Catalyzed ring opening of epoxides: Application to bioplasticizers synthesis

Article abstract of DOI:10.1016/j.apcata.2010.11.020

The ring opening of mono, di or tri-substituted epoxides by acetic anhydride to corresponding diacetates is catalyzed by weak bases such as hydrotalcite in the carbonated form. This reaction is performed at 423 K without solvent and the solid catalyst is reused after simple regeneration for 4 runs with constant conversion. Ring-opening of methyl oleate epoxide leads to the formation of useful diacetate methyl oleate. Starting from vegetable oils, polyacetate derivatives are prepared in three catalytic steps (methanolysis, epoxidation then ring opening). Plastisol was prepared by mixing these products with PVC and their rheological properties were evaluated. The recorded data show that they can act as bioplasticizer with similar behaviour as phthalate reference.

Full text of DOI:10.1016/j.apcata.2010.11.020

Applied Catalysis A: General 393 (2011) 1–8
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Catalyzed ring opening of epoxides: Application to bioplasticizers synthesis
Gabriella Fogassya,b, Pan Kea,b, Franc¸ ois Figuerasa,b, Philippe Cassagnaua,c
,
Sophie Rouzeaud, Valérie Couraultd, Georges Gelbarda,b, Catherine Pinela,b,∗
a Université de Lyon, F-69622, Lyon, France
b Institut de Recherches sur la Catalyse et l’Environnement de Lyon, CNRS-Université Lyon 1, UMR 5256, 2 avenue Albert Einstein, F-69626 Villeurbanne, France
c Laboratoire des Matériaux et Biopolymères, CNRS-Université Lyon 1, UMR, 15 boulevard André Latarjet, 69622 Villeurbanne Cedex, France
d TESF, BP54 38352 La Tour du Pin Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The ring opening of mono, di or tri-substituted epoxides by acetic anhydride to corresponding diacetates
is catalyzed by weak bases such as hydrotalcite in the carbonated form. This reaction is performed at
423 K without solvent and the solid catalyst is reused after simple regeneration for 4 runs with constant
conversion. Ring-opening of methyl oleate epoxide leads to the formation of useful diacetate methyl
oleate.
Received 16 September 2010
Received in revised form 26 October 2010
Accepted 4 November 2010
Available online 13 November 2010
Starting from vegetable oils, polyacetate derivatives are prepared in three catalytic steps (methanolysis,
epoxidation then ring opening). Plastisol was prepared by mixing these products with PVC and their
rheological properties were evaluated. The recorded data show that they can act as bioplasticizer with
similar behaviour as phthalate reference.
Keywords:
Hydrotalcite
Epoxide
Solvent free reaction
Bioplasticizer
PVC
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
an isosorbide diester developed by Roquette, showed excellent
plasticizing properties [5]. Danisco reported a plasticizer produced
Polyvinyl chloride (PVC) is one of the oldest and most com-
mon plastics in use today because of the low cost, durability, and
versatility with respect to fabrication and property modification.
It finds uses in one or both of two general forms: unplasticized
or plasticized PVC [1]. The latter one is widely used for numerous
applications such as household goods, cable covering, toys or as a
substitute for leather and introduced in clothes or various furnish-
ings. Plasticized PVC contains up to 50% plasticizers as an additive
to increase the flexibility, softness, processability and pliability of
the end product. Plasticizers are typically a high boiling, organic
liquid that reduces the glass temperature of the polymer where
the polymer changes from brittle to flexible [2]. Thus, the addi-
tion of plasticizer reduces the tensile strength and elastic modulus
of PVC but increases the elongation at tensile failure at ambient
temperature. The alkyl phthalate family represents the main PVC
plasticizers currently used in industrial processes. However, their
use is restricted in some applications (toys, food contact, medi-
cal devices) because they eventually migrate out of plastics [3,4].
Moreover, they are suspected to disrupt human endocrine activity.
So there is a tendency to the use of alternative plasticizers especially
prepared from renewable raw materials. Recently, Polysorb ID37®,
from castor oil that makes plastic products soft and flexible and that
is fully biodegradable [6]. Tall-oil derivatives were also described
as PVC plasticizers [7].
Large range of applications of oleochemicals in polymers has
been reported mainly as stabilizers or lubricants [8–12]. The cat-
alytic transesterification of vegetable oils with methanol to FAME is
now well established as well as the epoxidation of the unsaturated
fatty derivatives yielding to the production of reactive substrates
that can react with many nucleophiles leading to the opening of the
strained epoxide ring (Scheme 1) [13–16].
In previous communications, we reported the solvent-free ring
opening of epoxides by acetic anhydride catalyzed by quaternary
ammonium salts [17]. The diacetylation of 1,2-epoxydodecane took
place in good to excellent yields in a catalytic reaction in solvent free
conditions. Among the studied catalysts, tetrabutyl-ammonium
chloride and acetate were the best ones reaching full conversion at
373 K after 6 h. The replacement of soluble catalysts by solids would
provide cleaner and safer processes, due to the reduction of salts,
easier separation of the products and recycling of the catalyst which
is necessary for environmental preoccupations. Some functional-
ized polystyrene resins bearing ammonium groups were previously
evaluated, however, they exhibited much lower activities (only 85%
after 48 h). Considering these results, we studied alternative het-
erogeneous catalysts able to perform this transformation. Herein,
we report the development of basic solids to catalyze the ring open-
∗
Corresponding author. Tel.: +33 4 7244 5478.
