Reactions of hydroxylated sodium nitronates with acetic anhydride/pyridine

Article abstract of DOI:10.1016/S0040-4020(02)00487-8

Reactions with Ac₂O/Py of sodium nitronate salts derived from primary or secondary nitroalkanes bearing hydroxyl groups at γ or more remote positions have been studied. In all cases, the results could be explained through an acetic nitronic anhydride intermediate, whose evolution depends on conformational factors, and also on the type of the hydroxyl group.

Full text of DOI:10.1016/S0040-4020(02)00487-8

TETRAHEDRON
Pergamon
Tetrahedron 58 52002) 5327±5333
Reactions of hydroxylated sodium nitronates
with acetic anhydride/pyridine
p
M. V. Berrocal, M. V. Gil, E. Roman and J. A. Serrano
Â
 Â
Departamento de Quõmica Organica, Facultad de Ciencias, Universidad de Extremadura, 06071 Badajoz, Spain
Received 22 February 2002; accepted 6 May 2002
AbstractÐReactions with Ac2O/Py of sodium nitronate salts derived from primary or secondary nitroalkanes bearing hydroxyl groups at g
or more remote positions have been studied. In all cases, the results could be explained through an acetic nitronic anhydride intermediate,
whose evolution depends on conformational factors, and also on the type of the hydroxyl group. q 2002 Elsevier Science Ltd. All rights
reserved.
1. Introduction
To further explore the details of this type of reaction, we
have extended this method to include the use of other
secondary or primary alkyl sodium nitronates, with the
presence of g-hydroxyl groups as requisite. Herein, a full
account of our results is reported.
Although alkyl nitronate anions can be considered as very
important intermediates in synthetic organic chemistry,
studies on some of their reactions still remain relatively
lagged behind despite their extensive applications.1 In par-
ticular, reports in which those compounds react with acid
chlorides or anhydrides are scarce and rather old, most of
the results being fragmentary and sometimes divergent.2
The majority of explanations for these acylation processes
have proposed the initial formation of a mixed carboxylic±
nitronic acid anhydride;2,3 however, this type of substances
is generally very labile and only could be isolated 5usually
in poor yields) when metal salts of secondary nitroalkanes
were used as the starting materials.4
2. Results and discussion
First, with the aim to prepare the enantiomeric isoxazoline
N-oxides 4a and 4b, we carried out treatment with Ac2O/Py
of hydroxymethylene cyclohexene nitronates 5a and 5b. As
In relation with the above cited reactions, we have
described5 that on treatment with Ac2O/Py, the sodium
salts of d-galacto- and d-manno-5-glyco-4-nitrocyclo-
hexenes 1a and 1b led to highly stable chiral isoxazoline
N-oxides 3 in excellentyields. Based on setreochemical
arguments, we proposed a mechanism 5Scheme 1) in
which the not isolated acetic nitronic anhydride 2 could be
the intermediate that cyclizes to give the ®nal product.
Scheme 1.
Scheme 2. 5a) 4 M HCl/MeOH, re¯ux, 4 h; 5b) NaIO4, 1:2.3 MeOH/H2O,
08C, 15 min; 5c) NaBH4, MeOH, room temperature, 15 min; 5d) 2 M
NaOMe/MeOH, room temperature, 4 h; 5e) 1:20 Ac2O/Py, room tempera-
ture, 5 h; 5f) 1:1 Ac2O/Py, room temperature, 5 h.
Keywords: nitro compounds; nitronate salts; acetic nitronic anhydrides.
Corresponding author. Tel.: 134-924-289-381; fax: 134-924-271-149;
e-mail: roman@unex.es
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020502)00487-8

5328
M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
ratio 7:1) has been observed. However, we have not
attempted to separate them and also not assigned the signals
to their structures.
We think that the different behavior between nitronates 1
and 5 againstAc 2O/Py may be explained by considering two
circumstances: 5i) the hydroxyl group involved in the cycli-
zation is secondary in 1 and primary in 5; hence acetylation
of this group in the latter compound must be easier,7 thus
precluding, in part, the cyclization reaction leading to 4; 5ii)
the coupling constants between H-10 and H-5 in glyco-
cyclohexene nitronates 1a and 1b 59.7 and 3.1 Hz, respec-
tively) support conformations A and B 5Fig. 1), in which the
proximity between the hydroxyl group at C-10 and the nitro-
gen should favor the cyclization. On the contrary, this reac-
tion should be less favored in the case of nitronates 5,
Figure 1. Newman projections along the C10 ±C5 bonds in nitronates 1a
5A), 1b 5B) and 5 5C).
shown in Scheme 2, these latter compounds were obtained
from pentaacetylated trans-d-galacto- and d-manno-5-
glyco-4-nitrocyclohexenes6 6a,b through parallel processes
involving: 5a) acid-catalyzed deacetylation of their sugar
side-chains, leading to 7a,b;6 5b) oxidative cleavage of
these to give enantiomeric cyclohexene nitroaldehydes
8a,b;6 5c) sodium borohydride reduction of the carbonyl
group to yield 9a,b; and 5d), treatment of these with sodium
methoxide in methanol to produce the salts 5a,b.
0
00
because the intermediate values for J1 ,5 and J1 ,5 in these
compounds 5both of 4.6 Hz) are suggesting a conforma-
tional mixture C, in which the relative average positions
for the hydroxyl group and the nitrogen are more distant
than in A or B.
In order to extend the study of this reaction to the simplest
system bearing nitro and hydroxyl functions in 1,3-relative
positions, we have used 3-nitropropanol 128 as starting
material. As shown in Scheme 3, this compound has been
obtained by addition of sodium nitrite to acrolein, followed
by reduction of the resulting 3-nitropropanal with NaBH4;9
then, treatment of a methanolic solution of either 12, or it s
acetate 13, with an excess of sodium methoxide yielded
quantitatively sodium nitronate 14 as a white solid.
In contrast with what occurred from 1a,b, the reaction of
5a,b with Ac2O/Py 5Scheme 2, Step f) led to complex
mixtures, from which isoxazoline N-oxides 4a,b and acetic
nitronic anhydrides 11a,b were respectively isolated, as oils,
by preparative thin layer chromatography.
