Optical properties of hybrid dendritic-mesoporous titania nanocomposite films

Article abstract of DOI:10.1002/chem.200800606

New hybrid optical sensors have been prepared by grafting specifically designed fluorescent, functionalised, phosphorus-containing dendrimers onto a nanocrystalline mesoporous titania thin film formed by evaporation-induced self-assembly. The structural c

Full text of DOI:10.1002/chem.200800606

DOI: 10.1002/chem.200800606
Optical Properties of Hybrid Dendritic–Mesoporous Titania Nanocomposite
Films
Eugenia Martínez-Ferrero,[a] GrØgory Franc,[b] Serge Maz›res,[c] CØdric-Olivier Turrin,[b]
CØdric Boissi›re,[a] Anne-Marie Caminade,*[b] Jean-Pierre Majoral,*[b] and
ClØment Sanchez*[a]
Abstract: New hybrid optical sensors
have been prepared by grafting specifi-
cally designed fluorescent, functional-
ised, phosphorus-containing dendrim-
ers onto a nanocrystalline mesoporous
titania thin film formed by evapora-
tion-induced self-assembly. The struc-
tural characterisation and optical be-
haviour of these new fluorescent
probes have been studied both in solu-
tion and after being grafted onto an in-
organic network, which resulted in the
discovery of improved probing selectiv-
ity in the solid state. This new hybrid
sensor exhibits high sensitivity to phe-
nolic OH moieties (especially those
from resorcinol and 2-nitroresorcinol),
which induce the quenching of fluores-
cence more efficiently in the solid state
than in solution. This effect is a result
of the increased spatial proximity of
the fluorescent molecules, which is in-
duced by pore confinement that makes
the formation of hydrogen bonds be-
tween the hydroxyl moieties of the
quenchers and the carbonyl groups of
the dendrimer easier.
Keywords: dendrimers · fluorescent
probes
·
hybrid materials
·
mesoporous materials · titania
Introduction
ing field in the area of polymer science.[1] With tuneable
chemical composition and size, these perfectly controlled
structures are being developed more and more in the fields
of materials science, catalysis and biology. In materials sci-
ence, dendrimers participate in the elaboration of complex
hybrid organic–inorganic functional nanoarchitectures. Due
to the presence of specific functional groups at their surface
or in their internal cavities, they can act as controlled pock-
ets when embedded in a mineral matrix. In such a configura-
tion, they can act as sensitive probes, selective adsorbents,
drug-release controllers or nanoreactors.
Dendrimers are highly branched macromolecules with inter-
esting properties, the study of which constitutes an appeal-
[a] Dr. E. Martínez-Ferrero, Dr. C. Boissi›re, Dr. C. Sanchez
Laboratoire de Chimie de la Mati›re CondensØe
UniversitØ Pierre et Marie Curie
4 Place Jussieu
75252 Paris Cedex05 (France)
Fax: ( +33)144-274-769
E-mail: clems@ccr.jussieu.fr
As well as purely organic dendrimers, the phosphorus-
containing dendrimers also excel because of their simple
synthesis, their diverse chemistry and their easy attachment
to coupling agents that are able to form strong bonds with
inorganic materials. These striking features permit their use
in hybrid materials either as building blocks or by creating
supramolecular assemblies on the surface of materials.[2] The
first hybrid materials, prepared in 2000, were made by the
co-condensation of phosphorus dendrimers with a silica-
based inorganic matrix, which gave rise to porous materials
that tend to be used as absorbents or insulating materials.[3]
Second-generation hybrid materials were obtained shortly
afterwards by using non-silicate matrices. Two synthesis ap-
proaches have been tested: First, the co-assembly of pre-
[b] Dr. G. Franc, Dr. C.-O. Turrin, Dr. A.-M. Caminade, Dr. J.-P. Majoral
Laboratoire de Chimie de Coordination du CNRS
205 Route de Narbonne
31077 Toulouse cedex4 (France)
Fax: ( +33)561-553-003
E-mail: caminade@lcc-toulouse.fr
majoral@lcc-toulouse.fr
[c] Dr. S. Maz›res
Institut de Pharmacologie et de Biologie Structurale
CNRS-UPS, UMR 5089
205 Route de Narbonne
31077 Toulouse cedex4 (France)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800606 or from the author. It
contains the X-photoelectron spectroscopy studies of dendrimer 7
and TiO2–7 (Figure S1), and the FTIR spectrum of TiO2–7 quenched
by 4-nitrophenol (Figure S2).
formed nano building blocks (NBB), that is, [Ti16O16(OEt)32]
A
7658
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

FULL PAPER
clusters and dendrimers, gave rise to bicontinuous meso-
structured hybrid gels formed from periodically stacked in-
organic and organic NNBs.[4] Second, the mixof dendrimers
with titanium and cerium alkoxides allowed the control of
the reactivity of the inorganic precursors. After the appro-
priate treatments, porous materials made of dendrimers
locked in metal-oxide matrices were obtained.[5]
show a well-defined fluorescence pattern and thus behave as
a fluorescent organic probe for different hazardous nitro-
phenols and phenols. To compare grafting and detection ef-
ficiency, we also used (ZrO2)x(SiO2)1Àx MTS films with a 3D
R
mesoporous network as an alternative substrate. The study
of the optical properties of these systems was performed by
using quenching tests to determine the behaviour of the
dendrimers in the solid state and compare the results with
those obtained in solution.
For surface modification purposes, the main advantage of
using dendrimers instead of simple monomers is the number
of anchoring points provided by the dendrimer, which en-
hances the stability of the interfaces and give them a con-
trolled roughness. In particular, these properties were used
for the elaboration of stable and sensitive DNA chips.[6]
For optical applications, the dispersion of fluorescent den-
drimers in a hole-conducting polymer has permitted the
preparation of organic light-emitting diodes.[7] Fluorescent
dendrimers are also used for other applications, such as ana-
lytical probes, organic nanodots or biological labelling.[8] In
these functions, they can act as sensors by monitoring the
fluorescence response of the active entities. For example,
hazardous nitrophenols are known to be good quenchers of
some fluorescent substances due to their electron acceptor
capacity.[9] Hence, specifically functionalised dendrimers
have vast potential as components for the construction of
detectors.
However, practical detector applications require that the
probe should be immobilised into a robust matrixwith a
high surface area so that high sensitivity is reached and the
fast diffusion of the analytes is allowed. For this application,
micellar-templated materials (MTS) prepared by the sol-gel
route constitute a new type of extremely interesting matrix
for dendrimer immobilisation, due to their highly organised
and easily accessible 2D or 3D mesoporous structures and
huge surface area.[10] The recently developed evaporation-in-
duced self-assembly method permits their preparation in dif-
ferent forms, such as nanometric powders, monoliths, fibres
and especially thin films of optical quality and controlled
thickness.[11] These outstanding properties make these com-
pounds the ideal candidates for preparing new multifunc-
tional hybrid organic–inorganic optical sensors. However, as
demonstrated recently, the usual silica MTS have poor sta-
bility versus humidity.[12] A possible alternative is to use
MTS films made of amorphous (or more ideally nano-crys-
talline) transition-metal oxides, such as TiO2, ZrO2 or Al2O3,
or mixed metal oxides that are mechanically and chemically
much more stable.[13]
Results and Discussion
Dendrimer synthesis and characterisation: As stated in the
introduction, the dendrimer-like structures described herein
were specially designed to have two non-interfering func-
tionalities: First, to act as fluorescent probe, and second, to
possess distinct phosphorylated functional groups dedicated
to the strong anchoring of the dendrimers at the TiO2 sur-
face. In this regard, some of our previous studies reported
that the versatile functionalisation of cyclotriphosphazene
core dendrimers is a valuable route to functional dendritic
precursors, either for the design of fluorescent dendrimer
cores[16] or for the design of unprecedented divergent
points.[17] The synthesis of the dendrimer is shown in
Scheme 1. The first step in building the desired core consists
of the reaction of one equivalent of the sodium salt of 4-hy-
droxybenzaldehyde with N3P3Cl6 to give rise to a monosub-
stituted focal point (1).[18] The 31P NMR spectrum of 1 shows
one triplet at d=11.73 ppm (2J
(P,P)=62.4 Hz) for the phos-
N
phorus with the aldehyde derivative and one doublet at d=
22.53 ppm for the other two phosphorus atoms, which con-
firmed the formation of the desired structure. In parallel,
the maleimide-based phenolic chromophore 2 was synthes-
ised under the usual conditions (AcOH and high tempera-
ture)[19] from tyramine and diphenyl maleic anhydride (see
Scheme 1). Compound 2 was obtained in very good yield
(95%), which lead us to perform the same procedure with
phenylethylamine to prepare the same compound without
the OH group (3; 98% yield) to measure differences in the
fluorescence properties.
The substitution of the five remaining chlorine atoms of 1
by chromophore 2 under the usual conditions (THF and
Cs2CO3, 2 equiv per chlorine) gave 4. The reaction is rela-
tively slow and needs 2 days to complete, as shown by
31P NMR spectroscopic monitoring (multiplet at d=
8.47 ppm). After work-up, the purity of the dendrimer was
also confirmed by 1H, 13C NMR and FTIR spectroscopies
Herein, we report the synthesis of new hybrid optical sen-
sors shaped as thin films. For the inorganic matrix, we se-
lected nano-crystalline titania mesostructured films that
present excellent chemical stability[14] and an ordered grid-
like porous network that provides good accessibility and
permeability for external species.[15] Within these films, we
grafted new fluorescent phosphorus-containing dendrimers,
which have been specially designed to fulfil two key proper-
ties: First, to incorporate a reactive phosphonate group that
can graft onto titanium oxide mesoporous films, but that
would not impair the fluorescence properties, and second, to
1
and mass spectrometry. At this point, the H and 13C NMR
spectra started to show multiple signals due to the unsym-
metrical arrangement of the branches around the core. A
good example to describe this phenomenon concerns the
1H NMR spectrum of the tyramine moiety, in which each
a’
À
methylene group gave two multiplets at d=2.99 (C H)
b’
À
and 3.87 ppm (C H) in its instead of two triplets. Two dif-
ferent signals for each carbon also appear in the 13C NMR
spectrum at d=33.96 and 33.99 ppm for (Ca’) and at d=
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7659

A.-M. Caminade, J.-P. Majoral, C. Sanchez et al.
Scheme 1. The synthesis of chromophores 2 and 3 followed by the synthesis of fluorescent anchorlike dendrimers 7 and 9. a) 2 (5 equiv), Cs2CO3, THF.
b) H2NNMeP(S)Cl2, CHCl3.
39.51 and 39.54 ppm for (Cb’), respectively (see the Experi-
mental Section for the atom labelling system).
nate moieties and d=63.10 ppm for the P(S) group. Asym-
1
metry remained visible in the H and 13C NMR spectra, with
The dendritic core 4 was then functionalised to allow fur-
ther grafting onto the mesoporous films. Phosphonate was
chosen as the grafting entity because it is known to form
stable compounds with metal oxides.[20] Therefore, and to
have accessible grafting functional groups, we decided to
grow the specific branch with the aldehyde up to the first
generation. Thus, the next step consisted of a condensation
between the main structure and one equivalent of N-methyl-
dichlorophosphorhydrazide. This condensation reaction was
multiple signals for the methylene groups of the tyramine
moieties.
Hydroxyphosphonate is also a good candidate as a graft-
ing functional group on metal oxide, and was thus used on
the (ZrO2)x(SiO2)1Àx films. Consequently we slightly modi-
G
fied our strategy to reach this other target; in fact, a two-
step process starting from 5 readily gave access to this new
compound. First, aldehyde extremities (see compound 8)
were attached in a simple manner by nucleophilic substitu-
tion of the chlorine atom with the sodium salt of 4-hydroxy-
benzaldehyde. Second, a classic Abramov-type reaction is
undertaken with 8 in presence of dimethylphosphite in a
concentrated THF solution, which leads to the desired com-
pound 9 with two a-hydroxyphosphonates on the surface.[21]
1
monitored by 31P and H NMR spectroscopies that indicated
the total disappearance of the signal from the aldehyde
group, but FTIR spectroscopic monitoring could not be per-
formed because the carbonyl of the maleimide at 1703 cmÀ1
strongly hides the signal of the aldehyde group. This step
gave the dendritic core 5, with the P(S)Cl2 end group ap-
pearing in the 31P NMR spectrum at d=63.02 ppm, whereas
the phosphazene core remained unaffected. The molecule
has an unsymmetrical architecture, as shown, for example,
by two signals at d=170.51 and 170.53 ppm in the 13C NMR
spectrum for the carbonyl groups of the maleimides. Follow-
ing this, nucleophilic substitutions of the chlorine atoms by
6[16] gave 7 in very good yield (83%). After purification, the
chemical shifts of the 31P NMR spectrum appeared at d=
8.66 ppm for the core signal, d=26.89 ppm for the phospho-
1
Finally, H NMR spectroscopy and TLC were used to ascer-
tain the purity and the absence of residual impurities. More-
over, compounds were subjected to flash column chromatog-
raphy when possible during the synthesis to remove impuri-
ties.
The emissive properties of each resulting compound have
been characterised in solution by UV/Vis absorption spec-
troscopy and fluorescence measurements in distilled CH2Cl2,
and are summarised in Table 1. Note that the absorption in-
creases proportionally as the number of fluorescent groups
7660
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

Hybrid Nanocomposite Films
FULL PAPER
Table 1. Emissive properties of synthesised fluorescent molecules 2–4
and 7–9 in CH2Cl2.
e [molÀ1 cmÀ1
]
lex [nm]
lem [nm]
F
[a] [%]
2
3
4
7
8
9
3457
3470
17000
17650
18900
17900
320/380
300/375
300/375
315/375
307/376
302/375
506
506
506
506
506
506
8
52
64
50
61
52
[a] Coumarine 6 (F=78% in EtOH) was used as the reference com-
pound.
increases. They all possess two excitation centres, one for
the phenyl groups around l=300–320 nm and another at
l=375 nm for the maleimide moiety, which lead to a single
emitting species at l=506 nm because the chromophore has
a p-conjugated system. Another important point concerns
the quantum yields that were calculated with Coumarine 6
as the reference (F=78% in EtOH). The quantum yields
of monomers 2 and 3 were calculated to be 8 and 52%, re-
spectively. Therefore, the lack of an OH group for 3 enhan-
ces the fluorescence emission, which suggests that this spe-
cific group self-quenches the emission of 2.[22] Therefore, the
dendrimers have moderate to good quantum yields that
permit the formation of the fluorescent hybrid materials.
Structural characterisation of the functionalised films: The
mesoporous titania films were functionalised by dip-coating
in a dilute 1 mm solution of dendrimer 7 in dichloromethane
under an air fluxat 40 8C, to give TiO2–7. This technique
permits the homogeneous introduction and deposition of
the organic molecules inside the films, and the evaporation
of the solvent. Further washing in dichloromethane removed
non-bonded dendrimers. Because the phosphonates are
known to give rise to covalent bonds with titanium oxide,[23]
we can easily prepare functionalised films in this way.
Figure 1. Top: TEM image of the mesoporous films after being heated at
5508C; inset: the diffraction pattern. Bottom: Comparison of water ad-
sorption–desorption isotherms obtained from EEP measurements for the
TiO2 (c) and TiO2–7 (g) films. Inset: pore-size distributions for the
~
*
TiO2
( ) and TiO2–7 ( ) films, obtained from the adsorption region of
Transmission electron microscopy (TEM) reveals that the
films consist of an ordered structure, in which gridlike
porous structures, aligned parallel to each other, stand per-
pendicular to the substrate (see a TEM top view in Fig-
ure 1a) and the walls are composed of 10 nm anatase nano-
crystals, in accordance to previous published results (see the
diffraction pattern in Figure 1a, inset).[24,25] The structural
characterisation was completed by using environmental el-
lipsometric porosimetry (EEP), which is based on the mea-
surement of water adsorption–desorption isotherms. Fig-
ure 1b shows the variation of the normalised volume of ad-
sorbed water into the film (Vads/Vfilm), which is directly pro-
portional to the porous volume, with the relative water pres-
sure (P/P0). As the relative humidity (RH) increases to
70%, the non-functionalised titania film takes up water
slowly. Above 70% RH, a massive adsorption of water was
observed, which is characteristic of the narrow pore-size dis-
tribution of these mesostructured matrices.[24–26] Upon de-
sorption, the isotherm exhibits a hysteresis owing to the
presence of restrictions between the mesopores of the films.
Data are well fitted between 500 and 1000 nm with a
the isotherm.
Cauchy single-layer model, which indicates that the compo-
sition of the film is homogeneous. The parameters of the
porous network were calculated from the isotherm (Table 2)
by considering that mesopores are anisotropic and connect-
ed with nearest neighbours by four restrictions.[26] Data col-
lected from functionalised films were perfectly fitted with
the same model (a single-layer model), which suggested that
the organic molecules were homogeneously dispersed within
the film, and that no extra layer of dendrimers was detecta-
Table 2. Comparison of the structural parameters of the TiO2 and TiO2–7
films.
Thickness [nm] Small
Large
Porous Water contact
pore axis pore axis fraction angle
[nm]
[nm]
[%]
[8]
TiO2
197
4.5
3.0
7.6
5.0
29
25
5Æ2
TiO2–7 197
60Æ2
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7661

A.-M. Caminade, J.-P. Majoral, C. Sanchez et al.
ble on the top of the TiO2 layer. The isotherms display a
similar behaviour to that of the titania film. However, the
introduction of the dendrimers resulted in a porous network
with a decreased pore size (from 7.6 to 5.0 nm for ungrafted
and grafted TiO2, respectively) and a decreased porosity
(from 29 to 25%). These data are coherent with the known
flattening of dendrimers on surfaces.[1b] In addition, static
water contact-angle measurements proved that the high hy-
drophilicity of the TiO2 surface was lost upon dendrimer
grafting (Table 2). This strongly suggests that the surface hy-
droxyl groups of titania were barely accessible after grafting.
From the comparison of the parameters shown in Table 2,
we can conclude that functionalisation does not alter the
thickness or the meso structure of the film.
The presence of dendrimers in the films has also been
verified by FTIR spectroscopy, which compares the spectra
of functionalised and non-functionalised samples in different
regions (Figure 2). Between n˜ =3600 and 2600 cmÀ1 (Fig-
ure 2a), we observed a broad signal at n˜ =3420 cmÀ1 due to
the hydroxyl groups on the surface of the titania film. After
the introduction of the organic dendrimer, the broader
signal is centred at n˜ =3300 cmÀ1 and was assigned to the
contribution from the remaining hydroxyl groups of the in-
organic matrix. The weak band centred at n˜ =3050 cmÀ1 was
assigned to the aromatic alkene groups and the bands at n˜ =
2963, 2927 and 2855 cmÀ1 are due to the CH2 groups from
the organic molecules. At lower wavelengths (Figure 2b) we
observed bands at n˜ =1275, 1206, 1200, 1184, 1164, 1108,
1063 and 1030 cmÀ1 on the functionalised films, which were
assigned to the dendrimer. From this examination, we have
attempted to identify the bonding mode between the organ-
ic dendrimer probe and the titania film. Previous work using
phosphonate coupling molecules grafted to titania nanopar-
À
À
À
ticles showed that the appearance of P O bands is evidence
for extensive condensation and coordination of the phos-
phoryl oxygen, which mainly leads to bidentate phosphonate
units.[23] Consequently, we have focused our analysis on the
n˜ =1000–1350 cmÀ1 region, as shown in Figure 2c, which dis-
plays the spectra of the functionalised film and the free den-
drimer. Although they look quite similar, the band that ap-
pears at n˜ =1030 cmÀ1 in the spectrum of dendrimer 7 is
overlapped in the spectrum of the TiO2–7 film by new,
broad bands at n˜ =1002, 1018, 1045, 1055 and 1065 cmÀ1 that
À
were assigned to P O stretching bands, in agreement with
previously published work.[23]
In addition, and to shed more light on this subject, we
also performed X-photoelectron spectroscopy (XPS). The
survey of the dendrimer and the functionalised titania film
(see the Supporting Information) displays the expected
peaks for O, C, N and P (plus Ti for the film), which strong-
ly supports the presence of dendrimer in the films. A closer
analysis of the O1s region of the spectra did not afford un-
ambiguous proof of the covalent bonding of the dendrimers
Figure 2. FTIR spectra between a) 3600 and 2600 cmÀ1 and b) 1300 and
800 cmÀ1 of the TiO2 (c) and TiO2–7 (g) films. c) FTIR spectra of
free dendrimer 7 (g) and TiO2–7 (c) between 1000 and 1350 cmÀ1
.
533.6, 532.08 and 529.43 eV. In our case, the functional
groups that are able to react with the inorganic network are
À À
À À
onto the surface (Figure 3). According to literature, P O Ti
covalent-bond formation can be followed by observing the
P O CH3 and P=O. The literature binding energies of these
groups range from 531.3 to 532.5 eV.[27] The contribution
À À
À
evolution of P O Ti (531.4 eV) and P OH (533 eV)
from the additional oxygen of the C=O groups (532 eV) to
peaks.[27] Pure dendrimer 7 displays three adjacent peaks at
the same peak makes the separation of P O CH3 and C=O
À À
7662
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

Hybrid Nanocomposite Films
FULL PAPER
Figure 4. Normalised fluorescence spectra of the TiO2–7 film (c) and
dendrimer 7 (g) in solution. Left: excitation spectrum after emission at
520 nm and right: emission spectrum after excitation at 370 nm.
Figure 3. XPS spectra showing the O1s signals of the TiO2–7 film (c)
and free dendrimer 7 (g).
signals difficult. Upon functionalisation, the intense peak at
529.7 eV from TiO2 and the broadening of the two high-
of the dendrimer to the titania network. Energy transfer is
observed between titania and the dendrimer, a phenomenon
that has been previously reported in sol-gel-derived hybrid
(organic dye) titania matrices.[29] It is likely to be associated
with the partial overlap between the high-energy absorption
band of the dendrimers and the phosphonate-complexation-
driven shift in the band gap of titania.
À À
bonding-energy peaks of dendrimer 7 make the P O Ti
(531.4 eV) contribution difficult to detect.
Finally, the measurement of the water-contact angle of
films functionalised with dendrimer solutions of different
concentrations shows that higher angles are obtained for
higher solution concentrations (results not shown). Howev-
er, after washing in dichloromethane and drying, the result-
ing angle was the same for all the samples. Similarly, the re-
fractive indices of the impregnated films after washing were
found to be constant. Thus, we deduced that the final con-
centration of organic molecule that remained in the films
was the same regardless of the concentration of the initial
dendrimer solution.
From the EEP, FTIR and XPS measurements, we corro-
borated the presence of the dendrimer inside the films, but
definite proof of the covalent grafting of dendrimers onto
the TiO2 surface could not be obtained. However, there are
many examples in the literature of grafting under similar
conditions that attest the formation of covalent bonds be-
tween TiO2 surfaces and phosphonate functional
groups.[23,26,28] Therefore, we have assumed herein that the
dendrimers are covalently bonded to the inorganic frame-
work.
The influence of quenchers: The fluorescent emission of the
dendrimers is very strong even after functionalisation, which
makes these devices quite attractive for use as sensors. For
that reason, we performed quenching experiments at room
temperature to understand their optical properties. The
functionalised films were dipped into stock solutions of dif-
ferent compounds (see Scheme 2 and Table 3 for details)
and then dried in air. To verify the homogeneity and stabili-
ty of the hybrid films during the experiments, EEP, FTIR
Optical properties of the functionalised films: The films
were characterised by steady-state fluorescence (Figure 4).
The excitation spectrum of the dendrimer in solution shows
two maxima at l=315 and 375 nm that are due to the
phenyl and the maleimide moieties, respectively, whereas
the emission maximum is at l=506 nm. The corresponding
spectra of the solid-state samples show similar trends to
those in solution, although the maxima are slightly red-shift-
ed to l=373 and 515 nm, which is attributed to the bonding
Scheme 2. Molecular structure of the quenchers used herein.
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7663

A.-M. Caminade, J.-P. Majoral, C. Sanchez et al.
Table 3. Quenching molecules used in the fluorescence tests of the TiO2–
7 films, with the reaction conditions and the results for both the solid-
an alternative solvent for testing the quenching potential of
molecules that are insoluble in CHCl3.
state and solution phases, from
strong).
+ (moderate) to +++++ (very
Once the effect of OH moieties had been established, we
tested hazardous molecules that contain this group. There-
fore, we studied the response of our hybrid sensor to several
nitrophenol derivatives and observed spectacular modifica-
tions of the fluorescence (see Table 3). Hazardous 2-nitro-
phenol (2NP), 4-nitrophenol (4NP) and 2,4-nitrophenol
(24NP) have a stronger quenching effect on the fluores-
cence of the dendrimer than water and pentanol. 3-Nitro-
phenol (3NP) and 2,4,6-trinitrophenol (246TNP) presented
a stronger quenching effect, although the presence of water
in the latter reactant should be taken into account. Mole-
cules containing various phenols, such as 2-nitroresorcinol
(2NR) and 4-nitrocatechol (4NC), behaved in the same
manner; 2NR demonstrated the best quenching response.
Finally, we tested parent molecular structures containing no
nitro groups to determine their influence on quenching. As
expected, p-cresol (CS), resorcinol (RS) and catechol (CT)
showed the same quenching tendency previously observed.
RS had the strongest quenching effect determined in this
study (see Figure 6), with less than 1% of fluorescence re-
maining. Therefore, the hybrid material behaves as an excel-
lent sensor for resorcinol and 2-nitroresorcinol.
Quencher
Conditions
Quenching response Quenching response
(solid state) (solution)
AHCTREUNG
2NP
10 mm, CHCl3 ++
++++
3NP
4NP
10 mm, CHCl3 +++
10 mm, CHCl3 ++
+
+
24DNP
246TNP
2NR
4NC
RS
CT
CS
water
1-pentanol
ethanol
10 mm, EtOH ++
–
–
++
10 mm, EtOH +++
10 mm, CHCl3 +++++[a]
10 mm, EtOH ++
10 mm, EtOH +++++
10 mm, EtOH ++++
10 mm, CHCl3 ++++
–
–
–
0
–
–
–
–
–
–
+
+
0
[a] See Figure 6.
and contact angle measurements were performed at each
step.
The first test was performed in water to determine the
sensitivity of the devices towards humidity. Soaking and
drying the film had a quenching effect that moderately
modified the initial fluorescence values, as shown in
Figure 5. This is attributed to the formation of hydrogen
*
Figure 5. The effect of water ( ) and 1-pentanol (c) on the fluores-
cence emission (lex =373 nm) of TiO2–7 (a).
bonds between water and the carbonyl groups of the den-
drimers, as observed previously with chromophores 2 and 3
in solution (see above). To explore this effect, we also tested
a range of alkyl alcohols and obtained different results.
Tests with 1-pentanol produced the same quenching effect,
which confirmed our first observations. However, ethanol
did not quench the fluorescence, probably because to its
higher volatility. CHCl3 was our solvent of choice, however
this surprising result permitted us to use ethanol as well, as
Figure 6. a) Modification of the fluorescence response (lex =373 nm,
lem =520 nm) of the TiO2–7 film after soaking for 10 min in a 10 mm so-
lution of resorcinol in CHCl3 (TiO2–7+RS). b) Percentage of fluores-
cence remaining after the quenching experiments.
7664
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

Hybrid Nanocomposite Films
FULL PAPER
The results suggest that the fluorescence variation is due
to the interaction of the carbonyl group of the dendritic
core with the phenolic OH group. Indeed, the fluorescence
spectra of the hybrid films displayed no shift in the maxi-
mum of the emission wavelength. This observation suggests
that neither p stacking nor excimer formation occurs within
the dendritic-like structure. To shed light on this hypothesis
we measured the quenched films by using FTIR (see the
Supporting Information, Figure S2). The band at 1705 cmÀ1,
which was assigned to the carbonyl group of the maleimide
moiety, broadens and shifts towards lower frequency values
(between n˜ =1700 and 1690 cmÀ1, depending on the quench-
er), as expected for the formation of a hydrogen bond be-
tween the maleimide carbonyl and the phenolic OH group
of the quenching molecule.[30] In addition, UV/Vis spectros-
copy measurement of the quenched hybrid material displays
a hypsochromic shift of approximately 10 nm. The fact that
water and alkyl alcohols, such as pentanol, affect the fluo-
rescence in the same manner supports this idea.
We also tested the reversibility of the quenched system
under different conditions. The use of a protic polar organic
solvent like ethanol, with or without refluxing conditions,
did not give positive results. Washing attempts with acidic
water (one night at pH 0.3) resulted in the recovery of 40%
of the fluorescence initial values but, unfortunately, the
films deteriorated during the experiments. Less acidic solu-
tions gave poor signal recovery results. Basic conditions
gave the best results, with a fluorescence recovery of 60%
of the initial intensity while maintaining film homogeneity.
In Figure 7, the evolution of fluorescence intensity with dif-
ferent pH values for 2NR shows that a partial fluorescence
recovery starts at pH 9 and reaches a maximum at pH 10.
Given that water alone quenches between 40 and 50% of
the initial “dry film” fluorescence (Figure 5), these basic pH
recovery rates are very satisfactory.
Figure 7. a) Evolution of the fluorescence emission (lex =373 nm) of
quenched TiO2–7 devices after soaking in basic aqueous solutions with
pH values from 7.5 to 10. Results in water have been included for com-
parison purposes. b) Percentage of fluorescence recovery with pH for
2NR. Two consecutive measurements were performed for pH 9 and 10.
quantities of an identical solution of 8 that also contained
the quencher in higher concentrations were required, the
modification of the fluorescent emission by dilution being
negligible. The results, which are shown in Figure 8, clearly
reveal that there is no quenching effect in solution, except
for 2NP and for 2NR (only moderately).
Additionally, we performed measurements on functional-
ised (ZrO2)x(SiO2)1Àx (x=0.1 or 0.2) films with fluorescent
A
dendrimer 9 and observed the quasi-complete disappearance
of the fluorescence intensity after excitation at l=370 nm.
This can be explained by the presence of a higher density of
acidic hydroxyl groups bonded to the silicon and zirconium
centres of the amorphous inorganic network, which may in-
teract with the maleimide moiety by hydrogen bonding. Ad-
ditionally, the excitation of the phenyl moiety (l=290 nm)
gives a weak fluorescent signal, which demonstrates that, in
these amorphous matrices, the fluorescent centres of the
dendrimers are more affected by the hydroxyl groups from
the surface than from within the TiO2 nano-crystalline film.
Therefore, and in agreement with previous observa-
tions,[31] the hybrid material exhibits a strong improvement
in sensitivity. This can be explained by a local concentration
of dendrimer in the porous volume that is much higher than
a dispersion of the same quantity of dendrimers in solution
in dichloromethane. Indeed, when the phenol groups of the
dendrimer interact with the carbonyl groups on the surface
of the hybrid material, the probability of a phenyl group to
interact with another C=O bond or with the TiO2 surface—
and hence to become trapped through hydrogen bonding
with neighbouring dendrimers—is much higher in a confined
volume than in solution. As a consequence, we assumed that
grafted dendrimers behave like a supported phase that can
concentrate the probes because a partition coefficient exists
between the dendrimer and the solvent. Therefore, although
the interaction between the quencher carbonyl groups and
the dendrimer hydroxyl groups remains the same in solution
Comparison of optical properties and advantages of the
hybrid material: Once the hybrid devices were character-
ised, we focused our attention on dendrimers in solution.
For that, fluorescent dendrimer 8 (with aldehyde extremi-
ties) was tested with the quenchers that are soluble in di-
choromethane, namely 2NP, 3NP, 4NP, 2NR and CS, by fol-
lowing the Stern–Volmer procedure. In addition to an initial
solution of 8 (0.138 mm) in distilled dichloromethane, small
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7665

A.-M. Caminade, J.-P. Majoral, C. Sanchez et al.
The solvents were freshly dried and distilled (THF and diethyl ether over
sodium/benzophenone, pentane and CH2Cl2 over phosphorus pentoxide).
ESI mass spectrometry was carried out by using a PE SciexAPI 365 in-
strument. 1H, 13C and 31P NMR spectra were recorded by using Bruker
ARX250, AV300 and AV500 spectrometers, respectively. The references
for the chemical shifts were 85% H3PO4 for 31P NMR spectra and SiMe4
for 1H and 13C NMR spectra. The attribution of 1H and 13C NMR signals
was performed by using the J-modulated spin echo (Jmod), two-dimension-
al HMBC and HMQC and broadband or continuous-wave 31P decoupling
experiments where necessary. The numbering used in the NMR spectra is
shown in the Scheme 3. Compounds 2 and 3 are described elsewhere[22]
and compound 6 was synthesised according to a previously reported pro-
cedure.[16]
Figure 8. The Stern–Volmer procedure for dendrimer
8 in solution
(0.138 mm) and quenching solution (8, 0.138 mm and quencher,
Compound 1: The sodium salt of 4-hydroxybenzaldehyde (0.5 g,
3.47 mmol) was added to a solution of hexachlorocyclotriphosphazene
(5 g, 14.38 mmol) in THF (400 mL) at 08C, and the reaction mixture was
stirred at room temperature overnight. After removal of salts by centrifu-
gation, the clear solution was concentrated under reduced pressure and
1 mgmLÀ1). We performed two tests for each compound and all tests
~ ~
&
were performed in CH2Cl2. &, : 2NP (10 mg/10 mL),
,
: 3NP (10 mg/
* *
, : CS (10 mg/10 mL), *,*: 2NR
^ ^
10 mL),
,
: 4NP (10 mg/10 mL),
(10 mg/10 mL).
and in the solid state, the
device gives statistically more
chances to create the hydrogen
bond that extinguishes the fluo-
rescence of the dendrimers.
Scheme 3. Atom labelling system used for NMR spectral assignments.
Conclusion
subjected to flash chromatography (pentane/AcOEt, 1:0 to 6:1) to afford
as
white powder (65%). Uncorrected m.p. 538C; 1H NMR
(300.13 MHz,
4J
(H,P)=1.5 Hz, 2H;
1
a
Herein, we reported the con-
CDCl3):
d=7.42
(dd,
3J
(H,H)=8.4 Hz,
N
struction of a hybrid material composed of fluorescent den-
drimers anchored to nanocrystalline mesoporous titania thin
films, which provides the first example, as far as we know, of
such a compound. The new fluorescent dendrimers have
been specifically designed to act as fluorescent probes and
to be grafted to the substrate. At the same time, titania
films offer the perfect environment for the grafting of the
organic molecules due to their chemical composition and
stability, their mesoporous cavities and their optical proper-
ties that do not interfere with those of the dendrimers.
The characterisation of these new hybrid mesostructured
films has demonstrated that they have fluorescent proper-
ties. The dendrimers grafted to the films display higher
quenching efficiency towards phenolic OH moieties (espe-
cially those from resorcinol and 2-nitroresorcinol) than in
solution. This effect is attributed to the increased spatial
proximity of the fluorescent molecules, which is induced by
pore confinement that makes the formation of hydrogen
bonds between the hydroxyl moieties of the quenchers and
the carbonyl groups of the dendrimer easier.
2
3
G
C
H), 7.98 (d, 3J
E
À
À
10.03 ppm (s, 1H; CHO); 13C{1H} NMR (75.46 MHz, CDCl3): d=122.13
(d, 3J
C
N
d=11.73 (t, 2J
N
(P,P)=62.4 Hz,
À
+
À
P Cl2); ESIMS: m/z: 434 [M+H] .
Compound 4: A solution of 2 (1.73 g, 4.68 mmol) in THF (5 mL) was
added to a solution of 1 (390 mg, 0.91 mmol) in THF (20 mL) at 08C,
then caesium carbonate (3.04 g, 9.36 mmol) was added and the reaction
mixture was stirred at room temperature for two days. After removal of
salts by centrifugation, the clear solution was concentrated under reduced
pressure and subjected to flash chromatography (hexane/ether to THF)
to afford 4 as a yellow powder (86%). 1H NMR (500.33 MHz, CDCl3):
a’
b’
À
À
d=2.96–3.02 (m, 10H; C H), 3.85–3.88 (m, 10H; C H), 6.94–6.98 (m,
2’
3’
2
À
À
À
10H; C H), 7.07–7.17 (m, 12H; C H and C H), 7.30–7.41 (m, 30H;
Cm H and C H), 7.42–7.47 (m, 20H; C H), 7.70–7.73 (d, 3J
G
p
o
À
À
À
10 Hz, 2H; C H), 9.95 ppm (s, 1H; CHO); 13C{1H} NMR (125.8 MHz,
CDCl3): d=33.96 (s, Ca’), 33.99 (s, Ca’), 39.51 (s, Cb’), 39.54 (s, Cb’), 121.10
(m, C2’), 121.46 (m, C2), 128.55 (s, Cm), 128.57 (s, Cm), 128.62 (s, Ci),
128.64 (s, Ci), 129.80 (s, C3’), 129.84 (s, C3’), 129.89 (s, Co and Cp), 131.32
(s, C3), 133.14 (s, C4), 134.91 (s, C4’), 135.08 (s, C4’), 136.14 (s, C=C),
136.17 (s, C=C), 149.16–149.31 (m, C1’), 155.35–155.41 (m, C1), 170.53 (s,
C=O), 170.54 (s, C=O), 190.95 ppm (s, CHO); 31P{1H} NMR
(202.53 MHz, CDCl3): d=8.47 ppm (m, P=N); ESIMS (CH3CN): m/z:
2098 [M+H]+.
3
À
Therefore, the improvement in sensitivity resulting from
the immobilisation of these new probes in a film has allowed
the construction of materials with a potential application as
detectors for hazardous phenolic compounds.
Compound 5: A freshly prepared solution of N-methyldichlorothiophos-
phorhydrazide (0.16 mmol) in chloroform (6 mL) was added to a solution
of compound 4 (750 mg, 0.357 mmol) in chloroform (5 mL) at RT. The
reaction mixture was stirred overnight at RT, concentrated under reduced
pressure to about 4 mL and then precipitated with pentane. The powder
was filtered off, dissolved in the minimum amount of THF (ꢀ1 mL) and
precipitated with pentane. These washings were repeated three times to
afford 5 as a yellow powder (91%). 1H NMR (300.13 MHz, CDCl3): d=
Experimental Section
a’
3
ACHTREUNG
À
À
2.91–3.01 (m, 10H; C H), 3.40 (d, J(H,P)=14.1 Hz, 3H; N Me), 3.78–
Synthesis and characterisation of the dendrimers: All manipulations were
carried out by using standard high-vacuum and dry-argon techniques.
3.91 (m, 10H; Cb’ H), 6.88–6.99 (m, 12H; C H and C H), 7.07 (d,
2’
2
À
À
À
7666
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

Hybrid Nanocomposite Films
FULL PAPER
3’
3’
1
3J
G
G
was obtained as a yellow powder (85%). H NMR (200.13 MHz, CDCl3):
À
À
7.27–7.39 (m, 30H; Cm H and C H), 7.39–7.47 (m, 20H; C H), 7.54
d=2.85–3.05 (m, 10H; C H), 3.29 (d, J
(H,P)=10.3 Hz, 3H; Me N P),
(H,P)=7.0 Hz, 12H; P O Me), 3.80–4.10 (m, 10H; C H),
(H,P)=9.0 Hz, 2H; HO CH), 6.80–7.70 ppm (m, 83H; CH=N
p
o
a’
3
ACHTREUNG
À
À À
(d, 3J
(H,H)=8.7 Hz, 2H;
H), 7.69–7.70 ppm (m, 1H; CH=N);
3.50 (d, 3J
A
b’
À
À
À
13C{1H} NMR (75.46 MHz, CDCl3): d=31.77 (d, 3J
(C,P)=13 Hz,
4.95 (d, 2J
A
À
13
1
À À
CH3 N), 33.98 (s, C ), 39.49 (s, C ), 39.54 (s, C ), 121.16 (brs, C ),
121.34 (brs, C2), 128.52 (s, Cm), 128.55 (s, Cm and Ci), 128.64 (s, Ci), 129.76
(s, C3), 129.83 (s, C3’), 129.87 (s, Cp and Co), 130.94 (s, C4), 134.71 (s, C4’),
134.82 (s, C4’), 134.87 (s, C4’), 136.14 (s, C=C), 136.20 (s, C=C), 141.17 (d,
and Har); C{ H} NMR (50.323 MHz, CDCl3): d=33.96 (brs, CH3 N P
and Ca’), 39.52 (s, Cb’), 53.80 (d, 2J
(C,P)=7.0 Hz, P O Me), 70.03 (d,
E
À
À
1J
(C,P)=160.9 Hz, HO CH), 121.05–121.5 (m, C and C2’’), 121.85 (s,
2’
À
C2), 128.21 (s, C3), 128.55 (s, Cm), 128.60 (s, Ci), 129.70–129.90 (m, C3’, Cp
2
3J
(C,P)=18.9 Hz, CH=N), 149.26–149.37 (m, C1’), 151.99 (brs, C1), 170.51
and Co), 131.70 (d, J
(C,P)=5.7 Hz, C3’’), 133.41 (s, C4), 134.69–134.78 (m,
C4’’ and C4’), 136.07 (s, C=C), 139.20 (m, CH=N), 149.26 (m, C1’ and C1),
(s, C=O), 170.53 ppm (s, C=O); 31P{1H} NMR (121.49 MHz, CDCl3):
d=8.72 (m, P=N), 63.02 ppm (s, P=S).
150.81 (d, 2J(C,P)=8.9 Hz, C1’’), 170.49 (s, C=O), 170.54 ppm (s, C=O);
A
31P{1H} NMR (202.53 MHz, CDCl3): d=8.70 (m, P=N), 23.06 (m, P=O),
62.16 ppm (s, P=S).
Compound 7: Azabisdimethylphosphonate (0.742 mg, 1.94 mmol) was
added to a solution of 5 (2 g, 0.88 mmol) and caesium carbonate (1.269 g,
3.89 mmol) in THF (20 mL) at 08C. The reaction mixture was stirred
overnight at RT. Salts were then removed by centrifugation and the clear
solution was concentrated under reduced pressure. The residue was then
dissolved in the minimum amount of THF (ꢀ1 mL) and precipitated
with pentane. The resulting powder was filtered off and the procedure
was repeated twice to afford 7 as a yellow powder (83%). 1H NMR
Fluorescence characterisation of the dendrimers in solution: Absorption
and fluorescence measurements were performed in solution in distilled
dichloromethane at RT by using a Specord 205 spectrophotometer (Ana-
lytik Jena AG, Germany) and a PTI model QM-4 spectrofluorimeter
(Photon Technology International, USA), respectively. The quenching
experiments were also performed in solution in dichloromethane, accord-
ing to the Stern–Volmer procedure.[32]
(300.13 MHz, CDCl3): d=2.73 (t, J
(H,H)=8.1 Hz, 4H; C H), 2.91–2.99
(m, 10H;
H), 3.01 (t, 3J
H), 3.18 (d,
(H,P)=10.2 Hz,
(H,P)=10.2 Hz, 24H; P O CH3), 3.81–3.87
Synthesis of the films: Titania mesoporous films were prepared as previ-
ously reported in the literature.[14,24] The sol-gel solution was prepared
with TiCl4 as the inorganic precursor and Pluronic F127 block copolymer
2J
(H,P)=9.0 Hz, 8H; N CH2 P(O)
À
3H; CH3 N P), 3.72 (d, 3J
(m, 10H; C H), 6.90–7.00 (m, 12H; C H and C H), 7.01–7.18 (m,
18H; C H, C H and C H), 7.26–7.39 (m, 30H; C H and C H),
À
À
À
À
(HO
(CH2CH2 O)
G
N
3’’
p
À
turing agent in volatile ethanol/water as the solvent. All reagents were
used as provided by Sigma–Aldrich. A solution composition of TiCl4/
EtOH/H2O/F127 was adopted, with molar ratios of 1:40:10:0.005. This
transparent and slightly viscous solution was homogenised by stirring for
15 min at RT. The solution was then deposited by dip coating the surface
of the substrate (100 Si wafer from MEMC Electronic Materials and
quartz) at a constant rate of 2.5 mmsÀ1 in a controlled atmosphere. The
liquid layers were evaporated at a fixed RH of 15%, then transferred
and aged in a sealed environmental chamber with a fixed RH of 70% for
18 h. The following thermal treatment was then applied: 24 h at 1308C
and 3 h at 3508C to increase the density of the amorphous TiO2, then
flash heating at 5008C for 10 min followed by flash heating at 5508C for
5 min.
3
7.38–7.47 (m, 20H;
H), 7.59 (d, 3J
À
À
7.67 ppm (s, 1H; CH=N); 13C{1H} NMR (75.46 MHz, CDCl3): d=32.92–
a
À À
33.06 (m, CH3 N P and C ), 33.91 (s, C ), 33.97 (s, C ), 39.46 (s, C ),
39.53 (s, Cb’), 49.43 (dd, 1J
P(O)
C
(C,P)=7.1 Hz, N CH2
(C,P)=7.5 Hz, Cb),
À
3J
C1’), 151.40 (m, C1), 170.49 ppm (s, C=O); 31P{1H} NMR (121.49 MHz,
CDCl3): d=8.66 (s, P=N), 26.89 (s, P=O), 63.10 ppm (s, P=S).
Compound 8: The sodium salt of 4-hydroxybenzaldehyde (105 mg,
0.718 mmol) was added to a solution of 5 (736 mg, 0.325 mmol) in THF
(20 mL) at 08C, and the reaction mixture was stirred overnight at RT.
Salts were then removed by centrifugation and the clear solution was
concentrated under reduced pressure. The residue was then dissolved in
the minimum amount of THF (ꢀ1 mL) and precipitated with pentane.
The resulting powder was filtered off and the procedure was repeated
twice to afford 8 as a yellow powder (83%). 1H NMR (500.33 MHz,
For comparison purposes, mesoporous (ZrO2)x(SiO2)1Àx (x=0.1, 0.2)
T
films were synthesised according to the literature.[33] Physical characteri-
sation (TEM, XRD and EEP) was performed as for titania films.
Functionalisation of titania films: The functionalisation process was car-
ried out by dipping the titania films in a 1 mm solution of dendrimer 7 in
dichloromethane. The film was then removed from the solution at a con-
stant rate of 0.16 mmsÀ1 at 408C in a temperature-controlled chamber
under an air fluxto facilitate the solvent evaporation. The as-prepared
films were rinsed with dichloromethane to remove the non-bonded den-
a’
CDCl3): d=2.93 (t, 3J
À
3.37 (d, 3J
(H,P)=10.9 Hz, 3H; Me N P), 3.81 (t, 3J
(H,H)=8.1 Hz, 4H;
H), 3.83–3.87 (m, 6H; C H), 6.89–6.99 (m, 12H; C H and C H),
b’
2
À
C
drimers and then dried under air. The (ZrO2)x(SiO2)1Àx (x=0.1, 0.2) films
U
7.08 (d, 3J
(H,H)=8.5 Hz, 4H; C H), 7.12–7.15 (m, 6H; C H), 7.28–
were soaked in 0.1 mm solutions of dendrimer 9 in THF at RT for 2 h,
rinsed with THF and dried under an air flux. The characterisation was
performed by using EEP and FTIR, XPS, fluorescence and UV/Vis spec-
troscopies.
7.29, 7.29–7.45 (2m, 54H; C2’’ H, C H, C H and C H), 7.56 (d,
3J
3J
C
À
H), 7.71 (s, 1H; CH=N), 7.85 (d,
3
(H,H)=8.6 Hz, 4H; C3’’ H), 9.92 ppm (s, 2H; CHO); 13C{1H} NMR
(125.80 MHz, CDCl3): d=32.92 (d, 3J
(H,H)=13.3 Hz, CH3 N P), 33.97
Quenching experiments: The influence of different hydroxyl-containing
molecules on the fluorescence properties of the bonded dendrimers was
tested. These reactants are displayed in Scheme 2. They were purchased
from Sigma–Aldrich and used without further purification. Water, etha-
nol and 1-pentanol were used as received.
(s, Ca’), 33.99 (s, Ca’), 39.47 (s, Cb), 39.54 (s, Cb’), 121.05–121.20 (m, C2’),
121.44 (s, C2), 122.07 (d, 3J(C,P)=5.0 Hz, C2’’), 128.21 (s, C3), 128.53 (s,
A
Cm), 128.57 (s, Cm), 128.61 (s, Ci), 128.62 (s, Ci), 129.70–129.90 (m, C3’, Cp
and Co), 131.47 (s, C3’’) 133.46 (s, C4), 134.72 (s, C4’), 134.83 (s, C4’), 134.91
(s, C4’), 136.13 (s, C=C), 139.96–140.15 (m, CH=N), 149.33 (m, C1’),
151.76 (m, C1), 155.15 (d, 2J(C,P)=6.9 Hz, C1’’), 170.51 (s, C=O),
The functionalised films were soaked in a 10 mm solution of the quencher
in CHCl3 or EtOH for 10 min (see Table 3 for details) with constant agi-
tation at RT. They were then rinsed with the same organic solvent and
dried under air. The fluorescence properties were measured immediately
to avoid the effects of humidity on the readings.
190.76 ppm (s, CHO); 31P{1H} NMR (202.53 MHz, CDCl3): d=8.66 (m,
P=N), 60.69 ppm (s, P=S).
Compound 9: Compound 8 (780 mg, 0.321 mmol) was dissolved in THF
(0.8 mL) in presence of a catalytic amount of NEt3, then dimethyl phos-
phite (69.6 mL, 0.706 mmol) was added to the solution, and the reaction
mixture was stirred overnight at RT. After completion, water was added
dropwise and the organic residues were extracted twice with CH2Cl2. The
organic phase was dried over Mg2SO4, filtered and concentrated, then
dissolved in the minimum amount of THF (ꢀ1 mL), precipitated with
ether/pentane and filtered to remove excess dimethyl phosphite. After
evaporation of the residual solvent under reduced pressure, compound 9
Characterisation of the devices: EEP investigations were performed by
using a variable-angle ellipsometer 2000U from Woolam in the 500–
1000 nm range and by varying the relative humidity of the atmosphere
over the TiO2 films. The porous volume of the TiO2 films was calculated
by using the Brugemann effective medium approximation (BEMA) from
the optical properties of the pores and of a TiO2 reference film that was
prepared without structuring agent and that had the same thermal history
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7667

A.-M. Caminade, J.-P. Majoral, C. Sanchez et al.
as its mesoporous counterpart. The fractions of porous volume filled with
water during the analysis were calculated by using BEMA from a mixof
dry mesoporous film and completely water-saturated film. All details of
the experimental setup and data treatments have been described previ-
joral, M. Blanchard-Desce, Angew. Chem. 2006, 118, 4761–4764;
Angew. Chem. Int. Ed. 2006, 45, 4645–4648.
[9] a) H. Sohn, R. M. Calhoun, M. J. Sailor, W. C. Trogler Angew.
Chem. 2001, 113, 2162–2163; Angew. Chem. Int. Ed. 2001, 40, 2104–
2105; Angew. Chem. Int. Ed. 2001, 40, 2104–2105; b) X. Wang, H.
Zeng, L. Zhao, J. M. Lin, Talanta 2006, 70, 160–168; c) Y. Wang, K.-
M. Wang, G.-L. Shen, R.-Q. Yu, Talanta 1997, 44, 1319–1327.
[10] a) J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T.
Kresge, K. D. Schmitt, C. T-W. Chu, D. H. Olson, E. W. Sheppard,
S. B. McCullen, J. B. Higgins, J. L. Schlenker, J. Am. Chem. Soc.
1992, 114, 10834–10843; b) G. J. A. A. Soler-Illia, C. Sanchez, B.
Lebeau, J. Patarin, Chem. Rev. 2002, 102, 4093–4138.
[11] a) C. J. Brinker, Y. Lu, A. Sellinger, H. Fan, Adv. Mater. 1999, 11,
579–585; b) D. Grosso, F. Cagnol, G. J. de A. A. Soler-Illia, E. L.
Crepaldi, H. Amenitsch, A. Brunet-Bruneau, A. Bourgeois, C. San-
chez, Adv. Funct. Mater. 2004, 14, 309–322.
[12] J. D. Bass, D. Grosso, C. Boissiere, E. Belamie, T. Coradin, C. San-
chez, Chem. Mater. 2007, 19, 4349–4356.
[13] a) C. Sanchez, C. Boissiere, D. Grosso, C. Laberty, L. Nicole, Chem.
Mater. 2008, 20, 682–737; b) M. Kuemmel, D. Grosso, C. Boissiere,
B. Smarsly, T. Brezesinski, P. A. Albouy, H. Amenitsch, C. Sanchez,
Angew. Chem. 2005, 117, 4665–4668; Angew. Chem. Int. Ed. 2005,
44, 4589–4592; c) D. Grosso, C. Boissiere, B. Smarsly, T. Brezesin-
ski, N. Pinna, P. A. Albouy, H. Amenitsch, M. Antonietti, C. San-
chez, Nature Mater. 2004, 3, 787–792; d) E. L. Crepaldi, G. J. A. A.
Soler-Illia, D. Grosso, P. A. Albouy, C. Sanchez, Chem. Commun.
2001, 1582–1583.
[14] E. L. Crepaldi, G. J. d. A. A. Soler-Illia, D. Grosso, F. Cagnol, F.
Ribot, C. Sanchez, J. Am. Chem. Soc. 2003, 125, 9770–9786.
[15] M. Etienne, D. Grosso, C. Boissiere, C. Sanchez, A. Walcarius,
Chem. Commun. 2005, 4566–4568.
[16] a) L. Griffe, M. Poupot, P. Marchand, A. Maraval, C.-O. Turrin, O.
Rolland, P. MØtivier, G. Bacquet, J.-J. FourniØ, A.-M. Caminade, R.
Poupot, J.-P. Majoral, Angew. Chem. 2007, 119, 2575–2578; Angew.
Chem. Int. Ed. 2007, 46, 2523–2526; b) M. Poupot, L. Griffe, P.
Marchand, A. Maraval, O. Rolland, L. Martinet, F. E. L’Faqihi-
Olive, C.-O. Turrin, A.-M. Caminade, J.-J. FourniØ, J.-P. Majoral, R.
Poupot, FASEB J. 2006, 20, 2339–2351.
[17] V. Maraval, A. M. Caminade, J.-P. Majoral, J.-C. Blais, Angew.
Chem. 2003, 115, 1866–1870; Angew. Chem. Int. Ed. 2003, 42, 1822–
1826.
ously.[34] TEM images were collected by using
a JEOL 100 CX II
(120 kV) instrument. Samples obtained by scratching the films from the
substrate were suspended in ethanol and deposited on carbon–coated
copper grids. FTIR measurements were carried out by using a Nicolet
Magna 550 spectrometer in transmission mode through the samples de-
posited on the Si wafer substrates. XPS was performed at room tempera-
ture by using a SPECS PHOIBOS 100 hemispherical analyzer with non-
monochromatic MgKa radiation as the excitation source and under a base
pressure of ꢀ10À9 mbar. The samples were glued to the sample holder
with silver epoxy. The UV/Vis absorption spectra were recorded by using
a UVIKON XL SECOMAM (UVK-Lab) spectrometer directly on the
samples deposited on quartz substrates. The steady-state fluorescent be-
haviour of the samples on quartz was recorded by using a Fluoromax-3
instrument from Horiba Jobin Yvon with the slits fixed at 2 nm. The fluo-
rescence spectrum of each sample was obtained at an excitation wave-
length of 373 nm and an emission wavelength of 520 nm, always at RT.
No excitation was performed at 290 nm to avoid the decomposition of
the dendrimers promoted by the photocatalytic activity of the titania.
The samples were water sensitive (see above) so therefore, the initial ex-
citation–emission spectra for the titania films differed slightly.
Acknowledgements
The authors thank the NOE FAME for financial support and integration
activities. E.M.-F. thanks the Conselleria dꢁEmpresa, Universitat i Ci›n-
cia of the Generalitat Valenciana for a postdoctoral fellowship. G.F. is
thankful to the European Community for a Ph.D. grant and financial sup-
port (Fonds Social EuropØen).
[1] a) G. R. Newkome, C. N. Moorefield, F. Vçgtle, Dendrimers and
Dendrons, Wiley-VCH, Weinheim, 2001; b) Dendrimers and Other
Dendritic Polymers (Eds.: J. M. J. FrØchet, D. A. Tomalia), Wiley,
Chichester, 2001; c) J.-P. Majoral, A. M. Caminade, Chem. Rev.
1999, 99, 845–880.
[2] A. M. Caminade, J.-P. Majoral, Acc. Chem. Res. 2004, 37, 341–348.
[3] C.-O. Turin, V. Maraval, A. M. Caminade, J.-P. Majoral, A. Mehdi,
C. Reye, Chem. Mater. 2000, 12, 3848–3856.
[4] G. J. A. A. Soler-Illia, L. Rozes, M. K. Boggiano, C. Sanchez, C.-O.
Turrin, A.-M. Caminade, J.-P. Majoral, Angew. Chem. 2000, 112,
4419–4424; Angew. Chem. Int. Ed. 2000, 39, 4249–4254.
[5] A. Bouchara, L. Rozes, G. J. A. A. Soler-Illia, C. Sanchez, C.-O.
Turrin, A.-M. Caminade, J.-P. Majoral, J. Sol-Gel Sci. Technol. 2003,
26, 629–633.
[18] V. Chandrasekhar, A. Athimoolam, S. G. Srivatsan, P. S. Sundaram,
S. Verma, A. Steiner, S. Zacchini, R. Butcher, Inorg. Chem. 2002,
41, 5162–5173.
[19] L. M. Lima, F. C. F. de Brito, S. D. de Souza, A. L. P. Miranda, C. R.
Rodrigues, C. A. M. Fraga, E. J. Barreiro, Bioorg. Med. Chem. Lett.
2002, 12, 1533–1535.
[20] P. C. AngelomØ, S. Aldabe-Bilmes, M. E. Calvo, E. L. Crepaldi, D.
Grosso, C. Sanchez, G. J. A. A. Soler-Illia, New J. Chem. 2005, 29,
59–63.
[6] a) S. Slomkowski, B. Miksa, M. M. Chehimi, M. Delamar, E. Cabet-
Deliry, J.-P. Majoral, A. M. Caminade, React. Funct. Polym. 1999,
41, 45–57; b) B. Miksa, S. Slomkowski, M. M. Chehimi, M. Delamar,
J.-P. Majoral, A. M. Caminade, Colloid Polym. Sci. 1999, 277, 58–
65; c) V. Le Berre, E. TrØvisiol, A. Dagkessamanskaia, S. Sokol,
A. M. Caminade, J.-P. Majoral, B. Meunier, J. FranÅois, Nucleic
Acids Res. 2003, 31, e88/1-e88/8; d) E. TrØvisiol, V. Le Berre-Anton,
J. Leclaire, G. Pratviel, A. M. Caminade, J.-P. Majoral, J. M. Fran-
Åois, B. Meunier, New J. Chem. 2003, 27, 1713–1719; e) L. Nicu, M.
Guirardel, F. Chambosse, P. Rougerie, S. Hinh, E. TrØvisiol, J.-M.
FranÅois, J.-P. Majoral, A. M. Caminade, E. Cattan, C. Bergaud,
Sens. Actuators B 2005, 110, 125–136.
[21] D. PrØvôtØ, A. M. Caminade, J.-P. Majoral, J. Org. Chem. 1997, 62,
4834–4841.
[22] G. Franc, S. Maz›res, C.-O. Turrin, L. Vendier, C. Duhayon, A.-M.
Caminade, J.-P. Majoral, J. Org. Chem. 2007, 72, 8707–8715.
[23] a) P. H. Mutin, G. Guerrero, A. Vioux, J. Mater. Chem. 2005, 15,
3761–3768; b) G. Guerrero, P. H. Mutin, A. Vioux, Chem. Mater.
2001, 13, 4367–4373.
[24] D. Grosso, G. J. d. A. A. Soler-Illia, E. L. Crepaldi, F. Cagnol, C. Sin-
turel, A. Bourgeois, A. Brunet-Bruneau H. Amenitsch, P. A.
Albouy, C. Sanchez, Chem. Mater. 2003, 15, 4562–4570.
[25] Y. Sakatani, D. Grosso, L. Nicole, C. Boissiere, G. J. de A. A. Soler-
Illia, C. Sanchez, J. Mater. Chem. 2006, 16, 77–82.
[26] E. Martinez-Ferrero, D. Grosso, C. Boissiere, C. Sanchez, O. Oms,
D. Leclercq, A. Vioux, F. Miomandre, P. Audebert, J. Mater. Chem.
2006, 16, 3762–3767.
[27] N. Adden, L. J. Gamble, D. G. Castner, A. Hoffmann, G. Gross, H.
Menzel, Langmuir 2006, 22, 8197–8204, and references therein.
[28] P. C. AngelomØ, G. J. Soler-Illia, Chem. Mater. 2005, 17, 322–331.
[7] L. Brauge, G. VØriot, G. Franc, R. Deloncle, A.-M. Caminade, J.-P.
Majoral, Tetrahedron 2006, 62, 11891–11899.
[8] a) O. Mongin, T. R. Krishna, M. H. V. Werts, A.-M. Caminade, J.-P.
Majoral, M. Blanchard-Desce, Chem. Commun. 2006, 915–917; b) S.
Fuchs, H. Otto, S. Jehle, P. Henklein, A. D. Schlüter Chem.
Commun. 2005, 1830–1832; c) T. R. Krishna, M. Parent, M. H. V.
Werts, L. Moreau, S. Gmouh, S. Charpak, A. M. Caminade, J.-P. Ma-
7668
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7658 – 7669

Hybrid Nanocomposite Films
FULL PAPER
[29] a) T. A. Ostomel, G. D. Stucky, Chem. Commun. 2004, 1016–1017;
b) C. Sanchez, B. Lebeau, MRS Bull. MRS Bulletin 2001, 26, 377–
387.
[30] R. M. Silversteing, F. X. Webster, Spectrometric Identification of Or-
ganic Compounds, 6th ed., Wiley, 1998.
[33] G. J. de A. A. Soler-Illia, E. L. Crepaldi, D. Grosso, C. Sanchez, J.
Mater. Chem. 2004, 14, 1879–1886.
[34] C. Boissiere, D. Grosso, S. Lepoutre, L. Nicole, A. Brunet-Bruneau,
C. Sanchez, Langmuir 2005, 21, 12362–12371.
[31] R. Martinez-MaÇez, F. Sancenon, Coord. Chem. Rev. 2006, 250,
3081–3093.
[32] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluwer
Academic/Plenum, New York, 1999.
Received: April 1, 2008
Published online: July 30, 2008
Chem. Eur. J. 2008, 14, 7658 – 7669
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7669