Long-range ordering of highly charged self-assembled nanofilaments

Article abstract of DOI:10.1021/ja5082519

Charged nanoscale filaments are well-known in natural systems such as filamentous viruses and the cellular cytoskeleton. The unique properties of these structures have inspired the design of self-assembled nanofibers for applications in regenerative medicine, drug delivery, and catalysis, among others. We report here on an amphiphile of completely different chemistry based on azobenzene and a quaternary ammonium bromide headgroup that self-assembles into highly charged nanofibers in water and orders into two-dimensional crystals. Interestingly small-angle X-ray scattering (SAXS) shows that these fibers of 5.6 nm cross-sectional diameter order into crystalline arrays with remarkably large interfiber spacings of up to 130 nm. Solution concentration and temperature can be adjusted to control the interfiber spacings, and addition of salt destroyed the crystal packing indicating the electrostatic repulsions are necessary for the observed ordering. Our findings here demonstrate the universal nature of this phenomenon in systems of highly charged nanoscale filaments.

Full text of DOI:10.1021/ja5082519

Communication
pubs.acs.org/JACS
Long-Range Ordering of Highly Charged Self-Assembled
Nanofilaments
Liam C. Palmer,†,⊥ Cheuk-Yui Leung,‡,⊥ Sumit Kewalramani,§,⊥ Rohan Kumthekar,†
Christina J. Newcomb,§ Monica Olvera de la Cruz,†,‡,§ Michael J. Bedzyk,*,‡,§ and Samuel I. Stupp*,†,§,∥
‡
§
†Departments of Chemistry, Physics and Astronomy, and Materials Science and Engineering, Northwestern University, Evanston,
Illinois 60208, United States
∥Department of Medicine and Simpson Querrey Institute for BioNanotechnology, Northwestern University, Chicago, Illinois 60611,
United States
S
* Supporting Information
attractive interactions between neighboring nanofibers.12,13
ABSTRACT: Charged nanoscale filaments are well-
known in natural systems such as filamentous viruses
and the cellular cytoskeleton. The unique properties of
these structures have inspired the design of self-assembled
nanofibers for applications in regenerative medicine, drug
delivery, and catalysis, among others. We report here on an
amphiphile of completely different chemistry based on
azobenzene and a quaternary ammonium bromide head-
group that self-assembles into highly charged nanofibers in
water and orders into two-dimensional crystals. Interest-
ingly small-angle X-ray scattering (SAXS) shows that these
fibers of 5.6 nm cross-sectional diameter order into
crystalline arrays with remarkably large interfiber spacings
of up to 130 nm. Solution concentration and temperature
can be adjusted to control the interfiber spacings, and
addition of salt destroyed the crystal packing indicating the
electrostatic repulsions are necessary for the observed
ordering. Our findings here demonstrate the universal
nature of this phenomenon in systems of highly charged
nanoscale filaments.
Added salts, particularly divalent cations, can also induce
crystallization by bridging adjacent fibers through exponentially
decaying attractive forces.9,14 In contrast to the filament
bundling driven by attractive forces, a recent study of peptide
amphiphile nanofibers at ∼20 mM concentration showed the
formation of a fibrillar network driven by repulsions between the
negatively charged fibers using small-angle X-ray scattering
(SAXS).15 Interestingly, X-ray irradiation was found to create
additional charges on the surface of these fibers, causing bundle
formation at even lower concentrations (5−10 mM). It was
proposed that the long nanofibers formed at the early stages of
self-assembly create a stable network, which templates the
elongation of short nanofibers within the growing bundle. The
repulsion between these highly charged nanofibers is balanced
by spatial confinement imposed by neighboring bundles within
the network.
Inspired by the anisotropic hexagonal packing of nanostruc-
tures observed for peptide amphiphile nanofibers, we designed
amphiphilic molecules with a permanent charge (nonionizable)
in the headgroup based on aromatic groups in place of the
hydrogen-bonding peptides. In water, these molecules self-
assemble into 1D nanofibers that are expected to have a very
high surface charge density. It has been previously argued that if
the fibers form networks, the strong electrostatic repulsion
between the confined fibers leads to highly spaced fibers
arranged into a 2D hexagonal crystal.15,16 We investigate here
the ability of the fibers to form networks driven by electrostatic
repulsion at low concentration (2−16 mM). These networks
are studied by solution SAXS as a function of concentration and
temperature.
