Angewandte
Chemie
Molecular Switches
Photoswitching of Stereoselectivity in Catalysis
Using a Copper Dithienylethene Complex**
David Sud, Tyler B. Norsten, and Neil R. Branda*
Scheme 1. Interconversion of ring-open (o) and ring-closed (c) isomers
of 1 and 2.
Harnessing the changes in molecular structure and function
that occur when photoresponsive compounds are reversibly
toggled between thermally stable isomers is becoming
increasingly important to the development of molecular
devices containing switching elements.[1] Compounds that
undergo reversible photochemical transformations have been
investigated for use in optoelectronic technologies and to a
lesser extent, in influencing chemical reactivity. Although the
concept of using the geometric changes that accompany a
photoreaction to regulate chemical reactivity and catalysis
was introduced over 20 years ago, only a few examples have
been reported,[2–5] and all rely on azobenzene units as the
photoresponsive structure. The large molecular movements
that result from the cis–trans isomerization were shown to
influence ester hydrolysis within modified cyclodextrins[2] or
crowns ethers,[3] and amide formation in hydrogen-bonded
receptors.[4,5] Although they remain elegant examples of the
concept of photocontrolling catalysis, the photoresponsive
azobenzene derivatives are plagued by thermal reversibility,
which will significantly limit their practical use.
Our original work proved this anticipated metal-binding
mode to be incorrect and instead of forming a chelated
complex with copper(i), 1o forms a double helicate, [Cu2-
(1o)2], where two ligands are wrapped around two metals in a
stereospecific manner.[7] This result helps to explain why there
is no observable stereoselectivity when a metal-catalyzed
reaction (cyclopropanation, see entry 1 in Table 1) is carried
Table 1: Results of the cyclopropanation reactions of styrene with ethyl
diazoacetate using ligands 2 and 3, and CuOTf·0.5C6H6 in CH2Cl2.[a]
Entry
Ligand
ee [%][b]
d.r.[c]
trans
cis
Compounds constructed from the 1,2-dithienylethene
backbone represent a significant improvement over most
other photoresponsive structures primarily because they
undergo thermally irreversible photochemical ring-closing
and ring-opening reactions.[6] Our goals include using the
structural differences between the two photoisomers to
influence the outcome of metal-catalyzed reactions. We
prepared the chiral bis(oxazoline) ligand 1 with the expect-
ation that it can only chelate to a metal when the photoswitch
is in its flexible ring-open state, 1o (Scheme 1).[7] This
complexation places the metal center within a chiral environ-
ment ready to perform stereoselective reactions.[8] Irradiation
with UV light will generate the ring-closed form, 1c, which
cannot chelate the metal because the photochemically
produced rigid backbone forces the two oxazolines to exist
in a divergent relationship to each other.
1
2
3
4
2o
3o
0
30
11
5
0
50
37
5
–
55:45
70:30
63:37
23% 3c[d]
97% 3c[e]
[a] All values are averages of multiple runs using 10 mol% catalyst.[13]
[b] Determined by HPLC analysis using a CHIRACEL-OD column. The
values for the cis isomer are estimates owing to significant peak overlap
in the HPLC traces. [c] Determined by 1H NMR spectroscopic analysis of
the crude reaction mixture. The trans isomer was always found to be the
major product. [d] Represents the photostationary state generated by
irradiating a solution of the ligand with 313 nm light for 15 min.
[e] Isolated from the photostationary state by centrifugal chromatogra-
phy using 0.5% CH3OH in CH2Cl2. The remaining 3% is 3o.
out using the iso-propyl version of bis(oxazoline) 2o as the
ligand (10 to 25 mol%).[9] However, the fact that 2o may not
be conformationally well-defined around the metal center so
as to provide an appropriate chiral catalytic environment
cannot be ruled out.[10]
One of the appealing features of photoresponsive dithie-
nylethene compounds is the versatility with respect to what
can be attached onto the thiophene heterocycles. Relocating
the oxazoline groups onto the internal positions (thiophene
C2) provides ligand 3o that, according to computer modeling,
cannot form a helicate with metal ions such as copper(i). This
ligand is prepared from the known oxazoline 4 (Scheme 2) by
taking advantage of the oxazoline to direct the lithiation to
the C3 position on the thiophene.[11] Scheme 2 also illustrates
the significant differences between the metal-binding pocket
in the ring-open 3o and ring-closed 3c forms.
[*] D. Sud, Prof. N. R. Branda
Department of Chemistry
Simon Fraser University
8888 University Drive
Burnaby, BC V5A1S6 (Canada)
E-mail: nbranda@sfu.ca
T. B. Norsten
National Research Council of Canada
ICPET-Polymeric Materials Group
1200 Montreal Road, Ottawa, ON K1A0R6 (Canada)
[**] This work was supported by the Natural Sciences and Engineering
Research Council of Canada, the Canada Research Chair Program,
and Simon Fraser University. We thank Nippon Zeon Corporation
for supplying the octafluorocyclopentene needed to prepare the
photochromic compound.
Angew. Chem. Int. Ed. 2005, 44, 2019 –2021
DOI: 10.1002/anie.200462538
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2019