Environmentally friendly chemoselective oxidation of primary aliphatic amines by using a biomimetic eiectrocatalytic system

Article abstract of DOI:10.1002/chem.200700876

Environmentally friendly oxidation of primary aliphatic amines to imines has been successfully achieved, under metal-free conditions, by the use of diverse electrogenerated o-azaqui-none mediators. High catalytic performance, together with high chemoselec

Full text of DOI:10.1002/chem.200700876

DOI: 10.1002/chem.200700876
Environmentally Friendly Chemoselective Oxidation ofPrimary Aliphatic
Amines by Using a Biomimetic Electrocatalytic System
N
Martine Largeron,*[a] Ang›le Chiaroni,[b] and Maurice-Bernard Fleury[a]
Abstract: Environmentally friendly ox-
idation of primary aliphatic amines to
imines has been successfully achieved,
under metal-free conditions, by the use
of diverse electrogenerated o-azaqui-
none mediators. High catalytic perfor-
mance, together with high chemoselec-
tivity, were observed with electron-
poor o-azaquinone catalysts generated
from 2-aminoresorcinol derivatives.
Similar to copper amine oxidase en-
zymes, these mediators exhibited lower
reactivity toward a-branched primary
amines and no reactivity toward secon-
dary amines. In the case of 3,4-amino-
phenol derivatives lacking a 2-hydroxy
group, the generated o-azaquinone spe-
cies failed to catalyze the oxidation of
the amine to the corresponding imine.
Further mechanistic considerations al-
lowed a rationalization of the crucial
role of the 2-hydroxy group in convert-
ing a catalytically inert species into a
highly effective biomimetic catalyst.
Keywords: amines
tivity electrochemistry
enzyme models · oxidation
·
chemoselec-
A
·
·
Introduction
imines.[3] A noteworthy example is the biomimetic catalytic
aerobic oxidation of secondary amines involving ruthenium
amine complexes as key intermediates. This methodology,
which tolerates important substrate classes, affords both ke-
timines and aldimines in good yields and with high selectivi-
ty.[3c]
The oxidation of amines to imines is of current and intense
interest owing to the importance of imines as versatile syn-
thetic intermediates. In particular, imines can act as electro-
philic reagents in a plethora of reactions, including reduc-
tions, additions, condensations, and cycloadditions.[1] In
recent years, considerable efforts have been directed to-
wards the development of new, mild, and general oxidation
procedures for the synthesis of imines from secondary
amines. Among such procedures, the stoichiometric method
using the hypervalent iodine reagent 2-iodoxybenzoic acid
(IBX) allowed the direct oxidation of diverse secondary
amines to the corresponding imines under mild conditions
and in excellent yields.[2] Likewise, metal-catalyzed oxidation
reactions were found to be efficient and widely applicable
methods for converting various secondary amines into
In contrast to the extensive work on secondary amines,
comparatively little attention has been devoted to the oxida-
tion of primary amines, probably because the corresponding
imines, in which a second a-amino hydrogen is available,
usually constitute intermediate products that are rapidly de-
hydrogenated to nitriles.[3b,4] Furthermore, when primary
amines are treated with IBX, the corresponding carbonyl
species is isolated, even with the application of short reac-
tion times, following hydrolysis of the initially formed imine
product in situ.[2] However, it has recently been reported
that a series of uracil-annulated heteroazulene derivatives,
as well as some related compounds, were able to catalyze
the oxidation of some primary amines to produce imines in
situ, by photo-irradiation under aerobic conditions, whereas,
except in the case of benzylamine, no reaction took place
under aerobic and thermal conditions.[5]
The general interest in amine oxidation chemistry has
also stimulated efforts to mimic the biological activities of
amine dehydrogenases/oxidases toward primary amines.[6] In
these systems, the amine is not dehydrogenated but reacts
with a carbonyl group of a quinone cofactor leading to a
Schiff-base intermediate, which is then hydrolyzed to the
corresponding aldehyde and aminophenol products. The
[a] Dr. M. Largeron, Prof. Dr. M.-B. Fleury
UMR 8638 CNRS-UniversitØ Paris Descartes
FacultØ des Sciences Pharmaceutiques et Biologiques
4, avenue de LꢀObservatoire, 75270 Paris Cedex 06 (France)
Fax : (+33)01-4407-3588
E-mail: martine.largeron@univ-paris5.fr
[b] Dr. A. Chiaroni
Laboratoire de Cristallochimie
Centre National de la Recherche Scientifique
Institut de Chimie des Substances Naturelles
1, avenue de la Terrasse
91198 Gif-sur-Yvette Cedex (France)
996
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 996 – 1003

FULL PAPER
latter is oxidatively recycled to the starting quinone cofactor
with the elimination of ammonia.[7] Although this is not in
itself a suitable path to mimic the generation of imines, in
the absence of water, the Schiff-base intermediate can un-
dergo a direct addition of the amine affording the N-alkyli-
denealkylamine condensation product instead of the alde-
hyde, together with the aminophenol product. Consequently,
several synthetic models of naturally occurring quinones
have been developed, and good catalytic efficiency has been
observed in the catalytic oxidation of benzylamine to N-ben-
zylidenebenzylamine in organic media under metal-free con-
ditions.[8] However, these model systems failed to oxidize
unactivated primary amines under the same experimental
conditions, with the exception of the metal ion complex of a
tryptophan tryptophylquinone model compound, which was
able to oxidize aliphatic amines in anhydrous organic media.