etworks of one-dimensional (1D) cytoskeletal compo-
N
nents such as actin and microtubules are essential to
mediate important biological processes like cell division,
protein transport, and signaling.1,2 Self-assembled nanofibers3
have also attracted attention for their ability to mimic the
extracellular matrix and promote tissue regeneration. Recently,
a pathway was discovered to create monodomain hydrogels of
peptide amphiphile nanofibers that can align and guide cell
migration.4,5 These monodomain gels could also have
anisotropic mechanical and transport properties. Furthermore,
it may be possible to use aligned 1D nanostructures as
templates for inorganic semiconductor materials as pathways to
move charge carriers in transistors or photovoltaic devices.6
Advancing these important applications will require a deeper
understanding of the fundamental mechanisms and pathways to
orient 1D objects.7
The amphiphilic molecule studied here (1) is based on an
aromatic azobenzene group, with a hydrophobic tail of ten
carbons, and a permanent +1 charged headgroup (see Figure 1a
for molecular structure). The synthesis of this molecule with a
Br− counterion is described in detail in the Supporting
Information. In water, these molecules dissolve to form viscous,
yellow-orange solutions that are birefringent at or above 8 mM
(see Figures S2 and S3). Cryogenic transmission electron
microscopy (TEM) shows formation of high aspect ratio
It is known from filamentous viruses8−10 and self-assembling
cyclic peptides11 that like-charged nanofibers can form
crystalline bundles with wall-to-wall separation on the order
of their diameters. Counterintuitively, the crystallization of
these structures into bundles is driven by fluctuation-induced
Received: August 11, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja5082519 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Figure 1. (a) Molecular structure of amphiphile 1 and schematic
representation of self-assembled nanofiber. (b) Representative
cryogenic TEM images of nanofibers formed by self-assembly of
cation 1 (1 mM in water).
nanofibers in aqueous solution (Figure 1c). Despite the strong
electrostatic repulsions between the charged headgroups, the
self-assembled 1D nanostructure is likely stabilized by
attractions, such as van der Waals and π−π stacking interactions
or electrostatic screening by condensed Br− counterions.
Azobenzene amphiphiles in their trans form are known to
self-assemble17 and form liquid-crystalline phases18 at higher
concentration, suggesting that excluded volume interactions
characteristic of liquid crystals may also promote anisotropy
within a dense medium of nanofibers. Photoisomerization to
the cis configuration was not observed for these molecules
when assembled in water, suggesting that they are tightly
packed within the nanofiber structure.
Figure 2. (a) In situ SAXS data showing the background subtracted
scattered intensity versus the scattering vector q (log−log plot) for
amphiphile 1 in water as the solution is diluted from 16 to 2 mM. The
data sets are offset vertically for clarity. (b) The lattice constant of the
hexagonal lattice formed by the nanofibers versus the solution
concentration. (c) The variation of lattice constant with the volume
fraction (Vf) of amphiphile.
We used solution-phase SAXS19 to explore how these self-
assembled nanofibers ordered from 2 to 16 mM in water. The
processed SAXS data are shown in Figure 2a as a function of
the scattering vector q. For 4−16 mM solutions, the scattered
intensity profiles show sharp diffraction peaks in the 0.1−0.5
nm−1 q-range indicating long-range ordering in the assembly of
nanofibers. The positions of the peaks follow q/q* ratios of
1:√3:2:√7:3:√12 (where q* is the principal peak position),
corresponding to 2D hexagonal lattices. The center-to-center
spacing [lattice constant (l)] between the nanofibers varies
from l = 63 nm for c = 16 mM solution to l = 128 nm for c = 4
mM solution, in accordance with l ∝ c−1/2 (Figure 2b). The
exponent 1/2 is expected for a lattice of 1D objects that is
expanding in two dimensions. The domain size of the 2D lattice
can be estimated by the Scherrer equation to be on the order of
1 μm. The SAXS data also provide insights into the nanofiber
dimensions. Because the lattice is hexagonal, the lattice
parameter is related to the volume fraction (Vf) of 1 in
solution via
To understand the origin of the nanofiber packing, we use
the nanofiber radius and bulk density of amphiphile 1 to
estimate a surface charge density of about 0.22 C/m2 (which
corresponds to about +34e/nm). Charged filamentous viruses,
like tobacco mosaic virus (TMV), are known to pack with
liquid-like order in solution with a center-to-center spacing on
the order of 60 nm.20 The diameter of TMV is ∼18 nm, and
the magnitude of the surface charge density of TMV rods was
reported to be 0.04 C/m2,21 much lower than the value for our
nanofiber system. In that system, the lower linear charge
density for TMV results in a much smaller ratio of interfiber
separation to the particle diameter. Although the condensation
of bromide counterions may reduce the actual charge density in
our system,22 we expect the system is still highly charged.
