Influence of alkenyl structures on the epoxidation of unsaturated fatty acid methyl esters and vegetable oils

Article abstract of DOI:10.1039/c5ra11035a

Epoxidation of vegetable oils or fatty acid methyl esters (FAMEs) produce important monomers which are widely used as plasticizers or stabilizers in the polymer industry. However, little attention has been focused on the influence of the alkenyl structure of the fatty acid on the efficiency and selectivity of their epoxidation. In this work, the influence of the alkenyl structure (the number of double bonds) of the FAMEs on the epoxidation reaction has been investigated. Three model FAMEs with 1 to 3 double bonds were epoxidized using both a weak (formic acid) and a strong (sulfuric acid/acetic acid) acid system. It was found that FAMEs with more double bonds have higher reactivities toward the epoxidation reaction. In addition, the electron-donating effect of the double bonds on the fatty acid chain tends to stabilize the resulting epoxide adjacent to it with the weak acid system. Furthermore, FAMEs with more double bonds easily undergo side reactions with the strong acid system (H₂SO₄). Epoxidation of two vegetable oils with different fatty acid compositions were carried out with the same two acid catalyst systems. And the results were in agreement with those from the FAMEs. The current findings could provide useful guidance for the epoxidation of different vegetable oils with different alkenyl structure compositions.

Full text of DOI:10.1039/c5ra11035a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Influence of alkenyl structures on the epoxidation
of unsaturated fatty acid methyl esters and
vegetable oils†
Cite this: RSC Adv., 2015, 5, 74783
Yao-Bing Huang,‡ Meng-Yue Yao,‡ Ping-Ping Xin, Meng-Chao Zhou, Tao Yang
and Hui Pan*
Epoxidation of vegetable oils or fatty acid methyl esters (FAMEs) produce important monomers which are
widely used as plasticizers or stabilizers in the polymer industry. However, little attention has been focused
on the influence of the alkenyl structure of the fatty acid on the efficiency and selectivity of their
epoxidation. In this work, the influence of the alkenyl structure (the number of double bonds) of the
FAMEs on the epoxidation reaction has been investigated. Three model FAMEs with 1 to 3 double bonds
were epoxidized using both a weak (formic acid) and a strong (sulfuric acid/acetic acid) acid system. It
was found that FAMEs with more double bonds have higher reactivities toward the epoxidation reaction.
In addition, the electron-donating effect of the double bonds on the fatty acid chain tends to stabilize
the resulting epoxide adjacent to it with the weak acid system. Furthermore, FAMEs with more double
Received 10th June 2015
Accepted 27th August 2015
2 4
bonds easily undergo side reactions with the strong acid system (H SO ). Epoxidation of two vegetable
oils with different fatty acid compositions were carried out with the same two acid catalyst systems. And
the results were in agreement with those from the FAMEs. The current findings could provide useful
guidance for the epoxidation of different vegetable oils with different alkenyl structure compositions.
DOI: 10.1039/c5ra11035a
www.rsc.org/advances
chain of fatty acids to produce new value-added chemicals or
monomers for polymers has also attracted considerable interest
1
Introduction
The reliance on fossil feedstocks for the production of fuels and for many years. One of the most common approaches is the
chemicals has led to an energy shortage all over the world, as a epoxidation of vegetable oils or their fatty acid esters (FAMEs)
consequence of diminishing fossil resources and high energy which are useful intermediates or compounds that can serve as
6
7
8
demand. Moreover, CO2 emission from fossil fuels has also plasticisers, resins, stabilizers in coatings and biofuel
raised global warming problems and environmental pollution. additives.
9
Renewable biomass has been considered as a potential candi-
Epoxidation of vegetable oils is usually carried out with
date to solve this problem due to its large annual production peroxyacid formed in situ from organic carboxylic acid (e.g.
1
and rich carbon nature. Among different biomass resources, acetic acid) and hydrogen peroxide over an acidic catalyst (e.g.
vegetable oils are complex mixtures of triglycerides, which mineral acids) which is known as the Prileshaiev process
3
10
11
consist of a glycerol backbone and three fatty acid moieties (Fig. 1). For example, Goud and Okieimen et al. reported the
(
2
saturated or unsaturated). They are renewable, inherently epoxidation of Karanja oil and rubber seed oil with hydrogen
biodegradable, eco-friendly and inexpensive to serve as the peroxide and peroxyacetic acid over sulfuric acid catalysts. As
alternative sources to fossil feedstock for producing chemicals well as sulfuric acid (H
2
SO
4
), hydrochloric acid (HCl), nitric acid
3
or fuels.
(HNO
3
) and phosphoric acid (H
3
PO ) were also used as catalysts
4
12
Chemical modications of vegetable oils are important for the epoxidation of vegetable oils. The main drawbacks of
4
methods to obtain industrial products. Transesterication of the Prileshaiev process is that the resulted epoxides can easily
vegetable oils provides the rst generation of biodiesel fuels undergo ring-opening reactions due to the instability of the
5
13
which have already been industrialized for years. Besides, epoxide group under acidic conditions. Hence, tremendous
modication of unsaturated C]C bonds in the hydrocarbon research efforts have been devoted to the development of new
technologies to improve the efficiency and suppress the ring-
opening reactions, typically involving the exploration of new
College of Chemical Engineering, Nanjing Forestry University, 159 Longpan Road,
catalysts and optimization of the reaction conditions. Hetero-
2
†
1
‡
10037, Nanjing, China. E-mail: hpan@njfu.edu.cn
Electronic supplementary information (ESI) available. See DOI:
0.1039/c5ra11035a
14
15
geneous solid acid catalysts such as acidic resin, Ti–silicas,
16
17
18
Nb(V)–silica, sulfated-SnO2, polyoxometalate were devel-
oped for the reaction. These advances have further broadened
These authors equally contributed to the work.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74783–74789 | 74783

View Article Online
RSC Advances
Paper
the choice of catalytic systems for the epoxidation of vegetable purchased from local chemical reagent company and used
oils. without further purication.
Oleic (C18:1), linoleic (C18:2) and linolenic (C18:3) acids are
three typical unsaturated fatty acids in vegetable oils that
contain 1 to 3 double bonds. During our study on the epoxi- 2.2 Epoxidation of FAMEs and vegetable oils
dation of vegetable oils, we noted that vegetable oils containing
A typical procedure for the epoxidation of FAMEs was as follows:
mmol methyl oleate, 6 mmol HCOOH (or 0.41 mmol H SO , 6
more linolenic fatty acid composition tend to give inferior
epoxidation results (low epoxides yield and complex side
products). Vegetable oils from various plants contain different
fatty acid compositions including these three fatty acids. Thus,
it's important to nd out general principles of the unsaturated
fatty acid epoxidation regarding the relationship between the
unsaturated alkenyl structure and their epoxidation efficiency.
This would help to design more efficient catalytic systems for
the epoxidation of vegetable oils with different fatty acids
composition. However, to the best of our knowledge, there is
little in the literature related to the inuences of the alkenyl
skeleton with different numbers of C]C bonds on the effi-
ciency and product selectivity of the epoxidation of vegetable
oils.
4
2
4
mmol CH COOH) were rstly added to a 100 mL three necked
3
round-bottom ask equipped with a magnetic stirrer. The ask
was then purged with nitrogen to exclude air and placed in a
temperature controlled oil bath. The mixture was then stirred
ꢀ
and heated to 30 C. Then, 21 mmol H
2
O
2
(30 wt%) was added,
dropwise, to the reaction at a slow rate to maintain the
ꢀ
temperature of the reaction system within 30 ꢁ 1 C. The whole
reaction time was 5 h. Upon the end of the reaction, the reaction
mixture was extracted twice with ethyl acetate (2 ꢂ 20 mL). The
organic phase was washed with saturated sodium bicarbonate
(3 ꢂ 15 mL) and saturated sodium chloride (3 ꢂ 15 mL) for
three times, respectively. The resulting organic phase was then
dried with anhydrous magnesium sulfate. Aer ltration, the
solvent ethyl acetate was removed by vacuum evaporator.
The epoxidation of vegetable oils were followed the same
procedures as the epoxidation of FAMEs except that the amount
of reagents used were as follows: 3 g olive oil or linseed oil
reacted with 0.6 g (13 mmol) FA (or 0.06 g (6.1 mmol) SA, 0.79 g
Herein, we reported the epoxidation of three unsaturated
fatty acid methyl esters (methyl oleate, methyl linoleate, methyl
linolenate) derived from vegetable oils to investigate the inu-
ences of the alkenyl structure on the epoxidation reaction. From
the practical point of view, the two most commonly used liquid
acids catalysts of different acid strength, sulfuric acid (SA) and
formic acid (FA), were selected as the catalysts in this study
aiming to provide some insightful information on the common
methods of the epoxidation of vegetable oils. Different analysis
(
13.1 mmol) AA) and 5.19 g (ꢃ46 mmol) H
2 2
O (30 wt%).