In contrast, no reaction took place in the absence of the
metal ion.[9]
lysts, that closely approach the activity and specificity of
amine oxidase enzymes, might provide important guidelines
for designing inhibitors capable of regulating human
enzyme activity. In this paper, we present a full account of
the biomimetic catalytic oxidation of primary amines to al-
kylimines. In particular, through variation of the structure of
the electrogenerated 3,4-azaquinone mediator, we disclose
the specific role of the 1-carbonyl substituent (COR) and
the 2-hydroxy group in facilitating operation of the catalytic
process.
Results and Discussion
Choice ofthe reaction conditions : First, we performed opti-
mization studies of the 2ox-mediated catalytic process using
3,3-dimethylbutylamine as the amine substrate (Table 1). Al-
A few years ago, we showed that electrogenerated o-aza-
quinone 1ox (Scheme 1) acted as an effective biomimetic cat-
alyst for the oxidation of benzylamine under metal-free con-
Table 1. Representative screening conditions for the catalytic oxidation
of 3,3-dimethylbutylamine.[a]
Entry Anode
Hg
Solvent Supporting
electrolyte
Current
Yield[c] [%]
45
efficiency [%][b]
1
MeOH TEAHFP
98
2Carbon MeOH TEAHFP
90
40
3
4
5
6
7
8
9
10
Pt
Pt
Pt
Pt
Pt
Pt
Pt
Pt
MeOH TEAHFP
MeCN TEAHFP
CH2Cl2 TEAHFP
MeOH TEAP
MeOH LiClO4
MeOH TEAHFP
MeOH TEAHFP
MeOH TEAHFP
100
78
40
98
100
63
46
32
16
46
46
30
35
Scheme 1. Oxidation of primary amines mediated by the electrogenerat-
ed biomimetic catalyst 1ox.
74
9240
ditions, through the pyridoxal-like transamination process
reported for amine oxidase cofactors. The catalytic cycle
produced the reduced catalyst 1red and N-benzylidenebenzyl-
[a] Reagents: (2ox)=0.4 mm (entries 1–8), 0.2m m (entries 9 and 10); (3,3-
dimethylbutylamine)=2 0 mm (entries 1–7 and 9), 40 mm (entry 8), 10 mm
(entry 10). [b] The electrolysis time was 7 h and the controlled anodic po-
tential was +600 mV vs SCE. [c] The N-alkylidenealkylamine was isolat-
ed by conversion to the corresponding 2,4-dinitrophenylhydrazone
(DNPH) by aqueous acidic work-up of the oxidized solution with 2,4-di-
ACHTREUNG
ACHTREUNG
subsequent electrochemical reduction of the exhaustively
oxidized solution.[10a] Further, we demonstrated that, in con-
trast to other existing amine oxidase mimics,[8] 1ox was also
active toward aliphatic amines in the absence of a metal ion.
An expedient investigation of the performance of this bio-
mimetic electrocatalytic system led to a preliminary commu-
nication.[10b]
Then, we decided to explore further the potential of the
biomimetic electrocatalytic system with two objectives. First,
the 1ox-mediated catalytic oxidation of primary amines al-
lowed the generation of unstable N-alkylidenealkylamines,
without any stoichiometric reagents, under environmentally
friendly conditions. These conditions are highly favorable
from a synthetic viewpoint, in particular for using the imine
in situ for further reactions. Second, from a biological point
of view, we thought that the design of small artificial cata-
nitrophenylhydrazine.
TEAHFP: tetraethylammonium hexafluorophosphate; c.p.e.: controlled
potential electrolysis.
TEAP:
tetraethylammonium
perchlorate;
though the catalytic efficiencies of 1ox and 2ox were equiva-
lent (entries 1 and 2, Table 2), 2ox was considered to be the
more attractive compound because the synthesis of the re-
duced form 2red required only two steps from commercially
available 2-nitroresorcinol, whereas four steps were involved
in the preparation of the previously used compound 1red
For preparative-scale controlled potential electrolysis
(c.p.e.), the utilization of a platinum grid as the anode,
methanol as the solvent, and tetraethylammonium hexa-
fluorophosphate (TEAHFP) as the supporting electrolyte
gave optimal results (entry 3, Table 1). The Pt anode was
[10b]
.
Chem. Eur. J. 2008, 14, 996 – 1003
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
997

M. Largeron et al.
Table 2. Choice of the electrocatalyst.[a]
After determining the Pa potential by cyclic voltammetry,
controlled potential electrolysis was used as a preparative
method for isolation of the products resulting from the cata-
lytic oxidation of 3,3-dimethylbutylamine. When the control-
led potential of the Pt anode was fixed at +600 mV vs SCE,
the anodic current remained constant for a long time, and
the current efficiency obtained by electrolysis for 7 h was
100%, indicating that no side reaction took place under the
experimental conditions used (entry 3, Table 1). Note that a
high potential value was intentionally chosen, because of
the continuous shift of the peak Pa observed in the course
of the electrolysis to +500 mV vs SCE, when the amine con-
centration was no longer sufficient to ionize the 4-hydroxy
group of 2red. These results indicated that the 2red/2ox system
behaved as a redox mediator for the indirect electrochemi-
cal oxidation of 3,3-dimethylbutylamine to the correspond-
ing N-alkylidenealkylamine, according to the ionic transami-
nation mechanism previously reported.[10a] After exhaustive
controlled potential electrolysis, the unstable alkylimine was
isolated by converting it to the 2,4-dinitrophenylhydrazone
(DNPH) by aqueous acidic work-up of the oxidized solution
with 2,4-dinitrophenylhydrazine (see the Experimental Sec-
tion). Note that the yield could not exceed 50%, because
5 mmol of 3,3-dimethylbutylamine gave only 2.5 mmol of
the corresponding N-alkylidenealkylamine. Furthermore,
control studies indicated that the amount of N-alkylideneal-
kylamine produced either by simple autoxidation or by elec-
trochemical oxidation of 3,3-dimethylbutylamine in the ab-
sence of catalyst 2ox was negligible. Taken together, the re-
sults indicate that 2ox exhibited high catalytic efficiency in
the oxidation of this non-activated primary amine since the
yield of DNPH reached 46% (entry 3, Table 1). After ex-
haustive electrolysis, the catalyst 2ox was irreversibly con-
sumed, as corroborated by the anodic current, which re-
mained negligible upon further addition of the amine sub-
strate. This result was in agreement with the fact that lower-
ing the amount of catalyst 2ox from 2mol% to 1 mol% de-
creased the current efficiency as well as the yield of DNPH
(entries 8 and 9, Table 1).
