9152 J . Org. Chem., Vol. 62, No. 26, 1997
Bongini et al.
(2′S,3′R)-5b: IR (Nujol) ν 3260, 1720, 1650 cm-1; 1H NMR
(CDCl3) δ 0.81 (d, 3H, J ) 6.6 Hz), 1.02 (t, 3H, J ) 7.4 Hz),
1.45-1.62 (m, 2H), 1.87 (br.s, 1H), 1.94 (dt, 1H, J ) 2.5 Hz,
5.9 Hz), 2.86 (s, 3H), 3.76 (d, 1H, J ) 2.5 Hz), 3.98 (dq, 1H, J
) 6.6 Hz, 8.4 Hz), 5.25 (d, 1H, J ) 8.4 Hz), 7.09-7.41 (m, 5H);
13C NMR (CDCl3) δ 10.9, 14.7, 25.8, 27.9, 35.8, 42.0, 54.2, 59.6,
126.5, 128.1, 128.5, 136.1, 155.5, 171.1; MS m/z 287 (M+, 25),
272 (15), 258 (3), 231 (16), 217 (4), 191 (100), 175 (18), 132
(30); [R]D ) -75 (c 1.2, CHCl3); mp ) 140-145 °C. Anal.
Calcd for C16H21N3O2: C, 66.9; H, 7.4; N, 14.6. Found: C, 66.8;
H, 7.4; N. 14.5.
In conclusion, chiral 3′-benzyloxyamino imides are
suitable precursors of 3′-alkylaziridine-2-carboxylates,
which are obtained via chiral titanium or aluminum
enolates. A mechanicistic study by AM1 calculations
suggests that the complete trans diastereoselectivity can
be attributed to the preferred conformation displayed by
the intermediate cyclic enolate, which shows the alkyl
group and the leaving group trans to each other. In the
synthesis of 3′-unsubstituted aziridines, diastereoselec-
tivity is lower, probably owing to the absence of the
substituent in position 3′.
(2′S,3′R)-5c: IR (Nujol) ν 3270, 1720, 1660 cm-1; H NMR
1
(CDCl3) δ 0.83 (d, 3H, J ) 6.6 Hz), 0.95 (t, 3H, J ) 7.3 Hz),
1.46-1.56 (m, 4H), 1.72 (br.s, 1H), 2.00 (m, 1H), 2.89 (s, 3H),
3.80 (d, 1H, J ) 2.7 Hz), 4.00 (dq, 1H, J ) 6.6 Hz, 8.3 Hz),
5.26 (d, 1H, J ) 8.3 Hz), 7.10-7.41 (m, 5H); 13C NMR (CDCl3)
δ 13.7, 14.8, 20.2, 28.0, 34.8, 35.8, 40.6, 54.2, 59.7, 126.5, 127.1,
128.4, 128.5, 136.2, 155.5, 171.1; MS m/z 301 (M+, 10), 286
(6), 258 (3), 231 (15), 217 (6), 191 (100), 175 (10), 132 (24), 113
(26), 84 (27), 77 (13); [R]D ) -54 (c 1.4, CHCl3); mp ) 112-
116 °C. Anal. Calcd for C17H23N3O2: C, 67.8; H, 7.7; N, 13.9.
Found: C, 67.9; H, 7.7; N. 14.0.
Exp er im en ta l Section
Gen er a l Meth od s. 1H and 13C NMR spectra were recorded
at 300 and 75 MHz, respectively, and chemical shifts were
reported in ppm relative to the solvent peak of CHCl3. IR
spectra were recorded using a FT-IR spectrometer. Melting
points are uncorrected and are determined in open capillaries.
Flash chromatography was performed using Merck silica gel
60 (230-400 mesh). Solvents for flash chromatograpy were
simply distilled. THF was distilled from sodium benzophenone
ketyl. CH2Cl2 was distilled from P2O5. DMF was distilled
from activated molecular sieves. O-Benzylhydroxylamine was
used as a 0.5 M solution in CH2Cl2. This solution was obtained
by treatment of O-benzylhydroxylamine hydrochloride (2.0 g,
12.5 mmol) with NaOH (0.75 g, 18.8 mmol) in water (5 mL)
followed by extraction with CH2Cl2 (4 × 10 mL). The organic
layers were dried over Na2SO4 and partially evaporated at
reduced pressure. Finally the solution volume was adjusted
to 25 mL with dry CH2Cl2. AlMe2Cl was used as a 1 M solution
in cyclohexane.
Molecular mechanics MM+ and semiempirical AM1 calcula-
tions were achieved on the HYPERCHEM calculation pack-
age.24 MM+ was set on a dipole-dipole electrostatic force field.
Geometry optimizations were performed using a Polak-Ribiere
conjugate gradient algorithm until a gradient of 0.01 kcal/mol
was reached. The Monte Carlo method is available on the
CHEMPLUS24 package. The preferential geometry of chelate
and nonchelate Z enolate for (4S)-4-(methoxyamino)pentan-
2-one was fully optimized with semiempirical AM1 calcula-
tions. The AM1 optimization was performed over a collection
of stable conformations obtained with the Monte Carlo method
and minimized with MM+. The Monte Carlo method provided
100 random conformations by varying C3-C4, C4-N, and
N-O torsion angles. The result of HYPERCHEM/AM1 full
optimization was confirmed by MOPAC 6.028 AM1 calculations.
The preferential conformation of the aluminum complex 16
was fully optimized by means of AM1 calculations performed
on stable conformations obtained with a Monte Carlo method
and minimized with MM+. 300 Random conformations were
considered varying O-Ti, C4-Ph, N3-C1′, C1′-C2′, C2′-C3′,
C3′-N′, N′-O′, O′-CH2Ph, and CH2-Ph torsion angles.
(4R,5S)-1,5-Dim eth yl-3-a cr yloyl-4-p h en ylim id a zolid in -
2-on e (6). To a stirred solution of (4R,5S)-1,5-dimethyl-4-
phenylimidazolidin-2-one (0.7 g, 3.7 mmol) in anhydrous
CH2Cl2 (3 mL), were added diisopropylethylamine (0.95 mL,
5.5 mmol), acryloyl chloride (0.45 mL, 5.5 mmol) and catalytic
CuCl under an inert athmosphere at rt. The reaction was
refluxed for 2 h and then was quenched with 1 M NH4OH and
extracted with CH2Cl2 (3 × 25 mL). The organic layers were
dried over Na2SO4, and the solvent was evaporated at reduced
pressure. The residue was purified by crystallization from
EtOAc/cyclohexane affording 6 (0.87 g, 97%) as a solid. IR
(Nujol) 1710, 1670, 1620 cm-1; 1H NMR (CDCl3) δ 0.83 (d, 3H,
J ) 6.6 Hz), 2.86 (s, 3H), 3.94 (dq, 1H, J ) 6.6 Hz, 8.5 Hz),
5.37 (d, 1H, J ) 8.5 Hz), 5.77 (dd, 1H, J ) 10.4 Hz, 2.0 Hz),
6.40 (dd, 1H, J ) 17.0 Hz, 2.0 Hz), 7.40-7.15 (m, 5H), 7.72
(dd, 1H, J ) 10.4 Hz, 17.0 Hz); 13C NMR (CDCl3) δ 14.8, 28.1,
53.8, 59.3, 126.8, 127.9, 128.4, 128.6, 128.7, 128.8, 129.7, 136.3,
155.5, 164.4; MS m/z 244 (M+, 24), 229 (3), 189 (44), 175 (13),
143 (5), 132 (100), 105 (10), 91 (8), 77 (27), 58 (13), 55 (30);
[R]D ) -100.6° (c 1.0, CHCl3); mp 135-140 °C. Anal. Calcd
for C14H16N2O2: C, 68.8; H, 6.6; N, 11.5. Found: C, 69.0; H,
6.6; N. 11.5.
N-(ter t-Bu toxyca r bon yl)-O-ben zylh yd r oxyla m in e (9).
To a mechanically stirred solution of O-benzylhydroxylamine
hydrochloride (0.15 g, 0.94 mmol) in water (2 mL) and CH2-
Cl2 (2 mL) was added NaHCO3 (0.174 g, 2 mmol) in small
portions at 0 °C. When foaming ceased, di-tert-butyl dicar-
bonate (0.25 g, 1.1 mmol) was added at 0 °C and strirring was
continued for 3 h. The reaction was stopped by addition of 3
mL of a saturated solution of NaHCO3 and extracted three
times with CH2Cl2. The organic layers were dried over Na2-
SO4. After evaporation of the solvent at reduced pressure, 9
(0.21 g, 99%) was obtained without further purification. 1H
NMR (CDCl3) δ 1.48 (s, 9H), 4.86 (s, 2H), 7.11-7.48 (m, 6H).
(4R,5S)-1,5-Dim eth yl-3-[(3′-alkyl-2′-azir idin yl)car bon yl]-
4-p h en ylim id a zolid in -2-on e (4, 5). To a stirred solution of
2 or 3 (1 mmol) in dry CH2Cl2 (5 mL) under an inert
atmosphere either AlMe2Cl (1 M in cyclohexane, 1.1 mL, 1.1
mmol) or TiCl4 (0.12 mL, 1.1 mmol) was added at rt. After 15
min, the solution was transferred dropwise over 1 min at rt
and under an inert atmosphere by means of a Teflon cannula
to a solution of triethylamine (0.28 mL, 2 mmol) in CH2Cl2 (5
mL). The reaction was quenched with water after 30 min (5
mL) and extracted with CH2Cl2 (3 × 10 mL). The organic
layers were dried over Na2SO4, and the solvent was evaporated
at reduced pressure. The residue was purified by flash
chromatography on silica gel (for 4 cyclohexane/ethyl acetate
1/1, for 5 cyclohexane/EtOAc 1/2) yielding 4 or 5 as a solid
(70-97%).
(4R,5S)-1,5-Dim eth yl-3-[3′-[(ter t-bu toxyca r bon yl)(ben -
zyloxy)am in o]pr opan oyl]-4-ph en ylim idazolidin -2-on e (10).
To a suspension of NaH (0.002 g, 0.08 mmol) in dry DMF (1.5
m) was added 9 (0.055 g, 0.25 mmol) under an inert athmo-
sphere at rt. After 15 min, 6 (0.050 g, 0.2 mmol) was added,
and stirring was continued for 3 h. The reaction was diluted
with 30 mL of CH2Cl2 and washed four times with small
portions of water. The organic solvent was dried over Na2-
SO4 and evaporated at reduced pressure. The residue was
purified by flash chromatography on silica gel with cyclohex-
ane/EtOAc 70/30 affording 10 (0.090 g, 96%) as an oil. IR
1
(Nujol) 1730, 1680 cm-1; H NMR (CDCl3) δ 0.76 (d, 3H, J )
6.6 Hz), 1.49 (s, 9H), 2.80 (s, 3H), 3.23 (dt, 1H, J ) 7.1 Hz,
16.9 Hz), 3.43 (dt, 1H, J ) 6.9 Hz, 16.9 Hz), 3.70-3.87 (m,
3H), 4.80 (s, 2H), 5.18 (d, 1H, J ) 8.5 Hz), 7.09-7.44 (m, 10H);
13C NMR (CDCl3) δ 14.8, 28.2, 33.5, 45.2, 53.8, 59.2, 81.2, 126.8,
127.9, 128.2, 128.3, 129.3, 135.5, 136.5, 155.6, 156.3, 170.5;
MS m/z 244 (19), 189 (34), 175 (13), 149 (9), 132 (100), 117 (6),
References 9b and 11 include characterization data for
(2′R,3′S)-4, and (2′S,3′R)-5a , respectively.
(28) Stewart, J . J . P. QCPE455 Indiana University, Bloomington,
IN.