Effect of the molecular structure on the hierarchical self-assembly of semifluorinated alkanes at the air/water interface

Article abstract of DOI:10.1021/la201377f

Semifluorinated alkanes (C_nF_2n+1C_mH _2m+1), short FnHm display local phase separation of mutually incompatible hydrocarbon and fluorocarbon chain moieties, which has been utilized as a structure-forming motif in supramolecular architectures. The packing of semifluorinated alkanes, nominally based on dodecyl subunits, such as perfluoro(dodecyl)dodecane (F12H12) and perfluoro(dodecyl)eicosane (F12H20), as well as a core extended analogue, 1,4-dibromo-2-((perfluoroundecyl)methoxy)-5- (dodecyloxy)benzene) (F11H1-core-H12), was studied at the air/water interface. Langmuir monolayers were investigated by means of neutron reflectivity directly at the air/water interface and scanning force microscopy after transfer to silicon wafers. Narrowly disperse surface micelles formed in all three cases; however, they were found to bear different morphologies with respect to molecular orientation and assembly dimensionality, which gives rise to different hierarchical aggregate topologies. For F12H12, micelles of ca. 30 nm in diameter, composed of several circular or "spherical cap" substructures, were observed and a monolayer model with the fluorocarbon block oriented toward air is proposed. F12H20 molecules formed larger (ca. 50 nm diameter) hexagonally shaped surface micelles that were hexagonally, densely packed, besides more elongated but tightly interlocked wormlike structures. Conversely, F11H1-core-H12 films organized into linear rows of elongated surface micelles with comparable width, but an average length of ca. 400 nm, apparently formed by antiparallel molecular packing.

Full text of DOI:10.1021/la201377f

ARTICLE
pubs.acs.org/Langmuir
Effect of the Molecular Structure on the Hierarchical Self-Assembly
of Semifluorinated Alkanes at the Air/Water Interface
Laurence de Viguerie,† Rabea Keller,‡ Ulrich Jonas,*,† R€udiger Berger,‡ Christopher G. Clark, Jr.,‡
||
Christopher O. Klein,† Thomas Geue,§ Klaus M€ullen,‡ Hans-J€urgen Butt,‡ and Dimitris Vlassopoulos†,
†Bio-Organic Materials Chemistry Laboratory, Institute of Electronic Structure & Laser, Foundation for Research and
Technology—Hellas (FORTH), P.O. Box 1527, 71110 Heraklion, Greece
‡Max Planck Institute for Polymer Research, 55128 Mainz, Germany
§Laboratory for Neutron Scattering, Paul Scherrer Institut, 5232 Villigen, Switzerland
Department of Materials Science and Technology, University of Crete, Heraklion, Greece
S Supporting Information
b
ABSTRACT: Semifluorinated alkanes (CnF2nþ1CmH2mþ1),
short FnHm display local phase separation of mutually incom-
patible hydrocarbon and fluorocarbon chain moieties, which
has been utilized as a structure-forming motif in supramole-
cular architectures. The packing of semifluorinated alkanes,
nominally based on dodecyl subunits, such as perfluoro-
(dodecyl)dodecane (F12H12) and perfluoro(dodecyl)eicosane
(F12H20), as well as a core extended analogue, 1,4-dibromo-2-((perfluoroundecyl)methoxy)-5-(dodecyloxy)benzene) (F11H1-
core-H12), was studied at the air/water interface. Langmuir monolayers were investigated by means of neutron reflectivity directly at
the air/water interface and scanning force microscopy after transfer to silicon wafers. Narrowly disperse surface micelles formed in
all three cases; however, they were found to bear different morphologies with respect to molecular orientation and assembly
dimensionality, which gives rise to different hierarchical aggregate topologies. For F12H12, micelles of ca. 30 nm in diameter,
composed of several circular or “spherical cap” substructures, were observed and a monolayer model with the fluorocarbon block
oriented toward air is proposed. F12H20 molecules formed larger (ca. 50 nm diameter) hexagonally shaped surface micelles that
were hexagonally, densely packed, besides more elongated but tightly interlocked wormlike structures. Conversely, F11H1-core-
H12 films organized into linear rows of elongated surface micelles with comparable width, but an average length of ca. 400 nm,
apparently formed by antiparallel molecular packing.
to great detail at the air/water interface, e.g., by grazing-incidence
X-ray diffraction,5ꢀ8 the situation is more ambiguous for the
amphiphobic SFA structures without hydrophilic head groups.
Examination of diverse FnHm diblocks indicated that they tend to
self-assemble spontaneously at different characteristic length
scales into “surface micelles” (circular or elongated).5,6 This
aggregation behavior suggests that these molecules can be
exploited as motifs for hierarchical organization. A major question
in this context is the orientation of the molecules at the air/water
interface. Different models of film structures, where molecules are
oriented either parallel to each other or in a head-to-tail config-
uration, have been suggested on the basis of scanning force
microscopy (SFM), X-ray reflectivity, and grazing-incidence
small-angle X-ray scattering (GISAXS) measurements.9ꢀ11
One of the investigated molecules, F12H12 (see Figure 1) has
been previously studied for its propensity to form phase-sepa-
rated, lamellar structures in the solid bulk state.12ꢀ16 More recent
1. INTRODUCTION
Semifluorinated alkanes (SFAs, CnF2nþ1CmH2mþ1, FnHm)
are known to form stable Langmuir monolayers at the air/water
interface despite the hydrophobicity of both the alkyl chain and
the fluorinated moiety.1,2 They have been widely studied in the
literature because of their unique properties, determined by the
strong incompatibility of fluorocarbon and hydrocarbon blocks,
which drives molecular phase separation. In this respect, these so-
called amphiphobic molecules can be considered as molecular
analogues of incompatible diblock copolymers. In the bulk, they
exhibit a richness of solid state packing structures. Moreover,
they are applied for molecular interfacial film stabilization to
improve colloidal systems of biological interest, such as lipo-
somes employed as drug carriers, or in the fluorocarbon-in-water
emulsions used as blood substitutes.3,4
Significant efforts have been devoted to the investigation of
their structures at the air/water interface, and different structural
models have been proposed for various molecular architectures.
While the molecular orientation and packing of an SFA carrying a
hydrophilic head group at one end of the chain in contact with
the water phase is rather straightforward and could be elucidated
Received: April 15, 2011
Revised:
May 19, 2011
Published: June 14, 2011
r
2011 American Chemical Society
8776
dx.doi.org/10.1021/la201377f Langmuir 2011, 27, 8776–8786
|

Langmuir
ARTICLE
Figure 1. Chemical structures, molecular CPK models, and schematic representations of (a) F12H12 (with F12H20 being analogous, but with longer
H-alkyl chain) and (b) F11H1-core-H12.
Scheme 1. Synthetic Route to F11H1-core-H12 (1,4-Bibromo-
2-((perfluoroundecyl)methoxy)-5-(dodecyloxy)benzene))
monolayer morphology and neutron reflectivity to assess the
molecular orientation within the layer and determine the layer
thickness. To our knowledge, this is the first study to focus on
the structural properties (besides previously reported isotherm
data) of monolayers of semifluorinated alkane molecules with
symmetric F- and H-block lengths. As these SFA-type mol-
ecules are attractive for novel organization motifs in two dimen-
sions, it is of general interest whether similar block lengths
induce specific self-assembly properties different from the
reported asymmetric cases. Moreover, the results on the purely
alkyl-based SFAs are contrasted with those of the core-
extended molecule, which combines aromatic with alkyl moi-
eties and is expected to feature different modes of intermole-
cular interactions.