E-mail address: catherine.pinel@ircelyon.univ-lyon1.fr (C. Pinel).
0926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.11.020

2
G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
O
O
O
O
RO
O
OLE / LIN
O
O
RO
O
O
O
O
OLE-ox / LIN-ox
OAcOAc
OAc
RO
O
OAc
OAcOAc
OLE-ac / LIN-ac
Composition (%)
C16:0 C18:0 C18:1 C18:2 C18:3 Average C=C
OLE
LIN
6
6
2
3
58
32
21
57
9
2
1.2
1.5
Scheme 1. Synthetic route from vegetable oil to bioplasticizer.
ing reaction of oxirane with acetic anhydride. More specifically, we
compare a strong base such as Mg–La-mixed oxides [18,19], with
mild bases such as Mg–Al hydrotalcites [20–22], and the mixed
oxides obtained by calcinations, that are well known as basic cata-
lysts for a broad range of reactions. Furthermore, we wish to present
the development of this ring-opening reaction to epoxides derived
from fatty esters. The influence of the nature of vegetable oils and
the reaction conditions will be detailed as well as some properties
of the product.
2.2. Epoxydation of methyl oleate [24–26]
In a round bottom flask, 3 g of methyl oleate (10 mmol) and
0.1 ml of formic acid (3 mmol) were heated at 40 ◦C. 2.6 ml of 35%
H2O2 (30 mmol) were then added dropwise and the solution was
heated to 3.5 h at 70 ◦C. The solution was cooled to room temper-
ature. The epoxide was extracted by Et2O (30 ml) and the organic
layer was washed with H2O (30 ml), Na2CO3 sat (30 ml) and NaCl sat
(30 ml) and dried (MgSO4). After evaporation, quantitative yield of
epoxide was obtained which was used in the following step without
purification. NMR data are in accordance to literature.
2. Experimental
1,2-Epoxydodecane, methyl oleate and acetic anhydride were
purchased from Aldrich. Novance kindly provided two types of fatty
ester namely Lubrirob 926.65 containing mainly methyl oleate and
Lubrirob 301.01 containing mainly methyl linoleate (Scheme 1).
2.3. Epoxydation of Lubrirob 926.65 (OLE)
In a round bottom flask, 100 g of Lubrirob 926.65 (OLE) and
3.8 ml of formic acid (100 mmol) were introduced. 104 ml of 35%
H2O2 (30 mmol) were then added dropwise (30 min) and the solu-
tion was heated to 10.5 h at 70 ◦C (complete conversion monitored
by GC analysis). The solution was cooled to room temperature. The
epoxide was extracted by Et2O and the organic layer was washed
with H2O (30 ml), Na2CO3 sat (30 ml) and Na2SO3 sat until elimina-
tion of peroxide. The organic phase was dried over MgSO4 and the
solvent was evaporated to yield 86% of pale yellow liquid (OLE-ox)
which was used in the following step without purification.
2.1. Synthesis and characterization of the solid bases
Hydrotalcites (HDT–CO3, Mg/Al = 3) and Mg–La mixed oxides
(Mg/La = 3) were prepared by precipitation of a mixture of nitrate
salts at a constant pH = 10, according to previous reports [23].
Hydrotalcites were activated at 723 K in an air flow using a tem-
perature ramp of 2.5 K/min and maintained at this temperature for
1 h, while Mg–La mixed oxides were calcined at 923 K. HDT-calc and
Mg–La-calc were thus obtained. These calcined solids were eventu-
ally rehydrated overnight at room temperature in a flow of nitrogen
saturated by water, giving HDT–OH and Mg–La–OH.
The same procedure was applied for Lubrirob 301.01 (LIN)
unless 2 equiv. of formic acid and H2O2 were added (yield 60%, LIN-
ox). NMR and GC analysis are detailed in supplementary materials.
The surface area of the Mg–La mixed oxide is 38 m2 g−1 after
calcination at 923 K (60 m2 g−1 before calcination). The as-prepared
HDT shows a surface area of about 90 m2 g−1, while the surface area
of HDT-calc is 280 m2 g−1, and that of HDT-OH about 50 m2 g−1. The
basic strength of the solid and the number of basic sites were deter-
mined by microcalorimetry at room temperature, using CO2 as
probe molecule. Coupling microcalorimetry and volumetry allows
to determine the number of basic sites adsorbing CO2, and their
strength measured by the adsorption enthalpy, i.e. the energy of
the interaction of the probe with the surface.
