Use of mesoporous MCM-41 aluminosilicates as catalysts in the preparation of fine chemicals: A new route for the preparation of jasminaldehyde with high selectivity

Article abstract of DOI:10.1006/jcat.1998.1970

α-n-Amylcinnamaldehyde (Jasminaldehyde) has been prepared with high selectivity and using low ratios of benzaldehyde/heptanal by means of mesoporous molecular sieve catalysts and a process which involves the acetalization of heptanal with methanol, followed, in the same pot, by a slow hydrolysis of dimethylacetal and the aldolic condensation of the two aldehydes as the final step. A large pore zeolite (Beta) as well as mesoporous silica-aluminas with regular pore sizes such as MCM-41 and SAM have been used as catalysts. The results indicate that mesoporous silica-aluminas with a very narrow range of pore diameter such as MCM-41 are the most adequate catalysts to produce in one pot the three consecutive reactions, avoiding in a very large extent the self-condensation of heptanal and undesired consecutive reactions. This new route allows us to achieve Jasminaldehyde with high selectivity and with a relative high global reaction rates. The influence of the concentration of acid sites on MCM-41, reaction temperature, and molar ratio of the reactants have been also studied.

Full text of DOI:10.1006/jcat.1998.1970

JOURNAL OF CATALYSIS 175, 70–79 (1998)
ARTICLE NO. CA981970
Use of Mesoporous MCM-41 Aluminosilicates as Catalysts
in the Preparation of Fine Chemicals
A New Route for the Preparation of Jasminaldehyde with High Selectivity
M. J. Climent, A. Corma,1 R. Guil-Lo´pez, S. Iborra, and J. Primo
Instituto de Tecnolog´ıa Qu´ımica, Departamento de Qu´ımica (CSIC-UPV), Avenida de Los Naranjos s/n 46022, Valencia, Spain
Received August 20, 1997; revised November 4, 1997; accepted November 4, 1997
heptanalbeingadded slowlyto the reaction mixture at mod-
erate temperatures.
ꢀ-n-Amylcinnamaldehyde (Jasminaldehyde) has been prepared
with high selectivity and using low ratios of benzaldehyde/heptanal
by means of mesoporous molecular sieve catalysts and a process
which involves the acetalization of heptanal with methanol, fol-
lowed, in the same pot, by a slow hydrolysis of dimethylacetal and
the aldolic condensation of the two aldehydes as the final step. A
large pore zeolite (Beta) as well as mesoporous silica-aluminas with
regular poresizessuch asMCM-41and SAMhavebeen used ascata-
lysts. The results indicate that mesoporous silica-aluminas with a
very narrow range of pore diameter such as MCM-41 are the most
adequate catalysts to produce in one pot the three consecutive re-
actions, avoiding in a very large extent the self-condensation of
heptanal and undesired consecutive reactions. This new route al-
lows us to achieve Jasminaldehyde with high selectivity and with
a relative high global reaction rates. The influence of the concen-
tration of acid sites on MCM-41, reaction temperature, and molar
c
A general difficulty with aldol condensation reactions
and, more specifically, in the condensation of 1-heptanal
with benzaldehyde is that both reactants undergo side re-
actions which in this case reduce the yield of amyl cin-
namic aldehyde and form undesired by-products. During
the synthesis of (4) the most important undesired product
comes from the autocondensation of heptanal to form 2-n-
pentyl-2-n-nonenal (5) (Scheme [1] above). The formation
of (5) can be inhibited, to some extent by maintaining a very
low concentration of heptanal relative to benzaldehyde in
the reaction mixture. This methodology requires long ad-
dition times in batch reactions and renders impractical the
use of plug-flow continuous reactors for carrying out this
process.
The formation of further by-products results from the
limited stability of benzaldehyde under the reaction con-
ditions which tends to disproportionate via the Cannizzaro
reaction to yield benzyl alcohol and benzoic acid, forming
the benzoate with the base catalyst and loosing, therefore,
reactant and catalyst.
ratio of the reactants have been also studied.
ꢀ 1998 Academic Press
INTRODUCTION
The synthesis of ꢀ-n-amylcinnamaldehyde (jasminalde-
hyde) (4),2 a traditional perfumery material with violet
scent, involves the condensation of 1-heptanal (1) with ben-
zaldehyde (3).
Different solid base catalysts have been reported for the
synthesis of Jasminaldehyde. Among them, the most inter-
[1]
This aldol condensation is carried out in the presence of
alkali as catalyst (sodium or potassium hydroxide) (1, 2),
estingare:anionicexchange resins(3), potassium carbonate
in the presence of a solid–liquid phase transfer catalyst (4),
and solid–liquid phase transfer catalyst in dry media under
microwave irradiation (5). It is well known that this type
of reaction can also be catalyzed by solid acids. However,
the fact that lower selectivities are generally found when
1 To whom correspondence should be addressed. Fax: 96-387 78 09.
2 The bold face numbers in parentheses refer to the chemical formulas
used in the schemes given throughout the paper. Schemes are denoted by
bracketed numbers.
