Proton-and redox-controlled switching of photo-and electrochemiluminescence in thiophenyl-substituted boron-dipyrromethene dyes

Article abstract of DOI:10.1002/chem.200500729

A luminescent molecular switch in which the active thiol/disulfide switching element is attached to a meso-phenyl-substituted boron-dipyr-romethene (BDP) chromophore as the signalling unit is presented. The combination of these two functional units offers great versatility for multimodal switching of luminescence: 1)deprotonation/protonation of the thiol/thiolate moiety allows the highly fluorescent meso-p-thiophenol-BDP and its non-fluorescent thiolate analogue to be chemically and reversibly interconverted, 2) electrochemical oxidation of the monomeric dyes yields the fluorescent disulfide-bridged bichromophoric dimer, also in a fully reversible process, and 3) besides conventional photoexcitation, the well separated redox potentials of the BDP also allow the excited BDP state to be generated electro-chemically (i.e., processes 1) and 2) can be employed to control both photo-and electrochemiluminescence (ECL) of the BDP). The paper introduces and characterizes the various states of the switch and discusses the underlying mechanisms. Investigation of the ortho analogue of the dimer provided insight into potential chromophore-chromophore interactions in such bichromophoric architectures in both the ground and the excited state. Comparison of the optical and redox properties of the two disulfide dimers further revealed structural requirements both for redox switches and for ECL-active molecular ensembles. By employing thiol/disulfide switching chemistry and BDP luminescence features, it was possible to create a prototype molecular ensemble that shows both fully reversible proton-and redox-gated electrochemiluminescence.

Full text of DOI:10.1002/chem.200500729

FULL PAPER
DOI: 10.1002/chem.200500729
Proton- and Redox-Controlled Switchingof Photo- and
Electrochemiluminescence in Thiophenyl-Substituted Boron–
Dipyrromethene Dyes
[b]
Holger Rçhr,[a, c] Christian Trieflinger,[b] Knut Rurack,*[a] and JçrgDaub*
Abstract:
A
luminescent molecular
dimer, also in a fully reversible process,
and 3) besides conventional photoexci-
tation, the well separated redox poten-
tials of the BDP also allow the excited
BDP state to be generated electro-
chemically (i.e., processes 1) and 2) can
be employed to control both photo-
and electrochemiluminescence (ECL)
of the BDP). The paper introduces and
characterizes the various states of the
switch and discusses the underlying
mechanisms. Investigation of the ortho
analogue of the dimer provided insight
into potential chromophore–chromo-
phore interactions in such bichromo-
phoric architectures in both the ground
and the excited state. Comparison of
the optical and redox properties of the
two disulfide dimers further revealed
structural requirements both for redox
switches and for ECL-active molecular
ensembles. By employing thiol/disulfide
switching chemistry and BDP lumines-
cence features, it was possible to create
a prototype molecular ensemble that
shows both fully reversible proton- and
redox-gated electrochemiluminescence.
switch in which the active thiol/disul-
fide switching element is attached to a
meso-phenyl-substituted boron–dipyr-
romethene (BDP) chromophore as the
signalling unit is presented. The combi-
nation of these two functional units
offers great versatility for multimodal
switching of luminescence: 1) deproto-
nation/protonation of the thiol/thiolate
moiety allows the highly fluorescent
meso-p-thiophenol-BDP and its non-
fluorescent thiolate analogue to be
chemically and reversibly interconvert-
ed, 2) electrochemical oxidation of the
monomeric dyes yields the fluorescent
Keywords: boron–dipyrromethene ·
electrochemiluminescence · fluores-
cence · molecular switches · redox
chemistry · thiol/disulfide chemistry
disulfide-bridged
bichromophoric
Introduction
to access such information harbours the advantages of being
nondestructive, noninvasive and sensitive, and offers remote
access to a generated optical event. Examples of functional
supramolecular systems communicating through lumines-
cence include molecular-scale sensors,[1] switches,[2] “logic
gates”,[3] motors and machines,[4] wires,[5] or arrays, cascades
and cassettes that operate through energy- or electron-trans-
fer processes.[6] Besides the constant motivation to improve
the features that govern the performance of such systems in
terms of efficiency, functional supramolecules that can be
addressed or controlled through two or more different
inputs while delivering an optical output signal have recent-
ly focussed research efforts. Such advanced architectures
allow independent control of the various states of the
system and can be achieved by combinations of, for instance,
redox and photochromic units,[7] chemically sensitive and
redox units,[8] photochromic and chemically sensitive units[9]
or a combination of all the three types of stimuli-responsive
actions.[10]
The development of functional molecules capable of per-
forming chemically or physically controlled actions and re-
porting on or transducing these through luminescence sig-
nals has attracted considerable attention in recent years.
Employment of luminescence as the technique with which
[a] Dipl.-Phys. H. Rçhr, Dr. K. Rurack
Div. I.3, Bundesanstalt für Materialforschung und -prüfung (BAM)
Richard-Willstätter-Strasse 11, 12489 Berlin (Germany)
Fax : (+49)30-8104-5005
E-mail: knut.rurack@bam.de
[b] Dr. C. Trieflinger, Prof. J. Daub
Institut für Organische Chemie, Universität Regensburg
Universitätsstrasse 31, 93053 Regensburg (Germany)
Fax : (+49)941-943-4984
E-mail: joerg.daub@chemie.uni-regensburg.de
[c] Dipl.-Phys. H. Rçhr
Associated with: Fachbereich Biologie, Chemie, Pharmazie
Freie Universität Berlin, Takustrasse 3, 14195 Berlin (Germany)
A particularly attractive, intrinsically dual-mode switching
unit is the thiol group. This simple chemical group can be
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
689

deprotonated (i.e., chemically controlled) and can also be
oxidized to form disulfide bridges. Both reactions are rever-
sible (i.e., the thiolate can be reprotonated and the disulfide
bond can be broken by reduction). Although it is straight-
forward and versatile in nature, examples of switchable lu-
minescent molecules based on this control unit have, to the
best of our knowledge, yet to be reported.[11] This is surpris-
ing, as the thiol/disulfide redox pair plays an outstanding
role in protein biochemistry, most importantly as redox-
active switches through the two oxidation states of gluta-
thione or other cysteine-containing proteins.[12] Moreover,
the unravelling of the detailed mechanisms of natureꢁs thiol
redox chemistry has experienced significant progress in the
last years, basically because redox-active fluorescent tags
have become available.[13,14] These have been engineered by
attaching cysteine moieties to green fluorescent protein[13] or
yellow fluorescent protein.[14] Despite the results of first at-
tempts toward the design of artificial photochromic and flu-
orescent redox labels for cysteine-containing proteins,[15]
thiol/disulfide-active redox switches are assumed to be
promising candidates for biomolecular nanoelectronics.[16]
Here we present a simple luminescent molecular switch in
which the active thiol/disulfide element is attached to a
meso-phenyl-substituted boron–dipyrromethene (BDP)
chromophore as the signalling unit. BDP dyes possess valua-
ble spectroscopic features, such as intense absorption (e>
80000mꢀ1 cmꢀ1) and emission (Ff >0.70) bands in the visible
spectral range above 500 nm.[17] These dyes are readily solu-
ble in a large variety of solvents of different polarity, and
tuning of their spectroscopic properties towards the near in-
frared region is facile.[8a,18] If they contain an appropriate
unit—especially in the meso-position—that is either redox-
active or sensitive toward chemical inputs, such dyes can
show exceptionally powerful switching features, commonly
manifested in dramatic changes in fluorescence intensity or
lifetime.[19,20] The combination of thiol/disulfide chemistry
and BDP photophysics thus promised distinct ON/OFF be-
haviour through simple and rapid adjustment and intercon-
version of the chemical or physical states of the active con-
trol unit. Moreover, since the BDPsꢁ green emission of high
brightness can be generated not only photochemically but
also electrochemically,[8c,21] another stimulus for the research
reported here was the construction of a prototype molecular
system capable of displaying chemically and redox-modulat-
ed electrochemiluminescence (ECL).
Abstract in German: In dieser Arbeit wird ein lumineszieren-
der molekularer Schalter vorgestellt,bei dem das aktive
Thiol/Disulfid-Schaltelement an einen meso-phenylsubsti-
tuierten Bordipyrromethen (BDP) Farbstoff als Signal ge-
bende Komponente gekoppelt ist. Die Kombination dieser
beiden funktionellen Einheiten offeriert eine hohe Vielseitig-
keit für das multimodale Schalten der Lumineszenz: 1) die
Deprotonierung/Protonierung der Thiol/Thiolat-Gruppe er-
laubt es,das stark fluoreszierende meso-p-Thiophenol-BDP
und das analoge,nicht fluoreszierende Thiolat chemisch re-
versibel ineinander zu überführen,2) die elektrochemische
Oxidation der monomeren Farbstoffe ergibt das fluoreszier-
ende,disulfidverbrückte,bichromophore Dimer—dieser
Prozess ist ebenfalls vollständig reversibel,und 3) neben der
konventionellen Anregung mit Licht erlauben die deutlich ge-
trennten Redoxpotenziale des BDPs ebenfalls eine elektro-
chemische Generierung des angeregten BDP-Zustandes,d.h.,
Prozesse 1) und 2) kçnnen dazu eingesetzt werden,sowohl
die Photo- als auch die Elektrochemilumineszenz (ECL) des
BDP zu steuern. Diese Arbeit stellt die Charakteristika der
verschiedenen Zustände des Schalters vor und diskutiert die
zu Grunde liegenden Mechanismen. Die Untersuchung des
ortho-Analogen des Dimers vermittelte zudem Einsicht in
mçgliche Chromophor–Chro-
The thiophenyl-substituted BDP derivatives we studied in
this work are illustrated here. Compounds pSH-1 and (pS-
mophor-Wechselwirkungen im
Grund- wie im angeregten Zu-
stand von solchen bichromo-
phoren Architekturen. Der Ver-
gleich der optischen und Redox-
eigenschaften der beiden Disul-
fid-Dimere gab des Weiteren
Aufschluss über strukturelle
Voraussetzungen von Redox-
schaltern sowie ECL-aktiven
molekularen Ensembles. Unter
Einsatz schaltbarer Thiol/Disul-
fid-Chemie und mit den Lumi-
neszenzeigenschaften der BDPs
war es mçglich,ein erstes mole-
kulares Ensemble zu entwickeln,
das vollständig reversible proto-
nen- als auch redoxgesteuerte
Elektrochemilumineszenz zeigt.
690
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700

Proton- and Redox-Controlled Switching of Photo- and Electrochemiluminescence
FULL PAPER
1)2 were designed as the principle redox pair for multimode
switching of fluorescence and ECL, and the main intention
of this paper is to characterize and demonstrate the versatil-
ity of the system. To achieve better understanding of the
structural features of such compounds, the structurally more
demanding twin (oS-1)2 was also synthesized and investigat-
ed. With recent advances in protein labelling through thiol
functions[22] and the according future potential of pSH-1 to
act as a redox-active fluorescent tag in biosensory chemistry,
a comparative study of (pS-1)2 and (oS-1)2 also promised in-
sight into possible interchromophore interaction. These
might play an important role if, for instance, two BDP
labels were to come into close contact in a hydrophobic
pocket or on the surface of a protein, as recently shown by
Bergstrçm et al.[23]
Results and Discussion
Synthesis: The BDP-appended disulfide (pS-1)2 was ob-
tained from 4,4’-dithiobisbenzoic acid by conventional BDP
synthesis procedures (Scheme 1A),[10a] via in situ formation
of the acid chloride, while (oS-1)2 was prepared in a similar
way from 2,2’-dithiobisbenzoic acid. Reduction of (pS-1)2 af-
forded pSH-1 in 50% yield (Scheme 1A).[24] The synthesis
Figure 1. Normalized absorption and relative fluorescence spectra of
pSAc-1 (top), (pS-1)2 (middle) and (oS-1)2 (bottom) in MeCN (solid
line) and hexane (dotted line). The characteristic shoulders in the spectra
of reference compound pSAc-1 is depicted in Scheme 1B.
ꢀ
are due to the C C frame vibration indicated by the arrows in the top
Absorption and fluorescence spectroscopy: The para- and
ortho-thiophenyl-substituted BDPs were investigated both
by absorption and by steady-state and time-resolved fluo-
rimetry in their oxidized disulfide and reduced thiol forms
in a variety of solvents covering a large polarity range (from
hexane to DMSO). Compounds pSAc-1 and 1 were included
in the studies as models.
panel. Inset: Lowest-energy absorption band of (oS-1)2 in hexane con-
verted to the energy scale and results of the spectral deconvolution pro-
cedure.
internal decoupling of the chromophoric subunits (BDP and
phenyl moieties) and thus the weak influence of the pres-
ence or absence of para- or ortho-substituents. The full
widths
at
half-maximum
(fwhms) of the typical BDP
bands are virtually identical for
all the dyes, slightly increasing
from—for
instance—ca.
740 cmꢀ1 in hexane to ca.
790 cmꢀ1 in DMSO for (pS-1)2
and pSAc-1 and from 740 to
820 cmꢀ1 in the two solvents for
(oS-1)2. Besides these charac-
teristics,
the
cyanine-type
nature of the BDP chromo-
phore is further manifested in
the molecular C-C frame vibra-
tion typical of cyanine dyes[25]
Scheme 1. A) Synthetic route to (pS-1)2 and pSH-1. B) Synthetic route to pSAc-1.
at ca. 1300 cmꢀ1 and denoted by
arrows in Figure 1. These char-
acteristic features of BDP dyes
Independent of solvent polarity, the absorption maxima of
pSH-1, (pS-1)2, pSAc-1 and 1 show typical BDP features
and are centred at 525ꢁ2nm (Figure 1). The absorption
spectra of (oS-1)2 possess similar shapes and are slightly red-
shifted by 5 nm. These negligible differences indicate strong
have recently also been found by us for more extended
BDP-type p systems.[18b] The molar absorptivities support
the presence of two independent BDP moieties in the
dimers (i.e., emax ꢂ60000mꢀ1 cmꢀ1 for pSH-1 in all the sol-
vents studied and emax ꢂ119000mꢀ1 cmꢀ1 and 102000 mꢀ1 cmꢀ1
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
691

K. Rurack, J. Daub et al.
for (pS-1)2 and (oS-1)2, respectively). The steady-state emis-
sion features are consistent with these observations. The typ-
ical BDP emission of mirror-image shape is observed in all
cases, with maxima occurring at 538ꢁ2nm for pSH-1, (pS-
1)2, pSAc-1 and 1 and at 546ꢁ2nm for ( oS-1)2. Moreover,
the emission bands show fwhms identical to those of the ab-
sorption data within ꢁ10 cmꢀ1. No dual fluorescence was
observed during these studies, indicating the absence of the
formation of emissive charge-transfer states.[17b,18b,26]
With respect to pSH-1, these findings are consistent with
our previous report on a meso-(4-hydroxyphenyl)-substitut-
ed BDP in reference [27]. This analogue, with an OH group
instead of the SH group in pSH-1, also shows bright fluores-
cence with the typical BDP characteristics in its undissociat-
ed state in various solvents. While these straightforward
spectral characteristics are favourable for the facile opera-
tion of an optical switching system, what is essential for the
performance of a proton- and redox-active fluorescent
switch is controlled ON/OFF behaviour of the distinct
switching states. Ideally, in a two-state system, these states
should show dramatic differences in their fluorescence in-
tensities. In accordance with the work reported in refer-
ence [27], we found similar behaviour for pSH-1 and its de-
protonated form pS-1ꢀ.[28] Whereas the former, for instance,
emits green fluorescence with a high quantum yield compa-
rable to that of the model compound pSAc-1 (Table 1), the
rization of the entire chromophore, exemplified by a reduc-
tion in the interannular twist angle qcalcd from 898 in pSH-1
to 558 in pS-1ꢀ. This conformational effect might then give
rise to an oscillator-weak charge-transfer (CT) transition, lo-
cated at the low-energy side of the intense BDP band.[31]
The spectroelectrochemical measurements reported below
indeed support such an assumption, as a weak shoulder ap-
pears in the region around 570 nm during the first reduction
step of (pS-1)2 where pS-1ꢀ is formed (see left panel of
Figure 4 and Scheme 4; changes similar to those in the left
panel of Figure 4 were also observed during the titration of
pSH-1 with base). Deactivation of an ionic CT state then
most probably proceeds in radiationless fashion by the de-
scribed charge-shift mechanism. Furthermore, in the current
system, the fluorescence could be reversibly switched OFF
and ON by successive addition of base (e.g., 1,8-
diazabicyclo[5.4.0]undec-7-ene, DBU) and acid (e.g. tri-
fluoroacetic acid, TFA), respectively. The proton-induced
action is further confirmed by the fact that (pS-1)2 and
pSAc-1 do not show any spectroscopic changes upon base
addition.
For the goal of redox control over the switching features
of pSH-1, the fluorescence output of the oxidized disulfide
dimer (pS-1)2 is also important, as it presents the third cor-
nerstone in the cycle (Scheme 2). The data for the dimer are
Table 1. Spectroscopic data for the compounds studied in selected sol-
vents at 298 K; for additional data see Table S1 in the Supporting Infor-
mation.
Compound
Solvent
labs(max)
lem(max)
Ff
tf
[nm]
[nm]
[ns]
pSH-1[a]
pS-1ꢀ[a]
(pS-1)2
MeCN
MeCN
MeCN
THF
hexane
MeCN
THF
hexane
MeCN
THF
hexane
MeCN
hexane
522
518
524
526
527
530
533
532545
523
526
526
521
524
536
n.d[b]
537
539
539
544
547
0.64
<10ꢀ4
0.52[d]
4.72
[c]
3.44[d]
4.17
0.72
0.67
0.067
0.73
3.78
Scheme 2. Illustration of the switching states and processes of the title
compounds.
[e]
(oS-1)2
pSAc-1
1
[e]
[e]
0.78
537
539
538
535
537
0.72
0.72
0.73
0.87
0.82
4.68
4.11
4.02
5.25
4.77
included in Table 1. It is apparent that (pS-1)2 shows consid-
erable fluorescence with the typical BDP features in sol-
vents of any polarity, suggesting that the phenyl–disulfide–
phenyl bridge is large enough to prevent inter-BDP chromo-
phore communication. Results obtained recently by Wang
et al. on symmetric disulfide triads with the same spacer and
two naphthalimidyl fragments as chromophores support our
observations.[32] Nonetheless, the fluorescence quantum yield
of (pS-1)2 is reduced in relation to pSH-1, which is conceiva-
ble in view of the higher donor potential of the sulfur atom
in the latter compound.
For a three-state switch as shown in Scheme 2, with pSH-
1 and pS-1ꢀ as the proton-addressable and (pS-1)2 as the ad-
ditional redox partners, the moderately diminished fluores-
cence of the oxidized compound in relation to the reduced
form is important. In particular, the different fluorescence
[a] Measured only in degassed MeCN, because of rapid autooxidation in
the presence of air. [b] Not determined. [c] <3 ps. [d] Virtually identical
in degassed MeCN. [e] See description in the text and data in Table 2.
latter is virtually nonfluorescent. The distinctly higher donor
strength of the thiolate group obviously opens up an effi-
cient quenching pathway, most probably through rapid
back-electron transfer with charge-shift nature. Such nonra-
diative transitions to the ground state are well known for
ionic dye systems.[29] In the present case, preliminary quan-
tum chemical calculations at semiempirical level[30] suggest
that deprotonation might result in a certain degree of plana-
692
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700

