Septanose carbohydrates: Synthesis and conformational studies of methyl α-D-glycero-D-idoseptanoside and methyl β-D-glycero-D-guloseptanoside

Article abstract of DOI:10.1021/jo048932z

(Chemical Equation Presented). We report the synthesis of methyl α-D-glycero-D-idoseptanoside (1) and methyl β-D-glycero-D- guloseptanoside (2) and the characterization of their preferred solution conformations by computational chemistry and ¹H

Full text of DOI:10.1021/jo048932z

Septanose Carbohydrates: Synthesis and Conformational Studies
of Methyl r-D-glycero-D-Idoseptanoside and Methyl
â-D-glycero-D-Guloseptanoside
Matthew P. DeMatteo,† Nicole L. Snyder,‡ Martha Morton,‡ Donna M. Baldisseri,§
Christopher M. Hadad,*,† and Mark W. Peczuh*,‡
Department of Chemistry, The Ohio State University, 100 West 18th Avenue, Columbus, Ohio 43210,
Department of Chemistry, The University of Connecticut, 55 North Eagleville Road,
Storrs, Connecticut 06269, and Bru¨kerBioSpin Corporation, 15 Fortune Drive, Manning Park,
Billerica, Massachsetts 01821
mark.peczuh@uconn.edu
Received June 24, 2004
We report the synthesis of methyl R-D-glycero-D-idoseptanoside (1) and methyl â-D-glycero-D-
guloseptanoside (2) and the characterization of their preferred solution conformations by compu-
tational chemistry and 1H NMR 3JH,H coupling constant analysis. Central to the synthetic approach
was the epoxidation of glucose-derived oxepine 3 using DMDO. Nucleophilic attack on the resulting
1,2-anhydroseptanose using NaOCH3 in CH3OH followed by deprotection provided the 1,2-trans
diastereomers 1 and 2. The computational approach for determining the preferred low energy
septanose conformations began with a pseudo Monte Carlo search for each isomer using
minimization with the AMBER force field. Single-point energy calculations (HF/6-31G* and B3LYP/
6-31+G**) as well as full geometry optimizations in a model for aqueous solvent were then conducted
3
using the conformers within 5 kcal/mol of the AMBER global minimum. Calculated JH,H values,
based on a Boltzmann distribution of the computed low energy conformers, were compared to
3
experimental JH,H values from 1H NMR coupling constant analyses. The correlation between
calculated and observed values suggest that septanose carbohydrates are not so flexible and should
generally prefer one twist-chair (TC) conformation.
I. Introduction
processes and the manipulation of biological pathways
using designed molecules. Expanded nucleic acid2 and
protein3 analogues have been reported, and both have
yielded interesting new oligomeric structures with novel
biological activity. The common theme in these strategies
There is growing interest in the development of size-
expanded analogues of natural biopolymers that can
interact predictably with biological systems.1 The new,
unnatural oligomers that are the result of these inves-
tigations are instrumental for the exploration of cellular
(2) (a) Eschenmoser, A. Science 1999, 284, 2118. (b) Liu, H.; Gao,
J.; Lynch, S. R.; Saito, D.; Maynard, L.; Kool, E. T. Science 2003, 302,
868.
(3) (a) Porter, E. A.; Weisblum, B.; Gellman, S. H. J. Am. Chem.
Soc. 2002, 124, 7324. (b) Porter, E. A.; Wang, X.; Lee, H.-S.; Weisblum,
B.; Gellman, S. H. Nature 2000, 404, 585. (c) Gellman, S. H. Acc. Chem.
Res. 1998, 31, 173. (b) Gademan, K.; Ernst, M.; Hoyer, D.; Seebach,
D. Angew. Chem., Int. Ed. 1999, 38, 1223.
† The Ohio State University.
‡ The University of Connecticut.
§ Bru¨kerBioSpin Corp.
(1) (a) Rawls, R. L. Chem. Eng. News 2000, 78, 49. (b) Benner, S.
A. Nature 2003, 421, 118.
10.1021/jo048932z CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/08/2004
24
J. Org. Chem. 2005, 70, 24-38

Septanose Carbohydrates
CHART 1
to be ring-expanded glycals that allow for the synthesis
of a variety of septanose donors and glycoconjugates. In
fact, toward this end, we have recently reported the
oxidative coupling of oxepine 4 with a number of nucleo-
philes.10
At the outset, we endeavored to prepare monoseptano-
sides that contained an exocyclic hydroxymethyl group
at C6, such as 1 and 2, and to determine their preferred
low energy conformations. We were interested in eluci-
dating the dominant factors that govern septanose ring
conformational analysis, including the anomeric effect,
exo-anomeric effect, and a correlation with the different
hydroxymethyl rotamers around the C6-C7 bond. Par-
ticular among these interests was the orientation of the
hydroxymethyl group, an issue that has not been ad-
dressed prior to this report. The information gathered
here provides insight into the conformation of the sep-
tanose ring in general and will be used as a starting point
for the development of bioactive septanosides.
has been the homologation of natural monomer residues
by one carbon and their subsequent oligomerization. The
homologation of natural carbohydrates to create similar
size-expanded analogues has received less attention.
Septanose carbohydrates, such as 1 and 2, are homo-
logues of pyranoses in which the sugar ring is expanded
by one atom (Chart 1). Just as in the nucleic acid and
protein systems, septanose carbohydrates should provide
interesting new molecular structures and biological
activity. The major obstacles to the utilization of ring-
expanded (septanose) carbohydrates for chemical or
biological applications are the current inability to easily
prepare these molecules as well as our insufficient
understanding of their low energy conformations in
solution. Implicit to the synthetic challenge is to develop
septanose monosaccharides that can participate as do-
nors in glycosylation reactions. The synthesis and con-
formational analysis of methyl R-D-glycero-D-idoseptano-
side (1) and methyl â-D-glycero-D-guloseptanoside (2)
reported here is our starting point to address these
challenges.
Previous investigations on septanoses have focused
largely on the preparation of monoseptanosides.4 Cycliza-
tion of pyranoses through the C6 hydroxyl group to form
a seven-membered ring was central to many of the
strategies. Selective epimerization of protected glucosep-
tanosides5 prepared from the C6 cyclization route have
provided L-ido,6 L-altro, D-gulo, and D-galactoseptano-
sides.7 A variation of the cyclization approach has been
used to prepare pyranosyl septanosides.8 Acyclic chlo-
rothioethyl acetals were used as donors in glycosylation
reactions to give mixed O,S-acetals. The primary alcohol
at C6 of the erstwhile donor was then selectively depro-
tected and used in the cyclization to give pyranosyl
septanoside products. The structural consequence of
these examples is that the septanose products lack an
exocyclic hydroxymethyl group. The exocyclic hydroxy-
methyl group is characteristic of many natural pyranose
and furanose monosaccharides, and we suspect that it
may be important for the mimicry of natural pyranoses.
Our strategy for the homologation of pyranose sugars
relies on oxepines, such as 3, and leaves the C6 function-
ality intact.9 We consider carbohydrate-based oxepines
II. Results
A. Synthetic Methods. The broad utility of glycals
in carbohydrate chemistry has motivated our synthetic
strategy. Glycals can serve directly as donors under
appropriate conditions to form a variety of glycosides;11
they can also be used in the preparation of other classes
of donors.12 We have introduced a synthesis of carbohy-
drate-based oxepines using a ring-closing metathesis
(RCM) strategy.9 The route followed other successful
precedents of RCM with carbohydrate substrates.13 The
resulting oxepines (such as 3, 4) are seven-membered
ring cyclic enol ethers that can serve as glycosyl donors
in a similar manner as glycals. Our recent report of the
epoxidation of a related D-xylose-based oxepine (4) using
dimethyldioxirane (DMDO) selectively gave 1,2-anhydro-
â-D-idoseptanose and was trapped by a number of nu-
cleophiles to give the corresponding R-idoseptanosides.10
A key point is that the characterization of the epoxidation
stereochemistry in the D-xylose based oxepine (4) series
was facilitated by comparison of a methyl septanoside
derivative to a compound previously reported in the
literature.6c
The absence of reference compounds for reactions using
the D-glucose-based oxepine (3) meant that a similar
(9) Peczuh, M. W.; Synder, N. L. Tetrahedron Lett. 2003, 44, 4057.
(10) Peczuh, M. W.; Snyder, N. L.; Fyvie, W. S. Carbohydr. Res.
2004, 339, 1163.
(11) (a) Halcomb, R. L.; Danishefsky, S. J. J. Am. Chem. Soc. 1989,
111, 6661. (b) Honda, E.; Gin, D. Y. J. Am. Chem. Soc. 2002, 124, 7343.
(c) Kim, J. Y.; Di Bussolo, V.; Gin, D. Y. Org. Lett. 2001, 3, 303. (d)
Friesen, R. W.; Danishefsky, S. J. J. Am. Chem. Soc. 1989, 111, 6656.
(e) Griffith, D. A.; Danishefsky, S. J. J. Am. Chem. Soc. 1990, 112,
5811. (f) Di Bussolo, V.; Liu, J.; Huffman, L. G., Jr.; Gin, D. Y. Angew.
Chem., Int. Ed. 2000, 39, 204. (g) Liu, J.; Gin, D. Y. J. Am. Chem. Soc.
2002, 124, 9789. (h) Storkey, C. M.; Win, A. L.; Hoberg, J. O.
Carbohydr. Res. 2004, 339, 897-899. (i) Batchelor, R.; Hoberg, J. O.
Tetrahedron Lett. 2003, 44, 9043-9045. (j) Hoberg, J. O. J. Org. Chem.
1997, 62, 6615-6618.
(4) Pakulski, Z. Pol. J. Chem. 1996, 70, 667. An alternate route to
mono-septanosides has also been observed recently. See: Enright, P.
M.; Tosin, M.; Nieuwenhuyzen, M.; Cronin, L.; Murphy, P. V. J. Org.
Chem. 2002, 67, 3733.
(5) (a) Stevens, J. D. Aust. J. Chem. 1975, 28, 525. (b) Ng, C. J.;
Stevens, J. D. Carbohydr. Res. 1996, 284, 241.
