Supramolecular self-organization in constitutional hybrid materials

Article abstract of DOI:10.1039/b908641m

We present in this paper a new strategy to transcribe the supramolecular dynamic self-organization of the G-quadruplex and ureidocrown ether ion channel-type columnar architectures in constitutional hybrids. In particular, the use of a "dynamic reversible

Full text of DOI:10.1039/b908641m

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Supramolecular self-organization in constitutional hybrid materialsw
Simona Mihai, Adinela Cazacu, Carole Arnal-Herault, Gihane Nasr, Anca Meffre,
Arie van der Lee and Mihail Barboiu*
Received (in Montpellier, France) 14th April 2009, Accepted 4th September 2009
First published as an Advance Article on the web 25th September 2009
DOI: 10.1039/b908641m
We present in this paper a new strategy to transcribe the supramolecular dynamic
self-organization of the G-quadruplex and ureidocrown ether ion channel-type columnar
architectures in constitutional hybrids. In particular, the use of a ‘‘dynamic reversible hydrophobic
interface’’ can render the emerging hybrid mesophases self-adaptive. The reversible hydrophobic
interactions allow both supramolecular and inorganic silica components to mutually
(synergistically) adapt their spatial constitution during simultaneous (collective) formation of
micrometric self-organized hybrid domains. More generally, applying such consideration to
materials, leads to the definition of constitutional hybrid materials, in which organic
(supramolecular)/inorganic domains are reversibly connected. This might provide new insights
into the features that control the design of functional materials.
templates during the tetraethoxysilane (TEOS) sol–gel
process.7d We have recently reported nucleobase-based hybrids
1. Introduction
obtained under thermodynamic control.4d A second strategy
Convergent multidimensional self-assembly of molecular
components has been extensively used during the last few
presented in this paper is to use a reversible non-covalent
decades for the synthesis of functional supramolecular
interface between the supramolecular/organic and the siloxane/
devices.1 Kinetic or thermodynamic resolutions,2,3 self-assembly
inorganic constituents. As the self-organization of organic/
followed by covalent modification,4 and crystallization5,6 shed
inorganic domains has to explore the hypersurface of all
structure/energy combinations,1 the build-up of final hybrid
light on useful strategies to control and to make convergence
between self-organization and supramolecular functions. In
material upon a growing/self-reparation/termination sequence
this context, connecting the self-assembly processes with
might be able to select the correct spatial geometries for the
further covalent sol–gel modification might be considered
supramolecular organic and inorganic entities emerging from
as one straightforward approach used for the design of
a collection of building blocks. A dynamic reversible interface
functional hybrid materials.5–7 Considerable challenges still lie
might mediate the structural self-correlation of supramolecular
ahead and the more significant one is the ‘‘dynamic marriage’’
and inorganic domains by virtue of their basic constitutional
between supramolecular self-assembly and the sol–gel process.
behaviours. The resulting hybrid material might undergo
Both supramolecular and inorganic polymerization processes
continual change in its constitution, through dissociation/
might ‘‘correlate’’ in order to converge towards self-organized
reconstitution of different mesophases during the sol–gel
nanomaterials. By using silsesquioxane precursors (Scheme 1)
process. This concept can be related to a sort of ‘‘Chemical
in which the functional organic and siloxane moieties are
collectivism’’ of colloidal systems, as described some years ago
by Menger.3a Convergently, the self-replicating molecular and
covalently connected during the sol–gel process, the weak
supramolecular interactions positioning of the molecular
‘‘dynablock’’ systems very recently described by Giussepone
et al.3b,c are characterized by their aptitude to continuously
components are typically less robust and they usually are
destabilized when the cross-linked covalent siloxane network
organize at the macroscopic level the generation of their own
forms. The resulting hybrid architectures lead, in general, to
components, when different blocks form in coupled equilibria.
amorphous or polymorphic materials.7 One solution to overcome
The reversibility of the interactions between components of a
these difficulties is to improve the binding efficiency of
hybrid material might be a crucial factor and accordingly,
molecular components. An increased number of interactions
a dynamic reversible hydrophobic interface might render
between molecular components may improve the stability of
the emergence of organic/inorganic mesophases self-adaptive,
the templating supramolecular systems within the inorganic
which mutually (synergistically) may adapt their 3D spatial
siloxane network. Numerous successful examples, pioneered
distribution based on their own structural constitution
by Shinkai et al. concern the use of organogels acting as robust
during the simultaneous formation of hybrid self-organized
domains.
In this context, we will present in this paper two examples of
such constitutional hybrid materials prepared under thermo-
dynamic control, containing supramolecular ion channel-type
architectures combined with hydrophobic siloxanes as precursors
for the inorganic silica matrix:
Institut Europeen des Membranes, ENSCM/UMII/UMR CNRS 5635,
IEM/UMII, Place Eugene Bataillon CC047.
E-mail: mihai.barboiu@iemm.univ-montp2.fr;
Fax: (+33) 4 67 14 91 18
w Electronic supplementary information (ESI) available: Supplementary
FTIR, NMR and powder XRD data. See DOI: 10.1039/b908641m
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2335–2343 | 2335

View Article Online
Scheme 1 Sol–gel process of (i) classical and (ii) constitutional hybrid materials based on (a) G-quadruplex and (b) ureidocrown ether
architectures.
(a) The G-quartet, the hydrogen-bonded macrocycle formed
by the self-assembly of guanosine, is stabilized by alkali
cations.8 (Scheme 1a) The role of cation templating is to
stabilize by coordination to the eight carbonyl oxygens of
two sandwiched G-quartets, the G-quadruplex, the columnar
device formed by the vertical stacking of four G-quartets.
It represents a dynamic supramolecular system, in equilibrium
with linear oligomers. New strategies using reversible poly-
merization or phase-change-driven selection were successfully
used to generate G-quadruplex nanostructures.8 We have recently
reported a sol–gel route for preparing G-quadruplex-hybrids4a
or dynameric membranes.4b
(b) The heteroditopic ureidocrown ethers9a complexing
both cations and anions, and self-organizing into columnar
oligomeric superstructures in solution and the solid state
(Scheme 1b). Their self-organization in the liquid9b or bilayer
membranes9c provided evidence for a hybrid carrier/channel
transporting mechanism. Furthermore, these dynamic systems
have been ‘‘frozen’’ in a hybrid matrix by a sol–gel process,
transcribing their dynamic columnar self-organization in solid
dense hybrid nanomembranes.9d
performed on a Brucker ASX400 spectrometer. Mass spectro-
metric studies (60 1C, cone voltage of Vc = 5–10 V) were
performed by using a quadrupole Micromass, Platform II.
Samples were introduced through a Waters 616HPLC pump.
Solid state X-ray powder diffraction experiments were
performed with Cu-Ka radiation at 20 1C using a Philips
X’Pert Diffractometer equipped with an Xcelerator detector.
CD spectra (220–400 nm) were collected on a MOS-450/AF-
Biologic-CD spectrometer with a rate of 0.5 nm per step, using
10 mm quartz cells, at 25 1C. The measurements were carried
out on CHCl3 solutions (5 ꢁ 10ꢂ4 M of 2) and on mixtures of
1/2 (1 : 1 and 1 : 2 mol/mol) briefly ultrasonicated with K+ and
Ba2+ ions (0.25 or 1 eq.).
2.2 General procedure for synthesis of compounds 1* and 1
9-(20,30-O-Benzylidene-b-D-ribofuranosyl)-guanine, 1*. Guano-
sine (5.0 g, 17.65 mmol) was suspended in benzaldehyde
(50 ml) and dry zinc chloride (13.25 g, 97.18 mmol) was added.
The mixture was stirred at room temperature overnight. Then,
diethyl ether was added to precipitate the product. The crude
material was filtered off and washed with cold water. The crude
solid as mixture of diastereoisomers was recrystallized from
1
ethanol–water (3 : 1) to give a pale yellow product (84%). H
2. Experimental section
NMR (DMSO-d6, 300 MHz) d 10.65 (s, 1H); 7. 90 and 7.87 (2s,
1H); 7.42–7.35 (m, 5H); 6.46 (s, 2H); 6.11 and 5.94 (2s, 1H); 6.02
(2d, 1H); 5.31 (m, 1H); 5.06 (m, 2H); 4.23 (m, 1H); 3.55 (m, 2H).
13C NMR (DMSO-d6, 300 MHz): d 156.70; 153.66; 150.69;
136.15; 136.08; 129.75; 126.80; 116.74; 106.51; 102.99; 88.32;
87.68; 86.55; 84.49; 84.34; 83.16; 82.36; 80.80; 61.66.
2.1 Materials and methods
Guanosine, benzaldehyde stearoyl chloride, octadecylisocyanate,
octadecyltrimethoxy-silane (2), (triethoxysilyl)propan-1-amine
and 40-aminobenzo-15-crown-5 were purchased from Aldrich.
4a
9d
GSi and ECSi were synthesised as previously reported.
1H NMR and 13C NMR spectra were recorded on an ARX
300 MHz spectrometer in CDCl3 and DMSO-d6 with the use
of the residual solvent peak as reference. 29Si MAS NMR was
9-(20,30-O-Benzylidene-b-D-ribofuranosyl)-50stearoyl guanine, 1.
Compound 1* (2.0 g, 5.36 mmol), as mixture of diastereoisomers,
was dissolved in 40 ml of anhydrous dimethylformamide; then,
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2336 | New J. Chem., 2009, 33, 2335–2343

View Article Online
dimethylaminopyridine (0.2 g, 1.61 mmol) and triethylamine
(1.63 g, 16.08 mmol) were added. After 10 min, stearoyl chloride
(2.27 g, 7.5 mmol) was slowly added. The mixture was stirred at
room temperature overnight. The crude mixture was removed by
rotary evaporation and chloroform was added. The crude
material was purified by flash chromatography on silica gel
(gradient eluent CHCl3–MeOH 8 : 2). The fractions were
removed and the product was washed with methanol to give
pale yellow product (89%). 1H NMR (DMSO-d6, 300 MHz)
d 10.71 (s, 1H); 7.88 (s, 1H); 7.55–7.44 (m, 5H); 6.53 (s, 2H);
6.19 (s, 1H); 6.12 and 6.02 (2s, 1H); 5.41–5.31 (m, 2H); 4.48–4.18
(m, 3H); 2.26 (m, 2H); 1.47 (m, 2H); 1.23 (m, 28H); 0.84 (t, 3H).
13C NMR (DMSO-d6, 300 MHz): d 174.94; 159.11; 153.68;
149.83; 135.02; 129.44; 128.14; 126.15; 115.92; 107.16; 103.73;
91; 90; 62.62; 33.83; 31.22; 22.20; 13.63.
29.6, 29.7, 31.9, 42.5, 45.7, 58.4, 157.7; ES-MS: m/z (%):
531.45 M + H+.
2.5 General procedure for synthesis of Ai–Bi hybrid materials
i = 0, 10, 20, 40, 50, 60, 80% w/w can be achieved by mixing
compounds 3 and 4 in acetone (ESI, Table 1Sw), followed by
the sol–gel process performed for 7 d at room temperature and
using 6 equiv. of deionized water and 4 N HC for hybrid
materials Ai, and 3 equiv. of benzylamine for hybrid materials
Bi as catalysts (ESI, Table 2Sw). The sols were briefly stirred
and allowed to react at room temperature under static conditions
for 1 week and then samples were slowly evaporated at room
temperature to give hybrid materials as white powders.
3. Results and discussion
3.1. Constitutional G-quadruplex hybrid materials
+
2.3 General procedure for synthesis of AK+–DK
,
ABa2+–DBa and the reference E hybrid materials
2+
3.1.1 Design of hybrid molecular components. The lipophilic
stearoyl guanosine, 1, was prepared in two steps: the furanose
protection with benzaldehyde leading to acetal 1* as an equimolar
diastereoisomeric mixture,10a followed by the reaction with
stearyl chloride, to afford compound 1 as a white powder. The
1H and 13C NMR, and ESI-MS spectra are in agreement with
the proposed formula.
1H NMR experiments in d6-DMSO indicated low defined
self-association of ‘‘monomeric’’ 2, as previously reported.8
1H NMR spectra of mixtures 1/K+ and 1/Ba2+ in CDCl3
consist of broadened signals over temperatures from ꢂ60 1C
to 50 1C, corresponding to fast interexchanging species in
solution at the NMR time scale. The NMR signal of exchangeable
NH1 protons is broadened into the baseline at around d
9.5–10.5 ppm (ESI, Fig. 1Sw). After equilibration with K+
or Ba2+, the 1H NMR spectrum in CDCl3 presented sharp
resonances for NH1 amide protons (d 9.9 and 10.06 ppm) and
for NH2 amine protons (d 8.7 ppm), together with still present
broad signals (ESI, Fig. 1Sw). The fast exchange in solution
was confirmed by the NOESY spectrum of such mixtures,
which did not show clear NOEs between NH1 and the H8, as
previously observed by Davis et al.8l
In typical run, 1 equiv. of octadecyltrimethoxysilane (2) and
variable amounts of compound 1 together with the templating
metal salt (see the ESI, Table 1Sw) were dissolved in
chloroform–acetone (1 : 0.25). 6 equiv. of deionized water
and 3 equiv. of benzylamine (as catalyst) were added. Mixtures
were stirred at 60 1C for 24 h and then the samples were
slowly evaporated at room temperature to give sticky brown
products, which were isolated and characterized. A second
series of hybrids AK+–DK+, ABa2+–DBa2+, E were thermally
treated at 90 1C for 24 h. A reference sol–gel sample, E, was
prepared in the absence of salt triflates, under the same
reaction conditions.
2.4 General procedure synthesis of compounds 3 and 4
4-Octadecylurea-benzo-15-crown-5, 3, was prepared by
refluxing 0.27 g (0.92 mmol) of octadecylisocyanate and
0.2 g (0.71 mmol) of 40-aminobenzo-15-crown-5 (1.5 : 1
mol/mol) in acetonitrile (12 h). The crude product was
recrystallized in hexane to afford, as a white powder, 3 (90%
yield). 1H NMR (CDCl3, ppm) d 0.88 (t, 3H, CH3), 1.29
(m, 30H, CH2), 1.49 (m, 2H, CH2), 3.16 (m, 2H, CH2), 3.75
(m, 8H, –O–CH2–CH2–O–), 3.90 (m, 4H, –O–CH2–CH2–O–),
4.11 (m, 4H, –O–CH2–CH2–O–), 4.70 (t, 1H, NH-Alkyl), 6.17
(s, 1H, NH-Aryl), 6.68 (dd, 1H), 6.82 (d, 1H), 6.95 (d, 1H).
13C NMR (CDCl3 ppm) d 13.62, 22.19, 26.44, 28.87, 29.21,
29.68, 30.443, 31.42, 39.87, 68.13, 68.82, 69.95, 70.36, 108.58,
114.40, 122.90, 131.98, 145.66, 149.11, 156.04, 206.54. ES-MS:
m/z (%): 601.62 (100) M + Na+.
Octadecylureidopropyltrimethoxysilane, 4, was prepared by
refluxing 0.2 g (0.675 mmol) octadecylisocyanate and 0.179 g
(0.811 mmol) of (triethoxysilyl)propan-1-amine in chloroform
(12 h). The solvent was evaporated and the crude product was
recrystallized in hexane to afford, as a white powder, 4 (82%
1
yield). H NMR (CDCl3, ppm) d 0.84 (m, 2H, aCH2–Si–O–),
0.96 (m, 3H, CH3 terminal), 1.16 (m, 2H, bCH2–Si–O–), 1.22
(m, 6H, –Si–OR), 1.26 (m, 30H, CH2), 1.33 (m, 2H, aCH3
terminal), 3.24 (m, 2H, CH2–NH 1et m, 2H, CH2–NH2), 3.83
(m, CH3, –Si–OR), 3.98 (d, 1H, NH 1), 4.24 (d, 1H, NH 2);
13C NMR (CDCl3 ppm) d 7.5, 14.1, 11.4, 22.9, 24.8, 26.4, 29.4,
Fig. 1 The CD spectra at room temperature in CHCl3 of lipophilic
guanosine 1 (gray line), 1 + KOTf or 1 + BaOTf2 (gray dashed line);
mixture of 1/2 (black dotted line) and mixtures of 1/2 + KOTf or of
1/2 + BaOTf2 (black line).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2335–2343 | 2337

View Article Online
Next, we used CD spectroscopy to gain evidence that 1
forms G-quadruplex structures in the presence of ions in
CHCl3 (Fig. 1). In accordance with the NMR results, no
active CD bands specific to a G-quadruplex were observed
when 1 was mixed with ionic salts (KOTf and BaOTf2) or with
compound 2, indicating that the G-quadruplex is either
unstructured or exchanging very fast in solution. However,
the observed CD spectra of such mixtures presented strong
negative active CD bands at 295 and 305 nm as consequence of
equilibrium between the two diastereoisomers in solution.10a,b
Interestingly, mixtures of 1 with hydrophobic compound 2
(1 : 1 and 1 : 2 mol/mol) equilibrated with K+ and Ba2+ ions
gave the CD signature of a G-quadruplex with a positive band
at l = 270 nm and a negative band at l = 285 nm, closely
resembling the previously calculated CD spectra by Gray
et al.10c of G-quadruplexes whose adjacent G-quartets have
opposite polarities.
reactions of the alkoxysilane groups during the sol–gel
process. 29Si MAS NMR spectroscopic experiments are in
agreement with the B100% T3 C–Si(OSi)3 units, corresponding
to a totally condensed silica matrix in hybrid materials A–E
after the thermal treatment.
Further valuable insights into the self-organization
+
and reference E hybrid
2+
of AK+–DK or ABa2+–DBa
materials were obtained by X-ray powder diffraction
(powder XRD).
Firstly, the powder XRD patterns of lipophilic guanosine 1
derivative, equilibrated with salts (Na+, K+, Ba2+) display
broad peaks and the presence of the Bragg diffraction peak at
2y = 26.81 (d = 3.41 A) is representative for the p–p stacking
distance between two planar G-quartets (ESI, Fig. 5Sw).4a,b
The XRPD patterns of the both G-quadruplex hybrids
+
2+
AK+–DK and ABa2+–DBa
presents well-resolved peaks.
An indexation was attempted for several observed powder
diffraction diagrams (Fig. 2).
+
The stabilization of G-quadruplex structures in a hydro-
phobic environment occurred in a similar manner to that
already observed in a hydrophobic environment, such as
phospholipid liposomes.8i,k,11 Having shown that mixtures of
1/2 equilibrated with K+ and Ba2+ ions form G-quadruplex
superstructures, we next evaluated their ability to form higher
order superstructures in the solid state, by using a sol–gel
selection process.4a The generation of G-quadruplex hybrid
For DK (Fig. 2a,c), all peaks could be indexed with a
monoclinic unit cell of a = 15.35, b = 4.14, c = 23.97 A,
b = 94.011, possible space groups P2 or P21 (in the latter
case the b parameter should be doubled, the quality of
the diffractogram does not allow the two cases to be
distinguished). With a unit cell volume of 1520 A3, the
independent molecule in the asymmetric unit is expected
to contain 40–45 non-H atoms, slightly less than the
48 atoms of 1. In the case of a doubled b-parameter, and
hence a double cell volume, the asymmetric unit of the unit cell
incorporates 84 non-H atoms, which is very close to the 81
independent atoms of the complete hybrid material of 1 and 2
combination and which is thus more logic than the first
hypothesis.
+
2+
materials AK+–DK
or ABa2+–DBa
can be achieved
by mixing guanosine stearoyl derivative, 1, and octadecyl-
trimethoxysilane, 2, compounds with potassium or barium
triflates in chloroform/acetone, followed by the sol–gel process
performed at 60 1C and using benzylamine as catalyst
(Scheme 1a and ESI, Table 1Sw). We also carried out sol–gel
process, in the absence of K+ or Ba2+ cations, resulting in the
formation of a G-hybrid material, E. A portion of these
materials were then dried at 90 1C for 24 h in order to achieve
complete condensation of inorganic siloxane network.
2+
The diagrams in Fig. 2b,d corresponding to ABa2+–DBa
are easily indexed on the basis of a repeat unit of approxi-
mately 48 A between the individual stacks. This feature was
not observed for the XRPD patterns of the non-templated
G-hybrid material, E (Fig. 2, E patterns).
It is noted that the crystallinity of the Ba2+-templated
materials is extremely high compared with the K+-templated
analogs, in view of the harmonic signal peaks, indicative of
long-level structured material.12 A comparison between the
K+- and the Ba2+-templated hybrid materials thus shows that
the second is lamellar in nature with a high crystallinity in the
stacking direction, and that the first is more three-dimensional
in nature with a somewhat lower crystallinity.
3.1.2 Spectroscopic and X-ray diffraction characterization
of G-quadruplex hybrid materials. FTIR and NMR spectro-
scopic analyses demonstrate the formation of a self-organized
organic–inorganic network. FTIR spectra of hybrid materials
show the appearance of the broad vibrations of nSi–O–Si
=
900–1200 cmꢂ1 instead of the vibrations nSi–OEt = 950, 1070
and 1100 cmꢂ1 initially observed for the molecular precursor,
2. Evidence for H-bonding and K+ complexation was
obtained from the vibration shifts of the –CQO bonds,
detected after hydrolysis–condensation reactions: nCQO
=
We believe that in the first sol–gel step, the G-quadruplexes
are fixed via the hydrophobic interface in the siloxane network
of 2. These fixed (‘‘frozen’’) objects highly self-correlate, as
proven by powder XRD experiments, to generate anisotropic
mesophases interconnected via condensed lipophilic siloxane
bridges. With increasing amounts of 2, the observed
peaks are less well-defined and the peak width increases,
indicating smaller domains in which coherent scattering
occurs (Fig. 2c,d). After the thermal treatment and the
complete condensation of the siloxane network, the
crystallinity greatly decreases. The peak at 2y = 26.71
(d = 3.3 A) and representative for the p–p stacking distance
between two planar G-quartets, is amazingly amplified (ESI,
Fig. 6Sw).
1737–1699 cmꢂ1. This demonstrates that the self-organization
is preserved in the hybrid materials throughout the sol–gel
polymerization process (ESI, Fig. 3S and 4Sw). 29Si MAS
NMR spectroscopic experiments are in agreement for both
+
AK+–DK and ABa2+–DBa2+, and the reference E hybrid
materials, with a partially condensed hybrid material (61.3%),
with a lower percentage of cross-linked T3 C–Si(OSi)3 (8%)
units, and mostly composed of T1 C–Si(OSi)(OH)2 (24%) and
T2 C–Si(OSi)2(OH) (68%) units, showing a predominant 2D
arrangement.
This indicates that the self assembly processes occurs
differently in the presence or in the absence of K+ templating
ions, and seems to not influence the hydrolysis–condensation
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2338 | New J. Chem., 2009, 33, 2335–2343

View Article Online
3.2 Constitutional ureidobenzocrown ether hybrid materials
3.2.1 Design of hybrid molecular components 3 and 4. Our
efforts also involved the synthesis of constitutional hybrids
based on octadecylureido-15-crown-5 derivative 3, and octa-
decylureido-propyltrimethoxysilane 4 (Scheme 1b). For structural
(H-bonding and packing presented below) reasons, both
hybrid components have been designed bearing an urea
moiety. The octadecylisocyanate was treated with the amino-
benzo-15-crown-5
and 3-(triethoxysilyl)propan-1-amine,
(CH3CN, D, 5h) to afford, after crystallization, 3 and 4,
1
respectively, as white powders (90%). The H and 13C NMR,
and ESI-MS spectra are in agreement with the proposed
formula.
3.2.2 Spectroscopic and powder XRD characterization of
ureidocrown ether hybrid materials, Ai and Bi, i = 0–80.
The FTIR spectra of hybrid materials Ai and Bi showed
the appearance of broad vibrations of the nSi–O–Si
=
1110–1125 cmꢂ1 instead of the vibrations nSi–OEt = 950,
1070, 1100 cmꢂ1, initially observed for the molecular precur-
sor 4. Evidence for strong urea H-bonding was obtained from
the FTIR spectra of solids Ai and Bi; the H-bond vibration
shifts of the urea moiety were detected: nN–H = 3310
(Dn = 30), nCQO = 1630 (Dn = 15) and nC–N = 1570
(Dn = 25) cmꢂ1.6 Moreover, we observed a widening of the
nN–H = 3300 cmꢂ1 vibration band, resulting from super-
position of several signals. 29Si MAS-NMR spectroscopic
experiments were in agreement with the almost complete 3D
condensed Ai materials (B95 % T3 C–Si(OSi)3) and with the
2D condensed Bi materials (B70–90%). The Bi materials only
presented a lower percentage of T3 C–Si(OSi)3 (8%) and were
mostly composed of T2 C–Si(OSi)2 (OH) (20–80%) and T1
C–Si(OSi)(OH)2 (15–80%) units. Moreover, by increasing
the amounts of templating self-organized macrocycle 3, we
observed an increase in the amounts of T2 units in the Bi
materials (ESI, Table 3Sw). We presume that the hydrolysis–
condensation reactions of the siloxane network correlate with
a linear self-organization of 1 during the sol–gel process. We
have previously shown9 that in solution, the ureidocrown
Fig. 2 Powder XRD patterns of hybrids A–E templated in the
presence of KOTf: (a) low and (c) high angles and of BaOTf2: (b)
low and (d) high angles.
Fig. 3 Low angle powder XRD patterns of 4-octadecylurea-benzo-
15-crown-5, 3 (high angle powder XRD in detail).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2335–2343 | 2339

View Article Online
ethers form at least two types of hydrogen-bonded aggregates
that can be modelled as a dimer followed by discrete higher
oligomers. In the solid state, the macrocyclic compound 3
adopts a preferential anti-parallel columnar packing in the solid
state (Fig. 3).9c
octadecyl moieties are linear, interpenetrated layers of
antiparallel H-bonded oligomers generate an architecture in
For a similar hexylureabenzocrown ether compound, it was
possible to obtain full atomic structure determination and to
generate the corresponding diffractogram using the data of
single crystal measurements. By comparing the diffractogram
generated from the single crystal structure solution with the
experimental diffraction results of the bulk powder of 3 it can
be seen that the preferential antiparallel dimeric packing
co-exists with a second residual polymorph. It can be postulated
that in the solid state, the antiparallel packing corresponding to
the first intense peak is preferential to the residual parallel
packing corresponding to the second small peak. Fig. 3 shows
the powder XRD patterns of compound 3. The diffraction
pattern of 3 could be indexed using a monoclinic unit cell with
a = 37.65, b = 9.10, c = 9.25 A, b = 97.711, V = 3140 A3,
corresponding to 43–44 atoms in the asymmetric unit, and
thus accommodating the 40 non-H atoms of 3 plus 3–4 atoms
from either a solvent acetonitrile or disordered octadecyl unit
(possibly situated between two unit cells). The first intense
diffraction peak is related to the length of the molecule
obtained by molecular modelling of the structures of these
complexes on the basis of the geometry in the crystal structure,
using a linear geometry for the alkyl substituents. The
same premise was observed in the powder XRD patterns of
precursor 3.
The relative arrangement of 4 in the crystal of the anti-parallel
H-bonded aggregates is similar to the previously described five
single-crystal structures of different 3-(ureiodoarene)propyl-
triethoxysilane superstructures;4c in the crystal lattice these
compounds pack in one direction, into ureiodoarene and
triethoxysilane layers.4c These layers, which are alternatively
stratified, exhibit two types of interfaces in between: one
contact surface due to the aromatic interactions and the other
surface resulting from hydrophobic interactions of alkyl
groups that are in van der Waals contact. Assuming that the
Fig. 5 Powder XRD patterns of compounds 3 and 4, and hybrid
materials Ai and Bi, i = 0, 50, 60, 80% w/w prepared by using
benzylamine (basic) catalyst: (a) low and (b) high angles and HCl
(acidic) (c) low and (d) high angles.
Fig. 4 Low angle powder XRD patterns of octadecylureidopropyl-
trimethoxysilane, 4 (high angle powder XRD in detail).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2340 | New J. Chem., 2009, 33, 2335–2343

View Article Online
Scheme 2 Schematic representation of mixture of patterns of macrocyclic compound 3 and hydrophobic siloxane 4 in self-organized hybrid
materials Ai and Bi.
which the distance between urea ribbons of (4)n anti-parallel
H-bonded oligomers is compatible with the distance corres-
ponding to the first intense diffraction peak at 2y = 2.371
(d = 41.4 A) (Fig. 4).
interpenetrated H-bonded networks of 3 and 4 (Fig. 5a,c, B50).
Increasing the amount of 3 introduces a dominant behaviour
Further valuable insights into the self-organization of Ai
and Bi materials were obtained from the powder XRD (Fig. 5).
After the sol–gel process, the powder XRD data shows that
the long-range order of Ai and Bi hybrid materials is
less pronounced than in the precursors 3 and 4: there are less
well-defined peaks present than for the precursors and the
average peak width increases, indicating smaller domains in
which coherent scattering occurs.
However, it was noted that the crystallinity of Ai and Bi
materials is higher compared with the classical silsesquioxane-type
hybrid materials,4,7,9 in view of the more-defined signal
powder XRD peaks. The Bi (base-catalysed) materials
(Fig. 5a,c) are more structured than the Ai (acid-catalysed)
analogs (Fig. 5b,d), in correlation with 29Si MAS-NMR
experiments. After increasing the amounts of 3, from 10 to
50% w/w, the crystallinity of both Ai and Bi hybrids was not
extremely high in view of the largeness of the diffraction peaks
(ESI, Fig. 8Sw). However, for the concentrations c1 4 50 wt%,
the observed powder XRD peaks are indicative of the
self-organized columnar patterns of macrocyclic compound 3
(Fig. 5 Ai and Bi i = 60, 80). We believe that for the sol–gel
step, a critical concentration of 3 (C 4 50 wt%) is needed for
increasing the density and the orientation of the columnar-
packed architectures, which percolate in ‘‘highly concentrated’’
hybrid materials. These domains of dominant self-organization
behaviour of 3 are fixed via the hydrophobic interfaces in the
siloxane network of 4 (Scheme 2). The low-angle peaks in the
range 1.0–6.01 are indicative of the long-range organization
of the A50ꢂ80 and B50ꢂ80 hybrids. For A50ꢂ80 cross-linked
hybrids, one diffraction peak is located at 2y = 1.81
(d = 46.8 A), indicating that the materials express a cross-
over (mixed) behaviour of interfering networks of 3 and 4.
The robustness/stability of superstructures of 3 and 4 is
strongly expressed in an independent way (linear combination)
in the hybrid structure of B50, presenting two distinct diffraction
peaks at 2y = 1.851 (d = 46.82 A) and at 2y = 2.341
(d = 41.8 A), corresponding to the signature of a mixture of
Fig. 6 (a,b) SEM images and (c) powder XRD patterns of (a) classical
(GSi)4K+, and (b) of constitutional ABa2+G-quadruplex hybrids.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2335–2343 | 2341

View Article Online
in the structure of B80, such as the interpenetrated macrocyclic
network imposing its own output over the other one (4), as
only one peak at 2y = 1.851 (d = 46.82A) is present in powder
XRD of this material (Fig. 5a,c, B80).
d = 29.4 A, are compatible with the diameter of the hexagonal
packed G-quartet superstructure.4a The ABa2+–DBa
2+
and
AK+–DK constitutional hybrids showed high organisation,
+
with a = 64.4 A for AK+–DK+ and a = 47.2 A for ABa2+–DBa
2+
compatible with octadecyl-G-quartet superstructures.
It seems that these materials are more structured than the
classical (GSi)4K+ analogs in view of the sharpness and the
number of the diffraction peaks observed in XPRS patterns
(Fig. 6c). Moreover, scanning electron microscopy (SEM)
images of materials ABa2+–DBa2+ and AK+–DK+ show dense,
compact solids with a micrometric plate-like morphology.
The same premises hold true for the ureido-crown ether
hybrid systems Ai and Bi (Fig. 7). These constitutional hybrids
are reaching a better level of organisation giving access to a
large number of crystalline order in view of the higher number
of diffraction peaks observed in powder XRD patterns Ai and
Bi (Fig. 7c). Moreover, while the classical hybrids prepared
3.3 Comparative analysis of classical silsesquioxane and
constitutional dynamic hybrid materials
The present results showed that highly self-organized materials
can be synthesized by using a dynamic communicating
interface between organic and inorganic hybrid mesophases.
The last step of this work consists of comparing the novel
‘‘constitutional hybrid materials’’ with ‘‘classical covalent
hybrids’’ previously reported by our group.4a,9d The remarkable
G-quadruplex twisted hybrid nanorods can be obtained
as previously reported4a at the nanometric level by using
silsesquioxane GSi under the experimental conditions reported
in section 2.3 (Fig. 6a). These nanorods correlated with
an ordered hexagonal structure with the Bragg diffraction
peak at 2y = 3.01 corresponding to a characteristic distance
9b
from silsesquioxane ECSi under the same experimental
conditions as described in section 2.5 are compact dense solids
with a granular morphology (Fig. 7a), the constitutional
hybrids Ai and Bi consist of crystalline rods of micrometric
dimension. The organisation at micrometric length of
both ‘‘hybrid constitutional materials’’ studied in this paper,
compared with the classical ones showing nanometric
organization, is probably dependent of the superior emergence
of self-organizing collective domains adapting along the
hydrophobic reversible interfaces. This may increase the size
of self-organized domains, increasing the dimension of the
resulting G-quadruplex and ureidocrown ether constitutional
hybrids.
4. Conclusions
In conclusion, the results presented here reveal a new strategy
to transcribe and to fix the supramolecular ion channel-type
architectures in self-organized constitutional hybrids. In particular,
the use of a communicating hydrophobic interface between
organic/supramolecular and inorganic/siloxane networks is
noteworthy and represent a useful strategy for improving their
compatibility. Interpenetrated hybrid components led after
the sol–gel process to materials in which the features of
the supramolecular and inorganic networks are expressed via
cross-over and linear processing schemes. Dynamic self-
assembly of supramolecular systems prepared under thermo-
dynamic control may in principle be connected to a kinetically
controlled sol–gel process (under basic catalysis), in order to
extract and select the interpenetrated self-organized networks
under a specific set of experimental conditions. Such ‘‘dynamic
marriage’’ between supramolecular self-assembly and inorganic
sol–gel polymerization processes that synergistically communicate,
leads to higher self-organized hybrid materials with increased
micrometric scales. More generally, applying such consideration
to hybrid materials leads to the definition of constitutional
hybrid materials, in which organic (supramolecular)/inorganic
domains are reversibly connected. This might provide new
insights into the basic features that control the design
of functional constitutional architectures.1 Considering the
simplicity of this strategy, possible applications on the synthesis
Fig. 7 (a,b) SEM images and (c) powder XRD patterns of (a)
classical ECSi, and (b) of constitutional B50 ureidocrown ether hybrids.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
2342 | New J. Chem., 2009, 33, 2335–2343

View Article Online
of more complex hybrid architectures might be very effective,
coming close to novel expressions of complex matter.
(c) M. Barboiu, E. Petit, A. van der Lee and G. Vaughan, Inorg.
Chem., 2006, 45, 484–486.
7 (a) C. Sanchez, H. Arribart and M. M. Girau-Guille, Nat. Mater.,
2005, 4, 277–288; (b) J. J. E. Moreau, L. Vellutini, M. Wong Chi
´
Man, C. Bied, J. L. Bantignies, P. Dieudonne and J. L. Sauvajol,
Acknowledgements
J. Am. Chem. Soc., 2001, 123, 7957–7958; (c) R. J. P. Corriu, Eur.
J. Inorg. Chem., 2001, 1109–1121; (d) K. J. C. van Bommel,
A. Frigerri and S. Shinkai, Angew. Chem., 2003, 115, 1010–1030
(Angew. Chem., Int. Ed., 2003, 42, 980–999); (e) M. Barboiu,
C. Guizard, C. Luca, B. Albu, N. Hovnanian and J. Palmeri,
J. Membr. Sci., 1999, 161, 193–206; (f) M. Barboiu, C. Guizard,
C. Luca, N. Hovnanian, J. Palmeri and L. Cot, J. Membr. Sci.,
2000, 174, 277–286; (g) M. Barboiu, C. Guizard, N. Hovnanian,
J. Palmeri, C. Reibel, C. Luca and L. Cot, J. Membr. Sci., 2000,
172, 91–103; (h) O. Villamo, C. Barboiu, M. Barboiu, W. Yau-
Chun-Wan and N. Hovnanian, J. Membr. Sci., 2002, 204, 97–110;
(i) C. Guizard, A. Bac, M. Barboiu and N. Hovnanian, Sep. Purif.
Technol., 2001, 25, 167–180.
This work, conducted under the European Heads of Research
Councils and European Science Foundation EURYI
(European Young Investigator) Awards scheme in 2004,
was supported by funds from the Participating Organizations
of EURYI and the EC Sixth Framework Program. See
http://www.esf.org/euryi.
Notes and references
1 (a) J.-M. Lehn, Chem. Soc. Rev., 2007, 36, 151–160;
(b) J. A. Thomas, Chem. Soc. Rev., 2007, 36, 856–868.
2 (a) P. Vongvilai, M. Angelin, R. Larsson and O. Ramstrom,
¨
Angew. Chem., 2007, 119, 966–968 (Angew. Chem., Int. Ed.,
2007, 46, 948–950); (b) R. Larsson and O. Ramstrom, Eur. J.
¨
Org. Chem., 2006, 285–291; (c) R. Larsson, Z. Pei and
O. Ramstrom, Angew. Chem., 2004, 116, 3802–3804 (Angew.
¨
Chem., Int. Ed., 2004, 43, 3716–3718); (d) M. Angelin,
8 (a) J. T. Davis and G. P. Spada, Chem. Soc. Rev., 2007, 36,
296–313; (b) J. T. Davis, Angew. Chem., 2004, 116, 684–716
(Angew. Chem., Int. Ed., 2004, 43, 668–698); (c) T. Giorgi,
S. Lena, P. Mariani, M. A. Cremonini, S. Masiero, S. Pieracini,
J. P. Rabe, P. Samori, G. P. Spada and G. Gotarelli, J. Am. Chem.
Soc., 2003, 125, 14741–17479; (d) S. Pieraccini, S. Masiero,
O. Pandoli, P. Samori, G. P. Spada and G. Gotarelli, Org. Lett.,
2006, 8, 3125–3128; (e) A. Ghoussoub and J.-M. Lehn, Chem.
P. Vongvilai, A. Fischer and O. Ramstrom, Chem. Commun.,
Commun., 2005, 5763–5765; (f) V. Setnicka, M. Urbanova,
´
¨
2008, 768–770; (e) M. Angelin, P. Vongvilai, A. Fischer and
K. Volka, S. Nampally and J.-M. Lehn, Chem.–Eur. J., 2006, 12,
8735–8743; (g) S. Nampally and J.-M. Lehn, Chem.–Asian J., 2008,
3, 134–139; (h) W. C. Dreive and S. Neidle, Chem. Commun., 2008,
5295–5297; (i) M. S. Kaucher, W. A. Harrell and J. T. Davis,
J. Am. Chem. Soc., 2006, 128, 38–39; (j) L. Ma, M. Iezzi,
M. S. Kaucher, Y. F. Lam and J. T. Davis, J. Am. Chem. Soc.,
2006, 128, 15269–15277; (k) L. Ma, M. Melegari, M. Collombini
and J. T. Davis, J. Am. Chem. Soc., 2008, 130, 2938–2939;
(l) M. S. Kaucher, Y. F. Lam, S. Pieraccini, G. Gotarelli and
J. T. Davis, Chem.–Eur. J., 2005, 11, 164–163; (m) V. Shi,
K. M. Mullaugh, J. C. Fettinger, Y. Jiang, S. A. Hofstadler and
J. T. Davis, J. Am. Chem. Soc., 2003, 125, 10830–10841.
9 (a) M. Barboiu, G. Vaughan and A. van der Lee, Org. Lett., 2003,
5, 3073–3076; (b) M. Barboiu, J. Inclusion Phenom. Macrocyclic
Chem., 2004, 49, 133–137; (c) A. Cazacu, C. Tong, A. van der Lee,
T. M. Fyles and M. Barboiu, J. Am. Chem. Soc., 2006, 128,
9541–9548; (d) M. Barboiu, S. Cerneaux, A. van der Lee and
G. Vaughan, J. Am. Chem. Soc., 2004, 126, 3545–3550.
O. Ramstrom, J. Org. Chem., 2008, 73, 3593–3595.
¨
3 (a) F. Menger, Angew. Chem., 1991, 103, 1104–1118 (Angew.
Chem., Int. Ed. Engl., 1991, 30, 1086–1089); (b) S. Xu and
N. Giuseppone, J. Am. Chem. Soc., 2008, 130, 1826–1827;
(c) R. Nguyen, L. Allouche, E. Buhler and N. Giuseppone, Angew.
Chem., 2009, 121, 1113–1116 (Angew. Chem., Int. Ed., 2009, 48,
1093–1096) and references therein.
4 (a) C. Arnal-He
Angew. Chem., 2007, 119, 4346–4350 (Angew. Chem., Int. Ed.,
2007, 46, 4268–4272); (b) C. Arnal-Herault, A. Pasc-Banu,
´
rault, M. Michau, A. Pasc-Banu and M. Barboiu,
´
M. Michau, D. Cot, E. Petit and M. Barboiu, Angew. Chem.,
2007, 119, 8561–8565 (Angew. Chem., Int. Ed., 2007, 46,
8409–8413); (c) M. Michau, M. Barboiu, R. Caraballo, C. Arnal-
He
´
2008, 14, 1776–1783; (d) C. Arnal-He
rault, P. Periat, A. van der Lee and A. Pasc, Chem.–Eur. J.,
rault, M. Barboiu, A. Pasc,
´
M. Michau, P. Perriat and A. van der Lee, Chem.–Eur. J., 2007, 13,
6792–6800.
5 (a) P. N. W. Baxter, J.-M. Lehn and K. Rissanen, Chem. Commun.,
1997, 1323–1324; (b) M. Hutin, C. J. Cramer, L. Gagliardi, L. A.
R. Shahi, G. Bernardinelli, R. Cerny and J. R. Nitschke, J. Am.
Chem. Soc., 2007, 129, 8774–8780; (c) P. N. W. Baxter, J.-M. Lehn,
G. Baum and D. Fenske, Chem.–Eur. J., 2000, 6, 4510–4517;
(d) C. D. Pentecost, K. S. Chichak, A. J. Peters, G. W. V. Cave,
S. J. Cantrill and J. F. Stoddart, Angew. Chem., 2006, 119, 222–226
(Angew. Chem. Int. Ed., 2006, 45, 218–222); (e) C.-F. Chow,
S. Fujii and J.-M. Lehn, Chem. Commun., 2007, 4363–4365.
6 (a) F. Dumitru, E. Petit, A. van der Lee and M. Barboiu, Eur. J.
Inorg. Chem., 2005, 4255–4262; (b) Y. M. Legrand, A. van der Lee
and M. Barboiu, Inorg. Chem., 2007, 46, 9083–9089;
10 (a) A. Khaled, O. Piotrowska, K. Dominik and C. Auge,
´
Carbohydr. Res., 2008, 343, 167–178; (b) M. Iwasaki, S. Inue and
Y. Inue, Eur. J. Biochem., 1987, 168, 185–192; (c) D. M. Gray,
J.-D. Wen, C. W. Gray, R. Repges, C. Repges, G. Raabe and
J. Fleinschhauer, Chirality, 2008, 20, 431–440.
11 For recent examples where CD was used to detect a G-quadruplex
supramolecular assembly in a membrane, see: N. Sakai and
S. Matile, Chirality, 2003, 15, 766–771 and ref. 8i,k.
12 The Ba2+ G-quadruplex is thermodynamically and kinetically
more stable than the K+ G-quadruplex. See: X. Shi,
J. C. Fetiger and J. T. Davis, J. Am. Chem. Soc., 2001, 123,
6738–6739.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 2335–2343 | 2343