.
Angewandte
Communications
[a]
À
Table 1: Optimization of the cobalt-catalyzed C H amidation.
Entry Catalyst (mol%)
T [8C] t [h] Yield [%][b,c]
1
2
3
4
5
6
7
8
9
[{Cp*CoCl2}2] (5)/AgSbF6 (20)
RT
40
60
80
40
40
40
40
40
40
80
80
12
12
12
12
12
24
24
24
24
24
24
24
n.r.[d]
71
64
60
66
83 (79)
n.r.[d]
n.r.[d]
n.r.[d]
n.r.[d]
n.r.[d]
30
[{Cp*CoCl2}2] (5)/AgSbF6 (20)
[{Cp*CoCl2}2] (5)/AgSbF6 (20)
[{Cp*CoCl2}2] (5)/AgSbF6 (20)
[{Cp*CoCl2}2] (2.5)/AgSbF6 (10)
[{Cp*CoCl2}2] (1)/AgSbF6 (4)
[{Cp*CoCl2}2] (1)/AgOAc (4)
CoCl2 (2)/AgSbF6 (4)
[{Cp*CoCl2}2] (1)
AgSbF6 (4)
[{Cp*RhCl2}2] (2.5)/AgSbF6 (10)
[{Cp*IrCl2}2] (2.5)/AgSbF6 (10)
10
11
12
[a] Reaction conditions: 1a (0.2 mmol), 2a (1.1 equiv), 1,2-dichloro-
ethane (0.5 mL). [b] The yield was determined by NMR spectroscopy of
the crude reaction mixture with dibromomethane as an internal
standard. [c] The yield of the isolated product is given in parentheses.
[d] n.r.=no reaction.
None of the desired product was formed when other silver
salts, such as AgOAc or AgPF6, were used in place of AgSbF6
(Table 1, entry 7). CoCl2 was totally ineffective, and
[{Cp*CoCl2}2] or the Ag salt alone also did not display any
reactivity (Table 1, entries 8–10), thus indicating that the
cationic [Cp*CoIII] species is essential for the present
amidation reaction with 1,4,2-dioxazol-5-ones. Interestingly,
other Group 9 metal analogues of the type [Cp*MIII] (M = Rh
and Ir) were much less effective, even at a higher catalyst
loading and higher temperatures (Table 1, entries 11 and 12).
Under the optimized conditions with the cobalt catalyst,
Scheme 2. Catalytic activity of Group 9 [{Cp*MCl2}2] complexes in the
amidation of anilides. Reaction conditions: 1 (0.2 mmol), 2a
(1.1 equiv), ClCH2CH2Cl (0.5 mL). The yields shown for the cobalt-
catalyzed reactions are for the isolated product. [a] The yield was
determined by NMR spectroscopy of the crude reaction mixture with
dibromomethane as an internal standard; n.r.=no reaction.
À
the scope of the C H amidation with 1,4,2-dioxazol-5-one 2a
was subsequently evaluated in terms of the substituents on the
substrate (Scheme 2). Electronic variation of the anilide
aromatic ring barely affected the reaction efficiency, and the
desired amidated products 3b–g were obtained in satisfactory
yield. Trifluoromethyl, bromo, and chloro groups were
compatible with the present conditions (products 3c–e). The
reaction is regioselective in that the amidation took place at
the sterically less hindered site of a meta-substituted anilide to
give product 3e. Alteration of the N-acyl moiety of the anilide
substrate was also found to be feasible. For example, the
amidation of anilides bearing an isopropyl or a cyclohexyl
group proceeded to give products 3h and 3i, respectively, in
amidation of anilides with 1,4,2-dioxazol-5-one, but with
significantly lower efficiency as compared to the correspond-
ing cobalt catalyst system. However, the exact reason why the
[Cp*CoIII] catalyst is much superior to its rhodium and
iridium analogues is not clear at the present stage, thus
requiring more comprehensive mechanistic elucidation. This
comparative study clearly demonstrates that the cobalt
catalyst could be superior to rhodium or iridium in its
À
reactivity in certain C H functionalization reactions.
Next, we investigated an additional substrate type,
benzamides (Scheme 3). Although the optimized conditions
described above were found to be less effective in this case,
the addition of a catalytic amount of an acetate salt greatly
improved the reaction efficiency at higher temperature.[15]
When the [{Cp*CoCl2}2] precursor (1 mol%) was used in the
presence of AgSbF6 (4 mol%) and NaOAc (6 mol%), the
amidation of N-tert-butylbenzamide (4a) with 2a at 808C
(24 h) afforded the desired product 5a in good yield. This high
efficiency of amidation was also observed with benzamide
derivatives bearing methyl, methoxy, or chloro substituents at
the para position (products 5b–d), thus demonstrating again
À
high yield. Interestingly, the C H amidation of N-phenylurea
occurred to afford the desired product 3j, albeit in low yield.
To the best of our knowledge, this transformation is the first
successful functionalization of anilides by the action of
a [Cp*CoIII] catalyst.
Notably, when [{Cp*RhCl2}2] was employed as the catalyst
instead of the cobalt analogue, no catalytic reactivity was
observed under otherwise identical conditions. On the other
hand, [{Cp*IrCl2}2], another well-studied catalyst of the
À
Group 9 triad, was found to mediate the present C H
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 14103 –14107