J.A. Avenda n˜ o Villarreal et al.
Molecular Catalysis 509 (2021) 111631
column (30 m length, 250
μ
m internal diameter) and a flame ionization
1-propenylbenzene to optimize the reaction conditions. In the ethenol-
◦
detector (FID). GC analysis conditions: injection port, 310 C; split ratio
ysis of 1 (Scheme 2), two main products were observed: 4-vinylanisole
◦
◦
◦
′
5
0:1; FID, 320 C; oven temperature 50 C for 5 min, heating at 15 C/
(2) and 4,4 -dimethoxystyrene (3). The isomeric estragole was
◦
min to 280 C, kept for 5 min. Carrier gas: H
2
(UHP grade), make-up gas
observed in minor amounts and the formation of co-products propene
and 2-butene could be inferred, although they were not quantified due
to analytical limitations. In Table 1, a catalyst screening under various
reaction conditions is presented. The initial experiments were per-
formed with a catalyst loading of 1 mol% with respect to anethole in
2
N (UHP grade). Conversions and yields in catalytic runs were deter-
mined having undecane as internal standard. The GC/MS analyses were
performed using a Shimadzu GC2010/QP2010-GC/MS instrument
employing an electron impact detector at 70 eV. The GC/MS conditions
were identical to that of the GC/FID setup, except for the carrier gas
◦
THF, 60 C and 10 bar of ethylene.
(
helium). The products were purified by passing the reaction solution
The catalysts bearing tricyclohexylphosphine as ligand (GI, GII) gave
low conversion for ethenolysis of 1 (Table 1, Entries 1,2). GI is a useful
catalyst for, e.g., methyl oleate ethenolysis, but presents a very low
activity for anethole under the reaction conditions employed. Probably,
trough a small column containing 1 g of basic alumina (activated, basic,
Aluminum oxide, Brockmann I, Sigma-Aldrich), then, the solvent was
1
13
evaporated and the product analyzed by H NMRand C NMR.
the trans configuration of its propenylic C
omatic ring conjugation makes this substrate less reactive than methyl
oleate, which has a cis, non-conjugate C C double bond. GII presents a
–
C double bond and the ar-
2
.4. Analytical data
-vinyl anisole (2): Mass spectrometry (m/z/relative intensity): 134/
–
significantly better activity, which is expected as the catalyst containing
the N-heterocyclic carbene (NHC) ligand binds the olefin more effi-
ciently. Nevertheless, even with GII the conversion was not yet satis-
factory as only 30% was achieved after 16 h (entry 2). Conversely, HII
catalyst gave a much better result, reaching 94% yield (entry 3) under
the same reaction conditions. Although both GII and HII generate the
4
+
1
1
00 [M] , 119/60, 91/56, 65/20. H NMR (400 MHz, CDCl
3
): δ 7.37 (d,
= 8.8 Hz, Ar-H), 6.69
3
3
3
2
H,
J
H
–
H
= 8.8 Hz, 2 Ar-H), 6.89 (d, 2H,
J
H H
–
3
(
dd, 1H,
CH ), 5.15 (d, 1H,
C NMR (100 MHz, CDCl
CH=), 130.67 (aromatic carbon), 127.58 (2 aromatic carbons), 114.12
J
H
–
H
= 17.6 Hz, -CH=), 5.63 (d, 1H,
J
H
–
H trans = 18 Hz,
), 3.83 (s, 3H, Ar–OCH ).
): δ 159.6 (aromatic carbon), 136.44 (Ar-
3
=
2
J
H
–
H cis = 10.8 Hz, =CH
2
3
1
3
3
3
same catalytic species, the released PCy in GII can be detrimental for
). The 1H NMR
3
the catalyst activity and stability. Not only this ligand can rebind the
catalytically active species, but also it can attack Ru-methylidene in-
termediates leading to catalyst decomposition paths, as demonstrated by
Fogg and co-workers [36,37]. The analogous M73 is slightly less active,
showing that the leaving alkylidene can also influence on the system
performance. Perhaps more surprising is that M73Pr gives a better yield
for 2 than HII. Steric features of the stabilizing ligand seem to be critical
for a good performance of the catalyst in challenging metathesis re-
(
=CH
2
), 111.75 (2 aromatic carbons), 55.47 (Ar–OCH
1
3
and C NMR signals assignments were confirmed by comparison to the
literature values [34].
2
‑methoxy-4-vinylphenol (5): Mass spectrometry (m/z/relative in-
+
1
tensity): 150/100 [M] , 135/85, 107/40, 91/10, 77/36. H NMR
spectrum (400 MHz, CDCl ): δ 7.28 (m, 1H, 1 Ar-H), 6.98 (m, 2H, Ar-H),
= 14.1 Hz, -CH=), 5.94 (s, 1H, Ar-OH), 5.67 (d, 1H,
H trans = 17.6 Hz, =CH H cis = 10.8 Hz, =CH ),
.91 (s, 3H, Ar–OCH ). C NMR (100 MHz, CDCl ): δ 146.7, 145.7
3
.71 (dd, 1H, 3
6
3
J
H H
–
), 5.20 (d, 1H, 3
J
H
–
2
J
H
–
2
1
3
actions [38,39]. SIPr is a bulkier NHC ligand than H IMes and the
2
3
3
3
former may disfavor the coordination of anethole in an intermediate
(
aromatic carbons), 136.74 (Ar-CH=), 120.11 (aromatic carbon),
1
containing a bezylidene moiety, which is necessary to make the stilbene
derivative 3. Thus, the path of the cross metathesis involving the coor-
dination of the less encumbered ethylene is favored with this bulkier
version of the catalyst [40].
1
14.12 (=CH
2
), 111.75 (2 aromatic carbons), 55.87 (Ar–OCH
3
). The H
1
3
NMR and C NMR signals assignments were confirmed by comparison
to the literature values [10].
1
,2-dimethoxy-4-vinylbenzene (7): Mass spectrometry (m/z/relative
+
1
Considering that THF is nowadays a non-recommended solvent in
recent solvent guides [41], we decided to employ the more acceptable
and now standard toluene as solvent for olefin metathesis in the ethe-
nolysis of 1-propenylbenzenes. We were pleased to find that the
replacement was even advantageous as full conversion and near quan-
titative yield of 2 was achieved using the catalyst M73Pr (Table 1, entry
intensity): 164/100 [M] , 149/45, 121/20, 103/25, 91/37, 77/25.
NMR spectrum (400 MHz, CDCl ): δ 7.42 (m, 2H, 2 Ar-H), 7.18 (m, 1H,
= 14.2 Hz, -CH=), 5.89 (d, 1H,
H
3
Ar-H), 6.91 (dd, 1H, 3
3
J
H
–
H
J
H H trans
–
), 4.15 (s, 3H,
3
=
17.4 Hz, =CH
), 4.12 (s, 3H, Ar–OCH
aromatic carbon), 136.44 (Ar-CH=), 130.67 (aromatic carbon), 127.58
), 111.75 (2 aromatic carbons),
2
), 5.42 (d, 1H,
J
H
–
H cis = 10.8 Hz, =CH
2
). 1 C NMR (100 MHz, CDCl
3
Ar–OCH
3
3
3
): δ 159.6
(
6
). In order to verify whether or not 16 h were really necessary, we made
(
2 aromatic carbons), 114.12 (=CH
2
) The 1H NMR and C NMR signals assignments were
13
the reaction in a shorter reaction time (2 h) and no decrease in yield was
observed (Table 1, entry 7). Once obtained more satisfactory reaction
conditions, we addressed the issue of catalyst loading. It was possible to
reduce 20 times the catalyst loading (from 1 to 0.05 mol%; c.f. entry 7
and 8) without significant loss in yield. Further reduction to 0.01 mol%
still leaded to 87% yield (entry 9), resulting in a turnover number of
5
5.47 (Ar–OCH
3
confirmed by comparison to the literature values [35].
. Results and discussion
3
3
.1. Ethenolysis of anethole (1)
8
700, which is considered appropriate for fine chemicals applications.
Reducing ethylene pressure from 10 to 5 bar did not compromise the
yield (c.f. entry 8 and 10). This pressure reduction has important im-
plications from the technical point of view, as equipment to operate up
to 7 bar are considerably cheaper than for higher pressures. More
For the ethenolysis of 1-propenylbenzenes, we employed THF as
◦
solvent, 60 C, ethylene (10 bar) as initial conditions [10] for the
screening of the ruthenium-based catalysts presented in Chart 1 (GI, GII,
HII, M73, M73Pr). Anethole (1) was used as
a representative
Scheme 1. Ethenolysis of 1-propenylbenzenes to obtain functionalized styrenes.
3