Stereochemical study of 1,3-N,X-heterocycles derived from α-aminoacids and formaldehyde. Structural evidence for the existence of the anomeric effect

Article abstract of DOI:10.1016/S0040-4020(02)00405-2

We have shown that condensation of L-cysteine methyl ester (8) and L-threonine methyl ester (11) with formaldehyde provides a convenient and efficient access to novel heterocyclic 2:3 adducts. Depending on the amino acid, the condensation leads to either the N,N′-methylenebis(thiazolidine) (10, L-cys) or -(oxazolidine) (13, L-thr) derivative or to its bicyclo[4.4.1]undecane isomer (5, L-ser) as the major product of the reaction. The structure of 10, 12 and 13, was unambiguously confirmed by diffraction analysis and/or NMR spectroscopy. The latter proved to be a powerful tool to discriminate between the two possible isomers. X-Ray data emphasized the contribution of stereoelectronic effects to the structure of the above compounds.

Full text of DOI:10.1016/S0040-4020(02)00405-2

TETRAHEDRON
Pergamon
Tetrahedron 58 *2002) 4439±4444
Stereochemical study of 1,3-N,X-heterocycles derived from
a-aminoacids and formaldehyde. Structural evidence for the
existence of the anomeric effect
a
Jimmy S e lambarom, Francis Carr e , Alain Fruchier, Jean Pierre Roque and Andr e A. Pavia
b
c
a
a,p
a
Laboratoire de Chimie Organique Physique, Universit e de Montpellier II, CC020, Place Eug eÁ ne Bataillon,
F-34095 Montpellier Cedex 05, France
b
Laboratoire de Chimie Mol e culaire et Organisation du Solide UMR 5637, Universit e de Montpellier II, CC007, Place Eug eÁ ne Bataillon,
F-34095 Montpellier Cedex 05, France
c
Ecole Nationale Sup e rieure de Chimie de Montpellier, 8 rue de l'Ecole Normale, F-34095 Montpellier, France
Received 23 January 2002; revised 18 March 2002; accepted 11 April 2002
AbstractÐWe have shown that condensation of l-cysteine methyl ester *8) and l-threonine methyl ester *11) with formaldehyde provides a
0
convenient and ef®cient access to novel heterocyclic 2:3 adducts. Depending on the amino acid, the condensation leads to either the N,N -
methylenebis*thiazolidine) *10, l-cys) or -*oxazolidine) *13, l-thr) derivative or to its bicyclo[4.4.1]undecane isomer *5, l-ser) as the major
product of the reaction. The structure of 10, 12 and 13, was unambiguously con®rmed by diffraction analysis and/or NMR spectroscopy. The
latter proved to be a powerful tool to discriminate between the two possible isomers. X-Ray data emphasized the contribution of stereo-
electronic effects to the structure of the above compounds. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
The anomeric effect *and its generalized manifestations) is
well recognized as an important contributor to ground-state
conformational analysis of heteroatoms-containing systems.
1
±9
However, although it has been extensively studied
especially by NMR spectroscopy, X-ray diffraction
*
Scheme 1.
analysis and theoretical calculation), most of the efforts
were directed toward ®ve- and six-membered rings derived
from carbohydrates. So far, very little attention has been
paid to non-carbohydrate heterocycles containing N±C±O,
N±C±S or N±C±N sequences, especially when these are
incorporated by two in systems larger than the common
1
0
lography. The crystal structure of 2 showed that both
oxazolidine rings A and B were puckered. The E envelope
3
form of ring A was prone to electron delocalization from the
antiperiplanar lone-pair orbital on nitrogen to the s anti-
bonding orbital of the C-2±O-3 acceptor bond.
p
®ve- and six-membered rings. This paper aims to demon-
strate that anomeric effect applies to this type of situation.
On the other hand, the condensation of l-serine methyl ester
hydrochloride *3) with paraformaldehyde *4) in dichloro-
methane, in the presence of triethylamine, afforded two
novel isomeric 2:3 adducts 5 and 6 *Scheme 2). The struc-
Oxazolidine or thiazolidine derivatives were selected as
temporary protecting groups of tris*hydroxymethyl)-amino-
methane *TRIS), l-serine *l-ser), l-threonine *l-thr) and
l-cysteine *l-cys). When TRIS *1) was reacted with benzal-
dehyde, a 1:2 adduct was obtained which was assigned
the structure cis-2,8-diphenyl-5-hydroxymethyl-1-aza-3,7-
dioxabicyclo[3.3.0]octane *2) *Scheme 1), on the basis of
nuclear magnetic resonance spectroscopy and X-ray crystal-
1
13
ture of the major isomer 5 was established by H and C
NMR spectroscopy and further con®rmed by X-ray diffrac-
1
tion analysis to be [1S,2S,6S,7S]-1,6-diaza-4,9-dioxa-2,7-
1
dimethoxycarbonylbicyclo[4.4.1]undecane.
The crystal structure of compound 5 revealed *i) the
existence of two identical seven-membered rings, each
containing a N±C±O and a N±C±N group, *ii) the anti-
periplanar geometry of nN±C±O and *iii) the manifestation
of a strong anomeric effect occurring in both N±C±O
Keywords: anomeric effect; 1,3-N,X-heterocycles; a-aminoacids; crystal
structure.
Corresponding author. Tel./fax: 133-467-143-841;
e-mail: aapavia@univ-montp2.fr
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020*02)00405-2

4440
J. S e lambarom et al. / Tetrahedron 58 ꢀ2002) 4439±4444
Scheme 2.
groupings, as corroborated by bond distances and bond
angles.
behave similarly to serine methyl ester. In fact, whatever
the molar ratio of reactants, the reaction led to a mixture of
two isomeric 2:3 adducts which were separated on column
chromatography. The minor *15±20%) compound *12) was
obtained in crystalline state and assigned the structure
2
. Results and discussion
[1S,2S,3R,6S,7S,8R]-1,6-diaza-4,9-dioxa-2,7-dimethoxy-
carbonyl-3,8-dimethylbicyclo[4.4.1]undecane on the basis
A priori, the seven-membered ring forming process
1
2,13
mentioned above could agree with Baldwin's rules
which state that a 7-endo-trig cyclization is favored over a
-endo-trig process when the nucleophile is a ®rst-row
5
element, which is the case of serine. However, since the
normally disfavored 5-endo-trig process is facilitated
when sulfur *a second-row element) is involved, we
suspected that the reaction of l-cysteine methylester *8)
with formaldehyde could afford the thio-analogs of
compounds 6 and 7. Indeed, the reaction produced either
0
the methoxycarbonylthiazolidine 9 or the N,N -methylene-
bis*thiazolidine) derivative 10 depending on molecular ratio
1
see Scheme 3). Based on C and H NMR spectra, the
3
1
*
identi®cation of compound 9 *oil, m/z 147) was straightfor-
ward.
The 2:3 adduct which had never been reported before was
expected to be compound 10. This hypothesis was strongly
supported by the fact that 10 *m/z 306) was obtained by the
two convergent routes reported in Scheme 3 and also
because of strong similarities *in terms of signal patterns,
chemical shifts and coupling constants) between the proton
and carbon-13 NMR spectra of 10 and 9. X-Ray diffraction
analysis de®nitely con®rmed the structure of compound 10
*
Fig. 1*a)).
The reaction of l-threonine methyl ester *11) with formalde-
hyde was also investigated. We surmised that 11 might
Figure 1. ORTEP view of compounds 10 *a) and 12 *b) with atoms number-
ing. The thermal ellipsoids are drawn at the 30% probability level.
Scheme 3.

J. S e lambarom et al. / Tetrahedron 58 ꢀ2002) 4439±4444
4441
of a single-crystal diffraction analysis *see Fig. 1*b)). The
0
N,N -methylenebis*oxazolidine) structure of the major
component 13 *mp 32±338C) was established by *i) proton
and carbon-13 NMR spectroscopy, *ii) the facile isomeriza-
tion of 12 to 13 and vice versa *see latter on). However, as
1
1
for serine the oxazolidine 14 was never detected in the
reaction mixture.
