Intramolecular catalytic Friedel-Crafts reactions with allenyl cations for the synthesis of quinolines and their analogues

Article abstract of DOI:10.1021/ol049301u

(Equation Presented) This paper describes a novel method to synthesize a quinoline backbone by incorporating allenyl cations into a catalytic intramolecular Friedel-Crafts reaction. The initial products were isomerized and aromatized upon treatment with acid and base, respectively, to give quinolines. The basic concept also proved to be promising for 1-benzazepine, 1-benzazocine, or isoquinoline synthesis.

Full text of DOI:10.1021/ol049301u

ORGANIC
LETTERS
2004
Vol. 6, No. 14
2361-2364
Intramolecular Catalytic Friedel−Crafts
Reactions with Allenyl Cations for the
Synthesis of Quinolines and Their
Analogues
,†
Teruhiko Ishikawa,* Shinobu Manabe,‡ Toshiaki Aikawa,‡ Takayuki Kudo,‡ and
,‡
Seiki Saito*
Department of Bioscience and Biotechnology, School of Engineering and School of
Education, Okayama UniVersity, Tsushima, Okayama, Japan 700-8530
seisaito@biotech.okayama-u.ac.jp
Received April 16, 2004
ABSTRACT
This paper describes a novel method to synthesize a quinoline backbone by incorporating allenyl cations into a catalytic intramolecular
Friedel−Crafts reaction. The initial products were isomerized and aromatized upon treatment with acid and base, respectively, to give quinolines.
The basic concept also proved to be promising for 1-benzazepine, 1-benzazocine, or isoquinoline synthesis.
Synthetic routes to quinolines have been studied intensively
over many years due to the important role they play in
industry (e.g., pharmaceutical agents, ligands, and functional
materials). Since the first synthesis by Skraup over a century
ago,1 the aniline-based pathway has been the major route to
generate a quinoline backbone.2 Hence, heterocyclic rings
are usually constructed using substituted aniline derivatives
bearing an appropriate functional group on the nitrogen atom.
Thus, the efficacy of intramolecular Friedel-Crafts (IMFC)
reactions in terms of chemical yield, regioselectivity, and
reaction conditions usually determines the synthetic value
of such processes.3 Various reactive intermediates such as
acylium, iminium, or oxonium ions have been examined.4
However, further studies are required, in particular with
regard to the regiochemistry of the reactions when meta-
substituted aniline derivatives are employed.
We realized that an allenyl cation can be a good electro-
phile in the IMFC process as shown in Scheme 1.5 Since
Scheme 1
† School of Education.
this reaction fulfilled the criteria mentioned above, we
decided to examine its applicability to the synthesis of
‡ School of Engineering.
(1) Skraup, Z. H. Ber. 1880, 13, 2086-2087.
(2) Several modifications based on the Skraup method (ref 1) are well-
known as Doebner-von Miller, Combes, Conrad-Limpach, and Friedla¨nder
quinoline synthesis; for these, see: Gilchrust, T. L. Heterocyclic Chemistry,
3rd ed.; Addison-Wesley Longman: Essex, 1997; pp 158-164.
(3) For the most recent paper concerned with Friedla¨nder synthesis
employing o-nitrobenzaldehydes as a starting material, see: McNaughton,
B. R.; Miller, B. L. Org. Lett. 2003, 5, 4257-4259.
(4) (a) Mayr, H.; Ba¨uml, E. Tetrahedron Lett. 1983, 24, 357-360. (b)
Mayr, H.; Ba¨uml, E. Tetrahedron Lett. 1984, 25, 1127-1130. (c) Nicholas,
K. M. Acc. Chem. Res. 1987, 20, 207-214. (d) Mu¨ller, T. J. J.; Nets, A.
Organometallics 1998, 17, 3609-3614. (e) Olah, G. A.; Krishnamurti, R.;
Parakash, G. K. S. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I. Eds.; Pergamon: Oxford, 1991; Vol. 3, Chapter 1.8.
10.1021/ol049301u CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/09/2004

