On/off switching of perylene tetracarboxylic bisimide luminescence by means of substitution at the N-position by electron-rich mono-, Di-, and trimethoxybenzenes

Article abstract of DOI:10.1002/chem.201001489

A series of perylene tetracarboxylic bisimides, substituted at the N-position with methoxyphenyl groups, have been synthesized together with model compounds and their photophysical properties have been investigated by means of steady-state and time-resolv

Full text of DOI:10.1002/chem.201001489

DOI: 10.1002/chem.201001489
On/Off Switching of Perylene Tetracarboxylic Bisimide Luminescence by
Means of Substitution at the N-Position by Electron-Rich Mono-, Di-,
and Trimethoxybenzenes
Lucia Flamigni,*[a] Barbara Ventura,[a] Andrea Barbieri,[a] Heinz Langhals,*[b]
Fritz Wetzel,[b] Kerstin Fuchs,[b] and Andreas Walter[b]
Abstract: A series of perylene tetracar-
boxylic bisimides, substituted at the N-
position with methoxyphenyl groups,
have been synthesized together with
model compounds and their photo-
physical properties have been investi-
gated by means of steady-state and
time-resolved spectroscopic techniques.
The luminescence properties of the ex-
amined compounds vary remarkably
with the substitution pattern, with
emission quantum yields ranging from
1 to 10ꢁ2–10ꢁ3. The observed quenching
of the luminescence is assigned to a
photoinduced electron transfer (PET)
from the electron-rich methoxybenzene
unit to the perylene bisimide moiety.
The radical anion of perylene bisimide
has been detected by transient-absorp-
tion spectroscopy. The results could
satisfactorily be explained by taking
into consideration the redox potentials
of the partners and the electron-releas-
ing ability of each methoxy group in
relation to its position with respect to
N. Quantum-chemical calculations
were also performed.
Keywords: bisimides · charge sepa-
ration · electron transfer · lumines-
cence · supramolecular chemistry
Introduction
cal anion characterized by distinctive and intense absorption
bands in the NIR region of the spectrum.[6–8] This class of
molecules is therefore well suited to mechanistic studies on
electron-transfer processes. Excitation in the PI excited-
state manifold (450–550 nm) leads to an excited state that
has sufficient energy (around 2.3 eV) to accept in the semi-
occupied HOMO an electron from the HOMO of a nearby
electron-rich unit. Due to the short lifetime of the excited
state (3.8 ns) the potential electron donor has to be very
close, that is, it has to be either covalently or noncovalently
connected to the PI unit to be able to interact. On the other
hand, ground-state PIs can also accept in their LUMO elec-
trons from the LUMO of excited chromophores, provided
enough energy is stored in the donor excited state. In recent
years we, as other groups, have extensively explored several
molecular architectures of variable complexity, including
PIs, with the aim of generating, upon an efficient HOMO–
HOMO or LUMO–LUMO electron transfer, long-lived
charge-separated (CS) states that might store light energy as
electrochemical potential for useful reactions.[8,9]
On/off switching of luminophore emission by photoinduced
electron transfer (PET) is an extensively explored area of
research that has produced copious data and promising ap-
plications in the fields of sensing[1] and information process-
ing at the molecular level.[2] Perylene tetracarboxylic bisi-
mides (PIs) are a class of efficient luminophores, very stable
and easily processable, with intense absorption in the visible
range of the light spectrum; they can easily be functional-
ized to provide a variety of spectroscopic features.[3–5] PIs,
like other aromatic bisimides, are liable to a facile reduction
(E0 ꢀꢁ0.6 V vs. SCE) that leads to the formation of a radi-
[a] Dr. L. Flamigni, Dr. B. Ventura, Dr. A. Barbieri
Istituto ISOF-CNR
Via P. Gobetti 101, 40129 Bologna (Italy)
Fax : (+39)051-639-9844
E-mail: flamigni@isof.cnr.it
[b] Prof. Dr. H. Langhals, Dr. F. Wetzel, Dr. K. Fuchs, A. Walter
Department of Chemistry, LMU University of Munich
Butenandtstrasse 13, 81377 Munich (Germany)
Fax : (+49)89-2180-77640
In the present report we focus on the mechanism of lumi-
nescence quenching by a series of similar electron-rich sub-
stituents, namely, mono-, di-, and trimethoxybenzenes, di-
rectly connected to the N-position of a PI (Scheme 1). The
molecular design in this case has not been aimed toward
E-mail: Langhals@lrz.uni-muenchen.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001489.
13406
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416

FULL PAPER
connected to the PI core. Different substitution patterns
have been explored to understand the contribution of differ-
ent parameters to the luminescence quenching of the PI lu-
minophore. We will present the synthesis and the spectro-
scopic properties and the photophysical experiments de-
signed to clarify the mechanism of on/off luminescence
switching in the series of compounds and will discuss in
detail the results also with the aid of molecular modeling.
Results and Discussion
Design and synthesis: The perylene derivatives were pre-
pared in a one-step synthesis from the carboxylic anhy-
dride[10] 1 by condensation with primary amines in quinoline
and imidazole,[11] respectively, according to Scheme 2. Spe-
cial care was taken in the chromatographic purification of
the PI derivatives.
Absorption spectroscopy: The various compounds reported
in Scheme 1 are an exhaustive series of mono-, di-, and tri-
methoxybenzene-substituted PI with reference models PI-
B1 and PI-B2. The former is a model that is missing the me-
thoxy groups; the latter has been designed as model to test
the steric effect of ortho substitution and separate it from
the electronic effect induced by electron-rich methoxy deriv-
atives at the same position. The compounds have been ex-
amined in low-polarity toluene (TL) and in medium-polarity
dichloromethane (CH2Cl2), characterized by values of die-
lectric constant of around 2.4 and 8.9, respectively. This is
useful in view of the expected effect of polarity on the possi-
ble charge-transfer reactions.
The absorption parameters in the two solvents are report-
ed in the Supporting Information (Table S1), and a few ex-
amples of absorption spectra are reported in Figure 1 for
solutions in TL and CH2Cl2. The spectra of all compounds
in TL are very similar, with band maxima at 526 or 527 nm
and identical molar absorption coefficients within experi-
mental errors. This is true also for CH2Cl2 solutions, in
which the spectra display maxima shifted by 2 nm at 524 or
525 nm. The absence of, basically, any effect on the absorp-
tion spectra indicates that there are very little perturbations
in the electronic properties of the PI core by the presence at
the N-position of differently substituted methoxyphenyl
groups. This has been observed before[6,8] and is due to the
fact that orbital nodes in the HOMO and LUMO at the N
atom are present.[12] Due to the decoupling between the PI
core and the N-substituents, the present compounds can be
regarded as assemblies made of two independent compo-
nents, PI and methoxybenzene, and we will refer to them as
dyads.
Scheme 1. Molecular structures of the model compounds and substitution
pattern at the N with related nomenclature.
Scheme 2. Synthesis of PI derivatives; for R see Scheme 1.
Luminescence spectroscopy: The luminescence properties of
the series in the two solvents are collected in Table 1. The
luminescence maxima in CH2Cl2, at 530-532 nm, are hypso-
chromically shifted by 2–3 nm with respect to those in TL,
at 532–534 nm, in agreement with the absorption spectra.
generating long-lived CS states, but rather to increase the
electronic coupling between components. Therefore the
electron-rich methoxy benzene groups have been directly
Chem. Eur. J. 2010, 16, 13406 – 13416
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13407

H. Langhals, L. Flamigni et al.
10ꢁ2–10ꢁ3, thereby indicating an extremely efficient quench-
ing for some substitution patterns. The same substituent in-
duces higher quenching in CH2Cl2 compared to TL (the case
of PI-M2, unquenched in TL and slightly quenched in
CH2Cl2 is emblematic), with the exception of the weakest
emitters (ffl ꢂ10ꢁ2), for which the emission yields in the two
solvents are essentially the same within experimental errors
(see Figure 2 for some selected cases). The fluorescence
decays can be described by a single exponential in all cases,
with values of lifetimes varying from (3.8ꢃ0.1) ns for the
unquenched dyads and the models to approximately 5 ps,
which is the limit of the experimental resolution for emis-
sion experiments (Table 1). For both solvents, the decrease
in lifetime parallels reasonably well the luminescence quan-
tum yield, so that a radiative rate constant kr (kr =ffl/t) on
the order of 3–6ꢁ108 sꢁ1 can be calculated for all dyads. This
is a further indication that the PI intrinsic parameters are
very little affected by the methoxy substitution, and that the
donor and acceptor units are decoupled from an electronic
viewpoint. Nonetheless, the presence of methoxy deriva-
tives, in particular of the di- and trimethoxybenzenes with a
substituent in the para position with respect to the N-link-
age, leads to a strong quenching of the luminescence.
Quite remarkably, the fluorescence quenching is almost
completely blocked in a rigid TL glass at 77 K for all sys-
tems except PI-M4, PI-M8, and
Figure 1. Absorption spectra: A) PI-B1 (black line), PI-M4 (gray line),
and PI-M9 (circles) in TL; B) PI-B1 (black line), PI-M3 (gray line), and
PI-M8 (circles) in CH2Cl2.
