2,5-Di(2-pyridyl)phospholes: Model compounds for the engineering of π-conjugated donor-acceptor co-oligomers with a chemically tunable HOMO-LUMO gap

Article abstract of DOI:10.1039/a808408d

2,5-Di(2-pyridyl)phospholes possess an extended π-conjugated system with a charge transfer structure; high yielding chemical modifications involving the phosphorus atom allow fine tuning of the HOMO-LUMO gap.

Full text of DOI:10.1039/a808408d

2,5-Di(2-pyridyl)phospholes: model compounds for the engineering of
p-conjugated donor–acceptor co-oligomers with a chemically tunable
HOMO–LUMO gap
Caroline Hay,a Delphine Le Vilain,a Vale´rie Deborde,a Lo¨ıc Toupetb and Re´gis Re´au*a
a UMR 6509 CNRS-Universite´ de Rennes, Campus de Beaulieu, 35042 Rennes Cedex, France.
E-mail: regis.reau@univ-rennes1.fr
b UMR 6626 CNRS-Universite´ de Rennes, GMCM, Campus de Beaulieu, 35042 Rennes Cedex, France
Received (in Liverpool, UK) 28th October 1998, Accepted 23rd December 1998
2,5-Di(2-pyridyl)phospholes possess an extended p-conju-
gated system with a charge transfer structure; high yielding
chemical modifications involving the phosphorus atom
allow fine tuning of the HOMO–LUMO gap.
The control of both the structure and HOMO–LUMO gap of
linear p-conjugated oligomers is the focus of considerable
interest. These compounds have been investigated in order to
simulate the electronic and electrochemical properties of the
corresponding polymers and they are now emerging as efficient
molecular wires in electronics applications.1 Group 15 and 16
heterocyclopentadienes have been widely used as building
blocks for the design of well-defined linear p-conjugated
oligomers.1 However, there are only a few instances of
Scheme 1
phospholes having been used for such a purpose,2 although they
display particular properties which make them attractive
and 1b (94%). These compounds reacted at 278 °C with
synthons. Firstly, the aromaticity of phospholes is generally
much lower than that of furan, pyrrole or thiophene.3 This is a
consequence of the inherent pyramidal geometry of the
tricoordinate phosphorus atom which disrupts efficient orbital
interaction within the five-membered ring. This property should
have an impact on the molecular engineering of the HOMO–
LUMO separation of linear p-conjugated systems since p-
electron delocalization along the chain is favoured when the
backbone features heterocycles of low resonance energy.1a,3c
Secondly, and in contrast to thiophene or pyrrole, phospholes
possess a heteroatom which retains a versatile reactivity.3a,b
This may offer the possibility of tuning the HOMO–LUMO gap
via chemical modifications. Recently, Mathey et al. prepared
various a-oligophospholes (n = 2–4) and oligomers containing
a-thienyl- or a-furyl-a-phospholyl links.2a–e X-Ray diffraction
studies performed on these derivatives showed that they deviate
notably from planarity. This rotational disorder prevents
extended p-conjugation because the orbital overlap varies
approximately with the cosine of the twist angle. A recent
strategy to effect a coplanar arrangement in linear p-conjugated
polymers involved constructing alternating [A-B] derivatives
where the A and B units are electron-deficient and electron-rich,
respectively.4 These derivatives are considered to have a
charge-transfer structure with double bond character between
the A-B units. As the parent phosphole can be considered as a
very electron-rich heterocyclopentadiene,3c we decided to
investigate the preparation of derivatives based on alternating
pyridine and phosphole rings with a well defined 2,5-linkage.
Herein, we describe an efficient synthesis of the simplest model,
namely a 2,5-di(2-pyridyl)phosphole, and we show the possibil-
ity of tuning its HOMO–LUMO gap through chemical mod-
ifications.
‘zirconocene’6 via a regioselective ring-closing diyne coupling
to give the corresponding zirconacyclopentadienes which were
characterised by NMR spectroscopy. Isolation of the zirconium-
containing metallacycles was not necessary; addition of 1.1
equiv. of PhPCl2 to the reaction mixture at 278 °C afforded the
desired phospholes 2a,b in a ‘one-pot’ procedure. They were
isolated as air-stable yellow powders after purification by
chromatography (alumina) in 70 and 73% yield, respectively.
Phospholes 2a,b have been characterised by high resolution
mass spectrometry and elemental analyses and they exhibit the
expected NMR spectroscopic data. In the 13C NMR spectrum,
the endocyclic P–Ca carbon atoms of 2a give a singlet, which is
not unusual for phospholes,2e,3a,b while the other endocyclic
carbon and the Cipso atoms of the aromatic rings of compounds
2a,b all appear as doublets with classical JPC coupling
constants.
The crystal structure of derivative 2b (Fig. 1) reveals that, as
expected, the phosphorus atom is strongly pyramidalized
[S(CPC angles) = 299.3°] and that the endocyclic P–C bond
lengths [1.801(4) and 1.804(4) Å] approach that of a P–C single
bond (1.84 Å). These data compare well with those observed for
a 2,5-dialkynylphosphole2c [S(CPC angles) = 292.4°; P–C,
The target phospholes were prepared according to Fagan’s
method5 which involves the reaction of zirconacyclopenta-
dienes with dihalogenophosphines. In order to obtain the
desired 2,5-substitution pattern, diynes 1a,b possessing a
flexible spacer were employed (Scheme 1).6 A Sonogashira
coupling of hexa-1,7-diyne with 2-bromopyridine allowed an
efficient and large-scale preparation of derivatives 1a (92%)
Fig. 1 ORTEP view of 2b. Selected bond lengths (Å): P–C(1) 1.806(6),
C(1)–C(2) 1.354(8), C(2)–C(7) 1.478(9), C(7)–C(8) 1.356(9), C(8)–P
1.806(6).
Chem. Commun., 1999, 345–346
345

1.815–1.821 Å] and 1,2,5-triphenylphosphole7 [P–C, 1.822 Å],
two compounds for which it is proposed that ring delocalization
is almost switched off, with the endocyclic subunit being
conjugated with the C2 and C5 substituents.3a,b Of particular
importance, the two bond lengths between the pyridyl rings and
the slightly puckered phosphole ring [1.464(5) and 1.467(5) Å]
have values that lie between those associated with single and
double bonds, and the dihedral angles (25.6 and 7.0°) are
relatively small, allowing an extended p-conjugation. These
structural data clearly show that in 1-phenyl-2,5-di(2-pyr-
idyl)phospholes the dienic moiety of the phosphole ring is
conjugated with the two pyridine rings and not with the
phosphorus atom.
The UV–VIS data for compounds 2a,b and of the related
2,5-diphenylphosphole 2c are given in Table 1. Derivative 2c
was isolated in only 10% yield according to the strategy
presented in Scheme 1 starting from 1,8-diphenylocta-1,7-di-
yne.6b Interestingly, the lmax observed for 2,5-pyridylphos-
pholes 2a,b are comparable, and notably longer to that of the
2,5-diphenylphosphole 2c (Dlmax > 36 nm). This observation,
which is in line with the conclusion drawn from the solid state
structure, highlights the importance of the alternating electron-
deficient/electron-rich ring structure of derivatives 2a,b which
favours the delocalization of the p-system.
longer than those of 2b [1.801(4) and 1.804(4) Å]. The twist
angles between the pyridyl and the phosphole rings (11.3 and
0.0°) are smaller than those observed for the non-coordinated
phosphole 2b (25.6 and 7.0°). Since the pyridine–phosphole–
pyridine moiety of 2b is more planar when coordinated, the
origin of the blue shift observed on going from 2b to 4 is likely
to be due to electronic factors. It seems reasonable to propose
that both oxidation with sulfur and the coordination of the
phosphorus atom decreases the electron-density of the phos-
phole ring, leading to a decrease in the degree of charge transfer
in the p-linear conjugated pyridine–phosphole–pyridine sys-
tem. As expected, this effect is more pronounced for sulfide 3
than for complex 4.
Derivatives 2a,b and 4 are air-stable in the solid state and also
in THF solution for days. Preliminary stability tests by
thermogravimetric analysis and differential scanning calor-
imetry under nitrogen show that derivative 2a (mp 186 °C) and
2b (mp 192 °C) are stable up to 201 and 211 °C, respectively,
whereas complex 4 (mp > 204 °C) decomposes at 204 °C.
The synthesis of alternating pyridine–phosphole co-oligo-
mers with a well-defined 2,5-linkage is under active in-
vestigation.
We thank the Conseil Re´gional Bretagne for financial support
of this work.
The next question was whether chemical transformation
involving the phosphorus atom, which is not involved in the p-
delocalized system, would significantly modify the HOMO–
LUMO gap? Treatment of phosphole 2b with elemental sulfur
and W(CO)5(thf) afforded derivatives 3 and 4, respectively, in
near quantitative yields. The molecular formulae of 3 and 4
were established by multinuclear NMR spectroscopy, high
resolution mass spectrometry and elemental analyses. These
chemical transformations resulted in a blue shift of lmax relative
to phosphole 2b (Table 1), the effect being more pronounced for
the sulfide 3 (Dlmax = 26 nm) than for the complex 4 (Dlmax
= 17 nm). In order to evaluate the relative contribution of
geometric effects (degree of planarity) on values of lmax in this
series, derivative 4 was subjected to an X-ray diffraction study
(Fig. 2). The structural data for complex 4† are very similar to
those observed for the free phosphole 2b, suggesting that the
coordination of the metal did not result in a dramatic steric
perturbation. As observed for the free ligand, the three
heteroatoms are pointing in the same direction. Note that the
sum of the CPC angles is 298.7° (2b, 299.3°) and that the two
endocyclic P–C bonds [1.818(4) and 1.824(4) Å] are slightly
Notes and references
† Crystal data for 2b and 4: Samples were studied on a CAD4 NONIUS
diffractometer with graphite monochromatized Mo-Ka (l = 0.71073 Å) at
293(1) K. The whole structures were refined with SHELXL97. ORTEP
views with 50% probability were realized with PLATON98.
Crystal data for 2b: C24H21N2P, M = 368.42; crystal size 0.32 3 0.22 3
0.18 mm, monoclinic, space group P21/n, a = 8.619(2), b = 14.127(2), c
= 16.116(4) Å, b = 104.53(6)°, U = 1899.5(7) Å23, Z = 4, Dc = 1.288
g cm23, m = 1.55 cm21, F(000) = 776, 4398 reflections measured, 4128
were independent [1643 with I > 2s(I)], 245 variables refined, R1 = 0.0654
(wR2 = 0.1221).
Crystal data for 4: C29H21N2O5PW, M = 692.33; crystal size 0.40 3
0.22 3 0.12 mm, triclinic, space group P1; a = 9.943(2), b = 11.122(9), c
= 13.590(5) Å, a = 78.34(3), b = 72.90(2), g = 66.90(3)°, U = 1315(1)
Å23, Z = 2, Dc = 1.749 g cm23, m = 44.96 cm21, F(000) = 676, 5715
reflections measured, 5377 were independent [4764 with I > 2s(I)], 343
variables refined, R1 = 0.0278 (wR2 = 0.0686).
¯
CCDC 182/1131. Crystallographic data is available in CIF format from
the RSC Web site, see: http://www.rsc.org/suppdata/cc/1999/345/
1 (a) J. Roncali, Chem. Rev., 1997, 97, 173; (b) J. M. Tour, Chem. Rev.,
1996, 96, 537; (c) A. Kraft, A. Gimsdale and A. Holmes, Angew. Chem.,
Int. Ed., 1998, 37, 402; (d) I. Jestin, P. Fre`re, P. Blanchard and J. Roncali,
Angew. Chem., Int. Ed., 1998, 37, 942.
Table 1 UV–VIS data in THF for compounds 2a–c, 3 and 4
Compound
lmax/nm
log e
2 (a) E. Deschamps, L. Ricard and F. Mathey, Angew. Chem., Int. Ed.
Engl., 1994, 33, 1158; (b) M.-O. Bevierre, F. Mercier, L. Ricard and F.
Mathey, Angew. Chem., Int. Ed. Engl., 1990, 29, 655; (c) S. Holand, F.
Gandolfo, L. Ricard and F. Mathey, Bull. Soc. Chim. Fr., 1996, 133, 33;
(d) E. Deschamps and F. Mathey, J. Org. Chem., 1990, 55, 2494; (e) E.
Deschamps, L. Ricard and F. Mathey, Heteroatom Chem., 1991, 3, 377;
(f) S. H. Mao and T. Don Tilley, Macromolecules, 1997, 30, 5566.
3 (a) F. Mathey, Chem. Rev., 1988, 88, 437; (b) L. D. Quin, in
Comprehensive Heterocyclic Chemistry, ed. A. R. Katritzky, Pergamon,
Oxford, 1996, p. 757; (c) I. Albert, T. Marks and M. Ratner, J. Am. Chem.
Soc., 1997, 119, 6575; (d) A. Dransfeld, L. Nyulaszi and P. v. R.
Schleyer, Inorg. Chem., 1998, 37, 4413.
2a
2b
2c
3
390
395
354
364
373
3.96
4.02
4.20
3.28
4.01
4
4 (a) For recent contributions see: Q. T. Zhang and J. M. Tour, J. Am.
Chem. Soc., 1998, 120, 5355; F. Demanze, A. Yasser and F. Garnier,
Macromolecules, 1996, 29, 4267; (b) S. Yamaguchi, Y. Itami and K.
Tamao, Organometallics, 1998, 17, 4910.
5 P. J. Fagan, W. A. Nugent and J. C. Calabresse, J. Am. Chem. Soc., 1994,
116, 1880.
6 (a) E. Negishi and T. Takahashi, Acc. Chem. Res., 1994, 27, 124; (b) B.
Lucht, S. Mao and T. Don Tilley, J. Am. Chem. Soc., 1998, 120, 4354;
(c) B. Lucht and T. Don Tilley, Chem. Commun., 1998, 1645.
7 W. P. Ozbirn, R. A. Jacobson and J. C. Clardy, J. Chem. Soc., Chem.
Commun., 1971, 1062.
Fig. 2 ORTEP view of 4. CO ligands have been omitted for clarity. Selected
bond lengths (Å): P–C(1) 1.820(5), C(1)–C(2) 1.356(5), C(2)–C(7)
1.466(8), C(7)–C(8) 1.352(5), C(8)–P 1.819(4).
Communications 8/08408D
346
Chem. Commun., 1999, 345–346