Oxidation of organic substrates by molecular oxygen mediated by
N-hydroxyphthalimide (NHPI) and acetaldehyde
Cathy Einhorn,* Jacques Einhorn, Ce´line Marcadal and Jean-Louis Pierre
Laboratoire de Chimie Biomime´tique, UMR CNRS 5616, Universite´ J. Fourier, 38041 Grenoble, France
Various organic substrates, in particular hydrocarbons, are
efficiently oxidized under mild conditions using molecular
oxygen, N-hydroxyphthalimide and acetaldehyde, in the
absence of metal catalysts.
addition of a second equivalent of acetaldehyde and another 48
h standing period led to increased yields of 67 and 9%,
respectively. Obviously the rate of aldehyde autoxidation,
controlled by oxygen diffusion in the solution, plays a decisive
role in the outcome of the reaction. High aldehyde autoxidation
rates lead to fast NHPI degradation, no more ethylbenzene
being oxidized after its complete disappearance.
The use of molecular oxygen for the selective oxidation of
organic substrates, especially hydrocarbons under mild condi-
tions, is still a major challenge for organic chemistry.1 It
constitutes an environmentally safe alternative to more conven-
tional oxidants used in stoichiometric amounts and is therefore
of high economic value in industrial chemistry.2 Catalysis by
transition metal complexes is the main way of controlling the
partial oxidation of alkanes, alkenes and aromatic hydro-
carbons.3 Although they have been known for a long time,
autoxidations without transition metal catalysis are often of low
efficiency and selectivity. They proceed via a free radical chain
mechanism promoted by catalytic amounts of radical initiators.4
Aldehydes are exceptions among organic compounds as their
autoxidation rates are very high with long propagation chains
[reactions (1) and (2)] even at room temperature and atmos-
Passive diffusion of oxygen in bulk solutions is very sensitive
to geometric characteristics of the reactor and therefore
difficulties in the reproducibility and scaling up of such
experiments may be anticipated. We found that an efficient and
reproducible way to perform such oxidations is to add slowly the
aldehyde to the vigorously stirred reaction medium; when 1
equiv. of acetaldehyde was added over 5 h to a solution of
ethylbenzene and NHPI under oxygen, and fast stirring
maintained for an additional 5 h period, a 66% yield of
acetophenone and a 4% yield 1-phenylethanol were obtained.†
Table 1 shows further examples of oxidations performed under
the same conditions: 4-nitro-1-ethylbenzene is oxidized with
lower efficiency than ethylbenzene itself (run 3) whereas
4-methoxy-1-ethylbenzene is oxidized almost quantitatively
(runs 4, 5). 1-Phenylethanol is oxidized only slowly to
acetophenone (runs 6, 7) indicating that a direct pathway
leading from ethylbenzene to acetophenone is likely. In the case
of cumene, demethylation giving acetophenone competes with
hydroxylation (run 8). Indane and isochromane are oxidized
almost quantitatively (runs 9, 10). Surprisingly diphenyl-
methane is oxidized with a lower conversion (49%) than
ethylbenzene (70%), with a relatively high alcohol:ketone ratio
(run 11). A high conversion (94%) is obtained on the other hand
with xanthene (run 12). Inactivated hydrocarbons are oxidized
at a slow but nevertheless significant rate: 8% conversion of
cyclohexane is observed after a 72 h reaction time, with a 1:7
alcohol:ketone ratio (run 13). Adamantane is oxidized almost
exclusively to adamantan-1-ol with only trace amounts of
adamantan-2-one (runs 14, 15).
In order to obtain information about the reaction mechanism
of this oxidation, several parameters have been investigated. No
oxidation occurs in the presence of either hydroquinone or
TEMPO (1 mol%), confirming the free radical pathway of the
reaction. No oxidation is observed at room temperature with
NHPI alone.‡ Cooxidation of ethylbenzene and acetaldehyde (1
equiv.) in the absence of NHPI furnishes, after 48 h, only 7%
ketone and 2% alcohol, indicating a strong cooperative effect of
both NHPI and aldehyde under cooxidation conditions. Vir-
tually no oxidation is observed when acetaldehyde is replaced
by peracetic acid, clearly indicating that the species involved in
these oxidations is not the peracetic acid 2 (R = Me) formed in
situ, but is probably the acetylperoxy radical 1. This radical 1
may in turn oxidize NHPI to phthalimide N-oxyl 3 [reaction
(3)], a fairly stable but highly reactive free radical, which has
been proposed as a key intermediate in NHPI-mediated
oxidations.8d,8h,9c
O
•
+
(1)
R
C
O
R
O2
•
OO
1
O
H
O
O
•
R C O
+
R
+
(2)
R
R
•
OO
OOH
2
pheric pressure. This property has been used in the so-called
‘cooxidation’ processes, in which a mixture of an aldehyde and
another organic substrate is submitted to molecular oxygen. The
autoxidation of the aldehyde then promotes the oxidation of the
less reactive partner. Aldehyde-mediated cooxidations have
been used for the epoxidation of alkenes,5 Bayer–Villiger
oxidation of ketones6 and oxidation of alcohols and of some
hydrocarbons.7 We describe herein the oxidation of organic
substrates, in particular hydrocarbons, under aldehyde-pro-
moted cooxidation conditions, in the presence of
N-hydroxyphthalimide (NHPI). NHPI has been used previously
as an electron carrier for electrochemical oxidations.8 More
recently various oxidations have been conducted using molec-
ular oxygen and NHPI, generally combined with transition
metal salts, under nonelectrochemical conditions.9 Our new
combination of O2–aldehyde–NHPI reveals several interesting
features. When a 0.2 m solution of ethylbenzene, acetaldehyde
(1 equiv.) and NHPI (0.1 mol%) in MeCN was vigorously
stirred at room temperature under an oxygen atmosphere for 48
h, a 26% yield of acetophenone and a 6% yield of 1-phenyl-
ethanol were formed, with no further reaction after a longer
reaction time. No further oxidation was observed after the
addition of a second equivalent of acetaldehyde and an
additional 48 h stirring time. No NHPI was recovered at the end
of this experiment but an almost quantitative amount of
phthalimide was isolated. Surprisingly, the same experiment
carried out in the absence of stirring led to substantially
different results; after keeping the reaction medium under
oxygen without any stirring for 48 h the yields of ketone and
alcohol were 50 and 6%, respectively, while TLC analysis of the
solution indicated only a partial degradation of NHPI. The
O
O
O
O
•
(3)
O
+
+
R
N
OH
R
N
•
OO
OOH
O
O
3
Chem. Commun., 1997
447
Einhorn, Cathy
