The application of polymer-bound carbonylcobalt(0) species in linker chemistry and catalysis

Article abstract of DOI:10.1039/b303208f

Carbonylcobalt(0) species have been used as linkers between alkynes and a polymer support for the first time. The alkynes may be loaded indirectly onto a phosphine functionalised polymer via their hexacarbonyldicobalt(0) complex, or directly onto a cobalt

Full text of DOI:10.1039/b303208f

View Article Online / Journal Homepage / Table of Contents for this issue
The application of polymer-bound carbonylcobalt(0) species in
linker chemistry and catalysis
Alex C. Comely,a Susan E. Gibson,*a Neil J. Hales,b Craig Johnstoneb and Andrea Stevenazzia
a Department of Chemistry, King’s College London, Strand, London UK WC2R 2LS
b AstraZeneca, Mereside, Alderley Park, Macclesfield, UK SK10 4TG
Received 17th March 2003, Accepted 10th April 2003
First published as an Advance Article on the web 30th April 2003
Carbonylcobalt(0) species have been used as linkers between alkynes and a polymer support for the first time.
The alkynes may be loaded indirectly onto a phosphine functionalised polymer via their hexacarbonyldicobalt(0)
complex, or directly onto a cobalt coated polymer. The alkynes have been released either as alkynes, thus providing
a traceless method of immobilising alkynes, or by reaction with an alkene to generate a cyclopentenone via the
Pauson–Khand reaction. The cobalt coated polymers produced during this study were shown to catalyse the
Pauson–Khand reaction.
Multiple parallel synthesis and combinatorial chemistry are
important methods of generating large numbers of compounds
relatively quickly. Substrate molecules are generally attached to
the polymer via a ‘linker’ molecule using covalent bonds.1–3
Once attached, the immobilised substrate may be isolated
simply by filtration of the polymer from the reaction mixture.
Further organic transformations may then be carried out while
the substrate is bound to the polymer before chemical cleavage
of the final product from the solid support. We recently demon-
strated the potential of π-bound ligands in linker chemistry by
attaching substrates containing aromatic rings to polystyrene-
diphenylphosphine (PSDP) using a chromium carbonyl linker,
chemically manipulating the substrates, and then releasing the
products from the polymer.4,5
The chemistry of carbonyldicobalt(0) alkyne complexes is
well established in the solution phase and has been applied
to great effect in organic synthesis: alkyne protection,6 the
Nicholas reaction,7 and Pauson–Khand annelations,8 for
example, are amply documented. Given the ease of formation
of these complexes and their tolerance to diverse reaction con-
ditions it appeared that it may be possible to use these com-
plexes to attach alkynes to solid supports. (At present, there are
no methods available for attaching an alkyne to a solid support,
manipulating its substituents and then releasing the alkyne.) To
test this hypothesis, we proposed to immobilise an alkyne sub-
strate onto PSDP using a carbonyldicobalt(0) species 1, chem-
ically manipulate a pendant functionality to give 2 and release
the product from the resin to deliver a chemically-modified
alkyne-bearing substrate. This paper describes how we achieved
this goal, how the release strategy was subsequently modified to
include a Pauson–Khand reaction, and how some of the carb-
onylcobalt(0) coated polymers generated during the course of
this study were used as catalysts of the Pauson–Khand reac-
tion. Some of the work on the linker chemistry and the catalyst
discovery has been reported in communication form.9,10
Carbonyl exchange with an external ligand is a facile process
for (η2:η2-alkyne)hexacarbonyldicobalt(0) complexes.11 In
order to confirm this literature precedent prior to attempting
the chemistry on the solid support, and also to obtain readily-
characterisable samples for the purpose of data comparison,
complex 3 was treated with triphenylphosphine. Chromato-
graphic purification gave novel monophosphine complex 4 as a
deep red oil, and purple crystalline novel bisphosphine complex
5 in 42 and 15% yields respectively.
In order to attempt this chemistry on a solid support, PSDP
was swollen in THF under constant nitrogen agitation in a glass
frit-fitted vessel equipped with a condenser. Two equivalents of
alkyne complex 3 were added and after four hours at 50 ЊC the
mixture was filtered and washed with alternate aliquots of THF
and Et2O until the filtrate became colourless. The resulting
deep purple beads 6, significantly less swollen than the original
PSDP beads, were dried in vacuo.
Results and discussion
The ‘indirect’ loading approach
The first approach to loading an alkyne onto a polymer support
involved making a carbonylcobalt(0) derivative of the alkyne
and then attaching the complex thus formed to the polymer. A
commercially available alkyne with a pendant functional group,
hex-5-yn-1-ol was thus treated with octacarbonyldicobalt(0) in
hexane at ambient temperature to give the novel stable hexa-
carbonyldicobalt(0) complex 3 as a deep red oil in good yield.
The nature of the two distinct complexes on the resin, 6a and
6b, was ascertained primarily by comparison of the 31P NMR
and IR spectra of the resin with data obtained in solution for
mono- and bisphosphine complexes 4 and 5. Analysis of the IR
spectrum of the resin indicated the presence of the mono-
phosphine 6a as the major and the bisphosphine 6b as the
minor component in an approximate ratio of 6 : 4. Solution
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 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1959

View Article Online
Table 1 Characterisation of carbonyldicobalt(0) by IR and 31P NMR
ation, and no apparent change in the linker: the mono- and
bisphosphine composition was unchanged from that of 6. Inte-
gration of the 31P NMR spectrum of 7 indicated a reduction of
phosphorus site occupancy to 75% (within the assumed error
spectroscopy
Data from
solution-phase
analogue
margin) and polymer loading was calculated as 0.87
0.02
mmol[hex-5-yn-1-yl acetate]gϪ1. The product hex-5-yn-1-yl
acetate 8 was released by suspending a sample of resin 7,
ground to a fine brown powder, in DCM under white light for
72 h. Filtration of the resulting mixture and concentration of
the colourless filtrate gave 812 as the sole product in 66 2%
yield from 6.
Polymer-bound complex
Data from resin
(4, or 5)
υmax/cmϪ1
2055s
υmax/cmϪ1
2062s
2028w
2008ssh
2004s
1998vs
1979ssh
1990wsh
1967w
δP (ppm)
55.2
δP (ppm)
55.5
υmax/cmϪ1
1998vs
1979ssh
1950s
υmax/cmϪ1
2010s
1962ssh
1950s
1930wsh
δP (ppm)
53.4
δP (ppm)
55.5
calibration studies with complexes 4 and 5 verified this estimate.
31P NMR spectroscopy was extremely valuable in assessing the
composition of the polymer. A relatively well-resolved reson-
ance at 55 ppm integrated to an 80% complexation of phos-
phorus sites, the benign phosphine oxide, polymer-P(O)Ph2,
occupying the remaining 20%. Resolution in the polymer
spectrum was insufficient to distinguish between 6a and 6b,
however, solution-phase analogues 4 and 5 resonate at 55.2 and
53.4 ppm, respectively. In order to estimate the error margin
when working with these resins generated by indirect loading, it
was estimated that 5% inaccuracy accompanied analysis of IR
and 31P NMR spectra (Table 1). Consideration of both IR and
31P NMR spectra and the assumed error margins for both tech-
niques, therefore, allowed substrate loading to be calculated as
0.94
0.02 mmol[hex-5-yn-1-ol]gϪ1. The validity of the IR
calibration study was established from microanalysis of 6: both
cobalt and phosphorus analyses were in excellent accord with
values calculated on the basis of the loading calculation.
A polymer bearing a higher proportion of monophosphine
complex 6a would also be a less crosslinked one and one
more able to swell effectively. Attempts to generate just one
species on the resin by varying the stoichiometry of the reaction
between hexacarbonyl complex 3 and PSDP, however, proved
unsuccessful.
To explore further the stability of the linker to chemical
modification, an oxidation reaction was examined. Of the many
conditions available for oxidation of an alcohol, it was decided
to apply those of a Swern reaction. However, on addition
of an oxalyl chloride–DMSO mixture, cobalt(0) was instantly
stripped from the brown beads to give a deep green solution
indicating formation of liberated cobalt salts. It was suspected
that the HCl evolved during the Swern oxidation might be
responsible: indeed, addition of only a couple of drops of HCl
to a DCM suspension of 6 resulted in the same colour change.
A variety of “modified Swern” conditions exist.13 One set of
conditions that does not evolve HCl during the course of reac-
tion uses a sulfur trioxide–pyridine complex.14 Hence, alcohol 6
was suspended in DCM, and DMSO and triethylamine were
added sequentially, followed by a solution of sulfur trioxide–
pyridine complex. The resulting purple beads 9 were analysed
in the usual way. The IR spectrum displayed a diagnostic
aldehyde stretch at 1723 cmϪ1 and an unchanged mono-/
bisphosphine complex composition from that of 6. Further-
more, phosphorus site occupancy was reduced by only 10%,
thus demonstrating an encouraging stability of the linker to
these oxidising conditions. Hex-5-yn-1-al 1015 was isolated in
17 0.5% yield after decomplexation as before, the low yield
being attributed to the volatility of the aldehyde.
In order to start to probe the stability of the linker to chem-
ical modification, the pendant alcohol of complexes 6 was
acetylated. A suspension of 6 in THF was treated with triethyl-
amine and acetic anhydride and the mixture was left at ambient
temperature for 20 h under constant nitrogen agitation. Fil-
tration, washing and drying in vacuo gave deep purple beads of
acetate 7. Analysis by IR spectroscopy revealed an additional
absorption at 1734 cmϪ1 indicating successful acetate form-
The ‘direct’ loading approach
‘Direct’ loading of a substrate onto a support is an attractive
concept. The priming of the substrate by conventional solution
chemistry prior to attaching it to a support for solid phase
synthesis, often a multi-step process, can be laborious especially
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1960

View Article Online
Table 2 Characterisation of cobalt coated resins by IR and 31P NMR
spectroscopy
Data from
solution-phase
analogue
Data from
resin
Polymer-bound complex
(19, 18, 17)
υmax/cmϪ1
1995s
υmax/cmϪ1
1999s
1975msh
1888s
1985msh
1880vs
δP (ppm)
62
δP (ppm)
55.5
υmax/cmϪ1
2074w
υmax/cmϪ1
2077m
2012msh
1995s
1955w
2023w
1988s
1958w
δP (ppm)
62
δP (ppm)
65.3
υmax/cmϪ1
1977msh
1949s
υmax/cmϪ1
1969msh
1943msh
1898msh
1810w
1025msh
δP (ppm)
68
when the first degree of diversity in a combinatorial synthesis is
introduced at the loading stage. Although only one preliminary
step is required for the loading described in the previous
section, and that it is a simple and high yielding complexation,
it seemed that it might be possible to generate a suitably primed
polymeric support that would enable us to by-pass this step.
Literature precedent was encouraging: complexation of an
alkyne to a phosphine-substituted carbonyldicobalt(0) com-
plex, [(Bu3P)Co(CO)3]2 11, has been reported.16 At elevated
temperature, one carbon monoxide and one phosphine are dis-
placed resulting in monophosphine-substituted complex 12.
The phosphine released in this process then displaces a second
carbon monoxide ligand to deliver the bisphosphine 13.
1955 cmϪ1 indicating a presence of the monophosphine-
substituted complex 15 in a ratio estimated at 7 : 3.21 This is
largely consistent with literature precedent where one group
observed only ionic species 14 on the resin:19 that a proportion
of the monophosphine 15 was also observed in these experi-
ments possibly results from using a higher cobalt(0) : PSDP
ratio in the resin synthesis. Attempts to isolate only the resin-
bound ionic complex by decreasing this ratio to 1 : 2.5 failed,
however, resulting in an identical complex composition as
determined by IR spectroscopy. The neutral bisphosphine 16
A wide variety of transition metal carbonyl complexes have
been immobilised onto a diverse array of supports,17 both
organic and inorganic, and octacarbonyldicobalt(0) is no
exception. The reaction of PSDP with octacarbonyldicobalt(0)
to give resin-bound complexes 14,15 and 16 has been
reported.18,19 Thus, treatment of a swollen suspension of PSDP
in THF with octacarbonyldicobalt(0) at ambient temperature
gave a resin-bound mixture of ionic 14 and monophosphine 15
as deep purple beads. Suspension of this material in 1,4-
dioxane and heating to 75 ЊC gave the single resin-bound com-
plex 16 bearing only a trace of monophosphine complex 15, as
determined by IR spectroscopy. Alternatively, 16 was accessed
directly by treating PSDP with octacarbonyldicobalt(0) at
elevated temperature.
gave a broad IR absorption at 1949 cm .
Ϫ1 20 The assignments are
supported by 31P NMR resonances at ∼62 ppm for [14 ؉ 15]
and ∼68 ppm for 16; the broad signals for the latter were unsuit-
able for accurate quantitative analysis, however. In order to
estimate the error margin when working with resins generated
by direct loading, it was assumed that 5% inaccuracy accom-
panied analysis of IR spectra while a larger 10% margin needed
to be used for 31P NMR spectra of 16. Hence, loadings of
0.96
0.03 mmol[Co2(CO)7]gϪ1 and 0.22
0.05 mmol[Co2-
(CO)6]gϪ1 were calculated for [14 ؉ 15] and 16, respectively.
Resins [14 ؉ 15] and 16 are significantly more air-stable than
Co2(CO)8: half-lives for the complexes have not been quantified
but the resins can be left on the open bench for several days
without apparent change in composition as determined by IR
spectroscopy [the very labile Co2(CO)8 would exist for only
minutes under these conditions]. Substantial cross linking of
the polystyrene resin created by the bis-phosphine complexes is
manifested by a greatly-reduced swelling ability of the resins.
The purple, stable resins [14 ؉ 15] and 16 were characterised
by comparison of their IR spectroscopy data in the carbonyl
stretching region with those from the literature18,19 and from
ionic, monophosphine and bisphosphine complexes, 19,22 1821
and 1720 prepared using literature procedures for that purpose,
Table 2.
A very strong absorption at 1880 cmϪ1 ascribed to ionic
1418,22 was accompanied by weaker peaks at 2074, 2012 and
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1961

View Article Online
A stoichiometric solid-phase Pauson–Khand annelation
Having established that carbonylcobalt(0) linkers could be used
to immobilise alkynes, it occurred to us that a Pauson–Khand
reaction might be used as an alternative approach to the release
of the product alkyne.
Precedent for the solid phase PKA exists for substrates
immobilised by means of a side-chain ester linkage. Schore
describes the cyclisation of a tethered alkyne 20 with strained
alkenes such as norbornadiene.23 Cleavage of the linker in the
presence of potassium hydroxide and a phase-transfer catalyst
delivered the cyclopentenone bearing a pendant alcohol. In a
solid phase intramolecular PKA reported by Bolton, cleavage
of the product 21 with TFA afforded carboxylic acid-bearing
products.24 The use of carbonyldicobalt(0) complexes both as
alkyne linkers and as a vehicle for PKA-induced cleavage from
the polymer support would represent a new approach to solid
phase Pauson–Khand annelations. Hence, it was decided to
immobilise enyne 2225 onto PSDP via its carbonyldicobalt(0)
complex.
‘Direct’ loading of an alkyne onto the carbonylcobalt(0)
coated resin [14 ؉ 15] was investigated next. Suspension of
[14 ؉ 15] in 1,4-dioxane and heating to 70 ЊC in the presence of
hex-5-yn-1-ol generated the purple resin-bound alkyne com-
plexes 6, possibly via the intermediary carbonyl complex 16. As
before, overlapping sets of IR absorptions at 2055, 2009, 1967
cmϪ1 and 2009, 1948, 1930 cmϪ1 indicated the presence of both
mono- and diphosphine substituted alkyne complexes, 6a and
6b respectively, Table 1. The strongest band at 1948 cmϪ1 sug-
gested that the latter is the major component via this route in a
composition estimated to be 3 : 7. Solution calibration studies
with triphenylphosphine analogues 4 and 5 verified this esti-
mate. The broad resonance in the 31P NMR spectrum at 55 ppm
supports alkyne complexation. Further inspection of the
spectrum revealed a 20% phosphorus site occupancy leading to
a loading of 0.30 0.09 mmol[hex-5-yn-1-ol]gϪ1
.
The hexacarbonyldicobalt(0) complex 23 was generated in
89% yield as red–purple crystals. Using the “indirect” approach
described above, this material was reacted with THF-swollen
PSDP at 50 ЊC to give resin 24 bearing a mixture of mono- and
bisphosphine complexes, 24a and 24b respectively. Analysis of
the IR spectrum of 24 revealed a predominance of 24b accom-
panied by only a trace of monophosphine 24a, an observation
in contrast to the preparation of hex-5-yn-1-ol-loaded resin 6
where a 6 : 4 ratio of mono- to bisphosphine derivatives was
obtained. Heating resin 24 at 70 ЊC under 1.05 bar CO in THF
for 24 h was found to release the cyclopentenone product 25 in
26 1% isolated yield after chromatographic purification.
Catalytic Pauson–Khand annelations
Using an acetylation procedure identical to that described
above, hex-5-yn-1-yl acetate-bearing complexes 7 were gener-
ated from complexes 6 formed by direct loading of hex-5-yn-
1-ol. Analysis of the linker by means of 31P NMR and IR
spectroscopy was consistent with no reduction in phosphorus
site occupancy and with retention of monophosphine–
bisphosphine complex composition. An additional acetate
absorption at 1732 cmϪ1 indicated successful acetylation. Hex-
5-yn-1-yl acetate 8 was released from the support by aerial
oxidation as previously described. A yield of 70 30% from
resin 6 was recorded, a cleavage efficiency comparable to that
obtained from ‘indirect’ loading.