2. EXPERIMENTAL SECTION
1H, 13C, and 19F solution state NMR spectra were recorded on a
Bruker AMX500 NMR spectrometer using the residual proton of the
solvent (6.00 ppm for C2HDCl4) or the carbon signal of the deuterated
solvent (73.78 ppm for C2D2Cl4) as an internal standard. 19F NMR data
were externally referenced using hexafluorobenzene in C2D2Cl4
(ꢀ160.89 ppm). Field desorption mass spectra (FDMS) were per-
formed with a VG-Instruments ZAB 2-SE-FDP using 8 kV accelerating
voltage. Elemental analyses were performed by the Microanalytical
Laboratory of Johannes Gutenberg University.
2.1. Materials. Cesium carbonate was dried in vacuo overnight at
100 °C. All other materials were used as received from Aldrich.
Perfluoro(dodecyl)dodecane (F12H12) and perfluoro(dodecyl)eicosane
(F12H20) were synthesized as previously reported.14 1,4-Dibromo-2-
((perfluoroundecyl)methoxy)-5-(dodecyloxy)benzene) (F11H1-core-
H12) was synthesized, as outlined in Scheme 1, by alkylation of 2,
5-dibromo-4-dodecyloxyphenol20 with 1-perfluoroundecylmethylnonaflate21
in DMF using cesium carbonate as base. A methylene group adjacent to the
aryl ether was employed instead of a CF2 group, e.g., F11H1 vs F12, so
that the classical nucleophilic substitution could be carried out. The conse-
quence is a slightly more flexible ether linkage, but essentially preserving the
same overall helical conformations (left and right handed) as the SFA
fluorocarbon moiety. The synthesis of 1,4-dibromo-2-dodecyloxy-5-methox-
ybenzene is described elsewhere.20
investigations using different complementarytechniques (Brillouin
light scattering spectroscopy, shear rheometry, X-ray scattering,
solid state NMR, PVT measurements, etc.) have elaborated on the
mechanism of the lamellar first-order phase transitions in three
dimensions.17ꢀ19
Herein, we report the aqueous surface-mediated self-assembly
into molecularly thin films of SFAs F12H12, F12H20, and F11H1-
core-H12 (with a central 2,5-dibromo-1,4-phenyl core being
connected to each of the two chains via ether linkages), and their
structure investigation by scanning force microscopy to image the
2.1.1. 2,5-Dibromo-4-dodecyloxyphenol. A one-neck round-bottom
flask, equipped with a reflux condenser, was charged with 1,4-dibromo-
2-dodecyloxy-5-methoxybenzene (4.675 g, 10.38 mmol) and a solution
of lithium triethyl borohydride or Super-Hydride (1.0 M in THF,
8777
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
17.0 mL, 17.0 mmol). The reaction mixture was heated to reflux for
5 days under argon, cooled to 0 °C, and diluted with diethyl ether
(200 mL), and the resulting mixture was quenched with water. The ether
layer was washed with hydrochloric acid (aqueous, 5%) and brine. The
organic phase was dried over MgSO4, filtered, and concentrated in vacuo
to yield a colorless oil (4.30 g, 95%). The product was identified by 1H
NMR, identical to that of previously reported alternative syntheses,22,23
and was used without further purification.
cantilevers (Olympus OMCL-AC240TS, nominal tip radius <7 nm,
resonance frequency 50ꢀ90 kHz, spring constant 0.7ꢀ3.8 N mꢀ1).
3
The cantilevers were put on a glass coverslip with the tip side facing air
and were Ar plasma cleaned (Plasma Cleaner/Sterilizer PDC-002,
200W, Harrick Scientific Corp., United States) for 30 s at a pressure
of approximately 1.6 mbar.
For Kelvin probe force microscopy (KPFM) measurements the chip
was electrically connected to the chip holder using conductive silver
paste (Fluka, for electron microscopy). SFM and KPFM measurements
were performed under ambient conditions (room temperature, air).
Imaging of F12H12, F12H20, and F11H1-core-H12 samples was
performed at the lowest possible tapping forces characterized by the
ratio of free amplitude/amplitude set point26 of 0.96 and a peak offset of
ꢀ3%. The free amplitude was determined to be the amplitude of the
cantilever at peak maximum before approaching the surface. Even then
we observed in some cases that material was dislocated or removed by
the tip during the measurements, in particular in the case of F11H1-core-
H12 films. The generated “holes” in the surface micelle layer were used
to estimate the layer thickness. The SFM topographic images were
flattened and analyzed using Gwyddion Software (www.gwyddion.net).
KPFM is an operation mode, which is sensitive to local variations in
the surface potential of materials. Here, the detection of the surface
potential is achieved.27 First the topography is scanned, and then
the topographic profile is followed by a second scan at a defined distance
from the surface. During this scan the mechanical excitation is switched
off and an AC bias of 2000 mV is applied to the tip while the sample
remains grounded. An additional variable DC potential is applied at
the tip in order to nullify the oscillation amplitude of the cantilever by a
feedback loop. This specific DC potential value is recorded at each raster
point and is taken as the Kelvin potential difference between the tip and
the sample surface. The KPFM images were flattened using median line
correction.
2.1.2. 1,4-Dibromo-2-((perfluoroundecyl)methoxy)-5-(dodecyloxy)
benzene. A mixture of 2,5-dibromo-4-dodecyloxyphenol (2.0 g, 4.58
mmol) and 1-perfluoroundecylmethylnonaflate (4.45 g, 5.04 mmol) in
DMF (45.8 mL) was degassed by bubbling with argon for 40 min.
Cesium carbonate (2.25 g, 6.90 mmol) was added, and the reaction
mixture was heated to 65 °C for 12 h. The reaction mixture was
concentrated in vacuo and dissolved in refluxing ethyl acetate in water.
The aqueous layer was extracted a second time with refluxing ethyl
acetate. The organic layers were combined and washed with water,
aqueous sodium hydroxide (5%), and brine. The organic mixture was
dried over MgSO4, adsorbed to silica gel (9 g) in vacuo, and chromato-
graphed on silica gel using 15% ethyl acetate in hexane as eluent to yield a
white powder (3.9 g, 84%). The product was recrystallized from
perfluoromethylcyclohexane on cooling from reflux at 1 °C/h over 2
days to yield colorless crystals. 1H NMR (500 MHz, C2D2Cl4, 105 °C) δ
0.98 (t, 3H, 3JHH = 6.8 Hz, CH3), 1.3ꢀ1.5 (m, 22H, CH2), 1.58 (quintet,
2H, 3JHH = 7.1 Hz, OCH2CH2CH2), 1.89 (quintet, 2H, 3JHH = 6.9 Hz,
OCH2CH2), 4.07 (t, 2H, 3JHH = 6.4 Hz, OCH2CH2), 4.52 (t, 2H, 3JHF
=
13.1 Hz, OCH2CF2), 7.18 (s, 1H, ArH), 7.24 (s, 1H, ArH) ppm; 13C
NMR (125 MHz, C2D2Cl4, 105 °C) δ 13.50, 22.24, 25.68, 28.92, 28.93,
28.98, 29.17, 29.20, 29.27, 29.28, 31.56, 68.25 (t, 2JCF = 27.1 Hz), 70.88,
111.86, 112.53, 119.32, 121.83, 149.03, 152.67 ppm; 19F NMR (470
MHz, C2D2Cl4, 105 °C) δ ꢀ80.28 (3F, CF3), ꢀ118.08 (2F, CH2CF2),
ꢀ120.24 (8F, CF2), ꢀ120.45 (4F, CF2), ꢀ121.35 (2F, CF2), ꢀ121.82
(2F, CF2), ꢀ124.72 (2F, CF2CF3) ppm; FDMS: m/z [u eꢀ1] 1018.9
(100%), [Mþ]; calcd: 1018.0 (C30H2979Br81BrF23O2); Anal. calcd for
C30H29Br2F23O2: C, 35.38; H, 2.87; found: C, 35.43; H, 2.93.