70
0021-9517/98 $25.00
c
Copyright ꢀ 1998 by Academic Press
All rights of reproduction in any form reserved.

NEW PREPARATION OF JASMINALDEHYDE
71
using acid, instead of base, catalysts is responsible for the
TABLE 1
very few reports involving the use of heterogeneous acid
catalysts in the production of Jasminaldehyde (6).
Main Structural Characteristics of the Catalysts
Surfacea
area
Average (A)
pore
diameter
b
In this work we will present that, by working in a coupled
way with the catalyst and the process, it was possible to de-
velop a new route for the preparation of Jasminaldehyde us-
ing solid acid catalysts which gives higher selectivities than
solid base catalysts. The new route is based on our previous
observation that Al-MCM-41 materials were very active
catalysts for the acetalization reaction and the hydrolysis
of acetals (7). Then, we thought that these solids should
be able to catalyze, in one pot, three consecutive reactions:
acetalization of the aldehydes with methanol, followed by
a slow hydrolysis of acetal, and finishing with aldolic con-
densation. By operating in this way, it was possible, with
relatively high global reaction rates, to have low concentra-
tions of heptanal at the surface of the catalyst, decreasing,
therefore, the self-condensation of heptanal and achieving
high selectivity to Jasminaldehyde.
˚
Porea volume
Si/Al
ratio
Catalysts
(m2 ꢁ gꢂ1
)
(cm3 ꢁ gꢂ1
)
1.MCM-41
2.MCM-41
3.MCM-41
SAM-144
ꢁH
900
870
837
719
607
0.500
0.670
0.480
0.310
0.175
14
50
83
50
13
35
42
34
15
7.6 6.4
*
a Measured by N2.
b Measured by Ar.
temperature and, after saturation, the samples were de-
gassed at 423, 523, and 673 K and the IR spectra were re-
corded at room temperature.
Reaction Procedure
Activation of the catalyst, was performed by in situ heat-
ing at 373 K under vacuum (11 ꢃ 102 Pa) for 2 h. There-
after the system was left at room temperature and a solu-
tion of 10 mmol of heptanal (Technical grade, from Fluka)
EXPERIMENTAL
Materials
A large pore zeolite, H-Beta zeolite, was used, to- in methanol (Technical grade, from Panreac) (15 ml) was
gether with H-MCM-41 as the acid catalyst. The H-Beta poured onto the activated catalyst. The resulting suspen-
zeolite was prepared starting from a TEA-Beta sample sion was magnetically stirred at reflux temperature un-
(Si/Al = 13) and heating at 773 K in N2 stream, followed by til 80% conversion of the heptanal to dimethylacetal was
calcination in air at 823 K. The sample thereafter was twice achieved. At the end of the reaction the methanol was dis-
NH+4 -exchanged and calcined at 823 K.
tilled in a vacuum, and the benzaldehyde (10–50 mmol) was
Three samplesof Al-MCM-41mesoporousmaterial were added. The resulting suspension was magnetically stirred
synthesized following the procedure given in the literature under N2 atmosphere and the kinetics of the reaction was
(8) usinghexadecyltrimethyl-ammonium (Panreac98wt% ) followed at different reaction temperatures (100–140ꢄC).
cation as the template, and pseudoboehmite (Catapal B, Liquid samples of this suspension were extracted at reg-
Vista) as the aluminum source. The samples were activated ular times with a filtering syringe and analyzed by gas
by calcination in air at 813 K for 7 h, or calcination in N2 chromatography using a capillary column. At the end of
for 1 h, and in air for 6 h at 813 K.
the reaction the catalyst was filtered and washed with
For comparison purposes, one amorphous silica-alumina, dichloromethane and the filtrate was weighed and analyzed
was prepared according to Ref. (9), and it is named by gas chromatography–mass spectrometry (GC-MS) us-
SAM-144.
The main characteristics of the acid samples used are a 25-m capillary column of cross-linked 5% phenylmethyl-
summarized in Table 1. silicone; the ratios m/z and the relative intensities (% ) are
The solids were characterized by X-ray diffraction on a indicated for the corresponding peaks.
ing a Hewlett–Packard 5988A spectrometer provided with
Phillips PW diffractometer using CuKꢀ radiation. Surface
After the reaction was completed, the catalysts were
area measurements were obtained on an ASAP-2000 ap- submitted to continuous solid–liquid extractions with di-
paratus following the BET procedure. Pore diameter distri- chloromethane using a micro-soxhlet apparatus. After re-
bution was obtained using argon as adsorbate and follow- moval of the solvent the residue was analyzed by GC-MS
1
ing the Horvath–Kawazoe method (10). IR spectra were and H-NMR spectroscopy and included in the mass bal-
recorded on a Nicolet 710 FTIR spectrophotometer, us- ance. In allcasesthe recovered productsaccounted for more
ing 10 mg ꢁ cmꢂ2 wafers. The Bro¨nsted and Lewis acidity than 90% of the starting material. Isolation of pure com-
of the samples was measured by pyridine adsorption– pounds was accomplished by high performance liquid chro-
desorption, and the resultant sample was monitored by IR. matography, using an isocratic Waters apparatus provided
Thus, the samples were treated in an IR cell at 673 K and with a semi-preparative MicroprasilR column and mixtures
1
15 ꢃ 10ꢂ3 Pa for 16 h. Pyridine was then admitted at room of hexane-ethyl acetate as eluent. The H-NMR analyses

72
CLIMENT ET AL.