Entry Reduced
catalyst
R
R’
Current
efficiency
[%]
Yield[b]
[%]
1
2
3
4
5
6
7
8
1red
2red
3red
4red
5red
6red
7red
8red
OH COPh
OH COMe
OH COiBu
OH COC6H11
OH NO2
100
100
90
94
74
16
100
98
46
46
42
44
29
6
OH
H
OH CO
A
46
44
OH CO(4-SO2Me-
C6H4)
9
9red
OH CO(2,6-diMe-C6H3)
OH CO(2-MeO-C6H4)
OH CO(2-HO-C6H4)
96
100
2
44
46
0
0
0
10
11
12
13
10red
11red
12red
13red
7
[c]
H
COPh
–
–
[c]
Me COPh
[a] Reagents and c.p.e. conditions: (1ox–13ox)=0.4 mm, (3,3-dimethylbutyl-
amine)=2 0 mm, MeOH, RT, Pt anode (E=+600 mV vs SCE), 7 h (en-
tries 1–11), 2h (entries 12 and 13). [b] The N-alkylidenealkylamine was
isolated by conversion to the corresponding DNPH by aqueous acidic
work-up of the oxidized solution with 2,4-dinitrophenylhydrazine. [c] No
comparison can be made because no catalytic process developed in this
case.
more desirable from the point of view of green chemistry,
though an Hg anode could also be used without noticeable
change (entry 1, Table 1). MeCN (entry 4, Table 1) and
CH2Cl2 (entry 5, Table 1) proved not to be suitable solvents
for the catalytic oxidation of the amine, probably because
strong solvation of the o-azaquinone 2ox by methanol may
be required to enhance the electrophilicity of its quinonoid
moiety, thereby favoring the nucleophilic attack of the
amine. Among the supporting electrolytes tested, TEAHFP
was preferred over the others (entries 6 and 7, Table 1)
owing to the potential explosion risk associated with per-
chlorate anions. A combination of 5 mmol of 3,3-dimethyl-
butylamine with 0.1 mmol of 2red, which corresponds to
2mol% of the catalyst 2ox, was found to be ideal for the re-
action.
Under the optimized conditions, the cyclic voltammogram
of compound 2red (0.4 mm) in deaerated MeOH showed an
oxidation peak, Pa, at +500 mV vs SCE, due to a two-elec-
tron process, the sweep rate being 0.1 VsÀ1. The addition of
3,3-dimethylbutylamine (20 mm) had two effects: first, the
peak Pa was shifted to 0 mV vs SCE as a result of ionization
of the 4-hydroxy group; second, a slight increase in the
anodic peak intensity was noted, which suggested that 3,4-
azaquinone 2ox was protected from its subsequent polymeri-
zation reaction because it could act as a catalyst for the oxi-
dation of the amine. Similar effects have previously been
observed in relation to the catalytic activity of similar quino-
noid species.[11]
At this point, we suspected that the presence of both the
1-acetyl group and the 2-hydroxy substituent was of overrid-
ing importance for the catalytic efficiency of 2ox. In this con-
text, it may be noted that the 5-hydroxy proton of 2,4,5-tri-
hydroxyphenylalanine quinone (TPQ) and the 1-NH pyrrole
proton of pyrroloquinoline quinone were found to be pre-
requisites for the catalytic activity of these quinonoid cofac-
tors.[12] So, to evaluate the specific role of each substituent
in the course of the catalytic process, the autorecycling oxi-
dation of 3,3-dimethylbutylamine was examined using vari-
ously substituted o-azaquinone mediators.