Because of their enantiomeric relationship, each pair of
compounds 4 and 11 showed spectral identity and nearly
equal and opposite values for their optical rotations. The 1H
and 13C NMR spectra of isoxazolines were very similar to
those of their nitronate precursors; as expected, the most
signi®cant differences between chemical shifts of these
two types of compounds were found in the methylene
group that is involved in the cyclization; thus, signals for
protons H-10 and H-100 54.67 and 4.07 ppm) and carbon C-10
570.5 ppm) in 4 appear considerably deshielded when they
are compared with the same in nitronates 5 5H-10,100 at3.42
and 3.35 ppm; C-10 at61.6 ppm).
The reaction of 14 with Ac2O/Py proceeded in a different
manner with what is described above from cyclohexene
nitronates. In this case, there was no evidence of the forma-
tion of either hypothetical acetic nitronic anhydride or
1
isoxazoline-N-oxide;10 instead, H NMR examination of
the crude mixture product showed that it contained, almost
exclusively, compounds 13, 15 and 16 in relative ratio
1.6:2.2:1, respectively. By preparative thin layer chroma-
tography of this mixture, we could isolate a pure analytical
sample of hydroxamic acid derivative 15,11 butno complete
separation was achieved between 3-acetoxy-1-nitropropane
1
13 and 2,3-diacetoxypropanenitrile 16.12 The H and 13C
The acetic nitronic anhydrides 11 were obtained as greenish
blue-colored oils, which showed no decomposition after
several weeks of storage at room temperature in the dessi-
cator. The NMR spectra of these compounds agree with
their structures, thus appearing two distinct acetate groups
and three ole®nic carbons. Furthermore, the existence of
geometrical isomers at the anhydride groups could be
deduced, because duplication of the signals 5approximate
NMR spectra of 15 in deuteriochloroform showed the
presence of three methyl and four carbonyl groups; at
room temperature, methylenic carbon C-2 535.5 ppm) and
its protons H-2 and H-20 53.10 and 3.00 ppm) gave broad
signals, probably due to restricted rotation of bonds C-1±C-
2 and C-1±N. Concerning nitrile 16, the H-2 proton appears
as a triplet 55.58 ppm, J,5.0 Hz), due to its couplings with
H-3 5dd, 4.46 ppm) and H-30 5dd, 4.35 ppm); resonances for
CuN, C-2 and C-3 carbons were found at114.4, 59.1 and
60.6 ppm, respectively.
As shown in Scheme 4, we propose that formation of
compounds 15 and 16 could occur with common inter-
mediates at the ®rst steps; thus, the initially formed mixed
carboxylic±nitronic anhydride 17 would suffer a N!C
acetate migration, probably through cyclic intermediate
18, affording a nitrosoderivative 19, which tautomerizes to
oxime 20; then, either 5a) rearrangement or 5b) 1,2-elimi-
nation of acetic acid would lead, respectively, to O-acyl
Scheme 3. 5a) NaNO2/HOAc, THF, 08C, 4 h; 5b) NaBH4/MeOH, 08C,
15 min; 5c) 1:1 Ac2O/Py, 08C, 12 h; 5d) 2 M NaOMe/MeOH, room
temperature, 5 h; 5e) 1:1 Ac2O/Py, room temperature, 5 h.

M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
5329
Scheme 4.
Scheme 6.
hydroxamic acid 21 and O-acyl ketene oxime 22, from
which the products 15 and 16 were originated.
fragmentC-2±C-7 of 26b showed chemical shifts and
coupling constants that were very similar to those described
for acyclic peracetylated 1-deoxy-1-nitro heptitols with the
21
Finally, we have also included in our study the previously
described 1-deoxy-1-nitro-d-glycero-l-manno heptitol13
23a and 1-deoxy-1-nitro-d-glycero-d-galacto heptitol14
23b. Treatment of either of these both compounds with
2 M aqueous sodium hydroxyde, followed by reaction
of the presumably formed sodium nitronates with acetic
anhydride and pyridine 5Scheme 5) afforded crude mixtures,
same con®gurations;16 also, an IR band at2350 cm
5CuN) and HRMS data are in agreement with the proposed
nitrile oxide structure. As was found for other compounds of
this type,17 26b proved to be unstable and decomposition
was observed, even at0 8C, after a few days of storage.
The above results can be explained by the processes shown
in Scheme 6 for 23b; thus, protonation of the intermediate
salt and acetylation of the hydroxyl groups would lead, after
elimination of acetic acid in 28b, to 1-nitroalkene 25b. On
the other hand, nucleophilic attack of OH-4 on C-1 of acetic
nitronic anhydride 29b would give, through intermediate
30b, the 1,4-lactone oxime 24b.18 The formation of nitrile
oxide 26b could be also explained from anhydride 29b, by
acetic acid elimination between the nitrogen atom and C-1.
1
which were analyzed by H NMR spectroscopy. From the
spectra, we observed that 23a led to a mixture containing,
almostexclusively, 24a, 25a13 and 27a in respective ratio
3.6:2.6:1.0. When starting from 23b, the result was some-
whatsimilar, and 24b, 25b14 and 26b 5relative ratio
4.5:3.5:1.0), together with other minor unidenti®ed
products, were present in the resulting mixture.15
Structural assignment for the new compounds are supported
by spectroscopic data 5IR, NMR and HRMS) from pure
samples, isolated by preparative thin layer chromatography.
Thus, for both types of lactone oximes 24 and 27, the H-2
protons were those encountered at the lowest ®eld 5doublets
at ca. 5.9 ppm), due to their proximity with the carbon±
nitrogen double bond. The ring-sizes were deduced from
chemical shifts for H-4 or H-5, that revealed if their adjacent
carbons had been acetylated; for 1,4-lactone oximes 24a,b
the signals for H-4 5dd at ca. 4.6 ppm) appeared at
higher ®eld than the other protons on secondary carbons
55.6±5.3 ppm), whereas the same is true for proton H-5
5dd at4.30 ppm) in 1,5-lacotne oxime 27a. Furthermore,
the presence of oxime carbons accorded with 13C NMR
resonances atca. 155 ppm.
Since coupling constants between protons on C-2±C-6
backbones of both 23a and 23b supported extended zig-zag
planar conformations 5as it is showed for 29b in Scheme 6),
we postulate that a nucleophilic substitution reaction
between OH-3 and the nitrogen leading to an isoxazoline-
N-oxide, would be unfavored. On the contrary, processes in
which OH-4 5or OH-5) attack at C-1, or those in which
cyclization does not occur, would be more feasible.
