A new catalytic system for the selective aerobic oxidation of large ring cycloalkanes to ketones

Article abstract of DOI:10.1021/op0340199

The combination of cobalt with N-hydroxysaccharin proved to be an effective catalyst for the aerobic oxidation of large ring cycloalkanes to the corresponding ketones.

Full text of DOI:10.1021/op0340199

Organic Process Research & Development 2003, 7, 426−428
A New Catalytic System for the Selective Aerobic Oxidation of Large Ring
Cycloalkanes to Ketones
Xavier Baucherel,† Isabel W. C. E. Arends,† S. Ellwood,‡ and Roger A. Sheldon*,†
Laboratory for Biocatalysis and Organic Chemistry, Delft UniVersity of Technology, Julianalaan 136,
2628 BL Delft, The Netherlands, and Quest International, Ashford, Kent TN24 OLT, United Kingdom
Scheme 1
Abstract:
The combination of cobalt with N-hydroxysaccharin proved to
be an effective catalyst for the aerobic oxidation of large ring
cycloalkanes to the corresponding ketones.
Introduction
The oxidation of saturated hydrocarbons with molecular
oxygen is a reaction of considerable industrial importance.1
The selective transformation of large-ring cycloalkanes (e.g.,
cyclododecane 3) to the corresponding ketones is a particu-
larly important reaction since the oxidation products are
intermediates for the fragrances industry and for the produc-
tion of dicarboxylic acid precursors for polyamides and
polyesters.2 However, aerobic oxidations of (cyclic)alkanes
usually proceed in low selectivities, owing to overoxidation
of the ketone to complex mixtures of products. To obtain
reasonable selectivities the reaction is usually performed
according to the Bashkirov method.2,3 This involves aerobic
oxidation in the presence of stoichiometric amounts of B2O3
to give the borate ester of cyclododecanol as the major
product (80%) along with 10% of cyclododecanone at 30%
conversion. The borate ester is subsequently hydrolysed to
the alcohol and boric acid. A shortcoming of this method is
that it is circuitous, involving three stepssoxidation, hy-
drolysis, and subsequent dehydrogenation of the alcohol to
the ketonesand recycling of large quantities of boric acid.
Ishii and co-workers4 discovered that N-hydroxyphthalimide
(NHPI) 1 in combination with cobalt catalyses the autoxi-
dation of hydrocarbons under mild conditions (25-100 °C,
O2 1 atm). The promoting effect of NHPI was explained on
the basis of the mechanism shown in Scheme 1. NHPI is
converted into its corresponding phthalimide N-oxyl (PINO)
radical which is able to abstract a hydrogen atom from the
organic substrate, thus propagating the autoxidation chain.
In this way PINO is the actual chain carrier, which leads to
longer propagation chains and, hence, to higher rates and
selectivities compared to those from standard autoxidation.
The introduction of electron-withdrawing groups in the
aryl ring of NHPI was shown to have a beneficial effect on
the catalyst performance for the aerobic oxidation of alkyl-
benzenes5 and the electrocatalytic oxidation of alcohols.6 We
reasoned that the use of N-hydroxysaccharin7 (NHS) 2, in
which one carbonyl group (CO) is replaced by the more
electron-withdrawing sulfonyl (SO2) group, could provide
an even more effective promoter. This proved to be the case,
and we report herein our results on the aerobic oxidation of
large-ring cycloalkanes with metal catalysts in combination
with NHS (reaction 1).
* Author for correspondence. Fax: +31 15 2781415. Telephone: +31 15
2782675. E-mail: R.A.Sheldon@tnw.tudelft.nl.
† Laboratory for Biocatalysis and Organic Chemistry, Delft University of
Technology.
‡ Quest International.
(1) Sheldon, R. A.; Kochi, J. K. Metal-Catalysed Oxidations of Organic
Compounds; Academic Press: New York, 1981. Hill, C. L. ActiVation and
Functionalization of Alkanes; Academic Press: New York, 1989. Fisher,
W. B.; Van Peppen, J. P. Kirk-Othmer Encyclopaedia of Chemical
Technology, 4th ed.; Wiley: New York, 1996; Vol. 7, pp 851-859.
(2) Rademacher, H. Ullmann’s Encyclopedia Industrial Organic Chemical;
Wiley-VCH: Weinheim, 1999; Vol. 3, pp 1783-1787.
(3) Bashkirov, A. N.; Kamzolkin, V. V.; Sokova, K. M.; Andreyeva, T. P.
Oxidation of Hydrocarbons in Liquid Phase; Pergamon: Oxford, 1965; p
183.
(4) Ishii, Y.; Iwahama, T.; Sakaguchi, S.; Nakayama, K.; Takeno, M.;
Nishiyama, Y. J. Org. Chem. 1996, 61, 4520-4526. Ishii, Y.; Sakaguchi,
S. Catalysis SurVeys from Japan 1999, 3, 27-35. Ishii, Y. T. Eur. Pat., EP
0 824 962 A1. Ishii, Y.; Sakaguchi, S.; Iwahama, T. AdV. Synth. Catal.
2001, 343, 393-427.
Results and Discussion
Under the standard conditions described by Ishii et al.4
(10 mol % NHPI, 0.5 mol % Co(acac)2 in acetic acid at 100
(5) Wentzel, B. B.; Donners, M. P. J.; Alsters, P. L.; Feiters, M. C.; Nolte, R.
J. M. Tetrahedron 2000, 56, 7797-7803.
(6) Gorgy, K.; Lepretre, J.-C.; Saint-Aman, E.; Einhorn, C.; Einhorn, J.;
Marcadal, C.; Pierre, J.-L. Electrochim. Acta 1998, 44, 385-393.
(7) Nagasawa, H. T.; Kawle, S. P.; Elberling, J. A.; DeMaster, E. G.; Fukuto,
J. M. J. Med. Chem. 1995, 38, 1865-1871.
426
•
Vol. 7, No. 3, 2003 / Organic Process Research & Development
10.1021/op0340199 CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/04/2003

