Highly stereoselective synthesis of anti-N-protected-α-amino epoxides

Article abstract of DOI:10.1021/jo010319h

A simple and efficient method for the synthesis of anti-N-protected amino epoxides from carbamate-protected amino acids is described. The two key steps are the monobromination of a β-ketoester and chelation-controlled reduction of a bromomethyl ketone intermediate. Good overall yields, high diastereoselectivity, and excellent functional group compatibility are characteristic.

Full text of DOI:10.1021/jo010319h

5790
J . Org. Chem. 2001, 66, 5790-5795
High ly Ster eoselective Syn th esis of a n ti-N-P r otected -r-Am in o
Ep oxid es
Robert V. Hoffman,* Warren S. Weiner, and Najib Maslouh
Department of Chemistry and Biochemistry, New Mexico State University, North Horseshoe Drive,
Las Cruces, New Mexico 88003-8001
rhoffman@nmsu.edu
Received March 26, 2001
A simple and efficient method for the synthesis of anti-N-protected amino epoxides from carbamate-
protected amino acids is described. The two key steps are the monobromination of a â-ketoester
and chelation-controlled reduction of a bromomethyl ketone intermediate. Good overall yields, high
diastereoselectivity, and excellent functional group compatibility are characteristic.
Sch em e 1
In tr od u ction
Amino epoxides are versatile intermediates for the
synthesis of a variety of densely functionalized com-
pounds such as amino sugars, oxygenated amino acids,
and hydroxyethylene peptide isosteres.1 Although a
variety of methods for their synthesis have been re-
ported,2 a very direct route is the conversion of N-
protected amino acids into N-protected amino epoxides.
This conversion requires the addition of a carbon atom,
a reduction of the carboxylate oxidation level, and control
of the stereochemistry in the epoxide ring. Three strate-
gies have commonly been used to meet these require-
ments (Scheme 1).
Sch em e 2
Conversion of the carboxylate to a vinyl group (often
via the amino aldehyde) followed by peracid epoxidation
can be used to give the syn-amino epoxide.3 Reduction of
the carboxylate to the aldehyde and reaction with sulfur
ylids also has been used to produce the syn-amino-
epoxide.4 Finally, conversion of the carboxylate to a
halomethyl ketone, stereocontrolled reduction of the
ketone with Felkin-Anh control, and ring closure to the
syn-aminoepoxide have been utilized.1
To produce anti-amino epoxides, the oxygenated sec-
ondary carbon must have the opposite configuration. This
is usually achieved by using the N,N-dibenzyl protecting
group5 on the nitrogen of the amino acid (Scheme 2).
Reduction to the aldehyde and addition of a sulfonium
ylid thus produces the anti-amino epoxide.6 Alternatively,
conversion to a halomethyl ketone, stereocontrolled
reduction with chelation control, and ring closure is a
common way to access anti-amino epoxides.1,7
Although it is possible to prepare a variety of N-
protected amino epoxides by these methods, problems
with stereoselectivity, compatibility with side chain
functional groups, and the use of problematic reagent
systems (e.g., CH2N2, LiCH2Cl) highlight the need for a
more simple, general route to these compounds. Our
interest in the synthesis8 and reductions8b,9 of N-protected
R-amino ketones led to the synthetic plan shown in
Scheme 3 for the synthesis of anti-amino epoxides. There
are two key steps in the plan. Step a requires the
monobromination of a â-ketoester and step b requires the
stereoselective reduction of a protected aminoketone with
chelation control to give the anti-amino alcohol. This
synthetic approach was reported for a single compound
(1) (a) Barluenga, J .; Baragan˜a, B.; Concello´n, J . M. J . Org. Chem.
1995, 60, 6696 is an excellent summary of the literature. See also: (b)
Kurihara, M.; Ishii, K.; Kashara, Y.; Miyata, N. Tetrahedron Lett. 1999,
40, 3183. (c) Sengupta, S.; Sarma, D. S.; Das, D. Tetrahedron Assymetry
1999, 10, 1653.
(2) Albeck, A. Drug Dev. Res. 2000, 50, 425 is an extensive recent
summary.
(3) Albeck, A.; Persky, R. J . Org. Chem. 1994, 59, 653.
(4) (a) Ashton, W. T.; Cantone, C. L.; Meurer, L. C.; Tolman, R. L.;
Greenlee, W. J .; Patchett, A. A.; Lynch, R. J .; Schorn, T. W.; Strouse,
J . F.; Siegl, P. K. S. J . Med. Chem. 1992, 35, 2103. (b) Moore, W. J .;
Luzzio, F. A. Tetrahedron Lett. 1995, 36, 6599.
(5) The N,N-dibenzyl protecting group enforces Felkin-Anh stereo-
selection on reactions at an adjacent carbonyl group. Reetz, M. T.
Angew. Chem., Int. Ed. Eng. 1991, 30, 1531.
(6) (a) Reetz, M. T.; Binder, J . Tetrahedron Lett. 1989, 30, 5425.
See also: Concello´n, J . M.; Bernad, P. L.; Pe´rez-Andre´s, J . A. J . Org.
Chem. 1997, 62, 8902.
(7) (a) Rotella, D. P. Tetrahedron Lett. 1995, 36, 5453. (b) Albeck,
A.; Fluss, S.; Persky, R. J . Am. Chem. Soc. 1996, 118, 3591. (c) Albeck,
A.; Estreicher, G. I. Tetrahedron 1997, 53, 5325.
(8) Amino ketones in the synthesis of sphingosines: (a) Hoffman,
R. V.; Tao, J . Tetrahedron Lett. 1998, 39, 3953. (b) Hoffman, R. V.;
Tao, J . J . Org. Chem. 1998, 63, 3979. Aminoketones in the synthesis
of peptidomimetics: (c) Hoffman, R. V.; Tao, J . Tetrahedron 1997, 53,
7119. (d) Hoffman, R. V.; Tao, J . Tetrahedron, Lett. 1998, 39, 4195. (e)
Hoffman, R. V.; Tao, J . J . Org. Chem. 1999, 64, 126. (f) Hoffman, R.
