Hierarchical self-assembly of di-, tri- and tetraphenylalanine peptides capped with two fluorenyl functionalities: From polymorphs to dendrites

Article abstract of DOI:10.1039/c6sm00337k

Homopeptides with 2, 3 and 4 phenylalanine (Phe) residues and capped with fluorenylmethoxycarbonyl and fluorenylmethyl esters at the N-terminus and C-terminus, respectively, have been synthesized to examine their self-assembly capabilities. Depending on t

Full text of DOI:10.1039/c6sm00337k

View Article Online
View Journal
Soft Matter
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: E. Mayans, G.
Ballano, J. Casanovas, L. J. del Valle, M. M. Pérez-Madrigal, F. Estrany, A. I. Jimenez, J. Puiggali, C.
Cativiela and C. Aleman, Soft Matter, 2016, DOI: 10.1039/C6SM00337K.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/softmatter

Page 1 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
Hierarchical Self-Assembly of Di-, Tri- and
Tetraphenylalanine Peptides Capped
with Two Fluorenyl Functionalities:
From Polymorphs to Dendrites
Enric Mayans,1,2 Gema Ballano,3 Jordi Casanovas,4 Luis J. del Valle,1,2
Maria M. Pérez-Madrigal,1,2 Francesc Estrany,2,5 Ana I. Jiménez,3 Jordi
Puiggalí,1,2,* Carlos Cativiela,3,* and Carlos Alemán1,2,*
1 Departament d’Enginyeria Química, ETSEIB, Universitat Politècnica de Catalunya,
Diagonal 647, Barcelona E-08028, Spain
2 Centre for Research in Nano-Engineering, Universitat Politècnica de Catalunya,
Campus Sud, Edifici C’, C/Pasqual i Vila s/n, Barcelona E-08028, Spain
3 Department of Organic Chemistry and Instituto de Síntesis Quimica y Catalisis
Homogenea (ISQCH), University of Zaragoza–CSIC, 50009 Zaragoza, Spain
4 Departament de Química, Escola Politècnica Superior, Universitat de Lleida, c/
Jaume II nº 69, Lleida E-25001, Spain
5 Departament d’Enginyeria Química, EUETIB, Universitat Politècnica de Catalunya,
Comte d’Urgell 187, 08036 Barcelona, Spain
* jordi.puiggali@upc.edu, cativiela@unizar.es and carlos.aleman@upc.edu
1

Soft Matter
Page 2 of 47
View Article Online
DOI: 10.1039/C6SM00337K
ABSTRACT
Homopeptides with 2, 3 and 4 phenylalanine (Phe) residues and capped with
fluorenylmethoxycarbonyl and fluorenylmethyl ester at the N- and C-terminus,
respectively, have been synthesized to examine their self-assembly capabilities.
Depending on the conditions, the di- and triphenylalanine derivatives self-organize into
a wide variety of stable polymorphic structures, which have been characterized: stacked
braids, doughnuts-like, bundled arrays of nanotubes, corkscrew-like and spherulitic
microstructures. These highly aromatic Phe-based peptides also form incipient branched
dendritic microstructures, even though they are highly unstable, making their
manipulation very difficult. In opposition, the tetraphenylalanine derivative
spontaneously self-assemble into stable dendritic microarchitectures made of branches
growing from nucleated primary frameworks. The fractal dimension of these
microstructures is 1.70, which evidences self-similarity and two-dimensional diffusion
controlled growth. DFT calculations at the M06L/6-31G(d) level have been carried out
on model -sheets since it is the most elementary building block of Phe-based peptide
polymorphs. Results indicate that the antiparallel -sheet is more stable than the parallel
one, the difference between them growing with the number of Phe residues. Thus, the
cooperative effects associated with the antiparallel disposition become more favorable
when the number of Phe residues increases from 2 to 4, while those of the parallel
disposition remained practically constant.
2

Page 3 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
INTRODUCTION
Fluorenylmethoxycarbonyl (Fmoc)-based short peptides are widely studied because
of their particular supramolecular assembly capabilities. Thus, the Fmoc moiety
provides strong aromatic interactions that drive the peptide self-assembly into
nanofibers or nanotubes.1-11 In peptide sequences containing aromatic residues, which
intrinsically form -stacking interactions, the role of such type of interactions become
predominant. A very illustrative example corresponds to the self-assembly of
diphenylalanine (FF), as a minimal sequence to form peptide nanostructures, which
organizes forming peptide nanotubes stabilized by a combination of hydrogen bonding
and repeated phenyl stacking interactions.12-14 In contrast, Fmoc-FF forms peptide
fibrils15 and very stable hydrogels9,16,17 that were thought to arise from the stacking
between Fmoc groups and between phenyl groups. The remarkable importance of
stacking interactions induced by the Fmoc group at the N-terminus was also illustrated
using a series of dipeptides and amino acids,18 such aromatic moiety acting as a
consistent facilitator of gelation in comparison to other simple hydrophobic groups,
such as tert-butoxycarbonyl.
More recently, research on the self-assembly of triphenylalanine (FFF) and Fmoc-
FFF also evidenced some important differences.19,20 More specifically, FFF and Fmoc-
FFF were found to self-assemble into solid fibrillary plate-like nanostructures19 (also
named “nanoplates”) and hydrogels,20 respectively. In all cases, - stacking
interactions between aromatic rings were found to play a decisive role in the formation
of such supramolecular aggregates. In a very recent study we examined the self-
assembly of tetraphenylalanine (FFFF) and Fmoc-FFFF, which had never been reported
before.21 FFFF molecules were found to assemble into tubes, exhibiting structural
imperfections in comparison to FF. Theoretical calculations suggested that these
3

Soft Matter
Page 4 of 47
View Article Online
DOI: 10.1039/C6SM00337K
structural defects in FFFF tubes are due to the fact that the increment in the
conformational flexibility is accompanied by a reduction in the number of restrictions
associated with hydrogen bonds. In opposition, Fmoc-FFFF organizes into a variety of
polymorphs depending on the experimental conditions, which included ultra-thin
nanoplates, fibrils and star-like submicrometric aggregates.21
The applications of self-assembled peptides in the biomedical (e.g. as cargo to target
delivery of drugs and genes, scaffolds in tissue engineering and regenerative
biomedicine, and biosensors) and nanotechnological (e.g. fabrication of composite
materials by controlled nucleation, electronic and magnetic nanonowires) fields have
been extensively reviewed in the last years.22-25 However, many of these applications are
focused on a well-defined morphology. The control exerted by the environmental
conditions in the morphology of a given system, regulating the apparition of different
polymorphs, can enhance such applications, even leading to multifunctional systems. In
this work we synthesize and study the self-assembly of phenylalanine-based peptides
capped with Fmoc and 9-fluorenylmethyl ester (OFm) as N- and C-terminal aromatic
components, respectively. More specifically, the results for a peptide series formed by
FF, FFF and FFFF have been systematically compared. Furthermore, intermolecular
interactions formed by these peptides, named Fmoc-FF-OFm, Fmoc-FFF-OFm and
Fmoc-FFFF-OFm (Scheme 1), have been also investigated using theoretical
calculations. It should be remarked that the interest of these systems lies not only in the
high concentration of aromatic groups but also in the complete elimination of the
normally free basic (N-terminus) and acidic (C-terminus) ends that are often important
for gelation.2,26 Accordingly, no hydrogel is a priori expected for these systems while
the formation of multiple supramolecular self-assembled organizations may be reached
through stacking interactions. It is worth noting that in a very recent study it was
4

