Manganese(III) porphyrin catalysts for the oxidation of terpene derivatives: A comparative study

Article abstract of DOI:10.1006/jcat.2001.3493

A comparative study involving four manganese(III) porphyrin catalysts combined with two different oxidants (sodium hypochlorite and potassium monopersulfate) has been performed in the epoxidation of three terpene derivatives. The catalytic oxidation of α-pinene produces selectively 100% of the epoxide or 65% of allylic oxidation products, only by modification of the substituents on the meso-positions of the metalloporphyrin catalyst. The catalytic oxidation of the 5-vinyl-2-norbornene is regio- and stereoselective, producing only the exo-2,3-epoxy-5-vinylnorbornane. With α-terpinene, a conjugated di-olefin, an oxidative dehydrogenation reaction was surprisingly observed, producing p-cymene as a major compound.

Full text of DOI:10.1006/jcat.2001.3493

Journal of Catalysis 206, 349–357 (2002)
doi:10.1006/jcat.2001.3493, available online at http://www.idealibrary.com on
Manganese(III) Porphyrin Catalysts for the Oxidation of Terpene
Derivatives: A Comparative Study
∗
∗,1
Vale´ rie Maraval, Jean-Erick Ancel,† and Bernard Meunier
∗
Laboratoire de Chimie de Coordination du CNRS, 205 route de Narbonne, 31077 Toulouse cedex 4, France; and †Rhone-Poulenc
Industrialisation, 24 avenue Jean Jaure`s, 69153 De´cines-Charpieu Cedex, France
Received September 18, 2001; revised December 10, 2001; accepted December 13, 2001
pene derivatives involving four different manganese(III)
porphyrin complexes (Fig. 1) combined with two oxidants.
Oxidation products of terpenes are useful intermediates in
fragrance industries.
A comparative study involving four manganese(III) porphyrin
catalysts combined with two different oxidants (sodium hypochlo-
rite and potassium monopersulfate) has been performed in the
epoxidation of three terpene derivatives. The catalytic oxidation
of α-pinene produces selectively 100% of the epoxide or 65% of al-
lylic oxidation products, only by modification of the substituents
on the meso-positions of the metalloporphyrin catalyst. The cata-
lytic oxidation of the 5-vinyl-2-norbornene is regio- and stereose-
lective, producing only the exo-2,3-epoxy-5-vinylnorbornane. With
α-terpinene, a conjugated di-olefin, an oxidative dehydrogenation
reaction was surprisingly observed, producing p-cymene as a major
EXPERIMENTAL AND METHODS
Methods
The conversions of α-pinene, α-terpinene, and 5-vinyl-
2-norbornene and the yields of the differents oxidation
products were determined by gas chromatography (GC).
GC analyses were performed with an Intersmat IGC 120
FDL instrument equipped with a capillary column Alltech
AT-WAX 30 m × 0.25 mm. Peak areas were measured
with an integrator Spectra-Physics SP 4290. GC/MS analy-
ses were performed with a HP 6890/MSD 5973 instrument
equipped with a capillary column HP-5MS 30 m×0.25 mm.
1H NMR spectra were recorded on a Bruker AM 250
spectrometer. UV-visible spectra were recorded on a HP
8452A diode array spectrophotometer. Chemicals were
purchased from Aldrich and Fluka. Sodium hypochlorite
solution was supplied from Prolabo as an aqueous solu-
tion containing about 3.5% of active chlorine (0.5 M).
Curox(2KHSO5 · KHSO4 · K2SO4)wasagiftfromPeroxid-
Chemie, Germany.
c
compound. ꢀ 2002 Elsevier Science (USA)
Key Words: manganese(III) porphyrins; epoxidation; oxidative
dehydrogenation; α-pinene; 5-vinyl-2-norbornene; α-terpinene.
INTRODUCTION
Porphyrins are macrocyclic molecules consisting on the
assembly of four pyrrole rings. The large number of ac-
cessible synthetic porphyrin ligands has created a large re-
search field. Metalloporphyrins are widely used in the area
of oxidation catalysis (1). The key concept consists of using
simple oxidants as single oxygen atom donors to generate
high-valent metal–oxo species active in metalloporphyrin-
catalyzed oxidations. In 1979, Groves et al. published the
first article on the use of iodosylbenzene as a simple oxy-
genatomdonorinolefinepoxidationandalkanehydroxyla-
tion catalyzed by an iron(III) porphyrin complex (2). Since
then, several other oxidants have been used such as sodium
hypochlorite (3), potassium monopersulfate (4), hydrogen
peroxide (5), and t-butylhydroperoxide (6).
