Ni-Catalyzed [4+3+2] Cycloaddition
FULL PAPER
which could in turn react with the diene moiety of the com-
plex to give I. Since the reductive elimination from K to
give cyclopentene derivative is a slow process,[5h] the inser-
tion of the diene moiety would be favored. Although the
preference for the intermolecular cycloaddition over the in-
tramolecular cyclization at the first stage of the reaction ap-
pears improbable, Evans et al. suggested the preference for
the intermolecular metallacycle formation over the intramo-
lecular metallacycle formation in the Rh-catalyzed [4+2+2]
cycloaddition.[29] In this pathway, however, it is necessary to
suppose that the insertion and/or the rearrangement of the
diene moiety would proceed in highly selective manner,
which makes this pathway less likely.
Although it was not possible to eliminate some of the
pathways discussed here and provide a sharper focus to our
understanding of the reaction mechanism, we are inclined to
postulate that the nickelacycle E represents the key inter-
mediate for the progress of the [4+3+2] cycloaddition
(Scheme 1, pathway a). If a seven-membered intermediate
such as F was formed as the intermediate, the isomerization
of F to H would not proceed rapidly, and the reductive elim-
ination (i.e., the formation of F’) would be preferred. The
stereochemistry of the product (e.g., 6b and 6c) can be
clearly explained by the idea that this reaction proceeded
through complex E. The stability and equilibrium of the
nickelacycle species is generally dependent on the amount
and the structure of the ligand and the isolated nickelacycles
(53 or 55), which correspond to D, might represent a resting
species, which is only observed when PCy3 is used as the
ligand. Thus, the use of PPh3 or even 1 as the ligand, might
drive the equilibrium toward the formation of the s-complex
E, whereas trialkylphosphines and TOPP would stabilize the
p-allyl complex D.[30] In contrast, the use of PPh3 as the
ligand would be detrimental to the reductive elimination of
E, as suggested experimentally.[17] The reduced rate of the
reductive elimination might be advantageous to the inser-
tion of 1 to form a ten-membered metallacycle (H).
Experimental Section
A typical example (condition A): A solution of 1 (1 mmol) and 3a
(1.2 mmol) in dry toluene (0.5 mL) were added dropwise over 5 h to a
dark red mixture of [NiACHTNUGTRNEG(UN cod)2] (27.5 mg, 0.1 mmol) and PPh3 (52.5 mg,
0.2 mmol) in dry toluene (0.5 mL) at 508C under Ar. The progress of the
reaction was monitored by TLC and GC-MS, and the mixture was stirred
until the starting material 1 disappeared. The mixture was passed through
a short silica gel column (ether). Evaporation of the solvent gave an oil,
which was further purified by silica gel column chromatography to give
4Ea (and 4Za).
Stoichiometric reaction of 3a with [Ni
toluene solution of 3a (79 mg, 0.30 mmol) was added to a toluene solu-
tion (15 mL) of [Ni(cod)2] (83 mg, 0.30 mmol) and PPh3 (79 mg,
ACHTUGNRTEN(NUGN cod)2] in the presence of PPh3: A
AHCTUNGTRENNUNG
0.30 mmol), resulting in gradual change of the color to orange-yellow.
The reaction mixture was stirred for 2 h at room temperature, and then
all volatiles were removed under reduced pressure. The residue was
washed with hexane to give 46 (250 mg, 98%) as a yellow microcrystal-
line solid.
Gradual conversion of 46 to 49: Complex 46 (25.4 mg, 0.03 mmol) was
dissolved in C6D6 (0.4 mL), and the progress of the reaction was moni-
tored at room temperature by means of NMR spectroscopy. After 65 h,
formation of 49 was detected in 51% yield. Isolation of 49 was conducted
by dissolving 46 (0.3 mmol) in toluene solution, and then the crude prod-
uct was further purified by passing through a pad of silica gel column
chromatography (eluent; hexane) followed by HPLC (eluent; CHCl3),
giving 25 mg of 49 (0.095 mmol, 32% yield of the isolated product).
