed that heterogenous catalysts did
not favor the thermodynamic equilibrium toward (−)-isopulegol.
According to the authors, in most catalysts used, such as zirco-
nia, zeolites and mesoporous materials, the yields were around 4:1
The intramolecular cyclization of (+)-citronellal has been
attracting attention due to the importance of this reaction in the
fine chemical industry. The industrial synthesis of (−)-menthol is
well known [1,2] and starts with the acid-catalyzed isomeriza-
tion of citronellal resulting in a mixture of isopulegol intermediate
isomers, followed by the metal-catalyzed hydrogenation of the
(
−)-isopulegol. The formation of many isomers may lead to poor
selectivity and become the limiting step in (−)-menthol synthesis.
Trasarti et al. [3] performed consecutive cyclization of citronellal
on different solid acids. The data indicated that citronellal may be
(
isopulegol:other isomers).
Among the materials available with strong Brønsted acidity, het-
ଝ
Contribution from Laboratório de Catálise, Instituto de Química, Universidade
de Brasília, Brasília, DF 70910-970, Brazil, http://www.unb.br/iq/labcatalise.
∗
Corresponding authors. Tel.: +55 61 3107 3846; fax: +55 61 3368 6901.
E-mail addresses: sobral.patricia@gmail.com (P.R.S. Braga),
eropolyacids (HPAs) have occupied a distinguished position due to
their unique properties and catalytic results. Although its activity is
remarkable, the small surface area limits the direct use of HPAs with
Keggin structure in several reactions. For that reason, supported
HPAs [5–8] over classical porous materials (e.g., silica gel [9,10],
carbon [11], alumina [12–16] and mesoporous silica [7,17,18]) are
andreiaacosta@gmail.com (A.A. Costa), rafaelrocha@unb.br (R.O. Rocha),
julio@unb.br (J.L. de Macedo), araujo.ufrn@gmail.com (A.S. Araujo), scdias@unb.br
(
S.C.L. Dias).
1
Tel.: +55 61 3107 3857.
Tel.: +55 84 3211 9240.
2
1
381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2012.03.002

1
00
P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
◦
a graphite monochromator were employed with a 2ꢁ step of 0.02
and integration time of 1 s/step for 2ꢁ between 2 and 40 .
usually more effective for catalytic reactions. da Silva et al. [19]
◦
◦
reported the use of a silica-supported Pd/H PW12O40 catalyst in
3
◦
the citronellal cyclization under mild conditions (15–40 C) with
8
0% selectivity to (−)-isopulegol. On the other hand, the use of pure
2.4.2. Infrared spectroscopy
MCM-41 on this reaction revealed low conversions and no selec-
tivity for the desired isomer according to Imachi et al. [20] and Nie
et al. [21]. Thus, considering the low acidity presented by MCM-41,
the use of strategies to modify this material is highly recommended.
Due to the important influence of the acidity in this specific
reaction as well as the lack of kinetic studies in the literature for
the above-mentioned reaction, the goal of the present work was
FTIR spectra were recorded on a Nicolet 6700 spectrometer
(Thermo Scientific) with DTGS detector. FTIR transmittance mea-
surements were performed using the standard KBr pellet technique
−
1
(256 scans with a 4 cm resolution).
2.4.3. NMR spectroscopy
NMR experiments were performed at 7.05 T with a Varian
1
to study the synthesis and characterization of H PW12O40 (HPW)
Mercury Plus NMR spectrometer operating at 300 MHz ( H) and
3
121.4 MHz (31P) at 25 C. Solid state measurements (MAS NMR) in
the materials were performed in a 7 mm Varian CP/MAS probe and
silicon nitride rotors. 31P MAS spectra were obtained under 5 kHz
rate, pulse length of 8.0 ␮s, with recycle delay of 10 s and 512 scans.
The spectra were referenced to H3PO4 (ı = 0). The products of the
◦
supported on the mesoporous material Si-MCM-41. Modified
HPW/MCM-41 catalysts, with loadings in the range of 2–40 wt%,
were tested on the cyclization of (+)-citronellal including the
kinetic analysis of the formation of desired (−)-isopulegol. The cat-
alysts were characterized by 31P MAS NMR, XRD, FTIR, pyridine
gaseous adsorption, low temperature nitrogen adsorption and ther-
mal analysis. Structure, acidity and activity were evaluated as well
as the recycle of these catalysts.
