Ion-conduction pathways in self-organised ureidoarene-heteropolysiloxane hybrid membranes

Article abstract of DOI:10.1002/chem.200700943

This paper reports on hybrid organic-inorganic dense membrane materials in which protons and ions are envisioned to diffuse along the hydrophilic pathways. The hierarchical generation of functional hybrid materials was realised in two steps. First, the self-assembling properties of 3-(ureidoarene) propyltriethoxysilane compounds 1-5 in aprotic solvents were determined, revealing the formation of supramolecular oligomers. Compounds 1-5 generate organogels in chloroform or in acetone, leading in a second sol-gel transcription step to hybrid membrane materials on a nanoscopic scale. The crystal structures of 1-5 indicate that the arrangement is mainly defined by periodic parallel sheets, resulting from the alignment of hydrophobic organic and inorganic silica layers. Hybrid materials MB 1-MB 4, with a similar lamellar structure, define particularly attractive functional transport devices; they are oriented along the organic layers and sandwiched between the two siloxane layers. These systems have been employed successfully to design solid dense membranes and illustrate how the self-organised hybrid materials perform interesting and potentially useful functions.

Full text of DOI:10.1002/chem.200700943

DOI: 10.1002/chem.200700943
Ion-Conduction Pathways in Self-Organised Ureidoarene–
Heteropolysiloxane Hybrid Membranes
N
[a]
Mathieu Michau,[a] MihailBarboiu,* RØmi Caraballo,[a] Carole Arnal-HØrault,[a]
[b]
PascalPerriat, Arie van der Lee,[a] and A. Pasc[a]
Abstract: This paper reports on hybrid
organic–inorganic dense membrane
materials in which protons and ions are
envisioned to diffuse along the hydro-
philic pathways. The hierarchical gener-
ation of functional hybrid materials
was realised in two steps. First, the self-
assembling properties of 3-(ureidoare-
ne)propyltriethoxysilane compounds 1–
5 in aprotic solvents were determined,
revealing the formation of supramolec-
ular oligomers. Compounds 1–5 gener-
ate organogels in chloroform or in ace-
tone, leading in a second sol–gel tran-
scription step to hybrid membrane ma-
terials on a nanoscopic scale. The crys-
tal structures of 1–5 indicate that the
arrangement is mainly defined by peri-
odic parallel sheets, resulting from the
alignment of hydrophobic organic and
inorganic silica layers. Hybrid materials
MB1–MB4, with a similar lamellar
structure, define particularly attractive
functional transport devices; they are
oriented along the organic layers and
sandwiched between the two siloxane
layers. These systems have been em-
ployed successfully to design solid
dense membranes and illustrate how
the self-organised hybrid materials per-
form interesting and potentially useful
functions.
Keywords: membranes · molecular
devices · organic–inorganic hybrid
composites · self-assembly · sol–gel
processes
Introduction
according to different relaying and migration mecha-
nisms.[1–6]
Many fundamental biological processes appear to depend
on the unique properties of the hydrophilic domains of bio-
molecules. Such behaviour is illustrated by the functional
complexity of self-organised membrane proteins, which may
assist in proton and ion translocation through membranes.[1]
Gramicidin A,[2] the KCsA K+ channel,[3] and Aquaporin
water channels[4] are well-known non-exclusive examples of
functional systems in which protons, ions and water mole-
cules are envisioned to diffuse along a directional pathway
Numerous artificial transport systems utilizing carriers,
channel-forming or self-organised polymeric superstructures
able to orient, select, and pump the ionic transport across
membranes, have been developed in the last few decades.
Crown ethers,[7] cyclic peptides,[8] oligophenyl barrel–stave
structures[9] and oligoester bolaamphiphiles[10] have all been
used in this context, mostly in lipid bilayers. Functional crys-
talline organic[11] or inorganic materials,[12] polymeric[13] or
dendrimeric[14] self-organised devices, designed to mimic nat-
ural ion channel proteins, moved the field of bio-inspired
supramolecular frameworks to a nanometric scale.
