Reaction of epoxyketones with hydrogen peroxide - Ethane-1,1- dihydroperoxide as a surprisingly stable product

Article abstract of DOI:10.1002/chem.200800932

Reaction of epoxyketones with hydrogen peroxide, ethane-1,1-dihydroperoxide as a stable product was reported. Triacetone triperoxide and methylhydroperoxide were reported as highly explosive compounds. Thermogravimetric investigations showed decomposition in the temperature range 60-130°C with the highest decomposition rate at about 105°C. Using differently substituted epoxyketones as reactants, it was able to isolate and characterize propane-1,1-dihydroperoxide. Epoxyketones can also be attacked by H₂O₂ at the carbonyl C atom and at both epoxy C atoms as electrophilic centers. Initial results revealed that the acid-catalyzed reaction of 5- and 7-ring homologues 1 (n=0,2) with H₂O₂ runs similarly. They also show a lower tendency to form the geminal dihydroperoxide.

Full text of DOI:10.1002/chem.200800932

COMMUNICATION
DOI: 10.1002/chem.200800932
Reaction of Epoxyketones with Hydrogen Peroxide—Ethane-1,1-
dihydroperoxide as a Surprisingly Stable Product
Hans-Jürgen Hamann,* Alexander Bunge, and Jürgen Liebscher*[a]
Oxidation of organic compounds with hydrogen peroxide
has attracted attention for a long time and has been attribut-
ed to green chemistry. We used hydrogen peroxide in reac-
tions with cyclic alcohols and observed unusual rearrange-
ment reactions under ring enlargement giving new hydroper-
oxides and peroxides.[1–3] In this way we also obtained ali-
phatic primary geminal dihydroperoxides.[2] We report here
on reactions of cyclic epoxyketones with hydrogen peroxide.
Such epoxides are usually synthesised by the Weitz–Scheffer
reaction, that is, by reaction of a,b-unsaturated ketones with
H2O2 under basic conditions.[4] We treated such products
with 70% H2O2 in the presence of catalytic amounts of cam-
phor sulfonic acid at room temperature.
1,1-dihydroperoxide (2a) exhib-
its a signal at 107.9 ppm in the
13C NMR spectrum and one
typical hydroperoxide-H signal
(intensity two protons) at
1
9.71 ppm in the H NMR spectrum. In combination with the
HR-ESIMS result, the structure of 2a could be proven un-
ambiguously. The geminal dihydroperoxide 2a has the high-
est content of peroxide oxygen (68%) relative to hitherto
known organic peroxides and hydroperoxides.
The so-called triacetone tri-
peroxide (6; 43% peroxidic
oxygen) and methylhydroper-
Substituted 6-ring epoxyketones 1 at the 3-position gave
mixtures of products, interestingly consisting of geminal di-
hydroperoxides 2, dicarboxylic acids 3, carboxylic acids 4
and ketocarboxylic acids 5 (Table 1, Scheme 1).
oxide (7;[8] 66.7%) were report-
ed as highly explosive com-
pounds. Despite the very high
oxygen content compound 2a is
remarkably stable. It can be
kept at room temperature for
Table 1. Distribution of products in the reaction of epoxy ketones 1 with
several days and at À208C for several weeks without de-
composition. Thermogravimetric investigations (see Sup-
porting Information) showed decomposition in the tempera-
ture range 60–1308C with the highest decomposition rate at
about 1058C. Using differently substituted epoxyketones as
reactants we were also able to isolate and characterise pro-
pane-1,1-dihydroperoxide (2b), which is not yet described in
literature, and to obtain benzylidenedihydroperoxide (2c).
The latter was previously described by using a different
method from the reaction of benzaldehyde with H2O2.[5] Re-
markably, known syntheses of geminal dihydroperoxides
based on reactions of carbonyl compounds with H2O2 are re-
H2O2.