The modification of the substituents on the meso-
positions of the porphyrin macrocycle is known to induce
a change in the rate and the selectivity of catalytic oxida-
tions (the meso-positions of a porphyrin correspond to the
interpyrrolic methyne positions) (1). Thus, we decided to
carry out a comparative study in the oxidation of three ter-
Synthesis and Characterization of Catalysts
The (meso-tetraphenylporphyrinato) chloromangan-
ese(III) Mn(TPP)Cl is commercially available. The meso-
tetramesitylporphyrin H2TMP and the meso-tetrakis(2,6-
dichlorophenyl)porphyrin H2TDCPP were prepared using
a previously described procedure (7). The meso-tetrakis
(ortho-nitrophenyl)porphyrinH2TNPPwassynthesizedus-
ing the same procedure as for H2TMP. The crude product
was purified by chromatography over a dry column of ba-
sic alumina using first dichloromethane as eluent, and then
dichloromethane with 0.5% methanol (yield = 3%). 1H
NMR ([D6] DMSO) : δ (ppm) −2.70 (s, 2H, NH), 8.28 (m,
8H, Harom), 8.68 (m, 8H, Harom), 8.85 (s, 8H, Hpyr). UV-
visible (CH2Cl2) : λmax (nm) 422. C44H26N8O8 requires :
1 To whom correspondence should be addressed. E-mail: bmeunier
@lcc-toulouse.fr.
349
0021-9517/02 $35.00
c
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

350
MARAVAL, ANCEL, AND MEUNIER
tions were monitored by gas chromatography analyses of
aliquots of the organic phase.
General Procedure for Catalytic Oxidation Reactions
Using KHSO5 as Oxidant
All experiments were carried out at room tempera-
ture, under an air atmosphere, in a 50 ml round bottle
equipped with a magnetic stirring bar. Porphyrin complex
(5 µmol), BDTAC (0.015 mmol), 4-t-butylpyridine (none
or 0.15 mmol), olefin (1.0 mmol) and an internal stan-
dard for gas chromatography were dissolved in 0.5 ml of
dichloromethane. A solution of KHSO5 (1.1 or 1.75 mmol)
in 3 ml of a 0.25 M phosphate buffer pH 7 was then added.
The reactions were monitored by gas chromatography anal-
yses of aliquots of the organic phase.
RESULTS AND DISCUSSION
1. Catalytic Oxidation of a Cyclic Mono-Olefin, α-Pinene
FIG. 1. Structure of the four Mn(III) porphyrins used as catalysts.
Oxidation of α-pinene 1 has been studied using various
methods and generally gives a mixture of products such as
its epoxide 2, verbenol 3, verbenone 4 (Fig. 2).
C, 66.50; H, 3.30; N, 14.10. Found : C, 66.12; H, 3.15; N,
14.07.
For example, the oxidation of α-pinene with oxygen using
a homogeneous cobalt catalyst was described by Lajunen
and Koskinen, with a verbenone selectivity of over 70%
(10). Mn(III) salen complexes have been used, with io-
dosylbenzene as oxidant, leading to α-pinene oxide with
a 55% selectivity (11). Using methyltrioxorhenium and hy-
drogen peroxide, Jacobs et al. described a 90% selectivity in
α-pinene oxide (12). In addition, a number of studies have
investigated the oxidation of α-pinene using heterogeneous
catalysts such as resins (13), silica–titania co-gels (14), and
metallosilicates containing molybdenum (15).
The chloromanganese(III) porphyrin complexes Mn
(TMP)Cl and Mn(TDCPP)Cl were prepared using
dimethylformamide as a solvent as described (8, 9).
Mn(TNPP)Cl was synthesized using the same proce-
dure and purified by chromatography over a dry col-
umn of silica gel. The elution was performed first with
a mixture dichloromethane/hexane (90/10, v/v) and then
with a mixture of dichloromethane/methanol (98/2, v/v)
(yield = 60%). UV-visible (CH2Cl2) : λmax (nm) 484. De-
spite several attempts at purification, inorganic residue re-
mained in the sample. The elemental analyses indicated
that the C/H/N ratio was correct, but the loading of inor-
ganic impurities was estimated to be 10–15% based on the
comparison of the ε value of the Soret band of this metal-
loporphyrin with other similar complexes prepared in the
group.
1.1. Sodium hypochlorite as oxidant. The use of Mn-
(TPP)OAc as a catalyst for the epoxidation of α-pinene
with sodium hypochlorite as oxidant has been described in
previous papers (16, 17) leading to α-pinene oxide 2 with
80% yield. The results we obtained for the oxidation of
the same substrate with NaOCl in the presence of several
manganese(III) porphyrin complexes are summarized in
Table1. Thecatalyticoxidationreactionswerecarriedoutat
roomtemperatureusing0.5%molarofMn(III)catalystand
1.5% molar of 4-t-butylpyridine with respect to α-pinene.
The addition of pyridine derivatives acting as axial ligand
for the metal core of the porphyrin is known to improve
General Procedure for Catalytic Oxidation Reactions
Using NaOCl as Oxidant
All experiments were carried out at room temperature,
under an air atmosphere, in a 50 ml round bottle equipped
with a magnetic stirring bar. Either 1.1 or 1.75 mmol of
a 0.5 M NaOCl solution was added to a solution of por-
phyrin complex (5 µmol), benzyldimethyltetradecylammo-
nium chloride (BDTAC) (0.015 mmol), 4-t-butylpyridine
(0.15 mmol), olefin (1.0 mmol) and an internal standard
for gas chromatography in 0.5 ml of dichloromethane. The
internal standards used were alkanes or halogenated ben-
zene derivatives (n-undecane or chlorobenzene). The reac-
FIG. 2. Catalytic oxidation reaction of α-pinene 1.