[a] Dr. M. Michau, Dr. M. Barboiu, R. Caraballo, Dr. C. Arnal-HØrault,
Dr. A. van der Lee, Dr. A. Pasc
Hybrid organic–inorganic materials produced by sol–gel
processes are the subject of various investigations, offering
the opportunity to achieve nanostructured functional mate-
rials.[15] These materials reveal great potential on the level of
their chemical composition or organisation and in terms of
the applications concerned.[15a–c] Organogels resulting from
multiple self-assembly processes and acting as robust organ-
ic templates for the TEOS sol–gel process (TEOS=tetra-
Adaptative Supramolecular Nanosystems Group
Institut EuropØen des Membranes
ENSCM/UMII/UMR-CNRS 5635, Place Eug›ne Bataillon
CC 47, 34095 Montpellier, Cedex 5 (France)
Fax : (+33)467-14-91-19
E-mail: barboiu@iemm.univ-montp2.fr
[b] Dr. P. Perriat
MATEIS (GEMPPM) UMR-CNRS 5510, INSA de Lyon
Batiment, Blaise Pascal
7 Avenue Jean capelle, 69621 Villeurbane, Cedex (France)
ethoxysilane) have been used for the synthesis of different
inorganic structures.[15d] Many groups, including our own,
have found new methods for the elaboration of hybrid self-
organised materials based on silsesquioxanes, in which the
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
1776
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1776 – 1783

FULL PAPER
functional organic (supramolecular) and inorganic siloxane
networks are covalently connected. Rigid aromatic mole-
cules[15e–g] and urea ribbons[15h–k,16] are used to transcribe an
ic essential amino acids. The methoxybenzene 4 and hydrox-
ymethylbenzene 5 moieties were chosen for the structural
studies described later. The 3-isocyanatopropyltri-
oriented supramolecular self-organisation in
a
siloxane
A
matrix by a sol–gel process. Of particular interest is the
structure-directed function of hybrid materials and the con-
trol of their build-up by self-organisation. Recently, we
proved the possibility of creating hybrid macrocyclic com-
plexant membranes, self-organised in directional tubular su-
perstructures. They have been employed successfully to
design an ion-powered ATP2ꢀ pump.[16]
amine (CHCl3, 808C, 5 h) to afford after crystallisation 1–5
as white powders.
The generation of self-organised superstructures in solu-
tion, sols, gels, and in the solid state, is based on three en-
coded features:
1) They contain an hydrophilic arene moiety (except 1); the
importance of interactions between alkali cations and
side chains of aromatic amino acids has been known for
many years and they are of particular biological signifi-
cance;[3,5,7a]
2) The supramolecular guiding interaction is the urea head-
to-tail hydrogen-bond association,[9a] similar to the hy-
drogen bonding of the amide moieties in proteins
3) These ureido–aromatic residues are covalently bonded to
triethoxysilyl groups, allowing them, by a sol–gel process,
to transcribe their self-organised superstructures formed
in solutions and gels to the hybrid materials.
However, the structural functionality of membrane pro-
teins is not defined by the specific complexant receptors.
Very simple functional moieties (i.e., carbonyl, hydroxyl,
amide, etc.) point towards the protein core that surrounds
the transport direction. The end result when such moieties
come together is a supramolecular organisation of binding
sites collectively contributing to the selective translocation
of solutes along the protein wall.[2–4] Toward this objective,
our aim is to develop new functional hybrid membrane ma-
terials by using simple molecules bearing hydrophilic moiet-
ies that form patterns by collective self-assembly so as to
enable efficient translocation events. Herein we describe
several findings related to this aim, and we report new
hybrid organic–inorganic dense materials in which protons
and ions are envisioned to diffuse along hydrophilic path-
ways. The hierarchical generation of functional hybrid mate-
rials was realised in two steps. First, the self-assembling
properties of 3-(ureidoarene)propyltriethoxysilane com-
pounds 1–5 in aprotic solvents were determined, revealing
the formation of supramolecular oligomers. They generate
organogels in chloroform or acetone, leading in a second
sol–gel transcription step to hybrid membrane materials on a
nanoscopic scale. Finally they were tested in dialysis trans-
port experiments.
1H NMR spectra of solutions of 1–5 show a strong down-
field shift of both NH protons upon increasing the concen-
tration in CDCl3, which is indicative of self-association
through intermolecular hydrogen bonding.[18] The associa-
tion of species 1–5 diminishes from 1ffi5>4>2>3
(Figure 1). The most associated compounds (1, 4 and 5) con-
tain a hydrophobic sidearm. The formation of aggregates is
probably hindered when supplementary hydrogen-bonding
residues (2 and 3) are present.
