Methyl-Selective α-Oxygenation of Tertiary Amines to Formamides by Employing Copper/Moderately Hindered Nitroxyl Radical (DMN-AZADO or 1-Me-AZADO)

Article abstract of DOI:10.1002/anie.201909005

Methyl-selective α-oxygenation of tertiary amines is a highly attractive approach for synthesizing formamides while preserving the amine substrate skeletons. Therefore, the development of efficient catalysts that can advance regioselective α-oxygenation at the N-methyl positions using molecular oxygen (O₂) as the terminal oxidant is an important subject. In this study, we successfully developed a highly regioselective and efficient aerobic methyl-selective α-oxygenation of tertiary amines by employing a Cu/nitroxyl radical catalyst system. The use of moderately hindered nitroxyl radicals, such as 1,5-dimethyl-9-azanoradamantane N-oxyl (DMN-AZADO) and 1-methyl-2-azaadamanane N-oxyl (1-Me-AZADO), was very important to promote the oxygenation effectively mainly because these N-oxyls have longer life-times than less hindered N-oxyls. Various types of tertiary N-methylamines were selectively converted to the corresponding formamides. A plausible reaction mechanism is also discussed on the basis of experimental evidence, together with DFT calculations. The high regioselectivity of this catalyst system stems from steric restriction of the amine-N-oxyl interactions.

Full text of DOI:10.1002/anie.201909005

Angewandte
Research Articles
Chemie
International Edition: DOI: 10.1002/anie.201909005
German Edition: DOI: 10.1002/ange.201909005
Oxygenation Reactions
Methyl-Selective a-Oxygenation of Tertiary Amines to Formamides by
Employing Copper/Moderately Hindered Nitroxyl Radical (DMN-
AZADO or 1-Me-AZADO)
Satoru Nakai, Takafumi Yatabe, Kosuke Suzuki, Yusuke Sasano, Yoshiharu Iwabuchi, Jun-
ya Hasegawa, Noritaka Mizuno, and Kazuya Yamaguchi*
Abstract: Methyl-selective a-oxygenation of tertiary amines is
a highly attractive approach for synthesizing formamides while
preserving the amine substrate skeletons. Therefore, the devel-
opment of efficient catalysts that can advance regioselective a-
oxygenation at the N-methyl positions using molecular oxygen
(O2) as the terminal oxidant is an important subject. In this
study, we successfully developed a highly regioselective and
efficient aerobic methyl-selective a-oxygenation of tertiary
amines by employing a Cu/nitroxyl radical catalyst system. The
use of moderately hindered nitroxyl radicals, such as 1,5-
dimethyl-9-azanoradamantane N-oxyl (DMN-AZADO) and
1-methyl-2-azaadamanane N-oxyl (1-Me-AZADO), was very
important to promote the oxygenation effectively mainly
because these N-oxyls have longer life-times than less hindered
N-oxyls. Various types of tertiary N-methylamines were
selectively converted to the corresponding formamides. A
plausible reaction mechanism is also discussed on the basis of
experimental evidence, together with DFT calculations. The
high regioselectivity of this catalyst system stems from steric
restriction of the amine-N-oxyl interactions.
transforming tertiary amines to various valuable products.[3–5]
Particularly, the development of oxidative functionalization
of tertiary amines using molecular oxygen (O2) is highly
meaningful because traditional stoichiometric oxidants cause
undesirable oxidation of sensitive functional groups in
bioactive and natural compounds. As a continuation of our
interest in developing aerobic oxidative functionalization of
tertiary amines,[6,7] we focused on methyl-selective a-oxygen-
ation of tertiary amines. Formamides, products of a-methyl-
selective oxygenation of N-methylamines, are useful inter-
mediates for transforming to versatile functional groups[8–12]
and important building blocks for biologically active mole-
cules.[13–16] Furthermore, demethylation of N-methylamines
can be realized via the deformylation of formamides.[17,18]
Therefore, methyl-selective a-oxygenation would be a prac-
tical strategy for transforming N-methylamines to various
value-added compounds.
However, the presence of three a-carbons with different
intrinsic chemical environments made it difficult to selectively
produce desired amides from tertiary amines by a-oxygen-
ation. Particularly, regioselective methyl-selective a-oxygen-
ation is extremely difficult for the following reasons (i)–(iii). i)
Introduction
À
C H bonds of a-methyl groups are strong with high bond
Tertiary amines are recognized as important structural
motifs in versatile compounds, such as pharmaceuticals,
agricultural chemicals, bioactive compounds, and natural
alkaloids.[1,2] New functional molecules can be designed by
functionalization of tertiary amines while preserving their
dissociation energies compared with other a-alkyl groups,
such as a-benzyl groups.[19] For instance, a-benzyl positions
are reported to be regioselectively oxygenated by photo-
catalytic aerobic a-oxygenation of tertiary amines.[20] ii) Imi-
nium cation intermediates produced by oxidation at the a-
methyl groups are thermodynamically unstable compared
with the multi-substituted iminium cations. Previously, we
reported that cyclic and linear a-methylene positions are
selectively oxygenated in aerobic a-oxygenation of tertiary
amines catalyzed by supported gold nanoparticles.[21] iii) Un-
like cross-dehydrogenative coupling, a-oxygenation requires
additional oxidation of hemiaminal intermediates, produced
by hydration of the iminium cations. We previously developed
regioselective a-cyanation and a-alkynylation at N-methyl
positions of tertiary amines by MnO2-based catalysts.[6,7]
However, a MnO2-catalyzed methyl-selective a-oxygenation
strategy failed because of the poor catalytic activity of MnO2
for the hemiaminal oxidation. In previous Fe-[22–25] and Co-
based catalytic[26] and oxoammonium-oxidizing reactions,[27,28]
the formamide yields were also unsatisfactory, because the
unstable hemiaminal intermediates mainly decomposed to
undesired demethylated byproducts. Although formamides
are effectively synthesized by oxoammonium-mediated elec-
trochemical oxidation, the substrate scope is limited to aniline
À
skeletons. Functionalization of the C H bond adjacent to the
nitrogen atom through cross-dehydrogenative coupling is
gaining considerable attention as a powerful method of
[*] S. Nakai, Dr. T. Yatabe, Dr. K. Suzuki, Prof. Dr. N. Mizuno,
Prof. Dr. K. Yamaguchi
Department of Applied Chemistry, School of Engineering
The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656 (Japan)
E-mail: kyama@appchem.t.u-tokyo.ac.jp
Dr. Y. Sasano, Prof. Dr. Y. Iwabuchi
Department of Organic Chemistry, Graduate School of Pharmaceut-
ical Sciences, Tohoku University
6-3 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578 (Japan)
Prof. Dr. J. Hasegawa
Institute for Catalysis, Hokkaido University
Kita 21 Nishi 10, Kita-ku, Sapporo 001-0021 (Japan)
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under https://doi.org/10.
1002/anie.201909005.
&&&&
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
These are not the final page numbers!

Angewandte
Research Articles
Chemie
derivatives only.[29] As far as we know, there have been only
the amounts of an appropriate counter anion, such as ClÀ or
a few good reports on the methyl-selective a-oxygenation. For
example, the Pt black-catalyzed method showed preferential
N-methyl oxygenation in the presence of benzylic groups
BrÀ. With the optimized CuCl/Cu(OTf)2/bpy/DMN-AZADO
(or 1-Me-AZADO) catalyst system (OTf = triflate, bpy =
2,2’-bipyridyl), various tertiary aliphatic amines, including
cyclic N-methylamines, N,N-dimethylamines, a natural alka-
loid (pseudopelletierine), and pharmaceutical drugs (such as
N-methylparoxetine and amitriptyline), could be highly
regioselectively converted to the corresponding formamides
(30 examples). On the basis of several pieces of experimental
evidence, together with DFT calculations, we concluded that
the high regioselectivity of the proposed catalyst system
stemmed from steric restriction of the amine-N-oxyl inter-
actions.
though large amounts of expensive Pt were required.[30]
A
CuI-catalyzed oxygenation of dimethylamines under rela-
tively harsh conditions (1208C) was also reported.[31] How-
ever, the development of regioselective aerobic a-oxygen-
ation of tertiary amines at the N-methyl positions with
a broad substrate scope is still an extremely challenging task.
Cu/nitroxyl radical (N-oxyl) catalyst systems have gained
attention for various aerobic oxidation reactions under mild
conditions, particularly for alcohol oxidation.[32–39] Stahl et al.
recently reported that Cu/N-oxyl catalysts promoted hemi-
aminal oxidation in the aerobic oxidative acylation of amines
with primary alcohols.[40] Cu/N-oxyl-catalyzed oxidation of Results and Discussion
primary amines[41] and secondary amines[42–46] has also been
reported so far. However, to the best of our knowledge, there
is no precedent for Cu/N-oxyl-catalyzed oxidation of tertiary
amines. Since Cu/N-oxyl-catalyzed alcohol oxidation is be-
lieved to proceed in a bulky closed-shell nitroxyl-Cu ad-
duct,[47–50] we expected that tertiary amine oxidation would
proceed regioselectively at N-methyl-positions due to such
steric restriction.
In this study, we developed a practical route to various
formamides through regioselective a-oxygenation of tertiary
N-methylamines, using O2 (open air atmosphere) as the
terminal oxidant at room temperature, by employing the Cu/
N-oxyl catalyst system (Scheme 1). The reaction proceeded
efficiently in the presence of moderately hindered caged N-
oxyls, such as 1,5-dimethyl-9-azanoradamantane N-oxyl
(DMN-AZADO) and 1-methyl-2-azaadamanane N-oxyl (1-
Me-AZADO). Although the N-methyl-selective oxygenation
essentially proceeded even when using less hindered N-oxyls,
the reactions stopped at the early stage most likely due to the
deactivation of these N-oxyls. Additionally, the catalytic
activity was significantly improved by precisely controlling
Optimization of Reaction Conditions
Initially, we carried out the oxygenation of 1-methyl-
piperidine (1a) using 10 mol% of the CuBr2/bpy/9-azanor-
adamantane N-oxyl (nor-AZADO) catalyst system at room
temperature ( ꢀ 238C) in non-dehydrated acetonitrile
(MeCN) under open air conditions. Under these conditions,
the desired a-methyl-oxygenated product 1-formylpiperidine
(1b) was obtained selectively in 9% yield after 2 h, and N-
methylpiperidone (1c) formation was negligible (Figure 1,
Table S1, entry 1). Unfortunately, the formation of 1b almost
stopped within 2 h, and no significant increase in the yield of
1b was observed even when the reaction time was increased
to 24 h (Figure 1, Table S1, entry 2). Even after 24 h, 1c was
not detected. The undesirable termination of the reaction was
considered to be due to the deactivation of nor-AZADO.[51]
Then, other N-oxyls, such as 2-azaadamantane N-oxyl (AZA-
DO), 9-azabicyclo[3.3.1]nonane N-oxyl (ABNO), 9-
Figure 1. Effect of N-oxyls. Reaction conditions: 1a (0.25 mmol),
CuBr2/bpy/N-oxyl (10/10/10 mol%), MeCN (1 mL), air (open), room
temperature. Yields were determined by GC analysis. See also in
Table S1.
Scheme 1. Regioselectivity of a-oxygenation of tertiary amines.
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
&&&&
These are not the final page numbers!

Angewandte
Research Articles
Chemie
azabicyclo[3.3.1]nonane-3-one N-oxyl (keto-ABNO), 1-Me-
AZADO, DMN-AZADO, and 2,2,6,6-tetramethylpiperidine
N-oxyl (TEMPO), were also investigated for the oxygenation
of 1a rather than nor-AZADO. When using AZADO or
ABNO, the reactions were also N-methyl-selective and
stopped within 2 h, affording 1b in only ꢀ 10% yield after
24 h (Figure 1, Table S1, entries 3–6). Therefore, these less
hindered N-oxyls were also found to be ineffective from the
viewpoint of increasing the yield of 1b. Fortunately, we found
that moderately hindered caged N-oxyls, such as 1-Me-
AZADO and DMN-AZADO, exhibited high catalytic per-
formances and that 1b was produced effectively (Figure 1,
Table S1, entries 3–6). In the case of DMN-AZADO and 1-
Me-AZADO, byproducts due to the deactivation of these N-
oxyls were not detected.[51] Thus, steric hindrance of N-oxyls is
considered to be highly important for avoiding deactivation.