PI-M9, which display slightly
Table 1. Luminescence data for the examined compounds in CH2Cl2 and TL. The quenched dyads are indicat-
ed in gray.
quenched lifetimes of 3.4, 3.2,
and 3.5 ns, respectively, relative
to the model lifetime of (3.8ꢃ
0.1) ns. These dyads also exhibit
295 K
Ffl
77 K
[b]
lmax [nm][a]
t [ns][c]
lmax [nm][a]
t [ns][c]
E [eV][d]
PI-B1
PI-B2
PI-M1
PI-M2
PI-M3
PI-M4
PI-M5
PI-M6
PI-M7
PI-M8
PI-M9
TL
532, 574, 622
530, 572, 620
534, 576, 622
530, 570, 620
534, 574, 620
530, 572, 616
533, 574, 620
530, 572, 618
533, 574, 620
530, 571, 618
533, 574, 622
530, 572, 618
533, 574, 622
530, 570, 618
533, 574, 620
530, 570, 616
533, 574, 620
530, 570, 618
533, 574, 620
530, 571, 616
534, 574, 624
532, 572, 620
0.98
3.8
542, 586, 637
3.8
3.7
3.8
3.8
3.8
3.4
3.7
3.7
3.8
3.2
3.5
2.29
the
highest
fluorescence
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
CH2Cl2
TL
0.98
0.98
0.99
0.19
0.081
0.99
0.72
0.98
0.94
0.0069
0.0067
0.20
0.059
0.051
0.016
0.99
3.8
3.8
3.7
0.720
0.225
3.7
2.6
3.9
544, 588, 639
543, 586, 637
543, 586, 636
543, 585, 637
543, 587, 637
543, 586, 637
543, 585, 637
542, 585, 636
543, 587, 638
541, 585, 637
2.28
2.28
2.28
2.28
2.28
2.28
2.28
2.29
2.28
2.29
quenching at room tempera-
ture. For these, a low quenching
rate on the order of 107 sꢁ1 can
be calculated at 77 K by the
equation kq =1/tꢁ1/t0, in which
t and t0 represent the lifetime
of the excited state in the dyad
and in the model PI-B1, respec-
tively. As a general comment,
we can certainly state that the
process responsible for the lu-
minescence quenching is ther-
mally activated and disfavored
in a rigid medium.
3.7
0.010
0.010
0.840
0.123
0.260
0.046
3.9
CH2Cl2
TL
CH2Cl2
TL
0.93
3.6
0.0036
0.0030
0.013
0.016
ca. 0.005
ca. 0.005
0.033
0.052
Time-resolved absorption: Fur-
ther experiments were conduct-
ed
methods to detect possible non-
luminescent reaction intermedi-
ates. Picosecond time-resolved
CH2Cl2
[a] Data from uncorrected spectra. [b] Excitation at 490 nm in toluene and 489 nm in dichloromethane. For de-
tails, see the Experimental Section. [c] Excitation at 465 nm for nanosecond and at 532 nm for picosecond de-
terminations. [d] From spectra recorded at 77 K.
by
transient-absorption
Excitation spectra registered on the emission maxima per-
fectly match the absorption spectra, thereby indicating a
genuine emission of the lowest singlet excited state. The lu-
minescence quantum yield has values ranging from 0.99 to
absorption spectra of optically matched solutions with exci-
tation at 532 nm in the two solvents were performed. The
spectra for selected cases are reported in Figure 3 and
Figure 4 in TL and CH2Cl2, respectively, and the data are
13408
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416

Bisimide Luminescence
FULL PAPER
the lowest energy bands of the stimulated emission, and pos-
itive peaks at 690 and around 850 nm are detected. In
CH2Cl2, the stimulated luminescence bands in PI-B1 ap-
pears at around 575 and 620 nm, and a positive intense and
sharp peak is found at 710 nm with a broad, lower structure-
less absorption that extends from 750 to 950 nm. Though
different, the spectra were formerly identified as being due
to the same species, the singlet excited state of PI chromo-
phores, 1PI.[6] In agreement with this, they evolve very
slowly on the explored timescale (0–3.3 ns) with a lifetime
on the order of some nanoseconds, compatible with (3.8ꢃ
0.1) ns, the one derived from luminescence determinations.
A more precise determination is precluded by the too-
narrow time window explored during the experiment. The
same spectral shape and time profile as PI-B1 is displayed
by solutions of the unquenched dyads PI-B2, PI-M2, PI-M3,
and PI-M7 in TL (spectra not shown); see Table 1. Likewise,
the partially quenched dyads PI-M1, PI-M5, and PI-M6 (see
1
Figure 3B) display the same spectral shape as PI and the
decay lifetime is coincident, within experimental error, with
that determined by luminescence (compare Table 1 with
Table 2). On the contrary, the spectral shape displayed by
the most-quenched dyads PI-M4, PI-M8, and PI-M9, char-
acterized by a luminescence lifetime ꢂ30 ps, is different
from that of 1PI. The spectra do not have any stimulated
emission band but only positive
Figure 2. Uncorrected emission spectra: A) PI-B1 (black line), PI-M4
(gray line), and PI-M6 (circles) in TL; B) PI-B1 (black line), PI-M4
(gray line), and PI-M6 (circles) in CH2Cl2.
absorptions at 700, 800, and
950 nm (Figure 3C and D).
These spectral features are in
reasonably good agreement
with the anion radical PIꢁ in
TL, with bands at 710, 800, and
960 nm.[6,8a,9i] This is an indica-
tion that the observed species is
the reduced radical PIꢁ and, to
further support this, the mea-
sured lifetimes of the absorbing
species, 95 ps for PI-M8, 48 ps
for PI-M4, and 50 ps for PI-M9,
1
are different from those of PI
detected by luminescence tech-
niques, around 5, 10, and 33 ps,
respectively.
In CH2Cl2, the detected spe-
cies is that of the excited state
1PI in all cases except PI-M8
and, to some extent, PI-M4
(Figure 4), as confirmed by the
spectral shape. The time evolu-
tion of the absorbing intermedi-
ate is coincident with the lumi-
nescence lifetime, compare
Table 1 with Table 2, except for
PI-M8 (Figure 4C) and PI-M4
Figure 3. Time evolution of the absorption spectra registered after excitation with a laser pulse (35 ps, 532 nm,
3.5 mJ) of optically matched (A=0.65) solutions of A) PI-B1, B) PI-M6, C) PI-M8, and D) PI-M4 in TL. For
time delays of the spectra, see the insets in which time profiles at selected wavelengths and exponential fittings
are displayed.
collected in Table 2. Model PI-B1 displays different end of
pulse absorption spectra in TL and CH2Cl2 (Figure 3A and
4A). In TL, negative signals around 580 and 620 nm, due to
(Figure 4D), for which the decay lifetimes measured at
710 nm are 30 and 20 ps, respectively. It is evident that in
PI-M8 the species formed and decaying within the pulse is
Chem. Eur. J. 2010, 16, 13406 – 13416
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13409

H. Langhals, L. Flamigni et al.
the anion PIꢁ, with bands at
710, around 800, and 960 nm,
and the absence of the stimulat-
ed emission bands at 575 nm.
The situation is more ambigu-
ous for PI-M4 (see Figure 4D),
in which a trace of stimulated
emission at 575 nm with a very
fast decay (on the order of
10 ps) can still be detected, but
the spectrum has features that
resemble those of the anion
(see the band at 960 nm), and
the measured lifetime at
710 nm is around 20 ps. In this
case, we have clearly a mixture
1
of PI and PIꢁ, though a better
description of the system is pre-
cluded by the time resolution of
the apparatus. In all the exam-
ined cases, the transient spectra
recovers back to the baseline,
thus indicating that no further
intermediate product is formed
Figure 4. Time evolution of the absorption spectra registered after excitation with a laser pulse (35 ps, 532 nm,
3.5 mJ) of optically matched (A=0.65) solutions of A) PI-B1, B) PI-M5, C) PI-M8, and D) PI-M4 in CH2Cl2.
For time delays of the spectra, see the insets in which time profiles at selected wavelengths and exponential fit- after the decay of either the sin-
tings are displayed.
glet excited state or the anion.
Photoinduced processes: The
strong luminescence of PI can
Table 2. Transient-absorption parameters and assignment of the bands for the examined compounds in CH2Cl2
and TL at 295 K. The quenched dyads are indicated in gray. In the last columns, the oxidation potential and
the Hammett constant (s) of the pertinent methoxybenzene are collected.
be quenched to a great extent
by substitution at the nodal N-
position by a series of methoxy-
benzenes. Not all substitution
patterns are effective in
295 K
Substituents
Oxidation potential s[c]
(V vs. SCE)
+2.48[d]
Species lmax [nm][a]
1PI
t [ns][b]
quenching the fluorophore lu-
minescence; by inspection of
Table 1 and Scheme 1, it is evi-
dent that there are several pa-
rameters to consider. Both the
number of methoxy substituents
and the relative position of the
substituents with respect to
each other or with respect to
the attachment point to the PI
structure are critical in achiev-
ing quenching. We can rational-
ize the different substitution
patterns of the methoxyben-
zenes and the position of the at-
tachment point to the PI struc-
PI-B1
PI-B2
PI-M1
PI-M2
PI-M3
PI-M4
TL
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
580 (bleaching), 690, 850 ca. 4.0
575 (bleaching), 710 ca. 3.8
580 (bleaching), 690, 850 ca. 3.9
575 (bleaching), 710 ca. 3.6
580 (bleaching), 690, 850 0.670
575 (bleaching), 710 0.220
580 (bleaching), 690, 850 ca. 4.0
575 (bleaching), 710 ca. 2.5
580 (bleaching), 690, 850 ca. 3.7
575 (bleaching),710
700, 800, 950
575 (bleaching), 710,
850, 960
+2.06[d]
+1.76[e]
+1.76[e]
+1.76[e]
+1.45[e]
ꢁ0.10
ꢁ0.27
0.12
–
ca. 3.5
TL
PI-
0.048
0.12; ꢁ0.27
CH2Cl2 1PI,
PIꢁ
0.010, 0.020[f]
PI-M5
PI-M6
PI-M7
PI-M8
PI-M9
TL
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
TL
1PI
CH2Cl2 1PI
TL
PIꢁ
CH2Cl2 PIꢁ
TL
PIꢁ
CH2Cl2 1PI
1PI
580 (bleaching), 690, 850 0.810
575 (bleaching), 710 0.130
580 (bleaching), 690, 850 0.250
575 (bleaching), 710 0.060
580 (bleaching), 690, 850 ca. 4.0
+1.6[g]
+1.6[g]
+1.6[g]
+1.34[e]
+1.42[e]
0.12; 0.12
0.12; ꢁ0.27
–; –
–; 0.12
0.12; 0.12; ꢁ0.27
575 (bleaching), 710
700, 800, 950
ca. 3.9
0.095
0.030
0.050
0.065
ꢁ
ture with respect to the OCH3
710, 850, 960
700, 800, 950
group(s) if we examine: 1) the
oxidation potentials of each
single methoxybenzene sub-
stituent, and 2) the Hammett
constants (s) that describe the
electron-releasing (negative s)
575 (bleaching), 710
[a] Detected band maxima, typical of the identified species. [b] The picosecond time-resolved experiment has
a time window of 3.3 ns and allows the determination of lifetimes longer than 2 ns with rather low precision.
