Polymer supported cobalt carbonyl complexes as novel traceless alkyne linkers for solid-phase synthesis

Article abstract of DOI:10.1039/a906048k

The immobilisation of functionalised alkynes onto 'polymer-bound triphenylphosphine', their chemical manipulation and subsequent release has been demonstrated for the first time, thus illustrating that cobalt carbonyl complexes can be used as 'traceless' π-alkyne linkers.

Full text of DOI:10.1039/a906048k

Polymer supported cobalt carbonyl complexes as novel traceless alkyne linkers
for solid-phase synthesis
a
a
b
Alex C. Comely,* Susan E. Gibson (n e´ e Thomas) and Neil J. Hales
a
Department of Chemistry, King’s College London, Strand, London, UK WC2R 2LS. E-mail: susan.gibson@kcl.ac.uk
AstraZeneca, Mereside, Alderley Park, Macclesfield, UK SK10 4TG
b
Received (in Liverpool, UK) 22nd July 1999, Accepted 1st September 1999
The immobilisation of functionalised alkynes onto ‘polymer-
bound triphenylphosphine’, their chemical manipulation
and subsequent release has been demonstrated for the first
time, thus illustrating that cobalt carbonyl complexes can be
used as ‘traceless’ p-alkyne linkers.
conversion of (1 + 2) into 3 and complexation of an alkyne
becomes a feasible transformation (Scheme 2). Treatment of the
resin bearing complexes 1 and 2 with hex-5-yn-1-ol in
1,4-dioxane at 70 °C generated the purple resin-bound alkyne
complexes 4, presumably via the intermediary carbonyl com-
plex 3. Comparison of overlapping sets of IR absorptions with
Solid-phase synthesis continues to excite ever-greater interest,1
particularly as a means to facilitate the elaboration of compound
literature data, and data from samples of Ph
11
3
P(OC)
2
CoCo-
(CO) (hex-5-yn-1-ol) and [Ph P(OC) Co] (hex-5-yn-1-ol) pre-
3 3 2 2
2
libraries via combinatorial chemistry. Of especial importance
pared ourselves, indicates the presence of both mono- and
bisphosphine substituted alkyne complexes, 4a and 4b, and that
the latter is the major component.‡ The 31P NMR spectrum of
4a/4b supports alkyne complexation (with a 20% phosphorus
site occupancy).
to this technology is the linker, the structural motif which joins
the substrate under chemical manipulation to the polymeric
support. A legacy of solid-phase peptide synthesis is the release
of compounds bearing carboxylic acid or amide functionality
derived from an ester linkage. While appropriate for some target
molecules, this has rather limited appeal in a more general
synthetic sequence. Efforts to address this problem have
Acetylation of resin 4 to afford 5 using acetic anhydride–
3
NEt is supported by an additional acetate absorption at 1734
2
1
cm . Decomplexation was achieved by aerial oxidation in
CH Cl under white light for 72 h. Filtration and washing of the
3
stimulated the evolution of ‘traceless’ linkers, those which
2
2
leave minimal vestige of the solid support upon release of the
product. A good example is the tethering of compounds
containing an aromatic ring by means of an aryl–silicon bond.4
The linker is cleaved to reveal only a hydrogen atom at the
former aromatic linkage site.
brown polymeric residue delivered hex-5-yn-1-yl acetate as the
sole product in 70 ± 20% yield from 4.
A second approach with the same overall objective involved
the reaction of ‘polymer-bound triphenylphosphine’ with a
preformed alkyne complex (indirect loading) (Scheme 2).
Designed to permit comparison between the two routes, hex-
