Photoalignment of liquid crystals using a covalently attached self-assembled ultrathin film fabricated from diazoresin and cinnamate polyelectrolyte

Article abstract of DOI:10.1039/b703598e

Layer-by-layer films have been fabricated from diazoresin (DR) and poly[4-sulfonyl phenoxy(4-acryloyloxy)cinnamate, sodium salt] (PACSS) in aqueous solution via electrostatic attraction. The ionic bonds between the diazonium ions and sulfonate ions in the film were converted into covalent bonds by irradiation with 365 nm UV light. It was also confirmed that the vinylene units of the cinnamoyl group in PACSS did not change in this process. Covalently attached self-assembled ultrathin films containing photosensitive cinnamoyl groups could align liquid crystals (LC) uniformly with good stability after irradiation with linearly polarized ultraviolet light (LPUVL). In this work, polarized UV-vis spectroscopy was utilized to investigate the photochemical process of the covalent film. It was found that cinnamoyl moieties parallel to the polarization direction of the LPUVL were consumed by the photoreaction faster than those perpendicular to the polarization direction. It can be concluded that the selective photoreaction induced the anisotropy of the films. Experimental results revealed that the degree of photoreaction and the dichroic ratio of the films depended on the number of adsorbed layers and the irradiation time. These results were correlated with the LC homogeneous alignment behavior in the parallel LC cell. The reorientation behavior of the LC molecules was found to be associated with the covalent film thickness and irradiation time. The Royal Society of Chemistry.

Full text of DOI:10.1039/b703598e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Photoalignment of liquid crystals using a covalently attached self-assembled
ultrathin film fabricated from diazoresin and cinnamate polyelectrolyte
Lingli Zhang,ab Zenghui Peng,a Lishuang Yao,ab Fengzhen Lvab and Li Xuan*a
Received 9th March 2007, Accepted 4th May 2007
First published as an Advance Article on the web 15th May 2007
DOI: 10.1039/b703598e
Layer-by-layer films have been fabricated from diazoresin (DR) and poly[4-sulfonyl phenoxy(4-
acryloyloxy)cinnamate, sodium salt] (PACSS) in aqueous solution via electrostatic attraction. The
ionic bonds between the diazonium ions and sulfonate ions in the film were converted into
covalent bonds by irradiation with 365 nm UV light. It was also confirmed that the vinylene units
of the cinnamoyl group in PACSS did not change in this process. Covalently attached self-
assembled ultrathin films containing photosensitive cinnamoyl groups could align liquid crystals
(LC) uniformly with good stability after irradiation with linearly polarized ultraviolet light
(LPUVL). In this work, polarized UV-vis spectroscopy was utilized to investigate the
photochemical process of the covalent film. It was found that cinnamoyl moieties parallel to the
polarization direction of the LPUVL were consumed by the photoreaction faster than those
perpendicular to the polarization direction. It can be concluded that the selective photoreaction
induced the anisotropy of the films. Experimental results revealed that the degree of
photoreaction and the dichroic ratio of the films depended on the number of adsorbed layers and
the irradiation time. These results were correlated with the LC homogeneous alignment behavior
in the parallel LC cell. The reorientation behavior of the LC molecules was found to be associated
with the covalent film thickness and irradiation time.
are most suitable for durable LC photoalignment films because
Introduction
of the irreversible [2 + 2] photodimerization of cinnamoyl
groups.8 However, the photoreaction ratio of cinnamate
The liquid crystal (LC) alignment layer is one of the most
important technologies in forming high-performance LC
materials is low and the thermal stability of the photoalign-
devices, as it can directly affect the contrast ratio in display
ment film is not strong enough; both will limit the application
applications.1 Till now, the contact method based on surface
of the photoalignment technique in industry. Our previous
rubbing has dominated the mainstream of the LC cell
investigations showed that ordered cinnamate layer-by-layer
preparation process. Drawbacks to the contact method include
(LBL) ultrathin films as photoalignment films can readily
overcome these obstacles.10,11 LBL assembly of oppositely
unstable static electricity, residual stress, and dust pollution.
These drawbacks result in poor yield and reliability in
the micro-fabrication of small panel applications.2,3 Thus,
developing a non-contact technology has been a viable
solution to overcome the drawbacks of the rubbing method.
In the 1970s, the reorientation of a mesophase using
photoresponsive molecules was reported by Sackmann and
Haas et al.4,5 Ichimura et al. revolutionized this approach by
attaching a chromophore to a surface for LC alignment in
1988.6 Since then, the photoalignment technique of LC has
gained great attention as a non-contact alignment method.
Numerous investigations on the photoalignment of LC have
confirmed that the macroscopic orientation of LC molecules
can be controlled by photoinduced microscopic changes in
structure as well as orientation of photoreaction moieties,
including azobenzenes,6,7 cinnamates8 and polyimides.9
Among various photoreactive polymers, cinnamate materials
charged polyelectrolytes is a simple and powerful method for
the construction of self-assembled planar ordered composites
on the nanometer scale.12,13 Since the photosensitive cinna-
mates are oriented in the LBL films, the [2 + 2] photoreaction
can proceed easily. The coulombic interaction between layers
also enhances the film stability compared to the van der Waals
force in spin-coated films. The main advantage of these
cinnamate films over those prepared by spin coating is that
the concentration and layer sequence of aggregates can be
well characterized.14
However, the force between layers in such films is based on
coulombic interactions, which may be insufficient to stabilize
the films in certain cases. Cao and co-workers prepared a
family of stable ultrathin multilayer films from conventional
polymers. They reported the conversion of the ionic linkage
nature of the films from ionic bonds to covalent bonds under
ultraviolet irradiation.15,16
aState Key Laboratory of Applied Optics, Changchun Institute of Optics
Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun,
130033, China. E-mail: cundang@163.com; Fax: +86-431-86176316;
Tel: +86-431-86176316
In this investigation, new covalently attached multilayer
ultrathin films were prepared for photoalignment studies. The
ultrathin films were fabricated using diazoresin (DR) as
polycation and poly[4-sulfonate phenoxy(4-acryloyloxy)cinna-
mate, sodium salt] (PACSS) as polyanion (the chemical
bGraduate School of the Chinese Academy of Sciences, Beijing, 100039,
China
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3015–3022 | 3015

View Article Online
Fig. 1 Chemical structures of diazoresin (DR) and poly[4-sulfonate
phenoxy(4-acryloyloxy)cinnamate, sodium salt] (PACSS).
structures are shown in Fig. 1), by a self-assembly technique.
First, the self-assembled films were irradiated with 365 nm UV
light, which changed the linkage between the layers of the films
from ionic to covalent. Then, the covalent self-assembled
films were irradiated with 297 nm linearly polarized UV light
(LPUVL) to form anisotropic surfaces because of the axially
selective photoreaction of cinnamates in the films. The product
films showed excellent LC alignment performance. The stabi-
lities of the films were also enhanced by the linkage conversion
of the DR/PACSS layers from ionic to covalent bonds. The
films were also thermally stable to 180 uC, which was 80 uC
higher than conventional LBL photoalignment films.
Fig. 2 Synthetic route to poly[4-sulfonate phenoxy(4-acryloyloxy)
cinnamate, sodium salt].
1
134.72, 143.74, 152.38, 168.34, 171.62. H NMR (DMSO-d6):
d 6.17 (d, 1H, CH2LCH), 6.43 (m, 1H, CHLCH2), 6.52 (d, 1H,
CH2LCH), 6.6 (d, 1H, CHLCH), 7.23 (d, 2H, ArH), 7.6 (d, 1H,
CHLCH), 7.78 (d, 2H, ArH). IR (KBr powder): 979 (LCH),
1169 (C–O), 1688 (CLO), 1743 (C–O) cm21. m/z (ESI) 217
([M 2 H]2, 100%), 173.1 (17), 119.2 (66) and 71.0 (8). lmax
(ethanol)/nm 282 (e/dm3 mol21 cm21 25320), 217 (19080).
Elemental analysis, Calculated for C12H10O4: C, 66.06; H,
4.59. Found: C, 65.96; H, 4.37%.
Experimental
Synthesis
Diazoresin (DR) was synthesized from diphenylamine-diazo-
nium salt and formaldehyde in concentrated sulfuric acid,
Mn # 2000 g mol21, gsp/c = 0.12 g dL21 17
.
4-Hydroxy-
cinnamic acid and 4-hydroxybenzenesulfonic acid sodium
salt were supplied by Aldrich and used without further
purification. The typical procedure to prepare poly[4-sulfonate
phenoxy(4-acryloyloxy)cinnamate, sodium salt] (PACSS) is
shown in Fig. 2.
4-Sulfonate phenoxy(4-acryloyloxy)cinnamate sodium salt
(2). 4-Acryloyloxycinnamic acid (1) (8.72 g, 40 mmol) and
10 mL thionyl chloride (freshly distilled) were heated under
reflux in 50 mL toluene for 4 h with the addition of a few drops
of DMF as catalyst. The resultant solution was cooled and
needle-like crystals of 4-acryloyloxycinnamic acid chloride
were produced. After filtering, the crude product was
recrystallized using dry toluene. In a 250 mL flask, 4-hydroxy-
benzenesulfonic acid sodium salt (1.96 g, 10 mmol) and NaOH
(0.8 g, 20 mmol) were dissolved in 100 mL of water and 50 mL
THF, and cooled to 0–5 uC. Then, 4-acryloyloxycinnamic
acid chloride (4.73 g, 20 mmol) in 20 mL of THF was added
dropwise with stirring at the same temperature. After stirring
at room temperature for 24 h, some of the solvent was
removed by vacuum distillation. The precipitated was filtered
off, washed with water and recrystalllized from THF (1.58 g,
4 mmol, 10% yield). Mp 243 uC. 13C NMR (DMSO-d6): d
118.02, 121.88, 127.75, 129.55, 130.92, 131.65, 132.71, 146.22,
4-Acryloyloxycinnamic acid (1). To a well stirred mixture
solution of 4-hydroxycinnamic acid (13.12 g, 80 mmol) and
NaOH (6.4 g, 160 mmol) in 40 mL of water and 40 mL
dioxane, acryloyl chloride (9.8 mL, density = 1.11, 120 mmol)
was added dropwise at 5–10 uC. After stirring at room
temperature for 4 h, the reaction mixture was neutralized with
dilute aqueous HCl. The solid 4-acryloyloxycinnamic acid
precipitate was filtered off, washed with warm water, dilute
HCl and water, successively. The crude product was purified
via column chromatography (silica gel, THF) to give a white
solid product (11.3 g, 52 mmol, 65% yield). Mp 186 uC. 13C
NMR (DMSO-d6): d 120.27, 123.12, 128.38, 130.34, 131.26,
3016 | J. Mater. Chem., 2007, 17, 3015–3022
This journal is ß The Royal Society of Chemistry 2007

View Article Online
1
146.67, 153.01, 165.01. H NMR (DMSO-d6): d 6.18 (d, 1H,
of the film was 0.58 mW cm22 (at 297 nm). Linearly polarized
UV-vis absorption spectra were recorded on a Shimadzu UV-
3101PC spectrophotometer equipped with special polarizing
accessories and used to monitor the photoreaction process of
the films. IR spectra were collected on a Bio-Rad FTS3000
spectrophotometer. The surface morphology of the films was
obtained by using a Nanoscope III atomic force microscope
(AFM) (Dimention 3100 s, Digital Instrument Co.).
All measurements were performed in contact mode using
triangular Si3N4 cantilevers in air under ambient conditions.
CH2LCH), 6.46 (m, 1H, CHLCH2), 6.54 (d, 1H, CH2LCH), 6.6
(d, 1H, CHLCH), 7.16 (d, 2H, ArH), 7.23 (d, 2H, ArH), 7.3 (d,
2H, ArH), 7.61 (d, 1H, CHLCH), 7.77 (d, 2H, ArH). IR (KBr
powder): 1050 (SLO), 1688(CLO), 1739 (C–O) cm21. m/z (ESI)
373.1 ([M 2 Na]2, 100%), 293 (69), 226.9 (23), 172.1 (96) and
147.2 (28). lmax (ethanol)/nm 302 (e/dm3 mol21 cm21 26640),
220 (18980). Elemental analysis, calculated for C18H13O7SNa:
C, 54.5; H, 3.28. Found: C, 54.92; H, 3.68%.
Poly[4-sulfonate phenoxy(4-acryloyloxy)cinnamate, sodium
salt] (PACSS) (3). The compound 2 (1 g) was polymerized
in 20 mL DMF with 0.01 g AIBN as the initiator at 65 uC
under nitrogen atmosphere. The polymers were isolated after
polymerization for 12 h by adding the reaction solution to an
excess of methanol, purified by reprecipitation from DMF
solution to methanol, and dried under vacuum (56% yield).
Liquid crystal cells
A pair of substrates with photoirradiated covalently attached
films were assembled together in the parallel direction of
LPUVL irradiation by using 20 mm thick spacers. A com-
mercial LC material TEB30A (Slichem Co., China) was
injected into the cell at 71 uC, which is slightly higher than
the nematic-to-isotropic transition temperature (TNI = 61.2 uC)
of this LC. Then the LC cell was cooled to room temperature
slowly, to remove any flow-induced memory. Polarizing
microscopy (FOIC-2, China) with a digital camera was used
to evaluate the alignment quality of the LC and measure the
transmittance. The polarization directions of the two polar-
izers were crossed.
1
The GPC analysis gave Mn = 1317 and Mw = 1558. H NMR
(DMSO-d6): d 1.10 (d, 2H, CH2–CH), 2.65 (m, 1H, CH–CH2),
6.6 (d, 1H, CHLCH), 7.23 (d, 2H, ArH), 7.58 (d, 1H,
CHLCH), 7.61 (d, 2H, ArH), 7.73 (d, 2H, ArH), 7.89 (d, 2H,
ArH). lmax (ethanol)/nm 302 (e/dm3 mol21 cm21 28200), 220
(17410).
Preparation of covalently attached films
The self-assembled film were constructed on quartz slides for
UV-vis measurement, on silicon wafers for ellipsometry, on
CaF2 wafers for Fourier-transform (FT) IR spectroscopy and
on indium tin oxide (ITO) glass for the assembly of LC cells.
Before the film construction, the quartz wafers and the silicon
wafers were treated in fresh piranha solution (v/v = 1 : 3, 30%
H2O2 : 98% H2SO4) for 1 h, washed carefully with deionized
water and dried. The CaF2 wafers and ITO glass slides were
cleaned by washing with deionized water. The concentrations
of DR and PACSS in deionized water for dip coating were all
0.1 mg mL21. In all cases, DR was the first layer, and PACSS
was the last layer (outermost layer). For (DR/PACSS)n
multilayer films, the deposition time for each layer was
10 min, followed by washing with deionized water and
nitrogen gas blow drying. All fabrication processes were
performed in darkness. DR solution and films were also kept
in the dark before the irradiation. The quantity of material
deposited at each step was deduced from its absorption
spectrum, which was determined on a Shimadzu UV-3101PC
spectrophotometer. The thickness of the multilayer film on a
silicon wafer was determined with a Jobin Yvon UVISEL
ellipsometer (240–830 nm, 70u incident angle).
Results and discussion
Covalently attached self-assembled film formation
The layer-by-layer deposition of the DR/PACSS self-
assembled multilayer film was investigated by UV-vis absorp-
tion spectroscopic measurement (Fig. 3). The characteristic
peaks at 380 nm and 296 nm belong to the diazonium group of
the diazoresin18 and the p–p* transition of the double bond
conjugated with the phenyl group of the cinnamoyl group in
PACSS,19 respectively. The linear increase of the two peaks
with the number of bilayers clearly indicates the regular
growth of the DR/PACSS layers. The LBL film absorbance
was red-shifted 11 nm compared to the 1 mg mL21 PACSS
The well fabricated films were exposed to 365 nm UV light
(irradiation intensity: 0.23 mW cm22 at 365 nm) which was
obtained from a 300 W high pressure Hg–Xe lamp with glass
band-pass filters (365 nm ¡ 5 nm) for a given time to ensure
the photoreaction proceeded to completion.
Photoalignment characterization of covalently attached films
The covalent films were irradiated using a high pressure
Hg–Xe lamp system (300 W) with glass band-pass filters
(297 nm ¡ 5 nm) and a Glan-Talor lens to induce anisotropy
used for LC alignment. The intensity of LPUVL on the surface
Fig. 3 UV-vis absorption spectra of DR/PACSS film on quartz
substrate with increasing number of bilayers. The inset diagram shows
the increase of absorbance at 296 and 380 nm with the number of
bilayers.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3015–3022 | 3017

View Article Online
Fig. 4 The change in thickness as obtained by ellipsometry.
solution (lmax = 285 nm) indicating aggregates in the film. The
red-shift behavior for DR/PACSS shows the formation of
J-aggregates in each layer where the orientation of the
cinnamoyl groups could be described as end-to-end.20,21
Comparison of the lmax centred at about 296 nm values with
increasing number of layers revealed a small shift in lmax
position occurring intermittently. This implies that the degree
of aggregation changes with thickness. The reorganization of
the aggregates with increasing layer thickness is consistent
with a self-healing capacity for these types of films.22,23
Ellipsometric measurement was carried out to provide
information about the film thickness. The average thickness
was 2.3 nm bilayer21 when the concentration of DR (PACSS)
was 0.1 mg mL21. Fig. 4 shows that the film thickness
increased linearly with the number of bilayers. In summary,
the fact that the linear absorbance and thickness increase was
observed for each deposition cycle indicated a quantified
adsorption of polyelectrolyte.
Fig. 6 UV-vis absorption spectra of an 8 bilayer DR/PACSS film
under 365 nm UV irradiation for different times. Irradiation time (s)
(from top to bottom): 0, 10, 30, 60, 120, 240, 600, 900 and 1200.
Irradiation intensity (at 365 nm): 0.23 mW cm22. The inset plot shows
the decrease of absorbance at 380 nm with the irradiation time.
Fig. 5), which improved the stability of the film. Fig. 6 shows
the changes by means of UV-vis spectra of the films after
365 nm UV irradiation. As indicated in the figure, under
365 nm UV irradiation, the diazonium groups decomposed
gradually as indicated by the decrease of the absorbance at
380 nm and concomitant increase of the absorbance at 300 nm.
An isosbestic point at 340 nm appeared in accordance with the
results from previous investigations with DR as a component
in planar films.15,25 The decomposition was complete within
5 min. In order to further confirm this reaction, films of 100
bilayers of DR/PACSS were constructed on CaF2 plates, and
the IR spectra were recorded before and after irradiation as
shown in Fig. 7. As seen in Fig. 7, the asymmetric stretching
The polycation diazoresin and cinnamate polyanion were
linked by electrostatic interaction between the cationic
vibration of the –N2 at 2169 cm21 in the DR/PACSS film
+
2
diazonium group (–N2+) and the anionic –SO3 to form a
disappeared completely after 365 nm UV irradiation for 5 min,
indicating the complete decomposition of the diazonium
group.15 At the same time, the absorbance at 1580 cm21
corresponding to the vibration of the benzene ring conjugated
with an unsaturated group in the DR/PACSS film shifted to
layer-by-layer ultrathin film. The diazoresin was a photo-
sensitive polymer and decomposed readily by UV irradia-
tion;24 the ionic bond between –N2 and O2 could be easily
+
transformed into a covalent bond under UV light (as shown in
Fig. 5 Fabrication of self-assembled DR/PACSS multilayer films and conversion of the interlayer linkage from ionic to covalent under
UV irradiation.
3018 | J. Mater. Chem., 2007, 17, 3015–3022
This journal is ß The Royal Society of Chemistry 2007

View Article Online
direction of the LPUVL. Prior to 297 nm LPUVL irradiation,
the absorption was isotropic (A0 = A = A//). After the
H
irradiation, the intensity of the band at 296 nm decreased with
the irradiation of LPUVL, and a huge variance in the
absorbance was observed between A and A//. The result
H
indicated that the cinnamoyl groups with vinylene units
positioned parallel to the polarization direction of the
LPUVL (A//) were consumed by photoreaction faster than
those perpendicular to the polarization direction (A ) during
H
LPUVL irradiation. As a result of the direction selective
photoreaction, the vinylene CLC bonds of unreacted cinna-
moyl groups aligned perpendicular to the polarization direc-
tion of the LPUVL were left in a greater number relative to
those aligned parallel to the polarization direction. Therefore,
the structural anisotropy of the multilayer film was generated
from the selective photoreaction of the cinnamoyl group in the
film. The LPUVL-induced anisotropic film would give rise to
anisotropic interactions with the LC molecules, which might
possibly generate LC homogeneous alignment.
Fig. 7 IR spectra of a 100 bilayer DR/PACSS film on a CaF2
substrate (a) before and (b) after UV irradiation. Irradiation intensity
(at 365 nm): 0.23 mW cm22. Irradiation time: 5 min.
1600 cm21, which is the normal absorption of the benzene
ring.26 The band at 1633 cm21 due to the vinylene CLC of
the cinnamoyl group stretching vibration had no obvious
change.27 Thus the ionic bond between the diazonium and the
sulfonate group converted to a covalent bond by the first
irradiation, whereas the photosensitive double bond in the
PACSS did not change. Thus, covalently attached self-
assembled films containing photosensitive cinnamoyl groups
have been fabricated.
In order to verify the photoreaction process of the films, the
absorbance at 296 nm was plotted as a function of LPUVL
irradiation time. Fig. 9 shows the changes in absorbance
at 296 nm of DR/PACSS films with respect to LPUVL
irradiation. A0 and At are the absorbance before (0 min) and
after 297 nm LPUVL irradiation (t min) at 296 nm. As shown
in Fig. 9, the intensity of the absorbance at 296 nm decreased
rapidly with the increase of the irradiation time. However, the
change in slope was greater for thicker film. Clearly, there was
a significant effect of the bilayer number on the photoreaction
characteristics. The phenomena could be explained thus:
thicker films have a larger total amount of cinnamoyl groups
and should have more aggregated domains. As mentioned
above, the degree of aggregation changed with the thickness
of the film. The efficiencies of photoreaction were affected by
the micro-environmental properties including polarity, free
volume, molecule-to-molecule interactions, etc. So photoalign-
ment became a function of the total film thickness, where each
layer contributed a finite number of photoalignment domains.
Thickness dependence on photoreaction of LBL films has
also been reported previously for azobenzene materials.14,28
Like photochromic layers, we believe that the differences of
Polarized photoreaction of the covalently attached films
The covalent self-assembled films having one cinnamoyl
moiety per chemical repeat unit in the side chains in the
polyanion were then irradiated by 297 nm LPUVL to induce
anisotropy. The photoreaction procedure in the film was
examined by UV-vis spectroscopy.
Fig. 8 shows the linearly polarized UV-vis absorption
spectra of 3 bilayers DR/PACSS covalently attached film
irradiated with 297 nm LPUVL for 10 min. The polarized
absorbance AH (A//) was measured with a probing UV-vis light
linearly polarized perpendicular (parallel) to the polarization
Fig. 8 Polarized UV-vis absorption spectra of a 3 bilayer covalent
film of DR/PACSS irradiated with 297 nm LPUVL. The plot shows
the spectra before 297 nm LPUVL irradiation, and after irradiation
perpendicular (AH) and parallel (A//) to the polarization direction of
the LPUVL.
Fig. 9 Changes in absorbance at 296 nm of different thickness of
DR/PACSS films as a function of LPUVL irradiation time: 3 (%),
5 ($) and 7 (n) bilayers.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3015–3022 | 3019

View Article Online
Fig. 10 Dichroic ratios measured from different thick DR/PACSS
films irradiated by LPUVL with various time: 3 (%), 5 ($) and 7 (n)
bilayers.
photoreaction with different thicknesses of these self-
assembled films are due to matrix effects.7
Photoinduced optical anisotropy of the film was investigated
by measuring the absorbance at 296 nm parallel (A//) and
perpendicular (A ) to the electric vector of LPUVL. The
H
optical anisotropy is expressed here by the UV dichroic ratio
Fig. 11 Topographic AFM images of films of different thicknesses:
(a) 5 bilayers (3.3 nm rms); (b) 10 bilayers (4.5 nm rms).
[(A 2 A//)/(A + A//)]. The results are shown in Fig. 10. As
H
H
seen in the figure, all the measured dichroic ratios were positive
over the irradiation time range of ¡65 min. The dichroic
ratio reached 0.1, which is larger than that of spin-coated
photoalignment films.29,30 With increasing irradiation time,
the dichroic ratio increased rapidly to a maximum at 10 min
and then decreased. This could be reasonably explained by
assuming a possible mechanism, the preferred depletion of
cinnamic acid side chains parallel to the LPUVL.8 At the
beginning of linear photodimerization, those cinnamoyl
chromophores in the cinnamic acid side chains started to be
cross-linked which were parallel to the polarization direction
of LPUVL. The chromophores with off-axis transition
moments would at first be hardly affected. This led to the
observed initial increase of the dichroic ratio. With the
irradiation time increasing, the chromophore depletion parallel
to the LPUVL saturated and the probability that chromo-
phore pairs with off-axis orientation would undergo cross-
linking increased. Thus further irradiation would lead to the
gradual reduction of the dichroic ratio. From Fig. 10, we also
found that the thicker films had the larger dichroic ratio after
the LPUVL irradiation. This result is consistent with the
photoreaction efficiencies for films with different thicknesses.
As mentioned above, thicker films always have a larger
photoreaction degree for the same irradiation time. It was
already known that the photoreaction primarily induced the
anisotropy of the film. So the thicker film would have the
larger dichroic ratio after the LPUVL irradiation.
homogeneous film with the round-shaped domains varying in
size with thicker layers. The films were amorphous and
suggested an isotropic growth showing no preferred orienta-
tion. In general, the roughness increased with increasing thick-
ness. Roughness was determined from the histogram as rms
2–3 nm for thinner films (5 bilayer) and 4–5 nm for
thicker films (10 bilayer). The difference was largely due to
the presence of larger and more irregular patches with thicker
layers broadening the size histogram. It is not clear whether
the aggregates within the domains exhibit anisotropic orienta-
tion. However, it was confirmed that the aggregation
behaviour correlated with some of the photoreaction beha-
viour observed and the fact that the lmax centred at about
296 nm shifted with different thicknesses. There was no visible
morphological change after LPUVL irradiation as observed by
AFM. This result is in accord with spin-coated photoalign-
ment polymer films.31
LC alignment properties studies
A uniform homogeneous alignment of the LC molecules could
be obtained in a parallel cell modified by the irradiated
covalent self-assembled films. The transmittance intensity of
visible light (400–700 nm) through the LC cell between crossed
polarizers was monitored as a function of the rotation angle of
the cell in the plane under a polarizing microscope.
Fig. 12 shows the evolution of transmittance with irradia-
tion time as monitored through the cell. For a 7 bilayer DR/
PACSS film cell, the first few minutes did not produce
significant changes in the transmittance-angle dependence.
However, at 5 min birefringence began to be observed. The
maximum of transmittance was reached at 70% at 10 min. The
transmittance began to decrease with further irradiation. This
result suggests that the LPUVL irradiation time of 10 min was
AFM images
Fig. 11 showed micrographs of the DR/PACSS films at
different thicknesses. The morphology was comprised of
corrugated features with many round-shaped domains.
Clearly the surface topography showed a non-smooth but
3020 | J. Mater. Chem., 2007, 17, 3015–3022
This journal is ß The Royal Society of Chemistry 2007

View Article Online
photoalignment film’s thermal stability.11 Also the LC cell
was left for 1 month under ambient conditions, and the
transmittance was monitored. A slight decrease in intensity
was observed. These results clearly showed the stabilities of the
covalent photoalignment films were enhanced. This higher
stability was ascribed to the linkage conversion of the DR/
PACSS layers from ionic bond to covalent bond.
Conclusions
In this paper, the polyanion containing cinnamoyl group was
self-assembled with diazoresin to form a kind of stable covalent
ultrathin film upon exposure to 365 nm UV irradiation. As LC
alignment films, the photoalignment properties of DR/PACSS
covalent films were investigated. The covalent films were
found to have high anisotropy after irradiation with linearly
polarized UV light. It was also obvious in our experimental
results that the photoreaction properties were largely depen-
dent on the irradiation time and thickness of the film: the
dichroic ratio of the film increased rapidly to a maximum at
10 min and then decreased; and the thicker film had the
larger dichroic ratio after the LPUVL irradiation. A uniform
homogeneous alignment of the LC molecules could be
obtained in the parallel cell modified by the irradiated covalent
self-assembled film. The reorientation behaviors of the LC
molecules were also found to be associated with the film
irradiation time and thickness. Most importantly, the thermal
stability of the film is enhanced, reaching 180 uC. This is
sufficient for commercial requirements. Thus, the incorpora-
tion of the cinnamoyl moiety using the new covalent LBL film
is a promising technique for LC photoalignment studies.
Fig. 12 Angular transmittance intensity of LC cell under crossed
polarizers with 7 bilayer DR/PACSS covalent films with different
LPUVL irradiation times: 1 min (n), 5 min (#), 10 min (%) and
15 min ($).
sufficient to induce preferential reorientation of LC molecules,
which is consistent with the conclusion drawn from the
dichroic ratio measurements presented above. Similar results
were obtained for a 3 bilayer DR/PACSS film cell (data not
shown). However, the maximum transmittance of the 3 bilayer
cell was about 50% after 10 min, which means that the
birefringence of LC in the 3 bilayer cell is smaller than that in
a 7 bilayer DR/PACSS cell. Since the homogenous alignment
of LC molecules is controlled by the photo-orientation of
cinnamates in the DR/PACSS films, it is clear that the
alignment of LC molecules could be influenced by the
properties of the DR/PACSS film. In previous studies, we
found that the photoalignment behavior of a covalent self-
assembled film was influenced by the thickness of the sample,
i.e., the dichroic ratios were larger for the thicker samples after
the same irradiation time. Hence, the alignment properties of
the LC molecules could be influenced by the thickness of the
DR/PACSS film. This indicates that LC reorientation might
not simply be influenced by the contact photoreaction
molecule, but might be controlled by the overall thickness
and ordering of the film. It is well known that dipole force,
polarity, surface energy and topological factors influence the
surface anchoring energy of LC molecules.32 The differences
in the orientation and aggregation of cinnamates in the
covalently self-assembled films must influence these photo-
alignment properties of the film. Thus, the alignment of LC is
associated with the thickness of the DR/PACSS film. This is in
contrast to previously reported photoalignment behaviors in
LB which were thickness independent.33,34
Acknowledgements
The authors are grateful to the National Natural Science
Foundation of China (Grant No. 50473040, 60578035) and the
Natural Foundation of Jilin Province (Grant No. 20050520,
20050321-2) for financial support.
References
1 J. L. Janning, Appl. Phys. Lett., 1972, 21, 173.
2 J. Hoogboom, T. Rasing, A. E. Rowan and R. J. M. Nolte,
J. Mater. Chem., 2006, 16, 1305.
3 M. Nishikawa, B. Taheri and J. L. West, Soc. Inf. Disp. Int. Symp.
Dig. Tech. Pap., 1998, 29, 131.
4 E. Sackmann, J. Am. Chem. Soc., 1971, 93, 7088.
5 W. E. Haas, K. F. Nelson, J. E. Adams and G. A. Dir,
J. Electrochem. Soc., 1974, 121, 1667.
6 K. Ichimura, Y. Suzuki, T. Seki, A. Hosoki and K. Aoki,
Langmuir, 1988, 4, 1214.
The stability of the LC cell was also examined against heat
treatment. The cell was injected at 71 uC and heated to 75 uC
for 10 min, and then cooled to room temperature naturally.
The LC in the cell maintained good alignment. Then the cells
were sequentially heated at a temperature 5 uC higher than the
preceding heating step and cool as before. We discovered that
even when the temperature reached 180 uC the cells could
maintain good homogeneous alignment as before. These
results indicated that the covalent films maintained their
LC alignment properties up to at least 180 uC, which was
approximately 80 uC higher than the conventional LBL
7 K. Ichimura, Chem. Rev., 2000, 100, 1847.
8 M. Schadt, K. Schmitt, V. Kozinkov and V. Chigrinov, Jpn. J.
Appl. Phys., 1992, 31, 2155.
9 Y. H. Wang, C. Y. Xu, A. Kanazawa, T. Shiono, T. Ikeda,
Y. Matsuki and Y. Takeuchi, J. Appl. Phys., 1998, 84, 181.
10 Z. H. Peng and L. Xuan, Liq. Cryst., 2005, 32, 239.
11 L. L. Zhang, Z. H. Peng, L. S. Yao, C. H. Fei, F. Z. Lv and
L. Xuan, Appl. Surf. Sci., 2007, 253, 3372.
12 G. Decher and J. Hong, Makromol. Chem., Macromol. Symp.,
1991, 46, 321.
13 G. Decher, Science, 1997, 277, 1232.
14 S. Dante, R. Advincula, C. Frank and P. Stroeve, Langmuir, 1999,
15, 193.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3015–3022 | 3021

View Article Online
15 J. Sun, T. Wu, Y. P. Sun, Z. Wang, X. Zhang, J. Shen and W. Cao,
Chem. Commun., 1998, 1853.
16 J. Y. Chen and W. X. Cao, Chem. Commun., 1999, 1711.
17 S. G. Cao, C. Zhao and W. X. Cao, Polym. Int., 1998, 45, 142.
18 T. Cao, S. Yang, Y. Yang, C. Huang and W. Cao, Langmuir, 2001,
17, 6034.
19 G. Decher, J. D. Hong and J. J. Schmidt, Thin Solid Films, 1992,
210–211, 831.
20 M. Kasha, Radiat. Res., 1963, 20, 55.
21 M. Kasha, M. AshrafEl-Bayoumi and W. Rhodes, J. Chem. Phys.,
1961, 58, 916.
26 N. Nakanishi and P. H. Solomon, in Infrared Absorption
Spectroscopy, Holden-Day, Inc,San Francisco, 1977, vol. III,
pp. 19 and 26.
27 N. Kawatsuki, K. Matuyoshi, M. Hayashl, H. Takatsuka and
T. Yamamoto, Chem. Mater., 2000, 12, 1549.
28 M.-K. Park and R. C. Advincula, Langmuir, 2002, 18, 4532.
29 K. Ichimura, Y. Akita, H. Akiyama, K. Kudo and Y. Hayashi,
Macromolecules, 1997, 30, 903.
30 S. Perny, P. Le Barny, J. Delaire, T. Buffeteau and C. Sourisseau,
Liq. Cryst., 2000, 27, 329.
31 J. Liu, X. Liang, H. Tang, S. Xu and H. Gao, Thin Solid Films,
2001, 384, 212.
32 P. G. de Gennes and J. Prost, The Physics of Liquid Crystals,
Clarendon Press, Oxford, 1993.
22 R. Advincula, E. Fells and M. Park, Chem. Mater., 2001, 13, 2870.
23 R. Advincula, M.-K. Park, A. Baba and F. Kaneko, Langmuir,
2003, 19, 654.
24 J. Yue, Z. H. Wang, K. R. Cromack, A. J. Epstein and
A. G. Macdiarmid, J. Am. Chem. Soc., 1991, 113, 2665.
25 T. Cao, L. Wei, S. Yang, M. Zhang, C. Huang and W. Cao,
Langmuir, 2002, 18, 750.
33 T. Seki, M. Sakuragi, Y. Kawanishi, Y. Suzuki, T. Tamaki,
R. Fukuda and K. Ichimura, Langmuir, 1993, 9, 211.
34 N. Katayama, Y. Ozaki, T. Seki, T. Tamaki and K. Iriyama,
Langmuir, 1994, 10, 1898.
3022 | J. Mater. Chem., 2007, 17, 3015–3022
This journal is ß The Royal Society of Chemistry 2007