(12) Thiophenyl glycosides: (a) Boulineau, F. P.; Wei, A. Org. Lett.
2002, 4, 2281. (b) Gordon, D. M.; Danishefsky, S. J. Carbohydr. Res.
1990, 206, 361. Thioethyl glycosides: (c) Seeberger, P. H.; Eckhardt,
M.; Gutteridge, C. E.; Danishefsky, S. J. J. Am. Chem. Soc. 1997, 119,
10064. Glycosyl phosphates: (d) Plante, O. J.; Andrade, R. B.; See-
berger, P. H. Org. Lett. 1999, 1, 211.
(6) (a) Driver, G. E.; Stevens, J. D. Aust. J. Chem. 1990, 43, 2063.
(b) Ng, C. J.; Craig, D. C.; Stevens, J. D. Carbohydr. Res. 1996, 284,
249. (c) Tran, T. Q.; Stevens, J. D. Aust. J. Chem. 2002, 55, 171.
(7) Driver, G. E.; Stevens, J. D. Carbohydr. Res. 2001, 334, 81.
(8) McAuliffe, J. C.; Hindsgaul, O. Synlett 1998, 307.
(13) (a) Allwein, S. P.; Cox, J. M.; Howard, B. E.; Johnson, H. W.
B.; Rainier, J. D. Tetrahedron 2002, 58, 1997. (b) Postema, M. H. D.;
Piper, J. L.; Liu, L.; Shen, J.; Faust, M.; Andreana, P. J. Org. Chem.
2003, 68, 4748. (c) Liu, L.; Postema, M. H. D. J. Am. Chem. Soc. 2001,
123, 8602.
J. Org. Chem, Vol. 70, No. 1, 2005 25

DeMatteo et al.
SCHEME 1
CHART 2
product analysis for the determination of the stereose-
lectivity in epoxidation of 3 was not possible. The prospect
of using computational and spectroscopic methods in a
synergistic manner to assign the stereochemistry of 1 and
2 gave us additional impetus to use a computational
approach for determining the low-energy conformations
of 1 and 2.
among others,15c have been described by modeling and
spectroscopy. Low energy conformations of septanose
carbohydrates have been reported in the solid state18 and
in solution,19 but computational work on septanoses has
not been described to date. Seven-membered ring systems
in general show four low energy conformations; they are
the twist-chair (TC), chair (C), twist-boat (TB), and boat
(B) conformers (Chart 2) with the TC being most stable.
For symmetric species such as cycloheptane, all of the
TC conformers are energetically equivalent. Substituted
rings of lower symmetry, however, have multiple and
unique TC conformers; in fact, 14 unique TC conformers
are possible with septanose carbohydrates. The nomen-
clature20 for them is illustrated using the 3,4TC5,6 con-
former of 1 (Chart 2). For a given TC, three atoms define
a molecular plane, and the remaining atoms lie above or
below this plane. In the case of 1, the ring oxygen, C1,
and C2 are coplanar with each other, while C3 and C4
are above the plane and C5 and C6 are below the plane.
An axis of pseudosymmetry, present in all of the TC
conformers, is centered on C1 of the 3,4TC5,6 conformation.
The atom and its substituents at this position are said
to be isoclinal; because of the pseudosymmetry axis,
isoclinal substituents experience the similar steric envi-
ronments whether on the top or bottom face of the
ring.15b,16a,21
The synthesis of 1 and 2 is shown in Scheme 1.
Epoxidation of 3 using DMDO under standard conditions
gave the corresponding anhydroseptanose 5.14 The se-
lectivity in epoxidation of 3 was much lower than that
observed when using oxepine 4; the origin of the dete-
riorated stereoselectivity is uncertain, but is being in-
vestigated. Nucleophilic opening of anhydroseptanose 5
using NaOCH3/CH3OH gave methyl septanosides 6 and
7 in 63% combined yield in a 1:3 (6:7) ratio. The product
mixture was separated by column chromatography, and
the benzyl groups were removed by hydrogenolysis for
each isomer to give 1 and 2 in 95% and 98% yield,
respectively. Based on the relative similarity of the NMR
spectra and the paucity of reference data for D-glycero-
D-septanosides, assignment of compound 1 versus 2
proved somewhat problematic. As a result, we ap-
proached the assignment of 1 and 2 through a comparison
of computational and spectroscopic data.
B. General Conformational Analysis of Seven-
Membered Rings. Conformational preferences of cyclo-
heptane,15 oxepane,15d,16 and 1,3-/1,4-dioxepanes,15d,17
(18) (a) Bailey, C. J.; Craig, D. C.; Grainger, C. T.; James, V. J.;
Stevens, J. D. Carbohydr. Res. 1996, 284, 265. (b) Craig, D. C.; Stevens,
J. D. Aust. J. Chem. 1990, 43, 2083. (c) Foster, S. J. S. J.; James, V.
J.; Stevens, J. D. Acta Crystallogr. 1989, C45, 1329. (d) Wood, R. A.;
James, V. J.; Rae, D. A.; Stevens, J. D.; Moore, F. H. Aust. J. Chem.
1983, 36, 2269. (e) Foster, S. J.; James, V. J.; Stevens, J. D. Acta
Crystallogr. 1983, C39, 610. (f) Choong, W.; McConnell, J. F.; Stephen-
son, N.; Stevens, J. D. Aust. J. Chem. 1980, 33, 979. (g) James, V. J.;
Stevens, J. D. Carbohydr. Res. 1980, 82, 167. (h) McConnell, J. F.;
Stevens, J. D. J. Chem. Soc., Perkin Trans. 2 1974, 354. (i) Beale, J.
P.; Stephenson, N. C.; Stevens, J. D. Acta Crystallogr. 1972, B28, 3115.
(j) Stevens, J. D.; Beale, J. P.; Stephenson, N. C. J. Chem. Soc., Chem.
Commun. 1971, 10, 484.
(14) Due to its relative lability, anhydroseptanose 5 was not isolated.
It was routinely allowed to react with NaOCH3/CH3OH directly after
its preparation. See the Supporting Information for the NMR spectra
of 5.
(15) (a) Bocian, D. F.; Pickett, H. M.; Rounds, T. C.; Strauss, H. L.
J. Am. Chem. Soc. 1975, 97, 687. (b) Hendrickson, J. B. J. Am. Chem.
Soc. 1967, 89, 7043. (c) Hendrickson, J. B. J. Am. Chem. Soc. 1967,
89, 7047. (d) Bocain. D. F.; Strauss, H. L. J. Am. Chem. Soc. 1977, 99,
2876. (e) Favini, G. THEOCHEM 1983, 93, 139.
(16) (a) Entrena, A.; Campos, J.; Go´mez, J. A.; Gallo, M. A.;
Espinosa, A. J. Org. Chem. 1997, 62, 337. (b) Espinosa, A.; Gallo, M.
A.; Entrena, A.; Go´mez, J. A. J. Mol. Struct. 1994, 323, 247.
(17) Espinosa, A.; Entrena, A.; Gallo, M. A.; Campos, J.; Dom´ınguez,
J. F.; Camacho, E.; Garrido, R. J. Org. Chem. 1990, 55, 6018
(19) See refs 5-7 and also: Gonzalez, F. S.; Berenguel, A. V.; Molina,
J. M. Carbohydr. Res. 1991, 209, 155.
(20) Stoddart, J. F. Stereochemistry of Carbohydrates; Wiley and
Sons: New York, 1971.
26 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
An appropriately parametrized description of ring
puckering provides an itinerary for the interconversion
of low-energy conformations of flexible rings. For five-
membered (furanose) rings, the phase angle (P) and
puckering amplitude (τm) parameters define a pseudoro-
tational itinerary (wheel) in which individual conformers
are points on that wheel.22 For six-membered (pyranose)
rings, the generalized description defined by Cremer and
Pople correlates points on the surface of a sphere using
a puckering amplitude (Q) and two angular variables, θ
and φ.23 The Pople treatment proposes that the “surface
of a hypersphere in N-3 dimensions” will describe the
conformations in higher rings. Bocian and Strauss have
used similar variables (F, θ, φ1, φ2) to describe conforma-
tions on the surface of a torus.15 A cross-section of the
twist-chair (TC) to chair (C) surface shows the relative
conformational energy as a function of the pseudorota-
tional phase angle. A similar itinerary can be developed
for the TB-B interconversion; however, in substituted
systems, the TB and B conformers are high enough in
energy that they are not readily populated. Increased
steric interactions between substituents and the ring
atoms in the TB and B conformers explain this general
preference.16 The work reported here is concerned only
with identifying the low-energy conformations of 1 and
2; a description of their interconversion between alterna-
tive conformations will also require treatment of the
above considerations.
Some general principles that govern septanose ring
conformations can be derived from the systems noted
above and from the structural information gathered on
known examples of septanose carbohydrates. First, in
each case, a TC conformation is almost exclusively the
lowest energy conformation. Second, in the simpler
systems, a distribution of multiple low-energy conformers
of similar energies underscores their flexibility. Compu-
tational treatments of septanose carbohydrates are ab-
sent from the literature, but reported aqueous solution
structures assign only one conformer.19 This apparent
rigidity of monoseptanosides is therefore in contrast with
the conformational flexibility that has been noted in the
simpler seven-membered ring systems. Third, factors
such as steric interactions between substituents, and
electronic effects, such as the anomeric effect,24,25 can
significantly influence which TC conformer is most stable.
Of special note here are the steric factors associated with
substituents on the isoclinal carbon. Several examples
show that favored conformations will put large axial
substituents (such as the OCH3 group of the aglycon in
septanoside 1) into the isoclinal position so as to experi-
ence diminished steric interactions with the rest of the
ring. Restricted rotation about the C1-O1 bond is also
observed according to the exo-anomeric effect in calcu-
lated structures.26 In total, the principles operative in
determining preferred conformations of furanose and
pyranose rings seem to be operative in the septanose
series.