MANGANESE(III) PORPHYRIN CATALYSTS IN TERPENE OXIDATION
351
TABLE 1
In run 3, 73% of α-pinene oxide and 16% of verbenone
were obtained, with 100% substrate conversion. So, the ma-
terial balance indicates that 11% of one or several other oxi-
dation products are formed. But these compounds have not
been identified by GC analyses and might result from rear-
rangementofα-pineneoxide. Severalarticleshavereported
this weak stability of α-pinene oxide in oxidative conditions
leading to rearranged products such as campholenic alde-
hyde, trans-sobrerol, or trans-carveol for example (20, 21).
In order to reduce this phenomenon, a catalytic reaction
has been performed with only a slight excess of oxidant
(Table 1, run 5). As expected, no over-oxidation of α-pinene
oxide was observed and 88% yield of this compound was
obtained. It is important to specify that in all these catalytic
oxidations of α-pinene, a stereoselective oxidation was ob-
served leading to the formation of α-pinene oxide 2 and
verbenol 3. In all cases, the oxidation occurred on the less
sterically hindered side of the olefin, corresponding to an
oxygen addition of the high-valent metal–oxo species anti
to the bridge bearing the two methyl groups (as represented
in Fig. 2). In order to complete this comparative study, an-
other oxidant was used with the same Mn(III) catalysts.
Catalytic Oxidations of α-Pinene Using NaOCl as Oxidant
Eq. of
oxidant
vs substrate
Yield (%)a
Conversion
(%)
T.O.F.b
Run
Catalyst
2
3
4
(min−1
)
1
2
3
4
5
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Cl
1.75
1.75
1.75
1.75
1.1
100
98
100
97
50 36 10
0.8
1.6
0.5
0.5
0.5
67 29
0
16
8
73
86
88
0
3
0
Mn(TNPP)Cl
100
12
a Chromatographic yield determined by GC analyses.
b T.O.F. = Turn Over Frequency = moles of converted substrate/(moles
of catalyst × reaction time per minute).
the rate and the chemo- and stereoselectivity of catalytic
reactions (18, 19). An excess of oxidant is generally used,
in order to obtain a full conversion of the substrate. The
catalytic reactions were performed in a biphasic medium,
requiring the use of benzyldimethyltetradecylammonium
chloride as phase transfer agent.
First, one should note that the substrate is fully converted
in all experiments reported in Table 1 indicating that all the
1.2. Potassium monopersulfate as oxidant. Potassium
macrocyclic complexes used as catalysts are stable in the re- monopersulfate is a highly efficient oxygen atom donor.
action medium. The distribution of the oxidation products This inorganic peroxide is a stable triple salt 2KHSO5 ·
obtained depends upon the substituents present on the por- KHSO4 · K2SO4 which has a high capacity to generate
phyrin ring. The best yield in α-pinene oxide 2 was obtained metal–oxo species when added to metalloporphyrins. Since
with the less hindered porphyrin complex, Mn(TPP)Cl potassium monopersulfate is water-soluble, the oxida-
(Table 1, run 4). It is worth noting that when the steric hin- tion reactions were performed in a biphasic system, with
drance is increased by the substitution of the phenyl rings buffered water/dichloromethane in the presence of ben-
of the porphyrin (Table 1, runs 1 to 4), the yield of epox- zyldimethyltetradecylammonium chloride as a phase trans-
ide decreases in favor of the formation of allylic oxidation fer agent. Because a water solution of KHSO5 is very acidic,
products 3 and 4. When bulky substituents are present on a solution of this oxidant in a phosphate buffer at pH 7 was
the porphyrin ring, as for Mn(TMP)Cl (Table 1, run 1), the used in order to avoid the fast acid-catalyzed opening of the
approach of the substrate toward the metal–oxo species is oxirane rings.
tuned by small steric effects between the face of the olefin
When KHSO5 is used in metalloporphyrin-catalyzed ox-
opposite to the substrate methylene bridge and the mesityl idations, it is also known that the addition of pyridine
groups of the porphyrin ring leading to an increase of the derivatives in the reaction mixture increases the rate and
allylic oxidation compound to the epoxide formation. In modifies the oxidation selectivity (22). Consequently, cata-
association with the steric factor, the electronic properties lytic oxidation reactions were performed with and with-
of the substituents on the phenyl rings of the porphyrin out 4-t-butylpyridine. In a classical experiment, 0.5% molar
complexes must also be considered. For example, it is in- of manganese(III) catalyst and 0 or 1.5% molar of 4-t-
teresting to compare runs 2 and 3 of Table 1. The use of butylpyridine were used with regard to the substrate. The
Mn(TDCPP)Cl, bearing two bulky chloro substituents that reactions were generally performed with an excess of oxi-
are moderately electron-withdrawing groups, induces the dant and the results are reported in Table 2.
formation of trans-verbenol 3, but not that of verbenone
Asexpected, onecan firstnotethattheturnoverfrequen-
(Table 1, run 2). In contrast, the use of Mn(TNPP)Cl, cies (TOF) obtained without addition of 4-t-butylpyridine
bearing only one bulky nitro substituent that is strongly are lower than those observed in its presence in the same
electron-withdrawing, induces the formation of verbenone experimental conditions (comparison of runs 1 to 4 with
4, but no verbenol is observed (Table 1, run 3). One can sup- runs 6 to 9). The most remarkable increase of the reaction
pose that the verbenol formed is immediately reoxidized to rate has been obtained in the case of Mn(TPP)Cl (Table 2,
produce verbenone quantitatively. A catalytic experiment runs4and9):thecatalyticoxidationisalmost20timesfaster
realized using verbenol as substrate has confirmed this hy- with 4-t-butylpyridine than without. In the runs 1, 4, 7, and
pothesis, yielding verbenone as the only product.
