Solution DFWM χ(3) non-linear optical properties of poly [(arylene)silylene]s and poly [(arylene)(ethynylene)silylene]s containing tetra- or hypercoordinate silicon

Article abstract of DOI:10.1039/b003046p

The third-order optical nonlinearities of a series of conjugated poly[(arylene)(ethynylene)silylene]s, and also of a complementary series of poly[(arylene)silylene] statistical copolymers without acetylene groups, have been studied in chloroform solution

Full text of DOI:10.1039/b003046p

View Article Online / Journal Homepage / Table of Contents for this issue
Solution DFWM v(3) non-linear optical properties of
poly[(arylene)silylene]s and poly[(arylene)(ethynylene)silylene]s
containing tetra- or hypercoordinate silicon
William E. Douglas,*a Robert E. BenÐeld,b Oleg L. Antipov,c Larisa G. Klapshina,d
Alexander S. Kuzhelev,c Daniel M. H. Guy,a Richard G. Jones,b Alkay Mustafab and
George A. Domrachevd
a CNRS UMR 5637, Universite Montpellier II, 34095 Montpellier cedex 5, France
b School of Physical Sciences, University of Kent, Canterbury, UK CT 2 7NR
c Institute of Applied Physics, Russian Academy of Sciences, Uljanov Street 46, Nizhny Novgorod,
603600, Russia
d G. A. Razuvaev Institute of Metallo-organic Chemistry, Russian Academy of Sciences,
T ropinin Street 49, GSP-445, Nizhny Novgorod, 603600, Russia
Received 17th April 2000, Accepted 15th May 2000
Published on the Web 26th June 2000
The third-order optical nonlinearities of a series of conjugated poly[(arylene)(ethynylene)silylene]s, and also of
a complementary series of poly[(arylene)silylene] statistical copolymers without acetylene groups, have been
studied in chloroform solution by using the degenerate four-wave mixing technique with a Q-switched
Nd : YAG laserÈoscillator at 1064 nm and a pulse duration of ca. 6 ns. The polymers contain a variety of
arylene groups including carbazole and anthracene. The electronic and nuclear contributions of the s(3)
susceptibility and the thermal nonlinearity of the solutions were separated. The s(3) susceptibilities of some
poly[(arylene)(ethynylene)silylene]s were found to be as high as o Re s(3) o \ 0.95 ] 10~10 esu for solutions of
concentration 50 g l~1. The results show that the presence of a single 8-(dimethylamino)naphthyl ligand at
silicon a†ording pentacoordination has a beneÐcial e†ect on the s(3) properties. Comparison of the results for
the poly[(arylene)silylene]s with those for the poly[(arylene)(ethynylene)silylene]s suggests that the absence of
acetylene groups in the former case has in general a deleterious e†ect on the s(3) properties. ConÐrmation of
the order of magnitude of the non-linear response has been conÐrmed by Z-scan measurements with
picosecond laser pulses on hybrid solÈgel silica Ðlms of two of the poly[(arylene)(ethynylene)silylene]s
containing arylene amide groups.
High and fast optical nonlinearities of organic and
organometallic materials are attracting considerable attention
because of their potential uses in optical computation, tele-
communications, integrated optics, and opto-electronic
devices.1h3 A wide variety of organic polymers with conju-
gated p-electron systems have been studied for third-order
nonlinear optics.4,5 The magnitude of third-order nonlinear
optical susceptibilities (s(3)) has been found to be a†ected by
factors such as p-delocalization length, donorÈacceptor
groups, chain orientation and packing density, conformation
and dimensionality.6
describe
various
the
new
preparation
poly[(arylene)(ethynylene)silylene]s
and
characterization
of
and
the corresponding starting materials. Some of the
poly[(arylene)(ethynylene)silylene]s contain hypercoordinate
silicon resulting from the presence of the 8-(dimethylamino)
naphthyl ligand.11 Few such polymers are known although
the chemical and electronic behaviour of silicon is highly
dependent on the co-ordination number.12
Experimental
Poly[(arylene)ethynylene]s have interesting photo- and
electroluminescent properties and show large third-order non-
linear optical susceptibilities (s(3)) similar to those for
poly[diacetylene]s,7 giving rise to strong nonlinear optical
e†ects such as two- and four-wave mixings, and self-action.
Solution Z-scan s(3) measurements on the corresponding
Reactions were carried out under dinitrogen using Schlenk
tube techniques. Triethylamine was distilled over powdered
KOH, and toluene was distilled from CaH . NMR spectra
2
were run on a Bruker AVANCE DPX 200 spectrometer oper-
ating at 200 MHz (1H), 50.327 MHz (13C) or 39.763 MHz
silicon-containing
poly[(arylene)(ethynylene)silylene]s
of
(29Si). IR spectra were recorded on a Perkin Elmer 1600 FTIR
general structure È[ÈC3CÈSiR1R2ÈC3CÈArÈ] È have been
instrument in CCl solution or as Nujol mulls. Mass spectra
n
4
determined at 590 nm [Re(s(3)) \ [9.3 ] 10~13 esu)8 and at
were obtained by use of a Jeol JMS-DX300 instrument. Ele-
1064 nm in the near-resonant region [Re(s(3)) \ 4.0 ] 10~13
esu].9 Here, we report the considerably higher values obtained
by the degenerate four wave mixing (DFWM) technique for
solutions of a variety of poly[(arylene)(ethynylene)silylene]s,
and also the third-order nonlinear optical susceptibilities
mental analyses were done at the CNRS Service Central
dÏAnalyse. Size exclusion chromatography (SEC) was carried
out in THF at a Ñow rate of 0.9 ml min~1 by use of a Waters
510 system equipped with 100, 500, 1000 and 10 000 A
columns and
a UV detector operating at 254 nm. 1-
of
a
complementary series of poly[(arylene)silylene]
(Bromomethyl)-3,5-dibromobenzene and 2,5-dibromobenzoic
statistical copolymers without acetylene groups.10 We also
acid
were
commercial
samples
(Lancaster).