Catalyst screening: Having established a reliable set of con-
ditions, we examined the catalytic efficiency of a variety of
electrogenerated 3,4-azaquinone species for the oxidation of
3,3-dimethylbutylamine. The results are summarized in
Table 2. High catalytic performance was observed with elec-
tron-poor o-azaquinone entities generated from substituted
998
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 996 – 1003

Biomimetic Electrocatalytic Amine Oxidation
FULL PAPER
2-aminoresorcinol reduced catalysts 1red–4red bearing a car-
bonyl substituent (COR) at the 1-position, with high current
efficiencies (90–100%) and yields of DNPH ranging from
42to 46% (entries 1–4, Table 2). Replacing COR by a nitro
group decreased the yield to 29% (entry 5, Table 2), where-
as anodic oxidation of 2-aminoresorcinol 6red generated a
poorly reactive o-azaquinone catalyst, which mainly decom-
posed to melanin-like polymers, giving only 6% of DNPH
(entry 6, Table 2). In the specific case of 2-aminoresorcinol
derivatives bearing a benzophenone framework 7red–11red
(entries 7–11, Table 2), we sought to modulate the electro-
philic properties of the 3,4-azaquinone catalyst by the at-
tachment of substituents to the 1-benzoyl moiety. Surprising-
ly, the yield of DNPH was good, regardless of the electronic
and/or steric effects of the substituents. In particular, the
presence of an electron-donating group (entry 10, Table 2)
did not interfere with the catalytic process, except when a
hydroxy group was introduced at the 2’-position (entry 11,
Table 2). In this case, the catalytic efficiency of 3,4-azaqui-
none 11ox decreased markedly, and only 7% of DNPH was
obtained. This outcome could be rationalized by X-ray crys-
tallographic analyses of the reduced catalysts 7red, 9red, and
11red, which showed the presence of a hydrogen bond be-
tween the 2-hydroxy substituent and the carbonyl group of
the benzophenone skeleton (Figure 1).[13] As a consequence,
the substituted phenyl group was twisted out of the plane,
thereby affecting the transmission of the substituent effects.
Similarly, a hydrogen bond could be expected for the 3,4-
azaquinone oxidized form, which would produce the same
effects. In the case of compound 11red (entry 11, Table 2), a
second hydrogen bond was evidenced between the 2’-hy-
droxy substituent and the carbonyl group of the benzophe-
none framework (Figure 1(c)). Although the strengths of
these two hydrogen bonds were almost the same, it could be
expected that, in solution, the carbonyl group of the benzo-
phenone skeleton of the oxidized form 11ox would bind pref-
erentially to the 2’-phenolic group rather than to the more
acidic 2-phenolic substituent of the o-azaquinone moiety.
Entries 12and 13 in Table 2deserve special note because
the electrochemical oxidation of 3,4-aminophenol deriva-
tives 12red and 13red, which lack the 2-hydroxy substituent,
generated 3,4-azaquinone species 12ox and 13ox that were
devoid of catalytic efficiency toward the oxidation of 3,3-di-
methylbutylamine. Under the aforementioned conditions,
the cyclic voltammogram of compound 12red showed an oxi-
dation peak, Pa, due to a two-electron process at +200 mV
vs SCE, the sweep rate being 0.1 VsÀ1, indicating that the
3,4-aminophenol derivative 12red was less easily oxidized
than the corresponding 2-aminoresorcinol derivative 1red
.
When the controlled potential of the Pt anode was fixed
at +600 mV vs SCE, a potential at which 12red could be oxi-
dized to the 3,4-azaquinone form 12ox, a decrease in anodic
electrolysis current was observed immediately while the so-
lution became brown. At the end of the electrolysis, four
electrons had been transferred per molecule of 12red. These
results indicated that the electrogenerated 3,4-azaquinone
12ox did not act as a catalyst, but rather reacted with the
amine to yield a Michael adduct, which, after a subsequent
two-electron oxidation reaction, spontaneously decomposed
to melanin-like polymers. To confirm the ability of the 2-
substituent to prevent competing formation of the Michael
adduct, 3,4-aminophenol 13red, bearing a 2-methyl group,
was used as the reduced catalyst (entry 13, Table 2). When
subjected to the same experimental conditions, the produced
3,4-azaquinone 13ox also failed to catalyze the oxidation of
3,3-dimethylbutylamine to the corresponding N-alkylide-
nealkylamine. It was thus concluded that the presence of the
2-hydroxy substituent was an essential requirement for suc-
cessful operation of the catalytic process. The precise role of
the 2-hydroxy group in converting a catalytically inert o-aza-
quinone species into a highly effective biomimetic catalyst
will be disclosed below.
In view of the high catalytic efficiency of o-azaquinone 2ox
in the oxidation of 3,3-dimethylbutylamine and the facile
synthesis of its precursor 2red (only two steps), the synthetic
potential of this biomimetic electrocatalytic system was sub-
sequently explored through variation of the amine substrate.
Catalytic oxidation ofvarious amines using o-azaquinone
2ox: Results relating to the oxidation of representative
amines, under conditions optimized for 3,3-dimethylbutyl-
Figure 1. ORTEP views of 2-aminoresorcinol reduced catalysts: a) 7red
,
AHCTREUNG
b) 9red, and c) 11red. Displacement ellipsoids are drawn at the 30% proba-
bility level.[13]
Chem. Eur. J. 2008, 14, 996 – 1003
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
999

M. Largeron et al.
Table 3. Chemoselective oxidation of primary aliphatic amines mediated
times did not improve the yields of DNPH, but rather led to
the generation of a second product identified as the osazone
(1,2-bis-DNPH), as previously observed with o-azaquinone
by the biomimetic electrocatalyst 2ox.[a]
catalyst 1ox generated from 2-aminoresorcinol derivative
[10b]
1red
.
Further investigations would be necessary to ration-
alize the formation of the osazone, which seems to be fa-
vored when the starting amine is not sterically encumbered
by b- or g-branching. Interestingly, as for the copper amine
oxidase enzymes, a-branched amines (entries 10 and 11,
Table 3) were found to be inferior substrates for the bio-
mimetic electrocatalyst 2ox (compare, for example, entry 1
with entry 10), whereas secondary amines (entries 12and 13,
Table 3) were not reactive at all.
