Singlet oxygen generation from H2O2/MoO42-: Peroxidation of hydrophobic substrates in pure organic solvents

Article abstract of DOI:10.1016/S0040-4039(02)02108-1

Seventeen organic solvents are screened as reaction media to conduct the molybdate-catalyzed disproportionation of hydrogen peroxide into singlet molecular oxygen, ¹O₂ (¹Δ_g). The solvents are investigated by resorting to the detection of the infra-red luminescence of ¹O₂ at 1270 nm. Preparative peroxidations of representative substrates are carried out in the most efficient ones. The latter are protic and polar and constitute a simpler alternative to the well suited but more intricate microemulsion systems for the peroxidation of hydrophobic substrates with chemically generated ¹O₂.

Full text of DOI:10.1016/S0040-4039(02)02108-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 8731–8734
2−
Singlet oxygen generation from H2O2/MoO4 : peroxidation of
hydrophobic substrates in pure organic solvents
Ve´ronique Nardello,a Ste´phane Bogaert,a Paul L. Alstersb and Jean-Marie Aubrya,*
aLCOM, Equipe de Recherches ‘Oxydation et Formulation’, UMR CNRS 8009, ENSCL, BP 108 F-59652 Villeneuve d’Ascq
Cedex, France
bDSM Fine Chemicals-Advanced Synthesis & Catalysis, PO Box 18, 6160 MD Geleen, The Netherlands
Received 1 July 2002; revised 24 September 2002; accepted 26 September 2002
Abstract—Seventeen organic solvents are screened as reaction media to conduct the molybdate-catalyzed disproportionation of
hydrogen peroxide into singlet molecular oxygen, 1O2 (1Dg). The solvents are investigated by resorting to the detection of the
infra-red luminescence of 1O2 at 1270 nm. Preparative peroxidations of representative substrates are carried out in the most
efficient ones. The latter are protic and polar and constitute a simpler alternative to the well suited but more intricate
microemulsion systems for the peroxidation of hydrophobic substrates with chemically generated 1O2. © 2002 Published by
Elsevier Science Ltd.
1. Introduction
the aqueous microdroplets and then it diffuses before
deactivation to the organic phase where it reacts with
the substrate. Several drawbacks are, however, associ-
ated with these microemulsion systems: (i) these reac-
tion media are relatively intricate since they require
water, organic solvent, surfactant and co-surfactant, (ii)
all these components have to be mixed in well-defined
proportions, (iii) the thermodynamic properties of
microemulsions limit the amounts of H2O2 that can be
added without phase separation, (iv) the recovery of the
oxidation products is somewhat tedious.
Singlet molecular oxygen, 1O2 (1Dg), is quantitatively
generated in alkaline water through the disproportiona-
tion of hydrogen peroxide catalyzed by molybdate ions
(Eq. (1)).1
2−
MoO
4
2H2O2 ꢀꢀꢀꢀꢀꢀꢀꢁ 2H2O+1O2 (100%)
(1)
pH 9−11
In 1995, the mechanism of this reaction has been
elucidated and the different peroxo intermediates
formed as a function of pH and [H2O2] have all been
identified by 95Mo NMR.2 The oxotriperoxomolybdate,
MoO(O2)32−, was found to be the main precursor of 1O2
whereas the tetraperoxomolybdate, Mo(O2)42−, which is
formed with higher concentrations of H2O2, does not
generate the excited species.
In order to find a more easily applicable system for the
‘dark’ singlet oxygenation of lipophilic compounds, the
molybdate catalyzed disproportionation of H2O2 in
various organic solvents was investigated. Preparative
peroxidations of typical organic substrates were carried
out in organic solvents that were found to be most
efficient.
