Synthesis and characterization of new ferrocenyl heterobimetallic compounds with high NLO responses

Article abstract of DOI:10.1016/S0022-328X(98)00569-5

Several ferrocenyl based heterobimetallic compounds have been obtained and characterized, and their NLO responses have been measured. We find that the observed β values are among the highest of the organometallic based materials reported up to date.

Full text of DOI:10.1016/S0022-328X(98)00569-5

Journal of Organometallic Chemistry 562 (1998) 197–202
Synthesis and characterization of new ferrocenyl heterobimetallic
compounds with high NLO responses
Jose Mata a, Santiago Uriel a, Eduardo Peris a,*, Rosa Llusar b, Stephan Houbrechts c,
Andre´ Persoons c
a Departamento de Qu´ımica Inorga´nica y Orga´nica, Uni6ersitat Jaume I, E-12080 Castello´n, Spain
b Dpto. de Ciencies Experimentals, Uni6ersitat Jaume I, E-12080 Castello´n, Spain
c Centre for Research on Molecular Electronics and Photonics, Laboratory of Chemical and Biological Dynamics, Uni6ersity of Leu6en,
Celestijnenlaan 200D, B-3001 Leu6en, Belgium
Received 19 January 1998; received in revised form 11 March 1998
Abstract
Several ferrocenyl based heterobimetallic compounds have been obtained and characterized, and their NLO responses have been
measured. We find that the observed i values are among the highest of the organometallic based materials reported up to date.
© 1998 Elsevier Science S.A. All rights reserved.
Keywords: Ferrocenes; NLO; Heterobimetalic; Metal-carbonyl; Electrochemistry
1. Introduction
chromophore can allow us to tune both, electron-donor
and electron-acceptor fragments by modifying their
electronic properties.
In the last ten years, the incorporation of metals to
NLO systems have given a new dimension to the study
and design of new chromophores [1]. Although a large
number of organometallic complexes have been studied,
just a few examples can be found in which clear design
criteria are applied for the optimization of the nonlin-
ear responses [2]. It has been stated that the use of
organometallic compounds in NLO has a wide fan of
advantages [1] that make them appropiate in the field of
non linear optics. One of these advantages is that the
electronic properties of the metal fragment can be
tuned, so that the NLO response can be modulated.
The use of heterobimetallic organometallic complexes
in which the electron-accepting and donating properties
of two organometallic fragments are combined in order
to obtain high first hyperpolarizabilities is rare, and
only recently some examples have been reported [3].
The introduction of two metal centres in the NLO
In the search of new combinations of organometallic
fragments we have chosen a typical donor fragment
(ferrocene), and a series of electron accepting moieties
derived from M(CO)6 (M=Cr or Mo). The ferrocenyl
ligands were prepared according to conventional or-
ganic synthesis, and we have developed a synthetic
route to ligands with long conjugated chains. We have
studied the electrochemical properties of our com-
pounds and relate them to their NLO responses. These
bimetallic compounds exhibit static hyperpolarizabili-
ties up to 164×10−30 esu, the largest measured to date
for ferrocenyl derivatives.
2. Experimental section
2.1. General details
* Corresponding author. Fax: +34 964 345747.
All reactions were carried out under a nitrogen atmo-
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00569-5

198
J. Mata et al. / Journal of Organometallic Chemistry 562 (1998) 197–202
2.3. Syntheses of the ferrocenyl ligands 3, 5 and 9
sphere using standard Schlenck techniques. Solvents for
synthesis and electrochemical measurements were dried
and degassed by standard methods before use. Chro-
matographic work was performed on Silica gel 60 A or
Alumina columns.
Compound 1 was obtained according to literature
methods [4]. We report modified methods to the obtain-
tion of compounds 2 [5], 5, 8 and 9 [6] in which we have
simplified the work up and improved the yields.
Proton-NMR spectra were recorded on a Varian
Gemini 200 MHz, using CDCl3 as solvent unless other-
wise stated. IR spectra were recorded on a Perkin
Elmer System 2000 FT-IR using NaCl pellets. Cyclic
To an ice-cold solution of (1-(triphenylphosphi-
nomethyl)ferrocenyl]iodine (3 g, 5.1 mmol) in THF (50
ml) was added potassium tert-butoxide (850 mg, 7.6
mmol) and the resulting solution stirred for 30 min to
form the ylide. The corresponding aldehyde, t-cian-
namaldehyde (2.5 ml, 20.4 mmol for the synthesis of
3), 4-pyridinecarboxaldehyde (1.5 ml, 15.3 mmol for
the synthesis of 5) and 4-cyanobenzaldehyde (1.2 g,
15.3 mmol for the synthesis of 9), was added and the
resulting mixture stirred for ca. 24 h. After removing
the solvent under reduced pressure, the product was
extracted with CH2Cl2 and purified by column chro-
matography on silica gel with hexane/CH2Cl2 (8:2) for
compound 3 and hexane/CH2Cl2/acetone (12:8:1) for
compounds 5 and 9 as eluent. Recrystallization from
CH2Cl2/hexane mixtures afforded pure compounds.
Yields: 47% (for 3), 60% (for 5), 44% (for 9). Proton-
NMR spectrum for compound 3: l 7.41–7.18 (m, 5H,
C6H5), 6.88–6.76 (m, 2H, CHꢀCH); 6.60–6.36 (m, 2H,
CHꢀCH); 4.37 (s, 2H, C5H4); 4.25 (s, 2H, C5H4); 4.09
(s, 5H, C5H5); Proton-NMR spectrum for compound
˚
voltammetry experiments were performed with
a
ECHOCHEMIE PGSTAT 20 electrochemical analyzer.
All measurements were carried out at room tempera-
ture with a conventional three-electrode configuration
consisting of platinum working and auxiliary electrodes
and a Ag/AgCl reference electrode containing aqueous
3 M KCl. The solvent in all experiments was CH2Cl2,
which was obtained in HPLC grade from sds. The
supporting electrolite was 0.1 M tetrabutylammonium
hexafluorophosphate, obtained from Sigma or synthe-
sized by metathesis of tetrabutylammonium bromide
and HPF6, recrystallized from ethanol and dried under
vacuum. E1/2 values were determined as 1/2(Ep,a+Ep,c),
where Ep,a and Ep,c are the anodic and cathodic peak
potentials, respectively. All potentials reported are not
corrected for the junction potential.
3
5: l 8.45 (d, 2H, JH–H=5.04 Hz, C5H4N); 7.21 (d,
3
3
2H, JH–H=5.88 Hz, C5H4N); 7.06 (d, 1H, JH–H
=
16.12 Hz, CHꢀCH); 6.53 (d, 1H, 3JH–H=16.08,
CHꢀCH); 4.43 (s, 2H, C5H4); 4.29 (s, 2H, C5H4); 4.09
(s, 5H, C5H5); Proton-NMR spectrum for compound
3
9: l 7.53 (d, 2H, JH–H=8.43 Hz, C6H4); 7.42 (d, 2H,
3JH–H=8.36 Hz, C6H4); 7.20 CDCl3; 6.96 (d, 1H,
2.2. NLO measurements
3JH–H=16.12 Hz, CHꢀCH); 6.60 (d, 1H, JH–H=16.14
3
Hz, CHꢀCH); 4.43 (s, 2H, C5H4); 4.29 (s, 2H, C5H4);
4.09 (s, 5H, C5H5). Elemental Anal. Calc. for com-
pound 3, C20H18Fe, Mw=314.21: C, 76.5; H, 5.8.
Found: C, 76.6; H, 6.0. Elemental Anal. Calc. for
compound 5, C17H15FeN, Mw=289.26: C, 70.6; H,
5.2; N, 4.8. Found: C, 71.1; H, 5.0; N, 4.8. Elemental
Anal. Calc. for compound 9, C19H15FeN, Mw=
313.18: C,72.9; H, 4.8; N, 4.5. Found: C, 72.1; H, 5.0;
N, 4.4.