1
reactions were analyzed using CDCl3 as the solvent by H NMR in
a 5 mm ATB probe. 1H spectra were obtained under pulse length of
3.5 ␮s and acquisition time of 0.1 s. The spectra were referenced to
tetramethylsilane (TMS, ı = 0).
2
. Experimental
2.4.4. Nitrogen adsorption/desorption isotherms
2
.1. Materials
Surface area, pore size and pore volume were measured by nitro-
◦
gem adsorption/desorption isotherms at −196.14 C (77 K) using
Ammonium hydroxide (NH OH, 28–30%, Vetec), cetyltrimethy-
an ASAP 2020C from Micromeritics. Each sample (around 0.3 g)
4
◦
◦
lammonium chloride (CTAC, Sigma–Aldrich), tetraethylorthosil-
icate (TEOS, 98%, Sigma–Aldrich), HPW (Sigma–Aldrich) and
purified water (reverse osmosis system, Quimis) were directly used
for the Si-MCM-41 synthesis. (+)-Citronellal (98%, Sigma–Aldrich)
was distilled under vacuum conditions and dicloromethane (99.5%
CH Cl , Impex) was distilled over CaH prior to use.
was dried for 12 h at 200 C and 18 h at 100 C under vacuum
conditions, prior to the adsorption/desorption experiments. The
physico-chemical properties were obtained from the analysis of
the isotherms, using conventional BET and BJH equations available
in the Micromeritics software.
2
2
2
2.4.5. Thermal analysis (TG/TPD)
2.2. Synthesis of MCM-41 and HPW/MCM-41
Thermogravimetry (TG) was used to simulate the temperature-
programmed desorption (TPD) curves and was obtained in a
simultaneous DSC–TGA (TA Instruments) model 2960 using N2
The mesoporous material was synthesized using the co-
precipitation procedure. The exact method has been described
previously [22,23]. The final molar ratio of the Si-MCM-41 synthesis
−
1
(99.999%) as purge gas (100 mL min ). The analyses were made
◦
◦
◦
−1
.
from ambient temperature (∼26 C) to 800 C at 10 C min
was 525(H O):69(NH OH):0.125(CTAC):1(TEOS).
2
4
The aqueous impregnation method was used for the cata-
lyst syntheses (HPW/MCM-41) with the addition of determined
amount of precursor (HPW) on the support. The catalyst sam-
ples were prepared in order to yield HPW loadings in the range
of 2.0–40.0 wt%. A mixture of precursor with support was vigor-
2.4.6. Gas phase pre-adsorption of pyridine
For the pyridine (py) pre-adsorption experiment, platinum cru-
cibles loaded with the samples were placed in a shallow porcelain
plate and inserted into a glass tube adapted to a tubular furnace
(Model F21135, Thermolyne). The samples were dehydrated in
◦
−1
◦
◦
ously stirred (∼100 C) until complete evaporation of the solvent,
dried N2 (100 mL min ) at 300 C for 1 h, cooled to 100 C and then
gaseous pyridine diluted in N2 was allowed to pass through the
◦
followed by calcinations at 200 C for 6 h under air.
◦
samples for 60 min. The temperature was held at 150 C under N2
2
.3. Reaction of (+)-citronellal
for 30 min to remove all physically adsorbed pyridine. Then, the
samples were analyzed by FTIR and the total amount of pyridine
was quantified using TG [23].
To the intramolecular ene reaction of (+)-citronellal the cata-
◦
lysts were previously dried at 300 C for 4 h. After cooling under N2
atmosphere, each catalyst was transferred to a round bottom flask
2.4.7. Scanning electron microscopy (SEM)
(
10 wt% based on (+)-citronellal) with 5 mL of CH Cl . Next, 1 mmol
Scanning electron microscopy images were obtained with
Philips XL30-ESEM equipment operating at 10 kV and 90 mA of
the beam current. The samples were analyzed in a sample holder
through a thin carbon tape. Before the analyses, the samples were
subjected to pretreatment, which consisted of a deposition of a thin
gold monolayer.
2
2
of (+)-citronellal was added to that suspension and the mixture was
kept under magnetic stirring at room temperature. The time of the
reactions was controlled, varying from 30 min to 3 h. All products
were analyzed by 1H NMR. The best catalyst (20%HPW/MCM-41)
was recycled. The activation procedure was performed at 300 C for
4
◦
◦
−1
h at 10 C min under air in a muffle furnace (EDG, model 3000),
just before use in the cyclization reaction.
3. Results and discussion
2
2
.4. Characterization techniques
3.1. Structural and textural characterization of MCM-41 and
HPW/MCM-41
.4.1. X-ray diffraction (XRD)
The samples were analyzed in powder form in a Bruker model
D8 FOCUS at 40 kV and 30 mA. Radiation of Cu K␣ (ꢀ = 1.5418 A˚ ) and
The XRD pattern of calcined MCM-41 material showed three
◦
characteristic peaks, with a strong reflection at 2ꢁ = 2.2 and two

P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
101
Table 1
◦
◦
◦
Physico-chemical properties of MCM-41 (550 C/3 h/air), HPW (200 C/4 h/air) and HPW/MCM-41 catalysts (200 C/6 h/air), determined by XRD and N2 adsorption/desorption
measurements.
a0a (nm)
Pore sizeb (nm)
SBET (m2
g
−1
)
Pore volumec (cm3
g )
−1
Catalyst
d1 0 0 (nm)
HPW
MCM-41
–
1.22
4.09
3.89
3.82
3.98
3.89
4.21
1.9
2.4
2.0
2.1
2.4
2.4
2.8
4.2
831.5
743.1
817.3
647.9
619.3
740.9
0.002
0.51
0.37
0.43
0.38
0.34
0.53
3.5
3.4
3.3
3.5
3.4
3.6
2
5
1
2
4
%HPW/MCM-41
%HPW/MCM-41
5%HPW/MCM-41
0%HPW/MCM-41
0%HPW/MCM-41
√
a
b
c
a0 = 2d1 0 0/ 3 for MCM-41 catalysts.
Average pore size diameter (4 V/A) calculated by BJH desorption method.
BJH desorption cumulative volume of pores.
◦
◦
weak reflections at 4.0 and 4.6 related to the (1 0 0), (1 1 0) and
2 0 0) planes, respectively [23–25]. These results indicate that the
Textural measurements are depicted in Table 1. The synthesized
materials show a reversible type IV isotherm, which is character-
istic of mesoporous materials. The samples containing 15, 20 and
40% of HPW showed a slight type H1 hysteresis, which is associated
with porous materials comprised of ordered uniform spherical par-
ticles [23,31,32]. To analyze the total surface area (SBET) and pore
volume, a correlation was performed between the HPW distribu-
tion inside the pores (mesoporous area) and on the external surface
area. Considering that the mesoporous area contributes more to
the total surface area of the material than the external surface, a
distribution of the HPW inside the pores will affect more substan-
tially the total surface area. For the sample containing 2%, the HPW
species were mainly located inside the mesopores. Between 5 and
20% a simultaneous distribution of HPW inside the pores and on the
external surface was observed. For that reason, there was a consid-
erable decrease in the total surface area. Finally, for the 40% sample,
aggregates HPW species were located preferentially on the external
surface area.
(
synthesized material presents a highly ordered hexagonal array of
channels [23,26]. The calculated spacing distance (d1 0 0) did not
change with the HPW loading or with the unit cell parameter (ao),
which indicates that the mesostructure was not affected (Table 1)
27,28].
The X-ray powder diffraction analyses of the synthesized
HPW/MCM-41 catalysts were obtained in order to confirm the
presence of the Keggin anion on MCM-41. All samples showed, in
different degrees, changes on the planes related to the structural
long range ordering after HPW impregnation (Fig. 1). At higher
loadings of HPW (20 and 40%), two reasons could cause the loss
on the long range ordering of MCM-41: support hydrolysis, which
can cause a partial destruction on the mesopores structure due to
a transformation of MCM-41 siloxane into silanol groups, or a non-
[
3−
homogenous distribution of anionic species, such as [PW12O40
P2W21O71(H O)] or [P W18O62] , with silanol groups within
]
,
6−
6−
[
2
2
the walls of MCM-41 pores [29].