Silica coated aluminium thin layer chromatography (TLC)
plates were used for separating the reaction mixtures. The
eluent was a mixture of petroleum ether and ethyl acetate in
1
methods, iodine value, oxirane content, FTIR, H-NMR and ESI-
25 : 1 (vol/vol) ratio for epoxidized methyl oleate and in 10 : 1
MS, were used to characterize the epoxidation products. Finally,
two vegetable oils with different alkenyl fatty acids composi-
tions were epoxidized with the aforementioned acid catalyst
systems to further verify the results from the epoxidation of the
FAMEs.
(vol/vol) ratio for epoxidized methyl linoleate. The chromogenic
reagent was potassium permanganate solution.
Iodine values (IV ) of the FAMEs and vegetable oils were
O
expressed as gram per 100 gram sample (g per 100 g) and were
19
determined according to Wij's method. The amount of oxirane
oxygen (OO) was measured by the method with hydrobromic
20
acid in acetic acid solution.
The theoretical percentage of oxirane oxygen (OOthe) in 100 g
sample was determined by the following expression.
2
Experimental
21
8
9
>
;
2
.1 Materials
>
>
>
<
>
=
Methyl oleate (MO, Sigma-Aldrich, 99%), methyl linoleate (ML,
TCI, >95%) and methyl linolenate (MLN, Sigma-Aldrich, $99%)
are the model compounds used for this experiment. Glacial
acetic acids (AR), formic acids (AR), hydrogen peroxide (30 wt%)
and sulfuric acid (98%) were purchased from Nanjing Chemical
Reagent Co., Ltd. Wij's solution was bought from Tianjin weiyi
Chemical Technology Co., Ltd. Olive oil and linseed oil were
purchased from local market. All other reagents used were
ðIV
O
=2A
I
Þ
OOthe
¼
ꢀ
ꢁ
ꢂ
ꢃ
ꢂ A ꢂ 100
O
>
IV
O
>
>
:
100 þ
A
O
2A
I
where A
I
(126.9) and A
O
(16.0) are the atomic weights of iodine
and oxygen, respectively and IV is the initial iodine value of the
samples before epoxidation.
The conversion of substrate (C) of the epoxidation substrate
was determined as follows:
O
22
IV
O
ꢄ IVae
C ¼
ꢂ 100%
IV
O
where IV
O
is the initial iodine values of the samples before
epoxidation, IVae is the iodine values of substrate aer
epoxidation.
The yield of epoxidized products (Y) was calculated as
22
Fig. 1 The main reaction route of epoxidation.
follows. (Y):
74784 | RSC Adv., 2015, 5, 74783–74789
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
OOexp
OOthe
66.21%, 74.84% and 75.35%, respectively. Both the OOexp values
and the yield of the epoxidized FAMEs increased from 3.50 to
Y ¼
ꢂ 100%
8
.57 and from 50.51% to 69.45% as the number of double bonds
where OOexp was the actual percentage of the oxirane oxygen
obtained in the experiment.
The amount of the ring-opening products and all other
unidentied products from side reactions was determined as
the difference between the conversion of the epoxidation reac-
tion and the yield of the epoxidized products (CꢄY).
increased from 1 to 3 in the starting materials, respectively. It is
well known that the epoxidation reaction is always associated
with side reactions, such as hydroxylation, oxidation, oxygena-
tion, and dimer formation. It is worth noting that the yield of
the side products (CꢄY) decreased substantially as the number
of the double bonds increased. It has been reported that during
the epoxidation of alkenes, the electron-donating groups adja-
cent to the alkenyl groups could enhance the activity of the
13
2.3 Identication of reaction products
The resulting reaction mixtures were analyzed by gas alkenyl group toward the epoxidation, and moreover, stabilize
23
chromatography-mass spectroscopy (GC-MS) and GC. The GC- the resultant epoxides from decomposition. Similarly, it can
MS (Agilent 7890A, Agilent 5975C MSD) and GC (Agilent be speculated that FAMEs with more double bonds would have
7890A) instruments were equipped with a DB-5 capillary higher reactivities due to the electron-donating effects from the
column (30 m ꢂ 0.25 mm ꢂ 0.25 mm, Agilent). The temperature adjacent carbon and double bond. In addition, when one of the
ꢀ
ꢀ
ꢀ
ꢀ
ꢄ1
of column increased from 200 C to 250 C at a 5 C min rate. double bonds converted to epoxides, the adjacent double bond
The temperature of the injector was 250 C. Sample was diluted and the carbon in the middle (an allyl group) served as a weak
in ethyl acetate prior to injection.
