Heteropoly acid catalysis for the isomerization of biomass-derived limonene oxide and kinetic separation of the trans-isomer in green solvents

Article abstract of DOI:10.1016/j.apcata.2019.117173

Terpenes are an abundant class of natural products, which is important for flavor and fragrance industry. Many acid catalyzed reactions used for upgrading terpenes still involve mineral acids as homogeneous catalysts and/or toxic solvents. Heteropoly acids represent a well-established eco-friendly alternative to conventional acid catalysts. As these reactions are usually performed in the liquid phase, solvents play a critical role for the process sustainability. In the present work, we developed a catalytic route to valuable fragrance ingredients, dihydrocarvone and carvenone, from limonene oxide by its isomerization using silica-supported tungstophosphoric acid as a heterogeneous catalyst and dialkylcarbonates as green solvents. The reaction pathway can be switched between dihydrocarvone and carvenone (obtained in 90% yield each) simply by changing the reaction temperature. In addition, we developed an efficient method for kinetic separation of trans-limonene oxide from commercial cis/trans-limonene oxide mixture and stereoselective synthesis of trans-dihydrocarvone.

Full text of DOI:10.1016/j.apcata.2019.117173

Accepted Manuscript
Title: Heteropoly acid catalysis for the isomerization of
biomass-derived limonene oxide and kinetic separation of the
trans-isomer in green solvents
Authors: Rafaela F. Cotta, Rafael A. Martins, Matheus M.
Pereira, Kelly A. da Silva Rocha, Elena F. Kozhevnikova, Ivan
V. Kozhevnikov, Elena V. Gusevskaya
PII:
DOI:
Article Number:
S0926-860X(19)30328-X
https://doi.org/10.1016/j.apcata.2019.117173
117173
Reference:
APCATA 117173
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
15 May 2019
19 July 2019
22 July 2019
Please cite this article as: Cotta RF, Martins RA, Pereira MM, da Silva Rocha
KA, Kozhevnikova EF, Kozhevnikov IV, Gusevskaya EV, Heteropoly acid catalysis
for the isomerization of biomass-derived limonene oxide and kinetic separation
of the trans-isomer in green solvents, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.117173
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Heteropoly acid catalysis for the isomerization of biomass-derived
limonene oxide and kinetic separation of the trans-isomer in green solvents
a
a
b
b
Rafaela F. Cotta , Rafael A. Martins , Matheus M. Pereira , Kelly A. da Silva Rocha ,
c
c
a,
Elena F. Kozhevnikova , Ivan V. Kozhevnikov , Elena V. Gusevskaya *
aDepartamento de Química, Universidade Federal de Minas Gerais, 31270-901, Belo Horizonte, MG,
Brazil
bDepartamento de Química, Universidade Federal de Ouro Preto, 35400-000, Ouro Preto, MG, Brazil
cDepartment of Chemistry, University of Liverpool, Liverpool L69 7ZD, UK

Corresponding author. Tel.: +55 31 34095741; fax: +55 31 34095700.
E-mail address: elena@ufmg.br (E. V. Gusevskaya).
Graphical Abstract
Research Highlights

HPW/SiO
2
promotes selective isomerization of bimass-based limonene oxide into valuable fragrance
ingredients.
1




The reaction pathway can be switched between dihydrocarvone and carvenone (obtained in ca. 90%
unprecedented yield each) simply by changing the reaction temperature.
trans-Limonene oxide can be kinetically separated from the commercial isomeric mixture, with the cis-isomer
being stereoselectively converted into trans-dihydrocarvone.
Eco-friendly dimethylcarbonate and diethylcarbonate can replace toxic hazardous solvents commonly used
in this reaction.
Abstract
Terpenes are an abundant class of natural products, which is important for flavor and
fragrance industry. Many acid catalyzed reactions used for upgrading terpenes still involve
mineral acids as homogeneous catalysts and/or toxic solvents. Heteropoly acids represent a well-
established eco-friendly alternative to conventional acid catalysts. As these reactions are usually
performed in the liquid phase, solvents play a critical role for the process sustainability. In the
present work, we developed a catalytic route to valuable fragrance ingredients, dihydrocarvone
and carvenone, from limonene oxide by its isomerization using silica-supported
tungstophosphoric acid as a heterogeneous catalyst and dialkylcarbonates as green solvents. The
reaction pathway can be switched between dihydrocarvone and carvenone (obtained in 90% yield
each) simply by changing the reaction temperature. In addition, we developed an efficient method
for kinetic separation of trans-limonene oxide from commercial cis/trans-limonene oxide mixture
and stereoselective synthesis of trans-dihydrocarvone.
Keywords: Bio-renewables; isomerization; heteropoly acid catalysis; kinetic resolution; limonene
oxide.
_
______________
1
. Introduction
The sustainability of chemical processes and their environmental footprint are nowadays
among the main concerns of chemical industry. In this context, the use of biomass based raw
materials in the production of chemicals is particularly important [1]. The most abundant class of
natural products consists of terpenic compounds, which are available at a relatively low cost from
essential oils. Besides direct applications in flavor and fragrance industry [2], terpenes can be
2

