Efficient catalytic methods for the Baeyer-Villiger oxidation and epoxidation with hydrogen peroxide

Article abstract of DOI:10.1016/S0040-4039(01)00141-1

We have found that the Baeyer-Villiger oxidation of cyclohexanone with hydrogen peroxide in 1,1,1,3,3,3-hexafluoro-2-propanol as solvent is catalyzed by Br?nsted acids: in the presence of 1 mol% of p-TsOH, transformation to ε-caprolactone occurs rapidly and under mild conditions. Epoxidation of 1-octene can be performed just as efficiently in the presence of 1 mol% of benzenearsonic acid. Since Br?nsted acids do not effect epoxidation under these conditions, the formation of perarsonic acid appears to be the crucial feature of the latter catalyst.

Full text of DOI:10.1016/S0040-4039(01)00141-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 2293–2295
Efficient catalytic methods for the Baeyer–Villiger oxidation and
epoxidation with hydrogen peroxide
Albrecht Berkessel* and Marc R. M. Andreae
Institut fu¨r Organische Chemie der Universita¨t zu Ko¨ln, Greinstraße 4, D-50939 Cologne, Germany
Received 19 September 2000; revised 16 December 2000; accepted 24 January 2001
Abstract—We have found that the Baeyer–Villiger oxidation of cyclohexanone with hydrogen peroxide in 1,1,1,3,3,3-hexafluoro-2-
propanol as solvent is catalyzed by Brønsted acids: in the presence of 1 mol% of p-TsOH, transformation to o-caprolactone occurs
rapidly and under mild conditions. Epoxidation of 1-octene can be performed just as efficiently in the presence of 1 mol% of
benzenearsonic acid. Since Brønsted acids do not effect epoxidation under these conditions, the formation of perarsonic acid
appears to be the crucial feature of the latter catalyst. © 2001 Elsevier Science Ltd. All rights reserved.
Hydrogen peroxide is a highly attractive oxidant for a
number of reasons: it is a cheap, mild and an environ-
mentally benign reagent with a high content of ‘active’
oxygen, and water is formed as the only by-product.
However, catalysts are normally required to effect oxy-
gen transfer from hydrogen peroxide to the substrate.
From a synthetic point of view, the most attractive
oxygenations of organic compounds are the epoxida-
tion of olefins and the Baeyer–Villiger oxidation of
ketones. Particularly with regard to the epoxidation of
olefins, much effort has been spent on the development
of metal containing catalysts. Without going into detail,
complexes of Mo, W, Re and Mn have proven particu-
larly useful for the latter purpose.1–5 Furthermore,
arsonic acids were shown to catalyze the epoxidation of
olefins such as 1-octene with hydrogen peroxide.6 With
regard to the Baeyer–Villiger oxidation of ketones, the
use of hydrogen peroxide has mainly been restricted to
strained substrates such as cyclobutanones.7 For unacti-
vated ketones like cyclohexanone, it was shown that
again arsonic acids are capable of catalyzing lactone
formation.8 Furthermore, certain Pt complexes effect
the Baeyer–Villiger oxidation of unstrained ketones
with hydrogen peroxide.9
thermore, Neimann and Neumann recently reported
that in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), the
epoxidation of olefins and the Baeyer–Villiger oxidation
of ketones proceed efficiently in the absence of cata-
lysts.12 Their report prompts us to disclose our own
results in this area which can be summarized as follows:
(i) the Baeyer–Villiger oxidation of e.g. cyclohexanone
in HFIP is an acid-catalyzed process. Quite remarkable
acceleration can be effected by addition of as little as 1
mol% of Brønsted acids, such as p-toluenesulfonic acid.
(ii) the epoxidation of e.g. 1-octene is not significantly
accelerated by simple Brønsted acids. However, clean
and rapid epoxide formation is effected by addition of
1 mol% of arsonic acids, such as benzenearsonic acid.
Epoxidation: In a typical run, 144 mg (1.28 mmol) of
1-octene were mixed with 2 ml of HFIP in a 10 ml flask
equipped with a magnetic stir bar and a reflux con-
denser. Then, 12.8 mmol of the acid catalyst (1 mol%
rel. to the substrate olefin) was added. No precautions
were taken to exclude air or moisture. The reaction was
started by the addition of 94 ml (1.66 mmol, 1.3 equiv.)
of 50% aqueous hydrogen peroxide.13 The reaction
mixture was then heated to 60°C and stirred continu-
ously. The course of the epoxidation was followed by
capillary GC.14 The results for various acid additives
are summarized in Table 1.
Very recently, fluorinated alcohols have attracted con-
siderable attention as solvents for epoxidations with
hydrogen peroxide. For example, it was shown by
Sheldon et al. that the epoxidation of a number of
olefins, effected by HReO4 and Ph2AsMe as co-catalyst,
works best in 2,2,2-trifluoroethanol as solvent.10,11 Fur-
Inspection of Table 1 reveals that the arsonic acid is a
highly efficient catalyst in HFIP as solvent (entries 1
and 2). Neither one of the other acids effects any
acceleration beyond background (entries 1 and 3–6).
Furthermore, our catalytic system (Ph-AsO3H2/HFIP)
is far more effective than the combination of benzene-
* Corresponding author. E-mail: berkessel@uni-koeln.de
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00141-1

