N-Methoxy-1,2,4-Dioxazolidines by Ozonolysis Reactions

Article abstract of DOI:10.1021/jo971469o

Ozonolysis of ethyl vinyl ether (7a) in the presence of the O-methylated oxime of cyclohexanone (8) afforded N-methoxy-1,2-dioxa-4-azaspiro[4.5]decane (9a), and ozonolyses of the O-methylated dioximes of 1,4- and 1,5-dicarbonyl compounds (10a-e) afforded N-methoxylated bicyclic 1,2,4-dioxazolidines (12a-e).

Full text of DOI:10.1021/jo971469o

1086
J . Org. Chem. 1998, 63, 1086-1089
N-Meth oxy-1,2,4-Dioxa zolid in es by Ozon olysis Rea ction s
Karl Griesbaum,* Xuejun Liu, and Henning Henke1
Engler-Bunte-Institut, Bereich Petrochemie, Universita¨t Karlsruhe (TH) D-76128 Karlsruhe, Germany
Received August 7, 1997
Ozonolysis of ethyl vinyl ether (7a ) in the presence of the O-methylated oxime of cyclohexanone (8)
afforded N-methoxy-1,2-dioxa-4-azaspiro[4.5]decane (9a ), and ozonolyses of the O-methylated
dioximes of 1,4- and 1,5-dicarbonyl compounds (10a -e) afforded N-methoxylated bicyclic 1,2,4-
dioxazolidines (12a -e).
In tr od u ction
It has been reported that ozonolyses of a large variety
of vinyl ethers (1) in the presence of a number of imines
(3a ) afford the corresponding monocyclic 1,2,4-dioxazo-
lidines (4a ) by [3 + 2]-cycloadditions of the carbonyl
oxides 2 derived from 1 and the CdN moieties in the
imines 3a .2,3 Competitive reactions showed that the
unsubstituted carbonyl oxide reacts faster with imines
3a to give 4a than with ketones 5 to give ozonides 6,
whereas disubstituted carbonyl oxides did not react with
imines.2,3
F igu r e 1. Graphical representation of 12c by ORTEP with
thermal ellipsoids at the 50% probability density level. Hy-
drogen atoms are drawn as open circles of arbitrary size.
CH2Cl2 afforded 13% of the stable compound 9a . By
contrast, ozonolyses of 7b and of 7c in the presence of 8
under the same conditions did not provide 9b and 9c,
respectively. The structural assignment of 9a is based
on the appearance of 13C NMR signals at δ 89 and 102.68
for the C-atoms in the heterocyclic ring and of 17O NMR
signals at δ 269 and 308 for the peroxide group. Reduc-
tion of 9a with triphenyl phosphine (TPP) gave cyclo-
hexanone and the O-methyl oxime of formaldehyde in a
molar ratio of ca. 1:1.
O-methyl oximes (3b) are also cleaved by ozone to give
carbonyl oxides 2.4 Recently, we have made use of this
by ozonizing a variety of O-methylated ketooximes (3b)
in the presence of a large number of ketones (5) to
prepare otherwise inaccessible tetrasubstituted ozonides
(6).5 In none of these reactions could we find evidence
for the formation of O-methoxy substituted 1,2,4-dioxa-
zolidines (4b), i.e., in the presence of ketones, disubsti-
tuted carbonyl oxides apparently do not undergo cycload-
ditions with disubstituted O-methyl oximes. In the
present work, we have examined whether carbonyl oxides
can undergo cycloadditions with O-methyl oximes in the
absence of added ketones.
On the basis of the above experience, monoozonolyses
of O-methylated dioximes (10) of dicarbonyl compounds
would not be expected to give the corresponding bicyclic
products 12 since such ozonolyses generate the mono- or
disubstituted carbonyl oxides 11. However, ozonolyses
of 10a -e did provide the corresponding bicyclic dioxazo-
lidines 12 in yields of 24-50%. As neat samples, 12a -c
were stable, whereas 12d ,e decomposed slowly at room
temperature. In CDCl3 or upon prolonged contact with
silica gel, 12a also underwent gradual transformation
into a product of unknown identity. Ozonolysis of 10f
gave a peroxidic crude product which decomposed with
detonation immediately after the solvent was distilled
off.
