An efficient kinetic resolution of racemic betti base based on an enantioselective N,O-deketalization

Article abstract of DOI:10.1021/jo051328j

An efficient kinetic resolution of racemic Betti base with L-(+)-tartaric acid in acetone was developed based on a novel enantioselective N,O-deketalization, by which the enantiopure R- and S-enantiomers of Betti base were obtained as the corresponding N,

Full text of DOI:10.1021/jo051328j

SCHEME 1
An Efficient Kinetic Resolution of Racemic
Betti Base Based on an Enantioselective
N,O-Deketalization
Yanmei Dong, Rui Li, Jun Lu, Xuenong Xu,
Xinyan Wang, and Yuefei Hu*
Department of Chemistry, Tsinghua University,
Beijing 100084, P. R. China
yfh@mail.tsinghua.edu.cn
Received June 28, 2005
Structurally, Betti base [(()-1] is a 1,3-benzylamino-
hydroxy compound and is available in bulk as hydrochlo-
ride salt [(()-1‚HCl].6 Its nonracemic derivatives have
been proven to be both excellent ligands and auxiliaries
in asymmetric syntheses featuring its unique structure,
bulky size, and high reactivity.4,5 Although its classic
resolution has been described in two similar procedures
with L-(+)-tartaric acid in alcohols,4a,7 however, most
nonracemic Betti base derivatives reported in the litera-
ture were prepared by resolution of the corresponding
racemate or by Mannich condensation with chiral amines
rather than derived from Betti base enantiomers [(R)-1
and (S)-1].4 This unusual result strongly implies that the
resolution of racemic Betti base [(()-1] may encounter
some obstacles that are difficult to overcome thus far.
When we repeated the published procedures for the
resolution of (()-1, the enantiomers (R)-1 and (S)-1 were
obtained in unsatisfactory chemical and optical yields.
The controlled experiments revealed that Betti base [(()-
1] automatically carried out a retro-Mannich reaction in
protonic solvents at room temperature. For example, an
imine byproduct 1-[phenyl[(E)-(phenylmethylene)amino]-
methyl]-2-naphthalenol in about 5% yield was detected
in MeOH within 1 h (monitored by 1H NMR, Scheme 1).
Herein, we report an efficient kinetic resolution of
racemic Betti base [(()-1] with L-(+)-tartaric acid in
acetone based on a novel enantioselective N,O-deketal-
ization, by which the enantiopure R- and S-enantiomers
of Betti base were obtained as the corresponding N,O-
ketal (R)-2 and salt (S)-3 in excellent yields with a
practically foolproof operation.
An efficient kinetic resolution of racemic Betti base with
L-(+)-tartaric acid in acetone was developed based on a novel
enantioselective N,O-deketalization, by which the enan-
tiopure R- and S-enantiomers of Betti base were obtained
as the corresponding N,O-ketal compound and salt with
L-(+)-tartaric acid, respectively, in excellent yields with a
practically foolproof operation.
The catalytic and auxiliary-based asymmetric synthe-
ses have been recognized as two major methods for the
preparation of an enantioenriched compound in modern
organic synthesis, in which a chiral compound is essential
to serve as a ligand or an auxiliary in practice. Many
chiral aminohydroxy compounds are excellent ligands in
catalytic asymmetric synthesis,1 but only a few chiral
benzylaminohydroxy compounds are also excellent aux-
iliaries in auxiliary-based asymmetric synthesis due to
their ability to dissociate benzyl residue from the induced
products under N-debenzylation conditions.2-5 Prominent
among them is the artificial chiral pool compound Betti
base, which is gaining progressive importance.4,5
* Author to whom correspondence should be addressed. Phone: +86-
10-62795380. Fax: +86-10-62771149.
(1) For selected reviews, see: (a) Soai, K.; Niwa, S. Chem. Rev. 1992,
92, 833-856. (b) Ager, D. J.; Prakash, I.; Schaad, D. R. Chem. Rev.
1996, 96, 835-875. (c) Corey, E. J.; Helal, C. J. Angew. Chem., Int.
