X-ray crystallographic and 13C nuclear magnetic resonance studies of 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-2-ones derived from ephedrine and pseudoephedrine

Article abstract of DOI:10.1016/S0040-4020(01)00967-X

3,4,5,6-Tetrahydro-2H-1,3,4-oxadiazin-2-ones derived from (1R,2S)-ephedrine and (1S,2S)-pseudoephedrine have been synthesized and their conformational properties have been examined. The ephedrine heterocycles 5-7a appear to favor one set of equilibrating

Full text of DOI:10.1016/S0040-4020(01)00967-X

TETRAHEDRON
Pergamon
Tetrahedron 57 32001) 9789±9798
13
X-Ray crystallographic and C nuclear magnetic resonance
studies of 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-2-ones derived
from ephedrine and pseudoephedrineq
Shawn R. Hitchcock,a,p George P. Nora,a David M. Casper,a Michael D. Squire,a
Christopher D. Maroules,a Gregory M. Ferrence,b Lisa F. Szczepurab and Jean M. Standarda
aDepartment of Chemistry, Illinois State University, Normal, IL 61790-4160, USA
bDepartment of Chemistry, Structure Determination Laboratory, Illinois State University, Normal, IL 61790-4160, USA
Received 21 March 2001; accepted 26 September 2001
AbstractÐ3,4,5,6-Tetrahydro-2H-1,3,4-oxadiazin-2-ones derived from 31R,2S)-ephedrine and 31S,2S)-pseudoephedrine have been synthe-
sized and their conformational properties have been examined. The ephedrine heterocycles 5±7a appear to favor one set of equilibrating
conformers while the pseudoephedrine heterocycles 5±7b exist as multiple conformers at room temperature. The observed conformational
behavior of these heterocycles is attributed to allylic strain and a gauche effect arising from the torsional energy barrier between the lone pair
electrons of the N3- and N4-nitrogens. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
spectroscopy and semiempirical calculations suggest that
the ephedrine and pseudoephedrine heterocycles are both
involved in complex conformational equilibria. The
ephedrine heterocycle equilibrium appears to favor one set
of equilibria whereas the pseudoephedrine heterocycle equi-
librium appears to have multiple conformers present at
ambient temperatures. Ultimately, understanding the
fundamental conformational dynamics of this ring system
would enhance any investigations into their use as chiral
auxiliaries 3Fig. 1).
3,4,5,6-Tetrahydro-2H-1,3,4-oxadiazin-2-ones 31) represent
a class of heterocycles that have received little attention in
terms of their structural analysis and application in organic
synthesis. In 1968, Trepanier disclosed one of the ®rst
reports on 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-2-ones in
the synthesis of candidates for central nervous system
stimulant activity.1 In regard to their use in synthetic trans-
formations, Husson and co-workers2 successfully applied
this heterocycle family in asymmetric dipolar cyclo-
additions2a,b and in diastereoselective alkylations.2c There
is the possibility that these heterocycles may be further
exploited in asymmetric synthesis. This heterocycle family
can be considered to be nitrogen-homologs to the Evans's
oxazolidinone chiral auxiliaries which have enjoyed much
success since their initial development during the 1980s.3
We report on the preparation of a series of 3,4,5,6-tetra-
hydro-2H-1,3,4-oxadiazin-2-ones based on 31R,2S)-
ephedrine and 31S,2S)-pseudoephedrine and conformational
analysis of these heterocycles. The data collected from
X-ray crystallography, variable temperature 13C NMR
Figure 1.
2. Results and discussion
q Part of this work was presented at the 222nd National Meeting of the
American Chemical Society, Chicago, IL, August 2001, Abstract No.
417.
2.1. Preparation of 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-
2-ones 1a and b
Keywords: 3,4,5,6-tetrahydro-2H-oxadiazin-2-one; heterocycle; X-ray
crystallography; allylic strain; conformation.
Corresponding author. Tel.: 11-309-438-7854; fax: 11-309-438-5538;
e-mail: hitchcock@xenon.che.ilstu.edu
Both ephedrine 32a) and pseudoephedrine 32b) readily
formed N-nitrosamines 3a and b in yields of 95 and 97%,
respectively, when treated with NaNO2/HCl 3Scheme 1).4,5
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020301)00967-X

9790
S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
Scheme 1. Synthesis of 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-2-ones 5±7a and 5±7b.
The N-nitrosamines were reduced with lithium aluminum
hydride in THFto afford the corresponding b-hydrazino-
alcohols 4a and b.2b,6 The b-hydrazino-alcohols were
treated with 1,10-carbonyldiimidazole 3CDI)2b,7 to afford
the 3,4,5,6-tetrahydro-2H-1,3,4-oxadiazin-2-ones 1a and
1b in 74% yield in each case. The heterocycles were treated
with 4-dimethylaminopyridine 3DMAP), triethylamine and
an acyl chloride [3RCOCl; RˆCH3 35a, b); RˆCH2CH3 36a,
b); RˆCH2Ph 37a, b)] to generate the acylated oxadiazi-
nones 5±7a and 5±7b. The acylated heterocycles were
isolated in fair to excellent yield 343±95%) after puri®cation
by chromatography and recrystallization.
in conjunction with a gauche effect wherein the nitrogen
substituents adopt positions that diminish any interaction
between the lone pairs.10,11 As a consequence of this order-
ing, the N3-acyl group may be considerably distorted from
the ideal planarity with the ring system.12
As illustrated by the ORTEP diagram, the phenylacetyl
moiety lies in the plane of the heterocycle with the imide
carbonyls oriented in the same direction.13 This was
unexpected as the related oxazolidin-2-one system typically
exhibits the more stable dipole minimized anti-conforma-
tion in which the imide carbonyls are oriented in opposite
directions.14 Semiempirical calculations 3vide infra) suggest
that the barrier to interconversion between the anti- and syn-
conformations has a range of 2.8±5.4 kcal/mol wherein the
anti-conformer is the more stable of the two. This arrange-
ment would be favored as it would reduce any allylic strain
induced by the nearby N34)-methyl group.
2.2. Examination of heterocycle 7a by X-ray
crystallography
The structure of 7a was determined by single-crystal X-ray
diffraction 3Fig. 2).8 The crystallographic data revealed that
the conformation of 7a is that of a twist-boat with the N34)-
methyl and C35)-methyl substituents oriented in a trans 1,2-
diaxial arrangement as indicated by the 169.136)8 dihedral
angle [C313)±C35)±N34)±C314)]. The orientation of the
N34)-methyl group is likely a consequence of allylic strain9
Attempts to grow crystals suitable for X-ray crystallo-
graphic analysis for the pseudoephedrine based heterocycles
5±7b were unsuccessful. It is likely that the pseudo-
ephedrine heterocycles also adopt a similar twist boat

S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
9791
Figure 2. ORTEP diagram of 7a 3RˆCH2Ph), 30% probability ellipsoids shown. Hydrogen atoms are drawn arbitrarily small for clarity.
conformation and contain considerable allylic stain arising
from the proximity of the N3- and N4-substituents.
course of their variable temperature studies, they could
not observe more than one discrete conformation even
1
though there were changes in the peak shapes in the H
NMR spectra. A complex series of equilibria involving
slow and fast processes was believed to be the cause of
the observed phenomena. The evidence collected from
this research suggested that `passing ring inversion' was
the process being observed.17 In the case of the current
family of pseudoephedrine heterocycles 5±7b, the argument
is made that there is a similar complex set of equilibria that
2.3. Examination of the acylated heterocycles by NMR
spectroscopy
The collection of the 13C NMR spectroscopic data for the
ephedrine family of acylated heterocycles 5±7a was
straightforward and presented no dif®culties. In contrast,
the collection of 13C NMR spectroscopic data for the
pseudoephedrine series 5±7b was not straightforward as
some signals were signi®cantly broadened 3Table 1).
1
gives rise to the broadened signals observed in the H and
C-13 NMR spectra.
1
There was also evidence of a broadened signal in the H
1
In terms of the H NMR spectra of 5±7b, the only observ-
NMR spectra of pseudoephedrine derivatives 5±7b. This
signal was attributed to the C35)-methine. The 13C NMR
resonance signals that were broadened were determined to
be the methyl group appended to the N4-nitrogen 3±CH±
N4±CH3), the methine group appended to the N4-nitrogen
3±CH±N4±CH3), the alpha carbon of the acyl group 3±CO±
CH2R) and the urethane carbonyl group.15 In contrast, the
C-13 NMR spectrum of the non-acylated pseudoephedrine
heterocycle 1b does not have any resonance signals that are
broadened. It stands to reason that the addition of the
N3-acyl substituent is primarily responsible for the observed
conformational differences.
able broadening was in the C35)-methine proton at ambient
temperatures. This NMR signal resolves to a doublet at
2358C. Interestingly, the C35)-methine carbon shows
virtually no sign of broadening in the 13C NMR spectra of
5±7b. The 13C NMR signals 5±7b were recorded at 50, 25
and 2358C 3Table 1). The 13C NMR spectrum of the propio-
nyl pseudoephedrine heterocycle 6b is exemplary 3Fig. 3).
As the temperature was increased to 508C, the broadened
signals began to resolve, although not completely. In
contrast, when the temperature was lowered to 2358C, the
signals became well resolved and new, smaller companion
signals became apparent.
2.4. Variable temperature nuclear magnetic resonance
studies of 5±7b
The N4-methyl substituent 3CH±N4±CH3) resonates at
36.9 ppm at 258C. At 2358C, the same methyl group is
represented by a signal at 35.8 and another smaller, broader
signal at 46.7 ppm. The methine group appended to the N4-
nitrogen 3CH±N4±CH3) resonates at 60.2 ppm at ambient
Katritzky16 and Riddell17 independently investigated the
conformational behavior of the related ring system 3,4-
dimethyl-3,4,5,6-tetrahydro-2H-1,3,4-oxadiazine. In the
Table 1. 100 MHz 13C NMR spectroscopic data for pseudoephedrine heterocycles 5±7b
Substrate
Temperature 38C)
CH3±CH±N±CH3
Pseudoeq. Pseudoax.
35.8
CH3±CH±N±CH3
Pseudoeq. Pseudoax.
58.8
CH3±CH±N±CH3
Pseudoeq. Pseudoax.
5b
5b
5b
6b
6b
6b
7b
7b
7b
235
25
50
235
25
50
235
25
50
46.7 3broad)
64.9 3broad)
14.3
14.3
15.3
14.2
14.9
15.2
14.4
14.9
15.3
23.9
±
±
17.6
±
±
17.6
±
±
36.9 3broad)
38.1 3broad)
35.7
±
59.5 3broad)
60.6 3broad)
58.8
±
±
46.7 3broad)
±
65 3broad)
37.1 3broad)
38.1 3broad)
35.9
±
60.2 3broad)
60.6 3broad)
59.1
±
±
46.9 3broad)
±
65.6 3broad)
37.3 3broad)
38.4 3broad)
±
±
60.3 3broad)
60.9 3broad)
±
±

9792
S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
Figure 3. Comparative 13C NMR temperature study for heterocycle 6b.
temperatures. However, when the temperature is lowered to
2358C, the same methine group is represented by a well-
resolved signal at 58.8 ppm and a broadened signal at
65.0 ppm. The alpha carbon of the acyl group 3±CO±
CH2R) also shows broadening but not to the same degree.
The signal resolves easily when the temperature is reduced
to 2358C. The urethane carbonyl NMR signal resonates at
148.3 ppm at 258C and is signi®cantly broadened. At lower
temperatures, this signal resolves to a ®ne line width and
occurs at 148.1 ppm. However, no companion peaks could
be detected for the carbonyl. The companion signals are
believed to be either too weak to observe or at nearly the
same chemical shift.
suggest that the ephedrine heterocycles exist in an equi-
librium condition that is exactly the same as that of the
pseudoephedrine
2.5. Semi-empirical calculations
In an effort to gain insight into the observed conformational
changes in the pseudoephedrine heterocycles, AM1 semi-
empirical calculations were carried out.18 From the data
collected, it was determined that the anti-orientation
of the carbonyls was consistently more stable than the
syn-orientation in the gas phase as well as in polar and
non-polar media 3Scheme 2). The barrier of interconver-
sion between the two orientations ranges from 2.8 to
5.4 kcal mol21. Although interconversion between the two
conformers can readily occur, these calculations would
suggest that the solution phase of the heterocycles favor
the presence of the anti-orientation.13
Ephedrine based heterocycles 5±7a were also investigated
using variable temperature 13C NMR. Heterocycle 6a was
cooled to 2358C and its 13C NMR spectrum was recorded
and no change in peak shape was observed. Warming the
sample to 508C did not signi®cantly change the appearance
of the spectrum. There was no change in peak shape to
Charts 1 and 2 illustrate potential conformations A±D

S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
9793
Scheme 2. Carbonyl orientation for ephedrine heterocycle 5a.
1
that exist for the ephedrine heterocycle 5a and pseudo-
ephedrine heterocycle 5b. These charts describe two
basic conformational changes that are believed to be key
to understanding the oxadiazinone ring system. These
changes are ring ¯ipping 3A to C and B to D) and
pyramidal inversion about the N34)-nitrogen 3A to B and
C to D).
facile and cannot be directly observed by H or 13C NMR
spectroscopy. The energy barriers for the N4-pyramidal
interconversion of 5a-A to 5a-B and 5a-C to 5a-D suggest
that this process is less facile than ring ¯ipping. However,
the calculated barriers to pyramidal inversion are not high
enough to warrant the existence of 5a-D as the only confor-
mer.19 The 1H and 13C NMR spectra coupled with the calcu-
lated barriers lead to a number of different conclusions: 3a)
there is free equilibration among all four conformers; 3b) the
heterocycle is frozen in conformation 5a-D; 3c) all four
conformers are in equilibrium where the only experimen-
tally observable conformations are 5a-D and 5a-B 3via
facile ring ¯ipping). The conclusion adopted here is that
For the ephedrine class of compounds, conformer 5a-D,
which corresponds to X-ray crystal structure 7a, had the
greatest stability based on the AM1 predictions. The calcu-
lated barriers for interconversion of 5a-D to 5a-B as well as
5a-A to 5a-C suggest that the process of ring ¯ipping is
Chart 1. Potential conformations for ephedrine heterocycle 5a 3Rˆ±CH3).

9794
S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
Chart 2. Potential conformations for the pseudoephedrine heterocycle 5b 3Rˆ±CH3).
there is equilibration between the conformers which
marginally favors the thermodynamically more stable
isomer 5a-D which readily interconverts with 5a-B.
spectra would tend to support the idea that pyramidal inver-
sion at the N34)-nitrogen is the root cause of the observed
line broadening. The barriers for ring ¯ipping are too low
3,2 kcal mol21) for such events to be observed. The change
in position of the N4-methyl group is more likely to be
responsible for the observed signal broadening. The only
1H NMR signal that is broadened is that of the C35)-
methine. If this substituent is in an environment where the
N34)-methyl is motion, then an averaged signal would be
observed. In addition, the 13C NMR signal that is most
broadened is the N34)-methyl substituent. The 13C NMR
signal for the alpha carbon of the acyl substituent is also
broadened but to a smaller degree. Thus it would seem
unlikely that free rotation of N33)-acyl group is solely
responsible for the observed signal broadening. The inde-
pendent conformational factors active in the pseudo-
ephedrine family cannot necessarily explain the observed
phenomena. The origin of the line broadening observed in
pseudoephedrine family most likely stems from a combina-
tion of factors including N34)-pyramidal inversion, the
conformational mobility of the N33)-acyl group and other
events that may have not been fully elucidated.
The pseudoephedrine conformers 5b-A±D have barriers for
interconversion 3ring ¯ipping/pyramidal inversion) that are
on the same order as the ephedrine derivatives. The con-
former of highest stability predicted by AM1 calculations
was determined to be 5b-D conformer. However, based on
the 1H NMR spectra of pseudoephedrine heterocycles 5±7b
there is no direct evidence for the presence of conformer
5b-D within the detection limits of the 1H NMR
spectrometer. The geometry of this conformer places the
N34)-methyl group in the region proximal the p-cloud of
the aromatic residue at C36). This would lead to a chemical
shift of the N34)-methyl group that would be highly
shielded. In fact, the average chemical shift of the
ephedrine and pseudoephedrine N34)-methyl groups are at
2.97 and 2.85 ppm, respectively. This difference is not
substantial enough to support the existence of conformer
5b-D as the dominant conformation observed
experimentally.
Calculations were also carried out to investigate the possi-
bility of inversion at the N4-nitrogen. The barriers for N4-
pyramidal inversion between the pseudoephedrine hetero-
cycles 5±7b are listed in Chart 2. These barriers are on the
same level as that of the ephedrine heterocycles 5±7a and
similar ring systems.20 The variable temperature 13C NMR
3. Conclusion
In summary, we have synthesized a series of 3,4,5,6-tetra-
hydro-2H-1,3,4-oxadiazin-2-ones
available 31R,2S)-ephedrine and 31S,2S)-pseudoephedrine.
from
commercially

S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
9795
Based on the evidence gathered from the 13C NMR spectra,
X-ray crystallography and semiempirical calculations, it
appears that the ephedrine heterocycles exist in a complex
equilibria that favors one set of conformers 35a-B/5a-D).
The corresponding pseudoephedrine heterocycles exhibit a
complex series of equilibria that involve free equilibration
among all conformers. The conformation ¯exibility is most
likely due to the combination of several factors including
allylic strain and the gauche effect, dynamic mobility of
the N33)-acyl group and other molecular events yet to be
recognized.
either as a neat liquid or as a KBr window. Melting points
were recorded on Mel-Temp apparatus and are
uncorrected. Low-resolution gas chromatography±mass
spectrometry was performed on Hewlett-Packard
a
a
Instrument 3G1800A/GCD) with an ionization voltage of
70 eV; peaks are reported as m/z 3% intensity relative to
the base peak). High-resolution mass spectra were obtained
from the Mass Spectrometry Laboratory, School of
Chemical Sciences, University of Illinois, Urbana-
Champaign. The 70-VSE mass spectrometer used for
sample analysis was purchased in part with a grant from
the Division of Research Resources, National Institutes of
Health 3RR 04648). Elemental analyses were conducted by
either Galbraith Laboratories, Inc. Knoxville, TN or by the
Microanalytical Laboratory, School of Chemical Sciences,
University of Illinois, Urbana-Champaign.
4. Experimental
4.1. Crystal structure determination of 7a8
A single crystal suitable for X-ray diffraction was obtained
by layering an ethyl acetate solution of 7a with hexanes.
The colorless prismatic crystal was mounted on a glass ®ber
and used for data collection at 298 K on an Enraf-Nonius
4.2.1. 41R,2S)±2-Amino-N-nitroso-N-methyl-1-phenyl-1-
propanol 43a). This material was prepared as described in
an earlier report.4 Mp: 92±948C 3lit. mp: 92±948C).
1
Rfˆ0.55 3EtOAc±hexanes, 1:1). H NMR 3CDCl3/TMS):
Ê
MACH3 automatic diffractometer, Mo Ka 3lˆ0.70930 A).
1.47 3d, 3H, Jˆ7.0 Hz), 2.40 3bs, 1H), 2.96 3s, 3H), 4.69
3quintet, 1H, Jˆ7.0 Hz), 5.07 3d, 1H, Jˆ5.1 Hz), 7.35±7.37
3m, 5H). 13C NMR 3CDCl3): 13.4, 31.4, 65.3, 74.6, 126.4,
128.4, 128.8, 141.0.
Cell constants and an orientation matrix for data collection
were obtained by least squares re®nements of the diffraction
data from 25 re¯ections. 7a was found to crystallize in the
monoclinic crystal system with the following unit cell
parameters: aˆ9.38332), bˆ10.63932), cˆ9.51931) and
bˆ116.7231)8, Zˆ2. A total of 1719 re¯ections were
collected of which 1575 were unique and 1113 were
observed 3I.3sI).
4.2.2. 41S,2S)±2-Amino-N-methyl-N-nitroso-1-phenyl-1-
propanol 43b). In a round bottom ¯ask equipped with a
stir bar was placed 31S,2S)-31)-pseudoephedrine 3b
320.0 g, 121 mmol) and THF350 mL). An aqueous solution
of HCl 350.8 mL, 2.74 M, 139 mmol) was then added
followed by the addition of sodium nitrite 39.60 g,
139 mmol) in small portions over a period of 10 min. The
reaction was stirred for 24 h and the cloudy mixture was
then diluted with a saturated aqueous solution of sodium
bicarbonate. The reaction mixture was then extracted with
EtOAc 33£75 mL) and the extract was washed with a
saturated aqueous solution of brine 350 mL). The resulting
solution was dried 3Na2SO4) followed by the removal of the
solvent by rotary evaporation. This yielded a yellow solid
which was recrystallized with EtOAc and hexanes to give
4b as a yellow crystalline solid 322.9 g). Yield: 97%. Mp:
85±868C. Rfˆ0.47 3EtOAc±hexanes, 1:1). 1H NMR
3CDCl3): 1.29 3d, 3H, Jˆ6.6 Hz), 2.44 3bs, 1H), 3.07 3s,
3H), 4.75±4.86 3m, 2H), 7.32±7.41 3m, 5H). 13C NMR
3CDCl3): 16.0, 30.2, 65.1, 76.4, 126.8, 128.6, 128.9,
140.4. IR 3KBr): 3480, 3032, 1465 3NvO stretch). m/z
3EI): 107 370, M12C8H11), 77 335), 58 3100, M12C9H120).
Data reduction and re®nement were performed using maXus
2.0 package of programs.21 The space group was determined
to be P21. The structure was solved with direct methods and
all non-hydrogen atoms were re®ned with anisotropic
temperature factors. The H atoms of CH were ®xed at
Ê
dˆ0.96 A, allowed to ride on the C atoms and assigned
Ê 2
®xed isotropic temperature factor, Uˆ0.05 A . Scattering
factors are from Wassmaire and Kirfel.22 Full-matrix
least-squares re®nement on F led to convergence with
Rˆ3.51% and wRˆ4.29%. A ®nal difference Fourier
synthesis showed features in the range of 10.11 to
Ê 23
20.12 eA . Details of the X-ray data collection and of
the crystal structure determination can be obtained from
the Cambridge Crystallographic Data Center.8
4.2. General remarks
Tetrahydrofuran 3THF) and diethyl ether 3Et2O) were
distilled from a potassium/sodium amalgam with benzo-
phenone ketyl. Methylene chloride 3CH2Cl2) was distilled
from phosphorus pentoxide. All reactions were run under a
nitrogen atmosphere. Lithium aluminum hydride 3LAH)
was purchased from Aldrich chemicals. Flash chromato-
graphy was conducted with silica gel purchased from
Selecto scienti®c 332±63 mesh). Unless otherwise noted,
4.2.3. 41R,2S)-N-Amino-2-methylamino-1-phenyl-1-pro-
panol 44a). In a ¯ame-dried, nitrogen-purged 5 L, 3-neck
round bottom ¯ask ®tted with an addition funnel and a
condenser was placed lithium aluminum hydride 312.3 g,
325 mmol) and freshly distilled THF3300 mL). The mixture
was then heated to re¯ux and 31R,2S)-N-nitrosoephedrine
3a 321.0 g, 108 mmol), dissolved in freshly distilled THF
3250 mL), was added slowly through the addition funnel for
30 min. Once the addition was completed, the reaction
mixture was maintained under re¯ux for an additional 5 h.
The reaction mixture was cooled to 08C followed by the
cautious addition of NaOH 36 M) until the remaining
lithium aluminum hydride was consumed. The mixture
was diluted with a saturated aqueous solution of sodium
1
all H and 13C NMR spectra were recorded at 258C on a
Varian spectrometer in CDCl3 operating at 400 and
100 MHz, respectively. Chemical shifts are reported in
parts per million 3d scale), and coupling constants 3J values)
are listed in hertz 3Hz). Tetramethylsilane 3TMS) was used
as an internal standard 3dˆ0.00 ppm). Infrared spectra are
reported in reciprocal centimeters 3cm21) and are measured

9796
S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
potassium tartrate 3Rochelle's salt, 150 mL) and stirred for
45 min. The mixture was then extracted with EtOAc
34£125 mL), washed with saturated aqueous brine solution
3100 mL), dried 3MgSO4) and the solvent removed by rotary
evaporation. This process afforded a viscous yellow oil that
4.3. 45S,6R)-3-Acyl-4,5-dimethyl-6-phenyl-3,4,5,6-tetra-
hydro-2H-1,3,4-oxadiazin-2-ones 45a±7a) and 45S,6S)-3-
acyl-4,5-dimethyl-6-phenyl-3,4,5,6-tetrahydro-2H-1,3,4-
oxadiazin-2-ones 45b±7b): typical procedure
1
was determined to be ca. 95% pure by H NMR spectro-
In a ¯ame dried, nitrogen-purged round bottom ¯ask
equipped with a re¯ux condenser and stir bar was placed
ephedrine heterocycle 1a 31.00 g, 4.86 mmol) and
4-3dimethylamino)pyridine 30.593 g, 4.86 mmol) followed
by the addition of 1,2-dichloroethane 318 mL). Once the
solids were dissolved triethylamine 31.35 mL, 9.72 mmol)
were added. The acyl chloride 35.83 mmol) was added via
syringe and the reaction mixture was heated to re¯ux and
stirred for 15 h. The reaction mixture was then cooled to
room temperature and was diluted with a saturated aqueous
solution of sodium bicarbonate 350 mL) and extracted with
EtOAc 32£50 mL). The combined organic layers were
washed with brine 350 mL), dried 3MgSO4) and the solvents
were removed by rotary evaporation to afford the title
compound.
1
scopy 318.0 g). Yield: 93%. Rfˆ0.22 3EtOAc). H NMR
3CDCl3): 0.83 3d, 3H, Jˆ6.6 Hz), 2.59 3s, 3H), 2.76 3dq,
1H, Jˆ6.8, 1.5 Hz), 5.21 3s, 1H), 7.21±7.39 3m, 5H). 13C
NMR 3CDCl3): 3.1, 48.7, 64.9, 77.7, 125.9, 126.70, 142.4.
IR 3KBr): 3313, 2978, 1046.
4.2.4. 41S,2S)-N-Amino-2-methylamino-1-phenyl-1-pro-
panol 44b). 31S,2S)-N-Nitrosopseudoephedrine 3b 322.9 g,
118 mmol) was treated in a similar fashion as 3a. This
process afforded the title compound as a viscous yellow
oil 321.1 g). Yield: 99%. Rfˆ0.19 3EtOAc). 1H NMR
3CDCl3): 0.79 3d, 3H, Jˆ6.6 Hz), 2.58 3s, 3H), 2.63±2.69
3m, 1H), 4.42 3d, 1H, Jˆ9.2 Hz), 7.27±7.38 3m, 5H). IR
3neat): 3333, 2975, 754, 701. m/z 3EI): 133 31), 91 35), 73
3100).
4.3.1.
45S,6R)-3-Acetyl-4,5-dimethyl-6-phenyl-3,4,5,6-
4.2.5. 45S,6R)-4,5-Dimethyl-6-phenyl-3,4,5,6-tetrahydro-
2H-1,3,4-oxadiazin-2-one 41a). In a ¯ame-dried, nitrogen-
purged 500 mL round bottom ¯ask was placed 31R,2S)-
ephedrine hydrazine 5a 314.3 g, 79.4 mmol) and freshly
distilled THF3200 mL). To this solution was added
p-toluenesulfonic acid monohydrate 315.1 g, 87.4 mmol)
and 1,10-carbonyldiimidazole 314.2 g, 87.4 mmol). After
the addition was completed, the resulting mixture was
heated to re¯ux. The reaction mixture was cooled to room
temperature after 3 h and an aqueous saturated solution of
sodium bicarbonate 350 mL) was added. The resulting
mixture was extracted with EtOAc 33£50 mL) and the
extract was washed with saturated brine 350 mL), dried
3Na2SO4), and the solvent was removed by rotary evapora-
tion. This process yielded a yellow oil which was puri®ed by
column chromatography on silica gel 3EtOAc, column
dimensionsˆ22£5 cm) to yield 1a 312.1 g) as a white
tetrahydro-2H-1,3,4-oxadiazin-2-one 45a). The crude
product was then puri®ed by ¯ash chromatography on silica
3column dimensions: 14£5 cm, Rfˆ0.39, 55:45 hexanes/
ethyl acetate, 42 fractions at 3 mL each) to yield 1.15 g of
5a. Yield: 95%. Mp: 107±1088C. 1H NMR 3CDCl3) d 0.85
3d, 3H, Jˆ7.0 Hz), 2.59 3s, 3H), 2.97 3s, 3H), 3.41 3dq, 1H,
Jˆ7.0, 4.4 Hz), 6.05 3d, 1H, Jˆ4.4 Hz), 7.28±7.42 35H, m).
13C NMR 3CDCl3) d 12.6, 26.4, 43.5, 56.9, 78.1, 125.1,
128.4, 128.9, 135.8, 148.7, 170.9. IR 3KBr) 2978, 1760,
1731 cm21. HRMS calcd for C13H16N2O3 248.1161, found
248.1156. Anal. calcd for C13H16N2O3: C, 62.89; H, 6.50; N,
11.28. Found: C, 62.87; H, 6.48; N, 11.14.
4.3.2. 45S,6R)-4,5-Dimethyl-6-phenyl-3-propionyl-3,4,5,
6-tetrahydro-2H-1,3,4-oxadiazin-2-one 46a). The crude
product was recrystallized from hexanes/ethyl acetate to
yield 6a as white needles. Yield: 86%. Mp: 94±958C;
Rfˆ0.68 3EtOAc). 1H NMR 3CDCl3): 0.85 3d, 3H,
Jˆ6.6 Hz), 1.20 3t, 3H, Jˆ7.3 Hz), 2.97 3s, 3H), 2.97 3m,
2H), 3.41 3dq, 1H, Jˆ7.0, 4.7 Hz), 6.04 3d, 1H, Jˆ4.4 Hz),
7.29±7.42 3m, 5H). 13C NMR 3CDCl3): 9.2, 12.6, 31.5, 43.5,
56.9, 77.9, 125.1, 128.3, 128.8, 135.9, 148.5, 174.7. IR
3KBr): 3064, 2981, 1779, 1724 cm21. HRMS calcd for
C14H18N2O3 262.1317, found 262.1309. Anal. calcd for
C14H18N2O3: C, 64.11; H, 6.92; N, 10.68. Found: C,
64.10; H, 6.94; N, 10.68.
1
solid. Yield: 74%. Mp: 118±1208C. Rfˆ0.52 3EtOAc). H
NMR 3CDCl3): 0.93 3d, 3H, Jˆ7.0 Hz), 2.89 3s, 3H), 3.13
3bs, 1H), 5.81 3s,1H), 6.88 3bs, 1H), 7.32±7.41 3m, 5H). 13C
NMR 3CDCl3): 11.5, 46.5, 56.9, 74.1, 125.2, 127.9, 128.5,
136.2, 152.3. IR 3KBr): 3228, 2943, 1686. m/z 3EI): 118
375), 91 333), 77 3100). Anal. calcd for C11H14N2O2: C,
64.06; H, 6.84; N, 13.59. Found: C, 64.10; H, 6.90; N, 13.51.
4.2.6. 45S,6S)-4,5-Dimethyl-6-phenyl-3,4,5,6-tetrahydro-
2H-1,3,4-oxadiazine-2-one 41b). 31S,2S)-pseudoephedrine
hydrazine 4b 32.86 g, 15.9 mmol) was treated in a similar
fashion as 4a. This reaction yielded a yellow oil which was
puri®ed by column chromatography on silica 3EtOAc,
Rfˆ0.30, column dimensionsˆ5£16 cm) to yield 2.40 g of
the title compound as a white solid. Yield: 74%. Mp: 97±
4.3.3. 45S,6R)-4,5-Dimethyl-6-phenyl-3-phenylacetyl-3,4,
5,6-tetrahydro-2H-1,3,4-oxadiazin-2-one 47a). Chromato-
graphy of the crude reaction mixture on silica 3column
dimensions: 14£5 cm, Rfˆ0.41, 65:35 hexanes/ethyl
acetate, 79 fractions at 3 mL each, fractions 45±78
collected) afforded the title compound 7a. Yield: 60%.
Mp: 125±1268C. 1H NMR 3CDCl3): 0.68 3d, 3H,
Jˆ7.0 Hz), 2.93 3s, 3H), 3.37 3dq, 1H, Jˆ7.0, 4.4 Hz),
4.31 3AB spin system, 2H, Dnˆ35, Jˆ15 Hz), 6.01 3d,
1H, Jˆ4.8 Hz), 7.22±7.39 3m, 10H). 13C NMR 375 MHz;
CDCl3): 12.4, 43.4, 44.1, 57.0, 78.0, 125.0, 127.2, 128.3,
128.5, 128.8, 129.6, 134.1, 135.7, 148.3, 172.0. IR 3KBr):
2972, 1763 3broad CvO stretch) cm21. HRMS calcd for
C19H20N2O3 324.1474, found 324.1465. Anal. calcd for
1
988C; Rfˆ0.48 3EtOAc). H NMR 3CDCl3/TMS): 1.03 3d,
3H, Jˆ7.0 Hz), 2.78 3s, 3H), 3.32 3dq, 1H, Jˆ10.1, 7.0 Hz),
5.34 3d, 1H, Jˆ9.9 Hz), 7.32±7.42 3m, 5H) 7.45 3bs, 1H).
13C NMR 3CDCl3/TMS): 13.9, 39.8, 58.4, 80.3, 126.8,
128.4, 128.7, 136.4, 152.9. IR 3KBr): 3240, 1699, 755,
700. m/z 3EI): 191 319), 117 319), 57 3100). HRMS calcd
for C11H14N2O2 206.1055, found 206.1051. Anal. calcd for
C11H14N2O2: C, 64.06; H, 6.84; N, 13.59. Found: C, 63.80;
H, 6.69; N, 13.33.

S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
9797
1968, 11, 357. 3b) Trepanier, D. L.; Harris, J. N. US Patent
3,377,345, 1968; Chem. Abstr. 1969, 70, 78026c.
C19H20N2O2: C, 70.35; H, 6.21; N, 8.64. Found: C, 70.00; H,
6.29; N, 8.53.
2. 3a) Roussi, F.; Chauveau, A.; Bonin, M.; Micouin, L.; Husson,
H. Synthesis 2000, 1170. 3b) Roussi, F.; Bonin, M.; Chiaroni,
A.; Micouin, L.; Riche, C.; Husson, H. Tetrahedron Lett.
1999, 40, 3727. 3c) Roussi, F.; Bonin, M.; Chiaroni, A.;
Micouin, L.; Riche, C.; Husson, H. Tetrahedron Lett. 1998,
39, 8081.
4.3.4.
45S,6S)-3-Acetyl-4,5-dimethyl-6-phenyl-3,4,5,6-
tetrahydro-2H-1,3,4-oxadiazin-2-one 45b). Chromato-
graphy of the crude reaction mixture on silica 3column
dimensions: 12£5 cm, Rfˆ0.53, 50:50 hexanes/THF, 30
fractions at 3 mL each) afforded the title compound 5b.
1
Yield: 89%. Mp: 90.0±90.58C. H NMR 3CDCl3) d 1.15
3. 3a) Evans, D. A.; Bartoli, J.; Shih, T. L. J. Am. Chem. Soc.
1981, 103, 2127. 3b) Seyden-Penne, J. Chiral Auxiliaries and
Ligands in Asymmetric Synthesis; Wiley: New York, 1995.
3c) Lin, G.-Q.; Li, Y.-M.; Chan, A. S. C. Principles and
Applications of Asymmetric Synthesis; Wiley: New York,
2001. 3d) Ager, D. J.; Prakash, I.; Schaad, D. R. Aldrichim.
Acta 1997, 30, 3±12.
3d, 3H, Jˆ7.0 Hz), 2.55 3s, 3H), 2.79 3s, 3H), 3.32 3dq,
1H, Jˆ7.0, 4.0 Hz), 5.21 3bs, 1H), 7.29±7.41 35H, m). 13C
NMR 375 MHz, CDCl3) d 14.3, 24.6, 36.9, 59.5, 83.0,
126.1, 128.1, 128.3, 135.3, 148.3, 169.5. IR 3KBr) 2984,
1782, 1731, 1057 cm21. HRMS calcd for C13H16N2O3
248.1161, found 248.1165. Anal. calcd for C13H16N2O3:
C, 62.89; H, 6.50; N, 11.28. Found: C, 62.82; H, 6.45; N,
11.18.
4. The ephedrine N-nitrosamine was prepared previously:
Hitchcock, S. R.; Nora, G. P.; Hedberg, C.; Casper, D. M.;
Buchanan, L. S.; Squire, M. D.; West, D. X. Tetrahedron
2000, 56, 8799.
4.3.5. 45S,6S)-4,5-Dimethyl-6-phenyl-3-propionyl-3,4,5,
6-tetrahydro-2H-1,3,4-oxadiazin-2-one 46b). Chromato-
graphy of the crude reaction mixture on silica 3column
dimensions: 15£5 cm, Rfˆ0.28, 70:30 hexanes/ethyl
acetate, 47 fractions at 4 mL each) afforded 1.52 g of the
5. Caution. It should be noted that many N-nitrosamines are
potentially carcinogenic and should be handled with great
care. For more information on N-nitrosamines, please see:
Lawley, P. D. In Chemical Carcinogens, Searle, C. D., Ed.;
ACS Monograph 182; American Chemical Society:
Washington, DC, 1984.
6. The experimental procedure developed by Husson2b for the
reduction of the N-nitrosamine failed to work in our hands.
An older literature method was adopted for this reaction. See:
Corey, E. J.; McCaully, R. J.; Sachdev, H. S. J. Am. Chem.
Soc. 1970, 92, 2476.
1
title compound 6b as a viscous oil. Yield: 78%. H NMR
3CDCl3) d 1.13 3d, 3H, Jˆ6.6 Hz), 1.19 3t, 2H, Jˆ7.3 Hz),
2.78 3s, 3H), 2.86 3m, 2H), 3.30 3dq, 1H, Jˆ6.6, 4.4 Hz),
5.24 3d, 1H, Jˆ9.5 Hz), 7.29±7.42 35H, m). 13C NMR
3CDCl3) d 8.9, 14.9, 30.3, 37.2, 60.2, 83.8, 126.7, 128.8,
129.3, 135.9, 148.8, 173.9. IR 3KBr) 2981, 1780, 1726,
1209 cm21. m/z 3EI): 206 332), 118 3100), 91 322), 77
320). HRMS calcd for C14H18N2O3 262.1317, found
262.1323.
7. The experimental procedure for the synthesis of the hetero-
cycle was modi®ed from that of Husson2b due to the excessive
formation of the undesired urea product. Alternate methods
were developed from earlier literature. See: 3a) Sanemitsu, Y.;
Nakayam, Y. Synthesis 1984, 773. 3b) Townsend, L. B.;
Beylin, V. G. Heterocycles 1984, 22, 1693. 3c) Nachman, R.
J. Heterocycl. Chem. 1982, 1545.
4.3.6. 45S,6S)-4,5-Dimethyl-6-phenyl-3-phenylacetyl-3,4,
5,6-tetrahydro-2H-1,3,4-oxadiazin-2-one 47b). Chroma-
tography of the crude reaction mixture on silica 3column
dimensions: 14£5 cm, Rfˆ0.19, 45:55 hexanes/ethyl
acetate, 40 fractions at 4 mL each followed by 15 fractions
at 50 mL each, fractions 46±55 collected) afforded 1.015 g
8. Crystallographic data 3excluding structure factors) for struc-
ture 7a reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre. The atomic coordi-
nates and equivalent isotropic displacement coef®cients are
included in the deposited material 3CCDC 172027) as are a
complete list of bond distances and angles. Copies of available
material can be obtained, free of charge, on application to the
CCDC, 12 Union Road, Cambridge CB2, 1EZ, UK 3fax: 144-
1223-336033, email: deposit@ccdc.am.ac.uk).
1
of the title compound 7b as a yellow oil. Yield: 43%. H
NMR 3CDCl3/TMS) d 1.24 3d, 3H, Jˆ7.0 Hz), 2.94 3s, 3H),
2.97 3s, 3H), 3.27 3dq, 1H, Jˆ7.0, 2.9 Hz), 4.41 3AB spin
system, 2H, Dnˆ57.5 Hz, Jˆ14.7 Hz), 5.28 3bs, 1H), 7.26±
7.28 34H, m), 7.49±7.56 36H, m). 13C NMR 3CDCl3) d 14.9,
37.3, 43.3, 60.3, 83.8, 126.5, 126.9, 128.4, 128.6, 129.0,
129.2, 133.8, 135.6, 148.6, 171.1. IR 3KBr) 2981, 1775,
1722, 1207 cm21. Anal. calcd for C19H20N2O2: C, 70.35;
H, 6.21; N, 8.64. Found: C, 70.38; H, 6.12; N, 8.70.
9. For a general discussion of allylic strain, see: Johnson, F.
Chem. Rev. 1968, 68, 375.
10. Gilchrist, T. L. Heterocyclic Chemistry; Addison-Wesley,
Longman Printing, 1997 pp 44±46.
Acknowledgements
11. The gauche effect has also been observed by X-ray crystal-
lography in the related 3,4-dihydro-1,3,4-oxadiazin-2-one
system: Schildberg, M.; Debaerdemaeker, T.; Friedrichsen,
W. Chem. Ber. 1988, 121, 887.
We thank Dr R. W. Nagorski for assistance with the
VT-NMR investigation. We also thank the reviewers for
their helpful suggestions. We gratefully acknowledge ®nan-
cial support from the Department of Chemistry 3ISU) and
Illinois State University's University Research Grant
program and continued support from Merck and Co., Inc.
12. In the case of ephedrine-based 7a, the X-ray crystal structure
revealed that the imide carbonyls are twisted out of planarity
by approximately 358. The pseudoephedrine derivatives may
be even more distorted. Early attempts to conduct asymmetric
alkylation reactions with heterocycles 5±7a and 5±7b resulted
in the cleavage of the acyl group. The loss of the acyl group
may be due in part to the higher reactivity associated with
amides that are twisted from planarity. For example, see:
References
1. Trepanier, D. L.; Elbe, J. N.; Harris, G. H. J. Med. Chem.

9798
S. R. Hitchcock et al. / Tetrahedron 57 -2001) 9789±9798
Duspara, P. A.; Matta, F. C. F.; Jenkins, S. I.; Harrison, P. H. M.
Org. Lett. 2001, 3, 495.
19. Sibi and co-workers have employed pyrazolidinones as chiral
relays for enantioselective ampli®cation. This process
depends on the conformational mobility of the N31)-nitrogen
of the heterocycle. The substituent bound to this nitrogen
responds to the addition of a chiral Lewis acid by adopting a
position that is away from the steric bulk of the acid. The
barrier to pyramidal inversion in this series of compounds
must be on the same order of the barriers in this work until
such point where the Lewis acid becomes complexed to the
heterocycle. See: Sibi, M. P.; Venkatraman, L.; Liu, M.;
Jasperse, C. P. J. Am. Chem. Soc. 2001, 123, 8444.
13. The X-ray crystal structure obtained is representative of the
conformation that is adopted in the solid state by 7a but may
not entirely represent conformational preferences in the solu-
tion phase. This particular orientation is made more interesting
by the arrangement of the aromatic residue of the N3-acyl
group in the region proximal to the N4-methyl substituent. It
is possible that crystal-packing forces are predominantly
responsible for this spatial arrangement.
14. Evans, D. A.; Wu, L. D.; Wiener, J. J.; Johnson, J. S.; Ripin,
D. H.; Tedrow, J. S. J. Org. Chem. 1999, 64, 6411 and
references therein.
15. The assignment of the 13C NMR signals were determined by
13C±1H coupling.
³
20. Riddell and Kidd calculated a barrier of DG ˆ7.5 kcal mol21
for pyramidal inversion at nitrogen in the related 3,4,5,6-
tetrahydro-1,3,4-oxadiaizine system. See Ref. 17.
21. Mackay, S.; Edwards, C.; Henderson, A.; Gilmore, C.;
Stewart, N.; Shankland, K.; Donald, A. MaXus 2.0 data reduc-
tion and re®nement programs. University of Glasgow: Scot-
land, 1997.
16. Ferguson, I. J.; Katritzky, A. R.; Read, D. M. J. Chem. Soc.,
Perkin Trans. 2 1976, 1801.
17. Riddell, F. G.; Kidd, A. J. J. Chem. Soc., Perkin Trans. 2 1977,
1861.
22. Wassmaier, D.; Kirfel, A. Acta Crystallogr. 1995, A51, 416.
18. Calculations were performed using Spartan 5.1 on a SGI O2
workstation.