of the products were carried out with a 400 MHz Varian CH2(CH2)3CH3), 1.40–1.10 (m, 6H, CH2(CH2)3CH3), 0.83
VXR-400 S spectrometer in deuterated trichloromethane (t, 3H, J = 6 Hz, CH3); MS: 233 (M-1, 1), 174 (6), 121 (100),
with TMS as the internal standard. Chemical shifts are re- 104 (6), 91 (13).
ported in ꢂ (ppm) and are referenced to TMS. The chemical
structures of all the products obtained in this study are pre- Ph CH CH CH2), 6.22 (dt, 1H, Ph CH CH CH2 ),
sented:
trans-1-Phenyl-heptene (7). 1H-NMR: 6.36 (d, 1H,
–
==
–
–
==
–
–
[2]
+
–
==
–
–
2.19 (dc, 2H, Ph CH CH CH2 ); MS: 174 (M , 61), 161
(12), 117 (100), 104 (96), 91 (44).
1
2-Pentyl-3-methoxyindene (8). H-NMR: 7.43–7.12 (m,
–
== –
5H, ArH), 6.45 (s, 1H, Ph CH C ), 4.94 (s, 1H,
–
–
CHOCH3), 3.02 (s, 3H, OCH3), 2.40–2.30 (m, 2H,
–
CH2 (CH2)3CH3), 1.80–1.10 (m, 2H, CH2(CH2)2CH3),
1.45–1.20 (m, 4H, CH2(CH2)2CH3), 0.91 (t, 3H, J = 6 Hz,
CH3); MS: 216 (M+, 100), 185 (11), 159 (89), 131 (9).
2-Pentyl-3-phenyl-1,1,3-trimetoxy propane (9a, b). 7.50–
–
7.20 (m, 5H, ArH), 4.47 and 4.36 (d, 1H, J = 4 Hz, PhCH
–
OCH3), 4.13 and 4.11 (d, 1H, J = 4 Hz, CH(OCH3)2),
–
3.47, 3.38, 3.36, 3.30, 3.24, and 3.16 (s, 3H, OCH3), 2.10–
1.90 and 1.85–1.80 (m, 1H, CH(CH2)4CH3), 1.50–0.70 (m,
11H, n-pentyl); MS: 174 (10), 121 (100), 91 (10), 77 (15),
75 (100).
[3]
Spectroscopic-Spectrometry Data of the Reaction Products
Preliminaryexperimentswere carried out under the same
experimental conditions on MCM-41 (Si/Al = 14) catalyst,
but stirring at different speeds (500 and ꢂ1000 rpm), and no
differences in the initial rate of the reaction was observed.
Two series of experiments were also carried out at the same
reaction conditions but using 1.7% and 5 wt% of MCM-41
(Si/Al = 14) catalyst, the initial rate of the reaction was seen
to be directly proportional to the amount of catalyst, while
no differences on product selectivity were observed.
1
(E)-ꢀ-Pentylcinnamaldehyde (4). H-NMR: 9.54 (s, 1H,
–
–
==
CHO), 1.70–7.38 (m, 5H, ArH), 7.20 (s, 1H, Ph CH C),
–
–
–
2.52 (t, 2H, J = 8 Hz, CH2 (CH2)3 CH3), 1.50–1.39 (m,
–
–
–
–
–
2H, CH2 (CH2)2 CH3), 1.40–1.30 (m, 4H, CH2 (CH2)2
+
–
–
CH3), 0.89 (t, 3H, J = 8 Hz, CH3); MS: 202 (M , 12), 145
(9), 129 (48), 115 (100), 105 (28), 91 (14).
(E)-2-Pentyl-2-nonenaldehyde (5). 1H-NMR: 9.36 (s, 1H,
–
–
–
==
–
CHO), 6.44 (t, 1H, J = 8 Hz, CH2 CH CCHO ),
–
–
–
==
–
2.45 2.29 (m, 2H, CH2 CCHO CH ), 2.29–2.10 (m, 2H,
–
–
==
–
CH2 CH CCHO ), 1.60–1.10 (m, 14H, CH2), 1.10–0.80
RESULTS AND DISCUSSION
(m, 6H, CH3); MS: 210 (M+, 98), 153 (30), 139 (34), 125
(41), 41 (100).
Reaction Mechanism
3-Methoxy-2-pentyl-3-phenyl propanaldehyde (6). 6a.
