N,N′-Disubstituted Indigos as Readily Available Red-Light Photoswitches with Tunable Thermal Half-Lives

Article abstract of DOI:10.1021/jacs.7b08726

Some rare indigo derivatives have been known for a long time to be photochromic upon irradiation with red light, which should be advantageous for many applications. However, the absence of strategies to tune their thermal half-lives by modular molecular design as well as the lack of proper synthetic methods to prepare a variety of such molecules from the parent indigo dye have so far precluded their use. In this work, several synthetic protocols for N-functionalization have been developed, and a variety of N-alkyl and N-aryl indigo derivatives have been prepared. By installation of electron-withdrawing substituents on the N-aryl moieties, the thermal stability of the Z-isomers could be enhanced while maintaining the advantageous photoswitching properties upon irradiation with red light (660 nm LED). Both experimental data and computational results suggest that the ability to tune thermal stability without affecting the dyes' absorption maxima originates from the twisted geometry of the N-aryl groups. The new indigo photoswitches reported are expected to have a large impact on the development of optical methods and applications in both life and material sciences.

Full text of DOI:10.1021/jacs.7b08726

Article
pubs.acs.org/JACS
N,N′‑Disubstituted Indigos as Readily Available Red-Light
Photoswitches with Tunable Thermal Half-Lives
Chung-Yang Huang,† Aurelio Bonasera,† Lachezar Hristov,† Yves Garmshausen,† Bernd M. Schmidt,†
Denis Jacquemin,*,‡,¶ and Stefan Hecht*,†
†Department of Chemistry & IRIS Adlershof, Humboldt-Universitat zu Berlin, Brook-Taylor-Strasse 2, 12489 Berlin, Germany
̈
‡Laboratoire CEISAM, UMR CNRS 6230, Universite
́
de Nantes, 2 Rue de la Houssinier
¶Institut Universitaire de France, 1 rue Descartes, 75231 Paris, France
̀
e, BP 92208, 44322 Nantes Cedex 3, France
S
* Supporting Information
ABSTRACT: Some rare indigo derivatives have been known
for a long time to be photochromic upon irradiation with red
light, which should be advantageous for many applications.
However, the absence of strategies to tune their thermal half-
lives by modular molecular design as well as the lack of proper
synthetic methods to prepare a variety of such molecules from
the parent indigo dye have so far precluded their use. In this
work, several synthetic protocols for N-functionalization have
been developed, and a variety of N-alkyl and N-aryl indigo
derivatives have been prepared. By installation of electron-withdrawing substituents on the N-aryl moieties, the thermal stability
of the Z-isomers could be enhanced while maintaining the advantageous photoswitching properties upon irradiation with red
light (660 nm LED). Both experimental data and computational results suggest that the ability to tune thermal stability without
affecting the dyes’ absorption maxima originates from the twisted geometry of the N-aryl groups. The new indigo photoswitches
reported are expected to have a large impact on the development of optical methods and applications in both life and material
sciences.
molecules are oftentimes not red-shifted enough for biological
applications, even after delicate modifications of the azo
framework. Thus, it is (still) desirable to develop even more
red-shifted photoswitches. To our surprise, indigo, one of the
most ancient dyes known in human history, has been left out in
this recent wave of developments, although its photoisomeriza-
tion under the influence of red light has been described more
than half a century ago.8 This is presumably due to the fact that
the parent, unsubstituted indigo does not undergo photo-
induced E−Z isomerization because of its dominating and very
efficient deactivation via excited state intramolecular proton
transfer.9 In strong contrast, N,N′-disubstituted indigos,
carrying no amide protons, are capable of undergoing efficient
E−Z photoisomerization (Scheme 1).8,10−12
Photoswitches based on indigos in principle offer several
promising advantages. First and foremost, they could be
prepared from highly abundant, low-cost materials. Due to their
extended π-system and its inherent push−pull character, arising
from coupling two charge-transfer halves in a so-called H-
chromophore,13 indigos intrinsically absorb in the red region of
the spectrum, even without specific donor/acceptor derivatiza-
tion. Importantly, E−Z isomerization confers larger structural
changes, in particular with regard to substituents projected from
INTRODUCTION
■
In recent years the field of photochromism research has
witnessed exciting developments, and photoswitchable mole-
cules have been broadly applied in various disciplines of
science. In particular in the context of biomedical applications,
photoswitches have become a key tool to remotely and
noninvasively regulate processes1 in vivo, for example,
associated with photopharmacology,2 super-resolution fluores-
cence imaging,3 and optochemical genetics.4 To successfully
implement photoswitches in biological systems, they have to
meet three important criteria: First, the molecules should be
able to undergo efficient photoisomerization upon irradiation
with red or infrared light due to its superior ability to penetrate
tissues (“biological window”).5 Second, the two isomers should
possess significantly different structures to maximize the light-
induced activity modulation. Third, the thermal half-life of the
metastable, ideally biologically active, isomer should be tunable
to adjust its in vivo residence time according to circulation or
other parameters of interest.
