Halogen bonding versus hydrogen bonding in driving self-assembly and performance of light-responsive supramolecular polymers

Article abstract of DOI:10.1002/adfm.201200135

Halogen bonding is arguably the least exploited among the many non-covalent interactions used in dictating molecular self-assembly. However, its directionality renders it unique compared to ubiquitous hydrogen bonding. Here, the role of this directionality in controlling the performance of light-responsive supramolecular polymers is highlighted. In particular, it is shown that light-induced surface patterning, a unique phenomenon occurring in azobenzene-containing polymers, is more efficient in halogen-bonded polymer-azobenzene complexes than in the analogous hydrogen-bonded complexes. A systematic study is performed on a series of azo dyes containing different halogen or hydrogen bonding donor moieties, complexed to poly(4-vinylpyridine) backbone. Through single-atom substitution of the bond-donor, control of both the strength and the nature of the noncovalent interaction between the azobenzene units and the polymer backbone is achieved. Importantly, such substitution does not significantly alter the electronic properties of the azobenzene units, hence providing us with unique tools in studying the structure-performance relationships in the light-induced surface deformation process. The results represent the first demonstration of light-responsive halogen-bonded polymer systems and also highlight the remarkable potential of halogen bonding in fundamental studies of photoresponsive azobenzene-containing polymers. Copyright

Full text of DOI:10.1002/adfm.201200135

www.afm-journal.de
www.MaterialsViews.com
Halogen Bonding versus Hydrogen Bonding in Driving
Self-Assembly and Performance of Light-Responsive
Supramolecular Polymers
Arri Priimagi,* Gabriella Cavallo, Alessandra Forni, Mikael Gorynsztejn–Leben,
Matti Kaivola, Pierangelo Metrangolo,* Roberto Milani, Atsushi Shishido, Tullio Pilati,
Giuseppe Resnati,* and Giancarlo Terraneo
1. Introduction
Halogen bonding is arguably the least exploited among the many non-
covalent interactions used in dictating molecular self-assembly. However, its
Light-induced surface patterning is
a
unique feature of azobenzene-containing
materials, in which the photoisomeriza-
tion of the azobenzene chromophores
initiates mass transport over micrometer
distances.[1,2] This simple, all-optical, and
reversible process requires only a single
fabrication step to inscribe high-quality
surface relief gratings (SRGs), showing
potential for numerous applications in
photonics and nanotechnology.[2–6] The
grating formation has been studied in
directionality renders it unique compared to ubiquitous hydrogen bonding.
Here, the role of this directionality in controlling the performance of light-
responsive supramolecular polymers is highlighted. In particular, it is shown
that light-induced surface patterning, a unique phenomenon occurring in
azobenzene-containing polymers, is more efficient in halogen-bonded pol-
ymer–azobenzene complexes than in the analogous hydrogen-bonded com-
plexes. A systematic study is performed on a series of azo dyes containing
different halogen or hydrogen bonding donor moieties, complexed to poly(4-
vinylpyridine) backbone. Through single-atom substitution of the bond-donor,
control of both the strength and the nature of the noncovalent interaction
between the azobenzene units and the polymer backbone is achieved. Impor-
tantly, such substitution does not significantly alter the electronic properties
of the azobenzene units, hence providing us with unique tools in studying the
structure–performance relationships in the light-induced surface deforma-
tion process. The results represent the first demonstration of light-responsive
halogen-bonded polymer systems and also highlight the remarkable potential
of halogen bonding in fundamental studies of photoresponsive azobenzene-
containing polymers.
various
azobenzene-containing
sys-
tems,[7–9] and in the past few years,
supramolecular functionalization strat-
egies have emerged as facile tools to
design photoresponsive polymers that
undergo efficient light-induced mass
transport.[10–12] Furthermore, the inherent
tunability of supramolecular chemistry
provides unprecedented possibilities for
fundamental studies of light-induced
mass transport in photoresponsive poly-
mers, the fundamental mechanism and
Dr. A. Priimagi, M. Gorynsztejn–Leben, Prof. M. Kaivola
Department of Applied Physics
Dr. A. Forni, Dr. T. Pilati, Prof. G. Resnati
ISTM-CNR
Aalto University
P.O. Box 13500, FI-00076 Aalto, Finland
E-mail: arri.priimagi@aalto.fi
Università degli Studi di Milano
Via Golgi 19, I-20133 Milano, Italy
Prof. P. Metrangolo, Dr. R. Milani, Prof. G. Resnati,
Dr. G. Terraneo
Center for Nano Science and Technology@Polimi
Istituto Italiano di Tecnologia
Dr. A. Priimagi, Prof. A. Shishido
Chemical Resources Laboratory
Tokyo Institute of Technology
R1-12 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan
via Pascoli 70/3, I-20133 Milano, Italy
Dr. G. Cavallo, Prof. P. Metrangolo, Dr. T. Pilati, Prof. G. Resnati,
Dr. G. Terraneo
NFMLab, DCMIC “Giulio Natta”
Prof. P. Metrangolo, Dr. R. Milani
VTT, Technical Research Centre of Finland
Tietotie 2, FIN-02044 VTT, Finland
Politecnico di Milano
Via Mancinelli 7, I-20131 Milano, Italy
E-mail: pierangelo.metrangolo@polimi.it; giuseppe.resnati@polimi.it
DOI: 10.1002/adfm.201200135
©
wileyonlinelibrary.com
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1

www.afm-journal.de
www.MaterialsViews.com
CH3
N
CH3
N
structure–function relationships of which
remain unsettled up to date.[13,14] Herein, we
show that halogen bonding is a viable tool in
designing high-performance supramolecular
polymers for light-induced surface patterning.
In fact, halogen bonding, because of its direc-
tionality and tuneable interaction strength,
provides unique possibilities for fundamental
studies of photoresponsive polymers, and hal-
ogen-bonded supramolecular polymers com-
pare favourably and may even outperform
analogous hydrogen-bonded systems in the
SRG formation efficiency.
CH2
CH
N
H3
C
H3C
F
F
N
N
F
F
X
Z
N
N
N
O
H
N
1a, 1b, 1c
I, B r,
3a, 3b
P4VP
2
Z
= C2H4, C3H6
X
=
H
Scheme 1. Chemical structures of the azobenzene derivatives 1a-c and 2 and the bis-pyridine
compounds 1,2-di(4-pyridyl)ethane (3a) and 1,3-di(4-pyridyl)propane (3b) selected as model
compounds for the polymer P4VP.
