Alkane oxidation with molecular oxygen using a new efficient catalytic system: N-hydroxyphthalimide (NHPI) combined with Co(acac)n (n = 2 or 3)

Article abstract of DOI:10.1021/jo951970l

A novel class of catalysts for alkane oxidation with molecular oxygen was examined. N-Hydroxyphthalimide (NHPI) combined with Co(acac)_n (n = 2 or 3) was found to be an efficient catalytic system for the aerobic oxidation of cycloalkanes and alkylbenzenes under mild conditions. Cycloalkanes were successfully oxidized with molecular oxygen in the presence of a catalytic amount of NHPI and Co(acac)₂ in acetic acid at 100°C to give the corresponding cycloalkanones and dicarboxylic acids. Alkylbenzenes were also oxidized with dioxygen using this catalytic system. For example, toluene was converted into benzoic acid in excellent yield under these conditions. Ethyl- and butylbenzenes were selectively oxidized at their α-positions to form the corresponding ketones, acetophenone, and 1-phenyl-1-butanone, respectively, in good yields. A key intermediate in this oxidation is believed to be the phthalimide N-oxyl radical generated from NHPI and molecular oxygen using a Co(II) species. The isotope effect (k_H/k_D) in the oxidation of ethylbenzene and ethylbenzene-d₁₀ with dioxygen using NHPI/Co(acac)₂ was 3.8.

Full text of DOI:10.1021/jo951970l

4520
J . Org. Chem. 1996, 61, 4520-4526
Alk a n e Oxid a tion w ith Molecu la r Oxygen Usin g a New Efficien t
Ca ta lytic System : N-Hyd r oxyp h th a lim id e (NHP I) Com bin ed w ith
Co(a ca c)n (n ) 2 or 3)†
Yasutaka Ishii,* Takahiro Iwahama, Satoshi Sakaguchi, Kouichi Nakayama, and
Yutaka Nishiyama
Department of Applied Chemistry, Faculty of Engineering, Kansai University, Suita, Osaka 564, J apan
Received November 6, 1995X
A novel class of catalysts for alkane oxidation with molecular oxygen was examined. N-
Hydroxyphthalimide (NHPI) combined with Co(acac)n (n ) 2 or 3) was found to be an efficient
catalytic system for the aerobic oxidation of cycloalkanes and alkylbenzenes under mild conditions.
Cycloalkanes were successfully oxidized with molecular oxygen in the presence of a catalytic amount
of NHPI and Co(acac)2 in acetic acid at 100 °C to give the corresponding cycloalkanones and
dicarboxylic acids. Alkylbenzenes were also oxidized with dioxygen using this catalytic system.
For example, toluene was converted into benzoic acid in excellent yield under these conditions.
Ethyl- and butylbenzenes were selectively oxidized at their R-positions to form the corresponding
ketones, acetophenone, and 1-phenyl-1-butanone, respectively, in good yields. A key intermediate
in this oxidation is believed to be the phthalimide N-oxyl radical generated from NHPI and molecular
oxygen using a Co(II) species. The isotope effect (kH/kD) in the oxidation of ethylbenzene and
ethylbenzene-d10 with dioxygen using NHPI/Co(acac)2 was 3.8.
In tr od u ction
of molecular oxygen, as well as for the oxygenation of
benzylic compounds such as fluorene8 and the dehydro-
genation of alcohols9 with dioxygen under mild condi-
tions. However, it is difficult to oxidize alkanes such as
cyclohexane and toluene using NHPI alone.
We recently found that the catalytic activity of NHPI
is markedly enhanced by the presence of a very small
amount of Co(acac)n (n ) 2 or 3) (0.05 equiv with respect
to NHPI) as a cocatalyst. Thus, cycloalkanes and alkyl-
benzenes can be efficiently oxidized to the corresponding
carbonyl compounds in an oxygen atmosphere under
moderate conditions. This report presents our findings
on the catalysis of alkane oxidations with atmospheric
oxygen using NHPI combined with Co(acac)n (n ) 2 or
3).
Oxidative catalytic transformations of organic com-
pounds with molecular oxygen (dioxygen) play a very
important role in organic synthesis.1 In particular,
selective catalytic oxidation using dioxygen as the pri-
mary oxidant represents a critical technology and is an
area of continued research and development. Several
transition metal-catalyzed selective oxidations of alkanes
involving the combined use of dioxygen and reducing
agents such as H2,2 NaBH4,3 RCHO,4 etc., have been
reported. Recently, halogenated metalloporphyrins have
been shown to be efficient catalysts for the direct reaction
of alkanes with dioxygen without coreductants or sto-
ichiometric oxidants to give alcohols and/or carbonyl
compounds.5 However, the development of an aerobic
oxidation system in the absence of a reducing agent
remains a very important and challenging subject in
oxidation chemistry.
Resu lts
1. Oxid a tion of Cycloa lk a n es w ith Dioxygen
Usin g NHP I Com bin ed w ith Tr a n sition Meta ls. The
oxidation of cyclohexane (1) to a cyclohexanone (2)/
N-Hydroxyphthalimide (NHPI) was first used by Masui
et al. as an efficient electron carrier in the electrochemical
oxidation of secondary alcohols to ketones.6 In the course
of our study on the aerobic oxidation of benzylic com-
pounds catalyzed by vanadomolybdophosphates,7 we
found that NHPI is a unique catalyst for the activation
(5) (a) Lyons, J . E.; Ellis, P. E., J r. Metalloporphyrins in Catalytic
Oxidations; Sheldon, R. A., Ed.; Dekker: New York, 1994; p 291, and
references cited therein. (b) Ellis, P. E., J r.; Lyons, J . E. J . Chem. Soc.,
Chem. Commun. 1989, 1188. (c) Ellis, P. E., J r.; Lyons, J . E. J . Chem.