Page 5 of 47
Soft Matter
View Article Online
reported a double Fmoc-functionalized low molecular weight peptide, whichDObIe: 1h0a.1v03e9d/C6SM00337K
very differently from the corresponding single Fmoc-functionalized analogue.27 More
specifically, Fmoc-Lys-Fmoc was found to form pH-controlled gels whereas single
Fmoc-Lys failed under similar experimental conditions.
METHODS
Materials. Boc and Fmoc-aminoacids were supplied by PolyPeptide group, N-[3-
(dimethylamino)-propyl]-N’-ethylcarbodiimide hydrochloride was a product from
Bachem and all other reagents for peptide synthesis were purchased from Sigma-
Aldrich.
Peptide synthesis and characterization.
The preparation of Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-FFFF-OFm peptides
was carried out following standard procedures of peptide synthesis in solution starting
from the corresponding F-derivative and using the Boc or Fmoc group as protection for
the amino moieties. A general procedure for the coupling reactions is given in Figure 1.
Melting points were determined on a Gallenkamp apparatus and are uncorrected. IR
spectra were registered on a Thermo Nicolet Avatar 360 FTIR spectrophotometer; max
1
13
is given for the main absorption bands. H and C NMR spectra were recorded on a
Bruker AV-400 or ARX-300 instrument at room temperature unless otherwise indicated
and using the residual solvent signal as the internal standard; chemical shifts (δ) are
expressed in ppm and coupling constants (J) in Hertz. Optical rotations were measured
on a JASCO P-1020 polarimeter. High-resolution mass spectra were obtained on a
Bruker Microtof-Q spectrometer.
Preparation of initial solutions of peptides. Organic solvents were purchased from
Sigma-Aldrich, Fisher Scientific and Scharlab. The peptide concentration in the
5

Soft Matter
Page 6 of 47
View Article Online
DOI: 10.1039/C6SM00337K
prepared solutions ranged from 0.05 to 5 mg/mL. Solutions or dispersions (25 or 100
μL) of the peptides were prepared from 4-5 mg/mL stocks. The solvents used to
dissolve the synthesized peptides were hexafluoroisopropanol (HFIP) and
dimethylformamide (DMF). Milli-Q water, methanol (MeOH), ethanol (EtOH),
isopropanol (iPrOH) or acetone was added as co-solvents to reduce the peptide
concentration and alters the polarity of the environment. More specifically, as usual for
Phe-based aromatic peptides, the three investigates compounds were soluble in HFIP
and DMF, while they were only partially soluble or insoluble in alcohols, water and
acetone. Accordingly, self-assembly studies were conducted using as solvents pure
HFIP and mixtures HFIP:alcohol (alcohol= MeOH, EtOH and iPrOH), H2O being added
to complete the series of mixtures. As the main objective of these co-solvents was to
increase the hydrophilicity of the environment and, for this purpose, the ratio
solvent:co-solvent was systematically varied from 4:1 to 1:99 (i.e. 4:1, 2:3, 1:4, 1:9,
1:19; 1:24, 1:49 and 1:99). In addition, as peptides were also soluble in DMF and
partially soluble in acetone, some trials were performed using pure DMF, DMF:acetone
and HFIP:acetone mixtures, even though in this case the number of explored ratios was
lower because the amount of observed microstructures was relatively infrequent with
respect to HFIP and HFIP:alcohol. Finally, 10 or 20 μL aliquots were placed on
microscope coverslips or glass slides (glass sample holders) and kept at room
temperature (25 ºC) or inside a cold chamber (4 ºC) until dryness. The humidity was
kept constant in both laboratories at 50%.
Optical microscopy. Morphological observations were performed using a Zeiss
Axioskop 40 microscope. Micrographs were taken with a Zeiss AxiosCam MRC5
digital camera.
6

Page 7 of 47
Soft Matter
View Article Online
Scanning electron microscopy (SEM). SEM studies were performed in a DFOoI:c1u0.s10s3e9d/C6SM00337K
Ion Beam Zeiss Neon 40 scanning electron microscope operating at 5 kV and equipped
with an EDX spectroscopy system. Samples were mounted on a double-side adhesive
carbon disc and sputter-coated with a thin layer of carbon to prevent sample charging
problems.
Atomic Force Microscopy (AFM). Topographic AFM images were obtained using
either a Dimension 3100 Nanoman AFM or a Multimode, both from Veeco (NanoScope
IV controller) under ambient conditions in tapping mode. AFM measurements were
performed on various parts of the morphologies, which produced reproducible images
similar to those displayed in this work. Scan window sizes ranged from 55 m2 to
8080 m2.
Fractal characterization of dendritic microstructures. The dendritic assembly
behavior of Fmoc-FFFF-OFm was observed by optical microscopy and AFM. The
fractal dimension of the dendritic morphology was determined by the fractal box-
counting method28 using ImageJ software version 1.50e. (version: 2.0.0-rc-43/1.50e,
Fiji package).29 For this analysis, AFM images of dendritic microstructures were
converted into 8-bit binary format images, which were covered by square box arrays.
The number of boxes occupied by the underlying dendritic morphology (N) and the side
length of boxes (L) were plotted in logarithmic scale to determine the fractal dimension
(FD) of the dendrites.
Theoretical calculations. Density Functional Theory (DFT) calculations were
performed using the Gaussian 09 computer package.30 The geometries of the different
investigated systems were fully optimized using the M06L31,32 functional, which was
developed by Zhao and Truhlar to account for dispersion, combined with the 6-31G(d)
basis set. No symmetry constraints were used in the geometry optimizations.
7

