Impact of molecular flexibility on binding strength and self-sorting of chiral π-surfaces

Article abstract of DOI:10.1021/ja202696d

In this work, we have explored for the first time the influence of conformational flexibility of π-core on chiral self-sorting properties of perylene bisimides (PBIs) that are currently one of the most prominent classes of functional dyes. For this purpose, two series of chiral macrocyclic PBIs 3a-c and 4a-c comprising oligoethylene glycol bridges of different lengths at the 1,7 bay positions were synthesized and their atropo-enantiomers (P and M enantiomers) were resolved. Single crystal analysis of atropo-enantiomerically pure (P)-3a not only confirmed the structural integrity of the ethylene glycol bridged macrocycle but also illustrated the formation of π-stacked dimers with left-handed supramolecular helicity. Our detailed studies with the series of highly soluble chiral PBIs 4a-c by 1- and 2-D ¹H NMR techniques, and temperature- and concentration-dependent UV/vis absorption and circular dichroism (CD) spectroscopy revealed that in π-π-stacking dimerization of these PBIs chiral self-recognition (i.e., PP and MM homodimer formation) prevails over self-discrimination (i.e., PM heterodimer formation). Our studies clearly showed that with increasing conformational flexibility of PBI cores imparted by longer bridging units, the binding strength for the dimerization process increases, however, the efficiency for chiral self-recognition decreases. These results are rationalized in terms of an induced-fit mechanism facilitating more planarized π-scaffolds of PBIs containing longer bridging units upon π-π-stacking.

Full text of DOI:10.1021/ja202696d

ARTICLE
pubs.acs.org/JACS
Impact of Molecular Flexibility on Binding Strength and Self-Sorting
of Chiral π-Surfaces
Marina M. Safont-Sempere,† Peter Osswald,† Matthias Stolte,† Matthias Gr€une,† Manuel Renz,‡
Martin Kaupp,‡ Krzysztof Radacki,§ Holger Braunschweig,§ and Frank W€urthner*,†
†Universit€at W€urzburg, Institut f€ur Organische Chemie and R€ontgen Research Center for Complex Material Systems, Am Hubland,
97074 W€urzburg, Germany
‡Institut f€ur Chemie, Theoretische Chemie, Technische Universit€at Berlin, Sekr. C 7, Strasse des 17. Juni 135, 10623 Berlin, Germany
§Universit€at W€urzburg, Institut f€ur Anorganische Chemie, Am Hubland, 97074 W€urzburg, Germany
S Supporting Information
b
ABSTRACT: In this work, we have explored for the first time the
influence of conformational flexibility of π-core on chiral self-sorting
properties of perylene bisimides (PBIs) that are currently one of the
most prominent classes of functional dyes. For this purpose, two series
of chiral macrocyclic PBIs 3aꢀc and 4aꢀc comprising oligoethylene
glycol bridges of different lengths at the 1,7 bay positions were
synthesized and their atropo-enantiomers (P and M enantiomers)
were resolved. Single crystal analysis of atropo-enantiomerically pure
(P)-3a not only confirmed the structural integrity of the ethylene
glycol bridged macrocycle but also illustrated the formation of π-
stacked dimers with left-handed supramolecular helicity. Our detailed
studies with the series of highly soluble chiral PBIs 4aꢀc by 1- and 2-D
1H NMR techniques, and temperature- and concentration-dependent
UV/vis absorption and circular dichroism (CD) spectroscopy revealed that in πꢀπ-stacking dimerization of these PBIs chiral self-
recognition (i.e., PP and MM homodimer formation) prevails over self-discrimination (i.e., PM heterodimer formation). Our studies
clearly showed that with increasing conformational flexibility of PBI cores imparted by longer bridging units, the binding strength
for the dimerization process increases, however, the efficiency for chiral self-recognition decreases. These results are rationalized
in terms of an induced-fit mechanism facilitating more planarized π-scaffolds of PBIs containing longer bridging units upon
πꢀπ-stacking.
’ INTRODUCTION
the bay area leads to a twist of the two naphthalene units of the
initial planar perylene core caused by repulsive interactions of
these substituents.8 Such contortion of the aromatic cores has
been observed for other extended π-systems such as hexabenzo-
coronenes or helicenes.9 The extent of this twisting is related to
the bulkiness of the bay substituents varying from 4° for a
1,7-difluoro-substituted PBI8a,8h up to 37° for 1,6,7,12-tetra-
chloro derivatives.8d,8f,10 A most interesting feature of the twisted
perylene core is symmetry breaking leading to inherent chirality
of the aromatic core. However, the isolation of single atropo-
enantiomeric PBIs has been scarcely achieved as in most of the
reported bay-substituted PBIs the two enantiomers are in a
dynamic equilibrium.11 To prevent racemization and, thus, to
achieve configurationally stable atropo-enantiomeric and diaster-
eomeric PBIs, our group has attached one or two bridging units
shielding one or both of the PBI π-faces, respectively.5,12 Similar
approaches have been applied also for other dyes, for example,
Self-sorting1 is an important branch of systems chemistry2 that
can be defined as high fidelity recognition of self, from nonself,
between molecules in complex mixtures.1a,e Self-sorting can be
classified as self-recognition if affinity for itself is shown, or self-
discrimination if on the contrary affinity for others is observed.1d
Chiral self-sorting is classified as well into self-recognition or self-
discrimination,3 depending on whether an enantiomer recog-
nizes itself or its mirror image leading to the formation of homo-
or heterochiral species, respectively. The importance of chiral
self-sorting is evident, if we take into account the chiral nature of
biomolecules such as sugars or aminoacids and the unknown
origin of homochirality in life.4 Studies on chiral self-sorting in
perylene bisimide (PBI) dyes were recently started in our group
with stable atropo-enantiomeric macrocyclic PBIs.5 PBI dyes are
an outstanding class of chromophores, which due to their optical
and electronic properties6 have been widely used for the devel-
opment of functional organic materials.7 Moreover, by means of
X-ray diffraction analysis of different bay-substituted PBIs, it
could be demonstrated that the introduction of substituents in
Received: April 1, 2011
Published: May 09, 2011
r
2011 American Chemical Society
9580
dx.doi.org/10.1021/ja202696d J. Am. Chem. Soc. 2011, 133, 9580–9591
|

Journal of the American Chemical Society
ARTICLE
in Hertz (Hz). Prior to ROESY measurement the samples were degassed
with dry argon gas. Alternating-phase 180° pulses were applied during
the mixing time in the ROESY experiments to suppress undesired
TOCSY contributions.21 High resolution mass spectra (HRMS) were
recorded on a MicroTOF Focus from Bruker Daltonics. Analytical
HPLC was carried out on a JASCO system (PU 2080 PLUS) with a
diode array detector (MD 2015), equipped with a ternary gradient unit
(DG-2080ꢀ533) and inline-degasser (LG 2080ꢀ02). Semipreparative
HPLC was performed on a JASCO system (PU 2080 PLUS) with an
UV/vis detector (UV 2077 PLUS). Preparative recycling GPC LC-9105,
Japan Analytical Industry Co., Ltd. (JAI) was used in recycling mode for
the resolution of some of the racemates. Vapor pressure osmometry
(VPO) measurements were performed on a KNAUER osmometer with
a universal temperature measurement unit. Benzil was used as standard
and a calibration curve in terms of R (ohm) vs molal osmotic
concentration (moles per kg n-hexane) was accomplished up to 0.016 m.
X-Ray Analysis. The crystal data of enantiopure (P)-3a were
collected at a Bruker X8APEX diffractometer with CCD area detector
and multilayer mirror monochromated MoKR radiation. The structure
was solved using direct methods and expanded applying Fourier
techniques.22 All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were assigned to idealized positions and were included
in structure factor calculations. For details see the Supporting Informa-
tion and CCDC 818346.
Figure 1. Schematic representation of strapped atropo-enantiomeric
PBIs and variation of the naphthalene units twisting angle depending on
the bridge length.
porphyrins by Osuka and co-workers13 and oligothiophenes by
Takeuchi and co-workers,14 or for some special liquid crystalline
substances by Swager and co-workers15 to induce intriguing
mesophase properties. In a recent communication we have
shown by NMR and concentration-dependent UV/vis studies
the formation of πꢀπ-stacked dimers with preferential self-
recognition, that is, MM and PP dimers for a chiral PBI derivative
with a rather short bridging unit connected to the 1 and 7 bay
positions.5 Other studies of self-sorting (chiral and general) of
PBIs have been reported by Li and co-workers, however with
conformationally unstable atropisomeric PBIs.16a,b
UV/vis Absorption, Fluorescence, and Circular Dichroism
(CD) Spectroscopy. For all spectroscopic measurements, spectro-
scopic grade solvents (Uvasol) from Merck were used. UV/vis spectra
were recorded with a Perkin-Elmer Lambda 950 spectrometer equipped
with a PTP-1 peltier element. CD spectra were measured with a JASCO
J-810 spectrometer equipped with a CDF-242 peltier element. Fluores-
cence spectra were recorded with a PTI QM-4/2003 instrument. All
fluorescence measurements were performed at room temperature and
fluorescence spectra were corrected against photomultiplier and lamp
intensity. The fluorescence quantum yields were determined as the
average value for three different excitation wavelengths using N,N0-
di(2,6-diisopropylphenyl)-3,4:9,10-tetracarboxylic acid bisimide as re-
ference (Φfl = 100% in chloroform) by applying high dilution conditions
(A < 0.05).23 For the measurements at high concentrations, front-face
set up with 30° light beam angle deviation and 1 mm cells were used.
