Salen manganese (III) complexes as catalysts for R-(+)-limonene oxidation

Article abstract of DOI:10.1016/j.apcata.2009.10.041

The epoxidation of R-(+)-limonene using in situ generated dimethyldioxirane (DMD) as the oxidizing agent and four Jacobsen-type catalysts ((R,R)-Jacobsen, (S,S)-Jacobsen, racemic Jacobsen and achiral Jacobsen) was examined. The effect of the amount of KHSO₅ and acetone in the catalyzed and un-catalyzed reaction was also assessed. The main reaction products were diepoxide and endocyclic monoepoxide. In the absence of catalyst, the amount of KHSO₅ did not significantly influence conversion and selectivity. The catalyst can be segregated to a different phase and separated from the reaction media when the amount of KHSO₅ is above the stoichiometric ratio, R-(+)-limonene/KHSO₅ = 0.5 mmol/mmol, and acetone/mmol R-(+)-limonene = 2 mL/mmol. However, when the amount of KHSO₅ is below the stoichiometric ratio (R-(+)-limonene/KHSO₅ = 1.5 mmol/mmol) the catalyst is difficult to separate. Under the reaction conditions of this study, when the catalyst is segregated, no effect of the catalyst chiral center, (R,R)-Jacobsen or (S,S)-Jacobsen, was found on conversion and selectivity. Additionally, the (R,R)-Jacobsen's catalyst proved to be very stable to oxidative degradation.

Full text of DOI:10.1016/j.apcata.2009.10.041

Applied Catalysis A: General 373 (2010) 57–65
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Salen manganese (III) complexes as catalysts for R-(+)-limonene oxidation
*
´
Jairo Cubillos , Santiago Vasquez, Consuelo Montes de Correa
Environmental Catalysis Research Group, Universidad de Antioquia, Apartado Ae´reo 1226, Cra 53 No. 61-30, Medellı´n, Colombia
A R T I C L E I N F O
A B S T R A C T
Article history:
The epoxidation of R-(+)-limonene using in situ generated dimethyldioxirane (DMD) as the oxidizing
agent and four Jacobsen-type catalysts ((R,R)-Jacobsen, (S,S)-Jacobsen, racemic Jacobsen and achiral
Jacobsen) was examined. The effect of the amount of KHSO5 and acetone in the catalyzed and un-
catalyzed reaction was also assessed. The main reaction products were diepoxide and endocyclic
monoepoxide. In the absence of catalyst, the amount of KHSO5 did not significantly influence conversion
and selectivity. The catalyst can be segregated to a different phase and separated from the reaction media
when the amount of KHSO5 is above the stoichiometric ratio, R-(+)-limonene/KHSO5 = 0.5 mmol/mmol,
and acetone/mmol R-(+)-limonene = 2 mL/mmol. However, when the amount of KHSO5 is below the
stoichiometric ratio (R-(+)-limonene/KHSO5 = 1.5 mmol/mmol) the catalyst is difficult to separate.
Under the reaction conditions of this study, when the catalyst is segregated, no effect of the catalyst
chiral center, (R,R)-Jacobsen or (S,S)-Jacobsen, was found on conversion and selectivity. Additionally, the
(R,R)-Jacobsen’s catalyst proved to be very stable to oxidative degradation.
Received 9 May 2009
Received in revised form 18 October 2009
Accepted 24 October 2009
Available online 30 October 2009
Keywords:
R-(+)-limonene
Epoxides
Jacobsen-type catalysts
Homogeneous catalyst
Recoverable catalyst
ß 2009 Elsevier B.V. All rights reserved.
1. Introduction
with a biological receptor, mainly due to the orientation of the
substituents [5]. This is how a diasteromer can be effective in the
Limonene derivatives functionalized with oxygen are impor-
tant intermediates in the preparation of pharmaceuticals, food
additives, and agrochemicals [1]. Fig. 1 shows the main products
from the oxidation of limonene which are classified in two groups.
The first group comprises products from the epoxidation of the
double bonds denoted as epoxides [2]. Additionally, epoxides can
be classified as monoepoxides if a double bond is oxidized and
diepoxides if both double bonds are oxidized. If the double bond
between carbons 1 and 2 is oxidized, the monoepoxide corre-
sponds to the endocyclic epoxide. If the double bond between
carbons 8 and 9 is oxidized, the resulting monoepoxide is the
exocyclic epoxide [2]. The second group corresponds to allylic
oxidation products, which include carveol and carvone [2].
