Triacylglycerol self-metathesis over highly chemoselective methyltrioxorhenium supported on ZnCl2-promoted mesoporous alumina

Article abstract of DOI:10.1016/j.apcata.2014.04.016

Since triolein is the main constituent of most oils such as olive and high oleic sunflower oils, self-metathesis of this triacylglycerol at mild reaction conditions was investigated via a heterogeneous approach, in the presence of ZnCl₂-modified mesoporous alumina supported methyltrioxorhenium (MTO) as a catalyst. Both intra and intermolecular metathesis reactions were observed. While, 9-octadecene, dimers and trimers of the triacylglycerol were isolated, intermolecular cyclization metathesis products were also obtained. The MTO-based catalyst was found to be highly chemoselective affording only desired metathesis product formation. In contrast, more than 50% of triolein was converted to undesired solid product in the presence of the homogeneous Grubbs 2nd generation catalyst. Triolein metathesis products were effectively separated by chromatographic techniques and identified using mass spectrometric and NMR analyses. The reported results show that 3%MTO/ZnCl₂-Al ₂O₃-meso is an efficient and selective heterogeneous catalyst for bulky functionalized olefins metathesis such as triacylglycerols and edible oils, avoiding the use of costly homogeneous catalysts, or the use other rhenium-based catalysts with toxic promoters such as R₄Sn.

Full text of DOI:10.1016/j.apcata.2014.04.016

Accepted Manuscript
Title: Triacylglycerol Self-metathesis Over Highly
Chemoselective Methyltrioxorhenium Supported on
ZnCl2-Promoted Mesoporous Alumina
Author: Subha Kumaraswamy Pillai Abdelnasser Abidli
Khaled Belkacemi
PII:
S0926-860X(14)00221-X
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2014.04.016
APCATA 14789
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
18-11-2013
28-3-2014
5-4-2014
Please cite this article as: S.K. Pillai, A. Abidli, K. Belkacemi, Triacylglycerol
Self-metathesis Over Highly Chemoselective Methyltrioxorhenium Supported on
ZnCl2-Promoted Mesoporous Alumina, Applied Catalysis A, General (2014),
http://dx.doi.org/10.1016/j.apcata.2014.04.016
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Triacylglycerol Self-metathesis Over Highly Chemoselective
Methyltrioxorhenium Supported on ZnCl2-Promoted Mesoporous Alumina
Subha Kumaraswamy Pillai a,b,d, Abdelnasser Abidli a,c,d, Khaled Belkacemi a,c
*
aDepartment of Soil Sciences and Food Engineering; bDepartment of Chemical Engineering; cDepartment
d
of Food Sciences and Nutrition; Centre de Recherche en Catalyse et Chimie Verte (C3V ex: CERPIC),
Laval University, Quebec G1V0A6 (Canada).
*Corresponding author: Tel: 1 (418) 656 2131 ext: 6511; Fax: 1(418) 656 3723; E-mail address:
khaled.belkacemi@sga.ulaval.ca
Abstract
Since triolein is the main constituent of most oils such as olive and high oleic sunflower
oils, self-metathesis of this triacylglycerol at mild reaction conditions was investigated via
a
heterogeneous approach, in the presence of ZnCl2-modified mesoporous alumina supported
methyltrioxorhenium (MTO) as a catalyst. Both intra and intermolecular metathesis reactions
were observed. While, 9-Octadecene, dimers and trimers of the triacylglycerol were isolated,
intermolecular cyclization metathesis products were also obtained. The MTO-based catalyst
was found to be highly chemoselective affording only desired metathesis product formation.
In contrast, more than 50% of triolein was converted to undesired solid product in the
presence of the homogeneous Grubbs 2nd generation catalyst. Triolein metathesis products
were effectively separated by chromatographic techniques and identified using mass
spectrometric and NMR analyses. The reported results show that 3%MTO/ZnCl2-Al2O3-meso
is an efficient and selective heterogeneous catalyst for bulky functionalized olefins metathesis
such as triacylglycerols and edible oils, avoiding the use of costly homogeneous catalysts, or
the use other Rhenium-based catalysts with toxic promoters such as R4Sn.
Keywords: Self-metathesis; Triolein; Methyltrioxorhenium (MTO); ZnCl2-promoted mesoporous
alumina; Metathesis polymerization.
Page 1 of 39

1. Introduction
In an era of ever increasing global awareness for the needs of environmental liability,
sustainability and natural-resource management, there is a strong push toward the adoption of a bio-based
economy. Lowering the cost and improving the properties of bio-based materials are the challenging key
steps to achieve this goal.
Replacement of petroleum-based raw materials by renewable and sustainable resources is of great
interest, because of economic and environmental aspects [1] . Plant-based resources can be used for the
synthesis of a wide range of value-added products comprising pharmaceuticals [2], cosmetics, polymers
[3], fine chemicals [4] and bio-fuels [5]. This has subsequently gathered a higher attention in non-food
applications of seed oils during the limited supply and the escalating price of petroleum and petroleum-
based products [6].
Oilseed-processing, food manufacturing, animal-product rendering and meat packing generate large
quantities of oils/fats streams. As a result, the production systems for these fatty materials have been
underused and often have been operated below full capacities, leading to disposal problems or causing a
market surplus to many product sectors [7]. However, these low-cost fatty materials are suitable, readily
available starting materials for many gainful oleochemical processes. About 14% of the annual world
production of fats and oils (103 million tons) are used in the oleochemical industry as starting material
producing a wide range of interesting products [8]. More than 90% of all oleochemical reactions of fatty
acid esters are carried out at the carboxy function [9]. Moreover, carbon–carbon double bond
transformations by hydrogenation, epoxidation, ozonolysis, metathesis and hydroformylation are
becoming industrially more important [7]. Among these transformations, catalytic olefin metathesis is a
powerful tool in organic synthesis for the alkenes carbon–carbon double bond formation [10].
Catalytic metathesis of natural oils and fats and their derivatives is considered as a clean synthetic
route from the green chemistry aspects [11]. Fats and oils containing carbon-carbon double bonds can
undergo transformation by metathesis providing access to highly desired products [12-16], while various
olefin metathesis pathways are possible, including self-metathesis (SM), cross-metathesis (CM), ring
closing metathesis (RCM), ring-opening metathesis (ROM) and ROM-polymerization (ROMP) as well as
acyclic diene metathesis polymerization (ADMET) [17, 18]. Metathesis of triacylglycerols containing
unsaturated long-chain fatty acids occurs intramolecularly via SM [13-16], as well as intermolecularly via
CM predominantly [19-24].
A number of catalytic systems have been developed to carry out the metathesis of triacylglycerols
and vegetable oils. Homogeneous catalysts consisting of the transition metals 6-7 group were found to be
highly active metathesis catalysts. Catalytic systems such as WCl6/Sn(CH3)4 [25-28] or the commercially
available Grubbs or Hoveyda-Grubbs catalysts [29] have been reported to be efficient and selective
Page 2 of 39

catalysts for vegetable oils and triacylglycerols metathesis. However, their low stability and
decomposition hindered the recycling process [25]. In addition, their high costs limited their
implementation at commercial scale. This situation promoted intensive studies towards the development
of more cost-effective and recyclable catalysts, where heterogeneous catalysts could be a plausible
alternative.
Early discoveries in the field of homogeneous catalysis investigating functionalized alkenes widely
reported a high tolerance of tungsten-tin and molybdenum alkylidenes, then this approach was emerged
with other transition metals complexes as good initiators for functionalized olefin metathesis [30]. Most
of reported papers describing methyloleate, triacylglycerol and/or vegetable oil metathesis have been
achieved using homogeneous catalysts, whereas only few studies were reported using heterogeneous
catalysts, such as Al2O3 or SiO2-Al2O3 supported Re2O7-based solid catalysts promoted with SnR4
(R=Me, Et, Bu, Alkyl) [31, 32]. However, all these catalyst systems suffered from disposal of toxic
solvents and toxic spent catalysts due to the presence of tin-based promoter, as well as sensitivity to
oxygen which hampered their regeneration and prevent the metathesis reaction pathway [33].
To the best of our knowledge, this is the first report describing the self-metathesis of triacylglyceols
in the presence of heterogeneous mesostructured catalyst. Recently, we have reported for the first time
that methyltrioxorhenium (MTO) impregnated on ZnCl2-modified mesoporous alumina (MTO/ZnCl2-
Al2O3-meso) is an efficient catalyst for self-metathesis of methyloleate, a model molecule of
triacylglycerols present in vegetable oils, with high activity and good selectivity towards desired
metathesis products [34, 35].
The aim of this work is to study for the first time the self-metathesis of more bulky functionalized
molecules; triacylglycerols, like triolein over MTO/ZnCl2-Al2O3-meso as a heterogeneous catalyst in
comparison to the most used homogeneous commercial Grubbs’ second generation catalyst. Since all the
details about MTO//ZnCl2-Al2O3-meso catalyst synthesis and its characterization were already reported in
our previously reported work [34, 35], efforts will be focused at presenting comparatively the product
time course profiles, and elucidating the chemical structure of the products obtained from the triolein self-
metathesis over both catalysts.
Page 3 of 39

2. Experimental
2.1. Materials
All reagents were of the highest purity and were used without further purification. Aluminium-tri-
sec-butoxide (C12H27AlO3, 97%), zinc chloride (ZnCl2, 99.9%), methyltrioxorhenium (MTO, CH3O3Re
71-76%), glyceryl trioleate (triolein), 2,7-dichlorofluorescein, benzylidene [1,3-bis(2,4,6-
trimethylphenyl)-2-imidazolidinylidene] dichloro (tricyclohexylphosphine) ruthenium (2nd generation
Grubbs catalyst) were obtained from Sigma-Aldrich Canada Ltd (Oakville, ON, Canada). Trielaidin was
purchased from Nu-Chek Prep Inc. (Elysian, USA). Silica gel 60 F254 precoated glass-backed TLC plates
were used for thin layer chromatography (EMD Millipore Corp). Triblock copolymer of poly(ethylene
glycol)-poly(propylene glycol)-poly(ethylene glycol), (Pluronic P123, EO20PO70EO20, Mav = 5900 g/mol)
was graciously offered by BASF. All the solvents used were anhydrous.
2.2. Materials synthesis and characterization
2.2.1. Synthesis of Al2O3 -meso
Mesoporous alumina (Al2O3-meso) was synthesized according to the procedure described by Ray et
al. [36] using Pluronic acid as the templating agent with further modifications reported in our previous
work [34, 35].
2.2.2. Synthesis of ZnCl2-Al2O3-meso
Modification of Al2O3-meso by incorporation of zinc chloride was performed according to Oikawa
et al. [37] with further modifications as reported in our previous work [34, 35]. In a typical experiment,
the addition of ZnCl2 (0.668g, 4.9 mmol) in ethanol (5 mL) was done drop wise to Al2O3-meso (2g) under
stirring and the ethanol was allowed to dry (after 24h). The dried sample obtained was heated under an air
o
flow to 400˚C at ramping rate of 2 C/min and calcined at 400˚C for 4h to give ZnCl2-modified Al2O3-
meso with optimized Al/Zn ratio of 8.
2.2.3. Synthesis of the MTO/ZnCl2-Al2O3-meso catalyst
The MTO/ZnCl2-Al2O3-meso catalyst was synthesized as previously described in our work [34, 35],
where we reported that 3wt% is the optimum loading of MTO [34-35]; ZnCl2-Al2O3-meso support was
heated under nitrogen at 540 oC for 2h before impregnation of MTO.
Page 4 of 39

2.2.4. Catalyst characterization
The MTO catalyst was characterized using nitrogen physisorption, X-ray diffraction (XRD), X-ray
photoelectron spectroscopy (XPS) and Transmission electron microscopy (TEM) techniques. Further
elemental CNS composition analysis was performed for both fresh and spent MTO/ZnCl2-Al2O3-meso.
Nitrogen adsorption/desorption isotherms of calcined samples were obtained using a volumetric
adsorption analyzer (Model Autosorb-1, Quantachrome Instruments, Boyton Beach, FL) at -196°C (77
K). Before the adsorption analysis, the samples were degassed for 3 h at 200°C. Total pore volume was
estimated from the amount adsorbed at 0.99 relative pressures. Pore size distributions were calculated
using the desorption branch of the N2 adsorption/desorption isotherms and the Barrett-Joyner-Halenda
(BJH) [38].
Powder X-ray diffraction (XRD) patterns were obtained using an Ultima III Rigaku Monochromatic
Diffractometer using CuKα radiation (λ=1.5406 Å) at 40 kV and 44 mA. Small angle powder diffraction
patterns were acquired at a scanning rate of 0.5°/min in the 2θ range of 0.6-3°.
Transmission electron microscopy (TEM) images were taken on a JEM-3010 electron microscope
(JEOL, Japan) with an acceleration voltage up to 200 kV.
Elemental Carbon/Nitrogen/Sulfur (CNS) composition was carried out using TruMac series
instrument (LECO Corp). CNS analysis was performed for both fresh and spent MTO/ZnCl2-Al2O3-meso.
The combustion was carried out at 1350 °C under atmospheric gas flow, followed by a complete
oxidation under ultra pure oxygen within 5 min. The instrument system measurement accuracy was ± 1.0
%.
2.3. Metathesis reaction and analysis of the products
2.3.1. Metathesis reaction
Self-metathesis reactions were carried out in a small batch and controlled-inert atmosphere quartz
reactor. Typically, glyceryl trioleate or triolein (100 mg) in hexane (1 mL) was added to the catalyst (100
mg with 3wt% MTO loading) in dry conditions at the investigated temperature of 45°C. The stirring rate
of 400 rpm was taken constant throughout the reaction time. These conditions were found to be suitable
for optimum activity of the 3wt% MTO/ZnCl2-Al2O3-meso catalyst, as previously reported in our work
[34, 35]. Metathesis reaction over the homogeneous 2nd generation Grubbs catalyst was performed under
the same conditions described above using 2.67 mg of catalyst.
2.3.2. Product analysis
Bulkier molecules were obtained from the triolein metathesis, their separation and analysis by gas
chromatography was difficult. So the obtained products were subjected to column chromatography in
Page 5 of 39

order to separate them and the fractions obtained were identified by Thin Layer Chromatographic (TLC)
analysis. Among the isolated fractions, alkene and triolein were quantified using gas chromatographic
1
13
analysis and the remaining products were identified through H- and C-NMR analysis, in addition to
ESI-MS analysis.
2.3.2.1. Column chromatography
For the separation and purification of the reaction mixture obtained with either Grubbs’ catalyst or
MTO-based heterogeneous catalyst, a glass column (60 cm long x 1.5 cm diameter) was clamped
vertically and a small piece of cotton was used as plug to support the silica gel as adsorbent. Care should
be taken when placing the cotton compressed to support the packing but loose enough to allow the solvent
flow. The solvent used comprised hexane and diethylether mixed with a determined hexane:diethylether
volume ratio (80:20). Sand was added to form 1 cm layer on top of the plug and the column was gently
tapped to expel any trapped air. The silica gel was taken in a flask and enough solvent was added to form
slurry. The slurry was poured onto the column with opening of the stopcock to allow draining of excess
solvent.
The column packing must be uniform with no air bubbles trapped in it. A final layer of sand was
added to the column to protect the packed silica gel. After the column packing, the sample was introduced
in the column by adding carefully along the sides in order to avoid disturbing of the column packing.
Elution of the sample was done by adding the solvent and the fractions were collected. About 40
fractions were collected and each of them was tested by TLC.
2.3.2.2. TLC analysis
The isolated fractions from column chromatography were identified by TLC. In a typical
experiment, the commercially prepared 60-mesh Silica gel 60 F254 precoated glass-backed TLC plates
(Whatman K6F, Maidstone, UK) were used for the component identification in the different fractions.
Fractions were spotted in the plates and they were placed in the glass chamber saturated with the solvent
system for developing it. The solvent system used is the same as the eluting solvent in the column
chromatography, i.e. hexane:diethylether (80:20). The plates were visualized under UV light (254 nm).
For fractions that were in smaller concentration, the spots were visualized under UV by spraying the
fluorescent dye 2,7-dichloro fluorescein ethanol solution (0.2 w%).
2.3.2.3. Gas Chromatographic analysis
The fractions containing triolein were pooled and analyzed by gas chromatography in HP6890 gas
chromatograph with capillary inlet (split-split) mode with a constant split ratio (1:50). The capillary
column used was RTX-65 (30m long x 0.25mm i.d x 0.10μm film thickness) with H2 as carrier gas. The
Page 6 of 39