Results and Discussion
Synthesis of alkoxysilane precursors 1–5 and self-organised
hybrid materials MB1–MB4, MB1*–MB4*: Five 3-(ureido-
arene)propyltriethoxysilane compounds (1–5) were prepared
for the studies described here (Scheme 1). We restricted our
studies to benzene-, indole-, and phenol-derivatives 1–3, re-
spectively, which are the arene (Ar) termini of phenyl-
alanine, tryptophan, and tyrosine, respectively, three aromat-
ꢀ
Figure 1. Chemical shifts of the N H1 protons plotted against the concen-
tration of 1–5 at 258C.
Compounds 1–5 can gelate the solvents acetone and
chloroform (C>1, Figure 1). To limit the cross-linking of
the siloxane network favoured by hydrolysis, and to maxi-
mise the self-organisation of molecular organic moieties, we
conducted the sol–gel experiments using mild hydrolysis–
condensation conditions:
Scheme 1.
Chem. Eur. J. 2008, 14, 1776 – 1783
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1777

M. Barboiu et al.
1) The first strategy was to use heterogeneous acidic cataly-
sis in aprotic solvents, as previously reported by us[16c]
and by others.[15k]
2) The second method makes use of diphenyldichlorosilane
(Ph2SiCl2) as the organic initiator after reaction with the
residual water in the reaction medium.
The membranes MB1–MB4 (acidic catalysis) and MB1*–
MB4* (organic catalysis) were obtained by coating the sol
onto planar polyvinylidenedifluoride (PVDF) supports by
using a tape-casting method. By using similar experimental
conditions, the MB5 hybrid material precipitates after the
sol-gel process, and therefore only inhomogeneous mem-
brane films were prepared from compound 5.
FTIR and NMR spectroscopic analyses of precursors 1–5
and of solids MB1–MB4 and MB1*–MB4* demonstrate
the formation of self-organised hybrid materials. FTIR spec-
tra of the hybrid materials show the appearance of broad
vibrations of the nSi-O-Si =1000–1200 cmꢀ1 instead of the vi-
brations nSi-OEt =950, 1070, and 1100 cmꢀ1, initially observed
for the molecular precursors 1–5. Evidence for strong urea
hydrogen bonding was obtained from the FTIR spectra of
solids MB1–MB4 and MB1*–MB4*: hydrogen-bond vibra-
tion shifts of the urea moiety were detected: nN-H =3310
(Dn=30), nC O =1630 (Dn=15) and nC-N =1570 cmꢀ1 (Dn=
=
25).[17b,c] Moreover we observed a widening of the nN-H
=
3300 cmꢀ1 vibration band, resulting from the superposition
of several signals. 29Si MAS NMR spectroscopic experiments
are in agreement with the ꢁ70% condensed hybrid materi-
als MB1–MB4 and MB1*–MB4*.
Additional information concerning the self-organisation is
obtained from the single-crystal structures and the X-ray
powder diffraction analyses of compounds 1–5. Crystals suit-
able for X-ray structure determination were obtained after a
few days in acetone (c=0.1 gmLꢀ1) at ꢀ108C. The molecu-
lar structures and packing are presented in Figures 2 and 3.
They reveal the expected 3-(ureiodoarene)propyltriethoxy-
silanes structures of 1–5. Each unit cell contains four mole-
cules, except compound 3 (eight molecules). The urea
moiety and the aromatic ring are disposed in an angular po-
sition with respect to the arene plane. The H···O distances
of 2.04–2.08 Š are the same along the urea ribbon and are
consistent with other urea systems.[17] The relative arrange-
ment of the 3-(ureiodoarene)propyltriethoxysilane com-
pounds 1–5 in the crystal is similar to previously reported
benzocrown-ether-type superstructures.[16d] As shown in
Figure 1, the molecules of 1–5 are organised in the anti-par-
allel hydrogen-bonded aggregates. In the crystal lattice the
compounds 1–5 pack in one direction into ureiodoarene and
triethoxysilane layers, each of about 3 Š thickness, which
Figure 2. Crystal structures of compounds 1–5: side and top views of the
stick representation of H-bonded superstructures of a) 1, b) 2, c) 3, d) 4,
and e) 5; H atoms are omitted for clarity.
bonded to the hydroxyl of a neighbouring molecule of 3
(dH···O =1.86 Š). This results in the internal disruption of the
siloxane layers in which these hydrogen-bond interactions
occur (Figure 3b,c). Further valuable information was ob-
are alternately stratified and present
a
tight contact
tained from the crystal structures of the 3-(p-methoxyb
ureido)propyltriethoxysilane (4) and the 3-(p-hydroxy-
methylbenzeneureido)propyltriethoxysilane (5) derivatives.
