Modulating the photoluminescence of bridged silsesquioxanes incorporating Eu3+-complexed n, n′-diureido-2,2′-bipyridine isomers: Application for luminescent solar concentrators

Article abstract of DOI:10.1021/cm2019026

Two new urea-bipyridine derived bridged organosilanes (P5 and P6) have been synthesized and their hydrolysis-condensation under nucleophilic catalysis in the presence of Eu³⁺ salts led to luminescent bridged silsesquioxanes (M5-Eu and M6-Eu). An important loading of Eu³⁺ (up to 11% _w) can be obtained for the material based on the 6,6′-isomer. Indeed the photoluminescence properties of these materials, that have been investigated in depth (photoluminescence (PL), quantum yield, lifetimes), show a significantly different complexation mode of the Eu³⁺ ions for M6-Eu, compared with M4-Eu (obtained from the already-reported 4,4′-isomer) and M5-Eu. Moreover, M6-Eu exhibits the highest absolute emission quantum yield value (0.18 ± 0.02) among these three materials. The modification of the sol composition upon the addition of a malonamide derivative led to similar luminescent features but with an increased quantum yield (0.26 ± 0.03). In addition, M6-Eu can be processed as thin films by spin-coating on glass substrates, leading to plates coated by a thin layer (～54 nm) of Eu³⁺-containing hybrid silica exhibiting one of the highest emission quantum yields reported so far for films of Eu ³⁺-containing hybrids (0.34 ± 0.03) and an interesting potential as new luminescent solar concentrators (LSCs) with an optical conversion efficiency of ～4%. The ratio between the light guided to the film edges and the one emitted by the surface of the film was quantified through the mapping of the intensity of the red pixels (in the RGB color model) from a film image. This quantification enabled a more accurate estimation of the transport losses due to the scattering of the emitted light in the film (0.40), thereby correcting the initial optical conversion efficiency to a value of 1.7%.

Full text of DOI:10.1021/cm2019026

ARTICLE
pubs.acs.org/cm
Modulating the Photoluminescence of Bridged Silsesquioxanes
Incorporating Eu3+-Complexed n,n0-Diureido-2,20-bipyridine Isomers:
Application for Luminescent Solar Concentrators
Julien Graffion,†,‡ Xavier Catto€en,† Michel Wong Chi Man,*,† Vasco R. Fernandes,‡,§ Paulo S. Andrꢀe,§
Rute A. S. Ferreira,‡ and Luís D. Carlos*,‡
†
ꢀ
Institut Charles Gerhardt Montpellier, 8, Rue de l’Ecole Normale, 34296 Montpellier, France
‡Department of Physics and CICECO, and §Instituto de Telecomunicac-~oes and Department of Physics, Universidade de Aveiro,
3810-193 Aveiro, Portugal
S Supporting Information
b
ABSTRACT: Two new urea-bipyridine derived bridged organosilanes (P5 and P6) have been synthesized
and their hydrolysisꢀcondensation under nucleophilic catalysis in the presence of Eu3+ salts led to
luminescent bridged silsesquioxanes (M5-Eu and M6-Eu). An important loading of Eu3+ (up to 11%w) can
be obtained for the material based on the 6,60-isomer. Indeed the photoluminescence properties of these
materials, that have been investigated in depth (photoluminescence (PL), quantum yield, lifetimes), show a
significantly different complexation mode of the Eu3+ ions for M6-Eu, compared with M4-Eu (obtained
from the already-reported 4,40-isomer) and M5-Eu. Moreover, M6-Eu exhibits the highest absolute
emission quantum yield value (0.18 ( 0.02) among these three materials. The modification of the sol
composition upon the addition of a malonamide derivative led to similar luminescent features but with an
increased quantumyield(0.26 ( 0.03). Inaddition, M6-Eu can be processed asthin films byspin-coatingon
glass substrates, leading to plates coated by a thin layer (∼54 nm) of Eu3+-containing hybrid silica exhibiting
one of the highest emission quantum yields reported so far for films of Eu3+-containing hybrids (0.34 ( 0.03) and an interesting
potential as new luminescent solar concentrators (LSCs) with an optical conversion efficiency of ∼4%. The ratio between the light
guided to the film edges and the one emitted by the surface of the film was quantified through the mapping of the intensity of the red
pixels(intheRGBcolormodel) froma filmimage. This quantification enabled a moreaccurateestimationofthe transportlossesdueto
the scattering of the emitted light in the film (0.40), thereby correcting the initial optical conversion efficiency to a value of 1.7%.
KEYWORDS: bridged silsesquioxane, luminescent solar concentrators, solꢀgel, photoluminescence, thin films
substrate doped with fluorophores, such as organic dyes, quan-
tum dots (QDs), or Ln3+ complexes, that successfully convert
the ultraviolet (UV) component of sunlight into visible light.
Through total internal reflection, a large fraction of the emitted
light (theoretically ∼75%ꢀ80% for a phosphor layer with a
refractive index of 1.5111) is trapped within the plate and guided
to the edges, where it emerges in a concentrated form that can be
collected by solar cells.10ꢀ14 Connection of the coated LSCs with
monocrystalline silicon (c-Si) solar cells resulted in photovoltaic
outputs 10%ꢀ15% higher than those observed in systems using
the bare LSC plates.4 The Ln3+ complexes exhibit insignificant
reabsorption effects (nonoverlap between absorption and emission),
relatively to those typical of organic dyes and QDs, which is an
important advantage for LSC technology. However, the number
of papers involved with LSCs incorporating Ln3+ complexes is
very scarce.8ꢀ10
1. INTRODUCTION
Research on organicꢀinorganic hybrid materials containing
trivalent lanthanide ions (Ln3+) is a very active field that has
rapidly shifted in the last couple of years to the development of
eco-friendly, versatile, and multifunctional systems with applica-
tions spanning domains as diverse as optics, environment, energy,
and biomedicine.1ꢀ4
Particularly, highly efficient blue-, green-, and red-emitting Ln3+-
containing organicꢀinorganic hybrid phosphors are of widespread
interest in materials science, because of their important role in
displays, lighting technologies, and solar energy conversion. The
processability of the hybrid materials impacts significantly on the
integration, miniaturization, and multifunctionalisation of devices
and presently the focus is placed on large-scale thin film technology,
namely in the hybrid/substrate and multilayers adhesion, thick-
ness and homogeneity control, low surface roughness, and optical
transparency.
Thermally and chemically stable transparent thin films of hier-
archically organized Ln3+-containing organicꢀinorganic hybrids were
amply reported in the past decade. For instance, full-color
Transparent thin films of Ln3+-based hybrids find widespread
applicability in disparate contexts, spanning domains such as agri-
culture and horticulture,2 active coatings for improving the con-
version efficiency of Si-based solar cells,5,6 and for luminescent
solar concentrators (LSCs).7ꢀ10 LSCs make use of a transparent
Received: July 4, 2011
Revised:
September 19, 2011
Published: October 06, 2011
r
2011 American Chemical Society
4773
dx.doi.org/10.1021/cm2019026 Chem. Mater. 2011, 23, 4773–4782
|

Chemistry of Materials
ARTICLE
Figure 1. Chemical structure of the silylated precursors used in this study. The numbers denote the labeling of the atoms on the pyridine rings.
phosphors based on mesoporous silica incorporating Eu3+- and
Tb3+-based complexes and organic ligands were recently reported.15
Moreover, dense, homogeneous, well-oriented, and highly organized
open-channel monolayers of zeolite L microcrystals have been
synthesized using an organicꢀinorganic functional linker capable of
coordinating and sensitizing Ln3+ ions.16
ethanol (75 mL), and water (30 mL) was heated under reflux for 20 h.
The mixture was cooled to room temperature then basified to pH 12,
using a 1 M sodium hydroxide solution. After distilling off the triethy-
lamine and ethanol at atmospheric pressure, the residual mixture was
cooled to 4 ꢀC for 3 h, yielding off-white crystals of 5,50-diamino-2,20-
bipyridine dihydrate. These crystals were sublimed under vacuum to
afford the pure 5,50-diamino-2,20-bipyridine as a white powder (1.60 g,
84% yield).