2.4. Neutron Reflectivity. Neutron reflectivity experiments were
performed at the air/water interface using the dedicated Langmuir trough
on the AMOR reflectometer at the Swiss Spallation Neutron Source
(SINQ) at the Paul Scherrer Institute, Switzerland, with D2O as subphase.
The instruments and the procedure used are described elsewhere.28,29
Measurements were made at three incident angles (0.6, 1.4, and 2.4°),
which gave a range of momentum transfer Q (with Q = (4π sin θ)/λ,
where θ is the glancing angle of incidence and π is 3.14 and NOT the
Spreading solutions (1 mg mLꢀ1) were prepared in hexane
3
(analytical grade). Water was purified to a resistivity of 18.2 MΩ cm
3
using a Milli-Q filtration system (Millipore).
2.2. Monolayer Isotherms and Monolayer Transfer. Surface
pressure versus molecular area (πꢀA) isotherms were recorded on three
different Langmuir troughs made of hydrophobic Teflon, equipped with
two movable hydrophilic polyoxymethylene barriers: KSV Minitrough 4
for ISR 400 (KSV Instruments Ltd., Finland), appropriately modified
Langmuir-Trough RK1 (Riegler and Kirstein GmbH, Potsdam,
Germany) and a trough dedicated to scattering experiments.24 The
surface pressure, π, was measured using the Wilhelmy plate method.
Temperature was regulated to 20.0 °C. An equilibration time of 5 min
after spreading was used for solvent evaporation.
surface pressure (same symbol)) from 0.1 to 1.5 nmꢀ1
.
All data were analyzed using Parratt32 software, version 1.6 (Helmholtz-
Zentrum Berlin f€ur Materialien and Energie GmbH Germany).30
3. RESULTS
3.1. Isotherms. The surface pressure/molecular area iso-
therms for F12H12, F12H20, and F11H1-core-H12 are shown
in Figure 2. All isotherms comprise a first increase in surface
The monolayers were compressed up to the desired surface pressure
pressure that occurs at surface areas A0 of ca. 33ꢀ34 Å2
(compression speed 5 mm minꢀ1) and transferred onto plasma-cleaned
3
moleculeꢀ1 for F12H12, at 27ꢀ32 Å2 moleculeꢀ1 for
3
3
(10 min Ar plasma) silicon wafers. The silicon substrates were positioned
with a slight inclination in the aqueous subphase before spreading. The
monolayer was transferred by removal of the subphase from the trough
with a peristaltic pump (Minipuls 3, Abimed, Gilson Inc., United States)
F12H20, and at 35ꢀ37 Å2 moleculeꢀ1 for F11H1-core-H12.
3
A0 is the surface area at zero surface pressure extrapolated from
the slope between 2 and 6 mN mꢀ1. A stable monolayer is
3
with a pump rate of approximately 10 mL minꢀ1, thereby lowering the
formed up to the limiting area (collapse area) Ac = 31ꢀ33
3
Å2 moleculeꢀ1 for F12H12, Ac = 25ꢀ28 Å2 moleculeꢀ1 for
monolayer onto the wafer.25
3
3
F12H20, and Ac = 33ꢀ34 Å2 moleculeꢀ1 for F11H1-core-H12.
Monolayers without compression for SFM imaging were prepared by
direct evaporation (overnight) of one drop (submonolayer coverage) of
3
The isotherms were completely reversible, without hysteresis, for
compressionꢀexpansion cycles before the collapse of the mono-
layer, as we have recently shown.31 The measured Ac and A0
values for F12H12 are similar to those found in the literature for
other semifluorinated alkanes (F8H16 for example in ref 5),
which are consistent with the cross section (projection area) of a
fluorocarbon chain.
the amphiphobe solution (c = 1 mg mLꢀ1) spread on an aqueous film on
3
a freshly cleaned wafer.
2.3. SFM Measurements. The transferred films were analyzed
with a scanning force microscope (Dimension 3100, Veeco Instruments,
United States, and MFP-3D, Asylum Research, United States) in tapping
mode, equipped with sharpened, aluminum backside coated silicon
8778
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
Figure 2. Apparent surface pressure versus molecular area isotherms for
F12H20 (light gray), F12H12 (dark gray), and F11H1-core-H12 (black)
monolayers at the air/water interface at 20 °C.
After the collapse of the F12H12 monolayer, the surface
pressure remains almost constant at 10 mN mꢀ1 down to an
3
area of 13 Å2 moleculeꢀ1. The F12H20 and F11H1-core-H12
3
molecules show similar behavior, but with a slight increase in
pressure and slope in the “plateau region”. Then, in all cases, a
second steep pressure increase occurs at an area of ca. 10
Å2 moleculeꢀ1, which is approximately one-third of the mono-
Figure 3. (A) SFM topographic image of an F12H12 monolayer
3
transferred at 2 mN mꢀ1. (B) Zoom-in of the topographic image.
layer collapse area Ac. A similar behavior was reported recently by
Gracia Lux et al.32 for several semifluorinated alkanes with
different block lengths (F8H16, F8H18, F8H20, and F10H16),
which was explained by the buildup of a three-dimensional
aggregation system. The authors showed that the two- and
three-dimensional systems coexist in the plateau region.
3
Different micelle shapes and individual substructures marked by white
arrows are visible. (C) SFM topographic image of an F12H12 monolayer
deposited by direct evaporation of the subphase without compression.
The image shows closely packed micelles (dotted circle) as well as
disordered regions (dashed circle). (D) Line plot at the location marked
in image C. The plot shows the dimensions of one individual surface
micelle (height ∼ 2.2 nm, diameter ∼ 23 nm).
3.2. SFM Measurements. 3.2.1. F12H12 Monolayer Images. In
order to characterize the two-dimensional (2D) self-assembly
structure of the F12H12 monolayers, SFM measurements were
performed (Figure 3). Figure 3A shows the SFM topographic
image of a monolayer of F12H12 transferred to a silicon wafer at a
A peculiar feature at these surface pressures is the presence of
higher three-dimensional (3D) agreggates in the dense layer of
surface micelles (Figure 4), which do not show a specific contrast
in the surface potential measurements and have not been
previously reported in the literature. Up to the collapse, these
aggregates are virtually monodisperse (diameter ca. 40 nm;
thickness 4.3 nm with respect to the substrate).
lateral pressure of 2 mN mꢀ1. The image depicts an array of
3
surface micelles (circles, spirals, and wormlike structures) as
previously reported and rationalized for similar derivatives with
asymmetric fluorocarbon and hydrocarbon block lengths.5,33
These 2D micelles were observed at all surface pressures investi-
gated (from 1 mN mꢀ1 up to the collapse) with an average
In several regions distributed throughout the whole film, the
Kelvin probe measurements (image in Figure 4C and Figure S1
in the Supporting Information) reveal domains with lower
surface potential, which have only a faint contrast in the topo-
graphic (slightly brighter regions with higher layer thickness) and
the phase (slightly darker regions) images. These slightly higher
domain structures appear denser with less-defined micellar
boundaries. The corresponding more negative surface potential
difference is independent of the aggregates on top of the micelles,
and hints at a different structure/organization of the molecules in
these regions (further discussion in section 4.1).
