Nonlinear optical properties of some derivatives of salicylaldimine-based ligands

Article abstract of DOI:10.1016/j.molstruc.2004.06.017

A series of potential nonlinear optical (NLO) compounds, (-OH) and (-Cl) substituted derivatives of salicylaldimine-based ligands 1-6 (Fig. 2), was designed and synthesized. Compounds 1 and 2 were characterized by X-ray diffraction analysis, elemental analysis, IR, ¹H-NMR, ¹³C-NMR and UV-Visible spectroscopies. The electric dipole moments (μ) and the first hyperpolarizabilities (β) of the compounds 1-6 were calculated using ab-initio quantum calculations (finite field M?ller Plesset perturbation theory). The calculation results reveal that the substituent positions play a significant role in NLO properties of these compounds.

Full text of DOI:10.1016/j.molstruc.2004.06.017

Journal of Molecular Structure 702 (2004) 103–110
www.elsevier.com/locate/molstruc
Nonlinear optical properties of some derivatives
of salicylaldimine-based ligands
a
A. Karakas¸ , A. Elmali , H. Unver , I. Svoboda
b,
c
d
¨
*
aDepartment of Physics, Faculty of Arts and Sciences, Selc¸uk University, 42049 Campus, Konya, Turkey
b
˘
Department of Engineering Physics, Faculty of Engineering, Ankara University, 06100 Besevler, Tandogan, Ankara, Turkey
c
˘
Department of Physics, Faculty of Sciences, Ankara University, 06100 Tandogan, Ankara, Turkey
dInstitute for Materials Science, Darmstadt University of Technology, Petersenstrasse 23, D-64287 Darmstadt, Germany
Received 20 February 2004; accepted 9 June 2004
Abstract
A series of potential nonlinear optical (NLO) compounds, (–OH) and (–Cl) substituted derivatives of salicylaldimine-based ligands 1–6
(Fig. 2), was designed and synthesized. Compounds 1 and 2 were characterized by X-ray diffraction analysis, elemental analysis, IR,
1H-NMR, 13C-NMR and UV–Visible spectroscopies. The electric dipole moments ðmÞ and the first hyperpolarizabilities ðbÞ of the
compounds 1–6 were calculated using ab-initio quantum calculations (finite field Møller Plesset perturbation theory). The calculation results
reveal that the substituent positions play a significant role in NLO properties of these compounds.
q 2004 Elsevier B.V. All rights reserved.
Keywords: Nonlinear optical properties; Schiff base ligands; Electric dipole moment; First hyperpolarizability; Ab-initio calculation
1. Introduction
the molecule and therefore in nonlinear properties [5]. In
the present study, we studied the NLO properties of the
series of (–OH) and (–Cl) substituted derivatives of
salicylaldimine-based ligands 1–6 to get more insight and
to better understand the influence of donor–acceptor
substituents positions upon the first hyperpolarizability
values of these compounds. We carried out synthesis,
The molecules with large optical nonlinearities have
recently become the focus of most researches in view of
their potential applications in various photonic technol-
ogies, including all-optical switching and data processing
[1]. It is known that the molecular based macroscopic
p-electron systems possess many attractive nonlinear
optical (NLO) characteristics and show enhanced NLO
properties. The design of most efficient organic materials for
the nonlinear effect is based on molecular units containing
highly delocalized p-electron moieties and extra electron
donor and electron acceptor groups on opposite sides of the
molecule at appropriate positions on the ring to enhance the
conjugation.
1
X-ray structural analysis, elemental analysis, IR, H-NMR,
13C-NMR and UV–Visible spectroscopies of compounds
1 and 2 and ab-initio quantum calculations of compounds
1–6. The electric dipole moments and the first hyperpo-
larizabilities have been calculated in finite field (FF)
approach, which is currently one of the ultimate
procedures for obtaining numerically accurate NLO
responses. To get more reliable predictions of molecular
first hyperpolarizabilities and electric dipole moments, we
have used 3-21 þ G** polarized and diffused basis set
appeared to be a good compromise between accuracy and
calculation duration, and we have computed all these
properties with second-order Møller Plesset perturbation
theory (MP2) accounted sufficient treatment of electron
correlation.
The effect of electron withdrawing and electron
attracting substituents on the first hyperpolarizability b
of conjugated systems has received a great deal of attention
in recent years [2–4]. It was shown that the type of
substituent plays a major role in change transfer through
*
Corresponding author. Fax: þ90-312-2232395.
E-mail address: elmali@eng.ankara.edu.tr (A. Elmali).
0022-2860/$ - see front matter q 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2004.06.017

104
A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
2. Experiment
Table 1
Crystal and experimental data for the compounds 1 and 2
2.1. Reagent and techniques
Compound 1
13H10NOCl
Compound 2
2-Hydroxy-1-salicylaldehyde, 3-chloroaniline, 4-chlor-
oaniline, ethanol and CDCl3 were purchased from Merck
(Germany). Methanol and acetonitril were purchased from
Aldrich Chemical Co. Melting point was measured on a
Gallonkamp apparatus using a capillary tube. The elemen-
tal analysis was performed on a LECO CHNS-932 C-, H-,
N-analyser. Infrared absorption spectra were obtained from
a Mattson 1000 FT-IR spectrometer in KBr pellets and
reported in cm21. UV–Visible spectra were measured
using a Perkin Elmer Lambda 2 series spectrophotometer.
1H- and 13C-NMR spectra were recorded on a Bruker DPX
FT-NMR spectrometer (400 MHz) operating at a proton
frequency of 400.5 MHz and a carbon frequency of
100.7 MHz using CDCl3 solvent and tetramethylsilane as
internal standard.
Empirical formula
Colour
C
C13H10NOCl
Yellow
Yellow
Formula weight
Crystal system
Space group
231.67 g/M
Orthorhombic
Pca21
231.67 g/M
Monoclinic
P21/c
ꢀ
ꢀ
Unit cell dimensions
a ¼ 25:499ð5Þ A
a ¼ 13:432ð2Þ A
ꢀ
ꢀ
b ¼ 3:8980ð10Þ A
b ¼ 5:787ð1Þ A
b ¼ 106:12ð1Þ8
ꢀ
ꢀ
c ¼ 10:8060ð10Þ A
c ¼ 14:736ð2Þ A
3
˚
3
˚
Volume
1074.1(4) A
1100.4(4) A
Z
4
4
Calculated
1.433 g cm23
0.330 mm21
7897/2906
½RðintÞ ¼ 0:0691ꢀ
2906/1/149
1.398 g cm23
0.322 mm21
6568/2224
½RðintÞ ¼ 0:0464ꢀ
2224/1/149
Absorption coefficient
Reflections
collected/unique
Data/restraints/
parameters
Goodness-of-fit on F2
0.948
1.060
Final R indices ½I . 2sðIÞꢀ
R1 ¼ 0:0528;
wR2 ¼ 0:1303
R1 ¼ 0:0751;
wR2 ¼ 0:1459
0.269 and
R1 ¼ 0:0539;
wR2 ¼ 0:1550
R1 ¼ 0:0662;
wR2 ¼ 0:1683
0.264 and
2.2. Preparation of compounds 1 and 2
R indices (all data)
N-(3-chloro)-salicylaldimine 1 was prepared by con-
densation of 2-hydroxy-1-salicylaldehyde (0.01 mol
1.22 g) and 3-chloroaniline (0.01 mol 1.28 g) in 100 ml
of ethanol. The reaction mixture was stirred for 4 h, and
then placed in a freezer for 18 h. The yellow precipitate
was collected by filtration, and then washed with cold
ethanol: mp 89.5 8C, 2.11 g (89%) yields. Found: C, 67.89;
H, 4.71; N, 5.77. C13H10NOCl; C, 67.40; H, 4.35; N,
6.05%. N-(4-chloro)-salicylaldimine 2 was also syn-
thesized with same method: mp 101.6 8C, 2.02 g (85%)
yields. Found: C, 67.36; H, 4.58; N, 5.73. C13H10NOCl; C,
67.40; H, 4.35; N, 6.05%.
Largest diff. peak and hole
20.287 e A23
20.306 e A23
for the structures reported in this paper have been
deposited with the Cambridge Crystallographic Data
Centre as supplementary publication no. CCDC 231864
and 231866 [10].
Table 2
˚
Some selected bond lengths (A) and angles (8)
Compound 1
Compound 2
C(7)–N(1)
C(1)–N(1)
C(7)–C(8)
C(3)–C(4)
C(8)–C(13)
C(9)–C(10)
C(10)–C(11)
C(11)–C(12)
C(12)–C(13)
C(13)–O(1)
C(3)–Cl(1)
C(3)–Cl(1)
1.290(3)
1.401(3)
1.431(3)
1.366(4)
1.400(3)
1.372(4)
1.383(5)
1.372(4)
1.385(4)
1.353(3)
1.737(3)
1.271(2)
1.413(2)
1.449(3)
1.380(4)
1.406(3)
1.376(3)
1.381(3)
1.375(3)
1.388(3)
1.346(2)
2.3. X-ray structure determination
The data collection for both compounds was performed
on a Enraf-Nonius diffractometer employing graphite-
ꢀ
monochromatized Mo Ka radiation (l ¼ 0:71073 A). The
details of the X-ray data collection, structure solution and
structure refinement for both compounds were given in
Table 1. Data reduction and corrections for absorption
and crystal decomposition (1.2%) for compound 1 and
(0.9%) for compound 2 were achieved using the Nonius
Diffractometer Control Software [6]. The structure was
solved by SHELXS-97 [7] and refined with SHELXL-97 [8].
The positions of H atoms bonded to C atoms were
1.734(2)
C(6)–C(1)–C(2)
C(3)–C(4)–C(5)
N(1)–C(7)–C(8)
C(9)–C(8)–C(13)
C(9)–C(8)–C(7)
C(13)–C(8)–C(7)
C(9)–C(10)–C(11)
C(12)–C(11)–C(10)
O(1)–C(13)–C(12)
O(1)–C(13)–C(8)
C(12)–C(13)–C(8)
C(7)–N(1)–C(1)
118.6(2)
118.5(2)
122.2(2)
118.6(2)
119.3(2)
122.0(2)
119.5(3)
120.7(2)
118.6(2)
121.4(2)
119.9(2)
121.3(2)
118.6(2)
121.3(2)
122.2(2)
118.7(2)
119.8(2)
121.5(2)
119.2(2)
120.9(2)
119.0(2)
121.6(2)
119.4(2)
122.5(2)
˚
calculated (C–H distance 0.93 A), and included in the
structure factor calculation using a riding model. The
hydroxy atoms HO1 for the compounds 1 and 2 were
found from difference in Fourier map calculated at the
refinement process as a small positive electron density.
Bond distances and angles are listed in Table 2 and an
ORTEP view of the molecular structure is given in Fig. 1
[9]. Crystallographic data (excluding structure factors)

A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
105
Fig. 2. Chemical structure of the compounds 1–6.
4. Results and discussion
4.1. IR and NMR studies
The vibrational bands for compounds 1 and 2 are
observed in the 500–3500 cm21. The IR spectrum of the
Fig. 1. The molecular structure of the compounds 1 and 2. The displacement
ellipsoids are plotted at the 50% probability level [13].
compounds 1 and 2 reveals a band at 3415–3432 cm21
respectively, due to n(O–H). For the compounds 1 and 2,
,
3. Theoretical calculations
the absorption bands with the wave numbers of
3020–3070, (nC–H, Ar–H) and 1461–1597 cm21 (nCyC
)
As the first step of our ab-initio calculation, we performed
the geometry optimization with 3-21 þ G** polarized and
diffused basis set. And then, the electric dipole moments and
the b tensor components of the studied compounds were
calculated by using the FF MP2 method at the 3-21 þ G**
basis set level, which has been found to be more than
adequate for obtaining reliable trends in the first hyperpolar-
izability values. All m and b calculations of Schiff bases
studied here were performed using GAUSSIAN98W [11], on an
Intel Pentium IV 1.7 GHz processor with 512 MB RAM and
Microsoft windows as the operating system.
are observed. The absorption band assignable to the
stretching of CyN bond was observed at frequency 1620
(for 1) and 1624 (for 2) these values confirm with the
reported in the IR spectrum for substituted aromatic
Schiff base which possesses the formulation of C6H4-
OHCHyNC6H4Cl [12]. The observation of phenolic nC–O
at 1328 for the compound 1 and 1332 cm21 for the
compound 2 is the evidence for the existence of the enol
form (N· · ·H–O) intramolecular hydrogen bonding only
in the solid state [13,14].
In the solution state of compounds, the existence of the
intramolecular hydrogen bonding (N· · ·H–O) has been
1
confirmed by H-NMR spectroscopy [15,16]. The entire
We report btot (total first hyperpolarizability) for all the
molecules studied here. The components of the first hyperpo-
larizability can be calculated using the following equation:
compounds studied that the azomethine proton resonance are
a single broad signal. Spectroscopic data suggest that all
compounds exist mainly in the enol form in a weakly polar
solvent (CDCl3). The 1H-NMR data for the compound 1 are
d ¼ 12:69 ppm OH 1H s, d ¼ 8:40 ppm s CHyN 1H, Ar–H
7.26–6.77 ppm 8H m, and for the compound 2 ¼ 12.84 ppm
OH 1H s, d ¼ 8:40 ppm s CHyN 1H, Ar–H 7.24–6.76 ppm
8H m.
X
bi ¼ biii þ 1=3 ðbijj þ bjij þ bjjiÞ
ð1Þ
i–j
Using the x; y and z components, the magnitude of the first
hyperpolarizability tensor can be calculated by:
btot ¼ ðb2x þ by2 þ b2z Þ1=2
ð2Þ
According to the 13C-NMR spectra compounds 1 and 2
have 13 and 11 signals. 13C-NMR data; for the compound
1, 164.10 (C HyN), 161.57 ppm (Cipso, –OH) and
150.19, 135.45, 134.07, 133.01, 130.85, 127.22, 121.67,
120.17, 119.69, 119.35, 117.76 ppm (Ar–C) and for
the compound 2, 163.35 ppm (C HyN), 161.53 ppm
(Cipso, –OH) and 147.34, 133.88, 132.88, 130.22,
129.94, 123.16, 122.56, 119.65, 117.72 (Ar–C) are
The complete equation for calculating the magnitude of first
hyperpolarizability from GAUSSIAN98W output is given as
follows [1]:
2
2
btot ¼ ½ðbxxx þ bxyy þ bxzzÞ þ ðbyyy þ byzz þ byxx
2 1=2
Þ
þ ðbzzz þ bzxx þ bzyyÞ ꢀ
ð3Þ
To calculate all the electric dipole moments and all the first
hyperpolarizabilities, the origin of the cartesian coordinate
system ðx; y; zÞ ¼ ð0; 0; 0Þ has been chosen at own center of
mass of each Schiff bases in Fig. 2.
1
observed. In conclusion, H and 13C-NMR results show
that in deuterochloroform solution the compounds 1 and 2
exist in the enol form.

106
A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
4.2. UV–visible spectroscopy
region as the cut-off wavelength is around 398 nm, and
show more than 70% transmittance. The m-nitrobenzoic
acid crystals are transparent up to 390 nm. Both crystals
showed reasonable NLO peak intensity measured by Kurtz
powder technique at the cut-off wavelengths near the UV
region. Ren et al. [24] have shown that all absorption
maxima of the square-pyramidal zinc complexes are
located in UV region. All of the compounds exhibit
solvatochromism, i.e. their maximal absorption peaks show
hypsochromic or bathochromic shifts. It is well known that
solvatochromism is based on two-level model, and is valid
in a large number of organic NLO materials. It is
reasonable to say that the solvatochromism in the
investigated complexes indicates the change of dipole
moments in the ground and excited states verifying the
existence of intramolecular charge transfer and non-zero
NLO responses. The NLO properties and UV–Visible
studies of push–pull ferrocene complexes containing
heteroaromatic rings in the conjugation chain were
investigated by Justin Thomas et al. [18]. High bath-
ochromic shift generally considered as indicative of high b
values, and hence potential NLO properties has been
observed in dichloromethane. Di Bella et al. [25] have
reported bis(salicylaldiminato) nickel(II) compound exhib-
ited interesting linear optical spectroscopic features which
will be seen to be related to the second-order NLO
response. The optical absorption spectrum (.280 nm) of
archetypical Ni(salen) consists of two overlapping principal
features assigned to two different types of transitions. There
is a broad band in the region between 300 and 360 nm
involving mainly p ! pp transitions, and a relatively
intense structure in the 400–480 nm region. The band at
around 400 nm exhibits a solvatochromic shift, character-
istic of a large dipole moment, and frequently suggestive of
a large hyperpolarizability. Moreover, a negative solvato-
chromism, i.e. a hypsochromic shift is observed, thus
indicating a reduction of the dipole moment upon
electronic excitation.
It can be very helpful in the investigation of NLO
materials making it possible to check, apart from NLO
responses, also spectroscopic absorbance in the appropriate
wavelength. Thus, the wavelengths obtained by
UV–Visible spectral analysis can be helpful in planning
the synthesis of the promising NLO materials only [17].
The solution electronic absorption spectral studies of
compounds designed to possess NLO properties are
important for two specific reasons. Firstly, it is necessary
to know the transparency region. Secondly, the solvato-
chromic behavior of the sample is generally considered as
indicative of high molecular first hyperpolarizability [18].
Zhou [19] has concluded that the estimated electronic
transition wavelengths obtained by UV–Visible spectral
analysis for chain compounds are about 10 nm shorter than
that of the cycles, and the wavelengths for both groups
studied have been found to be shorter than 400 nm,
implying good transparency and NLO property. The
average values of the second-order hyperpolarizability
coefficients per molecular unit increase with increasing
units, in nonlinear manner. And the values of lmax
corresponding p ! pp transitions for both groups are
estimated to be shorter than 400 nm, i.e. almost within
354–371 nm, resulting solvatochromic behavior they
would possess in the future NLO applications. Albert
et al. [20] have reached the conclusion that with the correct
substitution of the push–pull system in the porphyrin ring,
characterized by strong intramolecular p ! pp charge
transfer transitions found through UV–Visible spectral
analysis, some specific electronic and structural properties
of this system could produce high NLO responses. Zhou
et al. [21] have found that the lmax results of novel para-
phenylenealkyne macrocycles are not accelerated with the
odd number of unit, even with 10 units the value of lmax is
360 nm. The bathochromic shift with the increase in
the number of unit is very small, suggesting their superior
optical transparency in the visible regime. Though the lmax
was estimated to be shorter than 400 nm in an enough large
sample, a strong increase in the hyperpolarizability value is
obtained with a sizable increase. The linear optical (UV–
Visible) and NLO properties of donor–acceptor substituted
diazabutadienes and hexatrienes have been investigated in
a combined theoretical and experimental study by Dworc-
zak et al. [22]. The theoretical calculations lead to a more
pronounced bathochromic effect and increased b values. If
one considers the azines described by Dworczak et al. as
polymethine-type compounds, replacement of a methine
carbon at an even position by an electronegative element
should not only lead to a bathochromic shift of the longest
wavelength absorption band ðlmax ¼ 394 nmÞ but also to
an increase in b: Rai et al. [23] measured the transparency
of the UNBA and m-nitrobenzoic acid crystals from 850 to
200 nm by using the UV–Visible spectrometer. The
UNBA crystals are quite transparent almost till the UV
The UV–Visible spectra of the compounds 1 and 2 have
been studied in polar solvents dimethylsulfoxide (DMSO),
methanol, tetrahydrofuran (THF), chloroform. Figs. 3 and 4
show the UV–Visible absorption spectra of the compounds
1 and 2, respectively, in the polar solvents used. It is seen
that the compounds 1 and 2 have given very strong
absorption band in the UV region. Table 3 shows the
maximum absorption wavelengths ðlÞ and molar extinction
coefficients ð1Þ obtained from the UV–Visible spectral
analysis of both compounds in all the solvents used. The
validity of FF MP2 approximation used in all the
computations here might also be illustrated by analyzing
the relationship between calculated values of b and
measured values of l: The compounds 1 and 2 exhibit
solvatochromism (see Figs. 3 and 4), i.e. maximal absorp-
tion peaks of both compounds show bathochromic behavior
with band shift ,400 nm and all the transitions are p ! pp
generally considered as indicative of molecular first

A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
107
Table 3
The maximum absorption wavelengths and molar extinction coefficients l;
nm (1; M21 cm21), respectively, obtained from the UV–Visible spectral
analysis of compounds 1 and 2 in solvents of different polarity
Compound Solvents [l; nm (1; M21 cm21)]
DMSO
Methanol
THF
Chloroform
1
2
341.2(8560) 340.0(6907)
320.0(8045) 318.3(6549)
274.6(9469) 300.0(6376)
270.9(8541)
342.3(7760) 341.7(6164)
318.3(6988) 320.0(5567)
300.0(6923) 268.9(237)
273.3(9562)
230.0(12035)
211.4(15423)
339.8(6620) 340.6(20156) 342.5(6564) 340.0(11594)
316.6(6092) 316.6(19036) 316.6(5985) 316.6(10724)
272.9(6092) 300.0(16990) 298.3(5917) 271.3(14460)
270.0(18264) 272.3(8060)
230.3(28227)
210.8(22628)
almost within 1–4 nm among the compounds 1 and 2 (see
Table 3). However, it is seen that the substituent position has
effected rather much the molecular polarities, and hence the
electric dipole moments (Table 5) and btot values (Table 4)
obtained by the numerical second-derivatives of electric
dipole moments according to the applied field strength in FF
approach. The maximum absorption wavelengths of the
synthesized compounds found shorter than 400 nm show
that such compounds might have non-zero rather high first
hyperpolarizabilities [19]. According to these results with a
great possibility, the compounds 3–6 having the same
donor–acceptor substituents with the compounds 1 and 2
could also yield similar properties for l and b values.
Fig. 3. UV–Visible absorption spectra of the compound 1 in DMSO,
chloroform, methanol and THF.
hyperpolarizabilities with non-zero values [19,20]. It is
found that the maximum wavelengths of p ! pp transition
for the compounds 1 and 2 are all almost within
211–342 nm (see Table 3). The substituent position has
not changed the l values so much, i.e. indicating differences
4.3. Description of the crystal structure
The extent of the planarity differs for both molecules 1
and 2. The molecule 1 is not planar and the two Schiff base
moieties (C7–C13, O1) [planar with a maximum deviation
˚
of 20.016(2) A for the O1 atom] and (N1, C1–C6, Cl1)
˚
[planar with a maximum deviation of 20.019(1) A for the
N1 atom] are inclined at angle of 6.44(9)8. However, the
molecule 2 is nearly planar and the two Schiff base moieties
(C7–C13, O1) [planar with a maximum deviation of
˚
0.016(1) A for the C7 atom] and (N1, C1–C6, Cl1) [planar
˚
with a maximum deviation of 20.010(1) A for the N1 atom]
are inclined at angle of 1.65(9)8.
Two types of intramolecular hydrogen bond (either
O–H· · ·N or O· · ·H–N) can occur in Schiff bases [26]. The
Schiff bases derived from salicylaldehyde always form
the O–H· · ·N type of hydrogen bonding regardless of
N-substituent (alkyl or aryl) [27]. The O· · ·N distances
˚
˚
[2.608(3) A for 1 and 2.617(2) A for 2] indicate a strong
intramolecular hydrogen bond for both compounds. The H
atoms are essentially bonded to the O atoms. Clearly, for the
compounds 1 and 2, the enol tautomer is favoured rather
Fig. 4. UV–Visible absorption spectra of the compound 2 in DMSO,
chloroform, methanol and THF.

108
A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
Table 4
All b components and btot value calculated at FF MP2 level using 3-21 þ G** basis set by GAUSSIAN98W for all Schiff bases studied
Compound bxxx
bxxy
bxyy
byyy
bxxz
bxyz
byyz
bxzz
byzz
bzzz
btot £ 10230 ðesuÞ
1
7.133
21040.779 2200.285
4.786 22.118 20.450 20.077
2249.692 2197.409
21.036 20.159 20.185
260.807
21.500
111.924 210.245 2669.898 267.472
0.183
44.760
0.523 239.345
112.565
2
3
4
5
6
16.797 21.156
906.511
266.744
270.365
259.932
384.978 2140.824
21.092
134.213
22.372
20.443
56.875
0.575 257.111
272.861
1.636 264.384
170.923 329.852
138.561
818.631
153.209
586.695
2.483 20.196
0.929 20.092
21.546 2295.275 289.538
30.730
1.183
22.011
2.271
54.985
473.493
255.163 23.579 20.644 2183.805 2209.777
294.132
than the keto form. These are evident from the observed
position (Fig. 2), we find that btot decreases in the order
meta . para . ortho in compounds 5, 3 and 1, respect-
ively, according to the positions of (–OH) groups (see
Table 4). According to the relative strength for ortho and
para resonance interactions, which are larger than for meta
interactions, one would expect best btot results for donor and
acceptor groups in positions ortho . para . meta, respect-
ively [29]. In the case of chloro substitution at the para
position, the resonance effects may dominate the inductive
effect. While the (–Cl) substituent is at the para position
(Fig. 2), btot decreases in the order ortho . para . meta in
compounds 2, 4 and 6, respectively, according to the
positions of (–OH) group (see Table 4).
˚
C13–O1 bond distances of 1.353(3) A for 1 and 1.346(2) A
˚
for 2, which are consistent with a single bond, and the
˚
˚
N1yC7 bond distances of 1.290(3) A for 1 and 1.271(2) A
for 2, which are indicative of a double bond. From the X-ray
diffraction experiment, we were only able to detect the
existence of the enol tautomer.
4.4. Computational results and discussion
Computational approach allows the determination of
molecular NLO properties as an inexpensive way to design
molecules by analyzing their potential before synthesis.
In addition to the well known empirical rules to estimate
qualitatively, the microscopic nonlinear response in organic
molecules, especially ab-initio MP2 calculations allow a
more accurate prediction of the NLO activity.
While the (–Cl) substituent is at the meta and para
positions (Fig. 2), we find that m decreases in the order
para . meta . ortho in compounds 3, 5, 1 and in the
order ortho . meta . para in compounds 2, 6, 4,
respectively, according to the positions of (–OH) group
(see Table 5). The presence of (–OH) group at the ortho
position, while the (–Cl) substituent is at the para
position in molecular structure, causes the drastic increase
of the ground state dipole moment and thus of the btot
value (see Tables 4 and 5). Our calculations indicate that
the compound 2 might be the b-interesting material. It is
shown that all the investigated compounds have great
non-zero m values (see Table 5). Because the studied
compounds 1–6 are the polar molecules having non-zero
dipole moments, it is seen that these m values in Table 5
have yielded non-zero btot values. Compounds 1 and 2
have the lowest and the greatest m values, respectively,
shown in Table 5 as well as btot values. The higher dipole
moment values are associated, in general, with even larger
All the aromatic systems studied have both donor
hydroxy (–OH) and acceptor chloro (–Cl) substituents
(Fig. 2). We calculated the electric dipole moments m and
the first hyperpolarizabilities b of salicylaldimine-based
Schiff base ligands with acceptor substituent (–Cl) at meta
and para positions, and with donor (–OH) group at ortho,
meta and para positions. The investigated compounds are
quite unique because they contain very similar structures
differing only by position of the donor and acceptor groups
of the aromatic ring.
It is shown that meta and ortho substituted compounds
are much less linear, and their btot values are even less than
the sum of the individual values of their mono substituted
derivatives [28]. In the case of chloro substitution, there is
an orbital–orbital replusive interaction between the (–OH)
bonding orbital and the (–Cl) atom. The nitrogen orbital is
ellipsoidal with a small cross-sectional area, combine to
produce a nearly spherical orbital. For chloro substituents at
the meta position in compounds 1, 3 and 5 (Fig. 2), the
inductive effects may be predominant as compared to
resonance effects. Also, the effect of polarization, size of
electron density distribution and configuration are so
balanced that the system is aligned for maximum charge
transfer. The nitrogen of Schiff bases with (–Cl) groups at
meta position affects the basicity of its lone-pair electrons to
a much smaller extent that they affect the acidity of the
(–OH) group. While the (–Cl) substituent is at the meta
Table 5
The ab-initio calculated the electric dipole moments m (total m) and dipole
moment components (mx; my and mz) for the compounds 1–6
Total dipole moments [m(Debye)] and dipole moment components
1
2
3
4
5
6
mx
my
0.5674
0.9446
8.7481 210.8506
3.9070 29.3852
5.2751
7.9870
0.2140
3.6698
0.0333 20.1102
11.4545 4.6251
2.4727
4.7508 21.9772
0.0589 20.1513
mz 20.1693
m
1.1149 11.8476
10.5192
5.6355

A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
109
projection of btot quantities [30]. The dipoles may oppose
or enhance one another or at least, bring the dipoles the
required or out of the required net alignment necessary for
NLO properties such as btot values. The connection between
the electric dipole moments of an organic molecule having
donor–acceptor substituents and first hyperpolarizability is
widely recognized in Refs. [2,31–34]. Several research
groups have tried to identify molecules with potentially
optimal nonlinearities through the two-level model. For
example, Marder et al. [35–38] used a four-site Hu¨ckel
model to examine how each of the two-level parameters
varies with the electron donating and electron accepting
abilities of appended substituents. The b responses derived
from this model were not optimized with maximal electronic
asymmetry unique to a given bridge structure. The maximum
was due to the behavior of non-zero m value. One of their
conclusions obtained from this work is that non-zero m value
might permit to find non-zero b value. In this study, so that
the first hyperpolarizabilities are computed by the numerical
second-derivatives of the electric dipole moments according
to the applied field strength in FF approach, there are rather
strong relationship among the calculated m and btot values of
the compounds 1–6. Therefore, these m values of the
compounds in Table 5 may be responsible for enhancing and
decreasing the btot values given in Table 4.
at MP2 3-21 þ G** level encourage the future use of these
compounds (Fig. 2) for electro-optics applications.
Our computational results yield that the donor–acceptor
substitution at appropriate positions enhances the btot and m
values (Tables 4 and 5). While the best btot results might be
obtained from the compounds having ortho and para
resonance interactions [29], Cheng et al. [28] have found
that the btot values of the compounds with substituents at
meta and ortho positions are rather small. In addition, we
have also computed that the compound 1 with substituents
at meta and ortho positions and, the compound 2 with ortho
and para resonance interactions have more interesting m and
btot values found as the smallest and the greatest for the
compounds 1 and 2, respectively, than the other investigated
compounds.
NLO properties for organic compounds are related to the
wavelength of absorption: molecules having a large l of
absorption show good hyperpolarizabilities. Thus, our
theoretical btot values could be clearly correlated with
spectroscopic data of UV–Visible in order to understand the
molecular structure–NLO property relationship in view of a
future optimization of the macroscopic NLO properties of
this family of Schiff bases. It has been found that the
synthesized the compounds 1 and 2 have rather great btot
and l values having shorter than 400 nm, implying good
optical nonlinearity.
The ab-initio calculated btot and m values for all the
compounds show that also it could be interesting to
synthesize the compounds 1 and 2 having the greatest and
the lowest, respectively, btot and m values. It is important to
stress that, in these btot investigations, we do not take into
account the effect of the field on the nuclear positions, i.e.
we evaluate only the electronic component of btot: The
vibrational contributions which, for conjugated systems, can
be important according to the NLO process are left for
further investigations.
Acknowledgements
The authors wish to acknowledge that this work was
supported by the Research Funds of Selc¸uk University and
Ankara University under grant numbers 2003/030 and CHE
2003 00 00 041, respectively.
References
5. Conclusions
[1] K.S. Thanthiriwatte, K.M. Nalin de Silva, J. Mol. Struct. (Theochem.)
617 (2002) 169.
In this paper, the electric dipole moments and the first
hyperpolarizabilities of the compounds studied have been
calculated with 3-21 þ G** basis set in FF approach. All
these properties have been computed by MP2 perturbation
theory. Computational experience on moderately sized
molecules as studied compounds here shows that the
determination of reference self-consistent field is an
essential part of the ab-initio calculation of (hyper)polariz-
abilities. With ab-initio FF MP2 computations, it is also
possible to predict btot values accurately for given
molecular structures, or to find good correlations between
computations and experiments. The ab-initio calculated
non-zero m values of the compounds 1–6 show that these
molecules might have microscopic first hyperpolarizabil-
ities with non-zero values obtained by FF MP2 method as
numerical derivatives of the dipole moments. Rather great
btot and m values derived from the theoretical calculations
[2] S.R. Marder, J.E. Sohn, G.D. Stucky (Eds.), Materials for Nonlinear
Optics: Chemical Perspectives, ACS Symposium Series 455, Amer-
ican Chemical Society, Washington, DC, 1991.
[3] M. Blanchard-Desce, C. Runser, A. Fort, M. Barzoukas, J.M. Lehn, V.
Bloy, V. Alain, Chem. Phys. 199 (1995) 253.
[4] G. Raos, M. Del Zoppo, J. Mol. Struct. (Theochem.) 439 (2002) 589.
[5] M. Jalali-Heravi, A.A. Khandar, I. Sheikshoaie, Spectrochim. Acta A
55 (1999) 2537.
[6] Enraf-Nonius diffractometer control software, Release 5.1., Enraf-
Nonius, Delft, Netherlands, (1993).
[7] G.M. Sheldrick, SHELXS-97, Program for the Solution of Crystal
Structures, University of Goettingen, Germany, 1997.
[8] G.M. Sheldrick, SHELXL-97, Program for the Refinement of Crystal
Structures, University of Goettingen, Germany, 1997.
[9] L.J. Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
[10] Further information may be obtained from: Cambridge Crystal-
lographic Data Center (CCDC), 12 Union Road, Cambridge CB21EZ,
UK, by quoting the depository numbers CCDC-231864 and 231866,
e-mail: deposit@ccdc.cam.ac.uk.

110
A. Karakas¸ et al. / Journal of Molecular Structure 702 (2004) 103–110
[11] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, Jr., R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi,
R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J.
Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K.
Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski,
J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith,
M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A.
Pople, GAUSSIAN98, Revision A.7, Gaussian Inc., Pittsburgh, PA,
1998.
[24] P. Ren, T. Liu, J. Qin, C. Chen, Spectrochim. Acta A 59 (2003) 1095.
[25] S. Di Bella, I. Fragala, I. Ledoux, M.A. Diaz-Garcia, P.G. Lacroix,
T.J. Marks, Chem. Mater. 6 (1994) 881.
[26] A.D. Garnowskii, A.L. Nivorozhkin, V.I. Minkin, Coord. Chem. Rev.
126 (1993) 1.
[27] M. Gavranic, B. Kaitner, E. Mestrovic, J. Chem. Crystallogr. 26
(1996) 23.
[28] L.T. Cheng, W. Tam, G.R. Meredith, G.L. Rikken, E.W. Meijer, Proc.
SPIE 1147 (1989) 61.
[29] S. Hillebrand, M. Segala, T. Buckup, R.B. Correia, F. Horowitz, V.
Stefani, Chem. Phys. 273 (2001) 1.
[30] V. Lukes, M. Breza, D. Vegh, P. Hrdlovic, V. Laurinc, Synth. Met.
138 (2003) 399.
[31] N.P. Prasad, D.J. Williams, Introduction to Nonlinear Optical Effects
in Molecules and Polymers, Wiley, New York, 1991.
[32] J. Messier, F. Kajar, P. Prasad, D. Ulrich (Eds.), Nonlinear Optical
Effects in Organic Polymers, Kluwer Academic Publishers, Dor-
drecht, 1989.
[12] G.Y. Yeap, S.T. Ha, N. Ishizawa, K. Suda, P.L. Boey, W.A.K.
Mahmood, J. Mol. Struct. 658 (2003) 87.
[13] P. Nagy, R. Harzfeld, Spectrosc. Lett. 31 (1998) 221.
[14] S.R. Salman, N.A.I. Salah, Spectrosc. Lett. 30 (1997) 1289.
[15] G.C. Perey, D.A. Thornton, J. Inorg. Nucl. Chem. 34 (1972) 3357.
[16] G.J.M. Fernandez, F. Del Rio-Portilla, B. Quiroz-Garcia, R.A.
Toscano, R. Salcedo, J. Mol. Struct. 561 (2001) 197.
[33] D.S. Chemla, J. Zyss (Eds.), Nonlinear Optical Properties of Organic
Molecules and Crystals, vols. 1 and 2, Academic Press, New York,
1987.
[17] J. Kulakowska, S. Kucharski, Eur. Polym. J. 36 (2000) 1805.
[18] K.R. Justin Thomas, J.T. Lin, Y.S. Wen, J. Organomet. Chem. 575
(1999) 301.
[34] G. Carters, J. Zyss (Eds.), Nonlinear Optical Processes in Organic
Materials, J. Opt. Soc. Am. B, vol. 4, 1987.
[35] S.R. Marder, D.N. Beratan, L.T. Cheng, Science 252 (1991) 103.
[36] S.M. Risser, D.N. Beratan, S.R. Marder, J. Am. Chem. Soc. 115
(1993) 7719.
[19] Y. Zhou, Mater. Sci. Eng. B 99 (2003) 593.
[20] I.D.L. Albert, T.J. Marks, M.A. Ratner, Chem. Mater. 10 (1998) 753.
[21] Y. Zhou, S. Feng, Z. Xie, Opt. Mater. 24 (2003) 667.
[22] R. Dworczak, W.M.F. Fabian, C. Reidlinger, A. Rumpler, J.
Schachner, K. Zangger, Spectrochim. Acta A 58 (2002) 2135.
[23] R.N. Rai, P. Ramasamy, C.W. Lan, J. Cryst. Growth 235 (2002) 499.
[37] S.R. Marder, L.T. Cheng, D.N. Beratan, AIP Conf. Proc. 262
(1992) 252.
[38] S.R. Marder, L.T. Cheng, B.G. Tiemann, D.N. Beratan, Proc. SPIE-
Int. Soc. Opt. Eng. 1560 (1991) 86.