1H-NMR: 9.71 (d, 1H, J = 4 Hz, CHO), 7.38–7.25 (m,
The production of Jasminaldehyde (4) in one pot from
–
–
5H, ArH), 4.32 (d, 1H, J = 8 Hz, CHOCH3), 3.16 (s, 3H, heptanal dimethyl acetal (2) and benzaldehyde (3) using
–
–
–
OCH3), 2.65–2.55 (m, 1H, CH CHO), 1.60–1.40 (m, 2H, heterogeneous solid acid catalysts, involves several reac-
–
–
–
–CH2 (CH2)3CH3), 1.30 1.10 (m, 6H, (CH2)3CH3); 0.80 tion steps (Scheme [2]). In the first one, heptanal dimethyl
–
(t, 3H, J = 7 Hz, CH3). 6b. 9.64 (d, 1H, J = 3 Hz, CHO), acetal (2) is formed by refluxing heptanal (1) with an excess
–
–
7.40–7.25 (m, 5H, ArH), 4.45 (d, 1H, J = 8 Hz, CH OCH3), of methanol in the presence of the solid acid catalyst. In the
–
–
2.52–2.50 (m, 1H, CH CHO), 1.80–1.60 (m, 2H, second step and when the yield of (2) is around 80% , the

NEW PREPARATION OF JASMINALDEHYDE
73
lyzed aldolic condensation between heptanal (1) and ben-
zaldehyde. Following this procedure and adjusting the rates
of the hydrolysis of the acetal and the aldolic condensation,
one can controlthe concentration ofheptanalon the surface
of the catalyst, and in this way it is possible to depress the
extension of the unwanted self condensation of heptanal.
When the reaction was carried out in the presence of
the mesoporous aluminosilicate 1.MCM-41 (Si/Al = 14)
(1.7 wt% ) using a molar ratio PhCOH/heptanal 1.5 : 1, at
398 K, we obtained after 2-h reaction time, a total conver-
sion of 99% with a yield of (E)-ꢀ-n-amylcinnamaldehyde
(jasminaldehyde) (4) of 68% . In the reaction mixture a
9% of self-condensation heptanal product (E)-2-n-pentyl-
2-n-nonenal (5), 17% of 3-methoxy-2-pentyl-3-phenyl pro-
panaldehyde (6), 2% of trans-1-phenylheptene (7) and
3% of 3-methoxy-2-pentyl indene (8) were also detected.
Furthermore, we also observed at the beginning of the
reaction the formation of a mixture of two diastereoiso-
mers of the dimethylacetal of (6), that is 2-pentyl-3-phenyl-
1,1,3-trimethoxy propane (9a, 9b), which disappeared with
the time (Scheme [3]). When the yields of the different
products were plotted versus the reaction time (Fig. 1b),
Jasminaldehyde (4) and the self-condensation product
(5) appeared as primary products, 3-methoxy-2-pentyl-3-
phenyl propanaldehyde (6) showed a primary but unstable
character, while alkene (7) and indene (8) are secondary
and stable products. The formation of (4) and (5) could be
very easily explained through a competitive acid catalyzed
aldolic condensation which involves the electrophilic at-
tack of a protonated carbonyl group, from benzaldehyde or
heptanal, to the enolic form of heptanal,
FIG. 1. Yields of different products (1 + 2) (᭹), 4 (᭡), 5 (ꢃ), 6 (1),
(7 + 8 + 9) (᭺) versus reaction time between benzaldehyde and hep-
tanal dimethylacetal using 1.MCM-41 (1.7 wt% ), a reactant molar ratio
PhCHO : heptanal = 1.5 : 1 at (a) 373 K, (b) 398 K, (c) 413 K.
methanol is distilled in vacuum, benzaldehyde is added and
the system is heated at the adequate temperature. Under
these reaction conditions heptanal dimethyl acetal under-
goes the deacetalization to (1), followed by the acid cata-
[4]

74
CLIMENT ET AL.
From Fig. 1b one can realize that the rate of the com-
petitive self-condensation reaction is very slow. This effect
being related to the low concentration of heptanal on the
surface of MCM-41. The deacetalyzation process of hep-
tanal dimethylacetal to produce the reactive heptanal can
occur by two different routes, namely, by hydrolysis and/or
by trans-acetalization between the two aldehydes involved
in the process. In our case, the detection of benzaldehyde
dimethyl acetal at the beginning of the reaction strongly
indicates that, besides the direct hydrolysis of the hep-
tanal dimethyl acetal, heptanal can also be formed by the
trans-acetalization reaction. The choice of a catalyst such
as MCM-41 with weak acid sites and large pores makes the
hydrolysis and transacetalization to proceed slowly, main-
taining in this way a low concentration of heptanal on the
catalyst surface.
With respect to product (6) we found that this is a mixture
of two diastereoisomers (6a, 6b) and their formation could
be explained, in a first approximation, by an acid catalyzed
electrophilic Michael addition of methanol to Jasminalde-
hyde (4). However, if this was the mechanism, product (6)
should appear as a secondary product, something which is
not reflected in Fig. 2, in where it appears that product (6)
has a primary character.
