Sonogashira,6 and amidation7 reactions. We wish to extend
this concept to incorporate a direct arylation as the second
step in the cascade (Scheme 1). This method could be used
Table 1. Optimization of the Domino C-N Coupling/Direct
Arylation Conditionsa
Scheme 1
.
Strategy for the Synthesis of Fused Indole
Derivatives
base
(2 equiv)
yield
(%)b
entry
ligand
PPh3
PPh3
BINAP
ligand/Pd ratio
1
10:1
10:1
4:1
Cs2CO3
Cs2CO3
Cs2CO3
K2CO3
Et3N
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
36
30
57
36
-
2c
3
4
5
BINAP
BINAP
BINAP
BINAP
P(2-MeOC6H4)3
P(2-MeOC6H4)3
P(2-MeOC6H4)3
P(2-MeOC6H4)3
P(2-MeOC6H4)3
4:1
4:1
4:1
4:1
to create novel polycyclic heteroaromatic molecules, which
can serve as templates for drug discovery8 or show utility
in organic electronic devices.9 These targets are inaccessible
by our previous methods.
6d
7e
8
41
-
2:1
70
80
-
9
10:1
10:1
2:1
10f
11g
12i
The reaction was initially screened using the parent
substrate 1a to investigate the effects of different Pd sources,
ligands, bases, solvents, and temperatures. We found that
Pd2(dba)3/P(2-MeOC6H4)3 in toluene with 2 equiv of Cs2CO3
as base at 120 °C (Table 1, entry 8) were optimal conditions.
However, when we attempted to expand the substrate scope,
the reactions became messy and yields decreased. We noted
that increasing the ratio of ligand to palladium partially
alleviated this problem (entry 9), and we postulated that
liberated halide may act as a catalyst poison.10 The fact that
no product was detected when the reaction was run in the
presence of 1 equiv of exogenous bromide in the form of
Bu4NBr (entry 10) supports this hypothesis. Gratifyingly, we
found that the addition of 1 equiv of cationic silver (in the
form of Ag2CO3) led to cleaner reactions and allowed us to
reduce the ligand/Pd ratio to 2:1 (entry 11).11,12 We
confirmed that this effect is not just that of adding an extra
equivalent of base by perfoming the reaction in the presence
of 3.0 equiv of Cs2CO3 (entry 12).
With these optimized conditions in hand, we set out to
investigate the scope of the domino process. We began by
altering the substitution of the aryl ring (Table 2). It was
found that tetracycles bearing a wide variety of functional
groups could be synthesized. Halide (entries 1 and 6),
electron-donating (entries 2, 3, and 5), and electron-
withdrawing (entries 4, 7, and 8) moieties were all well
tolerated. It was also possible to add a substituent at the
methylene group with no adverse effects (entry 9). Unfor-
90h
69
2:1
a All entries were performed with 3 mol % of Pd2(dba)3, ligand, and 2
equiv of base in toluene at 120 °C unless otherwise noted. b Yields
determined by NMR spectroscopy. c 6 mol % of Pd(OAc)2 was used as
catalyst. d Dioxane was used as solvent. e DMA was used as solvent. f 1
equiv of Bu4NBr was added g 0.5 equiv of Ag2CO3 was added. h Isolated
yield. i 3.0 equiv of Cs2CO3 was used.
tunately, replacement of the vinyl hydrogen with an aryl
group completely shuts down the reactivity (entry 10).
We anticipated that a more difficult modification would
be to change the heterocycle since this could have a profound
change on the reactivity of the site of direct arylation (Table
3). The reaction proceeded smoothly when benzothiophene
was used as the heterocycle instead of thiophene. Use of
nitro-substituted thiophene (entry 3) led to a sluggish
reaction, which did not reach full conversion even after
several days at 120 °C; still, compound 4b could be obtained
in 50% yield. The slower rate of this substrate may give
some insight as to the mechanism of the arylation step: if a
Heck-type pathway was operative for the arylation step,13
this reaction should proceed faster than the unsubstituted
thiophene (Table 1, entry 7). The retarded reactivity of this
substrate suggests that the reaction proceeds via a different
arylation pathway, possibly an electrophilic aromatic sub-
stitution. Furan and pyrrole derivatives (entries 4 and 5, Table
3) were also successful, albeit in lower yields.
We next modified the linker between the nitrogen atom
and the thiophene to examine the effect of ring size (Scheme
2). Synthesis of the six-membered ring 6a proceeded
smoothly, while the phenyl-fused seven-membered ring 6b
proved more difficult to form, yielding only 54% under our
standard conditions. However, the yield could be increased
to 71% by raising the catalyst loading to 10 mol %.
(6) Nagamochi, M.; Fang, Y.-Q.; Lautens, M. Org. Lett. 2007, 9, 2955.
(7) Yuen, J.; Fang, Y.-Q.; Lautens, M. Org. Lett. 2006, 8, 653.
(8) See: Ohno, H.; Iuchi, M.; Fujii, N.; Tanaka, T. Org. Lett. 2007, 9,
4813. and references cited therein.
(9) For a review on acene- and heteroacene-based OED’s, see: Anthony,
J. E. Chem. ReV. 2006, 106, 5028.
(10) For a review of halide effects in transition metal chemistry, see:
Fagnou, K.; Lautens, M. Angew. Chem, Int. Ed. 2002, 41, 26.
(11) For the use of silver salts to sequester iodide in direct arylation
reactions, see: (a) Campeau, L.-C.; Parisien, M.; Jean, A.; Fagnou, K. J. Am.
Chem. Soc. 2006, 128, 581. (b) Chiong, H. A.; Pham, Q.-N.; Daugulis, O.
One limitation of this method is that substrates bearing
an amide functional group, rather than an amine, fail to
J. Am. Chem. Soc. 2007, 129, 9879, and references therein
(12) The use of Ag2CO3 as a base in Rh-catalyzed direct arylation has
also been described. See: Yanagisawa, S.; Sudo, T.; Noyori, R.; Itami, K.
.
(13) (a) Glover, B.; Harvey, K. A.; Liu, B.; Sharp, M. J.; Tymoschenko,
M. F. Org. Lett. 2003, 5, 301. (b) Gozzi, C; Lavenot, L.; Ilg, K.; Penalva,
V.; Lemaire, M. Tetrahedron Lett. 1997, 38, 8867.
J. Am. Chem. Soc. 2006, 128, 11748
.
4634
Org. Lett., Vol. 10, No. 20, 2008