To date, azobenzenes have been the most promising
candidates due to their significant structural changes as a result
of the E−Z isomerization as well as their established
photochemistry and the straightforward synthesis.6 Recent
efforts by multiple groups have led to the development of novel
azo compounds addressable by visible light,7 clearly improving
their applicability. Despite the significant progress made, these
Received: August 19, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Photoswitchable N,N′-Disubstituted Indigos
Scheme 2. Approaches for Indigo Functionalization
the indigo N-atoms as compared to the commonly exploited
azobenzene para-positions, thus rendering indigos superior
steric switches. In addition, indigo photoswitches should
presumably exhibit reasonably low toxicity as evidenced by
their extensive use for centuries. Last but not least, indigos
display negative photochromism,14 providing a means to
enhance penetration depth through the course of irradiation
and therefore facilitating their application in pharmacology and
bulk materials.
method D),18 which subsequently can be alkylated or arylated
using the previously described methods A−C to obtain indigos
with two different N,N′-substituents. Although these reactions
have not yet been fully optimized, they serve as rapid and
straightforward entries into symmetrical and nonsymmetrical
N,N′-disubstituted indigos from the abundant and inexpensive
parent indigo dye. All of the prepared compounds were
obtained as dark blue or green solids, and, unlike their parent
indigo precursor, they are readily soluble in most organic
solvents, simplifying purification, characterization, and further
processing.
To date, symmetrical N,N′-diacyl,10 N,N′-dimethyl,11 and
N,N′-di(tert-butyloxycarbonyl)12 indigos are among the only
reported photoswitches in this family. Their absorption maxima
(λmax) range from 530 nm all the way to 670 nm, which is
strongly red-shifted when compared to conventional E−Z
photoswitches, such as stilbenes and azobenzenes.7a However,
these compounds suffer from either poor hydrolytic stability or
lack of direct functionalization, and more importantly,
molecular design rules allowing to rationally tune their thermal
half-lives remain elusive. Herein, we report a series of
symmetrical and nonsymmetrical N-alkyl- and N-aryl-substi-
tuted indigos, which allow for systematic modulation of their
thermal half-lives by direct structural modification of the parent
indigo while maintaining its advantageous red-shifted absorp-
tion (Scheme 1).
Our initial studies focused on compounds in series I, i.e.,
N,N′-dialkylindigos (Table 1). In accordance with the
Table 1. Photochemical Properties of Series I in CH3CN
RESULTS AND DISCUSSION
■
The initial hurdle that we had to overcome was the lack of
synthetic methods known to install N-substituents on the
parent indigo. According to scattered reports in the literature,
N-alkylation and N-arylation often require harsh conditions and
result in complex mixtures and low yields. We tried to
circumvent this problem by identifying appropriate function-
alization reagents that would provide enhanced reactivity yet
under milder conditions. To our delight, by using a weak base
and activated electrophiles, such as benzyl bromide or α-
bromoacetate, double N-alkylation of indigo can be achieved at
room temperature (Scheme 2, method A). Furthermore, we
developed two sets of complementary conditions that allow for
double N-arylation15 of indigo under mild conditions. For
installing electron-rich or electron-neutral aryl substituents, it
was found that Chan-Lam coupling is the most suitable
(Scheme 2, method B),16 while in the case of electron-deficient
aryl groups, Cu-catalyzed cross-coupling reaction with
aryliodonium salts was applied (Scheme 2, method C).17
Interestingly, both methods are extremely selective toward
double N,N′-arylation, and only minor amounts of mono-N-
arylated products were observed. However, to gain full
synthetic flexibility for functionalizing indigos, the ability to
prepare nonsymmetrical compounds is also critical. In this
context, a Goldberg−Ullmann reaction proved useful as it
reliably provides the mono-N-arylated indigos (Scheme 2,
series I: R1 =
Me (2)
Bn (3)
CH2CO2tBu (4)
λmax, E (nm)
λmax, Z (nm)
εE (M−1 cm−1
t1/2
641
648
623
561
a
a
n.a.