2.4. General method for ring opening reaction of
1,2-epoxydodecane with hydrotalcite
In a typical experiment, the reaction was carried out in a 5 ml
closed vial equipped with a magnetic stirrer. A mixture of 1 mmol
(0.2 g) of 1,2-epoxydodecane, 1.2 mmol (120 ␮l) acetic anhydride
and 20 mg catalyst (10 wt%) was introduced into the vial. Dode-
cane was introduced as internal standard. The reaction mixture
was heated to the desired temperature (typically 373 K) under stir-

G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
3
Table 1
ring for 24 h. After cooling, an aliquot was treated with water and
extracted with ether. The organic layer was dried over MgSO4 and
analyzed by GC (column HT5, 60 m).
Ring-opening of epoxydodecane by tertiary amine.
Entry
Catalyst
T (K)
Cat (mol%)
Conv. (%)
Sel (%)
1
2
3
4
5
NEt3
(C12H25)3N
373
373
373
373
353
5
5
20
5
5
93
87
98
97
69
94
92
90
95
99
2.5. General method for ring opening of epoxized oils catalyzed
with Bu4NCl
(C12
H
25)3N
C5H5N
C5H5N
In a typical experiment, 100 g of OLE-ox, 108 ml (1.1 mol) of
Ac2O and 5.5 g of Bu4NCl were added in a 500 ml round bottom flask
and heated for 18 h at 130 ◦C (complete conversion by GC analysis).
The solution was cooled down, extracted with AcOEt. The organic
layer was washed with H2O, Na2CO3 and then NaCl sat. The brown
solution was then filtered on active carbon (Acticarbon L3S) then
dried over MgSO4. After evaporation of the solvent, 228 g of slightly
yellow liquid were obtained (OLE-ac).
Reaction conditions: 1 equiv. 1,2-epoxydodecane, 1.2 equiv. Ac2O, catalyst, 5 h.
3. Results and discussion
3.1. Ring opening of epoxydodecane with base
Initially, the acylation of 1,2-epoxydodecane to the correspond-
ing diester product was used as a model reaction (Scheme 2). We
have previously shown that ammonium quaternary salts were very
efficient for this reaction [17]. Considering the mechanism, we
anticipated that the reaction could be catalyzed by bases so, eval-
uation of the corresponding amine was carried out using the same
reaction conditions: 373 K in the presence of a slight excess of acetic
anhydride (1.2 equiv.) and under solvent free conditions (Table 1).
Tertiary amines were very efficient to catalyze this reaction.
Almost complete conversions were achieved in the presence of 5%
triethylamine, fatty amine or pyridine (entries 1, 2 and 4). What-
ever the tertiary amines, high selectivity towards diacetate (>90%)
was obtained, the only by-product analyzed being the monoacetate
derivative. These results are in the same range to those obtained
with ammonium salts (97% conversion at 5 h [17]). Pyridine is the
most efficient catalyst and the reaction can be performed selec-
tively at 353 K but the conversion decreased (entry 5). However,
due to toxic properties of this aromatic amine, its use should be
avoided.
In parallel, several inorganic bases were evaluated for this reac-
tion and we found that hydrotalcite exhibited significant activity
achieving an almost complete conversion (95%) after 24 h. The reac-
tion was selective towards the diacetate, and only small amount of
monoacetate was formed. Considering this interesting result and
the easy availability of this catalyst family, we optimized the reac-
tion conditions with a range of solid oxides with different basic
strengths (Table 2).
The results of the determination of basic strengths of hydro-
talcites and Mg–La mixed oxide by microcalorimetry are reported
in Fig. 1. As-synthesized HDT, evacuated at 373 K, thus still in the
carbonate form, appears as a weak base, adsorbing CO2 with an
enthalpy of about 80 kJ mol−1. The number of active sites is about
150–200 ␮mol/g. Mg–La mixed oxide decarbonated at 923 K is a
strong base but with very low number of basic sites, adsorbing
about 10 ␮mol/g of CO2 with an enthalpy >140 kJ mol−1. Calcined,
then hydrated hydrotalcite (HDT–OH) adsorbs about 250 ␮mol/g
with a constant heat of adsorption of 100 kJ mol−1. Rehydration of
the solid shifts then the type of basicity from Lewis to Brönsted
with only marginal changes of basicity [27]. By contrast, Mg–La
mixed oxide appears to be a stronger base, and this series of cata-
lysts permits to investigate the effects of basicity on the catalytic
properties.
2.6. Analysis of fatty ester derivatives
The composition of the mixture was determined by GC analysis
using a DB-23 column (60 m) with methyl myristate as external
standard (Scheme 1).
The following data are obtained (supplementary materials):
LUBRIROB 926.65: methyl palmitate (6%), methyl stearate (2%)
methyl oleate (59%), methyl linoleate (21%), methyl linolenate
(9%).
LUBRIROB 301.01: methyl palmitate (6%), methyl stearate (3%)
methyl oleate (32%), methyl linoleate (57%), methyl linolenate
(2%).
