Bis(acetylacetonato)ruthenium(II) complexes containing bulky tertiary phosphines. Formation and redox behaviour of Ru(acac)2 (PR 3) (R = iPr, Cy) complexes with ethene, carbon monoxide, and bridging dinitrogen

Article abstract of DOI:10.1016/j.ica.2004.07.062

Reaction of cis-[Ru(acac)₂(η²-C₈H ₁₄)₂] (1) (acac = acetylacetonato) with two equivalents of PⁱPr₃ in THF at -25°C gives trans-[Ru(acac) ₂(PⁱPr₃)₂], trans-3, which rapidly isomerizes to cis-3 at room temperature. The poorly soluble complex [Ru(acac)₂(PCy₃)₂] (4), which is isolated similarly from cis-[Ru(acac)₂(η²-C₂H ₄)₂] (2) and PCy₃, appears to exist in the cis-configuration in solution according to NMR data, although an X-ray diffraction study of a single crystal shows the presence of trans-4. In benzene or toluene 2 reacts with PⁱPr₃ or PCy₃ to give exclusively cis-[Ru(acac)₂(η²-C₂H ₄)(L)] [L = PⁱPr₃ (5), PCy₃ (6)], whereas in THF species believed to be either square pyramidal [Ru(acac) ₂L], with apical L, or the corresponding THF adducts, can be detected by ³¹P NMR spectroscopy. Complexes 3-6 react with CO (1 bar) giving trans-[Ru(acac)₂(CO)(L)] [L = PⁱPr₃ (trans-8), PCy₃ (trans-9)], which are converted irreversibly into the cis-isomers in refluxing benzene. Complex 5 scavenges traces of dinitrogen from industrial grade dihydrogen giving a bridging dinitrogen complex, cis-[{Ru(acac)₂(PⁱPr₃)} ₂(μ- N₂)] (10). The structures of cis-3, trans-4, 5, 6 and 10 ? C₆H₁₄ have been determined by single-crystal X-ray diffraction. Complexes trans- and cis-3, 5, 6, cis-8, and trans- and cis-9 each show fully reversible one-electron oxidation by cyclic voltammetry in CH ₂Cl₂ at -50°C with E_1/2(Ru^3+/2+) values spanning -0.14 to +0.92 V (versus Ag/AgCl), whereas for the vinylidene complexes [Ru(acac)₂ (CCHR)(PⁱPr₃)] [R = SiMe₃ (11), Ph (12)] the process is irreversible at potentials of +0.75 and +0.62 V, respectively. The trend in potentials reflects the order of expected π-acceptor ability of the ligands: PⁱPr₃, PCy₃ 2H₄ < CCHR < CO. The UV-Vis spectrum of the thermally unstable, electrogenerated Ru^III-ethene cation 6⁺ has been observed at -50°C. Cyclic voltammetry of the μ-dinitrogen complex 10 shows two, fully reversible processes in CH ₂Cl₂ at -50°C at +0.30 and +0.90 V (versus Ag/AgCl) corresponding to the formation of 10⁺ (Ru^II,III) and 10²⁺ (Ru^III,III). The former, generated electrochemically at -50°C, shows a band in the near IR at ca. 8900 cm^-1 (w _1/2 ca. 3700 cm^-1) consistent with the presence of a valence delocalized system. The comproportionation constant for the equilibrium 10 + 10²⁺ ? 2 10⁺ at 223 K is estimated as 10 ^13.6.

Full text of DOI:10.1016/j.ica.2004.07.062

Inorganica Chimica Acta 358 (2005) 1692–1708
www.elsevier.com/locate/ica
Bis(acetylacetonato)ruthenium(II) complexes containing
bulky tertiary phosphines. Formation and redox behaviour of
Ru(acac)2 (PR3) (R = iPr, Cy) complexes with ethene,
carbon monoxide, and bridging dinitrogen
*
Martin A. Bennett , Matthew J. Byrnes, Guandolina Chung,
Alison J. Edwards, Anthony C. Willis
Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia
Received 29 June 2004; accepted 28 July 2004
Available online 13 November 2004
Dedicated to Professor Gordon Stone, FRS, on the occasion of his 80th birthday
Abstract
Reaction of cis-[Ru(acac)2(g2-C8H14)2] (1) (acac = acetylacetonato) with two equivalents of PiPr3 in THF at ꢀ25 ꢀC gives trans-
[Ru(acac)2(PiPr3)2], trans-3, which rapidly isomerizes to cis-3 at room temperature. The poorly soluble complex [Ru(acac)2(PCy3)2]
(4), which is isolated similarly from cis-[Ru(acac)2(g2-C2H4)2] (2) and PCy3, appears to exist in the cis-configuration in solution
according to NMR data, although an X-ray diffraction study of a single crystal shows the presence of trans-4. In benzene or toluene
2 reacts with PiPr3 or PCy3 to give exclusively cis-[Ru(acac)2(g2-C2H4)(L)] [L = PiPr3 (5), PCy3 (6)], whereas in THF species believed
to be either square pyramidal [Ru(acac)2L], with apical L, or the corresponding THF adducts, can be detected by 31P NMR spectr-
oscopy. Complexes 3–6 react with CO (1 bar) giving trans-[Ru(acac)2(CO)(L)] [L = PiPr3 (trans-8), PCy3 (trans-9)], which are con-
verted irreversibly into the cis-isomers in refluxing benzene. Complex 5 scavenges traces of dinitrogen from industrial grade
dihydrogen giving a bridging dinitrogen complex, cis-[{Ru(acac)2(PiPr3)} (l-N2)] (10). The structures of cis-3, trans-4, 5, 6 and
2
10ÆC6H14 have been determined by single-crystal X-ray diffraction. Complexes trans- and cis-3, 5, 6, cis-8, and trans- and cis-9 each
show fully reversible one-electron oxidation by cyclic voltammetry in CH2Cl2 at ꢀ50 ꢀC with E1/2(Ru3+/2+) values spanning ꢀ0.14 to
+0.92 V (versus Ag/AgCl), whereas for the vinylidene complexes [Ru(acac)2 (C@CHR)(PiPr3)] [R = SiMe3 (11), Ph (12)] the process
is irreversible at potentials of +0.75 and +0.62 V, respectively. The trend in potentials reflects the order of expected p-acceptor ability
of the ligands: PiPr3, PCy3 <C 2H4 < C@CHR < CO. The UV–Vis spectrum of the thermally unstable, electrogenerated RuIII
–
ethene cation 6+ has been observed at ꢀ50 ꢀC. Cyclic voltammetry of the l-dinitrogen complex 10 shows two, fully reversible proc-
esses in CH2Cl2 at ꢀ50 ꢀC at +0.30 and +0.90 V (versus Ag/AgCl) corresponding to the formation of 10+ (RuII,III) and 102+
(RuIII,III). The former, generated electrochemically at ꢀ50 ꢀC, shows a band in the near IR at ca. 8900 cmꢀ1 (w1/2 ca. 3700
cmꢀ1) consistent with the presence of a valence delocalized system. The comproportionation constant for the equilibrium
10 + 102+ ꢀ 2 10+ at 223 K is estimated as 1013.6
ꢁ 2004 Elsevier B.V. All rights reserved.
.
Keywords: Ruthenium (II) complex; Acetylacetonato complex; Redox behaviour; X-ray crystal structure
*
Corresponding author. Tel.: +61 2 6125 3639; fax: +61 2 6125
3216.
E-mail address: bennett@rsc.anu.edu.au (M.A. Bennett).
0020-1693/$ - see front matter ꢁ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2004.07.062

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1693
1. Introduction
we have used so far evidently replace the second alkene
in 1 or 2 more rapidly than the first, since mono-substi-
tution products could not be detected in these cases. It
seemed possible that the second step could be slowed
by use of bulky ligands such as PiPr3 and PCy3 and
we report here our results, which complement and ex-
tend those of Werner and coworkers [9].
Reduction of tris(acetylacetonato)ruthenium(III),
[Ru(acac)3], with zinc dust or zinc amalgam in the pres-
ence of a variety of ligands (L) provides a convenient
synthesis of ruthenium(II) complexes of the type [Ru(a-
ˆ
cac)2L2]. The procedure was employed first by Sato and
coworkers [1] to prepare the acetonitrile complex cis-
[Ru(acac)2(NCMe)2] and has been extended both to
complexes containing bidentate ligands such as S-BI-
NAP [2], chelating dienes [3,4] and conjugated dienes
[5–7], and to more labile complexes containing mono-
alkenes [cyclooctene (C8H14) and ethene (C2H4)] [4,8]
and monodentate Group 15 donors such as triisopropyl-
stibine (SbiPr3) and triisopropylphosphine (PiPr3) [9].
Werner and coworkers [9] have shown that one of the
stibine ligands of cis-[Ru(acac)2(SbiPr3)2] is easily re-
placed to give the mixed ligand species cis-[Ru
(acac)2(L)(SbiPr3)] (L = PiPr3, PCy3, C2H4, C@CHPh).
We have reported [4] that many ligands (L) replace the
labile alkenes in cis-[Ru(acac)2(g2-alkene)2] [alk-
ene = C8H14 (1), C2H4 (2)] to give, unexpectedly, the
trans-isomers of [Ru(acac)2L2] (L = PMe3, PEt3,
2. Results
2.1. General
The chemistry to be described is summarized in
Schemes 1–3. Elemental analyses and mass spectromet-
ric data for the new bis(acac) complexes are listed in Ta-
ble 1; selected IR and NMR data are collected in Tables
2 and 3. All the compounds show two or three intense
bands in the regions of 1590–1570 and 1525–1510
cmꢀ1 of their IR spectra (Table 2), which are character-
istic of O-bonded acac in its usual bidentate mode [10].
In general, the number of bands observed is unfortu-
nately not diagnostic of the cis- or trans-arrangement
of acac groups; this feature, however, is easily estab-
lished from the 1H and 13C NMR spectra (Table 3) [4,8].
t
PMe2Ph, PMePh2, P(OMe)3, P(OPh)3, BuNC), which
isomerize on heating to the more stable cis-complexes,
probably via an undetected fluxional square pyramidal
intermediate Ru(acac)2L. With some ligands, it is possi-
ble to isolate mono-substitution products cis-[Ru
(acac)2(g2-alkene)(L)] (alkene = C8H14, C2H4, L =
SbPh3, MeCN, NH3; alkene = C2H4, L = py) and, in
two cases, (alkene = C2H4, L = NH3, py), intermediate
trans-isomers have been identified [8]. Thus, the first
substitution in cis-[Ru(acac)2(g2-alkene)2] probably also
proceeds via a five-coordinate intermediate, Ru(a-
cac)2(g2-alkene). However, the tertiary phosphines that
2.2. [Ru(acac)2L2] complexes
Addition of two equiv of PiPr3 to a solution of 1 in
THF at ꢀ25 ꢀC causes the precipitation within 1 h of
trans-[Ru(acac)2(PiPr3)2] (trans-3) as a moderately air-
sensitive, rust-red solid, which can be isolated in ca.
45% yield (Scheme 1). The corresponding reaction with
2 is slower, no precipitate having formed after 1 h. As
expected, the H and 13C{1H} NMR spectra of trans-3
1
L
L
O
O
O
O
O
L / THF
O
O
L
O
O
O
O
L
Ru
L
Ru
O
Ru
L
THF
L = PiPr3
L = PCy3
trans-3
cis-3
cis-4
O
THF (-C2H4)
O
trans-4
L
Ru
O
O
O
alkene = C8H14 (1)
C2H4 (2)
L
O
O
Ru
O
L / C6H6
alkene = C2H4,
L = PiPr3 (5), PCy3 (6)
Scheme 1.

