cesses, thus producing mixtures of 7- and 8-membered rings,
and may also be complicated by allenylation.
resulted in a shortened reaction time and slightly higher yield.
Moderate heating at 80 °C greatly accelerated the reaction
but led to a decrease in yield. Microwave irradiation
produced further shortening of the reaction time and gave a
comparable yield. As a compromise between reaction rate
and yield the microwave-assisted protocol was chosen for
further elaboration. Experiments were carried out in a
monomode CEM Discover microwave reactor in sealed vials.
A number of amides 1a-m were synthesized11 starting
from the corresponding tryptamines and 3-substituted 2-pro-
pynoic acids. In most cases conditions initially applied for
the synthesis of 2a with minor variation of the reaction time
were suitable for fast and efficient conversion of the starting
amides (Table 2, entries 1-3 and 5-7). However, several
limitations on the structure of the starting material were
discovered. Thus, secondary amide 1d (Table 2, entry 4) was
found to be an unsuitable substrate for the cyclization
providing an ill-defined mixture of products. A higher
temperature and an increased amount of catalyst (15 mol
%) were required to drive to completion the cyclization of
1h bearing a bulky i-Pr group (Table 2, entry 8). Neverthe-
less, the corresponding product 2g was obtained in a high
yield (79%). The more bulky t-Bu group (Table 2, entry 9)
completely inhibited the reaction preserving the starting
amide 1i unchanged. The phenylpropiolic acid amide 1j
(Table 2, entry 10) had a reactivity comparable with that of
amide 1h and afforded the indoloazocinone 2h in 61% yield.
Amide 1k containing a highly electron-donating 3,4,5-
trimethoxyphenyl moiety (Table 2, entry 11) required
particularly harsh conditions for complete conversion. With
30 mol % of Hg(OTf)2 at a ceiling temperature of 120 °C,
the corresponding indoloazocinone 2i was obtained in only
40% yield after 50 min of irradiation.
On the other hand, we found out that amide 1a selectively
and efficiently undergoes cyclization into indoloazocinone
2a upon treatment with a stochiometric amount of Hg-
(O2CCF3)2 followed by workup with saturated aqueous KI
solution (Scheme 1). Apparently, this carbocyclization pro-
Scheme 1
ceeds8 via intermediates 3 and 4, the latter resulting from 3
after ion exchange. Nevertheless, we failed to isolate the
iodomercurate 4 or the analogous chloromercurate because
of rapid protodemercuration in aqueous media. The selective
formation of the 8-membered ring is undoubtedly the result
of electronic control during nucleophilic addition to the
conjugated triple bond.
Unfortunately, the cyclization failed to proceed under
catalytic conditions applying up to 15 mol % of Hg-
(O2CCF3)2. The reaction also did not occur in the presence
of Bro¨nsted acids like CF3CO2H and TfOH. Thus, we tested
several Lewis acids (5 mol %) which are often successfully
used as catalysts for alkyne carbocyclizations:9 AuCl, AuCl3,
AgOTf, AgO2CCF3, PdCl2, PtCl2, Hg(OTf)2. However, the
product was formed (Table 1) only in the case of Hg(OTf)2.10
Initially the reaction was performed at rt in MeCN and
required almost 2 days for completion. Switching to DCM
Surprisingly, 1l and 1m containing the terminal triple bond,
which is more susceptible toward an intramolecular nucleo-
philic attack, turned out to be very sluggish substrates in
the cyclization. The reaction mixtures had to be irradiated
for a long time (>60 min) even with up to 15 mol % of the
catalyst to reach complete conversion of the starting material.
The yields of the indoloazocinones 2j and 2k were, however,
extremely low.
Finally, we obtained some improvement applying con-
ventional heating at 50 °C (Table 2, entries 12 and 13); thus,
2h and 2i were isolated with respective yields of 25 and 21%.
N-Tosylated and N-acylated amides 1n and 1o (Table 2,
entries 14 and 15) were left unchanged, presumably, due to
the reduced nucleophilicity of the indole core.
Table 1. Optimization of the Reaction Conditionsa
(7) (a) Ferrer, C.; Echavarren, A. M. Angew. Chem., Int. Ed. 2006, 45,
1105. (b) Ferrer, C.; Amijs, C. H. M.; Echavarren, A. M. Chem.sEur. J.
2007, 13, 1358.
entry
solvent
T (°C)
time (h)
yield (%)
(8) (a) Bates, D. K.; Jones, M. C. J. Org. Chem. 1978, 43, 3856. (b)
Larock, R. C.; Harrison, L. W. J. Am. Chem. Soc. 1984, 106, 4218.
(9) For reviews on metal-catalyzed carbocyclization, see: (a) Ojima, I.;
Tzamarioudaki, M. L. Z.; Donovan, R. J. Chem. ReV. 1996, 96, 635. (b)
Negishi, E.; Coperet, C.; Ma, S.; Liou, S.-Y.; Liu, F. Chem. ReV. 1996, 96,
365. (c) Trost, B. M. Chem.sEur. J. 1998, 4, 2405.
1
2
3
4
MeCN
DCM
DCM
DCM
rt
rt
80
40
24
0.5
0.25
91
95
84
85
80 (MW)b
(10) (a) Nishizawa, M.; Takenaka, H.; Nishide, H.; Hayashi, Y.
Tetrahedron Lett. 1983, 24, 2581. (b) Nishizawa, M.; Morikuni, E.; Asoh,
K.; Kan, Y.; Uenoyama, K.; Imagawa, H. Synlett 1995, 169.
(11) See the Supporting Information for the exact experimental proce-
dures.
a Run on a 0.3 mmol scale in dry solvent (1.2 mL) in a sealed vial;
b Run in a monomode CEM Discover microwave reactor at a maximum
power of 60 W.
Org. Lett., Vol. 11, No. 16, 2009
3619