Mechanistic studies on a sulfoxide transfer reaction mediated by diphenyl sulfoxide/triflic anhydride

Article abstract of DOI:10.1002/chem.201102861

Sulfoxides are frequently used in organic synthesis as chiral auxiliaries and reagents to mediate a wide variety of chemical transformations. For example, diphenyl sulfoxide and triflic anhydride can be used to activate a wide range of glycosyl donors inc

Full text of DOI:10.1002/chem.201102861

FULL PAPER
DOI: 10.1002/chem.201102861
Mechanistic Studies on a Sulfoxide Transfer Reaction Mediated by Diphenyl
Sulfoxide/Triflic Anhydride
Martin A. Fascione,[a] Sophie J. Adshead,[a] Pintu K. Mandal,[a] Colin A. Kilner,[a]
Andrew G. Leach,[b] and W. Bruce Turnbull*[a]
In memory of Professor David Y. Gin
Abstract: Sulfoxides are frequently
used in organic synthesis as chiral aux-
iliaries and reagents to mediate a wide
variety of chemical transformations.
For example, diphenyl sulfoxide and
triflic anhydride can be used to activate
a wide range of glycosyl donors includ-
ing hemiacetals, glycals and thioglyco-
sides. In this way, an alcohol, enol or
sulfide is converted into a good leaving
group for subsequent reaction with an
acceptor alcohol. However, reaction of
diphenyl sulfoxide and triflic anhydride
with oxathiane-based thioglycosides,
and other oxathianes, leads to a differ-
ent process in which the thioglycoside
is oxidised to a sulfoxide. This unex-
pected oxidation reaction is very ste-
reoselective and proceeds under anhy-
drous conditions in which the diphenyl
sulfoxide acts both as oxidant and as
the source of the oxygen atom. Isotopic
labelling experiments support a reac-
tion mechanism that involves the for-
mation of oxodisulfonium (S-O-S) di-
cation intermediates. These intermedi-
ates undergo oxygen-exchange reac-
tions with other sulfoxides and also
allow interconversion of axial and
equatorial sulfoxides in oxathiane
rings. The reversibility of the oxygen-
exchange reaction suggests that the
stereochemical outcome of the oxida-
tion reaction may be under thermody-
namic control, which potentially pres-
ents a novel strategy for the stereose-
lective synthesis of sulfoxides.
Keywords: cations · oxidation · re-
action mechanisms · sulfoxides · tri-
fluoromethanesulfonic anhydride
Introduction
tion of reagents;[6,7] glycals can be oxidised to glycosyl
epoxides for reaction with alcohols;[8] and this method is
also often used to activate thioglycosides.[4f,7d,9]
Chemical transformations involving sulfoxides often involve
reaction with a strong electrophile such as triflic anhydride
(Tf2O) to activate the sulfoxide for reaction with nucleo-
philes.[1] For example, dimethylsulfide ditriflate will react
with alkenes, alkynes and arenes to provide vinyl and aryl
sulfides and sulfonium ions.[2] Alternatively, the same acti-
vated sulfoxide reagent can oxidise simple alcohols to ke-
tones.[3] Activated sulfoxides have also had a significant
impact on the synthesis of oligosaccharides,[4] ever since the
introduction of sulfoxide glycosyl donors by Kahne in
1989.[5] More recently, there has been a growing interest in
the use of diphenyl sulfoxide and triflic anhydride as a
system for activating glycosyl donors.[6] For example, glyco-
syl hemiacetals can be activated directly using this combina-
We sought to apply the diphenyl sulfoxide/Tf2O method
for activating oxathiane glycosyl donors 1 and 3 (Scheme 1)
that we have recently developed for the preparation of 1,2-
[a] Dr. M. A. Fascione, S. J. Adshead, Dr. P. K. Mandal, C. A. Kilner,
Dr. W. B. Turnbull
School of Chemistry, University of Leeds
Leeds, LS2 9 JT (UK)
Fax : (+44)1133436565
E-mail: w.b.turnbull@leeds.ac.uk
[b] Dr. A. G. Leach
AstraZeneca, Alderley Park
Macclesfield, Cheshire, SK10 4TF (UK)
Scheme 1. Activation of oxathiane glycosyl donors with diphenyl sulfox-
ide and triflic anhydride leading to oxidation of the thioglycoside.
DTBMP=2,6-di-tert-butyl-4-methylpyridine.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102861.
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2987

cis-a-glycosides.[10] However, no glycoside products were
formed when either donor was subjected to standard reac-
tion conditions (diphenyl sulfoxide (2.8 equiv) and Tf2O
(1.4 equiv))[9a,c] in the presence of an acceptor alcohol. In-
stead, both oxathianes were oxidised to the corresponding
sulfoxides 2 and 4 in high yield and excellent stereoselectiv-
ity.[11,12] Subsequent experiments on a non-carbohydrate oxa-
thiane showed that the oxidation reaction was not exclusive
to glycosides 1 and 3 (see below). Density functional theory
calculations (B3LYP/6-31G* and M06/6-31G*)[13] indicated
that in each case the major isomer was also the thermody-
namically preferred product (see the Supporting Informa-
tion). Very similar results were obtained when the oxidation
reactions were performed in the absence of an acceptor al-
cohol, thus demonstrating that the alcohols were superfluous
to the reactions. The lack of glycoside products in these re-
actions is consistent with the unusually high stability of sul-
fonium ions derived from glycosyl oxathiane com-
Results and Discussion
Reaction of diphenyl sulfoxide with triflic anhydride: Sever-
al methods for the preparation of 18O-enriched diphenyl
sulfoxide have been reported in the literature.[16] We chose
to activate the sulfoxide 5 with Tf2O (1 equiv), and then
quench the reaction with 18O-enriched water (97% 18O-
AHCTUNGTREGiNNNU ncorporation; Scheme 2). The diphenyl sulfoxide isolated
Scheme 2. Preparation of [18O]-diphenyl sulfoxide.
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
Oxygen-transfer reactions between sulfoxides and sul-
fides[14] (or between selenoxides and sulfides[15]) are well
documented in the literature. While spontaneous oxygen
transfer can occasionally occur at high temperature,[14c] in
general, the sulfoxide is usually activated by a strong acid
under aqueous conditions.[14a,b,f] A chlorosulfonium ion inter-
mediate has been implicated when HCl is used as the cata-
lyst.[14d,e,h] However, in the case of less nucleophilic anions,
for example, perchlorate or sulfate, it has been postulated
that the reaction may involve direct attack by the sulfide on
a protonated sulfoxide intermediate.[14f] Similar mechanisms
have also been proposed for oxygen transfer promoted by
trifluoroacetic anhydride.[14g]
from this reaction was found to have an 18O/16O ratio of
87:13 (see the Supporting Information). After re-subjecting
this product mixture to the same reaction conditions, the
18O incorporation increased to 92%. The efficiency of these
labelling reactions was found to be reproducible (Æ2%).