The pore size analysis showed small changes in the 2 and 5%
samples due to the formation of Keggin phases essentially inside
the pores [33]. For the 15–40% samples, the formation of external
surface aggregates of HPW increased the average pore size of the
final material.
No characteristic reflections of HPW were observed on the
supported catalysts containing 2–15%, evidencing a uniform dis-
tribution of the HPA on the MCM-41. In the samples containing
◦
◦
◦
2
0 and 40% of HPW, peaks at 2ꢁ = 10.3 , 25.3 and 26.3 (typical of
pure HPW) were observed, indicating a discrete dispersion of HPW
9,29]. Calculations were performed to evaluate the effect of the
impregnation procedure on the crystallite size of MCM-41. From
to 20%, the samples showed an average size of 23 nm, which is
Fig. 2 shows FTIR spectra of the samples in the range of
−
1
[
1500–400 cm . The main bands related to pure MCM-41 overlap
the main bands of HPW, especially at lower loading. However, in
2
the sample containing 40% HPW, a small band related to W Oc W
−
1
lower than the one calculated for pure MCM-41 (34 nm). This might
be related to the hydrolysis of siloxane groups, leading to smaller
crystallite sizes. On the other hand, the sample containing 40% of
HPW showed particles averaging 55 nm of MCM-41. According to
Llanos et al. [30], higher loadings of HPW can generate agglomer-
ation of MCM-41 particles through the formation of Keggin unit
bridges.
(corner sharing) stretching at 893 cm was observed [34–36]. As
previously reported by Dias et al. [9], an isolate FTIR analysis cannot
distinguish unequivocally the Keggin anion on supports involving
silica framework. Therefore, 31P MAS NMR experiments were per-
formed to confirm structural integrity of HPW. All catalysts showed
only a characteristic signal at ca. −14.8 ppm regard to hydrated
H PW
O after the impregnation and calcination processes. This
12 40
3
◦
Fig. 1. Powder XRD of the samples after impregnation and calcination at 200 C for 6 h under air.

1
02
P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
Fig. 5. Strength distribution of the acidic sites obtained by pyridine desorption
(TG/TPD).
◦
Fig. 2. FTIR spectra of calcined HPW/MCM-41 at 200 C/6 h/air.
◦
5 C. According to the literature, the range and the maximum peak
6
can change depending on the initial amount of water present in the
material due to preparation method and storage conditions [10].
◦
The second mass loss occurred at 159 C and was associated with
the formation of the anhydrous catalyst. The last step took place
◦
in a wide range of temperatures, from 308 to 524 C, related to
two simultaneous events: decomposition of Keggin structure and
dehydroxilation of the molecular sieve [10,23].
3.2. Acidity of materials
The acidity of the catalysts is a fundamental parameter to
determine its reactivity toward the chosen reaction, which uses
acid sites able to promote the isomerization of (+)-citronellal
to (−)-isopulegol. Pyridine gaseous adsorption experiment was
performed to measure the amount of acid sites. The amount of
adsorbed pyridine was obtained by TG/TPD calculations (Fig. 5).
Although the MCM-41 material presents only hydrogen bonding
sites, these sites are stronger than the analogous one on silica due
to neighboring silanol groups that are connected through bridges
in the hexagonal structure [23]. As the loading of HPW increased,
the HPA anions interacted with MCM-41 silanol groups to form new
acid sites. Initially, the total amount of acid sites decreased, because
HPW reacted with the stronger Si OH of MCM-41 to form isolated
Keggin anions. The HPW protons were tightly bound to the meso-
porous support. As the amount of HPW increased, Keggin anions
Fig. 3. 31P MAS NMR spectra of the HPW/MCM-41 catalysts.
signal is related to the maintenance of the Keggin structure and is
dependent on the hydration degree. The spectra indicated that no
decomposition occurred in the samples (Fig. 3) [10,36].
