Application of Baylis-Hillman methodology in a novel synthesis of quinoline derivatives

Article abstract of DOI:10.1039/a807827k

Reaction of 2-nitrobenzaldehyde with vinyl carbonyl compounds in the presence of 1,4-diazabicyclo[2.2.2]octane affords Baylis-Hillman products, catalytic reduction of which results in direct cyclisation to quinoline derivatives.

Full text of DOI:10.1039/a807827k

Application of Baylis–Hillman methodology in a novel synthesis of quinoline
derivatives
Oluwole B. Familoni, Perry T. Kaye* and Phindile J. Klaas
Department of Chemistry, Rhodes University, Grahamstown, 6140, South Africa. E-mail: chpk@hippo.ru.ac.za
Received (in Cambridge, UK) 27th May 1998, Accepted 8th October 1998
Reaction of 2-nitrobenzaldehyde with vinyl carbonyl com-
pounds in the presence of 1,4-diazabicyclo[2.2.2]octane
affords Baylis–Hillman products, catalytic reduction of
which results in direct cyclisation to quinoline derivatives.
below, room temperature† afforded the expected Baylis–
Hillman products 3a–c (Scheme 1) in moderate to good yield
(68–85%). Several methods of reducing the nitro compounds
3a–c were examined, the most efficient proving to be catalytic
hydrogenation using a 10% palladium on carbon catalyst in
EtOH.‡ Reduction of compound 3a afforded, in 56% yield, a
product initially presumed to be 2,3-dimethylquinoline but
subsequently identified as the N-oxide 4.§ Hydrogenation of the
methyl ester 3b afforded two cyclised products, viz. 3-methyl-
2-oxo-1,2,3,4-tetrahydroquinoline 5 (22%) and the 4-hydroxy
analogue 6 (59%), the latter as a diastereomeric pair, which
could be readily dehydrated (in 70% yield) to the conjugated,
achiral 3-methyl-2-quinolinone 7. The a,b-unsaturated car-
bonyl moiety in the Baylis–Hillman products 3 is, of course,
susceptible to conjugate addition, and treatment of the ethyl
ester 3c with piperidine¶ led to the diastereomers 8, reduction of
which afforded the 2-quinolinone derivative 9;∑ in this case,
cyclisation of the corresponding amino intermediate may only
occur via acyl substitution. In principle, cyclisation of the
reduced, or partially reduced, intermediates may be expected to
involve either conjugate addition or nucleophilic attack at the
carbonyl carbon. In practice, the latter path appears to be the
dominant, if not exclusive, mode of cyclisation—somewhat
surprisingly, given the lack of regioselectivity exhibited by
salicylaldehyde analogues.7,8
The Baylis–Hillman reaction has been the subject of two recent
reviews1,2 and continues to elicit attention.3,4 We have demon-
strated applications of this reaction in the synthesis of
substituted indolizines from pyridine-2-carbaldehydes5,6 and, in
analogous reactions of salicylaldehydes, have uncovered a
veritable cascade of transformations involving the formation of
chromene and coumarin derivatives.7,8 Extension of this general
methodology to 2-aminobenzaldehydes was expected to pro-
vide access to quinoline derivatives.
Numerous quinoline syntheses have been developed,9 includ-
ing the Friedlander synthesis (and modifications thereof) in
which use is made of 2-aminobenzaldehydes. A limiting factor
in the Friedlander methodology, however, is the relative
inaccessibility of substituted 2-aminobenzaldehydes. This lim-
itation, coupled with the fact that aldehyde electrophilicity is an
important factor in Baylis–Hillman reactions,10 prompted us to
explore the use of 2-nitrobenzaldehyde as an activated alter-
native to 2-aminobenzaldehyde, subsequent reduction of the
nitro group being expected to permit cyclisation via the
resulting amine. Quinolines have, in fact, been obtained
previously in yields of 27–30%, by passing mixtures of
2-nitrobenzaldehyde and various alcohols over a heterogeneous
catalyst at elevated temperature (300–320 °C).11 Here we report
preliminary results which clearly illustrate the potential of the
Baylis–Hillman approach to quinoline derivatives under re-
markably mild conditions.
Application of the methodology to the reaction of 2-nitro-
benzaldehyde with ethyl vinyl ketone afforded both 2-ethyl-
3-methylquinoline 10 (25%) and the N-oxide 11 (31%).
O
+
N
O
N
N
Treatment of 2-nitrobenzaldehyde 1 with methyl vinyl ketone
(MVK) 2a, methyl acrylate 2b and ethyl acrylate 2c in the
presence of 1,4-diazabicyclo[2.2.2]octane (DABCO) at, or
O–
10
11
12
O
O
N
H
+
NO2
R
N
H
O
a R = Me
b R = OMe
c R = OEt
1
2a–c
9
i
ii
OH
O
OH
CO2Et
N
R
ii
iii
For 3a
(R = Me)
For 3c
(R = OEt )
NO2
NO2
3a–c
8
ii
For 3b
(R = OMe )
OH
N
iv
+
N
N
O
O
N
O
O–
H
H
H
4
5
6
7
Scheme 1 Reagents and conditions: i, DABCO, CHCl3; ii, H2, Pd-C, EtOH; iii, piperidine, THF; iv, TsOH, toluene, reflux.
Chem. Commun., 1998, 2563–2564
2563