C. General Computational Methods. The System-
atic Pseudo Monte Carlo search protocol available in
Macromodel version 6.527 was used to generate 50 000
conformers of 1 and 2. These structures were then
optimized with the AMBER28 force field using the dielec-
tric constant of water; two Monte Carlo searches were
performed for each compound. The first search merely
varied the exocyclic dihedral angles and allowed the ring
to adopt a suitable conformation while the second search
varied both the exocyclic dihedral angles as well as the
endocyclic dihedral angles within the septanose ring.
Generating structural and conformational diversity was
very important, and we utilized these Monte Carlo
searches to guarantee a variety of conformations for
further analysis. The first Monte Carlo search of 1, using
the dielectric constant of water, generated 437 unique
conformations. The second search of 1 yielded 394 unique
conformations. The first search of 2 yielded 499 unique
conformations, while the second search yielded 341
conformations. Each conformer was given a unique
number, starting from number 1 for the calculated global
minimum at the AMBER level and then increasing up
the energy manifold. Conformations from the second
search have been denoted by an asterisk in the subse-
quent tables.
In the case of 1, the conformations within 4 kcal/mol
of the global minimum predicted by the AMBER force
field (approximately 40 conformers) were then further
optimized using the Minnesota Gaussian Solvation Model
(MN-GSM)29 at the SM5.42/HF/6-31G* level using pa-
rameters for water as a solvent. These and subsequent
calculations were performed with Gaussian 98.30 In the
case of 2, conformations (approximately 40 conformers)
within 5 kcal/mol were carried through for optimization
with MN-GSM because there were fewer conformations
within 4 kcal/mol than there were for 1. The geometry
(26) (a) Lemieux, R. U.; Pavis, A. A.; Martin, J. C.; Watanabe, K.
H. Can. J. Chem. 1969, 47, 427. (b) Booth, H.; Khedahair, K. A.;
Readshow, S. A. Tetrahedron 1987, 43, 4699.
(27) (a) MacroModel V6.5: Mohamadi, F.; Richards, N. G. J.; Guida,
W. C.; Liskamp, R.; Lipton, M.; Caulfield, C.; Chang, G.; Hendrickson,
T.; Still, W. C. J. Comput. Chem. 1990, 11, 440. (b) Goodman, J.; Still,
W. C. J. Comput. Chem. 1991, 12, 1110.
(28) (a) Homans, S. W. Biochemistry 1990, 29, 9112. (b) Weiner, S.
J.; Kollman, P. A.; Case, D. A.; Singh, U. C.; Ghio, C.; Alagona, C.;
Profeta, S., Jr.; Weiner, P. J. Am. Chem. Soc. 1984, 106, 765. (c)
Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M., Jr.;
Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman,
P. A. J. Am. Chem. Soc. 1995, 117, 5179. (d) Senderowitz, H.; Parish,
C.; Still, W. C. J. Am. Chem. Soc. 1996, 118, 2078. (e) Senderowitz,
H.; Still, W. C. J. Org. Chem. 1997, 62, 1427. (f) For an excellent
evaluation of the AMBER force field for carbohydrates, see Table 7 in:
Barrows, S. E.; Storer, J. W.; Cramer, C. J.; French, A. D.; Truhlar, D.
G. J. Comput. Chem. 1998, 19, 1111.
(29) (a) Xidos, J. D.; Li, J.; Hawkins G. D.; Liotard, D. A.; Cramer,
C. J.; Truhlar, D. G.; Frisch, M. J. MN-GSM, version 99.2; University
of Minnesota: Minneapolis, MN 55455. (b) Li, J.; Zhu, T.; Cramer, C.
J.; Truhlar, D. G. J. Phys. Chem. A 1998, 102, 1820. (c) Li, J.; Hawkins,
G. D.; Cramer, C. J.; Truhlar, D. G. Chem. Phys. Lett. 1998, 288, 293.
(d) Zhu, T.; Li, J.; Hawkins, G. D.; Cramer, C. J.; Truhlar, D. G. J.
Chem. Phys. 1998, 109, 9117. (e) Li, J.; Zhu, T.; Hawkins, G. D.;
Winget, P.; Liotard, D. A.; Cramer, C. J.; Truhlar, D. G. Theor. Chem.
Acc. 1999, 103.
(21) Hendrickson, J. B. J. Am. Chem. Soc. 1964, 86, 4854.
(22) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94,
8205.
(23) Cremer, D.; Pople, J. A. J. Am. Chem. Soc. 1975, 97, 1354.
(24) (a) Kirby, A. J. The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer-Verlag: New York, 1983. (b) Szarek, W.
A.; Horton, D.; Eds. Anomeric Effect. Origins and Consequences; ACS
Symposia Series 87; American Chemical Society: Washington, DC,
1979. (c) Deslongchamps, P. Stereoelectronic Effects in Organic Chem-
istry; Wiley: New York, 1983. (d) Thatcher, G. R. J., Ed. The Anomeric
Effect and Associated Electronic Effects in Organic Chemistry; ACS
Symposia Series 539; American Chemical Society: Washington, DC,
1993.
(25) For discussion of the anomeric effect in seven-membered rings
see ref 16a and: (a) De´silets, S.; St-Jacques, M. J. Am. Chem. Soc.
1987, 109, 1641. (b) De´silets, S.; St-Jacques, M. Can. J. Chem. 1992,
70, 2650.
J. Org. Chem, Vol. 70, No. 1, 2005 27

DeMatteo et al.
optimizations for 2 were also performed at the SM5.42/
HF/6-31G* level of theory with water as the solvent. Gas-
phase single-point energies were obtained at the
HF/6-31G* and B3LYP/6-31+G** levels of theory.
functional with the IGLO III basis set were used in all
of these calculations.33 These calculated one-bond and
three-bond coupling constants were then compared with
the experimental NMR coupling constants to aid in the
determination of the conformation of the septanose ring.
With
these
data
on
hand,
the
EB3LYP/6-31+G**//SM5.42/HF/6-31G*(solution) (referred to as the
DFT/MN-GSM energy) was determined from the fol-
lowing equation:31
D. Computational Results. 1. Monte Carlo Results
for Septanoside 1. The Monte Carlo searches for 1
resulted in approximately 800 conformations that were
shown to cover a very wide variety of internal dihedral
angles for the septanose ring (see Supporting Information
for details); therefore, extensive and diverse coverage of
the conformational space of this structure was obtained
by the Monte Carlo approach.
∆Gsolvation ) ESM5.42/X/6-31G* - EHF/6-31G*(gas)
EB3LYP/6-31+G**//SM5.42/HF/6-31G*(solution)
)
EB3LYP/6-31+G**//SM5.42/HF/6-31G*(gas) + ∆Gsolvation
Assignment of the conformation of the ring was made
by visual inspection of the conformations within 4
kcal/mol of the global minimum, according to the AMBER
force field (see the Supporting Information), and each
conformer was given a unique number, starting from
number 1 for the calculated global minimum at the
AMBER level. The majority of the lowest energy confor-
mations, especially those within 2 kcal/mol of the global
minimum, adopted a 3,4TC5,6 conformation. The next most
abundant conformation was found to be a 1,2TC3,4 con-
formation. Although a couple of chair structures as well
as one other type of twist-chair structure were found in
these lowest energy conformations, there appears to be
much less conformational flexibility for the septanose
ring than might have been originally anticipated. In fact,
the Monte Carlo search does adequately generate con-
formational diversity for the various conformations that
the septanose ring may adopt, but the majority of these
conformations appear to comprise very high energy
structures. The global minimum, according to the
AMBER force field, for the R anomer of 1 was determined
to be a 3,4TC5,6 conformation (Chart 2). Each individual
AMBER conformational search predicted the exact same
structure as the global minimum. The major difference
among the conformations within 4 kcal/mol of this global
minimum is due to the orientation of the exocyclic OH
bonds as well as the C6-C7 dihedral angle. Based on the
O-H‚‚‚O distances and angles, these interactions appear
to be apparent hydrogen bonds between the exocyclic OH
groups.34 All of the low energy conformations appear to
adopt conformations that allow for a high degree of
hydrogen bonding among the exocyclic OH bonds, even
with the inclusion of the dielectric constant of water for
the simulation. Indeed, computational studies have
shown that intramolecular hydrogen bonding is dimin-
ished, but not eliminated, in ethylene glycol35 or glycerol36
upon inclusion of aqueous solvation.
Many of the conformations, upon optimization, gener-
ated duplicate structures. The result of these optimiza-
tions at the SM5.42/HF/6-31G* level was that 16 unique
conformations were found within 5 kcal/mol of the global
minimum for compound 1, and 19 unique conformations
were obtained for compound 2 within 5 kcal/mol of its
global minimum. The energies of these conformations
were then refined to determine their contribution to the
overall Boltzmann distribution. A vibrational frequency
analysis was performed for each of the conformations that
contributed to greater than 2% of the Boltzmann distri-
bution in order to confirm that each was, in fact, a true
minimum.
Monte Carlo searches were also performed for each of
the septanosides and optimized in the absence of a
dielectric field with the AMBER force field. All conform-
ers within 5 kcal/mol of the global minimum (according
to the AMBER force field) were then optimized at the
HF/6-31G* level of theory in the gas phase. Vibrational
frequency analysis was performed for each of the con-
formers to verify that they were true minima. The
bottom-of-the-well energies at the HF/6-31G* level of
theory were then used to calculate the contribution of
each of these conformers to the Boltzmann distribution.
NMR coupling constants were determined for each of
the various conformations from the SM5.42/HF/6-31G*
optimizations using the deMon-NMR software package.32
The Perdew and Wang exchange and Perdew correlation
(30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J. A. Gaussian 98, Revision A.7, Gaussian, Inc.; Pitts-
burgh, PA, 1998.
2. MN-GSM Results for Septanoside 1. The confor-
mations within 4 kcal/mol of the global minimum, ac-
cording to the AMBER force field, were then fully
optimized at the SM5.42/HF/6-31G* level of theory (Table
(31) Houseknecht, J. B.; McCarren, P. R.; Lowary, T. L.; Hadad, C.
M. J. Am. Chem. Soc. 2001, 123, 8811.
(32) (a) St-Amant, A.; Salahub, D. R. Chem. Phys. Lett. 1990, 169,
387. (b) Casida, M. E.; Daul, C.; Goursot, A.; Koester, A.; Pettersson,
L.; Proynov, E.; St-Amant, A.; Salahub, D. R.; Duarte, H.; Godbout,
N.; Guan, J.; Jamorski, C.; Leboeuf, M.; Malkin, V.; Malkina, O.; Sim,
F.; Vela, A. deMon-KS, version 3.4; deMon Software, 1996. (c) Malkin,
V. G.; Malkina, O. L.; Salahub, D. R. Chem. Phys. Lett. 1994, 221, 91.
(d) Malkina, O. L.; Salahub, D. R.; Malkin, V. G. J. Chem. Phys. 1996,
105, 8793. (e) Salahub, D. R.; Fournier, R.; Mlynarski, P.; Papai, I.;
St-Amant, A.; Uskio, J. In Density Functional Methods in Chemistry;
Labanowski, J. K., Anzelm, J. W., Eds.; Springer: New York, 1991; p
77. (f) Malkin, V. G.; Malkina, O. L.; Casida, M. E.; Salahub, D. R. J.
Am. Chem. Soc. 1994, 116, 5898.
(33) (a) Perdew, J. P.; Wang, Y. Phys. Rev. B 1986, 33, 8800. (b)
Perdew, J. P. Phys. Rev. B 1986, 33, 8822. (c) Perdew, J. P. Phys. Rev.
B 1986, 34, 7406.
(34) The classification of these O-H‚‚‚O interactions as hydrogen
bonds is based on the calculated bond distances and angles. Topologi-
cally, hydrogen bonds can be characterized based on minima in the
overall electron density as demonstrated by: Klein, R. A. J. Am. Chem.
Soc. 2002, 124, 13931.
(35) Cramer, C. J.; Truhlar, D. G. J. Am. Chem. Soc. 1994, 116, 3892.
(36) Callam, C. S.; Singer, S. J.; Lowary, T. L.; Hadad, C. M. J. Am.
Chem. Soc. 2001, 123, 11743.
28 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
TABLE 1. Calculated Conformational Distribution for the r Septanoside Anomer (1)a
Erel
AMBER
Erel
Erel B3LYP/6-31+G**//
SM5.42/HF/6-31G* (solution)
Boltzmann
ring
conformation
C6-C7
rotamer
conformerb
SM5.42/HF/6-31G*
contributionc (%)
36
12
1
3.67
1.75
0.00
0.54
2.38
0.91
1.06
1.77
3.97
0.79
3.39
2.58
3.50
1.85
1.98
2.20
1.80
0.00
1.25
1.97
1.84
1.91
2.49
2.72
3.20
4.49
4.49
4.30
4.80
4.72
5.47
5.68
5.70
5.88
0.00
1.43
1.47
1.61
1.79
2.04
2.18
3.20
3.67
3.68
4.07
4.25
4.59
4.68
4.95
4.97
4.99
73.78
6.60
6.11
4.86
3.59
2.36
1.83
0.33
0.15
0.15
0.08
0.06
0.03
0.03
0.02
0.02
0.02
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
1,2TC3,4
1,2TC3,4
1,2TC3,4
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
1,2TC3,4
3,4TC5,6
3,4TC5,6
tg
gt
gt
gg
gt
gt
gt
gg
gt
gt
gg
gg
gg
gg
gg
gt
gg
3
22
5
9*
13
52*
4
30
23
37*
26*
18
20
17
a Relative energies for each conformer, at each level of theory, are in kcal/mol. b Conformers from the second Monte Carlo search (as
unique structures) are denoted with an asterisk. c The Boltzmann contribution, as a percentage, is determined based on the relative
energies calculated at the B3LYP/6-31+G**//SM5.42/HF/6-31G* (solution) level of theory.
1). One result of this optimization sequence became
readily apparent. At the SM5.42/HF/6-31G* level, con-
former 36 from the original Monte Carlo search, and not
conformer 1, became the global minimum. We, therefore,
set about to determine the origin of this discrepancy
between the two theoretical methods. Comparison of the
optimized geometries at the AMBER and SM5.42/HF/
6-31G* levels demonstrated that there was not a signifi-
cant change made to the geometry upon optimization at
the higher level of theory. Therefore, we postulated that
the energetic results by each theoretical method was
more critical. To verify this hypothesis, gas-phase, single-
point energies were calculated at the HF/6-31G* and
B3LYP/6-31G* levels of theory using the AMBER-
optimized geometries. The comparison of these single-
point energies to those calculated for the SM5.42/HF/
6-31G*-optimized geometries demonstrated that the
AMBER force field was fairly poor at determining the
correct order of the lowest energy conformations (see the
Supporting Information for details).
Little difference was observed between the AMBER-
optimized geometries and the SM5.42/HF/6-31G*-opti-
mized geometries; however, enough change was observed
such that the total number of unique conformations found
at the end of the MN-GSM optimization regimen was 16
(one structure was found to be a transition state and not
a true minimum). Many of the conformations from the
two Monte Carlo searches optimized to the same struc-
tures. The global minimum at the SM5.42/HF/6-31G*
level of theory was found to be a 3,4TC5,6 conformation.
Both AMBER and SM5.42/HF/6-31G* predict a global
minimum wherein all of the OH bonds form a cyclic
hydrogen-bonding array in which each OH unit is hy-
drogen bonded to the next (Figure 1). The difference
between the global minima predicted lies in the rotamer
orientation about the C6-C7 bond. SM5.42/HF/6-31G*
predicts a global minimum in which the C6-C7 bond lies
in a tg orientation (Chart 3) so as to allow the hydroxyl
group on C7 to hydrogen bond to the hydroxyl group on
C5. The AMBER force field predicted this C6-C7 rotamer
to be gt, allowing for hydrogen bonding with the ring
oxygen rather than O5. Conformer 12, which has a
relative energy (according to SM5.42/HF/6-31G*) just
between that of 36 and 1, has a nearly identical structure
to that of conformer 1 except for the orientation about
the C6-C7 bond.
Calculation of the contributions that each conformer
makes to the Boltzmann distribution demonstrated some
interesting details. The SM5.42/HF/6-31G* calculations
predict that the global minimum (conformation 36) would
comprise approximately 73% of the Boltzmann distribu-
tion. Therefore, this conformer is very highly favored to
the exclusion of nearly all others. It would appear from
these calculations that the R anomer of 1 should exist
almost exclusively in the 3,4TC5,6 conformation in water.
The C6-C7 rotamer should also exist preferentially in the
tg conformation. All other conformations should be very
unfavorable and make much less contribution to the
overall properties of this septanose species.
3. Gas-Phase HF/6-31G* Results for Septanoside
1. The conformations from the gas-phase Monte Carlo
search within 5 kcal/mol were optimized at the HF/6-
31G* level of theory. The optimization process resulted
in 20 unique conformations, spanning an energy range
of approximately 7 kcal/mol. The results of these calcula-
tions are presented in Table 2. As in the case of the MN-
GSM-optimized geometries, the ring conformations were
found to be overwhelmingly in the 3,4TC5,6 conformation
with the 1,2TC3,4 being the next most common conforma-
tion. Conformer 17 from the original Monte Carlo search
was found to be the global minimum with a ring
conformation of 3,4TC5,6 and a tg C6-C7 rotamer. This is
the same global minimum as found at the MN-GSM level.
The Boltzmann contribution of this conformer also cor-
relates very well with that found using the solvation
model. All of the conformers within 4 kcal/mol of the
global minimum were identical to those found using the
SM5.42/HF/6-31G* level of theory. Figure 2 depicts the
structures of the lowest energy conformers. The Boltz-
mann contributions of each of these conformers was,
likewise, very comparable to those calculated in the
presence of water as the solvent.
4. Monte Carlo Results for Septanoside 2. Monte
Carlo searches for 2 resulted in approximately 850 total
J. Org. Chem, Vol. 70, No. 1, 2005 29

DeMatteo et al.
CHART 3
conformations interspersed. The global minimum for 2,
according to the AMBER force field, was found to be a
6,OTC4,5 conformer with all of the exocyclic hydroxyl
groups arranged in a similar cyclic hydrogen bonding
array as was observed for the R anomer (1). It should be
noted that, again, all of these low-lying conformations
exhibit the same hydrogen-bonding pattern. There are
some differences in the conformation of the ring and in
the C6-C7 rotamer that give rise to the number of unique
conformers. In the case of 2, as noted in 1, both AMBER
conformational searches predicted the same global mini-
mum, and many of the same conformers are within 5
kcal/mol.
It is noteworthy that the 1,2TC3,4 was represented in
the Monte Carlo simulations of both 1 and 2. This
conformation puts C6, and subsequently the exocyclic
hydroxymethyl group, into the isoclinal position. Based
on the preference for large groups to occupy the isoclinal
position, this result is not surprising. The fact that the
1,2TC3,4 conformer is energetically accessible (∼3.5 kcal/
mol higher in energy for 1 and 2) from the global
minimum suggests that the 1,2TC3,4 conformer may be
preferred if the size of the C6 substituent is increased
(by glycosylation, for example).
5. MN-GSM Results for Septanoside 2. All struc-
tures within 5 kcal/mol of the global minimum of 2 were
then optimized using the SM5.42/HF/6-31G* approach
(Table 3). Many structures were found to be identical
after this optimization and resulted in a total of 19
unique conformers. For the case of 2, the AMBER force
field performed slightly better at predicting the global
minimum than it had for 1. The global minimum was
found to be conformer 40* (obtained from the second
Monte Carlo search). For the â anomer 2, conformer 1
was found to be very close in energy to conformer 40*
(approximately 0.05 kcal/mol), considerably closer than
the difference in energy between conformers 1 and 36
for the R anomer 1.
The global minimum (conformer 40*) at the SM5.42/
HF/6-31G* level has the same ring conformation of the
septanose ring as conformer 1, the global minimum at
the AMBER level. The only difference between the two
structures is the orientation of the C6-C7 bond (Figure
3). Once again, AMBER favors the gt orientation whereas
SM5.42/HF/6-31G* favors the tg orientation (Chart 3).