3
diameter of 25ꢀ35nm. Their averagethicknesswas determined to
be 2.1 ( 0.1 nm by surface profile analysis. The higher-resolution
image in Figure 3B exhibits features within the surface micelles
(see white arrows), which suggests that the surface micelles are
composed of individual more-uniform substructures (an assembly
and structure model will be discussed further below; see also
Figure 11 below). Their diameter of 5ꢀ15 nm has been estimated
by measuring the distance between the centers of adjacent
substructures in the SFM images.
At 2 mN mꢀ1, the micelles are partially arranged in quite
3
close-packed hexagonal arrays (see dotted circle in Figure 3C).
According to the literature, the micelle size varies slightly for
At 10 mN mꢀ1, additional larger aggregates appear as flat
3
islands on top of the micelle carpet, which are apparently 3D
layered structures associated with the monolayer collapse
(Supporting Information, Figure S2). Similar aggregates have
been reported for F8H20 diblocks, at high surface pressures, after
the collapse of the monolayer.32
other SFA molecules in the range of 1ꢀ10 mN mꢀ1, but the
3
micelles progressively move closer together with increasing
surface pressure.6 However, for the F12H12 molecule in this
surface pressure regime, the relative compression (ΔA/A0) is
only ca. 7% and the applied surface pressure barely influences the
micelle packing and size. This hints at a dense micellar organiza-
tion which results in solidlike viscoelastic properties resembling
the behavior of a 2D jammed colloidal system (e.g., glass-like),34,35
as measured by interfacial linear rheology.31
In order to see if the micelles self-assemble without application
of lateral surface pressure by the trough barriers, monolayers
were also prepared by simple solution evaporation on a water film
on top of a wafer at very low surface coverage and observed by
SFM (Figure 3C) after drying. The surface micelles are still
8779
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
Figure 4. SFM images of an F12H12 film transferred at 5 mN mꢀ1. (A) Topographic image showing the micelle carpet and larger 3D aggregates
3
(average diameter 40 nm; average thickness with respect to the substrate 4.3 nm) and domains of slightly higher layer thickness. (B) Corresponding
phase image showing a slight phase contrast (0.5°) for the domains. (C) Corresponding Kelvin potential image of the same region with domains of lower
surface potential, while the 3D aggregates do not show a significant surface potential difference.
only the substrate is visible where holes in the monolayer are
present.
3.2.3. F11H1-core-H12 Monolayer Images. SFM images
obtained for an F11H1-core-H12 monolayer transferred at
2 mN mꢀ1 are presented in Figure 7. A different type of surface
3
micelles is observed, consisting of straight ribbons with an
approximate length of up to ca. 400 nm (or smaller, depending
on the monolayer preparation conditions, Figure 7A,C,D) and an
average width of 30ꢀ40 nm, similar to the diameter of the
circular micelles formed by the F12H12 molecules. As for
Figure 5. SFM images of an F12H20 monolayer transferred at 2
F12H12 and F12H20, the micelles appear close-packed even at
low surface pressures. Striations of ca. 10 nm width are visible
within the ribbons parallel to their longitudinal axis (Figure 7B),
again reminiscent of the substructure dimension in the surface
micelles of F12H12. The micelles themselves exhibit a fan-like
structure with micelles spreading and branching from an initial
branching point (Figure 14 and Figure S5 in the Supporting
Information). Such surface micelles also form a dense layer in the
absence of external surface pressure, as in the case of F12H12
(Figure 7C). A similar surface-micellar structure can be observed
on the film obtained without compression, after evaporation, as
seen in Figure 7A, which shows a close-packed organization of
straight ribbons, but with a less-pronounced dendritic growth.
Again, in interfacial rheology experiments the interdigitated
micellar organization was found to influence the monolayer
viscoelastic properties, which reflect a strongly correlated, self-
similar viscoelastic solid-like material.31,39,40 KPFM images of the
samples do not show any inhomogeneities in the micelle carpet
(Figure 7F), but show the potential difference between sample
and substrate.
mN mꢀ1. (A) Topography of a densely packed, partially hexagonally
3
organized monolayer. (B) Topographic image of a different region
showing hexagonal as well as more elongated surface micelles.
present, but alternating between dense zones and more disordered
ones (as already observed for F8H16 monolayers).33 In the dense
zones the micelles are already packed in hexagonal arrays even
without compression of the layer. Larger and denser domains in the
topographic and phase images are also visible in the Kelvin probe
microscopy measurements by a lower surface potential (Supporting
Information, Figure S3), while the higher 3D aggregates are
indiscernible by KPFM. The surface potential difference between
individual surface micelles and the surrounding surface area has also
been determined as ꢀ0.6 V for the edge of a film without
compression (see Figure S4 in Supporting Information). This value
is consistent with KPFM measurements in the literature.36ꢀ38
3.2.2. F12H20 Monolayer Images. F12H20 monolayers were
also imaged by SFM: the molecules self-assemble in a hexagonal
pattern of hexagonally shaped surface micelles, besides extended,
wormlike surface micelles, which also form in various regions
(Figure 5). Compared to F12H12, the micelles are slightly larger
(ca. 50 nm diameter), but no fine structure (primary aggregates)
is discernible. Each surface micelle appears in very close contact
with all neighboring micelles, inducing hexagonal deformation.
A high-resolution SFM imageof anF12H20filmtransferred at2
Besides this difference of micellar morphology depending on
the preparation conditions, we also found a substantial reorga-
nization of the monolayer structure on the solid support after
extended periods of time. An example of the morphology change
of an F11H1-core-H12 monolayer on a silicon substrate after 2
months is shown in Figure S6 in the Supporting Information.
Such long-term film reorganization is expected to be induced by
the difference in interfacial energy between the F11H1-core-H12
monolayer and the silicon substrate, which is different from the
water subphase.
mN mꢀ1, shown in Figure 6, also demonstrates well the coex-
3
istence of hexagonal micelles and elongated structures in a more
interlocked 2D system, but it does not show any fine structure
within the individual micelles. This change in surface morphology
from nearly circular to stronger interlocked and more extended
micelles explains the shift in the rheological behavior measured by
us:31 F12H20 monolayers show a self-similar response typical of
gel-like systems, apparently due to the junctions and interlocking
of the extended micelles.39,40
3.3. Neutron Reflectivity. Monolayers of F12H12 equili-
brated at seven different surface pressures (from 2 to 10 mN
m
3
ꢀ1) were investigated by neutron reflectivity measurements, as
shown in Figure 8 (full set of data in the Supporting Information,
Figure S7). In neutron reflectivity measurements, the variation of
the reflectivity with the momentum transfer Q depends on the
scattering length density profile normal to the interface.
The corresponding KPFM image (Figure 6C) does not
show any contrast within the dense surface micelle film, and
8780
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
Figure 6. Higher-resolution SFM and KPFM images of an F12H20 monolayer transferred at 2 mN mꢀ1. (A) Topographic, (B) phase, and (C) Kelvin
3
probe images.
Figure 7. SFM images of F11H1-core-H12 films. (A) Topographic image of an F11H1-core-H12 film transferred at 2 mN mꢀ1. (B) Line plot at the
3
location marked in image A. Blue arrows mark the edges between neighboring micelles, and red arrows mark substructures inside each micelle. (C)
Topographic image of an F11H1-core-H12 film prepared by direct evaporation of the subphase at a nominal surface pressure of 0 mN mꢀ1. (D)ꢀ(F)
3
SFM images of an F11H1-core-H12 film transferred at 5 mN mꢀ1: (D) topographic, (E) phase, and (F) Kelvin potential images.
