Single-Step Glycosylations with 13C-Labelled Sulfoxide Donors: A Low-Temperature NMR Cartography of the Distinguishing Mechanistic Intermediates

Article abstract of DOI:10.1002/chem.202003850

Glycosyl sulfoxides have gained recognition in the total synthesis of complex oligosaccharides and as model substrates for dissecting the mechanisms involved. Reactions of these donors are usually performed under pre-activation conditions, but an experimentally more convenient single-step protocol has also been reported, whereby activation is performed in the presence of the acceptor alcohol; yet, the nature and prevalence of the reaction intermediates formed in this more complex scenario have comparatively received minimal attention. Herein, a systematic NMR-based study employing both ¹³C-labelled and unlabelled glycosyl sulfoxide donors for the detection and monitoring of marginally populated intermediates is reported. The results conclusively show that glycosyl triflates play a key role in these glycosylations despite the presence of the acceptor alcohol. Importantly, the formation of covalent donor/acceptor sulfonium adducts was identified as the main competing reaction, and thus a non-productive consumption of the acceptor that could limit the reaction yield was revealed.

Full text of DOI:10.1002/chem.202003850

Accepted Article
Accepted Article
Title: Single-step glycosylations with 13C-labelled sulfoxide donors:
a low-temperature NMR cartography of the distinguishing
mechanistic intermediates
Authors: Juan Luis Asensio, Andrés G. Santana, Laura Montalvillo-
Jiménez, Laura Díaz-Casado, Enrique Mann, Jesús Jiménez-
Barbero, and Ana M. Gómez
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202003850
Link to VoR: https://doi.org/10.1002/chem.202003850
01/2020

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Single-step glycosylations with 13C-labelled sulfoxide donors: a
low-temperature NMR cartography of the distinguishing
mechanistic intermediates
Andrés G. Santana,*§[a] Laura Montalvillo-Jiménez,§[a] Laura Díaz-Casado,[a] Enrique Mann,[a] Jesús
Jiménez-Barbero,[b,c] Ana M. Gómez[d] and Juan Luis Asensio*[a]
[a]
Dr. A. G. Santana, Dr. L. Montalvillo-Jiménez, L. Díaz-Casado, Dr. E. Mann, Dr. J. L. Asensio
Glycochemistry and Molecular Recognition group – Dpt. Bio-Organic Chemistry
Instituto de Química Orgánica General (IQOG–CSIC)
Juan de la Cierva 3. 28006 Madrid (Spain)
E-mail: andres.g.santana@csic.es; juanluis.asensio@csic.es
Prof. J. Jiménez-Barbero
Center for Cooperative Research in Biosciences (CIC-bioGUNE)
48160 Derio (Spain)
Prof. J. Jiménez-Barbero
[b]
[c]
[d]
Basque Foundation for Science
48013 Bilbao (Spain)
Dr. A. M. Gómez
Oligosaccharide and Glycosystems group – Dpt. Bio-Organic Chemistry
Instituto de Química Orgánica General – CSIC
Juan de la Cierva 3. 28006 Madrid (Spain)
Supporting information for this article is given via a link at the end of the document.
Abstract: Glycosyl sulfoxides have gained recognition in the total
synthesis of complex oligosaccharides and as model substrates for
dissecting the mechanisms involved. While these donors are usually
reacted under pre-activation conditions, an experimentally more
convenient single-step protocol has also been reported, whereby the
activation is performed in the presence of the acceptor alcohol; yet,
the nature and prevalence of the reaction intermediates formed in this
more complex scenario have comparatively received minimal
attention. Herein, we report a systematic NMR-based study employing
both 13C-labelled and unlabelled glycosyl sulfoxide donors for the
detection and monitorization of marginally populated intermediates.
Our results conclusively show that glycosyl triflates play a key role in
these glycosylations despite the presence of the acceptor alcohol.
Importantly, the formation of covalent donor/acceptor sulfonium
adducts has been identified as the main competing reaction, thus
defining a non-productive consumption of the acceptor that could limit
the reaction yield.
back in the mid-1990s with 4,6-O-benzylidene D-mannopyranosyl
sulfoxide donors, paved the way for the stereoselective
construction of 1,2-cis -D-mannopyranosides, one of the most
challenging glycosidic linkages to synthetically attain until then.