The separated products from TLC were characterized by the efficiency of the reaction and to stabilize the resulting epoxides.
nuclear magnetic resonance (NMR, BruckerAvance III 600 Hz)
The epoxidation of FAMEs with the SA/AA system oen
and high resolution mass spectroscopy (MS, Bruker VPEXII) showed signicant differences from those of the FA system. The
with EI mode.
conversions of the unsaturated double bonds were similar to
electron-donating group, which would help to promote the
Fourier transform infrared spectroscopy (FT-IR) spectra were that from the FA-catalyzed reaction, indicating the increasing
obtained by using Thermo Nicolet 380 coupled with Ezominic activities of the FAMEs with the increasing double bonds.
soware. Sample was smeared as a thin layer on KBr plates. However, the OOexp value and the yield of the epoxidized FAMEs
Then IR analysis was implemented between frequency range displayed the opposite trends to those from the FA system. They
ꢄ1
4
000–400 cm
.
both decreased sharply as the number of double bonds
increased, especially with the epoxidation of MLN. The OOexp
value and yield of epoxidized MLN were as low as 0.78 and 6.32,
respectively, while these values were 8.57 and 69.45 with the FA
system. In addition, all three FAMEs showed less efficient
epoxidation with the SA/AA system than that with the FA system.
Similar results have been reported in the literature which
attributed the low epoxides yield to the strong acidic nature of
SA, that could lead to severe side reactions when epoxidized
3
3
Results and discussion
.1 Epoxidation of three FAMEs
The epoxidation reactions of three unsaturated FAMEs (MO, ML
and MLN) were carried out using H as the oxidant over SA
and FA catalysts. The results of the epoxidation were listed in
Table 1. It should be noted that unlike acetic acid (AA) in the SA-
catalysed system, FA serviced as both the acid catalyst and the
active oxygen carrier for the epoxidation in the FA-catalysed
system. The epoxidation results exhibited obvious trends as
the number of double bonds varied in the FAMEs with the FA
2 2
O
24
FAMEs contained more epoxy groups. The above results
indicated that the alkenyl group with different numbers of
double bonds may have important inuence on the epoxidation
efficiency and product selectivity. To further investigate the
mechanism of the epoxidation of FAMEs with different
numbers of double bonds, the epoxidation reaction of the above
O
catalysed system. The IV increased with the increasing number
of double bonds in the FAMEs' skeleton, as expected. The
conversion of the double bonds in MO, ML, and MLN were
a
Table 1 Epoxidation of FAMEs with two acids
Entry
Systems
FA
Substrate
IV
O
(g/100 g)
IVae (g/100 g)
C (%)
OOthe (%/100 g)
OOexp (%/100 g)
Y (%)
CꢄY (%)
1
2
3
4
5
6
MO
ML
MLN
MO
ML
118.13
173.62
223.36
118.13
173.62
223.36
39.93
43.69
55.06
48.92
54.32
48.65
66.21
74.84
75.35
58.59
68.71
78.22
6.93
9.87
12.34
6.93
9.87
12.34
3.50
6.39
8.57
2.96
3.30
0.78
50.51
64.74
69.45
42.71
33.43
6.32
15.70
10.10
3.66
15.88
35.28
71.90
b
SA/AA
MLN
a
ꢀ
b
Conditions: (C]C) : FA/AA : H
2
O
2
¼ 1 : 1.5 : 5.25 (molar ratio), 30 C, 5 h. 2 wt% SA. FA: formic acid, SA: H
2 4
SO , MO: methyl oleate, ML: methyl
linoleate, MLN: methyl linolenate. IV
o
: iodine values. IVae iodine values aer epoxidation. C: conversion of substrate. OOthe theoretical percentage of
oxirane oxygen. OOexp experimental percentage of oxirane oxygen. Y: the yield of epoxidized products. CꢄY: others.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74783–74789 | 74785

View Article Online
RSC Advances
Paper
a
Table 2 Epoxidation of vegetable oils with different fatty acid compositions
Oil
Systems
FA
IV
O
(g/100 g)
IVae (g/100 g)
C (%)
OOthe (%/100 g)
7.42
OOexp (%/100 g)
Y (%)
Olive oil
Linseed oil
127.18
200.61
36.21
32.02
50.17
38.64
71.53
74.82
74.99
80.74
5.03
4.07
6.68
2.06
67.79
54.85
59.48
18.34
b
SA/AA
FA
SA/AA
11.23
b
a
ꢀ
b
Conditions: (C]C) : FA/AA : H
2
O
2
¼ 1 : 0.8 : 2.8 (molar ratio), 50 C, 2 h. 1 wt% SA was added.
three FAMEs were tracked and the reaction products were that most double bonds were converted while only a few
identied with IR, MS and NMR to evaluate the inuences of remained unreacted (spectrum d and g). The peak around 820–
ꢄ1
these alkenyl groups on the efficiency and selectivity of the 850 cm represented the oxirane ring of the epoxides and it
epoxidation reaction.
may be slightly changed due to different functional groups
27
adjacent to the oxirane group as reported in many literature.