G
N
U
N
N
E
R
T
G
N
U
T
H
C
A
enzene-
(Figure 3). For compounds 2 and 3, there is a very interest-
ing “communicative” interaction between the organic and
inorganic networks, corresponding to hydrogen-bonding
ACHTREUNG
A
ACHTREUNG
ꢀ
ꢀ
The substitution of the OH group by an OMe group in
compound 4 suppresses the hydrogen-bond communication
between the organic and silane networks, leading to a regu-
ꢀ
ꢀ
contacts between the NH (2) and the OH (3) groups and
neighbouring triethoxysilane moiety (dH···O =1.83 Š).
Moreover the phenolic group is simultaneously hydrogen
a
ꢀ
lar directional arrangement of MeO groups based on hy-
1778
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1776 – 1783

Self-Organised Hybrid Membranes
FULL PAPER
ꢀ
drophobic interactions (dO···O =5.0 Š). The OH group of 5
presents a statistical disorder in the crystal, being simultane-
ꢀ
ously hydrogen-bonded to the neighbouring CH2OH
(dO···O =2.88 Š) and alkoxysilane groups (dO···O =2.87 Š).
ꢀ
Accordingly, oriented arrays of hydrogen-bonded CH2OH
groups, mutually communicating with the inorganic silane
layers, are generated in the solid state of 5. This results in
the formation of continual inorganic alkoxysilane layers in
the crystal structures of 4 and 5 (Figure 3d,e).
To verify the compatibility of the bulk X-ray fingerprint
with the single-crystal structures described before, wet bulk
powders of 1–5 obtained from evaporation of gels before
the sol–gel process were investigated by X-ray powder dif-
fraction. In three cases (1, 3 and 5) a successful identifica-
tion could be made, which is shown by the Le Bail refine-
ment of the powder patterns in the Supporting Information,
Figure 3Sa–c. Impurity peaks were found in all cases, repre-
senting up to approximately 10% of the weight of the speci-
men. Since the compounds were found to be chemically
pure by NMR spectroscopy, we suppose that either poly-
morphic forms are present or that slight water uptake and/
or polymerisation have taken place. We also note that the
crystallinity is not very high in view of the large size of the
diffraction peaks. A full Rietveld refinement of the powder
diffractograms was attempted with the molecules 1, 3 and 5
parameterised as rigid bodies. Although the refinements
were stable, the Rietveld agreement factors were not very
low (ꢁ20%), presumably due to the quality of the diffrac-
tograms, which were degraded by the presence of impurity
phases and rather large peaks. For compounds 2 and 4 no
match could be made between the single-crystal structures
and the X-ray powder diffractograms. Therefore the cell pa-
rameters of a number of single crystals were tested in order
to verify that not an isolated polymorphic variant was
picked up to do the full single-crystal structure determina-
tion. Since this was not the case, ab initio indexing of the
powder patterns was attempted. In the case of 4, a success-
ful indexing was obtained (M16=27.1) with cell parameters
a=15.09, b=11.56, c=13.09 Š, a=90, b=98.29 and g=908,
3
for which the volume, V=2260.7 Š3, is interestingly 220 Š
larger than the room-temperature volume of the single-crys-
tal structure of 4; this could correspond to the uptake of
three water molecules. For 2 a successful indexation could
not be made, which is not surprising in view of its low crys-
tallinity. Compounds 2 and 4 containing indole and methoxy-
phenyl residues present a second labile interface defined by
weak hydrophobic interactions between aromatic units.
Simple precipitation of these powders in organic solvents
will induce formation of solvated or hydrated polymorphic
solids, of different orientations in the solid phase.
Figure 3. Crystal structures of compounds 1–5: stick representation of
crystal packing of a) 1, b) 2, c) 3, d) 4, and e) 5; hydrogen-bonds green
dotted lines; O1: crystal packing of the organic part; S1, S2: crystal pack-
ing of the alkoxysiloxane moieties. The thicknesses of O1, S1 and S2 are
about 3 Š.
Chem. Eur. J. 2008, 14, 1776 – 1783
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1779

M. Barboiu et al.
Unfortunately, the XPRD data show that the hybrid ma-
terials MB1–MB4 and MB1*–MB4* have a different struc-
ture compared with their crystalline precursors. The struc-
ture change is not surprising, considering that hydrolysis/
condensation reactions occur during the sol–gel process.[15,16]
For all the hybrids, a successful indexation could not be
made, which is not surprising in view of their low crystallini-
ty, confirming the partial loss of order of solvated powders.
