Sn-Beta zeolite as diastereoselective water-resistant heterogeneous Lewis-acid catalyst for carbon-carbon bond formation in the intramolecular carbonyl-ene reaction

Article abstract of DOI:10.1039/b313738d

The water-tolerant Lewis acid Sn-Beta isomerises citronellal to isopulegol with high diastereoselectivity working in batch or in fixed bed reactors with very high turnover numbers.

Full text of DOI:10.1039/b313738d

Sn-Beta zeolite as diastereoselective water-resistant heterogeneous
Lewis-acid catalyst for carbon–carbon bond formation in the
intramolecular carbonyl–ene reaction
Avelino Corma* and Michael Renz
Instituto de Tecnología Química, UPV-CSIC, Avda. de los Naranjos s/n, 46022 Valencia, Spain.
E-mail: acorma@itq.upv.es; Fax: 34 96 3877809; Tel: 34 96 3877800
Received (in Cambridge, UK) 30th October 2003, Accepted 22nd December 2003
First published as an Advance Article on the web 4th February 2004
The water-tolerant Lewis acid Sn-Beta isomerises citronellal to
isopulegol with high diastereoselectivity working in batch or in
fixed bed reactors with very high turnover numbers.
were less successful since side products resulting from etherifica-
tion, cracking or dehydration occur.8
Recently, we have introduced novel solid Lewis acids that
involve tin incorporated into a Beta zeolite network (Sn-Beta)9 or
into a MCM-41 molecular sieve.10 This type of catalysts has
already been applied successfully to the Baeyer–Villiger reaction
with hydrogen peroxide11 and to the Meerwein–Ponndorf–Verley
reduction.12 We will show here that this water-tolerant Lewis acid
(Sn-Beta) can be successfully applied to the carbonyl–ene reaction
in the cyclisation of citronellal to isopulegol.
The catalysts have been synthesized as reported elsewhere.11
Samples A, B and C of Sn-Beta have Si : Sn ratios of 250 (1 wt%
of SnO2), 123 (2 wt% of SnO2) and 82 (3 wt% of SnO2),
respectively. Activity tests in batch mode were carried out as
described in the following procedure: racemic citronellal (620 mg,
4 mmol) was dissolved in 3.0 g of solvent. A sample of the catalyst
(50 mg) was added and the reaction mixture heated to 80 °C for 1
h. The catalyst was filtered out and the reaction mixture analysed by
gas chromatography and the products were identified by GC-MS
spectroscopy and the isopulegol diastereomer by comparison with
the authentic sample.
When Sn-Beta zeolite was employed as catalyst in acetonitrile as
solvent the conversion was almost completed after one hour and the
desired product isopulegol was obtained with a selectivity higher
than 98% and an 85 : 15 diastereoselectivity ratio (Table 1, entry 1).
The influence of metal loading can be discussed from the results of
conversion obtained with catalysts A, B and C after 5 minutes
reaction time (Table 1, entries 1–3). It can be seen there that
conversion increases with Sn content, but the increase of activity is
not linear with Sn content. This is specially the case for the sample
containing 3 wt% SnO2, something which is not surprising since for
this level of metal substitution, not all of the metal is tetrahedrally
coordinated in framework positions.11 The reaction proceeded
smoothly in other polar solvents as dioxane, tert.-butanol or
nitromethane (Table 1, entries 5–7) and the yields of the desired
product were similar. The isomerisation of citronellal into iso-
pulegol could be carried out without solvent. However, in this case
the reaction stopped at a conversion level of 67% and the
diastereoselectivity was slightly lower (Table 1, entry 8). It should
(2)-Menthol is known for its refreshing diffuse odour character-
istic of peppermint and strong physiological cooling effect.1
Therefore, it has, contrary to other diastereo- or enantiomers, a
commercial importance for dentifrices, cosmetics and pharmaceuti-
cals. Synthetically it can be produced by Lewis-acid catalysed
cyclisation of citronellal with subsequent catalytic hydrogenation
(Scheme 1). The important feature of the cyclisation is the
diastereoselectivity towards the isopulegol since the other isomers
will produce commercially less interesting products. ZnBr2 has
been identified as the most suitable Lewis acid for this process
providing a 94 : 6 diastereoselectivity towards isopulegol2 and it is
employed as promoter in the actual industrial process for the
cyclisation yielding 92% of isopulegol.3 However, the stoichio-
metric amounts of zinc bromide required may cause waste disposal
problems since it acts as marine toxin.4 Another critical factor is the
hygroscopic character of anhydrous zinc bromide and the fact that
wet zinc bromide is much less active and selective than the
anhydrous form.
