Linear and non-linear optical properties of arene-Fe-Cp complexes

Article abstract of DOI:10.1016/S0022-328X(99)00497-0

Ionic [arene-Fe-Cp]⁺ fragments were used as electron acceptors in combination with dimethylamino or pyridoneiminyl donors to form push-pull substituted chromophores. While a tolane derivative showed high second-order non-linear optical activity (measured by hyper-Rayleigh scattering), an analogous pyridoneimine derivative did not. With the help of an X-ray crystal structure analysis this discrepancy was traced back to a cyanine-like structure of the imine. Semiempirical INDO/S calculations support this conclusion.

Full text of DOI:10.1016/S0022-328X(99)00497-0

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 592 (1999) 109–114
Linear and non-linear optical properties of areneꢀFeꢀCp
complexes
Christoph Lambert a,*, Wolfgang Gaschler a, Manfred Zabel b, Ralf Matschiner c,
Ru¨diger Wortmann c
a Institut fu¨r Organische Chemie, Uni6ersita¨t Regensburg, Uni6ersita¨tsstrasse 31, D-93040 Regensburg, Germany
b Instrumentelle Analytik der Fakulta¨t fu¨r Chemie und Pharmazie, Uni6ersita¨t Regensburg, Uni6ersita¨tsstrasse 31, D-93040 Regensburg, Germany
c Fachbereich Chemie, Uni6ersita¨t Kaiserslautern, Erwin-Schro¨dinger-Straße, D-67663 Kaiserslautern, Germany
Received 14 May 1999; received in revised form 16 June 1999; accepted 29 July 1999
Abstract
Ionic [areneꢀFeꢀCp]+ fragments were used as electron acceptors in combination with dimethylamino or pyridoneiminyl donors
to form push–pull substituted chromophores. While a tolane derivative showed high second-order non-linear optical activity
(measured by hyper-Rayleigh scattering), an analogous pyridoneimine derivative did not. With the help of an X-ray crystal
structure analysis this discrepancy was traced back to a cyanine-like structure of the imine. Semiempirical INDO/S calculations
support this conclusion. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Non-linear optics; Hyper-Rayleigh scattering; UV–vis spectroscopy; Iron; Chromophores; Palladium-catalysed cross coupling
chemical properties have led to intense investigations
1. Introduction
[3].
Recently, we have focused on ammonio (R3N+) and
The electron-donating and -accepting capabilities of
phosphonio (R3P+) acceptor groups that polarise p-
organoligandꢀmetal fragments have been successfully
electron systems in the p-plane in order to form one-di-
applied to push–pull substituted chromophores in or-
der to achieve high second-order non-linear optical
(NLO) responses [1]. Ferrocene, especially, is widely
used as an organometallic electron donor [2]. It com-
mensional NLO-chromophores [4]. In contrast, the
[CpꢀFe]+ fragment might polarise p-electron systems
perpendicular to the p-plane. The purpose of this paper
is to elucidate the linear and non-linear optical proper-
bines the advantage of high stability, electrochemical
ties of organometallic NLO chromophores where the
activity and the fact that it is easily incorporated into
larger organometallic molecules by standard synthetic
procedures. The electron-donor capacity of ferrocene in
terms of promoting charge transfer (CT) excitation and,
associated with this CT, quadratic hyperpolarisability,
acceptor functionality is a [benzeneꢀFeꢀCp]+ moiety
being combined with a conventional electron donor via
a p-electron spacer attached to the benzene ring.
has been estimated to be about that of a methoxy
group [2a,b]. In contrast, the electron-acceptor counter-
parts of ferrocene, ionic [areneꢀFeꢀCp]+ fragments,
have not been used in organometallic NLO chro-
mophores, although their electrochemical and photo-
2. Results and discussion
While chloroarenes are usually not susceptible to
palladium-catalysed Stille couplings under normal con-
ditions [5], the analogous [areneꢀFeꢀCp]+ complexes
readily undergo cross-coupling reactions with
a
dimethylamino-substituted phenylethynyltrimethylstan-
nane (see Scheme 1), leading to the one-dimensional
tolane chromophores 1 and the pentamethyl analogue
2. The latter compound was synthesised because of its
* Corresponding author. Tel.: +49-941-9434373; fax: +49-941-
9434984.
