Coupling of liquid crystals to silica nanoparticles

Article abstract of DOI:10.1021/ja0735608

Liquid crystal (LC) molecules were functionalized with alkoxysilane groups and grafted on silica nanoparticles by the sol-gel process, requiring condensation of hydrolyzed molecules with silanol groups at the surface of molecules. The grafting ratio was precisely quantified from a very simple procedure using in situ recording of ¹ H NMR spectra. We showed that LC-grafted nanoparticles can be dispersed at high concentration in organic solvents such as pure LC. They were used as nucleation centers to prepare polymer-dispersed liquid crystal (PDLC) samples with nanometric LC droplets and a complete phase separation between LC and polymer. Copyright

Full text of DOI:10.1021/ja0735608

Published on Web 07/11/2007
Coupling of Liquid Crystals to Silica Nanoparticles
Vincent Rachet, Khalid Lahlil, Mathieu Be´rard, Thierry Gacoin, and Jean-Pierre Boilot*
Laboratoire de Physique de la Matie`re Condense´e, Ecole Polytechnique, CNRS, 91128 Palaiseau, France
Received May 18, 2007; E-mail: jean-pierre.boilot@polytechnique.fr
Chart 1. Structures of LC Molecules (n ) 4-9) for Grafting on
Silica Particles
Electro-optical properties of polymer-dispersed liquid crystal
(PDLC)1,2 or holographic polymer-dispersed liquid crystal (H-
PDLC)3-7 are strongly related to optical index contrast between
polymer matrix and electro-optical nematic phase. In PDLC,
orientation of liquid crystal clusters located in droplets is controlled
by an external electric field which allows matching the liquid crystal
(LC) and polymer refractive indexes. When LC droplet diameters
are in a range of a micrometer, devices switch from light scattering
to transparent states providing indices are matched. Reducing the
size of liquid crystal droplets to the sub-wavelength scale leads to
nonscattering films, with a remaining electro-optic coefficient and
a polarization insensitivity for a normally incident wave.
These kinds of biphasic structures are usually obtained by
photoinduced phase separation (PIPS). Starting from a homogeneous
mixture of monomer and LC molecules, a curing process induces
a phase separation and formation of LC droplets. It is commonly
accepted that the phase separation is usually not complete and that
a part of LC molecules remains trapped in the polymer matrix.8,9
This kinetics effect is largely amplified for nanometric droplets
(50-100 nm) which only contain a small LC volume fraction,
increasing the driving voltage for PDLC, decreasing the spatial
modulation of the optical index, and leading to low diffraction
efficiency for H-PDLC recordings. The partial phase separation and
resulting morphologies have been recently explained using a general
Onsager theory of transport, highlighted the role of the initial
nucleation phase, and suggested the use of nucleation centers to
control the LC volume fraction in droplets.10
The same method of characterization was performed to follow
the covalent grafting reaction of L2 molecules at the surface of
silica nanoparticles. A commercial suspension of silica nanoparticles
in methanol was used (Nissan, MA-ST type, 14 nm in diameter).
After solvent transfer and evaporation of methanol, 162.5 mg of
silica in 1 ml of DMF were mixed to different quantities (from 18
to 64 mg) of L2 in a sealed NMR tube. The mixture was refluxed
for a few hundred of hours and 1H NMR spectra were periodically
recorded. As observed for pure L2 molecules, the concentrations
of starting monomer, hydrolyzed monomer, and dimer compounds
1
can be easily deduced from H NMR liquid spectra in the 0.40-
0.50 ppm chemical shift range (Figure 1, left). When time is
increased, the total concentration of these molecular species
progressively decreases to a plateau, corresponding to the progres-
sive grafting of L2 molecules at the surface of silica nanoparticles.
During 1H liquid NMR sequences, protons belonging to L2
molecules coupled to nanoparticles do not have sufficient time to
relax and are not detected. Consequently, the study of the decrease
of NMR signals assigned to molecular species allows calculation
of the fraction of grafted L2 molecules on silica particles, resulting
from condensation reactions between hydrolyzed monomers and
silanol groups located at the surface of particles.
In this paper, specific centers are prepared from the covalent
grafting of cyano-biphenyl LC molecules at the surface of silica
nanoparticles, allowing perfect stability of these particles into the
1
nematic phase. The grafting rate is precisely determined by H
NMR, with a simple and original procedure used for the first time
to quantify the functionalization of inorganic oxide particles. Finally,
in a preliminary study, LC-functionalized silica nanoparticles are
used as nucleation centers in a PIPS process with thiol-ene
monomers and cyano-biphenyl LC molecules. A large enhancement
of the phase separation is then observed from SEM observations
with a complete demixtion where all the initial LC molecules go
within the LC droplet volume. This should obviously increase the
refractive index spatial modulation and improve electro-optic
properties, especially the response time, in nanosized PDLC devices.
We designed and synthesized vinyl-substituted L1 and ethoxy-
dimethylsilane-substituted L2 LC molecules as shown in Chart 1.
In preliminary experiments, acid-catalyzed hydrolysis and conden-
Figure 1 (right) shows the time evolution of the different L2-
type species by using 57 mg (125.6 µmol) of LC molecules. After
introduction of silica particles and as shown by recording the first
NMR spectrum, monomers are rapidly hydrolyzed and it then
appeared to be a competition between the dimerization of hydro-
lyzed monomers and their condensation with silanol groups at the
1
sation reactions of L2 were followed in THF by H NMR, using
an internal reference constituted of a capillary tube filled of a
p-xylene solution in CDCl3. It was observed that the time evolution
of monomer and dimer species can be easily quantified from
chemical shift and intensity changes of the NMR peaks assigned
to methyl groups coupled to the silicon atom. The hydrolysis rate
was also deduced from the integration of the methylene triplet signal
due to the ethanol formation.
Figure 1. (Left) Typical 1H liquid NMR spectrum observed during the
coupling of L2 LC molecules to silica nanoparticles in DMF. (Right) Time
evolution of L2 LC species. The concentration of LC molecules grafted on
nanoparticles (G) was deduced from the decrease of the total intensity of
1H liquid NMR peaks assigned to molecular monomers (M), hydrolyzed
monomers (H), and dimer (D) species.
9
9274
J. AM. CHEM. SOC. 2007, 129, 9274-9275
10.1021/ja0735608 CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
surface of particles. In these experiments, the condensation kinetics
of LC molecules appears to be faster at the surface of particles. A
simple simulation using a kinetics model previously developed by
Pouxviel et al.11 for hydrolysis-condensation reactions of sol-gel
precursors indicates that the rate constant is about 5 times higher
than that for the self-condensation of hydrolyzed L2 monomers.
In fact, L2 monomers are preferentially hydrolyzed by confined
water molecules located near the surface of hydrophilic silica
particles, allowing rapid condensation of hydrolyzed monomers with
surface silanol groups. After about 60 h in our experimental
conditions, the grafting on particles is stopped while both hydrolysis
and self-condensation reactions of L2 monomers go on. This
indicates a saturation of the surface of particles, and this is a clear
check that the starting LC monomers are used in excess in these
experiments. The saturation effect was also highlighted by observing
the same plateau when increasing the initial concentration of L2
monomers. The concentration of grafted silanol sites at the surface
of nanoparticles can be easily deduced from these NMR experiments
with an accurate precision of about 5%. The value is 0.38 ( 0.01
mmol‚g-1. Taking into account both experimental measurement of
the density of silica nanoparticles (1.88 g‚cm-3) and of the average
size of nanoparticles (14 ( 4 nm) deduced from microscopy
observations, the grafting rate is ∼1 LC molecule/nm2 at the surface
of nanoparticles. Assuming the presence of 4-6 OH/nm2 at the
surface of silica particles, as previously deduced from ignition
measurements,12 shows that, due to steric effects, only about 20%
of the initial silanol groups are grafted by LC molecules. This is in
agreement with previous values determined by other techniques
(29Si NMR and thermal analysis) for grafting experiments performed
on the same type of silica particles.13
Figure 2. SEM images of typical PDLC samples obtained after curing a
mixture of thiol-ene monomers and nematic LCs. (a) PDLC from a standard
mixture without nanoparticles; (b) PDLC from a mixture containing LC-
grafted silica nanoparticles as nucleation centers.
transmission. While the nondoped sample is quite transparent in
the visible range, the doped one is white, indicating large diffusion
efficiency in spite of a lower nanometric size of droplets (Figure
S2). This suggests that light diffusion in the doped sample is
drastically increased by a larger optical index mismatch.
These results clearly demonstrate that LC-grafted nanoparticles
act as nucleation centers during the PIPS process leading to a
complete phase separation and a premature stop of the growing
process, which decreases the size of droplets. By reaching for the
first time a complete separation between liquid crystal and polymer,
LC-grafted nanoparticles offer new opportunities to the PDLC
community, allowing better diffraction efficiency with thinner films,
and then can advantageously decrease addressing voltage and
response time of future H-PDLC components. Beyond the PDLC
phase separation control, the ability to disperse silica particles at
high concentrations in pure LC can lead to verification of theoretical
predictions14 and allow new devices to be created.