injector and detector were set at 400°C. The oven temperature profile was: initial temperature of 250°C,
ramp at 10° C/min to 335°C and then to 360°C at ramp of 1.5° C/min and hold at this temperature for 3
min. The results obtained from the FID detector were analyzed using Chemstation software. The fractions
containing triolein and trielaidin were identified using commercially available pure triolein and trielaidin
triacylglycerol (TAG) standards. The quantification of triolein and trielaidin levels in the samples was
carried out using tripalmitin as an internal standard. The quantification of 9-octadecene was carried out by
HP6890 series gas chromatograph system equipped with a flame ionization detector using dodecane as
internal standard. The column used was BPX-70 (60m x 0.25mm x i.d x0.25μm film thickness) with
hydrogen as carrier gas. The oven temperature profile was: initial temperature 60°C, ramp at 10° C/min to
190°C, hold for 15 min; ramp at 5°C/min to 240 °C, hold for 10 min [14]. The collected data were
analyzed using Chemstation software. All the solvents and standards used were GC grade.
2.3.2.4. Mass spectroscopy (MS) Analysis
Electrospray ionization mass spectrometry (ESI-MS) analyses were performed using Agilent
technologies 1100 series LC/MS Trap by direct infusion of sample (10 µL/min) using KD Scientific
infusion pump using the procedure reported by Byrdwell and Neff [39] with little modifications. The
analysis was performed at positive mode with the following operating parameters: dry temperature
325°C, nebulizer 15 psi, dry gas (4L/min), average of 7 spectra with a scan range m/z (600-2200). The
sample was diluted in ethylacetate:acetonitrile (1:1) and the sensitivity of the analysis was enhanced using
ammonium acetate solution (20 mmol in 1:4 H2O:acetonitrile). Data acquisition and analysis were carried
out using LC-MSD Trap software 5.2 (Bruker Daltonik GmBH).
2.3.2.5. Nuclear magnetic resonance spectroscopy (1H- and 13C-NMR) analysis
1
One dimensional (1D) H- and 13C-NMR investigations were performed on a Varian 500DD2-
VNMRS 500 system (Agilent technologies) at 25°C; 1H-NMR spectra were acquired over 8 ppm spectral
window with a 1-s recycle delay with a pulse of 45 degrees and acquiring time of 4.089 s. 13C-NMR
spectra were recorded over 230 ppm spectral window with a 1-s recycle delay with a pulse of 45 degrees
on a power of 43 db continuously on WALTZ-16 modulated system. CDCl3 was used as solvent for
analysis and the processing of the acquired data was done using GRAMS software (Thermo Fisher
Scientific).
Page 7 of 39

3. Results and discussion
3.1. Characterization of textural and surface properties of the materials
Even though characterization of the MTO/Al2O3-meso was already reported in our previous works
[34, 35], it’s worth to demonstrate that the re-synthesized catalyst for the special needs of the present
work shows identical characterizations, confirming thereby, the repetitiveness of its formulation.
Indeed, nitrogen physisorption of the mesoporous Al2O3 support and MTO/ZnCl2-modified
mesoporous alumina catalyst reveals a mesoporous structure with high BET surface area up to 492 m2/g
with narrow pore size distribution centered at 8 ± 0.1 nm of average pore size and a total pore volume of
1.2 cm3/g, as presented in Table 1. By incorporation of MTO and after modification with ZnCl2, the BET
surface area slightly decreased to 418 m2/g. However the total pore volume and the average pore size
remained nearby 1 cm3/g and 8 nm, respectively.
Small-angle X-ray diffraction patterns of mesoporous alumina, ZnCl2-modified mesoporous
alumina as well as 3wt% MTO/ZnCl2-Al2O3-meso catalyst are shown in Fig. 1 which exhibit a strong
broad peak centered at 2Φ ≈ 0.65°. Similar low angle XRD patterns were previously reported by
Bagshaw and Pinnavaia [39] for mesoporous silica using amines as templates and also by Ray et al. [36]
using aluminum tri-sec-butoxide and TEOS in the presence of Pluronic 123 as a templating agent. These
authors associated these patterns to the presence of disordered worm-hole like structure. These results are
consistent with the TEM analysis depicted in Fig. 2a. The TEM micrographs covering various
agglomerates and closer regions confirm the presence of such a wormhole pore structure throughout the
material.
The porosities and surface areas of alumina samples were investigated by N2
adsorption−desorption at −196 °C. The prepared materials presented an IV-type isotherm which was
illustrated in our previously reported work [34, 35], approving the mesoporous structure with
interconnected cylinder pore. In good accordance with the N2 adsorption−desorption analysis results, the
TEM micrograph showed that they presented an interconnected worm-like pore network. Furthermore,
some slight changes were observed on this structure when the mesoporous alumina was modified with
ZnCl2 where less-identified interconnected pores regions were detected (Fig. 2b). This may be attributed
to the presence of Zn-Cl, and/or Cl-Zn-Cl attached to the OH groups of alumina.
A promoting effect of ZnCl2 in functionalized olefins metathesis such as methyloleate was
reported in our previous work [34, 35]. It is worth mentioning that these modifications were not
investigated in the present work. Surface study and characterization of alumina and ZnCl2-modified
alumina supports are currently underway in our laboratory, to establish the structure-activity relationship
of the ZnCl2-modified alumina throughout metathesis pathway, and results will be reported shortly.
Page 8 of 39

3.2. Self-metathesis of triolein
Metathesis is a clean catalytic reaction, where olefins are converted into new products through
scission and regeneration of C=C double bonds. It is a powerful tool in organic synthesis providing access
to many interesting products in an efficient and selective manner. Metathesis reaction mechanism was
first established by Herrison and Chauvin [40, 41] where a metal alkylidene forms a metallocyclobutane
with the alkene, then a rearrangement is occurred affording new alkene and metal alkylidene with a
different redistribution of alkene fragments [42]. The mechanism is depicted in Scheme 1.
R1
R2
R1
R2
R1
R2
[M]
[M]
R3
[M]
R3
R3
Scheme 1. The general mechanism of olefin metathesis reaction by Herrison and Chauvin [40].
Volatile alkenes are expected from triacylglycerols metathesis; diacylglycerols by intermolecular
reaction and monoacylglycerols by intramolecular reaction [42-44]. Since the triacylglycerols possess
three fatty acid fragments containing double bonds, which are subjected to metathesis conditions, there is
plausible intramolecular pathway for metathesis reaction. Hence both intra- and intermolecular reactions
are possible as depicted in Scheme 2.
+
Intra
+
Inter
Scheme 2. Intra- and intermolecular pathways for triolein metathesis [24].
In the intramolecular cyclization, a ring-closing metathesis can occur between two fatty acid
fragments of triolein, while the intermolecular pathway can occur via a metathesis between two identical
triolein molecules (self-metathesis) [42]. The formation of inter and intramolecular products is similar to
the mechanism initially proposed by Van Dam et al. [24].
The self-metathesis of triolein over MTO/ZnCl2-Al2O3-meso catalyst afforded oligomers (n=2 or
higher) and 9-Octadecene, the primary metathesis products.
Page 9 of 39

