Chiral and achiral phosphine derivatives of alkylidyne tricobalt carbonyl clusters as catalyst precursors for (asymmetric) inter- and intramolecular Pauson-Khand reactions

Article abstract of DOI:10.1039/b717698h

Phosphine derivatives of alkylidyne tricobalt carbonyl clusters have been tested as catalysts/catalyst precursors in intermolecular and (asymmetric) intramolecular Pauson-Khand reactions. A number of new phosphine derivatives of the tricobalt alkylidyne clusters [Co₃(μ₃-CR)(CO) ₉] (R = H, CO₂Et) were prepared and characterised. The clusters [Co₃(μ₃-CR)(CO) _9-x(PR′₃)_x] (PR′₃ = achiral or chiral monodentate phosphine, x = 1-3) and [Co₃(μ ₃-CR)(CO)₇(P-P)] (P-P = chiral diphosphine; 1,1′- and 1,2-structural isomers) were assayed as catalysts for intermolecular and (asymmetric) intramolecular Pauson-Khand reactions. The phosphine-substituted tricobalt clusters proved to be viable catalysts/catalyst precursors that gave moderate to very good product yields (up to ～90%), but the enantiomeric excesses were too low for the clusters to be of practical use in the asymmetric reactions. The Royal Society of Chemistry.

Full text of DOI:10.1039/b717698h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Chiral and achiral phosphine derivatives of alkylidyne tricobalt carbonyl
clusters as catalyst precursors for (asymmetric) inter- and intramolecular
Pauson–Khand reactions†
Viktor Moberg,a M. Abdul Mottalib,a De´sire´e Sauer,a Yulia Poplavskaya,a Donald C. Craig,b
Stephen B. Colbran,b Antony J. Deemingc and Ebbe Nordlander*a
Received 20th November 2007, Accepted 31st January 2008
First published as an Advance Article on the web 13th March 2008
DOI: 10.1039/b717698h
Phosphine derivatives of alkylidyne tricobalt carbonyl clusters have been tested as catalysts/catalyst
precursors in intermolecular and (asymmetric) intramolecular Pauson–Khand reactions. A number of
new phosphine derivatives of the tricobalt alkylidyne clusters [Co3(l3-CR)(CO)9] (R = H, CO2Et) were
prepared and characterised. The clusters [Co3(l3-CR)(CO)9−x(PRꢀ3)x] (PRꢀ3 = achiral or chiral
monodentate phosphine, x = 1–3) and [Co3(l3-CR)(CO)7(P–P)] (P–P = chiral diphosphine; 1,1ꢀ- and
1,2-structural isomers) were assayed as catalysts for intermolecular and (asymmetric) intramolecular
Pauson–Khand reactions. The phosphine-substituted tricobalt clusters proved to be viable
catalysts/catalyst precursors that gave moderate to very good product yields (up to ∼90%), but the
enantiomeric excesses were too low for the clusters to be of practical use in the asymmetric reactions.
Introduction
The Pauson–Khand reaction1,2 may be viewed as a [2 + 2 +
1] cycloaddition reaction between an alkyne, an alkene, and
carbon monoxide leading to the formation of a cyclopentenone
(Scheme 1), and inter- and intramolecular3 Pauson–Khand cy-
Scheme 1 General reaction scheme for the intermolecular catalytic
clizations are widely employed in synthetic organic chemistry. The
cyclization reaction is effected by polynuclear metal complexes
that may be used as either stoichiometric reagents or catalysts. In
the original Pauson–Khand reaction,1 the alkyne was coordinated
to [Co2(CO)8] forming [Co2(CO)6(alkyne)] which was heated in the
presence of a suitable alkene. It has been found that the cyclization
reaction can be accelerated by the addition of promoters, e.g.
tertiary amine oxides which effect oxidative decarbonylation1,4
or hard Lewis bases (alcohols, amines) which promote CO
dissociation.1,5 Other metal-based reagents, e.g. titanocenes,6 have
been used for cyclocarbonylation of enynes and new synthetic
methods for cyclopentenone synthesis have also been developed.7
Catalytic Pauson–Khand cyclization reactions were first de-
scribed by Pauson and co-workers.8 Sugihara et al.2,9 have
shown that alkylidyne tricobalt carbonyl clusters are efficient
catalysts/catalyst precursors for Pauson–Khand reactions, and
other transition metal clusters, e.g. [Ru3(CO)12],10 have also been
Pauson–Khand reaction.
used to catalyse this cycloaddition reaction. While cyclizations
leading to cyclopentenones that are substituted in positions 4 and
5 of the cyclopentenone invariably lead to the exo compound,
cyclopentenones that are substituted only in position 4 give rise
to enantiomers. Enantioselective and diastereoselective catalytic
Pauson–Khand reactions have been developed by adding auxiliary
chiral ligands, e.g. BINAP, to [Co2(CO)8].11,12 Considering the
excellent catalytic properties that have been reported for [Co3(l3-
CR)(CO)9] (R = H, CO2Et),2,9 we have prepared chiral derivatives
of alkylidyne tricobalt carbonyl clusters via coordination of
chiral phosphines to the cluster backbone in order to investigate
the catalytic properties of the new clusters in enantioselective
Pauson–Khand reactions. Here we report the preparation of a
number of achiral and chiral phosphine-substituted alkylidyne
tricobalt carbonyl clusters, viz. [Co3(l3-CR)(CO)9−x(PRꢀ3)x] (x =
1–3: R = CO2Et, PRꢀ3 = PEt3; R = H, PRꢀ3 = PMe2Ph), [Co3(l3-
CH)(CO)8{(S)-NMDPP}], [{Co3(l3-CH)(CO)8}2(l-P–P)] (P–P =
aInorganic Chemistry Research Group, Chemical Physics, Center for Chem-
istry and Chemical Engineering, Lund University, Box 124, SE-221 00, Lund,
Sweden. E-mail: Ebbe.Nordlander@chemphys.lu.se; Fax: (+46) 46 222 44
39
dppe, dppm) and [Co3(l3-CH)(CO)7(P–P)] [P–P
=
(R,R)-
DIPAMP, (S,S)-CHIRAPHOS, (R)-BINAP, (R)-(S)-Josiphos3,
(R,R)-Me-DUPHOS, (R)-PROPHOS, (R,R)-NORPHOS)],‡ and
their abilities to catalyze (enantioselective) Pauson–Khand cy-
clizations.
bSchool of Chemistry, University of New South Wales, Sydney, NSW, 2052,
Australia
cDepartment of Chemistry, University College London, 20 Gordon Street,
London, UK WC1H OAJ
† Electronic supplementary information (ESI) available: Figure S1. Two
views (with 30% thermal ellipsoids) of the ‘bent’ carbonyl ligand in
the crystal structure of [Co3(l3-CH)(CO)7(l-1,2-DIPAMP)] (12). CCDC
reference numbers 668301–668304. For crystallographic data in CIF or
other electronic format see DOI: 10.1039/b717698h
‡ These ligands are hereafter referred to without chirality notation, i.e.
DIPAMP = (R,R)-DIPAMP etc.
2442 | Dalton Trans., 2008, 2442–2453
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Results and discussion
New phosphine-substituted alkylidyne tricobalt carbonyl clusters
A number of mono- and bidentate phosphine derivatives of
[Co3(l3-CR)(CO)9] (R = H, CO2Et) of the general formulae
[Co3(l3-CR)(CO)9−x(PRꢀ3)x] (x = 1–3) and [Co3(l3-CRꢀ)(CO)7(P–
P)] (P–P = bidentate phosphine) were synthesized according to
established synthetic procedures (see below and Experimental
section). In addition, new clusters with the less common structures
that are depicted in Fig. 1 could also be isolated from these syn-
theses. The chiral ligands that have been used in this investigation
are depicted in Fig. 2.
Fig. 2 Structures of the chiral ligands used in this investigation.
[Co3(l3-CH)(CO)8(NMDPP)] (7) in 51% yield. The spectroscopic
data from IR, 1H and 31P NMR measurements, the mass spectrum
and the microanalysis are in full agreement with the suggested
formula. In addition to the typical IR bands found in monosub-
stituted alkylidyne tricobalt clusters, the spectrum reveals a signal
at 1858 cm−1 similar to that found in clusters 2 and 4.
Fig. 1 Schematic structures of the phosphine coordination modes in the
new cobalt clusters that have been prepared in this study.
Derivatives of bidentate phosphine ligands. The parent cluster
[Co3(l3-CH)(CO)9] was reacted with a stoichiometric amount
(1 eq.) of a diphosphine ligand P–P (P–P = dppe, dppm) in
toluene at room temperature. Both bridged dimers, [{Co3(l3-
CH)(CO)8}2(l-P–P)], and disubstituted monomeric clusters,
[Co3(l3-CH)(CO)7(P–P)], were obtained. The isolated yields of
[{Co3(l3-CH)(CO)8}2(l-P–P)] were low to moderate (P–P =
dppe (8), 25%; dppm (9), 4%) while the disubstituted [Co3(l3-
CH)(CO)7(P–P)] clusters were formed in relatively good yields
(P–P = dppe (10), 50%; dppm (11), 67%). The chiral bidentate
phosphine DIPAMP (see Fig. 2) was reacted with a stoichiometric
amount of the parent cluster, which led to the formation of [Co3(l3-
CH)(CO)7(DIPAMP)] (12) in 79% yield. The clusters 8–12 were
characterised by NMR and IR spectroscopy, mass spectrometry
and elemental analysis, all of which were consistent with the
proposed molecular formulae.