Thus,
a
first protocol based on the pre-activation[14] of
mannopyranosyl sulfoxides with triflic anhydride was shown to
take place through putative sulfonium triflate species ultimately
leading to -glycosyl triflates, which produced the final -
mannopyranosides through an SN2-like displacement by the
incoming nucleophile (Scheme 1a).[15,16] Low-temperature NMR
analysis of the reaction mixture after pre-activation and before
addition of the acceptor provided the required evidence for the α-
glycosyl triflate existence and its role as the key intermediate in
the process. Opposite to that, a close ion pair (CIP) intermediate
was proposed to account for the different stereoselective outcome
of the in situ activation variant of the same reaction performed in
Et2O, although the -stereoselective trend was restored in DCM,
thus signalling a more complex scenario (Scheme 1b).[7,17]
It is now well-accepted that glycosylation reactions can occur
within a mechanistic continuum spanning the gap between SN1
and SN2 archetypes, and that the identification and
characterization of the reactive species, e.g. oxocarbenium-like or
other neutral or cationic covalent intermediates, can help
understand the reaction pathway and eventually allow to control
its outcome.[18,19] In this context, kinetic isotope effects,[20] cation
clock methodologies[21] and more recently NMR[22] have been
used to study these chemical processes. Interestingly, in many
occasions triflate anions have been shown to facilitate the
reaction course, probably by promoting the formation of reactive
glycosyl triflate species through nucleophilic catalysis.[23] For
sulfoxide-promoted glycosylations, numerous activated entities
have been postulated to coexist upon pre-activation of the donor,
above which glycosyl triflates seem to occupy a hierarchical posi-
Introduction
The realization of the key role played by oligosaccharides in
biological processes has undoubtedly sparked the search for
more efficient stereoselective glycosylation methods.[1-4] These
reactions involve the coupling of a glycosyl donor and an acceptor,
and are the central transformation in saccharide synthesis. More
recently, it has also become clear that a better understanding of
the mechanistic aspects of the glycosylation process can
contribute to its optimization in terms of yield and
stereoselectivity.[5-11] A case in point in this regard has been
Crich’s -mannosylation reaction, one of the best-studied
glycosylations to date.[12,13] These investigations, which started
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Scheme 1. a) Proposed reaction mechanism and intermediates involved in glycosylations with sulfoxide donors under pre-activation conditions. b) The role played
by glycosyl triflates along with other alternative intermediates under single-step conditions is presently unknown and represents the central topic of our study.
tion, both mechanistically and concentration-wise, acting as either
true glycosylating species against the acceptor or as reservoirs of
even more reactive cationic intermediates.[24] However, this
scenario might not hold when the acceptor alcohol is present in
the reaction mixture from the beginning. Specifically, the critical
role of glycosyl triflates in the presence of other competing
nucleophiles has not been proven to date, and this assumption is
therefore questionable. Moreover, the occurrence of previously
undetected reaction intermediates involving donor and/or
acceptor could also be viable under these experimental
conditions. From all of the above, it became evident to us that the
study of reaction intermediates in single-step glycosylations with
in situ activation, as well as their intertwined reactivity dynamics,
is pertinent to clarify the main reaction pathways governing
glycoside formation, ultimately eliciting the development of more
efficient and stereoselective saccharide synthesis. It should be
noted that dissecting glycosylation mechanisms has proven a
challenging task given the high reactivity and marginal
concentration of many of the key reacting intermediates involved,
often present within complex mixtures. Indeed, the ability to
monitor the reaction progress with sufficient sensitivity and time
resolution is key to fulfil this purpose. Herein, we would like to
disclose our studies on the single-step variant of the glycosyl
sulfoxide glycosylation reaction triggered by triflic anhydride. The
employment of both 13C-labelled and unlabelled glycosyl sulfoxide
donors will enable, for the first time, the monitorization of in situ
promoted glycosylations by low-temperature NMR experiments.
The scope and limitations of this methodology have been
determined as a function of various structural and experimental
parameters, thus helping compose a detailed picture of the
ongoing mechanistic landscape.
display identical intensities throughout the reaction, pointing to the
accumulation of a single intermediate species (I-1a). To increase
the reaction rate, the temperature was raised to -45 °C till this
latter species reached an adequate population for 2D-NMR
analysis (40%).[25] At this point, we cooled the sample back to -
65 °C and acquired 2D-NOESY and TOCSY experiments. The
obtained data sets allowed assigning NMR peaks at 6.05 and 5.25
ppm to mannose and galactose anomeric protons, respectively.