DMN-AZADO showed a higher turnover number (TON)
(25.5) than 1-Me-AZADO (11.8) after 24 h, even when the
amount of N-oxyl was decreased to 2.5 mol% (Figure S1).
The reaction hardly proceeded with keto-ABNO or TEMPO,
possibly due to the high redox potential of keto-ABNO[42,52]
and the bulkiness of TEMPO (Figure 1, Table S1, entries 7, 8,
13, and 14). Hereafter, we mainly used the effective moder-
ately hindered DMN-AZADO or 1-Me-AZADO for further
detailed investigations.
CuBr2 showed the highest catalytic activity (Table 1, entry 1).
To investigate the effect of BrÀ in detail, 1a was oxygenated at
various BrÀ/Cu molar ratios by fixing Cu at 10 mol%. When
the BrÀ/Cu molar ratios were controlled in the range 0.6–1.4
by changing the amounts of Cu(OTf)2 and CuBr2, the yield of
1b was significantly higher than that obtained with Cu(OTf)2
(BrÀ/Cu = 0) (Figure 2a, Table 1, entries 8–10 vs. entry 3).
The oxygenation also proceeded efficiently when LiBr or
CuBr was added to Cu(OTf)2 with a BrÀ/Cu ratio in the range
The effect of Cu sources was then examined. In this
oxygenation, divalent Cu sources, such as CuBr2, CuCl2, and
Cu(OTf)2 (Table 1, entries 1–3), were more effective than
monovalent CuBr, CuCl, CuI, and Cu(OTf)·4MeCN (Ta-
ble 1, entries 4–7). Among the divalent Cu sources examined,
Table 1: Effect of Cu catalysts on the oxygenation of 1a.[a]
Entry
Catalyst (mol%)
Yield [%]
1b
1c
1
2
3
CuBr2 (10)
CuCl2 (10)
Cu(OTf)2 (10)
33
14
26
n.d.
n.d.
n.d.
4
5
6
7
CuBr (10)
CuCl (10)
CuI (10)
Cu(OTf)·4MeCN (10)
2
n.d.
1
n.d.
n.d.
n.d.
n.d.
16
8
9
Cu(OTf)2 (7) + CuBr2 (3)
Cu(OTf)2 (5) + CuBr2 (5)
Cu(OTf)2 (3) + CuBr2 (7)
Cu(OTf)2 (10) + LiBr (5)
Cu(OTf)2 (10) + LiBr (10)
Cu(OTf)2 (10) + LiBr (15)
Cu(OTf)2 (10) + LiBr (20)
Cu(OTf)2 (5) + CuBr (5)
83
88
78
83
87
71
45
91
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
10
11
12
13
14
15
16
17
18
19
Cu(OTf)2 (7) + CuCl2 (3)
Cu(OTf)2 (5) + CuCl2 (5)
Cu(OTf)2 (3) + CuCl2 (7)
Cu(OTf)2 (5) + CuCl (5)
92
96
52
98
n.d.
n.d.
n.d.
n.d.
Figure 2. a) Effect of counter anions on the yield of 1b. Reaction
conditions: 1a (0.25 mmol), Cu source/bpy/DMN-AZADO (10/10/
10 mol%), MeCN (1 mL), air (open), room temperature, 2 h. Conver-
sions and yields were determined by GC analysis. Detailed values are
shown in Table 1. b) Positive-ion CSI-MS spectra of catalyst solution.
Conditions: Cu/bpy/DMN-AZADO (1:1:1) in MeCN for (i) and (iii),
Cu(OTf)2/LiBr/bpy/DMN-AZADO (1:1:1:1) in MeCN for (ii). X=Br or
OTf. c) Formation of complex B.
20
Cu(OTf)2 (5) + CuI (5)
14
n.d.
[a] Reaction conditions: 1a (0.25 mmol), Cu/bpy/DMN-AZADO (10/10/
10 mol%), MeCN (1 mL), air (open), room temperature, 2 h. Yields were
determined by GC analysis.
&&&&
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
These are not the final page numbers!

Angewandte
Research Articles
Chemie
0.5–1.5 (Figure 2a, Table 1, entries 11–13 and 15). When the
to Cu center are effective. The reactions did not proceed when
reaction was performed in the presence of Cu(OTf)2 and LiBr
with a BrÀ/Cu ratio of 2.0, the result was similar to that using
CuBr2 (Figure 2a, Table 1, entry 14 vs. entry 1).
other metal sources such as MnCl2·4H2O, FeCl3, and CoCl2
were used (Table S3). The yield of 1b also drastically
decreased when the solvent was changed from MeCN to
tetrahydrofuran or 1,2-dichloroethane (Table S4).
To gain a mechanistic insight into the role of BrÀ, the
reaction solutions were characterized by cold spray ionization
mass spectrometry (CSI-MS, positive mode). The CSI-MS
spectrum of a MeCN solution of Cu(OTf)2, bpy, and DMN-
AZADO (Cu(OTf)2/bpy/DMN-AZADO = 1:1:1) presents
a series of signals at m/z 524.2, assignable to [Cu(bpy)2(OTf)]+
(calcd 524.0, complex A) (Figure 2b,i). Thus, without BrÀ,
a CuII species coordinated by two bpy molecules was mainly
observed, with no species coordinated by DMN-AZADO.
Conversely, the CSI-MS spectrum of a MeCN solution of
Cu(OTf)2, LiBr, bpy, and DMN-AZADO (Cu(OTf)2/LiBr/
bpy/DMN-AZADO = 1:1:1:1) exhibited series of signals at
m/z 456.2 and 466.2, assignable to [Cu(bpy)2Br]+ (calcd 456.0,
complex A) and [Cu(bpy)(DMN-AZADO)Br]+ (calcd 466.0,
complex B), respectively (Figure 2b,ii). The computational
calculation indicated that the formation of [Cu(bpy)(DMN-
AZADO)Br]+ (complex B) by replacement of MeCN in
[Cu(bpy)(MeCN)Br]+ is thermodynamically favorable by
7.6 kcalmolÀ1 (Figure 2c)[53] The CSI-MS spectrum of
a MeCN solution of CuBr2, bpy, and DMN-AZADO
(CuBr2/bpy/DMN-AZADO = 1:1:1) also exhibited series of
signals at m/z 456.2 and 466.2 (Figure 2b,iii). Consequently, it
is considered that BrÀ acts as a monodentate ligand to inhibit
coordination of two bpy ligands and allow coordination of
DMN-AZADO.[53] As mentioned above, the reaction pro-
ceeded efficiently with BrÀ/Cu = 0.5–1.5, and as the BrÀ/Cu
ratio was increased further, the reaction rate decreased.