The chiral diphosphines CHIRAPHOS, BINAP, NORPHOS,
Me-DUPHOS and PROPHOS (see Fig. 2) were reacted with
a stoichiometric equivalent of [Co3(l3-CH)(CO)9] in refluxing
(CHIRAPHOS, BINAP) or warm (40 ◦C) hexane, which gave
[Co3(l3-CH)(CO)7(P–P)] in good yields (P–P = BINAP (13), 51%;
CHIRAPHOS (14), 98%; NORPHOS (15), 55%; PROPHOS (16),
59%; Me-DUPHOS (17), 70%). Clusters 13–17 were analysed
and identified on the basis of NMR and IR spectroscopy and
mass spectrometry, which were in agreement with the proposed
molecular formulae. Cluster 17 shows distinct absorption bands at
Derivatives of monodentate phosphine ligands. The parent
clusters [Co3(l3-CR)(CO)9] (R = CO2Et, H) were reacted with
either a stoichiometric amount (1 eq.) or an excess of phos-
phine ligand, PEt3 or PMe2Ph, in dichloromethane at room
temperature. When one equivalent of the ligand was used, the
mono- and di-phosphine-substituted clusters were obtained in
moderate yields {1: [Co3(l3-CO2Et)(CO)8(PEt3)], 50%; 2: [Co3(l3-
CH)(CO)8(PMe2Ph)], 46%; 3: [Co3(l3-CH)(CO)7(PEt3)2], 11%;
4: [Co3(l3-CH)(CO)7(PMe2Ph)2], 17%}. When an excess of the
relevant phosphine was used, the corresponding trisubstituted
clusters [Co3(l3-CR)6(PRꢀ3)3] were the exclusive products and were
isolated in moderate to relatively good yields [R = H: PRꢀ3
=
PEt3 (5) (50%), PMe2Ph (6) (75%)]. Clusters 1–6 were identified
on the basis of NMR and IR spectroscopy, mass spectrometry
and elemental analyses, which were consistent with the proposed
molecular formulae. The IR spectra of clusters 1, 5 and 6 are
typical of phosphine-substituted alkylidyne tricobalt carbonyl
complexes.13,14 The IR spectra of 2, 3 and 4 in cyclohexane solution
exhibit weak bands in the region 1880–1824 cm−1 that indicate that
these clusters contain bridging carbonyl ligands but no such bands
could be detected in the solid state (KBr pellet) IR spectrum of 2.
The chiral monodentate ligand NMDPP (Fig. 2) was reacted
with [Co3(l3-CH)(CO)9] at elevated temperature for 4 h. Only one
product was isolated and identified as the mono-substituted cluster
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2443
©

View Article Online
1863 and 1819 cm−1, indicating the presence of bridging carbonyls,
which was verified by the molecular structure obtained from X-ray
diffraction analysis (vide infra).
the second phosphine moiety coordinating in an equatorial site
on the same cobalt. Bridging and chelating coordination modes
for Josiphos3 at a tetrahedral cluster core have been detected
for two isomers of [H4Ru4(CO)10(Josiphos3)]16 but in this case
no interconversion between the two isomers has been detected.
The fluxionality/equilibrium that has been observed for 18 may
be due to a poorer fit of the ligand in the bridging mode than
in the tetraruthenium cluster – the Co–Co distance in 18 is
Reaction of the chiral diphosphine Josiphos3 (Fig. 2) with
[Co3(l3-CH)(CO)9] in refluxing dichloromethane gave [Co3(l3-
CH)(CO)7(Josiphos3)] (18) in 83% yield. Cluster 18 was charac-
terised by NMR and IR spectroscopy and mass spectrometry. The
FAB mass spectrum confirmed the proposed molecular formula.
1
However, H and 31P NMR as well as IR spectra indicate the
approximately 0.5 A shorter than in the tetraruthenium cluster
˚
1
presence of isomers in solution. The peak areas in the H NMR
and this may cause a strain in the seven-membered ring that
is formed when the diphosphine bridges a metal–metal bond,
thus enabling interconversion of the two isomers at relatively low
energies. An alternative explanation is that the two interconverting
isomers have the same (bridging) coordination mode of the
diphosphine and the configurational isomerism consists of an
interconversion of terminal and bridging carbonyl ligands, while
the third, non-interconverting isomer contains the diphosphine
ligand in a different (chelating) coordination mode. It may be
expected that the energy barrier for such an isomerization reaction
is relatively low. Finally, it is possible that two diastereomeric forms
of 18 exist – the bridging coordination of a heterodidentate ligand
to a metal triangle leads to the formation of enantiomers17 which,
in combination with the chirality of the enantiomerically pure
phosphine ligand, leads to the formation of two diastereomers,
one of which may be involved in configurational isomerism.
spectra show some inconsistency. While the peaks for all protons
on the ligand integrate correctly, the peak for the l3-methylidyne
proton only measures 1/3 of the expected peak area. In the 31P
NMR, a set of six resonances are observed at d 32.7 (s), 37.2 (s)
39.9 (s), 54.2 (s), 61.2 (s) and 69.8 (s). A variable temperature 31
P
NMR experiment (Fig. 3) revealed that the intensities of the peaks
at d 37.2 and 39.9 decreased as the temperature was raised, until at
333 K (60 ◦C) they almost disappeared. In contrast, the two peaks
at d 54.2 and 69.8 increased significantly when the sample was
heated (Fig. 3). The observed isomerisation process was reversible;
once the temperature was decreased to ambient temperature (298
K), the intensities of peaks at d 37.2 and 39.9 increased, while the
peak intensities at d 54.2 and 69.8 decreased. The intensities of
the peaks at d 32.7 and 61.2 were unaffected by the temperature.
These results indicate that there are multiple isomers present in
the solution of cluster 18. The isomers may be configurational
isomers that involve different bonding modes of the diphosphine
and (or) carbonyl ligands. For example, [Co3(l-CR)(CO)7(P–P)]
clusters have been shown to undergo reversible isomerisation
reactions wherein the coordination mode of the diphosphine
changes between chelating and bridging.15 We postulate that such
an equilibrium exists for 18, with a chelating isomer, in which both
phosphine moieties are coordinated in equatorial positions on the
Co3(l3-CR) framework, interconverting with the corresponding
bridged dimer. As the two phosphine moieties of the Josiphos3
are inherently different, two 31P resonances are observed for each
isomer. The third isomer, which does not interconvert with the
two other isomers, is proposed to contain the diphosphine in
a chelating coordination mode, but with one phosphine moiety
coordinating in the (unique) axial site of one cobalt atom and
X-Ray crystallography
[Co3(l3-CH)(CO)8(PMe2Ph)] (2). The molecular structure of
cluster 2 is shown in Fig. 4 and relevant bond lengths and angles
are reported in Table 1. The three cobalt atoms form a triangle
which is capped by the l3-CH moiety, and the dimethylphenyl
phosphine ligand is equatorially coordinated to Co(3). The cobalt
atoms complete their pseudo-octahedral coordination spheres
with terminal carbonyl ligands. The Co–Co bonds are notably
˚
different in lengths, ranging from 2.4636(6) A [Co(2)–Co(3)] to
˚
2.4899(5) A [Co(1)–Co(3)], and are similar to those found in the
related cluster [Co3(l3-CC(O)NHPri)(CO)8(PPh3)].13 The average
˚
of the three Co–Co distances (2.4746 A) is similar to those
Fig.
4 An ORTEP drawing of the molecular structure of
Fig. 3 31P NMR spectra of [Co3(l3-CH)(CO)7{(R)-(S)-Josiphos3}] (18)
showing the increase/decrease of peak intensities upon heating of the
sample.
[Co3(l3-CH)(CO)8(PMe2Ph)] (2) showing the atom numbering scheme.
Thermal ellipsoids are drawn at the 30% probability level. For the sake of
clarity, all hydrogen atoms have been omitted.
2444 | Dalton Trans., 2008, 2442–2453
This journal is The Royal Society of Chemistry 2008
©

View Article Online
◦
˚
Table 1 (Contd.)
Cluster 17
Table 1 Relevant bond lengths (A) and angles ( ) for clusters [Co3(l3-
CH)(CO)8(PMe2Ph)] (2), [Co3(l3-CH)(CO)6(PMe2Ph)3] (6), [Co3(l3-
CH)(CO)7(l-1,2-DIPAMP)] (12) and [Co3(l3-CH)(CO)7(1,1-DUPHOS)]
(17)
Co(3)–Co(1)–P(1)
P(1)–Co(1)–P(2)
Co(3)–Co(1)–P(2)
Co(2)–Co(1)–P(2)
130.58(4)
88.90(5)
136.18(4)
112.41(5)
Co(2)–Co(1)–P(1)
P(1)–Co(1)–CH
P(2)–Co(1)–CH
132.95(5)
178.4(2)
90.9(1)
Cluster 2
Co(1)–Co(2)
Co(2)–Co(3)
Co(2)–C(1)
Co(1)–C(11)
Co(1)–C(13)
Co(2)–C(22)
Co(3)–C(31)
Co(3)–P(1)
P(1)–C(47)
2.4703(5)
2.4636(6)
1.913(3)
1.792(4)
1.814(3)
1.791(4)
1.773(3)
2.1981(9)
1.810(4)
Co(1)–Co(3)
Co(1)–C(1)
Co(3)–C(1)
Co(1)–C(12)
Co(2)–C(21)
Co(2)–C(23)
Co(3)–C(32)
P(1)–C(41)
P(1)–C(48)
2.4899(5)
1.888(3)
1.881(3)
1.761(3)
1.777(4)
1.807(4)
1.808(4)
1.819(3)
1.828(4)
found in tricobalt alkylidyne nonacarbonyl clusters, e.g. [Co3(l3-
18
˚
CMe)(CO)9] (Co–Comean = 2.467 A), [Co3(l3-CPh)(CO)9] (Co–
Comean = 2.467 A),19 [Co3(l3-CC(O)Ph)(CO)9] (Co–Comean = 2.470
˚
20
˚
A), in numerous phosphine derivatives of the type [Co3(l3-
CR)(CO)9−x(PR3)x]12,19–23 and [Co3(l3-CR)(CO)7(P–P)],13,24 and
also for Co–Co bonds found in various polynuclear cobalt
complexes.25 The PMe2Ph ligand coordinates in an equato-
rial position relative to the Co triangle, i.e. cis to a metal–
alkylidyne bond, as seen for other monosubstituted tricobalt
alkylidyne clusters, e.g. [Co3(l3-CC(O)NHPri)(CO)8(PPh3)]13 and
[Co3(l3-CMe)(CO)7{PPh2(C4H3S)}2].21 The Co–P bond length
P(1)–Co(3)–Co(1)
P(1)–Co(3)–Co(2)
C(1)–Co(3)–P(1)
100.05(3)
158.70(3)
112.40(10)
C(31)–Co(3)–P(1)
C(32)–Co(3)–P(1)
91.56(12)
97.30(11)
Cluster 6
Co(1)–Co(2)
Co(2)–Co(3)
Co(2)–C(7)
Co(1)–P(1)
Co(1)–C(3)
Co(2)–P(2)
Co(2)–C(2)
Co(3)–P(3)
Co(3)–C(3)
2.4445(9)
2.4377(9)
1.888(5)
2.1632(16)
1.920(6)
2.1660(17)
1.973(5)
2.1705(15)
1.972(6)
Co(1)–Co(3)
Co(1)–C(7)
Co(3)–C(7)
Co(1)–C(1)
Co(1)–C(4)
Co(2)–C(1)
Co(2)–C(5)
Co(3)–C(2)
Co(3)–C(6)
2.4436(9)
1.889(5)
1.895(5)
1.958(6)
1.751(6)
1.936(6)
1.754(6)
1.918(6)
1.756(7)
[2.1981(9) A] is normal for a Co–PR3 bond,26 e.g. [Co3(l3-
˚
˚
CMe)(CO)7{PPh2(C4H3S)}2] (Co–Pmean = 2.212 A), but sig-
nificantly shorter than the Co–P distance found in [Co3(l3-
i
13
˚
CC(O)NHPr )(CO)8(PPh3)] (2.2400 A). The average Co–C(1)
˚
distance (1.894 A) does not show any significant deviation
from other reported alkylidyne nonacarbonyl clusters and their
phosphine derivatives.13
Co(3)–Co(2)–Co(1)
Co(2)–Co(3)–Co(1)
P(3)–Co(3)–Co(2)
P(3)–Co(3)–Co(1)
C(5)–Co(2)–P(2)
P(2)–Co(2)–Co(3)
C(4)–Co(1)–P(1)
P(1)–Co(1)–Co(3)
60.07(3)
60.11(3)
133.38(5)
128.08(5)
102.0(2)
128.66(5)
101.5(2)
131.57(5)
Co(3)–Co(1)–Co(2)
C(6)–Co(3)–P(3)
C(7)–Co(3)–P(3)
C(7)–Co(2)–P(2)
P(2)–Co(2)–Co(1)
C(7)–Co(1)–P(1)
P(1)–Co(1)–Co(2)
59.83(3)
100.6(2)
97.31(16)
95.79(17)
130.85(5)
95.96(16)
128.28(6)
[Co3(l3-CH)(CO)6(PMe2Ph)3] (6). Fig. 5 presents a view of
cluster 6 and relevant bond lengths and angles are summarised
in Table 1. The three cobalt atoms in the cluster core form an
almost equilateral triangle capped by the l3-methylidyne moiety.