In conclusion, the foregoing new results complement
previous studies2,3 which were carried out for nitronates
under acetylation reaction conditions, where is proposed a
key mixed carboxylic±nitronic anhydride intermediate. In
addition, the presence of hydroxyl groups at g or more
remote positions allows the possibility of intramolecular
nucleophilic cyclizations, whose course is in¯uenced by
conformational effects and by the type of the involved
hydroxyl group. Other substrates are under current investi-
gation in order to extend the generality of and to improve the
methodology.
Besides the absence of signals for H-1, protons in the
3. Experimental
3.1. General
Solvents were evaporated under reduced pressure below
408C bath temperature. Melting points were determined
with an Electrothermal 8100 apparatus and are uncorrected.
Optical rotations were obtained at 20^28C with a
Scheme 5.

5330
M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
Perkin±Elmer 241 polarimeter. Infrared spectra were
recorded in the range 4000±600 cm21 with Perkin±Elmer
399 or Midac FT-IR spectrophotometers. NMR spectra
were recorded at20 8C on a Bruker spectrometer AM 400
colorless oil 50.15 g, 43%): Rf 0.68 53:1 hexane/ethyl ace-
tate); [a]Dˆ162.68, [a]578ˆ164.08, [a]546ˆ171.68,
[a]436ˆ1110.08 5c 0.50, CHCl3); nmax 5®lm) 5cm21) 2960,
2910, 2870, 2850 5C±H), 1740 5CvO), 1540, 1365 5NO2),
1220, 1040, 1020 5C±O); 1H NMR 5CDCl3) d 4.63 5td, 1H,
1
5400.13 MHz for H, 100.62 MHz for 13C) with TMS or
0
00
residual CHCl3/DMSO as internal standards. NMR assign-
ments were con®rmed by homonuclear double-resonance
experiments, and DEPT. Mass spectra were recorded on
a VG Autospec spectrometer. TLC was performed on
precoated plates of Silica Gel-60 GF254 5Merck), with
visualization of spots by UV light or iodine vapor, and the
solvent systems speci®ed. Preparative thin layer chroma-
tography was carried out on 0.20 mm Merck Silica Gel-60
PF254 plates.
J4,3a<J4,5ˆ9.9 Hz, J4,3bˆ5.3 Hz, H-4), 4.05 5dt, 2H, J1 ,1
ˆ
11.6 Hz, J1 ,5<J1 ,5ˆ5.0 Hz, H-10, H-100), 2.72 5br d, 1H,
H-3a), 2.64 5m, 1H, H-5), 2.49 5dd, 1H, J3a,3bˆ16.6 Hz,
H-3b), 2.18 5dd, 1H, J6a,6bˆ18.7 Hz, J5,6aˆ6.4 Hz, H-6a),
2.06 5br d, 1H, H-6b), 2.05 5s, 3H, OAc), 1.66 5s, 3H,
Me), 1.63 5s, 3H, Me); 13C NMR 5CDCl3) d 170.7
5OCOCH3), 124.3, 121.8 5C-1, C-2), 84.5 5C-4), 64.5
5C-10), 36.7 5C-5), 35.8, 32.9 5C-3, C-6), 20.6 5OCOCH3),
18.5 5Me-1, Me-2). CI MS m/z 5rel. int.): 228 5M1H, 19),
197 5M2NO, 8), 181 5M2NO2, 48), 168 5M2OAc, 13),
137 5M2NO±HOAc, 51), 121 5M2NO2±HOAc, 100).
HRMS 5CI) calcd for C11H17NO41H: 228.1235. Found
5M1H)1 228.1230.
0
00
3.1.1. +4S,5S)-5-Hydroxymethyl-1,2-dimethyl-4-nitro-
cyclohex-1-ene +9a). To a stirred solution of 54S,5S)-5-
formyl-1,2-dimethyl-4-nitrocyclohex-1-ene 8a6 50.41 g,
2.26 mmol) in methanol 512 mL) was added NaBH4
50.09 g, 2.26 mmol). After stirring for 15 min at room
temperature, the solution was diluted with water 512 mL)
and extracted with methylene dichloride 54£25 mL); then,
the extracts were washed with water 52£25 mL), dried
5MgSO4), and evaporated to yield 0.35 g 585%) of com-
pound 9a as a colorless oil: Rf 0.52 53:1 hexane/ethyl
acetate); [a]Dˆ168.38, [a]578ˆ170.98, [a]546ˆ178.58,
[a]436ˆ1116.78 5c 0.46, CHCl3); nmax 5®lm) 5cm21) 3550
and 3400 5OH), 2980, 2910, 2870, 2840 5C±H), 1545, 1365
5NO2), 1050, 1020 5C±O); 1H NMR 5CDCl3) d 4.72 5td, 1H,
3.1.4.
+4R,5R)-5-Acetoxymethyl-1,2-dimethyl-4-nitro-
cyclohex-1-ene +10b). Following the same procedure
above described for the preparation of its enantiomer 10a,
acetylation of 54R,5R)-5-hydroxymethyl-1,2-dimethyl-4-
nitrocyclohex-1-ene 9b 50.35 g, 1.89 mmol) led to the title
compound as a colorless oil 50.20 g, 46%): Rf 0.68 53:1 hex-
ane/ethyl acetate); [a]Dˆ261.08, [a]578ˆ262.38, [a]546
ˆ
270.48, [a]436ˆ2107.98 5c 0.50, CHCl3); IR, H and 13C
1
NMR data matched those for 10a.