Ed. 1998, 37, 1986-2012. (d) Pu, L.; Yu, H. B. Chem. Rev. 2001, 101,
757-824.
(2) For selected reviews, see: (a) Husson, H.-P.; Royer, J. Chem.
Soc. Rev. 1999, 28, 383-394. (b) Groaning, M. D.; Meyers, A. I.
Tetrahedron 2000, 56, 9843-9873.
(3) (a) Yamazaki, N.; Dokoshi, W.; Kibayashi, C. Org. Lett. 2001, 3,
193-196. (b) Ito, T.; Yamazaki, N.; Kibayashi, C. Org. Lett. 2002, 4,
2469-2472.
(4) For selected references, see: (a) Cardelliccnio, C.; Ciccarella, G.;
Naso, F.; Schingaro, E.; Scordari, F. Tetrahedron: Asymmetry 1998,
9, 3667-3675. (b) Cardellicchio, C.; Ciccarella, G.; Naso, F.; Perna, F.;
Tortorella, P. Tetrahedron 1999, 55, 14685-14692. (c) Palmieri, G.
Tetrahedron: Asymmetry 2000, 11, 3361-3373. (d) Liu, D.; Zhang, L.;
Wang, Q.; Da, C.; Xin, Z.; Wang, R.; Choi, M. C. K.; Chan, A. S. C.
Org. Lett. 2001, 3, 2733-2735. (e) Cimarelli, C.; Mazzanti, A.; Palmieri,
G.; Volpini, E. J. Org. Chem. 2001, 66, 4759-4765. (f) Wang, Y.; Li,
X.; Ding, K. Tetrahedron: Asymmetry 2002, 13, 1291-1297. (g) Ji, J.;
Qiu, L.; Yip, C. W.; Chan, A. S. C. J. Org. Chem. 2003, 68, 1589-
1590.
To avoid using protonic solvents in the resolution of
(()-1, some nonprotonic solvents, such as THF, CH3CN
or acetone, were scanned. When acetone was used as a
solvent in the presence of 1.0 equimolar of L-(+)-tartaric
acid at room temperature for 6 h, the desired enantiopure
salt (S)-3 as white crystal was collected in 48% yield by
(5) (a) Lu, J.; Xu, X.; Wang, C.; He, J.; Hu, Y.; Hu, H. Tetrahedron
Lett. 2002, 43, 8367-8369. (b) Lu, J.; Xu, X.; Wang, S.; Wang, C.; Hu,
Y.; Hu, H. J. Chem. Soc., Perkin Trans. 1 2002, 2900-2903. (c) Xu,
X.; Lu, J.; Li, R.; Ge, Z.; Dong, Y.; Hu, Y. Synlett 2004, 122-124. (d)
Xu, X.; Lu, J.; Dong, Y.; Li, R.; Ge, Z.; Hu, Y. Tetrahedron: Asymmetry
2004, 15, 475-479. (e) Dong, Y.; Sun, J.; Wang, X.; Xu, X.; Cao, L.;
Hu, Y. Tetrahedron: Asymmetry 2004, 15, 1667-1672. (f) Wang, X.;
Dong, Y.; Sun, J.; Xu, X.; Li, R.; Hu, Y. J. Org. Chem. 2005, 70, 1897-
1900.
(6) Betti, M. Organic Syntheses; Wiley & Sons: New York, 1941;
Collect. Vol. 1, pp 381-383.
(7) Betti, M. Gazz. Chim. Ital. 1907, 36, 392-395.
10.1021/jo051328j CCC: $30.25 © 2005 American Chemical Society
Published on Web 09/21/2005
J. Org. Chem. 2005, 70, 8617-8620
8617

SCHEME 2
SCHEME 4
SCHEME 3
TABLE 1. Effect of L-(+)-Tartaric Acid on the
Resolutiona
(R)-2
(S)-3
L-(+)-tartaric
acid (mol)
entry
yield (%) ee (%)b yield (%) ee (%)c
1
2
3
4
1.0
47
47
59
52
>99
>99
82
48
48
38
45
>99
>99
>99
>99d
0.75
0.55
0.55
88
a With 0.5 equimolar H2O for 6 h at room temperature. b ee was
determined by HPLC analysis using a chiral phase column
[Hypersil Pirkle(S) Napht]. c ee was determined as (S)-1 by 1H
NMR in the presence of (R)-(-)-3,5-(dinitrobenzoyl)-R-methylben-
zylamine. d 12 h was used.
filtration. Its structure and configuration were confirmed
by single-crystal X-ray analysis. Interestingly, instead of
the expected salt (R)-3, the enantiopure N,O-ketal com-
pound (R)-2 was obtained from the filtrate in 47% yield
(Scheme 2).
By carefully monitoring the whole process, we observed
that two white precipitates appeared successively in the
course of the resolution. The first appeared immediately
after the acetone solution of L-(+)-tartaric acid was
dropped into the acetone solution of (()-1. The other
formed once the first precipitate disappeared (about 30
min) and was easily recognized as the salt (S)-3. To our
great surprise, the first precipitate structure was as-
signed as racemic N,O-ketal compound (()-2 by its NMR
spectra and optical rotation.
enantiopure (R)-2 and (S)-3 were obtained in 47 and 48%
yields, respectively. As shown in Scheme 4, we propose
that N,O-ketal (S)-2 is initially converted into imine (S)-4
by a ring-chain tautomeric equilibrium9 (see NMR spec-
tra of (()-2 in the Supporting Information). Then,
another equilibrium between imine (S)-4 and (S)-Betti
base [(S)-1] was achieved in the presence of water under
acidic conditions. Since the salt (S)-3 is formed as an
extremely low soluble crystal immediately after (S)-1
appeared, it causes those two equilibriums to irreversibly
shift to the right. As a result, the (S)-2 is N,O-deketalized
enantioseletively in high yield.
Thus far, all the results have shown clearly that the
resolution of (()-1 carried out in acetone followed a
kinetic resolution mechanism8 based on a novel enantio-
selective N,O-deketalization rather than a classic resolu-
tion mechanism. We hypothesized that the resolution
may go through three stages, and L-(+)-tartaric acid
played a different role in each stage (Scheme 3). (1) As a
normal acid, L-(+)-tartaric acid promoted an acid-
catalyzed N,O-ketalization between (()-1 and acetone to
yield (()-2 as the first precipitate. (2) As a chiral acidic
reagent, L-(+)-tartaric acid enantioselectively catalyzed
the N,O-deketalization of (S)-2 to give (S)-1 as an
intermediate. (3) As a chiral acid, L-(+)-tartaric acid
enantioselectively captured (S)-1 to form the salt (S)-3
as the second precipitate.
To prove our hypothesis, a resolution using racemic
N,O-ketal [(()-2] (prepared in 96% yield from (()-1‚HCl)
as an initial substrate was tested in both anhydrous
acetone and aqueous acetone. As was expected, no
precipitate was found after stirring the solution of (()-2
and L-(+)-tartaric acid in anhydrous acetone for 3 h.
However, a white precipitate appeared approximately 10
min after 0.5 equimolar H2O was added. Six hours later,
Since N,O-ketalization of (()-1 into (()-2 promoted by
L-(+)-tartaric acid is very clear, the details for the reso-
lution of (()-2 to (R)-2 and (S)-3 in acetone were further
studied. We observed that the resolution with 0.55
equimolar L-(+)-tartaric acid gave (R)-2 in low enantio-
meric excess and (S)-3 in low yield, even though 0.5
equimoles were theoretically required (entry 3, Table 1).
Clearly, the problem must result from the uncompleted
N,O-deketalization of (S)-2 because the concentration of
L-(+)-tartaric acid in the reaction decreased sharply as
the salt (S)-3 precipitated out. Therefore, the resolution
result can be improved by using excess L-(+)-tartaric acid
or prolonged reaction time (entries 1, 2 and 4). Luckily,
the excess amount of L-(+)-tartaric acid did not affect the
enantioselectivity of N,O-deketalization at all.