n.a.
)
16700
<5 s
10800
<5 s
15700
2.8 min
77%
a
a
Z-content PSS
ϕE−Z (578 nm)
ϕZ−E (578 nm)
n.a.
n.a.
a
a
b
n.a.
n.a.
0.06
a
a
b
n.a.
n.a.
0.60
a
Not available due to extremely short lifetime of Z-isomer.
Determined at 25 °C.
b
literature,11 N,N′-dimethylindigo 2 and N,N′-dibenzylindigo 3
are among the most red-shifted compounds examined in this
study, whereas in the case of N,N′-di(benzyloxycarbonyl-
methyl)indigo 4 (Figure 1), the presence of electron-with-
drawing ester groups causes a slight blue shift by 20 nm. Upon
irradiation with a red LED (660 nm, 20 nm full width at half-
maximum (fwhm)), the intensity of the long-wavelength
absorption band decreased, and a shoulder below 600 nm
started to rise. When the light was turned off, 2 and 3 returned
B
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
the Z-isomer. However, in compound 7, where the tert-
butyloxycarbonyl group is attached directly to the indigo N-
atom, a much larger blue shift as well as a much longer thermal
half-life was observed. These findings are in line with the
behavior of N,N′-diacyl10 and N,N′-di(tert-butyloxy-carbonyl)12
indigos, in which the electron-withdrawing carbonyl-based N-
substituents result in thermally much more stable Z-isomers
and yet inevitable significantly blue-shifted absorption spectra,
rendering them no longer suitable for operation using red light.
In order to overcome this problem, we subsequently decided to
explore substituent effects on the N-aryl rings.
In series III we therefore fixed the CH2CO2tBu as the alkyl
group and studied the electronic effect of para-substituents on
the aryl ring (Table 3). We found that the nature of the
Table 3. Photochemical Properties of Series III in CH3CN
Figure 1. Photochromic and thermal behavior of indigo 4. (a)
Chemical structure. (b) Photographs of E-4 and PSS-4 solutions in
CH3CN. (c) Evolution of the PSS upon irradiation with 660 nm LED
with spectra taken every 30 s over 10 min at 25 °C (4.2 × 10−5 M in
CH3CN). (d) Thermal recovery at 25 °C monitored by absorbance at
625 nm in the dark.
to their initial state immediately, whereas 4 showed a slower
recovery with a thermal half-life (t1/2) of 2.8 min at room
temperature (see Figure 1c, d for a representative photo-
chemical forward and thermal back isomerization). The fact
that these indigo photoswitches exhibit negative photochrom-
ism allowed us to obtain the composition of the photosta-
tionary state (PSS) and the absorption spectrum of the pure Z-
isomer directly from the E and PSS spectra (see Section III of
the Supporting Information). Applying this methodology, we
found that the absorption maxima of E-4 and Z-4 are well
separated (Δλmax = 62 nm).