The stability of HPW supported on MCM-41 was examined by
TG/DTG/DTA measurements, under N2 atmosphere. For all cata-
lysts, the curves were similar. In Fig. 4, a curve of 20%HPW/MCM-41
can be observed. Three thermal events could be observed, analog to
◦
pure HPW. From room temperatureto 100 C, the first mass loss was
related to physically adsorbed water, with a peak maximum about
Fig. 4. TG curve of 20%HPW/MCM-41 sample.
Fig. 6. FTIR spectra of HPW/MCM-41 catalysts after adsorbed gaseous pyridine.

P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
103
Fig. 7. Proposed mechanism of the intramolecular cyclization of (+)-citronellal.
Table 2
Table 4
Conversion and selectivityb of 20%HPW/MCM-41 catalyst under different reaction
a
Ratio between Brønsted (B) and hydrogen bonding (HB) sites obtained by FTIR after
adsorption of pyridine.
times.c
Catalyst
Ratio B/HB
Time (min)
Conversion (%)
Stereoselectivity to
−)-isopulegol (%)
(
HPW
–
MCM-41
–
30
40
45
50
55
60
63.9
79.2
92.6
91.8
91.9
96.2
100.0
62.7
61.7
70.5
66.6
71.4
64.8
74.1
2%HPW/MCM-41
5%HPW/MCM-41
15%HPW/MCM-41
20%HPW/MCM-41
40%HPW/MCM-41
0.05
0.09
0.32
1.08
5.99
180
a
Experimental error associated to the method of calculation of the conver-
sion = 1.1%.
b
Experimental error associated to the method of calculation of the stereoselec-
tivity = 4.2%.
began to aggregate, forming a connected structure with terminal
oxygen linked to water molecules hydrogen bonded to the protons
c
Reaction conditions: 1 mmol de (+)-citronellal, 10 wt% of catalyst and 5 mL
CH2Cl2.
[
9]. With larger amounts of HPW (20 and 40%), larger clusters of
HPW formed, which led to enhancement in the overall acidity of
the materials because of the higher amount of available protons
that were not bonded directly to the MCM-41 framework.
3.3. Cyclization reaction
According to FTIR spectra after the pyridine adsorption exper-
iment (Fig. 6), the modified materials had two types of sites:
Brønsted and hydrogen bonding, with bands at 1540, 1490 and
The results of the cyclization reaction in 3 h for all studied
catalysts are shown in Table 3. The MCM-41 support exhibited
conversion around 35% and selectivity to (−)-isopulegol around
−
1
1
444 cm ; which were different from pure MCM-41 [23]. These
3
7%, which is associated with the strength of its hydrogen bonding
new sites originated with the presence of HPW inside the channels
and/or on the external surface of the MCM-41. The relative amount
of these sites can be estimated by the Brønsted (B) and hydrogen
bonding (HB) ratio (Table 2). As seen in Fig. 6, the bands related to
Brønsted sites increased, indicating that the weak sites of MCM-41
were tailored to stronger acid sites. The ratio of B/HB also increased,
evidencing the new properties of these acid sites. These are impor-
tant parameters to explain the activity of these new materials, as
discussed in the next section.
sites [23]. Although this result is not in accordance with the work
reported by Imachi et al. [20], which showed a yield of only 10%
in 18 h for MCM-41, we believe that the conversion and selectiv-
ity results are dependent on the structure, acidity and activation of
the catalyst. Different procedures used for catalyst synthesis affect
the textural properties, leading to different activities for the same
material. For instance, Imachi et al. [20] reported 71% selectivity
Table 3
Stereoselectivity for (−)-isopulegol formationa with different catalysts at 3 h
reaction.
Catalyst
Conversion (%)
Stereoselectivity to
−)-isopulegol (%)
(
MCM-41
35
100
100
100
100
100
37.3
65.0
65.9
62.2
74.1
61.7
2
5
1
2
4
%HPW/MCM-41
%HPW/MCM-41
5%HPW/MCM-41
0%HPW/MCM-41
0%HPW/MCM-41
a
Conditions of the reaction: 1 mmol of (+)-citronellal, 10 wt% of catalyst and 5 mL
CH2Cl2.