Unlike most routine kinetic resolutions, kinetic resolu-
tion of (()-2 is not sensitive to reaction time. As shown
in Table 2, the desired resolution not only can be achieved
in as short as 3 h (entry 1), but also can be kept for up to
another 9 h (entries 2 and 3). The enantioselective N,O-
deketalization of (()-2 catalyzed by L-(+)-tartaric acid
was so efficient that the relative rate kS/kR of this kinetic
resolution should be considered to be infinity.
(8) For selected reviews on kinetic resolution, see: (a) Kagan, H.
B.; Fiaud, J. C. Top. Stereochem. 1988, 18, 249-330. (b) Keith, J. M.;
Larrow, J. F.; Jacobsen, E. N. Adv. Synth. Catal. 2001, 1, 5-26.
(9) Szatmari, I.; Martinek, T. A.; Lazar, L.; Koch, A.; Kleinpeter,
E.; Neuvonen, K.; Fulop, F. J. Org. Chem. 2004, 69, 3645-3653.
8618 J. Org. Chem., Vol. 70, No. 21, 2005

TABLE 2. Effect of Reaction Time on the Resolutiona
(R)-2 (S)-3
yield (%) yield (%)
SCHEME 5
entry
time (h)
ee (%)
ee (%)
1
2
3
4
3
6
12
96
47
47
47
37
>99
>99
>99
>99
48
48
48
52
>99
>99
>99
90
a With 0.75 equimolar L-(+)-tartaric acid and 0.5 equimolar H2O
at room temperature.
TABLE 3. Effect of Amount of H2O on the Resolutiona
(R)-2
yield (%)
(S)-3
yield (%)
entry
H2O (mol)
ee (%)
ee (%)
1
2
3
0.5
1.0
10.0
47
47
46
>99
>99
>99
48
48
48
>99
>99
>99
a With 0.75 equimolar L-(+)-tartaric acid at room temperature
for 3 h.
TABLE 4. Effect of Temperature on the Resolutiona
(R)-2 (S)-3
yield (%) yield (%)
alization, by which the enantiopure R- and S-enantio-
mers were obtained as N,O-ketal compound (R)-2 and
L-(+)-tartaric acid salt (S)-3, respectively, in excellent
chemical and optical yields with a practically foolproof
operation.
entry
temp (°C)
ee (%)
ee (%)
1
2
3
4
0
25
35
45
47
47
47
37
>99
>99
>99
>99
48
48
48
43
>99b
>99
>99
>99
Experimental Section
a With 0.75 equimolar L-(+)-tartaric acid and 0.5 equimolar H2O
for 3 h. b 6 h was used.
Kinetic Resolution of (()-1. To a stirred solution of (()-1
(49.9 g, 200 mmol) in acetone (300 mL) was added a solution of
L-(+)-tartaric acid (22.6 g, 150 mmol) in acetone (400 mL). After
the resultant mixture was stirred at room temperature for 3 h,
the white solid was collected by filtration and was washed with
acetone (2 × 50 mL) to give (S)-3 (38.3 g, 48%); mp 186-188
Theoretically, the enantioselective N,O-deketalization
of (()-2 required 0.5 equimolar H2O only. In practice,
the tests in Table 3 proved that an excess amount of H2O
did not have any bad effects on the kinetic resolution.
The enantioselective ability of the N,O-deketalization for
enantiomer (S)-2 was so strong that enantiomer (R)-2
stayed intact even in the presence of 20 times more H2O
(entry 3). Thus, commercially available acetone in dif-
ferent grades can be used directly for the resolution.
Finally, the effect of temperature on the resolution was
tested. As shown in Table 4, the excellent resolution
results were obtained at a wide range of temperatures.
Room temperature gave the best results, while prolonged
time was noted when the resolution ran at 0 °C.
According to the above experimental results, the
optimized conditions were established in Scheme 5. The
method offered an excellent procedure for the kinetic
resolution of racemic Betti base [(()-1] under very mild
and undemanding conditions. Since the salt (S)-3 is more
suitable for storage than (S)-1 and can be used as an
equivalent of (S)-1 in most cases, it usually is not
necessary to convert salt (S)-3 into (S)-1. However, when
the suspension of salt (S)-3 in CH2Cl2 was neutralized
with saturated aqueous solution of Na2CO3 at 0 °C, the
desired product (S)-1 was obtained conveniently in 99%
yield without any loss of enantiomeric excess. Under the
same conditions, enantiopure (R)-1 was obtained in
almost quantitative yield by the conversion of (R)-2 into
salt (R)-5 with D-(-)-tartaric acid followed by neutraliza-
tion with aqueous Na2CO3.
°C; [R]25 +44.1 (c 1.0, DMSO); IR: ν 3489, 3264, 3139, 3040,
D
2936, 1618, 1522, 1507, 1438 cm-1; 1H NMR (DMSO-d6): δ 7.95-
7.99 (d, 1H), 7.78-7.80 (m, 2H), 7.15-7.48 (m, 9H), 6.18 (s, 1H),
4.05 (s, 2H); 13C NMR (DMSO-d6): δ 174.3 (2C), 155.1, 139.3,
131.9, 130.1, 128.7 (2C), 128.5, 127.9, 127.8, 127.3 (2C), 126.9,
122.6, 121.7, 119.5, 114.7, 72.0 (2C), 52.1; MS m/z (%): 249 (1.5),
232 (37), 231 (100), 144 (89), 105 (39), 104 (61); Calcd for C21H21
-
NO7: C, 63.15%; H, 5.30%; N, 3.51%. Found: C, 63.01%; H,
5.37%; N, 3.51%.
To the filtrate was added a saturated aqueous solution of
Na2CO3 (20 mL). After most of the acetone was removed by
rotavapor, the residue was diluted with CH2Cl2 (200 mL) and
H2O (200 mL). The separated aqueous layer was extracted with
CH2Cl2, and the combined organic layers were washed with H2O
and brine and dried over Na2SO4. Removal of the solvent gave
crude product, which was purified by recrystallization (EtOAc)
to give compound (R)-2 (27.2 g, 47%) as colorless crystals; mp
148-150 °C; [R]25 +33.7 (c 1.0, CHCl3), [R]25 +63.5 (c 4.0,
D
D
C6H6).
Preparation of (()-3,3-Dimethyl-1-phenyl-2,3-dihydro-
1H-naphtho[1,2-e][1,3]oxazine [(()-2]. To a stirred suspen-
sion of (()-1‚HCl (57.2 g, 200 mmol) in acetone (500 mL) was
added Et3N (20.2 g, 200 mmol) at 0 °C. Twenty minutes later,
H2SO4 (98%, 0.1 mL) was added. The resultant mixture was
stirred at room temperature for 1 h (monitored by TLC), and
saturated aqueous solution of Na2CO3 (10 mL) was added. After
most acetone was removed by rotavapor, the residue was diluted
with CH2Cl2 (200 mL) and H2O (200 mL). The separated aqueous
layer was extracted with CH2Cl2, and the combined organic
layers were washed with H2O and brine and dried over
Na2SO4. Removal of the solvent gave crude product, which was
purified by recrystallization (EtOAc) to give compound (±)-2
(55.6 g, 96%) as colorless crystals; mp 124-126 °C; IR: ν 2986,
1620, 1512, 1454 cm-1; 1H NMR (CDCl3): δ 7.69-7.72 (m, 2H),
7.06-7.25 (m, 9H), 5.57 (s, 1H), 1.54 (s, 3H), 1.53 (s, 3H); 13C
In summary, an efficient kinetic resolution of racemic
Betti base [(()-1] with L-(+)-tartaric acid in acetone was
developed based on a novel enantioselective N,O-deket-
J. Org. Chem, Vol. 70, No. 21, 2005 8619

NMR: δ 152.3, 143.5, 131.3, 129.2, 129.1, 128.8 (2C), 128.3 (2C),
128.2, 127.4, 125.7, 123.9, 122.7, 119.7, 114.0, 86.7, 54.0, 28.9,
23.2; MS m/z (%): 289 (M+, 6.7), 232 (42), 231 (100), 202 (20),
42 (24); Calcd for C20H17NO: C, 83.01%; H, 6.62%; N, 4.84%.