10, the α-pinene conversion was not complete despite the

352
MARAVAL, ANCEL, AND MEUNIER
TABLE 2
Catalytic Oxidations of α-Pinene Using KHSO5 as Oxidant
Eq. of
oxidant
vs substrate
Yield (%)a
Conversion
(%)
T.O.F.b
Run
Catalyst
t-Bu-Py
2
3
4
(min−1
)
1
2
3
4
5
6
7
8
9
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Cl
−
−
−
−
−
+
+
+
+
+
+
+
1.75
1.75
1.75
1.75
1.1
1.75
1.75
1.75
1.75
1.75
1.1
68
100
100
79
89
100
80
100
100
68
92
100
44
96
50
63
55
34
58
57
100
30
69
94
3
0
0
3
5
52
21
2
3
0
9
2
4
13
0
10
0
5
12
4
0.4
0.7
1.3
0.7
0.7
2.7
1.1
13.3
13.3
1.6
Mn(TNPP)Cl
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Cl
0
10
11
12
Mn(TMP)Clc
Mn(TNPP)Cl
Mn(TPP)Clc
17
10
2
3.1
16.7
1.75
a Chromatographic yield determined by GC analyses.
b
T.O.F. = Turn Over Frequency = moles of converted substrate/(moles of catalyst × reaction time per minute).
c
Use of 0.1% molar of catalyst with regard to the substrate instead of 0.5% molar in the others experiments.
use of an excess of oxidant. We decided in these four runs lar of Mn(TPP)Cl instead of the 0.5% molar generally
to stop these slow experiments before the full substrate used. The turn over frequency obtained in this case was
conversion: so these incomplete conversions must not be above 16 min−1, but by decreasing the catalyst concentra-
attributed to a degradation of the catalyst in the reaction tion the reaction selectivity was slightly reduced (Table 2,
medium.
run 12). Using Mn(TNPP)Cl as catalyst in the presence of
In the experiments realized without 4-t-butylpyridine 4-t-butylpyridine (Table 2, run 8), the TOF value obtained
(Table 2, runs 1 to 4), the distribution between the dif- was above 13 min−1 but in this case, the epoxide selectiv-
ferent oxidation products formed cannot be explained as ity was very weak. Some unidentified oxidation products
clearly as in the reactions performed with NaOCl consid- coming probably from rearrangement of the α-pinene ox-
ering the steric hindrance and the electronic properties of ide were formed in 30% yield. By reducing the excess of
the porphyrin substituents (Table 1). Products other than oxidant used, this phenomenon was eliminated (Table 2,
α-pinene oxide 2, trans-verbenol 3, and verbenone 4 were run 11) but then, the reaction became slower.
formed, making these reactions poorly selective, except
In the experiments performed with 4-t-butylpyridine,
with Mn(TDCPP)Cl as catalyst (Table 2, run 2) where a GC analyses of the reaction mixtures revealed the presence
96% yield in α-pinene oxide was obtained with 100% sub- of a small amount of 4-t-butylpyridine-N-oxide. However,
strate conversion.
it has been previously evidenced that the N-oxide formed
In contrast, in the presence of the pyridine derivative, is not a good oxygen atom donor in our experimental
the same steric and electronic effects as those obtained us- conditions (19).
ing NaOCl were observed. The selectivity in α-pinene ox-
Before ending this study, it is important to specify that the
ide increases when decreasing the steric hindrance on the stability of the different oxidation products was checked in
porphyrin core (Table 2, runs 6 to 9). A 100% yield of 2 the presence of the two oxidants tested: α-pinene oxide 2,
was obtained using the less hindered catalyst Mn(TPP)Cl trans-verbenol 3, and verbenone 4 were all stable. The sta-
(Table 2, run 9). On the other hand, a 65% yield of al- bility of α-pinene was also examined and weak conversions
lylic oxidation products (3 and 4) were observed using the below 10% were observed after six hours with NaOCl or
most bulky metalloporphyrin Mn(TMP)Cl (Table 2, run 6). KHSO5, showing the necessity to use a catalyst in combi-
It is interesting to note the large differences observed on nation with the oxidant to perform the full conversion of
selectivity which are due to the only modification of the the terpene derivatives.