Soft Matter
Page 8 of 47
View Article Online
DOI: 10.1039/C6SM00337K
The interaction energy, Eint, for each complex formed by three peptide molecules
was computed as the difference between the energy of the complex and the sum of the
energies calculated for each of the three peptide molecule:
Eint = E(complex) – E(peptide 1) – E (peptide 2) – E(peptide 3)
(1)
The cooperative energy, Ecoop, for the -sheets formed by three strands was
estimated as the difference between Eint and the expected interaction energies (Eq. 2).
The expected interaction energy, Eint(E), was supplied as the sum of the DFT
interaction energies of all dimers contained in the complex (Eqs. 3-6). Accordingly,
Ecoop provides an evaluation of the many-body (non-additive) effects.
Ecoop = Eint – Eint(E)
(2)
(3)
(4)
(5)
(6)
Eint(E) = Eint (1-2) – Eint (2-3) – Eint (1-3)
Eint (1-2) = E(dimer 1-2) – E(peptide 1) – E(peptide 2)
Eint (2-3) = E(dimer 2-3) – E(peptide 2) – E(peptide 3)
Eint (1-3) = E(dimer 1-3) – E(peptide 1) – E(peptide 3)
Interaction and cooperative energies were corrected with the basis set superposition
error (BSSE) by mean of the standard counterpoise method.
FTIR spectroscopy. Infrared transmittance spectra were recorded with a Jasco FTIR
4100 Fourier Transform spectrometer in a 4000-650 cm-1 interval. An MKII Golden
Gate attenuated total reflection (ATR) accesory from Specac was used. The
measurements were taken using 4 cm-1 resolution and 1000 scans averaging.
8

Page 9 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
RESULTS AND DISCUSSION
Results presented in this work only correspond to the conditions in which stable
microstructures (i.e. microstructures that remained formed upon manipulation for
optical microscopy, SEM and AFM observations) were formed. As the hydrophobicity
and degree of solubility of the peptides change with the number of Phe residues, in
many cases different peptides require different solvent:co-solvent ratios to self-assemble
into the microstructures with similar morphologies. This feature is reflected in Table 1,
which summarizes and compares the conditions required by Fmoc-FF-OFm, Fmoc-
FFF-OFm and Fmoc-FFFF-OFm for the formation of stable assemblies. It is worth
noting that for the three peptides under study the number of observed structures at 4 ºC
is significantly higher than that at 25 ºC (i.e. only one stable assembly was identified for
each peptide at 25 ºC). This result clearly indicates that the self-assembly of highly
hydrophobic peptide molecules to form microstructures is a thermodynamically
controlled process.
Peptide synthesis
To a solution of the appropriately N-protected -amino acid (4.00 mmol) in CH2Cl2,
1-hydroxybenzotriazole (HOBt) (4.40 mmol) was added, and the solution was cooled to
0 ºC in an ice bath. N-[3-(dimethylamino)-propyl]-N’-ethylcarbodiimide hydrochloride
(4.40 mmol) was added, followed by the solution of the amino component (4.40 mmol)
in CH2Cl2, obtained after acidolytic removal of the protecting group and N-
methylmorpholine (NMM) (4.40 mmol) or N-ethyldiisopropylamine (DIPEA) (4.40
mmol). The reaction mixture was stirred for 1 h at 0 ºC, then at room temperature for 24
h, by keeping the pH (moistened pH paper) at 8. The reaction mixture was repeatedly
washed with 5% KHSO4, 5% NaHCO3 and water. The organic phase was dried over
9

Soft Matter
Page 10 of 47
View Article Online
DOI: 10.1039/C6SM00337K
MgSO4 and evaporated to dryness. The peptide product was purified by flash
chromatography. Boc-F-OFm was purified using a 8:2 mixture of hexane/ethyl acetate.
Boc-FF-OFm and Boc-FFF-OFm were purified using a 98:2 mixture of
dichloromethane:methanol while a 98:3 mixture was used for Boc-FFFF-OFm. Fmoc-
containing peptides were very insoluble, precipitating upon solvent concentration. All
Fmoc-containing peptides were washed with ethyl acetate. For chemical
characterization, a small amount was purified using a 98:2 or 97:3 mixture of
dichloromethane:methanol.
Description of all intermediates is provided in the Electronic Supporting Information
(ESI).
Self-assembly of Fmoc-FF-OFm
Table 1 summarizes the different conditions required for the formation of the
morphologies identified in this work. Accordingly, assemblies with different
morphologies were obtained from solutions of Fmoc-FF-OFm in HFIP alone or mixed
with a co-solvent. Morphology drastically depends not only on the co-solvent but also
on the peptide concentration and temperature. This is evidenced in Figure S1, which
displays optical micrographs of the most representative morphologies. This enormous
morphological variability is in contrast with observations on Fmoc-FF, which tends to
self-assemble into a hydrogel based on - interlocked -sheets.6
Fmoc-FF-OFm assembles into stacked braid-like microstructures in 2:3 and 1:9
HFIP:water at 4 ºC and peptide concentrations ranging from 0.5 to 2 mg/mL (Figures 2a
and 2b). The length and width of the braids clearly decrease with the peptide
concentration in the solution, inducing a higher degree of compactness in the whole
aggregate. The very compact nature of the stacked aggregate obtained for a
10

Page 11 of 47
Soft Matter
View Article Online
concentration of 0.5 mg/mL is reflected in the AFM topographic images (FiDgOuI:r1e0.120b39)/.C6SM00337K
Sonication of the solutions with a peptide concentration of 2 mg/mL reduced the lateral
staking among the braids increasing the separation between them (Figure 2a), whereas
the system with 0.5 mg/mL remained practically unaltered. This observation is fully
consistent with the fact that the compactness of the supramolecular assemblies increases
with decreasing peptide concentration. Very similar assemblies were obtained in
mixtures with a high content of MeOH, e.g. 1:49 and 1:99 HFIP:MeOH, at 4 ºC and
concentrations ranging from 0.04 to 0.1 mg/mL as is evidenced by representative
optical (Figure S1c) and SEM (Figure S2a) micrographs.
Reduction of the proportion of water co-solvent in the mixtures with HFIP led to the
formation of doughnut-like microstructures (Figure S1d) of different diameters (i.e.
from 4 to 10 m). The most remarkable characteristic of this morphology is the
extremely high surface area, which is expected to facilitate its interaction with other
molecules or even with living organisms present in the environment. Detailed
inspection of these hollow microstructures (Figure 2c) reveals features very similar to
those previously described for polymeric systems prepared using a micelle formation
mechanism with gas bubbles templates.33 Also, macroporous honeycomb scaffolds were
successfully fabricated through the self-assembly of FF using the breath figure method,
in which moist air/water droplets were responsible for the pore formation.34 Although in
this work the central hole of doughnuts-like microstructures formed from Fmoc-FF-
OFm may be also due to nucleation of the assemblies around an air bubble, detailed
observation of the micrographs suggests an alternative mechanism: the coexistence of
small gelatinous rounded aggregates with irregular micrometric assemblies in the same
samples (Figures S1d-right and S2b) indicates that their collapse gives place to the
nucleation center and the subsequent growing of the doughnuts-like microstructures.
11