V.; Kim, H.-O. Tetrahedron Lett. 1992, 33, 3579. (g) Hoffman, R. V.;
Kim, H.-O. J . Org. Chem. 1995, 60, 5107.
(9) Hoffman, R. V.; Maslouh, N. J . Org. Chem. in press.
10.1021/jo010319h CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/01/2001

Synthesis of anti-N-Protected-R-Amino Epoxides
J . Org. Chem., Vol. 66, No. 17, 2001 5791
Ta ble 1. Ra tio of Mon obr om in a tion to Dibr om in a tion of
5 by NBS
Sch em e 3
entry
solvent
time (h)
T (°C)
ratio 6:7
1
2
3
4
5
6
CH2Cl2
EtOAc
acetone
MeOH
MeOH
MeOH
18
19
20
20
12
12
25
25
25
25
25
0
2.4:1
0.9:1
2.4:1
16:1
1.7:1:0.7a
2.3:1:0.7a
a
One equivalent of NBS. Ratio is 6:7:5.
material, monobrominated â-ketoester 3a , and dibromi-
nated product 4a (eq 2). Second, the ratios of products
in the patent literature, but the yield was modest (∼30%)
and the reduction was not stereoselective.10
We felt that the simplicity of the approach warranted
a second look. We are pleased to report that this strategy
has been developed into a general, efficient, and highly
stereoselective method for the synthesis of N-protected
anti-amino epoxides from amino acids.
Resu lts a n d Discu ssion
Using the method of J oullie´,11 a series of carbamate-
protected amino acids was converted to the corresponding
tert-butyl â-ketoesters (eq 1).8c,9 This procedure is efficient
and quite general, and it scales up very well.12
2a , 3a , and 4a were not consistent from run to run. Third,
it appeared that the product ratio for a given run changed
over time. At first this was attributed to instability of
2a , since it had been reported that monobrominated
â-ketoesters are unstable to storage.16 Thus it became
necessary to learn how to reliably carry out the mono-
bromination of a â-ketoester. These results also suggest
that the patent report of the bromination of γ-amino-b-
ketoesters may be irreproducible as well.10
Benzyl acetoacetate 5 was chosen as a model â-keto-
ester substrate because it has a simple NMR spectrum
(no multiplets) and is readily available. Moreover, the
monobromo product 6 and the dibromo product 7 are
easily distinguished and quantitated by the benzyl and
methyl singlets. A solution of 5 in various solvents was
treated with 0.8 equiv of NBS at room temperature, and
the ratio of monobrominated 6 to dibrominated 7 was
measured. The limited amount of NBS (0.8 equiv) was
Small scale brominations were used to develop a
bromination protocol. Ketoester 2a was reacted with
CuBr2 in refluxing ethyl acetate and in acetonitrile at
25 °C,13 CuBr2/ PhI(OH)OTs in acetonitrile at 0 °C,14 and
NBS in ethyl acetate at 25 °C.15 NBS was found to be
superior in terms of yield and convenience. Various
solvents were evaluated, and crude yields were very high
(>94%) for all of the solvents (EtOAc, THF, MTBE,
CH3CN, CH2Cl2, methanol, acetone, and toluene). NBS
in ethyl acetate at room temperature was chosen as the
standard bromination conditions.
Reactions run at larger scales revealed that significant
problems remained. First, use of 1 equiv of NBS with 2a
gave a product mixture consisting of unreacted starting
chosen in an attempt to avoid dibromination. The results
are shown in Table 1, entries 1-4. Methanol appeared
to be the superior solvent as it gave a 16:1 ratio of 6:7.
However, when 1 equiv of NBS was used (entry 5), the
product ratio was inferior and significant amounts of
unreacted 5 remained. Lowering the temperature (entry
6) improved the ratio only slightly. Different batches of
NBS gave slightly varying ratios as well.
Because the enol form is likely the species undergoing
bromination, it appeared that the enol forms of 5 and 6,
which both are present in solution, must undergo bro-
mination at somewhat comparable rates leading to
(10) Honda, Y.; Katayama, S.; Izawa, K.; Nakazawa, M.; Suzuki,
T.; Kanno, N. U.S. Patent 5,902,887, May 11, 1999. See also: Honda,
Y.; Katayama, S.; Izawa, K.; Nakazawa, M.; Suzuki, T.; Kanno, N. U.S.
Patent 5,767,316, J une 16, 1998.
(11) (a) Harris, B. D.; Bhat, K. L.; J oullie´, M. M. Tetrahedron Lett.
1987, 25, 2737. (b) Harris, B. D.; J oullie´, M. M. Tetrahedron 1988, 44,
3489.
(12) Hoffman, R. V.; Tao, J . Methods in Molecular Medicine, Vol.
23, Peptidomimetics Protocols; Kazmierski, W., Ed.; Humana Press:
Totowa, NJ , 1999; p 103.
(13) King, L. C.; Ostrum, G. K. J . Org. Chem. 1964, 29, 3459.
(14) Coats, S. J .; Wasserman, H. H. Tetrahedron Lett. 1995, 36,
7735.
(15) Levai, L.; Ritvay-Emandity, K. Chem. Ber. 1959, 92, 2775.
(16) Shi, X.-X.; Dai, L.-X. J . Org. Chem. 1993, 58, 4596.

5792 J . Org. Chem., Vol. 66, No. 17, 2001
Hoffman et al.
Ta ble 2. Equ ilibr a tion of a 6/7/5 P r od u ct Mixtu r e in
Ta ble 3. Red u ction of Br om om eth yl Keton es 8a -g
Meth a n ol a t 25 °C
LiAlH(OtBu)3
anti:syn (% de)
NaBH4
anti:syn
initial ratio (6:7:5)
additive
none
HOAc
2,6-lutidine
2,6-lutidine
final ratio
entry
ketone
1.7:1:0.7
1.7:1:0.7
1.7:1:0.7
10.4:1:0.9
0.9:1:0.8
0.3:1:0.7
1
2
3
4
5
6.