FIG. 2. Yield of Jasminaldehyde (4) (᭡) and 6 ( ꢃ ) versus total
conversion, obtained in the reaction between benzaldehyde and hep-
tanal dimethylacetal using 1.MCM-41 (1.7 wt% ), a reactants molar ratio
PhCHO : heptanal = 1.5 : 1 at 398 K.
Then, it can be expected that the intermediate (10) is more
stable than the intermediate formed by the direct protona-
tion of benzaldehyde.
In order to obtain further confirmation on this, Jasmi-
naldehyde was refluxed with methanol in the presence of
1.MCM-41 during 5 h. In these conditions, product (6) was
not formed being Jasminaldehyde dimethyl acetal, the only
product observed. From this it can be concluded that (6)
should be produced by a reaction other than the Michael
addition of methanol to Jasminaldehyde.
An alternative reaction which could explain the forma-
tion of (6), is the condensation between benzaldehyde and
heptanal which involves the carbocation (as electrophilic
species) (10) (Scheme [5]) which comes from the hemiac-
etal of benzaldehyde,
On the other hand, and as we indicated above, product
(6) appeared as an unstable product (Fig. 1b) whose maxi-
mum yield is achieved after 15 min of reaction time, being at
this point the conversion of the reactant close to 90% . After
this point, the yield of product (6) decreases, while the yield
to Jasminaldehyde (4) is still increasing. This effect can be
better seen when plotting the yield of Jasminaldehyde ver-
sus total conversion (Fig. 2). It can be seen that, at a high
level of conversion, Jasminaldehyde also forms, via an inter-
mediate, once depletion of the product (6) has taken place.
Thus, one can relate these two observations and propose
[5]
The high yield of product (6) observed at short reaction that under our reaction conditions the acid catalyzed elim-
times (Fig. 1b) indicates that both the transacetalization ination of methanol from (6) occurs to give Jasminalde-
between heptanal and benzaldehyde to produce the hemi- hyde (4) and that this is the main way to form Jasminalde-
acetal and the consecutive reaction of the carbocation (10) hyde.
with the heptanal, should take place at a much faster rate
The minority products (7) and (8) clearly come from
than the condensation of the protonated benzaldehyde with Jasminaldehyde. The former (1-phenylheptene) could be
heptanal to give Jasminaldehyde. This is not surprising if formed by oxidation of (4) followed by a fast decarboxyla-
one takes into account that the mesomeric donor effect of tion, while the indene (8) should be formed through a mech-
the metoxy group is larger than that of a hydroxy group. anism which involves an intramolecular acid-catalyzed

NEW PREPARATION OF JASMINALDEHYDE
75
TABLE 2
Results of the Reaction of Benzaldehyde with Heptanal Dimethylacetal (2) over Different Catalysts
Totala
conversion
(% )
Vdisappearance
ꢃ 10ꢂ3
Yielda (% )
Yielda (% )
Yielda (% )
Yielda (% )
Yielda (% )
Yielda (% )
Catalysts
(mol/min ꢁ g)
to 4
to 5
to 6
to 7
to 8
to 9
2.MCM-41
SAM-144
ꢁH
71
85
51
5.04
5.76
2.08
25
32
6
3
3
4
37
38
26
—
3
11
2
5
—
3
2
4
Note. Reaction conditions: 1.7% w/w of catalyst/398 K/ratio PhCHO : heptanal = 1.5 :1.
a Time of reaction: 2 h.
cyclization of the hemiacetal of Jasminaldehyde,
[6]
decarboxylation of (4) on the former catalyst. Since we have
not seen a large difference in the acidity of the two catalysts,
we think that the observed behavior can be a consequence
of the regular pore dimensions in H-MCM-41 which allow
faster diffusion of Jasminaldehyde avoiding consecutive re-
actions, such as decarboxylation and cyclization to indene,
to occur (Table 3). This hypothesis is in agreement with the
fact that for Beta zeolite, where the diffusion of the product
out is very slow, the autocondensation and decarboxylation
of Jasminaldehyde is strongly enhanced.
In conclusion, it becomes clear that an adequate catalyst
for the desired reaction involves the use of a mild acid cata-
lyst with large regular pores which allow fast diffusion of the
Jasminaldehyde decreasingand, therefore, the probabilities
for consecutive reactions to occur.
Finally, the very small amounts of product (9), i.e. 2-
pentyl-3-phenyl- 1,1,3-trimethoxy propane observed can be
formed through a transacetalization between the dimethy-
lacetal of heptanal and (6).
Influence of the Catalyst Structure
In order to study the influence of the catalyst structure
on the rate of formation of Jasminaldehyde we have car-
ried out the reaction using three different solid acid cata-
lysts. In this way the behavior of a large pore a microp-
orous zeolite such as Beta has been compared with that
of a mesoporous molecular sieve (H-MCM-41) and, also, a
mesoporous amorphous silica-alumina with a narrow pore
size distribution (SAM).