Soft Matter
Page 12 of 47
View Article Online
Thus, the formation of external filled region seems to precede the appearanDcOeI: o10f.10t3h9e/C6SM00337K
hole (Figure S2c). It should be noted that the presence of gelatinous rounded aggregates
is supported by the role played by aromatic end groups as facilitators of gelation.9,16-18
Accordingly, the majority of the peptide molecules located at the central region of such
gelatinous structures could migrate through diffusion towards the external region,
allowing the development of the hollow microstructure after solubilization or
precipitation of the remaining peptide molecules.
This proposed mechanism agrees with the recent observations of Ulijn and co-
workers35 for Fmoc-Ser-Phe-OMe (Fmoc-SF-OMe), which combines hydrophilic and
hydrophobic amino acids. Fmoc-SF-OMe gave rise to spherulitic structures of several
hundred microns in diameter surrounded by non-spherulitic areas of material, which
was constructed of layered lamellar structures. Interestingly, nucleation points were not
detected in such spherulitic structures, which is consistent with our observations for
Fmoc-FF-OFm. The main difference between the microspherulites reported for Fmoc-
SF-OMe35 and the doughnuts-like assemblies obtained in this work for Fmoc-FF-OFm
refers to the migration of the peptide molecules located at the center of the latter
microstructures. Taking into account the large hydrophobicity of Fmoc-FF-OFm, this
diffusive migration has been considered as a solvent-induced phenomenon to promote
favorable peptide-solvent interactions.
Besides, birefringent microtubes are obtained in 4:1 HFIP:iPrOH and 4:1
HFIP:acetone using 3-4 mg/mL peptide concentrations (Figures S1e and S1f,
respectively). Although the thickness of the tubes obtained in HFIP:iPrOH varies from 1
to 10 µm, all them exhibit a very smooth and regular surface. For instance, SEM
micrographs and AFM images displayed in Figure 3a correspond to a tube with a
thickness of 7 µm. As it can be appreciated in the high magnification SEM micrograph,
12

Page 13 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
tubes present a hexagonal-like symmetry that resembles that found for well-ordered
microtubes formed by self-assembled FF.36,37 The formation of robust hexagonal
microtubes was attributed to the confined organization of nanoscale tubular structures at
long range during slow crystallization or aggregation (i.e. kinetic control of nucleation
and growth).37,38 In general, FF-based nanotubes prepared at room temperatures by
rapid dispersion of molecules exhibit circular shape that is thermodynamically more
stable.12,21,39
Hexagonal-like peptide microtubes were also obtained in 4:1 HFIP:acetone at low
temperature (Figure 3b). Interestingly, high magnification SEM images recorded for the
broken end of a single microtube reveal substructures (i.e. nanotubes) that are confined
and uniaxially oriented along the longitudinal axis of the hexagonal microtube. Thus,
the microtube consists of bundled arrays of nanotubes with a diameter lower than 100
nm. Therefore, these microscale hexagonal tubes result from the supramolecular
organization of such nanotubular structures contained in the array. The hierarchical
evolution from the molecular level to hexagonal-like microtubes through the ordered
bundling of nanotubular substructures is in excellent agreement with the kinetic control
of nucleation and growth discussed above.
In addition to stacked braids, microdoughnuts and microtubes, incipient branched
dendritic-like microstructures were obtained from 2:3 HFIP:water and 1:4 HFIP:EtOH
solutions with relatively diluted peptide concentrations after sonication (Figures S3).
Nevertheless, the stability of these dendritic microstructures, which formed very rapidly
only after perturbing the equilibrium conditions (i.e. removal of the upper thin glass
cover of the glass slip), was very poor, thus disappearing after only 15 min. In spite of
such instability, these results suggest that Phe-based peptides could be used to tune the
morphology of macromolecules and inorganic materials. In a recent study Tendler and
13

Soft Matter
Page 14 of 47
40
View Article Online
DOI: 10.1039/C6SM00337K
co-workers described FF unstable dendritic structures obtained by spin-casting a HFIP
peptide solution (0.5 or 1 mg/mL) onto mica. However, such morphologies, which
transformed into needle-like crystals upon exposure to humid air, corresponded to star-
like dendritic assemblies rather than tree-like structures like those displayed in Figure
S3. Highly ordered dendritic assembly of FF was also reported by Kim and co-
workers,41 who used a buffer peptide solution with pH= 1 and a silicon wafer substrate.
In this case, the morphology of the self-assembled dendrites, which resembled ice
crystal structures in snowflakes,42 was also very different from the tree-like
arrangements achieved for Fmoc-FF-OFm. The instability of the dendritic structures
formed by Fmoc-FF-OFm has been attributed to the effects induced by the
environmental humidity and the surface charge, which experienced drastic changes
when the cover glass slide was removed.
Self-assembly of Fmoc-FFF-OFm
Although previous experiments on Fmoc-FFF proved that such peptide self-
assembles hydrogels,42 a variety of morphologies have been obtained for Fmoc-FFF-
OFm solutions using HFIP alone or mixed with a co-solvent (Table 1). Again, the
morphology changes not only with the co-solvent but also with the peptide
concentration and temperature. Optical micrographs of the most representative
morphologies are provided in Figure S4. A result that deserves special attention is the
apparition of dendritic-like microstructures in HFIP:EtOH mixtures (Figure S4e) that
resemble those observed for Fmoc-FF-OFm (Figure S3). Unfortunately, all the
dendritic-like microstructures achieved for Fmoc-FFF-OFm were very unstable,
disappearing when the glass slip were manipulated for AFM and SEM characterization.
14

Page 15 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
As occurred for the FF-derivative, Fmoc-FFF-OFm self-assemble into stacked braids
from 1:4 HFIP:water at 4 ºC (Figures 4a and S4a), even though in this case such
morphology was only detected for peptide concentrations relatively low in comparison
to those found for Fmoc-FF-OFm (i.e.  1.0 mg/mL). Despite this difference, the degree
of staking in the microstructures apparently increases with decreasing peptide
concentration: Figure 4a compares SEM micrographs of stacked braids formed at 1.0
and 0.05 mg/mL. Thus, the degree of compactness is significantly higher for the
aggregate formed using the lowest peptide concentration. Stacked braids with a similar
morphology were also found in HFIP:EtOH (Figure S5a) at 4 ºC for low peptide
concentrations. However, the most striking feature in these solvent mixtures is that
stacked braids frequently form supramolecular structures at higher peptide
concentrations, giving rise to three-dimensional morphologies. This is evidenced in
Figures 4b and 4c, which display corkscrew-like microstructures observed for peptide
concentrations of 2 and 1 mg/mL, respectively. In all the cases the evolution of the twist
defined by the spirals follow a counter-clockwise sense. It is worth noting that the
corkscrew-like microstructure shown in Figure 4c is attached to a hollow micro-conical
structure. Moreover, inspection of the magnified SEM micrograph (inset) clearly shows
that the stacked braids elements curl. Staked braids are also intuited in the 3D AFM
image. For this system, stable doughnuts-like microstructures were not formed reducing
the amount of co-solvent.
HFIP: iPrOH did not promote the formation of assemblies for peptide concentrations
higher than 0.5 mg/mL, whereas morphologies apparently similar to those displayed in
Figure 4a were obtained below this threshold (Figure S5b). However, detailed
inspection reveals that such morphology does not correspond to stacked braids but to
fused microfibers. In opposition, diluted solutions of Fmoc-FFF-OFm in HFIP:water at
15