2.7. Rheological properties
The plastisol formulation was 50 parts (by weight) of plasticizer
per hundred of PVC. Viscoelastic properties were measured with
a dynamic stress rheometer of TA Instruments (AR 2000). Rheol-
ogy measurements were carried out in the oscillatory mode with
a parallel-plate geometry (diameter = 25 mm). The thickness of the
sample (gap of the parallel plate) was about 1 mm. Furthermore,
◦
∼
a programmed heating rate ( 5 C/min) at a constant frequency
=
(ω = 6.28 rad/s) was carried out from room temperature (T) 150 ◦C
to follow the complete of gelation of various PVC plastisols. Finally,
the complex shear modulus (G*(ω) = Gꢀ(ω) + jGꢀꢀ(ω)) was recorded
versus temperature.
2.8. Procedure for recycling the catalyst
The first run was performed from 800 mg 1,2-epoxydodecane,
Ac2O (1.2 equiv.) and 80 mg catalyst (10%). At the end of the reac-
tion (24 h), the mixture was cooled to room temperature, the solid
was filtered. Several treatments were evaluated. Treatment 1: the
catalyst was reused without treatment. A mixture of 800 mg of 1,2-
epoxydodecane, 1.2 g acetic anhydride were introduced into the
vial containing the used catalyst. The reaction mixture was heated
at 373 K; treatment 2: the catalyst was washed with AcOEt, then
ether and air dried at room temperature; treatment 3: the catalyst
was washed with ether, air dried at room temperature and calcined
(723 K); Treatment 4: the catalyst was washed with ether (10 ml),
then treated with saturated NaHCO3 (30 ml). After filtration, the
solid was air dried.
The following parameters were studied: temperature of the
reaction, amount of catalyst, nature and activation of mixed oxides
(Table 2). The catalytic properties of the hydrotalcites (Mg/Al
OAc
Ac2O, Δ
O
OAc
Catalyst
Scheme 2. Catalyzed-ring opening of epoxides.

4
G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
Table 2
Influence of mixed oxide pre-treatment on the conversion of epoxide.
Entry
Catalyst
SBET (m2 g−1
)
Basic sites (␮mol/g)a
T (K)
Cat. (wt%)
t (h)
5/24
24/72
24
5/24
5/24
5/24
5/24
5/24
Conv. (%)
Sel24 h (%)
1
2
3
4
5
6
7
8
HDT–CO3
90
110
373
373
353
373
373
373
373
373
10
1
67/90
25/80
79
28/77
63/87
10/28
6/20
87
73
71
83
97
70
71
71
10
10
15
10
10
10
HDT-calc
HDT–OH
Mg–La–CO3
Mg–La-calc
Mg–La–OH
280
50
60
38
–
400
420
25
–
18/70
Reaction conditions: 1 equiv. 1,2-epoxydodecane, 1.2 equiv. Ac2O, catalyst; HDT-CO3 used as received; HDT-calc: after thermal treatment at 723 K (see Section 2); HDT-OH:
after rehydration (see Section 2).
a
Refers to number of basic sites with adsorption enthalpy higher than 40 kJ mol−1).
ratio = 3) are compared to those of stronger basic Mg–La mixed
oxides (ratio Mg/La = 3). All these solid catalysts are subjected to
different activation treatments. The original mixed oxides are pre-
pared in the carbonate form (–CO3). Decarbonatation occurs after
thermal treatment at 723 K for 1 h (-calc), finally the latter ones are
treated in a stream of nitrogen saturated with water to give the
hydrated form (–OH) [20,27].
The preliminary optimization of reaction conditions has
been performed with the as-synthesised hydrotalcite Mg/Al = 3
(HDT–CO3). Full conversion was achieved at 373 K after 24 h reac-
tion time in the presence of 10 wt% solid (Table 2, entry 1). The
reaction can be performed at 353 K without affecting strongly the
reaction rate (entry 3). As expected, decreasing the amount of cata-
lyst lead to a lower reaction rate, the conversion reaching only 80%
after 72 h (entry 2). In order to compare the activity of the different
solids, the reaction was carried out at 373 K in the presence of 10%
weight of basic solids (Table 2, entries 1, 4 and 5). After thermal acti-
vation (HDT-calc) and rehydration (HDT–OH), the solids exhibited
different behaviours the former one being much less active.
The kinetics of the catalyzed ring-opening of epoxide have been
determined on HDTs subjected to the different treatments. In all
cases, the diacetate is the main product of the reaction (>90%), the
by-product being the hydroxyl monoacetate [17]. The evolution of
the conversion with time is reported in Fig. 2.
It is clear that HDT–CO3 and HDT–OH presented similar per-
formances, while, HDT-calc is much less active. After 5 h reaction
time, 67, 63 and 28% yield of diacetate are achieved with HDT–CO3,
HDT–OH and HDT-calc respectively (Table 2, entries 1, 5 and 4).