Fig. 8. Recycle study of 20%HPW/MCM-41 catalyst.

1
04
P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
Fig. 9. SEM micrographs of (a): pure MCM-41, (b) 20%HPW/MCM-41 before reaction and (c) 20%HPW/MCM-41 after reaction.
and 40% conversion for SiO , while da Silva et al. [19] reported no
activity at all for SiO2.
1 h reaction, spherical morphology was preserved, but small aggre-
gates were observed [40,41].
2
All supported catalysts showed complete conversion of the
+)-citronellal in 3 h (Table 3). The main products formed
Finally, in order to check that no HPW species were dissolved,
a suspension of HPW/MCM-41 in dichloromethane was analyzed
by UV–Vis absorption spectroscopy and no leaching of HPW was
observed. In addition, to the solution separated from the solid cat-
alyst, fresh (+)-citronellal was added to the liquid and no conversion
was observed.
(
were the isomers: (−)-isopulegol, (+)-neoisopulegol, (+)-iso-
isopulegol and (+)-neoiso-isopulegol. Among the studied catalysts,
2
0%HPW/MCM-41 presented the highest selectivity for (−)-
isopulegol, the desired isomer (74.1%). The selectivities to
(
−)-isopulegol using the supported mesoporous catalysts varied
from 60 to 74%, showing notable stereoselectivity compared to the
literature [2,3,19,37]. Moreover, these results reveal good activity of
4. Conclusion
these catalysts as ZnBr , which is a traditional homogenous catalyst
for this reaction [38].
A schematic mechanism of the intramolecular cyclization of (+)-
citronellal is proposed in Fig. 7. The intramolecular cyclization of
2
Catalysts based on 12-tungstophosphoric acid supported on
mesoporous MCM-41 (2–40 wt%) were prepared and tested in the
cyclization of (+)-citronellal. The characterization of the mate-
rials confirmed maintenance of the Keggin anion, which was
preferentially anchored inside the mesochannels of the molecu-
lar sieve. These materials gave rise to acidity that corresponded
to the presence of Brønsted and hydrogen bonding sites. The
increase in the HPW loading is directly proportional to the
amount of Brønsted sites on the materials. All prepared catalysts
were active in the cyclization of (+)-citronellal forming the main
stereoisomer (−)-isopulegol, but a kinetic study demonstrated that
(
+)-citronellal is an ene reaction, for which the relative acidity of
the catalyst plays an important role. The (+)-citronellal carbonyl
group is activated by hydrogen bonding through the interaction of
the supported HPW proton with the oxygen atom of (+)-citronellal
molecule, which considerably increases the carbonyl group elec-
trophilic character. Therefore, the double bond nucleophilic attack
is favored; at the same time, terminal olefin forms when the oxygen
atom removes the hydrogen from the methyl group. Each stereoiso-
mer presents a different transition state: transition state A was the
less hindered and led to the formation of (−)-isopulegol; the other
transition (B, C and D) states showed at least one bulky group, which
increased the activation energy due to 1,3-diaxial steric repulsion.
Among the four carbocationic intermediates, transition state D was
the most energetic. This is in agreement with the work reported by
Mäki-Arvela et al. [37] where the (−)-isopulegol conformation is
more stable than the other stereoisomers. For that reason, the selec-
tivity obtained for (−)-isopulegol was higher for the most bulky
ones.
2
0%HPW/MCM-41 had the best performance with about 96% con-
version and 65% selectivity for 1 h reaction. This catalyst was reused
four times with low degree of deactivation, keeping the selectiv-
ity about the same. The difference in the activity was attributed
to the decrease in the dispersion degree into the pores of MCM-
4
1 by formation of small agglomerates in the channels. Thus, it
was demonstrate the successful application of supported HPW on
MCM-41 for this acid catalyzed cyclization and it was provided
detailed information on the stability of the catalyst as well as the
kinetics of that reaction.