1
[c] From ref. [15]. [d] Eox in acetonitrile (vs. SCE) from ref. [13]. [e] E = in acetonitrile (vs. SCE) from ref. [14].
2
[f] See text for details. [g] Irreversible, calculated according to the equation in ref. [14].
13410
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416

Bisimide Luminescence
FULL PAPER
or electron-withdrawing (positive s) properties of the me-
thoxy group(s) in the different positions with respect to N,
the attachment point. The literature oxidation potentials for
each methoxy benzene (or benzene) derivative[13,14] of the
dyad and the Hammett constants[15] of the group(s) con-
tained in the substituted benzene at the N-position are re-
ported in the last two columns of Table 2. It should be no-
ticed that the Hammett constants for ortho substitution are
liable to large errors because of the difficulty in sorting elec-
tronic from steric factors and are therefore not provided.
However, we will see that the simple parameters reported in
the last two columns of Table 2 are sufficient to discuss the
results.
As demonstrated by the absorption characteristics and by
1
the constancy in the PI radiative rate constants of the ex-
amined dyads, the two component units interact very little
both in the ground and in the excited states. Quenching,
when effective, therefore has to be ascribed to a “dynamic”
interaction between the PI unit and the methoxybenzene
unit that can transfer electrons to the partially empty
HOMO of the PI unit, as confirmed by the detection of the
anion radical PIꢁ. The mechanism of electron transfer is
also supported by the fact that the reaction is prevented at
77 K in a glass. This is typical for electron-transfer process-
es; in fact, whereas fluid solvents can stabilize the ionic
products of charge-transfer reactions by reorientation of the
solvent molecules, frozen solvents cannot do so and this
often prevents electron transfer from occurring because of
the insufficient or even negative driving force for the reac-
tion.[16] The discussion of the reaction dynamics within this
series of dyads can be done on the basis of a very simple
energy-level diagram (Figure 5). In this we report the
energy levels of the PI chromophore at approximately
2.3 eV (1PI) and at approximately 1.2 eV (3PI)[17]; those are
in fact the only excited states that can come into play upon
excitation in the energy range above 350–400 nm. A relative
scale of the energy levels of the charge-separated (CS)
states formed upon HOMO–HOMO electron transfer from
the methoxy derivative to the PI unit can be roughly esti-
mated from an Eo =ꢁ0.62 V versus SCE in CH2Cl2 for PI[6]
and from the oxidation potentials of the substituents report-
ed in Table 2. This allows one to derive approximate energy
levels of the CS states that involve reduction of PI and oxi-
dation of methoxybenzene (M), PIꢁ–M+, at an energy
ꢄ2.7 eV for PI-B1 and PI-B2; at around 2.3 eV for the
monosubstituted PI-M1, PI-M2, and PI-M3; at around
2.2 eV for the 1,3-derivatives PI-M5, PI-M6, and PI-M7; at
around 2.1 eV for PI-M4 and PI-M9; and below 2.0 eV for
PI-M8.[18] However, it is quite evident that the ability of the
methoxy group(s) to release electrons is also important in
the stabilization or destabilization of the methoxybenzene
cation, which is formed upon electron transfer and, as such,
the whole CS state. A higher stabilization of the CS state
means a higher driving force for electron transfer and, as a
consequence, a higher quenching rate. So, whereas mono-
substituted methoxybenzenes with ortho or meta substitu-
tion (PI-M2 and PI-M3) are not quenched, PI-M1, the one
Figure 5. Schematic representation of the energy-level diagram for the
examined dyads in CH2Cl2 and TL.
ꢁ
with an OCH3 group in the para position (negative s), has
a lower, more stabilized CS state and is quenched by elec-
tron transfer. A similar trend can be seen for the dime-
thoxy-substituted dyads, for which those with at least one
para substituent with respect to N are more strongly
quenched within the 1,2- and 1,3-substituted dyads. PI-M8,
on the other hand, has no para substitution, but the 1,4-de-
rivative can be oxidized more easily than all others (1.34 V
vs. SCE). It is worth noting that PI-M7 is not as liable to
electron transfer as the other 1,3-derivatives, very likely be-
ꢁ
cause the two OCH3 groups ortho to N are less electron-re-
leasing than the same groups meta to N. In the specific case
of P1-M7, the assignment of the energy level of the CS state
is based on kinetic evidence rather than on available data.
Finally the trisubstituted 1,2,3-methoxy derivative, PI-M9, in
spite of a slightly easier oxidation than the 1,2-methoxy de-
rivative PI-M4, has an extra meta substituent that apparent-
ly slightly destabilizes the CS state, as demonstrated by the
lower quenching. As is well known for electron-transfer re-
actions, a solvent polarity effect can also be noticed. This es-
sentially leads to the stabilization of CS states in more polar
CH2Cl2 (compared to TL), thus leading to a larger driving
force for electron transfer and to a faster quenching. This
trend can be clearly observed in dyads PI-M1, PI-M2, PI-
M5, and PI-M6. The positive effect of higher polarity on the
quenching levels off for more exoergonic reactions that are
in the “activationless” region, according to Marcus theory.[19]
This is the case for PI-M8, PI-M4, and PI-M9.
In conclusion, a HOMO–HOMO electron transfer from a
methoxybenzene unit to the excited PI unit can satisfactorily
explain all experimental observations. Spectroscopic signa-
Chem. Eur. J. 2010, 16, 13406 – 13416
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13411

H. Langhals, L. Flamigni et al.
tures that indicate the presence of the radical anion PIꢁ
have been distinguished in some cases. The detected spec-
trum should actually be due to the CS state PIꢁ–M+, but for
M+ spectral features at l<600 nm and molar-absorbance
coefficients on the order of 104 mꢁ1 sꢁ1 have been reported.[20]
This is an order of magnitude lower than the molar-absorp-
tion coefficient of PIꢁ,[21] so most probably the PIꢁ–M+
spectral features in the examined spectral range resemble
those of PIꢁ very closely. As stressed above, the CS-state
spectral features are only detected in a few cases, namely,
PI-M8 and PI-M4 in both solvents and PI-M9 in TL only,
and these are also the cases characterized by faster lumines-
cence quenching, that is, by a fast rate of charge separation.
This can be easily understood if we consider the following
consecutive kinetic scheme for the generic dyad PI–M.
of these orbitals. Fluorescence remains unaffected by these
substituents if their occupied orbitals are below the HOMO
of the chromophore that represents type I in Figure 6, such
CS
CR
1PI ꢁ M
PIꢁ ꢁ Mþ
PI ꢁ M
ð1Þ
ƒ!
ƒ!
Figure 6. Electronic terms of PI-M. Left: Electron-depleted substituent
(type I); right: electron-rich substituent (type II).
In the presence of a fast charge recombination (CR), the
species PIꢁ–M+ cannot accumulate when the rate of charge
separation (CS) is rather low and therefore cannot be de-
tected. Only when CS is extremely fast, or at least faster
than CR, does the species build up to the point where it can
be detected and its lifetime measured. This is the case of
PIꢁ–M+ in PI-M8, PI-M4, and PI-M9 in TL, which have CR
lifetimes of 95, 48, and 50 ps compared to CS lifetimes of
approximately 5, 10, and 33 ps, respectively. This is also true
for PI-M8 in CH2Cl2, which has a CR lifetime of 30 ps, com-
pared to a CS lifetime of around 5 ps. More ambiguous is
the case of PI-M4 in CH2Cl2 (see Figure 4D), for which CS
and CR rates are closer than in the other cases—10 and
20 ps, respectively—and the detected product is a mixture of
1PI-M and PIꢁ–M+ that evolves and decays within the pulse
length. As far as the CR rate is concerned, from the few
data available, the effect of the Marcus inverted region is
evident.[19] In fact, in CH2Cl2, in which the driving force for
recombination is lower (i.e., a more stabilized CS state) and
in which the polarity of the solvent enhances the reorganiza-
tional energy, recombination is faster than in TL (Table 2).