Incomplete transfer of the 18O-label indicated that either
a portion of diphenyl sulfoxide failed to react with Tf2O on
the timescale of the reaction, or the reaction was not com-
pletely anhydrous, or an activated sulfoxide intermediate
was hydrolysed on work-up with retention of the 16O-
atom.[17] IR spectroscopy was used to monitor the progress
of the reaction in situ. Analysis of the spectra by using
ConcIRTꢁ (Mettler Toledo) revealed that diphenyl sulfox-
ide was consumed fully within a minute of adding one equiv
Tf2O at À328C, and that the reaction mixture comprised at
least three species at equilibrium (see the Supporting Infor-
mation).
It is usually assumed that hydrolysis of a sulfoxonium in-
termediate would install the oxygen atom in the sulfoxide
product.[14d–h] However, a low temperature H NMR study of
1
the reaction involving oxathiane 1 revealed that the sulfox-
ide products 2R/2S were formed under anhydrous condi-
tions, and within minutes of the addition of Tf2O. Therefore,
in the diphenyl sulfoxide/Tf2O reaction reported here, it
would appear that the oxygen atom must originate from one
of these two activating agents. We reasoned that an isotope
labelling study could cast further light on the mechanism of
oxygen transfer, and possibly even provide insight to the
early steps of the glycosylation reactions promoted by di-
phenyl sulfoxide/Tf2O. Herein we will address three impor-
tant questions: 1) From where does the oxygen atom in the
sulfoxide product originate? 2) Which species acts as the ox-
idant in these reactions? 3) What is the basis for the high
stereoselectivity of these reactions? But first, it will be in-
structive to consider the first step of the oxidation process in
some detail: that is, the reaction between diphenyl sulfoxide
and Tf2O.
We reasoned that identification of these intermediates
would be helpful if we were to understand the oxidation re-
action in Scheme 1. The triflyloxy sulfonium ion
(Scheme 3a), which is presumably formed first in the re-
AHCTUNGTREGaNNUN ction mixture, could combine with its triflate anion to give
6
sulfurane 9, or it could react with another molecule of di-
phenyl sulfoxide 5 to yield dication 8, or the analogous
mono- or bis-sulfurane compounds 7 and 10. Oxo-bridged
disulfonium salt 8 has been proposed previously as a re-
AHCTUNGTREGaNNNU ction intermediate when an excess of diphenyl sulfoxide is
mixed with Tf2O.[8c] However, this intermediate could also
form when equimolar amounts of the sulfoxide and Tf2O
are mixed, if the Tf2O were added dropwise, and/or inter-
mediate 6 were more reactive as an electrophile than Tf2O;
for example, the phosphorus analogue, bis(triphenyl)-oxodi-
phosphonium
triflate
12
(Hendricksonꢂs
reagent,
Scheme 3b),[18] can be prepared by mixing triphenylphos-
phine oxide 11 and Tf2O in equimolar amounts.[18a] Kelly
and co-workers used X-ray crystallography studies to prove
that Hendricksonꢂs reagent is indeed a P-O-P dication.[18a]
2988
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2987 – 2997

Sulfoxide Transfer Mediated by a Diphenyl Sulfoxide-Triflic Anhydride
FULL PAPER
Scheme 3. a) Possible intermediates arising from the reaction of diphenyl
sulfoxide and triflic anhydride. b) Preparation of Hendricksonꢂs re-
agent.[18]
Density functional theory calculations (B3LYP/6-31G*
and M06/6-31G*)[13] support the hypothesis that all of the in-
termediates proposed in Scheme 3 are feasible structures
with the exception of bis-sulfurane 10 which exhibited no
barrier to dissociation in the gas phase. Monocation 6 and
À
dication 8 displayed very similar S O bond lengths and
À À
S O S bond angles which place the sulfur atoms in dication
8 only 2.94 ꢃ apart (Figure 1). Dications with the structure
[R2S-SR2]2+ are well documented in the literature[19] (see
below) and it would appear that there may be a similar
bonding interaction between the two positively charged
sulfur atoms in dication 8.[20]
Figure 2. All 1H NMR spectra [(a)–(d) 500 MHz; CD2Cl2], and 19F NMR
spectra [(e)–(h) 470 MHz; CD2Cl2] were recorded at À608C, except spec-
trum (d) which was acquired at À308C. In each case, several spectra
were acquired to check that the system had come to equilibrium. The
shoulder at À78 ppm in (h) is a shimming artefact.
ure 2b) and 2:1 (Figure 2c) mixtures were dominated by dif-
ferent species (presumably 6/9 and 7/8, respectively), the 1:1
mixture also displayed some minor signals (e.g., at 7.7 ppm)
that could be consistent with the partial formation of oxodi-
sulfonium ion 8. Addition of water to the NMR samples
converted each mixture to a single species (Figure 2d) that
differed from the original spectrum of diphenyl sulfoxide
(Figure 2a). We attribute these signals to protonation of the
sulfoxide by TfOH.[17]
Figure 1. Significant bond lengths and angles derived from density func-
tional theory calculations (B3LYP/6-31G*) on putative reaction inter-
mediates. Plan and elevation views of oxodisulfonium ion 8.
1H NMR spectroscopy was used to further characterise
1:1 and 2:1 mixtures of diphenyl sulfoxide 5 and freshly dis-
tilled Tf2O in CD2Cl2 at À608C (Figure 2). In each case, the
1H NMR spectrum revealed that all of the diphenyl sulfox-
ide had been consumed, to produce one or more products
displaying broadened NMR signals. While the 1:1 (Fig-
The reaction mixtures were also analysed by 19F NMR
spectroscopy. The 2:1 mixture showed only a single peak in
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2989

W. B. Turnbull et al.
its 19F NMR spectrum at À79.4 ppm (Figure 2 g), with a fre-
quency typical for a triflate anion, for example, sulfonium
triflate 13 (Figure 2 h).[21] In contrast, the 1:1 mixture dis-
played two signals at À72.0 ppm and À79.1 ppm (Figure 2 f).
The signal at À72.0 ppm corresponds to unreacted Tf2O
(Figure 2e), and is consistent with partial formation of the
dimeric sulfonium ion 8. While one might expect to see sev-
eral other fluorine signals corresponding to triflyl/triflate
groups in species 6–9, the sole broadened signal at
À79.1 ppm would suggest that all triflyl groups are in rapid-
intermediate exchange on the NMR timescale.[22] In summa-
ry, the 19F NMR spectra indicate that the triflyloxy group in
6 is in fast to intermediate exchange with the triflate counter
ion, presumably through an equilibrium with sulfurane 9.
However, any equilibrium between oxygen-bridged species
7/8 lies in favour of dication 8 and free triflate anions.
was allowed to react at À608C for 10 min before addition of
unlabelled water and neutralisation with triethylamine
(Scheme 5). In theory, when sulfoxides ACHTNUTRGENNUG HCATUNGTRENNGUN
[18O]-5, [16O]-15 and
As the NMR spectra (Figure 2) demonstrate that all of
the diphenyl sulfoxide was consumed under the conditions
of the labelling experiment (Scheme 2), the sub-optimal in-
ACTHNUTRGNEUNG
corporation of the 18O-label in [18O]-5 must arise from
either less than perfect anhydrous conditions or partial re-
tention of a 16O-atom present in one of the activated inter-
mediates (Scheme 4). In the case of dication 8, attack by
Scheme 5. Cross-over experiments in which 92%-enriched diphenyl sulf-
oxide ACHTNUGRTNEUNG
[18O]-5, unlabelled 15 and Tf2O are mixed according to the given
ratios.