substituents on the phenyl rings of the macrocyclic ligand
2. Catalytic Oxidation of a Nonconjugated Di-Olefin,
using exactly the same experimental conditions, the same
5-Vinyl-2-Norbornene
oxidant, and the same metal. The experiment performed
with Mn(TPP)Cl (Table 2, run 9) is remarkable because its
In the second part of this work, we focused our inter-
selectivity is totally orientated toward the α-pinene oxide, est in the study of a di-olefin oxidation and our choice
but also because this reaction is very fast (TOF > 13 min−1). has been directed toward the 5-vinylbicyclo[2.2.1]hept-2-
Then, a catalytic reaction was performed with 0.1% mo- ene (5-vinyl-2-norbornene as short name). This compound

MANGANESE(III) PORPHYRIN CATALYSTS IN TERPENE OXIDATION
353
TABLE 3
is attractive because it possesses two different types of
double bonds: one endocyclic and the other being exo-
cyclic and terminal. To our knowledge, there are only a
very few examples of catalytic epoxidation systems of 5-
vinyl-2-norbornene. Waegell et al. reported the use of pal-
ladium complexes with molecular oxygen to epoxidize this
substrate. They reported a low and slow conversion of
44% of the 5-vinyl-2-norbornene with the epoxidation of
the endocyclic double bond in 6 days at 60◦C (23, 24).
Heteropolyacids such as 12-tungstophosphoric acid or 12-
molybdophosphoric acid combined with cetylpyridinium
chloride have been reported as catalysts for the epoxidation
of 5-vinyl-2-norbornene in the presence of hydrogen perox-
ide (25). The authors reported a conversion of 90% of the
substrate and a 90% selectivity for the intracyclic epoxide.
Catalytic Oxidations of 5-Vinyl-2-Norbornene
Using NaOCl as Oxidant
Eq. of oxidant Conversion Yield of 6 T.O.F.b
Run
Catalyst
vs substrate
(%)
(%)a
(min−1
)
1
2
3
4
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Clc
1.1
1.1
1.1
1.1
98
85
25
21
98
85
20
21
0.6
0.1
0.1
1.4
a Chromatographic yield determined by GC analyses.
b T.O.F. = Turn Over Frequency = moles of converted substrate/
(moles of catalyst × reaction time per minute).
c Rapid degradation of the catalyst.
tion is formally an electrophilic addition, there is a prefer-
ence for the exo face over the endo face of the molecule. In
addition, the endo hydrogen at C6 and the vinyl substituent
at C5 are viewed as hindering the endo approach of the
electrophile.
2.1. Sodium hypochlorite as oxidant. In all experiments
realized on 5-vinyl-2-norbornene, only a slight excess of
oxidant was used. Our goal was to obtain selectively the
epoxidation of the more reactive norbornene intracyclic
double bond (Fig. 3).
The two hindered catalysts, Mn(TMP)Cl and Mn(TD-
CPP)Cl (Table 3, runs 1 and 2), surprisingly gave very good
substrate conversions. In addition, the catalytic reactions
performed with these two complexes were completely se-
lective and only produced the epoxide 6. This can be ex-
plained by steric considerations: the substrate 5 can only
approach the metal center, which is surrounded by bulky
substituents, by its less hindered exo face, producing then 6
exclusively. Using Mn(TPP)Cl, a rapid bleaching of the or-
ganic phase occurred, corresponding to the degradation of
the catalyst (Table 3, run 4). In contrast, Mn(TNPP)Cl gave
a weak conversion of the substrate but no degradation of
the porphyrin catalyst was observed: the reaction was just
very slow. For example, using Mn(TNPP)Cl as catalyst, the
conversion of the substrate is only 25% in 6.5 h when it is
of 98% in 5 h using Mn(TMP)Cl.
The experimental conditions used were the same as those
for α-pinene oxidation: 0.5% molar of catalyst and 1.5%
molar of 4-t-butylpyridine with respect to the substrate
in a biphasic medium with an ammonium salt as phase
transfer agent. The expected oxidation product, 2,3-epoxy-
5-vinylbicyclo[2.2.1]heptane 6, noncommercially available,
was synthesized as previously described (26). This epoxi-
dation was realized by using potassium monopersulfate, at
0–5◦C with acetone acting both as solvent and reactant. The
reaction occurs under nearly neutral conditions to avoid
epoxide ring opening. This method leads to a mixture of
mono- and diepoxide which can be separated chromato-
graphically. The pure sample of 6 was used as reference in
gas chromatography to analyze the catalytic reaction mix-
tures. The results of our experiments using NaOCl as oxi-
dant are summarized in Table 3.
As expected, the epoxidation reaction occurred regios-
electively on the most reactive intracyclic double bond of
the norbornene derivative, producing nearly exclusively the
compound 6. The reaction is not only regio- but also stere-
oselective, giving only the exo epoxide. This result can be
explained by a combination of steric and electronic effects.
It is known that the exo face of norbornene is somewhat
richer in π electrons than the endo face. Since the epoxida-
2.2. Potassium monopersulfate as oxidant. When using
potassium monopersulfate as oxidant, the catalytic reac-
tions were generally faster than with sodium hypochlorite
(Table 4).
As observed with NaOCl, Mn(TMP)Cl, and Mn(TD-
CPP)Cl (Table 4, runs 1 and 2) were the more active and
the more selective catalysts and the unique oxidation
product observed was the 2,3-epoxy-5-vinylnorbornane 6.
Mn(TNPP)Cl (Table 4, run 3) was less active and less se-
lective, but this complex was stable under the reaction con-
ditions, contrary to Mn(TPP)Cl (Table 4, run 4) which was
quickly degraded in the medium.