In spite of the presence of a low-lying triplet excited state at
as for the simple phenyl derivatives PI-B1 and PI-B2, and
even for the methoxy derivative PI-M3’ whose the orbitals
are shown in Figure 7. Moreover, two methoxy groups in
the ortho positions are not sufficient to interconvert the po-
sition of the orbitals (PI-M7’). As a consequence, no photo-
induced electron transfer (PET) can proceed and perturb
fluorescence so that unaltered high-fluorescence quantum
yields were found. This situation changes completely for the
trimethoxy derivative PI-M9’ in which two occupied orbitals
of the substituent are energetically above the HOMO of the
PI chromophore (see the right side of Figure 6, type II, and
the right side of Figure 7). As a consequence, an electronic
excitation (hn in Figure 6) leaves the HOMO of the chromo-
3
1.2 eV, PI-MB, recombination of PIꢁ–M+ occurs to the sin-
glet ground state, as testified by the baseline recovery of the
transient absorption spectrum. PI triplet is also known to
have, in addition to features around 500 nm, bands in the
NIR and should be detected in the present experiments, if
formed.[17,22]
Molecular modeling: The results of the complex spectro-
scopic behavior may be confirmed by means of quantum-
chemical calculations. We applied the DFT B3-LYP method
and simplified the structure of PI-M1–P1-M9 by means of
the replacement of the 1-hexylheptyl alkyl group by the
electronically similar methyl group to obtain PI-M1’–PI-
M9’. The calculations indicate energetically high-lying occu-
pied orbitals centered on the methoxyphenyl substituent
with energies close to the HOMO of the PI chromophore.
Slight structural variations can control the energetic position
Figure 7. Orbitals of the methyl analogues of PI-M (PI-M’). From bottom
to top (DFT B3-LYP): HOMOꢁ2, HOMOꢁ1, HOMO, LUMO. From
left to right: PI-M3’ and PI-M9’.
13412
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416

Bisimide Luminescence
FULL PAPER
phore semioccupied and a PET proceeds from the type II
side chain to the chromophore, fills the vacant position, and
inhibits fluorescence by this competition. One methoxy
group in the 4-position of the phenyl group to form PI-M1’
is sufficient to lift one filled orbital above the HOMO of the
chromophore and to dampen the fluorescence quantum
yield to 8% in CH2Cl2. The spectroscopic behavior becomes
more complex for PI-M2 in which both the filled orbitals of
the chromophore and of the methoxyphenyl substituent are
energetically similar and external influences, such as solvent
polarity, may switch from type I to type II and thus switch
fluorescence quantum yield.
Experimental Section
General: IR spectra were obtained using a Perkin–Elmer 1420 Ratio Re-
cording Infrared Spectrometer, FT 1000. UV/Vis spectra were obtained
using Varian Cary 5000 and Bruins Omega 20, and fluorescence spectra
were obtained using a Varian Eclipse instrument. NMR spectroscopy was
carried out using a Varian VNMRS 600 (600 MHz), and mass spectrome-
try was conducted using a Finnigan MAT 95 instrument.
Compound PI-B1: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (2.00 g, 3.48 mmol), formani-
lide (847 mg, 6.26 mmol), zinc acetate (100 mg, 0.45 mmol), and imida-
zole (10 g) under argon were heated at 1808C for 10 h and then allowed
to cool. While still warm, the solution was treated with 2n aqueous HCl
(55 mL), then collected by vacuum filtration, dried at 1208C in air
(2.15 g, 95%), purified by column separation (silica gel, chloroform),
evaporated, washed with distilled water, and dried under a medium
vacuum (508C, 4 h). Yield: 1.40 g (62%); Rf (silica gel, CHCl3)=0.53;
lH NMR (CDCl3): d=0.81 (t, 6H; 2ꢁCH3), 1.22 (mc, 16H; 8ꢁCH2), 1.87
(mc, 2H; lꢁa-CH2), 2.23 (mc, 2H; lꢁa-CH2), 5.17 (mc, lH; lꢁCH), 7.45
(mc, 5H; phenyl), 8.6 ppm (mc, 8H; perylene); 13C NMR (CDCl3): d=
14.03, 22.58, 26.95, 29.21, 31.76, 32.38, 54.85 (7 C-1-hexylheptyl), 123.03,
123.36, 123.31, 126.39, 126.65, 128.60, 128.88, 129.45, 129.53, 129.82,
131.77, 134.26, 135.08, 135.09 (14ꢁCaryl), 163.53 ppm (C=O); IR (KBr):
n˜ =2955 (m), 2927 (m), 2857 (m), 1699 (s), 1660 (s), 1594 (s), 1579 (m),
1505 (w), 1490 (w), 1455 (w), 1433 (w), 1405 (m), 1334 (s), 1255 (m),
1199 (w), 1177 (m), 1124 (w), 1110 (w), 1075 (w), 965 (m), 852 (w), 840
(w), 810 (s), 745 (s), 700 (w), 637 cmꢁl (w); UV (CHCl3): lmax (e)=492
(18750), 492 (52270), 528 nm (86730 molꢁ1 dm3 cmꢁ1); fluorescence
(CHCl3): lmax =538, 578 nm; MS (70 eV): m/z (%): 650 (5), 649 (23), 648
(47) [M+], 631 (7), 479 (4), 469 (3), 468 (18), 467 (60), 466 (100) [M+
ꢁC13H26], 465 (18), 449 (5), 422 (6), 421 (8), 373 (3), 345 (1), 233 (1), 69
(2), 55 (4), 41 (2), 28 (1); elemental analysis calcd (%) for C43H40N2O4: C
79.60, H 6.21, N 4.31; found: C 79.61, H 6.28, N 4.49.
The calculated structures concern the gas phase and all ef-
fects can be modulated by solvents. Thus, the comparably
high influence of the solvent polarity on the efficiency of
PET becomes understandable.
Conclusion
The effect of a series of electron-rich substituents directly
connected to the N-position of perylene tetracarboxylic bisi-
mide derivatives on the switching from a strong fluorescence
to an efficient generation of charge-separated excited states
by photoinduced electron transfer (PET) has been exam-
ined. The spectroscopic results have been rationalized in
terms of 1) the oxidation potentials of each single methoxy-
benzene substituent, 2) the Hammett constants of the me-
thoxy groups, and 3) quantum chemical calculations.
Compound PI-M3: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (350 mg, 0.61 mmol), 2-anisi-
dine (225 mg, 1.83 mmol), the amount of a microspatulum of zinc acetate
dihydrate, and imidazole (5 g) were heated at 1408C for 4 h. While still
warm, the solution was treated with 2n aqueous HCl (30 mL), stirred at
room temperature for 2 h, collected by vacuum filtration, washed with
distilled water, dried in air at 1208C, purified by column separation
(silica gel, chloroform to remove a forerun and then chloroform/acetone
10:1), concentrated by vacuum distillation, precipitated with methanol,
and dried in air at 1208C. Yield: 210 mg (51%); m.p. >2508C; Rf (silica
gel, CHCl3/acetone 10:1)=0.87; 1H NMR (300 MHz, CDCI3): d=0.83 (t,
The absorption spectra, very similar for all examined
compounds and with identical absorption coefficients, indi-
cate very little perturbations in the electronic properties of
the PI core by the presence at the N-position of differently
substituted methoxyphenyls, due to the presence of orbital
nodes at the N atom. In accordance, luminescence maxima
are the same for all compounds and are only slightly affect-
ed by the solvent polarity. On the other hand, the lumines-
cence quantum yields range from approximately 1 to 10ꢁ2–
10ꢁ3, thereby indicating an extremely efficient quenching for
some substitution patterns (in particular for the case of di-
and trimethoxybenzenes when a substituent in the para posi-
tion with respect to the N linkage is present). Transient-ab-
sorption experiments have allowed the detection of the
short-lived PI anion radical, thus indicating the population
of a charge-separated state, which was not the case for the
emitting derivatives. In all the examined cases, no further in-
termediate is formed after the decay of either the singlet ex-
cited state or the anion. A HOMO–HOMO electron trans-
fer from the methoxybenzene unit to the excited PI unit can
satisfactorily explain all experimental observations. The
analysis of the spectroscopic results has been confirmed by
quantum-mechanical calculations. In conclusion, the ob-
served switching is driven by the energy of the methoxyben-
zene-occupied orbitals with respect to the energy of the per-
ylene bisimide HOMO controlled by the positioning of the
methoxy substituents and may be further fine-tuned by
means of solvent effects.
