Activation of hydrogen peroxide through hydrogen-bonding interaction with acidic alcohols: Epoxidation of alkenes in phenol

Article abstract of DOI:10.1021/ol0344422

(Matrix presented) Electrophilic activation of hydrogen peroxide can be achieved in acidic alcohol solvents without the need for a metal catalyst. This concept is illustrated by the epoxidation of alkenes with H₂O ₂ employing phenol as a solvent. It is proposed that intermolecular hydrogen bonding between H₂O₂ and phenol activates H ₂O₂ for oxygen-atom transfer. In this interaction, the role of phenol is purely catalytic.

Full text of DOI:10.1021/ol0344422

ORGANIC
LETTERS
2003
Vol. 5, No. 10
1777-1780
Activation of Hydrogen Peroxide
through Hydrogen-Bonding Interaction
with Acidic Alcohols: Epoxidation of
Alkenes in Phenol
Joos Wahlen, Dirk E. De Vos, and Pierre A. Jacobs*
Department of Interphase Chemistry, Centre for Surface Chemistry and Catalysis,
Katholieke UniVersiteit LeuVen, Kasteelpark Arenberg 23, B-3001 HeVerlee, Belgium
pierre.jacobs@agr.kuleuVen.ac.be
Received March 13, 2003
ABSTRACT
Electrophilic activation of hydrogen peroxide can be achieved in acidic alcohol solvents without the need for a metal catalyst. This concept
is illustrated by the epoxidation of alkenes with H O employing phenol as a solvent. It is proposed that intermolecular hydrogen bonding
2
2
between H O and phenol activates H O for oxygen-atom transfer. In this interaction, the role of phenol is purely catalytic.
2
2
2 2
In the development of cost-efficient and environmentally
benign oxidation processes, hydrogen peroxide holds a
prominent place among other oxidants. Advantages of H2O2
are its low cost, high active oxygen content, high oxidation
potential, and the formation of only water as a reduction
product. However, a disadvantage is the high activation
energy required for the oxidation of many organic com-
pounds by H2O2. Therefore, catalysis is often required. Over
the years, a wide range of catalysts based on metals (Ti, W,
Mo, Mn, Re) or semimetals (As, Se) have been designed
for the in situ activation of H2O2 for the epoxidation of
olefins. In contrast, the number of organic, nonmetal
compounds capable of activating H2O2 for epoxidation is
rather limited, and especially catalytic cases are scarce.1 The
most prominent examples are R-halo carbonyl compounds
(e.g., hexafluoroacetone), which upon reaction with H2O2
yield R-hydroxy hydroperoxide intermediates capable of
effecting epoxidation.2 Recently, it was shown that per-
fluorinated alcohol solvents such as 2,2,2-trifluoroethanol and
particularly 1,1,1,3,3,3-hexafluoro-2-propanol are able to
activate H2O2 for the epoxidation of electron-rich alkenes.3
Moreover, significant rate enhancements were observed when
classical epoxidation catalysts were employed in these
“magic” fluoro alcohol solvents.4 It was postulated that the
fluoro alcohols activate H2O2 for oxygen-atom transfer
through hydrogen bonding.
Although structurally considerably different from fluoro
alcohols, phenol shows very similar physical properties, of
which the high hydrogen-bond donor strength is particularly
(2) (a) Kim, L. German Patent 2 239 681, 1973. (b) Heggs, R. P.; Ganem,
B. J. Am. Chem. Soc. 1979, 101, 2484-2486. (c) Ganeshpure, P. A.; Adam,
W. Synthesis 1996, 179-188. (d) Adam, W.; Degen, H.-G.; Saha-Mo¨ller,
C. R. J. Org. Chem. 1999, 64, 1274-1277.
(3) (a) Neimann, K.; Neumann, R. Org. Lett. 2000, 2, 2861-2863. (b)
van Vliet, M. C. A.; Arends, I. W. C. E.; Sheldon, R. A. Synlett 2001,
248-250. (c) Legros, J.; Crousse, B.; Bonnet-Delpon, D.; Be´gue´, J.-P. Eur.
