Boat conformations - Synthesis, NMR spectroscopy, and molecular dynamics of methyl 4,6-O-benzylidene-3-deoxy-3-phthalimido-α-D-altropyranoside derivatives

Article abstract of DOI:10.1016/S0008-6215(01)00036-2

Addition of the elements of phthalimide to methyl 2,3-anhydro-4,6-O-benzylidene-α-D-mannopyranoside (1) under fusion conditions has yielded methyl 4,6-O-benzylidene-3-deoxy-3-phthalimido-α-D-altropyranoside (2). The conformation of the pyranose ring of 2

Full text of DOI:10.1016/S0008-6215(01)00036-2

www.elsevier.nl/locate/carres
Carbohydrate Research 331 (2001) 461–467
Note
Boat conformationsꢀ Synthesis, NMR spectroscopy, and
molecular dynamics of methyl
4,6-O-benzylidene-3-deoxy-3-phthalimido-a-
D-
altropyranoside derivatives
Bruce Coxon,a,b,* Robert C. Reynoldsb,1
aNational Institute of Child Health and Human De6elopment, National Institutes of Health,
Bethesda MD 20892-2720, USA
bNational Institute of Standards and Technology, Gaithersburg MD 20899, USA
Received 17 October 2000; accepted 22 January 2001
Abstract
Addition of the elements of phthalimide to methyl 2,3-anhydro-4,6-O-benzylidene-a-
D
-mannopyranoside (1) under
fusion conditions has yielded methyl 4,6-O-benzylidene-3-deoxy-3-phthalimido-a- -altropyranoside (2). The confor-
D
1
mation of the pyranose ring of 2 has been shown to be non-chair by H NMR spectroscopy, in contrast to the
conformations of related derivatives having smaller substituents at C-3. Molecular dynamics simulations of 2 in
explicit chloroform-d solvent have indicated four principal conformational possibilities. Of these, the 7C5/1S5
chair/skew boat form 2d has the lowest potential energy, and is largely consistent with the observed vicinal H–1H
1
NMR coupling constants. © 2001 Elsevier Science Ltd. All rights reserved.
1
Keywords: Altropyranoside; Boat conformations; H NMR spectroscopy; Molecular dynamics; Phthalimido derivative; Simulated
annealing; Skew boat conformations; Vicinal coupling constants
1. Introduction
rings that are constrained to skew boat forms
by the presence of bridging anhydro or cyclic
acetal rings.1–3 These molecules are of particu-
lar interest for studies of the angular depen-
dence of homonuclear and heteronuclear
NMR coupling constants, because of elimina-
tion of contributions to the observed coupling
constants from alternative conformations.
Boat conformations are also of interest as
potential transition-state intermediates. The
data from such studies could be of particular
value in the structural and conformational
analysis of oligosaccharide components of
vaccines and aminoglycoside antibiotics. Also
In three previous papers in this series, we
have discussed some examples of pyranoid
ꢀ Part IV of a series, Boat conformations: for Parts I–III,
see Refs. 1–3.
* Corresponding author. Present address: National Insti-
tute of Child Health and Human Development, National
Institutes of Health, 6 Center Drive, MSC 2720, Bethesda
MD 20892-2720, USA. Tel.: +1-301-2954780; fax: +1-301-
2951435.
E-mail address: coxonb@mail.nih.gov (B. Coxon).
1 Present address: Southern Research Institute, 2000 Ninth
Avenue South, Birmingham AL 35205, USA.
0008-6215/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00036-2

462
B. Coxon, R.C. Reynolds / Carbohydrate Research 331 (2001) 461–467
of interest, is a determination of which types
of substituents have sufficiently large steric
repulsions to cause the pyranose ring to devi-
ate from the chair form. The 4,6-O-benzyli-
dene acetals of glycopyranosides having the
gluco, manno, altro, and allo configurations
(trans fused benzylidene ring) are usually
thought of as being conformationally homoge-
other example of non-chair conformations in
the methyl 4,6-O-benzylidene-3-deoxy-3-ph-
thalimido-a- -altropyranoside series. The
D
conformations of a number of related methyl
4,6-O-benzylidene-a- -altropyranoside deriva-
D
tives (2–7, see Scheme 1) have been studied by
1H NMR spectroscopy, and that of compound
2 also by molecular dynamics with simulated
annealing,11 with explicit treatment of the
chloroform-d solvent.