Soft Matter
Page 16 of 47
View Article Online
4 ºC produced well-defined individual tubes of submicrometric diameter,DOaI:s10.i1t039is/C6SM00337K
evidenced for a concentration of 0.05 mg/mL in Figure 4d. Moreover, as occurred for
Fmoc-FF-OFm, high magnification micrographs evidence that such structures are build
up through a hierarchical self-similar assembly process, according to which nanotubes
with a diameter lower than 100 nm organize by growing in submicrometic-sized tubular
patterns. The assembly or bundling of Fmoc-FFF-OFm nanotubes into submicroscale
tubes was also observed for concentrated peptide solutions (e.g. 6.1 mg/mL) in 38:11
DMF:MeOH (Figure S5c).
Reduction of the amount of co-solvent in the mixtures led to drastic morphological
changes, thus evidencing the strong effect of the environment on the self-assembly
process, especially at the supramolecular level. For example, although stacked braids
(Figure 4a) and microtubes (Figure 4d) were obtained in 1:4 and 1:99 HFIP:water at
4ºC, respectively, reduction of the water co-solvent to 4:1 HFIP:water leads to the
formation of spherulitic craters (Figure S4d). Inspection of the corresponding SEM and
AFM micrographs (Figure 5a) indicates that the internal and external diameters of those
craters are 3 and 11 m, respectively. Furthermore, the spherulitic morphology
actually consists of a dense packing of short and ultra-thin nanofibers, which group
around a central hole. Finally, birefringent spherulites are obtained after complete
removal of the co-solvent. Interestingly, these compact disc-like assemblies, which are
illustrated in Figure 5b for a 5 mg/mL peptide solution in HFIP at 4ºC, can be described
as a very dense tissue of nanofibers. Birefringent spherulitic structures without
nucleation points at their centers were obtained through enzymatically induced
hierarchical self-assembly of Fmoc-SF-OMe.35 Furthermore, peptide based spherulitic
structures have been also prepared by solvent-induced phase transitions43 and
hierarchical assembly of aromatic peptide amphiphiles.44
16

Page 17 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
Self-assembly of Fmoc-FFFF-OFm
Optical micrographs of the most representative assemblies identified for Fmoc-
FFFF-OFm are displayed in Figure S6. Although in this case the variety of polymorphic
structures decreases with respect to Fmoc-FFF-OFm and, especially, Fmoc-FF-OFm
(Table 1), it should be emphasized that Fmoc-FFFF-OFm exhibits a clear tendency to
form hierarchical dendritic assemblies. The stability and large surface area of the
assembled morphologies make the dendritic growth of this tetrapeptide potentially
useful for advanced micro- and nanofabrication.
Sonication of HFIP:water solutions at 4 ºC with low peptide concentrations (0.1-0.2
mg/mL) promoted the formation of ultra-thin dendritic-like structures, like those
displayed in Figures S6a and 6a. These morphologies, which are observed through the
dark-field mode in the reflected light microscopy, have been attributed to
heterogeneities (i.e. phase-separation phenomena) involved in polycrystalline growth
patterns.45,46 More specifically, dendritic-like structures reflect an interplay between the
ordering effect of crystallization and the disordering effect of interfacial instabilities.
Thus, in absence of growth front nucleation processes, crystallization is known to yield
highly symmetric dendrites, whereas intricately structured and locally disordered
polycrystalline spherulite patterns often form.47 Structures displayed in Figures S6a and
6a have intermediate complexity between these two extreme morphologies. The ultra-
thin nature of such dendritic structures is proved by the AFM amplitude images
included in Figure 6a. Moreover, inspection of the dewetting patterns in both Figures
S6a and 6a suggests the presence of fractal-like holes. Dewetting at the peptide surface
occurs via a nucleation and growth mechanism involved in the formation of the ultra-
thin dendritic-like structures, while the formation of fractal holes has been attributed to
17

Soft Matter
Page 18 of 47
View Article Online
DOI: 10.1039/C6SM00337K
the anisotropy of peptide mobility induced by the solvent. Similar dewetting patterns
were identified in thin films of poly(styrene)-block-poly(methyl methacrylate) diblock
copolymer after annealing in block selective solvent vapor.48
Besides, Fmoc-FFFF-OFm molecule self-assemble forming stacked braids
morphologies, similar to those reached for the smaller peptides, in 1:99 to 1:9
HFIP:EtOH at 4 ºC and peptide concentrations ranging from 0.05 to 0.5 mg/mL
(Figures S6b and 6b). The length of the braids, which ranged from 9 to 16 m, was
practically independent of the peptide concentration (Figure 6b and S7). Interestingly,
higher concentrations provoked the assembly of peptide molecules in soft and meta-
stable dendritic-like architectures (Figure S6c). Although the formation of such
assemblies was highly repetitive, they frequently disappeared in the conditions required
for AFM and, especially, SEM characterization.
The most spectacular assembly was obtained for peptide 1:4 HFIP:EtOH solutions at
room temperature. In this case, well-defined multidimensional dendritic
microarchitectures were obtained, as it is shown in Figure 7 for the 0.5, 1 and 2 mg/mL
Fmoc-FFFF-OFm solutions. These structures appeared very fast and spontaneously by
rubbing a dried mass of the material obtained from the drop-cast of the peptide in the
alcohol mixture with a glass slip. Although these branched structures present some
irregularities, inspection of the AFM images suggest that in the early stage of growth
primary frameworks were nucleated from the center. Thus, these framework structures
exhibit a 4-fold pseudo-symmetry with a typical branching angle of 90º. As growth
continued, the dendrites formed highly branched structures with a specific branching
angle of 45º. Branching angles are schematically depicted in Figure 7 for the dendritic
structure derived from the 1 mg/mL peptide solution, which exhibits the highest
regularity. Finally, it is worth noting that the length of branches decreases with
18