Soc., Chem. Commun. 1989, 1190. (d) Ellis, P. E., J r.; Lyons, J . E. J .
Chem. Soc., Chem. Commun. 1989, 1316. (e) Ellis, P. E., J r.; Lyons, J .
E. Catal. Lett. 1989, 3, 389. (f) Lyons, J . E.; Ellis, P. E., J r. Catal.
Lett. 1991, 8, 45. (g) Ellis, P. E. J r.; Lyons, J . E. Coord. Chem. Rev.
1990, 105, 181. (h) Lyons, J . E.; Ellis, P. E., J r.; Durante, V. A. Studies
in Surface Science and Catalysis; Grasselli, R. Ed.; Elsevier: New York,
1991; Vol. 67, p 99. (i) Ellis, P. E., J r.; Lyons, J . E. Prepr. Am. Chem.
Soc., Div. Pet. Chem. 1990, 35, 174. (j) Lyons, J . E.; Ellis, P. E., J r.;
Wagner, R. W.; Thompson, P. B.; Gray, H. B.; Hughes, M. E.; Hodge,
J . A. Prepr. Am. Chem. Soc., Div. Pet. Chem. 1992, 37, 307.
(6) (a) Masui, M.; Ueshima, T.; Ozaki, S. J . Chem. Soc., Chem.
Commun. 1983, 479. (b) Masui, M.; Hara, S.;Ueshima, T.; Ozaki, S.
Chem. Pharm. Bull. 1987, 31, 4209.
† Dedicated to Clayton H. Heathcock on the occasion of his 60th
birthday.
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
(1) (a) Sheldon, R. A.; Kochi, J . K. Metal-Catalyzed Oxidations of
Organic Compounds; Academic Press: New York, 1981. (b) Hill, C. L.
Activation and Functionalization of Alkanes; Academic Press: New
York, 1989. (c) Simandi, L. Catalytic Activation of Dioxygen by Metal
Complexes; Kluwer Academic Publisher: Boston, 1992. (d) The Activa-
tion of Dioxygen and Homogeneous Catalytic Oxidation; Barton, D. H.
R., Martell, A. E., Sawyer, D. T., Eds.; Plenum Press: New York, 1993.
(e) Mennier, B. Chem. Rev. 1992, 92, 1411. (f) Busch, D. H.; Alcock, N.
W. Ibid. 1994, 94, 585.
(2) Tabushi, I.; Yazaki, A. J . Am. Chem. Soc. 1981, 103, 7371.
(3) Tabushi, I.; Yazaki, A. J . Am. Chem. Soc. 1979, 101, 6456.
(4) (a) Kaneda, K.; Haruna, S.; Imanaka, T.; Kawamoto, K. J . Chem.
Soc., Chem. Commun. 1990, 1467. (b) Yamada, T.; Takai, T.; Rhode,
O.; Mukaiyama, T. Chem. Lett. 1991, 1. (c) Murahashi, S.-I.; Oda, Y.;
Naota, T. J . Am. Chem. Soc. 1992, 114, 7913. (d) Mukaiyama, T.
Yamada, T. Bull. Chem. Soc. J pn. 1995, 68, 17. (e) Hamamoto, M.;
Nakayama, K.; Nishiyama, Y.; Ishii, Y. J . Org. Chem. 1993, 58, 6421.
(7) (a) Hamamoto, M.; Nakayama, K.; Nishiyama, Y.; Ishii, Y. J .
Org. Chem. 1993, 58, 6421. (b) Nakayama, K.; Hamamoto, M.;
Nishiyama, Y.; Ishii, Y. Chem. Lett. 1993, 1699. (c) Fujibayashi, S.;
Nakayama, K.; Nishiyama, Y.; Ishii, Y. Chem. Lett. 1994, 1345.
(8) Ishii, Y.; Nakayama, K.; Takeno, M.; Sakaguchi, S.; Iwahama,
T.; Nishiyama, Y. J . Org. Chem. 1995, 60, 3934.
(9) Iwahama, T.; Sakaguchi, S.; Nishiyama, Y.; Ishii, Y. Tetrahedron
Lett. 1995, in press.