Excess BrÀ presumably inhibits the coordination of the amine
substrate, reducing the efficiency of the reaction.
The yield of 1b drastically decreased without a Cu source,
ligand, or N-oxyl (Table 2, entries 2–4 vs. entry 1). When the
reaction was performed under an Ar atmosphere, the yield of
1b was only 3%, which was below the amount of catalyst used
Table 2: Control experiments.[a]
Entry
Deviation from standard conditions
Yield [%]
1b
1c
1
2
3
4
5
6
No changes (standard conditions)
Without Cu
Without bpy
CuBr2 (10 mol%), without DMN-AZADO
Ar atmosphere
DMN-AZADO (1 mol%), 24 h
98
n.d.
11
n.d.
3
94
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
[a] Reaction conditions: 1a (0.25 mmol), Cu/bpy/DMN-AZADO (10/10/
10 mol%), MeCN (1 mL), air (open), room temperature, 2 h. Yields were
determined by GC analysis.
(10 mol%) (Table 2, entry 5). Therefore, O2 in air functioned
as the terminal oxidant in this a-oxygenation. Under the
optimized Cu(OTf)2/CuCl/bpy/DMN-AZADO reaction con-
ditions (standard conditions), the oxygenation of 1a was
completed within 2 h, affording 98% yield of 1b (Figure S2,
Table 2, entry 1). The reaction solution was dark red during
the catalytic turnover and became green when 1a was
completely converted to 1b. Even when the amount of
DMN-AZADO catalyst was reduced to 1 mol%, 1b was
obtained quantitatively after 24 h (Table 2, entry 6).
Furthermore, we found that ClÀ was more effective than
BrÀ; when ClÀ/Cu was controlled in the range 0.5–1.0 by using
Cu(OTf)2 and CuCl2, or Cu(OTf)2 and CuCl, 1b was obtained
in almost quantitative yields (Figure 2a, Table 1, entries 16,
17, and 19). The yield of 1b was more drastically decreased
when the ClÀ/Cu ratio exceeded 1.0 (Figure 2a, Table S1,
entry 18 vs. entry 10). CuI was ineffective for this oxygenation
(Table S1, entry 20). On the basis of the above-mentioned
investigations, we mainly used Cu(OTf)2/CuCl (5 mol%/
5 mol%) as the most effective copper catalyst for further
detailed investigations.
Next, various ligands were examined for the oxygenation
of 1a. The reaction proceeded effectively in the presence of
bpy, 4,4’-dimethyl-2,2’-bipyridyl (4Mebpy), 4,4’-dimethoxy-2,2’-
bipyridyl (4MeObpy), 1,10-phenanthroline (phen), and tetra-
methylethylenediamine (TMEDA) (Table S2, entries 1–5).
However, bulky 6,6’-dimethyl-2,2’-bipyridyl (6Mebpy) and
electron-deficient 4,4’-dibromo-2,2’-bipyridyl (4Brbpy) were
inferior to bpy (Table S2, entries 6 and 7). Whichever
bidentate ligand was used, 1b was selectively obtained
without formation of 1c. We also examined the effect of
monodentate ligands. 2,4,4-Trimethyl-2-oxazoline and pyri-
dine were not effective (Table S2, entries 8 and 9). 4-
Dimethylaminopyridine and 1-methylimidazole gave 1b in
79% and 70% yields, respectively (Table S2, entries 10 and
11). Therefore, it is likely that ligands that coordinate strongly
Substrate Scope for Formamide Synthesis
Under the optimized conditions, we examined the sub-
strate scope for the proposed Cu/N-oxyl-catalyzed methyl-
selective a-oxygenation of tertiary amines. It was confirmed
by NMR analyses of the isolated products or the crude
reaction mixtures (after removal of the catalysts by short
silica column) that all reactions in Table 3 occurred highly
regioselectively at the N-methyl positions. We first examined
the applicability to cyclic N-methylamines (Table 3a). Six-
membered cyclic N-methylamines, such as 1-methylpiperi-
dine and 4-methylmorpholine, were oxygenated to the
corresponding formamides in high yields (Table 3a, 1 and
2). Various six-membered cyclic N-methylamines with sub-
stituents at the 3- or 4-positions reacted efficiently to afford
the corresponding formamides in high yields (Table 3a, 3 and
4). Chloro, ketone, and amide groups were tolerated in this
oxygenation (Table 3a, 5–7). Although DMN-AZADO
showed poor catalytic activity for six-membered cyclic N-
methylamines with substituents at the 2-positions, the reac-
tion proceeded effectively with less bulky 1-Me-AZADO
(Table 3a, 8 and 9). The reaction of a five-membered N-
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
&&&&
These are not the final page numbers!

Angewandte
Research Articles
Chemie
Table 3: Substrate scope for a-methyl-selective oxygenation of tertiary amines.[a]
[a] Reaction conditions A: Amine substrate a (0.25 mmol), Cu(OTf)2/CuCl/bpy/DMN-AZADO (5/5/10/10 mol%), MeCN (1 mL), air (open), room
temperature; Conditions B: Amine substrate a (0.25 mmol), Cu(OTf)2/CuCl/bpy/2,4,4-trimethyl-2-oxazoline/DMN-AZADO (10/10/20/50/20 mol%),
MeCN (1 mL), air (open), room temperature. Isolated yields and GC yields (the values in parentheses) are shown. Limitation of substrate is described
in Table S5. [b] 1-Me-AZADO (10 mol%) rather than DMN-AZADO. [c] 0.5 mmol scale (two-fold scale-up). [d] Without the oxazoline ligand.
methylamine, 1-methylpyrrolidine, under the conditions A
was failed. When the amount of Cu/bpy/DMN-AZADO was
increased from 10 mol% to 20 mol%, and 2,4,4-trimethyl-2-
oxazoline was used, N-formylpyrrolidine was produced in
a quantitative yield (Table 3a, 10).[54] As previously men-
tioned, 2,4,4-trimethyl-2-oxazoline was not an effective ligand
(Table S2, entry 11). It is possible that 2,4,4-trimethyl-2-
oxazoline functions as a suitable base and promotes the
deprotonation (and/or protonation).[54] However, the details
are unclear and thus further studies are required. In all cases,
no byproducts derived from oxygenation at the ring positions
were detected.