Tf2O are mixed in a 1:1:1 ratio, there is an equal probability
that the Tf2O will react with diphenyl sulfoxide or ditolyl
sulfoxide to give intermediates 6a. The NMR experiments
in Figure 2 indicate that 6a would then react with the re-
maining 1:1 mixture of sulfoxides 5 and 15 to produce oxo-
disulfonium ions 8a. Half of the bridging oxygen atoms in
Scheme 4. Possible mechanisms for hydrolysis of activated sulfoxide inter-
mediates with retention of 16O. Reactions are depicted as SN2 processes
for convenience but may involve addition–elimination mechanisms via
sulfurane intermediates.
8a should thus be derived from ACTHNUTRGENNUG ACHTUNGTRENNUNG
[18O]-5, and half from [16O]-
[18O]-water at either sulfonium centre would result in reten-
tion of a 16O-atom, while for sulfonium ion 6, the labelled
water would have to attack the S(VI) centre to avoid intro-
ducing the label into the diphenyl sulfoxide product; [18O]-
TfOH 14 would also be produced as a by-product of this re-
15. Therefore, the 18O-enrichment in oxodisulfonium ions 8a
should be half that of the original [18O]-diphenyl sulfoxide
(i.e., 92%Ä2=46%). Upon aqueous work-up, intermedi-
ates 8a will be hydrolysed to reform sulfoxides 5 and 15.
Half of the molecules should retain the bridging oxygen
atom from 8a, while the other half would receive an oxygen
action. Gin has demonstrated that glycosyl hemiACHTNUTRGNEaNUG cetals react
with sulfonium ions such as 6 at the S(IV) centre only;[7a,23]
therefore, the reaction in Scheme 4b is unlikely to occur.
A cross-over experiment was performed to gain further
evidence for the formation of oxodisulfonium ion 8. A mix-
atom from H2O. Therefore, for a 1:1:1 mixture of ACHTUNGTRENNUNG
[18O]-5,
AHCTUNGTRENNUNG
[16O]-15 and Tf2O undergoing oxodisulfonium-mediated
18O-exchange, the theoretical 18O-enrichment in the product
sulfoxides 5 and 15 should be 23% (i.e., 46%Ä2).
ture of labelled diphenyl sulfoxide
N
In the case of the 1:1:2 mixture of [16O]-5, [18O]-15 and
Tf2O, one might expect species 6a to dominate the reaction
belled ditolyl sulfoxide
AHCTUNGTRENNUNG
2990
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2987 – 2997

Sulfoxide Transfer Mediated by a Diphenyl Sulfoxide-Triflic Anhydride
FULL PAPER
mixture. Transfer of 18O label between diphenylsulfoxide
and ditolyl sulfoxide could still occur through exchange of
labelled triflate ions. However, the label would only be re-
tained in the sulfoxide products 5 and 15 after aqueous
work-up if the triflyl groups could be removed as shown in
Scheme 4b. This process has already been discounted as un-
likely to occur,[7a] however, if it did happen, the 18O-incorpo-
ration in the product sulfoxides 5 and 15 would be very low.
Half of the Tf2O would react initially with
other half with
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
6a that are partially labelled (i.e., 92%Ä2=46% 18O) and
unlabelled triflate anions. Triflate exchange would scramble
all 12 oxygen atoms in the triflyloxy and triflate groups fur-
ther reducing the labelling in ions 6a.[24] Therefore, for a
1:1:2 mixture of ACHTUNGTRENNUNG ACHUTNGTRENNUGN
[18O]-5, [16O]-15 and Tf2O undergoing tri-
flate-mediated 18O-exchange, the theoretical 18O-enrichment
in product sulfoxides 5 and 15 should be less than 8% (i.e.,
92% 18OÄ12 oxygen atoms=7.7%).
Both experiments led to an even distribution of the 18O
label between diphenyl sulfoxide and ditolyl sulfoxide. Fur-
thermore, the incorporation of 18O label in the reaction
products was found to be higher than expected (32–33%
18O for 1:1:1 mixture; 26–28% for 1:1:2 mixture). In other
words, cross-over of the 18O-label occurred without any ap-
parent scrambling of oxygen atoms that would be associated
with a triflate-exchange mechanism. These results are only
consistent with a reversible 18O-exchange process involving
diaryl sulfoxides 5/15 and oxygen-bridged species 8a. The
formation of an oxodisulfonium ion intermediate 8a upon
mixing diaryl sulfoxide and Tf2O in equimolar amounts
would be in accord with formation of Hendricksonꢂs reagent
under similar conditions (Scheme 3b).[18a] Furthermore, the
enhanced label incorporation in the 1:1:1 reaction mixture
could be possible if intermediate 8a were to be quenched by
a side reaction prior to aqueous work-up (see below), thus
Figure 3. Mass spectra for a) [18O]-diphenyl sulfoxide reactant, b) oxa-
thaine 2-R and c) oxathiane 2-S products of the oxidation reaction.
Two things can be concluded from this experiment:
1) The oxygen atom arises solely from the diphenyl sulfox-
ide, and not from either the Tf2O or water on aqueous
work-up. 2) The reaction reaches completion before aqueous
work-up, that is, there can be no activated sulfoxide species
remaining at the time of work-up, or the percentage of 18O-
label incorporation would be reduced, at very least in the
case of the recovered diphenyl sulfoxide.
An identical labelling experiment using oxathiane ether 3
also afforded sulfoxide product 4-S with a very high level of
18O-incorporation (83%). On this occasion, it was not possi-
ble to isolate the equatorial sulfoxide isomer 4-R as it was
formed in very low concentration. It is not clear if the small
reduction in the isotopic enrichment (87% to 83%) is signif-
icant in this case. Therefore, we presume that both oxidation
reactions in Scheme 1 occur by the same mechanism.
ACTHNUTRGNEUNG
avoiding further dilution of the 18O label with [16O]-H2O.
Several conclusions can be drawn from these preliminary
studies on diphenyl sulfoxide and Tf2O: (1) triflyloxy sulfo-
nium ion 6 can react rapidly at its sulfur(IV) centre with di-
phenyl sulfoxide to generate oxodisulfonium ion 8; (2) the
oxodisulfonium ion 8 dominates the reaction mixture when
diphenyl sulfoxide and Tf2O are mixed in a ratio of 2:1; (3)
oxodisulfonium ions can undergo an exchange re
other sulfoxides.