After stopping the reaction by addition of acetone, with either 3% MTO/ZnCl2-Al2O3-meso or 2nd
generation Grubbs catalysts, two distinct phases (liquid and solid) were observed in the reaction mixture.
After centrifugation the liquid phase was separated, and then subjected to column chromatography (CC).
In the case of MTO/ZnCl2-Al2O3-meso catalyst, the solid phase contains mainly the recycled catalyst
which was stocked for other use. From the liquid phase, exactly 40 fractions were obtained by CC; each
fraction was analyzed by thin layer chromatography (TLC).
TLC analysis showed that the first six fractions corresponded to the solvent used, the fractions #7
to #14 contained 9-Octadecene which was compared with the prepared standards from self-metathesis
reaction of methyloleate. The fractions # 15 to 19 were identified as triolein and its isomer trielaidin
confirmed by commercial standards. The remaining fractions (i.e. # 20 to 40) were pooled together and
identified as soluble metathesis products and subjected to another set of CC and TLC analyses. Again, 40
sub-fractions were collected. Similarly the sub-fractions # 1-16 contained solvent. The sub-fractions #
17-19 contained traces of triolein and trielaidin mixture (denoted MIXTURE 1 in the text). The sub-
fractions # 20-25 (MIXTURE 2 in the text), # 26-29 (MIXTURE 3, in the text), # 30-33 (MIXTURE 4, in
the text), # 34-37 (MIXTURE 5, in the text) were attributed to the metathesis products while sub-fractions
# 37-40 contained only the solvent. Quantification of 9-Octadecene, triolein and trielaidin in the reaction
mixture was done using GC with appropriate internal standards.
Afterwards, the centrifuged solid phase was filtered and washed with hexane to remove any traces
of the liquid product. The white solid observed with Grubbs 2nd generation catalyst, has a sheet-like form
when dry (Fig. 3) which is insoluble in all organic solvents. On the other hand with the MTO-based
catalyst, the solid phase has a yellowish-brown colour, in an identical form of the prepared catalyst, while
no other solid products were observed. After centrifugation and washing, both fresh (before triolein
metathesis reaction) and spent (used) MTO/ZnCl2-Al2O3-meso catalysts as well as the catalyst support
(ZnCl2 modified Al2O3-meso) had undergone a CNS composition analysis, to see if any carbonaceous
product deposition on the catalyst was present after the metathesis reaction was completed.
The results showed that ZnCl2-modified mesoporous alumina exhibited traces of carbon (~ 2.3 ±
1.0 %) attributed to the remained pluronic acid template after the calcination step. The imprenation of
MTO on the support with 3% loading increased very slightly the carbon content to approximately 3.7
±1.0 %. The spent catalyst seemed to have a tiny little more carbon (~ 4.3 ± 1.0 %), about 0.6% more
than the fresh catalyst (< to the measurement accuracy of the elemental CNS analyzer, 1.0 %) which is
considered as an insignificant shift in carbon composition, even without taking into account the
instrument system measurement accuracy (± 1.0 %). This allows saying that no carbonaceous products
were deposited on the spent catalyst.
Page 10 of 39

Table 2 depicts the results obtained; triolein conversion, metathesis solid products, 9-Octadecene,
and soluble metathesis product yields of both catalysts for triolein self-metathesis, within 60 min at the
investigated conditions.
A conversion of 65 % was reached for the self-metathesis of triolein over 3% MTO/ZnCl2-Al2O3-
meso, achieving ~ 22 % yield of soluble metathesis products, and ~ 4% of 9-Octadecene, while ~ 40% of
trielaidin, the triolein trans-isomer, was obtained as the non-productive metathesis product. Noteworthy
that, the cis-trans isomerization route during metathesis reaction has been widely studied, and reported to
be occurring competitively with metathesis reaction mechanistic pathway, while the isomerization
products resulted from different active carbene bond scissions [43-50]. This is in a good agreement with
previous papers reporting the trielaidin as a non-productive triolein metathesis [43-50]. Moreover; all the
formed products were also detected in their trans conformation in higher proportion than cis
conformation [51, 52].
In the presence of the 2nd generation Grubbs catalyst, ~ 95 % triolein conversion was observed,
achieving ~ 20 % yield of soluble metathesis products, 20 % of 9-Octadecene and ~ 3% of trielaidin.
Remarkably, despite the lower conversion with the heterogeneous MTO catalyst, no solid product was
formed; meanwhile, the desired soluble metathesis products yield is slightly higher than obtained with
homogeneous catalyst. The white solid formation with the 2nd generation Grubbs catalyst was observed
within only 5 min consuming about 50% of the starting material. It may account to the non-desired highly
cross-linked polymers. Similar polymer formation was reported by Tian and Larock [51] on triolein
metathesis using 1st generation Grubbs catalyst as solid material produced in 87% yield, which was
insoluble in all organic solvents. Similarly, the formation of this product was also observed for self-
metathesis of glycerol trilinoleate using the same catalyst [51]. The cross-linking polymerization may be
associated with the additional polymerization due to the three double bonds of the triolein which behave
as trifunctional monomer to produce dendrimers [51].
It is worth noting that the results of carbon content by CNS analysis are in agreement with the
reaction mixture analysis (Table 2) indicating no traces (0.00%) of metathesis solid products formed
and/or deposited on the catalyst surface when using MTO/ZnCl2-Al2O3-meso as a catalyst.
3.3. Quantification and identification of soluble metathesis products
3.3.1. Product quantification
After column chromatography (CC) separation and thin layer chromatographic (TLC) analysis,
the sub-fractions denoted MIXTURE 1, 2, 3, 4, and 5 were obtained. The fraction containing triolein self-
metathesis products at the investigated reaction conditions over both catalysts were quantified by GC
analysis and summarized in Table 2.
Page 11 of 39

GC analysis of these sub-fractions containing soluble metathesis products showed almost similar
patterns with the appearance of four to five spots in TLC plates, when visualized under UV by spraying
the dye for both catalysts. The soluble metathesis products overall yield reached 20 to 26 wt%, while all
products structures (Components 1 to 5) are illustrated in Fig. 4. The yield of each detailed sub-fraction
did not exceed 9 %, when the triolein self-metathesis was carried out over MTO/ZnCl2- Al2O3-meso
catalyst, where Components 1 and 2 were the major products. In contrast, with the 2nd generation Grubbs
catalyst, the dimer of triacylglycerols (TAG) molecules obtained via intramolecular metathesis and
Component 3 is the major product.
Further analyses of these sub-fractions were accomplished by mass spectroscopy (ESI-MS) and
1H- and 13C-NMR for their structural identification.
3.3.2. Mass spectrometric analysis
ESI-MS analysis results of the metathesis products are presented in Table 3. Hence, the peak
corresponding to 902.8 [M + NH4]+ value in MIXTURE 1 collected with both catalysts (MTO/ZnCl2-
Al2O3-meso and 2nd generation Grubbs catalysts), represents a mass of 884.8, which is the triolein or
trielaidin molecular weight (Component 1). This is also is confirmed by the mass spectrometric analysis
of the standards. This is attributed to the triolein remained after reaction and the trielaidin formed by non-
productive metathesis over both catalysts as indicated in Table 2.
The MS spectra of MIXTURE 2 for both catalysts (see Figures 5 and 6) consisted primarily of
1536.3 [M + NH4]+ peak which corresponds to the dimer (Component 2) with actual mass of 1518.3
formed through the two triolein molecules metathesis and elimination of one octadecene molecule. The
ESI-MS molecular mass calculation ((884.8*2)-252=1517.6) fits well with the actual molecular weight of
the component 2 (1518.3) as indicated in Table 3 where 252 is the mass of octadecene eliminated during
the reaction.
The MS spectra of MIXTURE 3 for both catalysts revealed the presence of primarily 3 peaks with
2168.8, 1536.3 and 1283.1 [M + NH4]+ for which the actual masses are 2150.8, 1518.3, and 1265.1,
respectively. The first peak corresponds to the trimer (Component 3) formed through three triolein
molecules cross-metathesis and elimination of 2 ocatadecene molecules, while the ESI-MS molecular
mass calculation confirms ((884.8*3)-(2*252) =2150.4) the actual mass of 2150.8. The two remaining
peaks revealed the presence of dimer (Component 2), already discussed. The spectra revealed also a
modified triolein dimer in which an intramolecular ring-closing metathesis reaction occurred in one of the
two attached monomers ((884.8*2)-(2*252) =1265.6), corresponding to the synthesis of Component 4.
The possible triolein metathesis products with their corresponding masses were elucidated by Li et al.
Page 12 of 39

[52] which served as a reference to identify the products formed in the present work in the presence of
either 3% MTO/ZnCl2-Al2O3 -meso or 2nd generation Grubbs catalyst.
The MS-spectrum of MIXTURE 4, in the case of MTO/ ZnCl2- Al2O3-meso, revealed mainly 3
peaks at 2168.8, 1536.3 and 1263.1 [M + NH4]+ corresponding to the presence of same molecules found
in MIXTURE 3. On the other hand, the MS-spectrum of MIXTURE 4 obtained with Grubbs catalyst,
exhibited essentially 2 peaks at 1283.1 and 1664.3 [M + NH4]+. The first one indicates the presence of
Component 4, while the second one is in agreement with the actual mass of Component 5 (1646.3)
corresponding to the modified triolein trimer in which an intramolecular ring-closing metathesis reaction
occurred in two of the three attached monomers.
No discernible differences were detected between the MS-spectra of MIXTURE 5 and those of
MIXTURE 4 for both catalysts, that have been already discussed as presented in Fig. 5 and 6 as well as in
Table 3.
The products formed with the MTO catalyst are in agreement with the results reported by Refvik
and Larock [26] through his homogeneous approach for metathesis of Soybean oil. Also similar products
were obtained by Tian and Larock [51] for metathesis of triolein where formation of dimer, trimer,
tetramer and cyclic product was observed in the presence of Grubbs’ ruthenium catalyst. Furthermore,
similar results with structural identification were reported recently for metathesis of triolein using 2nd
Generation Grubbs catalyst [52].
It is worth to mention that when the mass spectrometric peaks were magnified, expected signal
were observed, i. e. 1535.4 [M + NH4]+, but multiple signals additionally appeared (before and after) at
1521.3, 1550.3, 1564.3, 1578.3 [M + NH4]+ which differ with 14 to 15 Da as shown in Fig. 7. Similar
results have been obtained for metathesis of vegetable oils over Grubbs 2nd generation catalyst where the
difference of 14 Da methylene units was assigned to be resulted from C-C double bonds isomerization
side reactions during self-metathesis. Such isomerization reactions give rise to shorter as well as longer
chain metathesis products due to the formation of metal-hydride intermediate species [53, 54].
If double bonds isomerize, the resulting new olefin can participate in self- and cross-metathesis
reactions, leading to observed distribution with mass difference of 14 Da [26]. Several studies describing
these products formation with methylene group during vegetable oils metathesis have been reported,
especially for ADMET polymerization. In case of internal alkenes several isomerization steps could be
occurred, leading to different terminal alkenes through the carbene formation, which will be involved in
several metallocyclobutane intermediates, providing access to many metathesis products [55, 56].
3.3.3. NMR Analysis
Page 13 of 39

Previous analysis demonstrated that same metathesis products were obtained using either MTO-
1
based catalyst or 2nd generation Grubbs catalyst. For this reason only H and 13C NMR spectra of
MIXTURES 1 to 5 that have been issued from triolein self-metathesis over MTO-based catalyst will be
displayed and discussed (Fig. 8 and Fig. 9 respectively).
The products of all MIXTURES analyzed contain the same backbone as they are made up from
repeating units of triolein. Therefore, in all 1H-NMR spectra, the olefinic protons –CH=CH– and methine
proton in glyceryl group –CH2-CH(O)-CH2– displayed signals at δ =5.50–5.26 ppm, while the methylene
protons in the glyceryl group –CH2-CH(O)-CH2– displayed the signal at δ=4.29-4.14 ppm, the signal at δ
=2.42–2.19 ppm were assigned to the α-methylene protons adjacent to carbonyl carbon –C(=O)- CH2-,
where the signal at δ = 2.09-1.79 ppm were assigned to allyl methylene protons =CH-CH2–. In addition,
the signals at δ =1.79-1.47 ppm, δ =1.47–1.12 ppm and δ =1.04-0.51 ppm were assigned to the β-
methylene protons from carbonyl carbon –C(=O)-CH2CH2–, methylene protons on saturated carbon
atoms –CH2–, and to the terminal methyl protons –CH3, respectively. Meanwhile, the peak at 7.35-7.18
ppm is assigned to the proton exchange with CDCl3 analysis solvent [57-59]. Little difference in the
relative area of the displayed signals was detected, depending on the number of repeating units.
1
The differences detected in H-NMR between linear oligomers and cyclic compounds of the
mixtures are mainly at the protons of the glyceryl group and the double bonds. The ratios of the α-
methylene protons adjacent to carbonyl on the methylene protons in the glyceryl group (–O-C(=O)-CH2–
:CH2-CH(O)-CH2–) and the ratio of the terminal methyl protons –CH3 on the methylene protons in the
glyceryl group were used to identify the structures of the pure compounds as listed in Table 4.
Similarly, in all 13C-NMR spectra of all MIXTURES analyzed, the carbonyl carbon -COO-
showed peak at δ ≈ 173 ppm, the olefin -CH=CH- at δ ≈ 131-129 ppm, the glyceryl group -CH2-CH(O)-
CH2- at δ ≈ 70-61 ppm, the α-methylene adjacent to carbonyl carbon –C(=O)-CH2- at δ ≈ 47-34 ppm,
allyl methylene =CH-CH2– at δ ≈ 32 ppm, the β-methylene from carbonyl carbon –C(=O)-CH2CH2– at δ
≈ 30 ppm, the saturated carbon atoms –CH2– at δ ≈ 23-27 ppm, and the terminal methyl –CH3 at δ ≈ 14
ppm, while the peak at circa 77 ppm is assigned to CDCl3 analysis solvent [59, 60]. All these peaks are
present in all MIXTURES with small differences observed in their intensities. The main structural
differences between the different compounds as exhibited in the 13C-NMR spectra are chemical shifts of
double bonds and the glyceryl group.
The 1H-NMR signals of MIXTURE 1 fit with those of triolein (100% cis) and/or trielaidin (100%
trans) [61] (Fig. 8). The ratios of the α-methylene protons adjacent to carbonyl on the methylene protons
in the glyceryl group (–O-C(=O)-CH2–:–CH2-CH(O)-CH2–) and that of terminal methyl protons –CH3 on
the methylene protons in the glyceryl group are of 1:1.43 and 1:2.61 respectively. These values are fairly
comparable to the expected ratios based on theoretical predictions for triolein and trielaidin structure 1:1.5
and 1:2.25 (Table 4).
Page 14 of 39

13
In C-NMR spectra of MIXTURE 1 described in Fig. 9, the olefin group -CH=CH- displayed a
signal at δ ≈ 131-129 ppm, and the chemical shift at δ < 130 ppm is assigned to cis-configuration while δ
> 130 ppm is assigned to trans-configuration which is a major one. The obtained results are consistent
with those obtained by ESI-MS analysis, and we can conclude that component 1 in MIXTURE 1 has
mainly a trielaidin structure. However, traces of triolein could also be present.
The 1H and 13C NMR spectra of MIXTURE 2 are similar to those reported and discussed by Li et
al. [52] where bis (1,3-bis(octadec-9-enoyloxy)propan-2-yl)octadec-9-enedioate and its isomers were
identified as metathesis products (Fig. 8). This product is a linear dimer resulted from two triolein
molecules metathesis reaction with elimination of octadecene as confirmed by ESI-MS analysis (denoted
Component 2). Moreover, the proton ratios are 1:1.48 and 1:1.44 which could reasonably correspond to
the theoretical ratios 1:1.50 and 1:1.50 characterizing a triolein dimer (Table 3). Indeed, in 13C-NMR
spectra of MIXTURE 2 described in Fig. 9, the glyceryl group with methine carbon -CH2-CH(O)-CH2- at
δ ≈ 69 ppm and methylene carbon -CH2-CHOR-CH2- at δ ≈ 62 ppm defines the linear triolein dimer [26].
The mass spectrometric results revealed the presence of two others molecules with component 2
1
in the MIXTURE 3. Difference in peaks intensities between H-NMR spectra of MIXTURES 2 and 3
confirm the presence of others compounds containing the same functional groups (Figure 8). There is a
linear trimer formed by the reaction of 3 triolein molecules and elimination of 2 octadecene molecules;
the 18-bis (1, 3-bis (octadec-9-enoyloxy) propan-2-yl)1,1-3-(octadec-9-enoyloxy)propane-1,2-
diyldioctadec-9-enedioate and its isomers representing the component 3. This compound structure was
elucidated and discussed by Li et al. [26], which was obtained via triolein metathesis in the presence of
2nd generation Grubbs catalyst, further confirmation was reported recently by Tian and Larock [52] during
ADMET polymerization of Soybean oil. The third molecule could be a dimer of triolein in which two of
the three oleate chains of one triolein fragment has undergone an intramolecular metathesis cyclization. Li
et al. [26] also isolated and identified this dimer by NMR analysis as 1-(2,3-bis(octadec-9-
enoyloxy)propyl)18-(5,22-dioxo-1,4-dioxacyclodocos-13-en-2-yl)methyloctadec-9-enedioate and/or 1-
(1,3-bis(octadec-9-enoyloxy)propan-2-yl)18-(6,23-dioxo-1,5-dioxacyclotricos-14-en-3-yl)octadec-9-
enedioate and its isomers (denoted component 4).
It is important to notice that the cyclic component 4 is a mixture of two positional isomers, one
with sn-1 and sn-3 ring structure (Component 4.a) so-called symmetric compound, and the other with a
sn-1 and sn-2 rings (Component 4.b) so-called asymmetric compound. The chemical environment of the
cyclic compound is more complex than linear oligomer environment, as illustrated by the differences in
1H-NMR spectra of MIXTURES 2 and 3 in Fig. 8. Particularly, the methine proton in glyceryl group –
CH2-CH(O)-CH2– signal at δ ≈ 5.2 ppm shifts to δ ≈ 5.4 ppm, and it is partially overlapped by the
olefinic protons –CH=CH– in the cyclic compound found in MIXTURE 3.
Page 15 of 39

1
MIXTURES 3, 4 and 5 exhibited identical signals for both H-NMR and 13C-NMR spectra,
showing the same composition, confirming the ESI-MS. For these MIXTURES, the protons ratios are not
determined because the compounds need to be isolated separately for an accurate integration of the
signals. In the case of the fractions separated from reaction mixture carried out with Grubbs catalyst, the
MIXTURE 2 contains the component 2 as underlined by the protons ratio values for the α-methylene
protons adjacent to carbonyl on the methylene protons in the glyceryl group (–O-C(=O)-CH2–:–CH2-
CH(O)-CH2–), and the ratio of the terminal methyl protons –CH3 on the methylene protons in the glyceryl
group, of 1:1.50 and 1:1.72, respectively which are close to the theoretical expected values of 1:1.50 and
1:1.50 (Table 4). The NMR spectra of MIXTURE 3 obtained with Grubbs catalyst are similar to the
MIXTURES 4 and 5 spectra obtained with MTO-based catalyst, where a little difference in peaks
intensities were observed.
3.3.4. Kinetic aspects of triolein and metathesis products
Metathesis of esters of glycerol, i.e. unsaturated fatty oils, proceeds via intra- as well as
intermolecular manners, while the latter strongly predominates [19]. Thus, glyceryl trioleate, yields
octadec-9-ene and polymeric glycerides (principally acyclic dimers and trimers and their cyclic
derivatives as demonstrated above).
In order to capture kinetically a representative picture of the triolein self-metathesis pathway, it is
worth to track down the evolution of triolein consumption as well as the formation of both solid and
liquid metathesis products during the investigated reaction time. It is worth pointing out that, the study of
the kinetic profile of triolein self-metathesis was hardly or not at all addressed in literature. To the best of
our knowledge, for the first time we are describing the kinetic evolution of triolein self-metathesis over
both homogeneous Grubbs catalyst and heterogeneous MTO-based catalyst.
Fig. 10 shows the time course profiles of triolein conversion and self-metathesis products
formation in the presence of 3% MTO/ZnCl2-Al2O3-meso heterogeneous catalyst. It can be observed that
the triolein consumption exhibits a fast kinetic period in the first 60 min, followed by a slower
consumption of triolein, and finally stabilized within 6h, reaching ≈ 75% of triolein conversion.
The trielaidin resulted from non-productive metathesis increased rapidly achieving 40 % yield
after 1h. Afterwards, it begins to undergo metathesis reaction. Indeed, it shows a moderate decreases
indicating its conversion into metathesis products, already initiated by the triolein metathesis itself.
When the triolein metathesis was carried out over the homogeneous 2nd generation Grubbs
catalysts, the obtained time course profiles for the products and triolein depletion were significantly
different as depicted in Fig. 11. In fact, the kinetics of triolein depletion through metathesis was faster, as
it is clearly observed that the reaction goes to completion within 5 min, where the catalyst has been found
Page 16 of 39

to be highly active for triolein metathesis. Unfortunately, the formation of the undesirable solid product
reached more than 40% yield within only 5 min, while 50 % yield was achieved after 60 min. The high
production of undesirable solid product is an important drawback, since it represents more than the half of
the isolated products, including desired metathesis products.
This phenomenon was also reported by Biermann et al. [62] when they carried out metathesis of
triolein as model-triacylglycerol of high oleic sunflower oil over 2nd generation Grubbs catalyst without
adding a chain polymerization stopper, methyl acrylate (MA). At least, the use of MA as additive was
required to control the formation of the polymeric solid product. Indeed, by using a high excess of MA,
the polymerization of triacylglycerol was completely suppressed. This was also observed by Tian and
Larock [51] when they investigated the ADMET metathesis of soybean oil giving several polymeric
materials such as sticky oils and rubbers-like products, while triolein metathesis afforded primarily
monocyclic oligomers as main products, which subsequently cross-linked into polymeric product [51].
Advantageously, in the heterogeneous MTO-based catalyst approach, no traces of cross
polymerization solid products were detected, and no addition of polymerization chain stopper was
required.
Despite the fact that 2nd generation Grubbs catalyst is more active, it has been found that it is less
selective towards the desired metathesis products, compared to the MTO-based heterogeneous catalyst.
4. Conclusions
Methyltrioxorhenium incorporated on the ZnCl2-modified mesoporous alumina (3% MTO/ZnCl2-
Al2O3 -meso with Al/Zn=8) was found to be a highly chemoselective heterogeneous catalyst for the self-
metathesis of triolein at optimized conditions, and found very efficient for metathesis of bulky
functionalized olefins such as triolein.
The MTO-based heterogeneous catalyst afforded good conversion up to 75% within 5h while the
desired metathesis products were isolated in ≈ 25-30% yield.
The undesired solid polymeric product formed by cross-linking polymerization during the triolein
self-metathesis over homogeneous Grubbs 2nd generation catalyst was advantageously avoided in the
presence of MTO-based catalyst.
Triolein self-metathesis gave rise selectively to the primary metathesis products; dimer, trimer and
their cyclic derivatives resulted from intramolecular metathesis pathway occurred simultaneously over
MTO/ZnCl2-Al2O3-meso, in addition to the 9-Octadecene formation, gathering a high industrial interest.
These promising results proved that the 3%MTO/ ZnCl2-Al2O3-meso catalyst can be successfully
applied for other vegetable oils metathesis reaction in a highly selective approach thus, simplifying access
to many industrially important products and polymer building blocks.
Page 17 of 39

Acknowledgements
The authors gratefully acknowledge the Natural Sciences and Engineering Research Council of
Canada who provided funds for this research. M. Alain Gaudreau and Pierre Audet are acknowledged for
Mass spectrometric and NMR analysis. N. Chorfa is acknowledged for the helpful NMR analysis
interpretations.
Page 18 of 39

References
[1] S.Z. Erhan, M.O. Bagby, J. Am. Oil Chem. Soc. 71 (1994) 1223-1226.
[2] J.K.C. Ma, R. Chikwamba, P. Sparrow, R. Fischer, R. Mahoney, R.M. Twyman, Trends Plant Sci.
10 (2005) 580-585.
[3] K.E. Perepelkin, Fibre Chem. 37 (2005) 417-430.
[4] J.E.G. van Dam, B. de Klerk-Engels, P.C. Struik, R. Rabbinge, Ind. Crops Prod. 21 (2005) 129-144.
[5] S. Pinzi, I.L. Garcia, F.J. Lopez-Gimenez, M.D. Luque de Castro, G. Dorado, M.P. Dorado, Energy
& Fuels 23 (2009) 2325-2341.
[6] M. Mosiewicki, M.I. Aranguren, J. Borrajo, J. Appl. Polymer Sci. 97 (2005) 825-836.
[7] H. Baumann, M. Bühler, H. Fochem, F. Hirsinger, H. Zoebelein, J. Falbe, Angew. Chem. Int. Ed.
27 (1988) 41-62.
[8] F. Gunstone, INFORM 11 (2000) 396-414.
[9] F.D. Gunstone, Eur. J. Lipid Sci. Technol. 103 (2001) 307-314.
[10] K.J. Ivin, J.C. Mol, Olefin Metathesis and Metathesis Polymerization , 2nd Edition, Academic
Press, London, 1997 [496] p.
[11] J.C. Mol, Green Chem. 4 (2002) 5-13.
[12] H. Ngo, K. Jones, T. Foglia, J. Amer. Oil Chem. Soc. 83 (2006) 629-634.
[13] J.C. Mol, J. Molec. Catal. 65 (1991) 145-162.
[14] J.C. Mol, Chemtech 13 (1983) 250-255.
[15] E. Verkuijlen, C. Boelhouwer, Fette, Seifen, Anstrichmittel 78 (1976) 444-447.
[16] C.P. Nicolaides, J.H. Opperman, M.S. Scurrell, W.W. Focke, J. Am. Oil Chem. Soc. 67 (1990) 362-
363.
[17] T.J. Katz, Angew. Chem. Int. Ed. 44 (2005) 3010-3019.
[18] K.C. Nicolaou, P.G. Bulger, D. Sarlah, Angew. Chem. Int. Ed. 44 (2005) 4490-4527.
[19] E. Verkuijlen, R.J. Dirks, C. Boelhouwer, J. Roy. Neth. Chem. Soc. 96 (1977) M86-M90.
[20] R.H.A. Bosma, G.C.N. Van Den Aardweg, J.C. Mol, J. Organometallic Chem. 255 (1983) 159-171.
[21] S. Warwel, A. Deckers, Fett Wissenschaft Technol. 91 (1989) 430.
[22] S. Warwel, G. Pütz, J. Organometallic Chem. 364 (1989) 323-329.
[23] A. Rybak, M.A.R. Meier, Green Chem. 9 (2007) 1356-1361.
[24] P.B. Van Dam, M.C. Mittelmeijer, C. Boelhouwer, J. Am. Oil Chem. Soc. 51 (1974) 389-392.
[25] N.A.M. Nordin, B.M. Yamin, M.A. Yarmo, K. Pardan, A.B. Alimuniar, J. Molec. Catal. 65 (1991)
163-172.
[26] M.D. Refvik, R.C. Larock, J. Am. Oil Chem. Soc. 76 (1999) 99-102.
Page 19 of 39

[27] N.A. Buang, N.A.M. Nordin, R.A. Kamarudin, A. Muda, J. Molec. Catal. A: Chem. 115 (1997) 95-
99.
[28] R.F.A. Teixeira, X.K.D. Hillewaere, S. Billiet, F.E. Du Prez, Chemistry of Crosslinking Processes
for Self-Healing Polymers, Self-Healing Polymers: From Principles to Applications, Wiley-VCH
Verlag GmbH & Co. KGaA2013, pp. 215-246.
[29] J. Patel, S. Mujcinovic, W.R. Jackson, A.J. Robinson, A.K. Serelis, C. Such, Green Chem. 8 (2006)
450-454.
[30] T.M. Trnka, R.H. Grubbs, Acc. Chem. Res. 34 (2001) 18-29.
[31] J.C. Mol, R. Buffon, J. Brazilian Chem. Soc. 9 (1998) 1-11.
[32] D. Mandelli, M.D.M. Jannini, R. Buffon, U. Schuchardt, J. Am. Oil Chem. Soc. 73 (1996) 229-
232.
[33] R. Buffon, U. Schuchardt, A. Abras, J. Chem. Soc. Faraday Trans. 91 (1995) 3511-3517.
[34] S.K. Pillai, S. Hamoudi, K. Belkacemi, Fuel 110 (2013) 32-39.
[35] S.K. Pillai, S. Hamoudi, K. Belkacemi, Appl. Catal. A: Gen. 455 (2013) 155-163.
[36] J.C. Ray, K.-S. You, J.-W. Ahn, W.-S. Ahn, Microporous Mesoporous Mater. 100 (2007) 183-190.
[37] T. Oikawa, Y. Masui, T. Tanaka, Y. Chujo, M. Onaka, J. Organometallic Chem. 692 (2007) 554-
561.
[38] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373-380.
[39] S.A. Bagshaw, T.J. Pinnavaia, Angew. Chem. Int. Ed. 35 (1996) 1102-1105.
[40] P. Jean-Louis Hérisson, Y. Chauvin, Die Makromolekulare Chem. 141 (1970) 161-176.
[41] Y. Chauvin, Angew. Chem. Int. Ed. 45 (2006) 3740-3747.
[42] C. Boelhouwer, J.C. Mol, J. Am. Oil Chem. Soc. 61 (1984) 425-430.
[43] F.C. Courchay, J.C. Sworen, K.B. Wagener, Macromolecules 36 (2003) 8231-8239.
[44] P.A. Fokou, M.A.R. Meier, J. Am. Chem. Soc. 131 (2009) 1664-1665.
[45] J.L. Bilhou, J.M. Basset, R. Mutin, W.F. Graydon, J. Chem. Soc., Chem. Commun. (1976) 970-971.
[46] J.M. Basset, G. Coudurier, R. Mutin, H. Praliaud, J. Catal. 34 (1974) 152-155.
[47] N. Calderon, E.A. Ofstead, J.P. Ward, W.A. Judy, K.W. Scott, J. Am. Chem. Soc. 90 (1968) 4133-
4140.
[48] L. Porri, P. Diversi, A. Lucherini, R. Rossi, Die Makromolekulare Chem. 176 (1975) 3121-3125.
[49] C.S. Consorti, G.L.P. Aydos, J. Dupont, Chem. Commun. 46 (2010) 9058-9060.
[50] D.M. Ohlmann, N. Tschauder, J.-P. Stockis, K. Gooßen, M. Dierker, L.J. Gooßen, J. Am. Chem.
Soc. 134 (2012) 13716-13729.
[51] Q. Tian, R. Larock, J. Am. Oil Chem. Soc. 79 (2002) 479-488.
[52] S. Li, L. Hojabri, S. Narine, J. Am. Oil Chem. Soc. 89 (2012) 2077-2089.
[53] U. Biermann, M.A.R. Meier, W. Butte, J.O. Metzger, Eur. J. Lipid Sci. Technol. 113 (2011) 39-45.
Page 20 of 39

[54] P.A. Fokou, M.A.R. Meier, Macromolecular Rapid Commun. 29 (2008) 1620-1625.
[55] S.E. Lehman Jr, J.E. Schwendeman, P.M. O'Donnell, K.B. Wagener, Inorg. Chimica Acta 345
(2003) 190-198.
[56] V.I. Petkovska, T.E. Hopkins, D.H. Powell, K.B. Wagener, Anal. Chem. 78 (2006) 3624-3631.
[57] U.N. Wanasundara, F. Shahidi, J. Food Lipids 1 (1993) 15-24.
[58] Y. Miyake, K. Yokomizo, N. Matsuzaki, J. Am. Oil Chem. Soc. 75 (1998) 15-19.
[59] M.a.D. Guillén, A. Ruiz, Trends in Food Sci. Technol. 12 (2001) 328-338.
[60] J.A. Hamilton, D.M. Small, Proc. National Academy Sci. 78 (1981) 6878-6882.
[61] L. Retief, J.M. McKenzie, K.R. Koch, Magnetic Resonance Chem. 47 (2009) 771-781.
[62] U. Biermann, J.O. Metzger, M.A.R. Meier, Macromolecular Chem. Phys. 211 (2010) 854-862.
Page 21 of 39

Figure Captions
Figure 1
Figure 2
: The small angle XRD patterns of the mesoporous alumina support, ZnCl2-modified
mesoporous alumina as well as 3wt% MTO/ZnCl2-Al2O3-meso catalyst.
: TEM image of the mesoporous alumina materials: a) mesoporous Al2O3 material; b)
ZnCl2-modified mesoporous alumina.
Figure 3
Figure 4
Figure 5
: Solid product formed in the reaction with 2nd generation Grubbs catalyst.
: Structures of Component 1 to 5 identified as metathesis products.
: Mass spectrometric analysis of the metathesis products obtained with MTO-based
catalyst.
Figure 6
Figure 7
Figure 8
: Mass spectrometric analysis of the metathesis products obtained with Grubbs catalyst.
: Magnified view of the peaks obtained in the mass spectrometric analysis.
1
: H-NMR spectra of MIXTURES issued from triolein metathesis over MTO-based
catalyst.
Figure 9
Figure 10
Figure 11
:
13C-NMR spectra of MIXTURES issued from triolein metathesis over MTO-based
catalyst.
: Triolein depletion and products formation time course profiles during self-metathesis of
triolein over MTO-based catalyst.
: Triolein depletion and products formation time course profiles during self-metathesis of
triolein over 2nd generation Grubbs catalyst.
Page 22 of 39

Table 1. Textural properties of Al2O3-meso and 3%MTO/ZnCl2-modified Al2O3-meso.
Materials
Al2O3-meso
3%MTO/ZnCl2-modified Al2O3-
meso
BET surface area (m2/g)
Pore size (nm)
Total pore volume (cm3/g)
490 ± 1
8.0 ± 0.1
1.2 ± 0.1
418 ± 1
7.6 ± 0.1
1.0 ± 0.1
Page 23 of 39

Table 2 Comparison of catalytic activity of MTO/ZnCl2-Al2O3 and Grubbs 2nd generation
catalysts for triolein self-metathesisa.
Catalyst
Triolein
conversi
on (%)
Trielaidin
yield
9-Octadecene
yield
Soluble
metathesis
product b
yield (%)
Solid product
(%)
(%)
(%)
MTO/
ZnCl2-Al2O3
65.01
95.24
40.19
2.80
3.98
22.20
20.09
0.00
2nd generation
20.25
50.38
Grubbs catalyst
a
o
b
Reaction conditions: 45 C, 60 min, 400 rpm; Metathesis products are products detected in
MIXTURES 2 to 5 excluding the remained triolein presents in the fractions and trielaidin
obtained as a non-productive metathesis product.
Page 24 of 39

Table 3 Yields and mass spectrometric detection by ESI-MS of molecular weight of ions for different products contained
in the fractions obtained after TLC analysis for triolein self-metathesis over both catalysts.
Catalysts
Fractions
MTO/ZnCl2-Al2O3 catalyst a
2nd generation Grubbs catalyst b
+ d
+
d
Components
Yields (%)c
Components
Yields (%)c
[M w + NH4
]
[
M w + NH 4
]
MIXTURE 1
MIXTURE 2
MIXTURE 3
Component
Component
1
4.03
9.50
902.8
Component
Component
1
0.40
902.8
2
1536.3
2
4.42
1535.4
Component
Component
Component
Component
Component
Component
Component
Component
Component
Total
2
3
4
2
3
4
2
3
4
5.57
0.99
0.14
3.51
1.35
0.18
0.76
0.03
0.25
26.33
1536.3
2168.8
1283.1
1536.3
2168.8
1283.1
1536.3
2168.8
1283.1
Component
Component
Component
Component
Component
4
5
2
4
5
2.34
0.26
0.12
3.73
0.21
1536.3
2168.8
1283.1
1664.3
2183.1
MIXTURE 4
MIXTURE 5
Component
Component
4
5
8.18
1.02
1663.3
1283.1
Total
20.49
a
b
c
d
After 5h of reaction; After 1h of reaction; Yield % = Product amount (mg)/Triolein amount (mg)*100; m/z of
[
M w + NH 4
]
+ ; The product amounts (mg) were determined by the collected mass of the MIXTURES obtained by TLC and
absolute abundances of ESI-MS analysis.
Page 25 of 39

Table 4. Ratio of selected protons identified from the 1H-NMR spectra.
(-CH2-CH(O)-CH2-)
(-CH2-CH(O)-CH2-)
(-CH3)
3% MTO/ZnCl2-Al2O3
MIXTURE 1
1.00
1.00
ND b
ND
1.43
1.48
ND
2.61
1.44
ND
MIXTURE 2
MIXTURE 3 b
MIXTURE 4 b
ND
ND
2nd generation Grubbs catalyst
MIXTURE 2
MIXTURE 3 b
MIXTURE 4 b
MIXTURE 5 b
1.00
ND
ND
ND
1.50
ND
ND
ND
1.72
ND
ND
ND
Expected values a
Component
Component
Component
Component
Component
1
2
3
4
5
1.00
1.00
1.00
1.00
1.00
1.50
1.50
1.50
1.50
1.50
2.25
1.50
1.12
0.75
0.25
a Expected values based on the theoretical prediction for the products as determined by Li et al.21b;
b
ND signifies Not Determined. Owing to the impure nature of the MIXTURES 3, 4, and 5, it
was not possible to determine the ratios of selected protons.
Page 26 of 39

Figure 1
Page 27 of 39

a)
Interconnected
porespores
b)
Unclear pores
Figure 2
Page 28 of 39

Figure 3
29
Page 29 of 39