Palladium catalysed cyclisation-carbonylation of enynes to give cyclic γ,δ-unsaturated acids

Article abstract of DOI:10.1039/b300719g

In the presence of acetic acid, trifurylphosphine and CO (2 atm), palladium catalyses the conversion of a range of enynes to cyclic δ,γ-unsaturated carboxylic acids in good yield.

Full text of DOI:10.1039/b300719g

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Palladium catalysed cyclisation–carbonylation of enynes to give cyclic
g,d-unsaturated acids†
Varinder K. Aggarwal,*a Mike Buttersb and Paul W. Daviesa
a School of Chemistry, Bristol University, Cantock’s Close, Bristol, UK BS8 1TS.
E-mail: v.aggarwal@bristol.ac.uk; Fax: +44 (0)117 929 8611; Tel: +44 (0)117 954 6315
b AstraZeneca R&D Avlon Works, Severn Road, Bristol, UK BS10 7ZE
Received (in Cambridge, UK) 21st January 2003, Accepted 28th February 2003
First published as an Advance Article on the web 2nd April 2003
In the presence of acetic acid, trifurylphosphine and CO (2
atm), palladium catalyses the conversion of a range of
enynes to cyclic d,g-unsaturated carboxylic acids in good
yield.
Cycloisomeristion of linear enynes using Pd(0) and an acid to
give cyclic dienes is a well established process and has been
extensively exploited in total synthesis, most notably by Trost.1
Scheme 3 Competing pathways: (i) oxidative addition by Pd(0), (ii)
carbonylation then transformation of the acyl palladium species, (iii)
complexation, (iv) carbopalladation.
Although we had previously shown that bromodiene 9 could
be converted to the same unsaturated acid 8 through careful
choice of conditions that limited b-hydride elimination from 4
(Scheme 3),5 we felt that the use of enynes offered several
advantages:
(i) The starting materials are less functionalised and would be
easier to prepare.
(ii) Increased atom economy.
(iii) If the enyne was complexed to Pd (Scheme 1,
intermediate 2), premature capture of the vinyl palladium
species 10 to give the linear acid 11 could be avoided. This side
reaction (Scheme 3) plagued the synthesis of unsaturated acids
8 from bromodiene substrates which did not have a high
proclivity towards cyclisation.5
Scheme 1 Cycloisomerisation of enynes: (i) complexation, (ii) hydrido-
palladation, (iii) carbopalladation, (iv) b-H elimination.
A possible catalytic cycle for this process is shown in Scheme
1. Palladium hydride (PdH), which is initally formed from Pd(0)
and acetic acid,2 complexes with the enyne 1 to give 2.
Following hydridopalladation and carbopalladation the alkyl
palladium intermediate 4 is formed which finally undergoes b-
hydride elimination to give the diene product 5 and reforms the
active catalyst.
We began our investigations of the enyne cyclisation process
on 1a using conditions that had successfully been used in
A valuable extension of this elegant process would be to
couple the alkyl palladium intermediate 4 with another reagent
to generate even more functionalised products.3 We considered
the possibility of trapping 4 with CO as the acyl palladium
intermediate 6 could react further with AcOH4 to furnish the
anhydride 7 and return the active PdH catalyst (Scheme 2). Thus
the acetic acid would be required to have a dual role in the
catalytic process as a source of H+ to form the PdH catalyst and
as a nucleophile to trap the acyl palladium intermediate 6. It is
likely that the anhydride would hydrolyse to the corresponding
acid 8 under the reaction conditions.
Scheme 4 Schematic for Table 1.
Table 1 Solvent and acid effects on product distribution (Scheme 4)
Yields (%)b
Entry Substrate Solvent Conditionsa Additive
8
1
12
1
2
3
4
5
6
1a
1b
1b
1b
1b
1b
DMF
DMF
DME
Dioxane
CH3CN
DME
A
B
B
B
B
B
—
—
—
—
—
17
10
55
32
—
76
78
23
8
—
—
—
—
< 10 32
14 18
TFA (1 eq.) 57
a All reactions were run at 80 °C for 24 h under 2 atm pressure of CO in the
presence of (A) Pd2(dba)3·CHCl3 (2.5 mol%), P(2-furyl)3 (0.3 eq.), HOAc
(2 eq.) at a substrate concentration of 0.25 M, (B) Pd(OAc)2 (10 mol%), P(2-
furyl)3 (60 mol%) and HOAc (10 eq.) at a substrate concentration of 0.33 M.
b Determined by NMR against an internal standard.
Scheme 2 Desired process: (i) carbonylation, (ii) transformation of
acylpalladium, (iii) hydrolysis.
† Electronic supplementary information (ESI) available: data for novel
compounds. See http://www.rsc.org/suppdata/cc/b3/b300719g/
1046
CHEM. COMMUN., 2003, 1046–1047
This journal is © The Royal Society of Chemistry 2003

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
these optimised conditions were applied to a number of
1,6-enynes good yields of the desired cyclised products were
obtained (Table 2, entries 1 to 3). Application to enynes bearing
non-terminal alkynes was not successful and starting material
was largely reisolated (entries 4 and 5). The 1,7-enyne 1f gave
the desired product 8g in reasonable yield together with 20% of
the linear product 11g (entry 6). This compares with the
bromodiene methodology in which the corresponding bromo-
diene 14 (Fig. 1) gave the acid products in 90% combined yield
as a 4.6+1 mixture of the linear to cyclic product. A less highly
substituted 1,7-enyne 1g gave linear acid 11g as the pre-
dominant product with only a small amount of 8g (entry 7)
indicating the limitation to the substrates which can be
employed.
There are two possible linear starting materials which lead to
cyclic g,d-unsaturated acids via the common palladium inter-
mediate 3: bromodiene 9 or enyne 1. The difference between the
two processes in relation to the ratio of cyclic+linear acids
formed appears to result from the method of formation of the
key vinyl palladium intermediate 3. Once this intermediate is
formed carbopalladation can ensue and the cyclic acid will be
produced. In the case of the bromodienes intermediate 3 is
reached only after formation of vinyl palladium species 10,
allowing carbonylation to compete with association and
cyclisation. In the case of enynes both the alkyne and alkene are
proposed to coordinate to the metal before addition of the PdH
across the alkyne and hence the vinyl palladium species 3 is
formed directly and not via intermediate 10. For premature
carbonylation to occur from the enyne substrate, dissociation of
the alkene from the palladium centre (3 ? 10) must proceed at
a rate comparable with carbopalladation. This clearly does not
occur with 1,6-enynes where carbopalladation is fast but does
occur with 1,7-enynes, resulting in formation of significant
quantities of linear acids.
Fig. 1
bromodiene cyclisation–carbonylation [Pd2(dba)3·CHCl3 (2.5
mol%), P(2-furyl)3 (0.3 eq.), NaOAc (2 eq.), DMF, 80 °C, CO
(2 atm)] except that NaOAc was replaced by AcOH. This first
attempt was indeed partially successful: the cyclised product
was formed, albeit in low yield, without formation of the linear
acid but a large quantity of starting material remained (Table 1,
entry 1). In fact when these conditions were employed with the
corresponding bromodiene 13 (Fig. 1), a 1+1 ratio of linear and
cyclised acids was formed.5 This highlights the considerable
advantage of the palladium complexed enyne intermediate 2, as
this clearly avoids premature capture of the vinyl palladium
intermediate 10 by CO. However the yield was low and large
amounts of starting material remained. In fact when these
conditions were applied to 1b, very little conversion occurred.
As 1b showed even less tendency to react, further investigations
were directed towards this substrate so that we could arrive at a
more general solution. Variation in source of Pd, amount of acid
and concentration finally led to a set of conditions (B) which did
indeed furnish a small amount of the desired product, but again
a large amount of starting material 1b remained (Table 1, entry
2). To promote the reaction, we needed to promote complexa-
tion of the PdH with the enyne, the first step in the cascade
process. We felt that DMF might be coordinating strongly to
Pd,2 thus preventing enyne complexation, and so we sought less
polar/coordinating solvents (Table 1). Indeed, changing from
DMF to DME produced a considerable improvement in yield of
the desired product (entry 3). Dioxane was also effective but
several side-products were observed (entry 4) and MeCN
changed the course of the reaction to give the diene 12 (entry
5).6 Attempts to increase the reactivity of the PdH species by
using a stronger acid did lead to increased conversion of the
enyne but also resulted in increased quantities of diene 12
(Table 1, entry 6).
In conclusion we have developed a novel cyclisation–
carbonylation–capture process for the conversion of enynes into
g,d-unsaturated cyclic acids in good yields. This study also
demonstrates that enynes may well be superior substrates to
bromodienes for cascade processes as one possible competing
pathway (premature capture) is severely retarded.
The authors thank AstraZeneca plc for support of a
studentship (PD).
Further optimisation of the enyne reaction in DME was
performed and it was found that changing from Pd(OAc)2 to
Pd2(dba)3·CHCl3 and increasing the amount of acid led to
complete consumption of enynes 1a/b (Table 2). Indeed when
Notes and references
1 B. M. Trost, Chem. Eur. J., 1998, 4, 2405; B. M. Trost, D. L. Romero and
F. Rise, J. Am. Chem. Soc., 1994, 116, 4268; G. C. Lloyd-Jones, Org.
Biomol. Chem., 2003, 1, 215.
2 Pd(PPh3)4 and HOAc have been shown to react in DMF to give a cationic
HPd(DMF)2+ species with an acetate counterion: C. Amatore, A. Jutand,
G. Meyer, I. Carelli and I. Chiaretto, Eur. J. Inorg. Chem., 2000, 1855.
3 When generated from an enyne, 4 has previously been trapped with: (a)
a hydride source such as PMHS to effect a cyclisation reduction strategy:
B. M. Trost and F. Rise, J. Am. Chem. Soc., 1987, 109, 3161; (b) an
organostannane to effect a cyclisation cross-coupling strategy: H.
Yamada, S. Aoyagi and C. Kibayashi, Tetrahedron Lett., 1997, 38,
3027.
Scheme 5 Schematic for Table 2.
4 K. Nagira, K. Kikukawa, F. Wada and T. Matsuda, J. Org. Chem., 1980,
45, 2365; G. P. Chiusoli and A. Cameroni, Chim. Ind. (Milan), 1964, 46,
1063.
5 V. K. Aggarwal, P. W. Davies and W. O. Moss, Chem. Commun., 2002,
972.
Table 2 Application of reaction conditions (Scheme 5)
Yields (%)a
Entry Substrate
X
n
R
R1
8
11
6 A likely pathway for the formation of diene 12 is shown in Scheme 6:
1
2
3
4
5
6
7
1a
1b
1c
1d
1e
1f
NTs
1
1
1
1
1
2
2
H
H
H
Me
TMS
H
H
H
Me
H
H
68
72
77
Trace
—
—
—
—
—
—
20
53
C(CO2Et)2
C(CO2Et)2
NTs
NTs
C(CO2Et)2
NTs
H
H
45
11
1g
H
Scheme 6 Formation of 12: (i) complexation, (ii) carbopalladation, (iii)
rearrangement, (iv) b-H elimination.
Reagents and conditions: All reactions were run at 80 °C for 38 h under 2
atm. pressure of CO in the presence of Pd2(dba)3·CHCl3 (5 mol%), P(2-
furyl)3 (60 mol%) and HOAc (15 eq.) at a substrate concentration of 0.33 M
in DME.a Purified by column chromatography.
See E. Negishi, C. Copéret, S. Ma, T. Mita, T. Sugihara and J. M. Tour,
J. Am. Chem. Soc., 1996, 118, 5904.
CHEM. COMMUN., 2003, 1046–1047
1047