This catalytic system was also applicable to the a-oxy-
genation of N,N-dimethylamines (Table 3b). For N,N-dime-
thylamines, both DMN-AZADO and 1-Me-AZADO were
effective. Various N,N-dimethylamines with benzyl, naph-
thylmethyl, and cyclohexyl groups reacted efficiently, afford-
ing the corresponding formamides in high yields (Table 3b,
11–14). The reaction proceeded smoothly when N,N-dime-
thylbenzylamine derivatives with electron-donating groups
were used as the substrates, and the corresponding forma-
mides were obtained in high yields (Table 3b, 15 and 16).
Although N,N-dimethylbenzylamine derivatives with elec-
tron-withdrawing groups were less effective, the oxygenation
was almost complete when the reaction time was increased to
24 h (Table 3b, 17–22). The yields of formamides were
unsatisfactory with some kinds of N,N-dimethylamines, such
as N,N-dimethylisopropylamine and N,N-dimethylbutyl-
amine, under the standard conditions. For these substrates,
the corresponding formamides were obtained in good yields
by using 2,4,4-trimethyl-2-oxazoline and increasing the
amount of catalyst as with 1-methylpyrrolidine (Table 3b, 23
and 24). The ester, acetal-protected aldehyde, and olefinic
groups were also tolerated, affording the corresponding
formamides in excellent yields (Table 3b, 25–27). Whichever
N,N-dimethylamine substrates were used, the oxygenation
occurred regioselectively at the N-methyl positions with N-
methylene, N-methine, and N-benzyl groups remaining intact,
as shown in Table 3b.
Notably, this catalyst system could be applied to a natural
alkaloid and commercially available pharmaceutical drugs
(Table 3c). Pseudopelletierine (28a), an alkaloid derived
from the pomegranate tree, was efficiently converted to the
corresponding formamide in a high yield (Table 3c, 28).
Furthermore, N-methylparoxetine (29a) and amitriptyline
(30a), medicines for mental illness, reacted smoothly, giving
the corresponding formamides (Table 3c, 29 and 30). Thus,
this catalyst system would enable a useful reaction for
functionalizing N-methyl groups of bioactive compounds
while preserving the original amine substrate skeletons.
&&&&
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
These are not the final page numbers!

Angewandte
Research Articles
Chemie
Reaction Mechanism and Origin of Regioselectivity
Various experiments were conducted to elucidate the
reaction mechanism. Firstly, to better understand the origin of
the formamide oxygen, a series of 18O-labeling experiments
was conducted. The a-oxygenation of 1a with 30 mol% of
H218O under 16O2 (purified air, 1 atm) afforded 18O-labeled 1b
and unlabeled 1b in 23% and 74% yields, respectively
(Figure S3a). The 18O-labeled 1b/total 1b ratio was 0.52 at the
initial stage of the reaction (after 15 min) and gradually
decreased as the yield of 1b increased (Figure S3b). When a-
oxygenation of 1a was performed with 500 mol% of H218O
under 16O2 (purified air, 1 atm), 67% 18O incorporation in the
product was observed after 24 h. These results suggest that the
formamide oxygen originates from water; thus, in the early
stages of the reaction, formamide is considered to be
produced from water (H218O) originally contained in the
reaction system, and water produced by 16O2 reduction is also
used in the later stage of the reaction.
In this a-oxygenation, two electrons and one proton of the
amine substrate are believed to transfer to the catalyst to form
an iminium cation intermediate which is subsequently reacted
with water to produce a hemiaminal intermediate. The
following investigations were performed to gain insight into
electron and proton transfer mechanisms. It is known that
ring-opening proceeds when radical species are generated
adjacent to a three-membered ring.[42,55,56] Among the follow-
ing three possible amine-oxidation mechanisms, path A:
single electron transfer (SET) from the nitrogen atom; path
B: hydrogen atom transfer (HAT) from the a-methyl
position; and path C: hydride transfer from the a-methyl
position, the ring-opening of three-membered cyclic amine
31a would proceed only via path A (Scheme 2a). In this
Cu(OTf)2/CuCl/bpy/1-Me-AZADO system, the conversion
of 31a reached > 99% after 4 h, and a-methyl-oxygenated
product 31b was obtained in only 17% yield with the
concomitant formation of many byproducts derived from
ring-opening including cinnamaldehyde (9% yield) (Sche-
me 2a). Therefore, path A (SET) is considered to be
dominant. Furthermore, the reaction of 31a hardly proceeded
when using DMN-AZADO (Scheme S1). Considering that
the N-oxyl-Cu adduct (complex B; see Figure 2) was detected
by CSI-MS as the possible active species, bulky 31a was
difficult to coordinate to the adduct when DMN-AZADO
was used rather than less bulky 1-Me-AZADO. Neither did
the reactions proceed without N-oxyls (Scheme S1), suggest-
ing that the reaction was initiated through SET by Cu only
when both the substrate and N-oxyl coordinated to the Cu
center.
Scheme 2. Mechanistic investigation of the radical intermediate. [a]
Reaction conditions: 31a (0.25 mmol), Cu(OTf)2/CuCl/bpy/N-oxyl (5/
5/10/10 mol%), MeCN (1 mL), air (open), room temperature, 4 h. [b]
Reaction conditions: 30a (0.25 mmol), Cu(OTf)2/CuCl/bpy/2,4,4-tri-
methyl-2-oxazoline/DMN-AZADO (10/10/20/50/20 mol%), MeCN
(1 mL), air (open), room temperature, 24 h.
The mechanism of electron and proton transfer from
a radical cation species formed through the SET was then
investigated. In stepwise proton and electron transfer (path
D), 5-exo-cyclization of amine 30a mainly proceeds via
formation of the carbon radical at the a-position (Sche-
me 2b).[55–57] However, when an electron and a proton are
concertedly transferred (path E), the 5-exo-cyclization of 30a
would be less likely to occur (Scheme 2b). In this Cu(OTf)2/
CuCl/bpy/DMN-AZADO system, the a-methyl-oxygenation
of 30a mainly proceeded without 5-exo-cyclization (Sche-
me 2b), suggesting that electron and proton transfer from the
radical cation would proceed concertedly (path E). This
concerted abstraction of the electron and proton from the
radical cation is considered to proceed soon after the initial
SET, since a small amount of 31b was produced even though
the ring-opening of 31a proceeded rapidly (Sche-
me 2a).[42,55,56]
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
&&&&
These are not the final page numbers!

Angewandte
Research Articles
Chemie
On the basis of the above-mentioned experimental
evidence, two amine-oxidation models were proposed
(Scheme 3): model A: SET and subsequent electron and
proton transfer in a closed-shell Cu complex coordinated by
Figure 3. Investigation into the regioselectivity of a-oxygenation of
tertiary amine 1a by DFT calculation based on amine-oxidation model
B. L=bpy, N-oxyl=DMN-AZADO, N-oxyl-H=hydroxylamine.
Scheme 3. Possible amine-oxidation models. L=bpy, N-oxyl=DMN-
AZADO or 1-Me-AZADO, N-oxyl-H=hydroxylamine.
when using nor-AZADO (Figure 1, Table S1, entries 1 and 2);
a-methyl oxidation was calculated to occur preferentially by
1.6 kcalmolÀ1 compared to a-methylene oxidation, and the
amine-N-oxyl interaction also determined the methyl selec-
tivity (Scheme S3b).
both the amine and the N-oxyl (complex CI); and model B:
SET in a closed-shell Cu complex CI and subsequent electron
and proton transfer in a N-oxyl-amine adduct (complex CII)
produced by radical-radical coupling of the N-oxyl and radical
cation derived from the amine substrate. These two amine-
oxidation mechanisms were examined by DFT computational
studies. Firstly, model A was found to be inappropriate; a-
methylene oxidation of five-membered cyclic amine 10a was
calculated to be favorable compared with a-methyl oxidation
with a difference exceeding 5 kcalmolÀ1, though the exper-
imental result revealed that a-oxygenation of 10a took place
regioselectively at the a-methyl position only (Scheme S2). In
computational studies based on model A, it was considered
that the steric hindrance of the amine and bpy ligand mainly
determined the regioselectivity of oxygenation of 10a at the
a-methylene position (Scheme S2, model A).
On the other hand, the experimental result was well-
supported by model B; a-methyl oxidation was calculated to
be more favorable than a-methylene oxidation (Figure 3,
Scheme S3a). As shown in Figure 3, the energy barrier of a-
methyl oxidation of 1a via the speculated intermediate CII
was calculated as 3.5 kcalmolÀ1. However, a-methylene
oxidation of 1a would require formation of the relatively
unstable conformation CII’ (2.1 kcalmolÀ1 higher than inter-
mediate CII) due to the steric hindrance originating from the
N-oxyl and six-membered ring structure of amine 1a. Thus,
the energy barrier for a-methylene oxidation was 6.2 kcal
molÀ1. In the oxygenation of 10a, a-methyl oxidation was also
calculated to be favourable because of the steric hindrance of
the amine substrate and N-oxyl compared with a-methylene
oxidation (Scheme S2, model B). Furthermore, we also
performed the DFT computational studies for the nor-
AZADO-catalyzed oxygenation of 1a using model B. The
calculation results well reproduced the experimental fact that
the reaction proceeded in a methyl-selective manner even
The regioselectivity of this a-oxygenation was then
further tested on the basis of computational model B. Unlike
amine 1a and 10a, a-methylene oxidation was calculated as
3.2 kcalmolÀ1 more favorable than a-methyl oxidation with 2-
methyl-1,2,3,4-tetrahydroisoquinoline (32a), possibly be-
cause 32a has no proton that protrudes into the bulky N-
oxyl due to the aromatic ring (Scheme S3c). Experimental
examination revealed that amide 32c, an a-methylene-oxy-
genated product, was obtained regioselectively in 42% yield
with this Cu(OTf)2/CuCl/bpy/1-Me-AZADO system, and
formamide 32b was hardly obtained (Scheme 4, Scheme S3c).
Consequently, we concluded that the regioselectivity of amine
oxidation was determined by the steric constraints of the
amine-N-oxyl interaction.
On the basis of the aforementioned experimental evi-
dence, DFT calculations, and reported literatures,[39,47–50]
a possible oxygenation mechanism is proposed as shown in
Scheme 5. A tertiary amine initially coordinates to complex B
to form complex CI (Scheme 5, step 1). A radical cation
intermediate formed through SET oxidation (TS-CI) is then
Scheme 4. Oxygenation of 32a. Reaction conditions: 32a (0.25 mmol),
Cu(OTf)2/CuCl/1-Me-AZADO/N-oxyl (5/5/10/10 mol%), MeCN
(1 mL), air (open), room temperature, 7 h. See also Scheme S3.
&&&&
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
These are not the final page numbers!

Angewandte
Research Articles
Chemie
Scheme 5. Plausible reaction mechanism. L=bpy, N-oxyl=DMN-AZADO or 1-Me-AZADO, N-oxyl-H=hydroxylamine.
immediately trapped by an N-oxyl to form complex CII
(Scheme 5, step 2) followed by the formation of an iminium
cation intermediate and a hydroxylamine (N-oxyl-H) through
concerted proton and electron transfer (TS-CII) (Scheme 5,
step 3). The regioselectivity of this oxidation is determined by
step 3. Reduced copper species D is then oxidized by O2 to
form species E followed by its reduction by N-oxyl-H to
generate CuII-OH species F (Scheme 5, steps 4 and 5).
Complex B is regenerated by the reaction of F with N-oxyl
to complete catalytic cycle I (Scheme 5, step 6). Meanwhile,
a hemiaminal intermediate is formed by nucleophilic addition
of water to the iminium cation species generated in catalytic
cycle I (Scheme 5, step 7). We believe that oxidation of the
hemiaminal intermediate proceeds according to the mecha-
nism proposed by Stahl and co-workers (Scheme 5, steps 8
and 9), in which two electrons and one proton transfer
concertedly in closed-shell transition state TS-CIII.[47–50] Re-
duced copper species D is regenerated to form active CuII-OH
species F by the reaction with O2 and N-oxyl-H as in catalytic
cycle I (Scheme 5, steps 10 and 11).
proceeds under extremely mild conditions (e.g., room tem-
perature and open air atmosphere), enabling excellent func-
tional group tolerance and a user-friendly protocol. We hope
that this a-oxygenation system will find wide application,
especially in fine chemical synthesis.