1
neous, and based on measurements of H–1H
NMR coupling constants, there is little doubt
that the vast majority of these derivatives
exists in a conformation in which the pyranose
ring is in the 4C1 form.4 Two interesting excep-
tions to this generalization are the observa-
tions that the pyranose rings of methyl
4,6-O-benzylidene-2-deoxy-2-C-pentachloro-
2. Results and discussion
Fusion of methyl 2,3-anhydro-4,6-O-ben-
zylidene-a-
ture of
phthalimide at 240 °C yielded methyl 4,6-O-
benzylidene-3-deoxy-3-phthalimido-a- -altro-
D
-mannopyranoside (1) with a mix-
phenyl-a-
-altropyranoside5 and methyl 4,6-
D
phthalimide
and
potassium
O-benzylidene-3-deoxy-3-diallylamino-a- -
D
altropyranoside6 adopt non-chair forms, and
some other examples of non-chair conforma-
tions are known that are due to the presence
of large substituent atoms or groups, and/or
the restrictive effects of acetal,7–9 and lactone
or lactam10 rings.
D
pyranoside (2) as a major product (see Scheme
1). The formation of a trans product by addi-
tion of the elements of phthalimide to the
epoxide ring of 1 parallels other reactions of
this type.2,12 The configuration of 2 was
proved by hydrazinolysis to the known methyl
During our studies of the synthesis of 15N-
labeled amino sugar derivatives in which the
3-amino-4,6-O-benzylidene-3-deoxy-a- -altro-
D
15
1
pyranoside (3). Phthalimido derivative 2 was
also characterized as its 2-O-acetyl (4) and
2-O-benzoyl (5) derivatives.
vicinal N and H nuclei have a defined stere-
ochemical orientation, we have discovered an-
1
NMR spectroscopy.—The H NMR spectra
of the a- -altropyranoside derivatives 2–7
D
were mostly well dispersed at 400 MHz and
1
were readily assigned. H chemical shifts and
coupling constants are reported in Tables 1
and 2, respectively. The small values of the
vicinal coupling constants J1,2, J2,3, and J3,4
(1.1–3.5 Hz) observed for the 3-amino-3-de-
oxy and 3-hydroxy derivatives 3 and 7, respec-
tively (see Table 2), are typical of the 4C1 chair
conformation.4 By contrast, the J1,2, J2,3, and
J3,4 values for the 3-deoxy-3-phthalimido
derivative 2 are much larger (6.8–9.2 Hz), and
these extreme values are consistent with either
2b, in which the pyranoid ring has the B2,5
1
form, or 2d with the pyranoid ring in a S5
skew form (structures 2a–2d).
The J1,2 and J2,3 values for the 2-O-acetyl 4
and 2-O-benzoyl 5 derivatives are intermedi-
ate in magnitude (4.4–4.6 and 5.7 Hz, respec-
tively) and probably reflect the effects of a
conformational equilibrium. However, in our
1
Scheme 1.
hands, measurements of the H NMR spectra

Table 1
1H NMR chemical shifts a of methyl 4,6-O-benzylidene-a-
D-altropyranoside derivatives
Derivative
R2
R3
H-1
H-2
H-3
H-4
H-5
H-6e
H-6a
OMe
OH
PhCH Ph
Phthalimido
7.689, 7.828
/
b
2
3
4
OH
OH
C8H4NO2
NH2
C8H4NO2
C8H4NO2
4.684
4.650
4.736
4.911
4.94
4.934
3.955
5.635
5.991
4.65
4.799
3.296
4.825
5.044
4.34
4.057
4.020
4.114
4.239
4.57
4.681
4.052
5.036
5.125
4.30
4.397
4.325
4.430
4.460
4.36
3.665
3.865
3.664
3.781
3.87
3.502
3.404
3.409
3.427
3.34
2.553
5.460
5.651
5.535
5.593
5.76
7.134–7.229
2.044 c
7.351–7.492
7.192–7.315
7.190–7.551
7.36, 7.53
OAc
OBz
C8H4NO2 OH
OH OH
7.694, 7.836
7.774
7.88
5
6 d
7 e
4.20
2.181, 2.883
4.663
4.012
4.101
3.963
4.206
4.339
3.829
3.442
5.630
7.337–7.514
a In ppm from internal Me4Si at 297 K. Estimated standard uncertainty 0.001 ppm.
3
b Phthalimido.
c Includes the NH2 signal.
(
d Literature data for an acetone-d6 (CD3COCD3) solution at 360 MHz.12
e Measured at 304 K.
