A critical evaluation of EFISH and THG non-linear optical responses of asymmetrically substituted meso-tetraphenyl porphyrins and their metal complexes

Article abstract of DOI:10.1016/S0020-1693(02)01027-7

The experimental second and third order non-linear optical (NLO) responses γ_EFISH and γ_THG, measured working with a non-resonant incident wavelength of 1.907 μm, of some push-pull meso-tetraphenyl porphyrins and their metal complexes

Full text of DOI:10.1016/S0020-1693(02)01027-7

Inorganica Chimica Acta 340 (2002) 70Á80
/
www.elsevier.com/locate/ica
A critical evaluation of EFISH and THG non-linear optical responses
of asymmetrically substituted meso-tetraphenyl porphyrins and their
metal complexes
Maddalena Pizzotti a,*, Renato Ugo a, Elisabetta Annoni a, Silvio Quici b,
Isabelle Ledoux-Rak c, Giuseppe Zerbi d, Mirella Del Zoppo d, PierCarlo Fantucci e,
Ivana Invernizzi e
a Dipartimento di Chimica Inorganica, Metallorganica e Analitica dell’Universita` e Istituto di Scienze e Tecnologie Molecolari del CNR (ISTM),
Universita` di Milano, Via G. Venezian 21, 20133 Milano, Italy
b Istituto di Scienze e Tecnologie Molecolari del CNR (ISTM), Via Golgi 19, 20133 Milano, Italy
c Ecole Normale Superieure de Cachan, 61 Avenue du President Wilson, 94230 Cachan, France
d Dipartimento di Chimica Industriale, Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy
e Dipartimento di Biotecnologie e Bioscienze, Universita` di Milano Bicocca, Piazza della Scienza 2, 20126 Milano, Italy
Received 28 January 2002; accepted 15 April 2002
Abstract
The experimental second and third order non-linear optical (NLO) responses gEFISH and gTHG, measured working with a non-
resonant incident wavelength of 1.907 mm, of some pushÁ
in the para position of one phenyl ring with a nitro group and with hydrogen or methyl or methoxy groups in the para position of
the other three phenyl rings, are critically discussed, also with the support of semiempirical MNDOÁTDHF calculations. Care must
be taken in evaluating the quadratic hyperpolarizabilities bvec from gEFISH being the contribution of the cubic term g0(ꢀ2v; v, v,
/pull meso-tetraphenyl porphyrins and their metal complexes, substituted
/
/
0) to gEFISH not negligible at least when the second order NLO response is relatively low, as in porphyrins and their metal complexes
investigated in this work. In addition experimental evidence has been produced that the value of gEFISH of the porphyrins and their
metal complexes is increased by resonance enhancement when working with an incident wavelength of 1.34 mm, due to the presence
of very strong absorption bands in the region 0.400Á
/
0.650 mm and that the gEFISH response is not significantly affected by 3dn (nꢁ
/
7Á10) metal coordination. On the contrary gTHG decreases by metal coordination due to the high resonance enhancement of gTHG of
/
free porphyrins. In fact the third harmonic 3v (0.636 mm) in free porphyrins is quite close to some strong Q absorption bands above
0.615 mm, which are lacking in the absorption spectra of their 3dn metal complexes. A vibrational method for the evaluation of the
order of magnitude of the static cubic g0 third order NLO response was successfully applied for the first time to a pushÁpull
/
porphyrinic system.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Porphyrin complexes; PushÁpull meso-tetraphenyl porphyrins; Non-linear optics
/
1. Introduction
pounds [5]. When compared to organic chromophores,
coordination and organometallic chromophores offer
additional electronic features which could act positively
on the second order NLO response, such as strong
charge transfer transitions (ligand to metal and vice-
versa) at relatively low energy, which can be tuned by
changing oxidation state, coordination sphere and
stereochemistry thus producing additional flexibility to
the design of a chromophore. In particular, cyclic
ligands with a large p electronic system could potentially
Dipolar organic molecules with significant second
order non-linear optical (NLO) properties have been
extensively investigated in the last decades [1Á3]. Only
/
more recently symmetrical octupolar molecules have
been studied with increasing interest [4], together with
asymmetrical coordination and organometallic com-
* Corresponding author. Fax: ꢂ
E-mail address: pizzotti@mailserver.unimi.it (M. Pizzotti).
/39-02-583 554405
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 1 0 2 7 - 7

M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á
/
80
71
act as a two dimensional highly polarizable electronic
framework able to transmit charge transfer processes in
contribution to this particular point we investigated, due
to their relatively easy synthesis, the experimental
determination of second and third order NLO responses
pushÁpull systems. However, these large p-conjugated
/
cyclic chromophores such as metallophthalocyanines [6]
and metallotriazolohemiporphyrazines [7] show signifi-
cant third order NLO responses, so their second order
NLO responses, when low, cannot be correctly mea-
sured by the EFISH (Electric Field Induced Second
Harmonic Generation) technique, due to a significant
third order electronic contribution to the EFISH
response [6,7].
of a series of asymmetrical pushÁpull meso-tetraphenyl
/
porphyrins (Fig. 1) with structural features similar to
those of porphyrins studied by Suslick et al. [15] and of
their metal complexes using the EFISH and THG
(Third Harmonic Generation) techniques. In few cases
we investigated also an easy and quick method for the
determination of the order of magnitude of static second
and third order NLO response, based on vibrational
measurements, developed by some of us [16]. In parallel
a theoretical investigation was carried out on the dipole
moment values and at least on the order of magnitude of
the EFISH and THG responses of these porphyrins and
their Zn(II) complexes, based on the semiempirical
MNDO (Modified Neglect of Diatomic Differential
Overlap) method [17] coupled with the TDHF (Time
Dependent Harthree-Fock) approach [18], as a tool to
confirm the experimental trends and the presence of
resonance enhancements in both EFISH and THG
experimental measurements and how these responses
can be affected by the presence of metal ions and their
electronic structure.
Among these large p-delocalised chromophores, por-
phyrins and their metal complexes could offer some
advantages for the synthesis of asymmetrical architec-
tures [8]. As in phthalocyanines the high polarizability of
the p-delocalised macrocycle of porphyrins is the origin
of significant third order non-linearities, which have
been extensively investigated [9]. Recent theoretical
investigations have reached the conclusion that the
correct substitution of the pushÁ
porphyrin ring, characterised by strong intramolecular
pÁp* charge transfer transitions [10], and some specific
/pull system in the
/
electronic and structural properties of the push and pull
groups could produce high second order responses [11],
as it has been found experimentally by Therien and co-
workers [12,13] and Ng and co-workers [14] who have
investigated 5,15-diphenylporphyrins and their Zn(II),
Cu(II) [12,13] or Ni(II) [14] complexes, carrying an
The asymmetrically substituted pushÁpull meso-tetra-
/
phenyl porphyrins, investigated in this work, are re-
ported in Fig. 1.
Only with the more asymmetrical porphyrin H2D the
series of complexes with Co(II) (3d7), Ni(II) (3d8),
Cu(II) (3d9), Zn(II) (3d10) ions and in parallel the series
of complexes with Ni(II), Pd(II), Pt(II) (3d8, 4d8, 5d8)
and Zn(II), Cd(II) (3d10, 4d10) ions were synthesized in
order to investigate a possible effect due to the metal d
electron releasing system (p-Me2NC6H4CÅ
series of different electron withdrawing systems (p-
NO2C6H4CÅCÃ [12,13] or ÃCHO, ÃCHÄC(CN)2, Ã
CHÄ CHCHO [14]) in the op-
/
CÃ/) and a
/
/
/
/
/
/
/
C(COOEt)2, trans-CHÄ
/
posite 10, 20 meso positions. The quadratic hyperpolar-
izabilities b of these porphyrins and their metal
electron configuration and to the energy of the nd8 (nꢁ
/
complexes have been measured by HyperÁ
/
Rayleigh
3, 4, 5) electron shell, respectively on the EFISH and
THG responses.
Scattering [12], Stark effect [13] and by EFISH [14].
Suslick et al. [15] have reported in a preliminary short
communication, rather low second order NLO re-
sponses, measured by the EFISH technique, of pushÁ
/
2. Experimental
pull meso-tetraphenyl porphyrins asymmetrically sub-
stituted in the para position of the phenyl rings with
amino and nitro groups. However, nobody has pointed
out clearly that due to the presence of strong absorption
2.1. Spectroscopic determinations, dipole moments and
NLO responses
bands in the region 0.400Á0.650 mm [10] the quadratic
/
1H NMR spectra were recorded on a Bruker AC-300
hyperpolarizability b measured with any technique with
an incident wavelength of 1.06 or 1.34 mm could be
affected by strong resonance enhancements, while in the
EFISH measurements, even if carried out with a non-
resonant incident wavelength of 1.907 mm, the third
order electronic contribution to gEFISH response could
not, as in the case of phthalocyanines [6,7], neglected
due to the already well-known large third order re-
sponses of porphyrins and their metal complexes [9].
Therefore, the experimental determination of the quad-
ratic hyperpolarizability of porphyrins requires a more
detailed and critical investigation. In order to give a
Spectrometer in CDCl3 as solvent, UVÁVis spectra were
/
Fig. 1. PushÁpull porphyrins studied in this work.
/

72
M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á80
/
obtained in CH2Cl2 with a Lambda 6 PerkinÁ
/
Elmer
2.2.1. 5-(4-Nitrophenyl)-10,15,20-tri-(4-
methoxyphenyl)-porphyrin (H2D)
Spectrometer. FABꢂ-MS mass spectrometric measure-
ments were performed with an analytical VG 7070 EQ
instrument. THG and EFISH experiments have been
performed in CHCl3 solutions, as a function of con-
A solution of 4-nitrobenzaldehyde (0.302 g, 2 mmol),
4-methoxybenzaldehyde (0.817 g, 6 mmol), pyrrole
(0.536 g, 8 mmol) and trifluoroacetic acid (CF3COOH)
(7.4 g, 65 mmol) in 800 ml of CHCl3 stabilised with 1%
EtOH was stirred at room temperature (r.t.) for 20 h
under nitrogen atmosphere and in the dark. After
addition of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone
(DDQ) (1.82 g, 8 mmol) the reaction mixture was stirred
for further 2 h, then Et3N (7.26 g, 72 mmol) was added
and the solvent evaporated. The residue was adsorbed
on 50 g of Florisil and purified by column chromato-
graphy (silica gel, CH2Cl2 as eluant).
Porphyrin H2D (320 mg, 21% yield) was separated as
a purple powder, from meso-tetrakis-(4-methoxyphe-
nyl)-porphyrin (21 mg) and meso-bis-(4-nitrophenyl)-
bis-(4-methoxyphenyl)-porphyrin (130 mg) as a mixture
of isomers.
centration (10ꢀ3Á10ꢀ4 M), working at 1.34 and 1.907
/
mm incident wavelengths, respectively, emitted by a Q-
switched Nd:YAG laser with 60 and 20 ns pulse
duration, respectively. The 1.907 fundamental wave-
length was obtained by Raman shifting of the 1.064 mm
emission of the Q-switched Nd:YAG laser in a high
pressure hydrogen cell (60 bar). A liquid cell with thick
windows in the wedge configuration was used to obtain
the Maker fringe pattern (harmonic intensity variation
as a function of liquid cell translation) [19]. In the
EFISH experiments the incident beam was synchronised
with a DC field applied to the solution in order to break
its centrosymmetry. From the concentration dependence
of the harmonic signal with respect to that of the pure
solvent, the NLO responses were determined (assumed
to be real because the imaginary part was neglected).
THG experiments provided the purely electronic cubic
2.2.2. 5-(4-Nitrophenyl)-10,15,20-tri-(4-methylphenyl)-
porphyrin (H2C)
Was prepared as H2D from 4-nitrobenzaldehyde
(0.302 g, 2 mmol), 4-methylbenzaldehyde (0.720 g, 6
mmol) and pyrrole (0.536 g, 8 mmol), yielding 370 mg of
H2C as a purple powder (26% yield), together with
meso-tetrakis-(4-methylphenyl)-porphyrin (35 mg) and
meso-bis-(4-nitrophenyl)-bis-(4-methylphenyl)-por-
phyrin (100 mg) as a mixture of isomers. It was
crystallized from CHCl3/n-hexane.
hyperpolarizability gTHG (ꢀ
EFISH experiments gave gEFISH, the sum of a cubic
electronic contribution g0(ꢀ2v; v, v, 0) and of a
quadratic orientational contribution m0bvec(ꢀ2v; v,
/3v; v, v, v) whereas
/
/
v)/5 kT, where m0 is the static dipole moment and bvec
the vectorial projection along the dipole moment direc-
tion of the tensorial quadratic hyperpolarizability (see
later Eq. (3)) [20].
2.2.3. 5-(4-Nitrophenyl)-10,15,20-triphenylporphyrin
(H2B)
A solution of 4-nitrobenzaldehyde (0.302 g, 2 mmol),
benzaldehyde (0.636 g, 6 mmol), pyrrole (0.536 g, 8
Experimental gTHG were corrected from the absorp-
tion of 3v by the Q absorptions bands [10], when these
latter were close to 3v (in particular in the case of free
porphyrins). It follows that the absolute value of gTHG
may be affected by significant errors; therefore in the
discussion we consider only the order of magnitude of
mmol) and BF3×/Et2O (0.84 g, 3 mmol) in 800 ml of
CH2Cl2 distilled over K2CO3, was stirred at r.t. for 20 h
under nitrogen atmosphere and in the dark. After
addition of DDQ (1.82 g, 8 mmol) the reaction mixture
was stirred for further 2 h, then Et3N (5 ml) was added
and the solvent evaporated. The residue was adsorbed
on 50 g of Florisil and purified by column chromato-
graphy (silica gel, CHCl3 as eluant). Porphyrin H2B (300
mg, 25% yield) was separated as a purple powder,
crystallized from CHCl3/n-hexane, from meso-tetrakis-
tetraphenylporphyrin (H2A) (112 mg) and meso-bis-(4-
nitrophenyl)-bis-(phenyl)-porphyrin (130 mg) as a mix-
ture of isomers.
gTHG
.
Experimental dipole moments determined in CHCl3
solutions following the Guggenheim method [21], were
too low and not reproducible probably due to some
aggregation of the dipolar disk-shaped porphyrin mole-
cules at the concentrations required for capacitance and
refractive index measurements.
2.2. Synthesis of asymmetrical porphyrins and their metal
complexes
Metal complexes of these porphyrins were obtained
by their well established reactions with metal chlorides
working in DMF as solvent [22].
In a typical preparation 100 mg of H2B were dissolved
in 20 ml of DMF, up to final dissolution. In a second
time 53 mg of PdCl2 were added and dissolved by
warming the mixture up to a solution which is kept
under reflux for 3 h. The final reaction mixture was
Porphyrins H2B, H2C and H2D, were prepared as
reported below. H2A was purchased from Aldrich
Chemicals and used as received, elemental analyses
were carried out in the Analytical Laboratories of Milan
University.

M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á
/
80
73
diluted with 20 ml of cold water, the solid residue was
filtered and washed with cold water. The solid was
dissolved in a small amount of CH2Cl2, the suspension
was filtered and the solvent evaporated under vacuum.
The red residue was washed few times with n-pentane
and then evaporated under vacuum, giving 106 mg (92%
yield) of the palladium complex. While some complexes
were isolated in enough pure form as red purple solids,
other complexes (PdC, CoD, ZnD, CdD, PdD) were
crystallized from CH2Cl2/n-hexane in order to obtain
pure compounds. The platinum complex, which was
obtained in relatively low yields starting from PtCl2, was
purified by column chromatography on SiO2 using
CH2Cl2 as eluant. The following yields (%, in parenth-
esis) have been obtained: PdB (92), PdC (93), PdD (95),
PtD (30), CoD (70), NiD (92), CuD (90), ZnD (90), CdD
(60). In Table 1 are reported the analytical data and the
relevant MS and 1H NMR data of both free porphyrins
and the metal complexes synthesized.
2.4. Vibrational method for evaluation of NLO properties
NLO properties of polyconjugated p-delocalised
structures may be evaluated by means of vibrational
observables due to the very large electronÁphoton
/
coupling which characterizes these peculiar highly p-
conjugated molecules. Under this condition a reliable
estimate of both second and third order NLO responses
is given just by the contribution of nuclear relaxations,
according to the Eqs. (1) and (2), derived under the
hypothesis of both mechanical and electrical harmoni-
city [16]:
ꢀ ꢁꢂꢀ
ꢁꢀ
ꢁ
X
1
1
nk2
dmn
damp
bGnmp
ꢁ
4p2c2
ꢁꢀ
dQk
dQk
ꢁꢃ
k
ꢀ
ꢁ
ꢀ
ꢁꢀ
dmm
danp
dmp
danm
ꢂ
ꢂ
(1)
dQk
dQk
dQk
dQk
ꢀ ꢁꢀ
ꢁꢀ
ꢁ
ꢁ
ꢀ
ꢁ
X
1
1
nk2
dmn
dbmps
dmm
gGnmps
ꢁ
ꢂ
4p2c2
dQk
dQk
ꢀ
dQk
k
ꢀ
ꢁ
ꢀ
ꢁꢀ
ꢁꢀ
ꢁ
dbnps
dmp
dbnms
dms
dbnmp
Â
ꢂ
ꢂ
2.3. Computational methods
dQk
dQk
ꢁ
dQk
dQk
ꢁ
dQk
ꢀ
ꢁꢀ
ꢀ
ꢁꢀ
ꢀ
ꢁꢀ
ꢁ
danm
daps
danp
dams
dans
damp
The geometrical electronic structure of the porphyri-
nic macrocycles and their Zn complexes were investi-
gated and dipole moments were calculated by the
semiempirical MNDO method [17]. The NLO proper-
ties, both static and frequency dependent, were calcu-
lated by the direct time dependent HF approach
(TDHF), as proposed by Sekino and Bartlett [18b,23]
and by Karna and Dupuis [18a]. This approach does not
require explicit calculations of excited states [24]. The
MNDO calculations have been carried out using a
standard version of MOPAC 6 program [25], to
compute the best molecular geometry and wavefunction
for the ground state, modified by us in order to include
the coupled perturbed HF equations [26]. The modified
NLO section of HONDO program [27], adapted by us
to work with basic MNDO integrals, was used to
calculate the NLO parameters.
ꢂ
ꢂ
ꢂ
dQk
dQk
dQk
dQk
dQk
dQk
(2)
The quantities such as dmn/dQk, danm/dQk, dbnms/dQk
(where n, m, s and p indicate the Cartesian compo-
nents), were obtained from infrared intensities and
Raman or Hyper Raman cross sections, respectively
(where nk is the vibrational frequency of the kth normal
mode Qk) giving purely static molecular NLO proper-
ties.
Raman data have been obtained with a Nicolet
Raman 910 with exciting line at 1.064 mm (resolution 4
cmꢀ1). Infrared spectra were recorded on a Nicolet
FTIR Spectrometer System 800 (resolution 1 cmꢀ1).
3. Results and discussion
Care must be taken when comparing calculated
microscopic NLO response to those obtained experi-
mentally by EFISH or THG techniques because the
definition gEFISH and gTHG is different if the Taylor
series or the perturbation series convention are used [28].
In our theoretical approach use is made of Taylor
expansion of both the energy and the induced dipole
moments, in terms of the applied electric field. In
addition the frequency dependent operators are all
treated in an exponential form, instead of a cosine
expansion. Both aspects require specific numerical
corrections (0.5 for bvec and 0.133 for gEFISH, gTHG
and g0) in order to compare the computed values with
the experimental EFISH or THG measurements, de-
fined according to so called ‘phenomenological’ conven-
tion [28].
3.1. Synthesis of porphyrins and their metal complexes
The synthesis of the asymmetrically substituted por-
phyrins H2B, H2C, H2D was achieved by performing the
general synthetic procedure described by Lindsey and
co-workers [29a] (see Section 2). The distribution of the
reaction products occurs in a statistical way, as recently
described by other authors [29b], giving place to a
mixture of porphyrins (Fig. 2) which can be easily
separated by chromatography. The metal complexes
were synthesized in DMF by reaction of the porphyrin
with the metal salt, following the reported general
synthetic procedure [22]. All attempts to isolate the
complex of H2D with the Hg(II) ion failed starting either

Table 1
Analytical data (elemental analyses, mass spectra and 1H NMR spectra) of porphyrins H2B, H2C, H2D and their metal complexes
a
b,c
Compound Elemental analysis
C(%) H(%)
79.51 (80.08) 3.98 (4.44) 10.45 (10.62) 659
MS-FAB(ꢂ)
(m/z)
1H NMR (CDCl3)
d(ppm)
N(%)
H2B
PdB
H2C
PdC
H2D
ꢀ2.8 (s, 2H, NH); 7.70Á
Hz]; 8.73 [d, 2H, Hb(Z)]; 8.86 [s, 4H, Hb (M)]; 8.89 [d, 2H, Hb(Z); JHb,Hoꢁ4.8 Hz]
/7.82 [m, 9H, Ho,Hp(M)]; 8.21 [dd, 6H, Hm(M)]; 8.39 [d, 2H, Hm(Z)]; 8.63 [d, 2H, Ho(Z); JHo,Hmꢁ8.6
69.42 (69.13) 3.06 (3.57) 8.67(9.23)
763
7.70Á7.82 [m, 9H, Ho, Hp(M)]; 8.16 [d, 6H, Hm(M)]; 8.33 [d, 2H, Hm(Z)]; 8.59 [d, 2H, Ho(Z); JHo,Hmꢁ8.3 Hz]; 8.68 [d, 2H,
Hb(Z)]; 8.82 [s, 4H, Hb(M)]; 8.85 [d, 2H, Hb(Z)]; JHb,Hoꢁ5.0 Hz]
/
80.20 (80.41) 4.59 (5.04) 10.10 (9.99) 701
69.72 (69.99) 3.92 (4.14) 8.72 (8.69) 806
74.83 (75.27) 4.40 (4.71) 9.40 (9.34) 750
ꢀ2.75 (s, 2H, NH); 2.70 (s, 9H, CH3); 7.56 [d, 6H, Ho(M)]; 8.09 [d, 6H, Hm(M); JHo,Hmꢁ7.7 Hz]; 8.38 [d, 2H, Hm(Z)]; 8.62
[d, 2H, Ho(Z); JHo,Hmꢁ8.6 Hz]; 8.71 [d, 2H, Hb(Z)]; 8.87 [s, 4H, Hb(M)]; 8.91 [d, 2H, Hb (Z); JHb,Hoꢁ4.8 Hz]
2.70 (s, 9H, CH3); 7.54 [d, 6H, Ho(M)]; 8.03 [d, 2H, Hm(M); JHo,Hmꢁ7.8 Hz]; 8.34 [d, 2H, Hm(Z)]; 8.60 [d, 2H, Ho(Z);
JHo,Hmꢁ8.6 Hz]; 8.66 [d, 2H, Hb(Z)]; 8.83 [s, 4H, Hb(M)]; 8.87 [d, 2H, Hb(Z); JHb,Hoꢁ5.0 Hz]
ꢀ2.75 (s, 2H, NH); 4.10 (s, 9H, OCH3); 7.29 [d, 6H, Ho(M)]; 8.11 (d, 6H, Hm(M); JHo,Hmꢁ8.6 Hz]; 8.39 [d, 2H, Hm(Z)]; 8.63
[dd, 2H, Ho(Z); JHo,Hmꢁ8.5; JHo,Hbꢁ4.0 Hz]; 8.70Á9.00 (m, 8H, Hb).
/
CoD
CuD
ZnD
69.80 (69.94) 4.34 (4.13) 8.75 (8.71) 807
69.08 (69.57) 3.71 (4.11) 8.72 (8.63) 811
70.11 (69.39) 3.99 (4.10) 8.86 (8.64) 813
Paramagnetic
Paramagnetic
4.10 (s, 9H, OCH3); 7.28 [d, 6H, Ho(M)]; 8.11 [(d, 6H, Hm(M); JHo,Hmꢁ7.9 Hz]; 8.39 [d, 2H, Hm(Z)]; 8.63 [dd, 2H, Ho(Z);
JHo,Hmꢁ8.4; JHo,Hbꢁ3.4 Hz]; 8.80Á9.08 (m, 8H, Hb)
/
CdD
NiD
PdD
PtD
a
65.55 (65.62) 4.09 (3.87) 8.57 (8.14) 860
69.57 (69.96) 4.24 (4.13) 8.45 (8.71) 805
65.94 (66.02) 3.78 (3.90) 8.25 (8.20) 854
59.78 (59.81) 3.04 (3.53) 7.39 (7.47) 942
4.10 (s, 9H, OCH3); 7.25Á7.30 [m, 6H, Ho(M)]; 7.85Á7.95 [m, 6H, Hm(M)]; 8.12Á
Ho(Z)]; 8.48Á8.62 (m, 8H, Hb)
4.05 (s, 9H, OCH3); 7.20 [d, 6H, Ho(M)]; 7.90 [d, 6H, Hm(M); JHo,Hmꢁ8.0 Hz]; 8.17 [d, 2H, Hm(Z)]; 8.54 [dd, 2H, Ho(Z);
JHo,Hmꢁ8.3; JHo,Hbꢁ3.5 Hz]; 8.56Á8.85 (m, 8H, Hb)
4.05 (s, 9H, OCH3); 7.27 [d, 6H, Ho(M)]; 8.06 (d, 6H, Hm(M); JHo,Hmꢁ8.1 Hz]; 8.34 [d, 2H, Hm(Z)]; 8.62 [dd, 2H, Ho(Z);
JHo,Hmꢁ8.2; JHo,Hbꢁ3.4 Hz]; 8.65Á8.92 (m, 8H, Hb)
4.10 (s, 9H, OCH3); 7.25 [dd, 6H, Ho(M); JHo,Hbꢁ3.0 Hz]; 8.04 (d, 6H, Hm(M) JHo,Hmꢁ8.3 Hz]; 8.32 [d, 2H, Hm(Z)]; 8.58
[dd, 2H, Ho(Z); JHo,Hmꢁ8.5; JHo,Hbꢁ3.6 Hz]; 8.65Á8.90 (m, 8H, Hb)
/
/
/
8.20 [m, 2H, Hm(Z)]; 8.35Á
/
8.45 [m, 2H,
/
/
/
/
Theoretical values in parenthesis.
b
c
Hb are the pyrrolic protons in the b position.

M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á
/
80
75
Table 2
Absorption electronic spectra in CH2Cl2
Compound lmax nm
(log o Mꢀ1 cmꢀ1
B bands
lmax nm (log o Mꢀ1 cmꢀ1
Q bands
)
)
H2A
411(5.11)
483sh (3.53), 514 (4.24), 549
(3.86), 588 (3.64), 646 (3.59)
515 (4.32), 550 (3.99), 590
(3.81), 647 (3.66)
H2B
412(5.34)
PdB
H2C
416(5.47)
416(5.23)
485 (3.52), 524 (4.48), 555 (3.52)
517 (4.20), 553 (3.95), 591
(3.67), 647 (3.61)
PdC
ZnC
H2D
415(5.29)
421(5.25)
419(5.17)
524 (4.34), 555 (3.48)
551 (4.29), 560 (3.89)
520 (4.04), 556 (3.87), 594
(3.54), 650 (3.56)
531 (4.13)
CoD
CuD
ZnD
CdD
NiD
PdD
PtD
414(5.26)
417(5.26)
423(5.39)
436(5.44)
419(5.29)
418(5.30)
406(5.35)
504 (3.48), 542 (4.24), 580(3.49)
515 (3.56), 552 (4.23), 593 (3.79)
530 (3.64), 570 (4.23), 614 (4.16)
531 (4.26), 615 (2.91)
488 (3.53), 526 (4.42), 560 (3.55)
512 (4.39), 540 (3.67)
Evidence of a fluorescent behaviour [10] can be
reached even visually: the CdD complex is redÁviolet
/
in the solid state, whilst in chloroform solution it
appears to be violet but with a strong dominating green
fluorescent emission.
Ground state dipole moments, which could not be
safely measured in CHCl3 solution by the standard
Guggenheim method [21] (see Section 2) were calculated
by using the MNDO semiempirical method [17]. Calcu-
lated dipole moments are quite similar (around 6D) and
they do not change when free and complexed porphyrins
are compared (Table 3). However, when the donor
group is NH2 the calculated dipole moment increases up
to 9 D (Table 3), a value higher than the experimental
Fig. 2. Distribution of products of a mixed condensation of two
aldehydes with pyrrole (d, donor group; a, acceptor group).
from HgO, Hg(CH3COO)2 or other Hg(II) compounds,
while the complex of the same porphyrin with the Cd(II)
ion was obtained only in low yield, although it was
reported that both complexes should be prepared easily
following this general synthetic procedure [22]. Also the
synthesis of the Pt(II) complex of H2D was not
straightforward. For instance it cannot be obtained in
acceptable yields starting from Na2[PtCl4]. Even using
as a starting material PtCl2 yields are rather low and
chromatographic purification is required (see Section 2).
value of 591 D reported by Suslick et al. [15]. This
/
latter experimental data however could be low due to
some aggregation of the porphyrin in CHCl3 solution
(see Section 2). Even if our calculated dipole moments
are probably overestimated, they seem to be quite low,
differently from the dipole moments of some asymme-
trical phthalocyanines with peripheral substitution of
the macrocycle with donor and acceptor groups [30].
For instance they are not much higher than just the
vectorial sum of the dipole moments of nitrobenzene
and of benzene substituted with the various donor
groups (NH2, CH3O, CH3) (Table 3) [31]. Therefore,
3.2. Electronic spectra and dipole moments
The electronic absorptions spectra of free porphyrins
H2A, H2B, H2C and H2D and their metal complexes,
characterized as expected [10] by a very strong Soret
absorption band (o about 105 Mꢀ1 cmꢀ1) at about 400
nm and by some other weaker, although still strong, Q
bands (o about 104 Mꢀ1 cmꢀ1) in the region 500Á
nm are not too influenced by the different molecular
asymmetry of our pushÁpull systems (see Table 2). The
/650
calculated dipole moments of these pushÁpull architec-
/
tures confirm a lack of significant electronic exchange
between the substituents on the phenyl rings via the
porphyrin macrocycle [15], in agreement with the
observation that in H2B, H2C, H2D, ZnB, ZnC, ZnD,
the calculated negative charge located on the NO2 group
/
introduction of a metal produces, as reported [10], a
lower number of Q absorption bands together with the
complete disappearance of those above 615 nm (Table
2).
(ꢀ0.18) remains quite independent upon the nature of
/

76
M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á80
/
Table 3
Calculated dipole moments
the region of Q absorption bands, particularly with free
porphyrins H2B, H2C and H2D (Table 2). Evidence for
resonance enhancements of the EFISH response was
found for these free porphyrins by carrying out some
measurements using also an incident wavelength of 1.34
mm. In this latter case the second harmonic 2v occurs at
0.670 mm, close enough to the region of Q absorption
bands of free porphyrins. We observed a significant
increase of gEFISH by going from 1.907 to 1.34 mm:
a,d
b,c
Compound mcalc(D)
Sum of m(C6H5NO2)ꢂm(C6H5X) (D)
4.03 (XꢁH)
H2B
ZnB
H2C
ZnC
H2D
ZnD
6.7
6.2
6.7
6.1
6.3
6.4
4.58 (XꢁCH3)
5.28 (XꢁOCH3)
gEFISH of H2B increases from 69
10ꢃ
10ꢀ34 esu and that of H2D from 4.29
to 1492ꢃ
10ꢀ34 esu, as expected for strong or sig-
/
0.5ꢃ
/
10ꢀ34 to 459
0.4ꢃ
10ꢀ34
/
a
Calculated by using the MNDO semiempirical method (see
/
/
/
Section 2.3).
b
/
/
Obtained from Ref. [31].
The sum of m(C6H5NO2)ꢂm(C6H5NH2) is 5.54 D quite in
c
nificant resonance enhancement originated by the strong
Q absorption band at 0.647 and 0.650 mm, respectively
(Table 2).
agreement with the experimental value of 591 D reported by Suslick
et al. [15] for their related pushÁpull meso-tetraphenyl porphyrin with
a pushÁpull system due to NH2 and NO2 groups.
The calculated value of the related pushÁpull meso-tetraphenyl
/
/
In agreement with this origin of the resonance
enhancement, when the porphyrin is coordinated to a
3dn metal, gEFISH values are not much affected by
shifting the incident wavelength from 1.907 to 1.34 mm:
/
d
/
porphyrin synthesized by Suslick et al. [15] (see c) is 8.99 D, to be
compared to the reported value of 591 D.
gEFISH of CuD moves from 109
/
3ꢃ
/
10ꢀ34 to 59
0.5ꢃ
/
0.5ꢃ
the donor group Y (see Fig. 1) on the opposite phenyl
ring and also upon coordination of the Zn(II) ion.
10ꢀ34 esu and PdD from 119
/
4ꢃ
/
10ꢀ34 to 6.59
/
/
10ꢀ34 esu. In these compounds, by complexation, any Q
absorption bands above 0.600 mm disappear so that the
third harmonic cannot be too resonant (see Table 2).
We also found that the order of magnitude of gTHG
and gEFISH is similar (10ꢀ34 esu), as it was reported to
occur in some asymmetrically substituted metal-free
phthalocyanines [31]. Absolute gTHG values in free
porphyrins are always higher probably due, as already
pointed out, to some resonance enhancement of 3v
(Table 4).
3.3. EFISH and THG NLO responses
In order to avoid or minimize resonance enhance-
ments of the NLO responses, EFISH measurements
(Table 4) were carried out with a non-resonant incident
wavelength of 1.907 mm, whose second harmonic 2v lies
at 0.953 mm in a transparent region of the absorption
spectra of all the pushÁ
complexes investigated, although the presence in this
region of extremely weak dÁd spin forbidden bands
/pull porphyrins and their metal
The comparable order of magnitude of gTHG and
gEFISH, as in the case of phthalocyanines [6], suggests a
significant contribution to gEFISH of the electronic cubic
/
cannot be totally excluded for Cu(II) or Co(II) macro-
cycles [30]. However, resonance enhancements cannot
be totally avoided in our THG measurements because
the third harmonic 3v occurs at 0.636 mm quite close to
term g0(ꢀ2v; v, v, 0) of Eq. (3):
/
m0bvec(ꢀ2v; v; v)
5 kT
gEFISH
ꢁ
ꢂg0(ꢀ2v; v; v; 0) (3)
Table 4
gEFISH and gTHG measurements in CHCl3 solutions using an incident
wavelength of 1.907 mm
It follows that the usual approximation, used for
classical pushÁpull organic systems in the evaluation
of bvec from gEFISH, neglecting the cubic term g0(ꢀ2v;
/
a
a
gEFISH (10ꢀ34 esu)
gTHG (10ꢀ34 esu)
/
Compound
v, v, 0) may give rise even in these porphyrins, as in
other similar macrocycles [6,7], to bvec values over-
estimated, particularly when they are basically low [15].
The not too different gEFISH values of H2D and its
H2A
H2B
PdB
H2C
PdC
H2D
CoD
NiD
PdD
PtD
(0.4)
(0.6)
690.5
5.190.6
1492
1796
4.290.4
691.2
792
1194
1492
1093
891
35910
1292
45910
691
65925
31910
2092
1594
1692
2095
2297
55930
complexes with various 3dn metal ions (nꢁ
/
7Á10)
/
(Table 4) indicate also the absence of relevant effects
on the gEFISH response upon coordination of the
porphyrin to 3dn metal ion in contrast with the marked
enhancement of the magnitude of the gEFISH response
reported to occur upon coordination to some open shell
3dn metal ions of octasubstituted phthalocyanines [32].
However, in this latter case the enhancement of
gEFISH by metal coordination was attributed to v
resonance of the 1.064 mm incident wavelength with
CuD
ZnD
CdD
1394
a
Values in parenthesis are calculated by semiempirical MNDOÁ
/
TDHF methods (see Section 2.3).

M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á
/
80
77
additional one-photon very weak forbidden dÁ
tions of some 3dn metal ions, which could occur in the
region 1.000Á1.300 mm. To a similar origin was attrib-
uted the dependence to metal coordination of g(v; v,
v, v) response obtained by using Degenerate Four
/
d transi-
by one or two order of magnitude, than the experi-
mental one (Table 5), producing additional evidence for
relevant resonance contributions to experimental gTHG
values of free porphyrins and probably, although at
minor extent, of their metal complexes. In fact our
/
ꢀ
/
Wave Mixing (DFWM) for phthalocyanines and naph-
tocyanines coordinated to some open shell 3dn metal
ions [33]. Obviously under our experimental conditions
(incident wavelength at 1.907 mm) such a v resonance
effect cannot occur.
MNDOÁTDHF approach does not take into considera-
tion resonance effects.
Our calculations also confirm that the cubic electronic
/
contribution g0(ꢀ2v; v, v, 0) to gEFISH response
/
cannot be safely neglected at least in porphyrins with a
relatively low second order NLO response as those
investigated in this work. In fact in agreement with a
small dipolar contribution m0bvec/5 kT to gEFISH and
3.4. Theoretical calculations of NLO responses
Theoretical calculations of the best geometries using
the MNDO semiempirical approach [17], have con-
firmed that in porphyrins H2A, H2B, H2C and H2D and
their Zn(II) complexes, the macrocycle is always essen-
tially planar, while the aromatic rings in the meso
position display, as expected, an arrangement nearly
perpendicular to the macrocycle due to the repulsion of
hydrogen atoms of the phenyl rings in the ortho position
by the b hydrogen of the two adjacent pyrrolic rings
[11,15]. This repulsion produces a torsional barrier
with a more relevant contribution of g0(ꢀ
measured bvec, neglecting g0(ꢀ2v; v, v, 0), is often
negative while calculated gEFISH is positive (Table 5).
Still, in line with a significant contribution of g0(ꢀ2v;
v, v, 0), the calculated gEFISH for the rather symme-
trical molecule H2A is similar, also as sign, to that of the
more asymmetrical porphyrins H2B, H2C and H2D
(Table 5), although for H2A the dipolar contribution
m0bvec/5 kT to gEFISH should be close to zero being the
local symmetry D2h. Calculations confirm also the
limited role of metal coordination: coordination to
Zn(II) of H2B, H2C and H2D produces only a small
increase of both calculated b0 and bvec (Table 5).
/
2v; v, v, 0)
/
/
about the C(meso)Ã
/
C(aryl) bond [34], as confirmed by
the X-ray crystallographic structure of H2A which
shows a twisting between the phenyl rings and the plane
of the porphyrin with an average dihedral angle of about
608 [35].
Differently from the ZINDOÁSOS (sum over state)
/
3.5. The order of magnitude of third order NLO
properties from vibrational data
semiempirical approach proposed by Marks and co-
workers [36], which is a time dependent perturbation
method leading to a SOS [37], we used a semiempirical
In the recent past it has been shown that when a
direct time dependent approach MNDOÁTDHF, as
/
strong electronÁphonon coupling occurs [39,40] the
/
proposed by Sekino and Bartlett [18b,23], and by Karna
and Dupuis [18a], which allows the computation of the
NLO responses without involving excited states (see
Section 2.3). Recent theoretical studies on the NLO
response of organic systems, using this approach, have
been carried out based on ab initio methods (mainly
Harthree-Fock and Density Functional Theory [38]),
which are expected to give results more accurate than
those based on semiempirical methods such as ZINDO
or MNDO. However, ab initio methods can be carried
out with reasonable computer time on relatively small
(or periodic) molecular systems, particularly when using
highly correlated methods. Up to now the systems
investigated by ab initio methods are considerably
smaller than those considered in the present study. Since
our theoretical investigation was carried out only with
the rather qualitative aim of confirming the experimen-
tal trends of gEFISH and gTHG the approach based on the
MNDO semiempirical method can give all the necessary
informations. With this limitation in mind we found that
the calculated order of magnitude of gEFISH values is in
good agreement with the experimental one, whilst the
order of magnitude of calculated gTHG is always lower,
vibrational hyperpolarizabilities can be extremely large.
Therefore, at least the order of magnitude of both static
second order b0 and third order g0 responses could be
evaluated from the contribution of the nuclear relaxa-
tion, in the hypothesis of an harmonic behaviour of both
atomic vibrations and electron displacements [16] (see
Section 2.4).
Table 5
Calculated NLO parameters
a,b
Compound b0 10ꢀ30
(esu)
bvec 10ꢀ30 gEFISH 10ꢀ34
gTHG 10ꢀ34
(esu)
(esu)
(esu)
H2A
ZnA
H2B
ZnB
H2C
ZnC
H2D
ZnD
ꢀ0.5
ꢀ4.3
ꢀ2.5
9.7
ꢀ6.5
ꢀ7.5
ꢀ3.2
16.1
ꢀ3.8
7.3
0.4
0.2
0.6
0.4
0.2 (690.5)
1.9
0.7 (35910)
0.4
ꢀ2.9
4.2
1.9
0.1 (1492)
1.1
0.7 (45910)
0.6
2.6
9.1
0.8 (4.290.4) 0.6 (65925)
1.3 (891) 0.5 (2297)
5.7
a
With the exception of b0, all the values are referred to an incident
wavelength of 1.907 mm.
Values under parenthesis are experimental values (Table 4).
b

78
M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á80
/
This vibrational approach was applied with a certain
error. Under the hypothesis that Hyper Raman inten-
sities are negligible [42] their static cubic hyperpolariz-
ability g0 were evaluated, neglecting the first four terms
of Eq. (2) (see Section 2.4), directly from measured
Raman cross-sections. The g0 values so determined are
of the same order of magnitude of g0 values calculated
success in the determinations of the static NLO proper-
ties of p-delocalised chain of organic systems such as
polyenes, with or without polar end groups, or ribbon
like molecules (rylenes) or even relatively smaller p-
delocalised cyclic systems such as naphthalene, perylene
and terylene [41].
with our semiempirical MNDOÁTDHF theoretical
/
Large static b0 values require simultaneous strong
infrared and Raman intensities for the same normal
mode (see Eq. (1)). Comparison of infrared and Raman
spectra of H2D (Fig. 3), shows that the strongest
infrared bands have negligible Raman intensities and
viceversa, as expected for a fundamentally centrosym-
metric porphyrinic skeleton not too much perturbed by
approach (Table 6), although they are much lower,
also as order of magnitude, when compared to experi-
mental gTHG values measured with an incident wave-
length of 1.907 mm (Table 4). This is an expected trend
because these latter values are affected by strong
resonance enhancements as experimentally and theore-
tically proved.
The most relevant contributions to g0 originate from
very low frequency vibrations (below 400 cmꢀ1), which
a careful spectroscopic work [43] assigned to skeletal
modes of low frequency. Since in the evaluation of g0
Raman intensities must be weighted by a factor 1/n2 (see
Eq. (1)), the contribution of the skeletal modes of low
frequencies is very important even if their Raman
intensities are not extremely strong.
the pushÁ
rings.
This trend is typical of all the pushÁ
investigated in this work in agreement with their rather
low second order NLO response. It was impossible to
obtain a quantitative estimate of b0 because the
measured absolute intensities of the few coincident
infrared and Raman bands are within the experimental
/
pull asymmetrical substitution of the phenyl
/pull porphyrins
Fig. 3. Infrared (top) and Raman (bottom) spectra of H2D in solid state.

M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á
/
80
79
Table 6
g0 values from Raman spectra
the irrelevant effect of metal coordination on the Soret
band [10]. Also with other porphyrins with a larger p
conjugated pushÁpull system [12,13], it was reported
/
Compound g0 10ꢀ34 a(esu) g0 10ꢀ34 b(esu) gTHG 10ꢀ34 c(esu)
that coordination of the porphyrin to Zn(II), Cu(II),
Ni(II) does not produce significant effects on the Soret
band [12,13]. Only when the macrocycle is originated by
multichelation, which can introduce some rigidity of the
organic skeleton and a totally new p electronic system
on ligands such as flexible tetradentate asymmetrical
Schiff bases, the effect of coordination to open shell 3dn
metal ions in increasing the second order NLO response
of the ligand is relevant and well established. As
expected in these latter cases significant changes occur
in the electronic absorption spectra upon coordination
[44]. In conclusion our investigation has produced some
evidence for a more careful evaluation of the experi-
mental EFISH measurements of second order responses
of porphyrins.
Finally, we have shown for the first time, although in
a rather preliminary way, that a vibrational method [16]
can give an acceptable evaluation of the order of
magnitude of at least the static cubic hyperpolarizabil-
ities g0 of p-delocalised macrocycles and their metal
complexes. With all its limitation this method, which is
unaffected by resonance enhancements or by fluorescent
contributions, could be in the future considered as the
first choice for a facile and quick determination of the
order of magnitude of the static second and third order
NLO responses of p-delocalised macrocycles such as
porphyrins and their metal complexes, which is the
relevant issue for any technological applications.
H2D
ZnD
NiD
CuD
0.99
0.67
0.83
0.53
2.54
1.7
65925
2297
2092
2095
n.d.
n.d.
a
Experimental from Raman spectra.
b
Static value calculated with the MNDOÁ
/
TDHF semiempirical
method.
c
Determined by THG experiments with 1.907 mm incident wave-
length.
4. Conclusion
In this work we have confirmed, taking as reference
asymmetrically substituted meso-tetraphenyl porphyr-
ins, characterized by a limited p conjugation between the
porphyrin and the aromatic rings and therefore by a
limited second order NLO response [15] that, due only
to the high polarizability of the p-delocalised porphyrin
ring, the third order NLO response is substantial. It
follows that their bvec values, when low, obtained from
gEFISH measurements could be affected by a significant
overestimation because, as in asymmetrical phthalocya-
nines [31], the cubic contribution g0(ꢀ
gEFISH cannot be neglected. This warning was supported
also by MNDOÁTDHF calculations of both gEFISH and
bvec. In addition we have experimentally and theoreti-
cally confirmed that in porphyrins and in their metal
complexes, showing strong absorption bands in the
/
2v; v, v, 0) to
/
region 0.400Á0.650 mm, the second order NLO response
/
[12], when working with an incident wavelength of 1.340
mm, is affected by relevant resonance enhancements due
to Q absorption bands. It follows that second order
NLO responses of porphyrins and their metal complexes
must be considered with some care particularly if these
measurements have not been carried out working with a
non-resonant incident wavelength of 1.907 mm.
Acknowledgements
This work was supported by the Ministero dell’Uni-
versita` e della Ricerca Scientifica e Tecnologica (Pro-
gramma di Ricerca MURST di tipo interuniversitario
nell’area delle scienze chimiche, ex 40%, 1997, Research
title: Molecole per materiali funzionali nanostrutturati)
and by Consiglio Nazionale delle Ricerche (Progetto
Finalizzato Materiali Speciali per Tecnologie Avanzate
II, 1998, Research Title: Sintesi e sviluppo di composti
molecolari organometallici e di coordinazione con
proprieta` ottiche non-lineari (NLO) e con proprieta`
elettriche anisotropiche e isotropiche). We thank Dr.
Francesca Tessore, Dr. Elena Cariati and Dr. Silvia
Bruni for some EFISH and THG measurements.
We also reached evidence that the coordination to a
metal ion does not perturb much the gEFISH response, at
least in the case of our porphyrins, when working with
an incident wavelength of 1.907 mm. This result is in
contrast with what reported for other p-delocalised
macrocycles such as octasubstituted phthalocyanines
[32] or triazolohemiporphyrazines [7a], working with
an incident wavelength of 1.064 mm. But in these latter
cases the increase of gEFISH by metal coordination to the
p-delocalised macrocycle was attributed to a v reso-
nance enhancement, due to some weak dÁ
possibly occurring around 1.064 mm.
MNDOÁTDHF calculations have produced addi-
/d absorption
References
/
tional evidence for a relatively small effect on the second
order NLO response upon coordination of our porphyr-
ins to Zn(II). This irrelevant effect could be related to
[1] (a) D.S. Chemla, J. Zyss, Nonlinear Optical Properties of Organic
Molecules and Crystals, vol. IÁ
/
II, Academic Press, Boston, 1987;
(b) I. Ledoux, J. Zyss, in: I.C. Khoo, F. Simoni, C. Umeton

80
M. Pizzotti et al. / Inorganica Chimica Acta 340 (2002) 70Á80
/
(Eds.), Novel Optical Materials and Applications, John Wiley and
[19] (a) I. Ledoux, J. Zyss, Chem. Phys. 73 (1982) 202;
(b) D.J. Williams, Angew. Chem., Int. Ed. Engl. 23 (1984) 690
(and references therein).
Sons, New York, 1997, p. 1.
[2] M. Blanchard-Desce, S.R. Marder, M. Barzoukas, in: D.N.
Reinhoudt (Ed.), Comprehensive Supramolecular Chemistry,
vol. 10 (and references therein), Pergamon Press, Oxford, 1996,
p. 833.
[20] (a) S.K. Saha, G.K. Wong, Appl. Phys. Lett. 34 (1979) 423;
(b) B.F. Levine, C.G. Bethea, Appl. Phys. Lett. 24 (1974) 445.
[21] E.A. Guggenheim, Trans. Faraday Soc. 45 (1949) 203.
[22] A.D. Adler, F.R. Longo, F. Kampas, J. Kim, J. Inorg. Nucl.
Chem. 32 (1970) 2443.
[3] D.R. Kanis, M.A. Ratner, T.J. Marks, Chem. Rev. 94 (1994) 195
(and references therein).
[23] H. Sekino, R. Bartlett, J. Chem. Phys. 94 (1991) 3665.
[24] D. Kanis, M.A. Ratner, T.J. Marks, J. Am. Chem. Soc. 114
(1992) 10338.
[4] (a) J. Zyss, I. Ledoux, Chem. Rev. 94 (1994) 77;
(b) J. Zyss, T. Chauvan, C. Dhenaut, I. Ledoux, Chem. Phys. 206
(1993) 409;
[25] J.J.P. Stewart, T.J. Seiler, MOPAC Manual, a program product,
6th ed., Research Laboratory, United States Air Force Academy,
1990.
(c) I. Ledoux, C. Dhenaut, I.D.W. Samuel, J. Zyss, Nonlinear
Opt. 14 (1995) 23.
[5] (a) N.J. Long, Angew. Chem., Int. Ed. Engl. 34 (1995) 21 (and
references therein);
[26] J. Gorrat, I.M. Mills, J. Chem. Phys. 49 (1968) 1719.
[27] M. Dupuis, A. Farazdel, S.P. Karna, S.A. Maluendes, HONDO-
8, in: E. Clementi (Ed.), Modern Techniques in Computational
Chemistry MOTECC, ESCOM, Leiden, 1990.
[28] A. Willetts, J.E. Rice, D.M. Burland, D.P. Shelton, J. Chem.
Phys. 97 (1992) 7590.
(b) H.S. Nalwa, Appl. Organomet. Chem. 5 (1991) 349;
(c) D.M. Roundhill, J.P. Fackler, Jr., Optoelectronic Properties of
Inorganic Compounds, Plenum Press, New York, 1999, p. 107.
[6] (a) G. De la Torre, P. Va`squez, F. Agullo`-Lo`pez, T. Torres, J.
Mat. Chem. 8 (1998) 1671 (and references therein);
(b) G. De la Torre, T. Torres, F. Agullo`-Lo`pez, Adv. Mat. 9
(1997) 265.
[29] (a) J.S. Lindsey, I.C. Schreiman, H.C. Hsu, P.C. Kearnay, A.M.
Marguerettez, J. Org. Chem. 52 (1987) 827;
(b) C.H. Lee, J. Lindsey, Tetrahedron 39 (2000) 11427.
[30] M.A. Diaz-Garcia, F. Agullo`-Lo`pez, A. Sastre, B. Del Rey, T.
Torres, C. Dhenaut, I. Ledoux, J. Zyss, Nonlinear Optics 15
(1996) 251.
[7] (a) M.A. D`ıaz-Garc`ıa, I. Ledoux, F. Ferna`ndez-La`zaro, A.
Sastre, T. Torres, F. Agullo`-Lo`pez, J. Zyss, J. Phys. Chem. 98
(1994) 4495;
(b) F. Ferna`ndez-La`zaro, T. Torres, B. Hanschel, M. Hanack,
Chem. Rev. 98 (1998) 563 (and references therein).
[8] J.S. Lindsey, in: F. Montanari, L. Casella (Eds.), Metallopor-
phyrins Catalysed Oxidations (and references therein), Kluwer
Academic Publishers, Dordrecht, 1994, p. 49.
[31] A.L. McLellan, Table of Experimental Dipole Moments, W.H.
Freeman and Company, San Francisco, 1963, p. 1.
[32] M.A. D`ıaz-Garc`ıa, I. Ledoux, J.A. Duro, T. Torres, F. Agullo`-
Lo`pez, J. Zyss, J. Phys. Chem. 98 (1994) 8761.
[33] J.S. Shirk, J.R. Lindle, F.J. Bartoli, Z.H. Kafafi, A.W. Snow, M.
Boyle, Int. J. Opt. Phys. 1 (1992) 699.
[9] (a) D.V.G.L.N. Rao, F.J. Arauda, J.F. Roach, D.E. Remy, Appl.
Phys. Lett. 58 (1991) 1241;
[34] S.S. Eaton, D.M. Fishnild, G.R. Eaton, J. Am. Chem. Soc. 99
(1977) 6594.
(b) H.L. Anderson, S.J. Martui, D.D.C. Bradley, Angew. Chem.,
Int. Ed. Engl. 33 (1994) 655;
[35] S.J. Sliver, A. Tulinsky, J. Am. Chem. Soc. 89 (1967) 3331.
[36] D.R. Kanis, P. Lacroix, M.A. Ratner, T.J. Marks, J. Am. Chem.
Soc. 116 (1994) 10089 (and references therein).
[37] B.J. Orr, J.F. Ward, Mol. Phys. 20 (1971) 513.
[38] (a) B. Champagne, D.H. Mosley, J.G. Fripiat, J.M. Andre, Adv.
Quantum Chem. 35 (1999) 95;
(c) S. Guha, K. Kang, P. Porter, J.F. Roach, D.E. Remy, F.J.
Arauda, D.V.G.L.N. Rao, Opt. Lett. 17 (1992) 264;
(d) K. Kandasauny, S.J. Shetty, P.N. Puntambekar, T.S. Srivas-
tava, T. Kundu, S.P. Singh, Chem. Commun. (1997) 1159.
[10] M. Gouterman, in: D. Dolphin (Ed.), The Porphyrins, vol. III,
Academic Press, New York, 1978, p. 1.
(b) B. Kirtman, B. Champagne, D.M. Bishop, J. Am. Chem. Soc.
122 (2000) 8007;
[11] I.D.L. Albert, T.J. Marks, M.A. Ratner, Chem. Mater. 10 (1998)
753.
(c) B. Champagne, P. Fisher, A.D. Buckingham, Chem. Phys.
Lett. 331 (2000) 83.
[12] S.M. LeCours, H.W. Guan, S.G. Di Magno, C.H. Wang, M.J.
Therien, J. Am. Chem. Soc. 118 (1996) 1497.
[39] C. Castiglioni, M. Del Zoppo, P. Zuliani, G. Zerbi, Synth. Met.
71 (1995) 171.
[13] L. Karki, F.W. Vance, J.T. Hupp, S.M. Le Cours, M.J. Therien,
J. Am. Chem. Soc. 120 (1998) 2606.
[40] P. Zuliani, M. Del Zoppo, C. Castiglioni, G. Zerbi, C. Andraud,
T. Brotin, A. Collet, J. Phys. Chem. 99 (1995) 16242.
[41] M.C. Rumi, G. Zerbi, K. Mullen, M. Rehalm, J. Chem. Phys. 106
(1997) 24.
[14] M. Yeung, A.C.H. Ng, M.G.B. Drew, E. Vorpagel, E.M.
Breitung, R.J. McMahon, D.K.P. Ng, J. Org. Chem. 63 (1998)
7143.
[42] C. Castiglioni, M. Tommasini, M. Del Zoppo, J. Mol. Struct. 521
(2000) 137.
[15] K.S. Suslick, C.T. Chen, G.R. Meredith, L.T. Cheng, J. Am.
Chem. Soc. 114 (1992) 6928.
[43] R.H. Felton, N.-T. Yu, in: D. Dolphin (Ed.), The Porphyrins, vol.
III, Academic Press, New York, 1978, p. 347.
[16] (a) C. Castiglioni, M. Gussoni, M. Del Zoppo, G. Zerbi, Solid
State Comm. 82 (1992) 13;
[44] (a) S. Di Bella, I. Fragala`, I. Ledoux, M.A. D`ıaz-Garc`ıa, T.J.
Marks, J. Am. Chem. Soc. 119 (1997) 9550;
(b) M. Del Zoppo, C. Castiglioni, G. Zerbi, Nonlinear Opt. 9
(1995) 73.
(b) S. Di Bella, I. Fragala`, T.J. Marks, M.A. Ratner, J. Am.
Chem. Soc. 118 (1996) 12747;
[17] M.J.S. Dewar, W. Thiel, J. Am. Chem. Soc. 99 (1977) 4899.
[18] (a) S.P. Karna, M. Dupuis, J. Comput. Chem. 12 (1991) 487;
(b) H. Sekino, R.J. Bartlett, J. Chem. Phys. 85 (1986) 976.
(c) P.G. Lacroix, S. Di Bella, I. Ledoux, Chem. Mater. 8 (1996)
541.