Acknowledgements
This work was financially supported by JSPS KAKENHI
Grant Numbers 15H05797, 15H05805, and 18H04232 in
Precisely Designed Catalysts with Customized Scaffolding.
Conflict of interest
The authors declare no conflict of interest.
Keywords: copper · formamides ·
methyl-selective a-oxygenation · nitroxyl radical · tertiary amines
Conclusion
In this study, we successfully developed a highly regiose-
lective and efficient aerobic a-oxygenation of tertiary N-
methylamines catalyzed by a Cu/bpy/N-oxyl catalyst system.
Various types of tertiary N-methylamines, including a natural
alkaloid and pharmaceutical drugs, could be transformed to
the corresponding formamides by the proposed method. The
key to realizing this efficient oxygenation system is the use of
caged N-oxyls with moderate steric hindrance together with
an appropriate amount of anion, such as ClÀ or BrÀ, as the
monodentate ligand to enhance the coordination of the N-
oxyls. We concluded that the steric constraints of the amine
and N-oxyl during amine oxidation determine the regiose-
lectively at the N-methyl positions. This catalytic oxygenation
[1] B. R. Brown, The Organic Chemistry of Aliphatic Nitrogen
Compounds, Oxford University Press, New York, 1994.
[2] M. B. Smith, J. March, Marchꢀs Advanced Organic Chemistry:
Reactions, Mechanisms, and Structure, 6th ed., Wiley, New York,
2007.
[3] C. S. Yeung, V. M. Dong, Chem. Rev. 2011, 111, 1215 – 1292.
[4] C. Liu, J. Yuan, M. Gao, S. Tang, W. Li, R. Shi, A. Lei, Chem.
Rev. 2015, 115, 12138 – 12204.
[5] J. W. Beatty, C. R. J. Stephenson, Acc. Chem. Res. 2015, 48,
1474 – 1484.
[6] X. Jin, K. Yamaguchi, N. Mizuno, RSC Adv. 2014, 4, 34712 –
34715.
[7] K. Yamaguchi, Y. Wang, N. Mizuno, ChemCatChem 2013, 5,
2835 – 2838.
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
&&&&
These are not the final page numbers!

Angewandte
Research Articles
Chemie
[8] S. Ding, N. Jiao, Angew. Chem. Int. Ed. 2012, 51, 9226 – 9237;
Angew. Chem. 2012, 124, 9360 – 9371.
[38] Y. Sasano, N. Kogure, S. Nagasawa, K. Kasabata, Y. Iwabuchi,
Org. Lett. 2018, 20, 6104 – 6107.
[9] N. Reddy, G. Kumar, P. Kumar, M. Kantam, K. Reddy, Synlett
2014, 25, 2133 – 2138.
[10] M. A. Ganiek, M. R. Becker, G. Berionni, H. Zipse, P. Knochel,
Chem. Eur. J. 2017, 23, 10280 – 10284.
[11] G. S. Kumar, R. A. Kumar, P. S. Kumar, N. V. Reddy, K. V.
Kumar, M. L. Kantam, S. Prabhakar, K. R. Reddy, Chem.
Commun. 2013, 49, 6686 – 6688.
[12] Y. Nakao, H. Idei, K. S. Kanyiva, T. Hiyama, J. Am. Chem. Soc.
2009, 131, 5070 – 5071.
[13] A. Jackson, O. Meth-Cohn, J. Chem. Soc. Chem. Commun. 1995,
1319.
[14] K. Kobayashi, S. Nagato, M. Kawakita, O. Morikawa, H.
Konishi, Chem. Lett. 1995, 24, 575 – 576.
[15] A. Kakehi, S. Ito, S. Hayashi, T. Fujii, Bull. Chem. Soc. Jpn. 2006,
68, 3573 – 3580.
[39] K. Kataoka, K. Wachi, X. Jin, K. Suzuki, Y. Sasano, Y. Iwabuchi,
J. Hasegawa, N. Mizuno, K. Yamaguchi, Chem. Sci. 2018, 9,
4756 – 4768.
[40] S. L. Zultanski, J. Zhao, S. S. Stahl, J. Am. Chem. Soc. 2016, 138,
6416 – 6419.
[41] J. Kim, S. S. Stahl, ACS Catal. 2013, 3, 1652 – 1656.
[42] T. Sonobe, K. Oisaki, M. Kanai, Chem. Sci. 2012, 3, 3249 – 3255.
[43] Z. Hu, F. M. Kerton, Org. Biomol. Chem. 2012, 10, 1618 – 1624.
[44] B. Han, X.-L. Yang, C. Wang, Y.-W. Bai, T.-C. Pan, X. Chen, W.
Yu, J. Org. Chem. 2012, 77, 1136 – 1142.
[45] Z. Chen, J. Chen, M. Liu, J. Ding, W. Gao, X. Huang, H. Wu, J.
Org. Chem. 2013, 78, 11342 – 11348.
[46] J. Yu, J. Xu, M. Lu, Appl. Organomet. Chem. 2013, 27, 606 – 610.
[47] J. M. Hoover, B. L. Ryland, S. S. Stahl, J. Am. Chem. Soc. 2013,
135, 2357 – 2367.
[16] R. G. Jones, J. Am. Chem. Soc. 1949, 71, 644 – 647.
[17] J. C. Sheehan, D.-D. H. Yang, J. Am. Chem. Soc. 1958, 80, 1154 –
1158.
[18] J. Martinez, J. Laur, Synthesis 1982, 11, 979 – 981.
[19] Y. R. Luo, Handbook of Bond Dissociation Energies in Organic
Compounds, CRC, Boca Raton, 2002.
[20] Y. Zhang, D. Riemer, W. Schilling, J. Kollmann, S. Das, ACS
Catal. 2018, 8, 6659 – 6664.
[21] X. Jin, K. Kataoka, T. Yatabe, K. Yamaguchi, N. Mizuno, Angew.