The results given in Table 2 show that the rate of disap-
pearance of heptanal on the different catalysts follows the
order SAM ꢅ H-MCM-41 > Beta. The similar initial rate
observed for the two silica-alumina catalysts, i.e., SAM and
H-MCM-41, is in agreement with the results given in the
literature showing similar rates for the oligomerization of
olefins (11) and acetalization of aldehydes (7). The lower
rate observed on the microporous molecular sieve can be
due to pore restrictions for the diffusion of the product out
of the catalyst. The combination of large pores and mild
acidity in the mesoporous materials makes these solids ad-
equate catalysts to carry out the acetalization–hydrolysis–
condensation reactions.
Influence of the Concentration of Acid Sites on H-MCM-41
In the case of bimolecular reactions catalyzed by acid
sites, not only the total number, but also the density of acid
TABLE 3
Yield of the Different Products and Selectivity to (4) in the Re-
action of Benzaldehyde with Heptanal Dimethylacetal (2) at 80%
Conversion over Different Catalysts
Selectivity Yield Yield Yield Yield Yield Yield
(% )
(% )
(% )
(% )
(% )
(% )
(% )
Catalysts
to 4
to 4
to 5
to 6
to 7
to 8
to 9
2.MCM-41
SAM-144
ꢁH
52
35
32
41
28
26
6
3
9
29
37
29
—
3
13
2
5
1
2
4
2
With respect to Jasminaldehyde selectivity (Table 3), it
can be seen that at high levels of conversion H-MCM-41 is
more selective than SAM, showing the better catalyst be-
havior of the former for the elimination of MeOH in prod-
uct (6) to give Jasminaldehyde (4), while producing less
Note. Reaction conditions: 1.7% w/w of catalyst/398 K/ratio PhCHO :
heptanal = 1.5 :1.

76
CLIMENT ET AL.
TABLE 4
Results of the Reaction of Benzaldehyde with Heptanal Dimethylacetal (2) in the Presence of H-MCM-41 with Different Si/Al Ratios
Totala
conversion
(% )
Vdisappearance
ꢃ 10ꢂ3
Yielda (% ) Yielda (% ) Yielda (% ) Yielda (% ) Yielda (% ) Yielda (% )
Catalyst
Si/Al
(mol/min ꢁ g) Vo/[Al/Al + Si]
to 4
to 5
to 6
to 7
to 8
to 9
1.MCM-41
2.MCM-41
3.MCM-41
14
50
83
99
70
83
19.71
5.04
1.11
0.35
0.30
0.33
68
25
29
9
3
5
17
37
43
2
—
—
3
2
2
—
3
4
Note. Reaction conditions: 1.7% w/w of catalyst/398 K/ratio PhCHO : heptanal = 1.5 :1.
a Time of reaction: 2 h.
sites are important. These two parameters have a strong is no doubt that if one increases further the frame-
influence on both the total number of active sites and on work Si/Al ratio, the moment arrives when the concen-
the adsorption properties of the material.
tration of active sites is so small that the reaction rates
It can be expected that for framework Si/Al ratios above are strongly influenced, so that the consecutive elimina-
10, most of the acid sites will be isolated and, consequently, tion reaction is proportionally influenced. This hypothe-
a further increase on the Si/Al ratio will only change the sis could explain the observation that the ratio of product
total number of acid sites and therefore total conversion, (6) to (4) increases, decreasing, therefore, the selectivity to
unless changes in the hydrophobic–hydrophillic properties Jasminaldehyde.
of the sample can play an important role on the reaction.
In our case, we have studied the influence of the num-
ber of acid sites by carrying out the reaction on three
H-MCM-41 samples with Si/Al ratios of 14, 50, and 83. The
Influence of the Reaction Temperature
Taking into account the nature of the products formed
initial rates, and the turnover numbers calculated by divid- during the studied reaction, it can be expected that the re-
ing the initial rate by the ratio Al/Al + Si (Table 4) indicate action temperature will have a strong influence on prod-
that the activity is directly proportional to the amount of uct selectivity. Indeed, an increase in the temperature of
acid sites, and furthermore, the activity per site is the same reaction, besides increasing the rate of disappearance of
regardless the amount of Al in the sample.
the reactant, should strongly favor the consecutive reac-
From the point of view of the product selectivity, the tions of Jasminaldehyde such as oxidation–decarboxylation
product distribution obtained on the three catalysts at the and the cyclization to indene, as well as the elimination of
same level of conversion (80% ) (Table 5) indicates that methanol from product (6). The two first reactions will tend
when increasing the framework Si/Al from 14 to 50, the to decrease the final selectivity to Jasminaldehyde, while the
selectivity to Jasminaldehyde increases, with the corre- elimination reaction will increase its final yield and selec-
sponding decrease in the molar ratio of product (6) to tivity. These expectations are confirmed by the experimen-
(4). This result can be explained by taking into account tal results given in Figs. 1a–c and Table 6. It can be seen
that the conversion of product (6) to Jasminaldehyde in- there that with the amount of catalyst used (1.7 wt% ), the
volves the elimination of a polar molecule (methanol), maximum selectivity is observed at 373 K, but under these
which certainly could be favored when increasing the hy- conditions 24 h of reaction time are needed to achieve only
drophobicity of the MCM-41 by increasing the Si/Al ra- 80% conversion. Then, in order to increase the rate of the
tio. Despite the beneficial effect of hydrophobicity, there reaction while still working at lower temperatures (373 K),
TABLE 5
Yield of the Different Products and Selectivity to 4 at 80% Conversion Obtained in the Reaction of Benzaldehyde with Heptanal
Dimethylacetal (2) over H-MCM-41 with Different Si/Al Ratio
Selectivity (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Ratio
Catalyst
Si/Al
to 4
to 4
to 5
to 6
to 7
to 8
to 9
6/4
1.MCM-41
2.MCM-41
3.MCM-41
14
50
83
25
52
30
20
41
24
2
6
4
57
29
46
—
—
—
1
2
2
—
2
4
2.85
0.71
1.92
Note. Reaction conditions: 1.7% w/w of catalyst/398 K/ratio PhCHO : heptanal = 1.5 :1.