quinolines and their analogues, for which there are no
previous reports. In addition, an allenyl moiety at C(4) may
allow further derivatization at this position. The work
described in this paper investigates IMFC reactions with
the allenyl cation6 for the synthesis of quinolines, in-
cluding extension to 1-benzazepins, 1-benzazocine, and
isoquinolines.
case shown in Scheme 1, which required a stoichiometric
amount of a Lewis acid such as TMSOTf.10 The results
shown in Table 1 indicate that the reaction afforded an
Table 1. Synthesis of 4-(Vinylidene)tetrahydroquinolines
A reaction similar to that shown in Scheme 1 would be
expected for propargyl silyl ethers (1). On the basis of this
idea, a synthetic route to quinolines and their analogues is
illustrated in Scheme 2. The initial product would be
yield of
Scheme 2
time
2 + 3
(%)a
ratio of
2:3b
entry substrate
R
X
Y
(min)
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1f
Ph
Ph
Ph
OMe
H
Me
H
OMe
H
H
H
H
H
H
10
10
10
10
10
10
60
60
99
98
95
90
86
10:1
3:1
3:1
3:1
i-Pr Me
Me OMe
H
Ph
Ph
OMe
H
Cl
1g
1h
68
60
3:1
a Yields for chromatographically pure products. b Determined by 1H
NMR analysis: isomers were inseparable.
inseparable mixture of 7-substituted tetrahydroquinolines
(2a,c-e,h) and 5-substituted isomers (3a,c-e,h) (entries
1,3-5, and 8) or 6-substituted isomers (2b,g) (entries 2 and
7) in moderate to excellent yields. Entry 6 using 1f bearing
R ) H resulted in the decomposition of the substrate,
indicating that the tertiary nature of the propargylic center
is crucial for the desired reaction to take place. Very high
yields were observed for compounds bearing activating
substituents on the ring (entries 1-5).
The ratio of regioisomers (2:3) for entries 1, 3, 4, and 5
listed in Table 1 suggest that the nature of the R group may
play a certain role in the regiochemical outcome of the IMFC
process. For instance, the ratio (10:1, entry 1) between 2
and 3 changed to 3:1 (entry 5) when the phenyl group was
replaced with a methyl group. The same trend was observed
when a methoxy substituent on the ring (entry 1) was
displaced with a methyl group (entry 3). It would appear
that the more sterically demanding the R or X moiety, the
greater the yield of 2.11
regioisomeric 4-[(3,3-disubstituted)vinylidene]tetrahydro-
quinolines (2 or 3) dependent upon the canonical form (II).
We also expected that the moiety at C(4) might play a
role in converting 2 or 3 to an aromatic system, for in-
stance by isomerization from the 1,2-dienes to 1,3-dienes7
and desulfonylation. Propargyl trimethylsilyl ethers (1a-h)
were readily prepared from aniline derivatives by a series
of routine synthetic reactions. This involved N-sulfonyla-
tion and N-alkylation by the Mitsunobu reaction8 with
3-butyn-1-ol, acetylide formation followed by coupling with
ketones or aldehydes, and then finally O-trimethylsilylation
(Scheme 3).9
(5) (a) Ishikawa, T.; Okano, M.; Aikawa, T.; Saito, S. J. Org. Chem.
2001, 66, 4635-4642. (b) Ishikawa, T.; Aikawa, T.; Mori, Y.; Saito, S.
Org. Lett. 2002, 5, 51-54.
Scheme 3
(6) For ruthenium-catalyzed propargylation of aromatic compounds,
see: (a) Nishibayashi, Y.; Yoshikawa, M.; Inaba, Y.; Hidai, M.; Uemura,
S. J. Am. Chem. Soc. 2002, 124, 11846-11847. (b) Hishibayashi, Y.; Inaba,
Y.; Hidai, M.; Uemura, S. J. Am. Chem. Soc. 2002, 124, 7900-7901.
(7) For isomerization of this class, see: Reinhard, R.; Glaser, M.;
Neumann, R.; Maas, G. J. Org. Chem. 1997, 62, 7744-7751.
(8) Mitsunobu, O.; Wada, M.; Sano, T. J. Am. Chem. Soc. 1972, 94,
679-680.
(9) Series of reactions shown in Scheme 3 led to 1a-h in 40-50%
overall yields from the corresponding aniline or its derivatives. This method
was unsuccessful for bromo or nitro derivatives.
(10) It should be noted that the reactions were completed by the use of
only a catalytic amount of BF3‚OEt2, whereas most of Friedel-Crafts
reactions required a stoichiometric amount of Lewis acid: see ref 4e.
Ring-closing Friedel-Crafts reactions of 1a-h proceeded
smoothly at 0 °C in the presence of a catalytic amount of
BF3‚OEt2 (20 mol %) in marked contrast to the previous
2362
Org. Lett., Vol. 6, No. 14, 2004