1694
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
L
O
O
O
O
Ru
L
trans-3
O
L
O
O
L
O
O
O
O
∆
CO
trans-4
Ru
O
Ru
OC
or
CO
O
L = PiPr3
L = PCy3
trans-8
cis-8
cis-9
L
O
Ru
O
trans-9
O
5,6
Scheme 2.
L
L
O
O
O
O
L
L
N2
N
N
Ru
O
O
Ru
N
N
Ru
O
O
Ru
O
O
O
5
+
O
O
O
O
O
L = PiPr3
10 ∆∆ / ΛΛ
10 ∆Λ / ΛΛ
-
= acetylacetonato anion (C5H7O2 )
O
O
Scheme 3.
Table 1
Elemental analysesa and mass spectra of Ru(acac)2(PR3) complexes (R = iPr, Cy)
Complex
%C
%H
%P
m/z (% relative abundance, assignment)
trans-[Ru(acac)2(PiPr3)2] (3)b
cis-[Ru(acac)2(PiPr3)2] (3)
460.2 (100, M ꢀ PiPr3)
54.26 (53.32)
8.38 (8.38)
620.0 (12, M), 519 (7, M ꢀ acac),
459.9 (100, M ꢀ PiPr3)
[Ru(acac)2(PCy3)2] (4)
cis-[Ru(acac)2(g2-C2H4)(PiPr3)] (5)
64.23 (64.23)
51.73 (51.10)
9.37 (9.57)
8.06 (7.61)
7.20 (6.93)
860.5 (6, M), 580.2 (100, M ꢀ PCy3)
460.1 (100, M ꢀ C2H4), 359.1 [10, Ru(acac)(PiPr3)],
300.0 [17, Ru(acac)2]
cis-[Ru(acac)2(g2-C2H4)(PCy3)] (6)
trans-[Ru(acac)2(CO)(PiPr3)] (8)
cis-[Ru(acac)2(CO)(PiPr3)] (8)
trans-[Ru(acac)2(CO)(PCy3)] (9)
cis-[Ru(acac)2(CO)(PCy3)] (9)
cis-[{Ru(acac)2(PiPr3)}2(l-N2)] (10)
59.29 (57.78)
49.27 (49.10)
49.27 (49.20)
57.31 (57.08)
57.31 (57.21)
48.19 (48.07)
8.46 (8.11)
7.24 (7.25)
7.24 (6.90)
7.79 (7.66)
7.79 (7.87)
7.45 (7.61)
860.5 [2, Ru(acac)2(PCy3)2], 580.3 (100, M ꢀ C2H4)
460.1 (100, M ꢀ CO)
6.35 (6.36)
6.35 (6.03)
5.10 (4.62)
5.10 (4.87)
2.96 (2.63) (N)
488.2 (45, M), 460.2 (100, M ꢀ CO), 389.1 (24, M ꢀ acac)
580.1 (53, M ꢀ CO), 297.1 [50, Ru(acac)2]
608.3 (62, M), 580.2 (100, M ꢀ CO), 509.2 (31, M ꢀ acac)
950.2, 948.2 (M), 460.2 [100, Ru(acac)2(PiPr3)]
a
Found values in parentheses.
Elemental analyses not obtained owing to thermal instability of compound.
b
in d8-toluene at ꢀ35 ꢀC show one sharp singlet each for
the acac methyl and methine groups and the 13C{1H}
NMR spectrum contains just one C@O resonance; how-
ever, the 1H and 13C resonances of the isopropyl groups
are broad, possibly because of restricted rotation about
the Ru–P bonds. Rotamers arising from hindered rota-
tion about M–P bonds have been observed at low tem-
perature for a variety of four- and five-coordinate
complexes containing bulky ligands such as t-butylphos-
phines and PCy3 [11–16].
When a solution of trans-3 in d8-toluene is warmed
from ꢀ35 to +5 ꢀC, quantitative trans- to cis-isomeriza-
tion occurs, as is evident from the disappearance of the
31P{1H} NMR singlet at d 29.6 and its replacement by a

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1695
Table 2
Characteristic IR and NMR data for Ru(acac)2(PR3) complexes (R = iPr, Cy)
b
Complex
IR (cmꢀ1 a
)
NMR (d)
3
1565, 1507 (acac)
1563, 1506 (acac)
4
5
1583, 1514 (acac)
1585, 1512 (acac)
1H: 4.00–4.60 (m, C2H4); 13C: 69.6 (s, C2H4)
1H: 3.78–4.38 (m, C2H4); 13C: 69.5 (s, C2H4)
13C: 206.5 (d, JPC 120, CO)
6
trans-8
cis-8
trans-9
cis-9
10
1565, 1517 (acac), 1931 (C„O)
1588, 1573, 1516 (acac), 1928 (C„O)
1572, 1510 (acac), 1950 (C„O)
1589, 1575, 1520 (acac), 1944 (CO)
1582, 1512 (acac), 2089 (N„N)c
13C: 209.4 (d, JPC 20, C„O)
13C: 206.8 (d, JPC 120, CO)
13C: 209.6 (d, JPC 18, C„O)
a
Bands refer to KBr disks, except where stated.
In C6D6, coupling constants J are in Hertz.
Raman.
b
c
singlet at d 47.7. The cis-isomer is also formed by direct
reaction of 1 or 2 with two equivalents of PiPr3 at room
temperature and can be isolated as an orange solid,
identical with the product reported by Werner and
coworkers [9]. The cis-configuration is evident from
the 1H and 13C{1H} NMR spectra, which show two acac
methyl singlets and one methine singlet; in addition,
there are two C@O resonances in the 13C{1H} NMR
spectrum. The structure of cis-3 has been confirmed by
X-ray crystallography (see below).
[PdX2L2], cis-[PdX2L2], and trans-[RhCl(CO)L2], are re-
lated linearly to the chemical shifts of the free phos-
phines. We find the same to hold for the complexes
cis- and trans-[Ru(acac)2L2, the equations for the best
straight lines being given in Eqs. (1) and (2):
1
cis : Dd ¼ 50:3 ꢀ 0:72dPðligandÞ;
ð1Þ
ð2Þ
trans : Dd ¼ 31:4 ꢀ 0:71dPðligandÞ;
where
Addition of two equiv of PCy3 to a solution of 1 or 2
in THF at room temperature gives a rust-red solid of
empirical formula [Ru(acac)2(PCy3)2] (4), which can be
isolated in ca. 80% yield. At room temperature it is al-
most insoluble in benzene or toluene and only slightly
soluble in THF, but on warming it dissolves completely
in all three solvents to give orange-brown solutions. The
NMR spectra of the red-brown solid that crystallizes
from toluene are identical with those of the original solid
and the X-ray structure of a single crystal selected from
the batch shows it to be the trans-isomer (see below).
However, the 1H and 13C{1H} NMR spectroscopic data
(two acac CH3 singlets, one acac methine singlet, and
two C@O resonances) are consistent with the presence
of cis-4 as the main species in solution.
Dd ¼ dPðcomplexÞ ꢀ dPðligandÞ:
On this basis the 31P chemical shifts calculated for cis-
and trans-4 are +43.9 and +26.0, respectively, and the
observed value is clearly closer to that expected for
cis-4. The similarity between the observed values for
the 31P chemical shifts for 4 and cis-[Ru(acac)2(g2-
C2H4)(PCy3)] (see below) also supports this assignment.
The matter could perhaps be settled by solid-state 31P
NMR spectroscopic measurements but these have not
been carried out.
There was no reaction between 1 and an excess of
PtBu3, even after a week at room temperature.
2.3. [Ru(acac)2(L)(L0)] complexes
The 31P{1H} NMR chemical shift also agrees better
with a cis-formulation, although the evidence is not deci-
sive. The spectrum in d8-toluene at room temperature
consists of a very broad peak (half-width 145 Hz) at d
38.5, which sharpens on warming to 100 ꢀC and sepa-
rates into two peaks of unequal intensity at d 37.7 and
The reactions of complex 2 with one equiv of PiPr3 or
PCy3 are solvent-dependent, as shown by 31P NMR
spectroscopy (Scheme 1). For PiPr3 in THF at ꢀ35 ꢀC
the first detectable product is the bis(ligand) complex
trans-3 (dP 29.6), but when the solution is allowed to
warm to ca. 10 ꢀC this disappears and is replaced by sin-
glets at d 87.3, 50.5 and 19.2, the last being due to free
PiPr3. Similarly, addition of one equiv of PCy3 to 2 in
THF at ꢀ35 ꢀC gives initially a singlet at d 21.0, whose
chemical shift suggests that it may be due to trans-4 (see
above), together with a peak at d 8.9 due to free PCy3.
When the solution is brought to ca. 0 ꢀC for 30 min, a
1
36.9 at ꢀ65 ꢀC. At the lowest temperature the H and
13C resonances of the cyclohexyl groups were broad
and provided no information. Similar behaviour has
been observed for cis-[Ru(O2CMe)2(PCy3)2] [9], trans-
[W(CO)3(PCy3)2] [14], and trans-[Re(CO)3(PCy3)2]+
[15], and, in the last two cases, was attributed to the
presence of rotamers. Shaw and coworkers [16–18] have
shown that the 31P shifts caused by coordination in a
series of tertiary phosphine complexes, such as trans-
The 31P chemical shifts are provided as supplementary material.
1

Table 3
NMR data for acac and PR3 in Ru(acac)2(PR3) complexes (R = iPr, Cy)a
Complex
acac
1H
PR3
1H
13C
13C
31P
CH3
CH
CH3
CH
C@O
trans-3b
cis-3
1.75
1.78, 1.91
5.10
5.30
27.4
27.8, 27.9
100.5
99.8
184.1
183.6, 186.8
1.32 (br, CH3), 2.30 (br, CH)
1.24, 1.37 (each dd, JPH 11,
JHH 7.2, CH3), 2.46 (m, CH)
1.15–1.45, 1.60–2.10, 2.20–2.40
(each m, CH2, CH)
20.0 (br, CH3), 24.8 (br, CH)
27.6 (d, JPC 9, CH3), 38.0 (br, CH)
29.6
47.7
4
1.82, 1.94
5.28
27.7 (d, JPC 2.2),
28.0
100.3
183.5, 186.7
27.5, 29.1 (m), 30.2 (br, CH2),
38.0 (br, CH)
38.5 (br)
50.5
5
1.84, 1.86, 1.88, 1.94 5.30, 5.35 27.2, 27.5, 28.1
(d, JPC 5.6), 28.3
98.1, 99.8
98.2, 99.8
98.9, 99.2
100.8
184.7, 185.0, 185.5 1.08, 1.23 (each dd, JPH 12, JHH 19.0, 19.4 (CH3), 24.8 (d, JPC 19, CH)
(d, JPC 2.2), 187.7 7.2, CH3), 2.45 (m, CH)
6
1.83, 1.86, 1.87, 1.98 5.30, 5.40 27.6, 27.7, 28.1
(d, JPC 5.6), 28.4
184.5, 184.9, 185.6 1.10–2.25 (br m, CH2, CH)
(d, JPC 2.2), 187.7
27.2, 28.5, 28.6, 28.7, 28.9, 29.2 (CH2), 40.0
35.4 (d, JPC 19, CH)
7
1.62, 1.64, 1.88, 1.96 5.07, 6.45 27.3, 28.0, 28.2,
28.3 (d, JPC 6.0)
184.2, 185.8,
186.0, 186.8
189.1
7.04–7.06 (m, m-, p-arom),
7.62–7.68 (m, o-arom)
1.23 (dd, JPH 12, JHH 7.1,
CH3), 2.25 (m, CH)
nm
54.7
18.9
61.5
8.4
trans-8
cis-8
trans-9
cis-9
1.70
5.08
27.0
19.1 (CH3), 23.1 (d, JPC 20, CH)
1.71, 1.79, 1.85, 1.91 5.16, 5.34 27.2, 27.7, 27.8, 28.0 99.1, 100.2 186.0, 186.5,
187.8, 189.1
1.13, 1.25 (each dd, JPH 13,
JHH 7.0, CH3), 2.25 (m, CH)
1.05–1.40, 1.50–2.30
18.8, 19.4 (CH3), 24.7 (d, JPC 22, CH)
1.73
5.09
27.0
100.6
188.9
27.1, 28.6 (d, JPC 8.8), 29.1
(CH2), 33.3 (d, JPC 10, CH)
27.0, 28.2 (d, JPC 2), 28.4
(d, JPC 3), 29.0
(br m, CH2, CH)
1.10–1.30, 1.50–2.00, 2.10–2.30
(br m, CH2, CH)
1.73, 1.86, 1.86, 1.98 5.17, 5.40 27.3, 27.8 (d, JPC
5.4), 27.9, 28.1
99.2, 100.3 186.0, 186.4,
187.9, 189.0
52.3
(d, JPC 2), 29.5 (CH2), 35.1
(d, JPC 21, CH)
19.2, 19.5 (CH3), 24.6 (d, JPC 20, CH)
10 DD/KK 1.81, 1.83, 1.85, 2.03 5.26, 5.33 27.2, 27.6, 27.8,
28.3 (d, JPC 5.6)
98.5, 100.3 184.6, 185.7,
186.4, 187.5
1.29, 1.40 (each dd, JPH 12.5,
JHH 7.5, CH3), 2.47 (m, CH)
60.6
10 DK/KD 1.80, 1.87, 2.08
5.29, 5.35 27.2, 27.6,
(d, JPC 5.6)
98.6, 100.3 184.5, 185.8, 187.6 1.29, 1.40 (each dd, JPH 12.5,
JHH 7.5, CH3), 2.47 (m, CH)
19.1, 19.4 (CH3), 24.6 (d, J
20, CH) 60.5
PC
a
Measured in C6D6 at 20.5 ꢀC, except where stated otherwise; resonances are singlets, except as indicated; coupling constants (J) in Hertz.
In d8-toluene at ꢀ40 ꢀC.
b

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1697
new species is formed that shows a singlet at d 74.7, and
at 25 ꢀC a second species of unknown origin character-
ized by a singlet at d 39.7 appears. The compounds
responsible for the highly deshielded resonances at d
87.3 (L = PiPr3) and 74.7 (L = PCy3) are likely to be
either the five-coordinate, probably square pyramidal
species [Ru(acac)2L], or possibly six-coordinate solvent
adducts [Ru(acac)2(L)(THF)]; similar deshielded reso-
nances have been observed for apical PPh3 in well-es-
tablished square pyramidal ruthenium(II) complexes
such as [RuCl2(PPh3)3] (dP 75.7) [19], [RuX(NN0N)(P-
Ph3)]OTf [dP 86.6 (X@Cl), 74.5 (X@OTf); NN0N@
duced by the cis-[Ru(acac)2] fragment is insufficient to
distinguish between the ethene carbon atoms. The
ethene C@C stretching bands could not be located in
the IR spectra of 5 and 6, probably because they are
hidden beneath the intense acac absorption in the
1500–1600 cmꢀ1 region. The complexes turn green on
exposure to air over several weeks but are apparently
stable indefinitely under argon at room temperature.
Ethene is not lost when a benzene solution of 6 is heated
in vacuo or under argon. However, although the
elemental analysis for 5 was satisfactory, that for 6
corresponded best with the empirical formula [Ru(a-
cac)2(PCy3)], possibly because ethene was lost before
combustion was complete. The most abundant ion in
the FAB-mass spectra of 5 and 6 corresponds in each
case to [M ꢀ C2H4]+, no molecular ion peak being de-
tected. Also, in THF, the 31P{1H} NMR spectra of 5
and 6 show the same highly deshielded resonances ob-
served in THF solutions of 3 and 4 (see above).
Evidently ethene is only weakly bound in 5 and 6.
In benzene, complex 2 also reacts with PPh3 (one
equiv) to give cis-[Ru(acac)2(g2-C2H4)(PPh3)] (7) as the
2,6-[bis(dimethylamino)methyl]pyridine]
[20],
and
[Ru{N(SPR2)2}(PPh3)] [dP 75.1 (R = iPr), 77.8 (R@Ph)]
[21].
The same highly deshielded resonances appear in
solutions of 3 and 4 in THF at room temperature, to-
gether with those for the original compounds and the
free ligands, evidently owing to ligand dissociation
1
(Scheme 1). Moreover, the H NMR spectrum of 4 in
THF shows two additional singlets at d 1.75 and 5.20,
consistent with the presence of a species having trans-
acac groups; correspondingly, the 13C{1H} NMR spec-
trum shows additional signals at d 27.4 (acac methyl),
100.2 (acac methine) and 183.8 (acac C@O).
1
main product, which was identified by its H, 13C and
31P{1H} NMR spectra (Table 3). It is noteworthy that
the 31P chemical shifts in C6D6 of cis-[Ru(acac)2(PPh3)2]
and 7 are very close (53.7 and 54.7, respectively); the
same is true for their PPri3 analogues, cis-3 and 5, and
for their PCy3 counterparts (cis-4 and 6). In THF at
ꢀ20 ꢀC the first species detectable by 31P NMR spectr-
oscopy when PPh3 (one equiv) is added to 2 is trans-
[Ru(acac)2(PPh3)2], but at room temperature this is
replaced by a singlet at d 54.7, due either to cis-[Ru
(acac)2(PPh3)2] or 7, and a less intense singlet at d
83.9. The latter is the only signal observed when a large
excess of 1 (ca. 10 equiv) is added to a solution of PPh3
in THF at room temperature. It probably belongs to the
pyramidal species [Ru(acac)2(PPh3)] or to its octahedral
THF adduct.
The reactions of 2 with PiPr3 or PCy3 (one equiv) in
C6D6 are slower than those in THF and the only detect-
able 31P NMR resonances are the singlets at d 50.5 and
49.0, respectively; the highly deshielded singlets ascribed
1
to [Ru(acac)2L] are absent. The H NMR resonances
due to acac consist of four methyl and two methine sin-
glets, characteristic of a cis-[Ru(acac)2] fragment at-
tached to two different ligands, and there is also a
symmetrical 4H-multiplet due to coordinated ethene in
the region d 3.7–4.8, the pattern being similar to that ob-
served in the complexes cis-[Ru(acac)2(g2-C2H4)(L)]
(L = SbPh3, NH3, MeCN).2 Thus, in C6D6, the com-
plexes cis-[Ru(acac)2(g2-C2H4)(L)] [L = PiPr3 (5), PCy3
(6)] are the exclusive products of reaction of 2 with
one equiv of the ligands (Scheme 1).
2.4. Reactions of [Ru(acac)2L2]
Complexes 5 and 6 can be isolated as, respectively,
red crystalline and yellow microcrystalline solids in 65–
70% yield from the reactions of 2 with the ligands in
benzene or toluene at room temperature over a period
The trans-isomers of [Ru(acac)2L2] [L = PiPr3 (3),
PCy3 (4)], and the cis-ethene derivatives [Ru(acac)2(g2-
C2H4)(L)] [L = PiPr3 (5), PCy3 (6)] react readily with
CO (1 bar) to give monocarbonyl complexes trans-
[Ru(acac)2(CO)(L)] [L = PPri3 (8), PCy3 (9)], which have
been isolated as yellow solids in yields of 50–70%
(Scheme 2). The carbonylation of trans-3 was performed
at ꢀ20 ꢀC to prevent competing isomerization to cis-3.
The latter also reacts with CO (3 bar) over a period of
days to give trans-8. Both trans-8 and trans-9 form the
corresponding cis-isomers when they are heated in
benzene. This behaviour is generally similar to that ob-
served in the triphenylphosphine series, except that cis-
[Ru(acac)2(PPh3)2] is completely unreactive toward CO
(1–3 bar) [4]. Also, in contrast with the lability of ethene
1
of hours. The H and 31P{1H} NMR spectra are identi-
cal with those measured in situ and the structures have
been confirmed by X-ray crystallography (see below).
The 13C{1H} NMR spectra support the structural
assignment: there are four C@O resonances in the region
of d 190, two methine resonances at d ca. 100, and four
acac methyl singlets at d 27–28, together with a singlet at
d ca. 69.5 due to coordinated ethene. At ꢀ95 ꢀC the
ethene proton multiplets broaden but the 13C resonance
remains sharp. Hence, as for other members of this class
[8,9], we do not know whether ethene rotation is fast
even at low temperature or whether the asymmetry in-

1698
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
in 5 and 6, the alkene in the triphenylstibine complexes
cis-[Ru(acac)2(g2-alkene)(SbPh3)] (alkene = C8H14 [4],
C2H4 [8]) is not replaced by CO after several days at
room temperature.
but this is accompanied by a peak at m/z 950.2 of un-
known origin.
The NMR spectra of 10 reveal the presence of two
isomers in solution, which are probably the homochiral
(DD/KK) and heterochiral (DK/KD) species arising from
the two chiral cis-[Ru(acac)2] fragments (Scheme 3).
Only the former is found in the solid state structure
(see below). Thus, the 1H NMR spectrum measured
ca. 10 min after dissolution of a solid sample in C6D6
shows, as expected, four acac methyl and two methine
singlets belonging to the major isomer; the correspond-
ing signals of the minor isomer are similar except that
two of the acac methyl singlets overlap. The isomer ratio
under these conditions is ca. 4:1 but this changes steadily
with time to ca. 2:1 after 14 h and 3:2 after 62 h. The
31P{1H} NMR spectrum, measured after 62 h, shows
two distinct singlets, at d 60.6 and 60.5, also in a ratio
of ca. 3:2, and in the 13C{1H} NMR spectrum there
are seven acac methyl, four methine, and seven C@O
resonances, the odd numbers presumably arising from
accidental overlap of a pair of signals. The isomer ratio
in a solution of 10, measured directly in situ by 1H
NMR spectroscopy after the reaction of 5 with dinitro-
gen, is ca. 1:1. Thus, the homochiral species must crys-
tallize preferentially and transform slowly in solution
to the diastereomeric mixture, possibly via the equilib-
rium shown in Eq. (3). A similar one-ended dissociation
of bridging dinitrogen has been observed in the nickel(0)
complex [{Ni(PCy3)2}2(l-N2)], the mononuclear species
in this case, [Ni(PCy3)2(N2)], being detectable by IR
spectroscopy [24]. Diastereomers of the binuclear os-
mium(II) complex cis-[{OsCl(bipy)2}2(l-N2)] have also
The structures of the isomers of 8 and 9 follow une-
quivocally from the number of acac resonances in the
1H and 13C{1H} NMR spectra (Table 2). The IR spectra
contain a single, intense m(CO) band in the region of
1950–1925 cmꢀ1 whose position does not differ signifi-
cantly for the cis- and trans-isomers. The highest identi-
fiable peak in the FAB-mass spectra of the trans-isomers
corresponds to the loss of one carbonyl ligand, whereas
a parent ion peak is observed in the case of the cis-
isomers.
In an attempt to generate either the five-coordinate
species [Ru(acac)2(PiPr3)] or its dihydrogen adduct, we
heated a benzene solution of the ethene complex 5 under
industrial grade hydrogen (3 bar) for 3 days. Unexpect-
edly, the product, isolated in ca. 90% yield, was the
yellow, crystalline, binuclear dinitrogen complex cis-
[{Ru(acac)2(PiPr3)}2(l-N2)] (10), which presumably
was formed from the small amount of nitrogen (ca.
100 ppm) present in the hydrogen gas. The presence of
nitrogen in 10 was confirmed by elemental analysis.
Complex 10 also resulted from an attempt to crystallize
5 in a nitrogen-filled ‘‘inert atmosphere’’ box over a per-
iod of several weeks. The displacement of ethene by
nitrogen is reversible, 5 being re-formed from 10 in the
presence of C2H4 (1 bar) in C6D6 at 25 ꢀC. Complex
10 also reacts readily with CO (1 bar) giving trans-8,
and with phenylacetylene to give the known phenylviny-
lidene complex cis-[Ru(acac)2(@C@CHPh)(PiPr3)] (11)
[9].
1
been observed by H NMR spectroscopy [25].
A single-crystal X-ray diffraction study of a n-hexane
solvate of 10, discussed in detail below, shows the pres-
ence of two cis-[Ru(acac)2(PiPr3)] units bridged symmet-
rically by linearly bonded dinitrogen (Scheme 3). When
the molecule is viewed down the Ru–N2–Ru axis, the
PiPr3 ligands appear to be mutually orthogonal. The Ra-
man spectrum of 10 contains a strong band at 2089
cmꢀ1 assigned to the symmetric m(NN) mode, similar
to those found for the symmetrical binuclear cations
[{Ru(NH3)5}2(l-N2)]4+ (2100 cmꢀ1) [22] and [{Ru(-
H2O)5}2(l-N2)]4+ (2080 cmꢀ1) [23]. The 2089 cmꢀ1 band
does not appear in the IR spectrum of 10, either in the
solid state or in benzene solution. Hence, the amount
of any mononuclear dinitrogen complex present in equi-
librium with 10 (Eq. (3)) is too small to be detected by
IR spectroscopy
2.5. X-ray structures
The molecular structures of cis-3, trans-4, 5, 6 and
10ÆC6H14, determined by single-crystal X-ray diffrac-
tion, are shown in Figs. 1–5; selected bond lengths and
angles are collected in Tables 4–8. The coordination
geometries are essentially octahedral, as expected. In
trans-4 the interbond angles are close to 90ꢀ, whereas
in cis-3 the P–Ru–P angle opens out to 105ꢀ, presumably
because of steric repulsion between the mutually cis-
PiPr3 ligands; similar angles have been observed in
[Rh(acac)(PCy3)2] [26], ½Rhðg2-O2CMeÞðPiPr3Þ ꢁ [27],
2
cis-[Mo(CO)4(PCy3)2] [28], and cis-[Mo(CO)4(PPh3)2]
[29]. In 5 the C@C vector of the coordinated ethene
eclipses the plane defined by the metal atom and the
mutually trans oxygen atoms O(1) and O(3); this orien-
tation is perpendicular to that found in cis-[Ru
(acac)2(g2-C2H4)(NH3)]2 and is presumably adopted to
minimize steric repulsion between ethene and the bulky
PiPr3 ligand. Surprisingly, however, the ethene ligand in
6, containing the even bulkier PCy3 ligand, displays two
mutually perpendicular orientations, the slightly more
½fRuðacacÞ ðPiPr3Þg2ðl-N2Þꢁ
2
ꢀ ½RuðacacÞ ðN2ÞðPiPr3Þꢁ þ ½RuðacacÞ ðPiPr3Þꢁ ð3Þ
2
2
The most abundant ion in the FAB-mass spectrum is at
m/z 460.2 corresponding to the fragment [Ru(acac)2
(PiPr3)]. A parent ion peak at m/z 948.2 is observed,

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1699
Fig. 1. Molecular structure of cis-[Ru(acac)2(PiPr3)2], cis-3, with
selected atom labelling. Displacement ellipsoids show 30% probability
levels. Hydrogen atoms have been omitted for clarity.
Fig. 3. Molecular structure of cis-[Ru(acac)2(g2-C2H4)(PiPr3)], 5, with
selected atom labelling. Displacement ellipsoids show 30% probability
levels. Hydrogen atoms have been omitted except those on the ethene
group, which are drawn as circles of small radii.
Fig. 2. Molecular structure of trans-[Ru(acac) (PCy3)2], trans-4, with
2
selected atom labelling. Displacement ellipsoids show 50% probability
levels. Hydrogen atoms have been omitted for clarity.
abundant one (ca. 55% occupation) being the same as in
5; in the minor one the C@C vector eclipses the plane
defined by Ru(1), P(1) and O(2). The Ru–C(C2H4) dis-
Fig. 4. Molecular structure of cis-[Ru(acac)2(g2-C2H4)(PCy3)], 6, with
selected atom labelling, showing the major orientation of coordinated
ethene. Displacement ellipsoids show 50% probability levels. Hydrogen
atoms have been omitted except those on the ethene group, which are
drawn as circles of small radii.
˚
tances in 5 and 6 are in the range 2.15–2.21 A found
in members of the series containing less bulky co-lig-
ands, viz., cis-[Ru(acac)2(g2-C2H4)(L)] (L = C2H4,
NH3) and trans-[Ru(acac)2(g2-C2H4)(py)] [8], and the
˚
C@C bond lengths are only ca. 0.02 A greater than that
˚
of free ethene [1.333(2) A] [30].
such as [{Ru(NH3)5}2(l-N2)]4+ [1.124(15) A] [31],
˚
In complex 10ÆC6H14 the Ru–N„N–Ru unit is al-
most linear, the Ru–N„N angle being 174ꢀ. The
0
˚
[{mer,trans-RuCl2(NN N)}2(l-N2)] [1.110(3) A] [20],
[{CpRu(dippe)}2(l-N2)]2+ [dippe = 1,2-bis-(diisopropyl-
˚
N„N distance [1.135(8) A] is slightly greater than that
˚
in free dinitrogen [1.0977 A] [30] and is similar to those
˚
phosphino)ethane]
[1.118(3)A]
˚
[32],
[{CpRu
(PEt3)2}2(l-N2)]2+ [1.114(5) A] [32], and [{RuH2-
observed in other l-dinitrogen ruthenium(II) complexes

1700
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
Table 6
Selected metrical parameters for complex 5
Ru(1)–O(1) 2.079(3) Ru(2)–O(5)
Ru(1)–O(2) 2.077(3)
2.082(3)
2.079(3)
2.064(3)
2.089(3)
2.322(1)
2.180(6)
2.185(5)
Ru(2)–O(6)
Ru(2)–O(7)
Ru(2)–O(8)
Ru(2)–P(2)
Ru(2)–C(32)
Ru(2)–C(33)
Ru (1)–O(3)
Ru(1)–O(4)
Ru(1)–P(1)
Ru(1)–C(11)
Ru(1)–C(12)
C(11)–C(12)
2.071(3)
2.094(3)
2.321(1)
2.172(5)
2.181(5)
1.350(9)
O(1)–Ru(1)–O(2)
O(1)–Ru(1)–O(3)
O(1)–Ru(1)–O(4)
O(1)–Ru(1)–P(1)
O(2)–Ru(1)–O(3)
O(2)–Ru(1)–O(4)
O(2)–Ru(1)–P(1)
O(3)–Ru(1)–O(4)
O(3)–Ru(1)–P(1)
O(4)–Ru(1)–P(1)
C(11)–Ru(1)–C(12)
P(1)–Ru(1)–C(11)
P(1)–Ru(1)–C(12)
89.1(1)
167.4(1)
81.4(1)
95.68(9)
79.9(1)
85.4(1)
91.63(9)
91.6(1)
90.70(9)
175.9(1)
36.1(2)
99.7(2)
91.3(2)
O(5)–Ru(2)–O(6)
O(5)–Ru(2)–O(7)
O(5)–Ru(2)–O(8)
O(5)–Ru(2)–P(2)
O(6)–Ru(2)–O(7)
O(6)–Ru(2)–O(8)
O(6)–Ru(2)–P(2)
O(7)–Ru(2)–O(8)
O(7)–Ru(2)–P(2)
O(8)–Ru(2)–P(2)
C(32)–Ru(2)–C(33)
P(2)–Ru(2)–C(32)
P(2)–Ru(2)–C(33)
89.5(1)
168.1(1)
81.4(1)
95.93(9)
80.3(1)
84.7(1)
91.14(9)
91.5(1)
90.48(9)
175.01(9)
36.0(2)
100.5(2)
91.0(2)
Fig. 5. Molecular structure of cis-[{Ru(acac)2(PiPr3)}(l-N2)], 10, with
selected atom labeling. Asterisks indicate atoms generated by the
symmetry operation (2 ꢀ x, y, 3/2 ꢀ z). Displacement ellipsoids show
30% probability levels. Only the major orientation of the disorder of
atoms C(17) and C(19) is shown. Hydrogen atoms have been omitted
for clarity.
Table 4
Selected metrical parameters for complex cis-3
Table 7
Selected metrical parameters for complex 6
Ru(1)–O(1)
Ru(1)–O(2)
Ru (1)–O(3)
2.068(2)
2.088(2)
2.059(2)
Ru(1)–O(4)
Ru(1)–P(1)
Ru(1)–P(2)
2.104(2)
2.3525(9)
2.3467(8)
Ru(1)–O(1)
Ru(1)–O(2)
Ru(1)–O(3)
Ru(1)–O(4)
Ru(1)–P(1)
2.06(2)
2.10(2)
2.07(2)
Ru(1)–C(121)
2.148(2)
1.005
1.344
0.987
1.097
1.356
0.738
C(110)–C(111)a
C(110)–C(120)a
C(110)–C(121)a
C(111)–C(120)a
C(111)–C(121)a
C(120)–C(121)a
2.074(18)
2.356(7)
2.167(2)
2.190(2)
2.151(2)
O(1)–Ru(1)–O(2)
O(1)–Ru(1)–O(3)
O(1)–Ru(1)–O(4)
O(1)–Ru(1)–P(1)
O(1)–Ru(1)–P(2)
O(2)–Ru(1)–O(3)
O(2)–Ru(1)–O(4)
O(2)–Ru(1)–P(1)
90.46(8)
174.02(7)
85.28(8)
90.20(6)
94.43(6)
84.62(8)
79.77(8)
167.23(6)
O(2)–Ru(1)–P(2)
O(3)–Ru(1)–O(4)
O(3)–Ru(1)–P(1)
O(3)–Ru(1)–P(2)
O(4)–Ru(1)–P(1)
O(4)–Ru(1)–P(2)
P(1)–Ru(1)–P(2)
87.25(6)
90.50(8)
93.86(6)
88.73(6)
87.57(6)
167.01(6)
105.42(3)
Ru(1)–C(110)
Ru(1)–C(120)
Ru(1)–C(111)
O(1)–Ru(1)–O(2)
O(1)–Ru(1)–O(3)
O(1)–Ru(1)–O(4)
O(1)–Ru(1)–P(1)
O(2)–Ru(1)–O(3)
O(2)–Ru(1)–O(4)
O(2)–Ru(1)–P(1)
O(3)–Ru(1)–O(4)
a
90.9(8)
89.2(6)
173.1(8)
90.2(6)
82.9(9)
83.9(8)
171.9(8)
90.2(8)
O(3)–Ru(1)–P(1)
89.2(6)
94.4(6)
O(4)–Ru(1)–P(1)
C(110)–Ru(1)–C(120)a
P(1)–Ru(1)–C(110)
P(1)–Ru(1)–C(120)
C(111)–Ru(1)–C(121)a
P(1)–Ru(1)–C(111)
P(1)–Ru(1)–C(121)
35.94(3)
116.33(18)
80.40(17)
36.78(3)
99.29(17)
93.92(17)
Table 5
Selected metrical parameters for complex trans-4
Ru(1)–O(1)
Ru(1)–O(2)
2.0677(13)
2.0658(12)
Ru(1)–P(1)
2.4273(4)
The atoms C(110), C(111), C(120) and C(121) were restrained
during refinement.
O(1)–Ru(1)–O(1)*
O(1)–Ru(1)–O(2)
O(2)–Ru(1)–O(2)*
P(1)–Ru(1)–P(1)*
*
180.0
P(1)–Ru(1)–O(1)
P(1)–Ru(1)–O(1)*
P(1)–Ru(1)–O(2)
P(1)–Ru(1)–O(2)*
90.94(4)
89.06(4)
91.06(4)
88.94(4)
91.87(5)
180.0
180.0
relatively unhindered coordination environment pro-
vided by two acac ligands and only one tertiary phos-
phine. The two octahedra in 10ÆC6H14 adopt a
mutually orthogonal orientation which, as suggested
also for the second and fourth of the compounds cited
above, may help to maximize d(p) back-donation into
the orthogonal sets of empty p*-orbitals on N2.
Atoms generated by the symmetry operation (1 ꢀ x, 1 ꢀ y, 1 ꢀ z).
1
i
˚
(g -N2)(P Pr3)2}2(l-N2)] [1.113(2) A for bridging N2,
˚
1.105(2) A for terminal N2] [33], consistent with
d(p) ! p* Ru–N2 back-bonding [31,34,35]. The Ru–N
The lability of complexes 3 and 4 can be correlated
with the Ru–P bond lengths. Thus, the Ru–P distance
˚
in trans-4, 2.4273(4) A, is significantly greater than in
distance to the nitrogen atom of N2 in 10ÆC6H14
˚
[1.919(14) A] is slightly shorter than those found in the
˚
first two compounds cited above [1.928(6), 1.953(2) A,
˚
trans-[Ru(acac)2(PMePh2)2] [2.343(1), 2.346(1) A for
independent molecules] [4]; the same trend is evident in
˚
a comparison of cis-3 [2.3525(9), 2.3467(8) A for inde-
pendent molecules] with cis-[Ru(acac)2(PMePh2)2]
respectively] and significantly shorter than those ob-
served in the last three [1.980(1), 1.977(3) and 2.050(2)
˚
A
(average), respectively], possibly reflecting the