Epoxyketone
n
R
2/[%][a]
3/[%][b]
3a/38
4 [%][b]
4a/29
5 [%][b]
5a/57
1a
1b
1c
1d
1e
0
1
1
1
2
Me
Me
Et
Ph
Me
2a/2
2a/12
2b/19
2c/19
2a/4
3b/53
4a/6
5b/15
[c]
[c]
[c]
3b/18
3c/66
–
5c/35
–
4a/20
[a] Yields of isolated products. [b] Yields determined by 1H NMR spec-
troscopy. [c] The aqueous phase consisted of a complex mixture and
could not be analysed.
The geminal dihydroperoxides 2 show significant signals
in the NMR spectra.[2,5–7] Thus, the hitherto unkown ethane-
[a] Dr. H.-J. Hamann, Dipl.-Chem. A. Bunge, Prof. J. Liebscher
Institute of Chemistry, Humboldt-University Berlin
Brook-Taylor-Str. 2, 12489 Berlin (Germany)
Fax : (+49)30-2093-7552
E-mail: Liebscher@chemie.hu-berlin.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800932.
Scheme 1. Acid-catalysed reaction of epoxy ketones 1 with H2O2
Chem. Eur. J. 2008, 14, 6849 – 6851
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6849

stricted mainly to ketones and aromatic aldehydes.[5,7,9] With
the exception of the special case of chloral[10] and one insuf-
ficiently described reaction of pentanal,[11] primary aliphatic
geminal dihydroperoxides have only been synthesised by
ozonolysis of enol ethers in the presence of excess of
H2O2.[12] Iskra et al. reported that aliphatic aldehydes form
1-hydroperoxy-1-hydroxyalkanes, that is, hemiacetal struc-
tures, in reactions with H2O2 in acetonitrile in the presence
of iodine.[7] We checked if by the application of more highly
concentrated H2O2 in this reaction with acetaldehyde wheth-
er the formation of the geminal dihydroperoxide could be
achieved; however, we could not detect 2a, even when using
70% H2O2.
droperoxy-3-hydroxyketone, which could then by protona-
tion and water elimination react in a Hock type reaction
under ring expansion to give intermediate 9. According to
pathway A, the R substitutent migrates to give ketolactone
10, which affords O,O-diacylacetal 11 by a further Baeyer–
Villiger oxidation. The latter is ring-opened by H2O2 afford-
ing dihydroperoxide 2 and glutaric acid 3b. The alternative
reaction pathways B and C do not provide geminal dihydro-
peroxides 2. Following pathway B the ring C atom of 9 mi-
grates. The resulting 5-acyl-d-lactone 12 undergoes
a
Baeyer–Villiger rearrangement by migration of the C5 atom
of the lactone ring, which again has a high migratory apti-
tude because of the ring O atom. Cleavage of the acetal in
the resulting intermediate 13 by H2O2 or water leads to the
formation of the carboxylic acid 4, while the remaining frag-
ment 14 is oxidised to glutaric acid. Finally, the formation of
ketocarboxylic acids 5 according to pathway C can be real-
ised by the addition of H2O2 to the cation 9 followed by a
Hock type hydroperoxide rearrangement.[1] The ring en-
largement occurs by migration of an O-atom-stabilised car-
benium ion. The resulting orthoester 16 is unstable and de-
composes into ketoacid 5 and formic acid, which is further
oxidised to carbon dioxide.
As can be seen by the product spectrum of the different
reactants used, the carbon skeleton of 2 is evidently derived
from the substituent R and the C2 atom of the epoxyketone
À
1, with the occurrence of a number of C C bond fissions.
To rationalise the complex reaction pathway we propose
the following mechanism as shown in Scheme 2 for 6-ring
We further wondered if this unusual reaction behaviour
can also be observed with structurally related epoxyketones.