S0022-3263(95)01970-0 CCC: $12.00 © 1996 American Chemical Society

Alkane Oxidation with Molecular Oxygen
J . Org. Chem., Vol. 61, No. 14, 1996 4521
Ta ble 1. Oxid a tion of Cycloh exa n e (1) w ith Dioxygen
Ca ta lyzed by NHP I in th e P r esen ce of Meta l Sa lts u n d er
Va r iou s Rea ction Con d ition sa
yield, %b
NHPI
temp conv
run (mol %) transition metal
(°C)
(%)
2
3c
1
2
3
4
5
10
-
10
10
5
10
10
10
10
10
10
10
10
10
10
-
100
100
100
100
100
75
1
trace
0
0
Co(acac)2
Co(acac)2
Co(acac)3
Co(acac)2
Co(acac)2
Co(acac)3
Co(acac)3
Co(acac)2
Co(acac)2
Co(OAc)2‚4H2O
Mn(acac)3
Cu(OAc)2‚H2O
Fe(acac)3
Ni(acac)2
trace trace
45
42
28
30
22
13
48
38
43
44
21
5
32
35
41
48
52
78
37
38
31
3
38
40
37
30
37
13
39
48
44
77
29
trace
-
cyclohexanol (4) mixture (ketone/alcohol (K/A) oil) is the
first step in the two-step process for the production of
adipic acid (3). Although there are several variants of
this oxidation, the principal method is the autoxidation
of cyclohexane in the presence of a metal catalyst such
as Co or Mn salt.10,11 An alternative cyclohexane oxida-
tion uses a higher concentration of Co(III) acetate under
oxygen pressure (20-30 atm).12 However, the homoge-
neous catalytic oxidations of 1 using these methods have
several problems, e.g., oxidative attack on the C-H bonds
is slow and requires vigorous reaction conditions. To
overcome these limitations, a new catalyst for selective
oxidation with dioxygen must be identified. Hence, the
catalytic oxidation of 1 with atmospheric oxygen under
mild conditions is a challenge in industrial catalysis.
The oxidation of 1 with dioxygen (1 atm) using a new
catalyst, NHPI, under selected conditions is shown in
Table 1. Although a benzylic compound such as fluorene
was efficiently oxidized to fluorenone in high yield (80%)
with molecular oxygen (1 atm) in the presence of NHPI
(10 mol %) in benzonitrile at 100 °C,8 1 was not oxidized
by NHPI alone under these conditions. Thus, the effects
of several transition metal salts on the NHPI-catalyzed
oxidation of 1 in a dioxygen atmosphere were examined.
An initial survey of representative transition metals
demonstrated that a very small amounts of cobalt salts
such as Co(acac)n (n ) 2 or 3) significantly facilitated the
NHPI-catalyzed aerobic oxidation of 1 under moderate
conditions. For example, the oxidation of 1 in the
presence of NHPI (10 mol %) and Co(acac)2 (0.5 mol %)
in acetic acid at 100 °C for 6 h (standard conditions) gave
cyclohexanone 2 (32%) and adipic acid 3 (38%) as the
main products in the 45% conversion of 1 (run 3).13
Oxidation did not take place in the absence of NHPI
under these conditions (run 2). When Co(acac)3 was used
in place of Co(acac)2 in this oxidation, 1 was converted
to 2 and 3 at a slightly lower conversion (42%) (run 4).
Similar effects of Co(acac)2 and Co(acac)3 on the oxidation
of 1 were also observed at 75 °C (runs 6 and 7). Oxidation
was considerably retarded by the use of acetonitrile as a
solvent, while the selectivity of 1 to 2 at 75 °C was
improved from 52% to 78% (run 8). Oxidation using
Co(OAc)2‚4H2O as a cocatalyst was similar to that with
Co(acac)2 (run 11). Mn(acac)3, which is often used as a
catalyst for autoxidation, was also effective in this
oxidation. It is interesting to note that the addition of
Mn(acac)3 led to 3 rather than 2 in high selectivity (77%).
This oxidation was slightly enhanced by adding Cu-
(OAc)2‚H2O. In contrast to the effect of Co and Mn salts
6
7
75
75
8d
9e
10f
11
12
13
14
15
100
100
100
100
100
100
100
43
90
-
0
a
1 (5 mmol) was allowed to react with dioxygen (1 atm) in the
presence of NHPI and transition metal (0.5 mol%) in acetic acid
(12.5 mL) for 6 h. Based on 1 reacted. c Yield of dimethyl adipate
b
d
after esterification with excess methanol. Acetonitrile was used
as solvent. Co(acac)2 (1 mol%) was used. f Co(acac)2 (0.25 mol%)
e
was used.
in the NHPI-catalyzed oxidation of 1, Fe(acac)3 had only
a slight effect on the oxidation, and Ni(acac)2 had no
effect.
When the quantity of NHPI was halved (5 mol %), 1
was oxidized at a slightly lower conversion (28%) to form
2 and 3 in 41% and 37% selectivities, respectively.
However, the effect of the concentration of Co(acac)2 as
the cocatalyst was clearly much less than that of NHPI.
For the oxidation of 1 by NHPI (10 mol %) in the presence
of Co(acac)2 (1 mol %) (run 9), the results were almost
the same as those with 0.5 mol % of Co(acac)2 (run 3). It
is noteworthy that the present aerobic oxidation of 1
resulted in 3 in higher selectivity, since the autoxidation
of 1 in the presence of Co and Mn salts is known to lead
to K/A oil (2 and 4) rather than 3 as the main product.12
In particular, it is interesting that 1 was converted into
3 in one step in 77% selectivity by the NHPI/Mn(acac)3
system under atmospheric oxygen at 100 °C.
On the basis of these results, several cycloalkanes were
oxidized under standard conditions, i.e., in the presence
of NHPI (10 mol %) and Co(acac)2 (0.5 mol %) in an
oxygen atmosphere (1 atm) using acetic acid at 100 °C
for 6 h (Table 2).
For the oxidation of cyclopentane (5) under these
conditions, the results were similar to those of 1 (run 1).