Stoichiometric reaction of 46 with 1: A C6D6 solution of 1 (3.8 mg,
0.03 mmol) was added to a C6D6 solution of 46 (25.6 mg, 0.03 mmol).
Monitoring of the reaction by 31P NMR spectroscopy demonstrated that
47% of 46 were consumed to generate 52 in 25% yield after 15 min.
Observation of a reaction intermediate of 3a with [Ni
ence of PCy3: A [D8]THF solution of 3a (7.9 mg, 0.03 mmol) was added
to [D8]THF solution of [Ni(cod)2] (8.3 mg, 0.03 mmol) and PCy3
ACHTUNGRTEN(NGNU cod)2] in the pres-
a
ACHTUNGTRENNUNG
(8.4 mg, 0.03 mmol) was added at À208C. The color of the solution
turned into orange yellow. NMR spectroscopic analysis of the sample at
À208C revealed the formation of a mixture of 54 and 55, the ratio of
which (76:24) was estimated by 31P NMR analysis. When the [D8]THF
solution was allowed to warm to room temperature, all of 54 was con-
verted to 55 quantitatively within 1 h.
Reaction of 55 with 1: Compound 1 (3.8 mg, 0.03 mmol) was added to a
C6D6 solution (0.4 mL) of 55 (18 mg, 0.03 mmol), resulting in a quantita-
tive formation of 56. Isolation of 56 was conducted by using 55
(0.3 mmol), and then the crude product was further purified by passing
through a pad of SiO2 (eluent; hexane) followed by HPLC (eluent;
CHCl3), giving 76 mg of 56 (0.19 mmol, 65% yield of the isolated prod-
uct).
Conclusion
X-ray crystallography: Single-crystal X-ray diffraction data of the crystal
of 21 was collected on a CCD diffractometer with graphite monochro-
mated MoKa (l=0.71073 ꢁ) radiation. The crystal structure was solved
by direct methods SHELXS-97 and refined by full-matrix least-squares
SHELXL-97.[31] All non-hydrogen atoms were refined anisotropically and
hydrogen atoms were included at their calculated positions. Single-crystal
X-ray diffraction data of the crystal of 48 and 52 were collected on a
imaging plate diffractometer with graphite monochromated MoKa (l=
0.71073 ꢁ) radiation.
We have successfully developed
a new Ni-catalyzed
[4+3+2] cycloaddition reaction of dienynes with ethyl cyclo-
propylideneacetate (1), and studied the scope and limita-
tions of the reaction in detail. Based on the results of the re-
actions and the studies of the Ni-dienyne complexes, the
mechanism of the reaction has been discussed. By postulat-
ing a mechanism that involves nickelacycloheptadiene as a
key intermediate, it is possible to explain the ligand and sub-
stituent effect and the stereochemical outcome of the reac-
tion. The linker moiety plays an important role in the effi-
ciency of the reaction by keeping the alkyne and diene moi-
eties in positions necessary to their effective reaction, and
the right choice of the linker improved the yield of the reac-
tion. The reaction provided a reliable and ready access to
the nine-membered carbocycles.
CCDC-843293 (21), CCDC-908193 (48), CCDC-908194 (52) and CCDC-
908195 (53) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystal data for 21: C24H31NO4S; Mw =429.56 gmolÀ1, monoclinic, P21/n,
colorless prism measuring 0.40ꢂ0.40ꢂ0.40 mm, T=150 K, a=
10.3279(13), b=18.556(2), c=11.6540(15) ꢁ, b=95.375(2)8, V=
2223.6(5) ꢁ3, Z=4, 1calcd =1.283MgmÀ3, GOF on F2 =1.024, R1 =0.0513,
wR2 =0.1373 [I>2s(I)], R1 =0.0764, and wR2 =0.1567 (all data).
Crystal data for 48: C50H44OP2Ni; M=781.50 gmolÀ1, triclinic, P1, yellow
¯
platelet measuring 0.26ꢂ0.18ꢂ0.07 mm, T=123 K, a=13.0298(14), b=
Chem. Eur. J. 2013, 19, 3415 – 3425
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3423