Solid-phase synthesis of pyrrolidines employing a cyclisation-cleavage strategy

Article abstract of DOI:10.1039/a904135d

The solid-phase synthesis of several pyrrolidines has been realised utilising imino-Sakurai and palladium-catalysed cyclisation-cleavage reactions as key steps.

Full text of DOI:10.1039/a904135d

Solid-phase synthesis of pyrrolidines employing a cyclisation–cleavage strategy
Richard C. D. Brown* and Martyn Fisher
Department of Chemistry, University of Southampton, Highfield, Southampton, UK SO17 1BJ.
E-mail: rcb1@soton.ac.uk
Received (in Liverpool, UK) 20th May 1999, Accepted 28th June 1999
The solid-phase synthesis of several pyrrolidines has been
realised utilising imino-Sakurai and palladium-catalysed
cyclisation–cleavage reactions as key steps.
Before attempting any solid-phase work the analogous
solution chemistry was investigated (Scheme 3). 2-Hydroxy-
methyl-3-trimethylsilylpropene 6 was esterified with 2-methyl-
3-phenylpropanoyl chloride to give the allylic silane 7a.
Addition of 7a to the preformed N-acyl imine 8 provided a
satisfactory yield of the homoallylic amine 9a. Attempted
cyclisation of either the Boc-protected homoallylic amine 9a or
the corresponding deprotected primary amine failed using 10
mol% Pd(acac)2 with 15 mol% diphenylphosphinoethane
(dppe). However, deprotection and reductive alkylation of 9a
provided secondary amine 10a which when subjected to the
same cyclisation conditions afforded the desired pyrrolidine 11
in 35% yield. Although some of the yields in the solution
pyrrolidine synthesis were rather low, we were sufficiently
satisfied with the overall approach to carry out further
optimisation on the solid phase.
The development of solid-phase combinatorial and multiple
parallel synthesis methods has stimulated considerable interest
in the design of new linkers and cleavage strategies which are
suitable for the release of small molecules from the solid
support.1–4 Linkers or cleavage conditions which activate a
polymer-bound molecule towards nucleophilic cleavage can be
particularly useful providing an additional point of variability in
the released compounds,4 or allowing the formation of a ring
upon cleavage.4,5 Here we describe the development of a novel
nucleophile-cleavable linker that is sufficiently robust to
survive a multi-step reaction sequence, yet can be cleaved under
mild conditions with incorporation of a nucleophile.
Several linkers which rely upon the palladium-catalysed
cleavage of allylic systems have been reported, releasing
carboxylic acids or amines (through an intermediate carbamic
acid) from the resin.6–9 We imagined that a reversed allylic
linker would provide a mild means of generating electrophilc p-
allyl palladium species which could be trapped with heteroa-
tom- or carbon-centred nucleophiles to release cyclic 2 or
acyclic 3 products from the solid phase (Scheme 1).
In order to demonstrate the viability of a palladium-catalysed
cyclisation–cleavage reaction we chose to investigate the solid-
phase synthesis of pyrrolidines 4 (Scheme 2).10,11 A key
intermediate is the homoallylic amine 5 which could be derived
from the reaction of a resin-bound allylic nucleophile with an
imine.11–13 The presence of an ester group in the linker meant
that the reactivity of the allyl metal component would have to be
attenuated, leading us to consider the use of an allylic silane as
the immobilised nucleophile.
Solid phase synthesis of pyrrolidine 11 started with a
carboxyethylated polystyrene resin† 1 which was prepared in
three steps from Merrifield resin. The 2-hydroxymethyl-
3-trimethylsilylpropene 6 was immobilised under standard
carbodiimide coupling conditions using an excess of DMAP to
prevent the well-known side reaction leading to a resin-bound
N-acylurea. The imino-Sakurai reaction required some optimi-
sation, turning out to be the most challenging step in the solid-
phase reaction sequence. The effect of changing several
reaction parameters including reaction time, amount of N-
acylimine and the amount of Lewis acid was investigated. The
efficiency of the reaction was measured by reduction of the ester
linkage followed by purification of the cleaved alcohol 12 and
quantification (Scheme 4). Our results are shown in Table 1.
Optimum conditions required a large excess of the N-acylimine
8 and an excess of BF3•OEt2 relative to the resin-bound
allylsilane 7b (entry 7), providing significantly better yield of
the homoallylic amine than was achieved in the analogous
reaction in solution.
Scheme 1
Scheme
3 Reagents and conditions: i, 1,3-diisopropylcarbodiimide,
DMAP, CH2Cl2, 1; or 2-methyl-3-phenylpropionyl chloride, CH2Cl2,
pyridine; ii, BF3•OEt2, CH2Cl2; iii, TFA, CH2Cl2; iv, PhCHO, AcOH,
CH2ClCH2Cl; v, Me4NB(OAc)3H, CH2ClCH2Cl; vi, Pd(acac)2, dppe, THF,
D.
Scheme 2
Chem. Commun., 1999, 1547–1548
1547

Scheme 4 Reagents and conditions: i, 8, BF3•OEt2, CH2Cl2; ii, LiBH4,
MeOH (1 equiv.), THF.
Table 1 Effect of reaction parameters on yield
Molar ratioa of Molar ratioa of Reaction
Yielda
(%)
Entry
N-acylimine 8
BF3•OEt2
time/h
1
2
3
4
5
6
7
8
8
12
4
8
8
8
12
12
12
28
12
3
16
25
61
49
70
70
70
the desired product over the release of by-products from earlier
steps. To test this theory further, several of the possible by-
products from the deprotection–reductive alkylation sequence
9b, 19 and 20 were subjected to the cyclisation–cleavage
reaction conditions. No significant quantity of material was
released from any of the resins 9b, 19 or 20. Interestingly,
treatment of the resin 9b recovered from the attempted
palladium-catalysed cleavage with LiBH4 in THF afforded a
reduced product 21 in 80% yield rather than the expected allylic
alcohol 12.
15
15
23
23
23
8
4
3
a
The yield represents the amount of purified isolated alcohol 12 relative
to the loading of the carboxylated resin 1. Molar ratios are also based upon
the loading of the carboxylated resin 1.
In summary, the viability of a palladium-catalysed cyclisa-
tion–cleavage reaction has been demonstrated by the solid-
phase synthesis of pyrrolidines 11 and 13–17. This strategy
should also be amenable to the solid-phase synthesis of other
heterocyclic and carbocyclic compounds. Intermolecular palla-
dium-catalysed nucleophilic cleavage, using LiBH4 as a source
of hydride, is also possible releasing substituted propene
derivatives such as 21.
We wish to thank the Royal Society for a University Research
Fellowship (R. C. D. B.), and the EPSRC for a QUOTA
studentship (M. F.). We wish to acknowledge the use of the
EPSRC’s Chemical Database Service at Daresbury.14
Removal of the carbamate protecting group from 9b and a
two-step reductive alkylation using benzaldehyde gave the
resin-bound cyclisation precursor 10b, which was also quanti-
fied by reductive cleavage from the resin. Initial attempts to
perform the cyclisation–cleavage reaction of 10b gave the
desired heterocycle 11 in low yield, even when a large quantity
of palladium complex was employed. However, efficient
catalytic cyclisation-cleavage was achieved by employing a
three-fold excess of the dppe ligand relative to palladium (10
mol% based upon the loading of the carboxylated resin 1),
returning an excellent yield of the desired pyrrolidine 11.‡
In order to demonstrate the scope of the palladium-catalysed
cyclisation–cleavage reaction, pyrrolidines 11 and 13–17 were
synthesised.§ In all instances the desired pyrrolidine was
obtained, furthermore the purity of the crude product, except in
Notes and references
† Merrifield resin was heated with the sodium salt of diethyl malonate at
60 °C for 14 h. Hydrolysis was carried out in refluxing THF–2 M KOH
(9+1). Finally, decarboxylation in THF–HCl gave the desired carboxyethy-
lated polystyrene 1.
1
the case of 13, was judged to be high by inspection of the H
NMR spectra. Notably, the crude pyrrolidines 11, 13–17 were
‡ Yields have been calculated based on the loading of the carboxylated
polystyrene resin 1.
§ All reactions were carried out on at least a 50 mM scale with respect to the
loading of the carboxylated polystyrene resin 1. All products were purified
by flash chromatography and gave satisfactory analytical data.
1 S. Booth, P. H. H. Hermkens, H. C. J. Ottenheijm and D. C. Rees,
Tetrahedron, 1998, 54, 15 385.
2 R. C. D. Brown, J. Chem. Soc., Perkin Trans. 1, 1998, 3293.
3 H.-M. Eggenweiler, Drug Discovery Today, 1998, 3, 552.
4 I. W. James, Tetrahedron, 1999, 55, 4855.
5 J. H. van Maarseveen, Comb. Chem. High Throughput Screening, 1998,
1, 185.
6 F. Guibé, O. Dangles, G. Balavoine and A. Loffet, Tetrahedron Lett.,
1989, 30, 2641.
7 K. Kaljuste and A. Undén, Tetrahedron Lett., 1996, 37, 3031.
8 H. Kunz and B. Dombo, Angew. Chem., Int. Ed. Engl., 1988, 27, 711.
9 X. Zhang and R. A. Jones, Tetrahedron Lett., 1996, 37, 3789.
10 B. M. Trost and P. J. Bonk, J. Am. Chem. Soc., 1985, 107, 1778.
11 B. M. Trost and C. M. Marrs, J. Am. Chem. Soc., 1993, 115, 6636.
12 S. J. Veenstra and P. Schmid, Tetrahedron Lett., 1997, 38, 997.
13 W. J. N. Meester, F. Rutjes, P. H. H. Hermkens and H. Hiemstra,
Tetrahedron Lett., 1999, 40, 1601.
cleaner than their acyclic precursors which were released from
the resin by reductive cleavage of 18 using LiBH4. This
significant result suggests that the cyclisation–cleavage reaction
acts as an in-built quality control step, favouring the release of
14 D. A. Fletcher, R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput.
Sci., 1996, 36, 746.
Communication 9/04135D
1548
Chem. Commun., 1999, 1547–1548