Next, we turned our attention to nonsymmetrical indigos
substituted with one aryl and one alkyl group each. First, we
compared series II, in which the aryl moiety (4-trifluorome-
thylphenyl) was kept constant and the alkyl group was varied
(Table 2). Similar to what we observed in N,N′-dialkylindigos,
the presence of the electron-withdrawing ester group in 6
causes a slight blue shift and an increased thermal half-life of
series III: R3 =
OMe (8)
H (9)
CF3 (6)
CN (10)
λmax, E (nm)
λmax, Z (nm)
εE (M−1 cm−1
t1/2
633
593
629
578
622
572
623
567
)
13500
59 s
11%
0.05
0.04
11900
1.9 min
62%
13200
3.5 min
56%
9800
5.8 min
71%
Z-content PSS
a
ϕE−Z
0.01
0.03
0.05
a
ϕZ−E
0.10
0.03
0.04
a
Determined at −5 °C and 578 nm (see Supporting Information for
full details).
substituents does not have a large influence on the absorption
maximum of the E-isomer, which for the entire series remains
in the red domain (623 ≤ λmax ≤ 633 nm). Introduction of
electron-withdrawing substituent leads to an increased spectral
separation of the absorption maxima of E- and Z-isomers. In
addition, we observed a sixfold increase in thermal half-life
when changing from an electron-donating methoxy (8) to an
electron-withdrawing cyano group (10). Thus, in acceptor-
substituted indigo 10, both enhanced spectral separation and
thermal lifetime lead to a much larger Z-content in the PSS as
compared to indigo 8. More importantly, this finding
demonstrates that the thermal half-life of indigo photoswitches
can indeed be tuned independently from their absorption
spectra by proper structural modification and encouraged us to
pursue N,N′-arylated indigo derivatives.
Table 2. Photochemical Properties of Series II in CH3CN
series II: R2 =
Me (5)
CH2CO2tBu (6)
Boc (7)
Utilizing the synthetic methods developed by us for indigo
N-arylation, we finally prepared two series of symmetrically as
well as nonsymmetrically substituted N,N′-diaryl indigos (series
IV and V). To our surprise, NMR analysis for these compounds
revealed that they all exist as a mixture of E- and Z-isomers.
Typically, we observe an E/Z ratio between 2:1 and 3:1 in
CD2Cl2 (see Section VIII of the Supporting Information). The
existence of such mixtures in the dark state implies that the
stability of both isomers is rather comparable and that the
separating barrier allows for thermal isomerization (in both
λmax, E (nm)
λmax, Z (nm)
εE (M−1 cm−1
t1/2
635
595
622
572
584
525
)
13000
57 s
38%
0.04
0.04
13200
3.5 min
56%
9900
193 min
80%
Z-content PSS
a
ϕE−Z
0.03
0.12
a
ϕZ−E
0.03
0.05
a
Determined at −5 °C and 578 nm (see Supporting Information for
full details).
C
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
directions) at room temperature. DFT-computed relative free
energies of the two isomers (see Table S1 in the Supporting
Information) confirm that the bis-arylated indigos exhibit
negligible ground-state energy differences between the two
isomers. Indeed, we determined free energy differences smaller
than 1 kcal/mol between the Z- and E-isomers of 11−18, a
value which is within the typical DFT error bar, indicating
nearly isoenergetic structures. A possible explanation involves
attractive π−π interactions between the two aryl moieties in the
Z-isomer on the one hand and repulsive interactions between
the aryl moiety and the indigo carbonyl group in the E-isomer
on the other hand. Both of these interactions are plausible by
comparing the molecular structures (Figure 2 for E-12 vs Z-
12), obtained by single-crystal X-ray diffraction and quantum
chemical calculation (see Table S2 of the Supporting
Information).
PSS to the thermodynamic equilibrium in the dark, we
introduce here the term “equilibration thermal half-life
t(eq)1/2”. Following the recovery of absorbance at 620 nm,
t(eq)1/2 was obtained for each compound (Table 4). A clear
trend of how the substitution pattern influences t(eq)1/2 was
observed as increasingly electron-withdrawing groups result in
longer t(eq)1/2, spanning a range from 58 s for OMe (11) to
408 min for NO2 (16). Importantly, this dramatic difference in
the rate of thermal recovery is accompanied by negligible
changes in the absorption maxima (λmax) of the molecules,
which remain red-shifted regardless of their substitution, and
thus a red LED (660 nm, fwhm = 20 nm) was utilized in all
cases to successfully induce the photoisomerization.
The observed correlation between the N-aryl substituents
and thermal half-lives led us to further investigate the process of
thermal relaxation. Comparing the compounds from series III
through V, it appears that in general N-arylation seems to
stabilize the Z-isomer relative to the E-isomer while at the same
time the transition state (TS) does not seem to be affected,
leading to a net increase of the thermal barrier. This is reflected
by the analysis of our kinetic data that shows a linear
correlation between the rate constants for thermal equilibration
and Hammett’s substituent parameters20 for both series III and
IV (Figure 4 and Section VIII of the Supporting Information).
