Tandem cycloaddition chemistry of nitroalkenes: Preparative and theoretical studies on the stereochemical course of [3 + 2] cycloaddition of cyclic nitronates

Article abstract of DOI:10.1021/jo9818374

Intermolecular [3 + 2] cycloadditions between two cyclic nitronates and a series of dipolarophiles are examined. High facial selectivity is observed in all cases and is analyzed with the aid of ab initio transition structure calculations. Monosubstituted dipolarophiles reacted with exclusive regiocontrol. Disubstituted dipolarophiles reacted with varying degrees of regiocontrol, which was dependent on the substituent. A theoretical approach for predicting regioselectivity is discussed. Exo selectivity was generally favored due to steric effects, and was especially high with cis-disubstituted dipolarophiles.

Full text of DOI:10.1021/jo9818374

884
J . Org. Chem. 1999, 64, 884-901
Ta n d em Cycloa d d ition Ch em istr y of Nitr oa lk en es: P r ep a r a tive
a n d Th eor etica l Stu d ies on th e Ster eoch em ica l Cou r se of [3 + 2]
Cycloa d d ition of Cyclic Nitr on a tes
Scott E. Denmark,* Mark Seierstad, and B. Herbert
Roger Adams Laboratory, Department of Chemistry, University of Illinois, Urbana, Illinois 61801
Received September 9, 1998
Intermolecular [3 + 2] cycloadditions between two cyclic nitronates and a series of dipolarophiles
are examined. High facial selectivity is observed in all cases and is analyzed with the aid of ab
initio transition structure calculations. Monosubstituted dipolarophiles reacted with exclusive
regiocontrol. Disubstituted dipolarophiles reacted with varying degrees of regiocontrol, which was
dependent on the substituent. A theoretical approach for predicting regioselectivity is discussed.
Exo selectivity was generally favored due to steric effects, and was especially high with
cis-disubstituted dipolarophiles.
In tr od u ction
The tandem [4 + 2]/[3 + 2] cycloaddition of nitroalk-
enes has emerged as a powerful method for the rapid and
stereoselective construction of complex polyheterocyclic
systems.1 A critical strategic feature that imparts great
versatility and stereocontrol to this sequence is the
[3 + 2] cycloaddition of the first-formed nitronate (Figure
1). Of the four fundamental permutations of this tandem
sequence, the two most studied are those which termi-
nate in an intramolecular [3 + 2] process.2-5 In these
cases, the regio- and stereochemical outcome of the
cycloaddition is primarily determined by the constraints
imposed by the tether.
Of equal synthetic interest are those tandem processes
that terminate in an intermolecular [3 + 2] cycloaddi-
tion.6-9 In these cases the stereochemical course of the
reaction is less predictable. Thus, to expatiate our
knowledge of the factors controlling selectivity in these
cycloadditions, we set out to systematically explore
[3 + 2] cycloadditions of representative cyclic nitronates
formed from a nitroalkene [4 + 2] cycloaddition with a
variety of dipolarophiles, to examine the effects of dipo-
larophile substitution on regio-, stereo-, and facial selec-
tivity.
F igu r e 1. Family of tandem [4 + 2]/[3 + 2] cycloadditions (A
) electron acceptor, D ) electron donor).
to compilations of experimental results, rationalizations
from theoretical studies have also been forwarded.13 Both
nitrones and nitrile oxides display very similar behavior,
so only the case of nitrones will be discussed here.
The most common method of rationalizing regioselec-
tivity is frontier molecular orbital (FMO) theory.14 The
frontier molecular orbitals of a simple nitrone are shown
below (Figure 2).15 The largest coefficient of the nitrone
HOMO is on the oxygen, while the largest coefficient of
the nitrone LUMO is at the carbon. Comparison of these
frontier orbitals with those of common monosubstituted
alkenes shows that the interaction with the smaller
energy gap is HOMOdipolarophile-LUMOnitrone, which leads
to 5-substituted isoxazolidines (a ).16 In the case of alkenes
with strongly electron-withdrawing groups, the LUMO-
dipolarophile-HOMOnitrone interaction becomes more signifi-
cant, and one would expect to see the other regioisomer
(b).
Ba ck gr ou n d
Regioselectivity. The regiochemical issues in cy-
cloadditions of the structurally similar nitrones10,11 and
nitrile oxides12 have been extensively studied. In addition
(1) Denmark, S. E.; Thorarensen, A. Chem. Rev. 1996, 96, 137.
(2) Denmark, S. E.; Middleton, D. S. J . Org. Chem. 1998, 63, 1604.
(3) Denmark, S. E.; Marcin, L. R. J . Org. Chem. 1997, 62, 1675.
(4) Denmark, S. E.; Dixon, J . A. J . Org. Chem. 1997, 62, 7086.
(5) Denmark, S. E.; Guagnano, V.; Dixon, J . A.; Stolle, A. J . Org.
Chem. 1997, 62, 4610.
(6) Avalos, M.; Babiano, R.; Cintas, P.; Higes, F. J .; J ime´nez, J . L.;
Palacios, J . C.; Silva, M. A. J . Org. Chem. 1996, 61, 1880.
(7) Liu, J .; Niwayama, S.; You, Y.; Houk, K. N. J . Org. Chem. 1998,
63, 1064.
(8) Denmark, S. E.; Thorarensen, A. J . Org. Chem. 1994, 59, 5672.
(9) Denmark, S. E.; Herbert, B. J . Am. Chem. Soc. 1998, 120, 7357.
(10) Tufariello, J . J . In 1,3-Dipolar Cycloaddition Chemistry; Padwa,
A., Ed.; J ohn Wiley & Sons: New York, 1984; Vol. 2, pp 83-168.
(11) Confalone, P. N.; Huie, E. M. In Organic Reactions; Kende, A.
S., Ed.; J ohn Wiley & Sons: New York, 1988; Vol. 36, pp 1-173.
(12) Caramella, P.; Gru¨nanger, P. In 1,3-Dipolar Cycloaddition
Chemistry; Padwa, A., Ed.; J ohn Wiley & Sons: New York, 1984; Vol.
1, pp 291-392.
(13) Houk, K. N.; Yamaguchi, K. In 1,3-Dipolar Cycloaddition
Chemistry; Padwa, A., Ed.; J ohn Wiley & Sons: New York, 1984; Vol.
2, pp 407-450.
(14) Fleming, I. Frontier Orbitals and Organic Chemical Reactions;
J ohn Wiley & Sons: New York, 1978.
(15) Houk, K. N.; Sims, J .; Duke, R. E., J r.; Strozier, R. W.; George,
J . K. J . Am. Chem. Soc. 1973, 95, 7287.
(16) Houk, K. N.; Sims, J .; Watts, C. R.; Luskus, L. J . J . Am. Chem.
Soc. 1973, 95, 7301.
10.1021/jo9818374 CCC: $18.00 © 1999 American Chemical Society
Published on Web 01/20/1999

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 885
tions were acrylonitrile, which gave an equal mixture of
stereoisomers, and dimethyl maleate, which formed only
the endo adduct. In all cases, the authors explained the
stereoselectivity as the effect of secondary orbital interac-
tions; INDO calculations of the frontier orbitals have been
used to support the rationale.
Studies from our laboratories have shown that an exo
orientation is generally preferred.1 In a complete reversal
of the precedent with acyclic nitronates, dimethyl male-
ate gave a single product resulting from exo approach to
nitronates 1,30 2,31 and 38 (Scheme 1). A fundamental
difference in these systems (compared to those of Carrie´)
is that there is no electron-withdrawing group on the
nitronate carbon, which affects the relative size of the
coefficients in the frontier orbitals, but not their signs;
any secondary orbital effect in the Carrie´ acyclic system
should therefore also be present in these cyclic systems.
F igu r e 2. Matching coefficients of nitrones and dipolaro-
philes.
Sch em e 1
In general, cycloadditions of nitrones with monosub-
stituted alkenes are consistent with this analysis.13
Highly electron-deficient dipolarophiles are more likely
to produce 4-substituted isoxazolidines (a ). Nitrones
substituted at carbon have a higher energy HOMO than
those without. With these nitrones, the LUMOdipolarophile
-
HOMOnitrone interaction becomes even more significant,
and an increase in the amount of isomer b is generally
observed. Examinations of disubstituted alkenes have
shown that the addition of alkyl groups can markedly
alter the regiochemical outcome.17
The frontier orbitals of nitronates with electron-
withdrawing groups attached to carbon have been cal-
culated and show the same pattern of coefficients as
nitrones and nitrile oxides.18-20 A different sort of ar-
rangement is observed with calculations on simple alkyl-
substituted nitronates.21 With these dipoles, the larger
coefficient is on carbon in both the HOMO and the
LUMO. When matched with electron-deficient dipolaro-
To better understand the nature of [3 + 2] cycloaddi-
tions of cyclic nitronates, a systematic study of dipolaro-
phile structure was undertaken. Nitronates 2 and 3 were
chosen for their ease of preparation and use; it was also
expected that the cycloadditions would be highly facially
selective, which would halve the number of possible
cycloadducts and ease identification.
philes, the smaller energy gap is the LUMOdipolarophile
-
Exp er im en ta l Resu lts
HOMOnitronate. FMO theory predicts 5-isoxazolidines would
always predominate; this has been the case with both
electron-deficient22-25 and electron-rich6 monosubstituted
dipolarophiles. With the exception of several intramo-
lecular cycloadditions1 and cycloadditions with sym-
metrical olefins,8,26 [3+2] cycloadditions between cyclic
nitronates and 1,2-disubstituted alkenes are rare.27,28
Ster eoselectivity. The stereochemical course of cy-
cloadditions of acyclic nitronates to mono-20 and di-
activated26,29 alkenes has been extensively studied by
Carrie´. In most cases, exclusive exo (with respect to the
electron-withdrawing group) selectivity was seen; excep-
P r ep a r a tion of 1,3-Dip oles. Nitronate 2 was pre-
pared in good yield and as a single diastereomer (d.r. >
25/1) by Ti(O-i-Pr)2Cl2-promoted cycloaddition of ni-
troalkene 78 with n-butyl vinyl ether (Scheme 2). This
Lewis acid generally promotes highly endo selective
cycloadditions.1 The crystalline nitronate 2 was easily
handled and stable indefinitely, but only when stored
rigorously free from air. Nitronate 3 was prepared as
previously reported from these laboratories.8
Sch em e 2
(17) Kouklovsky, C.; Dirat, O.; Berranger, T.; Langlois, Y.; Tran-
Huu-Dau, M. E.; Riche, C. J . Org. Chem. 1998, 63, 5123.
(18) Gre´e, R.; Tonnard, F.; Carrie´, R. Bull. Soc. Chim. Fr. 1975,
1325.
(19) Gre´e, R.; Tonnard, F.; Carrie´, R. Tetrahedron 1976, 32, 675.
(20) Gre´e, R.; Carrie´, R. J . Heterocycl. Chem. 1977, 14, 965.
(21) Schnute, M. E. Ph.D. Thesis, University of Illinois, 1995.
(22) Breuer, E.; Aurich, H. G.; Nielsen, A. Nitrones, Nitronates and
Nitroxides; J ohn Wiley & Sons: Chichester, 1989.
(23) Tartakovskii, V. A.; Chlenov, I. E.; Ioffe, S. L.; Lagodzinskaya,
G. V.; Novikov, S. S. J . Org. Chem. USSR (Engl. Trans.) 1966, 2, 1573;
Zh. Org. Khim. 1966, 2, 1593.
P r ep a r a tion of Mon osu bstitu ted Dip ola r op h iles.
The aldehyde precursors for several vinyl ketones were
prepared as outlined in Scheme 3. Aldehydes 11 and 12
were prepared in 2 steps from 1,4-butenediol 8 by
hydroxyl protection and ozonolysis of the alkene. The
addition of vinylmagnesium bromide to aldehyde 11
(24) Barco, A.; Benetti, S.; Pollini, G. P.; Spalluto, G.; Zanirato, V.
Tetrahedron Lett. 1991, 32, 2517.
(25) Marti, R. E.; Heinzer, J .; Seebach, D. Liebigs Ann. 1995, 1193.
(26) (a) Gre´e, R.; Carrie´, R. Tetrahedron 1976, 32, 683. (b) Kane-
masa, S.; Yoshimiya, T.; Wade, E. Tetrahedron Lett. 1998, 39, 8869.
(27) Brook, M. A.; Seebach, D. Can. J . Chem. 1987, 65, 836.
(28) Torssell, K.; Zeuthen, O. Acta Chem. Scand. B 1978, 32, 118.
(29) Gre´e, R.; Carrie´, R. Bull. Soc. Chim. Fr. 1975, 1319.
(30) Denmark, S. E.; Hurd, A. R. J . Org. Chem. 1998, 63, 3045.
(31) Guagnano, V. M.S. Thesis, University of Illinois, 1997.

886 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
product (18%). The use of 1 molar equivalent of quinoline
(relative to substrate) failed to suppress over-reduction.
The use of either 5% palladium on calcium carbonate
poisoned with lead (Lindlar catalyst)39 or palladium on
barium sulfate44 offered no advantages; the reduction
times were significantly extended (11 h vs 3 h) and
resulted in lower (2-4/1) Z/E selectivities. Fortunately,
the components were separable by MPLC but were
routinely carried into the next reaction as a mixture.
Ultimately, we found that the propargyl alcohol 23 could
be selectively reduced with dicyclohexylborane (followed
by protonolysis with acetic acid) to afford (Z)-25 in 71%
yield as a single isomer. Conversion of 25 and 26 to 27
and 28 proceeded in high yield and without isomerization
under the Swern oxidation conditions.
provided the allyl alcohol 13 in 79% yield. Oxidation of
the alcohol using the Swern procedure32 resulted in the
clean conversion of the alcohol into the ketone. The yields
of the process were widely variable due to the tendency
of the product to polymerize during distillation, even at
10-5 mmHg. We found no deleterious effect on the
subsequent [3 + 2] cycloaddition by using the ketone
immediately after initial chromatographic purification
and found the yields to fall within the 66-75% range.
(Benzyloxy)methyl vinyl ketone (16) was prepared by
Swern oxidation of the known alcohol 1433 in 68% yield
or by using the Dess-Martin reagent34,35 (87%).
Sch em e 3a
Sch em e 4
a
Key: (a) R ) TDS: TDSCl imidazole, DMF (97%); R ) Bn:
NaH, Bu4N+I-, BnCl, THF, 66 °C, (99%).
P r ep a r a tion of Disu bstitu ted Dip ola r op h iles. A
series of disubstituted acrylate derivatives was chosen
for study (Chart 1). (E)-2-Trimethylsilylpropenoate 17a 36
was prepared from oxidation and esterification of the
known allyl alcohol.37 (Z)-2-Substituted propenoates 17b36
and 1938 were synthesized by Lindlar39 hydrogenation of
the corresponding alkynes. A separable mixture of 20a
and 20b was prepared from O-acetylation of sodium ethyl
formyl acetate.40 The analogous methyl esters could be
synthesized by a similar procedure, but in much lower
yields.
[3 + 2] Cycloa d d ition s w ith Mon osu bstitu ted
Dip ola r op h iles. The results of the thermal cycloaddition
reaction between nitronates 2 or 3 and monosubstituted
dipolarophiles are collected in Table 1. The reactions were
facile; in general, the nitronates were converted into the
corresponding nitroso acetals in the presence of 5-6
equiv of the dipolarophile in benzene at room tempera-
ture. The cycloaddition was performed with 1.1-1.3
equivalents of the dipolarophile at a higher substrate
concentration (0.1 M vs 0.03 M) when 15 and 16 were
used in order to conserve material. The [3 + 2] cycload-
ditions occurred at ambient temperature within a rea-
sonable time span (1.5-12 h). Elevated reaction temper-
atures were required for the rapid and complete con-
sumption of nitronate 3 in the [3 + 2] cycloaddition with
allyl alcohol.
Ch a r t 1
Each of the nitroso acetals 29-36 were formed in high
yield (79-91%) as a pair of diastereomers. The effect of
(33) Trost, B. M.; Nu¨bling, C. Carbohydrate Res. 1990, 202, 1.
(34) Meyer, S. D.; Schreiber, S. L. J . Org. Chem. 1994, 59, 7549.
(35) The yield with the Swern reagent is for the separate preparation
and characterization of 16. The yield with the Dess-Martin reagent
is for the oxidation followed immediately by [3 + 2] cycloaddition. See
Experimental Section for details.
(36) Cunico, R. F.; Lee, H. M. J . Am. Chem. Soc. 1977, 99, 7613.
(37) J ones, T. K.; Denmark, S. E. Org. Synth. Coll. Vol. VII 1990,
524.
(38) Taschner, M. J .; Rosen, T.; Heathcock, C. H. Org. Synth. Coll.
Vol. VII 1990, 226.
(39) Lindlar, H.; Dubuis, R. Org. Synth. Coll. Vol. V 1973, 880.
(40) Freeman, P. K.; Stevenson, B. K.; Balls, D. M.; J ones, D. H. J .
Org. Chem. 1974, 39, 546.
Two (Z)-2-disubstituted enones were also examined.
The â-silylvinyl ketone 28 was prepared from (benzyl-
oxy)acetaldehyde (12) by the sequence depicted in Scheme
4. Metalation of silyl acetylene 2241 followed by the
addition of aldehyde 12 provided the propargyl alcohol
24 in 85% yield. Catayltic reduction of the alkynes 23
and 2442,43 was plagued with variability and lack of
reproducibility. For example, hydrogenation of alkyne 24
over 5% palladium on calcium carbonate in the presence
of quinoline (6 mol %) provided the allylic alcohols (Z)-
26 (42%) and (E)-26 (7%) in addition to the saturated
(41) Fleming, I.; Takaki, K.; Thomas, A. P. J . Chem. Soc., Perkin
Trans. 1 1987, 2269.
(42) Siegel, S. In Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 8; pp 417-442.
(43) Marvell, E. N.; Li, T. Synthesis 1973, 457.
(32) Mancuso, A. J .; Brownfain, D. S.; Swern, D. J . Org. Chem. 1979,
44, 4148.
(44) Bulman-Page, P. C.; Ley, S. V. J . Chem. Soc., Perkin Trans. 1
1984, 1847.

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 887
Ta ble 1. [3 + 2] Cycloa d d ition s w ith Mon osu bstitu ted Dip ola r op h iles
HC(2) ppm
C(2) ppm
nitronate
R2
CO2Me
CO2Me
CO2-t-Bu
COMe
COCH2OTDS
COCH2OBn
CHO
temp, °C
time, h
yield, %
nitroso acetal
a
b
a
b
d.r.a, a /b
2
3
3
3
3
3
3
3
rt
rt
rt
rt
rt
rt
rt
80
6
6
2.5
1.5
4
12
3.5
1.5
79
86
91
88
90
89
c
29
30
31
32
33
34
35
36
5.09
5.02
4.88
4.92
5.15
5.06
4.88
4.74
4.84
4.78
4.66
4.61
4.76
4.74
4.62
4.66
80.5
80.6
81.4
87.1
85.1
85.5
80.8
80.8
81.8
88.0
86.5
86.6
7.6/1
6.5/1
11.4/1
6.2/1b
5/1b
5.7/1
1/2
1.3/1
c
CH2OH
81
84.1
86.8
a
b
Product ratios determined by isolation. Determined by 1H NMR analysis of the mixture. c Converted to 36 by reduction with NaBH4
in 62% overall yield. Nitroso acetals 35 characterized only by 1H NMR analysis.
the vinyl ether substituent (OR1) on the diastereoselec-
tivity was minimal (cf., Table 1, entries 1 and 2).
Separation of the diastereomers was made easier using
the chiral auxiliary; therefore, the cycloaddition study
with monosubstituted dipolarophiles was conducted with
nitronate 3. The cycloadditions were found to be under
kinetic control. Nitroso acetal 30b was quantitatively
recovered after being stirred with tert-butyl acrylate in
benzene; no trace of 31a or 31b was detected. The rate
and selectivity of the cycloaddition were not subject to
significant solvent effects. Comparable yields for the
formation of nitroso acetals 29 were found in benzene
and acetonitrile, but the diastereomeric ratio was slightly
lower (5.5/1) in the latter. In all cases, the products were
assigned a configuration reflecting approach of the di-
polarophile to the face opposite the C(4) benzoate. This
assignment was proven unambiguously in the case of 30
(vide infra) and is consistent with previous studies8 and
X-ray crystal structure determinations.45
each of the exo-derived nitroso acetals was downfield by
0.12 ppm to 0.39 ppm relative to the endo-derived isomer.
Examination of the 13C NMR data revealed that the C(2)
resonance for 29b-36b was downfield relative to the
other isomer (0.2 ppm to 2.7 ppm).
P r oof of F a cia l Selectivity. In the [3 + 2] cycload-
dition there is an issue of regioselectivity in which the
carbon bearing the activating group of the dipolarophile
is attached to either the oxygen atom (head-to-head) or
to the carbon atom (head-to-tail) of the dipole. There are
also two issues of stereoselectivity. The first involves the
approach of the dipolarophile to either face of the dipole;
the second is the orientation of the activating group of
the dipolarophile (endo or exo) with respect to the dipole.
Overall this creates a set of eight possible isomers from
each nitronate. 2-D NMR analysis was consistent with
the head-to-head connectivity for all isomers, thereby
reducing the number of possible isomers to four. To
identify the origin of stereoisomerism, we sought a simple
method that would enable us to distinguish among the
four possibilities. That two diastereomers were formed
in the cycloaddition of both nitronates 2 and 3 indicated
that the stereochemical complexity was due to either
endo/exo isomerism or to facial indiscrimination. Since
two new stereocenters are created in the cycloaddition,
removal of one would result in the retention of sufficient
stereochemical information to eliminate one of the pos-
sible pathways. Independent hydrogenolysis of the ni-
troso acetals followed by removal of the hydroxy group
would provide a pair of deoxylactams. If the stereochem-
ical duplicity were due to facial indiscrimination, the
remaining stereogenic center would reflect that and the
lactams 38a and 38b would differ in the configuration
at the ring junction. The deoxylactams would be identical
if the dipolarophile approached from the same face; the
stereoisomerism would then arise from endo/exo path-
ways.
In accord with previous trends in the intermolecular
[3 + 2] cycloaddition of nitronates,8,18-20,46,47 the major
nitroso acetals (29a -36a ) arose from an exo orientation
of the dipolarophile controlling group (R2). The sole
exception was seen with acrolein; endo approach to
provide nitroso acetal 35b was slightly favored (an X-ray
crystal structure of 36b45 confirmed the assignment). A
noticeable correlation between the size of the dipolaro-
phile substituent and the diastereomeric ratio of nitroso
acetals was found. For example, the cycloaddition of
nitronate 3 with tert-butyl acrylate resulted in an 11.4/1
ratio of diastereomeric products. With methyl acrylate,
methyl vinyl ketone, allyl alcohol, and acrolein, the
diastereomeric ratios gradually decreased from 7/1 to 1/2.
A compilation of diagnostic NMR data for each of the
diastereomeric pairs of nitroso acetals is listed in Table
1. In all cases, the 2-D NMR analyses were consistent
with the attachment of the dipolarophile controlling
1
group (R2) at C(2). The H NMR resonance for HC(2) for
Deoxygenation of the lactam was accomplished using
the two-step procedure reported previously8 (Scheme 5).
The alcohol 37a 45 was first acylated with phenoxy chlo-
rothionocarbonate. Tributyltin hydride and AIBN were
then slowly added to a benzene solution of the thiono-
carbonate at 80 °C to furnish the deoxygenated lactam
38a in 58% yield. In a similar manner, the minor nitroso
acetal 37b was transformed to the deoxygenated lactam
(45) The authors have deposited atomic coordinates for the X-ray
crystal structures of 36b (no. 113363), 37a (no. 113364), 40bi (no.
113365), 42a ii (no. 113366), and 44bi (no. 113367) with the Cambridge
Crystallographic Data Centre. The data can be obtained, upon request,
from the Director, Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, U.K.
(46) Denmark, S. E.; Schnute, M. E. J . Org. Chem. 1994, 59, 4576.
(47) Gre´e, R.; Carrie´, R. J . Am. Chem. Soc. 1977, 99, 6667.