However, the XRPD patterns of the hybrids show well-re-
solved peaks, indicating that they are still ordered (Support-
ing Information, Figure 4S).
Scanning electron microscopy (SEM) reveals that the
thin-layer films (thickness of about 2–3 mm) were dense,
without pinholes (Figure 4a). Figure 4b depicts a typical
TEM image of an MB4 film and shows well-resolved and
highly ordered uniform rows with several orientations for
the bulk crystalline areas. Since crystallographic symmetry
could easily mask a particular TEM distance, it is difficult to
compare these distances with those obtained from XPRD
patterns of hybrid materials. However, it is clear that the re-
peating motif is formed by discrete alternative light (inor-
ganic siloxane matrix) and dark (organic self-assembled
molecules) rows, with a periodicity in one direction of about
3 Š. The match between these distances and the periodicity
of parallel sheets of the self-organised organic and siloxane
inorganic networks observed in the crystal packing indicates
that the hybrid materials have the same self-organised lay-
ered arrangement as observed in the crystal structures
(Figure 3).
The X-ray structural data and TEM analysis allow the fol-
lowing conclusions to be drawn:
1) The crystal structures of 1–5 indicate that the arrange-
ment is mainly defined by periodic parallel sheets, result-
ing from the alignment of hydrophobic organic and inor-
ganic silica layers. To our best knowledge, this is the first
example of preorganisation of hybrid precursors in which
a specific mutual communication between the organic
and inorganic networks of hybrid material is defined.
2) A very good compatibility between the wet bulk powders
X-ray fingerprint and the single-crystal structure has
been found for precursors 1, 3 and 5; some deviations
from the crystallinity found in the XPRD data of 2 and 4
might either arise from the different orientations of the
organic hydrophobic interfaces themselves or from inter-
calation of solvent or water molecules, defining different
polymorphic solids.
3) Hybrid materials MB1–MB4 and MB1*–MB4* define
particularly attractive systems: multioriented nano-
domains are well defined (see Figure 4a) starting from
molecular components that encode the required informa-
Figure 4. a) SEM image of a hybrid membrane cross-section: the thin-
layer dense film deposited onto polymeric PVDF support; b) TEM
images of the hybrid material MB4; c) Top view and d) side view of the
crystal packing of compound 4 in stick representation with a soft surface.
1780
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1776 – 1783

Self-Organised Hybrid Membranes
FULL PAPER
tion for ionic assisted diffusion within hydrophilic path-
ways (represented in blue in Figure 4c,d); they are ori-
ented parallel along the hydrophobic organic layers (rep-
resented in yellow in Figure 4c,d) and sandwiched be-
tween the two hydrophilic siloxane layers (represented
in magenta in Figure 4c,d).
The phenyl-type MB1 membrane can be considered as a
reference membrane in which the transport of ions occurs
through a siloxane network containing residual hydrophilic
ꢀ
Si OH groups. The transport of protons (Figure 5a) and
alkali cations (Figure 5b) is amplified when phenolic (MB3)
or methoxyphenyl (MB4) groups are present in the mem-
brane. This is consistent with the development of supple-
mentary organic ion-conducting pathways of self-organised
ꢀ
ꢀ
Dialysis membrane transport experiments: To demonstrate
the generation of hydrophilic pathways into hybrid mem-
branes we carried out dialysis transport experiments.
(Figure 5).
OH and OMe groups, assisting the diffusion into the
hybrid network. The phenolic groups form strong intramo-
ꢀ
ꢀ
lecular hydrogen-bonds with the adjacent OH, SiOH and
ꢀ
ꢀ
SiOSi moieties. This hampers the interactions with sur-
rounding water molecules, and makes the phenolic-type
membrane MB3 more lipophilic than the corresponding
methoxyphenyl one MB4. Consequently, the protons are
A
better transported along methoxyphenyl pathways within
the MB4 membrane. The dialysis transport experiments
were performed by equilibrating the membranes with water.
At equilibrium the water uptake might depend on the hy-
drophilic nature of aromatic group, but the thickness of our
active hybrid layers (2–3 mm) do not permit an accurate de-
termination of the amount of water with reference to the
supporting membrane. The proton transfer in the membrane
+
ꢀ
occurs by diffusion of O H !O bonds inside the network
of hydrophylic pathways, similar to proton transport in
water.[2]
+
The Na
A
ions are preferentially transported
through MB1, MB3 and MB4 membranes. Since the self-
[21a]
diffusion coefficients DNa
+
<DK ,
+
the transport
(H2O)n
N
of the hydrated cations is probably governed by the solva-
tion free energy, which is thermodynamically more favoura-
+
+ [21b]
ble for Na
G
(H2O)n
.
