Kinetic resolution of oxazinones: An organocatalytic approach to enantiomerically pure β-amino acids

Article abstract of DOI:10.1002/anie.200502003

(Chemical Equation Presented) A profitable split: An organocatalytic resolution converts readily available racemic oxazinones 1 into valuable enantiomerically pure β-amino acid derivatives 2 (> 99% ee of the remaining 1 at 53% conversion; 88% ee of the ester 2). This catalytic ring-opening reaction requires as little as 1 mol% of the modular and readily accessible thiourea organocatalyst 3.

Full text of DOI:10.1002/anie.200502003

Communications
Asymmetric Organocatalysis
DOI: 10.1002/anie.200502003
Kinetic Resolution of Oxazinones: An
Organocatalytic Approach to Enantiomerically
Pure b-Amino Acids**
Albrecht Berkessel,* Felix Cleemann, and
Santanu Mukherjee
The development of new methods for the enantioselective
synthesis of b-amino acids has been of considerable interest
during the last years.[1–4] Owing to their unique pharmaco-
logical properties in the free form,[5] as cyclized b-lactams,[6]
and as structural elements of naturally occurring compounds
such as taxol[7] and the dolastatins,[8] there is a high demand in
both academia and industry for enantiopure b-amino acids.[9]
Besides chiral-auxiliary-based strategies[4] and catalytic asym-
metric synthesis[1,10–13] the kinetic resolution of racemates is a
well-established method for obtaining b-amino acids with
high enantiomeric purity. The latter methodology is divided
into “classical” chemical resolution processes,in which chiral
resolving agents are used in stoichiometric amounts,[14] and
enzymatic resolutions,which require acylases,amidases,or
lipases.[15,16] Although the selectivity of enzymatic processes is
in most cases excellent,the narrow substrate scope and thus
the need for time-consuming enzyme engineering present a
major drawback. To overcome these difficulties,we envisaged
that the application of readily available modular organo-
catalysts[17] could greatly enhance the scope of these pro-
cesses. Motivated by our success in the organocatalytic
dynamic kinetic resolution (DKR) of azlactones rac-A,we
set out to expand this methodology beyond the synthesis of
(non-natural) a- to the even more challenging b-amino acids
(Scheme 1).[18,19]
Scheme 1. Alcoholytic DKR of azlactones (A, ent-A;top) and KR of
oxazinones (C, ent-C; bottom).
appears to have been reported so far.[21] We considered these
compounds to be promising substrates for organocatalytic
asymmetric ring opening. It was hoped that the bifunctional
thiourea catalysts 2a and 2b (Table 1),which proved to be
highly effective in the alcoholytic DKR of azlactones,[19]
would effect a kinetic resolution (KR) of racemic oxazi-
nones.[22,23] If the selectivity of the catalyst is sufficiently high,
a KR process will be of practical significance,provided that
the racemate is obtained in an easy and cost-effective manner
and that both the remaining and the converted substrate
enantiomers are valuable compounds.[24] Oxazinones can be
synthesized in a way similar to azlactones by cyclodehydra-
tion of the corresponding N-benzoyl amino acids
3
(Scheme 2).[25] We decided to use the one-step protocol of
Tan and Weaver for the synthesis of racemic b-amino acids 4
from inexpensive aldehydes,malonic acid,and ammonium
acetate (Scheme 2).[26]
Azlactones A racemize readily and can be converted into
the enantiomerically enriched N,C doubly protected a-amino
acid derivatives B. By formally inserting a methylene group
between the carbonyl group and the a-C atom of the
azlactone,the nitrogen atom moves to the b position. The
resulting compounds are known as 4,5-dihydro-1,3-oxazin-6-
ones C,which are—in contrast to azlactones—configuration-
ally stable. Although achiral and enantiomerically pure
oxazinones C were utilized in peptide chemistry for coupling
to amino acid esters,[20] no catalytic ring-opening reaction
The crude amino acids 4a–f were converted directly into
the N-benzoyl derivatives 3a–f without the need for preced-
ing purification. Condensation to give oxazinones 1a–f was
best effected by isobutyl chloroformate. Preliminary experi-
[*] Prof. Dr. A. Berkessel, Dipl.-Chem. F. Cleemann, MSc S. Mukherjee
Institut für Organische Chemie
Universität zu Köln
Greinstrasse 4, 50939 Köln (Germany)
Fax: (+49)221-470-5102
E-mail: berkessel@uni-koeln.de
[**] This work was supported by the Fonds der Chemischen Industrie,
particularly through a doctoral fellowship to F.C. The authors thank
Degussa AG, Hanau, for generous gifts of amino acids.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Scheme 2. Synthesis of the racemic oxazinones rac-1a–f.