ACHTUNGTRENUNaNG ction with
Identification of the oxidizing agent: Having determined the
origin of the oxygen atom in the product, the next question
to address was the nature of the oxidising agent; that is,
which species becomes reduced during the reaction. Mix-
tures of Tf2O and either dimethyl sulfoxide 16 or diphenyl
sulfoxide 5, have been reported to oxidise alcohols
(Scheme 6a),[25] and glycals,[8c] respectively. However, prece-
dent for oxidative transformations using Tf2O alone also
exists; for example, oxidation of aliphatic sulfides by Tf2O
can provide access to sulfoxides, unless an alcohol is added
to the mixture, in which case aldehydes or ketones will
result (Scheme 6b).[3]
Origin of the oxygen atom in the oxathiane-S-oxide product:
With [18O]-diphenyl sulfoxide in hand, we sought to test the
hypothesis that oxidation of the oxathiane involved transfer
of oxygen from diphenyl sulfoxide. [18O]-Diphenyl sulfoxide
5 (87% 18O) and Tf2O were used to oxidise oxathiane ketal
1 as described above. The resulting axial and equatorial sulf-
oxides were separated by silica gel chromatography for mass
spectrometric analysis. Both isomers of the oxathiane-S-
oxides (2-R and 2-S) were found to have 87% 18O-enrich-
ment (Figure 3). The excess diphenyl sulfoxide 5 that was
recovered from the reaction mixture also had the same 18O-
content as the starting material.
In cases where Tf2O acts as an oxidant, one sulfonate
group will become reduced to the sulfinate 20 (Scheme 6b).
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2991

W. B. Turnbull et al.
Scheme 6. Mechanisms for oxidation reactions promoted by Tf2O as a) a
sulfoxide activating agent and b) an oxidant. In reaction a)two equiv
TfOH 14 are produced, while reaction b) produces 1 equiv TfOH 14 and
1 equiv of sulfinic acid 20.
On the other hand, if the role of Tf2O is solely to activate a
sulfoxide, for example, 16 (Scheme 6a), then the corre-
sponding sulfide 19 should be formed as the reduction prod-
uct. Although we were never able to observe or isolate di-
phenyl sulfide from the sulfoxidation reaction, triaryl sulfo-
nium salt 13 was consistently observed in the mass spectra
of the crude reaction mixtures (Figure 4). The quantitative
formation of aryl sulfonium salt 13 was confirmed by
HPLC-mass spectrometric comparison of the crude product
mixture with authentic samples of sulfonium salt 13 of
known concentration (see the Supporting Information).
Sulfonium salt 13 has been synthesised previously from di-
phenyl sulfide and diphenyl sulfoxide activated with Tf2O
and other electrophiles.[26] Therefore, it is reasonable to con-
clude that an activated sulfoxide species is responsible for
oxidising the oxathiane compounds, while concomitantly
producing diphenyl sulfide as a by-product. Any excess acti-
vated sulfoxide is then sequestered by the diphenyl sulfide
to yield sulfonium ion 13.[27] This side-reaction has two con-
sequences: 1) The oxidation reaction is effectively quenched
by diphenyl sulfide prior to aqueous work-up (in agreement
with the results of 18O-labelling experiments). 2) The oxida-
tion reaction cannot become catalytic in Tf2O.
Figure 4. a) LC-MS ion chromatograms for sulfonium triflate 13 and the
crude product mixture from
a typical oxidation reaction in which
DTBMP is used as base forming pyridinium ion 21 as a by-product. Mass
spectra for b) ion 13 and c) the component of the reaction mixture elut-
ing between 1.6 and 1.7 min. In each case expansions of the major ions
are inset in b) and c).
pothesis, sulfoxide 2-R was treated first with Tf2O
(1.1 equiv), and then diphenyl sulfide (2 equiv). The major
product isolated from the reaction mixture was oxathiane 1
(62%). No evidence could be found for the formation of
sulfonium ion 25; however, triaryl sulfonium ion 13 was ob-
served in the mass spectrum of the crude product mixture.
This experiment demonstrated that even if triflyloxy sulfoni-
um ion 22 is formed during the oxidation reaction, it would
react preferentially with the sulfur atom of diphenyl sulfide,
rather than by electrophilic aromatic substitution. Further-
more, reduction of the oxathiane-S-oxide 2-R to give oxa-
thiane 1 suggests that the oxidation process is potentially re-
versible under the conditions of the reaction.
Reversibility of the oxidation reaction: We have reported
previously that oxathiane-S-oxide glycosyl donors can be
converted to aryl sulfonium ions (e.g., compound 23,
Scheme 7) by treatment with Tf2O in the presence of tri-
We were thus curious to know if the oxygen-transfer pro-
cess could also be reversible. Therefore equatorial sulfoxide
2-R was re-exposed to the conditions of the oxidation re-
AHCTUNGTRENNUNG
action (Scheme 8a). [18O]-Diphenyl sulfoxide (2.8 equiv)
A
was pre-activated with Tf2O (1.4 equiv) before adding sulf-
oxide 2-R. After 40 min at À608C, the reaction was
quenched by addition of diphenyl sulfide. The equatorial
sulfoxide isomer 2-R dominated the product mixture and
was recovered in 89% yield. Electrospray mass spectrome-
ion 25, derived from diphenyl sulfide 24, was not observed
as a by-product of the oxidation reaction. One possible ex-
planation would be that the triflyloxy sulfonium ion 22 is
not formed during the oxidation mechanism. To test this hy-
2992
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2987 – 2997

Sulfoxide Transfer Mediated by a Diphenyl Sulfoxide-Triflic Anhydride
FULL PAPER
try confirmed that the 18O-label from the diphenyl sulfoxide
had been transferred to oxathiane S-oxide 2-R (18O/16O=
48:52). Excess diphenyl sulfoxide recovered from the reac-
tion mixture (1.34 equiv) had a similar 18O-enrichment as 2-
R (18O/16O=44:56).[28] A very similar result was obtained for
oxathiane-S-oxide 4-S (Scheme 8b), except this time the
axial sulfoxide was the dominant product. Therefore both of
these examples demonstrate that oxygen transfer is a rever-
sible process that can occur with retention of configuration.
As pre-activation of labelled diphenyl sulfoxide results in
scrambling of the label with the oxathiane-S-oxide oxygen
atom, the reaction was also performed by pre-activation of
oxathiane-S-oxide 2-R with Tf2O (1.1 equiv; Scheme 8c).
Once again, NMR spectroscopy of the crude product mix-
ture showed that the equatorial sulfoxide was the dominant
product. Furthermore, by reversing the order of addition of
sulfoxides 2-R and ACTHNUTRGNEUNG
[18O]-5, 18O-enrichment close to the the-
oretical maximum was achieved for the equatorial isomer.
Incorporation of 18O with such high efficiency can only arise
through the formation of oxygen-bridged disulfonium ion 26
(Scheme 8c) as any other mechanism (e.g., 18O transfer via
triflate exchange) would result in significant dilution of the
label.[24]
Retention of configuration at the sulfoxide centre during
the 18O-crossover experiment requires a mechanism that can
allow inter-conversion of axial and equatorial intermediates
(Scheme 9). One can imagine two possible mechanisms for
such a process: an intermolecular mechanism, in which di-
phenyl sulfoxide 5 acts as a nucleophile to attack an activat-
ed oxodisulfonium ion 27; and an intramolecular mecha-
nism, in which the anomeric bond is broken to generate an
oxacarbenium ion intermediate 28 which would allow re-
Scheme 7. Reaction of oxathiane-S-oxide 2-R with Tf2O and electron-
rich aromatic compounds.
