Photostability of 4,4′-dihydroxythioindigo, a mimetic of indigo

Article abstract of DOI:10.1002/anie.201307016

The photochemical properties of indigo, a widely used industrial dye, has attracted both experimentalists and theoreticians from the beginning. Especially the high photostability of indigo has been the subject of intensive research. Recently, it was propo

Full text of DOI:10.1002/anie.201307016

Angewandte
Chemie
DOI: 10.1002/anie.201307016
Indigo Mimetic
Photostability of 4,4’-Dihydroxythioindigo, a Mimetic of Indigo**
Marc Dittmann, Franziska F. Graupner, Benjamin Maerz, Sven Oesterling,
Regina de Vivie-Riedle, Wolfgang Zinth,* Martin Engelhard,* and Wolfgang Lꢀttke
Abstract: The photochemical properties of indigo, a widely
used industrial dye, has attracted both experimentalists and
theoreticians from the beginning. Especially the high photo-
stability of indigo has been the subject of intensive research.
Recently, it was proposed that after photoexcitation an intra-
molecular proton transfer followed by a nonradiative relaxa-
tion to the ground state promote photostability. In indigo the
hydrogen bond and the proton transfer occur between the
opposing hemiindigo parts. Here, we provide experimental and
theoretical evidence that a hydrogen transfer within one
hemiindigo or hemithioindigo part is sufficient to attain
photostability. This concept can serve as an interesting strategy
towards new photostable dyes for the visible part of the
spectrum.
The photochemical properties of indigo, for example its
long-wavelength absorption maximum and its photostability,
have attracted the interest of experimentalists as well as
theoreticians from the beginning and have resulted in an
abundance of publications over the past decades. Quite early
on, in theoretical studies using LCAO molecular orbital
calculations (LCAO = linear combination of atomic orbitals)
a basic chromophore structure of indigo was proposed[5]
which was later confirmed by chemical synthesis.[6]
I
ndigo (1) has been used as brilliant dye for centuries. After
the determination of its structure and its chemical synthesis by
Adolf von Baeyer,[1] indigo was and still is produced indus-
trially because of its characteristic color and most importantly
its high photostability. In recent times new applications have
been proposed for this “old” dye. For example, an ambipolar
organic field-effect transistor on the basis of indigo has been
developed.[2] Derivatives of indigo like hemithioindigo[3] are
promising tools in the competitive field of chemical opto-
genetics.[4]
The photostability of indigo was attributed originally to
the lack of trans–cis photoisomerization.[7] In contrast, indigo
derivatives like thioindigo (2) readily isomerize about the
central double bond[8] and, unlike indigo, show strong
fluorescence.[9] It was proposed that the trans isomer of
indigo is stabilized by hydrogen bonds formed between the
NH and CO groups connecting the two halves of the
molecule.[7,10] It was also discussed that after photoexcitation
an intramolecular proton transfer followed by a nonradiative
relaxation to the ground state occurs.[11] However, in pico-
second infrared experiments Elsaesser et al. detected only
small changes and they concluded that a proton transfer does
not occur in the S1 state.[12] Recent work reexamined indigo
photochemistry. Iwakura et al.[13] reported on the fast photo-
excited proton transfer which should occur on the femto-
second time scale. Interestingly the back reaction should be
completed already at 0.5 ps, which explains the results of
Elsaesser et al.[12] Computational analysis on the photosta-
bilty of indigo addressed the role of isomerization and
excited-state proton transfer and found that rapid internal
conversion is induced by intramolecular single-proton trans-
fer.[14] In indigo the hydrogen bond and the proton transfer
occur between the opposing hemiindigo parts. The question
whether a hydrogen transfer within one hemiindigo (HI) or
hemithioindigo part would be sufficient to ascertain photo-
stability is addressed herein.
[*] M. Dittmann, Prof. Dr. M. Engelhard
Max Planck Institute of Molecular Physiology
Otto-Hahn-Strasse 11, 44227 Dortmund (Germany)
E-mail: martin.engelhard@mpi-dortmund.mpg.de
F. F. Graupner, B. Maerz, Prof. Dr. W. Zinth
Faculty of Physics and Center for Integrated Protein Science CIPSM
Ludwig-Maximilians-Universitꢀt Mꢁnchen
Oettingenstrasse 67, 80538 Mꢁnchen (Germany)
E-mail: wolfgang.zinth@physik.uni-muenchen.de
S. Oesterling, Prof. Dr. R. de Vivie-Riedle
Department of Chemistry
Ludwig-Maximilians-Universitꢀt Mꢁnchen (Germany)
Prof. Dr. W. Lꢁttke
Institut fꢁr Organische und Biomolekulare Chemie
Georg-August-Universitꢀt Gçttingen
Tammanstrasse 4, 37077 Gçttingen (Germany)
[**] This work was supported by the DFG through the SFB “Dynamics
and intermediates of molecular transformations” (SFB 749, A5 and
C2) and the Cluster of Excellence “Munich Center for Advanced
Photonics” (MAP). W.Z. and F.F.G. thank C. Nehls for help with the
time-resolved emission experiments. Scholarships from the Inter-
national Max Planck Research School (IMPRS) in Chemical Biology
and the Studienstiftung des deutschen Volkes are gratefully
acknowledged (M.D.).
Here we report on the synthesis and photophysical
properties of the newly synthesized 4,4’-dihydroxythioindigo
as well as 4,4’-dimethoxythioindigo (3), which was first
Supporting information for this article (including full experimental
details) is available on the WWW under http://dx.doi.org/10.1002/
anie.201307016.
Angew. Chem. Int. Ed. 2014, 53, 591 –594
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
591