Synthesis of Precursors Pn (n = 5 or 6). To a solution of n,
n0-diamino-2,20-bipyridine (1.50 g, 8.05 mmol) in freshly distilled pyri-
dine (26 mL) was added under argon ICPTES (4.39 mL, 17.7 mmol).
The mixture was heated at 60 ꢀC for 2 days in a sealed Schlenk tube.
After evaporation, washing with dry acetone (4 ꢁ 30 mL), and drying,
the precursors Pn were obtained as white powders. P5 (5.13 g, 94%
yield). 1H NMR (dmso-d6): δ (ppm): 0.57 (m, 4 H); 1.15 (t, J = 6.9 Hz,
18 H); 1.50 (m, 4 H); 3.08 (m, 4 H); 3.75 (q, J = 6.9 Hz, 12 H); 6.37 (br,
2 H); 7.98 (d, J = 8.6 Hz, 2 H); 8.14 (d, J = 8.6 Hz, 2 H); 8.57 (s, 2 H);
8.75 (s, 2 H). 13C NMR (dmso-d6): δ (ppm): 8.6; 19.5; 24.6; 43.2; 59.0;
120.9; 126.4; 138.2; 139.9; 149.6; 156.3. P6 (4.77 g, 87% yield). 1H NMR
(dmso-d6): δ (ppm): 0.62 (m, 4 H), 1.15 (t, J = 6.9 Hz, 18 H), 1.59
(m, 4 H), 3.25 (m, 4 H), 3.75 (q, J = 6.9 Hz, 12 H), 7.48 (d, J = 7.8 Hz,
2 H), 7.68 (d, J = 7.8 Hz, 2 H), 7.85 (t, J = 7.8 Hz, 2 H), 8.19 (s, 2 H), 9.31
(s, 2 H); 13C NMR (dmso-d6): δ (ppm): 11.2, 18.4, 24.8, 45.0, 58.5,
108.3, 113.6, 138.5, 151.7, 155.1, 158.6.
Synthesis of Hybrid Materials Mn-Ln (n = 5 or 6; Ln = Eu,
Tb). To a stirred suspension of Pn (200 mg, 0.294 mmol) in dry ethanol
(2 mL) at room temperature was added a mixture of water (155 μL,
8.61 mmol), ethanol (0.8 mL), and lanthanide chloride hexahydrate
(9.7 10ꢀ2 mmol). A solution of ammonium fluoride (1 M in H2O, 3 μL,
3 μmol) in ethanol (0.2 mL) was added after 10 min to this clear yellow
solution (final molar ratio Pn/LnCl3/H2O/NH4F = 1:0.33:32:0.01;
final concentration of Pn = 0.1 M). Precipitation of small particles
occurred within 30 min. After 3 days of aging, the precipitate was filtered
off, successively washed with water, ethanol and acetone (3 ꢁ 10 mL of
each solvent) and finally dried at 80 ꢀC for 12 h. White powders were
obtained in all cases. Elemental analyses: M5-Eu (Si, 10.45; Eu, 1.42);
M6-Eu (Si, 9.31; N, 12.95; Cl, 6.43; Eu, 9.25); M6-Tb (N, 13.38; Tb,
9.26).
M60-Eu. A synthesis similar to that of M6-Eu was performed, from
P6 (200 mg, 0.294 mmol) europium chloride hexahydrate (55 mg,
0.15 mmol) water (155 μL, 8.61 mmol), and ammonium fluoride (1 M
in H2O, 3 μL, 3 μmol) (final molar ratio Pn/EuCl3/H2O/NH4F =
1:0.50:33:0.01; final concentration of Pn = 0.1 M). Elemental analysis:
M60-Eu (Si, 8.11; N, 12.25; Eu, 11.20).
Thin films of Ln3+-based hybrids prepared without any external
template were also explored, although more scarcely. Silylated
ligands, such as carboxylates,17 diureasils,18 and pyridine,19 have
been employed in a pure form. Most of the other examples in the
literature reporting the fabrication of thin films of Ln3+-based
hybrids involve mixtures with TEOS or tetramethylorthosilicate
(TMOS) and an organosilylated precursor.20ꢀ27
In these organicꢀinorganic hybrids, the organic fragment has
a key role in complexing and sensitizing the Ln3+ ions. For this
purpose, nitrogen-based chelating heterocycles have been widely
used and, more particularly, Ln3+-containing 4,40-diureido-2,
20-bipyridine-based bridged silsesquioxanes obtained from
precursor P428 (Figure 1) have recently attracted considerable
interest, in view of their potential use as optically active hybrid
materials.16,29ꢀ32
Until now, only the 4,40-isomer of this system has been stud-
ied. As the position of the ureido substituents on the 2,20-bipyridine
fragments (basicity, steric hindrance, chelation by urea groups),33
should significantly impact on the light-emitting features of the
Ln3+-containing materials, we decided to investigate the effect of the
substituent position on the bipyridine ring by synthesizing two new
related precursors (denoted P5 and P6; see Figure 1). Here, we
report the preparation of Eu3+- and Tb3+-containing bridged
silsesquioxanes obtained by hydrolysisꢀcondensation of P5 or P6
in the presence of europium and terbium chloride and their photo-
luminescence (PL). In particular, we will focus on Eu3+-containing
hybrids derived from P6 that can be processed as transparent
and mechanically stable thin films with an intriguing potential as
new LSCs.
2. EXPERIMENTAL SECTION
Synthesis. All the manipulations were carried out using Schlenk
techniques under dry argon. Dry, oxygen-free solvents were employed.
(3-isocyanatopropyl)triethoxysilane (ICPTES, Aldrich) and water were
purified by distillation prior to use. Europium(III) chloride hexahydrate
(99.99%) and terbium(III) chloride hexahydrate (99.9%) were pur-
chased from Aldrich (Lot No. 203254) and Alfa Aesar, respectively. 6,
60-Diamino-2,20-bipyridine (purified by sublimation under vacuum),34
P4,28 and Pmal35,36 were synthesized according to previously reported
procedures.
M6-Eu-mal. A synthesis similar to that of M6-Eu was performed,
from P6 (200 mg, 0.294 mmol), Pmal (41 mg, 9.7 10ꢀ2 mmol),
europium chloride hexahydrate (36 mg, 9.7 10ꢀ2 mmol), water (210 μL,
11.7 mmol), and ammonium fluoride (1 M in H2O, 4 μL, 4 μmol) (final
molar ratio Pn/Pmal/EuCl3/H2O/NH4F = 1:0.33:0.33:42:0.01; final
concentration of Pn = 0.1 M). Elemental analysis: M6-Eu-mal (Si, 8.38;
N, 11.98; Cl, 6.97; Eu, 7.08).
5,50-Diamino-2,20-bipyridine.37,38 A mixture of 5,50-bis(2,
5-dimethylpyrrol-1-yl)-2,20-bipyridine38 (3.50 g, 10.2 mmol), hydroxy-
laminehydrochloride(21.3 g, 307 mmol), triethylamine(11.5 mL, 82 mmol),
4774
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
F6-Eu Thin Film. Borosilicate substrates (Normax microslides,
25 mm ꢁ 25 mm) were cleaned with acetone, immersed in a mixture
of hydrogen peroxide and sulfuric acid, and rinsed and stored with
distilled water. Prior to the deposition, they were dried by spin-coating
(5 000 rpm, 50 s), treated with ethanol, and dried again by spin-coating
just before use. Europium chloride hexahydrate (7.1 mg, 1.9 ꢁ 10ꢀ2 mmol)
and water (31 μL, 1.7 mmol) were incorporated into a suspension of P6
(40 mg, 5.9 ꢁ 10ꢀ2 mmol) in dry ethanol (2.5 mL). The mixture was
stirred for 20 min at room temperature, leading to a clear solution; then,
an ammonium fluoride solution (1 μL, 1 M in water) was added (final
molar ratio P6/EuCl3/H2O/NH4F = 1:0.33:32:0.01; final massic
(weight) concentration of P6 = 2%w). The substrates were held by
suction on a chuck, which was placed on the axis of the spin coater (SCS
Specialty Coating Systems, spin-coat G3-8). The F6-Eu thin films were
prepared by spin-coating two drops of the sol on glass substrates, with an
acceleration time of 10 s and a spin time of 50 s, at a spin rate of
5000 rpm. They were finally dried at 80 ꢀC for 10 h.