Initial results revealed that the acid-catalysed reaction of 5-
and 7-ring homologues 1 (n=0, 2) with H2O2 runs similarily
(Table 1). However, they show a lower tendency to form the
geminal dihydroperoxide 2a. We finally investigated in how
far 6-ring epoxyketones with additional alkyl substituents in
the ring exhibit a similar reaction behaviour. Piperitone
oxide 17 forms compound 2a to a similar level with the cor-
responding reaction products 4a, 18 and 19 (Scheme 3). Al-
though an isopropyl substituent with a relatively high migra-
tory aptitude is found in 17, the oxiranyl moiety still mi-
grates, as can be seen by the exclusive formation of 18 and
19.
Scheme 2. Proposed reaction mechanism of 6-ring epoxyketones with
H2O2.
epoxyketones 1 (n=1). Epoxyketones can be attacked by
H2O2 at the carbonyl C atom and/or at both epoxy C atoms
as electrophilic centers. We suppose that the primary attack
occurs at the carbonyl C atom causing a Baeyer–Villiger
type rearrangement in which the oxirane C atom migrates,
because of its higher migratory aptitude. Under acid condi-
tions the acetal moiety of the resulting lactone 8 can be
easily cleaved affording carbenium ion 9, which can react in
three different ways; Scheme 2 A, B and C. Alternatively,
hydrogen peroxide could add to the C2 atom to form a 2-hy-
Scheme 3. Reaction of piperitone oxide 17 with H2O2.
Interestingly, the hydroperoxylactone 21 and the acyllac-
tone 23 were obtained in addition to 2a[13] and 22 in the re-
action of isophorone oxide 20 with H2O2 (Scheme 4). These
types of products were not observed in the other cases.
Compound 21 (see Figure 1 for X-ray crystal analysis) is the
6850
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 6849 – 6851

Reactions of Epoxyketones with Hydrogen Peroxide
COMMUNICATION
Experimental Section
Warning! 70% H2O2 as well as peroxidic compounds are potentially ex-
plosive and should be handled with precautions.
Representative preparation of 2a: Compound 1b (923 mg, 8.38 mmol)
was dissolved in Et2O (2 mL) and treated with 70% H2O2 (2 mL) and d-
(+)-camphorsulfonic acid monohydrate (20 mg) under ice cooling. After
stirring at room temperature overnight, the reaction mixture was diluted
with water (30 mL) and extracted with Et2O (3”20 mL). The combined
organic layers were washed with satd. aq. NaHCO3 (20 mL). After drying
(Na2SO4) the solvent was removed with a rotatory evaporator and the re-
mainder purified by column chromatography (cyclohexane/EtOAc,
65:35) to yield 2a (95 mg, 12%) as colourless oil. Rf =0.29; 1H NMR:
(CDCl3): d=9.71 (s, 2H; OOH), 5.47 (q, J=5.7 Hz, 1H; CH), 1.42 ppm
(d, J=5.7 Hz, 3H; CH3); 13C NMR: (CDCl3): d=107.9 (CH), 14.6 ppm
Scheme 4. Reaction of isophorone oxide 20 with H2O2.
(CH3); HRMS
128.9958.
(ESI): m/z calcd for C2H6O4Cl: 128.9960 [M+Cl]+; found:
ACHTREUNG
Acknowledgements
We thank one referee for pointing out several interesting ideas and addi-
tions to our article, Dr. Burkhard Ziemer for X-ray crystal analysis and
Solvay Interox GmbH, Bayer Services GmbH & Co. OHG, BASF AG
and Sasol GmbH for the donation of chemicals. A. Bunge gratefully ac-
knowledges a grant from DFG.
Keywords: epoxides · hydrogen peroxide · hydroperoxides ·
oxidation · rearrangement
Figure 1. X-ray structure of the hydroperoxyperoxylactone 21.
first representative in the 7-ring series. So far, only analo-
gous 5-ring[14] and acyclic hydroperoxyperoxy esters[15] were
reported. They were obtained in totally different ways.
[1] H. J. Hamann, J. Liebscher, J. Org. Chem. 2000, 65, 1873–1876.