7
8a
8b
8c
8d
8e
8f
43:1 (96)
21:1 (92)
4.6:1
5.4:1
35:1
3.8:1
7.3:1
7.0:1
3.2:1
10.4:1:0.9
>100:1a (>99)
38:1 (95)
13.2:1:0.9a
36:1 (94)
35:1 (94)
30:1 (93)
a
After storage overnight at -20 °C.
8g
mixtures of mono- and dibrominated products. Since the
product ratios seemed to change with reaction time and
with different batches of NBS, it also seemed possible
that there is a disproportionation equilibrium between
5, 6, and 7 which changed the initial product ratios.
a
Only one isomer was observed by HPLC.
Although more experiments are needed (and planned)
to understand this intriguing system, the results in Table
2 provide a synthetic solution for monobrominating
â-ketoesters. After some fine-tuning, a solution of 2a in
methanol containing 2 mol % 2,6-lutidine was treated
with 1 equiv of NBS. After 2.5 h at room temperature,
the reaction mixture was placed in the freezer for 20 h.
After workup, the crude product was found to be a
mixture of monobrominated 3a and dibrominated 4a in
the ratio of 12.8:1. A single recrystallization from hexane
gave pure monobromo 3a . De-esterification/decarboxyl-
ation was carried in high yield out by refluxing 3a in
benzene/TFA (8:1) for 2 h. This procedure was used to
convert â-ketoesters 2 to bromomethyl ketones 8 in good
yields (eq 4). In practice the monobromo â-ketoester was
To test this hypothesis, three aliquots of a product
mixture which had a 6:7:5 ratio of 1.7:1:0.7 were redis-
solved in methanol. The first contained no additives, the
second contained ∼2% acetic acid, and the third con-
tained ∼2% of 2,6-lutidine. Since it seemed likely that
any equilibrium must be mediated by the enol forms, we
sought to catalyze enolization and thus speed up the
approach to equilibrium. The mixtures were allowed to
stand for 9 h at room temperature, and the product ratios
were redetermined (Table 2).
The results are unexpected and quite puzzling. In
methanol alone the product ratio shifts in the direction
of more dibrominated and unbrominated product. With
acid present this shift is even greater, presumably as a
result of the catalytic effect of acid in increasing the rate
of disproportionation. However, most surprisingly, the
presence of base causes a shift in the product ratio to
favor the monobrominated product by a large amount.
Simply put, the presence of acid shifts the product ratio
in one direction, the presence of base shifts it in the other
direction. If acid and base were merely serving as
catalysts to achieve equilibrium more rapidly, then the
change in product ratios should be in the same direction
and should ultimately converge. This is does not seem
to be the case.
not purified but merely isolated and then decarboxylated
to the bromomethyl ketone, which was recrystallized.
Chelation-controlled reduction of bromomethyl ketones
8a -g by LiAlH(OtBu)3 in ethanol at -78 °C gave the
anti-amino bromohydrins 9a -g in excellent yields (eq 5).
On the basis of stoichiometry the disproportionation
can be written as in eq 3. Not shown are the enol/enolate
This reagent has recently been shown reduce N-carbam-
ate-protected aminoketones to anti-amino alcohols with
very high anti diastereoselectivity.9 Only a single dia-
stereomer was observed in the NMR spectra of the crude
products, and this was assumed to be the anti isomer.
This assumption was proven valid by conversion to
known anti-amino epoxides (vide infra). We also assume
that the enantiomeric purity of the products is high as
well. We have prepared amino ketones from amino acids
by this general route in several different contexts and,
with the exception of phenyl glycine, have never observed
significant epimerization of the amino group.8
Diastereomeric mixtures of 9a -g were prepared for
comparison purposes by reduction of 8a -g with NaBH4
in ethanol at -78 °C. The anti isomer was the major
product, but the syn isomer could clearly be seen in the
NMR spectrum. The diastereoselectivity was quantitated
tautomers that are also present in the mixture. One
possible explanation for the divergent behavior with acid
or base catalyst is that in the presence of base, a greater
amount of the more acidic monobromo 6 is converted to
the enolate and thus the equilibrium is shifted to the left.
This system must be very sensitive to environmental
effects, because only small amounts of a relatively weak
acid or base apparently causes large changes in the
product ratios. Moreover, cooling the solution overnight
at -20 °C causes the 6:7:5 ratio to improve to 13.2:1:0.9.
Treatment of 5 with N-chlorosuccinimide and 2,6-
lutidine in methanol leads to a 1.1:1 ratio of monochlo-
rinated to dichlorinated products. Moreover, the ratio
does not change over the course of 13 h, indicating that
disproportionation is not occurring. Thus chlorination of
2 is not a viable alternative for bromination in the
preparation of amino epoxides by this strategy.17
(17) For an alternate route to analogous chloromethyl ketones, see:
Wang, X.; Thottathil, J . K.; Polniaszek, R. P. Synlett 2000, 902.

Synthesis of anti-N-Protected-R-Amino Epoxides
J . Org. Chem., Vol. 66, No. 17, 2001 5793
47.3, 37.2, 28.0. Anal. Calcd for C30H31NO5: C, 74.21; H, 6.43;
N; 2.88. Found: C, 74.25; H, 6.59; N, 3.00.
by HPLC, and the results are collected in Table 3. The
data in Table 3 are for crude products. Recrystallization
improves the ratio even more.
Treatment of amino bromohydrins 9 with potassium
carbonate in methanol gave the corresponding amino-
epoxide 10 in nearly quantitative yields (eq 6). Protected
Br om o â-Ketoester 3a . Solid NBS (178 mg, 1.31 mmol)
was added in one portion to a room-temperature solution of
2a (520 mg, 1.31 mmol) and 2,6-lutidine (15 mL, 0.13 mmol,
10 mol %) in methanol (2.6 mL). After stirring for 2.5 h, the
mixture was stored at -20 °C for 19 h. Ethyl acetate (20 mL)
was added, and the mixture was extracted with 50% saturated
NaCl (3 × 8 mL) and saturated NaCl (2 × 8 mL), dried with
sodium sulfate, and evaporated to give a white crystalline
product (96% yield) that was a 15.1:1:1.2 mixture of 3a :4a :
2a . Recrystallization from hexane gave pure 3a (71%) as a
1.7:1 mixture of diastereomers: mp 78.7-81.5 °C. The two
diastereomers had characteristic tert-butyl peaks at 1.43 and
1.47d. Anal. Calcd for C23H26BrNO5: C, 57.99; H, 5.50; N, 2.94.
Found: C, 58.13; H, 5.65; N, 3.09.
amino epoxides corresponding to 10a ,18 10b,19 10c,7a
10e,7c 10f,7c and 10g20 have all been reported recently in
the literature. Thus epoxide formation from 9 was
demonstrated for only a few representative examples,
and these were found to have the anti stereochemistry.
This procedure for the preparation of protected amino
epoxides is simple, efficient, and highly diastereoselec-
tive. It appears to be compatible with a variety of
functionalized side chains and can be used with Cbz and
Fmoc protecting groups (and presumably other acid-
stable carbamate protecting groups).
Br om om eth yl Keton e 8a . Ketoester 3a (601 mg, 1.26
mmol) was dissolved in 8:1 benzene/TFA (9 mL) and heated
at reflux for 2 h. After cooling to room temperature, the
mixture was extracted with water (2 × 4 mL), saturated
NaHCO3 (2 × 4 mL), and saturated NaCl (2 × 4 mL), dried
(MgSO4), and evaporated to give an orange crystalline product.
Flash chromatography (200:1, CH2Cl2/i-PrOH) furnished 8a
1
as a white solid (84%): mp 95.7-96.5, H NMR δ 7.4-7.1 (m,
10 H), 5.30 (d(br), J ) 8.1 Hz, 1 H), 5.08 (s, 2 H), 4.82 (m, 1
H), 3.92 (d, J ) 13.6 Hz, 1 H), 3.82 (d, J ) 13.7 Hz, 1 H), 3.12
(dd, J ) 5.5, 12.8 Hz, 1 H), 3.04 (dd, J ) 6.0, 12.7 Hz, 1 H);
13C NMR δ 200.4, 155.7, 135.9, 135.5, 129.1, 128.8, 128.4,
128.2, 128.0, 127.3, 67.1, 58.8, 37.6, 33.1. These data agree
with the literature data.18
Exp er im en ta l Section
Infrared spectra were taken as a KBr pellets, as solution in
CHCl3, or as neat liquids. 1H NMR and 13C NMR spectra were
recorded at 200 and 50 MHz, respectively, in CDCl3. Thin-
layer chromatography was performed on silica gel 60 F254
plates from EM reagents and visualized by UV irradiation or
by spraying with vanillan/H2SO4 and heating on a hot plate.
Flash chromatography was performed using silica gel 60 (230-
400 mesh). Tetrahydrofuran was distilled from benzophenone
ketyl. Other solvents were HPLC grade and were used without
further purification. Starting materials were purchased from
Acros, Aldrich, or Novabiochem and used as received. LiAlH-
(OtBu)3 was purchased from Fluka. Elemental analyses were
performed by MHW Laboratories, Phoenix, AZ. â-Ketoesters
Gen er a l P r oced u r e for th e Con ver sion of â-Ketoester s
2 to Br om om eth yl Keton es 8. Solid NBS (1 equiv) was
added to a solution of the â-ketoester 2 and 2,6-lutidine (10
mol %) in methanol (2 mL/mmol). After stirring for 2.5 h at
room temperature, the mixture was stored in the freezer (-20
°C) for 20 h. The reaction mixture was extracted with 50%
sat. NaCl (3 × 8 mL) and sat. NaCl (3 × 8 mL), dried (NaSO4),
and evaporated to give 2-bromo-â-ketoester 3 as a crude
product. Crude product 3 was dissolved in 8:1 benzene/TFA
(9 mL) and refluxed for 2 h. After cooling the reaction was
extracted with water (2 × 4 mL), saturated NaHCO3 (2 × 4
mL), and saturated NaCl (2 × 4 mL), dried (MgSO4), and
evaporated. The crude bromomethyl ketone was purified by
flash chromatography
9
2a -g were prepared by literature procedures.8c,
â-Ketoester 2e was obtained as a clear, colorless oil: [R]25
D
P r ep a r a tion of 8b fr om 2b (517 mg, 1.61 mmol) by the
above procedure (elution with CH2Cl2/MeOH 200:1) gave 8b
(73%) as a white crystalline solid: mp 77.0-79.1 °C (lit. mp
+10.7 (c 1.0, CH2Cl2); IR (neat) 3335, 1713, cm-1; H NMR δ
1
7.34 (m, 10 H), 5.54 (d(br), J ) 8.3 Hz, 1 H), 5.2-5.0 (m, 4 H),
4.84 (m, 1 H), 4.45 (m, 1 H), 3.48 (d, J ) 15.6 Hz, 1 H), 3.43
(d, J ) 15.5 Hz, 1 H), 3.18 (m, 2 H), 1.91 (m, 1 H), 1.7-1.2 (m,
5 H), 1.45 (s, 9 H); 13C NMR δ 202.0, 165.9, 156.6, 156.1, 136.7,
136.2, 128.5, 128.5, 128.2, 128.1, 128.0, 82.3, 67.1, 66.6, 59.9,
47.5, 40.4, 30.4, 29.5, 27.9, 22.1. Anal. Calcd for C28H36N2O7:
C, 65.61; H, 7.08; N, 5.46. Found: C, 65.58; H, 7.03; N, 5.44.
â-Ketoester 2f was obtained as a pale tan oil: [R]24D +10.5
1
83-84 °C); H NMR δ 7.36 (s, 5 H), 5.37 (d(br), J ) 9.8 Hz, 1
H), 5.11 (s, 2 H), 4.68 (m, 1 H), 4.07 (d, J ) 13.3 Hz, 1 H), 4.04
(d, J ) 13.3 Hz, 1 H), 1.43 (d, J ) 7.2 Hz, 3 H). These data
match the NMR spectrum reported for this compound.19 NMR
examination of the crude product indicated a 39:1:1.3 ratio of
3b:4b:2b in the bromination reaction.
P r ep a r a tion of 8c fr om 2c (1.050 g, 3 mmol) by the above
procedure (elution with toluene/EtOAc 20:1) gave 8c (87%) as
1
(c 1.1, CH2Cl2); IR (neat) 3344, 1736, 1719 cm-1; H NMR δ
7.36 (m, 5 H), 5.56 (d(br), J ) 7.5 Hz, 1 H), 5.11 (s, 2 H), 4.52
(m, 1 H), 3.66 (s, 3 H), 3.54 (d, J ) 15.6 Hz, 1 H), 3.48 (d, J )
15.6 Hz, 1 H), 2.6-2.2 (m, 3 H), 1.89 (m, 1 H), 1.45 (s, 9 H);
13C NMR δ 201.5, 173.2, 165.9, 145.1, 136.2, 128.5, 128.2,
128.1, 82.4, 67.1, 59.4, 51.8, 47.4, 29.7, 27.9, 26.0. Anal. Calcd
for C20H27NO7: C, 61.06; H, 6.92; N, 3.56. Found: C, 61.00;
H, 6.93; N, 3.61.
a white solid: mp 81.5-82.0 °C; [R]24 +41.2° (c 1.0, CH2Cl2);
D
1H NMR δ 7.36 (m, 5 H), 5.31 (d(br), J ) 8.8 Hz, 1 H), 5.11 (s,
2 H), 4.61 (dd, J ) 4.5, 8.7 Hz, 1 H), 4.09 (d, J ) 13.6 Hz, 1
H), 4.03 (d, J ) 13.6 Hz, 1 H), 2.23 (m, 1 H), 1.04 (d, J ) 6.8
Hz, 3 H), 0.84 (d, J ) 6.8 Hz, 3 H); 13C NMR δ 200.6, 156.4,
136.1, 128.6, 128.3, 128.1, 67.3, 63.0, 32.7, 30.3, 19.7, 17.0.
Anal. Calcd for C14H18BrNO3: C, 51.23; H, 5.53; N, 4.27.
Found: C, 51.13; H, 5.62; N, 4.12.
â-Ketoester 2g was obtained as a off white solid: mp
111.2-113.0 °C; [R]24 +6.4 (c 1.0, CH2Cl2); 1H NMR δ 7.76
D
P r ep a r a tion of 8d fr om 2d (986.7 mg, 1.96 mmol) by the
above procedure gave 8d as a white solid (71%) after silica
(m, 2 H), 7.6-7.1 (m, 11 H), 5.34 (d(br), J ) 8.0 Hz, 1 H), 4.68
(m, 1 H), 4.43 (dd, J ) 7.0, 10.6 Hz, 1 H), 4.36 (dd, J ) 6.6,
10.5 Hz, 1 H), 4.17 (m, 1 H), 3.38 (m, 2 H), 3.18 (dd, J ) 5.9,
14.1 Hz, 1 H), 3.02 (dd, J ) 6.8, 14.1 Hz, 1 H), 1.45 (s, 9 H);
13C NMR δ 201.6, 165.9, 155.7, 143.8, 141.4, 135.9, 129.3,
128.8, 127.8, 127.2, 127.1, 125.1, 120.0, 82.4, 67.0, 60.8, 48.2,
gel chromatography eluting with toluene/EtOAc 20:1: mp
1
136.9-137.3 °C; [R]24 +17.6 (c 1.0, CH2Cl2); H NMR δ 7.6-
D
7.2 (m, 10 H), 7.06 (m, 2 H), 6.90 (m, 2 H), 5.29 (d(br), J ) 7.6
Hz, 1 H), 5.09 (s, 2 H), 5.04 (s, 2 H), 4.79 (m, 1 H), 3.92 (d, J
) 13.6 Hz, 1 H), 3.82 (d, J ) 13.6 Hz, 1 H), 3.03 (m, 2 H); 13
C
NMR δ 200.5, 158.2, 155.8, 136.9, 136.1, 130.2, 128.6, 128.4,
128.2, 128.0, 127.7, 127.5, 115.4, 70.1, 67.3, 59.0, 37.2, 33.0.
Anal. Calcd for C25H24BrNO4: C, 62.25; H, 5.01; N, 2.89.
Found: C, 62.48; H, 5.09; N, 2.80.
(18) Albeck, A.; Persky, R. Tetrahedron 1994, 50, 6333.
(19) Harbeson, S. L.; Rich, D. H. J . Med. Chem. 1984, 32, 1378.
(20) Heinsoo, A.; Rudaru, G.; Linask, K.; J a¨rvu, J .; Zetterstro¨m, M.;
Langel, U. Tetrahedron Asymmetry 1995, 6, 2245.

5794 J . Org. Chem., Vol. 66, No. 17, 2001
Hoffman et al.
P r ep a r a tion of 8e fr om 2e (805 mg, 1.57 mmol) by the
above procedure gave 8e as a white solid (74%) after purifica-
tion by silica gel chromatography eluting with 7:2 toluene/
d4) δ 2.81 (dd, 1H, J ) 9.0, 14.1 Hz), 3.01 (dd, 1H, J ) 4.4,
14.0 Hz), 3.43 (dd, 1H, J ) 6.4, 9.9 Hz), 3.51 (dd, 1H, J ) 3.9,
10.5 Hz), 3.84 (m, 1H), 3.99 (m, 1H), 5.00 (s, 2H), 7.21-7.32
(m, 10H); 13C NMR (CDCl3 /few drops of methanol-d4) δ 35.5,
35.8, 55.3, 66.6, 72.9, 126.4, 127.7, 127.9, 128.3, 129.3, 137.6,
156.3. Anal. Calcd for C18H20BrNO3: C, 57.15; H, 5.33; N, 3.70.
Found: C, 57.28; H, 5.47; N, 3.65. These data match the
literature data.21 Use of Method B gave 9a in 95% crude yield
with an anti:syn ratio of 41.8:1. Method C gave 9a in quantita-
tive yield with a 4.41:1 anti:syn ratio.
EtOAc: mp 90.5-91.0 °C (lit. mp 89-90 °C);18 [R]25 +9.6 (c
D
1
1.0, CH2Cl2); H NMR δ 7.34 (m, 10 H), 5.56 (d(br), J ) 7.3
Hz, 1 H), 5.10 (s, 2 H), 5.07 (m, 2 H), 4.81 (m(br), 1 H), 4.61
(m, 1 H), 4.05 (m, 2 H), 3.19 (m, 2 H), 2.0-1.3 (m, 6 H). Anal.
Calcd for C23H27BrN2O5: C, 56.22; H, 5.54; N, 5.70. Found:
C, 56.50; H, 5.42; N, 5.58.
P r ep a r a tion of 8f fr om 2f (1.0 g, 2.54 mmol) by the above
procedure gave 8f (84%) as a white solid. Purification by
Br om oh yd r in 9b was prepared from 8b (77 mg, 0.26 mmol)
by Method A in 92% yield. The anti:syn ratio was 20.9:1.
Recrystallization from 10:1 hexanes:/EtOAc gave a pure
recrystallization from 5:2 hexanes/EtOAc gave an analytical
1
sample: mp 104.3-104.6 °C; [R]25 +11.5 (c 1.0, CH2Cl2); H
D
sample (71%): mp 110 °C; [R]21 -6.2 (c 1.00, CHCl3); FTIR
NMR δ 7.35 (m, 5 H), 5.57 (d(br), J ) 7.6 Hz, 1 H), 5.11 (s, 2
H), 4.71 (m, 1 H), 4.12 (d, J ) 13.4 Hz, 1 H), 4.08 (d, J ) 13.4
Hz, 1 H), 3.66 (s, 3 H), 2.6-2.2 (m, 3 H), 1.94 (m, 1 H); 13C
NMR δ 200.2, 173.2, 156.1, 136.0, 128.6, 128.3, 128.1, 67.4,
57.3, 51.9, 31.7, 29.6, 26.6. Anal. Calcd for C15H18BrNO5: C,
48.40; H, 4.87; N, 3.76. Found: C, 48.49; H, 4.87; N, 3.70.
P r ep a r a tion of 8g fr om 2g (1.0 g, 2.06 mmol) by the above
procedure gave 8g (62%) as an off-white solid after purification
by silica gel chromatography eluting with CH2Cl2: mp 132 °C
(dec); [R]24D +5.4 (c 1.0, CH2Cl2); 1H NMR δ 7.77 (m, 2 H), 7.7-
7.1 (m, 11 H), 5.28 (d(br), J ) 7.4 Hz, 1 H), 4.79 (m, 1 H), 4.43
(app d, J ) 6.8 Hz, 2 H), 4.18 (app t, J ) 6.8 Hz, 1 H), 3.88 (d,
J ) 13.6 Hz, 1 H), 3.78 (d, J ) 13.6 Hz, 1 H), 3.12 (dd, J )
6.6, 13.8 Hz, 1 H), 3.04 (dd, J ) 7.0, 13.9 Hz, 1 H); 13C NMR
δ 200.3, 155.8, 143.7, 141.4, 135.6, 129.2, 129.0, 127.9, 127.4,
127.2, 125.0, 120.1, 67.0, 58.9, 47.3, 37.9, 32.8. Anal. Calcd
for C25H22BrNO3: C, 64.66; H, 4.78; N, 3.02. Found: c, 64.83;
H, 5.00; N, 2.99.
D
(KBr) 3315, 1691 cm-1; H NMR (CDCl3) δ 1.05 (d, 3H, J )
1
6.6 Hz), 3.38 (dd, 1H, J ) 3.7, 10.5 Hz), 3.47 (dd, 1H, J ) 8.1,
10.4 Hz), 3.86 (m, 1H), 3.89 (m, 1H), 5.01 (broad, 1H), 5.10 (s,
2H), 7.35 (m, 5H); 13C NMR (CDCl3) δ 15.1, 36.1, 49.9, 67.1,
74.2, 128.2, 128.3, 128.6, 136.4, 156.1. Anal. Calcd for C12H16
-
BrNO3: C, 47.70; H, 5.34; N, 4.64. Found: C, 47.72; H, 5.55;
N, 4.74. Method C gave 9b in 94% yield with an anti:syn ratio
of 5.41:1.
Br om oh yd r in 9c was prepared from 8c (100 mg, 0.31
mmol) by Method A in 94% yield. Only a single diastereomer
was observed in the HPLC of the crude product. Recrystalli-
zation from hexanes gave an analytical sample (76%): mp 74
°C; [R]21 +2.3 (c 1.00, CHCl3); FTIR (KBr) 3425, 3345, 1691
D
cm-1; 1H NMR (CDCl3) d 0.87 (d, 3H, J ) 6.9 Hz), 0.95 (d, 3H,
J ) 6.9 Hz), 2.19 (m, 1H), 2.61 (s, 1H), 3.44 (dd, 1H, J ) 7.7,
10.4 Hz), 3.62 (dd, 1H), 3.65 (m, 2H), 4.78 (d, 1H, J ) 8.9 Hz),
5.10 (s, 2H), 7.35 (s, 5H); 13C NMR (CDCl3) d 16.3, 20.2, 27.9,
38.3, 58.9, 67.2, 72.5, 128.2, 128.3, 128.6, 136.3, 156.9. Anal.
Calcd for C14H20BrNO3: C, 50.92; H, 6.10; N, 4.24. Found: C,
51.19; H, 6.20; N, 4.35. Reduction of 8c using Method C also
gave only a single diastereomer (94%) by HPLC.
Gen er a l P r oced u r e for th e Red u ction of Am in o Ke-
ton es 8a -g to Br om oh yd r in s 9a -g. Meth od A. Solid
LiAlH(OtBu)3 (2 equiv) was placed in a round-bottomed flask
and cooled at -78 °C. Absolute ethanol (6 mL/mmol) was
syringed slowly down the side of the flask, and the mixture
was stirred vigorously. A cold solution of the bromomethyl
ketone 8 in absolute ethanol (22 mL/mmol) was syringed
slowly down the side of the reaction flask, and the reaction
mixture was stirred for 1 h at -78 °C. The reaction was
quenched with 5 mL of 1 M HCl and warmed to room
temperature. The solvent was evaporated, and the slushy
residue was triturated with ethyl acetate (8 mL). The organic
extract was washed with 1 M HCl (2 × 3 mL) and then
saturated NaCl (3 mL), dried (MgSO4), and evaporated to give
the bromohydrin 9. The diastereomeric excess of the crude
product was determinedby HPLC (normal phase, 20:1 dichloro-
ethane/acetonitrile), and the crude product had excellent purity
by NMR. A single recrystallization gave pure 9.
Br om oh yd r in 9d was prepared from 8d (100 mg, 0.21
mmol) by Method B in 96% yield. The anti:syn ratio was 38.4:
1. Recrystallization from chloroform gave a pure sample
(46%): mp 159 °C; [R]21 -13.3 (c 1.00, DMF); FTIR (KBr)
D
3326, 1691, 1621 cm-1; 1H NMR (CDCl3 /few drops of methanol-
d4) δ 2.73 (dd, 1H, J ) 9.2, 14.1 Hz), 2.97 (dd, 1H, J ) 4.2,
14.1 Hz), 3.41 (dd, 1H, J ) 7.1, 10.6 Hz), 3.50 (dd, 1H, J )
4.0, 10.6 Hz), 3.83 (m, 1H), 3.92 (m, 1H), 5.01 (s, 2H), 5.03 (s,
2H), 6.83-7.41 (m, 14H); 13C NMR (CDCl3 /few drops of
methanol-d4) δ 34.4, 35.6, 55.4, 66.4, 69.9, 72.8, 114.7, 127.3,
127.7, 128.2, 128.3, 130.2, 136.9, 156.3. Anal. Calcd for C25H26
-
BrNO4: C, 61.99; H, 5.41; N, 2.89. Found: C, 62.08; H, 5.51;
N, 3.00. Method C gave 9d (100%) with an anti:syn ratio of
3.77:1.
Meth od B. In some cases the aminoketone was rather
insoluble in cold ethanol and thus a cosolvent was needed to
increase solubility. Tetrahydrofuran/ethanol 1:1 was found to
serve this purpose well without diminution of either yield or
diastereoselectivity. Solid LiAlH(OtBu)3 (2 equiv) was placed
in a round-bottomed flask and cooled at -78 °C. Tetrahydro-
furan (3 mL/mmol) and then absolute ethanol (3 mL/ mmol)
were syringed slowly down the side of the flask, and the
mixture was stirred vigorously. A cold solution of the bromo-
methyl ketone 8 in a 1:1 mixture of tetrahydrofuran/absolute
ethanol (22 mL/mmol) was syringed slowly down the side of
the reaction flask, and the reaction mixture was stirred for 1
h at -78 °C. Quenching and workup were the same as in
Method A.
Br om oh yd r in 9e was prepared from 8e (100 mg, 0.20
mmol) by Method A in 97% yield. The anti:syn ratio was 36.1:
1. Recrystallization from toluene plus hexanes gave a pure
sample: mp 105 °C; [R]21 -9.4 (c 1.00, CHCl3); FTIR (KBr)
D
3316, 3026, 2957, 2907, 1696, 1546, 1465, 1312, 1292, 1262,
1152, 1067, 1033, 918, 848, 729, 699 cm-1; 1H NMR (CDCl3) δ
1.42 (m, 6H), 2.58 (broad, 1H), 3.11 (m, 2H), 3.33 (dd, 1H, J )
7.7, 10.6 Hz), 3.42 (dd, 1H, J ) 3.5, 10.4 Hz), 3.70 (m, 1H),
4.77 (broad, 1H), 5.01 (m, ovelapping, 5H), 7.32-7.33 (m, 10H);
13C NMR (CDCl3) δ 22.7, 29.0, 29.7, 36.4, 40.5, 54.3, 66.8, 67.1,
74.2, 128.1, 128.3, 128.6, 136.4, 136.7, 156.7. Anal. Calcd for
C
23H29BrN2O5: C, 55.99; H, 5.92; N, 5.68. Found: C, 56.14;
H, 6.14; N, 5.76. Method C gave 9e (100%) with an anti:syn
ratio of 7.26:1.
Meth od C. A solution of amino ketone 8 in 1:1 tetrahydro-
furan/absolute ethanol (30 mL/mmol) was cooled to -78 °C,
and solid sodium borohydride was added in one portion. The
mixture was stirred vigorously for 20 min at -78° and
quenched with 5 mL of 1 M HCl. Workup was the same as in
Method A.
Br om oh yd r in 9a was prepared from 8a (50 mg, 0.13 mmol)
by Method A. The crude product (97%) was pure by NMR. The
ratio of diastereomers was 42.7:1 anti:syn. Recrystallization
from hot EtOAc/hexane gave a pure sample (72%) with an anti:
syn ratio of 141:1: mp 141 °C; [R]21D -13.9 (c 1.00, DMF); FTIR
(KBr) 3336, 1696 cm-1; 1H NMR (CDCl3 /few drops of methanol-
Br om oh yd r in 9f was prepared from 8f (100 mg, 0.27
mmol) by Method B in 92% yield. The anti:syn ratio was 35.4:
1. Recrystallization from 1,2-dichloroethane plus hexanes gave
a pure sample (74%): mp 100 °C; [R]21 -9.4 (c 1.00, CHCl3);
D
1
FTIR (KBr) 3475, 3316, 1730, 1696 cm-1; H NMR (CDCl3) δ
1.81 (m, 1H), 1.96 (m, 1H), 2.41 (t, 2H, J ) 7.2 Hz), 2.90 (broad,
1H), 3.41 (dd, 1H, J ) 8.0, 10.5 Hz), 3.50 (dd, 1H, J ) 3.7,
(21) Parkes, K. E. B.; Bushnell, D. J .; Crackett, P. H.; Dunsdon, S.