The stoichiometric version of the Pauson–Khand reaction is
increasingly being replaced by catalytic versions of the reaction,
many of which are based on octacarbonyldicobalt(0) or
its derivatives.26 In view of the increasing awareness of the
environmental and handling advantages conferred by solid-
phase reagents and catalysts, it was decided to determine
whether or not the carbonylcobalt(0) containing polymers pre-
pared in the studies described above could be used to catalyse
the Pauson–Khand annelation.
Initially new batches of resins [14 ؉ 15] and 16 {[14 ؉ 15]Ј
and 16Ј} were prepared. The loading for 16Ј was calculated as
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1962

View Article Online
Table 3 CPKA conditions and version results using [14 ϩ 15]Ј and 16Ј
Entry
Catalyst precursor
T /ЊC
Solvent
Conversion (%)a
1
2
3
4
5
6
7
8
9
[14 ϩ 15]Ј
16Ј
[14 ϩ 15]Ј
16Ј
[14 ϩ 15]Ј
16Ј
16Ј
16Ј
16Ј recycled
70
70
65
65
75
75
70
70
70
THF
THF
THF
THF
THF
THF
1,4-Dioxane
Toluene
THF
37
(61)66
22
43
21
28
51
45
28
a Conversion as determined by 1H NMR spectroscopy (isolated yield following purification by flash column chromatography.)
the latter reaction. Thus both resins promoted a catalytic
conversion of enyne 22 into cyclopentenone 25.
To probe the effect of temperature, reactions were carried out
at 65 and 75 ЊC with both catalyst precursors [14 ؉ 15]Ј and 16Ј
(Table 3, entries 3 to 6). From these results it can be seen that
maximum conversion is obtained at 70 ЊC and the consequence
of a 5 ЊC increase or decrease in temperature leads to a drop in
conversion for both forms of the catalyst. Two other solvents
were applied to the system, both with moderate ability to swell
the resins. Neither 1,4-dioxane (Table 3, entry 7) nor toluene
(Table 3, entry 8), however, were as effective as THF. In one
further experiment, the catalyst used in Table 3, entry 2 was
recycled. Under identical conditions, a 28% conversion was
observed (Table 3, entry 9).
The phosphines present in PSDP are held very close to its
polystyrene backbone and thus site accessibility may be rather
difficult for incoming substrates and reagents. It was thus
decided to examine the Pauson–Khand reaction described
above using Novageldiphenylphosphine (NGDP), a poly-
styrene polyethyleneglycol resin. Cobalt species bound to the
phosphines present in this polymer should be sited further from
the polymer backbone than cobalt species bound to the phos-
phines present in PSDP. NGDP was thus used instead of PSDP
in the synthesis of resin [26 ؉ 27], an analogue of [14 ؉ 15].
The resulting resin was characterised by comparison of its IR
and 31P NMR spectra with those of [14 ؉ 15]. The 31P NMR
spectrum proved to be much better resolved than that of
[14 ؉ 15] (error margin = 1%) and thus the loading of resin
[26 ؉ 27] could be calculated more accurately.
0.36 0.04 mmol[Co2]gϪ1 from a phosphorus site complexation
of 50%, and the loading for [14 ؉ 15]Ј was calculated as 0.96
0.03 mmol[Co2]gϪ1 from a phosphorus site complexation of
75%. The resin loadings once again reflect the respective
uncertainties in the 31P NMR spectra and values of 0.4 mmol-
[Co2]gϪ1 and 0.95 mmol[Co2]gϪ1 were employed for 16Ј and
[14 ؉ 15]Ј respectively.
The experiments to probe catalytic activity were carried
out using 5 mol% of catalyst in THF, a solvent chosen for its
ability to swell the PSDP, under 1.05 atmospheres of carbon
monoxide, and with a reaction time of 24 hours. Preliminary
experiments demonstrated that a static atmosphere of carbon
monoxide was preferred over a continuous flow of carbon
monoxide, which led to solvent evaporation and safety prob-
lems, and that no stirring retarded the degeneration of the
beads into an amorphous gel. A temperature of 70 ЊC was
selected, in keeping with Livinghouse’s observation that there
is a thermal window between 60 and 70 ЊC for the catalytic
Pauson–Khand reaction.27 Using resins [14 ؉ 15]Ј and 16Ј,
conversions of 37 and 66% were observed (Table 3, entries 1
and 2), and cyclopentenone 25 was isolated in 61% yield from
Employing resin [26 ؉ 27] in the Pauson–Khand reaction
using substrate 22 gave a 68% conversion to cyclopentenone 25.
Thus the change in the environment around the cobalt centres
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1963

View Article Online
appears to have had little effect on the catalysis of the Pauson–
Khand reaction.
1.50 (1 H, br s, OH), 1.72 (4 H, br m, CH2CH2CH2OH),
2.90 (2 H, br m, HCCCH2), 3.71 (2 H, br m, CH2OH) and
6.03 (1 H, br m, CH); δC{1H}(90.6 MHz, CDCl3) 28.2
(CCH2CH2), 32.4 (CH2CH2OH), 34.2 (CCH2), 62.5 (CH2OH),
Conclusions
᎐
᎐
73.2 (HC᎐C), 97.1 (HC᎐C) and 200.1 [Co(CO)]; m/z (FAB
᎐
᎐
positive) 365 (Mϩ Ϫ H2O Ϫ H, 11%), 356 (M Ϫ CO, 10), 351
(M Ϫ H2O Ϫ CH2 Ϫ H, 19), 328 (M Ϫ 2CO, 100), 300
(M Ϫ 3CO, 74), 272 (M Ϫ 4CO, 59), 244 (M Ϫ 5CO, 21) and
216 (M Ϫ 6CO, 11).
The utility of carbonyldicobalt(0) complexes as linkers for the
immobilisation and subsequent release of alkyne substrates has
been successfully demonstrated for the first time. The linker is
readily characterisable by means of 31P NMR and IR spectro-
scopy, techniques which allow substrate loading to be quanti-
fied. Alkyne substrates can be loaded onto the polymeric
support both by ‘indirect’ and ‘direct’ means: it is not necessary
to prime an alkyne prior to polymer-loading by cobalt(0)
complexation. Substrate loading, however, is higher when the
alkyne complex is pre-formed, but the immobilisation–chemical
modification–release sequence proceeds with equal efficiency
thereafter.
As an alternative to oxidative release of the alkyne, a
Pauson–Khand reaction has been used to generate a cyclo-
pentenone. Thus, in principle, an alkyne may be bound to
a polymer, modified by solid-phase techniques and then
converted to a cyclopentenone on release from the polymer.
It has also been demonstrated that resins bearing carb-
onylcobalt(0) species catalyse the Pauson–Khand reaction.
Although it is as yet unclear whether catalysis is occurring on
the polymer or whether the resin acts as a source of catalytically
active carbonylcobalt(0) fragments (microanalysis of the resin
after catalysis on one occasion revealed a 13% loss of cobalt) it
is of note that from a practical viewpoint that the polymer-
bound carbonylcobalt(0) is more stable and easier to handle
than conventional solution catalysts of the Pauson–Khand
reaction such as octacarbonyldicobalt(0).
Pentacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-ol)(triphenylphosphine)-
dicobalt(0) 4 and tetracarbonyl-µ-(ꢀ2ꢀ2-hex-5-yn-1-ol)-
di(triphenylphosphine)dicobalt(0) 5
To a solution of hexacarbonyl-µ-(η2:η2-hex-5-yn-1-ol)dico-
balt(0) 3 (0.4 g, 1.04 mmol) in a solvent mixture of THF
(30 cm3) and Et2O (10 cm3) at 50 ЊC under nitrogen in the dark
was added dropwise over 30 min a solution of triphenyl-
phosphine (227 mg, 0.87 mmol) in THF (10 cm3). After 2.5 h
the mixture was concentrated in vacuo and purified by flash
column chromatography (SiO2, hexane–Et2O 4 : 1 to 0 : 1 gradi-
ent elution) to afford complexes 4 and 5. Data for 4: (hexane–
Et2O 2 : 1) deep red viscous oil (270 mg, 0.44 mmol, 42%)
(Found: C, 56.7; H, 4.2. C29H25Co2O6P requires C, 56.33; H,
4.08%); νmax(hexane)/cmϪ1 2062s, 2028w, 2008ssh, 2004s, 1990w,
᎐
1967w (C᎐O); δ (360.0 MHz, CDCl ) 1.29–1.56 (5 H, OH and
᎐
H
3
CH2CH2CH2OH), 1.98 (1 H, m, HCCCHH), 2.14 (1 H,
m, HCCCHH ), 3.50 (2 H, br t, J 5.7, CH2OH), 5.17 (1 H, d,
J 4.0, CH) and 7.44–7.52 (15 H, m, PPh3); δC{1H}(90.6 MHz,
CDCl3) 27.8 (CCH2CH2), 32.3 (CH2CH2OH), 32.5 (CCH2),
3
᎐
᎐
62.7 (CH OH), 72.2 (HC᎐C), 94.0 (HC᎐C), 128.6 (d, J 10,
᎐
᎐
2
CP
2
Ar C-m), 130.4 (Ar C-p), 133.1 (d, JCP 11, Ar C-o), 134.9 (d,
1JCP 41, Ar C-ipso), 202.4 [PCoCo(CO)], 205.8 [d, JPCoC 101,
2
PCo(CO)]; δP{1H}(145.8 MHz, CDCl3) 55.2 (CoPPh3); m/z
(FAB positive) 618 (Mϩ, 1%), 562 (M Ϫ 2CO, 3), 534
(M Ϫ 3CO, 26), 506 (M Ϫ 4CO, 8), 478 (M Ϫ 5CO, 100), 419
(M Ϫ 5CO Ϫ CH2CH2CH2OH, 3) and 321 (CoPPh3, 34). Data
for 5: (hexane–Et2O 0 : 1) red–purple solid (130 mg, 0.15 mmol,
15%), mp 161–163 ЊC (Found: C, 64.8; H, 4.8. C46H40Co2O5P2
requires C, 64.80; H, 4.73%); νmax(DCM)/cmϪ1 2010s, 1962ssh,
Experimental
Reactions under nitrogen and carbon monoxide were per-
formed using standard vacuum line and Schlenk tube tech-
niques.28 Diethyl ether, toluene and hexane were dried over
sodium wire. THF was distilled from sodium benzophenone
ketyl. DCM and 1,4-dioxane were distilled from calcium
hydride. All solvents intended for use in chromatography were
used as bought. SiO2 refers to chromatography grade silica of
particle size 40–63 µm. Enyne 22 was prepared using a literature
procedure.25
᎐
1950s and 1930wsh (C᎐O); δ (360.0 MHz, CDCl ) 0.70–0.95
᎐
H
3
(4 H, m, CH2CH2CH2OH), 1.36 (1 H, s, OH), 1.40 (2 H, br m,
HCCCH2), 3.18 (2 H, br m, CH2OH), 4.35 (1 H, br s, CH) and
7.27–7.37 (30 H, m, 2PPh3); δC{1H}(90.6 MHz, CDCl3) 27.0
(CCH2CH2), 31.3 (CCH2), 32.6 (CH2CH2OH), 63.0 (CH2OH),
Melting points, which are uncorrected, were determined
using a Gallenkamp capillary melting point apparatus. IR
spectra were obtained on a Perkin-Elmer 1710 FTIR spectro-
meter. NMR spectra were recorded at room temperature on
Bruker AM360 (360.0 MHz 1H, 90.6 MHz 13C, 145.8 MHz 31P),
Bruker AMX400 (400.0 MHz 1H, 100.6 MHz 13C) (King’s
᎐
᎐
71.9 (HC᎐C), 89.8 (HC᎐C), 128.3 (Ar C-m), 129.7 (Ar C-p),
᎐
᎐
2
133.1 (Ar C-o), 136.0 (m, Ar C-ipso), 207.4 [d, JPCoC 36,
PCo(CO)]; δP{1H}(145.8 MHz, CDCl3) 53.4 (CoPPh3); m/z
(FAB positive) 796 (Mϩ Ϫ 2CO, 4%), 768 (M Ϫ 3CO, 4), 740
(M Ϫ 4CO, 15), 642 [Co2(PPh3)2, 12], 534 (M Ϫ 2CO Ϫ PPh3,
6), 506 (M Ϫ 3CO Ϫ PPh3, 3), 478 (M Ϫ 4CO Ϫ PPh3, 100), 419
(M Ϫ 4CO Ϫ PPh3 Ϫ CH2CH2CH2OH, 5) and 321 (CoPPh3,
55).
College) and Bruker DPX300 (300.0 MHz H, 75.5 MHz 13C)
1
(AstraZeneca UK Ltd.). Mass spectra were recorded on a VG
70/250 SE spectrometer at AstraZeneca UK Ltd. and on Kratos
MS890MS and JEOL AX505W instruments at King’s College
London. Microanalyses were conducted by Steven Boyer
(S. A. C. S.) at the University of North London and by
MEDAC Ltd., Analytical and Chemical consultancy services,
Brunel Science Centre.
Pentacarbonyl-µ-(ꢀ2ꢀ2-hex-5-yn-1-ol)(polystyrenediphenyl-
phosphine)dicobalt(0) 6a and tetracarbonyl-µ-(ꢀ2ꢀ2-hex-5-yn-
1-ol)di(polystyrenediphenylphosphine)dicobalt(0) 6b
Polystyrenediphenylphosphine (1 g, 1.6 mmolP) was suspended
at ambient temperature in anhydrous THF (5 cm3) and, after
20 min under constant nitrogen agitation, a solution of hexa-
carbonyl-µ-(η2:η2-hex-5-yn-1-ol)dicobalt(0) 3 (1.2 g, 3.2 mmol)
in THF (5 cm3) was added. The mixture was heated to 50 ЊC
under constant nitrogen agitation for 4 h. The resulting deep
purple beads were filtered, washed with alternate aliquots
(20 cm3) of THF and diethyl ether until the filtrate became
colourless, and dried in vacuo to afford the title compound resin
(1.52 g, 0.94 0.02 mmol[hex-5-yn-1-ol]gϪ1) (Found: Co, 9.0; P,
3.6. Resin 6 requires Co, 8.90; P, 3.66%); νmax(Nujol)/cmϪ1 2055s
Hexacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-ol)dicobalt(0) 3
To a solution of octacarbonyldicobalt(0) (5.0 g, 14.6 mmol) in
anhydrous, deoxygenated hexane (40 cm3) at ambient tem-
perature under nitrogen was added hex-5-yn-1-ol (1.61 cm3,
14.6 mmol) and the solution was stirred for 24 h. Flash column
chromatography (SiO2, hexane–Et2O 1 : 0 to 4 : 1 gradient
elution) effected isolation of the title complex (4.78 g, 12.4
mmol, 85%) as a deep red oil (Found: C, 39.2; H, 2.8.
C12H10Co2O7 requires C, 39.16; H, 2.62%); νmax(hexane)/cmϪ1
2092w, 2052s, 2029s and 2020m (C᎐O); δ (360.0 MHz, CDCl )
(6a, C᎐O), 1998vs, 1979ssh (6a and 6b, C᎐O) and 1950s (6b,
᎐
᎐
᎐
᎐
1
᎐
᎐
᎐
C᎐O) [6a–6b 6 : 4]; δ { H}(145.8 MHz)(D O capillary lock)
᎐
H
3
P
2
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1964

View Article Online
Pentacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-al)(polystyrenediphenyl-
Table 4 IR spectroscopy calibration study for absorption intensities
of mixtures of mono- and di(triphenylphosphine) substituted alkyne
complexes 4 and 5
phosphine)dicobalt(0) 9a and tetracarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-
1-al)di(polystyrenediphenylphosphine)dicobalt(0) 9b
Carbonyl-µ-(η2:η2-hex-5-yn-1-ol)(polystyrenediphenylphos-
phine)dicobalt(0) [6a ؉ 6b] (500 mg, 0.47 mmol[5-hexyn-1-ol])
was suspended at ambient temperature in anhydrous DCM
(5 cm3) under constant nitrogen agitation for 20 min. DMSO
(3 cm3) and triethylamine (515 mg, 0.71 cm3, 5.1 mmol) were
added sequentially, followed by a solution of sulfur trioxide
pyridine complex (405 mg, 2.55 mmol) in DMSO (2 cm3). After
7 h at ambient temperature, the resulting deep purple beads
were filtered, washed thoroughly with alternate aliquots of
DMSO, DCM and diethyl ether and dried in vacuo to afford the
Complex mixture composition
4 : 5
4
5
a mg, b mmol
c mg, d mmol
6 : 4
5 : 5
4 : 6
3 : 7
2 : 8
6.4, 0.0104
6.4, 0.0104
6.4, 0.0104
3.2, 0.0052
1.6, 0.0026
5.9, 0.0069
8.8, 0.0104
13.2, 0.0155
10.3, 0.0121
8.8, 0.0104
title compound resin (495 mg, 0.85
0.03 mmol[hex-5-yn-1-
al]gϪ1); νmax(Nujol)/cmϪ1 2056s (9a, C᎐O), 1996vs and 1980ssh
᎐
᎐
᎐
᎐
(9a ؉ 9b, C᎐O), 1951s (9b, C᎐O) [9a–9b 6 : 4] and 1723m
᎐
᎐
(CHO); δP{1H}(145.8 MHz)(D2O capillary lock) 24.6 [poly-
styrene-P(O)Ph2, 30%] and 54.1 (polystyrene-PPh2–Co, 70%).
