N. Riache et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 344–348
345
Scheme 1. Self-metathesis (SM) of ethyl undecylenate catalyzed by silica supported W-based catalyst.
of E/Z internal olefins with the formation of trans ester thermody-
namically favored (molar ratio, 7:3).
olefin followed by -elimination of the possible metal alkyls).
These elementary steps are important to account for the ISOMET
reaction. Moreover, a dual Ru/W catalysts has been recently in
which the Ru catalyst serves only as the isomerization catalyst
in the ISOMET reaction [16]. As silica supported tungsten species
(1) acts as a multifunctional catalyst, we initially thought that (1)
would be ideal to produce a distribution of unsaturated diesters
of fatty acids with variable chain length from a functionalized
olefin (monounsaturated FAEs in our case), together with a distri-
bution of resulting olefins and monoesters. Indeed, it is known that
unsaturated fatty acid methyl esters (FAMEs) in presence of both
isomerizing and metathesis catalysts represent an elegant access
to a mixture of unsaturated diesters of fatty acids. France et al. dis-
closed one of the first tandem isomerization/metathesis processes
of unsaturated FAMEs by employing a dual catalytic Ir/Ag system
[17]. Later on, Ohlmann et al. employed a Pd/Ru system for this
transformation [18]. Starting from an unsaturated FAME, with an
internal double bond, they observed the formation of a distribu-
tion of olefins, unsaturated diesters and unsaturated monoesters
(resulting from both primary and secondary self-metathesis and
isomerization/cross metathesis reactions) (Scheme 2).
The conversion of ethyl undecylenate versus time was carried
◦
out at 150 C. The plots of TONs and conversion versus time are given
in Fig. 1. A final conversion of 12% is reached with 26 TONs. A plateau
is observed after 3 h which would correspond to a thermodynamic
equilibrium. The reaction is very fast reaching 8% conversion and 17
TONs at t = 15 min. The observable equilibrium must result from the
degenerative metathesis process and the accumulation of ethylene
in this closed system. The reaction conversion could further be opti-
mized by varying the concentration of the precatalyst. The study on
the concentration of the substrate is illustrated in Fig. S3. Varying
the W/substrate ratio (from 2.3% to 9.3%) gave a higher conversion
of 19% with TONs of 33 (see Fig. S3).
Despite the need of heating, the level of temperature of the
reaction was found to have a minor impact on the efficiency of
◦
◦
the catalyst. Indeed, for a given period of 18 h at 80 C, 150 C or
◦
2
00 C, the maximum unsaturated FAE conversion based on diester
◦
products reached respectively 14%, 12% and 11%. At 200 C, traces
amount of isomerized ethyl undecylenate were detected by GC–MS.
Employing Ru based catalyst, Grubbs first generation complex,
under similar reaction conditions, afforded self-metathesis prod-
ucts and two newly formed isomers of ethyl undecylenate with 28%
conversion towards expected diesters after 24 h reaction time (see
Fig. S4). Yet, our turnovers are comparable with the ones obtained
with the supported catalytic system WCl /Me Sn [6]. This suggests
With silica supported [( Si O)WMe5] (1) such diesters distri-
bution could not be observed. Interestingly, we also found that the
reaction of saturated FAE (as the ethyl undecylate) with 1 did not
◦
produce alkane metathesis products at 150 C for 3 days period.
This result can be rationalized by considering the possible coordi-
nation of a W-methylidene hydride, propagative species in alkane
metathesis [12], into the functional ester group. And, this would
prevent two crucial steps in the alkane metathesis: the C H alkane
activation required for the olefin intermediate and the release of
catalytic amount of hydrogen. These observations are consistent
with the absence of isomerization using unsaturated FAEs as sub-
strates, pointing out a detrimental effect from the ester group by
preventing the isomerization step in the olefin metathesis process.
6
4
that with the latter system, the in situ alkylation of WCl6 should
initially occur to form the desired propagative W carbene species.
To account for the low activity for this silica supported W sys-
tem, we assume that the geometry of the W metallocyclobutanes
plays an important role. Indeed, extensive DFT studies on metal-
lacyclobutanes with 5-coordinated W metal center have shown
that a square pyramidal (SP) geometry is likely to be responsible
for the deactivation of the catalyst [14,15]. While in the case for
the Grubbs Ru catalyst, calculations have shown that the trigonal
bipyramidal metallacyclobutane, (required for the cycloreversion
metathesis step) is generally much more stable than SP metalla-
cyclobutanes. Following the same catalytic procedure described
for ethyl undecylenate, we found that (1) catalyzes also FAEs with
internal double bond. For instance, ethyl oleate was converted to
the corresponding primary metathesis products: the symmetri-
cal unsaturated diesters and 9-octadecene (mixture of substituted
2.2. Cross metathesis study between ˛-olefin and an unsaturated
fatty acid ester
As species (1) is active in ␣-olefins ISOMET reaction but led
only to the self-metathesis products with unsaturated FAEs, we
investigated the reactivity of (1) in presence of both unsaturated
FAE and ␣-olefins. We examined the cross reaction of 1-decene
and ethyl undecylenate. The cross-metathesis involving a func-
tionalized and an unfunctionalized olefin gave only three types of
primary products: (i) self-metathesis product of 1-decene metathe-
sis, (ii) self-metathesis product of ethyl undecylenate metathesis
and (iii) cross metathesis products (with ethylene formation in
each metathesis reaction) as shown in Scheme 3. The reaction of
◦
symmetrical olefins, ratio of Z/E, 34:66) at 150 C for 3 h (see Fig.
S5).
Surprisingly, this reaction catalyzed by silica-supported pre-
cursor (1) is highly selective leading only to the primary
self-metathesis products of ethyl undecylenate or ethyl oleate
without apparent isomerization of the starting unsaturated FAE.
These observations strongly suggest that the ester group plays a
role in the inhibition of the isomerization process. In fact as men-
tioned earlier under the same reaction conditions, using solely
1
-decene (C10) with ethyl undecylenate (C11) (molar ratio of 1:1)
catalyzed by (1) produced the internal olefins, 9-octadecene (C18),
the higher molecular weight FAE (C19) and the symmetrical diester
of fatty acid (C20). By comparison with the linear ␣-olefins metathe-
sis, the conversions observed for the unsaturated FAE metathesis
or for the metathesis with a 1:1 molar ratio of ␣-olefin/unsaturated
FAE metathesis were found to be low. Metathesis of 1-decene and
ethyl undecylenate was also studied at different molar ratios of
ethyl undecylenate/1-decene; the results are depicted in Fig. 2 (see
1
-decene as a substrate gave a distribution of terminal and internal
olefins corresponding to the self-metathesis and secondary cross-
metathesis products (ISOMET) (see Fig. S6).
We previously proposed a mechanism involving in the metathe-
sis of ␣-olefins by 1, in which the isomerization steps take place
through the in situ formation of a W–H species (insertion of the