Influence of ligand substitution on molybdenum catalysts with tridentate Schiff base ligands for the organic solvent-free oxidation of limonene using aqueous TBHP as oxidant

Article abstract of DOI:10.1016/j.mcat.2017.09.033

The oxidation of limonene by aqueous TBHP has been analyzed in the presence of molybdenum complexes [MoO₂L]₂ as catalysts with five different tridentate ligands L in the absence of organic solvents (greener reaction conditions). The ligands are based on a common salicylidene amino(thio)phenolate, SA(T)P, backbone with differences in the coordination sphere (ONO for L = SAP vs. ONS for L = SATP) or in the salicyl moiety functionalization by OH groups for the ONO ligands. The process gives a regioselective endocyclic epoxidation to a kinetically controlled 1:1 mixture of the cis-LimO and trans-LimO epoxides and/or the isomeric diols ax-LimD and eq-LimD by the subsequent ring opening in the presence of water, with a product distribution that depends on the ligand, reaction time and temperature. In combination with control experiments of the cis/trans-LimO ring opening, the investigations demonstrate the catalytic action of the metal complexes in both the epoxidation and the ring opening steps, with the cis-LimO stereospecifically producing the ax-LimD product and the less reactive trans-LimO leading to a 4:3 mixture of ax-LimD and eq-LimD. The ONS system [MoO₂(SATP)]₂ exhibits the highest catalytic activity in both steps.

Full text of DOI:10.1016/j.mcat.2017.09.033

Molecular Catalysis 443 (2017) 52–59
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Editor’s choice paper
Influence of ligand substitution on molybdenum catalysts with
tridentate Schiff base ligands for the organic solvent-free oxidation of
limonene using aqueous TBHP as oxidant
Weili Wanga,b, Dominique Agustina,b,∗, Rinaldo Polia,c
a CNRS, LCC (Laboratoire de Chimie de Coordination), Université de Toulouse, UPS, INPT, 205, route de Narbonne, F-31077 Toulouse, France
b Université de Toulouse, Institut Universitaire de Technologie Paul Sabatier – Département de Chimie, Av. Georges Pompidou, BP 20258, F-81104 Castres
Cedex, France
c Institut Universitaire de France, 1 rue Descartes, 75231 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 July 2017
Received in revised form 5 September 2017
Accepted 29 September 2017
The oxidation of limonene by aqueous TBHP has been analyzed in the presence of molybdenum complexes
[MoO2L]2 as catalysts with five different tridentate ligands L in the absence of organic solvents (greener
reaction conditions). The ligands are based on a common salicylidene amino(thio)phenolate, SA(T)P,
backbone with differences in the coordination sphere (ONO for L = SAP vs. ONS for L = SATP) or in the
salicyl moiety functionalization by OH groups for the ONO ligands. The process gives a regioselective
endocyclic epoxidation to a kinetically controlled 1:1 mixture of the cis-LimO and trans-LimO epoxides
and/or the isomeric diols ax-LimD and eq-LimD by the subsequent ring opening in the presence of water,
with a product distribution that depends on the ligand, reaction time and temperature. In combination
with control experiments of the cis/trans-LimO ring opening, the investigations demonstrate the catalytic
action of the metal complexes in both the epoxidation and the ring opening steps, with the cis-LimO
stereospecifically producing the ax-LimD product and the less reactive trans-LimO leading to a 4:3 mixture
of ax-LimD and eq-LimD. The ONS system [MoO2(SATP)]2 exhibits the highest catalytic activity in both
steps.
Keywords:
Epoxidation
Molybdenum
Limonene
Solvent-free process
Ligand effect
© 2017 Elsevier B.V. All rights reserved.
Introduction
[11]. Both LimD isomers, eq-LimD and ax-LimD, generated from the
LimOs by epoxide ring opening with water [12], are also precursors
The use of natural feedstock as chemical source is of con-
stantly increasing interest [1]. Terpenes, due to their diverse
structures, represent an interesting available ore for several chem-
ical transformations [2]. As an example of terpene, limonene (Lim)
is an inexpensive substrate obtained as waste from the orange
juice industry [3,4]. The limonene epoxides (LimOs) and diols
(LimDs), obtained by oxidation of Lim (Scheme 1), are added-value
renewable-based compounds [5,6], with applications in the syn-
thesis of fragrances, flavors, food additives, or compounds with
biological and therapeutic activity [7,8]. Both LimO isomers, cis-
LimO and trans-LimO, find applications in medicine [9], metal
coatings, varnishes, printing inks, and in natural based polymers
[10] but exist in Nature in insufficient quantities to satisfy the needs
of interest for the pharmaceutical industry [13].
The Lim epoxidation may occur at the internal and/or terminal
alkene moiety and the presence of water facilitates the epoxide ring
opening. In most cases, the main products are cis-LimO and trans-
LimO deriving from the regioselective epoxidation of the endocyclic
double bond, which is more reactive because of its higher electron
density. The terms cis and trans, for this family of products, refer to
the relative position of the Me and iPr substituents relative to the
cyclohexane ring. Although four different stereoisomers may result
in principle from the ring opening of cis-LimO and trans-LimO, only
two of them, the diequatorial cis isomer (eq-LimD) and the diaxial
trans isomer (ax-LimD), have been obtained by this procedure as
shown in Scheme 1. Traces of other compounds such as carvone
and carveol, resulting from allylic oxidation, may also be observed.
For the non-catalyzed processes leading to endocyclic epoxida-
tion, organic peracids typically yield a 1:1 cis/trans-LimO mixture
[14], whereas N-sulfonyloxaziridines show stereoselectivities in
favor of cis-LimO [15]. In the extensive literature on metal catalyzed
∗
Corresponding author at: Université de Toulouse, Institut Universitaire de Tech-
nologie Paul Sabatier – Département de Chimie, Av. Georges Pompidou, BP 20258,
F-81104, Castres Cedex, France.
