Memory of chirality of tertiary aromatic amides: A simple and efficient method for the enantioselective synthesis of quaternary α-amino acids

Article abstract of DOI:10.1021/ja9039604

A new methodology for the asymmetric synthesis of quaternary R-substituted amino acids using memory of chirality has been developed. The strategy utilizes the dynamic axial chirality of tertiary aromatic amides to memorize the initial chirality of an α-amino acid during an enolization step. Starting from five different L-amino acids, the corresponding oxazolidin-5-ones containing a tertiary aromatic amide group have been synthesized in one step and then alkylated with various electrophiles, with good yields and enantioselectivities (up to 96% and up to >99% after recrystallization). One-step deprotection affords enantioenriched or enantiopure quaternary α-amino acids. We describe here the optimization process, the results obtained in each series and a plausible explanation, based on NMR studies, DFT calculations and crystallographic structures. The methodology presented herein constitutes an efficient synthesis of enantiopure quaternary R-amino acids (three steps only) starting from tertiary L-amino acids, without any external source of chirality.

Full text of DOI:10.1021/ja9039604

Published on Web 07/06/2009
Memory of Chirality of Tertiary Aromatic Amides: A Simple
and Efficient Method for the Enantioselective Synthesis of
Quaternary r-Amino Acids
Mathieu Branca, Se´bastien Pena, Re´gis Guillot, Didier Gori, Vale´rie Alezra,* and
Cyrille Kouklovsky*
Laboratoire de Chimie des Proce´de´s et Substances Naturelles, ICMMO, CNRS UMR 8182,
Baˆt 410, UniVersite´ de Paris-Sud, F-91405 Orsay, France
Received May 15, 2009; E-mail: valezra@icmo.u-psud.fr; cykouklo@icmo.u-psud.fr
Abstract: A new methodology for the asymmetric synthesis of quaternary R-substituted amino acids using
memory of chirality has been developed. The strategy utilizes the dynamic axial chirality of tertiary aromatic
amides to memorize the initial chirality of an R-amino acid during an enolization step. Starting from five
different L-amino acids, the corresponding oxazolidin-5-ones containing a tertiary aromatic amide group
have been synthesized in one step and then alkylated with various electrophiles, with good yields and
enantioselectivities (up to 96% and up to >99% after recrystallization). One-step deprotection affords
enantioenriched or enantiopure quaternary R-amino acids. We describe here the optimization process, the
results obtained in each series and a plausible explanation, based on NMR studies, DFT calculations and
crystallographic structures. The methodology presented herein constitutes an efficient synthesis of
enantiopure quaternary R-amino acids (three steps only) starting from tertiary L-amino acids, without any
external source of chirality.
chirality of the starting R-amino acids.5 Among them are the
Introduction
Self Regeneration of Stereocenters, developed by D. Seebach6
The conformation of a peptide is crucial for its biological
activity and most small natural peptides are conformationally
flexible, thus not able to adopt a secondary peptide structure.
One way to restrict peptide conformation is the introduction of
a side chain restricted amino acid, for instance a quaternary
R-amino acid.1 Thus, due to their biological importance, many
procedures have been developed for the stereoselective synthesis
of quaternary R-amino acids.2 Typical methods involve chiral
auxiliaries3 or chiral catalysts,4 but very few use only the
and exemplified by other groups,7 and more recently some
procedures applying a new principle, memory of chirality.8
A reaction proceeding with memory of chirality has been
defined by Carlier as “a formal substitution at a sp3 stereogenic
center that proceeds stereospecifically, even though the reaction
proceeds by trigonalization of that center, and despite the fact
that no other permanently chiral elements are present in the
system”. By applying this principle, several groups succeeded
in enantioselective quaternization of R-amino acids derivatives.9
The most successful examples have been reported by Carlier,
who used the intrinsic chirality of the benzodiazepine-2-one ring
to obtain R,R-dialkyl amino acids enantioselectively.12 In
addition, Kawabata has performed methylation and allylation
of various R-amino acids with high enantiomeric excesses.10
Although the electrophile range is quite limited for the
(1) (a) Maity, P.; Ko¨nig, B. Biopolymers 2008, 90, 8–27. (b) Toniolo, C.;
Crisma, M.; Formaggio, F.; Peggion, C. Biopolymers 2001, 60, 396–
419.
(2) Recent reviews: (a) Ohfune, Y.; Shinada, T. Eur. J. Org. Chem. 2005,
5127–5143. (b) Kang, S. H.; Kang, S. Y.; Lee, H.-S.; Buglass, A. J.
Chem. ReV. 2005, 105, 4537–4558. (c) Cativiela, C.; Diaz-de-Villegas,
M. D. Tetrahedron: Asymmetry 2007, 18, 569–623. (d) Vogt, H.;
Bra¨se, S. Org. Biomol. Chem. 2007, 5, 406–430.
(3) Recent examples: (a) Jones, M. C.; Marsden, S. P.; Mu˜noz Subtil,
D. M. Org. Lett. 2006, 8, 5509–5512. (b) Satoh, T.; Hirano, M.;
Kuroiwa, A.; Kaneko, Y. Tetrahedron 2006, 62, 9268–9279. (c)
Davies, S. G.; Garner, A. C.; Ouzman, J. V. A.; Roberts, P. M.; Smith,
A. D.; Snow, E. J.; Thomson, J. E.; Tamayo, J. A.; Vickers, R. J.
Org. Biomol. Chem. 2007, 5, 2138–2147. (d) Vassiliou, S.; Yiotakis,
A.; Magriotis, P. A. Tetrahedron Lett. 2006, 47, 7339–7341.