To overcome these drawbacks several soluble metal complexes
have been tested for the cyclisation.5 Thus, triethylaluminium and
2,6-diphenylphenol with a molar ratio of 1 : 3 has been used giving
yields of 95%. Unfortunately, the catalyst is destroyed with NaOH
after the reaction and cannot be recovered.5 Scandium tris(tri-
fluoromethylsulfonate) also works in homogeneous phase using a
halogenated solvent. It gave at 25 °C a diastereoselectivity of 80 :
20 with yield of 58%, while at 278 °C the diastereoselectivity
increased up to 94 : 6.6 Other soluble catalysts have also been used
for performing the selective cyclisation of citronellal, but the yields
obtained were lower than with ZnBr2.5
Finding a heterogeneous catalyst for this cyclisation that can give
similar yields as ZnBr2 under catalytic amounts, and could perform
with high turnover numbers (TON, moles of substrate converted
per mol of active sites) is a matter of much interest from the
academic as well as from the industrial point of view. With this
purpose zirconium-exchanged montmorillonite7 and zinc impreg-
nated commercial silicas3 have been reported. In both cases the
metal was employed in catalytic amounts and proceeded several
cycles. Furthermore, it has been reported that the montmorillonite
can be recycled without loss of activity and in five cycles, an overall
TON of 100 has been achieved. With respect to the diaster-
eoselectivity both systems reach high selectivities of 85–90%.†
Other heterogeneous catalysts — based on Brønsted acid sites —
Table 1 Sn-Beta catalysed cyclisation of citronellal in different solvents
Sn-Beta
Conv.a
[%]
Select.b
[%]
Diastereosel.c
[%]
Entry Sample Solvent
1
2
3
4
5
6
7
8
A
B
C
B
C
B
B
B
Acetonitrile
Acetonitrile
Acetonitrile
97 (47)
99 (63)
99 (66)
> 98
> 98
> 98
> 98
> 98
> 98
> 98
> 98
85
83
81
81
78
82
80
72
Acetonitriled 64
Dioxane
tert-Butanol
99
99
Nitromethane 99
67
—
a Conversion after 1 h and in parenthesis after 5 min. b Selectivity for the
four diastereomeric pulegols with respect to other products. c Selectivity for
the isopulegol diastereomer with respect to the other three diastereomers.
d At 40 °C reaction temperature.
Scheme 1 Industrial production of menthol from Lewis-acid catalysed
cyclisation of citronellal with subsequent catalytic hydrogenation.
550
C h e m . C o m m u n . , 2 0 0 4 , 5 5 0 – 5 5 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Online
be noted that water present in the substrate or the solvent does not
impede the reaction and special precautions were not necessary.
When the concentration of the substrate solution was increased
from 17 wt% to 29 wt% and 38 wt%, the TON increased from ca.
600 to 1000 and 1350, respectively. This is more than one order of
magnitude better than the best-known heterogeneous catalyst.
Superiority of the tin over other Lewis acids introduced into the
zeolite framework is seen in Table 2. Ti-Beta gave lower
conversion (Table 2, entry 3). Additionally, the diastereoselectivity
with respect to the isopulegol is much lower for the titanium-
containing material with only 56% vs. 85% for Sn-Beta (cf. Table
2, entries 1 and 3). Brønsted-acidic Beta zeolite with a comparable
hydrophobicity13 as Sn-Beta in the same reaction conditions also
gave only cyclisation without any side-products. However, again
the diastereoselectivity towards the desired product was lower
(63%; Table 2, entry 2) and the conversion was only 50% vs. full
conversion for Sn-Beta. The origin of the activity in the Sn-Beta
can be attributed to the tin since the all-silica Beta sample has no
activity (Table 2, entry 4).
Fig. 1 Conversion and diastereoselectivity for the cyclisation of citronellal
catalysed by Sn-Beta in a fixed bed.
a fixed bed continuous reactor. We are working at present on the
improvement of the diastereoselectivity.
The authors thank the CICYT (MAT2003–07945-C02–01) for
financial support. M. R. is grateful to the Spanish Ministry of
Science and Technology for a “Ramón y Cajal” Fellowship.
One attempt to improve the diastereoselectivity by lowering the
reaction temperature, as it has been done successfully in other
cases,6 did not produce any improvement. Thus, when the
temperature was lowered from 80 °C to 40 °C, similar selectivities
were obtained but the conversion was decreased significantly
(Table 1, entry 4).