Fourier Transform Infrared (FT-IR) Spectroscopy. The FT-IR
spectra were acquired between 400 and 4000 cmꢀ1 at room temperature
using a Mattson Mod 7000 spectrometer. The solid samples (3ꢀ4 mg)
were finely ground, mixed with 300 mg of dried KBr (Merck, spectroscopic
grade) and pressed into pellets, which were left for 24 h at 80 ꢀC in order to
reduce the level of absorbed water before use.
Nuclear Magnetic Resonance (NMR). Nuclear magnetic reso-
nance (NMR) spectra were recorded at 298 K on a Bruker Model
Advance 400 apparatus. 1H and 13C chemical shifts are reported in units
of ppm, relative to Me4Si.
Photoluminescence. The photoluminescence (PL) spectra were
recorded at 10 and 300 K and room temperature with a modular double-
grating excitation spectrofluorimeter with a TRIAX 320 emission mono-
chromator (Fluorolog-3 2-Triax, Horiba Scientific) coupled to a R928
Hamamatsu photomultiplier, using the front face acquisition mode. The
excitation source was a 450-W xenon arc lamp. The emission spectra
were corrected for detection and optical spectral response of the
spectrofluorimeter and the excitation spectra were weighed for the
spectral distribution of the lamp intensity using a photodiode reference
detector. The time-resolved measurements were acquired with the setup
previously described for the PL spectra, using a pulsed Xe-Hg lamp (6 μs
pulse at half width and 20ꢀ30 μs tail).
Absolute Emission Quantum Yields. The absolute emission
quantum yields were measured at room temperature using a quantum
yield measurement system (Model C9920-02, from Hamamatsu, with
a 150-W xenon lamp coupled to a monochromator for wavelength
discrimination), an integrating sphere as the sample chamber, and a
multichannel analyzer for signal detection. Three measurements were
made for each sample, so that the average value is reported. The method
is accurate to within 10%.
Ellipsometry. The spectroscopic ellipsometry measurements were
made using an AutoSE ellipsometer (Horiba Scientific) with a total of
250 points in the wavelength interval of 440ꢀ850 nm, an incidence
angle of 70ꢀ, an acquisition time of 22 ms per point, and an average of
10 measurements per point. Refractive index of the film were
calculated using the Lorentz model, which expresses the relative
complex dielectric constant as a function of the frequency (ω), which
is described by
Powder X-ray Diffraction (XRD). X-ray diffraction (XRD) mea-
surements of the dried powder samples were carried out in 1.5-mm-
diameter glass capillaries in a transmission configuration at the LCVN
Laboratory (Montpellier, France). A copper rotating-anode X-ray source
working at 4 kW with a multilayer focusing Osmic monochromator,
giving high flux and punctual collimation, was employed. A two-
dimensional (2D) image plate detector was used.
Scanning Electron Microscopy (SEM). Scanning electron mi-
croscopy (SEM) images were obtained with a Hitachi Model S-4800
apparatus after platinum metallization.
ðεs ꢀ ε∞Þ ꢁ ω20
ε ¼ ε∞
þ
ð1Þ
ðω20 ꢀ ω2Þ þ iΓω
where ε∞ is the high-frequency relative dielectric constant, εs the static
relative dielectric constant, ω0 (eV) the oscillator resonant frequency,
and Γ (eV) the damping factor.39 The fit method was detailed
elsewhere.40
3. RESULTS AND DISCUSSION
Optical Imaging. Optical micrographs were acquired with a Leica
Model DM6000 M microscope. The images were obtained through a
2592 ꢁ 1944 pixel CCD Leica Model DFC425C camera. Photographs
were taken with a digital camera (Canon Model EOS 400D), using an
exposure time between 6 and 20 s. Excitation of the film was performed
with a Spectroline E-Series UV lamp (Aldrich, Model Z169625) operat-
ing at 254 and 365 nm. The spectral intensity maps were performed
using the MATLAB programming language, considering the RGB
model to determine the intensity value of the red pixels (ranging from
0 to 255). In order to quantify the emission ratio factor (C), cross-
sectional intensity profiles of the pixels (red intensity values) were
measured in different planes of six film images. The photographs used to
The three organosilane precursors—P4ꢀP6—were obtained
in good to excellent yield from the corresponding n,n0-diamino-
2,20-bipyridine by reaction with (3-isocyanatopropyl)triethoxy-
silane in hot pyridine.28 Their hydrolysisꢀcondensation in the
presence of europium chloride was investigated for different
compositions. To evaluate the effect of the ureido substituent’s
positions on the luminescence properties, we will discuss, in the
first part of the manuscript, the synthesis and PL of the M5-Eu
and M6-Eu bulk materials obtained under the same conditions as
the previously reported M4-Eu.31 In the second part of the work,
we will optimize the optical properties of the materials, by modi-
fying the composition of the sol, describing the processing of
highly luminescent thin films obtained by spin-coating, with the
potential impact of new luminescent solar concentrators (LSCs).
3.1. Synthesis and Characterization of the M5-Eu and M6-
Eu Bulk Materials. For comparison purposes, we prepared a first
set of P4, P5, and P6-derived materials under similar conditions.
Materials Mn-Ln were obtained as bulk samples from a homo-
calculate the intensity maps were acquired with exposure times between
1
/
20 s and 2.5 s to avoid saturation of the intensity of the red pixels.
Elemental Analysis. Elemental analyses of the hybrids were per-
formed by the Service Central d’Analyses du CNRS (Lyon, France). The
experimental errors are (0.4% (absolute) for chlorine, (0.3% (absolute)
for nitrogen, and (2% (relative error) for europium and silicon. Europium
and silicon contents were determined by ICP-AES after alkaline diges-
tion, whereas chlorine and nitrogen contents were determined after
combustion.
UltravioletꢀVisible (UVꢀvis) Absorption Spectroscopy.
UltravioletꢀVisible (UVꢀvis) absorption spectra were measured on a
JASCO Model V-560 spectrometer with a scan range of 200ꢀ900 nm,
with a scan rate of 200 nm minꢀ1 and a resolution of 0.5 nm. A clean
substrate was used as reference in order to eliminate the substrate
contribution in the absorption spectra of the film.
geneous solution of Pn, LnCl3 6H2O, solvent (methanol or
3
ethanol), NH4F and water (Pn/LnCl3 6H2O/H2O/NH4F =
3
1:0.33:30:0.01 at 0.1 M in dry methanol (M4-Ln) or ethanol
(M5-Ln, M6-Ln)). Interestingly, whereas the precursors were
sparingly soluble in the alcoholic solvents, the introduction of the
lanthanide salt to these suspensions immediately led to complete
solubilization, indicating that complexation had occurred between
4775
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
Table 1. Elemental Composition and Calculated Molar Ratios for the Different Materials Studieda
Composition (%w)
hybrid
M6-Eu
M60-Eu
Si
N
Ln
Cl
Si/Eu
N/Ln
Cl/Eu
N/Si
9.31
8.11
ndb
12.95
12.25
13.38
11.98
9.25
11.20
9.26
7.08
1.42
4.98
6.43
5.4 (6)
3.9 (4)
ndb
15.2 (18)
11.9 (12)
16.7 (18)
18.4 (20)
3.0 (3)
2.8 (3)
3.0 (3)
ndb
2.9 (2.8)
ndb
3.4 (3)
ndb
ndb
6.97
ndb
ndb
ndb
ndb
4.2 (3)
ndb
ndb
M6-Tb
M6-Eu-mal
M5-Eu
8.38
10.45
8.29
6.4 (7)
39.9 (6)
9.0 (6)
ndb
ndb
M4-Eu (ref 31)
14.10
30.7 (18)
a Values given in parentheses indicate the theoretical ratio. b Not determined.
Figure 2. (Top) Schematic representation of the possible interactions between the Ln ion (full circle) and one diureido-2,20-bipyridine ligand.