NEW PREPARATION OF JASMINALDEHYDE
77
TABLE 6
Results of the Reaction of Benzaldehyde with Heptanal Dimethylacetal (2) at 80% Conversion at Different Reaction Temperatures
Temperature
(K)
Selectivity (% )
Vdisapp. ꢃ 10ꢂ3
(mol/min ꢁ g)
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
to 4
to 4
to 5
to 6
to 7
to 8
to 9
373
398
413
40
25
29
0.73
19.71
36.06
32
20
23
3
2
3
42
57
44
—
—
2
1
1
4
2
—
4
Note. Reaction conditions: 1.7% w/w of 1.MCM-41/ratio PhCHO : heptanal = 1.5 :1.
an experiment was done using 5 wt% of catalyst, since as relatively high ratio of benzaldehyde to heptanal, which re-
it was shown before that the amount of catalyst does not quires larger reactors.
have a direct influence on selectivitybut onlyon conversion.
We have shown here that through our reaction system it
The results obtained (Fig. 3) show that it is now possible to is possible to work with almost equimolecular mixtures of
achieve conversions above 90% and in all cases the selectiv- benzaldehyde and heptanal while producing very little of
ity achieved at 373 K is larger than when working at higher the self-condensation product. Nevertheless, we have also
temperatures. This is especially notorious at higher conver- studied the influence of the ratio of benzaldehyde to hep-
sions, where the consecutive reactions of Jasminaldehyde tanal and the results obtained are given in Table 7.
have a strong impact on the final selectivity observed.
Interestingly, when increasing the ratio of benzaldehyde
Finally, it is worth pointing out that an increase in the to heptanal from 1.5 to 5 mol ꢁ molꢂ1, the initial rate for the
reaction temperature has little effect on the condensation disappearance of heptanal (or even better for the corre-
product, while it strongly affects the decarboxylation and sponding acetal) decreases by threefold. This result could
cyclation to indene.
be explained on the basis of a competitive adsorption of
benzaldehyde on the acid sites responsible for the transac-
etalization and ulterior condensation, decreasing there-
fore, the concentration of the available active sites for
Influence of the Benzaldehyde/Heptanal Ratio
It was said in the Introduction that when the conden- the desired reaction. This hypothesis is supported by the
sation reaction is catalyzed by base catalysts, the concen- changes in selectivity observed when increasing the ben-
tration of heptanal in the reaction media should be kept zaldehyde/heptanal ratio. Indeed, while the selectivity for
very low in order to decrease the rate of the autoconden- the self-condensation decreases, as was expected, the se-
sation reaction. This is achieved in two ways: by decreasing lectivity to Jasminaldehyde decreases at conversions below
the rate of addition of heptanal and consequently involv- 80% (Fig. 4). This is mainly due to the higher concentra-
ing very long reaction time, or by directly working with a tion of product (6) (Fig. 5), possibly due to the competitive
FIG. 3. Yield of Jasminaldehyde versus total conversion obtained in
the reaction between benzaldehyde and heptanal dimethylacetal with a
reactants molar ratio PhCHO : heptanal = 1.5 : 1, using different amounts
of catalyst (1.MCM-41): 1.7 wt% (᭡) and 5 wt% (ꢃ) at 373 K.
FIG. 4. Yield of Jasminaldehyde versus total conversion obtained in
the reaction between benzaldehyde and heptanal dimethylacetal using
different molar ratios of the reactans: PhCHO : heptanal = 1.5 :1 (᭡), and
5 :1 (ꢃ), in the presence of 1.MCM-41 (5 wt% ) at 373 K.

78
CLIMENT ET AL.
TABLE 7
Results of the Reaction of Benzaldehyde with Heptanal Dimethylacetal (2) at 80% Conversion Using Different Reactant Ratio
PhCHO : heptanal
ratio
Selectivity (% )
Vdisapp. ꢃ 10ꢂ4
(mol/min ꢁ g)
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Yield (% )
Selecti.
