Cluster growth reactions with selenido-carbonyl clusters - Synthesis, characterisation and theoretical study of the dimetallic closo clusters [WRu3(μ4-Se2(μ-CO)4(CO)6(L)2] (L = Phosphane) and of the donor-acceptor adduct [(CO)5W(μ4-Se)Ru3(μ3-Se)(CO)7{P(CH

Article abstract of DOI:10.1002/ejic.200300386

The open-triangular, nido clusters of the type [Ru₃(μ ₃-Se)₂-(CO)₇(L)₂] [L = PPh ₃, PPh₂(OMe), PPh₂(Me), P(p-MeO-C ₆H₄)₃] react at roo

Full text of DOI:10.1002/ejic.200300386

FULL PAPER
Cluster Growth Reactions with Selenido-Carbonyl Clusters ؊ Synthesis,
Characterisation and Theoretical Study of the Dimetallic closo Clusters
[WRu3(µ4-Se)2(µ-CO)4(CO)6(L)2] (L ؍ Phosphane) and of the Donor-Acceptor
Adduct [(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7{P(CH2Ph)Ph2}2]
Daniele Cauzzi,[a] Claudia Graiff,[a] Roberto Pattacini,[a] Giovanni Predieri,*[a]
Antonio Tiripicchio,[a] Samia Kahlal,[b] and Jean-Yves Saillard[b]
Keywords: Ruthenium / Tungsten / Molybdenum / Cluster compounds / Selenium
The open-triangular, nido clusters of the type [Ru3(µ3-Se)2-
(CO)7(L)2] (L PPh3, PPh2(OMe), PPh2(Me), P(p-MeO-
C6H4)3] react at room temperature with [W(CO)3(MeCN)3] to
give the dimetallic closo clusters [WRu3(µ4-Se)2(µ-CO)4-
(CO)6(L)2]. When L is P(CH2Ph)Ph2 the donor-acceptor ad-
duct [(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7{P(CH2Ph)Ph2}2] is ob-
with the fragment W(CO)5 through a selenido ligand. DFT
calculations, performed on the model species [WRu3(µ4-
Se)2(µ-CO)4(CO)6(PH3)2] and [(CO)5W(µ4-Se)Ru3(µ3-Se)-
(CO)7(PH3)2], showed that the computed total energy of the
latter is 1.1 eV lower than that of the former.
=
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
tained where the nido cluster [Ru3(µ3-Se)2(CO)7(L)2] interacts Germany, 2004)
Introduction
cogenide metal clusters and suitable organometallic frag-
ments.
As regards route (b), selenido clusters [Ru3(µ3-Se)2-
In recent years we have carried out systematic investi-
gations on the selenium-transfer reactions of tertiary phos- (CO)7(L)2] were considered to be ideal candidates for
phane and diphosphane selenides with carbonyl clusters growth reactions. These starting compounds are open-tri-
MЈ3(CO)12 (MЈ ϭ Fe or Ru).[1Ϫ9] These reactions take ad- angular, nido clusters with seven skeletal electron pairs
vantage of the frailty of the PϭSe bond and provide a sim- (SEPs), according to the requirements of the polyhedral
ple, sometimes selective,[2] synthetic route to phosphane- skeletal electron pair (PSEP) theory. As a consequence they
substituted, mono- and diselenido clusters. When the start- could be prone to add a zero SEP fragment, such as
ing phosphane contains heterocyclic groups, such as the 2- M(CO)3, derived from [M(CO)3(CH3CN)3] (M ϭ Mo or
pyridyl and 2-thienyl moieties, PϪC bond cleavage is ob- W), giving the corresponding closo clusters [MRu3(µ4-Se)2-
served, affording selenido-phosphido clusters coordinating (CO)10(L)2]. In this regard Adams[11] and co-workers ob-
to heterocyclic fragments.[5Ϫ7]
tained the monophosphane sulfur derivative of formula
The synthesis of phosphane-substituted mixed-metal se- [WRu3(µ4-S)2(CO)11(PMe2Ph)] upon irradiating a solution
lenido clusters has also been investigated as the presence of [Ru3(µ3-S)2(CO)9] and [W(CO)5(PMe2Ph)].
In a preliminary communication,[12] the products of
the reactions between [Ru3(µ3-Se)2(CO)7(PPh3)2] and
[M(CO)3(CH3CN)3] were erroneously identified as the un-
expected, centrosymmetric closo clusters [M2Ru2(µ4-Se)2(µ-
CO)4(CO)6(PPh3)2] (M ϭ W or Mo) on the basis of misin-
terpreted X-ray structural analyses. However, preliminary
MO calculations, mass spectral and accurate elemental
analysis data stimulated further crystallographic studies,
which revealed that a disorder exists in the crystals of these
compounds involving the mutually opposite Ru and W
atoms of the closo core of the cluster. This results in a dis-
ordered Ru3M(µ4-Se)2 cluster core instead of a centrosym-
metric M2Ru2(µ4-Se)2 one. Definitive evidence for this was
provided by the solution of the crystal structure of
[WRu3(µ4-Se)2(µ-CO)4(CO)6{P(OMe)Ph2}2] which does not
of different metals can influence the selectivity of certain
processes,[8,9] leading to the formation of compounds with
new and not always easily predictable structures. For the
synthesis of dimetallic substituted selenido-carbonyl clus-
ters two different synthetic strategies have been developed:
(a) MЈM ϩ PSe, i.e. reactions between preformed dimetallic
clusters and tertiary phosphane selenides,[10] and (b) MЈSe
ϩ M, i.e. cluster growth reactions between preformed chal-
[a]
Dipartimento di Chimica Generale ed Inorganica, Chimica
`
Analitica, Chimica Fisica, Universita di Parma,
Parco Area delle Scienze 17/A, 43100 Parma, Italy
Fax: (internat.) ϩ 39-0521/905-557
E-mail: predieri@unipr.it
[b]
´
LCSIM-UMR 6511, Institut de Chimie de Rennes, Universite
de Rennes 1,
35042 Rennes-Cedex, France
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072
DOI: 10.1002/ejic.200300386
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1063

G. Predieri et al.
FULL PAPER
Scheme 1. (Carbonyl groups omitted)
exhibit the disorder described above. Consequently, this were obtained in quantitative yield by selenium transfer
paper deals with the synthesis of five [MRu3(µ4-Se)2(µ- from elemental Se to the parent phosphane. They exhibit a
CO)4(CO)6(L)2] clusters [M ϭ Mo, L ϭ PPh3; M ϭ W, L ϭ characteristic PϭSe IR stretching band in the range
PPh2(Me), PPh2(OMe), P(p-MeO-C6H4)3] and the struc- 520Ϫ530 cmϪ1 and a 31P NMR signal with 77Se satellites,
tural characterisation of four of them. In the case of the 1JP,Se being in the range 710Ϫ740 Hz as normally found in
reaction involving P(CH2Ph)Ph2, we obtained [(CO)5W(µ4- similar species.[13]
Se)Ru3(µ3-Se)(CO)7{P(CH2Ph)Ph2}2], which can be con-
sidered as the product of a donor-acceptor interaction be-
tween [Ru3(µ3-Se)2(CO)7{P(CH2Ph)Ph2}2] and the W(CO)5
fragment through a selenido ligand. Moreover, DFT calcu-
The reaction between [Ru3(CO)12] and the tertiary phos-
phane selenides in hot toluene and in the presence of
Me3NO gave a number of substituted selenidoruthenium
carbonyl clusters (some of them, 8Ϫ14, already known) be-
longing to three different families: the monoselenido trinu-
clear species with a Ru3(µ3-Se) core (9, 11, 14; type I), the
diselenido trinuclear species with a Ru3(µ3-Se)2 core (1, 2,
4, 5, 8, 10, 12, 13; type II), and the diselenido tetranuclear
species with a Ru4(µ4-Se)2 core (3, 6, 7; type III. See
Scheme 1). As the reaction conditions are the same for all
phosphanes, it appears that the product distribution is in-
fluenced by the nature of the phosphane.
lations
compounds
have
been
performed
on
the
model
and
[WRu3(µ4-Se)2(µ-CO)4(CO)6(PH3)2]
[(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7(PH3)2]. Besides these
major results, some new phosphane selenides have been
characterised, and used to produce the selenido clusters
shown in Scheme 1. They can be grouped in three different
families according to the cluster cores: Ru3Se (I), Ru3Se2
(II) and Ru4Se2 (III). The crystal structures of two of them,
[Ru3(µ3-Se)2(CO)7(L)2] [L ϭ PPh2(OMe) or P(CH2Ph)Ph2],
have been determined by X-ray methods.
Clusters belonging to the type I family could be regarded
as the primary product of the attack of one molecule of
the phosphane selenide on the starting reagent [Ru3(CO)12],
resulting in the transfer of a selenium atom to the metal
Results and Discussion
The new tertiary phosphane selenides Ph2(Me)PSe, triangle and the substitution of one or two carbonyl groups.
Ph2(MeO)PSe, (p-MeOC5H4)3PSe and Ph2(PhCH2)PSe These clusters undergo a second attack by another phos-
1064
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072

Cluster Growth Reactions with Selenido-Carbonyl Clusters
FULL PAPER
phane selenide molecule affording, under appropriate con-
ditions, the corresponding diselenido derivatives (clusters of
type II). Mono- and disubstituted diselenido species are al-
ways present even if the starting Ru3/PSe molar ratio is 1:1.
Type II clusters can be described as square pyramids with
two ruthenium and two selenium atoms alternating in the
basal plane and the third ruthenium atom at the apex of
the pyramid. Mono- and disubstitution by phosphanes oc-
curs only at the base of the pyramid.
Type III clusters belong to the family of tetranuclear
closo clusters with seven SEPs. They can be described as
octahedral clusters with four ruthenium atoms in the equa-
torial plane and two selenium atoms at the apices. Two (3,
6) or three (7) phosphane ligands substitute the same num- structure of 10 with the atom numbering scheme
ber of carbonyl groups. The coordination around each ru-
Figure 2. View (ORTEP, 30% probability level) of the molecular
thenium atom is completed by carbonyl groups.
and 10 are shown in Figure 1 and 2, respectively. The most
Slow crystallisation of compounds 2 and 10 from di-
chloromethane solutions afforded well-formed crystals suit-
able for an X-ray diffraction analysis. ORTEP views of 2
important bond lengths and angles are given in Table 1 to-
gether with those of a cluster of formula [Ru3(µ3-Se)2-
(CO)7(PPh3)2].[2]
The cluster cores are similar to that of the analogous
compound [Ru3(µ3-Se)2(CO)7(PPh3)2][2] in which the
˚
RuϪRu bond lengths are 2.801(2) and 2.855(2) A; those of
˚
the six RuϪSe bonds range from 2.491(1) to 2.536(2) A.
˚
The RuϪRu bond lengths are 2.831(2) and 2.847(2) A in 2
˚
and 2.799(2) and 2.806(2) A in 10, whereas the RuϪSe
˚
bond lengths range from 2.485(1) to 2.529(2) A in 2 and
˚
from 2.449(2) to 2.500(2) A in 10. In 10 the two diphenyl-
benzylphosphane ligands coordinate to two ruthenium
atoms, substituting two carbonyl groups in pseudo-equa-
torial positions [the P1ϪRu1 and P2ϪRu3 bond lengths are
˚
2.309(2) and 2.306(3) A]. In 2 the two diphenylmeth-
oxyphosphane ligands substitute two carbonyls in a pseudo-
equatorial and pseudo-axial position. The P1ϪRu1 and
˚
P2ϪRu3 bond lengths [2.327(2) and 2.274(3) A] differ re-
markably because of the different trans influence of the
RuϪRu and RuϪSe bonds.
Clusters 2 and 10 show the same NMR behaviour in
solution despite the different coordination of the phos-
Figure 1. View (ORTEP, 30% probability level) of the molecular
structure of 2 with the atom numbering scheme
˚
Table 1. Comparison of selected bond lengths (A) and angles (°) for compounds 2, 10 and [Ru3(µ3-Se)2(CO)7(PPh3)2]
2
10
[Ru3(µ3-Se)2(CO)7(PPh3)2]
Ru(1)ϪRu(2)
Ru(2)ϪRu(3)
Ru(1)ϪSe(1)
Ru(1)ϪSe(2)
Ru(2)ϪSe(1)
Ru(2)ϪSe(2)
Ru(3)ϪSe(1)
Ru(3)ϪSe(2)
Ru(1)ϪP(1)
Ru(3)ϪP(2)
2.8312(15)
2.847(2)
2.8061(15)
2.7991(17)
2.4491(16)
2.4761(16)
2.4757(18)
2.500(2)
2.4497(15)
2.4704(16)
2.309(2)
2.855(2)
2.801(2)
2.514(2)
2.491(2)
2.523(2)
2.536(2)
2.514(2)
2.505(2)
2.371(4)
2.326(3)
2.4853(12)
2.5143(15)
2.5240(12)
2.5290(15)
2.4872(16)
2.4987(12)
2.327(2)
2.274(2)
2.306(3)
Ru(1)ϪRu(2)ϪRu(3)
Se(1)ϪRu(3)ϪSe(2)
Se(1)ϪRu(1)ϪSe(2)
Se(1)ϪRu(2)ϪSe(2)
Ru(1)ϪSe(1)ϪRu(3)
Ru(3)ϪSe(2)ϪRu(1)
83.59(5)
80.98(4)
80.71(4)
79.69(5)
99.11(4)
98.03(4)
83.96(6)
80.30(6)
80.20(6)
79.23(7)
99.87(6)
98.56(6)
85.6(1)
79.4(1)
79.7(1)
78.7(1)
100.6(1)
99.7(1)
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1065

G. Predieri et al.
FULL PAPER
phane ligands in the solid state (eq-ax and eq-eq respec- dinated in an axial position and the other in an equatorial
tively). The room temperature 31P NMR spectra of clusters position. As one ligand moves from the axial to the equa-
2 and 10, as well as 5 and 13, show two broad peaks which torial position, the other moves from equatorial to axial.
collapse at a temperature higher than 40 °C, together with This concerted movement of both ligands is responsible for
two sharp singlets (see Figure 3 for 13).
the observed broadening of the major peaks.
Cluster Growth Reactions
Type II compounds are open-triangular, nido clusters
with seven SEPs, according to the PSEP rules. As a conse-
quence they could be prone to add a zero SEP fragment,
such as M(CO)3 (M ϭ Mo or W), giving the closo clusters
MRu3(µ4-Se)2(CO)10(L)2. We reacted clusters 2, 5, 10 and
13 with an excess of W(CO)3(CH3CN)3 in dichloromethane.
The reactions take place in a few hours, at room tempera-
ture, giving compounds [WRu3(µ4-Se)2(µ-CO)4(CO)6-
{P(OMe)Ph2}2]
{P(Me)Ph2}2] (16), [WRu3(µ4-Se)2(µ-CO)4(CO)6{P(p-MeO-
C6H4)3}2] (17) and [(CO)5W(µ4-Se)Ru3(µ3-Se)-
(15),
[WRu3(µ4-Se)2(µ-CO)4(CO)6-
(CO)7{P(CH2Ph)Ph2}2] (20) as the only isolable prod-
ucts. The IR spectra of 15Ϫ17, in the carbonyl region are
superimposable with [WRu3(µ4-Se)2(µ-CO)4(CO)6(PPh3)2]
(18) and [MoRu3(µ4-Se)2(µ-CO)4(CO)6(PPh3)2] (19), sug-
gesting the same molecular structure for all five com-
pounds.
Clusters 15, 17, 18 and 19 were unequivocally identified
by X-ray diffraction methods (18 and 19 are isomorphous).
Three of these clusters, namely 17, 18 and 19, show a crys-
tallographically imposed Ci symmetry, so the mutually op-
posite Ru and M (M ϭ W or Mo) atoms of the closo core
of the cluster must be disordered and distributed in the two
positions with the same occupancy factor in such a way
that a disordered [Ru3M(µ4-Se)2] cluster core forms instead
of a centrosymmetric [M2Ru2(µ4-Se)2] core. Differently
from the clusters 17, 18 and 19, cluster 15 does not show
this type of disorder.
Figure 3. 31P{1H} NMR spectra of compound 13 recorded at dif-
ferent temperatures evidencing the incipient coalescence of the
broad signals due to the different coordination modes of the phos-
phane ligands; the presence of four signals confirms the existence
of the two isomeric forms A (broad peaks) and B (sharp peaks)
depicted in Scheme 2
The two sets of peaks suggest the existence of two iso-
mers A and B, indicated below in Scheme 2, as observed
for [Ru3(µ3-Se)2(CO)7(µ-dppm)] (dppm
ϭ
Ph2PCH2-
PPh2).[14] They should be involved in an exchange process
as occurs for the dppm derivative, consisting of the mi-
gration of a RuϪRu bond from one side of the open Ru3
triangle to the basal plane of the square pyramid (see
Scheme 2). For all these complexes, an equilibrium takes
place in solution starting from pure crystals of isomers A.
The prevalence of isomer A (broader peaks) in solution is
suggested by the fact that the basal metal sites are the pre-
ferred ones for phosphane substitution.[1] For both isomers
the presence of two peaks suggests that the two P ligands
are not equivalent. In fact, in B the two phosphanes occupy
two different Ru sites, whereas in A (where the two Ru sites
are equivalent) the presence of two peaks is probably the
consequence of the fact that one phosphane ligand is coor-
The structural diagrams of clusters 15Ϫ20 are shown in
Scheme 3.
Scheme 3. (Carbonyl groups omitted)
Because of the disorder, compounds 18 and 19 were pre-
viously identified as the six SEP species [M2Ru2(µ4-Se)2(µ-
CO)4(CO)6(L)2] (M ϭ W or Mo).[12] According to the
PSEP theory, the observed closo arrangement is predicted
to be unstable for electron-deficient species. This was con-
Scheme 2
1066
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072

Cluster Growth Reactions with Selenido-Carbonyl Clusters
FULL PAPER
firmed by DFT calculations on the isoelectronic models
[Mo2Ru2(µ4-Se)2(µ-CO)4(CO)6(L)2] (L ϭ CO, PH3) as-
suming the observed closo architecture (geometry optimis-
ation carried out assuming C2h and Ci symmetry for L ϭ
CO and PH3, respectively). Indeed, small HOMO-LUMO
gaps were computed for these models (0.24 and 0.44 eV for
L ϭ CO and PH3, respectively). Consistently, the excited
triplet state was found to lie close to the singlet ground state
(by 0.12 eV and 0.26 eV for L ϭ CO and PH3, respectively).
All these results are indicative of a thermal and kinetic in-
stability for [M2Ru2(µ4-Se)2(µ-CO)4(CO)6(L)2] (M ϭ W or
Mo), a result at variance with the behaviour of 18 and 19.
On the other hand, calculations on [Mo2Ru2(µ4-Se)2(µ-
CO)4(CO)6(L)2]2Ϫ and [Mo2Ru2(µ4-SeH)2(µ-CO)4(CO)6-
(L)2] (L ϭ CO, PH3) indicated thermal stability for the reg-
ular seven SEP count (HOMO-LUMO gap 1.92 eV (L ϭ
CO) and 1.81 eV (L ϭ PH3) for the former and 1.69 eV
(L ϭ CO) and 1.61 eV (L ϭ PH3) for the latter). Consist-
ently, further crystallographic studies revealed that the cor-
rect formula for 18 and 19 is [MRu3(µ4-Se)2(µ-CO)4-
(CO)6(L)2] (M ϭ W or Mo).
Figure 4. View (ORTEP, 30% probability level) of the molecular
structure of 15 with the atom numbering scheme
ORTEP views of 15, 17, 18 and 19 are shown in Figure 4,
5, 6 and 7, respectively. The most important bond lengths
and angles are given in Table 2 together with those calcu-
lated for the complex [WRu3(µ3-Se)2(CO)10(PH3)2].
The four closo seven SEP complexes can be described as
distorted octahedra in which the three ruthenium and the
M atoms lie in the basal plane with two selenium atoms at
the apices. The RuϪRu bond lengths span from 2.676(7) to
˚
2.835(1) A. The MϪRu bond lengths range from 2.852(6)
Figure 5. View (ORTEP, 30% probability level) of the molecular
˚
˚
to 2.939(2) A for M ϭ W and from 2.938(8) to 3.027(5) A structure of one of the two disordered complexes 17 with the atom
for M ϭ Mo. Two phosphane ligands coordinate two non-
numbering scheme
bonded rutheniums through the phosphorus atoms. Four
carbonyls asymmetrically bridge the RuϪRu and RuϪM
edges. The coordination around each metal atom is com- (CO)6(PH3)2] model (assuming C1 symmetry) lead to a
pleted by terminal carbonyl groups. large HOMO-LUMO gap (1.65 eV), as expected for a closo
DFT calculations on the [WRu3(µ4-Se)2(µ-CO)4-
˚
Table 2. Comparison of selected bond lengths (A) and angles (°) found for compounds 15, 17, 18, 19 and calculated for complex
[WRu3(µ3-Se)2(CO)10(PH3)2]
15
17
18
19
WRu3(µ3-Se)2(CO)10(PH3)2
Ru(1)ϪRu(2)
Ru(1)ϪRu(2Ј)[a]
Ru(2)ϪM(1)[b]
Ru(2Ј)ϪM(1)
M(1)ϪSe(1)
M(1)ϪSe(1Ј)
Ru(1)ϪSe(1)
Ru(1)ϪSe(1Ј)
Ru(2)ϪSe(1)
Ru(2Ј)ϪSe(1Ј)
Ru(2)ϪSe(1Ј)
Ru(2Ј)ϪSe(1)
2.8349(11)
2.7894(16)
2.9399(16)
2.8639(9)
2.6736(13)
2.7028(10)
2.5841(13)
2.540(4)
2.6196(14)
2.5878(14)
2.6067(13)
2.5644(11)
2.792(10)
2.676(7)
2.935(4)
2.851(6)
2.658(5)
2.679(5)
2.512(9)
2.5894(10)
2.5967(14)
2.5967(14)
2.5894(10)
2.5894(10)
2.826(4)
2.792(3)
2.814(2)
2.855(2)
2.614(3)
2.626(2)
2.614(4)
2.569(4)
2.5793(14)
2.5793(14)
2.5976(15)
2.5976(15)
2.720(5)
2.781(7)
2.938(8)
3.027(5)
2.855(5)
2.729(8)
2.581(7)
2.505(5)
2.5782(17)
2.5782(17)
2.5971(17)
2.5971(17)
2.878
2.870
2.940
2.992
2.775
2.775
2.650
2.620
2.686
2.686
2.685
2.685
Ru(1)ϪRu(2)ϪM(1)
Ru(1)ϪRu(2Ј)ϪM(1)
Ru(2)ϪRu(1)ϪRu(2Ј)
Ru(2)ϪM(1)ϪRu(2Ј)
90.29(4)
92.81(3)
90.18(4)
86.67(3)
87.46(19)
87.46(19)
93.8(2)
90.73(8)
90.73(8)
89.88(11)
88.83(7)
94.2(2)
94.2(2)
91.72(16)
82.87(14)
92
93
86
85
87.27(11)
[a]
Symmetry transformations used to generate equivalent atoms: Ϫx ϩ 1, Ϫy, Ϫz ϩ 2 (for 17); Ϫx, Ϫy, Ϫz (for 18 and 19); Ru(2Ј) ϭ
[b]
Ru(3) (for 15), Se(1Ј) ϭ Se(2) (for 15). M ϭ W for 15, 17 and 18, Mo for 19.
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1067

G. Predieri et al.
FULL PAPER
A view of cluster 20 is given in Figure 8. A list of the
most important bond lengths and angles are given in
Table 3.
Figure 6. View (ORTEP, 30% probability level) of the molecular
structure of one of the two disordered complexes 18 with the atom
numbering scheme
Figure 8. View (ORTEP, 30% probability level) of the molecular
structure of 20 with the atom numbering scheme
˚
Table 3. Selected bond lengths (A) and angles (°) found for com-
pound 20 compared with those calculated for the complex
[(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7(PH3)2]
20
[(CO)5W(µ4-Se)Ru3-
(µ3-Se)(CO)7(PH3)2]
Ru(1)ϪRu(2)
Ru(2)ϪRu(3)
W(1)ϪSe(1)
Ru(1)ϪSe(2)
2.802(2)
2.907
2.902
2.757
2.607
2.582
2.608
2.698
2.618
2.649
2.8526(13)
2.6852(10)
2.5008(15)
2.5215(13)
2.5379(13)
2.5390(12)
2.4950(13)
2.5203(15)
Figure 7. View (ORTEP, 30% probability level) of the molecular
structure of one of the two disordered complexes 19 with the atom Ru(1)ϪSe(1)
numbering scheme
Ru(2)ϪSe(2)
Ru(2)ϪSe(1)
Ru(3)ϪSe(2)
Ru(3)ϪSe(1)
seven SEP species. The major optimised bond lengths and
angles are given in Table 2. Taking into account that such
calculations are known to overestimate the bond lengths
Ru(1)ϪRu(2)ϪRu(3)
Ru(1)ϪSe(1)ϪRu(3)
87.01(6)
101.09(5)
102.38(5)
77.13(5)
78.25(5)
87
100
100
81
between heavy atoms by approximately 2Ϫ6%,[15] the op- Ru(1)ϪSe(2)ϪRu(3)
Se(1)ϪRu(1)ϪSe(2)
Se(1)ϪRu(3)ϪSe(2)
timised geometry of [WRu3(µ4-Se)2(µ-CO)4(CO)6(PH3)2] is
in good agreement with the X-ray structures of 15, 18 and
80
19.
Surprisingly, in the case of the reaction between [W-
It consists of a [Ru3(µ3-Se)2(CO)7{P(CH2Ph)Ph2}2] moi-
(CO)3(MeCN)3] and 4, we obtained the compound ety linked to a W(CO)5 unit through a coordinative interac-
[(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7{P(CH2Ph)Ph2}2] (20) as tion between a selenido ligand and the tungsten atoms
˚
the only isolable product identified by X-ray diffraction [Se1ϪW1 ϭ 2.685(1) A]. The [Ru3(µ3-Se)2] core unit, de-
methods. It could be considered as the product of a donor- scribed as a square pyramid with two ruthenium and two
acceptor interaction between a W(CO)5 fragment and a se- selenium atoms at the base of the pyramid and a ruthenium
lenido ligand of 10. The presence of similar products is de- atom at the apex, is similar to those of 10 and 2 with similar
˚
tectable on the preparative TLC plates in all the growth RuϪRu [2.802(2) and 2.853(1) A], and RuϪSe bond lengths
˚
reactions described above, as red bands attributable to ad- [ranging from 2.495(1) to 2.539(1) A]. The two diphenyl-
ducts (which decompose during elution) are always ob- benzylphosphane ligands substitute two ruthenium atoms
served. In the case of PPh2(OMe), it was even possible to at the base of the pyramid in a pseudo-equatorial and
isolate and recognize, by FT-IR spectroscopy, a cluster pseudo-axial position [P1ϪRu1 and P2ϪRu3 bond lengths
˚
analogous to 20, which, however, rapidly decomposed. The of 2.315(1) and 2.345(2) A]. The quadruply bridging selen-
decomposition product was identified as the starting disub- ido ligand (µ4-Se) acts as a six-electron donor, two of which
stituted nido cluster 2.
serve to attain an 18-electron configuration at the tungsten
1068
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org Eur. J. Inorg. Chem. 2004, 1063Ϫ1072

Cluster Growth Reactions with Selenido-Carbonyl Clusters
FULL PAPER
atom. An analogous unsubstituted compound of formula to decompose over silica gel in the air in such a way that
[Os3(µ3-S)(µ4-S)(CO)9W(CO)5] was previously reported by they cannot be obtained by TLC as mentioned above. Only
Adams et al.[16] In contrast to this osmium derivative, the in the case of P(CH2Ph)Ph2 were we able to isolate and
heptacoordinate Ru2 atom is bound to a carbonyl group characterise the corresponding adduct 20. Moreover, the
essentially trans with respect to the (µ4-Se) ligand yield of these adducts, which does not exceed 40%, is influ-
[Se1ϪRu2ϪC8 176.9(2)°]. In the parent Os compound the enced by the scarcity of free CO, which could be the limit-
carbonyl group is almost trans to the (µ3-S) ligand ing factor in the process of formation of the adduct mol-
[SϪOsϪC 166.4(2)°] and in that case the better donor ecules. Adams et al. observed that these adduct species can
character of the µ3-ligand with respect to the µ4-ligand has be prepared by UV irradiation of solutions of
been invoked. In the case of 20 the steric hindrance of the [Os3(CO)9(µ3-S)2] and W(CO)6, giving [Os3(CO)9(µ3-S)-
two phosphane ligands predominates over the electronic (µ4-S){W(CO)5}] in 70% yield.[16] This could be a more con-
factors and forces the system to be disposed in the ob- venient route to obtain higher yields of similar adducts.
served way.
As regards the mechanism for the production of 20,
we are far from a unique solution, but it could involve a
labile intermediate of formula [Ru3(µ3-Se)(µ4-Se)-
Experimental Section
(CO)7{P(CH2Ph)Ph2}2W(CO)3(CH3CN)2], formed from
[W(CO)3(MeCN)3] by loss of one acetonitrile ligand. Sub-
sequently, two MeCN ligands could be substituted by two
carbonyls. On the other hand, if the labile intermediate
loses two acetonitrile ligands, the anchored W(CO)3 frag-
ment can approach the Ru2Se2 basal plane of the pyramid
(i.e. the cavity of the nido polyhedron), leading to the pre-
viously described closo clusters. These reaction sequences
are depicted in Scheme 4. In the case of PPh2(CH2Ph), the
corresponding dimetallic closo cluster has not been ob-
tained, probably because of the steric hindrance of the
benzyl group, which does not facilitate the approach of the
W(CO)3 moiety.
General Procedures: The starting reagents [Ru3(CO)12], elemental
selenium, phosphanes and Me3NO are pure commercial products
(Aldrich and Fluka) and were used as received. The solvents (C.
Erba) were dried and distilled by standard techniques before use.
[Ru3Se2(CO)7(PPh3)2][2] M(CO)3(CH3CN)3 (M ϭ W or Mo)[17] and
clusters 8Ϫ14[18] were prepared according to literature procedures.
All manipulations (prior to the TLC separations) were carried out
under dry nitrogen by means of standard Schlenk-tube techniques.
Elemental (C, H) analyses were performed with a Carlo Erba EA
1108 automated analyzer. IR spectra (KBr discs or CH2Cl2 solu-
tions) were recorded on a Nicolet ‘‘Nexus’’ spectrometer. 1H
(300 MHz) and 31P (162.0 MHz, 85% H3PO4 as external reference)
NMR spectra (CDCl3 solutions) were recorded on Bruker instru-
ments, AC 300 (1H) and AMX 400 (31P). NICI Mass spectra
(Negative-Ion Chemical Impact) were recorded on a Fennigan
MAT SSQ710; m/z values are reported as the average of the iso-
topic distributions.
Synthesis of the Phosphane Selenides: The phosphane selenides
were obtained as white powders by reacting the corresponding
phosphanes with elemental selenium (in excess) in toluene at 90 °C
with vigorous stirring. The solutions were dried in vacuo and the
residues were extracted with dichloromethane. All products gave
satisfactory elemental analysis. Their purity was checked by 1H
NMR spectroscopy.
Ph2(PhCH2)PSe: Yield 95%. FT-IR (KBr) ν(PϭSe): 526s cmϪ1
.
31P
ϭ
1
NMR (CDCl3): δ ϭ 34.3 ppm (s, with 77Se satellites, JP,Se
715 Hz).
Ph2(MeO)PSe: Yield 93%. FT-IR (KBr) ν(PϭSe): 520s cmϪ1
.
31P
ϭ
NMR (CDCl3): δ ϭ 88.0 ppm (s, with 77Se satellites, JP,Se
1
715 Hz).
Scheme 4. (Carbonyl groups omitted)
(p-MeO-C6H4)3PSe: Yield 97%. FT-IR (KBr) ν(PϭSe): 526s cmϪ1
.
1
31P NMR (CDCl3): δ ϭ 31.5 ppm (s, with 77Se satellites, JP,Se
ϭ
718 Hz).
DFT calculations on the [(CO)5W(µ4-Se)Ru3(µ3-Se)-
(CO)7(PH3)2] model (assuming C1 symmetry) lead to the
optimised bond lengths and angles given in Table 3. They
are consistent with those computed for [WRu3(µ4-Se)2(µ-
CO)4(CO)6(PH3)2]. It is noteworthy that the computed total
energy of [(CO)5W(µ4-Se)Ru3(µ3-Se)(CO)7(PH3)2] is 1.1 eV
lower than that of [WRu3(µ4-Se)2(µ-CO)4(CO)6(PH3)2] ϩ
2CO. Although this value has to be considered as a rough
approximation of the enthalpy of the reaction [(CO)5W(µ4-
Ph2(Me)PSe: Yield 97%. FT-IR (KBr) ν(PϭSe): 528s cmϪ1
.
31P
ϭ
NMR (CDCl3): δ ϭ 23.3 ppm (s, with 77Se satellites, JP,Se
731 Hz).
1
Reaction of Phosphane Selenides with [Ru3(CO)12]
Reaction with Ph2(MeO)PSe: [Ru3(CO)12] (300 mg, 0.47 mmol) was
reacted in a 1:2 ratio with Ph2(MeO)PSe (277 mg, 0.94 mmol) in
toluene at 70 °C for 1.5 h in the presence of Me3NO (35 mg,
0.47 mmol) until the solution turned deep reddish brown. The sol-
vents were evaporated to dryness and the residue was redissolved
in a small amount of dichloromethane. TLC separation on silica,
Se)Ru3(µ3-Se)(CO)7(PH3)2]
Ǟ
[WRu3(µ4-Se)2(µ-CO)4-
(CO)6(PH3)2] ϩ 2CO it is indicative of thermal stability of
20. However, the adduct compounds are labile and prone using a 2:1 dichloromethane/hexane as eluent, yielded three bands
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072
www.eurjic.org  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 1069

G. Predieri et al.
FULL PAPER
together with some minor decomposition. The three bands con-
tained clusters [Ru3(µ3-Se)2(CO)8{P(OMe)Ph2}] (orange, 1; yield
15%), [Ru3(µ3-Se)2(CO)7{P(OMe)Ph2}2] (red, 2; yield 30%) and
[Ru4(µ4-Se)2(µ-CO)2(CO)7{P(OMe)Ph2}2] (deep red, 3; yield 25%)
in order of elution.
1987vs, 1894m, 1910m, 1831w, 1799w cmϪ1
298 K): δ ϭ 3.49 (d), 3.58 (d), 3.66 (d) and 3.73 (d, 6 H, JH,P
.
1H NMR (CDCl3,
3
ϭ
13.5 Hz, CH3) (8:4:1:1); 7.10Ϫ7.80 (m, 20 H, 4Ph) ppm. 31P{1H}
NMR (CDCl3, 298 K): δ ϭ 130.0 (s) ppm. MS (NICI): m/z (%) ϭ
1356 (95) [Ru3W(µ4-Se)2(CO)10{P(OMe)Ph2}2]ϩ, 1300 (100)
[Ru3W(µ4-Se)2(CO)8{P(OMe)Ph2}2]ϩ, 1272 (42) [Ru3W(µ4-Se)2-
1: FT-IR (CH2Cl2): ν(CO): 2077s, 2045vs, 2027s, 2005s, 1984sh
1
2
cmϪ1. H NMR (CDCl3, 298 K): δ ϭ 3.60 (d, JH,P ϭ 14.1 Hz, 3 (CO)7{P(OMe)Ph2}2]ϩ,
1244
{P(OMe)Ph2}2]ϩ, 1216 (21) [Ru3W(µ4-Se)2(CO)5{P(OMe)Ph2}2]ϩ,
298 K): δ ϭ 140.0 (br. s) ppm. C21H13O9PRu3Se2: calcd. C 28.0, H 1188 (16) [Ru3W(µ4-Se)2(CO)4{P(OMe)Ph2}2]ϩ, 1160 (28)
(15)
[Ru3W(µ4-Se)2(CO)6-
H, CH3), 7.4Ϫ7.9 (m, 10 H, 2Ph) ppm. 31P{1H} NMR (CDCl3,
1.46; found C 27.9, H 1.41.
[Ru3W(µ4-Se)2(CO)3{P(OMe)Ph2}2]ϩ, 1104 (25) [Ru3W(µ4-
Se)2(CO){P(OMe)Ph2}2]ϩ,
1076 (10) [Ru3W(µ4-
Se)2{P(OMe)Ph2}2]ϩ. C36H26O12P2Ru3Se2W: calcd. C 31.9, H 1.93;
1
2: FT-IR (CH2Cl2): ν(CO): 2047s, 2015vs, 1973s, 1954sh cmϪ1. H
2
NMR (CDCl3, 298 K): δ ϭ 3.42 (d), 3.59 (d) and 3.61 (d, JH,P
ϭ
12.1 Hz, 6 H, CH3), 7.85Ϫ7.20 (m, 20 H, 4Ph) ppm. 31P{1H} NMR found C 31.8, H 2.00.
(CDCl3, 298 K): δ ϭ 146 (s), 141 (s) (isomer B); 139 (br), 137 (br)
Synthesis of [WRu3(µ4-Se)2(µ-CO)4(CO)6{P(Me)Ph2}2] (16): Com-
(isomer A) ppm. C33H26O9P2Ru3Se2: calcd. C 36.4, H 2.41; found
C 36.3, H 2.41.
pound 5 (80 mg, 0.076 mmol) and [W(CO)3(CH3CN)3] (60 mg,
0.153 mmol) were stirred in dry CH2Cl2 for 1.5 h at room tempera-
ture under N2. The resulting dark solution was evaporated to dry-
ness and the residue was dissolved in CH2Cl2 (10 mL). The product
(cluster 16, yield 45%) was separated and purified by TLC on silica
gel, using diethyl ether/light petroleum (b.p. 40Ϫ60 °C) (2:1) as
eluent mixture. FT-IR (CH2Cl2): ν(CO): 2043m, 2007vs, 1978vs,
3: FT-IR (CH2Cl2): ν(CO): 2041w, 2012vs, 2000s, 1966m, 1844w,
1807w cmϪ1 1H NMR (CDCl3, 298 K): δ ϭ 3.58 and 3.66 (d,
.
2JH,P ϭ 13.2 Hz, 6 H, CH3), 7.26Ϫ7.64 (m, 20 H, 4Ph). 31P{1H}
NMR (CDCl3, 298 K): δ ϭ 139 (s) ppm. C35H26O11P2Ru4Se2:
calcd. C 33.8, H 2.11; found C 33.5, H 2.12.
1894m, 1910m, 1831w, 1800w cmϪ1 1H NMR (CDCl3, 298 K):
.
Reaction with Ph2(Me)PSe: [Ru3(CO)12] (300 mg, 0.47 mmol) was
reacted in a 1:2 ratio with Ph2(Me)PSe (260 mg, 0.94 mmol) in tolu-
ene at 70 °C for 2 h in the presence of Me3NO (35 mg, 0.47 mmol)
until the solution turned deep reddish brown. The solvents were
evaporated to dryness and the residue was redissolved in a small
amount of dichloromethane. TLC separation on silica gel, using 3:1
dichloromethane/hexane as eluent, yielded four bands: an orange, a
red, a brown and a deep red one in order of elution. The four bands
contained compounds [Ru3(µ3-Se)2(CO)8{P(Me)Ph2}] (4; yield
15%), [Ru3(µ3-Se)2(CO)7{P(Me)Ph2}2] (5; yield 30%), [Ru4(µ4-
Se)2(µ-CO)2(CO)7{P(Me)Ph2}2] (6; yield 10%) and [Ru4(µ4-Se)2(µ-
CO)(CO)7{P(Me)Ph2}3] (7; yield 25%).
δ ϭ 2.08 (d), 2.03 (d) and 2.01 (d, 2JH,P ϭ 8.5 Hz, 6 H CH3) (1:3:9),
7.4Ϫ7.5 (m, 20 H, 4Ph) ppm. 31P{1H} NMR (CDCl3, 298 K): no
observable signals. MS (NICI): m/z (%)
[Ru3W(µ4-Se)2(CO)10{P(Me)Ph2}2]ϩ,
1296
(38)
ϭ
1324 (100)
(17)
[Ru3W(µ4-
Se)2(CO)9{P(Me)Ph2}2]ϩ,
1268
[Ru3W(µ4-Se)2(CO)8-
{P(Me)Ph2}2]ϩ, 1240 (35) [Ru3W(µ4-Se)2(CO)7{P(Me)Ph2}2]ϩ,
1212 (32) [Ru3W(µ4-Se)2(CO)6{P(Me)Ph2}2]ϩ, 1184 (31) [Ru3W(µ4-
Se)2(CO)5{P(Me)Ph2}2]ϩ,
1156
(33)
[Ru3W(µ4-Se)2(CO)4-
{P(Me)Ph2}2]ϩ, 1128 (28) [Ru3W(µ4-Se)2(CO)3{P(Me)Ph2}2]ϩ,
1100 (20) [Ru3W(µ4-Se)2(CO)2{P(Me)Ph2}2]ϩ, 1072 (25) [Ru3W(µ4-
Se)2(CO{P(Me)Ph2}2]ϩ, 1044 (27) [Ru3W(µ4-Se)2{P(Me)Ph2}2]ϩ.
