J. Vicente et al. / Journal of Organometallic Chemistry 648 (2002) 62–71
65
coordinate the two nitrogen atoms of the ligand with
varying results. Thus, N(1)-coordination of the quinox-
aline ligand in complex [PdCl(C,N4-CH2C8H5N2-
5)(PPh3)] (5) did not take place when this was reacted
with excess (1:1.3 or 1:2) or an equimolecular amount
of [PdCl2(NCPh)2] (dichloromethane, room tempera-
ture, 3–5 h). Instead, we obtained the poorly soluble
chloride-bridged dimer [Pd(m-Cl)(C,N4-CH2C8H5N2-
5)]2 (3) together with trans-[PdCl(m-Cl)(PPh3)]2 (Scheme
3), as confirmed by spectroscopic data and the X-ray
study of single crystals of the latter which grew from
the mixture.4 Although this compound was reported
some time ago [34], its X-ray crystal structure has not
been described previously. Cyclopalladated complexes
analogous to 5 are known to undergo solvolysis in
AcOH at 80 °C in the presence of LiCl to afford the
corresponding non-palladated ligand and [PdCl(m-
Cl)(PPh3)]2 [35]. Formation of complex 3 was also
detected, by 1H-NMR spectroscopy, when 6, 7 or 8
were reacted with excess [PdCl2(NCPh)2] [1:1.3 (6), 1:2
(7, 8)].
indicating the ionically bonded perchlorate. Also, the
PdꢀCl band that appears at 270 cm−1 in 6 was not
present in the reaction product. The above data, to-
gether with the insolubility of the product in a variety
of organic solvents such as acetone, chloroform,
dichloromethane, toluene or nitromethane, suggests
that it consists of multimetallic species of general for-
mula [Pd(C,N4,N1-CH2C8H5N2-5)(PR3)]n(ClO4)n. Un-
fortunately, its poor solubility also prevented further
purification and an analytically pure sample was not
obtained. The solid was moderately soluble in DMSO
1
and its H-NMR spectrum could be obtained, which
indicated that N(1)-coordination does not persist in
solution, since the expected downfield shift for the
aromatic hydrogens was not observed (see below). In
fact, both its 1H- and 31P-NMR spectra resembled
those of the starting complex 6, therefore suggesting the
formation of a monomeric compound of related nature,
i.e.
a
solvento complex [Pd(C,N4-C8H5N2CH2-
5)(DMSO)(PMe3)](ClO4). However, after precipitation
of the compound from a DMSO solution, no significant
amounts of sulfur were found in the elemental analysis,
indicating that coordination of DMSO does not occur
in the solid material. Other related method of prepara-
tion, e.g. reaction of 5 with Tl(CF3SO3) or AgX (X=
ClO4, CF3SO3) in acetone, gave comparable results.
It has been reported that complexes containing 2-(2%-
pyridyl)quinoxaline (L) as a bidentate chelating ligand,
[PtX2L] (X=Cl, Br), undergo thermal decomposition
to insoluble polymers (PtCl2L0.5–PtBr2L0.25), where L
acts as a tridentate bridging ligand [13]. Vibrational
spectroscopic studies have been used to prove the poly-
meric nature of these compounds. For example while a
single sharp band at 965–970 cm−1 is present in the IR
spectrum of the starting materials, this splits into a
triplet when both quinoxaline nitrogens participate in
coordination. Although thermogravimetric evidence of
formation of analogous Pd polymers was found, they
were too unstable to be isolated. In the IR spectra of
our complexes, only very weak ligand bands are found
in the region 940–980 cm−1, which do not allow clear
assignments. However, in one case (the product of the
reaction of 5 with AgClO4), a weak but distinct triplet
at 950 cm−1, corresponding closely to the above data,
was observed in place of a singlet in the starting
compound (5).
However, when the reaction between
5
and
[PdCl2(NCPh)2] was carried out in a 2:1 molar ratio,
the bis-C,N4,N1-coordinated tripalladium derivative
[PdCl2{PdCl(C,N4,N1-CH2C8H5N2-5)(PPh3)}2]
(10)
was obtained in quantitative yield (Scheme 3). Al-
though a number of triangular tripalladium complexes
with quinoline-based ligands have been described [19],
to our knowledge, there are no linear tripalladium
derivatives containing heterocyclic ligands. The only
fully characterized dipalladated quinoxaline-based lig-
and reported [12] coordinates two Pd atoms via double
cyclometallation.
We did not succeed in the synthesis of mixed-metal
Pt(II)ꢀPd(II) derivatives. For example, complex 5 did
not react with [PtCl2L2] [L=NCPh (2:1)], SMe2 (2:1)
(room temperature, acetone, 16 h) or [PtCl2(COD)]
[COD=1,5-cyclooctadiene (2:1), room temperature,
dichloromethane, 22 h].
In order to prepare polymeric species of the type
[Pd(C,N4,N1-CH2C8H5N2-5)(PR3)]nXn, we reacted 6
with AgClO4 in acetone. Immediate precipitation of a
pale yellow solid was observed. Upon evaporation of
the solvent in vacuum, removal of AgCl formed in the
reaction was possible by extracting the residue with
DMSO. Filtration and concentration of the extracts
gave a yellow solid, whose infrared spectrum in Nujol
emulsion showed two bands at ca. 1100 and 620 cm−1
,
2.3. Structures of complexes
1H- and 31P-NMR data for the isolated complexes
are presented in Table 1. In the H-NMR spectra, the
set of signals corresponding to H6, H7 and H8, consists
of one apparent triplet (H7) and two apparent doublets,
occurring within the range 6.81–10 ppm. However,
when the ligand is bound through N(1) (complexes 1
4 X-ray crystallographic data (excluding structure factors) for
trans-[PdCl(m-Cl)(PPh3)]2 have been deposited with the Cambridge
Crystallographic Data Centre as a Private Communication, J. Vi-
cente, M.C. Lagunas, E. Bleuel and M. Ram´ırez de Arellano, CCDC-
100877, 1997. Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
(Fax: +44-1223-336033 or e-mail: deposit@ccdc.cam.ac.uk).
1