In this respect, transition metals offer a very alluring solution:
the temporary and reversible immobilisation of an unsaturated
substrate via p-interactions to the molecular scaffold dispenses
with the functional group transformation necessarily required
by a covalent system. In recent months this concept was
illustrated for the first time with the immobilisation of arenes
2 8
5-yn-1-ol was complexed with Co (CO) to give 6, a deep red
oil (85%). Loading onto the polymer was effected in THF at 50
°C to afford 4§ possessing similar IR characteristics to the resin
prepared by the direct method. A comparison of the strengths of
1
1
bands attributed to the mono- and bisphosphine derivatives
5
identifies the former as the major component via this route.‡ 31P
NMR spectroscopy is extremely valuable here.¶ A well-
resolved resonance integrates to an 80% complexation of
phosphorus sites (the benign phosphine oxide, polymer-
through a chromium carbonyl linker. Liberation of the product
by oxidative decomplexation returns the aromatic ring un-
changed.
Surprisingly, there currently exists no method for the release
6
of an alkyne from a polymeric support. A system which allows
2
P(O)Ph , occupying the remaining 20%) and a loading of 0.94
2
1
the addition and later release of this functionality, without a
compromise to its integrity, would thus represent a valuable
extension to linker technology. The chemistry of cobalt
carbonyl alkyne complexes is well established in the solution
± 0.02 mmol[hexynol] g . Acetylation proceeded as before:
that the metal carbonyl stretches and the 31P NMR spectrum
undergo no change demonstrates the stability of this linker to
these conditions. Hex-5-yn-1-yl acetate (12 mg, 60 ± 10% from
4) was recovered from the resin following acetylation of resin 4
(150 mg) and decomplexation over 72 h.
7
phase and has been applied to great effect in organic synthesis:
7a
7b
alkyne protection, the Nicholas reaction and Pauson–Khand
cyclisations,7 for example, are amply documented. Given the
ease of formation and tolerance to diverse reaction conditions of
these complexes, it appeared to us that cobalt carbonyl species
offered considerable potential as alkyne linkers for solid-phase
synthesis.
c
Each loading technique has its own advantage: overall yields
via indirect loading are significantly higher than those obtained
from direct loading. The latter, however, obviates pre-formation
of a cobalt alkyne complex and thus offers convenience: once
Our initial investigations involved the reaction of ‘polymer-
bound triphenylphosphine’† with Co (CO) in THF at room
2 8
temperature to generate a cobalt carbonyl resin, a support to
which alkynes could be added directly (Scheme 1). The purple,
stable resin was characterised on the basis of its IR and 31
P
NMR spectra: strong absorptions ascribed to ionic 1 were
accompanied by weaker peaks indicating a minor presence of
8
the monophosphine-substituted complex 2. Heating of this
resin at 60 °C in 1,4-dioxane cleanly converted it into a second
form, assigned the structure of the neutral bisphosphine 3 on the
basis of its IR and 31P NMR spectra.9
The alkyne complexation of a phosphine-substituted cobalt
3 3 2
P)Co(CO) ]
, has been reported.10 In
carbonyl complex, [(Bu
view of the elevated temperature required, the simultaneous
Scheme 1
Chem. Commun., 1999, 2075–2076
This journal is © The Royal Society of Chemistry 1999
2075

Scheme 4
In conclusion, we have demonstrated for the first time the
utility of cobalt carbonyl complexes as linkers for the ‘traceless’
immobilisation and subsequent release of alkyne substrates
(
Scheme 4). The linker is stable to certain widely-used reaction
conditions and the work described herein is applicable to
various combinatorial methods.
The authors thank Jon Cobb for invaluable NMR assistance
and AstraZeneca for a studentship (ACC).
Footnotes and references
†
‘Polymer-bound triphenylphosphine’ (commercially available from
2
1
Fluka, ~ 1.6 mmol P g ) describes a diphenylphosphino polystyrene
polymer crosslinked with 2% divinylbenzene.
‡
Details of estimates used in this study are as follows: 4a+4b ratios were
estimated from IR intensities (3+7 and 6+4 for direct and indirect routes
respectively). Solution calibration studies verify this estimate. Solid phase
31
P NMR spectra can be accurately integrated but show no resolution of
mono- and bisphosphine derivatives. The error margins for yield and
31
loading calculations are estimated as follows: IR spectra, 5%; P NMR
spectra, direct method, 10%; 31P NMR spectra, indirect method 5%.
§
The experimental procedure for indirect formation of 4 is as follows:
polymer-bound triphenylphosphine (1 g, ~ 1.6 mmol P) was suspended in
3
oxygen-free anhydrous THF (10 cm ) and allowed to swell for 30 min under
nitrogen agitation. A solution of hexacarbonyl(hex-5-yn-1-ol)dicobalt(0)
3
complex (1.2 g, 3.2 mmol) in anhydrous THF (5 cm ) was added and the
black suspension was heated to 50 °C for 4 h. After cooling, the resin beads
were filtered, washed with alternate aliquots of THF and Et
2
O until the
filtrate became colourless, and dried in vacuo to afford deep purple beads of
2
1
4
¶
(1.46 g, 80% P site occupancy, 0.94 ± 0.02 mmol[hexynol] g ).
Polymer samples for 31P NMR were swollen in THF and scanned with a
D
2
O capillary lock at 145.7 MHz for 2 h. Different phosphine species are
-[Co], d = 50–80; polymer-P(O)Ph
= 26.3.
Scheme 2
readily distinguished: polymer-PPh
2
P
2
,
d
P
2 P
= 28.8; polymer-PPh , d
prepared, the resin-linker complex (1 + 2) can be stored and
used as required.
Finally, in preliminary experiments to explore the versatility
of this linker, alcohol 4 (prepared by direct loading) was
1
A. R. Brown, P. H. H. Hermkens, H. C. J. Ottenheijm and D. C. Rees,
Synlett, 1998, 817.
F. Balkenhohl, C. von dem Bussche-H u¨ nnefeld, A. Lansky and C.
Zechel, Angew. Chem., Int. Ed. Engl., 1996, 35, 2288.
2
12
oxidised in the presence of SO
3
·py-activated DMSO and NEt
3
3 P. H. H. Hermkens, H. C. J. Ottenheijm and D. C. Rees, Tetrahedron
Report No. 418, Tetrahedron, 1997, 53, 5643.
4 M. J. Plunkett and J. A. Ellman, J. Org. Chem., 1997, 62, 2885.
5 S. E. Gibson (n e´ e Thomas), N. J. Hales and M. A. Peplow, Tetrahedron
Lett., 1999, 40, 1417.
to afford the aldehyde complex 7 (Scheme 3), displaying a
2
1
diagnostic aldehyde stretch at 1723 cm . Significantly, all
other data (IR and 31P NMR) of resin 7 are unchanged from
those of 4, thus demonstrating the linker’s stability to these
oxidising conditions. Hex-5-yn-1-al was isolated in 17 ± 5%
yield after decomplexation as before (the low yield here is
attributed to the volatility of the aldehyde).
6
Alkynes have been immobilised via Michael addition but are liberated
as enones, N. W. Hird, K. Irie and K. Nagai, Tetrahedron Lett., 1997, 38,
7
111; or ketones, N. W. Hird, M. Crawshaw, K. Irie and K. Nagai,
Tetrahedron Lett., 1997, 38, 7115; and the cyclocondensation of
acetylenic ketones with a resin-bound thiouronium salt has generated
pyrimidine heterocycles, D. Obrecht, C. Abrecht, A. Grieder and J. M.
Villalgordo, Helv. Chim. Acta., 1997, 80, 65.
7
(a) K. M. Nicholas and R. Pettit, Tetrahedron Lett., 1971, 3475; (b) T.
Nakamura, T. Matsui, K. Tanino and I. Kawajima, J. Org. Chem., 1997,
6
2, 3032; (c) for a recent review of the asymmetric reaction: O. Geis and
H. G. Schmaltz, Angew. Chem., Int. Ed., 1998, 37, 911.
R. A. Dubois, P. E. Garrou, K. D. Lavin and H. R. Allcock,
Organometallics, 1986, 5, 460.
8
9
C. De-An and C. U. Pittman Jr., J. Mol. Catal., 1983, 21, 405.
1
1
0 A. J. Poe, J. Organomet. Chem., 1975, 94, 235.
1 For a representative monophosphine derivative see: G. V a´ radi, A. Vizi-
Orosz, S. Vastag and G. P a´ lyi, J. Organomet. Chem., 1976, 108, 225;
and for a bisphosphine derivative see, M. F. D’Agostino, C. S. Frampton
and M. J. McGlinchey, Organometallics, 1990, 9, 2972.
1
2 A. M. Fivush and T. M. Willson, Tetrahedron Lett., 1997, 38, 7151.
Scheme 3
Communication 9/06048K
2076
Chem. Commun., 1999, 2075–2076