View Article Online
(3H, s, 3-Me), 2.68 (3H, s, 2-Me), 7.45 (1H, s, 4-H), 7.51 (1H, t, 6-H), 7.63
Formation of the N-oxides 4 and 11 was established, in each
case, by FAB MS analysis. Further extension of the procedure
to the reaction of methyl vinyl ketone with 6-nitropiperonal
gave, amongst other products, the corresponding quinoline 12
(26%).
In summary, application of the Baylis–Hillman reaction to
2-nitrobenzaldehydes provides convenient access to substituted
quinoline derivatives which, in turn, constitute useful substrates
for further elaboration. The results of ongoing studies, aimed at
optimising reaction conditions for the selective formation of the
quinolines or their N-oxides and exploring the generality of the
method, will be reported fully in due course.
We thank the Foundation for Research Development (FRD)
and Rhodes University for generous financial support, the
University of Lagos, Nigeria, for study leave (to O. B. F.), Dr
W. E. Molema for assistance with NMR analysis and Dr L.
Fourie (University of Potchefstroom) for FAB MS data.
(1H, t, 7-H), 7.68 (1H, d, 5-H), 8.68(1H, d, 8-H); dC(100 MHz; CDCl3) 14.7
(2-Me), 20.2 (3-Me), 119.6 (C-5), 125.1 (C-4), 127.1 (C-8), 127.7 (C-7),
128.1 (C-4a), 129.2 (C-6), 130.8 (C-3), 139.9 (C-8a), 146.4(C-2)].
¶ A mixture of 2c (0.5 g), piperidine (0.5 ml) and THF (5 ml) was stirred in
a stoppered flask for 24 h. Excess piperidine was evaporated in vacuo and
the residue was chromatographed [flash chromatography on silica; elution
with hexane–EtOAc (2:1)] to give 8 (0.61g, 85%).
∑ Compounds 6, 8 and 9, which appear to be new, and the known quinoline
derivatives 4, 5, 7, 10–12 were characterised by elemental (high resolution
MS) and 1H and 13C NMR spectroscopic analyses.
1 S. E. Drewes and G. H. P. Roos, Tetrahedron, 1988, 44, 4653.
2 D. Basavaiah, P. Darma Rao and R. S. Hyma, Tetrahedron, 1996, 52,
8001.
3 G. P. Black, F. Dinon, S. Fratucello, P. J. Murphy, M. Nielsen and H. L.
Williams, Tetrahedron Lett., 1997, 38, 8561.
4 L. J. Brzezinski, S. Rafel and J. W. Leahy, Tetrahedron, 1997, 53,
16423.
5 M. L. Bode and P. T. Kaye, J. Chem. Soc., Perkin Trans. 1, 1990,
2612.
6 M. L. Bode and P. T. Kaye, J. Chem. Soc., Perkin Trans. 1, 1993,
1809.
7 P. T. Kaye and R. S. Robinson, Synth. Commun., 1996, 26, 2085.
8 J. Bacsa, P. T. Kaye and R. S. Robinson, S. Afr. J. Chem., 1998, 51,
47.
9 See, for example, G. Jones, in Comprehensive Heterocyclic Chemistry,
ed. A. J. Boulton and A. McKillop, Pergamon, Oxford, 1984, vol. 2,
p. 395.
Notes and references
† In a typical Baylis–Hillman reaction, a solution of 2-nitrobenzaldehyde 1
(5.0 g, 33 mmol), methyl acrylate 2b (2.95 g, 34.2 mmol) and DABCO (0.18
g, 1.6 mmol ) was stirred in a stoppered flask for 3–7 d. [In the case of
methyl vinyl ketone 2a, the reaction was noticeably exothermic; use of
CH2Cl2 as solvent and cooling the mixture (ca. 0 °C) during addition of the
reactants resulted in a significantly cleaner product.] The solvent was
evaporated in vacuo and the residue chromatographed [flash chromatog-
raphy on silica; elution with hexane–EtOAc (3:1)] to give 3a (6.78 g;
85%).
‡ Hydrogenation was effected in EtOH at atmospheric pressure using a 10%
Pd-C catalyst (wet, Degussa type; as supplied by Aldrich Chemical Co.)
§ Selected data for 4, mp 123–125 °C (Found, by FAB MS, MH+:
174.09179. Calc. for C11H12NO+, 174.09189. ); dH(400 MHz; CDCl3) 2.45
10 M. L. Bode and P. T. Kaye, Tetrahedron Lett., 1991, 21, 5611.
11 N. S. Koslov, Y. S. Chumakov and S. I. Kozintsev, Katal. Sint.
Prevrashch. Geterotsikl. Soedin., 1976, 57 (Chem. Abstr., 1978, 88,
22572).
Communication 8/07827K
2564
Chem. Commun., 1998, 2563–2564