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1701
Table 8
Selected metrical parameters for complex 10ÆC6H14
Table 9
Reduction potentials E1/2(Ru3+/2+) for complexes [Ru(acac)2(L)(L0)]a,b
Ru(1)–O(1)
Ru(1)–O(2)
Ru(1)–O(3)
Ru(1)–O(4)
2.046(4)
2.035(4)
2.050(4)
2.107(4)
Ru(1)–P(1)
Ru(1)–N(1)
N(1)–N(1)a
2.312(2)
1.919(4)
1.135(8)
Complex
E1/2
Complex
E1/2
trans-3
cis-5
trans-8
trans-9
11
ꢀ0.14
cis-3
cis-6
cis-8
cis-9
12
+0.02
+0.43
+0.92
+0.89c
+0.62d
+0.42
nm
+0.64c
+0.75d
+0.30, +0.90
O(1)–Ru(1)–O(2)
O(1)–Ru(1)–O(3)
O(1)–Ru(1)–O(4)
O(1)–Ru(1)–P(1)
O(1)–Ru(1)–N(1)
O(2)–Ru(1)–O(3)
O(2)–Ru(1)–O(4)
O(2)–Ru(1)–P(1)
a
92.4(2)
173.8(2)
82.5(2)
96.0(1)
90.3(1)
86.8(1)
86.4(1)
92.3(1)
O(2)–Ru(1)–N(1)
O(3)–Ru(1)–O(4)
O(3)–Ru(1)–P(1)
O(3)–Ru(1)–N(1)
O(4)–Ru(1)–P(1)
O(4)–Ru(1)–N(1)
P(1)–Ru(1)–N(1)
Ru(1)–N(1)–N(1)a
174.2(2)
91.3(2)
90.2(1)
90.0(2)
177.9(1)
88.9(2)
92.6(1)
174.2(4)
10
a
L0 = PiPr3, L = PiPr3 (3), C2H4 (5), CO (8), 1/2 N2 (10), C@CHPh
(11), C@CHSiMe3 (12); L0=PCy3, L = C2H4 (6), CO (9).
b
Measured in CH2Cl2 vs. Ag/AgCl at ca. ꢀ50 ꢀC, except where
stated.
c
Measured at ca. 20 ꢀC.
Irreversible process (see text).
d
Atoms generated by the symmetry operation (2-x, y, 3/2-z).
rates up to 500 mVsꢀ1, presumably because of the insta-
bility of the oxidation products. The anodic potentials
are +0.75 and +0.62 V, respectively. Although these val-
ues are not strictly comparable with the E1/2-values for
C2H4 and CO, the trend is similar to the order
i
˚
[2.2765(9) A] [4]. Replacement of one of the P Pr3 lig-
ands in cis-3 by C2H4 or N2 causes a contraction of
˚
ca. 0.04 A in the Ru–P distance.
2.6. Electrochemistry
C2H4 < C@CPh2 < CO
observed
for
trans-
RhClðLÞðPPri3Þ ꢁ, which is believed to reflect primarily
The half-wave potentials E1/2(Ru3+/2+) of the com-
plexes studied in this work are listed in Table 9. The
trans- and cis-isomers of 3 each show fully reversible re-
dox couples, the former being the more easily oxidized
by ca. 140 mV. Differences of a similar order have been
observed for trans- and cis-[Ru(acac)2(NCMe)2] (120
mV) and [Ru(acac)2(AsPh3)2] (190 mV) [36,37]. We
could not measure the E1/2-value for 4 because of the
insolubility of this compound in CH2Cl2 at room tem-
perature. Replacement of one of the PiPr3 ligands in
cis-3 by ethene shifts the redox potential by ca. 400
mV in favour of RuII, presumably because of the weaker
r-donor/stronger p-acceptor ability of ethene. The E1/2
values of 5 and 6 (+0.42, +0.43 V, respectively) are
slightly greater than that of cis-[Ru(acac)2(g2-
C2H4)(NH3)] (+0.38 V),2 but, surprisingly, are consider-
ably less than those for the corresponding acetonitrile
and SbPh3 complexes (+0.56, +0.59 V, respectively).
Substitution of CO for one of the PiPr3 ligands in cis-3
causes an even larger shift in favour of RuII than does
C2H4, clearly reflecting the preferred binding of the
strong p-acceptor to the d6 ion. The effect is most
marked for cis-8 and cis-9, which are less easily oxidized
to the RuIII level than their trans-isomers by ca. 300 mV,
in accord with the usual trend for substituted carbonyl
complexes [38,39]. The couples are fully reversible for
both cis- and trans-isomers.
2
the p-acceptor ability of these ligands [40].
The electronic spectra recorded at ꢀ50 ꢀC for the
one-electron
oxidation
of
cis-[Ru(acac)2(g2-
C2H4)(PCy3)] (6) are similar to those reported previ-
ously for cis-[Ru(acac)2(g2-C2H4)(L)] (L = C2H4,
NH3, SbPh3) [8] and for chelate alkene–amine com-
plexes such as [Ru(acac)2(o-CH2@CHC6H4NMe2)]
and [Ru(acac)2(2-CH2@CHC5H4N)] [41]. On electro-
oxidation of 6, a band at 29600 cmꢀ1 (e 6500), which
probably arises from charge transfer transitions within
the Ru(acac)2 unit, is replaced by absorptions at 14700
cmꢀ1 (e 1300) and 17600 cmꢀ1 (e 1300), due to acac(p)
! RuIII transitions in 6+, and by a band at 29400
cmꢀ1 (e 4700). The band in 6 at 36700 cmꢀ1 (e
14600), which probably arises from acac p ! p* tran-
sitions, shifts only slightly in position and intensity on
electro-oxidation. These changes can be completely re-
versed at ꢀ50 ꢀC by applying a potential of ca. 300 mV
less than the E1/2(Ru3+/2+) value. Isosbestic points are
observed during both oxidation and reduction proc-
esses showing that only two absorbing species are pre-
sent in solution. Addition of one equiv of AgPF6 to a
cold solution of 6 gives a deep blue solution whose CV
trace (6+ ! 6) is superimposable on that of 6 ! 6+.
Hydrodynamic voltammetric results show that, for 6,
no current is detected between 0 and +0.4 V (versus
Ag/AgCl) but, as the potential passes between +0.4
and +0.5 V, a positive current passes through the solu-
tion corresponding to the oxidation of the complex;
above +0.5 V a limiting current is reached. The poten-
tial at half-current height is ca. +0.45 V, in good agree-
ment with the value measured by cyclic voltammetry.
After addition of a slight excess of AgPF6, a negative
The vinylidene complexes [Ru(acac)2(@C@CHR)-
2
(PiPr3)] [R = Ph (11), SiMe3 (12)] [9] exhibit irreversi-
ble cyclic voltammograms, even at ꢀ50 ꢀC with scan
2
Complex 12 was prepared similarly to 11 by treatment of cis-3
with trimethylsilylacetylene.

1702
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
Table 10
Principal electronic band maxima (cmꢀ1) and molar absorptivities
(Mꢀ1 cmꢀ1) for RuII and in situ electrogenerated RuIII complexes
current is found between 0 and +0.4 V; above +0.5 V
no current is passing through the solution. The poten-
tial at half-current height is again ca. +0.45 V, indicat-
3
ing the presence of 6+ in solution.
Complex
MLCT or LMCT
acac p ! acac p*
6
29600 (6500)
27500 (16900)
36700 (14600)
37400 (53500)
36600 (12700)
The cyclic and AC voltammograms of the binuclear
l-dinitrogen complex 10 show two reversible oxidation
potentials, at +0.30 and + 0.90 V (versus Ag/AgCl) at
ꢀ50 ꢀC in CH2Cl2. All of the Ru(acac)2 complexes con-
taining p-acceptor ligands studied so far [8,36,37,41]
show just one one-electron oxidation in the potential
range 0–1.5 V and show no RuIII ! RuIV oxidation.
Hence, the two couples for 10 probably correspond to
sequential oxidation of the two metal centres, i.e.,
RuII,II ! RuII,III and RuII,III ! RuIII,III. At 25 ꢀC the
first oxidation is still fully reversible whereas the second
is irreversible. The difference in E1/2-values for the two
oxidation couples is similar to that found for
[{Ru(NH3)5}2(l-N2)]4+, which undergoes a reversible
one-electron oxidation at E1/2 + 0.73 V (versus NHE)
and a second, irreversible oxidation at ca. +1.2 V (versus
NHE) [42]. The electronic spectrum of 10 is very similar
to that of 6 and of the other Ru(acac)2(C2H4) complexes
[8]; application of a potential of ca. +1.2 V (versus Ag/
AgCl) at ꢀ50 ꢀC generates the spectrum typical of a
RuIII(acac)2 complex (Table 10) and is consistent with
the presence of the doubly oxidized RuIII,III species
cis-[{Ru(acac)2(PiPr3)}2(l-N2)]2+, 102+. This process is
fully reversible at ꢀ50 ꢀC. Application of a potential
of +0.60 V (versus Ag/AgCl) at ꢀ50 ꢀC generates a spec-
trum that differs from those of 10 and 102+ mainly in the
presence of a weak, broad, asymmetric near-IR band at
ca. 8900 cmꢀ1 (Table 10); again, the process is com-
pletely reversible at ꢀ50 ꢀC and isosbestic points are ob-
served, as shown in Fig. 6. The species responsible for
this spectrum is likely to be the half-oxidized RuII,III cat-
ion 10+; similar near IR-bands have been reported for
the mixed valence species [{Ru(NH3)5}2(l-N2)]5+ [42]
and for the Creutz–Taube ion containing bridging pyra-
zine, [{Ru(NH3)5}2(l-pyz)]5+ [43,44].
10
6+
14700 (1300),
17600 (1300)
10+
8900 (2900), 16300 (1800),
17700 (1800), 30100 (15500)
13900 (3000), 17800 (3000),
30000 (16700)
38200 (38800)
37600 (35700)
102+
IR band in 10+ (w1/2 ca. 3700 cmꢀ1) is broader than that
in [{Ru(NH3)5}2(l-N2)]5+ (w1/2 ca. 2630 cmꢀ1) but the
half-width is less than 4750 cmꢀ1, the value predicted
by Hushꢀs theory [48] for localized mixed-valence sys-
tems. Hush [49] has also suggested that fully delocalized
systems will show asymmetric bands. All these observa-
tions are consistent with the notion that 10+ belongs to
the class of fully delocalized mixed-valence ions (class
III of the Robin–Day classification).
3. Discussion
The tertiary phosphines PiPr3 and PCy3, like their less
bulky analogues, replace both alkenes from 1 or 2 to
give cis-[Ru(acac)2L2] as the final product. In contrast
with the behaviour of the less bulky ligands, however,
the half-substituted intermediates cis-[Ru(acac)2(g2-
C2H4)(L)] [L = PiPr3 (3), PCy3 (4)] can be isolated from
2, provided that the reaction is carried out in a poorly
coordinating solvent such as benzene. These compounds
may be formed via their trans-isomers, as is the case for
the corresponding compounds with L = NH3 or py, but
we have no strong supporting evidence. Although the
Ru–C(C2H4) and C@C bond lengths in 3 and 4 are nor-
mal, the ethene appears to be weakly bound, being dis-
placed readily by CO, N2 and even to some extent by
solvent THF; under the same conditions the analogues
cis-[Ru(acac)2(g2-C2H4)(L)] (L = NH3, py, SbPh3) are
inert towards these reagents.
We attribute this difference to the accessibility and
relative stability of the formally five-coordinate, 16-elec-
tron fragments Ru(acac)2L (L = PiPr3, PCy3), which can
be detected, possibly in the form of their weakly stabi-
lized THF adducts, by their characteristic, highly de-
shielded 31P NMR resonances. The exclusive
formation of the trans-isomers of the carbonyl com-
plexes 8 and 9 as the kinetic products of reaction of
CO with the cis-ethene complexes 5 and 6 is consistent
with a square pyramidal geometry, with L in the apical
position, for the 16-electron fragment. A similar expla-
nation was advanced previously to account for the for-
mation of the kinetic products trans-[Ru(acac)2L2]
Since the potentials for the stepwise oxidation of 10
differ by more than 250 mV [45], the value for the com-
proportionation constant K (Eq. (4)) at 223 K can be
estimated simply from the difference in E1/2 values (Eq.
(5)):
10 þ 102þ ꢀ 2 10þ
ð4Þ
ð5Þ
log10K ¼ 22:65ðDE1=2Þ:
The resulting value, 1013.6, can be compared with the
values at 298 K of ca. 108 for [{Ru(NH3)5}2(l-N2)]5+
and 4.106 for [{Ru(NH3)5}2(l-pyz)]5+ [46,47]. The near
3
The current is slightly greater after addition of the oxidant,
possibly because of a slight concentration of the solution during the
measurement, the presence of colloidal silver on the electrodes, or the
slight excess of Ag+.

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1703
Fig. 6. Electronic spectra recorded during the one-electron oxidation of cis-[{Ru(acac)2(PiPr3)} 2(l-N2)] (10) in 0.5 M ½BunNꢁPF6=CH2Cl2 at ca.
4
ꢀ50 ꢀC (Eappl = + 0.60 V vs. Ag/AgCl).
from the reactions of the cis-complex 1 with two equiv-
alents of L [4]. Square pyramidal geometry is favoured
for a five-coordinate d6-metal complex according to the-
oretical calculations [50,51] and has been observed for
several five-coordinate ruthenium(II) complexes, e.g.,
cates the high affinity of the fragment for N2. We have
no evidence for the existence of a mononuclear species,
[Ru(acac)2(g1-N2)(PiPr3)], under the reaction condi-
tions, although it may be an intermediate in the forma-
tion of 10 and, as noted above, in the interconversion of
the corresponding diastereomers. The high affinity
of Ru(acac)2(PiPr3) for N2 is reminiscent of the ability
of the 16-electron fragments [RuCpL2]2+ (L2 = dippe,
2PEt3, 2PMeiPr2) [32], Mo(CO)(Et2PCH2CH2PEt2)2
[56], and Ru{P(CH2CH2PPh2)3} [57] to scavenge traces
of dinitrogen from commercial argon or helium; we have
not attempted to generate Ru(acac)2(PiPr3) under these
gases. The apparently low affinity of Ru(acac)2(PiPr3)
for dihydrogen contrasts with the ability of numerous
16-electron fragments, such as [RuCpL2]2+, [RuH2Ln]
L = various tertiary phosphines; n = 2,3), and
[Os(NH3)5]2+, to bind both g1-N2 and H2 (either as
g2-H2 or a dihydride) ꢁ[58,59]; moreover, g2-H2 and
g1-N2 are supposed to have similar p-acceptor ability
[60]. The closest analogue to Ru(acac)2(PiPr3) may be
[Ru(H2O)5]2+, which reversibly binds both N2 and H2
[RuCl2(PPh3)3]
(Y = Cl, OTf) [20], and [Ru{N(SPR2)}2(PPh3)]
[19,52],
[RuY(NN0N)(PPh3)]OTf
i
(R = Ph, Pr) [21]. The proposed intermediate Ru(a-
cac)2L (L = PiPr3, PCy3) may be further stabilized by
weak intramolecular Ruꢂ ꢂ ꢂH–C interactions, as ob-
served in [RuCl2(PPh3)3] [52].
Complexes 3 and 4 undergo trans to cis isomerization
far more readily than their counterparts containing
smaller Group 15 donor ligands, a feature that can also
be traced both to the relative stability of the fragment
Ru(acac)2L and to the long (and presumably weak)
Ru–L bonds in the trans-isomers. Thermochemical
measurements of ligand exchange reactions 0of
[RuCl2(g6-p-cymene)(PR3)] [53] and ð½RuClðg5-C5R Þ-
5
ðPR3Þꢁ (R0 = H, Me) [54,55] have shown that PiPr3 or
PCy3 form weaker bonds, by ca. 45 kJ molꢀ1, than
PMe3, for example. Despite the steric crowding inherent
in the cis-arrangement of two such bulky ligands, the
competition between two strong r-donors for the same
metal-based orbital evidently disfavours the trans-
arrangement.
to give [Ru(H2O)5(g1-N2)]2+ and [Ru(H2O)5(g2-H2)]2+
,
respectively [61,62]. In both cases the back-bonding
interactions are thought to be weak [62,63] and the g1-
N2 complex readily decomposes to the l-N2 species
[{Ru(H2O)5}2(l-N2)]2+ [23,62]. Clearly, further work di-
rected to the detection or isolation of the mononuclear
entities [Ru(acac)2(g2-H2)(PiPr3)] and [Ru(acac)2(g1-
N2)(PiPr3)] is warranted.
The ability of the 16-electron fragment to form com-
plex 10 by scavenging dinitrogen from industrial grade
dihydrogen, itself a potential competitor ligand, indi-

1704
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
The ruthenium(III)–ethene cation 6+, like the analo-
dry-ice/acetone slush bath. Standard solutions of the
appropriate tertiary phosphine in d8-toluene were pre-
pared similarly. The solutions were mixed, placed imme-
diately in a NMR probe cooled to ꢀ40 ꢀC, and the
31P{1H} NMR spectra recorded immediately.
gous complexes [Ru(acac)2(g2-C2H4)(L)]+ (L = NH3,
SbPh3) [8], is stable indefinitely in solution at ꢀ50 ꢀC
but decomposes at room temperature, consistent with
weaker binding of ethene to the metal atom in its higher
oxidation state. Solutions of the half- and fully-oxidized
l-dinitrogen complexes, 10+ and 102+, also are stable at
ꢀ50 ꢀC. Evidently RuIII–N2 bonds have some stability
in the absence of competing donor solvents, despite
the poor p-donor ability of the trivalent metal ion; for
the ammine cation analogous to 10+, [{Ru(NH3)5}2(l-
N2)]5+, the half-life for aquation at 0 ꢀC is only 3.4
min [42].
4.3. Preparations
4.3.1. [Ru(acac)2(PiPr3)2] (3)
A solution of 1, prepared from [Ru(acac)3] (800 mg,
ca. 2.0 mmol) in THF (30 ml), was cooled to ca.
ꢀ25 ꢀC and PiPr3 (0.78 ml, 4.08 mmol) was added
from a gas-tight syringe. After the solution had been
stirred for 1 h, a fine solid had formed. The volume
of solvent was reduced in vacuo by about half and cold
hexanes were added by cannula. The mixture was stir-
red for ca. 5 min, the solid was allowed to settle, and
the supernatant was decanted. The process was re-
peated several times with cold pentane until the supern-
atant was almost colorless. Filtration gave trans-3 as a
rust-red solid (0.58 g, ca. 45%), which was stable in an
inert atmosphere at ca. ꢀ20 ꢀC but decomposed slowly
on exposure to air. The complex was converted into
cis-3 in toluene over 1 h at room temperature. The
product isolated from this solution by removal of sol-
vent in vacuo was identical to that obtained independ-
ently by Werner and coworkers [9] by reduction of
[Ru(acac)3] with Zn/Hg in the presence of PiPr3 or by
treatment of [Ru(acac)2(SbiPr3)2] with an excess of
PiPr3.
4. Experimental
4.1. General procedures
All operations were carried out under argon, unless
stated otherwise, with use of standard Schlenk tech-
niques. Solvents were dried by standard methods and
distilled under nitrogen. NMR spectra were recorded
at 20.5 ꢀC (unless otherwise indicated) on Varian Gemi-
ni 300 BB or Varian VXR 300 spectrometers (1H at 300
MHz, 13C at 75.4 MHz, 31P at 121.4 MHz) spectrome-
ters. Variable temperature NMR spectra were recorded
on the Varian VXR 300 instrument. Chemical shifts (d)
for 1H and 13C are given in ppm referenced to the resid-
ual protons and the carbon atoms of the deuterated sol-
vents, 31P chemical shifts are referenced to external 85%
H3PO4. Coupling constants (J) are given in Hertz. IR
spectra were recorded on either Perkin-Elmer 1800 FT
or Perkin-Elmer Spectrum One instruments with use
of KBr disks or 0.1 mm path length KBr cells. Fast
Atom Bombardment (FAB) mass spectra were meas-
ured on a VG ZAB2-SEQ mass spectrometer using
either 3-nitrobenzyl alcohol or (3-nitrophenyl)octylether
as the matrix. Microanalyses were performed in-house.
Electrochemical measurements in CH2Cl2 at various
temperatures were carried out on a Princeton Applied
Research 170 system as described in previous papers
[8,41,64]. At room temperature E1/2 for the [FeCp2]+/0
couple was +0.55 V versus a non-aqueous reference elec-
trode consisting of Ag/AgCl/CH2Cl2, 0.05 M [Bu4N]Cl,
0.45 M [Bu4N]PF6. Electronic spectra in the range 6000–
45000 cmꢀ1 were collected in a cryogenically controlled
OTTLE cell placed in the beam of either a Perkin-Elmer
Lambda 9 or a Cary 5E UV–Vis–NIR spectrometer as
described earlier.
4.3.2. [Ru(acac)2(PCy3)2] (4)
A solution of 2, prepared from [Ru(acac)3] (300
mg, 0.75 mmol) in THF (15 ml), was stirred with
PCy3 (500 mg, 1.78 mmol) at room temperature. After
the solution had been stirred for 30 min, a fine brown
solid had formed. The mixture was stirred for another
6 h and set aside overnight. The supernatant was
removed by decantation and the brown solid was
washed with pentane. The yield of 4 was 530 mg
(82%). The solid is air-stable and insoluble in benzene
and toluene at room temperature, though it is
soluble in the hot solvents. It is partly soluble in
THF at room temperature and decomposes in
CH2Cl2.
4.3.3. Cis-[Ru(acac)2(g2-C2H4)(PiPr3)] (5)
A magnetically stirred solution of 2 (freshly prepared
from [Ru(acac)3] (1.0 g, 2.51 mmol) in benzene (40 ml))
was treated with PiPr3 (0.48 ml, 2.51 mmol) at room
temperature and set aside overnight, causing the colour
to change from yellow to orange. Solvent was removed
in vacuo and the orange residue was redissolved in pen-
tane (ca. 5 ml). The solution was cooled to ꢀ20 ꢀC to
give a red, crystalline solid. The supernatant was re-
moved by decantation and evaporated to ca. half-vol-
4.2. In situ 31P NMR studies
Aliquots of freshly prepared solutions of 1 or 2 in
C6D6, or in THF containing a small amount of d8-tolu-
ene, contained in sealed NMR tubes were cooled in a

