Metal-metal charge transfer and solvatochromism in cyanomanganese carbonyl complexes of ruthenium and osmium

Article abstract of DOI:10.1039/b602334g

The complexes [(H₃N)₅Ru^II(-NC)Mn ^IL_x]²⁺, prepared from [Ru(OH ₂)(NH₃)₅]²⁺ and [Mn(CN)L _x] {L_x = trans-(CO)₂{P(OP

Full text of DOI:10.1039/b602334g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Metal–metal charge transfer and solvatochromism in cyanomanganese
carbonyl complexes of ruthenium and osmium†
Christopher J. Adams, Neil G. Connelly, Nicholas J. Goodwin, Owen D. Hayward, A. Guy Orpen and
Andrew J. Wood
Received 16th February 2006, Accepted 27th April 2006
First published as an Advance Article on the web 11th May 2006
DOI: 10.1039/b602334g
The complexes [(H3N)5RuII(l-NC)MnILx]2+, prepared from [Ru(OH2)(NH3)5]2+ and [Mn(CN)Lx] {Lx =
trans-(CO)2{P(OPh)3}(dppm); cis-(CO)2(PR3)(dppm), R = OEt or OPh; (PR3)(NO)(g-C5H4Me), R =
Ph or OPh}, undergo two sequential one-electron oxidations, the first at the ruthenium centre to give
[(H3N)5RuIII(l-NC)MnILx]3+; the osmium(III) analogues [(H3N)5OsIII(l-NC)MnILx]3+ were prepared
directly from [Os(NH3)5(O3SCF3)]2+ and [Mn(CN)Lx]. Cyclic voltammetry and electronic spectroscopy
show that the strong solvatochromism of the trications depends on the hydrogen-bond accepting
properties of the solvent. Extensive hydrogen bonding is also observed in the crystal structures of
[(H3N)5RuIII(l-NC)MnI(PPh3)(NO)(g-C5H4Me)][PF6]3·2Me2CO·1.5Et2O, [(H3N)5RuIII(l-
NC)MnI(CO)(dppm)2-trans][PF6]3·5Me2CO and [(H3N)5RuIII(l-NC)MnI(CO)2{P(OEt)3}(dppm)-
trans][PF6]3·4Me2CO, between the ammine groups (the H-bond donors) at the Ru(III) site and the
oxygen atoms of solvent molecules or the fluorine atoms of the [PF6]− counterions (the H-bond
acceptors).
Introduction
Results and discussion
Our investigations of polynuclear compounds in which a
cyanomanganese(I) carbonyl or nitrosyl group is N-bound to an-
other metal centre1–12 are part of a wider effort13 to understand the
structural, magnetic, electronic and optical properties of cyanide-
bridged metal species. In the course of our studies we noted14 that
certain compounds were solvatochromic, e.g. [X3Fe(l-NC)MnLx]
{X = Cl or Br; Lx = trans- or cis-(CO)2(PR3)(dppm), R = OEt or
OPh}. The present paper describes a detailed study of the synthe-
sis, structure, electrochemistry (including spectroelectrochemical
identification of redox products) and solvatochromic behaviour
of a second series of such compounds, namely [(H3N)5MIII(l-
NC)MnILx]3+ {M = Ru or Os; Lx = trans-(CO)2{P(OPh)3}(dppm);
cis-(CO)2(PR3)(dppm), R = OEt or OPh; (PR3)(NO)(g-C5H4Me),
R = Ph or OPh}. Our work is complementary to important studies
by Laidlaw et al.15–17 of [(g-C5R5)(Ph3P)2MꢀII(l-CN)MIII(NH3)5]3+
(M, Mꢀ = Ru or Os; R5 = H5, H4Me or Me5) and [{(CO)5Cr0(l-
CN)}n{RuIII(NH3)6−n}]3−n (n = 0–3)18 in providing further insight
into the electronic properties of mixed-valence complexes by
means of X-ray crystal structure determinations and the use
of carbonyl-containing fragments to enable IR spectroscopic
identification of the site of first oxidation in the heterobimetallic
precursors [(H3N)5MII(l-NC)MnILx]2+.
Synthesis and characterisation of cyanomanganese(I)
pentaammineruthenium(II) complexes
The reaction of [Ru(NH3)5(OH2)][PF6]2 with one equivalent of
trans-[Mn(CN)(CO)2{P(OPh)3}(dppm)] (trans-1), cis-[Mn(CN)-
(CO)2(PR3)(dppm)] (R = OPh, cis-1 or OEt, cis-2) or [Mn(CN)-
(PR3)(NO)(g-C5H4Me)] (R = Ph, 3 or OPh, 4) (Scheme 1)
in acetone gave orange or red solutions from which
[(H3N)5RuII(l-NC)MnILx][PF6]2 {Lx = trans-(CO)2{P(OPh)3}-
(dppm) 52+[PF6]2; Lx = cis-(CO)2(PR3)(dppm), R = OEt 62+[PF6]2
or OPh 72+[PF6]2; Lx = (PR3)(NO)(g-C5H4Me), R = Ph 82+[PF6]2
or OPh 92+[PF6]2} (Scheme 2) were isolated as air-sensitive yellow
or orange powders which dissolve in polar solvents such as
CH2Cl2, thf, MeCN and acetone to give air-sensitive solutions.
School of Chemistry, University of Bristol, Bristol, UK BS8 1TS. E-mail:
neil.connelly@bristol.ac.uk; Fax: +44 (0)117 929 0509; Tel: +44 (0)117
928 8162
† Dedicated to Professor Victor Riera, good friend and colleague, on his
70th birthday.
Scheme 1 P–P = Ph2PCH2PPh2 = dppm.
3584 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
complex suggesting that the increase in m(CN) expected on
N-bonding to the second metal (the kinematic effect21) is offset
by a decrease due to the relatively strong p-donating ability of
the RuII(NH3)5 group {cf. the 13 cm−1 reduction in m(CN) when
this group is attached to [Re(CN)(CO)3(phen)] in [(H3N)5Ru(l-
NC)Re(CO)3(phen)]2+ (phen = 1,10-phenanthroline)22}.
The cyclic voltammograms (CVs) of the dications 52+–92+ show
two oxidation waves in CH2Cl2. The first is always reversible
[(ip)red/(ip)ox = 1.0] and diffusion-controlled [(ip)ox/m1/2 is constant
at scan rates, m, between 50 and 500 mV s−1] with the potential
only slightly dependent on the ligand set at manganese (Table 2).
By contrast, the potential (and reversibility) of the second wave
is highly dependent on that ligand set. The first wave is therefore
assigned to the oxidation of Ru(II) to Ru(III) and the second to the
oxidation of Mn(I) to Mn(II).
For 52+, both oxidation waves are reversible suggesting that
the trication 53+ and the tetracation 54+ are stable on the cyclic
voltammetric time scale. The CVs of 82+ and 92+ are similar to that
of 52+ except that the second oxidation wave of the former appears
to be incompletely reversible, indicating that the tetracation 84+
is unstable on the CV time scale. The second oxidation wave for
92+ is at a more positive potential than that of 82+, consistent with
replacement of PPh3 by P(OPh)3, a better p-acceptor. The second
wave for 92+ shows a sharp ‘spike’ on the return sweep of the CV
suggesting that 94+ deposits as an insoluble conducting layer on
the electrode surface, redissolving on reduction back to 93+.
The salt 102+[PF6]2 is insoluble in CH2Cl2 and thf but its CV
in MeCN, at a glassy carbon electrode, shows two reversible oxi-
dation waves. The much lower oxidation potential for the second
wave (cf. 52+–92+) is due to the more electron-rich Mn(CO)(dppm)2
group. However, this wave is at a much more positive potential
(by over 0.4 V) than that for the uncoordinated cyanomanganese
complex trans-[Mn(CN)(CO)(dppm)2].
Scheme 2 Complexes [(H3N)5M(l-NC)MnLx]z+.
Attempts to prepare [(H3N)5Ru(l-NC)Mn(CO)2{P(OEt)3}-
(dppm)-trans][PF6]2 by a similar method were unsuccessful, IR
spectroscopy of the isolated yellow powder showing the presence
of significant quantities of the cis isomer 62+[PF6]2, presum-
ably formed by trans–cis isomerisation at the Mn(I) centre.19–20
The monocarbonyl complex [(H3N)5Ru(l-NC)Mn(CO)(dppm)2-
trans][PF6]2 102+[PF6]2 (solutions of which are more air-stable
than those of 52+[PF6]2–92+[PF6]2) was prepared as an or-
ange powder by reacting trans-[Mn(CN)(CO)(dppm)2] with
[Ru(NH3)5(OH2)][PF6]2.
The complexes 52+[PF6]2–102+[PF6]2 were characterised by ele-
mental analysis, IR spectroscopy (Table 1) and cyclic voltammetry
(Table 2). In all cases, m(CO) or m(NO) is shifted only minimally
(ca. 5 cm−1) to higher wavenumber when the cyanide complexes 1–
4 or trans-[Mn(CN)(CO)(dppm)2] are N-bound to the RuII(NH3)5
fragment. In each bimetallic complex m(CN) is almost unchanged
from that of the corresponding mononuclear manganese cyanide
The electrochemistry of the species containing cis-Mn(CO)2
fragments, 62+ and 72+, is complicated by redox-induced cis–trans
isomerisation at manganese, as observed19–20 for uncoordinated
cis-1 and cis-2 and for all cyanide-bridged complexes containing
Table 1 Analytical and IR spectroscopic data for [(H3N)5M(l-NC)MnLx]z+
Analysisb (%)
IRc/cm−1
Compounda
Colour
Yield (%)
C
H
N
m(CN)
m(CO)d
m(NO)
52+
Yellow
Yellow
Yellow
Orange
Orange
Orange
Green
Blue-green
Blue
Purple
Brown
Red-brown
Blue-green
Purple
48
46
63
79
57
78
77
75
85
63
83
73
64
67
51
41.8 (42.3)
32.4 (32.4)e
42.4 (42.3)
32.3 (32.3)
30.9 (30.8)
46.3 (46.1)
38.2 (38.0)
30.9 (31.2)
37.4 (38.0)
27.6 (28.0)
27.1 (26.8)
41.4 (41.7)
34.6 (34.8)h
29.6 (29.7)e
40.9 (41.3)
4.0 (4.0)
4.5 (4.2)
3.9 (4.0)
3.9 (4.0)
3.8 (3.8)
4.4 (4.3)
3.6 (3.6)
4.1 (4.0)
3.6 (3.6)
3.3 (3.5)
3.2 (3.3)
4.4 (4.0)
4.5 (4.7)
3.4 (3.6)
3.6 (3.7)
5.9 (6.4)
2086 vw
1930, (2012)
1959, 1901
1972, 1915
—
—
—
—
1736
1760
—
—
—
f
62+
6.1 (6.3)
6.1 (6.4)
10.3 (10.6)
9.7 (10.0)
6.6 (6.2)
6.0 (5.8)
6.6 (6.4)
5.9 (5.8)
8.8 (9.1)
8.4 (8.7)
5.6 (5.6)
5.7 (5.7)
5.3 (5.3)
5.3 (5.3)
72+
2093 vw
82+
2099 w
92+
—
f
102+
53+
1865g
f
2046 w
2045 w
2048 w
2060 mw
2071 mw
2000 mg
2035 w
2074 w
2025 m
1998 m, 1950
1958, 1928
1971 ms, 1941
—
63+
73+
—
83+
1756
1777
—
—
—
93+
—
103+
113+
123+
133+
1939 ms, 1890g
1990 m, 1948
1966, 1919
1886
Green
—
a Isolated as [PF6]− (M = Ru) or [CF3SO3]− (M = Os) salts. b Calculated values in parentheses. c Strong (s) absorptions in CH2Cl2 unless stated otherwise,
vw = very weak, m = medium. d Very weak A-mode given in parentheses. e Analysed as a 1 : 2 CH2Cl2 solvate. f Not observed. g In MeCN. h Analysed as
a 1 : 3 acetone solvate.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3584–3596 | 3585
©

View Article Online
Table 2 Electrochemical data for [(H3N)5M(l-NC)MnLx]3+a
E◦ꢀb/V
oxidant to produce [(H3N)5RuIII(l-NC)MnI(CO)(dppm)2-trans]3+
103+. Accordingly, the addition of [Fe(g-C5H5)2][PF6] to a solution
of [(H3N)5RuII(l-NC)MnI(CO)(dppm)2-trans][PF6]2 102+[PF6]2 in
acetone gave an immediate colour change from orange to darker
orange-brown; filtration of the reaction mixture, reduction of the
volume of the filtrate in vacuo and then layering it with diethyl ether
gave red-brown crystals of [(H3N)5RuIII(l-NC)MnI(CO)(dppm)2-
trans][PF6]3·5Me2CO, 103+[PF6]3·5Me2CO, at −10 ◦C. When these
crystals were carefully washed with diethyl ether, loss of solvent
from the lattice resulted in an air-sensitive red-brown powder
the elemental analysis of which was consistent with unsolvated
103+[PF6]3 (Table 1).
Compound
Solvent
M(II)/M(III)
Mn(I)/Mn(II)
DEc/V
53+
CH2Cl2
MeCN
0.34
0.17
1.12
0.78
0.87
0.87
0.85
0.91e
0.97e
—
1.04
0.16d
0.26
1.03d
thf
1.11
1.34(I) (0.93)
63+
CH2Cl2
MeCN
thf
CH2Cl2
MeCN
thf
CH2Cl2
MeCN
thf
CH2Cl2
MeCN
thf
0.37
0.24
1.27(I) (0.85)
f
0.26
73+
0.42d
0.28d
0.33
1.52(I) (1.13)d
1.07e
1.15e
—
1.47(I) (1.04)d
f
The trications 53+–93+ were isolated as their [PF6]− salts after
oxidation of the corresponding dications prepared in situ. Thus,
treatment of stirred solutions of [Ru(NH3)5(OH2)][PF6]2 in acetone
with one equivalent of trans-1, cis-1, cis-2, 3 or 4, also in acetone,
gave orange or red solutions of [(H3N)5RuII(l-NC)MnILx][PF6]2,
as above. After ten minutes, one equivalent of the oxidising
agent [N2C6H4F-p][PF6] was added to the mixture, resulting in
colour changes to deep blue-green for the reactions starting
from trans-1, cis-1 and cis-2, and deep purple-brown and deep
green for those involving 3 and 4 respectively. The complexes
[(H3N)5RuIII(l-NC)MnILx][PF6]3 {Lx = trans-(CO)2(PR3)(dppm),
R = OPh 53+[PF6]3; Lx = cis-(CO)2(PR3)(dppm), R = OEt 63+[PF6]3
or OPh 73+[PF6]3; Lx = (PR3)(NO)(g-C5H4Me), R = Ph 83+[PF6]3
or OPh 93+[PF6]3} were then isolated as air-stable solids.
83+
0.39 (0.18)h,i
0.19
1.16g,i
0.84e
0.93
—
1.12(I) (0.44)
f
0.22
0.37
0.22
0.25
93+
1.37i
1.00
1.16e
—
1.42(I) (0.92)
f
103+
113+
123+
MeCN
thf
dmf
CH2Cl2
thf
dmf
0.20
0.28
0.50
0.57
0.40
0.94
0.90
0.81
1.26(I) (0.83)
0.30
0.29
0.44
0.64
0.72
0.95
1.65e
1.70e
—
−0.04
0.30
0.18
−0.14
CH2Cl2 −0.42
MeCN
thf
−0.52
−0.49
−0.73
1.21(I) (0.75)
f
f
dmf
—
133+
CH2Cl2 −0.45d,j
0.52d
0.43
0.51
0.41
0.98
0.92
0.96
1.11
Attempts
to
prepare
[(H3N)5RuIII(l-NC)MnI(CO)2-
MeCN
thf
−0.50
−0.45
−0.71
{P(OEt)3}(dppm)-trans][PF6]3 in the same way gave a blue-
green powder which IR spectroscopy showed to contain a large
proportion of the isomer with the cis-MnI(CO)2 fragment, i.e.
63+. However, after layering the concentrated acetone reaction
mixture with diethyl ether, blue-green crystals of [(H3N)5Ru(l-
NC)Mn(CO)2{P(OEt)3}(dppm)-trans][PF6]3, which analysed (C,
H and N) as the acetone solvate 113+[PF6]3·3Me2CO, were isolated
after four weeks at −10 ◦C; the IR spectrum of the crystals showed
no sign of the cis isomer. These crystals are air-stable and do not
undergo trans-to-cis isomerisation on storage at −10 ◦C. However,
when crushed and placed in vacuo for 24 hours at room
temperature the IR spectrum of the resulting blue-green powder
showed the presence of a significant quantity of the cis isomer 63+.
Heating a mixture of [Os(NH3)5(O3SCF3)][CF3SO3]2 with cis-2
or trans-[Mn(CN)(CO)(dppm)2] under reflux in thf yielded
the air-stable solids [(H3N)5Os(l-NC)Mn(CO)2{P(OEt)3}(dppm)-
cis][CF3SO3]3 123+[CF3SO3]3 and [(H3N)5Os(l-NC)Mn(CO)-
(dppm)2-trans][CF3SO3]3 133+[CF3SO3]3; the latter is mildly light-
sensitive in the solid state and was therefore stored in the dark.
Attempts to form [(H3N)5Os(l-NC)Mn(CO)2{P(OEt)3}(dppm)-
trans][CF3SO3]3 from [Os(NH3)5(O3SCF3)][CF3SO3]2 and trans-2
were only partly successful; substitution of the coordinated triflate
anion required the use of high temperatures for an extended
period, conditions which led to trans-to-cis isomerisation at
manganese. The product obtained from this reaction was therefore
an isomeric mixture.
dmf
a Data reported from measurements on the tricationic complexes but
potentials identical (to within 10 mV◦) for measurements also carried out
ꢀ
on dicationic analogues (see text). b
E
for a reversible wave at a platinum
disc electrode at a scan rate of 200 mV s−1, unless stated otherwise. For an
irreversible (I) oxidation wave, the peak potential, (Ep)ox, is accompanied
by E◦ꢀ (in parentheses) for the reversible product wave. c DE is the difference
between the potentials for the Mn(I)/Mn(II) and M(II)/M(III) couples. d At
a glassy carbon electrode. e DE values calculated as described in text. f Not
observed; wave obscured by base electrolyte curve. g At a gold electrode.
h For the trication, the reduction wave is broad; (Ep)red is accompanied by
(Ep)ox in parentheses. i ‘Spike’ observed on reverse scan, with (ip)red/(ip)ox
ꢁ
1; unable to measure reversibility. j Presence of internal potential reference
compound [Fe(g-C5Me5)2] necessary for observation of reversible and
reproducible behaviour.
such units. Thus, the CV of 72+ in CH2Cl2 shows a reversible
oxidation wave at 0.42 V (at a scan rate of 200 mV s−1) and a
partially reversible oxidation wave at 1.52 V {E◦ is estimated as
ꢀ
1.46 V based on the average of the peak potentials, (Ep)ox and
(Ep)red}. The second wave becomes less reversible with decreasing
scan rates and at 50 mV s−1 the oxidation process is fully irreversible
[(Ep)ox = 1.51 V]. It is coupled to a product wave which, when
the CV of 72+ is scanned from −0.2 to 1.7 to 0.9 and finally to
1.4 V, is reversible and centred at 1.13 V, i.e. the potential of the
Mn(I)/Mn(II) couple of 52+. Thus, oxidative isomerisation to 54+
occurs only after formation of 74+.
The salts 53+[PF6]3–93+[PF6]3, 113+[PF6]3 and 123+[CF3SO3]3 are
readily soluble in polar solvents such as CH2Cl2, CHCl3, thf,
MeCN, acetone and dmf to give mildly air-sensitive solutions.
The monocarbonyl analogues 103+[PF6]3 and 133+[CF3SO3]3 are
rather less soluble, and in solution the former decomposes rapidly
in air.
Synthesis and characterisation of cyanomanganese(I)
pentaammineruthenium(III) and pentaammineosmium(III)
complexes
On the basis of the potentials noted above, the chemical oxidation
of the dication 102+ should require only a mild one-electron
3586 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
The salts of the trications have been characterised by elemental
analysis, IR spectroscopy (Table 1), cyclic voltammetry (Table 2)
and, in the cases of 83+[PF6]3, 103+[PF6]3 and 113+[PF6]3, by X-ray
crystallography (see below).
In contrast to the IR spectra of 52+–92+, the numbers and relative
intensities of the carbonyl and cyanide bands in the spectra of the
ruthenium complexes 53+–113+ are unusual (Table 1) and surpris-
ingly solvent dependent. For example, Fig. 1 shows the spectra
of [(H3N)5RuII(l-NC)MnI(CO)2{P(OPh)3}(dppm)-trans]2+, 52+, in
CH2Cl2 (2086vw, 2012vw, 1930s cm−1) and of [(H3N)5RuII(l-
NC)MnI(CO)2{P(OPh)3}(dppm)-trans]3+, 53+, in CH2Cl2 (2046w,
1998m, 1950s cm−1) and MeCN (2045w, 1989w, 1946s cm−1).
The two lower energy absorptions in the CH2Cl2 spectrum
of 53+ (at 1998 and 1950 cm−1) are both assigned to carbonyl
bands. However, their relative intensities (medium and strong
respectively) contrast markedly with those of 52+ (and for the
free ligand trans-[Mn(CN)(CO)2{P(OPh)3}(dppm)] and most of
its N-bound complexes) where the strong band for the trans-
Mn(CO)2 group is accompanied by a very weak absorption for
the IR-inactive A-mode.
It is noteworthy that the only other cyanomanganese carbonyl
complexes to show similarly unusual IR spectra are [X3FeIII(l-
NC)MnILx] [X = Cl or Br; Lx = trans- or cis-(CO)2(PR3)(dppm),
R = OEt or OPh]14 in which the FeIIIX3 fragments are also
strongly electron withdrawing. Thus, it is likely that N-bonding
of a strongly withdrawing group lowers m(CN) to bring it much
closer in energy to those of the carbonyl bands. Coupling between
the cyanide and carbonyl stretching vibrations then results in
the observed increased intensity for the absorption from the
IR-inactive A-mode of the trans-dicarbonyl group. {One might
visualise the Mn(CN)(CO)2 unit as being equivalent to a mer-
tricarbonyl fragment for which three bands of similar intensity
would be expected.} The osmium-containing trication 123+, with
the OsIII(NH3)5 fragment a weaker acceptor than RuIII(NH3)5,
displays carbonyl bands with the more usual relative intensities
(Table 1) {and with m(CN), at 2074 cm−1, shifted less to lower
wavenumber than in the ruthenium analogue 63+ (2045 cm−1)},
consistent with this explanation.
The X-ray structures of 83+[PF6]3·2Me2CO·1.5Et2O,
103+[PF6]3·5Me2CO and 113+[PF6]3·4Me2CO
Crystals of the solvated salts 83+[PF6]3·2Me2CO·1.5Et2O,
103+[PF6]3·5Me2CO and 113+[PF6]3·4Me2CO were grown by allow-
ing diethyl ether to diffuse into concentrated acetone solutions of
the complexes at −10 ◦C. The molecular structures of the trications
83+, 103+ and 113+ are shown in Fig. 2–4 respectively and selected
bond lengths and angles are given in Tables 3–5. In addition, the
crystal structure of 103+[PF6]3·5Me2CO is shown in Fig. 5.
Fig. 1 IR spectra of (a) 52+ in CH2Cl2, (b) 53+ in CH2Cl2, and (c) 53+ in
MeCN.
Of the three bands observed for 53+, that at highest energy
is assigned to the cyanide stretch; m(CN) is invariably higher
than m(CO) in complexes of cyanomanganese carbonyls. As is
expected when oxidation of a cyanide-bridged complex occurs
at the N-bound metal site, m(CN) for 53+ is observed at lower
wavenumber than for 52+. Moreover, m(CN) in 53+ is much lower in
energy (by ca. 40 cm−1) than in the mononuclear precursor trans-
[Mn(CN)(CO)2{P(OPh)3}(dppm)], suggesting that the N-bound
pentaammineruthenium(III) centre is strongly electron withdraw-
ing {cf. the N-bound Ru(II) centre in 52+}, as also deduced for
species such as [(g-C5H5)(Ph3P)2RuII(l-CN)RuIII(NH3)5]3+ where
a similar large shift in m(CN) to lower wavenumber, relative to that
of [Ru(CN)(PPh3)2(g-C5H5)], was observed.15
Fig. 2 Structure of the trication [(H3N)5RuIII(l-NC)MnI(PPh3)(NO)-
(g-C5H4Me)]3+ 3+. Hydrogen atoms and some disorder in the C5H4Me
8
ligand are omitted for clarity.
Dalton Trans., 2006, 3584–3596 | 3587
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Table 3 Selected bond lengths (A) and angles (◦) for
2Me2CO·1.5Et2O
8
3+[PF6]3·
˚
Mn(1)–C(1)
C(1)–N(1)
Ru(1)–N(1)
Mn(1)–N(7)
1.913(6)
1.162(7)
2.003(4)
1.646(5)
N(7)–O(1)
1.183(6)
2.293(2)
2.103(4)
2.111(4)
Mn(1)–P(1)
Ru(1)–N(4)
Ru(1)–Ncis(ave)
Mn(1)–C(1)–N(1)
C(1)–N(1)–Ru(1)
Mn(1)–N(7)–O(1)
N(7)–Mn(1)–C(1)
N(7)–Mn(1)–P(1)
C(1)–Mn(1)–P(1)
177.2(5)
177.0(4)
173.6(4)
99.6(2)
96.4(2)
87.6(2)
N(1)–Ru(1)–N(2)
N(1)–Ru(1)–N(3)
N(1)–Ru(1)–N(4)
N(1)–Ru(1)–N(5)
N(1)–Ru(1)–N(6)
88.5(2)
92.1(2)
178.2(2)
90.3(2)
89.1(2)
Fig.
4
Structure of the trication [(H3N)5RuIII(l-NC)MnI(CO)2-
{P(OEt)3}(dppm)-trans]3+ 113+. Hydrogen atoms are omitted for clarity.
◦
Table 5 Selected bond lengths (A) and angles ( ) for 113+[PF6]3·4Me2CO
˚
Mn(1)–C(1)
Mn(1)–C(2)
Mn(1)–C(3)
C(1)–N(1)
Ru(1)–N(1)
Ru(1)–N(6)
1.914(10)
1.797(13)
1.805(12)
1.198(10)
1.978(8)
2.104(7)
Ru(1)–Ncis (ave)
Mn(1)–P(1)
Mn(1)–P(2)
Mn(1)–P(3)
C(2)–O(1)
2.076(9)
2.211(3)
2.308(3)
2.285(3)
1.184(12)
1.184(11)
C(3)–O(2)
Mn(1)–C(1)–N(1)
Ru(1)–N(1)–C(1)
N(1)–Ru(1)–N(2)
N(1)–Ru(1)–N(3)
N(1)–Ru(1)–N(4)
N(1)–Ru(1)–N(5)
N(1)–Ru(1)–N(6)
C(1)–Mn(1)–P(1)
179.0(9)
179.7(8)
89.6(3)
92.3(3)
88.7(3)
89.6(3)
179.0(3)
93.0(3)
C(1)–Mn(1)–P(2)
C(1)–Mn(1)–P(3)
C(1)–Mn(1)–C(2)
C(1)–Mn(1)–C(3)
C(2)–Mn(1)–C(3)
P(1)–Mn(1)–P(2)
P(1)–Mn(1)–P(3)
P(2)–Mn(1)–P(3)
169.3(3)
96.4(3)
88.2(4)
88.9(4)
177.1(4)
97.7(1)
170.5(1)
72.9(1)
Fig.
3
Structure of the trication [(H3N)5RuIII(l-NC)MnI(CO)-
(dppm)2-trans]3+ 103+. Hydrogen atoms are omitted for clarity.
All of the trications are octahedral at ruthenium with deviations
from the ideal value of 90◦ limited to 3◦. The Ru–N bond
lengths are in the range 2.08 to 2.11 A and the bridging
˚
cyanide units are nearly linear, with Mn–C–N and C–N–Ru
angles of less than 74◦. In 103+ the carbonyl is trans to the
cyanide bridge and the four phosphorus atoms of the two dppm
ligands occupy the equatorial positions; in 113+ the CO ligands
are mutually trans, with the cyanide bridge trans to a phosphorus
atom of the bidentate dppm ligand.
Previous work has established that Mn–P bond lengths are
diagnostic of the oxidation state of manganese in both mononu-
clear and polynuclear octahedral cyanomanganese compounds;
Mn(II)–P bond lengths are substantially longer than Mn(I)–P bond
lengths.10,23–25 A comparison of the structures of 103+ and 113+ with
literature data established that both contain Mn(I).