JACHTUNRGTNEUNG(H,H)=6.6 Hz, 6H; 2ꢁCH3), 1.24–1.32 (m, 16H; 8ꢁCH2), 1.84–1.92
ꢁ
(m, 2H; b-CH2), 2.20–2.30 (m, 2H; b-CH2), 3.81 (s, 3H; OCH3), 5.14–
5.22 (m, 1H; a-CH), 7.10–7.52 (m, 4H; CHaryl), 8.58–8.71 ppm (m, 8H;
CHperylene); 13C NMR (75 MHz, CDCl3): d=14.1 (CH3), 22.6 (CH2), 27.0
ꢁ
(CH2), 29.2 (CH2), 31.8 (CH2), 32.4 (CH2), 45.8 (CH), 55.9 ( OCH3),
112.1, 121.1, 123.0, 123.2, 123.5, 123.9, 126.4, 126.7, 129.6, 129.9, 130.0,
10.5, 131.1, 131.7, 134.4, 134.9, 155.0, 163.2, 164.6 ppm; IR (KBr): n˜ =
3433 (brw), 3072 (brvw), 2954 (m), 2927 (m), 2856 (m), 1928 (brvw),
1712 (s), 1698 (s), 1658 (vs), 1595 (s), 1579 (m), 1502 (m), 1465 (w), 1434
(m), 1404 (m), 1344 (vs), 1305 (w), 1281 (w), 1255 (s), 1198 (w), 1177
(m), 1139 (vw), 126 (w), 1109 (w), 1045 (w), 1027 (w), 965 (w), 847 (w),
774 (vw), 747 (m), 729 (m), 696 (vw), 634 (w), 607 (vw), 572 (vw), 532
(vw), 495 (w), 430 cmꢁ1 (w); UV/Vis (CHCl3): lmax (Erel)=526.0 (l.0),
489.0 (0.60), 458.2 nm (0.21); fluorescence (CHCl3): lmax =534, 576 nm;
fluorescence quantum yield (lexc =489 nm, c=5.8ꢁ10ꢁ6 molLꢁ1 in chloro-
form, reference 2,9-bis-(1-hexylheptyl)anthra[2,1,9-def;6,5,10-d’e’f’]diiso-
quinoline-1,3,8,10-tetraone with F=1.0): 1.0; MS (70 eV): m/z (%): 678
(93) [M+], 661 (12) [M+ꢁOH], 647 (3) [M+ꢁOMe], 593 (3), 523 (2), 497
(64) [M+ꢁC13H25], 479 (6) [479ꢁH2O], 465 (100) [647ꢁC13H28], 437 (3)
[465ꢁCO], 390 (9) [465ꢁC6H3]; elemental analysis calcd (%) for
C44H42N2O5: C 77.85, H 6.24, N 4.13; found: C 77.70, H 6.13, N 4.10.
Compound PI-M1: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (460 mg, 800 mmol), 4-anisi-
dine (195 mg, 1.58 mmol), zinc acetate (37 mg, 0.20 mmol), and quinoline
Chem. Eur. J. 2010, 16, 13406 – 13416
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13413

H. Langhals, L. Flamigni et al.
(7 mL) were stirred at 2108C for 6 h under Ar atmosphere (dark red so-
lution) and then allowed to cool. While still warm and with strong stir-
ring, the solution was added dropwise to methanol/distilled water 1:1
(700 mL), stirred for 2 h, allowed to stand for 16 h, collected by vacuum
filtration, washed with boiling distilled water (300 mL) and a mixture of
distilled water and methanol (1:1, 300 mL), dried under vacuum at 908C
for 24 h, dissolved in the minimal amount of CHCl3/MeOH=80:1, and
purified by column separation (short column of neutral alumina, CHCl3/
MeOH=80:1, dark red first fraction). Yield: 504 mg (93%); m.p.
>3008C; Rf (CHCl3/MeOH=60:1): 0.4; Rf (silica gel, CHCl3/acetone
10:1)=0.79; 1H NMR (600 MHz, CDCl3, 278C, TMS): d=8.73–8.62 (m,
132.1, 135.3, 149.6, 154.2, 163.5, 163.9 ppm; IR (KBr): n˜ =3435 (brm),
3081 (brw), 2954 (m), 2927 (m), 2856 (m), 1711 (s), 1698 (s), 1659 (vs),
1618 (w), 1595 (s), 1579 (m), 1508 (s), 1466 (m), 1432 (m), 1405 (m), 1344
(vs), 1315 (m), 1275 (m), 1254 (s), 1232 (m), 1217 (m), 1198 (w), 1177
(m), 1148 (vw), 1127 (w), 1107 (w), 1042 (w), 966 (w), 854 (w), 810 (m),
777 (vw), 749 (m), 729 (vw), 648 (w), 608 (vw), 496 (w), 430 cmꢁ1 (w);
UV/Vis (CHCl3): lmax (Erel)=526.1 (l), 489.8 0.58), 458.2 nm (0.21); fluo-
rescence (CHCl3): lmax =533, 576 nm; fluorescence quantum yield (lexc
=
489 nm, c=1.0ꢁ10ꢁ6 molLꢁ1 in chloroform, reference 2,9-bis-(1-hexyl-
heptyl)anthra[2,1,9-def;6,5,10-d’e’f’]diisoquinoline-1,3,8,10-tetraone[4] with
F=1.0): 0.02; MS (70 eV): m/z (%): 708 (100) [M+], 691 (6) [M+ꢁOH],
677 (16) [M+ꢁOMe], 527 (34) [M+ꢁC13H25], 495 (78) [677ꢁC13H26], 390
(3) [495ꢁC6H2OMe], 373 (8) [390ꢁOH], 435 (3) [373ꢁCO]; elemental
analysis calcd (%) for C45H44N2O6: C 76.25, H 6.26, N 3.95; found C
76.41, H 6.23, N 3.91.
8H; CHperylene), 7.28 (d, 3J
(H,H)=8.9 Hz, 2H; CHarom.), 5.21–5.16 (m, 1H; CH), 3.89 (s, 3H; O-
CH3), 2.28–2.22 (m, 2H; CH2), 1.91–1.85 (m, 2H; CH2), 1.38–1.19 (m,
16H; 8ꢁCH2), 0.83 ppm (t, 3J(H,H)=7.1 Hz, 6H; CH3); 13C NMR
(H,H)=8.9 Hz, 2H; CHarom.), 7.09 (d, 3J-
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
(150 MHz, CDCl3, 27.08C): d=163.8, 159.7, 135.0, 134.3, 131.8, 131.1,
129.8, 129.5, 127.5, 126.6, 126.4, 123.4, 123.2, 123.0, 114.8, 55.5, 54.8, 32.4,
31.7, 29.2, 26.9, 22.6, 14.0 ppm; IR (ATR): n˜ =2954 (w), 2925 (m), 2856
(w), 1697 (s), 1655 (s), 1594 (s), 1577 (m), 1510 (s), 1484 (w), 1459 (w),
1434 (w), 1404 (m), 1342 (s), 1301 (m), 1249 (s), 1176 (m), 1145 (w), 1124
(w), 1107 (w), 1030 (m), 966 (m), 849 (m), 832 (m), 809 (s), 794 (s), 744
(s), 704 (w), 694 (w), 638 (w), 616 (w), 607 cmꢁ1 (w); UV/Vis (CHCl3):
lmax (Erel)=459.2 (0.22), 491.0 (0.60), 527.0 nm (1.00); fluorescence
(CHCl3): lmax (I)=534.3 (1.00), 577.5 (0.52); fluorescence quantum yield:
(CHCl3, lexc =491 nm, E491nm/1 cm =0.01587, reference: 2,9-bis-(1-hexylhep-
Compound PI-M6: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (57 mg, 0.30 mmol) and 2,4-
dimethoxyaniline (54 mg, 0.40 mmol) were allowed to react and the prod-
uct was purified as was described for PI-M3. Yield: 45 mg (63%); m.p.
>2508C; Rf (silica gel, CHCl3/acetone 10:1)=0.76; 1H NMR (300 MHz,
CDCl3): d=0.76 (t, J
ACHTUNGTREN(NNUG H,H)=6.6 Hz, 6H, 2ꢁCH3), 1.17–1.26 (m, 16H;
8ꢁCH2), 1.77–1.83 (m, 2H; a-CH2), 2.10–2.24 (m, 2H; a-CH2), 3.71 (s;
ꢁ
ꢁ
CH3, OCH), 3.81 (s, 3H; OCH3), 5.08–5.18 (m, 1H; N-CH), 6.58–6.60
(m, 1H; CHaryl), 7.13–7.19 (m, 2H; 2ꢁCHaryl), 8.52–8.65 ppm (m, 8H;
CHperylene); 13C NMR (75 MHz, CDCl3): d=4.4 (CH3), 23.0 (CH2), 27.4
tyl)anthra[2,1,9-def;6,5,10-d’e’f’]diisoquinoline-1,3,8,10-tetraone[4]
with
ꢁ
(CH2), 29.6 (CH2), 32.2 (CH2), 32.8 (CH2), 55.2 (CH), 56.0, ( OCH3),
F=1.0): 0.08; MS (DEP/EI): m/z (%): 678 (51) [M+], 496 (100) [M+
ꢁC13H26]; HRMS: m/z: calcd for C44H42N2O5: 678.3094; found: 678.3098,
D=0.0004; elemental analysis calcd (%) for C44H42N2O5: C 77.85, H 6.24,
N 4.13; found: C 77.80, H 6.17, N 4.13.
100.2, 105.3, 117.1, 123.4, 123.6, 126.8, 127.1, 130.0, 130.6, 131.5, 132.1,
134.8, 135.3, 156.2, 161.8, 163.8 ppm; IR (KBr): n˜ =3435 (brw), 2926 (m),
2856 (m), 1698 (vs), 1594 (m), 1513 (m), 1465 (m), 1437 (M9; 1405 (m),
1346 (vs), 1318 (m), 1288 (m), 1254 (s), 1210 (m), 1176 (m), 1117 (m),
1034 (w), 965 (w), 854 (w), 832 (w), 811 (m), 791 (w), 748 (m), 593 (w),
495 (w), 431 cmꢁ1 (w); UV/Vis (CHCl3): lmax (Erel)=526.1 (1.0), 489.5
(0.60), 459.0 nm (0.22); fluorescence (CHCl3): lmax =534, 575 nm; fluores-
cence quantum yield: (lexc =489 nm, c=7.2ꢁ10ꢁ6 molLꢁ1 in chloroform,
reference 2,9-bis-(1-hexylheptyl)anthra[2,1,9-def;6,5,10-d’e’f’]diisoquino-
line-1,3,8,10-tetraone[4] with F=1.0): 0.04; MS (70 eV): m/z (%): 708
(100) [M+], 691 (12) [M+ꢁOH], 677 [M+ꢁOMe], 527 (34) [M+
ꢁC13H25], 495 (78) [677ꢁC13H26], 390 (14) [495ꢁC6H2OMe], 373 (12)
[390ꢁOH], 345 (5) [373ꢁCO]; elemental analysis calcd (%) for
C45H44N2O6: C 76.25, H, 6.26, N 3.95; found: C 76.10, H 6.28, N 3.89.