3
Figure 8. Neutron reflectivity curves on D2O and corresponding πꢀA isotherms in the insets. Left: F12H12 monolayers at 2 mN mꢀ1 (blue circles), at
3
6 mN mꢀ1 (black triangles), and just before collapse (red squares). Right: F11H1-core-H12 monolayers at 10 mN mꢀ1 (black triangles) and at collapse
(red squares). The reflectivity curves for the neat D2O subphase are indicated by the asterisks (/).
3
3
Generally, oscillations (Kiessig fringes) may be visible in neutron
reflectivity profiles, depending on the thickness of the layer at the
air/water interface and with respect to the position and the
difference in scattering contrasts between the respective inter-
faces. For the particular F12H12 monolayers presented here, the
intensity of the neutron reflectivity curves decreases steeper with
increasing surface pressure, but no specific minimum is observed
(no Kiessig fringes visible), which may lead to difficulties in the
interpretation as discussed in section 4. However, neutron
reflectivity measurements present the advantage of investigating
the monolayer structure directly at the air/water interface with-
out transfer. Thus, even in the absence of characteristic minima,
fitting of the curves is attempted, taking into account different
possible packing models by variation of the structural parameters
8781
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
(number of layers, and for each layer, layer thickness, scattering
length density, and surface roughness).
These parameters can be estimated from data provided by
complementary techniques, such as the molecular area from
isotherms and the monolayer heights from SFM images:
• The scattering length density (SLD) of each compound is
estimated from ref 41 providing the bulk density of F12H12
(i.e., 1.6 g cmꢀ3 according to Nunez et al.17).
3
• The thickness of the monolayer can be calculated from a
simple bulk volume model as d = 25 Å, using the weight of
the deposited material, the bulk density of F12H12 mol-
ecules, and the molecular area in the isotherm at 5 mN mꢀ1
.
Figure 9. Different models for F12H12 structure at the air/water
interface (side view as cut through the film and the water surface). (a)
Horizontal (flat-lying molecules) (b) Vertical multilayer. (c)ꢀ(e)
Monolayer with vertical orientation: (c) F-block up, (d) H-block up,
and (e) head-to-tail orientation (left, antiparallel; right, random). The
white rectangles represent the F-blocks and the black ones represent the
H-blocks (for chemical structures, refer to Figure 1).
3
Moreover, based on a molecular structure model,12 a fully
stretched molecule with all-trans conformation will have the
fully extended length of d(FnHm) = 1.3n þ 1.265m þ 2.58
Å, which leads to a length of 33.4 Å for F12H12, consistent
with the computer model. Those two values (calculated
from the isotherm and the fully extended model) provide a
reasonable range of possible layer thicknesses. The thickness
estimated from the SFM measurements is around 21 Å.
Assuming the error in estimating the height for soft materi-
als is in the range of 20ꢀ30%, we obtain a corrected
thickness of 25ꢀ27 Å.
In the dense layer below the collapse surface pressure, the
existence of multilayer structures is excluded based on the height
data from the SFM measurements and the molecular surface
requirement from the isotherms. The area data provide direct
information on the film geometry and volume, which is consistent
with the deposited material volume in the spreading solution: no
material is lost during spreading and compression. The measured
• The molecular roughness of the subphase can be estimated
between one and two diameters of a water molecule, so
approximately 3ꢀ6 Å. A fit of the reflectivity curve obtained
for the neat D2O surface indeed provides a value of 5 Å for
the subphase roughness. Then the roughness of the different
layers is assumed to be above twice this value (roughness of
the subphase þ specific roughness of the layer). Consis-
tently, the SFM measurements provide a minimum rough-
ness value of 15 Å, which is used as a first guess for the fitting
parameter.
For F11H1-core-H12, the scattering profiles for the two surface
pressures are qualitatively similar (Figure 8, right). The reflectivity
curves show one reflectivity minimum at ca. Q = 0.07 Åꢀ1 and a
curve shape (Kiessing fringes) characteristic of a monolayer. The
same considerations (based on isotherms and SFM data) are used
to estimate the possible range of parameters (scattering length
density, roughness, and thickness). For the layer thickness, the
bulk estimation from the isotherm data provides an approximate
value around d = 31 Å. The fully extended length would be 39.4 Å,
assuming an approximate height of the core around 6 Å
(according to a computer-generated molecular structure model).
The thickness estimated from the SFM measurements is ca. 22 Å,
which here again must be considered with caution. It is possible
that flat-lying molecules cover the substrate in the “empty” zones
between the surface micelles, which are used as a reference level
for the height measurements. This value is clearly underestimat-
ing the extended chain length; thus we assume the possible layer
thickness to be in the range of 30ꢀ40 Å based on the isotherm
data and the fully extended chain model.
surface area of ca. 35 Å2 moleculeꢀ1 corresponds to the projec-
3
tionareaof the perfluoroalkyl chain along the molecular mainaxis.
The data suggest a monolayer with a vertical chain orientation
with the long molecular axis essentially oriented perpendicular to
the water surface. A flat phase (Figure 9a) may also be present,
coexisting with a vertical phase as proposed by Semenov et al.33
However, at the surface pressures investigated, this phase cannot
be predominant according to the measured molecular surface
area, and would not be detected in neutron reflectivity experi-
ments due to a too low thickness for the given scattering length
density. Indeed, the SFM topographic images indicate that the
substrate surface, which is visible between the surface micelles, is
partially covered with such flat-lying molecules. The surface
roughness of the free regions is enhanced compared to a blank
siliconwafer. Another hypotheticalcase of a highlybent molecular
structure with both the F- and the H-chains pointing toward the
air phase could be envisioned, where the polar CF2ꢀCH2ꢀ bond
would be in contact with the water phase. This model would
require a very strong deformation of the molecule with an
unfavorable proximity of the F- and H-chains, which is not
supported at all by the molecular surface requirement in the
isotherms and the layer height in the SFM images. This structure
is thus fully excluded from the discussion here.
Considering a monolayer structure with vertically oriented
molecules, three different packing models can be proposed
(Figure 9cꢀe). The fluorocarbon chains are a priori expected
to orient toward the air (Figure 9c), rather than toward the
aqueous phase, because of their larger hydrophobicity and higher
affinity for gases compared to hydrocarbon chains, which results
in a lower surface energy of the formed F-surface of the
monolayer. This configuration has been established for F8H16
by X-ray reflectivity.5 However, molecular dynamics simulations
of F12H18 molecules support the existence of antiparallel
molecular packing at short times (Figure 9e),42 which was also
proposed for F8H18 films based on X-ray reflectivity data,10 but
this remains a matter of discussion.2
For the asymmetric SFA F12H20, no neutron reflectivity data
were recorded due to the lack of resources (sample unavailability
at the time of the measurements).
4. DISCUSSION
4.1. Structural Model for F12H12. For semifluorinated
alkanes spread at the air/water interface several possible models
can be proposed, as presented in Figure 9.
8782
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
Figure 10. (a) Evolution of the scattering length density (SLD) for “semi-bilayer” models: (violet) F-block toward the air with a thickness of 14 Å and a
roughness of 15 Å for both half-layers; (orange) H-block toward the air with a thickness of 14 Å and a roughness of 15 Å for both half-layers. (b) Neutron
reflectivity curve (black squares) for F12H12 at 4 mN mꢀ1 and fits for both models: (violet) F-block toward the air (background 0.5 ꢁ 10ꢀ5); (orange)
3
H-block toward the air (background 1.5 ꢁ 10ꢀ5).
The fitting of the neutron reflectivity curves is used as a basis to
discriminate the different possible structural models cꢀe. Due to
the shape of the curves (absence of fringes, mentioned in section
3.3), several solutions can exist for the fitting of one curve. However,
an important constraint on the selection of model parameters is that
they should be physically reasonable based on the data from
complementary techniques. Providing the parameters discussed
in section 3.3 (thickness of the monolayer estimated between 25
and 33 Å, and roughness ca. 15 Å), three different models cꢀe in
Figure 9 have been tested for the curve simulation.
Hypothetical Model e: Assuming alternating head-to-tail
orientation, model e, the best fitting (with the best fit quality
and the most reasonable parameters) would be obtained for a
hypothetical thickness around 20 Å, with an average scattering
length density and a roughness around 15 Å. From an experi-
mental point of view this assumption would require that ca. 20%
of the material is lost after spreading, in order to yield the
measured molecular surface requirement in the isotherm, which
is not supported by experimental observations. Moreover, such a
molecular orientation would not lead to a lateral packing strain
(and associated layer curvature, see discussion of Figure 11
below) and thus should not induce any surface micelle formation,
but rather pack homogeneously in extended domains.
Hypothetical Models c and d: On the contrary, the data can
be fitted with a two-layer model (Figure 9c,d) with parameters
within the estimated ranges from the experimental results. It
appears then that the most accurate model that can be proposed
consists of a monolayer with parallel molecules, but as shown in
Figure 10, it is not possible to discriminate only with the neutron
reflectivity data between model c with F-block up, and model d
with H-block up: both models allow reasonably accurate curve
simulations.
In order to gain some insight on the molecular orientation of
the molecules, KPFM was employed. The KPFM technique is
sensitive to the effective dipole moment parallel to the tipꢀsur-
face axis (i.e., the surface normal) and the permittivity of the
material between the tip and the substrate.43 In the semifluori-
nated alkanes the major contributions to the effective dipole are
the molecular fragment dipoles of the CF3ꢀ group and the
central ꢀCF2ꢀCH2ꢀ bond.36 El Abed et al.10 and Broniatowski
et al.44 investigated the macroscopic surface potential of FnHm
molecules at the air/water interface. Both attributed a negative
surface potential to the fluorinated part of the respective mol-
ecules being oriented toward the air. KPFM studies and related
techniques on molecules with FnHm fragments have been
employed by Chi et al.,45 Sugimura et al.,37 Alexander et al.,36
and Magonov et al.,46 who interpreted differences in surface
potential in terms of ꢀCH2ꢀCF2ꢀ dipole orientation. A surface
potential difference at the air/water interface of ꢀ0.8 V has
been proposed for a vertical molecular orientation with the fluori-
nated part of the molecules oriented toward the air (or for
transferred layers ꢀ0.6 to ꢀ0.7 V relative to a Si wafer surface).36,38
Changes in the surface potential properties can have several
possible explanations. The presence of included water would lead
to slight changes in topographic, viscoelastic, and electric proper-
ties. A difference in the micellar density or a change in their
orientation would also induce modifications of the surface poten-
tial, as all three molecules F12H12, F12H20, and F11H1-core-H12
exhibit a strong dipole moment.
As observed in section 3.2.1, the majority of the F12H12
monolayer area does not show any local potential inhomogene-
ities in the Kelvin probe measurements on the micelle carpet: in
these regions the molecule distribution must be homogeneous at
least at the length scale of the KPFM lateral resolution (ca.
30 nm).27 A surface potential difference of about ꢀ0.6 V with
respect to the silicon substrate was measured for individual
micelles (see Supporting Information, Figure S4), in agreement
with Kelvin potential variations of F12H8 and F14H20 molecules
measured on a Si-wafer substrate.36ꢀ38 As a consequence, we
conclude that the F12H12 molecules inside the micelle carpet are
oriented vertically with their fluorocarbon part facing air. This
would be in good agreement with the literature data:2 the
molecular orientation is similar to the one established by X-ray
reflectivity measurements for F8H16 monolayers,5 which present
the same lateral organization with hexagonal arrays of disklike
surface micelles. Several domains with a more negative surface
potential with respect to the micelle carpet could also be
observed in some regions (see Figure 4C). It appears that the
surface micelles are packed more densely in these domains,
which is also supported by the slightly increased height in the
topographic images and the reduced intramicellar contrast. Such
a denser packing could induce a more vertical orientation of the
molecular axis and thus of the associated molecular dipole
ꢀCF2ꢀCH2ꢀ, which would result in a more negative surface
potential. A second possibility is the existence of domains
consisting of molecules with a different orientation (molecules
upside down or in an alternating packing), but this orientation is
expected to lead to a smaller surface potential difference. The
8783
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
most probable scenario is a denser packing of the monolayer, but
at the current state we cannot distinguish between these possi-
bilities with the SFM and KPFM data.
enabled by aggregate tilting toward neighboring aggregates and
the cluster center. This tilting will minimize contact with
surrounding aggregates outside the cluster and explains the
appearance of the surface micelles instead of a simple extended
hexagonal packing of the primary aggregates (Figure 11B). Slight
deformation in the contact regions between the primary aggre-
gates can occur, due to the mobility of the F12H12 molecule and
the flexibility of the aggregates.
This model is in accordance with other molecular assembly
models proposed in the literature.5,6,33 Both the micelles and
their arrangements are stabilized by a balance of factors that
minimize the system free energy: electrostatic interactions due to
dipole moments of fluorocarbons, interfacial tension, and hydro-
phobic interactions.33
The parallel orientation is apparently driven by phase separa-
tion between the F- and H-blocks. Considering this molecular
orientation, a simple model for the surface micelle formation can
be proposed as a hypothesis (Figure 11). The buildup of
isotropic packing strain due to the volume mismatch between
the thick F-block and the thin H-block leads to the formation of
“muffin” shaped structures (Figure 11A): the molecules aggre-
gate into a curved monolayer which is counteracted by the planar
water surface. The increasing packing strain between the layer
curvature and the planarizing water surface tension, as well as
electrostatic interactions (unfavorable orientation of the
dipoles), limit the size of the primary aggregates. These primary
aggregates, corresponding to the substructures observed in SFM,
assemble to bigger clusters containing on average five units
(spirals, circular, or elongated surface micelles, also discussed
in the literature) due to attractive capillary forces.47ꢀ49 As the
primary aggregates have inclined walls (due to the packing
strain/layer curvature), optimal contact between aggregates is
The orientation and packing structure of F12H20 molecules is
considered to be similar to that of F12H12, based on the
literature data, the analogous behavior in isotherms, and the
similar micellar organization imaged by SFM.
4.2. Structural Model for F11H1-core-H12. Considerations
similar to those for the F12H12 molecules can be proposed for the
packing models of F11H1-core-H12, consisting of monolayers
with vertical orientation of the F-block or H-block pointing
upward, or with antiparallel molecular orientation (alternating or
random). Here again neutron reflectivity measurements were used
to discriminate the different molecular orientations. A procedure
similar to that previously reported for F12H12 can be applied for
the core-extended molecule F11H1-core-H12. The parameters,
possible thickness and roughness, are discussed in section 3.3.