888 J . Org. Chem., Vol. 64, No. 3, 1999
Ta ble 2. [3+2] Cycloa d d ition s w ith Disu bstitu ted Acr yla tes
Denmark et al.
HC(2) ppm
no.
R1
R2
R3
solvent
temp, °C time, h
yield, % nitroso acetal
i
ii
d.r.a, a i/a ii/bi/bii
Me
H
H
H
SiMe3
H
H
OAc
benzene
CH3CN
CH3CN
CH3CN
CH2Cl2
CH3CN
rt
65
65
65
70b
65
rt
6
38
22
22
3
79
72
77
75
54
78
82
29
39
40
41
42
43
44
7.6/c/1/c
2.4/1/c/c
1/1.5/2.1/c
1/1.6/c/c
c/1/c/c
17b
17a
19
20b
20a
21
Me SiMe3
Me
Et
Et
Et
5.13
4.50
H
4.64, 4.60 4.16
5.07
Me
OAc
H
5.07
6.89
24
23
4.94, 4.80 6.99, 6.79
d
Me
H
CO2Me benzene
1/-/1/-
a
b
Product ratios determined by isolation or 1H NMR analysis of mixtures. Concentration ∼6 M in 20b; solution did not reflux. c Not
observed. Regioselectivity 1/2.7 i/ii by 1H NMR analysis of reaction mixture; stereoselectivity could not be determined.
d
Sch em e 5a
than benzene for cycloadditions that required heating.
These reactions usually gave lower yields of nitroso
acetals and a small amount of several decomposition
products, owing to the thermal instability of the nitr-
onate. Dimethyl fumarate (21) reacted rapidly at room
temperature. In contrast to nitroso acetals 29-36, cy-
cloadducts derived from disubstituted acrylates were
more easily separated when nitronate 2 was used.
In cycloadditions with nitronate 2, disubstituted dipo-
larophiles showed the high degree of diastereofacial
selectivity previously seen with monosubstituted dipo-
larophiles. The structure of three of these nitroso acetals
(40bi, 42a ii, and 44bi) was secured by X-ray crystal-
lography.45 In all cases, the structures revealed that the
dipolarophile had approached exclusively to the â-face
of the nitronate; the only relevant issues are regio- and
stereocontrol. Nitroso acetals i arise from approach of the
acrylate ester proximal to the nitronate oxygen (head-
to-head); nitroso acetals ii are the opposite regioisomers.
As before, nitroso acetals a and b derive from an exo and
endo (respectively) approach of the ester to the nitronate.
a
Reagents: (a) H2 (160 psi), Ra-Ni, MeOH; (b) PhOCSCl,
pyridine, THF; (c) Bu3SnH, AIBN, C6H6, 80 °C.
38b. Comparison of the melting points and H and 13C
NMR spectra revealed that lactams 38a and 38b were
identical.
1
Hydrogenolysis of nitroso acetals 29a -31a provided
the same R-hydroxy lactam (37a ). This outcome is
consistent with a similar stereostructure of the cycload-
ducts, arising from the approach of the dipolarophile to
the same face of the nitronate with the same orientation.
The 1H and 13C NMR data for C(2) also support the
assignment of 29a -36a as arising from the same facial
approach and orientation. A similar analysis of the NMR
data for the nitroso acetals 29b-36b suggests that all
of these isomers are of the same stereochemical sense.
We therefore conclude that in all cases, the diastereomers
from the [3 + 2] cycloaddition with monosubstituted
dipolarophiles with nitronates 2 and 3 are the result of
endo/exo isomerism with complete facial control.
[3 + 2] Cycloa d d ition s w ith Disu bstitu ted Dip o-
la r op h iles. With the exception of methyl crotonate (18),
all of the acrylate derivatives reacted smoothly with
nitronate 2 (Table 2). All attempts with 18 resulted only
in decomposition of the nitronate and no product forma-
tion. Unlike the monosubstituted dipolarophiles (vide
supra), disubstituted acrylates were unreactive at room
temperature; CH3CN and CH2Cl2 proved better solvents
In contrast to the earlier examples, disubstituted
acrylates reacted with poorer regiocontrol. With a silicon-
based â-substituent (nitroso acetals 39 and 40) the major
products still were those with the ester on C(2) of the
isoxazoline, but with one-third of the product being the
other regioisomer. Replacing a silicon-based substituent
with carbon-based substituent (nitroso acetals 41) re-
sulted in a reversal in regiocontrol, now favoring isomer
ii. The acetoxy acrylates were even more selective for the
head-to-tail pathway, with cis-acrylate 20b providing
nitroso acetal 42a ii as a single cycloadduct.
The cis-disubstituted dipolarophiles exhibited greatly
enhanced exo selectivity compared to the monosubsti-
tuted dipolarophiles. In all cases, only products arising
from exo approach could be detected. The trans-dipolaro-
philes were much less selective. The major product from
the cycloaddition of silane 17a was 40bi (45% of product),
the structure of which was determined by X-ray crystal-
lography,45 and which resulted from endo approach of the
ester (hence exo approach of the bulky silane). The other

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 889
Ta ble 3. [3+2] Cycloa d d ition s w ith Alk oxym eth yl Vin yl
Sch em e 6
Keton es
Ch a r t 2
HC(3) ppm C(3) ppm
time, yield, nitroso
d.r.,
no.
R
h
%
acetal
i
ii
i
ii
i/ii
To devise a basis for predicting the regiochemical
outcome of cycloadditions of nitronates of type 2 and 3
with disubstituted dipolarophiles (vide infra), we exam-
ined the dipolar cycloaddition of nitronate 2 with 2,3-
dihydrofuran (47, Scheme 6).
27 TDS
28 Bn
12
5
97
98
45a
46
2.30
32.0
26/1b
2.31 3.81 32.1 52.8 32/1c
a
b
Nitronate (+)-3 used in this case. Determined by 1H NMR
analysis of the mixture. c Determined by 1H NMR analysis and
isolation.
Nitronate 2 and 47 underwent ready cycloaddition to
provide two products (48a and 48b) in a ratio of 17/1 (as
head-to-head product, 40a i (22%), arose from a less
favorable endo approach of the silane. The sole head-to-
tail product isolated was determined by NOE analysis
(see Supporting Information) to be 40a ii, which came
from endo approach of the TMS group. The stereoselec-
tivity in the cycloaddition of acetoxy acrylate 20a could
not be determined, owing to the inseparability of the four
products. Dimethyl fumarate 21 displayed no stereose-
lectivity.
1
determined by H NMR analysis of the mixture). Sur-
prisingly, the minor isomer was not the head-to-tail
regioisomer, but rather derived from endo approach of
the dipolarophile, as assigned by NOE studies on both
diastereomers (see Supporting Information).
Th eor etica l Resu lts
Although different dipolarophiles combine with nitr-
onates 2 and 3 with varying levels of regio- and stereo-
selectivity, one constant factor is the high degree of facial
selectivity. This has previously been attributed to steric
repulsion from the C(4) benzoyloxy substituent.8 How-
ever, even nitronates with no substituent at that atom
(such as 1) have shown significant facial selectivity (6/1
The results of the thermal cycloaddition between
nitronate 3 and â-silylvinyl ketones 27 and 28 are listed
in Table 3. To conserve material, only 1.2 equiv of the
dipolarophile were used at higher substrate concentration
(0.1 M). The [3 + 2] cycloadditions occurred at ambient
temperature within five to 12 h. The reactions were
typically performed immediately following the Swern
oxidation of the precursor alcohols 25 and 26 and were
allowed to proceed overnight for convenience. Both ni-
troso acetals 45 and 46 were formed in high yield as a
pair of regioisomers.
to 8/1) in cycloadditions with various dipolarophiles.30
A
theoretical study using three model 1,3-dipoles (Chart
2) was undertaken to help understand the origin of facial
bias. Both nitrone 4948 and nitronate 5049 are known
compounds. Because they are achiral, experiments can-
not be used to test facial biases; theoretical studies can,
however, examine individual chiral conformations. Ni-
tronate 51 is a simplified model for nitronate 1.
Transition structures for cycloadditions of half-chair
conformations of 49-51 with ethylene were fully opti-
mized at the RHF and B3LYP levels (Figures 3-6). In
all cases, approach of ethylene distal to the C(6) axial
group (which is proximal to the C(6) methoxy group in
Figure 6) produced a transition structure wherein the six-
membered ring adopted a chairlike conformation. Ap-
proach from the opposite face forced the six-membered
ring to twist into a boatlike arrangement. Other boatlike
transition structures (which originated from approach of
ethylene to a half-boat conformation of the 1,3-dipole)
were observed but not studied systematically.
The cycloaddition between nitronate 3 and ketone 28
provided the corresponding nitroso acetal 46 in 98% yield
as a 32/1 ratio of regioisomers; cycloaddition with 27 was
only slightly less selective. The 1H NMR resonance of C(2)
of 46ii was shifted upfield by 0.36 ppm relative to 46i.
That difference was consistent with the range seen in
nitroso acetals 39 and 40, but was also within the range
noted in the cycloaddition studies on simple, monosub-
stituted dipolarophiles where it was determined that the
minor isomers arose from an endo orientation of the
dipolarophile. Comparison of the H and 13C NMR shifts
at C(3) was instrumental in deciphering the identity of
the isomers. The H NMR resonance of C(3) in the major
1
1
isomer was found 1.5 ppm upfield from that of the minor
suggesting that the silicon atom was directly attached
there. In the 13C NMR spectrum, C(3) of the major nitroso
acetal was located at 32.1 ppm while that of the minor
isomer was located at 52.8 ppm. The positional differ-
ences clearly indicate that an electron-withdrawing group
(carbonyl) was attached at C(3) of the minor isomer.
Comparison with data for similar (Z)-â-silyl-substituted
dipolarophile cycloadditions was also consistent with the
identity of the minor isomer as a regioisomer.
Absolute and relative energies for the calculated
transition structures are listed in Table 4. There was
little difference in geometry or energy between the HF/
6-31G* and HF/6-31+G* optimized structures; however,
(48) Tufariello, J . J .; Gatrone, R. C. Tetrahedron Lett. 1978, 2753.
(49) Chlenov, I. E.; Morozova, N. S.; Khudak, V. I.; Tartakovskii,
V. A. Bull. Acad. Sci. USSR, Div. Chem. Sci. (Engl. Trans.) 1970, 2492;
Izv. Akad. Nauk. SSSR, Ser. Khim. 1970, 2641.

890 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
the experimentally determined facial bias of nitronate
1.30 For nitronate 51, approach distal to the C(6) methoxy
group is preferred when the methoxy group is in an axial
orientation (Figure 5), while proximal approach is pre-
ferred when the methoxy group is equatorially disposed
(Figure 6).
Table 5 contains selected bond lengths in the calculated
transition structures of ethylene and the axial methoxy
conformation of nitronate 51. The structures denoting
chairlike and boatlike arrangements are very similar; one
important difference is that a boatlike arrangement
brings about slightly shorter (0.02-0.07 Å) atom dis-
tances representing the forming bonds.
F igu r e 3. Chairlike (a) and boatlike (b) transition structures
of nitrone 49 with ethylene.
Discu ssion
As with other cycloadditions, the key stereochemical
issues are facial selectivity (with respect to the axial C(6)
substituent), regioselectivity (head-to-head vs head-to-
tail approach), and stereoselectivity (exo vs endo orienta-
tion) (Figure 7).
F igu r e 4. Chairlike (a) and boatlike (b) transition structures
of nitronate 50 with ethylene.
F igu r e 5. Chairlike (a) and boatlike (b) transition structures
of nitronate 51 with ethylene (axial MeO-).
F igu r e 7. Selectivity issues for [3 + 2] cycloadditions of cyclic
nitronates.
F a cia l Selectivity. The enhanced facial selectivity
exhibited in nitronates 2 and 3 compared to nitronate 1
is a consequence of the C(4) benzoyloxy substituent.
Because nitronate 1 does react with modest selectivity,30
it is clear that the benzoyloxy group is not solely
responsible for the facial bias, but that it does augment
it.
The theoretical studies show that the conformation of
the nitronate is an important determinant for facial
selectivity. When ethylene approaches from the face
distal to the axial C(6) substituent (which is the only
facial approach observed in experiments with 2 and 3),
the six-membered ring may fold directly into a chair as
the reaction progresses from transition state to product.
Proximal approach would require the six-membered ring
to either flip to the other chair conformation, or fold
directly into a boat. The result is that proximal approach
is disfavored.
F igu r e 6. Chairlike (a) and boatlike (b) transition structures
of nitronate 51 with ethylene (equatorial MeO-).
the HF/3-21G optimized structures displayed a facial
preference significantly lower in magnitude.
For any given conformation of 1,3-dipole, approach to
afford a boatlike transition structure was about 3-5 kcal/
mol higher in energy than the corresponding chairlike
arrangement. The only exception was the B3LYP/3-21G
calculated structures with 51. The conformer in which
the C(6) methoxy group is axial favors distal approach
by only 0.9 kcal/mol. Although singular among the
calculated transition structures, this result best matches
A possible electronic source of facial bias is the kinetic
anomeric effect (KAE), a stereoelectronic stabilization
well-documented in cycloadditions50 and nucleophilic
(50) Vasella, A. Helv. Chim. Acta 1977, 60, 1273.

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 891
Ta ble 4. Ca lcu la ted En er gies (a u ), Rela tive En er gies (k ca l/m ol, in P a r en th eses) for Tr a n sition Str u ctu r es of
Cycloa d d ition s of Eth ylen e a n d 1,3-Dip oles
RHF
B3LYP
structure
49 chairlike
boatlike
50 chairlike
3-21G
6-31G*
6-31+G*
3-21G
6-31G*
-399.56896 (0.0)
-399.56333 (3.5)
-435.12446 (0.0)
-435.11876 (3.6)
-548.39020 (0.0)
-548.38522 (3.1)
-548.38306 (0.0)
-548.37722 (3.7)
-401.78868 (0.0)
-401.78184 (4.3)
-437.55037 (0.0)
-437.54222 (5.1)
-551.44195 (0.0)
-551.43454 (4.6)
-551.43861 (0.0)
-551.43008 (5.4)
-401.79864 (0.0)
-401.79166 (4.4)
-437.56281 (0.0)
-437.55432 (5.3)
-551.45681 (0.0)
-551.44910 (4.8)
-551.45442 (0.0)
a
-402.22103 (0.0)
-402.21623 (3.0)
-437.86951 (0.0)
-437.86493 (2.9)
-551.77298 (0.0)
-551.77153 (0.9)
-551.76618 (0.0)
-551.76155 (2.9)
-404.42659 (0.0)
-404.42064 (3.7)
-440.28109 (0.0)
-440.27460 (4.1)
-554.81020 (0.0)
-554.80574 (2.8)
-554.80653 (0.0)
a
boatlike
51 (ax. MeO-) chairlike
boatlike
51 (eq. MeO-) chairlike
boatlike
a
This structure would not optimize with this model.
Ta ble 5. Ca lcu la ted Atom Dista n ces (Å) for Tr a n sition
Str u ctu r es of Cycloa d d ition s of Eth ylen e a n d Nitr on a te
51 (Axia l Meth oxy)
RHF
B3LYP
bond
3-21G 6-31G* 6-31+G* 3-21G 6-31G*
Chairlike
ethylene CdC
forming C-O
forming C-C
N-O
1.36
2.10
2.31
1.34
1.30
1.37
2.07
2.18
1.26
1.32
1.38
2.07
2.17
1.26
1.32
1.36
2.22
2.34
1.30
1.33
1.38
2.21
2.17
1.25
1.35
F igu r e 9. Approach of 17a and 17b to 2.
pair inhabits a pseudoaxial position (this is analogous
to the conformational constraints imposed on tetrahy-
dopyrans by the “classic” anomeric effect).
N-C
Boatlike
1.38
2.04
2.16
1.26
ethylene CdC
forming C-O
forming C-C
N-O
1.36
2.06
2.28
1.34
1.31
1.38
2.04
2.15
1.26
1.32
1.37
2.20
2.27
1.30
1.33
1.38
2.18
2.13
1.25
1.35
Regioselectivity. With both cis- and trans-dipolaro-
philes, a general qualitative pattern of regioselectivity
is apparent. With regard to the â-substituent, head-to-
tail regioselectivity is favored by O > C > Si > H. This
pattern must, of course, be the result of a combination
of steric and electronic effects. Because both the HOMO
and the LUMO of nitronates such as 2 have the highest
coefficient on carbon, a switch in regioselectivity cannot
properly be explained by FMO theory; as a result, any
electronic contribution needs another rationale.
N-C
1.32
The replacement of hydrogen (methyl acrylate) with
silicon (17a ,b) is accompanied by a modest amount of
head-to-tail product (about 30% with each, see Table 2).
In the case of 17a , this can easily be explained with a
steric argument (Figure 9). Exo, head-to-head approach
of the dipolarophile to 2 places the TMS group in close
contact with the cyclic core, which destabilizes the
transition state relative to methyl acrylate. If a steric
effect were the dominant regiochemical controller, one
would expect the head-to-head approach of 17b to be less
destabilized relative to that of 17a ; in fact, both dipo-
larophiles display similar regioselectivity. Enones 27 and
28, which contain a bulkier â-silicon group, are actually
much more head-to-head selective than the â-silyl acry-
lates (Table 3). The dominance of steric effects would
require the enones to be at least as head-to-tail selective
as 17a ,b. The regiochemical preference of isocrotonate
19 is also not as easily explained by steric effects. By any
metric a methyl group is smaller than a trimethylsilyl
group, so a steric rationale would predict less head-to-
tail product; this is not the case, as the major product in
the [3 + 2] cycloaddition with 19 is the head-to-tail
adduct (41a ii, see Table 2). Overall, a rationale of
overriding steric effects is inconsistent with the regio-
chemical trends.
F igu r e 8. Dipolarophile approach to various nitronate con-
formations.
additions51 to nitrones. The KAE is defined in these cases
as an interaction in the transition state between the
incipient sp3 orbital on nitrogen and an adjacent electron-
deficient bond. Vasella50 has used this effect to rationalize
the high stereoselectivity in the cycloadditions of nitrones
with olefins; the favored transition structure displays an
anti-periplanar orientation of the forming lone pair on
nitrogen and the C-O bond of a sugar auxiliary.
A similar effect has been used to explain high stereo-
selectivity in [3 + 2] cycloadditions of certain nitronates
with activated dipolarophiles.26a,52,53 In these cases, the
favorable interaction in the transition structure is be-
tween the forming sp3 orbital on nitrogen and one (or
both) of the newly formed anti-periplanar bonds to
ethylene (Figure 8). The calculated transition structures
show that a nearly anti-periplanar orientation is possible
regardless of the face of approach (Figure 8 a,b); again,
however, this orientation is much more easily satisfied
with a chairlike conformation, wherein the nitrogen lone-
The regioselectivities seen with disubstituted dipolaro-
philes must be due more to an electronic perturbation.
Because FMO theory failed as a general predictive tool,
another method was sought that could also be easily
applied, but which was consistent with experiment. The
(51) Petrzilka, M.; Felix, D.; Eschenmoser, A. Helv. Chim. Acta 1973,
56, 2950.
(52) Denmark, S. E.; Dappen, M. S.; Cramer, C. J . J . Am. Chem.
Soc. 1986, 108, 1306.
(53) Denmark, S. E.; Cramer, C. J .; Sternberg, J . A. Helv. Chim.
Acta 1986, 69, 1971.

