"Dark" Singlet Oxygenation of Hydrophobic Substrates in Environmentally Friendly Microemulsions

Article abstract of DOI:10.1002/1615-4169(200202)344:2<184::AID-ADSC184>3.0.CO;2-Y

The molybdate-catalyzed "dark" singlet oxygenation of hydrophobic compounds with hydrogen peroxide proceeds efficiently with low catalyst loadings (10^-3 mol%) in chlorine-free w/o microemulsions. These micro-heterogeneous systems are composed of sodium dodecyl sulfate (SDS)/n-buta-nol/water/organic phase, the latter being either a "green" solvent such as ethyl acetate or a liquid substrate, such as α-terpinene or β-citronellol. Very high reactor yields with improved product/SDS ratio can be obtained for the "dark" singlet oxygenation of such liquid substrates.

Full text of DOI:10.1002/1615-4169(200202)344:2<184::AID-ADSC184>3.0.CO;2-Y

FULL PAPERS
™Dark∫ Singlet Oxygenation of Hydrophobic Substrates in
Environmentally Friendly Microemulsions
a
a
b,
a,
¬
¬
¬
Veronique Nardello, Melanie Herve, Paul L. Alsters, * Jean-Marie Aubry *
a
√
LCOM, Equipe de Recherches ∫Oxydation et Formulation∫, ESA CNRS 8009, ENSCL, Bat C7, BP 108, 59652
Villeneuve d×Ascq, Cedex, France
Fax: ( ‡ 33)-320-33-63-69, e-mail: veronique.rataj@univ-lille1.fr
DSM Fine Chemicals Advanced Synthesis & Catalysis, P.O. Box 18, 6160 MD Geleen, The Netherlands
Fax : ( ‡ 31)-46-476-76-04, e-mail: paul.alsters@dsm.com
b
Received: October 20, 2001; Accepted: January 6, 2002
Abstract: The molybdate-catalyzed ™dark∫ singlet substrate, such as a-terpinene or b-citronellol. Very
oxygenation of hydrophobic compounds with hydro- high reactor yields with improved product/SDS ratio
gen peroxide proceeds efficiently with low catalyst can be obtained for the ∫dark∫ singlet oxygenation of
3
loadings (10 mol %) in chlorine-free w/o micro- such liquid substrates.
emulsions. These micro-heterogeneous systems are
composed of sodium dodecyl sulfate (SDS)/n-buta-
nol/water/organic phase, the latter being either a Keywords: hydrogen peroxide; microemulsion; oxida-
∫green∫ solvent such as ethyl acetate or a liquid tion; simplex; singlet oxygen; sodium molybdate.
Introduction
efficiently in pure water or in aqueous mixtures such as
methanol/water.[7] To overcome this problem, water-in-
Singlet molecular oxygen, 1O2 (1Ag), is a powerful oil (w/o) microemulsions have been proposed as reac-
oxidizing agent that reacts selectively with various tion media to carry out the ™dark∫ singlet oxygenation of
electron-rich compounds. Industrially, singlet oxygen- hydrophobic substrates (Figure 1).[8]
ation is carried out by classical dye-sensitized photo-
Such media are able to dissolve both huge amounts of
oxidation based on molecular oxygen.[1-4] This method hydrophilic compounds (Na2MoO4, H2O2), which are
requires substantial investments in dedicated gas/liquid confined in the aqueous droplets, as well as non-polar
photo-reactors, which do not profit from an economy of organic molecules, which are localized in the organic
scale. Moreover, large-scale photo-oxidation entails continuous phase. In addition, the typical size of the
hazardous processing conditions because of the combi- microdroplets (ꢀ10 nm) is much smaller than the mean
nation of light, organic solvents, and dioxygen. As a travel distance of 1O2 in water (ꢀ200 nm).[8a,9] Hence, in
1
result of the foregoing disadvantages, industrial use of spite of its short lifetime in H2O (tæ ˆ 4.4 ms), O2 can
photo-oxidation is mainly limited to the manufacture of diffuse freely, before deactivation, from the aqueous
low-volume (1 5 tons/year), high-value flavour and droplets to the organic phase, where it can react with the
fragrance compounds. ™Dark∫ singlet oxygenation by substrate.
catalytic disproportionation of hydrogen peroxide pro-
vides a safe and potentially inexpensive alternative to
photo-oxidation that can be carried out in ordinary
multi-purpose plant stirred-tank reactors.[5] It is now
well established that the disproportionation of hydrogen
peroxide catalyzed by molybdate ions leads quantita-
tively to singlet molecular oxygen, 1O2 (1Ag), in alkaline
medium, Equation (1).[6]
Figure 1. Schematic representation of the w/o microemulsion
used to oxidize organic substrates with 1O2 chemically
However, most hydrophobic organic substrates cannot
be readily oxidized with the chemical source H2O2/
MoO4 since the generation of O2 regularly proceeds
2
*
generated by the system H2O2/MoO4
SDS).