Due to the existence of a chiral center in the chemical structure
of limonene (carbon 4, Fig. 1), this compound has two enantiomers
with optical configurations R and S [3]. Orange peels are the main
source of R-(+)-limonene, while lime peels mostly contain S-(–)-
limonene [4]. Either limonene enantiomer has eight optically pure
epoxides [4]. Fig. 2 shows the chemical structure of these epoxides
and their corresponding optical configuration. It is important to
note that each monoepoxide has two isomers (cis/trans) known as
diastereomers. Just like any pair of enantiomers, the cyclic
diasteromers cis/trans show different behavior when interacting
treatment of an illness; while its counterpart may be inactive or
even produce opposite effects. For this reason, there is much
interest on the development of efficient processes for obtaining
optically pure epoxides [6].
The Jacobsen–Katsuki epoxidation is used in the production of
optically pure epoxides [7]. The success of this method is based on
the use of manganese (III)–salen complexes, as catalysts. Fig. 3
shows the chemical structure of the Jacobsen’s catalyst [7]. The
most common solvents used are dichloromethane and acetonitrile
and the most used oxidizing agents are sodium hypochlorite,
iodosylbenzene, and meta-chlorobenzoic acid. In spite of the
satisfactory results achieved with this method (enantiomeric
excess ꢀ95%), no industrial process still exists because reaction
products and catalyst are in the same phase and are difficult to
separate by physical–mechanical methods [8]. Therefore, various
methods for the heterogenization of Jacobsen–Katsuki type
catalysts have been proposed. The immobilization of the catalyst
on the surface of a solid (support) by covalent chemical bond is
mostly used [9]. With this method, the complete separation of the
catalyst from the reaction products has been proved. However, no
successful immobilization method has been established that keeps
the catalytic properties of the homogeneous catalyst or the
preservation of the initial catalytic activity of the immobilized
catalyst [10].
It is generally recognized that the solid support prevents the
geometric reception of the active catalytic species attained in a
* Corresponding author. Tel.: +57 4 2196607; fax: +57 4 2196609.
E-mail address: jclobo@udea.edu.co (J. Cubillos).
homogeneous phase, which is interpreted as
a poor chiral
0926-860X/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2009.10.041

58
J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
Fig. 1. Main oxygenated products from limonene.
Fig. 2. Epoxides derived from R-(+)-limonene: (a) monoepoxides; (b) diepoxides.
communication between the catalyst and the olefin [11]. On the
other hand, it has been established that oxidizing agents such as
sodium hypochlorite and iodosylbenzene promote the deactiva-
tion of the catalyst because of the oxidative degradation of the
salen ligand [11]. To overcome these difficulties, the use of
dimethyldioxirane (DMD) generated in situ as an oxidizing agent
improves the stability of the catalyst because of the moderate
oxidation conditions that can be attained [12,13]. Additionally, the
catalyst initially soluble in the reaction medium, can be segregated
depending on reaction conditions.
In this paper we report on the asymmetric oxidation of R-(+)-
limonene using in situ generated DMD as oxidizing agent in the
Fig. 3. Jacobsen asymmetric epoxidation.

J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
59
presence of Jacobsen-type catalysts. Reaction conditions affect
product selectivity as well as catalyst recoverability. The catalyst
was easily recovered when the ratio R-(+)-limonene/KHSO5 = 0.5
and the ratio acetone/R-(+)-limonene was 2 mL/mmol. Under
these conditions, the chiral center of the catalyst did not appear to
influence catalytic activity. Additionally, the catalyst was success-
fully reused three times suggesting improvement of catalyst
stability to oxidative degradation.
with 100 mL of water at 5 8C and with 100 mL of methanol for
about five times. Next, the solid was dried, first in flowing air for 1 h
and then under vacuum at 40 8C.
The salen ligand ((R,R)-Jacobsen’s ligand) was obtained as
follows. In a 100 mL round-bottom vessel 1.11 g of (R,R)-1,2-
diaminocyclohexane mono-(+)-tartrate, 1.16 g of potassium car-
bonate and 6.0 mL of water was added. The resulting mixture was
magnetically stirred until complete dissolution. Then, 22 mL of
ethanol was added and the turbid mixture heated under reflux.
Next, a solution of 2.0 g of 3,5-di-tert-butyl-2-hidroxy-benzalde-
hyde dissolved in 10 mL of ethanol was added through the
condenser with a Pasteur pipette, and the pipette and the beaker
rinsed with 2.0 mL of hot ethanol. After heating under reflux for
1 h, 6 mL of water was added, the mixture allowed to cool and then
introduced into an ice bath for 30 min. The yellow solid was
recovered by filtration and washed with 5 mL of ethanol. This solid
was dissolved in 25 mL of dichloromethane and the solution
washed twice with 5 mL of water followed by 5 mL of a saturated
saline solution. The organic layer was dried with sodium sulfate
and the solvent removed by rotary evaporation to obtain a purified
yellow solid.
