ACS Catalysis
Research Article
(MeCN, Purple LEDs, 400 nm, shown as dark green bars).
Finally, an attempt was made to evaluate the photostability of
the complexes. Solutions of the complexes were irradiated for
24 h, and the %catalyst remaining was calculated by 1H
NMR25 (indicated as %Cu in light green bars, Figure 2).
Successful E ↔ Z photoisomerization relies on selectively
activating one geometric isomer over the other. Three
complexes were shown to promote >96% E → Z isomer-
ization.26 Although the Cu(dmp)(BINAP)BF4 complex
possessed promising photophysical properties and promoted
isomerization (83%Z), stability studies showed that the
complex had completely degraded in solution during the
time of irradiation (indicated as %Cu in light green bars in
Figure 2) and other complexes displaying greater stability were
prioritized. Neither the corresponding homoleptic complex
(Cu(dmp)2BF4) nor the Cu(dmp)(dppf)BF4 complex re-
sulted in any conversion of E- to Z-1 (Figure 2). Note that
complexes containing the dppf ligand have been reported to
possess competing photoinduced energy transfer from the
Cu(I) complex to the ferrocene moiety with little observable
emission at room temperature.25b When evaluating the four
criteria (ET, τ, %Z, and %Cu), both the Cu(dmp)(XantPhos)-
BF4 and Cu(dmp)(DPEPhos)BF4 complexes were identified
as most promising since they ranked consistently high in τ, %Z,
and %Cu. To distinguish between the complexes, isomer-
ization processes were evaluated at only 1 h of reaction time
(see insert in Figure 2). The %Z for the copper complex based
on DPEPhos dropped to 64%, while the %Z for the XantPhos
complex decreased to only 74%. As such, a slight edge was
given to the Cu-based complex bearing XantPhos, and further
optimization was conducted to investigate the effect of diimine.
Given the promising results with Cu-based complexes bearing
XantPhos as a bisphosphine, another screening was performed
using 17 different diimines with XantPhos as bisphosphine
(Figure 2b). Diimines included phenanthroline- and bipyr-
idine-based scaffolds and possessed different degrees of
substitution that influenced both steric and electronic
properties. Once again, all complexes were evaluated against
four criteria (ET, τ, %Z, and %Cu). A total of six sensitizers
demonstrated >90% E → Z isomerization (Figure 2b, dark
green bars). The three complexes showing the largest %Z were
based on the ligands dmp, bphen, and bathocup. The
bathocup-containing complex had low stability (24%Cu), so
all three complexes were evaluated in the same model
photoisomerization but for only 1 h of irradiation time.
Indeed, the Cu(bphen)(XantPhos)BF4 complex was the only
candidate to maintain the high %Z. As such, when considering
both high %Z and high %Cu, two complexes displayed
excellent photoisomerization and stability: Cu(bphen)-
(XantPhos)BF4 and Cu(quintri)(XantPhos)BF4. The former
was selected for further study, given the bphen diimine
afforded a longer τ (1798 ns) than the quintri complex. The
analogous triazole-based ligand complexes underwent signifi-
cant degradation in solution during irradiation (Cu(pytri)-
(XantPhos)BF4:0%Cu; Cu(iquintri)(XantPhos)BF4:54%Cu).
However, further study is warranted to understand the
structure/activity relationships in complexes bearing triazole-
based ligands.
Figure 1. Photocatalysis using Cu(NN)(PP)X complexes and
possible application to E ↔ Z isomerization.
through photochemistry. Not surprisingly, the interest in
visible-light-mediated photocatalysis has helped advance the
field considerably.15 Weaver and co-workers have studied the
use of Ir(ppy)3 for the isomerization of allylic amines,16 while
Gilmour and co-workers have exploited (−)-riboflavin for a
host of isomerization processes.17
The value of isomerization can be illustrated by tandem
processes utilizing the Z-isomer, which consist of isomerization
followed by C−C cross-coupling18,19 C−O cross-coupling,20
selective reduction,21 condensation/cyclization,22 and photo-
chemical [2 + 2] cycloaddition.23 Herein, we report on the
design of a heteroleptic copper-based photosensitizer for E ↔
Z isomerization via energy transfer. In addition, the first
isomerization of 1,3-enynes and macrocyclic alkenes, a new
tandem ATRA/photoisomerization process employing arylsul-
fonyl chlorides, and a rare example of photoisomerization with
halide-substituted olefins are reported.
2. RESULTS AND DISCUSSION
2.1. Optimization of Catalyst Structure. As previous
work has shown that diimine dmp affords complexes with
higher excited-state lifetimes,6 which would have the potential
as energy-transfer catalysts,17,24 initial investigations involved
screening Cu-based complexes bearing dmp and eight other
bisphosphines. Efficiency was judged against four criteria
(Figure 2a).
The complex Cu(bphen)(XantPhos)BF4 was crystallized
via vapor diffusion of Et2O into a CH2Cl2 solution. X-ray
crystallographic analysis (Figure 3) confirmed the tetrahedral
geometry about the copper center in the ground state. The
shielding provided by n-Bu groups (proposed to extend the
The first and second criteria included photophysical
properties such as triplet state energy (ET, shown as light
blue bars) and excited-state lifetime (shown as dark blue bars).
The third criterion was the efficiency to promote the
photochemical E ↔ Z isomerization of alkenyl ester E-1
8830
ACS Catal. 2021, 11, 8829−8836