As in the case of the R anomer (1), the difference is due
to the relative orientations of hydrogen bonds with the
ring oxygen and O5, respectively. Apparently, AMBER
dramatically overestimates the energetic difference be-
tween these two orientations.
Three additional conformers were found to lie within
1 kcal/mol of the global minimum (conformers 3, 9, and
26*). These three conformers were virtually identical by
visual inspection, and they differed only slightly in the
dihedral angles within the ring itself. All three are also
6,OTC4,5 conformations of the ring, much like conformers
FIGURE 1. Low energy conformers of R anomer (1). The
conformer number, ring conformation, relative energies at the
AMBER and B3LYP/6-31+G**//SM5.42/HF/6-31G* levels of
theory, as well as contribution to the Boltzmann distribution
are also shown.
conformations. All structures within 5 kcal/mol were
examined and visually assigned a configuration (see the
Supporting Information). It was observed that these low
energy structures consisted mainly of septanose rings in
the 6,OTC4,5 conformation (Chart 2). The next most
prevalent conformer in these structures is the 1,2TC3,4
conformation. As in the case of the R anomer, these
conformations make up the bulk of the low-lying confor-
mations with a couple of other twist-chair and chair
30 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
FIGURE 2. Low energy conformers of R anomer (1) in the gas phase (HF/6-31G*). The conformer number, ring conformation,
relative energies at the AMBER and HF/6-31G* levels of theory as well as the contribution to the Boltzmann distribution are
shown.
40* and 1. These conformers lie approximately 0.5 kcal/
mol above the global minimum. They differ from con-
formers 40* and 1 mainly in the C6-C7 orientation
(Figure 2). Conformers 3, 9, and 26* are gg rotamers with
the hydroxyl group pointed in such a way as to allow for
hydrogen-bonding interactions with the ring oxygen.
Calculations were made to determine the contributions
that each of these 19 conformers would make to the
Boltzmann distribution. It was found that conformers
40*, 1, 3, 9, and 26* made up the bulk (94%) of the
Boltzmann distribution. Although there is expected to be
very little conformational flexibility of the ring itself (it
should be predominantly fixed in the 6,OTC4,5 conforma-
tion), there is a significant amount of flexibility to be
expected for the orientation about the C6-C7 bond. This
is in stark contrast to the calculated results for the R
anomer 1 in which the tg rotamer was predicted to be
favored by at least 1.25 kcal/mol over the other rotamers.
For the â anomer 2, there is only a maximum of 0.5 kcal/
mol separating the rotamers, as predicted at the SM5.42/
HF/6-31G* level.
6-31G* level of theory. The optimization process resulted
in 22 unique conformations, spanning an energy range
of approximately 7 kcal/mol. The results of these calcula-
tions are given in Table 4. The results of the calculations
are, again, extremely similar to those of the MN-GSM
optimization sequence. The two lowest energy conformers
have a ring conformation of 6,OTC4,5 and correspond to
the gt and tg C6-C7 rotamers. The lowest energy con-
former is the gt rotamer with a Boltzmann contribution
of 33%. The tg rotamer is only 0.07 kcal/mol higher in
energy than the global minimum and contributes 29%
to the overall Boltzmann distribution. Both of the con-
formers are nearly identical to those found at the MN-
GSM level. Figure 4 depicts the structures of the lowest
energy conformers that contribute most significantly to
the Boltzmann distribution. Most of the conformers are
identical to those found under MN-GSM with the excep-
tion of conformers 49 and 42. Conformer 49 has a ring
conformation of 4,5TC2,3 and conformer 42 has a ring
conformation of 5,6TCO,1. Both of these ring conformations
were found not to exist in the lowest energy conformers
found at the MN-GSM level and represent about 13% of
the Boltzmann distribution. With the exception of these
6. Gas-Phase HF/6-31G* Results for Septanoside
2. The conformations from the gas-phase Monte Carlo
search for 2 within 5 kcal/mol were optimized at the HF/
J. Org. Chem, Vol. 70, No. 1, 2005 31

DeMatteo et al.
TABLE 2. Calculated Conformational Distribution for the r Septanoside Anomer (1) in the Gas Phasea
ring
conformation
C6-C7
rotamer
Boltzmann
conformer
Erel AMBER
Erel HF/6-31G*
contributionb (%)
17
24
1
12
7
6
36
3
22
28
37
8
44
35
40
39
23
15
31
41
2.79
3.41
0.00
2.57
1.36
1.34
4.02
0.90
3.14
3.55
4.09
2.27
5.03
3.94
4.15
4.15
3.30
2.64
3.72
4.28
0.00
1.51
1.59
1.71
2.55
3.31
3.38
4.35
5.30
5.44
5.53
5.60
5.67
5.99
6.22
6.56
6.84
6.93
7.10
7.24
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
1,2TC3,4
3,4TC5,6
1,2TC3,4
1,2TC3,4
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
3,4TC5,6
1,2TC3,4
1,2TC3,4
3,4TC5,6
1,2TC3,4
tg
gg
gt
gg
gt
gg
gg
gt
gt
gg
gt
gt
gg
gt
gt
gt
gg
gt
gt
gg
81.97
6.40
5.63
4.58
1.11
0.31
0.27
0.05
0.01
0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
a Relative energies for each conformer, at each level of theory, are in kcal/mol. b The Boltzmann contribution, as a percentage, is
determined based on the relative energies calculated at the HF/6-31G* level of theory.
TABLE 3. Calculated Conformational Distribution for the â Septanoside Anomer (2)a
Erel
AMBER
Erel
Erel B3LYP/6-31+G** //
SM5.42/HF/6-31G* (solution)
Boltzmann
ring
conformation
C6-C7
rotamer
conformerb
SM5.42/HF/6-31G*
contributionc (%)
40*
1
4.77
0.00
1.02
2.11
3.66
1.66
1.78
4.88
0.64
3.66
2.41
1.78
1.99
2.88
1.98
2.01
3.02
4.56
2.59
0.00
0.12
0.50
0.58
0.55
1.82
2.01
1.55
2.37
3.59
3.58
2.70
3.32
3.46
3.76
4.01
4.15
3.82
5.15
0.00
0.05
0.50
0.52
0.54
1.60
1.77
2.02
2.31
3.14
3.32
3.46
3.47
3.60
3.73
3.80
4.33
4.47
4.74
29.83
27.21
12.84
12.45
11.94
1.99
1.51
0.98
0.60
0.15
0.11
0.09
0.09
0.07
0.05
0.05
0.02
0.02
0.01
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
5C1,2
tg
gt
gg
gg
gg
gt
gt
gg
gt
gg
gt
gt
gt
gg
gt
gt
gg
gt
gt
3
9
26*
4
8*
49*
2
27*
11
6
11*
14
10*
7
17
37*
13
5,6TCO,1
6,OTC4,5
6,OTC4,5
1,2TC3,4
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
6,OTC4,5
3,4TC5,6
5,6TCO,1
a Relative energies for each conformer, at each level of theory, are in kcal/mol. b Conformers from the second Monte Carlo search (as
unique structures) are denoted with an asterisk. c The Boltzmann contribution, as a percentage, is determined based on the relative
energies calculated at the B3LYP/6-31+G**//SM5.42/HF/6-31G* level of theory.
3
3
two conformers, the gas-phase results are nearly identical
to those under the solvation model.
between 1 and 2 lies in the J2,3 and J3,4 coupling
constants. The R anomer (1) should have a 3J2,3 value of
approximately 8.9 Hz and a 3J3,4 value of approximately
7.5 Hz. The â anomer, 2, in contrast, should have a 3J2,3
7. Calculation of Theoretical NMR Coupling Con-
stants. NMR coupling constants were calculated for each
of the protons on the septanose ring. The coupling
constants were determined for each of the 17 unique
conformers determined for 1 and for each of the 19 unique
conformers of 2. These values were then weighted ac-
cording to the Boltzmann distribution of these conformers
at the B3LYP/6-31+G**//SM5.42/HF/6-31G* level and
summed to yield overall coupling constants for the R and
â anomers (1 and 2, respectively). The results of the
calculated coupling constants of the ring hydrogens are
summarized in Table 5.
3
value of approximately 5.4 Hz and a J3,4 value of
approximately 9.1 Hz. It appears that these coupling
constants are particularly indicative of the differences
between the 3,4TC5,6 conformation of the R anomer and
the 6,OTC4,5 conformation of the â anomer. Thus, it should
be very straightforward to assign the structures by NMR.
In fact, the values of these coupling constants are in good
agreement with the experimental observations, and the
3
largest deviations are observed for the J2,3 coupling in
2 and the J5,6 coupling in 1. In general, the coupling
3
Of particular note are the differences between the R
constants are within approximately 1 Hz of the experi-
mental values, and there does not appear to be a
consistent trend of the deMon-NMR calculated coupling
and â anomers. Both have very similar values for the
3
3
3J1,2, J4,5, and J5,6 coupling constants. The difference
32 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
FIGURE 3. Low energy conformers for â anomer 2. The conformer number, ring conformation, relative energies at the AMBER
and B3LYP/6-31+G**//SM5.42/HF/6-31G* levels of theory as well as contribution to the Boltzmann distribution are also shown.
constants as the calculated values are above, below and
close to the experimental values in the different cases.
2, the program NMRSim37 was used to simulate the one-
dimensional 1H NMR spectra. The experimental and
simulated spectra were then compared by visual inspec-
tion.
III. Discussion
The relative agreement between the experimental and
A. Ring Conformation. Table 5 collects the calcu-
lated, experimental, and simulated 3JH,H values for 1 and
3
simulated JH,H values provided an additional check on
the accuracy of the observed coupling constants. As noted
in Table 5, the errors in the measurement of the coupling
constants were (0.19 Hz for 1 and (0.38 Hz for 2. These
values were the result of the digital resolution in the
acquisition of the free induction decays for 1 and 2. The
3
1
2. Observed JH,H values were measured from H NMR
spectra of 1 and 2 dissolved in CD3OD. At 600 MHz, there
was sufficient spectral resolution of nearly all of the
resonances so that coupling constants could be conve-
niently extracted from the one-dimensional 1H NMR
spectra. The only exception was the H2, H3, and H5
resonances of 1 that showed significant chemical shift
overlap. To confirm the assigned coupling constants
associated with the H2, H3, and H5 nuclei in 1 and
generally to support the assigned 3JH,H couplings in 1 and
3
average deviation of the simulated JH,H values ((0.11
Hz for 1 and (0.21 Hz for 2) is within the coupling
constant error in each case. Despite this fact, for both 1
and 2, there are examples of significant discrepancies
(37) NMRSim v4.1 as part of TOPSPIN v1.1, Bru¨ker Biospin GmbH.
