Oxidative reactions of half-sandwich ruthenium compounds: Formation of cationic nitrosyl ruthenium(II) derivatives

Article abstract of DOI:10.1016/S0022-328X(01)00680-5

The electrochemical behaviour of the ruthenium(II) alkyl complexes [Ru(Me)Cp*(L)₂] (Cp=η⁵-C₅Me₅; L=PMe₃ 2a, PMe₂Ph 2b), [Ru(CH₂CMe₃)Cp*(PMe₃)₂] (3a), and the related [Ru(Me)Cp(PPh₃)₂] (4d) (Cp=η⁵-C₅H₅) in CH₂Cl₂ involves a one-electron process, yielding the corresponding ruthenium(III) paramagnetic cations, as shown by coupled electrochemical-EPR studies. Compounds 2-4 are oxidised by [FeCp₂]⁺ in benzene to unidentified paramagnetic products which may decompose giving the corresponding alkane. The Cp* compounds react with NOBF₄ affording the monocationic alkylnitrosyl derivatives [Ru(R)Cp*(NO)(L)]BF₄ (R=Me; L=PMe₃ 7a, PMe₂Ph 7b. R=CH₂CMe₃; L=PMe₃) and, when in excess of NO⁺, the ruthenium(II) dicationic complexes [RuCp*(NO)(L)₂](BF₄)₂ (L=PMe₃ 5a, PMe₂Ph 5b). The chloro complexes [Ru(Cl)Cp*(L)₂] (L=PMe₃ 1a, PMe₂Ph 1b, PPh₃ 1d) react analogously with NO⁺ to give [Ru(Cl)Cp*(NO)(L)]BF₄ (L=PMe₃ 6a, PMe₂Ph 6b, PPh₃ 6d) and [RuCp*(NO)(L)₂](BF₄)₂ (L=PMe₃ 5a, PMe₂Ph 5b). In contrast [Ru(Me)Cp(PPh₃)₂] gives only [Ru(Me)Cp(NO)(PPh₃)]BF₄. EPR spectroscopy suggests that these nitrosylation reactions are also oxidative in character proceeding through ruthenium(III) intermediates.

Full text of DOI:10.1016/S0022-328X(01)00680-5

Journal of Organometallic Chemistry 626 (2001) 145–156
www.elsevier.nl/locate/jorganchem
Oxidative reactions of ‘‘half-sandwich’’ ruthenium compounds:
formation of cationic nitrosyl ruthenium(II) derivatives
Pietro Diversi a,*, Marco Fontani b, Melania Fuligni a, Franco Laschi b,
Simona Matteoni a, Calogero Pinzino c, Piero Zanello b
a Dipartimento di Chimica e Chimica Industriale dell’Uni6ersita` di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
b Dipartimento di Chimica, Uni6ersita` di Siena, Via Aldo Moro, 53100 Siena, Italy
c Istituto di Chimica Quantistica e Energetica Molecolare del CNR, Area della ricerca di Pisa, Via Alfieri 1, Ghezzano, 56010 Pisa, Italy
Received 23 October 2000; accepted 5 January 2001
Abstract
The electrochemical behaviour of the ruthenium(II) alkyl complexes [Ru(Me)Cp*(L)2] (Cp*=h5-C5Me5; L=PMe3 2a, PMe2Ph
2b), [Ru(CH2CMe3)Cp*(PMe3)2] (3a), and the related [Ru(Me)Cp(PPh3)2] (4d) (Cp=h5-C5H5) in CH2Cl2 involves a one-electron
process, yielding the corresponding ruthenium(III) paramagnetic cations, as shown by coupled electrochemical–EPR studies.
Compounds 2–4 are oxidised by [FeCp2]+ in benzene to unidentified paramagnetic products which may decompose giving the
corresponding alkane. The Cp* compounds react with NOBF4 affording the monocationic alkylnitrosyl derivatives
[Ru(R)Cp*(NO)(L)]BF4 (R=Me; L=PMe3 7a, PMe2Ph 7b. R=CH2CMe3; L=PMe3) and, when in excess of NO+, the
ruthenium(II) dicationic complexes [RuCp*(NO)(L)2](BF4)2 (L=PMe3 5a, PMe2Ph 5b). The chloro complexes [Ru(Cl)Cp*(L)2]
(L=PMe3 1a, PMe2Ph 1b, PPh3 1d) react analogously with NO+ to give [Ru(Cl)Cp*(NO)(L)]BF4 (L=PMe3 6a, PMe2Ph 6b,
PPh3 6d) and [RuCp*(NO)(L)2](BF4)2 (L=PMe3 5a, PMe2Ph 5b). In contrast [Ru(Me)Cp(PPh3)2] gives only
[Ru(Me)Cp(NO)(PPh3)]BF4. EPR spectroscopy suggests that these nitrosylation reactions are also oxidative in character
proceeding through ruthenium(III) intermediates. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Nitrosyls; Rutheniumꢀcarbon bond cleavage; Electrochemistry; EPR spectroscopy
have been proved to be involved as intermediates in the
intra- and inter-molecular arene CꢀH activation by the
above iridium complexes in the presence of catalytic
amounts of one-electron oxidants (ETC catalysis). In
contrast the PPh3 derivative reacts in the polar acetoni-
trile solvent with [FeCp2]+ to give methane and
[Ir(Me)Cp*(PPh3)(MeCN)]+, no CꢀH bond activation
of the phosphine being observed [3]. Reaction of the
same iridium dimethyl complexes with NOBF4 affords
interesting cationic alkylnitrosyls [4], which form
through a mechanistic pathway involving a one-elec-
tron oxidation step. We have found, moreover, that the
ruthenium(II) dimethyl complexes [Ru(Me)2(h6-
C6Me6)(L)] are also able to activate the aromatic CꢀH
bonds in the presence of one-electron oxidants [5].
In this paper we describe the results obtained by
studying the reactivity, in particular the electrochemical
and chemical oxidation, of other ‘‘half-sandwich’’ sys-
tems, the alkyl ruthenium(II) compounds [Ru(Me)Cp*-
1. Introduction
Reactions involving the cleavage of metal–carbon
bonds, which are very important in organometallic
chemistry and catalysis, are oxidative in nature and are
promoted by electrophiles and oxidant agents [1]. In
particular the iridium(III) dimethyl complexes
[Ir(Me)2Cp*(L)] (L=PMe3, PMe2Ph, PMePh2, PPh3)
have been found to give a rich variety of reactions
under oxidative conditions [2]. Electrochemical oxida-
tion in CH2Cl2 involves a one-electron process, yielding
the corresponding iridium(IV) paramagnetic cations, as
shown by coupled electrochemical–ESR studies. AgBF4
oxidation in CH2Cl2 gives instead MeH and radical
species, having a ‘‘tucked-in’’ structure. Such species
* Corresponding author. Tel.: +39-50-918225; fax: +39-50-
918260.
E-mail address: div@dcci.unipi.it (P. Diversi).
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)00680-5