Entry
Amine
substrate
Current
efficiency [%]
Yield [%][b]
B
A
1
2
3
100
100
90
50
2500
46
2300
1250
25[c]
Mechanistic considerations: The above reported results indi-
cate that high catalytic performance, together with high che-
moselectivity, were observed with electron-poor o-azaqui-
none catalysts generated from 2-aminoresorcinol derivatives.
The question then arose as to the exact role of the 1-acetyl
and 2-hydroxy groups in converting a catalytically inert 3,4-
azaquinone species into a highly effective electrocatalyst.
For this purpose, we thoroughly re-examined the ionic trans-
amination mechanism that we reported previously
(Scheme 2).[10]
4
98
98
40
2000
5
6
100 422
8238
1900
[c]
7
8
9
6
922
60
1300
1100
1000
22
20
64
10
54
25
1250
11
24
7
350
[d]
12
13
–
0
0
0
0
[d]
–
[a] Reagents and c.p.e. conditions: (2ox)=0.4 mm, (amine substrate)=
2 0 mm, MeOH, RT, Pt anode (E=+600 mV vs SCE), 7 h (entries 1–11),
2h (entries 12 and 13). [b] The N-alkylidenealkylamine was isolated by
conversion to the corresponding DNPH by aqueous acidic work-up of
the oxidized solution with 2,4-dinitrophenylhydrazine; yields relative to
the amine substrate (A) and to the mediator (B). [c] The yield of DNPH
was lower than that expected from the current efficiency as a result of
partial conversion of the unstable alkylimine into volatile aldehyde on
the time scale of anodic electrolysis. [d] No comparison can be made be-
cause no catalytic process developed in this case.
ing N-benzylidenebenzylamine in quantitative yield since
the current efficiency and the yield of DNPH reached 100%
and 50%, respectively (entry 1, Table 3).
Scheme 2. Role of the 2-hydroxy group in the ionic transamination mech-
anism.
Non-activated aliphatic primary amines also proved to be
good substrates for the catalyst 2ox (entries 2–6, Table 3),
with isolated yields of DNPH ranging from 38 to 46% (1900
to 2300% relative to the mediator). In the specific cases of
isopentylamine (entry 3, Table 3) and ethanolamine (entry 7,
Table 3), the yields (25% and 26%) were lower than those
expected on the basis of the high current efficiencies as a
result of partial conversion of the unstable N-alkylideneal-
kylamines into volatile aldehydes on the time scale of
anodic electrolysis. Interestingly, the presence of the alcohol
group did not interfere with amine oxidation, indicating a
high degree of functional group tolerance (entry 7, Table 3).
3,4-Azaquinone 2ox was less effective in oxidizing more hy-
drophobic longer-chain amines such as phenylpropylamine
or hexylamine, as shown by the lower current efficiencies
obtained (entries 8 and 9, Table 3) and by the yields of
DNPH, which were roughly halved. Extended reaction
Obviously, the presence of the 1-acetyl group (or another
electron-withdrawing group) not only favors attack of the
amine at the 3-position of the electrogenerated 3,4-azaqui-
none species (step 2, Scheme 2), but also facilitates a-proton
abstraction and the subsequent electron flow from the a-
carbon to the o-azaquinone moiety, which aromatizes to the
Schiff base 2’ox (step 3, Scheme 2). Accordingly, 3,4-azaqui-
none 6ox, which is devoid of an electron-withdrawing group
at C-1, as well as 11ox, in which the electron-withdrawing
effect exerted by the carbonyl group of the benzophenone
skeleton is attenuated because of the conjugation of this
group with the 2’-hydroxyphenyl group, were found to be
poor catalysts for the oxidation of amines, promoting in-
stead the competitive polymerization reaction.
1000
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 996 – 1003

Biomimetic Electrocatalytic Amine Oxidation
FULL PAPER
Even more crucially, the presence of the 2-hydroxy group
proved to be an essential requirement for the successful op-
eration of the catalytic process. In fact, the activation of the
imine function for further nucleophilic attack by the amine,
leading to the extrusion of N-alkylidenealkylamine (step 4,
Scheme 2), is provided by an intramolecular hydrogen bond
between the 2-hydroxy group and the imine nitrogen, gener-
ating the highly reactive cyclic transition state 2’ox. This acti-
vated nucleophilic attack of amine, which leads to an aminal
intermediate (see Scheme 2in ref. [10b]), would constitute a
driving force for the overall transamination mechanism,
thereby preventing any competitive Michael addition reac-
tion. Similar effects of a 2-phenolic hydroxy group on the re-
activity of ketimine derivatives have recently been reported
in the literature.[14] Note that the activation of the imine
function through intramolecular hydrogen bonding also sup-
ports the preference for the use of methanol over MeCN
and CH2Cl2 as the solvent.
through variation of the structure of the o-azaquinone redox
mediator. High catalytic performance has been observed
with electron-poor o-azaquinone catalysts electrogenerated
from 2-aminoresorcinol derivatives, whereas o-azaquinone
species electrogenerated from 3,4-aminophenol derivatives
lacking the 2-hydroxy group proved to be devoid of any cat-
alytic efficiency. Given the facile synthesis of its precursor
2red (only two steps) and its high catalytic efficiency, 3,4-aza-
quinone 2ox, bearing 1-acetyl and 2-hydroxy substituents, is
considered to be the most promising biomimetic electrocata-
lyst of the studied series. Accordingly, 2ox exhibited the
same substrate specificity as copper amine oxidase enzymes,
that is, poor reactivity with a-branched amines and no reac-
tivity toward secondary amines. Finally, our biomimetic elec-
trocatalytic system displayed two features that are most
often associated with enzymatic systems. First, the reaction
was enhanced through the participation of 1-acetyl and 2-hy-
droxy substituents, as they prevented the competing forma-
tion of Michael adducts. Second, the presence of the active
2-hydroxy group (analogous to the 5-hydroxy group of
TPQ),[12a] which is engaged in an intramolecular hydrogen
bond with the imine nitrogen to form a highly reactive
Schiff-base cyclic transition state, proved to be an essential
requirement for successful operation of the catalytic process.