A kinetic study was also performed with the best catalyst,
0%HPW/MCM-41. The results are depicted in Table 4. The data
Acknowledgments
2
reveal that the best reaction time was 60 min with conversion
of 96.2% and selectivity to the isomer (−)-isopulegol of 64.8%.
Although the conversion changes with the reaction time, no signifi-
cant variation occurred for the selectivity relative to (−)-isopulegol.
The 20%HPW/MCM-41 catalyst was reused four times (Fig. 8)
in the (+)-citronellal intramolecular cyclization reaction. With a
reaction time of 60 min, a reduction of 25% in the (+)-citronellal
conversion was observed from the first cycle (96%) to the last
cycle (71%). However, differences in the selectivity for the (−)-
isopulegol isomer were small (13%). The reduction of catalytic
activity could be associated with the catalyst activation before
We acknowledge CNPq and CAPES for research and doc-
torate scholarships and also the financial support provided
by UnB/DPP/IQ, FINATEC, FAPDF, MCT/CNPq, FINEP/CTInfra,
FINEP/CTPetro and PETROBRAS. The authors also acknowledge the
Laboratório Institucional de Microscopia Eletrônica de Varredura
of Universidade do Rio Grande do Norte- UFRN for SEM analyses.
References
[
1] P. Mertens, F. Verpoort, A.-N. Parvulescu, D. Vos, J. Catal. 243 (2006) 7–13.
[2] F. Neatu, S. Coman, V.I. Pârvulescu, G. Poncelet, D. De Vos, P. Jacobs, Top. Catal.
52 (2009) 1292–1300.
◦
every cycle (300 C/4 h/air), which can partially affect the Keggin
[
[
3] A.F. Trasarti, A.J. Marchi, C.R. Apesteguía, J. Catal. 247 (2007) 155–165.
4] P. Virtanen, H. Karhu, G. Toth, K. Kordas, J.-P. Mikkola, J. Catal. 263 (2009)
209–219.
structure of HPW [39]. In addition, the formation of HPW aggregates
on the MCM-41 surface can reduce the accessibility of the reactants
to the catalyst acidic sites. Indeed, SEM studies indicate that the
[5] J. Haber, K. Pamin, L. Matachowski, D. Mucha, Appl. Catal. A 256 (2003) 141–152.
[6] I.V. Kozhevnikov, Russ. Chem. Rev. 56 (1987) 811–825.
[7] I.V. Kozhevnikov, K.R. Kloetstra, A. Sinnema, H.W. Zandbergen, H. Van Bekkum,
J. Mol. Catal. A: Chem. 114 (1996) 287–298.
2
0%HPW/MCM-41 sample showed a good dispersion of homoge-
nous particles in comparison with pure MCM-41 (Fig. 9a–c). After

P.R.S. Braga et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 99–105
105
[
[
8] W. Kuang, A. Rives, M. Fournier, R. Hubaut, Appl. Catal. A 250 (2003) 221–229.
9] J.A. Dias, E. Caliman, S.C.L. Dias, M. Paulo, A.T.C.P. de Souza, Catal. Today 85
[26] K.U. Nandhini, B. Arabindoo, M. Palanichamy, V. Murugesan, J. Mol. Catal. A:
Chem. 243 (2006) 183–193.
(
2003) 39–48.
10] J.A. Dias, M.C. Rangel, S.C.L. Dias, E. Caliman, F.A.C. Garcia, Appl. Catal. A 328
2007) 189–194.
[27] X.-M. Yan, J.-H. Lei, D. Liu, Y.-C. Wu, W. Liu, Mater. Res. Bull. 42 (2007)
1905–1913.
[28] B.R. Jermy, A. Pandurangan, Appl. Catal. A: Gen. 295 (2005) 185–192.
[29] J.C. Juan, J. Zhang, M.A. Yarmo, J. Mol. Catal. A: Chem. 267 (2007) 265–271.
[30] A. Llanos, L. Melo, F. Avenda n˜ o, A. Montes, J.L. Brito, Catal. Today 133–135 (2008)
1–8.
[
(
[
[
[
[
11] F. Levebvre, P. Dupont, A. Auroux, React. Kinet. Catal. Lett. 55 (1995) 3–9.
