Michael additions catalysed by N,N-dimethyl-3-aminopropyl - Derivatised amorphous silica and Hexagonal Mesoporous Silica (HMS)

Article abstract of DOI:10.1055/s-1998-1731

Solid bases prepared by derivatisation of amorphous silica and hexagonal mesoporous silica with dimethylaminopropyl groups are good catalysts for Michael addition reactions. Of the prepared catalysts, those based on hexagonal mesoporous silica show partic

Full text of DOI:10.1055/s-1998-1731

June 1998
SYNLETT
625
Michael Additions Catalysed by N,N-dimethyl-3-aminopropyl - Derivatised Amorphous Silica
and Hexagonal Mesoporous Silica (HMS)
James E. G. Mdoe, James H. Clark and Duncan J. Macquarrie*
Department of Chemistry, University of York, Heslington, York, YO1 5DD, England
Received 18 February 1998
T
Abstract: Solid bases prepared by derivatisation of amorphous silica
and hexagonal mesoporous silica with dimethylaminopropyl groups are
good catalysts for Michael addition reactions. Of the prepared catalysts,
those based on hexagonal mesoporous silica show particularly high
activities. The sol-gel method used in preparing the catalysts based on
hexagonal mesoporous silica enables loading to be increased to more
than twice that of those based on amorphous silica which are prepared
by the post-modification method.
A value of 0.74 for the surface E (Dimroth's parameter measured
N
using Reichardt's dye) has also been measured indicative of reduced
polarity compared to underivatised silica (0.967).
8
Two novel catalysts based on HMS and N,N-dimethyl-3-
aminopropylsilane were prepared using a templated sol-gel method
9,10
by varying the ratios of the amine to the silane. Catalysts 2 and 3, with
tetraethoxysilane : N,N-dimethyl-3-aminopropylsilane ratios of 9:1 and
4:1, respectively, were obtained. The products were characterised by
thermogravimetric analysis and showed good thermal stability. Both
materials lose the organic groups between 400 C and 650 C. N
physisorption and acid titration analyses gave surface areas of 925 m
The Michael addition is a carbon - carbon bond forming reaction
important in synthetic chemistry. Typically, the reaction involves a
nucleophilic addition of carbanions to α,β-unsaturated carbonyl
compounds. During the course of the reaction, electronic charge
migrates from the carbon in the donor molecule to a better acceptor
atom, usually oxygen, in the electrophile. A base is used to form the
carbanion by abstracting a proton from an activated methylene
precursor. The reaction is thus base catalysed and in addition to its
synthetic utility, may serve as a useful test reaction for studying the
catalytic activity of solid bases.
o
o
2
2
-1
2
-1
-1
-1
g
and 370 m
g
and loadings of 1.2 mmol g and 2.2 mol g for
catalyst 2 and 3, respectively. The pore diameter of 2 was 5.6 nm with
some larger pores of > 10 nm also being evident which are possibly due
to interparticle voids. 3 was predominantly amorphous (pore diameter >
10 nm) with a small amount of templated structure with a pore diameter
of 4 nm. Diffuse Reflectance FTIR (DRIFT) gave peaks at 2960 cm
(C-H .), 2890 cm [-N-(CH ) ], and 1480 cm (C-H ). The surface
-1
-1
-1
st
3 st.
def.
T
N
values were 0.75 and 0.68 for material 2 and 3, respectively.
E
Traditionally, the Michael reaction is catalysed by mild to moderately
1
strong bases. Bases such as diisopropylamine, potassium t-butoxide
All three materials were evaluated as catalysts in the reactions of
nitroalkanes with α,β-unsaturated carbonyl compounds. The results
2,3
10
and tetramethylguanidine have been used as homogeneous catalysts.
However, production of significant amounts of multiple Michael
adducts, other side reactions, difficult product separation and catalyst
recovery, have always been problematic. These problems, plus the
recent increase in the interest of developing environmentally friendly
solid catalysts, has led to the development of heterogeneous catalytic
systems for this reaction. Examples of such systems are those based on
are given in Table 1.
Generally, the catalysts based on hexagonal mesoporous silica show
higher activities than those prepared from amorphous silica, rates being
about 3 times faster on comparing catalysts with similar loadings (Fig.
1).
4
5
KF- and CsF-alumina, potassium t-butoxide on xonotlite, and
6
Amberlyst A-27, which have been used with different degrees of
success. The number of genuinely useful systems however, remains low.
As part of a programme aimed at developing heterogeneous base
catalysts for use in reactions of synthetic value, we have prepared two
catalysts by derivatising amorphous silica and hexagonal mesoporous
silica (HMS) with N,N-dimethyl-3-aminopropyl groups. The catalysts
were evaluated in the Michael additions of nitroalkanes and α,β-
unsaturated carbonyl compounds.
The first catalyst (N,N-dimethyl-3-aminopropyl-silica) 1, was prepared
7
according to a known post-modification method using N,N-dimethyl-
3-aminopropyltrimethoxysilane (Fluorochem, Derbys, 95%) and silica
gel (Kieselgel 100, Merck). Titration with dil. HCl showed a loading of
-1
about 0.85 mmol g . Increasing the amount of N,N-dimethyl-3-
aminopropylsilane did not increase the loading of the catalyst beyond 1
-1
mmol g , suggesting that this is the maximum loading with this type of
preparation method. Thermogravimetric analysis revealed the material
to be thermally stable and it does not lose the organic functions until
o
temperatures above 350 C. Surface area and pore size distribution were
determined by N physisorption. The material has a specific surface area
The HMS based catalysts do, however, have much higher surface areas
although when the amine loading was increased from 1.2 to about 2.2
mmol g on the HMS, the rate increased correspondingly despite the
fact that the surface area is much reduced and the catalyst structure is
essentially amorphous. The polarities of catalysts 1 and 2 as measured
2
2
-1
of 245 m g and a broad pore size distribution centred at 10 nm which
was essentially the same as for the parent silica. The Diffuse
Reflectance FTIR (DRIFT) spectrum of 1 displayed the expected peaks
-1
-1
-1
-1
at 2960 cm (C-H ), 2760 cm [-N-(CH ) ] and 1480 cm (C-H ).
st.
3 st.
def.