2. Experimental
2.1. Catalyst synthesis
2.1.1. Preparation of (R,R)-Jacobsen’s catalyst
The preparation of this catalyst is divided into three steps [14].
The first step is the racemic resolution of the optically active amine.
In the second step, the amine previously obtained reacts with 3,5-
di-tert-butyl-2-hidroxy-benzaldehyde to form the salen ligand. In
the last step, the salen ligand is treated with a manganese source
under an oxidizing atmosphere to produce (R,R)-Jacobsen’s
catalyst. The synthesis procedure is illustrated in Fig. 4.
The resolution of the optically active amine ((R,R)-1,2-
Finally, the (R,R)-Jacobsen’s catalyst was prepared as follows:
1.0 g of (R,R)-Jacobsen’s ligand and 25 mL of absolute ethanol was
added to a 100 mL round-bottom vessel. After heating and
magnetically stirring this mixture under reflux (105 8C) for
20 min, 2.0 equiv. of Mn(OAc)2ꢁ4H2O was added and the reflux
continued for 30 min. Afterwards, a low flow of air was bubbled
through the reaction mixture for 1 h in order to oxidize Mn2+ to
diaminocyclohexane
procedure. 150 g of
were mixed in
L-(+)-tartrate) was obtained by the following
-(+)-tartaric acid and 400 mL of distilled water
beaker with continuous stirring at room
L
a
temperature until complete dissolution. Next, 240 mL of a mixture
of cis/trans 1,2-diaminocyclohexane was added at a speed such
that the reaction temperature did not exceed 70 8C. Then, 100 mL
of glacial acetic acid was added at a speed such that the reaction
temperature did not exceed 90 8C. After the addition of acetic acid,
a white precipitate formed. This mixture was stirred vigorously for
2 h without heating. Then, the precipitate was filtered, washed
Mn3+
. Then, 3 equiv. of LiCl was added to neutralize the
organometallic cationic complex. After heating and stirring for
30 additional minutes, the solvent was removed by rotary
evaporation. The product was dissolved in 25 mL of dichloro-
Fig. 4. Preparation of (R,R)-Jacobsen’s catalyst.

60
J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
Fig. 7. Chemical structure of Jacobsen’s achiral catalyst.
Fig. 5. Chemical structure of (S,S)-Jacobsen’s catalyst.
2.1.2. Preparation of (S,S)-Jacobsen’s catalyst
The (S,S)-Jacobsen’s catalyst and (S,S)-Jacobsen’s ligand were
prepared following the same procedure described above, but
D-
(ꢂ)-tartaric acid was used instead of
L
-(+)-tartaric acid. Fig. 5
shows the chemical structure of the (S,S)-Jacobsen’s catalyst.
2.1.3. Preparation of the Jacobsen’s racemic catalyst
The Jacobsen’s racemic catalyst and Jacobsen’s racemic ligand
were prepared following a procedure similar to that of the pure
enantiomeric catalyst but, the first step was not performed.
Fig. 6 shows the chemical structure of the Jacobsen’s racemic
catalyst.
Fig. 6. Chemical structure of Jacobsen’s racemic catalyst.
methane, washed twice with water, and once with a saturated
saline solution. After drying the organic phase with sodium sulfate,
30 mL of heptane was added as the crystallization solvent.
Dichloromethane was removed by rotary evaporation and the
brown-colored mixture cooled in an ice bath for 30 min, to obtain
the (R,R)-Jacobsen’s catalyst.
2.1.4. Preparation of Jacobsen’s achiral catalyst
Jacobsen’s achiral catalyst and Jacobsen’s achiral ligand were
prepared following a procedure similar to that used for the
Jacobsen’s racemic catalyst. In this case, 1,2-diaminoethane was
used instead of 1,2-diaminocylohexane. Fig. 7 shows the chemical
structure of the Jacobsen’s achiral catalyst.
Fig. 8. Asymmetric oxidation of R-(+)-limonene with Jacobsen-type catalysts.

J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
61
Fig. 9. In situ formation of DMD from KHSO5 and acetone [15].
2.2. Catalyst characterization
0.08 MPa) in order to separate the catalyst from the reaction
products. When the catalyst was recoverable (Fig. 16b), the solid
and liquid phases were separated by filtration and/or centrifuga-
tion. The solid was washed with large amounts of deionized water
in order to remove inorganic salts produced from Oxone1 and
buffer solution. In both cases, the liquid mixture free of catalyst
was concentrated under vacuum and the reaction products
analyzed in a gas chromatograph coupled to a mass spectrometer
(Agilent Technologies 5975C GC–MS), and equipped with a FID and
Catalyst samples were characterized by infrared spectroscopy
in a FT-IR Nicolet Avatar 330 equipped with a dispersion cell.