Certainly, a substan-
tial contribution to sodium selectivity arises from exchange
features between the hydrated ions and hydrophilic moieties
located in the membrane. Both hydrated cations are better
transported along the more hydrophobic phenolic pathways
in the MB3 membrane. This suggests that they may stay
within the hydrophilic methoxy pathways of the MB4 mem-
brane.
Compared with the reference membrane MB1 the trans-
port rates of both cations across the indole-type membrane
+
MB2 decrease. MB2 is selective for K
A
tainly, a substantial contribution to this selectivity arises
from exchange features between the hydrated ions and the
hydrophobic indole pathways: the complexation of ions by
cation–p interactions is thermodynamically more favourable
for K+ ions than Na+.[21c] Many groups, including our
own,[7a,21c,22] have found that the pyrolo subunit of the indole
(i.e. C2=C3 bond) is a strong p-donor group for the K+ ion.
This is probably why K+ ions are preferentially transported
within the “indole pathway” of membrane MB2.
Figure 5. a) pH versus time transport profiles across membranes MB1,
MB3, and MB4; b) Permeabilities of MB1*–MB4* hybrid membranes
calculated from experimental data using the solution–diffusion model.[20]
The single and competitive H+, Na+ and K+ transport ex-
periments present similar features for both classes of materi-
als MB1–MB4 and MB1*–MB4*. In a classical diffusion
model the hydrated ions diffuse through the membrane and
the mechanism is based on a dual transport mode: solution–
diffusion.[20] The transport rate is thus governed by diffusion
coefficients and by the interaction equilibria of the hydrated
ions losing part of the hydration shell in exchange for the
specific complexant groups located in the membrane.
Conclusion
The controlled formation of such directional transporting
devices makes the hybrid membrane materials presented
Chem. Eur. J. 2008, 14, 1776 – 1783
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1781

M. Barboiu et al.
J. C. Raisaiah, G. Hummer, Proc. Natl. Acad. Sci. USA 2004, 101,
17002–17005.
[2] S. Cukierman, Biophys. J. 2000, 78, 1825–1834.
[3] R. Mackinnon, Angew. Chem. 2004, 116, 4363–4377; Angew. Chem.
Int. Ed. 2004, 43, 4265–4277.
here of interest for the development of a new supramolec-
ular approach to nanoscience through self-organisation,
toward hybrid materials of increasing addressability.
The hybrid membrane materials reported in this paper
are composed from nanodomains randomly ordered in the
hybrid matrix. These oriented nanodomains (Figure 4a)
result from the controlled self-assembly of simple molecular
components that encode the required information for ionic
assisted diffusion within hydrophilic pathways. The novelty
of these findings with respect to our previous work[16] with
complexant crown ether precursors is that simple molecules,
which collectively define transporting devices by self-assem-
bly, can be successfully used to transfer the overall function-
ality of their supramolecular self-organisation in hybrid
membrane materials by sol–gel transcription. Our results
imply that the control of molecular interactions can define
the self-organised supramolecular architectures presenting a
strong communication between the organic and the siloxane
layers. We have shown that specific hydrogen-bonding com-
munications between these networks have the ability to
create hydrophilic pathways of different chemical properties
in the self-organised superstructures of hybrid membrane
materials. They are essential in the diffusion process and in
the selectivity of the transport of hydrated alkali cations.
[4] P. Agre, Angew. Chem. 2004, 116, 4377–4390; Angew. Chem. Int.
Ed. 2004, 43, 4278–4290.
[5] D. A. Dougherty, Science 1996, 271, 163–168.
[6] J.P Gallivan, D.A Dougherty, Proc. Natl. Acad. Sci. USA, 1999, 96,
9459–9464.
[7] a) G. W. Gokel, A. Mukhopadhyay, Chem. Prod. Chem. Soc Rev.
2001, 30, 274–286; b) N. Voyer, Top. Curr. Chem. 1996, 184, 1–35.
[8] D. T. Bong, T. D. Clark, J. R. Granja, M. R. Ghadiri, Angew. Chem.
2001, 113, 1016–1041; Angew. Chem. Int. Ed. 2001, 40, 988–1011.
[9] a) S. Matile, Chem. Soc. Rev. 2001, 30, 158–167; b) N. Sakai, J.
Mareda, S. Matile, Acc. Chem. Res. 2005, 38, 79–87.
[10] a) P. K. Eggers, T. M. Fyles, ; K. D. D. Mitchell, T. Sutherland, J.
Org. Chem. 2003, 68, 1050–1058; K. D. D. Mitchell, T. Sutherland, J.
Org. Chem. 2003, 68, 1050–1058; b) T. M. Fyles, C. Hu, R. Knoy,
Org. Lett. 2001, 3, 1335–1337.
[11] a) L. E. Cheruzel, M. S. Pometum, M. R. Cecil, M. S. Mashuta, R. J.
Wittebort, R. M. Buchanan, Angew. Chem. 2003, 115, 5610–5613;
Angew. Chem. Int. Ed. 2003, 42, 5452–5455; b) Z. Fei, D. Zhao, T. J.
Geldbach, R. Scopelliti, P. J. Dyson, S. Antonijevic, G. Bodenhau-
sen, Angew. Chem. 2005, 117, 5866–5871; Angew. Chem. Int. Ed.
2005, 44, 5720–5725.
[12] L. Cronin, Angew. Chem. 2006, 118, 3656–3658; Angew. Chem. Int.
Ed. 2006, 45, 3576–3578.
[13] Chem Rev. 2005, 105, Special Issue on Functional Nanostructures.
[14] V. Percec, A. E. Dulcey, M. Peterca, M. Ilies, S. Nurnmelin, M. J.
Sienkowska, P. A. Heiney, Proc. Natl. Acad. Sci. USA 2006, 103,
2518–2523.
A
lthough these pathways do not merge to cross the micro-
metric films, they are well defined along nanometric distan-
ces. This is reminiscent of the supramolecular organisation
of binding sites in channel-type proteins collectively contri-
buting to the selective translocation of solutes along hydro-
philic pathways.[2–4]
Prospects for the future include the development of origi-
nal methodologies for such communicative transcription in
silica matrixes of the supramolecular functional devices, pre-
senting a greater degree of structural complexity.
[15] a) C. Sanchez, B. Julian, P. Belleville, M. Popall, J. Mater. Chem.
2005, 15, 3559–3592; b) C. Sanchez, F. Ribot, New J. Chem. 1994,
18, 1007–1047; c) N. Liu, D. R. Dunphy, M.A. Rodriguez, S. Singer,
J. Brinker, Chem. Commun., 2003, 1144–1145; d) K. J. C. van Bom-
mel, A. Frigerri, S. Shinkai, Angew. Chem. 2003, 115, 1010–1030;
Angew. Chem. Int. Ed. 2003, 42, 980–999; e) R. J. P. Corriu, Eur. J.
Inorg. Chem. 2001, 5, 1109–1121; f) F. Ben, B. Boury, R. J. P. Corriu,
P. Delord, M. Nobili, Chem. Mater. 2002, 14, 730–738; g) M. P.
Kapoor, Q. Yang, S. Inagaki, J. Am. Chem. Soc. 2002, 124, 15176–
11577; h) J. J. E. Moreau, L. Vellutini, M. Wong Chi Man, C. Bied,
J. L. Bantignies, P. DieudonnØ, J. L. Sauvajol, J. Am. Chem. Soc.
2001, 123, 7957–7958; i) J. J. E. Moreau, L. Vellutini, M. Wong Chi
Man, C. Bied, Chem. Eur. J. 2003, 9, 1594–1599; j) J. J. E. Moreau,
B. P. Pichon, H. Pritzkow, M. Wong Chi Man, C. Bied, J. L. Bantig-
nies, P. DieudonnØ, J. L. Sauvajol, Angew. Chem. 2004, 116, 205–
208; Angew. Chem. Int. Ed. 2004, 43, 203–206; k) O. J. Dautel, J. P.
L›re-Porte, J. J. E. Moreau, M. Wong Chi Man, Chem. Commun.
2003, 2662–2663.
[16] a) M. Barboiu, G. Vaughan, A. van der Lee, Org. Lett. 2003, 5,
3073–3076; b) M. Barboiu, J. Inclusion Phenom. Macrocyclic Chem.
2004, 49, 133–137; c) M. Barboiu, S. Cerneaux, A. van der Lee, G.
Vaughan, J. Am. Chem. Soc. 2004, 126, 3545–3550; d) A. Cazacu, C.
Tong, A. van der Lee, T. M. Fyles, M. Barboiu, J. Am. Chem. Soc.
2006, 128(29) 9541–9548; e) C. Arnal-HØrault, A. Banu, M. Bar-
boiu, M. Michau, A. van der Lee, Angew. Chem. 2007, 119, 4346–
4350; Angew. Chem. Int. Ed. 2007, 46, 4268–4272; f) C. Arnal-HØr-
ault, M. Barboiu, A. Pasc, M. Michau, P. Perriat, A. van der Lee,
Chem. Eur.J 2007, 13, 6792–6800.
ExperimentalSection
Experimental data are available in the Supporting Information. This ma-
terial includes the general procedure of synthesis and characterisation
data (NMR spectroscopy, ESI-MS, elemental analysis) of precursors 1–5
and hybrid membrane materials MB1–MB4 and MB1*–MB4*. In addi-
tion details of X-ray single-crystal diffraction and X-ray powder diffrac-
tion, and membrane transport measurements are given.
Acknowledgements
This work, conducted as part of the award “Dynamic adaptive materials
for separation and sensing Microsystems” (M.B.) made under the Euro-
pean Heads of Research Councils and European Science Foundation
EURYI (European Young Investigator) Awards Scheme in 2004, was
supported by funds from the Participating Organisations of EURYI and
the EC Sixth Framework Program (see www.esf.org/euryi). This research
was supported in part by CNRS and University of Montpellier 2 Mont-
pellier, France.
[17] a) M. C. Etter, Acc. Chem. Res. 1990, 23, 120–126; b) J. Van Esch, F.
Schoonbeek, M. De Loos, H. Kooijman, A. L. Spek, R. M. Kellog,
B. L. Feringa, Chem. Eur. J. 1999, 5, 937–950; c) F. Schoonbeek, J.
Van Esch, B. Wegewijs, D. B. A. Rep, M. P. De Haas, T. M. Klap-
wijk, R. M. Kellog, B. L. Feringa, Angew. Chem. 1999, 111, 1486–
1490; Angew. Chem. Int. Ed. 1999, 38, 1393–1397.
[18] Equilibration between hydrogen-bonded and non-hydrogen-bonded
[1] a) F. Hucho, C. Weise, Angew. Chem. 2001, 113, 3194–3211; Angew.
Chem. Int. Ed. 2001, 40, 3100–3116; b) S. Vaitheeeswaran, H. Yin,
ꢀ
states for a given N H proton is almost always fast on the NMR
1782
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1776 – 1783

Self-Organised Hybrid Membranes
FULL PAPER
timescale, and observed proton chemical shifts are weighted averag-
es of the chemical shifts of contributing states.
[21] Diffusion coefficients in water: DNa+ =1.334ꢁ10ꢀ9 m2 s, DK+ =1.957ꢁ
10ꢀ9 m2 s, a) Handbook of Chemistry and Physics, 76th ed., (Ed.:
D. R. Lide), 1995/1996, pp. 5–92; b) C. P. Kelly, C. J. Cramer, D. G.
Truhlar, J. Phys. Chem. B 2006, 110, 16066–16081; c) O. M. Caraba-
cos C. J. Weinheimmer, J. M. Lisy, J. Chem. Phys. 1999, 110, 8429–
8436.
[19] The hydrogen-bonding interactions between organic molecules and
the silica precursor are determinant for the morphology of the re-
sulted superstructures in the sol–gel transcription processes using
heterogeneous organogelator–TEOS systems. See, for instance, J. H.
Jung, M. Amaike, S. Shinkai, Chem. Commun. 2000, 2343–2344.
[20] a) M. Barboiu, C. Guizard, C. Luca, B. Albu, N. Hovnanian, J. Pal-
meri, J. Membr. Sci. 1999, 161, 193–206; b) M. Barboiu, C. Guizard,
N. Hovnanian, J. Palmeri, C. Reibel, C. Luca, L. Cot, J. Membr. Sci.
2000, 172, 91–103; c) M. Barboiu, C. Guizard, C. Luca, N. Hovnani-
an, J. Palmeri, L. Cot, J. Membr. Sci. 2000, 174, 277–286.
[22] C. Arnal-Herault, M. Barboiu, E. Petit, M. Michau, A. van der Lee,
New J. Chem. 2005, 29, 1535–1539.
Received: June 20, 2007
Revised: October 5, 2007
Published online: December 17, 2007
Chem. Eur. J. 2008, 14, 1776 – 1783
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1783