Scheme 9. a) Intermolecular and b) intramolecular mechanisms for the
inter-conversion of axial and equatorial isomers of oxodisulfonium ion
27-R/S. The intermolecular mechanism is depicted as an SN2 process for
simplicity, although it is possible it could involve addition–elimination
steps via a sulfurane intermediate.
ACTHNUTRGNEUNG
Scheme 8. Transfer of 18O from [18O]-5 to oxathiane-S-oxides 2 and 4:
a) and b) pre-activation of diphenyl sulfoxide (2.8 equiv) with Tf2O
(1.4 equiv). c) Pre-activation of oxathiane-S-oxide 2 with Tf2O (1.1 equiv)
before addition of [18O]-diphenyl sulfoxide (2.2 equiv).
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2993

W. B. Turnbull et al.
A
torial sulfoxides can be inter-converted under the conditions
of the oxidation reaction, and 2) the major 18O-labelled
product of the cross-over experiments is the thermodynami-
cally favoured isomer, it is possible that the high stereoselec-
tivity of the oxidation process could result from the reaction
occurring under thermodynamic control. Indeed, in each
case, re-equilibration of the major isomer results in a prod-
uct mixture that is very similar to that for the corresponding
oxidation reaction. On the other hand, the oxidation reac-
tion is effectively self-quenching; the diphenyl sulfide pro-
duced during the reaction sequesters activated sulfoxide in-
termediates to form sulfonium ion 13. Therefore, equilibra-
tion of sulfoxides and/or reaction intermediates, for exam-
ple, oxodisulfonium ion 27, would have to compete with re-
action quenching by diphenylsulfide under kinetic control.
While latter mechanism cannot be excluded for glycosyl
sulfoxides, we sought further evidence in support of the in-
termolecular mechanism.
A simple racemic oxathiane compound rac-32 lacking the
sugar ring was synthesised (Scheme 10). Alkylation of
b-mercaptoethanol 29 with bromoacetophenone 30 gave
Possible mechanistic pathways: The aforementioned experi-
ments provide evidence for a number of steps that must
occur during the oxidation reaction: the first committed step
in the mechanism must be the reaction of the oxathiane
sulfur atom with an activated diphenyl sulfoxide species;[29]
a diphenyl sulfoxide oxygen atom must become covalently
bound to the oxathiane sulfur atom; diphenyl sulfide must
be produced during the reaction, and then react with some
activated diphenyl sulfoxide species to produce the triaryl
sulfonium salt by-product.
Scheme 10. Synthesis of racemic oxathiane-S-oxide rac-33a and transfer
ACHTUNGTRENNUNG
of 18O from [18O]-5 with retention and inversion of configuration.
Several mechanistic pathways could be consistent with
these observations (Scheme 11). In pathway (a) oxathiane
34 initially attacks an electrophilic oxygen atom in triflyloxy
sulfonium ion 6 to produce activated oxathiane 35 and di-
phenyl sulfide 24. Activated oxathiane 35 would then react
with the excess diphenyl sulfoxide to provide oxodisulfoni-
um ion 27.[30] Intermediate 27 could also be formed via
route (b) which similarly involves reaction at an electrophil-
ic oxygen atom, but this time in dication 8. Although there
is evidence in the literature for activated sulfoxide species
reacting through electrophilic oxygen atoms,[31] we consider
routes (a) and (b) to be less likely than attack at the softer
ketone 31. Reductive cyclisation in the presence of
TMSOTf and Et3SiH led directly to oxathaine rac-32 in
good yield. Stereoselective oxidation with diphenyl sulfoxide
and Tf2O proceeded as observed previously for glycosyl oxa-
thiane 3 with the axial isomer rac-33a dominating the crude
reaction mixture (88:12, ax/eq).[12] Careful column chroma-
tography allowed the pure axial isomer rac-33a to be isolat-
ed in 64% yield.
The sulfoxide was added to a pre-formed mixture of la-
belled diphenyl sulfoxide
N
electrophilic sulfur atoms in intermediates
6 and 8
(1.4 equiv) at À608C. The cross-over experiment was
quenched after 1.5 h by addition of diphenyl sulfide and
NaHCO3. NMR spectroscopy demonstrated that the crude
product mixture was an 85:15 mixture of the axial and equa-
torial sulfoxides, while mass spectrometry gave an overall
18O-incorporation of 37%. As the level of 18O-incorporation
significantly exceeds the quantity of the minor isomer, it
must be possible to have oxygen exchange with retention of
configuration in this non-carbohydrate system. Therefore
the intermolecular mechanism for inversion of configuration
is possible. Furthermore, these experiments demonstrate
that the oxidation reaction is also applicable to sulfides that
lack the sugar ring.
(Scheme 11c-d).[32]
If oxathiane 34 were to react at the sulfonium centres of
cation 6 (route c) or dication 8 (route d), a dithiadication in-
termediate 36 would be produced. Dithiadications have
been the subject of several reviews,[19] and this type of func-
tional group is invariably synthesised by reaction between a
sulfide and an activated sulfoxide.[33] Dithiadications have
also been implicated in the mechanism of “remote Pummer-
er reactions” and other processes involving transfer of
oxygen from sulfoxides to sulfides.[14f,g,i,34]
If a dithiadication were to form during the mechanism de-
picted in Scheme 11, it would still be necessary to introduce
the sulfoxide oxygen atom and to release diphenyl sulfide as
a by-product. Both of these steps could be achieved by di-
phenyl sulfoxide attack at the oxathiane sulfur atom in inter-
mediate 36 to provide oxodisulfonium ion 27. Therefore, re-
gardless of the initial reaction step between thioglycoside 34
and the activated sulfoxide species, all four mechanistic
As noted above, the oxidation reaction is highly stereose-
lective and for each example in Scheme 1 and Scheme 10,
the thermodynamically favoured isomer is the dominant
product (see the Supporting Information for DFT calcula-
tions on each structure). Considering that 1) axial and equa-
2994
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2987 – 2997

Sulfoxide Transfer Mediated by a Diphenyl Sulfoxide-Triflic Anhydride
FULL PAPER
Scheme 11. Possible reaction pathways for the oxidation of generic oxathiane 34. Mechanisms are depicted as SN2 processes for simplicity, although it is
likely they may involve addition-elimination mechanisms via sulfurane intermediates.
routes could converge on the oxygen-linked dication inter-
mediate 27.
ported. NMR spectroscopy data and 18O-cross-over experi-
ments support the formation of oxodisulfonium ions when
sulfoxides and Tf2O are mixed at low temperature. Such
species undergo exchange reactions with other sulfoxides,
thus facilitating oxygen exchange and inter-conversion of
axial and equatorial sulfoxides in oxathiane rings.
