Simpler and more efficient strategy to stabilize the chromophore orientation in electro-optic polymers with copper-free thermal Huisgen reaction

Article abstract of DOI:10.1016/j.polymer.2011.02.042

A new strategy is proposed to stabilize the electro-optic (EO) activity of second-order materials using copper-free thermal Huisgen 1,3-dipolar cross-linking reaction. It consists in freezing the chromophores orientation after the poling process by a cross-linking reaction based on the 1,3-dipolar cycloaddition between an azide and an alkyne. To reach this goal, the synthesis of new methacrylate type polymers bearing a derivative of Disperse Red 1 chromophore was performed. The polymeric structure is bearing a cross-linkable function on its backbone and the complementary reactive function is brought by a small molecule called "doping agent" (DA), containing several complementary cross-linking groups, evenly distributed in the polymer film. Materials have been prepared and exhibit large second-order nonlinear optical coefficients (d₃₃) up to 60 pm/V at the fundamental wavelength of 1064 nm. Moreover, the thermal stability of the orientation of the chromophores could reach 150 °C upon cross-linking with such materials, which is higher than previously described cross-linkable EO polymers based on this reaction. Furthermore, this new strategy widens the possibilities offered by copper-free thermal Huisgen 1,3-dipolar cycloaddition as cross-linking reaction for EO polymers.

Full text of DOI:10.1016/j.polymer.2011.02.042

Polymer 52 (2011) 2286e2294
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Simpler and more efficient strategy to stabilize the chromophore orientation in
electro-optic polymers with copper-free thermal Huisgen reaction
,
1
Clément Cabanetos a, Errol Blart a, Yann Pellegrin a, Véronique Montembault b, Laurent Fontaine b
,
,
,
*
Frédéric Adamietz c, Vincent Rodriguez c 2, Fabrice Odobel a
a Université de Nantes, CNRS, Chimie et Interdisciplinarité: Synthèse, Analyse, Modélisation (CEISAM), UMR CNRS nꢀ 6230, 2, rue de la Houssinière, BP 92208,
44322 Nantes Cedex 3, France
b LCOM-Chimie des Polymères, UCO2M, UMR CNRS 6011, Université du Maine, Avenue O. Messiaen, 72085 Le Mans Cedex 9, France
c Université de Bordeaux, Institut des Sciences Moléculaires, UMR 5255 CNRS, 351 cours de la Libération, 33405 Talence Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
A new strategy is proposed to stabilize the electro-optic (EO) activity of second-order materials using
copper-free thermal Huisgen 1,3-dipolar cross-linking reaction. It consists in freezing the chromophores
orientation after the poling process by a cross-linking reaction based on the 1,3-dipolar cycloaddition
between an azide and an alkyne. To reach this goal, the synthesis of new methacrylate type polymers
bearing a derivative of Disperse Red 1 chromophore was performed. The polymeric structure is bearing
a cross-linkable function on its backbone and the complementary reactive function is brought by a small
molecule called “doping agent” (DA), containing several complementary cross-linking groups, evenly
distributed in the polymer film. Materials have been prepared and exhibit large second-order nonlinear
optical coefficients (d33) up to 60 pm/V at the fundamental wavelength of 1064 nm. Moreover, the
thermal stability of the orientation of the chromophores could reach 150 ꢀC upon cross-linking with such
materials, which is higher than previously described cross-linkable EO polymers based on this reaction.
Furthermore, this new strategy widens the possibilities offered by copper-free thermal Huisgen
1,3-dipolar cycloaddition as cross-linking reaction for EO polymers.
Received 6 January 2011
Received in revised form
9 February 2011
Accepted 24 February 2011
Available online 10 March 2011
Keywords:
Non linear optic
Cross-linking
Huisgen reaction
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
into thick and good optical quality films, their incorporation in
a polymer matrix is most often a prerequisite [8,9,11]. The chromo-
For the last past two decades, the development of organic
nonlinear optical (NLO) materials has generated a huge interest
owing to their potential applications in optical switching, informa-
tion storage or signal processing [1e4]. Compared to NLO inorganic
crystals, such as lithium niobate, organic NLO materials display faster
response times, larger bandwidths, higher electro-optic coefficients
and better processability [3,5,6]. Accordingly, polymeric materials
represent attractive candidates for the fabrication of electro-optic
devices, such as Mach Zendher interferometer [7e10]. Organic NLO
materials are generally composed of push-pull organic chromo-
phores are finallyoriented byapplying a high external electric fieldto
generate their noncentrosymmetric organization inside the matrix,
which is a key requirement to observe the bulk electro-optic activity.
The polymeric electro-optic material must also fulfill other require-
ments such as a high chemical, thermal and photochemical stability
and a high transparency at the laser wavelength, generally 1.55 mm
for telecommunication. However, the temporal instability of the
poled chromophores is certainly the most stringent limitation of the
organic electro-optic materials for their industrial utilization [12,13].
Indeed, in most organic materials, the chromophores tend to relax
and adopt a head-to-tail organization due to their high ground-state
dipole moment. As a result, the development of strategies to
permanently freeze the chromophores orientation after the poling
process is of a paramount practical significance for the applications
[13]. The development of new host matrices allowing a high stability
of the poled chromophores is certainly equally important as the
synthesis of new chromophores possessing large hyperpolarizability
phores, inwhich a p-conjugated bridge is end-cappedbya donorand
an acceptor moiety. For the practical processing of the chromophores
* Corresponding author. Tel.: þ33 (0)2 51 12 54 29; fax: þ33 (0)2 51 12 54 02.
E-mail addresses: laurent.fontaine@univ-lemans.fr (L. Fontaine), v.rodriguez@
ism.u-bordeaux1.fr (V. Rodriguez), Fabrice.Odobel@univ-nantes.fr (F. Odobel).
1
coefficients (b). The two most investigated strategies to prevent the
Tel.: þ33 (0)2 43 83 33 25; fax: þ33 (0)2 43 83 37 54.
2
chromophore relaxation are firstly to embed the chromophores into
Tel.: þ33 (0)5 40 00 69 71; fax: þ33 (0)5 40 00 84 02.
0032-3861/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2011.02.042

C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
2287
a high glass transition temperature (Tg) polymer and secondly to use
cross-linkable matrices. High glass transition temperature polymer
maintains the organization of the chromophores owing to the high
viscosity of the medium even at significant temperature [14e17],
while cross-linkable systems enable to freeze the chromophores
orientation by the formation of new covalent bonds inside the
material after the poling process. Cross-linkable NLO polymers can
be classified into two categories. The first one is generally based on
a condensation reaction conducted in two steps with monomers
substituted by multiple polymerization groups [18e22] while the
second is based on specific cross-linking groups which are distinct
from the polymerization groups, which only react above Tg. This
second strategy is much more convenient and more reproducible
thanthefirstonewhichdemands averydifficultcontrolof thecourse
of the condensation reaction. However, there are few cross-linking
reactions which fulfill the criteria to be used for such application.
First, the reacting groups need to be dormant below the glass tran-
sition temperature (Tg), but they must react quickly just above the Tg.
Second, they must be compatible with the polymerization reaction
so as the cross-linking reaction does not start during the preparation
of the polymer. Third, the reacting groups should be quite selective to
one another and should not react with the chromophores to prevent
their degradation. The very few cross-linking reactions successfully
implemented to electro-optic polymers are: the anthracene-mal-
eimide DielseAlder cycloaddition [7,23e25], the cyclodimerization
of trifluorovinylether [26e28], the thermal decomposition of cyclo-
butanone [29] and the epoxide opening with a carboxylic acid
[30e33]. However, someof themarenotfullysatisfying. Forexample,
the DielseAlder cycloaddition is a reversible reaction and the dien-
ophile tends to react with the double bonds of the chromophore
[12,23]. The shelf-storage of the systems employing an epoxide and
a carboxylic acid is modest owing to a premature cross-linking
occurring at room temperature and the cyclodimerization of tri-
fluorovinylether requires high temperature, which may not be
compatible with all polymer or chromophore thermal stabilities,
besides at high temperature the poling efficiency is decreased. Stable
thermal and temporal electro-optic materials were also designed by
layer-by-layer self-assembly methods or by supramolecular inter-
actions [22,34e43]. Although these materials exhibit exceptional
stabilities they are expensive to fabricate and/or these strategies are
not fully compatible with industrial developments. Recently, we
reported a new cross-linkable strategy based on copper-free thermal
Huisgen 1,3-dipolar reaction [44,45]. In the previous systems, one
cross-linkable group (azide) was borne by the chromophore and the
complementary group (acetylenic group) was placed on the polymer
backbone as a pendant group (PAS1 Fig. 1, and system A Fig. 2).
Despite the relatively high thermal stability of the electro-optic
activity of these new polymers, acquired thanks to the cross-linking
process, they present two potential limitations. First, we observed
a premature cross-linking process upon shelf-storage at room
temperature for several months. This could be cumbersome, since
even at a low percentage of cross-linking reaction, the initially high
solubility of the polymer could significantly diminish and prevents
the preparation of concentrated solution for spin-coating. Secondly,
the above strategy requires the double functionalization of the
chromophore, one for the attachment of the polymerization group
(methacrylatein thepresentcase)and the other for the cross-linking
group (azido or alkyne, Figs. 1 and 2). The introduction of two
functional groups could represent a tedious and costlysynthetic task
with powerful chromophores, whose synthesis is, itself, often
challenging [12,23]. A modification of our initial strategy is proposed
herein, which overcomes the above limitations. It is based on the
utilization of a complementary small molecule containing several
cross-linking groups, called doping agent (DA), which is mixed with
the electro-optic polymer just before the preparation of the film
(Fig. 2). Thus, the direct contact of the two complementary cross-
linkable groups is avoided during storage, since it occurs only in the
final stepof the process during the preparationof the solution for the
spin-casting of the two materials. Besides, with this new formula-
tion, the introduction ofa cross-linking group on the chromophore is
not strictly necessary. However, a system bearing cross-linkable
group both on the polymer and on the chromophore (system D) was
prepared to test its relative electro-optic stability compared to the
other materials (systems A, B and C; Fig. 2). The concept, based on the
mixing of a complementary cross-linking agent, was already
reported by Jen and coworkers, but using DielseAlder cycloaddition
as cross-linking reaction and proved to be successful [7,25,46,47].
In this paper, we describe the synthesis and the characteriza-
tions of three new polymeric matrices and doping agents (Schemes
1 and 2). The electro-optic stability of the mixtures corresponding
to the systems C and D is at least as high as that obtained with initial
system A, demonstrating the pertinence of this new strategy and
enlarging the scope of the copper-free Huisgen cycloaddition as
cross-linking reaction to stabilize the macroscopic electro-optic
activity of organic materials.
2. Experimental part
2.1. Materials
4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methyl-morpholinium
chloride DMTMM [48], compounds 1 [44], 3 [45], 5 [49], 6 [50], 10
[51], DA1 [52] and DA2 [53] were prepared according to the
procedures described in the literature.
2.2. Methods
NMR spectra were recorded using a Bruker 300 MHz instrument
(ARX-300, Bruker) with tetramethylsilane (TMS) as the internal
standard and CDCl3 as the solvent. The IR spectra were recorded
using an FT-IR spectrometer (Clark-MXR CPA). UVeVisible absorp-
tion spectra were recorded on a UV-2401PC Shimadzu spectropho-
tometer. Thermal analyses were performed using a TA instruments
Q100 in a nitrogen atmosphere at a heating rate of 10 ꢀC/min. The
number- and weight-average molecular weights and molecular
weight distributions were determined using a size exclusion chro-
matography (SEC) on a system equipped with a SpectraSYSTEM AS
1000 autosampler, with a guard column (Polymer Laboratories, PL
gel 5
mm guard column) followed by two columns (Polymer Labo-
ratories (PL), 2 PL gel 5
mm MIXED-D columns), with a Spec-
traSYSTEM RI-150 detector. Tetrahydrofuran (THF) was used as the
Fig. 1. Illustration of a previously studied cross-linkable EO polymer (PAS1).

2288
C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
Fig. 2. Illustration of the cross-linking strategies based on copper-free Huisgen cycloaddition.
mobile phase at a flow rate of 1 mL/min at 35 ꢀC. Polystyrene stan-
dards (580e483.103 g molꢁ1) were used to calibrate the SEC. High
resolution electro-spray mass spectra (HR-ESMS) were collected in
positive mode on an MS/MS ZABSpec TOF of Micromass equipped
with a geometry EBE TOF. The samples were injected in DCM.
Column chromatography was carried out with Merck 5735 Kieselgel
60F (0.040e0.063 mm mesh). Air sensitive reactions were carried
out under argon in dry solvents. Second harmonic generation (SHG)
measurements were performed using the optical setup described in
a previous study [54]. Polarized SHG Maker fringe patterns were
recorded before and, from time to time, after the poling process,
using a 1064 nm Nd:YAG laser operating at very low irradiance
(DMTMM), (0.78 mg, 0.281 mmol, 1.2 equiv.). The solution was
stirred one night at room temperature. Solvents were then evap-
orated under reduced pressure to give the crude product, which
was purified by silica gel column chromatography using dichloro-
methane as eluting system leading to 101 mg of compound 2 (96%).
1H NMR (300 MHz, CDCl3, MeOD),
d
(ppm): 8.48 (d,1H, J ¼ 1.8Hz);
8.25 (dd, 1H, J ¼ 8.4 Hz, J ¼ 2.1 Hz); 7.89 (d, 2H, J ¼ 9 Hz); 7.77 (d, 1H,
J ¼ 8.4 Hz); 6.81 (d, 2H, J ¼ 9 Hz); 6.10 (s,1H); 5.59 (s,1H); 4.38 (t, 2H,
J ¼ 6.6 Hz); 3.97 (s, 3H,); 3.75 (q, 2H, J ¼ 6 Hz); 3.53 (q, 2H, J ¼ 6.9 Hz);
1.94 (s, 3H); 1.25 (t, 3H, J ¼ 7.2 Hz).13C NMR (75 MHz, CDCl3),
d (ppm):
173.49; 167.26; 164.90; 151.70; 148.37; 146.82; 144.09; 135.77;
133.41; 129.85; 126.91; 126.27; 125.26; 118.64; 111.46; 61.59; 55,32;
52.67; 48.72; 45.63; 18.29; 11.22. HRMS-ESI: m/z calcd for
C22H24N4O6: 440.1696, found: 441.1774(MHþ). FT-IR (KBr, cmꢁ1):
(pulse energy <20 mJ; repetition rate 50 Hz; pulse width 15 ns). A
general SHG matrix method applicable to multilayered anisotropic
absorbing linear/nonlinear media has been applied, allowing
experimental determination of the resonance-enhanced NLO coef-
ficients dij, as well as the linear absorption coefficients of the
2954 (nst(CH2)); 1727 (nst(C
]O)); 1602, 1563 (nst(C]C)). UVeVis.
(CH2Cl2): lmax
(e
(molꢁ1 L cmꢁ1)) ¼ 478 (32 600).
harmonic wave (532 nm) in the parallel (a//) and perpendicular (
at)
2.3.2. DA3
directions with respect to the poling field. Thus, absorption of the
harmonic wave was explicitly taken into account and SHG coeffi-
cients were determined using the quartz reference with coefficient
d11 ¼ 0.3 pm/V at 1064 nm. Further details about the general
procedure to determine the linear and nonlinear optical constants
can be found elsewhere [54,55]. The indexes of refraction have been
assumed isotropic in accordance to the weak thicknesses of all
To a suspension of 2,4,6-trichloro-1,3,5-triazine (0.5 g,
2.74 mmol, 1 equiv.) in THF was added but-2-yn-1-ol (1.5 g,
27.4 mmol, 10 equiv.) and K2CO3 (1.2 g, 8.1 mmol, 3 equiv.). The
reaction mixture was stirred at 60 ꢀC overnight. Solvent was
removed and the crude was dissolved in 80 mL of dichloromethane.
Acetic acid and water were added and the two layers separated. The
organic one was washed with brine, dried over MgSO4 to give the
desired compound after evaporation of the solvents (1.7 g, 60%).
samples. Onlyanisotropic linear absorption coefficients (a// and at)
of the harmonic wave (close to resonance) have been considered by
measuring the absorbance maxima before (A0) and after poling (A).
Therefore, assuming no degradation of the chromophores during
1H NMR (300 MHz, CDCl3),
13C NMR (75 MHz, CDCl3),
(ppm): 172.9; 73.05; 57.14; 4.12. EI-MS:
m/z calcd for C15H15N3O3: 285.11 found: 286.11 (MHþ). FT-IR (KBr,
d (ppm): 5.01 (m, 6H); 1.82 (s, 9H).
d
the poling process, we have A0 ¼ <
a
> ¼ (2
a
tþa//)/3 before poling,
cmꢁ1): 2242 (
nst(C^C)); 1571 (nst(C]C)); 1333 (nst(C]N)). TGA-DSC
and, after poling we obtain directly the in-plane absorption term
a^ ¼ A. Finally, it is easy to determine the last anisotropic absorption
coefficient a//¼ 3A0ꢁ2A.
(10 ꢀC/min): Td ¼ 242 ꢀC.
2.3.3. DA4
In a 10 mL round bottom flask, a solution of 10 (0.25 g,
0.86 mmol, 1 equiv.) in 2 mL of TFA was cooled down to 0 ꢀC. A
mixture of sodium nitrite (0.235 mg, 3.44 mmol, 4 equiv.) in 0.5 mL
of water was added slowly. The solution was stirred at 0 ꢀC for
15 min. Then, a solution of sodium azide (0.220 mg, 3.44 mmol, 4
equiv.) in 0.5 mL of water was added. The reaction mixture was
stirred at 0 ꢀC during 1 h and warmed up to room temperature. Iced
water was added to dichloromethane, the two layers were
2.3. Synthesis of the materials
2.3.1. Compound 2
To
a solution of carboxylic acid compound 1 (100 mg,
0.235 mmol, 1 equiv.) and MeOH (V ¼ 5 mL) in THF (V ¼ 5 mL) was
added N-methylmorpholine (V ¼ 0.07 mL, 0.705 mmol, 3 equiv.)
and then dimethoxytriazine-N-methylmorpholine chloride

C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
2289
NH2.HCl
O
O
3
O
O
O
N
7
O
O
N
N
O
O
O
N3
N3
4
N
N3
AIBN
THF, 70°C
DMTMM, NMM
THF, r.t.
N
O
N
N
N
N
NO2
NO2
P3
O
NO2
NH
NH
1
O
OH
N3
3
N3
DMTMM, NMM
MeOH, r.t.
3
7
O
O
O
O
O
O
TMS
N
O
TMS
N
N
O
AIBN
P1
THF, 70°C
O
N
NO2
O
H3C
N
N
NO2
3
7
O
O
O
O
O
O
CH3
O
6
N
N3
O
N3
N
N
2
AIBN
THF, 70°C
P2
O
O
NO2
H3C
Scheme 1. Synthesis of EO polymers P1, P2 and P3.
separated and the organic one was washed with brine twice and
dried over anhydrous MgSO4. After filtration and evaporation of
solvent, the corresponding compound was dissolved in Et2O and
passed through a short plug of silica. After evaporation the expec-
ted product was obtained (0.25 g, 78%).
for 12 h at room temperature under reduced pressure. A red
powder was obtained.
2.3.4.1. P1 (yield: 60%). 1H NMR (300 MHz, CDCl3),
d (ppm): 8.44 (s,
1H); 8.19 (s, 1H); 7.80 (m, 3H); 6.78 (s, 2H); 4.57 (br s, 4H); 3.73 (m,
10.4H); 1.77 (m, 13H); 0.20 (br s, 24H). SEC (polystyrene): Mn (g/
mol): 10 700; PDI ¼ 1.4. FT-IR (KBr, cmꢁ1): 2958 (nst(CH2)); 2183 (nst
1H NMR (300 MHz, CDCl3),
d
(ppm): 7.03 (d, 6H, J ¼ 8.7 Hz); 6.92
(d, 6H, J ¼ 8.7 Hz). 13C NMR (75 MHz, CDCl3),
d (ppm): 144.44;
134.51; 125.12; 120.03. EI-MS: m/z calcd for C18H12N10: 368.12 (M),
found: 340.13 (M ꢁ N2); 312.14 (M ꢁ 2N2); 284.09 (M ꢁ 3N2). FT-IR
(C^C)); 1728 (nst(C]O)); 1598 (nst(C]C)). UVeVis (CH2Cl2): lmax (%
w) ¼ 480 nm (48). TGA-DSC (10 ꢀC/min): Tg ¼ 93 ꢀC; Td ¼ 250 ꢀC.
(KBr, cmꢁ1): 2101 (nst(N3)); 1618 (
n
st(C]C)). TGA-DSC (10 ꢀC/min):
Td ¼ 173 ꢀC.
2.3.4.2. P2 (yield: 72%). 1H NMR (300 MHz, CDCl3),
d (ppm): 8.46 (s,
1H); 8.18 (s, 1H); 7.70 (m, 3H); 6.76 (s, 2H); 4.00 (m, 5H); 3.35 (m,
8.4H); 1.20 (m, 8.5H); 0.99 (m, 14.4H). SEC (polystyrene): Mn
(g/mol): 11500; PDI ¼ 2.9. FT-IR (KBr, cmꢁ1): 2958 (nst(CH2)); 2108
2.3.4. General procedure of polymerization
Compound 2 or 3 (3 equiv., generally 600 mg), compound N3MA
or TMSMA (7 equiv.) and AIBN (2%m equiv.) were dissolved in THF
(8 mL) in a dry schlenk tube. The mixture was degassed by three
freeze-pump-thaw cycles, and heated for 18 h at 70 ꢀC in the dark.
After getting back to room temperature, the mixture was precipi-
tated when poured dropwise in methanol (80 mL). The solid was
washed twice and isolated by centrifugation. The solid was dried
(
nst(N3)); 1725 (nst(C]O)); 1580 (nst(C]C)). UVeVis (CH2Cl2): lmax (%
w) ¼ 478 nm (51). TGA-DSC (10 ꢀC/min): Tg ¼ 90 ꢀC; Td ¼ 244 ꢀC.
2.3.4.3. P3 (yield 65%). 1H NMR (300 MHz, CDCl3),
d (ppm): 8.48 (s,
1H); 8.10 (s, 1H); 7.70 (m, 5H); 6.86 (s, 2H); 6.71 (s, 2H); 9.91 (m,
5H); 3.42 (m, 8H); 1.77 (m, 9H); 0.90 (m, 15H). SEC (polystyrene):

2290
C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
N3
N3
Br
Br
NaN3
DMF, 60°C
7
DA1
N3
Br
O
N
OH
N
N
K2CO3, THF
60°C
O
O
DA2
Cl
N
N
N
CH3
Cl
Cl
8
CH3
HO
O
N
N
K2CO3, THF
60°C
DA3
CH3
O
N
O
H3C
N3
N3
H2N
NH2
O2N
NO2
N
NaNO2, NaN3
TFA, 0°C
N
N
hydrazine/Pd-C
Dioxane/EtOH
reflux
N3
DA4
NH2
NO2
10
9
Scheme 2. Synthesis of doping agents DA1, DA2, DA3 and DA4.
Mn (g/mol): 5400; PDI ¼ 3.3. FT-IR (KBr, cmꢁ1): 2932 (nst(CH2));
copolymerization initiated by azobisisobutyronitrile (AIBN) with
a 3/7 molar ratio of the chromophoric monomer and the suitable
complementary monomer (5 or 6) (Scheme 1). The number- and
weight average molecular weight (Mn) and the polydispersity
index (PDI) referenced versus polystyrene standards were deter-
mined by size exclusion chromatography (Table 1). The percentage
of chromophore incorporated in the polymer was determined by
two different methods, namely UVevisible spectrophotometry at
480 nm, corresponding to the charge transfer absorption band of
the chromophore and by integration of the surface of the signals in
the 1H NMR spectra. The results of these two analyses are consis-
tent with one another and gave a percentage of chromophore
around 48% in weight (29% in mole) for polymer P1, 51% (28% in
mole) for P2 and finally 55% (27% in mole) for P3. Interestingly, the
molar composition of the chromophores in the polymers is rela-
tively similar to that of the feed (30%). The FT-IR spectra clearly
2109 (nst(N3)); 1728 (Ester nst(C
]O)); 1599, 1507 (Amide nst(C]O));
1565 (
n
st(C]C)). UVeVis (CH2Cl2): lmax (%w) ¼ 480 nm (55). TGA-
DSC (10 ꢀC/min): Tg ¼ 110 ꢀC; Td ¼ 254 ꢀC.
3. Results and discussion
3.1. Synthesis and characterizations
Polymers P1, P2 and P3 were synthesized following the general
route laid out in Scheme 1. Disperse Red 1 (DR1) was used as NLO
chromophore, for its ease of synthesis and functionalization and
because there is a large body of informations about this compound,
allowing comparisons with previously reported systems [44,45].
Starting with the previously described monomer 1 [44], ester-
ification of the carboxylic acid group by methanol or amidification
by azidoaniline was carried out via activation with the 4-(4,6-
dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium
chloride
Table 1
(DMTMM) in presence of N-methylmorpholine (NMM) as a base
and led to the desired monomers (2 and 3) in quasi quantitative
yields (Scheme 1). The DMTMM presents several advantages such
as a cheap and scalable synthesis, a good storage stability and high
reactivity at room temperature and is fully compatible with
methacrylate derivatives [48]. The polymers P1, P2 and P3 were
synthesized in 60%, 72% and 65% yield respectively by free-radical
Specific properties of the polymers. Tg
Td ¼ decomposition temperature (Td measured at 5% wt decomposition).
¼
glass transition temperature;
Entry
Polymer
Mn (g molꢁ1
)
Mw (g molꢁ1
)
PDI
Tg (ꢀC)
Td (ꢀC)
1
2
3
P1
P2
P3
10,700
11,500
5400
15,000
33,000
17,800
1.4
2.9
3.3
93
90
110
250
244
254

C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
2291
Table 2
hydrazine monohydrate in presence of palladium on charcoal [51].
The freshly formed tri-aminophenylamine 10 was engaged in
a diazotation reaction with nitrite acid, affording a non-isolated
diazonium salt which was directly quenched by sodium azide to
give DA4 with 78% yield. This modified Sandmeyer reaction
produced DA4 in 50% yield over two steps.
To determine the glass transition (Tg), the decomposition (Td)
and the cross-linking temperatures (Tcross), polymers P1, P2 and P3
were first mixed in dichloromethane with the complementary DA
used in equimolar amount (versus the number of cross-linking
groups) and the homogenous solutions were vacuum-evaporated
at low temperature. The decomposition temperature (Td) of the
mixture was recorded by thermogravimetric analysis (TGA), while
the Tg and the Tcross were determined by differential scanning
calorimetry (DSC). The results of the analyses are gathered in
Table 2 and a typical DSC trace is showed in Fig. 3.
The DSC trace features a first endothermic breakdown around
90 ꢀC attributed to the glass transition temperature (Tg) of the
polymer. It should be noted that the DA does not significantly
modify the Tg of the polymer, inferring that the former displays
relatively insignificant plastifying effect. Then, a first exothermic
peak at around 140 ꢀC is assigned to the Huisgen cross-linking
reaction. A very intense exothermic peak, around 240 ꢀC, is
attributed to the decomposition of triazole probably into nitrene
[57]. A similar exothermic reaction was previously reported in tri-
azole based polymers by Ergin and coworkers [53]. TGA analysis
showed that all polymers exhibit high thermal stability since the
decomposition temperatures (Td), were never below 240 ꢀC
(Table 2). Interestingly, the cross-linking temperature of the
different mixtures, incorporating polymers P1 or P2, depends on
the nature of the doping agent. It is well-accepted and
Specific properties of the mixtures (polymer þ DA). Tcross ¼ cross-linking temper-
ature measured at the maximum of the exothermic peak on the DSC trace. Tg ¼ glass
transition temperature. Td ¼ decomposition temperature determined at 5% weight
loss.
Entry
Mixture
Polymer
DA
Tg (ꢀC)
Tcross (ꢀC)
Td (ꢀC)
1
2
3
4
5
M1
M2
M3
M4
M5
P1
P2
P2
P1
P3
DA1
DA2
DA3
DA4
DA3
93
90
90
93
110
150
143
174
157
182
250
244
252
251
238
revealed the presence of a weak C^C stretching band around
2260 cmꢁ1 for P1 and the characteristic intense azide stretching
band around 2180 cmꢁ1 for P2 and P3.
The three polymers P1-P3 exhibit high solubility in common
polar organic solvents, such as CH2Cl2, CHCl3, THF, DMF and DMSO,
and non-saturating concentrations up to 250 g/L in o-dichloro-
benzene have been reached. However, after cross-linking the
solubility of the polymer has been drastically decreased since the
polymer film cannot be dissolved upon soaking in a solvent.
In parallel to the polymer preparations, four doping agents were
synthesized following the routes depicted in Scheme 2.
It was decided to select DA whose preparation could be
accomplished with minimum synthetic efforts for scalable
preparative process in view of practical applications. The syntheses
of DA1 and DA2 were described earlier [52,56]. The doping agent
DA3 was prepared in a manner analogous to that for DA2 in 60%
yield by a nucleophilic addition of 2-butyn-1-ol on cyanuric chlo-
ride 8 in the presence of K2CO3. The synthesis of DA4 starts with the
reduction of commercial N,N,N-tris(4-nitrophenyl)amine 9 with
Fig. 3. Differential scanning calorimetry trace of the mixture M2 (P2þDA2) recorded under nitrogen flow with a scanning rate of 10 ꢀC/min. Blue trace: first scan, red trace: return
to RT, green trace second scan. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2292
C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
Table 3
experimentally verified that Huisgen cycloaddition reaction is
generally faster with electron-withdrawing substituent on the
alkyne group or with electron-releasing substituent on the azido
function [58]. The fact that Tcross for P1 mixed with DA4 (157 ꢀC) is
a slightly higher than that measured with DA1 (150 ꢀC) could be
explained by a higher reactivity and mobility of DA1 relative to DA4.
Indeed, large and rigid molecules, like DA4, certainly exhibit
reduced mobility than floppy and small molecules. Furthermore,
the higher cross-linking temperature measured in M3 (P2 þ DA3)
compared to that with M2 (P2 þ DA2), clearly shows that the
methyl group on the propargylic cross-linking group of the doping
agent decreases its reactivity by electron-releasing effect and
probably by steric hindrance (Table 2). These findings demonstrate
that structural and functional modifications on the cross-linking
groups can be used to tune the kinetics of this cross-linking
reaction.
Second harmonic generation coefficients determined by SHG technique; d: thick-
ness of the film; d33 resonance-enhanced coefficients measured at 1064 nm and
extrapolated (static value) following the two-states approximation [62,63]; stability:
temperature at which 5% of the initial SHG signal is lost.
Entry Material
d
d33 (pm/V) lmax(nm) d33 (pm/V)
Stability
(nm) @1064 nm
N nm (static) (ꢀC)
1
2
3
4
5
M2
M3
M4
M5
750 45.5
640 52.3
1010 42.1
735 60.0
1220 44.5
480
482
482
484
476
6.7
7.5
6.0
8.2
7.1
115
130
135
150
137
PAS1
described in references [54,55,61]. Note the strong dispersion of the
optical index at 532 nm which is close to the absorption maximum
for all materials.
The SHG coefficients (d33) of the different materials are rela-
tively similar to one another and within the same range than
previously measured with DR1-grafted methacrylate polymers
with similar chromophore loading such as PAS1 [30,44,45]. Inter-
estingly, mixture M5, which contains cross-linkable chromophores,
displays a much higher electro-optic coefficient certainly testifying
of a higher poling efficiency. The dynamic thermal stabilities of
poled materials were investigated by depoling experiments, in
which decay of the SHG signal was monitored as a function of the
temperature (Fig. 5).
3.2. Nonlinear optical measurements
Before measuring the nonlinear optical properties, the mixtures
were spun cast over glass sheets from filtered solutions of 200 g/L
in o-dichlorobenzene, through 0.2 mm PTFE membranes. Films were
then heated at 70 ꢀC on a heating plate for 2 h to ensure removal of
residual solvent. All the mixtures display high solubility in o-
dichlorobenzene allowing producing homogeneous films with high
optical quality (no light scattering) and with thickness up to 4.1 mm.
Several interesting conclusions can be drawn from these
depoling experiments. First, the cross-linking reaction systemati-
cally enhances the stability of the NLO activity of all the mixtures
relative to the uncross-linked polymer (P1 or P2). For a poled only
polymer (P1 or P2), the chromophores start to relax at around
75 ꢀC, which is close to Tg in agreement with what is generally
observed with other polymers, while cross-linked mixtures are
stable at least until 115 ꢀC (Table 3). Second, the mixtures M3 and
M4 display a similar NLO stability than that measured with PAS1
which corresponds to the most stable nonlinear optical polymer
synthesized with our initial approach (system A, Fig. 2). This result
indicates that the combination of simpler polymers, such as P1 or
P2, with a doping agent can lead to satisfying nonlinear optical
stability, which is in the same range as that obtained with a more
complex polymer (such as PAS1) with potential shelf-life insta-
bility. Finally polymer P3, combined with DA3, exhibits the highest
SHG stability among all the Huisgen-cross-linked based materials
we have tested so far [45]. Although, P3 still requires the double
functionalization of the chromophore, it most probably exhibits
longer shelf-life stability since the two complementary cross-link-
ing groups are not present in the polymer during storage and
In order to evaluate the EO activity of the mixtures, the chromo-
phores were poled with a 3.9 kV electric field for 1 h via corona
poling at 90 ꢀC. Temperature was then increased to 140 ꢀC for one
supplementary hour to induce the cross-linking process. The initial
high optical quality of films was conserved in the majority of cases
upon poling excepted for the mixture M1 composed of the polymer
P1 with the doping agent DA1. Whatever the different poling
conditions used, the systematic degradation of the films of M1 was
observed as illustrated in Scanning Electron Microscopy (SEM)
images (Fig. 4).
The comparison between M1 and M4 poled films clearly reveals
that the degradation is bound to the doping agent DA1. Although
not proved, the surface deformation causing scattering losses, could
be explained by a lack of resilience of the very sensitive benzylic
groups to the charges in motion generated by the electric field
[59,60].
The d33’s harmonic generation (SHG) coefficient of the films
M2eM5 were then measured by second harmonic generation using
Maker’s fringe techniques with a YAG Laser at 1064 nm and the
results are gathered in Table 3. The indexes of refraction at the
fundamental wavelength (1064 nm) and the harmonic wavelength
(532 nm) have been determined (during the SHG fitting procedure)
circa 1.5 ꢂ 0.1 and 1.9 ꢂ 0.1 respectively, following the method
therefore represents
a significant improvement over PAS1.
Although the NLO stability of the above materials is not the highest
Fig. 4. Scanning Electron Microscopy (SEM) images of an M1 film before and after poling.

C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
2293
higher second hyperpolarizability coefficient than DR1, in order to
prepare stable polymers with high EO coefficients.
1
0,8
0,6
0,4
0,2
0
M5
Acknowledgments
PAS1
M4
M3
Agence Nationale de la Recherche (ANR-Télécom with project
ModPol) and Région Pays de la Loire (MILES-MATTADOR program)
are gratefully acknowledged for the financial support of these
researches. V. R. thanks the Région Aquitaine for financial support
in optical, laser, and computer equipment. J.L. Moneger is
acknowledged for technical assistance (TGA analyses).
M2
P2
P1
50
65
80
95
110
125
140
155
Temperature (°C)
References
Fig. 5. Thermal stability of the studied matrices upon heating an initially poled film at
a rate of 2 ꢀC/min.
[1] Dalton L, Harper A, Ren A, Wang F, Todorova G, Chen J, Zhang C, Lee M. Ind Eng
Chem Res 1998;38(1):8e33.
ever reported in the field of electro-optic organic materials [10], the
Huisgen cross-linking reaction presented herein represents
a promising strategy since it enables to maintain the chromophore
orientation at a higher temperature than many reported cross-
linked NLO polymers [26,29,64,65] and as important the thermal
stability of these polymers lies above 85 ꢀC which is the lowest
threshold temperature for practical applications [10].
[2] Ostroverkhova O, Moerner WE. Chem Rev 2004;104(7):3267e314.
[3] Sullivan PA, Dalton LR. Acc Chem Res 2010;43(1):10e8.
[4] Prasad PN, Williams DJ. Introduction to nonlinear optical effects in molecules.
New York: John Wiley and Sons; 1991.
[5] Lee K-S, Dalton L. Adv Polym Sci; 2002:1581e6.
[6] Lee K-S, Kajzar F, Lee K-S, Jen A. Polymeric materials and their orientation
techniques for second-order nonlinear optics. In: Polymers for photonics
applications II, vol. 161. Berlin/Heidelberg: Springer; 2003. 1e85.
[7] Enami Y, Derose CT, Mathine D, Loychik C, Greenlee C, Norwood RA, Kim TD,
Luo J, Tian Y, Jen AKY, Peyghambarian N. Nat Photonics; 2007:1180e5.
[8] HaackG, Förster H, Büttiker M. PhysRev B 2010;82(15):155303e155303/155309.
[9] Dalton LR. Synth Met 2001;124(1):3e7.
4. Conclusions
[10] Dalton LR, Benight SJ, Johnson LE, Knorr DB, Kosilkin I, Eichinger BE,
Robinson BH, Jen AKY, Overney RM. Chem Mater 2011;23(3):430e45.
[11] Zyss J, Ledoux I. Chem Rev 1994;94(1):77e105.
[12] Luo J, Zhou X-H, Jen AKY. J Mater Chem 2009;19(40):7410e24.
[13] Burland DM, Miller RD, Walsh CA. Chem Rev 1994;94(1):31e75.
[14] Davey MH, Lee VY, Wu LM, Moylan CR, Volksen W, Knoesen A, Miller RD,
Marks TJ. Chem Mater 2000;12(6):1679e93.
[15] Jen AKY, Cai Y, Bedworth PV, Marder SR. Adv Mater 1997;9(2):132e5.
[16] Chen T-A, Jen AKY, Cai Y. J Am Chem Soc 1995;117(27):7295e6.
[17] Miller RD, Burland DM, Jurich M, Lee VY, Moylan CR, Thackara JI, Twieg RJ,
Verbiest T, Volksen W. Macromolecules 1995;28(14):4970e4.
[18] Jungbauer D, Reck B, Twieg R, Yoon DY, Willson CG, Swalen JD. Appl Phys Lett
1990;56(26):2610e2.
[19] Hubbard MA, Marks TJ, Lin W, Wong GK. Chem Mater 1992;4(5):965e8.
[20] Eich M, Bjorklund GC, Yoon DY. Polym Adv Technol 1990;1(2):189e98.
[21] Mao SSH, Ra Y, Guo L, Zhang C, Dalton LR, Chen A, Garner S, Steier WH. Chem
Mater 1998;10(1):146e55.
[22] Carella A, Centore R, Mager L, Barsella A, Fort A. Org Electron 2007;8
(1):57e62.
[23] Luo J, Huang S, Cheng Y-J, Kim T-D, Shi Z, Zhou X-H, Jen AKY. Org Lett 2007;9
(22):4471e4.
[24] Ma H, Liu S, Luo J, Suresh S, Liu L, Kang SH, Haller M, Sassa T, Dalton LR,
Jen AKY. Adv Funct Mater 2002;12(9):565e74.
[25] Haller M, Luo J, Li H, Kim T-D, Liao Y, Robinson BH, Dalton LR, Jen AKY.
Macromolecules 2004;37(3):688e90.
A new implementation of copper-free thermal Huisgen cyclo-
addition, as cross-linkable reaction for quadratic nonlinear optical
polymers, has been developed. It is based on the introduction of the
two complementary cross-linkable groups (azide and alkyne) in
two different materials, mixed just before the preparation of the
film. The new system consists in a polymer containing the chro-
mophore (DR1) and a single cross-linking group (azide or alkyne)
as pendant groups and a second material, corresponding to the
doping agent, containing several units of the complementary cross-
linkable group. In this study we have shown that mixtures of the
above mentioned polymer and a suitable doping agent are miscible
in the most commonly used organic solvents and allow to produce
homogeneous and thick films with good optical quality. Secondly,
this new strategy probably overcomes the potentially short shelf-
storage lifetime, since the cross-linkable groups are separated until
the utilization of the polymers. Second-harmonic generation
measurements demonstrate, first that the doping of such polymers
with a suitable polyazide or polyalkyne reagent does not alter the
orientation of the chromophores, since the SHG coefficients of the
mixtures M1-M5 (d33 up to 60 pm/V) were in the same range as
that reported for DR1-based polymers. Secondly, depoling experi-
ments revealed that the stability of the chromophore orientation in
these mixtures is not only enhanced compared to uncross-linked
polymer, but is also comparable to that obtained with the best
cross-linkable polymer PAS1 described in our previous studies. This
latter result demonstrates that the derivatization of the chromo-
phore with a cross-linking group is not an absolute requirement for
this copper-free Huisgen cross-linking reaction to be used. This
feature broadens the scope and versatility of this new cross-linking
methodology and most probably widens the possibilities offered by
copper-free Huisgen cross-linking reaction in the field of electro-
optic materials. Finally, when a cross-linking group (azide) is
positioned on both the chromophore and on the polymer back-
bone, the highest electro-optic stability was recorded (150 ꢀC). This
result tends to indicate that the mixture of two complementary
materials is certainly more efficient to raise the glass transition
temperature of a polymer by Huisgen reaction than when the two
cross-linking groups are grafted on a single polymer. The next step
is to apply this fully operational strategy to chromophores having
[26] Ma H, Chen B, Sassa T, Dalton LR, Jen AKY. J Am Chem Soc 2001;123(5):986e7.
[27] Luo J, Haller M, Ma H, Liu S, Kim T-D, Tian Y, Chen B, Jang S-H, Dalton LR,
Jen AKY. J Phys Chem B 2004;108(25):8523e30.
[28] Luo J, Liu S, Haller M, Liu L, Ma H, Jen AKY. Adv Mater 2002;14(23):1763e8.
[29] Bai Y, Song N, Gao JP, Sun X, Wang X, Yu G, Wang ZY. J Am Chem Soc 2005;127
(7):2060e1.
[30] Bosc D, Foll F, Boutevin B, Rousseau A. J Appl Polym Sci 1999;74(4):974e82.
[31] Monnereau C, Blart E, Illien B, Paris M, Odobel F.
J Phys Org Chem;
2005:181050e8.
[32] Apostoluk A, Nunzi JM, Boucher V, Essahlaoui A, Seveno R, Gundel HW,
Monnereau C, Blart E, Odobel F. Opt Commun 2006;260(2):708e11.
[33] Monnereau C, Blart E, Montembault V, Fontaine L, Odobel F. Tetrahedron
2005;61(42):10113e21.
[34] Jang SH, Jen AK. Chem Asian J 2009;4(1):20e31.
[35] Morotti T, Calabrese V, Cavazzini M, Pedron D, Cozzuol M, Licciardello A,
Tuccitto N, Quici S. Dalton Trans 2008;22:2974e82.
[36] Wang Y, Wang, Guo Y, Cui, Lin Q, Yu, Liu, Xu L, Zhang, Yang B. Langmuir
2004;20(21):8952e4.
[37] Heflin JR, Guzy MT, Neyman PJ, Gaskins KJ, Brands C, Wang Z, Gibson HW,
Davis RM, Van Cott KE. Langmuir 2006;22(13):5723e7.
[38] Facchetti A, Annoni E, Beverina L, Morone M, Zhu P, Marks TJ, Pagani GA. Nat
Mater 2004;3(12):910e7.
[39] Katz HE, Wilson WL, Scheller G. J Am Chem Soc 1994;116(15):6636e40.
[40] Kim T-D, Kang J-W, Luo J, Jang S-H, Ka J-W, Tucker N, Benedict JB, Dalton LR, Gray T,
Overney RM, Park DH, Herman WN, Jen AKY. J Am Chem Soc 2007;129(3):488e9.
[41] Caruso U, Centore R, Panunzi B, Roviello A, Tuzi A. Eur J Inorg Chem 2005;2005
(13):2747e53.

2294
C. Cabanetos et al. / Polymer 52 (2011) 2286e2294
[42] Carella A, Centore R, Sirigu A, Tuzi A, Quatela A, Schutzmann S, Casalboni M.
Macromol Chem Phys 2004;205(14):1948e54.
[54] Rodriguez V, Adamietz F, Sanguinet L, Buffeteau T, Sourisseau C. J Phys Chem B
2003;107(36):9736e43.
[43] Jin Y, Carr SH, Marks TJ, Lin W, Wong GK. Chem Mater 1992;4(5):963e5.
[44] Scarpaci A, Blart E, Montembault V, Fontaine L, Rodriguez V, Odobel F. Chem
Commun 2009;14:1825e7.
[55] Rodriguez V, Sourisseau C. J Opt Soc Am B 2002;19(11):2650e64.
[56] Wang X, Zhao Z, Tian J, Wan L, Hu Y, Huang F, Du L. Polym J 2009;41
(6):498e502.
[45] Scarpaci A, Cabanétos C, Blart E, Pellegrin Y, Montembault V, Fontaine L,
Rodriguez V, Odobel F. Polym Chem 2011;2:157e67.
[46] Kim TD, Luo J, Ka JW, Hau S, Tian Y, Shi Z, Tucker N, Jang SH, Kang JW, Jen AY.
Adv Mater 2006;18(22):3038e42.
[57] L’Abbe G. Chem Rev 1969;69(3):345e63.
[58] Parsons PJ. Synthetic applications of 1,3-dipolar cycloaddition chemistry
toward heterocycles and natural products, 8. New York, Chichester: Wiley;
2002. p. 293.
[47] Jianda X, Lihua H, Wenfang S, Xiaoxu D, Zhuangqi C, Qishun SJ. Polym Sci Part
B Polym Phys 2008;46(12):1140e8.
[59] Hong Min Y, Lee K-s, Sup Yoon C, Mi Do L. J Mater Chem 1998;8
(5):1225e32.
[48] Kunishima M, Kawachi C, Monta J, Terao K, Iwasaki F, Tani S. Tetrahedron
1999;55(46):13159e70.
[49] Scarpaci A, Cabanetos C, Blart E, Montembault V, Fontaine L, Rodriguez V,
Odobel F. J Polym Sci Part A: Polym Chem 2009;47(21):5652e60.
[50] Sumerlin BS, Tsarevsky NV, Louche G, Lee RY, Matyjaszewski K. Macromole-
cules 2005;38(18):7540e5.
[51] Xu K, Economy J. Macromolecules 2004;37(11):4146e55.
[52] Lu Y, Gervay-Hague J. Carbohydr Res 2007;342(12e13):1636e50.
[53] Ergin M, Kiskan B, Gacal B, Yagci Y. Macromolecules 2007;40(13):4724e7.
[60] Sang-Shin L, Garner SM, Chuyanov V, Hua Z, Steier WH, Fang W, Dalton LR,
Udupa A, Fetterman HR. IEEE J Quantum Electron 2000;36(5):527e32.
[61] Rodriguez V. J Chem Phys 2008;128(6):064707e10.
[62] Oudar JL, Chemla DS. J Chem Phys 1977;66(6):2664e8.
[63] Oudar JL. J Chem Phys 1977;67(2):446e58.
[64] Zhang C, Wang C, Yang J, Dalton LR, Sun G, Zhang H, Steier WH. Macromol-
ecules 2001;34(2):235e43.
[65] Sullivan PA, Olbricht BC, Akelaitis AJP, Mistry AA, Liao Y, Dalton LR. J Mater
Chem 2007;17(28):2899e903.