Decarboxylative isomerization of N-Acyl-2-oxazolidinones to 2-oxazolines

Article abstract of DOI:10.1021/jo800076f

(Chemical Equation Presented) N-Acyl-2-oxazolidinones are ring-opened by lithium iodide and decarboxylated in the presence of a mild proton source. Further reaction with an amine base provides 2-oxazolines. The transformation is general for oxazolidinones unsubstituted in the 5 position and occurs under mild conditions (25-50°C). These results complement the existing methods for this transformation by allowing lower temperatures and/or avoiding metal catalysts.

Full text of DOI:10.1021/jo800076f

SCHEME 1. Decarboxylative Cyclizations of
N-Acyloxazolidinones
Decarboxylative Isomerization of
N-Acyl-2-oxazolidinones to 2-Oxazolines
Aaron E. May, Patrick H. Willoughby, and
Thomas R. Hoye*
Department of Chemistry, UniVersity of Minnesota,
Minneapolis, Minnesota 55455
hoye@umn.edu
ReceiVed January 12, 2008
SCHEME 2. Lithium Iodide Initiated Decarboxylative
Cyclization
N-Acyl-2-oxazolidinones are ring-opened by lithium iodide
and decarboxylated in the presence of a mild proton source.
Further reaction with an amine base provides 2-oxazolines.
The transformation is general for oxazolidinones unsubsti-
tuted in the 5 position and occurs under mild conditions
(25-50 °C). These results complement the existing methods
for this transformation by allowing lower temperatures and/
or avoiding metal catalysts.
Oxazolines comprise a number of biologically active natural
products and medicinal compounds.1,2 Their well-documented
use as ligands in asymmetric catalysis, as well as their wide
use as monomers in living cationic polymerization, makes
multiple methods for their synthesis attractive.3,4 We describe
here a one-pot procedure for the conversion of N-acyl-2-
oxazolidinones to 2-substituted oxazolines by a lithium iodide/
ammonium chloride/DBU-mediated decarboxylative cyclization.
Previous one-pot transformations of this type are outlined in
Scheme 1. In the presence of an equal weight of CaO,
N-benzoyloxazolidinone 1a was converted to 2a in 65% yield
by heating with a Bunsen burner in the absence of solvent.5a
This method appears to have a limited substrate scope; for
example, N-acetyloxazolidinone 1b under these conditions gave
product 2b having purity of only 25%. Another high-temperature
method, applied only to the trifluoroacetyl derivative 1c, gave
2c in 76% yield.5b A palladium-catalyzed transformation of
5-vinyl-substituted oxazolidinones 3 to oxazolines 4 has been
reported.6 Yields and diastereoselectivities are high for this
process, but the requirement of a 5-alkenyl substituent is a
limitation of this chemistry. The drawbacks of the previous
methods caused us to search for an alternative that did not
require high temperatures or expensive catalysts while still
allowing a variety of N-acyl substituents.
The initial notion was that a decarboxylative cyclization
requiring only catalytic amounts of lithium iodide might be
achieved as outlined in Scheme 2. Attack at the 5 position of
activated oxazolidinone A by iodide would give the lithium
carboxylate B, which we hoped would decarboxylate to the
lithium amide C and O-alkylate to give oxazoline 6, thereby
regenerating lithium iodide.7 Indeed, 4-phenylmethyl-2-oxazo-
lidinone 5d (R ) Ph) reacted with lithium iodide under mild
(6) (a) Cook, G. R.; Shanker, S. P. J. Org. Chem. 2001, 66, 6818-
6822. (b) Cook, G. R.; Shanker, S. P. Tetrahedron Lett. 1998, 39, 3405.
(c) Cook, G. R.; Manivannan, E.; Underdahl, T.; Lukacova, V.; Zhang, Y.;
Balaz, S. Bioorg. Med. Chem. Lett. 2004, 19, 4935-4939.
(7) Intervention of N-alkylation to generate an N-acylaziridine cannot
be excluded since, via the Heine rearrangement, these are known to ring
open and isomerize to 2-oxazolines in the presence of iodide ion: (a) Heine,
H. W.; Kenyon, W. G.; Johnson, E. M. J. Am. Chem. Soc. 1961, 83, 2570-
2574. (b) Kurti, L.; Czako, B. Strategic Applications of Named Reactions
in Organic Synthesis; Elsevier Academic Press: London, 2005; pp 198-
199.
