R. Yamasaki et al. / Tetrahedron Letters 50 (2009) 1143–1145
1145
R1
might contribute to this selectivity since the 4c and 4f which can
form six-membered chelation cycle gave superior yield and
selectivity.10
In conclusion, 1,3-diyne was applied to Ni-catalyzed [3+2+2]
cocyclization reaction. We found that selective 1:1:1 coupling be-
tween 1,3-diyne, alkyne, and 1 afforded cycloheptadiene with an
intact triple bond. The study has expanded the scope of the
three-component [3+2+2] cocyclization reaction, and has provided
an efficient pathway for the synthesis of a range of multisubsti-
tuted cycloheptadienes.
R1
R1
R2
Ni
R2
R1
R1
CO2Et
R2
Ni (0)
R1
α
Ni
β
R1
R2
R1
Acknowledgment
CO2Et
Ni
We are grateful to Ms. Yukiko Fukusaki for the synthesis of 4f.
CO2Et
Ni
Supplementary data
R1
CO2Et
R2
R1
R1
R2 R1
Supplementary data associated with this article can be found, in
Scheme 2. A plausible mechanism for three-component cocyclization.
References and notes
homopropargyl ethers such as 4c and 4f were employed as the
substrates, the yields of the products were high (entries 3, 6, 9,
and 12). On the other hand, ethers with different lengths of the
methylene group turned out to be less effective substrates, and
the yields of the products decreased (compare entries 8–10). These
results imply that chelation of nickel by the oxygen atom and the
alkynyl group plays an important role for the efficient synthesis
of 5.10 The reaction of benzyl propargyl ether (4g) also proceeded,
and the product was isolated in 46% yield (entry 13). However,
other monosubstituted alkynes turned out to be inferior substrates
for this reaction (entries 14–16). This result also supports the exis-
tence of a chelation effect since all the alkynes that were superior
substrates for this reaction possess an oxygen atom. It is notewor-
thy that 2:1 cocyclization of the alkyne and 1 proceeded in a few
examples (entries 4, 10, and 14).
1. Multicomponent Reactions; Zhu, J., Bienaymé, H., Eds.; Wiley-VCH: Weinheim,
2005.
2. D’Souza, D. M.; Muller, T. J. J. Chem. Soc. Rev. 2007, 36, 1095–1108. and
references cited therein.
3. (a) Saito, S.; Masuda, M.; Komagawa, S. J. Am. Chem. Soc. 2004, 126, 10540–
10541; (b) Komagawa, S.; Saito, S. Angew. Chem., Int. Ed. 2006, 45, 2446–2449;
(c) Saito, S.; Takeuchi, K. Tetrahedron Lett. 2007, 48, 595–598; (d) Maeda, K.;
Saito, S. Tetrahedron Lett. 2007, 48, 3173–3176; (e) Saito, S.; Komagawa, S.;
Azumaya, I.; Masuda, M. J. Org. Chem. 2007, 72, 9114–9120; (f) Komagawa, S.;
Yamasaki, R.; Saito, S. J. Synth. Org. Chem. Jpn. 2008, 66, 974–982.
4. (a) Kulkarni, A. A.; Diver, S. T. Org. Lett. 2003, 5, 3463–3466; (b) Ni, Y.;
Montgomery, J. J. Am. Chem. Soc. 2006, 128, 2609–2614.
5. Low, P. J.; Bruce, M. I. Adv. Organomet. Chem. 2001, 48, 71–288.
6. Ni-catalyzed cocyclization of 1,3-diynes has been reported: (a) Deaton, K. R.;
Gin, M. S. Org. Lett. 2003, 5, 2477–2480; (b) Jeevanandam, A.; Korivi, R. P.;
Huang, I.-w.; Cheng, C.-H. Org. Lett. 2002, 4, 807–810; (c) Sato, Y.; Ohashi, K.;
Mori, M. Tetrahedron Lett. 1999, 40, 5231–5234; (d) Chalk, A. J.; Jerussi, R. A.
Tetrahedron Lett. 1972, 13, 61–62.
7. The choice of ligand, PPh3, was crucial in this reaction. Other ligands gave
poorer yield and selectivity.