J. Org. Chem, Vol. 70, No. 1, 2005 33

DeMatteo et al.
FIGURE 4. Low energy conformers of â anomer 2 in the gas phase (HF/6-31G*). The conformer number, ring conformation,
relative energies at the AMBER and HF/6-31G* levels of theory as well as the contribution to the Boltzmann distribution are
shown.
between the experimental and simulated values; specif-
mentioned, bulky substituents will prefer the isoclinal
position due to the diminished steric interactions at that
position. Additionally, the axial OCH3 group benefits
from stabilization via the anomeric affect in 1. In the
3,4TC5,6 conformer, the other ring substituents (C2-C6)
maintain an equatorial disposition, which alleviates
steric interactions and allows the possibility of intramo-
lecular hydrogen bonding. For â-septanoside 2, all of the
substituents are able to assume a similar equatorial
disposition when the ring adopts the 6,OTC4,5 conforma-
tion.
3
3
ically, the simulated J1,2 was -0.25 Hz and J2,3 was
-0.22 Hz for 1 and 3J2,3 was +0.46 Hz and 3J5,6 was -0.92
Hz for 2. Interestingly, in each of these cases, the
simulated values deviated in the direction of the deMon
calculated values. Overall, the consistency between the
calculated deMon values and the experimental or simu-
lated 3JH,H values supported both the stereochemical and
conformational assignment of 1 and 2.
Both 1 and 2 showed characteristic features that
governed their ring conformation. For example, the
3,4TC5,6 conformation adopted by 1 placed the aglycon
OCH3 group in the isoclinal position. Compound 1 is an
R-septanoside with an axial aglycon group. As previously
B. C6-C7 Rotamer Populations. Rotation about the
exocyclic hydroxymethyl group has been correlated with
ring conformations in several furanose systems.38 Our
34 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
TABLE 4. Calculated Conformational Distribution for the â Septanoside Anomer (2) in the Gas Phasea
ring
conformation
C6-C7
rotamer
Boltzmann
conformer
Erel AMBER
Erel HF/6-31G*
contributionb (%)
1
39
5
49
6
42
51
2
54
47
3
40
37
41
9
13
38
26
52
14
46
12
0.00
4.03
2.00
4.79
2.06
4.62
4.97
0.60
5.02
4.76
1.00
4.14
3.73
4.57
2.29
3.11
4.00
3.66
4.98
3.59
4.73
3.09
0.00
0.07
0.28
0.61
1.75
2.09
2.43
2.52
2.62
2.67
2.92
3.20
3.29
3.83
3.86
3.92
4.42
5.16
5.29
5.38
5.96
6.68
6,OTC4,5
6,OTC4,5
6,OTC4,5
4,5TC2,3
6,OTC4,5
5,6TCO,1
4,5TC2,3
5,6TCO,1
3,4TC5,6
6,OTC4,5
4,5TC2,3
6,OTC4,5
6,OTC4,5
5,6TCO,1
6,OTC4,5
6,OTC4,5
4,5TC2,3
4,5TC2,3
4,5TC2,3
4,5TC2,3
6,OTC4,5
5,6TCO,1
gt
tg
gg
gg
gt
tg
tg
gt
gt
tg
gt
gt
gg
gg
gt
gg
gt
gt
gg
gt
gt
gt
33.11
29.22
20.77
11.82
1.74
0.97
0.54
0.47
0.39
0.37
0.24
0.15
0.13
0.05
0.05
0.04
0.02
0.01
0.00
0.00
0.00
0.00
a Relative energies for each conformer, at each level of theory, are in kcal/mol. b The Boltzmann contribution, as a percentage, is
determined based on the relative energies calculated at the HF/6-31G* level of theory.
3
1
TABLE 5. Calculated and Experimental JH,H and JC,H Coupling Constants (Hz) for Septanosides 1 and 2
coupling
constant
calc 1a
expt 1b,d
NMRSim 1c
calc 2a
expt 2b,d
NMRSim 2c
3J1,2
5.81
8.86
7.48
7.17
8.5
3.08
6.41
6.2
9.8
5.95
9.58
5.92
5.42
9.11
7.38
8.32
3.68
3.85
5.4
3.3
5.44
3.76
9.11
7.32
8.68
6.7
3J2,3
3J3,4
8.3
8.41
9.2
3J4,5
8.3
10.2
5.2
2.8
11.7
167.9
8.36
10.14
5.26
2.77
11.76
7.2
3J5,6
9.6
3J6,7R
3J6,7S
2J7R,7S
1JC,H
6.7
2.5
2.55
11.76
11.8
160.1
154.40
143.39
a Values calculated using deMon and using the B3LYP/6-31+G**//SM5.42/HF/6-31G* Boltzmann weightings from that conformational
distribution. b Observed coupling constants from 1H NMR experiments. c Values used for simulation using NMRSim. d Coupling constant
error: (0.19 Hz for 1 and (0.38 Hz for 2.
TABLE 6. Comparison of C6-C7 Rotamer Populations
investigation of 1 and 2 was motivated in part by the
in 1 and 2
presence of the exocyclic hydroxymethyl groups in their
1
2
structures. Chart 3 shows the H-C-C-H dihedral
angles for the tg, gt, and gg conformers. The coupling
calcd
obsd
calcd
obsd
3
constants (3JH6,H7R and JH6,H7S) are diagnostic for the
Xtg
Xgt
Xgg
74
11
16
13
28
59
30
32
38
10
60
30
dihedral conformation and are described by eqs 1-3.39
0.57Xtg + 9.70Xgt + 4.11Xgg) 3JH6-H7R
(1)
eqs 1 and 2 is described in the Experimental Section.
Table 6 collects the rotamer populations of compounds 1
and 2 based on the conformer populations generated by
computation (calcd) and by solving eqs 1-3 using the
9.79Xtg + 3.74Xgt + 0.71Xgg) 3JH6-H7S
Xtg + Xgt + Xgg ) 1
(2)
(3)
3
observed J values (obsd).
Coupling constants were also determined theoretically
in order to identify the C6-C7 rotamer present in the
compounds. The coupling between H6 and H7R,S is indica-
tive of the C6-C7 rotamer. The coupling constants for the
individual conformers, as well as the Boltzmann-weighted
coupling constants, are summarized in Tables 7 and 8.
Comparison with experiment reveals that the SM5.42/
HF/6-31G* level of theory is rather poor at predicting the
C6-C7 rotameric distribution. As can be seen, the C6-
C7 rotamer population of the R anomer is particularly
poorly predicted.
The H7S chemical shift was assumed to be downfield
relative to H7R based on previous pyranose and furanose
examples.40,41 The determination of the coefficients for
(38) (a) Gordon, M. T.; Lowary, T. L.; Hadad, C. M. J. Am. Chem.
Soc., 1999, 121, 9682. (b) Gordon, M. T.; Lowary, T. L.; Hadad, C. M.
J. Org. Chem. 2000, 65, 4954. (c) McCarren, P. R.; Gordon, M. T.;
Lowary, T. L.; Hadad, C. M. J. Phys. Chem. A 2001, 105, 5911.
(39) The method for determining the coefficients in eqs 1-3 is
described in the Experimental Section.
(40) Bock, K.; Duus, J. Ø. J. Carbohydr. Chem. 1994, 13, 513.
(41) Serianni, A. S.; Barker, R. Can. J. Chem. 1979, 57, 3160.
J. Org. Chem, Vol. 70, No. 1, 2005 35

DeMatteo et al.
3
1
TABLE 7. JH,H Coupling Constants (Hz) for Individual
TABLE 9. JC,H at the C1-H1 Anomeric Position (Hz) in
r Septanoside 1
C6-C7 Rotamers for Septanoside 1
R conformera
3JH6,H7R
3JH6,H7S
rotamer
calcd coupling
constant (Hz)
C1-H1
R conformera
bond length (Å)
36
1
2.56
8.61
8.61
8.33
2.01
2.43
1.16
8.59
8.35
1.41
2.7
8.93
8.77
1.97
1.98
1.15
1.14
3.08
5.2
8.05
2.73
1.72
1.64
1.25
1.00
2.16
1.52
1.46
1.90
1.04
1.79
2.52
1.66
1.32
2.17
2.15
6.41
2.8
tg
gt
gt
gt
gg
gg
gg
gt
gt
gg
gg
gt
gt
gg
gg
gg
gg
36
12
1
154.53
154.3
153.9
1.082
1.082
1.082
1.082
1.083
1.083
1.083
1.082
1.087
1.087
1.082
1.083
1.082
1.082
1.082
1.087
1.085
5
9*
12
22
3
3
154.42
154.77
154.51
154.16
146.77
138.85
138.74
154.37
155.01
154.48
153.73
153.13
139.05
147.27
154.40
167.9b
22
5
9*
13
52*
4
13
52*
30
4
18
30
20
23
23
37*
26*
37*
26*
20
17
18
Overall Calculated
Experiment
17
Overall value (Hz)
Experiment
a Conformers from the second Monte Carlo search (as unique
structures) are denoted with an asterisk.
a Conformers from the second Monte Carlo search (as unique
structures) are denoted with an asterisk. b Coupling constant error
of (1.5 Hz.
3
TABLE 8. JH,H Coupling Constants (Hz) for Individual
C6-C7 Rotamers for Septanoside 2
â conformera
3JH6,H7R
3JH6,H7S
rotamer
between calculated and experimental C6-C7 rotamer
populations could be attained if the calculations included
at least one solvent molecule to provide some representa-
tion of such intrinsic interactions between the solute and
solvent.