ꢄ1
No obvious peak around 3400 cm (–OH stretching peak) was
detected in d and g, which indicated that the ring-opening side
diol products were negligible with the epoxidation of MO in
both systems. Based on the above results, it can be concluded
that both FA and SA/AA system were effective for the epoxidation
of the FAMEs with a single double bond and gave similar
performances in the reactions, albeit a little difference in the
conversions and product yields.
3.2 Tracking of epoxidation reaction
3.2.1 Epoxidation of methyl oleate. The epoxidation
process of MO with one double bond in the skeleton of the fatty
acid was tracked by GC and the results were shown in Fig. 2. The
spectrum on the le was the epoxidation catalyzed by FA while
the right was SA. Aer the rst 1 h, three major peaks were
detected in both systems and the peaks were identied by GC-
MS. Peak 1 was the starting material MO (Fig. S1 and S4†).
Peak 2 was the target epoxidized product (Fig. S2 and S5†). And
peak 3 was fatty ketones (Fig. S3 and S5†) which were result of
the rearrangement of the epoxidized product, as reported in
3.2.2 Epoxidation of methyl linoleate. Fig. 4 showed the
tracking of the epoxidation reaction of ML over FA and AA
catalytic systems. Two major products were detected by GC,
however, whose molecular ions cannot be detected by GC-MS.
Thus, these two products were separated by TLC and the sepa-
25
much literature. As the reactions proceeded, the amount of
MO gradually reduced in both systems. About 39.93% and
1
rated products were analyzed by H-NMR and MS (see ESI†).
48.92% of the double bond remained in the FA and SA/AA
According to the results, peak 4 was the starting material
systems, respectively, aer 5 h (Table 1, entry 1 and 4). Mean-
while, the amounts of the epoxidized products increased grad-
ually and reached 50.51% and 42.71% yield in the FA and SA/AA
systems, respectively. Comparing the two acid catalysed
systems, it can be seen that the FA system had a higher reaction
rate in the formation of epoxidized products than the SA system.
This could be attributed to the higher rate constant of the in situ
formation of the peroxyacidin the FA system than in the SA/AA
system, which was the rate-determining step of the epoxidation
(
Fig. S6†) and it gradually reduced as the reaction proceeded.
Peak 5 was a doublet peak corresponding to the isomerized
epoxidized product with only one epoxy group (Fig. S7 and S9†).
Peak 6 was also a doublet peak corresponded to the isomerized
epoxidized products with two epoxy groups (Fig. S8 and S10†).
26
reaction. The IR spectra of the reaction mixture is shown in
Fig. 3. For the starting material MO (spectrum a), the peak at
ꢄ1
3
004 cm was attributed to the C]C double bond in the fatty
acid skeleton. It became a small shoulder in the product
mixture aer the epoxidation in both systems, which implied
Fig. 3 FT-IR patterns of MO, ML, MLN, epoxidation of FAMEs with FA
Fig. 2 GC traces for the epoxidation of methyl oleate with FA and SA/ (EMO-FA, EML-FA, EMLN-FA) and SA/AA (EMO-SA/AA, EML-SA/AA,
AA. EMLN-SA/AA).
74786 | RSC Adv., 2015, 5, 74783–74789
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Peak 7 was the side products that cannot be identied. The
amount of the epoxidized products (peak 5) gradually increased
as the reaction proceeded. When the reaction was stopped aer
5
h, the FA system gave more epoxidized products than that
from SA system. As listed in Table 1, the conversion percentages
of the double bonds in ML were similar in the two acid catalytic
systems (74.84% and 68.71%), however much fewer of the
epoxidized products (peak 5) were detected by GC in the SA/AA
system than in the FA system. This is due to the severer ring-
opening reaction present in the epoxidized products in the
SA/AA system than in the FA system. Further conrmation via IR
analysis of the product mixture is shown in Fig. 3, the SA/AA
system provided a stronger –OH stretching peak around 3464
Fig. 5 GC traces for the epoxidation of methyl linolenate with FA and
SA/AA.
ꢄ1
cm compared to the FA system, thus indicating the epoxi-
dized products with two epoxy group easily underwent ring-
opening and formed –OH in the presence of strong acid
catalysts than FAMEs with two or single double bonds.
However, when the weak acid FA was used, the side reactions
were greatly suppressed and more epoxides were obtained.