J .; Freeman, A. C.; Gunn, M. P.; Hopkins, R. A.; Lambert, R. W.;
Martin, J . A.; Merrett, J . H.; Redshaw, S.; Spurden, W. C.; Thomas,
G. J . J . Org. Chem. 1994, 59, 3656.

Synthesis of anti-N-Protected-R-Amino Epoxides
J . Org. Chem., Vol. 66, No. 17, 2001 5795
10.6 Hz), 3.63 (s, 3H), 3.78 (m, 1H), 3.83 (m, 1H), 5.08 (s, 2H),
5.19 (d, 1H, J ) 8.6 Hz), 7.34 (m, 5H); 13C NMR (CDCl3) δ
24.7, 30.5, 36.4, 51.8, 54.0, 67.2, 74.0, 128.2, 128.3, 128.6, 136.3,
156.9, 174.0. Anal. Calcd for C15H20BrNO5: C, 48.14; H, 5.39;
N, 3.74. Found: C, 48.31; H, 5.45; N, 3.86. Method C gave 9f
(94%) with an anti:syn ratio of 6.98:1.
(2 mL) as a white solid (29.3 mg, 97%) crude): 1H NMR
(CDCl3) δ 0.97 (d, 3H, J ) 6.9 Hz), 1.02 (d, 3H, J ) 6.9 Hz),
1.85 (m, 1H), 2.74 (m, 2H), 2.89 (m, 1H), 2.29 (m, 1H), 4.74 (d,
1H, J ) 8.1 Hz), 5.09 (s, 2H), 7.30-7.35 (m, 5H); 13C NMR
(CDCl3) δ 17.8, 19.2, 30.8, 45.6, 52.5, 58.0, 67.0,128.1, 128.2,
128.6, 136.5, 156.3.
Br om oh yd r in 9g was prepared from 8g (100 mg, 0.22
mmol) by Method B in 99% yield. Because separation of the
diastereomers was incomplete, the anti:syn ratio could only
be determined as >30:1. Recrystallization from toluene gave
Ep oxid e 10e was prepared from 9e (40 mg, 0.081 mmol),
K2CO3 (22.4 mg, 0.16 mmol), and methanol (2 mL) as a white
solid (32.7 mg, 98% crude): 1H NMR (CDCl3) δ 1.44 (m, 6H),
2.76 (m, 2H), 2.83 (m, 1H), 3.14 (m, 2H), 3.42 (m, 1H), 4.86
(broad, 1H), 5.07 (m, 5H), 7.32-7.33 (m, 10H); 13C NMR
(CDCl3) δ 22.4, 29.7, 31.2, 40.5, 46.2, 52.7, 54.0, 66.7, 67.0,
a pure sample (68%): mp 172 °C; [R]21 -16.8 (c 1.00, DMF);
D
FTIR (KBr) 3326, 1696 cm-1; H NMR (CDCl3 / few drops of
1
methanol-d4) δ 2.71 (dd, 1H, J ) 9.1, 13.9 Hz), 2.92 (dd, 1H, J
) 4.0, 13.9 Hz), 3.29 (dd, 1H, J ) 7.6, 10.3 Hz), 3.37 (dd, 1H,
J ) 4.0, 10.6 Hz), 3.72 (m, 1H, overlapping), 3.84 (m, 1H), 4.01
(t, 1H, J ) 6.6 Hz), 4.23 (m,2H), 7.09-7.69 (m, 13H); 13C NMR
(CDCl3 / few drops of methanol-d4) δ 35.3, 36.0, 47.1, 55.2, 66.3,
72.8, 119.8, 124.8, 126.4, 126.9, 127.6, 128.1, 128.3, 129.3,
137.6, 141.2, 143.7, 156.3. Anal. Calcd for C25H24BrNO3: C,
64.38; H, 5.19; N, 3.00. Found: C, 64.50; H, 5.36; N, 2.97.
Method C gave 9g with an anti:syn ratio of 3.3:1.
Ep oxid e 10b was prepared from 9b to demonstrate the
method. A mixture of bromohydrin 9b (60 mg, 0.2 mmol) and
K2CO3 (55 mg, 0.4 mmol) in methanol (2 mL) was stirred at
room temperature for 30 min. After evaporation of the solvent,
ethyl acetate (10 mL) was added. Extraction with water (5 mL)
and saturated NaCl (5 mL), drying (Na2SO4), and solvent
evaporation gave 10b as a colorless oil (42 mg, 95%): 1H NMR
(CDCl3) δ 1.17 (d, 3H, J ) 6.8 Hz), 2.25 (m, 2H), 2.93 (m, 1H),
3.73 (m, 1H), 4.89 (broad, 1H), 5.09 (s, 2H), 7.34-7.36 (m, 5H);
13C NMR (CDCl3) δ 16.5, 46.0, 48.1, 54.5, 66.9, 128.1, 128.2,
128.6, 136.5, 155.8. These data are in accord with the literature
data for 10b.18
128.1, 128.5, 136.5, 136.7, 156.2, 156.7. Anal. Calcd for C23H29
-
BrN2O5: C, 55.99; H, 5.92; N, 5.68. Found: C, 56.14; H, 6.14;
N, 5.76.
Ep oxid e 10f was prepared from 9f (50 mg, 0.13 mmol),
K2CO3 (37 mg, 0.26 mmol), and methanol (2 mL) as a colorless
oil (37 mg, 97% crude): 1H NMR (CDCl3) δ 1.85 (m, 1H), 1.96
(m, 1H), 2.43 (m, 2H), 2.75 (m, 2H), 2.89 (m, 1H), 3.52 (m,
1H), 3.64 (s, 3H), 5.04 (broad, 1H), 5.08 (s, 2H), 7.34 (m, 5H);
13C NMR (CDCl3) δ 26.7, 30.4, 46.0, 51.7, 52.6, 53.7, 77.0, 128.1,
128.2, 128.6, 136.4, 156.1, 173.7.
Ack n ow led gm en t. This work was supported by a
grant from the National Science Foundation (CHE-
9900402). The authors also thank J oanna Szurmak for
assistance with the HPLC analyses.
Su p p or tin g In for m a tion Ava ila ble: 13C NMR spectra of
10c and 10f. This material is available free of charge via the
Internet at http://pubs.acs.org.
Ep oxid e 10c was prepared by the same procedure from 9c
(40 mg, 0.12 mmol), K2CO3 (37 mg, 0.24 mmol), and methanol
J O010319H