(4) Recent examples: (a) Maruoka, K.; Ooi, T.; Kano, T. Chem. Commun.
2007, 1487–1495. (b) Lee, Y.-J.; Kim, M.-J.; Kim, T.-S.; Park, H.-g.;
Jew, S.-s. Org. Lett. 2005, 7, 1557–1560. (c) Lee, Y.-J.; Lee, J.; Kim,
M.-J.; Jeong, B.-S.; Lee, J.-H.; Kim, T.-S.; Lee, J.; Ku, J.-M.; Jew,
S.-s.; Park, H.-g. Org. Lett. 2005, 7, 207–3209. (d) Achard, T. R. J.;
Clutterbuck, L. A.; North, M. Synlett 2005, 1828–1847. (e) Banti, D.;
Belokon, Y. N.; Fu, W.-L.; Groaz, E.; North, M. Chem. Commun.
2005, 2707–2709. (f) Chowdari, N. S.; Barbas, C. F., III. Org. Lett.
2005, 7, 867–870.
Hitchcock, S. R.; Powell, D. R.; Schrimpf, M. R. J. Am. Chem. Soc.
1999, 121, 2460–2470. (c) Pe´rez-Faginas, P.; O’Reilly, F.; O’Byrne,
A.; Garcia-Aparicio, C.; Martin-Martinez, M.; Pe´rez de Vega, J. J.;
Garcia-Lopez, M. T.; Gonzalez-Muniz, R. Org. Lett. 2007, 9, 1593–
1596.
(6) Seebach, D.; Swing, A. R.; Hoffmann, M. Angew. Chem., Int. Ed.
1996, 35, 2708–2748.
(7) Recent examples using the same principle: (a) Jimenez-Ose´s, G.;
Aydillo, C.; Busto, J. H.; Zurbano, M. M.; Peregrina, J. M.; Avenoza,
A. J. Org. Chem. 2007, 72, 5399–5402. (b) Shendage, D. M.; Fro¨hlich,
R.; Bergander, K.; Haufe, G. Eur. J. Org. Chem. 2005, 719–727.
(8) Reviews:(a) Fuji, K.; Kawabata, T. Chem.sEur. J. 1998, 4, 373–376.
(b) Kawabata, T.; Fuji, K. In Topics in Stereochemistry; Denmark,
S. E., Ed.; John Wiley & Sons: New York, 2003; Vol. 23, pp 175-
205. (c) Eames, J.; Suggate, M. J. Angew. Chem., Int. Ed. 2005, 44,
186–189. (d) Zhao, H.; Hsu, D. C.; Carlier, P. R. Synthesis 2005, 1–
16.
(5) (a) Ferey, V.; Toupet, L.; Le Gall, T.; Mioskowski, C. Angew. Chem.,
Int. Ed. 1996, 35, 430–432. (b) Vedejs, E.; Fields, S. C.; Hayashi, R.;
9
10.1021/ja9039604 CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 10711–10718 10711

A R T I C L E S
Branca et al.
Scheme 1. General Strategy for the Synthesis of Enantioenriched
Quaternary R-Amino Acids
Figure 1. Tertiary aromatic amides.
intermolecular reaction, a very efficient intramolecular alkylation
reaction was developed.11
In order to improve these results, we thought to use the axial
chirality of tertiary aromatic amides, which present two slow
bond rotations (Figure 1).13 These compounds are generally not
planar and even moderate steric hindrance forces a dihedral
angle of 90° on the Ar-CO bond. Subject to certain constraints
of substitution pattern (A * B), the two perpendicular conform-
ers about the Ar-CO bond are enantiomeric (R1 and R2 achiral)
or diastereomeric (R1 or R2 chiral), and these compounds can
therefore present axial chirality.
Scheme 2. Synthesis of Enantioenriched Quaternary L-Valine
Our strategy was to introduce a tertiary aromatic amide
function onto the starting R-amino acid in order to transfer the
initial central chirality to an axial chirality. Due to the high
rotation barriers of tertiary aromatic amides, the axial chirality
should be retained during the enolization/alkylation step at low
temperature and induce a stereoselective attack by the electro-
phile. The quaternary R-amino acid should be obtained after
deprotection (Scheme 1). In order to protect both the amino
and the acid groups, we chose an oxazolidin-5-one, knowing
that it is easy to synthesize and to cleave.
We have successfully applied this strategy to L-valine and
performed the synthesis of quaternary enantioenriched L-valine
3 in only three steps (Scheme 2).14 We have also proposed an
explanation for the observed stereoselectivity based on DFT
calculations and NMR studies.15 Herein, we report a full article
of our work, including the application of this strategy to other
amino acids.
(9) (a) Kolaczkowski, L.; Barnes, D. M. Org. Lett. 2007, 9, 3029–3032.
(b) Betts, M. J.; Pritchard, R. G.; Schofield, A.; Stoodley, R. J.; Vohra,
S. J. Chem. Soc., Perkin Trans. 1 1999, 1067–1072. (c) Giese, B.;
Wettstein, P.; Sta¨helin, C.; Barbosa, F.; Neuburger, M.; Zehnder, M.;
Wessing, P. Angew. Chem., Int. Ed. 1999, 38, 2586–2587. (d) Bonache,
M. A.; Lopez, P.; Martin-Martinez, M.; Garcia-Lopez, M. T.; Cativiela,
C.; Gonzalez-Muniz, R. Tetrahedron 2006, 62, 130–138. (e) Bonache,
M. A.; Cativiela, C.; Garcia-Lopez, M. T.; Gonzalez-Muniz, R.
Tetrahedron Lett. 2006, 47, 5883–5887.