chemically converted in a variety of value-added products [3-7]. Acid catalyzed reactions are
widely used for upgrading terpenes; however, many of these processes still involve mineral acids
as homogeneous catalysts and/or toxic solvents as reaction media.
Heteropoly acids of the Keggin series (HPAs) are a well-established eco-friendly
alternative to conventional acid catalysts for the clean synthesis of fine and specialty chemicals
[
8,9]. HPA catalysts have been successfully applied in various reactions of terpenes [7]. For
example, we have reported the isomerization [10-12] and etherification/esterification [13-15] of
terpenic compounds over the HPA catalysts as well as their coupling with aldehydes [16,17]. Due
to the high reactivity of terpenes in the presence of acids and their tendency to oligomerization,
most of these transformations require the use of solvents. Moreover, some of these reactions, for
instance, the isomerization of limonene oxide [10], have been found to be strongly dependent on
the solvent nature.
Limonene oxide is a natural compound available from citric essential oils and also from
the epoxidation of limonene, one of the most widespread terpenic compounds. In the presence of
acid catalysts, limonene oxide can be converted into various valuable products for flavor and
fragrance industry, such as carveol, exocarveol, dihydrocarvone and carvenone [18]. In most
cases, good selectivities were achieved only for para-menthenic allylic alcohols such as carveol
and exocarveol [19-25]. Reports on the selective synthesis of carbonyl compounds from
limonene oxide are scarce and usually describe the procedures which require the use of hazardous
solvents or drastic conditions. For example, dihydrocarvone was obtained in 50-70% yield in
benzene solutions using a homogeneous ZnBr
stoichiometric amounts of solid LiClO [26], BF
produced from limonene oxide in modest 50-60% yields in toluene solutions over H
26] or Cu/SiO -Al [27] and in 80-90% yields over zeolite [28] or montmorillonite [29] at
high temperature (140–190 C) or microwave radiation.[29] The first application of HPA
2
catalyst [18] or in toluene solutions using nearly
-etherate [26] or alumina [21]. Carvenone was
SO /SiO
4
3
2
4
2
[
2
2
O
3
o
catalysts for the isomerization of limonene oxide was reported in our previous work [10]. The
reaction resulted in dihydrocarvone as the main product; however, satisfactory yields (ca. 80%)
were achieved only in toxic and rigorously regulated solvents such as 1,4-dioxane,
dichloromethane and 1,2-dichloroethane. For this reason, our further efforts were directed to the
reduction of the environmental impact of the process by finding “greener” alternatives for these
solvents.
3

Here, we report the isomerization of limonene oxide over silica-supported
tungstophosphoric acid H 40 (HPW) in eco-friendly solvents, dimethylcarbonate (DMC)
3
PW12O
and diethylcarbonate (DEC). Low-toxic and biodegradable DMC and DEC are highly
recommended green solvents with an eco-friendly ranking, comparable to that of ethanol and
water [30-33]. We found that the reaction in these solvents can be directed to either
dihydrocarvone or carvenone simply by changing the reaction temperature. Moreover, a thorough
control of the reaction variables allowed for kinetic separation of trans-limonene oxide and
stereoselective synthesis of trans-dihydrocarvone to be developed. More reactive cis-limonene
oxide was selectively transformed to the highly valuable trans-dihydrocarvone, with trans-
limonene oxide remaining largely unconverted. Enantiomerically pure epoxides are versatile
building blocks in asymmetric synthesis. In particular, the trans-limonene oxide is used for the
production of special biodegradable polymers such as polycarbonates and polyesters [34-36].
Commercial limonene oxide is a mixture of cis and trans isomers (hardly separable by physical
means) and their kinetic separation has attracted much attention. Most separation methods
developed so far involve the cleavage of the epoxy ring by water or methanol to give limonene
diol or its methoxy derivatives [37-43]. Only a few reports on the kinetic separation of limonene
oxide isomers has been published, which are based on their difference in reactivity toward
isomerization reactions [25,27,44].
2
. Experimental
The reagents and solvents employed for the catalyst synthesis and catalytic tests were
from Sigma-Aldrich, except Aerosil 300 silica, which was acquired from Degussa. (+)-Limonene
oxide (a mixture of cis and trans isomers in a ratio of 43/57) was also from Aldrich. Anhydrous
dimethylcarbonate (DMC) was used without additional treatment. Diethylcarbonate (DEC) was
distilled under argon and stored over 4Å molecular sieves.
The silica-supported H
3
PW12
O
40 (HPW) catalyst (HPW/SiO ) was synthesized by wet
2
2
-1
impregnation of Aerosil 300 silica (BET surface area of 300 m g ) with an aqueous solution of
o
HPW and dried at 130 C/0.2-0.3 Torr for 1.5 h, as described elsewhere [45]. From nitrogen
physisorption (Micromeritics ASAP 2010 instrument), the HPW/SiO
2
catalyst had a BET surface
2
-1
3 -1
31
area of 200 m g , pore volume of 0.53 cm g and average pore diameter of 144 Å. P MAS
4

