Dynamic Kinetic Resolution for Asymmetric Synthesis of α-Alkyl Amino Acids via Dual-Function Catalysis of Modified Cinchona Alkaloids

Article abstract of DOI:10.1055/s-2003-41500

The electronic property of the N-protecting group of N-carboxyanhydrides (NCA) is found to have a significant influence on the rate of racemization of NCA with modified cinchona alkaloids. Consequently, modified cinchona alkaloids can be applied as dual-function catalysts to catalyze both the racemization and alcoholytic kinetic resolution of alkyl NCA bearing an electron-withdrawing N-protecting group, leading to a dynamic kinetic resolution converting racemic alkyl NCAs to the corresponding amino esters in 59-75% ee and 75-87% yield. Upon recrystallization, the amino ester can be obtained in greater than 99% ee and 53% yield from the racemic NCA.

Full text of DOI:10.1055/s-2003-41500

CLUSTER
1927
Dynamic Kinetic Resolution for Asymmetric Synthesis of a-Alkyl Amino Acids
via Dual-Function Catalysis of Modified Cinchona Alkaloids
A
symmetr
i
ic Syn
a
a
-
n
lkyl
A
mino
f
Acids eng Hang, Li Deng*
Department of Chemistry, Brandeis University, 415 South St., Waltham, MA 02454-9110, USA
Fax +1(781)362516; E-mail: deng@brandeis.edu
Received 21 July 2003
the overall efficiency of a DKR.6 The former is reflected
approximately by the ee of the product of a DKR at low
conversion while the latter is indicated by the change in
the ee of the product as the reaction proceeds from low to
Abstract: The electronic property of the N-protecting group of N-
carboxyanhydrides (NCA) is found to have a significant influence
on the rate of racemization of NCA with modified cinchona
alkaloids. Consequently, modified cinchona alkaloids can be
applied as dual-function catalysts to catalyze both the racemization complete conversion. Unless already ideal, the efficiency
and alcoholytic kinetic resolution of alkyl NCA bearing an electron-
withdrawing N-protecting group, leading to a dynamic kinetic
resolution converting racemic alkyl NCAs to the corresponding
racemization of NCA via a deprotonation of 1 to form an
amino esters in 59–75% ee and 75–87% yield. Upon
recrystallization, the amino ester can be obtained in greater than
of a DKR can be improved if either kf/ks or krac/ks
increases. Modified cinchona alkaloids catalyze the
enolate 3 and subsequently reprotonation of 3 (Scheme 1).
Thus the significantly less acidic a-hydrogen of NCAs 1
bearing an alkyl substituent, compared to that of NCAs 1
bearing an aryl substituent, leads to a reduced rate of
racemization, thereby rendering the development of an
efficient DKR of alkyl NCA a considerable challenge.
99% ee and 53% yield from the racemic NCA.
Key words: kinetic resolution, dynamic kinetic resolution,
cinchona alkaloid, amino acids, racemization
Converting both enantiomers of a starting racemic mix-
ture into a single optically active product, i.e. dynamic
kinetic resolution (DKR) is the most desirable form of ki-
netic resolution.1 However, the development of efficient
DKR represents one of the most challenging endeavors in
O-
O
O
H
H
*
*
R
R
R
NR3
NR3
+
*
+
HNR3
O
O
O
krac
krac
PN
PN
PN
O
O
O
O
S-1
3
R-1
asymmetric catalysis.2 In addition to
a
highly
*
*
R = Alkyl
NR3
R'OH
slow
NR3
R'OH
fast
enantioselective kinetic resolution, a racemization of the
starting material that is faster than, but does not interfer
with, the kinetic resolution needs to be established under
conditions that must not cause racemization of the
product. We have recently demonstrated that organic
catalysts such as modified cinchona alkaloids can serve
dual catalytic roles, mediating both the highly
enantioselective kinetic resolution and racemization of
cyclic N- and O-carboxyanhydrides such as urethane-
protected a-amino acid N-carboxyanhydrides (UNCAs)3
and 1,3-dioxolane-2,4-diones.4,5 Using this new organic
catalyst-based platform for DKR development, we have
developed highly enantioselective methods for the
synthesis of optically active aryl a-amino acids3 and aryl
a-hydroxy acids4 via efficient DKRs of aryl N- and O-car-
boxyanhydrides. A greater challenge is to extend this
approach to the DKR of the corresponding alkyl cyclic
anhydrides. We describe here progress toward the DKR of
a-alkyl N-carboxyanhydrides (NCA) with modified
cinchona alkaloids (Scheme 1).