Each dimethylphenyl phosphine ligand is equatorially bound
to one cobalt atom and is cis to an alkylidyne carbon–cobalt
bond. Cluster 6 contains three terminal carbonyls and three
edge-bridging carbonyls (l-CO), which is consistent with the l-
CO stretches seen in its IR spectrum (vide supra). The l-CO
groups are located in equatorial positions with no deviation
from the plane defined by the triangular cobalt cluster core.
Cluster 12
Co(1)–Co(2)
Co(1)–P(1)
Co(1)–CC(1)
Co(2)–Co(3)
Co(2)–CC(3)
Co(2)–CH
2.500(2)
2.171(3)
1.758(12)
2.506(2)
1.792(12)
1.912(10)
1.791(17)
1.935(11)
Co(1)–Co(3)
Co(1)–CC(1)
Co(1)–CH
2.444(2)
1.796(12)
1.839(10)
2.212(3)
1.773(11)
1.764(16)
1.572(12)
3.692
Co(2)–P(2)
Co(2)–CC(4)
Co(3)–CC(5)
Co(3)–CC(7)
P(1) · · · P(2)
Co(3)–CC(6)
Co(3)–CH
P(1)–Co(1)–CC(2)
Co(1)–Co(2)–P(2)
P(2)–Co(2)–CC(3)
P(2)–Co(2)–CH
93.7(4)
103.24(8)
100.2(4)
107.2(3)
140.7(18)
P(1)–Co(1)–CC(1)
P(1)–Co(1)–CH
Co(3)–Co(2)–P(2)
P(2)–Co(2)–CC(4)
101.4(3)
100.7(3)
156.35(9)
94.2(4)
Co(3)–CC(7) OC(7)
Cluster 17
Co(1)–Co(2)
Co(1)–P(1)
Co(1)–CC(1)
Co(1)–CH
Co(2)–CC(1)
Co(2)–CC(4)
Co(2)–CH
Co(3)–CC(5)
Co(3)–CC(7)
P(1) · · · P(2)
2.528(1)
2.218(1)
1.888(5)
1.889(5)
2.054(5)
1.785(6)
1.905(5)
1.769(6)
1.902(6)
3.074
Co(1)–Co(3)
Co(1)–P(2)
2.501(1)
2.171(1)
1.832(6)
2.432(1)
1.773(6)
2.008(5)
2.175(6)
1.797(6)
1.905(4)
Co(1)–CC(2)
Co(2)–Co(3)
Co(2)–CC(3)
Co(2)–CC(7)
Co(3)–CC2)
Co(3)–CC(6)
Co(3)–CH
Fig. 5 An ORTEP drawing of the molecular structure of [Co3(l3-
CH)(CO)6(PMe2Ph)3] (6) showing the atom numbering scheme. Thermal
ellipsoids are drawn at the 30% probability level. For the sake of clarity,
all hydrogen atoms have been omitted.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2445
©

View Article Online
The cobalt carbon distances for the three l-CO groups range
˚
˚
from 1.918(6) A [Co(3)–C(2)] to 1.973(5) A [Co(2)–C(2)], with
˚
an average distance of 1.946 A, and are somewhat longer than
those observed in [Co3(l3-CPh)(CO)7(1,1-bpcd)] [bpcd = 4,5-
27
˚
bis(diphenylphosphino)-4-cyclopenten-1,3-dione] (1.92 A). The
˚
three Co–Co bond lengths exhibit a mean distance of 2.4419 A,
which is significantly shorter than that found in 2. The three
˚
˚
Co–P distances vary from 2.163(2) A [Co(1)–P(1)] to 2.170(2) A
˚
[Co(3)–P(3)], and Co–Pmean (2.1665 A) is slightly shorter than that
˚
found in cluster 2 [2.1981(9) A]. All three phosphorus atoms in the
coordinated ligands occupy axial positions, cis to the capping l3-
methylidyne moiety. There is a pseudo-C3 axis going through the
alkylidyne C–H bond so that the phenyl moieties of the phosphines
form a cage around the l3-methylidyne moiety (Fig. 6).
Fig. 7 An ORTEP drawing of the molecular structure of [Co3(l3-
CH)(CO)7(l-1,2-DIPAMP)] (12) showing the atom numbering scheme.
Thermal ellipsoids are drawn at the 30% probability level. For the sake of
clarity, all hydrogen atoms have been omitted.
is leaning towards the two cobalt atoms bearing the phosphine
ligand. This phenomenon has been observed previously and has
been reported by Downard et al.30 The apical carbon adopts a
“semicapping orientation” with respect to the triangular array of
cobalt atoms. It is suggested that the observed displacement is a
result of the uneven increase of electron density on the CCo3 core,
as a consequence of the substitution of carbonyl ligands with
(di)phosphines. The di-equatorial diphosphine ligand prevents
the carbonyl ligands from adopting bridged coordination modes
that would assist in the dissipation of electron density from the
metals. An alternative, electron dispersion may occur through
the interaction of appropriate p-orbitals on the capped carbon
and the d-orbitals of Co(1) and Co(2).30 The Co–P distances
˚
are somewhat different in length, 2.171(3) A [Co(1)–P(1)] and
˚
2.212(3) A [Co(2)–P(2)]. The mean Co–P bond length in cluster
Fig. 6 An ORTEP drawing of the molecular structure of [Co3(l3-
CH)(CO)6(PMe2Ph)3] (6), giving a clear view of the pseudo-C3 axis through
the apical C–H bond.
˚
˚
12 (2.192 A) is similar to that found in 17 (2.194 A, vide infra),
and in range for what is acceptable for Co–PR3 bonds.26
[Co3(l3-CH)(CO)7(1,1-Me-DUPHOS)] (17). The molecular
structure of 17 is shown in Fig. 8 and relevant bond lengths and
angles are summarised in Table 1. Cluster 17 contains a chelating
DUPHOS ligand coordinated to Co(1), forming a five-membered
ring. The two phosphine moieties of the ligand are located in axial
positions, almost perpendicular to the Co3-plane, as previously
found in the cluster [Co3(l3-CPh)(CO)7(1,1-bpcd)].27 The Co–
[Co3(l3-CH)(CO)7(l-1,2-DIPAMP)] (12). The molecular
structure of 12 is shown in Fig. 7 (bond lengths and angles are
summarised in Table 1); it verifies the bridging coordination of
the diphosphine ligand DIPAMP across the Co(1)–Co(2) bond.
The Co–Co bond lengths in 12 are significantly different, ranging
˚
˚
from 2.506(2) A [Co(2)–Co(3)] to 2.444(2) A [Co(1)–Co(3)]
˚
˚
and show an average distance of 2.483 A. All carbonyls are
Co bond lengths in the cluster, which range from 2.432(1) A
˚
terminal and both phosphorus donors of the DIPAMP ligand are
coordinated in equatorial positions. The axial carbonyl group of
Co(3) deviates significantly from linearity, exhibiting a bond angle
of 140.7(18)◦ [Co(3)–Cc(7)–Oc(7)] suggestive of sp2 rather than
sp hybridization for carbon Cc(7). Bent metal–carbonyl bonds
occur rather frequently in transition metal carbonyl complexes.
In this case, it may be assigned to weak hydrogen bonding
present between the alkylidyne tricobalt carbonyl clusters in
the crystal lattice, as described by Desiraju28 and Braga et al.29
The cobalt to alkylidyne carbon bond distances are notably
[Co(2)–Co(3)] to 2.528(1) A [Co(1)–P(1)], differ significantly. The
two metal–metal bonds involving Co(1) are long [Co(1)–Co(2) =
˚
˚
2.528(1) A, Co(1)–Co(3) = 2.501(1) A], while the Co(2)–Co(3)
˚
distance is approximately 0.08 A shorter. The length of the two
former bonds may be due to a buildup of electron density on Co(1)
caused by the coordination of the diphosphine but the average
˚
length of the Co–Co bonds, 2.487 A, is in the same range as
other Co–Co single bonds found in related clusters.13,24,27 As in the
case of 6, cluster 17 contains three bridging CO ligands, which is
consistent with its IR spectrum. The cobalt–carbonyl carbon bond
˚
˚
˚
different, 1.839(10) A [Co(1)–CH], 1.912(10) A [Co(2)–CH] and
lengths exhibit a mean distance of 1.976 A, similar to that found
˚
˚
1.935(11) A [Co(3)–CH], i.e. the l3-capping methylidyne moiety
in 6 (1.946 A). Two of the bridging carbonyl ligands in cluster 17,
2446 | Dalton Trans., 2008, 2442–2453
This journal is The Royal Society of Chemistry 2008
©

View Article Online
Table 2 Results from catalytic intermolecular Pauson–Khand synthesis
Yieldb (%)
Entry
Catalysta
19
Triphenylbenzene
1
2
3
4
5
2
4
75
90
45
65
70
13
6
15
—
—
6
16
17
a Reaction conditions: n(substrate)/n(catalyst) = 50, p(CO) = 10 bar, T =
120 ◦C, solvent = toluene, duration = 10 h. b Isolated yields after prep.
TLC (hexane–ethyl acetate 4 : 1 v/v) (see Experimental section for further
details).
carried out in toluene at 120 ◦C under a pressure carbon monoxide
of 10 bar with tricobalt cluster and substrates. No additives
were used. The isolated yield of the cyclopentenone product,
4-phenyltricyclo[5.2.1.02,6]dec-4-en-3-one (19) (Scheme 2) varies
notably for the different catalysts, ranging from moderate to
good (see Table 2). The disubstituted cluster 4 gave the highest
isolated yield (90%) of 19. It was somewhat surprising to find
that the monosubstituted cluster 2 gave notably lower yield of
19 (75%) than 4, as multiple substitution of the metal carbonyl
cluster is expected to make the cluster less prone to further CO
loss/substitution. The results from the diphosphine-derivatised
clusters 16 and 17, 65 and 70% respectively, are notably lower than
that of 4 but somewhat higher than for 2. In the case of 16 and
17, no cluster was recovered after a complete catalytic experiment,
while experiments performed with 2, 4 and 6 gave a red, air-stable
crystalline compound which thus far has eluded identification. The
above-mentioned observations suggest that the cobalt–alkylidyne
clusters are indeed precursors to active catalysts.
Fig.
8 An ORTEP drawing of the molecular structure of
[Co3(l3-CH)(CO)7(1,1-DUPHOS)] (17) showing the atom numbering
scheme. Thermal ellipsoids are drawn at the 30% probability level. For
the sake of clarity, all hydrogen atoms have been omitted.
[Co(1)–Co(3) and Co(2)–Co(3)] coordinate in the same manner
as seen in cluster 6. The carbonyl bridging the Co(1)–C(2) edge is
notably affected by the bulky phosphine ligand, being forced out
of the triangular plane by 23.6◦, in relative trans position to the
capping l3-CH moiety.
˚
The two P–Co vectors are slightly different, 2.218(1) A [Co(1)–
˚
˚
P(1)] and 2.171(1) A [Co(2)–P(2)] (Co–Pmean = 2.194 A). The
observed [P(1)–Co(1)–P(2)] bond angle, 88.90(5)◦, is similar
to the P–Co–P angle found in the related structure [Co3(l3-
CPh)(CO)7(1,1-bpcd)] (89.2◦)27 but notably larger than that ob-
served in the ruthenium cluster [H4Ru4(CO)10(1,1-DUPHOS)]
(83.7◦ or 85.3◦).31 The two phosphorus atoms in the DUPHOS
ligand are separated by two relatively rigid sp2-hybridized carbons,
thus favouring a relatively close approach of the phosphorus atoms
˚
[P(1)–P(2) = 3.074 A] with their lone pairs directed so that a
chelating coordination mode of the ligand is favoured. On the
other hand, the phosphine moieties in DIPAMP are bridged by
two sp3-hybridized carbon atoms that make the backbone less
rigid/more flexible and permit the ligand to bridge a metal–metal
bond as in 12 (vide supra).
Catalysis experiments
The catalytic properties of the phosphine-substituted alkylidyne
tricobalt carbonyl clusters in inter- and intramolecular Pauson–
Khand reactions were investigated. The initial studies were carried
out on clusters with achiral phosphines, in order to establish
that Pauson–Khand cyclizations could indeed be achieved with
phosphine-substituted clusters. These studies were followed by at-
tempts to achieve enantioselective intramolecular Pauson–Khand
reactions. In addition, one attempt to hydrogenate 2-methyl-2-
butenoic (tiglic) acid using [Co3(l3-CH)(CO)7(PMe2Ph)2] (4) as
a catalyst was made, using conditions similar to those reported
for H4Ru4 clusters;31,32 however, no hydrogenation of the substrate
could be detected.
Scheme 2 The cyclisation reaction of norbornene, phenyl acetylene and
carbon monoxide in presence of a cobalt carbonyl catalyst to form
4-phenyltricyclo[5.2.1.02,6]dec-4-en-3-one (19) and 1,3,5-triphenylbenzene.
The cyclotrimerization of alkynes is a common side reaction that
is observed for the Pauson–Khand reaction, and triphenylbenzene
is the side-product formed from trimerization of phenyl acetylene.
Interestingly, no detectable amount of triphenylbenzene was
observed when 16 and 17 were used as catalyst precursors. Even
though the cluster most likely acts as a catalyst precursor, the chiral
diphosphine ligand may be present in the active catalyst, and the
steric bulk of the diphosphine may thus prevent the formation of
the sterically demanding by-product triphenylbenzene.
Intermolecular Pauson–Khand reaction. Clusters 2, 4, 6, 16
and 17 were successfully used as catalyst precursors in the
intermolecular cyclization of norbornene and phenyl acetylene
in the presence of carbon monoxide (Scheme 2). The reaction was
Intramolecular Pauson–Khand reaction. To test for stereose-
lectivity in intramolecular Pauson–Khand reactions, catalysis of
the cyclisation of diethyl allyl propargyl malonate (20) (Scheme 3)
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2447
©

View Article Online
the most poor product yield is the one that gives the highest
enantioselectivity (13% ee). Clusters 15, 16 and 17 gave slightly
higher yields but very low enantioselectivities. In contrast, high
enantioselectivies have been reported for intramolecular Pauson–
Khand reactions catalyzed by Co2(CO)8 with addition of a chiral
auxiliary ligand. Cyclisation of 20 catalyzed by Co2(CO)8/MeO-
BIPHEP33 resulted in 91.5% ee and in a similar reaction using
Co2(CO)8/BINAP11 91% ee was obtained.
Conclusions
Scheme 3 The cyclisation reaction of diethyl allyl propargyl malonate
(20) and carbon monoxide in presence of a cobalt catalyst, leading to
the formation of 7,7-bis(ethoxycarbonyl)bicycle[3.3.0]-1-octene-3-one (21)
and diethyl 2,2-diallylmalonate (22).
We have shown that clusters of the general formulae [Co3(l3-
CR)9−x(PRꢀ3)x] (PRꢀ3 = achiral or chiral monodentate phos-
phine) and [Co3(l3-CR)10(P–P)] (P–P = chiral diphosphine; two
structural isomers) are capable of acting as catalysts/catalyst
precursors for both inter- and intramolecular Pauson–Khand
reactions with moderate to good yields of cyclopentenones. How-
ever, although the intramolecular cyclization reactions catalysed
by [Co3(l3-CH)(CO)7(P–P) (P–P = CHIRAPHOS, NORPHOS,
Me-DUPHOS, PROPHOS) demonstrate the feasibility of using
such chiral clusters in Pauson–Khand reactions, the enantios-
electivities are too poor to make these specific clusters viable
catalysts/catalyst precursors for this type of asymmetric reaction.
using the chiral methylidyne tricobalt carbonyl clusters 4, 7, 12–
18 were performed at 120 C in toluene under 10 bar of carbon
monoxide for 10 h. As in the intermolecular reaction (vide supra),
no additives were used.
Similarly to the intermolecular cyclization experiments, there is
no evidence that the clusters act as catalysts, but rather as catalyst
precursors for the intramolecular cyclisation. Evidence for cluster
fragmentation and/or colloidal particle formation was observed,
such as a substantial colour change of the reaction solution during
the catalytic experiment, deposition of black metallic residues, and
no recovery of the original cluster.
◦
Experimental
The results of the catalytic intramolecular cyclisation reactions
are summarised in Table 3. The yield of the desired cyclopentenone
varies significantly between the different tests, ranging from very
low (14, 6%) to relatively good (7, 76%). The enantiomeric excess
(ee) of 7,7-bis(ethoxycarbonyl)bicycle[3.3.0]-1-octene-3-one (22)
(Scheme 3) was determined by HPLC analysis using a Chiralpak
AD-H analytical column. The enantiomeric excesses obtained,
if any, differ notably for the various cluster catalysts, but are
consistently low. The general trend in the catalysis experiments is
that the clusters that gave (relatively) high product yields generated
a racemate, while low conversion of the starting material gave
“higher” enantioselectivity. Thus, clusters 7, 12, 13 and 18 gave
yields of racemic cyclopentenone in the range of 31–76%. The
cobalt clusters 14, 15, 16 and 17 exhibit the lowest conversions
(slowest catalytic rates) but are able to generate enantioselectivity
in the intramolecular cyclisation reaction. Cluster 14 that gave
General experimental procedures
The parent clusters [Co3(l3-CH)(CO)9], [Co3(l3-CCO2Et)(CO)9]
and [Co3(l3-CCH3)(CO)9] were prepared by a literature method.34
The synthesis of diethyl allyl propargyl malonate (20) was also
carried out by a literature method.35 All reactions were performed
under an inert atmosphere of either argon or nitrogen, and
manipulations of the products were carried out in air. 1H and 31
P
NMR spectra were recorded on Varian Unity 300 MHz or Varian
Inova 500 MHz spectrometers at 298 K in CDCl3, unless stated
otherwise. The 1H NMR spectra were referenced to solvent signals
(CHCl3 = d7.25) while the 31P NMR spectra were referenced to
external 85% H3PO4. Infrared spectra were recorded on a Nicolet
Avatar 360 FT-IR spectrometer. Thin-layer chromatography was
performed on commercially available 20 × 20 cm glass plates,
covered with Merck Kieselgel 60 to 0.25 mm or 0.5 mm thickness.
In the catalysis experiments, a Parr autoclave fitted with a PTFE
liner (30 mL) was used as reaction vessel. Solvents used in synthesis
and catalysis experiments were distilled over appropriate drying
agents and were deoxygenated by bubbling with nitrogen gas prior
to use.
Table 3 Results from catalytic intramolecular Pauson–Khand synthesis
Entry
Catalysta
Yieldb (%)
ee (%)c
1
2
3
4
5
6
7
8
9
7
4
76d
66e
54
39
31
27
17
16
6
Racemate
Racemate
Racemate
Racemate
Racemate
2 (+)
12
18
13
16
17
15
14
Syntheses
[Co3(l3-CCO2Et)(CO)8(PEt3)] (1) and [Co3(l3-CCO2Et)-
(CO)7(PEt3)2] (3). A total of 100 mg (0.194 mmol) [Co3(l3-
CCO2Et)(CO)9] and 23 mg (0.194 mmol) triethylphosphine were
dissolved in 15 mL of dichloromethane and stirred at room
temperature for 18 h. The solvent was removed under reduced
pressure, the solid obtained was dissolved in a small quantity
of dichloromethane, and the products were separated using
preparative TLC (eluent: dichloromethane–hexane 3 : 2 v/v).
Except for unconsumed starting materials, two products were
removed from the TLC plates, extracted with dichloromethane
7 (−)
7 (−)
13 (+)
a Reaction conditions: n(substrate)/n(catalyst) = 50, p(CO) = 10 bar, T =
120 ◦C, solvent = toluene, duration = 10 h. b Isolated yield after prep.
TLC (hexane–ethyl acetate 3 : 2 v/v). c In cases where ee was obtained, +
indicates an excess of first product eluted and − indicates that the second
product was in excess. d Diethyl 2,2-diallylmalonate (22) was isolated (9%).
e Diethyl 2,2-diallylmalonate (22) was isolated (16%).
2448 | Dalton Trans., 2008, 2442–2453
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
1
J = 7.6 Hz, 18H), 1.33 (m, 3H), 1.01 (m, 27H); 31P{ H} NMR: d
and dried under vacuum. The first band gave a yellow–orange
solid identified as [Co3(l3-CCO2Et)(CO)8(PEt3)] (1) (47 mg,
40%). (Anal. Calc. for C18H20O10PCo3: C, 35.79 H, 3.34. Found:
C, 35.93; H, 3.38%); IR cyclohexane, mCO/cm−1: 2083 s, 2056 s,
30.74 (s). MS (FAB+, m/z): 784 (M+) and peaks for [M − nCO]+
(n = 1–6).
[Co3(l3-CH)(CO)6(PMe2Ph)3] (6). To
a dichloromethane
1
2040 vs, 2029 vs, 2018 s, 1994 w; H NMR: d 4.28 (q, J = 14.1,
(15 ml) solution of [Co3(l3-CH)(CO)9] (70 mg, 0.158 mmol),
PMe2Ph (90 mg, 0.652 mmol) was added and the reaction mixture
was stirred at room temperature for 2 h. The solvent was removed
under vacuum. The compound was purified by recrystallization
at −20 ◦C in hexane and afforded orange crystals identified as
[Co3(l3-CH)(CO)6(PMe2Ph)3] (6) (92 mg, 75%) (Anal. Calc. for
C31H34O6P3Co3: C, 48.21; H, 4.44, P, 12.03. Found: C, 48.02; H,
4.56, P, 11.53%); IR cyclohexane, mCO/cm−1: 2002 s, 1968 vs, 1808
vs; 1H NMR: d 10.67 (s, 1H), 7.72 (m, 6H), 7.47 (m, 9H), 1.71 (d,
7.2 Hz, 2H), 1.79 (dq, J = 15.9, 7.8 Hz, 6H) 1.31 (t, J = 7.2,
1
3H), 1.20 (dt, J = 7.2, 3.6 Hz, 9H); 31P{ H} NMR: d 39.95
(br s); m/z (FAB): 604 (M+), and peaks for [M − nCO]+ (n =
1–8), and the second band afforded a yellow solid identified as
[Co3(l3-CCO2Et)(CO)7(PEt3)2] (3) (15 mg, 11%) (Anal. Calc. for
C23H35O9P2Co3: C, 39.79; H, 5.08; P, 11. 85. Found: C, 40.33;
H, 5.38%); IR cyclohexane, mCO/cm−1: 2054 s, 2004 vs, 1998 vs,
1
1981 s, 1963 m, 1854 w, 1825 w; H NMR: d 4.19 (q, J = 14.4,
7.2 Hz, 2H), 1.78 (dq, J = 15.3, 7.5 Hz, 12H), 1.28 (t, J = 7.2 Hz,
18H); 31P{ H} NMR: d 14.99 (s); m/z (FAB): 772 (M+) and peaks
1
1
3H), 1.10 (dt, J = 15.3, 7.5 Hz, 18H); 31P{ H} NMR: d 52.3 (s);
for [M − nCO]+ (n = 1–6).
m/z (FAB): 694 (M+–1) and peaks for [M − nCO]+ (n = 1–7).
[Co3(l3-CH)(CO)8{(S)-NMDPP}] (7). To a toluene solu-
tion (10 mL) of [Co3(l3-CH)(CO)9] (100 mg, 0.226 mmol),
(S)-NMDPP [= ((1S)-2-isopropyl-5-methylcyclohexyl)diphenyl-
phosphine] (73 mg, 0.226 mmol) was added and the reaction
was stirred at 90 ◦C for 4 h. The solvent was removed under
vacuum and the residue was purified using preparative TLC. Be-
sides a stationary band (decomposed product(s)), elution with
hexane–dichloromethane (9 : 1 v/v) gave three bands. The three
products were removed from the TLC plates, extracted with
dichloromethane and dried under vacuum. The first product was
unconsumed [Co3(l3-CH)(CO)9] (15 mg). The second product
was only obtained in trace amounts and has not been character-
ized. The third product was identified as [Co3(l3-CH)(CO)7{(S)-
NMDPP}] (7) (85 mg, 51%), (Anal. Calc. for C31H30O8PCo3:
C, 50.43; H, 4.10. Found: C, 51.03; H, 4.23%); IR cyclohexane,
[Co3(l3-CH)(CO)8(PMe2Ph)] (2) and [Co3(l3-CH)-(CO)7-
(PMe2Ph)2] (4). To a dichloromethane (15 ml) solution of
[Co3(l3-CH)(CO)9] (100 mg, 0.226 mmol), PMe2Ph (31 mg,
0.226 mmol) was added and the reaction mixture was stirred
at room temperature for 3 h. The solvent was removed by
vacuum and the solid residue was dissolved in a small quantity of
dichloromethane and purified by preparative TLC. Elution with
hexane–CH2Cl2 (4 : 1 v/v) gave three bands, where the first band
was unconsumed parent cluster. The second band gave a brown–
black solid identified as [Co3(l3-CH)(CO)8(PMe2Ph)] (2) (57 mg,
46%) (Anal. Calc. for C17H12O8PCo3: C, 36.99; H, 2.19. Found:
C, 36.84; H, 2.21%); IR cyclohexane, mCO/cm−1: 2080 s, 2033 vs,
2025 vs, 2012 s, 1983 w, 1972 w, 1877 w, 1860 w; (solid state):
1
2074 s, 2033 m, 2019 s, 2004 vs, 1976 w, 1965 s; H NMR: d
11.01 (s, 1H), 7.36 (m, 2H), 7.23 (s, 3H), 1.21 (d, 3H); (at 203
1
mCO/cm−1: 2078 s, 2041 vs, 2015 vs, 1980 sh, 1952 w, 1858 m; H
1
K): d 10.60 (br s, 1H), 7.09 (m, 5H), 2.03 (s, 6H); 31P{ H} NMR:
d 16.14 (s, br) and (at 203 K) 18.70 (s); m/z (FAB): 552 (M+)
and peaks for [M − nCO]+ (n = 1–8). The third band afforded a
yellow–brown solid identified as [Co3(l3-CH)(CO)7(PMe2Ph)2] (4)
(25 mg, 17%) (Anal. Calc. for C24H23O7P2Co3: C, 43.53; H, 3.50.
Found: C, 43.63; H, 3.41%); IR cyclohexane, mCO/cm−1: 2046 s,
NMR: d 11.02 (s, 1H), 7.50 (m, 10H), 3.07 (t, J = 15.0 Hz, 1H),
2.15 (m, 2H), 1.75 (m, 4H), 1.60 (m, 1H), 1.22 (d, J = 15.0 Hz,
1H), 0.90 (q, J = 12.0, 6.0 Hz, 9H), 0.27 (d, J = 6.0 Hz, 1H);
1
31P{ H} NMR: d 60.0 (s); m/z (FAB): 738 (M+) and peaks for
[M − nCO]+ (n = 1–8).
1
1994 vs, 1988 sh, 1851 m, 1824 m; H NMR: d 11.03 (s), 7.75
[{Co3(l3-CH)(CO)8}2(l-dppe)] (8) and [Co3(l3-CH)(CO)7(l-
dppe)] (10). To a toluene (15 ml) solution of 100 mg (0.226 mmol)
[Co3(l3-CH)(CO)9], 90 mg (0.226 mmol) of dppe [= 1,2-
bis(diphenylphosphino)ethane] was added and the reaction mix-
ture was stirred at room temperature for 3 h. The solvent
was removed under vacuum and the solid residue was purified
using preparative TLC. Elution with a mixture of hexane–
dichloromethane (7 : 3 v/v) gave three bands, where the band
with highest Rf was identified as the unconsumed parent tricobalt
alkylidyne cluster. The second band gave [{Co3(l3-CH)(CO)8}2(l-
dppe)] (8) (69 mg, 25%). Anal. Calc. for C44H26O16P2Co6: C, 43.10;
H, 2.14. Found: C, 43.07; H, 2.15%); IR cyclohexane, mCO/cm−1:
1
(m, 4H), 7.52 (m, 6H), 1.75 (d, 12H); 31P{ H} NMR: d 34.28
(s); m/z (FAB): 662 (M+) and peaks for [M − nCO]+ (n = 1–7).
Single crystals of [Co3(l3-CH)(CO)8(PMe2Ph)] (2) suitable for X-
ray structure analysis were grown by slow evaporation at −20 ◦C
from a hexane–dichloromethane mixture.
[Co3(l3-CCO2Et)(CO)6(PEt3)3] (5). To a solution of [Co3(l3-
CCO2Et)(CO)9] (100 mg, 0.194 mmol) in 15 mL of
dichloromethane, an excess of triethylphosphine (71 mg,
0.6 mmol) was added. The reaction mixture was stirred at
room temperature for 18 h. The solvent was removed under
reduced pressure, the solid obtained was dissolved in a small
quantity of dichloromethane and the products were separated
using preparative TLC (eluent: dichloromethane–hexane 3 : 2
v/v). Except for unconsumed starting materials, one product was
removed from the TLC plates and extracted with dichloromethane
and dried under vacuum. The orange solid obtained was identified
as [Co3(l3-CCO2Et)(CO)6{PEt3}3] (5) (76 mg, 50%) (Anal. Calc.
for C28H50O8P3Co3: C, 42.87; H, 6.42; P, 11. 85. Found: C, 43.85;
H, 6.48; P, 11.47%); IR cyclohexane, mCO/cm−1: 2002 m, 1979 vs,
1966 s, 1946 w, 1814 s, 1803 w; 1H NMR: d 3.73 (m, 2H), 1.70 (t,
1
2080 s, 2037 s, 2024 vs, 2014 m, 1991 w, 1970 w, 1880 m; H
1
NMR: d 7.33 (m, 20H), 2.19 (br s, 4H); 31P{ H} NMR: d 46.02
(s); m/z (FAB): 1226 (M+) and peaks for [M − nCO]+ (n = 1–16).
The third band afforded [Co3(l3-CH)(CO)7(l-dppe)] (10) (90 mg,
51%) (Anal. Calc. for C34H25O7P2Co3: C, 52.07; H, 3.21. Found:
C, 51.79; H, 3.34%); IR cyclohexane, mCO/cm−1: 2051 vs, 2024 vs,
2000 s, 1981 m, 1973 w, 1941 m; 1H NMR: d 7.46 (m, 20H), 2.11
1
(m, 4H); 31P{ H} NMR: d 53.57 (s); m/z (FAB): 784 (M+) and
peaks for [M − nCO]+ (n = 1–7).
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2449
©

View Article Online
[{Co3(l3-CH)(CO)8}2(l-dppm)] (9) and [Co3(l3-CH)(CO)7(l-
dppm)] (11). To a toluene (15 ml) solution of [Co3(l3-CH)(CO)9]
(100 mg, 0.226 mmol), dppm (87 mg, 0.226 mmol) was added
and the reaction mixture was stirred at room temperature for 3 h.
The solvent was evaporated under vacuum and the solid residue
was dissolved in a small quantity of dichloromethane and purified
on preparative TLC plates. Two products, one yellow and the
other dark brown, were isolated from the TLC plates, extracted
with dichloromethane and dried under vacuum. The yellow solid
was identified as [{Co3(l3-CH)(CO)8}2(l-dppm)] (9) (15 mg, 4%),
(Anal. Calc. for C43H24O16P2Co6: C, 42.61; H, 2.00. Found: C,
43.17; H, 2.19%); IR cyclohexane, mCO/cm−1: 2075 s, 2036 s, 2008
vs, 1992 m, 1972 w; 1H NMR: d 7.25–7.43 (m, 20H), 2.86 (m, 2H);
black solid was obtained and identified as [Co3(l3-CH)(CO)7{(R)-
BINAP}] (29 mg, 51%) (13); IR hexane, mCO/cm−1: 2058 vs, 2010 s,
1
2005 vs, 1990 vw, 1973 w, 1966 w, 1950 vw, 1755 m, 1750 m; H
1
NMR: d 12.45 (s, 1H), 8.79–6.11 (m, 32H); 31P{ H} NMR: 59.8
(s), 46.2 (s); m/z (FAB): 983 (M+ − CO) and peaks for [M − nCO]+
(n = 2–7).
[Co3(l3-CH)(CO)7{(S,S)-CHIRAPHOS}] (14). A hexane so-
lution (5 ml) of [Co3(l3-CH)(CO)9] (10 mg, 22.6 lmol) and (S,S)-
CHIRAPHOS [= (2S,3S)-(−)-bis(diphenylphosphino)butane]
(10 mg, 23.5 lmol) was heated slowly to reflux under vigorous
stirring. The reaction was monitored by spot TLC (eluent: hexane–
CH2Cl2 2 : 1 v/v). When all starting cluster was consumed, the
reaction mixture was transferred to a Pasteur pipette prepared
with glass wool and a few centimetres of silica on top. Traces
of the parent cluster were eluted with pure hexane and then
the green product was eluted using hexane–CH2Cl2 (2 : 1 v/v).
The isolated green product was concentrated and dried under
vacuum, yielding a dark green product, [Co3(l3-CH)(CO)7(S,S)-
CHIRAPHOS] (14) (18 mg, 98%), IR hexane, mCO/cm−1: 2059 vs,
1
31P{ H} NMR: insufficient sample. Mass spectrum (m/z): 1212
(M+) and peaks for [M − nCO]+ (n = 1–6). The dark brown product
was identified as [Co3(l3-CH)(CO)7(l-dppm)] (11) (117 mg, 67%),
(Anal. Calc. for C33H23O7P2Co3: C, 51.46; H, 3.01. Found: C,
51.34; H, 3.10%); IR cyclohexane, mCO/cm−1: 2061 s, 2008 vs, 1992
w, 1970 w, 1958 vw; 1H NMR: d 12.10 (s, 1H), 7.74–7.14 (m, 20H),
4.04 (q, J = 20.0, 12.0 Hz, 1H), 3.40 (q, J = 24.0, 12.0 Hz, 1H);
1
2016 vs, 2002 vs, 1986 m, 1973 w, 1946 vw; H NMR: d10.86 (s,
1
31P{ H} NMR: d 42.22 (s). m/z (FAB): 770 (M+) and peaks for
1H), 7.78–7.12 (m, 20H), 2.30 (s, 2H), 0.82 (s, 6H); 31P{H} NMR:
77.5 (s); m/z (FAB): 812 (M+) and peaks for [M − nCO]+ (n =
1–7).
[M − nCO]+ (n = 1–7). An alternative synthetic procedure for the
preparation of 11 has been published by M. I. Bruce et al.13
[Co3(l3-CH)(CO)7{l-1,2-(R,R)-DIPAMP}] (12). To a tolu-
ene solution (10 mL) of [Co3(l3-CH)(CO)9] (30 mg,
67 lmol), DIPAMP [= (1R,2R)-bis(2-methoxyphenylphenylphos-
phino)ethane] (31 mg, 67 lmol) was added and the reaction
mixture was stirred at room temperature for 2.5 h. The solvent
was removed under vacuum and the black solid residue was
purified by preparative TLC. Besides a stationary brown band
(decomposed material(s)), elution with hexane-dichloromethane
(3 : 2 v/v) gave a single band, which was removed from the
TLC plates and extracted with dichloromethane. The solvent was
removed under vacuum, resulting in a black solid identified as
[Co3(l3-CH)(CO)7{(R,R)-DIPAMP}] (12) (40 mg, 70%), (Anal.
Calc. for C36H29O9P2Co3: C, 51.21; H, 3.46. Found: C, 51.11; H,
3.59%); IR hexane, mCO/cm−1: 2055 s, 2001 vs, 1992 sh, 1985 w,
[Co3(l3-CH)(CO)7{(R,R)-NORPHOS}] (15). [Co3(l3-CH)-
CO)9] (10 mg, 23 lmol) and (R,R)-NORPHOS [= (2R,3R)-
(−)-2,3-bis(diphenylphosphino)bicyclo[2.2.1]hept-5-ene] (11 mg,
24 lmol) were dissolved in hexane (10 mL). The reaction mixture
was heated gently to 40 ◦C. The reaction was monitored by
spot TLC (hexane–dichloromethane 1 : 1 v/v) and after 40 min,
when all ligand was consumed, the solvent was evaporated under
reduced pressure. The resultant green black-residue was dissolved
in a small quantity of dichloromethane and purified using a short
column packed with silica 60 (5 mm × 50 mm). The column
was eluted with hexane to remove unconsumed starting material
leaving a green–black solid, which eluted with a mixture of
hexane–dichloromethane (1 : 1 v/v). The green-black fraction
was collected and concentrated using rotary evaporation to
afford a green–black solid identified as [Co3(l3-CH)(CO)7{(R,R)-
NORPHOS}] (15) (11 mg, 55%). IR hexane, mCO/cm−1: 2058 vs,
1
1969 m; H NMR: d 7.38 (m, 10H), 6.93 (m, 8H), 3.86 (br, 3H),
1
3.41 (s br, 3H), 2.39 (br, 2H), 2.16 (br, 2H); 31P{ H} NMR: d 51.3
(s); m/z (FAB): 844 (M+) and peaks for [M − nCO]+ (n = 1–7).
Single crystals suitable for X-ray analysis were grown by slow
evaporation from a hexane-dichloromethane mixture at room
temperature.
1
2013 s, 2003 vs, 1990 w, 1970 w, 1936 vw; H NMR: d 10.91 (s,
1H), 8.07 7.11 (m, 20H), 5.96 (s, 1H), 4.15 (s, 1H), 3.06–1.33 (m,
1
6H); 31P{ H} NMR: d 60.8 (s), 51.3 (s); m/z (FAB): 821 (M+–CO)
and peaks for [M − nCO]+ (n = 2–7).
[Co3(l3-CH)(CO)7{(R)-BINAP}] (13). In a two-necked round
bottom flask fitted with a condenser [Co3(l3-CH)(CO)9] (25 mg,
57 lmol) and (R)-BINAP [= (R)-(+)-2,2ꢀ-bis(diphenylphosphino)-
1,1ꢀ-binapthyl] (35 mg, 56 lmol) were dissolved in CH2Cl2 (12 ml),
heated gently and maintained at reflux for 22 h. During the course
of reaction, the solution changed colour from deep purple to
black–brown. The solvent was removed using rotary evaporation.
The black–brown solid obtained was dissolved in approx. 2 ml
of CH2Cl2 and purified using preparative TLC (Merck Kieselgel
60, 0.5 mm thickness; eluent: hexane–CH2Cl2 3 : 1 v/v). Except
for a brown stationary band (probably decomposed material) and
traces of unconsumed [Co3(l-CH)(CO)9], three bands were eluted.
The green–black main product was scraped of the TLC plate and
dissolved in CH2Cl2. The silica was filtered off and the product
solution was concentrated and dried under vacuum. A green
[Co3(l3-CH)(CO)7{(R)-PROPHOS}] (16). In
a two-neck
round-bottom flask fitted with a condenser, [Co3(l3-CH)(CO)9]
(10 mg, 23 lmol) and (R)-PROPHOS [= (R)-(+)-1,2-
bis(diphenylphosphino)propane] (9 mg, 22 lmol) were dissolved
in hexane (12 ml) and stirred at room temperature (∼23 ◦C)
for 10 min. The reaction mixture was then heated at 40 ◦C,
whereupon the colour of the solution gradually changed from
deep purple to black–brown. After 1 h the solvent was removed
using rotary evaporation. The black–brown solid obtained was
purified using a short column packed with silica Kieselgel 60
(5 mm × 50 mm). The column was eluted with hexane to remove
any unconsumed starting material, leaving a dark green solid
which was eluted with a mixture of hexane–CH2Cl2 (1 : 1 v/v).
The green fraction was collected and concentrated using rotary
2450 | Dalton Trans., 2008, 2442–2453
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
evaporation, yielding a green–black solid identified as [Co3(l3-
CH)(CO)7{(R)-PROPHOS}] (16) (10.4 mg, 59%) (Anal. Calc. for
C35H27O7P2Co3: C, 52.66; H, 3.41. Found: C, 52.04; H, 3.67%);
IR hexane, mCO/cm−1: 2059 s, 2011 s, 2003 vs, 1991 w, 1971 m;
1H NMR: d 7.73–6.81 (m, 20H), 2.34 (s, 3H), 2.09 (m, 2H), 1.67
with 10 bar of CO. The reaction was continuously stirred with
a magnetic stirrer and heated at 120 ◦C for 10 h. The reaction
vessel was allowed to cool down to room temperature before it was
carefully opened. The reaction mixture was transferred to a round-
bottom flask and the solvent was removed under vacuum. The oily
residue was purified using preparative TLC, eluting with a hexane-
ethyl acetate mixture (4:1 v/v). Except for starting materials,
two products were removed from the TLC plates, extracted with
dichloromethane and dried under vacuum. The first product was
a white solid identified as 1,3,5-triphenylbenzene and the second
product was identified as 4-phenyltricyclo[5.2.1.02,6]dec-4-en-3-
1
(m, 1H), 1.26 (s, 3H), 0.96 (m, 2H), 0.85 (m, 1H); 31P{ H} NMR:
d 62.76 (s), 53.41 (s); m/z (FAB): 798 (M+) and peaks for [M −
nCO]+ (n = 1–7).
[Co3(l3-CH)(CO)7{(R,R)-Me-DUPHOS}] (17). To a hexane
solution (8 ml) of [Co3(l3-CH)(CO)9] (40 mg, 91 lmol)
was added (R,R)-Me-DUPHOS [= (−)-1,2-bis((2R,5R)-2,5-
dimethylphospholano)benzene] (27 mg, 88 lmol) and the reaction
mixture was heated slowly to 40 ◦C under vigorous stirring. After
30 min the reaction mixture was transferred to a Pasteur pipette,
prepared with glass wool and a few centimetres of silica on top.
Traces of the parent cluster were eluted with pure hexane and then
an orange compound was eluted using hexane–CH2Cl2 (1 : 1 v/v).
The isolated orange product was concentrated and dried under
vacuum, yielding a bright red product, [Co3(l3-CH)(CO)7{(R,R)-
Me-DUPHOS}] (17) (43 mg, 70%). IR hexane, mCO/cm−1: 2056 s,
2024 s, 1994 s, 1883 vs, 1863 m, 1819 m; 1H NMR: d 7.83–7.52 (m,
4H), 6.35 (s, 1H), 2.90 (s, 2H), 2.37 (m, 4H), 2.15 (m, 2H), 1.8–1.38
(m, 4H), 1.14 (dd, 1H, J = 6.7 Hz, J = 19.0 Hz), 0.65 (dd, 1H,
1
one (19), H NMR: d 7.68 (d, J = 7.5 Hz, 2H), 7.63 (d, J =
3.0 Hz, 1H), 7.36 (t, J = 7.5 Hz, 2H), 7.31 (d, J = 7.5 Hz, 1H),
2.69 (t, J = 3.75 Hz, 1H), 2.49 (d, J = 4.0 Hz, 1H), 2.36 (d, J =
5.0 Hz, 1H), 2.27 (d, J = 4.5 Hz, 1H), 1.69 (m, 1H), 1.61 (m, 1H),
1.33 (m, 2H), 1.11 (d, J = 10.5 Hz, 1H), 1.0 (d, J = 10.5 Hz, 1H).
Intramolecular Pauson–Khand reaction. In a typical experi-
ment, the autoclave was loaded with catalyst (14 mg) and diethyl
allyl propargyl malonate (20), in a 50-fold molar excess. Toluene
(5 mL) was added and the autoclave was purged several times
before being pressurized with 10 bar of CO. The reaction was
continuously stirred with a magnetic stirrer and heated at 120 ◦C
for 10 h. The reaction vessel was allowed to cool down to
room temperature before carefully opened. The reaction mixture
was transferred to a round-bottomed flask and the solvent
was removed under vacuum. The oily residue was purified on
preparative TLC, eluting with a hexane–ethyl acetate mixture (3 :
2 v/v). Except for starting material, two products were removed
from the TLC plates, extracted with dichloromethane and dried
under vacuum. The two transparent oils obtained were identified
1
J = 6.5 Hz, J = 14.1 Hz); 31P{ H} NMR: d 93.7; m/z (FAB): 692
(M+) and peaks for [M − nCO]+ (n = 1–7). Single crystals suitable
for crystallographic X-ray structure determination were grown by
slow evaporation at room temperature from a mixture of hexane
and CH2Cl2.
[Co3(l3-CH)(CO)7{(R)-(S)-Josiphos3}] (18). In a 25-ml two-
necked round-bottom flask fitted with a condenser and a drop-
ping funnel, [Co3(l3-CH)(CO)9] (16 mg, 36 lmol) was dissolved
in CH2Cl2 (5 ml) and heated to reflux. The dropping funnel
was charged with (R)-(S)-Josiphos3 [= (R)-(−)-1-[(S)-2-
(dicyclohexylphosphino)ferrocenyl]ethyldicyclohexylphosphine]
(20 mg, 33 lmol) dissolved in CH2Cl2 (5 ml) and the solution
was added dropwise to the refluxing cluster solution. After
5 min., the solution turned deep red–black and spot TLC (eluent:
hexane–CH2Cl2 1 : 1 v/v) showed one new orange–red product.
The solvent was evaporated and the resultant solid residue
was dissolved in a minimum amount of dichloromethane and
purified through a short plug of silica, eluting initially with
hexane, followed by hexane–dichloromethane 1 : 1 (v/v), and
finally CH2Cl2. One red fraction was collected and concentrated
under vacuum. The red solid residue obtained was identified as
[Co3(l3-CH)(CO)7{(R)-(S)-Josiphos3}] (18) (27 mg, 83%). IR
hexane, mCO/cm−1: 2050 s, 2018 vs, 1998 m, 1989 vs, 1982 m, 1956
m, 1924 w, 1879 w, 1854 m, 1818 m, 1755 m; 1H NMR: d 9.75 (s,
1
as 7,7-bis(ethoxycarbonyl)bicyclo[3.3.0]-1-octene-3-one (21), H
NMR: d 5.91 (s, 1H), 4.21 (dq, J = 14.0, 7.0 Hz, 4H), 3.33 (d,
J = 19.0 Hz, 1H), 3.23 (d, J = 19.0 Hz, 1H), 3.09 (ddd, J = 2.5,
3.5, 5.0 Hz, 1H), 2.78 (dd, J = 7.5, 13.0 Hz, 1H), 2.62 (dd, J =
6.5, 18.0 Hz, 1H), 2.13 (dd, J = 3.0, 17.5 Hz, 1H), 1.71 (t, J =
13.0 Hz, 1H), 1.25 (dt, J = 7.0, 14.0 Hz, 6H), and diethyl 2,2-
diallylmalonate (22), 1H NMR: d 5.64 (m, 2H), 5.09 (dd, J = 8.0,
2.0 Hz, 4H), 4.17 (q, J = 7.0 Hz, 4H), 2.62 (d, J = 7.5 Hz, 4H),
1.23 (t, J = 7.0 Hz, 6H).
HPLC analysis. The enantiomeric excess of the product, 7,7-
bis(ethoxycarbonyl)bicyclo[3.3.0]-1-octene-3-one (21), obtained
in catalytic experiments was assessed by HPLC analysis using a
Varian Prostar chromatograph (PDA detector) with a Chiralpak
AD-H analytical column. The eluent was n-hexane–isopropanol
(9 : 1 v/v) and the flow rate was 0.5 mL min−1. The retention times
for the first and second enantiomers were 16.9 and 17.8 min,
respectively.
1
0.32 H), 4.25 (m, 8H), 2.79–0.91 (m, 44H); 31P{ H} NMR: d 32.7
(s), 37.2 (s) 39.9 (s), 54.2 (s), 61.2 (s), 69.8 (s); m/z (FAB) 993 (M+
+1) and peaks for [M − nCO]+ (n = 1–7).
X-Ray crystallography
The diffraction data were collected at 293 K for clusters 2 and
6 using a Bruker SMART APEX diffractometer. The Bruker
SMART and SAINT programs36 were used for cell refinement
and data reduction. The structure solutions were carried out using
SHELXS 97,37 and SHELXL38 was used for structure refinements.