E-mail address: dominique.agustin@iut-tlse3.fr (D. Agustin).
http://dx.doi.org/10.1016/j.mcat.2017.09.033
2468-8231/© 2017 Elsevier B.V. All rights reserved.

W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
53
Scheme 1. Oxidation of limonene and potential LimO and LimD products.
Lim epoxidation, the cis/trans stereoselectivity is rarely indicated. A
variety of mononuclear molybdenum catalysts in combination with
the TBHP oxidant were reported to yield equimolar amounts of the
two isomers [16], while a greater stereoselectivity in favor of the
trans isomer has been recently reported for a dinuclear Mo catalyst
. A mononuclear iron catalyst yields predominantly the cis-LimO
product [17].
Concerning the selectivity of the ring opening step leading from
LimO to LimD, it has been reported that both cis- and trans-LimO
lead to the same ax-LimD product under acidic conditions (pH 3)
[18], which was attributed to a selective axial nucleophilic attack
that can be rationalized by the Fürst-Plattner rule [19]. However,
at an only slightly higher pH (4), only cis-LimO is opened stere-
ospecifically to yield ax-LimD while trans-LimO remains unreacted
[18]. Further studies have shown that the stereospecific opening of
the cis-LimO isomer to ax-LimD also occurs in the presence of lan-
thanide [20] or Mo [21] catalysts. This ring opening selectivity has
been exploited in some cases as a way to generate isomerically pure
ax-LimD and/or to kinetically accumulate pure trans-LimO from the
isomer mixture. The opposite preference for the trans-LimO ring
opening has been accomplished by use of other conditions such as
the use of nucleophilic amines [22], yielding again a trans-diaxial
addition product, or water in the presence of mercuric reagents at
pH 7, which leads, after treatment of an alkoxymercury intermedi-
ate with NaBH4, to the stereospecific generation of the eq-LimD
product [18,23]. The LimO ring opening has also been explored
with the use of enzymes. Using the Rhodotorula glutinis hydrolytic
enzyme, optically pure (1S,2R,4S)-cis-LimO was more rapidly and
stereospecifically hydrolyzed to (1S,2R,4S)-ax-LimD, whereas opti-
cally pure (1S,2R,4R)-trans-LimO was opened more slowly to yield a
mixture of (1R,2R,4R)-eq-LimD (65%) and (1S,2S,4R)-ax-LimD (35%)
[24]. By direct enzymatic oxidation of Lim using a chloroperoxi-
dase enzyme from Caldariomyces fumago, the LimD products were
directly generated without detection of the LimO intermediates
[25]. Only ax-LimD was selectively obtained when the procedure
was carried out in the absence of KCl and proposed to result from
the stereospecific opening of cis-LimO, whereas addition of KCl
yields a mixture of ax-LimD and eq-LimD, proposed to derive from
a mixture of cis-LimO and trans-LimO. These enzymatic studies
are apparently the only ones highlighting the direct formation of
eq-LimD by LimO ring opening.
Scheme 2. Pre-catalysts used in this study.
then further applied to terpenic compounds [31]. A particular class
of compounds that we have explored in some detail is the sali-
cylidene amino(thio)phenolate family shown in Scheme 2. It was
therefore of interest to apply these precatalysts to the Lim epox-
idation under simpler and cleaner operating conditions, without
organic solvents and with aqueous TBHP as oxidant. Preliminary
results on the greener epoxidation of Lim using a few related Mo
catalysts, yielding a mixture of cis-LimO, trans-Limo and ax-LimD,
have been included in a recent publication . Here, we report a fuller
investigation demonstrating how the ligand nature, particularly the
presence of the sulfur atom or an OH group in the R1 position, affects
the selectivity and the kinetics of the various transformations.
Results and discussion
Catalyzed limonene oxidation
The results of initial catalytic tests, run with all the pre-catalysts
shown in Scheme 2, are collected in Table 1. In the absence of cata-
lyst (run 1), aqueous TBHP leads only to a 7.6% conversion at 80 ◦C
after 4 h, with detected traces of LimOs and LimDs. Using a 0.5% cat-
alyst loading, Lim conversions of 79% (after 4 h with [MoO2(SAP)]2,
run 2) and 93% (after 30 min with [MoO2(SATP)]2, run 3) demon-
strated the catalytic activity of both complexes as well as the greater
activity of the sulfur-containing complex. The conversion monitor-
ing for runs 1, 2 and 3 is shown in Fig. 1. The much higher activity for
the catalyst with the ONS coordination sphere in Lim oxidation par-
allels that observed previously for the epoxidation of cyclooctene
.
Monitoring the evolution of the various products in runs 2 and 3
revealed additional interesting features (see Fig. 2). For the exper-
iment with [MoO2(SAP)]2 (Fig. 2, left), the two LimO isomers are
initially produced in a ca. 1:1 ratio (13.8% and 15.3% after 60 min),
but subsequently the amount of cis-LimO slowly decreased (2.6%
after 4 h) while that of trans-LimO continued to increase (19.8%
after 4 h). These data indicate that the rate of cis-LimO ring opening
exceeds that of its generation at longer reaction times, while trans-
LimO opens the ring more slowly. At the same time, the ax-LimD
amount increased from 4.7% to 26.5% while only traces of eq-LimD
It is also pertinent to highlight that the production of LimO and
LimD has typically been accomplished using non-green conditions
(oxidants such as peracids [14] and/or organic solvents [26]; or
use of mercuric reagents for separation as highlighted above). In
order to develop a greener access to these compounds, safer oxi-
dants and non-hazardous conditions are required [27]. Recently,
we have reported the solvent-free epoxidation of cyclooctene and
cyclohexene catalyzed by different molybdenum [28] and vana-
dium [29] complexes or by polyoxometalates [30]. The concept was

54
W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
Table 1
Oxidation of limonene in the presence of various Mo catalysts.a
Selectivity (%)b
LimO
TON
c
Highest TOF (h−1
)
Run
Catalyst
Time (h)
Lim Conv
(%)
LimD
Cis
trans
eq
ax
1
2
3
4
5
6
–
4
4
0.5
4
4
7.6
79
93
90
68
77
5.3
14.6
0
2.1
19.0
10.9
14.5
25.1
0
4.2
29.0
20.8
0.4
2.3
12.4
4.1
1.3
3.6
6.6
–
–
[MoO2(SAP)]2
[MoO2(SATP)]2
[MoO2La]2
[MoO2Lb]2
[MoO2Lc]2
33.5
40.6
44.3
30.3
35.7
158
185
180
136
154
130
371
308
77
4
110
a
b
c
Reaction conditions: [Mo]/limonene/TBHP = 0.5/100/200, T = 80 ◦C.
Select = nproduct formed/nLim transformed obtained from GC analysis of the organic phase.
TOF is calculated using the time interval with maximum slope in the conversion plot.
faster
ax-LimD
eq-
cis-LimO
75
50
25
0
?
Lim
trans-LimO
slower
Scheme 3. Selectivities for the LimO ring opening.
0
60
120 180
diol products cannot be established from these data (see Scheme 3),
but will be detailed later.
Fig. 1. Kinetic profiles for limonene oxidation: without catalyst (ꢀ), with
Concerning the experiment with [MoO2(SATP)]2 (Fig. 2, right),
only traces of the intermediate epoxide compounds were detected,
even at short reaction times, whereas the two isomeric diols were
rapidly produced in large amounts. A greater amount of eq-LimD is
generated relative to run 2, although the ax-LimD isomer remains
the prevalent product. These results suggest that not only the
epoxidation step but also the ring opening step is metal cat-
alyzed. These results are in line with a previous report, in which
a [MoO2Cl2L] complex (L = ␤-ketophosphonate derived from cam-
phor) was shown to catalyze the stereospecific ring opening of
cis-LimO to yield ax-LimD, whereas trans-LimO remained essen-
tially unreacted [21]. The catalytic activity of [MoO2(SAP)]2 and
[MoO2(SATP)]2 in the LimO ring opening has been further con-
firmed by separate investigations of the ring opening of cis- and
trans-LimO by water (see further details in the next section).
The mass balance (cis-LimO + trans-LimO + eq-LimD + ax-
LimD + residual Lim) is low in both experiments: 80% after 2 h
[MoO2(SAP)]2
(
), with [MoO2(SATP)]2 (᭿). [Mo]/limonene/TBHP = 0.5/100/200,
T = 80 ◦C.
were detected initially and significant amounts are only formed
after 3 h.
The initial observation of ax-LimD as the predominant diol in
solution while only the cis-LimO intermediate is being consumed
at significant rate suggests that the faster ring opening of cis-LimO
leads selectively to this product. This observation is consistent with
previous literature reports. On the other hand, the slower genera-
tion of eq-LimD suggests that this isomer results from the slower
ring opening of trans-LimO. We are unaware of other contribu-
tions where the direct generation of eq-LimD from metal-catalyzed
Lim epoxidation has been highlighted (see also the Introduction).
Whether the ring opening of trans-LimO is selective or leads to both
Fig. 2. Kinetic profile for the limonene oxidation catalyzed by [MoO2(SAP)]2 (left), [MoO2(SATP)]2 (right): [Mo]/limonene/TBHP = 0.5/100/200, T = 80 ◦C. Mass balance (. . .),
limonene (ᮀ), cis-LimO (
), trans-LimO (ꢁ), eq-LimD ( ), ax-LimD (ꢀ).

W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
55
Fig.
5. Influence
of
temperature
on
the
C
[MoO2(SATP)]2-catalyzed
Fig. 3. Kinetic profile for limonene oxidation: without catalyst (−), with
limonene oxidation: 30 ◦
C
(ꢀ), 50 ◦
C
(♦), 60 ◦
(ᮀ), 80 ◦
C
(ꢁ). Conditions:
[MoO2(SAP)]2
(
), with [MoO2La]2(
), with [MoO2Lb]2
(
), with [MoO2Lc]2(
[Mo]/limonene/TBHP = 0.5/100/200.
). [Mo]/limonene/TBHP = 0.5/100/200, reaction time = 4 h, T = 80 ◦C.
The kinetic profiles for the same oxidation processes in the
presence of the precatalysts [MoO2La–c
(runs 4–6) are shown
and 70% after 4 h for run 2; 56% after 0.5 h for run 3. This loss of
material can be attributed to the further oxidative transformations
of the LimD isomers to non-volatile products. Traces of carvone
and carveol, plus other unidentified products, were also visible in
the gas-chromatogram for both runs and were not quantified but
could be estimated as <5%, hence could not account for the lack of
mass balance. The final reaction mixture consisted of a single phase
in each case, presumably because the organic compounds and the
water originating from the TBHP reagent solution, which generates
two phases at the beginning of the reaction, are progressively
compatibilized by the LimD and tBuOH products.