J. Org. Chem. 2002, 3290-3293.
(4) (a) Shryne, T. M.; Kim, L. U.S. Patent 4 024 165, 1977. (b) van
Vliet, M. C. A.; Arends, I. W. C. E.; Sheldon, R. A. Tetrahedron Lett.
1999, 40, 5239-5242. (c) van Vliet, M. C. A.; Arends, I. W. C. E.; Sheldon,
R. A. Chem. Commun. 1999, 263-264. (d) van Vliet, M. C. A.; Arends, I.
W. C. E.; Sheldon, R. A. Chem. Commun. 1999, 821-822. (e) van Vliet,
M. C. A.; Arends, I. W. C. E.; Sheldon, R. A. Synlett 2001, 1305-1307.
(f) ten Brink, G.-J.; Fernandes, B. C. M.; van Vliet, M. C. A.; Arends, I.
W. C. E.; Sheldon, R. A. J. Chem. Soc., Perkin Trans. 1 2001, 224-228.
(g) Berkessel, A.; Andreae, M. R. M. Tetrahedron Lett. 2001, 42, 2293-
2295. (h) Iskra, J.; Bonnet-Delpon, D.; Be´gue´, J.-P. Tetrahedron Lett. 2002,
43, 1001-1003.
(1) Adam, W.; Saha-Mo¨ller, C. R.; Ganeshpure, P. A. Chem. ReV. 2001,
101, 3499-3548.
10.1021/ol0344422 CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/17/2003

relevant.5 Therefore, we reasoned that under certain condi-
tions, phenols should be capable of H2O2 activation. In a
preliminary experiment, the oxidation of cyclooctene (5
mmol) with aqueous hydrogen peroxide (50 wt %, 5 mmol)
was attempted in phenol (50 mmol) at 40 °C (Scheme 1).
(<1%) was observed using cyclohexanol or anisole as the
solvent. In benzyl alcohol, a conversion of only 5% was
obtained after 24 h. Although cyclooctene is known as a
reactive substrate, other alkenes could be oxidized as well
(Table 1). In the oxidation of those alkenes yielding acid-
Table 1. Epoxidation of Various Alkenes by H2O2 in Phenola
Scheme 1
Figure 1 shows the concentration of epoxide and H2O2 as a
function of time. After 45 h, the yield of cyclooctene oxide
Figure 1. Plot of the concentration of epoxide (GC analysis) and
H2O2 (iodometric titration)6 for cyclooctene epoxidation by H2O2
in phenol.
a Reaction conditions: 5 mmol of alkene, 1 mmol of n-alkane (internal
standard), 0.05 mmol of NaOAc, 7.5 mmol of H2O2 (50 wt %), 50 mmol
of phenol, GC(MS) analysis.11 b Diol and R-hydroxy phenyl ether were
formed in 5-15%. c Small amounts of terminal epoxides and diepoxides
were also formed. d p-Chlorophenol (50 mmol) was used. e 2,3-Epoxide:
6,7-epoxide ) 10:90.
reached 75% (complete selectivity), while the H2O2 ef-
ficiency was 90%. This intriguing observation of epoxidation
of alkenes promoted by phenol prompted us to investigate
in greater detail the origin of this unique reactivity.7 In
addition, the present study might give better insight into the
mechanism proposed for fluoro alcohol-promoted epoxida-
tions (e.g., are the fluorine atoms essential to the catalysis?).
Moreover, the use of phenol allows the study of the electronic
effect of substituents on the epoxidation. From a synthetic
point of view, phenol is an interesting solvent because it is
available at a much lower cost than fluoro alcohols and has
a high dissolving power, especially for apolar alkenes.
To establish the essential role of phenol in the reaction, a
set of control experiments was carried out. No conversion
sensitive epoxides, the product was mostly the epoxide,
accompanied by some diol and R-hydroxy phenyl ether,
arising from ring-opening of the primary epoxide product.