Chem. Int. Ed. 2016, 55, 7212 – 7217; Angew. Chem. 2016, 128,
7328 – 7333.
[48] J. M. Hoover, B. L. Ryland, S. S. Stahl, ACS Catal. 2013, 3, 2599 –
2605.
[49] B. L. Ryland, S. D. McCann, T. C. Brunold, S. S. Stahl, J. Am.
Chem. Soc. 2014, 136, 12166 – 12173.
[50] R. C. Walroth, K. C. Miles, J. T. Lukens, S. N. MacMillan, S. S.
Stahl, K. M. Lancaster, J. Am. Chem. Soc. 2017, 139, 13507 –
13517.
[51] We confirmed by GC-MS analysis that a byproduct possibly
derived from nor-AZADO (m/z = 151, assignable to N-formyl-
9-azanoradamantane) was formed during the reaction albeit in
a small amount (below 5%). There is no clear evidence for the
formation mechanism of this byproduct, but it is considered to
[22] D. D. Agarwal, D. Bhat, J. Porphyrins Phthalocyanines 2016, 20,
689 – 693.
[23] D. Narꢀg, U. Lechowicz, T. Pietryga, A. Sobkowiak, J. Mol.
Catal. A 2004, 212, 25 – 33.
À
be generated from the cleavage of the N O bond of nor-
AZADO. In contrast, such byproducts were not detected when
using DMN-AZADO and 1-Me-AZADO.
[24] S. Murata, M. Miura, M. Nomura, J. Org. Chem. 1989, 54, 4700 –
4702.
[52] M. Shibuya, F. Pichierri, M. Tomizawa, S. Nagasawa, I. Suzuki, Y.
Iwabuchi, Tetrahedron Lett. 2012, 53, 2070 – 2073.
[25] S. Murata, M. Miura, M. Nomura, J. Chem. Soc. Chem.
Commun. 1989, 116 – 118.
[26] A. Sobkowiak, D. T. Sawyer, J. Am. Chem. Soc. 1991, 113, 9520 –
9523.
[27] D. H. Hunter, J. S. Racok, A. W. Rey, Y. Zea-Ponce, J. Org.
Chem. 1988, 53, 1278 – 1281.
[28] D. H. Hunter, D. H. R. Barton, W. J. Motherwell, Tetrahedron
Lett. 1984, 25, 603 – 606.
[29] Y. Kashiwagi, J. Anzai, Chem. Pharm. Bull. 2001, 49, 324 – 326.
[30] G. T. Davis, D. H. Rosenblatt, Tetrahedron Lett. 1968, 9, 4085 –
4086.
[31] J. Genovino, D. Sames, B. B. Tourꢁ, Tetrahedron Lett. 2015, 56,
3066 – 3069.
[53] Stahl and co-workers reported that coordination of an N-oxyl to
form a [Cu(bpy)(N-oxyl)(OR)]+ species is quite unfavorable.[49]
In contrast, our DFT computational studies revealed that the
formation of [Cu(bpy)(DMN-AZADO)Br]+ was thermodynam-
ically favorable. Therefore, as one of the reviewers pointed out,
the replacement of an alkoxy ligand with BrÀ could change the
thermodynamics for binding of the N-oxyl.
[54] Even when the amount of 2,4,4-trimethyl-2-oxazoline was
reduced to 20 mol% for the oxygenation of 10a, 10b was
obtained quantitatively after 2 h. Furthermore, the oxygenation
of 10a was carried in the presence of bpy together with 1,8-
diazabicyclo[5.4.0]undec-7-ene (20 mol%). However, in this
case, the reaction hardly proceeded.
[32] J. M. Hoover, S. S. Stahl, J. Am. Chem. Soc. 2011, 133, 16901 –
16910.
[55] B. Maillard, D. Forrest, K. U. Ingold, J. Am. Chem. Soc. 1976, 98,
7024 – 7026.
[33] J. E. Steves, S. S. Stahl, J. Am. Chem. Soc. 2013, 135, 15742 –
15745.
[56] O. M. Musa, J. H. Homer, H. Shahin, M. Newcomb, J. Am.
Chem. Soc. 1996, 118, 3862 – 3868.
[34] B. L. Ryland, S. S. Stahl, Angew. Chem. Int. Ed. 2014, 53, 8824 –
8838; Angew. Chem. 2014, 126, 8968 – 8983.
[35] Y. Seki, K. Oisaki, M. Kanai, Tetrahedron Lett. 2014, 55, 3738 –
3746.
[57] Coupling reactions of carbon radical at a-methyl position with
O2 would be ignorable, considering that the main oxygen source
was water, rather than O2.
[36] Y. Sasano, S. Nagasawa, M. Yamazaki, M. Shibuya, J. Park, Y.
Iwabuchi, Angew. Chem. Int. Ed. 2014, 53, 3236 – 3240; Angew.
Chem. 2014, 126, 3300 – 3304.
[37] Y. Sasano, N. Kogure, T. Nishiyama, S. Nagasawa, Y. Iwabuchi,
Chem. Asian J. 2015, 10, 1004 – 1009.
Manuscript received: July 19, 2019
Revised manuscript received: September 11, 2019
Accepted manuscript online: September 11, 2019
Version of record online: && &&
,
&&&&
&&&&
www.angewandte.org
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
These are not the final page numbers!

Angewandte
Research Articles
Chemie
Research Articles
Oxygenation Reactions
Mild and selective: The methyl-selective
a-oxygenation of various tertiary amines
to the corresponding formamides (30
examples) is achieved by employing
copper and a moderately hindered
nitroxyl radical (DMN-AZADO or 1-Me-
AZADO) under mild conditions. The high
regioselectivity of this catalyst system
stems from steric restriction of the
amine–N-oxyl interactions.
S. Nakai, T. Yatabe, K. Suzuki, Y. Sasano,
Y. Iwabuchi, J. Hasegawa, N. Mizuno,
K. Yamaguchi*
&&&& — &&&&
Methyl-Selective a-Oxygenation of
Tertiary Amines to Formamides by
Employing Copper/Moderately Hindered
Nitroxyl Radical (DMN-AZADO or 1-Me-
AZADO)
Angew. Chem. Int. Ed. 2019, 58, 2 – 11
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
&&&&
These are not the final page numbers!