E-mail address: christoph.lambert@chemie.uni-regensburg.de (C.
Lambert)
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00497-0

110
C. Lambert et al. / Journal of Organometallic Chemistry 592 (1999) 109–114
Fig. 2. HRS signal S2ꢀ as a function of the fundamental laser
intensity Sꢀ for acetonitrile solutions of compound 2 (c=9.18 mol
Scheme 1.
m
−3), 4 (c=15.71 mol m−3) and the calibration standard pNA
(c=16.12 mol m−3). Each point represents an average over 300 laser
pulses. The solid lines are quadratic fits.
bands of 1 and of 2 are weakly solvatochromic (1: 352
nm in THF; 2: 344 nm in THF) that of 4 is not (365 nm
in THF).
The quadratic hyperpolarisability i of 2 and of 4
were determined by hyper-Rayleigh scattering (HRS) in
MeCN at 1064 nm fundamental energy [7]. The HRS
experiments were carried out as described with all
details in ref. [7b]. The experimentally detected HRS
signal S2ꢀ is a quadratic function of the incident inten-
sity Sꢀ. Typical results are shown in Fig. 2 for deriva-
tives 2 and 4 as well as for the reference standard
paranitroaniline (pNA). Such curves were taken and
fitted for several different concentrations. The i values
were finally calculated from the slopes of linear regres-
sion of the quadratic fit coefficients versus the concen-
tration of the solutes. pNA was used as a reference
compound {i1064=49.6×10−50 cm3 V−2 (134×
10−30 esu) in MeCN, iT convention used [7b, 8]}.
Measurements of 1 were hampered due to the increased
light sensitivity of 1 versus 2. The hexamethyl derivative
2 was stable under the conditions of the HRS experi-
ment. The i value of 2 is 120×10−50 cm3 V−2 (318×
10−30 esu) and the static value is 60×10−50 cm3 V−2
(168×10−30 esu) (static values calculated by extrapola-
tion via two-level model). This value is about that of
4-dimethylamino-4%-cyanotolane whose static i value is
66×10−50 Cm3 V−2 (174×10−30 esu, umax=373 nm
in CHCl3) [9]. In contrast, imine 4 showed no measur-
able SHG signal. This behaviour deserves an explana-
tion, as the CT bands of 2 and of 4 are at similar
energy and have almost equal intensity.
Fig. 1. UV–vis spectra of 1, 2, and 4 in MeCN.
higher light stability compared to 1. In addition,
[areneꢀFeꢀCp]+ was derivatised by nucleophilic substi-
tution with imine 3, which yielded the pyridoneimine 4.
These reactions demonstrate the strong electron-with-
drawing character of [areneꢀFeꢀCp]+ and thus
prompted us to investigate their optical properties.
Derivatives 1 and 2 combine the ionic iron acceptor
with the conventional dimethylamino donor, whereas
the donor in 4 is the N-methylpyrid-4-yl group, which
should act as an electron donor because it gains aro-
maticity upon electron loss.
Fig. 1 shows the UV–vis spectra of 1, 2 and 4 in
MeCN. All derivatives have intense bands at ca. 348
nm (1), 336 nm (2), 365 nm (4), which are attributed to
intraligand CT excitations. At longer wavelengths,
there are weaker MLCT bands at ca. 400–420 nm and
at shorter wavelengths there are LMCT or MLCT
bands at ca. 300 nm [6]. From deconvolutions of the
spectra with Gaussian curves we estimate the band
width at half-height of the intraligand CT bands to be
ca. 5700 cm−1 for 1, ca. 5400 cm−1 for 2, and to be
distinctly smaller (ca. 2900 cm−1) for 4. While the CT
The difference in hyperpolarisability of 2 and 4 might
be attributed to an unfavourable electronic structure of
4 which expresses itself in a small bond length alterna-
tion (BLA) in the conjugation path. It is known from

C. Lambert et al. / Journal of Organometallic Chemistry 592 (1999) 109–114
111
Table 1
a
,
Selected bond distances (A) of 4
N1ꢀC6
N1ꢀC12
N2ꢀC14
N2ꢀC15
N2ꢀC17
C6ꢀC7
1.368(3)
1.327(3)
1.350(3)
1.353(3)
1.461(3)
1.417(3)
1.426(3)
1.407(4)
C8ꢀC9
1.400(4)
1.402(4)
1.409(3)
1.422(3)
1.429(3)
1.355(3)
1.352(3)
C9ꢀC10
C10ꢀC11
C12ꢀC13
C12ꢀC16
C13ꢀC14
C15ꢀC16
Scheme 2.
C6ꢀC11
C7ꢀC8
the studies of Marder and co-workers that the BLA can
be used as an indicator for favourable electronic struc-
ture: e.g. for many olefins the optimal BLA at which i
displays a maximum is ca. 5 pm [10]. In cases where the
BLA is small or zero, the difference between excited-
and ground-state dipole moment vanishes (cyanine
limit) and, as can be seen by the two-level equation (Eq.
(1), where v01 is the transition moment, Dv is the dipole
moment difference between ground and excited state
and ꢀ01 is the CT transition energy), i approaches
zero. It is clear from the strong CꢁC triple bond force
constants that the tolane in 2 will not experience strong
alternation in bond lengths upon changing the donor or
acceptor functionalities. This is completely different in
4, where two mesomeric structures might equally con-
tribute to the electronic situation: one with a chinoid
pyridyl and a benzoid phenyl ring, the other with the
benzoid and chinoid structures exchanged (see Scheme
2). A single-crystal X-ray analysis of 4 revealed that the
pyridyl moiety has a chinoide distortion compared to
pyridinium structures (see Fig. 3 and Table 1). The
a Angle C12ꢀN1ꢀC6=124.7(2)°, angle between the phenyl ring and
the pyridinium ring=43.2°.
metrical cyanines show a zero BLA, this does not
necessarily hold true for unsymmetrical donor–accep-
tor systems where the dipole moment difference is zero
at a BLA"0, although this value might be quite small
[12].
6v201·Dv
izzz
=
(1)
'2ꢀ201
Our interpretation is supported by semiempirical
INDO/S calculations (Table 2) [13]. The CIS was based
on ten occupied and ten virtual orbitals resulting in a
CI matrix with 100 microstates. The calculated results
from the CI computations as well as the static i values
calculated by a sum-over-states method are given in
Table 2 and are in reasonable agreement with the
experiment. While the CT bands between ligands and
the metal centre are generally weak, the CI calculations
revealed strong intraligand CT excitations (LCT) at 341
nm (2) and 349 nm (4) in good agreement with the
experiment. However, the dipole moment difference
associated with the LCT excitation is large (44.0×
10−30 C m, 13.2 D) for 2 but rather small (4.0×10−30
C m, 1.2 D) for 4. Using the two-level approximation
(Eq. (1)), this explains the high izzz value for 2 and the
,
C12ꢀN1 bond (1.33 A) is ca. 5 pm longer and the
,
C6ꢀN1 bond (1.37 A) is ca. 5 pm shorter than in
aromatic Schiff bases [11]. The BLA in the imine bridge
thus is ca. 4 pm. Together with the distinctly smaller
band width at half-height, these features indicate a
cyanine-like structure with a too small BLA and a
negligible i value. It must be stressed that while sym-
Fig. 3. X-ray structure of 4. Hydrogen and counter anion are omitted for clarity.