The
DOI: 10.1039/b003046p
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201
This journal is ( The Owner Societies 2000
3195

View Article Online
poly[(arylene)(ethynylene)silylene]s 1,13,14 2,9 39 and 4,15 as
well as the poly[(arylene)silylene] statistical copolymers
(10aÈd, 11),10 were prepared as previously described.
130.59, 126.40, 125.66, 123.38, 121.05, 118.86, 117.42, 114.72.
M` 429 (EI). Calc. for C
H
Br NO: C, 53.18; H, 2.58; N,
19 11
2
3.26. Found: C, 52.42; H, 2.62; N, 3.27.
Preparation of 1-(trimethoxysilylmethyl)-3,5-dibromobenzene
Preparation of 1-bromo-4-(2-ethyl-1-hexyloxy)benzene
A mixture of 1-(bromomethyl)-3,5-dibromobenzene (5 g, 15.2
mmol) and trichlorosilane (1.6 ml, 15.9 mmol) in tri-n-propyl-
amine (4.5 ml) was kept at 95 ¡C for 20 h and then evaporated
to dryness. The resulting solid [1-(trichlorosilylmethyl)-3,5-di-
bromobenzene] was heated with methanol (15 ml) and tri-
ethylamine (15 ml) for 20 h under reÑux. The excess of meth-
anol and triethylamine was removed under vacuum and the
resulting mixture was extracted with n-pentane. Distillation of
the combined extracts gave 1.33 g (23% overall yield) of the
A mixture of 4-bromophenol (30.00 g, 0.173 mol), 1-bromo-2-
ethylhexane (33.47 g, 0.173 mol) and sodium carbonate (18.38
g, 0.173 mol) was heated under reÑux in DMF (80 ml) for 16
h. The resulting yellow suspension was Ðltered at RT and the
solid was washed with diethyl ether. The Ðltrate and washings
were extracted with water followed by NaOH solution (5%)
and then water again. The organic layer was dried over
K CO and evaporated to give a yellow liquid (33.34 g). Dis-
2
3
tillation (108È115 ¡C, 0.05 mbar) a†orded 1-bromo-4-(2-ethyl-
desired compound (89È95 ¡C, 0.05 mbar). IR [CCl , l/cm~1]:
1-hexyloxy)benzene (colourless liquid; 19.55 g, 40% yield). 1H
4
3109vw, 2944s, 2843s, 1925vw, 1756vw, 1726vw, 1702vw,
NMR (CDCl ): d 7.4 (m, 2 H), 6.8 (m, 2 H), 3.86 (d, 2 H), 1.8
3
1584s, 1552s, 1461m, 1423s, 1402m, 1361w, 1286w, 1192s,
(m, 1 H), 1.4 (m, 8 H), 1.0 (m, 6 H). 13C NMR (CDCl ): d
3
1174s, 1090vs. 1H NMR (CDCl ): d 7.5È7.3 (m, 3 H), 3.57 (s, 9
158.94, 132.58, 116.74, 112.90, 71.14, 39.76, 30.93, 29.51, 24.26,
3
H), 2.16 (s, 2 H). 13C NMR (CDCl ): d 142.01, 130.89, 130.86,
23.50, 14.55, 11.55.
3
123.09, 51.37, 19.59. 29Si NMR (CDCl ): d [49.20.
3
Preparation of phenyl[4-(2-ethyl-1-hexyloxy)phenyl]-
dichlorosilane
Preparation of 2,5-dibromobenzoyl chloride16
A mixture of 2,5-dibromobenzoic acid (5.0 g, 17.8 mmol) and
phosphorus pentachloride (3.7 g, 17.8 mmol) was heated at
65 ¡C for 30 min with evolution of hydrogen chloride. Phos-
phorus oxychloride (bp 106 ¡C) was distilled out of the
resulting brown liquid under atmospheric pressure. The distil-
lation was then continued under reduced pressure a†ording
2,5-dibromobenzoyl chloride (88È95 ¡C, 1 mbar) which crys-
tallized as a white solid (4.72 g, 89% yield). 1H NMR
(CDCl ): d 8.2 (m, 1 H), 7.6 (m, 2 H). 13C NMR (CDCl ): d
The Grignard reagent made by heating under reÑux for 3 h
1-bromo-4-(2-ethyl-1-hexyloxy)benzene (5.72 g, 20.1 mmol)
and magnesium (0.49 g, 20.1 mmol) in THF (50 ml) was added
drop by drop to a vigorously stirred solution of trichloro-
phenylsilane (4.25 g, 20.1 mmol) in THF (50 ml). Stirring was
continued overnight and the mixture was then pumped to
dryness and the residue extracted with n-pentane. Evaporation
of the combined extracts a†orded 7.51 g of the crude product
which was distilled under reduced pressure (145È158 ¡C, 0.15
mbar) to give phenyl[4-(2-ethyl-1-hexyloxy)phenyl]dichloro-
silane (colourless liquid; 5.12 g, 67% yield). 1H NMR
3
3
165.23, 139.97, 137.62, 136.52, 135.99, 121.69, 120.40.
(CDCl ): d 8.0È7.0 (m, 9 H), 3.96 (d, 2 H), 1.8 (m, 1 H), 1.4 (m,
Preparation of N,N-diethyl-2,5-dibromobenzamide
3
8 H), 1.0 (m, 6 H). 13C NMR (CDCl ): d 162.76, 136.36,
3
A solution of 2,5-dibromobenzoyl chloride (4.72 g, 15.3 mmol)
in freshly distilled diethyl ether (50 ml) was added rapidly to
diethylamine (3.89 g, 53.2 mmol). An exothermic reaction
resulted with formation of a copious white precipitate. Stirring
was continued for 1 h and the mixture was Ðltered. Evapo-
ration of the Ðltrate under reduced pressure gave the desired
product (5.3 g) as a white crystalline solid which was puriÐed
by distillation under reduced pressure (4.38 g, 85.4% yield).