kf
H
ks
O
H
R
R
OR'
OR'
NHP
R-2
NHP
S-2
Scheme 1
This is illustrated by results from dynamic kinetic
resolutions of a-phenyl (1a) and a-benzyl (1b) UNCAs,
respectively (Figure 1). Slow addition of a solution of
allyl alcohol in Et2O to a solution of a-phenyl UNCA 1a
and (DHQD)2AQN in Et2O at room temperature generates
the allyl amino ester 2a in nearly constant and excellent ee
(90–91%) as the reaction proceeds from low to complete
conversion. The nearly constant ee over the course of the
DKR indicates that the racemization of 1a is much faster
than the alcoholysis of 1a with (DHQD)2AQN (krac/ks >>
1). Under the same conditions, DKR of a-benzyl UNCA
1b afforded the corresponding amino ester 2b in 82% ee
at 9% conversion, but only 15% ee at 100% conversion.
This dramatic ee decrease over the course of the reaction
indicates that the (DHQD)2AQN-catalyzed racemization
is slow relative to the enantioselective alcoholysis of
UNCA 1b (krac/ks << 1).
Both the enantioselectivity of the kinetic resolution (kf/ks)
and the relative rates of racemization and kinetic
resolution (krac/ks) are important factors in determining
SYNLETT 2003, No. 12, pp 1927–1930
2
9
.0
9
.2
0
0
3
Advanced online publication: 19.09.2003
DOI: 10.1055/s-2003-41500; Art ID: Y01003ST
© Georg Thieme Verlag Stuttgart · New York

1928
J. Hang, L. Deng
CLUSTER
We envisaged that the introduction of a N-protecting with N-dichloroacetyl NCA 1d. The ee of amino ester 2d
group of NCA that is more electron-withdrawing than the was found to be 81% ee at 30% conversion and 69% at
carbamate group could enhance the acidity of the a- 100% conversion. Given the similar ee values of amino
hydrogen of NCA 1. This in turn could accelerate the esters 2a–d at low conversion, this increased efficiency in
modified cinchona alkaloid-catalyzed racemization of 1, the DKR of 2d is most likely due to the acceleration of the
thereby improving the overall efficiency of the DKR of racemization relative to the alcoholysis of NCA 1 and is
alkyl NCAs. To test this hypothesis, we prepared racemic the result of the electron-withdrawing effect of the N-
N-acetyl NCA 1c and N-dichloroacetyl NCA 1d and protecting group. These results provided experimental
subjected them individually to (DHQD)2AQN-catalyzed evidence supporting our hypothesis that the overall effi-
alcoholysis (Scheme 2).7 The ee of the corresponding ciency of DKR of alkyl NCAs can be significantly
products, amino esters 2c,d, were monitored during the improved via the introduction of a more electron-with-
course of each alcoholysis.
drawing N-protecting group to the NCA. Attempts to
synthesize alkyl NCA bearing other electron-
withdrawing protecting groups such as
triflouroacetamide, trichloroacetamide and various
sulfonamides were, however, unsuccessful.
1
O
O
(DHQD)2AQN (0.2 eq.)
R
O
R
Allyl alcohol (1.2 eq.)
O
P
O
P
N
HN
Et2O, 4 Å M.S., r.t.
Dynamic kinetic resolution of 1d with slow addition of
allyl alcohol as described above afforded the ester 2d in
69% ee but only in 60% isolated yield, even though the
reaction went to completion. The formation of significant
amount of unidentifiable side products was detected from
NMR analysis of the crude reaction mixture. On the other
hand, we found that formations of these side products can
be suppressed significantly by maintaining a high relative
ratio of the alcohol vs. the substrate. For example,
addition of a solution of NCA 1d in Et2O to a solution of
(DHQD)2AQN (0.2 equiv) and allyl alcohol (1.2 equiv) in
Et2O provided amino ester 2d in 85% isolated yield but in
45% ee at 100% conversion of 1d. This significant
reduction of ee is expected as the higher concentration of
alcohol in this procedure increases the rate of kinetic
resolution relative to racemization of NCA 1d. These
results suggest that maintaining both a low concentration
of alcohol and a relatively high ratio of the alcohol vs.
NCA 1d through the course of the reaction is important
for achieving an optimal DKR of 1d. This consideration
prompted us to examine a double addition procedure, in
which a solution of 1d and a solution of allyl alcohol, both
in Et2O, were added simultaneously to a solution of
(DHQD)2AQN in Et2O over a period of 2 hours.8 We were
pleased to observe that DKR at room temperature with
this double addition protocol resulted in the best overall
efficiency, affording 2d in 66% ee and 81% isolated yield
(Table 1, entry 1).
O
O
1
2
a
R = Ph,
P = OBn
b R = PhCH2, P = OBn
c
d
R = PhCH2, P = CH3
R = PhCH2, P = CHCl2
Et
Et
N
N
O
O
O
O
H
H
MeO
OMe
N
N
(DHQD)2AQN
Scheme 2
A number of a-alkyl NCAs protected with dichloro-
acetamide (1d–i) were subsequently investigated under
the optimized condition. As summarized in Table 1, all
the DKRs, carried out conveniently at room temperature,
proceeded to complete conversion in 3 hours to afford
the corresponding amino esters 2d–i in 59–75% ee and
75–87% yields. Although the DKR afforded amino ester
2 in moderate ee, it could still be synthetically useful since
the optical purity of amino esters 2 can be improved upon
recrystallization. For example, a single recrystallization
of amino ester 2h from EtOAc/hexanes upgrades its opti-
cal purity from 75% ee to greater than 99% ee. Overall, 2h
was prepared in high optical purity from racemic 1h in
53% isolated yield (Table 1, entry 5).
Figure 1 Procedure: At r.t., a 1.0% (v/v) solution of allyl alcohol
(0.12 mmol, 1.2 equiv) in Et2O was added over 1 h using a syringe
pump to a mixture of substrate 1 (0.10 mmol), (DHQD)2AQN (0.02
mmol, 0.2 equiv) and 4 Å molecular sieves (10 mg) in Et2O (7.0 mL).
As shown in Figure 1, under the same conditions
described above for the DKR of UNCAs 1a,b, the
(DHQD)2AQN-catalyzed alcoholysis of N-acetyl NCA 1c
gave amino ester 2c in 78% ee at 20% conversion, which
gradually decreased to 24% ee at 100% conversion. A
dramatically improved DKR of alkyl NCA was observed
Synlett 2003, No. 12, 1927–1930 © Thieme Stuttgart · New York

CLUSTER
Asymmetric Synthesis of a-Alkyl Amino Acids
1929
Table 1 Dynamic Kinetic Resolution of NCA (1d–i) with Alcohol-
References
ysis by (DHQD)2AQNa
(1) Reviews on dynamic kinetic resolutions: (a) Robinson, D.
E. J. E.; Bull, S. D. Tetrahedron: Asymmetry 2003, 14,
1407. (b) Huerta, F. F.; Minidis, A. B. E.; Bäckvall, J.-E.
Chem. Soc. Rev. 2001, 30, 321. (c) Faber, K. Chem.–Eur. J.
2001, 7, 5005. (d) Caddick, S.; Jenkins, K. Chem. Soc. Rev.
1996, 25, 447. (e) Ward, R. S. Tetrahedron: Asymmetry
1995, 6, 1475. (f) Noyori, R.; Tokunaga, M.; Kitamura, M.
Bull. Chem. Soc. Jpn. 1995, 68, 36.
O
O
(DHQD)2AQN (0.2 eq.)
Allyl alcohol (1.0 eq.)
R
O
R
O
O
Cl2HC
N
HN
CHCl2
Et2O, 4 Å M.S.
r.t., 3 h
O
O
2d–i
1d–i
(2) (a) Jurkauskas, V.; Buchwald, S. L. J. Am. Chem. Soc. 2002,
124, 2892. (b) Liang, J.; Ruble, J. C.; Fu, G. C. J. Org.
