A. S. Kyei et al. / Tetrahedron Letters 45 (2004) 8931–8934
8933
O
N
O
N
O
i, ii
iv
X
R
X
X
OH
X = O, NMe
PhSe
15 : X = O 80%
16 : X = NMe 83%
R = OH
R = Cl
iii
v
O
N
+
N
X
X
O
X = O
X = NMe
17 : 48%
18 : 49%
19 : 37%
20 : 33%
Scheme 4. Reagents: (i) (COCl)2, DCM; (ii) 2-(methylamino)ethanol, Et3N, DCM, quant. for two steps; (iii) MsCl, Et3N, DCM, 80–85%; (iv)
(PhSe)2, NaBH4, EtOH, reflux (see Ref. 11); (v) Bu3SnH, AIBN, benzene, reflux.
conjugation with the adjacent aryl ring and will not
undergo unwanted fragmentation prior to the hydrogen
transfer step.
17 and 18, respectively. The yields of isolated spirocy-
cles10 were similar to the vinyl radical addition cases
but significant amounts of reduced starting materials
19 and 20 were isolated in this case.
At first we investigated the cyclisations of the higher
energy aryl radicals. Radical cyclisation precursors
4a,b and 5 were prepared as shown in Scheme 2. To
our delight, when these were subjected to standard radi-
cal cyclisation conditions,9 the spirocycles 6a,b and 7
were formed in excellent yields with only traces of
reduced materials detectable by the 1H NMR of the
crude reaction mixture (Scheme 2).
In summary, we have investigated alkyl, vinyl and aryl
radical dearomatising spirocyclisations onto the C-2
position of benzofuran and indole. Additions of tethered
aryl radicals proceeded in near quantitative yields while
vinyl and alkyl radical cyclisations provided novel spiro-
cycles in moderate yields. Addition of a vinyl radical to
benzofuran proceeded with fragmentation due to instabi-
lity of the intermediate radical.
Next, we set out to investigate vinyl radical additions
and the cyclisation precursors 8 and 9 were prepared
via alkylation of secondary benzyl amides with 2,3-di-
bromo-1-propene as indicated in Scheme 3.
Acknowledgements
We would like to acknowledge the NMR staff of the
Chemistry Research Laboratory, especially Dr. Barbara
OÕDell for her help with establishing product structures.
Also, we are grateful to Dr. Jeremy Robertson for dis-
cussions and suggestions.
Under the radical cyclisation conditions, the indole
derivative 9 was converted into the spirocycle 10 while
the benzofuran derivative 8 underwent spirocyclisation
followed by a reductive fragmentation to 15. Our mech-
anistic proposal for the rearrangement is outlined in
Scheme 3. The spirocyclic radical 11 undergoes fragment-
ation with formation of a more stable phenoxy radical
12. Reduction of the quinone resonance form 13 by tribut-
yltin hydride results in the oxystannane 14, which is
hydrolysed to the phenol 15 on work up. The different
outcome of the two reactions is mostly attributed to
the greater stabilisation of the intermediate spirocyclic
radical by the lone pair of nitrogen in the indole case.
References and notes
1. Morris, B. D.; Prinsep, M. R. J. Nat. Prod. 1999, 62, 688–
693.
2. Takayama, H.; Kurihara, M.; Subhadhirasakul, S.; Kit-
ajima, M.; Aimi, N.; Sakai, S.-I. Heterocycles 1996, 42,
87–92; Takayama, H.; Ishikawa, H.; Kurihara, M.;
Kitajima, M.; Aimi, N.; Ponglux, D.; Koyama, F.;
Matsumoto, K.; Moriyama, T.; Yamamoto, L. T.;
Watanabe, K. J. Med. Chem. 2002, 45, 1949–1956.
3. Narasimhan, N. S.; Aidhen, I. S. Tetrahedron Lett. 1988,
29, 2987–2988; Estevez, J. C.; Villaverde, M. C.; Estevez,
R. J.; Castedo, L. Heterocycles 1994, 38, 1–4; Estevez, J.
C.; Villaverde, M. C.; Estevez, R. J.; Castedo, L. Tetra-
hedron 1994, 50, 2107–2114; Caddick, S.; Aboutayab, K.;
West, R. I. Chem. Commun. 1995, 13, 1353–1354;
Caddick, S.; Aboutayab, K.; Jenkins, K.; West, R. I. J.
Chem. Soc., Perkin Trans. 1 1996, 7, 675–682; Bonfand,
E.; Forslund, L.; Motherwell, W. B.; Vazquez, S. Synlett
2000, 475–478; Ohno, H.; Wakayama, R.; Maedo, S.-I.;
Iwasaki, H.; Okumura, M.; Iwata, C.; Mikamiyama, H.;
Tanaka, T. J. Org. Chem. 2003, 65, 5909–5916; Harrowven,
The yields of isolated radical cyclisation products were
lower in this case compared to the aryl radical additions,
however no reduced starting materials were observed in
the crude NMR spectra.
Finally, we moved on to investigate the alkyl radical
additions. Selenides 15 and 16 proved to be the best
models as the iodides and bromides were highly unstable
due to the presence of a nucleophilic amide. The cyclisa-
tion precursors were prepared via the sequence shown in
Scheme 4. When subjected to the radical conditions both
selenides underwent spirocyclisation forming spirocycles