Novel synthesis of 4,5-unsubstituted 2,3-dihydroisoxazoles from 5-acetoxyisoxazolidines

Article abstract of DOI:10.1055/s-0032-1317682

A new synthetic method for the preparation of 4,5-unsubstituted 2,3-dihydroisoxazoles from readily available 5-acetoxyisoxazolidines is presented. Elimination reactions are carried out in anhydrous N-methylpyrrolidin-2-one (NMP) with a catalytic amount of

Full text of DOI:10.1055/s-0032-1317682

PAPER
▌3783
paper
Novel Synthesis of 4,5-Unsubstituted 2,3-Dihydroisoxazoles from
5-Acetoxyisoxazolidines
Novel Synthesis of 4,5-Unsubstituted 2,3-Dihydroisoxazoles
Róbert Fischer,* Daniela Lackovičová, Lubor Fišera
Institute of Organic Chemistry, Catalysis and Petrochemistry, Slovak University of Technology, 812 37 Bratislava, Slovak Republic
E-mail: robert.fischer@stuba.sk
Received: 18.09.2012; Accepted after revision: 31.10.2012
Bn
O
Abstract: A new synthetic method for the preparation of 4,5-un-
substituted 2,3-dihydroisoxazoles from readily available 5-acetoxy-
isoxazolidines is presented. Elimination reactions are carried out in
anhydrous N-methylpyrrolidin-2-one (NMP) with a catalytic
amount of trimethylsilyl triflate in the presence of N,O-bis(trimeth-
ylsilyl)trifluoroacetamide (BSTFA) and provide the 2,3-dihy-
droisoxazoles in very good yields. The nature of the silylating agent
plays a very important role in elimination process. Anhydrous reac-
tion conditions are required, while trimethylsilanol, the product of
trimethylsilyl triflate and N,O-bis(trimethylsilyl)trifluoroacetamide
hydrolysis, can initiate reactions leading to deacetylation, giving
side products and thus decreasing the total yield of the elimination.
Bn
N
N
O
H
H
i-Pr
1
2
TBDPSO
BzO
O
Bn
N
O
3
O
Scheme 1
Key words: elimination, Lewis acids, cycloaddition, nitrones, 2,3-
dihydroisoxazoles
3,5-cis and 3,5-trans isomers and the chiral isoxazolidines
68 and 79 as a mixture of four isomers 3,5-cis/3,5-trans to-
gether with the corresponding 1′,3-syn/1′,3-anti isomers.
In the case of isoxazolidine 89 we only isolated and used
the major isomer 1′,3-anti-3,5-trans.
4,5-Substituted 2,3-dihydroisoxazoles (4-isoxazolines)
are interesting heterocycles used in various synthetic
strategies as suitable building blocks.1 On the other hand,
4,5-unsubstituted 2,3-dihydroisoxazoles 1 (Scheme 1)
have mostly been observed as byproducts from 1,3-dipo-
lar cycloadditions of nitrones with silyl enol ethers2 and p-
tolyl vinyl sulfoxides3 or in the Swern oxidation of 5-hy-
droxyisoxazolidines.4 Their chemical utilization is not
well-documented except for the transformation to aziri-
dines 2 (Scheme 1).5 Recently, the N-benzyl-2,3-dihy-
droisoxazole 3 (Scheme 1) was isolated as an undesired
byproduct in low yield in our laboratories during the
Vorbrüggen nucleosidation of 5-acetoxyisoxazolidines,
together with the corresponding isoxazolidinyl nucleo-
sides.6 To the best of our knowledge, this was the first
time that a 4,5-unsubstituted 2,3-dihydroisoxazole was
prepared directly from a 5-acetoxyisoxazolidine without
isolating an intermediate. Based on these observations we
focused our attention on the development of a new syn-
thetic method for the synthesis of 4,5-unsubstituted 2,3-
dihydroisoxazoles.
All elimination reactions were carried out in anhydrous
NMP with a catalytic amount of trimethylsilyl triflate in
the presence of N,O-bis(trimethylsilyl)trifluoroacetamide
(BSTFA) and provided 2,3-dihydroisoxazoles 3, 9–12 in
very good yields (Table 1).
The progress of the reactions was easily followed by TLC
(hexanes/ethyl acetate) and 2,3-dihydroisoxazoles 3, 9–12
were identified as readily oxidized nonpolar components
by permanganate detection at room temperature. After
isolating the products by flash column chromatography on
silica gel the structures were confirmed by 1H NMR spec-
troscopic analysis. The existence of the C4–C5 double
bond in 3, 9–12 is confirmed by characteristic chemical
shifts and multiplicity of the protons of H3, H4, and H5.
In all cases we found three narrow apparent triplets with
coupling constants in the range of 2–3 Hz.
The use of trimethylsilyl triflate/2,6-lutidine in the elimi-
nation reactions of methyl furanosides has been already
published by the Leumann group,10 but without mechanis-
tic rationale. A mechanism for the elimination reaction is
proposed according to the well-known Vorbrüggen
nucleosidation11 (Scheme 2). The oxocarbenium ion 13
created by trimethylsilyl triflate undergoes elimination to
the 4,5-unsubstituted 2,3-dihydroisoxazoles 3, 9–12 in the
absence of nucleophiles. Trifluoromethanesulfonic acid is
then rapidly silylated by N,O-bis(trimethylsilyl)trifluoro-
acetamide producing trimethylsilyl triflate and monosi-
lylated trifluoroacetamide 14, which is totally unreactive
as a nucleophile because of the strong electron-withdraw-
In accordance with our goal to study the transformation of
5-acetoxyisoxazolidines into 2,3-dihydroisoxazoles, we
prepared some representative isoxazolidines 4–8 (Table
1) via regioselective 1,3-dipolar cycloaddition of nitrones
with vinyl acetate as a mixture of two or four isomers de-
pending upon the nitrone. The isoxazolidines 47 and 54
were isolated and subsequently used as a mixture of both
SYNTHESIS 2012, 44, 3783–3788
Advanced online publication: 27.11.2012
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X
DOI: 10.1055/s-0032-1317682; Art ID: SS-2012-T0731-OP
© Georg Thieme Verlag Stuttgart · New York

3784
R. Fischer et al.
PAPER
Other Lewis acids such as boron trifluoride–diethyl ether
complex and tin(IV) chloride, which are commonly used
in the Vorbrüggen nucleosidation and are also able to
form the oxocarbenium ion did not cause any elimination.
Solvents used in the Vorbrüggen nucleosidations namely
acetonitrile or 1,2-dichloroethane as well as tetrahydrofu-
ran, N,N-dimethylformamide, and N-methylpyrrolidin-2-
one were also tested. While acetonitrile and 1,2-dichloro-
ethane were found to be unsuitable (no or low conversion
of the initial 5-acetoxyisoxazolidine, greater level of im-
purities in the reaction mixture), in the case of N,N-di-
methylformamide and N-methylpyrrolidin-2-one we
observed significant formation of 2,3-dihydroisoxazoles
3, 9–12. Moreover, the reaction in N-methylpyrrolidin-2-
one was cleaner than in N,N-dimethylformamide. The use
of less concentrated solutions led to higher yields with the
same reaction time (Table 1, entries 1–4).
Table 1 Elimination Reactions of 5-Acetoxyisoxazolidines 4–8 to
2,3-Dihydroisoxazoles 9–12, 3
TMSOTf (0.1 equiv)
BSTFA (1.2 equiv)
O
O
Bn
Bn
OAc
N
N
anhyd NMP
0 °C to r.t.
3 h
R
R
4–8
9–12,3
Entry Isoxazo-
lidine
R
2,3-Dihydro- Total yield
isoxazole
(%)
1
2
3
4
4
4
5
5
CO2Et
9
55a
76b
53a
75b
9
Ph
10
10
O
5
6
6
7
O
11a,bc
12a,bc
70a (29:71)d
73a (15:85)d
1'
During our search for the best reaction conditions, we
found that the nature of the silylating agent plays a very
important role in the process of the elimination. N,O-
Bis(trimethylsilyl)acetamide (BSA), which was originally
used for silylation of nucleoside bases, is also able to react
with the oxocarbenium ion.12 We examined N,O-bis(tri-
methylsilyl)acetamide in N-methylpyrrolidin-2-one, and
we observed the incorporation of the acetylamino group
into the C5 position of the isoxazolidinyl ring along with
2,3-dihydroisoxazole formation. To demonstrate the pos-
sible negative effect of N,O-bis(trimethylsilyl)acetamide
on the reaction course, we carried out the reaction of the
isoxazolidine 5 with monosilylated acetamide 15 in 1,2-
dichloroethane as a nonbasic solvent and we identified
acetylamino derivative 16 in the reaction mixture
(Scheme 3). Compound 16 was not observed when the re-
action was carried out with N,O-bis(trimethylsilyl)trifluo-
roacetamide alone, but in the presence of monosilylated
acetamide 15 subsequently a more nucleophilic N,O-
bis(trimethylsilyl)acetamide was produced in situ, which
is a crucial reagent for the incorporation of the acetylami-
no group. Analysis of the original reaction mixture by
NMR spectroscopy indicated the incorporation of the
acetylamino group into the C5 position of the isoxazolidi-
2'
O
1'
2'
O
3'
O
4'
O
BzO
1'
2'
TBDPSO
3'
4'
O
7
8
3
73a
O
a 1 M concentration.
b 0.2 M concentration.
c Mixture of two isomers a (1′,3-syn) and b (1′,3-anti).
d Ratios were determined by 1H NMR analysis of the crude reaction
mixture.
ing group effect of the trifluoromethyl group. Thus, only
a catalytic amount of trimethylsilyl triflate (0.1 equiv) and
a slight excess of N,O-bis(trimethylsilyl)trifluoroacet-
amide (1.2 equiv) is required, which is in agreement with
experimental observations.
OTf
O
O
O
O
O
OTf
TMSOTf
Bn
Bn
Bn
O
N
N
N
+
OTMS
O
O
R
R
R
4–8
13
TMS
NTMS
OTMS
F3C
O
Bn
NHTMS
O
(BSTFA)
N
TMSOTf
TfOH
+
+
F3C
R
14
9–12,3
Scheme 2
Synthesis 2012, 44, 3783–3788
© Georg Thieme Verlag Stuttgart · New York

PAPER
Novel Synthesis of 4,5-Unsubstituted 2,3-Dihydroisoxazoles
3785
nyl ring.13 2,3-Dihydroisoxazole 10 was not observed ow-
ing to the absence of basic N-methylpyrrolidin-2-one
which induces the elimination. Moreover, another silylat-
ing agent, 3-(trimethylsilyl)oxazolidin-2-one (17), which
could reform trimethylsilyl triflate from triflic acid,14 in
its reaction with isoxazolidines 4 and 5 produced the cor-
responding isoxazolidinyl oxazolidinones 18 and 19
(Scheme 3) as a mixture of 3,5-cis and 3,5-trans isomers,
which have been previously prepared by nitrone 1,3-dipo-
lar cycloaddition with N-vinylated oxazolidin-2-one.15
All of these obtained results indicate N,O-bis(trimethylsi-
lyl)trifluoroacetamide to be the best silylating agent for
our new method.