Details of the HRS set-up have been discussed previ-
ously [7,8]. All measurements are performed in chloro-
form and the known hyperpolarizability of
para-nitroaniline in this solvent (23×10−30 esu) is used
as a reference [9]. The samples are passed through a
0.45 mm filter for contaminated samples often produce
spurious signals, and are checked for multiphoton
fluorescence that can interfere with the HRS signal
[10–12]. IR-laser pulses generated with an injection
seeded, Q-switched Nd:YAG laser (Quanta-Ray GCR-
5, 1064 nm, 10 ns pulses, 10 Hz) are focused into a
cylindrical cell containing the solution (7 ml). The
fundamental intensity is altered by rotation of a half-
wave plate placed between crossed polarizers, and mea-
sured with a photodiode. An efficient condenser system
is used to collect the light scattered at the harmonic
frequency (532 nm) that is detected by a photomulti-
plier. Discrimination of the second-harmonic light from
the fundamental light is accomplished by a low-pass
filter and a 532 nm interference filter. Actual values for
the intensities are retrieved by using gated integrators.
In all experiments the incident light was vertically po-
larized along the z axis.
2.4. Syntheses of 2 and 4
Cr(CO)6 (500 mg, 2.3 mmol) and compound 1 (265
mg, 0.9 mmol) or compound 3 (283 mg, 0.9 mmol) were
refluxed in buthyl ether (40 ml) overnight. After cooling
the mixture to r.t., the solution was filtered and the
solvent removed under reduced pressure. Purification
by column chromatography on alumina with CH2Cl2/
hexane (4:1) afforded pure compounds 2 (Yield: 45%)
and 4 (Yield: 57%). Proton-NMR spectrum for com-
3
pound 2: l 6.82 (d, 1H, JH–H=16.25 Hz, CHꢀCH),
3
6.20 (d, 1H, JH–H=16.16 Hz, CHꢀCH); 5.47 (s, 5H,
C5H5); 4.44 (s, 2H C5H4); 4.34 (s, 2H C5H4); 4.18 (s,

J. Mata et al. / Journal of Organometallic Chemistry 562 (1998) 197–202
199
5H, C5H5); Carbon-NMR spectrum for compound 2: l
2.6. Syntheses of 7, 8 and 11
233.30 (3C, CO); 130.52 (1C, CHꢀCH); 121.24 (1C,
CHꢀCH); 107.66 (1C, C6H5); 93.42 (2C, C6H5); 90.56
(1C, C6H5); 89.87 (2C, C6H5); 81.27 (1C, C5H4); 69.78
(2C, C5H4); 69.56 (5C, C5H5); 67.30 (2C, C5H4)). Pro-
ton-NMR spectrum for compound 4: l 6.4–6.6 (m,
CHꢀCH); 6.45 (d, 1H, 3JH–H=16.16 Hz, CHꢀCH);
Compounds 7, 8 and 11 were prepared by the same
general method than 6 and 10 from Mo(CO)6 and
W(CO)6, respectively. Yield for 7, 10%, yield for 8, 63%
and yield for 11, 25%. Proton-NMR spectrum for
3
compound 7: l 8.41 (d, 2H, JH–H=6.23 Hz, C5H4N);
3
7.08 (d, 2H, 3JH–H=5.9 Hz, C5H4N); 7.13 (d, 1H,
6.05 (d, 1H, JH–H=16.16 Hz, CHꢀCH); 5.15–5.50 (m,
3
3JH–H=16.08 Hz, CHꢀCH); 6.46 (d, 1H, JH–H=16.1,
5H, C6H5); 4.37 (s, 2H C5H4); 4.27 (s, 2H C5H4); 4.18
(s, 5H, C5H5). Carbon-NMR spectrum for compound
4: l 233.24 (3C, CO); 134.83 (1C, CHꢀCH); 132.37 (1C,
CHꢀCH); 125.31 (1C, CHꢀCH); 125.18 (1C, CHꢀCH);
107.13 (1C, C6H5); 82.03 (1C, C5H4); 69.46 (2C, C5H4);
69.21 (5C, C5H5); 67.21 (2C, C5H4). Elemental Anal.
Calc. for compound 2, C21H16FeO3Cr, Mw=424.20: C,
59.5; H, 3.8. Found: C,59.3; H, 3.5. Elemental Anal.
Calc. for compound 4, C23H18FeO3Cr, Mw=450.24: C,
61.4; H, 4.0. Found: C,61.2; H, 4.4.
CHꢀCH); 4.45 (s, 2H, C5H4); 4.35 (s, 2H, C5H4), 4.09
(s, 5H, C5H5); Carbon-NMR spectrum for compound
7: l 214.57 (1C, trans CO); 204.50 (4C, cis CO); 154.78
(2C, Py); 146.30 (1C, Py); 135.65 (2C, Py); 120.86 (2C,
CHꢀCH); 80.96 (1C, C5H4); 70.52 (2C, C5H4); 69.54
(5C, C5H5); 67.85 (2C, C5H4); Proton-NMR spectrum
for compound 8: l 8.54 (d, 2H, 3JH–H=6.5 Hz,
3
C5H4N); 7.27 (d, 1H, JH–H=16.5 CHꢀCH); 7.23 (2H,
3
3JH–H=6.6 Hz, C5H4N); 6.49 (d, 1H, JH–H=16.4 Hz,
CHꢀCH); 4.47 (s, 2H, C5H4); 4.37 (s, 2H, C5H4); 4.10
2.5. Syntheses of 6 and 10
(s, 5H, C5H5); Proton-NMR spectrum for compound
3
11: l 7.56 (d, 2H, JH–H=8.06 Hz, C6H4); 7.47 (d, 2H,
Cr(CO)6 (176 mg, 0.8 mmol) and Me3NO (88.9 mg,
0.8 mmol) were dissolved in THF (20 ml) and the
resulting solution was stirred for 20 min. Compound 5
(231 mg, 0.8 mmol) or compound 9 (250 mg, 0.8 mmol)
was then added to the above solution for the prepara-
tion of 6 or 10, respectively. The reaction mixture was
stirred for another 20 min, the solution was filtered and
the solvent removed under reduced pressure. Purifica-
tion by column chromatography on alumina with
3
3JH–H=8.16 Hz, C5H4); 7.04 (d, 1H, JH–H=16.10 Hz,
3
CHꢀCH); 6.61 (d, 1H, JH–H=16.02, CHꢀCH); 4.45 (s,
2H, C5H4); 4.33 (s, 2H, C5H4); 4.11 (s, 5H, C5H5);
Carbon-NMR spectrum for compound 11: l 199.92
(1C, trans CO); 196.11 (4C, cis CO); 143.45 (1C, C6H4);
132.85 (2C, C6H4); 125.95 (2C, C6H4); 125.01 (1C,
C6H4); 123.18 (2C, CHꢀCH); 106.56 (1C, CN); 81.29
(1C, C5H4); 69.87 (2C, C5H4); 69.19 (5C, C5H5); 67.26
(2C, C5H4). Elemental Anal. Calc. for compound 7,
C22H15NFeO5Mo, Mw=525.15: C, 50.3; H, 2.9; N,
2.7. Found: C, 51.3; H, 3.4; N, 2.8. Elemental Anal.
Calc. for compound 8, C22H15NFeO5W, Mw=613.06:
C, 43.1; H, 2.5; N, 2.3. Found: C, 43.7; H, 2.5; N, 2.1.
CH2Cl2/hexane (1:1) afforded pure compounds
(Yield: 40%) and 10 (Yield: 20%). Proton-NMR spec-
6
3
trum for compound 6: l 8.34 (d, 2H, JH–H=5.7 Hz,
3
C5H4N); 7.08 (d, 2H, JH–H=5.9 Hz, C5H4N); 7.13 (d,
3
3
1H, JH–H=16.08 Hz, CHꢀCH); 6.46 (d, 1H, JH–H
=
Elemental
Anal.
Calc.
for
compound
11,
16.1, CHꢀCH); 4.45 (s, 2H C5H4); 4.35 (s, 2H, C5H4);
4,09 (s, 5H, C5H5); Carbon-NMR spectrum for com-
pound 6: l 214.55 (1C, trans CO); 211.53 (4C, cis CO);
155.29 (2C, Py); 146.30 (1C, Py); 135.65 (2C, Py);
120.86 (2C, CHꢀCH); 80.96 (1C, C5H4); 70.52 (2C,
C5H4); 69.54 (5C, C5H5); 67.85 (2C, C5H4); Proton-
NMR spectrum for compound 10: l 7.52 (d, 2H,
C24H15NFeO5W, Mw=637.08: C, 45.2; H, 2.4; N, 2.2.
Found: C, 44.3; H, 2.3; N, 2.1.
3. Results and discussion
3JH–H=8.07 Hz, C6H4); 7.41 (d, 2H, JH–H=8.10 Hz,
3
3.1. Synthesis and characterization of the ferrocenyl
compounds
3
C6H4); 6.96 (d, 1H, JH–H=15.98 Hz, CHꢀCH); 6.60
3
(d, 1H, JH–H=16.12, CHꢀCH); 4.43 (s, 2H, C5H4);
The new ferrocenyl compounds were prepared by
conventional organic synthetic procedures (Wittig reac-
tions), or modification of the literature methods. All the
4.29 (s, 2H, C5H4); 4.09 (s, 5H, C5H5); Carbon-NMR
spectrum for compound 10: l 219.41 (1C, trans CO);
214.05 (4C, cis CO); 143.31 (1C, C6H4); 132.78 (2C,
C6H4); 128.97 (1C, C6H4); 126.20 (2C, C6H4); 123.95
(1C, CHꢀCH); 123.60 (1C, CHꢀCH); 107.86 (1C, CN);
81.71 (1C, C5H4); 70.10 (2C, C5H4); 69.89 (5C, C5H5);
67.53 (2C, C5H4). Elemental Anal. Calc. for compound
6, C22H15NFeO5Cr, Mw=481.21: C, 54.9; H, 3.1; N,
2.9. Found: C, 54.3; H, 3.5; N, 3.1. Elemental Anal.
Calc. for compound 10, C24H15NFeO5Cr, Mw=
505.23: C, 57.1; H, 3.0; N, 2.8. Found: C, 57.3; H, 3.4;
N, 2.9.
1
compounds were characterized by means of IR-, H-
and 13C-NMR spectroscopy and satisfactory micro-
analyses. The y bonded p6 compounds (2 and 4) were
obtained refluxing the corresponding ferrocenyl ligands
(1 or 3) with chromium hexacarbonyl for 12 h in butyl
ether. Fig. 1 shows the general procedure to the obten-
tion of compounds 2, 4, 5–11. The coordination of the
ferrocenyl cyano and pyridyl derivatives were carried
out by previously reacting Me3NO with the correspond-

J. Mata et al. / Journal of Organometallic Chemistry 562 (1998) 197–202
201
Table 2
expected from the longer conjugated chain of 3). The
rest of the compounds show half wave potentials for
the ferrocene center, that are higher than that shown
for ferrocene, meaning that ferrocenyl ligands con-
nected to electronegative/electron-accepting fragments
show some electron transfer between the iron and the
pyridyl or cyano fragments. Coordination to the metal
carbonyl implies an additional increase on the half
wave pontential of the ferrocene center in all the cases,
indicating that the carbonyl metal attached to the ferro-
cenyl ligands promotes an additional increase of the
electron transfer in the direction Fe“M(CO)n. On the
other hand, the second oxidation step for 2 and 4
displays complete electrochemical irreversibility with an
anodic peak potential Ep,a showing a slight cathodic
shift with respect to E1/2 for (p6-C6H6)Cr(CO)3, in spite
of the positive charge in 2+ and 4+.
Experimental nonlinear response of the compounds synthetised previ-
ously; ai and i0 in 10−30 esu
Compound
umax (nm)
i
b0b
2
4
5
6
7
8
9
10
11
304
334
468
401
487
491
473
481
487
193
300
21
63
95
101
203
271
375
119
164
4
23
12
12
34
39
48
a Measured in chloroform by the HRS method and calculated with
respect to i of PNA in chloroform (23×10−30 esu). HRS at 1.06
mm. Values 915%.
b Calculated from the two level model.
In contrast, the cyclic voltammetric experiments for
CpFe(p5-C5H4)–CHꢀCH-py-Cr(CO)5 (6) show two
electrochemical quasireversible oxidation waves. Both
processes are coupled with a reduction on the reverse
sweep. The peak separation, 75 and 77 mV, and the
virtually equal peak currents suggest that we are ob-
serving two quasireversible one-electron oxidations.