892 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
Ta ble 6. HF /6-31G* Atom ic Ch a r ge Ca lcu la tion s
electro- electro-
static
static
atomic atomic
nitroso charge, charge,
(C(R)+1)/
dipolarophile acetal
C(R)
C(â)
i/ii
log(i/ii) (C(â)+1), P
47
48
46
39
40
41
42
43
0.08 -0.38 i only
-0.22 -0.42 32
-0.42 -0.33 2.4
-0.51 -0.28 2.1
1.74
28a
17b
17a
19
1.51
0.38
0.32
-0.23
-0.43
1.34
0.87
0.68
0.32
0.24
0.18
F igu r e 10. Relationship of product ratio to polarity index.
-0.64
-0.68
-0.76
0.12 0.59
0.31 0.37
0.36 ii only
20a
20b
other 1,3-dipoles); either electron-rich or electron-defi-
cient dipolarophiles can be used with equal efficiency.
However, to fully apply this modified allopolarization
principle would require dividing the list of charges as
they pass through zero; this would provide a list of
quotients which was unnaturally parabolic. To remedy
this, all of the charges were increased by 1 before division.
Although this is an arbitrary transformation, there is in
fact good agreement with this “polarity index” and the
log of the product ratio (Figure 10); the plot is linear with
r2 ) 0.97.
To test the ability of this method to predict the
regiochemical outcome of cycloadditions with nitronates
of type 2, 2,3-dihydrofuran (47) was chosen as a test
substrate. In the acrylate series, replacing a â-hydrogen
with a methyl group dramatically altered the regioselec-
tivity (see 41, Table 2). It was known6 that monosubsti-
tuted vinyl ethers react with high head-to-head selectiv-
ity; the analysis presented in Table 6 predicts that
replacing a â-hydrogen with a methylene would not alter
the regioselectivity. The experiment was consistent with
this prediction; nitroso acetals 48 were formed in good
yield, and no trace of the head-to-tail regioisomer could
be detected.
a
A smaller version of this dipolarophile was used. See Sup-
porting Information for details.
standard arsenal of substituent parameters calculated
by Taft54-56 and others57 did not lend itself to a simple
application in this case.
In an effort to formulate a general and quantifiable
theory to explain regioselectivity with ambident anions
(e.g., enolates and allyl anions), Gompper58 has intro-
duced the concept of allopolarization. For a given ambi-
dent anion, the relative charge density at each end is
calculated. From this is derived the “polarity index” (P)
which is the ratio of the relative charge densities; this is
then related linearly to the log of the product ratio. The
foundation of the allopolarization principle is that a
change in polarity of the ambident anion will affect a
change in its reactivity with an electrophile.
A similar analysis was made with the present system.
Table 6 lists the electrostatic atomic charges calculated
(with the HF/6-31G* model) for several of the disubsti-
tuted dipolarophiles; C(R) is the carbon adjacent to the
electronic “controlling” group (i.e., carboalkoxy, ketone,
or ether group). As before, nitroso acetal i arises from
head-to-head approach of the electronic controlling group,
while nitroso acetal ii derives from a head-to-tail ap-
proach.
It was found that the calculated atomic charges at C(R)
were an excellent qualitative predictor of regioselectivity;
they assign all of the disubstituted dipolarophiles into
the correct order, from most head-to-head selective to
most head-to-tail selective. Atomic charges at C(â) were
also mostly consistent with this ordering. What is most
relevant, of course, is the relationship between the charge
at both ends of the dipolarophile. The allopolarization
principle applied to ambident anions takes this into
account by dividing the atomic charges; the sign of the
charge at either end is always the same. With these
dipolarophiles, the sign of the atomic charge can change
as the substituents change. This is one of the strengths
of dipolar cycloadditions with nitronates (and also many
This method is not too far removed from the theory
underlying standard electronic substituent parameters.55,57
The primary difference here is that the effects of the
individual substituents are calculated not separately, but
taken together as a collection of field and resonance
effects within an entire molecule.
Ster eoselectivity. Exo approach of monosubstituted
dipolarophiles is generally favored and is correlated to
the size of the substituent (Table 1). The modest endo
preference observed in the case of acrolein (the smallest
of the carbonyl substituted alkenes) suggests that there
is a stabilizing electronic interaction associated with an
endo approach. However, this stabilizing interaction is
easily overpowered by the steric bulk of the dipolarophile.
The stereoselectivity observed in cycloadditions of cis-
1,2-disubstituted acrylate and enone derivatives to cyclic
nitronates is very high and is always exo. This is in
contrast to the report of Carrie´26a that dimethyl maleate
adds to acyclic nitronates with exclusive endo orientation.
Carrie´ ascribes his results to secondary orbital interac-
tions that are favorable in some cases, and unfavorable
in others; the results with cyclic nitronates cannot be
explained in a similar fashion. Because of the failure of
FMO theory to explain the regiochemical results in these
systems, it is difficult to justify an FMO approach to
(54) Taft, R. W.; Topsom, R. D. Prog. Phys. Org. Chem. 1987, 16, 1.
(55) Topsom, R. D. Prog. Phys. Org. Chem. 1987, 16, 125.
(56) Hehre, W. J .; Pau, C.-F.; Headley, A. D.; Taft, R. W.; Topsom,
R. D. J . Am. Chem. Soc. 1986, 108, 1711.
(57) Exner, O.; Ingr, M.; Ca´rsky, P. J . Mol. Struct. (THEOCHEM)
1997, 397, 231.
(58) Gompper, R.; Wagner, H.-U. Angew. Chem., Int. Ed. Eng. 1976,
15, 321.

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 893
4.65 mmol, 3.0 equiv) in CH2Cl2 (16 mL) was added Ti(O-i-
Pr)4 (1.38 mL, 4.68 mmol, 3.0 equiv). The clear solution was
stirred at room temperature for 30 min. To a solution of 7
(304.9 mg, 1.58 mmol) in CH2Cl2 (4 mL) was added butyl vinyl
ether (415 µL, 3.21 mmol, 2.0 equiv). The solution was cooled
to -82 °C, and the solution of “Ti(O-i-Pr)2Cl2” (6.0 equiv) was
added to it over 10 min. The resulting solution was allowed to
warm over 20 min to -76 °C and then was stirred for an
additional 3 h and 15 min. The reaction was quenched at -76
°C with 1 M NaOH in methanol (15 mL) and then was poured
into CH2Cl2 (200 mL), washed with water (3 × 75 mL), and
back extracted with CH2Cl2 (2 × 50 mL). The organic layer
was dried (Na2SO4), filtered through a pad of Celite, and
concentrated in vacuo. The residue was purified by silica gel
chromatography (hexane/EtOAc, 85/15, 80/20, 65/35, 50/50)
and recrystallized (EtOAc/hexane) to afford 323.6 mg (70%)
of 2 as a crystalline white solid. Data for 2: mp 74-75 °C
describing the stereochemical results. The outcome with
cis-dipolarophiles can, however, easily be explained using
steric arguments. Since methyl acrylate is moderately exo
selective, replacing the cis-â-hydrogen with a larger group
should enhance exo selectivity; this is exactly what is
observed.
Replacing the trans-â-hydrogen with a larger substitu-
ent tends to decrease the exo selectivity, which again
suggests that electronic effects alone do not determine
stereoselectivity. In fact, when the trans-â-substituent
is sterically demanding and has little electronic signifi-
cance (silane 17a ) the major approach is with the ester
group endo. When the â-substituent is approximately the
same size as the ester (acrylate 20a and fumarate 21),
essentially no stereoselectivity is observed.
There does appear to be an electronic contributor to
stereoselectivity in the case of 2,3-dihydrofuran (47).
Unlike the cis-disubstituted enones and acrylates, 47
reacted with nitronate 2 to give 48b as a minor (6%)
product, which arises from endo approach of the dipo-
larophile. There apparently exists a small stabilizing
effect when an electron-rich dipolarophile is oriented
endo.
1
(EtOAc/hexane); H NMR (400 MHz, CDCl3) δ 8.05 (dd, J )
8.4, 1.3, 2H), 7.60 (tt, J ) 7.4, 1.5, 1H), 7.46 (tt, J ) 8.1, 7.2,
2H), 6.62 (d, J ) 4.1, 1H), 5.71 (ddd, J ) 6.3, 4.1, 2.1, 1H),
5.49 (t, J ) 2.3, 1H), 4.02 (dt, J ) 9.4, 6.5, 1H), 3.63 (dt, J )
9.2, 6.3, 1H), 2.44 (ddd, J ) 15.3, 6.4, 3.2, 1H), 2.35 (dt, J )
15.3, 2.1, 1H), 1.69-1.62 (m, 2H), 1.48-1.39 (m, 2H), 0.93 (t,
J ) 7.3, 3H); 13C NMR (100 MHz, CDCl3) δ 165.5, 133.4, 129.6,
129.2, 128.3, 109.5, 101.2, 69.5, 61.3, 31.4, 29.5, 19.0, 13.7; IR
(CHCl3) 3018 (m), 1719 (s), 1625 (s) cm-1; MS (CI, CH4) m/z
294 (M+ + 1, 9), 57 (100); TLC Rf 0.20 (hexane/EtOAc, 65/35).
Anal. Calcd for C15H19NO5 (293.32): C, 61.42; H, 6.53; N, 4.78.
Found: C, 61.50; H, 6.56; N, 4.57.
Con clu sion s
Intermolecular [3 + 2] cycloadditions between two
disubstituted cyclic nitronates (2 and 3) and several
dipolarophiles were studied. Remarkably high facial
selectivity was observed; this was attributed to a com-
bination of steric shielding from the nitronate C(4)
substituent and an inherent facial bias from the chiral
conformation of the nitronate. Monosubstituted dipolaro-
philes provided cycloadducts with exclusive head-to-head
regioselectivity. Regioselectivity with disubstituted di-
polarophiles varied from low to excellent (>25/1). With
regard to the â-substituent, head-to-tail regioselectivity
was favored by O > C > Si > H and was qualitatively
consistent with an analysis based on electrostatic atomic
charge calculations. Stereoselectivity was found to be
mostly dependent on the steric nature of the dipolaro-
phile; an electronic contribution weakly favored an endo
approach, but was generally overruled by a more favor-
able exo approach of dipolarophiles bearing bulky sub-
stituents.
(Z)-1,4-Bis-(d im eth yl(1,1,2-tr im eth ylp r op yl)silyloxy)-
2-bu ten e (9). Diol 8 (3.52 g, 40.0 mmol) and imidazole (12.25
g, 180 mmol, 4.5 equiv) were dissolved in DMF (20 mL), and
the mixture was cooled to 5 °C, whereupon thexyldimethylsilyl
chloride (16.52 mL, 84 mmol, 2.1 equiv) was added. The
mixture was allowed to warm to room temperature and stir
for 18 h. The reaction mixture was diluted with tert-butyl
methyl ether (TBME) (200 mL) and was washed with sat. aq
NH4Cl solution (3 × 50 mL) and brine (25 mL). The aqueous
layers were back extracted with TBME (50 mL). The combined
organic layers were dried (MgSO4) and concentrated in vacuo.
The crude product was purified by silica gel chromatography
(hexane/EtOAc, 97.5/2.5) and distillation to afford 14.43 g
(97%) of 9 as a clear liquid. Data for 9: bp 180 °C (0.1 mmHg);
1H NMR (400 MHz, CDCl3) δ 5.53 (t, J ) 4.0, 2H), 4.20 (d, J
) 4.2, 4H), 1.62 (septet, J ) 7.0, 2H), 0.88 (d, J ) 6.8, 12H),
0.84 (s, 12H), 0.10 (s, 12H); 13C NMR (100 MHz, CDCl3) δ
130.1, 59.3, 34.1, 25.1, 20.3, 18.5, -3.3; IR (CCl4) 2950 (s), 2872
(s) cm-1; MS (CI, CH4) m/z 373 (M+ + 1, 26), 129 (100); TLC
Rf 0.74 (hexane/EtOAc, 97/3). Anal. Calcd for C20H44O2Si2
(372.74): C, 64.45; H, 11.90. Found: C, 64.42; H, 11.95.
2-[(Dim et h yl(1,1,2-t r im et h ylp r op yl)silyl)oxy]et h a n a l
(11). To a 250 mL, three-necked round-bottom flask equipped
with a thermometer and gas inlet were added 9 (5.0 g, 13.41
mmol) and CH2Cl2 (75 mL). After the solution was cooled to
-78 °C, ozone was bubbled through the solution until a blue
coloration appeared. The solution was then purged with
oxygen. Zinc (1.32 g, 20.12 mmol, 1.5 equiv) was added followed
by 50% aq acetic acid solution (35 mL). The mixture was
allowed to slowly warm to 0 °C. After 30 min the mixture was
allowed to warm to room temperature and then was stirred
for 1 h. The mixture was diluted with CH2Cl2 (200 mL) and
was washed with water (3 × 50 mL). The aq layers were back
extracted with CH2Cl2 (50 mL). The combined organic layers
were washed with sat. aq NaHCO3 solution (50 mL) and brine
(50 mL), dried (Na2SO4), and concentrated in vacuo. The crude
product was purified by silica gel chromatography (hexane/
EtOAc, 91/9) and distillation to afford 4.40 g (81%) of 11 as a
colorless oil. Data for 11: bp 100 °C (0.2 mmHg); 1H NMR
(400 MHz, CDCl3) δ 9.70 (t, J ) 1.0, 1H), 4.19 (d, J ) 0.7,
2H), 1.65 (septet, J ) 6.8, 1H), 0.89 (d, J ) 6.8, 6H), 0.88 (s,
6H), 0.13 (s, 6H); 13C NMR (100 MHz, CDCl3) δ 202.2, 69.3,
Exp er im en ta l Section
Gen er a l Exp er im en ta l. See Supporting Information.
Gen er a l Com p u ta tion a l. Ground-state atomic charge
calculations were performed on an Apple Power Macintosh
9500 with MacSpartan.59 Transition structure calculations
were performed on a Silicon Graphics Origin200 with Gauss-
ian94.60 See Supporting Information for coordinates. All
transition structures were fully optimized with the indicated
model and were characterized by harmonic frequency calcula-
tions and had one negative vibrational frequency.
r el-(4R,6R)-4-(Ben zoyloxy)-6-(b u t yloxy)-5,6-d ih yd r o-
4H-[1,2]oxa zin e N-Oxid e (2). To a solution of TiCl4 (510 µL,
(59) MacSpartan Version 1.1. Wavefunction, Inc. 18401 Von Karman
Avenue, Suite 370. Irvine, CA 92612.
(60) Gaussian 94, Revision B.3; Frisch, M. J .; Trucks, G. W.;
Schlegel, H. B.; Gill, P. M. W.; J ohnson, B. G.; Robb, M. A.; Cheeseman,
J . R.; Keith, T.; Petersson, G. A.; Montgomery, J . A.; Raghavachari,
K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.;
Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.;
Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.;
Defrees, D. J .; Baker, J .; Stewart, J . P.; Head-Gordon, M.; Gonzalez,
C.; Pople, J . A. Gaussian, Inc.: Pittsburgh, PA, 1995.
34.0, 25.2, 20.1, 18.4, -3.6; IR (CCl4) 2900 (s), 1742 (s) cm-1
MS (CI, CH4) m/z 203 (M+ + 1, 6), 117 (100); TLC Rf 0.56
;