461
467

464
B. Coxon, R.C. Reynolds / Carbohydrate Research 331 (2001) 461–467
Table 2
1H–1H NMR coupling constants a (Hz) of methyl 4,6-O-benzylidene-a-
D-altropyranoside derivatives
Derivative
R2
R3
J1,2
J1,3
J2,3
J3,4
J4,5
J5,6e
J5,6a
J6e,6a
b
c
2
3
4
5
OH
OH
OAc
OBz
C8H4NO2
OH
C8H4NO2
NH2
C8H4NO2
C8H4NO2
OH
6.8
1.1
4.4
4.6
2.0
1.2
8.7
2.4
5.7
5.7
2.2
2.4
9.2
3.5
8.3
8.5
3.9
3.0
10.0
9.9
10.1
10.0
9.5
5.1
4.4
5.0
5.1
5.2
5.1
9.9
10.0
10.0
10.2
9.8
10.4
10.2
10.4
10.4
9.8
0.9
6 d
7
OH
B1.1
9.8
10.0
10.2
a Measured for solutions in CDCl3 at 400 MHz, with an estimated standard uncertainty of 0.2 Hz.
b J2,OH 3.7 Hz
c Phthalimido.
d Literature data for an CD3COCD3 solution at 360 MHz; J3,OH 3.4 Hz.12
of 4 at temperatures down to −100 °C failed
to detect the signals of multiple conformers.
We have found previously12 that methyl 4,6-
mations, the 4,6-O-benzylidene ring exists in
7
the C5 chair form with the phenyl group
equatorial. According to this simulation for a
chloroform-d solution, the chair/skew boat
form 2d has the lowest potential energy, and
the 7C5/4C1 (2a), 7C5/B2,5 (2b), and 7C5/1,4B (2c)
conformations have higher relative energies of
58.2 (13.9), 28.9 (6.9), and 35.6 kJ mol−1 (8.5
kcal mol−1), respectively. By means of the
relationship DG= −RT ln K, equilibrium
constants were calculated from the potential
energy differences for the equilibria 2a X 2b,
2b X 2c, and 2c X 2d, giving K values of
1.26×105, 6.81×10−2, and 1.58×106, re-
spectively. From the K values, the BIOEQCALC
program was then used to calculate the con-
centrations of the conformational species,
from which the mole fractions of 2a, 2b, 2c,
O-benzylidene-2-deoxy-2-phthalimido-a- -al-
D
tropyranoside (6) shows small values of J1,2,
J2,3, and J3,4 (see Table 2), and, therefore,
exists predominantly with its pyranoid ring in
4
the C1 chair form. The difference with the
3-deoxy-3-phthalimido derivative 2 may be at-
tributed to the presence in the latter chair
form of a 1,3 syn–diaxial interaction between
the large phthalimido group and the methoxyl
group, an interaction that is absent in 6. In
marked contrast to the 2-phthalimido deriva-
tive 6, the 1H–1H coupling constants of
methyl 4,6-O-benzylidene-2-deoxy-2-C-pen-
tachlorophenyl-a- -altropyranoside 8 indicate
D
that it exists in a boat conformation.5 More-
over, our simulations show that the molecular
and 2d were calculated to be 7.38×10−11
,
volumes of the 2-C-pentachlorophenyl and 3-
9.29×10−6, 6.33×10−7, and 0.9999901, re-
spectively. It is apparent that the energetics of
the simulation of 2 in explicit chloroform-d
favors greatly the skew boat form 2d.
3
,
phthalimido derivatives are 341 and 297 A ,
respectively, a difference that must be due
mainly to the larger size of the 2-C-pen-
tachlorophenyl group.