However, cyclooctane (8) gave 1,4-cyclooctanedione (10)
(16%) in addition to cyclooctanone (9) (50%) and suberic
acid (11) (16%) in 93% conversion (run 2). Even when
the amount of NHPI was reduced to 5 mol %, 8 was
oxidized with a higher conversion (86%) (run 3). In
contrast to 1, which was slightly oxidized by NHPI alone,
8 could be oxidized to 9 (63%), 10 (8%), and 11 (17%) in
a 37% conversion by NHPI in the absence of any metals
(run 4). The 1,4-diketone 10 is believed to be formed via
intramolecular hydrogen abstraction by a transient per-
oxy radical generated from 8, as will be discussed later
(Scheme 1). Indeed, the successive oxidation of 9 under
these conditions led to dicarboxylic acid 11, while dike-
tone 10 was not formed at all. Cyclododecane (12) was
also converted into the corresponding ketone 13 and
dicarboxylic acid 14 at a satisfactory conversion (66%)
(run 5). During the oxidation of these cycloalkanes,
alcohols such as cyclohexanol and cyclooctanol are be-
(10) Parshall, G, W.; Ittel. S, D. Homogeous Catalysis; 2nd ed.; J ohn
Wiley and Sons: New York, 1992; p 246, and reference cited therein.
(11) Simandi, L. Catalytic Activation of Dioxygen by Metal Com-
plexes; Kluwer Academic Publisher: Boston, 1992; p 84, and reference
cited therein.
(12) (a) Steeman, J . W. M.; Kaarsemaker, S.; Hoftyzer, P. J . Chem.
Eng. Sci. 1961, 14, 139. (b) Miller, S. A. Chem. Process Eng. (London)
1969, 50, 63. (c) Tanaka, K. Chem. Tech. 1974, 555, and references
cited therein.
(13) By the GC-MS measurement of the reaction product of 1, the
formation of glutaric acid (up to 5%) and cyclohexyl acetate (up to 2%)
other than 2 and 3 was confirmed.

4522 J . Org. Chem., Vol. 61, No. 14, 1996
Ishii et al.
a
Ta ble 2. Oxid a tion of Cycloa lk a n es w ith Dioxygen Ca ta lyzed by NHP I in th e P r esen ce of Co(a ca c)2
lieved to be formed, but very small amounts of the
Sch em e 1. A P ossible Rea ction P a th for th e
Oxid a tion of Cycloocta n e (8) by th e NHP I-O2
System
corresponding acetates were detected using GC-MS. It
is believed that these alcohols are easily oxidized to
ketones and/or dicarboxylic acids since they are more
reactive than the starting cycloalkanes. In fact, cyclo-
hexanol 4 was oxidized to 2 in 92% yield in acetonitrile
at 75 °C.9
Methylcyclohexane (15) gave keto carboxylic acid 18
as a major product along with 2-methylcyclohexanone
(16) and 1-methylcyclohexanol (17) (run 6). The inde-
pendent oxidations of 16 and 17 under these reaction
conditions were carried out to reveal the reaction path
to 18. The oxidation of 16 gave 18, while 17 was a less
reactive substrate. Thus, 18 was a further oxidation
product of 16, but not of 17. The oxidation of adaman-
tane (19) at 75 °C for 3 h gave 1-adamantanol (20) (71%),
in which the tertiary C-H bond was selectively oxygen-
ated, together with small amounts of 2-adamantanone
(21) (9%) and 1,3-adamantanediol (22) (17%) in 65%
conversion (run 7). The product ratio of the tertiary C-H
bond to the secondary one was approximately 7.8. The
aerobic oxidation of 19 using the PW9Fe2Ni heteropolya-
nion reportedly gives 20 (76%), 21 (12%), and 2-adaman-
tanol (12%) in 29% conversion.14 On the other hand, the
oxidation of n-alkane such as n-octane (23) by this system
gave a mixture of 2-, 3-, and 4-octanols 24 and the
corresponding octanones 25 (run 8).
to be oxidized with dioxygen in the presence or absence
of transition metals.15
The oxidation of several alkylbenzenes with dioxygen
was attempted using the NHPI/Co(acac)2 system (Table
3, runs 1-11). The oxidation of toluene (26) with NHPI
2. Oxid a tion of Alk ylben zen es Usin g NHP I Com -
bin ed with Co(acac)2. Various alkylbenzenes are known
(14) Mizuno, N.; Tateishi, M.; Hirose, T.; Iwamoto, M. Chem. Lett.
1993, 2137.

Alkane Oxidation with Molecular Oxygen
J . Org. Chem., Vol. 61, No. 14, 1996 4523
a
Ta ble 3. Oxid a tion of Alk ylben zen es w ith Dioxygen Ca ta lyzed by NHP I in th e P r esen ce of Co(a ca c)2
run
substrate
toluene (26)
conv (%)
products (%)b
1
2c
3
4
5
92
9
benzoic acid (27) (99)
27 (trace)
p-tert-butylbenzoic acid (29) (95)
p-methoxybenzoic acid (31) (85)
acetophenone (33) (93)
26
p-tert-butyltoluene (28)
p-methoxytoluene (30)
ethylbenzene (32)
n-butylbenzene (34)
p-xylene (36)
36
o-xylene (39)
39
cumene (42)
97
98
91
90
94
99
92
99
31
6
7
1-phenyl-1-butanone (35) (67)
p-toluic acid (37) (78), terephthalic acid (38) (15)
37 (23), 38 (68)
8d,e
9
o-toluic acid (40) (80), phthalic acid (41) (14)f
40 (71), 41 (23)
10d,e
11
33 (54), 2-phenyl-2-propanol (43) (10), phenol (44) (17)
a
Substrate (5 mmol) was allowed to react with dioxygen (1 atm) in the presence of NHPI (10 mol %) and Co(acac)2 (0.5 mol %) in acetic
acid (12.5 mL) at 100 °C for 6 h. Based on substrate reacted. c In the absence of Co(acac)2. NHPI (20 mol %) was used. e Reaction was
b
d
carried out at 100 °C for 12 h. f Yield of dicarboxylic acid was estimated as dimethyl ester after esterification with excess methanol.