The final step of the reaction would involve intermediate
27 reacting with diphenyl sulfide 24 to release the oxa-
thiane-S-oxide product 37 and form the triaryl sulfonium
ion 13.[35] In this regard, it is worth reiterating that at no
point have we observed the formation of an aryl sulfonium
salt combining diphenyl sulfide and the oxathiane (e.g., 25,
Scheme 7). Therefore, if oxygen-linked dication 27 reacts di-
rectly with diphenyl sulfide, the reaction must occur very se-
lectively at the diphenyl sulfonium centre in intermediate
27. This may well be possible as alkyl sulfonium ions are
more stable than aryl sulfonium ions.[36]
It is also interesting to note that dithiadications can react
with electron-rich aromatic compounds to form aryl sulfoni-
um ions.[37] It is possible that an analogous reaction could be
responsible for the reduction reaction shown in Scheme 7;
reaction of diphenyl sulfide with activated sulfoxide 22
could give a dithiadication which might then react with a
second molecule of diphenyl sulfide to yield oxathiane 1 and
sulfonium salt 13.
The commonly used diphenyl sulfoxide/Tf2O reagent is
able to oxidise sulfides to sulfoxides. In this reaction diphen-
yl sulfoxide acts as the oxidant (following activation by
Tf2O) and also transfers the oxygen atom to the sulfide
through an oxodisulfonium ion intermediate 27
(Scheme 11). The first step of the oxidation process most
probably involves reaction of the sulfide with an activated
diphenyl sulfoxide species resulting in formation of a dithia-
dication intermediate 36. Either oxodisulfonium ion 8 or tri-
flyloxy sulfonium ion 6 could act as the activating agent. Al-
though oxodisulfonium ion 8 is the dominant species at
equilibrium for a 2:1 mixture of diphenyl sulfoxide and
Tf2O, triflyloxy sulfonium ion 6 is formed first and could
thus react directly with the sulfide. Either way, dithiadica-
tion 36 would then react with diphenylsulfoxide to introduce
the new sulfoxide oxygen atom via oxodisulfonium inter-
mediate 27. Diphenylsulfide produced in this step as an ini-
tial by-product, goes on to sequester activated sulfoxide spe-
cies, effectively quenching the reaction prior to aqueous
work-up.
Conclusion
Oxodisulfonium ions have been proposed previously as re-
action intermediates,[8c,31b] but to the best of our knowledge,
no experimental evidence for their existence has been re-
Sulfoxide exchange with the oxodisulfonium intermediate
27 presents an intriguing possibility that the stereochemical
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2995

W. B. Turnbull et al.
hydrates 2003, 121–145; d) W. B. Turnbull, M. A. Fascione, S. A.
Stalford, Sci. Synth. 2007, 29, 923–969; e) A. T. Carmona, A. J.
Moreno-Vargas, I. Robina, Curr. Org. Synth. 2008, 5, 33–60; f) D.
Crich, A. A. Bowers, in Handbook of Chemical Glycosylation, (Ed.:
A. V. Demchenko), Wiley-VCH, Weinheim, 2008, pp. 303–329;
g) M. C. Aversa, A. Barattucci, P. Bonaccorsi, Tetrahedron 2008, 64,
7659–7683; h) X. Zhu, R. R. Schmidt, Angew. Chem. 2009, 121,
1932–1967; Angew. Chem. Int. Ed. 2009, 48, 1900–1934.
outcome of the oxidation process could occur under thermo-
dynamic control. Of course, this mechanism would require
rapid equilibration of the oxodisulfonium intermediates
prior to deactivation by diphenylsulfoxide. Without a de-
tailed knowledge of the kinetics of the reaction it is not pos-
sible to conclude whether or not the original oxidation re-
ACHTUNGTRENNUNGaction is under kinetic or thermodynamic control. Neverthe-
[5] D. Kahne, S. Walker, Y. Cheng, D. Vanengen, J. Am. Chem. Soc.
1989, 111, 6881–6882.
[6] B. A. Garcia, J. L. Poole, D. Y. Gin, J. Am. Chem. Soc. 1997, 119,
7597–7598.
less, it may be possible that the system could be engineered
to slow or remove the diphenyl sulfide termination step, for
example by using a substituted diaryl sulfoxide. If so, this
oxygen-transfer reaction would present a novel strategy for
the stereoselective synthesis of sulfoxides.
[7] a) B. A. Garcia, D. Y. Gin, J. Am. Chem. Soc. 2000, 122, 4269–4279;
b) Y.-J. Kim, D. Y. Gin, Org. Lett. 2001, 3, 1801–1804; c) H. M.
Nguyen, J. L. Poole, D. Y. Gin, Angew. Chem. 2001, 113, 428–431;
H. M. Nguyen, J. L. Poole, D. Y. Gin, Angew. Chem. 2001, 113, 428–
431; Angew. Chem. Int. Ed. 2001, 40, 414–417; d) J. Dinkelaar,
J. D. C. Codꢄe, L. J. van den Bos, H. S. Overkleeft, G. A. van der
Marel, J. Org. Chem. 2007, 72, 5737–5742; e) J. M. Haberman, D. Y.
Gin, Org. Lett. 2003, 5, 2539–2541; f) J. D. C. Codꢄe, L. H. Hossain,
P. H. Seeberger, Org. Lett. 2005, 7, 3251–3254; g) D. Ye, W. Liu, D.
Zhang, E. Feng, H. Jiang, H. Liu, J. Org. Chem. 2009, 74, 1733–
1735.
Experimental Section
Full experimental details and analytical data for compounds prepared
and coordinates for compounds modelled using DFT calculations are in-
cluded in the Supporting Information. CCDC-733122 (2-R)[10a] and
CCDC-733123 (4-S)[10c] contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
[8] a) J.-Y. Kim, V. Di Bussolo, D. Y. Gin, Org. Lett. 2001, 3, 303–306;
b) L. Shi, Y.-J. Kim, D. Y. Gin, J. Am. Chem. Soc. 2001, 123, 6939–
6940; c) E. Honda, D. Y. Gin, J. Am. Chem. Soc. 2002, 124, 7343–
7352.
General experimental methods: All solvents were dried prior to use, ac-
cording to standard methods.[38] Trifluoromethanesulfonic anhydride
(Tf2O), was freshly distilled from phosphorus pentoxide under a N2(g) at-
mosphere and all other commercially available reagents were used as re-
ceived. All reactions were performed in oven dried glassware under a
N2(g) atmosphere.
[9] a) J. D. C. Codꢄe, R. E. J. N. Litjens, R. den Heeten, H. S. Over-
kleeft, J. H. van Boom and G. A. van der Marel, Org. Lett. 2003, 5,
1519–1522; b) J. D. C. Codꢄe, L. J. van den Bos, R. E. J. N. Litjens,
H. S. Overkleeft, J. H. van Boom, G. A. van der Marel, Org. Lett.
2003, 5, 1947–1950; c) J. D. C. Codꢄe, L. J. van den Bos, R. E. J. N.
Litjens, H. S. Overkleeft, C. A. A. van Boeckel, J. H. van Boom,
G. A. van der Marel, Tetrahedron 2004, 60, 1057–1064; d) R. E. J. N.
Litjens, L. J. van den Bos, J. D. C. Codꢄe, R. J. B. H. N. van den
Berg, H. S. Overkleeft, G. A. van der Marel, Eur. J. Org. Chem.
2005, 918–924; e) D. Cato, T. Buskas, G.-J. Boons, J. Carbohydr.
Chem. 2005, 24, 503–516; f) D. Crich, W. Li, Org. Lett. 2006, 8, 959–
962; g) J. Dinkelaar, L. J. van den Bos, W. F. J. Hogendorf, G.
Lodder, H. S. Overkleeft, J. D. C. Codꢄe, G. A. van der Marel,
Chem. Eur. J. 2008, 14, 9400–9411.
General procedure for oxidation reaction: Tf2O (1.4 equiv) was added to
a solution of sulfide (1 equiv), diphenyl sulfoxide (2.8 equiv) and 2,6-di-
tert-butyl-4-methylpyridine (DTBMP) (3 equiv) in CH2Cl2 (sulfide con-
centration: 250–500 mm) at À608C. After 40–90 min at that temperature
the reaction was quenched by addition of sat. NaHCO3 solution. The
mixture was extracted into CH2Cl2, dried (MgSO4) and concentrated in
vacuo to give the crude product which was purified by silica gel chroma-
tography.
[10] a) M. A. Fascione, S. J. Adshead, S. A. Stalford, C. A. Kilner, A. G.
Leach, W. B. Turnbull, Chem. Commun. 2009, 5841–5843; b) M. A.
Fascione, W. B. Turnbull, Beilstein J. Org. Chem. 2010, 6, No. 19;
c) M. A. Fascione, C. A. Kilner, A. G. Leach, W. B. Turnbull, Chem.
Eur. J. 2012, 18, 321–333.
Acknowledgements
This work was supported by The Royal Society (UF051621/UF090025),
EPSRC (EP/G043302/1), the University of Leeds and AstraZeneca.
W.B.T. is the recipient of a Royal Society University Research Fellow-
ship. We would like to thank Tanya Marinko-Covell, Dr. Stuart Warriner
and Simon Barrett for assistance with mass spectrometry and variable
temperature NMR experiments. We also thank Caroline Edwards,
Andrew Smith and Dr. Jon Goode (Mettler Toledo AutoChem) for ac-
quisition and analysis of the ReactIR data.
[11] In each case the stereochemistry of the major isomer was deter-
mined by X-ray crystallography.
[12] The 1H NMR geminal coupling constant for the methylene group
adjacent to the sulfoxide is diagnostic of the stereochemistry of sulf-
oxide groups in six-membered rings; Jax,eq =12 Hz for equatorial sulf-
oxide and 14 Hz for axial sulfoxides; a) J. B. Lambert, S. I. Feather-
man, Chem. Rev. 1975, 75, 611–625; b) J. B. Lambert, C. E. Mixan,
D. H. Johnson, J. Am. Chem. Soc. 1973, 95, 4634–4639.
[13] a) A. D. Becke, J. Chem. Phys. 1993, 98, 1372–1377; b) P. C. Hari-
haran, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222; c) C. Lee,
W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789; d) Y. Zhao, D.
Truhlar, Theor. Chem. Acc. 2008, 120, 215–241.
[1] I. L. Baraznenok, V. G. Nenajdenko, E. S. Balenkova, Tetrahedron
2000, 56, 3077–3119.
[2] a) V. G. Nenajdenko, P. V. Vertelezkij, E. S. Balenkova, Sulfur Lett.
1996, 20, 75–84; b) V. G. Nenajdenko, P. V. Vertelezkij, E. S. Balen-
kova, Synthesis 1997, 351–355; c) V. G. Nenajdenko, P. V. Vertelez-
kij, I. D. Gridnev, N. E. Shevchenko, E. Balenkova, Tetrahedron
1997, 53, 8173–8180.
[3] V. G. Nenajdenko, P. V. Vertelezkij, A. B. Koldobskij, I. V. Alabugin,
E. S. Balenkova, J. Org. Chem. 1997, 62, 2483–2486.
[4] a) T. Norberg, Front. Nat. Prod. Res. 1996, 1, 82–106; b) C. M.
Taylor in Solid Support Oligosaccharide Synthesis and Combinatorial
Carbohydrate Libraries, (Ed.: P. H. Seeberger), John Wiley & Sons,
New York, 2002, pp. 41–65; c) K. P. R. Kartha, R. A. Field, Carbo-
[14] a) D. Barnard, Chem. Ind. 1955, 565–566; b) F. G. Bordwell, B. M.
Pitt, J. Am. Chem. Soc. 1955, 77, 572–577; c) S. Searles, Jr., H. R.
Hays, J. Org. Chem. 1958, 23, 2028–2029; d) C. M. Hull, T. W.
Bargar, J. Org. Chem. 1975, 40, 3152–3154; e) A. Bovio, U. Miotti,
J. Chem. Soc. Perkin Trans. 2 1978, 172–177; f) V. Ruffato, U.
Miotti, Gazz. Chim. Ital. 1978, 108, 91–96; g) R. Tanikaga, K. Na-
kayama, K. Tanaka, A. Kaji, Chem. Lett. 1977, 395–396; h) U.
Miotti, J. Chem. Soc. Perkin Trans. 2 1991, 617–622; i) K. Kobaya-
shi, E. Koyama, C. Kono, K. Namatame, K. Nakamura, N. Furuka-
wa, J. Org. Chem. 2001, 66, 2085–2090.
2996
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2987 – 2997

Sulfoxide Transfer Mediated by a Diphenyl Sulfoxide-Triflic Anhydride
FULL PAPER
[15] a) D. Barnard, D. T. Woodbridge, Chem. Ind. 1959, 1603; b) F.
Ogura, H. Yamaguchi, T. Otsubo, H. Tanaka, Bull. Chem. Soc. Jpn.
1982, 55, 641–642; c) D. J. Procter, M. Thornton-Pett, C. M. Rayner,
Tetrahedron 1996, 52, 1841–1854.
[16] a) T. Nojima, Y. Hirano, K. Ishiguro, Y. Sawaki, J. Org. Chem. 1997,
62, 2387–2395; b) N. Furukawa, S. Ogawa, K. Matsumura, H. Fuji-
hara, J. Org. Chem. 1991, 56, 6341–6348; c) A. Okruszek, J. La-
belled Compd. Radiopharm. 1983, 20, 741–743; d) J. C. Martin, E. F.
Perozzi, J. Am. Chem. Soc. 1974, 96, 3155–3168; e) S. Oae, Y.
Onishi, S. Kozuka, W. Tagaki, Bull. Chem. Soc. Jpn. 1966, 39, 364–
366; f) Y. Sato, N. Kunieda, M. Kinoshita, Chem. Lett. 1972, 1023–
1026.
[17] The possibility of triflic acid-catalysed oxygen exchange was dis-
counted as no change in 18O-incorporation was observed in the la-
belling experiments if the reaction mixture was neutralised immedi-
ately or 30 min after aqueous work-up.