One important point which has to be taken into account in base-
catalyzed reaction is the acidity of the reactant. Indeed, when the
reactant is a ketone (pK ca. 19–20), strong basic sites are nec-
essary to deprotonate the substrate, while moderate strength is
enough to activate an acid anhydride. The calcined hydrotalcite
is the solid with the highest surface area (280 m2 g−1) and it also
possesses a large number of basic sites (ca. 400 ␮mol/g). Nev-
ertheless, it exhibits the lowest performance for the considered
reaction. Indeed, after the calcination of synthesized hydrotalcite,
the basic sites are Lewis sites which are not the most suitable for the
ring opening of epoxide with acetic anhydride [27]. On the other
hand, as-synthesized hydrotalcite (HDT–CO3) or rehydrated one
(HDT–OH) are solids with moderate surface area (50–90 m2 g−1
)
and 100–420 ␮mol/g Brönsted basic sites (Table 2). In the presence
of traces of water, the carbonates can be hydrolyzed to hydrogeno-
carbonates which are Brönsted bases [28]. We suggest that the
present reaction is mainly catalyzed by Brönsted type basic sites.
The role of the basic sites was confirmed by the use of Mg–La
mixed oxides. These solids are strong bases but with a small number
of sites compared to hydrotalcites (25 ␮mol/g, entry 7, Table 2) [23].
After 5 h, the calcined form which is considered as the most basic
one, yielded very low conversion (entry 7). This mixed oxide has
however a very low number of sites compared to HDT, as illustrated
in Fig. 1, and the higher strength does not compensate this small
amount of basic sites. After rehydration, Brönsted basic sites are
formed and the conversion increased from 20% with the calcined
sample to 70% with the hydrated one (entries 7 and 8), in agreement
with the proposal of a higher activity of Brönsted sites.
The comparison after 3 h on 10% and 1% HDT–CO3 shows that the
conversion is roughly proportional to the catalyst amount, which
means the process operates in the chemical regime. Several exper-
iments performed with 20 mg of catalyst exhibit reproducibility
better than 90%, thus showing that hydrotalcite has a homogeneous
composition.
160
HDT calcined 773K
HDT hydrated RT
140
HDT calcined 773K, hydrated RT
120
MgLa calcined 773K
100
80
60
40
20
0
0
50 100 150 200 250 300 350 400 450 500
Amount of CO2 adsorbed (µmol.g-1)
Fig. 1. Differential enthalpies of adsorption of CO2 on (a) HDT–CO3 evacuated at
373 K, (b) HDT calcined at 773 K, then evacuated at 773 K, (c) HDT calcined at 773 K,
hydrated at RT, evacuated at 373 K, and (d) Mg–La mixed oxide calcined at 923 K,
stored in air, evacuated at 773 K.
Fig. 2. Influence of the pretreatment of hydrotalcite on the kinetic of the ring
opening. Reaction conditions: 1 equiv. 1,2-epoxydodecane, 1.2 equiv. Ac2O, catalyst,
373 K.

G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
5
(a)
(b)
R
AcO
HDT -OAc
O
R
O
OAc
Al
AcO
O
O
O
O
R
AcO
Ac2O
Mg
Mg
HDT -O
Scheme 3. Schematic representation of the role of hydrotalcite in ring epoxide opening.
Considering the reaction mechanism (Scheme 3a), it can also
be assumed that strong bases such as Mg–La mixed oxides inter-
act too strongly with acetic anhydride or some intermediate acetic
derivatives, that could act as poison. High basicity is not necessary
for the ring-opening of epoxides therefore it is more convenient to
use solids with a larger number of Brönsted basic sites as present
in rehydrated hydrotalcite. The interest of hydrotalcite lies in their
role as acid–base bifunctional catalyst [29]. In the present work,
the mechanism could be a concerted reaction with nucleophilic
attack of the acetate on the epoxide activated by Mg2+ cations
(Scheme 3b). Similar concerted mechanism was reported by the
group of Kaneda in the condensation of epoxide with CO2 [30].
The comparison with usual liquid bases can be done under
similar reaction conditions, after 5 h of reaction: in the presence
of 5 mol% NEt3, the conversion reached 93% (entry 1, Table 1),
corresponding to a TOF ≈ 4 h−1, while in the presence of 10 wt%
HDT–CO3, the conversion reached 67% (entry 1, Table 2) that is
TOF ≈ 60 h−1 indicating that these solids are very efficient for that
reaction. The best performance achieved with solid catalysts could
be attributed to the concerted mechanism described in Scheme 3.
The possibility of reuse of the catalysts for the ring opening
of 1,2-epoxydodecane was then studied starting from the as-
synthesized HDT–CO3 solid (Fig. 3). At the end of the reaction, the
mixture was cooled down and the catalyst was separated by filtra-
tion and reused in a new run. A significant decrease in conversion
was observed under those conditions and the conversion reached
only 25% in the second run (treatment 1). Attempts to wash the
solid with ether to remove potential organics absorbed on the sur-
face failed as well (treatment 2). As expected, the calcination at
723 K did not allow reusing the solid (treatment 3, < 20%). Finally,
after the first run, the solid was filtered, washed with ether, treated
with aqueous solution of NaHCO3 in order to obtain the initial
hydrotalcite (treatment 4). This treatment was successful as sim-
ilar conversion and selectivity were achieved for the second run
(>90%). It is assumed that this work up reactivates the hydrotalcite
to the active carbonate form. Following this un-optimized treat-
ment, this catalyst was found to be reusable without appreciable
loss in activity for at least 4 runs.
