Observations on the activation of methyl thioglycosides by iodine and its interhalogen compounds

Article abstract of DOI:10.1016/S0957-4166(99)00501-7

Treatment of 'armed' methyl thiogalactosides with iodine in the absence of an acceptor alcohol results in thioglycoside epimerisation, whereas there is no effect on the corresponding 'disarmed' methyl thioglycosides. In contrast, iodine-hexamethyldisilane (which generates iodotrimethylsilane in situ) brings about epimerisation of 'disarmed' thioglycosides, ultimately giving rise to the corresponding α-glycosyl iodides on extended exposure. Cross-over experiments show the former iodine-promoted epimerisation process to be intermolecular, whereas the latter iodine-hexamethyldisilane-promoted epimerisation is intramolecular. Treatment of the same methyl thiogalactosides with iodine monobromide gives rise to the thermodynamically favoured α-glycosyl bromides, whereas reaction with iodine monochloride initially gives the kinetic β-glycosyl chlorides, which slowly epimerise to the thermodynamic α-linked products. Differences in the outcome of thioglycoside activation by I-I, I-Br and I-Cl suggest there may be scope for influencing the stereochemical course of thioglycoside-based glycosylation reactions through careful choice of promoter. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(99)00501-7

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 11 (2000) 581–593
Observations on the activation of methyl thioglycosides by iodine
and its interhalogen compounds1,2
K. P. Ravindranathan Kartha,∗ Peter Cura, Mahmoud Aloui, S. Kristy Readman,
Trevor J. Rutherford and Robert A. Field ∗
School of Chemistry, Biomolecular Sciences Building, University of St Andrews, St Andrews, Fife KY16 9ST, UK
Received 1 November 1999; accepted 10 November 1999
Abstract
Treatment of ‘armed’ methyl thiogalactosides with iodine in the absence of an acceptor alcohol results in
thioglycoside epimerisation, whereas there is no effect on the corresponding ‘disarmed’ methyl thioglycosides.
In contrast, iodine–hexamethyldisilane (which generates iodotrimethylsilane in situ) brings about epimerisation of
‘disarmed’ thioglycosides, ultimately giving rise to the corresponding α-glycosyl iodides on extended exposure.
Cross-over experiments show the former iodine-promoted epimerisation process to be intermolecular, whereas
the latter iodine–hexamethyldisilane-promoted epimerisation is intramolecular. Treatment of the same methyl
thiogalactosides with iodine monobromide gives rise to the thermodynamically favoured α-glycosyl bromides,
whereas reaction with iodine monochloride initially gives the kinetic β-glycosyl chlorides, which slowly epimerise
to the thermodynamic α-linked products. Differences in the outcome of thioglycoside activation by I–I, I–Br and
I–Cl suggest there may be scope for influencing the stereochemical course of thioglycoside-based glycosylation
reactions through careful choice of promoter. © 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Although major advances have been made in glycosylation chemistry over the past twenty-five years,3
precise mechanistic details about such processes remain limited.4 Given the lack of a general procedure
that can be used to obtain good control of anomeric stereochemistry, regardless of the sugar used, a
better understanding of the mechanism of glycosyl donor activation and subsequent reaction would be
invaluable. Whilst many different types of glycosyl donors are in common use, glycosyl halides and
thioglycosides are perhaps the most widely used.5 For some time we have been investigating iodine
and its interhalogen compounds as reagents for the activation of thioglycosides6 and glycosyl halides,7
and for the conversion of the former to the latter.8 In light of the observed conversion of thioglycosides
∗
Corresponding author. E-mail: robert.field@st-and.ac.uk
0957-4166/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00501-7
tetasy 3136

582
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
to glycosyl bromides and chlorides, we were drawn to investigate the possible formation of glycosyl
iodides from thioglycosides on treatment with iodine, and to question whether or not glycosyl halides are
intermediates in iodine- and interhalogen-promoted glycosylation reactions that employ thioglycoside
donors. In addition, in the course of our work on the iodine-promoted glycosylation of sugar alcohols with
‘armed’ methyl thio-β-D-galactosides,6 we observed formation of the corresponding α-thioglycosides
in the reaction mixture, presumably resulting from in situ epimerisation of the β-anomer. Although
depletion of both of the anomers occurred as the reaction progressed, sufficient quantities of the α-
thioglycoside could be isolated in some cases to enable unambiguous confirmation of its structure.9 These
results were not unexpected given the earlier work of the van Boom group,10 which has subsequently been
followed up by Boons and co-workers.11 The latter workers noted that the IDCP-promoted epimerisation
of thioglycosides is an intermolecular process and that the anomeric configuration of thioglycoside
donors can have an effect on the stereochemical outcome of glycosylation reactions.11
Taking all these points into consideration, we were drawn to investigate the importance, or otherwise,
of both thioglycoside epimerisation and glycosyl halide formation on the rate and stereochemical
outcome of glycosylation reactions employing thioglycoside donors (Scheme 1).
Scheme 1. Activation of thioglycosides with iodine and its interhalogen compounds. Possible roles for thioglycoside epimeri-
sation and glycosyl halide formation in the glycosylation process
Due to the potential mechanistic complexity of iodine- and interhalogen-promoted glycosylation with
thioglycosides, we first undertook a study of the activation of methyl thioglycosides with iodine, iodine
monobromide and iodine monochloride in the absence of an acceptor alcohol. Herein we report our
observations on this topic.
2. Results and discussion
2.1. Reaction of methyl thioglycosides with iodine
2.1.1. Epimerisation of ‘armed’ thioglycosides
In keeping with the report of Boons and co-workers,11 we observed the formation of an approximately
50:50 α:β mixture of thioglycosides at equilibrium for iodine-promoted epimerisation of both per-
benzylated methyl α- and β-thiogalactopyranosides, 1 and 2, in either acetonitrile or chloroform. How-
ever, in contrast to the Boons work, which used a stopped assay to study IDCP-promoted epimerisation
(monitored by mass spectrometry; reaction time approx. 1 min), we have been able to continuously
monitor the corresponding iodine-promoted reaction directly by 1H NMR spectroscopy.1 Epimerisation
of β-thioglycoside 2 was substantially slower in chloroform than in acetonitrile, suggesting a role for

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
583
the solvent in stabilisation of charged reaction intermediates in the epimerisation process. Using either
1 or 2 as the starting material an approximately 1:1 mixture of 1 and 2 was obtained in the first 1–2
min in the presence of 1–2 mol. equiv. of iodine at room temperature in acetonitrile. As expected,
the rate of epimerisation could be slowed by conducting the reaction at lower temperatures (−20°C
to ice-bath temperature). For short reaction times (up to 5–10 min) only epimerisation was observed;
at longer reaction times, degradation of reactants was noted, presumably due to quenching of reaction
intermediates by residual moisture, or conceivably due to iodine-promoted benzyl ether cleavage. Once
again these side reactions could be marginalised by carrying out the epimerisation at lower temperatures,
as mentioned above. Another important observation made during these studies was the formation of
dimethyl disulfide (δH: Me 2.43 ppm) as a by-product in these reactions, presumably arising from
spontaneous dimerisation of the methylsulfenyl iodide formed as outlined in Scheme 2.
Scheme 2. Iodine-promoted epimerisation of perbenzylated methyl α- and β-thiogalactopyranosides 1 and 2
These observations are consistent with the intermolecular epimerisation process outlined in Scheme 2.
Perhaps surprisingly, in no case was glycosyl iodide formation observed (δH: H-1α approx. 7.0 ppm)12
in the course of these particular epimerisation studies.
Further support for the intermolecular nature of the above process was obtained from cross-over
experiments employing per-O-benzylated methyl β-thiogalactopyranoside 2 and the corresponding ethyl
thioglucoside 3. When an equimolar mixture of 2 and 3 was reacted with iodine (one mol. equiv.) in
deuteroacetonitrile formation of all the possible thioglycosides was observed (Scheme 3), which is to
be expected on the basis of Boons’ work.11 Interestingly, when this reaction was carried out in the
presence of half mol. equiv. of iodine at approx. −20°C preferential epimerisation of thioglycoside 2 was
observed in the initial stages of the reaction. However, this observation in itself is insufficient to draw
any conclusion on the relative reactivities of methylthio- and ethylthioglycosides as the two glycosides
used are derived from two different sugars. It was also noted in the above reactions that the ratio of
the anomeric glucosides at equilibrium was different from the corresponding values for the anomeric
galactosides. Therefore epimerisation of methyl 2,3,4,6-tetra-O-benzyl-1-thio-β-D-glucopyranoside was
examined independently in the presence of 1 mol. equiv. of iodine in acetonitrile. Indeed the equilibrium
ratio of methyl-α/β-thioglucosides was found to be approximately 2.4, in contrast to the corresponding
galactoside.
2.1.2. Epimerisation of ‘disarmed’ thioglycosides
Iodine alone has little effect on ‘disarmed’ thioglycosides. However, iodine in combination with hexa-
methyldisilane (HMDS) (which generates iodotrimethylsilane in situ)13 gives rise to rapid epimerisation
of such compounds. In contrast to iodine-promoted epimerisation of ‘armed’ thioglycosides, this process

584
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
Scheme 3. Cross-over experiments to study the molecularity of iodine-promoted epimerisation of ‘armed’ thioglycosides
is intramolecular, as demonstrated by the non-appearance of cross-over products arising from treatment
of ‘disarmed’ thioglycosides 4 and 5 with I2–HMDS (Scheme 4). Epimerisation of a mixture of 4 and 5
led only to the formation of a mixture of four compounds, namely, methyl thio-α/β-galactosides and ethyl
1
thio-α/β-glucosides in approximately equimolar quantities (as determined by H NMR spectroscopy).
Moreover, when an equimolar mixture of these four compounds was treated with iodine–HMDS the
‘unchanged’ starting material mixture was recovered.
Scheme 4. I2–HMDS-promoted thioglycoside cross-over experiment using ‘disarmed’ methyl thiogalactoside 4 and ‘disarmed’
ethyl thioglucoside 5
These observations are also supported by the non-appearance of cross-over products in experiments
employing unlabelled and specifically di-13C-labelled methyl thiogalactosides 4 and 6 (Scheme 5). The
synthesis of the latter is outlined in Scheme 6.
Scheme 5. I2–HMDS-promoted thioglycoside cross-over experiment using ‘disarmed’ unlabelled and 13C-labelled methyl
thiogalactosides 4 and 6

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
585
Scheme 6. Synthesis of di-13C-labelled methyl thiogalactoside 6. (i) Ac2O, I2, rt, 6 h, followed by HBr/AcOH, rt, 2 h;14 (ii)
KSAc, DCM, rt, 18 h;15 (iii) NaOMe/MeOH, −40°C, 40 min;15 (iv) 13C-methyl iodide, DMF, rt, 12 h15
Furthermore, the experiments outlined in Scheme 5 did not produce any ‘mixed mass’/cross-over
products, as judged by MALDI-TOF analysis of the product mixture, so supporting the notion of an
intramolecular epimerisation process.
The epimerisation of 4/6 presumably occurs through TMS-I-promoted ring opening of the disarmed
glycopyranoside through interaction of the hard electrophilic silicon centre with the O-5 ring oxygen of
the sugar, followed by ring closure onto the opposite face of the sulfonium ion intermediate (Scheme
7). Accordingly, formation of dimethyl disulfide would not be expected. Indeed, in none of the NMR
experiments referred to in this section was the formation of this compound noted (δH: Me 2.43 ppm).
Scheme 7. Possible mechanism of I2–HMDS-promoted epimerisation of ‘disarmed’ thioglycosides
2.2. Reaction of methyl thioglycosides with iodine monobromide and iodine monochloride
We have previously demonstrated the use of I–Br for the conversion of thioglycosides to glycosyl
bromides.8 At temperatures ranging from 0°C to 22°C, treatment of thioaldosides with I–Br gave the
corresponding α-glycosyl bromides, irrespective of the presence or otherwise of a C-2 participating
group. With I–Cl, on the other hand, reaction of gluco- and galacto-configured thioglycosides possessing
a C-2 participating group invariably gave the corresponding β-glycosyl chloride as the initial product
(Scheme 8). More detailed studies on the reaction of I–Br and I–Cl with ‘disarmed’ β-thiogalactoside 4
are summarised in Table 1.
When β-thiogalactoside 4 was treated with I–Br (Table 1, entry 1) in dichloromethane at ice-bath
temperature, complete disappearance of the thioglycoside occurred in a few minutes and the α-bromide
7α was obtained in quantitative yield following an aqueous work-up. When the reaction was repeated
using acetonitrile as the solvent, α-halide 7α was once again obtained as the sole organic product (Table
1, entry 2). On the other hand, similar reactions using I–Cl as the activator gave anomeric mixtures of the
glycosyl chlorides 8 regardless of the solvent used (Table 1, entries 3–7). The thermodynamically less

586
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
Scheme 8. Transformation of thioglycosides to glycosyl halides with iodine monobromide and iodine monochloride
Table 1
Reaction of methyl thiogalactoside 4 with iodine monohalidesa
favoured β-anomer was found to be in approximately 10-fold excess in each case. This observation is not
without precedent; Glaudemans and co-workers16,17 noted formation of 6-O-acetyl-2,3,4-tri-O-benzyl-
β-D-glucopyranosyl chloride on treatment of the corresponding phenyl 1-thio-α-D-glucopyranoside
derivative with chlorine in carbon tetrachloride. These workers argued that their findings were at variance
with the results of Fugedi and co-workers,18 as well as Weygand and Ziemann,19 who obtained α-
glycosyl bromides on treatment of benzyl-protected α- and β-thioglycosides with bromine. In light of
the above observations, and the apparent conflict in the literature, a further study on the mechanism of
formation and epimerisation of glycosyl bromides and chlorides was undertaken.
2.2.1. Formation of ‘disarmed’ glycosyl bromides and chlorides
Addition of iodine monobromide to a solution of acetylated β-thiogalactoside 4β in a suitable solvent
presumably results in the attack of the halophilic thioglycoside sulfur atom on the electron deficient
iodine in the I–Br molecule, leading to formation of the corresponding iodo-sulfonium salt 9 (Scheme
9). It is not clear how the reaction proceeds subsequently. Formation of the thermodynamically favoured
glycosyl bromide 7α could arise through direct SN2 displacement of MeSI by bromide ion (path A).
Alternatively, anchimeric assistance from the 2-O-acetyl group could release MeSI to give dioxolenium

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
587
ion 10, which on trapping with bromide would give the less favoured anomer, 7β. The β-halide could then
undergo reaction with bromide to give the thermodynamic product 7α (path B). If path B is followed,
conversion of the bromide 7β to its α-anomer cannot be rate-limiting since no build-up of the former is
observed.
Scheme 9. Mechanisms for reaction of ‘disarmed’ methyl β-thiogalactoside 4β with I–Br and I–Cl
In contrast to the reaction of I–Br with β-thiogalactoside 4β, similar experiments with I–Cl gave
rise initially to the thermodynamically less favoured β-configured glycosyl chloride 8β. NMR time-
course experiments showed that the β-anomeric chloride underwent slow epimerisation to the thermody-
namically more stable α-chloride 8α in a few hours at room temperature. The rate of epimerisation
was dependent on both I–Cl concentration and temperature (Table 2). Taken together, these observations
suggest that path B (Scheme 9) is in operation, which is in agreement with the conclusions drawn from
polarimetric studies on the anomerisation of tetra-O-acetyl-β-D-glucopyranosyl chloride by Lemieux and
Hayami.20 Differences in the anomeric equilibria involving glycosyl bromides and glycosyl chlorides are
not entirely surprising considering the differences in the C–Br and C–Cl bond energies. We cannot, how-
ever, rule out the involvement of path A in the direct formation of α-chloride 8α from β-thioglycoside
4β. Although no direct evidence could be obtained for the intermediacy of dioxolenium ion 10 in this
study, we note the observation of such species (by NMR spectroscopy at low temperature) has recently
been made by Crich and co-workers.21
Activation studies were subsequently undertaken with the isomeric ‘disarmed’ α-thiogalactoside 4α.
Reaction of this compound with I–Cl gave rise to the β-chloride 8β (Table 2, entries 6–8), with only
traces of anomer 8α appearing after 24 h at room temperature in the presence of excess I–Cl (entry 8).
This latter observation is in sharp contrast to the results obtained in the reaction of the β-thioglycoside
4β with I–Cl, where formation of the β-chloride 8β was always accompanied by the formation of its
corresponding α-anomer, 8α, in the initial stages of the reaction, albeit in small quantities. The very low
quantities of 8α in these studies suggest that epimerisation of β-chloride 8β is inherently slow under
the reaction conditions. This adds weight to the argument that formation of at least some 8α from β-
thioglycoside 4β proceeds by a direct displacement mechanism (Scheme 9, path A). Formation of 8β
from 4α by direct displacement (Scheme 10, path A) is also considered likely, since an SN1 mechanism
(path B) would necessarily involve cyclic oxocarbonium ion 11. Presumably 11 would be liable to react
with chloride ion affording either both 8α and 8β, or preferentially 8α. Involvement of both 11 and

588
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
Table 2
Influence of I–Cl concentration and temperature on the reaction of ‘disarmed’ thiogalactoside 4a
dioxolenium ion 10 in formation of 8β from 4α cannot be ruled out, but it requires a rather convoluted
set of interconversions for what is, overall, a facile process.
Scheme 10. Mechanisms for reaction of ‘disarmed’ methyl α-thiogalactoside 4α with I–Cl
2.2.2. Formation of ‘armed’ glycosyl bromides and chlorides
Reactions of ‘armed’ thioglycosides bearing a non-participating protecting group at C-2 were next
investigated. The reaction of benzylated thiogalactosides 1 and 2 with I–Br and I–Cl (resulting in
formation of the glycosyl halides 12 and 13, respectively) was monitored directly by NMR spectroscopy
(Table 3).
When ‘armed’ β-thiogalactoside 2 was treated with iodine monohalides only the α-linked glycosyl
halides 12 and 13 could be detected by NMR spectroscopy (Table 3, entries 1–3). When less than
one mole equivalent of the interhalogen was employed the respective α-halide proportionate to the

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
589
Table 3
Reaction of ‘armed’ thiogalactosides 1/2 with iodine monohalidesa
reagent added was obtained, the remaining thioglycoside being returned unaffected. In contrast to the
reaction of the corresponding ‘disarmed’ thiogalactosides 4, the formation of β-linked halides in the
reaction of ‘armed’ β-thiogalactoside 2 with interhalogens was not observed. This suggests the classical
cyclic oxocarbonium ion (see Scheme 2) is not an intermediate in this particular glycosyl halide forming
process, since it would be expected to react to give both anomers of 12 and 13 in the presence of I–Br
and I–Cl, respectively. This is particularly true in the reaction with I–Cl, where we have already noted
that epimerisation of the β-halide is slow.
Reaction of ‘armed’ α-thiogalactoside 1 with I–Br gives exclusively the α-bromide 12α, which is
to be expected given the ease with which the ‘armed’ β-bromide 12β is expected to epimerise to its
thermodynamically more favoured α-isomer. Consistent with direct displacement of MeSI from the
iodosulfonium ion by halide, reaction of α-thiogalactoside 1 with 1.1 equivalent of I–Cl gives a 7:1
excess of the β-chloride 13β (Table 3, entry 6). Use of a large excess of I–Cl (Table 3, entry 7) results in
epimerisation of any initially formed β-chloride 13β, so giving rise exclusively to α-chloride 13α under
the conditions reported herein (Table 3, entry 7). These observations are consistent with substantial SN2
character to interhalogen-mediated activation of ‘armed’ α-thioglycosides.
3. Conclusions
The activation of thioglycosides with iodine and interhalogens is complex. For ‘disarmed’ thioglyco-
sides, iodine alone does not cause chemistry to take place, although thioglycoside epimerisation, and
eventually glycosyl iodide formation, is evident in the presence of iodine–HMDS (which generates
TMS-I in situ). Cross-over experiments show this epimerisation process to be intramolecular. Reaction
of the same ‘disarmed’ thioglycoside with I–Br gives rise to the thermodynamic α-bromide, whereas
reaction with I–Cl gives the kinetically favoured β-chloride. For ‘disarmed’ sugars, epimerisation of the
latter is slow and the β-chloride can easily be isolated. For ‘armed’ thioglycosides, reaction with iodine
results in intermolecular thioglycoside epimerisation, as noted from cross-over experiments, but not in

590
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
glycosyl iodide formation. In contrast, reaction of the same thioglycoside with I–Br gives rise to the
thermodynamic α-bromide, whereas reaction with I–Cl gives the kinetically favoured β-chloride. The
stereochemical outcome of glycosyl halide formation in these latter reactions differs by virtue of the
substantial difference in rate of epimerisation of β-bromides (fast) and β-chlorides (slow). We note that
these glycosyl halide forming processes proceed via reactions possessing substantial SN2 character. If this
can be extrapolated to oxygen rather than halogen nucleophiles, it is conceivable that one could devise
routes from a single thioglycoside donor to either α- or β-O-glycosides simply through appropriate
choice of promoter and reaction conditions. Further studies to investigate this point are in progress.
4. Experimental
4.1. General
All reagents were used as purchased without further purification. Reaction solvents were dried by
storage over activated molecular sieves (4 Å). HMDS, DCM and DMF refer to hexamethyldisilane
dichloromethane and N,N-dimethylformamide, respectively. The abbreviation rt refers to room temper-
ature (approximately 20 to 24°C). TLC was performed with 0.2 mm Merck pre-coated silica gel 60
F254 aluminium sheets. Compounds were detected by dipping the TLC plates in an ethanolic solution of
sulfuric acid (4% v/v) and heating. Sorbsil C60 40/60 A (Sorbsil Chromatography Media) was used for
column chromatography. Hexane refers to a mixture of isomeric hexanes. Melting points (uncorrected)
were determined on a Gallenkamp melting point apparatus. Optical rotations were recorded on an Optical
Activity Ltd. AA-1000 polarimeter at room temperature (approximately 22 to 24°C) for solutions in
1
dichloromethane. H NMR spectra were recorded at 300 MHz on a Bruker AM300 spectrometer or at
500 MHz on a Varian spectrometer in CDCl3 or CD3CN. Chemical shifts are expressed relative to that of
the residual proton in the deuterated solvents (δ 7.25 and 2.00 for CDCl3 and CD3CN, respectively).
13C NMR spectra were recorded at 75.47 MHz on a Bruker AM300 spectrometer. Assignments of
resonances are based on published data. The anomeric ratios of products reported were determined from
their NMR spectra. Mass spectrometry studies were conducted on a Perspective Biosystems Voyager-DE
STR Biospectrometry workstation. All compounds referred to in the text have been reported previously,22
many are commercially available and therefore only directly relevant spectral data are reported (see Table
4).
Table 4
Selected 1H NMR data for compounds 1–5

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
591
4.2. Typical procedure for the preparation of glycosyl halides from thioglycosides
Iodine monohalide (1–4 mol. equiv.; see Tables 1–3 for details) was added to a stirred solution of
the thioglycoside (for solvents see Tables 1–3) at the desired temperature (see Tables 1–4). Stirring was
continued until the reaction was complete, the mixture was then diluted with DCM and subjected to a
standard aqueous work up (see below) to yield the desired glycosyl halide.
For monitoring by NMR, reactions were conducted in standard 5 mm NMR tubes using neat
CDCl3/CD3CN as the solvent (obtained from Cambridge Isotope Laboratories, Inc.). For controlled
experiments the reagents were mixed in the NMR tube after cooling them to the desired temperature
(ice/ice–salt/acetonitrile–dry ice bath as desired). They were then allowed to warm up to rt in the NMR
probe while the reaction was being monitored. I–Br obtained as 1 M solution in DCM was used as such.
In the case of I–Cl, either neat I–Cl dissolved in the deuterated solvent or the commercially available 1
M solution in DCM was used for these studies.
4.3. Typical procedure for the thioglycoside epimerisation
HMDS (5 mol. equiv.) was added to a solution of the thioglycoside (e.g. compound 4β) (1–2 ml
acetonitrile/100 mg of thioglycoside) while being stirred at rt. Iodine (0.5 mol. equiv.) was then added
and stirring was continued until the equilibrium concentration of the epimers was achieved (e.g. 15–20
min for 4β). The mixture was then diluted with DCM and washed successively with 10% aqueous sodium
thiosulfate solution and 10% sodium carbonate solution, dried (Na2SO4), filtered and concentrated to
dryness under reduced pressure to yield the product as a mixture of the anomers in essentially quantitative
yield. Separation of the anomers was achieved by column chromatography on silica (e.g. in the case of
compound 4β: using Et2O:hexane, 3:2 as the eluent to obtain first 4α followed by 4β in 48% and 50%
yield, respectively, both as crystals). It should be noted that HMDS concentrations used in these studies
are unoptimised.
4.4. 13C-Methyl 2,3,4,6-tetra-O-acetyl-13C-1-thio-β-D-galactopyranoside 6
Iodine (2 mg) was added at rt to a stirred suspension of 1-13C-D-galactose (90.1 mg, 0.5 mmol) in acetic
anhydride (0.5 ml). Dissolution of the sugar was complete (approx. 6 h). TLC (EtOAc:hexane, 1:1) at
this stage revealed completion of the formation of the per-O-acetylated sugar.14 The reaction mixture was
then diluted with DCM (50 ml) and was washed successively with 10% aqueous sodium thiosulfate and
10% aqueous sodium carbonate solutions, dried (Na2SO4) and concentrated to dryness under reduced
pressure followed by drying under high vacuum to yield the penta-O-acetyl galactopyranose in virtually
quantitative yield. The resulting solid was dissolved in dry DCM (10 ml) and, after cooling in an ice-bath,
HBr/HOAc (1.5 ml of a 45% w/v solution) was added. The reaction was allowed to proceed for 2 h at
rt, by which time formation of the 13C-1-labelled acetobromogalactose was shown to be complete by
TLC (EtOAc:hexane, 2:3). The reaction mixture was then diluted with DCM (25 ml) and was washed
successively with ice-cold water (3×20 ml) and cold 10% aqueous sodium carbonate solution (2×20 ml)
in a separating funnel. The organic layer was then dried (Na2SO4), concentrated to dryness under reduced
pressure and finally dried under high vacuum to yield the glycosyl bromide in virtually quantitative
yield.1H NMR (CDCl3) δ: 6.95, 6.33, 2 d, 1H, 3.9 Hz and 186.5 Hz, H-1; 5.46, dd, 1H, 1.1 Hz and 3.3
Hz, H-4; 5.33, dddd, 1H, H-3; 4.98, dddd, 1H, H-2; 4.43, m, 1H, H-5; 4.09, m, 2H, H-6a and H-6b; 2.09,
2.06, 2.00 and 1.95, 4 s, 12H, 4×OCOCH3.

592
K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
The acetobromogalactose obtained above was then dissolved in a mixture of dry acetone (2 ml) and
dry DCM (2 ml) and was stirred at rt in the presence of potassium thioacetate (286 mg, 2.5 mmol)
for 24 h. TLC (EtOAc:hexane, 2:3) showed completion of the reaction. The reaction mixture was then
diluted with DCM (40 ml) and was washed with water (2×20 ml) in a separating funnel, dried (Na2SO4),
concentrated to a syrup and purified by column chromatography (silica gel, 25 ml; eluent, EtOAc:hexane,
1
2:3) to obtain 2,3,4,6-tetra-O-acetyl-13C-1-thioacetyl-β-D-galactopyranose (166 mg, 81.5%). H NMR
(CDCl3) δ: 5.46, 4.92, 2 d, 1H, 10.3 Hz and 160.7 Hz, H-1; 5.39, broad d, 1H, 3.3 Hz, H-4; 5.24, ddd,
1H, H-2; 5.05, dddd, 1H, H-3; 4.02, m, 3H, H-5, H-6a and H-6b; 2.33, near d, 0.7 Hz, 3H, SCOCH3;
2.08, 1.97, 1.96 and 1.92, 4 s, 12H, 4×OCOCH3.
To a stirred solution of the thioacetate thus obtained in dry MeOH (6 ml) cooled to −40°C was added
a solution (1 ml) of sodium methoxide (equiv. to 8.9 mg of sodium metal, 0.95 mol. equiv.) and stirring
was continued for 40 min at −40°C. Methanol was then removed by evaporation under reduced pressure
at ice-bath temperature and the residue was subsequently dried under high vacuum to yield an amorphous
mass. It was then taken up in dry DMF (2 ml) and was treated with 13CH3I (3 mol. equiv.) for 18 h at rt.
The reaction mixture was then concentrated to a syrup under reduced pressure and purified by column
chromatography (silica gel, 30 ml; eluent, EtOAc:hexane, 2:3) to yield 6 (139 mg, 90%). 13C NMR
(CDCl3) δ: 83.54, C-1; 11.45, SCH3.
Epimerisation studies with β-thioglycoside 6 gave the corresponding α-thioglycoside: 13C NMR
(CDCl3) δ: 83.20, C-1; 11.93, SCH3.
Acknowledgements
We thank the Wellcome Trust (grants: 042472 and 040331), the BBSRC/EPSRC, the Zeneca Strategic
Research Fund and the University of St Andrews for financial support. We are grateful to Professor
R. J. Ferrier for his interest in these studies and to Professor John Fothergill for assistance with mass
spectrometry experiments.
References
1. Iodine and its Interhalogen Compounds: Versatile Reagents in Carbohydrate Chemistry XI. For part X, see: Rutherford, T.
J.; Kartha, K. P. R.; Readman, S. K.; Cura, P.; Field, R. A. Tetrahedron Lett. 1999, 40, 2025–2028.
2. Presented in part at the XVIII International Carbohydrate Symposium, Milan, Italy, 1996 (abstract BO 57 and BP 271) and
at the XIX International Carbohydrate Symposium, San Diego, USA, 1998 (abstract BP 156).
3. Paulsen, H. In Modern Methods in Carbohydrate Synthesis; Khan, S. H.; O’Neill, R. A., Eds.; Harwood Academic
Publishers: Amsterdam, 1996; pp. 1–20.
4. For representative examples of studies on the mechanism of glycosylation reactions see the following articles and references
cited therein: Wulff, G.; Rohle, G. Angew. Chem., Int. Ed. Engl. 1974, 13, 157–170; Igarashi, K. Adv. Carbohydr. Chem.
Biochem. 1977, 34, 243–283; Paulsen, H. Angew. Chem., Int. Ed. Engl. 1982, 21 155–224; Schmidt, R. R. Angew. Chem.,
Int. Ed. Engl. 1986, 25, 212–235; Lemieux, R. U. Explorations with Sugars: How Sweet it Was; American Chemical Society:
Washington, 1990; Fraser-Reid, B.; Wu, Z. F.; Andrews, C. W.; Skowronski, E.; Bowen, J. P. J. Am. Chem. Soc. 1991, 113,
1434–1435; Gervay, J. In Organic Synthesis: Theory and Applicatons; JAI Press: London, 1998; Vol. 4, pp. 121–153; Crich,
D.; Cai, W. L. J. Org. Chem. 1999, 64, 4926–4930.
5. For review articles on the preparation of glycosyl halides and thioglycosides and their use in glycosylation reactions, see:
Modern Methods in Carbohydrate Synthesis; Khan, S. H.; O’Neill, R. A.; Eds.; Harwood Academic Publishers: Chichester,
1996; Garegg, P. J. Adv. Carbohydr. Chem. Biochem. 1997, 52, 179.
6. Kartha, K. P. R.; Aloui. M.; Field R. A. Tetrahedron Lett. 1996, 37, 5175–5178.
7. Kartha, K. P. R.; Aloui, M.; Field, R. A. Tetrahedron Lett. 1996, 37, 8807–8810.
8. Kartha, K. P. R.; Field, R. A. Tetrahedron Lett. 1997, 38, 8233–8236.

K. P. R. Kartha et al. / Tetrahedron: Asymmetry 11 (2000) 581–593
593
9. Kartha, K. P. R.; Field, R. A., unpublished observations.
10. Veeneman, G. PhD thesis, University of Leiden, The Netherlands, 1996.
11. Boons, G.-J.; Stauch, T. Synlett 1996, 906.
12. Gervay, J.; Nguyen, T. N.; Hadd, M. J. Carbohydr. Res. 1997, 300, 119–125.
13. Kartha, K. P.R.; Field, R. A. Carbohydr. Lett. 1998, 3, 179–186 and Ref. 4 cited therein.
14. Kartha, K. P. R.; Field, R. A. Tetrahedron 1997, 53, 11759–11766 (for acetylation) and Kartha, K. P. R.; Jennings, H. J. J.
Carbohydr. Chem. 1990, 9, 777–781 (for glycosyl bromide preparation).
15. Murase, T.; Kameyama, A.; Kartha, K. P. R.; Ishida, H.; Kiso, M.; Hasegawa, A. J. Carbohydr. Chem. 1989, 8, 265–283.
16. Glaudemans, C. P. J.; Kovac, P.; Nashed, E. M.; Padlan, E. A.; Arepalli, S. R. In Synthetic Oligosaccharides; ACS
Symposium Series 560; Kovac, P., Ed.; American Chemical Society: Washington, 1994; Chapter 10, pp. 157–183.
17. Kovac, P.; Lerner, L. Carbohydr. Res. 1988, 184, 87–112.
18. Fugedi, P.; Garegg, P. J.; Lonn, H.; Norberg, T. Glycoconjugate J. 1987, 4, 97–108.
19. Weygand, F.; Ziemann, H. Ann. 1962, 657, 179–198.
20. Lemieux, R. U.; Hayami, J. Can. J. Chem. 1965, 43, 2162–2173.
21. Crich, D.; Dai, Z. M.; Gastaldi, S. J. Org. Chem. 1999, 64, 5224–5229.
22. Refs. 5, 6 and 14 above and the references cited within; Kartha, K. P. R.; Field, R. A. J. Carbohydr. Chem. 1998, 17,
693–702 and the references cited within; Lemieux, R. U. In Methods in Carbohydrate Chem. 2; Whistler, R. L.; Wolfrom,
M. L., Eds.; Academic Press: New York, 1963, pp. 223–225; Pacsu, E.; Janson, J.; Lindberg, B. In Methods in Carbohydrate
Chem. 2; Whistler, R. L.; Wolfrom, M. L., Eds.; Academic Press: New York, 1963; pp. 376–385; Paulsen, H. In Methods in
Carbohydrate Chem. 6; Whistler, R. L.; Wolfrom, M. L., Eds.; Academic Press: New York, 1972; pp. 142–149.