There is closer agreement between the calculated and
observed populations for 2 (Table 8). The preference for
the gt rotamer here is primarily based on a gauche effect
between O6 and O7, and the gg conformer by hypercon-
jugation as was described for 1. The preference for the
gt conformer in 2 may also represent the formation of
an H-bond between the C7 hydroxyl group and the ring
(O6) oxygen.
C. Anomeric Orientation. Calculations were also
used to determine the one-bond C-H coupling constants
(1JC,H) at the anomeric (C1) position. This has been a very
useful tool for the assignment of R and â anomers in
pyranoses44 and some furanoses.45 We have postulated
that it may be a useful diagnostic for the identification
of the anomeric orientation in septanose rings. In the
larger ring, the value of this coupling constant is reason-
ably correlated to the C-H bond length.46 The results of
the calculations, as well as the C-H bond lengths, are
given in Tables 9 and 10. From these data, there is a
distinct difference in the value of these coupling constants
for the R anomer (1) as opposed to the â anomer (2). There
are, however, some discrepancies in the data; the devia-
tions appear to be in conformers which will not be
populated (due to their high energy) and which will
contribute very little to the actual Boltzmann distribu-
tion. It should be noted that these less stable species tend
to have noticeably different bond lengths from the rest
of the set, and this structural aspect may be the origin
of this discrepancy.
1
8.12
2.27
7.63
1.11
1.13
1.11
7.83
8.53
1.48
8.15
9.00
8.17
8.18
7.87
1.99
8.50
1.31
8.23
1.17
3.68
6.7
2.29
7.83
1.29
2.13
2.12
2.20
1.36
2.74
1.70
2.33
2.44
2.22
2.32
1.90
1.71
2.82
2.21
1.54
2.09
3.85
2.5
gt
tg
gt
gg
gg
gg
gt
gt
gg
gt
gt
gt
gt
gt
gg
gt
gg
gt
gg
40*
4
3
9
26*
8*
2
49*
11
6
11*
10*
7
27*
37*
14
13
17
Overall Calculated
Experiment
a Conformers from the second Monte Carlo search (as unique
structures) are denoted with an asterisk.
The discrepancies between the calculated and observed
C6-C7 rotamer populations in Tables 7 and 8 may be
explained by considering the effects that govern them.
Looking first at 1, the difference in populations is most
pronounced in changing from the tg rotamer (Chart 3)
in the calculated structures to gg in solution. The high
tg population may be attributed to an intramolecular
H-bond in the calculated structures that may not persist
in solution due to strong interactions with the solvent.
Dynamics calculations have shown that solvation can
play a significant role in governing rotamer populations.42
The preponderance of the gg rotamer observed for 1 may
be explained by invoking hyperconjugation between the
(44) (a) Perlin, A. S.; Casu, B. Tetrahedron Lett. 1969, 2921. (b) Bock,
K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2 1974, 293.
(45) Callam, C. S.; Gadikota, R. R.; Lowary, T. L. J. Org. Chem.
2001, 66, 4549.
(46) Wolfe, S.; Pinto, B. M.; Varma, V.; Leung, R. Y. N. Can. J.
Chem. 1990, 68, 1051
σ
(C6-H6) f σ*(C7-O7).
43 We anticipate that better agreement
(42) Kirschner, K., N.; Woods, R. J. Proc. Natl. Acad. Sci. U.S.A.
2001, 98, 10541.
(43) Juaristi, E.; Atu´nez, S. Tetrahedron 1992, 48, 5941.
36 J. Org. Chem., Vol. 70, No. 1, 2005

Septanose Carbohydrates
1
TABLE 10. JC,H at the C1-H1 Anomeric Position (Hz)
lack this functionality. We consider the exocyclic hy-
droxymethyl group to be an important factor for deter-
mining new low energy conformations, and for the
potential biological activity of septanosides in general.
The conformational analysis of 1 and 2 has allowed for
an assignment of their anomeric configuration and the
determination of their respective low energy conforma-
tions. This approach for determination of the low-energy
conformations involved a Monte Carlo search followed
by minimization with the AMBER force field and sub-
sequent geometric refinement at the SM5.42/HF/6-31G*
in â Septanoside 2
calcd coupling
constant (Hz)
C1-H1
bond length (Å)
â conformera
40*
143.8
1.086
1.086
1.086
1.086
1.086
1.086
1.087
1.086
1.088
1.086
1.086
1.082
1.086
1.086
1.086
1.081
1.081
1.087
1.088
1
143.1
3
9
143.87
143.93
143.97
142.74
142.43
144.04
144.18
142.67
141.81
151.7
142.25
143.83
143
149.64
149.36
151.01
143.43
143.39
160.1b
26*
4
8*
49*
2
3
and B3LYP/6-31+G** levels. Calculated JH,H coupling
27*
constants derived from a weighted Boltzmann population
of the low energy conformations using the deMon soft-
ware package provided a theoretical basis for discrimi-
nating the R and â anomers as 1 and 2, respectively. The
computational analysis indicated that only one ring
conformation was accessible by 1 (3,4TC5,6) and 2 (6,OTC4,5);
the main variation in the low-energy conformers was
through rotation about the C6-C7 exocyclic bond. This
finding supports previous reports that monoseptanosides
may reside in one dominant, low-energy conformation.
However, at this time, we are unable to provide informa-
tion about the energy barrier for interconversion between
the assigned low-energy conformers and their conforma-
tional variants. Overall, the results indicate that sep-
tanose carbohydrates may be less conformationally flex-
ible than other seven-membered ring structures, and that
electronic effects (anomeric, exo-anomeric, H-bonding)
common to carbohydrate chemistry contribute to deter-
mining the dominant energy minima.
11
6
11*
14
10*
7
17
37*
13
Overall value (Hz)
Experiment
a Conformers from the second Monte Carlo search (as unique
structures) are denoted with an asterisk. b Coupling constant error
of (1.5 Hz.
The experimental 1JC,H values for 1 and 2 (Tables 5, 9,
and 10) proved to be sufficiently distinct so that they were
diagnostic of the anomeric stereochemistry. The experi-
mental data were collected by 13C-coupled HMQC experi-
ments where the peak separation for C1 was measured
1
Gas-phase HF/6-31G* calculations predicted very little
difference in the final geometries as compared to opti-
mizations performed at the SM5.42/HF/6-31G* level of
theory and provided very similar contribution to the final
Boltzmann distribution. This indicates that the solvation
model does not significantly affect the geometries and
relative energetics of the conformers contributing to the
Boltzmann distribution. In both the gas-phase and aque-
ous-phase calculations, intramolecular hydrogen bonding
appears to be very significant in the most preferred
conformations of 1 and 2.
to yield the JC,H coupling constant. The magnitudes of
1JC,H of 1 (1JC,H ) 167.9 Hz) and 2 (1JC,H ) 160.1 Hz) were
comparable to reported pyranose and furanose values.44,45
1
The apparent sensitivity of the JC,H to the anomeric
stereochemistry further suggested that one ring con-
former was dominant for 1 and 2, and did not suggest
averaging due to conformational interconversion. Com-
parison of the measured with the calculated 1JC,H values
in Tables 9 and 10 show a systematic underestimation
in the magnitude of the one-bond coupling constants.
1
However, both the measured and calculated JC,H value
The computational results are supported by experi-
for 1 are greater than for 2. This consistency between
the relative differences in the measured and calculated
values indicated that there is some general agreement
between the two methods. Although preliminary, the
results show that the magnitude of 1JC,H values could be
diagnostic for the determination of septanoside anomeric
stereochemistry.
mental 1H NMR spectroscopic data. The correspondence
3
between the calculated, observed, and simulated JH,H
values allowed for the assignment of the R and â anomers
as 1 and 2, respectively, and confirm the accuracy of the
conformations provided by computation. Analysis of the
C6-C7R,S coupling constants allowed for the reassignment
of the rotamer populations. The discrepancy between the
calculated and experimental results for the C6-C7R,S
coupling constants was based on the fact that the
computational treatment over-predicted the tg rotamer
in both 1 and 2; a situation in which the tg rotamer was
involved in intramolecular hydrogen bonding with the
C5 hydroxyl group. On the other hand, in our NMR
experiments, the methanol solvent could effectively com-
pete as an H-bond acceptor and therefore favor alterna-
tive rotameric conformations around the C6-C7 bond.
We are currently exploring the synthesis of di- and
oligoseptanosides using the strategies we have described
here and the conformational analysis of these structures.
The goal of these investigations is to utilize septanose
carbohydrates as inhibitors of biologically important
IV. Conclusions
We have demonstrated the synthesis of methyl R-D-
glycero-D-idoseptanoside (1) and methyl â-D-glycero-D-
guloseptanoside (2). The route is highlighted by the use
of carbohydrate-based oxepine 4 as a glycosyl donor. In
contrast to our earlier report of the epoxidation of
carbohydrate-based oxepines using DMDO,10 the epoxi-
dation in the case of 4 was not stereoselective and gave
a mixture of 1,2-anhydroseptanoses. The product methyl
septanosides (1 and 2) are especially noteworthy due to
the hydroxymethyl groups at the C6 position of the ring.
Because most previous synthetic preparations cyclized
through the C6 hydroxy group, the resulting septanosides
J. Org. Chem, Vol. 70, No. 1, 2005 37

DeMatteo et al.
73.8, 72.4, 71.6, 64.1, 56.4; FAB-MS m/z (M + H)+ calcd
225.0974, found 225.0988.
carbohydrate processing enzymes and as the basis for
developing new protein-carbohydrate interactions that
are orthogonal to natural systems.
