oxidative addition such as bromobenzene and 3-bromotoluene
but also for the deactivated substrate 4-bromoanisole. When
3-bromotoluene is used as the substrate then, in principle, two
possible isomers can form. Indeed GC/MS analysis of the crude
product mixture obtained from the reaction showed the presence
of both the 2-methyl isomer (shown in entry 2) and the 4-methyl
isomer in a 5+1 ratio. By contrast when N-benzyl-2-chloro-
5-trifluoromethylaniline is used as a substrate there was no
evidence for the formation of the second isomer.
By contrast with the N-alkylated-2-chloroanilines, simple
unsubstituted 2-chloroanilines gave no or, under considerably
more forcing conditions (10 mol% catalyst, 60 h), trace amounts
of the desired carbazoles. In these cases the main product
observed was the corresponding 2-chloroaryl(aryl)amines 3 (R2
= H).
In all the cases of successful double coupling, small amounts
of the hydrodehalogenated products 4 (R2 = Me, Bn) are seen
(GC/MS) and in the reactions shown in entries 1–3 and 6 small
amounts of the compounds 3 are also observed. These latter
compounds are not seen when N-benzyl-2-chloro-5-trifluor-
omethylaniline is used as a substrate, presumably because the
CF3 group activates the chloride with respect to oxidative
addition.‡ These data indicate that, as anticipated, the amination
step occurs first, followed by a ring closing process. However
the presence of the hydrodehalogenation products 4 opens up
the possibility that ring closure may not occur via the
elimination of HCl but rather by the known oxidative coupling
of compounds of the type 4.6 In order to test this possibility we
subjected N-methyldiphenylamine to the same general catalytic
conditions. No N-methylcarbazole was produced, despite the
fact that this product is readily formed from N-methyl-
2-chloroaniline and bromobenzene (entry 6), demonstrating that
an oxidative coupling mechanism is not operative under these
conditions.
In order to confirm that the 2-chloroaryl(aryl)amines 3 are
indeed the intermediates that enter the second coupling step a
representative example 3a (R1 = 4A-OMe, R2 = Bn, R3 = H)
was synthesised by reacting the isolated compound 3b (R1 = 4A-
OMe, R2 = H, R3 = H, entry 7) with n-butyllithium and then
benzylbromide. When 3a was subjected to the standard reaction
conditions, the expected carbazole was formed in good yield
(entry 9). Therefore it seems reasonable to suggest that two
distinct catalytic manifolds operate; in the first cycle (cycle 1,
Scheme 2) a classical amination reaction occurs to generate an
intermediate 3 which then enters the second cycle by oxidative
addition to Pd(0). The resultant Pd(II) complex then undergoes
intramolecular C–H activation, presumably by an electrophilic
displacement mechanism,7 to give a six-membered palladacycle
which subsequently yields the carbazole by reductive elimina-
tion.
Since 2-haloanilines are strongly electronically deactivated
with respect to oxidative addition, we wondered whether there
is sufficient disparity in reactivity between N-benzyl-2-chlor-
oaniline and 4-chloroanisole for them to undergo the sequential
coupling reaction. GC/MS analysis of the reaction mixture after
24 h showed that whilst some double coupling had occurred to
give the desired carbazole, the major species proved to be
unreacted 4-chloroanisole, with substantial amounts of the
intermediate 3a and the hydrodehalogenated species 4a (R1 =
4-OMe, R2 = Bn, R3 = H) also present. In this case it seems
that the rate of the amination is too slow for substantial product
formation and that the conversion of the intermediate 3a to
product is probably limited by catalyst longevity. When the
reaction of N-benzyl-2-bromoaniline and 4-bromoanisole was
attempted, GC/MS analysis showed that very little double
Scheme 2 Probable catalytic pathway.
coupling had occurred, with only trace amounts of the carbazole
present along with substantial amounts of unreacted 4-bromoa-
nisole ( ~ 70%). In addition small amounts of the hydro-
dehalogenated species 4a and trace amounts of the 2-bromi-
nated intermediate 5 were seen. Here the low rate of amination
is probably due to steric hindrance caused by the bromide in the
2-position of the aniline.
In summary we have shown that it is possible to synthesise
carbazoles by a novel, sequential double coupling of 2-chlor-
oanilines with arylbromides. This method has the potential to
generate a large range of new carbazoles from simple, readily
available starting materials. We are currently investigating the
scope and limitations of this new methodology and its extension
to other heterocyclic systems.
We thank the university of Exeter for a student bursary (to C.
S. J. C.).
Notes and references
‡ GC/MS analysis of the crude reaction mixtures obtained with this
substrate showed the presence of small peaks with masses consistent with
products of the general formula BnN(C6H4R)(C6H3-2-OtBu-5-CF3) formed
by catalytic etheration of the aryl chloride function of the intermediates 3
with NaOtBu.
1 (a) T. Iwaki, A. Yasuhara and T. Sakamoto, J. Chem. Soc., Perkin Trans.
1, 1999, 1505; (b) D. E. Ames and A. Oplako, Tetrahedron., 1984, 40,
1919; (c) D. E. Ames and A. Oplako, Synthesis, 1983, 234.
2 For a recent review, see: A. R. Muci and S. L. Buchwald, Top. Curr.
Chem., 2002, 219, 131.
3 R. B. Bedford and C. S. J. Cazin, Chem. Commun., 2001, 1540.
4 For examples of amination reactions of aryl chlorides with these ligands,
see: J. P. Wolfe, H. Tomori, J. P. Sadighi, J. Yin and S. L. Buchwald, J.
Org. Chem., 2000, 65, 1158.
5 For examples of amination reactions of aryl chlorides with tri-tert-
butylphosphine, see: J. F. Hartwig, M. Kawatsura, S. I. Hauck, K. H.
Shaughnessy and L. M. Alcazar-Roman, J. Org. Chem., 1999, 64,
5575.
6 For a recent example, see: H. Hagelin, J. D. Oslob and B. Åkermark,
Chem. Eur. J., 1999, 5, 2413.
7 B. Martín-Matute, C. Mateo, D. J. Cárenas and A. M. Echavarren, Chem.
Eur. J., 2001, 7, 2341.
CHEM. COMMUN., 2002, 2310–2311
2311