Results and Discussion
1. Optimization of Reaction Conditions Starting from
L-Valine. We decided to develop our method starting from
L-valine because of the steric hindrance induced by the isopropyl
group and because of the biological interest of R-methylvaline.16
We have chosen oxazolidinones 4 arising from formaldehyde
(R2 ) H) or 5 from acetone (R2 ) Me) and also tested several
aromatic groups. Compounds 4 were synthesized following a
known two-step procedure: initial condensation of L-valine with
an aromatic acyl chloride,17 and then formation of the oxazo-
lidinone ring by heating the resulting compound 6 in toluene
with paraformaldehyde (Scheme 3).18 Unfortunately, we could
not avoid slight deterioration of the enantiomeric excess19
(10) Selected examples: (a) Kawabata, T.; Chen, J.; Suzuki, H.; Fuji, K.
Synthesis 2005, 1368–1377. (b) Kawabata, T.; Kawakami, S.-p.;
Shimada, S.; Fuji, K. Tetrahedron 2003, 59, 965–974. (c) Kawabata,
T.; Kawakami, S.-p.; Fuji, K. Tetrahedron Lett. 2002, 43, 1465–1467.
(d) Kawabata, T.; Suzuki, H.; Nagae, Y.; Fuji, K. Angew. Chem., Int.
Ed. 2000, 39, 2155–2157.
(11) (a) Moriyama, K.; Sakai, H.; Kawabata, T. Org. Lett. 2008, 10, 3883–
3886. (b) Kawabata, T.; Moriyama, K.; Kawakami, S.; Tsubaki, K.
J. Am. Chem. Soc. 2008, 130, 4153–4157. (c) Watanabe, T.; Kawabata,
T. Heterocycles 2008, 76, 1593–1606. (d) Kawabata, T.; Matsuda,
S.; Kawakami, S.; Monguchi, D.; Moriyama, K. J. Am. Chem. Soc.
2006, 128, 15394–15395. (e) Kawabata, T.; Kawakami, S.; Majumdar,
S. J. Am. Chem. Soc. 2003, 125, 13012–13013.
(15) Branca, M.; Alezra, V.; Kouklovsky, C.; Archirel, P. Tetrahedron 2008,
64, 1743–1752.
(12) (a) Carlier, P. R.; Zhao, H.; MacQuarrie-Hunter, S. L.; DeGuzman,
J. C.; Hsu, D. C. J. Am. Chem. Soc. 2006, 128, 15215–15220. (b)
MacQuarrie-Hunter, S. L.; Carlier, P. R. Org. Lett. 2005, 7, 5305–
5308. (c) Carlier, P. R.; Lam, P. C.-H.; DeGuzman, J. C.; Zhao, H.
Tetrahedron: Asymmetry 2005, 16, 2999–3002.
(16) (a) Bellanda, M.; Mammi, S.; Geremia, S.; Demitri, N.; Randaccio,
L.; Broxterman, Q. B.; Kaptein, B.; Pengo, P.; Pasquato, L.; Scrimin,
P. Chem.sEur. J. 2007, 13, 407–416. (b) Toniolo, C.; Crisma, M.;
Formaggio, F.; Peggion, C. Biopolymers 2001, 60, 396–419.
(17) o-Phenyl-benzoyl chloride was synthesized from the corresponding
acid (oxalyl chloride 3 equiv, catalytic DMF, dichloromethane, 0 °C,
1 h, 96%); o-tert-butyl-benzoyl chloride was synthesized by SNAr with
tert-butyllithium onto 2-fluorobenzo¨ıc acid: Gohier, F.; Castanet, A.-
S.; Mortier, J. Org. Lett. 2003, 5, 1919–1922. followed by reflux in
thionyl chloride (75% overall yield).
(13) (a) Betson, M. S.; Clayden, J.; Helliwell, M.; Johnson, P.; Lai, L. W.;
Pink, J. H.; Stimson, C. C.; Vassiliou, N.; Westlund, N.; Yasin, S. A.;
Youssef, L. H. Org. Biomol. Chem. 2006, 4, 424–443. (b) Bragg, R. A.;
Clayden, J.; Morris, G. A.; Pink, J. H. Chem.sEur. J. 2002, 8, 1279–
1289. (c) Ahmed, A.; Bragg, R. A.; Clayden, J.; Lai, L. W.; McCarthy,
C.; Pink, J. H.; Westlund, N.; Yasin, S. A. Tetrahedron 1998, 54,
13277–13294.
(18) Ben-Ishai, D. J. Am. Chem. Soc. 1957, 79, 5736–5738.
(19) All enantiomeric excesses are determined by chiral stationary-phase
HPLC.
(14) Branca, M.; Gori, D.; Guillot, R.; Alezra, V.; Kouklovsky, C. J. Am.
Chem. Soc. 2008, 130, 5864–5865.
9
10712 J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009

Quaternary R-Amino Acids by Memory of Chirality
Scheme 3. Synthesis of Oxazolidinone 4
A R T I C L E S
Figure 2. Four conformers of compound 4a.