Chem. 1998, 63, 3154. (c) Noyori, R.; Ikeda, T.; Ohkuma,
T.; Widhalm, M.; Kitamura, M.; Takaya, H.; Akutagawa, S.;
Sayo, N.; Saito, T.; Taketomi, T.; Kumobayashi, H. J. Am.
Chem. Soc. 1989, 111, 9134. (d) Schaus, S. E.; Jacobsen, E.
N. Tetrahedron Lett. 1996, 37, 7937. (e) Persson, B. A.;
Larsson, A. L. E.; Le Ray, M.; Bäckvall, J.-E. J. Am. Chem.
Soc. 1999, 121, 1645.
(3) (a) Hang, J.; Tian, S.-K.; Tang, L.; Deng, L. J. Am. Chem.
Soc. 2001, 123, 12696. (b) Hang, J.; Li, H.; Deng, L. Org.
Lett. 2002, 4, 3321.
(4) Tang, L.; Deng, L. J. Am. Chem. Soc. 2002, 124, 2870.
(5) Chen, Y.; McDaid, P.; Deng, L. Chem. Rev. 2003, 103, 2965.
(6) Kitamura, M.; Tokunaga, M.; Noyori, R. J. Am. Chem. Soc.
1993, 115, 144.
Entry
1
R
Ee of 2 (%)b,c Yield of 2 (%)d
66
67
68
70
81
82
75
82
d
2
3
4
e
F
f
Cl
Br
g
75 (99.4)e
59
87 (53)e
77
(7) Experimental Procedure for the Preparation of 1h: To a
suspension of the racemic b-(2-thienyl)-alanine (1.712 g,
10.0 mmol) in anhyd THF (16.0 mL) at 50 °C was added
triphosgene (1.01 g, 3.40 mmol, 1.02 equiv) in one portion.
After 1 h, 2 aliqouts of triphosgene (0.1 equiv/aliquot) were
added to the reaction mixture at 1 h intervals. The reaction
mixture was stirred at 50 °C for a total of 3 h, after which
time a clear solution was formed. The solution was cooled to
r.t., then concentrated to approximately 12 mL and then
poured into hexanes (40 mL). The resulting mixture was
stored in a freezer (–20 °C) overnight. The white crystals
formed during this time were collected and dried under
vacuum to give the corresponding NCA (1.763 g, 89%
yield), which was used for the next step without further
purification. To a solution of the NCA (8.94 mmol) in anhyd
THF (25.0 mL) at –30 °C, dichloroacetyl chloride (1.12 mL,
1.3 equiv) was added. A solution of N-methyl morpholine
(NMM, 1.47 mL, 1.5 equiv) in THF (5.0 mL) was introduced
dropwise to the reaction mixture over a period of 30 min.
The resulting mixture was stirred at –30 °C for a total of 3 h,
and then acidified by HCl (4.0 M in dioxane) until the pH of
the mixture was approximately 3. The resulting mixture was
allowed to warm to r.t. The precipitate (NMM
5
h
S
6f
i
a DKR procedure: At room temperature, a solution of allyl alcohol
(0.125 M, 1.60 mL, 0.20 mmol) in Et2O and a solution of 1 (0.125 M,
1.60 mL, 0.20 mmol) in Et2O were added simultaneously over 2.0 h
using a syringe pump to a mixture of (DHQD)2AQN (0.04 mmol, 0.2
equiv) and 4 Å molecular sieves (20 mg) in Et2O (12.0 mL). See ref.8
for details.
b Ee was determined using HPLC (Chiralcel OD or Chiralpak AS
column).
c The absolute configuration of 2d was assigned by comparing its
optical rotation with the value of authentic (S)-2d.
d Isolated yield.
e Results in the parentheses were obtained after recrystallization.
f Both the concentrations of allyl alcohol and 1i were 0.072 M; other
conditions were not changed.
In conclusion, we have extended the dual-function
catalysis of modified cinchona alkaloids to the dynamic
kinetic resolution of NCAs bearing alkyl substituents,
thereby establishing a synthetically useful dynamic
kinetic resolution with an organic catalyst. In order to
realize a general and broadly useful DKR approach
towards optically active alkyl amino acids, future
investigations in our laboratories will focus on the
development of new organic amine catalysts that are able
to efficiently discriminate and epimerize two enantiomers
of N-protected alkyl NCAs.
hydrochloride) was removed by filtration under N2
atmosphere through dry Celite 521 (3.0 g). The Celite was
washed with anhyd THF (10 mL). The filtrate was
concentrated and the residue was subjected to
recrystallization from Et2O/hexanes at –20 °C overnight.
The light brown solid was collected and dried under vacuum
to give the desired product 1h [2.63 g, 85% yield, 76%
overall yield from racemic b-(2-thienyl)-alanine]; mp 76–78
°C.
IR (CHCl3): 1356, 1434, 1738, 1799, 1869 cm–1.
1H NMR (400 MHz, CDCl3) d: 3.62 (dd, J = 15.3, 2.4 Hz, 1
H), 3.89 (dd, J = 15.3, 5.5 Hz, 1 H), 5.11 (dd, J = 5.5, 2.4
Hz, 1 H), 6.82–6.88 (m, 1 H), 6.94–7.00 (m, 1 H), 7.04 (s, 1
H), 7.22–7.28 (m, 1 H).
Acknowledgment
13C NMR (100 MHz, CDCl3): d = 28.7, 60.9, 63.8, 126.6,
127.8, 128.7, 131.9, 146.9, 162.2, 163.9.
We gratefully acknowledge the financial support of NIH (GM-
61591), Daiso Inc and an Alfred P. Sloan research fellowship (LD).
HRMS (EI): m/z calcd for C10H7NO4SCl2: 306.9473, found:
306.9472.
Synlett 2003, No. 12, 1927–1930 © Thieme Stuttgart · New York

1930
J. Hang, L. Deng
CLUSTER
(8) Experimental Procedure for the Dynamic Kinetic
Resolution of 1h: At r.t., a mixture of (DHQD)2AQN (0.04
mmol, 0.2 equiv) and 4 Å molecular sieves (20 mg) in anhyd
Et2O (12.0 mL) was stirred for 10 min, then a solution of 1h
in Et2O (0.125 M, 1.60 mL, 0.20 mmol) and a solution of
allyl alcohol in Et2O (0.125 M, 1.60 mL, 0.20 mmol) were
added to the mixture simultaneously over a period of 2 h by
using a syringe pump (SAGE instruments, model 355). The
resulting mixture was stirred at r.t for another hour, after
which the reaction mixture was washed with aq HCl (2.0 N,
2 × 3.0 mL) and brine (3.0 mL), and then concentrated. The
crude product was subjected to flash chromatography on
silica gel, eluting with hexanes–EtOAc (9:1). Amino ester
2h was isolated as a white solid (55.8 mg, 87% yield). The
ee was determined to be 75% by HPLC analysis [Daicel
Chiralpak AS, hexanes–isopropyl alcohol (90:10), 1.0 mL/
23
min, l 220 nm, tmajor = 19.02 min, tminor = 11.93 min]. [a]D
=
–32.7° (c 2.0, CHCl3); mp 103–105 °C.
IR (CHCl3): 1434, 1540, 1670, 1729, 3301 cm–1,
1H NMR (400 MHz, CDCl3): d = 3.40–3.54 (m, 2 H), 4.62–
4.74 (m, 2 H), 4.80–4.90 (m, 1 H), 5.26–5.44 (m, 2 H), 5.84–
6.00 (m, 2 H), 6.78–6.86 (m, 1 H), 6.90–7.00 (m, 1 H), 7.06–
7.16 (m, 1 H), 7.16–7.24 (m, 1 H).
13C NMR (100 MHz, CDCl3): d = 31.5, 53.7, 66.0, 66.7,
119.6, 125.2, 127.1, 127.2, 131.0, 136.1, 163.6, 169.6.
HRMS (EI): m/z calcd for C12H14NO3SCl2: 322.0071, found:
322.0061.
Compound 2h (55.8 mg, 75% ee) was recrystallized from
EtOAc (0.30 mL) and hexanes (1.0 mL) to give a white solid
(34.3 mg, 53% yield from 1h, 99.4% ee).
Synlett 2003, No. 12, 1927–1930 © Thieme Stuttgart · New York