Diffraction data were collected at 294 K for clusters 12 and
17 on a Nonius CAD-4 diffractometer. CAD4 v 5.039 were used
Catalysis experiments
Intermolecular Pauson–Khand reactions. In a typical experi-
ment, the autoclave was loaded with catalyst (14 mg) and the
substrates norbornene and phenyl acetylene, each in a 50 molar
excess relative to the catalyst. Toluene (5 mL) was added and
the autoclave was purged several times before being pressurized
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2451
©

View Article Online
Table 4 Crystallographic data for [Co3(l3-CH)(CO)8(PMe2Ph)] (2), [Co3(l3-CH)(CO)6(PMe2Ph)3] (6), [Co3(l3-CH)(CO)7(l-1,2-DIPAMP)] (12) and
[Co3(l3-CH)(CO)7(1,1-DUPHOS)] (17)
2
6
12
17
Empirical formula
C17H12Co3O8P
552.03
293(2)
0.71073
Monoclinic
P21
8.7435(7)
14.8061(12)
9.0608(8)
1053.31(15)
2
C31H34Co3O6P3
772.28
293(2)
0.71073
Triclinic
C36H29Co3P2O9
844.4
294
C26H29Co3O7P2
692.3
294
Mr
T/K
˚
k/A
0.71073
Orthorhombic
P212121
11.968(2)
12.618(2)
24.770(5)
3741(1)
4
0.71073
Orthorhombic
P212121
9.803(1)
16.155(2)
18.477(2)
2926.2(6)
4
Crystal system
Space group
P1
˚
a/A
10.1219(18)
10.4665(19)
10.5068(19)
859.3(3)
˚
b/A
˚
c/A
3
˚
V/A
Z
1
Dc/g/cm3
1.741
2.450
1.492
1.610
1.50
1.450
1.57
1.829
l/mm−1
Crystal size/mm
h range
0.40 × 0.20 × 0.15
−7 to 11
−19 to 18
−12 to 8
6738
4622
4622/1/262
0.872
0.37 × 0.25 × 0.06
−13 to 13
−13 to 13
−13 to 13
7355
6667
6667/3/394
0.985
0.36 × 0.24 × 0.17
0.25 × 0.18 × 0.12
0 to 14
0 to 11
k range
0 to 15
0 to 29
3673
3673
2687/0/324
1.24
0 to 19
0 to 22
2908
2907
2520/0/343
1.09
l range
Reflections collected
Independent reflections
Data/restraints/parameters
Goodness-of-fit on F2
R1
0.0272
0.0440
0.05
0.028
wR2
0.0696a
0.0802a
0.066b
0.033b
Weighting scheme
x = 0.0531, y = 0.0000
0.345/−0.189
0.282(13)
x = 0.0150, y = 0.0000
0.538/−0.357
0.015(15)
x = 0.0004
1.18/−0.69
0.00(4)
x = 0.0004
0.34/−035
0.00(2)
−3
˚
Dqmax/min/e A
Abs. struct. parameter
a w = 1/[r2(Fo2) + (xP)2 + yP], P = (Fo + 2Fc2)/3. b w = 1/[r2(Fo2) + xF2].
2
for cell refinement and data reduction. The structure solution was
carried out using SIR9240 and RAELS41 was used for structure
refinements. The absolute structures for clusters 2, 6, 12 and 17
were determined on the basis of the Flack index42 during the
refinement process.
Relevant crystal, data collection and refinement data for
the X-ray crystal structures of Co3(l3-CH)(CO)8(PMe2Ph)]
(2), [Co3(l3-CH)(CO)6(PMe2Ph)3] (6), [Co3(l3-CH)(CO)7(l-1,2-
DIPAMP)] (12) and [Co3(l3-CH)(CO)7(1,1-DUPHOS)] (17) are
summarised in Table 4.
2 T. Sugihara, M. Yamaguchi and M. Nishizawa, Chem.–Eur. J., 2001, 7,
1589.
3 N. E. Schore and M. C. Crondacy, J. Org. Chem., 1981, 46, 5436–
5438.
4 (a) S. Shambayati, W. E. Crowe and S. L. Schreiber, Tetrahedron Lett.,
1990, 31, 5289; (b) N. Jeong, Y. K. Chung, B. Y. Lee, S. H. Lee and
S.-E. Yoo, Synlett, 1991, 204.
5 (a) R. J. Dennenberg and D. J. Darensbourg, Inorg. Chem., 1972, 11,
72; (b) J. D. Atwood and T. L. Brown, J. Am. Chem. Soc., 1976, 98,
3160; (c) D. J. Darensbourg and R. L. Kump, Inorg. Chem., 1978, 17,
2680.
6 R. B. Grossman and S. L. Buchwald, J. Org. Chem., 1992, 57, 5803.
7 B. M. Trost and A. B. Pinkerton, Org. Lett., 2000, 11, 1603.
8 I. U. Khand, G. R. Knox, P. L. Pauson, W. E. Watts and M. I. Foreman,
J. Chem. Soc., Perkin Trans. 1, 1973, 977.
9 T. Sugihara and M. Yamaguchi, J. Am. Chem. Soc., 1998, 120, 10782.
10 P. J. Dyson, Chem. Rev., 2004, 248, 2443.
CCDC reference numbers 668301–668304.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b717698h
11 K. Hiroi, T. Watanabe, R. Kawagishi and I. Abe, Tetrahedron:
Asymmetry, 2000, 11, 797.
12 (a) X. Verdaguer, A. Moyano, M. A. Pericas, A. Riera, V. Bernardes,
A. E. Greene, A. Alvarez-Larena and J. F. Piniella, J. Am. Chem.
Soc., 1994, 116, 2153–2154; (b) X. Verdaguer, J. Vazquez, G. Fuster, V.
Bernardes-Genisson, A. E. Greene, A. Moyano, M. A. Pericas and A.
Riera, J. Org. Chem., 1998, 63, 7037–7052; (c) D. Konya, F. Robert,
Y. Gimbert and A. E. Greene, Tetrahedron Lett., 2004, 45, 6975–
6978.
13 M. I. Bruce, K. A. Kramarczuk, G. J. Perkins, B. W. Skelton, A. H.
White and N. N. Zaitseva, J. Cluster Sci., 2004, 15, 119.
14 G. A. Acum, M. J. Mays, P. R. Raithby and G. A. Solan, J. Organomet.
Chem., 1996, 508, 137.
15 W. H. Watson, S. G. Bodige, K. Ejsmont, J. Liu and M. G. Richmond,
J. Organomet. Chem., 2006, 691, 3609, and references therein.
16 V. Moberg, R. Duquesne and E. Nordlander, unpublished results.
17 S. P. Tunik, I. O. Koshevoy, A. J. Poe, D. H. Farrar, E. Nordlander,
T. A. Pakkanen and M. Haukka, Dalton Trans., 2003, 2457.
18 P. W. Sutton and L. F. Dahl, J. Am. Chem. Soc., 1967, 89, 261.
19 M. Ahlgre´n, T. T. Pakkanen and I. Tahvanainen, J. Organomet. Chem.,
1987, 323, 91.
Acknowledgements
This research has been sponsored by The Royal Swedish Academy
of Sciences, the Crafoord Foundation and the Swedish Research
Council (VR). A. J. Deeming thanks the Wenner-Gren Foundation
for a visiting professorship at Lund University. V. Moberg thanks
the Solander program for a travel award that enabled him to carry
out research at the University of New South Wales. We thank
Cecilia Olsson for helpful assistance with the HPLC analysis of
the potentially enantioselective Pauson–Khand reactions.
References
1 I. U. Khand, G. R. Knox, P. L. Pauson and W. E. Watts, J. Chem. Soc.
D, 1971, 36.
2452 | Dalton Trans., 2008, 2442–2453
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
20 D. C. Miller, R. C. Gearhart and T. B. Brill, J. Organomet. Chem., 1979,
169, 395.
21 J. D. King, M. Monari and E. Nordlander, J. Organomet. Chem., 1999,
573, 272.
22 S. J. Maginn, Acta Crystallogr., Sect. C, 1990, 46, 2339.
23 M. D. Brice, B. R. Penfold, W. T. Robinson and S. R. Tylor, Inorg.
Chem., 1970, 9, 362.
24 W. H. Watson, S. G. Bodige, K. Ejsmont, J. Liu and M. G. Richmond,
J. Organomet. Chem., 2006, 691, 3609.
25 L. Pauling, Proc. Natl. Acad. Sci. USA, 1976, 73, 4290.
26 (a) R. G. Cunningham, L. R. Hanton, S. D. Jensen, B. H. Robinson
and J. Simpson, Organometallics, 1987, 6, 1470; (b) R. P. Aggarwal,
N. G. Connelly, M. C. Crespo, B. J. Dunne, P. M. Hopkins and A. G.
Orpen, J. Chem. Soc., Dalton Trans., 1992, 655.
27 S. G. Bott, H. Shen and M. G. Richmond, Struct. Chem., 2001, 12, 225.
28 G. R. Desiraju, J. Chem. Soc., Dalton Trans., 2000, 3745.
29 (a) D. Braga, F. Grepioni, K. Biradha, V. R. Pedireddi and G. R.
Desiraju, J. Am. Chem. Soc., 1995, 117, 3156; (b) D. Braga and F.
Grepioni, Acc. Chem. Res., 1997, 30, 81.
30 A. J. Downard, B. H. Robinson and J. Simpson, Organometallics, 1986,
5, 1122.
31 V. Moberg, P. Homanen, S. Selva, R. Persson, M. Haukka, T. A.
Pakkanen, M. Monari and E. Nordlander, Dalton Trans., 2006, 279.
32 P. Homanen, R. Persson, M. Haukka, T. A. Pakkanen and E.
Nordlander, Organometallics, 2000, 19, 5568.
33 T. M. Schmid and G. Consiglio, Tetrahedron: Asymmetry, 2004, 15,
2205.
34 D. Seyferth, J. E. Hallgren and P. L. K. Hung, J. Organomet. Chem.,
1973, 50, 265.
35 (a) K. D. Do¨tz and M. Popall, Tetrahedron, 1985, 41, 5797; (b) A. K.
Pearson and R. A. Dubbert, Organometallics, 1994, 13, 1656.
36 SMART and SAINT Software Reference Manuals, Version 5.051
(Windows NT Version), Bruker Analytical X-ray Instruments Inc.,
Madison, WI, 1998.
37 G. M. Sheldrick, SHELXS97, Program for Crystal Structure Determi-
nation, University of Go¨ttingen, Germany, 1997.
38 G. M. Sheldrick, SHELXL97, Program for Crystal Structure Refine-
ment, University of Go¨ttingen, Germany, 1997.
39 J. D. Schagen, L. Strayer, F. van Meurs and G. Williams, CAD4
Operator’s Guide, Delft Instruments, Delft, Netherlands, 1989.
40 A. Altomare, M. C. Burla, G. Camalli, G. Cascarano, C. Giacovazzo,
A. Gualiardi and G. Polidori, J. Appl. Crystallogr., 1994, 27, 435.
41 A. D. Rae, RAELS96: A Comprehensive Constrained Least-Square Re-
finement Program, Australian National University, Canberra, A. C. T.,
Australia, 1996.
42 H. D. Flack, Acta Crystallogr., Sect. A, 1983, 39, 876.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 2442–2453 | 2453
©