(Fig. 3, h). Meanwhile, these epoxidized products were more
stable in the weak acid FA system and the –OH stretching peak
in the IR spectrum (Fig. 3, e) were relatively weak.
3
.2.4 The inuences of the structure on the epoxidation.
From the above results, it was found that the structure of FAMEs
different number of double bonds) had a signicant inuence
3.2.3 Epoxidation of methyl linolenate. The epoxidation of
(
MLN was carried out with the same two acid catalytic systems.
As shown in Fig. 5, the results from the tracking of the reaction
exhibited remarkable differences between the two acids
systems. When SA was used, two groups of product peaks (9 and
on the efficiency of their epoxidation reactions. The percentage
conversions of the double bonds increased as the number of
double bonds on the fatty acid skeleton increased, regardless
which acid catalysts was used (from 66.21% to 75.35% in FA
system and 58.59% to 78.22% in SA/AA system), indicating the
FAMEs with more unsaturated bonds exhibited higher reactiv-
ities towards epoxidation. This phenomenon can be ascribed to
the electron-donating effect from the adjacent double bonds. It
has been reported in the literature that epoxidation can be
promoted by electron-donating groups adjacent to the double
10) were detected by GC in the FA reaction system. However,
they were difficult to separate by TLC. Therefore, these two
groups of products were submitted to MS analysis. The molec-
ular ion peak at m/z 331.23 [M + Na] corresponded to mono-
epoxidized product 9 and the molecular ion peak at m/z
347.22 [M + Na] corresponded to di-epoxidized product 10
(Fig. S11†). Both of the two groups of products had three
23
bonds. In addition, the generated epoxy group can also be
isomers, thus they gave multiplet peaks in the GC spectrum. On
the other hand, no obvious product peaks were detected in the
SA/AA system, which indicated that most of the epoxidized
products might be further converted to ring-opening products,
such as polyols, that cannot be detected by GC due to their high
polarity. IR analysis of the reaction products of MLN conrmed
the above speculation by which the peak intensity around 3464
23b
stabilized by the adjacent electron-donating groups. In the
current systems, ML and MLN had two or three double bonds
on their fatty acid skeletons. When one double bond underwent
epoxidation, the nearest double bond and the carbon between
them can be viewed as an allyl group which was a weak electron-
donating group (Fig. 6). Therefore, FAMEs with more double
bonds would have higher activities when undergoing epoxida-
tion reaction either with a strong or weak acid catalyst.
Furthermore, the strength of the acid played an important
role in the nal product distributions of the epoxidation of the
FAMEs. When FA was used as the catalyst and oxygen-carrier, the
percentage conversion of double bonds and the yield of epoxi-
dized products increased as the number of double bonds
increased (Table 1, entries 1–3). However, when SA was used as
the catalyst and AA was the oxygen-carrier, the conversion of
double bonds still increased while the yield of the epoxidized
products decreased sharply. The strong acid SA caused severe
ring-opening reactions during this process, especially when the
products had two or three epoxy groups (Table 1, entries 5 and 6).
Based on the above results, it can be concluded that FAMEs
with more double bonds had higher activities towards epoxi-
dation (MLN > ML > MO) in both acid catalytic systems (Fig. 6).
Nevertheless, the weak acid FA can be used in the epoxidation of
all three FAMEs with different double bonds while the strong
ꢄ1
cm in the spectrum of epoxidized MLN with SA/AA system
(
Fig. 3, i) was stronger than those of all other reaction products.
ꢄ1
The peak of the epoxy group (820–850 cm ) was very weak
which is consistent with the lower OOexp in Table 1, entry 6.
These results revealed that FAMEs with three double bonds
easily underwent ring-opening reactions over strong acid
Fig. 4 GC traces for the epoxidation of methyl linoleate with FA and
with SA/AA.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74783–74789 | 74787

View Article Online
RSC Advances
Paper
undergo side reactions in a strong acidic system (SA/AA) and
lead to a lower oxirane oxygen value. Alternatively, a weak acid
(FA) exhibited better performance in the epoxidation of the
three FAMEs and the side reactions were greatly suppressed in
this weak acidic system. Finally, the epoxidation results of the
two vegetable oils with different fatty acid compositions further
conrmed the above conclusions. The current ndings provide
useful information on the inuences of the alkenyl structure on
the epoxidation reactions which may nd important application
in the design of new catalytic systems for the epoxidation of
vegetable oils with different fatty acid compositions. Future
studies will be focused on the development of some heteroge-
neous catalysts for the epoxidation of vegetable oils.
Fig. 6 Electron effect and reactivities of different FAMEs.
acid SA was more suitable for the epoxidation of FAMEs with a
single double bond.