Samples were diluted with potassium bromide to obtain a solid
mixture of approximately 3% (w/w). The spectra were registered in
the aromatic region 1200–1800 cmꢂ1
. UV–vis spectra were
recorded in the range 200–700 nm in a Lamda 4B PerkinElmer
spectrophotometer equipped with a diffuse reflectance attach-
ment, using BaSO4 as reference. Differential thermogravimetric
analysis (DTA) curves were obtained on a TGA 2950 thermal
analyzer. Samples were heated from room temperature up to
700 8C under flowing air using alumina sample holders. The sample
weight was ca. 10 mg and the heating rate was 5 8C/min. A
reference sample holder contained alumina. The bulk Mn content
was determined by atomic absorption spectroscopy (AAS) in a
Model S4 Thermo Electron Corporation spectrometer. Typically
one sample of 20 mg of solid, previously dried at 100 8C, was mixed
with 2 cm3 of HF and 2 cm3 of aqua regia. Finally, the sample was
adjusted to a known volume (100 mL) with deionized water.
a capillary column Beta-dex GTA (60 m length, 250
mm internal
diameter and 0.25 m film thickness). The reactions involved are
m
illustrated in Fig. 8. The in situ formation of DMD from the reaction
between KHSO5 and acetone in a slightly basic medium [15] is
illustrated in Fig. 9.
3. Results and discussion
3.1. Catalyst characterization
Fig. 10 shows the FT-IR spectra of the salen ligands and their
corresponding catalysts. The salen ligands show a characteristic
band around 1640 cmꢂ1, which is associated with the vibrations of
the imine group (HC55N) [16]. In the catalyst samples, this band is
displaced towards a lower wavelength (1620 cmꢂ1) as the first
evidence of the formation of the organometallic complex.
Additionally, characteristic bands at 1540 (C–O), 575 (Mn–O)
2.3. Catalytic tests
The catalytic activity of the synthesized catalysts was examined
on the asymmetric oxidation of R-(+)-limonene using in situ
generated dimethyldioxirane (DMD) as the oxidizing agent. The pH
control during the catalytic tests is important since the formation
of DMD by reaction between acetone and KHSO5 takes place in a
narrow pH range (8.0–8.5). pH measurements were performed
with a Metrohom 691 pH meter. Reproducibility of the pH
measurements was assured by calibrating the pH meter before
measurements, using standard buffer solution of pH 4.0 and 7.0. In
a standard procedure, 2.0 mmol of R-(+)-limonene, 1.2 mmol of
sodium bicarbonate and 0.05 mmol of catalyst were dissolved in
the required amount of acetone (mixture A). The pH was difficult to
measure due to the organic nature of this liquid mixture. For this
reason, a buffer solution (aqueous NaHCO3, 5 wt.%) was added to
bring the pH in the range between 8.0 and 8.5. In another vessel,
the required amount of KHSO5 (Oxone1) was dissolved in 4–8 mL
of water (mixture B). While mixture A was being stirred, mixture B
was slowly added, keeping the pH in the range 8.0–8.5 using
NaHCO3 solution (5% (w/w) aqueous). When mixture B was
completely added, stirring was stopped and the formed solids
separated by filtration and/or centrifugation. In homogeneous
phase experiments, 30 mL of dichloromethane was added to the
reaction mixture obtaining two phases. The aqueous phase was
discarded and the organic phase was vacuum distilled (160 8C and
Fig. 10. FT-IR spectra of (R,R)-Jacobsen’s ligand (a), (R,R)-Jacobsen’s catalyst (b),
(S,S)-Jacobsen’s ligand (c), (S,S)-Jacobsen’s catalyst (d), Jacobsen’s racemic ligand
(e), Jacobsen’s racemic catalyst (f), Jacobsen’s achiral ligand (g) and Jacobsen’s
achiral catalyst (h).

62
J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
Table 1
Chemical and thermal analysis of Jacobsen catalysts.
Catalyst
Mn loadinga
(wt.%)
Ligand loadingb
(wt.%)
Mn/ligandc
(R,R)-Jacobsen
(S,S)-Jacobsen
7.5
7.9
7.9
9.0
76
80
81
81
0.98
0.98
0.96
0.94
Jacobsen racemic
Jacobsen achiral
a
Determined by AAS.
b
Determined by TGA–DTA.
Molar ratio.
c
Table 1 collects the salen ligand loading of the four Mn(III)–salen
complexes.