Notes and references
† As-prepared Zn-silicas (ref. 3) gave 78–86% diastereoselectivities,
however, when washed bromide free the selectivity dropped to 70%. This
suggested that large amounts of ZnBr2 were able to modify the reaction
We have seen that the process can also be carried out in a fixed
bed continuous reactor. Thus, 500 mg of catalyst B were filled into
a stainless steel reactor (inner diameter 4 mm, length 13 cm), and
heated to 80 °C. A solution of citronellal in acetonitrile (50 wt%)
was passed through with a rate of 5 g h21. From Fig. 1 it can be seen
that both conversion and diastereoselectivity remain constant for at
least 48 h at a 99% and 83% level, respectively. Over this period
118 g of citronellal were converted which implies that each metal
site performed 11 500 reaction cycles on an average. In comparison
with the heterogeneous Zn or Zr materials this represents an
improvement of two orders of magnitude. No leaching of Sn was
detected when working in either fixed bed or in batch reactor.
In summary, Sn-Beta has been employed for the first time in a
reaction that involves carbon–carbon bond formation. The catalytic
performance with respect to conversion is much superior to
conventional heterogeneous catalysts used for this reaction.
Moreover, this catalyst does not require the usual precautions
against humidity needed for normal Lewis acids. The stability of
the Sn-Beta zeolite makes this catalyst suitable for applications in
selectivity, even though it did not influence the rate of reaction. In the Zr4+
-
montmorillonite case (ref. 7), for the diastereoselectivity only two isomers,
the isopulegol and the neo-isopulegol have been taken into account and a 90
: 10 ratio determined by 1H NMR spectroscopy at a 91% yield.
1 J. O. Bledsoe Jr., in Encyclopedia of Chemical Tecnology (Kirk-
Othmer), 4th Ed., Wiley-Interscience, New York, 1998, vol. 23, p. 858;
K. Bauer, D. Garbe and H. Suburg, Common Fragrance and Flavor
Materials, Wiley-VCH, Weinheim, 1997, p. 52.
2 Y. Nakatani and K. Kawashima, Synthesis, 1978, 147.
3 C. Milone, A. Perri, A. Pistone, G. Neri and S. Galvagno, Appl. Catal.
A: Gen., 2002, 233, 151.
4 Quest International B. V., Eur. Pat. Appl. 1053974, 1999.
5 T. Iwata, Y. Okeda and Y. Hori, Eur. Pat. Appl. 1225163 A2, 2002and
references therein.
6 C. P. Newman, C. S. Sell, P. N. Davy, V. K. Aggarwal and G. P.
Vennall, PTC Int. Appl. WO 9932422, 1999; V. K. Aggarwal, G. P.
Vennall, P. N. Davy and C. Newman, Tetrahedron, 1998, 39, 1997.
7 J. Tateiwa, A. Kimura and S. Uemura, J. Chem. Soc., Perkin Trans. 1,
1997, 2169.
8 G. D. Yadav and J. J. Nair, Chem. Commun., 1998, 2369; G. K. Chuah,
S. H. Liu, S. Jaenicke and L. J. Harrison, J. Catal., 2001, 200, 352; M.
Fuentes, J. Magraner, C. de las Pozas, R. Roque-Malherbe, J. Perez-
Pariente and A. Corma, Appl. Catal., 1989, 47, 367.
9 A. Corma, L. T. Nemeth, M. Renz and S. Valencia, Nature, 2001, 412,
423.
Table 2 Cyclisation of citronellal in acetonitrile catalysed by different Beta
zeolites
Metal
loading
Conv. Select.a Diastereosel.b
10 A. Corma, M. T. Navarro, L. Nemeth and M. Renz, Chem. Commun.,
2001, 2190; A. Corma, M. T. Navarro and M. Renz, J. Catal., 2003, 219,
242.
Entry Catalyst
[wt%]
[%]
[%]
[%]
1
2
3
4
Sn-Beta (B)
Al-Beta
Ti-Beta
2.0 SnO2
2.8 Al2O3
2.0 TiO2
99
50
35
0
> 98
> 98
> 98
83
63
56
11 M. Renz, T. Blasco, A. Corma, V. Fornés, R. Jensen and L. Nemeth,
Chem. –Eur. J., 2002, 8, 4708.
12 A. Corma, M. E. Domine, L. Nemeth and S. Valencia, J. Am. Chem.
Soc., 2002, 124, 3194; A. Corma, M. E. Domine and S. Valencia, J.
Catal., 2003, 215, 294.
13 M. A. Aramendía, V. Borau, C. Jiménez, J. M. Marinas, F. J. Romero
and F. J. Urbano, Catal. Lett., 2001, 73, 203; A. Corma, J. Catal., 2003,
216, 298.
Pure silica Beta—
a Selectivity for the four diastereomeric pulegols with respect to other
products. b Selectivity for the isopulegol diastereomer with respect to the
other three diastereomers.
C h e m . C o m m u n . , 2 0 0 4 , 5 5 0 – 5 5 1
551