In both cases the slope is clearly negative, indicating that
electron-deficient aryl moieties, in particular when attached to
both indigo N-atoms, slow down thermal return from the
photostationary to the dark state.
We performed broken-symmetry density functional theory
(BS-DFT) calculations aiming to determine and characterize
the TS for all dyes of series IV and V (see Section XII of the
Supporting Information). The obtained TS geometries exhibit a
dihedral angle close to 90° between the two halves of the
molecules, and the associated imaginary frequency is close to
50i cm−1 (Figure 5b). Gratifyingly, the correlation found
between our experimental and theoretical data is remarkable
(R2 = 0.97, Figure 5a), confirming the accuracy of our
computational model and thus providing a solid foundation and
encouraging prospect to design indigo derivatives with
improved thermal stability. More detailed analysis of the TS
(see Section XII of the Supporting Information) indicates that
the largest spin density of the unpaired electron on each side is
located on each of the neighboring bridging carbon atoms (ca.
0.5 e) whereas the spin density borne by the oxygen atoms is
smaller (ca. 0.3 e). In addition, the Mulliken electronic charges
borne by the oxygen atoms are only marginally larger in the TS
(ca. −0.6 e) as compared to the Z- and E-isomers (ca. −0.5 e).
These findings are in agreement with a predominant biradical
character of the TS.10h
Figure 2. (a) X-ray structure of E-12 (50% probability ellipsoids). (b)
DFT-computed structure of Z-12 (for details see Section XII of the
Supporting Information).
Irradiating these N,N′-diaryl indigos with a red LED (660
nm, fwhm = 20 nm) led to a different E/Z ratio in the PSS.
Based on 19F-NMR spectroscopy, compound 14 went from a
dark mixture containing 26% Z-isomer to a solution containing
81% Z-isomer in the PSS at room temperature, and when this
sample was left in the dark, it returned to the initial E/Z
mixture (Figure 3 and Supporting Information). Initial cycling
experiments with compound 12 indicated a robust switching
process without noticeable fatigue (see Section XI in the
Supporting Information).
Subsequently, we analyzed the kinetics of the thermal back
reaction.19 To describe the time required for returning from the
We have also characterized the derivatives of series IV and V
using time-dependent DFT. The theoretically predicted
absorption maxima (see Table S3 of the Supporting
Information) are well in line with the experimentally observed
values for both Z- and E-isomers. As expected, the absorption
maxima of all compounds and isomers are related to the lowest
excited state, which can be assigned to a HOMO → LUMO
transition. These frontier orbitals are centered on the H-
chromophore of the indigo core (Figure 5c, d), and the aryl
substituents contribute only little, explaining the relatively small
variations of the absorption spectra (see Table 4). This
decoupling of the aryl units is due to their almost perpendicular
orientation relative to the plane of the indigo. Consequently,
these groups can influence the central chromophore only via an
Figure 3. 19F-NMR spectra of dark vs irradiated indigo 14. Thermally
equilibrated E/Z mixture of 14 (ca. 10−4 M in CD3CN) in the dark
(top) and after irradiation with 660 nm LED for 3 h at 25 °C
(bottom). The initial spectrum (top) is recovered in the dark
overnight (see Supporting Information for 1H- and 2D NMR spectra).
D
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 4. Photochemical Properties of Series IV and V in CH3CN
series IV: R4 =
series V: R5 =
OMe (11)
H (12)
F (13)
CF3 (14)
CN (15)
NO2 (16)
OMe (17)
F (18)
λmax, E (nm)
λmax, Z (nm)
Z-content dark
t(eq)1/2
645
599
635
594
630
584
621
577
619
568
620
577
632
590
624
582
33%
58 s
52%
24%
34%
27%
25%
12%
39%
33%
3.1 min
64%
7.9 min
71%
81 min
83%
187 min
80%
408 min
82%
13 min
67%
19 min
76%
Z-content PSS
a
b
b
ϕE−Z
n.d.
n.d.