Scheme 4. Synthesis of Oxazolidinones 1 and 5a,b
Table 1. In Situ Methylation of Oxazolidines 1, 4 and 5
Figure 3. ORTEP plot for X-ray crystal structure of compound 4b.
Ellipsoids are drawn at the 50% probability level.
1
disappointing results, we obtained H NMR spectra in CDCl3
of the starting oxazolidinones 4a and 4b at low temperature.
For compound 4a, four conformers are observable (Figure 2),
the trans-conformers (carbonyl oriented toward the isopropyl
group) being the major ones (cis/trans assignment can be easily
made by chemical shift comparison, N-substituents cis to the
carbonyl group being shifted downfield relative to those of
N-substituents cis to the aromatic ring21). For compound 4b,
two conformers (ratio 1/0.2) are observable, the trans-conformer
being again the major one. Moreover, in the crystallographical
structure of compound 4b, the conformation of the tertiary
aromatic amide was determined to be (P,trans) (Figure 3),
suggesting once again that trans-conformers are favored. These
preferred trans-conformations force the aromatic group to be
oriented far away from the asymmetric center and could explain
the low enantioselectivities obtained. For compound 4c, the
situation is slightly different, because at room temperature, the
four conformers are already clearly defined, with one highly
favored one (trans-conformer). This indicates that the Ar-CO
and N-CO rotation barriers are much higher than in the former
cases. Thus, in this case, the higher enantiomeric excess of 56%
likely results from increased hindrance and/or increased rotation
barriers.
substrate
(ee %)
yielda eeb
entry
R2
Ar
product
base
(%)
(%)
1
2
3
4
5
H
H
H
1-naphthyl
o-biphenyl
4a (98%)
7a
7b
7c
8b
2a
LiHMDS 36
0
11
56
4b (>99%)
LDA
LDA
LDA
LDA
20
45
o-t-Bu-phenyl 4c (92%)
Me o-biphenyl
Me 1-naphthyl
5b (>99%)
1 (>99%)
5c 32
14
70
a Isolated yield. b Enantiomeric excess, determined by chiral
stationary-phase HPLC. c Conversion, determined by chiral stationary-phase
HPLC.
(except for compound 4b, because it was crystalline). We
obtained the expected compounds 4 in moderate yields (addition
of catalytic p-toluenesulfonic acid increased yields but also
racemization).
Oxazolidinones 1, 5a, b were synthesized without racemiza-
tion in one step from sodium L-valinate by modifying a
procedure described by Seebach (Scheme 4).20 Unfortunately,
we were unable to synthesize the o-tert-butyl-benzoyl derivative
by applying this method.
We next performed in situ methylation (introduction of
electrophile before base) to force alkylation to occur as soon as
the enolate is formed. In this case, racemization (i.e., Ar-CO
rotation) of the enolate should be limited. Deprotonation was
achieved with LiHMDS or LDA when LiHMDS was not strong
enough (Table 1). The major drawback of this method is the
likely competitive alkylation of the base, resulting in decreased
yields.
For compounds 5b and 1, introduction of a quaternary center
on the oxazolidinone (derived from acetone) induces a radical
change in cis/trans orientation. For compound 5b, at room
temperature, only two conformers are observable (1H NMR
spectrum in CDCl3), one of which is very predominant (ratio
100:4). As no crystallographic structure can be obtained for
Compound 7a was obtained in racemic form, and compound
7b, with a low enantiomeric excess. In order to understand these
(21) Lewin, A. H.; Frucht, M.; Chen, K. V. J.; Benedetti, E.; di Blasio, B.
(20) Seebach, D.; Fadel, A. HelV. Chim. Acta 1985, 68, 1243–1250.
Tetrahedron 1975, 31, 207–215.
9
J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009 10713

A R T I C L E S
Branca et al.
Table 2. Screening for Methylation Reaction Conditions of Compound 1a
entry
solvent
baseb (equiv)
additive (equiv)
t1 (min)
X (equiv)
t2 (min)
conv (%)c
ee (%)c
1
2
3
4
5
6
7
8
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
Et2O
Et2O
Et2Oi
Et2Oi
Et2Oi
Et2Oi
Et2Oi,j
Et2Oi,k
Et2Oi,l
LDA (3)
LDA (3)
LDA (3)
LDA (3)
-
-
-
-
-
0d
0d
0d
3
30
3
I (3)
I (6)
60
60
60
60
60
60
60
60
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
14
70
64
81
65
32
70
72
85
66
74
57
8
33
72
50
79
80
78
88
82
82
85
54
83
(9e)
(13e)
67(56e)
(56e)
49
I (30)
I (30)
I (30)
I (30)
I (30)
I (6)
I (30)
I (30)
I (30)
I (30)
I (30)
I (30)
I (30)
I (30)
I (5)
LDA (3)
LDA (3)
DME (6)
DME (6)
-
KDAf (3)
3
56g
20
49
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (3)
KHMDS (1.5)
KHMDS (1.5)
KHMDS (1.5)
KHMDS (1.5)
KHMDS (1.5)
KHMDS (1.5)
KHMDS (1.5)
0d
4
9
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
DME (6)
DME (6)
18-c-6 (3)
18-c-6 (3)
TMEDA (3)
-
4
57
10
10
1
10
10
10
10
10
3
8
10
8
8
8
100
100
95
64
78
DME (6)
DME (6) DMPU (6)h
DME (3) DMPU (3)h
DME (3)
DME (3)
DME (3)
DME (3)
DME (3)
DME (3)
87
71
I (5)
88(74e)
64
OTf (3)
OTf (5)
OTf (2)
OTf (5)
OTf (5)
OTf (5)
95(78e)
99(72e)
93
99
82
a Unless specified, the concentration of 1 was 0.07 mol ·L-1, and a mixture of base and additive was added Via canula at -78 °C to compound 1.
b KHMDS 0.5 M in toluene, LDA prepared from nBuLi in hexane and diisopropylamine. c Determined by chiral stationary-phase HPLC. d Electrophile
in situ. e Isolated yield. f Synthesized in situ from LDA and t-BuOK. g Nucleophilic attack of KDA was also observed. h Added with electrophile.
i Concentration of 1 was 0.15 mol ·L-1 j Reaction at -86 °C. k Reaction at -60 °C. l Crystal dissolution at -78 °C.
.
equiv) at -78 °C did not improve conversion (8%). However,
performing the deprotonation for 30 min with KHMDS at -56
°C and then addition of MeI (30 equiv) gave 35% conversion
but a lower ee (20%). We suppose that at higher temperature,
rotation of the aromatic becomes possible and allows deproto-
nation of the minor conformer.
Compound 1 gave the methylated compound 2a with the
highest enantiomeric excess (70%), despite a low yield. For
1
compound 1, four conformers are present on the H NMR
spectrum in CDCl3 at low temperature, and the cis-conformers
are now the major ones. We have previously noticed that the
crystal structure of compound 1 indicated that, in the solid state,
1 adopts a (P,cis)-conformation,14 which was consistent with
the major conformation adopted in CDCl3 at low temperature
(full 1H NMR assignment of the labile proton in the four
conformers was achieved by theoretical studies15). Thus, in
the major conformer, the carbonyl group is oriented toward the
more hindered quaternary center, placing the aromatic group
close to the asymmetric center. This phenomenon could explain
the better enantioselectivity observed.
Considering that compound 1 was easily accessible in
enantiopure form and that this oxazolidinone gave methylated
compound 2a with good enantiomeric excess (despite a low
yield), we decided to pursue our studies on this oxazolidinone.
Base, additives, solvents, electrophiles and reaction conditions
(temperature, concentration, deprotonation time or alkylation
time) were then screened (Table 2).
Figure 4. Minimized structure of compound 5b.
compound 5b, we attempted to determine the major conformer
observed (1H NMR spectrum in CDCl3) by performing com-
putational studies. In accordance with the experimental data,
all three semiempirical methods tested (MNDO, AM1 and PM3)
for energy minimization calculation22 gave the lowest energy
for the (M,cis)-conformer (Figure 4). Steric hindrance due to
the second aromatic ring (located right over the labile proton),
and high rotation barriers would explain why deprotonation of
compound 5b is so difficult (entry 4, Table 1).23 Sequential
deprotonation/methylation with KHMDS (3 equiv)/MeI (30
(23) At-78 °C the low ee and conversion could be eventually explained
by deprotonation/alkylation of the minor conformer, which is probably
a trans one.
(22) For details, see Supporting Information.
9
10714 J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009

Quaternary R-Amino Acids by Memory of Chirality
Table 3. Alkylation of Compound 1a
A R T I C L E S
Scheme 5. Access to Enantioenriched R-Substituted Amino Acids
recryst
ee
entry
electrophile
E
product yield (%) ee (%)b yield (%) (%)b
Table 4. Synthesis of Oxazolidinones 9-12
1
2
3
4
5
6
7
8
MeIc
MeOTf
EtOTf
allyl-I
Bn-I
Me
Me
Et
2a 74
2a 78d
2b 59
2c 88d
78
82
91
88
60
80
67
85
83
89
94
>99
99
>99
98
allyl
Bne
2d 72-98 73-91
4-OMeBn-I
ethyliodoacetate CH2CO2Et 2f
t-BuOD
4-OMeBn
2e 98d
80d
2g 76f
86
96
93
>99
D
a General procedure: KHMDS (1.5 equiv, 0.5 M in toluene) diluted in
DME (3 equiv) at -78 °C was added Via canula to a solution of 1 in
Et2O at -78 °C (concentration of 1 ) 0.15 mol ·L-1); after 8 min,
electrophile (5 equiv) was rapidly added and after 10 min at -78 °C,
reaction was quenched with acetic acid. b Determined by chiral
stationary-phase HPLC. c DMPU (3 equiv) added with electrophile.
d Complete conversion. e For unknown reasons, ee varies from 73 to
yield ee
product (%) (%)a
entry
amino acid
L-leucine
L-alanine
L-phenylalanine Bn
L-methionine
R
Lewis acid
1
2
3
4
i-Bu
Me
BF3.OEt2 (cat)
9
49 >99
52 >99
61 >99
46 >99
AlMe3 (1 equiv) 10
AlMe3 (1 equiv) 11
CH2CH2SMe AlMe3 (1 equiv) 12
f
1
a Determined by chiral stationary-phase HPLC.
91%. Percent deuteration determined by H NMR spectroscopy.
The in situ reaction gave the highest enantioselectivity but
low yields, and increased enolate formation time improved the
yield but caused deterioration of the enantioselectivity. KHMDS
gave higher yields and enantioselectivities than LDA (entries 2
and 8) and was therefore selected for further study. Yields and
enantioselectivities generally increased when an additive was
used. Among the additives tested, 1,2-dimethoxyethane (DME)
gave the best results. Using a crown ether boosted conversion
but was deleterious to the enantiomeric excess (entries 12 and
13), suggesting that the countercation has to remain in the
vicinity of the enolate. Therefore, we replaced THF by a less
coordinating solvent, diethyl ether. Conversion was, not surpris-
ingly, slightly decreased but racemization of the enolate was
also slowed down (entries 11 and 16). Good conversion was
restored when replacing methyliodide by the more reactive
methyltriflate (or by adding DMPU with methyliodide), and by
increasing the concentration. Lowering the temperature (-86
°C, entry 22) did not significantly improve the enantioselectivity,
and increasing the temperature was deleterious (-60 °C, entry
23). Moreover, dissolution of crystals at -78 °C (entry 24) just
before the reaction did not improve enantioselectivity. The best
conditions were the following: deprotonation with 1.5 equiv of
KHMDS (in toluene) in diethyl ether/DME (3 equiv) for less
than 10 min (a 0.15 mol ·L-1 concentration for the enolate
corresponds to a toluene/ether ratio of 55/45) and alkylation
with a very reactive electrophile for 10 min (entry 20), or with
a less reactive electrophile with DMPU (entry 18).
tively to R-methylvaline 3a or R-isopropylaspartic acid 3f
(Scheme 5). The S-configuration was confirmed by comparison
with the optical rotation of known compounds.24
3. Extension to Other Amino Acids. We next wanted to
extend this methodology to other amino acids. The correspond-
ing oxazolidinones were synthesized following the same pro-
cedure (in some cases, AlMe3 gave better yields than BF3.OEt2)
without racemization (Table 4). Crystal structures of compounds
10, 11 and 1225 show that in the solid state, they all adopt a
(P,cis)-conformation (Figure 5).
3.1. Extension to L-Leucine. Compound 9 was not very
soluble in ether, and therefore allylation of this oxazolidinone
under optimized conditions gave compound 13a in only 35%
yield. Solubilization with additional DME gave alkylated
compounds 13a-c with high yields and enantioselectivities
(Table 5).
Enantiopure quaternary R-amino acids can be synthesized
following the same strategy as described above: recrystallization
of compound 13c and deprotection in refluxing HBr led
quantitatively to enantiopure R-isobutylaspartic acid 14 (Scheme
6). Retention of configuration was confirmed by comparison of
the optical rotation with that of a known compound.26
3.2. Extension to L-Alanine. Compound 10 was not very
soluble in ether, and therefore supplementary DME (12 equiv)
was added to perform allylation. Enantioselectivity was disap-
pointingly low, and thus optimization of the allylation conditions
was necessary (Table 6, entry 1). Replacement of ether by THF
increased product solubility, and DME was no longer necessary.
This led to enantiomeric excess enhancement (entries 1 and 3),
and the best results were obtained by performing the reaction
without toluene (entry 5). Compared with oxazolidinone 1, the
enolate arising from oxazolidinone 10 is much more configu-
2. Alkylation of Compound 1 and Synthesis of r-Substituted
L-Valine. We next applied the optimized conditions to other
reactive electrophiles (Table 3): enantioselectivities exceed 73%,
and yields range from 59 to 98%. Enantioselectivity can be
enhanced by recrystallization for all compounds. For compounds
2d and 2e, the racemic compound crystallized and was removed,
leading to the enantiopure compound by evaporation of the
mother liquor. Moreover, analysis of deuterated compound 2g
by chiral stationary-phase HPLC indicated that deuteration
proceeded with retention of configuration.
(24) (a) Cativiela, C.; Diaz-de-Villegas, M. D.; Galvez, J. A.; Lapena, Y.
Tetrahedron 1995, 51, 5921–5928. (b) Fadel, A.; Salau¨n, J. Tetrahe-
dron Lett. 1987, 28, 2243–2246.
(25) For compound 12, the crystallographic structure reveals eight conform-
ers (because of the flexibility of the alkyl chain), all with (P,cis)-
conformation.
Finally, deprotection and isolation of the quaternary R-amino
acid was achieved in one step. Hydrolysis of compound 2a (ee
) 94%) or 2f (ee > 99%) in refluxing HBr 47% led quantita-
(26) Obrecht, D.; Bohdal, J.; Daly, J.; Lehman, C.; Shoenholzer, P.; Mueller,
K. Tetrahedron 1995, 51, 10883–10900.
9
J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009 10715

A R T I C L E S
Branca et al.
Figure 5. ORTEP plot for the X-ray crystal structure of compounds 10, 11 and 12. Ellipsoids are drawn at the 50% probability level.
Table 5. Alkylation of Oxazolidinone 9
Table 7. Alkylation of Oxazolidinone 10
entry
E-X
E
product
yield (%)
ee (%)a
entry
E-X
E
product
yield (%)
ee (%)a
1
2
3
allyl-I
Bn-Br
ethyliodoacetate
allyl
Bn
CH2CO2Et
15a
15b
15c
87
82
94
79
82
91
1
2
3
allyl-I
allyl
Me
13a
13b
13c
88
86
96
90
87
94
Me-OTf
ethyliodoacetate
CH2CO2Et
a Determined by chiral stationary-phase HPLC.
a Determined by chiral stationary-phase HPLC.
Scheme 6. Synthesis of (R)-R-Isobutylaspartic Acid
Scheme 7. Synthesis of (R)-R-Methylaspartic Acid
Table 8. Methylation of Oxazolidinone 11
Table 6. Allylation of Oxazolidinone 10
solvent
KHMDS (S2)
entry
solvent 1a (S1)
S1:S2
t1 (min) conv (%)b ee (%)b
entry
solvent 1a (S1)
solvent KHMDS (S2) S1:S2 t1 (min) conv (%)b ee (%)b
1
2
3
4
5
6
7
Et2O
THF
THF
THF
THF/DME (1:1)
THF/DMF (1:1)
toluenec 55:45
toluenec 55:45
toluenec 55:45
8
8
3
3
3
8
8
34
100
100
100
100
100
100
65
73
73
75
67
74
65
1
2
3
4
5
6
Et2O/DME (3:1)
Et2O/DME (3:1)
THF
THF
THF
THF
toluenec
toluenec
toluenec
toluenec
THFd
62:38
62:38
55:45
55:45
70:30
70:30
3
8
3
1
3
8
100
100
100
86
100
100
67
67
74
75
79
79
THFd
70:30
toluenec 55:45
toluenec 55:45
THFd
THF/toluene (1:1) toluenec 70:30
a Solvent used for oxazolidinone dissolution. b Determined by chiral
a Solvent used for oxazolidinone dissolution. b Determined by chiral
c
stationary-phase HPLC. Commercially available KHMDS 0.5 mol ·L-1
c
stationary-phase HPLC. Commercially available KHMDS 0.5 mol ·L-1
d
in toluene. KHMDS 0.5 mol ·L-1 in THF (evaporation of toluene and
d
in toluene. KHMDS 0.5 mol ·L-1 in THF (evaporation of toluene and
replacement by THF).
replacement by THF).
rationally stable: the enantiomeric excess is unchanged between
3 and 8 min deprotonation times (entries 1/2, or 5/6). Performing
direct deprotonation on crystals did not induce significant
changes, and we applied the optimized conditions (entry 5) to
other electrophiles (Table 7).
In all cases, alkylated products 15a-c were obtained in good
yields and enantioselectivities. Recrystallization of compound
15c and deprotection in refluxing HBr led quantitatively to
enantiopure R-methylaspartic acid 16 (Scheme 7). Retention of
configuration was confirmed by comparison of the optical
rotation with that of a known compound.24b
3.3. Extension to L-Phenylalanine. Application of the condi-
tions optimized for oxazolidinone 1 were not satisfying, and
therefore, a change of solvent was necessary (Table 8). Chiral
stationary-phase HPLC analysis showed directly (comparison
with the chromatogram of compound 15b obtained from
oxazolidinone 10) that compound 15b is formed with retention
of configuration. Compared with oxazolidinone 1, the enolate
arising from oxazolidinone 11 is much more configurationally
stable: the enantiomeric excess is unchanged between 3 and 8
min deprotonation time (entries 2 and 3). Addition of supple-
mentary DME, toluene or DMF did not improve enantioselec-
9
10716 J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009

Quaternary R-Amino Acids by Memory of Chirality
Table 9. Alkylation of Oxazolidinone 11
A R T I C L E S
entry
E-X
E
product
yield (%)
ee (%)a
1
2
3
MeOTf
allyl-I
ethyliodoacetate
Me
allyl
CH2CO2Et
(S)-15b
17a
17b
88
92
85
75
68
89
a Determined by chiral stationary-phase HPLC.
Figure 6. ORTEP plot for X-ray crystal structure of compound 18b.
Ellipsoids are drawn at the 50% probability level.
Table 10. Alkylation of Oxazolidinone 12
Scheme 8. Alkylation of Compound 5a
entry
E-X
E
product yield (%) ee (%)a
1
2
3
4
allyl-I
MeI
MeOTf
allyl
Me
Me
18a
18b
18b
18c
85
92
62
75
86
78
86
91
1
Table 11. Conformer Ratio Observed by H NMR Spectroscopy at
195 K of Oxazolidinones 1, 10-12
tert-butylbromoacetate CH2CO2t-Bu
starting
entry amino acid compound
number of
conformers (P,cis)/(M,cis)a cis/transa
a Determined by chiral stationary-phase HPLC.
solvent
THF-d8
Et2O-d10
toluene-d8/
Et2O-d10 (5/6)
THF-d8
1
2
3
L-Val
L-Val
L-Val
1
1
1
4
4
4
100.0/14.2 100.0/5.7
100.0/14.2 100.0/1.7
100.0/11.5 100.0/2.7
tivity (entries 5-7). As for oxazolidinone 10, best yields and
enantioselectivities were obtained by removal of toluene
(entry 4).
We next applied these conditions to other electrophiles (Table
9). Alkylated compounds 15b, 17a,b were obtained in good
yields but slightly lower enantioselectivities.
3.4. Extension to L-Methionine. For oxazolidinone 12, re-
placement of ether by THF gave better yields and enantiose-
lectivities. Alkylation of this compound with ethyliodoacetate
gave the expected product with only 16% yield because of
competitive sulfur alkylation.27 Other electrophiles gave the
quaternary compounds 18a-c in good yields (except for
methyltriflate, for probably the same reason) and enantioselec-
tivities (Table 10).