However, higher epoxide selectivity can be achieved by
buffering the reaction medium with a small amount of
sodium acetate. Phosphate, borate, or carbonate bases are
effective as well, proving that the added buffer is not essential
to the observed activity. Moreover, the reaction is not
inhibited by EDTA (5 mM), which excludes participation
of metal traces.
•
The intermediacy of free radicals (O2•-, HO2 , or HO•) or
1
other reactive oxygen species such as O2 is ruled out by
(5) (a) Middleton, W. J.; Lindsey, R. V., Jr. J. Am. Chem. Soc. 1964,
86, 4948-4952. (b) Filler, R.; Schure, R. M. J. Org. Chem. 1967, 32, 1217-
1219. (c) Purcell, K. F.; Stikeleather, J. A.; Brunk, S. D. J. Am. Chem. Soc.
1969, 91, 4019-4027. (d) Eberson, L.; Hartshorn, M. P.; Persson, O.;
Radner, F. Chem. Commun. 1996, 2105-2112.
(6) Iodometric titration was used for the determination of H2O2, as
titration with CeIV resulted in the immediate formation of a black precipitate.
(7) Detailed literature survey revealed one report that mentions the use
of phenolic compounds as cocatalysts in combination with a group Va
element (e.g., As) for the epoxidation of propene with H2O2. In one
comparative example, phenol was used in the absence of As, but only a
low yield was obtained. McMullen, C. H.; Fehskens, E. E.; Plotkin, J. S.
U.S. Patent 4 410 715, 1983.
several pieces of evidence.8 Thus, oxidation of cyclooctene
is not inhibited by the presence of 2,6-di-tert-butyl-p-cresol
(50 mM), a common radical scavenger. Epoxidation of
isomeric alkenes is fully stereospecific. cis-2-Hexene yields
only cis-2,3-epoxyhexane, while oxidation of trans-2-hexene
exclusively produces the trans-epoxide. Another indication
is the absence of allylic oxidation products in the epoxidation
of cyclohexene. In contrast, epoxidation via free radical
intermediates also yields 2-cyclohexenyl-1-hydroperoxide,
1778
Org. Lett., Vol. 5, No. 10, 2003

2-cyclohexen-1-ol, and 2-cyclohexen-1-one. Spectroscopic
evidence for the absence of radicals was obtained from ESR.
The ESR spectrum of a solution of H2O2 (1.75 M) in phenol
in the presence of the spin traps 5,5-dimethyl-1-pyrroline-
N-oxide (80 mM) or N-tert-butyl-R-phenylnitrone (60 mM)
does not show typical signals arising from adducts with HO•
or other radicals.9
Although rather large quantities of phenol are used, the
phenol-H2O2 system is catalytic in that phenol is not
consumed during the reaction. Indeed, a nearly complete
mass balance is observed for phenol, and no products
originating from phenol are detected by gas chromatography.
However, in less optimal reaction conditions (no buffering,
high temperature, poorly reactive alkenes), there seems to
exist a subtle balance between epoxidation of the alkene
substrate and hydroxylation of phenol itself. In such condi-
tions, small amounts of catechol, hydroquinone, and benzo-
quinone are detected. In separate control experiments,
catechol, hydroquinone, or benzoquinone was added to the
reaction mixture, but no significant rate enhancement was
observed, thus clearly demonstrating that these phenol-
derived products are not responsible for the observed
catalysis.10
Figure 2. Second-order rate constants for the epoxidation of
cyclooctene with H2O2 in differently substituted phenols. Reaction
conditions: 10 mmol of cyclooctene, 2 mmol of n-decane, 0.1 mmol
of NaOAc, 15 mmol of H2O2, 100 mmol of phenol, 16 mL of 1,2-
dichloroethane, 60 °C.14
clooctene. Although limited to five para-substituted phenols,
this was quantified in a Hammett-type plot, which shows
that log(kX/kH) correlates linearly (F ) 0.88, r2 ) 0.995) with
the σI value of the substituents (Figure 3).12,13 The higher
Indirect information on the nature of the oxidizing species
was deduced from the oxidation of differently substituted
alkenes (Table 1). Whereas electron-rich tetra- and trisub-
stituted alkenes show high reactivity, disubstituted alkenes
are reactive as well. For these substrates, high conversions
(>95%) were obtained using only 1.5 equiv of H2O2,
indicative of the efficient use of the oxidant. Geraniol, an
allylic alcohol containing two trialkyl-substituted double
bonds, affords a 90:10 mixture of 6,7-epoxygeraniol and 2,3-
epoxygeraniol, respectively. The observed ratio reflects the
preference for oxidation of the electron-rich 6,7-bond over
the allylic 2,3-bond, which experiences the electron-
withdrawing effect of the alcohol group. Electron-deficient
terminal alkenes such as 1-octene are oxidized more slowly.