Synthetic applications: The biomimetic catalytic oxidation
of primary aliphatic amines reported here produced chemi-
cally inaccessible alkylimines from amines, without any stoi-
chiometric reagents, under environmentally friendly condi-
tions. These conditions are particularly favorable for using
the imine in situ for further reactions. To this end, we have
recently shown that the tautomeric enamine form of the N-
alkylidenealkylamine generated by the catalytic oxidation of
an R1R2CHCH2NH2 amine (alone or in the presence of a
second amine, R3-NH2), could be efficiently deployed, under
well-defined conditions, as the dienophile in an inverse-elec-
tron-demand Diels–Alder (IEDDA) reaction with an o-aza-
quinone catalyst acting as the heterodiene (Scheme 3).[15]
Experimental Section
General considerations: 1H NMR spectra were recorded on a Bruker
AC-300 spectrometer operating at 300 MHz. Chemical shifts, d, are given
in ppm relative to TMS; coupling constants, J, are given in hertz. The
measurements were carried out using standard pulse sequences. Chemi-
cals were commercial products of the highest available purity and were
used as supplied. Reduced catalyst 1red was synthesized in four steps from
commercially available 2-nitroresorcinol,[16] while only two steps were re-
quired for the synthesis of reduced catalysts 2red–4red, 7red, and 10red using
the same starting material (see the supporting information of ref. [15d]).
The synthesis of the reduced catalysts 8red, 9red, and 13red is also described
in ref. [15d]. Reduced catalysts 5red, 6red, and 12red were synthesized ac-
cording to previously reported procedures.[17] Reduced catalyst 11red was
synthesized by demethylation of compound 10red by heating at 508C for
1.5 h with 6 equiv of AlCl3 in dry toluene according to a standard proto-
col.[16]
Scheme 3. 3,4-Azaquinone-mediated cascade reaction affording highly
functionalized 1,4-benzoxazine derivatives.
(3-Amino-2,4-dihydroxyphenyl)(2’-hydroxyphenyl)methanone
(11red):
Yellow solid (71 mg; 50%); m.p. 1768C (petroleum ether/diethyl ether);
1H NMR (300 MHz,[D6]DMSO, 258C, TMS): d=12.50 (brs, 1H), 9.86
(brs, 1H), 7.33 (t, J=7.5 Hz, 1H), 7.18 (dd, J=7.5, 1.5 Hz, 1H), 6.92(m,
2H), 6.55 (d, J=8.5, 1H), 6.31 ppm (d, J=8.5 Hz, 1H).
This cascade reaction, for which both cycloaddition partners
were generated in situ at room temperature under metal-
free conditions, allowed the one-pot regiospecific synthesis
of highly functionalized 2-alkylamino-1,4-benzoxazine deriv-
atives, which proved to be potent neuroprotective agents
both in vitro and in vivo.[15c]
X-ray analysis: A small plate of dimensions 0.25”0.20”0.075 mm was
used. Empirical formula C13H11NO4, M=245.23, T=293 K; monoclinic
system, space group P21/a, Z=4, a=7.712(5), b=7.712(6), c=
19.472(8) Š, b=100.82(4)8, V=1137.5(12) Š3, 1calcd =1.432gcm À3
,
F-
A
ACHTREUNG
tions was measured with a Nonius Kappa-CCD diffractometer, of which
2073 were unique. Refinement of 178 parameters against F2 led to
R1(F)=0.0475 calculated from 1367 observed reflections as I>2s(I), and
wR2(F2)=0.1228 considering all 2073 data. Goodness of fit=1.048.
Conclusion
X-ray crystallographic analysis of7 red: A small yellow plate of dimensions
0.50”0.50”0.025 mm, crystallized from a mixture of petroleum ether/di-
ethyl ether, was used. Empirical formula C13H10FNO3, M=247.22, T=
293 K; monoclinic system, space group P21/a, Z=4, a=8.233(4), b=
New insights into the scope and mechanism of the biomim-
etic catalytic oxidation of primary aliphatic amines to alkyli-
mines under metal-free conditions have been obtained
7.179(3), c=19.442(8) Š, b=91.76(2)8, V=1148.6(9) Š3, 1calcd
=
Chem. Eur. J. 2008, 14, 996 – 1003
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1001

M. Largeron et al.
1.430 gcmÀ3, F
G
A
Chem. Int. Ed. 2003, 42, 1480–1483; c) J. S. M. Samec, A. H. Ell,
J. E. Bäckvall, Chem. Eur. J. 2005, 11, 2327–2334; d) J. R. Wang, Y.
Fu, B.-B. Zhang, X. Cui, L. Liu, Q.-X. Guo, Tetrahedron Lett. 2006,
47, 8293–8297; e) H. Choi, M. P. Doyle, Chem. Commun. 2007,
745–747.
intensity data were measured with a Nonius Kappa-CCD diffractometer
giving 4725 monoclinic reflections, of which 2617 were unique. Refine-
ment of 176 parameters against F2 led to R1(F)=0.0448 calculated from
1698 observed reflections as I>2s(I), and wR2(F2)=0.1255 considering
all 2617 data. Goodness of fit=1.047.
