Highly regio- and chemoselective palladium(0)-mediated allylic substitution of difunctional allylic halides with phenols

Article abstract of DOI:10.1016/j.tetlet.2010.09.125

An efficient Pd(0)-mediated, base-assisted reaction of phenols with difunctionalised allylic halides results in the formation of coupled products in good yields. The reactions proceed with excellent stereo-, regio- and chemocontrol. An appropriately funct

Full text of DOI:10.1016/j.tetlet.2010.09.125

Tetrahedron Letters 51 (2010) 6378–6380
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Highly regio- and chemoselective palladium(0)-mediated allylic substitution
of difunctional allylic halides with phenols
Anant R. Kapdi a,b, , Richard J. K. Taylor a, Ian J. S. Fairlamb a,
⇑
⇑
a Department of Chemistry, University of York, York YO10 5DD, United Kingdom
b Department of Chemistry, Ramnarain Ruia College, Matunga, Mumbai 400 019, India
a r t i c l e i n f o
a b s t r a c t
Article history:
An efficient Pd(0)-mediated, base-assisted reaction of phenols with difunctionalised allylic halides results
in the formation of coupled products in good yields. The reactions proceed with excellent stereo-, regio-
and chemocontrol. An appropriately functionalised Weinreb amide, synthesised by this methodology,
undergoes halogen-lithium exchange and subsequent intramolecular 1,2-carbonyl addition/elimination
to give an advanced macrocyclic intermediate with potential use in the synthesis of likonide B.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 19 August 2010
Revised 6 September 2010
Accepted 24 September 2010
Keywords:
Pd(0)-mediated reactions
Allylic substitution
Bifunctional allylic halides
Regioselective
Chemoselective
Difunctional allylic sesquiterpenes and monoterpenes (Scheme
1, 1 and 2, respectively) are useful synthetic building blocks, which
have been successfully incorporated into various biologically active
and medicinally relevant farnesyl1,2 and geranylgeranyl deriva-
tives.3 Several synthetic methods have been described in the litera-
ture for accessing these types of difunctionalised allylic systems.4,5
We envisaged that a difunctionalised terpene containing different
allylic leaving groups (e.g., halide and acetate) would be amenable
to Pd(0)-mediated allylic substitution processes.6 This transforma-
tion requires the chemodifferentiation of two unique allyl leaving
groups by Pd(0) (exemplified by leaving groups X and Y in generic
compound2 in Scheme 1). Suchchemodifferentiation would depend
on: (i) the relative ease of oxidative addition (rate of C–X bond
cleavage) for the reaction of Pd(0) with 2 to give either 3 or 30;
(ii) the stability of the oxidative addition products, strength of
‘Pd(II)–X’ versus ‘Pd(II)–Y’ bonds; (iii) any reversible oxidative
addition reaction.
Pd(0)-catalysed allylic substitution leads to highly stereoselec-
tive substitution reactions and is unique for several reasons.6b It
is often possible to attain both high stereoselectivity and regiose-
lectivity, but chemoselectivity in reactions containing two types
of allylic leaving groups in an acyclic system has not been well
investigated.7 Herein, we report a novel Pd(0)-catalysed allylic
substitution of difunctionalised terpenyl halides, which proceeds
in a highly stereo-, regio- and chemoselective manner. Suitable
difunctionalised geranyl systems were identified as benchmark
substrates for the evaluation of Pd(0)-catalysed allylic substitution
reactions. For example, 8-hydroxygeranyl acetate 4 could be
converted into halo-substituted geranyl acetates 5a and 5b
(Scheme 2), using conventional procedures,8 in good yields with
high stereoselectivity (especially in the case of chloride; note that
the iodide compound 5a is light sensitive!). The presence of
distinct functional groups allows the chemodifferentiation in the
Pd(0)-catalysed allylic substitution to be tested.