(1) Bertram, A.; Pattenden, G. Nat. Prod. Rep. 2007, 24, 18-30.
(2) Pirrung, M. C.; Tumey, L. N.; McClerran, A. L.; Raetz, C. R. H. J.
Am. Chem. Soc. 2003, 125, 1575-1586.
(3) Desimoni, G.; Faita, G.; Jorgensen, K. A. Chem. ReV. 2006, 106,
3561-3651.
(4) Kobayashi, S.; Uyama, H.; Narita, Y.; Ishiyama, J. Macromolecules
1992, 25, 3232-3236.
(5) (a) Mundy, B. P.; Kim, Y. J. Heterocycl. Chem. 1982, 19, 1221-
1222. (b) Foris, A.; Neumer, J. F. Magn. Reson. Chem. 2005, 43, 867-
868.
10.1021/jo800076f CCC: $40.75 © 2008 American Chemical Society
Published on Web 03/18/2008
3292
J. Org. Chem. 2008, 73, 3292-3294

TABLE 1. Decarboxylative Cyclizations of
N-Acyloxazolidin-2-ones 5 to Oxazolines 6
entry
R
conditionsa
yield (%)b
a
b
c
d
e
f
Me
Et
A
A
A
B
B
B
B
62c
71
82
76
73d
45
80
iPr
Ph
FIGURE 1. Unsuitable decarboxylative cyclization substrates.
4-MeOC6H4
4-NO2C6H4
2-napthyl
Substrate scope was explored (Scheme 3 and Figure 1). The
aldol adduct 7 was a suitable substrate for ring opening, giving
the oxazoline 8 in 69% yield after chromatography.8 It is
noteworthy that no epimerization of the R stereocenter was
observed during the decarboxylation or cyclization steps,9
making this the method of choice for producing 2-oxazolines
with a stereocenter at the R position of the C2 substituent.10
Sulfonimide 9 gave rise to the iodoalkylsulfonamide 10 in 66%
yield following ring opening and decarboxylation.11,12
g
a A: LiI (4.0 equiv), NH4Cl (2.0 equiv), 0.2 M in CHCl3, 50 °C, 16 h;
DBU (3.0 equiv), rt, 6 h. B: LiI (2.0 equiv), NH4Cl (2.0 equiv), 0.2 M in
CH2Cl2, rt, 12 h; DBU (3.0 equiv), rt, 6 h. b Isolated yield after chroma-
tography. c With 10% recovered deacylated 5a. d Yield of 81% based on
recovered starting material.
SCHEME 3. Additional Lithium Iodide Initiated Reactions
Shown in Figure 1 are substrates that did not cleanly (if at
all) ring open with the iodide ion. In the case of attempted
aziridine formation with unactivated substrates 11 and 12, little
to no conversion of the starting material was observed, even
under forcing conditions.13 The lack of reactivity speaks to both
the chelating and electron-withdrawing property of the N-acyl
substituent in the successful cyclizations. In the case of
N-thiocarbonyl compound 13, poor conversion to product is
observed, presumably owing to the lower Lewis basicity of
sulfur.14 Substitution on the 5 position of the oxazolidinone as
in substrate 14 was found to effectively shut down the iodide-
initiated decarboxylation, a result that was not unexpected
considering the mechanism of the reaction. Interestingly, the
N-methoxycarbonyl compound 15 was found to be unsuitable
for the reaction; instead, it underwent demethylation by iodide
and decarboxylation to give methyl iodide (1H NMR) and 11.
In conclusion, the transformation of N-acyl-2-oxazolidinones
to 2-oxazolines is afforded by a one-pot reaction with lithium
iodide/NH4Cl followed by DBU. The conditions are mild and
general for oxazolidinones unsubstituted in the 5 position. These
results expand the methodology of decarboxylative cyclization
chemistry. This method is potentially valuable to those interested
in making oxazoline ligands for asymmetric catalysis, monomers
conditions (ambient temperature, DCM), but the overall plan
was thwarted when intermediate B failed to decarboxylate. That
is, in the presence of substoichiometric amounts of LiI, the
reaction never proceeded to completion, and (labile) intermediate
D (R ) Ph) could be observed (NMR spectroscopy) after protic
workup. This crude iodoamide could be subsequently cyclized
efficiently to oxazoline 6d (R ) Ph) by the action of DBU. We
were pleased to see that incorporation of a weak proton source
(e.g., NH4Cl) in the reaction mixture from the outset caused in
situ decarboxylation. We then found that, if DBU was added
after full conversion of 5 to D, all three steps could be performed
easily in one pot. If DBU was present from the outset, 5 was
consumed much more slowly, presumably because of the
reduced Lewis acidity resulting from competitive complexation
of Li+ by DBU. Finally, when pure 6d was subjected to LiI
(with and without added NH4Cl) in DCM in the absence of
DBU, it did not revert to D, demonstrating that the cyclization
of D to 6 is not reversible.
After screening for the best choice of solvent, concentration,
reaction time, proton source, metal additive, and follow-up base,
we settled upon one-pot conditions in which the 2-oxazolidinone
5 was first treated with LiI and NH4Cl in a chlorocarbon solvent
and, after complete consumption of 5 (TLC), DBU (3 equiv)
was added. Results from the decarboxylation/cyclization of
N-alkanoyloxazolidinones 5a-c (conditions A) and N-aroylox-
azolidinones 5d-g (conditions B) are summarized in Table 1.
For the latter set, less lithium iodide (2 vs 4 equiv) was used,
and the reaction was typically carried out at room temperature
rather than 50 °C.
(8) For the synthesis of anti-aldol adducts, see: Evans, D. A.; Tedrow,
J. S.; Shaw, J. T.; Downey, C. W. J. Am. Chem. Soc. 2002, 124, 392-393.
(9) As judged from the fact that the 1H NMR spectrum of the oxazoline
product 8 indicated that it was essentially one diastereomer.
(10) For general reviews on oxazoline formation and usage, see: (a)
Frump, J. A. Chem. ReV. 1971, 71, 483-505. (b) Reuman, M.; Meyers, A.
I. Tetrahedron 1985, 41, 837-860. (c) Kronek, J.; Luston, J.; Boehme, F.
Chem. Listy 1998, 92, 175-185.
(11) Hedayatullah, M.; Brault, J. F. Phosporus Sulfur 1981, 11, 303-
310.
(12) Treatment of 10 with DBU led to consumption and generation of
baseline TLC material. Although this was not further explored, the presumed
intermediate tosylaziridine would be expected to undergo ring opening by
the nucleophilic DBU to form an ammonium salt: Moon, B.; Han, S.; Kim,
D. Org. Lett. 2005, 7, 3359-3361.
(13) We are aware of a single example of aziridine formation from an
N-alkyloxazolidinone in which TMSI was used as the reagent. See:
Lakanen, J. R.; Pegg, A. E.; Coward, J. K. J. Med. Chem. 1995, 38, 2714-
2727.
(14) For the synthesis of thioacylated carbamates, see: Wagner, G.;
Leistner, S. Pharmazie 1972, 27, 547-552.
J. Org. Chem, Vol. 73, No. 8, 2008 3293

room temperature, 0.049 mL (0.329 mmol) of DBU was added.
After 6 h at room temperature, the mixture was filtered through a
plug of silica gel (EtOAc eluent) before being purified by medium
pressure liquid chromatography (30% EtOAc in hexanes) to afford
0.0231 g (69%) of the desired product as a clear oil: 1H NMR
(500 MHz, CDCl3) δ 7.33-7.25 (m, 7H), 7.22-7.17 (m, 3H), 5.75
(dq, J ) 15.0, 6.5 Hz, 1H), 5.26 (ddq, J ) 15.5, 9.0, 2.0, 2.0, 2.0
Hz, 1H), 4.81 (d, J ) 7.0 Hz, 1H), 4.675 (d, J ) 7.5 Hz, 1H),
4.674 (d, J ) 11.5 Hz, 1H), 4.45 (d, J ) 11.5 Hz, 1H), 4.38 (dddd,
J ) 9.5, 9.5, 7.5, 5.0 Hz, 1H), 4.22 (dd, J ) 9.0, 8.5 Hz, 1H), 4.09
(dd, J ) 9.5, 8.5 Hz, 1H), 3.94 (dd, J ) 8.5, 7.5 Hz, 1H), 3.14 (dd,
J ) 13.5, 4.5 Hz, 1H), 2.72 (dq, J ) 7.0, 7.5 Hz, 1H), 1.73 (dd, J
) 6.5, 2.0 Hz, 3H), and 1.11 (d, J ) 7.5 Hz, 3H); 13C NMR (125
MHz, CDCl3) δ 169.5, 138.2, 132.3, 129.4, 128.7, 128.6, 128.5,
128.1, 127.8, 126.6, 115.1, 91.4, 78.7, 71.7, 69.3, 67.3, 42.0, 39.1,
18.0, and 14.3; HRMS (ESI) calcd for (C24H30NO3+) 380.2220,
found 380.2289; GC-MS tr ) 14.7 min; m/z 379, 355, 341, 327,
288, 258, 243, 242, 228, 218, 207, 188, 161, 150, 117, 91, and 65;
IR (neat) 3030, 2939, 2885, 1667, 1654, 1505, 1455, 1383, 1174,
1100, 1036, 1027, 971, 924, 733, and 698 cm-1; TLC Rf ) 0.50 in
40% EtOAc in hexanes; [R]rt ) +70.5 (c ) 0.210, CHCl3).
for living cationic polymerization, or potential bioactive/
medicinal compounds.1-4
Experimental Section
Conditions B: The Decarboxylative Cyclization of 5d to give
6d. To a 6 mL screw cap vial equipped with a stirbar were added
0.0596 g (0.212 mmol) of (S)-3-benzoyl-4-phenylmethyl-2-oxazo-
lidinone (5d), 0.0590 g (0.440 mmol) of lithium iodide, 0.0233 g
(0.440 mmol) of ammonium chloride, and 1.1 mL of methylene
chloride. The vial was then capped, and the mixture was stirred at
room temperature for 12 h. DBU (0.095 mL, 0.637 mmol) was
added. After 6 h, the mixture was filtered through a plug of silica
gel (EtOAc eluent) before being purified by medium pressure liquid
chromatography (40% EtOAc in hexanes) to afford 0.0381 g (76%)
of the desired product as a clear oil: 1H NMR (500 MHz, CDCl3)
δ 7.59-7.94 (m, 2H), 7.50-7.47 (m, 1H), 7.43-7.40 (m, 2H),
7.33-7.29 (m, 2H), 7.26-7.23 (m, 3H), 4.58 (dddd, J ) 9.0, 9.0,
7.5, 5.0 Hz, 1H), 4.35 (dd, J ) 9.0, 9.0 Hz, 1H), 4.15 (dd, J ) 8.0,
8.0 Hz, 1H), 3.25 (dd, J ) 14.0, 5.5 Hz, 1H), and 2.73 (dd, J )
14.0, 9.0 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ 164.0, 137.9,
131.3, 129.6, 128.5, 128.3, 128.2, 126.5, 71.8, 67.9, and 41.8;
HRMS (ESI) calcd for (C16H15NO + Na+) 260.1046, found
260.1066; GC-MS tr ) 12.0 min; m/z 237, 218, 206, 146, 118,
105, 91, 77, 65, and 51; IR (neat) 3061, 3027, 2960, 2929, 2899,
1651, 1603, 1579, 1495, 1455, 1450, 1358, 1085, 1060, 1025, 967,
780, and 695 cm-1; TLC Rf ) 0.50 in 30% EtOAc in hexanes;
[R]rt ) +8.0 (c ) 0.670, CHCl3).
Acknowledgment. Dedicated to Professor Harold W. Heine
on the occasion of his 85th birthday. This work was supported
by the National Cancer Institute, National Institutes of Health
(CA-76497), and the Lando/NSF Chemistry Summer Research
Program. We thank Heidi A. Dahlmann for assistance with the
scaleup of compound 7.
Conditions A: The Decarboxylative Cyclization of 7 to give
8. To a 6 mL screw cap vial equipped with a stirbar were added
0.0374 g (0.0884 mmol) of (4R)-3-[(2S,3S,4E)-2-methyl-1-oxo-3-
phenylmethoxymethoxy-4-hexenyl]-4-phenylmethyl-2-oxazolidi-
none (7), 0.0575 g (0.429 mmol) of lithium iodide, 0.0105 g (0.198
mmol) of ammonium chloride, and 0.33 mL of chloroform. The
vial was then capped, and the mixture was stirred at 50 °C in an
oil bath for 12 h. After allowing the reaction mixture to cool to
Supporting Information Available: Experimental procedures
1
and full characterization data and copies of the H and 13C NMR
spectra for all compounds prepared. This material is available free
of charge via the Internet at http://pubs.acs.org.
JO800076F
3294 J. Org. Chem., Vol. 73, No. 8, 2008