Enantioselective synthesis of 3,6-dihydro-1H-pyridin-2-ones: Unexpected regioselectivity in the palladium-catalyzed decarboxylative carbonylation of 5-vinyloxazolidin-2-ones [1]

Full text of DOI:10.1021/ja993897c

2944
J. Am. Chem. Soc. 2000, 122, 2944-2945
Communications to the Editor
Scheme 1. Synthesis of 5-Vinyloxazolidinones 1
Enantioselective Synthesis of
3,6-Dihydro-1H-pyridin-2-ones: Unexpected
Regioselectivity in the Palladium-Catalyzed
Decarboxylative Carbonylation of
5-Vinyloxazolidin-2-ones
Julian G. Knight,*,† Simon W. Ainge,† Andrew M. Harm,†
Simon J. Harwood,† Haydn I. Maughan,† Duncan R. Armour,‡
David M. Hollinshead,§ and Albert A. Jaxa-Chamiec‡
Department of Chemistry, Bedson Building
Newcastle UniVersity
Newcastle upon Tyne, NE1 7RU, UK
GlaxoWellcome
Medicines Research Centre, Gunnels Wood Road
SteVenage, Hertfordshire, SG1 2NY, UK
Zeneca Pharmaceuticals, Mereside, Alderley Park
Macclesfield, Cheshire, SK10 4TG, UK
and five-membered lactams by decarboxylative carbonylation of
6-vinyltetrahydro-2H-1,3-oxazin-2-ones.13
Since the decarboxylative carbonylation of vinyltetrahydro-
oxazinones13 leads to ring contraction to form γ-lactams, we
envisaged a similar process leading from 5-vinyloxazolidin-2-
ones 1 to â-lactams. In this contribution we report that palladium-
catalyzed decarboxylative carbonylation of amino acid-derived
5-vinyloxazolidin-2-ones does not give the expected â-lactams.
Instead, the corresponding δ-lactams, 3,6-dihydro-1H-pyridin-2-
ones, are formed.
ReceiVed NoVember 3, 1999
Transition metal catalyzed carbonylation of organic substrates
has proved an important method for the synthesis of carbon-
carbon and carbon-heteroatom bonds.1 The synthesis of â-lactams
by transition metal catalyzed processes is well documented2,3 and
has been accomplished by the carbonylation of aziridines,4
2-bromoallylamines,5 propargylamines,6 4-amino-2-alkynyl car-
bonates,7 and allyl phosphates in the presence of imines.8
Transition metal catalyzed carbonylative syntheses of lactams with
ring sizes larger than 4 include: benzo-fused five-, six-, and seven-
membered lactams by cyclocarbonylation of 2-aminostyrenes and
2-allylanilines;9 four-, five-, six-, and seven-membered R,â-
unsaturated lactams from amino vinyl-halides10a,b and -triflates;10c
five-, and six-membered lactams by hydrocarbonylation of amino-
alkenes11a,b and -alkynes;11c five-, and six-membered lactams by
carbonylative ring-expansion of azetidines12a and pyrrolidines;12b
The required 5-vinyloxazolidin-2-ones 1 were synthesized from
the corresponding R-amino aldehydes14 2 (R2 ) H) or ketone 2d
(R1 ) i-Pr, R2 ) Me) (Scheme 1). Aldehydes 2 (R2 ) H) were
obtained by Swern oxidation15 of the corresponding N-BOC
protected R-amino alcohols. Ketone 2d (R1 ) i-Pr, R2 ) Me)
was prepared by the addition of MeMgBr to the corresponding
Weinreb amide.16 Grignard additions to 2 proceeded, as ex-
pected,14 with low diastereoselectivity to produce the alcohols 3
as 1-5:1 mixtures of diastereoisomers which were cyclized to
the oxazolidinones 1 by treatment with sodium hydride.
Attempted carbonylation of 1a under the conditions reported
for ring-expansion of aziridines4c (20 mol % Pd2(dba)3‚CHCl3,
160 mol % PPh3, 1 atm CO, rt, C6H6) gave complete recovery of
starting material. Indeed, we were unable to find any catalyst/
solvent system which would enable the carbonylation of vinyl-
oxazolidinones to proceed at 1 atm of CO. Even at higher
pressures (up to 60 atm) the carbonylation was unsuccessful in
aprotic solvents such as THF, DMF, and MeCN. However,
carbonylation was successful using Pd(OAc)2(PPh3)2 (5 mol %)
at a CO pressure of 65 atm in a protic solvent, ethanol.17 The
product from this reaction was not the expected â-lactam but the
δ-lactam 4a. The reaction was not catalyzed by either Pd(OAc)2
or PPh3 alone. Table 1 shows that this reaction is successful for
a range of oxazolidinones providing δ-lactams in good to excellent
yields.18 The reaction tolerates substitution at C-5 (R2 ) Me, see
entry for 1g) and on the central carbon of the allyl system (R3 )
Me, entries for 1d-f), but fails in the case of the terminally
disubstituted vinyl derivative 1h (R4 ) Me) probably due to the
requirement for carbonylation to form a quaternary center in this
case. Comparison of 4a and 4g with ent-4a and ent-4g (prepared
† Newcastle University.