0
00
J4,3a<J4,5ˆ10.1 Hz, J4,3bˆ5.6 Hz, H-4), 3.64 5dt, 1H, J1 ,1
ˆ
3.1.5. Sodium salt of +4S,5S)-5-hydroxymethyl-1,2-
dimethyl-4-nitrocyclohex-1-ene +5a). To a stirred solution
of 54S,5S)-5-hydroxymethyl-1,2-dimethyl-4-nitrocyclohex-
1-ene 9a 50.36 g, 1.94 mmol) in methanol 55 mL) was
added dropwise a solution of 2 M sodium methoxide in
methanol 51.22 mL, 2.44 mmol). After stirring for 4 h at
room temperature, evaporation of the solvent afforded 5a
as an amorphous hygroscopic solid, which was washed with
cold acetone and ®ltered 50.40 g, quantitative): mp 144±
1468C; [a]Dˆ122.28, [a]578ˆ124.18 5c 0.58, H2O); nmax
11.3 Hz, J1 ,5ˆ4.6 Hz, H-10), 3.59 5dt, 1H, J1 ,5ˆ4.6 Hz,
0
00
H-100), 2.73 5br dd, 1H, H-3a), 2.46 5dd, 1H, J3a,3b
17.1 Hz, H-3b), 2.38 5m, 1H, H-5), 2.14 5m, 2H, H-6a,
ˆ
0
00
H-6b), 1.74 5t, 1H, J1 ,OH<J1 ,OHˆ4.6 Hz, OH), 1.64 5s,
3H, Me), 1.63 5s, 3H, Me); 13C NMR 5CDCl3) d 124.8,
121.5 5C-1, C-2), 84.4 5C-4), 62.9 5C-10), 39.8 5C-5), 35.7,
30.1 5C-3, C-6), 18.5 5Me-1, Me-2). CI MS m/z 5rel. int.):
186 5M1H, 1), 168 5M2OH, 6), 138 5M2H±NO2, 19), 121
5M2H±OH±NO2, 28), 107 5M2OH±NO2±CH3, 100).
HRMS 5CI) calcd for C9H15NO31H: 186.1130. Found
5M1H)1 186.1122.
5KBr) 5cm21) 3500±3200 5OH), 2928 5C±H), 1155, 1020
1
0
00
5C±O); H NMR 5D2O) d 3.42 5dd, 1H, J1 ,1 ˆ10.6 Hz,
J1 ,5ˆ6.4 Hz, H-10), 3.35 5dd, 1H, J1 ,5ˆ8.7 Hz, H-100),
3.24 5m, 1H, H-5), 2.93 5d, 1H, J3a,3bˆ22.8 Hz, H-3a),
2.63 5d, 1H, H-3b), 2.19 5br d, 1H, H-6a), 2.02 5d, 1H,
J6a,6bˆ14.7 Hz, H-6b), 1.56 5s, 6H, Me-1, Me-2); 13C
NMR 5D2O) d 127.3 5C-4), 123.9, 122.6 5C-1, C-2), 61.6
5C-10), 37.6 5C-5), 32.4, 31.3 5C-3, C-6), 18.7, 18.0 5Me-1,
Me-2).
0
00
3.1.2. +4R,5R)-5-Hydroxymethyl-1,2-dimethyl-4-nitro-
cyclohex-1-ene +9b). Following the same procedure above
described for the preparation of its enantiomer 9a, reduction
of 54R,5R)-5-formyl-1,2-dimethyl-4-nitrocyclohex-1-ene
8b6 50.49 g, 2.67 mmol) led to the title compound as a color-
less oil 50.38 g, 77%): Rf 0.52 53:1 hexane/ethyl acetate);
[a]Dˆ267.58, [a]578ˆ270.38, [a]546ˆ278.68, [a]436
ˆ
2114.68 5c 0.47, CHCl3); IR, 1H and 13C NMR data matched
those for 9a.
3.1.6. Sodium salt of +4R,5R)-5-hydroxymethyl-1,2-
dimethyl-4-nitrocyclohex-1-ene +5b). Following the same
procedure above described for the preparation of its enan-
tiomer 5a, 54R,5R)-5-hydroxymethyl-1,2-dimethyl-4-nitro-
cyclohex-1-ene 9b 50.50 g, 2.70 mmol) led to the title
compound as an amorphous hygroscopic solid 50.56 g,
3.1.3. +4S,5S)-5-Acetoxymethyl-1,2-dimethyl-4-nitrocy-
clohex-1-ene +10a). To a stirred solution of 54S,5S)-5-
hydroxymethyl-1,2-dimethyl-4-nitrocyclohex-1-ene
9a
50.29 g, 1.57 mmol) in dry pyridine 53 mL) was added acetic
anhydride 50.15 mL). After stirring for 5 h at room tempera-
ture, the solution was poured onto ice cold water, extracted
with methylene dichloride 54£25 mL) and washed succes-
sively with 1 M hydrochloric acid 52£25 mL), saturated
aqueous sodium hydrogencarbonate 52£25 mL) and water
52£25 mL). The organic layer was dried 5MgSO4) and
the solvent evaporated, yielding the title compound as a
quantitative): mp 144±1468C; [a]Dˆ121.18, [a]578
ˆ
1
123.28 5c 0.57, H2O); IR, H and 13C NMR data matched
those for 5a.
3.1.7.
+3aS)-5,6-Dimethyl-+3,3a,4,7)-tetrahydrobenz-
isoxazoline N-oxide +4a) and 1,2-dimethyl-+5S)-
5-acetoxymethyl-4-+O-acetyl-aci-nitro)-cyclohex-1-ene
+11a). Method A. Following the same procedure above

M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
5331
described for the preparation of 10a, treatment of the
sodium saltof 54 S,5S)-5-hydroxymethyl-1,2-dimethyl-4-
nitrocyclohex-1-ene 5a 50.92 g, 4.42 mmol) with pyridine
59 mL) and acetic anhydride 59 mL) led to a crude oily
residue 50.68 g), which was submitted to preparative thin
layer chromatography 51:1 hexane/ethyl acetate), thus
affording analytically pure samples of the title compounds.
Data for 4a. Colorless oil 50.075 g, 10%); Rf 0.46 51:1
hexane/ethyl acetate); [a]Dˆ1148.68, [a]578ˆ1156.48,
[a]546ˆ1169.68 5c 0.28, CHCl3); nmax 5®lm) 5cm21) 2970,
[a]578ˆ2781.08, [a]546ˆ2626.48, [a]436ˆ2476.88 5c
1
0.40, CHCl3). IR, H and 13C NMR data matched those
for 11a.