3. Catalytic Oxidation of a Conjugated
Di-Olefin, α-Terpinene
To complete this comparative study, we decided to use a
conjugated diene, namely α-terpinene. Using only a slight
excess of oxidant, the selective formation of one of the two
FIG. 3. Catalytic epoxidation of the intracyclic double bond of the
5-vinyl-2-norbornene 5.

354
MARAVAL, ANCEL, AND MEUNIER
TABLE 4
In GC analyses of the reaction mixtures, the chromato-
graphic yield of 8 is the only one that was determined, the
other products being formed in small amounts (see Table 5).
With potassium monopersulfate and sodium hypochlo-
rite, a complete conversion of the substrate was obtained re-
gardless of the metalloporphyrin catalyst used. In the same
way as with the two precedently studied substrates, the cata-
lytic reactions were always more rapid with KHSO5 than
with NaOCl. Independant catalytic reactions indicated that
the yield of 8 ranged from 48 to 72% at full substrate con-
version. Nevertheless, GC/MS analyses of the reaction mix-
tures indicated that no epoxides and no oxygenation prod-
uctsofthesubstrateatm/z152wereformed. Incontrast, two
dehydrogenation compounds at m/z 134 were analyzed by
GC/MS. The major product was p-cymene and the second
one, less abundant, was identified as p-mentha-1,3,8-triene
9. In the case of p-cymene, a weight loss of 43 mass units cor-
responding to an isopropyl fragment was observed on mass
spectra, giving a peak at m/z 91. In the case of p-mentha-
1,3,8-triene, a weight loss of 41 mass units was obtained,
corresponding to a methylethenyl fragment, giving a signal
at m/z 93.
Catalytic Oxidations of 5-Vinyl-2-Norbornene
Using KHSO5 as Oxidant
Eq. of oxidant Conversion Yield of 6 T.O.F.b
Run
Catalyst
vs substrate
(%)
(%)a
(min−1
)
1
2
3
4
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Clc
1.1
1.1
1.1
1.1
99
94
76
44
99
94
65
43
1.1
1.0
0.4
0.5
a Chromatographic yield determined by GC analyses.
b T.O.F. = Turn Over Frequency = moles of converted substrate/
(moles of catalyst × reaction time per minute).
c Rapid degradation of the catalyst.
possible epoxides was expected. Monoepoxides are the
only products observed in the catalytic oxidation of 1,3-
cyclohexadiene using the NaOCl/Mn(TPP)OAc system
(19).
In the litterature, the catalytic photooxidation of α-
terpinene has been described using a TiO2 suspension
in acetonitrile under irradiation (27). In these experi-
mental conditions, a mixture of products is obtained, in-
cluding ascaridole resulting from the photooxygenation
reaction, isoascaridole (bis-epoxide) and para-cymene aris-
ing from the oxidative dehydrogenation reaction. Using a
molybdenum complex, with hydrogen peroxide and sodium
hydroxide, Woodward et al. described the selective oxida-
tion of α-terpinene in ascaridole (28). The catalytic ox-
idative dehydrogenation of α-terpinene to p-cymene was
also described using vanadium and molybdenum mixed
heteropolyanions in the presence of oxygen (29) or with
t-butylhydroperoxide (30).
All the experiments with α-terpinene 7 were realized
with 0.5% of catalyst and 1.5% of 4-t-butylpyridine with
respect to the substrate. The reactions were performed un-
der biphasic conditions using a slight excess of oxidant. In
all the catalytic experiments realized, no epoxidation prod-
ucts were observed. The major compound obtained was the
p-cymene 8, resulting in an oxidative dehydrogenation of
the substrate (Fig. 4).
For the oxidative dehydrogenation catalyzed by
=
Mn(V) O species, we propose a mechanism similar to
that previously described for cytochrome P450 with the
initial abstraction of an H-atom on the ring by porphyrin-
=
Mn(V) Ospeciesleadingtoanintermediateradical(Fig. 5,
for the generation of high-valent metal–oxo species in
P450 enzymes and related biomimetic models, see Ref. 32).
Then, the C-centered radical is oxidized by the Mn(IV)-
OH species generated by the H-atom transfer. The loss of a
β-proton from the resulting carbocation intermediate gen-
=
erates the new C C bond.
The yield of 9 was low and estimated to be below 3%.
But, if no oxygenation products of the substrate are formed,
TABLE 5
Catalytic Oxidations of α-Terpinene
Conversion Yield of 8 T.O.F.c
Run
Catalyst
Oxidanta
(%)
(%)b
(min−1
)
To our knowledge, only a few examples of oxidative de-
hydrogenation catalyzed by metalloporphyrins have been
previously described. Desaturation of aliphatic hydrocar-
bons have been reported with cytochrome P450 enzymes
as catalysts (31). No example of manganese(III) porphyrin
catalyzed oxidative dehydrogenation was found in the
litterature.