.
Angewandte
Communications
prepared by Friedlꢀnder.[15] The spectroscopic properties of
these compounds are similar to those of indigo (1) and
thioindigo (2), respectively. Like indigo, 4,4’-dihydroxythioin-
digo does not undergo photoisomerization and displays a fast
internal proton transfer. However, the possible proton trans-
fer occurs only within the hemithioindigo units.
4,4’-Dihydroxyindigo (3) and its dimethyl ether 4 were
synthesized in six and five steps, respectively, using a strategy
proven successful for the synthesis of thioindoxyls.[16,17]
Oxidation of thioindoxyls resulted in the formation of the
thioindigo derivatives with overall yields of 25% for 3 and
30% for 4. 4-Methoxy-4’-hydroxythioindigo (5) was obtained
by partial hydrolysis of (4) and purified by thin-layer
chromatography (see the Supporting Information).
Stationary irradiation experiments yielded the following
results. Table 1 summarizes the results from long-term
irradiation experiments. The samples were illuminated with
a lamp (15 W) equipped with a yellow filter (for experimental
Table 1: Photoisomerization of indigo dyes in different solvents.[a]
DMSO DMF CHCl3 Benzene
Thioindigo
ꢀ
+
ꢀ
ꢀ
ꢀ
ꢀ
+
ꢀ
ꢀ
ꢀ
ꢀ
+
4,4’-Dimethoxythioindigo
4,4’-Dihydroxythioindigo
Indigo
+
+
ꢀ
ꢀ
ꢀ
ꢀ
Figure 1. a) Time-dependent absorption change (Trans. Abs., red,
blue) and fluorescence emission (light red and light blue) for 4 (red)
and 3 (blue). b,c) Transient absorption difference data for the two
compounds. Excitation wavelengths 525 nm (emission) and 550 nm
(absorption experiments). For details of the experiments see the
Supporting Information.
4-Methoxy-4’-hydroxythioindigo
n.d.
n.d.
[a] “+”: photoisomerization trans!cis; “ꢀ”: no photoisomerization
trans!cis; n.d.: not determined.
details and use of another irradiation system see the
Supporting Information). As expected, indigo (1) was photo-
stable after 2 h of irradiation. Even after six weeks of
exposure to daylight dihydroxythioindigo (3) showed no
detectable signs of decomposition. A trans–cis isomerization
yielding a stable product was not observed. In contrast, under
identical conditions the dimethoxy derivative 4 and thioin-
digo (2) decomposed completely. In addition, the fluores-
cence of 3 and 5 was considerably weaker than that of 2 and 4.
Interestingly, isomerization of thioindigo (2) and dimethox-
ythioindigo (4) is quite dependent on the solvent (Table 1).
Generally, more polar solvents inhibit trans–cis isomerization.
When a strong acid (trifluoroacetic acid, TFA) was added to
a solution of thioindigo, practically no isomerization was
observed after two hours.
The strong change in fluorescence intensity is directly
related to the lifetime of the light-emitting S1 state, as can be
seen from the time dependence of the fluorescence emission
recorded with a streak camera (Figure 1a, light red and light
blue symbols; for experimental details see the Supporting
Information). For compound 4 we observe a decay of the
fluorescence emission with a time constant of approximately
11 ns (dots). The decay time of compound 3 of 50 ps
(triangles) is roughly 200 times faster. Data recorded at
618 nm (4) and 597 nm (3) show that the absorption transients
evolve with the same decay times as those found for
fluorescence emission (Figure 1a). Additional information
was obtained from time-resolved absorption spectra recorded
at distinct delay times. The transient difference spectra of
compound 3 (Figure 1b) evolve on the picosecond time scale
and finally disappear with the same time constant (50 ps) as
that found for the decay of the excited electronic state. No
absorption change remains at later times. At early times the
transient absorption spectrum displays a strong and narrow
peak at 600 nm due to excited-state absorption. Aweaker and
broader excited-state absorption band is found in the blue
part of the spectrum around 430 nm. At longer wavelengths
(630–700 nm) stimulated emission is visible and around
550 nm, the bleaching of the original ground state absorption
is found. A global fitting procedure reveals additional
absorption dynamics with a time constant of roughly 8 ps,
which can be related to motions on the excited-state potential
energy surface away from the originally populated Franck–
Condon region.
Dimethoxythioindigo (4) shows very similar spectral
features in the excited electronic state. Again there is
a pronounced excited-state absorption peak for compound 4
at 618 nm. The decay of the absorption changes occurs much
slower than in compound 3. This long-term behavior can be
well fitted with the same time constant as that found in the
emission experiments.
The experiments on the two thioindigo compounds are
summarized as follows: Dihydroxythioindigo (3) has higher
photostability, a much shorter excited-state lifetime, and
a weaker fluorescence quantum yield than dimethoxythioin-
digo (4). The excited electronic state of both compounds,
592
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 591 –594