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1705
ume in vacuo to give a second crop of red solid. Both
crops were washed by decantation with cold pentane.
The yield of 5 was 676 mg (55%). X-ray quality crystals
were selected from the first crop. The solid is stable
indefinitely under argon at 25 ꢀC and is very soluble in
benzene or toluene to give yellow solutions that turn
green immediately on exposure to air.
4.3.7. trans-[Ru(acac)2(CO)(PCy3)] (trans-9)
A suspension of 4 (200 mg, 0.23 mmol) in benzene (10
ml) was stirred under CO (1 bar). A clear orange solu-
tion was formed within 30 min. Most of the solvent
was removed in vacuo, the yellow residue was stirred
with pentane (5 ml), and the solution was evaporated
to a volume of ca. 2 ml in vacuo. The yellow solid that
precipitated was washed with ca. 2 ml of pentane. The
yield of trans-9 was 101 mg (72%). The properties of
the solid are similar to those of trans-8.
4.3.4. cis-[Ru(acac)2(g2-C2H4)(PCy3)] (6)
A solution of 2 (freshly prepared from [Ru(acac)3]
(420 mg, 1.05 mmol) in benzene (40 ml)) was treated
dropwise with a solution of PCy3 (292 mg, 1.04 mmol)
in benzene (10 ml) at room temperature. After 1 h the
colour had changed from yellow to orange. The mixture
was stirred overnight and worked up as for 5 to give 6 as
a yellow, microcrystalline solid (641 mg, 68%). Its prop-
erties are similar to those of 5. X-ray quality crystals
were obtained by slow evaporation of the supernatant
after the first crop of solid.
4.3.8. cis-[Ru(acac)2(CO)(PCy3)] (cis-9)
A solution of trans-9 (200 mg, 0.33 mmol) in benzene
(10 ml) was heated under reflux for 1 h. The solution
was evaporated almost to dryness and pentane (ca. 3
ml) was added to the yellow residue. The light yellow so-
lid that formed overnight was separated by filtration.
The yield of cis-9 was 123 mg (62%).
4.3.9. cis-[{Ru(acac)2(PiPr3)}2(l-N2)] (10)
4.3.5. trans-[Ru(acac)2(CO)(PiPr3)] (trans-8)
(i) A solution of trans-3 (246 mg, 0.40 mmol) in tolu-
ene (2 ml) and hexanes (5 ml) at ꢀ70 ꢀC was allowed to
warm to room temperature under CO (1 bar). When the
temperature had reached ꢀ15 ꢀC, the solution had be-
come yellow. Solvents were removed in vacuo leaving
a yellow solid, which was treated with pentane (5 ml).
The resulting solution was evaporated to ca. 2 ml and
set aside at ꢀ20 ꢀC to give a yellow solid, which was sep-
arated by filtration and washed with cold pentane (2 ml).
The yield of trans-8 was 105 mg (52%). The solid turns
green in air over several months and is readily soluble
in benzene, toluene, and THF.
(ii) A solution of cis-3 (197 mg, 0.32 mmol) in ben-
zene (10 ml) was stirred for 3 days under CO (3 bar)
in a medium pressure vessel behind a safety shield. Mon-
itoring by 31P{1H} NMR spectroscopy showed the
products to be trans-8 and free PiPr3, together with a
small amount of cis-8.
Benzene (10 ml) was added to cis-3 (167 mg, 0.34
mmol) in a medium pressure vessel under industrial
grade hydrogen gas (1 bar), which probably contains
at least 100 ppm N2 (this level is specified in Grade 3.8
high purity hydrogen). The pressure was increased to 3
bar and the yellow solution was heated at 60 ꢀC for 2
days. The solution was allowed to cool to room temper-
ature and the hydrogen pressure reduced to 1 bar. The
solution was cannulated into a Schlenk flask under
hydrogen and the solvent was removed in vacuo. The
yellow residue was dissolved in a small volume of hexa-
nes or pentane and the solution was set aside at ꢀ20 ꢀC
for 2 days. The yellow crystals were separated by filtra-
tion under argon and washed with small aliquots of hex-
anes or pentane. The yield of 10 was 149 mg (92%).
The solid compound is stable to air for short periods
but is best stored under argon. It forms thermally stable
solutions in benzene or toluene that immediately turn
green in air. Solutions in CH2Cl2 immediately turn
red. The products of these decompositions have not
been identified.
(iii) Carbon monoxide was bubbled gently through a
solution of 5 (ca. 15 mg, 0.03 mmol) in C6D6 (0.4 ml) for
30 min. Monitoring by 1H and 31P{1H} NMR spectros-
copy showed that trans-8 had been formed quantita-
tively after several hours.
4.4. X-ray crystallography
Crystal data and details of data collection are given in
Table 11. The data for complexes cis-3 and 5 were col-
lected on a Rigaku AFC6R diffractometer with graph-
4.3.6. cis-[Ru(acac)2(CO)(PiPr3)] (cis-8)
A solution of trans-8 (150 mg, 0.31 mmol) in ben-
zene (10 ml) was heated under reflux for 1 h, causing
the colour to fade. The solution was evaporated almost
to dryness and the residue was treated with pentane (2
ml). The solution was set aside at ꢀ20 ꢀC for 2 days,
and the resulting pale yellow, crystalline solid was sep-
arated by filtration and washed with cold pentane. The
yield of cis-8 was 85 mg (57%). The solid is air-stable
for several weeks and readily soluble in aromatic
solvents.
˚
ite-monochromated CuKa radiation (k = 1.54178 A)
and a rotating anode generator, those for 6 on a Rigaku
AFC6S diffractometer with graphite-monochromated
˚
Mo Ka radiation (k = 0.71069 A), and those for trans-
4 and 10 on a Nonius Kappa CCD area detector with
graphite-monochromated Mo Ka radiation. The struc-
tures of cis-3, 5 and 6 were solved by direct methods
[65] and those of trans-4 and 10 by heavy atom Patter-
son methods [66]. Least-squares refinements were

Table 11
Crystal and refinement data for cis-[Ru(acac)2(PiPr3)2] (cis-3), trans-[Ru(acac)2(PCy3)2] (trans-4), cis-[Ru(acac)2(g2-C2H4)(PiPr3)] (5), cis-[Ru(acac)2(g2-C2H4)(PCy3)] (6), and cis-[{Ru(acac)2-
a
(PiPr3)}2(l-N2)] (10ÆC6H14
)
cis-3
trans-4
5
6
10ÆC6H14
Empirical formula
Formula weight
Crystal system
C28H56O4P2Ru
619.77
C46H80O4P2Ru
860.16
C21H39O4PRu
487.58
triclinic
C30H51O4PRu
607.78
monoclinic
C44H84N2O8P2Ru2
1033.24
monoclinic
P21n (No. 14)
orange, block
10.660(3)
monoclinic
P21n (No. 14)
brown, plate
9.56650(10)
17.8085(3)
monoclinic
C 2/c (No. 15)
orange, block
20.087(1)
ꢀ
Space group
Crystal colour, habit
P1 (No. 2)
P21c (No. 14)
orange, triangular fragment
13.699(4)
orange, wedge
9.747(1)
15.645(2)
17.270(2)
66.391(7)
81.03(1)
89.79(1)
2378.5(5)
4
˚
a (A)
˚
b (A)
19.825(3)
15.504(3)
11.780(4)
19.130(3)
14.562(1)
18.065(1)
˚
c (A)
13.6210(2)
a (ꢀ)
b (ꢀ)
c (ꢀ)
98.89(2)
109.4660(10)
92.94(2)
106.538(4)
3
˚
V (A )
3237(1)
4
1.272
2187.90(5)
2
1.306
3076(2)
5065.5(5)
4
1.355
Z
4
1.31
Dcalc (g cmꢀ3
)
1.362
62.70
243
l (cmꢀ1
T (K)
)
50.76
193
4.70
200
5.90
296
7.07
200
Crystal dimensions (mm)
F(000)
0.22 · 0.18 · 0.15
1320.00
120.1
0.27 · 0.14 · 0.11
924.00
66.3
0.30 · 0.16 · 0.10
1024.00
120.1
0.50 · 0.30 · 0.18
1284.00
50.1
0.20 · 0.10 · 0.10
2176.00
55.1
2hmax (ꢀ)
Number of reflections measured
6272
45663
7548
5995
9739
Number of unique reflections (Rint
Number of observations
Number of variables
)
4813 (0.026)
4295 [I > 2r(I)]
317
8233 (0.047)
5758 [I > 3r(I)]
241
0.355; 0.156; 0.106b
0.035; 0.038
1.407
7064 (0.046)
5865 [I > 2r(I)]
511
5743 (0.018)
4319 [I > 3r(I)]
358
1.97; 0.322; 1.49b
0.028; 0.034
1.22
5751 (0.051)
3093 [I > 3r(I)]
254
p-Factor
R; RW
S
0.030
0.028; 0.034
1.71
0.020
0.036; 0.044
1.88
0.040
0.050; 0.059
1.44
q
max, qmin (eꢀ3
)
0.75, ꢀ0.52
0.46, ꢀ0.92
1=2
0.78, ꢀ0.74
0.28, ꢀ0.46
0.62, ꢀ0.55
P
P
P
P
P
Definitions: R = ||Fo|ꢀ|Fc||/ |Fo|; RW ¼ ½ wðjF oj ꢀ jF cjÞ = wF 2oꢁ ; w ¼ ½r2ðF oÞ þ 0:25p2F 2ꢁꢀ1; GOF = [
w(|Fo|ꢀ|Fc|)2/(NoꢀNv)]1/2; No = number of observations, Nv = number of
2
a
o
variables.
b
3-Term Chebychev polynomial: J.R. Carruthers, D.J. Watkin, Acta Crystallogr., Sect. A 35 (1979) 698.

M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
1707
carried out with use of the crystallographic software
packages teXsan [67] (for cis-3, 5 and 10) and CRYS-
TALS [68] (for trans-4 and 6).
associated with this article can be found, in the online
version, at doi:10.1016/j.ica.2004.07.062.
The non-hydrogen atoms of cis-3, trans-4 and 5 were
refined anisotropically, methyl groups being aligned to
best-fit peaks in difference electron-density maps. For
cis-3 and 5, the hydrogen atoms were included at geo-
metrically determined positions which were periodically
recalculated but not refined; for trans-4 the hydrogen
atoms rode on the carbon atom to which they were at-
tached. In 6 the coordinated ethene was disordered over
two sites with occupancies of 55(4) and 45(4)% (see text).
The optimal refinement required the imposition of re-
straints on the Ru–C distances, which were based on
those of the closely related complexes 2 and 5. Conse-
quently the e.s.d.s of the geometrical parameters of the
Ru–ethene fragment are underestimated and no e.s.d.s
are reported for these carbon atom positional parame-
ters. The remaining atoms were refined similarly to those
of cis-3, trans-4 and 5.
References
ˆ
[1] T. Kobayashi, Y. Nishina, K. Shimizu, G.P. Sato, Chem. Lett.
(1988) 1137.
[2] A.S. Chan, S.A. Laneman, C.X. Day, Inorg. Chim. Acta 228
(1995) 159.
[3] M.A. Bennett, H. Neumann, A.C. Willis, V. Ballantini, P. Pertici,
B.E. Mann, Organometallics 16 (1997) 2868.
[4] M.A. Bennett, G. Chung, D.C.R. Hockless, H. Neumann, A.C.
Willis, J. Chem. Soc., Dalton Trans. (1999) 3451.
[5] R.D. Ernst, E. Melendez, L. Stahl, M.L. Ziegler, Organometallics
10 (1991) 3635.
[6] E. Melendez, R. Ilarraza, G.P.A. Yap, A.L. Rheingold, J.
Organomet. Chem. 522 (1996) 1.
[7] K. Mashima, H. Fukumoto, K. Tani, M. Haga, A. Nakamura,
Organometallics 17 (1998) 410.
[8] M.A. Bennett, M.J. Byrnes, A.C. Willis, Organometallics 22
(2003) 1018.
[9] G. Grunwald, M. Laubender, J. Wolf, H. Werner, J. Chem. Soc.,
¨
In 10 a region of electron density around the twofold
axis at 0.50, y, 0.75 (y ca. 0.76) was modelled as a disor-
dered molecule of n-hexane with restraints on distances
and angles. Elongated ellipsoids on C(17) and C(19) im-
plied that these atoms also were disordered; each was
split over two sites and relative occupancies were refined.
All ordered non-hydrogen atoms were refined anisotrop-
ically; disordered atoms were refined isotropically.
Hydrogen atoms of the solvate and of the minor confor-
mation for the disordered PiPr3 were omitted; other
hydrogen atoms were included at calculated positions.
Neutral atom scattering factors [69], the anomalous
dispersion terms (Df0 and Df00) [70], and the mass atten-
uation coefficients [71] were taken from the appropriate
references.
Dalton Trans. (1998) 833.
[10] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, fifth ed., Wiley, New York, 1997, Part
B, p. 91.
[11] B.E. Mann, C. Masters, B.L. Shaw, R.E. Stainbank, Chem.
Commun. (1971) 1103.
[12] C.H. Bushweller, C.D. Rithner, D.J. Butcher, Inorg. Chem. 25
(1986) 1610.
[13] J.T. Poulton, M.P. Sigalas, K. Folting, W.E. Streib, O. Eisenstein,
K.G. Caulton, Inorg. Chem. 33 (1994) 1476.
[14] M.D. Butts, J.C. Bryan, X.-L. Luo, G.J. Kubas, Inorg. Chem. 36
(1997) 3341.
[15] D.M. Heinekey, C.E. Radzewich, M.H. Voges, B.M. Schomber,
J. Am. Chem. Soc. 119 (1997) 4172.
[16] B.E. Mann, C. Masters, B.L. Shaw, R.M. Slade, R.E. Stainbank,
Inorg. Nucl. Chem. Lett. 7 (1971) 881.
[17] B.E. Mann, C. Masters, B.L. Shaw, J. Chem. Soc., A (1971) 104.
[18] B.E. Mann, B.L. Shaw, R.M. Slade, J. Chem. Soc., A (1971)
2976.
[19] P.R. Hoffman, K.G. Caulton, J. Am. Chem. Soc. 97 (1975) 4221.
´
[20] R.A.T.M. Abbenhuis, I. del Rıo, M.M. Bergshoef, J. Boersma,
N. Veldman, A.L. Spek, G. van Koten, Inorg. Chem. 37 (1998)
1749.
Acknowledgements
[21] W.H. Leung, H. Zheng, J.L.C. Chim, J. Chan, W.T. Wong, I.D.
Williams, J. Chem. Soc., Dalton Trans. (2000) 423.
[22] J. Chatt, A.B. Nikolsky, R.L. Richards, J.R. Sanders, Chem.
Commun. (1969) 154.
We thank Professor Brian Mann, University of Shef-
field, UK, for assistance with some of the NMR exper-
iments, Dr. David Hockless for collecting the X-ray data
for complex 6, Dr. Vincent Otieno-Alego, University of
Canberra, for recording the Raman spectrum of
complex 10, and Dr. Graham Heath for helpful
discussions.
[23] C. Creutz, H. Taube, Inorg. Chem. 10 (1971) 2664.
[24] P.W. Jolly, K. Jonas, C. Kruger, Y.-H. Tsay, J. Organomet.
¨
Chem. 33 (1971) 109.
[25] G.M. Coia, K.D. Demadis, T.J. Meyer, Inorg. Chem. 39 (2000)
2212.
´
[26] M.A. Esteruelas, F.J. Lahoz, E. On˜ate, L.A. Oro, L. Rodrıguez,
P. Steinert, H. Werner, Organometallics 15 (1996) 3436.
[27] H. Werner, M. Scha¨fer, O. Nurnberg, J. Wolf, Chem. Ber. 127
¨
Appendix A. Supporting information available
(1994) 27.
[28] M. Watson, S. Woodward, G. Conole, M. Kessler, G. Sykara,
Polyhedron 13 (1994) 2455.
Tabulation of 31P NMR chemical shifts for cis- and
trans-[Ru(acac)2(PR3)2] and plots of DdP versus dP (free
ligand); tables giving X-ray data for cis-3, trans-4, 5, 6
and 10. This information is available free of charge from
the Internet at http://pubs.acs.org. Supplementary data
[29] F.A. Cotton, D.J. Darensbourg, S. Klein, B.W.S. Kolthammer,
Inorg. Chem. 21 (1982) 294.
[30] L.E. Sutton (Ed.), Tables of Interatomic Distances and Config-
urations in Molecules and Ions, Special Publication No. 18, The
Chemical Society, London, 1965.

1708
M.A. Bennett et al. / Inorganica Chimica Acta 358 (2005) 1692–1708
[31] I.M. Treitel, M.T. Flood, R.E. Marsh, H.B. Gray, J. Am. Chem.
Soc. 91 (1969) 6512.
[56] X-L. Luo, G.J. Kubas, C.J. Burns, R.J. Butcher, J.C. Bryan,
Inorg. Chem. 34 (1995) 6538.
´
[32] H. Aneetha, M. Jimenez-Tenorio, M.C. Puerta, P. Valerga, K.
Mereiter, Organometallics 21 (2002) 628.
[57] R. Osman, D.L. Pattison, R.N. Perutz, C. Bianchini, J.A. Casares,
M. Peruzzini, J. Am. Chem. Soc. 119 (1997) 8459.
[58] R.H. Morris, K.A. Earl, R.L. Luck, N.J. Lazarowych, A. Sella,
Inorg. Chem. 26 (1987) 2674.
[33] K. Abdur-Rashid, D.G. Guzev, A.J. Lough, R.H. Morris,
Organometallics 19 (2000) 1652.
[34] M.J. Andrechen, M.J. Ratner, D.E. Ellis, J. Am. Chem. Soc. 103
(1981) 1656.
[59] R.H. Morris, Inorg. Chem. 31 (1992) 1471.
[60] G.J. Kubas, in: Metal Dihydrogen and r-Bond Complexes,
Kluwer Academic/Plenum Publishers, New York, 2001, p. 75
(Chapter 4).
[35] J.N. Murrell, A. Al-Derzi, G.J. Leigh, M.F. Guest, J. Chem. Soc.,
Dalton Trans. (1980) 1425.
[36] L. Wallace, Ph.D. Thesis, Australian National University, 1991.
[37] D. Menglet, Ph.D. Thesis, Australian National University, 1996.
[38] F.L. Wimmer, M.R. Snow, A.M. Bond, Inorg. Chem. 13 (1974)
1617.
[61] G. Laurenczy, L. Helm, A.E. Merbach, A. Ludi, Inorg. Chim.
Acta 189 (1991) 131.
[62] N. Aebischer, U. Frey, A.E. Merbach, Chem. Commun. (1998)
2303.
[39] B.E. Bursten, M.R. Green, Prog. Inorg. Chem. 36 (1988) 393, and
references cited therein.
[63] N. Aebischer, E. Sidorenkova, M. Ravera, G. Laurenczy, D.
Osella, J. Weber, A.E. Merbach, Inorg. Chem. 36 (1997)
6009.
[40] I. Kovacik, O. Gevert, H. Werner, M. Schmittel, R. So¨llner,
Inorg. Chim. Acta 275–276 (1998) 435.
[64] C.M. Duff, G.A. Heath, Inorg. Chem. 30 (1991) 2528.
[65] SIR92: A. Altomare, M. Cascarano, C. Giacovazzo, A. Guag-
liardi, M.C. Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr.
27 (1994) 435.
[41] M.A. Bennett, G.A. Heath, D.C.R. Hockless, I. Kovacik, A.C.
Willis, J. Am. Chem. Soc. 120 (1998) 932.
[42] D.E. Richardson, J.P. Sen, J.D. Buhr, H. Taube, Inorg. Chem. 21
(1982) 3136.
[66] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, S.
Garcia-Granda, R.O. Gould, J.M.M. Smits, C. Smykalla, The
PATTY and DIRDIF-92 Program System, Technical Report of the
Crystallography Laboratory, University of Nijmegen, Nijmegen,
The Netherlands, 1992.
[43] C. Creutz, H. Taube, J. Am. Chem. Soc. 91 (1969) 3988.
[44] C. Creutz, H. Taube, J. Am. Chem. Soc. 95 (1973) 1086.
[45] D.E. Richardson, H. Taube, Inorg. Chem. 20 (1981) 1278.
[46] C. Creutz, Prog. Inorg. Chem. 30 (1983) 1.
[47] D.E. Richardson, H. Taube, Coord. Chem. Rev. 60 (1984) 107.
[48] N. Hush, Prog. Inorg. Chem. 8 (1967) 391.
[67] teXsan: Single-Crystal Structure Analysis Software, Version 1.8,
Molecular Structure Corp., The Woodlands, TX, 1997.
[68] D.J. Watkin, C.K. Prout, J.R. Carruthers, P.W. Betteridge,
Crystals, Chemical Crystallography Laboratory, Oxford, Issue
10, 1996.
[49] N. Hush, in: D.B. Brown (Ed.), Mixed-Valence Compounds,
Reidel Publishing Co, Dordrecht, Holland, 1980, p. 151.
[50] M. Elian, R. Hoffmann, Inorg. Chem. 14 (1975) 1058.
[51] J.K. Burdett, J. Chem. Soc., Faraday Trans. 2 70 (1974) 1599.
[52] S.J. La Placa, J.A. Ibers, Inorg. Chem. 4 (1965) 778.
[53] S.A. Serron, S.P. Nolan, Organometallics 14 (1995) 4611.
[54] M.E. Cucullu, L. Luo, S.P. Nolan, Organometallics 14 (1995)
289.
[69] D.T. Cromer, J.T. Waber, International Tables for X-ray Crys-
tallography, vol. IV, Kynoch Press, England, 1974.
[70] D.C. Creagh, W.J. McAuley, International Tables for Crystal-
lography, vol. C, Kluwer Academic, Boston, MA, 1992, p. 219.
[71] D.C. Creagh, J.H. Hubbell, International Tables for Crystallog-
raphy, vol. C, Kluwer Academic, Boston, MA, 1992, p. 200.
[55] L. Luo, S.P. Nolan, Organometallics 13 (1994) 4781.