Oligo(phenylenevinylene)s with terminal donor-acceptor substitution

Article abstract of DOI:10.1002/1521-3773(20020118)41:2<292::AID-ANIE292>3.0.CO;2-Q

Hypsochromic or bathochromic? Depending upon the acceptor group (A = H, CN, CHO, NO₂) the position and intensity of the absorption maxima of a series of substituted oligo(phenylenevinylene)s change (see picture). Supported by semiempirical calc

Full text of DOI:10.1002/1521-3773(20020118)41:2<292::AID-ANIE292>3.0.CO;2-Q

COMMUNICATIONS
[9] M. Huber-Wunderlich, R. Glockshuber, Folding Des. 1998, 3, 161
171.
[10] a) X. Lu, H. F. Gilbert, J. W. Harper, Biochemistry 1992, 31, 4205
4210; b) U. Grauschopf, J. R. Winther, P. Korber, T. Zander, P.
Dallinger, J. C. A. Bardwell, Cell 1995, 83, 947 955.
[11] A. Jacobi, M. Huber-Wunderlich, J. Hennecke, R. Glockshuber, J.
Biol. Chem. 1997, 272, 213692 21699; b) J. Hennecke, C. Spleiss, R.
Glockshuber, J. Biol. Chem. 1997, 272, 189 195.
[12] a) F. Siedler, S. Rudolph-Bˆhner, D. Masamitsu, H.-J. Musiol, L.
Moroder, Biochemistry 1993, 32, 7488 7495; b) F. Siedler, D.
Quarzago, S. Rudolph-Bˆhner, L. Moroder, Biopolymers 1994, 34,
1563 1572.
refolding mixtures. The samples with the cis isomer were irradiated at
360 nm for 4 min every 2 h over 14 h and for a further 4 min after 21 h
and 28 h. The time course of correctly folded RNase A was
determined by measuring on aliquots of the refolding mixtures the
initial rates of enzymatic cCMP hydrolysis (432 mm in 0.1m MOPS
buffer, pH 7.0) at 284 nm and at a final RNase A concentration of
1.7 mm.
[21] M. E. O×Donnel, H. W. Williams, Jr, J. Biol. Chem. 1983, 258, 13795
13805.
[13] L. Moroder, D. Besse, H.-J. Musiol, S. Rudolph-Bˆhner, F. Siedler,
Biopolymers (Pept. Sci.) 1996, 40, 207 234.
[14] R. Behrendt, M. Schenk, H.-J. Musiol, L. Moroder, J. Pept. Sci. 1999, 5,
519 529.
[15] a) R. Behrendt, C. Renner, M. Schenk, F. Wang, J. Wachtveitl, D.
Oesterhelt, L. Moroder, Angew. Chem. 1999, 111, 2941 2943; Angew.
Chem. Int. Ed. 1999, 38, 2771 2774; b) C. Renner, J. Cramer, R.
Behrendt, L. Moroder, Biopolymers 2000, 54, 501 514; c) C. Renner,
R. Behrendt, S. Spˆrlein, J. Wachtveitl, L. Moroder, Biopolymers
2000, 54, 489 500.
[16] 1H NMR experiments were acquired on 3 mm samples in 9:1 (v:v)
H2O/D2O at 500 MHz (Bruker DRX500 spectrometer) as described in
reference [15b,c] with water suppression achieved by presaturation or
Oligo(phenylenevinylene)s with Terminal
Donor Acceptor Substitution**
Herbert Meier,* J¸rgen Gerold, Heinz Kolshorn,
Wolfram Baumann,* and Michael Bletz
Conjugated oligomers[1] such as the oligo(phenyleneviny-
lene)s (OPV)s have been extensively investigated for a
number of years because they exhibit properties of interest
to materials science in regard to application in nonlinear
optics (NLO) and as photoconductors and electrolumino-
phores. Typical of such a class of compounds is the con-
vergence of absorption and fluorescence with the increasing
number n of repeating units.[2] The determination of the
effective conjugation length (ECL)[1q] is important for the
characterization of the oligomers, as well as for their function
as model compounds for the corresponding polymers. A
simple algorithm, based upon exponential functions as natural
growth functions, has been demonstrated to be effective for
the determination of convergence and ECL in more than
20 series of conjugated compounds.[3]
We recently demonstrated that with a terminal push pull
substitution of conjugated oligomers the expected monotonic
bathochromic shifts of absorption and fluorescence with
increasing number n of repeating units need not necessarily
be present.[4] Since such series are very important, especially
for NLO materials, we have investigated the effect more
thoroughly. OPV systems were prepared which have solubi-
lizing bis(2-hexyloctyl)amino residues as the donor (D).
Different acceptors (A) were introduced at the other end of
the OPV chain (Scheme 1). The involvement of zwitterionic
resonance structures should have a decisive influence upon
¬
WATERGATE technique (V. Sklenar, M. Piotto, R. Leppik, V.
Saudek, J. Magn. Reson. A 1993, 102, 241 245). Conversion of NMR
data to geometrical constraints and structure calculations (100
structures) were performed as described in reference [15b,c].
[17] The linear peptide precursors were synthesized on chlorotrityl resin by
using the 9-fluorenylmethoxycarbonyl (Fmoc)/tBu chemistry as
described in reference [14]. Cyclization of the resulting side chain
protected peptides was performed in DMF at a concentration of
0.5 mm by PyBOP/HOBt (benzotriazolyl-1-oxy-tris(pyrrolidino)pho-
phonium hexafluorophosphate/1-hydroxybenzotriazole), and depro-
tection was carried out with trifluoroacetic acid/CH2Cl2 (95:5)
containing triethylsilane as scavenger. Oxidation of the cyclic
bis(cysteinyl)peptide was performed by air-oxygen in ammonium
acetate buffer (pH 8) at a concentration of 10À4 m. 1: ESI-MS: m/z:
‡
1113.6 [M‡H ]; Mr ˆ 1112.4 calcd for C49H72N14O12S2; amino acid
analysis (6m HCl, 1108C, 24 h): Asp 1.00 (1), Thr 0.90 (1), Ala 0.97(1),
Cys 1.46 (2), Lys 2.95 (3); peptide content: 71%. 2: ESI-MS: m/z:
‡
1085.8 [M‡H ]; Mr ˆ 1082.6 calcd for C49H74N14O14; amino acid
analysis (6m HCl, 1108C, 24 h): Asp 1.00 (1), Thr 0.99 (1), Ser 1.98 (2),
Ala 0.98 (1), Lys 2.90 (3); peptide content: 75%.
[18] For the determination of the Kox values, the trans isomer and mixtures
of the cis/trans isomers of
1 (0.1 mm) in argon-saturated 0.1m
phosphate buffer (pH 7) containing 0.1m NaCl and 1 mm EDTA were
equilibrated at 258C for 3 4 h under argon with 100-fold excess
glutathione at varying GSH/GSSG ratios. Aliquots of the equilibrated
redox mixtures were quenched with 1m phosphoric acid and analyzed
by HPLC. Baseline separation of all components was achieved and
these were identified as the two sets of the cis and trans isomer species
by ESI-MS and by comparison with the redox mixture obtained for
the pure trans isomer (see Figure 2). The concentrations of the
oxidized and reduced forms of 1 at equilibrium were obtained by
integration of the corresponding HPLC peak areas.
[19] J. Rost, S. Rapoport, Nature 1964, 201, 185.
[*] Prof. Dr. H. Meier, Dipl.-Chem. J. Gerold, Dipl.-Chem. H. Kolshorn
Institut f¸r Organische Chemie
[20] Fully reduced and denatured RNase A was prepared from bovine
pancreatic RNase A (Aldrich, approximately 100 Kunitz unitsmgÀ1
protein) and refolding assays were performed following essentially the
protocols previously described (Y. Konishi, T. Ooi, H. A. Sheraga,
Biochemistry 1982, 21, 4734 4740). Reoxidation of the reduced
RNase A (24 mm) in 0.1m Tris-HCl buffer, 1 mm EDTA, pH 7.4, was
carried out at 308C under argon in the glutathione redox buffer at an
RNase A/GSH/GSSG molar ratio of 1:20:4 and in the presence of 1 at
an RNase A/GSH/GSSG/1 molar ratio of 1:20:3:1 and 1:20:2:2.