0.13
0.06
0.06
0.07
0.36
0.15
a
b
b
ϕZ−E
n.d.
n.d.
0.08
0.04
0.03
0.04
0.08
0.08
a
b
Determined at 25 °C and 578 nm (except for 17 at 546 nm; see Supporting Information for full details). Not determined due to short lifetime of
Z-isomer.
Figure 4. Hammett plots of the rate of thermal equilibration (from the
PSS to the dark state) for indigo series III and IV.
inductive effect, in contrast to the carbonyl-based N-
substituents, such as in 7, where resonance effects predominate,
resulting in more stable Z-isomers yet significantly blue-shifted
absorption spectra of the E-isomers.
An important merit of indigo photoswitches is their superior
ability to change the distance between two functional motifs by
light. A closer examination of both the DFT-optimized E- and
Z-isomers of indigos 4, 6, and 12 revealed that during the
course of photoisomerization a change of distance between the
two substituents of 2.7, 3.8, and 3.5 Å, respectively, can be
achieved (see Table S2 in the Supporting Information), which
is, for the arylated indigos, slightly superior to the ca. 3 Å
change that is generally exerted by azobenzenes. This finding
demonstrates the potential of indigos to serve as superior steric
switch. Furthermore, DFT calculations indicate that E−Z
photoisomerization is able to induce significant changes in
dipole moment of up to 7 D between the two isomers (see
Table S1 in the Supporting Information).
Figure 5. (a) Correlation between the experimentally determined
ln(t(eq)1/2) given in Table 4 and the PCM-(BS-)PBE0-D3BJ/6-
31G(d) computed free energy of the TS with respect to the Z-isomer
(ΔG‡). (b) Two views of the TS obtained for 13 (H-atoms omitted
for clarity). (c) HOMO of E-13. (d) LUMO of E-13 (contour
threshold: 0.04 au).
CONCLUSION
■
In conclusion, we have revisited the classical indigo
chromophore motif and developed a strategy to tune the
E
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
́
M. J.; van Dam, G. M.; Szymanski, W.; Feringa, B. L. Angew. Chem.,
thermal stability of the Z-isomers independently from the
advantageous optical spectra. The introduction of electron-
withdrawing N-aryl substituents has proven key to achieve
sufficiently long thermal half-lives of the Z-isomers, which are
attractive for biomedical applications, while maintaining the
compounds’ absorption in the red and hence photoisomeriza-
tion ability with 660 nm light. In contrast to previously
reported amide-containing indigos,10,12 the new indigo
derivatives display excellent hydrolytic stability in aqueous
solutions containing acetonitrile to facilitate solubility. Using a
combined experimental and theoretical approach, we have
gained general insight into the factors determining the thermal
stability of the indigo isomers and their electronic and
geometrical structure, encoding both their optical properties
and activity differences. The latter are expected to be significant
due to large geometry changes during photoisomerization and
promise that indigos might indeed perform as superior steric
photoswitches. Thus, our work introduces an exciting new
generation of indigo derivatives, which are readily derived from
one of the cheapest and most abundant dyes on the planet and
which should greatly expand the tool box available to
researchers, pursuing to exploit photoswitches in future life
and material science applications.
Int. Ed. 2016, 55, 10978−10999.
(3) Heilemann, M.; Dedecker, P.; Hofkens, J.; Sauer, M. Laser
Photonics Rev. 2009, 3, 180−202.
(4) Fehrentz, T.; Schonberger, M.; Trauner, D. Angew. Chem., Int. Ed.
̈
2011, 50, 12156−12182.
(5) Pansare, V. J.; Hejazi, S.; Faenza, W. J.; Prud’homme, R. K. Chem.
Mater. 2012, 24, 812−827.
(6) (a) Beharry, A. A.; Woolley, G. A. Chem. Soc. Rev. 2011, 40,
4422−4437. (b) Mart, R. J.; Allemann, R. K. Chem. Commun. 2016, 52,
12262−12277.