The half wave potential E1/2 for 6+/6 shows an anodic
shift of 30 mV relative to that of 5+/5 as expected
based on the electron-withdrawing nature of the
(C6H5)Cr(CO)3 moiety. Compounds 7 and 8 display
complete electrochemical irreversibility for the oxida-
tion of Mo and W respectively, as reported previously
for 8 ([5]b).
The chromium based compound 10, shows two elec-
trochemical quasireversible oxidation waves. Both pro-
cesses are coupled with a reduction on the reverse
sweep. The related tungsten based compound 11, dis-
plays complete electrochemical irreversibility for the
oxidation of W. This result is consistent with those
studies reported before where chromium and tungsten
ferrocenyl derivatives show quasireversible and com-
plete irreversible oxidation waves respectively [5,13].
3.3. NLO properties
The UV–vis spectra of compounds 1–11 show low
absorption in the region where second harmonic is
generated (1064 nm) in the solvent where the nonlinear
response was studied by the hyper Rayleigh scattering
method (chloroform). Table 2 shows the values of i
obtained for compounds 1–11. These values are com-
parable in magnitude to those obtained for the bimetal-
lic cationic compounds reported by Heck et al. ([3]a,b)
The simple two-level method allowed us to calculate the
static hyperpolarizabilities, i0; for compounds 7–11,
the differences between i and i0 are rather large be-
cause the maximum of absorption of these compounds
are relatively close to 532 nm. Since the molecules that
we have studied display several optical transitions, the
static hyperpolarizabilities given are only a crude esti-
mation. In their estimations (two level model) we have
only taken into account the transition closest to the
harmonic frequency, and damping has been neglected.
Compounds 2, 4, 6, 7, 8, 10 and 11, being a combina-
tion of two neutral fragments show the electron-accept-
ing capability of the chromium carbonyl moiety
through both y- and a |-bonding to the electron
donating ferrocene derivative. It is noteworthy pointing
out that previous studies have shown that a series of
(p6-arene)tricarbonylchromium derivatives show low
values of i ([2]d), while |-bonded derivatives of
(CO)5Cr(p-X-py) (X=H, NH2, C6H5, CHO, COCH3)
gave interesting results [14]. Compound 9 shows a
higher value of i than 5, probably because of the
longer chain separating the donor and acceptor frag-
ments. As seen from Table 2, y-donation affords a
much higer value of the hyperpolarizability than that
shown of the ligands upon |-donation. The result can
Table 1
Cyclic voltammetric data for compounds 1–11
Compound
E1/2(V) (DEp (mV))
Fe-based
E1/2(V) (DEp (mV))
M-baseda
Ferrocene
1
2
3
4
5
6
7
8
9
10
11
0.446 (106)
0.440 (85)
0.481 (74)
0.436 (87)
0.448 (91)
0.498 (70)
0.520 (75)
0.518 (83)
0.530 (77)
0.482 (82)
0.499 (72)
0.495 (68)
–
–
0.808b
–
0.784b
–
0.928 (77)
1.110b
1.125b
–
0.981 (66)
1.146b
a M=Cr, Mo, W.
b Irreversible peak, measured at 100 mV s−1
.

202
J. Mata et al. / Journal of Organometallic Chemistry 562 (1998) 197–202
Houbrechts, M. Kauranen, K. Clays, A. Persoons, J. Mater.
Chem. 7 (1997) 2175.
be interpreted in terms of the linkage to the carbonyl
fragment. In the case of the y-bonded compounds (2
and 4), the orbitals participating in the conjugated
delocalization of the charge are the same used in the
formation of the bond to chromium. The lower increase
of i in the |-bonded complexes can be explained in
terms of the role of the metal in lowering the energy of
the pyridyl acceptor orbital and this way to increase the
accepting properties of the pyridyl ligand [15].
There is a considerable increase in the value of i
when we change the metal, in the order W\Mo\Cr,
probably due to the decrease of the Pauling electroneg-
ativity in the same direction and to the higher backdo-
nation capability of W compared to Cr and Mo. As we
have previously mentioned, all the ferrocenyl-metalcar-
bonyl compounds that we report in the present work,
show higher half wave potentials (based on the iron
center), than that shown for ferrocene. We interprete
this result as a consequence of a partial charge transfer
to the metalcarbonyl fragment.
[2] See for example: (a) I.R. Whittall, M.G. Humphrey, A. Per-
soons, S. Houbrechts, Organometallics 15 (1996) 1935. (b) M.
Tamm, A. Grzegorzewski, Th. Steiner, Th. Jentzsch, W.
Werncke, Organometallics 15 (1996) 4984. (c) I.R. Whittall,
M.G. Humphrey, A. Persoons, S. Houbrechts, D.C.R. Hockless,
Organometallics 15 (1996) 5738. (d) L.T. Cheng, W. Tam, G.R.
Meredith, S.R. Marder, Mol. Cryst. Liq. Cryst. 189 (1990) 137.
(e) B.J. Coe, J.A. Harris, L.J. Harrington, J.C. Jeffery, L.H.
Rees, S. Houbrechts, A. Persoons, Inorg. Chem submitted,
(1997). (f) B.J. Coe, J.P. Essex-Lopresti, S. Houbrechts, A.
Persoons. J.C.S., Chem. Commun. (1997) 1645. (g) S. Di Bella,
I. Fragala`, T. Marks, M.A. Ratner, J. Am. Chem. Soc. 118
(1996) 12747. (h) S. Di Bella, I. Fragala`, I. Ledoux, T.J. Marks.
J. Am. Chem. Soc. 117 (1995) 9481.
[3] (a) U. Behrens, H. Brussard, U. Hagenau, J. Heck, E. Hen-
drickx, J. Kornich, J.G.M. van der Linden, A. Persoons, A.L.
Spek, N. Veldman, B. Voss, H. Wong, Chem. Eur. J. 2 (1996)
98. (b) U. Hagenau, J. Heck, E. Hendrickx, A. Persoons, T.
Schuld, H. Wong, Inorg. Chem. 35 (1996) 7863. (c) R. Loucif-
Saiba, J.A. Delaire, L. Bonazzola, G. Doisneau, G. Balavoine,
T. Fillebeen-Khan, I. Ledoux, G. Puccetti, Chem. Phys. 167
(1992) 369. (d) S. Houbrechrts, K. Clays, A. Persoons, V.
Cadierno, M.P. Gamasa, J. Gimeno, Organometallics 15 (1996)
5266.
In summary, we have presented a series of hetero-
bimetallic compounds showing large values of i. These
compounds are easy-to-make, stable and soluble in
most common organic solvents, showing low absorp-
tion in the second harmonic generation region. The
electronic and optical properties of these compounds
can be easily tuned by modifying the chainlength,
bonding to the metal and change of the metal.
[4] P.L. Pauson, W.E. Watts, J. Organomet. Chem. (1963) 2990.
[5] (a) M.M. Bhadbhade, A. Das, J.C. Jeffery, J. McCleverty, J.A.
Navas, M.D. Ward. J. Chem. Dalton Trans. (1995) 2769. (b) S.
Sakanishi, D.A. Bardwell, S. Couchman, J.C. Jeffery, J.A. Mc-
Cleverty, M.D. Ward. J. Organomet. Chem. 528 (1997) 35–45.
(c) S. Toma, A. Gaplovosky, P. Elecko, Chem. Pap. 39 (1985)
115.
[6] C. Degrand, A. Radecki-sudre, J. Organomet. Chem. 268 (1984)
63 and references cited therein.
[7] E. Hendrickx, K. Clays, A. Persoons, C. Dehu, J.L. Bre´das, J.
Am. Chem. Soc. 117 (1995) 3547.
[8] S. Houbrechts, K. Clays, A. Persoons, Z. Pikramenou, J.-M.
Lehn, Chem. Phys. Lett. 258 (1996) 485.
Acknowledgements
We thank the Generalitat Valenciana for the finan-
cial support of this work (GV-C-CN-09-071-96) and the
Belgian National Fund for Scientific Research-Flanders
(G.0308.96), the Belgian Government (IUAP-P4-11)
and the University of Leuven (GOA/1/95). We would
also like to thank Professor Robert H. Crabtree for his
valuable help and wise advices, and to Professor Miguel
Carda for the design of the organic pathways to obtain
the ligands. S.H. is a Research Assistant of the Belgian
National Fund for Scientific Research.
[9] K. Clays, A. Persoons, Rev. Sci. Instrum. 63 (1992) 3285.
[10] E. Hendrickx, C. Dehu, K. Clays, J.L. Bre´das, A. Persoons, in:
G.A. Lindsay, K.D. Singer, Polymers for Second-Order Nonlin-
ear Optics, American Chemical Society, Washington DC, 1995,
pp. 82–94.
[11] M.C. Flipse, R. de Jonge, R.H. Woudenberg, A.W. Marsman,
C.A. van Walree, L.W. Jenneskens, Chem. Phys. Lett. 245 (1995)
297.
[12] I.D. Morrison, R.G. Denning, W.M. Laidlaw, M.A. Stammers,
Rev. Sci. Instrum. 67 (1996) 1445.
[13] (a) L.K. Yeung, J.E. Kim, Y.K. Chung, P.H. Rieger, D.A.
Sweigart, Organometallics 15 (1996) 3298. (b) I.P. Gubin, V.S.
Khandkarova, J. Organomet. Chem. 22 (1970) 449.
[14] L.T. Cheng, W. Tam, D.F. Eaton, Organometallics 9 (1990)
2856.
References
[15] (a) D.R. Kanis, M.A. Ratner, T.J. Marks, J. Am. Chem. Soc.
114 (1992) 10338. (b) D.R. Kanis, P.G. Lacroix, M.A. Ratner,
T.J. Marks, J. Am. Chem. Soc. 116 (1994) 10089. (c) D.R.M.A.
Ratner, T.J. Marks, Chem. Rev. 94 (1994) 195.
[1] (a) N.J. Long, Angew. Chem. Int. Ed. Engl. 34 (1995) 21–38. (b)
I.R. Whittall, A.M. McDonagh, M.G. Humphrey, M. Samoc,
Adv. Organomet. Chem. 42 (1997) 291. (c) T. Verbiest, S.
.
.

200
J. Mata et al. / Journal of Organometallic Chemistry 562 (1998) 197–202
Fig. 1. Reaction scheme for the syntheses of compounds 2, 4, 5–11.
ing metal hexacarbonyl, and then adding the ferrocenyl
ligand. This method showed better results than any
other photochemical or thermal reaction used.
The pyridyl and cyano ferrocenyl ligands were ob-
tained in moderate high yields, with high selectivities on
the trans isomers, as stated in Section 2.
Comparison of the IR spectra of the |-bonded cyano
and pyridyl ferrocene derivatives (6, 7, 8, 10, 11) shows
that the compounds containing W have lower values of
w(CO), meaning that W is more capable than Cr and
Mo of reducing its electron density by y back-donation
to CO. As we comment in the NLO section, this fact
supports the observation of the larger hyperpolarizabil-
ities found for our compounds containing W. Similar
results have been reported for group 6 metal Fischer
type carbenes [14].
3.2. Cyclic 6oltammetry
Electrochemical data obtained for the compounds
studied are summarized in Table 1.
Cyclic voltammetric investigations for the ferrocene
derivatives CpFe(p5-C5H4)–(CHꢀCH)–(C6H5) (1), Cp-
Fe(p5-C5H4)–(CHꢀCH)2-(C6H5) (3), CpFe(p5-C5H4)–
(CHꢀCH)-py (5) and CpFe(p5-C5H4)–(CHꢀCH)–(C6-
H4)–CN (9) show a redox behavior analogous to the
one observed for ferrocene, that is, an electrochemically
quasireversible one-electron oxidation.
The half wave potential of the quasireversible wave
for compounds 1 and 3 is sligthly lower than that
shown for ferrocene, indicating that the oxidation is
favoured by the delocalization of the charge along the
conjugated system of the ligand (E1/2(3)BE1/2(1), as