More importantly, both scaffolds belong to a discrete structure,
which also integrates the -SO-Ph fragment, as revealed by the
presence of weak NOEs connecting both anomeric signals with
the sulfoxide phenyl ring (Figure 1c).[26] At higher temperatures
(ca. -20 °C), the detected intermediate evolves to produce a
mixture of / (1:5) glycosylation products (1a-/) with an
excellent yield (90%). Significantly, glycosylations carried out with
a pre-activation protocol furnished an identical stereochemical
outcome, indicating that the -mannosyl triflate must still be
playing an important role in both glycosylation processes.
This point was further confirmed with 13C-labelling of the anomeric
carbon of the mannopyranosyl sulfoxide 1. Thus, reaction courses
monitored through sequential 2D-HSQC experiments revealed
the presence of a minor transient concentration of the -mannosyl
triflate (6.18 and 97.4 ppm, respectively for anomeric 1H and 13
C
chemical shifts; indicated also in Figure 1d), together with a small
amount of the reduced thioglycoside (4.90 ppm and 88.7 ppm)
and the previously described intermediate (6.05 and 95.8 ppm;
1JHC = 168 MHz) which is now detected along with a minor isomer
(6.31 and 97.4 ppm) as judged by the similar 1H and 13C chemical
shifts and identical 1JHC anomeric coupling (168 MHz). Based on
the obtained spectroscopic data, these two latter species were
assigned as the sulfonium salts I-1a (two stereomers),
presumably formed from the triflate-activated sulfonium (B,
Scheme 1).[27] According to our data such intermediate undergoes
a triflate substitution by the acceptor alcohol at the sulfur atom, to
yield a more stable sulfonium intermediate (I-1a) which, even in
the presence of additional alcohol in the reaction mixture, evolves
to produce the key glycosyl triflate (1-Tf), perhaps by releasing
a neutral phenyl-sulfenate. It is worth mentioning that the
corresponding acceptor-sulfenate has been neither isolated nor
detected by us. Instead a sulfinate, as those reported by
Seeberger’s group[28] has been identified by HRMS directly from
the glycosidation crude. We hypothesised that the relative
population of the detected reaction intermediates (I-1a and 1-Tf)
along with their intrinsic reactivities should be strongly dependent
Results and Discussion
To have a first indication of the reactive species involved in the
one-step process scenario, unlabelled D-manno-donor 1 was
activated in the presence of diacetone-D-galactose (acceptor a,
2.2 equiv) and DTBMP (3 equiv) at -60 °C and the reaction
progress was followed by 1D-NMR experiments (Figures 1a,b
and S1). It can be observed that immediately after donor
activation two new signals appear in the anomeric region of the
spectrum (6.05 and 5.25 ppm, respectively). Interestingly, they
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 1. a) Schematic representation of glycosylation reactions involving donor 1 and acceptor a with a single-step (left) or a pre-activation (right) protocol. Obtained
yields for  and  products are indicated. b) Time evolution of a donor 1/acceptor a mixture upon activation with Tf2O in CDCl3 at -60 °C. NMR anomeric signals for
intermediate I-1a are indicated with a mauve circle. c) Through-space connections revealed by NOESY experiments (left) allowed the structural assignment of
intermediate I-1a to the represented donor/acceptor sulfonium adduct (right). d) Reactions carried out with 13C-labelled 1 and monitored by means of HSQC spectra
confirmed the coexistence of donor/acceptor sulfonium adducts (I-1a) and -glycosyl triflates (1-Tf) as the only detectable reaction intermediates. 1H and 13C
chemical shifts for the detected intermediates are also shown.
on both the donor and acceptor chemical properties. With this
idea in mind, we set out to investigate the structural and electronic
requirements governing the stability and reactivity of these
sulfonium adducts. As a first step, we focused our attention on the
influence exerted by the anomeric configuration. For this purpose,
parallel single-step glycosylations of -manno (1) and -manno
(2) sulfoxide donors with acceptor a were monitored by low-
temperature NMR under otherwise identical conditions. Fittingly,
significant differences between these two anomers were
immediately apparent, since -donor 1 exhibited the formation of
multiple activated species, as previously mentioned, while -
donor 2 exclusively provided the corresponding -mannosyl
triflate (1-Tf) as the only detectable intermediate. As a result, this
second glycosylation finished in a shorter reaction time and with
a considerable better yield than the one employing donor 1 as
starting material (Figure 2). Indeed, this experimental evidence
suggests a non-productive consumption of the acceptor alcohol
in the formation of the epimeric sulfonium species that cannot be
recovered later, and therefore presents a significant limitation of
the reaction’s overall performance. Consequently, this finding
clearly highlights the importance of the donor anomeric
configuration in the reaction evolution under one-step activation
conditions, and points toward much favourable sulfonium
formation with a beta anomeric configuration, either because of
grave steric hindrance of the sulfonium adduct with an -axial
configuration, or because of a stabilization of the cationic aglycon
in the  configuration by reverse anomeric effect.[29] However,
along this line, an alternative rationale based on the easiness of
the expelling of the leaving group in the axial alfa anomer,
compared to the conformationally locked beta isomer, could not
be excluded.
Encouraged by these results, we tested this point further to
generalise the scope of our conclusions by carrying out additional
glycosylation reactions employing 13C-labelled donors[30] with -
manno (1), -manno (2) and -gluco (3) configurations and
alcohols spanning a range of nucleophilicities (trifluoroethanol, b;
monofluoroethanol, c; and isopropanol, d). In all cases, acceptors
were employed in a significant excess (4 equiv) to ensure that this
reagent was not limiting, except for those glycosylations involving
diacetone-D-galactose as the acceptor counterpart, that were
carried out under more realistic conditions using only a slight
excess (1.3 equiv) of the alcohol (Figure 3). In Figure 3a, three
sets of HSQC experiments acquired 5-10 min after activation with
Tf2O are shown, one for each glycosyl donor. The obtained yields
for  and  products depicted correspond to a reaction time of 1
hour at the indicated temperature (time evolutions for some of the
analysed reaction mixtures are fully displayed in Figures 2 and
1
S2-S8). Anomeric 1H and 13C chemical shifts and JHC
heteronuclear coupling constants for some selected sulfonium
intermediates are also represented below (Figure 3b). According
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 2. Influence of the anomeric configuration in the reaction intermediate profile: comparative reaction evolution of -manno- (1) and -manno-donors (2) with
acceptor a. Time-evolution of 13C-labelled 1 (10 mM; right) and 2 (10 mM; left) in the presence of the galactose acceptor a (1.3 equiv) and DTBMP (3 equiv) after
addition of Tf2O (2 equiv) at the indicated temperature, monitored by 2D-HSQC NMR experiments. Assignments for key anomeric signals for the starting materials,
intermediates and products, together with their 1JHC coupling constants are shown. Final yields are indicated in cyan at the bottom of each panel.
to these data, it quickly became evident that the evolution path for
each glycosylation reaction is highly dependent on the structural
and electronic properties of both donor and acceptor species.
Thus, the impact of the electronic nature of the acceptor on
sulfonium adduct formation was apparent when more nucleophilic
alcohols (such as 2-fluoroethanol, MFE) were employed, which
translates into a paradoxically slower progression of the reaction
with richer rather than with weakly nucleophilic alcohols. Thus, the
balance between glycosyl-triflate and sulfonium transient
concentrations is highly dependent on the nucleophilic character
of the acceptor alcohol: while triflates are dominant with poor
nucleophiles (like TFE), these species progressively disappear in
favour of the sulfonium adducts with increasingly better
nucleophiles, till they become virtually undetectable (Figure 3,
from top to bottom within each panel). It is worth mentioning that
a set of reactions involving the same donors and acceptors were
performed under the pre-activation protocol, which provided a
mixture of the expected glycosylation products (Figure 3c; see the
supp. info. for experimental details). Interestingly, the observed
stereoselectivities seem relatively insensitive to the activation
protocol employed, showing only minor variations for single-step
vs pre-activation reactions with identical triflic anhydride
concentrations (this experimental parameter has been shown to
have an influence on the / ratios, increasing -product
population at larger concentrations of the activating agent).[19]
This remark points toward a convergent reaction pathway and
underlines the relevance of glycosyl triflates as the key common
reaction intermediate, whether detectable or not.
Regarding the donor structural features, the anomeric
configuration is shown to have a major influence on the steady
concentrations of reactive intermediates. Thus, while
stereoselectivity trends for both - and -mannose donors are
similar, the former exhibits a markedly reduced tendency to form
sulfonium intermediates (vide supra). Accordingly, the -mannose
donor reacts with a higher accumulation of sulfonium species and
a concomitant decrease in the obtained yields with respect to its
-epimer is observed. This effect is particularly clear with the
employed galactose acceptor, which, as previously mentioned,
produces significant transient concentrations of sulfonium
adducts only with the -mannose donor. On the contrary, for the
-sulfoxide the reaction seems to proceed cleanly through the
glycosyl triflate with an excellent yield. Hence, the different
propensity to form covalent donor/acceptor adducts renders -
and -donors not equivalent, being the former recommended
based on its improved yields. Furthermore, the axial/equatorial
configuration of position 2, neighbouring to the acetalic centre,
also exerts a strong influence on the formation of sulfonium
species. In fact, this process seems greatly facilitated for D-gluco-
with respect to D-manno-donors (Figure 3a). Thus, for donor 3,
exhibiting equatorial orientations for both C-1 and C-2
substituents, donor/acceptor adducts are detectable even with
weakly nucleophilic acceptors such as TFE. This observation
agrees with the overall poor performance exhibited by the -
glucosyl sulfoxide donor in all the glycosylations tested.