a
Table 1. Oxidation of cyclododecane 3 in AcOHa
sel. (%)b
Table 2. Oxidation of cyclododecane 3 in PhCF3
sel. (%)b
time conv.
time conv.
run
catalyst
T (°C) (h)
(%)
4
5
6
total
run
catalyst
T (°C) (h) (%)
4
5
6
total
1
2
3
4
5
Co(acac)2
Co(acac)2/NHS
Co(acac)2/NHPI 100
100
100
24
6
6
8
8
24
24
<4
0
0
8
5
0
0
1
2
3
4
5
6
7
Bu4NBr/NHS
Co(acac)2/NHS
Co(acac)2/NHPI 100
Co(acac)3/NHS 100
Co(acac)3/NHPI 100
Co(acac)3/NHS
Co(acac)3/NHPI
80
100
24
24
24
4
8
10
24
9
41 33 n.d. 74
64 31
58 29
16 55
30 64
32 54 20
43 48 8
35 64 17
23 63 21
24 72 18
23 42 16
0
6
5
5
0
5
74
62
86
89
90
63
Co(acac)2/NHS
Co(acac)2/NHPI
75
75
50
50
47 41 10 12 63
36 42 23 73
42 47 14 20 81
8
6c Co(acac)2/NHS
7c Co(acac)2/NHPI
80
80
0
-
-
-
0
a Cyclododecane 3 (3 mmol), NHS or NHPI (0.3 mmol), cocatalyst (0.015
mmol), AcOH (7.5 cm3), O2 1 atm. b Based on cyclododecane 3 reacted. c 5
cm3 AcOH.
a Cyclododecane (3 mmol), NHS or NHPI (0.3 mmol), cocatalyst (0.015
mmol), PhCF3 (9 cm3), O2 1 atm. b Based on cyclododecane 3 reacted.
Table 3. Oxidation of different cycloalkanes in PhCF3
catalysed by NHSa
°C and 1 bar O2) cyclododecane 3 afforded a mixture of
cyclododecanone 4 (29%), cyclododecanol 5 (5%), and 1,12-
dodecanedioic acid 6 (30%) at 58% conversion (Table 1,
entry 3). In addition to 6, other major byproducts were
identified by GC-MS analysis, as two isomeric hydroxy-
cyclododecanones and two isomeric cyclododecadienones.
The formation of these byproducts is attributed to competing
transannular hydrogen abstraction by an intermediate cy-
clododecylperoxy radical, analogous to that observed in the
autoxidation of cyclohexylbenzene.8 The products reported
in the tables plus the isomeric hydroxycyclododecanones and
cyclododecadienones accounted for more than 90% of the
substrate converted. Identification of the remaining (<10%)
byproducts is part of ongoing optimization studies and will
be reported in due course.
Substitution of NHPI by NHS (entry 2) afforded a slightly
higher conversion (64%) and a slightly higher selectivity to
ketone and alcohol (31 and 8%, respectively). Higher
selectivities were obtained by decreasing the reaction tem-
perature to 75 °C (runs 4 and 5) whereby the higher activity
of NHS became more pronounced. When the reaction
temperature was decreased further to 50 °C, no reaction was
observed with NHPI/Co(acac)2. In contrast NHS/Co(acac)2
gave 42% cyclododecane conversion and a selectivity to
ketone and alcohol of 47% and 14%.
We next turned our attention to the use of R,R,R-tri-
fluorotoluene as solvent. PhCF3 was used by Ishii9 as a sol-
vent for the metal-free autoxidation of adamantane catalysed
by NHPI in the presence of tetrabutylammonium bromide.
Oxidation of cyclododecane under similar conditions in the
presence of NHS (Table 2) led to a low conversion (9%)
with good overall selectivity (entry 1). Replacement of the
tetrabutylammonium salt by Co(acac)2 led to an increase in
rate and, with NHS, an increase in selectivity to ketone +
alcohol (74% at 32% conversion, entry 2).
The reaction rate and selectivity were further increased
by replacing Co(acac)2 with Co(acac)3 (35% conversion and
81% selectivity after 4 h, entry 4). The highest selectivity
(90% to a 4:1 mixture of ketone/alcohol at 24% conversion)
was observed with a combination of NHS and Co(acac)3 at
sel. (%)b
time conv
run
substrate
(h)
(%) ketone alcohol diacid
1
2
3
cyclododecane 10
cyclodecane
cyclooctane
24
31
30
72
61
55
18
24
22
0
2
20
5
1.5
a Cycloalkanes (3 mmol), Co(acac)3 (0.015 mmol), NHS (0.3 mmol), PhCF3
(9 cm3), 80 °C, O2 1 atm. b Based on substrate reacted.
80 °C in PhCF3. Under these conditions no formation of 1,12-
dodecanedioic acid was observed.
In all these experiments, the selectivity (for alcohol +
ketone) decreases with increasing conversion. At higher
conversions further oxidation of the ketone product starts to
become important, and we have demonstrated this in
competition experiments.10 The conditions developed for the
oxidation of cyclododecane in R,R,R-trifluorotoluene were
applied to the oxidation of different large-ring cycloalkanes,
and the results are presented in Table 3.
As a result of ring strain caused by transannular interac-
tion, cyclodecane and cyclooctane are more reactive under
autoxidation conditions than cyclododecane. In the oxidation
of cyclodecane, for example, 31% conversion was obtained
in 5 h to give cyclodecanone in 61% selectivity along with
24% cyclodecanol and only 2% 1,10-decanedioic acid. The
oxidation of cycloctane was even faster (30% conversion in
1.5 h) Cyclooctanone, cyclooctanol, and 1,8-octanedioic acid
were formed in 55, 22, and 20% selectivity, respectively.
Conclusions
Summarising, we have shown that the combination of
N-hydroxysaccharin with Co(acac)3 is an efficient catalyst
for the aerobic oxidation of large-ring cycloalkanes, display-
ing a higher activity than the previously described NHPI-
based catalyst. The selectivity toward ketone + alcohol (90%
of a 4:1 ketone alcohol mixture at 24% conversion) is as
(10) We have performed competition experiments where we added cyclopen-
tadecanone (0.42 mmol) after 1 h to a typical reaction of cyclododecane
in PhCF3 at 100 °C with NHS (as in run 4, Table 2). In this case 12% of
the added ketone was decomposed after 6 h into a variety of products. A
similar procedure was used to test the alcohol stability under these conditions
using a mixture of cyclopentadecanol and cyclododecane. It was demon-
strated that the alcohol is converted for almost 85% to the ketone and 9%
to the diacid.
(8) Arends, I. W. C. E.; Sasidharan, M.; Ku¨hnle, A.; Duda, M.; Jost, C.; Sheldon,
R. A. Tetrahedron 2002, 58, 9055-9061.
(9) Matsunaka, K.; Iwahama, T.; Sakaguchi, S.; Ishii, Y. Tetrahedron Lett. 1999,
40, 2165-2168.
Vol. 7, No. 3, 2003 / Organic Process Research & Development
•
427

