.
Angewandte
Communications
À
Table 2: Palladium-catalyzed C H amidation of 2-methylphenyl ethers.
Entry
PdII catalyst Source
Product
Yield [%][a]
1
2
3
4
5
6
7
8
Pd(hfacac)2
Pd(NCMe)2Cl2
Pd(OAc)2/bathocuproine
(bc)Pd(OAc)2
Pd(OAc)2/bathocuproine
(bc)Pd(OAc)2
6a
6a
6a
6a
6b
6b
6c
6c
0
10
68
81
55
45
<10
21
Pd(OAc)2/bathocuproine
(bc)Pd(OAc)2
[a] Yield of isolated product after column chromatography.
yield of isolated 6a could be improved further to 81%
(Table 2, entry 4). In contrast to the amidation of 8-methyl-
quinoline, the substrate 2-methylanisole tolerated only NFSI
as the oxidant and nitrogen source.
À
Scheme 2. Palladium-catalyzed C H amidation of 2-methylanisoles.
Yields refer to isolated material after column chromatography.
[a] 10 mol% [(bc)Pd(OAc)2], substrate (1 equiv), NFSI (2 equiv), diox-
ane/DMF (4:1) or dioxane/MeCN (3:1), 908C, 15 h. [b] 10 mol%
Pd(OAc)2, 10 mol% bc, substrate (1 equiv), NFSI (2 equiv), dioxane/
DMF (4:1), 908C, 15 h.
This combination of palladium(II) acetate, bathocu-
À
proine, and NFSI also effected the C H amidation of other
2-methylphenyl ethers such as 5b and 5c, albeit in lower
yields (Table 2, entries 5–8). As a result, a series of 2-
methylanisole derivatives were investigated, which under-
3
À
went selective C(sp ) H amidation under the optimized
conditions (Scheme 2). The reaction worked well for all
kinds of 2-methylanisoles bearing para, meta, and ortho
substituents on the arene ring. Notably, the oxidation
proceeded selectively at the methyl position, even for 6j,
that reductive elimination takes place rapidly even at
298 K.[19] In agreement with this assumption, a control experi-
ment with an equimolar mixture of 7 and the corresponding
complex deuterated at the methylene position proceeded
without detectable secondary kinetic isotope effect.[11] Oxi-
dation of 7 with a stoichiometric amount of NFSI led to 2b in
88% yield. While these experiments point to monomeric
complex 7 as an intermediate in the catalytic cycle,[20] the
exact nature of the high-oxidation-state intermediate could
not be determined experimentally.
À
which displays an additional biphenyl ether. The C H
functionalization was also selective in favor of the methyl
substituent over a potential ortho-tert-butyl group, as dem-
onstrated for 5k. The corresponding product 6k was charac-
terized unambiguously by X-ray crystal structure analysis.[13]
À
All these examples demonstrate the control of C H amida-
tion by weak metal coordination.[15]
Theory provides a more versatile tool to address the
underlying individual steps involved in the oxidation of 7
The successful realization of the oxidative direct amida-
tion of an alkyl group under palladium catalysis led us to
engage in a preliminary mechanistic investigation with 8-
methylquinoline (1a) as the substrate. Starting from the
(Figure 1).[11,21,22]
On
the
basis
of
N-
fluorobis(methylsulfonyl)imide [FN(SO2Me)2] as an electro-
philic two-electron oxidant, a linear transition state must be
involved in the oxidation from PdII to PdIV (TS7-A, Figure 1).
This results in the formation of a cationic fluorinated PdIV
intermediate A with a square-planar pyramidal geometry, in
which the methylene group occupies the apical position.[23]
The computed activation energy for the oxidation step in
dioxane using a continuum solvation model is 35.2 kcal
molÀ1.[24] The cationic PdIV intermediate A and bissulfonyli-
mide should not combine to a neutral Pd complex[9b] but
rather engage in direct nucleophilic substitution at the
À
palladium(II) salt, chelation-assisted C H activation forms
palladacycle 7, which was confirmed through a stoichiometric
control reaction. Product 7 was unambiguously characterized
by X-ray structure analysis,[13] which proved that its compo-
sition is indeed monomeric.
The formation of 7 represents the initial step of the
proposed catalytic cycle for this new intermolecular amida-
tion (Figure 1). Under catalysis conditions, a large primary
kinetic isotope effect kH/kD of 5.9 was determined for
competition between 1a and its [D3]methyl derivative,[11]
indicating that formation of 7 is either rate-limiting or
reversible. The next step, oxidation of 7 with NFSI to an
anticipated fluorinated high oxidation state intermediate,[18]
could not be monitored by NMR spectroscopy, suggesting
À
electrophilic carbon in the a position to install the new C N
bond (TSA-B, Figure 1), since the stabilization of cationic
PdIV is better accomplished by formal reductive elimination to
PdII than by anion recombination to neutral PdIV. The
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2225 –2228