Compound PI-M2: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (460 mg, 800 mmol), 3-anisi-
dine (194 mg, 1.58 mmol), zinc acetate (30 mg, 0.16 mmol), and quinoline
(7 mL) were allowed to react and the product was purified as was de-
scribed for PI-M1 (5 h, 2108C, dried at 908C). Yield: 476 mg (88%); m.p.
>3008C; Rf (CHCl3/MeOH=60:1)=0.5; Rf (silica gel, CHCl3/acetone
10:1)=0.78; 1H NMR (600 MHz, CDCl3, 278C, TMS): d=8.74–8.62 (m,
8H; CHperylene), 7.51–7.47 (m, 1H; CHarom.), 7.06 (ddd, 3J
4J(H,H)=2.5 Hz, 5J
(H,H)=0.9 Hz, 1H; CHarom.), 6.96–6.93 (m, 1H;
ACHTUNGTREN(NUNG H,H)=8.5 Hz,
A
ACHTUNGTRENNUNG
CHarom.), 6.91–6.90 (m, 1H; CHarom.), 5.22–5.15 (m, 1H; CH), 3.85 (s, 3H;
CH3), 2.29–2.20 (m, 2H; CH2), 1.91–1.83 (m, 2H; CH2), 1.38–1.20 (m,
ACTHNUTRGNEUNG
16H; 8ꢁCH2), 0.82 ppm (t, 3J(H,H)=6.9 Hz, 6H; CH3); 13C NMR
Compound PI-M4: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (57 mg, 0.10 mmol) and 3,4-
dimethoxyaniline (45 mg, 0.30 mmol) were allowed to react and the prod-
uct was purified as was described for PI-M3. Yield: 50 mg (71%), m.p.
>2508C; Rf (silica gel, CHCl3/acetone 10:1)=0.71; 1H NMR (300 MHz,
(150 MHz, CDCl3, 27.08C): d=163.5, 160.5, 136.1, 135.1, 134.3, 131.8,
130.1, 129.8, 129.5, 126.7, 126.4, 123.3, 123.0, 120.7, 115.0, 114.2, 55.4,
54.8, 32.4, 31.7, 29.2, 26.9, 22.6, 14.0 ppm; IR (ATR): n˜ =2956 (w), 2924
(m), 2856 (w), 1696 (s), 1658 (s), 1594 (s), 1577 (m), 1506 (w), 1487 (w),
1456 (w), 1435 (w), 1404 (m), 1345 (s), 1320 (w), 1256 (s), 1214 (w), 1180
(m), 1125 (w), 1110 (w), 1035 (w), 966 (w), 858 (m), 810 (s), 796 (s), 778
(m), 764 (w), 744 (s), 704 (w), 686 (m), 638 (w), 609 cmꢁ1 (w); UV/Vis
(CHCl3): lmax (Erel)=458.8 (0.24), 490.0 (0.61), 527.0 nm (1.00); fluores-
cence (CHCl3): lmax (I)=534.5 (1.00), 577.5 (0.51); fluorescence quantum
yield (CHCl3, lexc =490 nm, E490nm/1 cm =0.01423, reference: 2,9-bis-(1-hex-
ylheptyl)anthra[2,1,9-def;6,5,10-d’e’f’]diisoquinoline-1,3,8,10-tetraone[4]
with F=1.0): 0.77; MS (DEP/EI): m/z (%): 678 (51) [M+], 496 (100)
[M+ꢁC13H26]; HRMS: m/z: calcd for C44H42N2O5: 678.3094; found:
678.3097, D=0.0003; elemental analysis calcd (%) for C44H42N2O5: C
77.85, H 6.24, N 4.13; found C 78.06, H 6.32, N 4.15.
CDCl3): d=0.76 (t, JACHTNUTRGNENUG(H,H)=6.6 Hz, 6H, 2ꢁCH3), 1.17–1.26 (m, 16H;
ꢁ
8ꢁCH2), 1.78–1.85 (m, 2H; a-CH2), 2.15–2.22 (m, 2H; a-CH2), 3.83 (s;
ꢁ
ꢁ
OCH3), 3.89 (s; OCH), 5.01–5.16 (m, 1H; N-CH), 6.83–6.99 (m, 1H;
CHaryl), 7.19 (s, 1H; CHaryl), 8.47–8.63 ppm (m, 8H; CHperylene); 13C NMR
(75 MHz, CDCl3): d=14.4 (CH3), 23.0 (CH2), 27.4 (CH2), 29.6 (CH2),
ꢁ
ꢁ
30.1 (CH2), 32.2, (CH), 32.8, 55.3 ( OCH3), 56.4 ( OCH3), 56.5 (CH),
111.8, 112.2, 122.1, 123.4, 123.6, 126.6, 126.9, 128.1, 129.8, 130.0, 121.4,
132.0, 134.5, 135.3, 149.7, 150.1, 164.1 ppm; IR (KBr): n˜ =3436 (brm),
3078 (brw), 2956 (m), 2926 (m), 2856 (m), 1698 (s), 1660 (vs), 1595 (s),
1578 (m), 1514 (m), 1462 (w), 1433 (w), 1404 (m), 1346 (s), 1305 (w),
1254 (m), 1220 (w), 1176 (m), 1122 (m), 1028 (w), 966 (w), 860 (w), 810
(m), 794 (w), 746 (m), 619 (w), 490 (w), 431 cmꢁ1 (w); UV/Vis (CHCl3):
lmax (Erel)=526.7 (1), 490.2 (0.60), 459.0 nm (0.22); fluorescence (CHCl3):
lmax =534, 576 nm; fluorescence quantum yield: (lexc =489 nm, c=7.0ꢁ
10ꢁ6 molLꢁ1 in chloroform, reference 2,9-bis-(1-hexylheptyl)anthra[2,1,9-
def;6,5,10-d’e’f’]diisoquinoline-1,3,8,10-tetraone[4] with F=1.0): 0.02; MS
(70 eV): m/z (%): 708 (78) [M+], 691 (6), 526 (100) [M+ꢁC13H26], 373
(36) [526ꢁC6H2Me2OH], 345 (5) [373ꢁCO]; elemental analysis calcd
(%) for C45H44N2O6: C 76.25, H 6.26, N 3.95; found C 76.50, H 6.14, N
3.94.
Compound PI-M8: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (172 mg, 0.30 mmol) and 2,5-
dimethoxyaniline (230 mg, 1.50 mmol) were allowed to react and the
product was purified as was described for PI-M3. Yield: 140 mg (66%);
m.p. >2508C; Rf (silica gel, CHCl3/acetone 10:1)=0.79; 1H NMR
(300 MHz, CDCl3): d=0.85 (t, J
ACHTUNGTERN(NNUG H,H)=6.6 Hz, 6H; 2ꢁCH3), 1.26–1.35
(m, 16H; 8ꢁCH2), 1.85–1.92 (m, 2H; a-CH2), 2.22–2.32 (m, 2H; a-CH2),
ꢁ
ꢁ
3.78 (s; CH3, OCH3), 3.83 (s, 3H; CH3), 5.16–5.26 (m, 1H; N-CH),
6.94–7.28 (m, 3H; CHaryl), 8.60–8.74 ppm (m, 8H; CHperylene); 13C NMR
(75 MHz, CDCl3): d=14.4 (CH3), 23.0 (CH2), 27.3, (CH2), 29.6 (CH2),
Compound PI-M5: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (570 mg, 1.00 mmol) and 3,5-
dimethoxyaniline (770 mg, 5.00 mmol) were allowed to react and the
ꢁ
ꢁ
32.2 (CH2), 32.8 (CH2), 55.2 (CH), 56.2 ( OCH3), 56.8 ( OCH3), 113.4,
116.0, 116.1, 123.4, 123.6, 123.9, 124.7, 126.8, 127.1, 129.9, 130.4, 131.5,
13414
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416

Bisimide Luminescence
FULL PAPER
product was purified as was described for PI-M3. Yield: 540 mg (76%);
m.p. >2508C; Rf (silica gel, CHCl3/acetone 10:1)=0.81; 1H NMR
(w), 853 (w), 810 (m), 796 (w), 747 (m), 723 (w), 677 (m), 498 (w),
431 cmꢁ1 (w); UV (CHCl3): lmax (Erel)=527.6 (100), 490.9 (58.8),
459.8 nm (19.8); MS (70 eV): m/z (%): 738 (100) [M+], 569 (2.51) [M+
ꢁC9H11O3], 556 (49.5) [M+ꢁC13H26]; elemental analysis calcd (%) for
C46H46N2O7: C 74.82, H 6.46, N 3.73; found C 74.78, H 6.28, N 3.79.
(300 MHz, CDCl 3): d=0.85 (t, J
ACHTUNGTERN(NNUG H,H)=6.6 Hz, 6H; 2ꢁCH3), 1.26–1.36
(m, 16H; 8ꢁCH2), 1.84–1.96 (m, 2H; a-CH2), 2.21–2.34 (m, 2H; a-CH2),
ꢁ
ꢁ
3.85 (s, 6H; OCH3), 5.16–5.26 (m, 1H; NCH), 6.54 (d, JACTHUNTGRNEUNG(H,H)=
2.1 Hz, 2H; CHortho-aryl), 6.63 (t, 1H; CHpara-aryl), 8.63–8.76 ppm (m, 8H;
Spectroscopy and photophysics: Spectrophotometric-grade toluene and
dichloromethane (C. Erba) were used as supplied. Absorption spectra
were measured in 10 mm cells using a Perkin–Elmer Lambda 950 UV/
Vis spectrophotometer. A Spex Fluorolog II spectrofluorimeter was used
to acquire fluorescence spectra from air-equilibrated solutions in 10 mm
fluorescence cells. Experiments at 77 K made use of capillary tubes im-
mersed in liquid nitrogen contained in a homemade quartz dewar. The
reported luminescence spectra are uncorrected; emission quantum yields
were determined after correction for the photomultiplier response, with
reference to an air-equilibrated solution of N,N’-bis-(1-hexylheptyl)-
CHperylene); 13C NMR (75 MHz, CDCl3): d=1.4.4 (CH3), 23.0 (CH2), 27.3
ꢁ
(CH2), 29.6 (CH2), 32.1 (CH2), 32.8 (CH2), 55.2 (CH), 55.9 ( OCH3),
101.9, 107.2, 123.4, 123.7, 126.8, 127.1, 129.9, 130.2, 132.2, 134.7, 135.5,
137.1, 161.8, 163.8 ppm; IR (KBr): n˜ =3430 (brw), 3098 (brw), 2954 (m),
2927 (m), 2857 (m), 1770 (w); 1699 (s), 1662 (vs), 1594 (vs), 1578 (m),
1506 (m), 1465 (m), 1431 (m), 1405 (m), 1354 (vs), 1336 (s), 1322 (m),
1254 (m), 1218 (w), 1205 (m), 1176 (m); 1156 (m), 1127 (w), 1108 (w),
1063 (w), 966 w), 926 (vw), 852 (w), 811, (m), 796 (w), 747 (m), 728 w),
682 (w), 616 (vw), 590 (vw), 536 (vw), 496 (w), 431 cmꢁ1 (w); UV/Vis
(CHCl3): lmax (Erel)=526.8 (l), 489.8 (0.59), 458.9 (0.22) nm; fluorescence
(CHCl3): lmax =533, 575 nm; fluorescence quantum yield (lexc =489 nm,
3,4:9,10-perylenebis(dicarboxymide) in dichloromethane with
a Ffl =
0.99.[5] Luminescence lifetimes in the nanosecond range were obtained
using IBH single-photon counting equipment with excitation at 465 nm
from pulsed diode sources (resolution 0.3 ns). The lifetimes in the pico-
second range were measured using an apparatus based on a Nd:YAG
laser (35 ps pulse duration, 532 nm, 1.5 mJ) and a Streak Camera Hama-
matsu instrument. After deconvolution the resolution time is 5 ps. More
details can be found elsewhere.[23] Transient absorbances in the picosec-
ond range were measured using a pump and probe system based on a
Nd:YAG laser (Continuum PY62/10, 35 ps pulse, 532 nm, 3.5 mJ). Air-sa-
turated solutions with an absorbance of 0.65 at 532 nm were employed.
More details on the apparatus can be found elsewhere.[24] Estimated
errors are 10% on exponential lifetimes, 20% on quantum yields, 10%
on molar-absorption coefficients, 2 nm on emission- and ground- state ab-
sorption peaks, and 10 nm in transient-absorbance peaks. If not otherwise
specified, experiments were carried out at 295 K.
c=4.2ꢁ10ꢁ6 molLꢁ1
,
reference
2,9-bis-(1-hexylheptyl)anthra[2,1,9-
def;6,5,10-d’e’f’]diisoquinoline-1,3,8,10-tetraone[4] with F=1.0): 1.0; MS
(70 eV): m/z (%): 708 (48) [M+], 691 (6), 526 (100) [M+ꢁC13H26], 495
(2) [527ꢁC13H26OMe], 391 (4), 373 (8) [390ꢁOH]; elemental analysis
calcd (%) for C45H44N2O6: C 76.25, H 6.26, N 3.95; found: C 76.56, H
6.32, N 3.86.
Compound PI-M7: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (172 mg, 0.30 mmol) and 2,6-
dimethoxyaniline (138 mg, 0.90 mmol) were allowed to react and the
product was purified as was described for PI-M3. Yield: 210 mg (97%);
m.p. >2508C; Rf (silica gel, CHCl3/acetone 10:1)=0.79; 1H NMR
(600 MHz, CDCl3): d=0.88 (t, JACHTUNTRGNE(UNG H,H)=6.6 Hz, 6H; 2ꢁCH3), 1.28–1.39
(m, 16H; 8ꢁCH2), 1.90–1.94 (m, 2H; a-CH2), 2.27–2.33 (m, 2H; a-CH2),
ꢁ
3.84 (s; OCH3), 5.22–5.27 (m, 1H; N-CH), 6.78–6.79 (d, J
G
8.4 Hz, 1H; CHaryl), 7.31 (s, 1H; CHaryl), 7.46–7.49 (t, 1H; CHaryl),
8.73 ppm (dd, 8H; CHperylene); 13C NMR (150.9 MHz, CDCl3): d=14.4
(CH3), 23.0 (CH2), 27.4 (CH2), 29.6 (CH2), 32.2 (CH2), 32.8 (CH2), 55.2
Molecular modeling: Fully optimized using the DFT B3LYP basis set 3-
21G.
ꢁ
(CH), 56.5 ( OCH3), 104.9, 112.8, 123.4, 123.5, 126.9, 127.2, 130.0, 130.6,
130.8, 132.1, 135.2, 156.6, 163.2 ppm; IR (KBr): n˜ =3436 (brm), 2954 (m),
2927 (m), 2856 (w), 1714 (m), 1698 (s), 1658 (vs), 1595 (vs), 1579 (m),
1500 (w), 1481 (m), 1458 (w), 1434 (w), 1405 (m),l 1344 (vs), 1307 w),
1256 (m), 1200 (w), 1176 (w), 1140 (w), 1112 (m), 1034 (vw), 965 (w), 842
(w), 810 (m), 771 (w), 747 (m), 724 (w), 696 (vw), 614 (w), 494 (w),
432 cmꢁ1 (w); UV/Vis (CHCl3): lmax (Erel)=526.1 (1), 489.2 (0.61),
458.2 nm (0.21); fluorescence (CHCl3): lmax =534, 575 nm; fluorescence
quantum yield: (lexc =489 nm, c=4.2ꢁ10ꢁ6 molLꢁ1 in chloroform, refer-
Acknowledgements
This work was supported by CIPS cluster, the Fonds der Chemischen In-
dustrie, and by the CNR of Italy (PM.P04.010 MACOL).
ence
2,9-bis-(1-hexylheptyl)anthra[2,1,9-def;6,5,10-d’e’f’]diisoquinoline-
1,3,8,10-tetraone[4] with F=1.0): 1.0; MS (70 eV): m/z (%): 708 (100),
[M+], 691 (13) [M+ꢁOH], 623 (4), 539 (6), 527 (78) [M+ꢁC13H25], 495
(40) [677ꢁC13H26], 465 (9) [M+ꢁOMe, ꢁC13H28], 390 (67)
[495ꢁC6H2OMe], 373 (5) [390ꢁOH], 345 (3) [373ꢁCO]; elemental anal-
ysis calcd (%) for C45H44N2O6: C 76.25, H 6.26, N 3.95; found C 76.44, H
6.27, N 3.86.
[1] See the following reviews and references therein: a) A. P. de Silva,
T. S. Moody, G. D. Wright, Analyst 2009, 134, 2385–2393; b) T.
Gunnlaugsson, M. Glynn, G. M. Tocci, P. E. Kruger, F. M. Pfeffer,
Coord. Chem. Rev. 2006, 250, 3094–3117; c) C. W. Rogers, M. O.
Wolf, Coord. Chem. Rev. 2002, 233, 341–350.
[2] See the following reviews and references therein: a) A. Credi,
Angew. Chem. 2007, 119, 5568–5572; Angew. Chem. Int. Ed. 2007,
46, 5472–5475; b) A. P. de Silva, S. Uchiyama, Nat. Nanotechnol.
2007, 2, 399–410; c) A. P. de Silva, T. P. Vance, M. E. S. West, G. D.
Wright, Org. Biomol. Chem. 2008, 6, 2468–2481.
[3] a) H. Langhals, Helv. Chim. Acta 2005, 88, 1309–1343; b) H. Langh-
als, Heterocycles 1995, 40, 477–500; c) H. Langhals, Molecular Devi-
ces. Chiral Bichromophoric Silicones: Ordering Principles in Com-
plex Molecules in Silicon Based Polymers (Eds.. F. Ganachaud, S.
Boileau, B. Boury), Springer, Heidelberg, 2008, pp. 51–63.
Compound PI-M9: 9-(1-Hexylheptyl)-2-benzopyrano[6’,5’,4’:10,5,6]an-
thra[2,1,9-def]isoquinoline-1,3,8,10-tetraone (980 mg, 1.7 mmol), 3,4,5-tri-
methoxyaniline (930 mg, 5.1 mmol), a catalytic amount of zinc acetate di-
hydate, and imidazole (20 g) were heated at reflux (bath 1308C) for 4 h
and then allowed to cool. While still warm, the solution was treated with
2n aqueous HCl (110 mL), collected by vacuum filtration, dried in air
(1208C, 8 h), and purified by column separation (silica gel, chloroform/
aetone 15:1). Yield: 1.12 g (84%); m.p. >2508C; Rf (silica gel/chloro-
form–acetone 15:1): 0.4; 1H NMR (CDCl3): d=8.70 (d, J
2H; perylene), 8.65 (d, J(H,H)=8 Hz, 2H; perylene), 8.56 (d, J
3 Hz, 2H; perylene), 8.53 (d, J(H,H)=3 Hz, 2H; perylene), 6.64 (s, 2H;
ACHTUNGTRENNUNG
A
U
[4] H. Langhals, J. Karolin, L. B.-ꢂ. Johansson, J. Chem. Soc. Faraday
Trans. 1998, 94, 2919–2922.