́
(7) An overview is given in: (a) Bleger, D.; Hecht, S. Angew. Chem.,
Int. Ed. 2015, 54, 11338−11349. Examples include: (b) Siewertsen,
R.; Neumann, H.; Buchheim-Stehn, B.; Herges, R.; Nather, C.; Renth,
F.; Temps, F. J. Am. Chem. Soc. 2009, 131, 15594−15595. (c) Beharry,
A. A.; Sadovski, O.; Woolley, G. A. J. Am. Chem. Soc. 2011, 133,
19684−19687. (d) Samanta, S.; Qin, C.; Lough, A. J.; Woolley, G. A.
Angew. Chem., Int. Ed. 2012, 51, 6452−6455. (e) Yang, Y.; Hughes, R.
P.; Aprahamian, I. J. Am. Chem. Soc. 2012, 134, 15221−15224.
́
(f) Bleger, D.; Schwarz, J.; Brouwer, A. M.; Hecht, S. J. Am. Chem. Soc.
2012, 134, 20597−20600. (g) Samanta, S.; Beharry, A. A.; Sadovski,
O.; McCormick, T. M.; Babalhavaeji, A.; Tropepe, V.; Woolley, G. A. J.
Am. Chem. Soc. 2013, 135, 9777−9784. (h) Bushuyev, O. S.; Tomberg,
A.; Frisc
12559. (i) Knie, C.; Utecht, M.; Zhao, F.; Kulla, H.; Kovalenko, S.;
Brouwer, A. M.; Saalfrank, P.; Hecht, S.; Bleger, D. Chem. - Eur. J.
̌ ̌ ́
ic, T.; Barrett, C. J. J. Am. Chem. Soc. 2013, 135, 12556−
́
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b08726.
2014, 20, 16492−16501. (j) Weston, C. E.; Richardson, R. D.;
Haycock, P. R.; White, A. J. P.; Fuchter, M. J. J. Am. Chem. Soc. 2014,
136, 11878−11881. (k) Hansen, M. J.; Lerch, M. M.; Szymanski, W.;
Feringa, B. L. Angew. Chem., Int. Ed. 2016, 55, 13514−13518.
(l) Stricker, L.; Fritz, E.-C.; Peterlechner, M.; Doltsinis, N. L.; Ravoo,
B. J. J. Am. Chem. Soc. 2016, 138, 4547−4554. (m) Calbo, J.; Weston,
C. E.; White, A. J. P.; Rzepa, H. S.; Contreras-García, J.; Fuchter, M. J.
J. Am. Chem. Soc. 2017, 139, 1261−1274.
■
S
Experimental details including synthetic procedures,
compound characterization data, and photochemical
and kinetic experiments, computational details (PDF)
Crystallographic data (CIF)
(8) Wyman, G. M. Chem. Rev. 1955, 55, 625−657.
(9) Pina, J.; Sarmento, D.; Accoto, M.; Gentili, P. L.; Vaccaro, L.;
Galvao, A.; de Melo, J. S. S. J. Phys. Chem. B 2017, 121, 2308−2318.
̃
AUTHOR INFORMATION
■
and references therein.
Corresponding Authors
*denis.jacquemin@univ-nantes.fr
*sh@chemie.hu-berlin.de
(10) (a) Brode, W. R.; Pearson, E. G.; Wyman, G. M. J. Am. Chem.
Soc. 1954, 76, 1034−1036. (b) Wyman, G. M.; Zenhausern, A. F. J.
̈
Org. Chem. 1965, 30, 2348−2352. (c) Giuliano, C. R.; Hess, L. D.;
Margerum, J. D. J. Am. Chem. Soc. 1968, 90, 587−594. (d) Omote, Y.;
Imada, S.; Matsuzaki, R.; Fujiki, K.; Nishio, T.; Kashima, C. Bull. Chem.
Soc. Jpn. 1979, 52, 3397−3399. (e) Setsune, J.-I.; Wakemoto, H.;
Matsukawa, K.; Ishihara, S.-I.; Yamamoto, R.-I.; Kitao, T. J. Chem. Soc.,
Chem. Commun. 1982, 1022−1023. (f) Omote, Y.; Tomotake, A.;
Aoyama, H.; Nishio, T.; Kashima, C. Bull. Chem. Soc. Jpn. 1984, 57,
470−472. (g) Pouliquen, J.; Wintgens, V.; Toscano, V.; Jaafar, B. B.;
Tripathi, S.; Kossanyi, J.; Valat, P. Can. J. Chem. 1984, 62, 2478−2486.
