A Mild and Efficient One-Step Synthesis of Quinolines

Article abstract of DOI:10.1021/ol035333q

(Equation presented) The Friedlaender synthesis of quinolines is an extensively employed protocol, yielding the desired heterocycle in a two-step reduction-condensation sequence. We have developed a mild, efficient, high-yielding single-step variant of this methodology, which employs SnCl ₂ and ZnCl₂ to effect the reaction.

Full text of DOI:10.1021/ol035333q

ORGANIC
LETTERS
2
003
Vol. 5, No. 23
257-4259
A Mild and Efficient One-Step Synthesis
of Quinolines
4
Brian R. McNaughton and Benjamin L. Miller*,‡
§
Department of Dermatology, Department of Chemistry, and
The Center for Future Health, UniVersity of Rochester, Rochester, New York 14642
benjamin_miller@futurehealth.rochester.edu
Received July 18, 2003
ABSTRACT
The Friedl a1 nder synthesis of quinolines is an extensively employed protocol, yielding the desired heterocycle in a two-step reduction−
condensation sequence. We have developed a mild, efficient, high-yielding single-step variant of this methodology, which employs SnCl and
ZnCl to effect the reaction.
2
2
The Friedl a¨ nder synthesis of quinolines from o-aminobenz-
aldehydes is a staple reaction of organic synthesis. This
Because of their importance as substructures in a broad range
of natural and designed products, significant effort continues
to be given over to the development of new quinoline-based
1
synthesis is usually carried out via a two-step procedure, in
which reduction of an o-nitro aryl aldehyde (or 2-nitro vinyl
aldehyde, for the analogous preparation of pyridines) is
followed by condensation with an enolizable carbonyl
compound in the presence of a Brønsted or Lewis acid
catalyst (Scheme 1). A complicating factor in the reaction
3
4
structures and new methods for their construction.
As part of a continuing effort in our laboratory toward
the development of new methods for the expeditious
synthesis of biologically relevant heterocyclic compounds,5
we became interested in the possibility of developing a one-
pot analogue of the Friedl a¨ nder synthesis, in which the
intermediate o-aminobenzaldehyde was not isolated, but
rather immediately converted in situ to a quinoline. We
Scheme 1
6
anticipated that, by analogy to our earlier work, reaction of
2
2
-nitrobenzaldehyde with catalytic amounts of CrCl in the
presence of Mn or Zn dust and TMSCl would provide for
catalytic reduction of the nitroarene. If this were done in
the presence of an enolizable carbonyl compound, we
reasoned that formation of a Cr, Mn, or Zn enolate would
is the relative instability of the intermediate o-amino alde-
hyde, which can readily undergo self-condensation reactions.
Modifications such as those developed by Borsche, in which
the o-nitrobenzaldehyde is converted to an imine prior to
reduction of the nitro group, are helpful in reducing problems
due to o-aminobenzaldehyde instability but also increase the
(
3) Hoemann, M. Z.; Kumaravel, G.; Xie, R. L.; Rossi, R. F.; Meyer,
S.; Sidhu, A.; Cuny, G. D.; Hauske, J. R. Bioorg. Med. Chem. Lett. 2000,
0, 2675-2678.
4) A subset of recent work in this area includes: (a) Du, W.; Curran,
1
(
D. P. Org. Lett. 2003, 5, 1765-1768. (b) Lindsay, D. M.; Dohle, W.; Jensen,
A. E.; Kopp, F.; Knochel, P. Org. Lett. 2002, 4, 1819-1822. (c) Matsugi,
M.; Tabusa, F.; Minamikawa, J. Tetrahedron Lett. 2000, 41, 8523-8525.
2
number of synthetic operations that must be performed.
(d) Dormer, P. G.; Eng, K. K.; Farr, R. N.; Humphrey, G. R.; McWilliams,
J. C.; Reider, P. J.; Sager, J. W.; Volante, R. P. J. Org. Chem. 2003, 68,
467-477.
§
Department of Chemistry.
Department of Dermatology.
‡
(5) (a) Hari, A.; Rodrigues, W. C.; Karan, C.; Miller, B. L J. Org. Chem.
2001, 66, 991-996. (b) Hari, A.; Miller, B. L. Org. Lett. 2000, 2, 3667-
3670.
(
1) Cheng, C.-C.; Yan, S.-J. Org. React. 1982, 28, 37-200.
(2) Borsche, W.; Ried, W. Justus Liebigs Ann. Chem. 1943, 554, 269-
2
72.
(6) Hari, A.; Miller, B. L. Org. Lett. 2000, 2, 691-693.
1
0.1021/ol035333q CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/21/2003