The manganese contents of the Mn(III)–salen complexes are
summarized in Table 1. The Mn loadings are in the range 7.5–
9.0 wt.% Mn elemental analysis of the Mn(III)–salen complexes in
combination with TGA–DTA suggests that neither free ligand nor
free manganese appear to be present in the samples, because the
Mn/ligand molar ratio is close to 1.
Fig. 11. DR UV–vis spectra of (R,R)-Jacobsen’s ligand (a), (R,R)-Jacobsen’s catalyst
(b), (S,S)-Jacobsen’s ligand (c), (S,S)-Jacobsen’s catalyst (d), Jacobsen’s racemic
ligand (e), Jacobsen’s racemic catalyst (f), Jacobsen’s achiral ligand (g) and
Jacobsen’s achiral catalyst (h).
and 492 cmꢂ1 (Mn–N) are also associated with the complexation of
manganese by the salen ligand [16].
3.2. Catalytic activity
Fig. 11 shows DR UV–vis spectra of the salen ligands and their
corresponding catalysts. The salen ligands exhibit absorption bands
The choice of oxygen donor and solvent is crucial in the catalytic
oxidation of olefins. It is known that the best solvents for the
Jacobsen epoxidation are: dichloromethane, acetonitrile and
toluene when NaOCl, PhIO and m-CPBA are used as oxidizing
agents [8]. However, the use of these oxidizing agents leads to the
oxidative degradation of the catalyst [8]. On the other hand, the
formation of in situ dioxiranes requires a ketone derivative. In the
case of in situ DMD, acetone is used [15]. We conducted the test
reaction using dichloromethane, acetonitrile, toluene and equi-
molar mixtures of these together with KHSO5 as oxygen source, but
the reaction was poor and the catalyst was degraded by action of
KHSO5. In contrast, when acetone is used, KHSO5 does not act as
oxidizing agent. Instead, KHSO5 reacts with acetone in a slightly
basic medium to give DMD (Fig. 9). Thus, DMD as oxidizing agent
delivers the active oxygen towards the metal center of the catalyst
to give oxo-manganese (V) moieties and then it transfers the
atomic oxygen to the olefinic double bond. Therefore, KHSO5
deactivates the catalyst when it acts as oxidizing agent. On the
other hand, DMD improves catalyst stability. This finding can be
explained by the following two facts. First, DMD reacts with the
catalyst as it is formed so its life time is short. Second, KHSO5 reacts
preferentially with acetone instead of the catalyst.
The asymmetric oxidation of R-(+)-limonene using in situ
generated DMD was first examined in the absence of catalyst.
Fig. 11 shows the results for two different ratios of KHSO5; below
the stoichiometric amount (R-(+)-limonene/KHSO5 = 1.5) and
above the stoichiometric amount (R-(+)-limonene/KHSO5 = 0.5).
In these experiments, the acetone/R-(+)-limonene ratio was
maintained at 2 mL/mmol. As can be observed in Fig. 13 in the
absence of catalyst there is a moderate conversion of R-(+)-
limonene. Diepoxides were the main reaction products and
endocyclic monoepoxides were the secondary products. No
exocyclic epoxides were detected. Fig. 13 also suggests that there
is no significant effect of the amount of KHSO5 on reaction
parameters when the reaction occurs without catalyst.
at 265 nm and 335 nm. These bands are attributed to
transitions. The band at 265 nm has been assigned to the benzene
ring and the one at 335 nm, to the imino groups [17]. The imino
p
!
p*
p
!
p* transitions in the Mn salen complexes is shifted to larger
wavelengths due to metal coordination, confirming the formation of
Mn(III)–salen complex [17]. UV–Vis and FT-IR spectra revealed that
the salen ligands are unaffected and are not decomposed upon
coordination of the organo-funtional groups with manganese.
Thermogravimetric analysis (TGA–DTA) has been used to
monitor the decomposition profiles of all materials and to evaluate
the ligand loading of Mn(III)–salen complexes. The DTA curves of
Jacobsen’s achiral ligand and Jacobsen’s achiral catalyst materials
are shown in Fig. 12. The salen ligand exhibits one main step of
weight loss when heated under flowing air [18]. The main weight
loss (ca. 93.1%) centers at 290 8C. DTA profile indicates that the
Jacobsen’s achiral ligand is thermally stable up to about 170 8C. The
DTA profiles of the other three ligands are similar to that shown for
the Jacobsen’s achiral ligand. On the other hand, the Jacobsen’s
achiral catalyst is apparently stable up to about 280 8C. The
enhanced thermal stability of the Jacobsen’s achiral catalyst can be
assigned to the complexation of the salen ligand with manganese
[18]. The DTA curves of the other three Mn(III)–salen complexes
are similar to that of the Jacobsen’s achiral catalyst. Overall, the
decomposition profiles of Mn(III)–salen complexes are completed
at about 550 8C with the residues amounting to manganese oxides.