Acknowledgements
3.3 Epoxidation of vegetable oils
Two vegetable oils, olive oil and linseed oil, with different fatty
acid compositions were chosen for epoxidation to conrm the
effect of fatty acid structures on the efficiency of their epoxi-
dation reaction (Table 2). The olive oil mainly contains oleic and
linoleic acid (ꢃ63.7% and ꢃ25.2%, respectively) with almost no
linolenic acid. Linseed oil has a composition of oleic acid
ꢃ22.1%, linoleic acid ꢃ16.7%, and linolenic acid ꢃ52.5%.
When olive oil was submitted to epoxidation reaction, the
percentage conversions of the double bonds were 71.53% and
The authors are grateful for the nancial support by the Jiangsu
Specially-Appointed Professor program of the State Minister of
Education of Jiangsu Province and the start up foundation of
NJFU (GXL003).
Notes and references
1
(a) T. E. Bull, Science, 1999, 285, 1209; (b) A. J. Ragauskas,
C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney,
C. A. Eckert, W. J. Frederick Jr, J. P. Hallett, D. J. Leak,
C. L. Liotta, J. R. Mielenz, R. Murphy, R. Templer and
T. Tschaplinski, Science, 2006, 311, 484.
74.82% in the FA and SA/AA system, respectively. While the yield
of epoxidized products in the two acid systems were 67.79% (FA)
and 54.85% (SA/AA). The FA system gave a little higher product
yield than the SA/AA system. The linoleic acid composition in
olive oil should be the key factor for the side reactions in the SA/
AA system. When linseed oil was used as the starting material,
the conversions of the double bonds were higher than that from
olive oil, resulting from the higher reactivity of the linolenic
acid in linseed oil. The yield of the epoxidized products from
linseed oil with the FA system was still at a high level (59.48%),
which indicated that side reactions were suppressed to a
minimum level in the weak acidic systems. However, the
epoxidation of linseed oil with the SA/AA system gave a poor
yield of epoxidized products. All of the above results were
consistent with the results from the epoxidation of FAMEs.
Therefore, it can be concluded that a weak acidic system such as
FA was more preferable for the epoxidation reaction when the
vegetable oil contained more linolenic acid or linoleic acid.
2
3
Z. S. Petrovic, Polym. Rev., 2008, 48, 109.
´
(a) S. G. Tam and W. S. Chow, Polym.-Plast. Technol. Eng.,
2010, 49, 1581; (b) K. Dalai, Ind. Eng. Chem. Res., 2014, 53,
18668.
4
5
(a) A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
2
2
411; (b) M. Besson, P. Gallezot and C. Pinel, Chem. Rev.,
014, 114, 1827.
(a) R. Mythili, P. Venkatachalam, P. Subramanian and
D. Uma, Int. J. Energy Res., 2014, 38, 1233; (b) N. Narkhede,
S. Singh and A. Patel, Green Chem., 2015, 17, 89; (c)
Z. Helwani, M. R. Othman, N. Aziz, W. J. N. Fernando and
J. Kim, Fuel Process. Technol., 2009, 90, 1502; (d) L. Soh,
J. Curry, E. J. Beckman and J. B. Zimmerman, ACS
Sustainable Chem. Eng., 2014, 2, 387.
6
(a) H. Schuster, L. A. Rios, P. P. Weckes and W. F. Hoelderich,
Appl. Catal., A, 2008, 348, 266; (b) J. Salimon, N. Salih and
E. Yousif, Eur. J. Lipid Sci. Technol., 2010, 112, 519; (c)
J. O. Metzger, Eur. J. Lipid Sci. Technol., 2009, 111, 865.
4
. Conclusions
Three typical unsaturated FAMEs (methyl oleate, methyl lino-
leate and methyl linolenate) were used in the epoxidation
reaction with FA and SA/AA systems to investigate the inuences
of the alkenyl structure with different numbers of double bonds
on the epoxidation efficiency and selectivity. The epoxidation
reactions were tracked and the products were analyzed by
various technologies. It was found that FAMEs with more
7 M. T. Benaniba, N. Belhanceche-Bensemra and G. Gelbard,
Polym. Degrad. Stab., 2003, 82, 245.
8 (a) J. A. Sherringham, A. J. Clark and B. R. T. Keene, Lipid
Technol., 2000, 12, 129; (b) S. P. Bunker and R. P. Wool, J.
Polym. Sci., Part A: Polym. Chem., 2002, 40, 451.