Addition of even 2 mM NaCl or 1 mM CaCl2 or MgCl2 screens
the repulsion, and the hexagonal packing disappears (Figure S5
and S6). We conclude that the high linear charge density
contributes to strong repulsion among the nanofibers, enabling
formation of the observed hexagonally packed network with
exceptionally large d-spacings.
2π
f
l =
r
3V
(1)
where r is the radius of the nanofilament. Using the solid-state
density of 1.28 g/mL measured by gas pycnometry, the known
mass of the amphiphile in the solution and eq 1, we estimate a
nanofiber radius of 2.8 nm (Figure 1c), which is slightly less
than the fully extended molecular length of 3.1 nm, estimated
by molecular modeling. These observations imply that for 16
mM solutions, the interfiber spacing is ∼11 times greater than
the fiber diameter. In contrast to nanofiber radius, the scattered
intensity profile is not very sensitive to the persistence length L
of the nanofibers (Figures S4 and S6). At present, we can only
place a lower limit of L = 30 nm on the nanofiber persistence
length. (Figure S4 and accompanying text).
As we lower the concentration (c) of amphiphile 1 below 4
mM, the sharp diffraction peaks are replaced by broad intensity
modulations, indicating only short-range correlations among
the nanofibers at 2 mM (Figure 2a). It is worth noting that the
nanofiber morphology is maintained even at these low
concentrations (Figure 1c). The differences between high and
low concentrations can be more clearly seen in the radial
B
dx.doi.org/10.1021/ja5082519 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
distribution function (RDF) g(r) (Figure S7 and accompanying
text), which is a measure of the number density of nanofibers as
a function of radial distance from any given nanofiber.23 The
first maximum of the RDF for the 2 mM solution (Figure S7a)
gives the nearest-neighbor distance of ∼150 nm. The very
broad second maximum in g(r) coupled with the rapid decay of
g(r) indicates the loss of crystalline ordering. In contrast, for a
more concentrated sample with long-range order, sharp peaks
show up in the RDF (Figure S7c), and the amplitude of the
peaks remains strong over many oscillations. These results
indicate that above 2 mM the nanofibers are ordered in a
crystalline lattice; at lower concentrations, only short-range
correlations between the fibers are observed. The results of
sample dilution show that it is possible to control the ordering
and interfiber distances by varying concentration.
solutions are heated to 80 °C or higher, they maintain their
liquid-like order upon cooling, indicating an irreversible change
in the network structure. However, the similarity between the
SAXS intensity profiles at 50 and 80 °C suggests that the
nanofiber morphology is preserved at higher temperatures.
While the hexagonal packing can be observed up to 40 °C, the
network is actually stable up to 70 °C and can re-form the
hexagonal crystal packing with the same spacing upon cooling.
Qualitatively similar behavior was observed for sample
concentrations down to 4 mM. Optical microscopy at elevated
temperature does not show any birefringence (see Figure S8).
To understand if intermolecular interactions were respon-
sible for the changing line charge density with temperature, we
conducted variable-temperature 1H NMR spectroscopy of
1
amphiphile 1 in D2O (see Figure 3b). Up to 45 °C, the H
We examined the solutions as a function of temperature to
understand the thermal stability of the system. Changes in the
packing behavior were observed by heating a 16 mM solution
from 20 to 90 °C (see Figure 3a). From 20 to 40 °C, the
NMR spectrum shows only weak and very broad peaks, as
expected for a highly aggregated system. From 50 to 90 °C, the
proton resonances become sharper and more intense,
suggesting that the molecules have more degrees of freedom
within nanofiber. A similar transition temperature is observed
by variable-temperature absorbance spectroscopy (see Figure
S9). At higher temperatures, the lower packing density of the
molecules results in a decrease of the filament line charge
density. Therefore, the repulsion between the filaments in the
network is weakened, and the nanofibers can come closer as
suggested by the SAXS data.