The dihedral angles that were measured on
the energy-minimized, principal conforma-
tional possibilities are shown in Table 3. The
7C5/4C1 chair/chair conformation 2a has the
pyranose ring in a flattened chair form in
which ƒ is distorted from ꢀ60 to 90°, a
dihedral 2a,3ngle that is totally inconsistent with
the observed value J2,3 8.7 Hz. At least in the
molecular model 2a, the splaying out of the
syn–axial phthalimido and methoxyl groups,
and consequent flattening of the pyranose ring
may be attributed to steric repulsion between
Molecular modeling.—The 3-phthalimido
derivative 2 was simulated by molecular dy-
namics with explicit chloroform-d solvent and
the use of the CVFF forcefield.13 The appli-
cability of this and other forcefields to carbo-
hydrates has recently been reviewed.14 During
simulated annealing from 800 down to 300 K,
molecule 2 made excursions to four principal
7
conformations, including the C5/4C17 chair/
7
chair (2a), C5/B2,5 chair/boat (2b), C5/1,4B
7
chair/boat (2c), and C5/1S5 chair/skew boat
(2d, see Scheme 2)2. In each of these confor-
7
these groups. In the C5/B2,5 chair/boat con-
2 C-7 is the benzylic carbon atom.
formation 2b, all substituents on the pyranose

B. Coxon, R.C. Reynolds / Carbohydrate Research 331 (2001) 461–467
465
ring are equatorial, although the C-1ꢀO-5 and
C-3ꢀC-4 bonds are not quite parallel, with C-3
tilted up slightly above the C-1ꢀC-4ꢀO-5
atoms, a supposition that might be made on
the basis that the 2-O-acyl derivatives 4 and 5
(in which this type of bonding is not possible)
appear to exist as a mixture of conformations.
However, the molecular modeling of 2a, 2b,
2c, and 2d did not reveal any evidence of such
hydrogen bonding in terms of the close ap-
proach or overlap of the hydroxyl hydrogen
and carbonyl oxygen atomic spheres. Mea-
surements of the energy-minimized molecular
models indicated hydroxyl hydrogen atom–
carbonyl oxygen atom distances of 5.2 and
5.0, 4.0 and 3.3, 4.7 and 5.5, and 4.7 and 4.5
7
plane. Similarly, in the C5/1,4B chair/boat
conformation 2c, the C-2ꢀC-3 and C-5ꢀO-5
bonds are not quite parallel and C-2 is tilted
up slightly above the C-3ꢀC-5ꢀO-5 plane. In
2c, the phthalimido group occupies a less fa-
vorable axial orientation, but changes to a
more favorable quasi-equatorial orientation in
the chair/skew boat 2d.
In principle, the chair/boat conformations
2b and 2c and chair/skew boat conformation
2d could be stabilized by hydrogen bonding
between the 2-hydroxyl hydrogen atom and
one of the phthalimido carbonyl oxygen
,
A, respectively for 2a, 2b, 2c, and 2d. All of
these distances appear to be rather long for
strong hydrogen bonding, but it may be sig-
Scheme 2.
Table 3
Dihedral angles a (°) of methyl 4,6-O-benzylidene-3-deoxy-3-phthalimido-a-
D-altropyranoside (2) conformers obtained from
molecular dynamics models b
Atom pair
H-1, H-2
H-2, H3
H-3, H-4
H-4, H-5
H-5, H-6e
H-5, H-6a
Conformer
7C5/4C1 (2a)
7C5/B2,5 (2b)
7C5/1,4B (2c)
7C5/1S5 (2d)
−74.9
−161.8
−159.1
−171.0
90.0
178.7
121.0
158.7
37.4
−11.0
46.0
−177.9
−174.0
−174.0
176.1
−56.1
−57.8
−60.8
−52.2
−172.8
−175.0
−177.9
−168.6
21.9
a Estimated standard uncertainty 0.1°.
b Energy minimization with the CVFF forcefield.13

466
B. Coxon, R.C. Reynolds / Carbohydrate Research 331 (2001) 461–467
contained 10–11 mg of compound in chloro-
form-d (0.45 mL) containing Me4Si as an in-
ternal reference. Spectral assignments were
confirmed by matching of spin multiplet spac-
ings and homonuclear spin decoupling
experiments.
Molecular modeling.—Molecular modeling
computations were performed by using one
processor of a SGI Onyx 2 multiprocessor
system equipped with eight R-10000 250 MHz
CPUs and Molecular Simulations Inc (MSI)/
BIOSYM INSIGHT II/DISCOVER software, ver-
sion 980. Three-dimensional structures were
built in the MSI/BIOSYM SKETCHER program,
version 950 running on an SGI R4400 200
MHz workstation. Energy computations were
performed by using the CVFF forcefield.13
The calculations were done with explicit chlo-
roform-d solvent at a density of 1.488 g/cc,
under periodic boundary conditions, using a
nificant that the distance of closest approach
of the hydroxyl hydrogen and carbonyl oxy-
,
gen atoms is 3.3 A, in the chair/boat 2b.
1
Taken together, the results of H NMR spec-
troscopy and molecular modeling favor 2d as
the most stable conformer, with minute pro-
portions of 2a, 2b, and 2c.