(10 mol %) in the presence of Co(acac)2 (0.5 mol %) in
acetic acid in an oxygen atmosphere at 100 °C for 6 h
exclusively gave benzoic acid (27) (>99%) at 92% conver-
sion (run 1). The same oxidation with 26 by NHPI alone
produced only a trace amount of 27 (run 2). Toluene (26)
has been reported to be oxidized with air in the presence
of cobalt(II) 2-ethylhexanoate at 140-190 °C and up to
10 atm of pressure to give 27 in about a 80% yield at
40-65% conversion.15a Consequently, the NHPI/Co-
(acac)2 system is thought to be a useful catalytic system
for the aerobic oxidation of 26. Similarly, p-tert-butyl-
toluene (28) and p-methoxytoluene (30) were oxidized to
p-tert-butylbenzoic acid (29) and p-methoxybenzoic acid
(31), respectively, in good yields (runs 3 and 4). In the
oxidation of ethylbenzene (32), acetophenone (33) was
obtained in good yield (run 5). n-Butylbenzene (34) was
likewise oxidized to give 1-phenyl-1-butanone (35) in
slightly lower selectivity (67%). To evaluate the potential
for the oxidation of the disubstituted alkylbenzenes, the
oxidation of xylene was examined. p-Xylene (36) was
F igu r e 1. Time dependence curves of O2 uptakes for the
oxidation of ethylbenzene (32) under atmospheric pressure of
dioxygen by various catalysts. Conditions: Ethylbenzene (32)
converted into the corresponding mono- and dicarboxylic
acids 37 and 38, the ratio of which depended on the
reaction time. The oxidation of 36 under the standard
conditions gave p-toluic acid (37) in a 78% selectivity with
a 94% conversion. When the reaction time was prolonged
to 12 h, terephthalic acid (38) was obtained in 68% yield.
However, it was difficult to convert o-xylene (39) to
phthalic acid (41) in high selectivity (runs 9 and 10).
Cumene (42) is known to be oxidized by dioxygen to give
cumene hydroperoxide, which is converted to phenol and
acetone.16 However, 42 was oxidized with difficulty by
the NHPI/Co(acac)2 system to form acetophenone (33)
(54%), 2-phenyl-2-propanol (43) (10%), and phenol (44)
(17%) in 31% conversion. The low conversion of 42 was
due to the formation of phenol 44. The oxidation of 42
by this system in the presence of 44 (10 mol %) was
markedly inhibited, and only trace amounts of 33 and
43 were formed.
To gain additional insight into NHPI/Co(acac)2-cata-
lyzed aerobic oxidation, the absorption rate of dioxygen
during the oxidation of 32 with several catalytic systems
was measured using a constant-pressure absorption
apparatus. Figure 1 shows the time-dependence curves
of the O2 uptake of 32 under atmospheric pressure (1
atm) at 80 °C. It is interesting to compare the rate of O2
uptake by 32 in the NHPI/Co(acac)2 and NHPI/Co(acac)3
(10 mmol) was allowed to react with dioxygen (1 atm) in acetic
acid (25 mL) at 80 °C. (1) NHPI (10 mol %), Co(acac)2 (0.5 mol
%); (2) NHPI (10 mol %), Co(acac)3 (0.5 mol %); (3) NHPI (10
mol %); (4) Co(acac)2 (0.5 mol %); (5) AIBN (5 mol %), Co(acac)2
(0.5 mol %).
systems. No induction period was observed in O2 uptake
by 32 with NHPI/Co(acac)2, while that with the NHPI/
Co(acac)3 system did not occur until after about 1.5 h.
Almost no O2 uptake was observed in the oxidation of
32 by NHPI, Co(acac)2, or Co(acac)3 alone. In the same
oxidation using the AIBN/Co(acac)2 system, the results
were almost the same except for O2 uptake in the early
stage of the reaction by radicals generated from AIBN.
The isotope effect during the present oxidation was
estimated by measuring the oxygen uptake by ethylben-
zene (32) and ethylbenzene-d10 (32-d10) (Figure 2). The
observed isotope effect, kH/kD, was approximately 3.74.
In addition, the oxidation of 32 in the presence of
hydroquinone (1 mol %) under the standard conditions
did not occur at all. These results strongly suggest that
the present aerobic oxidation proceeds via a reaction
pathway similar to that in free radical autooxidation.
(15) (a) Pershall, G. W.; Ittel, S, D. Homogeous Catalysis, 2nd ed.;
J ohn Wiley and Sons: New York, 1992; p 256. (b) Sheldon, R. A.; Kochi,
J . K. Metal-Catalyzed Oxidations of Organic Compounds; Academic
Press: New York, 1981; p 315.
(16) Reichle, W. T.; Konrad, F. M.; Brooks, J . R. Benzene and its
Industrial Derivatives; Hancock, E. G., Ed.; Benn: London, 1975.
Discu ssion
Masui et al. suggested that phthalimide N-oxyl (45) is
a key species in the electrochemical oxidation of alcohols

4524 J . Org. Chem., Vol. 61, No. 14, 1996
Ishii et al.