3.1.9. 3-Acetoxy-1-nitropropane +13). Following the same
procedure above described for the preparation of 10a, treat-
mentof 3-nriot propanol
128 50.25 g, 2.21 mmol) with
pyridine 52.5 mL) and acetic anhydride 52.5 mL) led to the
title compound as a colorless oil 50.28 g, 93%): Rf 0.70 51:1
hexane/ethyl acetate); nmax 5®lm) 5cm21) 2992 5C±H), 1740
5CvO), 1557, 1385 5NO2), 1240 5C±O); 1H NMR 5CDCl3)
d 4.47 5t, 2H, J2,3ˆ6.7 Hz, H-3, H-30), 4.17 5t, 2H, J1,2ˆ
6.0 Hz, H-1, H-10), 2.34 5m, 2H, H-2, H-20), 2.04 5s,
3H, OAc); 13C NMR 5CDCl3) d 170.6 5OCOCH3), 72.2
5C-3), 60.6 5C-1), 26.3 5C-2). CI MS m/z 5rel. int.): 148
5M1H, 19), 130 5M1H±H2O, 8), 115 5M1H±H2O±CH3,
45), 101 5M2NO2, 53), 88 5M1H±HOAc, 100). HRMS
5CI) calcd for C5H9NO41H: 148.0609. Found 5M1H)1
148.0584.
2920, 2900, 2860 5C±H), 1700, 1650 5CvN), 1240, 1220
1
0
5C±O); H NMR 5CDCl3) d 4.67 5dd, 1H, J1 ,5ˆ9.3 Hz,
H-10), 4.07 5t, 1H, J1 ,5<J1 ,1 ˆ8.1 Hz, H-1 ), 3.54 5m,
1H, H-5), 3.02 5d, 1H, J3a,3bˆ21.5 Hz, H-3a), 2.90 5d, 1H,
H-3b), 2.36 5dd, 1H, J6a,6bˆ15.9 Hz, J5,6aˆ7.0 Hz, H-6a),
2.21 5br dd, 1H, J5,6bˆ11.7 Hz, H-6b), 1.73 5s, 3H, Me),
1.68 5s, 3H, Me); 13C NMR 5CDCl3) d 124.5, 122.3 5C-1,
C-2), 115.6 5C-4), 70.5 5C-10), 40.4 5C-5), 36.8, 29.1 5C-3,
C-6), 19.2, 18.7 5Me-1, Me-2). HRMS 5CI) calcd for
C9H13NO2: 167.0946. Found M1 167.0947. Data for 11a.
Greenish blue-colored oil 50.155 g, 13%); Rf 0.43 51:1
hexane/ethyl acetate); [a]Dˆ1852.08, [a]578ˆ1790.08,
[a]546ˆ1631.58, [a]436ˆ1481.18 5c 0.46, CHCl3); nmax
00
00
0
00
3.1.10. Sodium salt of 3-nitropropanol +14). Method A. To
a stirred solution of 3-nitropropanol 128 51.00 g, 9.51 mmol)
in methanol 514 mL) was added dropwise 2 M sodium
methoxide in methanol 55.91 mL, 11.80 mmol). After 5 h
at room temperature, the reaction mixture was treated
with activated charcoal, ®ltered and the solvent evaporated,
thus affording 1.07 g 5quantitative yield) of compound 14
as a crude white solid, which was washed with cold
acetone: Rf 0.40 510:1 ethyl acetate/ethanol); nmax 5KBr)
5cm21) 3000±3500 5OH), 2900±2800 5C±H), 1700, 1638
5®lm) 5cm21) 2915 5C±H), 1744 5CvN), 1225, 1042
1
0
5C±O); H NMR 5CDCl3) d 4.42 5dd, 1H, J1 ,5ˆ4.2 Hz,
J1 ,1 ˆ11.2 Hz, H-1 ), 3.93 5dd, 1H, J1 ,5ˆ7.6 Hz, H-100),
2.94 5d, 1H, J3a,3bˆ18.2 Hz, H-3a), 2.62 5m, 1H, H-5),
2.33 5dd, 1H, J5,6aˆ4.5 Hz, J6a,6bˆ17.9 Hz, H-6a), 2.27 5br
d, 1H, H-3b), 2.06 5dd, 1H, J5,6bˆ9.9 Hz, H-6b), 1.74 5s, 3H,
Me), 1.71 5s, 3H, Me); 13C NMR 5CDCl3) d 170.7, 168.4
5OCOCH3), 123.8, 122.5, 122.0 5C-1, C-2, C-4), 62.1
5C-10), 40.1 5C-5), 35.8, 32.4 5C-3, C-6), 21.1, 20.7
5OCOCH3), 18.7, 18.5 5Me-1, Me-2); HRMS 5CI) calcd
for C13H19NO51H: 270.1341. Found 5M1H)1 270.1354.
0
0
00
00
1
5CvN); H NMR 5D2O) d 6.20 5t, 1H, J2,3ˆ6.2 Hz, H-3),
3.72 5t, 2H, J1,2ˆ6.4 Hz, H-1, H-10), 2.50 5m, 2H, H-2,
H-20); 13C NMR 5D2O) d 117.1 5C-3), 59.2 5C-1), 30.8
5C-2).
Method B. To a stirred solution of 54S,5S)-5-hydroxy-
methyl-1,2-dimethyl-4-nitrocyclohex-1-ene 9a 50.10 g,
0.54 mmol) in methanol 52 mL) was added dropwise a solu-
tion of 2 M sodium methoxide in methanol 50.35 mL,
0.68 mmol). After stirring for 5 h at room temperature,
removal of the solvent afforded 0.11 g of an amorphous
solid, which was dissolved in dry pyridine 53 mL) and
treated with acetic anhydride 53 mL) for 5 h; then, work-
up of the reaction mixture as described for 10a, yielded a
crude oily residue 50.07 g) from which analytically pure
samples of the title compounds could be isolated, by
preparative thin layer chromatography 51:1 hexane/ethyl
acetate).
Method B. Treatment of 3-acetoxy-1-nitropropane 13 as
described above for 12 yielded quantitatively the salt 14.
3.1.11. 3-Acetoxy-1-nitropropane +13), 3-acetoxy-+N-ace-
tyl-N-acetoxy)-propionamide +15) and 2,3-diacetoxy-
propanenitrile +16). Following the same procedure above
described for the preparation of 10a, treatment of the
sodium saltof 3-nriot propanol 14 50.97 g, 7.63 mmol)
with pyridine 510 mL) and acetic anhydride 510 mL), led
1
to a crude oily residue 51.15 g), whose H NMR spectrum
showed it contained, almost exclusively, a mixture of the
title compounds 13, 15 and 16 in a relative ratio 1.6:2.2:1,
respectively. Preparative thin layer chromatography of this
mixture 51:1 hexane/ethyl acetate) afforded a fraction with
pure 15, whereas both 13 and 16 remain inseparable;
thus spectral data for 16 were obtained after subtracting
the spectra of 13 from the spectra of mixture 13116.
Data for 15. Colorless oil; Rf 0.60 51:1 hexane/ethyl
acetate); nmax 5®lm) 5cm21) 1734 5CvO), 1557, 1368
3.1.8.
+3aR)-5,6-Dimethyl-+3,3a,4,7)-tetrahydrobenz-
isoxazoline N-oxide +4b) and 1,2-dimethyl-+5R)-5-
acetoxymethyl-4-+O-acetyl-aci-nitro)-cyclohex-1-ene +11b).
Following the same procedure above described in Method A
for the preparation of their enantiomers 4a and 11a, treat-
mentof 5b 50.50 g, 2.40 mmol) with pyridine 55 mL) and
acetic anhydride 55 mL) led to an oily residue 50.38 g), from
which the title compounds could be isolated, by preparative
thin layer chromatography 51:1 hexane/ethyl acetate). Data
for 4b. Colorless oil 50.037 g, 9%); Rf 0.46 51:1 hexane/ethyl
acetate); [a]Dˆ2151.28, [a]578ˆ2159.78, [a]546ˆ2173.48
1
5NO2), 1240, 1042 5C±O); H NMR 5CDCl3) d 4.40 5t,
2H, J2,3ˆ6.2 Hz, H-3, H-30), 3.10 5m, 1H, H-2), 3.00 5m,
1H, H-20), 2.38 5br s, 3H, NAc), 2.33 5s, 3H, NOAc), 2.06
5br s, 3H, OAc); 13C NMR 5CDCl3) d 170.7 5OCOCH3),
167.6 5C-1, NCOCH3), 167.2 5NOCOCH3), 58.6 5C-3),
35.5 5C-2), 24.0 5NOCOCH3), 20.7 5OCOCH3), 17.8
5NCOCH3). CI MS m/z 5rel. int.): 232 5M1H, 90), 190
5M1H±C2H2O, 81), 172 5M1H±HOAc, 77), 130 5M1
H±HOAc±C2H2O, 43), 115 5M2C4H6NO3, 100). HRMS
1
5c 0.30, CHCl3). IR, H and 13C NMR data matched those
for 4a. Data for 11b. Greenish blue-colored oil 50.071 g,
11%); Rf 0.43 51:1 hexane/ethyl acetate); [a]Dˆ2842.08,

5332
M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
5CI) calcd for C9H13NO61H: 232.0811. Found 5M1H)1
232.0810. Data for 16. Rf 0.70 51:1 hexane/ethyl acetate);
3.1.13. N-Acetoxy-2,3,5,6,7-penta-O-acetyl-d-glycero-d-
galacto-1,4-lactone oxime +24b), +E)-3,4,5,6,7-penta-O-
acetyl-1-nitro-d-manno-hept-1-ene +25b) and 2,3,4,5,
6,7-hexa-O-acetyl-d-glycero-d-galacto-heptononitrile
oxide +26b). Following the procedure described in Section
3.1.12, 1-deoxy-1-nitro-d-glycero-d-galacto-heptitol 23b14
51.00 g, 4.15 mmol) led to an oily residue 51.25 g), whose
1H NMR spectrum showed that it contained a mixture of
24b, 26b, and known 25b14 54.5:1.0:3.5 respective ratio),
together with other minor unidenti®ed products. Analyti-
cally pure samples of these three products could be isolated
by preparative thin layer chromatography 51:1 hexane/ethyl
acetate). Data for 24b. Colorless oil; Rf 0.24 51:1 hexane/
n
max 5®lm) 5cm21) 2327 5CN); 1H NMR 5CDCl3) d 5.58 5t,
0
0
1H, J2,3ˆ4.5 Hz, J2,3 ˆ5.7 Hz, H-2), 4.69 5dd, 1H, J3,3
ˆ
12.0 Hz, H-3), 4.36 5dd, 1H, H-30); 13C NMR 5CDCl3) d
170.6, 169.7 5OCOCH3), 117.7 5CN), 61.6 5C-3), 59.1
5C-2), 20.3 5OCOCH3). CI MS m/z 5rel. int.): 172 5M1H,
4), 130 5M1H±C2H2O, 6), 112 5M1H±HOAc, 100),
88 5M1H±C2H2O±C2H2, 51). HRMS 5CI) calcd for
C7H9NO41H: 172.0609. Found 5M1H)1 172.0604.
3.1.12. N-Acetoxy-2,3,5,6,7-penta-O-acetyl-d-glycero-l-
manno-1,4-lactone oxime +24a), +E)-3,4,5,6,7-penta-O-
acetyl-1-nitro-d-galacto-hept-1-ene +25a) and N-acetoxy-
2,3,5,6,7-penta-O-acetyl-d-glycero-l-manno-1,5-lactone
oxime +27a). To a stirred solution of 1-deoxy-1-nitro-d-
glycero-l-manno-heptitol 23a13 51.00 g, 4.15 mmol) in
2 M NaOH 510 mL) was added cold methanol 540 mL).
After several minutes at 08C, there was apparition of a
white precipitate and the reaction mixture was kept in the
refrigerator for 24 h; then, the resulting hygroscopic solid
was ®ltered, washed on the ®lter with cold methanol and
dried in a vacuum dessicator 51.09 g). Subsequently, this
solid was treated with pyridine 54 mL) and acetic anhydride
58 mL) as described for 10a, thus affording a crude oily
residue 51.37 g), whose 1H NMR spectrum showed it
contained, almost exclusively, a mixture of 24a, 27a, and
known 25a,13 in a relative ratio 3.6:1.0:2.6, respectively.
Analytically pure samples of these three products could be
isolated by preparative thin layer chromatography 51:1
hexane/ethyl acetate). Data for 24a. Colorless oil; Rf 0.22
51:1 hexane/ethyl acetate); [a]Dˆ18.08, [a]578ˆ17.38,
[a]546ˆ18.98, [a]436ˆ118.08, [a]365ˆ128.98 5c 0.80,
CHCl3); nmax 5®lm) 5cm21) 2954 5C±H), 1755 5CvO),
ethyl acetate); [a]Dˆ140.78, [a]578ˆ143.48, [a]546
ˆ
149.18, [a]436ˆ186.38 5c 0.7, CHCl3); nmax 5®lm) 5cm21
)
2954 5C-H), 1746 5CvO), 1688 5CvN), 1215, 1047
5C±O); 1H NMR 5CDCl3) d 5.84 5d, 1H, J2,3ˆ2.9 Hz,
H-2), 5.51 5dd, 1H, J4,5ˆ3.8 Hz, J5,6ˆ6.4 Hz, H-5), 5.28
5m, 1H, H-6), 5.19 5t, 1H, J3,4ˆ3.2 Hz, H-3), 4.66 5t, 1H,
0
H-4), 4.41 5dd, 1H, J7,7 ˆ12.6 Hz, J6,7ˆ2.9 Hz, H-7), 4.22
0
0
5dd, 1H, J6,7 ˆ5.4 Hz, H-7 ), 2.18 5s, 3H, OAc), 2.16 5s, 3H,
OAc), 2.14 5s, 3H, OAc), 2.13 5s, 3H, OAc), 2.11 5s, 3H,
OAc), 2.07 5s, 3H, OAc); 13C NMR 5CDCl3) d 170.7, 169.9,
169.7, 169.2, 167.8 5OCOCH3), 159.6 5C-1), 85.1 5C-4),
75.4, 73.0, 69.9, 68.8 5C-2, C-3, C-5, C-6), 61.4 5C-7),
21.0, 20.9, 20.7, 19.3 5OCOCH3 and NOCOCH3). CI MS
m/z 5rel. int.): 476 5M1H, 75), 433 5M2C2H2O, 79), 416
5M2OAc, 14), 392 5M2C2H2O±C2H2N, 7), 374 5M2
OAc±C2H2O, 19), 373 5M2C2H2O±HOAc, 33), 332 5M2
2C2H2O±C2H2N±HOAc, 19), 314 5M2OAc±C2H2O±
HOAc, 13), 290 5M22C2H2O±C2H2N±HOAc, 19), 254
5M2OAc±C2H2O±2HOAc, 25), 211 5M2OAc±C2H2O±
2HOAc±NHCO, 50), 169 5M2OAc±2C2H2O±2HOAc±
NHCO, 100). HRMS 5CI) calcd for C19H25O13N1H:
476.1404. Found 5M1H)1 476.1385. Data for 26b. Color-
less oil; Rf 0.15 51:1 hexane/ethyl acetate); nmax 5®lm)
5cm21) 2940 5C±H), 2350 5CuN), 1753 5CvO), 1227,
1044 5C±O); 1H NMR 5CDCl3) d 5.54 5dd, 1H, J2,3ˆ
1.0 Hz, J3,4ˆ11.7 Hz, H-3), 5.47 5dd, 1H, J4,5ˆ1.3 Hz,
H-4), 5.35 5d, 1H, J5,6ˆ8.9 Hz, H-5), 5.20 5d, 1H, H-2),
0
1
1694 5CvN), 1260, 1036 5C±O); H NMR 5CDCl3) d
5.97 5d, 1H, J2,3ˆ4.5 Hz, H-2), 5.69 5dd, 1H, J3,4ˆ2.9 Hz,
H-3), 5.54 5dd, 1H, J4,5ˆ9.5 Hz, J5,6ˆ2.6 Hz, H-5), 5.49
5ddd, 1H, H-6), 4.65 5dd, 1H, H-4), 4.32 5dd, 1H, J6,7ˆ
0
0
5.4 Hz, J7,7 ˆ11.7 Hz, H-7), 4.01 5dd, 1H, J6,7 ˆ6.5 Hz,
H-70), 2.16 5s, 3H, OAc), 2.13 5s, 3H, OAc), 2.10 5s, 6H,
2OAc), 2.05 5s, 6H, 2OAc); 13C NMR 5CDCl3) d 170.3,
169.5, 169.4, 169.1 5OCOCH3), 167.6 5NOCOCH3), 158.0
5C-1), 78.9 5C-4), 68.7, 68.6, 66.4 5C-2, C-3, C-5, C-6), 61.5
5C-7), 20.6, 20.3, 19.1 5OCOCH3 and NOCOCH3). CI MS
m/z 5rel. int.): 476 5M1H, 100), 434 5M2C2H3N, 32),
415 5M2HOAc, 15), 392 5M2C2H3N±C2H2O, 7),
374 5M2HOAc±C2H3N, 11), 331 5M2HOAc±C2H3N±
COCH3, 15), 314 5M22HOAc±C2H3N, 5). HRMS 5CI)
calcd for C19H25O13N1H: 476.1404. Found 5M1H)1
476.1395. Data for 27a. Colorless oil; Rf 0.12 51:1 hexane/
ethyl acetate); [a]578ˆ20.88, [a]546ˆ21.68, [a]436ˆ24.08
5c 0.50, CHCl3); nmax 5®lm) 5cm21) 2959 5C±H), 1748
5CvO), 1661 5CvN), 1223, 1040 5C±O); 1H NMR
5CDCl3) d 6.01 5d, 1H, J2,3ˆ3.6 Hz, H-2), 5.43 5dd, 1H,
J4,5ˆ9.5 Hz, J3,4ˆ8.3 Hz, H-4), 5.33 5ddd, 1H, H-6), 5.29
5.05 5m, 1H, H-6), 4.21 5dd, 1H, J6,7ˆ2.6 Hz, J7,7
ˆ
0
0
12.5 Hz, H-7), 4.01 5dd, 1H, J6,7 ˆ5.4 Hz, H-7 ), 2.19 5s,
3H, OAc), 2.10 5s, 3H, OAc), 2.09 5s, 3H, OAc), 2.08 5s,
3H, OAc), 2.05 5s, 3H, OAc), 2.04 5s, 3H, OAc). 13C NMR
5CDCl3) d 170.5, 170.1, 170.0, 169.8, 169.4, 168.9
5OCOCH3), 67.9, 67.3, 67.0, 66.9, 66.7 5C-2, C-3, C-4,
C-5, C-6), 61.9 5C-7), 20.8, 20.6, 20.5, 20.3, 18.2
5OCOCH3). CI MS m/z 5rel. int.): 476 5M1H, 44), 433
5M2C2H2O, 67), 416 5M2OAc, 13), 374 5M2OAc±
C2H2O, 8), 359 5M2C2H2O±CH3CO±NOH, 100), 331
5M2OAc±C2H2O±CH3CO, 15), 328 5M22HOAc±NOH,
28), 314 5M2C2H2O±AcOH±OAc, 8). HRMS 5CI) calcd
for C19H25O13N1H: 476.1404. Found 5M1H)1 476.1424.
