Angewandte
Communications
Chemie
varying degrees of success.[11] A key criteria in reactor design
is to enable efficient mass transfer thus allowing the removal
of ethylene gas, this is especially key at larger scales.
Typically, gas-to-liquid-phase mass transfer in batch
vessels become less efficient as scale increases owing to
diminishing returns on surface area to volume ratios. Indeed,
in 2014 Skowerski and co-workers demonstrated efficient
mass transfer removal of ethylene gas with the application of
a vacuum across a tube-in-tube permeable membrane reac-
tor.[11k] Recent studies from Jamison, Bio and co-workers
elegantly demonstrated an alternative perevaporation
approach (“blowing” a stream of N2 gas across a permeable
membrane to carry the ethylene gas out of the liquid phase) as
an effective strategy for the mass transfer removal of ethylene
gas.[11o] Despite these key developments in reactor design,
a general approach to continuous flow Z-selective metathesis
remains elusive. In order to realise an effective continuous
process, careful attention must be given to both the catalyst
choice and the continuous reactor design (Figure 1b) but
success in this area paves the way to the continuous
preparation of a variety of pheromone and odorant mole-
cules.
We initiated our study by investigating the application of
a Teflon AF-2400 tube-in-tube semi-permeable membrane
reactor[12] to the Z-selective process, preliminary results
highlighted that application of a “vacuum-on” across the
membrane for ethylene removal, versus “vacuum-off” deliv-
ered a clear benefit in terms of yield (See Supplementary
Information (SI); Table S4 entry 5). Taking this further, the
self-metathesis (SM) of ethyl 9-decenoate under continuous
flow condition using a 1 mL Teflon AF-2400 tube-in-tube
reactor was explored in more detail against a focused
collection of catalysts Ru-1, Ru-4 and Ru-5 (Scheme 1. For
the gram-scale synthesis of Ru-4, -5 and X-ray character-
isation of Ru-5,[13] see SI, Scheme S1 and Figure S5). Despite
an excellent 95/5 Z/E ratio observed with Ru-1 within 0.5 h,
the selectivity dropped gradually as the conversion increased
reaching 78% after 3 h of residence time (Scheme 1a).[14]
Interestingly, 2,6-diisopropylphenyl (DIPP)-containing cyclo-
metalated Ru-4 demonstrated excellent catalytic perfor-
mance in the flow reactor affording the desired internal
olefin with 78% conversion and very high 97% Z-selectivity
after 3 hours of residence time (Scheme 1b).
Scheme 1. Catalytic performances of cyclometalated Ru-complexes Ru-
1, -4, and -5 in continuous flow self-metathesis of ethyl 9-decenoate 5.
hour residence time and 1 mol% catalyst loading led to
moderate to good conversions and yields. Notably, all CM
products were formed in excellent Z-selectivity, ranging from
94 to 98.5%, with the exception of allylic alcohol 11 which
afforded a Z/E ratio of 80/20[17] Furthermore, a 0.5 mol%,
Ru-4 loading was sufficient to promote the self-metathesis of
allyl-benzene and other unfunctionalized linear terminal
alkenes furnishing, after 2 hours, the corresponding internal
Z-olefins 12–14 in excellent selectivity (up to 98%) and
moderate to good isolated yields. Using the designed flow
reactor rig, highly valuable semiochemicals 15–20,[18] acting as
potential bio-pesticides against Lepidoptera (moth) and
Tenebrio Molitor (beetle), were efficiently produced with
excellent Z-selectivity (96–98.5%). Noticeable, a similar
efficiency was observed at 6 mmol scale (12 times the
standard substrate scope scale) with a slight alteration of Z-
selectivity (96–97%).
The novel Ru-5 catalyst, which is a structural link between
Ru-1 and Ru-4, deserved to be examined. Unfortunately,
a lower range of Z-selectivity over time (87 to 69%) was
observed, although the resulting diester 6 was produced in
a higher yield (84%, Scheme 1c).[14,15] Consequently, the
DIPP group appears to be the key structural moiety to deliver
the highest selectivity. It is worth underlining that continuous
flow metathesis can be conducted outside a glovebox while
batch conditions require an open vessel inside the glovebox to
efficiently remove the ethylene and reach high conversions.[16]
Having identified the combination of Ru-4 and a Teflon
AF-2400 vacuum-on tube-in-tube design as the most efficient
combination to achieve continuous Z-selective catalysis,
a range of several cross- and self-metathesis transformations
were explored in a larger 3 mL reactor (Scheme 2). Initially
running a range of substrates through the reactor with a two-
We next turned our attention to the macro ring-closing
metathesis (RCM) reaction of terminal olefins. Typically,
macro-RCM requires higher dilution than CM so as to
minimise the competitive oligomerization reaction; reaction
concentration therefore becomes a variable.[14] As depicted in
Table 1 entries 1 and 2, similar catalytic performances were
observed with Ru-1 and Ru-5 in the formation of the 16-
membered macrocycle 22 when the reaction was run at
20 mM[19] in 1,2-dichloroethane (708C, 3 hours residence
time, 70% and 75% isolated yield, respectively). Never-
theless, the Z-selectivity still remained moderate reaching
86% and 82% respectively. To our delight, Ru-4 showed an
impressive 97/3 Z/E ratio although a significantly lower yield
was observed (24% isolated yield, Table 1, entry 3).[20] By
increasing the residence time to 4.5 h (entry 4), the yield
could be slightly improved without any alteration of Z-
2
ꢁ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 1 – 7
These are not the final page numbers!