626
LETTERS
SYNLETT
4
in these reactions suffers from lack of reproducibility, and is therefore
unreliable.
Our results show that solid bases derived from chemically modified
mesoporous silicas are effective catalysts for Michael reactions. The
HMS-based materials are especially active. There is no simple
correlation between activities and any of the material properties of
surface area, surface polarities or loading.
We thank the Royal Academy of Engineering/EPSRC for a Clean
Technology Fellowship (to JHC), Norwegian Government NORAD
scheme (JEGM) and the Royal Society for a University Research
Fellowship (to DJM).
References
1.
2.
3.
4.
5.
6.
McMurry, J. E. and Melton, J. J. Org. Chem. 1973, 38, 436.
Miller, D. D. and Hamada, K. B. Tetrahedron Lett. 1983, 24, 555.
Pollini, G. P.; Barco, A. and De Giuli, G. Synthesis 1972, 44.
Clark, J. H. and Cork, D. G. Chem. Lett. 1983, 1145.
Laszlo, P. and Pennetreau, P. Tetrahedron Lett. 1985, 26, 2645.
Ballini, R.; Marziali, P. and Mozzicafreddo, A. J. Org. Chem.
1996, 61, 3209.
7.
a) Macquarrie, D. J.; Clark, J. H.; Lambert, A.; Mdoe, J. E. G. and
Priest, A. React. Funct. Polym. 1997, 35, 153.
b) Haginaka, J.; Yasuda, N.; Wakai,. J.; Matsunaga, H.; Yasuda, Y.
and Kimura, Y. Anal. Chem. 1989, 61, 2445.
8.
9.
Tavener, S. J.; Clark, J. H.; Gray, G. W.; Heath, P. A. and
Macquarrie, D. J. Chem. Commun. 1997, 1147.
Macquarrie, D. J. Chem. Commun. 1996, 1961.
T
10. Typical procedure for catalyst preparation: In
preparation of catalyst 2; 18.75 g of tetraethoxysilane and 2.08 g
N,N-dimethyl-3-aminopropyltrimethoxysilane were added
a typical
by their surface E values are similar and higher than that of catalyst 3.
N
All three catalysts show high reaction selectivities especially in the
reactions between nitroalkanes and but-3-en-2-one. In the reaction
between 2-cyclohexen-1-one and the nitroalkanes, nitroethane and
nitropropane, two diastereoisomers were obtained as products in a 1:1
ratio in both cases. In neither reactions, however, were any Knoevenagel
type products observed.
simultaneously over a few seconds to a stirred mixture of n-
dodecylamine (5.09 g) in ethanol (52 g) and distilled water (53 g)
at room temperature. The reaction was allowed to continue for 18
h. The resulting white solid was filtered and washed with ethanol
3
3
(50 cm ). The powder was then refluxed in 3 x 100 cm ethanol to
remove the template, drying initially at room temperature
No leaching of the catalysts was observed in the reaction mixtures, and
the reactions did not proceed when they were interrupted before their
completion by removing the catalyst. This observation has also been
noted in Knoevenagel reactions catalysed by aminopropyl-silica
o
overnight, then at 110 C.