O
O
O
Bn
OAc
Bn
OH
Bn
OTMS
a
b
N
N
N
EtO2C
EtO2C
EtO2C
21a,b
4
20
Scheme 4 Reagents and conditions: (a) TMSOH (2.5 equiv),
TMSOTf (0.2 equiv), MeCN, 40 °C, 6 h, 80%, 2 isomers with ratio
60:40; (b) BSTFA (4 equiv), DCE, r.t., 2 h, 69%, ratios were deter-
mined by 1H NMR analysis of the crude reaction mixture.
reaction mixture when N-methylpyrrolidin-2-one was not
stored over molecular sieves and, thus, probably con-
tained traces of water.
In conclusion, a new synthetic method for the preparation
of 4,5-unsubstituted 2,3-dihydroisoxazoles 3, 9–12 from
readily available 5-acetoxyisoxazolidines 4–8 is present-
ed. Elimination reactions are carried out in anhydrous N-
methylpyrrolidin-2-one with a catalytic amount of tri-
methylsilyl triflate in the presence of N,O-bis(trimethyl-
silyl)trifluoroacetamide and provide the 2,3-dihydro-
isoxazoles 3, 9–12 in very good yields (70–76%). The na-
ture of the silylating agent plays a very important role in
the process of the elimination. Anhydrous reaction condi-
tions are required, while trimethylsilanol, the product of
trimethylsilyl triflate and N,O-bis(trimethylsilyl)trifluoro-
acetamide hydrolysis, can initiate reactions leading to
deacetylation, giving side products and so decreasing the
total yield of the elimination.
O
O
TMS
O
Bn
OAc
N
O
NHTMS
N
17
15
R = Ph
R
4 R = CO2Et
5 R = Ph
a
b
O
O
H
O
O
Bn
N
Bn
N
N
N
O
Ph
R
16
18 R = CO2Et
19 R = Ph
Scheme 3 Reagents and conditions: (a) TMSOTf (0.1 equiv),
BSTFA (1.5 equiv), DCE (1 M), r.t., 24 h; (b) TMSOTf (0.1 equiv),
DCE (1 M), r.t., 24 h, 18 75%, ratio 3,5-cis/3,5-trans, 25:75, 19 82%,
1
ratio 3,5-cis/3,5-trans, 33:67, ratios were determined by H NMR
analysis of the crude reaction mixture.
All melting points were measured on a Melting Point B-540 appa-
ratus (Büchi) and are uncorrected. Elemental analyses were carried
out on the analyzer Vario Macro Cube (Elementar). Optical rota-
tions were measured on a Jasco P-2000 polarimeter (concentration
c is given as g/100 mL). NMR spectra were recorded on a Varian
VRX-300 spectrometer (1H, 300 MHz and 13C, 75.4 MHz) and
Varian Inova-600 spectrometer (1H, 600 MHz and 13C, 150.8 MHz)
in CDCl3 using TMS as the internal standard. TLC analysis was car-
ried out using TLC silica gel 60 F254 (Al sheets, Merck) and visual-
ized by UV light or with permanganate soln followed by heating.
Column chromatography was performed on Büchi system (Pump
Manager C-615 and Fraction Collector C-660) using Normasil 60
silica gel (0.040–0.063 mm) (VWR). All solvents were dried and
distilled according to conventional methods. NMP, DCE, and
MeCN were stored over molecular sieves and handled under an in-
ert atmosphere. All reagents were purchased from Aldrich, Acros
Organics, Alfa-Aesar, Merck, and Mikrochem and were used with-
out further purification.
Next we focused our attention on the preparative perfor-
mance under anhydrous conditions since the both key
reagents trimethylsilyl triflate and N,O-bis(trimethylsi-
lyl)trifluoroacetamide are highly water-sensitive and can
rapidly hydrolyze to form trimethylsilanol. On the basis of
our observations, we suppose that the presence of trimeth-
ylsilanol along with trimethylsilyl triflate can initiate the
reactions leading to deacetylation and producing 5-hy-
droxyisoxazolidines. The hydroxy group is subsequently
silylated by excess N,O-bis(trimethylsilyl)trifluoroacet-
amide. The 5-(silyloxy)isoxazolidines are unreactive by-
products and decrease the total yield of the elimination.
This was confirmed by reaction of 5-acetoxyisoxazolidine
4 (ratio 3,5-cis/3,5-trans, 20:80) with trimethylsilanol
(2.5 equiv) and trimethylsilyl triflate (0.2 equiv) followed
by the silylation of 5-hydroxyisoxazolidine 20 with N,O-
bis(trimethylsilyl)trifluoroacetamide (Scheme 4). The fi-
nal 5-(trimethylsilyloxy)isoxazolidine 21 was isolated in
55% yield over two steps as a mixture of 3,5-cis 21a and
3,5-trans 21b isomers with a ratio of 21a/21b 30:70.
2,3-Dihydroisoxazoles; General Procedure
The reaction flask was charged with 5-acetoxyisoxazolidine, sealed
with a rubber septum and filled with argon. Anhyd NMP was added
(0.2 M or 1 M) followed by BSTFA (1.2 equiv). The soln was
cooled in an ice/water bath (2–5 °C) and TMSOTf (0.1 equiv) was
added. The mixture was subsequently stirred at r.t. for 3 h in the
dark. When the initial 5-acetoxyisoxazolidine had been consumed
(TLC), the mixture was cooled in an ice/water bath, sat. aq NaHCO3
and Et2O were added, and the mixture was stirred for 5 min. The
aqueous layer was removed, the organic layer was washed with H2O
(3 ×) and dried (Na2SO4), and the solvent was evaporated in vacuo
(bath temperature < 35 °C). 2,3-Dihydroisoxazole was isolated by
column chromatography (silica gel, hexanes–EtOAc, 90:10).