(h) Sueishi, Y.; Ohtani, K.; Nishimura, N. Bull. Chem. Soc. Jpn. 1985,
ORCID
Denis Jacquemin: 0000-0002-4217-0708
Stefan Hecht: 0000-0002-6124-0222
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work is dedicated to the memory of Lachezar Hristov. The
authors thank Dr. Benjamin Koeppe for initial discussions and
Dr. Andre Dallmann for help with NMR analysis. C.-Y.H. and
B.M.S. are indebted to the Alexander von Humboldt
Foundation for providing postdoctoral fellowships. L.H. was
supported by a Fulbright predoctoral fellowship. Generous
support by the European Research Council (ERC via ERC-
2012-STG_308117 “Light4Function”) and the German
Research Foundation (DFG via SFB951, project A3) is
gratefully acknowledged. This work used the computational
facilities of the CCIPL center installed in Nantes.
58, 810−814. (i) Gorner, H.; Pouliquen, J.; Kossanyi, J. Can. J. Chem.
̈
1987, 65, 708−717.
(11) (a) Weinstein, J.; Wyman, G. M. J. Am. Chem. Soc. 1956, 78,
4007−4010. (b) Miehe, G.; Susse, P.; Kupcik, V.; Egert, E.; Nieger,
̈
M.; Kunz, G.; Gerke, R.; Knieriem, B.; Niemeyer, M.; Luttke, W.
̈
Angew. Chem., Int. Ed. Engl. 1991, 30, 964−967.
(12) (a) Głowacki, E. D.; Voss, G.; Demirak, K.; Havlicek, M.;
Sunger, N.; Okur, A. C.; Monkowius, U.; Gąsiorowski, J.; Leonat, L.;
̈
Sariciftci, N. S. Chem. Commun. 2013, 49, 6063−6065. (b) Farka, D.;
Scharber, M.; Głowacki, E. D.; Sariciftci, N. S. J. Phys. Chem. A 2015,
119, 3563−3568.
(13) (a) Klessinger, M.; Luttke, W. Tetrahedron 1963, 19, 315−335.
̈
REFERENCES
(b) Dahne, S.; Leupold, D. Angew. Chem., Int. Ed. Engl. 1966, 5, 984−
̈
■
993.
(1) Szyman
́
(14) Bouas-Laurent, H.; Durr, H. Pure Appl. Chem. 2001, 73, 639−
A.; Feringa, B. L. Chem. Rev. 2013, 113, 6114−6178.
(2) (a) Velema, W. A.; Szyman
Soc. 2014, 136, 2178−2191. (b) Broichhagen, J.; Frank, J. A.; Trauner,
D. Acc. Chem. Res. 2015, 48, 1947−1960. (c) Lerch, M. M.; Hansen,
̈
665.
(15) Matsumoto, Y.; Tanaka, H. Heterocycles 2003, 60, 1805−1810.
(16) Qiao, J. X.; Lam, P. Y. S. Synthesis 2011, 6, 829−856.
F
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(17) Yoshimura, A.; Zhdankin, V. V. Chem. Rev. 2016, 116, 3328−
3435.
(18) Kim, I. K.; Li, X.; Ullah, M.; Shaw, P. E.; Wawrzinek, R.;
Namdas, E. B.; Lo, S.-C. Adv. Mater. 2015, 27, 6390−6395.
(19) (a) Delbaere, S.; Vermeersch, G.; Micheau, J.-C. J. Photochem.
Photobiol., C 2011, 12, 74−105. (b) Hemmer, J. R.; Poelma, S. O.;
Treat, N.; Page, Z. A.; Dolinski, N. D.; Diaz, Y. J.; Tomlinson, W.;
Clark, K. D.; Hooper, J. P.; Hawker, C.; Read de Alaniz, J. J. Am. Chem.
Soc. 2016, 138, 13960−13966.
(20) (a) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96−103.
A
review is given in: (b) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev.
1991, 91, 165−195.
G
DOI: 10.1021/jacs.7b08726
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX