Synthesis, characterization and third-order nonlinear optical properties of symmetrical ferrocenyl Schiff base materials

Article abstract of DOI:10.1016/j.cplett.2015.01.057

Six symmetrical ferrocenyl Schiff base materials were synthesized and characterized by UV, ¹H NMR, mass spectrometry (MS) and elemental analysis. Their off-resonant third-order nonlinear optical properties were measured using femtosecond laser

Full text of DOI:10.1016/j.cplett.2015.01.057

Chemical Physics Letters 624 (2015) 47–52
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Synthesis, characterization and third-order nonlinear optical
properties of symmetrical ferrocenyl Schiff base materials
Weiguo Yua,b, Jianhong Jiaa, Jianrong Gaoa,∗, Liang Hana, Yujin Lia
a State Key Laboratory Breeding Base of Green Chemistry-Synthesis Technology, Zhejiang University of Technology, Hangzhou 310014, PR China
b Zhejiang Pharmaceutical College, Ningbo 315100, PR China
a r t i c l e i n f o
a b s t r a c t
Six symmetrical ferrocenyl Schiff base materials were synthesized and characterized by UV, 1H NMR,
mass spectrometry (MS) and elemental analysis. Their off-resonant third-order nonlinear optical prop-
erties were measured using femtosecond laser and degenerate four-wave mixing (DFWM) technique. The
third-order nonlinear optical susceptibilities ꢀ(3) were 1.961–6.363 × 10−13 esu. The nonlinear refractive
indexes n2 were 3.609–11.716 × 10−12 esu. The second-order hyperpolarizabilities ␥ of these molecules
were 1.967–6.388 × 10−31 esu. The response time were 45.759–73.079 fs. The results indicate that these
materials have potential nonlinear optical applications.
Article history:
Received 10 September 2014
In final form 31 January 2015
Available online 11 February 2015
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
double bonds ( C N ). The electron donating units in the
push–pull system are usually aromatic rings, heteroaromatics and
other chromophores. However, few studies have been carried
out on the synthesis and optical nonlinearities of symmetrical
ferrocene derivatives, which could become a good optical appli-
cation. Therefore, it is essential for this research to be thoroughly
explored. Our interest is particularly focused on investigating
whether introducing various aryl symmetrical Schiff base groups
into ferrocene molecule will have huge impact on the nonlinear
optical response. Herein, We report the design and synthesis of
a series of six symmetrical ferrocenyl Schiff base compounds 1–6
(Figure 1) which possess the bridge of carbon–nitrogen double
There is a flurry of interest in the research and development
with large third-order nonlinear optical (NLO) materials due to
their potential applications prospects [1–4]. Metal organic NLO
materials are of typical ones amongst them. A great deal of metal
organic nonlinear optical materials have been investigated in the
past decades and well known for their third-order nonlinear prop-
erties, i.e. metal-substituted phthalocyanines and poly(metal-ynes)
[5–9]. These materials have been investigated as alternatives to
inorganic species due to their low cost, fast and large nonlinear
response over broad frequency range, inherent synthetic flexibility,
high optical-damage threshold, and intrinsic tailorability [10–12].
As one type of metal organic materials, ferrocene metal organic
compounds have unique electrochemical and optical properties,
and thus showing great potential applications in optical [13–17]
and other high-tech fields. NLO properties are key indexes in study-
ing organic NLO materials. The third-order nonlinear properties of
materials have attracted extensive interest [18–22].
bonds (
C N ), being expected to greatly enhance their third-
order optical nonlinearities. Our six representative compounds
contain not only phenyl, substituted phenyl with electron-donating
and electron-withdrawing groups, but also heterocycles. These
compounds were characterized by UV, 1H NMR, MS and elemental
analysis. Their third-order NLO properties were measured using
DFMW technique (Figure 2). The results indicate that these fer-
rocenyl derivatives possess large and ultrafast third-order NLO
responses, showing great potential applications in photonics.
Ferrocene derivative NLO materials are generally formed by
the connection of the electron donating and withdrawing groups
through the ␲-electron bridge, and this unique structural feature
helps them obtain a higher nonlinear optical response [23,24].
For this reason, researchers have focused on synthesizing var-
ious compounds containing ␲-electron bridge which includes
2. Experimental
2.1. Preparation and structural characterization
carbon–carbon double bonds
( C C ) and carbon-nitrogen
2.1.1. Synthesis of ferrocene-1,1ꢀ-dicarbaldehyde
The synthetic method was based on the literature reports [25].
1H NMR (500 MHz, CDCl3) ı 9.950 (s, 1H, –CHO), 4.887 (s, 2H, Cp-H),
4.676 (s, 2H, Cp-H).
∗
Corresponding author.
E-mail address: 183010318@qq.com (J. Gao).
http://dx.doi.org/10.1016/j.cplett.2015.01.057
0009-2614/© 2015 Elsevier B.V. All rights reserved.

48
W. Yu et al. / Chemical Physics Letters 624 (2015) 47–52
Figure 1. The synthetic routes of ferrocenyl Schiff base derivatives.
2.1.2. Procedure for synthesis of Compounds 1–6
Compound 3 was washed with petroleum ether and ethyl
Ferrocene-1,1ꢀ-dicarbaldehyde (0.22 g, 0.90 mmol), aromatic
amine(2.14 mmol), and 75 mL xylene were sequentially added to
a 150 mL three-neck round bottom flask. When the ferrocene-1,1ꢀ-
dicarbaldehyde was completely dissolved with stirring vigorously,
0.2 g basic Al2O3 (1.96 mmol) was added. The reaction mixture was
heated to reflux for 8 h and then filtered. Solvent was evaporated
and the product was washed with solvent, affording a correspond-
ing product in a certain yield.
acetate to afford a red solid product in 89.9% yield. m.p. 90–91 ◦C. 1H
NMR (500 MHz, CDCl3): ı 8.282 (s, 2H, CH
N ), 7.236 (t, J = 8 Hz,
2H, Ar-H), 7.163 (d, J = 8 Hz, 2H, Ar-H), 7.101 (d, J = 1.5 Hz, 2H, Ar-H),
6.975 (d, J = 8 Hz, 2H, Ar-H), 4.887 (s, 4H, Cp-H), 4.574 (s, 4H, Cp-H).
MS (ESI): m/z = 461.04(M+H)+. Elem. Anal. Calcd for C24H18N2Cl2Fe:
C, 62.51; H, 3.93; N, 6.07. Found: C, 62.41; H, 3.89; N, 6.14.
Compound 4 was washed with petroleum ether to give a red
solid product in 49.9% yield. m.p. 114–115 ◦C.1H NMR (500 MHz,
Compound 1 was synthesized according to the general proce-
dure above. After the reaction was completed, the reaction mixture
was concentrated to give red oil product which was washed with
petroleum ether and dichloromethane, giving a red needle crys-
talline solid in 89.9% yield. m.p. 92–93 ◦C. 1H NMR (500 MHz,
CDCl3): ı 8.249 (s, 2H, CH N ), 6.620 (d, J = 1.5 Hz, 4H, Ar-H),
6.605 (s, 2H,A-rH), 4.892 (s, 4H, Cp-H), 4.591 (s, 4H,Cp-H). MS (ESI):
m/z = 465.2 (M+H)+. Elem. Anal. Calcd for C24H16N2F4Fe: C, 62.09;
H, 3.47; N, 6.03. Found: C, 62.18; H, 3.58; N, 6.01.
Compound 5 was washed with petroleum ether and ethyl
CDCl3): ı 8.325 (s, 2H, CH
N ), 7.333 (t, J = 8 Hz, 4H, Ar-H), 7.204
acetate to afford a red solid product in 45.2% yield. m.p. 188–189 ◦C.
(t, J = 7.5 Hz,2H, Ar-H), 7.132 (t, J = 7.5 Hz, 4H, Ar-H), 4.890 (t, J = 2 Hz,
4H, Cp-H), 4.556 (t, J = 4 Hz, 2H, Cp-H). MS (ESI): m/z = 393.1 (M+H)+.
Elem. Anal. Calcd for C24H20N2Fe: C, 73.48; H, 5.14; N, 7.14. Found:
C, 73.29; H, 5.26; N, 7.09.
1H NMR (500 MHz, CDCl3):ı 8.341 (s, 2H, CH
N ), 7.977 (d,
J = 8 Hz, 2H, Ar-H), 7.833 (d, J = 1.5 Hz, 2H, Ar-H), 7.410 (t, J = 8 Hz,
2H, Ar-H), 7.365 (d, J = 8 Hz,1H, Ar-H),4.963(d, J = 1.5 Hz, 4H, Cp-H),
4.631 (t, J = 1.5 Hz, 4H,Cp-H). MS (ESI): m/z = 482.98(M+H)+. Elem.
Anal. Calcd for C24H18N4O4Fe: C, 59.77; H, 3.76; N, 11.62. Found: C,
59.64; H, 3.79; N, 11.55.
Compound 2 was washed with petroleum ether to give reddish
brown oil product in 24.3% yield. 1H NMR (500 MHz, CDCl3): ı 8.326
(s, 2H, CH
N
), 7.230 (t, J = 8 Hz, 2H, Ar-H), 6.802 (d, J = 10 Hz, 2H,
Compound 6 was synthesized following the general procedure
described above. After the reaction was completed, the reaction
mixture was concentrated under vacuum to afford red oil crude,
the crude product was washed with petroleum ether and ethanol
to give a black powder-like solid product in 20.5% yield. m.p.
Ar-H), 6.761 (d, J = 8 Hz, 2H, Ar-H), 6.710 (d, J = 9.5 Hz, 2H, Ar-H),
4.879 (s, 4H, Cp-H), 4.558 (s, 4H,Cp-H), 3.809 (s,6H, –OCH3). MS
(ESI): m/z = 463.09 (M+H)+. Elem. Anal. Calcd for C26H24N2O2Fe: C,
69.04; H, 5.35; N, 6.19. Found: C, 68.95; H, 5.41; N, 6.13.
Figure 2. Experimental setup of DFWM.

W. Yu et al. / Chemical Physics Letters 624 (2015) 47–52
49
91–93 ◦C. 1H NMR (500 MHz, CDCl3): ı 9.955 (s, 2H, CH
N ),
7.775–6.962 (m, 8H,Ar-H), 6.605 (s, 2H, Ar-H), 4.896 (s, 2H, Cp-
H), 4.642 (s,2H,Cp-H). MS (ESI):cacld for (M+H)+:473.1169 found
m/z = 473.1174. Elem. Anal. Calcd for C26H18N6Fe: C, 66.40; H, 3.86;
N, 17.87. Found: C, 66.29; H, 3.82; N, 17.92.
2.2. Nonlinear optical properties measurement
DFWM technique is a conventional method for measuring both
electronic and dynamic nonlinearities. The theories and working
procedures for DFWM have been shown and discussed in many
literature reports [26–32]. In our letter, the third-order NLO prop-
erties were measured using femtosecond DFWM technique, with
a Ti: Sapphire laser. The laser power of our measurements was
0.8 mj, and the pulse width was determined to be 40 fs on a SSA25
autocorrelator. The operating wavelength was centered at 800 nm,
and the repetition rate of the pulses was 1 kHz. Figure 2 shows
the experimental setup. In this setup, ‘BS’ stands for beam split-
ter, M1–M5 represent specific mirrors in the experiments. There
are also four beams in this setup, and the wave-vectors k1, k2,
k3, k4 represent the two split beams and the other two gener-
ated beams, respectively. It is essential that the ωi of these four
beams are approximately the same (␻ indicates the frequency of
i
a beam), that is ω1 = ω2 = ω3 = ω4. k3 and k4 were generated from
the two split beams (k1,k2) being overlapped and mixed spatially in
the samples. And the relations of these four wave-vectors were as
follows: k3 = 2k1 − k2, k4 = 2k2 − k1 (2k1, 2k2 represent the second-
harmonic generation of k1, k2). The intensity of k3 or k4 indicates
the third-order nonlinear optical properties of the corresponding
material.
During the measurement, the laser was set to be stable. The
input beam was split into two beams k1 and k2 with nearly equal
energy by use of a beam splitter, then focused on a plot of the sam-
ple. The beam k2 passed through a delay line driven by a stepping
motor so that the optical path length difference between k2 and
k1 beams could be adjusted during the measurement. The angle
between beams k1 and k2 was around or less 5◦. Beams k3 and k4
were generated when k1 and k2 were overlapped spatially in the
sample. k4 passed through an aperture, recorded by a photodiode
and subsequently analyzed by a Lock-in amplifier and computer.
In our experiments, the third-order nonlinear optical suscepti-
bility ꢀ(3) was measured via a comparison with that of a reference
sample CS2. This value was calculated from the DFWM signal (I),
which originated from the beams k4 processed by Lock-in ampli-
fier and computer. During experiments, we found that the signal
intensities values were quite different for various samples, how-
ever it did litter impacts on the calculations and final results. For
these six compounds, the measuring intensities ranged from 20 a.u.
to 50 a.u. in our work. The DFWM signal curve for CS2 and signal for
solution (dichloromethane) are available in Support Information.
Figure 3. The UV–vis absorption spectra of Compounds 1–6.
the sample thickness (L) and absorption correction factor using the
following equation [33]
ꢀ ꢁ
ꢀ
ꢁ
Is 1/2 Lr ns 2 ˛L exp(˛L/2)
ꢀ(s3)
=
ꢀ(r3)
Ir
Ls nr
1 − exp(−˛L)
where the subscripts ‘s’ and ‘r’ represent the parameters for the
sample and CS2. And ˛ is for the linear absorption coefficient.
The fraction (˛Lexp(˛L/2)/1 − exp(− ˛L)) comes from the sample
absorption and is equal to 1 approximately while the sample has
little absorption around the used laser wavelength. Therefore, the
above equation can be simplified as the following,
ꢀ ꢁ
ꢀ
ꢁ
Is 1/2 Lr ns
2
ꢀ(s3)
=
ꢀ(r3)
.
Ir
Ls nr
(3)
The values of ꢀr and nr for CS2 are 6.8 × 10−13 esu and 1.632,
respectively [34]. The nonlinear refractive index n2 in isotropic
media is estimated through the equation [35] n2(esu) = 12ꢁꢀ(3)/ns
,
2
3. Results and discussion
where ns is the linear refractive index of the measured samples
Figure 3 displays the UV–visible absorption spectra for six
compounds dissolved in dichloromethane. The laser wavelength
(800 nm) used in the experiment of DFWM was out of the
absorption region. In order for the off-resonant third-order opti-
cal nonlinearities of the compounds to be measured. The solvent
dichloromethane has no nonlinear signal under the light intensity
used, so the third-order optical nonlinearities measured come from
those six compounds only.
solution, measured by 2WAJ Abbe refractometer.
The second-order hyperpolarizability
ꢂ of a molecule in
isotropic media is related to the solutionꢀ(3) by [36]: ꢂ = ꢀ(3)/(Nf4),
f = (ns2 + 2)/3 where N is the number density of the solute per
milliliter, and f4 is the local field correction factor (ns is the linear
refractive index of the measured samples solution as above).
The temporal response of the phase conjugate signal as a func-
tion of the delay time of the input beam is shown in Figure 4. The
curves are obtained via fitting the time convolution between auto-
correlation function of pulse and single exponent decline function
exp[t/(−T2)]. The response time (ꢃ) of the samples is corresponded
The third-order nonlinear optical susceptibility ꢀ(3) is measured
via a comparison with that of a reference sample CS2. This value was
calculated from the DFWM signal (I), the linear refractive index (n),

50
W. Yu et al. / Chemical Physics Letters 624 (2015) 47–52
Figure 4. DFWM signal versus delay time for Compounds 1–6 in dichloromethane solution.
Table 1
The relevant parameters values for Compounds 1–6.
Compounda
ꢄmax (nm)
εmax (L mol−1 cm−1
)
n
ꢀ(3) (10−13 esu)
n2 (10−12 esu)
ꢂ (10−31 esu)
ꢃ (fs)
1
2
3
4
5
6
335
365
363
368
285
265
4702
4684
4781
4859
4902
4895
1.4306
1.4308
1.4309
1.4306
1.4305
1.4309
3.307
1.961
3.935
5.713
6.363
5.470
6.088
3.609
7.242
10.518
11.716
10.067
3.319
1.967
3.946
5.733
6.388
5.483
49.842
49.705
45.759
60.836
51.342
73.079
a
The experimental data except ꢄmax and εmax were obtained in 5.0 × 10−4 mol/L solution.

W. Yu et al. / Chemical Physics Letters 624 (2015) 47–52
51
to the half-peak width, and the intensity of the optical conjugate
wave (I) is corresponded to the peak height. The values of ꢀ(3), n2,
ꢂ and response time for the samples were deduced and calculated
from the experimental results are listed in Table 1.
thus being fully consistent with the experimental results listed in
Table 1.
In summary, these six compounds exhibit good third-order
nonlinear optical properties with different ꢀ(3), ꢂ, n2 values and
response time due to their individual structure features.
Saswati Ghosal et al. [18] used the femtosecond laser (the
light source is Nd:YAG laser, wavelength 602 nm, pulse width
400 fs) and the optical path of the DFWM experiment for the mea-
surement, and obtained ꢂ (ferrocene) was 1.61 0.18 × 10−35 esu.
Other than that, the representative ꢂ values for 1-styrylferrocene
(FcCH = CH-C6H5) and 1,1ꢀ-distyrylferrocene (Fc(CH = CHC6H5)2)
which possessed carbon–carbon double bond as the bridge group
were 8.55 1.98 × 10−35 esu, 2.70 0.26 × 10−34 esu, respectively.
4. Conclusions
Six organometallic third-order nonlinear optical materials of
the symmetrical ferrocenyl Schiff base were synthesized and char-
acterized. Their off-resonant third-order NLO were investigated
by the DFWM technique. Their large and ultrafast nonlinearities
are believed to originate from the large increase for the degree
of the conjugation system, extensive electronic delocalization and
polarization. The factors such as long conjugation system, for-
mation of ␲–D–␲ structure, strong electron-donating abilities of
substituents, and symmetry of the molecule all contribute to the
high third-order optical nonlinearity.
Moreover, the
ꢂ values for some similar materials (single-
substituted ferrocenyl Schiff bases) reported in the same exper-
imental conditions ranged from 2.10 × 10−31 to 3.15 × 10−31 esu
[37]. However, our results showed that the ␥ values of our
six compounds were in the range of 1.967–6.388 × 10−31 esu.
Obviously, the second-order hyperpolarizabilities of our molecules
(except 2) are several orders or much higher than those results
by researchers listed above. The enhancement of the NLO prop-
erties for our materials by comparison with most reported
compounds is believed to the increasing of the conjugation length
and nitrogen atoms within the molecules possessing lone pair
electrons.
Acknowledgements
The work was supported by the Ningbo Natural Science Founda-
tion, China (grant no. 2013A610095) & Zhejiang Provincial Young
Teachers Program Project, China (grant no. kzyx2010006).
In addition to these ferrocene derivatives 1–6 containing a
conventional electron-donating centered ferrocene ring, the sym-
metrical bis-imino (conjugate ␲ system)groups introduced into
ferrocene display a structure feature of ␲–donor–␲ (␲–D–␲),
which leads to a higher degree conjugation, thus providing a
better chance for charge transfer. As a result, the designed com-
pounds gave better optical properties than the starting material
ferrocene as well as those reported single substituted ferrocene
derivatives. On the other hand, those with 3-chrolobenzene or
3,5-difluorobenzene moieties which contain electron-withdrawing
functionalities, together with carbon–nitrogen double bonds
Appendix A. Supplementary data
Supplementary material related to this article can be found, in
the online version, at doi:10.1016/j.cplett.2015.01.057.
References
[1] J.S. Shirk, J.R. Lindle, F.J. Bartoli, C.A. Hoffman, Z.H. Kafafi, A.W. Snow, Appl. Phys.
Lett. 55 (1989) 1287.
[2] V.A. Margulis, O.B. Tomilin, Synth. Met. 79 (1996) 207.
[3] A.S.L. Gomes, L. Demenicis, D.V. Petrov, C.B. deAraujo, C.P. deMelo, R.
SoutoMaior, Appl. Phys. Lett. 69 (1996) 2166.
(
C N ) can, in theory, significantly interact with the electron-rich
ferrocenyl at the center of the molecules, thus forming a structure
feature of acceptor–␲–donor–␲–acceptor (A–␲–D–␲–A),which
significantly increases the probability of charge transfer. In prin-
ciple, the better the electron acceptor is, the higher degree for
the hyperpolarizability of the molecule will be. Due to the halo-
gen’s dual electronic nature of electron-withdrawing via induction
and electron-donating via conjugation within the molecules, the
decreasing order of electron acceptor is as the following 3,5-
difluro > 3-Cl, which is consistent with the order of 4 > 3. Since
nitro group is a strong electron-withdrawing group, Compound 5
gave high ␥ value without surprise. Compared with the structure
of Compound 1 (␲–D–␲), Compounds 3–5 gave higher ꢂ val-
ues as expected. Whereas compound 2 with an electron-donating
methoxyl group and double aryl imino groups with ferrocene is
believed to adopt a structure feature of donor–␲–donor–␲–donor
(D–␲–D–␲–D). By comparison of D–␲–D–␲–D with A–␲–D–␲–A
structure unit, the latter is obviously more conducive to the flow of
the intramolecular electrons. Therefore, the ꢂ values of compounds
3–5 are all higher than that of compound 2 as expected. Moreover,
D–␲–D–␲–D structure is theoretically more disadvantageous with
regard to the flow of intramolecular electrons than ␲–D–␲ sys-
tem, which is verified by the experimental results, and the ꢂ value
of Compound 2 is even lower than that of Compound 1. Because
of the fused heterocyclic structure unit, compound 6 has higher
degree of electron delocalization with lower delocalization energy
than the benzene ring. As a result, intramolecular charge transfer
with Compound 6 is more prone to occur under the action of exter-
nal electric field. Therefore, compound 6 shows comparatively high
ꢂ value. Consequently, the decreasing order of ꢂ values in princi-
ple for these material molecules is as follows: 5, 4 > 3 > 1 > 2; 6 > 1,
[4] I. Papagiannouli, et al., Chem. Phys. Lett. 554 (2012) 107.
[5] S. Takahashi, H. Morimoto, E. Murata, S. Kataoka, K. Sonogashira, N. Hagihara,
J. Polym. Sci. Polym. Chem. Ed. 20 (1982) 565.
[6] C.C. Frazier, S. Guha, W.P. Chen, M.P. Cockerham, P.L. Porter, E.A. Chauchard,
C.H. Lee, Polymer 28 (1987) 553.
[7] D. Dini, M.J.F. Calvete, M. Hanack, M. Meneghetti, J. Phys. Chem. A 112 (2008)
8515.
[8] M.C. Larciprete, R. Ostuni, A. Belardini, M. Alonzo, G. Leahu, E. Fazio, C. Sibilia,
M. Bertolotti, Photon. Nanostruct. 5 (2007) 73.
[9] M.C. Larciprete, S. Mangialardo, A. Belardini, C. Sibilia, M. Bertolotti, J. Appl.
Phys. 104 (2008) 073109.
[10] W. Sun, M.M. Bader, T. Carvalho, Opt. Commun. 215 (2003) 185.
[11] G. Fu, T. Yoda, K. Kasatani, H. Okamoto, S. Takenaka, Synth. Met. 155 (2005) 68.
[12] P. Aloukos, G. Chatzikyriakos, I. Papagiannouli, N. Liaros, S. Couris, Chem. Phys.
Lett. 495 (2010) 245.
[13] J.A. Delaire, K. Nakatani, Chem. Rev. 100 (2000) 1817.
[14] G. de la Torre, P. Vázquez, F. Agulló-López, T. Torres, Chem. Rev. 104 (2004)
3723.
[15] H.S. Nalwa, Mater. Lett. 33 (1997) 23.
[16] R. Alicante, R. Cases, P. Forcen, L. Oriol, B. Villacampa, J. Polym. Sci. Pol. Chem.
48 (2010) 232.
[17] C.-B. Yao, Y.-D. Zhang, H.-T. Yin, Q.-Y. Meng, C.-Q. Yu, J. Li, P. Yuan, Chem. Phys.
Lett. 576 (2013) 35.
[18] S. Ghosal, M. Samoc, P.N. Prasad, J.J. Tufariello, J. Phys. Chem. 94 (1990) 2847.
[19] G. Li, et al., Inorg. Chem. 42 (2003) 913.
[20] L. Tamayo-Rivera, R. Rangel-Rojo, Y. Mao, W.H. Watson, Opt. Commun. 281
(2008) 5239.
[21] F.E. Hernández, A. Marcano, O. Ysaias Alvarado, A. Biondi, H. Maillotte, Opt.
Commun. 152 (1998) 77.
[22] R. Rangel-Rojo, K. Kimura, H. Matsuda, M.A. Mendez-Rojas, W.H. Watson, Opt.
Commun. 228 (2003) 181.
[23] T.B. Issa, P. Singh, M.V. Baker, J. Power Sources 140 (2005) 388.
[24] T. Kondo, H. Hoshi, K. Honda, Y. Einaga, A. Fujishima, T. Kawai, J. Phys. Chem. C
112 (2008) 11887.
[25] G.G.A. Balavoine, G. Doisneau, T. Fillebeen-Khan, J. Organomet. Chem. 412
(1991) 381.
[26] J.L. Brédas, C. Adant, P. Tackx, A. Persoons, Chem. Rev. 94 (1994) 243.
[27] P.W. Smith, J. Tomlinson, W.J. Eilenberger, D.J.P.J. Maloney, Opt. Lett. 6 (1981)
581.

52
W. Yu et al. / Chemical Physics Letters 624 (2015) 47–52
[28] T.J. Bunning, L.V. Natarajan, M.G. Schmitt, B.L. Epling, R.L. Crane, Appl. Opt. 30
(1991) 4341.
[33] D.F. Eaton, Science 253 (1991) 281.
[34] M.E. Orczyk, M. Samoc, J. Swiatkiewicz, P.N. Prasad, J. Chem. Phys. 98 (1993)
[29] M. Zhao, Y. Cui, M. Samoc, P.N. Prasad, M.R. Unroe, B.A. Reinhardt, J. Chem. Phys.
95 (1991) 3991.
2524.
[35] S.A. Jenekhe, S.K. Lo, S.R. Flom, Appl. Phys. Lett. 54 (1989) 2524.
[30] P.A. Fleitz, R.L. Sutherland, L.V. Natarajan, T. Pottenger, N.C. Fernelius, Opt. Lett.
17 (1992) 716.
[36] B.K. Mandal, B. Bihari, A.K. Sinha, M. Kamath, L. Chen, Appl. Phys. Lett. 66 (1995)
932.
[31] C. Maloney, H. Byrne, W.M. Dennis, W. Blau, Chem. Phys. 121 (1988) 21.
[32] Hari Singh Nalwa, Appl. Organomet. Chem. 5 (1991) 349.
[37] J. Jia, X. Tao, Y. Li, W. Sheng, L. Han, J. Gao, Y. Zheng, Chem. Phys. Lett. 514 (2011)
114.