The isomers 21a and 21b were separated by the column
chromatography on silica gel and their cis/trans relative
configuration was assigned based on NOE experiments.
During the separation process, the unstable trimethylsil-
oxy group was partially hydrolyzed. The undesired isoxa-
zolidines 20 and 21a,b were also detected in the crude
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2012, 44, 3783–3788

3786
R. Fischer et al.
PAPER
2-Benzyl-3-phenyl-2,3-dihydroisoxazole (10); Typical Proce-
dure
4.19 (d, J = 12.9 Hz, 1 H, NCH2Ph), 4.78 (dd, J = 2.3, 2.9 Hz, 1 H,
H4), 6.54–6.56 (m, 1 H, H5), 7.26–7.39 (m, 5 H, Ph).
The reaction flask was charged with 5-acetoxyisoxazolidine 5 (1.5
g, 5.0 mmol), sealed with a rubber septum, and filled with argon.
Anhyd NMP (25 mL) was added followed by BSTFA (1.59 mL, 6.0
mmol). The soln was cooled in an ice/water bath (2–5 °C) and
TMSOTf (0.09 mL, 0.5 mmol) was added. The mixture was subse-
quently stirred at r.t. for 3 h in the dark. When the initial 5-
acetoxyisoxazolidine had been consumed (TLC, hexanes–EtOAc,
83:17), the mixture was cooled in an ice/water bath, sat. aq NaHCO3
(30 mL) and Et2O (50 mL) were added, and the mixture was stirred
for 5 min. The precipitated solid was dissolved by addition of H2O
(50 mL). The aqueous layer was removed, the organic layer was
washed with H2O (3 × 30 mL) and dried (Na2SO4), and the solvent
was evaporated in vacuo (bath temperature < 35 °C). 2,3-Dihy-
droisoxazole 10 was isolated by column chromatography (silica gel,
hexanes–EtOAc, 90:10) as a colorless oil that became darker over
time; yield: 0.89 g (3.75 mmol, 75%).
13C NMR (150.8 MHz, CDCl3): δ = 25.1, 26.3, 63.3, 65.5, 70.7,
77.5, 94.9, 109.3, 127.6, 128.4, 129.5, 135.9, 143.9.
Anal. Calcd for C15H19NO3: C, 68.94; H, 7.33; N, 5.36. Found: C,
69.08; H, 7.10; N, 5.09.
(3S)-2-Benzyl-3-[(4S)-2,2-dimethyl-1,3-dioxolan-4-yl)]-2,3-di-
hydroisoxazole (11b)
Colorless solid; yield: 385 mg (52%); mp 42–43 °C; [α]D20 –265.73
(c 1.02, CHCl3).
1H NMR (600 MHz, CDCl3): δ = 1.31, 1.38 [2 s, 2 × 3 H, C(CH3)2],
3.50–3.53 (m, 1 H, H2a′), 3.82 (d, J = 12.9 Hz, 1 H, NCH2Ph), 3.89–
3.91 (m, 1 H, H3), 3.92–3.97 (m, 2 H, H1′, H2b′), 4.15 (d, J = 12.9
Hz, 1 H, NCH2Ph), 5.00 (dd, J = 2.3, 2.9 Hz, 1 H, H4), 6.51–6.53
(m, 1 H, H5), 7.25–7.38 (m, 5 H, Ph).
13C NMR (150.8 MHz, CDCl3): δ = 25.3, 26.7, 63.3, 66.8, 70.6,
78.6, 97.1, 109.3, 127.8, 128.4, 129.5, 136.1, 143.2.
1H NMR (600 MHz, CDCl3): δ = 4.02 (d, J = 12.9 Hz, 1 H,
NCH2Ph), 4.30 (d, J = 12.9 Hz, 1 H, NCH2Ph), 4.90–4.92 (m, 1 H,
H3), 5.01 (dd, J = 2.3, 2.9 Hz, 1 H, H4), 6.56–6.58 (m, 1 H, H5),
7.20–7.40 (m, 10 H, Ph).
Anal. Calcd for C15H19NO3: C, 68.94; H, 7.33; N, 5.36. Found: C,
68.76; H 7.06; N 5.12.
13C NMR (150.8 MHz, CDCl3): δ = 63.3, 72.1, 100.1, 127.0, 127.5
(2 C), 128.3, 128.5, 129.3, 136.5 (2 C), 142.0.
(3S)-2-Benzyl-3-(1,2:3,4-di-O-isopropylidene-D-xylo-tetritol-1-
yl)-2,3-dihydroisoxazole (12b)
Colorless viscous oil, becomes darker over time; yield: 215 mg
20
Anal. Calcd for C16H15NO: C, 80.98; H, 6.37; N, 5.90. Found: C,
80.91; H, 6.14; N, 6.19.
(50%); [α]D –166.66 (c 1.03, CHCl3). Additionally, a mixture of
unseparated isomers 12a and 12b (100 mg, 23%) was also isolated.
1H NMR (600 MHz, CDCl3): δ = 1.35, 1.37, 1.39, 1.40 [4 s, 4 × 3
H, C(CH3)2], 3.63 (dd, J = 4.1, 7.0 Hz, 1 H, H2′), 3.85 (d, J = 12.9
Hz, 1 H, NCH2Ph), 3.85 (dd, J = 7.0, 6.5 Hz, 1 H, H1′), 3.88–3.94
(m, 2 H, H4′), 4.04 (td, J = 1.8, 2.3 Hz, 1 H, H3), 4.12 (dt, J = 7.0,
4.1 Hz, 1 H, H3′), 4.16 (d, J = 12.9 Hz, 1 H, NCH2Ph), 5.00 (dd, J =
2.3, 2.9 Hz, 1 H, H4), 6.52–6.54 (m, 1 H, H5), 7.28–7.38 (m, 5 H,
Ph).
13C NMR (150.8 MHz, CDCl3): δ = 25.8, 26.3, 27.1, 27.5, 63.3,
66.0, 70.5, 76.3, 78.6, 79.6, 96.7, 109.5, 109.6, 127.8, 128.5, 129.6,
135.9, 143.4.
(3S)-3-(3-O-Benzoyl-4-O-tert-butyldiphenylsilyl-1,2-O-isopro-
pylidene-D-xylo-tetritol-1-yl)-2-benzyl-2,3-dihydroisoxazole (3)
Colorless oil, becomes darker over time; yield: 200 mg (73%);
[α]D20 –84.64 (c 0.86, CHCl3).
1H NMR (600 MHz, CDCl3): δ = 0.96 [s, 9 H, OSiC(CH3)3], 1.39,
1.41 [2 s, 2 × 3 H, C(CH3)2], 3.84 (dd, J = 4.1, 7.6, 1 H, H1′), 3.87
(dd, J = 5.3, 10.6 Hz, 1 H, H4a′), 3.92 (d, J = 13.5 Hz, 1 H,
NCH2Ph), 3.95 (dd, J = 8.2, 10.6 Hz, 1 H, H4b′), 4.15 (d, J = 13.5
Hz, 1 H, NCH2Ph), 4.18–4.20 (m, 1 H, H3), 4.29 (dd, J = 1.8, 7.6
Hz, 1 H, H2′), 4.94 (dd, J = 2.3, 2.9 Hz, 1 H, H4), 5.49 (ddd, J = 5.3,
8.2, 1.8 Hz, 1 H, H3′), 6.45–6.47 (m, 1 H, H5), 7.23–7.41 (m, 11 H,
Ph), 7.44 (t, 2 H, Ph), 7.56 (t, 1 H, Ph), 7.62 (d, 2 H, Ph), 7.67 (d, 2
H, Ph), 8.09 (d, 2 H, Ph).
Anal. Calcd for C20H27NO5: C, 66.46; H, 7.53; N, 3.88. Found: C,
66.23; H, 7.48; N, 3.63.
13C NMR (150.8 MHz, CDCl3): δ = 19.1, 26.6, 27.1, 27.3, 63.4,
64.0, 70.0, 73.2, 76.1, 78.1, 95.6, 109.4, 127.5, 127.6 (2 C), 128.3
(2 C), 129.1, 129.6 (2 C), 129.8, 130.2, 133.0, 133.3, 133.4, 135.6
(2 C), 136.6, 143.5, 166.0.
Ethyl 2-Benzyl-5-(2-oxooxazolidin-3-yl)isoxazolidine-3-carbox-
ylate (18)
The reaction flask was charged with the 5-acetoxyisoxazolidine 4
(150 mg, 0.51 mmol), sealed with a rubber septum and filled with
argon. Anhyd DCE was added (0.50 mL) followed by N-(trimethyl-
silyl)oxazolidin-2-one (17, 0.12 mL, 0.76 mmol). The soln was
cooled in an ice/water bath (2–5 °C) and TMSOTf (10 μL, 0.05
mmol) was added. The mixture was subsequently stirred at r.t. for
24 h. When the initial 5-acetoxyisoxazolidine had been consumed
(TLC, hexanes–EtOAc, 40:60), the mixture was cooled in an
ice/water bath, sat. aq NaHCO3 (5 mL) and CH2Cl2 (5 mL) were
added, and the mixture was stirred for 5 min. The aqueous layer was
removed and extracted with CH2Cl2 (5 mL). The combined organic
layers were washed with H2O and dried (Na2SO4), and the solvent
was evaporated in vacuo. The products were separated by column
chromatography (silica gel, hexanes–EtOAc, 40:60).
Anal. Calcd for C40H45NO6Si: C, 72.37; H, 6.83; N, 2.11. Found: C,
72.65; H, 7.09; N, 2.32.
Ethyl 2-Benzyl-2,3-dihydroisoxazole-3-carboxylate (9)
Colorless oil, becomes darker over time; yield: 610 mg (76%).
1H NMR (600 MHz, CDCl3): δ = 1.23 (t, J = 7.3 Hz, 3 H,
OCH2CH3), 3.97 (d, J = 13.0 Hz, 1 H, NCH2Ph), 4.16 (q, J = 7.3 Hz,
2 H, OCH2CH3), 4.29 (d, J = 13.0 Hz, 1 H, NCH2Ph), 4.48 (dd, J =
2.4, 2.8 Hz, 1 H, H3), 4.94 (dd, J = 2.4, 2.8 Hz, 1 H, H4), 6.55–6.57
(m, 1 H, H5), 7.28–7.41 (m, 5 H, Ph).
13C NMR (150.8 MHz, CDCl3): δ = 14.1, 61.3, 63.4, 70.1, 94.8,
127.8, 128.5, 129.3, 135.4, 143.8, 170.4.
3,5-cis-18a
Anal. Calcd for C13H15NO3: C, 66.94; H, 6.48; N, 6.00. Found: C,
66.99; H, 6.21; N, 6.27.
Colorless oil; yield: 33 mg (0.1 mmol, 20%).
1H NMR (300 MHz, CDCl3): δ = 1.25 (t, J = 7.0 Hz, 3 H,
OCH2CH3), 2.57 (ddd, J = 4.1, 7.6, 13.8 Hz, 1 H, H4a), 2.87 (ddd,
J = 8.2, 9.1, 13.8 Hz, 1 H, H4b), 3.55 (dd, J = 7.6, 9.1 Hz, 1 H, H3),
3.69–3.90 (m, 2 H, NCH2CH2O), 4.01 (d, J = 13.5 Hz, 1 H,
NCH2Ph), 4.09–4.16 (m, 2 H, OCH2CH3), 4.13 (d, J = 13.5 Hz, 1 H,
NCH2Ph), 4.22–4.35 (m, 2 H, NCH2CH2O), 5.92 (dd, J = 4.1, 8.2
Hz, 1 H, H5), 7.27–7.36 (m, 5 H, Ph).
(3R)-2-Benzyl-3-[(4S)-2,2-dimethyl-1,3-dioxolan-4-yl)]-2,3-di-
hydroisoxazole (11a)
Colorless solid; yield: 130 mg (18%); mp 47–48 °C; [α]D20 +262.49
(c 1.01, CHCl3).
1H NMR (600 MHz, CDCl3): δ = 1.30, 1.31 [2 s, 2 × 3 H, C(CH3)2],
3.71 (dd, J = 8.8, 11.2 Hz, 1 H, H2a′), 3.89 (d, J = 12.9 Hz, 1 H,
NCH2Ph), 3.98–4.01 (m, 2 H, H1′, H2b′), 4.06–4.07 (m, 1 H, H3),
13C NMR (75.4 MHz, CDCl3): δ = 14.1, 36.3, 40.9, 61.6, 61.7, 62.4,
66.7, 81.6, 127.7, 128.3, 129.2, 135.9, 157.7, 169.9.
Synthesis 2012, 44, 3783–3788
© Georg Thieme Verlag Stuttgart · New York

PAPER
Novel Synthesis of 4,5-Unsubstituted 2,3-Dihydroisoxazoles
3787
Anal. Calcd for C16H20N2O5: C, 59.99; H, 6.29; N, 8.74. Found: C,
60.25; H, 6.11; N, 8.59.
mixture was cooled in an ice/water bath, sat. aq NaHCO3 (5 mL)
was added, and the mixture was stirred for 5 min. MeCN was par-
tially removed by evaporation under reduced pressure. H2O (10
mL) was added and the mixture was extracted with CH2Cl2 (3 × 10
mL). The combined organic layers were washed with H2O and dried
(Na2SO4), and the solvent was evaporated in vacuo. The mixture of
two isomers 20 was isolated by column chromatography (silica gel,
hexanes–EtOAc, 60:40).
3,5-trans-18b
Colorless oil; yield: 90 mg (0.3 mmol, 55%).
1H NMR (300 MHz, CDCl3): δ = 1.28 (t, J = 7.0 Hz, 3 H,
OCH2CH3), 2.55 (ddd, J = 4.7, 7.6, 13.5 Hz, 1 H, H4a), 2.83 (ddd,
J = 6.4, 8.2, 13.5 Hz, 1 H, H4b), 3.54 (dd, J = 7.6, 8.2 Hz, 2 H,
NCH2CH2O), 3.74 (dd, J = 7.0, 7.6 Hz, 1 H, H3), 4.09 (d, J = 13.5
Hz, 1 H, NCH2Ph), 4.16 (d, J = 13.5 Hz, 1 H, NCH2Ph), 4.18 (q, J =
7.0 Hz, 2 H, OCH2CH3), 4.30 (dd, J = 7.0, 9.4 Hz, 2 H,
NCH2CH2O), 5.84 (dd, J = 4.7, 8.2 Hz, 1 H, H5), 7.28–7.37 (m, 5
H, Ph).
Unseparated 3,5-cis and 3,5-trans isomers of 20
Colorless oil; yield: 200 mg (0.8 mmol, 80%).
1H NMR (300 MHz, CDCl3): δ = 1.21 (t, J = 7.0 Hz, 3 H), 1.22 (t,
J = 7.0 Hz, 3 H), 2.39 (ddd, J = 1.2, 4.1, 13.5 Hz, 1 H), 2.53–2.64
(m, 2 H), 2.72 (ddd, J = 5.0, 8.2, 12.9 Hz, 1 H), 3.61 (dd, J = 4.1,
9.4 Hz, 1 H), 3.95 (dd, J = 7.6, 7.9 Hz, 1 H), 4.03 (d, J = 12.9 Hz, 1
H), 4.07–4.16 (m, 5 H), 4.25 (d, J = 12.9 Hz, 1 H), 4.38 (d, J = 12.9
Hz, 1 H), 5.51 (d, J = 4.1 Hz, 1 H), 5.63 (d, J = 4.7 Hz, 1 H), 7.28–
7.42 (m, 10 H, Ph).
13C NMR (75.4 MHz, CDCl3): δ = 14.2, 35.4, 40.4, 60.3, 61.5, 62.0,
65.3, 82.1, 127.7, 128.3, 129.3, 135.8, 157.1, 169.3.
Anal. Calcd for C16H20N2O5: C, 59.99; H, 6.29; N, 8.74. Found: C,
59.75; H, 6.48; N, 8.47.
13C NMR (150.8 MHz, CDCl3): δ = 14.0, 14.1, 40.4, 40.6, 61.4,
61.9, 62.5, 64.7, 65.3, 65.4, 96.7, 98.0, 127.5, 127.8, 128.3, 128.4,
129.2, 129.4, 135.8, 136.8, 170.7, 172.3.
3-(2-Benzyl-3-phenylisoxazolidin-5-yl)oxazolidin-2-one (19)
The reaction flask was charged with the 5-acetoxyisoxazolidine 5
(200 mg, 0.67 mmol), sealed with a rubber septum, and filled with
argon. Anhyd DCE was added (0.70 mL) followed by N-(trimethyl-
silyl)oxazolidin-2-one (17, 0.16 mL, 1.0 mmol). The soln was
cooled in an ice/water bath (2–5 °C) and TMSOTf (13 μL, 0.07
mmol) was added. The mixture was subsequently stirred at r.t. for
24 h. When the initial 5-acetoxyisoxazolidine had been consumed
(TLC, silica gel, hexanes–EtOAc, 60:40), the mixture was cooled in
an ice/water bath, sat. aq NaHCO3 (5 mL) and CH2Cl2 (5 mL) were
added, and the mixture was stirred for 5 min. The aqueous layer was
removed and extracted with CH2Cl2 (1 × 5 mL). The combined or-
ganic layers were washed with H2O and dried (Na2SO4), and the sol-
vent was evaporated in vacuo. The products were separated by
column chromatography (silica gel, hexanes–EtOAc, 70:30).
The mixture of 3,5-cis and 3,5-trans isomers 20 (185 mg, 0.74
mmol) was dissolved in anhyd DCE (0.80 mL), the reaction flask
was flushed with argon, and BSTFA (0.79 mL, 2.98 mmol) was
added dropwise. The mixture was stirred at r.t. for 2 h. The solvent
was evaporated to dryness in vacuo. The products were separated by
column chromatography (silica gel, hexanes–EtOAc, 90:10).
3,5-cis-21a
Colorless oil; yield: 50 mg (0.15 mmol, 21%).
1H NMR (600 MHz, CDCl3): δ = 0.06 [s, 9 H, OSi(CH3)3], 1.26 (t,
J = 6.9 Hz, 3 H, OCH2CH3), 2.53 (ddd, J = 2.0, 6.9, 13.0 Hz, 1 H,
H4a), 2.62 (ddd, J = 5.7, 8.9, 13.0 Hz, 1 H, H4b), 3.50 (dd, J = 7.3,
8.9 Hz, 1 H, H3), 3.99 (d, J = 13.4 Hz, 1 H, NCH2Ph), 4.11–4.20 (m,
3 H, NCH2Ph, OCH2CH3), 5.52 (dd, J = 2.0, 5.7 Hz, 1 H), 7.26 (t, 1
H, Ph), 7.30 (t, 2 H, Ph), 7.40 (d, 2 H, Ph).
3,5-cis-19a
Colorless solid; yield: 60 mg (0.18 mmol, 28%); mp 157–159 °C.
1H NMR (300 MHz, CDCl3): δ = 2.32 (ddd, J = 4.8, 9.9, 13.6 Hz, 1
H, H4a), 2.99 (ddd, J = 7.7, 8.1, 13.6 Hz, 1 H, H4b), 3.66 (d, J =
14.3 Hz, 1 H, NCH2Ph), 3.73 (td, J = 5.9, 8.4 Hz, 1 H, NCH2CH2O),
3.85 (dd, J = 7.7, 9.9 Hz, 1 H, H3), 3.88 (q, J = 8.4 Hz, 1 H,
NCH2CH2O), 3.97 (d, J = 14.3 Hz, 1 H, NCH2Ph), 4.23–4.36 (m, 2
H, NCH2CH2O), 5.92 (dd, J = 5.1, 8.4 Hz, 1 H, H5), 7.24–7.45 (m,
10 H, Ph).
13C NMR (150.8 MHz, CDCl3): δ = 0.1, 14.1, 42.5, 61.2, 62.0, 65.8,
95.7, 127.4, 128.1, 129.6, 136.4, 170.1.
Anal. Calcd for C16H25NO4Si: C, 59.41; H, 7.79; N, 4.33. Found: C,
59.69; H, 7.56; N, 4.60.
3,5-trans-21b
Colorless oil; yield: 115 mg (0.36 mmol, 48%).
13C NMR (75.4 MHz, CDCl3): δ = 40.6, 43.0, 59.8, 62.2, 70.8, 81.3,
127.2, 127.4, 128.2, 128.3, 128.7, 129.0, 137.4, 138.0, 157.8.
1H NMR (600 MHz, CDCl3): δ = 0.17 [s, 9 H, OSi(CH3)3], 1.19 (t,
J = 7.0 Hz, 3 H, OCH2CH3), 2.49 (ddd, J = 1.2, 7.0, 12.3 Hz, 1 H,
H4a), 2.70 (ddd, J = 4.7, 8.2, 12.3 Hz, 1 H, H4b), 3.91 (dd, J = 7.6,
8.2, 1 H, H3), 4.09 (q, J = 7.0 Hz, 2 H, OCH2CH3), 4.16 (d, J = 12.9
Hz, 1 H, NCH2Ph), 4.37 (d, J = 12.9 Hz, 1 H, NCH2Ph), 5.58 (d, J =
4.7 Hz, 1 H), 7.26 (t, 1 H, Ph), 7.31 (t, 2 H, Ph), 7.40 (d, 2 H, Ph).
Anal. Calcd for C19H20N2O3: C, 70.35; H, 6.21; N, 8.64. Found: C,
70.07; H, 6.18; N, 8.46.
3,5-trans-19b
Colorless solid; yield: 118 mg (0.36 mmol, 54%); mp 153–155 °C.
