Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
.
Angewandte
Communications
vesicles showed an appreciable kinetic stability as no signifi-
cant leakage of entrapped calcein was observed even after six
days (Figure S9).[12]
However, dynamic-light-scattering studies (for the raw
data, see Table S1–S4) estimated a larger particle size for
NDI-2 (average Dh = 180 nm) than for NDI-1 (Dh = 140 nm;
Figure 1e). Nevertheless, the most pronounced difference is
noticed for the z value of NDI-2, which exhibits a close to
neutral value of À10.0 Æ 2.0 mV, suggesting that the surface is
almost completely nonionic in contrast to the negatively
charged surface of NDI-1. The adsorption of the hydroxide
ions onto the otherwise non-ionic oligooxyethylene (OE)
surface is likely to contribute to such insignificantly negative
z values at basic pH values.[13] To further confirm and quantify
the difference in surface functionality, we conducted cationic
methyl viologen (MV) assisted quantitative precipitation
experiments similar to what is routinely done to probe lectin–
carbohydrate binding.[14] In fact, when an aqueous solution of
NDI-1 (pH 9) was added to MV, the solution instantaneously
turned turbid, and after some time, a precipitate formed. For
NDI-2, however, the solution remained clear for several days
in the presence of MV under identical conditions. After
removing the insoluble solid by centrifugation, the UV/Vis
spectra of the two supernatant solutions differed significantly
(Figure 2). For the NDI-2 solution, almost no change was
noticed in the absorbance spectrum (Figure 2b) whereas for
NDI-1, the intensity was reduced by approximately 80%
(Figure 2a), confirming the presence of effective interactions
of MV selectively with NDI-1 and not with NDI-2. When the
precipitate that was isolated by centrifugation was re-
dissolved in THF (Figure 2a), sharp absorption peaks con-
firmed the presence of NDI-1 in the precipitate. To further
push the extent of precipitation, another experiment was
carried out with FeCl3 as the Fe3+ ion has the ability to form
strong chelate complexes with negatively charged ligands.[15]
The addition of an aqueous FeCl3 solution to a solution of
NDI-1 or NDI-2 (0.5 mm, pH 9) resulted in very different
outcomes (Figure 2c,d). After removal of the precipitate, the
supernatant solution of NDI-1 indicated quantitative precip-
itation (Figure 2c) as its absorption intensity was reduced by
approximately 95% whereas no significant changes were
observed for the NDI-2 solution (Figure 2d), confirming that
Fe3+ selectively interacts with a vesicular assembly of NDI-1.
In case of a 1:1 mixed stacking of NDI-1 and NDI-2 in
alternate fashion, interestingly, almost full precipitation was
observed (Figure S10) as for NDI-1 alone, suggesting that
mostly/fully one type of assembly with an anionic surface
charge is formed by the mixed assembly of NDI-1 and NDI-2.
The model depicted in Scheme 1 can rationalize the
experimental observation presented in the above discussion.
According to this model, the two most prominent common-
alities between the self-assembly of these twin amphiphiles
are: 1) Parallel alignment allows the hydrazides to remain on
the same side of the monolayer. 2) Owing to a preferred
direction of curvature of the monolayer, the hydrazides (and
consequently the functional groups attached to the same arm)
remain at the inner wall of the vesicle. The reason is obvious
as hydrazides, which are known[7] to exhibit strong intermo-
lecular hydrogen bonding, drive the unidirectional alignment
Figure 1. a,b) Structures of NDI-1 and NDI-2. c) Absorption-normal-
ized emission spectra of calcein in NDI-1 (black line) and NDI-2 (gray
line); dashed lines indicate the corresponding emission spectra of
calcein in water without NDI. d) Solvent-dependent (solid line: THF;
dashed line: water, pH 9.0) UV/Vis absorption spectra of NDI-
1 (black) and NDI-2 (gray); c=0.5 mm. e) DLS data of NDI-1 and
NDI-2 (c=0.5 mm).
filled space inside the vesicles. UV/Vis spectra suggest
pronounced solvent effects (Figure 1d); in THF, sharp
absorption bands (lmax = 376 and 383 nm) indicate mono-
meric species while in H2O (pH 9.0), a bathochromic shift of
approximately 7.0 nm with around 60% reduction in absorb-
ance and reversed intensities of the two absorption bands
(lmax = 376 and 383 nm) suggest p–p interactions.[10] Interest-
ingly, zeta potential (z) measurements revealed a highly
negative value of À45.0 Æ 3.0 mV, indicating a negatively
charged outer surface. To investigate the role of the location
of the hydrazide group in displaying the anionic phenoxides
on the outer surface, we compared the self-assembly of
NDI-1 with that of NDI-2 (Figure 1b), which has the same
structure as NDI-1 except for the position of the hydrazide
group, which is now located on the anionic arm. HRTEM
images (Figure S5) along with XRD (Figure S3) confirmed
the presence of monolayer vesicles (diameter ca. 140 nm; wall
thickness: 42 Æ 5 ꢀ) similar to those of NDI-1. Successful
calcein encapsulation (Figure 1c) ascertained the vesicular
structure. Pleasingly, an estimation of the amount of entrap-
ped calcein dye revealed encapsulation efficiencies of
approximately 13% and 10% for the NDI-1 and NDI-2
vesicles, respectively, which are in line with those calculated
for liposomes of comparable dimensions in earlier reports[11]
(see Figure S4 and a related discussion in the Supporting
Information). UV/Vis studies revealed identical p–p inter-
actions (Figure 1d). Variable-temperature UV/Vis studies
indicated a remarkably high thermal stability for the vesicular
assembly of these two p-amphiphiles as their spectral patterns
remained unaltered even at 908C (Figure S6). Concentration-
dependent UV/Vis studies (Figure S7, S8) measured a slightly
higher value for the critical aggregation concentration (CAC)
for NDI-2 (3.3 ꢂ 10À5 m) than for NDI-1 (2.7 ꢂ 10À5 m). Both
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 7
These are not the final page numbers!