Fig. 14 shows the effect of acetone/R-(+)-limonene ratio on the
oxidation of R-(+)-limonene using in situ generated DMD in the
absence of catalyst. In these experiments, the ratio of R-(+)-
limonene/KHSO5 was kept at 0.5. As can be observed in Fig. 14,
conversion and selectivity to diepoxides slightly improved when
the acetone/R-(+)-limonene ratio was 10 mL/mmol instead of
2 mL/mmol. Acetone contributes to DMD formation (see Fig. 9),
which act as oxidizing agent. At the same time, acetone favors the
Fig. 12. DTA curves of Jacobsen’s achiral ligand (a) and Jacobsen’s achiral catalyst
(b).

J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
63
Fig. 15. Effect of the amount of KHSO5 on the catalytic activity of (R,R)-Jacobsen’s
catalyst. Reaction conditions: R-(+)-limonene = 2.0 mmol; NaHCO3 = 1.2 mmol;
catalyst = 0.05 mmol; acetone = 4 mL; H2O = 4 mL; room temperature; pH ꢃ 8.0–
8.5; reaction time ꢃ 20–25 min.
Fig. 13. The effectof the amount of KHSO5 on the oxidation of R-(+)-limonene using in
situ generated DMD in the absence of catalyst. Reaction conditions: R-(+)-
limonene = 2.0 mmol; NaHCO3 = 1.2 mmol; catalyst = 0.05 mmol; acetone = 4.0 mL;
H2O = 4 mL; room temperature; pH ꢃ 8.0–8.5; reaction time ꢃ 20–25 min.
obtained. Besides, diepoxide selectivity increased by 22%. A very
important result from these experiments was the total recovery of
the catalyst, which was favored by increasing the amount of
KHSO5. This way, when an amount of KHSO5 below the
stoichiometric ratio was used (R-(+)-limonene/KHSO5 = 1.5), the
catalyst and the reaction products remained in the same phase, i.e.
homogenous catalysis (Fig. 16a). In this case the reaction products
and the catalyst cannot be separated by physical–mechanical
methods, so it was necessary to use a vacuum distillation process
in order to separate the catalyst. When the stoichiometric ratio (R-
(+)-limonene/KHSO5 = 1.0) was used there was a partial distribu-
tion of the catalyst between the solid and liquid phases, which was
entitled as partial recovery of the catalyst. Above the stoichio-
metric ratio (R-(+)-limonene/KHSO5 = 0.5), total recovery of the
catalyst was achieved (Fig. 16b). In this case, the catalyst is present
in the solid phase (brown color), which also contains inorganic
salts coming from the oxygen source and the buffer solution. These
salts are easily removed from the solid phase by washing with
enough water. After this, the catalyst was dried in flowing air at
60 8C and reused.
Fig. 16 shows pictures of the homogeneous and heterogeneous
systems. In the homogeneous system, both the catalyst and the
reaction products coexist in the upper phase. The bottom phase is
primarily composed of water. The photograph on the right side
correspondstoreactionconditions wherethecatalyst isrecoverable.
Here a solid phase that adheres to the walls of the tube can be
observed, as well as a liquid phase. The solid phase was composed of
catalyst and inorganic salts that accompany the process, while the
liquid phase contains the reaction products. This was determined by
extraction, vacuum distillation, and chromatographic analysis. The
catalyst can be recovered from the solid phase by washing it with
abundant water to remove inorganic salts.
Fig. 14. EffectoftheamountofacetoneontheoxidationofR-(+)-limoneneusinginsitu
generated DMD in the absence of catalyst. Reaction conditions: R-(+)-limonene = 2.0
mmol; KHSO5 = 4 mmol; NaHCO3 = 1.2 mmol; catalyst = 0.05 mmol; H2O = 4 mL;
room temperature; pH ꢃ 8.0–8.5; reaction time ꢃ 20–25 min.
interaction between DMD and R-(+)-limonene in the organic
phase.
Fig. 15 shows the effect of the amount of KHSO5 on the
oxidation of R-(+)-limonene in the presence of the (R,R)-Jacobsen’s
catalyst. In these experiments, we used acetone/R-(+)-limone-
ne = 2 mL/mmol. Similarly to the reaction without catalyst, there is
no significant effect of the amount of KHSO5 on conversion and
selectivity to diepoxides. The best selectivity towards diepoxides
(90%) was reached when the stoichiometric ratio R-(+)-limonene/
KHSO5 = 1 was used; above this value the selectivity did not further
improve. A total conversion of the R-(+)-limonene (100%) was
reached with the (R,R)-Jacobsen’s catalyst when R-(+)-limonene/
KHSO5 = 0.5, while in the absence of catalyst 38% conversion was
Fig. 16. (a) Catalyst in homogeneous phase; (b) recoverable catalyst in heterogeneous phase.

