Full Paper
equality of left- and right-handed arrangements explains the
absence of a corresponding CD signal.
To comprehend the supramolecular organization of the pure
enantiomers, which differs from that of the meso form by
forming uniform twisted ribbons with a defined handedness,
we again started the minimization with a linear arrangement
of molecular stacks (see the Supporting Information), but now
with the attachment of R,R-configured head groups. Thus,
a head-group starting conformation was chosen that facilitated
the formation of intermolecular hydrogen bonds between OH
groups of neighboring molecules; this is only slightly different
from an all-trans conformation. Interestingly, in this case,
energy minimization led to only a slight twist of the strand,
which was in contrast to the strong twist obtained for the
meso form (Figure 8, right top). This result should be traced
back to the stabilizing effect of the additional intermolecular
OH hydrogen bonds (Figure 9) between hydroxyl groups of
the head-group moieties. Such an almost linear strand should
be prone to rather planar (two-dimensional) growth in the
manner of ribbons instead of highly curved cylinders. The mini-
mal bend of the strand, however, can explain the overall twist
and is thus an expression of the molecular chirality of the ob-
served ribbons.
Figure 8. Left: Top: A strongly twisted ribbon results from energy minimiza-
tion of a 26-molecule stack of R,S-[G1]16. A preferred direction of handed-
ness is not defined. The colors of the inserted head group moieties corre-
spond to the dendron (red), aromatic platform (blue), and amide linker
(green). Bottom: Model of a cylindrical assembly consisting of seven helical
molecular strands with meso head groups (R,S). The hydrophobic core of
such a cylinder (the diameter was taken from TEM data) is sized to accom-
modate the hexadecyl tails of the participating molecules in the all-trans bi-
layer conformation. For clarity, only the head groups and aromatic platforms
that contribute high density in the TEM data were used in this model. The
overall sizes and density distributions are in accordance with the dimension
of the molecular bilayer. Right: Top: Energy minimization of a 26-molecule
stack of R,R-[G1]16 results in a only slightly twisted linear stack. Bottom: The
lateral packing and bilayered arrangement (head groups: yellow, hydropho-
bic chains: gray) results in the formation of twisted ribbons in agreement
with the TEM data. The handedness of the twist is determined by the chirali-
ty of the head group.
of a chiral molecular component, however, should allow for
both equivalent designs of left- and right-handed twists.
In the experimental data, however, instead of highly twisted
helical fibers, we observe uniform and smooth cylindrical struc-
tures for the meso compound with a diameter in the order of
a molecular bilayer and a bilayer density profile. How can we
explain these differences?
We have to assume that, beyond the twisted arrangement
of individual chains observed in the simulation, the additional
lateral packing of molecules is an important factor to explain
the formation of closed cylindrical assemblies. We will see that
this lateral growth phenomenon plays a similar role for chiral
enantiomers, but with far less influence on the ultrastructural
twist.
Figure 9. Schematic representation of the G1 head groups of the pure R,R
enantiomer and possible inter- (red) and intramolecular (blue) OH hydrogen-
bonding motifs detected by energy minimization calculations (green tubes
and red lines depict the orientation of the amide hydrogen bonds and aro-
matic platform, respectively). It is easily comprehensible that such kinds of
cross-linked hydrogen bond patterns might impede the primary twist of the
molecular strands, and thus, may lead to the formation of only slightly twist-
ed ribbons.
It can be shown that a smooth cylindrical arrangement can
be obtained if seven helically twisted strands are arranged in
a shifted fashion (Figure 8, bottom left). Taking the diameter of
the cylinders from the TEM data, we could construct a cylindri-
cal model strand with molecular bilayer dimensions that corre-
lated well with those of the observed aggregates. By consider-
ing the dimensions of the less electron dense inner core
volume of the cylinder, we could roughly estimate that the
hexadecyl chains of seven such helices would perfectly fill the
interior of its volume. It has to be noted that the resolution in
the data is clearly too low to visualize individual helices within
the assembly structure. Molecular dimensions and structural
parameters, however, leave little space for an alternative mo-
lecular arrangement. It is reasonable to assume that the high
edge energy of an individual strand is compensated for by the
cyclical side-by-side arrangement of helices. The formation of
amide hydrogen-bonded molecule strands combined with the
simultaneous growth of several interlacing strands would ex-
plain the stiffness of the fibers. Although the individual molec-
ular ultrastructure is chiral (=twisted ribbons), the statistical
Looking closer at the resulting R,R strand, we found a repeat-
ing hydrogen-bonding motif (Figure 9) along the strand.
Herein, the intermolecular hydrogen bonds between the 5-
and 5’-OH groups (Figure 9, red) seem to be responsible for
the ability of the R,R- and S,S-configured head groups to defy
the original twisting tendency of the hydrophobic molecule
parts, while the intramolecular hydrogen bonds might stabilize
their respective conformations. In this pattern, the 6- and 6’-
OH groups establish lateral interactions with neighboring
strands to eventually form elongated twisted ribbons (Figure 8,
right bottom).
Conclusion
Herein, we have designed and synthesized nonionic chiral (R,R
and S,S) and achiral (R,S) G1 dendritic amphiphiles with diaro-
matic spacer coupled to single hydrophobic hexadecyl chains
Chem. Eur. J. 2016, 22, 5629 – 5636
5635
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim