Oxidative addition of iodo-acetonitrile and of elemental halogens to [Pt3(μ-CO)3(PCy3)3]

Article abstract of DOI:10.1039/b415311a

The reaction of [Pt₃(μ-CO)₃(PCy₃) ₃] (1) with one mole-equivalent of iodo-acetonitrile was quantitative at - 70 °C giving the oxidative addition product [Pt ₃(μ-CO)₃(PCy₃)₃(I)(CH ₂CN)] (2). Fragmentation of 2 was observed in solution giving [Pt₂I(CH₂CN)(CO)₂(PCy₃) ₂] (3) which is the major product at room temperature if the starting cluster/reactant ratio is equal to or less than 1 to 1.5. Dimer 3 decomposes slowly in solution giving [Pt₂I₂(CO) ₂(PCy₃)₂] (4a) and succinonitrile. Monomer [PtI(CH₂CN)(CO)(PCy₃)] was the final product of the reaction when using excess of iodo-acetonitrile. The reactions of 1 with one mole-equivalent of halogens X₂ gave the new 44-electron clusters [Pt₃X(μ-CO)₂(μ-X)(PCy₃)₃] (X = I₂ (7a) or Br₂ (7b)) by oxidative addition followed by substitution of CO by X^-. Fragmentation of 7a and 7b took place in solution when using one and a half mole-equivalents of X₂ giving dimers 4a and [Pt₂Br₂(CO)₂(PCy ₃)₂] (4b) as well as [Pt₂X ₂(μ-X)₂(CO)₂(PCy₃)₂]. Monomers cis-[PtX₂(CO)(PCy₃)] were the final products of the reaction of 1 with excess of halogens. Insertion of SnCl₂ was observed into the Pt-Pt bond but not into the Pt-X bond, when equimolar amounts of SnCl₂·2H₂O were added to a solution of 4a or its chloro-analogue giving [Pt₂X₂(μ-SnCl ₂)(CO)₂(PCy₃)₂]. The Pt(I) dimers have unusually small J(Pt-Pt) values as observed by ¹⁹⁵Pt NMR and calculated by DFT. These values showed periodic changes comparing 4a and its analogues with other halides and mixed halide dimers.

Full text of DOI:10.1039/b415311a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Oxidative addition of iodo-acetonitrile and of elemental halogens to
[Pt3(l-CO)3(PCy3)3]
Zolta´n Be´ni,a Renzo Ros,*b Augusto Tassan,b Rosario Scopellitia and Raymond Roulet*a
a Institute of Chemical Sciences and Engineering, Swiss Federal Institute of Technology,
Lausanne, BCH, CH-1015, Lausanne, Switzerland. E-mail: raymond.roulet@epfl.ch
b Dipartimento di Processi Chimici dell’Ingegneria, Universita` di Padova, Via Marzolo 9,
I-35131, Padova, Italy
Received 4th October 2004, Accepted 12th November 2004
First published as an Advance Article on the web 30th November 2004
The reaction of [Pt3(l-CO)3(PCy3)3] (1) with one mole-equivalent of iodo-acetonitrile was quantitative at −70 ◦C giving
the oxidative addition product [Pt3(l-CO)3(PCy3)3(I)(CH2CN)] (2). Fragmentation of 2 was observed in solution giving
[Pt2I(CH2CN)(CO)2(PCy3)2] (3) which is the major product at room temperature if the starting cluster/reactant ratio is
equal to or less than 1 to 1.5. Dimer 3 decomposes slowly in solution giving [Pt2I2(CO)2(PCy3)2] ( 4a) and succinonitrile.
Monomer [PtI(CH2CN)(CO)(PCy3)] was the final product of the reaction when using excess of iodo-acetonitrile. The
reactions of 1 with one mole-equivalent of halogens X2 gave the new 44-electron clusters [Pt3X(l-CO)2(l-X)(PCy3)3]
(X = I2 ( 7a) or Br2 ( 7b)) by oxidative addition followed by substitution of CO by X−. Fragmentation of 7a and 7b took
place in solution when using one and a half mole-equivalents of X2 giving dimers 4a and [Pt2Br2(CO)2(PCy3)2] ( 4b) as
well as [Pt2X2(l-X)2(CO)2(PCy3)2]. Monomers cis-[PtX2(CO)(PCy3)] were the final products of the reaction of 1 with
excess of halogens. Insertion of SnCl2 was observed into the Pt–Pt bond but not into the Pt–X bond, when equimolar
amounts of SnCl2·2H2O were added to a solution of 4a or its chloro-analogue giving [Pt2X2(l-SnCl2)(CO)2(PCy3)2].
The Pt(I) dimers have unusually small J(Pt–Pt) values as observed by 195Pt NMR and calculated by DFT. These
values showed periodic changes comparing 4a and its analogues with other halides and mixed halide dimers.
to reproduce tendencies in the observed, unusually small J(Pt–
Pt) values and to verify stereochemistries in solution.
Introduction
Oxidative addition reactions involving mononuclear platinum
complexes are widely studied since they are elementary steps
Results and discussion
in some catalytic cycles.1–4 In contrast not much is known on
how platinum clusters act in this type of reaction.5–7 Continuing
a study of the reactivity of triangulo phosphino-carbonyl
platinum clusters,8–12 we report here on reactions with the
electrophilic reagents iodo-acetonitrile and elemental halogens.
Iodo-acetonitrile was chosen for being more reactive than methyl
iodide which reacts extremely slowly at low temperature with
[Pt3(l-CO)3(PCy3)3] (1). It is already known that elemental halo-
gens react readily with clusters isostructural with 1 giving Pt(I)
dimers and Pt(II) monomers,13,14 but no adduct complex having
an intact Pt triangle has been reported so far. We investigated
the reactions of 1 with I2 and Br2 targeting these clusters and
the synthesis of [Pt2I2(CO)2(PCy3)2] (4a) which was observed
as a decomposition product of [Pt2I(CH2CN)(CO)2(PCy3)2] (3)
formed in the reaction of 1 with iodo-acetonitrile.
On the basis of X-ray analysis and in situ low temperature NMR
and IR spectroscopic studies the following reaction schemes
could be drawn for the reactions of [Pt3(l-CO)3(PCy3)3] (1) with
iodo-acetonitrile (Fig. 1) and iodine (Fig. 2).
The reaction of 1 with one mole-equivalent iodo-acetonitrile
is quantitative at −70 ◦C, giving the oxidative addition
product [Pt3(l-CO)3(PCy3)3(I)(CH2CN)] (2). Upon warm-
ing, fragmentation occurred and mixtures of 2 and dimer
[Pt2I(CH2CN)(CO)2(PCy3)2] (3) were observed, besides starting
cluster 1. Dimer 3 was the major product of the reaction at room
temperature if the starting cluster/reactant ratio was equal to or
less than 1 to 1.5. Upon standing in solution slow decomposition
of 3 occurred giving the Pt(I) dimer [Pt2I2(CO)2(PCy3)2] (4a) and
succinonitrile. Excess of the electrophilic reagent converted 3
to Pt(II) monomer [PtI(CH2CN)(CO)(PCy3)] (5), which was the
final product of the reaction. An in situ NMR study indicated
that 4a slowly reacted with ICH2CN and showed that this
reaction was complete at 60 ◦C giving cis-[PtI2(CO)(PCy3)]
besides succinonitrile. Formation of the latter product suggests
that scission of the Pt–CH2CN bond is homolytic.
The irregular behaviour of J(Pt–Pt) with respect to metal–
metal distances was the object of a recent theoretical study.15
This study indicated that in Pt(I) dimers bearing CO and Cl−
ligands, r interactions with CO ligands in axial positions were
responsible for the reduced coupling constants. A DFT study
was undertaken on model molecules of new Pt(I) dimers in order
Fig. 1 Proposed reaction scheme for the reaction of 1 with iodo-acetonitrile.
T h i s j o u r n a l i s T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5 D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 1 5
©

View Article Online
◦
˚
Table 1 Selected distances (A) and angles ( ) of 2
Pt(1)–Pt(2)
Pt(1)–Pt(3)
Pt(2)–Pt(3)
Pt(1)–P(1)
Pt(2)–P(2)
Pt(3)–P(3)
Pt(1)–C(4)
2.7369(6)
2.7340(6)
2.6864(6)
2.360(3)
2.287(3)
2.273(3)
2.177(12)
Pt(1)–I(1)
Pt(1)–C(1)
Pt(2)–C(1)
Pt(2)–C(2)
Pt(3)–C(2)
Pt(1)–C(3)
Pt(3)–C(3)
2.7289(10)
2.330(12)
1.925(12)
2.055(12)
2.159(12)
2.288(11)
1.983(12)
Fig. 2 Proposed reaction scheme for the reaction of 1 with iodine.
Pt(3)–Pt(1)–Pt(2)
Pt(2)–Pt(3)–Pt(1)
Pt(3)–Pt(2)–Pt(1)
C(4)–Pt(1)–Pt(2)
I(1)–Pt(1)–Pt(3)
P(1)–Pt(1)–I(1)
C(4)–Pt(1)–Pt(3)
C(4)–Pt(1)–I(1)
C(4)–Pt(1)–C(1)
58.819(16) C(4)–Pt(1)–P(1)
60.645(16) P(1)–Pt(1)–Pt(2)
60.536(16) P(1)–Pt(1)–Pt(3)
94.0(3)
152.90(7)
145.77(7)
155.71(8)
143.76(8)
152.20(8)
146.85(8)
79.38(2)
98.4(3)
The reactions of 1 with one mole equivalent of elemental
halogens (I2 or Br2) are more complex. After a possible
oxidative addition step substitution of a CO ligand by a X−
takes place giving the new triangulo-clusters [Pt3X(l-X)(l-
CO)2(PCy3)3] (X = I2 (7a) or Br2 (7b)) bearing both bridging
and terminal halide ligands. Besides these clusters, dimers
4a or [Pt2Br2(CO)2(PCy3)2] (4b) and the already known16–23
Pt(II) dimer [Pt2X2(l-X)2(PCy3)2](X = I, 8) were observed
as fragmentation products. Dimers 4a or 4b were the major
products in dichloromethane when reacting 1 with one and a
half mole equivalents of halogens. These dimers are analogues of
complexes synthesized from clusters isostructural with 1 having
phosphine ligands other than PCy3.14,24 The chloro-complex
[Pt2Cl2(CO)2(PCy3)2] (4c) was synthesized following the method
reported by Mingos et al.24 Dimers with mixed halide contents
were formed by simple mixing of the corresponding dimers 4a, 4b
or 4c, indicating that intermolecular exchange of halide ligands
is a fast process in solution. When using excess of halogens the
final products were monomers cis-[PtX2(CO)(PCy3)] (X = I (6a),
Br (6b)), which are already known from literature.
92.1(3)
77.14(2)
94.31(8)
97.2(3)
171.3(3)
84.0(4)
P(2)–Pt(2)–Pt(1)
P(2)–Pt(2)–Pt(3)
P(3)–Pt(3)–Pt(1)
P(3)–Pt(3)–Pt(2)
I(1)–Pt(1)–Pt(2)
C(3)–Pt(3)–P(3)
Complex 2 crystallized from toluene as a solvate of compo-
sition [Pt3(l-CO)3(PCy3)3(I)(CH2CN)]·3.5C7H8 (Fig. 3, left). X-
Ray analysis showed that 2 is a Pt(II)Pt(0)Pt(0) cluster compound
in which one Pt atom inserted into the iodo–alkyl bond of iodo-
acetonitrile. The metal triangle remains intact and is almost
˚
isosceles with Pt–Pt distances of 2.73 and 2.69 A (selected bond
lengths and angles are listed in Table 1). These distances are
very similar to the metal–metal bond lengths observed in the
44–electron [3:3:4] cluster [Pt3(l-CO)3(PCy3)4].25
The phosphine ligands P2 and P3 remain almost coplanar
˚
with the metal triangle, with Pt–P distances of ∼2.28 A. The Pt1–
˚
P1 distance is 0.05 A longer than the other two in accordance
with the increased coordination number of the Pt1 center
and is very similar to the corresponding bond lengths in the
[3:3:4] cluster. For the same reason the Pt–C(carbonyl) distances
involving the Pt1 center are significantly elongated (2.29 and
2.33 A) compared to Pt2–C1 (1.93 A) and Pt3–C3 (1.98 A). The
mean value of the Pt–CO distances is similar to that found in
the starting cluster.25,26
X–Ray analysis of 7a (Fig. 3, right) shows that the Pt3 triangle
remains intact as well and is almost isosceles. The Pt–Pt distances
(Table 2) are comparable with the L2Pt–PtL metal–metal bond
lengths in the [3:3:4] analogue of the starting cluster25 or with
the L3Pt–PtL distances in 2. The increase in Pt–Pt bond lengths
from 1 to 7a is consistent with theoretical calculations on the
transformation of a 42–electron to a 44–electron Pt cluster.27,28
The phosphine ligands remain coplanar with the metal core and
the Pt–P bond lengths are similar to those found in the [3:3:4]
analogue of 1.25 Oxidative addition followed by substitution
of a CO ligand by I− has caused the remaining two carbonyl
ligands to flip out from the metal plane and occupy opposite
sides with respect to the two halide ligands. The Pt–P and
Pt–C(carbonyl) distances involving the Pt1 centre (having the
Adducts [Pt3(l-CO)3(PCy3)3(I)(CH2CN)] (2) and
[Pt3X(l-X)(l-CO)2(PCy3)3] (X = I ( 7a), Br (7b))
˚
˚
˚
The three adducts are 44-electron clusters. Adduct 2 is the first
example of a cluster where a PtL3 moiety is bonded to two
PtL units. In contrast to our previous observation where the ox-
idative addition of the low polarity reagent Ph3SnH converts the
same starting cluster to a cis adduct, the two added ligands of 2
occupy opposite sides of the Pt3 plane in an almost 180◦ arrange-
ment. For mononuclear transition metal complexes, it has al-
ready been shown that similar reactions give trans (electrophilic
reagents) and cis (low polarity reagents) products.4 The reactions
with elementary halogens are more complex. After the oxidative
addition substitution of one of the bridging CO by a l-X (X =
I or Br) unit occurs which forces the remaining CO bridges to
flip from the plane defined by the metal core to the other side
compared to the two halide ligands. Cluster 2 is fairly stable in
solution below 0 ◦C, although no conditions were found where
this complex was the sole product. In solution clusters 7a or 7b
are thermally stable with respect to fragmentation up to 90 ◦C.
Fig. 3 ORTEP view of the core of 2 (left) and 7a (right).
3 1 6
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5

View Article Online
◦
Table 4 Selected NMR data of 7a (7b) at 90 ◦C (d in ppm, J in Hz)
˚
Table 2 Selected bond lengths (A) and angles ( ) of 7a
Pt(1)–Pt(2)
Pt(1)–Pt(3)
Pt(2)–Pt(3)
Pt(1)–I(1)
Pt(2)–I(2)
Pt(3)–I(2)
Pt(1)–P(1)
2.7366(6)
2.7231(7)
2.7215(6)
2.7524(11)
2.7448(10)
2.6748(10)
2.328(3)
Pt(2)–P(2)
Pt(3)–P(3)
Pt(1)–C(1)
Pt(3)–C(1)
Pt(1)–C(2)
Pt(2)–C(2)
Pt(3)–Pt(1)–Pt(2)
2.258(3)
2.309(3)
2.174(12)
1.898(12)
2.034(13)
1.963(12)
59.796(17)
d(Pt)
d(P)
−4098 (−4029)
55.4 (50.5)
183
1J(Pt–Pt)
1J(Pt–P)
2J(Pt–P)
3J(P–P)
763
4805 (4814)
191 (202)
43.3 (43.0)
d(C)
1J(Pt–C)
725
2 using the gNMR 4.1 software package did not completely
converge, but a good fit was observed for the central peaks.
The spectrum presents two distinct phosphine environments
centered at 51.4 (P1) and 47.6 ppm (P2(3)). The one–bonded
Pt–P coupling constants involving distinct phosphines differ
Pt(2)–Pt(3)–Pt(1)
Pt(3)–Pt(2)–Pt(1)
Pt(2)–Pt(1)–I(1)
Pt(3)–Pt(1)–I(1)
Pt(1)–Pt(2)–I(2)
Pt(3)–Pt(2)–I(2)
P(1)–Pt(1)–I(1)
Pt(2)–I(2)–Pt(3)
P(1)–Pt(1)–Pt(3)
P(1)–Pt(1)–Pt(2)
60.349(17) P(2)–Pt(2)–Pt(3)
59.854(17) P(2)–Pt(2)–Pt(1)
151.31(9)
145.24(9)
108.92(9)
147.27(8)
148.08(8)
105.09(9)
61.14(2)
105.88(3)
86.4(5)
114.14(3)
77.83(3)
103.60(3)
58.59(2)
97.29(9)
60.27(6)
156.66(8)
139.93(8)
P(2)–Pt(2)–I(2)
P(3)–Pt(3)–Pt(1)
P(3)–Pt(3)–Pt(2)
P(3)–Pt(3)–I(2)
I(2)–Pt(3)–Pt(2)
I(2)–Pt(3)–Pt(1)
Pt(2)–C(2)–Pt(1)
Pt(3)–C(1)–Pt(1)
markedly (1J(Pt1–P1) 2589 Hz and J(Pt2(3)–P2(3)) 4474 Hz),
1
which indicates that the oxidative addition affected mostly
the Pt1 centre. All coupling constants involving only Pt2(3)
and P2(3) are similar to those found in 1, supporting again
similar stereochemistries in solution and in the solid state. Two-
and three-bonded coupling constant values are similar to the
corresponding ones in 1b. The 195Pt spectrum of 2 presents broad
signals and only chemical shift values are given in Table 3.
NMR measurements showed that clusters 7a and 7b are
fluxional in solution. The terminal and bridging halide ligands
and the two bridging COs are in fast exchange at the NMR scale.
A limiting temperature for this dynamic process could not been
reached down to −90 ◦C, only a significant line broadening was
observed indicating that the dynamic process slowed down.
Accordingly, 31P NMR spectra of 7a in CD2Cl2 presented only
one phosphine environment at 55.4 ppm (50.5 ppm in 7b) with
satellite peaks due to scalar coupling with three equivalent 195Pt
nuclei (Table 4), and with 1J(Pt–P) and 2J(Pt–P) values similar
to those found in the starting cluster.29
83.6(5)
highest coordination number) are longer than the corresponding
distances involving Pt2 and Pt3. The carbonyl and iodide ligands
bridge asymmetrically the Pt–Pt edges (e.g. 2.17 (Pt1–C2) and
˚
1.90 A (Pt3–C2)). These asymmetries may be related with the
fact that 7a is stereochemically non-rigid in solution (see below).
Structures of 2 and 7a (7b) in solution
The spectroscopic properties of 2 in solution (Table 3) are in
accordance with its geometry in the solid state. Its IR spectrum
presents two strong carbonyl stretching bands at 1872(b) and
1772(vs) cm−1 typical of bridging COs. In addition a weak
absorbance at 2242 cm−1 indicates the presence of a terminal
non-coordinated CN group.
The 13C NMR spectrum of 7a is in accordance with the
31P spectrum, as it showed a quartet central peak at 183 ppm
with satellite peaks from coupling with three equivalent 195Pt
nuclei. 195Pt NMR spectra showed a broad signal at −4090 ppm
(−4029 ppm for 7b). Raising the temperature resulted in
narrower signals, but also◦induced a slow decomposition of 7a
(b) to 4a (b) and 8. At 90 C, the 195Pt NMR spectra presented
a pseudo doublet of doublet of triplets in accordance with
the ligand movement about the metal core and a proposed
mechanism of the dynamic processes in solution is given in Fig. 4.
Samples of 2 enriched to 99% 13CO consist of six isotopomers
differing in the isotopic distribution of 195Pt and all NMR
spectra appear as superimpositions of subspectra. The 13C NMR
spectrum presents two chemically different CO environments
centered at 224 (C3(1)) and 257 ppm (C2) with relative intensities
1 : 2, respectively. The chemical shift values confirm the bridging
mode of the carbonyls, and the d(C2) value is similar to the
observed d(C) in 1 (259 ppm).11 The most intense features in the
spectrum have a triplet (C2) and a doublet of doublets (C3(1))
structure, due to 2J(C–C) and 2J(P–C) scalar couplings (Table 3).
The 1J(Pt–C) and 2J(Pt–C) couplings involving the two distinct
carbonyls differ slightly and are in accordance with the observed
asymmetry of these carbonyls in the solid state. Thus e.g. the
Pt(I) dimers [Pt2(CO)2(PCy3)2(I)(CH2CN)] (3) and
[Pt2X2(CO)2(PCy3)2] (X = I (4a), Br (4b), Cl (4c))
One and a half mole-equivalents of iodo-acetonitrile convert
1 to the asymmetric Pt(I) dimer [Pt2(CO)2(PCy3)2(I)(CH2CN)]
(3). This dimer slowly decomposes upon standing in solution
(even at −20 ◦C) giving back the starting cluster 1, as well as 4a
and [PtI(CO)(PCy3)(CH2CN)] (5). Due to this decomposition
value of 1J(Pt1–C1(3)) (187 Hz) corresponding to bond lengths
of about 2.3 A are significantly smaller than those of J(Pt2(3)–
1
˚
˚
C1(3)) which correspond to distances of about 2.0 A. Due to
inherent complexity, simulation of the 31P NMR spectrum of
Table 3 Selected NMR data of 2 from gNMR simulation (d in ppm, J in Hz)
d(C2)
d(C3(1))
d(P1)
d(P2(3))
d(Pt1)
d(Pt2(3))
257.2
224.4
51.4
1J(Pt1–C1(3))
1J(Pt2(3)–C1(3))
1J(Pt2(3)–C2)
2J(Pt2(3)–C1(3))
2J(Pt1–C2)
187
1093
832.5
26
38
30.1
∼3–4
1J(Pt1–P1)
2589.8
4474.2
385
412
229
81
34.9
1J(Pt2(3)–P2(3))
2J(Pt2(3)–P1)
2J(Pt2(3)–P3(2))
2J(Pt1–P2(3))
3J(P1–P2(3))
3J(P2–P3)
47.6
−4410
−4560
2J(C2–C1(3))
2J(C–P)
Fig. 4 Proposed dynamic process of 7a in solution.
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 1 7

View Article Online
Table 5 NMR spectral data for 3 (d in ppm, J in Hz)
d(C1)
d(C2)
d(P2)
d(P1)
d(Pt1)
d(Pt2)
1J(Pt1–Pt2)
174.9
178.8
50.0
1J(Pt1–C1)
1J(Pt2–C2)
1J(Pt1–P1)
1J(Pt2–P2)
2J(Pt2–P1)
2J(Pt1–P2)
3J(P1–P2)
2009
1260
2675
2300
561
115
144
55.3
−4244
−4010
166.5
all attempts to isolate 3 failed, hence only a structure in solution
can be proposed (Fig. 5), which is based on the NMR data
shown in Table 5 and on IR spectroscopic properties.
Fig. 7 Ball and stick representation of disordered dimer 4a.
Dimer 3 decomposed upon standing in solution giving
4a besides succinonitrile and monomer 5. The presence of
succinonitrile was confirmed by IR spectroscopy (m(CN) at
2254 cm−1) and by a GC-MS measurement of a MeOH extract
of the reaction residue.
Fig. 5 Proposed structure for dimer 3.
A satisfactory computer simulation of the 13C NMR spectrum
of 3 has been implemented on the basis of the superimposition
of subspectra belonging to the four isotopomers differing in 195Pt
content. The 13C NMR spectrum showed two chemically non-
equivalent, terminal COs at 179 (C1) and 175 ppm (C2) with
equal intensities, in accordance with the IR spectrum where
two strong carbonyl absorbances were present at 2007 and
2035 cm−1. In addition the latter spectrum presented a weak
peak at 2205 cm−1 which could be attributed to a terminal,
non-coordinated nitrile group. Computer simulation of the 31P
spectrum (Fig. 6, left) of 3 could be implemented with the
support of the 195Pt data. The 195Pt NMR spectrum (Fig. 7, right)
shows two well separated peaks appearing at −4244 ppm (similar
to the d observed in 4a, see Table 6) and at −4010 ppm with
expected 1J(Pt–P), 2J(Pt–P) and 1J(Pt–C) couplings. Compari-
son of corresponding 1J(Pt–C) and 1J(Pt–P) values of dimers 3
and 4a, which share a common [Pt(CO)I(PCy3)] unit, allowed
the assignment of each Pt fragment of 3, while the similarity
of 1J(Pt–C) values in the [Pt(CO)(PCy3)(CH2CN)] fragment
of 3 and in 5 (1260 and 1120 Hz, respectively) suggested a
similar trans arrangement of the CO and CH2CN ligands about
Pt2. The positive 2J(Pt1–P2) value of 564 Hz strongly suggests
direct metal–metal interaction and that the two phosphine
ligands occupy equivalent sites in a possibly linear P–Pt–Pt–
P arrangement.30 This suggestion was confirmed by a DFT
calculation of coupling constants (see below) in model dimers. In
contrast, 2J(Pt2–P1) is very small (115 Hz) as well as 1J(Pt1–Pt2)
(166.5 Hz), but it is known that 1J(Pt–Pt) values are not always
indicative of direct interaction between two platinum centers.15,31
Platinum(I) dimers similar to 4a have already been
reported.13,14,24,32,33 All these complexes contain terminally
bonded halides. In these species the two Pt centres (having both
square-planar geometry) associate almost perpendicularly with
a direct metal–metal bond, as shown by the crystal structure of
[Pt2Cl2(CO)2(PtBu2Ph)2]14 and [Pt2Cl2(CO)2(PFc2Ph)2].24 Treat-
ment of 1 with 1.5 equivalents of elemental halogens led to the
formation of the Pt(I) dimers [Pt2X2(CO)2(PCy3)2] (X = I 4a, Br
4b). The analogous complex containing chloride as ligand 4c was
synthesized by reduction of [Pt(COD)Cl2] following the method
developed by Mingos and co-workers.24 Simple mixing of the
corresponding dimers 4a, 4b or 4c resulted in the immediate
formation of the analogous complexes with mixed halide ligands.
Dimer 4a crystallized from a toluene solution as a solvate,
[Pt2I2(CO)2(PCy3)2]·3.5C7H8. X–Ray analysis (an ORTEP rep-
resentation is shown in Fig. 7) showed that 4a is isostructural
with the complexes already cited.14,24 The dimer consists of two
[PtI(CO)(PCy3)] units (both of them close to planarity) linked
via a direct metal–metal bond. The planes defined by the ligands
about the two Pt centres are approximately perpendicular with
a dihedral angle of 76.5◦ which is larger than the observed value
of 66.5◦ in the related complex with a polyaromatic phosphine,24
and the value of 70.1◦ observed in the dimer with PtBu2Ph.14 This
arrangement of the two planes renders the complex chiral. The
iodide and the CO group around the Pt2 centre are disordered,
in the sense that the iodide ligand occupies the position of the
CO ligand in 20% of the molecules which are present in the
˚
crystal. The Pt1–Pt2 distance of 2.64 A (see Table 7) is within
the usual range (2.56–2.65) found for platinum(I) dimers,34
Fig. 6 Observed and simulated 31P (left) and 195Pt (right) NMR spectra of 3 in d8-toluene at 283 K.
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 1 8

View Article Online
Table 6 Selected NMR data of 4a and its analogues
X
Y
1J(Ptx–Pty)
1J(Ptx–Px)
1J(Pty–Py)
2J(Pty–Px)
2J(Ptx–Py)
3J(Px–Py)
4a
I
I
I
Br
Br
Cl
I
0
54
110
217
304
535
30
2617
2628
2641
2415
2419
2268
2617
2434
2318
2415
2280
2268
198
172
177
250
238
344
198
236
291
250
317
344
154
66
79
168
97
181
Br
Cl
Br
Cl
Cl
4b
4c
◦
˚
Table 7 Selected bond lengths (A) and angles ( ) of 4a
Pt(1)–Pt(2)
Pt(1)–P(1)
Pt(1)–I(1)
Pt(1)–C(1)
Pt(2)–P(2)
Pt(2)–I(2A)
Pt(2)–I(2B)
Pt(2)–C(2A)
Pt(2)–C(2B)
2.6374(5)
2.3535(16)
2.6353(7)
1.905(10)
2.3440(17)
2.6246(11)
2.604(6)
P(1)–Pt(1)–Pt(2)
P(2)–Pt(2)–Pt(1)
I(1)–Pt(1)–Pt(2)
I(2A)–Pt(2)–Pt(1)
I(2B)–Pt(2)–Pt(1)
C(1)–Pt(1)–Pt(2)
C(2A)–Pt(2)–Pt(1)
C(2B)–Pt(2)–Pt(1)
C(1)–Pt(1)–P(1)
C(1)–Pt(1)–I(1)
P(1)–Pt(1)–I(1)
177.83(4)
170.22(5)
83.242(18)
84.27(2)
78.7(2)
83.1(2)
81.4(4)
85.7(8)
96.1(2)
166.3(2)
97.60(4)
C(2A)–Pt(2)–P(2)
98.1(4)
93.6(8)
3.4(4)
164.0(8)
164.7(4)
11.2(10)
96.92(5)
101.1(2)
165.9(9)
161.4(2)
C(2B)–Pt(2)–P(2)
C(2A)–Pt(2)–I(2B)
C(2B)–Pt(2)–I(2B)
C(2A)–Pt(2)–I(2A)
C(2B)–Pt(2)–I(2A)
P(2)–Pt(2)–I(2A)
P(2)–Pt(2)–I(2B)
C(2A)–Pt(2)–C(2B)
I(2B)–Pt(2)–I(2A)
1.845(8)
1.856(8)
which is shorter than that observed in Pt(0) cluster complexes.
˚
The Pt–C(carbonyl) distances (1.90 and 1.85 A) are very similar
to those observed in isostructural complexes.
The IR spectrum of 4a shows two strong CO bands at 2010
and 2030 cm−1, indicative of terminally bonded CO groups.
Analogous complexes showed similar absorption frequencies
(around 2012 and 2035 cm−1). An absorption band due to
the Pt–X bond was observed at 172 and 216 cm−1 for 4a and
4b, respectively, which are typical of terminal Pt–X stretching
frequencies.35,36
The 31P NMR spectra show a single resonance at 57.5
(4a), 56.1 (4b) and 55.6 ppm (4c), respectively, with satellites
characteristic for two chemically but not magnetically equivalent
phosphine environments. For complexes with mixed halide
content the phosphine peaks are somewhat lower fielded. As ex-
pected, the spectra are the superimposition of subspectra related
to the three isotopomers differing in 195Pt content. Satisfactory
computer simulations of these spectra have been implemented
using gNMR 4.1 and have ascertained the assignments of all
coupling constants (Table 6).
Fig. 8 Observed (top) and simulated (bottom) 195Pt NMR spectra of
4a in CD2Cl2.
The13C NMR spectrum of 4a confirmed the presence of
terminally bonded CO ligands with a chemical shift value of
167.7 ppm, which is typical of this bonding mode. The 195Pt
spectra of 4a (Fig. 8), 4b or 4c showed doublet of doublets
as the most intense features centred at −4243.2, −3998.7 and
−3913.5 ppm, respectively.
These small 1J(Pt–Pt) coupling constants may be rationalized
on the basis of theoretical calculations due to Ziegler et al.15
Thus, the smaller value of J(Pt–Pt) observed when replacing
1
a polyaromatic phosphine (748 Hz) by PCy3 (535 Hz) is in
accordance with the increased r bonding ability of PCy3. The
decrease of this coupling along the sequence Cl > Br > I
is in accordance with the relative r bonding ability of these
ligands and is in accordance with the trend observed in e.g.
[Pt2X4(CO)2]2−.37
Three dynamic processes take place in solution. First, a rapid
CO exchange was observed when a solution of 4a was stirred
under 13CO atmosphere. The lability of the CO ligands suggests
that exchange may occur intermolecularly as well, but there is
no direct evidence. Secondly, the rapid formation of the dimer
complexes with mixed halide content by simple mixing of the
corresponding dimers 4a, 4b or 4c suggests that fast substitution
of halide takes place in solution. This process probably occurs
via the formation of a solvato-complex which liberates a small
The 1J(Pt–P) and 1J(Pt–C) values of 2617 and 1945 Hz,
respectively, are similar to corresponding values in isostructural
complexes.14,24 Complexes with other halides show the expected
decrease in values of 1J(Pt–P) in accordance with the increased
electronegativity of the halide and the decreased s-electron den-
sity of the Pt–P bond. The observed 2J(Pt–P) value (198 Hz) of 4a
is significantly smaller than that of the above referred complexes,
or in 4b (250 Hz) and in 4c (344 Hz), but is close to the value
of 219 Hz found in the Pt(I) dimer [Pt2I2(CO)2(PtBu2Ph)2].13
Furthermore 1J(Pt–Pt) coupling constants are small (0
30 Hz in 4a, 217 Hz in 4b and 535 Hz in 4c) compared to
the estimated value of 748 Hz of the Mingos chloro-dimer
containing polyaromatic phosphines.24
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 1 9

View Article Online
Fig. 9 Equilibrium of enantiomers.
quantity of halide. The third process is that already proposed by
Mingos and co-workers24 which occurs between enantiomers as
presented in Fig. 9.
The suggested mechanism involves CO bridges in the tran-
sition state, based on a low temperature NMR study and MO
calculation, which showed that the rotation about the Pt–Pt
bond is not favoured since it is has a high energy barrier
of about 130 kJ mol−1 (calculated for [Pt2Cl2(CO)2(PH3)2]).
Our low temperature 31P and 13C NMR measurements of
complexes 4a, b and c showed that all spectra in dichloromethane
solution with or without the presence of a chiral environment
(1(+)R-camphor) were identical at each temperature. Changes
in coupling constants showed the same tendency as that found
by Mingos et al.24 Coalescence points of the two lines belonging
to diastereoisomers in the 31P NMR spectra are estimated to
be −72 and −85 ◦C for 4a and 4b, respectively while in the
case of 4c, a limiting temperature could not be observed down
to −105 ◦C. At these temperatures intermolecular substitution
of halides still occurred. A mechanism of interconversion which
also takes into account the observed intermolecular substitution
of the halide ligands could start with the formation of a halide
bridge and involve a nucleophilic attack of halide (Fig. 10).
Fig. 11 Observed (top) and simulated (bottom) 119Sn spectra of
complex 9a.
Table 8 Selected coupling constants of complexes 8a and 8b
9a
9b
dP
38.5
43.7
dPt
dSn
−4655.7
50.5
415
2163
104
45
2050
6458
−4272.5
241.8
385
1950
128
54
2041
7555
2J(Pt–Pt)
1J(Pt–P)
3J(Pt–P)
4J(P–P)
2J(Sn–P)
1J(Pt–Sn)
smaller). The observed nJ(Pt–P) values (415 in 9a and 385 Hz in
9b) were larger than those of 3 and 4a (and somewhat smaller
than that of 4c).
Fig. 10 Possible reaction mechanism for the interconversion of the
enantiomers of 4a presented with Newmann projections.
All this information points towards the probable formation of
[Pt2(l-SnCl2)(CO)2(PCy3)2X2] (X = I for 9a or Cl for 9b). Pt2(l-
Sn) groups have been prepared earlier by insertion of SnCl2 into
the Pt(I)–Pt(I) bond of [Pt2X4(CO)2]2−,38 by addition of Sn(acac)2
to [(PPh3)2Pt(p-C2H4)],39 or more recently by oxidative addition
of tin(IV) halides to Pt(II) complexes.40 It is also known that in the
case of [Pt2Cl4(PEt3)2] insertion of the Sn(II) unit proceeds into a
Pt–X bond41 rather than into the Pt–Pt bond. No single crystals
of good quality were obtained because of slow decomposition
of these complexes in solution. However X-ray analysis could
ascertain the atomic connectivity of 9a (Fig. 12).
Dimers [Pt2(l-SnCl2)(CO)2(PCy3)2X2] (X = I 9a or Cl 9b)
In order to get additional information about the low tem-
perature behaviour of complexes similar to 4 but containing
unambiguously terminal ligands, complexes 4a and 4c were
reacted with SnCl2·2H2O, as we expected the formation of
[Pt2(SnCl2X)2(CO)2(PCy3)2].
Surprisingly, the 119Sn NMR spectra (Fig. 11) of the products
2
show bridging Sn units with two large J(119Sn–31P) (2050 and
2041 Hz in 9a and 9b, indicating trans Sn–P arrangement) and
DFT study of coupling constants in Pt(I) dimers
1
two relatively small J(119Sn–195Pt) coupling constants (6459 in
9a and 7555 Hz in 9b).
Dimers 3 and 4a and its analogues show very small one-bonded
Pt–Pt couplings reflecting small s-electron density between
metallic centres and indicate that in the case of platinum there
is no obvious correlation between bond lengths and nuclear
spin–spin coupling constants. In the literature there is only one
example (supported by a crystal structure) where the observed
scalar coupling in a direct Pt–Pt bond is smaller than 200 Hz, i.e.
triangulo [Pt3(l-CNtBu)3(CNtBu)3].42 Recently Ziegler and co-
workers published a theoretical investigation of the irregular
behaviour of Pt–Pt coupling constants.15 They studied three
Pt(I) dimers, [Pt2(CO)6]2+ (1J(Pt–Pt) = 551 Hz experimentally
These spectral information exclude the formation of com-
plexes with terminal SnCl2X (X = I or Cl) groups. Conduc-
timetric measurement in CH3NO2 showed that the complexes
are neutral, which excludes substitution of halides by l-SnCl2.
The IR spectrum indicated the presence of terminal COs, with
stretching bands at 2050 and 2077 cm−1.
The observed 31P and 195Pt NMR spectra of 9a and 9b and the
spectroscopic data (Table 8) were very similar to those of starting
dimers 4a and 4c, indicating similar Pt(I) environments (only the
2
observed J(Pt–P) value of 127 Hz (107 for 9b) was somewhat
3 2 0
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5

View Article Online
constants in complexes with axial COs or with axial phosphine
ligands are comparable. This would be in accordance with the
supposed small effect of p back donation of COs, which in
itself is greater than that of phosphines, while r interactions
of these two types of ligands are similar. In order to verify these
suggestions we investigated nuclear scalar couplings in model
molecules which are analogues of 3, 4a and 4c containing PMe3
to reduce the cost of the calculation.
Three model complexes were used for the calculation of
nuclear coupling constants: [Pt2I(CH2CN)(CO)2(PMe3)2] (3m),
[Pt2I2(CO)2(PMe3)2] (4am) and [Pt2Cl2(CO)2(PMe3)2] (4cm). The
optimized geometry of 4am was in good agreement with the solid
state structure of 4a, and the calculated geometry of 3m has a
smaller total bonding energy than that of a possible isomer
of 3 having one phosphine ligand trans to the CH2CN group
(Fig. 13).
Fig. 12 Ball and stick representation of 9a (R = 0.120).
The calculated 1J(Pt–Pt) and 2J(Pt–P) values of 3m, 4am
and 4cm, using all electron TZP basis sets were overestimated
observed and 874 Hz calculated), [Pt2Cl4(CO)2]2− (5250 Hz
observed and 6397 Hz calculated, COs in equatorial positions,
linear Cl–Pt–Pt–Cl arrangement) and its hypothetical isomer
where the COs occupy axial positions (−963 Hz calculated).
They found that r interactions, particularly in the axial posi-
tions, are responsible for the reduced coupling constant in the
hexacarbonyl dimer compared to the other dimer, and that p
back donation of the COs has little effect. They showed that
the relativistic effect has a great influence on the r bonding
capacity of Pt. The expected large magnitude of the Fermi
contact contribution to coupling is to a large extent compensated
by increasing r interactions with axial CO ligands. Using these
findings and comparing experimentally observed values of the
Pt–Pt coupling constants in complexes 3, 4a and some analogues,
one would expect that the reduction of metal–metal coupling
1
while J(Pt–Pt) was underestimated (Table 9) compared to the
corresponding experimental values, although comparing them
to models with PH3 as ligands they were largely improved. Only
the calculated Pt–C coupling constants were in good agreement
with the experimentally observed values. Unfortunately further
increment of the size of the model molecules (using larger
phosphines) was not possible with an acceptable increase of the
cost of the calculations. Solvent effect has already been shown43
to have a major role in the magnitude of metal–metal nuclear
coupling constants in heterometallic Pt–Tl dimers, therefore
computation of scalar couplings were repeated on 3m and 4am,
applying the conductor like screening model (COSMO)44–46 of
solvation. Using this solvation model improved significantly
the quality of the calculated results (Table 9) compared to the
Fig. 13 Optimized geometries of model dimers 3m, 4am and 4cm.
Table 9 Experimentally observed (for 4a, 3 and 4c) and calculated (full electron, COSMO) coupling constants for 4am, 3m and 4cm
Exp 4a
Exp 3
Exp 4c
Calc. 4am
Calc. 3m
Calc. 4cm
Calc. 4am COSMO
Calc. 3m COSMO
1J(Pt1–Pt2)
1J(Pt1–P1)
2J(Pt1–P2)
1J(Pt1–C1)
1J(Pt2–P2)
2J(Pt2–P1)
1J(Pt2–C2)
0
2617
198
1945
2617
198
167
2675
115
2009
2300
561
535
2268
344
—
2268
344
—
1094
1632
427
2065
1632
427
1355
1530
352
2115
1468
837
1695
1323
650
—
863
1709
222
2198
1709
222
1001
1633
217
2233
1513
585
1945
1260
2065
1396
2198
1475
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 2 1

View Article Online
Table 10 Differences in observed and calculated coupling constants (Hz)
Dexp (J3 − J4a)
Dcalc (J3m − J4am
)
Dcalc (J3m − J4am) COSMO
Dexp (J4c − J4a)
Dcalc (J4cm − J4am
)
1J(Pt1–Pt2)
1J(Pt1–P1)
2J(Pt1–P2)
1J(Pt1–C1)
1J(Pt2–P2)
2J(Pt2–P1)
1J(Pt2–C2)
167
58
261
−102
−75
50
138
−76
−4
535
−349
146
601
−309
223
−83
34
36
−317
363
−715
−163
−196
−349
−309
410
363
146
223
−669
−723
previous calculations. Calculated 1J(Pt–Pt) and 2J(Pt–P) values
1
1
decreased while J(Pt–P) and J(Pt–C) values increased. The
numerical values of 1J(Pt–Pt) and 1J(Pt–P) are still not in good
agreement with the experimental ones but 1J(Pt–C) and 2J(Pt–P)
values are.
Comparison of differences between observed and calculated
coupling constants (Table 10) show that both calculations (using
COSMO or not) were able to reproduce trends in these values.
Thus, for example, the calculations (not applying COSMO)
1
predict the J(Pt–Pt) value to be smallest in 4am and largest
in 4cm, which is indeed the case in the experimentally observed
metal–metal coupling constants. The calculated large differences
in the values of two different pairs 2J(Pt–P) and 1J(Pt–C) in 3m
fit well with the corresponding observed differences between
these couplings in complex 3. This confirms that the structure
supposed for 3 based on NMR data is most probably right and
indicates that the cyano-alkyl group is responsible for significant
changes in these values.
In conclusion these calculations clearly showed that PH3
fails to model PCy3 in dimeric Pt complexes. The calculation
confirmed the suggested structure of complex 3. Trends in corre-
sponding coupling constants were nicely reproduced with model
systems containing PMe3 as ligands, although the numerical
values differed significantly from experimentally observed ones.
These differences were probably due to the improper electronic
description of these systems. Autsbach et al.15 have already
shown that standard ADF basis sets are not flexible enough
in the core region in the case of heavy elements such as platinum
for calculating coupling constants. They suggested that one has
to add steep basis functions with high exponents to reach the
required flexibility. In the absence of such basis sets we had to
settle for conventional ADF sets.
Fig. 14 ORTEP view of monomer 5.
◦
˚
Table 11 Important bond lengths (A) and angles ( ) in 5
Pt1–P1
Pt1–C2
I1–Pt1–P1
C1–Pt1–I1
2.298(3)
2.181(12)
Pt1–I1
Pt1–C1
2.648(12)
1.919(16)
85.9(4)
178.24(9)
84.6(4)
C2–Pt1–I1
C1–Pt1–C2
169.9(5)
doublet of doublet centered at −4738 ppm, a value in the typical
chemical shift range for Pt(II) complexes.
Elemental halogens converted
1
to monomers cis-
[PtX2(CO)(PCy3)] (X = I, 6a and Br, 6b). These types of com-
plexes with different tertiary phosphines are well known from
the literature.18,19,23,48–52 X-ray analyses of cis–[PtBr2(CO)PCy3]
(6b) and [PtICl(CO)PCy3] (6c) has been undertaken (see
supplementary crystallographic data). The X-ray structure of
the known16–23 Pt(II) dimer [Pt2I2(l–I)2(PCy3)2] 8 (an ORTEP
representation is shown in Fig. 15) which is a fragmentation
product of 7a has been determined as well.
Monomers [PtI(CO)(PCy3)(CH2CN)] (5), cis-[PtI2(CO)(PCy3)]
(6a) and its analogues
Excess of the electrophilic reagents converted the starting
cluster 1 to monomers as shown in Fig. 1 and 2. In the
case of iodo-acetonitrile the new 18 electron Pt(II) complex
[PtI(CO)(PCy3)(CH2CN)] (5) was observed. This complex is
barely soluble in toluene, thus the pure product could be
separated from the solution as an ivory solid. X-Ray analysis
(an ORTEP representation is shown in Fig. 14) showed that 5
is a square planar platinum complex which has a trans CO–
CH2CN arrangement (selected bond lengths and angles are
listed in Table 11). The Pt–P, Pt–C(carbonyl) and Pt–I distances
are within the usual ranges. Due to the larger sterical demand
of the cyclohexyl group compared to phenyl, the Pt–C2 value of
˚
˚
2.181 A is longer than the corresponding bond lengths (2.121 A)
observed in cis–[Pt(CH2CN)2(PPh3)2].47
NMR and IR spectroscopic measurements showed that stere-
ochemistries in solution and in the solid state are similar. The
31P and 13C NMR spectra show one phosphine (d = 28.7 ppm)
and one carbonyl (173.5 ppm) environment, with the expected
satellites due to coupling with one 195Pt nuclei (3179 and 1120 Hz,
respectively). The magnitude of the 2J(C–P) coupling constant
(5.9 Hz) confirmed the cis arrangement of CO and PCy3 ligands.
The 195Pt spectrum of the 100% 13C labelled complex 5 showed a
Fig. 15 ORTEP (50% probability level) representation of 8.
3 2 2
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5

View Article Online
Syntheses
Conclusions
[Pt3I(CH2CN)(l-CO)3(PCy3)3] (2). A solution of [Pt3(l-
CO)3(Cy3)3] (0.29 g, 0.19 mmol) in toluene (25 ml) was cooled to
−70 ◦C in a N2(l)/acetone bath and ICH2CN (14 ll, 0.19 mmol)
was added with a Hamil◦ton syringe. The mixtur◦e was stirred for
30 min, 10 min at −50 C and 10 min at −20 C. The solvent
was evaporated to dryness at this temperature over a period
of 3 h. The product is a mixture of 2 and 3. Single crystals
were obtained from a toluene solution of this mixture by slow
diffusion of ethanol. IR and NMR data: Table 3. X-Ray data:
see Table 1, Table 12 and supplementary crystallographic data.
The formation of a Pt–alkyl bond by oxidative addition to a plat-
inum cluster compound is not documented. We found that the
oxidative addition of iodo-acetonitrile to [Pt3(l-CO)3(PCy3)3] in
equimolar ratio gives the adduct [Pt3(l-CO)3(PCy3)3I(CH2CN)]
(2) where one Pt atom is inserted into the I–CH2CN bond. The
reaction, however, is more complicated when excess of iodo-
acetonitrile is added to the starting cluster, as fragmentation to
dimers such as [Pt2(CO)2(PCy3)2I(CH2CN)] (3) and monomers
takes place. Since one monomer cis-[PtI2(CO)(PCy3)] and succi-
nonitrile are formed, the fragmentation must take place partly
by homolytic scission of the Pt–CH2CN bond.
[Pt2I(CO)2(PCy3)2(CH2CN)] (3).
A solution of [Pt3(l-
Oxidative addition of iodine to [Pt3(l-CO)3(PCy3)3] resembles
that of iodo-acetonitrile in the sense that an adduct is formed
when using 1 : 1 mole ratio of reactant, i.e. [Pt3(l-CO)2(l-
I)I (PCy3)3] where one platinum atom inserted into the I–
I bond as it is does into the I–CH2CN bond giving [Pt3(l-
CO)3I(CH2CN)(PCy3)3]. In the former case, insertion was fol-
lowed by CO substitution by I−, as iodide can have two binding
modes (terminal and bridging), whereas in the latter case the -
CH2CN group can only be terminal and no CO substitution can
occur. A complicated fragmentation is observed in the presence
of excess of elemental iodine (or bromine) giving monomers
as final products and dimers such as [Pt2(CO)2I2(PCy3)2] as
intermediates. The crystal structure of the latter indicated that
the dimer is chiral with all ligands being terminally bonded. The
existence of enantiomers in solution could not be established
with certainty when trying to replace the iodide ligand with an
NMR active atom as structural indicator i.e. when reacting the
dimer with SnCl2 (to give Pt–SnCl2I), it was found that SnCl2
inserted into the metal–metal bond giving the new compound
[Pt2(l-SnCl2)(CO)2(PCy3)2I2] which evidently is not chiral.
CO)3(Cy3)3] (0.29 g, 0.19 mmol) in toluene (25 ml) was cooled to
−50 ◦C in N2(l)/acetone bath and ICH2CN (21 ll, 0.29 mmol)
was added with a Hamilton syringe. The mixture was stirred for
20 min, for another 10 min at −20 and 0 ◦C and finally for 1 h at
room temperature. Concentration of the resulting solution gave
a red–brown solid. (Yield: 0.245 g, 72%). An analytically pure
sample could not be obtained. Selected IR data (cm−1): toluene
solution m(C≡O) 2007; 2035vs; m(C≡N) 2205 m. NMR data: see
Table 5.
[PtI(CO)(PCy3)(CH2CN)] (5).
A
solution of [Pt3(l-
CO)3(Cy3)3] (0.29 g, 0.19 mmol) in toluene (25 ml) was cooled to
−20 ◦C in N2(l)/acetone bath and ICH2CN (45 ll, 0.62 mmol)
was added. The mixture was stirred for 10 min in the cold,
for another 10 min at 0 ◦C, and finally for a night at room
temperature. Filtration of the concentrated toluene solution
resulted in a whitish-yellow solid, which could be recrystallized
from dichloromethane/heptane mixture. (Yield: 0.220 g, 57.2%).
Selected IR data (cm−1): in THF solution m(C≡O) 2080vs;
m(C≡N) 2214w. Anal. Calc. for C21H35INOPPt: C, 37.62; H,
5.26; N, 2.09. Found: C, 37.57; H, 5.31; N, 2.05. NMR data: see
text. Crystal data of 5: see supplementary crystallographic data
and Table 12.
Experimental
All manipulations were carried out under a nitrogen atmosphere
using standard Schlenk techniques. All solvents were dried and
distilled prior to use. Infrared spectra (CaF2 cells) were recorded
on a Perkin–Elmer 983 spectrometer, and NMR spectra on
Bruker AC 200 (1H at 200.13 MHz, 31P at 81.02 MHz and 13C
at 50.32 MHz), Bruker WH 360 (13C at 90.55 MHz) and
Bruker DRX 400 (31P at 161.93 MHz, 13C at 100.6 MHz and
195Pt at 84.5 MHz) spectrometers. The chemical shifts (d) are
referred to external TMS (1H and 13C) external 85% H3PO4
Evidence for the formation of succinonitrile
A THF solution of 3 was magnetically stirred for one day at
60 ◦C under N2. The solvent was evaporated at reduced pressure.
A MeOH extract of the product was analyzed by GC-MS
◦
with a source temperature of 150 C. Characteristic m/z were
found at 81 (34.4%), 53 (100%), which were consistent with the
spectral features presented in the GC-MS spectrum of a MeOH
solution of succinonitrile recorded under the same conditions
(m/z: 81 (20.9), 53 (100)). A third measurement was made on
a MeOH solution of ICH2CN to exclude the direct formation
of succinonitrile under the measurement conditions. The m/z
lines found in the latter spectrum are: 167 (81.38), 144 (100), 127
(16.48). The IR spectrum shows a CN band at 2254 cm−1.
1
(31P), and Na2PtCl6 (195Pt). The spectra of all nuclei were H
decoupled. Simulations were performed with the gNMR 4.1
program, considering only the relative populations of platinum
1
isotopomers since all samples for 13C{ H} NMR were enriched
to ≥99% 13CO. [Pt3(CO)3(PCy3)3] was prepared by the literature
method.53 The 13C isotopically labelled derivative was prepared
similarly, using 13CO enriched to 99%. Iodo-acetonitrile, I2, Br2,
NaBEt3H, SnCl2·2H2O (Aldrich) and carbon monoxide (>99%
13CO) were used as received (Aldrich).
[Pt2I2(CO)2(PCy3)2] (4a). A THF solution (25 ml) of [Pt3(l-
CO)3(PCy3)3] (0.20 g, 0.124 mmol) was cooled to −60 ◦C in
a N2(l)/EtOH bath and I2 (dissolved in 5 ml toluene, 51 mg,
0.2 mmol) was added with a syringe. The mixture was stirred
for 10 min at low temperature, for another 10 min at 0 ◦C and
finally for a short time at room temperature. Partial evaporation
of solvent and addition of MeOH to the solution gave a
yellow–brown solid. (Yield: 0.130g, 55%). Selected IR data
(cm−1): THF solution m(C≡O) 2031vs; 2010vs. Anal. Calc. for
C38H66I2O2P2Pt2: C, 36.20; H, 5.28. Found: C, 35.70; H, 5.46.
Yellow single crystals were grown from toluene at 5 ◦C in a
10 mm NMR tube over 2 months. NMR data: see Table 6. X-Ray
data: see Table 7, Table 12 and supplementary crystallographic
data.
The relevant details of the crystal, data collection, and struc-
tural refinement are listed in the supplementary crystallographic
data. Diffraction data were collected using Mo Ka radiation.
The structure was refined using the full-matrix least-squares
on F2. All structures have been refined anisotropically, except
2 and 7a for which light atoms (C, N and O) have unstable
behaviour and then retained isotropic. The same problem
occurred in the refinement of compound 8a for which also
P and Cl atoms have been retained isotropic. Isotropic is
also the refinement of a disordered solvent molecule in the
crystal structure of compound 4a. All hydrogens were made
to ride isotropically on their associated carbons. Structure
refinement and geometrical calculations were carried out on all
structures with the SHELXL-97.54 Graphical representations of
the molecular structures in the crystal were generated with the
program ORTEP.55
[Pt2Br2(CO)2(PCy3)2] (4b). As for 4a starting from 1 (0.2 g,
0.124 mmol) and treated with Br2 (10 ll, 0.129 mmol). Yellow–
brown solid, 0.130 g (60%). IR (cm−1): THF solution m(C≡O)
2032; 2012vs. Anal. Calc. for Br2C38H66O2P2Pt2: C, 39.11; H,
5.70. Found: C, 39.28; H, 5.60. NMR data: see Table 6.
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 2 3

View Article Online
[Pt2Cl2(CO)2(PCy3)2] (4c). This complex was prepared
following the method reported by Mingos et al.24 for
[Pt2Cl2(CO)2(PNpPh2)2] starting with [PtCl2(COD)] (0.3 g,
0.8 mmol). Yield 0.218 g (51%) as brown microcrystals. An
analytically pure sample could not been obtained as it contained
small amounts of 1 and 6. NMR data: see Table 6.
[PtI2(CO)PCy3] (6a), [PtBr2(CO)PCy3] (6b) and [PtCl(I)-
(CO)PCy3] (6c). Complex 6a was prepared following the
literature method.19 6b and 6c were obtained as minor products
in the preparation of 4b and 9a, respectively. X-Ray data for the
latter two complexes: see supplementary crystallographic data
and Table 12.
Pt3I(l-I)(l-CO)2(PCy3)3] ( 7a). A toluene solution (25 ml) of
[Pt3(l-CO)3(PCy3)3] (0.10 g, 0.062 mmol) was cooled to −60 ◦C
in a N2(l)/EtOH bath and I2 (dissolved in 5 ml toluene, 17 mg,
0.066 mmol) was added with a syringe. The mixture was stirred
for 10 min at low temperature, for another 10 min at 0 ◦C and
finally for a short time at room temperature. A mixture of 7a and
4a was formed. The mixture was separated on a 20 × 20 mm TLC
plate (Kieselgel 60 F254). Elution with dichloromethane/hexane
1 : 2 gave a dark red band following a yellow band of 4a (yield:
20 mg, 17.4%). Anal. Calc. for C56H99I2O2P3Pt3: C, 38.74; H,
5.75. Found: C, 39.18; H, 5.90. Selected IR data (cm−1): THF
solution m(C≡O) 2066vs and 2042sh. NMR data: see Table 4.
Dark red single crystals were obtained from THF solution of
the mixture of 7a and 4a at −20 ◦C with slow diffusion of
heptane. X-Ray data: see Table 2, Table 12 and supplementary
crystallographic data.
[Pt3Br(l-Br)(l-CO)2(PCy3)3] ( 7b). As for 7a using Br2
instead of I2. No attempts to separate the mixture of 7b and
4b were made. NMR data: see Table 4.
[Pt2I2(l-I)2(PCy3)2] (8). This complex was prepared follow-
ing the method reported by Al-Najjar.19 Yellow single crystals
were obtained from a CH2Cl2 solution standing at −20 ◦C
with the formula [Pt2I2(l-I)2(PCy3)2]·2CH2Cl2. X-Ray data: see
supplementary crystallographic data and Table 12.
[Pt2I2(SnCl2)(CO)2(PCy3)2] (9a). SnCl2·2H2O (0.013 g,
0.576 mmol) was suspended in a CH2Cl2 solution of 4a (0.070 g,
0.55 mmol) and vigorously stirred for 10 min. The reaction
progress was monitored by IR (disappearance of m(C≡O) at
2031vs and 2010vs, and appearance of m(C≡O) at 2058 cm−1).
Addition of MeOH following concentration of the resulting
solution gave a yellow solid. (Yield: 0.060 g, 73%). Slow diffusion
of MeOH into a dichloromethane solution at −20 ◦C gives
yellow crystals. Anal. Calc. for C38Cl2H66I2O2P2Pt2Sn: C, 31.47;
H, 4.58. Found: C, 31.23; H, 4.69. NMR data: see Table 8. A
supplementary structure is available with a bad R factor (see
supplementary crystallographic data and Table 12).
[Pt2Cl2(SnCl2)(CO)2(PCy3)2] (9c). As for 9a starting from
4c. Analytically pure sample could not been obtained due to
decomposition. NMR data: see Table 8.
CCDC reference numbers 246661–246668.
See http://www.rsc.org/suppdata/dt/b4/b415311a/ for cry-
stallographic data in CIF or other electronic format.
DFT calculations
All computations were carried out with the Amsterdam Density
Functional (ADF) software package.56–58 The computation of
nuclear coupling constants employed the cpl code implemented
by Autsbach and Ziegler,59,60 which calculates analytically nu-
clear spin–spin coupling constants based on the relativistic zero-
order regular approximation (ZORA).61,62 The Vosko–Wilk–
Nusair (VWN)63 local density functional and the Becke–Lee–
Yang–Parr64–67 exchange and correlation functional have been
used for geometry optimization. As basis sets ZORA/TZ2P
were used for all elements.58 Electrons in the lower shells either
3 2 4
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5

View Article Online
were (frozen core: Pt–4f, C–1s, O–1s, N–1s, Cl–2p, I–4p) or were
not (full electron) treated with the frozen core approximation.
Preliminary calculations on the model systems containing PH3
as ligand have shown that coupling constants are dominated
by the scalar relativistic Fermi contact (FC) term while the
other three, i.e. spin–dipole (SD), paramagnetic orbital (OP),
and diamagnetic orbital (OD) terms, were found to be relatively
negligible. Based on this observation, only the FC term was
calculated for the larger systems (PMe3).
30 C. E. Briant, D. L. Gilmour and D. M. P. Mingos, J. Organomet.
Chem., 1986, 308, 381.
31 P. S. Pregosin, Annu. Rep. NMR Spectrosc., 1986, 17, 285.
32 C. S. Brownings and D. H. Farrar, Organometallics, 1989, 8,
813.
33 A. Albinati, K.-H. Dahmen, F. Demartin, J. F. Forward, C. J.
Longley, D. M. P. Mingos and L. M. Venanzi, J. Chem. Soc., Chem.
Commun., 1992, 924.
34 A. Dedieu and R. Hoffman, J. Am. Chem. Soc., 1978, 100, 2074.
35 R. J. Goodfellow, P. L. Goggin and L. M. Venanzi, J. Chem. Soc. A,
1967, 1897.
36 D. Adams, P. J. Chandler and R. G. Churchill, J. Chem. Soc. A, 1967,
1072.
References
37 N. M. Boag, P. L. Goggin, R. J. Goodfellow and I. R. Herbert,
J. Chem. Soc., Dalton Trans., 1983, 1101.
1 R. A. Periana, D. J. Taube, S. Gamble, H. Taube, T. Satoh and H.
Fujii, Science, 1998, 280, 560.
2 T. M. Gilbert, I. Hristov and T. Ziegler, Organometallics, 2001, 20,
1183.
3 S. Sakaki, B. Biswas, and M. Sugimoto, J. Chem. Soc., Dalton Trans.,
1997.
38 R. J. Goodfellow and I. R. Herbert, Inorg. Chim. Acta, 1982, 65,
L161.
39 G. W. Bushnell, D. T. Eadie, A. Pidcock, A. R. Sam, R. D. Holmes-
Smith and S. R. Stobart, J. Am. Chem. Soc., 1982, 104, 5837.
40 M. C. Janzen, M. C. Jennings and R. J. Puddephatt, Organometallics,
2001, 20, 4100.
41 P. S. Pregosin and H. Rueger, Inorg. Chim. Acta, 1984, 86, 55.
42 M. Green, J. A. K. Howard, M. Murray, J. L. Spencer and F. G. A.
Stone, J. Chem. Soc., Dalton Trans., 1977, 1509.
43 J. Autsbach and B. Le Guennic, J. Am. Chem. Soc., 2003, 125, 13585.
44 A. Klamt, J. Phys. Chem., 1995, 99, 2224.
45 A. Klamt and V. Jones, J. Chem. Phys., 1996, 105, 9972.
46 A. Klamt and G. Schuurmann, J. Chem. Soc. ,Perkin Trans., 1993, 2,
799.
4 J. P. Collman, L. S. Hegedus, J. R. Norton, and R. G. Finke,
Principles and Applications of Organotransition Metal Chemistry,
Oxford University Press, 1987.
5 M. C. Jennings and R. J. Puddephatt, Inorg. Chem., 1988, 27, 4280.
6 G. J. Spivak, L. Hao, J. J. Vittal and R. J. Puddephatt, J. Am. Chem.
Soc., 1996, 118, 225.
7 L. Hao, L. Manojlovic-Muir, K. W. Muir, R. J. Puddephatt, G. J.
Spivak, J. J. Vittal and D. Yuffitt, Inorg. Chim. Acta, 1997, 265, 65.
8 Z. Beni, R. Scopelliti and R. Roulet, Inorg. Chem. Commun., 2004,
7/8, 935.
9 R. Ros, A. Tassan, R. Roulet, G. Laurenczy, V. Duprez and K.
Schenk, J. Chem. Soc., Dalton Trans., 2002, 3565.
10 R. Ros, A. Tassan, R. Roulet, V. Duprez, S. Detti, G. Laurenczy and
K. Schenk, J. Chem. Soc., Dalton Trans., 2001, 2858.
11 R. Ros, A. Tassan, G. Laurenczy and R. Roulet, Inorg. Chim. Acta,
2000, 303, 94.
47 W. Henderson and A. G. Oliver, Acta Crystallogr., Sect. C, 1999,
C55(9), 1408.
48 C. Y. Mok, S. G. Tan and G. C. Chan, Inorg. Chim. Acta, 1990, 176,
43.
49 D. G. Evans, M. F. Hallam, D. M. P. Mingos and R. W. M. Wardle,
J. Chem. Soc., Dalton Trans., 1987, 1889.
50 H. C. Clark, A. S. Goel and C. S. Wong, Inorg. Chim. Acta, 1979, 34,
12 R. Ros, G. Facchin, A. Tassan, R. Roulet, G. Laurenczy and F.
Lukacs, J. Cluster Sci., 2001, 12, 99.
13 A. S. Goel and S. Goel, Inorg. Nucl. Chem. Lett., 1980, 16, 397.
14 C. Couture, D. H. Farrar, D. S. Fisher and R. R. Gukathasan,
Organometallics, 1987, 6, 532.
159.
51 L. Manojlovic-Muir, K. W. Muir and R. Walker, J. Chem. Soc.,
Dalton Trans., 1976, 1379.
52 L. Manojlovic-Muir, K. W. Muir and T. Solomun, J. Organomet.
Chem., 1977, 142, 265.
15 J. Autsbach, D. I. Ciprian and T. Ziegler, J. Am. Chem. Soc., 2003,
125, 1028.
16 M. Black, R. H. B. Mais and P. G. Owston, Acta Crystallogr., Sect.
B, 1969, 25, 1760.
17 S. F. Watkins, J. Am. Chem. Soc. A, 1970, 168.
18 G. K. Anderson, H. C. Clark and J. A. Davis, Organometallics, 1982,
1, 550.
19 I. M. Al-Najjar, Inorg. Chim. Acta, 1987, 128, 93.
20 B. L. Simms, M. Shang, J. Lu, W. J. Youngs and J. A. Ibers,
Organometallics, 1987, 6, 1118.
53 K.-H. Dahmen, A. Moor, R. Naegeli and L. M. Venanzi, Inorg.
Chem., 1991, 30, 4285.
54 G. M. Sheldrick, SHELXL-97, Institut for Anorganische Chemie der
¨
Universitat, Tmmanstarsse 4, D-3400 Gottingen, Germany, 1997..
55 L. J. Faruggia, Appl. Crystallogr., 1997, 30, 565.
56 C. Fonseca Guerra, O. Visser, J. G. Snijders, G. te Velde, and E. J.
Baerends, in Parallelization of the Amsterdam Density Functional
Program, Cagliari, 1995.
¨
57 G. te Velde, F. M. Bickelhaupt, E. J. Baerends, S. J. A. van Gisbergen,
C. Fonseca Guerra, J. G. Snijders and T. Ziegler, J. Comput. Chem.,
2001, 22, 931.
21 A. J. Blake, R. O. Gould, A. M. Marr, D. W. H. Rankin and M.
Schroder, Acta Crystallogr., Sect. C, 1989, 45, 1218.
22 N. M. Boag and M. S. Ravetz, Acta Crystallogr., Sect. C, 1996, 52,
1942.
58 ADF, Amsterdam Density Functional, Theoretical Chemistry, Vrije
Universiteit, Amsterdam..
59 J. Autsbach and T. Ziegler, J. Chem. Phys., 2000, 113, 9410.
60 J. Autsbach and T. Ziegler, J. Chem. Phys., 2000, 113, 936.
61 E. van Lenthe, E. J. Baerends and J. G. Snijders, J. Chem. Phys., 1993,
99, 4597.
23 G. K. Anderson and R. J. Cross, J. Chem. Soc., Dalton Trans., 1980,
1988.
24 T. E. Mu¨ller, F. Ingold, S. Menzer, D. M. P. Mingos and D. J. Williams,
J. Organomet. Chem., 1997, 528, 163.
25 A. Albinati, G. Carturan and A. Musco, Inorg. Chim. Acta, 1976, 16,
L3.
26 A. Albinati, Inorg. Chim. Acta, 1977, 22, L31.
27 D. J. Underwood, R. Hoffman, K. Tatsuni, A. Nakamura and Y.
Yamamoto, J. Am. Chem. Soc., 1985, 107, 5968.
28 D. M. P. Mingos and T. Slee, J. Organomet. Chem., 1990, 394, 679.
29 A. Moor, P. S. Pregosin and L. M. Venanzi, Inorg. Chim. Acta, 1981,
48, 153.
62 E. van Lenthe, The ZORA equation, Vrije Universiteit, Amsterdam,
1996..
63 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1989, 58, 1200.
64 A. D. Becke, Phys. Rev. A, 1988, 38, 3098.
65 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
66 B. G. Johnson, P. M. W. Gill and J. A. Pople, J. Chem. Phys., 1993,
98, 5612.
67 T. V. Russo, R. L. Martin and P. J. Hay, J. Chem. Phys., 1994, 101,
7729.
D a l t o n T r a n s . , 2 0 0 5 , 3 1 5 – 3 2 5
3 2 5