Angewandte
Chemie
with a seven-membered metallacycle. Therefore, the relief of
the nonplanar strain of the curved corannulene skeleton
should be another key driving force for the formation of
intermediate I. In the same way, more curved compounds
(with a deeper cavity), such as a half fullerene (C30H12) and
À
open-cage fullerenes, are known to undergo C C bond
activation through oxidative addition,[9] and their nonplanar
strain energy is much higher than that of corannulene
according to a recent calculation.[10] Our observation of
À
Figure 5. Plausible mechanism of the iridium-catalyzed reductive C C
bond cleavage reaction of 1. X=solvents, ClÀ, and/or hydrides, y=1–
À
3.
aromatic C C bond cleavage in a carbon nanomaterial with
a shallower curvature, namely corannulene, thus constitutes
a rare example.[7]
The first step of this reaction, the reduction of IrCl3·nH2O,
was experimentally confirmed by X-ray photoelectron spec-
troscopy (XPS). When we analyzed black solid samples
obtained by microwave heating of IrCl3·nH2O in ethylene
glycol at 25088C for five minutes, two distinct peaks corre-
sponding to Ir 4f7/2 and Ir 4f5/2 were observed at 61.4 and
64.4 eV, respectively. These binding energies were signifi-
cantly lower than those of IrIIICl3 (62.7 and 65.6 eV),
suggesting that the IrIIICl3 salt was reduced to a lower-valent
iridium species (Supporting Information, Figure S19). The
disappearance of the peaks that are due to the Cl atoms also
supports the reduction of IrIIICl3. The binding energies of the
reduced Ir components are close to those of metallic iridium,
Ir0 (61.35 and 64.35 eV),[13] indicating that the resulting
iridium species should be based on Ir0 or IrI. Indeed, these
reaction conditions are very similar to those of a known
method, the so-called polyol process, where metallic nano-
particles are produced from a transition-metal salt by heating
in ethylene glycol at high temperature.[14] According to the
literature, ethylene glycol reacts with a metal salt to afford
elemental metal, 2,3-butanedione, and HCl.[14d] As iridium
nanoparticles are also synthesized by a similar process,[14e] the
formation of Ir0 species is likely to occur under these
conditions.
The final step in the proposed mechanism entails reduc-
tion of the vinyl moiety and protonation to give product 2. It
was experimentally confirmed that our reaction conditions
allowed for the hydrogenation of styrene to form ethyl-
benzene, even though the conversion was only approximately
50% (Table S1). Accordingly, the possibility of direct hydride
insertion from the iridium center to the metalated vinyl group
cannot be excluded. The following protonation is likely to
take place because HCl should be generated according to the
mechanism of the polyol process. Consequently, the released
iridium species with a higher oxidation state, IrII or IrIII,
should again be reduced by the polyol process to enter the
catalytic cycle once again. Another possible pathway, the
reductive elimination of IrII or IrIII from a hydride complex
and subsequent dissociation of Ir0 or IrI, might also be
involved.
Finally, we investigated the metal specificity of this
reaction by conducting screening experiments with various
transition-metal chlorides (FeCl3, CoCl2·6H2O, NiCl2·6H2O,
CuCl2·2H2O, RuCl3, RhCl3·3H2O, PdCl2, AgCl, IrCl3·nH2O,
PtCl2, and Na[AuCl4]·2H2O) under almost identical condi-
tions (with ca. 1 equiv or an excess amount of the metal salts).
Interestingly, it was found that only IrCl3·nH2O provided full
conversion of 1 into 2. Other metal salts (except for RuCl3,
which afforded a rather complicated mixture including 1 and
2) did not react with 1 at all, and the starting material was fully
recovered after the reaction (Figures S23–S33). These results
As electron-rich transition metals in low oxidation states
À
generally prefer to insert into C C bonds through oxidative
addition,[8a] we propose that the resulting low-valent iridium
species, Ir0 or IrI,[15] inserted into the aromatic C C bond of
strongly indicate that the aromatic C C bond cleavage
À
À
the corannulene skeleton of 1 to form intermediate I with an
IrII or IrIII center (Figure 5). This mechanism is well supported
by a recent theoretical study, in which metal insertion into
reaction of 1 is specific to IrCl3·nH2O among the late
transition metal chlorides that we tested, which is consistent
À
À
with recent progress in C C and C H bond activation with
Group 9 metals (Rh, Ir). The superiority of Ir over Rh has
À
a C C bond at the edge of a carbon nanotube was proposed as
À
the first step of the transition-metal-catalyzed unzipping of
carbon nanotubes.[3j] Furthermore, the proposed structure of
intermediate I also meets several conditions that need to be
been reported for some C C activation processes and was
[8c]
À
explained in terms of the stability of the M C bonds.
In summary, we have demonstrated the efficient and site-
À
À
fulfilled for C C bond activation to occur, such as relieving
selective C C cleavage reaction of the corannulene skeleton
strain energy, inducing aromatic stabilization, forming stable
metallacyclic complexes, and chelation-assisted activation.[8a]
For instance, intermediate I comprises two metallacycles, one
of which is stabilized by chelation with the 2-pyridyl
substituent. In fact, the 2-pyridyl moieties often serve as
of 1 by using IrCl3·nH2O and ethylene glycol as catalyst and
solvent, respectively, with the aid of microwave heating at
2508C.[17] Several control experiments suggest that an Ir
species obtained by in situ reduction reacted with 1 to form
intermediate I in a process that is driven by both 2-pyridyl
chelation and the relief of the nonplanar strain in the
À
chelating auxiliaries in transition-metal-mediated C H and
C C bond activation reactions.
By the formation of intermediate I, the curved p-system of
the corannulene skeleton of 1 is likely to be flattened to some
extent through the iridium insertion because the planar
benzo[ghi]fluoranthene skeleton of I is only loosely tethered
[16]
À
À
À
corannulene. Although several C H and C C bond activa-
tion reactions have been developed to date by means of
elaborate iridium catalysts, the commercially available iri-
dium salt IrCl3·nH2O used here is one of the simplest iridium
catalysts. We envision that a further expansion of this reaction
Angew. Chem. Int. Ed. 2015, 54, 5351 –5354
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5353