(Bottom) Mesomeric formulas involving the pyridine nitrogen atoms.
the europium salt and the bipyridine-based precursors. After
addition of the ammonium fluoride catalyst, gelation (M4-Ln) or
precipitation (M5-Ln, M6-Ln) occurred.
depicted in Figure 2), and also from the good packing ability of
the linear monomers that would disfavor the formation of bulkier
europium complexes. In contrast, the full incorporation of the
Ln3+ salt in M6-Eu and M6-Tb might come from the chelation of
the Ln ion by both the urea and bipyridine moieties, with the
formation of a cagelike complex, as illustrated in Figure 2.
Moreover, the preservation of a Cl/Eu molar ratio of 3.0 in
M6-Eu demonstrates that no anion exchange occurred during
the solꢀgel process.
Figure S1 in the Supporting Information depicts the FT-IR
spectra of all the materials, which are all very similar. The for-
mation of the siloxane skeleton was ascertained by the appearance
of a broad band (975ꢀ1140 cmꢀ1) with two main components
located at 1020and 1140 cmꢀ1 (νas(SiOSi)). A characteristicband
at 905 cmꢀ1, attributable to noncondensed SiꢀOH groups, is also
observed, giving evidence that the condensation is not complete.
However, since no 29Si solid-state NMR spectrum could be
recorded, because of the paramagnetism of the Eu3+ ions, it is
difficult to precisely evaluate the condensation at Si. The broad
bands centered at 1550, 1605, and 1668 cmꢀ1 result from the
contributions of the vibrations in the pyridine rings and of the
amide I and amide II modes of the urea groups. This gives clear
evidence for the presence of the organic fragments within the
material.
The elemental analyses (Table 1) performed on the different
materials indicate that the N/Si ratio is similar (within the
experimental error) to the expected values, thus further confirm-
ing the preservation of the organic component within the hybrid
materials. These analyses give also very good indications of the
complexing ability of precursors Pn (n = 4, 5, or 6): whereas M6-
Eu contains as much as 9.2%w of europium (the theoretical value
for all these europium-containing materials in the case of a
complete condensation is 9.3%w), the corresponding amount is
reduced to 5.0%w in M4-Eu, and only 1.4%w in the case of M5-
Eu. This corresponds to bpy/Eu ratios of 5:1, 20:1, and 3:1, for
M4-Eu, M5-Eu, and M6-Eu, respectively. The low amount of
europium incorporated in M5-Eu may result from the low
basicity of the 5,50-diureido-2,20-bipyridine ligand, compared to
the other isomers (as inferred from the mesomeric formulas
XRD experiments performed on the dried powders exhibit
two very broad bands, peaking at ∼5.7ꢀ5.9 nmꢀ1 and
∼15.5ꢀ15.7 nmꢀ1 (see Figure S2 in the Supporting Infor-
mation), which indicates nonorganized structures, even for M5-Eu.
Notably, no sharp lines typical of inorganic crystals that would
show a phase segregation between lanthanide salts and the hybrid
materials41 could be detected.
Different morphologies were observed for these materials
by SEM (Figure 3): whereas M4-Eu exhibits a featureless
morphology,31 the structure of M5-Eu can be described as an
assembly of tiny spheres (<100 nm in size). Interestingly, M6-Eu
and M6-Tb are obtained as aggregates of microspheres (1ꢀ4 μm
in diameter) that are also observable by optical microscopy (see
Figure S3 in the Supporting Information).
The PL of these hybrid materials was investigated using
various techniques (steady-state and time-resolved emission,
excitation, and emission quantum yield measurements). For
M6-Tb, the intra-4f lines could not be discerned and the emission
spectrum is formed only by a host-related broad component (see
Figure S4 in the Supporting Information and discussion below).
The host-to-metal energy transfer is hindered by the high-energy
location of the Tb3+ 5D4 emitting level, relative to the ligands
exciting states, and then we will focus on the Eu3+-containing
hybrids. Figure 4A depicts the emission features of M5-Eu and
4776
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
5
M6-Eu, under the excitation wavelength that maximizes the
emission intensity. In order to study the effect of the substituents’
position on the PL features the spectrum of M4-Eu is also
presented.29,31 All the spectra display the Eu3+ 5D0f7F0,4
transitions and a broad band (400ꢀ550 nm), most evident for
M5-Eu, whose energy depends on the excitation wavelength (see
Figure S5 in the Supporting Information). The broad band
resembles that observed for the M4-Eu31 and pristine M442
hybrids, being ascribed to the superposition of three distinct
components: the bpy-related triplet state and electronꢀhole
recombinations originated in the urea groups cross-linkages and
in the siliceous nanoclusters.42 The Eu3+-excitation paths were
selectively studied by measuring the excitation spectra monitored
within the D0f7F2 transition (see Figure 4B). The spectra
reveal a large band in the UV/blue spectral region with two main
components, at 277 nm and 363ꢀ373 nm, and a low-relative
intensity band between 360 and 450 nm (more evident in the
excitation spectrum of M5-Eu). A similar excitation spectrum
was already observed for the analogous M4-Eu31 hybrid (curve c
inFigure4B). Whereasthelow-wavelength region (240ꢀ360 nm)
is related to the excited states of the bpy-ligands, the high-
wavelength one (360ꢀ450 nm) is associated with the excited
states of the NH/CdO groups and of the siliceous nanoclusters.43,44
The presence of both the bpy and hybrid host excited states in the
excitation spectra monitored within the Eu3+ excited states
points out the presence of bpy/hybridfEu3+ energy transfer,
as also noted for M4-Eu.31 The different energy values of the
excitation components and the distinct relative intensity of the
intra-4f lines suggest that the nature of the isomer plays an active
role in the Eu3+ sensitization.
The emission features were quantified through the measure-
ment of the absolute emission quantum yields (Table 2). Max-
imum quantum yield values of 0.01 ( 0.001 (280 nm) and
0.18 ( 0.02(350 nm) weremeasured under excitation via the bpy-
related excited states for M5-Eu and M6-Eu, respectively. The
quantum yield value previously reported for the M4-Eu is 0.08 (
0.01 (320 nm). The significant difference between those values
points out the crucial role of the ureido substituents’ positions in
the PL features.
In order to gain some insight into the Eu3+-local coordination
of the M5-Eu and M6-Eu hybrid hosts, the Eu3+-lines were
measured with high resolution (Figure 5). The energy and full
width at half-maximum (fwhm) of the intra-4f 6 transitions in the
Figure 3. SEM images of (A) M4-Eu, (B) M5-Eu, (C) M6-Eu, and
(D) M6-Eu-Mal.
Figure 4. (A) Emission and (B) excitation spectra (300 K) of (a) M6-Eu, (b) M5-Eu, and (c) M4-Eu monitored at 612 nm and excited at 373, 363, and
320 nm, respectively. The solid line with squares corresponds to the excitation spectra of M6-Eu monitored at 616 nm. The insets show a magnification
of the broad band emission of M6-Eu (spectrum a) and of the 7F0,1f5D1 transitions of M4-Eu (spectrum c).
Table 2. E00, Full Width at Half Maximum (fwhm00), 5D0 Lifetime (τ), and Maximum Absolute Emission Quantum Yieldsa of All
the Eu3+-Containing Bridged Silsesquioxanes
E00 (cmꢀ1
)
Full Width at Half Maximum, fwhm00 (cmꢀ1
)
τ (ms)
hybrid
Site 1
Site 2
Site 1
Site 2
Site 1
Site 2
ϕ(λex (nm))
M4-Eu (ref 31) 17270.8 ( 0.4
0.384 ( 0.016
0.483 ( 0.005b
0.08 ( 0.02 (320)
0.01 ( 0.001(280)
M5-Eu
17270.5 ( 0.5
56.1 ( 1.2
61.5 ( 5.9
42.6 ( 1.9
45.1 ( 1.1
50.4 ( 1.4
M6-Eu
17270.0 ( 3.3 17224.2 ( 0.2
17242.6 ( 1.9 17223.0 ( 0.11
17237.7 ( 1.1 17223.9 ( 0.2
17239.4 ( 1.1 17223.6 ( 0.1
25.0 ( 0.5
19.3 ( 0.3
21.6 ( 0.5
21.0 ( 0.3
0.313 ( 0.009 0.863 ( 0.013 0.18 ( 0.02 (350)
0.300 ( 0.020 0.850 ( 0.039 0.19 ( 0.02 (350)
0.309 ( 0.014 0.871 ( 0.016 0.26 ( 0.03 (350)
0.301 ( 0.014 0.860 ( 0.071 0.34 ( 0.03(350)
M60-Eu
M6-Eu-mal
F6-Eu
a The excitation wavelength is indicated in parentheses. b Average lifetime (Æτæ).
4777
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
Figure 5. High-resolution emission spectra (300 K) of (A) M5-Eu and (B) M6-Eu excited at 277 nm (black line), 363 nm (red line), and 373 nm (blue line).
emission spectra of M5-Eu are almost independent of the
therefore ascribed to Site 1, and another component character-
ized by lower values of E00 and fwhm00 (E00 = 17224.2 ( 0.2,
fwhm00 = 25.0 ( 0.5 cmꢀ1), which will be denoted hereafter as
5
excitation wavelength. Moreover, a single D0f7F0 line and a
J-degeneracy splitting of the 7F1,2 levels into three and four Stark
components, respectively, are observed, indicating that the Eu3+
ions occupy a low symmetry site without an inversion center, in
accordance with the higher intensity of the electric-dipole
5D0f7F2 transition.45 Nevertheless, the emission spectra of
M6-Eu show major changes in the energy, fwhm, and relative
intensity of the 5D0f7F0ꢀ4 lines under distinct excitation wave-
lengths (see Figure 5), pointing out the presence of at least two
Eu3+ average local environments. This emission dependence on
the excitation wavelength is mostly evident in the relative
intensity variation of the 5D0f 7F2 Stark components at 612 and
616 nm (marked with an asterisk in Figure 5) as the excitation
wavelength increases from 277 nm to 373 nm. In order to study
the excitation path of each Eu3+ local site selectively, the excitation
spectrum monitored at 612 nm was compared with that mon-
itored at 616 nm (see Figure 4B). Although the low-wavelength
regions of the two spectra overlap, the high-wavelength compo-
nents are blue-shifted (ca. 2050 cmꢀ1), supporting the presence
of two distinct Eu3+ local environments, as previously mentioned.
The Eu3+ local coordination, as function of the position of the
two urea substituents on the bipyridine fragments was further
analyzed through the estimation of the energy (E00) and full-
width at half maximum (fwhm00) of the 5D0f 7F0 transition (see
Figure S6 in the Supporting Information). The 5D0f7F0 transi-
tion of the M5-Eu spectrum is well-reproduced by a single
Gaussian function, yielding E00 = 17270.5 ( 0.5 cmꢀ1 and
fwhm00 = 56.1 ( 1.2 cmꢀ1 (see Table 2). To simplify the
discussion, this Eu3+ local site will be hereafter referenced as “Site 1”.
Despite the presence of a single 5D0f 7F0 line, the large fwhm00
value suggests a higher nonhomogeneous distribution of similar
Eu3+ Site 1 chemical environments, because of changes outside
the coordination polyhedron.
5
“Site 2”. The lower energy of the D0f7F0 transition of Site 2
indicates that the Eu-ligand bonds have (on average) a high cov-
alent character, relative to that found in Site 1.46,47 This is in
good agreement with the main coordination type proposed
in Figure 2 for the P6-related hybrids, which involves both
the N atoms of the bpy-ligands (which are also involved in Site
1 coordination) and the O atoms of the urea groups.
5
The D0 lifetime values were estimated by measuring the
emission decay curves within the 5D0f7F2 transition excited via
the bpy/hybrid-related broad band (277 nm). The 5D0 emission
decay curve of M5-Eu reveals a nonsingle exponential behavior
(see Figure S7 in the Supporting Information), which can be
attributed to the large distribution of Eu3+ local sites, as men-
tioned above. Therefore, an average lifetime value (Æτæ) was esti-
mated, considering
Z
t1
IðtÞt dt
t0
Æτæ ¼ Z
ð2Þ
t1
IðtÞ dt
t0
where t0 = 0 and t1 is the time interval where the luminescence
intensity (I(t)) reaches the background, yielding ∼0.483 (0.005 ms.
For M6-Eu, the emission decay curves are well-reproduced by a
biexponential function, in good agreement with the presence of
two Eu3+ local sites, yielding lifetime values of 0.313 ( 0.009 ms
and 0.863 ( 0.013 ms (see Figure S8 in the Supporting
Information).
The emission features of M5-Eu and M6-Eu were also studied
at a temperature of 10 K. Whereas the PL features of M5-Eu
resemble those acquired at room temperature, for M6-Eu, the
emission data at 10 K is almost independent of the excitation
The E00 value is the same as that reported for M4-Eu
(E00 = 17270.8 ( 0.4; see Table 2), suggesting an analogous Eu3+
local coordination in both M4 and M5 hybrids, in agreement with
the proposed schematic representation of the possible Eu3+-to-
diureido-2,20-bipyridine ligand interactions (see Figure 2). Never-
theless, the significant increase in the fwhm00 value, relative to
that of M4-Eu (fwhm00 = 41.6 ( 0.9 cmꢀ1; see Table 2), reflects a
larger distribution of related Eu3+ local coordination sites.
wavelength (see Figure S9 in the Supporting Information), and
5
only the D0f7F0 component of Site 2 is detected (E00
=
17224.5 ( 0.1, fwhm00 = 21.4 ( 0.1 cmꢀ1), suggesting that Site
1 is thermally activated in M6-Eu. The selective emission of Site 2
at 10 K is corroborated by the emission decay curves (monitored
within the 5D0f7F2 transition) which are well-reproduced by a
single exponential function (see Figure S10 in the Supporting
Information), yielding a lifetime value of 1.064 ( 0.008 ms.
Comparing this value with those measured at 300 K, we
The 5D0f7F0 transition of the M6-Eu spectrum was modeled
by the sum of two Gaussian functions displaying one component
with E00 = 17270.0 ( 3.3 cmꢀ1 and fwhm00 = 61.5 ( 5.9 cmꢀ1
,
tentatively assign the low and high lifetime values to the D0
5
comparable to that measured for M4-Eu and M5-Eu and
level of Eu3+ in Site 1 and Site 2 (see Table 2).
4778
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
Figure 6. (A) Photograph and (B) intensity map of the red pixel of F6-Eu excited at 365 nm. The top edge of the photograph corresponds to the more-
intense profile in the intensity map.
Figure 7. (A) UVꢀvis absorption spectrum and (B) high-resolution emission spectra (300 K) excited at 277 nm (black line) and 373 nm (red line) of
F6-Eu. The inset shows the excitation spectrum (300 K) monitored at 612 nm.
3.2. Optimization of the Materials and Processing as Thin
Films. Since M6-Eu is the most efficient material among the
three isomers of the Eu3+-containing bridged silsesquioxanes
investigated in this work, we focused our efforts on materials
derived from P6 to optimize the formulation of the luminescent
hybrids. The first modification implies an increase in the Eu3+
contents in the materials: the hybrid material M60-Eu was syn-
thesized with the bpy/EuCl3 ratio of 2:1 instead of 3:1 for M6-
Eu. Elemental analysis (Table 1) evidences the full Eu3+ incor-
poration into the hybrid host, where a loading as high as 11.2%w
was reached without any detectable phase separation by XRD.
The second modification consists of the incorporation of an
additional silylated precursor featuring the tetraethylmalonamide
moiety (Pmal; see Figure 1),35,36 to increase the UV-harvesting
absorption and trying to saturate the Ln3+ coordination shell
yielding material M6-Eu-mal. Indeed, the malonamide fragment
is a good ligand for lanthanides and actinides and can coordinate
the cation in a monodentate or bidentate fashion.48
from powder precipitation on a glass substrate. The emission
quantum yield measured for F6-Eu (0.34 ( 0.03) is among the
highest values ever reported for films of Eu3+-containing hybrids
(see Table 3 of ref 3). To date, only very few examples of organicꢀ
inorganic hybrids, all made with TMOS, are known: (TMOS)/
3-glycidoxypropyltrimethoxysilane and TMOS/diethoxydimethyl-
silane doped with [Eu(tta)3(H2O)2] (tta = thienoyltrifluoro-
acetonate), with values between 0.12 ( 0.02 and 0.34 ( 0.02,20
and also with TMOS, pluronic P123, EuCl3 6H2O, 1,10-phe-
3
nanthroline, and salicylic acid, with values between 0.04 and
0.29.15 The absorption coefficient, α = A/d (and the correspond-
ing UVꢀvis absorbance A of F6-Eu), is very low in the visible
spectral range (wavelengths of >400 nm), reaching a maximum at
∼350 nm (see Figure 7A) (α is not recorded for wavelengths
lower than 297 nm, because the glass substrate displays a very
high absorbance in that spectral region).