After elimination of remaining molecular species (see Supporting
Information), the grafted particles can be dispersed at high con-
centration in organic solvents (20-25 wt% in dichloromethane and
in the pure cyano-biphenyl LC), while starting silica nanoparticles
are only soluble in polar solvents.
Acknowledgment. We are grateful to P. Le Barny and P.
Feneyrou (Thales Research Technology) for fruitful discussions.
Supporting Information Available: Experimental details for the
synthesis of L1 and L2 LC molecules, characterization of silica
nanoparticles. This material is available free of charge via the Internet
at http://pubs.acs.org.
LC-grafted nanoparticles were tested in PIPS experiments to
prepare PDLC with submicrometer sized droplets. PDLC samples
were prepared from a standard thiol-ene polymerizable mixture
belonging to the Norland series (NOA81, 70 wt %) with a typical
nematic LC (BL24 from Merck, 30 wt %). Note that the densities
of the monomer and LC are equal, allowing volume and weight
fractions to be assimilated. Curing process was performed in cells
(20 µm in thickness) under UV light (150 mW‚cm-2) for 4 min.
One set of samples was initially doped with about 5 wt % of LC-
grafted nanoparticles. After curing, cells were opened and metallized
with a Au/Pd alloy for SEM observations. Assuming spherical
droplets, the 3D size distribution can be computed from 2D SEM
pictures (Figure 2), leading to the whole volume occupied by LC
droplets. For samples without particles (Figure 2a), droplets (average
size of 420 nm) only fill 14.8% of the total volume. This means
that about 50% of the initial LC molecules are trapped into the
polymer walls corresponding to a partial phase separation. Concern-
ing the sample containing LC-grafted nanoparticles (Figure 2b),
the microstructure is very different with a larger number of droplets
of smaller sizes (average of 170 nm). In this case, the droplet
volume is 28%, showing that the phase separation is complete with
about all of the LC molecules in droplets. Another significant
difference between the two types of samples is given by the optical
References
(1) Bouteiller, L.; Le Barny, P. Liq. Cryst. 1996, 21, 157-174.
(2) Bouteiller, L.; Le Barny, P.; Martinot-Lagarde, Ph. Liq. Cryst. 1994, 17,
709-716.
(3) Qi, J.; De Sarkar, M.; Warren, G. T.; Crawford, G. J. Appl. Phys. 2002,
91, 4795-4800.
(4) Suntherland, R. L.; Tondiglia, V. P.; Natarajan, L. V.; Bunning, T. J.
Appl. Phys. Lett. 2001, 79, 1420-1422.
(5) Suntherland, R. L. J. Opt. Soc. Am. 2002, 19, 2995-3012.
(6) Suntherland, R. L.; Tondiglia, V. P.; Natarajan, L. V.; Chandra, S.; Tomlin,
D.; Bunning, T. J. Opt. Express 2002, 10, 1074-1082.
(7) Divlianski, I. S.; Mayer, T. S.; Holliday, K. S.; Crespi, V. H. Appl. Phys.
Lett. 2003, 82, 1667-1669.
(8) Bunning, T. J.; Natarajan, L. V.; Tondiglia, V. P.; Sutherland, R. L.; Vezie,
D. L.; Adams, W. W. Polymer 1995, 36, 2699-2708; 1996, 37, 3147-
3150.
(9) Rachet, V.; Feneyrou, P.; Loiseaux, B.; Machke, U.; Le Barny, P.;
Huignard, J. P. Mol. Cryst. Liq. Cryst. 2004, 421, 165-174.
(10) Rachet, V.; Feneyrou, P.; Le Barny, P.; Loiseaux, B.; Huignard, J. P.
Proc. SPIE 2004, 5518, 211-219.
(11) Pouxviel, J. C.; Boilot, J. P.; Beloeil, J. C.; Lallemand, J. Y. J. Non-
Cryst. Solids 1987, 89, 345-350.
(12) Iller, R. K. The Chemistry of Silica; Wiley-Interscience: New York, 1979.
(13) Douce, J.; Boilot, J. P.; Biteau, J.; Scodellaro, L.; Jimenez, A. Thin Solid
Films 2004, 1-2, 114-122.
(14) (a) Poulin, P. Science 1997, 275, 1770-1773. (b) Mondain-Monval, O.;
Poulin, P. J. Phys.: Condens. Matter 2004, 16, 1873-1885.
JA0735608
9
J. AM. CHEM. SOC. VOL. 129, NO. 30, 2007 9275