12] Y. Izumi, R. Hasebe, K. Urabe, J. Catal. 84 (1983) 402–409.
13] K.M. Rao, R. Gobetto, A. Iannibello, A. Zecchina, J. Catal. 119 (1989) 512–516.
14] K. Nowinska, R. Fiedorow, J. Adamiec, J. Chem. Soc. Faraday Trans. 87 (1991)
[31] G. Giannetto, Zeolitas – Caracteristicas, propriedades y aplicaciones industrials,
EdIT, Caracas, 1990.
7
49–753.
a
[
[
15] L.R. Pizzio, C.V. Cáceres, M.N. Blanco, Appl. Catal. A 167 (1998) 283–294.
16] L.R. Pizzio, P.G. Vázquez, C.V. Cáceres, M.N. Blanco, Appl. Catal. A 256 (2003)
[32] J.L. Figueiredo, F.R. Ribeiro, Catálise Heterogénea, 1 ed., Funda c¸ ão Calouste
Gulbenkian, Lisboa, 1989.
1
25–139.
[33] K.U. Nandhini, J.H. Mabel, B. Arabindoo, M. Palanichamy, V. Murugesan, Micro-
por. Mesopor. Mater. 96 (2006) 21–28.
[
[
[
17] P.A. Jalil, M.A. Al-Daous, A.A. Al-Arfaj, A.M. Al-Amer, J. Beltramini, A.I. Barri,
Appl. Catal. A 207 (2001) 159–171.
18] T. Blasco, A. Corma, A. Martínez, P. Martínez-Escolano, J. Catal. 177 (1998)
[34] A.V. Ivanov, T.V. Vasina, V.D. Nissenbaum, L.M. Kustov, M.N. Timofeeva, J.I.J.I.
Houzvicka, Appl. Catal. A 259 (2004) 65–72.
[35] C. Pazé, S. Bordiga, A. Zecchina, Langmuir 16 (2000) 8139–8144.
[36] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171–198.
[37] P. Mäki-Arvela, N. Kumar, V. Nieminen, R. Sjöholm, T. Salmi, D.Y. Murzin, J.
Catal. 225 (2004) 155–169.
[38] E.J. Lenardão, G.V. Botteselle, F. Azambuja, G. Perin, R.G. Jacob, Tetrahedron 63
(2007) 6671–6712.
[39] C. Milone, C. Gangemi, G. Neri, A. Pistone, S. Galvagno, Appl. Catal. A 199 (2000)
239–244.
3
06–313.
19] K.A. da Silva, P.A. Robles-Dutenhefner, E.M.B. Sousa, E.F. Kozhevnikov, I.V.
Kozhevnikov, E.V. Gusevskaya, Catal. Commun. 5 (2004) 425–429.
20] S. Imachi, K. Owada, M. Onaka, J. Mol. Catal. A: Chem. 272 (2007) 174–181.
21] Y. Nie, W. Niah, S. Jaenicke, G.-K. Chuah, J. Catal. 248 (2007) 1–10.
22] A.A. Costa, G.F. Ghesti, J.L. Macedo, V.S. Braga, M.M. Santos, J.A. Dias, S.C.L. Dias,
J. Mol. Catal. A: Chem. 282 (2008) 149–157.
[
[
[
[23] P.R.S. Braga, A.A. Costa, J.L. Macedo, G.F. Ghesti, M.P. Souza, J.A. Dias, S.C.L. Dias,
Micropor. Mesopor. Mater. 139 (2011) 74–80.
24] A. Corma, Chem. Rev. 97 (1997) 2373–2419.
[40] A.C. Carmo Jr., L.K.C. Souza, C.E.F. Costa, E. Longo, J.R. Zamian, G.N.R. Filho, Fuel
88 (2009) 461–468.
[
[25] F. Thbault-Starzyk, B. Gil, S. Aiello, T. Chevreau, J. Gilson, Micropor. Mesopor.
Mater. 67 (2004) 107–112.
[41] B. Li, W. Ma, J. Liu, S. Zuo, X. Li, J. Colloid Interface Sci. 362 (2011) 42–49.