2294
A. Berkessel, M. R. M. Andreae / Tetrahedron Letters 42 (2001) 2293–2295
Table 1. Catalytic epoxidation of 1-octene with hydrogen peroxide
Entry
Solvent
Catalyst
Reaction time (h)
Conversion of 1-octenea (%)
Epoxide yielda (%)
1
2
3
4
5
6
7
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
Dioxane
None
4.5
4.5
4.5
4.5
4.5
4.5
4.5
9
9
95b,c
7
7
8
Ph-AsO3H2
Ph-PO3H2
HCl
CF3CO2H
p-TsOH
Ph-AsO3H2
Quant.
7
8
9
8
7
B1
B1
a GC: HP 5890 Series II Plus; 50 m×0.2 mm HP-1
b Up to full conversion of the olefin, no by-products such as 1,2-dihydroxyoctane were observed by GC.
c In a small-scale preparative run, 89% of pure 1,2-epoxyoctane was isolated by distillation (see Ref. 17).
arsonic acid and dioxane as solvent, originally des-
cribed by Jacobson et al. (entries 2 and 7).6 Although
not yet strictly proven, it appears reasonable to assume
that the formation of perarsonic acid may account for
the effectiveness of Ph-AsO3H2. It is not clear at the
moment by which mechanism HFIP potentiates the
activity of this catalyst compared to dioxane as the
solvent (entries 2 and 7). Under identical conditions,
1-dodecene was also smoothly and quantitatively epoxi-
dized.
The result shown in Table 2, entry 2 leaves open the
question of whether the catalytic activity of benzenear-
sonic acid is based on peracid formation or whether it
just acts as a general acid catalyst. On the other hand,
the activity of HCl (entry 4) indicates that an alterna-
tive non-peracid, but proton-catalyzed process exists
that also leads to lactone formation. Since neither
Ph-PO3H2, CF3CO2H or p-TsOH are active in the
epoxidation of 1-octene (Table 1, entries 3, 5 and 6), it
is reasonable to assume that they catalyze the Baeyer–
Villiger oxidation by acid catalysis. It may be specu-
lated that an acid-catalyzed rearrangement of an
initially formed peroxyacetal is involved. This assump-
tion is consistent with the known ionizing power of
HFIP and will be the subject of further studies.16
Baeyer–Villiger oxidation: In just the same way as
described above for the epoxidation of 1-octene, cyclo-
hexanone was treated with 50% hydrogen peroxide in
the presence of various additives.13 The results are
summarized in Table 2. Inspection of Table 2 reveals
that in the absence of a catalyst, basically no oxygena-
tion of the ketone was observed (entry 1). Furthermore,
the known activity of benzenearsonic acid in the
Baeyer–Villiger oxidation of ketones is again remark-
ably increased in HFIP as solvent (entries 2 and 7).8
Exchanging Ph-AsO3H2 for the corresponding phos-
phonic acid led to the interesting observation that the
latter material is even more active than the arsonic acid
(entries 2 and 3). We rapidly realized that, unlike the
epoxidation described above, even simple Brønsted
acids such as HCl or TFA are catalytically much more
active than benzenearsonic acid (entries 2, 4 and 5). In
our hands, 1 mol% of p-toluenesulfonic acid proved
best (entry 6): under these conditions, cyclohexanone is
converted to o-caprolactone in good yield in less than
one hour.15
In summary, we have presented two simple but highly
efficient protocols for catalytic epoxidation and
Baeyer–Villiger oxidation with hydrogen peroxide. We
demonstrated the efficiency of the methods on two
notoriously difficult-to-oxidize substrates, i.e. 1-octene
and cyclohexanone. We believe that our method has
great potential for synthetic application: firstly, 50%
hydrogen peroxide is employed which is not only safe
to handle, but also represents a very economic form of
this oxidant. The catalysts used, i.e. p-toluenesulfonic
acid and benzenearsonic acid, are cheap and commer-
cially available materials. In the course of the distilla-
tive work-up procedure, the fluorinated solvent can be
recovered basically quantitatively.17 Furthermore, the
reaction times are short and the oxygenation products
Table 2. Catalytic Baeyer–Villiger oxidation of cyclohexanone with hydrogen peroxide
Entry
Solvent
Catalyst
Reaction time (min)
Conversion of cyclohexanonea (%)
Lactone yielda (%)
1
2
3
4
5
6
7
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
Dioxane
None
240
240
75
65
65
B1
Quant.
86
90
95
B1
85b
Ph-AsO3H2
Ph-PO3H2
HCl
CF3CO2H
p-TsOH
Ph-AsO3H2
84b
85b
90b
40
Quant
92b,c
B1d
240c
B1d
a GC: HP 5890 Series II Plus; 50 m×0.2 mm HP-1.
b Up to full conversion of the ketone, no by-products were observed by GC.
c In a small-scale preparative run, 73% of o-caprolactone was isolated by distillation (see Ref. 17).
d After 24 h, 4% of o-caprolactone was formed.