]
2
in Fig. 3 in comparison with those of runs 1 and 2. The con-
version after 4 h was very good for each catalyst (see Table 1),
in the order [MoO2La]2 (90%) > [MoO2(SAP)]2 (79%) ∼ [MoO2Lc]2
(77%) > [MoO2Lb]2 (68%). The [MoO2La]2 complex (R1 = OH) exhib-
ited a much higher activity than the other pre-catalysts. The initial
cis/trans ratio for the LimO intermediates for the experiments with
the [MoO2Lb]2 and [MoO2Lc]2 complexes was similar (ca. 1:1, see
Fig. 4) to that observed for the experiment with [MoO2(SAP)]2
(Fig. 2), whereas the most active [MoO2La]2 catalyst yields an initial
cis/trans ratio of ca. 1:2 (Fig. 4). Once produced, the LimO isomers
Fig. 4. Kinetic profile for the limonene oxidation catalyzed by [MoO2La]2 (up), [MoO2Lb]2 (down left), [MoO2Lc]2 (down right),: [Mo]/limonene/TBHP = 0.5/100/200, T = 80 ◦C.
Mass balance (. . .), limonene (ᮀ),-LimO cis-LimO (
), trans-LimO (ꢁ), eq-LimD ( ), ax-LimD (ꢀ).

56
W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
Fig. 6. Influence of the catalyst loading on the [MoO2(SATP)]2-catalyzed limonene oxidation at 50 ◦C (left) and 80 ◦C (right): [Mo]/limonene/TBHP = ×/100/200, x = 0 (♦), 0.1
(ꢁ), 0.25 (ᮀ), 0.5 (ꢀ).
Fig. 7. Kinetic profile for the limonene oxidation catalyzed by [MoO2(SATP)]2 at 30 ◦C (left) and 50 ◦C (right): [Mo]/limonene/TBHP = 0.5/100/200. Mass balance (. . .), limonene
(ᮀ), cis-LimO (
), trans-LimO (ꢁ), eq-LimD ( ), ax-LimD (ꢀ).
are consumed to yield the two LimD products, mainly the axial
isomer. The ring opening data are also in line with the behavior
reported above for the parent [MoO2(SAP)]2 pre-catalyst: rapid for
cis-LimO with selective formation of ax-LimD and slower for trans-
LimO (see further details later). In addition, the LimD selectivity
correlates with the Lim conversion, with the relative amount of
the minor eq-LimD product increasing with the rate of Lim con-
version. These data further confirm that the epoxide ring opening
process is, like the epoxidation, metal-catalyzed and that the bet-
ter epoxidation catalyst is also a better ring opening catalyst. The
same trends for the epoxide/diol selectivity vs. the rate of con-
version, although quantitatively less pronounced, have also been
previously observed for the oxidation of cyclohexene catalyzed by
the same ONO complexes but no conclusion concerning the epox-
ide ring opening catalysis was made . The Green metrics for these
catalyzed reactions have been analyzed in terms of the influence of
catalyst and temperature on the LimO and LimD production (See
SI).
The results of the individual intermediates and products moni-
toring at 30 and 50 ◦C are shown in Fig. 7. At the end of the catalytic
runs, the LimO intermediates (mainly the trans isomer) could be
detected only at temperatures up to 50◦ when in the presence
of catalyst. Interestingly, while the cis-LimO was the main epox-
ide product in the absence of catalyst at 50 ◦C, the selectivity was
inverted in the presence of catalyst in favor of trans-LimO, proba-
bly because of the faster catalyzed ring opening of cis-LimO opening
(mainly yielding ax-LimD as already discussed above). The highest
eq-LimD yield was observed in the presence of catalyst at 80 ◦C (run
15).
The effect of the TBHP/substrate ratio has also been investigated,
both at 50 ◦C and at 80 ◦C, with a 0.5% loading of the [MoO2(SATP)]
catalyst (Table 3). Lowering the TBHP amount leads, as expected,
to a slower conversion (cf. run 16 with run 11 of Table 2 and run 17
with run 15 of Table 2). It is of interest to note that this also results
in a slight decrease of the eq-LimD selectivity and an increase of the
ax-LimD selectivity.
In order to better understand the catalytic epoxidation and
epoxide ring opening for this substrate, additional experiments
were carried out for the more active [MoO2(SATP)]2 systems at
variable temperature and catalyst loading (Table 2 and Figs. 5–7). A
temperature increase without catalyst leads to greater Lim trans-
formation although the conversion and selectivity remained poor
(runs 9 and 12). With 0.5% molar ratio [MoO2(SATP)] and 2 eq TBHP
per substrate in the 30–80 ◦C range, the Lim conversion after 4 h
increased dramatically, from 36% at 30 ◦C after 5 h to 93% at 80 ◦C
after 30 min (runs 8, 11 and 15; Fig. 5). At both 50 ◦C (runs 6–11)
and 80 ◦C (runs 12–15) the conversion becomes faster as expected
upon increasing the catalyst loading, see also Fig. 6.
Catalyzed limonene oxide ring opening
As shown in the previous section, the Lim epoxidation produced
LimOs and/or LimDs in a catalyst- and temperature-dependent pro-
cess and there is clear evidence that the more active epoxidation
catalysts (e.g. the SATP complex) also have higher activity in the
ring opening process. It was therefore of interest to carry out more
detailed investigations on the catalyzed epoxide ring opening pro-
cess. For this purpose, both a commercial LimO isomer mixture
(cis/trans = 47:53) and the pure individual isomers have been inves-
tigated (Table 4).

W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
57
Table 2
Influence of temperature and catalyst loading for limonene epoxidation catalyzed by [MoO2(SATP)]2.a
T (◦C)
time (h)
X (%)
Lim Conv.
(%)
c
TOF (h−1
)
Run
Selectivity (%)b
LimO
TON
LimD
eq
Cis
trans
ax
7
8
30
50
5
5
0
0.5
3.2
36
9.4
3.7
9.4
32.2
1.6
1.7
5.6
35.3
–
70
–
26
9
10
11
0
0.1
0.5
5.3
62.5
83
10.4
0.3
0.1
0.6
20.3
10.4
0.6
4.6
5.8
5.8
42.2
50
627
166
312
161
12
13
14
15
80
4
2
0.5
0.5
0
7.7
85
87
93
5.2
0
0
14.3
0.4
7.9
11.4
12.4
6.5
0.1
0.25
0.5
0
0
0
72.9
46.3
40.6
850
348
185
1042
1743
371
0
a[Mo]/limonene/TBHP = ×/100/200.b
bSelect = nproduct formed/nLim transformed obtained from GC analysis of the organic phase.
cTOF is calculated using the time interval with maximum slope in the conversion plot.