3.2. Ring opening of substituted epoxides
The performances of the catalysts for the ring opening of 2,3-
epoxynorbornene 1, limonene epoxide 2, ␣-pinene oxide 3 and
methyl oleate epoxide 4 were tested (Fig. 4). For comparison,
NBu4Cl reported as efficient catalyst for this reaction was also eval-
uated (Table 3).
Regardless of the treatment of hydrotalcite, in the presence of
the cyclic epoxides 1–3, a mixture of products was obtained, show-
ing that the catalyst is not stereoselective. As a general feature, we
observed that HDT-calc exhibited the lowest activity while simi-
lar conversions were observed with HDT–CO3 or HDT–OH solids
(entries 2–4). An increase of temperature allowed an improvement
of conversion. For the less reactive ␣-pinene oxide 3, almost com-
plete conversion was achieved after 24 h at 403 K in the presence
of HDT–OH solid (entries 6 and 7).
Due to their readily availability, HDT could be applied to the ring
opening of methyl oleate epoxide 4 providing access to the cor-
responding diacetate that could find application as bioplasticizers
[31,32]. The most active catalyst (HDT–OH) was evaluated for this
reaction. For this disubstituted cis-epoxide, only 60% conversion
was achieved at 373 K for 24 h in the presence of 10% catalyst (entry
8). Increasing the temperature allowed almost complete conver-
sion using either HDT–OH or HDT–CO3 (entries 9 and 10). The
reaction was very selective only small amount (<5%) of monoac-
etate being detected as a by-product. As observed for the previous
substrates, the performance of HDT-calc was lower (entry 11).
Table 3
Acylation of epoxide catalyzed with HDT-OH.
Entry
Substrate
Catalyst
T (K)
Conv (%)
1
2
3
4
5
6
7
8
1
2
HDT-OH
HDT-CO3
HDT-calc
HDT-OH
Bu4NCl
HDT-OH
HDT-OH
HDT-OH
HDT-OH
HDT-CO3
HDT-calc
Bu4NCl
373
373
373
373
373
373
403
373
403
403
403
373
423
85
100
36
98
88
72
100
61
94
99
3
4
9
10
11
12
13
Fig. 3. Reuse of the catalysts. Reaction conditions: 1 equiv. 1,2-epoxydodecane,
1.2 equiv. Ac2O, 10% HDT, 373 K, 24 h. Treatment 1: the catalyst was reused without
treatment; treatment 2: the catalyst was washed with ether and air dried at room
temperature; treatment 3: the catalyst was washed with ether, air dried at room
temperature and calcined (723 K); treatment 4: the catalyst was washed with ether,
then NaHCO3.
71
83
97
Bu4NCl
Reaction conditions: 1 equiv. epoxide, 1.2 equiv. Ac2O, catalyst 10 wt.% HDT or
5 mol% Bu4NCl, 24 h.

6
G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
C8H17
O
O
O
O
O
O
1
2
3
4
Fig. 4. Structure of the epoxides evaluated in this work.
Table 5
We have previously shown that NBu4Cl was also efficient cat-
Gelation temperature and maximal storage modulus for selected plastisol.
alyst for the ring opening of 1,2-epoxydodecane [17]. Indeed, we
evaluated this ammonium salt for the ring opening of di or trisub-
stituted epoxides. Some results are reported in Table 3 indicating
that high conversions were achieved in the presence of 5 mol% of
homogeneous catalyst (entries 5 and 13) but at 403 K.
Then, we carried out the ring opening of epoxides synthetized
from vegetable oil. Hydrotalcite catalyst was evaluated for the ring
opening of epoxidized fatty ester and compared with homogeneous
ammonium salts. We investigated different sources of vegetable oil
having from 1 to 3 double bonds in the fatty chain (Table 4). In the
following, two types of fatty ester are considered: OLE containing
methyl oleate as the main constituent (58%) and LIN constituted
of mainly methyl linoleate (57%) (Scheme 1). The epoxidation of
fatty ester was carried out using Prileschajew reaction conditions
(HCOOH, cat. H2O2) [24]. After 10 h at 343 K, complete conversion
was achieved.
The reaction conditions necessary to get complete conversion
using different anhydride and catalyst are reported in Table 4.
Hydrotalcites were active for the ring opening of OLE-ox (Table 4,
entry 1). However, to reach complete conversion, an excess of Ac2O
was used and the reaction was carried out for 24 h at 403 K. Using
NBu4Cl as catalyst, complete conversion was achieved after 6 h at
403 K with 1.2 equiv. of anhydride. This latter catalytic system was
used to synthesize large scale of substituted oil (Scheme 1) for the
evaluation of their physicochemical properties. In order to estab-
lish the influence of the chain length not only acetic anhydride was
used as reactant but also heavier anhydride from propionic up to
valeric (Table 4). With these anhydrides the reactions were carried
out overnight for complete conversion.