13C NMR (150.8 MHz, CDCl3): δ = 0.2, 14.1, 42.4, 61.2, 65.5, 65.6,
98.0, 127.4, 128.3, 129.2, 137.0, 171.0.
1H NMR (300 MHz, CDCl3): δ =2.53–2.69 (m, 2 H, H4a, H4b),
3.55–3.69 (m, 2 H, NCH2CH2O), 3.71 (d, J = 14.7 Hz, 1 H,
NCH2Ph), 3.84–3.95 (m, 1 H, H3), 4.01 (d, J = 14.7 Hz, 1 H,
NCH2Ph), 4.27–4.39 (m, 2 H, NCH2CH2O), 5.92 (dd, J = 4.0, 7.7
Hz, 1 H, H5), 7.23–7.43 (m, 10 H, Ph).
Anal. Calcd for C16H25NO4Si: C, 59.41; H, 7.79; N, 4.33. Found: C,
59.65; H, 7.67; N, 4.55.
13C NMR (75.4 MHz, CDCl3): δ = 40.0, 41.5, 58.9, 62.0, 68.6, 81.5,
127.3, 127.9, 128.1, 128.3, 128.8, 129.0, 136.6, 137.9, 157.2.
Acknowledgment
The authors are grateful to the Slovak Grant Agency (No. 1/0236/09
and No. 1/0115/10) and EU Structural Funds (ITMS 262 401
20001, 262 401 20025).
Anal. Calcd for C19H20N2O3: C, 70.35; H, 6.21; N, 8.64. Found: C,
70.10; H, 6.31; N, 8.34.
Ethyl 2-Benzyl-5-(trimethylsilyloxy)isoxazolidine-3-carboxyl-
ate (21)
The reaction flask was charged with the 5-acetoxyisoxazolidine 4
(300 mg, 1.0 mmol), sealed with a rubber septum, and filled with ar-
gon. Anhyd MeCN was added (2 mL) followed by TMSOH (0.28
mL, 2.5 mmol), and TMSOTf (0.04 mL, 0.2 mmol) was added. The
mixture was subsequently stirred at 40 °C for 6 h. Afterwards, the
References
(1) (a) Freeman, J. P. Chem. Rev. 1983, 83, 241. (b) Pinho e
Melo, T. M. V. D. Eur. J. Org. Chem. 2010, 3363.
© Georg Thieme Verlag Stuttgart · New York
Synthesis 2012, 44, 3783–3788

3788
R. Fischer et al.
PAPER
(9) Fischer, R.; Drucková, A.; Fišera, L.; Hametner, C.
(c) Chukanov, N. V.; Reznikov, V. A. Russ. Chem. Bull.
2011, 60, 379.
ARKIVOC 2002, (viii), 80.
(2) (a) Dhavale, D. D.; Trombini, C. J. Chem. Soc., Chem.
Commun. 1992, 1268. (b) Camiletti, Ch.; Dhavale, D. D.;
Gentilucci, L.; Trombini, C. J. Chem. Soc., Perkin Trans. 1
1993, 3157.
(3) Koizumi, T.; Hirai, H.; Yoshii, E. J. Org. Chem. 1982, 47,
4005.
(4) Keirs, D.; Moffat, D.; Overton, K.; Tomanek, R. J. Chem.
Soc., Perkin Trans. 1 1991, 1041.
(5) Ishikawa, T.; Kudoh, T.; Yoshida, J.; Yasuhara, A.; Manabe,
S.; Saito, S. Org. Lett. 2002, 4, 1907.
(6) Hýrošová, E.; Medvecký, M.; Fišera, L.; Hametner, C.;
Fröhlich, J.; Marchetti, M.; Allmaier, G. Tetrahedron 2008,
64, 3111.
(7) Chiacchio, U.; Gumina, G.; Rescifina, A.; Romeo, R.;
Uccella, N.; Casuscelli, F.; Piperno, A.; Romeo, G.
Tetrahedron 1996, 52, 8889.
(8) Merino, P.; Del Alamo, E. M.; Franco, S.; Merchan, F. L.;
Simon, A.; Tejero, T. Tetrahedron: Asymmetry 2000, 11,
1543.
(10) (a) Luisier, S.; Leumann, C. J. ChemBioChem 2008, 9, 2244.
(b) Šilhár, P.; Leumann, C. J. Bioorg. Med. Chem. 2010, 18,
7786.
(11) Vorbrüggen, H.; Ruh-Pohlenz, C. Org. React. 2000, 55, 1–
630.
(12) (a) Ochoa, C.; Provensio, R.; Jimeno, M. L.; Balzarini, J.; De
Clercq, E. Nucleosides Nucleotides 1998, 17, 901. (b) Liao,
J.; Sun, J.; Yu, B. Tetrahedron Lett. 2008, 49, 5036.
(13) Selected NMR data for 16: 1H NMR (300 MHz, CDCl3): δ =
2.35 (m, 2 H, H4), 2.45 (s, 3 H, COCH3), 3.87 (m, 1 H, H3),
6.25 (m, 1 H, H5); 13C NMR (75 MHz, CDCl3): δ = 24.0
(COCH3) and 172.0 (NHCOCH3).
(14) (a) Ballester, M.; Palomo, A. L. Synthesis 1983, 571.
(b) Aizpurua, J. M.; Palomo, C.; Palomo, A. L. Can. J.
Chem. 1984, 62, 336.
(15) Nguyen, T. B.; Martel, A.; Dhal, R.; Dujardin, G. J. Org.
Chem. 2008, 73, 2621.
Synthesis 2012, 44, 3783–3788
© Georg Thieme Verlag Stuttgart · New York