Overall, increased accumulation of transient donor/acceptor
sulfonium adducts directly correlates with decreased yields. This
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 3. a) NMR detection of key intermediates in glycosylations carried out with 13C-labelled donors 1-3 and acceptors a-d (structures shown on the right). HSQC
spectra (anomeric region) acquired with the unmodified donors and 5-10 min after activation of the different reaction mixtures are shown. Key signals are coloured
according to the code shown on the right for intermediates and products. Conversions for - and -glycosylation products, after 60 min of evolution are indicated.
1
b) Representative examples of the detected donor/acceptor sulfonium adducts, together with their anomeric 1H and 13C chemical shifts and JHC values. c)
Stereoselectivities obtained with acceptors a-d employing a pre-activation protocol are represented for comparison.
negative influence on the glycosylation process results from a
combination of two effects: first, stable sulfonium formation
implies a non-productive consumption of the acceptor, which
might be limiting when used in stoichiometric or sub-
stoichiometric amounts (e.g. acceptor a cases). This further
aggravates the problems derived from direct triflation of the
acceptor with triflic anhydride, especially relevant for more
nucleophilic alcohols (e.g. acceptor d). Secondly, good
nucleophiles lead to exceedingly stable adducts (e.g. I-3d) whose
evolution require temperatures at which parasitic reactions might
be operative, perhaps involving the own sulfonium adduct, the
glycosyl triflate or even more reactive glycosyl oxocarbenium-like
species. According to this view, sulfonium adducts would be
acting as a kinetic trap for a successful glycosylation process.
On the other hand, taking into account that some literature
precedents had already shown that the absolute configuration at
sulfur, in anomeric sulfoxides, has some effect in the outcome of
their glycosylation reactions, we have designed some
experiments to test if that would be the case in our systems.[31–33]
Thus, glycosylations carried out employing both glucose donors 3
with opposite stereochemistry at the sulfur atom confirmed that
identical sulfonium adducts, with the same stereomeric ratio, are
formed in both cases (Figure S9). On the contrary, -mannose
donors (strongly affected by the modulatory influence of the axial
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 4. NMR detection of key intermediates in glycosylations with 13C-labelled donors. Reactions carried out with acceptor a and diastereomeric sulfoxides (R)-1
(up) and (S)-1 (down). In the latter case, the influence of the temperature and the number of acceptor equivalents was also tested (yellow and green arrows,
respectively). HSQC spectra (anomeric region) acquired 5-10 min after activation of the different reaction mixtures are shown. Conversions for - and -glycosylation
products after 60 min of evolution are indicated.
substituent at position C-2) lead to rather different ratios of the
same activated sulfur species, being one of them barely
detectable in one of the cases (compound (S)-1, Figure 4).[34]
Moreover, glycosylations performed with either of the sulfoxide
stereomers separately showed that the formation of
donor/acceptor sulfonium adducts is greatly facilitated from one
of the isomers with respect to the other (Figure 4). Indeed,
reactions with donor (R)-1 proceed with significantly larger
transient concentrations of the sulfonium intermediates, ultimately
leading to somewhat decreased reaction yields when a moderate
excess of acceptor alcohol is employed.
counterion already positioned within the coordination sphere of
the anomeric centre (Fig. S10) Finally, regarding the influence of
temperature on the reaction outcome, the obtained data show that
detected steady glycosyl triflate concentrations gradually
decrease in the -60 to -40 °C range, consistent with the more
limited stability of these intermediates. On the contrary, covalent
donor/acceptor sulfonium adducts remain relatively minor (<10%),
yet constant.