Found: C, 83.23%; H, 6.85%; N, 4.36%.
Kinetic Resolution of (()-2. To a stirred solution of (±)-2
(55.0 g, 190 mmol) in acetone (300 mL) was sequentially added
a solution of L-(+)-tartaric acid (21.4 g, 142.5 mmol) in acetone
(400 mL) and H2O (1.71 g, 95 mmol). After the resultant mixture
was stirred at room temperature for 3 h, the white solid was
collected by filtration and was washed with acetone (2 × 50 mL)
to give (S)-3 (36.4 g, 48%). By using the exact same procedure
in the kinetic resolution of (()-1, compound (R)-2 was obtained
(25.8 g, 47%).
Preparation of (R)-(-)-Betti Base [(R)-1]. To a stirred
solution of (R)-2 (5.8 g, 20 mmol) in acetone (30 mL) was
sequentially added a solution of D-(+)-tartaric acid (4.5 g, 30
mmol) in acetone (40 mL) and H2O (36 mg, 20 mmol). After the
resultant mixture was stirred at room temperature for 3 h, the
white solid was collected by filtration and was washed with
acetone (2 × 20 mL) to give (R)-5 (7.8 g, 98%); mp 186-189 °C;
[R]25 -44.5 (c 1.0, DMSO). Its spectra data are same as those
D
of (S)-3.
To a stirred suspension of (R)-5 (4.0 g, 10 mmol) in CH2Cl2
(30 mL) was added an aqueous solution of Na2CO3 (10%, 30 mL)
at 0 °C. The resultant mixture was stirred until the solid
disappeared completely (30 min). The separated aqueous layer
was extracted with CH2Cl2, and the combined organic layers
were washed with H2O and brine and dried over Na2SO4.
Removal of the solvent gave compound (R)-1 (2.46 g, 99%) as
Preparation of (S)-(+)-Betti Base [(S)-1]. To a stirred
suspension of (S)-3 (6.0 g, 15 mmol) in CH2Cl2 (30 mL) was
added an aqueous solution of Na2CO3 (10%, 30 mL) at 0 °C. The
resultant mixture was stirred until the solid disappeared
completely (30 min). The separated aqueous layer was extracted
with CH2Cl2, and the combined organic layers were washed with
H2O and brine and dried over Na2SO4. Removal of the solvent
gave compound (S)-1 (3.7 g, 99%) as colorless crystals; mp 133-
134 °C (lit.3a 136-137 °C); [R]25 +94.1 (c 1.0, CHCl3), [R]25
colorless crystals; mp 133-135 °C; [R]25 -94.3 (c 1.0, CHCl3).
D
Its spectra data are same as those of (S)-1.
Acknowledgment. We are grateful to the National
Natural Science Foundation of China for financial
support.
D
D
+56.6 (c 4.0, C6H6); [lit.3a +56 °C (c 4.4, C6H6)]; IR: ν 3360, 3290,
1
3010, 1620, 1600, 1462, 1450 cm-1; H NMR (CDCl3): δ 7.60-
Supporting Information Available: 1H NMR and 13C
NMR spectra for all compounds and CIF file for single-crystal
X-ray analysis structures of (S)-3. This material is available
free of charge via the Internet at http://pubs.acs.org.
7.70 (m, 3H), 7.10-7.40 (m, 8H), 5.95 (s, 1H), 2.39 (br. s, 2H);
13C NMR (CDCl3): δ 156.9, 142.3, 131.9, 129.6, 128.9 (2C), 128.7,
128.4, 127.8, 127.2 (2C), 126.4, 122.4, 121.2, 120.4, 115.1, 55.8;
MS m/z (%): 249 (M+, 0.05), 232 (37), 231 (100); Calcd for C17H15-
NO: C, 81.90%; H, 6.06%; N, 5.62%. Found: C, 81.72%; H,
5.88%; N, 5.70%.
JO051328J
8620 J. Org. Chem., Vol. 70, No. 21, 2005