Synthesis, characterization, optical spectroscopic, electronic structure, and second-order nonlinear optical (NLO) properties of a novel class of donor-acceptor bis(salicylaldiminato)nickel(II) Schiff base NLO chromophores

Article abstract of DOI:10.1021/ja971349y

The synthesis, characterization, thermal stability, optical spectroscopic, electronic structure, and second-order nonlinear optical (NLO) response of a series of donor-acceptor bis(salicylaldiminato)nickel(II) Schiff base complexes and the free ligand precursors are reported. The effect of the metal center in such complexes is manifold: it templates the formation of acentric molecular structures, imparts high thermal stability to the chelate ring, and both 'switches on' and enhances NLO response. Metal complexation imparts new linear optical spectroscopic features, having metal-to-ligand charge transfer character, which are responsible for the second-order nonlinearity. Moreover, the present synthetic strategy represents a novel route to inorganic NLO chromophores. Solution-phase hyperpolarizability values, deduced by electric field-induced second-harmonic-generation experiments are as high as -79 x 10^-30 cm⁵ esu^-1 (hω = 0.92 eV). Experimental linear and nonlinear optical features are fully consistent with INDO/SCI-SOS theoretical calculations. They provide a rationale for the NLO response of these materials and are attractive for designing new, highly efficient second-order nonlinear optical inorganic chromophores.

Full text of DOI:10.1021/ja971349y

9550
J. Am. Chem. Soc. 1997, 119, 9550-9557
Synthesis, Characterization, Optical Spectroscopic, Electronic
Structure, and Second-Order Nonlinear Optical (NLO)
Properties of a Novel Class of Donor-Acceptor
Bis(salicylaldiminato)nickel(II) Schiff Base NLO Chromophores
Santo Di Bella,*,† Ignazio Fragala`,*,† Isabelle Ledoux,‡ Maria A. Diaz-Garcia,‡,| and
Tobin J. Marks*,§
Contribution from the Dipartimento di Scienze Chimiche, UniVersita` di Catania, 95125 Catania,
Italy, Molecular Quantum Electronics Department, France Telecom, CNET, 92225 Bagneux,
France, and Department of Chemistry and the Materials Research Center, Northwestern
UniVersity, EVanston, Illinois 60208-3113
ReceiVed April 28, 1997X
Abstract: The synthesis, characterization, thermal stability, optical spectroscopic, electronic structure, and second-
order nonlinear optical (NLO) response of a series of donor-acceptor bis(salicylaldiminato)nickel(II) Schiff base
complexes and the free ligand precursors are reported. The effect of the metal center in such complexes is
manifold: it templates the formation of acentric molecular structures, imparts high thermal stability to the chelate
ring, and both “switches on” and enhances NLO response. Metal complexation imparts new linear optical spectroscopic
features, having metal-to-ligand charge transfer character, which are responsible for the second-order nonlinearity.
Moreover, the present synthetic strategy represents a novel route to inorganic NLO chromophores. Solution-phase
hyperpolarizability values, deduced by electric field-induced second-harmonic-generation experiments are as high
as -79 × 10-30 cm5 esu-1 (hω ) 0.92 eV). Experimental linear and nonlinear optical features are fully consistent
with INDO/SCI-SOS theoretical calculations. They provide a rationale for the NLO response of these materials and
are attractive for designing new, highly efficient second-order nonlinear optical inorganic chromophores.
Introduction
yond early studies focused on understanding and rationalizing
the microscopic molecular nonlinearity, â(-2ω;ω,ω), attention
has more recently been directed toward optimizing NLO
response, toward achieving enhanced chemical and thermal
stability, as well as toward improving processability for device
fabrication.1-3 Crucial prerequisites for achieving large bulk
second-order NLO responses are that the individual constituents
have large molecular responses and that they be arranged in a
noncentrosymmetric architecture. Considerable effort has been
directed toward the molecular engineering of such structures,
and a variety of strategies has emerged. To date, however, this
activity has primarily focused on purely organic systems1-3 and,
to a far lesser extent, on organometallic and coordination
complexes.4,5 Compared to organic molecules, metal complexes
offer a larger variety of structures, the possibility of high
environmental stability, and a diversity of electronic properties
tunable by virtue of the coordinated metal center.
Molecule-based second-order nonlinear optical (NLO) materi-
als have recently attracted much interest because they involve
new scientific phenomena and because they offer potential
applications in emerging optoelectronic technologies.1-3 Be-
† Universita` di Catania.
‡ France Telecom.
§ Northwestern University.
| Present address: Departamento Materiales C-IV, Universidad Auto´noma
de Madrid Cantobianco, Madrid 28049, Spain.
X Abstract published in AdVance ACS Abstracts, September 1, 1997.
(1) (a) Polymers for Second-Order Nonlinear Optics; Lindsay, G. A.,
Singer, K. D., Eds.; ACS Symposium Series 601; American Chemical
Society: Washington, DC, 1995. (b) Molecular Nonlinear Optics; Zyss,
J., Ed.; Academic Press: New York, 1994. (c) Prasad, N. P.; Williams, D.
J. Introduction to Nonlinear Optical Effects in Molecules and Polymers,
Wiley: New York, 1991. (d) Materials for Nonlinear Optics: Chemical
PerspectiVes; Marder, S. R., Sohn, J. E., Stucky, G. D., Eds.; ACS
Symposium Series 455; American Chemical Society: Washington, DC,
1991. (e) Nonlinear Optical Properties of Organic Molecules and Crystals;
Chemla, D. S., Zyss, J., Eds.; Academic Press: New York, 1987; Vols 1
and 2.
We present here a new synthetic strategy for obtaining
thermally stable noncentrosymmetric coordination complexes
(2) For recent reviews on NLO materials see, for example: (a) Ledoux,
I.; Zyss, J. In NoVel Optical Materials and Applications; Khoo, I. C., Simoni,
F., Umeton, C., Eds; John Wiley & Sons: New York, 1997. (b) Dalton, L.
R.; Harper, A. W.; Ghosn, R.; Steir, W. H.; Ziari, M.; Fetterman, H.; Shi,
Y.; Mustacich, R. V.; Jen, A. K.-Y.; Shea, K. J. Chem. Mater. 1995, 7,
1060. (c) Benning, R. G. J. Mater. Chem. 1995, 5, 365. (c) Marks, T. J.;
Ratner, M. A. Angew. Chem., Int. Ed. Engl. 1995, 34, 155. (d) Optical
Nonlinearities in Chemistry; Burland, D. M., Ed., Chem. ReV. 1994, 94
(no. 1), 1-278.
(3) (a) Nonlinear Optical Properties of Organic Materials; Proceedings
of SPIE 1988-1994, Vols. 971, 1147, 1337, 1560, 1775, 2025, 2285. (b)
Organic Materials for Nonlinear Optics III; Hann, R. A., Bloor, D., Eds.;
Royal Society of Chemistry: London, 1992. (c) Organic Materials for
Nonlinear Optics II; Hann, R. A., Bloor, D., Eds.; Royal Society of
Chemistry: London, 1991. (d) Organic Molecules for Nonlinear Optics
and Photonics; Messier, J., Kajzar, F., Prasad, P., Eds.; Kluwer Academic
Publishers: Dordrecht, 1991. (e) Organic Materials for Nonlinear Optics;
Hann, R. A., Bloor, D., Eds.; Royal Society of Chemistry: London, 1989.
(4) (a) For a recent review of NLO studies involving organometallic and
coordination complexes, see: Long, N. J. Angew. Chem., Int. Ed. Engl.
1995, 34, 21. (b) For a review including early NLO studies, see: Nalwa,
H. S. Appl. Organomet. Chem. 1991, 5, 349.
(5) For more recent NLO studies on organometallic and coordination
complexes see, for example: (a) Cummings, S. D.; Cheng, L. T.; Eisenberg,
R. Chem. Mater. 1997, 9, 440. (b) Di Bella, S.; Fragala`, I.; Marks, T. J.;
Ratner, M. A. J. Am. Chem. Soc. 1996, 118, 12747. (c) Lacroix, P. G.; Di
Bella, S.; Ledoux, I. Chem. Mater. 1996, 8, 541. (d) Houbrechts, S.; Clays,
K. Persoons, A.; Cadierno, V.; Gamasa, M. P.; Gimeno, J. Organometallics
1996, 15, 5266. (e) Whittall, I. R.; Humphrey, M. G.; Houbrechts, S.;
Persoons, A.; Hockless, D. C. R. Organometallics 1996, 15, 5738; 1935.
(f) Di Bella, S.; Fragala`, I.; Ledoux, I.; Marks, T. J. J. Am. Chem. Soc.
1995, 117, 9481. (g) Kanis, D. R.; Lacroix, P. G.; Ratner, M. A.; Marks,
T. J. J. Am. Chem. Soc. 1994, 116, 10089. (h) Bourgault, M.; Mountassir,
C.; Le Bozec, H.; Ledoux, I.; Puccetti, G.; Zyss, J. J. Chem. Soc., Chem.
Comm. 1993, 1623.
S0002-7863(97)01349-8 CCC: $14.00 © 1997 American Chemical Society

Schiff Base NLO Chromophores
J. Am. Chem. Soc., Vol. 119, No. 40, 1997 9551
Chart 1
Scheme 1
the corresponding diamine (Scheme 1; see Experimental Section
for details).10 The yellow imines were readily purified by
recrystallization from ethanol. Ligands 1-6 were characterized
by 1H NMR spectroscopy and satisfactory microanalyses. The
crystal structures of unsubstituted H2(salen)11 (1) and H2-
(salophen)12 (3) indicate an enolimine tautomeric (nonplanar)
structure, also evident in solution.13 Thus, an analogous,
predominantly enoliminic structure (I in Scheme 1) can be
assumed for the related donor-acceptor substituted imines 2
and 4-6.14 In support of this formulation, a broad resonance
1
in the low field δ ) 12-13 ppm region is evident in the H
NMR spectra of all the present imine ligands. It disappears
upon exchange with D2O and can be associated with the -OH
protons doubtless involved in intramolecular hydrogen bonding
with iminic nitrogen atoms.
The reaction between ligands 1-6 and Ni(II) solutions leads
to the formation of stable noncentrosymmetric Ni(II) complexes
(1a-6a; Chart 1, Scheme 1). Yields are 60-70% for 1a-4a
and 40-50% for 5a and 6a. They were characterized by EI,
FAB mass spectrometry, 1H NMR spectroscopy, and satisfactory
microanalyses. FAB mass spectra indicate the existence of
monomeric species in the gas phase. Furthermore, available
X-ray structures of the unsubstituted Ni(salen)15 (1a) and Ni-
(salophen)16 (3a) complexes indicate monomeric, planar species
in the solid state. Thus, analogous planar, monomeric molecular
structures can be safely assumed for the remaining donor-
acceptor-substituted Ni(II) complexes 2a and 4a-6a. Therefore,
the formation of planar complexes 1a-6a is expected to effect
greater delocalization of out-of-plane π electron density over
the chelate and aromatic rings than in the free ligands.
Thermal Analysis. Ligands 1-6 exhibit relatively low
melting points (120-160 °C). In contrast, complexes 1a-6a
are more thermally stable materials which decompose, without
melting, at temperatures as high as 380 °C, as judged by
combined DSC/TGA thermal analysis (Table 1). The data
indicate generally good thermal stability, comparable to or
higher than current generation high-efficiency, thermally stable
NLO organic chromophores.17 Within the salophen series, the
having sizable second-order NLO responses, tunable by the
metal center, and a comparative experimental/theoretical inves-
tigation of the molecular second-order NLO properties,
â(-2ω;ω,ω), of a series of donor-acceptor bis(salicylaldimi-
nato)nickel(II) Schiff base complexes and related ligands (Chart
1).6 This study focuses on the synthesis, characterization,
thermal properties, linear optical spectroscopic, and second-order
NLO response, sampled by the electric field-induced second-
harmonic-generation (EFISH) technique,7 in combination with
a quantum chemical analysis, within the proven INDO/S-SOS
(ZINDO) formalism,5b,c,e-g,8,9 to describe/understand the elec-
tronic structure and structure NLO property relationships of this
new class of inorganic NLO chromophores.
Results
Synthesis, Characterization, and Molecular Structure.
Schiff base ligands 1-6 were prepared in high yields (80-
90%) via condensation of the appropriate salicylaldehyde with
(10) See, for example: (a) Hoss, H.; Elias, H. Inorg. Chem. 1993, 23,
317. (b) Chen, D.; Martell, A. E. Inorg. Chem. 1987, 26, 1026.
(11) Bresciani Pahor, N.; Calligaris, M.; Nardin, G.; Randaccio, L. Acta
Crystallogr. 1978, B34, 1360.
(12) Bresciani Pahor, N.; Calligaris, M.; Delise, P.; Dodic, G.; Nardin,
G.; Randaccio, L. J. Chem. Soc., Dalton Trans. 1976, 2478.
(13) ComprehensiVe Coordination Chemistry; Wilkinson, G., Ed.; Per-
gamon Press: Oxford, 1987; Vol. 2, p 730.
(14) Although a mixture of three stereoisomers could in principle be
formed in the condensation reaction to obtain ligands 4-6, the stereoisomer
having both trans-imine linkages, i.e., with the enolimine tautomeric
structure I in Scheme 1, is expected to be predominant. Moreover, since
complexation involves only the dianion of stereoisomer I, no attempts to
separate the mixture of the three stereoisomers were made.
(15) Montgomery, H.; Morosin, B. Acta Crystallogr. 1961, 14, 551.
(16) Gaetani Manfredotti, A.; Guastini, C. Acta Crystallogr. 1983, C39,
863.
(6) (a) For a preliminary communication of some aspects of this work,
see: Di Bella, S.; Fragala`, I.; Ledoux, I.; Diaz-Garcia, M. A.; Lacroix, P.
G.; Marks, T. J. Chem. Mater. 1994, 6, 881. (b) Salen coordination polymers
have previously investigated for NLO applications: Chiang, W.; Thompson,
M. E.; Van Engen, D. In ref 3c, p 210.
(7) For a general discussion of the EFISH tecnique, see: (a) Oudar, J.
L. J. Chem. Phys. 1977, 67, 446. (b) Levine, B, F.; Bethea, C. G. J. Chem.
Phys. 1976, 65, 1989.
(8) Kanis, D. R.; Ratner, M. A.; Marks, T. J. Chem. ReV. 1994, 94, 195.
(9) (a) There are several definitional issues associated with comparing
calculated microscopic responses to those obtained by experiment.9b In the
present contribution, both derived theoretical and experimental âµ values
employ the Ward definition,9c within a perturbation series expansion. (b)
Willetts, A.; Rice, J. E.; Burland, D. M.; Shelton, D. P. J. Chem. Phys.
1992, 97, 7590. (c) Ward, J. F. ReV. Mod. Phys. 1965, 37, 1.

9552 J. Am. Chem. Soc., Vol. 119, No. 40, 1997
Di Bella et al.
Table 1. Experimentala,b and ZINDO-Derived Spectroscopic and Nonlinear Optical Properties (hω ) 0.92 eV) for Donor-Acceptor Schiff
Base Ni(II) Complexes and Related Ligands
λ
max, nm
µg, D
exptl (∆µge)
µâ, ×10-48 esu
âµ, × 10-30 esu
c
compound
exptl (ꢀ × 10-3
)
calc (f)
calc (∆µge)
exptl
calc
exptl
calc
Td
325
1
1a
2
318 (8.4)
414 (6.9)
308
398 (8.6)
362 (13.4)
480 (9.0)
358
450 (16.3)
418
570 (9.3)
420
2.8
8.9 (-3.4)
3.9
8.6
6.0
7.0 (-5.1)
5.8
5.5
5.9
e
6.9
e
0
0
-9.3
1.2
360 (0.26)
352 (0.31)
387 (0.44)
381 (0.50)
485 (0.13)
491(0.18)
8.4 (-4.1)
8.4
-85
-85
-83
-10.1
-12.9
-17.3
-22.0
-38.7
-41.5g
4.7
2a
3
3a
4
-108
-9.6
0
219 (mp)
360
0
8.5 (-5.2)
7.5
-147
29
-144
-165
-298
-286
-20.5
5.0
4a
5
-126
-23
380
32
5.5
5ad
6
7.7
-331
54
-546
-43f
7.8
220
6ad
570 (13.1)
6.9g
-79f
260
a Chloroform solution. b Estimated uncertainties are (5% in µ and (10% in âµ. c Decomposition temperature (°C) as assessed by the onset of
the thermolysis endotherm in the DSC. d Acetone solution. e Complex insufficiently soluble for accurate measurement. f Estimated using calculated
dipole moment. g The calculated µ and âµ values for compound 6a differ from those previously reported6a because of the different conformation
presently adopted for the 4-OMe substituents (vide infra).
Table 2. Optical Absorption Maxima of the Lowest Energy
Charge-Transfer Optical Transition for Ni(salen) and Ni(salophen)
Complexes in Solvents Having Different Polarities
λ
max, nm
ethyl
complex
CCl4 benzene chloroform acetate acetone methanol
Ni(salen)
Ni(salophen) 494
422
488
414
480
414
480
410
478
400
466
to “d-d” transitions,19 is present in the 500-600 nm region
(Figure 1).
The 4-methoxy substitution of the salicylidene rings (H2(4-
OMe-salen), 2) results in a new, relatively intense band in the
longer wavelength region (388 nm) and in a more intense and
blue-shifted absorption band at 308 nm, compared to the 318
nm band of 1. This alteration is reflected in the optical spectrum
of 2a, which is slightly blue-shifted and more intense than that
of 1a (Figure 1). Substitution of the ethylenediamine bridge
of 1 with o-phenylenediamine (3) effects a larger π charge
delocalization over the salicylidene and phenylene aromatic rings
than in 1, resulting in a more intense, broader and red-shifted
πfπ* band in the optical spectrum (Figures 1 and 2). This
structural modification is also reflected in the optical spectrum
of Ni(II) complex 3a which exhibits two new intense features
in the longer wavelength region (a band at 381 nm and structure
in the 420-550 nm region), the latter of which is more intense
and red-shifted compared to that of Ni(salen) (Figures 1 and
2). Moreover, in analogy to 1a, the band centered at 480 nm
(ꢀ ) 9000) in 3a exhibits an analogous negative solvato-
chromism (Table 2), but a stronger dipole moment change (∆µ
) -5.1, Table 1), thus indicating greater CT character of the
associated optical transition.
Figure 1. Optical absorption spectra of H2(salen) and complexes 1a
and 2a in chloroform solution.
present thermal stabilities increase with 4-methoxysalicylidene
ring substitution and decrease with 3-nitrophenylene ring
substitution.
Optical Spectroscopy. The solution-phase optical absorption
spectrum (>280 nm) of ligand 1 consists (Figure 1) of a
relatively intense band centered at 318 nm (ꢀ ) 8400), involving
πfπ* transitions,18 and of a low intensity feature (ꢀ ≈ 100) in
the 380-420 nm region, responsible for the yellow color, and
presumably involving nfπ* excitation.18 Complexation with
Ni(II) results in two overlapping principal features assigned19
to two different types of transitions. There is a broader band
in the region between 300 and 360 nm still involving principally
intraligand πfπ* transitions and a new, relatively intense
structure in the 380-480 nm region, absent in the absorption
spectrum of the free ligand, involving both the ligand and the
metal center. Moreover, the band at 414 nm (ꢀ ) 6900) exhibits
a significant solvatochromic shift, characteristic of a large dipole
moment change (∆µ) between the ground and the excited state,
and indicative of charge transfer (CT) character. In particular,
a negative solvatochromism, i.e., a hypsochromic (blue) shift
with increasing solvent polarity, is observed (Table 2), indicating
a reduction in the dipole moment upon electronic excitation (∆µ
) -3.4; Table 1). Finally, a weak feature (ꢀ ≈ 100) assigned
Either 4-methoxy substitution in the salicylidene fragment,
or 3-NO2 substitution in the phenylene fragment of the salophen
ligand series (4-6), results in a new spectroscopic structure in
the longer wavelength region (a relatively intense shoulder at
440 nm for 4 and an intense band at ∼420 nm for 5 and 6).
These substituent modifications are in turn reflected in the
optical spectra of 5a and 6a which exhibit new spectral features
in the 500-600 nm region, compared to that of the unsubstituted
3a complex (Figure 2). In contrast, the longer wavelength band
in the optical spectrum of 4a is slightly blue-shifted, but
substantially more intense, compared to that of 3a (Figure 2).
(17) See, for example: Moylan, C. R.; Twieg, R. J.; Lee, V. Y.; Swanson,
S. A.; Betterton, K. M.; Miller, R. D. J. Am. Chem. Soc. 1993, 115, 12599.
(18) Crawford, S. M. Spectrochim. Acta 1963, 19, 255.
Ground-State Dipole Moment. Ground-state dipole mo-
ments, µg, of the present ligands and Ni(II) complexes measured
(19) Bosnich, B. J. Am. Chem. Soc. 1968, 90, 627.

Schiff Base NLO Chromophores
J. Am. Chem. Soc., Vol. 119, No. 40, 1997 9553
Table 3. INDO Eigenvalues and Mulliken Population Analysis of
the Outermost MO’s for Ni(salen)
population, %
a
b
MO
ꢀ, eV
Ni 2O 2N 2Cim
12Csal
characterb
dxy
Csal
Csal
CdN
CdN
62
59
58
57
56c
55d
54
53
52
51
50
1.87
1.21
1.19
51
0
0
8
4
4
12
0
0
5
0
2
24
96
94
36
39
55
74
49
24
82
49
-0.47
-0.51
-7.49
-7.94
-8.45
-8.55
-8.68
-9.57
2
1
13
5
35
60
7
4
2
22
18
5
13
1
20
22
10
2
11
3
38
36
0
1
0
0
1
dxz, O2p
O2p, Csal
dyz, N2p, Csal
2
dz
9
dxz, Csal
dyz, Csal
42
4
1
4
b
a Cim ) imine carbon atoms. Csal ) salicylidene carbon atoms.
c LUMO. d HOMO.
Figure 2. Optical absorption spectra of H2(salophen) and complexes
3a-6a in chloroform (3, 3a, 4a) or acetone (5a, 6a) solution.
ligands 2 and 4-6, positive âµ values, ∼ /5 to 1/7 the magnitude
1
of the related Ni(II) complexes, are observed. These observa-
tions illustrate the efficacy of metal complexation in “switching
on” and tuning the NLO response, upon formation of noncen-
trosymmetric structures.
in chloroform solution are collected in Table 1. Within the salen
series, the dipole moments of ligands 1 and 2 are consistently
lower than the corresponding values of Ni(II) complexes 1a
and 2a. This is consistent with a pseudo-centrosymmetric
structure of the free salen ligands (trans conformation with
respect to the ethylenediamine bridge),11 which becomes
constrained and noncentrosymmetric in the subsequent Ni(II)
complexes, thus resulting in an increased dipole moment.
However, for the present salophen series in which a noncen-
trosymmetric enoliminic structure is expected (vide infra),
complexation actually results in a smaller dipole moment
change. In both complexes of the salen and salophen series,
salicylidene ring 4-methoxy substitution results in a smaller
dipole moment with respect to the unsubstituted species.
Theoretical Results. Electronic Structure and Optical
Spectroscopy. The electronic structure of the H2(salen) ligand
1 can be described in terms of two weakly interacting N-
salicylideneamine subunits. The frontier molecular orbitals
(MO’s) of the latter molecule consists of symmetry combinations
of filled out-of-plane O2pz, π2 benzene, and π CdN imine
orbitals and of unfilled π* CdN MO’s.
The metal-ligand bonding in the present Ni(II) complexes
can be described analogously to that of d8 square-planar
complexes studied previously.24 Thus, the metal-ligand bond-
ing involves metal 3d, 4s, 4p valence orbitals and more external
filled MO’s of the ligand dianion. The latter consist of in-plane
N2p and O2p lone pairs and of out-plane π2 benzene, π CdN
and O2pz orbitals of appropriate symmetry combinations (a2 and
b2 in C2V). All these orbitals are strongly admixed and perturbed
by the interaction with the metal ion; however, the net metal-
ligand stabilization arises from the σ donation of the in-plane
N2p and O2p lone pairs to the empty metal 3dxy,4s orbitals.
INDO eigenvalues and atomic population analyses of the
frontier MO’s for prototypical Ni(salen) (1a) are reported in
Table 3. They consist of out-of-plane π orbitals mainly
delocalized over the salicylidene rings which have some CdN
and O2pz contributions, admixed to varying extents with metal
3d orbitals of appropriate symmetry. In particular, the HOMO
and SHOMO possess a large O2pz contribution, while the LUMO
and SLUMO possess a predominant CdN character. The
calculated ground-state dipole moment (8.4 D) is in very good
agreement with the experimental value (8.9 D; Table 1) and is
directed between the imine groups (Scheme 2). INDO/S
calculations account well for the experimental linear optical
spectroscopic features, and a satisfactory agreement between
calculated and observed transition energies is found (Table 4).
In particular, the aforementioned band at 414 nm may be
characterized as principally πfπ* (HOMOfLUMO) in char-
acter, essentially involving the metal dxz + O2pz and the CdN
orbitals, and is mainly responsible for the second-order NLO
response (vide infra). Moreover, the calculated negative ∆µge
(-4.1 D) value associated with this transition is in good
agreement with that determined from the negative solvato-
chromism (-3.4 D; Table 1).
Second-Order NLO Response. The molecular hyperpolar-
izabilities, â(-2ω;ω,ω), of the present compounds were mea-
sured by the EFISH technique.7,20 This technique allows the
determination of the µ‚â dot product when an electric field is
applied to a solution of an NLO-active species. The âµ value,
i.e., the vector component of the âijk tensor along the dipole
moment direction,21 can be thus extracted if the ground-state
dipole moment is known. Experimental âµ values of the present
ligands and complexes are compared in Table 1. It can be seen
that complexation dramatically enhances and, in some cases,
“switches on” the second-order NLO response. The âµ values
of 1a-6a are rather large for coordination complexes,4,5 ranging
from ∼-10 × 10-30 cm5 esu-1 (hω ) 0.92 eV), similar to that
of the classical donor-acceptor organic chromophore p-nitroa-
niline,22 to ∼-79 × 10-30 cm5 esu-1, rivaling âµ values reported
for efficient, second-order organic chromophores such as 4,4′-
donor-acceptor-substituted stilbenes.22 The observed NLO
response along the 1a-6a series parallels the aforementioned
red-shift and intensity increase of the lowest CT excitation, as
usually observed in typical donor-acceptor organic chro-
mophores.23
EFISH measurements on unsubstituted ligands 1 and 3
indicate a vanishingly small NLO response, while for substituted
(20) (a) Ledoux, I.; Zyss, J. Chem. Phys. 1982, 73, 203. (b) Barzoukas,
M.; Josse, D.; Fremaux, P.; Zyss, J.; Nicoud, J. -F.; Morley, J. O. J. Opt.
Soc. Am. B 1987, 4, 977.
(21)
3
µiâi
âµ(-2ω;ω,ω) )
∑
|µ|
i)1
1
On passing to Ni(salophen) (3a), substitution of the ethyl-
enediamine bridge of 1a with the phenylenediamine ring leads
where âi ) âiii + /3∑j*i(âijj + âjij + âjjj).
(22) Cheng, L. -T.; Tam, W.; Stevenson, S. H.; Meredith, G. R.; Rikken,
G.; Marder, S. R. J. Phys. Chem. 1991, 95, 10631.
(23) Di Bella, S.; Marks, T. J.; Ratner, M. A. J. Am. Chem. Soc. 1994,
116, 4440.
(24) See, for example: Di Bella, S.; Casarin, M.; Fragala`, I.; Granozzi,
G.; Marks, T. J. Inorg. Chem. 1988, 27, 3993.

9554 J. Am. Chem. Soc., Vol. 119, No. 40, 1997
Di Bella et al.
Table 4. Comparison of INDO/S-Deriveda and Experimental Optical Transition Energiesb of Ni(salen)
expt
exc state
S1
λcalc
f calc c
λmax
fexpt c
compositiond of CI expansion
principal character
dz2dxy
52f62
55f57
52f62
55f56
55f57
53f62
422
0.008
540
0.00
-0.68ø
+0.35ø
0.38ø
-0.45ø
0.56ø
+ 0.39ø
+ 0.35ø
54f56
S2
368
364
0.043
dxz, O2pCdN
446
0.03
0.07
48f62
51f62
55f62
54f57
55f57
50f62
53f62
51f56
53f56
49f62
55f56
S3
0.029
-0.39ø
+0.26ø
0.64ø
-0.41ø
-0.53ø
+0.59ø
+0.60ø
-0.27ø
0.38ø
+ 0.37ø
- 0.32ø
dxz, O2pdxy
}
{
55f56
52f62
S4
S5
S8
360
342
293
0.257
0.005
0.508
414
- 0.62ø
+ 0.27ø
dxz, O2pCdN
dyzdxy
392 (sh)
51f57
54f56
+ 039ø
+ 0.54ø
ππ*
332
0.15
2
50f56
53f56
52f56
S11
277
0.180
-0.32ø
+ 0.43ø
dz , dyzCdN
}
{
0.55ø
a Singlet excited states (S). b Transition energies (λmax) in nm. c Oscillator strength. See Table 3.
d
Scheme 2
an electron-donor group such as the methoxy in the salicylidene
rings tends to balance the charge separation with respect to the
unsubstituted complex, thus resulting in a reduced dipole
moment (see Scheme 2). Moreover, note that while in the C2V-
symmetric complexes 1a-4a, the total dipole moment (µg) is
oriented along the y axis (i.e., µg ) µy) and, in turn along the
µe CT axis, in the asymmetrically 3-NO2 substituted complexes
5a and 6a, the µg and µy vectors are not parallel, since the µx
vector component also makes an appreciable contribution (Table
5, Scheme 2).
Second-Order Molecular Nonlinearity. The calculated
second-order hyperpolarizabilities5,8,9 of the present molecules
are collected in Table 1 and compared with the experimental
data. The agreement between experimental and theoretical
values is very good in terms of both absolute values as well as
in the trends observed within the series.
For C2V complexes 1a-4a, the dominant contribution to âµ
derives from the yyy component of the âijk hyperpolarizability
tensor (where y is the direction along the principal dipole
moment axis of the molecule), which is parallel to the CT axis
(Scheme 2). In contrast, for unsymmetrical 3-NO2 substituted
complexes 5a and 6a, µg and µe are not parallel and while âyyy
is still the largest tensor, the âxxy ) âxyx tensors also contribute
to a larger π charge delocalization over the entire molecule,
and the frontier MO’s acquire a significant phenylene ring
character. This is reflected in a reduced energy gap between
the frontier MO’s as well as in a calculated and observed red
shift of the lowest CT transition (1a vs 3a; Table 1). In addition,
both the observed and calculated intensity and dipole moment
change associated with this transition are larger compared to
those of Ni(salen) (Table 1), because of the increased π charge
delocalization over the phenylene ring.
The 4-methoxy donor substitution in the salicylidene rings
and/or the 3-NO2 acceptor substitution in the phenylene ring
results in some donor character of the frontier filled MO’s and
a dominant acceptor character of the lowest unoccupied MO’s.
This, in turn, leads to a larger energy gap between filled and
unoccupied MO’s in the case of 4a and in a reduced gap for
complexes 5a and 6a, which is also reflected in stronger
calculated blue- and red-shifted lowest energy CT transitions
for 4a and 5a + 6a, respectively, in agreement with the
experimental data (Table 1).
significantly to the nonlinearity, and the total hyperpolarizability
25
(âvec
)
is substantially larger than âµ (Table 5).
In accord with experiment, an almost vanishing nonlinearity
(âµ ) 0.4 × 10-30 cm5 esu-1; hω ) 0.92 eV; on the basis of
the crystal structure11) is calculated for the H2(salen) ligand,
even assuming a cis (relative to the ethylene bridge) conforma-
tion. Analogously, in the case of unsubstituted H2(salophen),
the small calculated âµ value (0.9 × 10-30 cm5 esu-1; hω )
0.92 eV; assuming the distorted crystal structure12) agrees well
with the observed vanishing âµ. In addition, even assuming an
idealized planar structure, a small âµ value (3.1 × 10-30 cm5
esu-1; hω ) 0.92 eV) is still predicted. This observation further
underscores the effect of the metal center in “switching on” the
NLO response, because of a greater π charge delocalization over
the chelate ring and the existence of new excited states created
upon metal coordination.
Discussion
Metal complexation of ligands 1-6 with Ni(II) leads to
formation of thermally stable, constrained acentric planar
The 4-methoxy substitution in the salicylidene rings of 3a
leads to a lower calculated total dipole moment in 4a, in
agreement with the experiment (Table 1). Since the dipole
moment is directed between the imine groups, substitution of
(25) âvec ) (âx + ây + âz2)1/2. Note that âvec is always a positive
quantity regardless of the sign of the individual vectorial âi components.
2
2

Schiff Base NLO Chromophores
J. Am. Chem. Soc., Vol. 119, No. 40, 1997 9555
Table 5. Computed Second-Order NLO Responsea and Hyperpolarizability-Determining Parameters (eq 1) for Donor-Acceptor Schiff-Base
Ni(II) Complexes
c
compd
µx, D
µy, D
µg, D
stateb
hωeg, eV
rge, D
∆µge, D
âyyy,t
âyyy
âµ
âvec
â0,vec
1a
2a
3a
4a
5a
6a
7a
8a
9a
0.0
0.0
0.0
0.0
4.8
4.8
0.0
3.2
0.0
8.4
8.4
8.5
7.5
6.1
5.0
19.4
1.9
8.4
8.4
8.5
7.5
7.7
6.9
19.4
3.7
S4
S3
S2
S2
S3
S2
S4
S3
S3
3.44
3.52
3.21
3.26
2.55
2.52
3.01
3.12
2.09
4.36
4.70
5.97
6.38
3.58
4.10
5.99
6.16
5.88
-4.1
-3.7
-5.2
-4.9
-10.7
-11.6
0.2
-5.6
-6.0
-18.2
-17.6
-29.5
-43.0
0.7
-6.9
-7.2
-16.1
-17.3
-52.6
-58.8
3.8
-10.1
-12.9
-17.3
-22.0
-38.7
-41.5
14.2
10.1
12.9
17.3
22.0
59.0
68.7
14.2
55.6
447.6
7.6
9.8
11.5
15.1
27.3
32.5
7.6
-6.6
14.9
-33.0
373.0
-43.8
424.8
-14.5
447.6
31.7
88.0
0.22
0.22
a In 10-30cm5 esu-1 (hω ) 0.92 eV). b S ) singlet excited state involved in the two-state contribution. c Total static (zero frequency)
hyperpolarizability.
structures 1a-6a, which possess new linear optical features
responsible for the pronounced second-order optical nonlinearity.
The effect of the metal center is thus manifold: it templates
the formation of acentric structures, imparts high thermal
stability to the chelate ring, and both “switches on” and enhances
the NLO response. Moreover, the good agreement between
NLO experimental and theoretical data for the present complexes
suggests that theory can be used with confidence in rationalizing
nonlinearity and in designing new molecular architectures having
optimized NLO response.
A detailed analysis of the computational results indicates that
the second-order NLO response of the present complexes is
dominated in all instances by the intense, low-energy HOMO
f LUMO CT excitation mentioned above, in which the metal
acts as a electron donor. In such cases, the quadratic hyper-
polarizability can be simply related to the two-state contribution
(eq 1),7a where hω is the incident (laser) radiation frequency,
phenylene ring, would increase the ground-state dipole moment
but would decrease the CT character of the lowest optical
transition and, hence, the nonlinearity (Table 5). On the other
hand, 4-NO2 (instead of 3-NO2) phenylene ring substitution in
Ni(salophen) (8a) results (8a vs 5a) in a lower ground-state
dipole moment and a smaller ∆µge associated with the involved
lowest CT transition, in addition to an increased oscillator
strength and a blue shift of the same transition. The balancing
effects of these contributions leads to a slightly smaller âvec
than for 3-NO2, but to a substantially smaller âµ (Table 5).
These observations have interesting implications for the
design of donor-acceptor metal complex architectures having
optimized NLO response. For example, the combination of a
stronger (than the methoxy) 4-donor substituent in the sali-
cylidene rings, such as the diethylamino group, and 3,6-NO2
phenylene ring substitution, leads to a donor-acceptor structure
(9a) which is predicted (Table 5) to possess an almost vanishing
dipole moment (µg ) µy ) 0.22 D, having an opposite direction
than in 4a-8a), but a very large CT character (∆µge ) 14.9 D)
and a Very large NLO response (âµ = âyyy ) 448 × 10-30 cm5
esu-1; hω ) 0.92 eV; â0 ) 88.0 × 10-30 cm5 esu-1).26
hωegr2ge ∆µge
((hωeg)2 - (hω)2)((hωeg)2 - (2hω)2)
(1)
3e2
2
âyyy(-2ω;ω,ω) )
and hωeg, rge, and ∆µge, are, respectively, the energy, the
transition dipole moment, and the dipole moment variation,
between the ground and first CT excited state. Therefore, the
increasing âµ values observed on passing from 1a to 6a can be
directly related to the increasing CT character, hence, greater
∆µge values and to the bathochromic shift of the â-determining
CT transition (Table 5).
The negative experimental and calculated âµ values are a
consequence of the experimental and calculated negative
solvatochromism, hence negative ∆µge values, associated with
the â-determining CT transition (eq 1). This is due to a different
orientation of the dipole moment on passing from the ground
to the excited state. In particular, while in the ground state the
dipole moment is directed between the imine groups, in the
excited state it has an opposite orientation, i.e., between the
oxygen donor atoms (Scheme 2). The larger NLO response of
substituted donor-acceptor complexes 4a-6a with respect to
unsubstituted Ni(salophen) (3) results from an accurate design
of ring substitution in order to maximize ∆µge, i.e., the CT
character of the relevant optical transition. For example, an
inverted ring-substitution pattern in 3, that is 4-NO2 acceptor
substitution in the salicylidene rings (7a) instead of in the
Conclusions
This paper presents the synthesis and characterization of a
novel class of thermally stable, geometrically constrained
acentric NLO inorganic chromophores the second-order response
of which is “switched on” and tuned by the coordination of a
metal center. Even if the NLO response of these new materials
is modest compared to latest generation organic NLO chromo-
phores,2a-c the present synthetic strategy represents a novel route
to metal-organic NLO chromophores which are intriguing
candidates for further studies and applications.27
Experimental Section
Starting Materials. Salicylaldehyde, 2-hydroxy-4-methoxybenzal-
dehyde, ethylenediamine, and nickel(II) acetate tetrahydrate (reagent
grade, Aldrich) were used without further purification. 1,2-Phenylene-
(26) In this case, positive ∆µge and âµ values are computed because the
ground- and excited-state dipole moments are both directed between the
oxygen donor atoms.
(27) (a) Preliminary studies indicate that they can be easily incorporated
in thermally stable main-chain polymers for poled polymer NLO
applications.27b (b) Vitalini, D.; Mineo, P.; Di Bella, S.; Fragala`, I.;
Maravigna, P.; Scamporrino, E. Macromolecules 1996, 29, 4478.

9556 J. Am. Chem. Soc., Vol. 119, No. 40, 1997
Di Bella et al.
diamine and 3-nitro-1,2-phenylenediamine (Aldrich) were purified by
sublimation in vacuo.
Ni(4-OMe-sal-3-NO2-ophen) (6a). EI MS (70 eV) m/z: 477, 479
(M+).1H NMR (250 MHz, DMSO-d6) δ 3.75 (s, 6H, OMe), 6.2-6.4
(m, 5H, Ph), 7.45 (d, J ) 8.6 Hz, 1H, Ph), 7.64 (d, J ) 8.6 Hz, 2H,
Ph), 7.80 (d, J ) 7.3 Hz, 1H, Ph), 8.32 (s, 2H, CHdN). Anal. Calcd
for C22H17N3O6Ni: C, 55.26; H, 3.58; N, 8.79. Found: C, 54.9; H,
3.48; N, 8.48.
Synthesis of Ligands and Complexes. All Schiff bases were
prepared using standard procedures10 involving reaction of the ap-
propriate salicylaldehyde with the corresponding diamine (2:1 molar
ratio) in ethanol. The yellow imines were purified by recrystallization
from ethanol (yields 80-90%). The complexes were prepared10,28 by
reaction of an aqueous (for 1a-4a) or methanolic (for 5a and 6a)
solution of nickel(II) acetate with a boiling alcoholic solution of the
corresponding Schiff base ligand (1:1 molar ratio). The precipitated
complexes were collected by filtration, washed with an H2O/ethanol
mixture, and then recrystallized from absolute ethanol and/or chloroform
(yields, 60-70%). In the case of complexes 5a and 6a, the precipitates
were washed with methanol (yields 40-50%). While complexes 1a-
4a are soluble in most common organic solvents, 4a and 5a are soluble
in strongly polar solvents and moderately soluble in acetone.
H2(salen) (1). Melting point: 125-126 °C. Anal. Calcd for
C16H16N2O2: C, 71.62; H, 6.01; N, 10.44. Found: C, 71.40; H, 5.98;
N, 10.02.
Physical Measurements. Elemental analyses were performed on
a Carlo Erba 1106 elemental analyzer. EI and FAB mass spectra were
recorded on a Kratos MS 50 double-focusing mass spectrometer
equipped with a standard FAB source. FAB MS spectra were obtained
by using 3-nitrobenzyl alcohol (3NBA) as the matrix. 1H NMR spectra
were recorded on a Bruker AC-250 spectrometer. Thermal measure-
ments were performed with a Mettler 3000 system equipped with a
TG 50 thermobalance, a TC 10 processor, and a DSC 30 calorimeter.
The weight of samples was 15-20 mg for thermogravimetric analysis
(TGA), and 4-8 mg for differential scanning calorimetry (DSC).
Thermal analyses were made under prepurified nitrogen using a 10
°C/min heating rate. Thermal stabilities were estimated by combined
DSC and TG analyses. The intercept of the leading edge of the
decomposition endotherm with the base line of each DSC scan was
assigned as the decomposition temperature (Td).17 At this Td, the
corresponding weight loss in each thermogram was always less than
2%.
Ni(salen), 1a. Anal. Calcd for C16H14N2O2Ni: C, 59.13; H, 4.34;
N, 8.62. Found: C, 58.99; H, 4.30; N, 8.29.
H2(4-OMe-salen) (2). Melting point: 162-163 °C. 1H NMR (250
MHz, CDCl3): δ 3.79 (s, 4H, CH2), 3.85 (s, 6H, OMe), 6.34-6.41
(m, 4H, Ph), 7.08 (d, J ) 8.00 Hz, 2H, Ph), 8.20 (s, 2H, CHdN),
13.74 (br, s, 2H, OH). Anal. Calcd for C18H20N2O4: C, 65.84; H,
6.14; N, 8.53. Found: C, 65.41; H, 5.99; N, 8.15.
Spectroscopy. UV-vis spectra were recorded with a Beckman DU
650 spectrophotometer, and λmax values are considered accurate to (1
nm. Optical absorption spectra of complexes 1a-4a in chloroform
solution, recorded in the 5 × 10-3 to 2 × 10-6 M range, followed the
Lambert-Beer law, thus indicating that the complexes were monomeric
in solution.
Ni(4-OMe-salen) (2a). Melting point: 217-219 °C. EI MS (70
eV) m/z: 384, 386 (M+). 1H NMR (250 MHz, CDCl3): δ 3.35 (s,
4H, CH2), 3.74 (s, 6H, OMe), 6.16 (dd, J ) 8.7, 2.5 Hz, 2H, Ph), 6.51
(d, J ) 2.4 Hz, 2H, Ph), 6.91 (d, J ) 8.7, 2H, Ph), 7.29 (s, 2H, CHdN).
Anal. Calcd for C18H18N2O4Ni: C, 56.15; H, 4.71; N, 7.27. Found:
C, 55.82; H, 4.50; N, 6.93.
H2(salophen) (3). Melting point: 164-165 °C. 1H NMR (250
MHz, DMSO-d6): δ 6.9-7.0 (m, 4H, Ph), 7.3-7.6 (m, 8H, Ph), 8.85
(s, 2H, CHdN), 13.02 (br, s, 2H, OH). Anal. Calcd for C20H16N2O2:
C, 75.93; H, 5.10; N, 8.85. Found: C, 75.80; H, 4.99; N, 8.71.
Ni(salophen) (3a). FAB MS (3NBA) m/z: 373 (MH+). 1H NMR
(250 MHz, DMSO-d6): δ 6.67 (t, J ) 7.3 Hz, 2H, Ph), 6.88 (d, J )
8.5 Hz, 2H, Ph), 7.33 (m, 4H, Ph), 7.60 (d, J ) 8.0 Hz, 2H, Ph), 8.15
(dd, J ) 6.2, 3.0 Hz, 2H, Ph), 8.90 (s, 2H, CHdN). Anal. Calcd for
C20H14N2O2Ni: C, 64.39; H, 3.78; N, 7.51. Found: C, 64.01; H, 3.70;
N, 7.32.
H2(4-OMe-salophen) (4). Melting point: 167-168 °C. 1H NMR
(250 MHz, CDCl3): δ 3.83 (s, 6H, OMe), 6.45-6.55 (m, 4H, Ph),
7.22-7.32 (m, 6H, Ph), 8.54 (s, 2H, CHdN), 13.62 (s, 2H, OH). Anal.
Calcd for C22H20N2O4: C, 70.20; H, 5.35; N, 7.44. Found: C, 69.89;
H, 5.31; N, 7.15.
Ni(4-OMe-salophen) (4a). FAB MS (3NBA) m/z: 433 (MH+). 1H
NMR (250 MHz, CDCl3): δ 3.80 (s, 6H, OMe), 6.32 (dd, J ) 8.89,
2.43 Hz, 2H, Ph), 7.15-7.22 (m, 4H, Ph), 7.65 (dd, J ) 7.65, 3.33,
Hz, 2H, Ph), 8.09 (s, 2H, CHdN). Anal. Calcd for C22H18N2O4: C,
61.01; H, 4.19; N, 6.47. Found: C, 60.69; H, 3.98; N, 6.18.
H2(sal-3-NO2-ophen) (5). Melting point: 156-158 °C. 1H NMR
(250 MHz, DMSO-d6): δ 6.72 (t, J ) 8.10 Hz, 2H, Ph), 7.0 (m, 3H,
Ph), 7.4-7.5 (m, 3H, Ph), 7.8-8.0 (m, 3H, Ph), 8.91 (s, 2H, CHdN),
11.75 (br, s, 2H, OH). Anal. Calcd for C20H15N3O4: C, 66.48; H,
4.18; N, 11.63. Found: C, 66.01; H, 3.98; N, 11.20.
The dipole moment variation between the ground and the excited
state, ∆µge, was determined from the solvatochromism of the relevant
absorption band by means of the Lippert-Mataga equation29 where
n2 - 1
2n2 + 1
D - 1 n2 - 1
D + 2
νcr ) νgCT + C1
+ C2
-
(2)
(
)
n2 + 2
νCT is the frequency (cm-1) of the optical transition in a particular
solvent, D and n are the dielectric constant and refractive index of the
solvent, respectively. The intercept, νgCT, is the frequency of the
transition in the gas phase, and C2 ) 2µg∆µge/hca3; where µg is the
ground-state dipole moment, h and c are Planck’s constant and the
speed of light, respectively, a is the Onsager radius, and ∆µge is the
dipole moment change. The constant C2 was determined from the least-
squares fit of eq 4 to the absorption maxima of the CT band in 13
different solvents. ∆µge was calculated from C2 with the measured
value of the ground-state dipole moment and a value of the Onsager
radius (6.0 Å and 6.5 Å for 1a and 3a, respectively) estimated from
the solute molar volume.23
EFISH Measurements.
Second-order molecular hyper-
polarizability analyses were performed using a Q-switched mode-locked
Nd:YAG laser operating at 1.34 µm (hω ) 0.92 eV) by the electric
field-induced second-harmonic-generation method.20 The laser delivers
pulse trains of total duration envelope around 90 ns, each pulse duration
being 160 ps. The molecules to be measured were dissolved in
chloroform (used as a reference) or acetone (for complexes 5a and 6a)
at various concentrations, x, between 5×10-3 and 5 × 10-4 M, and
the solutions were placed in the wedge-shaped measurement cell. A
high voltage pulse (around 5 kV), synchronized with the laser pulse,
breaks the centrosymmetry of the liquid by dipolar orientation of the
molecules. Translation of this cell perpendicular to the beam direction
modulates the second-harmonic signal into Maker fringes. The
amplitude and periodicity of the fringing pattern are related to the
macroscopic susceptibility, Γ(x), of the solution and to the coherence
length, lc(x). Calibrations were made with respect to the pure solvent.
Γ(x) is related to the microscopic hyperpolarizability of the solvent,
γs, and of the dissolved molecule, γm, by
Ni(sal-3-NO2-ophen) (5a). FAB MS (3NBA) m/z: 418 (MH+). 1H
NMR (250 MHz, DMSO-d6): δ 6.32 (t, J ) 7.5 Hz, 2H, Ph), 6.69 (t,
J ) 7.5 Hz, 1H, Ph), 6.82 (d, J ) 8.1 Hz, 2H, Ph), 7.38 (m, 2H, Ph),
7.59 (d, J ) 7.5 Hz, 1H, Ph), 7.72 (d, J ) 8.3 Hz, 2H, Ph), 7.92 (d, J
) 7.1 Hz, 1H, Ph), 8.31 (s, 2H, CHdN). Anal. Calcd for C20H13N3O4-
Ni: C, 57.46; H, 3.13; N, 10.05. Found: C, 56.9; H, 3.26; N, 9.82.
H2(4-OMe-sal-3-NO2-ophen) (6). Melting point: 176-178 °C. 1H
NMR (250 MHz, DMSO-d6): δ 3.81 (s, 6H, OMe), 6.50-6.75 (m,
5H, Ph), 7.36 (dd, J ) 7.45, 1.37 Hz, 1H, Ph), 7.72 (d, J ) 8.62 Hz,
2H, Ph), 7.91 (dd, J ) 8.71, 1.36 Hz, 1H, Ph), 8.78 (s, 2H, CHdN),
12.22 (br, s, 2H, OH). Anal. Calcd for C22H19N3O6Ni: C, 62.70; H,
4.54; N, 9.97. Found: C, 62.40; H, 4.32; N, 9.61.
(28) Olszewski, E. J.; Martin, D. F. J. Inorg. Nucl. Chem. 1964, 26,
1577.
(29) Mataga, N.; Kubota, T. Molecular Interactions and Electronic
Spectra; Marcel Dekker: New Work, 1970; p 371.

Schiff Base NLO Chromophores
J. Am. Chem. Soc., Vol. 119, No. 40, 1997 9557
interaction (CIS) approximation was employed to describe the excited
states. In all calculations, the lowest 160 energy transitions between
SCF and CIS electronic configurations were chosen to undergo CI
mixing and were included in the SOS. This SOS truncation was found
to be sufficient for complete convergence of the second-order response
in all cases considered. All calculations were performed using the
ZINDO program5b,8 implemented on an IBM ES/9000 system.
Metrical parameters used for the calculations were taken from related
crystal structure data.15,16 Standard bond distances and bond angles
were adopted for donor-acceptor substituents.32 A syn conformation
of methoxy groups with respect the Csal-OMe bonds was adopted for
substituted 4-OMe 2a, 4a, and 6a complexes.33 This conformation was
found always to be more stable than the anti one. Calculations were
performed assuming a C2V planar (pseudo-planar for 1a and 2a; local
C2V for asymmetrically 3-NO2 substituted complexes 5a and 6a)
geometry. In the case of free ligands 1-6, no attempts to perform a
complete conformational analysis were made, thus no calculated
spectroscopic and NLO data were reported, except those for unsubsti-
tuted 1 and 3 species, for which the crystallographic structure was
assumed.11,12
γs
γm
NfF
Γ(x) )
+ x
(3)
(
)
1 + x Ms
Mm
where F is the density of the solvent, N is Avogrado’s number, f is an
averaged local field factor, and Ms and Mm are the molecular masses
of the solvent and of the solute, respectively. γm is given by
µg‚âµ
γm ) γe +
(4)
5kT
where γe is the scalar part of the third-order hyperpolarizability, µg the
ground-state dipole moment, and âµ ) âyyy + âyxx + âyzz ≈ âyyy for
one-dimensional charge-transfer molecule. In this work, we have
neglected the small γe contribution. The sign of âµ was determined
by studying the variation of Γ(x) as a function of x. If âµ is negative,
Γ first decreases when increasing x and then cancels out when the
contribution from the solvent is exactly compensated by that of the
dissolved NLO molecule. The Γ(x) value then increases because it is
dominated by the contribution of the solute. Further details of the
experimental methodology and data analysis are reported elsewhere.20
The ground-state dipole moment was determined by the standard
method of Guggenheim.30
Theoretical Methods. The all-valence INDO/S (intermediate
neglect of differential overlap) formalism,31 in connection with the sum-
over excited particle-hole-states (SOS) formalism,5b,8 was employed.
Details of the computationally-efficient ZINDO-SOS-based method for
describing second-order molecular optical nonlinearities have been
reported elsewhere.5b,8 Standard parameters and basis functions were
used.31 In the present approach, the closed-shell restricted Hartree-
Fock (RHF) formalism was adopted. The monoexcited configuration
Acknowledgment. This research was supported by the
Consiglio Nazionale delle Ricerche (CNR Rome), by the
Ministero dell’Universita` e della Ricerca Scientifica e Tecno-
logica (MURST, Rome), by the NSF-MRL Program through
the Materials Research Center of Northwestern University (Grant
DMR9632427) and by the MURI program of the DOD through
the CAMP research collaboration (ONR N00014-95-1-1319. We
thank Dr. P. G. Lacroix (CNRS, Toulouse) for helpful discus-
sions and Professor F. Castelli (Universita` di Catania) for
thermal measurements.
(30) Guggenheim, E. A. Trans. Faraday Soc. 1949, 45, 203.
(31) (a) Zerner, M.; Loew, G.; Kirchner, R.; Mueller-Westerhoff, U. J.
Am. Chem. Soc. 1980, 102, 589. (b) Anderson, W. P.; Edwards, D.; Zerner,
M. C. Inorg. Chem. 1986, 25, 2728. (b) Bacon, A. D.; Zerner, M. C. Theor.
Chim. Acta (Berlin) 1979, 53, 21. (c) Ridley, J.; Zerner, M. C. Theor. Chim.
Acta (Berlin) 1973, 32, 111.
JA971349Y
(32) Kanis, D. R.; Marks, T. J.; Ratner, M. A. Int. J. Quantum Chem.
1992, 43, 61.
(33) In the previously reported calculations6a on 6a, an anti conformation
was adopted, thus resulting in a slightly different calculated NLO response.