Compound 1 was used as trans isomer (100% after thermal relaxation
at 508C overnight) and, upon irradiation at 360 nm, in the cis-azo
configuration, that is in the photostationary state (cis/trans ratio:
80:20 Æ 3). The reactivation reactions were initiated by addition of
RNase A to the preequilibrated (at least 2 h at room temperature)
Johannes Gutenberg-Universit‰t Mainz
Duesbergweg 10 14, 55099 Mainz (Germany)
Fax : (‡49)6131-39-25396
E-mail: hmeier@mail.uni-mainz.de
Prof. Dr. W. Baumann, Dipl.-Chem. M. Bletz
Institut f¸r Physikalische Chemie
Johannes Gutenberg-Universit‰t Mainz
Duesbergweg 10 14, 55099 Mainz (Germany)
Fax : (‡49)6131-39-22980
E-mail: wbaumann@mail.uni-mainz.de
[**] This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie.
292
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0292 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 2

COMMUNICATIONS
simple protecting-group technique.[5] Each new aldehyde
stage 3d 5d was then converted into the target compounds
4a c, 5a c, and 6a c with 2a c.
Figure 1 shows the long-wavelength maxima of the donor-
substituted series 3a 6a and the donor acceptor substituted
series 3b 6b, 3c 6c, and 3d 6d. With each additional n the
terminal acceptors lead to a marked bathochromic shift of the
push pull systems relative to the purely donor-substituted
Scheme 1. Donor acceptor substituted OPV systems
the electronic transitions; the solely donor substituted series
(A ˆ H) serves as a reference series.
The preparation of monodisperse, configurationally pure
oligomers was carried out by a convergent-synthesis strategy
outlined in Scheme 2. 4-Bis(2-hexyloctyl)aminobenzaldehyde
(1) was converted into the monostilbenes 3a c with the
phosphonates 2a c, and also used for the construction of the
∫styrylogous∫ aldehydes 3d 6d, the latter was by a repetitive
Wittig Horner reaction with the phosphonate 2d and a
Figure 1. Long-wavelength absorption maxima of the OPV series 3a 6a
~
!
&
*
(
), 3b 6b ( ), 3c 6c ( ), and 3d 6d ( ) in CHCl3 and fitting of the
Equation (1).[7] nÄ ˆ common convergence limit.
1
series 3a 6a. Most pronounced is the effect of the nitro
group, the strongest acceptor. In contrast to the OPV series
previously investigated,[2, 6] with terminal push pull substitu-
ents a charge transfer occurs during the electron transition.
The monotonic reduction in excitation energies (bathochro-
mic effect) with increasing n is in line with expectations in the
two series 3a 6a and 3b 6b; in the series 3d 6d nÄmax is
surprisingly almost independent of n, and in the series 3c 6c
the effect is even reversed: with 3c 6c an increase in
conjugation leads to a pronounced hypsochromic shift. With
increasing n all four series converge towards the convergence
limit E1 of the donor-substituted series 3a 6a.
How may the different behavior of the donor acceptor
substituted series be explained? It may be concluded from the
common convergence limit that with longer OPV chains the
acceptor group plays no role in the absorption. Therefore, we
have split the energy values of the electron transitions EDA(n)
into two components. The first term ED(n) corresponds to the
normal behavior of the purely donor-substituted series 3a
6a, that is, to the expected bathochromic shift with increase in
conjugation. The second term DEDA(n) is a correction term for
the superimposition of the intramolecular charge transfer
(ICT) in the push pull substituted series. Both terms can be
described by exponential functions[2] [Eq. (1)].
EDA(n)
ˆ
ED(n) À DEDA(n) ˆ E1 ‡ [ED(1) À E1]eÀa(nÀ1)
À [ED(1) À EDA(1)]eÀDa(nÀ1)
(1)
Table 1 gives an overview of the measured absorption
maxima and the parameters from the fitting for convergence
with increasing n. The charge-transfer correction term
DEDA(n) is always greatest at n ˆ 1. With increasing separa-
Scheme 2. Preparation of the OPV series
b) 1) KOC(CH3)3, THF; 2) HCl.
3 6. a) NaH, DME;
Angew. Chem. Int. Ed. 2002, 41, No. 2
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0293 $ 17.50+.50/0
293

COMMUNICATIONS
Table 1. Long-wave absorption maxima (nÄmax [cmÀ1], emax [cm2 mmolÀ1]) of the OPV series 3a 6a, 3b 6b, 3c 6c, and 3d 6d in CHCl3 and parameters
from the fitting of Equation (1).[a]
Series a
Series b
Series c
Series d
Compound
n
nÄmax
emax
nÄmax
emax
nÄmax
emax
nÄmax
emax
3
4
5
6
1
2
3
4
27248
24814
23866
23474
19483
29711
68944
24691
23529
23364
23256
21711
48107
63341
86164
21645
22321
22936
23148
28789
35425
58760
95860
23641
23256
23256
23256
20462
46822
67320
84882
Parameters
E1 [cmÀ1
]
23231 Æ 12
0.93 Æ 0.01
23231 Æ 12
25575603
23231 Æ 12
3607
23231 Æ 12
ED(1) À EDA(1) [cmÀ1
]
a
Da
0.93 Æ 0.01
0.93 Æ 0.01
0.93 Æ 0.01
0.87 Æ 0.02
0.76 Æ 0.04
0.86 Æ 0.03
[a] Curve fitting was carried out with the program Origin5.0.
tion of donor and acceptor it falls exponentially towards zero,
for example, from 5603 cmÀ1 for 3c, through 2493 and 930 to
326 cmÀ1 for 6c in the NO2 substituted series. This reduction
cannot be offset by the opposing effect which is a result of the
extension of conjugation and is represented by ED(n). The
nitro-substituted series consequently shows a hypsochromic
shift for increasing n. In contrast the cyano-substituted series
exhibits significantly smaller correction terms DEDA(n), which
moreover decrease more slowly with increasing n, from
2557cm À1 at n ˆ 1 to 218 cmÀ1 at n ˆ 4. In this case the
bathochromic effect on extending conjugation can overcom-
pensate the decreasing charge-transfer term and in the
summation the expected bathochromic shift is observed. In
the series 3d 6d the conjugation effect and the charge-
transfer effect offset each other almost completely so that nÄmax
is almost independent of n.
How may this theory, on the basis of empirical fit functions,
be supported quantum mechanically? We have calculated the
frontier orbitals of 3a c, 4a c, 5a c, and 6a c by AM1 and
INDO/S methods.[8, 9]. A selection of these is reproduced in
Figure 2. The long-wavelength electron transitions of the
OPV systems are essentially HOMO !LUMO transitions for
n ˆ 1. The HOMO !LUMO energy difference decreases for
n ˆ 1 (and also for each higher n) in the sequence a > b > c.
Since the exchange integral for the electron correlation also
decreases in this sequence because of the decreasing tran-
sition density, a bathochromic shift in the absorption for every
n in the change from A ˆ H, through A ˆ CN, to A ˆ NO2
results. Within the series a, b, and c the HOMO !LUMO
energy difference decreases with increasing n. However, the
electron correlation leads then to smaller Coulomb integrals
for the repulsion with increasing size of the chromophore, and
therefore, to an opposing tendency. Thus, in principle, with
increasing n in the oligomer series either a bathochromic or a
hypsochromic shift of the absorption can occur.
With increasing n the fraction of HOMO !LUMO tran-
sitions in the long-wavelength absorption falls drastically;
above all the electron transitions NHOMO !LUMO, HO-
MO !NLUMO, and NHOMO !NLUMO intermix (where
N ˆ next-to-). However, as Figure 2 shows, the HOMO !LU-
MO fraction is essentially determinative for charge transfer. If
the ICT correction term DEDA(n) is now plotted against the
involvement of the HOMO !LUMO transition calculated by
the INDO/S method a simple correlation is recognized: the
greater DEDA(n) for the charge transfer is, the greater the
HOMO !LUMO contribution to the long-wavelength elec-
tron transfer is (Figure 3). Thus semi-empirical quantum
mechanics reproduces correctly the observed trend: with
increasing n the HOMO !LUMO contribution decreases,
and with it the DEDA ; this effect dominates in the nitro-
substituted series 3c 6c and hence leads to the hypsochromic
shift–in contrast in the cyano-substituted series 3b 6b the
term ED(n), which expresses the extension of the conjugation,
dominates such that a bathochromic shift occurs.
Since the nitro-substituted series clearly plays a special role
we carried out electro-optical absorption measurements
(EOAM) on the compounds 3c 5c; a number of the results
of these measurements are summarized in Table 2. Assuming
that polarizabilities may be neglected with respect to dipole
moments, and m(S0), m(S1), and the transition moment m01 are
parallel, EOAM[10, 11] gives m0 as well as the terms m0(m1 À m0)
and (m1 À m0)2, which are obtained as different regression
coefficients from the first and second derivation of e(nÄ)
according to the wavenumber nÄ. Both statistically as well as by
the EOAM model the first term m0(m1 À m0) leads to the more
reliable values.
Figure 2. Frontier orbitals of compounds 3b and 3c, as well as 6b and 6c
calculated by the INDO/S method. The arrows illustrate the shift of the
electron densities during the HOMO !LUMO transition.
As expected, the intensity I, the oscillator strength f, and the
transition moment m01 increase with increasing n. Surprisingly
294
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0294 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 2

COMMUNICATIONS
Table 2. Optical and electro-optical data of 3c, 4c, and 5c.[a]
[b]
Compound
n
I [1010 cmmolÀ1
]
f
m01 [10À30 Cm]
m0(S0) [10À30 Cm][c]
m1(S1) [10À30 Cm][c,d]
3c
4c
5c
1
2
3
15.48
19.70
31.79
„
0.68
0.86
1.40
27.23
30.30
38.01
29.8
26.4
25.3
112.0
128.8
108.2
[a] Measurements in dioxan. [b] I ˆ e(nÄ)dnÄ. [c] The statistical regression error is 1 5% for m0 and 2 6% for m1. [d] Determined from de/dn; d2e/dn2 gives the
less reliable values 106.6, 128.5, and 151.3 Cm for 1030 m1 with 3c 5c.
h) H. S. Nalwa in Nonlinear Optics of Organic Molecules and
Polymers (Eds.: H. S. Nalwa, S. Miyata), CRC, New York, 1997,
p. 611 797; i) S. R. Marder, B. Kippelen, A. K.-Y. Jen, N. Peygham-
barian, Nature 1997, 388, 845 851; j) J. M. Tour, Chem. Rev. 1996, 96,
537 553; J. M. Tour, Acc. Chem. Res. 2000, 33, 791 804; k) J. L.
Bredas, Adv. Mater. 1995, 7, 263 274; l) P. B‰uerle, Adv. Mater. 1993,
5, 879 886; m) W. R. Salaneck, I. Lundstrˆm, B. Ranby, Conjugated
Polymers and Related Materials, Oxford University Press, Oxford,
1993; n) K. M¸llen, Pure Appl. Chem. 1993, 65, 89 96; o) H. Meier,
Angew. Chem. 1992, 104, 1425 1446; Angew. Chem. Int. Ed. Engl.
¬
1992, 31, 1399 1420; p) J. L. Bredas, R. Silbey, Conjugated Polymers,
Kluwer, Dordrecht, 1991; q) H.-H. Hˆrhold, M. Helbig, D. Raabe, J.
Opfermann, U. Scherf, R. Stockmann, D. Wei˚, Z. Chem. 1987, 27,
126 137.
[2] H. Meier, U. Stalmach, H. Kolshorn, Acta Polym. 1997, 48, 379 384,
and references therein.
[3] See refs. [1d, 2] and H. Meier, H. Aust, D. Ickenroth, H. Kolshorn, J.
Prakt. Chem. 1999, 341, 529 535.
[4] H. Meier, R. Petermann, J. Gerold, Chem. Commun. 1999, 977 978.
[5] The Wittig Horner reaction is sufficiently stereoselective in the
stilbenoid series; small cis fractions (ꢀ5%) may be separated by
column chromatography or by recrystallization. Compounds 3a and
3b are oily, compounds 4a, 5c, 6a, and 6c are highly viscous to waxy;
the other oligomers have the following melting points: 3b: 608C, 3c:
628C, 4b: 1038C, 4c: 948C, 4d: 988C, 5a: 1588C, 5b: 1628C, 5d:
1658C, 6b: >2408C, 6d: 1958C. The 1H and 13C NMR and mass
spectrometric data will be reported elsewhere.
Figure 3. Charge-transfer terms DEDA in relationship to the HOMO !LUMO
fraction calculated by the INDO/S method for the long-wavelength electron
transitions in the OPV series 3b 6b (A ˆ CN) and 3c 6c (A ˆ NO2).
the experimental dipole moment m0 of the ground state S0
decreases with the increasing distance between donor and
acceptor. The dipole moments m1 of the electronically excited
state S1 are significantly higher. Clearly, a simple relationship
with the benzenoid and quinoid (zwitterionic) resonance
structures formulated in Scheme 1 does not exist. The
calculation of resonance parameters known in the literature
[6] R. Schenk, H. Gregorius, K. M¸llen, Adv. Mater. 1991, 3, 492 493.
[7] Initial measurements for n ˆ 5 confirm the reliability of the function
used. For A ˆ NO2 the function gives nÄmax ˆ 23151 cmÀ1, measured in
[12, 13]
with the values for m0(S0), m1S1, and m01 given in Table 2
CHCl3 23148 cmÀ1
.
shows that for the compounds 3c 5c the zwitterionic
structure–independent of n[14]–contributes modestly to the
ground state at about 10%. The charge transfer on electronic
excitation indeed does increase the dipole moment consid-
erably (Dm ˆ 92.3(Æ10.1)  10À30 Cm), it has, however, an
increasingly smaller influence on the long-wavelength ab-
sorption with increasing n since with increasing distance
between donor D and acceptor A at almost constant Dm,
increasingly less negative charge is transferred to the acceptor
part of the molecules. For the conjugated systems with large
[8] Dimethylamino groups were used in the calculations in place of the
branched dialkylamino groups. Differences in the frontier orbitals
from the AM1 and INDO/S calculations were insignificant. For the
INDO/S method see W. P. Anderson, W. D. Edwards, M. C. Zerner,
Inorg. Chem. 1986, 25, 2728 2732. For the calculation of 3c see M.
Dekhtyar, W. Rettig, Phys. Chem. Chem. Phys. 2001, 3, 1602 1610.
[9] According to measurements and calculation the oscillator strengths f
increases with increasing n for all series. However, as with the absolute
values of lmax, the absolute values of f can not be calculated
satisfactorily by semi-empirical quantum mechanical methods.
[10] W. Baumann in Physical Methods of Chemistry, Vol. 3B (Eds.: B. W.
Rossiter, J. F. Hamilton), Wiley, New York, 1989.
[11] W. Liptay in Excited States, Vol. 1 (Ed.: E. C. Lim), Academic Press,
New York, 1974.
[12] S. Beckmann, K.-H. Etzbach, P. Kr‰mer, K. Lukaszuk, R. Matschiner,
A. J. Schmidt, R. Wortmann, F. W¸rthner, Adv. Mater. 1999, 11, 536
541.
D
A distances described here the frequently used model of a
dominating electrically neutral and a less involved terminal
zwitterionic resonance structure is usable for the ground state
S0. The electronic excited state S1 does have a significantly
higher dipole moment, but the participation of resonance
structures never reverses. The simple VB model is not
consistent with this.[15]
[13] M. Blanchard-Desce, M. Barzoukas, J. Opt. Soc. Am. B 1998, 15, 302
307.
[14] With short push pull substituted polyenes it was found that the
participation of zwitterionic forms falls with increasing n.[13]
[15] The frequently used criterion for the convergence of bond lengths is
also not suitable for the series investigated here. Neither semi-
empirical quantum mechanics nor the experimental investigation
show a change in bond lengths within a series. The experimental
investigation was carried out with the coupling constants 3J of the
olefinic protons which are of the order of 16.2(Æ0.2) Hz.
Received: July 24, 2001 [Z17579]
[1] a) J. Roncali, Acc. Chem. Res. 2000, 33, 147 156; b) U. H. F. Bunz,
Chem. Res. 2000, 100, 1605 1644; c) J. M. Tour, Acc. Chem. Res. 2000,
33, 791 804; d) R. E. Martin, F. Diederich, Angew. Chem. 1999, 111,
1440 1469; Angew. Chem. Int. Ed. 1999, 38, 1350 1377; e) U. Scherf,
Top. Curr. Chem. 1999, 201, 163 222; f) K. M¸llen, G. Wegner,
Electronic Materials: The Oligomer Approach, Wiley-VCH, Wein-
heim, 1998; g) A. Kraft, A. C. Grimsdale, A. B. Holmes, Angew.
Chem. 1998, 110, 416 443; Angew. Chem. Int. Ed. 1998, 37, 403 428;
Angew. Chem. Int. Ed. 2002, 41, No. 2
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4102-0295 $ 17.50+.50/0
295