Initial experiments were performed with both 5a and 5b and
4-methoxyphenol 6 (as the nucleophile) with NaH (base) under
Pd(0)-catalysed conditions. It was found that only halide displace-
ment was observed with allyl iodide 5a giving higher regioselectiv-
ity (a-7 was the only observed regioisomer) as compared to allyl
chloride 5b (where a ratio of 2.5:1,
a
-7/ -7, was obtained). At
c
the outset of our studies using 5a, various phosphine ligands were
screened in combination with [Pd02(dba-3,5,30,50-OMe)3] as the
Pd(0) source.9–11 The Pd(0) catalyst systems derived from simple
phosphine ligands 8a–d displayed higher reactivity (Table 1,
entries 1–4).12 Complete stereo- and regiocontrol was made possi-
ble by using such simple ligand systems, with 8c identified as the
most effective. Unfortunately, bulky electron-rich ligands 8e–g
(Table 1, entries 5–7) and electron-deficient ligand 8h (entry 8)
gave poor results also bringing about a reduction in the stereose-
lectivity. The allylic substitution reaction with phenols and base
in the absence of Pd(0) proceeds with poor regio- and stereocontrol
(lower yields were obtained too), highlighting the importance of
the developed protocol with catalytic Pd(0) (entry 9).
⇑
Corresponding authors.
E-mail address: ijsf1@york.ac.uk (I.J.S. Fairlamb).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.09.125

A. R. Kapdi et al. / Tetrahedron Letters 51 (2010) 6378–6380
6379
Scheme 1. Difunctionalised allylic derivatives, and proposed chemodifferentiation of allylic halide and acetate leaving groups by Pd(0).
Scheme 2. Synthesis of bifunctional geranyl acetates 5a and 5b.
With an effective catalytic system in hand, using 8c as the li-
gand of choice, we explored its scope in the coupling of a more ste-
rically hindered phenol 9a with 5a to give -10 in 93% yield and
Scheme 3. Evaluation of a highly substituted phenol coupling partner 9a.
a
16:1 stereoselectivity (Scheme 3). Phenol 9a contains a labile silyl
protecting group, which was tolerated in this reaction highlighting
the mildness of the synthetic protocol.
mediated allylic substitution protocol (Scheme 4). In line with
the earlier results, using 8c as the ligand, complete regio- and ste-
reocontrol (from allylic starting material 11 to product, either
a-
Farnesyl pyrophosphates (FPP) are important intermediates
used by organisms for the biosynthesis of various classes of natural
products.13 Likonides A and B are one such class, which possess an
embedded ‘farnesyl’ motif (vide infra).14–16 Indeed, we recognised
that the Pd(0)-mediated allylic substitution protocol could be used
in the synthesis of potentially useful intermediates to these com-
pounds. Accordingly, a functionalised farnesyl iodide 11 was cou-
pled with both substituted phenols 9a and 9b using the Pd(0)-
12a or -12b) was observed. In these reactions, neither the vinyl
a
ester nor the aromatic bromides caused any problems in terms of
the catalysis. Synthetically more relevant functional groups could
also be incorporated via this methodology. Accordingly, a farnesyl
iodide containing a Weinreb amide functionality 13 was reacted
with substituted phenol 9a giving the coupled product
a-14 with
complete regio- and stereocontrol (Scheme 4).
Table 1
Pd(0)-catalysed allylic alkylation catalyst screening
Entry
Ligand {P(Ar-X)3}
E:Z ratioa
Yield a
-7b (%)
1
PPh3
8a
8b
8c
8d
8e
8f
8g
8h
—
18:1
18:1
18:1
15:1
15:1
15:1
12:1
15:1
3:1
79
89
94
81
54
68
32
84
59
2
P(4-FC6H4)3
P(4-MeOC6H4)3
P(4-MeC6H4)3
P(2,4,6-Me3C6H2)3
PCy3
3c
4
5
6
7
t-Bu3PꢀHBF4
8
P(OPh)3
9d
—
a
b
c
E:Z ratio calculated from the 1H NMR spectra of the crude products (>99%
a-7b; c-7b not observed).
Isolated yield following purification by flash chromatography on silica gel with the product showing complete chemoselectivity for iodide over acetate.
Under otherwise identical conditions, using Pd2(dba-H)3, a 76% yield was recorded.
Absence of palladium precursor and ligand (1:1 ratio of a-7b/c-7b).
d

6380
A. R. Kapdi et al. / Tetrahedron Letters 51 (2010) 6378–6380
Acknowledgements
I.J.S.F. thanks the University of York for a Ph.D. studentship
(A.R.K.) and the Royal Society for funding.
References and notes
1. Nadin, A.; Nicolaou, K. C. Angew. Chem., Int. Ed. 1996, 35, 1622–1656.
2. (a) Bhide, R. S.; Patel, D. V.; Patel, M. M.; Robinson, S. P.; Hunihan, L. W.;
Gordon, E. M. Bioorg. Med. Chem. Lett. 1994, 4, 2107–2112; (b) Clarke, S. Ann.
Rev. Biochem. 1992, 61, 355–386; (c) Schafer, W. R.; Rine, J. Ann. Rev. Genet.
1992, 30, 209–237; (d) Omer, C. A.; Gibbs, J. B. Mol. Microbiol. 1994, 11, 219–
225; (e) Casey, P. J.; Seabra, M. C. J. Biol. Chem. 1996, 271, 5289–5292; (f)
Labadie, G. R.; Viswanathan, R.; Poulter, C. D. J. Org. Chem. 2007, 72, 9291–9297.
3. Selected examples: (a) Tae Kim, G.; Wenz, M.; Park, J.; Hasserodt, J.; Janda, K. D.
Bioorg. Med. Chem. 2002, 10, 1249–1262; (b) Li, J.; Lan, J.; Liu, Z.; Li, Y. J. Nat.
Prod. 1998, 61, 92–95.
4. Fairlamb, I. J. S.; Dickinson, J. M.; Pegg, M. Tetrahedron Lett. 2001, 42, 2205–
2208.
5. (a) Mechelke, M.; Wiemer, D. F. Tetrahedron Lett. 1998, 39, 783–786; (b)
Mechelke, M.; Wiemer, D. F. J. Org. Chem. 1999, 64, 4821–4829.
6. (a) Trost, B. M. Pure Appl. Chem. 1994, 66, 2007–2014; (b) Trost, B. M.; Fullerton,
T. J. J. Am. Chem. Soc. 1973, 95, 292–294.
Scheme 4. Synthesis of farnesyl-substituted compounds.
7. Use of two allylic leaving groups has been well investigated in cyclic systems:
(a) Bäckvall, J. E.; Nordberg, R. E. J. Am. Chem. Soc. 1981, 103, 4959–4960; (b)
Grennberg, H.; Gogoll, A.; Bäckvall, J. E. J. Org. Chem. 1991, 56, 5808–5811; (c)
Hupe, E.; Aranyos, A.; Szabo, K. J.; Bäckvall, J. E. Tetrahedron 1998, 54, 5375–
5384.
8. Paz, J. L.; Rodrigues, J. A. R. J. Braz. Chem. Soc. 2003, 14, 975–981.
9. [Pd0(dba-3,5,30,50-OMe)3] is available commercially through Sigma–Aldrich
(als2o known as [Pd(dm-dba)2]; Cat. No. 65693-3, and is a non-proprietary
material). This Pd0 precursor was found to be superior to Pd0(dba-H)3, see
2
Table 1, entry 3 and footnote c.
10. Fairlamb, I. J. S.; Kapdi, A. K.; Lee, A. F. Org. Lett. 2004, 6, 4435–4438.
11. (a) Fairlamb, I. J. S.; Kapdi, A. R.; Lee, A. F.; McGlacken, G. P.; Weissburger, F.; de
Vries, A. H. M.; van de Vondervoort, L. S. Chem. Eur. J. 2006, 12, 8750–8761; (b)
Fairlamb, I. J. S. Org. Biomol. Chem. 2008, 6, 3645–3656.
