Using the transformation of 1 þ 2f3, we assessed
the performance of this catalyst, referred to as cata-
lyst I, thereby revealing two outstanding properties
(Scheme 2). First, catalyst I dramatically suppresses
homodimer formation. Second, catalyst I promotes
the coupling at a faster rate than that with the cata-
lysts derived from 5 and 6.
stored on the benchtop at rt without noticeable loss of
catalytic activity.12 Importantly, catalyst II allows us to
utilize ∼95% of nucleophile 2 for the carbonyl addition,
with an operationally simple procedure.13
Despite extensive efforts, we were not successful in
crystallizing catalyst II. At present, we have no direct
evidence to prove that catalyst II is a Ni,Cr-hetero-
bimetallic catalyst. However, the observation that
the overall reactivity profile of catalyst II is similar
to that of catalyst I supports that catalyst II is a Ni,
Cr-heteobimetallic catalyst, although it is undoubtedly
contaminated with bis-Ni- and bis-Cr-complexes.
During this study, we obtained a single crystal of the bis-
Ni-complex derived from 11. An X-ray analysis of this
complex (Figure 1B) revealed an interesting structural
feature, in connection with the NiTCr proximity required
for the transmetalation.
We then conducted two structural modifications of 11.
First, we examined the effect of the tether chain length and
found, once again, that 11 is the best catalyst among
11ꢀ13 (Scheme 2).
Second, the scyllo-inositol orthoester scaffold offers
several sites for structural modification. To explore
whether such a structural modification might result in
a change in the overall reactivity profile of the Ni,
Cr-heterobimetallic catalysts, we studied the coupling
reaction in the presence of six orthoesters (Scheme 3).
Overall, these structural modifications resulted in
more or less anticipated effects on the catalyst activity.
For example, catalysts derived from 15, 16, and 17
exhibited reactivity almost identical to that of 11,
except for better organic solvent solubility of the
catalyst derived from 17. The performance of the
catalyst prepared from 18, which bears an axial meth-
oxy group, was poor, suggesting the importance of the
space to accommodate substrates at the catalytic sites.
In addition, we prepared ligand 19 bearing three di-
pyridyls in close proximity, with the hope of exploring
the potential that a Ni/Cr-heterometallic catalyst with
a Ni/Cr = 1/2 or 2/1 ratio might offer. However, a
preliminary study indicated that the coupling rate with
this catalyst was slow.
Figure 1. X-ray structure of (A) 7•NiCl2,CrCl3 complex and
(B) 11•2ꢁNiCl2 complex.
Having demonstrated the usefulness of the orthoe-
ster scaffold derived from scyllo-inositol, we studied
the effect of the tether chain length. For this study, we
prepared the Ni/Cr-heterobimetallic catalyst from
homologous ligand 8. The catalyst thus obtained
was less effective in suppressing homodimerization
than catalyst I, indicating that the proximity of two
catalytic sites is important to achieve efficient NifCr
transmetalation.
Being encouraged by the attractive reactivity profile
observed for catalyst I, we became interested in a
tethered ligand containing two 6-methyl-2,20-dipyri-
dyls. This ligand does not contain differentiated liga-
tion sites for selective complexation with Ni or Cr. Yet,
we were interested in this ligand primarily because of
the ease of synthesis. Indeed, 11 was straightforwardly
prepared from scyllo-inositol in three steps.
On treatment with CrCl3 3THF (1.0 equiv) for 6 h in
THF,10,11 then with NiCl2 (MeOCH2)2 (1.0 equiv), 11
We selected 11 representative nucleophiles to test the
scope of Ni,Cr-heterobimetallic catalyst II (Scheme 4).
Overall, the catalyst performed well for all the substrates
tested, including a tetrasubstituted iodoolefin.14 Note-
worthily, di- and trisubstituted trans-iodoolefins furnished
the expected coupled products without geometrical isomer-
ization. On the other hand, di- and trisubstituted cis-
iodoolefins gave the expected coupled products, but with
3
3
gave green amorphous solid, referred to as catalyst II.
This catalyst was found to exhibit the excellent catalytic
property (Scheme 2). Interestingly, catalyst II shares the
overall profile of catalyst I, yet the performance of
catalyst II is clearly superior to catalyst I in two respects:
suppression of homodimerization and a faster coupling
rate.
Catalyst II (green amorphous solid) can be prepared
with excellent reproducibility. It is air-stable and can be
(13) Typical procedurefor couplings: A pear-shape flask was charged
with a magnetic stir bar, catalyst II (0.01ꢀ0.02 equiv, solid), Zr(cp)2Cl2
(1.2 equiv, solid, Aldrich), Mn powder (2.0 equiv, solid, Aldrich), and
LiCl (2.0 equiv, solid, Aldrich). With stirring, 1,2-dimethoxyethane
(Aldrich, sure-sealed) was added up to the final concentration = 0.5
M at rt under nitrogen. To this mixture were added aldehyde (1.0 equiv)
and iodide (1.1 equiv) successively. The reaction mixture was stirred at rt
(10) TLC analyses showed only a trace amount of free ligand 11
present.
(11) Four different methods were tested for preparation of Ni,Cr-
catalyst from 11. For details, see Supporting Information.
(12) There was no noticeable loss of activity observed for the catalyst
kept on benchtop for several months.
under nitrogen, typically for
Information.
(14) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama, T.; Nozaki, H.
2 h. For details, see Supporting
Tetrahedron Lett. 1983, 24, 5281.
88
Org. Lett., Vol. 14, No. 1, 2012
Peng, Jianbiao