2.1. Crystal structure of compounds 10 and 12.
Stereoelectronic effects
All crystal data for both compounds are presented in Section
4.2 *see Section 4). It must be noted that the e.s.d.s for bond
Ê
lengths are #0.003 A and bond angles #0.28.
Figure 2. Perspective view of the X-ray crystal structure of 12.
three-dimensional structure of 10 reveals a signi®cant
deviation from theoretical alignment. On the basis of
difference in electronegativity between nitrogen
*xˆ3.0) and sulfur *xˆ2.5) one could have envisage
2
.1.1. Compound 10. X-Ray crystal structure and atoms
labeling are shown in Fig. 1*a). The R con®guration of the
a-carbon of l-cysteine is preserved at C-2 *C-7), whereas
N-1*N-6) also displays the R con®guration. In the solid
state, 10 exhibits a rigorous *crystallographic) two-fold
axis of symmetry. This axis lies in the plane N-1±C-11±
N-6 passing through C-11 and bisecting the bond angle
N-1±C-11±N-6.
p
electron delocalization to occur from n*S-4) to the s
antibonding orbital of the C-5±N-1bond. Crystallo-
graphic data de®nitely rules out such an eventuality.
1
1,14,15
*b) As already observed, anomeric effect seems to
be absent in the N-1±C-11±N-6 aminal motif. The C±N
3
bond distance is identical to standard Csp ±Nsp bond
3
Ê
length, i.e. 1.469 A.
The three-dimensional representation of 10 *Fig. 1*a))
clearly shows the EN envelope conformation adopted by
both ®ve-membered rings. Nitrogen atoms were found
Ê
2.1.2. Compound 12. Bond distances and bond angles are
listed in Section 4.2. The structure and atom labeling are
shown in Figs. 1*b) and 2.
0
.506 A below the plane de®ned by the remaining four
atoms. The dihedral angle between this plane and the
envelope tip *36835) indicates that both ®ve-membered
rings are puckered. This envelope conformation may be
explained as a way to minimize inter- and intra-thiazolidine
ring non-bonding interactions. Indeed, molecular modeling
showed that the dihedral angle between the ester group and
the pseudo axial bond at C-3 increases to approximately
Con®guration of the a-*S) and b-*R) carbons of l-threonine
is preserved at C-2 *C-7) and C-3 *C-8). In the solid state 12
exhibits also a rigorous *crystallographic) two-fold axis of
symmetry. Conformation of the seven-membered ring A
Ê
1
*
508. In addition, rotation occurring at the inter-thiazolidine
or oxazolidine) linkage plays an important role in relieving
for instance is chair-like. Indeed, atom O-4 is 0.676 A
away from the P1plane which is de®ned by atoms N-1,
Ê
C-3, C-5, N-6 while atom C-11 is 0.679 A away from the
same plane. The dihedral angle between P1and P2 *de®ned
by atoms C-3, O-4, C-5) is equal to 6085 while P1forms a
unfavorable interactions, including those involving nitrogen
lone-pairs. On the basis of torsion angles, nitrogen lone pairs
were shown to be roughly orthogonal.
6685 dihedral angles with P3 being de®ned by atoms N-1,
C-11, N-6. Data above show that the sequence N-1±C-3±O-
Stereoelectronic effects. The most signi®cant crystal data in
relation with the possible contribution of the anomeric effect
to the structure of compound 10 are the following:
1
1
4±C-5±N-6±C-11 describes an almost perfect
conformation *see Fig. 2). This conformation accounts for
C chair
4
1
the speci®c features of the H NMR spectrum that will be
discussed later on.
*
a) A signi®cant contraction of C-5±N-1bond is
Ê
Ê
observed: 1.450 A compared to 1.469 and 1.470 A for,
respectively *N-1±C-2) and *N-1±C-11). Although less
Ê
pronounced *Dd 0.020 A) than in serine/threonine deriva-
Stereoelectronic effects. The contribution of stereo-
electronic effects to the three-dimensional structure of 12
can be inferred from the following crystal features:
Ê
tives 5 *Dd 0.033±0.039 A) and 12 *see below), this
contraction seems to indicate that anomeric effect occurs
in the sequence N-1±C-5±S-4. This hypothesis is corro-
Ê
*a) The N-1±C-10 bond *1.428 A) is unquestionably
shorter than either N-1±C-2 or N-1±C-11, respectively,
borated by the concomitant increase of the C-5±S-4 bond
Ê
Ê
1.474 and 1.465 A.