Page 19 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
increasing distance from the primary framework, evidencing that the self-similarity of
these hierarchical microarchitectures is not very high.
It should be mentioned that polymorphism was also reported for Fmoc-FFFF.21 In
HFIP:water this N-Fmoc-protected peptide was found to assemble into ultra-thin
nanoplates that aggregate in microclusters. Replacement of water by EtOH as co-
solvent resulted in the formation of peptide fibrils at peptide concentrations < 0.5
mg/mL, while irregular star-like structures of submicrometric dimensions appeared at
higher peptide concentrations. Finally, poorly defined nanospherical aggregates were
obtained at Fmoc-FFFF concentrations of 1 mg/mL in HFIP:water.
Fractal analysis of Fmoc-FFFF-OFm dendritic microarchitectures
In this section we introduce a fractal analysis of the geometrical structures of the
dendritic microarchitectures obtained for Fmoc-FFFF-OFm. Fractal objects are self-
similar structures for which increasing magnifications reveal similar features on
different length scales.49 The fractal dimension (FD), which indicates how a fractal
pattern changes with the scale at which it is measured, was determined by the box-
counting method9 analyzing binary images (Figure S8) derived from the AFM images
displayed in Figure 7.
Figure 8 shows logarithm plots of the number of squares boxes that occupy at least
part of the dendritic microstructure (N) as a function of the side length of the square box
(L). The scaling relationship N(L)  L-FD relates FD with slope in the logarithmic scale
(m) through the FD = –m. From the slopes in Figure 8, FD ranges from 1.69 to 1.71
reflecting the self-similarity (i.e. dimensional consistency at different length scales) of
the dendritic microstructures formed by Fmoc-FFFF-OFm. Furthermore, these FD
values are very close to the ideal value of 1.67 expected for microstructures generated
19

Soft Matter
Page 20 of 47
by diffusion limited aggregation (DLA) onto a 2D substrate surface.50,51 AccDoOrId: 1i0n.1g03t9o/C6SM00337K
this, in Fmoc-FFFF-OFm 1:4 HFIP:EtOH solutions, peptide molecules or small
aggregates diffuse by Brownian motion and a random walk process occurs until they
contact and adhere to another one. This process, which occurs multiple times to form
clusters, is dominated by steric constraints and short-range forces and, therefore, a
diffusing molecule or small aggregate is more likely to adhere to the outer edges of the
growing cluster than to the internal regions. The steric shielding of the internal regions
of the clusters is responsible for the formation of dendritic microstructures that are
fractal.
View Article Online
Theoretical calculations
Density Functional Theory (DFT) calculations at the M06L/6-31G(d) level have
been performed on Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-FFFF-OFm assemblies
to evaluate both the effect of the peptide length in the binding energy and the
preferences of the strands within the -sheet. It should be remarked that the M06L is a
meta-hybrid generalized gradient approximation (GGA) functional with a very good
response under dispersion forces, improving one of the biggest deficiencies in DFT
methods.31,32 Thus, the M06L functional is able to describe the geometry and interaction
energy of complexes stabilized by non-covalent interactions, including -staking, with
accuracy close to that of couple cluster calculations with both single and double
substitutions.
Starting conformations for Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-FFFF-OFm
were taken from a previous DFT study on n-Phe homopeptides with n ranging from 1 to
4.52 As those calculations were performed considering acetyl (Ac) and N-methylamide
(NHMe) as N- and C-terminal groups, these have been replaced by Fmoc and OFm,
20

Page 21 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
respectively, all the resulting structures being re-optimized at the M06L/6-31G(d) level.
After this, the three lower energy conformations of each peptide were used to construct
assemblies involving three identical molecules arranged in parallel or antiparallel
orientations defined with respect to the molecule. Around 20-30 starting complexes
were obtained as for each peptide, complete geometry optimizations being conducted
for all them at the M06L/6-31G(d) level.
Although the three minimum energy conformations considered for Fmoc-FF-OFm,
Fmoc-FFF-OFm and Fmoc-FFFF-OFm peptides in the starting complexes did not
correspond to the semi-extended -strands,52 geometry optimizations led to the -sheet
alignment of three -strands as the most stable complex for both parallel and
antiparallel assemblies. The most important characteristics associated with such lowest
energy -sheet dispositions are summarized in Table 2 for each peptide, while Figures
9, S9 and S10 represent such structures for Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-
FFFF-OFm, respectively. In the parallel disposition the phenyl side groups of all strands
are perfectly packed forming aromatic -ladders, while the phenyl groups of
consecutive strands point in opposite directions in the antiparallel arrangement. The
number of intermolecular hydrogen bonds in Fmoc-FF-OFm, Fmoc-FFF-OFm and
Fmoc-FFFF-OFm is 2, 3, and 4, respectively, per pair of interacting molecules either in
parallel or antiparallel. However, in all cases hydrogen bonding parameters, especially
the H···O distance, are more favourable for the antiparallel than for the parallel
assembly. This feature agrees with the fact that the antiparallel sheet is energetically
favoured with respect to the parallel one for the three investigated peptides. Moreover,
the instability of the parallel assembly increases from 1.8 to 9.8 kcal/mol when the
number of Phe residues in the peptide increases from 2 to 4. This behavior is fully
consistent with the interaction energies (Eint) listed in Table 2, which only take into
21

Soft Matter
Page 22 of 47
View Article Online
DOI: 10.1039/C6SM00337K
account the non-covalent interactions between the three peptides of each system. Thus,
the Eint values calculated for the antiparallel / parallel assemblies decrease from -52.8 /
-51.0 kcal/mol to -88.6 / -78.7 kcal/mol when the number of Phe per molecule increases
from 2 to 4.
The cooperative energy (Ecoop) values for all complexes, which were calculated as it
is described in the ESI, are included in Table 2. It should be noted that Ecoop provides
an evaluation of the many-body non-additive effects. Amazingly, the Ecoop values
obtained for the parallel assembly of the three peptides are very similar, ranging from -
7.5 (Fmoc-FF-OFm and Fmoc-FFF-OFm) to -8.3 kcal/mol (Fmoc-FFFF-OFm). In
contrast, for the antiparallel assembly, Ecoop gradually decreases from a repulsive value
of +1.5 kcal/mol for Fmoc-FF-OFm, which is consistent with the presence of anti-
cooperative effects, to -4.9 kcal/mol for Fmoc-FFFF-OFm. Accordingly, representation
of both the relative energy between the parallel and antiparallel arrangements (E) and
the difference between their Ecoop values [Ecoop= Ecoop(antiparallel) –
Ecoop(parallel)] against the number of Phe residues (Figure S11) clearly indicates that
the stability of the antiparallel assembly with respect to the parallel one increases with
the size of the homopeptide.
The antiparallel preferences of the three peptides studied in this work have been
proved by FTIR spectroscopy. More specifically, FTIR spectra were registered
considering two different samples of each peptide: (i) fibrous self-assembled structures;
and (ii) quenched samples coming from (i) (i.e. fibrous structures were melted at 210-
260 ºC and immediately cooled using liquid nitrogen to avoid the crystallization and/or
formation of ordered structures). Comparison of the spectra recorded for the two
samples of Fmoc-FF-Fmoc (Figure S12) confirms the tendency of this peptide to self-
assemble into antiparallel -sheets. Thus, fibrous samples exhibit sharp peaks at 1617
22