[18] a) S.-L. You, H. Razavi, J. W. Kelly, Angew. Chem. 2003, 115, 87–
89; Angew. Chem. Int. Ed. 2003, 42, 83–85; S.-L. You, H. Razavi,
J. W. Kelly, Angew. Chem. 2003, 115, 87–89; b) M. Mossotti, L.
Panza, J. Org. Chem. 2011, 76, 9122–9126.
yl sulfide is unreactive to Tf2O, even at room temperature, see
Ref. [3]. Therefore, it must be an activated diphenyl sulfoxide spe-
cies that reacts with diphenyl sulfide to produce sulfonium ion 13.
[28] An attempt to repeat the oxygen cross-over experiment starting
from the axial oxathiane-S-oxide 2-S led only to decomposition
products.
[29] Direct reaction of the oxathiane sulfur atom with Tf2O would not
be a productive step as it would lead to reduction of Tf2O to tri-
fluoromethanesulfinate rather than reduction of diphenyl sulfoxide
to diphenyl sulfide (see Scheme 6b).
[30] Activated oxathiane 35 could alternatively react with diphenyl sulf-
oxide at its sulfur(VI) centre to release the oxathiane-S-oxide prod-
uct 37 directly. However, our 18O-crossover experiments
(Scheme 8c) suggest that if intermediate 35 is produced during the
reaction, its major fate would be to react with diphenyl sulfoxide at
its sulfur(IV) centre to produce oxodisulfonium ion 27. Further-
more, Gin and co-workers have observed that other oxygen nucleo-
philes react selectively at the sulfur(IV) centre of triflyloxy sulfoni-
um ion 6 (Ref. [7a]) and related systems (Ref. [23])
[31] For a mechanism for self-immolative reduction of sulfoxides by trifl-
ic anhydride in the absence of competing nucleophiles, see: Y.
Macꢄ, J.-C. Blazejewski, C. Pradet and E. Magnier, Eur. J. Org.
Chem. 2010, 2010, 5772–5776. In principle this process could pro-
vide a fifth possible mechanism for the oxidation reaction reported
here. The mechanism would start with attack of diphenyl sulfoxide
on an electrophilic oxygen in activated sulfoxide 6 to generate a per-
oxysulfonium ion and diphenyl sulfide 24. These species would react
together to form sulfonium ion 13 and peroxytriflic acid which could
then oxidise oxathiane 34 to sulfoxide 37. However, we would con-
sider this route very unlikely as the self-immolative reduction reac-
tion is very slow taking days to reach completion and would thus
not be competitive with the very rapid reaction that we observe be-
tween a sulfide and activated sulfoxide.
[32] Honda and Gin considered intermediate 8 as an oxygen electrophile
during studies into the mechanism of glycal C2-hydroxyglycosylation
(Ref. [8c]); however, in that case, experimental data favoured initial
reaction of the glycal at an electrophilic sulfur atom.
[33] a) F. Iwasaki, N. Toyoda, R. Akaishi, H. Fujihara, N. Furukawa,
Bull. Chem. Soc. Jpn. 1988, 61, 2563–2567; b) V. G. Nenajdenko,
N. E. Shevchenko, E. S. Balenkova, Tetrahedron 1998, 54, 5353–
5362.
[34] K. Kobayashi, E. Koyama, K. Namatame, T. Kitaura, C. Kono, M.
Goto, T. Obinata, N. Furukawa, J. Org. Chem. 1999, 64, 3190–3195.
[35] Alternatively, one could also imagine a mechanism in which dication
27 reacts first with diphenyl sulfoxide to regenerate tetraphenyloxo-
disulfonium ion 8, which might then be sequestered by diphenyl sul-
fide.
[19] a) W. K. Musker, Acc. Chem. Res. 1980, 13, 200–206; b) N. Furuka-
wa, K. Kobayashi, S. Sato, J. Organomet. Chem. 2000, 611, 116–126;
c) V. G. Nenajdenko, N. E. Shevchenko, E. S. Balenkova, I. V. Ala-
bugin, Chem. Rev. 2003, 103, 229–282.
[20] In contrast, the corresponding sulfur atoms in sulfurane 7 are over
À
3.5 ꢃ apart, and its S O bond lengths are similar to those in diphen-
yl sulfoxide 5 (ca. 1.5 ꢃ) and triflyloxy sulfonium ion 6 (ca. 1.7 ꢃ);
as such, sulfurane 7 more closely resembles an adduct in which sulf-
À
oxide 5 and monocation 6 are connected through a long S O bond
(2.36 ꢃ). Sulfurane 9 is similar in structure to sulfurane 7 except in
À
the lengths of the apical S O bonds. It is possible that sulfuranes 7
and 9 are intermediates in addition-elimination process that would
allow the interconversion of monocation 6 and dication 8, or triflate
exchange in 6: N. Furukawa, S. Sato, Top. Curr. Chem. 1999, 205,
89–129.
[21] 19F NMR chemical shifts are reported relative to the IUPAC stan-
dard CFCl3 (0 ppm), by using hexafluorobenzene (À162.2 ppm) as
an internal reference; W. R. Dolbier, Jr., Guide to Fluorine NMR
for Organic Chemists, John Wiley & Sons, New York, 2009, p. 5.
[22] The signal at À79.1 ppm has an integration three times greater than
the signal for Tf2O at À72.0. There is also a significant difference
between the chemical shifts for the species in Figure 1 f when com-
pared to the triflate signals in Figure 1 g and h.
[23] a) T. A. Boebel, D. Y. Gin, J. Org. Chem. 2005, 70, 5818–5826;
b) T. A. Boebel, D. Y. Gin, Angew. Chem. 2003, 115, 6054–6057;
Angew. Chem. Int. Ed. 2003, 42, 5874–5877.
[24] M. J. P. Harger, Org. Biomol. Chem. 2006, 4, 2842–2844.
[25] J. B. Hendrickson, S. M. Schwartzman, Tetrahedron Lett. 1975, 16,
273–276.
[36] M. G. Beaver, S. B. Billings, K. A. Woerpel, J. Am. Chem. Soc. 2008,
130, 2082–2086.
[26] a) J.-H. Park, D.-C. Seo, S.-J. Park, S.-J. Kim (Korea Kumho Petro-
chemical Co. Ltd.), EP 0972761, 2000; b) S. R. Akhtar, J. V. Crivello,
J. L. Lee, J. Org. Chem. 1990, 55, 4222–4225.
[37] H. Fujihara, R. Akaishi, N. Furukawa, Chem. Lett. 1988, 709–712.
[38] W. L. F. Armarego, D. D. Perrin, Purification of Laboratory Chemi-
cals, Butterworth-Heinemann, Oxford, 1996, p. 529.
[27] Aliphatic sulfides can also be activated by Tf2O to undergo electro-
philic aromatic substitution with aromatic compounds: a) V. G. Ne-
najdenko, P. V. Vertelezkij, E. S. Balenkova, Sulfur Lett. 1996, 20,
75–84; b) N. E. Shevchenko, V. G. Nenajdenko, E. S. Balenkova,
Synthesis 2003, 1191–1200. However, it has been noted that diphen-
Received: September 13, 2011
Revised: November 25, 2011
Published online: January 31, 2012
Chem. Eur. J. 2012, 18, 2987 – 2997
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2997