134.53, 132.98, 132.06, 128.73, 122.85, 115.05, 70.82, 39.73,
30.94, 29.52, 24.29, 23.50, 14.55, 11.56.
Preparation of phenyl[4-(2-ethyl-1-hexyloxy)phenyl]-
diethynylsilane
Monoacetylene Grignard reagent was prepared by bubbling
acetylene through a solution of the Grignard reagent prepared
from magnesium (0.91 g, 37.4 mmol) and bromoethane (4.09 g,
37.5 mmol) in THF (70 ml). A solution of phenyl[4-(2-ethyl-1-
hexyloxy)phenyl]dichlorosilane (5.00 g, 13.12 mmol) in THF
(50 ml) was added rapidly to the monoacetylene Grignard
reagent at [78 ¡C. The mixture was allowed to warm to room
temperature overnight, and stirring was continued for 3 d. The
reaction mixture was pumped to dryness and the residue
taken up in diethyl ether to give a solution which was washed
successively with aqueous acetic acid (10%), aqueous sodium
bicarbonate, and water. The organic layer was dried over
Mp 52.4È53.7 ¡C. 1H NMR (CDCl ): d 7.4 (m, 3 H), 3.8 (m, 1
3
H), 3.4 (m, 1 H), 3.2 (m, 2 H), 1.27 (t, 3 H), 1.10 (t, 3 H). 13C
NMR (CDCl ): d 167.28, 140.88, 134.67, 133.41, 130.75,
3
121.95, 118.35, 43.18, 39.46, 14.36, 12.91. Calc. for
C
H
Br NO: C, 39.43; H, 3.91; N, 4.18. Found: C, 39.79;
11 13
2
H, 3.89; N, 4.26.
Preparation of N-(2,5-dibromobenzoyl)carbazole
Following the method used to prepare N-benzoylcarbazole,17
carbazole (2.51 g, 15.0 mmol) and sodium hydride (0.75 g of
60% dispersion in mineral oil, 17.5 mmol) were stirred in a
mixture of DMF (60 ml) and toluene (60 ml) at 0 ¡C for 15
min. A solution of 2,5-dibromobenzoyl chloride in DMF (40
ml) was then added drop by drop. After being stirred for 4 h
the reaction mixture was hydrolysed with 7% H SO and was
MgSO and evaporated to give a brown oil (4.28 g). The latter
4
was distilled under reduced pressure (170È185 ¡C, 0.1 mbar) to
a†ord
phenyl[4-(2-ethyl-1-hexyloxy)phenyl]diethynylsilane
(red oil; 2.49 g, 53% yield). IR [neat, l/cm~1]: 3951vw,
3584vw, 3275s, 3071w, 3052w, 3023w, 3001w, 2959vs, 2928vs,
2872s, 2859s, 2526vw, 2040vs [l(C3C)], 1899w, 1820vw,
1778vw, 1595vs, 1563m, 1503s, 1488w, 1466m, 1430m, 1400w,
1380m, 1311m, 1279vs, 1249vs, 1182s, 1159w, 1119vs, 1090m,
1030m, 998w, 971w, 938vw, 928vw, 896vw, 826s, 814m, 789w,
2
4
extracted with toluene. The combined extracts were dried over
MgSO and the solvents were pumped o† to a†ord the crude
4
product. Recrystallization from absolute ethanol gave beige
crystals of the compound (3.40 g, 53.2% yield). Mp 125.2È
741m, 697vs, 630m, 601s. 1H NMR (CDCl ): d 8.0È7.0 (m, 9
3
126.0 ¡C. IR [CCl , l/cm~1]: 3065w, 2986w, 1898vw, 1685vs,
H), 3.93 (d, 2 H), 2.79 (s, 2 H), 1.8 (m, 1 H), 1.4 (m, 8 H), 1.0 (m,
4
1600w, 1490m, 1479m, 1454s, 1445vs, 1418vw, 1388s, 1356m,
6 H). 13C NMR (CDCl ): d 162.00, 136.84, 135.18, 132.24,
3
1342m, 1330vs, 1303s, 1268vw, 1238w, 1214m, 1156w, 1138vw,
130.94, 128.61, 121.93, 115.04, 97.54 (SiC3CH), 84.27
1122w, 1089w, 1034m. 1H NMR (CDCl ): d 8.1È7.3 (m, 11 H).
(SiC3CH), 70.69, 39.75, 30.96, 29.52, 24.30, 23.50, 14.55, 11.56.
3
13C NMR (CDCl ): d 164.48, 138.91, 137.91, 133.94, 133.89,
29Si NMR (CDCl ): d [48.2. M` 360 (FAB`, no matrix).
3
3
3196
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201

View Article Online
phenylphosphine (81.9 mg, 0.312 mmol) in toluene (60 ml) and
triethylamine (60 ml) was stirred at 85 ¡C for 20 h and then
Ðltered. The residue was washed with THF and the Ðltrate
was evaporated to dryness a†ording a red solid (1.64 g) which
was dissolved in THF and the solution Ðltered. Addition of
n-pentane precipitated the polymer (1.04 g, 32% yield). IR
[CCl , l/cm~1]: 2158, 2132 [l(C3C)]. 1H NMR (CDCl ): d
4
3
9.0È7.3 (m, 18.3 H), 3.63 (s, 1 H; OCH ), 2.27 (s, 0.22 H; CH ).
3
2
13C NMR (CDCl ): d 135.65, 135.53, 135.15, 133.24, 131.19,
130.29, 128.97, 128.83, 128.59, 128.03, 127.77, 118.94, 106.21
3
Fig. 1 Experimental optical bench for the DFWM study.
M \ mirror, BS \ beamsplitter, PD \ photodiode, P \ glass prism,
PR \ 90¡ polarization rotator, D \ diaphragm.
(C3C), 102.28 (C3C), 51.47. 29Si NMR (CDCl ) [INVGATE;
3
delay time 15 s]: d [47.13 (7.4 Si; SiC3C), [47.33 (1.6 Si;
SiC3C), [48.68 (0.5 Si; SiCH ), [48.89 (0.5 Si; SiCH ).
2
2
Calc. for C
7.73.
H
OSi: C, 79.95; H, 7.83. Found: C, 79.90; H,
24 28
Preparation of 6
A mixture of diethynyldiphenylsilane (0.29 g, 1.25 mmol),
N,N-diethyl-2,5-dibromobenzamide (0.42 g, 1.25 mmol),
bis(triphenylphosphine)palladium(II) dichloride (1.8 mg, 0.0026
mmol), copper(I) iodide (5.3 mg, 0.0278 mmol) and tri-
phenylphosphine (13.3 mg, 0.0507 mmol) in toluene (15 ml)
and triethylamine (10 ml) was stirred at 80 ¡C for 20 h and
then Ðltered. The Ðltrate was evaporated to dryness a†ording
a brown solid which was dissolved in CH Cl and the
Preparation of 5
A mixture of diethynyldiphenylsilane (1.84 g, 7.92 mmol), 9,10-
dibromoanthracene (2.40 g, 7.14 mmol), 1-(trimethoxy-
silylmethyl)-3,5-dibromobenzene (0.29 g, 0.78 mmol),
bis(triphenylphosphine)palladium(II) dichloride (8.7 mg, 0.0124
mmol), copper(I) iodide (35.7 mg, 0.187 mmol) and tri-
2
2
resulting solution Ðltered. Addition of n-pentane to the Ðltrate
precipitated the desired product which was isolated by cen-
trifugation (0.46 g, 91% yield). IR [CCl , l/cm~1]: 2163
4
[l(C3C)], 1643 [l(C2O)]. 1H NMR (CDCl ): d 8.1È7.2 (m, 13
3
H), 4.0È3.0 (m, 4 H), 1.5È0.8 (m, 6 H). 29Si NMR (CDCl ): d
3
[47.6.
Preparation of 7
Fig. 2 Wave vectors and polarization diagrams for the two series of
DFWM experiments. (a) experimental measurements of s(3) suscepti-
bility of the polymer solutions, (b) experiment for separation of elec-
tronic and nuclear contributions. Pumping waves \ E and E , probe
The polymer (brown solid) was obtained in 93% yield starting
from the corresponding pentacoordinate diethynylsilane9
under the same experimental conditions as those used for 6.
1
2
wave \ E , phase conjugated wave \ E (E is the result of di†rac-
3
4
4
tion of the pumping waves on the reÑecting (dn ) and transmitting
IR [CCl , l/cm~1]: 2158 [l(C3C)], 1646 [l(C2O)]. 1H NMR
1
4
(dn ) refractive index gratings).
(CDCl ): d 8.5 (s, 1 H), 8.0È7.2 (m, 8 H), 3.9È3.0 (m, 4 H;
2
3
NCH ), 2.7 (s, 3 H; NCH ), 2.6 (s, 3 H; NCH ), 1.4È0.8 (m, 6
2
3
3
H; NCH CH ), 0.6 (s, 3 H; SiCH ). 29Si NMR (CDCl ): d
2
3
3
3
[56.8È [ 57.8 (m).
Preparation of 8
The polymer (light brown solid) was obtained in 68% yield
starting from N-(2,5-dibromobenzoyl)carbazole under experi-
mental conditions essentially the same as those used for 6. IR
[Nujol, l/cm~1]: 2162 [l(C3C)], 1682 [l(C2O)]. 1H NMR
(CDCl ): d 8.1È6.6 (m). 29Si NMR (CDCl ): d [47.5È [ 48.0
3
3
(m).
Preparation of 9
The polymer (red solid) was obtained in 57% yield starting
from phenyl[4-(2-ethyl-1-hexyloxy)phenyl]diethynylsilane and
9,10-dibromoanthracene under experimental conditions essen-
tially the same as those used for 6. IR [Nujol, l/cm~1]: 2126
[l(C3C)]. 1H NMR (CDCl ): d 8.9È6.5 (m, 17 H; Ar), 3.79 (d,
3
2 H; OCH ), 1.61 (d, 1 H; CH), 1.5È1.0 (m, 8 H; CH ), 0.9È0.6
2
2
(m, 6 H; CH ). 13C NMR (CDCl ): d 105.92 (C3C), 102.75
(C3C). 29Si NMR (CDCl ): d [47.4.
3
3
3
DFWM measurements of v(3)
DFWM experiments were made by using a Q-switched
Nd : YAG laser-oscillator at 1064 nm with a pulse duration of
ca. 6 ns. The original beam was passed through a Faraday
isolator, followed by an Nd : YAG ampliÐer and was then
directed into the DFWM optical bench (Fig. 1). Two beam
splitters BS1 and BS2 provided two counter-running pumping
beams E and E (with equal energy of ca. 10 mJ) and probe
Fig. 3 The DFWM reÑection coefficient as
a function of the
pumping wave intensity for the Ðrst (a) and second (b) series of experi-
1
2
beam E (with energy of ca. 1 mJ). The probe beam entered
3
ments for a chloroform solution of polymer 1.
the polymer cell at an angle of 5 ] 10~3 rad with respect to
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201
3197

View Article Online
Scheme 1
the pump-wave directions. Two series of experiments with dif-
ferent mutual polarization orientations of the pumping and
probe beams were performed (Fig. 2). The phase conjugated
(PC) wave was recorded for both polarization orientations.
The reÑection coefficient of the PC wave was found to
increase quadratically with increasing pump intensity (Fig. 3).
To determine the absolute value of the susceptibility of the
polymers, the DFWM reÑection coefficient was calibrated
with respect to that for neat chloroform.
The brown polymer 6 was synthesized from N,N-diethyl-
2,5-dibromobenzamide, the latter being prepared from 2,5-di-
bromobenzoyl chloride (Scheme 4). The low molecular
weight corresponds to a chain length of about 9 units. The
brown polymer 7 containing pentacoordinate silicon was pre-
pared in an analogous way, the molecular weight in this case
corresponding to only 6 units. Several 29Si NMR resonances,
in the same region as for polymer 3 also containing penta-
coordinate silicon,9 are observed with the multiplicity reÑec-
ting the short chains as well as the unsymmetrical nature of
the benzamide moieties.