The experimental reaction parameters tested, such as the initial
acceptor/donor ratio or the temperature at which the reaction is
performed, both seem to have a minor influence on the steady
concentrations of sulfonium species: while more acceptor
translates into a higher build-up of the donor/acceptor adducts,
these eventually react to yield the desired glycosylation products
thanks to the excess of alcohol employed. Similarly, higher
reaction temperatures have almost no effect on the concentration
of sulfonium adducts, but translate into lower steady
concentrations of the corresponding glycosyl triflates. Finally, the
chirality of the sulfur atom at the initial sulfoxide can pose a
different reactivity profile for each epimer, as in manno-derivative
1, or have no impact at all, as in gluco-donor 3.
Based on these results, we wondered whether the identified
reactivity trends were maintained for more reactive glycosyl
donors. To complete the scope of this study, the in situ activation
methodology was also applied to 2-deoxy--D-glucosyl sulfoxide
(4), which due to the lack of an electron-withdrawing substituent
at the neighbouring C-2 position, generally exhibits a more
reactive profile than the corresponding oxygenated counterparts
(Figure 5). In a first instance, the reaction mixtures of donor 4
against the four different acceptors contemplated in this study (a-
d) were monitored upon activation with triflic anhydride. To our
surprise, no significant build-up of any of the reaction
intermediates already seen for donors 1-3 (sulfonium salts or
glycosyl triflates) was apparent regardless of the alcohol
employed. Instead, a new set of intermediates was instantly
detected and identified, according to their spectroscopic features,
as the corresponding - and -phenyl sulfenates, 4-Sf/4-Sf,
which somehow did not seem to hamper the formation of
glycosylation products over time, thus probably acting as glycosyl
Next, the influence of other experimental parameters, such as the
donor/acceptor ratio or the temperature, on the extent by which
reactive intermediates are formed throughout the reaction course
and the resulting outcome, was also evaluated (Figure 4). As
expected, increased fractions of acceptor alcohol translate into a
slightly larger accumulated concentration of the sulfonium
intermediates, with a concomitant reduction in that of the
corresponding glycosyl triflate. However, under these
circumstances the acceptor is not limiting, and consequently, the
reactions afford somewhat improved yields in the end.
Interestingly, the observed / stereoselectivities are, in all cases,
in the 1:4-1:5 range, almost identical to those obtained with a pre-
activation protocol (Figure 3c). As a result, the direct nucleophilic
attack of the acceptor to the reactive sulfonium species to furnish
the -glycosylation product is not operative even in the presence
of a 7-fold excess of the alcohol. Instead, reactive donor/acceptor
sulfonium adducts seem to preferentially react with the more
nucleophilic triflate anions[23] to yield the corresponding -glycosyl
triflates. The dominance of this process is probably facilitated by
the reduced reactivity and more discriminative character of the
detected sulfonium species, and by the fact that a cationic
anomeric leaving group will favourably interact with the negatively
charged triflate acting as counterion, whereas such coulombic
attraction cannot occur with the neutral acceptor. Intriguingly, the
addition of exogenous tetra-butylammonium triflate (TBAOTf) to
the activated reaction mixture does not perturb the sulfonium
adduct/glycosyl triflate ratio in favour of the second, possibly
signalling the preferential attack of the endogenous triflate
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 5. a) Glycosylations carried out with 2-deoxy--D-glucosyl donor (4, 10 mM) and acceptors a (1.3 equiv) or b-d (4 equivalents) at -65 ºC monitored by HSQC
experiments. Three snapshots for each reaction are shown. Assignments for key anomeric signals (starting material, intermediates and products) are displayed.
Final yields are indicated in cyan. b) Time evolution of an equimolecular 3/4 mixture (10 mM each) in the presence of TFE (left) or MFE (right) upon activation with
triflic anhydride at -65 ºC. Four equivalents of alcohol and three equivalents of DTBMP per donor were employed. Assignments for key anomeric signals (starting
material, intermediates and products) and final yields are indicated.
