Chemical dynamics and reverse saturable absorption in di-furfuryl ether solutions

Article abstract of DOI:10.1016/S0009-2614(01)00760-6

We report on the chemical dynamics and nonlinear optical absorption of di-furfuryl ether (DFE) solutions. DFE molecules were derivatized from the room temperature furfuryl alcohol self-reaction, whose reaction kinetics was investigated with the Z-scan technique and laser-induced fluorescence. These molecules are shown to exhibit a pronounced reverse saturable absorption process. The magnitude of the excited-state absorption cross-sections and the intersystem crossing lifetime have been evaluated with 8 ns pulses at 532 nm.

Full text of DOI:10.1016/S0009-2614(01)00760-6

1
0 August 2001
Chemical Physics Letters 343 +2001) 499±506
www.elsevier.com/locate/cplett
Chemical dynamics and reverse saturable absorption
in di-furfuryl ether solutions
*
C.R. Mendonßca, P.S. Batista, M.F. de Souza, S.C. Zilio
ꢀ
Instituto de Fisica de S a~ o Carlos, Universidade de S a~ o Paulo, Caixa Postal 369, 13560-970 S a~ o Carlos, SP, Brazil
Received 27 December 2000; in ®nal form 25 April 2001
Abstract
We report on the chemical dynamics and nonlinear optical absorption of di-furfuryl ether +DFE) solutions. DFE
molecules were derivatized from the room temperature furfuryl alcohol self-reaction, whose reaction kinetics was in-
vestigated with the Z-scan technique and laser-induced ¯uorescence. These molecules are shown to exhibit a pro-
nounced reverse saturable absorption process. The magnitude of the excited-state absorption cross-sections and the
intersystem crossing lifetime have been evaluated with 8 ns pulses at 532 nm. Ó 2001 Published by Elsevier Science B.V.
1
. Introduction
of compounds is based on the furfuryl alcohol
FA), a natural and renewable source of mono-
+
Organic materials have attracted considerable
mers that originates from the chemical transfor-
mation of polyoses present in corn-cobs, cotton,
rice-husk and sugar-cane [8,9]. Although FA
conjugated polymers constitute a new class of
conductive materials with potential technological
applications [10,11], their most interesting prop-
erties cannot be fully explored because the back-
bone usually presents rami®cations that are
deleterious to the linear conjugation [6,7]. There-
fore, the understanding of the FA polymerization
reaction, which is assisted by the condensation
induced by an acid [12], is important to avoid the
rami®cation mechanism, whose reduction could
lead to the improvement of the polymer's con-
ductive and mechanical properties. Moreover, the
control of the oligomer's size [13], and conse-
quently of the conjugation length, could allow the
enhancement of the nonlinear optical response. In
the course of investigations on the rami®cation
mechanism, we veri®ed the presence of molecules
arising from a self-reaction process that could
interest because their large nonlinear susceptibili-
ties and fast response times are promising for use
in optical signal processing devices [1±3], and also
due to their potential application as passive optical
limiters to protect sensors and eyes from intense
light pulses [4]. Among organic materials, p-con-
jugated polymers such as poly+aniline), poly+thy-
ophene) and poly+furylene vinylene) have been the
subject of extensive theoretical and experimental
investigation aiming at understanding and im-
proving their design, synthesis and optical prop-
erties [5]. A class of p-conjugated systems
containing furan rings in the backbone have also
received considerable attention because they are
employed to synthesize polymers and resins that
can be used in coating processes [6,7]. This family
*
Corresponding author. Fax: ꢀ55-16-271-3616.
E-mail address: zilio@if.sc.usp.br +S.C. Zilio).
0
009-2614/01/$ - see front matter Ó 2001 Published by Elsevier Science B.V.
PII: S 0 0 0 9 - 2 6 1 4 + 0 1 ) 0 0 7 6 0 - 6

500
C.R. Mendon ßc a et al. / Chemical Physics Letters 343 +2001) 499±506
a€ect the polymerization evolution. Besides, we
casually found that the self-reaction product pre-
sents a strong saturable absorption +RSA) mech-
anism. This observation motivated us to
investigate in detail the formation of derivatized
molecules arising as consequence of the FA self-
reaction and their behavior as passive optical
limiters. The use of several analytical techniques
allowed us to identify the newly formed species as
the di-furfuryl ether +DFE) and its function iso-
mer, 5-furfuryl-furfuryl alcohol +5FFA). In order
to gain further insight on the self-reaction process,
we employed the Z-scan technique with Fourier
analysis [14] and laser-induced ¯uorescence +LIF)
to study its kinetics. To our knowledge, this is the
Fig. 1. Absorbance +À log ꢀI=I †) spectra of freshly distilled
0
10
FA +solid line) and after aging for several days at room tem-
perature +dashed line). The inset shows molecules of FA +a),
DFE +b) and 5FFA +c).
®
rst time that the sensitive nonlinear Z-scan tech-
nique was used to follow the time evolution of a
chemical reaction. With the combination of these
two techniques we can monitor the presence of
di€erent products during the FA self-reaction;
while the Z-scan measurement detects all types of
molecules that absorbs in the visible and give rise
to thermal lensing, the LIF measurement is more
selective because it detects just the DFE. We were
also able to ®nd that in a later stage, DFE mole-
cules combine to form tetramers in such way that
their concentration remains constant. Finally, we
performed nonlinear absorption measurements to
analyze the reverse saturable absorption process in
DFE molecules, which is an important feature for
the manufacture of optical limiting devices.
spectra were measured with a Hitachi F-4500 ¯u-
orescence spectrometer, by pumping the sample
with light at 500 nm. In order to identify the de-
rivatized molecules arising at the earlier stage of
the self-reaction process, we carried out chro-
matographic analyses with a Shimadzu GC-17A,
using hydrogen as the mobile phase and thermal
conductive detection +TCD) in column DB1, and
mass spectrometry performed in a Shimadzu GC-
MS QP5000.
The self-reaction kinetics was studied with both
LIF and Z-scan techniques. The former uses an
argon ion laser operating at 514.5 nm to excite the
sample. The resulting ¯uorescence is collected by a
15 cm focal length lens and ®ltered by a high-pass
2
. Experimental
®lter placed in front of a photomultiplier tube to
collect the wavelength-integrated signal. On the
other hand, the Z-scan technique [15] was used as
a probe for non-luminescent molecules. Basically,
the method consists in translating the nonlinear
sample through the focal plane of a tightly focused
Gaussian beam and monitoring the changes in the
far ®eld intensity pattern. For a purely refractive
nonlinearity, the light ®eld induces an intensity
dependent nonlinear phase and as consequence of
the transverse Gaussian intensity pro®le, the
sample presents a lens-like behavior. The induced
self-phase modulation has the tendency of defo-
cussing or recollimating the incident beam, de-
pending on its z position with respect to the focal
The commercially available furfuryl alcohol
Sigma-Aldrich) was puri®ed by low-pressure dis-
+
tillation, resulting in a colorless liquid that was
subsequently dried in Na SO . Soon after, the in-
2
4
13
frared absorption and C-NMR spectra were,
respectively, measured with a Bomem-Mb102 and
a Brucker UX-200 spectrometers. The results ob-
tained were identical to those already reported in
the literature [6,7], meaning that we are indeed
working with the pure FA schematically shown in
inset of Fig. 1. UV-Vis spectra were obtained with
a Hitachi U2001 spectrometer, with the solution
placed in a 2 mm quartz cell, while emission

C.R. Mendon cß a et al. / Chemical Physics Letters 343 +2001) 499±506
501
plane. By monitoring the transmittance change
through a small circular aperture placed at the
far-®eld position, one is able to determine the
sorption band centered at 270 nm, which is known
Ã
to be related to the furan ring p±p transition [12].
However, when the solution ages for several days
at room temperature and illumination, it presents
a bathochromic shift due to its self-reaction [6,9],
as shown by the dashed line in Fig. 1. By means of
chromatographic analyses carried out after 6 h of
aging, we found 99.56 wt% of pure FA, but we
also found 0.30 and 0.14 wt% of two other com-
pounds. These unknown compounds were isolated
and concentrated to have their masses determined.
The resulting fragmentation spectra are similar to
those reported in [17,18], and the structures ob-
served were directly associated with the DFE and
its function isomer, 5FFA, both shown in the inset
of Fig. 1.
2
nonlinear refractive index, n . Any nonlinear ab-
sorption present at the sample can be found in a
measurement where the aperture is removed.
However, our samples did not present any non-
linear absorption when exposed to a low power cw
laser beam and we restricted our measurements to
the nonlinear refractive index.
For the present measurements, we employed the
Z-scan technique with Fourier analysis [14,16] to
investigate the nonlinear refractive index of the FA
solution. The experimental setup is fully described
in [14] and only a few details are presented here. As
light sources we have used an argon ion laser +488
and 514.5 nm) and a cw dye laser operating with
rodhamine 590 +570±615 nm). The laser beam is
modulated at a frequency f with a mechanical
chopper and the transmission through the aperture
located at the far ®eld is monitored as a function
of the sample z-position. The Fourier components
at f and 2f, measured with two independent lock-
ins, are respectively, related to linear and nonlinear
refraction, and their ratio, Rꢀz; f †, allows to elim-
inate spurious linear e€ects. The dependence of R
on f can be used to distinguish nonlinear processes
with di€erent relaxation times and allows the
determination of the thermal di€usion coecient
in thermal processes.
Each isolated compound was diluted in DCM
to allow the transmission of light through the
sample, which otherwise would be opaque at the
UV region. The UV-Vis spectra of the isolated
compounds show that only DFE molecules absorb
appreciably in the visible, as depicted in the left-
hand side of Fig. 2. This spectrum can be decon-
voluted and results in three bands centered at 275,
340 and 400 nm, as shown by dashed lines in the
®gure. On the other hand, the 5FFA presents
negligible absorption in the visible and the only
important band observed was associated to the
Ã
furan ring p±p transition lying at the ultraviolet.
The light absorption by DFE molecules results in
¯uorescence emission, whose lineshape strongly
Nonlinear absorption measurements were car-
ried out with a Q-switched Nd:YAG laser pro-
ducing 8-ns-duration +FWHM) pulses at 532 nm.
The beam with a 3 mm diameter was transmitted
through a 10 mm quartz cell containing a dilute
À4
solution of DFE ꢀ3  10 M† in dichloro-meth-
ane +DCM). The transmission was characterized
as a function of the input laser ¯uence by rotating
the ®rst of two Glan laser-polarizers placed in
front of the cell.
3
. Results and discussion
3
.1. Derivatized molecules characterization
Fig. 2. Absorbance +À log ꢀI=I †) and ¯uorescence spectra of
0
10
The UV-Vis spectrum for freshly distilled FA is
depicted in Fig. 1. It presents just a broad ab-
DFE in a DCM solution. This particular emission lineshape is
obtained for excitation light at 500 nm.

502
C.R. Mendon ßc a et al. / Chemical Physics Letters 343 +2001) 499±506
depends on the excitation wavelength due to the
existence of three absorption bands. As an exam-
ple, the right-hand side of Fig. 2 shows the emis-
sion spectrum obtained when the sample is
pumped by light at 500 nm, which excites just the
tail of the longer wavelength absorption band.
When the excitation is performed at 450 nm, two
emission bands are observed. On the other hand,
since the absorption of the 5FFA molecule is
negligible at the visible, no ¯uorescence is ob-
served. Accordingly, when the aging FA solution
is exposed to light from an argon ion laser oper-
ating at 514.5 nm, the ¯uorescence intensity can be
used to determine the number of DFE derivatized
molecules. However, since the aging sample con-
tains mostly FA, we have to make sure that the
Ã
emission is not related to a direct p±p transition
due to a possible two-photon absorption process
occurring in these molecules. For this purpose, we
measured the ¯uorescence intensity as a function
of the argon laser power. The linear behavior ob-
served indicates that no two-photon process is
present and the ¯uorescence can be attributed to a
small amount of DFE molecules absorbing and
emitting in the visible region.
Fig. 3. +a) Dependence of DZpv on the chopper frequency and
theoretical ®t +solid line) with a function 50=f ‡ 1:7 and +b)
DRpv versus the modulation frequency and theoretical ®t +solid
line) provided by Eq. +4) of [14]. The excitation wavelength is
514.5 nm.
3
.2. Z-Scan measurements
As we just mentioned, LIF can be used to
monitor the number of DFE derivatized molecules
because they have a suciently high ¯uorescence
quantum yield. On the other hand, the Z-scan
technique was chosen as a probe for non-lumi-
nescent molecules because of its high sensitivity to
measure any light-induced refractive index change.
To determine the origin of the nonlinear index, we
performed a set of Z-scan measurements at 514.5
nm with di€erent modulation frequencies. We
used a FA sample that aged for 6 h after distilla-
tion, as in the chromatographic analysis. The ob-
quencies, the heat di€usion is considerable and the
thermally induced lens becomes broader than the
Gaussian pro®le of the incident beam, resulting in
a larger DZ . However, at the high frequency re-
pv
gime there is not enough time for the heat di€u-
sion to occur and the thermal lens follows the
Gaussian pro®le of the beam. In this case, DZpv is
narrower and tends to the value 1:7 z , predicted
0
by Sheik-Bahae et al. [15] as the frequency be-
comes much larger than the inverse of the thermal
relaxation time. De®ning DRpv as the di€erence of
Rꢀz; f † between the peak and valley positions, we
served nonlinearity is negative ꢀn < 0†, which is
2
typical for thermal nonlinearities in liquid sam-
ples. Fig. 3a shows the dependence of the peak
and valley separation, DZpv, on the modulation
measured the dependence of DR on f, as shown
pv
in Fig. 3b.
frequency. The solid line represents a function
À1
The thermal di€usion coecient, D ˆ K =qC
T
p
f
dependence indicates that the nonlinear e€ect has
‡ constant, used to ®t the data as in [16]. This
+KT being the thermal conductivity, q the mass
density and C the heat capacity) and the nonlinear
p
indeed a thermal origin. For low modulation fre-
refractive index, n th, can be obtained using Eq. +4)
2

C.R. Mendon cß a et al. / Chemical Physics Letters 343 +2001) 499±506
503
À4
2
of [14]. We found D ˆ 7:0  10 cm =s, which is
in good agreement with values presented in the
ˆ
literature for the pure furfuryl alcohol +K
T
À4
À1 À1
À3
4
:28  10 cal cm s , q ˆ 1:12 g cm
and
À1
À1
C ˆ 0:502 cal g °C ) [19], and the thermal
p
À7
2
nonlinear index as n th ˆ À6:9  10 cm =J.
2
The thermal lensing monitored by the Z-scan
technique can be attributed to species formed
during the FA self-reaction, which are responsible
for the absorption of light and subsequent heat
generation. We have also carried out Z-scan
measurements as a function of the wavelength to
con®rm the thermal origin of the observed e€ect.
We observed that n enhances in the blue region
2
where the energy of the incident photon increases,
as expected for a thermal nonlinearity arising from
a radiative processes. As the light frequency in-
creases, the quantum eciency becomes smaller
and the excess energy carried by the photon is
converted into heat.
3
.3. Self-reaction kinetics
The concentration and number of species de-
pends on the FA aging stage. After several days or
weeks, a large number of di€erent species are
produced and this makes the analysis quite com-
plex. However, we are interested in the initial stage
of degradation, where just the two compounds
observed in gas chromatography are present and
are responsible for the optical process observed.
As the self-reaction evolves, the concentration of
derivatized molecules increases in a way that we
want to characterize. As discussed previously, the
Fig. 4. +a) Fluorescence and +b) Z-scan signal in an FA sample
aging at room temperature. The excitation wavelength is 514.5
nm.
observed. The time for which the Z-scan signal
changes its growth rate corresponds to that where
the ¯uorescence intensity becomes constant.
To explain this behavior, we propose a model
where two molecules of FA initially combine to
produce either one DFE or one 5FFA molecule.
The linear increase in the DFE concentration ob-
served in both probe signals is related to the fact
that the concentration of FA is not depleted be-
cause the amount of derivatized molecules is small.
The rate equation describing the time-evolution of
the DFE concentration is given by:
5
FFA is invisible to the optical probes employed
because it neither absorb nor emit when pumped
by 514.5 nm light. Therefore, both Z-scan and
¯
uorescence signals are proportional just to the
concentration of DFE. Fig. 4+a) shows that the
uorescence intensity increases linearly for the ®rst
10 h of aging and after that it becomes constant.
¯
1
From the chromatographic analysis carried out
after 6 h of aging we estimate that the ¯uorescence
stabilizes at a DFE concentration corresponding
to about 5 wt% of that of FA. Fig. 4b presents the
d‰NDFEŠ
2
ˆ
k
1
‰NFAŠ ;
ꢀ1†
dt
where ‰N Š and ‰N Š are, respectively, the DFE
DFE
FA
time evolution of DR , measured at f ˆ 50 Hz. In
pv
and FA concentrations, and k is the rate constant
1
this case, the signal always increases linearly in
time, but two di€erent growth rates can be clearly
for dimerization. Assuming that the NFA concen-
tration is constant, we obtain the linear time-

504
C.R. Mendon ßc a et al. / Chemical Physics Letters 343 +2001) 499±506
2
dependent behavior as ‰NDFEŠ ˆ k
1
‰NFAŠ t. Addi-
3.4. Reverse saturable absorption
tional terms accounting for the reaction between
DFE and FA, and between two DFE are not
considered because they would produce a nonlin-
ear time-dependence, which clearly does not re-
produce our results. Since the DFE concentration
clamps at about 5% after 110 h, we can estimate
The ¯uence-dependent transmission of the di-
lute DFE/DCM solution is represented by full
circles in Fig. 5. This result can be explained by the
®ve-level model depicted in the insert, where the
relaxation from T to S is neglected because it is
1
0
À9
À1 À1
the value of k
1
as being 7 Â 10 l mol s .
much longer than the 8 ns pulse duration. The
upper excited singlet and triplet levels S and T are
assumed to be too short-lived to present any
population build up. The rate equations used to
describe the population densities, n
3), at each level are:
After a critical concentration of dimers is
reached, new reactions will take place, but in this
case, the newly formed DFE and 5FFA are reac-
tants and not products as in the earlier case. In this
second stage of the self-reaction, the new com-
pounds originating from the dimers structures
stabilize the DFE concentration, as seen in the
n
n
i
+i ˆ 0, 1 and
dn
dt
0
n
T
1
ˆ Àn
0
W
01
‡
;
ꢀ3†
ꢀ4†
ꢀ5†
10
¯
uorescence signal depicted in Fig. 4+a). The in-
crease in the growth rate of the Z-scan signal may
have two origins. First, the new derivatized mole-
cules could have an absorption coecient higher
than DFE, but we do not have a way to con®rm
this because the absorbed light fraction is always
very small and cannot be detected. Second, and
probably more important, since the tetramers
formed do not ¯uoresce, the heat emitted through
radiationless transitions is much higher and the
thermal lensing is more notable. The rate equation
for the DFE concentration in this second stage is
more complex because we have, in principle, to
take into account the reaction between two DFE,
and between one DFE and one 5FFA. Again, we
do not consider the reaction between DFE and FA
because it already did not exist in the ®rst stage.
Therefore the new rate equation is:
dn1
dt
n1
n1
ˆ
n
0
W
01 À T À T
;
10
isc
dn
dt
n
Tisc
3
1
ˆ
;
where W ˆ r I=hm is the transition rate for
01
01
S ! S , with r being the ground state cross-
0
1
01
section and I the intensity. T10 is the lifetimes of
level S and Tisc is the intersystem crossing lifetime.
1
This set of equations was numerically solved using
a Gaussian temporal intensity pro®le for the
Q-switched pulse used in our experiment, yielding
the population dynamics, n ꢀt†: The time evolution
i
d‰NDFE
Š
2
2
0
2
ˆ 1
k ‰N Š Àk ‰N Š Àk ‰NDFEŠ‰N5FFAŠˆ0;
FA
2
DFE
dt
ꢀ
2†
0
where k
2
and k are rate constants for the DFE
2
reaction, and just the stationary case was consid-
ered because the ¯uorescence intensity is constant.
Eq. +2) cannot be directly solved because we do not
know the time-evolution of [N5FFA ] and neither if
becomes constant after the ®rst hundred hours.
However, in order to make a rough estimate of k
we will neglect the last term in Eq. +2), which lead
us to the ratio k =k
self-reaction of DFE is more e€ective than that of
FA.
2
,
2
1
ˆ 900. This means that the
Fig. 5. Fluence-dependent transmission of the dilute DFE/
DCM solution +full circles) and theoretical ®tting with the pa-
rameters given in the text +solid line).

C.R. Mendon cß a et al. / Chemical Physics Letters 343 +2001) 499±506
505
of the nonlinear absorption was calculated ac-
cording to:
polymerization process. However, owing to the
low vapor pressure of ethers, DFE molecules can
be removed if the polymerization is carried out in
vacuum. We already tried this approach and ob-
tained good results that will be published else-
where.
Finally, the nonlinear optical absorption of a
DFE/DCM solution was investigated with 8 ns
pulses at 532 nm. The sample presented an RSA
process and the magnitude of the excited-state
absorption cross-sections as well as the intersystem
crossing lifetime have been evaluated.
aꢀt† ˆ Nfn r ‡ n r ‡ n r g;
ꢀ6†
0
01
1
12
3 34
where N is the DFE concentration, and r12 and r34
are the excited state cross-sections. r01 was deter-
mined by measuring the linear absorption at
1
7
À3
5
32 nm +a
0
ˆ Nr01, with N ˆ 2:5  10 cm ) re-
À19
2
sulting in r01 ˆ 3:5  10
cm . The numerical
calculation was carried out with T10 ˆ 2:7 ns, ob-
tained through time-resolved ¯uorescence mea-
surements. The solid line in Fig. 5 represents the
theoretical ®tting obtained with r ˆ 7:5Â
12
À19
2
À17
2
1
0
cm , r ˆ 9  10
cm and T ˆ 150 ns.
34
isc
The intersystem crossing lifetime is almost 10 times
higher than those of other organic molecules tra-
ditionally investigated for optical limiting pur-
poses [20,21]. However, this time could be lowered
with the introduction of a metal atom to increase
the spin-orbit interaction. On the other hand, the
ratio between the excited and ground states cross-
section is nearly 200, a factor of 10 higher than in
metallophthalocyanines and tetrabenzpophyrins,
which is quite convenient for optical limiters. We
are presently characterizing the dynamic optical
nonlinearities in DFE solutions with the Z-scan
technique with pulse trains [22].
Acknowledgements
This research was supported by Fundaßc ~a o de
Amparo ꢁa Pesquisa do Estado de S ~a o Paulo
+FAPESP) and Conselho Nacional de Desen-
ꢀ
volvimento Cienti®co e Tecnol oꢀ gico +CNPq). The
authors thanks Marcel Tabak and Luiz A.O.
Nunes for helping in the course of ¯uorescence
experiments.
References
[
[
1] N. Bloembergen, J. Nonlinear Opt. Phys. Mat. 5 +1996) 1.
2] G.T. Boydm, in: C.P. Wong +Ed.), Polymers for Electronic
and Photonic Applications, Academic, Boston, MA, 1993,
p. 467.
4
. Conclusions
[
3] M.G. Kuzyk, in: C.P. Wong +Ed.), Polymers for Electronic
and Photonic Applications, Academic, Boston, MA, 1993,
p. 507.
We have employed gas chromatography and
mass spectroscopy to determine the derivatized
molecules arising at the earlier stage of the FA self-
reaction. Then, we used two complementary opti-
cal probes to analyze the kinetics followed by the
self-reaction. LIF measurement permitted a direct
determination of the DFE concentration while the
non-selective Z-scan technique monitored the
thermal lensing arising as consequence of the ab-
sorption of any derivatized molecules in the
wavelength used. It also provided information on
the thermal properties of the FA. Although an
equivalent method has already been used to sense
the thermal lens in a chemical reaction [23], this is
the ®rst time that the Z-scan technique is employed
for this purpose. The presence of DFE molecules
and its high reaction rate may be deleterious to the
[
[
4] T.H. Wei, D.J. Hagan, M.J. Sence, E.W. Van Stryland,
J.W. Perry, D.R. Coulter, Appl. Phys. B 54 +1992) 46.
5] P.N. Prasad, D.J. Williams, Introduction to Nonlinear
Optical E€ects in Molecules and Polymers, Wiley, New
York, NY, 1991.
[6] M. Choura, N.M. Belgacem, A. Gandini, Macromolecules
9 +1996) 3839.
7] I.S. Chuang, G.E. Maciel, G.E. Myers, Macromolecules 17
1984) 1087.
2
[
[
+
8] A. Gandini, in: S.L. Aggarwal, S. Russo +Eds.), Compre-
hensive Polymer Science, First Supplement, Pergamom
Press, Oxford, UK, 1992, p. 527.
[
9] A.P. Dunlop, F.N. Peters, in: The Furans, American
Chemical Society Monograph Series, vol. 119, Reinhold
Publishing Corporation, New York, NY, 1953.
[
10] F.B. Benvenuti, A.M.R. Galletti, C. Carlini, G. Sbrana,
P. Brushi, A. Nannini, Polymer 38 +1997) 4973.
[11] P.S. Batista, M.F. De Souza, Synth. Met. 101 +1999) 635.

506
C.R. Mendon ßc a et al. / Chemical Physics Letters 343 +2001) 499±506
ꢀ
ꢀ
[
12] R. Gonz ꢀa lez, R. Martinez, P. Ortis, Makromol. Chem.
Rapid. Commun. 13 +1992) 517.
[18] E.D. Loughran, E.M. Wewerka, G.J. Hammons, J. Het-
erocyclic. Chem. 9 +1972) 57.
[
[
13] P.S. Batista, M.F. De Souza, Polymer 41 +2000) 8263.
14] C.R. Mendonßca, L. Misoguti, S.C. Zilio, Appl. Phys. Lett.
[19] E.W. Flick +Ed.), Industrial Solvents Handbook, Noyes
Data Corporation, New Jersey, 1991, p. 397.
[20] T.C. Wen, L.D. Lian, Synt. Metals 83 +1996) 111.
[21] C.R. Mendonßca, L. Ga€o, L. Misoguti, W.C. Moreira,
O.N. Oliveira Jr., S.C. Zilio, Chem. Phys. Lett. 323 +2000)
300.
7
1 +1997) 2094.
[
[
[
15] M. Sheik-Bahae, A.A. Said, E.W. Van Stryland, Opt. Lett.
4 +1989) 955.
1
16] C.R. Mendonca, M.M. Costa, J.A. Giacometti, F.D.
Nunes, S.C. Zilio, Electron. Lett. 34 +1998) 116.
[22] L. Misoguti, C.R. Mendonßca, S.C. Zilio, Appl. Phys. Lett.
74 +1999) 1531.
[23] R.D. Snook, R.D. Lowe, Analyst 120 +1995) 2051.
17] Q.N. Porter, in: Mass Spectrometry of Heterocyclic Com-
pounds, second ed., Wiley, New York NY,1985, p. 147.