Retention of configuration was confirmed by the crystal-
lographic structure of compound 18b (Figure 6): absolute
configuration was determined by the refinement of the Flack
parameter (Flack, 1983).28
In conclusion, all five oxazolidinones gave alkylated com-
pounds in high yields and enantioselectivities, with retention
of configuration. Best results were obtained with ethyliodoac-
etate or tert-butylbromoacetate as electrophile, and in these
cases, yields range from 75 to 96% and enantioselectivities from
89 to 96%.
4. Explanation for Observed Stereoselectivity. In order to
confirm our initial hypothesis, we used the optimized conditions
and performed an alkylation reaction on compound 5a (which
cannot present axial chirality around the Ar-CO bond). Only
racemic product 19 was obtained in 59% yield (Scheme 8). We
can conclude that there is no complex between the starting
4
5
6
L-Ala
L-Phe
L-Met
10
11
12
4
3
4
100.0/29.0 100.0/2.0
100.0/49.8 100.0/9.9
100.0/30.1 100.0/2.2
b
THF-d8
THF-d8
a Determined by integration of the labile proton in each conformer.
b Chemical shifts are slightly different because of the second aromatic
group.
oxazolidinone and the corresponding enolate that could be
responsible for asymmetric induction.
We have also previously noticed that enantioselectivity
depends on temperature, deprotonation time, and solvent.
Performing deprotonation on crystals or dissolution of crystals
at -78 °C just before reaction produced no significant change.
All these observations suggest the formation of a dynamic axial
chiral enolate.
In addition, we have obtained 1H NMR spectra of compounds
1, 10, 11 and 12 in Et2O-d10 or THF-d8 at 195 K. For compound
1, we had previously achieved complete assignment of the labile
proton in the four conformers in CDCl3 by theoretical studies,
and because of similarity between the spectra (for CDCl3, THF,
Et2O and Et2O/toluene), we have extended the preceding
assignment to the other solvents and then to the other oxazo-
lidinones. Therefore, we deduced the following points: in all
cases, the (P,cis)-conformer is observed in solid state and
corresponds to the major conformer in solution; the (P,trans)-
and (M,trans)-conformers can often be neglected (except for
compound 11, derived from phenylalanine), and various (P,cis):
(M,cis) ratios can be determined (Table 11). We also supposed
that oxazolidinone 9 (derived from L-leucine) behaved in the
same manner.
(27) (a) Labuschagne, A. J.; Malherbe, J. S.; Meyer, C. J.; Schneider, D. F.
J. Chem. Soc., Perkin Trans. 1 1978, 955–961. (b) Lawecka, J.;
Bujnicki, B.; Drabowicz, J.; Rykowski, A. Tetrahedron Lett. 2008,
49, 719–722.
Since these ratios cannot explain the high conversion rates
and enantioselectivities and also since reactions on crystals were
unsuccessful, we suggest that a dynamic kinetic resolution is
(28) See crystallographic data in Supporting Information.
9
J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009 10717

A R T I C L E S
Branca et al.
Scheme 9. Explanation of Observed Stereoselectivity
sible in only one step starting from the corresponding amino
acid, and their synthesis uses only simple organic compounds
(naphthoyl chloride and acetone). We have developed suitable
conditions for quaternization of these compounds and performed
efficient alkylation with reactive electrophiles. This methodology
was found to be general to several amino acids included the
functionalized methionine. Thus, we have developed an efficient
three-step synthesis of enantioenriched R-methylvaline (ee )
94%), R-isopropylaspartic acid (ee > 99%), R-methylaspartic
acid (ee > 99%) and R-isobutylaspartic acid (ee > 99%) without
an external source of chirality. Therefore, we have performed
alkylation of oxazolidinones by memory of chirality as defined
by Carlier, and shown the efficiency of using dynamic axial
chirality of tertiary aromatic amides in the field of memory of
chirality. We have also proposed an explanation for the observed
stereoselectivities, based on 1H NMR spectra at low temperature,
DFT calculations and crystallographic structures. Further NMR
and DFT studies are under investigation in our laboratory to
quantify rotation barrier, as well as extension of this methodol-
ogy to other amino acids, and to aldolization reactions.
responsible for stereochemical induction (Scheme 9). Ar-CO
rotation in compound 20 (between the two diastereomeric
conformers (P,cis,S) and (M,cis,S)) should be faster than
metalation and lead to preferential deprotonation of the (P,cis)-
conformer in which the labile proton is more accessible.
Racemization of the resulting enolate 21 (Ar-CO rotation)
should be relatively slow (even very slow for enolates derived
from compounds 10 and 11), and alkylation should occur
opposite to the second aromatic ring, leading to global retention
of configuration. For compound 11, the presence of a trans-
conformer can perhaps explain the lower enantioselectivities
observed.
Acknowledgment. This research was supported by Universite´
de Paris-Sud, CNRS and Ministe`re de l’Enseignement Supe´rieur
et de la Recherche (doctoral grant to M.B.). We thank J.-P. Baltaze
and A. Dos Santos for NMR analysis.
Supporting Information Available: Crystallographic data and
1H NMR at low temperature of compounds 10, 11 and 12,
experimental details and characterization of all new compounds.
This material is free of charge via the Internet at http://
pubs.acs.org.
Conclusion
In conclusion, we have demonstrated that amino acid-derived
N-aroyloxazolidinone are valuable substrates for alkylation
according to Memory of Chirality. These substrates are acces-
JA9039604
9
10718 J. AM. CHEM. SOC. VOL. 131, NO. 30, 2009