894 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
(hexane/EtOAc,91/9). Anal. Calcd for C10H22O2Si (202.37): C,
59.35; H, 10.96. Found: C, 59.34; H, 10.94.
1-[(Dim e t h yl(1,1,2-t r im e t h ylp r op yl)silyl)oxy]-4-(d i-
m eth ylp h en ylsilyl)-3-bu tyn -2-ol (23). A solution of ethyl
bromide (2.30 mL, 30.8 mmol, 1.25 equiv) in Et2O (12 mL) was
added dropwise to a suspension of magnesium turnings (737
mg, 30.3 mmol, 1.23 equiv) in Et2O (13 mL) at a rate such as
to maintain a gentle reflux. Upon complete addition, the
reaction mixture was heated at reflux for 1.5 h and then was
allowed to cool to room temperature. A solution of alkyne 22
(4.95 g, 30.9 mmol, 1.25 equiv) in THF (13 mL) was added
dropwise over 15 min, and the reaction mixture was heated
at reflux for 1 h. Upon cooling to room temperature, a solution
of the aldehyde 11 (5.00 g, 24.7 mmol) in THF (13 mL) was
added dropwise over 15 min. The reaction mixture was then
heated at reflux for 1 h and allowed to cool to room temper-
ature, and the reaction was quenched with sat. aq NH4Cl
solution. The reaction mixture was poured into water (300 mL)
and extracted with pentane (3 × 100 mL). The organic phase
was washed with water (100 mL) and brine (100 mL) and dried
(Na2SO4). The crude material was purified by chromatography
on silica (hexane/EtOAc, 94/6) and distilled (bulb-to-bulb) to
provide 7.71 g (86%) of 23 as a colorless liquid. Data for 23:
1-[(Dim e t h yl-1-(1,1,2-t r im e t h ylp r op yl)silyl)oxy]-3-
bu ten e-2-ol (13). A solution of aldehyde 11 (5.01 g, 24.8 mmol)
in THF (25 mL) was added dropwise over 15 min to the THF
solution of vinylmagnesium bromide (1.0 M, 32 mL, 32 mmol,
1.3 equiv) at 0 °C. Upon complete addition, the reaction
mixture was heated at reflux for 1 h and then cooled to 0 °C,
and the reaction was quenched with sat. aq NH4Cl solution
(10 mL). The contents of the reaction were poured into water
(70 mL) and extracted with Et2O (2 × 50 mL). The combined
organic phases were washed with water (50 mL) and brine
(50 mL) and dried (MgSO4). The crude material was purified
by chromatography on silica (hexane/EtOAc, 91/9) and distilled
(bulb-to-bulb) to provide 4.53 g (79%) of 13 as a colorless liquid.
Data for 13: bp 150 °C (0.2 mmHg); 1H NMR (400 MHz,
CDCl3) δ 5.81 (ddd, J ) 17.3, 10.5, 5.6, 1H), 5.34 (dt, J ) 17.1,
1.7, 1H), 5.18 (dt, J ) 10.5, 1.7, 1H), 4.19-4.13 (m, 1H), 3.64
(dd, J ) 10, 3.7, 1H), 3.43 (dd, J ) 9.8, 7.6, 1H), 2.55 (bs, 1H),
1.62 (septet, J ) 6.8, 1H), 0.88 (d, J ) 6.8, 6H), 0.85 (s, 6H),
0.11 (s, 6H); 13C NMR (100 MHz, CDCl3) δ 136.7, 116.3, 72.9,
66.7, 34.1, 25.1, 20.2, 18.4, -3.6; IR (neat) 2959 (s) cm-1; MS
(CI, CH4) m/z 231 (M+ + 1, 4), 129 (100); TLC Rf 0.46 (hexane/
EtOAc, 80/20). Anal. Calcd for C12H26O2Si (230.43): C, 62.55;
H, 11.37. Found: C, 62.28; H, 11.48.
1
bp 175 °C (8.5 × 10-5 mmHg); H NMR (500 MHz, CDCl3) δ
7.63-7.35 (m, 5H), 4.42 (ddd, J ) 6.2, 5.5, 3.8, 1H), 3.77 (dd,
J ) 10.1, 3.8, 1H), 3.66 (dd, J ) 10.1, 6.4, 1H), 2.60 (d, J )
5.5, 1H), 1.61 (septet, J ) 6.8, 1H), 0.88 (d, J ) 6.8, 6H), 0.86
(s, 6H), 0.42 (s, 6H), 0.13 (s, 3H), 0.13 (s, 3H); 13C NMR (125
MHz, CDCl3) δ 136.5, 133.7, 129.4, 127.8, 105.4, 88.0, 66.5,
63.5, 34.1, 25.1, 20.3, 18.5, -1.0, -3.5; IR (neat) 3479 (bw),
2959 (s) cm-1; MS (CI, CH4) m/z 364 (M+ + 1, 1), 261 (100);
TLC Rf 0.28 (hexane/EtOAc, 90/10). Anal. Calcd for C20H34O2-
Si2 (362.66): C, 66.24; H, 9.45; Si, 15.49. Found: C, 66.07; H,
9.41; Si, 15.74.
1-[(Dim eth yl(1,1,2-tr im eth ylp r op yl)silyl)oxy]-3-bu ten -
2-on e (15). A solution of dimethyl sulfoxide (350 µL, 4.93
mmol, 2.3 equiv) in CH2Cl2 (17 mL) was cooled to -74 °C and
then treated with oxalyl chloride (210 µL, 2.41 mmol, 1.1 equiv)
in one portion. After 45 min, a solution of alcohol 13 (500 mg,
2.17 mmol) in CH2Cl2 (5 mL) was added dropwise via cannula.
The resultant cloudy mixture was maintained for 1 h before
triethylamine (1.5 mL, 11 mmol, 5 equiv) was introduced
dropwise. The reaction mixture was maintained at -74 °C for
10 min and then allowed to warm to room temperature over
15 min. The reaction mixture was diluted with Et2O (100 mL)
and washed with 0.1 M hydrochloric acid (35 mL), water (35
mL), and brine (35 mL). The combined aq layers were back
extracted with Et2O (50 mL), and the organic layers were dried
(Na2SO4). The residue was purified by rapid chromatography
on neutral alumina (III) (hexane/EtOAc, 91/9) to provide the
ketone 15 as a faint gold liquid. The ketone was further
purified by distillation (bulb-to-bulb) to provide 325 mg (66%)
of 15 as a colorless liquid that rapidly polymerized. Data for
15: bp 165 °C (0.15 mmHg); 1H NMR (500 MHz, CDCl3) δ 6.67
(dd, J ) 17.6, 10.8, 1H), 6.35 (dd, J ) 17.6, 1.6, 1H), 5.78 (dd,
J ) 10.6, 1.5, 1H), 4.34 (s, 2H), 1.65 (septet, J ) 6.8, 1H), 0.90
(d, J ) 7.0, 6H), 0.88 (s, 6H), 0.13 (s, 6H); TLC Rf 0.37 (hexane/
EtOAc, 80/20).
4-(Dim eth ylp h en ylsilyl)-1-(p h en ylm eth oxy)-3-bu tyn -2-
ol (24). A solution of ethyl bromide (1.85 mL, 24.8 mmol, 1.24
equiv) in Et2O (12 mL) was added dropwise to a suspension of
magnesium turnings (598 mg, 24.6 mmol, 1.23 equiv) in Et2O
(12 mL) at a rate such as to maintain a gentle reflux. Upon
complete addition, the reaction mixture was heated at reflux
for 1.5 h and then allowed to cool to room temperature. A
solution of alkyne 22 (4.00 g, 25.0 mmol, 1.25 equiv) in THF
(12 mL) was added dropwise over 15 min, and the reaction
mixture heated at reflux for 1 h. Upon cooling to room
temperature, a solution of aldehyde 12 (3.00 g, 20.0 mmol) in
THF (12 mL) was added dropwise over 15 min, and the
reaction mixture was heated at reflux for 1 h and then was
cooled to 0 °C. The reaction was then quenched with saturated
aq NH4Cl solution. The reaction mixture was poured into sat.
aq NH4Cl solution and extracted with Et2O (2 × 100 mL). The
organic phase was washed with water (100 mL) and brine (100
mL). The aq phase was back extracted with Et2O (50 mL), and
the combined organic layers were dried (Na2SO4). The crude
material was purified by chromatography on silica (hexane/
EtOAc, 89/11) and distilled (bulb-to-bulb) to provide 5.26 g
(85%) of 24 as a colorless liquid. Data for 24: bp 250 °C (3.4
1-(P h en ylm eth oxy)-3-bu ten -2-on e (16). A solution of
dimethyl sulfoxide (440 µL, 6.20 mmol, 2.2 equiv) in CH2Cl2
(22 mL) was cooled to -74 °C and then treated with oxalyl
chloride (270 µL, 3.10 mmol, 1.1 equiv) in one portion. After
45 min, a solution of alcohol 14 (506 mg, 2.84 mmol) in
CH2Cl2 (6 mL) was added dropwise via cannula. The resultant
cloudy mixture was maintained for 1 h before triethylamine
(2.0 mL, 14 mmol, 5 equiv) was introduced dropwise. The
reaction mixture was maintained at -74 °C for 10 min and
then allowed to warm to room temperature over 15 min. The
reaction mixture was diluted with CH2Cl2 (60 mL) and washed
with 0.1 M hydrochloric acid (40 mL), water (40 mL), and brine
(40 mL). The combined aq layers were back extracted with
CH2Cl2 (40 mL), and the organic layers were dried (Na2SO4).
The residue was purified by rapid chromatography on neutral
alumina (III) (hexane/EtOAc, 86/14) to provide 355 mg (68%)
of ketone 16 as a light gold liquid. The ketone was further
purified by distillation (bulb-to-bulb) to provide 144 mg (29%)
as a colorless liquid that rapidly polymerized. Data for 16: bp
140 °C (1.2 × 10-5 mmHg); 1H NMR (400 MHz, CDCl3) δ 7.37-
7.31 (m, 5H), 6.55 (dd, J ) 17.6, 10.5, 1H), 6.34 (dd, J ) 17.6,
1.5, 1H), 5.83 (dd, J ) 10.7, 1.5, 1H), 4.62 (s, 2H), 4.28 (s, 2H);
13C NMR (100 MHz, CDCl3) δ 197.0, 137.1, 132.3, 129.2, 128.4,
127.9, 127.9, 73.7, 73.2; IR (neat) 1714 (s), 1698 (s) cm-1; TLC
Rf 0.30 (hexane/EtOAc, 80/20).
1
× 10-5 mmHg); H NMR (500 MHz, CDCl3) δ 7.62-7.29 (m,
10H), 4.63 (d, J ) 12, 1H), 4.61 (d, J ) 12, 1H), 4.62-4.59 (m,
1H), 3.68 (dd, J ) 9.9, 3.7, 1H), 3.60 (dd, J ) 9.7, 7.3, 1H),
2.47 (d, J ) 4.9, 1H), 0.42 (s, 6H); 13C NMR (125 MHz, CDCl3)
δ 137.6, 136.5, 133.7, 129.5, 128.5, 127.9, 127.7, 104.8, 88.6,
73.5, 73.4, 62.2, -1.1; IR (neat) 2959 (m) cm-1; MS (CI, CH4)
m/z 311 (M+ + 1, weak), 91 (100); TLC Rf 0.44 (hexane/EtOAc,
67/33). Anal. Calcd for C19H22O2Si (310.47): C, 73.50; H, 7.14;
Si, 9.05. Found: C, 73.47; H, 7.19; Si, 9.23.
(Z)-1-[(Dim eth yl(1,1,2-tr im eth ylp r op yl)silyl)oxy]-4-(d i-
m eth ylp h en ylsilyl)-3-bu ten -2-ol (25). Freshly distilled cy-
clohexene (3.4 mL, 34 mmol, 4 equiv) was added dropwise over
5 min to a solution of BH3‚THF (1.0 M, 16.5 mL, 16.5 mmol,
2 equiv) maintained at 0 °C. The reaction mixture was
maintained for 1 h during which it gradually thickened. A
solution of alkyne 23 (3.00 g, 8.27 mmol) in THF (8 mL) was
added dropwise to the white suspension over 15 min. The
reaction mixture was maintained at 0 °C for 1 h (reaction
mixture clarified) and then allowed to warm to room temper-
ature. After 1 h, acetic acid (950 µL, 16.6 mmol, 2 equiv) was
added dropwise over 5 min, and the reaction mixture was

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 895
maintained overnight (12 h). The contents of the reaction were
diluted with Et2O (100 mL) and washed with water (50 mL)
and brine (50 mL). The combined aq layers were back extracted
with Et2O (50 mL), and the organic layers were dried (Na2-
SO4). The crude material was purified by chromatography on
silica (hexane/EtOAc, 96/4) to give a colorless oil which was
distilled (bulb-to-bulb) to provide 2.14 g (71%) of cis-alkene
25 as a colorless liquid. Data for 25: bp 165 °C (7.0 × 10-5
mmHg); 1H NMR (500 MHz, CDCl3) δ 7.55-7.34 (m, 5H), 6.32
(dd, J ) 14.5, 8.2, 1H), 5.91 (d, J ) 14.3, 1H), 4.19-4.16 (m,
1H), 3.43 (dd, J ) 10.1, 3.8, 1H), 3.33 (dd, J ) 9.7, 7.9, 1H),
2.48 (d, J ) 2.2, 1H), 1.59 (septet, J ) 6.8, 1H), 0.85 (d, J )
6.8, 6H), 0.82 (s, 6H), 0.42 (s, 3H), 0.40 (s, 3H), 0.05 (s, 3H),
0.03 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 147.4, 139.2, 133.6,
131.5, 129.0, 127.9, 72.3, 66.2, 34.1, 25.1, 20.2, 18.5, -0.6, -0.9,
-3.5, -3.6; IR (neat) 3486 (bw), 2958 (s) cm-1; MS (CI, CH4)
m/z 365 (M+ + 1, weak), 147 (100); TLC Rf 0.35 (hexane/EtOAc,
90/10). Anal. Calcd for C20H36O2Si2 (364.68): C, 65.87; H, 9.95;
Si, 15.40. Found: C, 65.90; H, 10.28; Si, 15.59.
over 15 min. The reaction mixture was diluted with pentane
(300 mL) and washed with 0.1 M hydrochloric acid (100 mL),
water (100 mL), and brine (100 mL). The combined aq layers
were back extracted with pentane (100 mL), and the organic
layers were dried (Na2SO4). The residue was purified by
chromatography on silica (pentane, then pentane/Et2O, 97.5/
2.5) to provide 3.51 g (97%) of ketone 27 as a light gold oil.
Data for 27: 1H NMR (500 MHz, CDCl3) δ 7.60-7.34 (m, 5H),
7.22 (d, J ) 14.3, 1H), 6.66 (d, J ) 14.3, 1H), 4.24 (s, 2H), 1.66
(septet, J ) 6.8, 1H), 0.91 (d, J ) 7.0, 6H), 0.89 (s, 6H), 0.47
(s, 6H), 0.12 (s, 6H); 13C NMR (125 MHz, CDCl3) δ 199.5, 150.2,
139.5, 138.3, 133.7, 128.7, 127.6, 69.3, 34.1, 25.2, 20.2, 18.5,
-2.6, -3.5; IR (neat) 1715 (w), 1694 (m) cm-1; MS (CI, CH4)
m/z 364 (M+ + 1, 4), 201 (100); TLC Rf 0.57 (hexane/EtOAc,
94/6). Anal. Calcd for C20H34O2Si2 (362.66): C, 66.24; H, 9.45;
Si, 15.49. Found: C, 66.40; H, 9.36; Si, 15.32.
(Z)-4-(Dim eth ylph en ylsilyl)-1-(ph en ylm eth oxy)-3-bu ten -
2-on e (28). A solution of dimethyl sulfoxide (170 µL, 2.40
mmol, 2.2 equiv) in CH2Cl2 (6 mL) was cooled to -74 °C and
treated with oxalyl chloride (100 µL, 1.15 mmol, 1.1 equiv) in
one portion. After 30 min, a solution of alcohol (Z)-26 (341 mg,
1.09 mmol) in CH2Cl2 (5 mL) was added dropwise via cannula.
The resultant cloudy mixture was maintained for 1 h before
triethylamine (760 µL, 5.45 mmol, 5 equiv) was introduced
dropwise. The reaction mixture was maintained at -74 °C for
5 min and then allowed to warm over 15 min. The reaction
mixture was diluted with CH2Cl2 (50 mL) and washed with
0.1 M hydrochloric acid (20 mL), water (20 mL), and brine (20
mL). The combined aq layers were back extracted with CH2-
Cl2 (20 mL), and the organic layers were dried (Na2SO4). The
residue was purified by rapid chromatography on neutral
alumina (III) (hexane/EtOAc, 92/8) to provide 270 mg (80%)
of ketone 28 as a light gold oil. Data for 28: 1H NMR (500
MHz, CDCl3) δ 7.59-7.28 (m, 10H), 7.06 (d, J ) 14.3, 1H),
6.65 (d, J ) 14.1, 1H), 4.55 (s, 2H), 4.14 (s, 2H), 0.46 (s, 6H);
13C NMR (125 MHz, CDCl3) δ 197.3, 151.0, 139.2, 138.3, 137.0,
133.7, 128.5, 128.0, 128.0, 127.6, 75.0, 73.2, -2.6; IR (neat)
1709 (s), 1694 (s) cm-1; MS (CI, CH4) m/z 311 (M+ + 1, 13),
233 (100); TLC Rf 0.25 (hexane/EtOAc, 94/6). Anal. Calcd for
(Z)-4-(Dim eth ylph en ylsilyl)-1-(ph en ylm eth oxy)-3-bu ten -
2-ol (26). Palladium on calcium carbonate (102 mg, 5% (no
lead content)) was suspended in benzene (1 mL) and prehy-
drogenated for 20 min. Quinoline (23 µL, 0.19 mmol, 0.06
equiv) was added followed by a solution of alkyne 24 (1.02 g,
3.29 mmol) in benzene (2.3 mL). The progress of the reaction
was monitored by HPLC, and after 2 h and 40 min, the
starting material had been consumed. The catalyst was
removed by filtration (SiO2/EtOAc), and the crude material
was prepurified by chromatography on silica (hexane/EtOAc,
80/20). The three components were separated by MPLC
(hexane/EtOAc, 90/10) and individually distilled (bulb-to-bulb)
to provide 429 mg (42%) of the cis-alkene (Z)-26, 76.1 mg (7%)
of the trans-alkene (E)-26, and 183 mg (18%) of the saturated
alkane 26b as colorless liquids. Data for (Z)-26: bp 170 °C
1
(3.2 × 10-5 mmHg); H NMR (500 MHz, CDCl3) δ 7.55-7.25
(m, 10H), 6.32 (dd, J ) 14.3, 8.2, 1H), 5.91 (dd, J ) 14.3, 0.9,
1H), 4.44 (d, J ) 13.0, 1H), 4.41 (d, J ) 13.0, 1H), 4.39-4.34
(m, 1H), 3.30 (d, J ) 5.3, 2H), 2.29 (d, J ) 2.7, 1H), 0.41 (s,
3H), 0.38 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 147.0, 139.1,
137.8, 133.6, 131.7, 129.0, 128.4, 127.9, 127.8, 127.7, 73.5, 73.1,
70.9, -0.8, -1.0; IR (neat) 3448 (bs), 2954 (s) cm-1; MS (CI,
CH4) m/z 330 ([M + CH5]+, 49), 219 (100); TLC Rf 0.44 (hexane/
EtOAc, 67/33). Anal. Calcd for C19H24O2Si (312.49): C, 73.03;
H, 7.74; Si, 8.99. Found: C, 72.92; H, 7.83; Si, 8.74. Data for
(E)-26: bp 190 °C (6.0 × 10-5 mmHg); 1H NMR (500 MHz,
CDCl3) δ 7.52-7.28 (m, 10H), 6.15 (dd, J ) 18.8, 1.3, 1H), 5.91
(ddd, J ) 18.8, 4.4, 2.0, 1H), 4.56 (s, 2H), 4.39-4.37 (m, 1H),
3.56 (dd, J ) 9.5, 3.3, 1H), 3.36 (ddd, J ) 9.5, 7.9, 1.6, 1H),
2.52-2.50 (m, 1H), 0.34 (s, 6H); 13C NMR (125 MHz, CDCl3) δ
145.5, 138.3, 137.8, 133.8, 129.3, 129.0, 128.5, 127.8, 127.8,
127.7, 73.8, 73.3, 72.9, -2.7, -2.7; IR (neat) 3434 (bs), 2954
(s) cm-1; MS (CI, CH4) m/z 330 ([M + CH5]+, 6.4), 91 (100);
TLC Rf 0.43 (hexane/EtOAc, 67/33). Anal. Calcd for C19H24O2-
Si (312.49): C, 73.03; H, 7.74; Si, 8.99. Found: C, 72.76; H,
7.76; Si, 9.04. Data for 26b: bp 190 °C (6.5 × 10-5 mmHg); 1H
NMR (500 MHz, CDCl3) δ 7.56-7.32 (m, 10H), 4.58 (d, J )
19.4, 1H), 4.54 (d, J ) 19.4, 1H), 3.78-3.74 (m, 1H), 3.53 (dd,
J ) 9.3, 2.7, 1H), 3.37-3.33 (m, 1H), 2.51 (m, 1H), 1.55-1.47
(m, 2H), 0.98-0.91 (m, 1H), 0.79-0.69 (m, 1H), 0.41 (s, 3H),
0.38 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 138.9, 137.9, 133.5,
128.8, 128.4, 127.7, 127.7, 127.7, 74.0, 73.2, 72.5, 27.3, 11.1,
-3.3; IR (neat) 3435 (bs) cm-1; MS (CI, CH4) m/z 332 ([M +
CH4]+, 2.0), 91 (100); TLC Rf 0.40 (hexane/EtOAc, 67/33). Anal.
Calcd for C19H26O2Si (314.50): C, 72.56; H, 8.33; Si, 8.93.
Found: C, 72.45; H, 8.14; Si, 9.22.
C
19H22O2Si (310.47): C, 73.50; H, 7.14; Si, 9.05: Found: C,
73.79; H, 7.25; Si, 9.22.
r el-(2S,3a S,4R,6R)-4-(Ben zoyloxy)-6-(b u t yloxy)h exa -
h ydr oisoxazolo[2,3-b][1,2]oxazin e-2-car boxylic Acid Meth -
yl Ester (29a ) a n d r el-(2R,3a S,4R,6R)-4-(Ben zoyloxy)-6-
(bu tyloxy)h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-2-ca r -
boxylic Acid Meth yl Ester (29b). Methyl acrylate (1.2 mL,
13 mmol, 6 equiv) was added to a solution of nitronate 2 (666
mg, 2.27 mmol) and several crystals of 1,4-di-tert-butylhydro-
quinone in benzene (75 mL), and the reaction mixture was
maintained at room temperature for 6 h. Upon concentration
of the solution in vacuo, the residue (931 mg) was purified by
radial chromatography (hexane/TBME, 60/40, 4 mm SiO2
plate). The major nitroso acetal was recrystallized from TBME/
pentane to provide 599 mg (70%) of 29a as colorless, spiny
needles. The minor nitroso acetal was recrystallized from
TBME/pentane to provide 78.5 mg (9%) of 29b as colorless,
spiny needles. Data for 29a : mp 61-62 °C (TBME/pentane);
1H NMR (400 MHz, CDCl3) δ 8.08 (d, J ) 7.1, 2H), 7.59 (t, J
) 7.6, 1H), 7.46 (t, J ) 7.8, 2H), 5.27 (ddd, J ) 8.3, 4.6, 3.7,
1H), 5.09 (dd, J ) 9.0, 5.0, 1H), 4.97 (t, J ) 5.1, 1H), 3.97 (dt,
J ) 9.3, 6.6, 1H), 3.85-3.80 (m, 1H), 3.77 (s, 3H), 3.46 (dt, J
) 9.3, 6.6, 1H), 2.69-2.63 (m, 2H), 2.30 (ddd, J ) 14, 5.6, 5.1,
1H), 2.16 (ddd, J ) 14, 8.1, 4.9, 1H), 1.67-1.60 (m, 2H), 1.46-
1.37 (m, 2H), 0.93 (t, J ) 7.6, 3H); 13C NMR (100 MHz, CDCl3)
δ 170.3, 165.7, 133.2, 129.7, 129.4, 128.3, 98.2, 80.5, 72.0, 68.9,
(Z)-1-[(Dim eth yl(1,1,2-tr im eth ylp r op yl)silyl)oxy]-4-(d i-
m eth ylp h en ylsilyl)-3-bu ten -2-on e (27). A solution of dim-
ethyl sulfoxide (1.55 mL, 21.5 mmol, 2.2 equiv) in CH2Cl2 (80
mL) was cooled to -74 °C and treated with oxalyl chloride (955
µL, 10.9 mmol, 1.1 equiv) in one portion. After 45 min, a
solution of alcohol 25 (3.63 g, 9.95 mmol) in CH2Cl2 (18 mL)
was added dropwise via cannula. The resultant cloudy mixture
was maintained for 1 h before triethylamine (7.0 mL, 50 mmol,
5 equiv) was introduced dropwise. The reaction mixture was
maintained at -74 °C for 5 min and then allowed to warm
67.5, 52.4, 32.9, 31.5, 29.5, 19.2, 13.8; IR (CCl4) 1725 (s) cm-1
;
MS (CI, CH4) m/z 380 (M+ + 1, 11), 105 (100); TLC Rf 0.32
(hexane/EtOAc, 67/33). Anal. Calcd for C19H25NO7 (379.41): C,
60.15; H, 6.64; N, 3.69. Found: C, 60.33; H, 6.60; N, 3.71. Data
1
for 29b: mp 79-80 °C (TBME/pentane); H NMR (400 MHz,
CDCl3) δ 8.08 (d, J ) 8.3, 2H), 7.58 (t, J ) 7.3, 1H), 7.45 (t, J
) 8.1, 2H), 5.34 (ddd, J ) 8.1, 4.6, 3.2, 1H), 4.95 (t, J ) 4.6,
1H), 4.84 (t, J ) 8.1, 1H), 4.00 (dt, J ) 9.5, 6.6, 1H), 3.82 (s,
3H), 3.74 (td, J ) 9.3, 2.9, 1H), 3.44 (dt, J ) 9.3, 6.6, 1H), 2.72
(t, J ) 8.3, 2H), 2.28 (dt, J ) 14.4, 5.4, 1H), 2.09 (ddd, J )

896 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
14.4, 6.8, 4.2, 1H), 1.66-1.59 (m, 2H), 1.46-1.37 (m, 2H), 0.92
(t, J ) 7.3, 3H); 13C NMR (100 MHz, CDCl3) δ 171.1, 165.8,
133.2, 129.8, 129.6, 128.3, 98.4, 80.8, 72.3, 68.6, 67.7, 52.6, 31.6,
31.0, 29.1, 19.2, 13.9; IR (CCl4) 1722 (s) cm-1; MS (CI, CH4)
m/z 380 (M+ + 1, 13), 105 (100); TLC Rf 0.28 (hexane/EtOAc,
67/33). Anal. Calcd for C19H25NO7 (379.41): C, 60.15; H, 6.64;
N, 3.69. Found: C, 60.18; H, 6.65; N, 3.61.
25.1; IR (CCl4) 1726 (s) cm-1; MS (CI, CH4) m/z 525 (M+ + 1,
39), 159 (100); TLC Rf 0.34 (hexane/EtOAc, 80/20). Anal. Calcd
for C30H37NO7 (523.63): C, 68.81; H, 7.12; N, 2.67. Found: C,
68.92; H, 7.16; N, 2.61. Data for 31b: mp 143-144 °C (Et2O/
pentane); 1H NMR (500 MHz, CDCl3) δ 8.03 (dd, J ) 8.2, 1.1,
2H), 7.58 (t, J ) 7.5, 1H), 7.45 (t, J ) 7.7, 2H), 7.25-7.14 (m,
5H), 4.98 (ddd, J ) 9.3, 4.8, 4.6, 1H), 4.66 (dd, J ) 9.0, 7.0,
1H), 4.19 (t, J ) 6.0, 1H), 3.63-3.56 (m, 2H), 2.68-2.46 (m,
4H), 1.89-1.74 (m, 5H), 1.64-1.29 (m, 4H), 1.50 (s, 9H); 13C
NMR (125 MHz, CDCl3) δ 169.9, 165.8, 144.3, 133.3, 129.7,
129.7, 128.4, 128.0, 127.8, 126.2, 98.7, 82.2, 81.8, 80.8, 73.5,
70.0, 51.2, 34.5, 32.7, 32.2, 30.1, 27.9, 25.8, 25.3; IR (KBr) 1739
(s), 1719 (s), 1713 (s) cm-1; MS (CI, CH4) m/z 524 (M+, 7), 292
(100); TLC Rf 0.28 (hexane/EtOAc, 80/20). Anal. Calcd for
C30H37NO7 (523.63): C, 68.81; H, 7.12; N, 2.67. Found: C,
68.93; H, 7.30; N, 2.71.
r el -(2S ,3a S ,4R ,6R )-4-(B e n z o y lo x y )-6-[(1S ,2R )-(2-
p h en ylcycloh exyl)oxy]h exa h yd r oisoxa zolo[2,3-b][1,2]-
oxa zin e-2-ca r b oxylic Acid Met h yl E st er (30a ) a n d
r el-(2R,3aS,4R,6R)4-(Ben zoyloxy)-6-[(1S,2R)(2-ph en ylcyclo-
hexyl)oxy]hexahydroisoxazolo[2,3-b][1,2]oxazine-2-carboxy-
lic Acid Meth yl Ester (30b). Methyl acrylate (1.6 mL, 18
mmol, 5 equiv) was added to a solution of nitronate 3 (1.42 g,
3.59 mmol) and several crystals of 1,4-di-tert-butylhydro-
quinone in benzene (118 mL), and the reaction mixture was
maintained at room temperature for 6 h. Upon concentration
of the solution in vacuo, the residue was purified by chroma-
tography on silica (hexane/EtOAc/, 80/20). The major nitroso
acetal (1.43 g) was recrystallized from CH2Cl2/hexane to
provide 1.29 g (75%) of 30a as colorless, spiny needles. The
minor nitroso acetal was recrystallized from CH2Cl2/hexane
to provide 197 mg (11%) of 30b as colorless, spiny needles.
Data for 30a : mp 125-126 °C (CH2Cl2/hexane); 1H NMR (400
MHz, CDCl3) δ 8.03 (dd, J ) 8.3, 1.2, 2H), 7.59 (tt, J ) 7.5,
1.5, 1H), 7.46 (t, J ) 7.8, 2H), 7.27-7.15 (m, 5H), 5.02 (dd, J
) 9.8, 3.7, 1H), 4.95 (dt, J ) 9.3, 4.9, 1H), 4.12 (t, J ) 5.9,
1H), 3.74 (s, 3H), 3.71 (ddd, J ) 11.5, 7.8, 3.9, 1H), 3.63 (dd,
J ) 10.5, 4.4, 1H), 2.62 (ddd, J ) 12.2, 7.8, 3.9, 1H), 2.56-
2.48 (m, 2H), 2.42-2.38 (m, 1H), 1.90-1.83 (m, 3H), 1.78-
1.71 (m, 2H), 1.67-1.27 (m, 5H); 13C NMR (100 MHz, CDCl3)
δ 170.3, 165.6, 144.1, 133.3, 129.7, 129.4, 128.3, 128.0, 127.8,
126.2, 98.7, 81.9, 80.6, 72.6, 69.5, 52.4, 51.2, 34.5, 33.5, 32.4,
29.8, 25.7, 25.2; IR (CCl4) 1726 (s) cm-1; MS (CI, CH4) m/z 482
(M+ + 1, 7.8), 159 (100); TLC Rf 0.41 (hexane/EtOAc, 67/33).
Anal. Calcd for C27H31NO7 (481.55): C, 67.35; H, 6.49; N, 2.91.
Found: C, 67.49; H, 6.62; N, 2.94. Data for 30b: mp 107-108
°C (CH2Cl2/hexane); 1H NMR (400 MHz, CDCl3) δ 8.06 (dd, J
) 8.1, 1.2, 2H), 7.59 (tt, J ) 7.3, 1.2, 1H), 7.46 (t, J ) 7.8, 2H),
7.26-7.14 (m, 5H), 5.05 (ddd, J ) 7.1, 5.9, 3.7, 1H), 4.78 (dd,
J ) 9.0, 7.3, 1H), 4.17 (t, J ) 5.4, 1H), 3.80 (s, 3H), 3.66-3.60
(m, 2H), 2.71-2.44 (m, 4H), 1.88-1.85 (m, 2H), 1.79-1.71 (m,
3H), 1.67-1.27 (m, 4H); 13C NMR (100 MHz, CDCl3) δ 171.0,
165.6, 144.1, 133.2, 129.7, 129.6, 128.3, 127.9, 127.7, 126.1,
99.0, 82.1, 80.8, 72.7, 69.0, 52.4, 51.2, 34.4, 32.5, 31.5, 29.2,
25.8, 25.1; IR (CCl4) 1724 (s) cm-1; MS (CI, CH4) m/z 483 (M+
+ 1, 4), 159 (100); TLC Rf 0.31 (hexane/EtOAc, 67/33). Anal.
Calcd for C27H31NO7 (481.55): C, 67.35; H, 6.49; N, 2.91.
Found: C, 67.21; H, 6.53; N, 2.90.
r el-(2S,3a S,4R,6R)-4-(Ben zoyloxy)-6-[(1S,2R)-(2-p h e-
n ylcycloh exyl)oxy]h exah ydr oisoxazolo[2,3-b][1,2]oxazin e-
2-ca r b oxylic Acid 1,1-Dim et h ylet h yl E st er (31a ) a n d
r el-(2R,3a S,4R,6R)-4-(Ben zoyloxy)-6-[(1S,2R)-(2-p h en -
ylcycloh exyl)oxy]h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-
2-ca r boxylic Acid 1,1-Dim eth yleth yl Ester (31b). tert-
Butyl acrylate (1.6 mL, 11 mmol, 5 equiv) was added to a
solution of nitronate 3 (803 mg, 2.03 mmol) and several
crystals of 1,4-di-tert-butylhydroquinone in benzene (64 mL),
and the reaction mixture was maintained at room temperature
for 2.5 h. Upon concentration of the solution in vacuo, the
residue (1.23 g) was purified by chromatography on silica
(hexane/EtOAc, 91/9, 89/11, 86/14). The major nitroso acetal
(960 mg) was recrystallized from TBME/pentane to provide
891 mg (84%) of 31a as colorless, spiny needles. The minor
nitroso acetal was recrystallized from Et2O/pentane to provide
78.0 mg (7%) of 31b as colorless, spiny needles. Data for 31a :
mp 138-139 °C (TBME/pentane); 1H NMR (400 MHz, CDCl3)
δ 8.04 (dd, J ) 8.3, 1.2, 2H), 7.59 (tt, J ) 7.6, 1.2, 1H), 7.46 (t,
J ) 7.8, 2H), 7.29-7.15 (m, 5H), 4.95 (dt, J ) 9.0, 4.6, 1H),
4.88 (dd, J ) 9.8, 4.2, 1H), 4.12 (t, J ) 5.9, 1H), 3.70 (ddd, J
) 10.3, 8.1, 3.9, 1H), 3.63 (td, J ) 10.3, 4.2, 1H), 2.57-2.39
(m, 4H), 1.89-1.82 (m, 3H), 1.77-1.70 (m, 2H), 1.63-1.26 (m,
4H), 1.44 (s, 9H); 13C NMR (100 MHz, CDCl3) δ 168.9, 165.6,
144.1, 133.2, 129.7, 129.5, 128.3, 127.9, 127.8, 126.2, 98.6, 82.2,
81.9, 81.4, 72.6, 69.6, 51.2, 34.5, 33.4, 32.4, 29.8, 27.8, 25.7,
r el-(2S,3a S,4R,6R)-4-(Ben zoyloxy)-6-[(1S,2R)-(2-p h e-
n ylcycloh exyl)oxy]-2-(1-oxoet h yl)h exa h yd r oisoxa zolo-
[2,3-b][1,2]oxa zin e (32a ) a n d r el-(2R,3a S, 4R,6R)-4-(Ben -
zo y lo x y )-6-[(1S ,2R )-(2-p h e n y lc y c lo h e x y l)o x y ]-2-(1-
oxoet h yl)h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e (32b ).
Methyl vinyl ketone (0.5 mL, 6.0 mmol, 8 equiv) was added to
a solution of nitronate 3 (302 mg, 0.764 mmol) and several
crystals of 1,4-di-tert-butylhydroquinone in benzene (25 mL),
and the reaction mixture was maintained at room temperature
for 1.5 h. Upon concentration of the solution in vacuo, the
residue was purified by chromatography on silica (hexane/
EtOAc, 86/14). The inseparable mixture of diastereomers (345
mg) was recrystallized from CH2Cl2/hexane to provide 306 mg
(88%) of a 6.2/1 mixture of ketones 32a and 32b as colorless,
spiny needles. Data for 6.2/1 mixture of 32a and 32b: mp 165-
1
166 °C (dec); H NMR (500 MHz, CDCl3) δ 8.04 (d, J ) 7.1,
2H), 7.59 (t, J ) 7.3, 1H), 7.46 (t, J ) 7.9, 2H), 7.27-7.15 (m,
5H), 4.95 (dt, J ) 9.0, 4.6, 1H), 4.92 (dd, J ) 10.2, 4.2, 1H),
4.61 (dd, J ) 9.5, 6.2, 1H), 4.16 (t, J ) 6, 1H), 3.64 (td, J )
10.4, 4.4, 1H), 3.57 (ddd, J ) 10.4, 8.4, 4.2, 1H), 2.58 (ddd, J
) 12.4, 8.2, 4.2, 1H), 2.57-2.51 (m, 1H), 2.43-2.39 (m, 1H),
2.42 (dt, J ) 12.6, 10.2, 1H), 2.36 (s, 3H), 2.22 (s, 3H), 1.90-
1.71 (m, 6H), 1.67-1.54 (m, 2H), 1.52-1.43 (m, 1H), 1.38 (m,
1H); 13C NMR (125 MHz, CDCl3) δ 205.9, 200.7, 165.6, 144.0,
133.3, 129.7, 129.5, 128.3, 128.3, 128.0, 128.0, 127.8, 127.7,
126.3, 99.0, 98.9, 88.0, 87.1, 82.3, 72.5, 69.5, 51.3, 34.6, 32.4,
31.9, 29.7, 26.6, 25.7, 25.2; IR (KBr) 1721 (s) cm-1; MS (CI,
CH4) m/z 466 (M+ + 1, 3), 159 (100); TLC Rf 0.15 (hexane/
EtOAc, 80/20). Anal. Calcd for C27H31NO6 (465.55): C, 69.66;
H, 6.71; N, 3.01. Found: C, 69.36; H, 6.66; N, 3.05.
rel-(2S,3aS,4R,6R)-4-(Benzoyloxy)-2-{2-[(dimethyl(1,1,2-tri-
m eth ylpr opyl)silyl)oxy]-1-oxoeth yl}-6-[(1S,2R)-(2-ph en yl-
cycloh exyl)oxy]h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e
(33a ) a n d r el-(2R,3a S,4R,6R)-4-(Ben zoyloxy)-2-{2-[(d im -
e t h yl(1,1,2-t r im e t h ylp r op yl)silyl)oxy]-1-oxoe t h yl}-6-
[(1S,2R)-(2-p h en ylcycloh exyl)oxy]h exa h yd r oisoxa zolo-
[2,3-b][1,2]oxa zin e (33b). A solution of dimethyl sulfoxide
(340 µL, 4.79 mmol, 2.2 equiv) in CH2Cl2 (20 mL) was cooled
to -74 °C and treated with oxalyl chloride (210 µL, 2.41 mmol,
1.1 equiv) in one portion. After 30 min, a solution of alcohol
13 (501 mg, 2.17 mmol) in CH2Cl2 (2 mL) was added dropwise
via cannula. The resultant cloudy mixture was maintained for
1 h before triethylamine (1.5 mL, 11 mmol, 5 equiv) was
introduced dropwise. The reaction mixture was maintained at
-74 °C for 10 min and then allowed to warm over 15 min.
The reaction mixture was diluted with Et2O (120 mL) and
washed with 0.1 M hydrochloric acid (30 mL), water (30 mL),
and brine (30 mL). The aq layers were back extracted with
Et2O (50 mL), and the combined organic layers were dried
(Na2SO4). The residue was purified by rapid chromatography
on silica (pentane/Et2O, 91/9) and dried at 0.1 mmHg to
provide 374 mg (75%) of ketone 15 as a golden oil. The ketone
was not purified further but was used directly in the next
reaction. Nitronate 3 (490 mg, 1.24 mmol) was added to a
solution of ketone 15 (369 mg, 1.62 mmol, 1.3 equiv) and
several crystals of 1,4-di-tert-butylhydroquinone in benzene (12
mL), and the reaction mixture was maintained at room
temperature for 4 h. Upon concentration of the solution in
vacuo, the residue was prepurified by gradient chromatogra-

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 897
phy on silica (pentane/Et2O, 91/9, then Et2O) to provide 770
mg of ketones 33a and 33b in a 5/1 ratio. The diastereoisomers
were separated by radial chromatography (hexane/EtOAc, 89/
11, 4 mm SiO2 plate) to give 607 mg (79%) of nitroso acetal
33a and 87.2 mg (11%) of nitroso acetal 33b as foams after
drying at 7.5 × 10-5 mmHg. Data for 33a : 1H NMR (500 MHz,
CDCl3) δ 8.03 (dd, J ) 8.2, 1.1, 2H), 7.58 (tt, J ) 7.5, 1.3, 1H),
7.45 (t, J ) 8.2, 2H), 5.15 (dd, J ) 10.2, 4.0, 1H), 4.95 (ddd, J
) 9.2, 4.9, 4.8, 1H), 4.43 (s, 2H), 4.16 (t, J ) 5.9, 1H), 3.63 (td,
J ) 10.4, 4.4, 1H), 3.55 (ddd, J ) 10.6, 8.2, 4.2, 1H), 2.61 (ddd,
J ) 12.4, 8.1, 4.2, 1H), 2.53 (ddd, J ) 13.0, 10.4, 3.7, 1H), 2.45
(dt, J ) 12.6, 10.6, 1H), 2.41-2.38 (m, 1H), 1.89-1.72 (m, 6H),
1.66-1.53 (m, 2H), 1.51-1.42 (m, 1H), 1.37-1.29 (m, 1H), 0.88
(d, J ) 7.0, 6H), 0.85 (s, 6H), 0.10 (s, 3H), 0.09 (s, 3H); 13C
NMR (125 MHz, CDCl3) δ 205.6, 165.6, 144.1, 133.3, 129.7,
128.4, 128.1, 127.8, 126.3, 98.9, 85.1, 82.3, 72.3, 69.5, 67.4, 51.3,
34.6, 34.0, 32.5, 32.1, 29.8, 25.8, 25.2, 25.2, 20.1, 18.4, -3.6;
IR (CCl4) 1726 (s) cm-1; MS (CI, CH4) m/z 624 (M+, 2.5), 159
(100); TLC Rf 0.41 (hexane/EtOAc, 80/20). Anal. Calcd for
C35H49NO7Si (623.87): C, 67.38; H, 7.92; N, 2.25. Found: C,
67.22; H, 7.97; N, 2.34. Data for 33b : 1H NMR (500 MHz,
CDCl3) δ 8.04 (dd, J ) 8.1, 1.1, 2H), 7.58 (tt, J ) 7.5, 1.3, 1H),
7.46 (t, J ) 8.1, 2H), 7.24-7.14 (m, 5H), 4.96 (ddd, J ) 8.6,
4.9, 3.8, 1H), 4.87 (d, J ) 18.8, 1H), 4.76 (dd, J ) 9.5, 6.2,
1H), 4.57 (d, J ) 18.7, 1H), 4.10 (t, J ) 5.7, 1H), 3.59 (m, 1H),
3.54 (td, J ) 10.6, 4.6, 1H), 2.62 (ddd, J ) 12.6, 9.5, 8.6, 1H),
2.55-2.47 (m, 2H), 2.41-2.38 (m, 1H), 1.91-1.87 (m, 2H),
1.79-1.55 (m, 6H), 1.47-1.27 (m, 2H), 0.93 (d, J ) 7.0, 6H),
0.91 (s, 6H), 0.16 (s, 3H), 0.14 (s, 3H); 13C NMR (125 MHz,
CDCl3) δ 207.2, 165.7, 144.0, 133.3, 129.8, 129.7, 128.4, 128.1,
127.8, 126.4, 99.2, 86.5, 82.7, 73.1, 69.5, 67.6, 51.2, 34.6, 34.1,
32.6, 30.6, 29.5, 25.8, 25.4, 25.3, 20.4, 20.3, 18.5, 18.5, -3.4,
-3.6; IR (CCl4) 2931 (s), 1720 (s), 1716 (s) cm-1; MS (CI, CH4)
m/z 624 (M+, 1), 159 (100); TLC Rf 0.31 (hexane/EtOAc, 80/
20). Anal. Calcd for C35H49NO7Si (623.87): C, 67.38; H, 7.92;
N, 2.25. Found: C, 67.10; H, 7.97; N, 2.19.
(dd, J ) 10.4, 4.2, 1H), 4.93 (ddd, J ) 9.1, 4.8, 4.2, 1H), 4.59
(d, J ) 11.9, 1H), 4.56 (d, J ) 11.9, 1H), 4.31 (s, 2H), 4.15 (t,
J ) 5.9, 1H), 3.61 (td, J ) 10.4, 4.4, 1H), 3.51 (ddd, J ) 10.2,
8.2, 3.8, 1H), 2.61 (ddd, J ) 12.4, 8.2, 4.2, 1H), 2.55-2.50 (m,
1H), 2.46 (dt, J ) 12.8, 10.4, 1H), 2.38-2.35 (m, 1H), 1.89-
1.70 (m, 5H), 1.66-1.42 (m, 3H), 1.37-1.28 (m, 1H); 13C NMR
(125 MHz, CDCl3) δ 204.3, 165.6, 144.0, 136.8, 133.3, 129.7,
129.5, 128.4, 128.3, 128.1, 128.0, 128.0, 127.8, 126.3, 98.9, 85.5,
82.3, 73.3, 72.7, 72.3, 69.4, 51.3, 34.5, 32.5, 32.2, 29.7, 25.7,
25.1; IR (CCl4) 1726 (s) cm-1; MS (CI, CH4) m/z 572 (M+, 1),
91 (100); TLC Rf 0.51 (hexane/EtOAc, 67/33). Anal. Calcd for
C34H37NO7 (571.68): C, 71.44; H, 6.52; N, 2.45. Found: C,
71.31; H, 6.57; N, 2.52. Data for 34b : 1H NMR (500 MHz,
CDCl3) δ 8.04 (d, J ) 7.1, 2H), 7.59 (t, J ) 7.3, 1H), 7.46 (t, J
) 7.9, 2H), 7.41-7.12 (m, 10H), 4.97 (ddd, J ) 8.8, 5.3, 3.8,
1H), 4.76 (d, J ) 18.1, 1H), 4.74 (dd, J ) 9.5, 6.2, 1H), 4.63 (s,
2H), 4.43 (d, J ) 18.1, 1H), 4.09 (t, J ) 5.7, 1H), 3.59 (ddd, J
) 10.1, 8.6, 3.5, 1H), 3.45 (td, J ) 10.4, 4.4, 1H), 2.61 (dt, J )
12.6, 9.3, 1H), 2.53-2.47 (m, 2H), 2.34-2.31 (m, 1H), 1.88-
1.86 (m, 2H), 1.78-1.67 (m, 3H), 1.62-1.51 (m, 2H), 1.41-
1.28 (m, 2H); 13C NMR (125 MHz, CDCl3) δ 206.2, 165.6, 143.9,
137.4, 133.3, 129.7, 129.6, 128.5, 128.3, 128.0, 128.0, 127.9,
127.9, 127.7, 126.3, 99.2, 86.6, 82.7, 73.2, 72.9, 72.9, 69.3, 51.1,
34.5, 32.5, 30.5, 29.2, 25.7, 25.2; IR (CCl4) 1727 (s), 1719 (s)
cm-1; MS (CI, CH4) m/z 572 (M+, 2), 91 (100); TLC Rf 0.31
(hexane/EtOAc, 67/33). Anal. Calcd for C34H37NO7 (571.68): C,
71.44; H, 6.52; N, 2.45. Found: C, 71.51; H, 6.53; N, 2.52.
rel-(2R,3aS,4R,6R)-4-(Benzoyloxy)-2-(hydroxymethyl)-6-[(1S,-
2R)-(2-phenylcyclohexyl)oxy]hexahydroisoxazolo[2,3-b][1,2]-
oxa zin e (36b) a n d r el-(2S,3a S,4R,6R)-4-(Ben zoyl-oxy)-2-
(h yd r oxym et h yl)-6-[(1S,2R)-(2-p h en yl-cycloh exyl)oxy-
]h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e (36a ). [3 + 2]
Cycloa d d ition w ith Acr olein . Acrolein (440 µL, 6.59 mmol,
6.5 equiv) was added to a solution of nitronate 3 (401 mg, 1.01
mmol) and several crystals of 1,4-di-tert-butylhydroquinone in
benzene (30 mL), and the reaction mixture was maintained
at room temperature for 3.5 h. Upon concentration of the
solution in vacuo, the residue was purified by chromatography
on silica (hexane/EtOAc, 80/20, 67/33, 50/50). The major nitroso
acetal 35b (296 mg) and minor nitroso acetal 35a (158 mg)
were repurified by chromatography two additional times and
dried azeotropically with benzene to afford 251 and 97.3 mg,
respectively. The hydrated aldehyde pair were combined,
dissolved in EtOH (2 mL), and treated with NaBH4 (61.3 mg,
1.62 mmol, 2.1 equiv) in portions at 0 °C. After 1 h, the reaction
mixture was poured into water (25 mL) and extracted with
CH2Cl2 (3 × 10 mL). The extracts were washed with water
(10 mL) and brine (10 mL) and dried (Na2SO4). The residue
was purified by radial chromatography (CH2Cl2/methanol, 99/
1, 4 mm SiO2 plate). The major nitroso acetal was recrystal-
lized from TBME/pentane to provide 193 mg (42%) of 36b as
colorless, spiny needles. The minor nitroso acetal was recrys-
tallized from TBME/pentane to provide 93.6 mg (20%) of 36a
as colorless, spiny needles. Data for 1/2 mixture of 35a and
35b:1H NMR (500 MHz, CDCl3) δ 9.76 (d, J ) 1.1, 1H), 9.65
(d, J ) 1.8, 1H), 8.06 (dd, J ) 8.2, 1.1, 2H), 8.02 (dd, J ) 8.2,
1.1, 2H), 7.61-7.57 (m, 1H), 7.48-7.44 (m, 2H), 7.27-7.15 (m,
5H), 5.00 (ddd, J ) 8.4, 4.9, 3.8, 1H), 4.93 (dt, J ) 9.0, 4.9,
1H), 4.88 (ddd, J ) 10.2, 4.0, 1.8, 1H), 4.62 (ddd, J ) 9.7, 5.9,
1.1, 1H), 4.17 (t, J ) 6.0, 1H), 4.14 (t, J ) 5.5, 1H), 3.68-3.53
(m, 2H), 2.65-2.30 (m, 4H), 1.90-1.29 (m, 9H). Data for 36b:
mp 138-139 °C (TBME/pentane);1H NMR (500 MHz, CDCl3)
δ 8.05 (dd, J ) 8.3, 1.2, 2H), 7.59 (tt, J ) 7.6, 1.2, 1H), 7.46 (t,
J ) 8.1, 2H), 7.26-7.15 (m, 5H), 5.01 (ddd, J ) 8.8, 4.9, 3.9,
1H), 4.66 (tt, J ) 8.5, 2.4, 1H), 4.20 (t, J ) 5.6, 1H), 3.89 (dt,
J ) 12.7, 2.2, 1H), 3.65 (ddd, J ) 11.5, 7.6, 3.7, 1H), 3.61 (td,
J ) 10.7, 4.4, 1H), 3.57 (ddd, J ) 12.7, 10.0, 2.9, 1H), 2.66
(dd, J ) 10.0, 2.4, 1H), 2.52 (ddd, J ) 12.9, 10.3, 3.4, 1H), 2.39
(dt, J ) 12.0, 7.8, 2H), 2.30 (td, J ) 12.0, 8.8, 1H), 1.89-1.74
(m, 5H), 1.66-1.28 (m, 4H); 13C NMR (125 MHz, CDCl3) δ
165.8, 144.1, 133.3, 129.8, 129.7, 128.4, 128.1, 127.9, 126.3,
99.1, 87.0, 82.4, 73.9, 69.4, 61.9, 51.3, 34.6, 32.6, 29.9, 28.8,
25.8, 25.2; IR (KBr) 1721 (s) cm-1; MS (CI, CH4) m/z 454 (M+
+ 1, 8), 159 (100); TLC Rf 0.35 (EtOAc/hexane, 50/50). Anal.
Calcd for C26H31NO6 (453.54): C, 68.86; H, 6.89; N, 3.09.
r el -(2S ,3a S ,4R ,6R )-4-(B e n z o y lo x y )-6-[(1S ,2R )-(2-
p h e n ylcycloh e xyl)oxy]-2-[2-(p h e n ylm e t h oxy)-1-oxo-
et h yl]h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e (34a ) a n d
r el-(2R,3a S,4R,6R)-4-(Ben zoyloxy)-6-[(1S,2R)-(2-p h en yl-
cycloh exyl)oxy]-2-[2-(ph en ylm eth oxy)-1-oxoeth yl]h exah y-
d r oisoxa zolo[2,3-b][1,2]oxa zin e (34b). A solution of water
(56 µL, 3.1 mmol, 1.1 equiv) in CH2Cl2 was added dropwise
over 70 min to a solution of the Dess-Martin periodinane (1.40
g, 3.30 mmol, 1.2 equiv) and the allyl alcohol 14 (503 mg, 2.82
mmol) in CH2Cl2 (14 mL). After an additional 20 min, the
cloudy reaction mixture was diluted with Et2O (100 mL) and
filtered (Watman 1). The organic phase was washed with 1/1
(v/v) saturated aq NaHCO3/0.1 N aq sodium thiosulfate
solution (50 mL), water (50 mL), and brine (50 mL). The aq
layers were back extracted with Et2O (25 mL), and the
combined organic layers were dried (Na2SO4). The residue was
purified by rapid chromatography on silica (pentane/Et2O, 67/
33) and dried at 0.1 mmHg to provide 430 mg (87%) of ketone
16 as an oil. The ketone was not purified further but was used
directly in the next reaction. Nitronate 3 (720 mg, 1.82 mmol)
was added to a solution of ketone 16 (424 mg, 2.41 mmol, 1.3
equiv) and several crystals of 1,4-di-tert-butylhydroquinone in
benzene (19 mL), and the reaction mixture was maintained
at room temperature for 12 h. Upon concentration of the
solution in vacuo, the residue was prepurified by gradient
chromatography on silica (hexane/EtOAc, 86/14, 80/20, 67/33).
The diastereoisomers were separated by radial chromatogra-
phy (hexane/EtOAc, 87/13, 4 mm SiO2 plate) to give 795 mg
(76%) of the major nitroso acetal 34a as a foam. The minor
nitroso acetal was recrystallized from TBME/hexane to provide
140 mg (13%) of 34b as colorless, spiny needles. Data for 16:
1
bp 140 °C (1.2 × 10-5 mmHg); H NMR (400 MHz, CDCl3) δ
7.37-7.31 (m, 5H), 6.55 (dd, J ) 17.6, 10.5, 1H), 6.34 (dd, J )
17.6, 1.5, 1H), 5.83 (dd, J ) 10.7, 1.5, 1H), 4.62 (s, 2H), 4.28
(s, 2H); 13C NMR (100 MHz, CDCl3) δ 197.0, 137.1, 132.3,
129.2, 128.4, 127.9, 127.9, 73.7, 73.2; IR (neat) 1721 (s) cm-1
;
TLC Rf 0.30 (hexane/EtOAc, 80/20). Data for 34a : 1H NMR
(500 MHz, CDCl3) δ 8.02 (dd, J ) 8.2, 1.3, 2H), 7.59 (tt, J )
7.3, 1.5, 1H), 7.46 (t, J ) 7.7, 2H), 7.33-7.16 (m, 10H), 5.06

898 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
Found: C, 68.88; H, 6.90; N, 3.11. Data for 36a : mp 140-141
°C (TBME/pentane);1H NMR (500 MHz, CDCl3) δ 8.04 (dd, J
) 8.3, 1.2, 2H), 7.58 (tt, J ) 7.6, 1.2, 1H), 7.45 (t, J ) 8.1, 2H),
7.26-7.16 (m, 5H), 4.97 (dt, J ) 8.8, 4.9, 1H), 4.76-4.72 (m,
1H), 4.15 (t, J ) 5.9, 1H), 3.84 (ddd, J ) 12.5, 4.9, 2.7, 1H),
3.66 (ddd, J ) 9.3, 8.3, 4.4, 1H), 3.64 (td, J ) 11.0, 4.4, 1H),
3.60 (ddd, J ) 12.5, 8.8, 3.9, 1H), 2.52 (ddd, J ) 12.9, 10.3,
3.4, 1H), 2.40-2.36 (m, 1H), 2.36 (ddd, J ) 12.7, 8.3, 4.2, 2H),
2.26 (dt, J ) 12.0, 9.8, 1H), 1.89-1.28 (m, 10H); 13C NMR (125
MHz, CDCl3) δ 165.8, 144.2, 133.3, 129.8, 129.7, 128.4, 128.1,
127.9, 126.3, 98.8, 84.1, 82.1, 73.7, 70.1, 63.1, 51.4, 34.7, 32.6,
(ddd, J ) 15.1, 13.9, 8.1, 1H), 2.23 (dddd, J ) 15.4, 7.6, 3.9,
1.7, 1H), 1.87 (dt, J ) 11.5, 10.7, 1H); 13C NMR (100 MHz,
CDCl3) δ 175.7, 165.9, 133.2, 129.4, 129.2, 128.3, 77.9, 71.9,
61.6, 40.8, 36.6, 31.6; IR (CCl4) 1723 (s) cm-1; MS (CI, CH4)
m/z 262 (M+ + 1, 100); TLC Rf 0.17 (EtOAc). Anal. Calcd for
C
14H15NO4 (261.28): C, 64.36; H, 5.79; N, 5.36. Found: C,
64.35; H, 5.96; N, 5.34.
r el-(2R,7R,7a S)-7-(Ben zoyloxy)-2-h yd r oxyh exa h yd r o-
1H-p yr r olizid in -3-on e (37b). To a suspension of W-2 Raney
nickel (680 mg) [washed with water (3 × 10 mL) and methanol
(2 × 20 mL)] in methanol (90 mL) was added nitroso acetal
30b (425 mg, 0.883 mmol). The suspension was maintained
under an atmosphere of H2 (160 psi) with vigorous stirring
for 48 h, filtered through Celite (methanol, CH2Cl2), and
concentrated in vacuo. The crude material was purified by
chromatography on silica (EtOAc) to provide 131 mg (84%) of
the chiral auxiliary and 185 mg of lactam 37b. Recrystalliza-
tion from CH2Cl2/hexane provided 159 mg (69%) of lactam 37b
as fine, colorless needles. Data for 37b: mp 198-199 °C (CH2-
Cl2/hexane);1H NMR (400 MHz, Me2SO-d6) δ 7.98 (d, J ) 7.1,
2H), 7.67 (t, J ) 7.6, 1H), 7.53 (t, J ) 7.6, 2H), 5.74 (d, J )
5.1, 1H), 4.98 (dt, J ) 7.3, 5.1, 1H), 4.08 (m, 1H), 4.01 (q, J )
6.3 1H), 3.60 (ddd, J ) 11.7, 8.5, 4.9, 1H), 3.14 (dt, J ) 11.5,
7.6, 1H), 2.34 (ddd, J ) 15.6, 7.6, 7.3, 1H), 2.21-2.05 (m, 3H);
13C NMR (100 MHz, Me2SO-d6) δ 174.7, 165.6, 133.6, 129.4,
129.3, 128.8, 77.6, 73.0, 63.9, 40.2, 34.8, 32.2; IR (CCl4) 1703
(s), 1699 (s) cm-1; MS (CI, CH4) m/z 262 (M+ + 1, 100); TLC
Rf 0.45 (CH2Cl2/methanol, 90/10). Anal. Calcd for C14H15NO4
(261.28): C, 64.36; H, 5.79; N, 5.36. Found: C, 64.41; H, 5.82;
N, 5.39.
31.1, 29.9, 25.8, 25.3; IR (KBr) 3488 (s), 2927 (s), 1727 (s) cm-1
;
MS (CI, CH4) m/z 454 (M+ + 1, 14), 278 (100); TLC Rf 0.33
(EtOAc/hexane, 50/50). Anal. Calcd for C26H31NO6 (453.54): C,
68.86; H, 6.89; N, 3.09. Found: C, 69.05; H, 6.98; N, 3.08.
rel-(2R,3aS,4R,6R)-4-(Benzoyloxy)-2-(hydroxymethyl)-6-[(1S,-
2R)-(2-phenylcyclohexyl)oxy]hexahydroisoxazolo[2,3-b][1,2]-
oxa zin e (36b ) a n d r el-(2S,3a S,4R,6R)-4-(Ben zoyloxy)-
2-(h yd r oxym eth yl)-6-[(1S,2R)-(2-p h en ylcycloh exyl)oxy]-
h exa h yd r oisoxa zolo[2,3b][1,2]oxa zin e (36a ). Allyl alcohol
(0.5 mL, 7.4 mmol, 6 equiv) was added to a solution of
nitronate 3 (485 mg, 1.23 mmol) and several crystals of 1,4-
di-tert-butylhydroquinone in benzene (40 mL), and the reaction
mixture was maintained at 80 °C for 1.5 h. Upon concentration
of the solution in vacuo, the residue was purified by radial
chromatography (CH2Cl2/methanol, 99/1, 4 mm SiO2 plate).
The minor nitroso acetal (235 mg) was recrystallized from
TBME/pentane to provide 199 mg (36%) of 36b as colorless,
spiny needles. The major nitroso acetal (280 mg) was recrys-
tallized from TBME/pentane to provide 253 mg (45%) of 36a
as colorless, spiny needles. Data for 36b: mp 137-138 °C
(TBME/pentane);1H NMR (500 MHz, CDCl3) δ 8.05 (dd, J )
8.4, 1.3, 2H), 7.59 (tt, J ) 7.5, 1.3, 1H), 7.46 (t, J ) 7.5, 2H),
7.26-7.16 (m, 5H), 5.01 (ddd, J ) 8.6, 4.8, 3.8, 1H), 4.66 (tt, J
) 8.2, 2.2, 1H), 4.20 (t, J ) 5.5, 1H), 3.89 (dt, J ) 12.6, 2.2,
1H), 3.65 (ddd, J ) 11.2, 7.3, 3.5, 1H), 3.65-3.55 (m, 2H), 2.65
(dd, J ) 10.1, 2.2, 1H), 2.55-2.50 (m, 1H), 2.42-2.36 (m, 2H),
2.30 (td, J ) 11.5, 9.0, 1H), 1.89-1.74 (m, 5H), 1.66-1.28 (m,
4H); 13C NMR (125 MHz, CDCl3) δ 165.7, 144.0, 133.2, 129.7,
129.6, 128.3, 128.0, 127.8, 126.2, 99.1, 86.8, 82.3, 73.8, 69.3,
62.0, 51.2, 34.5, 32.5, 29.8, 28.8, 25.7, 25.1; IR (CCl4) 1724 (s)
cm-1; MS (CI, CH4) m/z 454 (M+ + 1, 1), 123 (100); TLC Rf
0.35 (EtOAc/hexane, 50/50). Anal. Calcd for C26H31NO6
(453.54): C, 68.86; H, 6.89; N, 3.09. Found: C, 68.98; H, 6.89;
N, 3.27. Data for 36a : mp 128-129 °C (TBME/pentane);1H
NMR (400 MHz, CDCl3) δ 8.04 (d, J ) 8, 2H), 7.58 (t, J ) 8,
1H), 7.45 (t, J ) 8, 2H), 7.25-7.15 (m, 5H), 4.97 (ddd, J ) 9.0,
8.6, 4.9, 1H), 4.74 (dtd, J ) 7.1, 4.5, 3.9, 1H), 4.16 (t, J ) 5.6,
1H), 3.84 (ddd, J ) 12.5, 4.9, 2.7, 1H), 3.68-3.52 (m, 3H), 2.52
(ddd, J ) 12.9, 9.5, 3.4, 1H), 2.41-2.36 (m, 1H), 2.36 (ddd, J
) 11.7, 8.5, 4.6, 1H), 2.26 (dt, J ) 12.0, 9.8, 1H), 1.89-1.30
(m, 10H); 13C NMR (100 MHz, CDCl3) δ 165.6, 144.1, 133.1,
129.7, 129.6, 128.2, 128.0, 127.8, 126.2, 98.7, 84.1, 81.9, 73.4,
70.0, 62.8, 51.2, 34.5, 32.4, 30.8, 29.7, 25.7, 25.1; IR (CCl4) 2932
(s), 1725 (s) cm-1; MS (CI, CH4) m/z 454 (M+ + 1, 17), 160
(100); TLC Rf 0.33 (EtOAc/hexane, 50/50). Anal. Calcd for
r el-(7R,7aS)-7-(Ben zoyloxy)h exah ydr o-1H-pyr r olizidin -
3-on e (38a ). Phenoxy chlorothionocarbonate (160 µL, 1.16
mmol, 2 equiv), pyridine (93 µL, 1.15 mmol, 2 equiv), and
4-(N,N-dimethylamino)pyridine (35.5 mg, 0.291 mmol, 0.5
equiv) were added to a solution of lactam 37a (150 mg, 0.574
mmol) in CH2Cl2 (7.5 mL), and the reaction mixture was
maintained at room temperature for 5.5 h. Upon concentration
in vacuo, the residue was purified by chromatography on silica
(hexane/EtOAc, 67/33) to give the solid thionocarbonate (226
mg) which was recrystallized from EtOAc/hexane. The thiono-
carbonate [162 mg, 71% (mp 151-152 °C)] was dissolved in
benzene (43 mL) and heated at reflux. A solution of tributyltin
hydride (143 µL, 0.532 mmol, 1.3 equiv) and azobisisobuty-
ronitrile (13 mg, 0.079 mmol, 0.2 equiv) in benzene (5 mL) was
added dropwise to the reaction mixture over 45 min via syringe
pump. After an additional 2.5 h at reflux, the reaction mixture
was allowed to cool to room temperature and concentrated in
vacuo in the presence of 100 mg of 1/1 SiO2-KF. Chromato-
graphic purification on silica (EtOAc/hexane, 90/10) provided
an oil (85.8 mg) which was triturated with hexanes and
recrystallized (TBME/hexane) to give 81.6 mg (58%) of lactam
38a as a white solid. Data for 38a : mp 82-83 °C (TBME/
hexane);1H NMR (500 MHz, CDCl3) δ 8.04 (dd, J ) 8.4, 1.3,
2H), 7.60 (tt, J ) 7.3, 1.3, 1H), 7.47 (t, J ) 7.7, 2H), 5.06 (dt,
J ) 7.3, 4.6, 1H), 4.00 (ddd, J ) 8.2, 7.1, 4.6, 1H), 3.90 (ddd,
J ) 12.3, 8.4, 4.4, 1H), 3.22 (dtd, J ) 11.9, 7.9, 1.1, 1H), 2.68
(dtd, J ) 16.5, 10.4, 1.1, 1H), 2.50 (dddd, J ) 13.0, 8.8, 7.1,
1.6, 1H), 2.42 (ddd, J ) 16.7, 9.3, 1.8, 1H), 2.38 (m, 1H), 2.21
(ddt, J ) 13.7, 8.1, 4.6, 1H), 1.99 (ddt, J ) 13.0, 11.5, 9.0, 1H);
13C NMR (125 MHz, CDCl3) δ 175.6, 166.2, 133.3, 129.6, 129.5,
C
26H31NO6 (453.54): C, 68.86; H, 6.89; N, 3.09. Found: C,
68.97; H, 6.89; N, 3.17.
r el-(2S,7R,7a S)-7-(Ben zoyloxy)-2-h yd r oxyh exa h yd r o-
1H-p yr r olizid in -3-on e (37a ). To a suspension of W-2 Raney
nickel (1.88 g) [washed with water (6 × 10 mL) and methanol
(3 × 15 mL)] in methanol (90 mL) was added nitroso acetal
30a (1.20 g, 2.49 mmol). The suspension was maintained under
an atmosphere of H2 (160 psi) with vigorous stirring for 48 h,
filtered through Celite (methanol, CH2Cl2), and concentrated
in vacuo. The crude material was purified by chromatography
on silica (EtOAc) to provide 437 mg (100%) of the chiral
auxiliary and 517 mg of lactam 37a . Recrystallization from
EtOAc/hexane provided 441 mg (68%) of lactam 37a as fine,
colorless needles. Data for 37a : mp 111-112 °C (EtOAc/
hexane);1H NMR (400 MHz, CDCl3) δ 8.03 (d, J ) 7.1, 2H),
7.60 (t, J ) 7.3, 1H), 7.47 (t, J ) 8.1, 2H), 5.12 (dt, J ) 6.8,
3.9, 1H), 4.60 (t, J ) 8.1, 1H), 3.97 (ddd, J ) 11.7, 8.1, 3.9,
1H), 3.81 (ddd, J ) 9.8, 5.6, 4.2, 1H), 3.77-3.52 (bs, 1H), 3.24
(dt, J ) 11.7, 7.8, 1H), 2.99 (ddd, J ) 12.5, 7.3, 6.1, 1H), 2.32
128.4, 78.1, 66.5, 40.5, 33.9, 32.6, 26.2; IR (KBr) 1703 (s) cm-1
;
MS (CI, CH4) m/z 246 (M+ + 1, 100); TLC Rf 0.15 (EtOAc/
hexane, 90/10). Anal. Calcd for C14H15NO3 (245.28): C, 68.56;
H, 6.16; N, 5.71. Found: C, 68.38; H, 5.87; N, 5.85.
r el-(7R,7aS)-7-(Ben zoyloxy)h exah ydr o-1H-pyr r olizidin -
3-on e (38b). Phenoxy chlorothionocarbonate (105 µL, 0.759
mmol, 2 equiv), pyridine (62 µL, 0.77 mmol, 2 equiv), and
4-(N,N-dimethylamino)pyridine (23.5 mg, 0.192 mmol, 0.5
equiv) were added to a solution of lactam 37b (99.1 mg, 0.379
mmol) in CH2Cl2 (5 mL) and the reaction mixture maintained
at room temperature for 5 h. Upon concentration in vacuo, the
residue was purified by chromatography on silica (hexane/
EtOAc, 67/33) to give the thionocarbonate as an oil (147 mg).
The thionocarbonate was dissolved in benzene (37 mL) and
heated at reflux. A solution of tributyltin hydride (197 µL,

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 899
0.732 mmol, 1.9 equiv) and azobisisobutyronitrile (12 mg, 0.073
mmol, 0.2 equiv) in benzene (5 mL) was added dropwise to
the reaction mixture over 35 min via syringe pump. After an
additional 2.5 h at reflux, the reaction mixture was allowed
to cool to room temperature and concentrated in vacuo in the
presence of 100 mg of 1/1 SiO2-KF. Chromatographic purifica-
tion on silica (EtOAc) provided an oil (86.8 mg) which was
triturated with hexanes and recrystallized (TBME/hexane) to
give 60.7 mg (65%) of lactam 38b as a white solid. Data for
38b: mp 82-83 °C (TBME/hexane);1H NMR (500 MHz, CDCl3)
δ 8.04 (dd, J ) 8.4, 1.3, 2H), 7.60 (tt, J ) 7.3, 1.3, 1H), 7.47 (t,
J ) 7.7, 2H), 5.06 (dt, J ) 7.3, 4.6, 1H), 4.00 (ddd, J ) 8.2,
7.1, 4.6, 1H), 3.90 (ddd, J ) 12.1, 8.4, 4.4, 1H), 3.22 (dtd, J )
11.7, 7.7, 1.1, 1H), 2.68 (dtd, J ) 16.5, 10.4, 1.1, 1H), 2.50
(dddd, J ) 13.0, 8.8, 7.1, 1.6, 1H), 2.42 (ddd, J ) 16.7, 9.3,
1.8, 1H), 2.38 (m, 1H), 2.21 (ddt, J ) 13.7, 8.1, 4.6, 1H), 1.99
(ddt, J ) 13.0, 11.5, 9.0, 1H); 13C NMR (125 MHz, CDCl3) δ
175.6, 166.2, 133.6, 129.6, 129.5, 128.4, 78.1, 66.5, 40.6, 33.9,
32.6, 26.2; IR (KBr) 1703 (s), 1698 (s), 1693 (s) cm-1; MS (CI,
CH4) m/z 246 (M+ + 1, 100); TLC Rf 0.15 (EtOAc/hexane, 90/
10). Anal. Calcd for C14H15NO3 (245.28): C, 68.56; H, 6.16; N,
5.71. Found: C, 68.56; H, 6.15; N, 5.74.
r el-(2S,3R,3a R,4R,6R)-4-(Ben zoyloxy)-6-(b u t yloxy)-3-
(tr im eth ylsilyl)h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-2-
ca r boxylic Acid Meth yl Ester (39a i) a n d r el-(2R,3R,-
3a S ,4R ,6R )-4-(B e n zo y lo x y )-6-(b u t y lo x y )-2-(t r im e t h -
ylsilyl)h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-3-ca r boxy-
lic Acid Meth yl Ester (39a ii). To a solution of 2 (303.0 mg,
1.03 mmol) in CH3CN (4.5 mL) was added 17b (340 mg, 2.15
mmol, 2.1 equiv). The solution was degassed, heated at 65 °C
(ext) for 38 h, and then concentrated and chromatographed
on silica (hexane/EtOAc, 95/5) to afford 337 mg of a 2.4/1
mixture of 39a i (51%) and 39a ii (21%) as a colorless, viscous
oil. Data for 2.4/1 mixture of 39a i and 39a ii: 1H NMR (500
MHz, CDCl3) δ 8.10-8.07 (m, 2H), 8.09 (dd, J ) 7.3, 0.6, 2H),
7.62-7.58 (m, 1H), 7.60 (t, J ) 7.4, 1H), 7.49-7.45 (m, 2H),
7.47 (t, J ) 7.9, 2H), 5.28 (dt, J ) 9.1, 4.6, 1H), 5.19 (dt, J )
8.7, 4.3, 1H), 5.13 (d, J ) 9.5, 1H), 5.01 (t, J ) 5.8, 1H), 4.94
(t, J ) 5.8, 1H), 4.50 (d, J ) 10.9, 1H), 4.10 (dd, J ) 8.3, 4.7,
1H), 3.99 (dd, J ) 13.1, 4.3, 1H), 3.99-3.94 (m, 1H), 3.96 (dt,
J ) 9.5, 6.6, 1H), 3.75 (s, 3H), 3.68 (s, 3H), 3.63 (dd, J ) 10.8,
8.1, 1H), 3.50-3.44 (m, 1H), 3.47 (dt, J ) 9.4, 6.8, 1H), 2.41
(dt, J ) 13.5, 5.3, 1H), 2.34-2.14 (m, 2H), 2.25 (dd, J ) 12.7,
9.8, 1H), 2.10-2.04 (m, 1H), 1.68-1.60 (m, 2H), 1.46-1.36 (m,
2H), 0.93 (t, J ) 7.2, 3H), 0.11 (s, 9H); 13C NMR (125 MHz,
CDCl3) δ 172.3, 170.5, 165.7, 165.4, 133.2, 133.1, 129.6, 129.5,
129.4, 128.4, 128.3, 128.2, 98.1, 98.0, 85.2, 77.8, 77.4, 73.9, 69.4,
69.1, 67.7, 67.6, 52.0, 50.4, 32.8, 31.4, 31.4, 30.2, 30.1, 19.1,
19.1, 13.7, -1.9, -3.1; IR (CHCl3) 2958 (s), 1743 (s), 1722 (s)
cm-1; MS (CI, CH4) m/z 452 (M+ + 1, 33), 378 (100); TLC Rf
0.23 (hexane/EtOAc, 85/15). Anal. Calcd for C22H33NO7Si
(451.59); C, 58.51; H, 7.37; N, 3.10; Si, 6.22. Found: C, 58.48;
H, 7.29; N, 3.24; Si, 6.01.
r el-(2R,3R,3a R,4R,6R)-4-(Ben zoyloxy)-6-(b u t yloxy)-3-
(t r im e t h y ls ily l)h e x a h y d r o is o x a zo lo [2,3-b ][1,2]o x a -
zin e-2-ca r boxylic Acid Meth yl Ester (40bi), r el-(2S,3S,-
3aR,4R,6R)-4-(Ben zoyloxy)-6-(bu tyloxy)-3-(tr im eth ylsilyl)-
h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-2-ca r boxylic Acid
Met h yl E st er (40a i), a n d r el-(2S,3R,3a S,4R,6R)-4-
(Benzoyloxy)-6-(butyloxy)-2-(trimethylsilyl)hexahydroisoxa-
zolo[2,3-b][1,2]oxa zin e-3-ca r boxylic Acid Meth yl Ester
(40a ii). To a solution of 2 (306.1 mg, 1.04 mmol) in CH3CN
(4.5 mL) was added 17a (264.0 mg, 1.67 mmol, 1.6 equiv). The
solution was degassed, heated at 65 °C (ext) for 21.5 h, and
then concentrated and chromatographed on silica MPLC
(hexane/EtOAc, 95/5) to afford 197.0 mg of a 1.5/1 mixture of
40a ii (25%) and 40a i (17%) as a colorless oil, and 166 mg (35%)
of 40bi as a white solid. Approximately 10 mg of 40a ii was
separated as the less polar component from approximately 20
mg of the mixture via preparative HPLC (hexane/EtOAc, 95/
5). A single crystal of 40bi was grown from Et2O/pentane. Data
for 1.5/1 mixture of 40a ii and 40a i: 1H NMR (500 MHz, CDCl3)
δ 8.07-8.04 (m, 2H), 7.61-7.56 (m, 1H), 7.48-7.43 (m, 2H),
5.38-5.32 (m, 1H), 5.02 (t, J ) 6.5, 1H), 4.90 (t, J ) 5.9, 1H),
4.64 (d, J ) 10.5, 1H), 4.16 (d, J ) 11.2, 1H), 3.97-3.89 (m,
2H), 3.65 (s, 3H), 3.62 (s, 3H), 3.50-3.31 (m, 2H), 2.35-2.28
(m, 1H), 2.18 (ddd, J ) 13.7, 9.3, 5.9, 1H), 2.04 (ddd, J ) 13.4,
10.4, 6.4, 1H), 1.65-1.59 (m, 2H), 1.44-1.38 (m, 2H), 0.93 (t,
J ) 7.4, 3H), 0.13 (s, 9H), 0.11 (s, 9H); 13C NMR (125 MHz,
CDCl3) δ 171.8, 171.2, 165.7, 133.3, 129.7, 129.7, 129.6, 128.4,
128.4, 97.8, 97.6, 81.9, 77.5, 75.6, 69.0, 68.0, 67.7, 67.0, 52.4,
50.1, 48.9, 31.7, 31.6, 31.4, 30.7, 19.2, 13.9, 10.6, -3.6, -4.0;
IR (CCl4) 2957 (s), 1738 (s), 1725 (s) cm-1; MS (CI, CH4) m/z
452 (M+ + 1, 2), 105 (100); TLC Rf 0.27 (hexane/EtOAc, 85/
15). Anal. Calcd for C22H33NO7Si (451.59): C, 58.51; H, 7.37;
N, 3.10; Si, 6.22. Found: C, 58.74; H, 7.46; N, 3.00; Si, 6.02.
Data for 40a ii: 1H NMR (500 MHz, C6D6) δ 8.11 (dd, J ) 8.4,
1.3, 2H), 7.11 (tt, J ) 7.3, 1.2, 1H), 7.04 (t, J ) 7.5, 2H), 5.45
(dt, J ) 8.9, 4.6, 1H), 4.51 (t, J ) 5.9, 1H), 4.28 (d, J ) 11.3,
1H), 4.06 (dd, J ) 10.3, 3.8, 1H), 3.94 (dt, J ) 9.3, 6.6, 1H),
3.45 (t, J ) 10.8, 1H), 3.29 (s, 3H), 3.27 (dt, J ) 9.3, 6.5, 1H),
2.23 (ddd, J ) 13.5, 9.3, 5.8, 1H), 2.00 (dt, J ) 13.5, 5.5, 1H),
1.57-1.51 (m, 2H), 1.42-1.30 (m, 2H), 0.84 (t, J ) 7.3, 3H),
0.14 (s, 9H). Data for 40bi: mp 150-151 °C (Et2O/pentane);
1H NMR (500 MHz, CDCl3) δ 8.11-8.09 (m, 2H), 7.60-7.56
(m, 1H), 7.46-7.42 (m, 2H), 5.37 (td, J ) 5.1, 3.0, 1H), 4.96 (t,
J ) 4.0, 1H), 4.60 (d, J ) 8.5, 1H), 4.01 (dt, J ) 9.2, 6.6, 1H),
3.81 (s, 3H), 3.64 (dd, J ) 12.4, 2.7, 1H), 3.43 (dt, J ) 9.5, 6.7,
1H), 2.33-2.27 (m, 1H), 2.30 (dd, J ) 12.2, 8.4, 1H), 1.98 (ddd,
J ) 14.5, 5.4, 3.3, 1H), 1.67-1.60 (m, 2H), 1.48-1.38 (m, 2H),
0.93 (t, J ) 7.5, 3H), 0.15 (s, 9H); 13C NMR (125 MHz, CDCl3)
δ 171.3, 165.8, 133.2, 129.9, 129.8, 128.3, 98.5, 85.0, 74.6, 68.7,
67.7, 52.6, 31.7, 30.1, 28.9, 19.3, 13.9, -2.6; IR (CCl4) 2956
(s), 1737 (s) cm-1; MS (CI, CH4) m/z 452 (M+ + 1, 11), 105
(100); TLC Rf 0.20 (hexane/EtOAc, 85/15). Anal. Calcd for
C
22H33NO7Si (451.59): C, 58.51; H, 7.37; N, 3.10; Si, 6.22.
Found: C, 58.37; H, 7.55; N, 2.95; Si, 6.41.
r el-(2R,3R,3a S,4R,6R)-4-(Ben zoyloxy)-6-(b u t yloxy)-2-
m eth ylh exah ydr oisoxazolo[2,3-b][1,2]oxazin e-3-car boxylic
Acid E t h yl E st er (41a ii) a n d r el-(2S,3R,3a S,R,6R)-4-
(B e n zo y lo x y )-6-(b u t y lo x y )-3-m e t h y lh e x a h y d r o is o x -
a zolo[2,3-b][1,2]oxa zin e-2-ca r boxylic Acid Eth yl Ester
(41a i). To a solution of 2 (273.7 mg, 0.933 mmol) in CH3CN
(5.0 mL) was added 19 (212.7 mg, 1.86 mmol, 2.0 equiv). The
solution was degassed, heated at 65 °C (ext) for 22 h, and then
concentrated and chromatographed on silica MPLC (hexane/
EtOAc, 95/5) to afford 175.4 mg (46%) of 41a ii as a white
crystalline solid (recrystallized from Et2O/hexane), and 111.0
mg (29%) of 41a i as a white crystalline solid (recrystallized
from Et2O/hexane). Data for 41a ii: mp 83-84 °C (Et2O/hexane);
1H NMR (500 MHz, CDCl3) δ 8.03 (d, J ) 7.9, 2H), 7.55 (t, J
) 7.2, 1H), 7.42 (d, J ) 7.9, 2H), 5.19 (dt, J ) 9.3, 4.7, 1H),
5.07 (dq, J ) 9.7, 6.5, 1H), 4.96 (t, J ) 5.9, 1H), 4.21 (dd, J )
8.6, 4.6, 1H), 4.17 (q, J ) 7.1, 2H), 3.93 (dt, J ) 9.3, 6.7, 1H),
3.57 (t, J ) 9.3, 1H), 3.44 (dt, J ) 9.4, 6.8, 1H), 2.28 (dt, J )
14.0, 5.4, 1H), 2.21 (ddd, J ) 14.0, 9.0, 5.6, 1H), 1.64-1.58
(m, 2H), 1.43-1.35 (m, 2H), 1.26 (d, J ) 6.5, 3H), 1.23 (t, J )
7.2, 3H), 0.90 (t, J ) 7.4, 3H); 13C NMR (125 MHz, CDCl3) δ
170.3, 165.9, 133.3, 129.8, 129.6, 128.4, 98.0, 80.8, 76.3, 69.8,
67.8, 61.3, 51.9, 31.6, 30.3, 19.3, 15.2, 14.2, 13.9; IR (CHCl3)
1724 (s) cm-1; MS (CI, CH4) m/z 408 (M+ + 1, 3), 184 (100);
TLC Rf 0.23 (hexane/EtOAc, 85/15). Anal. Calcd for C21H29
-
NO7 (407.46): C, 61.90; H, 7.17; N, 3.44. Found: C, 61.92; H,
7.30; N, 3.40. Data for 41a i: mp 75.5-76.5 °C (Et2O/hexane);
1H NMR (500 MHz, CDCl3) δ 8.08 (d, J ) 7.6, 2H), 7.59 (t, J
) 7.4, 1H), 7.45 (d, J ) 7.8, 2H), 5.35 (dt, J ) 6.8, 4.1, 1H),
5.07 (d, J ) 9.6, 1H), 4.98 (t, J ) 4.8, 1H), 4.24 (q, J ) 7.1,
2H), 3.96 (dt, J ) 9.3, 6.7, 1H), 3.52 (dd, J ) 10.2, 3.5, 1H),
3.45 (dt, J ) 9.4, 6.7, 1H), 3.04 (tq, J ) 9.9, 7.0, 1H), 2.27 (dt,
J ) 14.2, 5.3, 1H), 2.08 (ddd, J ) 14.2, 7.0, 4.3, 1H), 1.66-
1.60 (m, 2H), 1.45-1.37 (m, 2H), 1.29 (t, J ) 7.1, 3H), 1.15 (d,
J ) 7.0, 3H), 0.92 (t, J ) 7.4, 3H); 13C NMR (125 MHz, CDCl3)
δ 168.9, 165.8, 133.3, 129.8, 129.7, 128.4, 98.4, 83.8, 76.8, 68.1,
67.6, 61.3, 39.0, 31.6, 29.8, 19.3, 14.2, 13.8, 12.2; IR (CHCl3)
1747 (s), 1719 (s) cm-1; MS (CI, CH4) m/z 408 (M+ + 1, 17),
184 (100); TLC Rf 0.14 (hexane/EtOAc, 85/15). Anal. Calcd for
C
21H29NO7 (407.46): C, 61.90; H, 7.17; N, 3.44. Found: C,
61.83; H, 7.25; N, 3.38.
r el-(2S,3R,3a S,4R,6R)-2-(Acetyloxy)-4-(ben zoyloxy)-6-
(bu tyloxy)h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-3-ca r -