Proton- and Redox-Controlled Switching of Photo- and Electrochemiluminescence
FULL PAPER
decay times—4.72ns for pSH-1 and 3.44 ns for (pS-1)2—
make it possible to distinguish between the two forms with
any commercial fluorescence lifetime instrumentation that
can be operated in the picosecond time domain. All the
states of the switch can thus be accessed by simple spectro-
scopic experiments.
Before we discuss the electrochemical side of the chemi-
cally and redox-active switching system, we will remain with
the dynamic fluorescence properties of the dimers and com-
pare the behaviour of (pS-1)2 and (oS-1)2 in more detail. As
mentioned in the Introduction, an important aspect in the
use of dyes as fluorescent tags in labelling or imaging appli-
Figure 2. Representative lifetime distribution functions of (oS-1)2 in
DMSO (solid, minor mode indicated by the arrow) and THF (dashed),
as well as (pS-1)2 in MeCN (dotted), obtained for 494/564 nm excitation
and emission wavelengths. For data, see Table 2.
cations is the behaviour of the dye under high-loading con-
ditions or under circumstances in which binding sites are sit-
uated in close proximity to one another. In both cases, inter-
chromophore communication can arise and can produce
either aggregation effects or fluorescence quenching. The
latter can be particularly important when such closely locat-
ed binding sites are contained within a hydrophobic pocket
of, for example, a protein, which might then facilitate ac-
commodation and cross-talk of (two) fluorophores. Firstly, it
is important to note that we did not observe any other fea-
tures for concentrated solutions of (pS-1)2 beyond those re-
ported in Table 1. Secondly, as mentioned above, the spec-
tral characteristics of (oS-1)2 are virtually identical to those
of the other BDPs investigated here.[33] However, the data
for (oS-1)2 in Tables 1 and S1 further reveal that the fluores-
cence quantum yield decreases as a function of solvent po-
larity for solvents more polar than THF and that the fluores-
cence of (oS-1)2 is considerably lower than that of (pS-1)2 in
solvents of high polarity. Moreover, whereas (pS-1)2 exhibits
single-exponential fluorescence lifetimes throughout the
spectrum of solvents studied and (oS-1)2 largely shows mon-
oexponential fluorescence decay kinetics in apolar solvents,
strongly nonexponential decays are found for (oS-1)2 in all
the polar solvents. This concomitant reduction of Ff and
tf—when the latter is expressed as the average lifetime—
points to activation of a nonradiative channel as the quench-
ing process rather than a change of the emitting state to one
with a lower emissivity. Apparently, whereas weakly polar
solvents produce sufficient isolation of both BDP units to
prevent any kind of pronounced interaction, in polar sol-
vents fluorescence quenching occurs, most probably through
chromophore–chromophore interaction within the dimer.
Because the spectral absorption and emission features of
(oS-1)2 are independent of the solvent and remain BDP-
like, H- or J-aggregate formation is less likely to be the
cause of the quenching. In this case, new bands at the high-
emission band. To acquire access to the underlying photo-
physical mechanisms it is important to keep in mind that
only the typical BDP emission band is observed (Figure 1)
and that these nonexponential decays can thus only be at-
tributed to emitting (monomeric) BDP moieties. However,
the occurrence of three distinct excited state species for a bi-
chromophoric molecule seems rather unusual. If we consider
further that the formation of intramolecular sandwich-type
conformations—presumably preexisting in the ground
state—-is possible for (oS-1)2, it might be helpful to invoke
another formalism for the analysis of the fluorescence
decays: lifetime distribution analysis (LDA; Figure 2). LDA
is based on the assumption that an ensemble of molecular
emitters showing a quasicontinuous distribution of lifetimes
or reaction rates is present in the solution.[35] In contrast to
distinct fluorophores in liquid solution, such an ensemble
might constitute of molecules in confined media such as
polymers, zeolites, or inclusion complexes in which the mi-
croenvironment of the discrete emitters is heterogeneous.[36]
Other examples in which this strategy has successfully been
applied to recover the decay kinetics include surface-bound
fluorophores[37] or flexible bichromophoric systems in which
the relative orientation of the two chromophores can vary
around one or more certain preferred conformation(s).[38]
The molecular structure of (oS-1)2, which is designed in
such a fashion that the two chromophore units can adopt a
sandwich-type conformation, indicates that this dye might
fall into the latter category (Scheme 3). In LDA of the
decays of such systems, the envelope of a lifetime distribu-
tion function usually shows several peaks, corresponding to
the average lifetimes <t> (or centres of gravity) of certain
preferred conformations or environmental situations, and
each of the peaks is characterized by a specific width s. The
width of the function, then, is a measure of the intramolecu-
lar spatial distribution of, or microheterogeneity around, the
emitting species.
!
or low-energy side of the S1 S0 band of the BDP spectrum
should occur.[23,34]
A closer look at the fluorescence decays of (oS-1)2 in
polar solvents reveals that analysis of the decay traces only
yields acceptable fits when three discrete exponentials are
employed. Furthermore, these three components are only
found as decay times, and their relative amplitudes remain
virtually constant when monitored at three different emis-
sion wavelengths between 540 and 600 nm, over the entire
The data in Table 2show two trends. A bimodal distribu-
tion with a centre at ca. 0.5 ns and a second centre in the 5–
6 ns range, the latter being largely independent of the sol-
vent, is found in the region of high solvent polarity. In the
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
693