Reaction (1) is usually carried in aqueous medium
because of the hydrophilicity of reactants. Such
aqueous systems allow the peroxidation of various
hydrophilic or low molecular weight organic com-
pounds, such as tiglic acid, 1,4-cyclohexadiene deriva-
tives or allylic alcohols.3 Unfortunately, the synthetic
1
2. Detection of the IR luminescence of O2 at 1270 nm
1
2−
The generation of O2 from the system H2O2/MoO4
1
scope of this aqueous source of O2 with regard to the
was studied in 17 organic solvents by resorting to the
detection of the IR luminescence at 1270 nm. This
technique constitutes an unambiguous proof of 1O2
formation since the detection of a signal at 1270 nm is
specific of this excited species.5 The solvents were com-
pared in terms of maximal signal intensity (Imax), time
required to decompose a given amount of H2O2 (Dt)
peroxidation of hydrophobic compounds is only very
limited. Water-in-oil microemulsions4 have been devel-
1
oped to overcome this problem: O2 is generated into
* Corresponding
author.
Tel.:
00.33.3.20.33.63.64;
fax:
00.33.3.20.33.63.64; e-mail: jean-marie.aubry@univ-lille1.fr
0040-4039/02/$ - see front matter © 2002 Published by Elsevier Science Ltd.
PII: S0040-4039(02)02108-1

8732
V. Nardello et al. / Tetrahedron Letters 43 (2002) 8731–8734
1
and yield of cumulated O2. The results are listed in
Table 1. The concentrations of molybdate ions and
hydrogen peroxide were respectively equal to 0.1 and
0.5 mol/L. This molybdate concentration was chosen in
order to ensure a significant signal, in other words a
relatively high reaction rate, and the amount of H2O2
was adjusted in order to favor the formation of the
oxotriperoxomolybdate. The temperature was chosen
such that the reaction proceeded within a reasonable
time in the solvent under study.
1
The yield of O2 generation was determined by addition
of a known amount of a-terpinene 1 or furfuryl alcohol
2, depending on their solubility in the solvent under
study. These highly reactive chemical traps are known
1
Figure 1. IR luminescence signal of O2 generated by a solu-
tion containing 0.1 M Na2MoO4 and 2×0.5 M H2O2 in
methanol at T=25°C.
1
to react with O2 by a pure chemical process.6 Hence,
the missing area in the luminescence signal corresponds
to the amount of 1O2 trapped by a-terpinene 1 or
furfuryl alcohol 7. As the total area under the curve is
same luminescence signal. Addition of a-terpinene 1
leads, as expected, to a decrease of the signal. Then, as
the substrate is oxidized, the signal rises again to the
value measured in the absence of substrate. The missing
area corresponds to an amount of 1O2 equal to the
initial concentration of a-terpinene 1, and the cumu-
1
directly related to the cumulated amount of O2 gener-
ated, the comparison of the two areas provides the yield
1
of O2.
1
lated amount of O2 is calculated by comparison with
Fig. 1 represents the luminescence signal of 1O2
obtained in methanol through the disproportionation
of H2O2 0.5 M by Na2MoO4 0.1 M at 25°C. We can
notice that, as soon as H2O2 is added in the reaction
medium, an important signal with a maximum intensity
of about 470 mV is detected. Most of the H2O2 is
decomposed within 40 min and the process is shown to
be catalytic since a second batch of H2O2 provides the
the total area.
Among
the
solvents
studied,
only
THF,
dichloromethane and ethyl acetate do not lead to a
significant formation of 1O2 in the organic phase.
Indeed, in these solvents, addition of H2O2 (]0.1 M)
results in a biphasic medium and the detected signal
corresponds to 1O2 generated in the aqueous phase.
1
2−
Table 1. Luminescence detection of O2 generated from the systems H2O2/MoO4 in various pure organic solvents.