Halogen bonding[15] occurs between the positive region of the
electrostatic potential surface of halogen atoms, which function
as electrophilic species, and neutral or anionic Lewis bases.[16]
Among the many noncovalent interactions that are commonly
used to direct the formation of supramolecular assemblies, hal-
ogen bonding is arguably the least exploited, which is surprising
given its potentially powerful analogy to ubiquitous hydrogen
bonding.[17] In spite of this, halogen bonding has been convinc-
ingly used to control the self-assembly of diverse host-guest
solids, with applications ranging from liquid-crystalline,[18–21]
porous,[22,23] magnetic,[24] and organic phosphorescent mate-
rials,[25] to ion pair recognition,[26] biomolecular engineering,[27]
and chemical separation.[28] In particular, halogen bonding is
nowadays considered among the most valued tools in the field
of crystal engineering because of its high strength, specificity,
and directionality, which has allowed designing very complex
structures with a high degree of accuracy and precision.[29]
Examples of halogen bonding involving polymeric donors
and/or acceptors are extremely rare. In 2002 we reported
the self-assembly of poly(4-vinylpyridine) with various α,ω-
diiodoperfluoroalkanes into fluorous supramolecular poly-
mers.[30] Later on, applications of polymeric halogen-bonded
materials have appeared in the fields of molecularly imprinted
polymers,[31] topochemical polymerization,[32] and layer-by-layer
assembly.[33]
Given that hydrogen bonding has been widely and success-
fully exploited in the design of functional and stimuli-respon-
sive polymeric materials,[34–38] it appears reasonable to predict
that halogen bonding holds a great potential for the design of
a new class of analogous systems. Importantly, these two non-
covalent interactions also display significant differences. Firstly,
halogen bonding is more directional than hydrogen bonding[39]
and has already proven effective in the design and synthesis
of topologically complex structures.[40] Secondly, its interaction
strength can be fine tuned by properly choosing the halogen
atom that takes part in the bond formation without significantly
changing the electronic structure of the compound.[41]
to previously established phenol–pyridine hydrogen-bonded
systems, and showed that as a likely consequence of the higher
directionality of halogen bonding, halogen-bonded complexes
exhibit higher patterning efficiency than the hydrogen-bonded
systems. The study is complemented by theoretical calculations
and crystallographic investigations, which further enlighten the
different performances displayed by halogen- and hydrogen-
bonded systems.
2. Results
2.1. Systems Design
The similarities of hydrogen bonding and halogen bonding
allowed us to use the same polymer host, poly(4-vinyl pyridine)
(P4VP; Mw ≈ 1000 g mol−1 ), as a bond acceptor matrix for both
interactions. It has already been demonstrated that P4VP effi-
ciently self-assembles with hydrogen and halogen bond-donor
molecules.[30,38] In the chromophore design, we made use of
the unique possibility provided by halogen bonding to fine
tune the nature and strength of the polymer–dye interaction
by single halogen atom mutation on the tetrafluorobenzene
ring. We synthesized the azobenzene derivatives 1a-c shown
in Scheme 1. The derivatives 1a and 1b are expected to form
medium-to-weak halogen bonds with P4VP,[30,41] while 1c is
expected to form a medium-to-weak hydrogen bond with P4VP
because of the acidic hydrogen atom of the tetrafluorobenzene
ring. Hence the compound 1c functions as a control molecule
unable to form halogen bonds. Moreover, in order to compare
the performance of halogen bonding and hydrogen bonding in
SRG formation, the commercially available azobenzene deriv-
ative 2 was also chosen as a reference. Despite its somewhat
different chemical structure with respect to 1a-c, the phenol-
containing 2 serves as an excellent reference since supramo-
lecular polymers based on the phenol/pyridine supramolecular
synthon have already proven to be efficient for photoinduced
SRG formation,[42–45] and since its binding strength to P4VP is
still higher than that shown by the halogenated dyes (cf. below),
even when lacking fluorination of the benzene ring.
In the present work, the above-mentioned features of halogen
bonding are shown to provide unique possibilities in under-
standing and controlling the properties of stimuli-responsive
supramolecular polymers for light-induced surface patterning.
We have designed halogen-bonded supramolecular polymers,
which, by proper single atom mutation in the halogen bonding-
donor moiety, allowed us to probe the role of both the nature
and the strength of the noncovalent bonding on the SRG forma-
tion efficiency. We also compared the halogen-bonded systems
2.2. Theoretical Modelling
In order to validate our design assumptions and verify the
occurrence, nature, and strength of the involved noncovalent
©
2
wileyonlinelibrary.com
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135

www.afm-journal.de
www.MaterialsViews.com
The photoinduced surface patterning is
driven by trans–cis–trans photoisomerization
cycling of the azobenzene moieties, which
in turn is affected by their dipole moments
and spectroscopic properties.[49] The UV-vis
absorption spectra of the dyes 1a-c in dilute
dimethylformamide (DMF) solutions are
very similar (see Supporting Information,
Figure S2), and significantly red-shifted
when compared to the spectrum of 2. This
suggests that due to the strong electron-
withdrawing character of the fluorinated
aromatic ring, the dyes 1a–c all have similar
dipole moments, which are higher than that
of 2. This is supported by DFT calculations,
according to which the dipole moments
are 7.3, 7.5, 6.4, and 3.8 D in vacuum, and
10.2, 10.5, 8.9, and 5.7 D in DMF solu-
tion for 1a, 1b, 1c, and 2, respectively. The
absorption spectra were then computed by
time-dependent DFT, which reproduced the
experimentally observed trend with the max-
Figure 1. Plots of the electrostatic potential of the compounds 1a–c and 2. Potentials are
mapped on the respective isosurfaces (0.001 a.u.) of electron density. Values of electrostatic
potential range from –0.03 (red) to 0.03 (blue) a.u. Atom color scheme: C, gray; H, light gray;
N, dark blue; O, red; F, sky blue, I, magenta.
imum absorption wavelength decreasing in the order 1a ≅ 1b >
1c >> 2 (see Supporting Information, Table S1). We note that
the similar electronic structure of the dyes 1a-c is central for
the present study, allowing us to distinguish the roles of nature
and strength of the polymer–dye interaction from other factors
affecting the photoinduced surface patterning.