A. Berkessel, M. R. M. Andreae / Tetrahedron Letters 42 (2001) 2293–2295
2295
are formed in good yields and purities. Finally, the
current procedure for the epoxidation of olefins is most
suitable for hydrophobic substrate, e.g. long-chain ter-
minal olefins. It is thus complementary to our highly
efficient epoxidation method using MnTMTACN/
ascorbate as catalyst.5 For the latter oxidation, more
hydrophilic and in particular electron deficient olefins
such as acrylates are the best substrates.5
10. van Vliet, M. C. A.; Arends, I. W. C. E.; Sheldon, R. A.
Tetrahedron Lett. 1999, 40, 5239–5242.
11. Van Vliet, M. C. A.; Arends, I. W. C. E.; Sheldon, R. A.
J. Chem. Soc., Perkin Trans. 1 2000, 377–380.
12. Neimann, K.; Neumann, R. Org. Lett. 2000, 2, 2861–
2863.
13. In all experiments, non-stabilized 50% aqueous hydrogen
peroxide was used. HFIP was purchased from Aldrich or
ABCR, both materials gave the same experimental
results. Benzenearsonic acid was purchased from
Avocado. All reagents were used without further
purification.
Acknowledgements
14. Samples (100 ml) were withdrawn from the reaction flask
and dichloromethane (1 ml) was added. This mixture was
filtered through a plug of neutral alumina/MnO2 and
analyzed by GC.
This work was supported by the BASF AG and by the
Fonds der Chemischen Industrie.
15. Similarly, 2-methyl- and 2-phenylcyclohexanone as well
as cyclopentanone were cleanly converted to the corre-
sponding lactones.
16. Eberson, L.; Hartshorn, M. P.; Persson, O.; Radner, F. J.
Chem. Soc., Chem. Commun. 1996, 2105–2112.
References
1. Wahl, G.; Kleinhenz, D.; Schorm, A.; Sundermeyer, J.;
Stowasser, R.; Rummey, C.; Bringmann, G.; Fickert, C.;
Kiefer, W. Chem. Eur. J. 1999, 5, 3237–3251.
2. (a) Dickman, M. H.; Pope, M. T. Chem. Rev. 1994, 94,
569–584; (b) Roma˜o, C. C.; Ku¨hn, F. E.; Herrmann, W.
A. Chem. Rev. 1997, 97, 3197–3246.
3. (a) Sato, K.; Aoki, M.; Ogawa, M.; Hashimoto, T.;
Noyori, R. J. Org. Chem. 1996, 61, 8310–8311; (b) Sato,
K.; Aoki, M.; Ogawa, M.; Hashimoto, T.; Panyella, D.;
Noyori, R. Bull. Chem. Soc. Jpn. 1997, 70, 905–915.
4. De Vos, D. E.; Sels, B. F.; Reynaers, M.; Subba Rao, Y.
V.; Jacobs, P. A. Tetrahedron Lett. 1998, 39, 3221–3224.
5. Berkessel, A.; Sklorz, C. A. Tetrahedron Lett. 1999, 40,
7965–7968.
6. Jacobson, S. E.; Mares, F.; Zambri, P. M. J. Am. Chem.
Soc. 1979, 101, 6946–6950.
7. Bogdanowicz, M. J.; Ambelang, T.; Trost, B. M. Tetra-
hedron Lett. 1973, 14, 923–926.
8. Jacobson, S. E.; Mares, F.; Zambri, P. M. J. Am. Chem.
Soc. 1979, 101, 6938–6946.
17. Small-scale preparative Baeyer–Villiger oxidation of
cyclohexanone and epoxidation of 1-octene: 19.3 mmol of
cyclohexanone (1.89 g) [1-octene (2.17 g)] were mixed
with 24 ml HFIP, and 1 mol% of p-toluenesulfonic acid
monohydrate (36.7 mg) [benzenearsonic acid (39.0 mg)]
was added. Hydrogen peroxide (50%, 1.4 ml, 24.7 mmol,
1.3 equiv.) was added in portions of ca. 100 ml with
cooling (CAUTION: significant evolution of heat in the
case of the Baeyer–Villiger oxidation). The reaction mix-
ture was stirred at 60°C. Monitoring by GC indicated
complete conversion after ca. 1 h in the case of the
Baeyer–Villiger reaction, whereas the epoxidation
required heating to 60°C for ca. 2–3 h. After removal of
the solvent [22 ml (92%) recovered], the crude product
was taken up in dichloromethane, washed with aq.
NaHCO3, and stirred over anhydrous Na2SO4 and
MnO2. Removal of the solvent and fractionating vacuum
distillation afforded 1.61 g (73%) of o-caprolactone and
2.20 g (89%) of 1,2-epoxyoctane, respectively.
9. Strukul, G. Angew. Chem. 1998, 110, 1256–1267; Angew.
Chem., Int. Ed. Engl. 1998, 37, 1198–1209.
.
.