The different possible models were used to fit the neutron
reflectivity curves. For the hypothetical “head-to-tail”configura-
tion, two models can be anticipated with either isotropic scattering
length density (model A in Figure 12) or a well-defined aromatic
core position as intermediate layer (model B in Figure 12).
Figure 13 depicts the best fitting that can be obtained for each
model. It appears that only a structural model consisting of
molecules with mixed/alternating orientation can be used. Both
mixed-orientation models A and B (from Figure 12) lead to an
accurate simulation, providing layer thicknesses of 40 and 38 Å,
respectively. As shown in the inset in Figure 13b, it is not possible
to further discriminate these two models due to the poor
measurement statistics obtained at high momentum transfer (Q).
From the fitting results the molecular orientation appears to be
different from that of the F12H12 molecule as a consequence of
the central aromatic core (PhBr2).
Figure 11. Model proposed for the hierarchical formation of the surface
micelles. (A) Self-assembly of the amphiphobic F12H12 molecules in
parallel orientation with the F-block pointing up to the air into curved
primary aggregates with ca. 10 nm diameter. (B) Secondary assembly
(side view) of the primary aggregates into clusters of the surface micelles
driven by capillary forces. (C) Surface micelle aggregation (top view)
with slight deformation of the primary aggregates and comparison to the
SFM image details.
A second noticeable difference between the two types of
amphiphobes is the surface micelle shape as observed in section
Figure 12. Models for antiparallel orientation (left: alternating and random chain packing; for chemical structures, refer to Figure 1). Model A: globally
averaged scattering length density. Model B: triple layer with the upper and lower parts having the same average scattering length density (identical to an
averaged F12H12) and the center region corresponding to a well-defined aromatic core layer.
8784
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
Figure 13. (a) Evolution of the scattering length density (SLD) for different models. Antiparallel packing: model A (green line), thickness of 40 Å and
roughness of 7 Å; model B (blue line), thickness of 16.5 Å and roughness of 5 Å for both the upper and lower alkyl chain regions (F- and H-blocks mixed),
thickness of 5 Å and roughness of 5 Å for the core layer. Parallel packing: three-layer model with F-block up (violet dashed line), thickness of 15 Å and
roughness of 5 Å for both H-block and F-block layers and thickness of 5 Å and roughness of 5 Å for the core layer; three-layer model with H-block up
(orange dashed line), thickness of 15 Å and roughness of 10 Å for both H-block and F-block layers and thickness of 5 Å and roughness of 5 Å for the core
layer. (b) Neutron reflectivity curve at 10 mN mꢀ1 for F11H1-core-H12 (black squares) and fits for the different models proposed: (green line)
3
antiparallel model A (background 8 ꢁ 10ꢀ6), random model (blue line) B (background 1 ꢁ 10ꢀ5), and F-block up and down without background
(violet and orange dashed lines as in (a)).
expected to limit the width of the surface micelle along the
short axis.
5. CONCLUSION
The original 2D structure of non-classical amphiphobic mol-
ecules with symmetric F- and H-block lengths was studied by
combination of neutron reflectivity and scanning force micro-
scopy, coupled with Kelvin probe force microscopy measure-
ments. As reported in the literature for asymmetric block lengths
and shown here for F12H20, both molecules with symmetric
Figure 14. Left: Model proposed for the hierarchical packing structure
of the F11H1-core-H12 molecules into a ribbon element (yellow-green
block). Center: SFM image with observed dendritic ribbon structure.
chain length (F12H12 and its core extended analogue F11H1-
core-H12) form surface micelles when spread at the air/water
interface. The symmetry does not appear to play a crucial role as
the F12H12 molecules were found to behave like other asym-
metric derivatives, e.g., F12H20. With the molecular main axis
perpendicular to the air/water interface and the F-block pointing
toward the air, they self-organize into 30 nm disk-like surface
micelles (or spirals slightly extended), composed of circular
10 nm substructures size-controlled by free-energy minimization.
Analogous micellar organization was found for F12H20 mol-
ecules, but with a more close packed self-assembly, which induces
the coexistence of hexagonally deformed micelles with inter-
locked extended ones. Conversely, the phenyl ring insertion
induces dramatic structural changes with respect to the probable
molecule orientation and the resulting micelle shape, consisting
of branched elongated structures, but which again feature a
10 nm striation substructure (which appears to be a more general
structure motif). These results further illustrate how chemical
modifications can be used to tailor the hierarchical self-assembly
of such semifluorinated alkanes and to extend their possible
applications for driving supramolecular assembly. The changes in
micellar organization, i.e., the micelle shape, but also their close
packing, are strongly reflected in the viscoelastic properties of the
monolayer.31 A rigorous structural characterization complemen-
ted with interfacial rheological measurements is suggested as a
powerful methodology to yield detailed insight into the mono-
layer characteristics for the rational design of novel molecular
architectures in order to target the desired film properties.
Right: Line illustration of a single dendritic domain with a fingering of
the ribbons.
3.2. The F11H1-core-H12 self-assembles into ribbons with a
length about 10 times larger than the average diameter of the
circular micelles observed for F12H12 monolayers. The aromatic
core induces a structural as well as a dipole asymmetry in the
molecule, compared to the rotational symmetry of the F12H12
along the molecular main axis, which induces a strong packing
anisotropy. While the antiparallel orientation of the molecules is
suggested by the fitting of the neutron reflectivity data, such
packing should not lead to a strain buildup by volumetric
mismatch if only the F- and H-blocks would be considered.
Extended monolayer domains would be expected for this mole-
cule type, ignoring the anisotropic core structure. Since well-
defined surface micelles exist, obviously, a different mechanism
limiting the layer extension due to the presence of the aromatic
core in the layer must be present, which along the direction of
the phenyl face or the phenyl edge leads to a packing struc-
ture with elongated micelles. The packing structure model in
Figure 14 may be proposed as a hypothesis. The aromatic core
has about the same thickness as the alkyl chain, so no distortion,
but rather a stabilization of the alkyl packing, is expected along
the stacking direction of the cores, which is expected to be the
long axis of the elongated surface micelle. On the other hand,
the core is much wider than the cross section of the alkyl chain; thus
along the direction of edge-on stacking a strong packing mismatch is
8785
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786

Langmuir
ARTICLE
’ ASSOCIATED CONTENT
(24) Als-Nielsen, J. In Handbook on Synchrotron Radiation; Brown,
G., Moncton, D. E., Eds.; North Holland: Amsterdam, 1991; Vol. 3, pp
471ꢀ504.
S
Supporting Information. Additional SFM images of
b
(25) Vogel, N.; de Viguerie, L.; Jonas, U.; Landfester, K.; Weiss,
C. K. Adv. Funct. Mater. 2011, DOI: 10.1002/adfm.201100414.
(26) Magonov, S. N. In Encyclopaedia of Analytical Chemistry;
Meyers, R. A., Ed.; John Wiley & Sons Ltd: Chichester, U.K., 2000;
pp 7432ꢀ7491.
F12H12 and F11H1-core-H12 films; neutron reflectivity curves
for F12H12 monolayers at the air/water interface. This material
is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
(27) Berger, R.; Butt, H. J.; Retschke, M. B.; Weber, S. A. L.
Macromol. Rapid Commun. 2009, 30, 1167–1178.
(28) Gupta, M.; Gutberlet, T.; Stahn, J.; Keller, P.; Clemens, D.
Pramana 2004, 63 (1), 57–63.
Corresponding Author
*E-mail: ujonas@iesl.forth.gr.