triflate reservoirs or as true glycosyl donors by their own merit
(Figures 5a and S11-S15). Formation of anomeric
phenylsulfenates has previously been reported in the course of
the glycosylation of activated sulfoxide donors by Khane and col.,
which showed that these intermediates, being stable at low
temperatures, could be further activated and coupled with
acceptors at warmer temperatures.[18] Of note, identical mixtures
of activated species, with similar / ratios, also formed under
pre-activation conditions (data not shown). To check the relative
glycosylation rate of 2-deoxy-glucosyl sulfoxide in comparison
with its 2-oxygenated analogue, competition experiments were
designed whereby donors 2 and 4 were simultaneously reacted
against TFE or MFE (acceptors b and c, respectively) (Figure 5b
and S16). By doing so, we were able to observe that indeed the
2-deoxy donor reacts with no apparent accumulation of any
reaction intermediates, unlike glucose donor 2, which ultimately
translated into a shorter reaction time and a better yield for the
former donor species, a trend that is further accentuated for more
nucleophilic alcohols, such as MFE. Overall, glycosylations
involving highly reactive donors, such as 2-deoxy-glycosyl
sulfoxides, seem to evolve through a differing reaction pathway
that entails glycosyl sulfenates as the only detectable
intermediates. However, it should be pointed out that this
observation does not rule out the participation of glycosly triflates,
or the derived close-ion-pairs, as key reactive species. As
expected, the glycosylations performed with donor 4 in this study
maintained neither the -stereoselectivity typical of manno-donor
1 nor the -selectivity of glucose donor 3, regardless of the
alcohol employed, which were generally able to sway the /
stereoselectivity from 2.5:1 with TFE to its opposite, 1:2.5 with 2-
propanol.[10a]
In summary, the one-step glycosylation protocol works best for
reactions involving weak acceptors, or donors less prone to form
and stabilise sulfonium adducts, such as -sulfoxides or 2- deoxy-
derivatives. The opposite scenario is presented by donors with -
D-gluco configuration, whose performance is poor even with
moderate nucleophiles, despite forming more reactive glycosyl
triflates. This counterintuitive observation is due to the effective-
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Figure 6. Summarized conclusions indicating the recommended glycosylation protocol for each model donor/acceptor pair.
ness of the mechanistic kinetic trap herein identified for this
particular example. In these cases, the alternative two-step pre-
activation protocol provides a much better alternative. To sum up,
all the experimental findings previously commented about have
been gathered in the following chart that is aimed at managing a
better informed choice of glycosylation protocol depending on the
structural and electronic nature of both donor and acceptor
(Figure 6).
As a corollary, it should be noted that most disaccharide-forming
reactions will fall within the acceptor range loosely illustrated by
the left half of the table (Figure 6), defined by moderately poor
and/or relatively hindered acceptors. Thus, according to our
collective data, the single-step activation variant of the sulfoxide
glycosylation presents significant experimental advantages over
the more widely spread pre-activation protocol for particular
donor/acceptor combinations. In this regard, the employment of
the in situ activation protocol herein studied should be definitively
considered a robust, yet simpler method for the efficient
construction of glycosidic linkages for 2-deoxy-glycosyl sulfoxide
donors.
by-standing glycosyl-intermediate must eventually transform,
even if transiently, into the corresponding glycosyl-triflate for the
glycosylation to come to fruition. Certainly, the formation of
covalent donor/acceptor sulfonium adducts has been identified as
the main competing reaction, which determines an irreversible
non-productive consumption of the alcohol and consequently has
the potential to curtail the reaction yield. The prevalence of such
species, which can pose a kinetic trap for the glycosylation to
proceed, seems to mainly depend on the structure of the donor
and the electronic nature of the acceptor, although other
experimental factors have been shown to have a minor influence,
too; to circumvent this issue, single-step glycosylations should be
given enough time and, whenever possible, higher reaction
temperatures. Moreover, the use of super-stoichoimetric amounts
of acceptor alcohol, while economically inconvenient, can also
help alleviate this problem. The extent of this detrimental pathway
and its progression as a function of different experimental and
structural parameters has provided valuable guidelines for the
selection of an optimal glycosylation protocol for each
donor/acceptor dyad.
Conclusion
Experimental Section
General experimental methods: All solvents and reagents were obtained
commercially and used as received unless stated otherwise. Residual
water was removed from starting compounds by repeated coevaporation
with toluene. Reactions were carried out at ambient temperature unless
stated otherwise. All moisture-sensitive reactions were performed in dry
flasks fitted with glass stoppers or rubber septa under a positive pressure
of argon. Air- and moisture-sensitive liquids and solutions were transferred
by syringe or stainless steel cannula. Anhydrous MgSO4 or Na2SO4 were
used to dry organic solutions during workup, and evaporation of the
solvents was performed under reduced pressure using a rotary evaporator.