146
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
(L)2] (L=PMe3, PMe2Ph) [6]. As far as we know, only
the redox behaviour of the related [RuRCp(PPh3)2]
(R=Me, CH2Ph; Cp=h5-C5H5) was explored in the
1980s [7]. Since it was already known that such systems
are able to thermally activate arene CꢀH bonds [6c], we
reasoned that the presence of one-electron oxidants
could catalyse arene activation analogously to what had
been observed for the above-cited iridium and ruthe-
nium complexes [M(Me)2)(hn-CnMen)(L)] (M=Ir, n=
5; M=Ru, n=6; L=phosphine) [2,5]. An
encouragement in this direction was provided by Tilset
et al. [8] where the oxidation of the related
[Ru(Me)Cp*(CO)(PPh3)] is reported to give MeH by
hydrogen abstraction from a ligand spectator. While
our original hopes eventually proved misplaced, since
no ETC catalysis was observed, we have found that
these ruthenium alkyls show an interesting chemistry
and in particular react with NOBF4 giving a variety of
cationic nitrosyls. The results of these studies are re-
ported below.
reaction mixture is avoided and the alkyl derivative is
isolated by extraction and crystallisation.
We have obtained [Ru(Me)Cp*(PMe3)2] (2a) (70%),
[Ru(Me)Cp*(PMe2Ph)2] (2b) (68%) and [Ru(CH2-
CMe3)Cp*(PMe3)2] (3a) (40%) carrying out the alkyla-
tion reactions with an excess of Grignard (6–10 per
mole of ruthenium), and by treating the reaction mix-
ture with 1,4-dioxane in order to precipitate the magne-
sium salts prior the hydrolysis. On carrying out the
hydrolysis without this pre-treatment it came out that
the isolated product is the iodo derivative [Ru(I)-
Cp*(L)2], probably obtained by further reaction of the
ruthenium alkyl complex with iodo salts, as it had been
previously observed [6a]. The yields are in general
comparable to those reported in the literature.
2.2. Electrochemical oxidation of the ruthenium(II)
alkyl deri6ati6es [Ru(R)Cp*L2] (R=Me, L=PMe3 2a,
PMe2Ph 2b; R=CH2CMe3, L=PMe3 3a) and
[Ru(Me)Cp(PPh3)2] 4d
Fig. 1 shows the cyclic voltammetric behaviour of 2b
in dichloromethane solution, at low temperature (0°C)
[10].
2. Results and discussion
2.1. Preparation of the ruthenium(II) alkyl deri6ati6es
[Ru(R)Cp*L2] (R=Me; L=PMe3 2a, PMe2Ph 2b.
R=CH2CMe3; L=PMe3 3a)
Two oxidation processes are displayed, only the first
of which exhibits features of chemical reversibility.
Controlled potential coulometry (Ew=0.0 V) shows
that the first oxidation consumes one electron/molecule.
As a consequence of the exhaustive oxidation, the
original pale-yellow solution turns pale-pink and gives
rise to a cyclic voltammogram quite complementary to
the original one, thus confirming the chemical re-
versibility of the Ru(II)–Ru(III) 2b/2b+ process. Based
on the relative peak-heights, we assume that also the
second oxidation process is a one-electron process likely
involving the irreversible Ru(III)/Ru(IV) oxidation.
Complex 4d exhibits a voltammetric profile substan-
tially similar to that of 2b, but for the presence of a
minor spurious peak between the two main oxidation
processes, due to some impurity of [Ru(X)Cp(PPh3)2]
The alkyl ruthenium(II) complexes of general for-
mula [Ru(R)Cp(L)2] and [Ru(R)Cp*(L)2] (R=Me,
CH2CMe3, L=phosphine) [6] have been prepared by
alkylation of the corresponding chlorocompounds, for
which several synthetic methods are available in the
literature [9] depending on the nature of the phosphine
and cyclopentadienyl ligands.
In particular [Ru(Me)Cp*(PMe3)2] (2a) and [Ru-
(CH2CMe3)Cp*(PMe3)2] (3a) have been prepared (73–
76%) [6b] by using equimolar amounts of the appropri-
ate Grignard reagent. [Ru(Me)Cp*(PMe2Ph)2] (2b) has
been synthesised (78%) by employing an excess of LiMe
(3:1) [6c]. In all the above cases hydrolysis of the
1
(X=Cl or I) (as proved by the H-NMR spectrum).
Also in this case exhaustive one-electron oxidation fol-
lowed by cyclic voltammetric tests confirms the re-
versibility of the 4d/4d+ process. As a consequence of
the one-electron removal the original yellow solution
turns orange.
Fig. 2 shows the cyclic voltammogram of 3a,
recorded at low temperature.
As seen, in this case the first oxidation step is fol-
lowed by a series of minor peaks, which cannot be
attributed to impurities, in that they arise from decom-
position of the electrogenerated monocation 3a+. As a
matter of fact the 3a/3a+ process is not completely
reversible since in controlled potential electrolysis it
leads to the formation of new products just coincident
with the above spurious processes. As a proof of the
Fig. 1. Cyclic voltammogram recorded at a platinum electrode on
CH2Cl2 solution containing 2b (2.4×10−3 dm−3) and [NBu4]PF6
(0.2 dm−3). Temperature=0°C. Scan rate: 0.2 V s−1
.

P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
147
The electrode potentials for the oxidation processes
are consistent with the electronic characteristics of the
ligands with the exception of 2a and 3a, for which they
are identical (as above discussed, the figure for 2a might
be affected by error). The partial chemical reversibility
of the Ru(II)–Ru(III) process of some of the present
complexes is in agreement with previous findings on
related derivatives [7].
The Ru(III) systems 2b+ and 4d+, generated by
electrolysis a −20°C in solution of CH2Cl2 under inert
atmosphere, have been characterised by EPR spec-
troscopy at different temperatures. Fig. 3 shows the
X-band spectra at 100 K of 4d+ as an example.
The X-band EPR spectra of 2b+ and 4d+ at 100 K
exhibit an S=1/2 well-resolved rhombic structure
(gx\gy\gelectron=2.0023\gz) typical of low-spin
Ru(III) paramagnetic complexes [11]. The 4d+ species
evidences a significant superhyperfine resolution in the
gz area, with a triplet of relative intensity 1:2:1, deriving
from the coupling of the unpaired electron with two
magnetically equivalent phosphorus nuclei. The rele-
vant hyperfine couplings reported in Table 2 are com-
puter evaluated [12]. There is no evidence for Ru
satellite hyperfine couplings arising from the magnetic
interaction of the unpaired electron with the magneti-
cally active ruthenium nuclei.
Fig. 2. Cyclic voltammogram recorded at a platinum electrode on
CH2Cl2 solution containing 3a (1.0×10−3 dm−3) and [NBu4]PF6
(0.2 mol dm−3). Temperature=0°C. Scan rate: 0.2 V s−1
.
Table 1
Formal electrode potentials (in V vs. SCE) and peak-to-peak separa-
tions (in mV) for the oxidation processes of the Ru(II) complexes
under study in CH2Cl2 solution
a
Compound
T (°C)
E°%
(0/+)
DEp
Ep(+/2+)
a,b
2a
2b
2b
4d
3a
3a
20
0
20
20
20
0
−0.34
−0.27
−0.27
+0.09
−0.34 c
−0.34
95
92
90
+0.72 b
+0.69 b
+0.91 b
134 c
105
a Measured at 0.2 V s−1
b Peak potential for irreversible processes.
c Measured at 1 V s−1
.
Table 2 compiles the X-band EPR parameters as
derived from computer simulation of the corresponding
experimental parameters.
.
For both the electrogenerated 2b+ and 4d+ com-
plexes, the anisotropic lineshape disappears at the tran-
sition temperature from frozen to fluid solution
(T=178 K). 4d+ exhibits, up to 200 K, the correspond-
ing broad and unresolved isotropic signal, with giso
=
2.126(5) and DHiso=95(6) G, without any evidence for
the relevant phosphorus nuclei superhyperfine split-
tings. This is a consequence of the overall DHiso
=
105(5) G broadening largely overlapping the relevant
31P couplings, so that an upper limit for such a mag-
netic interaction can be evaluated: DHiso]2aiso(2P).
Temperatures higher than 200 K induce the isotropic
signal disappearing, and the solution remains EPR
mute in the whole temperature range explored (200–
300 K). This derives from the increase of the inter/in-
tramolecular dynamics which become extremely rapid,
rising the temperature, and results in a dramatic broad-
ening of the fluid solution signal [12]: actually refreez-
ing the solution below the fluid/glassy temperature
restores quantitatively the original signal. The related
giso and Žg values are in good agreement, suggesting
that the main geometry of the electrogenerated 4d+
complex is maintained in different experimental condi-
tions. On the contrary, in the case of the electrogener-
ated complex 2b+, the rhombic spectrum drops out at
the glassy-fluid transition phase and the paramagnetic
solution becomes EPR mute. The lineshape analysis of
Fig. 3. Experimental (upper) and simulated (lower) EPR spectra of
4d+ at 100 K.
instability of the monocation 3a+, the voltammogram
at room temperature of the 3a/3a+ process appears
irreversible and the associated reduction peak can be
detected only at high scan rate.
Finally, complex 2a undergoes an oxidation process
which is only in part chemically reversible, but the
electrochemical results are not completely reliable be-
cause of the low stability of the starting complex.
The redox potentials for all the above processes are
reported in Table 1.

148
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
the two electrogenerated 2b+ and 4d+ complexes
strongly points out the large metal-in character of the
actual EPR features and reflects the significant Ru(III)
spin–orbit contribution in the present spectra. Accord-
ingly, for both the complexes the singly occupied
molecular orbital (SOMO) results from the significant
participation of the atomic orbitals of the two phospho-
rus atoms but has a fundamental ruthenium 4d orbital
character, as indicated by the actual giso values and the
relevant linewidth DHiso. It is also to be noted that the
nature of the cyclopentadienyl ligand does not influence
substantially the shapes of the signals, the correspond-
ing DHiso, giso and aiso values being only slightly
affected.
the reaction is carried out in deuteriated solvents, hy-
drogen abstraction has to occur from one of the lig-
ands. This behaviour is analogous to that observed, in
the presence of catalytic amounts of one-electron oxi-
dants, for the related iridium and ruthenium systems
[Ir(Me)2Cp*(PMe3)] [2] and [Ru(Me)2Cp*(PMe)3] [5],
which, however, are able to activate aromatic CꢀH
bonds yielding intermolecular activation products. This
difference is perplexing, unless steric factors are crucial:
actually it is possible that the presence of two phos-
phines in the coordination sphere of the ruthenium
centres in the [Ru(R)Cp*(phosphine)2] systems, instead
of only one in the case of the iridium [2] and ruthenium
[5] cases cited above, inhibits the interaction with the
arene. In the thermal activation presumably this is
overcome by preliminary phosphine dissociation [6c].
2.3. Chemical oxidation of the ruthenium(II) alkyl
complexes [RuRCp*(L)2] (R=Me, L=PMe3 2a,
PMe2Ph 2b; R=CH2CMe3, L=PMe3 3a) and
[Ru(Me)Cp(PPh3)2] 4d
2.4. Synthesis of cationic nitrosyl deri6ati6es by
reaction of the chloro and alkyl ruthenium complexes
[Ru(X)(p5-C5Me5)(phosphine)2] (X=Cl, Me,
CH2CMe3) with NOBF4
When 2a and 3a were reacted with equimolar
amounts of [FeCp2]PF6 in C6D6, a gradual change in
the colour of the solution from yellow to maroon and
formation of a maroon solid was observed. The spec-
trum recorded after a few hours revealed the formation
of methane (l=0.14) (2a) or neopentane (l=0.89)
(3a) and ferrocene (l=4.00), both of them correspond-
ing to 15–20% of the starting product. No signals
attributable to new soluble organometallic products
were observed.
Compounds 2b and 4d react quite slowly with
ferrocenium as shown by the rate of formation of
ferrocene. Neither methane nor new soluble
organometallic products was observed.
These data, together with the electrochemical results,
suggest that the compounds undergo one-electron oxi-
We have recently reported [4] that the iridium(III)
dimethyl derivatives [Ir(Me)2Cp*(L)] (L=PMePh2,
PMe2Ph, PMe3) react with NOBF4 to give the cationic
methylnitrosyl compounds [Ir(Me)2Cp*(NO)]BF4 or
[Ir(Me)Cp*(NO)(L)](BF4)2 (L=PMePh2, PMe2Ph,
PMe3) depending on the NO+/Ir molar ratio employed.
We have now studied the reaction of the alkyl ruthe-
nium derivatives 2a, 2b and 3a with NOBF4 and we
have found an analogous behaviour: the nitrosyl com-
plexes [Ru(R)Cp*(NO)(L)]BF4 or
a
mixture of
[Ru(R)Cp*(NO)(L)]BF4 and [RuCp*(NO)(L)2](BF4)2
form depending on the phosphine and on the NO+/Ru
ratio (Scheme 1).
dation reactions to give the corresponding radical
Some previous work had been reported in the litera-
ture for the reaction of the chloro complexes
[M(Cl)Cp(PR3)2] (M=Ru, Os; R=Me, Ph) [13] and
[Ru(Cl)Cp*(PPh3)2] [9e] with NO+ to give the corre-
’
cations [Ru(R)Cp%(L)2] + (Cp%=C5Me5, C5H5) which,
depending on their thermal stability, may decompose to
RꢀH through the homolysis of the RuꢀC bond. Then
the unstable 2a+ and 3a+ decompose to give methane
and neopentane, respectively, while 2b+ and 4d+ are
further oxidised to give insoluble dicationic species.
In the case of 2a and 3a, since decomposition to
alkane occurs without incorporation of deuterium when
sponding dications [M(h5-C5R%)(NO)(L)2]2+ (R%=H,
5
Me; L=Me, Ph) by replacement of the chloride ion.
However nothing has been reported, as far as we know,
on the reaction of the corresponding alkyl compounds
with NO+.
Table 2
X-band EPR parameters of the electrogenerated compounds 2b+ and 4d+, recorded at liquid nitrogen temperature (T=100 K)
a
b
b
b
b,c
Species
gx
gy
gz
Žg
giso
ax
ay
az
Ža
2b+
4d+
2.256
2.230
2.111
2.145
1.982
1.984
2.116
2.120
23
19
26
21
31
24
27
21
2.126
a
Žg=(gx+gy+gz)/3.
b In Gauss.
c
Ža=(ax+ay+az)/3.

P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
149
together with other peaks for the isotopic cluster of
ruthenium, characterised by a mass difference of Dm/2.
1
The H-NMR spectrum in CD3NO2 shows the presence
of C5Me5 as a triplet (l=2.23, JHP=2.0 Hz) and two
2
PMe3 groups as a virtual triplet (l=2.0, JHP+4JHP
=
11.2 Hz). The presence of a linear terminally bonded
NO ligand is indicated by a strong IR absorption at
1845 cm−1. The spectrum contains also a large IR
band at 1060 cm−1 indicative of the presence of the
BF−4 anion.
An analogous behaviour was observed in the case of
2b which reacts with equimolar amounts or with an
excess of NOBF4 giving [Ru(Me)Cp*(NO)(PMe2-
Ph)]BF4 (7b) or 7b and [RuCp*(NO)(PMe2Ph)2](BF4)2
(5b), respectively. OPMe2Ph (d, l=2.06, JHP=13.5
Hz) was also formed.
Changing the nature of the alkyl group does not alter
the reaction course: 3a gives in fact [RuCp*(NO)-
(PMe3)2](BF4)2 (5a) and [Ru(CH2CMe3)(h5-C5Me5)-
Scheme 1.
1
(NO)(PMe3)]BF4. Monitoring the reaction by H-NMR
spectroscopy revealed the formation of OPMe3 and
neopentane (l=0.91).
In a typical experiment an equimolar amount of
NOBF4 was added to 2a in CH2Cl2. The reactant
dissolved causing gas evolution (NO) and rapid colour
change from maroon to deep-red. From this solution
an oil (7a) was obtained which is soluble in
dichloromethane, acetone, nitromethane, acetonitrile
and methanol. The 1H-NMR spectrum in CD2Cl2
showed the presence of an organometallic product
characterised by a C5Me5 group (l=1.94, JHP=1.7
Hz), a s-bonded methyl group (l=0.85, JHP=6.2 Hz)
In general the dicationic salt precipitates from the
dichloromethane solution and may be easily separated
from the soluble monocationic one. However in the
case of 5b and 7b the solubilities are not different
enough to allow their separation by simple decantation.
Only after several crystallisations from CH2Cl2/pentane
it was possible to obtain the two compounds in a pure
form.
In all the above reactions formation of the alkanes
(methane, neopentane) occurs without incorporation of
deuterium, then apparently the alkyl group extracts an
hydrogen atom from one of the groups in the coordina-
tion sphere of ruthenium. As demonstrated in the case
of related reactions of some iridium systems [2c,4b], it is
possible that the hydrogen derives from a Cp* methyl
group to give ‘‘a tucked-in’’ intermediate.
Another interesting aspect of these reactions is the
origin of phosphine oxide. It could be tempting to
attribute its formation to the adventitous presence of
dioxygen, but we disregard this hypothesis since we
have never observed the presence of phosphine oxides
in reactions of these systems with other reagents. A
second hypothesis is that the dissociated phosphine
reacts with NO as already observed in the literature
[14]:
and one PMe3 ligand (l=1.64,
JHP=10.9 Hz).
Trimethylphosphine oxide (d, l=1.85, JHP=13.5 Hz)
was also found. The IR spectrum shows two strong
bands at 1800 and 1059 cm−1, which indicate the
presence of coordinated NO and BF−4 , respectively. In
particular the value of the NO stretching indicates a
linear coordination of the nitrosyl ligand. The MS-IS
spectrum presents the typical isotopic cluster of peaks
(m/e 358 for the 102Ru isotope) corresponding to
[Ru(Me)Cp*(NO)(PMe3)]+ and a signal at m/e 87 for
BF−4 . Moreover the fragmentation of the molecular ion
indicates the loss of CH4 (m/e 342). On the basis of all
these data 7a has been formulated as [Ru(Me)Cp*
(NO)(PMe3)]BF4, which results from the displacement
of trimethylphosphine by NO+.
When two moles of NO+ per mole of complex were
used, the reaction followed a different course: methane
(CH4) was immediately formed (l=0.20, CD2Cl2), the
colour of the solution changed from maroon to deep-
red and a light-orange solid precipitated. From the
solution 7a was recovered. The precipitate, which is
soluble in acetone, nitromethane, acetonitrile,
methanol, analysed correctly for [RuCp*(NO)(PMe3)2]-
(BF4)2 (5a). The MS-IS spectrum indicates a bivalent
cation, showing a parent ion at m/2 (209.5=419/2)
PR3+2NO“N2O+OPR3
A third hypothesis implies an electron transfer from
the phosphine to NO+ to generate a phosphine radical
cation which in addition to dimerise may be intercepted
by solvent impurities such as ethanol or water acting as
nucleophiles to give the corresponding phosphine oxide
[15].

150
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
Finally we turned our attention to the reaction of the
cationic carbonyl derivative [RuCp*(CO)(PPh3)2]+ [9e]
via loss of the chloride. In fact, when the reaction was
monitored by 1H-NMR spectroscopy in CD3OD, we
observed a doublet (l=1.54, JHP=1.8 Hz) for the
C5Me5 protons and multiplets in the aromatic region
(l=7.35–7.65), whose integrated areas are in the 1:1
ratio. The product is only partially soluble in methanol,
corresponding chloro compounds of formula [Ru(Cl)-
Cp*(L)2] in order to investigate if also for this series the
NO+–Ru ratio influenced the reaction pattern. In the
only cases described in the literature, [Ru(Cl)Cp*-
(PPh3)2] [9e] and the related [Ru(Cl)Cp(PR3)2] (R=
PMe3, PPh3) [13], which are reported to give only the
corresponding dicationic compound by substitution of
the chloride ligand with NO+ [9e,13], no mention of
such dependence is done.
1
and well soluble in benzene. The H-NMR spectrum in
C6D6 showing a doublet at l=1.38 (JHP=1.8 Hz) and
multiplets at l=7.0–7.8, is identical to that reported
for [Ru(Cl)Cp*(CO)(PPh3)], prepared by a different
route [9e].
In all the reactions of the chloro compounds with
NOBF4, as well as for the alkyl series, the phosphine
displacement results in the formation of the corre-
sponding oxide.
Apparently the formation of the dicationic com-
pounds depends on the properties of the phosphine
ligands: in the series 1a, 1b and 1d the dicationic
product becomes less favoured as the ligand ability to
donate electrons to the metal decreases [16]; a similar
trend was found for the methyl derivatives 2a and 2b,
where the yields of the dication are 30 and 15%,
respectively (Scheme 1).
Finally, as for the two compounds of the Cp series,
i.e. [Ru(X)Cp(PPh3)2] (X=Cl, Me), the reaction with
NO+ appears to be insensitive to the NO+/Ru ratio,
although they give different products: in fact it has been
reported that the chloro derivative gives [RuCp(NO)-
(PPh3)2]2+ [13], while from the methyl derivative we
have obtained [Ru(Me)Cp(NO)(PPh3)]+ as the only
product.
The mechanism of the above substitution reactions is
apparently a complex one since, as already noted, for-
mation of gas (NO) and phosphine oxide is observed in
addition to the organometallic products. In general
only a few mechanistic studies [17,18] are reported in
the literature, and nothing has been said in the reports
dealing with the reaction of the ruthenium chloro com-
plexes [Ru(Cl)Cp(PR3)2] (R=Me, Ph) [13] and
[Ru(Cl)Cp*(PPh3)2] [9e] with NO+. In particular it has
not been established whether a direct ligand substitu-
tion by NO+ occurs or whether one-electron oxidation
of the starting chloro compound takes place and is then
followed by reaction of the oxidised product with NO
generated in the redox process.
In an effort to acquire some mechanistic insight the
reaction was monitored by EPR spectroscopy which
showed that odd-electron species are involved. Actu-
ally, when 1a and NOBF4 were mixed in CH2Cl2 in an
EPR tube, a signal arose after a few minutes at 203 K
in fluid solution, consisting of three lines due to the
hyperfine interaction of the unpaired electron with the
14N nucleus of the nitrosyl ligand (giso=1.974,
A(N)iso=18.78 G) (Fig. 4a). The EPR spectrum of the
frozen solution is rhombic and is shown in Fig. 4b
Indeed it was sorted out that the NO+/Ru ratio does
influence the nature of the products in the case of the
chloro compounds series also. By using one equivalent
of NO+, [Ru(Cl)Cp*(PMe3)2] (1a) gives both the chlo-
ride (5a) and phosphine (6a) substitution products
(Scheme 1), while 1b and 1d yield exclusively the mono-
cations 6b and 6d, respectively. Instead when two
equivalents of NO+ were used, 1b gives both
[Ru(Cl)Cp*(NO)(PMe2Ph)]+ (6b) and [RuCp*(NO)-
(PMe2Ph)2]2+ (5b). Surprisingly we have not been able
(even by using an excess of NO+) to prepare the
dicationic nitrosyl compound [RuCp*(NO)(PPh3)2]2+
which, as already mentioned, is reported [9e] to have
been prepared just by reacting [Ru(Cl)Cp*(PPh3)2] (1d)
with a slight excess of NO+ (NO+/Ru molar ratio=
1.35). We have used a NO+/Ru ratio=1.35 and 2.0,
obtaining in both cases a light-red solid, which was
identified as [Ru(Cl)Cp*(NO)(PPh3)]BF4 (6d) by 1H-
NMR and IR spectroscopy as well as by mass spec-
trometry (ion spray) and elemental analysis. The
1H-NMR spectrum in CD3COCD3 shows the C5Me5
resonance as a doublet at l=1.84 (JHP=2.2 Hz) and
the aromatic protons of the PPh3 ligand as a multiplet
at l=7.5–7.8, whose integrated areas are in the 1:1
ratio. The C5Me5 signal occurs at the same chemical
shift reported in the literature for the dication, but the
multiplicity is described as a triplet (JHPꢀ2 Hz) [9e].
The IR spectrum of 6d in nujol shows a strong band at
1806 cm−1 for the NO stretching, which is almost
identical to the reported value (1805 cm−1) for the
dicationic compound [9e]. Finally the mass spectrum in
methanol, although we were unable to detect the
mother ion, is consistent with that of 6d. In fact it
presents a peak at m/e 566 (the cluster of peaks has the
intensity expected from the contribution of the ruthe-
nium and chlorine isotopes), corresponding to
[Ru(Cl)Cp*(MeOH)(PPh3)]+, which can be rationalised
as deriving from 6d by NO displacement by methanol.
Even by reacting 1d in the presence of an excess of
triphenylphosphine (PPh3/Ru=3) with the aim to in-
hibit the displacement of the phosphine, we have ob-
tained 6d as the only product.
Then, at least in our hands, the loss of the phosphine
is the preferential reactivity pattern of 1d, as confirmed
also by the reaction with CO in methanol in the pres-
ence of NH4PF6, which is reported to give instead the

P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
151
Fig. 4. EPR experimental (upper) and simulated (lower) spectra of the species generated by reaction of [Ru(Cl)Cp*(PMe3)2] (1a) with NOBF4 in
CH2Cl2: (a) fluid solution; and (b) frozen solution.
together with the computer simulated spectrum [19].
The parameters are the following: gx=2.00520, gy=
1.99919, gz=1.91541; A(N)x=11.77 G, A(N)y=33.34
G, A(N)z=13.60 G, A(2P)x=7.04, A(2P)y=6.97 G,
A(2P)z=7.68 G.
sity of an excess of NO+ (a part from the case of 1a) in
order that the dicationic product forms. We have not
been able to observe a signal attributable to the other
crucial odd-electron species such as [Ru(Cl)Cp*-
’
(PMe3)2] +, but there may be strong reasons for this
These parameters are consistent with the structure
[RuCp*(NO)(PMe3)2] +, and are identical to those ob-
’
tained for the one-electron reduction product of 5a with
cobaltocene [20]. Finally on raising the temperature the
intensity of the signal decreased, disappearing at room
+
’
temperature. The detection of [RuCp*(NO)(PMe3)2]
is probably of some mechanistic significance since it is
possible that the formation of 5a and/or 6a occurs by
an electron-transfer mechanism as illustrated in Scheme
2. Here it is proposed that the starting chloro com-
pound undergoes a one-electron oxidation to A, as
already demonstrated for reactions of related iridium
compounds with NO+ [4b]. A can be trapped by NO to
give 6a or alternatively can eliminate HCl by extracting
a hydrogen from a Cp* methyl group giving a ‘‘tucked-
in’’ intermediate from which, after extraction of an
hydrogen from the solvent to restore the hapticity of
the Cp* ligand B is obtained. In principle B could form
’
also by direct loss of Cl , but we tend to disregard this
possibility in order to explain the behaviour of
+
’
[Ru(Cl)Cp(PPh3)2] for which a different intermediate
has been intercepted (see below). Then B could react
directly with a second equivalent of NO+ to give the
dicationic compound 5a, or alternatively could be inter-
cepted by NO to give the spectroscopically observed
intermediate C, which after oxidation by NO+ [17b]
generates the final product 5a.
Then B might be the crucial intermediate for the
formation of 5a, and the mechanism depicted in
Scheme 2 may obviously account for the general neces-
Scheme 2.