Fluorescence lifetimes were measured under ambient conditions by
using a PTI LaserStrobe fluorescence lifetime spectrometer system
containing a GL-3300 nitrogen laser (337.1 nm, pulse width 600 ps,
pulse energy 1.45 mJ) coupled with a dye laser GL-302 (pulse width
500 ps, pulse energy 220 μJ at 550 nm) as an excitation source and a
stroboscopic detector. Laser output was tuned within the emission
curves of the laser dyes supplied with the instrument (PLD 421, 500,
579, 665, 735). The time resolution following deconvolution of experi-
mental decays is 200 ps. The instrument response function was collected
by scattering the exciting light of a dilute, aqueous suspension of Silica
(LUDOX). Decay curves were evaluated using the software supplied
with the instrument applying least-squares regression analysis. The
quality of the fit was evaluated by analysis of χ2 (0.9 ꢀ 1.1), DW factor
(>1.75) and Z value (<ꢀ1.96, confidence level 0.95) as well as by
inspection of residuals and autocorrelation function.
Here we present the synthesis of a series of chiral PBIs, bearing
oligoethylene glycol (OEG) bridges of different lengths between
the 1 and 7 bay positions. The substituents directly attached at
the 1,7 positions are in all cases phenoxy units, which afford
identical repulsive interactions between the bay substituents and,
thus, an equal twist of the naphthalene entities derived thereof.
However, elongation of the OEG bridging units provides the
molecules with greater conformational flexibility, leading to less
“imprinted” inherent chirality (see Figure 1), which profoundly
affects two determinant features: chiral self-sorting of these
molecules and binding strength in the dimerization process.
We have observed that the increase of flexibility in the PBI cores
adversely affects the self-recognition process. On the contrary, an
increase of the binding strength by this dimerization process
upon elongation of the bridging units has been observed, which
revealed that induced-fit17 plays a determinant role in our system.
The induced-fit principle has been widely applied to explain the
complexation process in various natural17,18 and synthetic19
hostꢀguest systems, in which molecular flexibility allows under-
going structural changes to form more stable hostꢀguest com-
plexes or selective binding of specific guests. Quite similarly, our
differently flexible macrocyclic PBIs undergo structural changes
in order to form more stable self-assembled entities by optimiz-
ing the πꢀπ-contact surfaces. These changes are more pro-
nounced for the most flexible structures, a rather unprecedented
feature in π-stacking dye aggregates.
’ EXPERIMENTAL SECTION
Computational Details. Ground-state DFT structure optimiza-
tions for 4a have been done with the Turbomole program package
(version 5.9.1)24 at the RI-BLYP-D/TZV(P) level25 including the
empirical dispersion corrections by Grimme et al.26 The same method
has recently been shown to provide accurate structures and binding
energies for unbridged PBI dimers.27 The C12H25 chains in 4a were
simplified to hydrogen atoms for computational efficiency.
Synthesis of PBIs 3b,c. N,N0-Dicyclohexyl-1,7-(3-hydroxyphe-
noxy)perylene-3,4:9,10-tetracarboxylic acid bisimide 15 (230 mg,
0.30 mmol for 3b and 400 mg, 0.52 mmol for 3c) and cesium carbonate
Cesium carbonate (99%) and potassium carbonate (g99%) were
obtained from commercial sources. All chemicals and reagents were
used as received, unless otherwise stated. Potassium carbonate was dried
in vacuo at 100 °C for 24 h. Diethylene glycol ditosylates were prepared
according to literature procedures.20 PBI derivatives 1 and 2 were prepared
as previously reported.5 Flash column chromatography was perfor-
1
med using silica gel (Si60, mesh size 40ꢀ63 μm). H NMR spectra
were recorded with Bruker Avance 400 MHz and Bruker DMX 600 MHz
instruments. Chemical shifts are given in parts per million (ppm) and are
referred to TMS as internal standard. 1H coupling constants J are given
9581
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
(1.7 g, 5.21 mmol for 3b and 3 g, 9.20 mmol for 3c) were dissolved in
DMSO (150 mL for 3b and 500 mL for 3c) and heated to 120 °C for 7 h.
The addition of the respective tri or tetra(ethylene glycol) distosylate
(280 mg, 0.57 mmol and 520 mg, 1.00 mmol, respectively) dissolved in
DMSO (20 mL) was carried out with a syringe pump over a period of
3 h. The reaction mixture was cooled to ambient temperature and
dropped into 0.33 M hydrochloric acid (150 mL) under stirring. The
resulting precipitate was collected by filtration and dried in vacuo
(10 mbar). Column chromatography with silica gel and dichloro-
methane (DCM) for 3b and alox (neutral) and DCM for 3c, with
subsequent precipitation with methanol from DCM solution afforded
61.0 mg (0.069 mmol) of 3b (23% yield) as an orange solid and 74.2 mg
(0.078 mmol) of 3c (15% yield) as a red solid.
dried in vacuo (10 mbar). Silica gel (4b: DCM /n-hexane, 1/1; 4c: DCM/
n-hexane, 3/2) column chromatography and precipitation with methanol
from DCM solution afforded 146 mg (0.078 mmol) of 4b (22% yield)
and 50.1 mg (0.026 mmol) of 4c (15% yield) as an orange solid.
4b. Mp 180ꢀ190 °C. 1H NMR (CDCl3): δ 9.52 (d, 2H, J = 8.2), 8.64
(d, 2H, J = 8.2), 8.48 (s, 2H), 7.33 (t, 2H, J = 8.3), 7.10 (m, 2H), 6.93
(m, 2H), 6.50 (s, 2H), 5.87 (m, 2H), 3.90ꢀ3.62 (m, 4H), 3.55ꢀ3.48 (m,
8H), 2.64 (t, 12H, J = 7.8), 1.59ꢀ1.55 (m, 12H), 1.32ꢀ1.18 (m, 108H),
0.83ꢀ0.78 (m, 18H). HRMS (ESI-TOF, pos. mode, DCM/acetonitrile =
1:1): m/z calc. for C126H181N2O10 [M þ H]þ: 1882.3716, found:
1882.3711. UV/vis (DCM): λmax/nm (εmax/Mꢀ1 cmꢀ1) 524 (77 700),
488 (48 100), 457 (18 100), 263 (43 000). Fluorescence (DCM): λmax
534 nm (λex = 525 nm), Φfl = 0.14.
=
3b. Mp 318ꢀ319 °C. 1H NMR (CDCl3): δ 9.42 (d, 2H, J = 8.2), 8.55
(d, 2H, J = 8.2), 8.38 (s, 2H), 7.32 (t, 2H, J = 8.3), 7.10 (m, 2H), 6.48 (m,
2H), 5.80 (s, 2H), 5.00 (m, 2H), 3.82ꢀ3.73 (m, 4H), 3.63ꢀ3.42 (m, 8H),
2.53 (m, 4H), 1.89 (bd, 4H), 1.74 (bd, 6H), 1.53ꢀ1.25 (m, 6H). HRMS
(ESI-TOF, pos. mode, DCM/acetonitrile = 1:1): m/z calc. for
C54H49N2O10 [M þ H]þ: 885.3309, found: 885.3382. UV/vis (DCM):
λmax/nm (εmax/Mꢀ1cmꢀ1) 522 (64 000), 486 (39 900), 454 (14 800), 266
(33 200). Fluorescence (DCM): λmax = 532 nm (λex = 525 nm), Φfl = 0.33.
3c. Mp 210ꢀ211 °C. 1H NMR (CDCl3): δ 9. 53 (d, 2H, J = 8.5), 8.58
(d, 2H, J = 8.5), 8.34 (s, 2H), 7.38 (t, 2H, J = 8.2), 7.11 (m, 2H), 6.63 (m,
2H), 6.03 (t, J = 2.3, 2H), 5.00 (m, 2H), 3.88ꢀ3.78 (m, 4H), 3.65ꢀ3.61
(m, 4H), 3.51ꢀ3.44 (m, 8H), 2.53 (m, 4H), 1.89 (bd, 4H), 1.74 (bd,
6H), 1.53ꢀ1.25 (m, 6H). HRMS (ESI-TOF, pos. mode, DCM/
acetonitrile =1:1): m/z calc. for C56H52N2O11Na [M þ Na]þ:
951.3469, found: 951.3463, λmax/nm (εmax/Mꢀ1 cmꢀ1) 525 (55 800),
4c. Mp 148 °Cꢀ155 °C. 1H NMR (CDCl3): δ 9.62 (d, 2H, J = 8.3),
8.67 (d, 2H, J = 8.3), 8.44 (s, 2H), 7.37 (t, 2H, J = 8.2), 7.11 (m, 2H),
6.93 (m, 2H), 6.03 (t, J = 2.3, 2H), 6.10 (m, 2H), 3.92ꢀ3.82 (m, 4H),
3.67ꢀ3.65 (m, 4H), 3.55ꢀ3.48 (m, 8H), 2.63 (t, 12H, J = 8.0),
1.59ꢀ1.55 (m, 12H), 1.32ꢀ1.18 (m, 108H), 0.83ꢀ0.78 (m, 18H).
HRMS (ESI-TOF, pos. mode, DCM/acetonitrile = 1:1): m/z calc. for
C127H182N2O11 [M þ H]þ: 1926.39001, found: 1926.39729, λmax/nm
(εmax/Mꢀ1 cmꢀ1
) 526 (64 900), 491 (41 300), 461 (15 800),
263 (43 000). Fluorescence (DCM): λmax = 551 nm (λex = 525 nm),
Φfl = 0.25.
Resolution of P-(þ)-4 and M-(ꢀ)-4 Enantiomers by HPLC.
Separation of the enantiomers was achieved on a semipreparative
column (Trentec Reprosil 100 chiral-NR) using different mixtures of
n-hexane and chloroform.
4b. 4b was eluted in a 65/35 mixture of chloroform/n-hexane using a
flow of 3.5 mL minꢀ1. The separation was done in the JAI recycling
system, after 7 cycles and two iterations on a scale of 10 mg.
M-(ꢀ)-4b: CD (DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 522 (ꢀ43), 487
(ꢀ30), 456 (ꢀ12), 286 (þ69), 263 (ꢀ67).
488 (35 500), 458 (13 800), 265 (35 200). Fluorescence (DCM): λmax
551 nm (λex = 525 nm), Φfl = 0.57.
=
Resolution of P-(þ)-3 and M-(ꢀ)-3 Enantiomers by HPLC.
Separation of the enantiomers was achieved on a semipreparative
column (Trentec Reprosil 100 chiral-NR) using different mixtures of
n-hexane and DCM as eluent.
P-(þ)-4b: CD (DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 522 (þ45), 487
(þ30), 456 (þ12), 286 (ꢀ69), 263 (þ69).
4c. 4c was eluted in a 7/3 mixture of chloroform/n-hexane using a
flow of 10 mL minꢀ1. The separation was done in the JAI recycling
system, after 7 cycles and five iterations on a scale of 5 mg.
M-(ꢀ)-4c: CD (DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 525 (ꢀ33), 488
(ꢀ18), 456 (ꢀ7), 286 (þ 32), 263 (ꢀ37).
3b was eluted in a 3/2 mixture of DCM/ n-hexane using a flow of
8 mL/min. The separation was done on a scale of 15 mg.
M-(ꢀ)-3b: Retention time (Trentec Reprosil 100 chiral-NR, DCM/
n-hexane (3/2), flow: 8.0 mL minꢀ1; Ø = 2.5 cm): 22.0 min. CD
(DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 518 (ꢀ47), 484 (ꢀ31), 454
(ꢀ12), 286 (ꢀ86), 262 (þ64).
P-(þ)-4c: CD (DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 525 (þ37), 488
(þ24), 456 (þ11), 286 (ꢀ 39), 263 (þ37).
P-(þ)-3b: Retention time (Trentec Reprosil 100 chiral-NR, DCM/
n-hexane (3/2), flow: 8.0 mL minꢀ1; Ø = 2.5 cm): 32.0 min. CD
(DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1): 518 (þ44), 484 (þ30), 454
(þ11), 286 (þ86), 262 (ꢀ64).
’ RESULTS
3c was eluted in a 4/1 mixture of DCM/n-hexane using a flow of
8 mL/min. The separation was done on a scale of 15 mg.
M-(ꢀ)-3c: Retention time (Trentec Reprosil 100 chiral-NR, DCM/
n-hexane (4/1), flow: 8.0 mL minꢀ1; Ø = 2.5 cm): 10.0 min. CD
(DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1) 522 (ꢀ34), 486 (ꢀ23), 458 (ꢀ9),
287 (ꢀ59), 264 (þ36).
Synthesis. PBIs 3 and 4 were synthesized from previously
reported5 precursors 1 and 2, respectively (Scheme 1). The 3,4,5-
trisdodecylphenyl substituents in PBIs 4aꢀc were introduced to
provide the molecules with improved solubility properties, while
the cyclohexane substituents in PBIs 3aꢀc facilitate the single
crystal growth. The free OH groups of PBIs 1 and 2 are available
for a two-fold Williamson etherification with differently sized
ditosylates.28 The macrocyclization reaction takes place in the
presence of cesium carbonate, which is known to facilitate ring
closures.29 A solution of the respective ditosylate was slowly
dropped over several hours by using a syringe pump into a
solution of the precursors 1 or 2 in DMSO or THF, respectively.
After work up of the reaction mixtures, the desired macrocycles
3aꢀc and 4aꢀc could be isolated as racemates. The chiral
resolution of these PBIs could be achieved by HPLC on a chiral
column (Trentec Reprosil 100 chiral-NR) using different experi-
mental conditions for each macrocycle in a purity of over 99%.
Optical Properties of the Monomers. The optical properties
of monomeric macrocycles 3 and 4 were investigated by UV/vis,
P-(þ)-3c: Retention time (Trentec Reprosil 100 chiral-NR, DCM /
n-hexane (4/1), flow: 8.0 mL minꢀ1; Ø = 2.5 cm): 12.0 min. CD
(DCM): λmax/nm (Δε/Mꢀ1 cmꢀ1) 522 (þ37), 486 (þ24), 458 (þ9),
287 (þ47), 264 (ꢀ39).
Synthesis of PBIs 4b,c. N,N0-Di(3,4,5-tridodecylphenyl)-1,7-
di(3-hydroxyphenoxy)perylene-3,4:9,10- tetracarboxylic acid bisimide
25 (300 mg, 0.17 mmol for both 4b and 4c), and cesium carbonate (1 g,
3.06 mmol) were dissolved in dry THF (400 mL) and refluxed at 90 °C
for 7 h. The addition of the correspondent tri or tetra(ethylene glycol)
distosylate (150 mg, 0.31 mmol and 170 mg, 0.32 mmol, for 4b and 4c
respectively) dissolved in THF (20 mL) was carried out with a syringe
pump over a period of 6 h. The reaction mixture was cooled to ambient
temperature and dropped into 0.33 M hydrochloric acid (150 mL)
under stirring. The resulting precipitate was collected by filtration and
9582
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Scheme 1. Synthetic Route and Chiral Resolution of Macrocyclic PBIs 3 and 4
CD spectra of their respective enantiomers in dichloromethane
(DCM) are shown in Figure 2 and Figure S1 in the Supporting
Information, and their optical data are summarized in Tables 1
and 2. The absorption spectra of macrocyclic PBIs 3aꢀc and
4aꢀc reveal S0ꢀS1 transition with three vibronic progressions
between 527ꢀ524 nm, 492ꢀ490 nm and 462ꢀ459 nm in DCM
(Figures 2 and S1 in the Supporting Information). The higher
energetic S0ꢀS2 transition has diminished in the case of 3a and
4a, while for 3c and 4c it is clearly visible. The absorption
coefficients are lower for 3aꢀc derivatives than for the respective
derivatives 4aꢀc (Table 1 and 2). This appears to be a general
absorption phenomenon and has been observed for other PBIs
with alkyl or aryl substituents in imide positions.30 Some minor
changes in the shape of the UV/vis spectra are observed for 3a/
4a, 3b/4b and 3c/4c that should be attributed to structural
changes related to the different bridge lengths. The CD spectra of
M and P enantiomers of 3aꢀc and 4aꢀc (top panels, Figure 2
and S1 in the Supporting Information) are mirror images as
expected for pure enantiomers. The stereochemical assignment
of these enantiomers was achieved by comparison of the
respective CD spectra with those of previously reported structu-
rally similar epimerically pure P- and M-configured macrocyclic
PBIs.31
Figure 2. (Bottom) UV/vis absorption spectra (solid lines) of racemic
PBIs 4aꢀc in DCM at 5 ꢁ 10ꢀ6 M and normalized emission fluores-
cence spectra (∼10ꢀ7 M, dashed lines, λex = 525 nm) in DCM of 4aꢀc:
4a (black line), 4b (red line) and 4c (blue line) at RT. (Top) CD spectra
of P enantiomer (solid lines) and M enantiomers (dashed lines) of 4aꢀc
in DCM at 5 ꢁ 10ꢀ6 M at RT.
Bisignate Cotton effects are not observed in the visible region.
This observation and the absence of further transitions in
excitonic coupled states in the spectral region of the monomer
S0ꢀS1 absorption support the exclusive presence of monomeric
species at this concentration (c = 5 ꢁ 10ꢀ6 M) in DCM. The fact
that the CD signal decreases in intensity upon increasing the
bridge length (top panels, Figure 2, Table 1 and 2) provides
further evidence that this CD band solely originates from the
fluorescence and circular dichroism (CD) spectroscopy. The
UV/vis absorption of the racemates (no changes are observed in
comparison with the UV/vis spectra of their enantiomers) and
9583
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Table 1. Optical Properties of Macrocyclic PBIs 3aꢀc in
DCM at RT
λmax(abs)
/
ε/
ꢀ1 cmꢀ1
λmax(em)
/
Φfl/
Δε (P)a/
Δε (M)a/
nm
M
nm
%
M
ꢀ1 cmꢀ1
M
ꢀ1 cmꢀ1
3a 525
3b 522
3c 525
56 000
64 000
56 000
538
533
550
27
33
57
51
43
36
ꢀ49
ꢀ44
ꢀ33
a Determined for maximum emission of S0-S1 transition.
Table 2. Optical Properties of Macrocyclic PBIs 4aꢀc in
DCM at RT
λmax(abs)
/
ε/
ꢀ1 cmꢀ1
λmax(em)
/
Φfl/
Δε (P)a/
Δε (M)a/
nm
M
nm
%
M
ꢀ1 cmꢀ1
M
ꢀ1 cmꢀ1
4a 525
4b 522
4c 525
65 000
76 000
65 000
545
534
553
9
14
25
49
43
37
ꢀ47
ꢀ45
ꢀ33
Figure 3. (a) πꢀπ interaction between two PBI molecules in a dimer
observed in the solid state structure of (P)-3a: Crystal packing (left) and
corresponding structural representation (right) of dimer A2 and dimer
B2. In the structural representation, the top molecule is highlighted in
red and the substituents in the bay area were omitted for clarity. (b)
Molecular packing of (P)-3a in the solid state. View along the a-axis
(left) and b-axis (right). The red line represents the a-axis, the green line
the b-axis, and the blue line the c-axis of the unit cell. Di(ethylene glycol)
bridges and included solvent molecules are omitted for clarity.
a Determined for maximum emission of S0-S1 transition.
twisted core. The drop of the signal intensity is about 33% from
3a,4a (shortest bridge) to 3c,4c (largest bridge) in the visible
region. The latter effect is even more pronounced in the
ultraviolet region with a descent in the intensity of about 58%.
Here, a distinct bisignate Cotton effect is observed that can be
attributed to the chiral excitonic coupling between transition
dipole moments that are polarized perpendicular to the long
molecular axis within the two naphthalene subunits in the PBI
aromatic core. The magnitude of Δε for such excitonic coupling
is directly related to the twist angle, which is a consequence of
repulsive interactions between the sterically encumbered bay
substituents.8dꢀf,12a Although racemization via interconversion
of atropo-enantiomers is not possible for PBIs 3 and 4 due to the
bridging unit on one π-surface, the decrease of the intensity of
the CD signal especially in this ultraviolet region is a clear
indication that on average a less pronounced twisting angle is
adopted upon increasing the bridge length, pointing to partial
planarization or even temporary flip of the naphthalene units as
suggested in Figure 1 for the most flexible PBIs 3c and 4c. For
these two macrocyclic PBIs with the longest OEG chains, a
partial interconversion between M and P forms can be imagined
based on molecular modeling studies. However, on average the
chiral sense is preserved as evidenced by the sign of the CD
spectra. Therefore, all these macrocyclic PBIs can be considered
as configurationally stable chiral molecules.
The fluorescence spectra of 3 and 4 show similar progressions as
their UV/vis spectra, however, are less pronounced upon increasing
bridge length. The observed emission maxima are between 532 and
553 nm (Figure 2 and S2 in the Supporting Information). The
emission maxima at the longest wavelengths are concomitant with
the largest Stockes shifts observed for 3c and 4c, suggesting more
pronounced structural relaxation processes of these more flexible
molecules in the excited state. Similarly large or even larger Stokes
shifts have been observed for tetraphenoxy-substituted PBIs that
were attributed to pronounced reorientation of those substituents.32
Again, no significant spectral differences are found for the racemic
mixtures and their respective pure enantiomers. The fluorescence
quantum yields have been determined in DCM (Table 1 and 2)
using N,N0-di(2,6-diisopropylphenyl)-3,4:9,10-tetracarboxylic acid
bisimide as reference (Φfl = 100% in CHCl3). It can be noticed that
the quantum yield increases upon enlarging the bridging unit. This
phenomenon has been observed previously and explained for
double-bridged PBIs.12a
X-Ray Diffraction Analysis. Single crystals of enantiopure
(P)-3a suitable for X-ray diffraction were obtained by dissolving
this compound in dichloromethane and addition of equal amounts
of methanol, followed by slow evaporation of the solvent. The
enantiomerically pure (P)-3a crystallized in the monoclinic space
group C2 and two crystallographically independent molecules were
found in the unit cell (Figure 3a).
The torsion of the central six-member ring was determined for
one type of molecules (denoted as B) to 16.9° and 18.5°, and for
the other (denoted as A) to 16.8° and 9.8°. Molecule A possesses
a bowl-like structure whereas the second type of molecule (B)
does not show any distortion along the N,N-axis of the perylene
bisimide. The low torsion of 9.8° as well as the bowl-like bending
of A might be due to packing effects. Even more interesting than
the molecular structures of A and B is the solid state packing of
the enantiopure macrocycle (P)-3a, as exclusive formation of
π-dimers for these homochiral macrocycles is observed, provid-
ing interesting structural information about the supramolecular
arrangement. The dimers are arranged in two perpendicular
layers with the perylene plane of 3a being either perpendicular to
the a-axis (dimer B2) or the c-axis (dimer A2) of the unit cell
(Figure 3b). The closest distance between the molecules was
determined to 3.27 Å for dimer B2 whereas for both π-dimers the
mean distance between two perylene bisimide π-surfaces was
elucidated to 3.47 Å (see Figure S4 in the Supporting In-
formation), which matches well with the distance observed in
the solid state for the πꢀπ interactions (3.34ꢀ3.55 Å) of the
fully planar core-unsubstituted PBIs.33 Both dimers A2 and B2 are
characterized by a rotational offset of 43° and 58°, respectively.
The rotational offset is accompanied by a longitudinal
offset along the N,N-axis of the perylene bisimide which is
negligible for dimer A2 but pronounced for dimer B2. The
helicity of the dimers induced by the rotational offset can be
9584
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
to the contemporaneous presence of homo and heterochiral dimers
in the racemic solution.5 For macrocycle 4a it was found that for
solutions with % ee not equaling 0 or 100% at c = 2 ꢁ 10ꢀ3 M two
0
sets of signals appear for each PBI core proton pair (Ha,Ha , Hb,
0
0
Hb , and Hc,Hc , see Figure 4). The signal splitting of these
protons is related to the different chemical environments of Hx
0
and Hx (x = a, b, or c) when they are in a homodimer or in a
heterodimer, or what is the same, due to the diastereomeric
relationship between homo- and heterodimers. The relative
0
intensities of the two signals related to one proton (Hx Hx ) of
4a correspond to the enantiomeric ratio, which can be directly
related to the % ee, and thus the integration of the NMR signals
can be used as direct measure of the % ee. Such correlation of the
intensities of NMR signals and the ee indicates that the system is
under fast exchange conditions.34 Similarly, splitting of the
signals at % ee different from 0 and 100% was found for 4b and
4c, pointing to diastereomeric relationships of the dimers (homo
and heterochiral) present in solution (Figure 4). However, for 4b
the splitting of the signals is less pronounced than for 4a, and the
splitting of 4c is even less pronounced than for 4b (Figure 4).
This phenomenon indicates that upon increasing the bridge
length and hence the flexibility of the molecules, the diastereo-
meric relationship between the homo and heterochiral dimers
becomes less evident.
Consequently, the observed splitting of NMR signals is also
less pronounced for the macrocyclic PBIs with larger bridging
units (4b and 4c), which prohibits a proper analysis of the data
Figure 4. 600 MHz 1H NMR spectra of solutions with different ee in
n-hexane-d14 at 331 K (c = 2 ꢁ 10ꢀ3 M) (range of PBI protons) for
(a) 4b and (b) 4c.
over the entire concentration range. Nevertheless, at least for 4b,
0
the % ee can be determined from Hc,Hc signals between 100 and
26% (Table S1 in the Supporting Information).
ROESY and COSY NMR Spectroscopy. Several attempts were
made to obtain proper ROESY spectra of macrocycles (P)-4aꢀc
(or (M)-4aꢀc) at 331 and 298 K in n-hexane-d14 at a concentra-
tion of 2 ꢁ 10ꢀ3 M. The advantage of the measurements at high
temperature (331 K) is the sharpening of the NMR signals.
Unfortunately, no intermolecular cross peaks at this temperature
could be observed in any of the ROESY experiments performed
for these pure enantiomeric PBIs. We attribute this fact to the fast
exchange conditions at this temperature. To overcome this draw-
back, experiments at 298 K were performed with (P)-4aꢀc. At this
temperature, the NMR signals of macrocyclic (P)-4a and (P)-4c are
too broad, especially in the aromatic region for ROESY measure-
ments. Fortunately, for (P)-4b signals, which are sharp enough
for recording, a successful ROESY spectrum was found at RT
(Figure 5a).
As expected, intramolecular cross peaks between neighboring
aromatic protons Ha and Hb can be observed in the ROESY
spectrum. More interestingly, a much weaker cross peak appears
between the distant aromatic protons Ha and Hc (Figure 5a). To
rule out the possibility that this cross peak originates from the
indirect coupling between Ha and Hc protons through the
aromatic naphthalene unit, a COSY experiment was carried out
under the same conditions as for the ROESY one. As can be
appreciated in Figure 5b, expected intramolecular cross peaks
between aromatic protons Ha and Hb can be observed. How-
ever, no trace of the cross peak resulting from the coupling
between aromatic protons Ha and Hc is present. Accordingly,
this cross peak between Ha and Hc is attributed to an inter-
molecular contact in the homochiral dimer where these protons
come into close proximity (see DFT calculations also).35
Optical Properties of the Dimers. The aggregation behavior
of PBIs 4aꢀc (racemate and pure enantiomers) was investigated
assigned as left-handed for both dimers A2 and B2. This indicates
that the (P)-configuration of the twisted perylene core enforces
the formation of dimers with left-handed supramolecular chi-
rality. In contrast to enantiopure (P)-3a, no π-stacked dimers
could be observed in single crystals of racemic PBI 3a.5 Attempts
to grow single crystals of PBIs 3b and 3c using the same
procedure as applied for PBI 3a were not successful, probably
due to the more flexible bridging units of the former structures.
Dimer Formation in Solution. Dimer formation of highly
concentrated solutions of 4aꢀ4c could be demonstrated by
vapor pressure osmometry (VPO) measurements. VPO experi-
ments of 4b and 4c (c = 1 ꢁ 10ꢀ3 to 5 ꢁ 10ꢀ3 M) in n-hexane at
328 K using benzil for calibration gave aggregation numbers N
(N = Kcal/Kmean) of 1.81 and 2.05, respectively (Figure S5 in the
Supporting Information).
1
Comparative H NMR studies of 4b,c (racemates and their
respective pure enantiomers) in n-hexane-d14 at 331 K were
performed to elucidate structural differences between homo- and
heterochiral dimers. At the measured concentration, dimers are
the major species as indicated by VPO and concentration-
dependent UV/vis studies (about 80 and 90% of dimers are
present in 4b and 4c, respectively, according to the data shown in
Figures 8 and S7bꢀS12b in the Supporting Information). In
contrast to the spectra measured at room temperature (RT), the
1H NMR signals obtained at 331 K are sharper, thus allowing a
quantitative evaluation of the NMR data. The superposed spectra
of the racemates and their respective enantiomers revealed some
differences in the shift of the aromatic protons, especially the
ones corresponding to the protons of the PBI core. We have
previously demonstrated that the observed differences between
the racemates and their respective pure enantiomers are related
9585
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Figure 5. 600 MHz (a) ROESY and (b) COSY NMR data of (P)-4b in n-hexane-d14 at 298 K (c = 2 ꢁ 10ꢀ3 M) (range of PBI protons). The ROESY
spectrum was recorded with a mixing time of 200 ms. Positive and negative signals are represented by black and red lines, respectively.
by concentration-dependent UV/vis, CD and fluorescence spec-
troscopy in a concentration range between 10ꢀ6 and 10ꢀ3 M in
n-hexane (for CD and fluorescence) and n-heptane (for UV/vis)
at RT.36
dependence indicates that these excimers are not formed from excited
monomeric PBIs by diffusion but originated from photoexcited PBI
dimers.
To further elucidate the fluorescence properties of these
macrocyclic PBIs 4aꢀc lifetime experiments were carried out.
The fluorescence decay times monitored at the monomer band
(535 nm) and the aggregate band (610 nm) are collected in
Table S2 in the Supporting Information. Lifetime decay values of
about 4 and 32 ns for the monomer and dimer bands, respec-
tively, were found in all cases, in good accordance with values
found for extended self-assembled π-stacks of PBIs40 and
covalently fixed dimeric π-stacks.41 It is remarkable, that neither
the shape of the fluorescence spectra nor the fluorescence decay
values are strongly affected by the length of the bridge unit in
4aꢀc which is in striking contrast to UV/vis and CD spectra. The
rationale behind these differences might be that the dimer
aggregates of these PBIs exhibit more similar structures in the
relaxed excited state than in the ground state. Quantum chemical
calculations have suggested a structural rearrangement of PBI
dimers in the excimer state,27 which might indeed afford quite
similar structures for all three PBI derivatives.
Concentration-dependent UV/vis studies of 4aꢀc in n-hep-
tane show pronounced spectral changes, including a hypsochro-
mic shift of the absorption maxima upon increasing concen-
tration (Figure 6, bottom panels), indicating the formation
of H-type aggregated species.37 Concentration-dependent
UV/vis spectra show no evident differences between racemates
and pure enantiomers (Figures 6 and S7ꢀS12 in the Supporting
Information). Spectral changes, however, are observed in the
shape of the absorption bands for differently bridged PBIs 4aꢀc,
being most pronounced for the aggregate spectrum of 4c
(Figure 6c). Two clear isosbestic points are observed at about
470 and 525 nm (more or less shifted for 4a, 4b or 4c),
confirming the presence of two species, that is, monomeric and
dimeric, in equilibrium. CD spectra of 4aꢀc M and P enantio-
mers were measured at concentration of 1 ꢁ 10ꢀ3 (strongly
aggregated) and 5 ꢁ 10ꢀ6 M (not aggregated) in n-hexane at RT
(see Figure 6, top panels). Hypsochromic shifts of the CD
maxima are observed in accordance with the spectral changes
recorded in the UV/vis spectra. Characteristic bisignate Cotton
signals are present for all studied macrocyclic PBIs, indicating the
helical nature of the formed dimers in solution, as it has been
previously demonstrated for PBI 4a.5 The negative sign of the
exciton couplet for (P)-4aꢀc points to a left-handed helicity (the
opposite for (M)-4aꢀc) of the respective dimeric aggregate
according to exciton chirality method.38 Interestingly, the in-
tensity of the bisignate signal in the ultraviolet region corre-
sponding to the chiral exciton coupling between the transition
localized in the two naphthalene subunits of the PBI core is
reduced in the case of the dimeric aggregates compared to the
respective monomeric dyes 4aꢀc. This result can be attributed to
a decrease of the core twisting angle of these macrocyclic PBIs
upon πꢀπ-stacking after optimization of the contact surfaces of
the interacting naphthalene subunits, and thus a special case of
“induced-fit” mechanism can be implicated for the π-stacking
dimerization of PBIs 4aꢀc.
’ DISCUSSION
Quantification of Chiral Self-Sorting and Effect of Tem-
perature. The binding constants KD for the dimerization of
(rac)-4aꢀc, (P)-4aꢀc and (M)-4aꢀc were determined by
concentration-dependent UV/vis spectroscopy at 331 K (Figure 8
and S7ꢀS12 in the Supporting Information and ref 5) in n-heptane.
The concentration-dependent UV/vis absorption spectra of
the racemates and their respective pure enantiomers (all experi-
ments were repeated twice) of these three macrocyclic com-
pounds appear virtually identical, indicating similar aggregation
behavior. The fitting of the experimental data obtained from the
concentration-dependent UV/vis absorption studies can be
carried out by nonlinear regression analysis using the dimeriza-
tion model42 or applying a multilinear analysis of the data carried
out by a custom-made program.43 Preferential self-recognition,
that is, formation of homo dimers, at 331 K in n-heptane could
recently be proven for compound 4a with such methods.5 Thus,
similar studies on 4b and 4c will provide enough information to
unravel the influence of the increasing molecular flexibility of the
PBI derivatives on their chiral self-sorting. Moreover, all mea-
surements were repeated at 298 K (see Figures S13ꢀS21 in the
Concentration-dependent fluorescence spectra of 4aꢀcare shown
in Figures 7 and S6 in the Supporting Information. The most
remarkable feature is that fluorescence spectra of highly concentrated
solutions (c = 5 ꢁ 10ꢀ4 M) show similar excimer-type emission as
observed for extended π-stacks.37,39 The pronounced concentration
9586
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Figure 6. Concentration-dependent UV/vis absorption (racemates, bottom panels) in n-heptane and CD spectra (enantiomers, top panels; (M)-
enantiomer: blue lines, (P)-enantiomer: red lines) of 4aꢀc in n-hexane, in the range from 5 ꢁ 10ꢀ6 M (dashed lines) to 1 ꢁ 10ꢀ3 M (solid lines) at RT
for (a) 4a, (b) 4b, and (c) 4c. Arrows indicate changes upon increasing concentration.
Figure 7. Normalized concentration-dependent fluorescence spectra of
4a in n-hexane at 5 ꢁ 10ꢀ4 M (dashed line) to 5 ꢁ 10ꢀ6 M (solid line) at
Figure 8. Concentration-dependent UV/vis absorption spectra of (rac)-
4b in n-heptane at 331 K ([M]0 = 1 ꢁ 10ꢀ6 to 8 ꢁ 10ꢀ4 M) (dotted lines)
RT measured in a front face setup.
and results from the multilinear fit routine showing the derived pure
monomer (red) and dimer (blue) spectra. (Inset) Molar fraction of
monomer species obtained from nonlinear regression analysis at selected
wavelengths. Arrows indicate changes upon increasing concentration.
Supporting Information) to show the possible influence of
temperature in the efficiency of chiral self-sorting events. The
data obtained at 298 K are collected together with those obtained
at 331 K for better comparison (Table 3).
analysis, we consider these results to be more reliable in compar-
ison to those obtained by single wavelength fit of the data.
Therefore, for further analysis, only the KD values obtained through
multilinear routine will be considered.
In Figures 8 and S7ꢀS12 in the Supporting Information,
concentration-dependent UV/vis studies of (rac)-4b,c, (P)-4b,c,
and (M)-4b,c at 331 K in n-heptane in a concentration range
between 10ꢀ3 and 10ꢀ6 M are shown, as well as their fit by
nonlinear regression analysis at selected wavelengths and their
multilinear fit. Analogous to what was recently reported for 4a,5
macrocyclic PBIs 4b and 4c show well-defined isosbestic points.
Nonlinear regression analysis of the data at selected wavelengths
show coherence between the dimerization constants KD ob-
tained at each wavelength. Moreover, consistent results are
obtained through multilinear analysis of the data, which is
confirmed by the good accordance between the calculated and
the experimental UV/vis spectra at each measured concentration.
Due to the larger amount of data processed in the multilinear
The results obtained from nonlinear regression analysis at
selected wavelength as well as from multilinear regression
analysis (Figures 8 and S7ꢀS12 in the Supporting Information)
are summarized in Tables S3ꢀS6 in the Supporting Information.
The KD values obtained after multilinear analysis of the concen-
tration-dependent UV/vis data of 4b and 4c at 331 and 298 K in
n-heptane for the quantitative analysis of chiral self-sorting are
collected in Table 3 (some data of 4a already presented in ref 5
are also included for clarity in further discussion).
Applying eq 15 to the values collected in Table 3 afforded
quantitative information about the ratio of homo and heterochiral
9587
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Table 3. Dimerization Constants KD Determined for (rac)-
4aꢀc, (P)-4aꢀc and (M)-4aꢀc in n-Heptane at 331 and 298 K
by Multilinear Fit of the Concentration-Dependent UV/vis
Absorptiona
Table 5. Calculated Equilibrium Constants (K0) and Related
Gibbs Energies (ΔG0) for the Formation of Heterodimers
from a Racemic Mixture of Homodimers of 4aꢀc at 298 and
331 K
sample
T/K
KD (P)b/Mꢀ1
KD (M)b/ Mꢀ1
KD (rac)b/Mꢀ1
T/ K
sample
K0
ΔG0/kJ molꢀ1
4ac
331
298
331
298
331
298
2800
12 000
5200
2800
12 100
5000
1500
6100
298
4a
4b
4c
4a
4b
4c
7 ꢁ 10ꢀ5
0.073
23.8
6.5
4b
4c
3100
0.293
3.0
18 100
9100
19 700
9300
12 000
7000
331
0.005
14.5
8.4
0.047
38 400
39 200
29 900
0.272
3.6
a Values are averaged from two independent data sets. b Error of the
multilinear analysis is about (15%. c Values for 4a are taken from ref 5.
naphthalene units, being more pronounced for larger bridges.
The flipping process of molecules 4b and 4c (being more
pronounced for 4c, due to the larger bridging unit) provides
the molecules with more stacking possibilities, some of them
allowing a better overlapping of the aromatic surfaces in the
heterochiral dimers (temporary planarization and/or flipped
structures). These results point to larger amounts of hetero
dimers upon increasing the bridge length owing to the flexibility
of the molecules, which is associated with a decrease of the
average twisting angle of the PBI core. Moreover, the quantity of
homochiral dimers at both temperatures for the three studied
racemic compounds is nearly the same. In the light of this data,
we can conclude that temperature has a very small effect on the
fidelity of the chiral self-recognition of 4aꢀc. This is in good
accordance with other reported systems, in which temperature
only plays an important role when the thermodynamic equilib-
rium has not been achieved.1b,44
It is also notable that the binding constants increase progres-
sively for larger bridging units, being largest for the nonbridged
precursor 2. From 4a to 4b and from 4b to 4c, the binding
constants double for each successive increase of the OEG bridge
length. This trend is lost for precursor 2, which has a much higher
binding constant than 4c (about six times higher than KD(homo)
and eight times higher than KD(rac)). This progression can be
rationalized by the induced-fit effect, by which the PBIs undergo
an appreciable change in their three-dimensional structure upon
formation of a π-stacked dimer. The extent of the induced fit is
directly related to the flexibility of the molecules. The most
flexible molecule, precursor 2, is able to accomplish the most
pronounced structural changes to fit the other PBI molecule
within the dimer complex. The tremendous difference between
the binding constants of 2 and (rac)-4c (Table 3) can be
explained by three features of molecule 2: (i) the presence of
two free π-surfaces available for πꢀπ-stacking (purely entropic
contribution), (ii) the possibility of complete flipping of the
naphthalene units from M to P atropo-enantiomers, and (iii) the
possibility of the phenoxy groups attached to the 1,7-bay posi-
tions to freely rotate and adopt the most favorable conformation
to stabilize the self-assembled π-stack. The conformational
changes of these phenoxy units are considerably limited by the
presence of the OEG bridges, thus only the effect of the more or
less restricted flipping of the naphthalene units will contribute to
the induced fit process for the dimerization process of 4aꢀc.
Consequently, the increasing flexibility in PBI aromatic cores has
a positive influence on the induced-fit effects in the dimerization
process as reflected in the calculated binding constants, while it
clearly compromises the quality of the chiral self-sorting process.
Table 4. Dimerization Constants KD(hetero) Obtained From
the Values of Table 3 for (rac)-4aꢀc, (P)-4aꢀc and (M)-4aꢀc
in n-Heptane at 331 and 298 K According to eq 1, the Ratio of
Homo- versus Heterodimers Derived Thereof, and Corre-
sponding Percentage of Homodimers Present in Each of the
Studied Racemates
a
a
T/ KD(homo)
/
KD(hetero)
/
[homo]/[hetero] =
sample
K
Mꢀ1
Mꢀ1
2KD(homo)/KD(hetero) % homo
4a
331
298
331
298
331
298
331
2800
12 100
5100
400
200
14.0
120.5
4.6
93
99
82
79
66
65
4b
4c
2
2200
18 900
9200
10 200
9600
3.7
1.9
38 800
55000 ( 5400b
42 000
1.8
a Error of the multilinear analysis is about (15%. b See Figure S22 in the
Supporting Information.
dimers in a racemic solution (Table 4). The KD of enantiopure
(P)-4aꢀc (or (M)-4aꢀc) is related to the formation of homo-
chiral dimers (i.e., KD = KD(homo)), while the obtained KD(rac)
from (rac)-4aꢀc is the average value of homo and hetero
dimerization processes.
K
DðheteroÞ þ 2KDðhomoÞ
KDðracÞ
¼
ð1Þ
4
The results in Table 4 clearly show that the increasing
flexibility from PBIs 4a to 4c achieved by the introduction of
more extended bridging units between 1 and 7 positions
decreases the fidelity of chiral self-recognition in the dimerization
process. Thus, upon increasing the bridge length the amount
of homochiral dimers present in the racemate decreases from
99ꢀ93% in 4a, to 82ꢀ79% in 4b, and to 66ꢀ65% in 4c.
Macrocyclic PBI 4a with the shortest bridge shows the highest
fidelity for chiral self-recognition, which must be directly related
to the higher rigidity of its twisted structure. The shortness of the
bridge does not allow the twist angle of the PBI core to flip and
hence no temporary planar conformation can be achieved, which
restricts the stacking possibilities for the dimerization of PBIs 4a.
It has already been shown by CD studies that the twisting angle in
the aromatic core region decreases on average upon increasing
the length of the bridging unit, which can be explained by a partial
planarization of the molecules resulting from the flipping of the
9588
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
Figure 9. Energy diagram of the different supramolecular diastereomers of 4a. The OEG are represented in gray, the phenoxy units in blue, and the PBI
core in red for M-enantiomer and yellow for P-enantiomer.
Equilibrium Constant K0 for the Formation of Heterodi-
mers from Homodimers. The equilibrium constant (K0) and
related Gibbs energy (ΔG0) for the formation of heterodimers
from a racemic mixture of homodimers can be calculated
applying the values obtained for KD(homo) and KD(hetero) into
eq 2, and the obtained K0 into ΔG0 = ꢀRT In K0 (where R is the
ideal gas constant and T is the temperature in Kelvin). The
calculated values for 4aꢀc at 298 and 331 K are summarized in
Table 5.5
data at high concentrations in n-hexane, whose bisignate Cotton
effects indicate the presence of dimers with rotational displace-
ments. Thus, only these dimeric species will be considered for
further discussion (the preference for the eclipsed structure
in the computation by 5 kJ molꢀ1 is entirely due to a larger
dispersion contribution of about 15 kJ molꢀ1, whereas the
rotated structure is favored by electrostatics27).
The energies obtained for both homochiral dimers are, as
expected, equal within possible errors arising from incomplete-
ness of structure optimization convergence on very shallow
potential energy surfaces (Figure 9). The computed dimerization
KD2 ðheteroÞ
4K0 ¼
ð2Þ
energy for the rotated homochiral dimer is ca. 173 kJ molꢀ1
,
KD2 ðhomoÞ
which is larger than obtained previously at the same level for an
unsubstituted PBI dimer (125 kJ molꢀ1).27 The heterochiral
dimer was calculated to be 14.5 kJ molꢀ1 (19.2 kJ molꢀ1 for the
ecliptic conformer) less stable than the homodimers in excellent
accordance with the experimental values (Table 5), supporting
the validity of the calculations. Thus, the higher stability of
homochiral dimers over their heterochiral counterparts has been
demonstrated both experimentally and theoretically. The reason
of this extra stabilization in the case of the homodimers is clearly
the better overlapping of the π-surfaces in comparison to the
heterochiral dimers as can be appreciated from Figure 9. The
better fit between the, in this case self-complementary aromatic
surfaces of the homodimer is apparent. The heterodimer exhibits
less tight πꢀπ-stacking and some distortion of the monomer
structures in the dimer (Figure 9). Such geometrical factors have
been previously proven to be decisive in the efficiency of both
chiral and “general” self-sorting processes.13c,16b,45
The data in Table 5 show that ΔG0 values are steadily positive
for the whole series. Thus, the formation of heterodimers from
the respective homodimers remains in all cases a disfavored
process. The highest fidelity of chiral self-recognition, i.e. pre-
ference for homo versus heterodimer formation, is found for the
shortest bridging unit. Upon increasing the bridge length and
hence the flexibility of the structures of 4aꢀc, the formation of
heterodimers becomes energetically less unfavorable, enabling an
increasing amount of heterochiral dimers in the racemic solution.
The tendency shown by these data anticipates that a long enough
bridging unit might provide a system in which the formation of
heterodimers is favored (for entropic reasons) over the forma-
tion of homodimers, or at least equally favorable.
For a better understanding of the experimental results, DFT
structure optimization of homo and heterodimers formed by 4a
were performed at the RI-BLYP-D/TZV(P) level (cf. Computa-
tional Details). For the different diastereomers (MM, PP and
MP) two different structure types with energetically similar
minima could be found (Figure S23 in the Supporting In-
formation). One of the structures shows an ecliptic arrangement,
with the PBIs almost superimposed and very small rotational
displacement (Figure S23a in the Supporting Information). The
second, marginally less favored structure exhibits two stacked
PBIs with rotational displacement of about 30° (Figure S23b in
the Supporting Information). The latter arrangement is consis-
tent with previous experimental and theoretical results for
unsubstituted PBIs27 in their ground state, and with our CD
’ CONCLUSIONS
This series of core-twisted dyes enabled for the first time the
elucidation of the effect of conformational parameters (rigidity of
a chiral π-scaffold) on binding constants for dimerization by
πꢀπ-stacking and selectivity of chiral recognition (homo- versus
heterochiraldimerization). Already CD spectraof monomeric pure
enantiomers revealed a greater flexibility of the PBI structures
equipped with larger bridging units owing to a more pronounced
flipping of the chiral PBI scaffold. Concentration-dependent
9589
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
spectroscopic studies of racemic and enantiopure PBI dyes 4aꢀc
provided unambiguous evidence for preferential chiral self-
recognition over self-discrimination in π-stacking dimerization
of these PBIs. However, the amount of heterochiral dimers
present in the corresponding racemates increases for more
extended bridging units. This demonstrates that the fidelity of
chiral self-sorting in PBIs is compromised by an increasing
flexibility of the structures. Less rigid scaffolds bearing longer
bridging units were shown to enable more planarized π-scaffolds
upon πꢀπ-stacking by means of a induced-fit mechanism
leading to higher binding strength but lower enantioselectivity
for the recognition of the homo or heterochiral partner molecule.
To conclude, we have revealed for the first time the impact of
rigidity on chiral self-sorting processes for core-twisted π-scaf-
folds. Conformational flexibility was shown to improve the
binding strength via induced-fit mechanism but at the expense
of the quality of chiral self-recognition between chiral π-surfaces.
Funct. Mater. 2010, 20, 50–58. (c) Weil, T.; Vosch, T.; Hofkens, J.;
Peneva, K.; M€ullen, K. Angew. Chem., Int. Ed. 2010, 49, 9068–9093.
(8) (a) Schmidt, R.; Oh, J. H.; Sun, Y. S.; Deppisch, M.; Krause,
A. M.; Radacki, K.; Braunschweig, H.; K€onemann, M.; Erk, P.; Bao,
Z. A.; W€urthner, F. J. Am. Chem. Soc. 2009, 131, 6215–6228. (b)
Sautter, A.; Thalacker, C.; W€urthner, F. Angew. Chem., Int. Ed. 2001,
40, 4425–4429. (c) Jones, B. A.; Ahrens, M. J.; Yoon, M. H.; Facchetti,
A.; Marks, T. J.; Wasielewski, M. R. Angew. Chem., Int. Ed. 2004,
43, 6363–6366. (d) Chen, Z.; Debije, M. G.; Debaerdemaeker, T.;
Osswald, P.; W€urthner, F. ChemPhysChem 2004, 5, 137–140. (e)
W€urthner, F.; Stepanenko, V.; Chen, Z.; Saha-M€oller, C. R.; Kocher,
N.; Stalke, D. J. Org. Chem. 2004, 69, 7933–7939. (f) Leroy-Lhez, S.;
Baffreau, J.; Perrin, L.; Levillain, E.; Allain, M.; Blesa, M. J.;
Hudhomme, P. J. Org. Chem. 2005, 70, 6313–6320. (g) Chao, C. C.;
Leung, M. K.; Su, Y. O.; Chiu, K. Y.; Lin, T. H.; Shieh, S. J.; Lin, S. C.
J. Org. Chem. 2005, 70, 4323–4331. (h) W€urthner, F.; Osswald, P.;
Schmidt, R.; Kaiser, T. E.; Mansikkamaki, H.; K€onemann, M. Org. Lett.
2006, 8, 3765–3768.
(9) (a) Xiao, S.; Myers, M.; Miao, Q.; Sanaur, S.; Pang, K.;
Steigerwald, M. L.; Nuckolls, C. Angew. Chem., Int. Ed. 2005,
44, 7390–7394. (b) Fasel, R.; Parschau, M.; Ernst, K. H. Nature 2006,
439, 449–452. (c) Phillips, K. E. S.; Katz, T. J.; Jockusch, S.; Lovinger,
A. J.; Turro, N. J. J. Am. Chem. Soc. 2001, 123, 11899–11907.
(10) Gs€anger, M.; Oh, J. H.; K€onemann, M.; Hoffken, H. W.; Krause,
A. M.; Bao, Z. N.; W€urthner, F. Angew. Chem., Int. Ed. 2010, 49, 740–743.
(11) (a) Xie, Z.; W€urthner, F. Org. Lett., 2010, 12, 3204ꢀ3207. (b)
Osswald, P.; W€urthner, F. J. Am. Chem. Soc. 2007, 129, 14319–14326.
(12) (a) Osswald, P.; Leusser, D.; Stalke, D.; W€urthner, F. Angew.
Chem., Int. Ed. 2005, 44, 250–253. (b) Osswald, P.; Reichert, M.;
Bringmann, G.; W€urthner, F. J. Org. Chem. 2007, 72, 3403–3411.
(13) (a) Hwang, I. W.; Kamada, T.; Ahn, T. K.; Ko, D. M.;
Nakamura, T.; Tsuda, A.; Osuka, A.; Kim, D. J. Am. Chem. Soc. 2004,
126, 16187–16198. (b) Kamada, T.; Aratani, N.; Ikeda, T.; Shibata, N.;
Higuchi, Y.; Wakamiya, A.; Yamaguchi, S.; Kim, K. S.; Yoon, Z. S.; Kim,
D.; Osuka, A. J. Am. Chem. Soc. 2006, 128, 7670–7678. (c) Mizumura,
M.; Shinokubo, H.; Osuka, A. Angew. Chem., Int. Ed. 2008,
47, 5378–5381. (d) Maeda, C.; Kamada, T.; Aratani, N.; Sasamori, T.;
Tokitoh, N.; Osuka, A. Chem.—Eur. J. 2009, 15, 9681–9684.
(14) (a) Sugiyasu, K.; Honsho, Y.; Harrison, R. M.; Sato, A.; Yasuda,
T.; Seki, S.; Takeuchi, M. J. Am. Chem. Soc. 2010, 132, 14754–14756. (b)
Sugiyasu, K.; Takeuchi, M. Chem.—Eur. J. 2009, 15, 6350–6362.
(15) (a) Long, T. M.; Swager, T. M. J. Mater. Chem. 2002,
12, 3407–3412. (b) Lohr, A.; Swager, T. M. J. Mater. Chem. 2010,
20, 8107–8111. (c) Lim, J.; Swager, T. M. Angew. Chem., Int. Ed. 2010,
49, 7486–7488.
’ ASSOCIATED CONTENT
Supporting Information. 1H NMR spectra of new com-
S
b
pounds, additional UV/vis spectra and binding constants, and
X-ray crystallographic data for (P)-3a. This material is available
free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
wuerthner@chemie.uni-wuerzburg.de
’ ACKNOWLEDGMENT
We acknowledge financial support from the DFG (GRK 1221).
’ REFERENCES
(1) (a) Kramer, R.; Lehn, J. M.; Marquisrigault, A. Proc. Natl. Acad.
Sci. U.S.A. 1993, 90, 5394–5398. (b) Wu, A. X.; Isaacs, L. J. Am. Chem.
Soc. 2003, 125, 4831–4835. (c) Mukhopadhyay, P.; Wu, A. X.; Isaacs, L.
J. Org. Chem. 2004, 69, 6157–6164. (d) Shivanyuk, A.; Rebek, J., Jr. J. Am.
Chem. Soc. 2002, 124, 12074–12075. (e) Taylor, P. N.; Anderson, H. L.
J. Am. Chem. Soc. 1999, 121, 11538–11545. (f) Jiang, W.; Winkler,
H. D. F.; Schalley, C. A. J. Am. Chem. Soc. 2008, 130, 13852–13853.
(2) (a) Corbett, P. T.; Sanders, J. K. M.; Otto, S. Angew. Chem., Int.
Ed. 2007, 46, 8858–8861. (b) Ludlow, R. F.; Otto, S. Chem. Soc. Rev.
2008, 37, 101–108. (c) Peyralans, J. J. P.; Otto, S. Curr. Opin. Chem. Biol.
2009, 13, 705–713. (d) Nitschke, J. R. Nature 2009, 462, 736–738. (e)
Wagner, N.; Ashkenasy, G. J. Chem. Phys. 2009, 16907 (1).
(16) (a) Wang, W.; Shaller, A. D.; Li, A. D. Q. J. Am. Chem. Soc. 2008,
130, 8271–8279. (b) Shaller, A. D.; Wang, W.; Gan, H. Y.; Li, A. D. Q.
Angew. Chem., Int. Ed. 2008, 47, 7705–7709.
(17) (a) Thoma, J. A.; Koshland, D. E., Jr. J. Am. Chem. Soc. 1960,
82, 3329–3333. (b) Koshland, D. E. Angew. Chem., Int. Ed. 1995,
33, 2375–2378.
(18) (a) James, L. C.; Roversi, P.; Tawfik, D. S. Science 2003,
299, 1362–1367. (b) Martin Schmeing, T.; Huang, K. S.; Strobel,
S. A.; Steitz, T. A. Nature 2005, 438, 520–524. (c) Ringe, D.; Petsko,
G. A. Science 2008, 320, 1428–1429.
(3) (a) K€uhnle, A.; Linderoth, T. R.; Hammer, B.; Besenbacher, F.
Nature 2002, 415, 891–893. (b) Huang, W. H.; Zavalij, P. Y.; Isaacs, L.
Angew. Chem., Int. Ed. 2007, 46, 7425–7427. (c) Zehnacker, A.; Suhm,
M. A. Angew. Chem., Int. Ed. 2008, 47, 6970–6992.
(19) (a) Kasai, K.; Aoyagi, M.; Fujita, M. J. Am. Chem. Soc. 2000,
122, 2140–2141. (b) Gasparrini, F.; Pierini, M.; Villani, C.; Filippi, A.;
Speranza, M. J. Am. Chem. Soc. 2007, 130, 522–534. (c) Filby, M. H.;
Dickson, S. J.; Zaccheroni, N.; Prodi, L.; Bonacchi, S.; Montalti, M.;
Paterson, M. J.; Humphries, T. D.; Chiorboli, C.; Steed, J. W. J. Am.
Chem. Soc. 2008, 130, 4105–4113. (d) Hiraoka, S.; Harano, K.;
Nakamura, T.; Shiro, M.; Shionoya, M. Angew. Chem., Int. Ed. 2009,
48, 7006–7009. (e) Mꢀenand, M.; Leroy, A.; Marrot, J.; Luhmer, M.;
Jabin, I. Angew. Chem., Int. Ed. 2009, 48, 5509–5512. (f) Han, J.-M.;
Pan, J.-L.; Lei, T.; Liu, C.; Pei, J. Chem.—Eur. J. 2010, 16,
13850–13861.
(4) (a) Pasteur, L. C. R. Hebd. Seanc. Acad. Sci. 1848, 26, 538. (b)
Kondepudi, D. K.; Kaufman, R. J.; Singh, N. Science 1990, 250, 975–976.
(c) Crusats, J.; Veintemillas-Verdaguer, S.; Ribꢀo, J. M. Chem.—Eur. J.
2006, 12, 7776–7781. (d) Fletcher, S. P.; Jagt, R. B. C.; Feringa, B. L.
Chem. Commun. 2007, 2578–2580. (e) Blackmond, D. G. Chem.—Eur. J.
2007, 13, 3290–3295. (f) McBride, J. M.; Tully, J. C. Nature 2008,
452, 161–162.
(5) Safont-Sempere, M. M.; Osswald, P.; Radacki, K.; W€urthner, F.
Chem.—Eur. J. 2010, 16, 9366–9373.
(6) W€urthner, F. Chem. Commun. 2004, 1564–1579.
(7) (a) Zhan, X. W.; Facchetti, A.; Barlow, S.; Marks, T. J.; Ratner,
M. A.; Wasielewski, M. R.; Marder, S. R. Adv. Mater. 2011, 23, 268–284.
(b) Letizia, J. A.; Rivnay, J.; Facchetti, A.; Ratner, M. A.; Marks, T. J. Adv.
(20) Marquis, D.; Desvergne, J. P.; Bouaslaurent, H. J. Org. Chem.
1995, 60, 7984–7996.
9590
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591

Journal of the American Chemical Society
ARTICLE
(21) (a) Hwang, T. L.; Shaka, A. J. J. Am. Chem. Soc. 1992,
114, 3157–3159. (b) Hwang, T. L.; Kadkhodaei, M.; Mohebbi, A.;
Shaka, A. J. Magn. Reson. Chem. 1992, 30, 24–34.
(22) Sheldrick, G. Acta Crystallogr. 2008, A64, 112–122.
(23) (a) Gvishi, R.; Reisfeld, R.; Burshtein, Z. Chem. Phys. Lett. 1993,
213, 338–344. (b) Sens, R.; Drexhage, K. H. J. Lumin. 1981, 24ꢀ5,
709–712.
Lemmetyinen, H.; Ihalainen, P.; Peltonen, J.; Dehm, V.; W€urthner, F.
Langmuir 2009, 26, 6630–6637.
(41) Yoo, H.; Yang, J.; Yousef, A.; Wasielewski, M. R.; Kim, D. J. Am.
Chem. Soc. 2010, 132, 3939–3944.
(42) (a) Connors, K. A. Binding Constants: The Measurement of
Molecular Complex Stability; Wiley: New York, 1987. (b) Lagona, J.;
Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L. Angew. Chem., Int. Ed.
2005, 44, 4844–4870.
(24) Ahlrichs, R.; B€ar, M.; H€aser, M.; Horn, H.; K€ohmel, C. Chem.
Phys. Lett. 1989, 162, 165.
(43) Wortmann, R.; R€osch, U.; Redi-Abshiro, M.; W€urthner, F.
(25) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100. Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988,
37, 785–789. (b) Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995,
102, 346–354. (c) Sch€afer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys.
1994, 100, 5829–5835. (d) Eichkorn, K.; Weigend, F.; Treutler, O.;
Ahlrichs, R. Theor. Chim. Acta 1997, 97, 119–124.
Angew. Chem., Int. Ed. 2003, 42, 2080–2083.
(44) Zheng, Y. R.; Yang, H. B.; Ghosh, K.; Zhao, L.; Stang, P. J.
Chem.—Eur. J. 2009, 15, 7203–7214.
(45) (a) Amemiya, R.; Yamaguchi, M. Org. Biomol. Chem. 2008,
6, 26–35. (b) Trzaska, S. T.; Hsu, H. F.; Swager, T. M. J. Am. Chem. Soc.
1999, 121, 4518–4519.
(26) (a) Grimme, S. J. Comput. Chem. 2004, 25, 1463–1473.
See also(b) Piacenza, M.; Grimme, S. ChemPhysChem 2005, 6, 1554.
(c) Antony, J.; Grimme, S. Phys. Chem. Chem. Phys. 2006, 8, 5287–5297.
(d) Grimme, S.; Antony, J.; Schwabe, T.; M€uck-Lichtenfeld, C. Org.
Biomol. Chem. 2007, 5, 741–745.
(27) (a) Fink, R. F.; Seibt, J.; Engel, V.; Renz, M.; Kaupp, M.;
Lochbrunner, S.; Zhao, H.-M.; Pfister, J.; W€urthner, F.; Engels, B. J. Am.
Chem. Soc. 2008, 130, 12858–12859. (b) Zhao, H.-M.; Pfister, J.; Settels,
V.; Renz, M.; Kaupp, M.; Dehm, V. C.; W€urthner, F.; Fink, R. F.; Engels,
B. J. Am. Chem. Soc. 2009, 131, 15660–15668.
(28) Marquis, D.; Desvergne, J. P.; Bouaslaurent, H. J. Org. Chem.
1995, 60, 7984–7996.
(29) Dijkstra, G.; Kruizinga, W. H.; Kellogg, R. M. J. Org. Chem.
1987, 52, 4230–4234.
(30) Kaiser, H.; Lindner, J.; Langhals, H. Chem. Ber. 1991,
124, 529–535.
(31) For the details on the determination of absolute configurations
of similar PBI marcocycles see ref. 12b.
(32) Fron, E.; Schweitzer, G.; Osswald, P.; W€urthner, F.; Marsal, P.;
Beljonne, D.; M€ullen, K.; De Schryver, F. C.; Van der Auweraer, M.
Photochem. Photobiol. Sci. 2008, 7, 1509–1521.
(33) (a) Klebe, G.; Graser, F.; Hadicke, E.; Berndt, J. Acta Crystal-
logr., Sect. B: Struct. Sci 1989, 45, 69–77. (b) Hino, K.; Mizuguchi, J. Acta
Crystallogr., Sect. E: Struct. Rep. Online 2005, 61, O672–O674.
(34) (a) Bergman, S. D.; Kol, M. Inorg. Chem. 2005, 44, 1647–1654.
(b) Gut, D.; Rudi, A.; Kopilov, J.; Goldberg, I.; Kol, M. J. Am. Chem. Soc.
2002, 124, 5449–5456. (c) Giordano, C.; Restelli, A.; Villa, M.;
Annunziata, R. J. Org. Chem. 1991, 56, 2270–2272. (d) Luchinat, C.;
Roelens, S. J. Am. Chem. Soc. 1986, 108, 4873–4878. (e) Dobashi, A.;
Saito, N.; Motoyama, Y.; Hara, S. J. Am. Chem. Soc. 1986, 108, 307–308.
(f) Harger, M. J. P. J. Chem. Soc., Perkin Trans. 2 1977, 1882–1887.
(35) As COSY experiments are much more sensitive than ROESY
ones, if the cross peak signal arises from an intramolecular coupling of
the protons, it must be present in both ROESY and COSY spectra. As it
is not the case, this cross peak must be originated intermolecularly.
(36) Such studies are not possible for macrocycles 3aꢀc due to poor
solubility at the required concentrations in nonpolar solvents.
(37) Chen, Z.; Baumeister, U.; Tschierske, C.; W€urthner, F. Chem.—
Eur. J. 2007, 13, 450–465.
(38) (a) Berova, N.; Nakanishi, K. Circular Dichroism: Principles and
Applications; Wiley-VCH: New York, 2000. (b) Berova, N.; Di Bari, L.;
Pescitelli, G. Chem. Soc. Rev. 2007, 36, 914–931. (c) Buss, V.; Reichardt,
C. Chem. Commun. 1992, 1636–1638.
(39) (a) Rybtchinski, B.; Sinks, L. E.; Wasielewski, M. R. J. Phys.
Chem. A 2004, 108, 7497–7505. (b) Ahrens, M. J.; Sinks, L. E.;
Rybtchinski, B.; Liu, W. H.; Jones, B. A.; Giaimo, J. M.; Gusev, A. V.;
Goshe, A. J.; Tiede, D. M.; Wasielewski, M. R. J. Am. Chem. Soc. 2004,
126, 8284–8294. (c) van der Boom, T.; Hayes, R. T.; Zhao, Y. Y.;
Bushard, P. J.; Weiss, E. A.; Wasielewski, M. R. J. Am. Chem. Soc. 2002,
124, 9582–9590. (d) Giaimo, J. M.; Gusev, A. V.; Wasielewski, M. R.
J. Am. Chem. Soc. 2002, 124, 8530–8531.
(40) (a) W€urthner, F.; Chen, Z.; Dehm, V.; Stepanenko, V. Chem.
Commun. 2006, 1188–1190. (b) Tolkki, A.; Vuorimaa, E.; Chukharev, V.;
9591
dx.doi.org/10.1021/ja202696d |J. Am. Chem. Soc. 2011, 133, 9580–9591