*1.843 A) which is longer than the reference C-3±S-4
bond *1.808 A). Such observations imply that, most
likely, electron delocalization from n*N) to C-5±S-4 s
Ê
*b) The signi®cant contraction of N-1±C-10 *N-6±C-5)
Ê
p
bond *Dd 0.039±0.046 A, i.e. 14£e.s.d. on the average)
antibonding orbital occurs. However, the differences in
bond shortening between compound 10 and compound 5
may re¯ect a less ef®cient overlapping in 10. This could
be due to the less electron-acceptor character of the C±S
bond compared to its C±O analog and possibly to an
imperfect antiperiplanar conformation between the nitro-
gen lone-pair and the C-5±S-4 acceptor bond. Indeed, the
reveals delocalization of the unshared pair of electrons on
nitrogen atom to the C±O s antibonding orbital in both
p
N±C±O groupings. The antiperiplanar geometry of the
nN±C±O sequence favors overlapping of n*N) and s
p
orbitals. Torsion angles involving bridgehead nitrogen
atoms *6685 and 7289) show that the C-5±O-4 bond
bisects fairly well the C-11±N-6±C-7 bond angle, as

4442
J. S e lambarom et al. / Tetrahedron 58 ꢀ2002) 4439±4444
does O-9±C-10 with respect to the C-11±N-1±C-2 bond
angle.
c) Bridgehead nitrogen atoms have gained some sp
both seven-membered rings A and B. They also attested that
chair conformation was maintained in solution.
2
*
character as illustrated by the signi®cant widening of
N-1bond angles. Of special interest is the sum of these:
3
value should be compared to the one measured in
NMR spectra of compound 10 were compared with those of
compound 5 *see Table 1). The salient features that emerged
from the comparison were the following: *i) the signal for
the methylene protons at C-11 *d 3.28) was considerably
different from the equivalent signal in compound 5 *d 4.56),
*ii) neither long-range coupling between H-5e *H-3e) and
H-11 *H-11 ) nor those between H-8e *H-10e) and H-5e
3
4587 compared with sp standard value 32885. This
compound 10 *33185) as well as those reported by
Lemoine et al. for acyclic N,N -methylenebis[*S)-5-
0
phenyloxazolidine] *32585 and 32484) and by Engel
et al. for N,N -methylenebis[4-methyl-5-phenyloxa-
1
6
B
A
0
*H-3e) was detected in 10, *iii) however, long-range
coupling was observed between H-5e and H-2 *0.5 Hz),
1
7
zolidine] *3248 and 32886).
d) None of the above data can reasonably support the
hypothesis of a stereoelectronic effect occurring in the
1
3
*
*iv) in the C NMR spectra, the most signi®cant difference
arose from the chemical shift of carbon-11 which was much
more deshielded in compound 10 *d 73.09) than in
compound 5 *d 68.08).
1
1,14,15
N±C±N aminal group as already observed.
The above observations allow to foresee successful identi-
®cation of both classes of compounds on the basis of NMR
spectroscopy alone. This eventuality was unambiguously
con®rmed by the careful examination of NMR spectra of
12 and 13 *Table 1) which underlined the following
peculiarities:
2
2
.2. Structure in solution
.2.1. NMR study. Rough NMR data were unable to dis-
criminate between the two possible structures for the 2:3
adducts. A priori, data could ®t both isomers. However, in
compound 5 for instance, careful analysis of H signals and
1
spectra simulations revealed the existence of two long-range
coupling constants between H-5e *H-10e) and H-11B
1
. Signi®cant chemical shift differences were observed for
C-11 *Dd 8.5), CH -11 *Dd 0.82) and methyl groups at
2
*
*
H-11 ) as well as between H-5e *H-10e) and H-3e
A
H-8e), respectively *see Table 1), supporting the existence
C-3 and C-8 *Dd 0.22).