Table 3
Influence of the TBHP amount on the [MoO2(SATP)]2-catalyzed limonene epoxidation.a
Selectivity (%)b
LimO
TON
T (◦C)
Time (h)
Conversion
(%)
TOFc(h−1
)
y
Run
LimD
cis
trans
eq
ax
16
17
50
80
5
0.5
110
150
61
89
0
0
4.9
0
9.5
12.2
58.5
47.4
123
178
109
357
a
Conditions: [Mo]/limonene/TBHP = 0.5/100/y. Reaction time = 4 h. T = 50 ◦C.
Select = nproduct formed/nLim transformed obtained from GC analysis of the organic phase.
TOF is calculated using the time interval with maximum slope in the conversion plot.
b
c
Table 4
Ring opening of limonene oxides catalyzed by molybdenum complexes.a
LimDselectivityb (%)
TON
c
TOF (h−1
)
Run
Conv (%)
substrate
Time (h)
catalyst
cis
trans
eq
ax
18
19
20
4
4
0.5
No catalyst
[MoO2(SAP)]2
[MoO2(SATP)]2
60.7
93.3
100
12.0
26.2
100
11.7
2.6
12.1
58.7
58.5
45.0
–
119
200
–
45
400
LimO
mixtured
21
22
23
4
4
0.5
No catalyst
[MoO2(SAP)]2
[MoO2(SATP)]2
13.5
26.9
100
0
0
0
6.4
28.4
25.3
–
52
200
–
27
400
cis-LimO
24
trans-LimO
1.5
[MoO2(SATP)]2
100
15.8
20.8
200
130
a
[Mo]/limonene/TBHP = 0.5/100/200; T = 80 ◦C.
Select = nproduct formed/nLimO(cis+trans) transformed obtained from GC analysis of the organic phase.
TOF is calculated using the time interval with maximum slope in the conversion plot.
Commercially available sample (cis/trans = 47/53).
b
c
d
The epoxidation of the trans/cis LimO mixture confirms that
The separate investigations of pure cis-LimO and trans-LimO
provide additional useful information on the stereochemistry of
the ring opening process. None of the eq-LimD was generated
by the ring opening of cis-LimO, either with or without catalyst
(runs 21–23). Thus, both non-catalyzed and catalyzed cis-LimO ring
openings lead stereospecifically to the ax-LimD isomer, in agree-
ment with previous reports.18,21,24 However, the ring opening of
pure trans-LimO (run 24) provides a mixture of ax-LimD and eq-
LimD in a ca. 4:3 ratio. To the best of our knowledge, this is the
first report of the metal-catalyzed opening of trans-LimO. Inter-
estingly, the stereoselectivity of this Mo-catalyzed ring opening
process is quite similar to that reported for the opening catalyzed
by the hydrolytic enzyme Rhodotorula glutinis (65:35) [24].
cis-LimO is opened faster than trans-LimO, whether in the pres-
ence or absence of catalyst. The rate of the transformation
increases in the order: no catalyst (run 18) < [MoO2(SAP)]2 (run
19) < [MoO2(SATP)]2 (run 20). Both LimO isomers are opened faster
by the more active [MoO2(SATP)]2 catalyst (cf. runs 19 and 20). The
greater catalytic activity of the ONS ligand in ring opening is con-
firmed by the study of the isomerically pure cis-LimO (cf. run 22 and
23), which results in the stereospecific production of ax-LimD. Only
the most active [MoO2(SATP)]2 catalyst was tested in the opening
of pure trans-LimO (run 24): in spite of the more sluggish ring open-
ing of this isomer, quantitative conversion could be achieved after
1.5 h. This run confirms that the eq-LimD isomer results from the
opening of trans-LimO, but the product ratio is still in favor of the
ax-LimD isomer. The mass balance suggests that either other reac-
tions also occur, or that the LimD products are partially transferred
in aqueous phase.
In terms of mechanistic interpretation, the generation of the eq-
LimD product cannot result by an effect of the Mo complex identical
to that proposed for Hg2+ in the selective generation of this diol
from trans-LimO, for two reasons: i) the Hg2+ reagent accelerates

58
W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
Scheme 4. Proposed mechanism for the Mo-catalyzed ring opening of the LimO isomers.
the opening of trans-LimO relative to cis-LimO, whereas the Mo cat-
alyst leads to faster opening of the cis isomer; ii) the Hg2+-promoted
opening of the trans-LimO is stereospecific, leading solely to eq-
LimD, whereas the Mo catalyst generates a mixture of both LimD
isomers. Indeed, the action of Hg2+ was rationalized though a con-
formational change induced by the interaction between the soft
Hg2+ ion and the isopropenyl substituent, which is not possible
for the Lewis-acidic MoVI center because this d0 metal ion has no
affinity toward C C unsaturations. We can rather invoke a mecha-
nistic scheme with formation of MoVI-glycolate intermediates (see
Scheme 4), based on that proposed by Cole-Hamilton et al. and sup-
ported by NMR investigations of the catalyst-substrate interactions
[21]. Activation of the epoxide by the Lewis acidity of the mononu-
clear 5-coordinate [MoO2L] complex would result in equilibria with
a molybdate-substituted carbocation and a Mo-glycolate interme-
diate (A and B from the cis isomer; C and D for the trans one). These
intermediates are then attacked by water at the more substituted
C atom with inversion of configuration. The smaller catalytic activ-
ity of the previously used camphor-based [MoO2Cl2L] catalyst [21]
did not allow the stereoselectivity of the water addition to D to
be established. Indeed, formation of intermediate D itself was not
considered possible. We now find that the greater catalytic activity
of [MoO2(SATP)] (and to a certain extent also the other catalysts
containing the SAP, La, Lb and Lc ligands) allows sufficiently rapid
access to this intermediate. While the reactivity of the tertiary C
atom in intermediate B dominates and leads to the selective gener-
ation of the trans-diaxial ax-LimD product, the lower reactivity of
the same ring C atom in intermediate D allows competitive attack
at the secondary C atom with production of a mixture of the two
LimD products.