The excitation (monitored within the 5D0f7F2 transition) and
emission spectra of M60-Eu (see Figures S12 and S13 in the Sup-
porting Information), M6-Eu-mal (Figures S12 and S14 in the
Supporting Information), and F6-Eu (Figure 7B) resemble those
acquired for M6-Eu (see Figures 4 and 5). The only difference is
observed in the excitation spectrum of F6-Eu (the inset in
Figure 7B), which displays a blue-shift (∼2200 cmꢀ1), relative to
that measured for M6-Eu, which could be ascribed to differences in
the gelification and condensation rates between the spin-coating
process in the film and the solꢀgel reactions in the monoliths. For
thin films, the forced solvent extraction is much faster, compared
with that of bulk monoliths, resulting in a structure with a lower
degree of organization. This distortion on the symmetry could cause
the change in the energy levels positions of the hybrid host and,
consequently, the blue-shift in the thin film excitation spectrum.18
Interestingly, all the spectroscopic features of M60-Eu and M6-
Eu-mal are very similar to those of M6-Eu (see Figures S1 and S2
in the Supporting Information). Furthermore, although the
morphology of the former is very similar to that of M6-Eu, the
latter is obtained in a wormlike form, probably resulting from the
agglomeration of small spheres (Figure 3). In addition, the sol
yielding M6-Eu could easily be spin-coated into transparent
luminescent thin films (F6-Eu) (see Figure 6 and Figure S11 in
the Supporting Information) with a thickness value (d) of ∼54 nm,
determined by ellipsometry measurements, vide infra. As far as
we know, only Dong et al.17,49 and Chamas et al.19 spin-coated
pure silylated precursors, as we describe here, whereas Wang
et al.50,51 reported films of Ln3+-containing bridged silsesquioxanes
4779
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
In order to discuss the effects on the Eu3+ local coordination in
more detail, the energy and fwhm of the two components of the
5D0f7F0 transitionwerefitted by a sum of two Gaussian functions
(see Figure S6 in the Supporting Information). It is observed
that the variation of the bpy/Eu3+ ratio (9.2%wꢀ11.2%w Eu), the
incorporation of the malonamide fragments and the thin-film
processing induce no changes in the E00 and fwhm00 values of Site
2. In contrast, for Site 1, a decrease in the fwhm00 and a red-shift in
the E00 are detected, pointing out that the Eu3+ local coordination
at Site 1 is moreaffected than the local coordination involving both
bpy and urea ligands (Site 2). The similarities between the
emission features are reinforced by the 5D0 emission decay curves
(see Figures S15ꢀS17 in the Supporting Information), which are
well-reproduced by a biexponential function with time decay
constants similar to those measured for M6-Eu (see Table 2).
The effect of the three modifications was also monitored in the
emission quantum yield values: whereas no significant changes
were observed with the increase in Eu3+ concentration, M6-Eu-
mal and the F6-Eu film display a significant increase in the
maximum values of the emission quantum yield (0.26 ( 0.03 and
0.34 ( 0.03, respectively; see Table 2), relative to M6-Eu. We
should note that as the excitation spectra of M6-Eu and M6-Eu-
Figure 8. Refractive index variation, as function of the wavelength of
F6-Eu. The inset shows the depolarization spectrum.
fraction occupied by the nanometer-size particles is very small
(the contribution of the remaining surface is largely predominant)
and (ii) the dimension of the nanometric particles itself is also very
small, and, therefore, its effect could be neglected. Consequently,
the values of the roughness measured by ellipsometry and the
presence of the cubic nanoparticles in the surface of the hybrid are
not in contradiction.
5
mal that are monitored within the D0f7F2 transition are
identical (see Figure S12 in the Supporting Information), thus
indicating comparable 5D0 excitation paths in the presence and
absence of malonamide, we may suggest the presence of efficient
malonamidefbpyfEu3+ energy transfer, which confirms that
the malonamide fragments act as active additional UV-harvesting
moieties. Contrary to its expected role, the PL features of M6-Eu-
mal tend to indicate that the malonamide fragment is not directly
coordinated to the Eu3+ ions, although it may figure in the second
coordination shell.
The photograph of the F6-Eu film under UV light at 365 nm
(Figure 6A) shows the emitted intra-4f6 red light to be guided to
the film edges through internal reflection, revealing that the F6-
Eu film can be used as LSC. To quantify the emission ratio factor
C, we performed an intensity map of the red pixel (in the RGB
color model) in nonsaturated photographs (not shown). Figure 6B
displays one example of such three-dimensional (3D) intensity
maps. In order to separate the contribution of the thin film emission
and that of the scattered light at the surface, part of the film was
removed from the substrate and the red pixel intensity at the film
and at the substrate was estimated. This experiment enables a
quantification of the scattered light of ∼20% at 350 nm. Therefore,
the average intensity ratio between corrected Isf and Ie yields C = 6.
This value is consistent with that reported for LSCs based on
poly(methyl methacrylate) encapsulating organic dyes and quan-
tum dots (QDs), with C values of 5ꢀ10.12
In addition, the F6-Eu thin films were characterized by spec-
troscopic ellipsometry. The film thickness and the refractive
index dispersion curve were measured, considering a layered
structure model that was composed of the substrate, an organicꢀ
inorganic hybrid layer, and air as ambient medium (see Figure
S18A in the Supporting Information). The thickness of the
substrate was considered infinite, and the refractive index was
obtained by direct inversion of the ellipsometric data (not
shown). The experimental ellipsometric parameters and the
respective fit using the Lorentz model are shown in Figure
S18B in the Supporting Information. For an illustrative example,
the best-fit solution yielded a thickness value of d = 53.84 (
1.57 nm and the refractive index variation represented in Figure 8.
The presence of surface roughness was modeled by the Brugge-
mann effective medium approximation,52 using the thickness
(average roughness) of a layer composed of air and hybrid
material. The improvement in the goodness of fit is <5%, yielding
roughness values of ∼0.1 nm, pointing out that surface irregula-
rities can be neglected in the present case, despite the presence of
small cubic nanoparticles shown in the SEM image (see Figure
S19 in the Supporting Information) of the surface of the hybrid.
The potential effect of the cubic nanometric particles and
roughness in the ellipsometric spectra is depolarization of the
reflected light, because of the scattering of the incident light by
these surface features, which originates light reflected in more
than one polarization states.53 The measured depolarization
spectrum shows a very small depolarization effect with a polar-
ization degree very close to the unit (see inset in Figure 8). The
poor effect of the surface nanoparticles in the ellipsometric
analysis can be explained by two main reasons: (i) the surface
The performance of the F6-Eu thin film as a LSC is given by
the optical conversion efficiency ηopt of the collector, which is
defined as8,11,13
ηopt ¼ ð1 ꢀ RÞηabsηf ηsηtηtrηself
ð3Þ
where R is the reflectance, ηabs is the ratio of photons absorbed by
the plate to the number of photons falling on the plate, ηf is the
emission quantum yield, ηs is the Stokes efficiency (ratio of the
average energy of emitted photons to the average energy of the
absorbed ones), ηt is the trapping efficiency, ηt = (1 ꢀ 1/n2)1/2
(where n is the refractive index of the light-emitting medium), ηtr
is the transport efficiency (which takes into account the transport
losses due to matrix absorption and scattering), and ηself is the
self-absorption efficiency arising from self-absorption of the emitters.