Considering the structures of the products and the previously
proposed mechanism,3 we postulate a similar one for the three-
component cocyclization of 1,3-diyne, alkyne, and 1 as shown in
Scheme 2. It is proposed that the catalytic cycle is initiated by
the reaction of the alkyne and diyne with the Ni(0) species to form
nickelacycle. Then, 1 is inserted into the metallacycle to occur
cyclopropylmethyl to butenyl rearrangement,11 affording the se-
ven-membered nickelacycle. Subsequently, nickelacyclooctadiene
is transformed to the cycloheptadiene derivative by the reductive
elimination of the Ni complex.
8. X-ray data were collected on a Bruker Smart1000 CCD detector. The crystal
structure was solved by direct methods SHELXS-97 (Sheldrick, 1997) and was
refined by full-matrix least-squares SHELXL-97 (Sheldrick, 1997). All non-
hydrogen atoms were refined anisotropically. All hydrogen atoms were
included as their calculated positions. Crystal data for 5yaE: C22H34O4Si2;
M = 418.67 g molꢀ1, triclinic, P1, colorless prism measuring 0.4 ꢁ 0.3 ꢁ 0.2 mm,
ꢀ
T = 150 K, a = 10.3539(7), b = 11.0209(7), c = 11.5075(7) Å,
a = 96.523(1)°,
b = 93.299(1)°,
= 0.164 mmꢀ1, Tmax = 0.9679, Tmin = 0.9371, GOF on F2 = 0.979, R1 = 0.0419,
wR2 = 0.1003 ([I > 2 (I)], R1 = 0.0506, and wR2 = 0.1067 (all data).
c ,
= 106.417(1)°, V = 1245.87(14) Å3, Z = 2, Dc = 1.116 Mg mꢀ3
l
r
Crystallographic data (excluding structure factors) for the structures in this
Letter have been deposited at the Cambridge Crystallographic Data Center
(CCDC-709808). Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK, (fax: +44-
(0)1223-336033 or e-mail: deposit@ccdc.cam.ac.uk).
As discussed previously,3e regioselectivity and chemo selectiv-
ity can be explained in terms of the steric and electronic effect
on the metallacycle intermediate. The steric factor explains the
9. The reaction of 1, 1,6-dimethoxy-2,4-hexadiyne, and 4a gave the desired
product albeit in low yield (21%). The three-component coupling reactions of
other diynes, such as 1,4-diphenyl-1,3-butadiyne and terminal 1,3-diynes,
resulted in the formation of a complex mixture.
10. Huffman, M. A.; Liebeskind, L. S. J. Am. Chem. Soc. 1991, 113, 2771–2772.
11. Binger, P.; Doyle, M. J.; Benn, R. Chem. Ber. 1983, 116, 1–10.
12. (a) Mori, N.; Ikeda, S.-i.; Sato, Y. J. Am. Chem. Soc. 1999, 121, 2722–2727; (b)
Wakatsuki, Y.; Nomura, O.; Kitamura, K.; Morokuma, K.; Yamazaki, H. J. Am.
Chem. Soc. 1983, 105, 1907–1912.
13. Stockis, A.; Hoffmann, R. J. Am. Chem. Soc. 1980, 102, 2952–2962.
14. (a) Cho, C.-W.; Krische, M. J. Org. Lett. 2006, 8, 3873–3876; (b) Varela, J. A.;
Castedo, L.; Saá, C. J. Am. Chem. Soc. 1998, 120, 12147–12148.
preference of a,b-substituted nickelacycle over a, a- or b, b-disub-
stituted nickelacycle12 and alkyne over 1,3-diyne. As for the elec-
tronic factor, Hoffmann et al. proposed that the largest lobe of
*
p ) is prone to form a C–C bond (at the b-position)
LUMO of alkyne (
in order to maximize the orbital overlap.13 Considering that 1,3-
diyne possesses its largest LUMO at the terminal carbon,14 it is rea-
sonable that the remaining ethynyl group derived from the 1,3-
diyne is located at the
a-position. Although the chemoselectivity
of heterometallacycle between alkyne and 1,3-diyne over homom-
etallacycle is not clear at this stage, chelation between Ni and ether