Angewandte
Chemie
which is reached after 8 ps, shows very similar absorption and
emission spectra. The major difference in the two molecules is
the lifetime of this state. No indication of a long-lasting
absorption change is found for compound 3. Apparently
isomerization at the central double bond of the molecule does
not take place. Since compounds 3 and 4 differ only by the
substitution at positions 4 and 4’, the hydroxy group and
a transient proton transfer as suggested for indigo should play
a major role in the differences in photostability and reaction
dynamics. The nearly identical features of the excited-state
spectra of the two compounds suggest that the first relaxation
out of the Franck–Condon region leads to similar intermedi-
ate states. A reaction along the proton-transfer coordinate of
compound 3 should only occur afterwards and mediates the
relaxation out of the S1 state. For dimethoxythioindigo (4) the
methyl group prevents this reaction path and relaxation to the
ground state occurs predominantly by fluorescence emission.
The long lifetime of the S1 state may also increase the
importance of other, destructive decay routes in compound 4.
To support these qualitative interpretations, quantum-
chemical computations were performed on suitable model
molecules (for details see the Supporting Information). For
both model molecules a conical intersection with the ground-
state surface is found, located slightly below the energy at the
Franck–Condon geometry. To reach this conical intersection
the molecules must undergo a proton/methyl cation transfer
and an out-of-plane deformation of the phenyl ring. The
methyl cation transfer (red line), however, involves a large
barrier (TSScan), which obstructs access to this conical
intersection. In contrast, the proton transfer (blue line)
occurs without a barrier in the model system and exhibits
only a small barrier in the full system (compound 3). Thus
theoretical modeling indicates that both molecules 3 and 4
leave the Franck–Condon range in the direction of a nearby
local minimum (MinS1) on the S1 potential surface due to the
rearrangement of the p system. Only for compound 3 is the
conical intersection accessible after passing the global mini-
mum on S1 (MinPS1; see Figure 2). The shape of the potential
energy surfaces in the vicinity of this conical intersection
guides the molecule back to the planar structure. Back
transfer of the proton from a first and shallow local ground-
state minimum (MinPS0) completes the photophysical process.
For compound 4 the large barrier in the S1 state prohibits
access to the conical intersection, the excited electronic state
becomes long-lived, and a high fluorescence quantum yield
results.
Figure 2. Critical points of the ground- and excited-state potential
energy surfaces for the photoreaction of model systems of 3 and 4.
The process is illustrated along two reaction coordinates, the proton
transfer or methyl cation transfer and the out-of-plane deformation of
the phenyl ring. Although both conical intersections (CoIn) are
situated energetically below the Franck–Condon point (FC), the CoIn
can only be reached by compound 3. The excited-state intramolecular
proton/methyl+ transfer is prerequisite to reach the CoIn. In com-
pound 4 the related methyl cation transfer is hindered by a substantial
barrier.
hemithioindigo moiety improves photostability. To prove this
point the monomethoxy derivative of thioindigo (5) was
synthesized by partial hydrolysis of 4. Indeed, the photo-
stabilty of 5 in CHCl3 was comparable to that of dihydrox-
ythioindigo (3). No photo-isomerization was found and the
fluorescence emission was as weak as that of 3 (see the
Supporting Information). The basic promoting mechanism—
excited-state proton transfer—is the same as that in indigo
and in many UV stabilizers (see Ref. [18] for the photo-
chemistry of highly photostable compounds). However, the
results on 3 show that suitably arranged proton donating and
accepting groups promote excited-state proton transfer even
if they are within rigid ring systems. This concept can serve as
an interesting strategy towards new photostable dyes for the
visible part of the spectrum.
A direct consequence of the long lifetime of the excited
electronic state of compound 4 is the possibility that slow and
destructive reaction channels in the nanosecond time domain
may gain importance. Another implication may be photo-
decomposition induced by the absorption of another photon
during the long population of the S1 state. The possibility of an
excited-state proton transfer in compound 3 opens a new path
for internal conversion, shortens the lifetime of the excited
electronic state, and finally leads to the observed large gain in
photostability.
Received: August 9, 2013
Revised: September 27, 2013
Published online: November 26, 2013
Keywords: dyes/pigments · photochemistry · proton transfer ·
.
thioindigo · ultrafast spectroscopy
[1] A. Baeyer, A. Emmerling, Ber. Dtsch. Chem. Ges. 1870, 3, 514 –
517.
The vastly different photochemical properties of the two
thioindigo compounds 3 and 4 demonstrate that excited-state
proton transfer within the rigid ring system of only one
[2] M. Irimia-Vladu, E. D. Glowacki, P. A. Troshin, G. Schwabeg-
ger, L. Leonat, D. K. Susarova, O. Krystal, M. Ullah, Y. Kanbur,
Angew. Chem. Int. Ed. 2014, 53, 591 –594
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
593