(
n-butanol,
2
1
184
¹ VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1615-4150/02/34402-184 191 $ 17.50-.50/0
Adv. Synth. Catal. 2002, 344, No. 2

™Dark∫ Singlet Oxygenation of Hydrophobic Substrates
FULL PAPERS
We have shown previously that a microemulsion industrially photooxidized with 1O2 for the synthesis of
based on sodium dodecyl sulfate (SDS) as the surfac- rose oxide.[4]
tant, n-butanol as the cosurfactant, and methylene
chloride as the oil meets all the requirements in order
to be a preparatively useful medium: (i) no phase
separation during storage and during the oxidation
Results and Discussion
process, (ii) high solubility of the reactants, (iii) chemical
inertness of the microemulsion components towards Kinetic Study by Laser Flash Photolysis
H2O2, Na2MoO4 or the intermediates derived thereof
(1O2 and peroxomolybdates), (iv) relatively high life- Influence of the Nature of the Solvent on the Lifetime of
time of 1O2 (tD ª 96 ms),[10] and (v) simple recovery of the 1O2
oxidized products, surfactant and catalyst at the end of
the reaction. In spite of its demonstrated synthetic The organic solvent, which constitutes the continuous
usefulness, however, industrial use of this SDS/n- phase of the w/o microemulsion, must meet several
BuOH/CH2Cl2-based microemulsion is hampered for requirements: (i) to be non-chlorinated and non-toxic,
several reasons: (i) environmental legislation aims at (ii) to solubilize high amounts of hydrophobic sub-
reducing the use of chlorinated solvents, (ii) large-scale strates, (iii) to have a boiling point lower than 120 8C and
handling of CH2Cl2, which also includes measures to (iv) to have a relatively low physical quenching constant
avoid emission of its vapour, is not straightforward (kd) towards 1O2 (tD > 20 ms). Based on the first three
because of its volatility, (iii) a large amount of SDS criteria, four solvents have been selected: toluene (bp ˆ
relative to the substrate is required in order to obtain a 111 8C), cyclohexane (bp ˆ 81 8C), ethyl acetate (bp ˆ
stable microemulsion.
77 8C) and dipropyl ether (bp ˆ 90 8C). The influence of
1
This work describes our attempts to develop new the nature of the solvent on the lifetime of O2 in the
microemulsions suitable for large-scale ™dark∫ singlet corresponding microemulsions was investigated by
oxygenation. The development of such new micro- resorting to laser flash photolysis,[11b,12] using (w/o)
emulsion media that meet all the demands from a microemulsions of identical compositions (71.2% of
preparative and industrial point of view is not straight- organic solvent, 15.2% of BuOH, 7.6% of SDS, and
forward. In particular, a major challenge is finding a 6.0% H2O). For comparison, CH2Cl2- and CCl4-based
substitute for CH2Cl2, in which 1O2 has a relatively long microemulsions were also studied and lifetimes of O2
1
lifetime, thus reducing the required amount of aqueous were measured in the pure solvents as well. It has
1
H2O2 and the risk of phase separation. By resorting to previously been shown that, in microemulsions, O2
the simplex method, suitable compositions for stable behaves kinetically in the same way as in homogeneous
microemulsions based on ethyl acetate as organic media.[8a]
1
solvent in the highly complex, seven-component system
The lifetimes of O2 (tD) measured in the different
(organic solvent, H2O, SDS, BuOH, substrate, Na2MoO4 solvents and in the corresponding microemulsions are
and H2O2) have been found that allow ™dark∫ singlet given in Table 1.
oxygenation of lipophilic substrates. For two liquid
The lifetime of 1O2 strongly depends on the nature of
substrates, i.e., a-terpinene (a-Terp) 1 and b-citronellol the solvent. As expected, the highest values for the
(b-Cit) 3, solvent-free microemulsion conditions were lifetime of 1O2 are obtained in the chlorinated solvents
developed that are compatible with both the preparative CH2Cl2 and CCl4.[13] It is also noteworthy that the
1
as well as the industrial demands. b-Citronellol 3 is lifetimes of O2 in the microemulsions are longer than
particularly relevant as a model since this substrate is expected if the media were considered as homogene-
1
Table 1. Lifetimes of O2, overall rate constants (kr ‡ kq) and Foote reactivity index kd/(kr ‡ kq) of b-citronellol 3 in pure
solvents and in the corresponding w/o microemulsions constituted of solvent 71.2%/BuOH 15.2%/SDS 7.6%/H2O 6.0%.
Organic phases
Pure solvents
w/o Microemulsions
tD
(ms)
kr ‡ kq
kd/(kr ‡ kq)
tD (ms)
kr ‡ kq
kd/(kr ‡ kq)
1
1
1
1
1
1
(mol ¥ L ¥ s )
(mol ¥ L )
(mol ¥ L ¥ s )
(mol  L )
Toluene
Cyclohexane
Ethyl acetate
Propyl ether
Methylene chloride
Carbon tetrachloride
23
23
38
21
97
4.4 Â 105
1.6 Â 105
5.8 Â 105
2.1 Â 105
9.2 Â 105
2.5 Â 105
0.10
0.27
0.05
0.23
0.01
0.003
25
23
30
19
41
59
4.0 Â 105
1.7 Â 105
4.4 Â 105
2.4 Â 105
5.1 Â 105
3.4 Â 105
0.10
0.25
0.07
0.22
0.05
0.05
1300
Adv. Synth. Catal. 2002, 344, 184 191
185

¬
Veronique Nardello, et al.
FULL PAPERS
ous.[8a] Except for CCl4, the lifetimes measured in the cyclohexane, or toluene. This choice results from their
pure solvents are of the same order of magnitude than first large-scale application, i.e., the enhanced oil
those measured in the microemulsions. Actually, once recovery and also from the fact that physico-chemists
1
generated in the aqueous droplet, O2 can either be prefer to use simple apolar solvents to investigate the
quenched by water or diffuse out the droplet into the properties of microemulsions.[16-18] Hence, few micro-
organic continuous phase where it can either be dispersed reaction media based on polar solvents have
deactivated by the molecules of solvent or enter back been described in the literature.[8,16,18-20] In particular,
into the aqueous micelle. The equilibrium reaction of only very few publications describe ethyl acetate-based
1O2 between the two phases is much faster than the microemulsions and, in most cases, they concern bio-
decay processes. Therefore, the w/o microemulsion chemical reactions involving enzymes, amino acids and
behaves kinetically as an ordinary solvent. In particular, proteins.[16a,19] The formulation of microemulsions based
1
the O2 lifetime in the AcOEt-based microemulsion is on polar solvents is critical because polar molecules are
1
only slightly lower than the O2 lifetime in the CH2Cl2 likely to interact with the interfacial film impeding a
microemulsion. Because the latter has already proved to straightforward formation of the microdispersed me-
be useful for preparative ∫dark∫ singlet oxygenations, dium. In addition, most reported reactions cannot be
AcOEt may also be a suitable solvent, provided that considered as really preparative. Actually, the concen-
stable AcOEt-based microemulsions could be prepared. trations of substrates are generally limited to 0.1 mol Â
1
L
in order to maintain the stability of the micro-
emulsion throughout the reaction. Only Menger has
1
Influence of the Solvent on the Reactivity of b-
Citronellol (b-Cit) 3
described the oxidation of 1 mol  L of mustard oils
with hypochlorites.[20] This high concentration could
nevertheless be achieved by compensating the influence
In order to quantitatively assess the reactivity of b- of the reactants on the stability of the microemulsion by
citronellol 3 towards 1O2 in different media, Table 1 lists high surfactant concentration. In Table 2, we have
the Foote×s reactivity index b ˆ kd/(kr ‡ kq) of b-Cit 3.[14] reported the chosen substrates for performing prepara-
This useful parameter provides the minimal concentra- tive peroxidation in chlorine-free microemulsions. For
tion of substrate required so that the interaction of 1O2 each of them, the main physical and chemical features of
with the substrate becomes predominant over the the substrate and the formulas of the expected oxidation
deactivation process by the solvent. Thus, it is preferable products are given.
to use an initial concentration of the substrate higher
To be used as a reaction medium for preparative
than the b value in order to avoid the wasting of the peroxidations, a microemulsion should allow the simul-
major part of available 1O2 by physical quenching by the taneous solubilization of high concentrations of reac-
2
solvent (kd). We can notice that the values of (kr ‡ kq) of tants (H2O2, MoO4 , substrate) without demixing dur-
b-Cit 3 depend on the polarity of the solvent. Actually, ing the process. Every constituent present in the reaction
the lowest overall rate constants are obtained in strictly medium considerably influences the stability of the
apolar solvents such as cyclohexane or CCl4 in agree- microemulsion. For instance, the introduction of an
ment with the polar character of the perepoxide electrolyte such as sodium molybdate strongly restricts
intermediate formed in the ene reaction.[2b,15] On the the microemulsion domain owing to the interaction
other hand, they do not change a lot if the solvent is pure between the anionic head R-SO4 of the surfactant SDS
‡
or in microemulsion suggesting that, under our chosen with the cations Na of the electrolyte, which makes the
conditions, b-Cit 3 is mainly located in the continuous surfactant less hydrophilic. The determination of the
organic phase. The b values reported in the last column composition of the optimal microemulsion is then
of Table 1 show that AcOEt is the best candidate as a tedious on account of the high number of interacting
substitute for chlorinated solvents.
parameters.
In order to localize rapidly the optimal microemulsion
in the quaternary phase diagram (oil, surfactant, co-
Design of an Ethyl Acetate-Based w/o Microemulsion surfactant, water), we resorted toasimplex technique.[21]
for Preparative Peroxidation of Organic Substrates
This stepwise method of optimization allows us to
determine the best compromise between conflicting
Since more than 20 years, microemulsions have been aims by varying simultaneously several parameters. A
employed as reaction media for various organic reac- similar approach has already been applied to the
tions.[16] Actually, such thermodynamically stable micro- optimization of the preparative peroxidation of b-
dispersed media allow one to solve the problem of citronellol 3 in a microemulsion SDS/BuOH/CH2Cl2/
reacting water-soluble inorganic reactants with water- H2O.[8c] The progression ofthe simplex was carried out in
insoluble organic compounds. However, in most report- a pseudo-ternary phase diagram in which the ratio
ed studies, the oil phase of the microemulsions is based BuOH/SDS was maintained at a constant value. In the
on a limited range of hydrocarbons, such as heptane, present work, the amounts of SDS, AcOEt, BuOH and
186
Adv. Synth. Catal. 2002, 344, 184 191

™Dark∫ Singlet Oxygenation of Hydrophobic Substrates
FULL PAPERS
Table 2. Substrates peroxidized in chlorine-free microemulsions.
Substrates
Features
Products
log(kr ‡ kq)[8a]
kr/(kr ‡ kq)[8a]
liquid
hydrophobic
very reactive
[4 ‡ 2] cycloaddition
8.0
1.0
solid
very hydrophobic
moderately reactive
[4 ‡ 2] cycloaddition
6.2
6.0
0.34
0.87
liquid
polar
moderately reactive
ene reaction
water were all changed but the concentration of the to solubilize simultaneously maximal amounts of cata-
model substrate, a-terpinene 1, was maintained at lyst and water without phase separation. The amount of
1
1 mol  kg in order to ensure preparative conditions. catalyst present in the microemulsion is directly related
Thus, four starting mixtures A, B, C and D (Table 3), to the reaction time whereas the amount of added water,
delimit the initial simplex, i.e., a small regular tetrahe- resulting from H2O2 decomposition limits the maximal
1
dron within the quaternary phase diagram (Figure 2).
amount of O2 that can be produced in the micro-
This initial simplex was chosen on the basis of the emulsion through reaction (1). Based on these criteria,
following requirements: (i) low amount of surfactant the ™worst∫ point (mixture C) was identified, discarded
(SDS), (ii) high proportion of organic solvent (AcOEt) and replaced by its symmetric (mixture E) in relation to
and (iii) no demixing of the microemulsion. The the centre of gravity of the three remaining points
mixtures were then compared according to their ability (mixtures A, B and D).[21] This method of simplex
progression was applied until no better conditions could
be obtained. For all studied points (A to L), proportions
of components and maximal amounts of sodium mo-
lybdate and added water are given in Table 3. Thus, the
microemulsions Jand L were those for which the highest
amounts of water and sodium molybdate could simulta-
neously be added.
During the oxidation process, hydrogen peroxide is
added in batches to the microemulsion in order to
2
promote the formation of MoO(O2)3 , the precursor of
1O2.[6a] The total amount of H2O2 that can be added is
limited by the demixing of the microemulsion due to the
accumulation of H2O coming from H2O2 (disproportio-
nation and water of dilution). By using 50% H2O2, the
1
oxidation of 1 mol  kg of a-terpinene 1 requires at
1
1
least 2 mol  kg of H2O2 which generate 5.8 mol  kg
of H2O as by-product. It is noteworthy that the most
useful microemulsions are those containing high con-
centrations of the catalyst MoO4 , in order to minimize
2
1
reaction time. Although the rate of O2 formation is
Figure 2. Quaternary phase diagram and initial simplex used
to optimize microemulsion.
rather intricate since it depends on the temperature, pH,
Adv. Synth. Catal. 2002, 344, 184 191
187

¬
Veronique Nardello, et al.
FULL PAPERS
Table 3. Compositions (% in weight) of w/o microemulsions SDS/BuOH/AcOEt/H2O containing 1 mol  kg 1 of a-terpinene 1
and the maximum amount of sodium molybdate and H2O. Mixtures F, H, K and M do not form stable microemulsions.
1
2
1
Mixtures
SDS [%]
BuOH [%]
AcOEt [%]
H2O [%]
[H2O]max [mol  kg ]
[MoO4 ]×(mol  kg )
A
B
C
D
E
G
I
7.5
7.5
7.5
7.5
7.5
12.5
7.5
2.5
4.2
10.3
7.1
6.9
5.9
8.1
70.0
75.0
70.0
70.0
73.5
67.2
64.8
64.7
61.1
62.3
63.5
15.0
10.0
10.0
10.0
13.2
15.5
13.7
11.2
17.0
14.2
14.0
2.70
5.20
6.70
11.70
3.70
6.40
10.80
14.50
10.40
13.75
13.00
0.030
0.030
0.000
0.060
0.080
0.110
0.040
0.045
0.070
0.085
0.055
12.5
10.8
13.1
11.2
17.0
15.0
17.6
14.4
J
L
O
P
2
[MoO4 ] and [H2O2],[6b] the reaction time (Dt) can be 12.5% of SDS and only 7.5% of BuOH. In addition, the
predicted aspH is constantand [H2O2]is adjustedso that amount of water that can be added is relatively high
1
the triperoxomolybdate, main precursor of 1O2, is (11.7 mol  kg ) for a concentration of catalyst equal to
1
always the prevalent species in solution, Equation (2):
0.06 mol  kg . This formulation provides a good com-
promise between the conditions of oxidation (reaction
time and amount of substrate) and the treatment of the
reaction medium for the recovery of the products.
In order to validate the optimal ethyl acetate-based
1
According to Equation (2) where Dt is expressed in microemulsion, O, the peroxidation of 1 mol  kg of a-
2
2
1
hours, we can predict that for [MoO4 ] ˆ 10 M, 10 M terpinene 1 was performed. The reaction was carried out
1
or 1 M, the peroxidation of 1 mol  kg of a-Terp 1 will at 25 8C and the disappearance of the substrate was
require 10 h, 1 h or 6 min, respectively.
monitored by HPLC at 266 nm after dilution of the
Finally, for a further optimization, we investigated two reaction medium in MeOH. The peroxidation of a-Terp
other microemulsions, O and P, localized at the same 1 was completed in 1.5 h with addition of a total amount
1
1
distance between J and L (see Table 3). For both of H2O2 equal to 2.2 mol  kg . Thus, 91% of the O2
microemulsions, the amounts of H2O that could be formed in the reaction medium was trapped chemically
added before demixing were higher than the one by the substrate. The NMR analysis confirmed the
1
required (5.8 mol  kg ) for the peroxidation of formation of pure ascaridole 4, the expected oxidation
1
1 mol  kg of a-Terp 1. However, the reaction time, product.
2
which is inversely proportional to [MoO4 ], must also be
Finally, the microemulsion O was applied to the
considered as an important factor in the determination oxidation of 9,10-diphenylanthracene (DPA) 2, which is
of the optimal microemulsion. Therefore, according to about 100 times less reactive than a-terpinene 1 and
this secondary criterion, the microemulsion O, would be much less soluble in polar solvents. The concentration of
1
1
the best one since 13.85 mol  kg of H2O could be 2 was equal to 0.1 mol  kg , which is higher than its
2
added before demixing for a concentration of MoO4
solubility in the reaction medium. Under such condi-
1
equal to 0.085 mol  kg . Under these conditions, a tions, a yellow suspension of 2 was present at the
predicted reaction time of 1.2 h can be calculated beginning of the reaction. The disappearance of DPA 2
instead of 2.5 h for the microemulsion J.
and the appearance of its endoperoxide DPAO2 5 were
Such an approach allows us to define two other followed by HPLC at 392 nm and 220 nm, respectively.
interesting microemulsions, B and D (see Table 3). The The peroxidation of DPA 2 was completed in 2.0 h with
first one contains the highest amount of organic phase addition of a total amount of H2O2 equal to 3.3 mol Â
1
(75%) and the lowest amount of SDS (7.5%). Such a kg . This result shows that 36% of the 1O2 formed in the
composition is particularly favourable to solubilize high reaction medium was trapped chemically by the sub-
amounts of substrate and to make easier the recovery of strate. The NMR analysis confirmed the formation of
the oxidation product since elimination of SDS and pure DPAO2 5, the expected oxidation product. It is
BuOH is troublesome. However, this microemulsion B noteworthy that, at the end of the reaction, the reaction
would be more suitable for highly reactive substrates, medium is still a microemulsion, with a white suspension
with no physical quenching of 1O2, since the amount of of DPAO2, suggesting that a higher amount of starting
H2O and catalyst that can be added are only of 5.2 and material could have been oxidized. This result clearly
1
0.03 mol  kg , respectively. The microemulsion D is shows that the oxidation can well be performed even
also of some interest since it contains 70% of AcOEt, when the substrate is not completely solubilized in the
188
Adv. Synth. Catal. 2002, 344, 184 191

™Dark∫ Singlet Oxygenation of Hydrophobic Substrates
FULL PAPERS
microemulsion. Such a procedure presents a tremen- water that could be added to these microemulsions
dous advantage in terms of oxidation product recovery before demixing, the proportions of BuOH and of the oil
since it requires a simple filtration of solid products.
were consequently adjusted. The amount of SDS was
maintained in the range 16 17%. On the other hand, in
order to add a sufficient amount of sodium molybdate
likely to provide an acceptable reaction time in compari-
son to the high concentration of substrate, the propor-
tion of the oil should be decreased to 57% for b-Cit 3
Design of Solvent-Free w/o Microemulsions for
Preparative Peroxidations of Liquid Organic
Substrates
1
(i.e., 3.6 mol  kg ) and 59% for a-Terp 1 (i.e. 4.3 mol Â
1
The concept of ™solvent-free∫ microemulsions has al- kg ), respectively. In the same way, microemulsions
ready been applied to some reactions involving the could be obtained by addition of much higher amounts
copolymerization of vinyl acetate,[22a] the catalytic of BuOH, 16 and 13% for a-Terp 1 and b-Cit 3,
oxidation of alkanes with H2O2 and iron salts[22b,22c] and respectively, instead of 5.8% for the AcOEt-based
the oxidation of olefins and styrene by the Wacker microemulsion.
1
process.[22d,22e] For all these systems, the reacting sub-
The quantitative peroxidation of 4.3 mol  kg a-
strate, which is always an unpolar compound, constitutes Terp 1 in the solvent-free microemulsion was limited by
the organic continuous phase of the microemulsion. For the demixing of the reaction medium which was
1
singlet oxygen peroxidations of organic hydrophobic observed when 4.1 mol  kg of H2O2 were added.
substrates, such solvent-free microemulsions exhibit Under these conditions, 45% of the starting material
1
several advantages: (i) huge amounts of reacting sub- could be oxidized providing 1.94 mol  kg of ascari-
strate, (ii) no toxicity problem arising from the solvent dole 4. After treatment of the final reaction medium, a
1
and (iii) maximum chemical trapping of O2 since no mixture of a-terpinene 1 and ascaridole 4 remains that
physical quenching of 1O2 by the solvent (kd) takes place. can be readily separated by distillation under vacuum.
1
The only requirement is that both substrate and product On the other hand, 3.6 mol  kg of b-Cit 3 could be
1
have to be liquid. In order to apply this concept to 1O2 completely oxidized with 8.6 mol  kg of H2O2 without
oxidations, a-terpinene 1 and b-citronellol 3 were demixing of the microemulsion, leading to equal
chosen as models since they illustrate nicely the typical amounts of the two expected hydroperoxides 6 and 7.
1
[4 ‡ 2] cycloaddition and the ene reaction. The first Such a result provides a fixation rate of O2 equal to
1
substrate, which is very reactive towards O2, should 84%. In order to bring out the usefulness and the
rather be considered as a hydrophobic oil whereas the efficiency of environmentally friendly microemulsions,
second one is polar through the presence of the hydroxy we have summarized in Table 4 all the peroxidations of
groupand presents a moderate reactivity towards the
excited species (see Table 1).
a-terpinene 1 and b-citronellol 3 described in the
literature and carried out in microemulsion systems by
The composition of the solvent-free microemulsions using chemically generated 1O2.
was defined from the composition of the optimized ethyl
This table reveals two important results: (i) the
acetate-based microemulsion O, previously determined replacement of CH2Cl2 by the more environmentally
with the simplex method. According to the amount of friendly AcOEt allows a-terpinene 1 to be quantita-
Table 4. Peroxidations of a-terpinene 1 and b-citronellol 3 performed in microemulsions.
a-Terpinene 1
2
Oil
[a-Terp]0
[mol  kg ]
[MoO4 ]/[prod]
[H2O2]/[prod]
[SDS]/[prod]
Conversion [%]
Ref.
1
CH2Cl2
CH2Cl2
AcOEt
a-Terp4.3
0.32
0.49
1.0
0.047
0.098
0.085
2.12
2.44
2.25
2.20
0.44
1.30
1.03
0.94
45
100
100
100
[5b]
[8a]
this work
0.0049
this work
b-Citronellol 3
2
Oil
[b-Cit]0
[mol  kg ]
[MoO4 ]/[prod]
[H2O2]/[prod]
[SDS]/[prod]
Conversion [%]
Ref
1
CH2Cl2
CH2Cl2
CH2Cl2
b-Cit
0.29
0.39
1.5
0.052
0.039
0.023
0.0036
2.96
4.70
3.33
2.39
1.08
1.58
0.47
0.25
100
82
100
100
[5b]
[8a]
[8c]
this work
3.6
Adv. Synth. Catal. 2002, 344, 184 191
189

¬
Veronique Nardello, et al.
FULL PAPERS
tively converted at a higher concentration (twice more) Instrumentation
and with less SDS and H2O2 per product, thus signifi-
Flash Photolysis: (PICR, Manchester). The photosensitized
generation of O2 was carried out by energy transfer from a
cantly increasing efficiency and utility of microemul-
sions as media for ∫dark∫ singlet oxygenation, (ii)
solvent-free microemulsions allow the preparative per-
oxidation of liquid substrates at very high concentration
('4 mol/kg) with low amounts of SDS and molybdate
catalyst per product. However, with some substrates
such as a-terpinene 1, the conversion is incomplete and
the product must be separated from the substrate at the
end of the reaction.
1
photosensitizer (TPP or TPPS) to 3O2. In a quartz cell, a 2-mL
sample of an air-saturated solvent or microemulsion contain-
ing the photosensitizer (TPP for organic solvent, TPPS for
microemulsions) was irradiated with a short (6 ns) flash of light
at 532 nm, emitted by a Nd-YAG laser. Infrared phosphor-
escence of 1O2 at 1270 nm was detected perpendicularly with a
germanium photodiode (Judson J16) and the signal was
recorded with an oscilloscope and analyzed according to a
first-order decay.[11]
HPLC: High-performance liquid chromatography analyses
were carried out with a reversed-phase column (Nova-pack
C18, 4 mm, 4.6 Â 250 mm) using a 600 controller pump from
Waters, a mixture of CH3OH/H2O/H3PO4 as the eluent, and
UV detection with a Waters 490E programmable multiwave-
length detector.
Conclusion
When a liquid substrate has to be peroxidized with
chemically generated 1O2, solvent-free microemulsions
are the most convenient media. They allow one to
convert huge amounts of starting materials using low
1H and 13C NMR Spectroscopy: Spectra were recorded in
1
CDCl3 at 300 MHz for the H and 75.46 MHz for the 13C by
using a Bruker AC 300P FT spectrometer. All chemical shifts
are relative to the TMS signal (d ˆ 0 ppm) as reference.
2
SDS/product and MoO4 /product ratios with minimum
loss of 1O2 through physical quenching. For solid
substrates, the alternative solution is to resort to ethyl
acetate-based microemulsions which also provide high
substrate conversions on the preparative scale. In
addition, such media do not require complete solubili-
zation of the substrate and permit a straightforward
recovery of poorly soluble products such as DPAO2 by a
simple filtration.
General Procedure for Preparative Peroxidations in
Microemulsion
Microemulsions were prepared in a thermostatted bath at
25 8C by adding dropwise an aqueous solution of sodium
molybdate to a magnetically stirred slurry of SDS, n-BuOH,
and organic phase (AcOEt or liquid substrate). The trans-
parent, mobile and isotropic medium was obtained after 5 min
of mixing and can be kept unchanged in a capped flask for
several weeks. To this system was added the substrate for
AcOEt-based microemulsions. At the initial time, a fraction of
50% H2O2 was introduced and the red-brown reaction medium
was stirred at 25 8C until the colour faded to yellow. Several
other fractions of H2O2 were then allowed to react in the same
way until complete conversion of the substrate or demixing of
the microemulsion. The solvents (BuOH, water and AcOEt)
were rotary evaporated at 40 8C under vacuum. The residue
was stirred vigorously in 100 mL of toluene for 15 min and the
suspension was filtered to recover sodium molybdate and
sodium dodecyl sulfate. It is noteworthy that, if an environ-
mentally friendly process has to be investigated, toluene can
readily be replaced by any other solvent like cyclohexane, able
to dissolve the oxidation products and to leave as precipitates
both sodium molybdate and SDS. The resulting organic phase
was dried with Na2SO4 and passed through a silica layer in
Experimental Section
Chemicals
Sodium molybdate dihydrate (99%), a-terpinene (a-Terp, 1,
98%) 9,10-diphenylanthracene (DPA, 2, 98%), b-citronellol
(b-Cit, 3, 95 %), methylene chloride (99.9%), toluene (99%),
cyclohexane (99%), ethyl acetate (99.5%), dipropyl ether
(99%), 5,10,15,20-tetraphenyl-21H,23H-porphine (TPP, 98%)
and 5,10,15,20-tetraphenyl-21H,23H-porphine-p, p',p'',p'''-tet-
rasulfonic acid, tetrasodium salt dodecahydrate (TPPS, 98%)
were purchased from Aldrich Chemie and used as received.
Carbon tetrachloride (Rectapur), n-butanol (BuOH) (Norma-
pur), sodium dodecyl sulfate (SDS, 98%), hydrogen peroxide
(Rectapur, 50%) were obtained from Prolabo.
Table 5. Compositions of chlorine-free microemulsions and amounts of reactants used for the preparative peroxidation of
organic compounds.
Microemulsion compositions
Substrates
H2O2 (50%)
Conversion [%]
2
Solvent
SDS
BuOH
H2O
MoO4
AcOEt
AcOEt
a-Terp 1
b-Cit 3
31.2 g
31.2 g
17.7 g
17.0 g
8.8 g
8.8 g
4.8 g
5.25 g
2.9 g
2.9 g
4.8 g
4.0 g
7.1 g
7.1 g
2.7 g
3.75 g
1.03 g
1.03 g
0.069 g
0.097 g
a-Terp 1
DPA 2
a-Terp 1
b-Cit 3
6.85 g
1.7 g
8 Â 0.8 mL
12 Â 0.8 mL
70 Â 0.1 mL
150 Â 0.1 mL
100
100
45
100
190
Adv. Synth. Catal. 2002, 344, 184 191

™Dark∫ Singlet Oxygenation of Hydrophobic Substrates
FULL PAPERS
order to eliminate traces of SDS and catalyst. Finally, toluene
was rotary evaporated under vacuum at yielding the oxidation
products. Compositions of the microemulsions and amounts of
reactants are listed in Table 5.
[9] P. C. Lee, M. A. J. Rodgers, J. Phys. Chem. 1984, 88,
4385 4389.
[10] F. Wilkinson, W. P. Helman, A. B. Ross, J. Phys. Chem.
Ref. Data, 1995, 24, 663 1021.
[11] a) C. Vever-Bizet, M. Dellinger, D. Brault, M. Rougee,
R. V. Bensasson, Photochem. Photobiol. 1989, 50, 321
325; b) M. A. J. Rodgers, P. T. Snowden, J. Am. Chem.
Soc. 1982, 104, 5541 5543.
[12] A. A. Krasnovsky, Jr., Chem. Phys. Lett. 1981, 443 445.
[13] a) R. Schmidt, H. D. Brauer, J. Am. Chem. Soc. 1987,
109, 6976 6981; b) R. Schmidt, E. Afshari, J. Phys.
Chem. 1990, 94, 4377 4378; c) J.-M. Aubry, B. Mandard-
Cazin, M. Rougee, R. V. Bensasson, J. Am. Chem. Soc.
1995, 117, 9159 9164.
[14] M. El Bouamri, J. P. Gorrichon, A. M. Braun, E.
Oliveros, Photochem. Photobiol. 1991, 54, 619 623.
[15] M. Hild, H. D. Brauer, Ber. Bunsen-Gesel. Phys. Chem.
Chem. Phys. 1996, 100, 1814 1817.
[16] a) M. J. Schwuger, K. Stickdorn, Chem. Rev. 1995, 95,
849 864; b) F. M. Menger, Angew. Chem. Int. Ed. Engl.
1991, 30, 1086; c) J. Klier, C. J. Tucker, T. H. Kalantar,
D. P. Green, Advanced Materials, 2000, 12, 1751 1757;
d) K. Holmberg, Advances in Colloid and Interface
Science 1994, 51, 137 174.
Acknowledgements
Flash photolysis experiments were performed at the Paterson
Institute for Cancer Research, Free Radical Research Facility
(Manchester, UK) with the support of the European Commis-
sion through access to large-scale activity of the TMR program.
References
[1] a) M. Prein, W. Adam, Angew. Chem. Int. Ed. Engl.
1996, 35, 447 494; b) H. H. Wasserman, J. L. Ives,
Tetrahedron 1981, 37, 1825 1852.
[2] a) E. L. Clennan, Tetrahedron 2000, 56, 9151 9179;
b) M. Stratakis, M. Orfanopoulos Tetrahedron 2000, 56,
1595 1515.
[3] P. Esser, B. Pohlmann, H.-D. Scharf, Angew. Chem. Int.
Ed. Engl. 1994, 33, 2009 2023.
[4] a) K. Gollnick, Chim. Ind. 1982, 64, 156 166; b) A. M.
Braun, M. T. Maurette, E. Oliveros, Technologie Photo-
chimique, Presses Polytechniques Romandes, 1986, p p .
429 481.
[5] a) A. E. Gekhman, G. E. Amelichkina, N. I. Moiseeva,
M. N. Vargaftik, I. I. Moiseev, J. Mol. Catal. A: Chemical
2000, 162, 111 124; b) J.-M. Aubry, French Patent
2,612,512, 1987; Chem. Abstr. 111, 133340; c) P. L.
Alsters, V. Nardello, J.-M. Aubry, WO 00/61524 (PCT/
EP00/02552); Chem. Abstr. 133, 281926; d) P. L. Alsters,
V. Nardello, J.-M. Aubry, WO 00/64842 (PCT/EP00/
02553); Chem. Abstr. 133, 334856.
[6] a) V. Nardello, J. Marko, G. Vermeersch, J.-M. Aubry,
Inorg. Chem. 1995, 34, 4950 4957; b) J.-M. Aubry, B.
Cazin, Inorg. Chem. 1988, 27, 2013 2014; c) Q. Niu, C. S.
Foote, Inorg. Chem. 1992, 31, 3472 3476; d) K. Bˆhme,
H.-D. Brauer, Inorg. Chem. 1992, 31, 3468 3471; e) T.
Kawatani, J. M. Lin, M. Yamada, Analyst 2000, 125,
2075 2078.
[7] a) V. Nardello, J.-M. Aubry, T. Linker, Photochem.
Photobiol. 1999, 70, 524 530; b) V. Nardello, S. Bout-
temy, J.-M. Aubry, J. Mol. Catal. A: Chemical 1997, 117,
439 447; c) J.-M. Aubry, B. Cazin, F. Duprat, J. Org.
Chem. 1989, 54, 726 728; d) J.-M. Aubry, J. Am. Chem.
Soc. 1985, 107, 5844 5849.
[17] a) D. O. Shah, R. S. Schechter, (Eds.), Improved Oil
Recovery by Surfactant and Polymer Flooding, Academic
Press, New York, 1977; b) D. O. Shah, (Ed.), Surface
Phenomena in Enhanced Oil Recovery, Plenum, 1979.
[18] a) C. Solans, R. Pons, H. Kunieda, Industrial Applica-
tions of Microemulsions, Marcel Dekker, Inc. New York,
1997, pp. 1 19; b) R. Zana, J. Lang, in Microemulsions:
Structure and Dynamics, (Eds.: S. E. Friberg, P. Bothor-
el), Chemical Rubber Company: Boca Raton, FL; 1987.
[19] a) X. Jorba, P. Clapes, J. L. Torres, G. Valencia, J.
Mataalvarez, Colloids and Surfaces A-Physicochemical
and Engineering Aspects 1995, 96, 47 52; b) G. F. Hall,
P. F. Turner, Electroanalysis 1994, 6, 217 220; c) K. L.
Lee, J. F. Biellmann, Tetrahedron 1988, 44, 1135 1139.
[20] F. M. Menger, A. R. Elrington, J. Am. Chem. Soc. 1991,
113, 9621.
[21] a) G. A. Lewis, D. Mathieu, R. Phan-Tan-Luu, Pharma-
ceutical experimental design, in Drugs and the pharma-
ceutical sciences, Marcel Dekker, New York, Vol. 92,
1999; b) D. Mathieu, E. Puech-Costes, M. T. Maurette, R.
Phan Tan Luu, Chemometrics and Intelligent Laboratory
Systems 1993, 20, 25 34.
[22] a) D. Donescu, L. Fusulan, J. Dispersion Sci. Technol.
1996, 17, 701 715; b) T. Briffaut, C. Larpent, J. Chem.
Soc. Chem. Commun. 1990, 1193 1194; c) C. Larpent,
H. Patin, J. Mol. Catal. 1992, 72, 315 329; d) N. Alandis,
I. Rico-Lattes, New J. Chem. 1994, 18, 1147 1149; e) N.
Alandis, I. Rico, A. Lattes, Bull. Soc. Chim. Fr. 1988, 2,
252 255.
[8] a) J.-M. Aubry, S. Bouttemy, J. Am. Chem. Soc. 1997,
119, 5286 5294; b) V. Nardello, M. J. Marti, C. Pierlot,
J.-M. Aubry, J. Chem. Ed. 1999, 76, 1285 1288; c) S.
Bouttemy, J.-M. Aubry, M. Sergent, R. Phan Than-Luu,
New J. Chem. 1997, 21, 1073 1084.
Adv. Synth. Catal. 2002, 344, 184 191
191