For the propargyl silyl ethers with an activated substituent
on the ring (entries 1-5), the reaction was usually complete
within 10 min under the given conditions. Interestingly,
however, when the reaction time was extended (i.e., 60 min),
the product composition changed. For example, prolonged
reaction of 1a resulted in an inseparable mixture of 4a and
5a (10:1) instead of a mixture of 2a and 3a (Scheme 4).
Scheme 5
Scheme 4
However, eight-membered ring formation using 10 afforded
12 in only 40% yield13 probably because of the more
entropically demanding nature of the reaction.
To apply this methodology to the synthesis of substituted
isoquinolines, efficient synthesis of the substrates is required.
For this purpose we employed O-methylhydroxylamine14 as
a dual nucleophile that can undergo SN2 reactions twice with
arylmethyl halides and propargyl halides. We also expected
that a methoxy substituent on the nitrogen atom would trigger
aromatization of the initial Friedel-Crafts products through
an acid-promoted elimination process.
The outline of the preparation of substrates for isoquinoline
synthesis is shown in Scheme 6. Although it was difficult
Although 1g gave only 2g even after 60 min (see entry 7),
2g led to 4g on treatment with p-TsOH in EtOH (reflux, 24
h) (Scheme 4). Hence, the 1,2-diene function of this class
can easily be isomerized to the 1,3-diene function by the
action of BF3‚OEt2 when the benzene rings were activated
or by the action of acid (TsOH) when not activated.
Desulfonylation of 4 (+ 5) to 6 (+ 7) was easily achieved
by exposure to KOH in MeOH (reflux) (Scheme 4).
Fortunately, the regioisomeric mixtures could be separated
at this stage by column chromatography. It should be noted
that the attempted desufonylation of allene derivatives 2 (or
3) resulted in quantitative recovery of 2 (or 3). Oxidative
cleavage of the diphenylvinyl moiety of 6g (KMnO4)
followed by esterification afforded 8g that may be a useful
intermediate for further synthetic studies.
Scheme 6
These successful results prompted us to examine the
applicability of the present methodology for synthesizing
seven- or eight-membered ring heterocycles.12 The results
are summarized in Scheme 5. Propargyl silyl ethers 9 and
10 were prepared by a procedure similar to that described
for 1. Ring-closing IMFC reactions of 9 using BF3‚OEt2
(20 mol %) under the reaction conditions identical to those
used for 1 gave the seven-membered ring product 11 in very
high yield. In this case, no regioisomeric product (6-MeO
isomer) was detected at all. This is probably because one of
the phenyl groups linked to the propargylic carbon becomes
sterically more demanding at the transition state than is the
case during the formation of 3a (entry 1 in Table 1).
to suppress N,N-diarylmethylation, the use of excessive
MeONH2 led to 60% yield of the desired product. The
(11) Benzophenone bearing appropriate substituents on the rings should
be examined in this context, which, however, has not been done yet.
(12) Most previous reports for the synthesis of 1-benzazepine backbone
from aniline derivatives featured the formation of a seven-membered
heterocyclic ring relying on a functional group incorporated into an ortho
substituent of the aniline derivatives: see, for example: (a) Zheng, Z. B.;
Dowd, P. Tetrahedron Lett. 1993, 34, 7709-7712. (b) Anastasiou, D.;
Jackson, W. R. Austr. J. Chem. 1992, 45, 21-37. (c) Sato, T.; Ito, T.;
Ishibashi, H.; Ikeda, M. Chem. Pharm. Bull. 1990, 38, 3331-3334. On the
other hand, to the best of our knowledge, there has been no report with
respect to the formation of 1-benzazocine derivatives featuring an IMFC
process for heterocyclic ring closure employing aniline derivatives as a
starting material.
Org. Lett., Vol. 6, No. 14, 2004
2363

following N-propargylation gave the corresponding [N-(aryl)-
methyl-N-propargyl]-O-methylhydroxylamine in high yield.
A series of reactions similar to those employed for the
preparation of 1 involving generation of acetylides, their
addition to benzophenone, and final O-trimethylsilylation led
to 13a-d.15
Scheme 7
The results of the isoquinoline synthesis are tabulated in
Scheme 7. In these cases, a stoichiometric amount of
TMSOTf in place of BF3‚OEt2 was required for the desired
Friedel-Crafts process to take place (see entries 1 and 2).
In addition, the reactivity of 13 depended on the substrate
structure: 13a,b bearing an electron-donating group in the
meta position were reactive enough to afford 14a,b in high
yield (entries 1 and 3), whereas 13c or 13d, bearing either a
p-methoxy group or no substituent, resulted in decomposition
of the substrate.16 Under the reaction conditions described,
both 14a and 14b were accompanied by a small amount of
byproduct such as 16a or 16b, respectively. This was
probably the result of isomerization followed by elimination
(MeOH) of the initial products (14a or 14b). Tetrahydroiso-
quinolines, a mixture of 14 and 15, also led to 4-vinyliso-
quinoline (16a or 16b) in high yield upon exposure to acid,
resulting in isomerization followed by elimination, as shown
at the bottom of Scheme 7.
a For chromatographically pure products. bAccompanied by
c
separable 16a (5% yield). Inseparable by SiO2 column chroma-
tography. d13a, recovered in 85% yield. eAccompanied by separable
16b (9% yield).
In conclusion, we have demonstrated the synthetic poten-
tial of the allenyl cations in IMFC reactions for quinoline,
isoquinoline, 1-benzazepine or 1-benzazocine syntheses. In
this strategy, the cations were generated by the action of a
catalytic amount of BF3‚OEt2, although the isoquinoline
synthesis required a stoichiometric amount of TMSOTf as a
promoter. The ease of substrate synthesis from substituted
N-tosylanilines for 1, 9, or 10 and N-methoxybenzylamines
for 13 indicates the practical aspect of this synthetic scheme.
Acknowledgment. This research was supported by a
Grant-in-Aid for Scientific Research from the Ministry of
Education, Culture, Sports, Science and Technology, Japan.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for the substrates (1, 9-10,
13) and the products (2-8, 11, 12, 14-16) and copies of
1H and 13C NMR spectra for representative compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(13) In this case, unsaturated ketone was obtained probably through the
Meyer-Schuster rearrangement due to a small amount of water inevitably
present in the reaction medium. Hence, prolonged reaction did not lead to
an increase in the yield of 12.
(14) We have attempted to prepare N-tosyl versions of 13 by reacting
N-tosylarylmethylamines with propargyl bromide (NaH, DMSO) or prop-
argyl alcohol (Mitsunobu conditions), which, however, was unsuccessful.
This was why we paid attention to MeONH2.
(15) Series of reactions shown in Scheme 6 led to 13a-d in 40-50%
overall yields from the corresponding benzyl halides.
OL049301U
(16) No identifiable product was obtained. This probably means that the
side chain was subject to decomposition due to the slow rate of the IFCR
process in these cases.
2364
Org. Lett., Vol. 6, No. 14, 2004