.
Angewandte
Communications
M. A. Bodea, V. F. Razumov, H. Sitter, S. Bauer, N. S. Sariciftci,
Adv. Mater. 2012, 24, 375 – 380.
[9] G. Haucke, R. Paetzold, Nova Acta Leopoldina, Supplement 11,
1978.
[10] W. Madelung, O. Wilhelmi, Ber. Dtsch. Chem. Ges. 1924, 57,
234 – 241.
[11] G. M. Wyman, J. Chem. Soc. D 1971, 1332 – 1334.
[12] T. Elsaesser, W. Kaiser, W. Lꢁttke, J. Phys. Chem. 1986, 90,
2901 – 2905.
[13] I. Iwakura, A. Yabushita, T. Kobayashi, Chem. Phys. Lett. 2010,
484, 354 – 357.
[14] S. Yamazaki, A. L. Sobolewski, W. Domcke, Phys. Chem. Chem.
Phys. 2011, 13, 1618 – 1628.
[15] P. Friedlꢀnder, Ber. Dtsch. Chem. Ges. 1916, 49, 955 – 966.
[16] C. Mukherjee, A. De, Synlett 2002, 325 – 327.
[17] C. Mukherjee, S. Kamila, A. De, Tetrahedron 2003, 59, 4767 –
4774.
[18] P. F. McGarry, S. Jockusch, Y. Fujiwara, N. A. Kaprinidis, N. J.
Turro, J. Phys. Chem. A 1997, 101, 764 – 767.
[3] N. Regner, T. T. Herzog, K. Haiser, C. Hoppmann, M. Beyer-
mann, J. Sauermann, M. Engelhard, T. Cordes, K. Rꢁck-Braun,
W. Zinth, J. Phys. Chem. B 2012, 116, 4181 – 4191.
[4] T. Fehrentz, M. Schçnberger, D. Trauner, Angew. Chem. 2011,
123, 12362 – 12390; Angew. Chem. Int. Ed. 2011, 50, 12156 –
12182.
[5] M. Klessinger, W. Lꢁttke, Tetrahedron 1963, 19 (Suppl. 2), 315 –
335.
[6] W. Lꢁttke, H. Hermann, M. Klessinger, Angew. Chem. 1966, 78,
638 – 639; Angew. Chem. Int. Ed. Engl. 1966, 5, 598 – 599.
[7] W. R. Brode, E. G. Pearson, G. M. Wyman, J. Am. Chem. Soc.
1954, 76, 1034 – 1036.
[8] G. M. Wyman, W. R. Brode, J. Am. Chem. Soc. 1951, 73, 1487 –
1493.
594
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 591 –594