2
. Long-range coupling *J5e,11BˆJ
ˆ2.0 Hz) was
10e,11A
of a rigid W-conformation for the appropriate bond
sequences. These observations were in accordance with
the chair-like conformation observed in its solid state for
observed in compound 12 whereas it was absent in 13.
Interestingly, the E envelope conformation *see Fig. 1*a)
N
and discussion below) of the ®ve-membered rings
0
appears to be preserved in solution in both N,N -
methylenebis derivatives 10 and 13, as attested by the
coupling between protons 5eq and 2eq *J, 0.5 Hz).
Indeed, crystal data and molecular models reveal that
1
13
Table 1. H and C NMR data for compound 5, 10, 12 and 13 in CDCl
3
*d
ppm, J Hz)
such E conformation is prone to long-range coupling
N
Compounds
5
10
12
13
as a consequence of the appropriate sequence displaying
a clear W-rigid arrangement.
a
Protons
H-2 *H-7)
d
4.00
4.43
4.26
4.24
3.92
4.56
±
4.40
4.28
4.24
3.29
3.14
3.28
±
3.59
4.44
4.19
±
4.06
4.54
1.23
3.73
3.51
4.58
4.46
±
3.99
3.72
1.48
3.75
3
3
. Finally, noticeable differences were observed in
J
H-5eq *H-10eq
H-5ax *H-10ax
H-3eq *H-8eq
H-3ax *H-8ax
)
)
coupling constants: J ˆ9.8 Hz in 12 versus 6.6 Hz in
2
,3
13.
)
)
b
H-11
CH
OCH
A B
, H-11
3
2.2.2. Equilibration in solution. The con®rmation of struc-
tural relationship between 5 and 6 on one hand, 12 and 13 on
3
3.74
3.74
J *Hz)
the other, was obtained through isomerization experiments.
Interconversion of 5 to 6 and vice versa in CDCl was
2
2
3
5
5
5
5
,3eqˆ7,8eq
3.0
9.8
±
211.4
1.6
0
3.5
7.6
±
29.9
0
9.8
±
6.6
±
6.2
6.2
0
3
,3axˆ7,8ax
1
investigated. At 408C, H NMR monitoring indicated that
,CH
ax, 5eqˆ10eq, 1 0ax
3
6.2
211.8
2.0
the equilibrium was reached within 24 h. Isomerization of 5
was accompanied by the decrease of the signal ascribed to
the methylene protons of the aminal bridge *d 4.56) and by
the concomitant increase of the equivalent signal *d 3.81) in
b
eq, 1 1Bˆ10eq, 1 1
A
eq, 2ˆ10eq, 7
0.5
0
0
0
0.5
0
eq, 3eqˆ10eq, 8eq
0.6
a
0
Carbons
CvO
the N,N -methylenebis*oxazolidine) derivative 6. This
change parallels the decrease of the OCH signal at d 3.74
172.0
87.44
70.03
64.58
68.08
52.28
±
171.45
57.32
32.60
171.06
87.94
76.45
172.47
85.38
76.83
3
C-5, C-10
C-3, C-8
C-2, C-7
C-11
and the increase of the equivalent signal in 6 at d 3.62.
Relative concentrations of 5 and 6 after equilibration were
80:20.
67.0170.20
73.09
52.35
±
69.26
68.17
52.5152.07
20.42
76.61
20.09
OCH
CH
3
3
A similar study was performed on CDCl solutions of pure
3
compounds 12 and 13. Interconversion of 12 to 13 and vice
versa in CDCl was investigated. H NMR monitoring indi-
cated that equilibrium was reached within 5 h at room
temperature. The equilibrium mixture was quantitated
CDCl
3
solution *solvent at 7.280 ppm vs TMS).
See Schemes 2 and 3 and Fig. 1for atoms numbering.
1
a
3
b
H-11
A
and H-11
B
represent proton-11 directed above ring A and ring B,
respectively *compounds 5 and 12).

J. S e lambarom et al. / Tetrahedron 58 ꢀ2002) 4439±4444
4443
using the relative intensity of the following signals: *i) that
of the methylene protons of the aminal bridge, *ii) that of the
methyl groups at C-3 *C-8), *iii) that of OCH . At equi-
very speci®c probes to discriminate between the two
possible isomeric structures.
3
librium, the molar ratio 12/13 *5:95) was drastically differ-
ent from that observed with serine. In contrast, under the
same conditions, compound 10 remained stable over a long
period of time *several weeks). This is in accordance with
4. Experimental
4.1. General procedure
1
8
the higher stability of the 1,3-thiazolidine ring. In conclu-
sion, the above data allow to establish the usefulness of
1
3
C
To a suspension of l-aminoacid methylester hydrochloride
*6.5 mmol) in anhyd. dichloromethane *50 mL) were added
triethylamine *7.8 mmol) and paraformaldehyde *6.5 or
13) mmol). After stirring at room temperature for 1±3
days *t.l.c. monitoring), the solvent was evaporated to
dryness and the residue, dissolved in diethylether
*100 mL), was ®ltered on sintered glass. The ®ltrate was
dried over anhyd. Na SO and evaporated to dryness. The
1
and H NMR spectroscopy to identify isomeric heterocyclic
:3 adducts resulting from the condensation of l-serine,
l-threonine and l-cysteine methylester with formaldehyde.
2
Surprisingly, the condensation of l-threonine methyl ester
0
with formaldehyde produces predominantly the N,N -
methylenebis*oxazolidine) derivative 13, a ®nding that
contrasts sharply with the observations concerning serine.
Due to the fact that compound 13 was the thermo-
dynamically stable isomer, this unexpected result might be
explained in the following ways:
2
4
residue was either recrystallized or puri®ed on silica gel
column chromatography to afford a solid or an oily material.
4.1.1. [1S,2S,6S,7S]-1,6-Diaza-4,9-dioxa-2,7-dimethoxy-
carbonylbicyclo[4.4.1]undecane ꢀ5). 90%; mp 91±928C;
2
0
1
1
. In compound 12, the presence of a methyl at C-3 gener-
ates additional non-bonding interactions that are absent
in 13. In this connection, the following points should be
[a]
[M1H] , 144; H and C NMR data see Table 1. Anal.
calcd for C11 *274): C, 48.17; H, 6.56; N, 10.22; O,
D
ˆ154.5 *c 1.0, CHCl
3
); FAB MS *NBA) m/z 275
1
1
13
H O N
18 6 2
made: *i) CH belongs to the chair materialized by the
3
35.03 Found: C, 48.10; H, 6.64; N, 10.27; O, 35.61.
N-1±C-3±O-4±C-5±N-6±C-11 sequence *see Fig. 2),
*
ii) consequently, it experiences two gauche interactions
4.1.2. 2ꢀR)-2-Methoxycarbonylthiazolidine ꢀ9). 60%;
1
1
1
1
with, respectively, the N-1pseudo-equatorial lone pair
and the C-2±N-1bond, *iii) in addition, it displays a
further unfavorable gauche interaction with the ester
group at C-2 *dihedral angle C-21±C-2±C-3±C-
FAB MS *NBA) m/z 148 [M1H] , 160 [M1Na] ; H
NMR *200.13 MHz, CDCl , J *Hz)): 2.50 *br. s, 1H, NH),
2.90 and 3.27 *ABX, 2H, 2H-3, Jˆ7.7 Hz), 3.80 *s, 3H,
CH
O), 3.88 *ABX, 1H, H-2, Jˆ7.5 Hz), 4.14 and 4.40
*AB, 2H, 2H-5, Jˆ9.6 Hz); C NMR *100.60 MHz,
CDCl , J *Hz)): 37.37 *C-3), 52.94 *CH O), 54.82 *C-5),
3
3
1
3
3
1ˆ438).