64
J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
Fig. 19. Reuse of (R,R)-Jacobsen’s catalyst.
Fig. 17. Effect of the amount of acetone on the catalytic activity of (R,R)-Jacobsen’s
catalyst. Reaction conditions: R-(+)-limonene = 2.0 mmol; KHSO5 = 4 mmol;
NaHCO3 = 1.2 mmol; catalyst = 0.05 mmol; H2O = 4 mL; room temperature;
pH ꢃ 8.0–8.5; reaction time ꢃ 20–25 min.
the main product is the diepoxide. In the absence of catalyst,
monoepoxide was also obtained with a poor selectivity (28%) and a
poor diasteromeric excess (35%).
The effect of the amount of acetone on the oxidation of R-(+)-
limonene in the presence of the (R,R)-Jacobsen’s catalyst was also
examined. In these experiments, the ratio of KHSO5 was above
stoichiometry, i.e. R-(+)-limonene/KHSO5 = 0.5. Fig. 17 shows
almost complete conversion when either amount of acetone was
used. However, the selectivity was different. Selectivity to
diepoxides was 86% when the ratio of acetone/R-(+)-limone-
ne = 2 mL/mmol while the endocyclic monoepoxide was obtained
(85% selectivity) when the ratio of acetone/R-(+)-limone-
ne = 10 mL/mmol. Comparing the catalyzed and un-catalyzed
reaction using the same ratio of acetone/R-(+)-limonene (10 mL/
mmol) it can be seen that the presence of the catalyst favors the
endocyclic monoepoxide with a diasteromeric excess of 58% for the
cis-epoxide. In the absence of catalyst diepoxides were the main
products. Also, in these experiments the recovery of the catalyst
was dependent on the amount of acetone. The catalyst was totally
recovered when acetone/R-(+)-limonene = 2 mL/mmol. On the
other hand, when acetone/R-(+)-limonene = 10 mL/mmol the
catalyst and reaction products remained in the same phase.
The oxidation of R-(+)-limonene using the (R,R)Jacobsen’s
catalyst, (S,S)-Jacobsen’s catalyst, Jacobsen’s racemic catalyst, and
Jacobsen’s achiral catalyst at reaction conditions under which the
catalyst was recoverable (R-(+)-limonene/KHSO5 = 0.5 mmol/
mmol and CH3COCH3/R-(+)-limonene = 2 mL/mmol) are shown
in Fig. 18. As can be seen in Fig. 18, similar conversion and
selectivity to diepoxide were observed with tested catalysts
suggesting that the chiral center of the enantiomerically pure
catalysts ((R,R) and (S,S)-Jacobsen) appear to have little or no
influence on conversion and selectivity to diepoxide; rather the
state of coordination given by the salen ligand to the manganese
appears to be crucial [19]. Here, the contribution of the catalyst is
proven once again by an increase of R-(+)-limonene conversion but
Finally, the reuse of (R,R)-Jacobsen’s catalyst was investigated
at reaction conditions under which the catalyst was recoverable. As
can be observed in Fig. 19 the catalyst experienced a slight
decrease in conversion through three consecutive runs. However,
the high level of conversion remained. On the other hand, the
obtained selectivity values are close to the initial selectivity. Given
that the percent recovery of catalyst varied between 85 and 90%,
the slight loss of catalytic activity is mainly associated with the
physical loss of the catalyst during the recovery process. These
results indicate the good stability of the catalyst towards the
oxidative degradation.
We recently reported that using in situ generated dimetildiox-
yrane (DMD) as oxidizing agent the catalyst is stable to the
oxidative degradation [20]. By FT-IR and DR UV–vis it was
demonstrated that the chemical structure of the catalyst was
retained after reusing it [20]. In the present work, the lack of
demetalation is proved by AAS since no Mn is detected in the solid-
free liquid phase (Fig. 16b). This fact makes unlikely the occurrence
of the hydrolysis. It is well known that the salycilidene imine group
is prone to undergo acid-catalyzed hydrolysis, reverting to the
corresponding salicylaldehyde and diamine in the presence of
water [11]. However, the stability of the salycilidene imine group
increases considerably upon coordination with a manganese ion
and formation of the Mn(III)–salen complex [11]. Therefore, in
contrast to the ligand salen, the Mn(III)–salen complex can be used
in wet solvents or even in aqueous media without undergoing
hydrolysis.