C36H26O10P2Ru3Se2W: calcd. C 32.7, H 1.98; found C 32.8, H 2.00.
4: FT-IR (CH2Cl2): ν(CO): 2077s, 2043vs, 2025s, 2066s, 1975m
cmϪ1. 1H NMR (CDCl3, 298 K): δ ϭ 2.49 (br., 3 H, CH3), 7.4Ϫ7.8
(m, 10 H, 2Ph) ppm. 31P{1H} NMR (CDCl3, 298 K): δ ϭ 37.2
(s) ppm. C21H13O8PRu3Se2: calcd. C 28.6, H 1.48; found C 28.6,
H 1.47.
Synthesis of [WRu3(µ4-Se)2(µ-CO)4(CO)6{P(p-MeO-C6H4)3}2] (17):
Compound 13 (100 mg, 0.073 mmol) and [W(CO)3(CH3CN)3]
(57.3 mg, 0.147 mmol) were stirred in dry CH2Cl2 for 1.5 h at room
temperature under N2. The resulting dark solution was evaporated
to dryness and the residue was dissolved in CH2Cl2 (10 mL). The
product (cluster 17, yield 45%) was separated and purified by TLC
on silica gel, using diethyl ether/light petroleum (b.p. 40Ϫ60 °C)
(2:1) as eluent mixture. FT-IR (CH2Cl2): ν(CO): 2039m, 2005vs,
1976s, 1888m, 1831w, 1800w cmϪ1. 1H NMR (CDCl3, 298 K): δ ϭ
5: FT-IR (CH2Cl2): ν(CO): 2045vs, 2009vs, 1967m, 1946w cmϪ1
.
1H NMR (CDCl3, 298 K): δ ϭ 2.20 (br), 2.32 (d) and 2.52 (d,
2JH,P ϭ 9.3 Hz, 6 H, CH3), 7.3Ϫ7.7 (m, 20 H, 4Ph) ppm. 31P{1H}
NMR (CDCl3, 298 K): δ ϭ 52 (s), 43 (s) (isomer B); 47 (br), 40
(br) (isomer A) ppm. C33H26O7P2Ru3Se2: calcd. C 37.6, H 2.48;
found C 37.3, H 2.41.
3
4
31
3.87 (s, 18 H, CH3), 6.92 (dd, JHα
ϭ 9, JHβ
31
ϭ 1.8 Hz, 12
,H
,
P
β
6: FT-IR (CH2Cl2): ν(CO): 2056w, 2038w, 2010vs, 1992sh, 1961sh,
H, HβAr), 7.29 (dd, 3JHα
ϭ 9, 3JH P) ϭ 10.8 Hz, 12 H, HαAr)
,
α
,H
β
1
1839w, 1800w cmϪ1. H NMR (CDCl3, 298 K): δ ϭ 2.07 (d) and
ppm. 31P{1H} NMR (CDCl3, 298 K): δ ϭ 39.0 (s) ppm. MS
2
2.12 (d, JH,P ϭ 9 Hz, 6 H, CH3), 7.3Ϫ7.6 (m, 20 H, 4Ph) ppm.
(NICI): m/z (%)
{P(p-MeO-C6H4)3}2]Ϫ,
{P(p-MeO-C6H4)3}2]ϩ,
{P(p-MeO-C6H4)3}2]ϩ,
{P(p-MeO-C6H4)3}2]ϩ,
{P(p-MeO-C6H4)3}2]ϩ,
{P(p-MeO-C6H4)3}2]ϩ,
ϭ
1361 (95) [Ru3W(µ4-Se)2(CO)10-
31P{1H} NMR (CDCl3, 298 K): δ ϭ 35.15 (s), 25.1 (s) (3:2).
C35H26O9P2Ru4Se2: calcd. C 34.7, H 2.16; found C 34.5, H 2.14.
7: FT-IR (CH2Cl2): ν(CO): 2017s, 1995vs, 1984s, 1955m, 1789w
cmϪ1. 1H NMR (CDCl3, 298 K): δ ϭ 2.02 (d) and 2.30 (d, 2JH,P ϭ
8.4 Hz, 9 H, CH3), 7.2Ϫ7.9 (m, 30 H, 6Ph) ppm. 31P{1H} NMR
(CDCl3, 298 K): δ ϭ 23.3 (s), 30.0 (s) (2:1). C47H39O8P3Ru4Se2:
calcd. C 40.8, H 2.84; found C 40.3, H 2.81.
1574
1546
1518
1490
1462
1432
(100)
(86)
(72)
(63)
(62)
(43)
[Ru3W(µ4-Se)2(CO)8-
[Ru3W(µ4-Se)2(CO)7-
[Ru3W(µ4-Se)2(CO)6-
[Ru3W(µ4-Se)2(CO)5-
[Ru3W(µ4-Se)2(CO)4-
[Ru3W(µ4-Se)2(CO)2-
{P(p-MeO-C6H4)3}2]ϩ, 1410 (32) [Ru3W(µ4-Se)2(CO{P(p-MeO-
C6H4)3}2]ϩ, 1382 (24) [Ru3W(µ4-Se)2{P(p-MeO-C6H4)3}2]ϩ.
C52H42O16P2Ru3Se2W: calcd. C 38.4, H 2.60; found C 38.0, H 2.60.
Synthesis of [WRu3(µ4-Se)2(µ-CO)4(CO)6{P(OMe)Ph2}2] (15):
Compound
2 (50 mg, 0.089 mmol) and [W(CO)3(CH3CN)3]
(50 mg, 0.128 mmol) were stirred in dry CH2Cl2 for 1 h at room Mass Spectroscopic data of [MRu3(µ4-Se)2(µ-CO)4(CO)6(PPh3)2]
temperature under N2. The resulting dark solution was evaporated [M ؍ W (18), Mo (19)]: The synthesis of the closo clusters
to dryness and the residue was dissolved in CH2Cl2 (10 mL). The [WRu3(µ4-Se)2(µ-CO)4(CO)6(PPh3)2] (18) and [MoRu3(µ4-Se)2(µ-
product (cluster 15, yield 40%) was separated and purified by TLC
on silica gel, using diethyl ether/light petroleum (b.p. 40Ϫ60 °C)
(2:1) as eluent mixture. FT-IR (CH2Cl2): ν(CO): 2043m, 2007vs,
CO)4(CO)6(PPh3)2] (19) has been reported previously.[12]
18: MS (NICI): m/z (%) 1449 (100) [Ru3W(µ4-
Se)2(CO)10{PPh3}2]ϩ, 1421 (20) [Ru3W(µ4-Se)2(CO)9{PPh3}2]ϩ,
ϭ
1070
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072

Cluster Growth Reactions with Selenido-Carbonyl Clusters
FULL PAPER
1393 (33) [Ru3W(µ4-Se)2(CO)8{PPh3}2]ϩ, 1365 (37) [Ru3W(µ4-
graphite monochromated Mo-Kα radiation, λ ϭ 0.71073 A), those
˚
Se)2(CO)7{PPh3}2]ϩ, 1337 (31) [Ru3W(µ4-Se)2(CO)6{PPh3}2]ϩ, of 15, 17, and 18 on a Philips PW 1100 single-crystal diffractometer
1309 (32) [Ru3W(µ4-Se)2(CO)5{PPh3}2]ϩ, 1281 (35) [Ru3W(µ4-
Se)2(CO)4{PPh3}2]ϩ, 1253 (22) [Ru3W(µ4-Se)2(CO)3{PPh3}2]ϩ, those of 19 on an Enraf Nonius CAD 4 single-crystal dif-
1225 (25) [Ru3W(µ4-Se)2(CO)2{PPh3}2]ϩ, 1197 (27) [Ru3W(µ4-
fractometer (graphite monochromated Cu-Kα radiation,
Se)2(CO){PPh3}2]ϩ, 1169 (21) [Ru3W(µ4-Se)2{PPh3}2]ϩ. C46H30O10- 1.54184 A). Crystallographic and experimental details for the struc-
(graphite monochromated Mo-Kα radiation, λ ϭ 0.71073 A) and
˚
λ
ϭ
˚
P2Ru3Se2W: calcd. C 38.1, H 2.07; found C 38.2, H 2.10.