2
. The above interactions are either removed or drastically
reduced, in compound 13. Indeed, puckering of the ®ve-
membered oxazolidine rings increases the dihedral angle
between the methyl and the ester groups to approxi-
mately 1508 and concomitantly relieves the gauche inter-
action. The sum of the above non-bonding interactions
may reasonably account for the energy difference
observed between compounds 12 and 13 and very likely
for the speci®c behavior of l-threonine regarding its
condensation with formaldehyde.
3
3
65.60 *C-2), 172.21 *CvO).
0
4
.1.3. N,N -Methylenebis[ꢀ2R)-2-methoxycarbonylthia-
2
0
zolidine] 10. 80%; mp 36±378C; [a] ˆ2175.8 *c 1.0,
D
1
CHCl ); FAB MS *NBA) m/z 305 [M2H] , 160; H and
1
1
3
1
3
C NMR data see Table 1. Anal. calcd for C H O N S
2 2
11 18
4
*306): C, 43.13; H, 5.88; N, 9.15; O, 20.91; S, 20.91. Found:
C, 43.02; H, 5.92; N, 9.24; O, 21.21; S, 21.58.
4
.1.4.
[1S,2S,3R,6S,7S,8R]-1,6-Diaza-4,9-dioxa-2,7-di-
methoxycarbonyl-3,8-dimethylbicyclo[4.4.1]undecane
2
0
ꢀ12). 66%; mp 104±1068C; [a] ˆ228.1* c 1.0, CHCl );
D
3
1
1
1
FAB MS *NBA) m/z 303 [M1H] , 158; H and C NMR
13
3
. Summary and conclusion
data see Table 1. Anal. calcd for C H O N *302): C,
6
1
3
22
2
5
7
1.65; H, 7.28; N, 9.27; O, 31.78. Found: C, 51.60; H,
.39; N, 9.34; O, 31.76.
Elucidation of the structure of compounds 5, 10, 12, 13 by
NMR spectroscopy and/or X-ray crystallography
unambiguously established the following points: *i)
0
methyloxazolidine]
p
4
.1.5. N,N -Methylenebis[ꢀ2S,3R)-2-methoxycarbonyl-3-
31±328C;
n*N)!s *C±N) delocalization is absent in the N±C±N
ꢀ13).
10%;
mp
aminal group whereas strong anomeric effect is present in
both N±C±O groups as corroborated by bond distances and
bond angles, *ii) electron delocalization from n*N) to C±S
2
0
1
[
[
a]D ˆ272.5 *c 1.0, CHCl ); FAB MS *NBA) m/z 303
3
1
M1H] , 158; H and C NMR data see Table 1. Anal.
1
13
p
calcd for C H O N *302): C, 51.65; H, 7.28; N, 9.27; O,
2
s antibonding orbital occurs also in the sequence N±C±S
although to a lesser extent than in its N±C±O counterpart.
13 22
6
31.78. Found: C, 51.63; H, 7.25; N, 9.14; O, 31.94.
*
membered rings of N,N -methylenebis derivatives 10 and
iii) conformation of both oxazolidine and thiazolidine ®ve-
0
3, and seven-membered ring of bicyclo[4.4.1]undecane
4.2. Supplementary material
1
derivatives 5 and 12 are preserved in solution as con®rmed
Full crystallographic data *tables of crystallographic details,
non-hydrogen coordinates, bond distances and bond angles,
anisotropic thermal parameters, hydrogen coordinates and
isotropic thermal parameters) have been deposited with the
1
by NMR data. In that respect, it should be noted that the C
3
1
and H NMR chemical shifts of the N±CH ±N aminal
bridge as well as long-range coupling have proved to be
2

4444
J. S e lambarom et al. / Tetrahedron 58 ꢀ2002) 4439±4444
Cambridge Crystallographic Data Centre *CCDC no.
58098 and 158097 for compounds 10 and 12, respec-
4. *a) Deslonchamps, P. Stereoelectronic Effects in Organic
Chemistry; Pergamon: New York, 1983. *b) Praly, J. P.;
Lemieux, R. U. Can. J. Chem. 1987, 65, 213±223.
5. Stoddart, J. F. Sterochemistry of Carbohydrates; Wiley Inter-
science: New York, 1971; pp 55±58.
1
tively). This material is available on request from The
Director, CCDC, 12 Union Road, Cambridge CB21EZ,
UK *tel.: 144-1223-336408; fax: 144-1223-336033;
e-mail: deposit@ccdc.cam.ac.uk or http://www/ccdc.cam.
ac.uk).
6. Durette, P. L.; Horton, D. Adv. Carbohydr. Chem. Biochem.
1971, 26, 49±125.
7
. Juaristy, E.; Cuevas, G. Tetrahedron 1992, 48, 5019±5087.
. Tvaroska, I.; Carver, P. J. Phys. Chem. 1996, 100, 11305±
11313.
8
Acknowledgements
9
. Coss e -Barni, A.; Dubois, J. E. J. Am. Chem. Soc. 1987, 109,
1503±1511.
10. Monge, S.; S e lambarom, J.; Carr e , F.; Verducci, J.; Roque,
J. P.; Pavia, A. A. Carbohydr. Res. 2000, 328, 127±133.
11. S e lambarom, J.; Monge, S.; Carr e , F.; Fruchier, A.; Roque,
J. P.; Pavia, A. A. Carbohydr. Res. 2001, 330, 43±51.
The authors are indebted to the Minist eÁ re de l'Education
Nationale de la Recherche et de la Technologie for a grant
*
Allocation de recherche MENRT) to J. S e lambarom as well
as to R. Astier *Laboratoire Commun de Mesures
Physiques, Universit e de Montpellier II) for X-ray data
collection.
12. Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976, 734±736.
1
3. Baldwin, J. E.; Cutting, J.; Dupont, W.; Kruse, L.; Silberman,
L.; Thomas, R. C. J. Chem. Soc., Chem. Commun. 1976, 736±
References
7
38.
1
4. Alder, R. W.; Heilbronner, E.; Honegger, E.; MacEven, A. B.;
Moss, R. E.; Ole®rowicz, P.; Petillo, A.; Sessions, R. B.;
Weisman, G. R.; White, J. W.; Yang, Z. Z. J. Am. Chem.
Soc. 1993, 115, 6580±6591.
1
. *a) Lemieux, R. U.; Ch uÈ , N. J. In Abstract of Paper, 133rd
Meeting; American Chemical Society, 1958; p 31N. See
Lemieux, R. U. Rearrangements and Isomerisations in Carbo-
hydrate Chemistry. In Molecular Rearrangements; de Mayo,
P., Ed.; Intersciences: New York, 1964; p 709. *b) Lemieux,
R. U. Explorations with Sugars. How Sweet It Was. In
Pro®les, Pathways and Dreams. Autobiographies of Eminent
Chemists; Seeman, J. I., Ed.; American Chemical Society:
Washington, DC, 1990; pp 75±102 185p. *c) Lemieux,
R. U.; Koto, S. Tetrahedron 1974, 30, 1933±1944.
1
5. Alder, R. W.; Carniero, T. M. G.; Mowlan, R. W.; Orpen,
A. G.; Petillo, P. A.; Vachon, D. J.; Weisman, G. R.; White,
J. M. J. Chem. Soc., Perkin Trans. 2 1999, 589±599.
6. Lemoine, P.; Viossat, B.; Martin, P. G.; Aitken, D. J. Acta
Crystallogr. 1998, C54, 1036±1038.
1
1
1
7. Engel, V. J.; Tr oÈ mer, H. G.; Sheldrick, W. S. Chem. Ztg 1982,
1
06, 427±429.
8. *a) F uÈ lop, F.; Mattinen, J.; Pihlaja, K. Tetrahedron 1990, 46
18), 6545±6552. *b) Canle, M.; Lawley, A.; McManus, E. C.;
More O'Ferrall, R. A. Pure Appl. Chem. 1996, 68 *4), 813±
18.
2
. *a) Eliel, E. L. Angew. Chem., Int. Ed. Engl. 1972, 11, 739±
7
45. *b) Eliel, E. L.; Wilen, W. H. Stereochemistry of Organic
Compounds, Wiley Interscience: New York, 1994 pp 749±
54.
*
7
8
3
. Kirby, A. J. The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer: New York, 1983.