Experimental part
Materials and methods
All manipulations were carried out in air. Water was deion-
ized twice before use. Organic solvents (ethanol, methanol,
diethylether: synthesis grade, Aldrich) were employed as received
without any purification. All molybdenum complexes were syn-
thesized as previously described . Limonene (98% Aldrich), the
LimO mixture (Aldrich cis/trans 47/53, 98%), ax-LimD (Aldrich,
98%)) and TBHP (70% in water, ACROS) were used as received. The
pure cis-LimO, trans-LimO and eq-LimD were synthesized accord-
ing to literature procedure [18,22,23]. The catalytic reactions were
monitored by gas chromatography on an Agilent 6890A chromato-
graph equipped with an FID detector, a HP5-MS capillary column
(0.30 m × 0.25 mm × 0.25 m) and an autosampler, or on a Fisons GC
8000 chromatograph equipped with an FID detector and with a
SPB-5 capillary column (30 m × 0.32 mm × 0.25 m). Authentic sam-
ples of all reactants and products were used for calibration. The Lim
conversion and the formation of LimOs and limDs were calculated
from the calibration curves (r2 = 0.999) and an internal standard.
Catalytic procedure
In a typical experiment, limonene (1 equiv) and catalyst (x equiv,
see tables) were mixed together in air in a round bottom flask.
Acetophenone was added as internal standard. The flask was then
immersed in a thermostated bath set at the desired temperature
(see Tables) and magnetically stirred. After thermal equilibration,
aqueous THBP (70% in water, y equiv see tables) was added to
the mixture, starting the reaction. Samples were periodically with-
drawn and quenched by the addition of MnO2, followed by the
addition of diethylether and removal of the manganese oxide and
residual water by filtration through silica before GC analysis.
Conclusions
Lim epoxidation is catalyzed very efficiently by [MoO2L]2
with salicylidene amino(thio)phenolate ligands, particularly
[MoO2(SATP)]2. These complexes promote not only the Lim regios-
elective endocylic epoxidation to yield both cis- and trans-LimO
isomers in a kinetically controlled 1:1 ratio, but also the subse-
quent epoxide ring opening in the presence of water to obtain a
mixture of ax- and eq-LimD. The high activity of the [MoO2(SATP)]2
precatalyst promotes the efficient ring opening of the less reactive
trans-LimO isomer, which is shown to proceed unselectively to
a mixture of ax- and eq-LimD by analogy with a previous report
of hydrolytic enzymatic ring opening, whereas the more reactive
cis-LimO is stereospecifically opened to ax-LimD.
Acknowledgements
We acknowledge the Centre National de la Recherche Scien-
tifique (CNRS) and the Institut Universitaire de France (IUF) for
financial support, the Université Paul Sabatier and its Institut
Universitaire Technologique for the facilities and the Chinese Schol-
arship Council for the fellowship of W.W. No.201208370239.

W. Wang et al. / Molecular Catalysis 443 (2017) 52–59
59
Appendix A. Supplementary data
[12] (a) C.K. Modi, J.A. Chudasam, H.D. Nakum, D.K. Parmar, A.L. Patel, J. Mol. Catal.
A Chem. 395 (2014) 151–161;
(b) A.M.A. Santa, J.C.G. Vergara, S. Palacio, I. Catal. Today 133–135 (2008)
80–86.
[13] M. Blair, K.L. Tuck, Tetrahedron: Asymmetry 20 (2009) 2149–2153.
[14] (a) W.F. Newhall, J. Org. Chem. 24 (1959) 1673–1676;
(b) W.F. Newhall, J. Org. Chem. 29 (1964) 185–187;
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.mcat.2017.09.
033.
(c) W. Knoll, C. Tamm, Helv. Chim. Acta 58 (1975) 1162–1171.
[15] (a) F.A. Davis, J.C. Towson, D.B. Vashi, R. Thimmareddy, J.P. McCauley, M.E.
Harakal, D.J. Gosciniak, J. Org. Chem. 55 (1990) 1254–1261;
(b) F.A. Davis, R. Thimmareddy, J.P. McCauley, R.M. Przeslawski, M.E. Harakal,
P.J. Carroll, J. Org. Chem. 56 (1991) 809–815.
[16] (a) R. Clarke, M. Gahagan, R.K. Mackie, D.F. Foster, D.J. Cole-Hamilton, M.
Nicol, A.W. Montford, J. Chem. Soc., Dalton Trans. (1995) 1221–1226;
(b) J.A. Brito, S. Ladeira, E. Teuma, B. Royo, M. Gomez, Appl. Catal. A 398 (2011)
88–95.