NMR of HPW/SiO
5% H PO as the reference) revealed only one signal at ca. –15 ppm characteristic of the Keggin
structure of HPW (Figure S1) [46]. X-ray diffraction analyses (Rigaku Geigerflex-3034
diffractometer with CuK radiation) showed that most of HPW was finely dispersed on the
2
(Bruker Avance DSX 400 NMR, room temperature, spinning rate of 4 kHz,
8
3
4
α
surface, with only small amounts of crystalline phase being detected (Figure S2). The content of
tungsten and phosphorus was determined by inductively coupled plasma atomic emission
spectroscopy (ICP, SpectroCiros CCD instrument) and corresponded to 20 wt% of HPW loading
in the catalyst. The characterization data for the HPW/SiO
Information (Figures S1 – S3) and also documented in ref [8]. The acid properties of the
HPW/SiO catalyst have been discussed in detail elsewhere [47], including the number and the
2
catalyst are presented in Supporting
2
nature of acid sites and their strength.
Catalytic experiments were performed under air in a 10 mL glass reactor. In a typical run,
a mixture of limonene oxide (0.45–3.00 mmol), dodecane (GC internal standard, 0.30 mmol) and
HPW/SiO
2
catalyst (1–25 mg, 0.07–1.75 µmol of HPW) in a specified solvent (total volume of
o
3
.0 mL) was stirred with a magnetic stirrer at 25–90 C for a required time. The stirring was
stopped periodically and, after a quick catalyst settling, the samples of the supernatant liquid
were taken for the gas chromatography (GC) analysis (Shimadzu 17 instrument, Carbowax20 M
capillary column, flame ionization detector). Due to the small volume of the samples taken for
the GC analysis (5 µL) the changes in catalyst bulk density were insignificant. The reaction rate
was not dependent on the intensity of stirring within the range used, suggesting the absence of
mass transfer limitations. Conversions and selectivities were calculated from GC analysis using
dodecane as the internal standard. Reaction products were identified by GC-MS (Shimadzu
1
13
QP2010-PLUS spectrometer, 70 eV) by comparison with authentic samples and by H and C-
NMR (Bruker 400 MHz spectrometer, CDCl , TMS) after the separation by a column
chromatography (silica gel 60, hexane/CH Cl mixtures). The product characterization data are
3
2
2
presented in Supplementary material (Figures S4-S10). To control the leaching of HPW from
silica, the reaction mixture after reaction was submitted to centrifugation (1.5 h, 18000 rpm) or
vacuum filtration (10–20 microns filter) to separate the catalyst then the supernatant was allowed
to react further under the same conditions. The absence of any additional conversion of limonene
oxide indicated the absence of dissolved HPW, i.e., the absence of any significant HPW leaching.
5

3
. Results and discussion
3
.1. Synthesis of dihydrocarvone
The data on the isomerization of limonene oxide (1) over the HPW/SiO catalyst are
2
presented in Tables 1 and 2. In the absence of catalyst or in the presence of pure silica, the
substrate was not converted at all in blank reactions in all solvents used. The structures of
identified reaction products such as dihydrocarvone (2), carvenone (3), 1-methyl-3-isopropenyl-
cyclopentyl-1-carboxaldehyde (4), and limonene-1,2-diol (5) are presented in Schemes 1 and 2.
Previously, we have reported that the isomerization of limonene oxide over HPA catalysts
resulted mainly in dihydrocarvone [10]. Among various solvents tested, 1,4-dioxane gave the
best result, however 1,4-dioxane is a hazardous solvent. Organic carbonates, in particularly
dimethylcarbonate (DMC) and diethylcarbonate (DEC), have recently attracted our attention as
alternative eco-friendly solvents for the catalytic transformations of terpenic substrates.
In the present work, it was found that DMC and DEC could be successfully used to
perform the isomerization of limonene oxide over the silica-supported HPW. It is especially
important that HPW is not soluble in DMC and DEC in contrast to 1,4-dioxane, hence the
reaction can be conducted under truly heterogeneous conditions.
The reaction in anhydrous DMC occurred rapidly in the presence of small amounts of
HPW/SiO (0.2 wt% based on the whole reaction mixture) at room temperature, reaching
2
complete conversion in 15 min (Table 1, run 1). Cis and trans isomers of dihydrocarvone were
formed in total 84% yield derived from the trans and cis isomers of the original limonene oxide,
respectively (Scheme 1). This result is better than that obtained in homogeneous 1,4-dioxane
solutions of HPW previously [10]. After reaction (runs 1 and 9, Table 1), the solid catalyst was
removed and the supernatant was allowed to react with a new portion of limonene oxide under
the same conditions. No further reaction was observed to confirm the heterogeneous nature of the
process without any significant HPW leaching.
In an attempt to improve the catalyst efficiency in terms of turnover number (TON), the
concentration of the substrate was increased (Table 1, run 2). Although the reaction was fast, the
selectivity for dihydrocarvone dropped to 64% probably due to concomitant oligomerization of
the substrate. The GC analysis of the supernatant after the reaction showed the presence of high-
6