152
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
’
eliminate Cl because the route via the ‘‘tucked-in’’
intermediate is unavailable, and then undergoes substi-
tution of Cl− with NO . This could explain well the
’
formation of the dicationic product even when equimo-
lar amounts of NO+ are used. Alternatively it is possi-
ble that the classical even-electron substitution takes
place in addition to the odd-electron pathway. Finally
the above picture could explain also why
[Ru(Me)Cp(PPh3)2] gives only the phosphine substitu-
tion product: in fact formation of methane would be
prohibited.
Although further work is necessary in order to vali-
date these ideas, all the above results already show that
the mechanistic pathways of these substitution reac-
tions are less obvious than one could suspect, being
finely tuned by the nature of the ligands.
3. Conclusions
Fig. 5. EPR experimental (upper) and simulated (lower) spectrum of
the species generated by reaction of [Ru(Cl)Cp(PPh3)2] in CH2Cl2
with NOBF4 (frozen solution).
The chemical and electrochemical oxidation of the
ruthenium alkyl derivatives described here is reminis-
cent of what has been found for the iridium complexes
[Ir(Me)2Cp*(phosphine)] [2], apart from the inability of
these ruthenium systems to give ETC-catalysed CꢀH
bond activation. In particular the reaction with NO+
produces new cationic nitrosyl compounds that for-
mally derive from alkyl or phosphine substitution by
NO+. The corresponding chloro derivatives [Ru(Cl)-
Cp*(phosphine)2] undergo the formally analogous
chloro or phosphine substitution by NO+. In all these
reactions the nature of the products depends essentially
on the NO+/Ru ratio employed. The mechanism of the
reaction is intriguing, but EPR studies show that odd-
electron species are partially or fully involved even in
the seemingly simple case of chloride substitution by
NO+.
lack of observation: for instance the life time could be
quite short for the EPR acquisition times.
Furthermore, although we are not in possession of
spectroscopic evidences for the reactions of the corre-
sponding alkyl derivatives with NO+, we propose that
a similar mechanistic pattern can hold good. In fact
since the alkyl ruthenium derivatives respect to the
chloro compounds are more easily oxidised, the odd-
electron mechanism should be even more important.
A different intermediate was observed in the case of
the reaction of [RuClCp(PPh3)2] with NOBF4: a signal
arose in the fluid solution at g=2.13, but no hyperfine
coupling with 14N was observed. The frozen solution
spectrum (Fig. 5) has rhombic parameters: gx=
2.39573, gy=2.12136, gz=1.95085.
On the basis of computer simulation [19] we assign
this spectrum to the one-electron oxidation product of
the starting chloro compound [RuClCp(PPh3)2] +. The
’
4. Experimental
signal disappeared rapidly before reaching 0°C. On this
basis, and by considering that no cation radical equiva-
lent to the intermediate C of Scheme 2 was detected, we
propose the simplest mechanism for the formation of
the final dication from this radical species as shown in
Scheme 3.
The reactions and manipulation of organometallics
were carried out under dinitrogen or argon using stan-
dard techniques. The solvents were dried and distilled
prior to use. The compounds [Ru(Cl)Cp*(L)2] (L=
PMe3 (1a), PMe2Ph (1b)) were prepared from
[RuCl2Cp*]2 by reaction with an excess of phosphine,
according to the procedure described by Tilley [6b] for
the trimethylphosphino derivative (1a). Some problems
The reason why reactions of these chloro compounds
with NO+ involve different species is not clear: a
+
’
plausible hypothesis is that [RuClCp(PPh3)2] cannot
Scheme 3.

P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
153
with this method arise in the separation of
[Ru(Cl)Cp*(PMe2Ph)2] from the secondary product
trans-[Ru(Cl)2(PMe2Ph)4]: in fact when the phosphine
contains aromatic groups the formation of trans-
[RuCl2L4] becomes competitive (50–60%) and, since the
solubilities of the two products are not so different to
allow an easy separation by simple extraction, several
crystallisations are required. Then [Ru(Cl)Cp*(PMe2
Ph)2] (1b) was obtained in lower yield (27%) than
[Ru(Cl)Cp*(PMe3)2] (1a) (53%). As for [Ru(Cl)Cp*-
(PPh3)2] (1d) several routes have been reported
[6c,9b,d–i]. We have followed three different literature
procedures [9d,f,g,i] preparing a product having the
croanalisi of the Istituto di Chimica Organica, Facolta`
di Farmacia, University of Pisa.
4.1. Preparation of [Ru(Me)Cp*(PMe3)2] 2a
[Ru(Cl)Cp*(PMe3)2] (1a) (0.21 g, 0.496 mmol) was
treated under stirring with MeMgI (2.8 ml of a 0.91 M
solution in diethyl ether, 2.48 mmol). The mixture was
stirred for 24 h at room temperature (r.t.), then evapo-
rated to dryness. The residue was extracted with pen-
tane and treated with 1,4-dioxane (1 ml). After removal
of the magnesium salts the resulting solution was evap-
orated to dryness. The residue was extracted with pen-
tane and hydrolysed with water. The organic extracts
were dried over anhydrous sodium sulfate and the
solvent was removed in vacuo to give (0.14 g, 70%) of
[Ru(Me)Cp*(PMe3)2] (2a) as a pale-yellow oil [6c].
Anal. Found: C, 49.7; H, 9.3. Calc. for C17H36P2Ru: C,
1
following NMR data: H-NMR (C6D6): l 1.13 (t, 15H,
1
J
HP=1.5 Hz, C5Me5), 6.90–7.80 (m, 30H, PPh3); H-
NMR (THF-d8): l 1.05 (t, 15H, C5Me5), 7.03–7.43
(bm, 30H, PPh3). Contradictory spectroscopic data are
found in the literature: the C5Me5 resonance is reported
either as a singlet at l=1.33 [9b] or as a triplet
(JHP=1 Hz) at l=1.02 [9e] in CDCl3, and as a singlet
at l=1.03 in THF-d8 [9i]. We have found that the
product is quite unstable in solution: fortuitous or
deliberate introduction of air causes a gradual darken-
ing starting from the liquid surface in contact with the
atmosphere and a growing of several Cp* signals. In
particular the chlorocompound decomposes almost
completely in CDCl3 solution soon after its solubilisa-
tion, even in the absence of air. Then the literature
discrepancies for the 1H-NMR data in this solvent must
be attributed to this behaviour. [Ru(Cl)Cp(PPh3)2] (2d)
was prepared by direct synthesis from RuCl3,
triphenylphosphine and cyclopentadiene in refluxing
ethanol according to the published procedure [3a].
[Ru(Me)Cp(PPh3)2] was prepared according to the liter-
ature [5c,d]. AgBF4 and [FeCp2]PF6 were Aldrich prod-
ucts. NOBF4 (Aldrich product) was treated under
50.6; H, 8.9%. 1H-NMR (C6D6): l −0.29 (t, 3H,
2
J
HP=6.9 Hz, RuMe), 1.1 (vt, 18H, JHP+4JHP=10.6
Hz, PMe3), 1.7 (t, 15H, JHP=1.2 Hz, C5Me5); 1H-
NMR (CD2Cl2): l −0.75 (t, 3H, RuMe), 1.19 (vt, 18H,
PMe3), 1.67 (t, 15H, C5Me5).
4.2. Preparation of [Ru(Me)Cp*(PMe2Ph)2] 2b
[Ru(Cl)Cp*(PMe2Ph)2] (1b) (0.241 g, 0.44 mmol) was
treated under stirring with MgMeI (2.9 ml of a 0.91 M
solution in diethyl ether, 2.64 mmol). After 72 h at r.t.
the reaction mixture was worked out as above. A
pale-yellow oil (0.159 g, 68%) was obtained having the
same properties as [Ru(Me)Cp*(PMe2Ph)2] (2b) [6c].
1H-NMR (C6D6): l 0.05 (t, 3H, JHP=6.0 Hz, RuMe),
2
1.15 (vt, 6H, JHP+4JHP=7.2 Hz, PMe), 1.35 (vt, 6H,
2JHP+4JHP=7.4 Hz, PMe), 1.5 (t, 15H, JHP=1.4 Hz,
C5Me5), 7.05 (bm, 6H, Ph), 7.45 (bm, 4H, Ph); 1H-
NMR (C6D5CD3): l −0.07 (t, 3H, RuMe), 1.14 (vt,
6H, PMe), 1.32 (vt, 6H, PMe), 1.46 (t, 15H, C5Me5),
7.0–7.5 (bm, 10H, Ph).
1
2
vacuum before use. H-, 31P-, H-NMR spectra were
recorded on Varian Gemini 200 and VXR 300 instru-
ments. Mass spectroscopy was performed on a Perkin–
Elmer Sciex API III plus instrument. Mass values are
given for the 102Ru isotope. EPR spectra for the chem-
ical oxidation reactions were obtained by using EPR
Varian E 112 instrument equipped with a Varian E 257
for temperature control. The spectrometer was inter-
faced to an AST Premium 486/25 by means of a data
acquisition system capable of acquiring up to 500 000
12-bit samples s−1, including 32-bit add to memory,
thus giving on-line signal averaging and a software
package specially designed for EPR experiments [21].
Material and apparatus used for electrochemistry and
coupled EPR measurements have been described else-
where [22]. All the potential values are referred to the
saturated calomel electrode (SCE): under the present
experimental conditions the one-electron oxidation of
ferrocene occurs at +0.44 V (CH2Cl2). Elemental
analyses were performed by the Laboratorio di Mi-
4.3. Preparation of [Ru(CH2CMe3)Cp*(PMe3)2] 3a
Compound 1a (0.30 g, 0.708 mmol) was reacted with
(Me3CCH2)MeCl (6.94 ml of a 0.51 M solution in
diethyl ether, 3.54 mmol) following the same procedure
described for 2a to give 0.234 g (72%) of a yellow oil
having the same properties as [Ru(CH2CMe3)Cp*-
(PMe3)2] (2a) prepared as reported in the literature [6c].
Anal. Found: C, 55.4; H, 8.9. Calc. for C21H44P2Ru : C,
1
54.9; H, 9.6%. H-NMR (C6D6): l 1.01 (t, 2H, JHP
=
5.8 Hz, RuCH2), 1.15 (vt, 18H, JHP+4JHP=7.2 Hz,
2
PMe3), 1.30 (s, 9H, CMe3), 1.67 (t, 15H, JHP=1.4 Hz,
1
C5Me5); H-NMR (CD2Cl2): l 0.74 (t, 2H, RuCH2),
0.91 (s, 9H, CMe3), 1.28 (vt, 18H, PMe3), 1.68 (t, 15H,
C5Me5).

154
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
4.4. General procedure for the reaction of [RuRCp*L2]
(R=Me, L=PMe3 2a, PMe2Ph 2b; R=CH2CMe3,
L=PMe3 4a and [RuMeCp(PPh3)2] 4d with AgPF6 or
[FeCp2]PF6
Compound 5a: Anal. Found: C, 31.8; H, 5.9; N, 2.1.
Calc. for C16H33B2F8NOP2Ru: C, 32.5; H, 5.6; N 2.4%.
1H-NMR (CD3NO2): l 2.23 (t, 15H, JHP=2.0 Hz,
2
C5Me5), 2.00 (vt, 18H, JHP+4JHP=11.2 Hz, PMe3);
1H-NMR (CD3OD): l 2.23 (t, 15H, C5Me5), 2.00 (vt,
1
In an NMR tube the appropriate amount of
[FeCp2]PF6 or AgBF4 was added to 0.030 g of the
ruthenium complex dissolved in C6D6 or CD3CN (1
ml), and the reaction was followed by 1H-NMR
spectroscopy.
18H, PMe3); H-NMR (D2O): l 1.98 (t, 15H, C5Me5),
1
1.76 (vt, 18H, PMe3); H-NMR ((CD3)2CO): l 2.3 (t,
15H, C5Me5), 2.11 (vt, 18H, PMe3); 1H-NMR
(CD3CN): l 2.06 (t, 15H, C5Me5), 1.84 (vt, 18H, PMe3);
31P-NMR (CD3OD): l −1.97; 13C-NMR (CD3OD): l
10.86 (C5Me5), 18 (vt, 1JCP+3JCP=36.0 Hz, PMe3),
115.09 (C5Me5); FT-IR (Nujol) (cm−1): ws (NO) 1845;
ws (BF) 1060; IS-MS (CH3OH): m/e 209.5 [M]++; 87
[BF4]−.
4.5. Reaction of [Ru(Me)Cp*(PMe3)2] (2a) with
equimolar amounts of NOBF4: formation of
[Ru(Me)Cp*(NO)(PMe3)]BF4 7a
NOBF4 (0.044 g, 0.376 mmol) was added under
stirring to a solution of 2a (0.15 g, 0.372 mmol) in 10
ml of dichloromethane. After the mixing of the reac-
tants NOBF4 gradually dissolved to give a ruby solu-
tion and a moderate formation of gas. After an hour at
r.t., the solution was evaporated under vacuum to give
a residue which was washed repeatedly with pentane
and dried. A deep-red oil (0.125 g, 76%) was obtained
which was identified as [Ru(Me)Cp*(NO)(PMe3)]BF4
(7a). Anal. Found: C, 37.0; H, 5.9; N, 3.0. Calc. for
4.7. Preparation of [Ru(Me)Cp*(NO)(PMe2Ph)]BF4 7b
by reaction of [Ru(Me)Cp*(PMe2Ph)2] 2b with
equimolar amounts of NOBF4
The preparation is analogous to that of 7a: 0.102 g
(0.19 mmol) of 2b in 6 ml CH2Cl2 were treated with
0.023 g (0.19 mmol) of NOBF4 to give after 2 h at r.t.
a maroon oil (70%), identified as [Ru(Me)Cp*(NO)-
(PMe2Ph)]BF4 (7b). Anal. Found: C, 44.8; H, 6.1; N,
2.1. Calc. for C19H29BF4NOPRu: C, 45.1; H, 5.8; N,
1
1
C14H27BF4NOPRu: C, 37.8; H, 6.1; N, 3.2%. H-NMR
2.8%. H-NMR (CD2Cl2): l 1.03 (d, 3H, JHP=6.1 Hz,
(CDCl3): l 1.93 (d, 15H, JHP=1.7 Hz, C5Me5), 1.64 (d,
9H, JHP=10.9 Hz, PMe3), 0.86 (d, 3H, JHP=6.1 Hz,
RuMe), 1.84 (d, 3H, JHP=10.7 Hz, PMe), 1.86 (d, 3H,
J
HP=10.3 Hz, PMe), 1.73 (d, 15H, JHP=1.8 Hz,
1
C5Me5), 7.2–7.6 (bm, 5H, Ph); 1H-NMR ((CD3)2CO): l
1.84 (d, 3H, RuMe), 2.01 (d, 15H, C5Me5), 2.43 (d, 6H,
PMe), 2.47 (d, 6H, PMe), 7.5–7.9 (bm, 5H, Ph); FT-IR
(Nujol) (cm−1): ws (NO) 1834; ws (BF) 1060.
RuMe); H-NMR (CD2Cl2): l 1.94 (d, 15H, C5Me5),
1.64 (d, 9H, PMe3), 0.9 (d, 3H, RuMe); FT-IR (Nujol)
(cm−1): ws(NO) 1800; ws (BF) 1059; IS-MS (CH3OH):
m/e 358 [M]+; 342 [M−CH4]+; 87 [BF4]−.
4.6. Reaction of [Ru(Me)Cp*(PMe3)2] 2a with an
excess di-NOBF4: formation of
[RuCp*(NO)(PMe3)2](BF4)2 5a and
[Ru(Me)Cp*(NO)(PMe3)]BF4 7a
4.8. Reaction of [Ru(Me)Cp*(PMe2Ph)2] 2b with an
excess of NOBF4: Preparation of
[RuCp*(NO)(PMe2Ph)2](BF4)2 5b and
[Ru(Me)Cp*(NO)(PMe2Ph)]BF4 7b
NOBF4 (0.191 g, 1.64 mmol) was added under stir-
ring to a solution of 2a (0.33 g, 0.819 mmol) in 15 ml of
dichloromethane. Soon after mixing of the reagents
methane was evolved (l=0.20, CD2Cl2), while a ruby-
red solution formed with a dark-orange precipitate. The
mixture was stirred at r.t. for 1 h, then the solid was
separated from the solution, and washed with
dichloromethane, benzene, pentane, and finally dried
under vacuum. The residue was purified by adding
dropwise a concentrated solution in acetone (1 ml) to a
large excess of diethyl ether (10 ml). Pale-yellow crys-
tals (0.14 g, 30%) of [RuCp*(NO)(PMe3)2](BF4)2 (5a)
were obtained. The above solution was evaporated
under vacuum and the residue was washed with pen-
tane and dried under reduced pressure. The resulting oil
was purified by adding a concentrated solution in
dichloromethane to a large excess of diethyl ether to
give a deep-red oil (0.18 g, 50%) which was identified as
7a.
The reaction is analogous to that of 2a. 0.1 g (0.19
mmol) of 2b in 5 ml of CH2Cl2 were reacted with 0.044
g
(0.38 mmol) of NOBF4 to give after 2 h
[RuCp*(NO)(PMe2Ph)2](BF4)2 (5b) as a light-orange
solid (0.02 g, 15%) and [Ru(Me)Cp*(NO)(PMe2Ph)]BF4
(7b) as a dark-red oil (0.05 g, 50%). 5b Anal. Found: C,
43.1; H, 5.8; N, 2.0. Calc. for C26H37B2F8NOP2Ru: C,
1
43.6; H, 5.2; N, 1.96%. H-NMR (CD3)2CO): l 1.95 (t,
15H, JHP=2.0 Hz, C5Me5), 2.10 (vt, 6H, 2JHP+4JHP
=
10.8 Hz, PMe), 2.35 (vt, 6H, 2JHP+4JHP=11.0 Hz,
PMe), 7.6–8.1 (bm, 10H, Ph); FT-IR (Nujol) (cm−1):
ws (NO) 1800; ws (BF) 1055.
4.9. Reaction of [Ru(CH2CMe3)Cp*(PMe3)2] 3a with
an excess of NOBF4 in CH2Cl2: formation of
[RuCp*(NO)(PMe3)2](BF4)2 5a and
[Ru(CH2CMe3)Cp*(NO)(PMe3)]BF4
To 3a (0.22 g, 0.48 mmol) dissolved in CH2Cl2 (15