The effects of solution concentration and temperature are
summarized in Figure 4. When the solution concentration is
Figure 4. SAXS-determined phase diagram showing the hexagonal
(open hexagons) and liquid-like ordering (filled circles) of nanofibers
formed by amphiphile 1 as a function of temperature and
concentration.
equal to (see Figure S10) or greater than 4 mM and the
temperature of the system is below the transition temperature
of 50 °C observed by 1H NMR and UV−vis, the highly charged
nanofibers organize into a hexagonal lattice. In this regime, the
nearest-neighbor distance between these nanofibers increases as
the solution is diluted. When the solution is heated above the
melting temperature or diluted to 2 mM, only short-range
correlations are observed between nanofibers by SAXS. For the
2 mM sample, the transition temperature is below the melting
temperature of molecule 1. Therefore, the collapse of the
hexagonal packing is due to dilution of the nanofibers rather
than changes on the molecular level within the nanofibers. It is
likely that at this diluted concentration, the electrostatic
repulsion between the charged fibers is barely sufficient to
maintain the crystal structure within the confinement volume.
Therefore, fluctuations of the fibers even at room temperature
Figure 3. (a) SAXS data showing the background subtracted scattered
intensity versus the scattering vector q (log−log plot) for amphiphile 1
in water as the solution temperature changes from 20 to 90 °C. The
data sets are offset vertically for clarity. (b) Variable-temperature H
NMR of 16 mM amphiphile 1 in D2O shows a transition at 50 °C.
1
hexagonal peak positions remain constant, i.e., the lattice
parameter is unchanged. When the temperature is increased to
50 °C, the hexagonal packing transforms into liquid-like
ordering of nanofibers, as indicated by the broad peaks in the
SAXS data. As we heat the solution further, this correlation
peak position shifts to higher q, which indicates a reduction in
the average spacing between the nanofibers. For samples heated
up to 70 °C, the change is reversible, and the hexagonal packing
is restored upon cooling. In contrast, if these nanofiber
C
dx.doi.org/10.1021/ja5082519 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
make the crystal unstable. In contrast, at high concentration,
the loss of crystalline ordering upon heating is connected to
decreased linear charge density caused by molecular level
changes within the nanofiber.
1048773), the International Institute for Nanotechnology
(IIN) and Northwestern University. The authors thank S.
Weigand for assistance with the X-ray scattering, M. Seniw for
graphical assistance, M. McClendon for help with rheology, R.
Korkosz in the Kanatzidis group for help with the gas
pycnometer, and J. Boekhoven, J. Ortony and Y. Velichko for
helpful discussions.
CONCLUSIONS
■
We have reported the crystallization of highly charged cationic
supramolecular nanofibers into a hexagonally packed lattice
with remarkably large interfiber spacings, more than 11 times
the fiber diameter. The high charge density required for this
packing distance is made possible by close association of
molecules within the nanofiber and the permanent +1 charge
on each amphiphile. The strong electrostatic repulsions
between molecules in the fibers can be counterbalanced by
strong intermolecular attractions among aromatic groups and
hydrophobic tails in the covalent structure of molecules. The
packing structure and distance between the nanofibers can be
controlled by either temperature or concentration. The
dimensions of these crystalline lattices of supramolecular
filaments could allow the templated growth of hybrid
organic−inorganic hybrid materials on the appropriate length
scale to separate excitons for energy applications.24,25
REFERENCES
■
(1) Needleman, D. J.; Ojeda-Lopez, M. A.; Raviv, U.; Miller, H. P.;
Wilson, L.; Safinya, C. R. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 16099.
(2) Wang, N.; Butler, J. P.; Ingber, D. E. Science 1993, 260, 1124.
(3) Palmer, L. C.; Stupp, S. I. Acc. Chem. Res. 2008, 41, 1674.
(4) Zhang, S. M.; Greenfield, M. A.; Mata, A.; Palmer, L. C.; Bitton,
R.; Mantei, J. R.; Aparicio, C.; Olvera de la Cruz, M.; Stupp, S. I. Nat.
Mater. 2010, 9, 594.
(5) Berns, E. J.; Sur, S.; Pan, L.; Goldberger, J. E.; Suresh, S.; Zhang,
S.; Kessler, J. A.; Stupp, S. I. Biomaterials 2014, 35, 185.
(6) Dang, X.; Yi, H.; Ham, M.-H.; Qi, J.; Yun, D. S.; Ladewski, R.;
Strano, M. S.; Hammond, P. T.; Belcher, A. M. Nat. Nanotechnol.
2011, 6, 377.
(7) Aida, T.; Meijer, E. W.; Stupp, S. I. Science 2012, 335, 813.
(8) Parsegian, V. A.; Brenner, S. L. Nature 1976, 259, 632.
