Organic Letters
Letter
Scheme 2. (A) Design Plan. (B) Optimization of the Structure of the Template. (C) Optimization of the Reaction Conditions
and Sensitivity Tests
thermodynamically favored E-isomer into the contra-thermo-
dynamic Z-isomer using an EnT manifold under simple
reaction conditions. In addition, the use of copper salt in
photochemistry has a long history, particularly in [2 + 2]
cycloaddition under UV irradiation.9 Surprisingly, copper has
been scarcely used in alkene isomerization.10 In that respect,
we should mention the work from Deyrup, who studied the
Cu-photocatalyzed cis → trans isomerization of cyclooctene
using CuCl2, albeit with a moderate efficiency (Scheme 1, eq
4).10b As part of our program to develop an efficient copper-
catalyzed and photocatalyzed process,11 we sought to develop
a practical method for the E → Z isomerization of polarized
alkenes under visible-light irradiation. In the early 1980s, Lewis
brought an alternative concept to permit the E → Z
isomerization of polarized alkenes using a Lewis acid catalyst
(Scheme 1, eq 5).12 The addition of boron- or aluminum-
based Lewis acid favored the contra-thermodynamic con-
version of E-cinnamates into the Z-isomer under UV
irradiation (λ = ca. 300 nm). This facilitated interconversion
results from a change of the spectroscopic properties and
therefore of the photochemical behavior of the substrate.
With the aim of providing a straightforward protocol to
perform E → Z isomerization, we conjectured that the use of a
copper salt coordinated with an appropriate ligand, in a
catalytic amount, should allow this endergonic process under
visible-light irradiation by forming a transient chromophore
species (Scheme 1, eq 6).13 The coordination, by changing the
HOMO and LUMO levels of the substrate, will induce a
bathochromic shift in the absorption (π → π* transition) of
the complexed substrate and therefore a lowest energy
transition from the ground state (S0) to the existing state
(S1). Importantly, apart the compulsion to proceed under
visible-light irradiation, the success of the approach has several
prerequisites. (1) The E → Z quantum yield (ΦE→Z) should be
enhanced upon complexation and higher than the Z → E
quantum yield (ΦZ→E) of both complexed and noncomplexed
olefins, (2) the complexed E substrate should have a stronger
light absorption of than the Z-one, and (3) the E-isomer
should have a larger equilibrium constant for complexation
than the Z-isomer (Scheme 2A). To address these challenges,
we sought to take benefit from the ability of Cu complexes to
coordinate a cinnamate derivative bearing an appropriate
achiral template to carry out this E → Z isomerization
(Scheme 2A).14 These templates have been widely used in
Lewis acid catalysis and photocatalytic [2 + 2] cycloaddition to
perform catalytic enantioselective transformations.12e−m,14 At
the outset of the project, we studied the α-methyl cinnamate
backbone as a model. Indeed, to our knowledge, no efficient
and general isomerization protocol of these substrates was
reported yet.15 The contra-thermodynamic isomerization of the
E-methyl α-methylcinnamate 1 into the Z-isomer was observed
in the presence of 5 mol % of Cu(OAc)2 and rac-BINAP in
71% yield after irradiation with blue LEDs. Importantly, no
significant isomerization occurred when (E)-1 was irradiated in
the absence of Cu(OAc)2 and/or rac-BINAP, demonstrating
the ability of this complex to promote the E → Z geometrical
inversion. Encouraged by this promising proof of concept, we
evaluated various achiral templates that would enhance the
coordination of the Cu-complex (Cu(OAc)2/rac-BINAP) and
favor the isomerization. Amide (E)-2 was less efficient than the
ester, while the introduction of a phosphate coordination site
increased the isomerization into (Z)-3 to 72%. Pleasingly,
oxazolidinone proved to be an efficient template and allowed
the isomerization of (E)-4 into (Z)-4 in 85% yield. Other
templates, N-methyl-2-aminopyridine, azaindoline, and pyr-
azole, were slightly less efficient than the oxazolidinone.16
Then the reaction conditions were optimized, and the use of
EtOAc instead of THF permitted the isomerization of (E)-4
into (Z)-4 in 89% yield (Scheme 2C).16 It is worth mentioning
that the Cu(OAc)2/rac-BINAP was the best combination as all
other ligands tested were less efficient in this isomerization
process.16 The replacement of Cu(OAc)2 by other metal salts,
Pd, Zn, Ni, and Mn, did not afford any isomerization
reaction.16 These results clearly demonstrated the real added
value of using copper in terms of efficiency, cost, and
sustainability. Finally, control experiments demonstrated that
the reaction was inefficient in the absence of light, while the
use of thermal conditions (80 °C) in the dark was
unproductive as well. To demonstrate the versatility and to
ensure the reproducibility of our protocol, the sensitivity
assessment regarding the reaction parameters was carried out
(Scheme 2C).17 Interestingly, this contra-thermodynamic
isomerization protocol is robust and insensitive toward
temperature, light intensity, scale, moisture, and oxygen.
Note that a slight decrease of the amount of (Z)-4 was
observed under diluted conditions (35% drop).16
Having these optimized conditions in hand, we evaluated the
scope and limitations of this contra-thermodynamic isomer-
ization protocol (Scheme 3A). The introduction of an ethyl
substituent at the α-position reduced the efficiency of the
isomerization process as the obtained Z/E ratio dropped from
89:11 to 78:22. The substitution pattern was evaluated with
the introduction of a methyl substituent at the para, meta, and
ortho positions of the aromatic ring (9−11). The Z/E ratio
B
Org. Lett. XXXX, XXX, XXX−XXX