19: MS (NICI): m/z (%) 1361 (100) [Ru3Mo(µ4-
tures are summarised in Table 4. An empirical correction for ab-
ϭ
sorption was made for 15, 17, 18 and 19 [maximum and minimum
Se)2(CO)10{PPh3}2]ϩ, 1333 (18) [Ru3Mo(µ4-Se)2(CO)9{PPh3}2]ϩ, value for the transmission coefficient was 1.000 and 0.7262 (15),
1305 (43) [Ru3Mo(µ4-Se)2(CO)8{PPh3}2]ϩ, 1277 (41) [Ru3Mo(µ4- 1.000 and 0.6872 (17), 1.000 and 0.5947 (18) and 1.000 and 0.460
Se)2(CO)7{PPh3}2]ϩ, 1249 (37) [Ru3Mo(µ4-Se)2(CO)6{PPh3}2]ϩ,
(19)].[19a,19b] For complexes 2, 10 and 20·2CH2Cl2 [maximum and
1221 (33) [Ru3Mo(µ4-Se)2(CO)5{PPh3}2]ϩ, 1193 (22) [Ru3Mo(µ4- minimum effective transmission value was 1.000 and 1.000 (2),
Se)2(CO)4{PPh3}2]ϩ, 1165 (34) [Ru3Mo(µ4-Se)2(CO)3{PPh3}2]ϩ, 1.000 and 0.8532 (10) and 1.000 and 0.5104 (20·2CH2Cl2)] the raw
1137 (18) [Ru3Mo(µ4-Se)2(CO)2{PPh3}2]ϩ, 1109 (31) [Ru3Mo(µ4- frame data were processed using SAINT and SADABS to yield
Se)2(CO){PPh3}2]ϩ,
1081
(23)
[Ru3Mo(µ4-Se)2{PPh3}2]ϩ. the reflection data file and the Bruker software was used for the
C46H30MoO10P2Ru3Se2: calcd. C 40.5, H 2.20; found C 40.6, H
2.22.
absorption correction.[19cϪ19e] The structures were solved by Pat-
terson and Fourier methods and refined by full-matrix least-
squares procedures (based on F2o) (SHELX-97)[20] first with iso-
tropic thermal parameters and then with anisotropic thermal par-
ameters in the last cycles of refinement for all the non-hydrogen
atoms. In the crystals of 20 dichloromethane molecules of solvation
were found. The hydrogen atoms were introduced into the geo-
metrically calculated positions and refined as riding on the corre-
sponding parent atoms.
In the crystals of 17, 18 and 19 a disorder exists which involves the
mutually opposite Ru and W atoms of the closo core of the cluster,
resulting in a disordered [Ru3M(µ4-Se)2] core, distributed in two
positions with the same occupancy factor, instead of the centro-
symmetric [M2Ru2(µ4-Se)2] core.
CCDC-208915 (2), -208916 (10), -208917 (15), -208920 (17),
-208918 (18), -208919 (19) and -208921 (20·2CH2Cl2) contain the
supplementary crystallographic data for this paper. These data can
be obtained free of charge at www.ccdc.cam.ac.uk/conts/
retrieving.html [or from the Cambridge Crystallographic Data
Centre, 12, Union Road, Cambridge CB2 1EZ, UK; Fax: (in-
ternat.) ϩ44-1223/336-033; E-mail: deposit@ccdc.cam.ac.uk].
Synthesis
of
the
Adduct
[(CO)5W(µ4-Se)Ru3(µ3-Se)-
(CO)7{P(CH2Ph)Ph2}2] (20): Compound 10 (100 mg, 0.083 mmol)
and [W(CO)3(CH3CN)3] (65 mg, 0.166 mmol) were stirred in dry
CH2Cl2 for 1 h at room temperature under N2. The resulting dark
solution was evaporated to dryness and the residue was dissolved
in CH2Cl2 (10 mL). The product (cluster 20, yield 40%) was sepa-
rated and purified by TLC on silica gel, using dichloromethane/
light petroleum (b.p. 40Ϫ60 °C) (1:1) as eluent. FT-IR (CH2Cl2):
ν(CO): 2071m, 2050vs, 2022vs, 1987m, 1972m, 1931s, 1895w cmϪ1
.
1H NMR: δ ϭ 3.81 (br., 4 H, CH2 Bz), 7.7Ϫ6.5 (m, 24 H, Ph)
ppm. 31P{1H} NMR: δ ϭ 43.5 (br.) ppm. C50H34O12P2Ru3Se2W:
calcd. C 39.2, H 2.24; found C 39.1, H 2.23.
Crystal Structure Determination of 2, 10, 15, 17, 18, 19 and
20·2CH2Cl2: Suitable crystals for the X-ray analysis for all com-
plexes were obtained by layering methanol on dichloromethane
solutions. The intensity data, at room temperature, of 2, 10 and
20·2CH2Cl2 were collected on a Bruker AXS Smart 1000 single
crystal diffractometer (equipped with an area detector using a
Table 4. Crystal data and structure refinement for 2, 10, 15, 17, 18, 19 and 20·2CH2Cl2
2
10
15
17
18
19
20·2CH2Cl2
Formula
Mol. wt.
Crystal system triclinic
C33H26O9P2Ru3Se2 C45H34O7P2Ru3Se2 C36H26O12P2Ru3Se2W C52H42O16P2Ru3Se2W C46H30O10P2Ru3Se2W C46H30MoO10P2Ru3Se2 C52H38Cl4O12P2Ru3Se2W
1089.61
1209.79
triclinic
¯
P1
1357.49
monoclinic
P21/n
1629.78
monoclinic
P21/c
1449.62
triclinic
¯
P1
1361.71
triclinic
¯
P1
1703.54
triclinic
¯
P1
¯
P1
Space group
˚
a (A)
8.792(3)
10.701(4)
21.163(5)
79.90(5)
85.36(5)
71.63(5)
1860(1)
2
11.804(5)
14.109(5)
14.885(5)
65.23(5)
77.34(5)
72.54(5)
2135(1)
2
14.533(4)
29.729(5)
9.572(3)
90
98.76(5)
90
12.054(3)
15.643(5)
14.458(4)
90
94.29(5)
90
10.102(5)
11.156(5)
11.205(5)
95.86(2)
97.33(2)
110.16(2)
1161(1)
1
10.229(5)
11.147(5)
11.151(5)
95.56(2)
97.55(2)
110.29(2)
1168(1)
13.819(3)
18.333(5)
13.259(3)
77.36(5)
66.38(5)
80.49(5)
2992(1)
˚
b (A)
˚
c (A)
α (°)
β (°)
γ (°)
3
˚
V (A )
4087(1)
4
2719(1)
2
Z
1
2
Dcalcd (g cmϪ3
F(000)
)
1.946
1.882
2.206
1.991
2.073
1.935
1.891
1052
1180
2568
1572
690
658
1636
Crystal size
µ, cmϪ1
Reflns. collected 19813
0.11 ϫ 0.12 ϫ 0.20 0.18 ϫ 0.22 ϫ 0.25 0.15 ϫ 0.21 ϫ 0.19
0.12 ϫ 0.15 ϫ 0.11
43.92
5945
0.14 ϫ 0.31 ϫ 0.15
51.18
5606
0.12 ϫ 0.21 ϫ 0.09
127.59
4393
0.22 ϫ 0.11 ϫ 0.20
41.64
43030
32.96
28.79
8705
58.11
7629
Reflns. unique 9941 (Rint ϭ 0.0545) 7003 (Rint ϭ 0.0527) 7183 (Rint ϭ 0.0456) 5945 (Rint ϭ 0.0245) 5606 (Rint ϭ 0.000)
4393 (Rint ϭ 0.000)
2991
16317 (Rint ϭ 0.0690)
9452
Obs. reflns.
[I Ͼ 2σ(I)]
R indices[a]
[I Ͼ 2σ(I)]
R indices
5117
3367
5539
3914
4072
R1 ϭ 0.0436
wR2 ϭ 0.0754
R1 ϭ 0.1152
R1 ϭ 0.0416
wR2 ϭ 0.0684
R1 ϭ 0.1129
R1 ϭ 0.0480
wR2 ϭ 0.1229
R1 ϭ 0.0671
R1 ϭ 0.0372
wR2 ϭ 0.0943
R1 ϭ 0.0719
R1 ϭ 0.0588
wR2 ϭ 0.1248
R1 ϭ 0.0847
R1 ϭ 0.0634
wR2 ϭ 0.1718
R1 ϭ 0.0936
R1 ϭ 0.0395
wR2 ϭ 0.0882
R1 ϭ 0.0946
(all data)
wR2 ϭ 0.0927
wR2 ϭ 0.0823
wR2 ϭ 0.1386
wR2 ϭ 0.1145
wR2 ϭ 0.1405
wR2 ϭ 0.2210
wR2 ϭ 0.1034
[a]
2
2 2
2 2
R1 ϭ Σ||Fo| Ϫ |Fc||/Σ|Fo|; wR2 ϭ {Σ[w(Fo Ϫ Fc ) ]/Σ[w(Fo ) ]}1/2
.
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1071

G. Predieri et al.
FULL PAPER
[10] [10a]
C. Graiff, A. Ienco, C. Massera, C. Mealli, G. Predieri, A.
Computational Details: Density functional theory (DFT) calcu-
lations were carried out using the Amsterdam density functional
(ADF) program[21] developed by Baerends and co-workers.[22] The
VoskoϪWilkϪNusair parameterization[23] was used for the local
density approximation (LDA) with gradient correction for ex-
change (Becke88)[24] and correlation (Perdew86).[25] Relativistic
corrections were added using the ZORA (Zeroth Order Regular
Approximation) scalar Hamiltonian.[26] The atom electronic con-
figurations were described by a triple-ξ Slater-type orbital (STO)
basis set for H 1s, C, O 2s and 2p, P 3s and 3p and Se 4s and 4p
augmented with a 3d single-ξ polarisation function for C, O, P,
with a 2p single-ξ polarisation function for H and 4d single-ξ
polarisation function for Se. A triple-ξ STO basis set was used for
Mo, Ru 4d and 5s and for W 5d and 6s augmented with a single-
ξ 5p for Mo and Ru and 6p for W. A double-ξ STO basis set was
used for W 5s and 5p. A frozen-core approximation was used to
treat the core electrons of C, O, P, Se, Mo, Ru and W.
[10b]
Tiripicchio, F. Ugozzoli, Inorg. Chim. Acta 2002, 330, 95.
´
P. Braunstein, C. Graiff, C. Massera, G. Predieri, J. Rose, A.
Tiripicchio, Inorg. Chem. 2002, 41, 1372.
[11]
[12]
R. D. Adams, T. A. Wolfe, W. Wu, Polyhedron 1991, 10, 447.
D. Cauzzi, C. Graiff, C. Massera, G. Mori, G. Predieri, A.
Tiripicchio, J. Chem. Soc., Dalton Trans. 1998, 321.
D. Cauzzi, C. Graiff, M. Lanfranchi, G. Predieri, A. Tiripic-
chio, J. Organomet. Chem. 1997, 536Ϫ537, 497.
D. Cauzzi, C. Graiff, G. Predieri, A. Tiripicchio, C. Vignali, J.
Chem. Soc., Dalton Trans. 1999, 237.
[13]
[14]
[15]
[16]
[17]
[18]
M. T. Garland, J.-F. Halet, J.-Y. Saillard, Inorg. Chem. 2001,
40, 3342.
R. D. Adams, I. T. Horvath, S. Wang, Inorg. Chem. 1985, 24,
1728.
H. Werner, K. Dechelmann, U. Schoenenbergher, Helv. Chim.
Acta 1970, 53, 2002.
D. Belletti, D. Cauzzi, C. Graiff, A. Minarelli, R. Pattacini, G.
Predieri, A. Tiripicchio, J. Chem. Soc., Dalton Trans. 2002,
3160.
[19] [19a]
N. Walker, D. Stuart, Acta Crystallogr., Sect. A 1983, 39,
[19b]
[19c]
Acknowledgments
Financial support from Ministero dell’Universita e della Ricerca
Scientifica e Tecnologica (Cofin 2000) is gratefully acknowledged.
The facilities of the Centro Interfacolta di Misure ‘‘G. Casnati’’
158.
F. Ugozzoli, Comput. Chem. 1987, 11, 109.
SMART Software Users Guide, Version 5.0; Bruker Analytical
X-ray Systems, Madison, WI, 1999. [19d] SAINT Software Users
Guide; Bruker Analytical X-ray Systems, Madison, WI, 1999.
`
`
[19e]
G. M. Sheldrick, SADABS, Bruker Analytical X-ray Sys-
`
(Universita di Parma) were used to record the NMR and mass
tems, Madison, WI, 1999.
[20] [20a] G. M. Sheldrick, SHELXL-93, a program for crystal struc-
ture determination; University of Göttingen, Germany, 1993.
spectra. Computing facilities were provided by the Centre de
Ressources Informatiques (CRI) of Rennes and the Institut de De-
veloppement et de Ressources en Informatique Scientifique du
Centre National de la Recherche Scientifique (IDRIS-CNRS).
´
[20b]
G. M. Sheldrick, SHELXL-97, Program for crystal struc-
ture refinement; University of Göttingen, Germany, 1997.
Amsterdam Density Functional program (ADF, 2000.02); Div-
[21]
ision of Theoretical Chemistry, Vrije Universiteit, De Boelelaan
1083, 1081 HV Amsterdam, The Netherlands; http://
www.scm.com.
[1]
D. Cauzzi, C. Graiff, G. Predieri, A. Tiripicchio, in Metal Clus-
ters in Chemistry (Eds.: P. Braunstein, L. Oro, P. Raithby), vol.
1, VCH, Weinheim 1999, 193.
P. Baistrocchi, D. Cauzzi, M. Lanfranchi, G. Predieri, A. Tirip-
icchio, M. Tiripicchio Camellini, Inorg. Chim. Acta 1995, 235,
173.
D. Cauzzi, C. Graiff, C. Massera, G. Predieri, A. Tiripicchio,
D. Acquotti, J. Chem. Soc., Dalton Trans. 1999, 3515.
D. Cauzzi, C. Graiff, M. Lanfranchi, G. Predieri, A. Tiripic-
chio, J. Chem. Soc., Dalton Trans. 1995, 2321.
D. Cauzzi, C. Graiff, C. Massera, G. Predieri, A. Tiripicchio,
Inorg. Chim. Acta 2000, 300Ϫ302, 471.
D. Cauzzi, C. Graiff, C. Massera, G. Predieri, A. Tiripicchio,
J. Cluster Sci. 2001, 12, 259.
D. Cauzzi, C. Graiff, C. Massera, G. Predieri, A. Tiripicchio,
[22] [22a] E. J. Baerends, D. E. Ellis, P. Ros, Chem. Phys. 1973, 2, 41.
[2]
[22b]
E. J. Baerends, P. Ros, Int. J. Quantum Chem. 1978, S12,
[22c]
169.
P. M. Boerringter, G. te Velde, E. J. Baerends, Int. J.
[22d]
Quantum Chem. 1988, 33, 87.
Int. J. Quantum Chem. 1992, 99, 84.
G. te Velde, E. J. Baerends,
[3]
[23]
S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200.
[4]
[24] [24a]
[24b]
A. D. Becke, J. Chem. Phys. 1986, 84, 4524.
A. D.
Becke, Phys. Rev. A 1988, 38, 3098.
[5]
[25]
J. P. Perdew, Phys. Rev. B 1986, 33, 8822.
[26] [26a] E. van Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys.
1993, 99, 4597. [26b] E. van Lenthe, E. J. Baerends J. G. Snijders,
[6]
[26c]
J. Chem. Phys. 1994, 101, 9783.
E. van Lenthe, R. van
[7]
Leeuwen, E. J. Baerends, Int. J. Quantum Chem. 1996, 57, 281.
Eur. J. Inorg. Chem. 2001, 721.
[26d] E. van Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys.
[8]
[26e]
´
P. Braunstein, J. Rose, Heterometallic Clusters in Catalysis, in
1996, 105, 6505.
J. Chem. Phys. 1999, 110, 8943.
E. van Lenthe, A. Ehlers, E. J. Baerends,
Metal Clusters in Chemistry (Eds.: P. Braunstein, L. A. Oro, P.
R. Raithby), VCH, Weinheim, 1999, pp. 616Ϫ677.
P. Braunstein, E. Sappa, A. Tiripicchio, Coord. Chem. Rev.
1985, 65, 219.
Received June 21, 2003
Early View Article
Published Online February 3, 2004
[9]
1072
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1063Ϫ1072