Methyl â-D-glycero-D-Guloseptanoside (2). Pd/C (10%,
0.009 g) was added to a solution of 7 (0.028 g, 0.047 mmol) in
CH3OH (10 mL). The reaction was placed under an H2
atmosphere via a balloon, and the mixture was stirred for 4 h
at rt. The balloon was removed from the flask, and the mixture
was filtered through a short pad of Celite. The Celite was
washed with additional CH3OH (4 × 5 mL). The solvent was
removed from the combined filtrates by rotary evaporation
under reduced pressure to give a clear, colorless oil (0.010 g,
98%): [R]D +7.8 (c 1.01, CH3OH); 1H NMR 600 MHz (CD3OD)
δ 4.33 (d, J ) 5.4 Hz, 1H), 3.83 (dd, J ) 11.8, 2.6 Hz, 1H),
3.80 (dd, J ) 5.4, 4.0 Hz, 1H) 3.70 (dd, J ) 9.1, 7.3 Hz, 1H),
3.67 (dd, J ) 9.1, 4.0 Hz, 1H), 3.57 (dd, J ) 11.6, 6.7 Hz, 1H),
3.48 (ddd, J ) 8.3, 6.7, 2.6 Hz, 1H), 3.43 (s, 3H), 3.29 (dd, J )
8.3, 7.3 Hz, 1H); 13C NMR 150 MHz (CD3OD) δ 110.0 82.0,
75.7, 75.6, 73.9, 72.0, 64.5, 56.7; FAB-MS m/z (M + H)+ calcd
225.0974, found 225.0992.
V. Experimental Section
1,2-Anhydro-3,4,5-tetra-O-benzyl-r/â-D-glycero-D-ido/
guloseptanose (5). Oxepine 3 (0.029 g, 0.054 mmol) was dried
via azeotropic distillation from toluene (3 × 5 mL) under
reduced pressure and dissolved in dry DCM (2 mL). The
solution was cooled in an ice bath to 0 °C and a DMDO (0.310
mL, 0.35 M) solution was added dropwise. The mixture was
stirred at 0 °C for 30 min and the solvent was removed under
reduced pressure. NMR showed quantitative conversion: 1H
NMR (CDCl3) δ 7.44-7.32 (m, 15H), 4.88 (d, J ) 11.2 Hz, 1H),
4.85 (d, J ) 13.2, 1H), 4.80 (d, J ) 2.2 Hz, 1H), 4.75 (d, J )
11.5 Hz, 1H), 4.69 (d, J ) 9.8 Hz, 1H), 4.60 (d, J ) 11.5 Hz),
3.85 (dd, J ) 13.1, 3.5 Hz, 1H), 3.77 (m, 2H), 3.67 (dd, J )
13.1, 6.6 Hz, 1H), 3.56 (m, 1H), 2.99 (s, 1H); 13C NMR (CDCl3)
δ 138.6, 138.1, 128.7, 128.6, 128.3, 128.1, 127.9 (2), 82.3, 80.6,
79.8, 78.3, 75.4, 73.6, 73.0, 64.1, 58.4.
Methyl 3,4,5,7-Tetra-O-benzyl-r-D-glycero-D-idosep-
tanoside (6) and Methyl 3,4,5,7-Tetra-O-benzyl-â-D-glyc-
ero-D-guloseptanoside (7). DMDO epoxidation of 3 (0.039
g, 0.09 mmol) in CH2Cl2 (2 mL) at 0 °C over 30 min was
followed by solvent removal under reduced pressure. To the
residue was added NaOCH3 (0.006 g) in CH3OH (3 mL), and
the mixture was stirred overnight (12 h). The reaction was
quenched with water (2 mL), and the solvent was removed
under reduced pressure. The residue was dissolved in CH2Cl2
(15 mL), washed with water (2 × 15 mL), and dried (Na2SO4),
and the solvent was removed under reduced pressure. The
residue was purified by column chromatography (3:1 hexanes/
EtOAc) to give two products.
Methyl 3,4,5,7-Tetra-O-benzyl-r-D-glycero-D-idosep-
tanoside (6). The first fraction gave 6 (0.0084 g, 16%) as a
white solid: Rf 0.32 (3:1 hexanes/EtOAc); mp 92-94 °C; [R]D
+23.6 (c 1.43, CHCl3); 1H NMR 400 MHz (CDCl3) δ 7.34-7.26
(m, 18H), 7.14-7.12 (m, 2H), 4.99 (d, J ) 11.1 Hz, 2H), 4.81
(d, J ) 11.1 Hz, 1H), 4.80 (d, J ) 10.8 Hz, 1H), 4.67-4.55 (m,
4H), 4.51 (d, J ) 6.0 Hz, 1H), 3.86-3.78 (m, 2H), 3.77 (dd,
J ) 10.2, 3.3 Hz, 1H) 3.66-3.60 (m, 4H), 3.46 (s, 1H), 2.99 (s,
1H); 13C NMR 100 MHz (CDCl3) δ 138.8, 138.5, 138.1, 137.8,
128.9, 128.6(3), 128.3(2), 128.0, 127.9(3), 127.7, 127.6, 104.3,
88.8, 79.8(2), 76.6, 76.2, 75.3, 73.7, 72.9, 70.5, 69.7, 56.1; FAB-
MS m/z (M - H)+ calcd 583.2696, found 583.2674.
Methyl 3,4,5,7-Tetra-O-benzyl-â-D-glycero-D-gulosep-
tanoside (7). The second fraction gave 7 (0.025 g, 47%) as a
clear colorless oil: Rf 0.24 (3:1 hexanes/EtOAc); [R]D +23.9 (c
0.45, CHCl3); 1H NMR 400 MHz (CDCl3) δ 7.33-7.27 (m, 18H),
7.18-7.17 (m, 2H), 4.72 (d, J ) 11.5 Hz, 1H), 4.60-4.54 (m,
4H), 4.51-4.46 (m, 3H), 4.31 (d, J ) 11.3 Hz, 1H) 4.04-4.03
(m, 2H), 3.96-3.95 (m, 2H), 3.69 (d, J ) 9.1 Hz, 1H) 3.61-
3.57 (m, 2H), 3.54 (s, 3H); 13C NMR 100 MHz (CDCl3) δ 138.6,
138.3, 138.1, 128.6(2), 128.5, 128.2(2), 128.1, 128.0, 127.9(2),
127.7, 106.9, 81.3, 79.2, 77.8, 77.6, 73.9, 73.5, 72.9, 72.8(2),
71.7, 56.5; FAB-MS m/z (M + H)+ calcd 585.2852, found
585.2883.
NMR Spectroscopy for Conformational Analysis. NMR
spectra used in the conformational analysis of 1 and 2 were
recorded on samples at 5-10 mM concentration in 0.75 mL
3
CD3OD. The JH,H values were measured from 600 MHz 1H
1
NMR spectra. Simulation of the H NMR spectra for 1 and 2
was done using NMRSim from Bru¨ker.37 Overlays comparing
the simulated to the observed spectra are provided in the
1
Supporting Information. The JC,H were measured from 13C-
coupled HMQC spectra. These HMQC spectra for 1 and 2 are
also provided in the Supporting Information.
Determination of the C6-C7 Rotamer Populations.
Equations 1-3 were used to determine the rotamer popula-
3
tions about the C6-C7 bond by analysis of the JH,H coupling
constants between H6 and H7R (3JH6,H7R) and H6 and H7S
(3JH6,H7S). The coefficients for these equations were determined
using eq 4.47 Coefficients derived using other methods48 were
too large to fit the observed coupling constants and gave
negative rotamer populations. Measured values of the H-C-
C-H dihedral angles (φ) from the calculated low energy
conformers of 1 and 2 were used to define the calculated
3
3JH6,H7R and JH6,H7S for the tg, gt, and gg rotamers.
Jcalcd ) J0 cos2φ - 0.28 Hz; J0 ) 9.27 for 0° ) φ ) 90°,
J0 ) 10.36 for 90° ) φ ) 180° (4)
Acknowledgment. M.P.D. acknowledges partial
support by a GAANN fellowship and a Procter and
Gamble fellowship. C.M.H. gratefully acknowledges
generous computational resources at the Ohio Super-
computer Center where all of these calculations were
performed and thanks the National Science Foundation
and the OSU Environmental Molecular Science Insti-
tute (CHE-0089147) for general support. M.W.P. thanks
the University of Connecticut for financial support. We
thank Bru¨kerBioSpin Corp., Billerica, MA, for gener-
ously supplying NMR time and technical assistance in
the collection of data on compounds 1 and 2.
Methyl r-D-glycero-D-Idoseptanoside (1). Pd/C (10%,
0.006 g) was added to a solution of 6 (0.021 g, 0.036 mmol) in
CH3OH (10 mL). The reaction was placed under an H2
atmosphere via a balloon, and the mixture was stirred for 4 h
at rt. The balloon was removed from the flask, and the mixture
was filtered through a short pad of Celite. The Celite was
washed with additional CH3OH (4 × 5 mL). The solvent was
removed from the combined filtrates by rotary evaporation
under reduced pressure to give a clear, colorless glass (0.0078
g, 95%): [R]D +106.8 (c 0.24, CH3OH); 1H NMR 600 MHz (CD3-
OD) δ 4.35 (d, J ) 6.2 Hz, 1H), 3.79 (dd, J ) 11.7, 2.8 Hz,
1H), 3.64 (dd, J ) 11.7, 5.2 Hz, 1H), 3.56 (ddd, J ) 10.2, 5.2,
2.8 Hz, 1H) 3.46 (s, 3H), 3.47-3.41 (m, 3H), 3.24 (dd, J ) 8.3,
8.3 Hz, 1H); 13C NMR 100 MHz (CD3OD) δ 106.2, 81.3, 74.7,
Supporting Information Available: 1H and 13C NMR
spectra for 1, 2, and 5-7. 13C-Coupled HMQC and NMRSim
overlays for 1 and 2. Summary of energies, geometries,
vibrational frequencies, and NMR coupling constants for each
conformer. This material is available free of charge via the
Internet at http://pubs.acs.org.
JO048932Z
(47) Blackburn, B. J.; Grey, A. A.; Smith, I. C. P.; Hruska, F. E.
Can. J. Chem. 1970, 48, 2866.
(48) (a) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; de Leeuw, H. P.
M.; Altona, C. Recl. Trav. Chim. Pays-Bas 1979, 98, 576. (b) Altona,
C.; Sundaralingam, M. J. Am. Chem. Soc. 1973, 95, 2333.
38 J. Org. Chem., Vol. 70, No. 1, 2005