The relative reactivity series for epoxidation by H2O2 in
phenol is qualitatively similar to that expected for electro-
philic oxygen-atom transfer reactions.
Figure 3. Hammett-type plot for cyclooctene epoxidation by H2O2
in p-substituted phenols. Reaction conditions were the same as in
Figure 2.
Electrophilic activation of H2O2 is further supported by
the observation that electron-withdrawing substituents on
phenol increase the epoxidation rate (Figure 2). Whereas
phenols substituted with electron-withdrawing groups in a
meta or para position show high activity, phenols bearing
alkyl groups in the para position and ortho-substituted
phenols show lower activity in the epoxidation of cy-
epoxidation rate observed for electron-poor phenols is also
illustrated in Table 1 for the epoxidation of citronellol. In
p-chlorophenol as the solvent, about the same conversion
(95%) is obtained in 2 h instead of 8 h in phenol.
Regarding the mechanism of oxygen-atom transfer, a
tentative explanation for the activity of acidic alcohols in
alkene epoxidation might be their ability to increase the
electrophilic character of a peroxy oxygen atom of H2O2 and,
at the same time, assist the leaving group (H2O) in departing
from the reactive intermediate (Scheme 2).15,16 In these
(8) It is well-known that small amounts of phenolic compounds play a
key role in the protection of living organisms and synthetic materials against
oxidative damage. In such reactions, a phenol acts as a radical scavenger
by hydrogen atom transfer to a peroxyl radical. For example, see: (a) Avila,
D. V.; Ingold, K. U.; Lusztyk, J.; Green, W. H.; Procopio, D. R. J. Am.
Chem. Soc. 1995, 117, 2929-2930. (b) Wright, J. S.; Johnson, E. R.;
DiLabio, G. A. J. Am. Chem. Soc. 2001, 123, 1173-1183.
(9) Reference radicals were generated by Fenton’s reagent: (a) Harbour,
J. R.; Chow, V.; Bolton, J. R. Can. J. Chem. 1974, 52, 3549-3553. (b)
Janzen, E. G.; Nutter, D. E., Jr.; Davis, E. R.; Blackburn, B. J.; Poyer, J.
L.; McCay, P. B. Can. J. Chem. 1978, 56, 2237-2242.
(10) Presence of benzoquinone may induce some dioxirane-like chem-
istry. To disprove this possibility, various sets of control experiments were
carried out. Details are shown in Supporting Information.
(11) Authentic epoxides were prepared by oxidation with m-CPBA in
CHCl3.
(12) Although a Hammett correlation for substituted solvents might be
regarded as nonclassical, interaction of phenol with H2O2 in a transition
state as depicted in Scheme 2 would justify such an approach. CN and
COOH substituents were excluded because these functional groups might
react with H2O2 to form reactive peroxy species.
(13) (a) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165-
195. (b) Taft, R. W., Jr.; Lewis, I. C. J. Am. Chem. Soc. 1958, 80, 2436-
2443.
Org. Lett., Vol. 5, No. 10, 2003
1779

higher water content (30 wt % aqueous solution) or the
addition of cyclooctene oxide to the reaction mixture at the
start of the reaction also has a negative effect on the rate of
epoxidation. The importance of the degree of association of
phenol is further demonstrated by the observation that
epoxidation rates increase with increasing phenol concentra-
tion. Although a high yield of cyclooctene oxide is already
obtained using only 2 mol equiv of phenol relative to cy-
clooctene, a higher rate is observed using 5 or 10 mol equiv.