ꢀ
Entry
Plasticizers
T (◦C)gelation
Gmꢀ ax (×105 Pa)
TG
(◦C)
max
1
2
3
4
5
6
DINP
OLE
LIN
OLE-ox
OLE-ac
LIN-ac
80
80
78
52
72
64
3.6
1.3
1.3
8.3
4.3
4.7
125
150
150
100
130
120
All the samples were miscible with PVC indicating a good compat-
ibility of the plasticizers. In Fig. 5a and b, storage modulus Gꢀ and
loss modulus Gꢀꢀ of plastisol prepared from oleate derivatives (OLE)
were recorded as a function of the temperature. For comparison, a
reference plastisol was prepared from phthalate derivative (DIPN).
At room temperature, the storage modulus and loss modulus
values of commercial fatty ester (OLE) are one order of magnitude
lower than the phthalate indicating low plastic properties probably
due to modest interactions between fatty ester and PVC. The prop-
erties of the epoxide derivative (OLE-ox) largely used as stabilizer
in PVC were also determined. With such compound, the gelation
temperature (52 ◦C) corresponding to the transformation from the
liquid paste to solid is too low (Table 5, entry 4).
After acylation (OLE-ac) the storage modulus and loss modulus
profiles are very similar to those recorded for phthalate plasticizer
with a Gmꢀ ax of same order of magnitude (4 × 105 Pa) at 125–130 ◦C.
The gelation temperatures are 80 ◦C and 72 ◦C for DIPN and OLE-ac
respectively (entries 1 and 5).
The properties recorded with LIN series (supplementary
material) are similar to those reported with OLE series both for the
non-substituted substrate (entries 2 and 3) and for peracetylated
one (entries 5 and 6). Moreover, the plastisol was stable and after
one month storage, similar profiles were recorded (Fig. 5a and b).
The ring-opening of OLE-ox was carried out with several anhy-
drides and the properties of the resulting products were evaluated
(Fig. 6). The higher the chain length is, the higher the viscosity of
the product is. Moreover, the maximum storage modulus is lower
for butyric or valeric derivatives and has been reached for a higher
temperature corresponding to unfavourable gelation properties.
To summarize, the viscoelastic properties of plastisol prepared
with LIN-ac and OLE-ac exhibited similar properties to fossil-based
phthalate derivatives and are good candidates to replace the lat-
ter one. Further works are under progress to determine extended
physicochemical properties.
3.3. Physicochemical properties of fatty ester derivatives
To evaluate the physicochemical properties of the potential
plasticizers synthetized from vegetable oils, we prepared plastisol
which is a liquid suspension of PVC particles finely dispersed in the
liquid plasticizer matrix (1 part of plasticizer and 2 parts of PVC)
at room temperature. When the PVC plastisol is heated, the plas-
ticizer diffuses into the resin particles up to give a homogeneous
mass corresponding to the gelation [33,34]. The most commonly
used methods for characterizing gelation behaviour are the deter-
mination of the gelation temperature where a significant change is
brought about by heating. The storage modulus Gꢀ and loss modu-
lus Gꢀꢀ were recorded as the function of the temperature [35–37].
Table 4
Reaction conditions used to get complete conversion of epoxide.
Epoxide
OLE-ox
LIN-ox
Composition
1.2 C C/chain
1.5 C C/chain
(RCO)2O (n equiv.)
t (h)
Catalyst
Ac2O (2)
Ac2O (1.2)
(C2H5CO)2O (1.2)
(C3H7CO)2O (1.2)
(C4H9CO)2O (1.2)
24
6
18
18
18
10 w% HDT–CO3
5 mol% nBu4NCl
5 mol% nBu4NCl
5 mol% nBu4NCl
5 mol% nBu4NCl
Ac2O (2)
6
5 mol% nBu4NCl

G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
7
(a)
(b)
LOSS MODULUS
STORAGE MODULUS
1.E+06
1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
1.E-01
1.E+06
1.E+05
1.E+04
1.E+03
1.E+02
1.E+01
1.E+00
1.E-01
DIPN
OLE
OLE-ox
OLE-ac
OLE-ac 1 month
20
40
60
80
100
120
140
160
20
40
60
80
100
120
140
160
T (°C)
T (°C)
Fig. 5. (a) Storage modulus (Gꢀ) profile of plastisol as a function of temperature and (b) Loss modulus (Gꢀꢀ) profile of plastisol as a function of temperature.
Acknowledgements
We thank Agrice-ADEME (grant no. 2005-11) and Novance for
gift of Lubrirob.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2010.11.020.
References
[1] L.I. Nass, C.A. Heiberger, Encyclopedia of PVC, vol. 2, second edition, Marcel
Dekker Inc., New York, 1987, p. 202.
[2] L.G. Krauskopf, A. Godwin, in: C.E. Wilkes, J.W. Summers, C.A. Daniels (Eds.),
PVC Handbook, Hanser Gardner, 2005, p. 173.