K. Rurack, J. Daub et al.
pothesis of sandwich formation, and due to the lack of rise
times, the behaviour of (oS-1)2 cannot be described in terms
of classical exciplex/excimer or charge-transfer formation
models, which are often found for bichromophoric com-
pounds, and the involvement of molecular aggregates is also
unlikely (vide ante). In contrast, we tentatively attribute
these results to certain prearrangements of the molecules al-
ready present in the ground state. If we assume that the
molecules can adopt folded conformations, such as the U-
form, that bring the two BDP moieties into closer proximity,
fast intramolecular fluorescence energy transfer can occur
upon excitation.[39] Actually, ground-state geometry optimi-
zation performed for (oS-1)2 at the AM1 level converged to
two different conformations: the U-shaped, quasi-“sand-
wich”-type structure and a more Z-shaped conformation
that resembles a zigzag analogue of (pS-1)2 (Scheme 3). The
dipole moments of both conformations were calculated to
be 6.6 D for the U- and 3.0 D for the Z-form, suggesting
that polar solvents might stabilize the U-form better. The ef-
ficiency of the quenching process then strongly depends on
the mutual preorientation and distance between the two
identical fluorophores, with the spatial distribution of the
apparent torsion angle around the disulfide bond (see cap-
tion of Scheme 3 for definition) in the U-form in particular
being able to entail broader lifetime distributions. Further-
more, the bimodality of the distributions in the highly polar
solvents suggests that both sets of conformations actually
seem to be present and play a role for the reduced fluores-
cence. Molecules in the Z-form, with a conformation that
more closely resembles (pS-1)2, may thus be responsible for
the slow decays of about 5–6 ns, whilst molecules in the U-
form—apparently decoupled in the ground state but suffi-
ciently preoriented for interaction in the excited state—
would accordingly be responsible for the fast decay times.
Another interesting feature of substituted BDP dyes, the
distinctly longer decay times of ortho-donor-phenyl substi-
tuted compounds in relation to their unsubstituted or para-
substituted analogues, is also evident from Table 2. Whereas
the fluorescence of pSAc-1 decays in 4.68 ns in MeCN, 1
shows a lifetime of 5.25 ns and (oS-1)2 a lifetime of 6.47 ns.
Similar observations have previously been made by us for a
1,1’-binaphthyl-appended bichromophoric BDP[20a] and for
2-methoxy-3-BDP-naphthalene, which show decay times of
6.17 and 6.55 ns in acetonitrile. These features are not yet
completely understood, and current theoretical and experi-
mental work is directed at more fundamental understanding.
A possible explanation might involve hindered rotation in
ortho-substituted derivatives, which might suppress a dark
deexcitation channel of the BDP chromophore through a
“butterfly”-type structure. Lindsey et al. recently postulated
that population of such a “dark state” involves coplanariza-
tion of the meso-substituent with a part of the dipyrrome-
thene ring system.[40,41]
Scheme 3. A) Two possible conformers of (oS-1)2 with maximum (U-
form) and minimum (Z-form) chromophore–chromophore interaction.
The apparent torsion angles between atoms B1-C4-C4’-B1’ (see B) ob-
tained for the two stable U- and Z-conformations of (oS-1)2 by ground-
state geometry optimization by the AM1 method are ꢀ618 for the U-
form and ꢀ1348 for the Z-form.
Table 2. Fluorescence lifetime data for (oS-1)2 in various solvents at
298 K; for additional data see Table S2 in the Supporting Information.[a]
Compound
Solvent
<t1 > [ns]
<t2 > [ns]
fr2el[b]
(oS-1)2
DMSO
MeCN
acetone
THF
hexane
MeCN
MeCN
5.23ꢁ1.00
6.47ꢁ0.74
3.29ꢁ2.35
5.44ꢁ1.52–
5.99ꢁ1.37
3.41ꢁ0.66
4.63ꢁ0.84
0.41ꢁ0.16
0.49ꢁ0.09
0.33ꢁ0.09
0.64
0.68
0.06
–
–
–
–
–
–
–
(pS-1)2
pSAc-1
[a] The data are averages from three independent fits of three decays re-
corded between 540 and 600 nm for two samples. [b] Relative amplitude
of the short decay mode <t2 >.
weak and medium polar solvents, monomodal distributions
are obtained. For the first group of highly polar solvents, the
relative contributions of the two modes vary as a function of
solvent polarity, reflecting the overall decrease in fluores-
cence quantum yield: fast decay times predominate in
DMSO and MeCN, while slow ones present the major mode
in acetone. In the less polar solvents, only the slow decay
times are found. For a better illustration, Figure 2shows
representative LDA results for (oS-1)2 in THF and DMSO.
In the less polar solvents, the width of the monomodal dis-
tribution of lifetimes of (oS-1)2 is still considerably broader
than that of the corresponding single exponential decays of
the para analogue (pS-1)2, as well as model pSAc-1
(Table 2). In accordance with the previously outlined hy-
Low-temperature measurements in the time domain at
liquid nitrogen temperature yielded single-exponential
decays for all the derivatives investigated. Compounds (pS-
1)2 and pSAc-1, which decay with 4.2and 4.8 ns lifetimes in
694
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700

Proton- and Redox-Controlled Switching of Photo- and Electrochemiluminescence
FULL PAPER
ethanol at 298 K, showed lifetimes of 6.1 and 6.2 ns at 77 K.
Furthermore, the bimodal distribution of (oS-1)2 in metha-
nol at 298 K is not maintained in the glass, in which the dye
shows a monoexponential decay with tf = 7.1 ns. These re-
sults indicate that any dark relaxation pathway that relies on
larger molecular motions is closed at 77 K.
[electron transfer–chemical step–electron transfer (or ECE)
mechanism; vide infra],[44,45] which is electrochemically re-
versed at ꢀ265 mV. As pS-1ꢀ is nonfluorescent, the ON and
OFF positions of the fluorescent switch can also be adjusted
electrochemically. The large peak displacement of disulfide
cleavage and formation, of ca. 1300 mV (DE
ꢂ30 kcalmolꢀ1), indicates the bistability of the (pS-1)2/pS-1ꢀ
couple over this potential range. Both forms can exist in this
potential range under similar conditions, the nature of the
species depending on the previous process [i.e., whether it
was formed by oxidation or reduction (higher or lower elec-
trochemical potential)]. This behaviour is closely related to
the memory effect of bistable photochromic molecules, an
important aspect in molecular data storage and molecular
energy storage.
Electrochemistry and spectroelectrochemistry: Returning to
the redox-active switching process at room temperature,
cyclic voltammetry clearly reveals that the thiol/disulfide
transition of the BDPs can be performed and controlled
electrochemically. Whereas pSAc-1 shows only the typical
reversible reduction and quasi-reversible oxidation of the
BDP moiety (ꢀ1600 mV and +620 mV vs. Fc+/Fc in
MeCN),[42] an additional redox process arises in the cyclic
voltammogram of (pS-1)2, with a reduction at ꢀ1545 mV
and a significantly displaced back oxidation at ꢀ265 mV
(Figure 3).[43] As can be shown by calibration against ferro-
The cyclic voltammogram of (oS-1)2 is very similar to that
of the para dimer, with the disulfide cleavage being facilitat-
ed by 45 mV (Table 3) because of steric crowding in the
Table 3. Electrochemical data of the compounds in MeCN/0.2m TBAH.
Er1ed BDP
Eo1 x BDP
Erped
Eopx
=
2
=
2
[mV]
[mV]
[mV]
[mV]
p-SAc-1
(pS-1)2
(oS-1)2
1
ꢀ1600
ꢀ1735
ꢀ1885
ꢀ1630
+620
+645
+690
+620
–
–
ꢀ1545
ꢀ1500
–
ꢀ265
ꢀ235
–
ortho derivative. Moreover, because of the close vicinity of
the thiolate group formed in the first reduction, the second
reduction on the BDP core is aggravated, occurring only at
ꢀ1885 mV (Table 3). With regard to a possible chromo-
phore–chromophore interaction in the ground state, the
cyclic voltammogram of (oS-1)2 [and also that of (pS-1)2]
does not show any splitting of the redox waves, thus stress-
ing the conclusion implied by the similarity of the absorp-
tion spectra (i.e., both BDP moieties are independent and
do not interact in the ground state).
To obtain a better understanding of the ECE mechanism
involved in the reduction of (pS-1)2 to pS-1ꢀ as described
above, as well as the reoxidation process, the cyclic voltam-
mogram of the para dimer was simulated[46] under the as-
sumptions detailed in Scheme 4. Electron transfer to the
diaryl disulfide subunit results in the dissociation of the di-
sulfide bridge in the intermediately formed (pS-1)2ꢀ·, yield-
ing pS-1ꢀ and the thio radical pS-1·. Under the prevailing
potential conditions, the latter species is instantly reduced to
a second pS-1ꢀ, so that the net reaction of the first reduction
is a transfer of two electrons. The corresponding formation
of the disulfide bond (i.e., the formation of (pS-1)2) pro-
ceeds in a first step from pS-1ꢀ by oxidation to pS-1·. Two of
these radicals can then directly combine to yield (pS-1)2, or
pS-1· and pS-1ꢀ can react to (pS-1)2ꢀ·, which is then oxidized
to the para dimer. The resulting calculated CV is compared
to the experimentally measured cyclic voltammogram in
Figure 5.
Figure 3. Cyclic voltammetry of (pS-1)2 and pSAc-1 in MeCN/0.2m
TBAH.
cene, each process represents a two-electron transfer for
(pS-1)2, due to its dimeric structure. The high similarity of
the UV/Vis spectra of reduced (pS-1)2 obtained in spectro-
electrochemical measurements (Figure 4) and of pS-1ꢀ, ob-
tained by deprotonation of pSH-1, strongly supports the at-
tribution of the first reduction step at ꢀ1545 mV to a disso-
ciative conversion of (pS-1)2 into two molecules of pS-1ꢀ
Figure 4. Spectroelectrochemistry of (pS-1)2. First reduction (left) and
second reduction to the typical BDP radical dianion (right) in MeCN/
0.2m TBAH.
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
695