[MoO42−]=0.1 M, [H2O2]=0.5 M, Dt=time corresponding to 5% of Imax
Solvents
Chemiluminescence
Name
~D (ms)7
Molybdate
T (°C)
Imax (mV)
Dt (min)
1O2 Yield (%)
1001
Watera,d
Ethanolb
4
14
Na
Li
Na
Li
Na
K
Na
Na
Na
Na
Li
Na
Na
Na
Na
Li
25
40
40
25
25
25
25
40
35
25
40
50
40
50
50
50
50
20
25
25
25
78
502
80
441
470
27
410
166
930
850
122
150
221
172
17
21
25
186
40
40
42
20
71
24
20
172
80
23
61
21
180
330
2
92
45
81
71
36
70
69
65
42
42
27
26
26
25
25
22
11
Methanola,b
10
Ethylene glycola,d
N,N-Dimethylformamidea
Sulfolaneb
7.5
19
24c
12
N-Methylformamidea
1,2-Dimethoxyethaneb
Propylene carbonateb
Formamidea,d
–
39
11.5
15
Pyridineb
N-Methyl pyrrolidoneb
Acetonitrileb
12c
67
348
48
285
143
166
115
1,4-Dioxaneb
27
19
22
43
Li
Na
Li
Li
Li
Dimethylsulfoxideb
Tetrahydrofuranb
Ethyl acetateb
43
23
28
B5
B5
B5
Dichloromethaneb
97
a Homogeneous medium.
b Heterogeneous medium.
c Our work.
d Furfuryl alcohol. 1O2 yields are given at 5%.

V. Nardello et al. / Tetrahedron Letters 43 (2002) 8731–8734
8733
When a-terpinene 1 is added for yield measurement, no
Na2MoO4 precipitates at the end of the reaction when
hydrogen peroxide has been consumed completely. This
facilitates catalyst recycling by simple filtration.
1
decrease of the signal is observed because O2 does not
interact with the trap located in the organic phase.
Therefore, it appears that one required condition for
peroxidation of organic substrates by 1O2 chemically
generated is that added hydrogen peroxide remains
miscible with the organic solvent during the reaction. In
acetonitrile, pyridine, propylene carbonate and 1,4-
Besides catalyst solubility, another required factor for
1
efficient O2 generation is the amount of H2O2 relative
to molybdate in the reaction mixture. Care should be
taken to avoid a large excess of H2O2 since in that case
the reaction rate is considerably slowed down because
of the formation of the tetraperoxomolybdate, which
does not generate 1O2. Therefore, H2O2 should be
added gradually or batch wise in small portions. As it is
commonly observed for molybdate catalyzed H2O2 dis-
proportionation, the rate is strongly influenced by tem-
perature. In ethanol, for example, the reaction at 25°C
is very slow (more than 12 h to decompose 0.5 M H2O2
by 0.1 M Li2MoO4), but if the temperature is increased,
the reaction time considerably decreases and at 40°C,
only 28 min are required to decompose 0.5 M of H2O2
by 0.1 M of catalyst.
1
dioxane, the yields of O2 are a little bit higher but do
not exceed 27%. In these solvents, a biphasic medium is
obtained as soon as 0.5 M H2O2 is added.
The best solvents are ethanol and methanol, which give,
1
under optimized conditions, 92 and 81% of O2, respec-
tively. In the case of MeOH, the 1O2 generation
efficiency was found to be influenced by the nature of
the cation of the molybdate salt. Thus, when lithium,
sodium or potassium molybdates are used as the cata-
lyst, the yields of 1O2 generated are respectively equal to
81, 71 and 36%. Such a result shows that the solubility
of the catalyst in the organic solvent, which decreases
from lithium to potassium, is a key factor. Na2MoO4,
which is poorly soluble in MeOH, dissolves in the
presence of hydrogen peroxide because of the forma-
tion of more soluble peroxymolybdate species. An
attractive feature of the Na2MoO4 catalyzed ‘dark’
singlet oxygenation in MeOH is that most of the
Other organic solvents such as ethylene glycol, sul-
folane, 1,2-dimethoxyethane and formamide derivatives
1
provide fair yields of O2 (>40%). In these solvents, the
molybdate salt and the peroxides are more or less
completely soluble and the nature of the cation has no
1
2−
Table 2. Preparative peroxidations of typical substrates in organic solvents by the mineral source of O2, H2O2/MoO4
.
Conditions: [Substrate]=0.5 M, [Catalyst]=0.1 M. [1]=0.05 M, b[7]=0.1 M, Conversion=100% except for : 80%
a
c

8734
V. Nardello et al. / Tetrahedron Letters 43 (2002) 8731–8734
1
influence on the reaction of O2 generation. Neverthe-
less, such solvents present high boiling points and can-
not be removed easily under vacuum. In these cases, the
work up of the reaction medium consists in adding a
given amount of water and the resulting aqueous phase
is extracted with methylene chloride. After drying on
Na2SO4 and filtration, the organic phase is then rotary
evaporated under vacuum providing the oxidation
products.
Three conditions are required for an efficient reaction:
(i) the peroxomolybdates must be soluble in the reac-
tion medium, (ii) the reaction mixture must remain
miscible with water generated by H2O2 decomposition
and (iii) the rate of H2O2 addition must be equal to the
rate of its decomposition in order to form preferentially
1
the oxotriperoxomolybdate, the precursor of O2, since
an excess of H2O2 leads to the formation of the tetra-
peroxomolybdate which is very stable in organic sol-
vent. The ‘dark’ singlet oxygenation proceeds most
efficiently in MeOH or EtOH with Li2MoO4 as the
catalyst.
3. Preparative peroxidations
The most efficient solvents providing a good yield for
1O2 generation were applied to the preparative (0.5 M)
peroxidation of typical organic compounds of different
reactivities: a-terpinene 1, b-citronellol, 3, mesitol 5 and
2,3-dimethyl-but-2-ene 7. H2O2 was added in several
batches in order to favor the formation of the
oxotriperoxomolybdate. The results are listed in Table
2. All the substrates were peroxidized on the prepara-
tive scale (0.5 M) providing the expected oxidation
products through a straightforward recovery. It is note-
worthy that mesitol 5, which is difficult to oxidize
under its phenolic form on account of high physical
quenching of 1O2, could be efficiently peroxized into the
corresponding perquinol 6 provided that 1 equiv. of
NaOH was added in order to form the more chemically
reactive phenoxide form.
References
1. (a) Aubry, J. M. J. Am. Chem. Soc. 1985, 107, 5844–5849;
(b) Aubry, J. M.; Cazin, B. Inorg. Chem. 1988, 27, 2013–
2014.
2. Nardello, V.; Marko, J.; Vermeersch, G.; Aubry, J. M.
Inorg. Chem. 1995, 34, 4950–4957.
3. (a) Nardello, V.; Aubry, J. M.; Linker, T. Photochem.
Photobiol. 1999, 70, 524–530; (b) Nardello, V.; Bouttemy,
S.; Aubry, J. M. J. Mol. Catal. A: Chem. 1997, 117,
439–447.
4. Aubry, J. M.; Bouttemy, S. J. Am. Chem. Soc. 1997, 119,
5286–5294.
5. Bo¨hme, K.; Brauer, H. D. Inorg. Chem. 1992, 31, 3468–
3471.
6. (a) Schenk, G. O.; Ziegler, K. Naturwissenschaften 1944,
32, 157; (b) Braun, A. M.; Dahn, H.; Gassmann, E.;
Gerothanassis, I.; Jakob, L.; Kateva, J.; Martinez, C. G.;
Oliveros, E. Photochem. Photobiol. 1999, 70, 868–874.
7. Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem.
Ref. Data 1995, 24, 663–1021.
8. (a) Van Laar, F. M. P. R.; De Vos, D.; Vanoppen, D.;
Sels, B.; Jacobs, P. A.; Del Guerzo, A.; Pierard, F.;
Kirsch-de Mesmaeker, A. Chem. Commun. 1998, 2, 267–
268; (b) Sels, B.; De Vos, D.; Grobet, P. J.; Piet, J.; Jacobs,
P. A. Chemistry-A European Journal 2001, 7, 2547–2556.
9. Goran, Mc.; Ernest, C.; Wyborney, M. Tetrahedron Lett.
1989, 30, 783–786.
4. Conclusions
The disproportionation of H2O2 by molybdate ions in
organic solvents is by far the simplest method to oxi-
dize organic compounds with singlet oxygen. Some
molybdate-exchanged layered double hydroxides
(LDH) and resin-bound peroxomolybdates have been
developed acting as heterogeneous catalysts for H2O2
decomposition into 1O2 in organic solvents such as
acetone or dioxane but such catalysts need to be pre-
pared.8,9