The isolation and purification of tris-2,2′-bipyridine complexes of ruthenium(II) containing unsymmetrical ligands

Article abstract of DOI:10.1039/b104365j

Monomeric ruthenium(II) complexes [Ru(L)₃]²⁺ containing unsymmetric bipyridine ligands [Where L = 5-methyl-2,2′-bipyridine (L¹), 5-ethyl-2,2′-bipyridine (L²), 5-propyl-2,2′-bipyridine (L³), 5-(2-methylpropyl)-2,2′-bipyridine (L⁴), 5-(2,2-dimethylpropyl)-2,2′-bipyridine (L⁵) or 5-(carbomethoxy)-2,2′-bipyridine (L⁶)] have been studied and the meridional and facial isomers isolated by the use of cation-exchange column chromatography (SP Sephadex C-25) eluting with either sodium toluene-4-sulfonate or sodium hexanoate. The relative yield of the facial isomer was found to decrease with increasing steric bulk, preventing the isolation of fac-[Ru(L⁵)₃]²⁺. The two isomeric forms were characterized by ¹H NMR spectroscopy, with the complexes [Ru(L^1-3)₃]²⁺ demonstrating an unusually large coupling between the H⁶ and H⁴ protons. Crystals suitable for X-ray structural analysis of [Ru(L¹)₃]²⁺ were obtained as a mixture of the meridional and facial isomers, indicating that separation of this isomeric mixture could not be achieved by fractional crystallisation. The optical isomers of the complex [Ru(L³)₃]²⁺ were chromatographically separated on SP Sephadex C-25 relying upon the inherent chirality of the support. It is apparent that chiral interactions can inhibit geometric isomer separation using this technique.

Full text of DOI:10.1039/b104365j

View Online / Journal Homepage / Table of Contents for this issue
The isolation and purification of tris-2,2Ј-bipyridine complexes of
ruthenium(II) containing unsymmetrical ligands
Nicholas C. Fletcher,* Mark Nieuwenhuyzen and Sharon Rainey
School of Chemistry, The Queen’s University of Belfast, David Keir Building, Belfast,
Northern Ireland, UK BT9 5AG. E-mail: n.fletcher@qub.ac.uk
Received 17th May 2001, Accepted 11th July 2001
First published as an Advance Article on the web 24th August 2001
2
ϩ
Monomeric ruthenium() complexes [Ru(L) ] containing unsymmetric bipyridine ligands [Where L = 5-methyl-
3
1
2
3
2
(
,2Ј-bipyridine (L ), 5-ethyl-2,2Ј-bipyridine (L ), 5-propyl-2,2Ј-bipyridine (L ), 5-(2-methylpropyl)-2,2Ј-bipyridine
4
5
6
L ), 5-(2,2-dimethylpropyl)-2,2Ј-bipyridine (L ) or 5-(carbomethoxy)-2,2Ј-bipyridine (L )] have been studied and the
meridional and facial isomers isolated by the use of cation-exchange column chromatography (SP Sephadex C-25)
eluting with either sodium toluene-4-sulfonate or sodium hexanoate. The relative yield of the facial isomer was found
5
2ϩ
to decrease with increasing steric bulk, preventing the isolation of fac-[Ru(L ) ] . The two isomeric forms were
3
1
1–3
2ϩ
characterized by H NMR spectroscopy, with the complexes [Ru(L ) ] demonstrating an unusually large coupling
3
6
4
1
2ϩ
between the H and H protons. Crystals suitable for X-ray structural analysis of [Ru(L ) ] were obtained as a
3
mixture of the meridional and facial isomers, indicating that separation of this isomeric mixture could not be
3
2ϩ
achieved by fractional crystallisation. The optical isomers of the complex [Ru(L ) ] were chromatographically
3
separated on SP Sephadex C-25 relying upon the inherent chirality of the support. It is apparent that chiral
interactions can inhibit geometric isomer separation using this technique.
Introduction
The development of polynuclear coordination species has been
the subject of extensive research due to the large number of
1
potential applications. Of particular interest have been metallo-
supramolecular assemblies of the polypyridine complexes of
ruthenium(), osmium() and rhodium() due to their poten-
2–4
tial as photoactivated charge separated species. As a con-
sequence, there is a vast wealth of literature describing the
7,8
5
6
fascinating synthetic, photophysical and electrochemical
properties of simple monomeric systems. By the judicious
choice of ligands and metal centres the electronic potentials can
be controlled to direct electrons towards specific sites for light
1
2ϩ
Fig. 1 Schematic illustration of (a) fac- and (b) mer-[Ru(L ) ]
.
3
9–12
activated secondary reactions; i.e. artificial photosynthesis.
In order to bring these individual units together within a
larger assembly, it is of considerable importance to be able to
control the spatial arrangement of both the ligands and the
metal centres in relation to one another. It has been demon-
strated that the distance between and the orientation of
individual groups within larger assemblies can have small but
significant effects on the electrochemistry and on the excited-
The presence of the mer/fac-isomerism in tris-unsymmetric
diimine complexes of ruthenium() has been observed in a
number of papers, where the individual isomers have been
1
13 19,20
99
21–23
identified by H,
C
and even Ru NMR spectroscopy.
However the number of examples where such species have been
prepared isomerically pure are scarce. Kaim and co-workers
have separated, using HPLC methods, the tris-2,2Ј-azobis-
(pyridine) complex of ruthenium(), demonstrating that the
13,14
state lifetimes,
and consequently their potential use as
2
4
photosensitisers. The isolation of single isomeric forms also
has great implication on the ease of characterisation by NMR
spectroscopy and simplifies the art of crystal growth for X-ray
structural analysis.
two isomers have different UV–VIS and IR absorbances,3
ϩ
while Tresoldi et al. isolated the two isomers of [Rh(bpp)3]
[bpp = 2,3-bis(2-pyridyl)pyrazine] by column chromatography
25
on alumina. A more comprehensive study of the possibility of
separating geometric isomers of unsymmetrically functional-
ised bipyridine complexes of ruthenium() was performed by
Rutherford and Keene using the ligand 4-(2,2-dimethylpropyl)-
Significant progress has been made in the control of the
chirality at the metal centres, using either individually isolated
2
ϩ 15
chiral building blocks such as ∆- or Λ- cis-[Ru(bpy) (py) ]
2
2
2ϩ 16
17,18
26
or cis-[Ru(bpy) (CO) ]
among others.
has been paid to the potential of meridional (mer) and facial
fac) isomerism (Fig. 1). In a typical synthesis, using three
Far less attention
4Ј-methyl-2,2Ј-bipyridine. To date however, nobody has
2
2
described the use of any of these species in the preparation of
larger polynuclear assemblies with controlled structural
integrity.
Using a similar methodology as that described by Rutherford
and Keene, a detailed investigation is described here exploring
the possible strengths of cation-exchange chromatography
upon 5-functionalised bipyridine ligands. Substitution at the
5-position has not been explored to the same extent as the
(
unsymmetrically functionalised ligands, a statistical distribu-
tion of three parts of the mer-isomer will be produced to one of
the fac-isomer, arising from the potential orientation of the
ligands in a stepwise addition to the metal centre. However, due
to the steric requirements of the ligand, it is expected that the
ratio will be further tipped in favour of the mer-form.
DOI: 10.1039/b104365j
J. Chem. Soc., Dalton Trans., 2001, 2641–2648
This journal is © The Royal Society of Chemistry 2001
2641

View Online
analogous 4-functionalised bipyridine species, probably because
of the increased steric constraints which are considered within
this study and the commercial availability of 4,4Ј-dimethyl-
Complexes. All of the ligands were coordinated to ruthen-
ium() by reaction with RuCl ؒxH O in either ethanol or
3
2
2ϩ
DMF to give the tris-chelate [Ru(L) ] . The complexes were
3
2
,2Ј-bipyridine. The orientation of the substituents relative
initially purified using cation-exchange chromatography upon
SP Sephadex® C-25, eluting with aqueous sodium chloride
solution and characterised by elemental analysis, mass spectro-
scopy and the characteristic UV–VIS absorption spectra. The
yields of the products were generally found to be higher in
DMF, presumably because of the higher reaction temperature.
However, the yields were generally lower than would have been
expected, especially with the more sterically hindered ligands.
In many of the reactions a brown precipitate was isolated
directly from the reaction mixture, which was assumed to be a
bis-chelate species indicating that the final ligation step limits
the reaction. Difficulties were experienced in the purification of
to each other is of considerable interest to us, the fac-isomer
of 5-functionalised bipyridines provides three groups placed
orthogonal. By the use of suitable functionality, the question is
posed as to whether these pseudo-octahedral complexes can be
used as building blocks for the preparation of larger networks,
opening a new route for polynuclear assemblies with controlled
geometrical isomerism.
Results and discussion
Synthesis
5
2ϩ
[
Ru(L ) ] , with the complex repeatably analyzing for an excess
3
Ligand. A series of unsymmetrical ligands were synthesised
of ammonium hexafluorophosphate despite repeated recrystal-
lisation and repeated passage down a Sephadex LH20 column.
It is assumed that the hydrophobic functionality is capable of
retaining hydrophobic species.
1
(
Scheme 1) from 5-methyl-2,2Ј-bipyridine L , prepared by a lit-
1
X-Ray structural analysis of [Ru(L ) ](PF )
3
6
2
From a partially resolved fraction taken from a cation-
exchange column (described subsequently) crystals of
1
[
Ru(L ) ](PF ) suitable for X-ray structure determination were
3
6 2
obtained from an aqueous methanol mixture. The X-ray struc-
ture determination shows the asymmetric unit contains one
1
1
[
Ru(L ) ] cation and two PF anions. The [Ru(L ) ] cations are
3
6
3
arranged into columns within the lattice and involved in weak
28,29
C–H ؒ ؒ ؒ π interactions, as observed for [Ru(bipy) ](PF ) .
3
6 2
The PF anions within the lattice are involved in a significant
6
number of short contacts (2.3–2.9 Å) with the C–H and the
1
methyl moieties of the [Ru(L ) ] cations. This is consistent with
3
the formation of C–H ؒ ؒ ؒ F hydrogen bonds between the
cations and anions and in agreement with a database study on
organometallic compounds containing fluorinated anions by
Braga and co-workers in which they showed that both BF and
4
30
PF6 anions can act as hydrogen bond acceptors. They
concluded that C–H ؒ ؒ ؒ F hydrogen bonds to BF and PF6
4
anions may be significant in the formation of many crystalline
materials.
1
6
Scheme 1 Synthetic route to ligands L –L .
The average Ru–N bond lengths and N–Ru–N angles corre-
2
7
erature synthesis and further purified by vacuum distillation.
28,29
late with those of published structures
lengths and angles are given in Table 1) The [Ru(L ) ] cations
(selected bond
Additional functionalisation was achieved, by deprotonation
with lithium diisopropylamide (LDA) and subsequent quench-
ing of the carboanion with the appropriate bromoalkane
or methyliodide for ligand L ) in a similar procedure described
for the synthesis of 4-(2,2-dimethylpropyl)-4Ј-methyl-2,2Ј-
bipyridine. The rate of the nucleophilic attack was found to
decrease significantly with the size of the incoming alkane, with
the bulky 2-bromo-2-methylpropane requiring several days at
1
3
are randomly disordered throughout the lattice as a mixture of
mer- and fac-isomers with respect to the relative orientation of
2
(
1
the methyl substituents of a single L group. The two conform-
ations refined to 64(1) and 36(1)% respectively, the major con-
formation is shown in Fig. 2. By examination of the structure,
the disorder of the structure can be accounted for by the
similarity of the two forms, differing by a single methyl group.
The orientation of the rigid cation provides suitable space in
the lattice to allow the methyl group to position either side of
one of the ligand numbered A. Since the two isomers appear
to be isostructural, separation by crystallisation would prove
extremely difficult.
26
5
room temperature to react and giving a poor yield of L .
In order to allow further reactions once the ligand has been
coordinated to a ruthenium() metal centre, (i.e. to act as a
supramolecular building block), a group needs to be appended
that is unreactive during coordination to a metal centre, but can
offer subsequent reactivity to create larger assemblies. While
aliphatic ligands could potentially be considered, the conditions
required to append secondary groups are extreme requiring the
use of strong bases. Appended bromo, carboxy and aldehyde
groups were also considered, but problems were encountered
upon complexation to the metal centre. Many bipyridine ester
complexes are known however, and simple base hydrolysis
opens up the possibility to add further functionality. 5-
Carbomethoxy)-2,2Ј-bipyridine L was investigated being the
least sterically demanding example. The ligand synthesis
was achieved from L via a modified literature procedure to
-carboxy-2,2Ј-bipyridine. From this the acid chloride was
prepared in thionyl chloride and subsequently reacted with dry
Stereochemical explorations and isomer separation
The ratio of the two isomers from a simple statistical analysis
should be three mer to one fac. Direct indication to confirm this
1
by H NMR proved impossible because of the complex nature
of the spectra of the two isomers with many signals occupying
very similar regions of the spectrum. In order to judge the rel-
ative percentage of each isomer, separation of the two species
was required.
Attempts were initially made to separate the two geometric
isomers synthesised using cation-exchange chromatography on
Sephadex C25 as has been previously described by Keene and
co-workers, eluting with aqueous sodium toluene-4-sulfonate
6
(
1
27
5
6
methanol to give L .
2
642
J. Chem. Soc., Dalton Trans., 2001, 2641–2648

View Online
1
Table 1 Selected bond lengths (Å) and angles (Њ) for [Ru(L ) ](PF )
3
6
2
Ru(1)–N(1)A
Ru(1)–N(1)B
Ru(1)–N(1)C
2.055(4)
2.048(4)
2.074(5)
Ru(1)–N(1)2A
Ru(1)–N(1)2B
Ru(1)–N(1)2C
2.063(4)
2.059(4)
2.052(4)
N(1)A–Ru(1)–N(1)B
N(1)A–Ru(1)–N(1)C
N(1)A–Ru(1)–N(1)2A
N(1)A–Ru(1)–N(1)2B
N(1)A–Ru(1)–N(1)2C
N(1)B–Ru(1)–N(1)C
N(1)B–Ru(1)–N(1)2A
N(1)B–Ru(1)–N(1)2B
172.24(18)
91.91(18)
79.06(19)
94.64(17)
96.31(17)
94.90(17)
96.33(18)
79.03(17)
N(1)B–Ru(1)–N(1)2C
N(1)C–Ru(1)–N(1)2A
N(1)C–Ru(1)–N(1)2B
N(1)C–Ru(1)–N(1)2C
N(1)2A–Ru(1)–N(1)2B
N(1)2A–Ru(1)–N(1)2B
N(1)2B–Ru(1)–N(1)2C
88.66(16)
97.53(18)
170.86(17)
78.94(18)
89.98(16)
174.15(19)
93.98(19)
Fig. 3 Schematic illustration of the possible difference in hydrophobic
1
–5
2ϩ
interactions between the complexes [Ru(L ) ] with hexanoate.
3
Fig. 2 Molecular connectivity and labeling scheme for the major
conformation of the cation. The hydrogen atoms and anions have been
removed for clarity.
the mer-isomer, the three functional groups are not aligned as
well, with two of the functional groups diametrically opposed
(
trans) to each other, and so the hydrophobic cavity is not as
pronounced, hence provides a poorer interaction with the
aliphatic anion. The toluene-4-sulfonate, allows for additional
π-stacking interactions with the pyridine rings, which negates
the interaction with the three functional groups, and so the
differentiation of the two geometric forms is much less
pronounced.
Ϫ3 31,32
solution (0.15 mol dm ).
While there was clear indication
1 2
that the complexes of ligands L and L were separable, an
effective column length of over 8 m was required. As a con-
sequence aqueous sodium hexanoate (0.125 mol dm ) was
considered, having been shown previously to have a suitable
hydrophobic interaction with polypyridine complexes of
ruthenium(). With ligands L and L the effective column
length required to separate the isomers was reduced to 5 and 3
m respectively. While ligands L and L were both separated
rapidly upon the column within 2 m.
In each case the isomer travelling fastest on the cation-
exchange column was tentatively assigned as the fac-isomer
being expected in smaller quantities), as the colour of the band
on the column was considerably less intense than the following
product and consequently less concentrated. With ligands L
and L , the approximate ratio of products obtained from the
column were three parts mer to one of fac. However, using the
more sterically demanding ligands L and L , the ratio was
significantly increased, while none of the fac- isomer of ligand
Ϫ3
6
The ruthenium() complex of ester ligand L proved a little
31
1
2
more difficult to separate. While resolution of two isomers was
achieved using sodium hexanoate after 6 m, the isolation of the
products from the eluent mixture proved to be impossible. It is
assumed that either the hexanoate associates too well with the
3
4
Ϫ
Ϫ
cation to allow exchange of the anion (such as PF , or BPh )
6 4
from the aqueous solution preventing both precipitation and
extraction into methylene chloride. Alternatively, it is possible
that the basicity of the eluent has facilitated deesterification
and the zwitterionic complex can not be extracted or precipi-
tated from aqueous solution. Using aqueous sodium toluene-4-
(
1
2
Ϫ3
sulfonate solution (0.15 mol dm ), significant broadening of
3
4
the band on the column was observed. However, after an effect-
ive 10 m column length complete separation had not been
achieved. By systematically removing the front and tail of the
band, two different products were isolated. The slower moving
fraction was subsequently identified as the fac-isomer, contrary
to the results observed with the aliphatic isomers.
5
L was observed.
With these aliphatic isomers, a greater differentiation of the
two isomeric forms is observed with the hexanoate rather than
the toluene-4-sulfonate anion. Since the fac-isomer had a
greater rate of passage through the column, the association
with the anion must be greater, reducing the effective charge on
the complex, when compared to the mer-isomer. This can be
justified by a simple consideration of the structure of the com-
plexes (Fig. 3). In the fac-isomers, the three hydrophobic groups
are all aligned along the three-fold axis of the pseudo-
octahedral structure. As a consequence it provides a hydro-
phobic pocket to receive the aliphatic tail of the hexanoate. In
6
The separation of the ester functionalised complex L
appeared to be contrary to the above hypothesis, however the
carboxyl groups do not provide a good hydrophobic cavity, and
so the differentiation of the two geometric forms must rely
upon a different structural feature. The preliminary studies
indicate that the mer-isomer has the most favourable inter-
actions with both toluene-4-sulfonate and hexanoate anions.
Further studies are required to rationalise the nature of these.
J. Chem. Soc., Dalton Trans., 2001, 2641–2648
2643

View Online
3
2ϩ
Fig. 4 Circular dichroism spectrum of Λ- (——) and ∆-[Ru(L ) ]
3
(
----).
1
Fig. 5 H NMR spectra (500 MHz, 25 ЊC, d -acetone) of (a) fac- and
6
3
2ϩ
1
The fraction of [Ru(L ) ] , identified as the mer-isomer, was
(b) mer-[Ru(L ) ](PF ) .
3 6 2
3
repeatably passed down the cation-exchange column eluted
Ϫ3
with 0.125 mol dm sodium hexanoate over the period of
6
or an unsubstituted ring. In the fac-isomer, all three of the H
several days. Once an effective column length of 12 m had been
achieved, two resolved bands of equal intensity were observed.
Preliminary H NMR studies indicated that the two products
protons are positioned over the methylpyridine rings, causing a
stronger upfield position. While the H Ј are all positioned over
unsubstituted pyridine systems, causing a relative downfield
position. With the mer-isomer, one H is over a methylpyridine
ring, giving a signal similar to that of the fac-isomer (Fig. 1),
while two are over unsubstituted pyridine systems giving rise to
a larger signal downfield of H . A similar argument can be
given to account for the order of the signals of H Ј and the
6
1
were identical. By comparison of the two circular dichroism
spectra of the two fractions, the isomer has separated into the
two enantiomers (Fig. 4). Since the anion is achiral, the chirality
of the polydextrose Sephadex support must be responsible for
the chiral induction. While similar behaviour has been observed
with dinuclear species eluted with sodium toluene-4-sulfonate,
this is the first instance where it has been observed using sodium
hexanoate with a mononuclear species.
While the use of cation-exchange chromatography on a
Sephadex support for the separation of geometric isomers of
thermodynamically stable transition metal bipyridine com-
plexes appears to be possible, the efficacy of the procedure can
be reduced due to the inherent chirality of the support itself.
After repeated passage down the column, the enantiomeric
resolution will shadow the desired delicate geometric separ-
6
6
6
32
three signals for the methyl groups. The system is however made
6
4
more complex by exceptional couplings between H and H ,
6
4
and H Ј and H Ј, as observed in the fac-isomer.
2–5
The spectra for the complexes mer-[Ru(L ) ](PF ) demon-
strate similar behaviour to that of mer-[Ru(L ) ](PF ) , how-
3
1
6 2
3
6 2
ever, as the aliphatic chain length increases, the splitting
between the three sets of signals for each ligand system
2
3
4
becomes more apparent. In the complex of L , the H and H
5
are partially resolved, while ligand L has 21 independent
3
3
identifiable aromatic signals. H and H Ј show significant differ-
ences in their relative position, and yet do not experience ring
currents of adjacent aromatic groups. As a consequence, the
σ-electron donation from the aliphatic groups must play a
significant role in the relative position of all of the aromatic
signals. The three bipyridine ligands are identical, only their
relative orientation differs, with respect to the groups situated
trans to each coordinated pyridine group through the ruthen-
ium() centre. The electron donating groups increase the
Lewis basicity of the pyridine groups. This has a consequence
upon the other coordinated groups giving rise to a difference
in the chemical shifts observed. While the connectivity of the
6
2ϩ
ation and as in the case of [Ru(L ) ] could potentially prevent
complete separation of the two geometric isomers.
3
1
H NMR studies
1
The analysis of the H NMR spectra of the fac-isomers is rela-
tively straightforward arising from the C symmetry (Fig. 1). As
a consequence the three ligands are all equivalent giving rise to
seven aromatic signals. The mer-isomers on the other hand
possess C symmetry, with all three of the ligands inequivalent,
potentially giving rise to 21 aromatic signals. However, due
to the similarity of the signals many of these occupy very
similar regions of the spectrum. By the use of H correlation
3
1
1
1
three ring systems was established using H-COSY techniques,
spectroscopy (COSY), signals were isolated into individual ring
systems (Table 2).
analysis of the relative peak position is unduly complex and
identification of individual ring systems was not achieved.
1
1
6
1
The complex fac-[Ru(L ) ](PF ) [Fig. 5(a)] shows the H
fac-[Ru(L ) ](PF ) demonstrates a significantly different H
3
6
2
3 6 2
NMR spectrum expected, with seven aromatic signals. A point
NMR spectrum [Fig. 6(a)] to that of the other ligands con-
sidered. As a consequence of the electron-withdrawing nature
of the ester substituent, the functionalised pyridine ring has
chemical shifts significantly downfield of those of the other6
complexes examined, with a more typical coupling between H
6
worthy of note however is that H was not the expected tight
4
doublet, but had a strong coupling to H , of 20 Hz. Similar
behaviour was also observed upon the unfunctionalised pyr-
6
idine ring, with an unusually large coupling between H Ј and
4
4
H Ј. These couplings appear to be solvent independent, being
and H observed (1.4 Hz).
6
observed in deuterated acetone, acetonitrile and DMSO. An
mer-[Ru(L ) ](PF ) has a spectrum with great similarity to
3
6 2
2
analogous coupling was also present in the complex of L and
that of the analogous fac-isomer. However, the three sets of
ligand signals offer a significant difference in their chemical
3
to a lesser extent with L . As the size of the aliphatic functional-
4
5
ity increased to that of L the effect disappeared, and was not
shifts. In particular the H Ј signals are completely resolved into
6
present in the ester functionalised complex fac-[Ru(L ) ](PF ) .
three sets of multiplets [Fig. 6(b)]. This large difference in
the three ligand environments is not only a consequence of
the strong intraligand electronic effects developed from the
attached carbonyl function, but also the interligand effects of
the carbonyl group to orthogonal ring systems. The strength
of this latter effect appears to be considerable, by comparison
3
6 2
1
The complex mer-[Ru(L ) ](PF ) [Fig. 5(b)] has a remarkably
3
6 2
similar spectrum to that of the fac-isomer. Only the signals
6
6
assigned as H and H Ј showed any significant difference in
their chemical shifts as a consequence of slight ring current
effects depending on whether they are positioned over a methyl
2
644
J. Chem. Soc., Dalton Trans., 2001, 2641–2648

View Online
J. Chem. Soc., Dalton Trans., 2001, 2641–2648
2645

View Online
6
the ligand L was successfully coordinated to the metal centre
and separation of the mer/fac-isomers attempted. In order to
demonstrate further functionality can be achieved, the separ-
ated complexes were deesterified under basic conditions to yield
the acid. Unfortunately, the improved water solubility has as yet
prevented characterisation, although it has been assumed that
the metal centre stereochemistry is maintained. Similarly, the
resolved esters were reacted in thionyl chloride to give the tris-
acid chloride complex. While this species appeared to readily
react with several amines, insufficient quantities of the products
were isolated to facilitate characterisation. Because of the prob-
lems of isomeric separation of the ester-functionalised system
6
2ϩ
[
Ru(L ) ] , it is apparent that this preparative route is not suf-
3
ficient to allow appreciable quantities of these two potential
new building-blocks to be prepared and explored. Alternative
synthetic approaches to overcome this problem are currently
being explored.
1
Fig. 6 H NMR spectra (500 MHz, 25 ЊC, d -acetone) of (a) fac- and
Conclusions
6
6
(
b) mer-[Ru(L ) ](PF ) .
3
6 2
The isolation of several new fac- and mer-isomers of 5-
functionalised-2,2Ј-bipyridine complexes has been achieved
using cation-exchange chromatography, using either sodium
hexanoate or toluene-4-sulfonate. However, the isolation of the
two forms of the complexes appears to be hindered, especially
in the species taking a longer passage down the column, by the
simultaneous chiral resolution induced by the polydextrose
support. As expected, the increasing steric bulk of the
appended group has a significant effect on the ratio of the fac
to mer isomers. With an appended butyl group, only the mer
isomer was isolated. As a consequence a planned synthesis of
this isomer can be considered by the use of large bulky groups.
However, due to the statistical and steric considerations a route
to the fac-isomer using this method does not appear to be prac-
tical as the separation of a mixture of products appears to be
extremely inefficient.
Fig. 7 Possible synthetic route to new supramolecular building blocks.
of the spectrum of the mer-isomer to that of the fac-isomer.
As with the previous systems, a consideration of the relative
6
6
position of the H and H Ј signals is instructive. In the fac-
6
6
isomer, the signal for H is downfield relative to all of the H
signals of the mer-complex. In the former, the functionalised
ring experiences a cooperative effect of the three ester moieties
acting solely upon each other. Due to the electronegativity
of the oxygen atom, the proton signals for these three rings
are all deshielded. In the mer-isomer, one functionalised ring
is adjacent to a carbonyl function, while the other two are
adjacent to unsubstituted pyridine systems. As a result, the
deshielding effect on the rings with appended ester groups is
less pronounced. A similar set of arguments are possible to
explain the reason for the upfield shift demonstrated by the
Experimental
Instrumentation
1
13
H and C NMR spectra were recorded on a Brüker DPX 300
and DRX500 using the solvent as an internal reference,
electronic spectra were recorded on a Perkin-Elmer Lambda 9
spectrophotometer, circular dichroism (CD) spectra were
recorded on a Jasco J-720 spectropolarimeter and normalised
to the electronic spectra. Microanalyses and EI mass spectro-
scopy were performed by A.S.E.P., The School of Chemistry,
The QueenЈs University of Belfast. The LSIMS(FAB) and mass
spectroscopy was performed by the EPSRC mass spectrometry
Service, The University of Wales, Swansea.
6
signal for H Ј.
The potential construction of supramolecular building blocks
As discussed previously, polypyridine complexes of ruthen-
ium() have been widely used to add photophysical function
in supramolecular assemblies. In order to ensure that structural
integrity can be maintained, careful consideration of the
individual monomeric species is required. From this study, it is
apparent that geometric isomers can be separated into the two
forms. In theory, the two discussed isomers can potentially be
viewed as rigid building blocks (Fig. 7), bringing a controlled
architecture to supramolecular structures such as dendrimers.
The mer-isomer is effectively T-shaped while the fac-isomer
with its three-fold symmetry and three orthogonal functional
groups can be viewed as a corner piece to a molecular cube, or
an end unit to a helical structure. Further, with each unit being
chiral and enantiomeric separation using cation-exchange
methods discussed previously, further structural control can be
envisaged in the development of interesting new materials.
To create larger assemblies, suitable groups to facilitate fur-
ther functionalisation are required. Simple attempts to prepare
ruthenium complexes with the ligands 5Ј-bromomethyl-2,2Ј-
bipyridine and 5-(2,2Ј-bipyridine)carbaldehyde were initially
considered, however these ligands proved to be impossible to
coordinate to the transition metal. To overcome this problem,
Materials
Lithium diisopropylamide 2.0 M in cyclohexane (LDA;
Aldrich) and ruthenium trichloride hydrate (Johnson Matthey)
were used as received without further purification. Laboratory
grade solvents were used unless otherwise specified. Tetrahydro-
furan and diethylether (BDH) were distilled under N from
2
potassium and sodium respectively with benzophenone as an
indicator. SP-Sephadex C-25 (Aldrich) and Sephadex LH20
(
Fluka) were used for chromatographic purification of the
1
metal complexes. 5-Methyl-2,2Ј-bipyridine (L ) was prepared via
a Kröhnke synthesis from 2-acetylpyridine (99ϩ%; Aldrich).
33
Ligand synthesis
2
5
The aliphatically substituted ligands L –L were synthesised fol-
26
lowing a modified literature method. In a typical reaction a
solution of 5-methyl-2,2Ј-bipyridine (0.500 g, 2.94 mmol) in dry
3
THF (50 cm ) (or diethylether) was cooled to below Ϫ40 ЊC
2
646
J. Chem. Soc., Dalton Trans., 2001, 2641–2648

View Online
3
under nitrogen and a large excess of LDA (6.0 cm , 12 mmol)
ther adjustment of the pH allowed second and third fractions
to be obtained. Yield 4.20 g, 38%. Characterisation in keeping
with published results. Melting point 217–220 ЊC.
added over 30 min. After stirring at Ϫ40 ЊC for 3 h, a large
excess of the appropriate bromoalkane (16 mmol) (or methyl-
iodide in the preparation of L ) was added in dry THF (20 cm )
and the reaction allowed to slowly warm to room temperature
while stirring overnight. The reaction was quenched with water
1–2 cm ) and the solvent removed in vacuo. After the addition
of a saturated aqueous solution of sodium hydrogen carbonate
10 cm ), the mixture was extracted with dichloromethane
3 × 30 cm ). The organic phase was dried over magnesium
sulfate and filtered. Following the removal of the solvent, the
brown residue was purified firstly by column chromatography
on silica gel with dichloromethane–methanol (50 : 1), collecting
the first band as a yellow oil.
34
2
3
6
5
-(Carbomethoxy)-2,2Ј-bipyridine (L ). 5-Carboxy-2,2Ј-
3
bipyridine (0.20 g, 1.0 mmol) was refluxed in thionyl chloride
30 ml) for 3 h. The thionyl chloride was removed under
(
(
3
reduced pressure and the yellow solid dried in vacuo for 3 h.
This was dissolved in dry THF (30 ml) to which was added
triethylamine (2 ml) followed by methanol dried over molecular
sieves (5 ml) and stirred for 16 h. The solvent was removed and
the solid suspended in water (50 ml) and extracted into dichloro-
methane (3 × 30 ml). The organic layer was dried over mag-
nesium sulfate and the solvent removed in vacuo to give a white
solid. Yield 0.202 g, 94% (Found: C, 62.60; H, 4.66; N, 11.83.
C H N O ؒH O requires C, 62.06; H, 5.21; N, 12.06%). EIMS:
(
3
(
2
5
-Ethyl-2,2Ј-bipyridine (L ). Yield 86% (Found: C, 75.60; H,
1
2
10
2
2
2
ϩ
ϩ
ϩ
6
1
.83; N, 13.17. C H N ؒ1/3H O requires C, 75.76; H, 6.71; N,
12 12 2 2
ϩ 1
m/z 214 [M] , 183 [M Ϫ OMe] , 155 [MϪCO Me] , NMR
2
1 6
4.72%), EIMS: m/z 184 [M ]. NMR (CDCl , 300 MHz): H,
3
(CDCl , 300 MHz): H, δ 9.27 (1 H, s, H ), 8.72 (1 H, d, J 4.4
3
6 3 3
6
6
δ 8.67 (1 H, d, J 4.2 Hz, H Ј), 8.53 (1 H, s, H ), 8.37 (1 H, d, J 8.0
Hz, H Ј), 8.32 (1 H, d, J 8.1 Hz, H ), 7.79 (1 H, dd, J 7.8, 7.8 Hz,
H Ј), 7.65 (1 H, d, J.8.1 Hz, H ), 7.17 (1 H, d, J 4.7, 7.3 Hz,
H Ј), 2.74 (2 H, q, J 7.6 Hz, CH ), 1.29 (3 H, t, J 7.6 Hz, CH );
C, δ 156.1 (Cq), 153.9 (Cq), 149.1 (CH), 148.9 (CH), 139.5
Cq), 136.8 (CH), 136.2 (CH), 123.5 (CH), 120.8 (CH), 120.7
CH), 25.9 (CH ), 15.2 (CH ).
Hz, H Ј), 8.51 (1 H, d, J 8.1 Hz, H ), 8.48 (1 H, d, J 8.1 Hz, H Ј),
3
3
4
4
8
7
.41 (1 H, d, J.8.2 Hz, H ), 7.85 (1 H, dd, J 7.8, 7.8 Hz, H Ј),
.36 (1 H, d, J 4.8,7.8 Hz, H Ј), 3.98 (2 H, s, CH ), C, δ 164.8
4
4
5
1
3
3
5
2
3
(CO ), 158.5 (Cq), 154.0 (Cq), 149.5 (CH), 148.4 (CH), 137.0
2
13
(
(
CH), 136.1 (CH), 124.6 (Cq), 123.5 (CH), 120.9 (CH), 119.5
CH).
(
(
2
3
Complex syntheses
3
5
-Propyl-2,2Ј-bipyridine (L ). Yield 84% (Found: C, 77.89; H,
1
6
7
1
8
.47; N, 13.09. C H N ؒ1/6H O requires C, 77.58; H, 7.18; N,
The ruthenium() complexes of ligands L –L were synthesised
by similar reactions in either DMF or ethanol. In a typical
reaction a solution of 5-ethyl-2,2Ј-bipyridine (120 mg, 0.65
13
14
2
2
ϩ
1
3.92%). EIMS: m/z 198 [M ]. NMR (CDCl , 300 MHz): H, δ
3
6
6
.67 (1 H, d, J 4.2 Hz, H Ј), 8.50 (1 H, s, H ), 8.36 (1 H, d, J 8.0
3
3
Hz, H Ј), 8.31 (1 H, d, J 8.1 Hz, H ), 7.80 (1 H, dd, J 7.6, 7.8 Hz,
H Ј), 7.63 (1 H, d, J.8.1 Hz, H ), 7.23 (1 H, d, J 4.8, 7.5 Hz,
H Ј), 2.65 (2 H, q, J 7.6 Hz, CH ), 1.66 (2 H, m, CH ) 0.97 (3 H,
t, J 7.4 Hz, CH ); C, δ 155.3 (Cq), 153.1 (Cq), 148.3 (CH),
mmol) and RuCl ؒxH O (38.6 mg, 0.15 mmol) were dissolved in
3 2
4
4
DMF (30 ml) and heated at 100 ЊC for 16 h. The mixture was
cooled, diluted with water (150 ml) and added to SP Sephadex®
C-25 cation-exchange support. Once the complex was entirely
loaded on the product was eluted with 0.3 M aqueous sodium
chloride solution, collecting the orange–red fraction. The
product was reclaimed from the brine solution, by the addition
of saturated aqueous ammonium hexafluorophosphate sol-
ution and subsequent extraction with dichloromethane (5 × 50
ml) and dried in vacuo. Yield 86.7 mg. (Samples for micro-
analysis were further crystallized from acetone–water and
passed down an Sephadex® LH20 column, eluted with a 50%
methanol–acetone mixture.)
5
2
2
13
3
1
48.1 (CH), 137.0 (Cq), 135.8 (2 × CH), 127.2 (CH), 122.4
(
CH), 119.7 (CH), 33.8 (CH ), 22.9 (CH ), 18.1 (CH ).
2
2
3
4
5
-(2-Methylpropyl)-2,2Ј-bipyridine (L ). Yield 60% (Found:
C, 77.37; H, 7.60; N, 12.79. C H N ؒ1/4H O requires C, 77.56;
H, 7.67; N, 12.92%). EIMS: m/z 212 [M ]. NMR (CDCl , 300
MHz): H, δ 8.62 (1 H, d, J 5.0 Hz, H Ј), 8.45 (1 H, s, H ), 8.31
14
16
2
2
ϩ
3
1
6
6
3
3
(
1 H, d, J 7.7 Hz, H Ј), 8.22 (1 H, d, J 8.0 Hz, H ), 7.78 (1 H, dd,
4 4
J 7.5, 7.5 Hz, H Ј), 7.41 (1 H, d, J.8.0 Hz, H ), 7.24 (1 H, d,
5
J 5.0, 7.8 Hz, H Ј), 2.45 (2 H, d, J 7.2 Hz, CH ), 1.84 (1 H, m,
CH) 0.85 (6 H, d, J 6.6 Hz, CH ); C, δ 156.6 (Cq), 154.2 (Cq),
2
13
1
3
[Ru(L ) ](PF ) . Yield 62% (prepared in ethanol) [Found: C,
45.53; H, 3.98; N, 7.83. C H N RuP F ؒ2(CH ) CO requires
33 30 6 2 12 3 2
ϩ
C, 46.02; H, 4.16; N, 8.26%]. LSIMS: m/z 757 [M Ϫ PF ] , 616
[MH Ϫ 2PF ] . UV–VIS absorption (MeCN): λmax/nm (ε/dm
mol cm ) 253 (29600), 289 (80400), 448 (12000).
3
6 2
1
1
2
50.2 (CH), 149.5 (CH), 137.8 (CH), 137.4 (Cq), 137.2 (CH),
23.7 (CH), 121.1 (CH), 120.9 (CH), 42.5 (CH ), 30.4 (CH),
2.6 (CH ).
2
6
ϩ
3
3
6
Ϫ1
Ϫ1
5
5
-(2,2-Dimethylpropyl)-2,2Ј-bipyridine (L ). Yield 50%
(
Found: C, 78.21; H, 8.18; N, 12.38. C H N ؒ1/5H O requires
2
15
18
2
2
[Ru(L ) ](PF ). Yield 75% (prepared in DMF) (Found: C,
3
6
ϩ
C, 78.36; H, 8.07; N, 12.18%). EIMS: m/z 226 [M ]. NMR
4
5.84; H, 3.92; N, 8.54. C H N RuP F requires C, 45.82; H,
36 36 6 2 12
ϩ
3.92; N, 8.91%). LSIMS: m/z 799 [M Ϫ PF ] , 654
MH Ϫ 2PF ] . UV–VIS absorption (MeCN): λmax/nm (ε/dm
mol cm ) 252 (31700), 290 (88000), 447 (12600).
1
6
(
CDCl , 300 MHz): H, δ 8.67 (1 H, d, J 4.2 Hz, H Ј), 8.46 (1 H,
3
6
6
3
3
s, H ), 8.37 (1 H, d, J 8.0 Hz, H Ј), 8.30 (1 H, d, J 8.1 Hz, H ),
7
7
CH ); C, δ 155.2 (Cq), 152.8 (Cq), 149.7 (CH), 148.1 (CH),
1
1
ϩ
3
[
6
4
4
.80 (1 H, dd, J 7.6, 7.8 Hz, H Ј), 7.59 (1 H, d, J.8.1 Hz, H ),
Ϫ1
Ϫ1
5
.28 (1 H, d, J 5.3, 7.8 Hz, H Ј), 2.55 (2 H, s, CH ), 0.94 (9 H, s,
2
13
3
3
[
Ru(L ) ](PF ). Yield 65% (prepared in DMF) (Found: C,
41.5 (Cq), 137.6 (CH), 135.9 (CH), 122.4 (CH), 119.8 (CH),
19.1 (CH), 46.0 (CH ), 29.9 (Cq), 28.2 (CH ).
3
6
4
4
[
7.70; H, 4.62; N, 7.75. C H N RuP F requires C, 47.52; H,
39 42 6 2 12
ϩ
.29; N, 8.54%). LSIMS: m/z 841 [M Ϫ PF ] , 695
2
3
6
ϩ
3
MH Ϫ 2PF ] . UV–VIS absorption (MeCN): λmax/nm (ε/dm
6
Ϫ1 Ϫ1
5
-Carboxy-2,2Ј-bipyridine. 5-Methyl-2,2Ј-bipyridine (8.75 g,
mol cm ) 252 (29300), 290 (84200), 447 (12700).
5
1.45 mmol) was mixed with water (60 ml) and 35 g of KMnO4
was added in 7 portions over 7 h, heated initially at 70 ЊC for 3 h
and then at 90 ЊC for the subsequent 4 h. The mixture was
filtered hot through Celite® and washed with hot water (3 × 50
ml). The pale pink solution was then slowly acidified (dropwise)
with 1 M HCl to obtain a while precipitate at pH ca. 4–5. The
precipitate was collected by filtration and dried in vacuo. Fur-
4
[Ru(L ) ](PF ). Yield 94% (prepared in DMF) (Found: C,
3
6
47.40; H, 4.48; N, 7.24. C H N RuP F ؒ2H O requires C,
47.42; H, 4.92; N, 7.90%). LSIMS: m/z 883 [M Ϫ PF ] , 738
[MH Ϫ 2PF ] . UV–VIS absorption (MeCN): λmax/nm (ε/dm
mol cm ) 252 (30900), 290 (91300), 446 (14100).
42
48
6
2
12
2
ϩ
6
ϩ
3
6
Ϫ1
Ϫ1
J. Chem. Soc., Dalton Trans., 2001, 2641–2648
2647

View Online
5
[
Ru(L ) ](PF ). Yield 54% (prepared in ethanol) (Found: C,
Acknowledgements
3
6
1
4
4
4
[
8.38; H, 4.84; N, 8.17. C H N RuP F ؒ
8.66; H, 4.99; N, 7.88%). LSIMS: m/z 925 [M Ϫ PF ] , 780
NH PF requires C,
45
54
6
2
12
4
6
ϩ
We would like to thank Prof. F. Richard Keene for useful dis-
cussions, funding provided by the EPSRC, the EPSRC Mass
Spectrometry Service in Swansea for the LSIMS spectra and
Johnson Matthey Ltd. for the loan of RuCl ؒxH O.
6
ϩ
3
MH Ϫ 2PF ] . UV–VIS absorption (MeCN): λmax/nm (ε/dm
6
Ϫ1 Ϫ1
mol cm ) 253 (29600), 290 (88600), 447 (13200).
3
2
6
[
Ru(L ) ](PF ). Yield 62% (prepared in ethanol) (Found:
3
6
C, 40.91; H, 2.97; N, 7.80. C H N O RuP F ؒH O requires
References
36
30
6
6
2
12
2
ϩ
C, 41.11; H, 3.07; N, 7.99%). LSIMS: m/z 888 [M Ϫ PF ] ,
6
ϩ
ϩ
1 J. P. Sauvage, Transition Metals in Supramolecular Chemistry, Wiley,
Chichester, 1999.
2 V. Balzani, A. Juris, M. Venturi, S. Campagna and S. Serroni, Chem.
Rev., 1996, 96, 759.
8
(
73 [M Ϫ (PF ϩ CH )] , 742 [MH Ϫ 2PF ] , 729 [MH Ϫ
2PF ϩ CH )] . UV–VIS absorption (MeCN) λmax/nm (ε/dm
6
ϩ
3
6
3
6
Ϫ1
3
Ϫ1
mol cm ) 251 (30500), 290 (94800), 477 (11300).
3
V. Balzani and F. Scandola, Supramolecular Photochemistry, Ellis
Horwood, Chichester, 1991.
Separation of mer/fac-isomers. The mixture of the geometric
isomers (typically 70–100 mg) were converted to the chloride
salt by metathesis with LiCl in acetone solution and the solid
collected by filtration on Celite®. The red solid was extracted
with water and the resulting solution introduced onto a SP
Sephadex C-25 column (dimensions 26 mm × 1.6 m). Eluent
flow was regulated by use of a peristaltic pump. On elution of
4
5
C. Kaes, A. Katz and M. W. Hosseini, Chem. Rev., 2000, 100, 3553.
P. A. Anderson, G. F. Strouse, J. A. Treadway, F. R. Keene and T. J.
Meyer, Inorg. Chem., 1994, 33, 3863.
6 A. Juris, S. Barigelletti, S. Campagna, V. Balzani, P. Belser and
A. von Zelewsky, Coord. Chem. Rev., 1988, 84, 85.
7
E. S. Dodsworth and A. B. P. Lever, Chem. Phys. Lett., 1986, 124,
1
52.
8
9
A. B. P. Lever, Inorg. Chem., 1990, 29, 1271.
V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni and M.
Venturi, Coord. Chem. Rev., 1994, 132, 1.
Ϫ3
Ϫ3
0
.125 mol dm sodium hexanoate (0.15 mol dm sodium
6
toluene-4-sulfonate with ligand L ), the intense red band was
typically repeatedly passed through the column until separation
was clearly observed. The material was collected from the col-
umn and a saturated aqueous solution of NH PF added to the
10 A. Juris, V. Balzani, S. Campagna, G. Denti, S. Serroni, G. Frei and
H. U. Güdel, Inorg. Chem., 1994, 33, 1491.
1 M. A. Haga, M. M. Ali and R. Arakawa, Angew. Chem., Int. Ed.
Engl., 1996, 35, 76.
2 M. A. Haga, M. M. Ali, S. Koseki, K. Fujimoto, A. Yoshimura,
K. Nozaki, T. Ohno, K. Nakajima and D. J. Stufkens, Inorg. Chem.,
1
4
6
resulting orange solution and extracted with dichloromethane
1
3
(
3 × 20 cm ). The solvent was removed under reduced pressure.
Excess inorganic salts were removed by the passage through a
short Sephadex LH20 column (eluted with 50% methanol–
acetone) and the product isolated by removal of the solvent and
1
996, 35, 3335.
13 B. D. Yeomans, L. S. Kelso, P. A. Tregloan and F. R. Keene, Eur. J.
Inorg. Chem., 2001, 239.
1
14 T. J. Rutherford and F. R. Keene, J. Chem. Soc., Dalton Trans., 1998,
dried in vacuo. H NMR data is given in Table 2.
1
155.
5 X. Hua and A. von Zelewsky, Inorg. Chem., 1995, 34, 5791.
16 T. J. Rutherford, M. G. Quagliotto and F. R. Keene, Inorg. Chem.,
1
X-Ray structural analysis
1
995, 34, 3857.
Data were collected a Bruker-AXS SMART diffractometer
1
7 M. J. Kim, F. M. MacDonnell, M. E. GimonKinsel, T. DuBois, N.
35
using the SAINT-NT software with graphite monochromated
Mo-Kα radiation. A crystal was mounted on to the diffract-
ometer at low temperature under dinitrogen at ca. 120 K.
Crystal stability was monitored and there were no significant
variations (< ± 2%). Cell parameters were obtained from
Asgharian and J. C. Griener, Angew. Chem., Int. Ed., 2000, 39, 615.
18 D. Hesek, Y. Inoue, S. R. L. Everitt, H. Ishida, M. Kunieda and
M. G. B. Drew, Inorg. Chem., 2000, 39, 317.
1
9 M. J. Cook, A. P. Lewis, G. S. G. McAuliffe and A. J. Thomson,
Inorg. Chim. Acta, 1982, 64, 25.
2
0 M. J. Cook, A. P. Lewis and G. S. G. McAuliffe, Org. Magn. Res.,
4
00 accurately centred reflections in range 3–50Њ. ω–θ scans
1
984, 22, 388.
were employed for data collection and Lorentz and polarisation
corrections and empirical absorption corrections were applied.
The structure was solved using direct methods and refined
with the SHELXTL version 5.0 and SHELXL-98 program
21 C. Brevard and P. Granger, Inorg. Chem., 1983, 22, 532.
22 G. Orellana, A. Kirsch-De Mesmaeker and N. J. Turro, Inorg.
Chem., 1990, 29, 882.
2
3 G. Predieri, C. Vignali, G. Denti and S. Serroni, Inorg. Chim. Acta,
1993, 205, 145.
4 M. Krejcik, S. Zalis, J. Klima, D. Sykora, W. Matheis, A. Klein and
W. Kaim, Inorg. Chem., 1993, 32, 3362.
36
packages and the non-hydrogen atoms were refined with
anisotropic thermal parameters. Hydrogen-atom positions were
added at idealised positions and a riding model with fixed
thermal parameters (U = 1.2U for the atom to which they are
2
25 G. Tresoldi, S. LoSchiavo and P. Piraino, Inorg. Chim. Acta, 1997,
254, 381.
ij
eq
bonded (1.5Ueq for CH )), was used for subsequent refinements.
2
6 T. J. Rutherford, D. A. Reitsma and F. R. Keene, J. Chem. Soc.,
Dalton Trans., 1994, 3659.
3
2
2
The function minimised was Σ[w(|F | Ϫ |F | )] with reflection
weights w = [σ |F | ϩ (g P) ϩ (g P)] where P = (max. |F | ϩ
o
c
Ϫ1
2
2
2
2
27 K. D. Bos, J. G. Kraaijkamp and J. G. Noltes, Synth. Commun.,
979, 9, 497.
28 M. Biner, H.-B. Burgi, A. Ludi and C. Rohr, J. Am. Chem. Soc.,
992, 114, 5197.
29 D. P. Rillema, D. S. Jones, C. Woods and H. A. Levy, Inorg. Chem.,
o
1
2
o
1
2
2
|F | )/3.
c
1
1
Crystal data for [Ru(L ) ](PF ) . M = 901.64, monoclinic,
3
6 2
1
992, 31, 2935.
space
group
P21,
a = 10.5990(15),
b = 13.5789(18),
3
0 F. Grepioni, G. Cojazzi, S. M. Draper, N. Scully and D. Braga,
3
c = 13.3087(18) Å, β = 105.815(3)Њ, U = 1842.9(4) Å , Z = 2,
µ = 0.607 mm , Rint = 0.0704, transmission range (max,
Organometallics, 1998, 17, 296.
31 N. C. Fletcher and F. R. Keene, J. Chem. Soc., Dalton Trans., 1999,
683.
32 N. C. Fletcher, P. C. Junk, D. A. Reitsma and F. R. Keene, J. Chem.
Soc., Dalton Trans., 1998, 133.
33 T. L. J. Huang and D. G. Brewer, Can. J. Chem., 1981, 59, 1689.
Ϫ1
min.) = 0.848, 0.759. A total of 14232 reflections were measured
for the angle range 3 < 2θ < 57Њ and 8088 independent reflec-
tions were used in the refinement. The final parameters were
I
wR2 = 0.1328 and R1 = 0.0509 [I > 2σ ].
3
4 M. R. Ghadiri, C. Soares and C. Choi, J. Am. Chem. Soc., 1992, 114,
CCDC reference number 166368.
See http://www.rsc.org/suppdata/dt/b1/b104365j/ for crystal-
lographic data in CIF or other electronic format.
8
25.
3
5 SAINT-NT, Bruker, Madison, WI, 1998.
36 G. M. Sheldrick, SHELXTL Version 5.0, Madison, WI, 1998.
2
648
J. Chem. Soc., Dalton Trans., 2001, 2641–2648