4. Conclusions
The main conclusions from this work are:
1. Under the studied conditions, diepoxides and endocyclic
monoepoxides are the main products of the oxidation of R-
(+)-limonene using in situ generated DMD as the oxidizing
agent.
2. In the presence of the Jacobsen-type catalysts, the R-(+)-
limonene/KHSO5 and CH3COCH3/R-(+)-limonene ratios affect
the selectivity and recovery of the catalyst. In this way, the
selectivity towards diepoxides and the total recovery of the
catalyst is favored above the stoichiometric ratio of R-(+)-
limonene/KHSO5. On the other hand, the increase of the amount
of acetone favors the selectivity towards the endocyclic epoxide
and the diasteromeric excess of the endocyclic cis-epoxide.
However, under these conditions catalyst is not recoverable. In
contrast, decreasing the amount of acetone favors the selectivity
towards diepoxides and catalyst recovery.
Fig. 18. Effect of the catalysts’ chirality on the catalytic activity. Reaction conditions:
R-(+)-limonene = 2.0 mmol; KHSO5 = 4 mmol; NaHCO3 = 1.2 mmol; catalyst =
0.05 mmol; acetone = 4 mL; H2O = 4 mL; room temperature; pH ꢃ 8.0–8.5;
reaction time ꢃ 20–25 min.
3. No effect of catalyst chirality on the catalytic activity was found.
Therefore, the origin of the activity of these catalysts is mostly

J. Cubillos et al. / Applied Catalysis A: General 373 (2010) 57–65
[3] L. Saikia, D. Srinivas, P. Ratnasamy, Appl. Catal. A 309 (2006) 144.
65
associated with the high coordination state of the manganese
instead of the chiral center of the salen ligand.
4. The (R,R)-Jacobsen’s catalyst exhibited good stability to
oxidative degradation.
[4] R.M. Carman, K.D. Klika, Aust. J. Chem. 44 (1991) 1803.
[5] R.A. Sheldon, Chirotechnology, Marcel Dekker, Inc., New York, 1993.
[6] H. Caner, E. Groner, L. Levy, Drug Discov. Today 9 (2004) 105.
[7] L. Canali, D.C. Sherrington, Chem. Soc. Rev. 28 (1999) 85.
[8] H.-U. Blaser, Chem. Commun. (2003) 293.
[9] Q.-H. Fan, Y.-M. Li, A.S.C. Chan, Chem. Rev. 102 (2002) 3385.
[10] Q.-H. Xia, H.-Q. Ge, C.-P. Ye, Z.-M. Liu, K.-X. Su, Chem. Rev. 105 (2005) 105.
Acknowledgments
´
[11] C. Baleiza˜o, H. Garcıa, Chem. Rev. 106 (2006) 3987.
[12] J.A. Cubillos, W.F. Hoelderich, Rev. Fac. Ing. Univ. Antioquia 41 (2007) 31.
The authors gratefully acknowledge the financial support from
CENIVAM-Universidad de Antioquia (RC No. 432) and 2008–2010
sustainability project.
[13] J. Reyes, J.A. Cubillos, C. Montes de Correa, A.L. Villa, Ing. Invest. 28 (2008) 37.
[14] J.F. Larrow, E.N. Jacobsen, J. Org. Chem 59 (1994) 1939.
[15] K.S. Webb, S.J. Ruszkay, Tetrahedron. 54 (1998) 401.
[16] B. Bahramian, V. Mirkhani, M. Moghadam, A.H. Amin, Appl. Catal. A 315 (2006)
52.
[17] V.D. Chaube, S. Shylesh, A.P. Singh, J. Mol. Catal. A: Chem. 241 (2005) 79.
[18] L. Frunza, H. Kosslick, H. Landmesser, E. Hijft, R. Fricke, J. Mol. Catal. A: Chem. 123
(1997) 179.
References
[1] P. Oliveira, M.L. Rojas-Cervantes, A.M. Ramos, I.M. Fonseca, A.M. Botelho do Rego
c, J. Vital, Catal. Today 118 (2006) 307.
[2] S.G. Casuscelli, M.E. Crivello, C.F. Perez, G. Ghione, E.R. Herrero, L.R. Pizzio, P.G.
[19] W. Adam, C. Mock-Knoblauch, C.R. Saha-Molller, M. Herderich, J. Am. Chem. Soc.
122 (2000) 9685.
´
´
Vazquez, C.V. Caceres, M.N. Blanco, Appl. Catal. A 274 (2004) 115.
[20] J. Cubillos, I. Montilla, C. Montes de Correa, Appl. Catal. A 366 (2009) 348.