Trimetallic complexes featuring Group 10 tetracyanometallate dianions as bridging ligands

Article abstract of DOI:10.1016/j.ica.2005.10.001

The trimetallic complexes {Ru(PPh₃)₂Cp}₂{μ-M(CN)₄} and {Ru(dppe)Cp*}₂{μ-M(CN)₄} (M = Ni, Pd, Pt) have been prepared from reactions of RuCl(PPh₃)₂Cp or RuCl(dppe)Cp* with the appropriate tetracyanometallate salt, and structurally characterised. While a similar reaction of FeCl(dppe)Cp with K₂[Pt(CN)₄] afforded {Fe(dppe)Cp}₂{μ-Pt(CN)₄}, the iron cyanide complex Fe(CN)(dppe)Cp was isolated as the only iron containing product from reaction of FeCl(dppe)Cp with K₂[Ni(CN)₄]. The trimetallic complexes can be oxidised in two sequential one-electron steps. Spectroelectrochemical experiments reveal weak NIR absorption bands in the mono-oxidised complexes which are not present in the binuclear complex K[Ru(dppe)Cp*{Pt(CN)₄}], and are therefore attributed to Ru^II → Ru^III charge transfer processes. The coupling parameter, V_ab, extracted using Hush-style analysis falls in the range 250 ± 50 cm^-1, consistent with the weak interaction between the Group 8 metal centres. The energy of the IVCT process is dominated by reorganisation energy of the Group 8 metal-ligand fragment.

Full text of DOI:10.1016/j.ica.2005.10.001

Inorganica Chimica Acta 359 (2006) 3459–3466
www.elsevier.com/locate/ica
Trimetallic complexes featuring Group 10 tetracyanometallate
dianions as bridging ligands
´
´
Richard L. Cordiner, Matthew P. Feroze, Carlos Lledo-Fernandez, David Albesa-Jove,
*
Judith A.K. Howard, Paul J. Low
Department of Chemistry, University of Durham, South Road, Durham, DH1 3LE, UK
Received 30 August 2005; received in revised form 3 October 2005; accepted 3 October 2005
Available online 15 November 2005
Dedicated to Professor D.M.P. Mingos, whose works have been a source of instruction, inspiration and discussion in so many areas of chemistry.
Abstract
The trimetallic complexes {Ru(PPh3)2Cp}2{l-M(CN)4} and {Ru(dppe)Cp*}2{l-M(CN)4} (M = Ni, Pd, Pt) have been prepared from
reactions of RuCl(PPh3)2Cp or RuCl(dppe)Cp* with the appropriate tetracyanometallate salt, and structurally characterised. While a
similar reaction of FeCl(dppe)Cp with K2[Pt(CN)4] afforded {Fe(dppe)Cp}2{l-Pt(CN)4}, the iron cyanide complex Fe(CN)(dppe)Cp
was isolated as the only iron containing product from reaction of FeCl(dppe)Cp with K2[Ni(CN)4]. The trimetallic complexes can be
oxidised in two sequential one-electron steps. Spectroelectrochemical experiments reveal weak NIR absorption bands in the mono-oxi-
dised complexes which are not present in the binuclear complex K[Ru(dppe)Cp*{Pt(CN)4}], and are therefore attributed to RuII ! RuIII
charge transfer processes. The coupling parameter, Vab, extracted using Hush-style analysis falls in the range 250 50 cmꢀ1, consistent
with the weak interaction between the Group 8 metal centres. The energy of the IVCT process is dominated by reorganisation energy of
the Group 8 metal–ligand fragment.
ꢀ 2005 Elsevier B.V. All rights reserved.
Keywords: Cyanide; Spectroelectrochemistry; Metal–metal interactions
1. Introduction
ear [4,5], and bent polymetallic complexes [5,6], molecular
squares and polyhedra [7] and materials with extended
structures [8]. We have recently been drawn to this work
as part of a larger study concerned with electron transfer
between Group 8 metal centres through unsaturated
bridges of general form Cn [9] and CnN [10], and had cause
to consider the influence that the size and electronic struc-
ture of the Group 10 metal might play in moderating the
interactions between remote metal groups in a mixed
valence complex featuring the linear M(II)–NC–M0–CN–
M(III) (M = Fe, Ru; M0 = Ni, Pd, Pt) assembly. We report
herein descriptions of the syntheses, molecular structures
and electrochemical responses of complexes {M(PP)Cp0}2-
{l-M0(CN)4} [M = Fe, Ru; PP = (PPh3)2, dppe; Cp0 = Cp,
Cp*; M0 = Ni, Pd, Pt (not all combinations)], and the char-
acteristics of the mixed-valance complexes that can be
derived from them.
The cyanide ligand is almost ubiquitous in studies of
intramolecular electron transfer in metal complexes.
Although the tricentennary of the discovery of Prussian
Blue was recently celebrated [1], investigations of cyanide
mediated electron transfer continue to raise interest [2],
and ‘‘designer’’ derivatives of cyanide bridged polymetallic
complexes featuring a wide-variety of molecular frame-
works and polygonal structures are a source of consider-
able research effort driven by the electronic, magnetic
and/or photonic properties such materials may offer [3].
The tetracyanometallate dianions derived from Ni(II),
Pd(II) and Pt(II) are useful reagents for the assembly of lin-
*
Corresponding author.
E-mail address: p.j.low@durham.ac.uk (P.J. Low).
0020-1693/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2005.10.001

3460
R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
2. Experimental
2.1. General conditions
Cipso); 133.8 (t, JCP = 5 Hz, Cortho); 129.3 (s, Cpara); 128.3
(t, JCP = 5 Hz, Cmeta); 83.4 (s, Cp). ES(+)-MS (m/z): 1568
[M + Na]+, 691 [Ru(PPh3)2Cp]+.
Reactions were carried out under an atmosphere of dry
nitrogen using standard Schlenk methods. No special pre-
cautions were taken during work-up and isolation. The
complexes RuCl(PPh3)2Cp [11], RuCl(dppe)Cp* [12], and
FeCl(dppe)Cp [13] were prepared by literature methods.
Other reagents were purchased and used as received.
NMR spectra were recorded on Varian Unity-300 (1H,
299.91 MHz; 31P, 121.40 MHz) or INOVA 500 (13C,
125.69 MHz) spectrometers from solutions in CDCl3, and
referenced against residual solvent resonances or an exter-
nal H3PO4 reference. Electrochemical measurements were
made from solutions in CH2Cl2 or THF containing 0.1 M
NBu4BF4 supporting electrolyte, using a conventional
three electrode cell and recorded on an AutoLab
PGSTAT-30 potentiostat. A platinum dot working elec-
trode was employed together with Pt wire counter and
pseudo reference electrodes. All potentials are reported
against SCE, being referenced against an internal ferro-
cene/ferrocenium [Fc/Fc+ = +0.46 V versus SCE (CH2Cl2)
or +0.56 V versus SCE (THF)] or decamethylferrocene/
decamethylferrocenium [Fc*/Fc*+ = ꢀ0.02 V versus SCE
(CH2Cl2) or 0.13 V versus SCE (THF)] couples [14]. Spec-
troelectrochemical studies were carried out in an OTTLE
cell of standard design [15], cooled at ꢀ30 ꢁC by cold nitro-
gen gas, using sample solutions (0.1 M NBu4BF4/CH2Cl2)
approximately 0.1 mM in analyte. The reported extinction
coefficients have been corrected to allow for the compro-
portionation equilibria [16].
2.2.3. [{Ru(PPh3)2Cp}2{Pd(CN)4}] (1b)
(0.100 mmol, 48%). Found: C, 63.38; H, 4.39; N, 3.57.
C86H70P4N4Ru2Pd Æ 0.5CH2Cl2 requires: C, 63.56; H,
1
4.38; N, 3.43. H NMR: d 4.30 (s, 10H, Cp); 7.33–7.10
(m, 72 H, PPh3). 13C NMR: d 136.0 (m, JCP = 20 Hz,
Cipso); 132.5 (t, JCP = 5 Hz, Cortho); 128.1 (s, Cpara); 127.0
(t, JCP = 5 Hz, Cmeta); 82.2 (s, Cp). ES(+)-MS (m/z):
1614, [M + Na]+; 691, [Ru(PPh3)2Cp]+.
2.2.4. [{Ru(PPh3)2Cp}2{Pt(CN)4}] (1c)
(0.095 mmol, 46%). Found C, 60.64; H, 4.12; N, 3.37.
C86H70P4N4Ru2Pt Æ 0.5(CH2Cl2) requires: C, 60.30; H,
1
4.15; N, 3.25. H NMR: d 4.30 (s, 10H, Cp); 7.30–7.26
(m, 72 H, PPh3). 13C NMR: d 137.3 (m, JCP = 22 Hz,
Cipso); 133.7 (t, JCP = 5 Hz, Cortho); 129.4 (s, Cpara); 128.3
(t, JCP = 4 Hz, Cmeta); 83.4 (s, Cp). ES(+)-MS (m/z): 1705
[M + Na]+; 691 [Ru(PPh3)2Cp]+.
2.2.5. [{Ru(dppe)Cp*}2{Ni(CN)4}] (2a)
(0.129 mmol, 68%). Found: C, 63.37; H, 5.45; N, 3.97.
C76H78P4N4Ru2Ni requires: C, 63.74; H, 5.49; N, 3.91.
1H NMR: d 1.45 (s, 15H, Cp*); 2.06, 2.62 (m, 4H, dppe);
7.15–7.70 (m, 20H, Ph). 13C NMR: d 142.9 (s, CN); 136.9
(m, Cipso); 134.2 (m, Cipso); 133.6 (m, 2 · Cortho); 130.0 (s,
CN); 129.9 (s, Cpara); 129.7 (s, Cpara); 128.6 (t, JCP = 5 Hz,
Cmeta); 128.0 (t, JCP = 5 Hz, Cmeta); 91.7 (s, C5Me5); 28.8–
28.4 (m, dppe); 10.1 (s, C5Me5). ES(+)-MS (m/z): 1433
[M + H]+; 635 [Ru(dppe)Cp*]+.
2.2. General procedure
2.2.6. [{Ru(dppe)Cp*}2{Pd(CN)4}] (2b)
(0.068 mmol, 37%). Found: C, 60.81; H, 5.25; N, 3.70.
C76H78P4N4Ru2Pd Æ 2(C2H5OH) requires: C, 61.21; H,
2.2.1. Preparation of [{RuL2Cp0}2{l-M(CN4)}] (L = PPh3,
Cp0 = Cp; L2 = dppe, Cp0 = Cp*; M = Ni, Pd, Pt)
1
5.77; N, 3.56. H NMR: d 1.48 (s, 15H, Cp*); 2.00, 2.57
A 50 ml, two-necked Schlenk flask was charged with
K2[M(CN)4] (0.208 mmol) and RuClL2Cp0 (0.415 mmol).
The mixture was suspended in MeOH (30 ml) and heated
at reflux for 1 h, after which time the solution was cooled
and the precipitate was collected and washed with cold
methanol to afford [{RuL2Cp0}2{l-M(CN4)}] as a bright
yellow solid, which was purified by column chromatogra-
phy (alumina). Crystals of compounds 1a–c suitable for
X-ray diffraction studies were obtained by slow diffusion
of MeOH into a CH2Cl2 solution. Crystals of compounds
2a and 2c suitable for X-ray diffraction studies were
obtained by slow diffusion of EtOH into CHCl3 solutions.
Crystals of compound 2b were obtained by slow diffusion
of EtOH into a CH2Cl2 solution.
(m, 4H, dppe); 7.10–7.70 (m, 20H, Ph). 13C NMR: d
140.9 (s, CN); 136.7 (m, Cipso); 134.0 (m, Cipso); 133.64 (t,
JCP = 5 Hz, Cortho); 133.5 (t, JCP = 5 Hz, Cortho); 130.0 (s,
Cpara); 129.8 (s, Cpara); 128.7 (t, JCP = 5 Hz, Cmeta); 128.0
(t, JCP = 5 Hz, Cmeta); 127.2 (s, CN); 91.7 (s, C5Me5);
28.8–28.4 (m, dppe); 10.1 (s, C5Me5). ES(+)-MS (m/z):
1481 [M + H]+; 635 [Ru(dppe)Cp*]+.
2.2.7. [{Ru(dppe)Cp*}2{Pt(CN)4}] (2c)
(1.08 mmol, 57%). Found: C, 58.16; H, 4.98; N, 3.54.
C76H78P4N4Ru2Pt requires: C, 58.19; H, 5.01; N, 3.57.
1H NMR: d 1.48 (s, 15H, Cp*); 2.08, 2.63 (m, 4H, dppe),
7.15–7.75 (m, 20H, Ph). 13C NMR: d 136.7 (m, Cipso);
134.0 (m, Cipso); 133.7 (t, JCP = 5 Hz, Cortho); 133.6 (s,
CN); 133.5 (t, JCP = 5 Hz, Cortho); 130.0 (s, Cpara); 129.8
(s, Cpara); 128.7 (t, JCP = 5 Hz, Cmeta); 128.0 (t, JCP = 4 Hz,
2.2.2. [{Ru(PPh3)2Cp}2{Ni(CN)4}] (1a)
(0.149 mmol, 72%). Found: C, 65.78; H, 4.52; N, 3.59.
C86H70P4N4Ru2Ni Æ 0.5(CH2Cl2) requires: C, 65.48; H,
C
meta); 121.6 (s, CN); 91.7 (s, C5Me5); 28.8–28.5 (m, dppe);
1
4.51; N, 3.53. H NMR: d 4.29 (s, 10H, Cp); 7.30–7.25
10.1 (s, C5Me5). ES(+)-MS (m/z): 1569 [M + H]+; 635
(m, 72 H, PPh3). 13C NMR: d 137.4 (m, JCP = 22 Hz,
[Ru(dppe)Cp*]+.

R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
3461
2.2.8. K[Ru(dppe)Cp*{Pt(CN)4}] (3)
refined crystal structure. In 2b, the Squeeze tool was also
applied to remove approximately six disordered dichloro-
methane molecules per unit cell, giving a total of ca.
244.9 eꢀ within the void volume, and leaving two ordered
ethanol molecules per formula unit, which are hydrogen
bonded to cyanide groups. In 2c, Squeeze was applied to
remove approximately six disordered chloroform mole-
cules per unit cell, giving a total of ca. 357.8 eꢀ within
the void volume, and leaving two molecules of ethanol
per formula unit, which are hydrogen bonded to cyanide
groups in a manner similar to that in 2b. Crystal data
and experimental details are listed in Table 2.
A 50 ml, two-necked Schlenk flask was charged with
K2[Pt(CN)4] (127 mg, 0.337 mmol) and RuCl(dppe)Cp*
(150 mg, 0.224 mmol). The mixture was suspended in
MeOH (10 ml) and heated at reflux for 1.5 h after which
time the solution was cooled and the solvent removed.
The yellow residue was dissolved in the minimum quantity
of CH2Cl2 and filtered. Removal of solvent afforded 3 as a
yellow powder (168 mg. 0.172 mmol, 77%). Found: C,
48.85; H, 4.07; N, 5.28. C40H39P2N4RuPtK requires: C,
1
49.38; H, 4.04; N, 5.76. H NMR: d 1.47 (s, 15H, Cp*);
2.08, 2.64 (m, 4H, dppe); 7.20–7.68 (m, 20H, Ph). ES(+)-
MS (m/z): 635 [Ru(dppe)Cp*]+. ES(ꢀ)-MS (m/z): 932
[Ru(dppe)Cp*{Pt(CN)4}]ꢀ; 273 [Pt(CN)3]ꢀ.
3. Results and discussion
2.2.9. [{Fe(dppe)Cp}2{Pt(CN)4}] (4c)
The compounds 1a–c and 2a–c were prepared by reac-
tion of the half-sandwich metal chlorides RuCl(PPh3)2Cp
or RuCl(dppe)Cp* with one half-equivalent of the appro-
priate potassium tetracyanometallate (Scheme 1). The
KCl by-product was removed by extraction and filtration,
and the pure trimetallic compounds were obtained as yel-
low crystalline solids in moderate to good yield after col-
umn chromatography and crystallisation.
A 50 ml, two-necked Schlenk flask was charged with
K2[Pt(CN)4] (102 mg, 0.27 mmol) and FeCl(dppe)Cp
(300 mg, 0.54 mmol). The mixture was suspended in
MeOH (30 ml) and heated at reflux for 90 min after which
time reaction was cooled and the solvent removed. The res-
idue was then dissolved in a minimum volume of CH2Cl2,
filtered, loaded onto a silica column and eluted with 60:40
acetone:hexane. The resultant red band was collected and
the solvent removed. Subsequent crystallisation from slow
diffusion of MeOH into a CH2Cl2 solution resulted in the
formation of 4c as red crystals (156 mg, 0.12 mmol, 43%).
1H NMR: d 2.08, 2.63 (2 · br, 4H, dppe); 4.19 (s, 5H,
Cp); 7.20–7.79 (m, 20H, Ph).
The compounds were readily characterised from solu-
tion spectroscopic data (Table 1). In each case the IR spec-
trum clearly revealed two m(C„N) bands which were
assigned to bridging (2136–2157 cmꢀ1) and non-bridging
(2118–2130 cmꢀ1) CN moieties by comparison with the
spectra of the precursor K2[M(CN)4] species (Table 1).
The Cp and Cp* ligands gave rise to the expected reso-
1
2.3. Crystallography
nances in the H and 13C NMR spectra, with additional
resonances arising from the phosphine ligands. The 31P
NMR spectra were also unremarkable and simply served
to confirm the presence of the phosphine ligands. The com-
plexes co-crystallised with solvent, and the tenacity of the
crystalline samples to retain portions of these solvents
was apparent in the microanalytical results. The positive-
ion electrosopray mass spectra (ES(+)-MS) displayed iso-
topic envelopes arising from the [M + Na]+ (1a–b) or
[M + H]+ ions (2a–c) as well as fragment ions correspond-
ing to the [Ru(PPh3)2Cp]+ and [Ru(dppe)Cp*]+ at m/z 691
and 635, respectively.
Diffraction data were collected on Bruker three-circle
diffractometers with SMART 6K (for 1c), SMART 1K (for
1a, 1b and 2c) or APEX (for 2a and 2b) CCD area detectors,
using graphite-monochromated sealed-tub Mo Ka radia-
tion. The data collection was carried out at 120 K using
cryostream (Oxford cryosystem) open flow N2 cryostats.
Reflection intensities were integrated using the SAINT
V6.22 program [17], for 2a–c, and SAINT V6.02a [18] for
1a–c.
The crystal structures were solved using direct-methods
and refined by full matrix least-squares against F2 of all
data using SHELXTL [19] software. All non-hydrogen atoms
where refined in anisotropic approximation, except a chlo-
rine atom of a dichloromethane molecule in 1b, which was
isotropically refined. Hydrogen atoms were either located
by a difference map (for 2a–c) or placed in calculated posi-
tions (1a–c) and refined isotropically using a riding model.
Solvent molecules were initially present in all the struc-
tures, and in cases where all or some of the solvent mole-
cules were disordered (for 1a, 2b and 2c) Platonꢀs Squeeze
[20] tool was applied to remove them. Thus, in 1a approx-
imately four molecules of dichloromethane molecules per
unit cell were removed, resulting in a total of ca. 152.5 eꢀ
within the void volume, as calculated by Squeeze, and leav-
ing no molecules of the crystallisation solvent in the final
The bimetallic anion [Ru(dppe)Cp*{Pt(CN)4}]ꢀ was
obtained as its potassium salt (3) from a 1.5:1 reaction of
K2[Pt(CN)4] with RuCl(dppe)Cp* in refluxing methanol,
with subsequent work-up affording the material as a pow-
der. This salt was characterised by the usual spectroscopic
techniques (Table 1), including negative-ion electrospray
mass spectrometry (ES(ꢀ)-MS), which displayed isotopic
envelopes at m/z = 932 ([Ru(dppe)Cp*{Pt(CN)4}]ꢀ) and
273 ([Pt(CN)3]ꢀ). Infra-red spectroscopy revealed a single
m(C„N) band at 2130 cmꢀ1
.
With a range of ruthenium complexes in hand, analo-
gous species featuring the more electron-rich/p-donating
[Fe(dppe)Cp]+ fragment were also sought. Reaction of
the iron complex FeCl(dppe)Cp with K2[Pt(CN)4] in
refluxing methanol (90 min) resulted in the formation of

3462
R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
M
L
R
H
H
H
PPh3 Ni
PPh3 Pd
PPh3 Pt
1a
1b
1c
2a
2b
2c
R5
R5
R5
CN
M
K2[M(CN)4]
Ru NC
CN Ru
L
Ru Cl
+
-2 KCl
L
L
L
CN
L
Me dppe Ni
Me dppe Pd
Me dppe Pt
L
Scheme 1. Synthesis of [{Ru(L)Cp0}2{l-M(CN)4}].
Fe(CN)(dppe)Cp at 2063 cmꢀ1. Attempts to separate this
mixture have, as yet, been unsuccessful.
Table 1
Selected spectroscopic data for compounds 1a–c, 2a–c, 3 and 4c
Compound
m(CN)/cmꢀ1
d(Cp or Cp*) (1H/13C)
d(PR3)
K2[Ni(CN)4]
K2[Pd(CN)4]
K2[Pt(CN)4]
2124a
3.1. Molecular structures
2136a
2133a
The structure of each of the ruthenium complexes 1a–c,
2a–c was determined by X-ray diffraction studies, the struc-
ture of 4c having already been reported [5]. The crystallo-
graphic data are summarised in Table 2, whilst selected
bond lengths and bond angles are summarised in Tables
3 and 4. A representative molecular structure is shown in
Fig. 1, illustrating the atom labelling scheme.
1a
1b
1c
2a
2b
2c
3
2143, 2119
2157, 2130
2157, 2129
2136, 2118
2146, 2129
2150, 2128
2130
4.29/83.43
4.30/82.16
4.30/83.43
1.45/91.68, 10.08
1.48/91.68, 10.13
1.48/91.70, 10.07
1.47/90.40, 8.79
4.19/–
43.40
43.21
43.12
71.71
75.79
75.73
75.67
100.34
4c
a
2149, 2128
For each compound, the Group 10 metal sits on a centre
of inversion in the molecule, and as such for any given
complex the parameters at each Group 8 metal centre are
identical. The geometry about the ruthenium centres is
not unusual. The geometry about the central metal M
(M = Ni, Pd, Pt) is square planar, with the sum of angles
C(1)–M–C(2) being 360ꢁ. Direct comparisons of the metri-
cal parameters about the Group 10 metal are complicated
by the different size of the metal atoms. The effective ionic
radii of square planar Ni2+, Pd2+ and Pt2+ are 0.49, 0.64
In water [22].
a red solution from which [{Fe(dppe)Cp}2{l-Pt(CN)4}]
(4c) was obtained as a red crystalline solid in 43% isolated
yield. This method of preparation is a minor modification
of that described previously by the Vahrenkamp group,
which involved the reaction of PPN2[Pt(CN)4] with
FeCl(dppe)Cp in CH2Cl2 for 3 days to afford 4c in 72%
yield [5]. The lower yield of the reaction reported here
is counterbalanced by the shorter reaction time and the
commercial availability of the potassium tetracyanoplati-
nate reagent.
Attempts to form complexes [{Fe(dppe)Cp}2{l-
M(CN)4}] [M = Ni (4a), Pd (4b)] using similar conditions
were not so successful. Reaction of FeCl(dppe)Cp with
K2[Ni(CN)4] in refluxing methanol resulted in a red/orange
solution from which orange crystals were obtained upon
work-up. Spectroscopic analysis of this material revealed
a single m(C„N) band in the IR spectrum at 2063 cmꢀ1
as well as 1H NMR resonances at 4.31 ppm (Cp) and broad
signals at 2.62 and 2.37 ppm (dppm). Comparison of these
data with literature values [10a,21], revealed this com-
pound to be Fe(CN)(dppe)Cp (5), formed by cyanide
ligand abstraction from the nickel precursor and isolated
in ca. 40% yield. The relatively high yield precludes the
notion that 5 is formed from trace amounts of KCN pres-
ent in the nickel reagent. Reaction of FeCl(dppe)Cp with
K2[Pd(CN)4] in refluxing methanol resulted in a red/orange
powder, with IR spectroscopy revealing m(C„N) bands at
2134 and 2127 cmꢀ1, consistent with [{Fe(dppe)Cp}2-
˚
and 0.60 A, respectively [23], and when allowance is made
for this variation the experimentally determined M–C(1)
and M–C(2) bond lengths display no statistically significant
difference.
The Ru–N(2) bond lengths for in 2a–c [2.037(3)–
˚
2.049(2) A] are marginally shorter than those in 1a–c
˚
[2.060(8)–2.069(3) A], probably a consequence of electro-
static factors. Similar effects are observed on the Ru–N
bonds in benzonitrile complexes of these same RuL2Cp0
fragments [24]. Attachment of the Group 8 metal centre
has little effect on the CN moiety, there being no significant
variation in the C(1)–N(1) and C(2)–N(2) bond lengths
across the series. While the M–C(2)–N(2) bond angles are
close to linear, there is a greater deviation from linearity
in the Ru–N(2)–C(2) bond angles (Table 4). Similar effects
have been observed in other cyanide-bridged species, and
generally attributed to the differences in p-back-bonding
at the C and N termini of the cyanide moiety.
The crystal structure of the iron/platinum complex 4c
has been reported previously [5]. As with the ruthenium
complexes this structure was found to be centrosymmetric
about the [M(CN)4] moiety. The only differences of note in
{Pd(CN)4}] (4b), as well as
a band arising from

R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
3463
Table 2
Crystallographic details for compounds 1a–c and 2a–c
Compound
Formula
1a
1b
1c
2a
2b
2c
C86H70N4P4-
Ru2Ni
C86H70N4P4Ru2Pd Æ
2CH2Cl2/2MeOH
1825.82
10.5596(10)
22.500(2)
17.1648(13)
90
99.559(6)
90
4021.6(6)
1.508
C86H70N4P4Ru2Pt Æ
2CH2Cl2/2MeOH
1914.51
10.5697(5)
22.5080(9)
17.1733(7)
90
99.628(2)
90
4028.0(3)
1.578
C76H78N4P4Ru2- C76H78N4P4Ru2P Æ
Ni2CHCl3
1551.5
C76H78N4P4Ru2Pt Æ
2C2H5OH
1661
8.5326(12)
25.012(3)
20.777(3)
90
113.577(5)
90
4064.0(10)
1.357
2C2H5OH
1572
8.5279(8)
24.917(2)
19.1022(16)
90
90.971(4)
90
4058.5(6)
1.286
M
1544.19
9.8590(3)
23.3226(7)
17.7015(5)
90
92.1780(10)
90
4067.3(2)
1.261
˚
a (A)
11.3214(10)
12.3886(11)
14.3895(11)
71.639(3)
75.692(4)
78.916(4)
1841.6(3)
1.507
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
q (mg/m3)
T (K)
120(2)
120(2)
120(2)
120(2)
120(2)
120(2)
Crystal system
Space group
Z
monoclinic
P2(1)/n
2
monoclinic
P2(1)/n
2
monoclinic
P2(1)/n
2
triclinic
monoclinic
P2(1)/n
2
monoclinic
P2(1)/c
2
ꢀ
P1
1
l (mmꢀ1
)
0.718
0.856
2.366
1.009
0.710
2.207
Reflections collected
Independent
29102
45343
43801
12290 [0.0559]
22020
10996 [0.0276]
50442
12325 [0.0281]
28553
11443 [0.1025]
10001 [0.0594] 10830 [0.0987]
reflections [Rint
]
Goodness-of-fit (on F2) 1.177
1.003
0.1001, 0.2149 0.0544, 0.1281
1.028
1.127
1.079
0.936
R [F, I > 2r(I)]
0.0424, 0.1132
0.0563, 0.1532
0.0304, 0.0800
0.0745, 0.1732
wR (F2, all data)
Table 3
˚
Selected bond lengths (A) for the complexes 1a–c, 2a–c and 4c [5]
Compound
Ru–N(2)
N(2)–C(2)
M–C(2)
M–C(1)
C(1)–N(1)
Ru–P(1)
Ru–P(2)
1a
1b
1c
2a
2b
2c
4ca
a
2.060(6)
2.069(3)
2.065(3)
2.037(3)
2.048(2)
2.049(6)
1.914(5)
1.151(9)
1.141(5)
1.150(5)
1.148(5)
1.149(2)
1.139(9)
1.131(8)
1.871(7)
2.002(4)
1.994(4)
1.854(4)
1.990(2)
1.992(7)
1.993(6)
1.885(8)
2.010(5)
1.991(4)
1.875(4)
2.004(2)
2.004(9)
1.990(8)
1.129(10)
1.133(6)
1.149(5)
1.148(5)
1.148(3)
1.170(10)
1.153(10)
2.317(2)
2.3333(19)
2.3221(11)
2.3231(9)
2.3127(9)
2.2927(5)
2.275(2)
2.3160(12)
2.3183(9)
2.2922(9)
2.2783(5)
2.293(2)
2.210(2)
2.202(2)
For Ru read Fe.
Table 4
Selected bond angles (ꢁ) for the complexes 1a–c, 2a–c and 4c [5]
Compound
Ru–N(2)–C(2)
N(2)–C(2)–M
M–C(1)–N(1)
C(1)–M–C(2)
C(2)–M–C(10)
P(1)–Ru–P(2)
1a
1b
1c
2a
2b
2c
4ca
a
169.1(6)
171.3(4)
171.0(3)
169.2(3)
175.82(14)
174.7(6)
168.5(5)
177.1(6)
178.4(4)
178.2(3)
172.5(3)
174.6(2)
175.9(6)
176.0(6)
178.7(8)
177.2(5)
177.6(4)
177.3(3)
177.8(2)
177.1(7)
178.8(7)
93.0(3)
88.5(2)
88.7(1)
88.0(2)
91.55(7)
91.3(3)
91.8(2)
87.0(3)
91.5(2)
91.3(1)
92.0(2)
88.45(7)
88.7(3)
88.2(2)
101.11(7)
100.74(4)
100.68(3)
82.65(3)
82.87(2)
82.81(7)
85.59(7)
For Ru read Fe.
the structure of 4c when compared with the ruthenium ana-
logues 1c and 2c arise from the smaller size of the iron cen-
tre, which results in shortening of the Fe–N and Fe–P
distances relative to the Ru–N and Ru–P. In summary,
the molecular structures of the compounds 1a–c, 2a–c
and 4c are similar. There is no variation in the C(2)–N(2)
bond lengths that might be expected to reflect variation
in back-bonding interactions, nor is there any change in
the C(2)–M bond lengths with variation of M0 (M0 = Ni,
Pd, Pt).
3.2. Electrochemistry
The great interest in the properties of compounds in which
redox active groups are separated by some bridging moiety
prompted an examination of the tetracyanometallate

3464
R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
in the supporting ligands. Unsurprisingly, the very elec-
tron-rich iron end-caps in 4c result in the lowest oxidation
potentials observed across the series.
A variety of solvation factors and ion-paring phenom-
ena can influence oxidation potentials, and this point has
recently been highlighted by Keene as a potential compli-
cation in the determination of ‘‘electronic coupling’’ on
the basis of electrochemical measurements alone [25].
The electrochemical responses of 1a–c and 2a–c were
recorded in two different solvents in an effort to distin-
guish through-bond from through-space effects. In gen-
eral, the difference between the two oxidation potentials
(DE1/2) is somewhat smaller in THF than CH2Cl2, with
the decreased separation of the oxidation potentials aris-
ing from the greater relative shift of the first oxidation
event to higher potentials. Thus, it would appear that
there is some contribution to the thermodynamic stability
of the one-electron oxidation products arising from solva-
tion factors. The redox behaviour of compounds 1c and
4c have already been described by Vahrenkamp [5], with
two successive oxidations being observed at +0.89 and
+0.99 V and +0.36 V and +0.47 V, respectively, relative
to Ag|AgCl.
Fig. 1. A plot of a molecule of [{Ru(dppe)Cp}2{l-Pd(CN)4}] (2b)
showing the atom labelling scheme, which is representative of that used
in the series 1a–c, 2a–c.
bridged complexes 1a–c, 2a–c and 4c by cyclic and differen-
tial pulse voltammetry (Table 5).
For compounds 1a–c, 2a–c and 4c, two oxidation events
were observed at a platinum electrode. Whilst these events
were fully chemically and electrochemically reversible at
room temperature for compounds 2a–c and 4c, sub-ambi-
ent temperatures (ꢀ30 ꢁC) were required to improve the
chemical reversibility of these anodic processes for 1a–c.
The small difference between the half-wave potentials,
DE, necessitated that in some cases differential pulse vol-
tammetry be used to determine this parameter more accu-
rately. The electrochemical response of the bimetallic anion
3 was characterised by a single reversible oxidation at
0.74 V. The lower oxidation potentials of the compounds
2a–c relative to 1a–c can be attributed to the variations
The bimetallic complex 3 only exhibits a single anodic
event within the potential window accessible in CH2Cl2.
The most closely related trimetallic complex, 2c, is more
readily oxidised than 3, despite the anionic charge on the
latter. Taken together, these observations point to some
interaction between the Group 8 metal centres.
3.3. Mixed–valence complexes
The electronic spectra of the oxidised versions of the
complexes in this study would be expected to show IVCT
bands characteristic of the metal–metal interactions, as
demonstrated for 4c [5]. Furthermore band shape analysis
would be expected to reveal the extent of the communica-
tion between the metal centres and any trends that occur
as a function of the central metal would become apparent.
While the mixed-valence derivative [4c]+ and the dication
[4c]2+ have been prepared by stoichiometric oxidation of
4c with AgPF6, and the spectroscopic properties of these
species studied in detail [5], for convenience we chose to
use spectro-electrochemical methods to analyse the other
compounds in the series.
Table 5
Electrochemical data, given against SCE by correction against internal
ferrocene/ferrocinium or decamethylferrocene/decamethylferrocinium
couples
c
Compound
E1/2(1) (V)
E1/2(2) (V)
DE1/2 (V)
Kc
1aa
1ab
1ba
1bb
1ca
1cb
2aa
2ab
2ba
2bb
2ca
2cb
3a
0.95
1.11
1.00
1.14
0.97
1.08
0.69
0.81
0.72
0.82
0.71
0.82
0.74
0.34
1.06
1.20
1.08
1.20
1.08
1.20
0.82
0.90
0.84
0.90
0.85
0.92
n/a
0.11
0.09
0.08
0.06
0.11
0.12
0.13
0.09
0.12
0.08
0.14
0.10
n/a
190
75
45
20
190
310
155
35
105
20
230
50
n/a
70
The UV–Vis–NIR spectra were collected from CH2Cl2
solutions approximately 0.1 mM in analyte and contain-
Table 6
UV–Vis spectroscopic data for compounds 2a–c, 3 and 4c [5]
Compound
ꢀmmax=cmꢀ1 and (e/molꢀ1 dm3 cmꢀ1
)
4ca
0.45
0.11
2a
2b
2c
3
32680 (12280); 25900 (3900)
31250 (3860); 27250 (2480)
31060 (14540); 26390 (3920)
31150 (2680); 26460 (1370)
a
CH2Cl2.
THF.
b
c
Kc = exp{DEoxF/RT}, where F/RT = 47.76 Vꢀ1 at 243 K (1a–c) and
38.92 Vꢀ1 at 298 K (2a–c).
4c
29850 (8673); 21740 (1360); 19690 (1120)

R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
3465
Table 7
NIR spectroscopic parameters and associated values from [2a]+, [2b]+, and [2c]+
mmax ðcmꢀ1
Þ
Dm1=2ðcalcÞ ðcmꢀ1
Þ
Dm1=2ðfoundÞ ðcmꢀ1
Þ
e (molꢀ1 dm3 cmꢀ1
)
r (A)
Vab
k
˚
ꢀ
ꢀ
ꢀ
Compound
[2a]+
[2b]+
[2c]+
[4c]+
6380
6780
5680
6410
3800
4000
3600
3560
4800
4000
5100
4720
730
500
530
600
10.00
10.35
10.34
10.00
300
230
250
270
5300
5800
4900
5520
ing 0.1 M [NBu4]BF4 as a supporting electrolyte. Whilst
on the CV timescales (<10 s) the anodic events from
the whole series were chemically reversible, the first oxi-
dation states of 1a and 1b were insufficiently reversible
on the longer timeframe required for bulk electrolysis
in the OTTLE cell for acquisition of meaningful spectra.
Therefore, in the interests of obtaining a complete data
set, only the Ru(dppe)Cp* derived complexes 2a–c were
studied in detail. However, even in these cases, only the
first oxidation product could be obtained with complete
reversibility and then only at sub-ambient temperatures
(ꢀ30 ꢁC).
The reorganisation energy associated with the electron-
transfer process, k, can be estimated from the electrochem-
ical and spectroscopic data, with
ꢀ
k ¼ mmax ꢀ 8064DE1=2
.
ð3Þ
The electronic parameters extracted from the spectroscopic
data show little sensitivity to the nature of the central me-
tal, and given the approximations made, the metal–metal
coupling or ‘‘communication’’ in these systems are identi-
cal, and, as noted by Vahrenkamp for related systems,
the energy of the IVCT band depends heavily on the reor-
ganisation energy of the electron-transfer event. The reor-
ganisation energies and coupling parameters are similar,
displaying less than 10% variation with the change in
Group 10 metal atom, suggesting that the nature of the
Group 8 metal containing fragment and the through-space
electron-transfer distance are predominant factors in the
electron-transfer process.
The neutral compounds 2a–c and 3 displayed two
absorption bands in the UV (35000–20000 cmꢀ1) region
of the spectrum, the lower energy band appearing as a
shoulder. Compound 4c displays an absorption band at
29850 cmꢀ1 as well as a broader absorption envelope with
apparent band maxima at 21740 and 19690 cmꢀ1
(Table 6).
4. Conclusion
Upon oxidation of 2a–c, these electronic absorption
bands moved to slightly lower energy, while in contrast oxi-
dation of 4c resulted in the reverse behaviour was seen.
Oxidation of the bimetallic complex 3 gave a new absorp-
The tetracyanometallates of the Group 10 metals are
convenient dianionic metalloligands, and readily give rise
to linear trimetallic complexes. Oxidation of the complexes
tion band at 20330 cmꢀ1
.
{Ru(dppe)Cp*}2{l-M(CN)4} affords Class
2 mixed-
In addition to the bands already described, an absorp-
tion band envelope was observed in the NIR region of
the spectra of the trimetallic cations [2a]+, [2b]+ and
[2c]+. The high energy side of this absorption envelope
overlapped the tails of bands from the higher energy end
of the visible spectrum. Critically, the NIR region of 3
remained transparent during the oxidation and therefore
the NIR absorptions observed in [2a]+, [2b]+ and [2c]+
can be confidently assigned to a genuine RuII/RuIII IVCT
transition.
valence complexes, which demonstrate electron-transfer
˚
over distances of ca. 10 A. A combination of electrochem-
ical and spectroelectrochemical studies reveal the energy of
the photoinduced electron-transfer process to be dependent
on the reorganisation energy of the terminal Group 8 metal
fragments and the through-space separation.
5. Supporting information
The NIR band in [2a]+, [2b]+, [2c]+ is well approximated
by a single Gaussian-shaped curve, which is broader than
that calculated from the Hush expression (Eq. (1)), suggest-
ing that each complex is well described a classical Class 2
mixed-valence complex
Crystallographic data for the structural analysis has
been deposited with the Cambridge Crystallographic Data
Centre, CCDC Nos. 280645–280650 for compounds 1a—c,
2a–c, respectively. Copies of this information may be
obtained free of charge from The Director CCDC, 12
Union Road, Cambridge, CB2 1EZ UK (fax (int code):
+44 1223 336 033 or e-mail: deposit@ccdc.cam.ac.uk or
www: http://www.ccdc.cam.ac.uk).
1=2
ꢀ
ꢀ
Dm1=2 ¼ ð2310mmax
Þ
.
ð1Þ
Assuming the crystallographically determined Ru–Ru sep-
aration approximates the electron-transfer distance, r,
modest coupling parameters, Vab, are calculated in the fol-
lowing equation (Table 7):
Acknowledgements
We thank the EPSRC and the University of Durham for
funding this work. R.L.C. received studentship support from
the Department of Chemistryꢀs Doctoral Training Account.
1=2
ꢀ
ꢀ
0:0205ðemmaxDm1=2
Þ
V ab
¼
.
ð2Þ
r

3466
R.L. Cordiner et al. / Inorganica Chimica Acta 359 (2006) 3459–3466
[7] (a) L.R. Falvello, M. Tomas, Chem. Commun. (1999) 273;
References
(b) J.Y. Yang, M.P. Shores, J.J. Sokol, J.R. Long, Inorg. Chem. 42
(2003) 1403;
(c) M.P. Shores, J.J. Sokol, J.R. Long, J. Am. Chem. Soc. 124 (2002)
124.
[1] A. Ludi, J. Chem. Ed. 58 (1981) 1013.
[2] H. Vahrenkamp, A. Geiß, G.N. Richardson, J. Chem. Soc., Dalton
Trans. (1997) 3643.
[3] (a) D.J. Darensbourg, W.-Z. Lee, M.J. Adams, D.L. Larkins, J.H.
Reibenspies, Inorg. Chem. 38 (1999) 1378;
[8] (a) R.W. Gable, B.F. Hoskins, R. Robson, Chem. Commun. (1990)
762;
(b) K.-M. Park, T. Iwamoto, J. Chem. Soc., Dalton Trans. (1993) 1875;
(c) J. Cernak, J. Skorsepa, K.A. Abboud, M.W. Meisel, M. Orendac,
A. Orendacova, A. Feher, Inorg. Chim. Acta 326 (2001) 3.
[9] (a) M.I. Bruce, K. Costuas, T. Davin, B.G. Ellis, J.F. Halet, C.
Lapinte, P.J. Low, M.E. Smith, B.W. Skelton, L. Toupet, A.H.
White, Organometallics 24 (2005) 3864;
(b) H. Oshino, M. Yamamoto, T. Ito, Inorg. Chem. 41 (2002)
5817;
(c) R.C. Millward, C.E. Madden, I. Sutherland, R.J. Mortimer, S.
Fletcher, F. Marken, Chem. Commun. (2001) 1994;
(d) W.E. Buschmann, J. Ensling, P. Gutlich, J.S. Miller, Chem. Eur.
¨
J. 5 (1999) 3019;
(e) S. Bordoni, L. Busetto, P. Colucci, A. Palazzi, V. Zanotti, J.
Organomet. Chem. 545–546 (1997) 117;
(f) N.G. Naumov, D.V. Soldatov, J.A. Ripmeester, S.B. Artemkina,
V.E. Fedorov, Chem. Commun. (2001) 571;
(b) M.I. Bruce, B.G. Ellis, P.J. Low, B.W. Skelton, A.H. White,
Organometallics 22 (2003) 3184;
(c) M.I. Bruce, P.J. Low, Adv. Organomet. Chem. 50 (2004) 179.
´
[10] (a) M.E. Smith, R.L. Cordiner, D. Albesa-Jove, D.S. Yufit, F. Hartl,
J.A.K. Howard, P.J. Low, Can. J. Chem. (in press);
(b) R.L. Cordiner, M.E. Smith, A.S. Batsanov, D. Albesa-Jove´, F.
Hartl, J.A.K. Howard, P.J. Low, Inorg. Chim. Acta (in press);
(c) R.L. Cordiner, D. Corcoran, D.S. Yufit, A.E. Goeta, J.A.K.
Howard, P.J. Low, Dalton Trans. (2003) 3541;
(d) M.I. Bruce, M.A. Buntine, K. Costuas, B.G. Ellis, J.F. Halet,
P.J. Low, B.W. Skelton, A.H. White, J. Organomet. Chem. 689
(2004) 3308.
(g) C. Ornelas, C. Gandum, J. Mesquita, J. Rodrigues, M.H. Garcia,
N. Lopes, M.P. Robalo, K. Na¨ttinen, K. Rissanen, Inorg. Chim.
Acta 358 (2005) 2482;
´
(h) M.J. Calhorda, M.G.B. Drew, V. Felix, L.P. Fonseca, C.A.
Gamelas, S.S.M.C. Godinho, I.S. Gonc¸alves, E. Hunstock, J.P.
Lopes, A.J. Parola, F. Pina, C.C. Romao, A.G. Santos, J. Organo-
˜
met. Chem. 632 (2001) 94;
(i) M.P. Shores, L.G. Beauvais, J.R. Long, J. Am. Chem. Soc. 121
(1999) 775;
(j) C.P. Berlinguette, A. Dragulescu-Andrasi, A. Sieber, J.R. Galan-
Mascaros, H.U. Gudel, C. Achim, K.R. Dunbar, J. Am. Chem. Soc.
126 (2004) 6222;
[11] M.I. Bruce, C. Hameister, A.G. Swincer, R.C. Wallis, Inorg. Synth.
28 (1990) 270.
[12] M.I. Bruce, B.G. Ellis, P.J. Low, B.W. Skelton, A.H. White,
Organometallics 22 (2003) 3184.
[13] M.J. Mays, P.L. Sears, J. Chem. Soc., Dalton Trans. 1873 (1973).
[14] N.G. Connelly, W.E. Geiger, Chem. Rev. 96 (1996) 877.
[15] C.M. Duff, G.A. Heath, Inorg. Chem. 30 (1991) 2528.
[16] J.-P. Launay, Chem. Soc. Rev. 30 (2001) 386.
(k) E.J. Schelter, A.V. Provirin, K.R. Dunbar, J. Am. Chem. Soc.
126 (2004) 15004.
[4] (a) S.-M. Kuang, P.E. Fanwick, R.A. Walton, Inorg. Chem. 41
(2001) 147;
[17] SAINT V6.22, Bruker AXS, Madison, WI, USA, 2001.
[18] SAINT V6.02a, Bruker AXS, Madison, WI, USA, 2000.
[19] SHELXTL, Version 5.10, Bruker AXS, Madison, WI, USA, 1997.
[20] A.L. Spek, J. Appl. Crystallogr. 36 (2003) 7.
[21] G.J. Baird, S.G. Davies, J. Organomet. Chem. 262 (1984) 215.
[22] G.J. Kubas, L.H. Jones, Inorg. Chem. 13 (1974) 2816.
[23] R.D. Shannon, Acta Crystallogr. Sect. A 32 (1976) 751.
(b) I. Muga, J.M. Gutierrez-Zorrila, P. Vitoria, P. Roman, F. Lloret,
Polyhedron 21 (2002) 2631;
(c) T. Sheng, H. Vahrenkamp, Eur. J. Inorg. Chem. (2004) 1198;
(d) M.L. Flay, H. Vahrenkamp, Z. Anorg. Alleg. Chem. 630 (2004)
2508.
[5] G.N. Richardson, U. Brand, H. Vahrenkamp, Inorg. Chem. 38
(1999) 3070.
[6] (a) G.N. Richardson, H. Vahrenkamp, J. Organomet. Chem. 597
(2000) 38;
´
[24] R.L. Cordiner, D. Albesa-Jove, R.L. Roberts, J.D. Farmer, H.
Puschmann, D. Corcoran, A.E. Goeta, J.A.K. Howard, P.J. Low, J.
Organomet. Chem. 690 (2005) 4908.
[25] D.M. Dꢀ Alessandro, F.R. Keene, Dalton Trans. (2004) 3950.
(b) M.-L. Flay, H. Vahrenkamp, Eur. J. Inorg. Chem. (2003)
1719.