K. Rurack, J. Daub et al.
the instability of the species in-
volved in the processes necessa-
ry to generate ECL.
Toward fluorescent tags: The
validity of p-thiophenyl-substi-
tuted BDPs as both chemically
and electrochemically addressa-
ble switches that can be con-
veniently operated in the visible
spectral range having been es-
tablished, recent progress in the
redox labelling of proteins, as
described in the Introduction,
suggests that the thiol title com-
pound pSH-1 might be a versa-
tile and promising candidate. In
the manner reported above, the
labelling reaction should be ac-
complishable by a redox step,
with the optical spectroscopic
features of the dye remaining
largely unperturbed [i.e., the
fluorescence properties should
change in accordance with the
Scheme 4. ECE mechanism of the first reduction of (pS-1)2 (top) and oxidation of pS-1ꢀ with subsequent di-
merization (bottom).
Figure 5. Simulated (dotted line) and measured (solid line) CVs of (pS-
1)2.
Figure 6. ECL spectra of (pS-1)2 in CH2Cl2/0.2m TBAH, with switch
times of 1 sꢀ1 (discrete lines) and 50 msꢀ1 (curve). Inset: corresponding
CV, with the applied potentials indicated by the arrows.
Electrochemiluminescence: Keeping in mind that electroac-
tive fluorophores are interesting candidates for several opto-
electronic applications, mainly with respect to organic light-
emitting devices (OLEDs), we investigated the electroche-
miluminescence behaviour of pSAc-1 and (pS-1)2.[8c,21,47,48]
step from pSH-1 to (pS-1)2]. Furthermore, the results pre-
sented above for (oS-1)2 indicate that a typical BDP fluores-
cence of moderate intensity (Ff ꢂ0.1) could still be moni-
tored even under high loading conditions, or if two labels
were in close proximity. This is a valuable criterion for imag-
ing applications. Additionally, in the case of pSH-1, two dif-
ferent steps could be invoked to remove background signals
from unreacted stains: either conventional washing out of
pSH-1 or, through simple pH adjustment, easy and quantita-
tive conversion of pSH-1 into the fluorimetrically silent
form pS-1ꢀ.
By applying alternating potentials in the range of Epred=ox
-
(BDP), we obtained BDP-luminescence in both cases (illus-
trated for (pS-1)2 in Figure 6).[49] Since the redox processes
of the thiol/disulfide couple, for instance, are between the
applied potential range for (pS-1)2, one can assume that the
excited state is generated by recombination of pS-12ꢀ and
(pS-1)22+, the mechanism being depicted in Scheme 5. The
impact of the dyeꢁs structure is evident from the fact that we
did not find ECL activity for (oS-1)2, presumably because of
To test the suitability of this system for a broader range of
labelling reactions, pS-1ꢀ was coupled to cyanuric chloride
to obtain a reactive BDP dye. Cyanuric chloride is widely
696
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700

Proton- and Redox-Controlled Switching of Photo- and Electrochemiluminescence
FULL PAPER
pSH-1 is virtually revived
during the reaction), and its
typical BDP absorption band is
bathochromically shifted by ca.
10 nm in relation to that of pS-
1ꢀ. Both dyes, pSH-1 and
pSCC-1, thus seem to show
promise as fluorescent labels,
and future work in our labora-
tories is currently being direct-
ed at detailed investigation of
their staining properties.
Conclusion
This work offers a new concept
of using the well investigated
BDP core as a proton- and
redox-addressable
molecular
switch, preserving all the fa-
vourable features of the family
of BDP dyes while exhibiting
full reversibility. With pSH-1,
its deprotonated, fluorescently
silent form pS-1ꢀ and the oxi-
dized dimer (pS-1)2, the chemi-
cally and electrochemically
switchable states of the system
are introduced and character-
ized. By employing thiol/disul-
fide switching chemistry, it was
possible to create a first molec-
ular ensemble that shows both
proton- and redox-controlled
electrochemiluminescence. The
potential fields of application of
the versatile molecular system
range from multimode molecu-
lar switching via chemically
and/or redox-gated OLEDs to
pH monitoring, protein label-
ling or other applications as a
reactive dye. The ortho dimer
(oS-1)2 allowed us to acquire
deeper insight into structural
requirements in redox switches,
in the sense that excessive prox-
imity of the redox centres pre-
cludes a potentially ECL-active
ꢀ
Scheme 5. Mechanism of ECL generation by annihilation in (pS-1)2. The involvement of (pS-1)2 as a minority
electron donor cannot be ruled out.
Scheme 6. Nucleophilic substitution of cyanuric chloride with pS-1ꢀ.
used as a cross-linking agent for coloration—in the textile
industry or for optical brighteners, for example—and has
long been used as a linker for reactive fluorescent stains.[50]
The reaction with pSH-1 is straightforward and proceeds
after base addition via the fluorimetrically silent pS-1ꢀ to
yield pSCC-1 as shown in Scheme 6. The exclusively formed
monoproduct is highly fluorescent (i.e., the fluorescence of
chromophore from showing luminescence. On the other
hand, the comparative investigation of the ortho- and the
para-dimers revealed that a simple meso-o-thiophenyl group
is a large enough spacer to prevent interchromophore com-
munication of bichromophoric BDP dyes in the ground
state, yet it retains the possibility for interaction in the excit-
ed state by energy transfer.
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
697