At 350 nm, the ηopt value of F6-Eu is initially estimated as ∼4.3%,
considering R = 0.067 (with the refractive index value at 350 nm,
n = 1.7006, extrapolated from the dispersion curve of Figure 8),
ηabs = (1 ꢀ 10ꢀA) = 0.30 (where A is the absorbance (0.153; see
Figure 8)), ηf = 0.34, ηs = 0.572 (the emitted photons at 612 nm
dominate), ηt = 0.786 (n = 1.6161 at 612 nm), and ηtr= ηself = 1,
4780
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
because the hybrid matrix is almost transparent to UVꢀvis light
and Eu3+ ions have negligible self-absorption (no losses). The
optical conversion efficiency of F6-Eu is smaller than that
recorded for LSCs formed by a PMMA layer containing Red
305 dyes on top of a transparent glass support (∼12%) and the
upper estimate reported for films of Tb3+-based poly(vinyl
alcohol) incorporating salicylic acid (8.8%),8 the only value that
we found in the literature for LSCs based on Ln ions. However, in
this latter example, ηabs is assumed to be equal to 1 (3-fold larger
than that in F6-Eu). We certainly can improve the optical
conversion efficiency of the F6-Eu film, by increasing its thick-
ness and, consequently, its absorbance.
these materials, particularly the energetic position of the 5D0f7F0
transition, evidence a significantly different complexation mode
of the Eu3+ ions for M6-Eu, compared to M4-Eu and M5-Eu.
Moreover, M6-Eu exhibits the highest absolute emission quan-
tum yield value (0.18 ( 0.02) among these three materials. This
value can also be increased (0.26 ( 0.03) upon the addition of a
malonamide derivative that promotes light harvesting without
modification of the emission and excitation spectra. In addition,
M6-Eu can be processed as thin films by spin-coating on glass
substrates, leading to plates that are covered by a thin layer
(∼54 nm) consisting of luminescent bridged silsesquioxanes
exhibiting an emission quantum yield (0.34 ( 0.03) that is
among the highest values reported for films of Eu3+-containing
hybrids and that may be used as luminescent solar concentrators
(LSCs) with an optical conversion efficiency of ∼4%. The emission
ratio factor (C) was quantified through the mapping of the intensity
of the red pixels, which allowed for a more consistent estimation for
the transport losses, because of the scattering of the emitted light in
the film and substrate (0.40), and then for the effective optical
conversion efficiency (1.7%). This is a very rare example of the use
of Ln3+-containing hybrid thin film for LSCs. Moreover, the optical
conversion efficiency of the F6-Eu films can be optimized increasing
the Eu3+ concentration, the film thickness, and C, which opens the
possibility of increasing the photocurrent generated by solar cells
coupled to the F6-Eu LSC edges, relative to the same cells facing
the sun.
The emission ratio factor (C) could be also defined as
!
ꢀ
ꢁ
ηopt
Asf
Asf
2ηt
C ¼
¼
ð4Þ
ηsf =2 Ae
1 ꢀ ηt Ae
where
ηsf ¼ ð1 ꢀ RÞηabsηf ηsηtð1 ꢀ ηtrÞηself
ð5Þ
is the conversion efficiency of the signal emitted at the surface of the
film (in which the trapping efficiency is replaced by its complemen-
tary value, (1 ꢀ ηt)); Asf and Ae are the surface area and the area of
the plate edges, respectively; and the factor 1/2 takes into account
that we are considering the emission trapping in only one film
surface. For F6-Eu, the ratio between Asf and Ae, known as the LSC
geometrical factor, is 7.7. From eq 4, an emission ratio factor of 57 is
estimated, which is substantially higher than the experimental C
value (6) that was derived from the intensity map (Figure 7B).
To interpret such a discrepancy, we may consider that the signal
trapped in the waveguide will lose part of its intensity due to
scattering effects along the propagation in the film (being the
scattered signal emitted by the surface), in a similar way that was
performed in the estimation of the losses incurred by self-absorption
in LSCs of liquid solutions of PbS QDs.54 Considering a thin layer
(as is the casefor theF6-Eu film, in which a signal that is not trapped
does not undergo any scattering process), we can rewrite the top
surface optical conversion efficiency ηsf as
’ ASSOCIATED CONTENT
S
Supporting Information. Figure S1, Fourier transform
b
infrared (FTIR) spectroscopy data; Figure S2, X-ray diffraction
(XRD) patterns; Figures S3 and S11, optical micrographs;
Figures S4ꢀS6, S9, S13, and S14, emission spectra; Figures S7,
S8, S10, and S15ꢀS17, 5D0 emission decay curves; Figure S12,
excitation spectra; Figure S18, refractive index dispersion curve
and experimental ellipsometric parameters; and Figure S19, SEM
image. This information is available free of charge via the Internet
at http://pubs.acs.org.
1
’ AUTHOR INFORMATION
ηsf
¼
ð1 ꢀ RÞηabsηf ηs½ð1 ꢀ ηtÞ þ ð1 ꢀ ηtrÞꢂ
ð6Þ
2
Corresponding Author
*E-mail: lcarlos@ua.pt michel.wong-chi-man@enscm.fr.
Inserting eqs 6 and 3 into eq 4, we obtain the effective emission ratio
factor (Ceff):
ꢂ
ꢃ
Asf
2ηtrηt
Ceff
¼
ð7Þ
’ ACKNOWLEDGMENT
ð1 ꢀ ηtÞ þ ð1 ꢀ ηtrÞ Ae
We acknowledge the financial support of Fundac-~ao para a
Ci^encia e a Tecnologia (FCT, Portugal), COMPETE, andFEDER
programs (PTDC/CTM/101324/2008). The authors thank
D. Cot (IEM Montpellier), P. Dieudonnꢀe (L2C Montpellier),
and S. S. Nobre and P. P. Lima (Universityof Aveiro) for theircon-
tribution in the SEM, PXRD, and photoluminescence measure-
ments, respectively. J.G. gratefully acknowledges the “Ministꢁere de
l’Enseignement Supꢀerieur et de la Recherche” and EMMI for their
financial support.
To attain an emission ratio factor of 6, the scattering factor in
eq 7 (ηtr) is 0.40 (instead of 1, as initially considered), which
yields an effective optical conversion efficiency of 1.7%.
3. CONCLUSIONS
In this work, the syntheses of two new urea-bipyridine-derived
bridged organosilanes (P5 and P6) have been achieved. Their
hydrolysisꢀcondensation under nucleophilic catalysis in the
presence of Eu3+ salts led to luminescent bridged silsesquioxanes
(M5-Eu and M6-Eu), which were compared with M4-Eu. The
maximum value of incorporated Eu3+ in the materials strongly
depends on the isomer used. A value of 11%w can even be reached
without phase segregation for M60-Eu as the 6,60-isomer is
strongly basic and is expected to form a cagelike complex involving
urea and bipyridine fragments. Indeed, the photoluminescence of
’ REFERENCES
(1) Escribano, P.; Juliꢀan-Lꢀopez, B.; Planelles-Aragꢀo, J.; Cordoncillo,
E.; Viana, B.; Sanchez, C. J. Mater. Chem. 2008, 18, 23.
(2) Binnemans, K. Chem. Rev. 2009, 109, 4283.
(3) Carlos, L. D.; Ferreira, R. A. S.; de Zea Bermudez, V.; Ribeiro,
S. J. L. Adv. Mater. 2009, 21, 509.
4781
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782

Chemistry of Materials
ARTICLE
(4) Carlos, L. D.; Ferreira, R. A. S.; deZea Bermudez, V.; Juliꢀan-Lꢀopez,
B.; Escribano, P. Chem. Soc. Rev. 2011, 40, 536.
(39) Fox, M. Optical Properties of Solids; Oxford University Press:
New York, 2002.
(5) Jin, T.; Inoue, S.; Machida, K.; Adachi, G. J. Electrochem. Soc.
1997, 144.
(40) Fernandes, V. R.; Vicente, C. M. S.; Wada, N.; Andrꢀe, P. S.;
Ferreira, R. A. S. Opt. Express 2010, 18.
(6) Wang, H. Q.; Batentschuk, M.; Osvet, A.; Pinna, L.; Brabec, C. J.
Adv. Mater. 2011, 23.
(7) Machida, K.; Li, H.; Ueda, D.; Inoue, S.; Adachi, G. J. Lumin.
(41) Nobre, S. S.; Brites, C. D. S.; Ferreira, R. A. S.; de Zea
Bermudez, V.; Carcel, C.; Moreau, J. J. E.; Rocha, J.; Wong Chi Man,
M.; Carlos, L. D. J. Mater. Chem. 2008, 18, 4172.
2000, 87ꢀ9, 1257.
(42) Nobre, S. S.; Catto€en, X.; Ferreira, R. A. S.; Wong Chi Man, M.;
Carlos, L. D. Phys. Status SolidiꢀRRL. 2010, 4, 55.
(43) Carlos, L. D.; Ferreira, R. A. S.; de Zea Bermudez, V.; Ribeiro,
S. J. L. Adv. Funct. Mater. 2001, 11.
(44) Nobre, S. S.; Lima, P. P.; Mafra, L.; Ferreira, R. A. S.; Freire,
R. O.; Fu, L. S.; Pischel, U.; de Zea Bermudez, V.; Malta, O. L.; Carlos,
L. D. J. Phys. Chem. C 2007, 111, 3275.
(8) Misra, V.; Mishra, H. J. Chem. Phys. 2008, 128, 244701.
(9) Moudam, O.; Rowan, B. C.; Alamiry, M.; Richardson, P.;
Richards, B. S.; Jones, A. C.; Robertson, N. Chem. Commun. 2009, 6649.
(10) Wilson, L. R.; Rowan, B. C.; Robertson, N.; Moudam, O.;
Jones, A. C.; Richards, B. S. Appl. Opt. 2010, 49, 1651.
(11) Reisfeld, R. Opt. Mater. 2010, 32, 850.
(12) van Sark, W. G. J. H. M.; Barnham, K. W. J.; Slooff, L. H.;
Chatten, A. J.; Buchtemann, A.; Meyer, A.; McCormack, S. J.; Koole, R.;
Farrell, D. J.; Bose, R.; Bende, E. E.; Burgers, A. R.; Budel, T.; Quilitz, J.;
Kennedy, M.; Meyer, T.; Donega, C. D. M.; Meijerink, A.; Vanmaekelbergh,
D. Opt. Express 2008, 16, 21773.
(45) Carlos, L. D.; Ferreira, R. A. S.; de Zea Bermudez, V.; Molina,
C.; Bueno, L. A.; Ribeiro, S. J. L. Phys. Rev. B 1999, 60, 10042.
(46) Carlos, L. D.; Malta, O. L.; Albuquerque, R. Q. Chem. Phys. Lett.
2005, 416, 238.
(47) Malta, O. L.; Batista, H. J.; Carlos, L. D. Chem. Phys. 2002, 282.
(48) Thuꢀery, P.; Nierlich, M.; Charbonnel, M. C.; Den Auwer, C.;
Dognon, J. P. Polyhedron 1999, 18, 3599.
(13) Dienel, T.; Bauer, C.; Dolamic, I.; Bruhwiler, D. Sol. Energy
2010, 84, 1366.
(14) Calzaferri, G. Top. Catal. 2010, 53, 130.
(49) Dong, D. W.; Yang, Y. S.; Jiang, B. Z. Mater. Chem. Phys. 2006,
(15) Zhao, D.; Seo, S. J.; Bae, B. S. Adv. Mater. 2007, 19, 3473.
(16) Wang, Y.; Li, H. R.; Feng, Y.; Zhang, H. J.; Calzaferri, G.; Ren,
T. Z. Angew. Chem., Int. Ed. 2010, 49, 1434.
95, 89.
(50) Wang, Y.;Wang, Y. G.;Cao, P. P.;Li, Y. N.;Li, H. R.CrystEngComm
2011, 13, 177.
(17) Dong, D. W.; Jiang, S. C.; Men, Y. F.; Ji, X. L.; Jiang, B. Z. Adv.
Mater. 2000, 12, 646.
(18) Pecoraro, E.; Ferreira, R. A. S.; Molina, C.; Ribeiro, S. J. L.;
Messsaddeq, Y.; Carlos, L. D. J. Alloys Compd. 2008, 451, 136.
(19) Chamas, Z. E.; Guo, X. M.; Canet, J. L.; Gautier, A.; Boyer, D.;
Mahiou, R. Dalton Trans. 2010, 39, 7091.
(51) Wang, Y.; Wang, Y.; Gan, Q. SolꢀGel Sci. Technol. 2010, 56, 141.
(52) Azzam, R. M. A.; Bashara, N. M., Ellipsometry and Polarized
Light; Elsevier North-Holland: Amsterdam, New York, 1977.
(53) Ramsey, D. A.; Ludema, K. C. Rev. Sci. Instrum. 1994, 65, 2874.
(54) Shcherbatyuk, G. V.; Inman, R. H.; Wang, C.; Winston, R.;
Ghosh, S. Appl. Phys. Lett. 2010, 96.
(20) Lenaerts, P.; Storms, A.; Mullens, J.;D'Haen, J.; Gorller-Walrand,
C.; Binnemans, K.; Driesen, K. Chem. Mater. 2005, 17, 5194.
(21) Li, Y. H.; Zhang, H. J.; Wang, S. B.; Meng, Q. G.; Li, H. R.;
Chuai, X. H. Thin Solid Films 2001, 385, 205.
(22) Li, H. R.; Fu, L. S.; Lin, J.; Zhang, H. J. Thin Solid Films 2002,
416, 197.
(23) Yan, B.; Sui, Y. L.; Liu, J. L. J. Alloys Compd. 2009, 476, 826.
(24) He, H. S.; Dubey, M.; Sykes, A. G.; May, P. S. Dalton Trans.
2010, 39, 6466.
(25) Sui, Y. L.; Yan, B. J. Photochem. Photobiol. A 2006, 182, 1.
(26) Armelao, L.; Bottaro, G.; Quici, S.; Scalera, C.; Cavazzini, M.;
Accorsi, G.; Bolognesi, M. ChemPhysChem 2010, 11, 2499.
(27) Oh, E. O.; Kim, Y. H.; Whang, C. M. J. Electroceram. 2006,
17, 335.
(28) Benyahya, S.; Monnier, F.; Taillefer, M.; Wong Chi Man, M.;
Bied, C.; Ouazzani, F. Adv. Synth. Catal. 2008, 350, 2205.
(29) Li, H. R.; Lin, N. N.; Wang, Y. G.; Feng, Y.; Gan, Q. Y.; Zhang,
H. J.; Dong, Q. L.; Chen, Y. H. Eur. J. Inorg. Chem. 2009, 519.
(30) Feng, Y.; Li, H. R.; Gan, Q. Y.; Wang, Y. G.; Liu, B. Y.; Zhang,
H. J. J. Mater. Chem. 2010, 20, 972.
(31) Nobre, S. S.; Ferreira, R. A. S.; Catto€en, X.; Benyahya, S.;
Taillefer, M.; de Zea Bermudez, V.; Wong Chi Man, M.; Carlos, L. D.
Spectrosc. Lett. 2010, 43, 321.
(32) Franville, A. C.; Zambon, D.; Mahiou, R.; Troin, Y. Chem.
Mater. 2000, 12, 428.
(33) Caix-Cecillon, C.; Chardon-Noblat, S.; Dꢀeronzier, A.; Haukka,
M.; Pakkanen, T. A.; Ziessel, R.; Zsoldos, D. J. Electroanal. Chem. 1999,
466, 187.
(34) Thibault, M. E.; Luska, K. L.; Schlaf, M. Synthesis 2007, 791.
(35) Broudic, J. C.; Conocar, O.; Moreau, J. J. E.; Meyer, D.; Wong
Chi Man, M. J. Mater. Chem. 1999, 9, 2283.
(36) Bourg, S.; Broudic, J. C.; Conocar, O.; Moreau, J. J. E.; Meyer,
D.; Wong Chi Man, M. Chem. Mater. 2001, 13, 491.
(37) Janiak, C.; Deblon, S.; Wu, H. P.; Kolm, M. J.; Kl€ufers, P.;
Piotrowski, H.; Mayer, P. Eur. J. Inorg. Chem. 1999, 1507.
(38) Albrecht, M.; Janser, I.; L€utzen, A.; Hapke, M.; Fr€ohlich, R.;
Weis, P. Chem. Eur. J. 2005, 11, 5742.
4782
dx.doi.org/10.1021/cm2019026 |Chem. Mater. 2011, 23, 4773–4782