Clearly, it is the juxtaposition of an axial,
large substituent at C-3 with the syn–axial
methoxyl and H-5 that causes steric instability
in the chair/chair conformations of the N,N-
disubstituted methyl 3-amino-4,6-O-benzyli-
dene-3-deoxy-a- -altropyranoside derivatives,
D
thus leading to the adoption of a non-chair
conformation. In this sense, ‘large substituent’
implies that the nitrogen atom bonded to C-3
has two bulky non-hydrogen, substituents
such as two allyl groups6 or the carbonyl
groups of the phthalimido moiety. Notewor-
thy is the fact, that if these large groups
attached to nitrogen are absent (as in the
3-amino derivative 3), then the chair/chair
conformation is favored. In compound 8, the
larger size of the pentachlorophenyl group
causes the adoption of a non-chair form by
the pyranose ring,5 even when this group is
syn–diaxial with only one hydrogen atom (H-
4) in the chair/chair conformation. At this
juncture, no value is available for the energy
barrier between the skew boat cycle and the
chair form.
,
unit cell size of 35×35×35 A, a cutoff dis-
tance of 17 A, and a dielectric constant of 1.0.
,
Initially, a steepest descent energy minimiza-
tion was conducted for 3000 steps, followed
by a VA09A minimization for 1000 steps,
until a maximum rms derivative of 4.2 kJ
−1
mol−1
(1.0 kcal mol−1
A
−1) was
,
,
A
reached. Molecular dynamics with simulated
annealing was then performed over the tem-
perature range 800–300 K, with 50 K decre-
ments.
At
each
temperature,
1000
equilibration steps of 1 fs each were run,
followed by 5000 dynamics steps of 1 fs each,
and then a VA09A energy minimization, typi-
cally with B400 iterations. The conformation
and the potential energy value were inspected
after each minimization. The energy of each
major conformer was then exhaustively mini-
mized by further VA09A computations, fol-
lowed by conjugate gradient minimization (up
3. Experimental
General.—Reactions were monitored by
thin layer chromatography (TLC) on Silica
Gel G plates (E. Merck) that were eluted with
97:3 (v/v) CH2Cl2–MeOH. Compounds were
detected by charring with 5% aq H2SO4 and
the amino derivative was deleted by reaction
to 3000 steps) to a maximum rms derivative of
with
ninhydrin.
Melting
points
are
−1
0.0004 kJ mol−1
A
(0.0001 kcal mol−1
,
A
uncorrected.
−1). Finally, H–1H dihedral angles were
1
NMR spectroscopy.—1H NMR spectra
were recorded at 400 MHz and 297 K by
using a Bruker Instruments WM 400 spec-
trometer. A 32,768 point data set was used,
together with a spectral width of 3.26 kHz, a
90° pulse (30 ms), and a pulse recycle time of
8.52 s, giving a digital resolution of 0.2 Hz/
point in the Fourier transform. The solutions
,
measured on the annealed, energy-minimized
final structure of each major conformer by
means of the INSIGHTII/DISCOVER software.
The configuration of each chiral carbon atom
was monitored for constancy during the com-
putations using the INSIGHTII/DISCOVER
software.

B. Coxon, R.C. Reynolds / Carbohydrate Research 331 (2001) 461–467
467
Calculations on equilibrium constants were
performed by using the BIOEQCALC module
running within the MATHEMATICA program
on a personal computer.15
(116 mg) in dry Py (1 mL). The solution was
kept at rt for 16 h, poured into ice–NaHCO3,
and then extracted with CH2Cl2 (3×20 mL).
Drying (Na2SO4) and concentration of the
combined extracts, followed by crystallization
from EtOH, yielded the 2-O-acetyl derivative
4 as needles (106 mg, 83%): mp 186–187 °C,
unchanged by further recrystallization. Anal.
Calcd for C24H23NO8: C, 63.57; H, 5.11; N,
3.09. Found: C, 63.36; H, 5.26; N, 3.07. The
2-O-benzoyl derivative 5 was prepared (as an
oil) in a similar way by treatment of 2 with
Py–BzCl.
Methyl 4,6-O-benzylidene-3-deoxy-3-phthal-
imido-h- -altropyranoside (2).—A mixture of
D
phthalimide (1.34 g, 9.11 mmol) and potas-
sium phthalimide (0.082 g, 0.44 mmol) was
heated at 238 °C for 5 min, then treated with
methyl
2,3-anhydro-4,6-O-benzylidene-a- -
D
mannopyranoside (1)16 (1.00 g, 3.79 mmol).