AHCTUNGTRENNUNG
phenyl), 5.21–5.16 (m, 1H; CH), 3.96 (s, 3H; phenyl), 3.90 (s, 6H;
phenyl), 2.32–2.24 (m, 2H; CH2), 1.96–1.88 (m, 2H; CH2), 1.41–1.25 (m,
[5] a) H. Langhals, A. J. Esterbauer, Chem. Eur. J. 2009, 15, 4793–4796;
b) F. Wꢃrthner, Chem. Commun. 2004, 1564–1579; c) F. Ilhan, D. S.
Tyson, D. J. Stasko, K. Kirschbaum, M. A. Meador, J. Am. Chem.
Soc. 2006, 128, 702–703; d) S. Vajiravelu, L. Ramunas, G. J. Vidas,
G. Valentas, J. Vygintas, S. Valiyaveettil, J. Mater. Chem. 2009, 19,
4268–4275; e) J. Fortage, M. Sꢄverac, C. Houarner-Rassin, Y. Pelle-
grin, E. Blart, F. Odobel, J. Photochem. Photobiol. A 2008, 197,
156–169; f) H. Langhals, A. Obermeier, Eur. J. Org. Chem. 2008,
16H; CH2), 0.86 ppm (t, JACTHNUGTRNENUG ACHTUNGTERN(NUGN CDCl3):
(H,H)=7 Hz, 6H, 2ꢁCH3); 13C NMR
ꢁ
d=163.9 (C O), 154.3, 138.6, 135.4, 134.4, 132.1, 132.1, 131.0, 130.0,
129.8, 126.9, 126.6, 123.6, 123.5, 123.4, 106.3, 61.3, 56.6, 55.3, 32.8, 32.2,
29.6, 27.4, 23.0, 14.4 ppm; IR (KBr): n˜ =3436 (s), 2928 (s), 2857 (m), 1699
(s), 1661 (s), 1595 (s), 1578 (w), 1506 (w), 1464 (w), 1433 (w), 1422 (w),
1405 (m), 1355 (m), 1328 (s), 1254 (m), 1176 (w), 1125 (m), 1008 (w), 965
Chem. Eur. J. 2010, 16, 13406 – 13416
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13415

H. Langhals, L. Flamigni et al.
6144–6151; g) H. Langhals, S. Poxleitner, O. Krotz, T. Pust, A.
Walter, Eur. J. Org. Chem. 2008, 4559–4562.
[13] P. B. Merkel, P. Luo, J. P. Dinnocenzo, S. Farid, J. Org. Chem. 2009,
74, 5163–5173.
[6] H. Langhals, A Obermeier, Y. Floredo, A. Zanelli, L. Flamigni,
Chem. Eur. J. 2009, 15, 12733–12744.
[14] A. Zweig, W. G. Hodgson, W. H. Jura, J. Am. Chem. Soc. 1964, 86,
4124–4129.
[7] a) A. S. Lukas, Y. Zhao, S. E. Miller, M. R. Wasielewski, J. Phys.
Chem. B 2002, 106, 1299–1306; b) F. Wꢃrthner, A. Sautter, D.
Schmid, P. J. A. Weber, Chem. Eur. J. 2001, 7, 894–902; c) W. E.
Ford, H. Hiratsuka, P. V. Kamat, J. Phys. Chem. 1989, 93, 6692–
6696.
[8] a) L. Flamigni, B. Ventura, M. Tasior, T. Becherer, H. Langhals,
D. T. Gryko, Chem. Eur. J. 2008, 14, 169–183; b) L. Flamigni, B.
Ventura, C.-C. You, C. Hippius, F. Wꢃrthner, J. Phys. Chem. C 2007,
111, 622–630; c) A. I. Oliva, B. Ventura, F. Wꢃrthner, A. Camara-
Campos, C. A. Hunter, P. Ballester, L. Flamigni, Dalton Trans. 2009,
4023–4037; d) M. Tasior, D. T. Gryko, J. Shen, K. M. Kadish, T.
Becherer, H. Langhals, B. Ventura, L. Flamigni, J. Phys. Chem. C
2008, 112, 19699–19709.
[9] a) S. Xiao, Y. Li, Y. Li, J. Zhuang. N. Wang, H. liu, B. Ning, Y. liu, F.
Lu, L. Fan, C. Yang, Y. Li, D. Zhu, J. Phys. Chem. B 2004, 108,
16677–16685; b) W. Xu, H. Chen, Y. Wang, C. Zhao, X. Li, S.
Wang, Y. Weng, ChemPhysChem 2008, 9, 1409–1415; c) B. Rybtch-
inski, L. E. Sinks, M. R. Wasielewski, J. Am. Chem. Soc. 2004, 126,
12268–12269; d) C. Flors, I. Oesterling, T. Schnitzler, E. Fron, G.
Schweitzer, M. Sliwa, A. Herrmann, M. van der Auweraer, F. C. de
Schryver, K. Mꢃllen, J. Hofkens, J. Phys. Chem. C 2007, 111, 4861–
4870; e) R. F. Kelley, W. S. Shin, B. Rybtchinski, L. E. Sinks, M. R.
Wasielewski, J. Am. Chem. Soc. 2007, 129, 3173–3181; f) S. I. Yang,
S. Prathapan, M. A. Miller, J. Seth, D. F. Bocian, J. S. Lindsey, D.
Holten, J. Phys. Chem. B 2001, 105, 8249–8258; g) M. J. Arhens,
R. F. Kelley, Z. E. X. Dance, M. R. Wasielewski, Phys. Chem. Chem.
Phys. 2007, 9, 1469–1478; h) A. Prodi, C. Chiorboli, F. Scandola, E.
Iengo, E. Alessio, R. Dobrawa, F. Wꢃrthner, J. Am. Chem. Soc.
2005, 127, 1454–1462; i) F. J. Cꢄspedez -Guirao, K. Ohkubo, S. Fu-
kuzumi, A. Sastre-Santos, F. Fernꢅndez-Lꢅzaro, J. Org. Chem. 2009,
74, 5871–5880; j) N. Van Anh, F. Schlosser, M. M. Groeneveld,
I. H. M. van Stokkum, F. Wꢃrthner, R. Williams, J. Phys. Chem. C
2009, 113, 18358–18368.
[15] M. Montalti, A. Credi, L. Prodi, M. T. Gandolfi, Handbook of Pho-
tochemistry, 3rd ed., CRC Press, Boca Raton, 2006, pp. 624–627.
[16] a) G. L. Gaines III, M. P. OꢆNeil, W. A. Svec, M. P. Niemczyk, M. R.
Wasielewski, J. Am. Chem. Soc. 1991, 113, 719–721; b) L. Flamigni,
E. Baranoff, J.-P. Collin, J.-P. Sauvage, Chem. Eur. J. 2006, 12, 6592–
6606.
[17] W. E. Ford, P. V. Kamat, J. Phys. Chem. 1987, 91, 6373–6380.
[18] In this case, the CS-state energy level relative to the ground state,
DG0CS
,
is calculated very simply from DG0CS =EoxꢁEred
, without
taking into account: 1) the Coulombic interaction and 2) the effect
of the solvents on redox potentials used. There is a partial compen-
sation of the two terms. The first is in fact a term that stabilizes the
CS state, higher in low-dielectric constant media; the second is a de-
stabilizing term, higher in low-dielectric media. The result is an
overall increase of the CS-state energy level in TL compared to
CH2Cl2, that is, of a decrease of the driving force for charge separa-
tion and an increase of the driving force for charge recombination
in TL relative to CH2Cl2. Given the level of approximation involved
in the present discussion, we consider the simplification acceptable
and convenient.
[19] R. A. Marcus, Angew. Chem. 1993, 105, 1161–1172; Angew. Chem.
Int. Ed. Engl. 1993, 32, 1111–1121, and references therein.
[20] P. O’Neill, S. Steenken, D. Schulte-Frohlinde, J. Phys. Chem. 1975,
79, 2773–2779.
[21] a) J. Salbeck, H. Kumkely, H. Langhals, R. W. Saalfrank, J. Daub,
Chimia 1989, 43, 6–9; b) T. Kircher, H. G. Lçhmannsrçben, Phys.
Chem. Chem. Phys. 1999, 1, 3987–3992.
[22] A. A. Rachford, S. Goeb, F. N. Castellano, J. Am. Chem. Soc. 2008,
130, 2766–2767.
[23] L. Flamigni, B. Ventura, A. I. Oliva, P. Ballester, Chem. Eur. J. 2008,
14, 4214–4224, and references therein.
[24] L. Flamigni, A. M. Talarico, S. Serroni, F. Puntoriero, M. J. Gunter,
M. R. Johnston, T. P. Jeynes, Chem. Eur. J. 2003, 9, 2649–2659, and
references therein.
[10] H. Kaiser, J. Lindner, H. Langhals, Chem. Ber. 1991, 124, 529–535.
[11] H. Langhals, Chem. Ber. 1985, 118, 4641–4645.
[12] H. Langhals, S. Demmig, H. Huber, Spectrochim. Acta 1988, 44A,
1189–1193.
Received: May 28, 2010
Published online: October 7, 2010
13416
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13406 – 13416