(29) Clemens, D.; Gross, P.; Keller, P.; Schlumpf, N.; Koennecke,
M. Physica B 2000, 140, 276–278.
’ ACKNOWLEDGMENT
(30) Braun, C. Parratt32 Fitting Routine for Reflectivity Data; HMI:
Berlin, 1997ꢀ1999.
The authors thank the EU for financial support via the projects
ITN COMPLOIDS (Grant FP7-PEOPLE-ITN-2008-234810)
and Nanodirect (Grant CP-FP 213948). We thank J. Ochsmann
for assistance with X-ray experiments, S. Weber and M. Retschke
for assistance with KPFM measurements and valuable discus-
sions, and B. Loppinet for many fruitful discussions.
(31) Klein, C. O.; de Viguerie, L., Christopoulou, C.; Jonas, U.;
Clark, C. J.; Muellen, K.; Vlassopoulos, D. Soft Matter 2011, in press;
DOI:10.1039/C1SM05357D.
(32) Gracia Lux, C. de; Gallani, J. L.; Waton, G.; Krafft, M. P. Chem.
—Eur. J. 2010, 16, 7186–7198.
(33) Semenov, A. N.; Gonzalez-Perez, A.; Krafft, M. P.; Legrand, J. F.
Langmuir 2006, 22 (21), 8703–8717.
(34) Ozon, F.; Petekidis, G.; Vlassopoulos, D. Ind. Eng. Chem. Res.
2006, 45 (21), 6946–6952.
’ REFERENCES
(35) Di Cola, E.; Moussaid, A.; Sztucki, M.; Narayanan, T.; Zaccarelli,
E. J. Chem. Phys. 2009, 131 (14), 144903.
(36) Alexander, J.; Magonov, S.; Moeller, M. J. Vac. Sci. Technol., B
2009, 27, 903–911.
(37) Sugimura, H.; Hayashi, K.; Saito, N.; Nakagiri, N.; Takai, O.
Appl. Surf. Sci. 2002, 188, 403–410.
(38) Magonov, S.; Alexander, J. Beilstein J. Nanotechnol. 2011, 2,
15–27.
(39) Romeo, G.; Filippone, G.; Fernꢀandez-Nieves, A.; Russo, P.;
Acierno, D. Rheol. Acta 2008, 47 (9), 989–997.
(40) Shah, S. A.; Chen, Y.-L.; Schweizer, K. S.; Zukoski, C. F. J. Chem.
Phys. 2003, 119 (16), 8747–8760.
(1) Gaines, G. L. Langmuir 1991, 7 (12), 3054–3056.
(2) Krafft, M. P.; Riess, J. G. Chem. Rev. 2009, 109 (5), 1714–1792.
(3) Riess, J. G. Curr. Opin. Colloid Interface Sci. 2009, 14, 294–304.
(4) Broniatowski, M.; Dynarowicz-Latka, P. Adv. Colloid Interface Sci.
2008, 138, 63–83.
(5) Maaloum, M.; Muller, P.; Krafft, M. P. Angew. Chem., Int. Ed.
2002, 41 (22), 4331–4334.
(6) Zhang, G. F.; Marie, P.; Maaloum, M.; Muller, P.; Benoit, N.;
Krafft, M. P. J. Am. Chem. Soc. 2005, 127 (29), 10412–10419.
(7) Jacquemain, D.; Grayer Wolf, S.; Leveiller, F.; Lahav, M.;
Leiserowitz, L.; Deutsch, M.; Kjaer, K.; Als-Nielsen, J. J. Am. Chem.
Soc. 1990, 112 (21), 7724–7736.
(41) National Institute of Standards and Technology Center for
Neutron Research. Scattering Length Density Calculator; http://www.
ncnr.nist.gov/resources/sldcalc.html (accessed SeptemberꢀDecember
2010).
(8) Grayer Wolf, S.; Deutsch, M.; Landau, E. M.; Lahav, M.;
Leiserowitz, L.; Kjaer, K.; Als-Nielsen, J. Science 1988, 242, 1286–1290.
(9) Zhang, G. F.; Maaloum, M.; Muller, P.; Benoit, N.; Krafft, M. P.
Phys. Chem. Chem. Phys. 2004, 6, 1566–1569.
(42) Kim, N.; Shin, S. J. Chem. Phys. 1999, 110, 10239–10242.
(43) Taylor, D. M.; Bayes, G. F. Phys. Rev. E 1994, 49 (2), 1439–1449.
(44) Broniatowski, M.; Minones, J.; Dynarowicz-Latka, P. J. Colloid
Interface Sci. 2004, 279, 552–558.
(10) El Abed, A.; Faurꢀe, M. C.; Pouzet, E.; Abillon, O. Phys. Rev. E
2002, 65, 051603.
(11) Broniatowski, M.; Sandez Macho, I.; Minones, J., Jr.; Dynarowicz-
ya€utka, P. J. Phys. Chem. B 2004, 108, 13403–13411.
(12) Lo Nostro, P. Curr. Opin. Colloid Interface Sci. 2003, 8, 223–226.
(13) Hoepken, J.; Pugh, C.; Richtering, W.; Moeller, M. Makromol.
Chem. 1988, 189, 911–925.
(45) Chi, L. F.; Jacobi, S.; Fuchs, H. Thin Solid Films 1996, 284ꢀ
285, 403–407.
(46) Magonov, S.; Alexander, J.; Jeong, S.-H.; Kotov, N. J. Nanosci.
Nanotechnol. 2010, 10 (11), 7060.
(14) Rabolt, J. F.; Russell, T. P.; Twieg, R. J. Macromolecules 1984, 17
(47) Kralchevsky, P. A.; Nagayama, K. Langmuir 1994, 10 (1), 23.
(48) Kralchevsky, P. A. Adv. Colloid Interface Sci. 2000, 85, 145–192.
(49) Kralchevsky, P. A. Curr. Opin. Colloid Interface Sci. 2001, 6 (4),
383–401.
(12), 2786–2794.
(15) Russell, T. P.; Rabolt, J. F.; Twieg, R. J.; Siemens, R. L.; Farmer,
B. L. Macromolecules 1986, 19, 1135–1143.
(16) Hoepken, J.; Moeller, M. Macromolecules 1992, 25, 2482–2489.
(17) Nunez, E.; Clark, C. G.; Cheng, W.; Best, A.; Floudas, G.;
Semenov, A. N.; Fytas, G.; Muellen, K. J. Phys. Chem. B 2008, 112 (21),
6542–6549.
(18) Lee, Y. J.; Clark, C. G., Jr.; Graf, R.; Wagner, M.; Muellen, K.;
Wolfgang Spiess, H. W. J. Phys. Chem. B 2009, 113, 1360–1366.
(19) Clark, C. G., Jr.; Floudas, G. A.; Lee, Y. J.; Graf, R.; Spiess,
H. W.; Muellen, K. J. Am. Chem. Soc. 2009, 131, 8537–8547.
(20) Ruiz, J. P.; Dharia, J. R.; Reynolds, J. R.; Buckley, L. J. Macro-
molecules 1992, 25, 849–860.
(21) Nenov, S.; Clark, C. G., Jr.; Klapper, M.; Muellen, K. Macromol.
Chem. Phys. 2007, 208, 1362–1369.
(22) Baskar, C.; Lai, Y. H.; Valiyaveettil, S. Macromolecules 2001,
34, 6255–6260.
(23) Kandre, R. M.; Kutzner, F.; Schlaad, H.; Schl€uter, A. D.
Macromol. Chem. Phys. 2005, 206, 1610–1618.
8786
dx.doi.org/10.1021/la201377f |Langmuir 2011, 27, 8776–8786