1
2
3
4
5
6
7
8
Mn(TMP)Cl
Mn(TDCPP)Cl NaOCl
NaOCl
100
100
100
96
100
100
96
54
54
66
72
49
1.3
1.3
1.3
1.3
2.7
4.4
12.8
12.7
Mn(TNPP)Cl
Mn(TPP)Cl
Mn(TMP)Cl
Mn(TDCPP)Cl
Mn(TNPP)Cl
Mn(TPP)Cl
NaOCl
NaOCl
KHSO5
KHSO5
KHSO5
KHSO5
48
57(49)d
65(47)d
95
a In all the experiments, 1.1 equivalents of oxidant versus the substrate
are used.
b Chromatographic yields determined by GC analyses.
c T.O.F. = TurnOverFrequency = molesofconvertedsubstrate/(moles
of catalyst × reaction time per minute).
FIG. 4. Catalytic oxidative dehydrogenation of the α-terpinene 7.
d Yields observed 30 minutes after the end of the conversion.

MANGANESE(III) PORPHYRIN CATALYSTS IN TERPENE OXIDATION
355
how can we explain that the yield of p-cymene is only about
50%? In the runs 7 and 8 of Table 5, the conversion of
the substrate was almost complete; so the oxidant has been
quasi totally consumed because only a slight excess was
used. After a reaction time of 15 minutes, the yield of 8
was 57% using Mn(TNPP)Cl (Table 5, run 7) and 65% us-
ing Mn(TPP)Cl (Table 5, run 8). When stirring was main-
tained, after a period of 30 minutes, we observed that the
yield of 8 decreased to 49% for the reaction performed
with Mn(TNPP)Cl and to 47% with Mn(TPP)Cl. This phe-
nomenon can not be explained by a catalytic oxidation reac-
tion requiring a stoechiometric amount of sodium monop-
ersulfate since the oxidant excess was too weak (with only
1.1 equivalent of oxidant the substrate was fully converted).
Then, we propose that a catalytic amount of Mn(V) = O
species in the medium was able to initiate an autocatalytic
free radical chain reaction partially consuming the gener-
ated p-cymene 8. This free radical chain reaction can lead
to several products and its presumed mechanism is depicted
in Fig. 6.
In this mechanism, a catalytic amount of Mn(V) = O
species initiates the reaction, inducing the abstraction of a
hydrogen atom from 8. The alkyl radical formed can react
with oxygen to give the corresponding alkylperoxy radi-
cal generating the alkoxy radical 10 after dimerization and
–
Russell fragmentation. The homolytic cleavage of the C C
bond between the ring and the quaternary carbon of 10
(β fragmentation) leads to the formation of acetone and p-
cresol 11 as previously described in the case of cumene (33).
The presence of p-cresol was not detected in the GC/MS
FIG. 5. Proposed mechanism for the oxidative dehydrogenation re-
action of the α-terpinene 7. It should be noted that the abstraction on the
other allylic position (on the side of the methyl group) is also possible.
Only one route is depicted.
FIG. 6. Proposed mechanism for the autocatalytic free radical chain reaction involving the p-cymene 8.

356
MARAVAL, ANCEL, AND MEUNIER
substituents on the selectivity of the catalytic oxidation re-
action (epoxide formation versus allylic oxidation).
• Regio- and stereoselective epoxidation is observed in
the case of the 5-vinyl-2-norbornene (nonconjugated di-
ene). The intracyclic double bond is the only one to be
epoxidized.
• An original oxidative dehydrogenation reaction takes
place in the case of the α-terpinene (conjugated diene).
Such hydrocarbon desaturation is an unusual oxidation
pathway when using metalloporphyrin catalysts.
FIG. 7. Catalytic oxidation reaction of the 1,2-dihydronaphthalene 14.
analyses of the reaction mixtures. In fact, compound 11 is
unstable under the oxidative reaction conditions and prob-
ably gives the corresponding quinone and polymers. The in-
termediate alkoxy radical 10 also produced the correspond-
ing alcohol 12 which was observed in GC/MS at m/z 150.
The mass spectra of this compound exhibited a fragmenta-
tion peak at m/z 91 corresponding to the elimination of the
2-propylalcohol fragment. A further oxidation of this prod-
uct 12 produced traces of the diol 13 which was observed at
m/z 166 in GC/MS.
Thisfreeradicalchainoxidationexplainsthelimitedyield
in p-cymene observed in these biomimetic reactions and
also the formation of the different hydroxylated products
arising from its further oxidation. A blank experiment has
been performed using p-cymene 8 as substrate leading to
the formation of the same oxidation products 12 and 13.
In order to complete this study, a catalytic experiment was
performed with the 1,2-dihydronaphthalene 14 as substrate
(Fig. 7).
ACKNOWLEDGMENTS
The financial support of CNRS and Aventis is gratefully acknowledged.
V.M. is indebted to Aventis-Nutrition Animale for a post-doctoral fellow-
ship.
REFERENCES
1. (a) Meunier, B., “Biomimetic Oxidations Catalyzed by Transition
Metal Complexes,” Imperial College Press, 2000; (b) Meunier, B.,
Robert, A., Pratviel, G., and Bernadou, J., in “The Porphyrin Hand-
book,” Vol. 4, p. 119. Academic Press, San Diego, 1999.
2. Groves, J. T., Nemo, T. E., and Myers, R. S., J. Am. Chem. Soc. 101,
1032 (1979).