The mixture was heated at 240 °C for 30 min,
and the resulting dark residue was extracted
with CH2Cl2 (60 mL). The dark extract was
cooled in the freezer, washed with ice-cold 5%
aq NaOH (50 mL), followed by water (5×
200 mL), dried (anhyd Na2SO4), and decol-
orized (Norit A charcoal). Filtration of the
extract through a Norit A/Celite pad, fol-
lowed by evaporation, yielded crude 2 as a
pale-yellow syrup (0.71 g, 45%) that was crys-
tallized from aq MeOH to give 2 as colorless
plates (0.56 g): mp 168–169 °C. IR wmax
(CHCl3) 3675w and 3590m (OH), 1785m and
1720s (fused lactam ring CꢁO), and 1619w
cm−1 (Ar). Anal. Calcd for C22H21NO7: C,
64.22; H, 5.15; N, 3.41. Found: C, 64.09; H,
5.24; N, 3.77. Seed crystals were obtained
from a preliminary preparation in which chro-
matography on a silicic acid column was used
to purify the product.
Acknowledgements
Thanks are due Dr Robert Goldberg for
help with the equilibrium calculations and Ms
M.L. Luzarraga for technical assistance.
References
[1] Coxon, B. Carbohydr. Res. 1970, 13, 321–330.
[2] Coxon, B. Carbohydr. Res. 1999, 322, 120–127.
[3] Coxon, B. Carbohydr. Res. 2000, 329, 131–139.
[4] Coxon, B. Tetrahedron 1965, 21, 3481–3503.
[5] Lee, J. B.; Scanlon, B. J. Chem. Soc., Chem. Commun.
1969, 955–956.
[6] Picq, D.; Carret, G.; Anker, D. Carbohydr. Res. 1986,
149, 458–463.
[7] Lin, G. H.-Y.; Sundaralingam, M.; Jackobs, J. Carbo-
hydr. Res. 1973, 29, 439–449.
[8] Bissember, B. B.; Wightman, R. H. Carbohydr. Res. 1981,
91, 89–92.
[9] Miethchen; R., Rentsch; D., Michalik; M. Liebigs Ann.
Chem. 1994, 219–222.
[10] Kovarikova, R.; Ledvina, M.; Saman, D. Collect. Czech.
Chem. Commun. 1999, 64, 673–684.
[11] Homans, S. W.; Forster, M. Glycobiology 1992, 2, 143–
151.
[12] Coxon, B.; Reynolds, R. C. Carbohydr. Res. 1982, 110,
43–54.
[13] Dauber-Osguthorpe, P.; Roberts, V. A.; Osguthorpe, D.
J.; Wolff, J.; Genest, M.; Hagler, A. T. Proteins 1988, 4,
31–47.
[14] Pe´rez, S.; Imberty, A.; Engelsen, S. B.; Gruza, J.;
Mazeau, K.; Jimenez-Barbero, J.; Poveda, A.; Espinosa,
J.-F.; van Eyck, B. P.; Johnson, G.; French, A. D.;
Kouwijzer, M. L. C. E.; Grootenuis, P. D. J.; Bernardi,
A.; Raimondi, L.; Senderowitz, H.; Durier, V.; Vergoten,
G.; Rasmussen, K. Carbohydr. Res. 1998, 314, 141–155.
[15] Goldberg, R. personal communication.
[16] Robertson; G. J., Griffith; C.F. J. Chem. Soc. 1935,
1193–1201.
Methyl
3-amino-4,6-O-benzylidene-3-de-
oxy-h- -altropyranoside (3).—A solution of
D
the phthalimido derivative 2 (0.21 g) in warm
EtOH (10 mL) was treated with 85% aq hy-
drazine hydrate (0.07 mL), and the mixture
was boiled under reflux for 2 h. Evaporation
of the solution yielded a solid that was treated
with 5% aq KOH. The mixture was extracted
with CH2Cl2 (3×15 mL), and the combined
extracts were washed with water (2×20 mL),
dried (Na2SO4), and evaporated to clear
syrup. Crystallization from warm EtOH af-
forded two crops of plates (81 mg, 56%): mp
183–186 °C, which on recrystallization from
EtOH yielded pure 3, mp 188–190 °C. Lit.17
mp 188–190 °C.
Methyl 2-O-acetyl-4,6-O-benzylidene-3-de-
oxy-3-phthalimido-h- -altropyranoside (4).—
D
[17] Myers, W. H.; Robertson, G. J. J. Am. Chem. Soc. 1943,
65, 8–11.
Ac2O (1 mL) was added to a solution of 2
.
.