The manganese centre in 83+ is significantly distorted from
pseudo-tetrahedral geometry, with angles between the monoden-
tate ligands closer to 90◦ than 109.5◦. Comparison of the structure
of 83+ with that of [Mn(CNBPh3){P(OPh)3}(NO)(g-C5H4Me)]26
again suggests a Mn(I) oxidation state. Thus, the assignment of
manganese(I) centres to 83+, 103+ and 113+ indicates Ru(III)Mn(I)
core oxidation states for each complex, as deduced from the
CV studies on the corresponding dications (i.e. ruthenium-based
oxidation is the first step).
angles between 175.7◦ and 179.4◦. As in previous structures
4,7–10,14,23–24
containing the N-bound units (NC)Mn(CO)(dppm)2
and (NC)Mn(CO)2{P(OEt)3}(dppm),6–7,23 the geometry around
manganese is essentially octahedral in both 103+ and 113+; the
largest angular distortions from regular geometry are due to the
small bite of the dppm chelate which leads to Pdppm–Mn–Pdppm
Table 4 Selected bond lengths (A) and angles (◦) for 103+[PF6]3·
˚
5Me2CO
Mn(1)–C(1)
Mn(1)–C(2)
C(1)–N(1)
Ru(1)–N(1)
Ru(1)–N(3)
Ru(1)–Ncis(ave)
1.937(7)
1.790(7)
1.182(7)
1.998(6)
2.106(5)
2.111(5)
C(2)–O(1)
1.156(7)
2.302(3)
2.285(3)
2.300(3)
2.295(3)
Mn(1)–P(1)
Mn(1)–P(2)
Mn(1)–P(3)
Mn(1)–P(4)
Mn(1)–C(1)–N(1)
Ru(1)–N(1)–C(1)
C(1)–Mn(1)–C(2)
Mn(1)–C(2)–O(1)
P(1)–Mn(1)–P(2)
P(1)–Mn(1)–P(3)
P(2)–Mn(1)–P(4)
176.6(6)
175.4(5)
174.9(3)
178.0(7)
73.6(1)
106.3(1)
107.2(1)
P(3)–Mn(1)–P(4)
N(1)–Ru(1)–N(2)
N(1)–Ru(1)–N(3)
N(1)–Ru(1)–N(4)
N(1)–Ru(1)–N(5)
N(1)–Ru(1)–N(6)
72.9(1)
90.7(2)
179.1(2)
88.2(2)
88.9(2)
92.8(2)
One of the main features common to the crystal structures of
83+, 103+ and 113+ is the extensive hydrogen bonding involving
3588 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Fig. 5 Crystal structure of 103+[PF6]3·5Me2CO. Hydrogen atoms are omitted for clarity. Dotted lines indicate the interactions between the NH3 groups
and acetone molecules or [PF6]− counterions.
the ammine groups (the H-bond donors) at the Ru(III) site and
the oxygen atoms of solvent molecules (acetone or diethyl ether)
in the crystal lattice, or the fluorine atoms of the [PF6]− counterions
(the H-bond acceptors). Some of the H · · · O and H · · · F distances
orange suspension of 102+[PF6]2 in CH2Cl2 with 2.5 equivalents
of [NO][PF6] under argon, and with light excluded, gave a purple
solution from which a purple solid was isolated. The IR spectrum
of the solid showed a cyanide band at 2092 cm−1 and one carbonyl
band at 1956 cm−1 in CH2Cl2; the increase in energy of the
carbonyl band (of ca. 90 cm−1) with respect to that of the dication
102+ (1865 cm−1 in MeCN) is consistent with a carbonyl ligand
bound to Mn(II), i.e. as in [(H3N)5RuIII(l-NC)MnII(CO)(dppm)2-
trans][PF6]4 104+[PF6]4. However, the purple solid (for which
an acceptable elemental analysis could not be obtained) was
extremely air-sensitive and slowly decomposed even when stored
under argon at −10 ◦C.
˚
are less than 2 A, corresponding to strong hydrogen bonds (cf.
the sum of the van der Waals radii of hydrogen and oxygen, or
hydrogen and fluorine, both of which can be approximated as
˚
2.75 A).
In the case of 103+ the hydrogen bonding network orders
the structure such that distinct channels contain the acetone
solvent molecules and the [PF6]− anions (Fig. 5). The hydrophilic
pentaammineruthenium(III) centres are orientated towards the
channels while the hydrophobic manganese carbonyl fragments
are disposed away from the channels. Thus, a network of
hydrophilic channels is interspersed with an equal number of
hydrophobic channels. Because of this ordering, the structure has a
degree of anisotropy which may account for the red-green dichroic
nature of the crystals.
Further evidence for the formation of 104+ was obtained by
an IR spectroelectroch◦emical study of the monocarbonyl dication
102+, in MeCN at −40 C. After electrochemical generation of the
trication 103+, which reaches completion at 0.4 V, further oxidation
at a potential more positive than that of the Mn(I)/Mn(II) couple
yielded new bands at 1958 and 2084 cm−1, in good agreement
with the IR spectrum observed on chemical oxidation of 103+ in
CH2Cl2. Lowering the potential led to the reformation of 103+ and
then 102+, showing the trication and tetracation to be stable at
−40 ◦C.
The electrochemistry of the trications 53+–133+
As expected from the electrochemistry of the dicationic complexes,
each of the CVs of the trications 53+–133+ shows one reduction
wave, associated with the Ru(II)/Ru(III) or Os(II)/Os(III) couple,
and one oxidation wave associated with the Mn(I)/Mn(II) couple
(Table 2). Treatment of the ruthenium trications 53+–103+ with the
mild reducing agent [Fe(g-C5Me5)2] regenerates the Ru(II)Mn(I)
dications 52+–102+ but similar reactions with the Os(III) complexes
123+ and 133+ did not result in reduction, consistent with the much
more negative potentials (by ca. 0.6–0.8 V) for the Os(II)/Os(III)
couple.
A similar IR spectroelectrochemical◦study of the reduction of
113+ at −0.2 V, in CH2Cl2 or thf at −40 C, resulted in the collapse
of the three bands observed for the trication to give a very weak
cyanide band and a single carbonyl band [m(CO) = 1916 cm−1 in
CH2Cl2] consistent with the formation of 112+ (which could not
be made directly from [Ru(NH3)5(OH2)][PF6]2 and trans-2 or by
the chemical reduction of 113+ – see above); reoxidation at 0.4 V
regenerated 113+.
Applying
a potential more positive than that of the
Mn(I)/Mn(II) couple of 113+ in CH2Cl2 (but not in thf in which
the tetracation decomposes) resulted in a single carbonyl band
at 2010 cm−1 (cf. 1993 cm−1 for free trans-1+), and returning the
potential to 0.4 V led to the reformation of the three absorptions
characteristic of 113+, again implying the formation of 114+. (These
The reversibility of the oxidation waves for the trans-dicarbonyl
trications 53+ and 113+ and for the monocarbonyl 103+ suggested
that the tetracations 54+, 104+ and 114+ might also be synthet-
ically accessible using a moderately strong oxidant. Attempts
to prepare 54+ and 114+ were unsuccessful but treatment of an
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3584–3596 | 3589
©

View Article Online
results also confirm that IR spectra with unusual relative intensities
for the carbonyl and cyanide bands are confined to those of the
mixed-valence trication 113+.)
123+, where the oxidation wave for the Mn(I)/Mn(II) couple was
irreversible because of redox-induced cis–trans isomerisation, and
the cases of 83+ in CH2Cl2 and 93+ in MeCN, where the oxidation
waves were incompletely defined, values of DE were estimated.
The variation in DE implies that the energy barrier to electron
transfer from manganese(I) to ruthenium(III) or osmium(III),
In CH2Cl2, 133+ showed only a poorly resolved reduction wave
for the Os(II)/Os(III) couple at either a Pt or glassy carbon working
electrode. However, in the presence of a small amount of [Fe(g-
C5Me5)2], added as an internal calibrant, fully reversible and
reproducible CVs were obtained, possibly because of mediation
of the heterogeneous electron transfer to the Pt electrode.
The electrochemical studies reported above serve to identify the
redox processses undergone and the potentials of the various cou-
ples present {i.e. Mn(I)/Mn(II), Ru(II)/Ru(III) and Os(II)/Os(III)}.
However, voltammetric studies of the trications in a wider range
of solvents provided insight into the extent of MMCT (metal-
to-metal charge transfer), complementary to the results from the
UV/visible spectra described below.
E
MMCT, in the trications is higher in dmf than CH2Cl2, for example.
Moreover, DE in CH2Cl2 for the Mn(I)/Mn(II) and Os(II)/Os(III)
couples of 123+ (1.65 V) and 133+ (0.98 V) is greater than for the
Mn(I)/Mn(II) and Ru(II)/Ru(III) couples of the analogues 63+ (0.91
V) and 103+ (ca. 0.30 V) because of the significantly more negative
potentials for the Os(II)/Os(III) couples. Hence MMCT occurs at
a higher energy for osmium than ruthenium complexes.
Overall, plots of DE vs. EMMCT are linear (e.g. for both CH2Cl2
and MeCN, R2 = 0.91). However, attempts to quantify the extent
of electronic interaction between the two metal sites should be
treated with caution given the substantial medium effects (e.g.
solvent and base electrolyte salt) on DE.28–29 A more quantitative
approach has therefore been taken which is based on an extensive
study of electronic spectra in a wide range of solvents.
The CVs of 113+ in CH2Cl2, thf and dmf (Fig. 6, Table 2)
show that the potentials for both the oxidation and the re-
duction waves are solvent dependent (referenced against [Fe(g-
C5H5)(g-C5H4COMe)] as an internal potential standard). Both
the Mn(I)/Mn(II) and Ru(II)/Ru(III) couples are shifted to more
negative potentials as the solvent Gutmann donor number (DN)
increases.27 However, the effect of the solvent on the Ru(II)/Ru(III)
couple is much greater, shifting to a more negative potential by
ca. 400 mV from CH2Cl2 to dmf, cf. a shift of only 90 mV for the
Mn(I)/Mn(II) couple. Thus, DE (Table 2), the difference between
the Mn(I)/Mn(II) and Ru(II)/Ru(III) couples, for 113+ is greater in
dmf than in CH2Cl2.
Optical properties of cyanomanganese(I)
pentaammineruthenium(III) and pentaammineosmium(III)
complexes
The bimetallic M(III)Mn(I) (M = Ru or Os) species 53+–133+
are strongly solvatochromic, a characteristic of Robin–Day Class
II mixed-valence compounds and stemming from a metal-to-
metal charge transfer (MMCT) band in the UV-visible region
(Table 6).
Qualitatively, the band shapes and relative energies of the
absorption spectra of the trications are similar regardless of the sol-
vent. In CH2Cl2, the spectra of the cis-dicarbonyl trications 63+, 73+
and 123+ are the simplest, showing a single broad band at 843 nm
(e = 2370 dm3 mol−1 cm−1), 780 nm (e = 2130 dm3 mol−1 cm−1)
and 531 nm (e = 1980 dm3 mol−1 cm−1) respectively. The trans-
dicarbonyl trications 53+ and 113+ show lower energy absorptions
at 1134 nm (e = 4730 dm3 mol−1 cm−1) and 1338 nm (e =
7760 dm3 mol−1 cm−1), which are significantly stronger than those
of the cis-dicarbonyl analogues, and have shoulders towards lower
wavelength. Fig. 7(a) shows the superimposed spectra of the cis-
and trans-dicarbonyl complexes 53+ and 73+, indicating that the
shoulders on the absorptions for the trans-dicarbonyl complexes
are at a similar energy to the single absorption bands found for
their cis-dicarbonyl analogues. (However, other characterisation
methods described above, namely cyclic voltammetry and IR
spectroscopy, imply these trans complexes are not contaminated
by their cis isomers.)
In CH2Cl2, 83+ and 93+ show two distinct absorptions [Fig. 7(b)].
In each case, a band in the visible region is accompanied by a
stronger band at lower energy, in the near-IR region. For example,
83+ shows absorptions at 501 nm (e = 930 dm3 mol−1 cm−1) and
1098 nm (e = 3640 dm3 mol−1 cm−1). In MeCN, the monocarbonyl
103+ shows a single broad asymmetric absorption at 1453 nm (e =
1360 dm3 mol−1 cm−1), Fig. 7(c). For the analogous osmium species
133+ this asymmetric peak is resolved into an absorption band at
746 nm (e = 1940 dm3 mol−1 cm−1) in CH2Cl2, with a shoulder at
higher energy.
Fig. 6 Cyclic voltammograms of 113+ in (a) CH2Cl2, (b) thf, and (c) dmf.
A similar dependence on solvent of the Ru(II)/Ru(III) redox
potential is observed for 53+ though the oxidation wave for the
Mn(I)/Mn(II) couple could not always be observed (for example,
the solvent window in thf is limited to ca. 1 V) and DE could
not be calculated. For the cis-dicarbonyl trications 63+, 73+ and
3590 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
to describe solvent polarisability, and the magnitude of c4 is related
to c1. (Indeed, c4 is more properly c1d, where d is either zero or
a small and usually negative number, depending on the nature of
the solvatochromic process being described.32)
The magnitudes of the coefficients c1–c4 reflect the relative
importance of their respective parameters in solvatochromism,
which in turn sheds light on which solvent properties stabilise or
destabilise the ground state relative to the MMCT excited state.
Values of c1–c5 were obtained for 63+, 123+ and 133+ by measuring
their UV-visible spectra in a wide variety of solvents (solvent
parameters were taken from ref. 31 and 32), and using multilinear
regression analysis to fit the results to the Kamlet–Taft expression
(eqn (1)). Variables which were not statistically valid were omitted
stepwise from the analysis.
Not all solvents proved suitable because of problems of com-
pound solubility or reactivity, the air-sensitivity of solutions or,
in the case of some protic solvents, self-association. The final set
of solvents used in the analysis of each of the three compounds
studied was decided upon by studying correlation plots of EMMCT
for pairs of the three compounds, by the observation of sample
solution reactivity or air-sensitivity, and in some instances by the
removal of outliers. In general, alcohols proved largely unsuitable,
but more particular details are given below for each complex.
The final form of the Kamlet–Taft expression obtained for
63+ is shown in eqn (2), where data for all non-halogenated
alcohols (reactive) and aniline (reactive, with poor correlation
plots) were excluded from the regression analysis. Equations
displaying statistically stronger correlations could be obtained
by excluding different solvents, but it was unclear whether such
exclusions were judicious or arbitrary. At any rate, parameter b
was the only statistically significant variable influencing EMMCT
;
using slightly different solvent sets in the analysis did not change
this, nor did it produce large changes in the value of c3.
EMMCT (× 10−3 cm−1) = 6.19b + 11.24
(R2 = 0.90, F = 144.8)
(2)
(3)
(4)
EMMCT (× 10−3 cm−1) = 3.29b − 0.73a + 18.56
(R2 = 0.83, F = 39.8)
EMMCT (× 10−3 cm−1) = 3.80b + 12.61
Fig. 7 UV-visible-NIR spectra of (a) 53+ and 73+ in CH2Cl2, (b) 83+ and
93+ in CH2Cl2, and (c) 103+ in MeCN.
(R2 = 0.81, F = 53.7)
The Kamlet–Taft expressions for 123+ and 133+ are shown in
eqn (3) and (4) respectively. Eqn (3) (in which the statistical
significance of a appears marginal, but see below) is also based
on the exclusion of non-halogenated alcohols and aniline from
the solvent set on the basis of their reactivity; diethylamine was
omitted from the analysis because of low compound solubility.
Eqn (4) is based on the exclusion of non-halogenated alcohols,
glacial acetic acid, thf and aniline, all of which were either reactive
or formed air-sensitive solutions. As for 63+, b was the only
significant solvent parameter for 123+ and 133+, regardless of the
solvent set used.
On the basis of the UV/visible spectroscopic studies on 63+, 123+
and 133+, the nature of the solvatochromism in the trications, with
M(III)Mn(I) (M = Ru or Os) cores, can be explained. In all three
cases the positive values of c3 show that the greater the hydrogen
bond accepting ability of the solvent, the higher the energy of
MMCT. The MIII(NH3)5 fragment in the ground state is relatively
In order to determine which solvent properties most strongly
influence solvatochromic behaviour, a more detailed investigation
was made of the UV-visible spectra of 63+, 123+ and 133+ (Table 7).
Previous studies of solvatochromism30 have shown that the
effects of different solvent–solute interactions on EMMCT are
additive, i.e. solvent properties such as polarity and propensity for
hydrogen bonding can be added in a linear relationship to predict
the energy of the UV-visible absorption band. The Kamlet–Taft
expression31 (eqn (1)) is one such relationship:
EMMCT = c1p* + c2a + c3b + c4d + c5
(1)
where p* provides a measure of the ability of the solvent to stabilize
a solute dipole, and is nearly proportional to solvent dipole
moment, a is a measure of solvent hydrogen bond donor ability,
and b is a measure of solvent hydrogen bond acceptor ability. The d
term is a correction factor sometimes required in order adequately
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3584–3596 | 3591
©

View Article Online
Table 6 UV-visible spectroscopic data, and calculated values of J and a2, for [(H3N)5M(l-NC)MnLx]3+
Compound
Solvent
kmax/nm
E/cm−1
e/dm3 mol−1 cm−1
m1/2/cm−1
J/cm−1
a2 (%)
53+a
CH2Cl2
MeCN
thf
1134
963
946
574
740
843
746
725
629
780
685
689
562
501
1098
472
940
476
944
784d
476
883
430
760
431
751
634d
1453
578f
1364
1012
531
524
499
477
746
742
712
647
8820
10390
10580
17430
13510
11870
13400
13790
15900
12830
14590
14525
17810
19945
9110
21210
10640
21010
10600
12755d
21005
11330
23255
13155
23180
13310
15770d
6885
4730
2350
2510
920
1130
2370
1790
1810
1310
2130
1510
1730
1080
930
3640
890
2250
920
2650
1350
880
2570
890
3980
5360
4400b
—
1640
1460
1380b
—
3.48
1.98
1.71b
—
dmf
5630b
4580
4750
4800
4910
4630
4860
4750
5650
—
1180b
1450
1360
1400
1290
1430
1320
1390
1330
—
0.77b
1.50
1.03
1.03
0.66
1.25
0.82
0.92
0.56
—
63+a
73+a
83+c
CH2Cl2
MeCN
thf
dmf
CH2Cl2
MeCN
thf
dmf
CH2Cl2
3060
—
3380
—
3210
3710
—
3630
—
3820
—
3780
5870b
2490
1290
—
1150
—
1210
1020
—
1310
—
1100
—
1180
1140b
610
2.01
—
1.17
—
1.32
0.64
—
1.35
—
0.71
—
0.79
0.52b
0.79
MeCN
thf
dmf
CH2Cl2
93+c
MeCN
thf
1480
860
1700
860
dmf
MeCN
thf
103+e
123+h
133+h
1360
17300f
7330
9890
18820
19100
20020
20950
13410
13480
14050
15450
—
—
—
—
g
g
g
g
dmf
810
1980
2030
1780
1680
1940
1720
840
5600
5130
5030
6280
5440
2770
3210
2990
3660
880
1760
1780
1900
1760
1080
1100
760
0.79
0.88
0.87
0.91
0.71
0.65
0.67
0.29
0.57
CH2Cl2
MeCN
thf
dmf
CH2Cl2
MeCN
thf
dmf
1470
1160
a
2
Metal–metal separation, r, defined as 5.093 A in calculations of J and a . b Bandwidth at half height estimated as in text. c Metal–metal separation, r,
˚
defined as 5.078 A in calculations of J and a . d Higher energy band observed as shoulder at edge of solvent window; kmax not obtained. e Metal–metal
2
˚
separation, r, defined as 5.118 A in calculations of J and a . f Absorption for decomposition product. g Values not obtained because of decomposition.
2
˚
h
˚
Metal–metal separation, r, defined as 5.1 A.
electron-poor when compared with the MII(NH3)5 fragment of
the photoexcited state [i.e. after electron transfer from Mn(I) to
M(III)], leading to stronger electrostatic interactions in the ground
state between the N–H bonds and hydrogen bond accepting groups
in the solvent. Thus, the greater the interaction with a hydrogen
bond acceptor (larger b), the greater the stabilization of the ground
state relative to the excited state by solvent interactions, leading in
turn to higher EMMCT
.
Previous studies15 of solvatochromic pentaammine metal com-
plexes such as [(g-C5R5)(Ph3P)2MꢀII(l-CN)MIII(NH3)5]3+ (M, Mꢀ =
Ru or Os; R5 = H5, H4Me or Me5) have shown a similar dependence
of EMMCT on hydrogen bond accepting ability, with linear plots
against the Gutmann donor number.27 Similar plots for 63+, 123+
and 133+ (Fig. 8) also show positive correlations (R2 = 0.92, 0.78
and 0.89 respectively), consistent with the results of the Kamlet–
Taft analysis. Moreover, as described above, extensive hydrogen-
bonding involving the ammine ligands is observed in the solid state
structures of 83+[PF6]3·2Me2CO·1.5Et2O, 103+[PF6]3·5Me2CO and
113+[PF6]3·4Me2CO.
Fig. 8 Plots of EMMCT vs. Gutmann donor number for 63+, 123+ and 133+
.
3592 | Dalton Trans., 2006, 3584–3596
This journal is The Royal Society of Chemistry 2006
©

View Article Online
Table 7 Solvent parameters and UV-visible spectroscopic data for 63+, 123+ and 133+
kmax/nm
a
b
p*
d
DNa
63+
123+
133+
Acetone
Water
Nitromethane
Methanol
Chloroform
Ethyl acetate
dmf
Toluene
thf
Propan-2-ol
Acetonitrile
Methyl formate
CH2Cl2
0.08
1.17
0.22
0.98
0.20
0.00
0.00
0.00
0.00
0.76
0.19
0.00
0.13
0.00
0.00
0.00
0.00
0.05
0.03
0.26
1.12
0.84
1.51
1.96
0.00
0.43
0.47
0.06
0.66
0.10
0.45
0.69
0.11
0.55
0.84
0.40
0.37
0.10
0.76
0.76
0.30
0.37
0.05
0.70
0.50
0.45
0.90
0.00
0.00
0.47
0.71
1.09
0.85
0.60
0.58
0.55
0.88
0.54
0.58
0.48
0.75
0.62
0.82
1.00
0.88
1.01
0.90
0.62
0.24
0.73
0.64
0.52
0.73
0.65
0.27
0.0
0.0
0.0
0.0
0.5
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.5
0.0
0.0
1.0
1.0
0.5
0.0
1.0
0.0
0.0
0.5
0.5
0.0
17.0
18.0
2.7
30.0
4.0
17.1
26.6
0.1
20.0
36.0
14.1
711.6
683.9
843.7
715.8
826.4
710.1
628.8
796.4
725.4
730.4
746.2
723.6
842.5
613.0
629.4
845.3
763.5
803.2
505.0
501.4
542.8
511.8
523.6
497.0
477.3
509.2
499.4
513.9
523.6
502.6
531.4
469.4
478.4
535.0
522.2
516.8
430.0
515.5
504.3
516.4
577.1
606.0
707.0
781.2
742.6
732.2
689.0
647.1
780.0
711.6
732.8
741.8
699.7
745.8
624.6
654.6
770.6
737.6
738.9
1.0
29.8
27.8
4.4
dmso
dma
Nitrobenzene
Benzonitrile
Bromoform
Diethylamine
Aniline
Glacial acetic acid
Propan-1-ol
2,2,2-Trifluoroethanol
1,1,1,3,3,3-Hexafluoropropan-2-ol
Diethyl ether
11.9
50.0
35.0
20.0
568.5
708.6
728.0
915.0
946.0
754.4
737.0
834.6
19.2
a DN = Gutmann donor number.
If the small negative value (−0.73) of the a coefficient (c2)
in eqn (3) is real, it may result from the influence of solvents
displaying both hydrogen bond donor and acceptor properties.
For such solvents, self-association might act partially to diminish
the interactions between the solvent and the N–H groups of the
solute.
The UV-visible spectroscopic data can also be used to calculate
values for the parameters J (eqn (5)) and a2 (eqn (6), unrelated
to the solvent parameter a discussed above) which measure the
degree of electronic communication between the metal centres in
the dinuclear mixed valence complexes.
Conclusions
The complexes [Mn(CN)Lx] {Lx = trans-(CO)2{P(OPh)3}(dppm);
cis-(CO)2(PR3)(dppm), R = OEt or OPh; (PR3)(NO)(g-C5H4Me),
R = Ph or OPh} react with [Ru(OH2)(NH3)5]2+ to give
[(H3N)5RuII(l-NC)MnILx]2+ which undergo two sequential one-
electron oxidations, the first at the ruthenium centre to give
[(H3N)5RuIII(l-NC)MnILx]3+. These trications, and their os-
mium(III) analogues [(H3N)5OsIII(l-NC)MnILx]3+, are strongly
solvatochromic, the energy of the metal-to-metal charge transfer
band, EMMCT, varying with the ligand set, Lx, at manganese and
the identity of the second metal (M = Ru or Os).
J = 2.05 × 10−2[e × m1/2 × E]1/2 × r−1 (cm−1)
(5)
(6)
Multilinear regression analysis of the electronic spectra of
[(H3N)5RuIII(l-NC)MnI(CO)2{P(OEt)3}(dppm)][PF6]3, [(H3N)5-
Os(l-NC)Mn(CO)2{P(OEt)3}(dppm)-cis][CF3SO3]3 and [(H3N)5-
Os(l-NC)Mn(CO)(dppm)2-trans][CF3SO3]3 in a wide range of
solvents showed EMMCT to depend almost exclusively on the
hydrogen-bond accepting properties of the solvent. This is
consistent with the extensive hydrogen bonding observed between
the ammine groups (the H-bond donors) at the Ru(III) site and the
oxygen atoms of solvent molecules or the fluorine atoms of the
[PF6]− counterions (the H-bond acceptors) in the crystal
structures of [(H3N)5RuIII(l-NC)MnI(PPh3)(NO)(g-C5H4Me)]-
[PF6]3·2Me2CO·1.5Et2O, [(H3N)5RuIII(l-NC)MnI(CO)(dppm)2-
a2 = 4.24 × 10−4[(e × m1/2)/(E × r2)] (%)
Here, e is the extinction coefficient for the MMCT band in
dm3 mol−1 cm−1, m1/2 is the width of the band at half-height in cm−1,
E is the energy of the band maximum in cm−1 and r is the metal–
3+
3+
˚
metal separation in A. [For the Ru(III)Mn(I) complexes 8 , 10
and 113+, r was taken from the X-ray structures; in the absence
of suitable crystals of analogous osmium complexes the Os–Mn
˚
distance was taken as 5.1 A, i.e. the average of the Ru · · · Mn
distance for 83+, 103+ and 113+ (given that the ionic radii of second
and third row transition metals are similar).]
trans][PF6]3·5Me2CO
and
[(H3N)5RuIII(l-NC)MnI(CO)2{P-
The values of J and a2 for 53+–93+ (Table 6) fall in the
ranges expected for Robin–Day Class II compounds (typically
between several hundred and a few thousand cm−1, and 0.5–5%
respectively), i.e. there is appreciable electronic communication
between metal centres, but not to the degree where there is
complete delocalisation of charge.
(OEt)3}(dppm)-trans]][PF6]3·4Me2CO.
Experimental
The preparation and purification of the complexes described
was carried out under an atmosphere of dry nitrogen unless
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3584–3596 | 3593
©

View Article Online
82+[PF6]2 or OPh 92+[PF6]2} were prepared similarly. The salt
52+[PF6]2 was further purified by dissolution in the minimum
volume of acetone and dropwise addition to stirred diethyl ether
as described above. The salt 62+[PF6]2 was purified by dissolution
in CH2Cl2 and filtration through Celite before reduction of the
solvent volume to ca. 5 cm3 in vacuo and dropwise addition of
the solution to stirred diethyl ether (100 cm3). For 92+[PF6]2 the
extraction step required a larger volume of CH2Cl2 (40 cm3) and
the product was further purified by dissolution in CH2Cl2 and
dropwise addition to stirred diethyl ether. In the preparation of
82+[PF6]2 the extraction step required MeCN (30 cm3) instead of
CH2Cl2.
otherwise stated, using dried and deoxygenated solvents purified
either by distillation or by using Anhydrous Engineering double
alumina or alumina/copper catalyst drying columns. Reactions
were monitored by IR spectroscopy where necessary. Reactions
involving [Ru(NH3)5(OH2)][PF6]2 were carried out under argon;
the heterodinuclear products are light-sensitive in solution, so
that light was excluded during their synthesis. The compounds
trans- and cis-[Mn(CN)(CO)2(PR3)(dppm)] (R = OEt1 and
OPh33), trans-[Mn(CN)(CO)(dppm)2],2 [Mn(CN)(PR3)(NO)(g-
C5H4Me)] (R = Ph26,34 and OPh26), [Ru(NH3)5(OH2)][PF6]2,35
[Os(NH3)5(O3SCF3)][CF3SO3]2,36 [N2C6H4F-p][PF6],37–38 [Fe(g-
C5H5)2][PF6] and [Fe(g-C5H5)(g-C5H4COMe)][BF4]38 were pre-
pared by published methods or variations thereof. The salts
[NO][PF6] and Tl[PF6] were purchased from Fluorochem and
Strem Chemicals respectively.
[(H3N)5Ru(l-NC)Mn(CO)(dppm)2-trans][PF6]2 102+[PF6]2. Addi-
tion of trans-[Mn(CN)(CO)(dppm)2] (400 mg, 0.456 mmol) in
CH2Cl2 (25 cm3) to [Ru(NH3)5(OH2)][PF6]2 (225 mg, 0.456 mmol)
in acetone (100 cm3) gave an orange solution which was stirred for
15 min before the solvent was removed in vacuo. The residue was
extracted into MeCN (40 cm3) and filtered through Celite. The
volume of the filtrate was reduced to ca. 5 cm3 before diethyl ether
(60 cm3) was added to give an orange solid. The air-stable solid
was collected, washed with diethyl ether (3 × 20 cm3) and then
dried in vacuo for 24 h, yield 483 mg (78%).
IR spectra were recorded on a Nicolet 5ZDX FT spectrometer
and UV-visible spectra on a Perkin Elmer Lambda 19 UV/VIS
spectrometer. Electrochemical studies were carried out using an
EG & G model 273A potentiostat linked to a computer using EG &
G Model 270 Research Electrochemistry software in conjunction
with a three-electrode cell. The auxiliary electrode was a platinum
wire and the working electrode a platinum disc (1.6 or 2.0 mm
diameter) or glassy carbon disc (3.0 mm diameter). The reference
was an aqueous saturated calomel electrode separated from the
test solution by a fine porosity frit and an agar bridge saturated
with KCl. Solutions were 5 × 10−4 or 1 × 10−3 mol dm−3 in
[(H3N)5Ru(l-NC)Mn(CO)2{P(OPh)3}(dppm)-cis][PF6]3 73+[PF6]3.
Addition of cis-[Mn(CN)(CO)2{P(OPh)3}(dppm)] (300 mg,
0.361 mmol) in acetone (30 cm3) to a stirred solution of
[Ru(NH3)5(OH2)][PF6]2 (178 mg, 0.361 mmol) in acetone (30 cm3)
gave an orange solution. After 10 min [N2C6H4F-p][PF6] (97 mg,
0.361 mmol) was added to the mixture to give an immediate colour
change to deep blue. The reaction mixture was stirred for a further
1 h before it was evaporated to dryness in vacuo. The residue was
extracted into CH2Cl2 (40 cm3) and filtered through Celite. The
filtrate was reduced in volume to ca. 10 cm3 and diethyl ether was
added to give a blue powder which was washed with diethyl ether
(3 × 30 cm3). The powder was redissolved in the minimum volume
of acetone (less than 10 cm3) and the solution was added dropwise
to stirred diethyl ether (125 cm3) to precipitate the product. The
air-stable blue solid was washed with diethyl ether (3 × 30 cm3)
and dried in vacuo for 24 h, yield 447 mg (85%).
the test compound and 0.1 mol dm−3 in [NBun ][PF6] as the
4
supporting electrolyte. Under the conditions used, E◦ for the
ꢀ
one-electron oxidation of [Fe(g-C5H4COMe)2], [Fe(g-C5H5)(g-
C5H4COMe)], [Fe(g-C5H5)2] and [Fe(g-C5Me5)2], and for the one-
electron reduction [Co(g-C5H5)2]+, added to the test solutions
as internal calibrants, are 0.97, 0.74, 0.47, 0.08 and −0.87 V
respectively in CH2Cl2; the E◦ values for [Fe(g-C5H5)2] in thf
ꢀ
and MeCN are 0.54 and 0.39 V respectively; the E◦ values for
ꢀ
[Fe(g-C5H5)(g-C5H4COMe)] in thf and MeCN are 0.77 and 0.64 V
respectively. IR spectroelectrochemical measurements were made
using an EG & G model 273A potentiostat in conjunction with a
three-electrode system linked to the Bruker IFS25 spectrometer.
Microanalyses were carried out by the staff of the Microanalysis
Service of the School of Chemistry, University of Bristol.
The complexes [(H3N)5Ru(l-NC)MnLx][PF6]3 {Lx = trans-
(CO)2{P(OPh)3}(dppm) 53+[PF6]3, cis-(CO)2{P(OEt)3}(dppm)
63+[PF6]3; Lx = (PR3)(NO)(g-C5H4Me), R = Ph 83+[PF6]3 or OPh
93+[PF6]3} were prepared similarly. In the case of 53+[PF6]3 and
63+[PF6]3 the extraction step was effected with CHCl3. For 93+[PF6]3
the slightly oily product was vigorously stirred with diethyl ether
(40 cm3) for 2 h to give a dark brown powder.
Syntheses
[(H3N)5Ru(l-NC)Mn(CO)2{P(OPh)3}(dppm)-cis][PF6]2 72+[PF6]2.
Addition of cis-[Mn(CN)(CO)2{P(OPh)3}(dppm)] (200 mg,
0.241 mmol) in acetone (30 cm3) to [Ru(NH3)5(OH2)][PF6]2
(119 mg, 0.241 mmol) in acetone (10 cm3), under argon and in
the absence of light, gave an orange solution which was stirred for
20 min and then evaporated to dryness in vacuo. The residue was
extracted into CH2Cl2 (10 cm3) and filtered through Celite and
the filtrate was evaporated to dryness in vacuo. The orange residue
was redissolved in the minimum volume of acetone (ca. 5 cm3) and
added dropwise to stirred diethyl ether (125 cm3) to precipitate the
product as a pale yellow solid. The air-sensitive solid was collected,
washed with diethyl ether (3 × 20 cm3) and dried in vacuo for 24 h,
yield 198 mg (63%).
[(H3N)5Ru(l-NC)Mn(CO)(dppm)2-trans][PF6]3 103+[PF6]3. To
a
stirred solution of [(H3N)5Ru(l-NC)Mn(CO)(dppm)2-
trans][PF6]2 (300 mg, 0.200 mmol) in acetone (30 cm3) was added
[Fe(g-C5H5)2][PF6] (66 mg, 0.200 mmol) to give an immediate
colour change from orange to orange-brown. The reaction
mixture was stirred for 10 min before the solvent volume was
reduced to ca. 10 cm3 and the mixture filtered. The concentrated
acetone solution was layered with diethyl ether and stored at
−10 ◦C for 3 days. The resulting dark red-brown crystals were
washed with diethyl ether (3 × 20 cm3) to give an air-sensitive
red-brown powder, yield 219 mg (73%).
The air-sensitive complexes [(H3N)5Ru(l-NC)MnLx][PF6]2
{Lx
=
trans-(CO)2{P(OPh)3}(dppm) 52+[PF6]2, cis-(CO)2-
{P(OEt)3}(dppm) 62+[PF6]2, (PR3)(NO)(g-C5H4Me), R = Ph
3594 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Table 8 Crystal and refinement data for [(H3N)5Ru(l-NC)Mn(PPh3)(NO)(g-C5H4Me)][PF6]3·2Me2CO·1.5Et2O
8
3+[PF6]3·2Me2CO·1.5Et2O,
[(H3N)5Ru(l-NC)Mn(CO)(dppm)2-trans][PF6]3·5Me2CO 103+[PF6]3·5Me2CO and [(H3N)5Ru(l-NC)Mn(CO)2{P(OEt)3}(dppm)-trans][PF6]3·4Me2CO
113+[PF6]3·4Me2CO
Compound
8
3+[PF6]3·2Me2CO·1.5Et2O
103+[PF6]3·5Me2CO
113+[PF6]3·4Me2CO
Formula
M
C37H63F18MnN7O4.5P4Ru
1299.83
C67H89F18MnN6O6P7Ru
1789.24
C46H76F18MnN6O9P6Ru
1540.96
Crystal system
Space group (no.)
Triclinic
Triclinic
Monoclinic
P21/n (no. 14)
14.091(4)
31.216(5)
16.167(4)
90
93.48(3)
90
293(2)
7098(3)
¯
¯
P1 (no. 2)
P1 (no. 2)
˚
a/A
9.972(2)
13.579(2)
20.985(5)
81.27(1)
81.48(2)
76.76(1)
173(2)
2715.0(11)
2
0.735
11.596(8)
14.943(8)
25.237(15)
104.96(2)
100.72(4)
98.59(3)
173(2)
4060(4)
2
0.571
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
T/K
3
˚
U/A
Z
4
l/mm−1
0.622
Reflections collected
Independent reflections (Rint
27761
12195 (0.0602)
0.0656, 0.1488
21254
13903 (0.0807)
0.0710, 0.0997
36451
12454 (0.1689)
0.0883, 0.1893
)
Final R indices [I > 2r(I)]: R1,wR2
diffuse slowly into an acetone solution of the powder gave the
product as light-sensitive green microcrystals after 14 d at 4 C,
yield 82 mg (51%).
[(H3N)5Ru(l-NC)Mn(CO)2{P(OEt)3 }(dppm)-trans][PF6 ]3·
3Me2CO, 113+[PF6]3·3Me2CO. Addition of [Ru(NH3)5(OH2)]-
[PF6]2 (216 mg, 0.436 mmol) in acetone (25 cm3) to a stirred
solution of trans-[Mn(CN)(CO)2{P(OEt)3}(dppm)] (300 mg,
0.436 mmol) in acetone (25 cm3) gave an orange solution. After
10 min [N2C6H4F-p][PF6] (117 mg, 0.436 mmol) was added to
give an immediate colour change to blue. The reaction mixture
was stirred for a further 45 min before the solvent volume was
reduced to ca. 10 cm3 and the reaction mixture filtered. The
filtrate was layered with diethyl ether and stored at −10 ◦C for
4 weeks. The mother liquors were removed from the resulting
dark blue-green crystals which were washed with diethyl ether
(2 × 10 cm3) and then separated from a small amount of white
powder, yield 412 mg (64%).
◦
Structure determinations of [(H3N)5Ru(l-NC)Mn(PPh3)(NO)(g-
C5H4Me)][PF6]3·2Me2CO·1.5Et2O 83+[PF6]3·2Me2CO·1.5Et2O,
[(H3N)5Ru(l-NC)Mn(CO)(dppm)2-trans][PF6]3·5Me2CO
103+[PF6]3·5Me2CO and [(H3N)5Ru(l-NC)Mn(CO)2-
{P(OEt)3}(dppm)-trans][PF6]3·4Me2CO 113+[PF6]3·4Me2CO
Dark purple crystals of [(H3N)5Ru(l-NC)Mn(PPh3)(NO)(g-
C5H4Me)][PF6]3·2Me2CO·1.5Et2O 83+[PF6]3·2Me2CO·1.5Et2O,
red-green dichroic crystals of [(H3N)5Ru(l-NC)Mn(CO)(dppm)2-
trans][PF6]3·5Me2CO 103+[PF6]3·5Me2CO and dark blue crys-
tals of [(H3N)5Ru(l-NC)Mn(CO)2{P(OEt)3}(dppm)-trans]-
[PF6]3·4Me2CO 113+[PF6]3·4Me2CO were grown by allowing
diethyl ether to diffuse into a concentrated acetone solution of
[(H3N)5Os(l-NC)Mn(CO)2{P(OEt)3 }(dppm)-cis][CF3SO3 ]3·
2CH2Cl2 123+[CF3SO3]3·2CH2Cl2. A solution of [Os(NH3)5-
(O3SCF3)][CF3SO3]2 (88 mg, 0.122 mmol) and cis-[Mn-
(CN)(CO)2{P(OEt)3}(dppm)] (84 mg, 0.122 mmol) in thf (20 cm3)
was heated under reflux for 48 h and then stirred for a further
48 h. The solution was filtered through Celite, and the solution
volume reduced in vacuo to ca. 5 cm3. Diethyl ether (40 cm3) was
slowly added to the stirred solution to precipitate a purple powder
which was washed with diethyl ether (2 × 20 cm3) and then dried
in vacuo. Further purification, by slow evaporation under vacuum
of a solution of the complex in CH2Cl2 (30 cm3) and n-hexane
(30 cm3), gave a purple powder which was washed with diethyl
ether (2 × 20 cm3) and dried in vacuo, yield 129 mg (67%).
[(H3N)5Os(l-NC)Mn(CO)(dppm)2-trans][CF3SO3]3 133+[CF3SO3]3.
A mixture of [Os(NH3)5(O3SCF3)][CF3SO3]2 (73 mg, 0.100 mmol)
and trans-[Mn(CN)(CO)(dppm)2] (89 mg, 0.101 mmol) in thf
(20 cm3) was heated under reflux for 24 h and then stirred for
another 24 h. The solution was then evaporated to dryness in
vacuo, and the residue extracted into CH2Cl2 (1 × 10 cm3 then 2 ×
5 cm3). The extract was filtered, n-hexane was added (25 cm3), and
the product was precipitated by slow evaporation in vacuo. The
green powder was washed with hexane (2 × 10 cm3), before drying
under vacuum in the absence of light. Allowing diethyl ether to
◦
the complex at −10 C. The crystals of 113+[PF6]3·4Me2CO were
very poorly diffracting, leading to a high value of Rint (0.17) and
ultimately to a relatively poor value of R1 (0.0883).
Many of the details of the structure analyses are listed in Table 8.
CCDC reference numbers 603279–603281.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b602334g
Acknowledgements
We thank the EPSRC for Studentships (to O.D.H. and A.J.W.)
and the New Zealand Foundation for Research, Science and
Technology for a Postdoctoral Fellowship (to N.J.G.).
References
1 G. A. Carriedo, N. G. Connelly, M. C. Crespo, I. C. Quarmby, V. Riera
and G. H. Worth, J. Chem. Soc., Dalton Trans., 1991, 315.
2 A. Christofides, N. G. Connelly, H. J. Lawson, A. C. Loyns, A. G.
Orpen, M. O. Simmonds and G. H. Worth, J. Chem. Soc., Dalton
Trans., 1991, 1595.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3584–3596 | 3595
©

View Article Online
3 G. A. Carriedo, N. G. Connelly, S. Alvarez, E. Perez-Carreno and S.
Garcia-Granda, Inorg. Chem., 1993, 32, 272.
4 F. L. Atkinson, N. C. Brown, N. G. Connelly, A. G. Orpen, A. L.
Rieger, P. H. Rieger and G. M. Rosair, J. Chem. Soc., Dalton Trans.,
1996, 1959.
5 M. Bardaji, N. C. Brown, A. Christofides and N. G. Connelly, J. Chem.
Soc., Dalton Trans., 1996, 2511.
17 G. U. Bublitz, W. M. Laidlaw, R. G. Denning and S. G. Boxer, J. Am.
Chem. Soc., 1998, 120, 6068.
18 W. M. Laidlaw and R. G. Denning, Inorg. Chim. Acta, 1996, 248, 51.
19 N. G. Connelly, K. A. Hassard, B. J. Dunne, A. G. Orpen, S. J. Raven,
G. A. Carriedo and V. Riera, J. Chem. Soc., Dalton Trans., 1988, 1623.
20 C. F. Hogan, A. M. Bond, A. K. Neufeld, N. G. Connelly and E.
Llamas-Rey, J. Phys. Chem. A, 2003, 107, 1274.
6 N. C. Brown, G. B. Carpenter, N. G. Connelly, J. G. Crossley, A. Martin,
A. G. Orpen, A. L. Rieger, P. H. Rieger and G. H. Worth, J. Chem.
Soc., Dalton Trans., 1996, 3977.
7 N. G. Connelly, G. R. Lewis, M. T. Moreno and A. G. Orpen, J. Chem.
Soc., Dalton Trans., 1998, 1905.
8 N. G. Connelly, O. M. Hicks, G. R. Lewis, M. T. Moreno and A. G.
Orpen, J. Chem. Soc., Dalton Trans., 1998, 1913.
21 D. A. Dows, A. Haim and W. K. Walmarth, J. Inorg. Nucl. Chem., 1961,
21, 33.
22 C. A. Bignozzi, R. Argazzi, J. R. Schoonover, K. C. Gordon, R. B.
Dyer and F. Scandola, Inorg. Chem., 1992, 31, 5260.
23 F. L. Atkinson, A. Christofides, N. G. Connelly, H. J. Lawson, A. C.
Loyns, A. G. Orpen, G. M. Rosair and G. H. Worth, J. Chem. Soc.,
Dalton Trans., 1993, 1441.
9 N. G. Connelly, O. M. Hicks, G. R. Lewis, A. G. Orpen and A. J. Wood,
J. Chem. Soc., Dalton Trans., 2000, 1637.
10 K. M. Anderson, N. G. Connelly, N. J. Goodwin, G. R. Lewis, M. T.
Moreno, A. G. Orpen and A. J. Wood, J. Chem. Soc., Dalton Trans.,
2001, 1421.
24 G. A. Carriedo, N. G. Connelly, E. Perez-Carreno, A. G. Orpen, A. L.
Rieger, P. H. Rieger, V. Riera and G. M. Rosair, J. Chem. Soc., Dalton
Trans., 1993, 3103.
25 D. Bellamy, N. C. Brown, N. G. Connelly and A. G. Orpen, J. Chem.
Soc., Dalton Trans., 1999, 3191.
11 K. M. Anderson, N. G. Connelly, E. Llamas-Rey, A. G. Orpen and
R. L. Paul, Chem. Commun., 2001, 1734.
12 C. J. Adams, K. M. Anderson, M. Bardaji, N. G. Connelly, N. J.
Goodwin, E. Llamas-Rey, A. G. Orpen and P. H. Rieger, Dalton Trans.,
2004, 683.
26 D. Bellamy, N. G. Connelly, O. M. Hicks and A. G. Orpen, J. Chem.
Soc., Dalton Trans., 1999, 3185.
27 V. Gutmann, Electrochim. Acta, 1976, 21, 661.
28 F. Barriere, N. Camire, W. E. Geiger, U. T. Mueller-Westerhoff and R.
Sanders, J. Am. Chem. Soc., 2002, 124, 7262.
13 See, for example: K. R. Dunbar and R. A. Heintz, Prog. Inorg. Chem.,
1997, 45, 283; H. Vahrenkamp, A. Geiss and G. N. Richardson, J. Chem.
Soc., Dalton Trans., 1997, 3643; C. C. Chang, B. Pfennig and A. B.
Bocarsly, Coord. Chem. Rev., 2000, 208, 33; M. L. Kuhlman and T. B.
Rauchfuss, Inorg. Chem., 2004, 43, 430; P. Albore´s, L. D. Slep, T.
Weyhermu¨ller and L. M. Baraldo, Inorg. Chem., 2004, 43, 6762; S.
Nishikiori, H. Yoshikawa, Y. Sano and T. Iwamoto, Acc. Chem. Res.,
2005, 38, 227; L. M. C. Beltran and J. R. Long, Acc. Chem. Res., 2005,
38, 325; E. Coronado, M. C. Gime´nez-Lo´pez, G. Levchenko, F. M.
Romero, V. Garc´ıa-Baonza, A. Milner and M. Paz-Pasternak, J. Am.
Chem. Soc., 2005, 127, 4580; C. P. Berlinguette, A. Dragulescu-Andrasi,
A. Sieber, H.-U. Gu¨del, C. Achim and K. R. Dunbar, J. Am. Chem.
Soc., 2005, 127, 6766.
14 N. G. Connelly, O. M. Hicks, G. R. Lewis, A. G. Orpen and A. J. Wood,
Chem. Commun., 1998, 517.
15 W. M. Laidlaw and R. G. Denning, J. Chem. Soc., Dalton Trans., 1994,
1987.
16 W. M. Laidlaw, R. G. Denning, T. Verbiest, E. Chauchard and A.
Persoons, Nature, 1993, 363, 58.
29 D. M. D’Alessandro and F. R. Keene, Dalton Trans., 2004, 3950.
30 E. S. Dodsworth, M. Hasegawa, M. Bridge and W. Linert, Comprehen-
sive Coordination Chemistry II, ed. J. A. McCleverty and T. J. Meyer,
Elsevier, Amsterdam, 2004, ch. 2.27.
31 M. J. Kamlet, J.-L. M. Abboud, M. H. Abraham and R. W. Taft, J. Org.
Chem., 1983, 48, 2877.
32 Y. Marcus, Chem. Soc. Rev., 1993, 22, 409.
33 G. A. Carriedo, M. C. Crespo, V. Riera, M. G. Sanchez, M. L. Valin,
D. Moreiras and X. Solans, J. Organomet. Chem., 1986, 302, 47.
34 D. L. Reger, D. J. Fauth and M. D. Dukes, J. Organomet. Chem., 1979,
170, 217.
35 D. E. Harrison, H. Taube and E. Weissberger, Science, 1968, 159, 320;
R. W. Callahan, G. M. Brown and T. J. Meyer, Inorg. Chem., 1975, 14,
1443; J. A. Baumann and T. J. Meyer, Inorg. Chem., 1980, 19, 345.
36 P. A. Lay, R. H. Magnuson and H. Taube, Inorg. Chem., 1989, 28,
3001; P. A. Lay, R. H. Magnuson and H. Taube, Inorg. Synth., 1986,
24, 271.
37 A. Roe, Org. React., 1949, 5, 193.
38 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877.
3596 | Dalton Trans., 2006, 3584–3596
This journal is
The Royal Society of Chemistry 2006
©