Typical procedure for the Michael reaction: To a 100 ml two-
necked round-bottomed flask was added 0.5 g catalyst 2, 25 ml
nitroethane, 20 mmol of but-3-en-2-one and 0.3 ml n-dodecane
(GC internal standard). The reaction mixture was stirred, refluxed
and monitored continuously by GC. When no significant increase
in product was observed, the reaction was stopped and the reaction
mixture filtered and evaporated at reduced pressure to give the
crude product. The crude product was purified by chromatography
7a
catalyst. The reuse of the catalysts after decantation of the supernatant
and recharging with fresh substrate mixtures has been studied. The two
groups of catalysts, i.e., those based on amorphous silica and those
based on HMS, showed different results. Catalysts 2 and 3 continued to
catalyse the reactions close to completion, though after progressively
longer times. Effective reuse of the catalysts for up to 3 and 5 times, for
catalyst 2 and 3, respectively, was possible. Catalyst 1, on the other
hand, showed an induction period of about 3 h on reuse before any
product was observed. The rate started to pick up when additional but-3-
en-2-one was introduced. High yields were obtained even on the second
or third re-use of the catalyst but again only if additional but-3-ene-2-
one was introduced. Our catalysts show good activity and stability in
relatively small amounts, and outperform other known catalysts such as
over silica gel with ethyl acetate/n-hexane (1:4) as eluent giving
1
2.58 g (89%) of pure product. H NMR (CDCl ): δ 1.6 (d, 3H),
3
2.2 (s, 3H superimposed on m, 2H), 2.6 (t, 2H), and 4.7 (m, 1H).
+.
GC-MS: m/z (%) 146 (M , 4), 115(8), 99(100), 83(5), 76(3),
-1
55(12). IR: 1730 (υ ), 1560 cm (υNO ).
CO
2
11. Lewellyn, M. E. and Tarbell, D. S. J. Org. Chem. 1974, 39, 1407.
12. Sapi, J.; Dridi, S.; Laronze, J.; Sigaut, F.; Patigny, D. Laronze, J-Y.
14
and Levy, J. Tetrahedron 1996, 52, 8209.
Amberlyst A-27 which is used in relatively large amounts, requires
1
longer reaction periods and suffers from limited thermal stability. Basic
alumina has also been used for such reactions, but again is required in
13. The analytical data for 3-(1-nitropropyl)cyclohexanone:
H
C
15
13
NMR (CDCl ) δ 1.1 (t, 3H), 1.25 - 2.6 (m, 11H), 4.5 (m 1H).
3
large amounts to ensure good rates. KF-Alumina, a very active catalyst
NMR (CDCl ) δ 208.6 (CO), 94.1 (CHNO ), 43.6, 43.2, 41.2,
3 2

June 1998
SYNLETT
627
40.7, 27.4, 27.3, 24.2, 23.9, 10.1 (-CH CH ). IR: 1730 (υ ),
14. Bellini, R., Marziali, P., Mozzicafreddo A. J. Org. Chem. 1996, 61
2
3
CO
+.
1560 cm-1 (υNO ); and GC-MS: m/z (%) 186 (M , 28), 168(5),
3209.
2
155(10), 139(80), 121(65), 109(10), 97(90), 81(25), 69(55),
55(100).
15. Rosini, G., Marotta, E., Bellini, R., and Petrini, M., Synthesis
1986, 237.