Resu lts
Ozonolysis of ethyl vinyl ether (7a ) in the presence of
one mol equiv of O-methyl cyclohexanoneoxime (8) in
(1) Institut fu¨r Anorganische Chemie, Universita¨t Karlsruhe (TH).
(2) Mori, M.; Nojima, M.; Kusabayashi, S.; McCullough, K. J . J .
Chem. Soc., Chem. Commun. 1988, 1550.
(3) McCullough, K. J .; Mori, M.; Tabuchi, T.; Yamakoshi, H.;
Kusabayashi, S.; Nojima, M. J . Chem. Soc., Perkin Trans. 1995, 41.
(4) Ito, Y.; Yokoya, H.; Umehara, Y.; Matsuura, T. Bull. Chem. Soc.
J pn. 1980, 53, 2407.
(5) Griesbaum, K.; Liu, X.; Kassiaris, A.; Scherer, M. Liebigs Ann.
1997, 1381.
S0022-3263(97)01469-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/03/1998

Dioxazolidines by Ozonolysis Reactions
J . Org. Chem., Vol. 63, No. 4, 1998 1087
to give diradical 17d , by analogy with the postulated
course of decomposition of compounds 4a .3 Further
reaction of diradical 17d may produce the oxaziridine
18d , whichsin line with previously postulated reactions
of oxaziridines4smay be converted into 15d and 16d . The
fact that the isomeric amides 15 and 16 with R1 ) C6H5
and R2 ) CH3 were not formed may be due to resonance
stabilization of the incipient ketone moiety by the phenyl
group.
In conclusion, the present investigation has revealed
that the known principle of cycloadditions of carbonyl
oxides (2) with imines (3) to give dioxazolidines (4)2,3 can
be extended to the cycloaddition of the unsubstituted
carbonyl oxide 7a , yet not of the substituted carbonyl
oxides 7b and 7c with the O-methylated oxime 8 to give
a N-methoxy-substituted dioxazolidine (9a ). This, to-
gether with the fact that no N-methoxydioxazolidines
(4b) had been obtained in ozonolysis reactions of O-
methylated oximes (3b) in the presence of ketones (5),4
indicates that, in contrast to alkyl or aryl imines,2,3
O-methylated oximes are poorer dipolarophiles than
ketones. Furthermore, ozonolyses of O-methylated di-
oximes (10) opened for the first time a one-step synthesis
for bicyclic N-methoxy-substituted 1,2,4-dioxazolidines.
Compounds 9a and 12a -f are the first examples of
N-methoxy substituted 1,2,4-dioxazolidines.
Exp er im en ta l Section
Gen er a l. 1H,5 13C,5 and 17O8 NMR spectra were recorded
on a Bruker AC 250, as reported previously. 15N NMR spectra
(CDCl3, NH4NO3 external) were collected with a Bruker DRX-
500. Chromatographic separations were done with flash
chromatography on silica gel.
The structural assignments of compounds 12a -e are
based on the appearance of 13C NMR signals in the range
of δ 95-105 for the carbon atoms in the heterocyclic
rings. The assignments of 12a -c were further supported
by 15N NMR signals in the range of δ 203-210 and by
17O NMR signals in the range of δ 304-311 for the
peroxide groups. The structure of 12c was additionally
proven by X-ray diffraction (Figure 1). Atoms C(3), H(3a),
H(3b), N(1), O(3), C(8), and H(8c) lie on a local mirror
plane that is not necessitated by space group require-
ments, while all other atoms have a mirror-related
counterpart in accordance with the NMR spectra; for
instance, C(1) is equivalent to C(5). This intrinsic
symmetry is further reflected by the observed pattern of
interatomic distances and angles (see Supporting Infor-
mation) and by the envelope conformation of the dioxa-
zolidine ring. Atom N(1) deviates by 63 pm from the
least-squares plane through the atoms C(1), O(1), O(2),
and C(5). With a length of 148.2(1) pm the peroxide bond
is at the upper limit of the expected range,6 which may
be due to the unfavorable ecliptic orientation of the
electron lone pairs at O(1) and O(2).7
Substrates 7a -c are commercially available. Substrates 8
and 10a -f were prepared according to a published procedure9
by reactions of the corresponding carbonyl compounds with
an excess of O-methyl hydroxylamine hydrochloride, and
isolated by flash chromatography. As shown by GLC analysis,
compounds 10a -c,e,f were obtained as mixtures of three
isomers, each, whereas only two isomers of 10d were obtained.