(9) Li, T.; Winans, R. E.; Lee, B. Langmuir 2011, 27, 10929.
(10) Li, T.; Zan, X. J.; Winans, R. E.; Wang, Q.; Lee, B. Angew. Chem.,
Int. Ed. 2013, 52, 6638.
ASSOCIATED CONTENT
* Supporting Information
■
S
́
(11) Valery, C.; Paternostre, M.; Robert, B.; Gulik-Krzywicki, T.;
Detailed synthesis, spectroscopic characterization, and X-ray
scattering curves. This material is available free of charge via the
Internet at http://pubs.acs.org.
Narayanan, T.; Dedieu, J.-C.; Keller, G.; Torres, M.-L.; Cherif-Cheikh,
R.; Calvo, P.; Artzner, F. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10258.
(12) Oosawa, F. Biopolymers 1968, 6, 1633.
(13) Safinya, C. R.; Li, Y. Science 2010, 327, 529.
(14) Solis, F. J.; Olvera de La Cruz, M. Phys. Rev. E 1999, 60, 4496.
(15) Cui, H.; Pashuck, E. T.; Velichko, Y. S.; Weigand, S. J.;
Cheetham, A. G.; Newcomb, C. J.; Stupp, S. I. Science 2010, 327, 555.
(16) Yao, Z. W.; Olvera de la Cruz, M. Phys. Rev. E 2013, 87, 042605.
(17) Kunitake, T.; Okahata, Y.; Shimomura, M.; Yasunami, S.;
Takarabe, K. J. Am. Chem. Soc. 1981, 103, 5401.
AUTHOR INFORMATION
■
Corresponding Authors
bedzyk@northwestern.edu
s-stupp@northwestern.edu
Author Contributions
⊥These authors contributed equally.
(18) Ikeda, T.; Tsutsumi, O. Science 1995, 268, 1873.
(19) Svergun, D. I.; Koch, M. H. J. Rep. Prog. Phys. 2003, 66, 1735.
(20) Maier, E.; Krause, R.; Deggelmann, M.; Hagenbuchle, M.;
Weber, R.; Fraden, S. Macromolecules 1992, 25, 1125.
Notes
The authors declare no competing financial interest.
(21) Ermolina, I.; Morgan, H.; Green, N. G.; Milner, J. J.; Feldman,
Y. Biochim. Biophys. Acta, Gen. Subj. 2003, 1622, 57.
(22) Gelbart, W.; Bruinsma, R.; Pincus, P.; Parsegian, V. Phys. Today
2000, 53, 38.
(23) Spaar, A.; Salditt, T. Biophys. J. 2003, 85, 1576.
(24) Sofos, M.; Stone, D. A.; Goswami, D. K.; Okasinski, J. S.; Jin, H.;
Bedzyk, M. J.; Stupp, S. I. J. Phys. Chem. C 2008, 112, 2881.
(25) Briseno, A. L.; Holcombe, T. W.; Boukai, A. I.; Garnett, E. C.;
ACKNOWLEDGMENTS
■
Materials design and synthesis was supported by the Non-
equilibrium Energy Research Center (NERC) at Northwestern
University, funded by the US Department of Energy (DOE),
Office of Basic Energy Sciences under award no. DE-
SC0000989. Characterization was supported by the U.S.
Department of Energy, Office of Basic Energy Sciences under
award no. (DE-FG02-00ER45810). The X-ray scattering
experiments and analysis were supported by DOE-BES DE-
FG02-08ER46539. S.K. was funded by AFOSR (FA9550-11-1-
0275). SAXS experiments were performed at the DuPont-
Northwestern-Dow Collaborative Access Team (DND-CAT)
located at Sector 5 of the Advanced Photon Source (APS).
DND-CAT is supported by E.I. DuPont de Nemours & Co.,
The Dow Chemical Company, and Northwestern University.
Use of the APS, an Office of Science User Facility operated for
the U.S. Department of Energy (DOE) Office of Science by
Argonne National Laboratory, was supported by the U.S. DOE
under contract no. DE-AC02-06CH11357. This work made use
of the Biological Imaging Facility (TEM) and the Keck
Biophysics Facility (optical spectroscopy). This work also relied
on NMR and mass spectrometry instrumentations in the
Integrated Molecular Structure Education and Research Center
(IMSERC) which were funded by the NSF (NSF CHE-
́
Shelton, S. W.; Frechet, J. J. M.; Yang, P. Nano Lett. 2010, 10, 334.
D
dx.doi.org/10.1021/ja5082519 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX