All-optical pulsed writing in azobenzene copolymer films in the sub-millisecond regime

Article abstract of DOI:10.1039/c2jm30596h

Block copolymers P(MA4-block-MMA)10 and P(MA4-block-MMA)20 and a random copolymer P(MA4-ran-MMA)10 were prepared from azobenzene methacrylate (MA4) and methyl methacrylate (MMA). The former monomer was introduced as both a photoresponsive and a mesogenic component, and the latter as a transparent non-mesogenic component. Thin films of the copolymers were used to investigate all-optical writing by linear-polarized illumination in the pulsed regime to modify the local birefringence at the microscopic scale. The optical properties of P(MA4-block-MMA)10 were compared with those of P(MA4-ran-MMA)10 having the same chemical composition (10 mol% MA4) but different distributions of the two components. It was found that in the block copolymers relatively intense pulses as short as 100 μs produced a remarkable local increase in the birefringence, stable in the time-scale (up to several days) explored in the experiment. In contrast, efficient and stable optical modifications could not be achieved in the random copolymer.

Full text of DOI:10.1039/c2jm30596h

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 14510
www.rsc.org/materials
PAPER
All-optical pulsed writing in azobenzene copolymer films in the sub-
millisecond regime†
a
a
a
Sara Menghetti, Michele Alderighi, Giancarlo Galli, Francesco Tantussi,b Massimo Morandini,b
*
Francesco Fusob and Maria Allegrinib
Received 31st January 2012, Accepted 16th May 2012
DOI: 10.1039/c2jm30596h
Block copolymers P(MA4-block-MMA)10 and P(MA4-block-MMA)20 and a random copolymer
P(MA4-ran-MMA)10 were prepared from azobenzene methacrylate (MA4) and methyl methacrylate
(MMA). The former monomer was introduced as both a photoresponsive and a mesogenic component,
and the latter as a transparent non-mesogenic component. Thin films of the copolymers were used to
investigate all-optical writing by linear-polarized illumination in the pulsed regime to modify the local
birefringence at the microscopic scale. The optical properties of P(MA4-block-MMA)10 were
compared with those of P(MA4-ran-MMA)10 having the same chemical composition (10 mol% MA4)
but different distributions of the two components. It was found that in the block copolymers relatively
intense pulses as short as 100 ms produced a remarkable local increase in the birefringence, stable in the
time-scale (up to several days) explored in the experiment. In contrast, efficient and stable optical
modifications could not be achieved in the random copolymer.
oriented perpendicular to the polarization of the incident light.6
1. Introduction
This process has a cooperative character and photoorientation
Among photoactive polymers, those consisting of an azobenzene
by a photoselection process is especially relevant in liquid crys-
talline polymers, other than amorphous polymers.7 The trans
chromophore have attracted numerous studies of both funda-
mental significance and practical potential.1 The photoresponse
form of the azobenzene is rodlike and linear and supports the
of these materials resides in the photoisomerization of the
structure of a liquid crystalline mesophase. In contrast, the
constituent azobenzene units that undergo large configurational
photoinduced cis form has a bent shape and is not compatible
changes between two states in response to the absorption of light
with the liquid crystalline mesophase, which is destabilized.8
at two different wavelengths.2 Thus, the photochemical inter-
Thereby, anisotropy of the optical properties of azobenzene
conversion between the thermally relaxed trans isomer and the
polymers can be induced by illumination with polarized light
depending on the polarization of the pump light,1 and their
higher energy cis isomer results in marked changes in shape and
dipole moment/polarity of the azobenzene molecules. This
potential for reversible optical data storage, especially holo-
special feature has inspired many diverse concepts for exploita-
graphic optical storage, has become an area of vast interest.9,10
tion.1 Some of the most recent advances are in the areas of DNA-
The photoorientation process is typically accompanied by
based optomechanical molecular motors,3 electrospun nano-
extensive mass migration phenomena leading, for instance, to
fibers with photodriven wettability,4 and photoswitching
surface relief formation.11–18 Local topographical modifications
photonic crystals.5
can cause variations of the optical properties, but adversely affect
In addition, the photoisomerization is only activated when the
material performance for optical data storage purposes. Mass
transition dipole moment axis of the chromophore has
migration is inherently associated with relatively long exposure
a component parallel to the linear light polarization; the
times, which must be shortened significantly in order to make the
perpendicular direction is excluded from optical activation and
process competitive with other data storage mechanisms.
becomes enriched in chromophores. Within the steady state of
Moreover, the photoinduced formation of surface topographies
the trans–cis–trans repeated excitation–relaxation cycles accom-
can prevent the achievement of full reversibility in the written
panied by rotational diffusion, the azobenzene groups become
features and reduce the spatial resolution of the optical bit.
Block copolymers consisting of a polymer block with photo-
aDipartimento di Chimica e Chimica Industriale, UdR INSTM Pisa,
Universitaꢀ di Pisa, 56126 Pisa, Italy. E-mail: gallig@dcci.unipi.it
active groups can enlarge the potential for application in new
smart technologies, as they undergo microphase separation
bDipartimento di Fisica ‘Enrico Fermi’ and CNISM, Universitaꢀ di Pisa,
leading to well-defined morphologies with nanoscale domains
56127 Pisa, Italy
and favoring optically driven nanomodifications.19 In fact, phase
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2jm30596h
separation in nanosized domains is expected to be relevant in
14510 | J. Mater. Chem., 2012, 22, 14510–14517
This journal is ª The Royal Society of Chemistry 2012

View Online
azobenzene block copolymers,20 when low chromophore
contents are used for efficient photoorientation processes.21–33 In
these geometries, higher densities and shorter writing times of
optical bits may be achieved by taking advantage of the highly
homogeneous chemical substrate within the azobenzene nano-
sized domains.
2.2. Block and random copolymers
We describe in detail a typical two-step preparation of P(MA4-
block-MMA)10 (Scheme 2).
In the first step, 18.72 g (187 mmol) of MMA, 0.2 mg (0.9
mmol) of PMDETA, and 0.18 g (0.9 mmol) of EBrIB were
introduced into a dry Schlenk flask under nitrogen. The solution
was purged with nitrogen for 15 min and then 0.129 g (0.9 mmol)
of CuBr was added. After three freeze–thaw–pump cycles, the
polymerization was let to proceed under nitrogen for 4.5 h at
90 ꢀC. The polymer solution was diluted with THF and then
eluted several times on neutral alumina. The polymer was then
purified by repeated precipitations from THF solutions into
methanol. Yield 68%; Mn ¼ 20 000 g molꢁ1 (by SEC).
Optical writing at the sub-micrometer scale is frequently per-
formed13,15,17,34–38 by using quasi-point-like sources of non-
propagating electromagnetic fields as in scanning near-field
optical microscopy (SNOM).39 Thanks to the simultaneous
acquisition of optical and topographical maps, SNOM can
provide evidence or even investigate the dynamics of mass
migration phenomena, and we have recently demonstrated
negligible topographical modifications in azobenzene block
copolymers written in the near-field at relatively short exposure
times.38 Despite powerful insight into the local features, SNOM
is not suited for systematic studies or for investigation of the
long-term persistence of the optical bit.
In the second step, 0.843 g (0.043 mmol Br terminal) of
PMMA-Br, 1.00
g (2.146 mmol) of MA4 and 4.6 mg
(0.043 mmol) of CuCl were introduced into a dry Schlenk flask,
which was then evacuated and flushed with nitrogen three times.
Then, 20 mL of anisole and 9.9 mg (0.043 mmol) of HMTETA
were added under nitrogen and four freeze–thaw–pump cycles
were applied. The polymerization was let to proceed for 48 h at
80 ꢀC. When the reaction was stopped, the polymer solution was
washed thoroughly with water. The polymer was finally purified
by repeated precipitations from dichloromethane solutions into
The main objective of this work was to write single optical bits,
that is to modify the birefringence of the material upon optical
control under optimal conditions in azobenzene block copoly-
mers that were specifically designed to contain a low amount (10
and 20 mol%) of an azobenzene methacrylate block (with
a methyl methacrylate block). For this purpose, a comprehensive
analysis as a function of writing time, energy and substrate
temperature was carried out addressing in particular the persis-
tence of the written optical bit. A random copolymer with an
equal composition (10 mol%) of the azobenzene methacrylate
units was also prepared for the sake of comparison of the optical
response. We found that information could be optically stored in
the block copolymers, resulting in a well contrasted local modi-
fication of the birefringence, using 1–10 mW writing pulses as
short as 100 ms. Furthermore, the use of these operational
conditions allowed us to reliably verify the persistence of the
optically written features on a time-scale of up to several days,
suggesting long-term stability of the optical bit, and to assess
reversible writing. In contrast, the random copolymer underwent
a poor and unstable modification of its birefringence.
methanol at 40 ꢀC. Yield 56%; Mn ¼ 27 000 g molꢁ1, Mw
¼
35 000 g molꢁ1 (by SEC).
FT-IR (film on KBr): ꢁn (cmꢁ1) ¼ 3072, 2942, 2867, 1733, 1600,
1251 and 840. 1H NMR (CDCl3): d (ppm) ¼ 7.9 (2H, aromatic),
7.7 (2H, aromatic), 6.9 (3H, aromatic), 3.9 (6H, CH2O), 3.6
(27H, COOCH3), 2.2 (3H, aromatic CH3), 2.1–1.2 (62H, CH2
and CH2C(CH3)COO), 1.0 (3H, CH3). 13C NMR (CDCl3):
d (ppm) ¼ 177.8, 174.5, 161.7, 161.3, 147.3, 147.2, 124.4, 114.8,
114.3, 68.1, 64.5, 55.4, 51.7, 44.4, 36.0, 29.0–25.5, 15.8. UV-Vis
(CHCl3, c ¼ 5 ꢂ 10ꢁ5 M azobenzene): lmax (nm) ¼ 362 (p–p*),
450 (n–p*).
For the preparation of P(MA4-ran-MMA)10, MMA (54.84
mmol), MA4 (9.677 mmol) and AIBN (0.609 mmol) in 60 mL of
benzene were reacted at 65 ꢀC for 48 h under typical conditions of
high vacuum nitrogen atmosphere. Yield 98%; Mn ¼ 24 000 g
molꢁ1, Mw ¼ 56 000 g molꢁ1 (by SEC).
2. Experimental
2.3. Polymer films
2.1. Materials
N,N,N0,N00,N000,N000-Hexamethyltriethylenetetramine (HMTETA),
Thin polymer films (200–600 nm) were obtained by spin-coating
toluene solutions (2–5 wt%) onto cleaned glass cover slides. Prior
N,N,N0,N00,N00-pentamethyldiethylenetriamine
(PMDETA),
ethyl 2-bromoisobutyrate (EBrIB), 4-pentoxyaniline (1), o-
cresol, and 6-chloro-1-hexanol were used without further puri-
fication (Sigma-Aldrich, >98–99%). Methacryloyl chloride was
distilled at 400 mbar prior to use. Methyl methacrylate (MMA)
was washed with 5% NaOH, water and dried over Na2SO4 and
distilled. Benzene, tetrahydrofuran, and anisole were refluxed
and distilled over Na–K alloy at atmospheric pressure under
nitrogen. Triethylamine was refluxed and distilled over KOH.
Dimethylsulfoxide was refluxed and distilled under vacuum over
CaH2. a,a0-Azobis(isobutyronitrile) (AIBN) was crystallized
from methanol.
The synthesis of the azobenzene methacrylate monomer, 4-
pentoxy-30-methyl-40-((6-methacryloyloxy)hexyloxy)azobenzene
(MA4), starting from (1) (Scheme 1) is described in the ESI†.
Scheme 1 Reaction pathway for the synthesis of monomer MA4.
J. Mater. Chem., 2012, 22, 14510–14517 | 14511
This journal is ª The Royal Society of Chemistry 2012

View Online
spot by a 10ꢂ microscope objective. Monitoring of the writing
process (readout) was accomplished by measuring the thin film
birefringence in the same spot. To this aim, a polarization-
modulated red laser beam (low power radiation at 690 nm,
a wavelength far from the absorption band of the material) was
superposed on the writing laser beam, allowing evaluation of the
optical retardation induced by the writing process. The use of
polarization-modulation40 replicates the configuration employed
for birefringence measurements in the near field, where the
technique is frequently used to analyze optical activity at
the local scale. Due to the relatively small size of the focal spot,
the writing radiation intensity can reach the kW cmꢁ2 range with
an input power as low as a few mW. Such an intensity range is
typical for near-field optical nanowriting,13–15,35,38 but larger than
in experiments aiming at material modifications on the extended
scale.9–11,18,20,22,24
Scheme 2 Reaction pathway for the preparation of block copolymers
P(MA4-block-MMA)z.
to experiments, the films were annealed for 12–24 hours at 130 ꢀC
to attain equilibrium morphology and enhance homogeneity.
Optical bits were produced in the pulsed regime that allows for
exploring a wide range of writing energy doses and verifying the
performance of the material in terms of writing speed, while
preventing effects associated with material heating. The laser
power was adjusted by a set of neutral filters in the range P ¼
0.05–10 mW (measured at the entrance of the microscope
objective). Depending on the pulse duration Dt, typically 100 ms
through 100 ms, an energy dose D ¼ PDt ¼ 5 ꢂ 10ꢁ3 to 103 mJ
could be achieved. A custom-made sample holder equipped with
a temperature-controlled stage enabled varying the sample
temperature in the range T ¼ 25–70 ꢀC.
2.4. Characterization of materials
1H and 13C NMR spectra were recorded with a Varian Gemini
VRX 300 spectrometer on solutions of the polymer in CDCl3.
FT-IR spectra were recorded with a Spectrum One Perkin-Elmer
spectrophotometer on polymer films cast from solution on a KBr
crystal plate. UV-Vis absorption spectra were recorded on
polymer solutions and spin-coated films with a UV-Vis-NIR
Varian Cary-500 spectrophotometer in the spectral range of 300–
600 nm.
Size exclusion chromatography (SEC) was carried out with
a Jasco PU-1580 liquid chromatograph equipped with two PL gel
5 mm Mixed-D columns, a Jasco 830-RI138 refractive index
detector and a Perkin-Elmer LC75 UV detector. PMMA stan-
dards (1000–80 000 g molꢁ1) were used for calibration.
In the readout, the polarization state of the reading radiation
was periodically changed (at 50 kHz frequency) by a photoelastic
modulator.41 Radiation transmitted from the sample at 690 nm
was analyzed by a fixed axis polarizer and then collected by
a photodiode. Synchronous detection provides with an output
dependent on the sample birefringence.42 By an accurate model
of all optical components, accounting also for residual birefrin-
gence and/or dichroism in mirrors and beam splitters used to
guide the light onto the sample, the optical retardation Df
induced by the writing laser pulse on the material is evaluated
(see ESI†).
Differential scanning calorimetry (DSC) measurements were
performed with a Mettler DSC 30 instrument. Samples (5–10 mg)
were scanned (10 ꢀC minꢁ1 heating/cooling rate) under a pure
nitrogen flow. Temperature and energy were calibrated using
standard samples of cyclohexane, tin, indium and zinc. The
nematic–isotropic transition temperature (TNI) of the polymer
was taken as corresponding to the maximum temperature in the
relevant endothermic peak. The glass transition temperature (Tg)
was taken at the inflection point in the second heating scan.
Polarized optical microscopy (POM) observations were per-
formed on polymer powder sandwiched between glass slides with
a Reichert-Jung Polyvar microscope equipped with a program-
mable Mettler FP52 heating stage (3–10 ꢀC minꢁ1 scanning rate).
The surface morphology of polymer films was investigated
using a Veeco Multimode atomic force microscope (AFM)
equipped with a Nanoscope IV controller in a tapping mode.
Topographic and phase contrast imaging experiments were
performed on 1 ꢂ 1–5 ꢂ 5 mm2 regions. The roughness of the film
surface was evaluated as the root-mean-square roughness,
Rq ¼ (Sizi2/N)0.5, where zi is the height value and N the number of
points measured on the surface analyzed.
3. Results and discussion
3.1. Synthesis
Many examples of polymers based on different azobenzene units
are known,1,43 including MA4-based polymers.44,45 However, the
synthesis of the azobenzene monomer MA4 and its incorpora-
tion into block copolymers, e.g., by a controlled radical poly-
merization, have never been described before. The chemical
structure of MA4 was in fact designed in such a way that azo-
benzene should act as both a photoresponsive and a mesogenic
unit over a wide temperature range above the room temperature
in a nematic side chain polymer.44 Accordingly, a polymerizable
methacrylate group was linked to
a hindered, laterally
substituted azobenzene core carrying relatively short spacer and
tail segments (Scheme 1).
2.5. Optical analysis
MA4 was found to form a monotropic nematic mesophase,
that is a metastable nemaꢀtic mesophase with respect to the
crystalline phase, below 39 C by polarizing optical microscopy
(POM) observations and differential scanning calorimetry (DSC)
Optical writing, i.e., production of single optical bits, on copoly-
mer thin films spin-coated onto glass was verified by focusing
a 473 nm linear-polarized laser radiation on a ꢃ7 mm diameter
14512 | J. Mater. Chem., 2012, 22, 14510–14517
This journal is ª The Royal Society of Chemistry 2012

View Online
Table 1 Thermal characterization (by DSC at 10 ꢀC minꢁ1 scanning
rate) of the copolymers
measurements. MA4 homopolymer samples with varied molar
mass have previously been shown to form a nematic mesophase
directly above the glass transition temperature (Tg ꢃ21 to 32 ꢀC)
up to the nematic–isotropic transition temperature (TNI ꢃ78 to
a
b
Copolymer
Tg (ꢀC)
TNI (ꢀC)
84 C).45 Therefore, the nematic mesophase is stabilized in the
P(MA4-ran-MMA)10
P(MA4-block-MMA)
83
55, 124c
—
80
ꢀ
polymer with respect to the monomer and extends over a rela-
tively wide temperature window. This in turn can allow for
10
P(MA4-block-MMA)
20
51, 121c
85
freezing an optically imprinted information in the MA4 polymer
46
below Tg and for erasing it above TNI
.
a
b
Glass transition temperature. Nematic–isotropic phase transition
In this work we introduced MA4 into two kinds of copolymer
structures along with methyl methacrylate (MMA), namely
random copolymer and block copolymer structures. One random
copolymer P(MA4-ran-MMA)10 (10 mol% MA4) was prepared
by conventional free radical, AIBN initiated polymerization.
Two block copolymers, P(MA4-block-MMA)10 and P(MA4-
block-MMA)20 (10 mol% and 20 mol% MA4, respectively), were
prepared by a two-step procedure involving sequential atom
transfer radical polymerizations (ATRP) (Scheme 2). MMA was
chosen as a photochemically stable, transparent component in
order to reduce the light absorption at the recording
wavelength.22,23,47,48
c
temperature. Glass transition temperatures of the MA4 and MMA
polymer blocks, respectively.
glassy phase of the MMA block and was frozen down to room
temperature in a glassy state below Tg ꢃ 50. Morphological
constraints from the rigid MMA block served to increase the Tg
of the MA4 block over that of the MA4 homopolymer.45 This in
turn could help to improve the long term stability of an imprinted
optical information by preventing smearing out of the photoin-
duced orientation. In contrast, dilution of the MA4 units by non-
mesogenic MMA units in the random copolymer prevented the
onset of a mesophase above its highly increased Tg ¼ 83 ꢀC
relative to the MA4 homopolymer.
Following a macroinitiator route, a bromo-terminated poly-
(methyl methacrylate) macroinitiator (PMMA-Br) was firstly
prepared by ATRP of MMA using ethyl 2-bromoisobutyrate
(EBrIB) as an initiator with CuBr–pentamethyldiethylenetri-
Atomic force microscopy (AFM) investigations on thin films
(thickness 300–600 nm) revealed a cylinder-like morphology for
both block copolymers, with the MA4 photoresponsive nano-
domains (diameter ꢃ15 to 30 nm) being dispersed in the trans-
parent incompatible MMA matrix and lying parallel to the
surface (Fig. 1). Such morphology of phase separated polymer
blocks was stable up to the order–disorder transition at 130–140
ꢀC. AFM scans on the random copolymer thin films (not shown
here) were featureless with no domain morphology. The polymer
films were found to be very smooth, the AFM root-mean-square
roughness being Rq ¼ 2–5 nm. This was unaltered by the optical
writing in the photoaddressed films.
amine
(CuBr–PMDETA)
as
a
catalyst
(MMA : EBrIB : CuBr : PMDETA 200 : 1 : 1 : 1 mole ratio).
The size exclusion chromatography (SEC) molar mass of
PMMA-Br was Mn ¼ 20 000 g molꢁ1, corresponding to a degree
of polymerization of 200. PMMA-Br was then used to initiate
and grow the ATRP block of MA4 using CuCl–hexamethyl-
triethylenetetramine (1 : 1 mole ratio) (CuCl–HMTETA) as
a
catalyst. Block copolymers P(MA4-block-MMA)10 and
P(MA4-block-MMA)20 were prepared by varying the
MA4 : PMMA-Br mole ratio and reaction time, 50 : 1 (for 48 h
at 80 ꢀC) and 70 : 1 (for 96 h at 95 ꢀC), respectively. The 1H NMR
composition of the two block copolymers was 10 and 20 mol%
MA4 (z in Scheme 2), corresponding to degrees of polymeriza-
tion of the MA4 polymer block of ꢃ22 and ꢃ51, respectively.
Thus, the controlled nature of the ATRP49 enabled us to prepare
block copolymers in which the length of the MMA polymer
block was constant but the length of the MA4 polymer block was
changed in a tailored way. Moreover, it was possible to compare
the properties of the two copolymers having one same chemical
composition (10 mol% MA4), but completely different distri-
butions of the two constituent units in alternatively random
sequences or polymer blocks.
3.2. Optical writing
The UV-Vis absorption spectra of spin-coated films of all the
copolymers exhibited two partially overlapping bands, typical of
the electronic transitions of the azobenzene chromophore, p–p*
at 362 nm and n–p* around 450 nm (ref. 10, 17, and 35)
(Fig. S2†). Therefore, by using a suitable wavelength, complete
trans–cis–trans photoisomerization cycles can be induced. Such
photoisomerization cycles are used in the writing process to drive
the photoorientation of the azobenzene units. To this aim, in our
experiments we used laser radiation at 473 nm. The use of 488 nm
light is also reported in the literature.10,17,35 Once the optical bit is
produced by illumination of the thin film with pulsed 473 nm
laser light, the optical retardation Df, representative of the
material birefringence, is measured on the same spot.
DSC and POM investigations of the copolymers evidenced
different thermal and phase behavior (Table 1). Whereas the
random copolymer exhibited one glass transition temperature
and no mesophase formation, both block copolymers showed
two distinct glass transitions at ꢃ50 ꢀC and ꢃ120 ꢀC in addition
to a nematic–isotropic phase transition at ꢃ80 ꢀC (Fig. S1†). The
nematic nature of the liquid crystalline mesophase was confirmed
by optical observation of the birefringent textures up to TNI. The
two incompatible MA4 and MMA polymer blocks were micro-
phase separated and underwent their individual thermal transi-
tions, including formation of a nematic mesophase for the MA4
polymer block. One should note that this coexisted with the
Fig. 2 shows as an example Df of written block copolymer
films as a function of the writingꢀ energy dose at various
temperatures in the range T ¼ 25–70 C. Different powers P of
the writing laser (data are shown in separate plots) and different
choices of the pulse duration Dt (as indicated in the plots) were
used. The results demonstrate the achievement of remarkable
birefringence in the films, with a maximum Df above 0.4 rad.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 14510–14517 | 14513

View Online
Fig. 1 AFM phase images of block copolymers P(MA4-block-MMA)10 (left) and P(MA4-block-MMA)20 (right).
Assuming a linear relationship, the optical retardation is
proportional to the film thickness d and to the difference in the
refractive index Dn for two mutually orthogonal directions: Df ¼
Dn2pd/l. For the films considered in Fig. 2, d ¼ (550 ꢄ 50) nm, as
ascertained through AFM imaging of scratched samples.
Therefore, Dn can get values on the order of 0.05–0.1 with
a writing energy dose in the mJ range.
photoorientation process and in the subsequent persistence of the
optical bit.
In block copolymers, our data show that the imprinted
birefringence was a generally increasing function of the dose
with a strongly sub-linear dependence; note the logarithmic
horizontal axis in the plots. Looking, for instance, at the data
acquired at T ¼ 40 ꢀC (filled triangles in the plots), an increase
by three decades in the dose, obtained by varying the pulse
duration in the range 1–100 ms, led to Df increased by less
than a factor of two. Moreover, in some cases, e.g., at large
laser power and moderate temperature, the birefringence
tended to saturate at large doses, indicating the attainment of
an almost complete photoorientation in the written optical
bit.
Such an efficient writing process was found only for block
copolymers. Data acquired on the random copolymer (not dis-
played in graphs) were in fact markedly different. Only for
relatively long writing pulses, e.g., Dt > 10 ms, and energy doses
above 102 mJ, a Df ꢃ 0.1 rad, comparable to the measurement
uncertainty, could be achieved. This clearly demonstrates that
the block architecture plays an essential role in the
Fig. 2 Optical retardation Df achieved in thin films of the block copolymers P(MA4-block-MMA)10 (left) and P(MA4-block-MMA)20 (right) as
a function of the writing energy dose. Different plots correspond to different powers of the writing laser; the various symbols refer to different
temperatures, as specified in the insets. The floating horizontal axis shows the pulse duration Dt, varied in the range 0.1–100 ms. A representative error
bar is shown for a selected data point in every plot.
14514 | J. Mater. Chem., 2012, 22, 14510–14517
This journal is ª The Royal Society of Chemistry 2012

View Online
Any temperature-related change in the material properties can
reflect in a modification of the writing efficiency. We found out
ꢀ
that, in the block copolymer samples, heating above 50 C, i.e.,
above Tg of the MA4 block, drastically decreased the writing
efficiency. The behavior in the range between room temperature
and T ¼ 50 ꢀC, studied in steps of 5 ꢀC, exhibited a slight increase
(around 10%) in Df for 35 ꢀC < T < 45 ꢀC, followed by a more
marked decrease when Tg was approached. Our finding can be
relevant in view of practical exploitation in the context of optical
data storage, since it suggests that no temperature conditioning
or heating above room temperature is required.
A comparison between results obtained with the two block
copolymers shows that the general behavior was not strongly
affected by the copolymer composition. In particular, both
materials could be efficiently written and the increasing trend of
Df as a function of the writing dose, as well as the value Df
reached at large doses, was similar. However, the block copol-
ymer with the largest azobenzene content demonstrates a distinct
ability to effectively modify its birefringence with short pulses.
This is well illustrated in Fig. 3, where the Df induced in
Fig. 4 Temporal behavior of the optical retardation Df measured after
writing P(MA4-block-MMA)10 and P(MA4-ran-MMA)10 films (filled
dots and solid line, respectively) with P ¼ 10 mW, Dt ¼ 20 ms. Data are
normalized to the Dft¼1 s values measured 1 s after the laser shot. Data
for the block copolymer were obtained by averaging the measurements
over different optical bits written at discrete intervals. Data for the
random copolymer were produced by direct acquisition of the experi-
mental signals through a digital oscilloscope.
ꢀ
a P(MA4-block-MMA)20 sample kept at T ¼ 40 C is plotted
versus the writing energy dose. A remarkable optical retardation
was attained with writing pulses as short as 100 ms and 1 mW
power, which makes this polymer an excellent candidate for
optical data storage applications.
substantial stability of the birefringence induced by optical
writing. Since no decreasing trend can be observed, stability at
even longer times is anticipated. Dramatically different results
were obtained with the P(MA4-ran-MMA)10 copolymer after
writing under the same conditions. In this case, due to the rapid
decrease in Df, a digital oscilloscope was used to acquire the
temporal behavior, producing the solid line plotted in the graph
of Fig. 4.
In general, for a certain writing dose, the imprinted birefrin-
gence was larger for shorter writing pulses, i.e., for stronger laser
powers. Such a behavior is particularly clear at relatively low
doses, where Df was smaller than the saturation value, and in the
P(MA4-block-MMA)20 sample. Under these conditions, the
photoinduced alignment appears to be governed by the power
rather than the total energy deposited in the sample, suggesting
that for fast writing processes the influence of mechanisms
competing with photoorientation is less pronounced.
In the random copolymer the birefringence vanished in
a rather short time-scale. The observed decay can be fitted by
a double exponential function leading to two characteristic times
(around 20 s and 200 s, respectively). Decays described by
a superposition of different exponential functions can be ascribed
to the presence of different relaxation processes.
The temporal evolution of the birefringence measured on the
optical bit can take place on a longer time-scale due to slow
phenomena, for instance, involving interaction with the thermal
bath and canceling the effects of the photoinduced orientation.
As an example, Fig. 4 shows (filled dots) the temporal behavior
of Df, normalized to the value Dft ¼ 1 s measured 1 s after the
writing pulse, in optical bits written on a P(MA4-block-MMA)10
film at a dose D ¼ 200 mJ (P ¼ 1 mW, Dt ¼ 20 ms). The explored
time range is on the order of a few days and data demonstrate the
The lack of stability in random copolymers can be ascribed to
decreased interactions among azobenzene units,10,24 which can be
conversely enhanced in block copolymers due to phase separa-
tion in nanodomains. Our results demonstrate that the random
copolymer cannot be used to obtain persistent optical
modifications.
Optical encoding of information requires the ability both to
control the features of the written bit by the polarization direc-
tion of the writing radiation37 and to optically erase, or over-
write, a previously imprinted feature. In practical terms, this
entails the possibility to modify the birefringence of an optical
bit, produced by irradiation with a certain polarization, through
subsequent irradiation polarized along an orthogonal direction.
In order to assess this ability, we performed experiments where
the polarization of the writing laser was set to a certain direction
and a sequence of writing pulses, all delivering the same energy
dose, was sent to the same spot on the sample surface. Then
a new sequence of pulses was sent to the same spot with polari-
zation rotated by 90ꢀ. Fig. 5 shows the birefringence signal
acquired by a digital oscilloscope during a so-conceived over-
writing experiment on a P(MA4-block-MMA)10 film. The hori-
zontal axis represents time, while the vertical axis reports the
Fig. 3 Optical retardation Df as a function of the writing energy dose
measured after writing on a P(MA4-block-MMA)20 film at T ¼ 40 ꢀC by
using different pulse durations, as shown in the inset. Solid lines are
guides for the eyes.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 14510–14517 | 14515

View Online
variation of Df referenced to the original value (arbitrary units
are used here; the unity roughly corresponds to Df ¼ 0.5). The
top trace shows the firing of the laser pulses, which are sequen-
tially numbered; after pulse 10, the polarization was rotated by
90ꢀ and after pulse 20 it was rotated back to the initial direction.
In the first pulse sequence, Df follows an incremental rise, the
increment being smaller once repeated irradiation has been per-
formed due to saturation in the birefringence locally induced in
the thin film. After the writing laser polarization was rotated by
90ꢀ, the measured Df re-started to markedly change, following,
however, an opposite trend. The incremental decrease in bire-
fringence during this stage can be explained by considering that
writing consists now in aligning the trans azobenzene units along
a direction orthogonal to the previous one. This leads to
canceling the previously written optical bit, by overwriting along
an orthogonal direction.
birefringence occurs in a short time-scale owing to the large
intensity achieved in the experiment, that leads to a fast rate of
trans–cis–trans photoisomerization cycles. Mechanisms ruled by
different time-scales are typically involved in optical writing, as
also ascertained by investigations at the local scale.36,50 Mass
migration and other topographical effects are known to occur on
longer time-scales and to play a marginal role with sub-ms
writing pulses. Cooperative orientation effects can instead be
involved. Our results indicate that such effects determine efficient
and fast writing of the optical bit. Furthermore, due to the small
film thickness, hence to the limited amount of material involved
in the process, saturation of the effect is achieved on a relatively
short time-scale. This is not the case of holographic (volume)
optical data storage, where formation of surface reliefs due to
mass migration normally takes place. Our conditions are,
however, typical for writing of single nanosized optical bits,
where thin films are necessarily used in order to allow coupling
with the near-field. As we have previously demonstrated,38 the
use of short exposure times of block copolymers featuring a small
azobenzene content can lead to pure optical modifications, with
negligible contributions from topographical modifications.
Remarkably, the first laser pulse after the polarization was
rotated (pulse 11) produced the largest birefringence variation.
In fact, this laser pulse interacts with a sample almost completely
aligned along the parallel direction with respect to its polariza-
tion. Hence, optical absorption is enhanced and the photo-
orientation process is more efficient than, for instance, in the very
first laser pulse of the whole sequence (pulse 1), which interacted
with a randomly oriented sample. The dependence of the writing
efficiency on the status of the sample is confirmed by the varia-
tion induced by pulse 21, having a polarization rotated back to
the initial direction. Also in this case, the laser pulse on a pre-
aligned sample yielded a variation similar (but obviously
reversed in sign) to that produced by pulse 11. The ability to tune
the optical retardation Df according to the polarization direction
of the writing light can open the way to possible applications
where multi-bit encoding is accomplished through control of the
polarization in the writing laser beam.37
4. Conclusions
The investigated azobenzene block copolymers are suitable
recording media for writing optical bits in a fast regime. Using 1–
10 mW writing pulses as short as 100 ms local high-contrast
changes in birefringence were produced, which remained unaf-
fected over time. Owing to microphase separation, the homo-
geneous chemical environment of the discrete azobenzene
nanodomains favours cooperative reorientation to a direction
perpendicular to the incident light polarization. The strong
driving force for photoselection of the chromophores is enhanced
in the nematic phase of the block copolymers. In contrast, this
cannot be effected in the non-liquid crystalline random copol-
ymer, in which more modest and unstable optical writing was
only achieved over time. While the present experiment, being
based on micrometer-sized optical spots, cannot shed light on
material homogeneity and morphological modifications at the
nanometer scale, the operational conditions identified in this
work may be specifically implemented for high-resolution, all-
optical nanowriting, e.g., by SNOM, in azobenzene block
copolymers with no concomitant, negative changes in surface
topography, as we have recently demonstrated.38
Our findings confirm that illumination with polarized light
induces anisotropy in the material optical properties,1 that is
responsible for the measured birefringence. Build-up of the
Acknowledgements
The authors thank the Regione Toscana, POR OB3 Misura D4,
for partial financial support of the work.
Notes and references
Fig. 5 Acquisition by a digital oscilloscope of the optical retardation
imprinted by a sequence of writing pulses (shown in the top trace) with P
¼ 10 mW and Dt ¼ 100 ms with a P(MA4-block-MMA)10 film. Polari-
zation of the writing laser is rotated by 90ꢀ after pulse 10 (around 40 s
after the beginning of the acquisition) and then rotated back after pulse
20 (around 80 s after the beginning of the acquisition). We note that the
temporal behavior of the signal cannot be considered representative of
the birefringence build-up, due to the finite integration time of the used
instrumentation determining an artificial exponential trend.
1 Smart Light-responsive Materials: Azobenzene-containing Polymers
and Liquid Crystals, ed. Y. Zhao and T. Ikeda, Wiley, New York,
2009.
2 H. Rau, Photoisomerization of Azobenzenes, in Photochemistry and
Photophysics, ed. J. F. Rabek, CRC Press, Boca Raton, 1990, vol.
2, p. 119.
3 M. McCullagh, I. Franco, M. A. Ratner and G. C. Schatz, J. Am.
Chem. Soc., 2011, 133, 3452.
4 M. Chen and F. Besenbacher, ACS Nano, 2011, 5, 1549.
14516 | J. Mater. Chem., 2012, 22, 14510–14517
This journal is ª The Royal Society of Chemistry 2012

View Online
5 M. Moritsugu, S.-N. Kim, S. Kubo, T. Ogata, T. Nonaka, O. Sato
and S. Kurihara, React. Funct. Polym., 2011, 71, 30.
6 I. Zebger, M. Rutloh, U. Hoffman, J. Stumpe, H. W. Siesler and
S. Hvilsted, J. Phys. Chem. A, 2002, 106, 3454.
7 K. Anderle, R. Birenheide, M. J. A. Werner and J. H. Wendorff, Liq.
Cryst., 1991, 9, 691.
8 O. Tsutsumi, T. Shiono, T. Ikeda and G. Galli, J. Phys. Chem. B,
1997, 101, 1332.
ꢂ ꢂ
9 S. Hvilsted, C. Sanchez and R. Alcala, J. Mater. Chem., 2009, 19,
29 W. Deng, P.-A. Albouy, E. Lacaze, P. Keller, X. Wang and M.-H. Li,
Macromolecules, 2008, 41, 2459.
30 G. Wang, X. Tong and Y. Zhao, Macromolecules, 2004, 37, 8911.
31 X. Tong, G. Wang, A. Soldera and Y. Zhao, J. Phys. Chem. B, 2005,
109, 20281.
32 K. Moriya, T. Seki, M. Nakagawa, G. Mao and C. K. Ober,
Macromol. Rapid Commun., 2000, 21, 1309.
33 W. Chen, J.-Y. Wang, X. Wei, J. Xu, A. C. Balazs, K. Matyjaszewski
and T. Russell, Macromolecules, 2011, 44, 278.
6641.
10 P. Forcen, L. Oriol, C. Sanchez, R. Alcala, S. Hvilsted, K. Jankova
34 L. Andreozzi, M. Faetti, G. Galli, M. Giordano and D. Palazzuoli,
Macromol. Symp., 2004, 218, 323.
35 V. Likodimos, M. Labardi, L. Pardi, M. Allegrini, A. Arena,
ꢂ
and J. Loos, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 1899;
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢀ
P. Forcen, L. Oriol, C. Sanchez, F. J. Rodrıguez, R. Alcala,
S. Hvilsted and K. Jankova, Eur. Polym. J., 2008, 44, 72.
11 D. Kim, S. Tripathy, L. Li and J. Kumar, Appl. Phys. Lett., 1995, 66,
1165.
M. Giordano and S. Patane, Appl. Phys. Lett., 2003, 82, 3313.
36 D. Garrot, Y. Lassailly, K. Lahlil, J. Boilot and J. Peretti, Appl. Phys.
Lett., 2009, 94, 033303.
ꢀ
37 M. Savoini, P. Biagioni, L. Duo and M. Finazzi, Opt. Lett., 2009, 34,
761.
38 F. Tantussi, S. Menghetti, E. Caldi, F. Fuso, M. Allegrini and
G. Galli, Appl. Phys. Lett., 2012, 100, 083103.
39 See, for instance L. Novotny and B. Hecht, Principles of Nano-optics,
Cambridge University Press, New York, 2006.
12 X. Jiang, L. Li, J. Kumar, D. Kim, V. Shivshankar and S. Tripathy,
Appl. Phys. Lett., 1996, 68, 2618.
13 N. Landraud, J. Peretti, F. Chaput, G. Lampel, J.-P. Bilot, K. Lahlil
and V. I. Safarov, Appl. Phys. Lett., 2001, 79, 4562.
14 A. Natansohn and P. Rochon, Chem. Rev., 2002, 102, 4139.
ꢀ
15 A. Ambrosio, A. Camposeo, P. Maddalena, S. Patane and
M. Allegrini, J. Microsc., 2008, 229, 307.
16 S. Manickasundaram, P. Kannan, Q. M. A. Hassan and
P. K. Palanisamy, J. Mater. Sci.: Mater. Electron., 2008, 19, 1045.
40 See, for instance P. G. Gucciardi, R. Micheletto, Y. Kawakami and
M. Allegrini, Polarization-Modulation Techniques in Near-Field
Optical Microscopy for Imaging of Polarization Anisotropy in
Photonic Nanostructures, in Applied Scanning Probe Methods, ed.
B. Bhushan and H. Fuchs, Springer, Heidelberg, 2006, vol. 2, ch. 10.
41 J. C. Kemp, J. Opt. Soc. Am., 1970, 59, 950.
ꢂ
17 S. Patane, A. Arena, M. Allegrini, L. Andreozzi, M. Faetti and
M. Giordano, Opt. Commun., 2002, 210, 37.
€
18 K. Gharagozloo-Hubmann, O. Kulikovska, V. Borger, H. Menzel
42 B. Wang, Proc. SPIE, 1998, 3424, 120.
and J. Stumpe, Macromol. Chem. Phys., 2009, 210, 1809.
19 J.-M. Schumers, C.-A. Fustin and J.-F. Gohy, Macromol. Rapid
Commun., 2010, 31, 1588.
20 H. Yu, T. Kobayashi and H. Yang, Adv. Mater., 2011, 23, 3337.
21 H. Yu and T. Ikeda, Adv. Mater., 2011, 23, 2149.
43 For one very recent example, see A. Bobrovsky, V. Shibaev,
A. Bubnov, V. Hamplova, M. Kaspar, D. Pociecha and
M. Glogarova, Macromol. Chem. Phys., 2011, 121, 342.
44 A. S. Angeloni, D. Caretti, M. Laus, E. Chiellini and G. Galli, J.
Polym. Sci., Part A: Polym. Chem., 1991, 29, 1865.
45 L. Andreozzi, M. Faetti, M. Giordano, D. Palazzuoli and G. Galli,
Macromolecules, 2001, 34, 7325; L. Andreozzi, M. Faetti, G. Galli,
M. Giordano and D. Palazzuoli, Mol. Cryst. Liq. Cryst., 2004, 411,
515.
46 G. Galli and E. Chiellini, Liq. Cryst., 2006, 33, 1297 and references
therein.
47 C. Pakula, C. Hanisch, V. Zaporojtchenko, T. Strunskus,
C. Bornholdt, D. Zargarani, R. Herges and F. Faupel, J. Mater.
Sci., 2011, 46, 2488.
ꢂ
22 S. Gimeno, P. Forcen, L. Oriol, M. Pinol, C. Sanchez, F. Rodriguez,
R. Alcala, K. Jankova and S. Hvilsted, Eur. Polym. J., 2009, 45,
~
ꢂ
ꢂ
262.
23 J. Del Barrio, L. Oriol, R. Alcala and C. Sanchez, J. Polym. Sci., Part
A: Polym. Chem., 2010, 48, 1538.
24 H. Yu, Y. Naka, A. Shishido and T. Ikeda, Macromolecules, 2008, 41,
7959.
25 H. Yu, A. Shishido, T. Ikeda and T. Iyoda, Macromol. Rapid
Commun., 2005, 26, 1594.
€
48 T. Breiner, K. Kreger, R. Hagen, M. Hackel, L. Kador,
26 C. Frenz, A. Fuchs, H.-W. Schmidt, U. Theissen and D. Haarer,
Macromol. Chem. Phys., 2004, 205, 1246.
27 S. Kadota, K. Aoki, S. Nagano and T. Seki, J. Am. Chem. Soc., 2005,
127, 8266.
28 K. Aoki, T. Iwata, S. Nagano and T. Seki, Macromol. Chem. Phys.,
2010, 211, 2484.
€
A. H. E. Muller, E. J. Kramer and H.-W. Schmidt,
Macromolecules, 2007, 40, 2100.
49 K. Matyjaszewski and J. H. Xia, Chem. Rev., 2001, 101, 2921.
50 M. Juan, J. Plain, R. Bachelot, P. Royer, K. Gray and P. Wiederrecht,
ACS Nano, 2009, 3, 1573.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 14510–14517 | 14517