9 B. K. Sharma, K. M. Doll and S. Z. Erhan, Green Chem., 2007,
9, 469.
double bonds have higher reactivities towards the epoxidation 10 V. V. Goud, N. C. Pradhan and A. V. Patwardhan, J. Am. Oil
reaction with both acid catalysts, which may be due to the Chem. Soc., 2006, 83, 7.
electron-donating effect of the adjacent double bonds. In 11 F. E. Okieimen, O. I. Bakare and C. O. Okieimen, Ind. Crops
addition, epoxidized FAMEs with more double bonds can easily Prod., 2002, 15, 139.
74788 | RSC Adv., 2015, 5, 74783–74789
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
1
1
2 S. Dinda, A. V. Patwardhan, V. V. Goud and N. C. Pradhan, 20 C. Paquot. Standard Methods for the Analysis of Oils, Fats and
Bioresour. Technol., 2008, 99, 3737. Derivatives Part -1, Pergamon Press, Germany, 1979.
3 (a) A. Campanella and M. A. Baltan ´a s, Lat. Am. Appl. Res., 21 A. P. Chavan and P. R. Gogate, J. Ind. Eng. Chem., 2015, 21,
2
005, 35, 205; (b) A. Campanella and M. A. Baltan ´a s, Lat.
842.
Am. Appl. Res., 2005, 35, 211.
22 (a) E. Milchert, A. Smagowicz and G. Lewandowski, J. Chem.
Technol. Biotechnol., 2010, 85, 1099; (b) E. Milchert,
A. Smagowicz and G. Lewandowski, Org. Process Res. Dev.,
2010, 14, 1094.
1
4 (a) S. Dinda, V. V. Goud, A. V. Patwardhan and N. C. Pradhan,
Asia-Pac. J. Chem. Eng., 2011, 6, 870; (b) A. Campanella and
M. A. Baltan ´a s, Chem. Eng. Process., 2007, 46, 210.
1
5 (a) A. Campanella, M. A. Baltan ´a s, M. C. Capel-S ´a nchez, 23 (a) W. Zheng, R. Tan, L. Zhao, Y. Chen, C. Xiong and D. Yin,
J. M. Campos-Mart ´ı n and J. L. G. Fierro, Green Chem.,
004, 6, 330; (b) L. A. Rios, P. Weckes, H. Schuster and
RSC Adv., 2014, 4, 11732; (b) A. H. Hoveyda, D. A. Evans and
G. C. Fu, Chem. Rev., 1993, 93, 1307.
2
W. F. Hoelderich, J. Catal., 2005, 232, 19; (c) D. Kumar and 24 (a) A. Campanella, C. Fontanini and M. A. Baltan ´a s, Chem.
A. Ali, Energy Fuels, 2012, 26, 2953; (d) M. Guidotti,
E. Gavrilova, A. Galarneau, B. Coq, R. Psaro and
N. Ravasio, Green Chem., 2011, 13, 1806.
6 C. Tiozzo, C. Bisio, F. Carniato, L. Marchese, A. Gallo,
N. Ravasio, R. Psaro and M. Guidotti, Eur. J. Lipid Sci.
Technol., 2013, 115, 86.
Eng. J., 2008, 144, 466; (b) M. R u¨ sh, G. Klaas and
S. Warwel, ed. G. Knothe and J. T. P. Derksen, New
Oxidation Methods for Unsaturated Fatty Acids, Esters, and
Triglycerides, Recent Developments in the Synthesis of Fatty
Acid Derivatives, AOCS Press, Champaign, IL, 1999, p. 157.
25 (a) L. A. Rios, B. A. Llano and W. F. Hoelderich, Appl. Catal.,
A, 2012, 445, 346; (b) R. O. C. Norman and J. M. Coxon,
Principles of Organic Synthesis, 1993, 3, 590.
1
1
1
1
7 A. K. R. Somidi, R. V. Sharma and A. K. Dalai, Ind. Eng. Chem.
Res., 2014, 53, 18668.
8 Y. Leng, J. Zhao, P. Jiang and J. Wang, ACS Appl. Mater. 26 L. H. Gan, S. H. Goh and K. S. Ooi, J. Am. Oil Chem. Soc., 1992,
Interfaces, 2014, 6, 5947. 69, 347.
9 C. Paquot. Standard Methods for the Analysis of Oils, Fats and 27 (a) N. Kim, Y. Li and S. Sun, Ind. Crops Prod., 2015, 64, 1; (b)
Derivatives Part I, Pergamon, Oxford, United Kingdom, 6th
edn, 1979.
P. K. Gamage, M. O'Brien and L. Karunanayake, J. Natn, Sci.
Foundation Sri Lanka, 2009, 37, 229.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74783–74789 | 74789