Results and discussion
The light brown polymer 8 was synthesized from N-(2,5-
dibromobenzoyl)carbazole, the latter being prepared from car-
bazole and 2,5-dibromobenzoyl chloride (Scheme 4). The
l(C3C) IR absorbance and 29Si NMR chemical shifts are in
the same range as for 6. The molecular weight corresponds to
a weight-average chain length of 90 units (Table 1). However,
the true molecular weight may in fact be lower since although
8 is much bulkier than 6 the calculations are made with refer-
ence to polystyrene standards.
Preparation and properties of new monomers and polymers
The new starting monomer phenyl[4-(2-ethyl-1-hexyloxy)-
phenyl]diethynylsilane was prepared by treatment of the cor-
responding dichlorosilane with monoacetylene Grignard
reagent (Scheme 1). Its spectral properties are very similar to
those of diphenyldiethynylsilane13,14 with the same IR l(C3C)
frequency and 29Si NMR chemical shift, the 13C NMR acety-
lene resonances being shifted slightly downÐeld.
The red polymer 9 has a molecular weight slightly lower
than that of the analogous polymer 1 without the side-chain.
The spectral characteristics of both polymers are similar sug-
gesting that the 2-ethyl-1-hexyloxyphenyl side-chain at Si has
little e†ect on the electron distribution within the polymer
backbone.
The new dibromoarene monomer 1-(trimethoxysilylmethyl)-
3,5-dibromobenzene was prepared from 1-(bromomethyl)-3,5-
dibromobenzene (Scheme 2). N,N-Diethyl-2,5-dibromobenza-
mide and N-(2,5-dibromobenzoyl)carbazole were prepared
from 2,5-dibromobenzoic acid (Scheme 3).
Following the route previously described,13 the new poly-
mers 5È9 (Table 1) were prepared by palladium-catalysed
cross-coupling polymerization between the appropriate di-
ethynylsilane and dibromoarene (Scheme 4).
The red polymer 5 was synthesized from a mixture of 10%
1-(trimethoxysilylmethyl)-3,5-dibromobenzene (Scheme 3) and
90% 9,10-dibromoanthracene, the molecular weight being half
that of the corresponding polymer 1 with anthryl groups.13,14
Two l(C3C) IR absorbances are observed both at higher fre-
quency than that of the single absorption observed for 1, the
29Si and acetylene 13C NMR chemical shifts being the same
as in 1. Several 29Si NMR resonances are observed, the
number and relative intensities suggesting that the
trimethoxysilyl-containing monomer is distributed non-
randomly within the polymer chains.
DFWM measurements
The laser pulse duration was much longer than the time of
termalization of polymer molecular excitation (of the order of
Scheme 3
Scheme 4
Scheme 2
3198
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201

View Article Online
Table 1 Properties of polymers -[-C3C-SiR1R2-C3C-Ar-] (1È9)
n
Polymer substituents
Polymer (R1R2)Sia
Si
coord.
no.
Molecular weight data
Solution DFWM (s(3)) measurementse
Arb
M c
M /M c nd
Colour
o n o /cm2 W~1 o Re s(3) o /esu
w
w
n
2
1
2
3
4
Ph Si
(1-Np) Si
Me(NpN)Si
(NpN) Si
An
An
An
An
4
4
5
6
23500
8800
12400
20000
3.2
56
17
28
34
red
orange
red
190 ] 10~14f
1.5 ] 10~14
110 ] 10~14g
63 ] 10~14g
95 ] 10~12f
0.75 ] 10~12
55 ] 10~12g
31 ] 10~12g
2
2.8
2.7
3.2
2
orange
2
5
Ph Si
(10%)
4
13000
2.9
36
red
15 ] 10~14h
7.5 ] 10~12h
2
An (90%)
6
7
Ph Si
4
5
3390
2310
1.8
1.5
9
6
brown
brown
6.3 ] 10~14h
170 ] 10~14g
3.1 ] 10~12h
85 ] 10~12g
2
Me(NpN)Si
8
9
Ph Si
4
4
40500
19300
3.4
2.3
90
cream
5.7 ] 10~14
2.8 ] 10~12
2
An
40
red
10 ] 10~14
10 ] 10~12
a NpN \ 1-(8-dimethylamino)naphthyl. b An \ 9,10-anthrylene. c Determined by SEC with reference to polystyrene standards. d Number of
units calculated from value of M . e At 1064 nm with 6 ns pulse duration in CHCl solution (measured at, or linearly extrapolated to, 50 g l~1).
w
3
f Concentration, 2.5 g l~1. g Concentration, 5 g l~1. h Concentration, 12.5 g l~1.
tens of picoseconds5). Therefore, the thermally-induced
grating of the solvent refractive index must be taken into con-
sideration for the theoretical description of the DFWM
process. To interpret the experimental results we used the
plane wave approximation for the interacting beams E \ E
by the interference Ðeld of the pumping wave E and the co-
1
running probe E . The spatial period of the large-scale inter-
3
ference grating (K) was about 200 lm. The delay time (q ) for
s
the isobar formation of this thermal grating is estimated to be
q \ 2p/Kc \ 34 ns (the speed of sound in chloroform is ca.
m
m
s
s
exp(i[u t [ k r]) (m \ 1, 2, 3, 4), and the approximation of
c \ 105 cm s~1). Under the present experimental condi-
m
m
S
constant intensities for the probe and pumping beams. The
latter approximation is correct because the PC-wave intensity
in the nonlinear cell was much less than the probe and
pumping wave intensities, o E o2 @ o E o2 @ o E o2 \ o E o2. For
tions the laser pulse duration was much less than this delay
time. Therefore, the large-scale thermally-induced grating of
the solution density variation (do) was determined by the full
thermodynamic equation system which can be written in the
following form18,19
4
3
1
2
these reasons the equation for the PC wave amplitude can be
written in the following form
A
d2
dt2
B
d(do)
dt
Fc2 4p2
S
[ c2 +2
\ [ a
dE
4i
CAdn
B
S
o C K2
\ [ ik
*o(E E*)E
1 3 2i
0
p
0
dz
do
n c
4p
A
i2px
K
B
1
D
0
]
E E* exp
,
(2)
]
(s(3) E
E
E* ] s(3) E
E
E* )
(1)
1 3
ijkl 1j 2k 3l
ikjl 2k 1j 3l
2n
0
where a is the absorption coefficient of the polymer, F \ (do/
dT ) is the volume coefficient of thermal expansion of chloro-
form, C is the speciÐc heat at the constant pressure, o is the
where k is the undisturbed wave number of the optical
beams, n is the refractive index of the medium, s(3) is the fast
0
0
ijkl
optical non-linear susceptibility of the polymer, (i,j,k,l are the
Cartesian coordinates x,y,z indicating the directions of the
wave polarizations), (dn/do) is the gradient of the refractive
index changes due to thermal variation of the chloroform
density, and *o(E E*) is the amplitude of the chloroform
p
0
undisturbed chloroform density, and c and c are the velocities
S
of sound and light, respectively.
The solution of eqn. (2) for optical pulses with the temporal
form (1 [ exp([t/q )) exp([t/q ) gives the following expres-
on
off
1
3
sion for the maximum of the thermally-induced grating ampli-
density grating thermally-induced by the interference Ðeld of
the optical waves.
tude (*o \ do exp([ix2p/K))
To decrease the contribution of the thermally-induced grat-
ings to the total DFWM reÑectivity, in the Ðrst series of
F
4p2 n c
0
*o B aq
E E*
max
S
1 3
experiments the polarization of the probe beam E was the
o C K2 4p
3
same as that for the co-running pump E being cross-
1
0
p
C
q3 /q3
q3/q3
1 S
D
polarized with respect to the polarization of the other pump
off S
]
[
,
(3)
E (Fig. 2a). In this case only a large-scale grating was induced
1 ] (q2 /q2) 1 ] (q2/q2)
2
off S
1 S
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201
3199

View Article Online
where q \ q q /(q ] q ), q and q being the switch-o†
off on on off off on
and switch-on times of the optical pulse.
The expression (3) indicates that the amplitude of the
thermally-induced grating strongly decreases as the ratio
consistent with the predominance of the electronic contribu-
tion of the nonlinear susceptibility over the orientational
(nuclear) component, the latter e†ect being more sensitive to
intermolecular interactions.21
1
q
/q decreases. To make the contribution of the thermal
In order to separate the electronic and nuclear contribu-
tions in the fast Kerr-like nonlinear s(3) susceptibility we have
compared the results for two experimental series [Fig. 2(a)
and 2(b)]. The second series of experiments was performed for
the chloroform solution of polymer 1 containing tetra-
off S
grating negligible it is necessary to operate with pulses of
duration much less than q \ 34 ns. Under our experimental
S
conditions the ratio was of the order of q /q B 0.17. By using
off S
eqns. (1), (2) and (3) and taking into account the values of the
chloroform parameters (dn/dT ) \ (dn/do) (do/dT ) \ 6 ] 10~4
coordinate Ph Si groups, which had shown the highest sus-
2
K~1, a O 0.05 cm~1, oC \ 1 J cm~3 K~1, it is possible to
ceptibility in the previous series. The separation of electronic
p
estimate that the fast grating (due to the Kerr-like
and nuclear (orientational) contributions is based on di†erent
spatial symmetries of the s(3) tensor.22,23 In isotropic media
the four independent s(3) tensor elements reduce to three
according to the relation s(3) \ s(3) ] s(3) ] s(3) , where
nonlinearities) predominates over the thermally-induced
grating if the nonlinear susceptibility s(3) is greater than
2.4 ] 10~13 esu. This condition is achieved even in the case of
the polymer solutions of relatively low nonlinear susceptibility
(Tables 1 and 2).
In the second series of experiments (with the probe wave
cross-polarized to the pumps [Fig. 2(b)] a scalar density
grating was not induced. Therefore, the thermal grating under
both our experimental conditions is estimated to be negligibly
small in comparison with the grating induced due to the fast
Kerr-like non-linearity, the DFWM reÑection coefficient being
determined by the formula
xxxx
xxyy
xyxy
xyyx
by convention the order of the indices corresponds to the dif-
fracted Ðeld (E ), the pump (E ), the other pump (E ) and the
4
1
2
probe Ðeld (E ).22 If the response is purely nuclear
3
(orientational) the relative magnitude of the elements becomes
s(3) \ [0.5s(3) , s(3) \ s(3) \ 0.75s(3) .22,24 For a purely
xyyx
xxxx xxyy
xyxy
xxxx
electronic response in an isotropic medium the tensor ele-
ments have the next relative values: s(3) \ s(3) \ s(3) \
xyyx
xxyy
xyxy
s(3) /3.22,23 The s(3) susceptibility was determined in the
xxxx
xyyx
experiment with the probe wave cross-polarized to both
pumps. By comparing the result of the last experiment with
the results of the previous experiments (in which the probe
wave was co-polarized to one of the pumps giving rise to the
K
E
E
K
2
A
kl
B
2
4i
3l
R \
il
\
o s(3) E
E
] s(3) E
E o2, (4)
ijkl 1j 2k
ikjl 2k 1j
2n
0
s(3) susceptibility) it is possible to determine the ratio of the
(l is the length of the nonlinear medium).
xyxy
electronic and nuclear (orientational) contributions to the fast
Two terms in the expression are indicative of two possible
refractive index gratings: the large-scale reÑection grating,
and the small-scale transmission grating. For both series of
nonlinearity: o s(3) (electronic) o / o s(3) (nuclear) o \ 3È6. It can
xxxx
xxxx
hence be concluded that the electronic contribution is much
larger than the nuclear (orientational) contribution for the s(3)
susceptibilities of the present series of poly[(arylene)(ethynyl-
ene)silylene]s and poly[(arylene)silylene]s.
experiments the s(3) -susceptibility components were the same:
ijkl
s(3) , for the Ðrst series; s(3) , for the second.
xyxy
xyyx
xyxy
Eqn. (4) allows the s(3) or s(3) (and n ) values for the
xyyx
2
polymer to be determined by using the results of the measure-
ments of the DFWM reÑection coefficient in the solution. The
value of the DFWM efficiency in neat chloroform was used as
a calibration value for the n determinations. The o s(3) o value
was calculated from the n value by using the relationship n
(cm2 W~1) \ 5.2 ] 10~2 s(3)[esu]/n2 .20 Measurements on
chloroform solutions of the various polymers were made
v(3) properties of the polymers
For the poly[(arylene)(ethynylene)silylene]s the DFWM mea-
surements were made on chloroform solutions at concentra-
2
xyxy
tions ranging from 50 to 2.5 g l~1 (Table 1). The value of n
2
2
2
measured for the monomer Ph Si(C3CH) as well as for the
0
2
2
previously reported9 model monomer Ph Si(C3CM9-anthrylN)
2
2
(Tables 1 and 2). For all the polymers the values measured
are much higher than that found for neat chloroform, ca.
n \ 3 ] 10~15 cm2 W~1.6 The determined o s(3) o value is
were similar to that observed for the solvent CHCl
3
(3 ] 10~15 cm2 W~1), being about one order of magnitude
less than that observed for the polymers. Except for 6 and 7,
the polymers were of sufficient chain length for the s(3) e†ects
to be at a maximum25 the highest value being observed for 1
2
xyxy
much greater than the literature s(3) value for CS ,
2
s(3) \ 6.8 ] 10~13 [esu].21
In order to study the inÑuence of the solvent on the nonlin-
ear susceptibility of the polymers, the DFWM experiment was
also performed for solutions of polymer 3 in THF and
containing tetracoordinate Ph Si groups (s(3) \ 3.1 ] 10~12
2
esu at a concentration of 2.5 g l~1). Based on number density
considerations (B0.25 wt.% polymer in the solution), this
value might be expected to be greater by a factor of 400 for
the neat polymer suggesting s(3) activity as high as that for
derivatives of poly[(1,4-vinylene)phenylene].26
toluene, the results of the s(3) (and n ) measurements being
2
essentially the same as in CHCl . It should be noted that the
3
molecular interactions of the various solvents with the
polymer are di†erent: the THF molecules with a large dipole
moment might be expected to interact with the polymer mol-
ecules more strongly than do chloroform molecules. However,
the experimental results indicate only a weak e†ect of the
solution molecular interaction on the s(3) susceptibility. This is
For purposes of comparison, where necessary the values of
n
and s(3) have been linearly extrapolated to the reference
2
concentration (50 g l~1), the absorbances for all the solutions
being negligible (\0.05 cm~1). The results obtained are
greater by an order of magnitude than those previously
Table 2 s(3) Solution properties of È[È(SiMePh) (Ar) È] (10 Ar \ 1,4-C H , 11 Ar \ 9,10-anthrylene)a
1.0 1.3
n
6 4
Polymer
M
M
M /M
o n o/cm2 W~1
o Re s(3) o/esu
w
n
w
n
2
10a
10b
10c
10d
11
6783
9946
13515
22774
21700
2663
3135
4125
4019
1350
2.5
3.2
3.3
5.7
16.1
1.2 ] 10~14
4.7 ] 10~14
23 ] 10~14
9.8 ] 10~14
5.6 ] 10~14
0.6 ] 10~12
2.4 ] 10~12
11.5 ] 10~12
4.9 ] 10~12
2.9 ] 10~12
a Measured by DFWM technique at 1064 nm with 6 ns pulse duration in CHCl solution (concentration 50 g l~1).
3
3200
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201

View Article Online
reported for 1, 3 and 4 in CHCl solution (50 g l~1), as deter-
groups designed to facilitate incorporation of the materials
into gels by formation of covalent bonds. Polymers 6È8
contain amide groups which are known to give rise to homo-
geneous organicÈinorganic polymer hybrids,27 and 8 also con-
tains a carbazole group of potential use for its photorefractive
properties. Initial results28 obtained by the Z-scan technique
with picosecond laser pulses on Ðlms of 7 and 8 (5% in a silica
matrix) are of the same order of magnitude as those reported
here for the solutions.
3
mined by the Z-scan technique at 1064 nm (pulse duration 140
ps, intensity 20 GW cm~2).9 Comparison of results obtained
using the two techniques is known to be difficult5 but may in
this case be partly explained by the much lower light beam
intensity (B1 MW cm~2) used here than that used for the
Z-scan study9 (20 GW cm~2). At the greater intensities used
in the Z-scan experiment higher order nonlinear suscep-
tibilities (s(5) and s(7)) giving negative contributions could
decrease the overall nonlinear e†ect of phase modulation.8
Another reason for the small values of n obtained by Z-scan
2
Acknowledgements
measurement is possible competing e†ects such as two photon
absorption and stimulated scattering appearing at high beam
intensity.
Support from the International Association for the promotion
of co-operation with scientists from the New Independent
States of the former Soviet Union (grant no. 97-1785), Bruss-
els, and The International Centre for Advanced Studies (grant
no. 98-3-03), Nizhny Novgorod, is gratefully acknowledged.
The authors thank Prof. Ya. I. Khanin for helpful discussions.
As previously reported,9 for the series of polymers 2È4 con-
taining naphthyl or naphthylamine groups at silicon smooth
trends in the 29Si and SiC3C 13C NMR chemical shifts as well
as the energy of the l(C3C) IR absorbance are observed with
increasing co-ordination number at silicon (reÑecting the
change in the electronic environment), unlike 1 with phenyl
groups at silicon. However, in the case of the s(3) results no
such trend is observed, the e†ect increasing from 2 to 3 and
then decreasing on going from 3 to 4, whereas the value for 1
is extremely high. This is consistent with there being a steric
e†ect on the s(3) results perhaps resulting from unfavourable
conformations forced on the polymers containing naphthyl or
naphthylamine groups. This e†ect may be o†set to some
extent by the favourable electronic e†ect of the naphthylamine
group. Indeed, on comparing 6 and 7, the presence of the
naphthylamine group can be seen to have a most positive
e†ect in the case of pentacoordinated 7.
References
1
2
3
4
5
6
7
8
9
D. S. Chemla and J. Zyss, Nonlinear Optical Properties of Organic
Molecules and Crystals, Academic Press, Orlando, 1987.
K. Toda and H. S. Hilton, Photonic Switching 2, Springer-Verlag,
Berlin, 1990.
J. Zyss, Molecular Nonlinear Optics: Materials, Physics and
Devices, Academic Press, Boston, 1993.
J. L. Bredas and R. Silbey, Conjugated Polymers, Kluwer, Boston,
1991.
J. L. Bredas, C. Adant, P. Tackx and A. Persoons, Chem. Rev.,
1994, 94, 243.
H. S. Nalwa, T. Hamada, A. Kakuta and A. Mukoh, Nonlinear
Optics, 1994, 7, 193.
Comparison of the results for 1 and 9 shows that the pres-
ence of a long chain ether group in the para position on one of
the silicon phenyl substituents is deleterious to the s(3) proper-
ties.
R. Giesa, J. Macromol. Sci. Rev. Macromol. Chem. Phys., 1996, 4,
631.
R. K. Meyer, R. E. Bender, Z. V. Vardeny, Y. Ding and T.
Barton, Mol. Cryst. L iq. Cryst., 1994, 256, 597.
W. E. Douglas, D. M. H. Guy, A. K. Kar and C. Wang, Chem.
Commun., 1998, 2125.
The DFWM measurements for the complementary series of
poly[(arylene)silylene] statistical copolymers10 without acety-
lene groups were made on chloroform solutions at a concen-
tration of 50 g l~1. The polymers contain silylene and arylene
units in random sequences of single units and short blocks,
electron delocalization along the backbone thus involving
both r-conjugated polysilylene groups and n-conjugated poly-
arylene blocks. Comparison of the results (Table 2) with those
for the poly[(arylene)(ethynylene)silylene]s (Table 1) suggests
that the absence of acetylene groups in the former case has in
general a deleterious e†ect on the s(3) properties. An increase
in molecular weight on going from 10a to 10d gives rise to an
increase in s(3) properties, whereas replacement of a phenylene
group (10aÈd) by an anthrylene group (11) has little e†ect. The
result for 10c shows that low polydispersity is particularly
favourable to high s(3) values.
In conclusion, the very high s(3) properties of some of the
poly[(arylene)(ethynylene)silylene]s conÐrm the presence of
extensive through-Si conjugation along the backbone.9 In par-
ticular, the high s(3) values observed for 1 should be noted.
The results of this study suggest that the presence of a single
8-(dimethylamino)naphthyl ligand at silicon a†ording penta-
coordination has a beneÐcial e†ect on the s(3) properties.
Further incorporation of such ligands to give hexacoordinate
species is deleterious. In particular, the high s(3) values
observed for 1 should be noted.
10 R. E. BenÐeld, U. Budnik, T. Niklas, S. Henke and R. G. Jones,
Polymer, 1996, 37, 5761.
11 J. T. B. H. Jastrzebski, C. T. Knaap and G. van Koten, J.
Organomet. Chem., 1983, 255, 287.
12 C. Chuit, R. J. P. Corriu, C. Reye and J. C. Young, Chem. Rev.,
1993, 93, 1371.
13 R. J. P. Corriu, W. E. Douglas and Z.-X. Yang, J. Polym. Sci.,
Poly. L ett. Ed., 1990, 28, 431.
14 R. J. P. Corriu, W. E. Douglas, Z.-X. Yang, Y. Karakus, G. H.
Cross and D. Bloor, J. Organomet. Chem., 1993, 455, 69.
15 K. Boyer-Elma, F. H. Carre, R. J.-P. Corriu and W. E. Douglas,
J. Chem. Soc., Chem. Commun., 1995, 725.
16 S. C. J. Olivier, Rec. T rav. Chim., 1929, 48, 227.
17 A. Chakrabarti, G. K. Biswas and D. P. Chakraborty, T etra-
hedron, 1989, 45, 5059.
18 D. Smith, Proc. IEEE, 1977, 65, 76.
19 H. J. Ho†man, J. Opt. Soc. Am. B, 1986, 3, 258.
20 C. Bubeck, in Electronic materials: the oligomer approach, ed. K.
Mullen and G. Wegner, Wiley-VCH, Weinheim, 1998, pp. 449È
478.
21 R. W. Boyd, Nonlinear Optics, Academic Press, New York, 1992.
22 D. J. McGraw, A. E. Siegman, G. W. Wallraf and R. D. Miller,
Appl. Phys. L ett., 1989, 54, 1713.
23 S. Kielich, Molecularna Optyka Nieliniowa, Nauka, Moscow,
Russ. edn., 1981, p. 197.
24 S. M. Arakeljan and Y. S. Chilingarjan, Nonlinear Optics of
L iquid Crystals, Nauka, Moscow, 1984, p. 275.
25 I. D. W. Samuel, I. Ledoux, C. Delporte, D. L. Pearson and J. M.
Tour, Chem. Mater., 1996, 8, 819.
The nonlinear optical and photorefractive properties of
organicÈinorganic hybrids made from the polymers by the
solÈgel technique are at present being studied and are
expected to be of interest in view of the very high solution s(3)
values reported here. Polymer 5 contains trimethoxysilyl
26 H. K. Shim, K. S. Lee and J. I. Jin, Macromol. Chem. Phys., 1996,
197, 3501.
27 T. Saegusa and Y. Chujo, Makromol. Chem. Macromol. Symp.,
1992, 64, 1.
28 B. A. Bushuk, personal communication.
Phys. Chem. Chem. Phys., 2000, 2, 3195È3201
3201