‡ GlaxoWellcome.
§ Zeneca Pharmaceuticals.
(1) Colquhoun, H. M.; Thompson, D. G.; Twigg, M. V. In Carbonylation;
Plenum Press: New York, 1991; pp 191-203.
(2) Barrett, A. G. M.; Sturgess, M. A. Tetrahedron 1988, 44, 5615-5652.
(3) Khumtaveeporn, K.; Alper, H. Acc. Chem. Res. 1995, 28, 414-422.
(4) (a) Piotti, M. E.; Alper, H. J. Am. Chem. Soc. 1996, 118, 111-116. (b)
Calet, S.; Urso, F.; Alper, H. J. Am. Chem. Soc. 1989, 111, 931-934. (c)
Spears, G. W.; Nakanishi, K.; Ohfune, Y. Synlett 1991, 91-92. (d) Tanner,
D.; Somfai, P. Bioorg. Med. Chem. Lett. 1993, 3, 2415-2418.
(5) Mori, M.; Chiba, K.; Okita, M.; Kayo, I.; Ban, Y. Tetrahedron 1985,
41, 375-385.
(6) (a) Bonardi, A.; Costa, M.; Gabriele, B.; Salerno, G.; Chiusoli, G. P.
Tetrahedron Lett. 1995, 36, 7495-7498. (b) Matsuda, I.; Sakakibara, J.;
Nagashima, H. Tetrahedron Lett. 1991, 32, 7431-7434.
(7) Mandai, T.; Ryoden, K.; Kawada, M.; Tsuji, J. Tetrahedron Lett. 1991,
32, 7683-7686.
(8) (a) Torii, S.; Okumoto, H.; Sadakane, M.; Hai, A. K. M. A.; Tanaka,
H. Tetrahedron Lett. 1993, 34, 6553-6556. (b) Zhou, Z.; Alper, H. J. Org.
Chem. 1996, 61, 1256-1260.
(9) El Ali, B.; Okuro, K.; Vasapollo, G.; Alper, H. J. Am. Chem. Soc. 1996,
118, 4264-4270.
(10) (a) Mori, M.; Washioka, Y.; Urayama, T.; Yoshiura, K.; Chiba, K.;
Ban, Y. J. Org. Chem. 1983, 48, 4058-4067. (b) Cope´ret, C.; Ma, S.; Sugihara,
T.; Negishi, E.-I. Tetrahedron 1996, 52, 11529-11544. (c) Crisp, G. T.; Meyer,
A. G. Tetrahedron, 1995, 51, 5585-5596.
(13) Bando, T.; Tanaka, S.; Fugami, K.; Yoshida, Z.-I.; Tamaru, Y. Bull.
Chem. Soc. Jpn 1992, 65, 97-110.
(14) Jurczak, J.; Golebiowski, A. Chem. ReV. 1989, 89, 149-164.
(15) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43, 3,
2480-2482.
(11) (a) Krafft, M. E.; Wilson, L. J.; Onan, K. D. Tetrahedron Lett. 1988,
29, 6421. (b) Zhou, J. Q.; Alper, H. J. Org. Chem. 1992, 57, 3328-3331. (c)
Campi, E. M.; Chong, J. M.; Jackson, W. R.; van der Schoot, M. Tetrahedron
1994, 50, 2533-2542.
(16) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
(17) These conditions are similar to those reported by Bando for the
decarboxylative carbonylation of vinyloxazinones.13
(12) (a) Roberto, D.; Alper, H. J. Am. Chem. Soc. 1989, 111, 7539. (b)
Wang, M. D.; Alper, H. J. Am. Chem. Soc., 1992, 114, 7018.
(18) See the Supporting Information for experimental procedures.
10.1021/ja993897c CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/10/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2945
Scheme 2. Carbonylation of N-BOC-Protected
Table 1. Palladium Catalyzed Decarboxylative Carbonylations of
5-vinyloxazolidinones 1a
5-Vinyloxazolidinone 5
1
R1
R2
R3
R4
yieldb (%)
a
i-Pr
Bn
Ph
i-Pr
Bn
Ph
H
H
H
H
H
H
Me
H
H
H
H
Me
Me
Me
H
H
H
H
H
H
H
H
Me
87
78
58
74
66
57
86
0d
b
cc
d
e
Scheme 3. Proposed Catalytic Cycle for Carbonylation
Reaction
fc
g
i-Pr
i-Pr
h
H
a Pd(Ph3P)2(OAc)2 (5 mol %), EtOH, CO (65 atm), 65-70 °C, 120
h. b Isolated yield. c The (4R)-isomer (derived from R-phenylglycine)
was used. d Starting material was recovered.
from (R)-valine) by chiral GC19,20 showed that no loss of
stereochemical integrity had occurred during the syntheses.
The reason for the unexpected regioselectivity of carbonylation
is unclear. Endo-type cyclizations of π-allyl palladium intermedi-
ates to give the larger of the two available ring sizes are commonly
seen in macrolide and medium ring carbocycle synthesis.21
However, when the selectivity is between formation of four- and
six-membered rings, kinetic control usually leads to the smaller
ring size (as seen in the formation of carbocycles21 and
â-lactams4c,d). With heteroatom nucleophiles, such as nitrogen,
formation of the larger ring size is possible due to thermodynamic
control resulting from reversible cyclization.21,22 It is therefore
possible that kinetic formation of the â-lactam is followed by
equilibration to the more stable δ-lactam in our case. It should
be noted that treatment of N-tosyl vinyloxazolidinones with Pd-
(PPh3)4 catalyst is reported to lead to the corresponding vinyl-
aziridines by loss of CO2.23 Treatment of the vinyloxazolidinone
1a with Pd(PPh3)4 catalyst in either THF or EtOH leads to
recovery of the oxazolidinone as a single (trans) diastereoisomer.
Thus, ring opening of the oxazolidinone to form the π-allyl
palladium species is reversible and is not accompanied by fast
decarboxylation.
showed no evidence of â-lactam formation (in particular, no
signals between δ 3.5-4 for the azetidinone 3- and 4-protons4c).
The π-allyl intermediate formed in this reaction clearly does not
decarboxylate since this would produce an intermediate identical
to that formed in the carbonylation of vinylaziridines. Subjecting
5 to our carbonylation conditions did not lead to the N-BOC-
protected δ-lactam but gave the ethoxycarbonylated allylic amine
6 in 48% yield (Scheme 2).
Formation of the anti complex 7 (Scheme 3) and decarboxy-
lation may be favored by interaction between the nucleophilic
amine and Pd; anti intermediate 8 has the required geometry to
form the six-membered ring. In the case of 5, the electron-
withdrawing nature of the BOC group may lead to decarboxy-
lation to form allyl species with no N-Pd coordination. Proto-
nation of the nitrogen by ethanol will render it a rather poor
nucleophile; hence, the formation of the ethoxycarbonylated
species 6.
In the formation of â-lactams from vinylaziridines, the nitrogen
carries an electron-withdrawing group (BOC or Ts). We therefore
prepared the N-BOC-protected vinyloxazolidinone 5 and subjected
it to carbonylation under Ohfune’s conditions.4c The only identifi-
able compound from this reaction was unreacted 5 (35%);
In summary, we have demonstrated that palladium-catalyzed
decarboxylative carbonylation of 5-vinyloxazolidin-2-ones, which
are readily prepared from amino acid precursors, leads to 3,6-
dihydro-1H-pyridin-2-ones in good yields. Further studies into
the mechanism of this reaction and synthetic applications of the
lactam products are underway and will be reported in due course.
1
however, the H NMR spectrum of the crude product mixture
(19) Stationary phase: heptakis(2,6-di-O-methyl-3-O-pentyl)-â-cyclodex-
trin.20
(20) Konig, W. A.; Gehrcke, B.; Icheln, D.; Evers, P.; Donnecke, J.; Wang,
W. C. J. High Resolut. Chromatogr. 1992, 15, 367.
(21) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1989, 28, 1173-1192 and
references therein.
(22) (a) Trost, B. M.; Geneˆt, J. P. J. Am. Chem. Soc. 1976, 98, 8516-
8517. (b) Grellier, M.; Pfeffer, M.; van Koten, G. Tetrahedron Lett. 1994,
35, 2877-2880.
(23) Ibuka, T.; Mimura, N.; Aoyama, H.; Akaji, M.; Ohno, H.; Miwa, Y.;
Taga, T.; Nakai, K.; Tamamura, H.; Fujii, N.; Yamamoto, Y. J. Org. Chem.
1997, 62, 999.
Acknowledgment. We thank the EPSRC, Zeneca Pharmaceuticals,
and GlaxoWellcome for studentship support and the Royal Society of
Chemistry for funds to purchase an autoclave.
Supporting Information Available: Experimental procedures, spec-
troscopic and analytical data for compounds 1a-h, 3a-h, 4a-g, 5, and
6 (PDF). This material is available free of charge via the Internet at
http://pubs.acs.org.
JA993897C