Furthermore, 2-norbornene (46) was allowed to react
with NHPI under an oxygen atmosphere in acetonitrile
at 60 °C. The reaction gave N-(2-hydroperoxybicyclo-
[2.2.1]heptan-2-yloxy)phthalimide (47) in 52% yield (eq
2). Treatment of triphenylphosphine with 47 produced
triphenylphosphine oxide (94%) and N-(hydroxybicyclo-
[2,2,1]heptan-2-yloxy)phthalimide (48) (83%) (eq 3). This
F igu r e 2. Time dependence curves of O2 uptakes for the
oxidation of ethylbenzene (32) and ethylbenzene-d10 (32-d10
)
by NHPI/Co(acac)2-O2 system. Conditions: Ethylbenzene (32)
(10 mmol) was allowed to react with dioxygen (1 atm) in the
presence of NHPI (10 mol %) and Co(acac)2 (0.5 mol %) in acetic
acid (25 mL) at 80 °C.
result definitely shows that the radical species 45 is
smoothly generated from NHPI and molecular oxygen in
the absence of transition metals under mild conditions.
Hence, the oxidation process of alkanes such as cyclooc-
tane 8, which is oxidized by dioxygen using NHPI alone,
can be outlined as in Scheme 1. The first step of the
reaction is thought to involve the generation of the
phthalimide N-oxyl radical 45 from NHPI and dioxygen.
The resulting 45 abstracts a hydrogen atom from the
substrates to form alkyl radicals, subsequent oxygenation
of which by dioxygen produces peroxy radicals, which are
converted to ketones and/or dicarboxylic acids. In some
cases, intramolecular hydrogen abstraction by the result-
ing alkyl peroxy radical occurs, and leads to the formation
of the diketone 10.
Unfortunately, we are currently unable to clearly
explain the role of Co(acac)n as a cocatalyst in NHPI-
catalyzed aerobic oxidation. However, we can make
several proposals which seem to agree with the experi-
mental results.
F igu r e 3. ESR spectrum of N-oxyl radical 45 under dioxygen
in benzonitrile at 80 °C.
to ketones using NHPI as an electron carrier.17 To
confirm the formation of phthalimide N-oxyl (45) in the
present NHPI-catalyzed aerobic oxidation, electron spin
resonance (ESR) measurements were carried out under
selected conditions. When a benzonitrile solution of
NHPI was exposed to dioxygen at 80 °C for 1 h,18 the
ESR spectrum attributed to 45 was clearly observed as
a triplet signal based on hyperfine splitting (hfs) by the
nitrogen atom (g ) 2.0074, aN ) 0.43 mT) (Figure 3). The
g-value and hfs constant observed for 45 were consistent
with those (g ) 2.0073, aN ) 0.423 mT) reported by
Mackor et al.19 No ESR signal was observed under argon.
It is noteworthy that 45 is easily formed by exposing
NHPI to molecular oxygen under moderate conditions.
As mentioned earlier, no induction period was observed
in the oxidation of ethylbenzene 32 with the NHPI/Co-
(acac)2 system. However, oxygen uptake by 32 did not
occur until after about 1.5 h with the NHPI/Co(acac)3
system. This finding is in accord with the results of
adding NHPI alone to Co(acac)2 or Co(acac)3. When
NHPI was added to an acetic acid solution of Co(acac)2,
the color of the solution immediately changed from pink
to violet, and a mixture of several complexes 49 was
obtained. On the other hand, the same procedure with
Co(acac)3 led to no color change of the solution, and most
of the starting materials were recovered unchanged.
However, when ethylbenzene 32 was added to this
solution, the color of the solution gradually changed to
violet and finally became the same as that in the NHPI/
Co(acac)2 system to form complexes similar to those
derived from NHPI and Co(acac)2. This result indicates
that Co(acac)3 is gradually reduced to Co(II) with 32 via
(17) (a) Masui, M.; Hosomi, K.; Tsuchida, K.; Ozaki, S. Chem.
Pharm. Bull. 1985, 33, 4798. (b) Ueda, C.; Nayama, M.; Ohmori, H.;
Masui, M. Chem. Pharm. Bull. 1987, 35, 1372.
(18) ESR spectra were obtained under the following conditions:
sweep width 327 ( 2.5 mT; modulation 0.1 mT; and microwave power
1 mW. Benzonitrile containing 10-2 mmol of NHPI was exposed under
oxygen atmosphere at 80 °C for 1 h. The air in the ESR tube was
replaced by dioxygen gas by means of the freeze-pump-thaw method.
The ESR parameter was determined by using solid Mn2+ (g ) 2.034)
as a standard.
(19) Mackor, A.; Wajer, Th. A. J . W.; de Boer, Th. J . Tetrahedron
1968, 24, 1623.

Alkane Oxidation with Molecular Oxygen
J . Org. Chem., Vol. 61, No. 14, 1996 4525
a well-known one-electron transfer process,20 and the
resulting Co(II) species readily reacts with NHPI to
produce complexes similar to those derived from the
NHPI/Co(acac)2 system. These results well reflect the
differences in the induction period between the NHPI/
Co(acac)2 and NHPI/Co(acac)3 systems during O2 uptake
by 32. The induction period of 1.5 h observed with the
NHPI/Co(acac)3 system may represent the time required
to reach a threshold concentration of Co(II) by one-
electron transfer from 32 to Co(III).
24.4, 40.6, 41.3, 213.7; IR (NaCl) 2941, 2865, 1697, 1445, 1334,
1108, 1091, 946, 822 cm-1
.