[4] For recent examples, see: a) K. Mori, K. Yamaguchi, T. Mizugaki,
K. Ebitani, K. Kaneda, Chem. Commun. 2001, 461–462; b) Y.
Maeda, T. Nishimura, S. Uemura, Bull. Chem. Soc. Jpn. 2003, 76,
2399–2403; c) K. Yamaguchi, N. Mizuno, Chem. Eur. J. 2003, 9,
4353–4361; d) S. Kamiguchi, A. Nakamura, A. Suzuki, M. Kodo-
mari, M. Nomura, Y. Iwasawa, T. Chihara, J. Catal. 2005, 230, 204–
213, and references therein.
X-ray crystallographic analysis of9 red: A small orange prismatic crystal of
dimensions 0.50”0.35”0.20 mm, crystallized from a mixture of petrole-
um ether/diethyl ether, was used. Empirical formula C15H15NO3, M=
257.28, T=293 K; orthorhombic system, space group P212121, Z=4, a=
9.256(3), b=11.940(4), c=12.006(4) Š, b=908, V=1326.9 Š3, 1calcd
1.288 gcmÀ3, F
A
of 11236 reflections was measured with a Nonius Kappa-CCD diffrac-
tometer, of which 3027 were unique. Refinement of 187 parameters
against F2 led to R1(F)=0.0425 calculated from 2440 observed reflections
as I>2s(I), and wR2(F2)=0.1156 considering all 3027 data. Goodness of
fit=1.044.
[5] a) Y. Mitsumoto, M. Nitta, J. Org. Chem. 2004, 69, 1256–1261; b) S.-
I. Naya, M. Nitta, Tetrahedron 2004, 60, 9139–9148; c) M. Nitta, D.
Ohtsuki, Y. Mitsumoto, S.-I. Naya, Tetrahedron 2005, 61, 6073–
6081; d) S.-I. Naya, Y. Yamaguchi, M. Nitta, Tetrahedron 2005, 61,
7384–7391.
[6] M. Mure, Acc. Chem. Res. 2004, 37, 131–139, and references there-
in.
[7] M. Mure, S. A. Mills, J. P. Klinman, Biochemistry 2002, 41, 9269–
9278, and references therein.
CCDC 649690 (11red), 649691 (7red), and 649692( 9red) contain the supple-
mentary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[8] a) M. Mure, J. P. Klinman, J. Am. Chem. Soc. 1995, 117, 8707–8718;
b) Y. Lee, L. M. Sayre, J. Am. Chem. Soc. 1995, 117, 11823–11828,
and references therein; c) S. Itoh, N. Tanaka, S. Haranou, T. Ando,
M. Komatsu, Y. Ohshiro, S. Fukuzumi, J. Org. Chem. 1996, 61,
8967–8974, and references therein; d) A. C. Rinaldi, A. Rescigno,
A. Rinaldi, E. Sanjust, Bioorg. Chem. 1999, 27, 253–288, and refer-
ences therein; e) K. Q. Ling, J. Kim, L. M. Sayre, J. Am. Chem. Soc.
2001, 123, 9606–9611.
Electrochemistry: Cyclic voltammetry measurements were made with a
Radiometer-Tacussel PRG 5 multipurpose polarograph, which was used
only as a rapid-response potentiostat. Triangular waveforms were sup-
plied by a Tacussel GSTP 4 function generator. Current–potential curves
were recorded with a Schlumberger SI 8312instrument. The cell was a
Radiometer-Tacussel CPRA water-jacketed cell operating at a tempera-
ture of 258C. The working electrode was a platinum disk, which was care-
fully polished with an aqueous suspension of alumina before acquiring
each voltammogram. The counter electrode was a Tacussel Pt 11 plati-
num electrode. The reference electrode, to which all quoted potentials
are referred, was an aqueous saturated calomel electrode (SCE), which
was isolated from the bulk solution in a glass tube by a fine-porosity frit.
[9] S. Itoh, M. Tanaguchi, N. Takada, S. Nagatomo, T. Kitagawa, S. Fu-
kuzumi, J. Am. Chem. Soc. 2000, 122, 12087–12097.
[10] a) M. Largeron, M.-B. Fleury, J. Org. Chem. 2000, 65, 8874–8881;
b) M. Largeron, A. Neudçrffer, M.-B. Fleury, Angew. Chem. 2003,
115, 1056–1059; Angew. Chem. Int. Ed. 2003, 42, 1026–1029.
[11] K. Kano, M. Nakagawa, K. Takagi, T. Ikeda, J. Chem. Soc. Perkin
Trans. 2 1997, 1111–1119, and references therein.
[12] a) M. Mure, J. P. Klinman, J. Am. Chem. Soc. 1993, 115, 7117–7127;
b) S. Itoh, M. Mure, M. Ogino, Y. Ohshiro, J. Org. Chem. 1991, 56,
6857–6865.
[13] It may be noted that, in the three analyzed compounds, all the hy-
drogen atoms belonging to the nitrogen atom N3 and to the differ-
ent hydroxy groups are engaged in either inter- or intramolecular
hydrogen bonds. As shown in Figure 1, a strong intramolecular hy-
drogen bond is established between the hydroxy group O2-H and
the carbonyl oxygen atom O7, with the following characteristics: for
General procedure for the o-azaquinone-mediated autorecycling oxida-
tion ofamines : Controlled-potential electrolysis was carried out in a cy-
lindrical, three-electrode divided cell (9 cm diameter), using an electronic
potentiostat. In the main compartment, a platinum grid (60 cm2 area)
served as the anode (working electrode). A platinum sheet was placed in
the concentric cathodic compartment (counter electrode), which was sep-
arated from the main compartment by a glass frit. The SCE was as de-
scribed above. The electrolyte solution (0.02molL À1 tetraethylammoni-
um hexafluorophosphate in methanol) was poured into the anodic and
cathodic compartments, as well as into the glass tube that contained the
SCE. Reduced catalyst (0.1 mmol) and an excess of primary aliphatic
amine (5 mmol) were then added to the solution in the main compart-
ment (250 mL), and the resulting solution was oxidized under nitrogen at
room temperature at +600 mV vs SCE (initial current 30–40 mA). After
exhaustive electrolysis, that is, when a negligible current was recorded
(0.5–1.0 mA), the solution was worked-up by the addition of 2,4-dinitro-
phenylhydrazine reagent (2.5 mmol in 5 mL of H2SO4, 15 mL of EtOH,
and 5 mL of water),[18] the stoichiometry reflecting the fact that 5 mmol
of the primary amine gave only 2.5 mmol of the N-alkylidenealkylamine.
After 1 h, the resulting solution was concentrated to a volume of 40 mL.
The solid was collected by filtration, washed with water, and dried in a
vacuum desiccator. The identity and purity of the 2,4-dinitrophenylhydra-
zone (DNPH) was confirmed by TLC and 1H NMR, after comparison
with an authentic sample.
7
red, distances O2···O7=2.524(2) Š, O2–H=0.92Š, H···O7 =1.68 Š,
angle O2···H···O7=151.48; for 9red, distances O2···O7=2.573(2) Š,
O2–H=0.84 Š, H···O7=1.81 Š, angle O2···H···O7=151.58; for 11red
,
distances O2···O7=2.562(2) Š, O2–H=0.82Š, H···O7 =1.80 Š,
angle O2···H···O7=153.48, and in this last compound, there is a
second hydrogen bond between the hydroxy group O2’–H and the
oxygen atom O7, with the respective geometry as follows: distances
O2’···O7=2.584(2) Š, O2’–H=0.93 Š, H···O7=1.76 Š, angle
O2’···H···O7=145.98, aligning the three oxygen atoms (angle
O2···O7···O2’=178.28). In the three compounds 7red, 9red, and 11red
,
the two phenyl rings are tilted by 43.9, 72.6, and 42.38, respectively.
[14] a) H. Miyabe, Y. Yamaoka, Y. Takemoto, Synlett 2004, 2597–2599;
b) H. Miyabe, Y. Yamaoka, Y. Takemoto, J. Org. Chem. 2006, 71,
2099–2106.
[15] a) M. Largeron, A. Neudçrffer, M. Vuilhorgne, E. Blattes, M.-B.
Fleury, Angew. Chem. 2002, 114, 852–855; Angew. Chem. Int. Ed.
2002, 41, 824–827; b) E. Blattes, M.-B. Fleury, M. Largeron, J. Org.
Chem. 2004, 69, 882–889; c) E. Blattes, B. Lockhart, P. Lestage, M.-
B. Fleury, M. Largeron, J. Med. Chem. 2005, 48, 1282–1286; d) D.
Xu, A. Chiaroni, M.-B. Fleury, M. Largeron, J. Org. Chem. 2006, 71,
6374–6381.
[1] For a review on the synthetic use of imines, see: J. P. Adams, J.
Chem. Soc. Perkin Trans. 1 2000, 125–139.
[2] a) K. C. Nicolaou, C. J. N. Mathison, T. Montagnon, Angew. Chem.
2003, 115, 4111–4117; Angew. Chem. Int. Ed. 2003, 42, 4077–4082;
b) K. C. Nicolaou, C. J. N. Mathison, T. Montagnon, J. Am. Chem.
Soc. 2004, 126, 5192–5201.
[3] For selected recent examples, see: a) A. H. Ell, J. S. M. Samec, C.
Brasse, J. E. Bäckvall, Chem. Commun. 2002, 1144–1145; b) K. Ya-
maguchi, N. Mizuno, Angew. Chem. 2003, 115, 1518–1521; Angew.
[16] R. Larget, B. Lockhart, B. Pfeiffer, A. Neudçrffer, M.-B. Fleury, M.
Largeron, Bioorg. Med. Chem. Lett. 1999, 9, 2929–2934.
1002
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 996 – 1003

Biomimetic Electrocatalytic Amine Oxidation
FULL PAPER
[17] a) H. Lindner, Chem. Ber. 1923, 56, 1870–1871; b) P. Piateck, J. Ka-
lisiak, J. Jurcka, Tetrahedron Lett. 2004, 45, 3309–3311; c) G. M.
Tseiltlin, B. V. Tokarev, V. N. Kulagin, J. Org. Chem. USSR (Engl.
Transl.) 1982, 931–934.
[18] P. Lue, W.-Q. Fan, X.-J. Zhou, Synthesis 1989, 692–693.
Received: June 8, 2007
Revised: August 23, 2007
Published online: November 8, 2007
Chem. Eur. J. 2008, 14, 996 – 1003
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1003