4 + 6
to 4
to 4
to 5
to 6
to 7
to 8
to 9
1.5 : 1
5 : 1
42
37
9.9
3.4
35
30
4
—
36
48
—
—
1
—
4
2
83
97
Note. Reaction conditions: 5% w/w of 1.MCM-41/373 K.
adsorption of benzaldehyde on the acid sites, which de-
The results obtained with the Beta zeolite are a clear con-
creasesthe rate ofelimination ofMeOH from product (6) to sequence of the faster diffusion of heptanal with respect to
give Jasminaldehyde. Nevertheless, by increasing the ratio benzaldehyde, and especially of the smaller size of the tran-
of benzaldehyde to heptanal, we decrease the autoconden- sition state for the self-condensation reaction with respect
sation and increase the yield of products (6) + (4) which to that necessaryto form the Jasminaldehyde, and the better
finally will end up giving product (4) as the only one, in- fit of the former in the pores of the zeolite.
creasing the selectivity to Jasminaldehyde (4). Indeed this
is observed from the results in Fig. 4, where the selectivity
to Jasminaldehyde can be as high as 90% for conversions
larger than 80% .
CONCLUSIONS
We have shown here that it ispossible to carryout the syn-
At this point it is reasonable to ask what will be the prod- thesisofJasminaldehyde, an industriallyimportant product,
uct distribution under acceptable reaction conditions if one with high selectivities to the desired product and using low
tries to carry out the synthesis of Jasminaldehyde by di- ratios of benzaldehyde to heptanal by the combination of
rect condensation of benzaldehyde and heptanal, without the appropriate process and catalyst. In this case it is pos-
forming in the same pot the acetal of the heptanal.
sible by adjusting the acidity of the catalyst to carry out
When this reaction was carried out on 1. MCM-41 using a the reaction using a unique catalyst in a single pot reac-
molar ratio of benzaldehyde to heptanal of Fig. 5, the results tion and forming the acetal of the heptanal in a first step.
given in Fig. 6 show that the process was much less selective Then, by using the adequate catalyst one can balance the
with more than 20% of self-condensation being produced. rates of the hydrolysis of the acetal and the condensation of
If one uses Beta zeolite instead of MCM-41 (Fig. 7), the the two aldehydes, avoiding to a very large extent the self-
selectivity to the self-condensation product is much higher, condensation. Meanwhile, if the pores of the catalyst are
being the resultant selectivity to Jasminaldehyde of order in a very narrow range of pore diameter and in the meso-
25% .
porous region (MCM-41) the product can readly diffuse
FIG. 5. Yield of product (6) versus total conversion obtained in the
reaction between benzaldehyde and heptanal dimethylacetal using differ-
FIG. 6. Time-conversion plot for reaction of benzaldehyde (5 mol)
ent molar ratios of the reactants: PhCHO : heptanal = 1.5 :1 (᭡), and 5 : 1 and heptanal (1 mol) in the presence of 1.MCM-41 (1.7 wt% ) at 398 K:
(ꢃ) in the presence of 1.MCM-41 (5 wt% ) at 373 K. 1 (᭹), 4 (᭡), 5 (ꢃ), 7 (+), heptanoic acid (ᮀ).

NEW PREPARATION OF JASMINALDEHYDE
ACKNOWLEDGMENTS
79
Financial support by the Direcccio´n General de Investigacion Cient´ıfica
y Te´cnica of Spain (Projects MAT94-0359-C02-02 and MAT97-1016-C02-
01) is gratefully acknowledged.
REFERENCES
1. Farbenindustrie, I. G., DRP, 284, 458/1927.
2. Payne, L. S., E.P. 0 392 579 A2 (1990).
3. Mastagli, P., and Durr, G., Bull. Soc. Chim. France, 268 (1955).
4. Sarkar, A., Dey, P. K., and Datta, K., Indian J. Chem. B 25, 656 (1986).
5. Abenhaim, D., Ngoc Son, C. P., Loupy, A., and Ba Hiep, N., Synthetic
Comm. 24, 1199 (1994).
6. Mastagli, P., and Lagrange, G., Compt. Rend. 244, 207 (1957).
7. Corma, A., Climent, M. J., Iborra, S., Navarro, M. C., and Primo, J.,
J. Catal. 161, 783 (1996).
FIG. 7. Time-conversion plot for reaction of benzaldehyde (5 mol)
and heptanal (1 mol) in the presence of Beta zeolite (10 wt% ) at 398 K:
1 (᭹), 4 (᭡), 5 (ꢃ), 7 (+), heptanoic acid (᭺).
8. Corma, A., Forne´s, V., Navarro, M. T., and Perez-Pariente, J., J. Catal.
148, 569 (1994).
9. Bellusi, G., Perego, C., Carati, A., Peratello, S., Previde Massara, E.,
and Perego, G., Stud. Surf. Sci. Catal. 84, 85 (1994).
10. Horvath, G., and Kawazoe, K., J. Chem. Eng. Jpn 16, 470 (1983).
11. Bellusi, G., Cavani, F., Arrigoni, V., and Ghezzi, R., E.P. 463 673
(1992).
out, avoiding consecutive reactions which can decrease the
final yield of Jasminaldehyde.