bonds, we modelled the P4VP–azobenzene systems using
density functional theory (DFT). We calculated the interaction
energies between the dyes in Scheme 1 and 4-methylpyridine,
which functions as a model compound for the polymeric elec-
tron donor. The interaction strengths were computed as the dif-
ference between the energy of the bonded dimers and the sum
of the energies of the single monomers. The energies calculated
for the dimers of the iodine- and bromine-containing azo-dyes
1a and 1b (ΔEBSSE = –5.135 and –3.501 kcal mol−1 , respectively)
are perfectly in line with those calculated for some analogous
pyridine–halotetrafluorobenzene complexes.[20,46] Similarly, the
interaction energies between 4-methylpyridine and the dyes 1c
and 2 (ΔEBSSE = –3.930 and –10.053 kcal mol−1 , respectively)
correspond to values typical of medium-to-strong hydrogen
bonds.[47]
The calculated interaction energies imply that the hydrogen-
bonded dimer of 2 with 4-methylpyridine is more stable than
the analogous halogen-bonded dimer formed with 1a. How-
ever, as brought out by Politzer,[39] halogen bonding is more
directional than hydrogen bonding. The reason for this is
illustrated in Figure 1: Both the iodine and hydrogen atoms
of the dyes 1a and 2 display positively charged outer regions,
which explains their capability to interact with negatively
charged sites. But while the dark blue positively charged
region is narrowly localized along the extension of the C–I
bond of the iodine atom of 1a, it is hemispherically distrib-
uted around the phenolic hydrogen atom of 2 and even spread
over the nearby hydrogen atom in the ortho position of the
phenol ring. Similarly, a narrower positively charged region is
found on the bromine atom of 1b compared to the hydrogen
atom in the para position of the tetrafluorobenzene ring of
1c, although both dyes were found to have similar interac-
tion energies with 4-methylpyridine. This is highly pertinent
because halogen bonding has been reported to successfully
compete with hydrogen bonding in building up supramo-
lecular architectures, even in spite of its lower or comparable
interaction strength.[48]
2.3. Single Crystal X-ray Diffraction Studies
To have a model of the P4VP–azobenzene systems, we co-crys-
tallized 1a–c and 2 with the bis-pyridine derivatives shown in
Scheme 1. Good-quality single crystals of the halogen-bonded
complex 4 were obtained upon evaporation of a 2:1 chloroform
solution of 1a and 3a. The differential scanning calorimetry
(DSC) study of those crystals revealed a unique sharp melting
endotherm at 465 K, which is higher than those of the starting
materials (1a: 456 K; 3a: 383 K), thus confirming the formation
of a new chemical species rather than a mechanical mixture of
the two starting compounds. The single crystal X-ray analysis of
4 revealed details of the supramolecular organization of the two
components in the crystal lattice (Figure 2, top).
The components are held together in trimers by halogen
bonds between the two nitrogen atoms of 3a and the iodine
atoms of two distinct 1a molecules, explaining the observed 1:2
ratio. The N···I distances are 2.807(2) Å and 2.818(2) Å, corre-
sponding to ca. 20% contraction with respect to the sum of the
van der Waals radii for N and I.[50] The two C–I···N angles are
173.47(7)° and 170.45(7)°, which are perfectly consistent with
the expected directionality of halogen bonds.[51]
The hydrogen-bonded complex 5 between 2 and 3b was iso-
lated as pale orange crystals with a melting point at 446 K, in
between those of the starting compounds (2: 479 K; 3b: 336 K).
The supramolecular structure of 5 is sustained by the N···H–O
interaction where 3b functions as a bidentate hydrogen bond
acceptor bridging two different molecules of 2. The N···O and
N···H distances are 2.703(2) Å and 1.73(3) Å, respectively,
©
wileyonlinelibrary.com
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3

www.afm-journal.de
www.MaterialsViews.com
bonds, it is reasonable to discuss vibrational
spectra of the complexes in terms of modi-
fied modes of the starting azobenzenes and
pyridyl moieties. Pure compounds 3a and 3b
show a symmetric ring stretching mode at
990 cm−1 , which exhibits a clear blue-shift of
9 cm−1 and 21 cm−1 once complexed with 1a
and 2, respectively. Similar shifts have been
attributed to a reduced electron density over
the N atom of the pyridine ring when it is
involved as a bond acceptor in the forma-
tion of hydrogen and halogen bonding.[52–54]
The larger the shift, the stronger the interac-
tion the pyridyl is involved in. The blueshifts
observed in the IR spectra of the complexes
4-6 suggest that the hydrogen bond given by
the azobenzene 2 is stronger than the hal-
Figure 2. Halogen and hydrogen bonding drive the self-assembly of the azobenzenes 1a and 2
and the bis-pyridine derivatives 1,2-di(4-pyridyl)ethane (3a) and 1,3-di(4-pyridyl)propane (3b)
into the trimeric supramolecular structures 4 and 5, respectively. Colors are as follows: C, gray;
H, light gray; N, sky blue; O, red; F, green; I, magenta.
ogen bond of 1a (see Supporting Information, Figure S3), thus
confirming the energy trend calculated for the 4-methylpyridine
dimers of the azobenzenes 1a–c and 2.
with an O–H···N angle of 168(2)° (Figure 2, bottom) and a
C–O···N angle of 117.70(9)°.
Based on the solid-state structures of the two model systems
4 and 5 described above, it is evident that the geometries of the
halogen-bonded and hydrogen-bonded complexes are quite dif-
ferent. The bis-pyridine molecule is almost collinear with the
halogen-bonded iodotetrafluorobenzene of 1a, as a consequence
of the fact that halogen bonding essentially occurs along the
extension of the C–I bond, i.e., the angle C–I···N is usually
close to 180° (see Supporting Information, Figure S9). Instead,
no such alignment is observed in the corresponding hydrogen-
bonded system (the angle C–O···N is usually close to 120°, see
Supporting Information, Figure S10).
We did not obtain good quality single crystals for the com-
plex between 1a and 3b. However, from a 2:1 acetone solution
of the starting components, we obtained the microcrystalline
powder 6, melting at 445 K. Also in this case, the melting tem-
perature is different from those of the two starting compounds
(1a: 456 K; 3b: 336 K), which again proves that 6 is not a simple
mechanical mixture of 1a and 3b. Moreover, the 1H-NMR anal-
ysis in chloroform solution confirmed the expected 2:1 ratio
between the starting compounds in the complex 6.
The dyes 1b and 1c did not afford azobenzene–bipyridine
complexes under the same co-crystallization procedures,
instead, 1b crystallized as a pure compound. This is consistent
with theoretical modelling that 1b and 1c are
2.4. Thin Film Characterization and Photoinduced
Surface Patterning
In order to study the SRG formation in halogen-bonded and
hydrogen-bonded complexes, we prepared dimethyl formamide
(DMF) solutions of the polymer–dye mixtures and spin-cast
them as thin films on silicon or glass substrates. In order to
prevent dye aggregation and phase segregation, we used rela-
tively low dye/polymer ratios of 0.1 (Figure 3) or 0.2 (Figure 4).
