Full Paper
sion spectra upon cooling from 908C to RT, which supports
the formation of an H-type supramolecular polymer.
ure S20 in the Supporting Information). By monitoring chemi-
cal shifts of proton HA versus concentration, a sigmoidal aggre-
gation curve was observed and fitted to the isodesmic model
(Figure S21).
In order to understand the specific process in more detail,
cooling curves of 2 (cooling rate=1 Kminꢀ1) extracted from
temperature-dependent UV/Vis studies in MCH at different
concentrations were measured, as shown in Figure 2 f. Each
cooling curve could be perfectly fitted to the cooperative nu-
cleation–elongation model.[22] The thermodynamic parameters
are shown in Table 1 and Table S1 in the Supporting Informa-
tion. The high binding constants of the supramolecular elonga-
tion step (up to 2.0ꢃ105 mꢀ1), which is around three times
higher than the K value of 1, along with the low s values (be-
tween 4.8ꢃ10ꢀ3 and 5.7ꢃ10ꢀ3) correspond to a cooperative
self-aggregation process. By comparing the self-assembly be-
havior as well as the supramolecular polymerization mecha-
nism, one can notice significant differences between 1 and 2.
In order to understand this dissimilarity, the reason for it has to
lie on the molecular level, since the linking group is the only
difference between both molecules. An amide group is, in con-
trast to the ester linker, both a hydrogen bond acceptor and
donor, which provides an additional intermolecular force. The
UV/Vis spectra of 2 are reminiscent of the H-type aggregates
reported by the Wꢀrthner group[23] and us[18] in terms of
a blue-shifted transition as well as a red-shifted shoulder rela-
tive to the monomer spectrum. The blue-shifted peak at
520 nm indicates a face-to-face H-type stacking whereas the
red-shifted shoulder at 620 nm correlates with a rotational dis-
placement of the BODIPY chromophores in the stack.[21] Thus,
to provide intermolecular hydrogen bonds between the amide
groups, the molecules undergo a rotational face-to-face stack-
ing on the account of the grade of the p–p stacking’s overlap.
In sharp contrast ester-containing BODIPY 1 shows also a blue-
shift in the spectra but no red-shifted shoulder, which indicates
that the rotational displacement becomes less prominent for
this system. Ultimately, the additional intermolecular force as-
sisting hydrogen bonding tunes the isodesmic to a cooperative
H-type self-aggregation process.
We next attempted to analyze the influence of temperature
on the chemical shifts of 2 in [D14]MCH, as it was previously
observed that temperature-dependent spectroscopic studies
worked well for this system. However, the stronger tendency
of 2 to self-assemble compared to 1 leads to a severe broaden-
ing of the proton NMR signals even at elevated temperatures.
Thus, the selection of a more appropriate solvent mixture was
required in order to monitor the self-assembly of 2 through
NMR spectroscopy experiments. CD2Cl2 was the first solvent of
choice, however, the good solvation of the units of 2 in this
solvent, results in minor changes in the chemical shifts upon
increasing concentration. In order to find the right conditions
providing the balance between both a strong tendency to ag-
gregate and a sufficient solubility, we selected CCl4 as a “poor
solvent” in combination with CD2Cl2. Previous UV/Vis and fluo-
rescence results (Figure 1c and d) had suggested the initial
stages of a self-assembly process in this solvent. However, this
was not as strong as in hydrocarbon solvents. By mixing
CD2Cl2 and CCl4 in different ratios, we found that the solvent
mixture CCl4/CD2Cl2 (9:1, v/v) was the most appropriate one in
To investigate the participating intermolecular forces during
the aggregation process in more detail, concentration-depen-
1
dent H NMR measurements were recorded. A concentration-
dependent shift of the proton signals in the NMR spectrum in-
dicates participation of the very same during the supramolec-
ular polymerization process, caused by a change of the spatial
environment of the particular protons. For instance, the shield-
ing of the aromatic protons is commonly attributed to the p–p
stacking of the aromatic planes whereas hydrogen-bond inter-
actions usually cause a downfield shift of the involved proton
NMR signals.[24] For the target systems 1 and 2, concentration-
dependent 1H NMR measurements were recorded (Figure 3).
Upon increasing the concentration of 1 ([D14]MCH, Fig-
ure 3a), both proton signals of the aromatic plane (HA, HB and
HC) and those of the methyl groups of the BODIPY core (HD, HE
and HF) undergo upfield shifts, thus supporting p–p stacking
of the molecules (for proton labeling see Scheme 1; capital let-
ters 1, small letters 2). A similar shielding effect upon cooling
a solution of 1 from 343 K to RT was observed by tempera-
1
Figure 3. Concentration-dependent H NMR spectroscopy of: a) 1 (400 MHz,
1
ture-dependent H NMR experiments (17.5 mm, 600 MHz; Fig-
298 K, [D14]MCH), and b) 2 (500 MHz, 298 K, CCl4/CD2Cl2 9:1, v/v).
&
&
Chem. Eur. J. 2016, 22, 1 – 7
4
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!