25.6 [polystyrene-P(O)Ph2, 20%] and 55.5 (polystyrene-PPh2–
Co, 80%).
Infra-red spectroscopy calibration study for mixtures of penta-
carbonyl-µ-(ꢀ2ꢀ2-hex-5-yn-1-ol)(triphenylphosphine)dicobalt(0) 4
and tetracarbonyl-µ-(ꢀ2ꢀ2-hex-5-yn-1-ol)di(triphenylphosphine)-
dicobalt(0) 5
Hex-5-yn-1-al 1015
Carbonyl-µ-(η2:η2-hex-5-yn-1-al)(polystyrenediphenylphos-
phine)dicobalt(0) [9a ؉ 9b] (420 mg, 0.36 mmol[hex-5-yn-1-yl
acetate]) was ground to a powder using a pestle and mortar
and suspended in DCM (15 cm3) in a 25 cm3 round bottomed
flask equipped with a condenser and CaCl2 drying tube and
stirred at ambient temperature in air under white light (100 W)
for 72 h. The resulting brown suspension was filtered through
Celite and the polymeric residue was washed with DCM. Care-
ful evaporation of the filtrate and washings afforded the colour-
less volatile product (5.7 mg, 0.06 mmol, 17 0.5%); νmax(neat)/
Complex 4 (a mg, b mmol) and complex 5 (c mg, d mmol)
were dissolved in DCM (6 cm3) and analysed by IR spectro-
scopy using a solution cell (Table 4). Qualitative comparison
of absorption intensities with those from resin-bound
mixtures provided an estimate of complex composition
[6a ؉ 6b].
Pentacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-yl acetate)(polystyrenedi-
phenylphosphine)dicobalt(0) 7a and tetracarbonyl-µ-(ꢀ2:ꢀ2-hex-5-
yn-1-yl acetate)di(polystyrenediphenylphosphine)dicobalt(0) 7b
cmϪ1 3310m (CC–H stretch), 2145w (C᎐C stretch), 1723s (HC᎐
᎐
᎐
᎐
O) and 643m (CCH bend); δH (360.0 MHz, CDCl3) 1.79 (2 H,
qn, J 7, CH2CH2CHO), 1.92 (1 H, t, J 3, HCCCH2), 2.21 (2 H,
td, J 7 and 3, HCCCH2), 2.55 (2 H, td, J 7 and 1, CH2CHO)
and 9.81 (1 H, t, J 1, CHO) (lit.15 1.85, 1.98, 2.27, 2.61 and
9.80); δC{1H}(90.6 MHz, CDCl3) 17.8 (CCCH2), 20.8
(CCCH2CH2), 42.6 (CH2CHO), 69.4 (HCCCH2), 79.8
(HCCCH2) and 201.8 (CHO).
Carbonyl-µ-(η2:η2-hex-5-yn-1-ol)(polystyrenediphenylphos-
phine)dicobalt(0) [6a ؉ 6b] [300 mg, 0.28 mmol(hex-5-yn-1-ol)]
was suspended at ambient temperature in anhydrous THF
(10 cm3) under constant nitrogen agitation and, after 20 min,
triethylamine (2.5 cm3) and acetic anhydride (1.5 cm3) were
added sequentially. After 20 h at ambient temperature, the
resulting deep purple beads were filtered, washed thoroughly
with alternate aliquots of THF and diethyl ether (20 cm3) and
dried in vacuo to afford the title compound resin (325 mg, 0.87
0.02 mmol[hex-5-yn-1-yl acetate]gϪ1); νmax(Nujol)/cmϪ1 2055s
[Tricarbonyldi(polystyrenediphenylphosphine)cobalt(1؉)
tetracarbonylcobaltate(1؊) 14 and heptacarbonyl(polystyrene-
diphenylphosphine)dicobalt(0)] 1519b
Polystyrenediphenylphosphine (2 g, 3.2 mmol P) was suspended
at ambient temperature in anhydrous THF (15 cm3) and a
solution of octacarbonyldicobalt(0) (2.2 g, 6.4 mmol) in THF
(5 cm3) was added via filter cannula. After 1.5 h under constant
nitrogen agitation, the mixture was filtered and washed with
alternate aliquots of THF and Et2O until the filtrate became
colourless. The resulting solid was dried in vacuo to afford the
product complex (2.37 g, 0.96 0.03 mmol[Co2]gϪ1) as purple–
᎐
᎐
(7a, C᎐O), 1994vs and 1981ssh (7a and 7b, C᎐O), 1950s (7b,
᎐
᎐
1
᎐
C᎐O) [7a–7b 6 : 4] and 1734m [O(C᎐O)CH ]; δ { H}(145.8
᎐
᎐
3
P
MHz)(D2O capillary lock) 24.7 [polystyrene-P(O)Ph2, 25%] and
54.3 (polystyrene-PPh2-Co, 75%).
Hex-5-yn-1-yl acetate 812
Carbonyl-µ-(η2:η2-hex-5-yn-1-yl
acetate)(polystyrenedi-
red beads; νmax(Nujol)/cmϪ1 2074w, 2012msh (15, C᎐O), 1995s
᎐
᎐
phenylphosphine)dicobalt(0) [7a ؉ 7b] (150 mg, 0.132 mmol-
[hex-5-yn-1-yl acetate]) was ground to a powder using a pestle
and mortar and suspended in DCM (15 cm3) in a 25 cm3 round
bottomed flask equipped with a condenser and CaCl2 drying
tube and stirred at ambient temperature in air under white light
(100 W) for 72 h. The resulting brown suspension was filtered
through Celite and the polymeric residue was washed with
DCM. The combined filtrate and washings were concentrated
in vacuo to afford the product as a colourless oil (12 mg,
᎐
᎐
᎐
(14 ؉ 15, C᎐O), 1985msh (15, C᎐O), 1955w (14, C᎐O) and
᎐
᎐
᎐
19b
᎐
1880vs (15, C᎐O) [14–15 7 : 3] (lit. 2000s, 1955m, 1800s);
᎐
δP{1H}(145.8 MHz)(D2O capillary lock, THF-swollen resin)
32 [polystyrene-P(O)Ph2, 25%] and 62 (polystyrene-PPh2–Co,
75%).
Hexacarbonyldi(polystyrenediphenylphosphine)dicobalt(0)-
(Co–Co) 1619b
0.09 mmol, 66 2%); νmax(neat)/cmϪ1 3296w (C᎐C) and 1732
The resin-bound complex 16 was prepared both a) by thermal
treatment of resin-bound complexes [14 ؉ 15] and; b) by direct
reaction between polystyrenediphenylphosphine and octacarb-
onyldicobalt(0) at elevated temperature.
᎐
᎐
(OC᎐O); δ (400.0 MHz, CDCl ) 1.54 (2 H, m, CH CH CH O
᎐
H
3
2
2
2
or CH2CH2CH2O), 1.68 (2 H, m, CH2CH2CH2O or CH2CH2-
CH2O), 1.90 (1 H, t, J 3, HCCCH2), 1.98 (3 H, s, COCH3), 2.17
(2H, td, J 7 and 3, HCCCH2) and 4.02 (2 H, t, J 7, CH2O) (lit.12
1.62, 1.77, 1.96, 2.1, 2.22 and 4.1); δC{1H}(90.6 MHz, CDCl3)
18.1 (CCCH2), 21.0 (COCH3), 25.0 (CH2CH2O), 27.7
(CCCH2CH2), 64.0 (CH2O), 68.8 (HC), 83.9 (HCCCH2) and
171.2 (CO).
a) Thermal treatment of resin-bound complexes [14 ؉ 15].
Resin-bound complex mixture [14 ؉ 15] (490 mg, 0.47 mmol-
[Co2]gϪ1) was suspended in anhydrous 1,4-dioxane (10 cm3) and
was heated to 75 ЊC for 16 h in the dark. The black mixture was
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1965

View Article Online
cooled, filtered and washed with alternate aliquots of THF and
Et2O until the filtrate became colourless. The resulting solid was
dried in vacuo to afford the product complex (482 mg, 0.22
0.05 mmol[Co2]gϪ1) as deep purple beads; νmax(Nujol)/cmϪ1
by filtration, washed with hexane (3 × 10 cm3) and dried in
vacuo to afford the title complex (575 mg, 0.69 mmol, 78%) as a
pale brown crystalline solid, mp 135–140 ЊC (dec.) (Found: C,
61.6; H, 3.8. C43H30Co2O7P2 requires C, 61.59; H, 3.61%);
19b
22
νmax(Nujol)/cmϪ1 1999s, 1975msh and 1888sbr (C᎐O) (lit.
᎐
᎐
2032w, 1977msh, 1949s and 1925msh (C᎐O) (lit. 2041vw,
᎐
᎐
1963sh, 1952vs and 1901vw); δP{1H}(145.8 MHz)(D2O capil-
lary lock, THF-swollen resin) 31 [polystyrene-P(O)Ph2, 70%]
and 68 (polystyrene-PPh2–Co, 30%).
2002s, 1883vs); δH(360.0 MHz, CDCl3) 7.44 (12 H, br m, Ar–H)
and 7.63 (18 H, br m, Ar–H); δP{1H}(145.8 MHz, CDCl3) 55.5
(CoPPh3); m/z (FAB positive) 667 [Co(CO)3(PPh3)2ϩ, 50%], 611
[Co(CO)(PPh3)2, 16], 583 [Co(PPh3)2, 100] and 321 (CoPPh3,
90).
b) Direct reaction between polystyrenediphenylphosphine and
octacarbonyldicobalt(0) at elevated temperature. Polystyrene-
diphenylphosphine (1 g, 1.6 mmol P) was suspended at ambient
temperature in anhydrous 1,4-dioxane (15 cm3) for 30 min,
after which time a solution of octacarbonyldicobalt(0) (383 mg,
1.12 mmol) in 1,4-dioxane (5 cm3) was added via filter cannula.
After 30 min under constant nitrogen agitation at ambient tem-
perature, the mixture was heated to 75 ЊC for 16 h in the dark.
The black mixture was cooled, filtered and washed with alter-
nate aliquots of THF and Et2O until the filtrate became colour-
less. The resulting solid was dried in vacuo to afford the product
Pentacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-ol)(polystyrenediphenyl-
phosphine)dicobalt(0) 6a and tetracarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-
1-ol)di(polystyrenediphenylphosphine)dicobalt(0) 6b (direct
method)
To a suspension of complex mixture [14 ؉ 15] (400 mg,
0.38 mmol[Co2]) in anhydrous dioxane (10 cm3) was added
hex-5-yn-1-ol (254 mg, 0.29 cm3, 2.6 mmol). After 2 h constant
nitrogen agitation at 70 ЊC, the deep purple beads were filtered,
washed thoroughly with alternate aliquots of THF and diethyl
ether and dried in vacuo to afford the product complex (408 mg,
0.30 0.09 mmol[hex-5-yn-1-ol]gϪ1); νmax(Nujol)/cmϪ1 2055w
complex (1.15 g, 0.42
0.04 mmol[Co2]gϪ1) as deep purple
beads (Found: Co, 7.7; P, 3.7. Resin 16 requires Co, 4.98; P,
3.61%); νmax(Nujol)/cmϪ1 2032w, 1977msh, 1949s and 1925msh
᎐
᎐
᎐
(6a, C᎐O), 2009s (6a and 6b, C᎐O), 1967ssh (6a, C᎐O) and
᎐
᎐
᎐
19b
(C᎐O) (lit. 2041vw, 1963sh, 1952vs and 1901vw); δP{1H}-
᎐
1
᎐
᎐
1948vs and 1930ssh (6b, C᎐O) [6a–6b 3 : 7]; δ { H}(145.8 MHz,
᎐
P
(145.8 MHz, D2O capillary lock, THF-swollen resin) 29 [poly-
styrene-P(O)Ph2, 40%] and 69 (polystyrene-PPh2–Co, 60%).
D2O capillary lock) Ϫ5 (polystyrene-PPh2, 20%), 25 [poly-
styrene-P(O)Ph2, 60%] and 55 (polystyrene-PPh2–Co, 20%).
For acetylation and decomplexation of the product alkynyl
ester the procedures were identical to those employed for the
indirect loading method.
Hexacarbonyldi(triphenylphosphine)dicobalt(0) 1720
Octacarbonyldicobalt(0) (1.1 g, 3.21 mmol), triphenyl-
phosphine (1.69 g, 6.43 mmol) and 1,4-dioxane (20 cm3) were
combined under nitrogen and heated to 60 ЊC in the dark for
6 h. The red–brown precipitate was filtered, washed with THF
(3 × 10 cm3) and dried in vacuo to give the title complex (2.28 g,
2.81 mmol, 96%) as a brown powder, mp 125–130 ЊC (dec.)
(lit.20 130 ЊC) (Found: C, 62.4; H, 3.8. C42H30Co2O6P2 requires
C, 62.24; H, 3.73%); νmax(Nujol)/cmϪ1 1969msh, 1943s,
Pentacarbonyl-µ-(ꢀ2:ꢀ2-hex-5-yn-1-yl acetate)(polystyrenedi-
phenylphosphine)dicobalt(0) 7a and tetracarbonyl-µ-(ꢀ2:ꢀ2-hex-5-
yn-1-yl acetate)di(polystyrenediphenylphosphine)dicobalt(0) 7b
Carbonyl-µ-(η2:η2-hex-5-yn-1-ol)(polystyrenediphenylphos-
phine)dicobalt(0) [6a ؉ 6b] [300 mg, 0.09 mmol(hex-5-yn-1-ol)]
was suspended at ambient temperature in anhydrous THF
(10 cm3) under constant nitrogen agitation and, after 20 min,
triethylamine (2.5 cm3) and acetic anhydride (1.5 cm3) were
added sequentially. After 20 h at ambient temperature, the
resulting deep purple beads were filtered, washed thoroughly
with alternate aliquots of THF and diethyl ether (20 cm3) and
dried in vacuo to afford the product resin (296 mg, 0.30 0.09
mmol[hex-5-yn-1-yl acetate]gϪ1); νmax(Nujol)/cmϪ1 2055w (7a,
ϩ
᎐
1898msh and 1810w (C᎐O); m/z (FAB positive) 321 (Ph PCo ,
᎐
3
6%).
Heptacarbonyl(triphenylphosphine)dicobalt(0) 1821
To a solution of octacarbonyldicobalt(0) (2 g, 5.85 mmol) in
THF (15 cm3) under nitrogen in the dark was added a solution
of triphenylphosphine (1.38 g, 5.26 mmol) over 30 min. After a
further 30 min the mixture was concentrated in vacuo and the
brown residue was purified by flash column chromatography
(SiO2, hexane–Et2O 1 : 0 to 9 : 1 gradient elution) and recrystal-
lisation (hexane–Et2O) to give the title complex (1.97 g, 3.42
mmol, 65%) as deep purple crystals, mp 120–125 ЊC (dec.)
(Found: C, 52.1; H, 2.7. C25H15Co2O7P requires C, 52.11; H,
2.62%); νmax(THF)/cmϪ1 2077m, 2023w, 1988s and 1958w
᎐
᎐
᎐
C᎐O), 2010s (7a and 7b, C᎐O), 1967ssh (7a, C᎐O), 1949vs and
᎐
᎐
᎐
᎐
1930ssh (7b, C᎐O) [7a–7b 3 : 7] and 1732m [O(C᎐O)CH ];
᎐
᎐
3
δP{1H}(145.8 MHz, D2O capillary lock) Ϫ5.2 (polystyrene-
PPh2, 20%), 24.8 [polystyrene-P(O)Ph2, 60%] and 53.4
(polystyrene-PPh2–Co, 20%).
Hex-5-yn-1-yl acetate 812
21
Carbonyl-µ-(η2:η2-hex-5-yn-1-yl
acetate)(polystyrenedi-
᎐
(C᎐O) (lit. 2079, 2026, 2010sh, 1996, 1964); δH(360.0 MHz,
᎐
CDCl3) 7.48 (m, Ar–H); δC{1H}(90.6 MHz, CDCl3) 129.5 (d,
3JCP10, Ar C-m), 131.3 (s, Ar C-p), 132.6 (Ar C-ipso), 133.1 (d,
2JCP10, Ar C-o) and 201.1 [br s, Co(CO)]; δP{1H}(145.8 MHz,
CDCl3) 65.3 (CoPPh3); m/z (FAB positive) 548 (Mϩ Ϫ CO, 2%),
492 (M Ϫ 3CO, 21), 405 (M Ϫ Co Ϫ 4CO, 7), 377 (M Ϫ Co Ϫ
5CO, 13), 349 (M Ϫ Co Ϫ 6CO, 17) and 321 (CoPPh3, 100).
phenylphosphine)dicobalt(0) [7a ؉ 7b] (280 mg, 0.08 mmol-
[hex-5-yn-1-yl acetate]) was ground to a powder using a pestle
and mortar and suspended in DCM (15 cm3) in a 25 cm3 round
bottomed flask equipped with a condenser and CaCl2 drying
tube and stirred at ambient temperature in air under white light
(100 W) for 72 h. The resulting brown suspension was filtered
through Celite and the polymeric residue was washed with
DCM. The combined filtrate and washings were concentrated
in vacuo to afford the product as a colourless oil (8.5 mg,
0.06 mmol, 72 30%).
Tricarbonyldi(triphenylphosphine)cobalt(1؉) tetracarbonyl-
cobaltate(1؊) 1922
To a solution of octacarbonyldicobalt(0) (300 mg, 0.88 mmol)
in THF (50 cm3) under nitrogen at ambient temperature was
added triphenylphosphine (1.84 g, 7.02 mmol) as one portion.
After stirring for 20 min, the solution was cooled to 0 ЊC
whereupon hexacarbonyldi(triphenylphosphine)dicobalt(0) 17
(104 mg, 0.13 mmol, 15%) precipitated. The mixture was fil-
tered via filter cannula under nitrogen and to the filtrate was
added hexane (50 cm3). The resulting precipitate was collected
Hexacarbonyl[N-(prop-2-enyl)-N-µ-(ꢀ2:ꢀ2-prop-2-ynyl)-p-
toluenesulfonamide]dicobalt(0) 23
To a solution of N-(prop-2-enyl)-N-(prop-2-ynyl)-p-toluene-
sulfonamide 22 (1 g, 4.02 mmol) in DCM (20 cm3) at ambient
temperature under nitrogen was added octacarbonyldicobalt(0)
(1.51 g, 4.42 mmol). After 2 h the mixture was concentrated
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1966

View Article Online
Table 5 Conditions and reagents for catalytic Pauson–Khand annelation
Entry
Catalyst precursor
Catalyst precursor (mg, mmol)
Solvent
T / ЊC (z)
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
[14 ϩ 15]Ј
16Ј
[14 ϩ 15]Ј
16Ј
[14 ϩ 15]Ј
16Ј
26, 0.025
63, 0.025
26, 0.025
63, 0.025
26, 0.025
63, 0.025
63, 0.025
63, 0.025
THF
THF
THF
THF
THF
THF
1,4-Dioxane
Toluene
70
70
65
65
75
75
70
70
16Ј
16Ј
in vacuo and purified by flash column chromatography (SiO2,
hexane–Et2O 1 : 0 to 19 : 1 gradient elution) to give the title
complex (1.92 g, 3.59 mmol, 89%) as a deep red solid, mp 100–
101 ЊC (Found: C, 42.6; H, 2.9; N, 2.6. C19H15Co2NO8S requires
brown residue by flash column chromatography (SiO2, hexane–
EtOAc 4 : 1 to 3 : 2 gradient elution) gave the title compound 25
(19 mg, 0.069 mmol, 26 1%) as a white solid. (See below for
spectroscopic data.)
C, 42.64; H, 2.82; N, 2.62%); νmax(DCM)/cmϪ1 2096w, 2058s,
᎐
2029s (C᎐O) and 1159 [S(᎐O) ]; δ (360.0 MHz, CDCl ) 2.44 (3
Characterisation of batches of resins [14 ؉ 15]Ј and 16Ј used for
᎐
᎐
2
H
3
H, s, CH ), 3.97 (2 H, d, J 7, CH CH᎐CH ), 4.49 (2 H, s,
catalysis experiments
᎐
3
2
2
CH CCH), 5.16 [1 H, dd, J 11, 1, C᎐CHH], 5.20 [1 H, dd, J 7, 1,
᎐
2
Employing a method identical to that for [14 ؉ 15] described
above, the resin complex [14 ؉ 15]Ј was isolated as purple–red
beads:
C᎐CHH ], 5.54 (1 H, m, CH᎐CH ), 6.01 (1 H, s, CH CCH ) and
᎐
᎐
2
2
7.31 (2 H, d, J 8, Ar–H) and 7.72 (2 H, d, J 8, Ar–H);
δC{1H}(90.6 MHz, CDCl3) 21.6 (CH3), 48.8 (CH2CCH), 50.5
[14 ؉ 15]Ј: (1.19 g, 0.96 0.03 mmol[Co2]gϪ1) [0.95 mmol-
(CH CH᎐CH ), 73.7 (CH CCH), 89.7 (CH CCH), 120.1 (CH᎐
᎐
᎐
[Co2]gϪ1 applied to catalysis]; νmax(Nujol)/cmϪ1 2074w, 2014msh
2
2
2
2
CH ), 127.3 (o-Ar), 129.9 (m-Ar), 132.1 (CH᎐CH ), 137.4 (p-
᎐
2
2
᎐
᎐
᎐
᎐
(14, C᎐O), 1995s (14؉15, C᎐O), 1982msh (1, C᎐O), 1955w (14,
᎐
᎐
Ar), 143.7 (ipso-Ar) and 199.1 [Co(CO)]; m/z (FAB positive ϩ
NaI) 557 (MNaϩ Ϫ H, 22%), 501 (MNa Ϫ H Ϫ 2CO, 38), 473
(MNa Ϫ H Ϫ 3CO, 47), 417 (MNa Ϫ H Ϫ 5CO, 52), 366 (MH
Ϫ 6CO, 40), 272 (MNa Ϫ 2Co Ϫ 6CO, 59), 143 [Co(CO)3, 23]
and 59 (Co, 100).
19b
᎐
᎐
C᎐O) and 1878vs (14, C᎐O) [14–15 7 : 3] (lit. 2000s, 1955m,
᎐
᎐
1800s); δP{1H}(145.8 MHz)(D2O capillary lock, THF-swollen
resin) 31 [polystyrene-P(O)Ph2, 25%] and 68 (polystyrene-
PPh2–Co, 75%).
Employing a method identical to that for 16 described
above, the resin complex 16Ј was isolated as deep purple beads
(1.15 g, 0.36 0.04 mmol[Co2]gϪ1) [0.4 mmol[Co2]gϪ1 applied
to catalysis]; νmax(Nujol)/cmϪ1 2032w, 1977msh, 1949s and
Pentacarbonyl(polystyrenediphenylphosphine)[N-(prop-2-enyl)-
N-µ-(ꢀ2:ꢀ2-prop-2-ynyl)-p-toluenesulfonamide]dicobalt(0) 24a
and tetracarbonyldi(polystyrenediphenylphosphine)[N-(prop-2-
enyl)-N-µ-(ꢀ2:ꢀ2-prop-2-ynyl)-p-toluenesulfonamide]dicobalt(0)
24b
19b
᎐
1925msh (C᎐O) (lit. 2041vw, 1963sh, 1952vs and 1901vw);
᎐
δP{1H}(145.8 MHz)(D2O capillary lock, THF-swollen resin) 29
[polystyrene-P(O)Ph2, 50%] and 69 (polystyrene-PPh2–Co,
50%).
Polystyrenediphenylphosphine (1 g, 1.6 mmolP) was suspended
at ambient temperature in THF (10 cm3) under constant nitro-
gen agitation for 20 min. To the suspension was added a solu-
tion of hexacarbonyl[N-(prop-2-enyl)-N-µ-(η2:η2-prop-2-ynyl)-
p-toluenesulfonamide]dicobalt(0) 23 (860 mg, 1.6 mmol) in
THF (5 cm3) via cannula under nitrogen and the mixture was
heated to 50 ЊC for 4 h in the dark. The beads were filtered and
washed with alternate aliquots of THF and Et2O until the
filtrate became colourless and dried in vacuo to give the title
compound resin (1.61 g, 0.44 0.03 mmol[N-(prop-2-enyl)-N-
(prop-2-ynyl)-p-toluenesulfonamide]gϪ1) as deep purple beads;
2,3,3a,4-Tetrahydro-2-[(4-methylphenyl)sulfonyl]cyclopenta[c]-
pyrrol-5(1H)-one 25 via catalysis27b
A mixture of Catalyst Precursor ([14 ؉ 15]Ј, or 16Ј) and
N-(2-propenyl)-N-(2-propynyl)-4-methylphenylsulfonamide 22
(125 mg, 0.5 mmol) in anhydrous solvent (5 cm3) under an
atmosphere of CO (1.05 bar) was heated to z ЊC for 24 h in the
dark without stirring (for details refer to Table 5). The mixture
was filtered, the resin was washed with alternate aliquots of
THF and Et2O (4 × 1 cm3) and the combined filtrates were
concentrated in vacuo. The brown residue was redissolved in
CDCl3 and the extent of reaction was calculated from 1H NMR
spectroscopy.
For the experiment listed in Table 5, entry 2, purification of
the residue by flash column chromatography (SiO2, hexane–
EtOAc 4 : 1 to 3 : 2 gradient elution) gave the title compound
(85 mg, 0.31 mmol, 61%) as a white solid; mp 147–149 ЊC
(lit.27b 145–148 ЊC); νmax(DCM)/cmϪ1 3058w [C–H(Ar)], 1716s
νmax(Nujol)/cmϪ1 2075w and 2047w (24a, C᎐O), 2031m, 2008m,
᎐
᎐
᎐
1976m, 1949s and 1926msh (24b, C᎐O) and 1350w and 1157w
᎐
[S(᎐O) ]; δ {1H}(145.8 MHz, D O capillary lock) 28 [polymer-
᎐
2
P
2
P(O)Ph2, 30%], 63 [polymer-PPh2Co, 70%]; m/z (FAB positive)
437
(tetracarbonyl[N-(prop-2-enyl)-N-µ-(η2:η2-prop-2-ynyl)-
p-toluenesulfonamide]dicobalt(0)ϩ
CH CH᎐CH H,
1%), 410 (tricarbonyl[N-(prop-2-enyl)-N-µ-(η2:η2-prop-2-ynyl)-
Ϫ
Ϫ
᎐
2
2
p-toluenesulfonamide]dicobalt(0) Ϫ CH CH᎐CH , 2%) and
᎐
2
2
394 (monocarbonyl[N-(prop-2-enyl)-N-µ-(η2:η2-prop-2-ynyl)-p-
(C᎐O) and 1350s and 1164 (NSO2); δH(360.0 MHz, CDCl3) 1.99
᎐
toluenesulfonamide]dicobalt(0)ϩ Ϫ H, 1%).
[1 H, dd, J 18, 4, CHH(7)], 2.32 (3 H, s, CH3), 2.48–2.58 [2 H,
m, CHH(7) and CHH(4)], 3.08 [1 H, m, CH(6)], 3.93–3.98 [2 H,
m, CHH(4) and CHH(5)], 4.27 [1 H, d, J 17, CHH(5)], 5.92
Procedure for stoichiometric Pauson–Khand annelation of
polymer-bound complexes [24a ؉ 24b]:
(1 H, s, C᎐CH ), 7.28 (2 H, d, J 8, m-Ar–H ), 7.66 (2 H, d,
᎐
2,3,3a,4-Tetrahydro-2-[(4-methylphenyl)sulfonyl]cyclopenta[c]-
pyrrol-5(1H)-one 2527b
J 8, o-Ar–H ); δC{1H}(90.6 MHz, CDCl3) 21.6 (CH3), 39.8
[CH2(7)], 44.0 [CH(6)], 47.7 [CH2(5)], 52.5 [CH2(4)], 126.2
Resin [24a ؉ 24b] (600 mg, 0.26 mmol[N-(prop-2-enyl)-N-
(prop-2-ynyl)-p-toluenesulfonamide]) was suspended in THF
(10 cm3) under an atmosphere of CO (1.05 mbar) and was
heated to 70 ЊC for 24 h in the dark. The resulting deep brown
mixture was filtered, the purple beads were washed with alter-
nate aliquots of THF and Et2O (4 × 5 cm3) and the combined
filtrates were concentrated in vacuo. Purification of the resulting
(C᎐CH), 127.5 (o-Ar-C), 130.1 (m-Ar-C), 144.2 (ipso-Ar–C),
᎐
178.6 (C᎐CH), 207.2 (C᎐O); m/z (FAB positive) 278 (MHϩ,
᎐
᎐
38%), 154 (M Ϫ C7H7O2, 100), 122 (M Ϫ C7H7O2S, 100) and 91
(C7H7, 23).
On one occasion, the deep purple resin collected after
catalysis was dried in vacuo and sent for microanalysis (Found:
Co, 6.7; P, 3.48).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1967

View Article Online
4 S. E. Gibson, N. J. Hales and M. A. Peplow, Tetrahedron Lett., 1999,
40, 1417.
5 A. C. Comely, S. E. Gibson, N. J. Hales and M. A. Peplow, J. Chem.
Soc., Perkin Trans. 1, 2001, 2526.
[Tricarbonyldi(Novageldiphenylphosphine)cobalt(1؉) tetracarb-
onylcobaltate(1؊) 26 and heptacarbonyl(Novageldiphenyl-
phosphine)dicobalt(0)] 27
6 K. M. Nicholas and R. Pettit, Tetrahedron Lett., 1971, 3475.
7 T. Nakamura, T. Matsui, K. Tanino and I. Kawajima, J. Org. Chem.,
1997, 62, 3032.
8 K. M. Brummond and J. L. Kent, Tetrahedron, 2000, 56, 3263.
9 A. C. Comely, S. E. Gibson and N. J. Hales, Chem. Commun., 1999,
2075.
10 A. C. Comely, S. E. Gibson and N. J. Hales, Chem. Commun., 2000,
305.
11 G. Váradi, A. Vizi-Orosz, S. Vastag and G. Pályi, J. Organomet.
Chem., 1976, 108, 225.
12 S. Sharma and A. C. Oehlschlager, J. Org. Chem., 1989, 54, 5064.
13 A. J. Mancuso and D. Swern, Synthesis, 1981, 165.
14 A. M. Fivush and T. M. Willson, Tetrahedron Lett., 1997, 38, 7151.
15 G. B. Jones, R. S. Huber and B. J. Chapman, Tetrahedron:
Asymmetry, 1997, 8, 1797.
16 A. J. Poe, J. Organomet. Chem., 1975, 94, 235.
17 D. C. Bailey and S. H. Langer, Chem. Rev., 1981, 81, 109.
18 R. A. Dubois, P. E. Garrou, K. D. Lavin and H. R. Allcock,
Organometallics, 1986, 5, 460.
The resin ‘Novageldiphenylphosphine’ (NGDP) (250 mg,
0.53 mmol PgϪ1) was suspended in oxygen-free anhydrous THF
(5 cm3) and allowed to swell for 10 min under nitrogen
agitation. A solution of commercial octacarbonyldicobalt(0)
(226 mg, 0.66 mmol) in anhydrous, deoxygenated THF (5 cm3)
was added under nitrogen agitation and the dark mixture was
left at RT for 10 min. The resin beads were filtered, washed with
alternate aliquots of THF and Et2O until the filtrate became
colourless and dried in vacuo to afford deep purple beads;
{100% P site complexation, 60% ionic complex (26), 40% mono
complex (27), 0.47 mmol [Co2(CO)7] gϪ1}, νmax(Nujol)/cmϪ1
᎐
2074w, 2012msh, 1995s, 1955w and 1880vs (C᎐O); δ (145 MHz;
᎐
P
THF [D2O capillary lock]) 66 (polymer-PPh2–[Co], mono com-
plex (16), 40%), 75 (polymer-PPh2–[Co], ionic complex (15),
60%).
19 (a) G. O. Evans, C. U. Pittman Jr., R. McMillan, R. T. Beach and
R. Jones, J. Organomet. Chem., 1974, 67, 295; (b) C. De-An and
C. U. Pittman Jr., J. Mol. Catal., 1983, 21, 405.
20 A. R. Manning, J. Chem. Soc. (A), 1968, 1135.
21 P. Szabó, L. Fekete, G. Bor, Z. Nagy-Magos and L. Markó,
J. Organomet. Chem., 1968, 12, 245.
Application in the Pauson–Khand annelation
Resin [26 ؉ 27] was subjected to identical reaction conditions
to those used for entry 1 in Table 3 (see Protocol above
and Table 4) and was found to produce a conversion to
cyclopentenone 25 of 68%.
22 (a) W. Hieber and W. Freyer, Chem. Ber., 1958, 91, 1230;
(b) O. Vohler, Chem. Ber., 1958, 91, 1235.
23 (a) N. E. Schore and S. D. Najdi, J. Am. Chem. Soc., 1990, 112, 441;
(b) J. L. Spitzer, M. J. Kurth, N. E. Schore and S. D. Najdi,
Tetrahedron, 1997, 53, 6791.
24 (a) G. L. Bolton, Tetrahedron Lett., 1996, 37, 3433; (b) G. L. Bolton,
J. C. Hodges and J. R. Rubin, Tetrahedron, 1997, 53, 6611.
25 (a) F. E. Scully Jr. and K. Bowdring, J. Org. Chem., 1981, 46, 5077;
(b) W. Oppolzer, A. Pimm, B. Stammen and W. E. Hume, Helv.
Chim. Acta, 1997, 80, 623.
26 S. E. Gibson and A. Stevenazzi, Angew. Chem., Int. Ed., 2003, 42,
1800.
27 (a) D. B. Belanger, D. J. R. O’Mahoney and T. Livinghouse,
Tetrahedron Lett., 1998, 39, 7637; (b) B. L. Pagenkopf and
T. Livinghouse, J. Am. Chem. Soc., 1996, 118, 2285.
28 D. F. Shriver and M. A. Dredzon, The Manipulation of Air Sensitive
Compounds, Wiley, Chichester, 1986.
Acknowledgement
The authors are very grateful to AstraZeneca for the provision
of two fully funded studentships (to ACC and AS).
References
1 R. E. Sammelson and M. J. Kurth, Chem. Rev., 2001, 101, 137 and
references therein.
2 M. T. Reetz, Angew. Chem., Int. Ed., 2001, 40, 284 and references
therein.
3 A. C. Comely and S. E. Gibson, Angew. Chem., Int. Ed., 2001, 40,
1012 and references therein.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 5 9 – 1 9 6 8
1968