Photoinduced orientational dynamics of azophenylcarbazole molecules in polycarbonate

Article abstract of DOI:10.1016/j.dyepig.2011.08.004

The synthesis and new optical features of four azophenylcarbazoles are described. All-optical poling of the azophenylcarbazoles with different donor groups was investigated. The highest optical poling efficiency (χ( ²⁾?0.72pm/V) was measured fo

Full text of DOI:10.1016/j.dyepig.2011.08.004

Dyes and Pigments 92 (2012) 1204e1211
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Photoinduced orientational dynamics of azophenylcarbazole molecules
in polycarbonate
a
a
a
,
b
c
c
*
_
ꢀ ꢁ
ꢀ
ꢀ
ꢀ
ꢀ
_
G. Navickaite , G. Seniutinas , R. Tomasiunas , R. Petruskevicius , V. Getautis , M. Daskeviciene
a
_
Institute of Applied Research, Vilnius University, Sauletekio 10, LT-10223 Vilnius, Lithuania
b Center for Physical Sciences and Technology, Savanoriu Av. 231, 02300 Vilnius, Lithuania
c
_
Kaunas University of Technology, Radvilenu pl. 19, LT-50270 Kaunas, Lithuania
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and new optical features of four azophenylcarbazoles are described. All-optical poling of
the azophenylcarbazoles with different donor groups was investigated. The highest optical poling effi-
Received 12 February 2011
Received in revised form
29 July 2011
Accepted 1 August 2011
Available online 10 August 2011
ciency ð
c
ð2Þx0:72 pm=VÞ was measured for relative larger molecules (branched). Parameters describing
mobility of the azophenylcarbazoles (orientational diffusion coefficient vary from 2.2 ꢀ 10ꢁ5 sꢁ1 to
8 ꢀ 10ꢁ5
s
ꢁ1) were extracted using an adopted model. Photoinduced erasure while reading the poled
media turned out to be more pronounced for less developed structures (less branched donor). Different
polymer materials were investigated as the host matrix for the most interesting azophenylcarbazole
molecule.
Keywords:
Azophenylcarbazole
All-optical poling
Relaxation kinetics
Orientational diffusion
Photoinduced erasure
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
linked with carbazole chromophore embedded into polyimides
show highly stable optically induced birefringence [3]. To gain
The angular redistribution of azo moieties in chromophores
already attracts much attention from the interdisciplinary science
community. Chemists are searching for new materials featuring
original intramolecular links or diverse coupling with the sur-
rounded matrix. Physicists look for non-linear optical properties
introduced by non-centrosymmetry. Technologists have fun with
profiling the films, eg. creating photoinduced surface gratings.
Finally, information technologists gain from developing photo-
functional devices such as optical memory and photoswitching
components. The photoinduced orientation process in media
containing azo junctions comprises trans-cis-trans photo-
isomerization cycles followed by a change of orientation of chro-
mophores toward the perpendicular direction with respect to the
polarized light, or in brief - angular hole burning [1]. It is well
known that an all-aromatic azo chromophore has higher thermal
stability than the alkyl-linked azo chromophore (eg. Disperse Red
1), which in turn informs us about improved optical stability [2].
An aromatic azo chromophore composed of an azo chromophore
better non-linear susceptibility crosslinking of the aligned mole-
cules was performed [4]. Azocarbazoles with an attached epoxide
ring have been immobilized via a surrounding matrix made of
epoxy resin. Much potential for the optical non-linearity of
carbazoles have been shown, when attaching them to multifunc-
tional inorganic chains as polysiloxane [5] or polyphosphazene [6].
However, non-centrosymmetry of the media containing the
carbazole chromophores was achieved while applying an external
electric field (corona poling) [4e7]. Despite the effectiveness to
align chromophores with a steady-state electric field the con-
tactless all-optical poling method has many advantages. The quasi-
phase matched condition for optical generation of harmonics (eg.
second), or the measurement of the time resolved dynamics of
polar media are noteworthy [8].
In this work we extend our knowledge of the non-linear optical
features of azocarbazole chromophores by investigating their
alignment dynamics and consequent orientational relaxation by
means of the contactless all-optical poling technique. We show
how the reading of polarity influences the re-orientation of the
chromophores. Investigations done with different chromophores
enables us to draw relationships between the embranchment of
molecule and it’s optical susceptibility.
* Corresponding author.
ꢀ ꢁ
E-mail address: rolandas.tomasiunas@ff.vu.lt (R. Tomasiunas).
0143-7208/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.dyepig.2011.08.004

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1205
2. Material and methods
Elemental analysis. Calcd. for C18H12N4O3 (%): C 65.06; H 3.64; N
16.86. Found (%): C 64.88; H 3.59; N 16.42.
2.1. Synthesis of chromophores
2.1.2. 9-Methyl-2-methoxy-3-(4-nitrophenyl)azo-9H-carbazole
(2, C20H16N4O3)
The synthetic procedure for azophenylcarbazole-based dyes
1e4 is summarized in Fig. 1. The synthetic strategy displays the
chemical reaction scheme that was used to synthesize the key
compound 3-(4-nitrophenyl)azo-9H-carbazol-2-ol (1) and is based
on these concepts. Thus, azocarbazole-based dye 1 was prepared by
coupling reaction of commercially available 9H-carbazol-2-ol
(Aldrich) with a diazonium salt prepared from 4-nitroaniline.
Then, the OH and NH groups of 1 were alkylated with different
halogenalkanes for better solubility to yield the desired guest
chromophore molecules 2e4, namely 9-methyl-2-methoxy-3-(4-
nitrophenyl)azo-9H-carbazole (2), 3-(4-nitrophenyl)azo-9-propyl-
2-propoxy-9H-carbazole (3) and 3-(4-nitrophenyl)azo-9-(2-ethyl)
hexyl-2-(2-ethyl)hexyloxy-9H-carbazole (4). All new azo dyes were
identified by various spectroscopy and elemental analysis.
A mixture of azo compound 1 (1.0 g, 3.0 mmol), iodomethane
(2.6 g, 18.0 mmol), 85% powdered KOH (1.2 g, 18.0 mmol), and
anhydrous Na2SO4 (0.17 g, 1.2 mmol) was stirred at room temper-
ature in DMF (6 mL) until the starting and monoalkylated inter-
mediate disappeared (24 h). After termination of the reaction (TLC)
the reaction mixture was extracted with ethyl acetate. The organic
layer was dried (MgSO4) and filtered off. Ethyl acetate and excess of
iodomethane were removed under vacuum and a red solid product
2 was obtained in 74% yield (0.8 g); m.p.: 234e236 ꢂC (recrystal-
lized from THF).
IR (KBr),
n
, cmꢁ1: 3045 (CHarom), 2958, 2925, 2854 (CHaliph), 1630
(N]N), 1515, 1335 (NO2), 854, 745 (CH]CH 1,4- and 1,2-
disubstituted benzenes).
1H NMR (300 MHz, DMSO,
d, ppm): 8.50 (s, 1H, 4-H of
2.1.1. 3-(4-Nitrophenyl)azo-9H-carbazol-2-ol (1, C18H12N4O3)
4-Nitroaniline (2.07 g, 15 mmol) was dissolved in a 12%
solution of HCI (13.5 mL) by heating till 50 ꢂC. The obtained
solution was cooled with an ice bath to bellow 5 ꢂC, then an
aqueous solution containing sodium nitrite (1.05 g, 65.4 mmol)
was slowly added to the obtained suspension with vigorous
stirring. During the addition of the solution of sodium nitrite,
the temperature of the mixture was not allowed to rise above
12e15 ꢂC. After the addition was completed the stirring was
continued for about 30 min to afford a clear solution. The
resulting diazonium salt solution was added with vigorous
stirring to the mixture of acetone (40 mL) and ethanol (25 mL)
solution of 9H-carbazol-2-ol (2.2 g, 12 mmol) at 0e5 ꢂC. Then
the mixture was allowed to reach room temperature and was
stirred for about 4 h. The separated brownish solid was filtered
off and washed with water and then with ethanol. The product
was purified by recrystallization from the mixture of THF and 2-
propanol (1:1). The azo compound 1 was obtained in 88% yield
(3.5 g); m.p.: 314e315 ꢂC (THF: 2-propanol, 1:1).
carbazole); 8.42 (d, J ¼ 9.0 Hz, 2H, of p-subst. Ph); 8.16 (d,
J ¼ 7.7 Hz 1H, 5-H of carbazole); 8.03 (d, J ¼ 9.0 Hz, 2H, of p-
subst. Ph); 7.60 (d, J ¼ 7.7 Hz, 1H, 8-H of carbazole); 7.50e7.42
(m, 1H, 7-H of carbazole); 7.38 (s, 1H, 1-H of carbazole);
7.28e7.18 (m, 1H, 6-H of carbazole); 4.15 (s, 1H, OCH3); 3.95 (s,
1H, NCH3).
Elemental analysis. Calcd. for C20H16N4O3 (%): C 66.66; H 4.48; N
15.55. Found (%): C 66.38; H 4.2; N 14.92.
2.1.3. 3-(4-Nitrophenyl)azo-9-propyl-2-propoxy-9H-carbazole (3,
C24H24N4O3)
Azo compound 3 was prepared according to the procedure
described above for 2 except that 1-bromopropane (2.21 g,
18 mmol) instead of iodomethane was used. A red solid product 3
was obtained in 80% yield (1.0 g); m.p.: 183.5e185 ꢂC.
IR (KBr),
n
, cmꢁ1: 3045 (CHarom), 2967, 2938, 2877 (CHaliph), 1630
(N]N), 1515, 1324 (NO2), 859, 740 (CH]CH 1,4- and 1,2-
disubstituted benzenes).
1H NMR (300 MHz, CDCI3,
d, ppm): 8.53 (s, 1H, 4-H of
IR (KBr),
(N]N), 1518, 1340 (NO2), 854 (CH]CH 1,4-disubstituted benzene).
1H NMR (300 MHz, DMSO,
, ppm): 12.04 (s, 1H, NH); 11.60 (s,
n
, cmꢁ1: 3385 (OH, NH), 3101, 3061 (CHarom), 1645
carbazole); 8.35 (d, J ¼ 9.0 Hz, 2H, of p-subst. Ph); 8.02 (d,
J ¼ 9.0 Hz, 2H, of p-subst. Ph); 7.49e7.34 (m, 2H, 8-H, 7-H of
carbazole); 7.28e7.21 (m, 1H, 6-H of carbazole); 6.90 (s, 1H, 1-
H of carbazole); 4.42e4.12 (m, 4H, OCH2, NCH2); 2.18e1.85
(m, 4H, CH2CH3); 1.18 (t, J ¼ 7.4 Hz, 3H, CH3); 1.03 (t, J ¼ 7.4 Hz,
CH3).
d
1H, OH); 8.54 (s, 1H, 4-H of carbazole); 8.35 (d, J ¼ 8.8 Hz, 2H, of
p-subst. Ph); 8.15e7.95 (m, 3H, 5-H of carbazole and p-subst. Ph);
7.45e7.28 (m, 2H, 7-H, 8-H of carbazole); 7.17 (t, J ¼ 7.2 Hz, 1H, 6-H
of carbazole); 6.92 (s, 1H, 1-H of carbazole).
Elemental analysis. Calcd. for C24H24N4O3 (%): C 69.21; H 5.81; N
13.45. Found (%): C 69.08; H 5.62; N 13.04.
2.1.4. 3-(4-Nitrophenyl)azo-9-(2-ethyl)hexyl-2-(2-ethyl)hexyloxy-
9H-carbazole (4, C34H44N4O3)
NO2
Azo compound 4 was prepared according to the procedure
described above for 2 except that 2-ethylhexylbromide (3.51 g,
18 mmol) instead of iodomethane was used. Product 4 was purified
by column chromatography (eluent: acetone:n-hexane ¼ 1:24).
Yield of 4 was 1.2 g (75%); m.p.: 76.5e78 ꢂC (acetone:ethanol, 1:1).
N
-
+
N
N2CI
O2N
OH
OH
N
H
N
H
1
n
, cmꢁ1: 3068 (CHarom), 2958, 2925, 2872, 2857
IR (KBr),
NO2
(CHaliph), 1628 (N]N), 1521, 1322 (NO2), 857, 743 (CH]CH 1,4- and
1,2-disubstituted benzenes).
1H NMR (300 MHz, CDCI3,
d, ppm): 8.45 (s,1H, 4-H of carbazole);
N
N
RX, KOH
2-4
8.37 (d, J ¼ 8.9 Hz, 2H, of p-subst. Ph); 7.98 (d, J ¼ 8.9 Hz, 2H, of
p-subst. Ph); 7.58e7.32 (m, 2H, 8-H, 7-H of carbazole); 7.31e7.11 (m,
2H, 1-H, 6-H of carbazole); 4.41e4.06 (m, 4H, OCH2, NCH2);
2.11e1.75 (m, 2H, CH); 1.65e1.06 (m,16H, CH(CH2)CH2CH2CH2CH3);
1.03e0.68 (m, 12H, CH3).
O
R
N
R
2
R = CH3;
3
R = CH2CH2CH3; 4 R = CH2CH(C2H5)CH2CH2CH2CH3.
Elemental analysis. Calcd. for C34H44N4O3 (%): C 73.35; H 7.97; N
10.06. Found (%): C 73.08; H 7.62; N 9.64.
Fig. 1. Synthetic procedure of azophenylcarbazole-based dyes 1e4.

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1206
2.2. Characterization of chromophores
NO2
NO2
1H NMR spectra were measured on a Varian Unity Inova spec-
trometer (300 MHz) in DMSO-d6 or CDCl3. The IR spectra of the
samples in KBr pellets were recorded on a Perkin Elmer Spectrum
GX FT-IR System spectrometer. The absorption spectra of the
sample THF solutions (10ꢁ4 M) placed in a 1 mm path length
microcell were recorded on a Perkin Elmer Lambda 35 UV/VIS
spectrometer. The course of the reactions was monitored by TLC
on Merck TLC aluminum Silica gel 60 F254 sheets and developed
with UV light. Silica gel (grade 62, 60e200 mesh, 150 Å, Aldrich)
was used for column chromatography. Elemental analyses were
performed with an Exeter Analytical CE-440 Elemental Analyzer.
Melting points were determined in capillary tubes on capillary
melting point apparatus Electrothermal MEL-TEMPÒ. Optical char-
acterization via measuring optical absorption spectra of the films
was performed using dual beam spectrophotometer Shimadzu UV-
3101PC (Fig. 2).
N
N
N
N
N
OH
OMe
N
H
N
Me
1
2
NO2
NO2
N
O
N
N
O
N
N
4
3
2.3. Preparation of films
Fig. 3. Molecular structure of azophenylcarbazole samples.
The four different azophenylcarbazoles 1e4 (Fig. 3) were
dispersed in a polycarbonate matrix for optical poling experiments.
The samples were prepared by a spin-coating technique involving
a 500 cycles per min spin for the first 6 s followed by a 2000 cycles
per min spin for the next 20 s. Silica glass plates were used as
substrates and were degreased in ethanol, washed in distilled
water and dried. Polycarbonate Iupilon Z-200 with Tg ¼ 189 ꢂC was
chosen as the polymer matrix for these molecules to create guest-
host system.
In order to avoid aggregation, the loading density of chromo-
phores in all samples was chosen as 10% (wt%). Drying via two steps
(25 min at room temperature followed up by 45 min at 80 ꢂC) was
performed for finishing.
photon of the second harmonic to the same energy level can be
achieved. The coherent pumping of the NLO polymer film results
in an orientational hole burning [8,9]. Thus, provided there is
a phase difference between these seeding beams, a poling
electric field inside the material results. The orientational hole
burning is followed by a reversible trans-cis-trans isomerization
process, which drives optical orientation motion. Finally, it
leads to a net permanent molecular polar order. This induced
molecular orientation breaks the initial centrosymmetry of the
structure inside the material creating a quasi-phase-matched
grating of non-linear second-order susceptibility applicable to
the second harmonic generation. The highest poling yield is
achieved when polarization of the seeding beams is both linear
and parallel and the probability of the material to absorb
1 photon of second harmonic frequency and two photons of
fundamental frequency is the same [8]. We used a two-branch
experimental setup [10]. The light source was a Nd:YAG laser
(1064 nm, 20 ns, 10 Hz). From the two branches frequency
2.4. Technique for optical poling
The all-optical poling technique permits polar orientation of
molecules. If the material with the initial centrosymmetry of the
structure is irradiated with fundamental
harmonic 2 beams it is able to absorb these beams, quantum
interference between two photons of fundamental and one
u and its second
doubled (2
beams. These two beams of Gaussian profile were focused on the
sample at the same spot of 200 m diameter. A pair of shutters
u) and first harmonic u beams served as seeding
u
m
was used to alter the seeding and probing processes. Optically
induced second-order susceptibility of the sample was
measured via second harmonic generation of the
During the reading process the 2 beam was blocked and only
the beam was left, so its energy and spot size were the same as
u beam.
u
u
during the seeding period. The decay of second harmonic
generation was measured either using permanent pulsed irra-
diation by the reading beam (“light”), or not permanent pulsed
irradiation (“dark”), when probing was performed by measuring
a few single laser shots only, thus, having minimal influence on
the decay process from the reading beam. No photothermal
heating of samples was observed because of low loading density
of chromophores and relatively high glass transition tempera-
ture for the polycarbonate polymer matrix used.
3. Theory and calculation
The physical origin of all-optical poling is attributed to orien-
tational hole burning and molecular re-orientation of azo-dye
chromophores in a polymer host matrix. This process can be
described by two coupled orientational diffusion equations for
Fig. 2. Spectral dependence of absorption coefficient
samples. Arrow indicates quantum energy (2.34 eV) for the second harmonic light.
a for azophenylcarbazole

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1207
photoexcitation and re-orientation of azo-dye molecules in trans
and cis states [11,14,15]:
where m01 is the transition dipole moment and Dm
difference between the dipole moments in the excited and ground
¼
m1
ꢁ
m0 is the
8
>
:
Z Z
Z Z
Z Z
ꢀ
ꢁ ꢀ ꢁ
ꢀ
ꢁ
ꢀ
ꢁ
ꢀ
ꢁ
vn ð
U
Þ
1
sc
1
>
>
<
0 þ DtP2ntð
Þ
U
t
¼
¼
ꢁ
x
Ið
q
Þntð
Þncð
U
Þ þ
Þ þ
f
F U0
/
U
I
q0 nc U0
d
U0
þ
ꢁ
G
U0
/
U
nc U0
d
U
vt
;
(1)
ꢀ
ꢁ ꢀ ꢁ
ꢀ
ꢁ
vn ð
U
Þ
>
ꢁ
f
Ið
q
U
x
Q
U0
/
U
I
q0 nt U0
d
U0
ncð
U
Þ þ DcP2ncð
U
Þ
c
vt
sc
where nt(
isomers, respectively, with dipole momentum direction in the solid
angle ). and are the quantum efficiency of trans-to-cis and
cis-to-trans photoisomerization, respectively. sc represents the
lifetime of cis isomer defined by thermal relaxation rate. G(U0
U
) and nc(
U
) are the molecular density of trans and cis
state for two-level molecule approximation of rod-like azo-dye
chromophore [8]. Parameters I1, I2, I3 correspond respectively to
U
(q
,
f
x
f
excitation terms of one-photon absorption at frequency 2
amplitude E2u), two-photon absorption at frequency
u
(field
(field
u
/
U
)
amplitude Eu) and the interference between these two terms. DJ
corresponds to the phase difference between two beams on the
is the probability for molecules to rotate from
U
0 to
U
F
in the process
of cis-to-trans thermal recovery. Q(U0
/
U
) and
(
U0
/
U
) are
incident surface of the sample. (DJ
between two beams after propagation over distance z. The term
containing I3cos3
bears polarity, which is the origin of photoin-
þ
Dkz) is the relative phase
probabilities of trans-to-cis and cis-to-trans optical transition,
respectively. Dt and Dc are thermal-induced orientational diffusion
constants for trans and cis isomers, respectively.
q
duced polar orientation [8,9].
The first term in Eq. (2) (upper) and the second term in Eq. (2)
(lower) define isomerization of trans to cis state under photoexci-
tation. Isomerization of cis to trans state describe the first term in
Eq. (2) (lower) and the second term in Eq. (2) (upper). The third and
the forth terms stand for thermal-induced relaxation of cis isomer
back to trans state and for molecular orientational diffusion,
In a simple case of rod-like molecules of the trans isomer and the
ball-like molecules of the cis isomer excited with linear polarization
of the seeding beams at frequencies
process preserves symmetry around the polarization direction. As
a consequence, the angular distributions nt( ) and nc( ) depend
only on the polar angle and can be projected on Legendre poly-
nomials Pj(cos ) by introducing expansion coefficients Tj and Cj
u and 2u, the all-optical poling
U
U
q
respectively. Further, we neglect the term I(
q
)nc(
U
) by setting
f
¼ 0,
q
because the orientational hole burning mechanism for the ball-like
cis isomer molecules is less important [1]. Thus, Eq. (2) simplifies
into:
called the Order Parameters for trans and cis isomers [11,15]:
N
X
1
2j þ 1
ntð
U
Þ ¼ ntð
q
Þ ¼
TjPjðcos
q
Þ;
(4)
8
Z Z
2
p j ¼ 0
2
ꢀ
ꢁ
ꢀ
ꢁ
vn ð
U
Þ
1
sc
>
>
<
t
¼ ꢁ
x
Ið
q
Þntð
U
Þþ
G
U0
/
U
nc U0
d
U0 þDtV2ntð
Þ
U
vt
Z Z
;
ꢀ ꢁ
ꢀ
ꢁ
vn ð
U
Þ
1
N
X
>
>
:
c
ÞI q0 nt U0
d
U
0 ꢁ ncð
U
ÞþDcV2ncð
Þ
U
1
2j þ 1
¼
x
Qð
U
0/
U
ncð
U
Þ ¼ ncð
q
Þ ¼
q
CjPjðcos Þ;
(5)
vt
sc
2
p j ¼ 0
2
(2)
The probabilities of redistribution processes G(U0
Q(U0
) depend only on the rotation angle
U0 and
, therefore we can expand them in terms of the Legendre
polynomials Pj(cos
/
U
) and
what would correspond to a transition scheme depicted in Fig. 4.
/
U
c between directions
The orientational hole burning mechanism for trans isomer in
U
all-optical poling is represented by the term I(
q)nt(
U
), where I(q) is
c) as well [11,15].
azo-dye excitation rate:
N
X
ꢀ
ꢁ
1
2m þ 1
Ið
q
Þ ¼ I1cos2
q
þ I2cos4
q
þ I3cos3
q
;
(3)
G
U0
/
U
¼ Gð
c
Þ ¼
GmPmðcos
c
Þ;
(6)
(7)
2
p m ¼ 0
2
with
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
N
X
ꢀ
ꢁ
1
2
2k þ 1
I1 a m2 E2
;
U0
/
U
¼ Qð
c
Þ ¼
QkPkðcos Þ;
c
ꢂ
Q
01
2
u
p k ¼ 0
2
ꢂ
ꢂ
ꢂ
m201Dm2
ꢂ
ꢂ
I2
I3
a
a
E4
;
ꢂ
u
ð2Z
u
Þ2
where Gm and Qk describe the shape of probability functions.
In the case of an assembly of non-interacting molecules, the
macroscopic polarizability can be theoretically obtained from the
corresponding microscopic polarizability
possible orientations, that is,
ꢂ
ꢂ
ꢂ
m201Dm
ꢂ
ꢂ
E2 E*ucosðDJ
þ
D
kzÞ ;
ꢂ
u
2
Z
u
b
after averaging over all
Z
cð2Þ
¼
nð
Þ
d ;
U b U
(8)
where n(U) is the molecular density oriented in the solid angle U.
Considering the azo-dye polymeric system dominated by rod-like
molecules of trans isomer, the induced c(2) can be expressed in
terms of the trans isomer order parameters Tj [11,14,15].
For the first time to our knowledge Eq. (2) is converted into
a system of 10 ordinary differential equations for Order Parameters
T0, T1, T2, T3, T4 of trans and C0, C1, C2, C3, C4 of cis isomers,
respectively, using Eqs. (4)e(7) for sets of orthogonal Legendre
Fig. 4. Scheme of transitions.

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1208
polynomials up to the fourth order. The system was solved using
Matlab software for the CW (continuous wave) excitation case. It
was assumed that before poling process starts all azo-dye mole-
cules are in trans isomer state and oriented isotropically. Initial
values of parameters required for numerical simulation were
chosen close to real ones obtained in our previous experiments
with azo-dye doped polymer materials [10]. The quantum effi-
After poling is stopped, a domain characterized by relatively fast
relaxation occurs, which could relate to thermal back-reaction of
cis-to-trans isomerization, and is common for all chromophores
measured (see Fig. 6). The much slower second time domain
concerns mainly the loss of induced polar order by thermal-
induced orientational diffusion [2]. The complexity of the process
people describe via multi-exponential decay [11]. In our case the
difference in the slow part of the kinetics is obvious, what features
longer lasting poled media with more branched chromophores.
To investigate the density of molecules in the trans state,
measurements at different seeding energies were performed.
Basically no difference in the magnitude of the signal was obtained,
when reading the poled media with the same energy beam (Figs. 7
and 9). It indicates that almost all the molecules in the trans state
available for excitation have performed trans-to-cis transition even
for the lowest energy of the seeding beam applied (curves c). The
result shows that depletion of molecules in the trans states is
achievable for chromophores of different embranchment (samples
4 and 2). The dynamic poled media loses it’s orientation does not
greatly depend upon the embranchment, however, some deviation
to faster decay appears for the smaller chromophore (sample
ciency of trans-to-cis photoisomerization
x
¼ 0.11 was taken from
the literature [14]. It was assumed that orientational diffusion
coefficients for the trans and cis isomer Dt and Dc, respectively, are
equal. Initial values for expansion parameters of probability
function G1 ¼ Q1 ¼ 0.85, G2 ¼ Q2 ¼ 0.61, G3 ¼ Q3 ¼ 0.38,
G4 ¼ Q4 ¼ 0.21, are defined by Boltzman-like probability distri-
bution and similar to ones obtained in literature [16].
4. Results
Four different azophenylcarbazole-based chromophores in
a polycarbonate matrix were investigated, though only three of
them (2, 3 and 4) showed an optical poling effect. No measurable
second harmonic signal was obtained for chromophore 1 (donor
termination by H and OH). Poling graphs for the three samples
using the same excitation condition (pulse energy) for seeding and
reading are shown in Fig. 5. Independently from the embranch-
ment of molecules poling proceeds in a quite similar way and
reaches saturation after w10 min. The saturation informs us about
some quasi-stationary situation, when the balance between poling
and depoling processes occurs, i.e. trans-to-cis transition rate
approaches the decay rate of poled media defined by re-
orientational diffusion of molecules in the trans and cis states,
thermal and photoinduced cis-to-trans transition. The efficiency to
generate a second harmonic strongly differs and depends on the
embranchment of molecules as expected. The highest value for the
most branched chromophore 4 obtained may be explained by
larger partial volume and dipole moment of the molecule consid-
ering higher polarizability. For the smallest chromophore 1 the
polarizability was probably too low to detect a second harmonic
signal. Reabsorption of the second harmonic radiation shouldn’t be
considered since the poling efficiency goes in the opposite direction
as the absorbance at 2.34 eV for chromophores investigated (see
Fig. 2).
2) under the highest excitation condition Iu
¼
2.8 J/cm2,
I2 ¼ 0.45 mJ/cm2, what may refer to increased flexibility due to
u
heating the media (compare Figs. 8 and 10).
Photoinduced orientational relaxation of the poled media was
investigated by measuring the decay in “light” and “dark”
regimes. The reading beam induces more rapid decay, i.e. effec-
tive ”washing-out” of the orientation alignment of molecules
(Figs. 11e13) for all chromophores investigated.
To evaluate the macroscopic second-order optical non-linearity
of the material a comparison of second-order non-linear suscepti-
bility c(2) between azophenylcarbazole and z-cut quartz plate was
performed using equation [12]:
sffiffiffiffiffiffi
u
cð2Þ
cðq2Þ
2 lc;q I2
¼
;
(9)
u
Iq2
p
d
where the coherence length lc,q
refractive index n are parameters for quartz and d is the thickness of
¼
lu/(4(nq(2 ꢁ nq(u))) and the
u)
the azophenylcarbazole film.
u and 2u denote the fundamental and
u
the second harmonic beam. I2 and Iq2 are the second harmonic
u
Fig. 5. Second harmonic radiation generated by optically poled azophenylcarbazole of
different embranchment (seeding beam: I1 - 1.82 J/cm2, I2 - 0.16 mJ/cm2; reading
Fig. 6. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole of different embranchment (seeding beam: I1 - 1.82 J/cm2, I2u - 0.16 mJ/
u
u
u
beam: I1u - 1.82 J/cm2) (“dark” regime for reading process). Lines are modeling results
cm2; reading beam: I1 - 1.82 J/cm2) (“dark” regime for reading process) (normalized
u
according Eq. (2) explained in Section 3.
signal). Lines are modeling results according Eq. (2) explained in Section 3.

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1209
a
a
b
c
d
b
c
Fig. 7. Second harmonic radiation generated by optically poled azophenylcarbazole
sample 4 for different seeding beam values: a) I1u - 2.8 J/cm2, I2u - 1.4 mJ/cm2; b) I1
Fig. 9. Second harmonic radiation generated by optically poled azophenylcarbazole
-
u
sample 2 for different seeding beam values: a) I1u - 2.8 J/cm2, I2u - 0.45 mJ/cm2; b) I1
-
-
1.82 J/cm2, I2 - 1.2 mJ/cm2; c) I1 - 1.5 J/cm2, I2 - 0.9 mJ/cm2; d) illustration of non-
u
u
u
u
u
1.82 J/cm2, I2 - 0.16 mJ/cm2; c) I1 - 1.5 J/cm2, I2 - 0.1 mJ/cm2; (reading beam: I1
optimal beam polarization case, conditions same as b) (reading beam: I1 - 1.8 J/cm2)
u
u
u
u
1.82 J/cm2) (“dark” regime for reading process).
(“dark” regime for reading process).
using AFM) best featuring optical poling the obtained
c
intensities measured from azophenylcarbazole film and quartz
plate, respectively. The values of cðq2Þ for quartz measured over the
past two decades had a wide range from 1.0 to 0.6 pm/V [13]. The
latter value claimed as more accurate was used in our calculation
ð2Þx0:72 pm=V is only slightly higher than that of z-cut quartz.
The theoretical model developed was used to extract parameters
for azophenylcarbazole material via best fit of the experimental data.
It’s important to note that comparison between “dark” and “light”
for z-cut quartz plate. For sample 4 (film thickness 1.5 mm measured
Fig. 8. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole sample 4 for different seeding beam values (reading beam: I1u - 1.8 J/cm2)
(“dark” regime for reading process) (normalized signal).
Fig. 10. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole sample 2 for different seeding beam values (reading beam: I1 - 1.82 J/
u
cm2) (“dark” regime for reading process) (normalized signal).

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1210
Fig. 11. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole sample 4 for different reading regimes (“dark” and “light”) (seeding
beam: I1 - 1.8 J/cm2, I2 - 1.4 mJ/cm2; reading beam: I1 - 1.8 J/cm2) (normalized
Fig. 13. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole sample 2 for different reading regimes (“dark” and “light”) (seeding
beam: I1 - 1.8 J/cm2, I2 - 1.4 mJ/cm2; reading beam: I1 - 1.8 J/cm2) (normalized
u
u
u
signal). Lines are modeling results according Eq. (2) explained in Section 3.
u
u
u
signal). Lines are modeling results according Eq. (2) explained in Section 3.
reading regimes allowed us to evaluate relative one- and two-photon
excitation rates for these carbazole based chromophores in poly-
carbonate. Parameters obtained from fitting the data in Figs.11e13 are
presented in Table 1. The same parameters also align well with the
poling kinetics measured using a larger pulse energy ratio between
seeding beams (see Figs. 5 and 6). We obtained the relationship that
the larger the chromophore size (material 4) the smaller the value of
it’s orientational diffusion coefficient Dt.
Probably, too rapid orientational diffusion and back-reaction of cis-
to-trans isomerization in the relative fluid substance has defined the
very fast decay of poled media. Following this disappointing result
the search for new host matrices are in progress.
5. Discussion
To investigate how thermal-induced orientational diffusion of
azophenylcarbazole chromophores and the trans-to-cis photo-
isomerization are dependent on the guest-host system and
whether it is possible to fix better the poled media created in poly-
carbonate, different polymer host matrices were used. By use of
photopolymerization of organically modified silica (ormosil)
and commercially available inorganic-organic hybrid polymers
(ORMOCER) it should be possible to fix the oriented media. Experi-
ments were performed with chromophore 4 which showed the
greatest optical poling inpolycarbonate. However, without a success;
no signal indicating poling condition in the matrices was obtained.
Experimental investigation revealed accordance between the
strength of optical absorbance and efficiency of optical poling for
azophenylcarbazoles in polycarbonate with the most significant
effect noted for the chromophore with the highest branched donor
(see Figs. 2 and 5).
The experiment and theoretical modeling disclose polar order
relaxation kinetics of carbazoles roughly split into two time periods,
short and long. The long term period (over 30 min) is fully charac-
terized by only trans isomer orientational diffusion coefficient Dt and
this coincides well with previous data [8,10,11]. For the short time
period (up to 30 min) the relaxation kinetics of polar order is defined
by many parameters of the azo-polymer system. First of all, the
importance of the parameters Gm and Qk standing for probabilities of
redistribution process, which was mainly not noted in previous
works. In Ref. [8], it was assumed that G (U0
/
U
) ¼ 1/4 and G0 ¼ 1,
p
G1 ¼ G2 ¼ G3 ¼ G4 ¼ 0. This implies that after excitation, each chro-
mophore may rotate from its initial direction by any angle in any
azimuthal direction and with the same probability. In Refs. [11,15], it
was assumed that G1 ꢀ Q1 << 1 and all expansion coefficients are
small and does not influence the relaxation kinetics. Rather strong
influence on the fitting procedure for the short time scale was
observed, when the finite values of Gm and Qk were taken into
account. Thus, the Gm values obtained demonstrate the trans-to-cis
thermal recovery input in orientational rotation, however, without
significant difference for our samples.
Table 1
Parameters used for modeling the experimental optical poling kinetics for ”light”
and ”dark” reading regimes of azocarbazole chromophores dispersed in
polycarbonate.
Material
I1 (sꢁ1
)
I2 (sꢁ1
)
G1
Dt (sꢁ1
)
Dc (sꢁ1
)
sc (s)
Fig. 12. Decay of second harmonic radiation generated by optically poled azophe-
nylcarbazole sample 3 for different reading regimes (“dark” and “light”) (seeding
beam: I1 - 1.8 J/cm2, I2 - 1.1 mJ/cm2; reading beam: I1 - 1.8 J/cm2) (normalized
4
3
2
4 ꢀ 10ꢁ5
4 ꢀ 10ꢁ5
4 ꢀ 10ꢁ5
2 ꢀ 10ꢁ3
2 ꢀ 10ꢁ3
1 ꢀ 10ꢁ3
0.85
0.8
0.85
2.2 ꢀ 10ꢁ5
4 ꢀ 10ꢁ5
8 ꢀ 10ꢁ5
2.2 ꢀ 10ꢁ5
4 ꢀ 10ꢁ5
8 ꢀ 10ꢁ5
143
167
167
u
u
u
signal). Lines are modeling results according Eq. (2) explained in Section 3.

_
G. Navickaite et al. / Dyes and Pigments 92 (2012) 1204e1211
1211
It is defined that the read-out of a poled structure by a high
intensity fundamental wave induces additional decay of the
recorded micro-pattern. The physics behind rests on different
excitation condition while seeding or reading. At the time of
(different donor) defines their orientational mobility. Thus, the
larger the chromophore the weaker the orientational diffusion
(diffusion coefficient vary from 2.2 ꢀ 10ꢁ5 sꢁ1 to 8 ꢀ 10ꢁ5 sꢁ1
)
was obtained. The highest optical poling efficiency was measured
for large molecules, however, photoinduced erasure while
reading the poled media was more pronounced for less devel-
oped structures (weakly branched donor). The second-order
non-linear optical susceptibility for azophenylcarbazoles evalu-
reading using one beam (u) no resulting poled electric field is
provided. Alternating electric field via hole burning affects all the
molecules, independently of their transition dipole moment
orientation, including those responsible for the poled media and
quasi-phase matched condition, thus, stimulating the erasure
effect. From the modeling point of view it appears via the second
term in Eq. (3), experimentally - from Figs. 11e13.
ated ranges to
c
ð2Þx0:72 pm=V. Modeling revealed that expan-
sion parameters up to the fourth order for probability functions
of redistribution process for both trans and cis states have to be
taken into account, when interpreting the experimental results.
Among different polymer matrices investigated only poly-
carbonate kept azocarbazole molecules relatively oriented.
Another important fact revealed concerns the relaxation kinetic
dependence on the history of seeding process. There is some
“memory” effect related to seeding conditions, i.e. the pulse energy
(first and second harmonic) and the seeding time. For different
seeding conditions we obtain different relaxation kinetics of polar
order (see Fig. 8). This fact could be interpreted using our all-optical
poling model for trans and cis isomer components. The generated
density of cis isomer component, which strongly depends on
seeding conditions, even in full “dark” free relaxation regime has
a back-reaction to trans isomer. Cis to trans thermal relaxation
created an additional source of polar order for trans isomer and this
process changes considerably the decay kinetic. The short term
relaxation time starts to depend on generated density of the cis
isomer, meanwhile the long term relaxation time of polar order is
defined only by orientational diffusion constant of the trans isomer.
This process is described by second term in Eq. (2).
Acknowledgments
This work was partially funded by a grant (No. AUT-06/2010)
from the Research Council of Lithuania, by a grant from the Lithu-
anian State Science and Studies Foundation (project Mulatas 2), by
the Agency of International Science and Technology Development
in Lithuania and by the EU COST Action MP0702, MP0604. The
ꢀ
_
authors wish to express their appreciation to G.Medeisiene for
ꢀ
_
sample preparation and L.Kucinskaite for numerical simulations.
References
In comparison to other dyes azophenylcarbazoles show relative
low value of second-order non-linear susceptibility
c
(2). Several
[1] Sekkat Z, Dumont M. Photoinduced orientation of azo dyes in polymeric films -
characterization of molecular angular mobility. Synth Met 1993;54:373e81.
[2] Verbiest T, Burland DM, Jurich MC, Lee VY, Miller RD, Volksen W. Excep-
tionally thermally stable polyimides for 2nd-order nonlinear-optical appli-
cations. Science 1995;268:1604e6.
[3] Chen JP, Labarthet FL, Natansohn A, Rochon P. Highly stable optically induced
birefringence and holographic surface gratings on a new azocarbazole-based
polyimide. Macromolecules 1999;32:8572e9.
[4] Niziol J, Pielichowski J. Usability of epoxy resins in conjunction with carbazole
dyes in non-linear optics applications. Opt Mater 2010;32:673e6.
[5] Li J, Qin JG, Ye C. Three series of multifunctional polysiloxane attached with
charge-transporting agents and electro-optical chromophores. Synth Met
2005;152:305e8.
[6] Zhang L, Huang MM, Jiang ZW, Yang Z, Chen ZJ, Gong QH, et al. A carbazole-
based photorefractive polyphosphazene prepared via post-azo-coupling
reaction. React Funct Polym 2006;66:1404e10.
aspects are important to explain. The
c
(2) basically depends on two
parameters: the molar concentration and the first-order hyper-
polarizability (Eq. (8)). The molar concentration we used (10% (wt
%)) corresponds well to the mid of range used for carbazoles (2% -
[17], 5% - [4], 10% - [6], 38% - [18] depending on host material). From
the quantum-mechanical two-level model following expression for
the first-order hyperpolarizability is valid [19]:
2
p2Dm
ð
m2eg
Þ
3
u2eg
ꢀ
ꢁꢀ
ꢁ
b
¼
;
(10)
3
0h2
u2eg ꢁ 4u2 u2eg
ꢁ
u2
where Dm
¼
me
ꢁ
mg is the difference between the excited and
[7] Qin Z, Fang C, Pan Q, Gu Q, Chen F, Li F, et al. Optical properties of NAEC-
PMMA nonlinear polymeric thin film. J Mater Sci 2002;37:4849e52.
[8] Fiorini C, Charra F, Nunzi JM, Raimond P. Quasi-permanent all-optical encoding of
noncentrosymmetry in azo-dye polymers. J Opt Soc Am B 1997;14:1984e2003.
[9] Dumont M. Dynamics of all-optical poling of photoisomerizable molecules. I.
Symmetries of tensorial properties. J Opt Soc Am B 2009;26:1057e75.
ground state dipole moment, meg is the transition dipole moment
between the ground and excited state, 0 is the vacuum permittivity
3
and ueg is the transition frequency. meg is related to the overall
strength of the transition and according to general quantum
mechanics can be determined from the area under absorption
band. This is quite well represented in our case for azophe-
nylcarbazoles: the ten times difference in effect of optical poling
could be attributed to the strength of transition, what basically
depends on donor architecture (Figs. 2 and 5). It is most probable
that both, the dipole moment and the transition dipole moment,
ꢀ
ꢀ
ꢀ
ꢀ
[10] Seniutinas G, LaipnieceL, Kreicberga J, Kampars V, Grazulevicius J, Petruskevicius R,
et al. Orientational relaxation of three different dendrimers in polycarbonate
matrix investigated by optical poling. J Opt A Pure Appl Opt 2009;11:034003.
[11] Liu X, Xu G, Si J, Ye P, Li Z, Shen Y. Erasure effect of the reading beam on the
decay process of c(2) in all-optical poling. Appl Phys B 2000;71:539e43.
[12] Sahraoui B, Luc J, Meghea A, Czaplicki R, Fillaut JL, Migalska-Zalas A. Nonlinear
optics and surface relief gratings in alkynyleruthenium complexes. J Opt A
Pure Appl Opt 2009;11:024005.
are responsible for the relative low
c
(2). Same for the screening
[13] Flueraru C, Grover CP. Relative measurements of second-order susceptibility
with reflective second-harmonic generation. Appl Opt 2003;42:6666e71.
[14] Sekkat Z, Knoll W. Creation of second-order nonlinear optical effects by
photoisomerization of polar azo dyes in polymeric films: theoretical study of
steady-state and transient properties. J Opt Soc Am B 1995;12:1855e67.
[15] Xu G, Liu XC, Si JH, Ye PX, Li Z, Shen YQ. Modified theory of photoinduced
molecular polar alignment in azo polymers. Opt Lett 2000;25:329e31.
[16] Singer KD, Kuzyk MG, Sohn JE. Second-order nonlinear-optical processes in
orientationally ordered materials: relationship between molecular and
macroscopic properties. J Opt Soc Am B 1987;4:968e76.
[17] Angiolini L, Benelli T, Giorgini L, Mauriello F, Salatelli E. Synthesis of optically
active photoresponsive multifunctional polymer containing the side-chain
azocarbazole chromophore. Macromol Chem Phys 2006;207:1805e13.
[18] Cho MJ, Choi DH, Sullivan PA, Akelaitis AJP, Dalton LR. Recent progress in
second-order nonlinear optical polymers and dendrimers. Prog Polym Sci
2008;33:1013e58.
effect provided by photoinduced cis isomer with typically strong
donoreacceptor interaction, which affects (lowers) the non-
centrosymmetry of the optically poled material. It is worth
emphasizing that all-optical poling revealed azophenylcarbazole as
material suitable for studying molecule orientation dynamics,
other photoinduced phenomena and having potential in applica-
tion for optical storage.
6. Conclusions
The ability of azophenylcarbazoles to create a media featuring
non-centrosymmetry was demonstrated. All-optical poling
investigations revealed that the architecture of molecules
[19] Pan Q, Fang C, Li F, Zhang Z, Qin Z, Wu X, et al. Thermally stable chromophores
for nonlinear optical applications. Mater Res Bull 2002;37:523e31.