P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
155
ml) NOBF4 (0.112 g, 0.96 mmol) was added. The
mixing of the reagents is followed by a gas efferves-
cence and by the formation of an orange solid. The
colour of the solution changed from yellow-green to
deep-red. The reaction mixture was maintained under
stirring for 1 h. The solid precipitated was washed with
dichloromethane, benzene and pentane, dried and
purified by dropping an acetone solution to a large
excess of diethyl ether (11:1). 0.081 g (29%) of 5a were
obtained. The solution was evaporated at reduced pres-
sure and the residue was purified by dropping a
dichloromethane solution to an excess of diethyl ether
(20:1). 0.1 g (40%) of a deep-red solid, identified as
[Ru(CH2CMe3)Cp*(NO)(PMe3)]BF4, were obtained.
Anal. Found: C, 42.9; H, 6.8; N, 2.5. Calc. for
respect to the residual protons of the deuteriated sol-
vent. H-NMR (CD2Cl2): l 1.74 (d, 15H, JHP=2.2 Hz,
C5Me5), 2.07 (d, 3H, JHP=11.9 Hz, PMe), 2.17 (d, 3H,
1
J
HP=11.6 Hz, PMe), 7.5–7.9 (bm, 5H, PPh); IS-MS
(CH3OH): m/e 440 [M]+; 87 [BF4]−; FT-IR (Nujol)
(cm−1): ws (NO) 1809; ws (BF) 1059.
4.12. Reaction of [Ru(Cl)Cp*(PMe2Ph)2] 1b with an
excess of NOBF4: formation of
[RuCp*(NO)(PMe2Ph)2](BF4)2 5b and
[Ru(Cl)Cp*(NO)(PMe2Ph)]BF4 6b
NOBF4 (0.025 g, 0.21 mmol) was added to 1b (0.12 g,
0.22 mmol) dissolved in 20 ml of CD2Cl2. Soon after
the mixing of the reagents the colour changed from
orange to deep-red and an orange solid precipitated.
After 2 h of stirring at r.t., the solid was separated and
washed repeatedly with dichloromethane, benzene and
pentane and dried to give [RuCp*(NO)(PMe2Ph)2]-
(BF4)2 (5b) (0.016 g, 10%) as a light-orange solid. The
solution, previously decanted, was evaporated under
vacuum and the residue was washed with pentane and
dried under vacuum to give [Ru(Cl)Cp*(NO)(PMe2-
1
C18H35BF4NOPRu: C, 43.2; H, 7.0; N, 2.8%. H-NMR
(CD2Cl2): l 0.59 (d, 2H, JHP=7.8 Hz, CH2), 0.86 (s,
9H, CMe3), 1.83 (d, 9H, JHP=11.9 Hz, PMe3), 1.97 (d,
15H, JHP=2.1 Hz, C5Me5); FT-IR (Nujol) (cm−1): ws
(NO) 1806; ws (BF) 1062.
4.10. Reaction of [Ru(Cl)Cp*(PMe3)2] 1a with NOBF4:
formation of [RuCp*(NO)(PMe3)2](BF4)2 5a and
[Ru(Cl)Cp*(NO)(PMe3)]BF4 6a
1
Ph)]BF4 (6b) (0.08 g, 70%) as a dark-red oil. 5b H-
NMR (CD3)2CO): l 1.95 (t, 15H, JHP=2.0 Hz,
C5Me5), 2.10 (vt, 6H, 2JHP+4JHP=10.8 Hz, PMe),
2.35 (vt, 6H, 2JHP+4JHP=11.0 Hz, PMe), 7.6–8.1
(bm, 10H, Ph); FT-IR (Nujol) (cm−1): ws (NO) 1800; ws
(BF) 1055.
The reaction proceeds without variations by using a
molar ratio NO+/Ru of 1 or 2. As an example we
report the reaction carried out by using equimolar
amounts of reactants. NOBF4 (0.0065 g, 0.055 mmol)
was added under stirring to a solution of 1a (0.023 g,
0.055 mmol) in 1 ml of CH2Cl2. A gas evolved, the
solution underwent a rapid colour change from orange
to deep-red and a precipitate formed. After stirring 2 h
at r.t., the precipitate was separated from the solution,
washed with benzene and pentane, then dried. A beige
solid (0.008 g, 26%) was obtained identified as
[RuCp*(NO)(PMe3)2](BF4)2 (5a). Removal of the sol-
vent from the decanted CH2Cl2 solution, followed by
washing with pentane, gave [Ru(Cl)Cp*(NO)-
(PMe3)]BF4 (6a) (0.013 g, 50%) as a deep-red oil. 6a
1H-NMR (CD2Cl2): l 1.48 (d, 9H, JHP=12.9 Hz,
PMe3), 1.97 (d, 15H, JHP=1.9 Hz, C5Me5); IS-MS
(CH3OH): m/e 372 [M]+; 87 [BF4]−; FT-IR (Nujol)
(cm−1): ws (NO) 1806; ws (BF) 1059.
4.13. Reaction of [Ru(Cl)Cp*(PPh3)2] 1d with NOBF4
in CH2Cl2: formation of [Ru(Cl)Cp*(NO)(PPh3)]BF4 6d
The reaction proceeds without substantial differences
utilizing a molar ratio NO+/Ru=1 or 2. As described
for 6a, 0.180 g (0.23 mmol) of 1d were treated with
0.053 g (0.46 mmol) of NOBF4 in 11 ml of CH2Cl2 to
give light-red crystals of [Ru(Cl)Cp*(NO)(PPh3)]BF4
(6d) (0.117 g, 80%). Anal. Found: C, 52.26; H, 4.80; N,
2.05. Calc. for C28H30BClF4NOPRu: C, 51.6; H, 4.6; N,
1
2.15%. H-NMR ((CD3)2CO): l 1.84 (d, 15H, JHP=2.3
Hz, C5Me5), 7.5–7.8 (bm, 18H, PPh3); IS-MS
(CH3OH): m/e 566 [M−NO+MeOH]+; 87 [BF4]−;
FT-IR (Nujol) (cm−1): ws (NO) 1805, ws (BF) 1059.
4.11. Reaction of [Ru(Cl)Cp*(PMe2Ph)2] 1b with
equimolar amounts of NOBF4: formation of
[Ru(Cl)Cp*(NO)(PMe2Ph)]BF4 6b
4.14. Reaction of [Ru(Me)Cp(PPh3)2] (4d) with
NOBF4: formation of [Ru(Me)Cp(NO)(PPh3)]BF4
NOBF4 (0.004 g, 0.034 mmol) was added in an NMR
tube to a solution of of 1b (0.019 g, 0.034 mmol) in
CD2Cl2. A gas evolved and the colour changed from
orange to deep-red. After 3 h at r.t. a product formed
which was identified as [Ru(Cl)Cp*(NO)(PMe2Ph)]BF4
Reactions carried out using a NO+/Ru molar ratio
of 1 or 2 afford identical results. As an example we
describe the following procedure. 4d (0.012 g, 0.017
mmol) dissolved in CD2Cl2 (0.75 ml) was treated with
NOBF4 (0.004 g, 0.034 mmol). A gas was evolved and
a variation of colour from yellow to red was observed.
1
(6b) (92%) from the H-NMR spectrum. The yield was
evaluated by comparison of the C5Me5 integrated area
1
After 6 h at r.t., the H-NMR spectrum showed the

156
P. Di6ersi et al. / Journal of Organometallic Chemistry 626 (2001) 145–156
Lucherini, C. Pinzino, P. Zanello, J. Organomet. Chem. 584
(1999) 73.
quantitative formation of [Ru(Me)Cp(NO)(PPh3)]BF4.
The solution was evaporated under vacuum and the
residue was purified by addition of a solution in
CH2Cl2 to a large excess of diethyl ether to give pale-
yellow crystals (67%) of [Ru(Me)Cp(NO)(PPh3)]BF4.
1H-NMR (CD2Cl2): l 1.42 (d, 3H, JHP=5 Hz, RuMe),
5.7 (s, 5H, C5H5), 7.3–7.7 (bm, 15H, PPh3); FT-IR
(Nujol) (cm−1): ws (NO) 1811; ws (BF) 1066; IS-MS
(CH3OH): m/e 474 [M]+; 458 [M−CH4]+; 429 [M−
MeNO]+; 87 [BF4]−.
[5] A. Ceccanti, P. Diversi, G. Ingrosso, F. Laschi, A. Lucherini, S.
Magagna, P. Zanello, J. Organomet. Chem. 526 (1996) 251.
[6] (a) T. Blackmore, M.I. Bruce, F.G.A. Stone, J. Chem. Soc. A
(1971) 2376. (b) T.D. Tilley, R.H. Grubbs, J.E. Bercaw,
Organometallics 3 (1984) 274. (c) H. Lehmkuhl, M. Bellenhaum,
J. Grundke, H. Mauermann, C. Kru¨ger, Chem. Ber. 121 (1988)
1719 (and references cited therein).
[7] M.F. Joseph, J.A. Page, M.C. Baird, Organometallics 3 (1984)
1749.
[8] T. Aase, M. Tilset, V.D. Parker, J. Am. Chem. Soc. 112 (1990)
4974.
[9] (a) M. Bruce, C. Hameister, A.G. Swincer, R.C. Wallis, Inorg.
Synth. 28 (1991) 270. (b) P.M. Treichel, D.A. Komar, P.J.
Vincenti, Synth. React. Inorg. Met.-Org. Chem. 14 (1984) 383.
(c) N. Oshima, H. Suzuki, Y. Moro-oka, Chem. Lett. (1984)
1161. (d) T. Arliguie, B. Chaudret, J. Chem. Soc. Chem. Com-
mun. (1986) 985. (e) F.M. Conroy-Lewis, S.J. Simpson, J.
Organomet. Chem. 322 (1987) 221. (f) T. Arliguie, C. Border, B.
Chaudret, J. Devillers, R. Poiblanc, Organometallics 8 (1989)
1308. (g) M.S. Chinn, D.M. Heinekey, J. Am. Chem. Soc. 112
(1990) 5166. (h) P.J. Fagan, W.S. Mahoney, J.C. Calabrese, I.D.
Williams, Organometallics 9 (1990) 1843. (i) L. Luo, S.P. Nolan,
P.J. Fagan, Organometallics 12 (1993) 4305. (j) R.T. Lubia´n,
M.A. Paz-Sandoval, J. Organomet. Chem. 532 (1997) 17. (k)
D.C. Smith, Jr., C.M. Haar, L. Luo, C. Li, M.E. Cucullu, C.H.
Mahler, S.P. Nolan, Organometallics 18 (1999) 2357.
[10] E.R. Brown, J.R. Sandifer, in: B.W. Rossiter, J.F. Hamilton
(Eds.), Physical Methods of Chemistry: Electrochemical Meth-
ods, vol. Vol. 2, Wiley, New York, 1986 (Chap. 4).
4.15. EPR spectral studies of the reactions of the
chloro and methyl deri6ati6es with NOBF4
A weighed amount of metal complex (typically, 5
mg) and NOBF4 (1–2 mg) was placed into a quartz
tube (o.d. 3 mm; i.d. 2 mm) fitted with a quartz-Pyrex
joint and a Corning Rotaflo Teflon tap (DISA, Milan,
Italy). The tube was attached to a vacuum line and
degassed by standard vacuum/argon techniques; then it
was immersed in a dry ice-cooled acetone bath and
charged with dichloromethane. The quartz tube was
then introduced into the spectrometer cavity, pre-
cooled to the desired temperature, and the reagents
were allowed to mix gradually. The hyperfine coupling
constants and linewidths were obtained by computer
simulation of the EPR spectra.
[11] F.E. Mabbs, D. Collison, Electron Paramagnetic Resonance of d
Transition Metal Compounds. In: Studies in Inorganic Chem-
istry, vol. Vol. 16, Elsevier, NewYork, 1992.
[12] J.P. Lozos, B.M. Hoffman, C.G. Franz, QCPE 243 (1973) 11.
[13] M.I. Bruce, I.B. Tomkins, F.S. Wong, B.W. Skelton, A.H.
White, J. Chem. Soc. Dalton Trans. (1982) 687.
Acknowledgements
[14] (a) M. Halmann, L. Kugel, J. Chem. Soc. (1962) 3273. (b) A.R.
Middleton, G. Wilkinson, J. Chem. Soc. (1981) 1898.
[15] (a) S. Yasui, K. Shioji, M. Tsujimoto, A. Ohno, J. Chem. Soc.
Perkin Trans. 2 (1999) 855. (b) S. Yasui, M. Fujii, C. Kawano,
Y. Nishimura, A. Ohno, Tetrahedron Lett. 32 (1991) 5601.
[16] (a) C.A. Tolman, Chem. Rev. 77 (1977) 313. (b) R. Longhi, R.O.
Ragsdale, R.S. Drago, Inorg. Chem. 1 (1962) 768. (c) M.M.
Rahman, H.Y. Liu, K. Eriks, A. Prock, W.P. Giering,
Organometallics 6 (1987) 650.
[17] N.G. Connelly, J.D. Davies, J. Organomet. Chem. 38 (1972) 385.
[18] (a) K.G. Caulton, Coord. Chem. Rev. 14 (1975) 317. (b) N.G.
Connelly, W.E. Geiger, Chem. Rev. 96 (1996) 877.
[19] (a) The software was provided by the Illinois EPR Research
Center, NIH Division of research, Resonances Grant. RR01811.
(b) K.J. Mattson, R.B. Clarkson, R.L. Belford, in 11th Interna-
tional EPR Symposium, 30th Rocky Mountains Conference,
Denver, CO, August 1988.
We thank CNR (Rome) for financial support. P.Z.
acknowledges the technical assistance of Mrs G.
Montomoli.
References
[1] D. Astruc, Electron Transfer and Radical Processes in Transi-
tion Metal Chemistry, VCH Publishers, New York, 1995 (and
references cited therein).
[2] (a) P. Diversi, S. Iacoponi, G. Ingrosso, F. Laschi, A. Lucherini,
P. Zanello, J. Chem. Soc. Dalton Commun. (1993) 351. (b) P.
Diversi, S. Iacoponi, G. Ingrosso, F. Laschi, A. Lucherini, C.
Pinzino, G. Uccello-Barretta, P. Zanello, Organometallics 14
(1995) 3275. (c) F. de Biani Fabrizi, P. Diversi, A. Ferrarini, G.
Ingrosso, F. Laschi, A. Lucherini, C. Pinzino, G. Uccello-Bar-
retta, P. Zanello, Gazz. Chim. It. 126 (1996) 391.
[20] P. Diversi, M. Fuligni, C. Pinzino, unpublished results.
[21] R. Ambrosetti, D. Ricci, Rev. Sci. Instrum. 62 (1991) 2281;
EPR-ENDOR, ICQEM-CNR, 1992.
[22] D. Osella, M. Ravera, C. Nervi, C.E. Housecroft, P.R. Raithby,
P. Zanello, F. Laschi, Organometallics 10 (1991) 3253.
[3] A. Pedersen, M. Tilset, J. Am. Chem. Soc. 13 (1994) 4887.
[4] (a) P. Diversi, V. Ermini, G. Ingrosso, A. Lucherini, C. Pinzino,
F. Simoncini, J. Organomet. Chem. 555 (1998) 135. (b) P.
Diversi, F. Fabrizi de Biani, G. Ingrosso, F. Laschi, A.
.