3. Guilmet, E., and Meunier, B., Tetrahedron Lett. 4449 (1980).
4. Meunier, B., New J. Chem. 16, 203 (1992).
5. Battioni, P., Renaud, J. P., Bartoli, J. F., and Mansuy, D., J. Chem. Soc.,
Chem. Commun. 888 (1985).
6. Yuan, L. C., and Bruice, T. C., J. Am. Chem. Soc. 108, 1643
(1986).
In this case, the oxidative dehydrogenation reaction com-
petes with the epoxidation reaction. Substrate 14 reacts in
the same way as styrene, which gives the epoxidation prod-
uct on the vinyl group in high yield (22). In contrast with
α-terpinene 7, substrate 14 already possesses an aromatic
ring, thus the epoxidation of the nonaromatic double bond
becomes the major reaction (65% of 16). The naphthalene
15 arising from the dehydrogenation reaction was obtained
in 30% yield; 15 was stable in the oxidative conditions, in
contrast with p-cymene 8. This difference of stability can
be explained by the fact that 15 has no substituents that can
undergo hydrogen abstraction in opposition to p-cymene
which possesses hydrogen atoms in two different allylic
positions.
7. Hoffmann, P., Robert, A., and Meunier, B., Bull. Soc. Chim. Fr. 129,
85 (1992).
8. Adler, A. D., Longo, F. R., Kampas, F., and Kim, J., J. Inorg. Nucl.
Chem. 32, 2443 (1970).
9. Robert, A., Momenteau, M., Loock, B., and Meunier, B., Inorg. Chem.
30, 706 (1991).
10. Lajunen, M. K., and Koskinen, A. M. P., Tetrahedron Lett. 35, 4461
(1994).
11. Games, M. F. T., and Antures, O. A. C., Catal. Lett. 42, 213 (1996).
12. Villa de P., A. L., De Vos, D. E., Montes de C., C., and Jacobs, P. A.,
Tetrahedron Lett. 39, 8521 (1998).
13. Villa de P., A. L., Sels, B. F., De Vos, D. E., and Jacobs, P. A., J. Org.
Chem. 64, 7267 (1999).
14. McMorn, P., Roberts, G., and Hutchings, G. J., Catal. Lett. 67, 203
(2000).
15. Arnold, U., Sepa da Cruz, R., Mandelli, D., and Schuchardt, U., J. Mol.
Catal. A 165, 149 (2001).
16. De Carvalho, M. E., and Meunier, B., Tetrahedron Lett. 24, 3621
(1983).
CONCLUSION
17. De Carvalho, M. E., and Meunier, B., New J. Chem. 10, 223 (1986).
18. Meunier, B., Chem. Rev. 92, 1411 (1992).
19. Meunier, B., Guilmet, E., De Carvalho, M. E., and Poilblanc, R.,
J. Am. Chem. Soc. 106, 6668 (1984).
20. Lajunen, M. K., Maunula, T., and Koskinen, A. M. P., Tetrahedron 56,
8167 (2000).
21. Lajunen, M. K., J. Mol. Catal. A 169, 33 (2001).
22. Robert, A., and Meunier, B., New J. Chem. 12, 885 (1988).
23. Heumann, A., Chauvet, F., and Waegell, B., Tetrahedron Lett. 23, 2767
(1982).
24. Chauvet, F., Heumann, A., and Waegell, B., J. Org. Chem. 52, 1916
(1987).
Manganese(III) porphyrin complexes are efficient oxi-
dation catalysts for the three studied terpene derivatives.
This comparative study allowed us to illustrate the differ-
ent oxidation products that can be generated using three
differently substituted metalloporphyrin catalysts:
• Epoxidation and allylic oxidation reactions were ob-
served with α-pinene (mono-unsaturated substrate). The
experiments reported with this terpene derivative allowed
us to evidence the strong influence of the metalloporphyrin

MANGANESE(III) PORPHYRIN CATALYSTS IN TERPENE OXIDATION
357
25. Ishii, Y., Yamawaki, K., Ura, T., Yamada, H., Yoshida, T., and Ogawa,
M., J. Org. Chem. 53, 3587 (1988).
30. Hill, C. L., Zeng, H., and Zhang, X., J. Mol. Catal. A 113, 185
(1996).
26. Crivello, J. V., and Narayan, R., Macromolecules 29, 433 (1996).
27. Fox, M. A., Sackett, D. D., and Younathan, J. N., Tetrahedron 43, 1643
(1987).
28. Backhouse, J. R., Lowe, H. M., Sinn, E., Suzuki, S., and Woodward, S.,
J. Chem. Soc., Dalton Trans. 1489 (1995).
31. Rettie, A. E., Rettenmeier, A. W., Howald, W. N., and Baillie, T. A.,
Science 235, 890 (1987).
32. Meunier, B., and Bernadou, J., Struct. Bonding 97, 1 (2000).
33. (a) Hock, H., and Lang, B., Chem. Ber. 77, 257 (1944). (b) Sheldon,
R. A., and Kochi, J. K., in “Metal-Catalyzed Oxidations of Organic
Compounds,” p. 18. Academic Press, San Diego, 1981.
29. Neumann, R., and Levin, M., J. Am. Chem. Soc. 114, 7278 (1992).