Page 23 of 47
Soft Matter
and 1695 cm-1, which have been related with the presence of -sheet structuresDaOnI:d10.w10i3Vt9ihe/wC6ASrtMicl0e0O3n3li7nKe
an antiparallel arrangement of the -sheets, respectively.53 However, although the
positions of these two peaks are typical to an antiparallel arrangement of the β-sheets,
their relative intensity is not. Thus, for ideally defined antiparallel β-sheet arrangements,
usually find in longer peptide sequences and proteins, the band at 1695 cm-1 is expected
to be much weaker.54 This unusual feature was also detected for the Fmoc-FF26 and, by
analogy, has been attributed to the short peptide sequence. Thus, the antiparallel β-sheet
obtained for these two amino acids peptide (Fmoc-FF-Fmoc and Fmoc-FF) is not
ideal.26 Interestingly, the spectra recorded for fibrous Fmoc-FFF-Fmoc and Fmoc-
FFFF-Fmoc samples also display the peak at 1695 cm-1, even though in these cases a
very intense peak appears at 1640-1645 cm-1 (Figure S13). The latter absorption, which
has been also identified in fibrous amyloids organized in antiparallel -sheets,55 is
related with the twist angle of -sheets formed by a large number of -strands. These
observations are fully consistent with previously discussed theoretical calculations.
Significant findings are also detected from the comparison of the antiparallel
preferences found in this work for Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-FFFF-
OFm with those of their unblocked homologues or with a single Fmoc group. Thus, X-
ray crystallography proved a parallel -sheet assembly for FF,56 while which was
confirmed by FTIR studies.57 In contrast, intermolecular hydrogen bonds and anti-
parallel sheets were found for Fmoc-FF.6 This change was attributed to the interlocking
the Fmoc groups from alternate -sheets to create -stacked pairs. On the other hand,
Tamamis et al.42,58 and Guo et al.59 used atomistic and coarse-grained molecular
dynamics (MD) simulations, respectively, to study the assembly mechanism and the
molecular basis for the structural features of FFF-based peptides nanostructures.
Authors found that FFF-based peptides spontaneously assembled into solid nanometer-
23

Soft Matter
Page 24 of 47
View Article Online
sized nanospheres and nanorods with substantial content of anti-parallelDOI:-1s0h.1e03e9t/.C6SM00337K
Furthermore, theoretical calculations on FFFF and Fmoc-FFFF using the methodology
that in the present study revealed a parallel -sheets for FFFF and Fmoc-FFFF.21 The
overall of those previous observations combined with the results presented in work
indicate that the -sheet assembly of Phe-based peptides depends on their
hydrophobicity, which in turn varies with both the number of Phe residues and the
presence of one or two Fmoc groups. Accordingly, it can be concluded that -stacking
interactions are the driving force for the formation of -sheet assemblies in Fmoc-
containing peptides.
Quantitative prediction of 2D and 3D assemblies combining a bottom-up approach
with atomistic computer simulations is a very complex task. Thus, although the rational
design of molecules for biological and pharmaceutical applications is rather well
established, only few works predict the complete self-assembly of soft materials using
computational tools.30,31,60,61 Experimental observations presented in this work for small
Phe-homopeptides with two fluorenyl functionalities reveal that their self-assembly is
greatly influenced by both the solvent and the peptide concentration, which further
complicates the prediction of their 3D structure. Because of this, no attempt has been
carried out in this work to extend the antiparallel 1D model predicted for Fmoc-FF-
OFm, Fmoc-FFF-OFm and Fmoc-FFFF-OFm to 2D packing of -sheets.
CONCLUSIONS
Self-assembled structures in the nano- and microscale are central for future
technological applications. Within this context peptide assemblies are of special interest
for the nanoscience and biomaterials science communities. Following our previous
work, which proved the importance of N- and C-termini capping groups in the ordered
24

Page 25 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
assemblies of di-, tri- and tetraphenylalanine motifs, the current work has explored the
self-assemblies of highly aromatic peptides by capping the N- and C-terminal ends of
Phe-derivatives with fluorenylmethoxycarbonyl and fluorenylmethyl ester, respectively.
Specifically, we have reported the spontaneous formation of a large number of stable
polymorphic structures for Fmoc-FF-OFm, Fmoc-FFF-OFm and Fmoc-FFFF-Fmoc.
Thus, depending on the solvent:co-solvent conditions, peptide concentration and
temperature, molecules can organize into stacked braids, doughnuts-like, bundled arrays
of nanotubes, corkscrew-like and/or spherulitic microstructures. In addition to these
well-organized structures, such three small bioinspired molecules exhibit some ability to
assemble into dendritic assemblies of various sizes. Unfortunately, the dendritic
structures formed by Fmoc-FF-OFm and Fmoc-FFF-OFm were unstable; they rapidly
disappeared when the equilibrium conditions were slightly perturbed. On the contrary,
very stable well-defined dendritic structures are derived from 0.5-2 mg/mL Fmoc-
FFFF-OFm solutions in 1:4 HFIP:EtOH at room temperature. AFM images have
revealed that they consisted of a central framework with 4-fold pseudo-symmetry that
grows forming highly branched structures. The fractal dimension analysis used in this
work reflects not only self-similarity but also that the dendritic assembly occurs through
a diffusion limited aggregation mechanism onto a plane (i.e. the surface substrate).
Theoretical calculations considering a model -sheet with three interacting strands
indicate that the studied peptides adopt an antiparallel arrangement, which is more
stable than the parallel one, with intermolecular hydrogen bonds and - interactions.
Moreover, such stability increases with the length of the Phe-segment. Unfortunately,
extension of these theoretical studies to 2D and/or 3D nano-architectures represents a
very complex task that, in addition, is severely limited by the large influence of the
25