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2005, 44, 7466 –7469
7466

Angewandte
Chemie
Table 1: Optimization of the reaction conditions
ments indicate that 1 can also be obtained
directly in a single step from b-amino acids 4
by treatment with excess benzoyl chloride.
The first experiments with 4,5-dihydro-2,4-
diphenyl-1,3-oxazin-6-one (rac-1a) as sub-
strate and allyl alcohol as nucleophile in the
presence of the bifunctional thiourea 2a
(5 mol%) already showed the potential of
this process. As expected for a kinetic
resolution,the enantiopurity of the remain-
ing oxazinone 1a and the product ester 5a is
a function of the conversion (Figure 1).
At 57% conversion,the S enantiomer of
the oxazinone substrate rac-1a is com-
pletely consumed (99% ee in favor of the
remaining R enantiomer),and the enantio-
meric excess of the product ester (S)-5a is
86% (Table 1,entry 3). Calculated from
Entry
Catalyst
loading
[mol%]
Cat.
Solvent
Alcohol
(R)
t
[h]
Conv.
[%][a]
ee
ee
(1a)
(5a–7)
[%][a]
[%][a]
1
2
3
4
5
5
5
5
2a
2a
2a
2b
2a
2a
2a
2a
2a
2a
2a
2a
toluene
toluene
toluene
toluene
toluene
toluene
toluene
THF
allyl
allyl
allyl
allyl
allyl
allyl
allyl
allyl
allyl
allyl
Me
0.5
1.5
6.5
12
4.5
26
26
45
57
59
57
61
61
58
66
28
63
4
25
58
99
96
94
86
81
84
87
85
81
62
69
71
23
5
97
10
2.5
1
5
5
5
5
5
>99
98
these kinetic data,the selectivity factor
S
6
of the reaction is 68. At a conversion lower
than 45%,the ester is produced with more
than 94% ee (Figure 1 and Table 1,entries 1
and 2). Slightly lower selectivity and activity
were observed when using 5 mol% of the
cyclohexyl-substituted catalyst 2b (Table 1,
entry 4). As a consequence we chose cata-
lyst 2a for all further experiments. Because
7[b]
8
10
126
24
98
84
9
DCM
>99
18
>99
<2
10
11
12
CH3CN
toluene
toluene
24
24
24
iPr
[a] Conversions and enantiomeric excesses were determined by chiral HPLC. [b] This reaction was
carried out with a substrate concentration of 0.5m; 0.1m concentration in other reactions.
Scheme 3. Hydrolytic work-up procedure for the separation of the
resolved oxazinone (R)-1a from the product ester (S)-5a.
enantiomeric purity during this hydrolysis reaction
(Scheme 3). The filtrate contains the pure ester product (S)-
5a.
Figure 1. Time course of the kinetic resolution of rac-1a with allyl
alcohol (1.0 equiv) and catalyst 2a (5 mol% relative to rac-1a).
An increase in the catalyst loading to 10 mol% decreased
the reaction time to 4.5 h,without affecting the selectivity
(Table 1,entry 5). At a catalyst loading of 2.5 mol%,the
reaction time is longer but the selectivity remains virtually
unaffected (Table 1,entry 6). This decrease in the reaction
rate could be compensated for by raising the concentration of
the substrates from 0.1 to 0.5m. Under these conditions,a
catalyst loading of only 1 mol% was sufficient to catalyze the
reaction effectively (Table 1,entry 7). Comparison with
authentic samples prepared from enantiomerically pure (S)-
3-amino-3-phenylpropanoic acid allowed us to determine the
absolute configuration of both the oxazinone 1a and the ester
product 5a by chiral HPLC.
of the inherent difficulty in the accurate determination of
selectivity factors when S > 50,we decided to characterize the
quality of the kinetic resolution described herein solely in
terms of conversion and ee values of the substrate and
[27]
product,respectively (Table 1).