The thin films are denoted as P4VP(i)x, where i corresponds
to the dye molecule in question and x is the dye/pyridyl molar
ratio. The thin films exhibited similar absorption spectra as
measured for the dyes in dilute solutions (see Supporting Infor-
mation, Figure S2), showing that no excessive dye aggregation
takes place in the films.[55]
The formation of supramolecular complexes in spin-cast
thin films was verified by X-ray photoelectron spectroscopy
(XPS), which is a powerful tool for investigating halogen-
bonded and hydrogen-bonded complexes in thin films.[20,56,57]
The binding energies for the I3d doublet of the dye 1a were
less effective than 1a in establishing non-
covalent bonds with pyridyl moieties. A fur-
ther indication of this came from the solid-
state ball milling syntheses of complexes
involving the azobenzene derivatives 1a–c, 2
and the bis-pyridine 3b: Melting point anal-
ysis revealed that the complexes 5 and 6 were
formed after ball milling the starting com-
pounds for 60 min at 30 Hz, whereas simple
mechanical binary mixtures were obtained
when the compounds 1b and 1c were used.
The noncovalent interactions between
the compounds 1a and 2 and the pyridyl
moieties were also probed by infrared spec-
troscopy. Since both halogen and hydrogen
Figure 3. Comparison between a) the first-order diffraction efficiency evolution over time and
b) the surface profiles of thin films of P4VP(1a)0.1, P4VP(1b)0.1, P4VP(1c)0.1, and P4VP(2)0.1
The sample thicknesses were 90 nm 5 nm. Both the diffraction efficiency and the modulation
.
bonding are weaker than covalent and ionic depth are seen to increase in the order 1c < 1b < 2< 1a.
©
4
wileyonlinelibrary.com
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135

www.afm-journal.de
www.MaterialsViews.com
Figure 3. The complex with 1c is the only
exception, as it yields essentially no diffrac-
tion and only very weak surface modulation,
whereas in the halogen-bonded complexes of
P4VP with 1a and 1b the first-order diffrac-
tion efficiency in reflection mode is approxi-
mately 3.8% and 2.1% after 20 min of irradia-
tion, respectively. The diffraction efficiencies
correlate excellently with the surface profiles
(Figure 3b), which show that the modulation
depths are ca. 55 nm, 30 nm, and 10 nm for
the 1a, 1b, and 1c-based complexes, respec-
tively. The dashed line in Figure 3a corre-
sponds to the diffraction curve for the analo-
gous phenol–pyridine complex P4VP(2)0.1
,
the SRG inscription rate of which is slower
than for the P4VP(1a)0.1 complex. On the
other hand, the inscription performance of
the dye 2 exceeds that of the bromine-substi-
tuted chromophore 1b.
Next, we further compared the photopat-
terning in 1a and 2-based complexes using
0.2 dye/pyridyl molar ratios. Both samples
displayed high optical quality (Figure 4a
Figure 4. Characterization of spin-cast thin films of P4VP(1a)0.2 and P4VP(2)0.2 complexes. The
sample thicknesses were ca. 110 nm and 100 nm for the films containing 1a and 2, respec-
tively. a,b) Optical microscopy images of thin films of P4VP(1a)0.2 and P4VP(2)0.2, respectively. and 4b), having an average surface rough-
c) Atomic force microscopy (AFM) surface profile of the SRG obtained from P4VP(1a)0.2. d) The
surface-modulation depths and e) the time evolution of the first-order diffraction efficiency of
a He–Ne probe beam upon the SRG formation.
ness of around 1 nm as measured by AFM.
The surface profile for a grating inscribed
on P4VP(1a)0.2 fi lm is shown in Figure 4c.
As seen from Figure 4d, the modulation
depths of the gratings were ca. 90 nm and 70 nm in favor of
the halogen-bonded complex. However, the most striking dif-
ference between the halogen- and hydrogen-bonded samples
lies in the time evolution of first-order diffraction efficiency,
which is commonly used to monitor the build-up of the SRG
(Figure 4e). In the case of the halogen-bonded film, the initial
slope of the curve, i.e., the SRG inscription efficiency, is signifi-
cantly higher than for the hydrogen-bonded film, even if their
overall diffraction efficiencies in reflection mode upon 20 min
of irradiation time are comparable, ca. 5% for P4VP(1a)0.2 and
4% for P4VP(2)0.2. We note here that when the inscription
was performed using s-polarized light, essentially no gratings
were formed. This evidences the fact that under the experi-
mental conditions used (inscription wavelength 488 nm, irra-
diation intensity 300 mW cm−2 , irradiation time 20 min), the
grating formation is driven by light-induced mass transport,
as opposed to thermal effects, ablation, and photodegradation,
which start playing a role at higher irradiation intensities but
are polarization-independent.[13,14]
620.95 eV/632.40 eV in its pure form, and 620.62 eV/632.08 eV
in the complex P4VP(1a)0.2, i.e., they shifted by 0.33 and 0.32 eV
to lower energy (Supporting Information, Figure S4a). Similarly,
the binding energy of the O1s electron of the dye in the complex
P4VP(2)0.2 shifted by 0.66 eV to lower energy upon hydrogen
bond formation, from 532.32 eV to 531.66 eV (Supporting
Information, Figure S4b). These energy shifts are in excellent
agreement with literature values for similar interactions,[20,56,57]
and they prove that the spin-cast thin films contain halogen-
or hydrogen-bonded species. A similar conclusion can also be
derived from the analysis of the Fourier transform infrared
(FTIR) spectra of the films P4VP(1a)0.2 and P4VP(2)0.2, which
showed several band shifts to lower wave numbers compared to
the spectra of the pure dyes (Supporting Information, Table S2).
In order to establish a connection between the SRG inscrip-
tion efficiency and the nature and strength of the polymer–dye
interaction, we compared the grating formation in P4VP(1a)0.1
,
P4VP(1b)0.1, P4VP(1c)0.1, and P4VP(2)0.1. First, we estimated
the rate of thermal cis–trans isomerization for the complexes
following literature procedures,[58,59] and obtained values of
(11.8 0.2) × 10−3 s−1 , (13.4 0.1) × 10−3 s−1 , (12.2 0.2) ×
10−3 s−1 , and (13.3 0.1) × 10−3 s−1 for P4VP(1a)0.1, P4VP(1b)0.1
,
3. Discussion
P4VP(1c)0.1, and P4VP(2)0.1, respectively. The fact that there is
no significant difference in the cis-isomer lifetime between the
different dyes is important for meaningful comparison of their
SRG formation efficiencies.
The films afforded high-quality SRGs upon irradiation with
an interference pattern produced with circularly polarized
(or p-polarized) light. The diffraction efficiencies and grating
modulation depths for the x = 0.1 complexes are shown in
Based on the results shown above, the grating formation effi-
ciency develops in the order 1c < 1b < 2 < 1a, which allows us to
make two important conclusions. First, the comparison between
the complexes involving the dyes 1a, 1b, and 1c attests that the
grating formation efficiency is related both to the strength (cf.
1a and 1b) and to the nature (cf. 1b and 1c) of the noncovalent
interactions used. Importantly, the single atom mutation does
©
wileyonlinelibrary.com
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
5

www.afm-journal.de
www.MaterialsViews.com
not influence the photophysical and electronic properties of the
three dyes, which are very similar. Hence, these features do not
account for the remarkably different optical performances of
the complexes shown in Figure 3. The comparison between 1a
and 1b is unique, representing to the best of our knowledge
the first experimental demonstration of the effect of polymer–
azobenzene interaction strength on SRG formation efficiency,
even if it has been previously established that covalent/noncov-
alent bond between the polymer and the azo dyes is required
in order for the mass transport to take place.[12,60] We empha-
size that this is made possible by the unique nature of halogen
bonding, which allows one to control the interaction strength
by halogen atom substitution while leaving the electronic struc-
ture and the photophysical properties of the dyes essentially
unaltered.[41]
the hydrogen-bonded complex. The above-described results
prompt us to propose that there is a relationship among the
positive charge localization on the dye, the directionality and
rigidity of the polymer–dye interaction, and the SRG forma-
tion and efficiency in the studied systems. The sharp localiza-
tion of the positively charged region in 1a and 1b asks for a
highly directional interaction with the host polymer matrix. In
contrast, the positive charge is hemispherically distributed on
the hydrogen atoms of 1c and 2 resulting in a less directional
and, more importantly, less rigid interaction between the pol-
ymer and the dye. The conclusion is that not only the strength,
but even more importantly the directionality of the noncovalent
interaction between the polymer and the photoactive chromo-
phore are to be seen as dominating factors in determining the
SRG inscription efficiency in the systems under study.
On the other hand, the neatly superior performance of 1b
with respect to 1c cannot be explained on the basis of interac-
tion strength and photoswitching efficiency alone, since our
theoretical studies showed that the interaction energies of
the two dyes with the pyridyl moieties as well as their dipole
moments have comparable values. We therefore suggest that
the higher directionality of halogen bonding with respect to
hydrogen bonding provides a more rigid dye–polymer junction,
resulting in enhanced mass transport and more efficient SRG
inscription. This statement is further supported by the compar-
ison between 1a- and 2-based complexes (Figure 4), in which
1a displays a better performance in spite of a lower binding
strength to P4VP. Therefore, our results suggest that halogen-
bonded complexes can perform at least as well as analogous
hydrogen-bonded polymer–azobenzene complexes in the photo-
patterning efficiency, and potentially even outperform them.
However, we note that a clear-cut comparison is difficult to
make in the latter case, because of the different positioning and
bulkiness of 1a and 2, and because their electronic structures
differ significantly (the absorption maxima of 2 and 1a are at
415 nm and 470 nm, respectively), which can play a role in the
SRG formation.[49] Moreover, one might invoke a stronger plas-
ticizing effect on the part of the fluorinated ring, which might
ease the polymer chain mobility and result in more efficient
grating inscription. As far as the latter effect is concerned, the
wide span of performances covered by the fluorinated dyes 1a–c
strongly suggests that this is unlikely to be a major factor. The
different positioning of 1a and 2 with respect to the pyridine
moieties may indeed affect the inscription performance, but it
should be noted that this factor is not expected to account for
the better results obtained for 1b in comparison to 1c. Simi-
larly, even if the bulkiness and electronic structure differences
are likely to play a role in the SRG formation, they cannot be
accounted for as the main elements determining the higher
SRG inscription efficiency of P4VP(1a)0.2. In order to verify
whether a higher polymer/dye ratio would play in favour of
the hydrogen-bonded complex, we challenged the SRG forma-
tion efficiency of P4VP(1a)0.2 with that of P4VP(2)0.33 samples,
i.e., higher content of hydrogen bonding dye in the polymer.
Moreover, we decided to use in this experiment an irradia-
tion wavelength of 458 nm, instead of the 488 nm used in
the previous experiments, since this wavelength is equally
absorbed by both dyes (see Supporting Information, Figure S5).
Even in those conditions, the 1a-based complex outperforms
4. Conclusions
Our results not only show that halogen bonding can be used
to design supramolecular polymer–azobenzene complexes
for photoinduced surface patterning, but also point out that
i) the higher directionality of halogen bonding as compared
to hydrogen bonding promotes the patterning efficiency and
ii) the unique property of halogen bonding to allow fine-tuning
of the polymer–dye interaction strength via single halogen
atom mutation provides us with fundamental tools for under-
standing the phenomenon of SRG formation not available in
other materials. Such implications are remarkably important in
this context, as the underlying mechanism of this fascinating
and high-potential photomechanical effect is still under debate.
To the best of our knowledge, this is one of the very few reports
on the successful use of halogen bonding in the self-assembly
of polymer-based systems and the first application of this
non-covalent interaction in controlling the properties of light-
responsive supramolecular polymers.
5. Experimental Section
Materials and Methods: The starting materials were purchased from
Sigma-Aldrich, Acros Organics, and Apollo Scientific. The dye 2 was
purchased from TCI Europe. Commercial high-performance liquid
chromatography (HPLC)-grade solvents were used without further
purification, except for acetonitrile, which was dried over molecular
sieves before use.
1H and 19F NMR spectra were recorded at room temperature on a
Bruker AV500 spectrometer, using CDCl3 or acetone-d6 as solvents. H
NMR chemical shifts were referenced to tetramethylsilane (TMS) using
the residual proton impurities of the deuterated solvents as standard
reference, while 19F NMR chemical shifts were referenced to an internal
CFCl3 standard.
The molecular structures of dyes 1a–c and 2 were optimized
for vacuum and for DMF solution using DFT by applying the PBE0
functional and the 6-311++G∗∗ basis set for all atoms. Absorption
spectra were computed by time-dependent DFT. Accurate interaction
energies for the molecular dimers composed of the dyes and
4-methylpyridine, corrected for basis set superposition error (ΔE BSSE),
were computed at PBE0/6-311++G∗∗ level as the difference between
the energy of the dimer and the sum of the energies of the single
monomers. Further details about the theoretical calculations can be
found in the Supporting Information.
1
©
6
wileyonlinelibrary.com
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135

www.afm-journal.de
www.MaterialsViews.com
DSC analysis was performed on
a Mettler Toledo DSC823e
acetonitrile (10 mL) at –30 °C. After 1h of additional stirring at –30 °C
the N,N-dimethylaniline (24 mmol) was added dropwise. The resulting
solution was stirred overnight at room temperature and then water (30
mL) was added. The mixture was extracted three times with CH2Cl2.
The organic layers were collected and dried over Na2SO4. The solvent
was removed under reduced pressure. The residue was purified by
recrystallization from methanol.
instrument, using aluminium light 20 μL sample pans and Mettler
STARe software for calculation. The melting points were also determined
on a Reichert instrument by observing the melting process through an
optical microscope. Attenuated total reflectance FTIR (ATR-FTIR) spectra
were obtained with a Nicolet Nexus FTIR spectrometer. The values were
given in wave numbers and were rounded to 1 cm−1 upon automatic
assignment. The X-ray crystal structures were determined on a Bruker
Smart Apex II diffractometer. Crystallographic data (excluding structure
factors) for the structures reported in this paper have been deposited
with the Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC-829039; 829040; 829041, 829042. These data can
be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif.
Experimental details about the crystalline structures can be found in the
Supporting Information.
The surface-relief gratings were inscribed on thin films spin-cast
from freshly prepared DMF solutions of the dye–polymer mixtures on Si
substrates. The sample thicknesses were measured with a spectroscopic
ellipsometer (Sentech Instruments GmbH, SE805) and double-checked
using a DEKTAK 6M surface profiler. The UV-vis spectra were collected
using Perkin Elmer Lambda 950 spectrophotometer from both thin films
(spin-coated on glass substrates) and diluted (10−6 M) DMF solutions.
XPS measurements were performed using an ULVAC-PHI Inc. 1700R
ESCA spectrometer equipped with MgK α X-ray source (1253.6 eV) and
hemispherical analyzer. The resolution used was 0.05 eV. The reference
spectra of the pure dyes were measured from powders mounted on
double-sided carbon tape. The spectra of the complexes were measured
from thin films spin coated on silicon substrates. All spectra were
referenced to the C1s neutral carbon peak at 284.6 eV.[55]
1
1a: Yield 42%; m.p. = 456 K; H NMR (500 MHz, CDCl3, δ) = 7.90
(d 2 H, J = 9.2 Hz), 6.77 (d 2 H, J = 9.2 Hz), 3.15 (s 6 H); 19F NMR (470
MHz, CDCl3, δ) -122.48 (2 F), -151.12 (2 F); FTIR: νmax 2910, 2857, 2804,
2740, 2667, 1596, 1523, 1476, 1393, 1359, 1332, 1308, 1278, 1230, 1144,
1051, 976, 940, 887, 818, 806, 728, 664, 619.
1
1b: Yield 40%; m.p. = 444 K; H NMR (500 MHz, CDCl3, δ) 7.90 (d
2 H, J = 9.4 Hz), 6.74 (d 2 H, J = 9.4 Hz), 3.13 (s 6 H); 19F NMR (470
MHz, CDCl3, δ) -135.48 (2 F), -151.77 (2 F); FTIR: νmax 2918, 2857, 2810,
2667, 1596, 1519, 1484, 1445, 1395, 1360, 1331, 1309, 1282, 1229,1176,
1142, 1052, 981, 939, 889, 829, 818, 730.
1c: Yield 25%; m.p. = 421 K; 1H NMR (500 MHz, CDCl3, δ) 7.90 (d 2
H, J = 8.4 Hz), 6.74 (d 2 H, J = 8.4 Hz), 6.99 (m 1 H), 3.13 (s 6 H); 19F
NMR (470 MHz, CDCl3, δ) -140.92 (2 F), -153.46 (2 F); FT- FTIR: νmax
3082, 2913, 2817, 2667, 1596, 1502, 1359, 1333, 1308, 1278, 1230, 1140,
1035, 954, 865, 819, 758, 734, 714, 687, 663, 629.
Co-Crystallization Experiments: In a typical co-crystallization procedure,
the azobenzene derivative and the bis-pyridine compound were separately
dissolved in CHCl3 or acetone, at room temperature in a 2:1 ratio, under
saturated conditions. The two saturated solutions containing the donor
and the acceptor were then mixed in a clear borosilicate glass vial, which
was left open in a closed cylindrical wide-mouth bottle containing paraffin
oil. Solvents were allowed to slowly evaporate at room temperature for a
few days until the formation of crystals.
The thermal cis–trans isomerization was studied by exciting the
chromophores with a circularly polarized pump beam (457 nm,
50 mW cm−2 ) and monitoring the transmittance changes after blocking
the pump. The probe was a fiber-coupled xenon lamp equipped with
proper bandpass filters. The signal was detected with a photodiode and
a lock-in amplifier. The isomerization rate constants were estimated
using methods proposed in the literature.[58,59] The surface-relief gratings
were inscribed using a spatially filtered circularly polarized beam from
an Ar+-laser (λ = 488 nm unless otherwise stated), with an irradiation
intensity of 300 mW cm−2 . The interference pattern was created using
a Lloyd mirror configuration, with half of the beam directly incident on
the sample and the other half reflected from a mirror set at 90° with the
sample. The angle between the plane of the mirror and the incident beam
was set to 14° to yield an interference pattern with a periodicity of 1 μm.
The resulting diffraction gratings were monitored by measuring the first-
order diffraction of a low power (<300 μW; 633 nm) He–Ne laser beam
in reflection mode. AFM images used for surface profile characterization
were taken with a Veeco Dimension 5000 SPM in tapping mode.
Synthesis of 4-Iodo-2,3,5,6-tetrafluoroaniline:[61] Yellow HgO (1.2 g,
5.5 mmol) was added to a solution of 2,3,5,6-tetrafluoroaniline (1.18 g,
7.2 mmol) in ethanol (20 mL). The solution was vigorously stirred for
30 min at room temperature, then iodine (1.82 g, 7.2 mmol) was added.
The mixture was stirred overnight and filtered over celite. The solvent
was removed under reduced pressure then the residue was dissolved
in CH2Cl2. The solution was washed several times first with an aqueous
solution of Na2S2O3 (saturated solution), then with pure water. After
drying on Na2SO4 (1 g) the solvent was removed under reduced pressure
giving the pure product (1.90g, yield 90%). m.p. 350 K; 1H NMR
(500 MHz, CDCl3, δ) = 4.1 (-NH2 broad signal); 19F NMR (470 MHz,
CDCl3, δ) -124.41 (2 F); -159.90 (2 F); FT-IR: νmax = 3481, 3387, 1625,
1615, 1598, 1480, 1408, 1347, 1327, 1267, 1171, 1146, 1090,1051, 975,
915, 800, 712.
Co-crystal 4: m.p. = 465 K; FTIR: νmax 3076, 3057, 3034, 2994, 2915,
2862, 2806, 2741, 2698, 2667, 1597, 1556, 1524, 1471, 1396, 1362, 1333,
1309, 1276, 1229, 1143, 1071, 1052, 1044, 999, 975, 940, 887, 819, 763,
726.
Co-crystal 5: m.p. = 446 K; FTIR: νmax 3070, 2991, 2941, 2900, 2801,
2678, 1586, 1516, 1470, 1423, 1363, 1275, 1254, 1229, 1219, 1195, 1144,
1095, 1070, 1011, 943, 837, 817, 793, 730.
1
Co-crystal 6: m.p. = 445 K; H NMR (500 MHz, CDCl3, δ) 8.56 (dd 4
H), 7.90 (dd 4 H), 7.23 (dd 4H), 6.75 (dd, 4H), 3.15 (s 12 H), 2.73 (t 4
H), 2.08 (m, 2H); 19F NMR (470 MHz, CDCl3, δ) -122.50 (4 F), -151.13
(4 F); FTIR: νmax 3072, 3034, 2995, 2916, 2859, 2806, 2744, 2666, 1597,
1557, 1521, 1472, 1444, 1419, 1394, 1362, 1335, 1309, 1273, 1231, 1146,
1067, 1041, 1000, 972, 943, 887, 818, 805, 798.
Mechanochemical Synthesis of the Azobenzene Complexes: Solid-state
reactions were carried out using a Retsch MM400 ball mill with 5.0 mL
vessels. 1,3-di(4-pyridyl)propane (3b) was mixed with 2 equivalents
of azobenzenes 1a–c and 2 in a 5 mL grinding jar. The reactants were
ground at 30 Hz for 60 min. This procedure resulted in the quantitative
formation of the co-crystals 5 and 6, starting from 1a and 2, while
grinding 3b with 1b or 1c gave just a mechanical mixture of the two
starting compounds.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
General Procedure for the Synthesis of 1a–c:[62] The general procedure
for the synthesis of the azobenzene derivatives 1a–c involves the
diazobenzene coupling of a tetrafluoroaniline with N,N-dimethylaniline.
A.P. acknowledges the financial support provided by the Japanese
Society for the Promotion of Science and the Foundations’ Post Doc
Pool in Finland. This work was also partially funded by the Academy
of Finland (PHORMAT project; 135106). P.M. and G.R. acknowledge
the Cariplo Foundation (projects 2150 and 2010-1351) for financial
support. A.F. acknowledges the CINECA Award N. ∗∗HP10BFJG1H 2011
Reactions were carried out in oven-dried glassware under
a
nitrogen atmosphere, using dry solvents. A solution of the proper
2,3,5,6-tetrafluoroaniline (6 mmol) in dry acetonitrile (10 mL) was added
dropwise to a solution of nitrosonium tetrafluoroborate (6 mmol) in
©
wileyonlinelibrary.com
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
7

www.afm-journal.de
www.MaterialsViews.com
‘IscrB_HEChro’, for the availability of high performance computing
resources and support. Prof. Baba and Prof. Motokura from Tokyo Tech are
greatly acknowledged for their kind assistance with XPS spectroscopy.
[28] A. Farina, S. V. Meille, M. T. Messina, P. Metrangolo, G. Resnati,
G. Vecchio, Angew. Chem. Int. Ed. 1999, 38, 2433.
[29] a) P. Metrangolo, F. Meyer, T. Pilati, G. Resnati, G. Terraneo, Angew.
Chem. Int. Ed. 2008, 47, 6114; b) K. Rissanen, CrystEngComm 2008,
9, 1107.
[30] R. Bertani, P. Metrangolo, A. Moiana, E. Perez, T. Pilati, G. Resnati,
I. Rico-Lattes, A. Sassi, Adv. Mater. 2002, 14, 1197.
[31] T. Takeuchi, Y. Minato, M. Takase, H. Shinmori, Tetrahedron Lett.
2005, 46, 9025.
Received: January 15, 2012
Revised: February 14, 2012
Published online:
[1] P. Rochon, E. Batalla, A. Natansohn, Appl. Phys. Lett. 1995, 66, 136.
[2] N. K. Viswanathan, D. Y. Kim, S. Bian, J. Williams, W. Liu, L. Li,
L. Samuelson, J. Kumar, S. K. Tripathy, J. Mater. Chem. 1999, 9,
1941.
[3] Y. Morikawa, S. Nagano, K. Watanabe, K. Kamata, T. Iyoda, T. Seki,
Adv. Mater. 2006, 18, 883.
[4] C. Hubert, A. Rumyantseva, G. Lerondel, J. Grand, S. Kostcheev,
L. Billot, A. Vial, R. Bachelot, P. Royer, Nano Lett. 2005, 5, 615.
[5] S. Lee, H. S. Kang, J.-K. Park, Adv. Funct. Mater. 2011, 21, 1770.
[6] A. Kravchenko, A. Shevchenko, V. Ovchinnikov, A. Priimagi,
M. Kaivola, Adv. Mater. 2011, 23, 4174.
[7] T. Ubukata, T. Seki, K. Ichimura, Adv. Mater. 2000, 12, 1675.
[8] B. Darracq, F. Chaput, K. Lahlil, Y. Levy, J. P. Boilot, Adv. Mater.
1998, 10, 1133.
[32] a) N. S. Goroff, S. M. Curtis, J. A. Webb, F. W. Fowler, J. W. Lauher,
Org. Lett. 2005, 7, 1891; b) A. Sun, J. W. Lauher, N. S. Goroff, Science
2006, 312, 1030.
[33] T. Shirman, D. Freeman, Y. D. Posner, I. Feldman, A. Facchetti,
M. E. van der Boom, J. Am. Chem. Soc. 2008, 130, 8162.
[34] L. Brunsveld, B. J. B. Folmer, E. W. Meijer, R. P. Sijbesma, Chem.
Rev. 2001, 101, 4071.
[35] M. Weck, Polym. Int. 2007, 56, 453.
[36] T. Kato, N. Mizoshita, K. Kishimoto, Angew. Chem. Int. Ed. 2006, 45,
38.
[37] M. R. Hammond, R. Mezzenga, Soft Matter 2008, 4, 952.
[38] G. ten Brinke, O. Ikkala, Chem. Rec. 2004, 4, 219.
[39] P. Politzer, J. S. Murray, T. Clark, Phys. Chem. Chem. Phys. 2010, 12,
7748.
[9] H. Nakano, T. Tanino, T. Takahashi, H. Ando, Y. Shirota, J. Mater.
Chem. 2008, 18, 242.
[40] P. Metrangolo, F. Meyer, T. Pilati, G. Resnati, G. Terraneo, Chem.
Commun. 2008, 1635.
[10] J. Gao, Y. He, F. Liu, X. Zhang, Z. Wang, X. Wang, Chem. Mater.
2007, 19, 3877.
[41] P. Politzer, P. Lane, M. C. Concha, Y. Ma, J. S. Murray, J. Mol. Model.
2007, 13, 305.
[11] O. Kulikovska, L. M. Goldenberg, J. Stumpe, Chem. Mater. 2007, 19,
3343.
[42] A. Priimagi, K. Lindfors, M. Kaivola, P. Rochon, ACS Appl. Mater.
Interfaces 2009, 1, 1183.
[12] N. Zettsu, T. Ogasawara, N. Mizoshita, S. Nagano, T. Seki, Adv.
Mater. 2008, 20, 516.
[43] J. Vapaavuori, A. Priimagi, M. Kaivola, J. Mater. Chem. 2010, 20,
5260.
[13] O. N. Oliveira, J. Kumar, L. Li, S. K. Tripathy, in Photoreactive Organic
Thin Films (Eds: Z. Sekkat, W. Knoll) Academic Press, San Diego,
CA 2002, Ch. 14.
[14] K. G. Yager, C. J. Barrett, in Smart Light-Responsive Materials, (Eds.
Y. Zhao, T. Ikeda), John Wiley & Sons Inc., Hoboken, NJ 2009, Ch. 4.
[15] O. Hassel, Nobel Lecture, 9 June 1970.
[16] An IUPAC Task Group set up to examine the definition of halogen
bonding has not yet reported, so that given here should be taken
as temporary. See the IUPAC website www.iupac.org/web/ins/2009-
032-1-100 and the website of the project on halogen bonding: www.
halogenbonding.eu, (accessed March 2012).
[44] T. Alasaarela, D. Zheng, L. Huang, A. Priimagi, B. Bai, A. Tervonen,
S. Honkanen, M. Kuittinen, J. Turunen, Opt. Lett. 2011, 36, 2411.
[45] J. Vapaavuori, V. Valtavirta, T. Alasaarela, J. Mamiya, A. Priimagi,
A. Shishido, M. Kaivola, J. Mater. Chem. 2011, 21, 15437.
[46] A. Forni, J. Phys. Chem. A 2009, 113, 3403.
[47] M. D. Joeston, R. S. Drago, J. Am. Chem. Soc. 1962, 84, 3817.
[48] E. Corradi, S. V. Meille, M. T. Messina, P. Metrangolo, G. Resnati,
Angew. Chem. Int. Ed. 2000, 39, 1782.
[49] C. J. Barrett, A. L. Natansohn, P. L. Rochon, J. Phys. Chem. 1996,
100, 8836.
[50] A. Bondi, J. Phys. Chem. 1964, 68, 441.
[17] P. Metrangolo, H. Neukirch, T. Pilati, G. Resnati, Acc. Chem. Res.
2005, 38, 386.
[18] H. Loc Nguyen, P. N. Horton, M. B. Hursthouse, A. C. Legon,
D. W. Bruce, J. Am. Chem. Soc. 2004, 126, 16.
[19] J. Xu, X. Liu, C. He, Macromolecules 2005, 38, 3554.
[20] J. Xu, X. Liu, J. K.-P Ng, T. Lin, C. He, J. Mater. Chem. 2006, 16,
3540.
[21] D. W. Bruce, P. Metrangolo, F. Meyer, T. Pilati, C. Präsang,
G. Resnati, G. Terraneo, S. Wainwright, A. C. Whitwood, Chem. Eur.
J. 2010, 16, 9511.
[51] P. Metrangolo, G. Resnati, T. Pilati, S. Biella, in Halogen Bonding
Fundamentals and Applications, (Eds: P. Metrangolo, G. Resnati),
Springer, Berlin, Germany 2008, Ch. 4, pp. 105–136.
[52] L. C. Cesteros, J. R. Isasi, I. Katime, Macromolecules 1993, 26, 7256.
[53] J. Ruokolainen, G. ten Brinke, O. Ikkala, M. Torkkeli, R. Serimaa,
Macromolecules 1996, 29, 3409.
[54] A. Priimagi, J. Vapaavuori, F. J. Rodriguez, C. F. J. Faul, M. T. Heino,
O. Ikkala, M. Kauranen, M. Kaivola, Chem. Mater. 2008, 20, 6358.
[55] A. Priimagi, S. Cattaneo, R. H. A. Ras, S. Valkama, O. Ikkala,
M. Kauranen, Chem. Mater. 2005, 17, 5798.
[22] P. Metrangolo, Y. Carcenac, M. Lahtinen, T. Pilati, K. Rissanen,
A. Vij, G. Resnati, Science 2009, 323, 1461.
[56] X. Zhou, S. H. Goh, S. Y. Lee, K. L. Tan, Appl. Surf. Sci. 1997, 119,
60.
[23] A. Abate, M. Brischetto, G. Cavallo, M. Lahtinen, P. Metrangolo,
T. Pilati, S. Radice, G. Resnati, K. Rissanen, G. Terraneo, Chem.
Commun. 2010, 46, 2724.
[24] M. Fourmigue, P. Batail, Chem. Rev. 2004, 104, 5379.
[25] O. Bolton, K. Lee, H. J. Kim, K. Y. Lin, J. Kim, Nat. Chem. 2011, 3, 205.
[26] A. Mele, P. Metrangolo, H. Neukirch, T. Pilati, G. Resnati, J. Am.
Chem. Soc. 2005, 127, 14972.
[57] F. Wang, N. Ma, Q. Chen, W. Wang, L. Wang, Langmuir 2007, 23,
9540.
[58] C. S. Paik, H. Morawetz, Macromolecules 1972, 5, 171.
[59] C. Barrett, A. Natansohn, P. Rochon, Chem. Mater. 1995, 7, 899.
[60] C. Fiorini, N. Prudhomme, G. de Veyrac, I. Maurin, P. Raimond,
J.-M. Nunzi, Synth. Met. 2000, 115, 121.
[61] H. H. Wenk, W. Sander, Angew. Chem. Int. Ed. 2002, 41, 2742.
[62] R. Weiss, F. G. Puhlhofer, J. Am. Chem. Soc. 2007, 129, 547.
[27] A. R. Voth, P. Khuu, K. Oishi, P. S. Ho, Nat. Chem. 2009, 1, 74.
©
8
wileyonlinelibrary.com
2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Funct. Mater. 2012,
DOI: 10.1002/adfm.201200135