900 J . Org. Chem., Vol. 64, No. 3, 1999
Denmark et al.
1
boxylic Acid Eth yl Ester (42a ii). To a solution of nitronate
2 (78.8 mg, 0.269 mmol) in CH2Cl2 (450 µL) was added 20b
(437.9 mg, 2.77 mmol, 10.3 equiv). This solution was heated
at 69-70 °C (internal) for 3 h, and then concentrated. The
residue was chromatographed on silica (hexane/EtOAc, 80/20)
and crystallized from EtOAc/hexane. The solid was recrystal-
lized (EtOAc/hexane) to afford 65.1 mg (54%) of 42a ii as white
crystals. Data for 42a ii: mp 96.5-97.5 °C (hexane/EtOAc); 1H
NMR (400 MHz, CDCl3) δ 8.07 (d, J ) 8.2, 2H), 7.58 (t, J )
7.5, 1H), 7.45 (t, J ) 7.7, 2H), 6.89 (d, J ) 6.3, 1H), 5.38 (dt,
J ) 9.5, 4.7, 1H), 4.92 (t, J ) 6.1, 1H), 4.26 (dd, J ) 10.9, 4.5,
1H), 4.13 (q, J ) 7.1, 2H), 3.92 (dt, J ) 9.4, 6.7, 1H), 3.85 (dd,
J ) 11.1, 6.3, 1H), 3.45 (dt, J ) 9.4, 6.7, 1H), 2.31 (ddd, J )
14.0, 6.1, 4.9, 1H), 2.19 (ddd, J ) 13.8, 9.6, 6.0, 1H), 2.01 (s,
3H), 1.64-1.57 (m, 2H), 1.43-1.33 (m, 2H), 1.17 (t, J ) 7.0,
3H), 0.90 (t, J ) 7.2, 3H); 13C NMR (100 MHz, CDCl3) δ 168.7,
166.2, 165.7, 133.3, 129.8, 129.5, 128.4, 98.2, 98.2, 70.9, 68.1,
68.0, 61.7, 51.1, 31.5, 30.5, 20.6, 19.2, 14.0, 13.8; IR (CHCl3)
1721 (m) cm-1; MS (CI, CH4) m/z 452 (M+ + 1, 3), 270 (100);
hexane); H NMR (500 MHz, CDCl3) δ 8.06 (d, J ) 7.3, 2H),
7.55 (t, J ) 7.4, 1H), 7.42 (t, J ) 7.8, 2H), 5.54 (ddd, J ) 5.9,
4.9, 3.1, 1H), 5.20 (d, J ) 5.6, 1H), 4.94 (t, J ) 4.3, 1H), 4.00
(dd, J ) 10.2, 5.7, 1H), 3.94 (dt, J ) 9.5, 6.5, 1H), 3.91 (dd, J
) 10.0, 3.0, 1H), 3.81 (s, 3H), 3.74 (s, 3H), 3.39 (dt, J ) 9.2,
6.6, 1H), 2.26 (dt, J ) 14.5, 5.0, 1H), 2.08 (ddd, J ) 14.6, 5.8,
3.5, 1H), 1.61-1.55 (m, 2H), 1.41-1.34 (m, 2H), 0.88 (t, J )
7.3, 1H); 13C NMR (125 MHz, CDCl3) δ 170.6, 170.2, 165.6,
133.2, 129.7, 129.6, 128.2, 98.7, 83.4, 75.1, 67.8, 67.7, 52.9, 52.8,
49.1, 31.5, 28.6, 19.1, 13.8; IR (CHCl3) 1744 (s), 1724 (s) cm-1
;
MS (CI, CH4) m/z 438 (M+ + 1, 4), 214 (100); TLC Rf 0.13
(hexane/EtOAc, 80/20). Anal. Calcd for C21H27NO9 (437.45): C,
57.66; H, 6.22; N, 3.20. Found: C, 57.76; H, 6.28; N, 3.35.
(2S ,3R ,3a R ,4R ,6R )-4-(B e n zo y lo x y )-2-{2-[d im e t h y l-
(1,1,2-t r im et h ylp r op yl)silyloxy]-1-oxoet h yl}-3-(d im et h -
ylp h en ylsilyl)-6-[(1S,2R)-(2-p h en ylcycloh exyl)oxy]h exa -
h yd r oisoxa zolo[2,3-b][1,2]oxa zin e (45a i) a n d (2R,3R,3a S-,
4R,6R)-4-(Ben zoyloxy)-3-{2-[d im et h yl-(1,1,2-t r im et h yl-
p r op yl)silyloxy]-1-oxoet h yl}-2-(d im et h ylp h en ylsilyl)-6-
[(1S,2R)-(2-ph en ylcycloh exyl)oxy]h exah ydr oisoxazolo[2,3-
b][1,2]oxa zin e (45a ii). Nitronate (+)-3 (1.06 g, 2.68 mmol)
was added to a solution of ketone 27 (1.17 g, 3.21 mmol, 1.2
equiv) in benzene (25 mL), and the reaction mixture was
maintained at room temperature for 12 h. Upon concentration
of the solution in vacuo, the residue was purified by chroma-
tography on silica (pentane/Et2O, 90/10, 70/30) to provide 1.96
g (97%) of a 26/1 mixture of (-)-45a i and 45a ii as a colorless
foam. Data for 26/1 mixture of (-)-45a i and 45a ii: 1H NMR
(500 MHz, CDCl3) δ 8.01 (dd, J ) 8.4, 1.3, 2H), 7.59 (tt, J )
7.5, 1.3, 1H), 7.47 (t, J ) 7.9, 2H), 7.36 (dd, J ) 8.1, 1.3, 2H),
7.30 (tt, J ) 7.5, 1.3, 1H), 7.21 (t, J ) 7.5, 4H), 7.15 (td, J )
7.9, 1.3, 3H), 5.11 (d, J ) 9.5, 1H), 4.68 (d, J ) 11.0, 1H), 4.62
(ddd, J ) 8.8, 4.6, 4.2, 1H), 4.16 (t, J ) 5.7, 1H), 4.13 (d, J )
18.8, 1H), 4.08 (d, J ) 18.8, 1H), 3.73 (dd, J ) 12.4, 4.0, 1H),
3.59 (td, J ) 10.4, 4.2, 1H), 2.51 (ddd, J ) 13.0, 10.2, 3.5, 1H),
2.42-2.38 (bm, 1H), 2.30 (dd, J ) 12.6, 9.9, 1H), 1.86 (bd, J )
12.4, 2H), 1.75 (bd, J ) 13.0, 1H), 1.68-1.40 (m, 6H), 1.31 (tt,
J ) 12.8, 3.5, 1H), 0.88 (d, J ) 6.8, 6H), 0.84 (s, 6H), 0.39 (s,
3H), 0.37 (s, 3H), 0.05 (s, 3H), 0.03 (s, 3H); 13C NMR (125 MHz,
CDCl3) δ 205.9, 165.0, 144.1, 137.3, 133.7, 133.2, 129.8, 129.7,
129.4, 129.4, 128.1, 128.0, 127.7, 126.3, 99.2, 88.2, 82.5, 74.6,
69.3, 68.2, 51.2, 34.7, 34.0, 32.8, 32.0, 29.7, 25.8, 25.2, 20.2,
20.2, 18.5, 18.4, -2.5, -3.4, -3.5, -3.6; IR (CCl4) 1720 (m)
cm-1; MS (FAB) m/z 759 (M+ + 1, 1), 105 (100); [R]2D3 -32.3°
(CHCl3, c ) 1.07); TLC Rf 0.49 (hexane/EtOAc, 80/20). Anal.
Calcd for C43H59NO7Si2 (758.12): C, 68.13; H, 7.84; N, 1.85;
Si, 7.41. Found: C, 68.02; H, 7.87; N, 1.90; Si, 7.27.
TLC Rf 0.16 (hexane/EtOAc, 80/20). Anal. Calcd for C22H29
-
NO9 (451.47): C, 58.53; H, 6.47; N, 3.10. Found: C, 58.52; H,
6.58; N, 3.07.
Isom er s of r el-(3aS,4R,6R)-2-(Acetyloxy)-4-(ben zoyloxy)-
6-(bu tyloxy)h exah ydr oisoxazolo[2,3-b][1,2]oxazin e-3-car -
boxylic Acid Eth yl Ester an d r el-(3aS,4R,6R)-3-(Acyloxy)-
4-(b e n zoyloxy)-6-(b u t yloxy)h e xa h yd r oisoxa zolo[2,3-
b][1,2]oxa zin e-2-ca r boxylic Acid Eth yl Ester (43a i, a ii, bi,
bii). To a solution of nitronate 2 (305.7 mg, 1.04 mmol) in CH3-
CN (4.5 mL) was added 20a (344.8 mg, 2.18 mmol, 2.1 equiv).
This solution was degassed, then heated at 65 °C (ext) for 24
h, and then concentrated and chromatographed on silica
(hexane/EtOAc, 85/15) to afford 365 mg (78%) of a mixture
43a i, a ii, bi, bii as a light brown oil. Data for 43a i, a ii, bi, bii:
1H NMR (500 MHz, CDCl3) δ 8.11-8.04 (m, 2H), 7.62-7.58
(m, 1H), 7.49-7.44 (m, 2H), 6.99 (d, J ) 2.3, 0.36H), 6.79 (d,
J ) 4.5, 0.17H), 5.79 and 5.76 (dt, J ) 10.7, 5.4 and dd, J )
7.2, 3.4, 0.66H), 5.48-5.45 (m, 0.42H), 5.41 (dd, J ) 8.5, 2.6,
0.11H), 5.06-5.01 (m, 0.81H), 4.94 (d, J ) 3.6, 0.17H), 4.80
(d, J ) 3.3, 0.03H), 4.30-4.07 (m, 2.47H), 4.04-3.90 (m,
1.76H), 3.73-3.67 (m, 0.43H), 3.59 (dd, J ) 7.5, 2.1, 0.43H),
3.52-3.45 (m, 0.96H), 2.97 (dd, J ) 15.9, 8.3, 0.13H), 2.91 (d,
J ) 3.2, 0.11H), 2.49 (dt, J ) 13.4, 6.6, 0.49H), 2.43-2.35 (m,
0.57H), 2.19 (s, 0.63H), 2.17 (s, 0.72H), 2.11 (s, 0.32H), 2.09
(s, 1.15H), 2.06-2.00 (m, 0.68H), 1.66-1.58 (m, 2.27H), 1.45-
1.37 (m, 2.00H), 0.96-0.91 (m, 2.63H); IR (CCl4) 2961 (s), 1741
(s), 1737 (s), 1727 (s) cm-1; MS (CI, CH4) m/z 452 (M+ + 1, 3),
270 (100); TLC Rf 0.08 (hexane/EtOAc, 85/15). Anal. Calcd for
C
22H29NO9 (451.47): C, 58.53; H, 6.47; N, 3.10. Found: C,
r el-(2S,3R,3aR,4R,6R)-4-(Ben zoyloxy)-3-(dim eth ylph en -
ylsilyl)-6-[(1S,2R)-(2-phenylcyclohexyl)oxy]-2-[2-(phenylmethoxy)-1-
oxoeth yl]h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e (46a i)
and rel-(2R,3R,3aS,4R,6R)-4-(Benzoyloxy)-2-(dimethylphenyl-
silyl)-6-[(1S ,2R )-(2-p h e n ylcycloh e xyl)oxy]-3-[2-(p h e n -
ylmethoxy)-1-oxoethyl]hexahydroisoxazolo[2,3-b][1,2]oxazine
(46a ii). Nitronate 3 (469 mg, 1.19 mmol) was added to
a solution of ketone 28 (443 mg, 1.43 mmol, 1.2 equiv)
in benzene (12 mL) and the reaction mixture was
maintained at room temperature for 5 h. Upon concentration
of the solution in vacuo, the residue was purified by
chromatography on silica (pentane/Et2O, 80/20) to provide
806 mg (96%) of 46a i as a colorless foam and 14.5 mg,
(1.7%) of the sensitive minor nitroso acetal 46a ii. Data for
46a ii: 1H NMR (500 MHz, CDCl3) δ 7.90 (d, J ) 8.3,
2H), 7.55 (t, J ) 7.6, 1H), 7.47 (d, J ) 7.8, 2H), 7.40
(t, J ) 8.1, 2H), 7.37-7.15 (m, 14H), 4.83 (dt, J ) 10.5,
4.6, 1H), 4.70 (d, J ) 10.7, 1H), 4.18 (d, J ) 12.0, 1H),
4.16 (t, J ) 6.6, 1H), 4.12 (d, J ) 12.0, 1H), 3.87 (dd,
58.29; H, 6.61; N, 3.06.
r el-(2S ,3S ,3a S ,4R ,6R )-4-(Be n zoyloxy)-6-(b u t yloxy)-
h exa h yd r oisoxa zolo[2,3-b][1,2]oxa zin e-2,3-d ica r b oxyl-
ic Acid Dim eth yl Ester (44a i) a n d r el-(2R,3R,3a S,4R,6R)-
4-(Be n zoyloxy)-6-(b u t yloxy)h e xa h yd r oisoxa zolo[2,3-
b][1,2]oxazin e-2,3-dicar boxylic Acid Dim eth yl Ester (44bi).
To a solution of 2 (266.3 mg, 0.908 mmol) in benzene (20 mL)
was added dimethyl fumarate (520.8 mg, 3.613 mmol, 4.0
equiv). This clear solution was degassed, stirred at room
temperature for 23 h, and then concentrated and separated
on silica MPLC (hexane/EtOAc, 85/15) to afford 162.1 mg (41%)
of 44a i as a light yellow oil and 163.1 mg (41%) of 44bi as a
white crystalline solid. A single crystal of 44bi was grown from
Et2O/hexane. Data for 44a i: 1H NMR: (400 MHz, CDCl3) δ
8.04 (dd, J ) 8.4, 1.3, 2H), 7.58 (tt, J ) 7.5, 1,4, 1H), 7.44 (t,
J ) 7.5, 2H), 5.62 (d, J ) 6.8, 1H), 5.47 (dt, J ) 9.7, 5.0, 1H),
4.96 (t, J ) 6.0, 1H), 4.15 (dd, J ) 10.3, 5.3, 1H), 3.91 (dt, J )
9.5, 6.8, 1H), 3.78 (s, 3H), 3.78 (dd, J ) 10.2, 7.0, 1H), 3.69 (s,
3H), 3.45 (dt, J ) 9.5, 6.6, 1H), 2.22 (ddd, J ) 13.7, 6.0, 5.1,
1H), 2.01 (ddd, J ) 13.7, 9.1, 5.9, 1H), 1.63-1.56 (m, 2H),
1.43-1.33 (m, 2H), 0.90 (t, J ) 7.4, 1H); 13C NMR (100 MHz,
CDCl3) δ 169.6, 168.7, 165.4, 133.3, 129.6, 129.3, 128.4, 97.8,
82.7, 73.9, 67.9, 66.3, 52.8, 52.8, 48.8, 31.4, 30.5, 19.2, 13.8;
IR (CHCl3) 1743 (s) cm-1; MS (CI, CH4) m/z 438 (M+ + 1, 8),
145 (100); TLC Rf 0.15 (hexane/EtOAc, 80/20). Anal. Calcd for
J
J
) 8.3, 4.6, 1H), 3.81 (dd, J ) 10.7, 8.5, 1H), 3.60 (td,
) 10.3, 4.2, 1H), 3.44 (d, J ) 17, 1H), 3.40 (d, J
)
17, 1H), 2.52 (ddd, J ) 12.9, 10.3, 3.4, 1H), 2.37 (dd, J
) 12.7, 3.4, 1H), 1.92-1.86 (m, 3H), 1.77 (bd, J ) 12.9,
1H), 1.68-1.42 (m, 6H), 1.37-1.28 (m, 1H), 0.38 (s, 3H),
0.36 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 98.4, 82.1,
78.8, 78.6, 75.7, 72.4, 69.5, 52.8, 51.1. Data for 46a i: 1H
NMR (500 MHz, CDCl3) δ 8.03 (d, J ) 7.6, 2H), 7.59
C
21H27NO9 (437.45): C, 57.66; H, 6.22; N, 3.20. Found: C,
57.71; H, 6.41; N, 2.92. Data for 44bi: mp 91-92 °C (Et2O/

[3 + 2] Cycloaddition of Cyclic Nitronates
J . Org. Chem., Vol. 64, No. 3, 1999 901
(t, J ) 7.3, 1H), 7.47 (t, J ) 7.6, 2H), 7.39-7.13 (m, 15H),
5.06 (d, J ) 9.8, 1H), 4.69 (ddd, J ) 8.5, 4.4, 4.2, 1H),
4.40 (s, 2H), 4.15 (t, J ) 5.6, 1H), 3.94 (d, J ) 18.6,
1H), 3.89 (d, J ) 18.6, 1H), 3.71 (dd, J ) 12.7, 3.9, 1H),
3.59 (td, J ) 10.5, 4.4, 1H), 2.53-2.48 (m, 1H), 2.39-2.36
(m, 1H), 2.31 (dd, J ) 12.7, 9.8, 1H), 1.87-1.84 (m, 2H),
1.76-1.73 (m, 1H), 1.66 (ddd, J ) 13.9, 5.6, 5.4, 1H),
1.63-1.23 (m, 5H), 0.41 (s, 3H), 0.37 (s, 3H); 13C NMR
(125 MHz, CDCl3) δ 205.0, 165.0, 144.1, 137.1, 137.0, 133.7,
133.3, 129.8, 129.7, 129.4, 128.4, 128.4, 128.1, 128.0, 128.0,
127.9, 127.7, 126.3, 99.2, 88.6, 82.5, 74.6, 73.8, 73.1, 69.2,
51.2, 34.7, 32.7, 32.1, 29.6, 25.8, 25.2, -2.7, -3.2; IR (CCl4)
1725 (m) cm-1; MS (FAB) m/z 706 (M+ + 1, 6), 104 (100);
TLC Rf 0.30 (hexane/EtOAc, 80/20). Anal. Calcd for
C42H47NO7Si (705.93): C, 71.46; H, 6.71; N, 1.98; Si, 3.98.
Found: C, 71.19; H, 6.70; N, 1.93; Si, 4.27.
1, 7), 168 (100); TLC Rf 0.49 (hexane/EtOAc, 65/35). Anal.
Calcd for C19H25NO6 (363.41): C, 62.80; H, 6.93; N, 3.85.
Found: C, 62.91; H, 6.86; N, 3.93. Data for 48b: 1H NMR (500
MHz, C6D6) δ 8.06 (d, J ) 7.5, 2H), 7.13 (t, J ) 7.5, 1H), 7.05
(t, J ) 7.5, 2H), 5.90 (d, J ) 5.8, 1H), 5.04 (dt, J ) 12.4, 5.2,
1H), 4.41 (t, J ) 7.2, 1H), 4.24 (dt, J ) 10.8, 7.3, 1H), 3.90 (dt,
J ) 9.4, 6.7, 1H), 3.78 (t, J ) 8.3, 1H), 3.27 (dt, J ) 9.3, 6.6,
1H), 3.18 (dd, J ) 8.4, 5.3, 1H), 2.57 (dddd, J ) 11.0, 8.3, 5.8,
2.5, 1H), 2.43 (td, J ) 12.6, 7.8, 1H), 2.09 (ddd, J ) 12.5, 6.7,
5.5, 1H), 1.92 (dd, J ) 13.4, 6.7, 1H), 1.65 (dtd, J ) 13.2, 11.0,
9.0, 1H), 1.57-1.51 (m, 2H), 1.39-1.30 (m, 2H), 0.84 (t, J )
7.3, 3H); 13C NMR (125 MHz, C6D6) δ 165.5, 133.2, 130.1,
129.9, 128.5, 115.7, 97.6, 74.4, 67.8, 67.4, 67.3, 45.2, 31.7, 31.6,
27.2, 19.3, 13.8; TLC Rf 0.51 (hexane/EtOAc, 65/35). Data for
48a : 1H NMR (500 MHz, C6D6) δ 8.22 (d, J ) 8.3, 2H), 7.13-
7.07 (m, 3H), 6.31 (d, J ) 5.5, 1H), 5.13 (dt, J ) 7.8, 4.0, 1H),
4.62 (t, J ) 5.0, 1H), 4.03 (dt, J ) 9.3, 6.6, 1H), 3.57 (t, J )
8.4, 1H), 3.46 (ddd, J ) 12.6, 8.7, 4.4, 1H), 3.31 (dt, J ) 9.4,
6.5, 1H), 3.27 (dd, J ) 8.1, 3.8, 1H), 2.59 (td, J ) 8.0, 5.7, 1H),
2.06 (ddd, J ) 13.8, 7.6, 4.8, 1H), 1.80 (dt, J ) 14.0, 5.2, 1H),
1.60-1.54 (m, 2H), 1.46 (tt, J ) 12.6, 8.1, 1H), 1.43-1.33 (m,
2H), 1.15 (dd, J ) 12.7, 4.7, 1H), 0.86 (t, J ) 7.3, 3H); 13C
NMR: (125 MHz, C6D6) δ 165.8, 133.1, 130.5, 130.1, 128.4,
114.5, 97.9, 76.4, 70.0, 67.8, 67.0, 46.4, 31.6, 30.5, 29.5, 19.5,
13.8; TLC Rf 0.47 (hexane/EtOAc, 65/35).
r el-(3a S,3bS,4R,6R,9a R)-4-(Ben zoyloxy)-6-(bu tyloxy)-
octa h yd r o(fu r o[1,2-d ]isoxa zolo)[4,5-b][1,2]oxa zin e (48b)
an d r el-(3aR,3bS,4R,6R,9aS)-4-(Ben zoyloxy)-6-(bu tyloxy)-
octa h yd r o(fu r o[1,2-d ]isoxa zolo)[4,5-b][1,2]oxa zin e (48a ).
To a solution of nitronate 2 (297.0 mg, 1.0 mmol) in CH3CN
(5.0 mL) was added 2,3-dihydrofuran (750 µL, 695 mg, 10.0
mmol, 10.0 equiv). This was heated at 65 °C (ext) for 9 h and
then concentrated and chromatographed on silica (hexane/
EtOAc, 80/20) to afford 302.8 mg (82%) of a 17/1 mixture of
48a and 48b as a pale yellow oil. A portion of the mixture was
separated by preparative HPLC (hexane/EtOAc, 90/10) to
afford 7.0 mg of 48b and 203.8 mg of 48a . Data for 17/1
mixture of 48a and 48b: 1H NMR (500 MHz, CDCl3) δ 8.09
(d, J ) 7.7, 2H), 8.04 (d, J ) 7.5, 2H), 7.60 (t, J ) 7.5, 1H),
7.47 (t, J ) 7.7, 2H), 6.31 (d, J ) 5.4, 1H), 6.15 (d, J ) 5.5,
1H), 5.36 (dt, J ) 8.1, 4.2, 1H), 4.98 (t, J ) 5.2, 1H), 4.92 (t, J
) 7.0, 1H), 4.09 (t, J ) 8.4, 1H), 3.96 (dt, J ) 9.4, 6.7, 1H),
3.88 (ddd, J ) 12.3, 8.9, 4.9, 1H), 3.50 (dd, J ) 8.0, 4.2, 1H),
3.47 (dt, J ) 9.4, 6.7, 1H), 3.17 (td, J ) 7.9, 5.5, 1H), 2.30 (dt,
J ) 14.2, 5.2, 1H), 2.18-2.08 (m, 2H), 1.87 (dd, J ) 12.8, 4.7,
1H), 1.67-1.61 (m, 2H), 1.46-1.39 (m, 2H), 0.93 (t, J ) 7.5,
3H); 13C NMR (125 MHz, CDCl3) δ 166.0, 133.4, 129.8, 128.4,
114.5, 98.2, 76.4, 69.6, 67.8, 67.3, 46.8, 31.6, 30.9, 29.8, 19.3,
Ack n ow led gm en t. We are grateful to the National
Institutes of Health (GM-30938) for financial support.
We thank Dr. Scott R. Wilson for help with crystal
structure refinement and Dr. Mark E. Schnute for the
characterization of compounds 9 and 11.
Su p p or tin g In for m a tion Ava ila ble: General experimen-
1
tal, a full list of H NMR and 13C NMR listings with assign-
ments, IR absorbances, and MS fragments for all new com-
pounds, as well as coordinates for all calculated structures.
This material is available free of charge via the Internet at
http://pubs.acs.org.
13.9; IR (CHCl3) 1720 (m) cm-1; MS (CI, CH4) m/z 364 (M+
+
J O9818374