Increasing the amount of phenol beyond 10 mol equiv does
not result in a further increase in the reaction rate.
Scheme 2. Hypothetical Mechanism for Epoxidation of
Alkenes by H2O2 in Phenol
interactions, hydrogen bonding plays an important role.17,18
Moreover, aggregation of phenol molecules could be an
important factor in stabilizing the transition state.
The involvement of hydrogen bonding is further illustrated
by the observation that the addition of cosolvents with high
hydrogen-bond acceptor strength dramatically decreases the
reaction rate. In Figure 4, this is illustrated for tetrahydro-
In conclusion, intermolecular hydrogen bonding between
phenol and H2O2 yields an electrophilic peroxy species that
is able to epoxidize alkenes. In these interactions, the role
of phenol is purely catalytic.
Acknowledgment. This work was supported by the
Belgian Federal Government in the frame of an IUAP-PAI
project on Supramolecular Chemistry and Catalysis.
Supporting Information Available: Experimental pro-
cedures and detailed results of the control experiments. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL0344422
(14) 1,2-Dichloroethane was used as a cosolvent to dissolve the high-
melting phenols. Under these conditions, but in the absence of phenol,
epoxidation hardly occurred (conversion was <3% after 24 h).
(15) Similar cyclic hydrogen-bonded interactions of peroxides with
alcohols have been proposed: (a) Dankleff, M. A. P.; Curci, R.; Edwards,
J. O.; Pyun, H.-Y. J. Am. Chem. Soc. 1968, 90, 3209-3218. (b) Bruice, T.
C.; Noar, J. B.; Ball, S. S.; Venkataram, U. V. J. Am. Chem. Soc. 1983,
105, 2452-2463. (c) Richardson, W. H. In The Chemistry of Functional
Groups, Peroxides; Patai, S., Ed.; John Wiley & Sons: New York, 1983;
Chapter 5. (d) Bach, R. D.; Su, M.-D.; Schlegel, H. B. J. Am. Chem. Soc.
1994, 116, 5379-5391.
Figure 4. Epoxidation of cyclooctene with H2O2 in phenol in the
presence of different amounts of tetrahydrofuran. Reaction condi-
tions: 10 mmol of cyclooctene, 2 mmol of n-decane, 0.1 mmol of
NaOAc, 10 mmol of H2O2, 100 mmol of phenol, 40 °C, x mmol of
tetrahydrofuran (see legend at right of graph).
(16) Epoxidation of cyclooctene using tert-butyl hydroperoxide as the
oxidant at 60 °C gave 20% conversion after 24 h.
(17) Effect of hydrogen bonding is also observed in the natural abundance
17O NMR spectrum of H2O2 in phenol (δ 180 ppm, with H2O as an external
reference). This corresponds to a small but significant shift in comparison
with H2O2 in other solvents (e.g., δ 176 ppm in methanol).
(18) Reduction of charge separation in transition states by intramolecular
hydrogen bonding and proton transfer is a generally accepted concept, e.g.,
the “butterfly” transition state in epoxidations mediated by typical per-
acids: (a) Bartlett, P. D. Rec. Chem. Prog. 1950, 11, 47-51. (b) Bach, R.
D.; Owensby, A. L.; Gonzalez, C.; Schlegel, H. B.; McDouall, J. J. W. J.
Am. Chem. Soc. 1991, 113, 2338-2339.
furan. Addition of an equimolar amount of tetrahydrofuran
relative to phenol completely suppresses epoxidation. Also,
the oxidation fails when other cosolvents such as methanol
are employed.19 It is believed that these solvents disrupt the
hydrogen-bonded phenol network. The use of H2O2 with a
(19) Chlorinated cosolvents, which are less coordinating, have a much
less negative effect on the rate of epoxidation.
1780
Org. Lett., Vol. 5, No. 10, 2003