K. Rurack, J. Daub et al.
Compound (oS-1)2: A mixture of commercially available 2,2’-dithiobis-
benzoic acid (500 mg, 1.6 mmol) in SOCl2 (15.0 mL) was heated at reflux
until a homogenous solution was obtained (18 h). Excess SOCl2 was re-
moved under vacuum. The remaining solid was dissolved in CH2Cl2
(200 mL), and 3-ethyl-2,4-dimethylpyrrole (0.8 mL, 760 mg, 6.24 mmol)
was added. After the mixture had been heated at reflux for 4 h, BF3·OEt2
and ethyl-diisopropyl-amine (3 mL each) were added. After stirring for
30 min, the reaction mixture was quenched with water and extracted with
CH2Cl2. The organic layer was dried over anhydrous Na2SO4 and evapo-
rated. Column chromatography on silica gel with CH2Cl2 as an eluent
gave (oS-1)2 as a red solid with a green lustre (11% yield, 150 mg,
0.18 mmol). M.p.>3608C; 1H NMR (300 MHz, CDCl3): d = 7.67–7.62
(m, 2H), 7.20–7.16 (m, 4H), 2.52 (s, 12H; CH3), 2.31 (q, J = 7.6 Hz, 8H;
CH2), 1.25 (s, 12H; CH3), 1.00 (t, J = 7.6 Hz, 12H; CH3) ppm; 13C NMR
(75.5 MHz, CDCl3): d = 154.6 (q), 137.9 (q), 136.1 (q), 136.0 (q), 133.6
(q), 133.0 (q), 130.4 (q), 129.7 (+), 129.1 (+), 17.1 (ꢀ), 14.6 (+), 12.6 (+),
11.2( +) ppm; FTIR (KBr): n˜ = 2964, 2926, 1544, 1478, 1320, 1192, 1073,
980 cmꢀ1; MS (FD, 70 eV): m/z: 822 (100) [M]+; HRMS (EI, 70 eV):
calcd: 822.3771; found 822.3770.
Experimental Section
General: Melting points were recorded on a Reichert-Thermovar micro
melting apparatus and are not corrected. 1H NMR and 13C NMR spectra
were measured on an Avance 300 instrument. Mass spectra were record-
ed on a Varian CH-5 and a Finnigan MAT 95 machine. IR spectra were
obtained with a Biorad FTS 155 instrument in KBr disks.
Materials and synthesis: All solvents and chemicals were of reagent
grade quality, were obtained commercially and were used without further
purification except as noted below. The tetrabutylammonium hexafluoro-
phosphate (TBAH) used as a supporting electrolyte and acetonitrile and
dichloromethane for the electrochemical measurements were refined by
previously described procedures.[51]
Compound pSAc-1: A mixture of 4-acetylsulfanylbenzoic acid[52] (200 mg,
1.0 mmol) in SOCl2 (10.0 mL) was heated at reflux until a homogenous
solution was obtained (1 h). Excess SOCl2 was removed under vacuum.
The remaining solid was dissolved in CH2Cl2 (100 mL) and 3-ethyl-2,4-di-
methylpyrrole (0.3 mL, 285 mg, 2.34 mmol) was added. After the mixture
had been heated at reflux for 4 h, BF3·OEt2 and ethyl-diisopropyl-amine
(3 mL of each) were added. After stirring for 30 min, the reaction mix-
ture was quenched with water and extracted with CH2Cl2. The organic
layer was dried over anhydrous Na2SO4 and evaporated. Column chro-
matography on silica gel with CH2Cl2/ethyl acetate (1:1) as an eluent
gave pSAc-1 as a red solid with a green lustre (9% yield, 40 mg,
0.09 mmol). M.p.>222–2238C; 1H NMR (300 MHz, CDCl3): d = 7.55–
7.52 (m, 2H), 7.37–7.33 (m, 2H), 2.53 (s, 6H; CH 3), 2.46 (s, 3H; CH3),
2.30 (q, J = 7.6 Hz, 4H; CH2), 1.33 (s, 6H; CH3), 0.98 (t, J = 7.6 Hz,
6H; CH3) ppm; 13C NMR (75.5 MHz, CDCl3): d = 193.3 (q), 154.1 (q),
138.8 (q), 138.3 (q), 137.1 (q), 135.2( +), 133.0 (q), 130.5 (q), 129.2 (+),
128.9 (q), 30.3 (+), 17.1 (ꢀ), 14.6 (+), 12.5 (+), 11.8 (+) ppm; FTIR
(KBr): n˜ = 2968, 2926, 2856, 1717, 1536, 1474, 1320, 1193, 1119, 1065,
976, 760 cmꢀ1; MS (EI, 70 eV): m/z: 454 (100) [M]+; HRMS (EI, 70 eV):
calcd: 454.2066; found 454.2057.
Compound 1: This was synthesized as reported by us previously.[17a]
Steady-state absorption and fluorescence spectroscopy: Steady-state ab-
sorption measurements were carried out on
Omega 10 and a Cary 5000 UV/Vis/NIR spectrophotometer. The steady-
state fluorescence spectra were recorded with Spectronics Instru-
a Bruins Instruments
a
ment 8100 spectrofluorimeter. For all measurements, the temperature
was kept constant at 298ꢁ2K. Unless otherwise noted, only dilute solu-
tions of an optical density of less then 0.1 at the absorption maximum
were used. Fluorescence experiments were performed with a 908 stan-
dard geometry, with polarizers set at 54.78 for emission and 08 for excita-
tion. The fluorescence quantum yields (Ff) were determined relative to
fluorescein 27 in 0.1n NaOH (Ff = 0.90ꢁ0.03).[53] All the fluorescence
spectra presented here were corrected for the spectral response of the de-
tection system (calibrated quartz halogen lamp placed inside an integrat-
ing sphere; Gigahertz-Optik) and for the spectral irradiance of the exci-
tation channel (calibrated silicon diode mounted at a sphere port; Giga-
hertz-Optik). The uncertainties of the fluorescence quantum yields were
determined to ꢁ5% (for Ff >0.2) and ꢁ10% (for 0.2>Ff >0.02).
Spectral fittingprocedure : The full widths at half-maximum (fwhms) of
the reported absorption and emission bands were calculated from the de-
convoluted absorption and emission spectra, respectively, which had been
converted to the energy scale and subsequently fitted by a Gaussian fit.
Independent of the compound, five components were necessary in both
absorption and fluorescence spectra to yield acceptable fitting results
with regard to the lowest energy transition. The fwhms of the compo-
nents were linked and no other constrains were set.
Compound (pS-1)2: A mixture of commercially available 4,4’-dithiobis-
benzoic acid (500 mg, 1.6 mmol) in SOCl2 (15.0 mL) was heated at reflux
until a homogenous solution was obtained (30 h). Excess SOCl2 was re-
moved under vacuum. The remaining solid was dissolved in CH2Cl2
(200 mL), and 3-ethyl-2,4-dimethylpyrrole (0.8 mL, 760 mg, 6.24 mmol)
was added. After the mixture had been heated at reflux for 4 h, BF3·OEt2
and ethyl-diisopropyl-amine (3 mL each) were added. After stirring for
30 min, the reaction mixture was quenched with water and extracted with
CH2Cl2. The organic layer was dried over anhydrous Na2SO4 and evapo-
rated. Column chromatography on silica gel with CH2Cl2 as an eluent
gave (pS-1)2 as a red solid with a green lustre (11% yield, 150 mg,
0.18 mmol). M.p.>265–2678C; 1H NMR (300 MHz, CDCl3): d = 7.63–
7.60 (m, 4H), 7.27–7.24 (m, 4H), 2.52 (s, 12H; CH3), 2.28 (q, J = 7.5 Hz,
8H; CH2), 1.27 (s, 12H; CH3), 0.96 (t, J = 7.5 Hz, 12H; CH3) ppm;
13C NMR (75.5 MHz, CDCl3): d = 154.1 (q), 138.8 (q), 138.0 (q), 137.5
(q), 135.2(q), 133.0 (q), 130.6 (q), 129.2( +), 128.6 (+), 17.0 (ꢀ), 14.6
(+), 12.5 (+), 11.8 (+) ppm; FTIR (KBr): n˜ = 2964, 2929, 1544, 1474,
1320, 1193, 1073, 980 cmꢀ1; MS (FD, 70 eV): m/z: 822 (100) [M]+; HRMS
(EI, 70 eV): calcd: 822.3771; found 822.3767.
Time-resolved fluorescence spectroscopy: Fluorescence lifetimes (tf)
were determined by a unique customized laser impulse fluorimeter with
picosecond time resolution, which we have described in earlier publica-
tions.[18b,54] The fluorescence was collected at right angles (polarizer set at
54.78; monochromator with spectral bandwidths of 4, 8 and 16 nm) and
the fluorescence decays were recorded with a modular single-photon
timing unit as described in reference [18b]. While producing typical in-
strumental response functions of fwhm of ca. 25–30 ps, the time division
was 4.8 ps per channel and the experimental accuracy amounted to
ꢁ3 ps. The laser beam was attenuated with a double prism attenuator
(LTB) and typical excitation energies were in the nanowatt to microwatt
range (average laser power). The fluorescence lifetime profiles were ana-
lysed on a PC with the Global Unlimited V2.2 software package (Labora-
tory for Fluorescence Dynamics, University of Illinois). The goodness of
fit of the single decays as judged by reduced chi-squared (cR2) and the au-
tocorrelation function C(j) of the residuals was always below cR2 <1.2.
For all the dyes, decays were recorded at three different emission wave-
lengths over the BDP-type emission spectrum and were analysed global-
ly. Such a global analysis of decays recorded at different emission wave-
lengths implies that the decay times of the species are linked while the
program varies the preexponential factors and lifetimes until the changes
Compound pSH-1: A mixture of (pS-1)2 (60 mg, 0.07 mmol) and Zn
(2.0 g) was stirred in AcOH (20.0 mL). After ca. 2.5 h, excessive Zn was
removed by filtration, and water was added to the remaining solution.
After extraction with CH2Cl2, the organic layer was dried over anhydrous
Na2SO4 and evaporated. Column chromatography on silica gel with
CH2Cl2 as an eluent gave pSH-1 as a red solid with a green lustre (50%
1
yield, 30 mg, 0.07 mmol). M.p.>210–2128C; H NMR (300 MHz, CDCl3):
d = 7.39–7.37 (m, 2H), 7.16–7.13 (m, 2H), 3.59 (s, 1H), 2.52 (s, 6H;
CH3), 2.30 (q, J = 7.5 Hz, 4H; CH2), 1.33 (s, 6H; CH3), 0.98 (t, J =
7.5 Hz, 6H; CH3) ppm; 13C NMR (75.5 MHz, CDCl3): d
= 153.9 (q),
139.3 (q), 138.3 (q), 133.1 (q), 132.9 (q), 132.1 (q) 130.8 (q), 129.7 (+),
129.1 (+), 17.1 (ꢀ), 14.6 (+), 12.5 (+), 11.9 (+) ppm; FTIR (KBr): n˜ =
2968, 2929, 1532, 1474, 1316, 1185, 1069, 976, 803 cmꢀ1; MS (DCI, NH3):
m/z: 412(100)
412.1960.
[
M]+; HRMS (EI, 70 eV): calcd: 412.1949; found
in the error surface (c2 surface) are minimal; that is, convergence is
2
reached. The fitting results are judged for every single decay (local cR
)
698
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700

Proton- and Redox-Controlled Switching of Photo- and Electrochemiluminescence
FULL PAPER
and for all the decays (global cR2). The errors for all the global analytical
results presented here were below a global cR = 1.2.
[6] a) T. G. Goodson, III, Acc. Chem. Res. 2005, 38, 99–107; b) J. M. J.
Frechet, J. Polym. Sci. Part A: Polym. Chem. 2003, 41, 3713–3725;
c) V. Balzani, P. Ceroni, M. Maestri, C. Saudan, V. Vicinelli, Top.
Curr. Chem. 2003, 228, 159–191; d) M. A. Miller, R. K. Lammi, S.
Prathapan, D. Holten, J. S. Lindsey, J. Org. Chem. 2000, 65, 6634–
6649; e) A. Burghart, L. H. Thoresen, J. Chen, K. Burgess, F. Berg-
strçm, L. B.-Š. Johansson, Chem. Commun. 2000, 2203–2204;
f) J. Y. Ju, A. N. Glazer, R. A. Mathies, Nucleic Acids Res. 1996, 24,
1144–1148.
[7] a) G. M. Tsivgoulis, J.-M. Lehn, Adv. Mater. 1997, 9, 627–630; b) H.
Spreitzer, J. Daub, Chem. Eur. J. 1996, 2, 1150–1158; c) J. Daub, C.
Fischer, J. Salbeck, K. Ulrich, Adv. Mater. 1990, 2, 366–369.
[8] a) K. Rurack, M. Kollmannsberger, J. Daub, Angew. Chem. 2001,
113, 396–399; Angew. Chem. Int. Ed. 2001, 40, 385–387; b) G.
Hennrich, H. Sonnenschein, U. Resch-Genger, J. Am. Chem. Soc.
1999, 121, 5073–5074; c) M. Kollmannsberger, T. Gareis, S. Heinl, J.
Breu, J. Daub, Angew. Chem. 1997, 109, 1391–1393; Angew. Chem.
Int. Ed. Engl. 1997, 36, 1333–1335.
[9] a) L. Gobbi, P. Seiler, F. Diederich, V. Gramlich, C. Boudon, J.-P.
Gisselbrecht, M. Gross, Helv. Chim. Acta 2001, 84, 743–777; b) L.
Gobbi, P. Seiler, F. Diederich, Angew. Chem. 1999, 111, 737–740;
Angew. Chem. Int. Ed. Engl. 1999, 38, 674–678.
[10] a) C. Trieflinger, K. Rurack, J. Daub, Angew. Chem. 2005, 117,
2328–2331; Angew. Chem. Int. Ed. 2005, 44, 2288–2291; b) J.
Achatz, C. Fischer, J. Salbeck, J. Daub, J. Chem. Soc. Chem.
Commun. 1991, 504–507.
[11] A number of redox-active thiol/disulfide switches that can be
probed by various other instrumental methods have been reported;
see, for example: a) H. Graubaum, F. Tittelbach, G. Lutze, K. Gloe,
M. Mackrodt, J. Prakt. Chem. 1997, 339, 55–58; b) M. Irie, O. Miya-
take, K. Uchida, T. Eriguchi, J. Am. Chem. Soc. 1994, 116, 9894–
9900; c) S. Shinkai, K. Inuzuka, O. Miyazaki, O. Manabe, J. Org.
Chem. 1984, 49, 3440–3442; d) S. Shinkai, K. Inuzuka, O. Manabe,
Chem. Lett. 1983, 747–750; e) T. Minami, S. Shinkai, O. Manabe,
Tetrahedron Lett. 1982, 23, 5167–5170. An example of a thiol/disul-
fide-controlled molecular machine based on rotaxanes has also been
published recently, f) Y. Furusho, T. Hasegawa, A. Tsuboi, N.
Kihara, T. Takata, Chem. Lett. 2000, 18–19.
2
Lifetime distribution analysis (LDA): LDA of the fluorescence decay
traces that could not be sufficiently described by one or two exponentials
in the global analysis was performed with the FLA900 software package
(Edinburgh Analytical Instruments, level 2, version 1.6). The lifetime
range was set to a reasonable value, the starting channel shift was set to
0.5 (not fixed), all available channels were used, and the maximum possi-
ble number of 100 individual exponential lifetimes was employed. No
other constraints were made. The quality of the fit was again reviewed by
the c2 analysis. For all results c2 was below 1.3.
Cyclic voltammetry, spectroelectrochemistry, electrochemiluminescence:
CV measurements were performed with solutions of the appropriate
compound (ca. 1 mm) in highly pure solvents buffered with TBAH (0.2m,
vide ante) on a potentiostat/galvanostat (EG&G 283A). The measure-
ment cell had a three-electrode set-up (Pt working electrode, gold coun-
ter-electrode and Ag/AgCl pseudo-reference electrode) and the measure-
ments were referenced against ferrocenium/ferrocene (Fc+/Fc) as the in-
ternal standard. A Perkin–Elmer Lambda 9 UV/Vis/NIR spectrophotom-
eter in combination with an Amel 2053 potentiostat/galvanostat and a
custom-build quartz cuvette with a minigrid gold net as transparent
working electrode was employed for the spectroelectrochemical experi-
ments (for a detailed description see reference [55]). ECL measurements
were carried out on a Hitachi F-4500 fluorimeter fitted with a customized
ECL cell and a set-up as described in reference [56]. For the ECL experi-
ments, solvents of similar grade to those used in the other electrochemi-
cal studies were employed, and the concentrations of the compounds
were adjusted to about 1 mm. TBAH (0.2m) served as supporting electro-
lyte, and the measurements were performed without stirring of the solu-
tions. The potential was switched with alternation between the oxidation
and reduction potentials (see inset of Figure 6, electrochemical switching
frequency 1 sꢀ1, scan rate of the spectrometer 240 nmminꢀ1). The spiked
spectrum is the result of the “slow” switching frequency of 1 sꢀ1, whilst
the other spectra were obtained with n = 50 and 20 msꢀ1 in the case of
pSAc-1 and (pS-1)2.
[12] a) A. P. Fernandes, A. Holmgren, Antioxid. Redox Signal. 2004, 6,
63–74; b) D. Barford, Curr. Opin. Struct. Biol. 2004, 14, 679–686;
c) K. Linke, U. Jakob, Antioxid. Redox Signal. 2003, 5, 425–434;
d) H. Sies, Free Radic. Biol. Med. 1999, 27, 916–921.
[13] a) G. T. Hanson, R. Aggeler, D. Oglesbee, M. Cannon, R. A. Capa-
ldi, R. Y. Tsien, S. J. Remington, J. Biol. Chem. 2004, 279, 13044–
13053; b) H. Østergaard, A. Henriksen, F. G. Hansen, J. R. Winther,
EMBO J. 2001, 20, 5853–5862.
Acknowledgment
Financial support by the German Research Foundation (DFG), the
German National Academic Foundation (C.T.) and BAMꢁs Ph.D. pro-
gramme (H.R.) is greatly appreciated. This work is part of the Graduate
College “Sensory Photoreceptors in Natural and Artificial Systems”
granted by the DFG (GRK 640, University of Regensburg).
[14] H. Østergaard, C. Tachibana, J. R. Winther, J. Cell Biol. 2004, 166,
337–345.
[15] a) T. Sakata, Y. Yan, G. Marriott, J. Org. Chem. 2005, 70, 2009–
2013; b) P. Ge, P. R. Selvin, Bioconjugate Chem. 2003, 14, 870–876.
[16] R. Rinaldi, G. Maruccio, A. Biasco, P. Visconti, V. Arima, R. Cingo-
lani, Ann. N.Y. Acad. Sci. 2003, 1006, 187–197.
[1] a) L. Pu, Chem. Rev. 2004, 104, 1687–1716; b) A. P. de Silva, B.
McCaughan, B. O. F. McKinney, M. Querol, Dalton Trans. 2003,
1902–1913; c) R. Martinez-Manez, F. Sancenon, Chem. Rev. 2003,
103, 4419–4476.
[17] a) P. Toele, H. Zhang, C. Trieflinger, J. Daub, M. Glasbeek, Chem.
Phys. Lett. 2003, 368, 66–75; b) M. Kollmannsberger, K. Rurack, U.
Resch-Genger, J. Daub, J. Phys. Chem. A 1998, 102, 10211–10220;
c) J. Karolin, L. B.-A. Johansson, L. Strandberg, T. Ny, J. Am. Chem.
Soc. 1994, 116, 7801–7806; d) J. A. Pardoen, J. Lugtenburg, G. W.
Canters, J. Phys. Chem. 1985, 89, 4272–4277.
[18] a) W. Zhao, E. M. Carreira, Angew. Chem. 2005, 117, 1705–1707;
Angew. Chem. Int. Ed. 2005, 44, 1677–1679; b) Z. Shen, H. Rçhr,
K. Rurack, H. Uno, M. Spieles, B. Schulz, G. Reck, N. Ono, Chem.
Eur. J. 2004, 10, 4853–4871; c) K. Rurack, M. Kollmannsberger, J.
Daub, New J. Chem. 2001, 25, 289–292; d) H. Kim, A. Burghart,
M. B. Welch, J. Reibenspies, K. Burgess, Chem. Commun. 1999,
1889–1890; e) J. Chen, J. Reibenspies, A. Derecskei-Kovacs, K. Bur-
gess, Chem. Commun. 1999, 2501–2502.
[2] a) F. M. Raymo, M. Tomasulo, Chem. Soc. Rev. 2005, 34, 327–336;
b) Molecular Switches (Ed.: B. Feringa), Wiley-VCH, Weinheim,
2002; c) A. P. de Silva, D. B. Fox, T. S. Moody, S. M. Weir, Trends
Biotechnol. 2001, 19, 29–34.
[3] a) A. P. de Silva, N. D. McClenaghan, Chem. Eur. J. 2003, 10, 574–
586; b) V. Balzani, A. Credi, M. Venturi, Chem. Phys. Chem. 2003,
3, 49–59; c) F. M. Raymo, Adv. Mater. 2002, 14, 401–414.
[4] a) Special Issue (6) on “Molecular Machines” Acc. Chem. Res. 2001,
34, 409–522; b) C. P. Mandl, B. Kçnig, Angew. Chem. 2004, 116,
1650–1652; Angew. Chem. Int. Ed. 2004, 43, 1622–1624; c) Molec-
ular Devices and Machines,A Journey into the Nanoworld (Eds.: V.
Balzani, M. Venture, A. Credi), Wiley-VCH, Weinheim, 2003;
d) A. R. Pease, J. F. Stoddart, Struct. Bonding 2001, 99, 189–236.
[5] a) J. R. Heath, Pure Appl. Chem. 2000, 72, 11–20; b) D. T.
McQuade, A. E. Pullen, T. M. Swager, Chem. Rev. 2000, 100, 2537–
2574; c) A. Harriman, R. Ziessel, Chem. Commun. 1996, 1707–
1716.
[19] a) Y. Gabe, Y. Urano, K. Kikuchi, H. Kojima, T. Nagano, J. Am.
Chem. Soc. 2004, 126, 3357–3367; b) C. Goze, G. Ulrich, L. Char-
bonni›re, M. Cesario, T. PrangØ, R. Ziessel, Chem. Eur. J. 2003, 9,
Chem. Eur. J. 2006, 12, 689 – 700
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
699

K. Rurack, J. Daub et al.
3748–3755; c) R. W. Wagner, J. S. Lindsey, Pure Appl. Chem. 1996,
68, 1373–1380.
[40] F. Li, S. I. Yang, Y. Ciringh, J. Seth, I. C. H. Martin, D. L. Singh, D.
Kim, R. R. Birge, D. F. Bocian, D. Holten, J. S. Lindsey, J. Am.
Chem. Soc. 1998, 120, 10001–10017.
[41] Note that the involvement of triplet states in the deactivation of the
BDPꢁs S1 state was found to be of negligible importance (e.g.
refs. [17a, 18b], J. Banuelos Prieto, F. López Arbeloa, V. Martinez -
Martinez, I. López Arbeloa, Chem. Phys. 2004, 296, 13–22) and sig-
nificantly lower than in rhodamine dyes; see, for example, F. Liang,
H. P. Zeng, Z. R. Sun, Y. Z. Yuan, Z. G. Yao, Z. Z. Xu, J. Opt. Soc.
Am. B 2001, 18, 1841–1845.
[20] a) G. Beer, K. Rurack, J. Daub, Chem. Commun. 2001, 1138–1139;
b) G. Beer, C. Niederalt, S. Grimme, J. Daub, Angew. Chem. 2000,
112, 3385–3388; Angew. Chem. Int. Ed. 2000, 39, 3252–3255; c) M.
Kollmannsberger, K. Rurack, U. Resch-Genger, W. Rettig, J. Daub,
Chem. Phys. Lett. 2000, 329, 363–369.
[21] a) J. M. Brom, Jr., J. L. Langer, J. Alloys Compd. 2002, 338, 112–
115; b) R. Y. Lai, A. J. Bard, J. Phys. Chem. B 2003, 106, 5036–5042.
[22] a) L. Milanesi, G. D. Reid, G. S. Beddard, C. A. Hunter, J. P. Waltho,
Chem. Eur. J. 2004, 10, 1705–1710; b) W. M. Macindoe, A. H. va-
n Oijen, G.-J. Boons, Chem. Commun. 1998, 847–848.
[42] Corresponding data in CH2Cl2: ꢀ1750 mV and +580 mV versus
Fc+/Fc.
[23] F. Bergstrçm, I. Mikhalyov, P. Hägglçf, R. Wortmann, T. Ny, L. B.-
Š. Johansson, J. Am. Chem. Soc. 2002, 124, 196–204.
[24] Various attempts to obtain the corresponding monomer oSH-1 from
(oS-1)2 failed, and cyclovoltammetric measurements suggested that
the close vicinity of nucleophilic thiol group and BDP residue most
probably renders the monomer rather unstable.
[43] These two processes belong together as confirmed by multisweep
experiments under thin-film conditions. The corresponding data in
CH2Cl2 are ꢀ1750 mV and +590 mV; the BDPꢁs redox potentials of
(pS-1)2 lie at ꢀ1875 mV and +590 mV in CH2Cl2.
[44] S. Antonello, K. Daasbjerg, H. Jensen, F. Taddei, F. Maran, J. Am.
Chem. Soc. 2003, 125, 14905–14916.
[25] Y. Fujimoto, N. Katayama, Y. Ozaki, S. Yasui, K. Iriyama, J. Mol.
Struct. 1992, 274, 183–195.
[45] Diaryl disulfides are known to be reductively cleaved to thiolate
anions, see refernce [44] and references therein.
[26] K. Rurack, M. Kollmannsberger, U. Resch-Genger, J. Daub, J. Am.
Chem. Soc. 2000, 122, 968–969.
[27] T. Gareis, C. Huber, O. S. Wolfbeis, J. Daub, Chem. Commun. 1997,
1717–1718.
[28] Because of the sensitivity of such thiolates towards oxidation, these
measurements were performed under Ar. Note that the use of such
conditions has a negligible influence on the fluorescence features of
the title dyes: Ff = 0.64 and 0.65 for pSH-1 in aerated and de-aer-
ated MeCN, for example.
[46] By use of the Digisim 3.03a program; M. Rudolph, S. W. Feldberg,
Digisim 3.03a, Bioanalytical Systems, Inc., West Lafayette, IN, USA,
1994–2000.
[47] a) K. Kelnhofer, A. Knorr, Y. H. Tak, H. Bässler, J. Daub, Acta
Polym. 1997, 48, 188–192; b) J. Daub, K. Kelnhofer, T. Gareis, A.
Knorr, M. Kollmannsberger, Y.-H. Tak, H. Bässler, Polym. Preprints
1997, 38, 339–340; c) J. Strauss, J. Daub, Adv. Mater. 2002, 14, 1652–
1655; d) J. Strauss, J. Daub, Adv. Mater. 2003, 15, 258; e) C. Trieflin-
ger, H. Rçhr, K. Rurack, J. Daub, Angew. Chem. 2005, 117, DOI:
10.1002/ange. 200501573; Angew. Chem. Int. Ed. 2005, 44, DOI:
10.1002/anie.200501573.
[48] M. M. Richter, Chem. Rev. 2004, 104, 3003–3036.
[49] The bathochromic shift of ca. 20 nm between the fluorescence and
the ECL spectrum is well known and can be attributed to reabsorp-
tion effects due to the higher dye concentration used in the ECL ex-
periment and to the different dielectric properties of the supporting
electrolyte.
[29] W. Rettig, Top. Curr. Chem. 1994, 169, 253–299.
[30] AMPAC V6.55, Semichem, Inc. M. J. S. Dewar, E. G. Zoebisch, E. F.
Healy, J. J. P. Stewart, J. Am. Chem. Soc. 1985, 107, 3902–3909.
[31] Such a dependence of the intensity of the bathochromically shifted
transition on the interannular twist angle has also been found for
the phenolate anion of 9-(p-hydroxyphenyl)acridine, V. Zanker, A.
Reichel, Z. Elektrochem. 1959, 63, 1133–1140 and V. Zanker, A.
Reichel, Z. Elektrochem. 1960, 64, 431–437.
[32] a) Y. Wang, C.-P. Chang, H. Tian, Dyes Pigm. 2002, 54, 265–274;
b) Y. Wang, C.-P. Chang, Y. Wu, H. Tian, Dyes Pigm. 2001, 51, 12 7–
136.
[33] The findings that both the absorption and emission bands of (oS-1)2
are only slightly broader (on average by ꢂ15 cmꢀ1) than those of
(pS-1)2 suggest that exciton coupling such as has been observed for
p-conjugated pre-twisted BDPs by Falk and Schoppel (see ref. [39])
does not play a major role in the present case.
[50] a) C. Kempter, W. Pçtter, N. Binding, H. Kläning, U. Witting, U.
Karst, Anal Chim. Acta 2000, 410, 47–64; b) D. Blakeslee, M. G.
Baines, J. Immunol. Methods 1976, 13, 305–320; c) Reactive Dyes in
Protein and Enzyme Technology (Eds. Y. D. Clonis, A. Atkinson,
C. J. Bruton, C. R. Lowe), Macmillan, Basingstoke, UK, 1987.
[51] a) J. Salbeck, Dissertation, Universität Regensburg, 1988; b) J. Sal-
beck, J. Electroanal. Chem. 1993, 340, 169–175.
[52] Synthesized according to P. R. Menard, J. T. Suh, H. Jones, B. Love,
E. S. Neiss, J. Wilde, A. Schwab, W. S. Mann, J. Med. Chem. 1985,
28, 328–332.
[34] R. F. Khairutdinov, N. Serpone, J. Phys. Chem. B 1997, 101, 2602–
2610.
[35] W. R. Ware in Photochemistry in Organized and Constrained Media
(Ed.: V. Ramamurthy), VCH, New York, 1991, pp. 563–602.
[36] a) K. Rurack, K. Hoffmann, W. Al-Soufi, U. Resch-Genger, J. Phys.
Chem. B 2002, 106, 9744–9752; b) R. Fritz, A. Kungl, W. Rettig, J.
Springer, Chem. Phys. Lett. 1996, 260, 409–417; c) F. V. Bright, G. C.
Catena, J. Huang, J. Am. Chem. Soc. 1990, 112, 1343–1346.
[37] D. R. James, Y.-S. Liu, P. De Mayo, W. R. Ware, Chem. Phys. Lett.
1985, 120, 460–465.
[53] J. Olmsted, III, J. Phys. Chem. 1979, 83, 2581–2584.
[54] U. Resch, K. Rurack, Proc. SPIE-Int. Soc. Opt. Eng. 1997, 3105, 96–
103.
[55] J. Salbeck, Anal. Chem. 1993, 65, 2169–2173, M. Bueschel, C. Sta-
dler, C. Lambert, M. Beck, J. Daub, J. Electroanal. Chem. 2000, 484,
24–32.
[56] S. Hien, Ph.D. Thesis, University of Regensburg, 1995.
[38] S. G. Tarasov, J. R. Casas-Finet, W. M. Cholody, C. J. Michejda, Pho-
tochem. Photobiol. 1999, 70, 568–578.
Received: June 24, 2005
[39] H. Falk, G. Schoppel, Monatsh. Chem. 1990, 121, 67–76.
Published online: October 17, 2005
700
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 689 – 700