Orthoesters versus 2-O-acyl glycosides as glycosyl donors: Theorectical and experimental studies

Article abstract of DOI:10.1002/chem.200304856

n-Pentenyl orthoesters (NPOEs) undergo routine acid catalyzed rearrangement into 2-O-acyl n-pentenyl glycosides (NPGs). The reactant and product can both function as glycosyl donors affording 1,2-trans linked glycosides predominantly. However, both donors

Full text of DOI:10.1002/chem.200304856

FULL PAPER
Orthoesters versus 2-O-Acyl Glycosides as Glycosyl Donors: Theorectical and
Experimental Studies
Bert Fraser-Reid,*[a] Stefan Grimme,*[b] Manuel Piacenza,[b] Mateuz Mach,[a] and
Urs Schlueter[a]
Abstract: n-Pentenyl
orthoesters
counterparts. The calculations show that
in the case of a manno NPG and NPOE
pair, each donor goes initially to a
different cationic intermediate. Thus,
the former goes to a high-energy oxo-
carbenium ion before descending to a
trioxolenium ion in which the charge is
distributed over the pyrano ring oxygen,
as well as the carbonyl and ether oxygen
atoms of the putative C2 ester. On the
other hand, ionization of the NPOE
produces a dioxolenium ion lying slight-
ly above the more stable trioxolenium
counterpart. For the gluco pair, the NPG
also goes to a very high-energy oxo-
carbenium ion, which also descends to a
trioxolenium ion. However, unlike the
manno analogue, the gluco NPOE does
not give a dioxolenium ion; indeed, the
dioxolenium is not energetically distin-
guishable from the trioxolenium coun-
terpart. The theoretical observations
have been tested experimentally. Thus,
it was found that with manno deriva-
tives, the orthoester is a more reactive
donor than the corresponding NPG
donor, whereas, for gluco derivatives,
there is no advantage to using one over
the other, unless one resorts to carefully
selected promoters.
(NPOEs) undergo routine acid cata-
lyzed rearrangement into 2-O-acyl n-
pentenyl glycosides (NPGs). The reac-
tant and product can both function as
glycosyl donors affording 1,2-trans
linked glycosides predominantly. How-
ever, both donors differ in their rates of
reactions, the yields they produce, and
the nature of their byproducts, indicat-
ing that the NPOE/NPG pair may not be
reacting through the same intermedi-
ates. We have therefore applied quan-
tum chemical calculations using DFT
methods and MP second order pertur-
bation theory to learn more about
orthoesters and their 2-O-acyl glycosidic
Keywords: carbohydrates ¥ density
functional calculations ¥ glycosides
Introduction
2-O-Acyl glycosyl bromides, 1, are readily converted[1] into
cyclic 1,2-orthoesters such as 2, the OR' unit of which may be a
simple alkoxy residue, or a complex oligosaccharide.[2] Under
acid catalysis, rearrangement of the latter to a 2-O-acyl
glycoside, 3, is the major reaction pathway,[3] although stereo-
electronic controlled decomposition[4, 5] to give glycosyl esters
such as 4 has been reported (Scheme 1).[6]
n-Pentenyl orthoesters (NPOEs) (i.e., 2: R' ˆ pent-4-enyl)
are unique in that either they or their n-pentenyl glycoside
(NPG) rearrangement products (i.e., 3: R' ˆ pent-4-enyl) can
serve as glycosyl donors leading to the same glycosidation
[a] Prof. Dr. B. Fraser-Reid, Dr. M. Mach, Dr. U. Schlueter
Natural Products and Glycotechnology Research Institute Inc. (NPG)
4118 Swarthmore Road, Durham, NC 27707 (USA)
Fax : (‡1)919-493-6113
Scheme 1.
E-mail: Dglucose@aol.com
[b] Prof. Dr. S. Grimme, M. Piacenza
Organisch Chemisches Institut der Universitaet M¸nster
Corrensstrasse 40, 48149 M¸nster (Germany)
Fax : (‡49)251-83-36515
product for example 10.[7] Thus, when treated with an
halonium ion, both release a halomethyl furan 5; this reaction
results in the initial formation of cations 6 and 8, respectively,
E-mail: grimmes@uni-muenster.de
Chem. Eur. J. 2003, 9, 4687 4692
DOI: 10.1002/chem.200304856
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4687

B. Fraser-Reid, S. Grimme et al.
FULL PAPER
which are interconvertible via a highly delocalized species
symbolized by 7. Thus, the charge can reside on one, two or
three oxygen atoms in the oxocarbenium, dioxolenium and
trioxolenium ions 8, 6 and 7, respectively.
However, studies in our laboratory suggest that, in terms of
rates, yields, and so on, donors such as NPOE 2 and NPG 3 are
not ™the same∫. Thus, there may be occasions when one serves
better than the other. We are anxious to know whether this
outcome is related to the cation 6 or 8 generated initially by
the donor in question. We have
of the NBS, the molar amounts of the unreacted donors are in
direct proportion to their relative reactivity.[17]
The donors of interest are shown in Scheme 2a. We planned
to use preparative thin-layer chromatography to isolate the
unreacted donors, but the Rf values of the gluco pair, 11 and
12, 0.48 and 0.43, respectively, were too close to permit
convenient separation. It was therefore decided to use a
™reporter∫ donor as a reference substrate. Compound 17, the
6-O-methyl analogue of 12, seemed an appropriate candidate,
therefore applied quantum
chemical calculations, and pro-
vided experimental support, to
learn more about orthoesters
and their 2-O-acyl glycosidic
counterparts, and to determine
if there may be advantages to
using one or other on occasion.
Computational Methods
All calculations were performed on a
800 MHz LINUX-PC (2GB RAM) us-
ing the TURBOMOLE 5.3[8] program
Scheme 2.
suite. Geometry optimizations were
carried out using the BP[9, 10] and B3-
LYP,[11, 12] density functional theory
a
valence triple basis-set (TZVP).[13] B3-LYP
and the compound was readily prepared as shown in
Scheme 2b. Thus, the gluco NPOE derivative 15, prepared
in the usual way,[7] was converted by routine transformations
into 16a d, and rearrangement led to 17, which was found to
have a substantially different Rf value of 0.33, and was
therefore an ideal ™reporter∫substrate.
(DFT) methods, and
optimizations for the very large 3,4,6-tri-O-benzyl-2-O-benzoyl-pyrano-
sides were carried out using the smaller SV(P)[14] basis-set. For all
optimized structures, M˘ller Plesset second order pertubation theory[15]
single-point calculations were performed using the RI-approximation
(RIMP2)[16] and the TZVPP[13] basis set which includes additional polar-
ization functions on all atoms. All calculations were performed without
symmetry restrictions (C1 point group) and all stationary points were
verified as minima on the energetic potential surfaces by the absence of
imaginary frequencies from vibrational normal mode calculations at the
B3LYP/TZVP level of theory (12 and 14: BP/SV(P)). The relative energies
are given with respect to the lowest lying donors (14 and 19). For the
reaction energies from the neutral reactants to the cationic products,
separate calculations for a free chloride anion were performed on the
corresponding level of theory. The thus received energy was then
substracted from the donor cation difference.
Calculations based on model donors: In order to calibrate our
methodology, we carried out DFT calculations with two
different basis sets to determine the relative energies of the
anomeric n-pentenyl 3,4,6-tri-O-benzyl-2-O-benzoyl-b-gluco-
and a-manno-pyranosides, 12 and 14 (Table 1). The pyranosyl
donors were modeled by tetrahydropyrans having a C1
chloride as leaving group and a C2 acetoxy (sugar numbering)
as illustrated for 18 and 19 (Table 1). The DFT energies
(entries i and ii) refer to fully optimized calculations, whereas
the MP2 energies (entry iii) are single-point calculations using
the DFT geometries of the preceding row. Both methods
show the b-glucoside 12 to be of higher energy than the
Erx ˆ Edonor À Ecation À EClÀ
Results and Discussion
We have previously shown that
competitive oxidative hydroly-
sis provides a ready procedure
for evaluating the relative reac-
tivity of a pair of NPG do-
nors.[17] The success of the meth-
od relies on the competitive
oxidative hydrolysis of one
equivalent each of the two do-
nors with one equivalent of N-
Table 1. Relative stabilities of experimental and model pyranosyl donors [kcalmolÀ1].
Calculation
method
Experimental substrates
Model donors
a-manno
b-gluco
a-manno
b-gluco
i
ii
iii
B3-LYP/SV(P)
B3-LYP/TZVP
MP2/TZVPP//
B3-LYP/TZVP
2.7
2.8
0.0
0.0
4.0
3.5
0.0
0.0
bromosuccinimide
(NBS).
Upon complete disappearance
4688
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 4687 4692

Glycosyl Donors
4687 4692
a-mannoside 14 in keeping with established experimental
ing column. There are differences, albeit very small, between
the manno and gluco oxocarbenium ions 24a and b, with
either calculation method. However, for the di- and triox-
olenium ions, the DFT and MP2 energies are very similar
owing, presumably, to the rigid geometry of the bicyclic
structures.
The energy values in Table 3 show that the di- and
trioxolenium ions 25a and 26a, were found as discrete
structures in the case of the manno derivatives, whereas for
the gluco derivatives, there was no energy difference between
25b and 26b by either MP2 or DFT calculation.
20]
observations and earlier theoretical calculations.[18
Orthoester donors: The cyclic 1,2-orthoesters were modeled
by tetrahydropyrans with a 1,3-dioxolane ring fused at C1 and
C2, with geminal methyl and chloride groups in the endo or
exo orientation at C3 of the dioxolane moiety as shown in
Table 2.
With regard to the orthoesters, the energies for the trans
series 22/23 (Table 2) are substantially higher than those for
the cis analogues 20/21, as is to be expected for such fused 6/5
bicyclic systems.[21] For the cis-gluco (20a, b) derivatives, the
exo isomers are higher in energy than the endo (20a > 20b),
whereas for the cis manno (21a, b) derivatives, the situation is
reversed, endo being higher in energy than exo (21b > 21a).
For the trans derivatives all endo isomers are slightly higher in
energy, (i.e., 22b > 22a and 23b > 23a) in all calculations.
Experimentally, the n-pentenyl analogues of the exo isomers
are obtained as the major substrates.
These energy data are supported by the MP2 derived
geometries depicted in Table 4. Thus, in the manno derivative
À
25a, both C O bond lengths in the five-membered ring are
the same (1.49Æ 0.01ä) as expected for a pure dioxolenium
ion, whereas in 26a the bond to the anomeric center is much
longer (1.57 versus 1.48 ä), as expected for the trioxolenium
ion. However, the gluco analogues 25b and 26b show
À
identical C O bond lengths of 1.60 and 1.48 ä; this indicates
that both structures are the same. Furthermore, the fact that
À
the anomeric C O bond is longer (1.60 versus 1.50 ä) assigns
the structure as a tri- rather than a dioxolenium ion.
Calculation of the energy and structure of cationic products:
The energies for the manno and gluco cationic intermediates
in Table 3 are shown, which arise from the donor models
shown in Tables 1and 2. The DFT values refer to fully
optimized calculations, whereas the MP2 values are single-
point calculations using the DFT geometries of each preced-
From donors to cations: The reaction energies for conversion
of the donors into the cationic intermediates are shown over
the arrows in Figure 1. In both the manno and gluco cases, the
Table 2. Relative energies[a] of model orthoester donors [kcalmolÀ1].
Calculation method
cis
trans
a-gluco
b-manno
b-gluco
a-manno
20a
20b
21a
21b
22a
22b
23a
23b
X ˆ CH3/Yˆ Cl X ˆ Cl/Yˆ CH3 X ˆ CH3/Yˆ Cl X ˆ Cl/Yˆ CH3 X ˆ Cl/Yˆ CH3 X ˆ CH3/Yˆ Cl X ˆ CH3/Yˆ Cl X ˆ Cl/Yˆ CH3
exo
endo
exo
endo
endo
exo
endo
exo
i B3-LYP/TZVP
ii MP2/TZVPP//
B3-LYP/TZVP
17.3>16.2
13.4>12.5
15.6<18.7
11.0<13.4
25.3>24.9
19.7>19.0
29.7>29.3
24.4>23.8
[a] With respect to derivative 19.
Table 3. Relative energies[a] [kcalmolÀ1] of model cationic substrates.
Calculation method
Oxocarbenium ions
Dioxolenium ions
Trioxolenium ions
manno
gluco
manno
gluco
manno
gluco
i
ii
B3-LYP/TZVPP
MP2/TZVPP//
63-LYP/TZVP
145.6
159.9
144.5
158.7
135.3
142.9
133.1
142.1
133.1
141.9
133.1
142.1
[a] With respect to derivative 19 À ClÀ1
.
Chem. Eur. J. 2003, 9, 4687 4692
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4689

B. Fraser-Reid, S. Grimme et al.
FULL PAPER
3, is (usually) 1,2-trans, and so we need to be concerned
ONLY with b glucosides, (e.g. 18) and a mannosides (e.g.
19). (However, this is not a major issue since the reaction
energies for the unrepresented anomers are virtually the
same.)
iii) Only the DFT energies in Figure 1(lower values over the
arrows) are shown for the transitions in Figure 2.
The oxocarbenium ions 24a and b are very high in energy–
as is to be expected in view of the absence of charge
delocalization. The relative energies in Table 4 show that the
trioxolenium ion 26a is 1.5 3.4 kcalmolÀ1 more stable than
the corresponding dioxolenium derivative 25a, the DFT
results being higher than the MP2.
For experimental verification, the ™reporter∫ 17, was first
tested against its 6-O-benzyl analogue 12. In a typical
experiment, 1:1:1 molar amounts of 12, 17, and NBS were
allowed to react for at least 10 h; this is the length of time
which had previously been shown to be adequate for complete
hydrolysis of reporter 17. Unreacted 17 and 12 were isolated
by preparative thin-layer chromatography. On the basis of the
recovered amounts, their relative reactivity, k12/17, was found
to be 0.93 (Table 5, entry i). Thus, compounds 17 and 12 have
virtually the same reactivity, which confirms that for our
purposes, 6-O-methyl and 6-O-benzyl protecting groups were
equivalent for our measurements.
The same procedure was applied to determine the relative
reactivity of other pairs of donors: 17 and 11, 17 and 13, and 17
and 14, which were found to be 0.99, 1.15 and 3.19,
respectively (Table 5, entries ii iv). The results, presented
in Table 5, are consistent with the theoretical findings:
1) first, the ratios in entries iii) and iv) indicate that the
mannosides are ™more stable∫ than the glucosides, as
suggested by the energies in Table 1;
Figure 1. Calculated transition energies [kcalmolÀ1
cationic intermediates (upper values from MP2, lower values from B3-
LYP).
] from donors to
direct reaction from pyranosyl donor to trioxolenium ions is
favored over proceeding to the corresponding oxocarbenium
ions. For example in the manno case, transition 19 ! 26a is
more energetically favorable than the alternative 19 ! 24a.
Similarly in the gluco case, transition 18 ! 26b is preferred to
18s ! 24b. These DFT calculations (lower values over the
arrows in Figure 1) give a difference of ꢀ12 kcalmolÀ1 for
both manno and gluco cases in favor of the trioxolenium ions.
(Notably the energy differen-
2) second, the experimental value of 0.99 for the NPOE
and NPG, 11 and 17, confirms the theoretical finding
that these gluco counterparts have the same reactivities,
whereas,
ces, according to MP2 calcula-
tions, are slightly higher, be-
tween 16 18 kcalmolÀ1.)
Comparison of theoretical and
experimental results
Figure 2 gives another repre-
sentation of the data in Figure 1
with the following simplifying
modifications:
i) The trans orthoesters have
been ignored, because the
manno derivatives 23a and
b are impossible, and the
gluco analogues, 22a and b
are so highly strained as to
be uselessly unstable.
ii) As indicated in Scheme 1,
the rearrangement product
Figure 2. B3-LYP/TZVP Transition energies [kcalmolÀ1] from donors to cations.
www.chemeurj.org
of orthoesters, for example
4690
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2003, 9, 4687 4692

Glycosyl Donors
4687 4692
Table 4. Relative energies[a] [kcalmolÀ1] and optimized geometries of
dioxolenium and trioxolenium ions.
Experimental Section
Structure
MP2
TZVP
BP
TZVP
B3-LYP
TZVP
B3-LYP
TZVPP
General: All NMR spectra were recorded on GE 300 or Varian 400 MHz
NMR spectrometers and chemical shifts are reported relative to internal
TMS. Mass spectrometry was performed at the Duke University Depart-
ment of Chemistry Mass Spectrometry Facility. Chemical Ionization (CI)
was done on a Hewlett-Packard 5988A GC/MS using 1% ammonia in
methane as the reagent gas, with a source temperature of 1008C, at 1Torr.
High resolution mass spectra (HRMS) and fast atom bombardment (FAB)
analyses were recorded with a JEOL JMS-SX102A mass spectrometer
operating at 10 K resolution, using a dithiothreitol/dithioerythritol or m-
nitrobenzyl alcohol as the matrix with xenon as the fast atom. All reactions
were conducted under argon atmosphere. Thin-layer chromatogra-
phy(TLC): Riedel-de Haen, coated with silica gel 60F 254 and were
detected by UV or by spraying or dipping in a solution of ammonium
molybdate (6.25 g) and cerium(iv) sulfate (25 g) in 10% aqueous sulfuric
acid (250 mL) and subsequent heating. Flash column chromatography was
performed on silica gel (spectrum SIL 58, 230 400 mesh, grade 60) using
mixtures of hexane and ethyl acetate as eluants. Dichloromethane and
toluene were distilled from CaH2. N-Bromosuccinimide was purchased
from Aldrich and recrystallized from hot water and dried on vacuum.
manno
i
0.0
1.5
0.0
3.4
0.0
2.2
0.0
2.2
ii
gluco
The n-pentenyl orthoeters, NPOEs 11, 13 and 15, and 2-O-benzoyl n-
pentenyl glycosides NPG 12 and 14 were prepared as previously
described.[22]
iii
0.5
0.5
À 0.10.0
À 0.0(1)
À 0.0(4)
Pent-4-enyl 3,4-di-O-benzyl-6-O-methyl-b-d-glucopyranose (17): a-d-Glu-
copyranose 1,2-(pent-4-enyl orthobenzoate), 15,[22] (2.2 g, 6.2 mmol),
diisopropylethylamine, (2.2 mL, 12.7 mmol), 90% triisopropylsilyl chloride
(2 mL, 8.4 mmol) and DMAP (50 mg, 0.4 mmol) were dissolved in dry
dichloromethane (20 mL) and stirred overnight at room temperature.
Water was added and the product was extracted into ethyl acetate. The
organic layer was washed with water, brine, dried and chromatography on
silica (hexanes/ethyl acetate 6:1 ! 1:1) provided the syrupy diol 16a
(2.67 g, 84%). The compound was dissolved in dimethylformamide
(50 mL). Sodium hydride (50% suspension in mineral oil, 2.0 g, 41.2 mmol)
was added and the reaction mixture was stirred at 08C for 30 min. Benzyl
bromide (2.0 mL, 16.8 mmol) was then added dropwise, the temperature of
the reaction being maintained below 108C. Then, cooling bath was
removed and the reaction mixture was stirred at room temperature until
TLC (hexanes/ethyl acetate 4:1) showed full disappearance of the starting
material and formation of a new, less polar product (ca. 1h). The reaction
mixture was diluted with diethyl ether, cooled to 08C, and water was
carefully added to decompose the excess sodium hydride. The product was
extracted into diethyl ether, the organic layer was washed with water, brine,
dried over sodium sulfate and concentrated. Column chromatography
(hexanes/ethyl acetate 9:1 ! 5:1) provided product 16b (3.29 g, 91%). The
material was dissolved in THF (10 mL), and added to a mixture of 2,6-
lutidine (0.2 mL, 1.7 mmol) and TBAF (1m in THF, 10 mL). The reaction
mixture was stirred at room temperature overnight. Water was added and
the product was extracted into ethyl acetate. The organic layer was washed
with water, brine, dried over sodium sulfate and concentrated. The crude
product, 16d, was treated with methyl iodide under the same conditions as
used for the above-described benzylation, except that cooling was omitted.
After column chromatography the product 16d was obtained (3.29 g,
91%). Compound 16c was directly dissolved in dry dichloromethane
(10 mL) under argon, 4-pentenol (50 mL, 0.49 mmol) and TBDMSOTf
(10 mL, 0.044 mmol) were added. The reaction mixture was stirred at room
temperature for 2 min and then diluted with diethyl ether. Water was added
and the product was extracted with diethyl ether. The organic layer was
washed with 2% sulfuric acid, water, saturated NaHCO3, water, brine,
dried over sodium sulfate and concentrated. Column chromatography
(hexanes/ethyl acetate 9:1 ! 3:1) provided provided NPG 17 (1.83 g, 70%)
‰bŠ
iv
À 0.1
À 0.0(4)
[a] With respect to derivative 19 À ClÀ1. [b] Not stable.
Table 5. Competitive oxidative hydrolysis between ™reporter∫ 17 and other
donors.
Competitors Before reaction Unreacted amounts Relative ratio
[mol  10À4
]
[mol  10À4
]
17
other donor 17
other donor other donor/17
gluco
i
ii
17 vs 12b
17 vs 11
1.6061 1.6058
1.6061 1.6058
0.9723 0.9025
0.8067 0.7985
12b/17 ˆ 0.93
11/17 ˆ 0.99
manno
iii
iv
17 vs 13
17 vs 14a
1.6061 1.6058
1.6061 1.6058
0.8086 0.9266[a]
0.4336 1.3626
13/17 ˆ 1.15
14a/17 ˆ 3.14
[a] On the basis of 1H NMR analysis of the reaction mixture containing 17 and
13.
3) third, the manno NPG and NPOE donors display very
different relative reactivities (entries iii and iv), the ratios
of 1.15 and 3.14 indicating that orthoester 13 is 2.73 times
more reactive than the NPG counterpart 14.
In summary, the experimental results support the theoret-
ical findings, that the manno NPOE and NPG pair initially go
to different intermediates while the gluco counterparts go to
the same intermediate. In other words, the manno derivatives,
the orthoester is a better or more reactive donor than the
corresponding glycoside, whereas for gluco derivatives, there
is no advantage to using one or the other.
1
as a syrup. H NMR (300 MHz, CDCl3): d ˆ 8.01(d, 2H, J ˆ 7.2 Hz, ortho
protons from benzoate), 7.59 7.33 (m, 13H, arom.), 5.69 5.55 (m, 1H, H-4
from pent.), 5.26 (dd, 1H, J ˆ 8.1, 8.7 Hz, H-2), 4.88 4.62 (m, 6H), 4.48 (d,
1H, J ˆ 7.2 Hz, H-1), 3.91 3.60 (m, 5H), 3.52 3.39 (m, 2H), 3.40 (s, 3H,
OCH3), 2.00 1.86 (m, 2H), 1.64 1.48 (m, 2H); HR-LSIMS:m/z: calcd for
‡
C33H38O7Na: 569.2515; found: 569.2523 [M ‡Na].
Conditions for competition reactions: Our recently described procedure
for determining the relative reactivity of two NPGs[17] was used. Thus, the
hydrolysis solution was prepared from acetonitroile (49.5 mL), NBS
Chem. Eur. J. 2003, 9, 4687 4692
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4691

B. Fraser-Reid, S. Grimme et al.
FULL PAPER
(0.143 g, 0.8038 mmol) and water (0.5 mL). The ™reporter donor∫ 17
(87.8 mg, 1.606 mmol) and the substrate donor (1.605 mmol) were dissolved
in acetonitrile (2 mL), and the hydrolysis solution (10 mL) was added. The
reaction mixture was left for 10 h, previous tests having shown that this
time is sufficient to hydrolyse the ™reporter donor∫ 17 completely. Diethyl
ether was then added and the reaction mixture was quenched with 10%
Na2S2O3, washed with water, brine, dried, concentrated and subjected to a
silica gel column to clean up the material. A portion of the product was
then separated by preparative layer chromatography, the zones of interest
being detected by UV light, scraped off, extracted with purified ethyl
acetate and weighed.
[4] J. F. King, A. D. Allbutt, Tetrahedron Lett. 1967, 49 53; J. F. King,
A. D Allbutt, Can. J. Chem. 1970, 48, 1754 1769.
[5] P. Deslongchamps, Stereoelectronic Effects in Organic Chemistry,
Pergamon Press, New York, 1983, Chapter.
[6] R. U. Lemieux, A. R. Morgan, IUPAC Sym. on the Chemistry of
Natural Products Kyoto, April 1964.
[7] C. Roberts, C. L. May, B. Fraser-Reid, Carbohydr. Lett. 1994, 1, 89
93; J. G. Allen, B. Fraser-Reid, J. Am. Chem. Soc. 1999, 121, 468 469.
[8] Electronic structure calculations on workstation computers, The
program system TURBOMOLE: R. Ahlrichs, M. Bar, M. H‰ser, H.
Horn, C. Kolmel, Chem. Phys. Lett. 1989, 162, 165 169.
[9] A. D. Becke, Phys. Rev. A 1988, 38, 3098 3100.
[10] J. Perdew, Phys. Rev. B 1986, 33, 8822 8824.
[11] A. D. Becke, J. Chem. Phys. 1993, 98, 5648 5652.
[12] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785 789.
[13] A. Schafer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829
5835.
Acknowledgements
B.F.-R. is grateful to the Human Frontier Science Program Organization,
The Mizutani Foundation and The National Science Foundation for
support of this work.
[14] A. Schafer, A. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571 2577.
[15] C. M˘ller, M. S. Plesset, Phys. Rev. 1934, 46, 618.
[16] F. Weigend, M. Haser, Theor. Chem. Acc. 1997, 97, 331 340.
[17] B. G. Wilson, B. Fraser-Reid, J. Org. Chem. 1995, 60, 317 320.
[18] Anomeric Effect, Origin and Consequences, Vol. 87 (Eds.: W. A.
Szarek, D. Horton), ACS Symposium Series Washington, DC, 1979.
[19] A. J. Kirby in The Anomeric Effect and Relate Stereoelectronic Effects
at Oxygen, Spektrum Verlag, Heidelberg, 1983, Chapter 2.
[20] The Anomeric Effect and Associated Stereoelectronic Effects, Vol. 539
(Ed.: G. R. J. Thatcher), ACS Symposium Series, Washington, DC,
1993.
[1] See for example: R. Lemieux, T. Takeda, B. Y. Chung, ACS Symp. Ser.
1976, 39; N. E. Franks, R. Montgomery, Carbohydr. Res. 1968, 6, 286
298; H. B. Boren, G. Ekborg, K. Eklind, P. J. Garegg, A. Pilotti, C. G.
Swahn, Acta Chem. Scand. 1973, 27, 2639 2648; N. K. Kochetkov,
A. Y. Khorlin, A. F. Bochkov, Tetrahedron 1967, 23, 693 699; S. E.
Zurabyan, M. M. Tikhomirov, V. A. Nesmeyanov, A. Y. Khorlin,
Carbohydr. Res. 1973, 26, 117 123; A. F. Bochkov, Y. V. Voznyi,
Carbohydr. Res. 1974, 32, 1 8.
[2] T. Ogawa, T. Nukada, Carbohydr. Res. 1988, 136, 135 153; W. Wang,
F. Kong, J. Org. Chem. 1998, 63, 5744 5745; L. Sznaidman, C.
Johnson, C. Crasto, M. Hecht, J. Org. Chem. 1995, 60, 3942 3943; T.
Ogawa, K. Katano, K. Sasajima, M. Matsui, Tetrahedron 1981, 37,
2779 2786.
[21] E. L. Eliel in Stereochemistry of Carbon Compounds, McGraw Hill
New York, 1962, Chapter 10.
[22] M. Mach, U. Schlueter, F. Mathew, B. Fraser-Reid, K. C. Hazen,
Tetrahedron 2002, 58, 7345 7354.
[3] See for example: A. F. Bochkov, G. E. Zaikov, Chemistry of the
O-Glycosidic Bond, Pergamon Press, Oxford (UK), 1979.
Received: February 14, 2003 [F4856]
4692
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 4687 4692