lead to concomitant condensation with the reduced arene,
providing the quinoline in a single synthetic operation.
obtained. However, reduction of nitrobenzaldehyde followed
by condensation with 2-butanone in the presence of excess
ZnCl
2
was successful, albeit low yielding (Table 1, entry
To our surprise, however, reaction of 2-nitrobenzaldehyde
4
). Incorporation of 4 Å molecular sieves into the second
with CrCl
2
, Mn(0), and TMSCl in DMF in the presence of
-butanone produced none of the desired quinoline (Table
, entry 1). Similarly, substitution of Zn(0) for Mn(0) (Table
step provided a substantial increase in the yield. Most
gratifyingly, highest yields were obtained when the reaction
was carried out as a single operation (Table 1, entry 6).
2
1
2 2
Reduction in the number of equivalents of SnCl and ZnCl
to 2 each resulted in a complex mixture of unreacted starting
material, o-nitrobenzaldehyde, and quinoline products.
We next examined the utility of this process to synthesize
a range of quinolines (Table 2). Simple ketones, both n-alkyl
Table 1. Reagent Screening for One-Pot Quinoline Synthesis
2 2
Table 2. Synthesis of Quinolines via the SnCl /ZnCl
Procedure
entry
1
conditions
yield of 2 + 3
CrCl2 (0.25 equiv)
Mn(0) (16 equiv)
TMSCl (16 equiv)
DMF, 50 °C
0
2
CrCl2 (0.25 equiv)
Zn(0) (16 equiv)
TMSCl (16 equiv)
DMF, 50 °C
0
3
4
SnCl2‚2H2O (10 equiv)
EtOH, 70 °C
(1) SnCl2 (5 equiv)
EtOH, 70 °C
0
40%
(2) ZnCl2 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C
5
(1) SnCl2 (5 equiv)
EtOH, 70 °C
80%
(2) ZnCl2 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C, 4 Å mol sieves
SnCl2 (5 equiv)
6
7
8
9
90%
23%
42%
51%
ZnCl2 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C, 4 Å mol sieves
CrCl2 (5 equiv)
ZnCl2 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C, 4 Å mol sieves
SnCl2 (5 equiv)
FeCl2 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C, 4 Å mol sieves
SnCl2 (5 equiv)
AlCl3 (5 equiv)
2
-butanone (1 equiv)
EtOH, 70 °C, 4 Å mol sieves
1
, entry 2) was unsuccessful. Numerous other metal salts
7
have been employed in the reduction of nitroarenes; SnCl
2
8
is one in particular that has found wide usage. We found
that while SnCl indeed caused the disappearance of the
starting nitrobenzaldehyde, none of the desired quinoline was
2
(
(
7) Wei, G. P.; Phillips, G. B. Tetrahedron Lett. 1998, 39, 179-182.
8) Bellamy, F. D.; Ou, K. Tetrahedron Lett. 1984, 25, 839-842.
4258
Org. Lett., Vol. 5, No. 23, 2003

and cyclic, give uniformly high yields. Since the number of
commercially available o-nitrobenzaldehydes is exceedingly
small, we have only begun to examine the substrate scope
of the reaction with regard to this moiety. However, as may
be seen in entries 5-7, the reaction readily tolerates both
methoxy and chloro substitutions, as well as hydroxy and
chloro monosubstitution. Likewise, methyl pyruvate works
well (entry 8), potentially providing a simple entry into
In conclusion, we have developed a rapid, high-yielding
procedure for the conversion of o-nitrobenzaldehydes to
10
quinolines. This method provides a significant improvement
over the venerable Friedl a¨ nder synthesis, by obviating the
need to prepare and isolate o-aminobenzaldehydes. We
anticipate that this methodology will be readily applicable
to solid-phase and combinatorial library synthesis; efforts
along those lines are currently in progress.
compounds related to xanthurenic acid (4), a critical regulator
of malaria development.9
Acknowledgment. The authors thank the sponsors of the
Center for Future Health, University of Rochester, for
generous financial support, and thank Prof. Steven Ley,
Cambridge University, for helpful discussions.
Supporting Information Available: Spectral data for
compound 9. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL035333Q
To verify that the ketone partner was not at fault in our
inability to prepare quinolines using a Cr(II)-based process,
we treated a mixture of o-nitrobenzaldehyde and cyclohex-
(
10) Representative experimental procedure: The procedure for the
reaction of cyclohexanone with o-nitrobenzaldehyde (entry 3) is illustrative.
A 150-mL round-bottom flask equipped with a stir bar and reflux condenser
was flame dried under an atmosphere of N2. o-Nitrobenzaldehyde (0.5 g,
.3 mmol) and cyclohexanone (0.325 g, 3.3 mmol) were added, followed
by 20 mL of anhydrous ethanol. SnCl2 (3.138 g, 16.55 mmol, 5 equiv),
anone as described above, replacing the 5 equiv of SnCl
2
+
3
ZnCl with 5 equiv of CrCl + ZnCl . This provided 7 in a
2
2
2
modest 23% yield, with unreacted o-nitrobenzaldehyde and
o-aminobenzaldehyde constituting the bulk of the remaining
material. This difference in reactivity is striking, and is likely
a result of differences in the Lewis acidity of Cr(II) and
Sn(II).
ZnCl (2.257 g, 16.55 mmol, 5 equiv), and approximately 0.5 g of 4 Å
2
molecular sieves were added to the solution. This mixture was then heated
at 70 °C under an atmosphere of nitrogen for 3 h. The reaction was then
cooled to room temperature and rendered basic (pH 8) with 50 mL of 10%
sodium bicarbonate (aq). The mixture was transferred to a separatory funnel,
and was extracted with 3 × 20 mL of ethyl acetate. Organics were combined
and washed thoroughly with saturated NaCl (aq), dried over Na2SO4, and
filtered through Celite. Following reduction of the solvent in vacuo, the
material remaining was subjected to chromatography on silica (20% ethyl
acetate in hexane). The desired quinoline (7) was obtained in 94% yield as
an orange oil.
(
9) Billker, O.; Lindo, V.; Panico, M.; Etienne, A. E.; Paxton, T.;
Dell, A.; Rogers, M.; Sinden, R. E.; Morris, H. R. Nature 1998, 392, 289-
92.
2
Org. Lett., Vol. 5, No. 23, 2003
4259