The remaining oxazinone 1a can be separated from the
converted product ester 5a by treating the reaction mixture
with dilute aqueous HCl (Scheme 3; see Supporting Infor-
mation for details). This hydrolytic procedure quantitatively
converts the remaining oxazinone (R)-1a into the insoluble
N-benzoyl b-amino acid (R)-3a. There is virtually no loss in
Angew. Chem. Int. Ed. 2005, 44, 7466 –7469ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7467

Communications
An increase in the solvent polarity,for example,by
examples demonstrate that the procedure is not limited to b-
amino acids with aromatic substituents but can also be used
equally well for the enantioselective synthesis of alkyl-
substituted derivatives.
changing from toluene to THF or acetonitrile,retarded the
reaction quite significantly (Table 1,entries 8 and 10). Based
on the assumption that hydrogen bonds between the catalyst
and the substrate are crucial for the activation of the latter,
this effect can be rationalized by the ability of THF to act as a
competing hydrogen-bond acceptor. In the case of acetoni-
trile,hydrogen bonds are weakened by dipolar interaction
with the solvent. In dichloromethane (DCM) the reaction
proceeded with a satisfying rate,but the selectivity was
somewhat lower (Table 1,entry 9). Prolonged reaction times
resulted from the variation of the alcohol nucleophile to
methanol or 2-propanol (Table 1,entries 11 and 12). In the
case of methanol,the reaction was still practically useful,
although the selectivity was somewhat lower. In contrast,the
sterically more demanding 2-propanol reacts much more
slowly. Only 4% conversion was observed after 24 h (Table 1,
entry 12).
In summary,we have described a novel and practical
organocatalytic method for the synthesis of enantiomerically
pure b-amino acids. To the best of our knowledge,this is the
first time that oxazinones were applied in a catalytic ring-
opening reaction. The resolutions are carried out at ambient
temperature; the applied catalyst is modular,readily avail-
able,and can be used at loadings as low as 1 mol%. This
process provides efficient access to highly valuable protected
b-amino acids from inexpensive bulk chemicals. The resolved
starting material and the product can be separated by
filtration in a simple workup procedure. It is particularly
noteworthy that the remaining enantiomerically pure oxazi-
nones are activated b-amino acid derivatives that can,in
principle,be applied directly in coupling reactions,for
example,in the synthesis of b-peptides. The scope of this
novel resolution process and its mechanism are currently
under investigation.
To demonstrate the broad substrate scope of our KR
process,we chose the electron-poor p-chlorophenyl- and m-
nitrophenyl-substituted oxazinones 1b and 1d as well as the
electron-rich p-methoxyphenyl-substituted oxazinone 1c as
substrates (Table 2,entries 2,4,and 3). Furthermore,to show
Experimental Section
All commercially available chemicals were used without further
purification. (S)-3-Amino-3-phenylpropionic acid was a gift from
Degussa AG,Hanau. Solvents were distilled prior to use and dried,if
necessary,by using standard techniques. HPLC analysis was per-
formed on Merck–Hitachi HPLC equipment with HPLC-grade
solvents from Fisher Scientific. GC analysis was performed on
Hewlett-Packard equipment. Catalysis runs were carried out under
inert atmosphere.
Table 2: Asymmetric alcoholytic ring opening of oxazinones
Entry
Substrate (R)
t
Conv.
[%][a]
ee (1)
ee (5)
[h]
[%][a,b]
[%][a]
Resolution of oxazinones: The alcohol nucleophile (1.00 equiv)
was added to a solution of the catalysts 2a or 2b (8.33 mmol,
0.05 equiv) in absolute toluene (670 mL). After the addition of a
solution of the oxazinone 1a–f (167 mmol,1.00 equiv) in absolute
toluene (1.00 mL),the homogeneous reaction mixture was stirred at
ambient temperature. For analysis,50- mL samples were withdrawn
and diluted with acetonitrile (450 mL). The conversion and enantio-
meric excess were determined immediately by HPLC or GC (see
Supporting Information). Quantification was based on UV detection
at l = 230 nm. The conversion was determined by comparison with
the peak areas of stock solutions of the oxazinones 1 and the
corresponding N-benzoyl amino acid esters 5–7 in dichloromethane.
1
2
3
4
5
6
1a (Ph)
6.5
15
10.5
3.0
48
57
59
57
64
54
53
99
99
98
99
97
98
86
83
87
81
82
88
1b (p-ClC6H4)
1c (p-OMeC6H4)
1d (m-NO2C6H4)
1e (tBu)
1 f (iPr)
48
[a] Conversions were determined by chiral HPLC, the enantiomeric
excesses either by chiral HPLC or chiral GC. [b] The absolute config-
urations of the products 1b–f were assigned by analogy to 1a.
that the method is not limited to oxazinones with aromatic
substituents,we also included the tert-butyl derivative 1e and
the isopropyl derivative 1 f (Table 2,entries 5 and 6) in our
study. We were pleased to see that all substrates were resolved
with a high degree of enantioselectivity. In the presence of
catalyst 2a (5 mol%),the oxazinones 1a–f were generally
obtained with excellent enantiomeric excess (Table 2),
whereas the esters were still produced with 80–90% ee.
Comparison of 1a with the p-chlorophenyl-substituted
oxazinone 1b and the p-methoxy-substituted oxazinone 1c
reveals that the enantioselectivity is not significantly affected
by the substituent on the aromatic ring (Table 2,entries 1–3).
The same holds for the oxazinone 1d,which bears a nitro
substituent in the meta position (Table 2,entry 4). With the
latter substrate,the reaction is complete within only 3 h. The
sterically more-demanding tert-butyl residue and the isopro-
pyl group of oxazinones 1e and 1 f are also tolerated at
reaction times of 48 h (Table 2,entries 5 and 6). The latter two
Received: June 10,2005
Published online: October 24,2005
Keywords: beta amino acids · kinetic resolution ·
.
organocatalysis · oxazinones · thiourea
[1] J.-A. Ma, Angew.Chem. 2003, 115,4426 – 4435; Angew.Chem.
Int.Ed. 2003, 42,4290 – 4299.
[2] E. Juaristi,H. Lopez-Ruiz, Curr.Med.Chem. 1999, 6,983 – 1004.
[3] A. F. Abdel-Magid,J. A. Cohen,C. A. Maryanoff, Curr.Med.
Chem. 1999, 6,955 – 970.
[4] E. Juaristi, Enantioselective Synthesis of b-Amino Acids,Wiley,
New York, 2005.
[5] S. Shinagawa,T. Kanamaru,S. Harada,M. Asai,H. Okazaki, J.
Med.Chem. 1987, 30,1458 – 1463.
[6] G. I. Georg, The Organic Chemistry of b-Lactams,VCH,New
York, 1993.
7468 www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7466 –7469

Angewandte
Chemie
[7] I. Ojima,S. Lin,T. Wang, Curr.Med.Chem. 1999, 6,927 – 953.
[8] R. B. Bates,K. G. Brusoe,J. J. Burns,S. Caldera,W. Cui,S.
Gangwar,M. R. Gramme,K. J. McClure,G. P. Rouen,H.
Schadow,C. C. Stessman,S. R. Taylor,V. H. Vu,G. V. Yarick,
J. Zhang,G. R. Pettit,R. Bontems, J.Am.Chem.Soc. 1997, 119,
2111 – 2113.
[9] D. Seebach,S. Abele,K. Gademann,G. Guichard,T. Hinter-
mann,B. Jaun,J. L. Matthews,J. V. Schreiber, Helv.Chim.Acta
1998, 81,932 – 982.
[10] S. Kobayashi,H. Ishitani, Chem.Rev. 1999, 99,1069 – 1094.
[11] H. Ishitani,M. Ueno,S. Kobayashi, J.Am.Chem.Soc. 1997, 119,
7153 – 7154.
[12] M. P. Sibi,J. J. Shay,M. Liu,C. P. Jasperse, J.Am.Chem.Soc.
1998, 120,6615 – 6616.
[13] a) J. K. Myers,E. N. Jacobsen, J.Am.Chem.Soc.
1999, 121,
8959 – 8960; b) A. G. Wenzel,E. N. Jacobsen, J.Am.Chem.Soc.
2002, 124,12964 – 12965.
[14] J. A. Zablocki,J. G. Rico,R. B. Garland,T. E. Rogers,K.
Williams,L. A. Schretzman,S. A. Rao,P. R. Bovy,F. S. Tjoeng,
R. J. Lindmark,M. V. Toth,M. E. Zupec,D. E. McMackins,S. P.
Adams,M. Miyano,C. S. Markos,M. N. Milton,S. Paulson,M.
Herin,P. Jacqmin,N. S. Nicholson,S. G. Panzer-Knodle,N. F.
Haas,J. D. Page,J. A. Szalony,B. B. Taite,A. K. Salyers,L. W.
King,J. G. Campion,L. P. Feigen, J.Med.Chem. 1995, 38,2378 –
2394.
[15] P. Flores-Sanchez,J. Escalante,E. Castillo, Tetrahedron: Asym-
metry 2005, 16,629 – 634.
[16] V. A. Soloshonok,N. A. Fokina,A. V. Rybakova,I. P. Shishkina,
S. V. Galushko,A. E. Sorochinsky,V. P. Kukhar,M. V. Sav-
chenko,V. K. Svedas, Tetrahedron: Asymmetry 1995, 6,1601 –
1610.
[17] A. Berkessel,H. Gröger, Asymmetric Organocatalysis,Wiley-
VCH,Weinheim, 2005.
[18] A. Berkessel,F. Cleemann,S. Mukherjee,T. N. Müller,J. Lex,
Angew.Chem. 2005, 117,817 – 821; Angew.Chem.Int.Ed. 2005,
44,807 – 811.
[19] A. Berkessel,S. Mukherjee,F. Cleemann,T. N. Müller,J. Lex,
Chem.Commun. 2005,1898 – 1900.
[20] a) C. N. C. Drey,J. Lowbridge,R. J. Ridge, J.Chem.Soc.Perkin
Trans.1 1973,2001 – 2006; b) C. M. C. Drey,E. Mtetwa, J.Chem.
Soc.Perkin Trans.1 1982,1587 – 1592.
[21] For a report on ring-opening polymerization of oxazinones,see:
S. Kobayashi,Y. Tsukamoto,T. Saegusa, Macromolecules 1990,
23,2609 – 2612.
[22] The tert-leucine amide motif was employed before in the design
of chiral organocatalysts: a) D. E. Fuerst,E. N. Jacobsen, J.Am.
Chem.Soc. 2005, 127,8964 – 8965; b) T. P. Yoon,E. N. Jacobsen,
Angew.Chem. 2005, 117,470 – 472; Angew.Chem.Int.Ed. 2005,
44,466 – 468; c) M. S. Taylor,E. N. Jacobsen, J.Am.Chem.Soc.
2004, 126,10558 – 10559; d) G. D. Joly,E. N. Jacobsen, J.Am.
Chem.Soc. 2004, 126,4102 – 4103; e) P. Vachal,E. N. Jacobsen, J.
Am.Chem.Soc. 2002, 124,10012 – 10014; f) A. G. Wenzel,M. P.
Lalonde,E. N. Jacobsen, Synlett 2003,1919 – 1922.
[23] Catalysts of the thiourea-tert-amine type were applied before for
asymmetric Michael additions and aza-Henry reactions: a) T.
Okino,Y. Hoashi,Y. Takemoto, Angew.Chem. 2005, 117,4100 –
4103; Angew.Chem.Int.Ed. 2005, 44,4032 – 4035; b) T. Okino,
T. Furukawa,X. Xu,Y. Takemoto, J.Am.Chem.Soc. 2005, 127,
119 – 125; c) T. Okino,S. Nakamura,T. Furukawa,Y. Takemoto,
Org.Lett. 2004, 6,625 – 627; d) T. Okino,Y. Hoashi,Y. Take-
moto, J.Am.Chem.Soc. 2003, 125,12672 – 12673.
[24] J. M. Keith,J. F. Larrow,E. N. Jacobsen, Adv.Synth.Catal. 2001,
343,5 – 26.
[25] W. Baker,W. Ollis, J.Chem.Soc. 1949,345 – 349.
[26] C. Y. K. Tan,D. F. Weaver, Tetrahedron 2002, 58,7449 – 7461.
[27] C. S. Chen,Y. Fujimoto,G. Girdaukas,C. J. Sih, J.Am.Chem.
Soc. 1982, 104,7294 – 7299.
Angew. Chem. Int. Ed. 2005, 44, 7466 –7469ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7469