References
[1] (a) G.A. Cordell, T.L.G. Lemos, F.J.Q. Monte, M.C. de Mattos, J. Nat. Prod. 70
(2007) 478–492;
L. Filiciotto, A. M. Balu, J. C. Van der Waal, R. Luque, Catal Today (2017) proofs
online http://dx.doi.org/10.1016/j.cattod.2017.03.008;
(b) A. Arbenz, L. Averous, Green Chem. 17 (2015) 2626–2646;
(c) I.K.M. Yu, D.C.W. Tsang, Bioresour. Technol. 238 (2017) 716–732;
(d) A.K. Kumar, S. Sharma, Bioresources and Bioprocessing 4 (2017) 7;
(e) G. Chatel, K. De Oliveira Vigier, F. Jérôme, ChemSusChem 7 (2014)
2774–2787;
(f) P. Spannring, P.C.A. Bruijnincx, B.M. Weckhuysen, R.J.M. Klein Gebbink,
Catal. Sci. Technol. 4 (2014) 2182–2209;
(g) C.O. Tuck, E. Perez, I. Horvath, R. Sheldon, M. Poliakoff, Science 337 (2012)
695–699;
(h) J. Kudakasseril Kurian, G. Raveendran Nair, A. Hussain, G.S. Vijaya
Raghavan, Renew. Sust. Energ. Rev. 25 (2013) 205–219;
(j) L. Wu, T. Moteki, A.A. Gokhale, D.W. Flaherty, F.D. Toste, Chemistry 1
(2016) 32–58;
(k) F.D. Pileidis, M.-M. Titirici, ChemSusChem 9 (2016) 562–582;
(l) V. Negro, G. Mancini, B. Ruggeri, D. Fino, Bioresour. Technol. 214 (2016)
806–815;
(m) E.V. Gusevskaya, ChemCatChem 6 (2014) 1506–1515;
(n) T.I.J. Dugmore, J.H. Clark, J. Bustamante, J.A. Houghton, A.S. Matharu, Top.
Curr. Chem. 375 (2017) 46;
(o) M.P. Adhikari, R. Adhikari, R.G. Shrestha, R. Rajendran, L. Adhikari, P. Bairi,
R.R. Pradhananga, L.K. Shrestha, K. Ariga, Bull. Chem. Soc. Jpn. 88 (2015)
1108–1115;
[17] D. Clemente-Tejeda, A. López-Moreno, F.A. Bermejo, Tetrahedron 69 (2013)
2977–2986.
[18] M. Blair, P.C. Andrews, B.H. Fraser, C.M. Forsyth, P.C. Junk, M. Massi, K.L. Tuck,
Synthesis (2007) 1523–1527 (and references cited therein).
[19] (a) A. Fürst, P.A. Plattner, Helv. Chim. Acta 32 (1949) 275–283;
(b) E.E. Royals, J.C. Leffingwell, J. Org. Chem. 31 (1966) 1937–1944.
[20] P.C. Andrews, M. Blair, B.H. Fraser, P.C. Junk, M. Massi, K.L. Tuck, Tetrahedron:
Asymmetry 17 (2006) 2833–2838.
[21] L. Salles, A.F. Nixon, N.C. Russell, R. Clarke, P. Pogorzelec, D.J. Cole-Hamilton,
Tetrahedron: Asymmetry 10 (1999) 1471–1476.
[22] D. Steiner, L. Ivison, C.T. Goralski, R.B. Appell, J.R. Gojkovic, B. Singaram,
Tetrahedron: Asymmetry 13 (2002) 2359–2363.
[23] M.J. van der Werf, H. Jongejan, M.C.R. Franssen, Tetrahedron Lett. 42 (2001)
5521–5524.
[24] C. Weijers, Tetrahedron: Asymmetry 8 (1997) 639–647.
[25] S. Aguila, R. Vasquez-Duhalt, R. Tinoco, M. Rivera, G. Pecchi, J.B. Alderete,
Green Chem. 10 (2008) 647–653.
[26] (a) A.A. Valente, I.S. Goncalves, A.D. Lopes, J.-E. Rodriguez-Borges, M. Pillinger,
C.C. Romao, J. Rocha, X. Garcia-Mera, New J. Chem. 25 (2001) 959–963;
(b) P. Oliveira, M.L. Rojas-Cervantes, A.M. Ramos, I.M. Fonseca, A.M. Botelho
do Rego, J. Vital, Catal. Today 118 (2006) 307–314;
(p) V.S. Sikarwar, M. Zhao, P. Clough, J. Yao, X. Zhong, M.Z. Memon, N. Shah,
E.J. Anthony, P.S. Fennell, Energy Environ. Sci. 9 (2016) 2939–2977;
(q) A. Scoma, S. Rebecchi, L. Bertin, F. Fava, Crit. Rev. Biotechnol. 36 (2016)
175–189.
(c) N. Mizuno, S. Hikichi, K. Yamaguchi, S. Uchida, Y. Nakagawa, K. Uehara, K.
Kamata, Catal. Today 117 (2006) 32–36;
(d) C.C.L. Pereira, S.S. Balula, F.A. Almeida Paz, A.A. Valente, M. Pillinger, J.
Klinowski, I.S. Goncalves, Inorg. Chem. 46 (2007) 8508–8510;
(e) T.R. Amarante, P. Neves, C. Tomeı´, M. Abrantes, A.A. Valente, F.A. Almeida
Paz, M. Pillinger, I.S. Gonc¸ alves, Inorg. Chem. 51 (2012) 3666–3676;
(f) T.R. Amarante, P. Neves, A.A. Valente, F.A. Almeida Paz, A.N. Fitch, M.
Pillinger, I.S. Gonc¸ alves, Inorg. Chem. 52 (2013) 4618–4628;
(g) Z. Strassberger, E.V. Ramos-Fernandez, A. Boonstra, R. Jorna, S. Tanase, G.
Rothenberg, Dalton Trans. 42 (2013) 5546–5553;
(h) A. Gallo, C. Tiozzo, R. Psaro, F. Carniato, M. Guidotti, J. Catal. 298 (2013)
77–83;
(i) J. Mlodzik, A. Wroblewska, E. Makuch, R.J. Wrobel, B. Michalkiewicz, Catal.