112
C. Lambert et al. / Journal of Organometallic Chemistry 592 (1999) 109–114
Table 2
INDO/S calculated optical properties of 2 and 4
umax (nm)
f
Band type
Dv (10−30 C m) (D)
i0 (10−50 cm3 −2) (10−30 esu)
V
2
4
374
341
317
0.012
0.986
0.023
MLCT
LCT
LMCT
2.7 (0.8)
44.0 (13.2)
16.7 (5.0)
38 (102) a
367
349
329
0.015
0.722
0.022
MLCT
LCT
MLCT
2.7 (0.8)
4.0 (1.2)
20.4 (6.1)
B7.2 (20) b
a izzz
b All tensor components are smaller.
.
absence of any measurable SHG signal for 4 in accor-
dance with the calculations (see Table 2). In addition,
the INDO/S calculations suggest that in 2 the only CT
excitation contributing to i is that within the tolane
system; the MLCT excitation at longer wavelength and
the LMCT at shorter wavelength are polarised perpen-
dicular to the tolane (zzz) axis and no other i tensor
component besides izzz is significant in 2.
mmol) and aluminiumchloride (20.0 g, 150 mmol) were
stirred under a nitrogen atmosphere at 120°C for 14 h.
The melt was cooled and hydrolysed gently with ice.
The resulting suspension was filtered, the filtrate was
washed with petrol ether and the product was precipi-
tated by adding an NH4PF6 solution (7 g in 50 ml of
water) to the aqueous phase. The crude product was
filtered off, dissolved in acetone, dried over Na2SO4 and
precipitated by adding diethyl ether. Yield 7.20 g (43%)
1
of a yellow powder; m.p. (dec.) 200°C; H-NMR (250
3. Conclusions
MHz, acetone-d6): l=4.95 (s, 5 H, Cp), 2.78 (s, 6 H,
o-CH3), 2.67 (s, 6 H, m-CH3), 2.60 (s, 3 H, p-CH3);
13C-NMR (62.9 MHz, acetone-d6): l=108.7, 100.9,
100.5, 99.3, 81.0, 18.6, 17.8, 17.5; MS (PI-FD, acetone-
d6) m/z: 303 (cation+); C16H20ClF6FeP (448.6), Anal.
Calc. C 42.84, H 4.49, found C 42.61, H 4.31%.
[AreneꢀFeꢀCp]+ substituents are effective acceptor
units that can be used in push–pull substituted p-elec-
tron systems for non-linear optics. The quadratic hy-
perpolarisability of 2 is comparable to that of e.g.
4-dimethylamino-4%-cyanotolane at somewhat shorter
CT absorption wavelength. In this respect, one can
conclude that the electron-acceptor strength of a
4.2. p6-[4-(N,N-Dimethylamino)phenylethynyl-
benzene]-p5-cyclopentadienyl-iron-(II)-hexafluorophos-
phate (1)
[areneꢀFeꢀCp]+ moiety is about that of
a
4-
cyanophenyl group. A major drawback of the com-
pounds studied are MLCT bands at longer wavelengths
than the intraligand CT band. These MLCT bands do
not contribute to the hyperpolarisability and limit the
blue transparency as is often observed for organometal-
lic derivatives [1]. One possible way to improve the
efficiency–transparency trade-off is to use more ex-
tended p-systems with LCT bands at similar or lower
energy than the MLCT bands and which then display
intrinsically higher optical non-linearities. Although
this will improve the non-linearity only because of the
red shift of the intraligand CT, there is at least no loss
of transparency due to ineffective MLCT bands.
h6 - Chlorobenzene - h5 - cyclopentadienyl - iron - (II) -
hexafluorophosphate [15] (404 mg, 0.90 mmol), [4-
(N,N - dimethylamino)phenylethynyltributylstannane
(prepared by the reaction of lithio-4-(N,N-dimethyl-
amino)phenylacetylene with chlorotributylstannane in
Et2O at −78°C) (677 mg, 1.58 mmol) and Pd(PPh3)4
(78 mg, 0.068 mmol) were dissolved in 5 ml of dry
DMF under nitrogen atmosphere. This solution is
stirred at 110°C for 2.5 h under exclusion of light. The
solvent was removed in vacuo and the resulting dark oil
was washed with petrol ether. The residue was dis-
solved in acetone and precipitated with a large excess of
diethyl ether. The crude product was again dissolved in
acetone, filtered over neutral alumina and precipitated
with diethyl ether. Yield 340 mg (78%) of a yellow
4. Experimental
1
powder; m.p. (dec.) 202°C; IR (KBr) 2212 cm−1; H-
4.1. p6-Chloropentamethylbenzene-p5-cyclopentadienyl-
iron-(II)-hexafluorophosphate
NMR (250 MHz, acetone-d6): l=7.48 (m, AA%, 2 H),
6.78 (m, BB%, 2 H), 6.59 (5 H, phenyl), 5.29 (s, 5 H,
Cp), 3.05 (s, 6 H, NMe2); 13C-NMR (62.9 MHz, ace-
tone-d6): l=152.4, 134.3, 112.6, 107.7, 96.0, 90.1, 89.9,
89.2, 87.7, 83.5, 79.2, 40.0; MS (PI-FD, MeOH) m/z:
Chloropentamethybenzene [14] (6.8 g, 37 mmol), fer-
rocene (14.9 g, 80 mmol), aluminium powder (1.0 g, 37

C. Lambert et al. / Journal of Organometallic Chemistry 592 (1999) 109–114
113
342 (cation+); C21H20F6FeNP (487.2), Anal. Calc. C
4.5. X-ray crystallographic data of 4
51.77, H 4.14, N 2.87, found C 51.71, H 4.32, N
3.03%.
C17H17F6FeN2P, M=450.2. Data were collected on a
STOE-Imaging Plate System with monochromatic Mo–
4.3. p6-[4-(N,N-Dimethylamino)phenylethynyl-
pentamethylbenzene]-p5-cyclopentadienyl-iron-(II)-
hexafluorophosphate (2)
Ka radiation (u=0.71069 A); T=120(2) K; mono-
,
,
clinic, space group P21/c; a=7.3316(6) A, b=
,
,
14.1435(9) A, i=99.809(9)°, c=16.7866(14) A; V=
1715.2(2) A ; Z=4; zcalc.. =1.743 g cm−3; red plates
3
,
(0.12×0.20×0.44 mm3); scan range 1.89BqB25.61°;
11 492 collected, 3155 independent and 2247 observed
(I\2|(I)) reflections; the structure was solved with
direct methods (SHELXS-93) and refined against F2. All
non-hydrogen atoms were refined anisotropically, the
hydrogen atoms isotropically in fixed idealised positions
(244 parameters). Goodness-of-fit=0.862; R1=0.0292
h6 -Chloropentamethylbenzene-h5 -cyclopentadienyl-
iron-(II)-hexafluorophosphate (717 mg, 1.60 mmol), [4-
(N,N-dimethylamino)phenylethynyltributylstannane
(814 mg, 1.90 mmol) and Pd(PPh3)4 (100 mg, 0.087
mmol) were dissolved in 5 ml of dry DMF under
nitrogen atmosphere. This solution was stirred at 110°C
for 3 h under exclusion of light. The solvent was
removed in vacuo and the resulting dark oil was washed
with petrol ether. The residue was dissolved in acetone
and precipitated by a large excess of diethyl ether. The
crude product was purified by column chromatography
on neutral alumina first with CH2Cl2 and then with
CH2Cl2+5% acetone. The product was precipitated
from concentrated solutions with diethyl ether. Yield
260 mg, (29%) of a yellow powder; m.p. (dec.) 258–
(I\2|(I)), wR2=0.0616; min/max residual electron
−3
,
density: −0.47/0.52 e A
.
5. Supplementary material
Crystallographic data for the structural analysis has
been deposited with the Cambridge Crystallographic
Data Centre, CCDC no 120568 for compound 4.