boiling products with the mass spectra similar to that of limonene oxide (probably, limonene
oxide dimers or trimers). Optimization of the catalyst loading and substrate concentration allowed
to achieve high TONs (6000–13000) while maintaining excellent selectivity for dihydrocarvone
(Table 1, runs 3–6). Minor products in all these runs were ring-contracted aldehyde 4 and diol 5,
with the latter being probably formed due to the residual water in the system. After reaction, the
catalyst was separated by centrifugation from the reaction mixture, washed with chloroform and
reused two times without a significant decrease in activity and selectivity (Table 1, runs 2, 2a and
2
b).
o
The reactions in DMC at 40 C gave dihydrocarvone in nearly 90% yield, the best result
ever reported for this product as far as we know (Table 1, runs 7 and 8). DEC, a solvent ranked in
sustainability guides even better that DMC and close to water [30], can also be used as the
solvent for the isomerization of limonene oxide. In the DEC, dihydrocarvone was obtained in 80–
8
5% yields (Table 1, runs 9 and 10). The limonene oxide conversion occurred slower in DEC
than in DMC (Table 1: run 9 vs. run 6; run 10 vs. run 7).
Thus, both organic carbonates, DMC and DEC, can perfectly substitute hazardous 1,4-
dioxane, dichloromethane and 1,2-dichloroethane as the reaction media for the synthesis of
dihydrocarvone from limonene oxide over the HPW catalyst.
Furthermore, an unexpected observation allowed us to change the reaction selectivity in
these solvents and obtain another valuable limonene oxide isomer in excellent yields. When the
o
reaction mixture was left in contact with the catalyst for 5 hours at 40 C, a new reaction product
appeared in appreciable concentrations at the expense of dihydrocarvone (Table 1, run 8). The
product was identified as carvenone (compound 3 in Scheme 1), which is a high-value natural
compound used as a constituent for perfume, flavor and fragrance formulations. Therefore, we
further attempted to develop an efficient and eco-friendly synthesis of carvenone from limonene
oxide (Table 2).
3
.2. Synthesis of carvenone
The selectivity for the formation of carvenone was found to be strongly dependent on the
reaction temperature. The yield of carvenone in a 5-hour reaction was 9% at 40 ºC (Table 1, run
), 18% at 50 ºC, 35% at 60 ºC and 68% at 70 ºC (Table 2, runs 1, 2 and 3). Keeping the reaction
8
7

mixture in contact with the catalyst for 10 h at 70 ºC resulted in 84% yield of carvenone along
with 5% of remaining dihydrocarvone (Table 2, run 3). All the reactions presented in Table 2
reached a nearly 100% conversion of limonene oxide at first sampling in 15 min, with
dihydrocarvone being the major product at that moment. The formation of carvenone was
accompanied by the gradual decrease in the concentration of dihydrocarvone. It is important that
the combined selectivity for dihydrocarvone and carvenone was nearly the same (85–90%) along
the whole reaction course. Thus, the catalytic process consists of two consecutive steps including
fast formation of dihydrocarvone from limonene oxide and slow isomerization of dihydrocarvone
into carvenone, with both steps catalyzed by HPW (Scheme 1).
The results obtained clearly showed that higher temperatures were required to accelerate
the second reaction step. In order to be able to increase the reaction temperature, we substituted
DMC (b.p. 90 ºC) by a less volatile DEC (b.p. 126 ºC). The catalytic performance of HPW/SiO
was nearly the same in DMC and DEC at 70 ºC (Table 2, runs 3 and 4). At 90 ºC in DEC
solutions, carvenone was obtained in 93% yield in only 2 hours, which is the best result ever
2
reported for this product as far as we know (Table 2, run 5). The catalyst was reused two times
Table 2, runs 5, 5a and 5b). A complete conversion of limonene oxide was observed in these
(
runs in 15 min, without a significant decrease in the yield of carvenone after 2 hours (ca. 90%).
The initial rate of carvenone formation, the second slower step in Scheme 1, was
estimated and the TOF values are presented in Table 2. It was found that the reaction had the
-1
same activation energy of 82 kJ mol in both DMC and DEC solutions (in the range of 40–70 ºC
in DMC and 70–90 ºC in DEC). It should be emphasized that turnover frequencies given in Table
2
are refered to the formation of carvenone from dihydrocarvone.
In run 10 (Table 2), catalyst was removed from the reaction mixture after 60 min and the
supernatant was allowed to react further under the same conditions with no additional conversion
of dihydrocarvone being observed thereupon to confirm the absence of significant HPW leaching.
HPW is insoluble in low polar DEC and DMC and thermally stable under the conditions used,
thus ensuring the stability of the catalyst.
The effectiveness of organic carbonates as the reaction medium for the synthesis of
carvenone from limonene oxide appears particularly remarkable when compared to that of other
solvents (Table 2, runs 6–9). In 1,4-dioxane, not a trace of carvenone was detected under the
same conditions used for the reactions in DMC and DEC (Table 2, runs 6 and 7 vs. runs 3 and 4).
8