[3] S.E. Djilai, T. Bouchami, F. Krid, N. Boudiaf, D. Messadi, Eur. Polym. J. 26 (2000)
1981.
[4] A. Marcilla, S. Garcia, J.C. Garcia-Queseda, Polym. Test. 27 (2008) 221.
[5] http://www.biohub.fr/.
[6] B. Nielsen, F.V. Sparso, J.K. Kristiansen, WO 01/14466 A1 (2001).
[7] R. Hinault, H. Lê Chien, J. Barbier, WO 06/075071 (2006).
[8] P. Bondioli, in: D. Gunstone, R.J. Hamilton (Eds.), Oleochemical Manufacture
and Applications, Sheffield Academic Press, 2001, p. 74.
[9] H. Schuster, L.A. Rios, P.P. Weckes, W.F. Hoelderich, Appl. Catal. A 348 (2008)
266.
[10] P.S. Lathi, B. Mattiasson, Appl. Catal. B 69 (2007) 207.
[11] J. Hancsok, M. Bubalik, M. Törö, J. Baladincz, Eur. J. Lipid Sci. Technol. 108 (2006)
644.
[12] H. Wagner, R. Luther, T. Mang, Appl. Catal. A 221 (2001) 429.
[13] J.A. Melero, J. Iglesias, G. Morales, Green Chem. 11 (2009) 1285.
[14] M. Di Serio, R. Tesser, L. Pengmei, E. Santacesaria, Energy Fuels 22 (2008) 207.
[15] F.R. Ma, M.A. Hanna, Bioresour. Technol. 70 (1999) 1.
[16] M. Guidotti, R. Psaro, M. Sgobba, N. Ravasio, in: G. Centi, R.A. van Santen (Eds.),
Catalysis for Renewables, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim,
2007, p. 257.
Fig. 6. Infuence of lateral chain length on storage modulus (Gꢀ) profile of plastisol
as a function of temperature.
[17] G. Fogassy, C. Pinel, G. Gelbard, Catal. Commun. 10 (2009) 557.
[18] B. Veldurthy, J.M. Clacens, F. Figueras, Adv. Synth. Catal. 347 (2005) 767.
[19] B. Veldurthy, F. Figueras, Chem. Commun. (2004) 734.
[20] D.P. Debecker, E.M. Gaigneaux, G. Busca, Chem. Eur. J. 15 (2009) 3920.
[21] B.F. Sels, D.E. DeVos, P.A. Jacobs, Catal. Rev. 43 (2001) 443.
[22] A. Vaccari, Appl. Clay Sci. 14 (1999) 161.
[23] J. Palomeque, J.M. Clacens, F. Figueras, J. Catal. 211 (2002) 103.
[24] B. Rangarajan, A. Havey, E.A. Grulke, P.D. Culnan, J. Am. Oil Chem. Soc. 72 (1995)
1161.
4. Conclusion
In conclusion, we have found that solid bases such as cheap and
easily available hydrotalcites are suitable catalysts for ring open-
ing of epoxides with acetic anhydride. Moreover, these solid bases
are easy to separate, non-toxic and show a good stability versus
temperature.
[25] A. Campanella, M.A. Baltanas, J. Am. Oil Chem. Soc. 76 (2002) 373.

8
G. Fogassy et al. / Applied Catalysis A: General 393 (2011) 1–8
[26] M.T. Benaniba, N. Belhaneche-Bensemra, G. Gelbard, Energy Ed. Sci. Technol.
21 (2008) 71.
[27] F. Figueras, in: A. Gil, S.A. Korii, R. Trujilane, M.A. Vicente (Eds.), Pillared Clays
and Related Catalysts, Springer, 2010, p. 422.
[28] D. Tichit, M.H. Lhouty, A. Guida, B.H. Chiche, F. Figueras, A. Auroux, D. Bartalini,
E. Garrone, J. Catal. 151 (1995) 50.
[31] L.G. Krauskopf, J. Vinyl Additive Tech. 9 (2003) 159.
[32] O. Fenollar, D. Garcia, L. Sanchez, J. Lopez, R. Balart, Eur. Polym. J. 45 (2007)
2674.
[33] S.Y. Kwak, J. Appl. Polym. Sci. 39 (1995) 3507.
[34] H. Boudhani, C. Laine, R. Fulchiron, P. Cassagnau, Rheol. Acta 46 (2007)
825.
[29] R.S. Varma, Tetrahedron 58 (2002) 1235.
[30] K. Yamaguchi, K. Ebitani, T. Yoshida, H. Yoshida, K. Kaneda, J. Am. Chem. Soc.
121 (1999) 4526.
[35] N. Nakajima, D.W. Ward, J. Appl. Polym. Sci. 28 (1983) 807.
[36] A. Marcilla, J.C. Garcia, M. Beltran, Eur. Polym. J. 33 (1997) 753.
[37] H. Boudhani, R. Fulchiron, P. Cassagnau, Rheol. Acta 48 (2009) 135.