Today 268 (2016) 111–120;
[2] (a) W. Schwab, C. Fuchs, F.-C. Huang, Eur. J. Lipid Sci. Technol. 115 (2013) 3–8;
(b) N. Ravasio, F. Zaccheria, M. Guidotti, R. Psaro, Top. Catal. 27 (2004)
157–168;
(c) J.L.F. Monteiro, C.O. Veloso, Top. Catal. 27 (2004) 169–180.
[3] G.A. Burdock, Fenaroli’s Handbook of Flavor Ingredients, 3rd ed., CRC Press,
Boca Raton, 1995, p.107.
[4] (a) R.A. Sheldon, in: F.E. Harkes (Ed.), Catalysis of Organic Reaction, USA, 1998;
(b) R.A. Sheldon, in: G. Centi, F. Trifiro (Eds.), New Developments in Selective
Oxidation, Elsevier, Amsterdam, 1990.
[5] R. Ciriminna, M. Lomeli-Rodriguez, P. Demma Cara, J.A. Lopez-Sanchez, M.
Pagliaro, Chem. Commun. 50 (2014) 15288–15296.
(j) A. Wroblewska, E. Makuch, P. Miadlicki, Catal. Today 268 (2016) 121–129.
[27] R.A. Sheldon, Green Chem. 7 (2005) 267–278.
[28] (a) J. Morlot, N. Uyttebroeck, D. Agustin, R. Poli, ChemCatChem 5 (2013)
601–611;
[6] (a) K.A.D. Swift, Top. Catal. 27 (2004) 143–155.
[7] (a) J.M. Derfer, M.M. Derfer, Encyclopedia of Chemical Technology
Kirk-Othmer, vol. 22, third ed., Wiley, New York, 1978, p. 709;
(b) E. Herrero, S. Casuscelli, J. Fernandez, C. Poncio, M. Rueda, O. Oyola,
Molecules 5 (2000) 336–337;
(c) K. Bauer, D. Garbe, H. Surburg, Common Fragrance and Flavor Materials,
Preparation, Properties and Uses, Wiley VCH, New York, 1997.
[8] (a) D. Comins, L. Guerra-Weltzien, J.M. Salvador, Synlett (1994) 972–974;
(b) W. Chrisman, J.N. Camara, K. Marcellini, B. Singaram, C.T. Goralski, D.L.
Hasha, P.R. Rudolf, L.W. Nicholson, K.K. Borodychuk, Tetrahedron Lett. 42
(2001) 5805–5807;
(c) H. Lebel, E.N. Jacobsen, J. Org. Chem. 62 (1998) 9624–9625.
[9] (a) S.K. Kulkarni, A. Dhir, Prog. Neuropsychopharmacol. Biol. Psychiatry 31
(2007) 1248–1254;
(b) T. Umezu, Pharmacol. Biochem. Behav. 64 (1999) 35–40;
(c) A.A.C. de Almeida, J. Pereira Costa, R.B.F. de Carvalho, D. Pergentino de
Sousa, R. Mendes de Freitas, Brain Res. 1448 (2012) 56–62.
[10] (a) C.M. Byrne, S.D. Allen, E.B. Lobkovsky, G.W. Coates, J. Am. Chem, Soc 126
(2004) 11404–11405;
(b) W. Wang, T. Vanderbeeken, D. Agustin, R. Poli, Catal. Commun. 63 (2015)
26–30;
(c) W. Wang, J.-C. Daran, R. Poli, D. Agustin, J. Mol. Catal. A 416 (2016)
117–126;
(d) J. Pisk, D. Agustin, J.C. Daran, V. Vrdoljak, R. Poli, Adv. Synth. Catal. 353
(2011) 2910–2914;
(e) J. Pisk, B. Prugovecˇki, D. Matkovic´-Calogovic´, R. Poli, D. Agustin, V.
ˇ
Vrdoljak, Polyhedron 33 (2012) 441–449;
ˇ
(f) J. Pisk, B. Prugovecˇki, D. Matkovic´-Calogovic´, T. Jednacˇak, P. Novak, D.
Agustin, V. Vrdoljak, RSC Adv. 4 (2014) 39000–39010;
(g) V. Vrdoljak, J. Pisk, D. Agustin, P. Novak, J. Parlov Vukovic´, D.
ˇ
Matkovic´-Calogovic´, New J. Chem. 38 (2014) 6176–6185.
[29] (a) C. Cordelle, D. Agustin, J.C. Daran, R. Poli, Inorg. Chim. Acta 364 (2010)
144–149;
(b) J. Pisk, J.-C. Daran, R. Poli, D. Agustin, J. Mol. Catal. A: Chem. 403 (2015)
52–63.
(b) A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502;
(c) O. Hauenstein, M. Reiter, S. Agarwal, B. Rieger, A. Greiner, Green Chem. 18
(2016) 760–770;
(d) E. Hosseini Nejad, A. Paoniasari, C.G.W. van Melis, C.E. Koning, R.
Duchateau, Macromolecules 46 (2013) 631–637.
[30] B. Guérin, D. Mesquita Fernandes, J.-C. Daran, D. Agustin, R. Poli, New. J. Chem.
37 (2013) 3466–3475.
[31] (a) M. Loubidi, D. Agustin, A. Benharref, R. Poli, C. R. Chim. 17 (2014) 549–556;
(b) M. Cindric, G. Pavlovic, R. Katava, D. Agustin, New J. Chem. 41 (2017)
594–602.
[11] Jd. Moraes, A.A. Almeida, M.R. Brito, T.H. Marques, T.C. Lima, D.P. Sousa, E.
Nakano, R.Z. Mendonc¸ a, R.M. Freitas, Planta Med. 79 (2013) 253–258.