0
5dd, 1H, H-3), 4.46 5dd, 1H, J6,7ˆ6.0 Hz, J7,7 ˆ11.5 Hz,
H-7), 4.30 5dd, 1H, J5,6ˆ2.7 Hz, H-5) 4.41 5dd, 1H,
Acknowledgements
0
0
J6,7 ˆ7.0 Hz, H-7 ), 2.16 5s, 3H, OAc), 2.08 5s, 3H, OAc),
2.02 5s, 3H, OAc); 13C NMR 5CDCl3) d 170.3, 169.8, 169.4,
169.1, 168.8, 167.4 5OCOCH3 and NOCOCH3), 154.2
5C-1), 69.5, 66.8, 65.2, 64.3 5C-2, C-3, C-4, C-5, C-6),
61.2 5C-7), 20.7, 20.6, 20.5, 19.3 5OCOCH3 and
NOCOCH3).
This work was supported by the Spanish Ministerio de
Â
Educacion y Cultura 5CICYT, PB98-0997) and the Junta
de Extremadura-Fondo Social Europeo through grant
Â
IPR00C021. We also thank the Unidad de Espectrometrõa
de Masas de la Universidad de Cordoba 5Spain) for mass
Â

M. V. Berrocal et al. / Tetrahedron 58 02002) 5327±5333
5333
10. The ease of the acetylation of primary hydroxyl group,
together with the long distance between the hydroxyl and
the nitrogen may be responsible for the lack of isoxazoline-
N-oxide.
spectra, and the Universidad de Extremadura 5Programa
Propio) for a postdoctoral fellowship to M. V. Gil.
References
11. Hydroxamic acid derivatives as 15 have been detected, or
isolated, in reactions between either primary nitroalkanes or
their nitronate salts with acylating agents, such as acetyl
chlorides and acid anhydrides. Several mechanisms to justify
their formation are given in Refs. 2, 3.
1. 5a) Kornblum, N. Angew. Chem., Int. Ed. 1975, 14, 734.
5b) Hand, E. S.; Jencks, W. P. J. Am. Chem. Soc. 1975, 97,
6221. 5c) Rosini, G.; Ballini, R. Synthesis 1988, 833. 5d) Baer,
H. H. Adv. Carbohydr. Chem. 1969, 24, 67. 5e) Torssell,
K. B. G. Nitrile Oxides, Nitrones and Nitronates in Organic
Synthesis; VCH: New York, 1988.
12. As far as we are aware, compound 16 has never been described
in the literature. Its deacetylated derivative appears cited as
cyanoethylene glycol, as a part of a polymer used in the
preparation of synthetic ®bres: Hiroshi, O.; Tetsuya, T.;
Tetsunori, M.; Masami, O.; Jpn Patent 10,096,118, 1998;
Chem. Abstr. 1998, 128, 309410.
2. Cherest, M.; Lusinchi, X. Tetrahedron 1986, 42, 3825 and
references cited therein.
3. 5a) Ste¯, E. P.; Dull, M. F. J. Am. Chem. Soc. 1947, 69, 3037.
5b) McKillop, A.; Kobylecki, R. J. Tetrahedron 1974, 30,
1365. 5c) Stermitz, F. R.; Neiswander, D. W. Tetrahedron
1975, 31, 655. 5d) Guillaumel, J.; Demerseman, P.; Clavel,
J. M.; Royer, R.; Platzer, N.; Brevard, C. Tetrahedron 1980,
36, 2459.
13. Sowden, J. C.; Strobach, D. R. J. Am. Chem. Soc. 1960, 82,
954.
14. Sowden, J. C.; Schaffer, R. J. Am. Chem. Soc. 1951, 73, 4662.
15. Although it had been described that 1,3-dipolar cycloadditions
occur between benzonitrile, bromoformonitrile or acetonitrile
oxides and sugar nitro ole®ns, we have not detected, under our
conditions, any product from the reaction between 25b and
4. Nielsen, A. T. In Nitronic Acids and Esters in The Chemistry
of Nitro and Nitroso Groups, Feuer, H., Ed.; Interscience:
New York, 1970; pp. 461±473 Part1.
Â
26b: Mancera, M.; Roffe, I.; Galbis, J. A. Tetrahedron 1995,
Â
5. Gil, M. V.; Roman, E.; Serrano, J. A. Tetrahedron Lett. 2000,
51, 6349.
16. Bueno, M.; Eliel, E. L.; Serrano, J. A.; Roman, E. Magn.
Reson. Chem. 1987, 25, 1021.
41, 3221.
6. Serrano, J. A.; Moreno, M. Ch.; Roman, E.; Arjona, O.;
Â
Â
Â
Plumet, J.; Jimenez, J. J. Chem. Soc., Perkin Trans. 1 1991,
3207.
17. 5a) Tronchet, J. M. J.; Le-Hong, N. Carbohydr. Res. 1973, 29,
311. 5b) Dawson, M.; Johnson, T.; Paton, R. M.; Rennie,
R. A. C. J. Chem. Soc., Chem. Commun. 1988, 1339.
5c) Yao, C.; Chen, W.; Lin, Y. Tetrahedron Lett. 1996, 37,
6339.
7. Wolfrom, M. L.; Szarek, W. A. The Carbohydrates:
Chemistry and Biochemistry; Pigman, W., Horton, D., Eds.;
Academic: New York, 1972; Vol. IA, p. 218.
8. Baer, H. H.; Chin, S. H. L.; Shields, D. C. Can. J. Chem. 1973,
51, 2828.
18. A similar attack of HO-5 on C-1 of the hypothetic acetic
nitronic anhydride from 23a would explain the formation of
the 1,5-lactone oxime 27a.
È
9. Griesser, H.; Ohrlein, R.; Schwap, W.; Ehrler, R.; Jager, V.
È
Org. Synth. 1999, 77, 236.