Soft Matter
Page 26 of 47
View Article Online
DOI: 10.1039/C6SM00337K
environmental conditions (e.g. solvent, peptide concentration and temperature) in the
assembly.
In summary, our results suggest that Phe-homopeptides capped with two fluorenyl
functionalities are a novel class of material that can be used to achieve a wide variety of
desirable structures at the very small length-scale by simply controlling the assembly
conditions. In particular, the well-defined and stable dendritic structures formed by
Fmoc-FFFF-OFm indicate that highly aromatic Phe-homopeptides with four, or even
more, residues should be considered as powerful building blocks for the fabrication of
complex and relatively infrequent structures. Potential applications of peptide
assemblies in nanotechnology and nanobiology should be explored along next decades.
Acknowledgements
Authors thank supports from MINECO and FEDER (MAT2012-34498, MAT2012-
36205 and CTQ2013-40855-R), Generalitat de Catalunya (XRQTC) and CESCA and
Gobierno de Aragón - Fondo Social Europeo (research group E40). Support for the
research of C.A. was received through the prize “ICREA Academia” for excellence in
research funded by the Generalitat de Catalunya.
References
1.
2.
3.
X. Yan, P. Zhu and J. Li, Chem. Soc. Rev., 2010, 39, 1877.
S. Flemming and R. V. Ulijn, Chem. Soc. Rev., 2014, 43, 8150.
S. Bai, C. Pappas, S. Debnath, P. W. J. M. Frederix, J. Leckie, S. Fleming and R.
V. Ulijn, ACS Nano, 2014, 8, 7005.
4.
J. Shi, Y. Gao, Z. Yang and B. Xu, Beilstein J. Org. Chem., 2011, 7, 167.
26

Page 27 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
5.
6.
7.
R. J. Williams, A. M. Smith, R. Collins, N. Hodson, A. K. Das and R. V. Ulijn,
Nat. Nanotechnol., 2009, 4, 19.
A. M. Smith, R. J. Williams, C. Tang, P. Coppo, R. F. Collins, M. L. Turner, A.
Saiani and R. V. Ulijn, Adv. Mater., 2008, 20, 37.
V. Castelletto, C. M. Moulton, G. Cheng, I. W. Hamley, M. R. Hicks, A.
Rodger, D. E. López-Pérez, G. Revilla-López and C. Alemán, Soft Matter, 2011,
7, 11405.
8.
9.
A. K. Das, R. Collins and R. V. Ulijn, Small, 2008, 4, 279.
V. Jayawarna, M. Ali, T. A. Jowitt, A. F. Miller, A. Saiani, J. E. Gough and R.
V. Ulijn, Adv. Mater., 2006, 18, 611.
10.
11.
D. E. López-Pérez, G. Revilla-López, I. W. Hamley and C. Alemán, Soft Matter,
2013, 9, 11021.
Z. Yang, H. Gu, D. Fu, P. Gao, J. K. Lam and B. Xu, Adv. Mater., 2004, 16,
1440.
12.
13.
14.
15.
16.
M. Reches and E. Gazit, Science, 2003, 300, 625.
M. Reches, Y. Porat and E. Gazit, J. Biol. Chem., 2002, 277, 35475.
M. Reches and E. Gazit, Nano Lett., 2004, 4, 581.
N. Amdursky, E. Gazit and G. Rosenman, Adv. Mater., 2010, 22, 2311.
A. Mahler, M. Reches, M. Rechter, S. Cohen and E. Gazit, Adv. Mater., 2006,
18, 1365.
17.
18.
J. Raeburn, C. Mendoza-Cuenca, B. N. Cattoz, M. A. Little, A. E. Terry, A. Z.
Cardoso, P. C. Griffiths and D. J. Adams, Soft Matter, 2015, 11, 927.
S. Debnath, A. Shome, D. Das and P. K. Das, J. Phys. Chem. B, 2010, 114,
4407.
27

Soft Matter
Page 28 of 47
View Article Online
19.
20.
21.
22.
23.
24.
T. H. Han, T. Ok, J. Kim, D. O. Shin, H. Ihee, H.-S. Lee and S. O. Kim, Small,
DOI: 10.1039/C6SM00337K
2010, 6, 945.
L. Chronopoulou, S. Lorenzoni, G. Masci, M. Dentini, A. R. Togna, G. Togna,
F. Bordi and C. Palocci, Soft Matter, 2010, 6, 2525.
E. Mayans, G. Ballano, J. Casanovas, A. Díaz, A. Pérez-Madrigal, F. Estrany, J.
Puiggalí, C. Cativiela and C. Alemán, Chem. Eur. J., 2015, 21, 16895.
P. P. Sharma, B. Rathi, J. Rodrigues and N. Y. Gorobets, Curr. Topics Med.
Chem., 2015, 15, 1268.
V. E. Santo, M. E. Gomes, J. F. Mano and R, L. Reis, Nanomedicine, 2012, 7,
1045.
A. Amdursky, P. Beker, J. Schklovsky, E. Gazit and G. Rosenman,
Ferroelectrics, 2010, 399, 107.
25.
26.
M. Reches and E. Gazit, Phys. Biol., 2006, 3, S10.
C. Tang, A. M. Smith, R. F. Collins, R. V. Ulijn and A. Saiani, Langmuir, 2009,
25, 9447.
27.
28.
29.
S. M. M. Reddy, G. Shanmugam, N. Duraipandy, M. S. Kiran and A. B.
Mandal, Soft Matter, 2015, 11, 8126.
H. Takayasu, Fractals in the Physical Sciences, Manchester University Press,
Manchester and New York, 1990, pp 11–17.
W. Rasband, NIH, Fractal box counting method using ImageJ,
http://www.nist.gov/lispix/doc/fractal/image-java--box-count.htm, accessed:
December, 2015.
30.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
28

Page 29 of 47
Soft Matter
View Article Online
DOI: 10.1039/C6SM00337K
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J.
A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E.
Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N.
Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09, Revision A.02.
Gaussian Inc. 2009.
31.
32.
33.
Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215.
Y. Zhao and D. G. Truhlar, J. Chem. Phys., 2006, 125, 194101.
B. Teixeira-Dias, C. Alemán, F. Estrany, D. S. Azambuja and E. Armelin,
Electrochim. Acta, 2011, 56, 5836.
34.
35.
M. Du, P. Zhu, X. Ya, Y. Su, W. Song and J. Li, Chem. Eur. J., 2011, 17, 4238.
M. Hughes, H. Xu, P. W. J. M. Frederix, A. M. Smith, N. T. Hunt, T. Tuttle, I.
A. Kinloch and R. V. Ulijn, Soft Matter, 2011, 7, 10032.
36.
37.
38.
39.
C. H. Görbitz, Chem. Commun., 2006, 2332.
X. Yan, J. Li and H. Möhwald, Adv. Mater., 2011, 23, 2796.
H. Cölfen and S. Mann, Angew. Chem. Int. Ed., 2003, 42, 2350.
X. H. Yan, Y. Cui, Q. He, K. W. Wang, J. B. Li, W. H. Mu, B. L. Wang and Z.
C. Ou-yang, Chem. Eur. J., 2008, 14, 5974.
40.
V. V. Korolkov, S. Allen, C. J. Roberts and S. J. B. Tendler, Faraday Discuss.,
2013, 166, 257.
29