Both solid HPW/SiO and dissolved HPW have been tested in the corresponding amounts in 1,4-
2
dioxane (HPW is soluble in 1,4-dioxane). In 1,2-dichloroethane and toluene, the formation of
carvenone occurred 3 times slower than in DMC and DEC (Table 2, runs 8 and 9 vs. runs 3 and
4
). In all cases, the main product rapidly formed from limonene oxide was dihydrocarvone, which
was then slowly converted into carvenone (except for 1,4-dioxane). The rate of the second
reaction step was strongly dependent on the solvent nature.
The data obtained demonstrate that the solvent effect in dihydrocarvone isomerization is
complex and does not correlate with the solvent polarity and basicity. The reaction in the most
polar solvent studied, 1,2-dichloroethane (ε = 10.4), had a similar rate as in nonpolar toluene (ε =
2
.4) (Table 2, run 8 vs. run 9). Both 1,2-dichloroethane and toluene are nonbasic solvents. On the
other hand, weakly basic and weakly polar DMC and DEC (ε = 3.1 and 2.8, respectively) showed
much better efficiency than 1,2-dichloroethane and toluene. Finally, the reaction did not occur at
o
all at 70 C in nonpolar 1,4-dioxane, the solvent with the highest basicity among the solvents
tested in the present work. This may be due to a decrease of HPA acid strength in 1,4-dioxane
solutions. Further studies are required for better understanding the solvent effects in this reaction.
3
.3. Kinetic separation of trans-limonene oxide
Epoxidation of limonene usually results in an almost equimolar mixture of cis and trans
diastereomers of limonene oxide. Working with a commercial sample of limonene oxide
cis/trans ratio of 43:57) we noticed a great difference in reactivity of the two isomers toward the
(
isomerization into dihydrocarvone. A larger steric hindrance of the oxygen atom in the trans
isomer and consequently more difficult access to the active acid sites of the catalyst probably
accounts for its lower reactivity in comparison to the cis-limonene oxide [27]. This difference
provides the possibility for the kinetic separation of trans-limonene oxide from the commercial
mixture over the HPW catalyst and for the selective synthesis of trans-dihydrocarvone (2a)
(
Scheme 1).
The reactions aiming at the kinetic separation of trans-limonene oxide were performed
under ambient conditions in the presence of small amounts of HPW/SiO (0.015 mol% of HPW,
2
corresponding to a TON of ca. 6500). The data presented in Table 3 clearly show much higher
reactivity of cis isomer 1a as compared to trans isomer 1b. In DMC solutions, complete
9

conversion of 1a was observed in 10 min, whereas only 15% of 1b was converted.
Dihydrocarvone2 (85% trans) was formed in 91% chemoselectivity, with the remaining limonene
oxide having exclusively trans geometry. However, under such conditions the reactivity of the
trans-isomer was too high to allow sufficient time for the catalyst separation in order to quench
the reaction at this step. In 15 min, the conversion of the trans-isomer was already 27%
decreasing the diastereomeric purity of the dihydrocarvone product (Table 3, run 1). The reaction
rate could be diminished at lower temperatures, but this would bring about a technical problem
related to the low-temperature catalyst separation (centrifugation or filtration). In a 5-fold diluted
solution of DMC, the reaction occurred slower and showed complete conversion of the cis isomer
in 5 h (Table 3, run 2). Nearly 80% of the trans isomer remained unconverted ensuring the
diastereoisomeric purity of 82% of the resulting dihydrocarvone. Another way to decelerate the
reaction could be the use of DEC as the solvent, as in DEC the reaction occurred slower than in
DMC (Table 1, runs 6 vs. run 9)
It is important that the kinetic diastereo-discrimination between the cis and trans isomers
of limonene oxide in DEC solutions was found not only as efficient as in DMC but also more
convenient technically due to the lower reactivity of both isomers (Table 3, run 3 vs. run 1). Best
purity of the remaining limonene oxide (100% trans) and dihydrocarvone (84% trans) was
obtained in 30 min. To achieve this, the reaction must be terminated at a very low conversion of
trans-limonene oxide. Conversely, a long contact with the catalyst resulted in nearly complete
conversion of both isomers, with the final isomeric composition of dihydrocarvone corresponding
to the isomeric ratio in the original limonene oxide (Table 3, runs 1 and 3). It should be taken into
account that the cis isomer of limonene oxide gives trans-dihydrocarvone, whereas trans-
limonene oxide gives cis-dihydrocarvone (Scheme 1).
In a special experiment (Table 3, run 4), the reaction in DEC was stopped in 30 min and
the catalyst was quickly separated by vacuum filtration. No further reaction was observed in the
supernatant for the next two hours confirming the technical viability of isomer separation under
these conditions and also the absence of catalyst leaching problems. The remaining trans-
limonene oxide after fractional distillation from the reaction mixture can be further processed
over the HPW/SiO
2
catalyst to produce pure cis-dihydrocarvone. The individual isomers of
dihydrocarvone have different organoleptic characteristics and would be much more valuable
1
0

than their mixture. The trans isomer has the odor of spearmint, whereas the cis isomer has a
musty and woody scent [44].
3
.4. Comments on reaction mechanism
The suggested mechanism of the transformations of limonene oxide 1 in the presence of
HPW catalyst is schematically represented in Scheme 2. Protonation of the oxygen atom opens
the epoxy ring in limonene oxide resulting in carbenium ion A, which can further undergo several
competing rearrangements. The C2-C1 hydride shift in A gives carbenium ion B, which then
loses a proton to give dihydrocarvone 2. As mentioned above, the formation of carvenone 3 was
accompanied by the gradual decrease in the concentration of dihydrocarvone 2. The
isomerization of dihydrocarvone into carvenone can be induced by its protonation resulting in the
formation of intermediate carbenium ions C and D. Alternatively, carbenium ion A can rearrange
into carbenium ion E via the transfer of the electron pair of the C2– C3 σ-bond to C3 carbon
atom. Proton abstraction from E would result in ring-contracted aldehyde 4. Finally, the residual
water can react with carbeniumion A to give limonene diol 5.
4
. Conclusions
In has been demonstrated that dimethylcarbonate (DMC) and diethylcarbonate (DEC) are
effective and eco-friendly alternatives for conventional hazardous solvents for the isomerization
of limonene oxide to valuable fragrance ingredients. The HPA catalyst was applied for the first
time for the efficient synthesis of carvenone from limonene oxide. The reaction can be selectively
directed to either dihydrocarvone or carvenone simply by changing the reaction temperature.
These compounds are obtained in ca. 90 yield each using silica-supported tungstophosphoric acid
3
H PW12O40 as a heterogeneous catalyst. This is the highest yield reported for these compounds so
far, to the best of our knowledge. The catalyst and the solvents can be easily recovered and
reused. Different reactivity of cis- and trans-limonene oxide can be effectively utilized for kinetic
separation of trans-limonene oxide and stereoselective synthesis of trans-dihydrocarvone.
Acknowledgments
1
1

The authors acknowledge financial support and scholarships from CNPq, CAPES,
FAPEMIG, INCT-Catálise (Brazil) and PROPP-UFOP.
1
2

References
[
1]
A. Behr, L. Johnen, Alternative Feedstocks for Synthesis, in Ch-J. Li (Ed.), Handbook
of Green Chemistry, Volume 7: Green Synthesis, first ed., Wiley-VCH, Weinheim, 2012,
pp. 69–92.
[
[
2]
3]
E. Breitmaier, Terpenes. Flavors, Fragrances, Pharmaca, Pheromones, Wiley-VCH,
Weinheim, 2006.
A. Behr, A.J. Vorholt, K.A. Ostrowski, T. Seidensticker, Green Chem. 16 (2014) 982–
1
006.
[
[
[
[
[
4]
5]
6]
7]
8]
E.V. Gusevskaya, J. Jimènez-Pinto, A. Börner, ChemCatChem 6 (2014) 382–411.
K.A.D. Swift, Top. Catal. 27 (2004) 143–155.
J.L.F. Monteiro, C.O. Veloso, Top. Catal. 27 (2004) 169–180.
E.V. Gusevskaya, ChemCatChem 6 (2014) 1506–1515.
I.V. Kozhevnikov, Catalysts for Fine Chemicals, Catalysis by Polyoxometalates, Vol. 2,
Wiley, Chichester, 2002.
[
[
[
[
[
[
[
[
9]
Y. Kamiya, T. Okuhara, M. Misono, A. Miyaji, K. Tsuji, T. Nakajo, Catal. Surv. Asia 12
(2008)101–113.
10] V.V. Costa, K.A. da Silva Rocha, I.V. Kozhevnikov, E.F. Kozhevnikova, E.V.
Gusevskaya, Catal. Sci. Technol, 3 (2013) 244–250.
11] A.L.P. de Meireles, M. dos Santos Costa, K.A. da Silva Rocha, E.V. Gusevskaya, Appl.
Catal. A 502 (2015) 271–275.
12] K.A. da Silva Rocha, P.A. Robles-Dutenhefner, I.V. Kozhevnikov, E.V. Gusevskaya,
Appl. Catal. A 352 (2009) 188–192.
13] A.L.P. de Meireles, M. dos Santos Costa, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V.
Kozhevnikov, E.V. Gusevskaya, ChemCatChem 6 (2014) 2706–2711.
14] K.A. da Silva Rocha, N.V.S. Rodrigues, I.V. Kozhevnikov, E.V. Gusevskaya, Appl.
Catal. A 374 (2010) 87–94.
15] A.L.P. de Meireles, M. dos Santos Costa, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V.
Kozhevnikov, E.V. Gusevskaya, ChemCatChem 6 (2014) 2706–2711.
16] V.V. Costa, K.A. da Silva Rocha, R.A. Mesquita, E.F. Kozhevnikova, I.V. Kozhevnikov,
E.V. Gusevskaya, ChemCatChem 5 (2013) 3022–3026.
1
3

[
17] R.F. Cotta, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V. Kozhevnikov, E.V.
Gusevskaya, Appl. Catal. B 217 (2017) 92–99.
[
[
[
[
[
[
18] R. L. Settine, G. L. Parks, G. L. K. Hunter, J. Org. Chem. 29 (1964) 616–618.
19] I. C. Nigam, L. Levi, Can. J. Chem. 46 (1968) 1944–1947.
20] E. H. Eschinasi, Israel J. Chem. 6 (1968) 713–721.
21] K. Arata, K. Tanabe, Chem. Lett. (1976) 321–322.
22] J. Jayasree, C. S. Narayanan, Res. Chem. Intermed. 23 (1997) 169–178.
23] G. G. Kolomeyer, J. S. Oyloe (Millennium Specialty Chemicals), EP Patent 1404635
(2003).
[
[
24] C. Raptis, H. Garcia, M. Stratakis, Angew. Chem. Int. Ed. 48 (2009) 3133–3136.
25] G. G. Kolomeyer, D. A. Ferone (LyondellBasell Flavors & Fragrances Company), US
Patent 7884252 (2011).
[
[
[
26] K. Arata, S. Akutagatva, K. Tanabe, J. Catal. 41 (1976) 173–179.
27] N. Ravasio, F. Zaccheria, M. Guidotti, R. Psaro,Top. Catal. 27 (2004) 157–168.
28] Y. Fujiwara, M. Nomura, K. Igawa (Toyo Soda Mfg. Company), JP Patent 62114926
(1987).
[
[
29] T.-T.T. Nguyen, D.-K.N. Chau, F. Duus, T.N. Le, Int. J. Org. Chem. 3 (2013) 206–209.
30] C.M. Alder, J.D. Hayler, R. K. Henderson, A. M. Redman, L. Shukla, L. E. Shuster, H. F.
Sneddon, Green Chem. 18 (2016) 3879–3890.
[
31] D. Prat, A.Wells, J. Hayler, H. Sneddon, C.R. McElroy, S. Abou-Shehada, P.J. Dunn,
Green Chem. 18 (2016) 288–296.
[
[
32] F. Aricò , P. Tundo, Rus. Chem. Rev. 79 (2010) 479–489.
33] B. Schäffner, F. Schäffner, S. P. Verevkin, A. Börner, Chem. Rev. 110 (2010) 4554–
4
581.
34] C.M. Byrne, S.D. Allen, E.B. Lobkovsky, G.W. Coates, J. Am. Chem. Soc. 126 (2004)
1404–11405.
35] O. Hauenstein, M. Reiter, S. Agarwal, B. Riegerb, A. Greiner, Green Chem. 81 (2016)
60–770.
[
[
1
7
[
[
36] R.C. Jeske, A.M. DiCiccio, G.W. Coates, J. Am. Chem. Soc. 129 (2007) 11330–11331.
37] L. Salles, A.F. Nixon, N.C. Russell, R. Clarke, P. Pogorzelec, D.J. Cole-Hamilton,
Tetrahedron: Asymmetry 10 (1999) 1471–1476.
1
4

[
38] M.J. van der Werf, H. Jongejan, M.C.R. Franssen, Tetrahedron Lett. 42 (2001) 5521–
524.
5
[
39] D. Steiner, L. Ivison, C. T. Goralski, R.B. Appel, J.R. Gojkovic, B. Singaram,
Tetrahedron: Asymmetry 13 (2002) 2359–2363.
[
[
40] B.K. Bettadaiah, P.J. Srinivas, Photochem. Photobiol. A 167 (2004) 137–140.
41] P.C. Andrews, M. Blair, B.H. Fraser, P.C. Junk, M. Massi, K.L. Tuck, Tetrahedron:
Asymmetry 17 (2006) 2833–2838.
[
[
42] M. Blair, P.C. Andrews, B.H. Fraser, C.M. Forsyth, P.C. Junk, M. Massi, K.L. Tuck,
Synthesis (2007) 1523–1527.
43] C.S. Kumar, J.R. Manjunatha, P. Srinivas, B.K. Bettadaiah, J. Chem. Sci. 126 (2014)
8
75–880.
[
[
44] C. G. Cardenas, Z.U. Din (SCM Corporation), US Patent 4 296 257 (1981).
45] I.V. Kozhevnikov, A. Sinnema, A.J.A. van der Weerdt, H. van Bekkum, J. Mol. Catal. A
1
20 (1997) 63–70.
[
[
46] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171–198.
47] E.F. Kozhevnikova, I.V. Kozhevnikov, J. Catal. 224 (2004) 164–169.
1
5

Scheme and Table captions
Scheme 1. Main products of limonene oxide isomerization of over HPW/SiO .
2
Scheme 2. Schematic presentation of acid catalyzed transformations of limonene oxide (1).
Table 1. Isomerization of limonene oxide (1) over HPW/SiO
Table 2. Isomerization of limonene oxide (1) over HPW/SiO
2
2
: synthesis of dihydrocarvone (2).
: synthesis of carvenone (3).
a
Table 3. Kinetic separation of trans-limonene oxide induced by HPW/SiO
2
1
6

Table 1. Isomerization of limonene oxide (1) over HPW/SiO
Run Catalyst/HPW T
Time Conversion Selectivity (%)
mmol) (mg/µmol) (ºC) (min) (%)
Solvent: dimethylcarbonate
2
: synthesis of dihydrocarvone (2)a
1
TONb
(
2
3
4
5
1c
2d
0.45
3.00
5/0.35
5/0.35
25
25
15
15
100
81
84
64
64
65
65
62
64
62
64
65
63
82
81
88
83
87
88
77
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9
8
7
1290
11 14
6
0
100
76
11 14 8570
12 11
2
2
3
4
5
6
ad
bd
3.00
3.00
3.00
1.50
0.90
0.45
5/0.35
5/0.35
1/0.07
1/0.07
1/0.07
1/0.07
25
25
25
25
25
25
15
6
0
100
75
10 13 8570
13 10
15
6
0
100
63
12 11 8570
12 10
15
6
0
99
10 12 42400
15
65
9
8
9
9
3
7
6
2
2
8
6
9
9
4
7
6
6
0
6
0
94
20100
12900
15
62
6
0
100
60
15
6
0
98
6040
6430
7
8
0.45
0.45
1/0.07
40
40
15
15
100
100
100
25/1.75
3
00
257
Solvent: diethylcarbonate
9c
0.45
1/0.07
25
40
15
34
94
89
82
87
85
0
0
0
0
0
6
5
7
10
1
20
11 6040
8
1
0
0.45
1/0.07
15
76
6
0
100
9
6430
a
Total volume 3.0 mL. Conversion and selectivity (based on converted limonene oxide) were determined by GC. b
c
Turnover number in moles of substrate converted per mole of a total amount of HPW in the catalyst. After the run,
the catalyst was removed, the solution was recharged with fresh substrate (0.45 mmol) and the reaction was allowed
d
to proceed further under the same conditions for 1 h, with no further conversion observed thereupon. After run 2,
the catalyst was reused two times (runs 2a and 2b).
1
7

Table 2. Isomerization of limonene oxide (1) over HPW/SiO
2
: synthesis of carvenone (3)a
Run Solvent
T
Time Conversion Selectivity (%) TONb TOFc
-
1
(ºC) (min) (%)
2
3
(min )
1
2
3
dimethylcarbonate 50
dimethylcarbonate 60
dimethylcarbonate 70
15
00
15
100
100
100
85
72
84
61
66
20
6
3
0.15
0.30
1.62
3
18
4
257
257
3
00 100
35
23
68
84
27
75
77
41
93
38
91
35
90
0
15
100
100
100
100
100
100
100
3
00
00
15
00
00
15
6
257
4
diethylcarbonate
70
66
20
11
47
2
1.63
3
6
257
257
257
257
257
257
257
257
257
5d
diethylcarbonate
diethylcarbonate
90
90
90
70
70
7.08
6.50
6.00
0
1
20 100
15 100
20 100
15 100
20 100
5
5
6
ad
52
4
1
bd diethylcarbonate
1,4-dioxane
54
4
1
15
00
15
00
15
00
15
00
100
100
100
100
100
100
100
100
100
100
85
86
88
88
83
51
81
46
45
45
3
3
3
3
0
7e
8
1,4-dioxane
0
0
0
1,2-dichloroethane 70
3
0.56
0.50
1.70
37
1
9
toluene
70
70
34
52
52
1
0f
diethylcarbonate
60
6
00
a
Limonene oxide, 0.45 mmol; catalyst 20%HPW/SiO
2
, 25 mg (1.75 µmol of HPW); total volume 3.0 mL.
b
Conversion and selectivity (based on converted limonene oxide) were determined by GC. Turnover number in
1
8

c
moles of substrate converted per mole of a total amount of HPW in the catalyst. Initial turnover frequency for the
d
formation of carvenone 3 (up to 25–40% conversion of dihydrocarvone). After run 5, the catalyst was reused two
f
times (runs 5a and 5b). [e] Homogeneous HPW catalyst, 5 mg (1.75 µmol). After 60 min, the catalyst was removed
from the reaction mixture and the supernatant was allowed to react further under the same conditions with no
additional conversion of dihydrocarvone being observed thereupon.
1
9

a
Table 3. Kinetic separation of trans-limonene oxide induced by HPW/SiO
Run Solvent Time Conversion (%) Composition of remaining 1 Selectivity for 2 (%)
min) 1a (cis) 1b (trans) (1b trans/1c cis)
2
(
(2a trans/2b cis)
91 (85/15)
88 (80/20)
85 (65/35)
83 (52/48)
76 (47/53)
95 (72/28)
90 (83/17)
85 (82/18)
89 (89/11)
87 (87/13)
87 (84/16)
84 (67/33)
82 (52/48)
87 (84/16)
87 (84/16)
1
DMC
10
100
100
100
100
100
24
15
27
51
95
100
5
100/0
100/0
100/0
100/0
-
1
5
0
0
3
6
3
00
30
2b
DMC
DEC
61/39
84/16
100/0
77/23
92/8
1
3
80
00
78
9
100
64
20
3
3
15
2
5
0
0
90
9
3
6
100
100
100
100
100
19
52
90
19
20
100/0
100/0
100/0
100/0
100/0
1
20
4c
DEC
30
1
50
a
Limonene oxide (1) (commercial mixture of cis and trans isomers in a ratio of 43/57), 0.45 mmol; catalyst
o
2
0%HPW/SiO
2
, 1 mg (0.07 mol of HPW); 25 C; total volume 3.0 mL. Conversion and selectivity (based on
b
c
converted limonene oxide) were determined by GC. Total volume 15.0 mL After 30 min, the catalyst was
centrifuged from the reaction mixture and the supernatant was allowed to react further under the same conditions
with no additional conversion of limonene oxide being observed thereupon.
2
0

Scheme 1. Main products of limonene oxide isomerization over HPW/SiO .
2
2
1

Scheme 2. Schematic presentation of acid catalyzed transformations of limonene oxide (1).
2
2