[2] H. J. Hamann, J. Liebscher, Synlett 2001, 96–98.
[3] H. J. Hamann, A. Wlosnewski, T. Greco, J. Liebscher, Eur. J. Org.
Chem. 2006, 2174–2180.
The hydroperoxyperoxylactone 21 can be considered as a
dimethyl analogue of ketoacids 5 being converted into a per-
acid, which cyclises and forms 5 by substitution of the hemi-
acetal OH group with H2O2. This peculiar behaviour of iso-
phorone oxide 20 could be explained by the Thorpe–Ingold
effect (geminal dialkyl effect) supporting the ring closure of
the d-oxocarboxylic acid to the hyroperoxylactone 21. Ac-
cordingly, acid treatment of the 3,3-dimethyl-5-caproic acid
with H2O2 gave 21 independently, while other ketoacids (e.
g., 5b) lacking the two methyl groups failed to give peroxi-
dic products. The acyllactone 23 is likely to be formed from
21 by a ring-contracting peroxide rearrangement. Interest-
ingly, an industrial process for the preparation of dimethyl-
glutaric acid 22 from isophorone and H2O2 was claimed.[16]
In summary, we have shown that cyclic, 3-substituted ep-
oxyketones react with H2O2 under acid catalysis by an un-
known, complex reaction mechanism that includes several
[4] E. Weitz, A. Scheffer, Ber. Dtsch. Chem. Ges. 1921, 54, 2327–2344.
[5] a) K. Zmitek, M. Zupan, S. Stavber, J. Iskra, Org. Lett. 2006, 8,
2491–2494; b) K. Zmitek, M. Zupan, J. Iskra, Org. Biomol. Chem.
2007, 5, 3895–3908.
[6] A. O. Terentꢁev, A. V. Kutkin, M. M. Platonov, I. I. Vorontsov, M. Y.
Antipin, Y. N. Ogibin, G. I. Nikishin, Russian Chemical Bulletin
2004, 53, 681–687.
[7] K. Zmitek, M. Zupan, S. Stavber, J. Iskra, J. Org. Chem. 2007, 72,
6534–6540.
[8] A. Rieche, F. Hitz, Ber. Dtsch. Chem. Ges. 1929, 62, 2458–2474.
[9] B. Das, M. Krishnaiah, B. Veeranjaneyulu, B. Ravikanth, Tetrahe-
dron Lett. 2007, 48, 6286–6289.
[10] A. A. Popova, A. P. Khardin, I. V. Shreibert, Mater. Nauchn. Konf.
Volgogr. Politekh. Inst., 1965, 2, 18–23; Chemical Abstracts 1967, 66,
85413.
[11] E. Antonelli, R. DꢁAloisio, M. Gambaro, T. Fiorani, C. Venturello,
J. Org. Chem. 1998, 63, 7190–7206.
[12] H. S. Kim, Y. Nagai, K. Ono, K. Begum, Y. Wataya, Y. Hamada, K.
Tsuchiya, A. Masuyama, M. Nojima, K. J. McCullough, J. Med.
Chem. 2001, 44, 2357–2361.
À
[13] N. A. Milas, A. Golubovic, J. Am. Chem. Soc. 1959, 81, 3361–3364.
[14] R. C. P. Cubbon, C. Hewlett, J. Chem. Soc. C 1968, 2983.
[15] B. De Vries, A. P. Van Swieten, E. A. Syed, WO 2003070699, 2003.
[16] P. Lehky, V. Franzen, EP 64633, 1981.
C C bond cleavages by Baeyer–Villiger oxidation and Hock
type hydroperoxide rearrangements. In this way ethane-1,1-
dihydroperoxide could be obtained; this compound contains
the highest peroxidic oxygen content of all hitherto known
organic compounds and is remarkably stable.
Received: May 15, 2008
Published online: July 4, 2008
Chem. Eur. J. 2008, 14, 6849 – 6851
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6851