In each case, the isomer with the shorter elution time was the
predominant one and has been isolated and characterized: 10a
[1H NMR: δ 1.70 (m), 2.20 (q), 3.81 (s), 7.36 (t). 13C NMR: δ
23.77, 28.90, 61.54, 149.74.]; 10b [1H NMR: δ 1.70 (m), 1.83
(s), 2.21 (m), 3.81 (s), 3.82 (s), 7.36 (t). 13C NMR: δ 13.87,
23.50, 25.11, 35.16, 61.10, 61.54, 150.05, 156.20.]; 10c [1H
NMR: δ 1.73 (m), 1.82 (s), 2.18 (t), 3.82 (s). 13C NMR: δ 13.78,
23.78, 35.24, 60.95, 156.71.]; 10d [1H NMR: δ 1.78 (m), 2.22
(t), 2.75 (t), 3.82 (s), 3.96 (s), 7.30-7.40 (m), 7.60-7.70 (m).
13C NMR: δ 13.77, 23.19, 26.06, 35.78, 61.06, 61.80, 126.24,
128.40, 128.99, 135.68, 156.73, 157.95.]; 10e [1H NMR: δ 1.84
(s), 2.38 (s), 3.82 (s). 13C NMR: δ 13.98, 32.68, 61.12, 156.25;
identical with lit. data.10 ]; 10f [1H NMR: δ 1.80 (s), 3.01 (s),
3.86 (s). 13C NMR: δ 13.42, 42.30, 61.41, 153.76.].
Ozon olysis P r oced u r e. A solution of the respective
substrate(s) in pentane and/or in dichloromethane was treated
with ozone until the solution turned blue. Residual ozone was
flushed off with nitrogen, the solvent was distilled off at room
temperature and reduced pressure, and from the remaining
residue, the products were isolated by flash chromatography.
Red u ction Rea ction s. A solution of a dioxazolidine in 1
mL of CDCl3 was admixed with excess TPP in a NMR tube
and kept at room temperature until 1H NMR analysis showed
the disappearance of the substrate. The products were as-
Reductions of 12a -c and 12e with TPP gave the
corresponding monooximes 13, whereas reduction of 12d
afforded both of the two possible monooximes 13d and
14. Decomposition of the labile 12e in CDCl3 at room
temperature also produced monooxime 13e, but decom-
position of the labile 12d in CDCl3 gave the substituted
amides 15d and 16d in a ratio of 7:3. Their formation
may be explained by cleavage of the peroxide bond of 12d
(8) Hock, F.; Ball, V.; Dong, Y.; Gutsche, S.-H.; Hilss, M.; Schlind-
wein, K.; Griesbaum, K. J . Magn. Res. 1994, Ser. A 111, 150.
(9) Vogel, I.; Cresswell, W. T.; J effery, G. H.; Leicester, J . J . Chem.
Soc. 1952, 514.
(6) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen,
A. G.; Taylor, R. In International Tables for Crystallography; Kluwer
Academic Publishers: Dordrecht, 1992; Vol. C, Table 9.5.1.1, p 705.
(7) Henke, H.; Keul, H. Cryst. Struct. Commun. 1975, 4, 451.
(10) Shatzmiller, S.; Lidor, R. Synthesis 1983, 7, 590.

1088 J . Org. Chem., Vol. 63, No. 4, 1998
Griesbaum et al.
signed on the basis of comparison of their 1H and 13C NMR
data with those of authentic samples.
Red u ction of 12b. After 12 h, 12b had disappeared and
13b was shown to be the sole product of reduction by 1H NMR
[δ 1.77 (m), 2.13 (s), 2.17 (q), 2.48 (t), 3.80 (s)]5 and 13C NMR
analysis [δ 20.40, 28.62, 29.78, 42.33, 61.07, 207.67].
Ozon olysis of 10c. Ozonolysis of 0.74 g (4.0 mmol) of 10c
in 60 mL of pentane gave 0.73 g of a viscous residue, from
which 0.34 g (50%) of 12c was isolated [petroleum ether/ether,
9:1 (1000 mL), 1:1 (500 mL)].
Ozon olysis of 7a in th e P r esen ce of 8. A solution of 0.88
g (12.2 mmol) of 7a and 1.55 g (12.2 mmol) of 8 in 60 mL of
dichloromethane was ozonized at -78 °C. From the liquid
residue of 1.77 g, we isolated 0.27 g (13%) of 9a [solvent:
petroleum ether/ether, 19:1 (1000 mL), 1:1 (500 mL), acetone
(500 mL)].
N-Meth oxy-1,2-d ioxa -4-a za sp ir o[4.5]d eca n e (9a ): color-
less liquid. 1H NMR: δ 1.40-1.55 (m, 2 H), 1.55-1.75 (m, 6
H), 1.85-2.15 (m, 2 H), 3.62 (s, 3 H), AB-system δA 4.86, δB
4.95 (J ) 7.0 Hz, 2 H). 13C NMR: δ 23.29, 23.83, 25.15, 28.73,
35.15, 62.07, 89.00, 102.68. 17O NMR: δ 109 (s), 269 (s), 308
(s). Anal. Calcd for C8H15NO3 (173.2): C, 55.47; H, 8.73; N,
8.09. Found: C, 55.58; H, 8.60; N, 8.08.
1,5-Dim et h yl-N-m et h oxy-6,7,8-d ioxa zob icyclo[3.2.1]-
octa n e (12c): colorless solid, mp 36 °C. 1H NMR: δ 1.55 (s,
3 H), 1.40-1.63 (m, 1 H), 1.80-2.10 (m, 5 H), 3.66 (s, 3 H).
13C NMR: δ 17.23, 17.56, 37.66, 62.79, 102.67. 17O NMR: δ
91 (s), 311 (s). 15N NMR: δ 210 (s). Anal. Calcd for C8H15
-
NO3 (173.2): C, 55.47; H, 8.73; N, 8.09. Found: C, 55.44; H,
8.87; N, 8.07.
Reduction of 9a for 24 h gave cyclohexanone [1H NMR: δ
2.28-2.38 (m)] and the O-methylated oxime of formaldehyde11
[1H NMR: δ 3.89 (s, 3 H), AB-system δA 6.41, δB 7.00 (J ) 8.3
Hz, 2 H). 13C NMR: δ 61.71, 136.86.], as evidenced with the
help of authentic samples.
Ozon olysis of 7b in th e P r esen ce of 8. A solution of 0.90
g (6.7 mmol) of 7b and 0.85 g (6.7 mmol) of 8 in 60 mL of
dichloromethane was ozonized at -78 °C. From the liquid
residue of 1.93 g we isolated 0.44 g (49%) of 7b and 0.70 g
(82%) of 8 (solvents as above).
Ozon olysis of 7c in th e P r esen ce of 8. A solution of 0.60
g (7.0 mmol) of 7c and 0.89 g (7.0 mmol) of 8 in 60 mL of
dichloromethane was ozonized at -78 °C. The viscous residue
of 1.25 g was treated with pentane to give 0.30 g of a peroxidic
precipitate. From the pentane solution the solvent was
distilled off at room temperature and reduced pressure, and
from the liquid residue of 0.91 g, we isolated 0.72 g (81%) of 8
(solvents as above).
Ozon olysis of 10a . Ozonolysis of 0.67 g (4.2 mmol) of 10a
in 50 mL of pentane and 10 mL of dichloromethane gave 0.80
g of a viscous residue, from which 0.15 g (24%) of 12a and
0.04 g of an unassigned byproduct were isolated (petroleum
ether/ether, 6:1).
N-Meth oxy-6,7,8-dioxazo-bicyclo[3.2.1]octan e (12a): col-
orless liquid. 1H NMR: δ 1.30-1.60 (m, 1 H), 1.60-2.00 (m,
5 H), 3.64 (s, 3 H), 5.36 (d, J ) 4.6 Hz, 2 H). 13C NMR: δ
14.34, 30.82, 60.74, 95.45. 17O NMR: δ 117 (s), 304 (s). 15N
NMR: δ 203 (s). Anal. Calcd for C6H11NO3 (145.2): C, 49.65;
H, 7.64; N, 9.65. Found: C, 49.60; H, 7.67; N, 9.60.