1-Meth ylcycloh exa n ol (17): 1H NMR (CDCl3) δ 1.21 (s,
3H), 1.36 (s, 1H), 1.44-1.63 (m, 10H); 13C NMR (CDCl3) δ 22.6,
25.6, 29.5, 39.4, 69.9; IR (NaCl) 3352, 2930, 2859, 1170, 1120,
967, 911 cm-1
.
6-Oxoh ep ta n oic a cid (18): 1H NMR (CDCl3) δ 1.62-1.65
(m, 4H), 2.16 (s, 3H), 2.36-2.40 (m, 2H), 2.45-2.48 (m, 2H),
11.02 (s, 1H); 13C NMR (CDCl3) δ 23.0, 24.0, 29.9, 33.8, 43.2,
179.8, 209.0; IR (NaCl) 3046, 2945, 1712, 1414, 1369, 1235,
1176 cm-1
.
1,3-Ad a m a n ta n d iol (22): 1H NMR (CDCl3) δ 1.61-1.77
On the other hand, 32 and 1 were oxidized using the
complexes 49 obtained from NHPI and Co(acac)2 (eq 4).
32 was oxidized to acetophenone 33 in 69% yield, while
1 failed to be oxidized by these complexes.
(m, 14H), 2.34 (s, 2H); 13C NMR (CDCl3) δ 31.1, 34.5, 43.7,
52.5, 70.3; IR (KBr) 3349, 2931, 1333, 1030 cm-1
.
Gen er a l P r oced u r e for Oxid a tion of Alk ylben zen es.
An acetic acid (12.5 mL) solution of alkylbenzenes (5 mmol),
NHPI (82 mg, 10 mol %), and Co(acac)2 (7.5 mg, 0.5 mol %)
was placed in a three-necked flask equipped with a ballon filled
with O2. The mixture was stirred at 100 °C for 6 h. After
removal of the solvent under reduced pressure, the products
were purified by column chromatography on silica gel to give
the corresponding oxygenated products.
Products 27, 29, 31, 33, 37, 38, 40, 41, and 44 were
identified by comparing of the isolated products with authentic
samples.
1-P h en yl-1-bu ta n on e (35): 1H NMR (CDCl3) δ 0.98-1.04
(t, J ) 7.3 Hz, 3H), 1.71-1.85 (m, 2H), 2.92-2.98 (t, J ) 7.6
Hz, 2H), 7.43-7.58 (m, 3H), 7.95-7.98 (d, J ) 7.9 Hz, 2H);
13C NMR (CDCl3) δ 200.4, 137.1, 132.8, 128.5, 128.0, 40.5, 17.7,
Although the role of the Co(II) species in the NHPI-
catalyzed aerobic oxidation is not fully understood, the
Co(II) species may be related to the generation of the
phthalimide N-oxyl radical 45 from NHPI. However, it
appears that the present results do not detract from the
importance of the discovery of a new mode of dioxygen
activation by the NHPI/Co(acac)n (n ) 2 or 3) system.
13.9; IR (NaCl) 2963, 1688, 1449, 1214, 692 cm-1
.
2-P h en yl-2-p r op a n ol (43): 1H NMR (CDCl3) δ 1.55 (s, 6H),
2.40 (s, 1H), 7.19-7.50 (m, 5H); 13C NMR (CDCl3) δ 31.6, 72.4,
124.4, 126.5, 128.1, 149.1; IR (NaCl) 3374, 2976, 2359, 1446,
1363, 764, 699, 544 cm-1
.
Exp er im en ta l Section
Gen er a l P r oced u r e for Mea su r in g Oxygen -Absor p tion
Ra tes. Oxygen-absorption rates were measured with an
isobaric gas-absorption apparatus in a closed-flow system (2
( 0.1 L oxygen/h) equipped with an electrolyzer using 25 mL
of acetic acid containing ethylbenzene (1.06 g, 10 mmol), NHPI
(163 mg, 10 mol %) and Co(acac)2 (14.7 mg, 0.5 mol %) at 80
°C. Oxygen absorption was periodically measured in the
constant-pressure closed system.
Rea ction of 46 w ith NHP I. An acetonitrile (5 mL)
solution of 2-norbornene (564 mg, 6 mmol) and NHPI (489 mg,
3 mmol) was placed in a three-necked flask equipped with a
balloon filled with O2. The mixture was stirred at 60 °C for
20 h. After the reaction, acetonitrile was removed under
reduced pressure to give a white crystal, which was purified
by diethyl ether (30 mL) to give the hydroperoxide 47 in 52%
yield.
N-(2-Hyd r op er oxybicyclo[2.2.2]h ep ta n -2-yloxy)p h th a l-
im id e (47): 1H NMR (CDCl3) δ 1.16-1.41 (m, 2H), 1.48-1.65
(m, 2H), 2.04 (d, J ) 10.0 Hz, 1H), 2.31 (s, 1H), 2.94 (s, 1H),
4.14 (d, J ) 5.3 Hz, 1H), 4.36 (d, J ) 5.3 Hz, 1H), 7.79-8.55
(m, 4H), 10.65 (s, 1H); 13C NMR (CDCl3) δ 22.9, 25.9, 33.6,
39.8, 41.7, 89.0, 93.6, 123.9, 128.6, 134.9, 164.3; IR (KBr) 3381,
2950, 1789, 1732, 1379, 1188, 993, 878, 699, 520 cm-1. Anal.