Copies of this information may be obtained free of
charge from The Director, CCDC, 12 Union Road,
Cambridge, CB2 1EZ, UK (Fax: +44-1223-336-033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://
www.ccdc.cam.ac.uk).
260°C; IR (KBr) 2203 cm−1 1H-NMR (250 MHz,
;
acetone-d6): l=7.54 (m, AA%, 2 H), 6.80 (m, BB%, 2 H),
4.87 (s, 5 H, Cp), 3.06 (s, 6 H, CH3), 3.05 (s, 6 H,
NMe2), 2.63 (s, 6 H, CH3), 2.60 (s, 3 H, CH3); 13C-
NMR (62.9 MHz, acetone-d6): l=152.3, 134.0, 112.7,
108.3, 101.5, 100.1, 100.0, 99.9, 88.4, 83.4, 80.4, 40.1,
19.4, 17.3, 17.3; MS (PI-FD, CH2Cl2) m/z: 412 (cat-
ion+); C26H30F6FeNP (557.4), Anal. Calc. C 56.03, H
5.43, N 2.51, found C 55.75, H 5.44, N 2.43%.
Acknowledgements
4.4. p6-[(N-Methyl-4-pyridoneiminyl)benzene]-
We are grateful to the Fonds der Chemischen Indus-
trie (Liebig grant to C.L.), the Deutsche Forschungsge-
meinschaft (Habilitandenstipendium), and the Stiftung
Volkswagenwerk for financial funding, and especially to
Professor J. Daub for his kind support at Regensburg.
-p5-cyclopentadienyl-iron-(II)-hexafluorophosphate (4)
N-Methyl-4-pyridoneimine [16] (6.2 g, 5.7 mmol),
h6-chlorobenzene-h5-cyclopentadienyl-iron-(II)-hexa-
fluorophosphate (1.9 g, 5.0 mmol) and K2CO3 (2.8 g)
were dissolved in 10 ml of DMF and stirred at 60°C for
14 h. Water (100 ml) was added and the mixture
extracted with CH2Cl2. The solvents were removed in
vacuo; the residue was purified by column chromatog-
raphy on neutral alumina first with acetone–water (4:1)
and then with acetone–water (1:1). The eluents were
concentrated and extracted with CH2Cl2. The product
was precipitated by adding diethyl ether. Yield 600 mg
(27%) of an orange powder, m.p. (dec.) 200–201°C; IR
References
[1] For a general introduction to non-linear optics of organic com-
pounds, see (a) J.J. Wolff, R. Wortmann, Adv. Phys. Org.
Chem. 32 (1999) 121. For organometallic NLO compounds, see
(b) I.R. Whittall, A.M. McDonagh, M.G. Humphrey, M.
Samoc, Adv. Organomet. Chem. 42 (1998) 291l. (c) N.J. Long,
Angew. Chem. 107 (1995) 37.
[2] See e.g. (a) J.C. Calabrese, L.-T. Cheng, J.C. Green, S.R.
Marder, W. Tam, J. Am. Chem. Soc. 113 (1991) 7227. (b) D.R.
Kanis, M.A. Rantner, T.J. Marks, J. Am. Chem. Soc. 114 (1992)
10338. (c) S.R. Marder, J.W. Perry, B.G. Tiemann, W.P. Schae-
fer, Organometallics 10 (1991) 1896. (d) U. Behrens, H. Brus-
saard, U. Hagenau, J. Heck, E. Hendrickx, J. Ko¨rmich, J.G.M.
van der Linden, A. Persoons, A.L. Spek, N. Veldman, B. Voss,
H. Wong, Chem. Eur. J. 2 (1996) 98.
(KBr) 1650 cm−1 1H-NMR (250 MHz, acetone-d6):
;
l=7.56 (m, AA%, 2 H), 6.53 (m, BB%, 2 H), 6.17 (5 H,
phenyl), 5.00 (s, 5 H, Cp), 3.76 (s, 6 H, NMe); 13C-
NMR (62.9 MHz, acetone-d6): l=161.0, 141.5, 128.9,
113.1, 87.9, 83.1, 80.1, 77.1, 43.6; MS (PI-FD, CH2Cl2)
m/z: 305 (cation+); C17H17F6FeN2P (450.2), Anal. Calc.
C 45.36, H 3.81, N 6.22, found C 45.41, H 3.99, N
6.22%.
[3] (a) A. Darchen, J. Organomet. Chem. 302 (1986) 389. (b) W.A.
Hendrickson, M.C. Palazzotto, Photosensitive metal–organic
systems, in: C. Kutal, N. Serpone (Eds.), Advanced Chemistry,

114
C. Lambert et al. / Journal of Organometallic Chemistry 592 (1999) 109–114
Series 238, ACS, Washington, DC, 1993, pp. 411. (c) D. Astruc,
Acc. Chem. Res. 30 (1997) 383.
[4] (a) C. Lambert, S. Stadler, G. Bourhill, C. Bra¨uchle, Angew.
Chem. 108 (1996); Angew. Chem. Int. Ed. Engl. 35 (1996) 644.
(b) C. Lambert, E. Schma¨lzlin, K. Meerholz, C. Bra¨uchle, Chem.
Eur. J. 4 (1998) 512.
[5] M.T. Reetz, G. Lohmer, R. Schwickardi, Angew. Chem. 110
(1998) 492.
[6] A. Le Beuze, R. Lissilour, J. Weber, Organometallics 12 (1993)
47.
Blanchard-Desce, J.W. Perry, J. Skindhoj, Science 263 (1994)
511. (b) G. Bourhill, J.-L. Bredas, L.-T. Cheng, S.R. Marder, F.
Meyers, J.W. Perry, B.G. Tiemann, J. Am. Chem. Soc. 116
(1994) 2619. (c) M. Blanchard-Desce, V. Alain, P.V. Bedworth,
S.R. Marder, A. Fort, C. Runser, M. Barzoukas, S. Lebus, R.
Wortmann, Chem. Eur. J. 3 (1997) 1091.
[11] C.H. Warren, G. Wettermark, K. Weiss, J. Am. Chem. Soc. 93
(1971) 4658.
[12] I.D.L. Albert, T.J. Marks, M.A. Ratner, J. Phys. Chem. 100
(1996) 9714.
[7] For a general introduction to the hyper-Rayleigh scattering
technique, see (a) E. Hendrickx, K. Clays, A. Persoons, Acc.
Chem. Res. 31 (1998) 675. The experimental set-up used in the
present study is described in (b) R. Wortmann, C. Glania, P.
Kra¨mer, R. Matschiner, J.J. Wolff, S. Kraft, B. Treptow, E.
Barbu, D. La¨ngle, G. Go¨rlitz, Chem. Eur. J. 3 (1997) 1765.
[8] A. Willetts, J.E. Rice, D.M. Burland, D.P. Shelton, J. Chem.
Phys. 97 (1992) 7590.
[13] MOS-F V4, Fujitsu Ltd, Japan, 1998. For a general introduction
to quantum chemical calculations on NLO properties of organic
compounds, see D.R. Kanis, M.A. Ratner, T.J. Marks, Chem.
Rev. 94 (1994) 195.
[14] M. Kodomari, H. Satoh, S. Yoshitomi, Bull. Chem. Soc. Jpn. 61
(1988) 4149.
[15] I.U. Khand, P.L. Pauson, W.E. Watts, J. Chem. Soc. C (1968)
2257.
[16] A.E. Tschitschibabin, E.D. Ossetrowa, Chem. Ber. 58 (1925)
1708.
[9] L.-T. Cheng, W. Tam, S.R. Marder, A.E. Stiegman, G. Rikken,
C.W. Spengler, J. Phys. Chem. 95 (1991) 10643.
[10] (a) S.R. Marder, L.-T. Cheng, B.G. Tiemann, A.C. Friedli, M.
.
.