good as or better than that observed with the Bashkirov
method,2 and the procedure is considerably less circuitous.
Co(acac)3, and 200 mg of 1,2,4-trichlorobenzene were
weighed in the flask before addition of 7.5 or 9 mL of acetic
acid. A series of vacuum/molecular oxygen were applied
before the reaction was started. The reaction was monitored
by gas chromatography using a CP-WAX 52 CB column
(50 m × 0.53 mm). A typical temperature program starts at
50 °C (5 min) with a ramp of 8°/min up to 220 °C. The
reported yields are thus GC yields, and are determined using
1,2,4-trichlorobenzene as the internal standard. 1,12-Dode-
canedioic acid, 1,10-decanedioic acid, and 1,8-octanedioic
acid were analysed by HPLC (Waters Symmetry C18
reversed phase column; eluent, water/methanol 50:50, 0.1%
TFA) using octanoic acid as an internal standard.
Experimental Section
N-Hydroxysaccharin (NHS) was synthesised according to
a seven-step synthesis starting from 2-sulfobenzoic anhy-
dride, published by Nagasawa et al.7 in 1995. The overall
yield was 17%. Numerous attempts to synthesise NHS via a
different route failed. Direct reaction of the 2-sulfobenzoic
anhydride with hydroxylamine in the presence of base,
analogous to the synthesis of N-hydroxyphthalimide deriva-
tives,5 gave no NHS. Also direct oxidation of saccharin with
electrophilic oxidants such as H2O2 (basic conditions), H2O2/
WO42-, KHSO5 (oxone), dimethyloxirane, or methyl(trifuo-
romethyl)dioxirane failed.
Acknowledgment
In a typical experiment, the catalytic run was carried out
in a two-necked 25-cm3 round-bottom flask containing a
magnetic stirrer, equipped with a condenser and connected
to a gas buret filled with molecular oxygen. In a typical
experiment 504 mg (3 mmol) of cyclododecane, 60 mg (0.3
mmol) of N-hydroxysaccharin, 5.3 mg (0.015 mmol) of
The project was financed by ICI, and the contribution is
kindly acknowledged.
Received for review February 4, 2003.
OP0340199
428
•
Vol. 7, No. 3, 2003 / Organic Process Research & Development