Calcd for C15H15NO5: C, 62.28; H, 5.23; N, 4.84. Found: C,
62.17; H, 5.18; N, 4.82.
Rea ction of 47 w ith Tr ip h en ylp h osp h in e. An ethanol
(30 mL) solution of 47 (289 mg, 1 mmol) and Ph3P (262 mg, 1
mmol) was placed in a three-necked flask, and the mixture
was stirred at 50 °C for 3 h under an Ar atmosphere. After
the reaction, ethanol was removed under reduced pressure to
give a white crystal, which was purified by diethyl ether (30
mL) to give triphenylphosphine oxide in 94% yield along with
alcohol 48 (83%).
1H and 13C NMR were measured at 270 and 67.5 MHz,
respectively, with tetramethylsilane as an internal standard.
Infrared (IR) spectra were measured using NaCl or KBr
pellets. A GC analysis was performed with a flame ionization
detector using a 0.2 mm × 25 m capillary column (OV-1).
Oxygen-absorption rates were measured with an isobaric gas-
absorption apparatus under a closed-flow system (2 ( 0.1 L
oxygen/h) equipped with an electrolyzer. ESR measurements
were performed on a J EOL-FE-1X (X-band) with 100-kHz field
modulation.
All starting materials and catalysts were purchased from
commercial sources and used without further treatment. The
yields of products were estimated from the peak areas based
on the internal standard technique.
Gen er a l P r oced u r e for Oxid a tion of Cycloa lk a n es. An
acetic acid (12.5 mL) solution of cycloalkane (5 mmol), NHPI
(82 mg, 10 mol %), and Co(acac)2 (7.5 mg, 0.5 mol %) was
placed in a three-necked flask equipped with a balloon filled
with O2. The mixture was stirred at 100 °C for 6 h. After
removing the solvent under reduced pressure, methanol (25
mL) and a catalytic amount of concd H2SO4 were added to the
resulting mixture and stirred at 65 °C for 15 h. The resulting
solution was extracted with diethyl ether (20 mL × 3). The
combined extracts were dried over anhydrous MgSO4. Re-
moval of solvent under reduced pressure gave a clean liquid,
which was purified by column chromatography on silica gel
(n-hexane/AcOEt ) 5/1) to give the corresponding oxygenated
products.
Ketones 2, 6, 9, 13, 16, 21, and 24, alcohols 20 and 25, and
dicarboxylic acids 3, 7, 11, and 14 were identified by comparing
the isolated products with authentic samples.
1,4-Cycloocta n ed ion e (10): 1H NMR (CDCl3) δ 1.83-1.87
(m, 4H), 2.40-2.44 (m, 4H), 2.71 (s, 4H); 13C NMR (CDCl3) δ
N-(2-H yd r oxyb icyclo[2.2.2]h ep t a n -2-yloxy)p h t h a lim -
id e (48): 1H NMR (CDCl3) δ 1.05-1.12 (m, 3H), 1.21 (d, J )
10.5 Hz, 1H), 1.46-1.63 (m, 4H), 2.05 (d, J ) 10.5 Hz, 1H),
2.29 (s, 1H), 2.66 (s, 1H), 3.91 (s, 1H), 4.00 (s, 1H), 7.75-7.87
(m, 4H), 10.65 (s, 1H); 13C NMR (CDCl3) δ 24.0, 24.9, 32.7,
41.3, 43.2, 75.8, 93.2, 123.7, 128.6, 134.7, 164.0; IR (KBr) 3451,
(20) The oxidation of alkylbenzenes21 and cyclohexane22 by Co(III)
ion is known to involve one-electron transfer from substrates to Co-
(III), yielding Co(II) ion and radical cations.
(21) Heiba, E. I.; Dessau, R. M.; Koehl, W. J . J r. J . Am. Chem. Soc.
1969, 91, 6830.
(22) Onopchenko, A.; Schulz, J . G. D. J . Org. Chem. 1973, 38, 3729.
2964, 1783, 1730, 1379, 1186, 999, 878, 782, 703, 518 cm-1
.

4526 J . Org. Chem., Vol. 61, No. 14, 1996
Ishii et al.
These values were consistent with those reported in the
literature.23
mixture was stirred at 80 °C for 3 h. The workup was
performed using the same method as previously described.
P r ep a r a tion of Com p lexes 49. A mixture of NHPI (294
mg, 2 mmol) and Co(acac)2 (235 mg, 0.8 mmol) in acetic acid
(15 mL) was stirred at 80 °C under an oxygen atmosphere.
After 0.5 h, the reaction mixture was evaporated, and an
orange solid was obtained. The resulting solid was washed
using acetonitrile, and then the complexes 49 were obtained
(271 mg).
Oxid a tion of 1 a n d 32 Ca ta lyzed by Com p lexes 49. To
a stirred solution of complex (49) (21 mg, 4 wt %) in acetic
acid (10 mL) was added 1 or 32 (5 mmol), and the reaction
mixture was fitted with a balloon filled with oxygen. The
Ack n ow led gm en t. This work was supported by a
Grant-in-Aid for Scientific Research (No. 06453143)
from the Ministry of Education, Science and Culture,
J apan, and J apan private University foundation.
Su p p or tin g In for m a tion Ava ila ble: Copies of spectra of
compounds 10, 17, 18, 22, 35, 43, 47, and 48 (24 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
(23) Ozaki, S.; Hamaguchi, T.; Tsuchida, K.; Kimata, Y. J . Chem.
Soc., Perkin Trans 2 1989, 951.
J O951970L