Phenylene ethynylene-tethered perylene bisimide folda-dimer and folda-trimer: Investigations on folding features in ground and excited states

Article abstract of DOI:10.1002/chem.201405231

In this work, we have elucidated in detail the folding properties of two perylene bisimide (PBI) foldamers composed of two and three PBI units, respectively, attached to a phenylene ethynylene backbone. The folding behaviors of these new PBI folda-dimer a

Full text of DOI:10.1002/chem.201405231

DOI: 10.1002/chem.201405231
Full Paper
&
Foldamers |Hot Paper|
Phenylene Ethynylene-Tethered Perylene Bisimide Folda-Dimer
and Folda-Trimer: Investigations on Folding Features in Ground
and Excited States
Benjamin Fimmel,[a] Minjung Son,[b] Young Mo Sung,[b] Matthias Grꢀne,[a] Bernd Engels,*[c]
Dongho Kim,*[b] and Frank Wꢀrthner*[a]
Abstract: In this work, we have elucidated in detail the fold-
ing properties of two perylene bisimide (PBI) foldamers com-
posed of two and three PBI units, respectively, attached to
a phenylene ethynylene backbone. The folding behaviors of
these new PBI folda-dimer and trimer have been studied by
solvent-dependent UV/Vis absorption and 1D and 2D NMR
spectroscopy, revealing facile folding of both systems in
tetrahydrofuran (THF). In CHCl3 the dimer exists in extended
(unfolded) conformation, whereas partially folded confor-
mations are observed in the trimer. Temperature-dependent
1H NMR spectroscopic studies in [D8]THF revealed intramo-
lecular dynamic processes for both PBI foldamers due to, on
the one hand, hindered rotation around CꢀN imide bonds
and, on the other hand, backbone flapping; the latter pro-
cess being energetically more demanding as it was observed
only at elevated temperature. The structural features of
folded conformations of the dimer and trimer have been
elucidated by different 2D-NMR spectroscopy (e.g., ROESY
and DOSY) in [D8]THF. The energetics of folding processes
for the PBI dimer and trimer have been assessed by calcula-
tions applying various methods, particularly the semiempiri-
cal PM6-DH2 and the more sophisticated B97D approach, in
which relevant dispersion corrections are included. These
calculations corroborate the results of NMR spectroscopic
studies. Folding features in the excited states of these PBI
foldamers have been characterized by using time-resolved
fluorescence and transient absorption spectroscopy in THF
and CHCl3, exhibiting similar solvent-dependent behavior as
observed for the ground state. Interestingly, photoinduced
electron transfer (PET) process from electron-donating
backbone to electron-deficient PBI core for extended, but
not for folded, conformations was observed, which can
be explained by a fast relaxation of excited PBI stacks in the
folded conformation into fluorescent excimer states.
Introduction
into the backbone to allow for coiling into a structure with in-
tramolecular contacts, a multitude of possible conformations
usually show up and the supramolecular arrangement of the
constituent building blocks becomes out of control. Thus, the
co-existence of a large variety of conformers becomes ap-
parent.[3] In contrast, for selective backbones a transition from
randomly coiled structures (unfolded state) into a small set of
almost identical conformers (folded state) can be observed
due to intramolecular ordering of the repeat units attached to
the backbone.[4] The issue of folding of polypeptides into
precise three-dimensional structures has been intensively eluci-
dated, but still the de novo design of synthetic proteins with
Although the continuing progress in synthetic methodology
and analytical methods enables the construction of defined
molecules of greater size, chemists have to admit that the vast
majority of their elaborate large-sized molecules are of irregu-
lar shape, even if the molecules are monodisperse due to the
precise synthetic protocols applied. Among the few exceptions
are the rather stiff conjugated oligomers[1] and shape-persis-
tent macrocycles,[2] in which intramolecular contacts between
non-neighboring building blocks are prohibited by the stiffness
of the molecular backbone. When flexible units are introduced
[a] B. Fimmel, Dr. M. Grꢀne, Prof. Dr. F. Wꢀrthner
Institut fꢀr Organische Chemie & Center for Nanosystems Chemistry
Universitꢁt Wꢀrzburg
E-mail: dongho@yonsei.ac.kr
[c] Prof. Dr. B. Engels
Institut fꢀr Physikalische und Theoretische Chemie
Universitꢁt Wꢀrzburg
Emil-Fischer-Strasse 42, 97074 Wꢀrzburg (Germany)
Fax: (+49)931-31-85331
Am Hubland, 97074 Wꢀrzburg (Germany)
Fax: (+49)931-31-84756
E-mail: wuerthner@chemie.uni-wuerzburg.de
[b] M. Son, Y. M. Sung, Prof. Dr. D. Kim
E-mail: bernd.engels@uni-wuerzburg.de
Spectroscopy Laboratory for Functional p-Electronic Systems (FPIES) and
Department of Chemistry, Yonsei University
Shinchon-dong 134, Seoul 120-749 (Korea)
Fax: (+82)2-2123-2434
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405231.
Chem. Eur. J. 2014, 20, 1 – 17
1
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
functions that are prevalent in natural systems remains a big
challenge.[5] Nevertheless, such natural biomacromolecules
have strongly inspired chemists and motivated the search for
synthetic backbones that may fold into desirable three-dimen-
sional architectures, being reminiscent of their natural counter-
parts. Among the first examples of such synthetic systems are
the naphthalene diimide dihydroxynaphthalene aedamers of
Lokey and Iverson,[6] the phenylene ethynylene oligomers by
Moore et al.,[7] and oligopyridinecarboxamides by Huc and
Lehn.[8] Recently, Huc and co-workers have convincingly dem-
onstrated by combined NMR spectroscopy and single-crystal
X-ray analysis that large synthetic scaffolds may also organize
into defined three-dimensional architectures, which are consid-
ered to be crucial if specific and highly sophisticated functions
beyond simple molecular recognition or light harvesting were
to be achieved.[9]
However, the polydispersity restricted the structural
elucidation, particularly by NMR spectroscopic methods.
Therefore, we altered our approach and decided to synthe-
size small but monodisperse PBI oligomers and study their
folding properties as model systems to elucidate structure–
property relationships. Here, we present our comprehensive in-
vestigation on folding features of newly designed folda-dimer
2 and folda-trimer 3 (Figure 1). We have explored in detail the
folding properties of PBI 2 and 3 in the ground and excited
state in different solvents by steady state and time-resolved
spectroscopy, and theoretical calculations as well. Our studies
revealed that, through external control over the spatial
arrangement of dye ensembles by the nature of solvent, folda-
dimer 2 and folda-trimer 3 can be triggered to adopt desired
and predictable conformations in which the PBI moieties are
arranged either randomly or in H-type stacked states.
Self-assembly of perylene bisimide (PBI) dyes has been a sub-
ject of extensive investigations over more than a decade.[10]
Whereas the intermolecular self-assembly approach enabled us
to synthesize large PBI-based supramolecular architectures
with fascinating functions such as photoinduced charge
separation in supramolecular p–n-heterojunctions,[11] charge
transport in columnar stacks,[12] and exciton migration in
J-aggregates,[13] we realized that it is rather elusive with
currently available analytical tools to achieve unambiguous
information on the exact supramolecular structure as desired
to derive structure–property relationship guidelines for a ration-
al design of advanced functional materials. Therefore, we paid
attention to covalent scaffolds that preorganize PBI dyes in
a suitable way with regard to intramolecular folding into
specific three-dimensional architectures.[14] Previously, we have
investigated a phenylene ethynylene-tethered PBI oligomer
1 (Figure 1) and demonstrated solvent-dependent p–p-stack-
ing of the backbone-embedded PBI units. This was attributed
to a folding into a helical PBI p-stack surrounded by the alter-
nating ortho–meta-(phenylene ethynylene) conjugated back-
bone (with structural resemblance to DNA double helices).
Results and Discussion
Synthesis
The target phenylene ethynylene tethered folda-dimer 2 and
folda-trimer 3 were synthesized according to the routes
depicted in Scheme 1. Starting with the literature-known 4,5-
diiodo-1,2-di-n-hexylbenzene 4,[15] the monosilyl-protected
1,2-diethynyl-benzene derivative 7 was synthesized by two
sequential Pd/Cu-catalyzed Sonogashira–Hagihara reactions
with 2-methylbut-3-yne-2-ol and triisopropylsilylacetylene in
diisopropylamine under inert conditions, followed by
quantitative cleavage of the tertiary alcohol protective group
in precursor 6 with sodium hydroxide in a boiling toluene/
dioxane mixture in an overall yield of 63%.
Compound 7 was previously synthesized by a different
route;[14b] however, the present pathway using a polar tagging
strategy[16] resulted in significantly higher yield. In the next
step, 12-tricosanyl-substituted[17] (first mentioned as “swallow-
tailed”[18] alkyl substituents for smaller branched alkyl chains
Figure 1. Chemical structures of PBI oligomer 1 (nꢁ8),[14a] and new folda-dimer 2 and folda-trimer 3 investigated here.
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
2
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 1. Synthetic routes to folda-dimer 2 and folda-trimer 3 as well as reference molecule 15.
attached to steroids by Louw et al.) perylene-3,4:9,10-tetra-
carboxylic acid-3,4-anhydride-9,10-imide 8[19] was treated with
3-iodoaniline to obtain PBI 9 in a yield of 97% in analogy to
the literature-known reaction of symmetric 3,5-diiodoaniline
with 8 affording PBI 12,[14a] which was used to build up the
inner PBI moiety in folda-trimer 3. The unsymmetrical PBI dye
9 was subjected to Sonogashira–Hagihara coupling reaction
with 7 in degassed triethylamine (NEt3) and dry tetrahydrofur-
an (THF) in the presence of triphenylphosphine to obtain the
key intermediate 10 (65% yield) for the target compounds 2
and 3. The Sonogashira–Hagihara reaction of in situ-generated
terminal alkyne 11, which could also be isolated for characteri-
zation purposes, from 10 by using tetra-n-butylammonium
fluoride (TBAF) with either 9 or 12 under similar reaction con-
ditions as applied for the synthesis of 10, afforded the desired
folda-dimer 2 and folda-trimer 3 in 69 and 44% yields, respec-
tively. The compounds 2 and 3 were separated from byprod-
ucts and purified by gel permeation chromatography (GPC).
The syntheses of the reference compounds 9 and 15 (for struc-
tures, see Scheme 1) are described in the Supporting Informa-
tion. All new compounds, including the target compounds 2
and 3, were characterized by NMR spectroscopy, high-resolu-
tion mass spectrometry, and elemental analysis. The synthetic
details and the characterization data of new compounds folda-
dimer 2 and folda-trimer 3 as well as those of 5, 6, 7, 9, 10, 11,
13, 14, and 15 are given in the Supporting Information.
Solvent-dependent folding properties in the ground state
We have investigated the folding properties of the newly syn-
thesized folda-dimer 2 and folda-trimer 3 in the ground state
by UV/Vis absorption and 1H NMR spectroscopy. To ascertain
the suitable solvents or solvent combinations for these studies,
we have first measured UV/Vis absorption spectra of PBI 2 in
a broad variety of solvents ranging from nonpolar methylcyclo-
hexane to considerably polar chlorinated solvents and to
highly polar benzonitrile (Figures S1 and S2 in the Supporting
Information). Based on these solvent-dependent studies (for
details, see the Supporting Information) we have chosen
chloroform (CHCl3) and tetrahydrofuran (THF), and mixtures of
these two solvents for the elucidation of ground state folding
properties of 2 and 3.
The UV/Vis absorption spectra of folda-dimer 2 and folda-
trimer 3 in pure CHCl3 and THF and in different mixtures of
these solvents are shown in Figure 2. The spectrum of folda-
dimer 2 in pure CHCl3 exhibits two sharp absorption bands at
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
3
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
away from each other, this value was reported to be 1.3 in
CHCl3. For the photoswitchable PBI dimer of Feringa and co-
workers, this ratio was changed upon irradiation from 1.4 to
0.8 in THF.[22] The fact that PBI folda-dimer 2 shows A0ꢀ0/A1ꢀ0
values of 1.3 in CHCl3 and 0.8 in THF implies that this dimeric
PBI system exists in folded state in THF, whereas in the halo-
genated solvent it is unfolded and the data are in good
accordance with the literature-reported values mentioned
above.
Folda-trimer 3 exhibits similar solvent-dependent changes of
absorption spectra (Figure 2b) as observed for 2. In CHCl3,
again two dominant absorption bands can be seen at 531 and
495 nm with e values of 118300 and 124100mꢀ1 cmꢀ1, respec-
tively, along with a shoulder at 468 nm (Figure 2b). However,
the intensities of bands at 531 and 495 nm are already re-
versed compared with that of the corresponding bands of
folda-dimer 2 and the ratio of the two bands, A531nm/A495nm, for
3 is 0.95. These observations can be taken as an indication for
a stronger coupling between the PBI units in partially folded
conformations of 3 in CHCl3. Upon increasing the content of
THF both bands are hypochromically shifted (De down to
51600 and 14800mꢀ1 cmꢀ1, which correspond to an intensity
loss of 44 and 12% for the bands at 531 and 495 nm, respec-
tively). The ratio of A535nm/A495nm in pure THF is reduced to 0.61,
which is markedly smaller than that observed for folda-dimer
2. For a cofacial PBI trimer with a rigid xanthene spacer, a
Figure 2. Solvent-dependent UV/Vis absorption spectra of a) 2
(c=5ꢂ10ꢀ6 m) and b) 3 (c=6ꢂ10ꢀ6 m) in CHCl3/THF mixtures at 258C start-
ing in pure CHCl3 (d) changed to pure THF (c) by increasing THF con-
A
0ꢀ0/A1ꢀ0 ratio of 0.6 was reported by Wasielewski and
co-workers.[20] Thus, a folded conformation of 3 in THF can
be implied, whereas in CHCl3 this molecule is apparently in a
more unfolded state.
! ~
( ) and
tent in steps of 10 vol%. Inset: Plot of the absorption maxima A1
0
!
&
A0 ( ) against the content of THF (0 to 100 vol%) in CHCl3.
0
To assess whether any further self-assembly by intermolecu-
lar interactions between individual molecules occurs, tempera-
ture- and concentration-dependent UV/Vis studies were carried
out for folda-dimer 2 (the Supporting Information, Figures S4–
S6) and folda-trimer 3 (Figures S7 and S8) in CHCl3 and THF. In
both solvents, the temperature- and concentration-dependent
spectral changes for 2 and 3 are almost negligible. These ob-
servations suggest that, on the one hand, the PBI molecules
are appropriately solvated and do not build larger supramolec-
ular ensembles and, on the other hand, the intramolecularly
folded dimer 2 and trimer 3 are held together by rather strong
p–p interactions in THF as no spectral changes were observed
at higher temperatures (see the Supporting Information,
Figures S4a, S6, S7a, and S8).
529 and 493 nm with extinction coefficients of 123600 and
96100mꢀ1 cmꢀ1, respectively, along with a shoulder at 462 nm
and a broad slope in the region of 300 to 370 nm. Upon in-
creasing the content of THF in CHCl3/THF mixtures from 0 to
100% the lowest-energy band (at 529 nm) shows significant
changes with large hypochromic shifts (De down to
49800mꢀ1 cmꢀ1, which corresponds to a loss in intensity of
about 40%) and a small hypsochromic shift of 3 nm, whereas
the band at 493 nm remains nearly unchanged (De up to
4700mꢀ1 cmꢀ1, loss in intensity <5%). The plot of extinction
coefficients e of PBI 2 against the contents of THF in CHCl3
(Figure 2a, inset) reveals a gradual decrease of the e value for
the band at 529 nm with increasing of the THF content from 0
to 60%. For solvent mixtures consisting of>60% THF, the in-
tensities of both bands (at 529 and 493 nm) remain virtually
unchanged. Accordingly, the ratio of absorption maxima of
both prominent bands A529nm/A493nm of folda-dimer 2 is 1.29 in
CHCl3 which gradually decreases down to 0.80 upon increasing
the THF content from 0 to 100%. It is noteworthy that for the
reference PBI dye 15, which contains one PBI unit and thus
cannot fold, this ratio is close to 1.65 in both solvents (the Sup-
porting Information, Figure S3). In this regard it should be
mentioned that for a rigid cofacial stacked PBI dimer of Wasie-
lewski, a A0ꢀ0/A1ꢀ0 ratio of around 0.9 in CHCl3 was reported,[20]
whereas for a rigid biphenyl-bridged PBI dimer of Langhals
and co-workers,[21] in which both chromophores are arranged
To get more insight into the solvent-dependent conforma-
tions of folda-dimer 2 and folda-trimer 3, and to explore
whether additional dynamic processes are involved, we have
1
conducted temperature-dependent H NMR studies in [D8]THF
and CDCl3. It has to be pointed out here that, in contrast to
the research on biomolecules, application of NMR methods is
much rarer for the elucidation of conformational preferences
of synthetic oligomer backbones. However, in the field of
naphthalene and perylene bisimide dyes a few studies have
been reported, for example, by the groups of Li,[23]
Castellano,[24] Sanders,[25] and Wasielewski.[26]
In CDCl3, which is known as a good solvent for the solvation
of p-extended aromatic systems, in particular PBI dyes,[27] a sur-
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
4
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1
Figure 3. Aromatic regions of temperature-dependent H NMR spectra (600 MHz) of folda-dimer 2 (c=5ꢂ10ꢀ4 m) in [D8]THF at 274 to 333 K and in CDCl3 at
293 K (top). The temperature is indicated above the respective spectrum. Solvent signals are marked with asterisks (*). In the proton spectrum of 2 in [D8]THF
at 333 K (bottom) signals of the low-populated conformer (see the discussion below) are marked with plus symbols (+). Inset: Chemical structure of folded 2
with the assignment of aromatic protons of PBI core (P1, P2, P3, and P4) and phenylene ethynylene scaffold (2, 4, 5, 6, and 3’). Long alkyl chains are represent-
ed by methyl substituents for simplicity. Strong interactions between adjacent protons observed in the 2D ROESY NMR spectrum (see the Supporting
Information, Figure S10) are indicated by a grey arrow.
prisingly complex signal pattern was observed in the aromatic
region (d=8.8 to 7.2 ppm) for folda-dimer 2 at 293 K (Figure 3,
top). Such signal pattern suggests a coexistence of different
conformations of 2 in CDCl3, whose interconversion is not fast
enough on the NMR timescale for signal averaging. Whereas
the signals for PBI core protons (between d=8.8 and 8.2 ppm)
are not reliably assignable due to signal overlapping, the
major signals (between d=7.7 and 7.2 ppm) of phenylene
ethynylene scaffold could be assigned (Figure 3, top spectrum).
Variation of temperature in the range from 274 to 323 K did
not simplify the spectra of 2 in CDCl3 (the Supporting Informa-
tion, Figure S9). On the other hand, variable temperature (from
274 to 333 K) 1H NMR spectra of 2 in [D8]THF showed compara-
tively simpler signal pattern with perylene core protons locat-
ed in a narrow range (d=8.4 to 8.2 ppm). At lower tempera-
ture (274 to 298 K) some dynamic broadening of the signals of
perylene core protons at d=8.4 to 8.2 ppm was observed.
Upon increasing the temperature to 313 K and higher, the
broadened perylene proton signals got increasingly sharper
and nicely resolved in four doublets. Compared with the spec-
tra in CDCl3, the signals of the perylene protons of folda-dimer
2 in [D8]THF are significantly higher-field shifted, which is in
stark contrast to the monomeric PBI reference compound 15
since for the latter an opposite trend was observed in these
solvents (see the Supporting Information, Figure S11). The
observed highfield shift for the perylene protons of PBI 2 in
[D8]THF can be taken as an indication for the formation of
p–p-stacked PBI dimer, that is, folding of 2. Since the temp-
erature increase (from 274 to 333 K) resulted in a sharpening
of the aromatic proton signals without any considerable shift
or change of coupling pattern (Figure 3), an intramolecular
dynamic process that preserves the p–p stacking should be
responsible for the observed temperature-dependent changes
of the NMR spectra of 2 in [D8]THF.
To shed light on this dynamic process, the variable tempera-
ture NMR study of reference PBI 15 (Figure 4a), which is a sim-
pler monomeric analogue of 2, in [D8]THF is very instructive. A
1
selection of temperature-dependent H NMR spectra of 15 in
[D8]THF is shown in Figure 4d (for the complete set of spectra
measured at temperature range 274 to 333 K, see the Support-
ing Information, Figure S12). In the spectrum at lower tempera-
ture (274 K), four nicely separated doublets were observed at
d=8.94, 8.71, 8.68, and 8.64 ppm in a signal integral ratio of
4:2:1:1, respectively. These doublets could be assigned to the
perylene protons as shown in Figure 4a.
Upon increasing the temperature, the two doublets at d=
8.68 and 8.64 ppm (for protons P4/P4’) become gradually
broader and merge into a broad signal at around 303 K (see
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
5
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 4. a) Chemical structure of 15 with the assignment of aromatic protons; b) representative CPK models of the two rotational conformations (A and B)
arising from the hindered rotation of swallow tail around the CꢀN bond in PBI 15 schematically shown for N-(isopropyl)-1,8-naphthalene imide (view along
the CꢀN bond, R=n-C11H23). The two different perylene core protons P4 and P4’, distinguishable at low temperature, are labelled in both sketches;
c) schematic representation of the energy profile for the dynamic interconversion of the two rotamers; d) aromatic regions of the selected temperature-
1
dependent H NMR spectra (600 MHz) of 15 (c=5ꢂ10ꢀ4 m) in [D8]THF at 274 to 333 K (below). Temperatures are indicated above the respective spectrum.
Signals are assigned as depicted in (a).
also the Supporting Information, Figure S12). Upon further in-
creasing of temperature, this broad signal gets again narrower
and above 323 K a doublet arises, which becomes very sharp
at 333 K. With the concomitant separation of the signals for
protons P2 and P3 the expected four doublets with intensity
ratio of 2:2:2:2 are now realized. The observed temperature-
dependent spectral changes for 15 in [D8]THF can be attribut-
ed to a hindered rotation around the CꢀN bond with a
swallowtail (12-tricosanyl) imide substituent. Indeed, perylene
bisimides and similar imide compounds that exhibit restricted
rotation around CꢀN imide bond with considerably high
rotational barriers have been reported previously.[24,28] Thus,
such rotational isomers can be frozen at low temperature. For
PBI 15 we have observed two doublets at d=8.68 and
8.64 ppm for the two perylene ortho-protons P4/P4’ at low
temperature (274 K) in [D8]THF, whereas for the other two
ortho-protons (P1) only one doublet at d=8.71 ppm was ob-
served. This spectral feature can be attributed to a slower rota-
tion of the swallowtail substituent around the respective CꢀN
bond on the NMR timescale, and thus freezing of the two ro-
tamers A and B shown in Figure 4b. Upon increasing the tem-
perature, these doublets are gradually broadened, and at
around 303 K show coalescence, revealing a dynamic equilibri-
um of these rotamers on the NMR timescale. At temperatures
above 303 K, the rotation gets faster than the NMR timescale
and thus the two perylene protons P4/P4’ become chemically
equivalent, giving rise to a sharp doublet. An energy barrier
(DG°) of about 64 kJmolꢀ1 (see Figure 4c) could be estimated
for this intramolecular rotation process by using the coales-
cence method.[29] Similar values have been reported for the ro-
tational barrier of PBIs bearing branched alkyl substituents at
the imide positions.[28a] By employing a smaller reference
system, namely N-(isopropyl)-1,8-naphthalene imide, we have
estimated an energy barrier of about 61 kJmolꢀ1 by DFT calcu-
lations. For comparison, an energy barrier of only 25 kJmolꢀ1
was calculated for a related 1,8-naphthalene imide bearing
a meta-ethynyl phenylene (instead of isopropyl) imide substitu-
ent.[30] Accordingly, also DFT calculations support our
conclusion that the hindered rotation around the CꢀN
bond with a swallowtail imide substituent, and not that of
ethynyl phenylene substituent is responsible for the observed
1
temperature-dependent H NMR spectral changes in the 274–
333 K temperature regime in [D8]THF (also see the Section:
Theoretical calculations).
In light of the above discussion on variable-temperature
1H NMR spectroscopic study of monomeric reference PBI 15,
the temperature-dependent spectral changes observed for the
folda-dimer 2 in [D8]THF (Figure 3) can be ascribed to a similar
hindered rotation around the CꢀN bonds of two PBI units
bearing swallowtail substituents. The broad signal observed at
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
6
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
d=8.4 to 8.2 ppm for the perylene protons of PBI 2 at low
temperature (274 K) can be attributed to a slow rotation of the
swallowtail substituents around CꢀN bonds on the NMR time-
scale. At higher temperatures (>313 K) the rotation around Cꢀ
N bond gets much faster than the NMR timescale, and thus
the perylene protons (P1, P2, P3, and P4) of the two PBI units
are nicely resolved in four sharp doublets, which could be
clearly assigned (as shown in Figure 3) based on the 2D NMR
spectra. The observation of only four doublets at higher
temperatures (323 K) for 16 perylene protons of folda-dimer 2,
implies that the respective proton pairs (P1, P2, P3, and P4) in
both PBI units of 2 are chemically equivalent. This can be
taken as an indication for a tightly packed and symmetric
folding of PBI 2 in [D8]THF.
8.41 and 8.37 ppm) are gradually broadened and merge into
a broad signal at around 313 K, and at higher temperatures
(333 K) again become a doublet.
Likewise, with increasing temperature, the broad signals at
d=8.2 ppm first change into a broad singlet and become final-
ly resolved to a sharp doublet at higher temperature (333 K).
Again the hindered rotation around the CꢀN bonds bearing
branched alkyl substituents (swallowtail) may account for this
dynamic spectral behavior of folda-trimer 3 as discussed
before for the folda-dimer 2. As in the case of dimeric ana-
logue 2, for folda-trimer 3 the existence of a second, low-
populated conformer can be anticipated from the additional
signal set (labelled with symbol +) observed at 333 K (Fig-
ure 5a, bottom spectrum). A ratio of 1:4 between the two con-
formers was determined by signal integration in the higher
field (CꢀH proton of the alkyl substituent, see the expanded
spectrum in the Supporting Information, Figure S16). By apply-
ing Boltzmann analysis, an energy difference of approximately
4 kJmolꢀ1 could be estimated for this distribution. This value is
slightly smaller than that determined for 2. The reason for this
might be that in 3 the two PBI bridging scaffold subunits are
influencing and preorganizing each other.
1
A closer look at the high temperature H NMR spectrum of 2
in [D8]THF at 333 K reveals the appearance of an additional set
of weaker signals (Figure 3, bottom). These signals presumably
arise from a second folded conformation of 2 with quite differ-
ent 1H NMR chemical shifts, which are difficult to assign reliably
as they are low in intensity and, moreover, overlap partially
with the prominent signals. From signal integration of
a-proton resonances of imide substituents at d=5.25 to
5.10 ppm (for an enlarged spectrum see the Supporting Infor-
mation, Figure S13) a 1:10 ratio of the conformers can be esti-
mated. For this population distribution, an energy difference of
approx. 6 kJmolꢀ1 between the two conformers can be esti-
mated by applying Boltzmann analysis. Backbone flapping in
folded PBI 2 may indeed result in two different conformations
as revealed below by theoretical calculations (local optimiza-
tion of one gradually distorted phenylene unit in folded 2 with
PM6-DH2), which provide an energy barrier of approximately
50 kJmolꢀ1 for the backbone flapping, which might be
significantly underestimated by a factor of 2 (for details, see
the Section: Theoretical calculations).
Moreover, the 2D-NMR spectroscopic studies were per-
formed to gain more insight into the structural features of
folded conformations of folda-dimer 2 and trimer 3. Since the
neighboring PBI units are arranged close to each other in the
folded state, prominent cross-couplings would provide infor-
mation on spatial arrangement of PBI dye units in conformers.
However, due to the potentially high C2 symmetry of the
folded structures, the nuclear Overhauser effects (NOE) visual-
ized by 2D-ROESY NMR spectroscopy can only partially be
used to reveal the PBI interactions upon folding. The most ex-
pressive through-space proton–proton cross-couplings in the
folded states are given between the two H2 protons in 2 (see
the inserted structure in Figure 3) and the protons at the H2
and H2’’ positions in 3 of the phenylene imide substituents in
[D8]THF (see the inserted structure in Figure 5a). These cross-
couplings cannot be identified because of the overlapping
with the diagonal signals in the ROESY spectra. In particular,
couplings between perylene core protons P1 and P2 and H6 of
the phenylene ethynylene scaffold can be observed for 2 in
[D8]THF (see the Supporting Information, Figure S10), which
supports the proposed folded structure for dimer 2 (the evalu-
ated distances between these protons, P1ꢀH6 and P2ꢀH6, re-
spectively, by our calculations are 2.4 and 3.3 ꢃ). On the other
hand, the dominant signals of 2 in CDCl3 do not show any
cross-couplings (the Supporting Information, Figure S17). This
corroborates our suggestion that unfolded states of 2 prevail
in chlorinated solvent. As mentioned, PBI 2 adopts different
conformations in CDCl3, whereas few of them bring both chro-
mophores in sufficient contact to each other to communicate.
Although not completely folded, this PBI–PBI distance
narrowing results in appreciable ROESY cross-couplings (see
the Supporting Information, Figure S17). Likewise, the 2D
ROESY spectrum of folda-trimer 3 in CDCl3 (the Supporting
Information, Figure S18) does not exhibit any significant cross
coupling as the PBI dye units are apparently not arranged in
Temperature-dependent 1H NMR spectroscopic studies of
folda-trimer 3 in [D8]THF and CDCl3 were performed as well
(Figure 5a and the Supporting Information, Figure S15). In ad-
dition, the 2D-ROESY NMR spectra were measured (Figure 5b
and the Supporting Information, Figure S14). In CDCl3 a broad
signal pattern spreading over the complete aromatic range
from d=8.9 to 6.9 ppm is observed, which is, as expected,
more complex than that observed for 2 in this solvent. Al-
though the assignment of signals to the protons of 3 appears
rather elusive, the large number of peaks provides an indica-
tion for the co-existence of different conformations with vary-
ing spatial arrangement of the PBI subunits. The variation of
temperature from 274 to 333 K did not appreciably affect the
spectral shape in CDCl3 (the Supporting Information, Fig-
1
ure S15). However, the temperature-dependent H NMR spectra
of 3 in [D8]THF exhibit, as in the case of 2, a simpler signal pat-
tern (Figure 5a). Thus, the signals of all aromatic protons be-
tween d=8.4 and 7.2 ppm become increasingly narrower with
increasing temperatures from 274 to 333 K. Most significant
temperature-dependent changes are observed for the two
doublets at d=8.41 and 8.37 ppm, and for the broad signals
at d=8.2 ppm that appear in the spectrum at 274 K. Upon in-
creasing the temperature, the signals of the two doublets (d=
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
7
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
1
Figure 5. a) Temperature-dependent H NMR spectra of folda-trimer 3 in [D8]THF at 274 to 333 K and in CDCl3 at 293 K (top) and b) 1H,1H-2D-ROESY NMR
spectrum of 3 in [D8]THF at 328 K (for (a) and (b): c=4ꢂ10ꢀ4 m, 600 MHz). In (a) the solvent signals are marked with asterisks (*). In the proton spectrum in
[D8]THF at 333 K (bottom) smaller signals are marked with plus symbols (+). Inset: The chemical structure of folded 3 with the assignment of aromatic phen-
ylene protons. Alkyl chains are represented by methyl substituents for simplicity. Perylene protons of outer PBI units (PBIo, bold black and orange in the
sketch) are marked with Po, and those of the inner PBI (PBIi, bold red in the sketch) with Pi. The blue arrow indicates a strong interaction between adjacent
protons in the ROESY spectrum. The changes in chemical shifts of the signals in the region of d=7.5 to 7.3 ppm upon increasing temperature are illustrated
1
by dashed lines. In (b), signals which are marked with circles correspond to cross-couplings that are not present in the H,1H-2D COSY NMR spectrum (see the
Supporting Information, Figure S19).
spatial proximity. However, in the [D8]THF spectra of 3 promi-
nent through-space interactions are present (Figure 5b, for an
enlarged spectrum see the Supporting Information, Fig-
ure S14). These include signals that arise from couplings be-
tween protons of the two outer PBI moieties (Po) with those of
the inner PBI (Pi) as well as those between the outer PBI moiet-
ies (Po) and the backbone H2’’ protons. These through-space
couplings can take place only in the folded state (depicted by
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
8
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
the arrow in the structure shown in Figure 5a). A comparison
of the 2D-COSY NMR spectrum of 3 in [D8]THF (the Supporting
Information, Figure S19) with the 2D-ROESY spectrum (Fig-
ure 5b) reveals such through-space couplings. Additionally, the
number of carbon atoms and CꢀH units comply with the
signals present in the 13C NMR spectra (besides overlapping of
signals with similar shifts) as well as in the 2D-HSQC NMR
spectra both for PBI 2 and PBI 3 in [D8]THF (see the Supporting
Information, Pages S51–S54).
ranged p-stacks. To obtain more detailed information on the
energetically most favorable structures for the PBI systems 2
and 3 and thus to corroborate the results of NMR studies, we
performed calculations by applying various methods. We first
used the force fields OPLS-AA, AMBER, and MM3* to determine
the equilibrium structures of 2 and 3. The resulting structures,
which are summarized in Figure 6, indicate strong variations
depending on the force fields. The force fields MM3* and
AMBER predict folded structures, whereas the OPLS-AA force
field computed open structures as the most favorable
conformers.
DOSY NMR spectroscopic studies were performed to assess
the molecular dimensions of the foldamers. For folda-dimer 2
in [D8]THF at 298 K, a diffusion coefficient (D) of 6.3ꢂ
10ꢀ10 m2 sꢀ1 was estimated, which corresponds to a diameter
of 22.6 ꢃ according to the Stokes–Einstein Equation[31] by as-
suming a spherical molecule (see also the Supporting Informa-
tion, Figure S20). For the folded state of dimer 2, the geometry
optimized structure revealed a distance of 22.4 ꢃ between the
first carbon atom of the imide substituent and the alkyl chain
on the middle phenylene spacer unit, which is in good accord-
ance with the experimentally obtained value of 22.6 ꢃ. Notably,
the other set arising from the weaker NMR signals shows
almost the same diffusion behavior, which is further in accord-
ance with a second folded conformer of PBI 2. For folda-trimer
3, a diffusion coefficient of 3.3ꢂ10ꢀ10 m2 sꢀ1 was determined
by DOSY NMR spectroscopy in [D8]THF at 293 K (see the Sup-
porting Information, Figure S21). Based on this data, a diameter
of 27.8 ꢃ was calculated for the folded trimer 3, which is, as
expected, slightly higher than the value determined for 2 in
[D8]THF. As in the case of PBI 2, similar diffusion constants are
obtained for the NMR signal set with low intensity for PBI 3.
Therefore, both signal sets of 3 appear to correspond to the
same state, that is, the folded state.
In the next step, the structural moieties responsible for flexi-
bility in dimer 2 and trimer 3 were identified and analyzed in
terms of their impact on the overall energy. The most flexible
linkages, which are capable of driving the molecules to folded
or unfolded states, are the single and triple bonds between
the inner imide moieties and the phenylene imide substituents
as well as in the tolan subunits as schematically depicted for
the reference PBI 15 (the Supporting Information, Figure S23a).
To test the accuracy of various approaches, we have compared
the relative energies of gradual distortion (0 to 1808) of
dihedrals for the related, simple model compounds tolan and
N-phenyl-1,8-naphthalimide determined by different force
fields, semiempirical and DFT approaches (the Supporting
Information, Figure S23b–d). On the basis of these theoretical
studies, we decided to use the PM6-DH2 approach to perform
molecular dynamic (MD) simulations for 2 and 3 to elucidate
the flexibility of the systems.
PM6-DH2 includes dispersion forces and should be suffi-
ciently flexible for the description of unusual conformations,
which might be energetically overestimated by force field
approaches. In addition, the PM6-DH2 approach is an ideal
compromise between accuracy and efforts. As starting struc-
tures for the MD simulations we used the minimum structures
predicted by the force fields (Figure 6).
Theoretical studies on the conformational properties
Our NMR spectroscopic studies revealed that in CHCl3 the PBI
units of folda-dimer 2 and folda-trimer 3 adopt a larger confor-
mational space due to good solvation of the aromatic scaf-
folds, whereas in THF the conformational diversity is apprecia-
bly reduced due to intramolecular folding into helically ar-
The energy distributions given in Figure 7a and b reflect the
relative energies of the minima, which are obtained when the
MD frames of the last 50 ps of the MD simulations are opti-
mized. The abbreviation i–iii denotes that the MD run started
from the structures I–III (see Figure 6), respectively. The
Figure 6. Geometry optimized structures (side view) of a) folda-dimer 2 and b) folda-trimer 3 obtained by different force field optimizations (I: MM3*; II:
AMBER; III: OPLS-AA). The respective top views of these conformations are shown in the Supporting Information, Figure S22.
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
9
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 7. Relative energies in kJmolꢀ1 of a) folda-dimer 2 and b) folda-trimer 3 obtained from molecular dynamic simulations (PM6-DH2, 300 K, 50 ps) and
subsequent local optimization. Lowest-energy structures (side view) of c) PBI dimer 2-i to 2-iii and d) PBI trimer 3-i to 3-iii obtained after MD simulations. The
respective top views of these conformations are shown in Figure S24 (the Supporting Information).
energies summarized in Figure 7a and b are relative to the
lowest-energy structures of folda-dimer 2 and folda-trimer 3,
respectively. All MD runs were performed for the gas phase.
The MD simulations revealed the following important
results: First, both unfolded conformations III of PBI 2 and 3
already form completely folded structures after a time period
of 8 ps (PBI dimer 2) and 25 ps (PBI trimer 3; see conforma-
tions iii in Figure 7c,d), implying that the intramolecular folded
arrangements are much more favored than the unfolded con-
formations. For dimer 2 and trimer 3, these folding processes
from III to iii can nicely be visualized by the distance narrow-
ing from 26 and 23 ꢃ (in unfolded conformations III) to around
3.7 and 3.6 ꢃ (in folded conformations iii) of adjacent PBI cen-
ters in 2 and averaged for 3 as shown in Figure S25 (the Sup-
porting Information). Although gas phase calculations cannot
take into account explicit solute–solvent interactions, that is,
the difference between CHCl3 and THF for the folding process,
the outcome of our MD indicates a very flexible bridging unit
of PBI 2 and 3. For the DFT-D geometry optimized structure of
the unfolded conformation 2-III (the Supporting Information,
Figure S26), an energy difference to 2-I of 165 kJmolꢀ1 was
computed on the B97D/STO-3G level of sophistication. Notably,
this value is very similar to the binding energy of two PBI mon-
omers forming the energetically most favorable p–p-stacked
dimer (125 kJmolꢀ1, BLYP-D/TZV(P)).[32] The higher energy dif-
ference obtained in the present computations can be reduced
to basis set superposition errors of the smaller STO-3G basis.
The small variation in the calculated values strongly underlines
that the phenylene ethynylene backbone is sufficiently flexible
that both PBI units are not hindered in the formation of favora-
ble p-stacked structures. In the case of PBI dimer 2-III, the fold-
ing leads to the conformation 2-iii. The latter one is very simi-
lar to the conformation 2-i obtained from 2-I, that is, their
energy difference is 0 kJmolꢀ1, with 2-i being almost identical
with the MM3* geometry-optimized structure 2-I (DE (2-Iꢀ2-
i)=4 kJmolꢀ1). Notably, the relative energies of all structures
adopted during MD simulations (after local optimizations with
PM6-DH2 of the last 50 ps structures) of 2-i and 2-iii are in
narrow distribution (0–8 kJmolꢀ1) as the structures are nearly
congruent. In contrast, the MD simulation starting from 2-II
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
10
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
leads to structure 2-ii. It differs from 2-i and 2-iii by the orien-
tation of the bridge relative to the PBI planes. The average en-
ergies of the minima obtained by optimizing the correspond-
ing MD frames are shifted to somewhat higher values (Fig-
ure 7a, 8 kJmolꢀ1). This indicates that the minima 2-II and 2-ii
are separated from the other minima (2-I, 2-I, and 2-iii), by bar-
riers that are not crossed during the MD simulations.
an energy difference between 2-I and 2-II (see the Supporting
Information, Figure S25) of about 8 kJmolꢀ1 is obtained. Ac-
cordingly, the calculated values are in reasonable agreement
with the NMR data (ca., 6 kJmolꢀ1).
For folda-dimer 2, a subsequent distortion of one phenylene
spacer in conformation II was used to estimate the energy
barrier of the flapping process. For this approach, gradually
distorted structures were locally optimized using PM6-DH2.
Small distortions did not significantly change the relative
energies of the system since the conformation II, especially the
mutual orientation of both PBI moieties, did not show a large
impact. The barrier of the flapping process was calculated to
be about 50 kJmolꢀ1 (Figure 8b). This value might be signifi-
cantly underestimated based on comparison of experimental
and calculated energy barriers for the rotation of the swallow-
tail imide substituent. For the latter one, the energy barrier
was computed to be about 33 kJmolꢀ1 using local optimiza-
tion on PM6-DH2 level, which is only about half of the experi-
mental value of 64 kJmolꢀ1 based on temperature-dependent
NMR analysis (Figure 4). Accordingly, the barrier for the flap-
ping process is expected to be about 100 kJmolꢀ1, which ex-
plains why no coalescence of NMR signals of the major and
the minor conformations of backbone-flapped folded species
of PBI 2 could be observed in our accessible temperature
range (Figure 3). The same situation holds true for PBI 3 in
which the two conformations 3-i and 3-iii (Figure 7d) exhibit
similar energies and the NMR experiment (Figure 5a) also
revealed the coexistence of two conformations that show no
coalescence in the accessible temperature range in [D8]THF.
As expected, computations in which the solvent is approxi-
mated by a continuum cannot explain the differences in the
folding behavior of our systems in THF and CHCl3. Because ac-
curate simulations describing the complete system in a large
solvent box are very costly and this topic is not the central
focus of this work, we only roughly estimated enthalpy effects
for folda-dimer 2 (see discussion in the Supporting Informa-
tion, Pages S34–S38, Figures S27–S30, and Tables S1 and S2).
These estimates indicate that the differences in the folding
behavior do not stem from the strengths of the interaction
between individual solvent molecules with a PBI moiety. The
number of solvent molecules that fit within the first
solvent shell seems to be more important. Notably, entropy
effects always support the folding process because solvent
molecules are set free during the folding processes.
For the folda-trimer 3, the MD simulations gave similar re-
sults as for the folda-dimer 2. Also for the conformation 3-i the
relative energies are within 0 and 4 kJmolꢀ1, which suggests
that the ideal conformation has already been adopted. Al-
though conformations 3-i and 3-ii possess virtually the same
spatial arrangement after approximately 8 ps MD simulation
and local optimization, the relative energies of structures from
conformation 3-ii show a wider distribution spreading from 0
to 24 kJmolꢀ1. However, the main portion (ꢁ80%) of the
structures is again within 0 and 4 kJmolꢀ1 (see Figure 7b). For
the conformation 3-iii, which is obtained from the MD
simulation starting with the unfolded conformation 3-III, the
primarily populated structures are higher in energy (energy dif-
ference about 4 kJmolꢀ1) and the lowest energy conformation
3-iii resembles more the non-optimal structure of 3-II.
These results explain the appearance of a second low-popu-
1
lated signal set observed in H NMR spectra of folda-dimer 2
and folda-trimer 3 in [D8]THF, which is most expressive at
higher temperatures (see Figure 3 and 5, bottom spectra). Be-
cause 2-i and 2-ii are nearly identical to 2-I and 2-II, further
considerations focus on these two structures. The schematic
energy profile as well as the suggested dynamic process
caused by backbone flapping between conformation 2-I and
2-II are shown in Figure 8. Notably, the conformations 2-I and
2-I’ with the backbone directed upwards and downwards,
respectively, are chemically and energetically equivalent and
thus are not distinguishable by NMR spectroscopy.
The energy difference between 2-I and 2-II is computed to
approximately 4 kJmolꢀ1 (by PM6-DH2), which enables a de-
tectable population of II by NMR spectroscopy. By using geom-
etry optimizations on the B97D/STO-3G level of sophistication
Folding properties in the excited state
Optical excitation of folda-dimer 2 and folda-trimer 3 afford ex-
cited molecules whose initial conformational versatility equals
the one given in the ground state (Franck–Condon principle).
Subsequently, however, the molecules may relax into ener-
getically more favorable excited-state conformations such as
excimers[12,33] (in which two PBI units exhibit a more coplanar
arrangement) or interact with the electron-donating phenylene
ethynylene backbone.[34] To elucidate these processes we
performed steady state and time-resolved fluorescence and
transient absorption spectroscopy experiments.
Figure 8. a) Schematic representation of the dynamic interconversion be-
tween conformations I, II, and I’ of folded PBI 2 caused by flapping of phen-
ylene ethynylene scaffold; b) schematic illustration of the energy profile for
the dynamic interconversion of the flapping with relative energies calculated
by local optimization (PM6-DH2) of gradually distorted phenylene spacer
unit in the scaffold.
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
11
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
The fluorescence emission and excitation spectra as well as
the absorption spectra of folda-dimer 2 and folda-trimer 3 in
CHCl3 and THF are shown in Figure 9. The emission spectrum
of folda-dimer 2 in CHCl3 consists of two sharp bands located
at 537 and 580 nm, which are hypsochromically shifted to 533
and 575 nm in THF (Figure 9a and b). Moreover, a strong third
band was observed at 615 nm in THF, whereas a small, barely
recognizable shoulder appears in CHCl3 at 630 nm. For each
solution, fluorescence excitation spectra were recorded at the
detection wavelengths (ldet) of 530 and 650 nm. In both sol-
vents, the excitation spectra of folda-dimer 2 for lex =530 nm
perfectly resemble the absorption spectrum of 2 in CHCl3,
whereas the ones for ldet =650 nm match with the absorption
spectrum of folda-dimer 2 in THF. This result suggests that the
observed fluorescence can be related to the ground state con-
formational populations of unfolded and folded molecules,
that is, similar conformations prevail for excited molecules.
The fluorescence spectra of folda-trimer 3 in CHCl3 and THF
(Figure 9c and d) show stronger changes when compared with
those of folda-dimer 2 or reference compounds 9 and 15 (the
Supporting Information, Figure S3). The emission spectrum of
PBI 3 in CHCl3 is composed of a broad band with a maximum
at 600 nm and a smaller signal at 540 nm, in which the former
band could be ascribed to the emission of excited folded mol-
ecules. Upon solvent change from CHCl3 to THF, both bands
are hypsochromically shifted (532 and 617 nm, respectively)
and the intensity of the smaller band significantly decreased,
which indicates the prevalence of folded molecules. The fluo-
rescence excitation spectra of 3 for ldet =530 nm in CHCl3 and
THF again resemble the absorption spectra of unfolded mole-
cules in CHCl3, whereas those at ldet =650 nm resemble the ab-
sorption spectrum of 3 in THF. Moreover, the fluorescence
quantum yields (Ffl)[35] were determined and are collected to-
gether with other significant optical properties of folda-dimer
2, folda-trimer 3, and reference compounds 9 and 15 in
Table 1.
As shown in Table 1, the fluorescence quantum yields of
monomeric PBI dye 9 are close to unity (around 97%), which is
in accordance with the generally observed outstanding fluores-
cence of monomeric PBI dyes. For the reference compound 15,
which bears a phenylene ethynylene substituent, the quantum
yields are significantly reduced to 21% in CHCl3 and 36% in
THF, indicating an intramolecular electron-transfer proc-
ess[26,34b,36] between the electron-rich phenylene ethynylene
backbone and the electron-deficient PBI chromophore, which
becomes more efficient in CHCl3 than in THF (see below). For
folda-dimer 2 and folda-trimer 3, a contrasting behavior is ob-
served, that is, the Ffl values are always lower in THF com-
pared with those in CHCl3, which suggests the presence of an-
other non-radiative deactivation pathway in THF like folding
into less emissive excimers.[37] In THF this process is obviously
more prominent because the individual dye molecules are al-
ready in a p-stacked folded conformation in the ground state.
Likewise, the lower fluorescence quantum yields and more
pronounced excimer emission of 3 compared to 2 suggest
that only a very minor portion of unfolded molecules are
present in the excited state.
Figure 9. Absorption (solid line, c=10ꢀ6 m), fluorescence emission
(lexc =480 nm; d) and fluorescence excitation spectra (ldet =530 nm: b;
ldet =650 nm: g) of 2 in a) CHCl3 and b) THF, and 3 in c) CHCl3 and d) THF
at room temperature. The fluorescence excitation spectra are normalized to
the absorption spectra at approximately 490 nm.
The time-resolved fluorescence decay profiles of folda-dimer
2 and folda-trimer 3 along with those of 9 and 15 as reference
compounds were measured by using the time-correlated
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
12
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
713, 800, and 960 nm, we fo-
cused our attention on the de-
tection of this spectral feature in
the transient excited-state ab-
sorption (ESA) signals. The TA
spectra of all compounds
probed at visible region showed
negative ground-state bleaching
(GSB) signals at 450–530 nm and
stimulated emission (SE) up to
Table 1. Optical properties of folda-dimer 2 and folda-trimer 3 as well as those of the reference compounds
PBI 9 and PBI 15 in different solvents.
PBI
Solvent
labs [nm]
(emax [mꢀ1 cmꢀ1])[a]
lem
[nm][b,c]
Ffl
[%][b,d]
CHCl3
THF
CHCl3
THF
CHCl3
THF
CHCl3
THF
529 (123600), 493 (96100), 462 (37300)
526 (73800), 493 (91400)
531 (118300), 495 (124100)
537, 580
533, 575, 615
540, 600
28ꢂ2 (31)
16ꢂ2 (16)
22ꢂ2 (24)
12ꢂ1 (13)
97ꢂ3 (89)
97ꢂ6 (94)
21ꢂ1 (17)
36ꢂ5 (38)
2
3
534 (66700), 495 (109300)
532, 617
528 (83100), 491 (50000), 460 (18100)
522 (81900), 486 (49700), 456 (18100)
528 (84200), 491 (50800), 460 (18800)
523 (80400), 487 (49200), 457 (18200)
532, 574, 622
531, 571, 618
537, 579, 627
531, 571, 617
9
ꢁ600 nm
(for
THF,
see
15
Figure 10; for CHCl3, see the
Supporting Information, Fig-
ure S32). Strong positive signals
which come from the excited-
state absorption from S1 state to
[a] Measurements at 258C, c=10ꢀ6 m; [b] using high dilution method (ODmax <0.05) at RT; [c] lexc =480 nm;
[d] under magic angle conditions (q=54.78), lexc =470–490 nm; values in brackets are the results obtained
from measurements without magic angle conditions.
single photon counting (TCSPC) technique and their fitted
fluorescence lifetimes are collected in the Supporting Informa-
tion, Figure S31 and Table S3. In the case of monomeric PBI
dye 15, the observed fluorescence lifetimes (2.7 ns in CHCl3
and 2.3 ns in THF) were relatively short in both solvents as
compared with those of 9 (3.9 ns in CHCl3 and 4.0 ns in THF)
and other PBI monomers.[38] In addition to the long decay com-
ponents, shorter components with the time constants of 80 ps
(81%) in CHCl3 and 160 ps (36%) in THF were observed for PBI
15, which was absent in the backbone-free monomer 9. This
can be attributed to the existence of a non-radiative pathway
originating from the photoinduced electron-transfer (PET)
process from the electron-donating backbone to the electron-
deficient PBI core. Here, the much higher amplitude value
(81%) of the short time constant corroborates our previous
interpretation of the trends in fluorescence quantum yields
that this process is more efficient in CHCl3, that is, in the
extended form.
higher (Sn) states were revealed in the 600–850 nm spectral
range.[34a,39] Interestingly, the characteristic sharp structures
with peaks at ꢁ705 nm, attributed to one of the absorption
bands of a PBI radical anion,[34a,39] turn progressively broader
and lose their structures as the molecules go from the extend-
ed forms to the compact folded forms in THF, suggesting a de-
creased efficiency of the CS process from the monomer to the
trimer (Figure 10a–c). In CHCl3 this trend is also visible but is
not as pronounced (see the Supporting Information, Fig-
ure S32).
Further analysis of the PET dynamics was performed by as-
sessing the TA spectra at near-infrared (NIR) wavelengths of
860–1100 nm, which are partly shown in Figure 10d–f (THF; for
comparison with the spectra in CHCl3, see Figure S33 in the
Supporting Information). The spectra of 15 and 2 show charac-
teristic features in which the broad (ꢁ850–1100 nm) ESA of
the neutral PBI decays in time and a sharper band that peaks
at ꢁ960 nm emerges at the same time, which is another evi-
Cꢀ
For folda-dimer 2 and folda-trimer 3, fluorescent species
with about 10-fold longer lifetimes (20–26 ns) grew dominant,
which can be assigned as the enhanced lifetimes of molecules
in relaxed excited states, that is, excimer states.[37] In particular,
for the folded structure of 3 in THF, the relative amplitude of
the long-lived species was found to be as high as 77% when
the excimer emission band was directly probed. Meanwhile,
when the 0–0 fluorescence peaks (530–540 nm) were moni-
tored in 2 and 3, short components with the time constants of
approximately 100 ps were observed in all cases in addition to
the singlet excited-state lifetimes of the PBI units (2.0–3.1 ns).
The shorter components here can be related to the fast
fluorescence quenching by the PET process as discussed
above.
dence of the photoinduced formation of PBI by the PET pro-
cess. Interestingly, whereas this prominent signature for the
PBI radical anion of the monomer 15 is already decreased for
folda-dimer 2, it is almost entirely vanished for folda-trimer 3.
A similar tendency is given in CHCl3, in which the radical anion
signatures are even more prominent than in THF. From these
results, we can conclude that the efficiency of the PET process
is closely related to the spatial configuration of the
chromophores within the systems. In other words, the more
the molecules are folded, the less efficient the PET process is.
To further substantiate our findings, we have analyzed the
time constants (see Table 2) extracted from the kinetic traces,
which gave rise to double exponential decay profiles for PBI 2
and PBI 3 in both solvents. Here, the two time components
are ascribable to CS and CR times, respectively. Since it was im-
To acquire an in-depth understanding of the fast dynamics
responsible for the reduced fluorescence of the PBI foldamers,
that is, photoinduced charge separation (CS) and charge
recombination (CR), we performed femtosecond transient
absorption (TA) spectroscopy for 2, 3, and 15 in CHCl3 and THF.
As it is already known that formation of PBI radical anions
Cꢀ
possible to directly probe the rising PBI band, which is largely
immersed in the strong ESA signal, we instead monitored the
decay rate of the ESA of neutral PBI (ꢁ920 nm), which was
gradually converted into the PBI radical anion absorption band
(ꢁ960 nm) with time (see the Supporting Information, Fig-
ure S34). In dimer 2, the CS time was about three times faster
in the extended form in CHCl3 (1.5 ps) than that of its folded
counterpart in THF (4.3 ps). A similar trend was observed in
Cꢀ
(PBI ) through chemical reduction is easily recognized by the
decrease of PBI absorption in the range of 400–550 nm and
the concomitant increase of new absorption bands[34a,39] at
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
13
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 10. Femtosecond transient absorption spectra of reference compound 15 (a, d), folda-dimer 2 (b, e), and folda-trimer 3 (c, f) in THF probed at the
visible region (left column) and at the near-infrared region (right column; lex =520 nm; c=1ꢂ10ꢀ5 m).
Conclusion
Table 2. Charge separation (CS) and recombination (CR) time constants
of folda-dimer 2 and folda-trimer 3 from TA measurements in CHCl3 and
THF.
By attaching PBI chromophores to a defined phenylene
ethynylene scaffold, we were able to build up new oligomeric
PBI foldamer systems folda-dimer 2 and folda-trimer 3. As an
outstanding characteristic, these ensembles are capable of
adopting different conformations triggered by the solvents.
With these solvent-dependent conformational preferences, the
PBI chromophore units are either brought in close p–p-contact
to each other or appear in unfolded conformations, in which
the dye moieties are arranged randomly. The folding/unfolding
equilibria governed by solvation could be monitored by UV/Vis
absorption and 1D and 2D NMR spectroscopy (ROESY, DOSY)
and rationalized by theoretical calculations. Because the same
conformations, that is, folded and unfolded also prevail in the
excited state, we were able to elucidate the different excited-
PBI
CHCl3
THF
tCS [ps][a]
tCR [ps][a]
tCS [ps][a]
tCR [ps][a]
2
3
1.5
3.9
50
57
4.3
4.5
60
96
[a] Probe wavelength of l=920 nm was applied.
folda-trimer 3 despite much less clear difference in the CS
rates (3.9 and 4.5 ps in CHCl3 and THF, respectively). This
complies with our previous findings that 3 already resides in
a more folded structure even in CHCl3.
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
14
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
D. A. Klein, D. R. Powell, X. Huang, J. J. Barchi, S. H. Gellman, Nature
1997, 387, 381–384.
[6] R. S. Lokey, B. L. Iverson, Nature 1995, 375, 303–305.
[7] J. C. Nelson, J. G. Saven, J. S. Moore, P. G. Wolynes, Science 1997, 277,
1793–1796.
state dynamics of folded and unfolded PBI oligomers. For the
extended structures, the intrinsic PET process from the
electron-donating phenylene ethynylene backbone to the elec-
tron-deficient PBI units is very prominent. As long as the mole-
cules are in the folded state, this pathway becomes replaced
by the preferential relaxation of the PBI stack into a fluorescent
excimer state. Accordingly, in this latter state, in which the dis-
tance between dyes is short enough, the charge separation is
highly suppressed. This interesting switching of excited state
processes by conformational control through external stimuli[40]
will be investigated in more depth in the near future.
[8] V. Berl, I. Huc, R. G. Khoury, M. J. Krische, J.-M. Lehn, Nature 2000, 407,
720–723.
[9] a) Y. Ferrand, A. M. Kendhale, J. Garric, B. Kauffmann, I. Huc, Angew.
Chem. 2010, 122, 1822–1825; Angew. Chem. Int. Ed. 2010, 49, 1778–
1781; b) N. Delsuc, S. Massip, J.-M. Lꢄger, B. Kauffmann, I. Huc, J. Am.
Chem. Soc. 2011, 133, 3165–3172; c) Q. Gan, Y. Ferrand, C. Bao, B. Kauff-
mann, A. Grꢄlard, H. Jiang, I. Huc, Science 2011, 331, 1172–1175; d) M.
Kudo, V. Maurizot, B. Kauffmann, A. Tanatani, I. Huc, J. Am. Chem. Soc.
2013, 135, 9628–9631.
[10] a) F. Wꢀrthner, Chem. Commun. 2004, 1564–1579; b) M. R. Wasielewski,
J. Org. Chem. 2006, 71, 5051–5066; c) A. Ajayaghosh, V. K. Praveen, C.
Vijayakumar, Chem. Soc. Rev. 2008, 37, 109–122; d) T. F. A. De Greef,
M. M. J. Smulders, M. Wolffs, A. P. H. J. Schenning, R. P. Sijbesma, E. W.
Meijer, Chem. Rev. 2009, 109, 5687–5754; e) D. Gçrl, X. Zhang, F. Wꢀrth-
ner, Angew. Chem. 2012, 124, 6434–6455; Angew. Chem. Int. Ed. 2012,
51, 6328–6348; f) T. Seki, X. Lin, S. Yagai, Asian J. Org. Chem. 2013, 2,
708–724.
Experimental Section
See the Supporting Information for the experimental details.
Acknowledgements
[11] F. Wꢀrthner, Z. Chen, F. J. M. Hoeben, P. Osswald, C.-C. You, P. Jonkheijm,
J. v. Herrikhuyzen, A. P. H. J. Schenning, P. P. A. M. van der Schoot, E. W.
Meijer, E. H. A. Beckers, S. C. J. Meskers, R. A. J. Janssen, J. Am. Chem.
Soc. 2004, 126, 10611–10618.
[12] Z. Chen, V. Stepanenko, V. Dehm, P. Prins, L. D. A. Siebbeles, J. Seibt, P.
Marquetand, V. Engel, F. Wꢀrthner, Chem. Eur. J. 2007, 13, 436–449.
[13] H. Lin, R. Camacho, Y. Tian, T. E. Kaiser, F. Wꢀrthner, I. G. Scheblykin,
Nano Lett. 2010, 10, 620–626.
[14] a) V. Dehm, M. Bꢀchner, J. Seibt, V. Engel, F. Wꢀrthner, Chem. Sci. 2011,
2, 2094–2100; b) F. Schlosser, M. Moos, C. Lambert, F. Wꢀrthner, Adv.
Mater. 2013, 25, 410–414.
[15] Q. Zhou, P. J. Carroll, T. M. Swager, J. Org. Chem. 1994, 59, 1294–1301.
[16] D. Sahoo, S. Thiele, M. Schulte, N. Ramezanian, A. Godt, Beilstein J. Org.
Chem. 2010, 6, 57.
We are grateful to the Deutsche Forschungsgemeinschaft
(DFG) for financial support for this work within the research
graduate school GRK 1221 “Control of Electronic Properties of
Aggregated p-conjugated Molecules” and the Research Group
FOR 1809 “Light-Induced Dynamics in Molecular Aggregates”
at the University of Wꢀrzburg. We are indebted to Elfriede
Ruckdeschel for the NMR measurements. We would also like to
extend our thanks to Daniel Bellinger from the computational
chemistry group of Bernd Engels for his support. The work at
Yonsei was financially supported by the Mid-career Researcher
Program (2005-0093839) administered through the National
Research Foundation of Korea (NRF) funded by the Ministry of
Education, Science and Technology (MEST).
[17] a) S. Demmig, H. Langhals, Chem. Ber. 1988, 121, 225–230; b) H. Kaiser,
J. Lindner, H. Langhals, Chem. Ber. 1991, 124, 529–535.
[18] D. F. Louw, J. Strating, H. J. Backer, Recl. Trav. Chim. Pay. B 1954, 73,
649–654.
[19] M. J. Tauber, R. F. Kelley, J. M. Giaimo, B. Rybtchinski, M. R. Wasielewski,
J. Am. Chem. Soc. 2006, 128, 1782–1783.
[20] J. M. Giaimo, J. V. Lockard, L. E. Sinks, A. M. Scott, T. M. Wilson, M. R. Wa-
sielewski, J. Phys. Chem. A 2008, 112, 2322–2330.
Keywords: dyes/pigments · foldamers · molecular dynamics ·
NMR spectroscopy · time-resolved spectroscopy
[21] H. Langhals, A. Hofer, S. Bernhard, J. S. Siegel, P. Mayer, J. Org. Chem.
2011, 76, 990–992.
[22] J. Wang, A. Kulago, W. R. Browne, B. L. Feringa, J. Am. Chem. Soc. 2010,
132, 4191–4196.
[1] a) U. H. F. Bunz, Chem. Rev. 2000, 100, 1605–1644; b) M. D. Watson, A.
Fechtenkçtter, K. Mꢀllen, Chem. Rev. 2001, 101, 1267–1300; c) A. C.
Grimsdale, K. Leok Chan, R. E. Martin, P. G. Jokisz, A. B. Holmes, Chem.
Rev. 2009, 109, 897–1091; d) A. L. Kanibolotsky, I. F. Perepichka, P. J. Ska-
bara, Chem. Soc. Rev. 2010, 39, 2695–2728.
[23] A. D. Shaller, W. Wang, A. Li, G. Moyna, J. J. Han, G. L. Helms, A. D. Q. Li,
Chem. Eur. J. 2011, 17, 8350–8362.
[2] a) W. Zhang, J. S. Moore, Angew. Chem. 2006, 118, 4524–4548; Angew.
Chem. Int. Ed. 2006, 45, 4416–4439; b) D. Mçssinger, J. Hornung, S. Lei,
S. De Feyter, S. Hçger, Angew. Chem. 2007, 119, 6926–6930; Angew.
Chem. Int. Ed. 2007, 46, 6802–6806; c) T. Chen, G.-B. Pan, H. Wettach, M.
Fritzsche, S. Hçger, L.-J. Wan, H.-B. Yang, B. H. Northrop, P. J. Stang, J.
Am. Chem. Soc. 2010, 132, 1328–1333; d) A. D. Finke, D. E. Gross, A.
Han, J. S. Moore, J. Am. Chem. Soc. 2011, 133, 14063–14070.
[24] M. Myahkostupov, V. Prusakova, D. G. Oblinsky, G. D. Scholes, F. N. Cas-
tellano, J. Org. Chem. 2013, 78, 8634–8644.
[25] T. W. Anderson, G. D. Pantos¸, J. K. M. Sanders, Org. Biomol. Chem. 2011,
9, 7547–7553.
[26] B. Rybtchinski, L. E. Sinks, M. R. Wasielewski, J. Phys. Chem. A 2004, 108,
7497–7505.
[27] Z. Chen, B. Fimmel, F. Wꢀrthner, Org. Biomol. Chem. 2012, 10, 5845–
5855.
[3] S. Hecht, I. Huc, Foldamers: Structure, Properties and Applications, Wiley-
VCH, Weinheim, 2007.
[28] a) H. Langhals, S. Demmig, H. Huber, Spectrochim. Acta Part A 1988, 44,
1189–1193; b) K. D. Shimizu, T. M. Dewey, J. Rebek, Jr., J. Am. Chem. Soc.
1994, 116, 5145–5149; c) S. Kohmoto, K. Sakayori, K. Kishikawa, M. Ya-
mamoto, J. Chem. Soc. Perkin Trans. 2 1999, 833–836; d) L. D. Wescott,
D. L. Mattern, J. Org. Chem. 2003, 68, 10058–10066; e) C. F. Degenhardt,
J. M. Lavin, M. D. Smith, K. D. Shimizu, Org. Lett. 2005, 7, 4079–4081;
f) Y. Zhong, B. Kumar, S. Oh, M. T. Trinh, Y. Wu, K. Elbert, P. Li, X. Zhu, S.
Xiao, F. Ng, M. L. Steigerwald, C. Nuckolls, J. Am. Chem. Soc. 2014, 136,
8122–8130.
[4] a) D. J. Hill, M. J. Mio, R. B. Prince, T. S. Hughes, J. S. Moore, Chem. Rev.
2001, 101, 3893–4012; b) I. Saraogi, A. D. Hamilton, Chem. Soc. Rev.
2009, 38, 1726–1743; c) G. Guichard, I. Huc, Chem. Commun. 2011, 47,
5933–5941; d) D.-W. Zhang, X. Zhao, J.-L. Hou, Z.-T. Li, Chem. Rev. 2012,
112, 5271–5316.
[5] a) D. H. Appella, L. A. Christianson, I. L. Karle, D. R. Powell, S. H. Gellman,
J. Am. Chem. Soc. 1996, 118, 13071–13072; b) D. Seebach, P. E. Ciceri,
M. Overhand, B. Jaun, D. Rigo, L. Oberer, U. Hommel, R. Amstutz, H.
Widmer, Helv. Chim. Acta 1996, 79, 2043–2066; c) D. Seebach, M. Over-
hand, F. N. M. Kꢀhnle, B. Martinoni, L. Oberer, U. Hommel, H. Widmer,
Helv. Chim. Acta 1996, 79, 913–941; d) D. H. Appella, L. A. Christianson,
[29] M. Hesse, H. Meier, B. Zeeh, Spectroscopic Methods in Organic Chemistry,
2nd ed., Thieme, Stuttgart, 2007.
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
15
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[30] J.-E. Lee, V. Stepanenko, J. Yang, H. Yoo, F. Schlosser, D. Bellinger, B.
Engels, I. G. Scheblykin, F. Wꢀrthner, D. Kim, ACS Nano 2013, 7, 5064–
5076.
[31] Y. Cohen, L. Avram, L. Frish, Angew. Chem. 2005, 117, 524–560; Angew.
Chem. Int. Ed. 2005, 44, 520–554.
[32] R. F. Fink, J. Seibt, V. Engel, M. Renz, M. Kaupp, S. Lochbrunner, H.-M.
Zhao, J. Pfister, F. Wꢀrthner, B. Engels, J. Am. Chem. Soc. 2008, 130,
12858–12859.
[33] a) M. J. Ahrens, L. E. Sinks, B. Rybtchinski, W. Liu, B. A. Jones, J. M.
Giaimo, A. V. Gusev, A. J. Goshe, D. M. Tiede, M. R. Wasielewski, J. Am.
Chem. Soc. 2004, 126, 8284–8294; b) J. B. Birks, Rep. Prog. Phys. 1975,
38, 903–974.
[34] a) C. Hippius, I. H. M. van Stokkum, E. Zangrando, R. M. Williams, F.
Wꢀrthner, J. Phys. Chem. C 2007, 111, 13988–13996; b) K. M. Lefler, K. E.
Brown, W. A. Salamant, S. M. Dyar, K. E. Knowles, M. R. Wasielewski, J.
Phys. Chem. A 2013, 117, 10333–10345.
[35] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, 2nd ed., Kluwer
Academic/Plenum Publishers, New York, 1999.
[36] a) E. H. A. Beckers, S. C. J. Meskers, A. P. H. J. Schenning, Z. Chen, F.
Wꢀrthner, R. A. J. Janssen, J. Phys. Chem. A 2004, 108, 6933–6937; b) C.
Hippius, I. H. M. van Stokkum, E. Zangrando, R. M. Williams, M. Wykes,
D. Beljonne, F. Wꢀrthner, J. Phys. Chem. C 2008, 112, 14626–14638;
c) C. C. Hofmann, S. M. Lindner, M. Ruppert, A. Hirsch, S. A. Haque, M.
Thelakkat, J. Kçhler, Phys. Chem. Chem. Phys. 2010, 12, 14485–14491.
[37] a) A. T. Haedler, H. Misslitz, C. Buehlmeyer, R. Q. Albuquerque, A. Kçhler,
H.-W. Schmidt, ChemPhysChem 2013, 14, 1818–1829; b) J. B. Birks, D. J.
Dyson, I. H. Munro, Proc. R. Soc. London Ser. A 1963, 275, 575–588; c) T.
Fçrster, K. Kasper, Z. Elektrochem. 1955, 59, 976–980.
[38] a) R. Gvishi, R. Reisfeld, Z. Burshtein, Chem. Phys. Lett. 1993, 213, 338–
344; b) M. W. Holman, R. Liu, L. Zang, P. Yan, S. A. DiBenedetto, R. D.
Bowers, D. M. Adams, J. Am. Chem. Soc. 2004, 126, 16126–16133; c) C.-
C. You, R. Dobrawa, C. R. Saha-Mçller, F. Wꢀrthner in Supermolecular Dye
Chemistry, Vol. 258 (Ed.: F. Wꢀrthner), Springer, Berlin/Heidelberg, 2005,
pp. 39–82.
[39] a) J. Salbeck, J. Electroanal. Chem. 1992, 340, 169–195; b) R. O. Marcon,
S. Brochsztain, J. Phys. Chem. A 2009, 113, 1747–1752; c) N. Vꢅn Anh, F.
Schlosser, M. M. Groeneveld, I. H. M. van Stokkum, F. Wꢀrthner, R. M.
Williams, J. Phys. Chem. C 2009, 113, 18358–18368; d) M. Wolffs, N.
Delsuc, D. Veldman, N. Vꢅn Anh, R. M. Williams, S. C. J. Meskers, R. A. J.
Janssen, I. Huc, A. P. H. J. Schenning, J. Am. Chem. Soc. 2009, 131, 4819–
4829.
[40] a) A. Khan, C. Kaiser, S. Hecht, Angew. Chem. 2006, 118, 1912–1915;
Angew. Chem. Int. Ed. 2006, 45, 1878–1881; b) C. Song, T. M. Swager,
Org. Lett. 2008, 10, 3575–3578.
Received: September 11, 2014
Published online on && &&, 2014
&
&
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
16
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
&
Foldamers
B. Fimmel, M. Son, Y. M. Sung, M. Grꢀne,
B. Engels,* D. Kim,* F. Wꢀrthner*
&& – &&
Phenylene Ethynylene-Tethered
Perylene Bisimide Folda-Dimer and
Folda-Trimer: Investigations on
Folding Features in Ground and
Excited States
Where are the electrons? A detailed in-
vestigation on the behavior of two fold-
able perylene bisimide (PBI, see figure)
systems was carried out with regard to
the structural and optical peculiarities of
the phenylene ethynylene-tethered PBI
chromophores. By solvent regulation
the dye–dye arrangement can be
influenced externally thus enabling
switching between solvated monomers
or small p-stacks. Consequently, either
excimer formation or photoinduced
electron transfer dominates the proper-
ties of the photoexcited systems.
Chem. Eur. J. 2014, 20, 1 – 17
www.chemeurj.org
17
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