12. Representative procedure for allylic substitution: A solution of [Pd02(dba-3,5,30,50-
OMe)3]9 (0.0031 mmol, 0.5 mol %) and phosphine (0.024 mmol, 4 mol%) in THF
(2 mL) at 25 °C was stirred for 10 min under N2. To this was added 8-iodo-3,7-
dimethyl-(2E,6E)-octadien-1-yl acetate (5a) (0.20 g, 0.62 mmol) and the
reaction mixture was allowed to stir for 5 min. In a separate operation, a
solution of phenol (6) (0.62 mmol) in THF (2 mL) was added to NaH (60%
emulsion in oil) (2 equiv, 1.24 mmol) at 0 °C under N2. The resulting solution
was added to the reaction mixture and stirred at 25 °C for the desired time. On
completion, the reaction was quenched with H2O (5 mL) and then extracted
with Et2O (10 mL). The organic layer was separated, dried over MgSO4 and
filtered. Concentration in vacuo gave the desired products as oils. All
compounds were separated by flash chromatography on silica gel using
hexane/EtOAc mixtures as the eluent. Representative data for (2E,6E)-8-(4-
Scheme 5. Macrocyclic ketone 15 formation from Weinreb amide a-14.
To the best of our knowledge Weinreb amides have not been
used to form macrocyclic ketones. We postulated that the aromatic
bromide component of
a-14 could undergo halogen-lithium ex-
change at low temperature and high dilution, followed by an intra-
molecular 1,2-carbonyl addition/elimination sequence with the
suitably disposed vinyl amide motif (Scheme 5). Rather pleasingly,
methoxyphenoxy)-3,7-dimethyl-2,6-octadienyl acetate (a-7). Colourless oil.
the in situ generated aromatic lithium derivative of
a-14 was
1H NMR (400 MHz, CDCl3) 1.69 (s, 3H), 1.71 (s, 3H), 2.04 (s, 3H), 2.08–2.25 (m,
4H), 3.75 (s, 3H), 4.31 (s, 2H), 4.58 (d, 2H, J = 7.1 Hz), 5.35–5.38 (m, 1H), 5.48–
5.51 (m, 1H), 6.75–6.91 (m, 4H); 13C NMR (100 MHz, CDCl3) 16.3, 18.8, 23.4,
28.2, 41.3, 58.0, 63.8, 78.5, 114.2, 116.9, 118.2, 123.5, 134.0, 144.2, 152.4,
156.1, 173.7; LRMS (CI) m/z (rel.%): 336 (M++NH4þ, 20%), 318 (M+, 30), 275
found to undergo macrocyclisation (at ꢁ100 to 25 °C) to afford
15 in 23% yield. In this case, only the 16-membered macrocyclic
product was isolated from the reaction mixture (by flash chroma-
tography on silica gel).
In summary, a convenient protocol for the synthesis of geranyl
and farnesyl derivatives by Pd(0)-mediated allylic substitution,
which proceeds with excellent regio-, stereo- and chemocontrol,
has been developed. In preliminary studies, a macrocyclic ketone 15,
which is, potentially, an advanced intermediate for the synthesis
of likonide B15,16 could be synthesised using a novel macrocyclisa-
(50), 259 (60), 135 (100); HRMS (CI) m/z exact mass calculated for
þ
C
19H26O4+NH4 336.2188; found 336.2177. Selected data for regioisomer,
c
-
7: 1H NMR (400 MHz, CDCl3) 1.69 (s, 3H), 1.73 (s, 3H), 2.04 (s, 3H), 2.08–2.12
(m, 2H), 2.14–2.19 (m, 2H), 3.75 (s, 3H), 4.59 (d, 2H, J = 6.6 Hz), 4.61 (t, 1H,
J = 6.7 Hz), 5.01 (d, 1H, J = 1.3 Hz), 5.02 (d, 1H, J = 1.3 Hz), 5.28–5.31 (m, 1H),
6.72–6.89 (m, 4H).
13. Cane, D. E.; Tsantrizos, Y. S. J. Am. Chem. Soc. 1996, 118, 10037–10040.
14. Rudi, A.; Benayahu, Y.; Kashman, Y. Org. Lett. 2004, 6, 4013–4016.
15. Olson, B. S.; Trauner, D. Synlett 2005, 700–702.
16. Bruder, M.; Moody, C. J. Synlett 2008, 575–577.
tion strategy from a-14.