Un a ssign ed byp r od u ct: colorless liquid. 1H NMR: δ 1.70
(m, 2 H), 2.02 (m, 2 H), 2.45 (m, 2 H), 3.80 (s, 3 H), 5.28 (t, J
) 5.4 Hz, 1 H). 13C NMR: δ 16.33, 30.85, 33.28, 62.60, 81.20,
168.73.
Red u ction of 12a . After 5 d, 12a had disappeared and
two isomers of 13a as well as the unassigned byproduct were
present in relative amounts of 54%, 18%, and 28%, respec-
tively, as evidenced by 1H NMR analysis with the help of
authentic samples [13a , isomer 1:5 δ 3.81 (OCH3), 7.35
(HCdN). 13a , isomer 2:5 δ 3.86 (OCH3), 6.63 (H-CdN).
Unassigned product: δ 3.80 (OCH3).].
Decom p osition of 12a in CDCl3. One drop of 12a in 1
mL of CDCl3 was kept at room temperature for 2 weeks. 1H
NMR analysis showed the presence of only the unassigned
byproduct.
Red u ction of 12c. After 15 d, 12c had disappeared and
two isomers of 13c were present as the sole products of
reduction as evidenced by 1H NMR [13c, isomer 1 (32%): δ
1.85 (s), 3.78 (s). 13c, isomer 2 (68%): δ 1.81 (s), 3.81 (s).]
and 13C NMR analysis [13c, isomer 1: δ 157.04 (CdN), 207.55
(CdO). 13c, isomer 2: δ 156.69 (CdN), 208.10 (CdO).] with
the help of authentic samples.5
Cr ysta l Str u ctu r e Deter m in a tion of 12c. Long pris-
matic needles are formed if a concentrated petroleum ether
solution of 12c is cooled to -20 °C. A single-crystal of suitable
size (0.9 × 0.5 × 0.4 mm3) was cut from such a needle. Crystal
data: C8H15NO3, fw ) 173.2 g mol-1, monoclinic, C2/c (no. 15),
a ) 1538.4(3) pm, b ) 1000.8(2) pm, c ) 1376.6(2) pm, â )
119.87(1)°, Z ) 8, Dcalcd ) 1.252 mg mm-3, T ) -90(2) °C.
Intensity data collection was performed with a Stoe four-
circle diffractometer equipped with a graphite monochromator
(Mo KR radiation) and a low-temperature attachment LT-1 of
Nicolet. Altogether 4374 reflections (h,k,(l were measured
up to the limit of 2θ ) 55° applying the method of learnt
profiles12 and ω/θ scans: basic width 34 steps, extra steps for
R1/R2 dispersion, counting times between 0.5 and 1.5 s per step
depending on intensity I/σ(I ). Corrections were made for
background rate, Lorentz and polarization factors, and the
intensity loss (<1.1%) of three standard reflections recorded
every 2 h.
The structure was solved by direct methods (SHELXS-86)
2
and refined on Fo data with anisotropic temperature factors
for C, N, and O (SHELXL-93). All hydrogen atoms appeared
in a difference Fourier synthesis. Their positions were refined
with individual isotropic temperature parameters giving the
final residuals R1 ) 0.031 for 1878 reflections, Fo > 4σ (Fo),
R1 ) 0.036 for all 2101 independent reflections, wR2 ) 0.084,
169 structure parameters. At this stage a difference density
map showed no feature above 0.28e/106 pm3 (see Supporting
Information).
Ozon olysis of 10d . Ozonolysis of 0.80 g (3.2 mmol) of 10d
in 40 mL of pentane and 20 mL of CH2Cl2 gave 0.82 g of a
viscous residue, from which 0.18 g (24%) of 12d was isolated
[petroleum ether/ether, 9:1 (1000 mL), 1:1 (500 mL)].
N-Met h oxy-1-m et h yl-5-p h en yl-6,7,8-d ioxa zob icyclo-
[3.2.1]octa n e (12d ): colorless liquid. 1H NMR: δ 1.60 (s, 3
H), 1.65 (m, 1 H), 1.90-2.50 (m, 5 H), 3.25 (s, 3 H), 7.20-7.50
(m, 3 H), 7.50-7.70 (m, 2 H). 13C NMR: δ 17.56, 17.65, 37.56,
38.65, 62.31, 103.58, 105.10, 126.06, 127.88, 127.94, 136.66.
Red u ction of 12d . After 14 h, 12d had disappeared. 1H
NMR analysis showed the presence of 13d [17%; δ 2.10 (s,
CH3CdO)], 14 [26%; δ 1.85 (s, CH3CdN)], 15d [36%; δ 3.16
(s, CH3N)], and 16d [21%; δ 2.14 [s, CH3CdO)].
Ozon olysis of 10b. Ozonolysis of 0.45 g (2.6 mmol) of 10b
in 60 mL of pentane gave 0.47 g of a viscous residue, from
which 0.12 g (28%) of 12b was isolated [petroleum ether/ether,
17:3 (1000 mL), 1:1 (500 mL)].
Decom p osition of 12d . A solution of one drop of 12d in 1
mL of CDCl3 was heated to 40 °C in a sealed NMR tube. 1H
NMR analysis after 30 h showed the presence of 15d (69%)
and 16d (31%).
Ozon olysis of 10e. Ozonolysis of 0.71 g (4.1 mmol) of 10e
in 60 mL of pentane gave 0.65 g of a viscous residue, from
which 0.22 g (34%) of 12e was isolated (petroleum ether/ether,
9:1).
N -Me t h o x y -1-m e t h y l-6,7,8-d io x a zo b ic y c lo [3.2.1]-
octa n e (12b): colorless liquid. 1H NMR: δ 1.35-1.75 (m, 1
H), 1.56 (s, 3 H), 1.80-2.10 (m, 5 H), 3.66 (s, 3 H), 5.33 (d, J
) 4.6 Hz, 1 H). 13C NMR: δ 16.09, 16.66, 31.27, 37.18, 61.78,
97.36, 101.16. 17O NMR: δ 106, 308. 15N NMR: δ 207. Anal.
Calcd for C7H13NO3 (159.2): C, 52.82; H, 8.23; N, 8.80.
Found: C, 52.60; H, 8.27; N, 8.71.
(11) J ensen, K. A.; Buus, L.; Hohn, A. Acta Chem. Scand. Ser. B,
1977, 31, 28.
(12) Clegg, W. Acta Crystallogr. 1981, A37, 22.

Dioxazolidines by Ozonolysis Reactions
J . Org. Chem., Vol. 63, No. 4, 1998 1089
N-Met h oxy-1,4-d im et h yl-2,3,7-d ioxa zob icyclo[2.2.1]-
h ep ta n e (12e): colorless liquid. 1H NMR: δ 1.53 (s, 6 H),
1.80-2.05 (m, 4 H), 3.69 (s, 3 H). 13C NMR: δ 12.99, 30.80,
62.89, 101.68.
solvent was not completely removed to leave 7 mL of a
concentrated solution. When this was placed on a chromato-
graphic column, a spontaneous, exothermic decomposition
occurred.
Red u ction of 12e. After 1 h, 12e had disappeared and
13e was shown to be the sole product of reduction by 1H NMR
[δ 1.82 (s), 2.19 (s), 2.45 (t, J ) 6.9 Hz), 2.68 (t, J ) 7.3 Hz),
3.80 (s)] and 13C NMR analysis (δ 14.55, 29.90, 30.01, 39.17,
61.12, 155.51, 207.43) with the help of an authentic sample.5
Decom p osition of 12e. One drop of 12e in 1 mL of CDCl3
was kept at room temperature for 7 h to give 13e as the sole
Su p p or tin g In for m a tion Ava ila ble: Tables of the frac-
tional crystallographic coordinates of 12c, thermal parameters,
and bond lengths and angles (4 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
1
product as shown by H and 13C NMR analysis.
Ozon olysis of 10f. Ozonolysis of 0.60 g (3.8 mmol) of 10f
in 60 mL of pentane gave a residue which decomposed with
detonation at room temperature. In a second experiment, the
J O971469O