Flash column chromatography was performed using 230–400 mesh silica
gel. Thin-layer chromatography was conducted on Kieselgel 60 F254. 1H-
and 13C-NMR spectra were recorded in CDCl3 at 300, 400 or 500 MHz and
75, 101 or 126 MHz, respectively. Chemical shifts are expressed in parts
per million (δ scale) downfield from tetramethylsilane and are referenced
In conclusion, our results show that glycosyl triflate intermediates
remain the key players in glycosylations involving glycosyl
sulfoxides, even when donor activation is performed in the
presence of the acceptor alcohol. Despite the fact that the
observed steady concentration for this intermediate can be
influenced by a variety of structural and experimental factors, the
similar stereoselectivity obtained for the same donor/acceptor
pairs regardless of the protocol employed rules out the
participation of other activated species downstream the reaction
pathway (particularly at the critical stereoselectivity-determining
step) and consolidates the role of glycosyl triflates as ubiquitous
true donors whenever triflate anions partake in the reaction
mixture. This is due to the fact that almost any other preceding or
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
to residual protium in the NMR solvent (CHCl3: δ 7.26 ppm). High
Resolution Mass spectra were recorded by direct injection with an
Accurate Mass Q-TOF LC/MS spectrometer equipped with an electrospray
ion source in positive mode.
For the preparation of donors 1-4 and their 13C-labelled versions see
Supporting Information.
H-5´), 1.51 (s, 3H, CH3), 1.45 (s, 3H, CH3), 1.33 (s, 3H, CH3), 1.32 (s, 3H,
CH3); 13C-NMR (126 MHz, CDCl3) δ 137.6 (C, Ar), 129.0 (CH, Ar), 128.3
(2 CH, Ar), 126.2 (2 CH, Ar) 109.6 (O-C-O), 108.8 (O-C-O), 102.7 (CH-1´),
101.7 (O-CH-O), 96.4 (CH-1), 79.9 (CH-2´), 78.7 (CH-3´), 78.7 (CH-4´),
71.7 (CH-4), 70.8 (CH-3), 70.6 (CH-2), 70.2 (CH2-6), 68.6 (CH2-6´), 68.3
(CH-5), 67.3 (CH-5´), 62.2 (OCH3), 58.8 (OCH3), 26.1 (CH3), 26.0 (CH3),
25.1 (CH3), 24.5 (CH3). HRMS (ESI) m/z calcd for C27H38NaO11 [M+Na]+:
561.23063, found: 561.23013.
General Procedures:
Procedure I. General method for the glycosylation experiments with
sulfoxide donors 1-4 and acceptor a: To an oven-dried 50 mL round-
bottomed flask containing the corresponding glycosyl sulfoxide (1.0 mmol),
equipped with a stirring bar, was added 4 Å m.s. (100% w/w), DTBMP (3
mmol), and 1,2:3,4-di-O-isopropylidene- -D-galactose (1.3 mmol), and
then, the flask was purged three times with vacuum/Ar. Next, the mixture
was dissolved in dry DCM (25 mL) under Ar atmosphere and cooled down
to the designated temperature with an acetone/dry ice bath. The triflic
anhydride (1.2 mmol) was injected into the solution and the resulting
mixture was stirred during 1 h, at which point TEA (0.1 mL) was added,
and the quenched reaction mixture was allowed to warm up to room
temperature, filtered and evaporated under vacuum. The resulting crude
material was purified through a silica gel column chromatography.
Synthesis of 2,2,2-trifluoroethyl 4,6-O-benzylidene-2,3-di-O-methyl-
-D-mannopyranoside (1b-/ and 2b-/): Glycosyl sulfoxide 1 (205
mg, 0.51 mmol) and 2,2,2-trifluoroethanol (0.15 mL, 2.03 mmol) were
coupled following the general procedure II (-45 ºC, 1 h). After work up, the
residue was purified by flash silica gel chromatography (hexane/ethyl
acetate 9:1 to 7:3) to give glycoside 1b- (59 mg, 31 %) along with
glycoside 1b- (94 mg, 49 %).
Similarly, glycosyl sulfoxide 2 (167 mg, 0.41 mmol) and and 2,2,2-
trifluoroethanol (0.12 mL, 1.65 mmol) were also coupled following the
general procedure II (-50 ºC, 1 h). After work up, the residue was purified
by flash silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
glycoside 1b- (42 mg, 27 %) along with glycoside 1b- (82 mg, 53 %).
1
For 1b-: H-NMR (400 MHz, CDCl3) δ 7.68-7.62 (m, 3H, Ar), 7.50-7.42
Procedure II. General method for the glycosylation experiments with
sulfoxide donors 1-4 and acceptors b-d: To an oven-dried 50 mL round-
bottomed flask containing the corresponding glycosyl sulfoxide (1.0 mmol),
equipped with a stirring bar, was added 4 Å m.s. (100% w/w) and DTBMP
(3 mmol), and then the flask was purged three times with vacuum/Ar. Next,
the mixture was dissolved in dry DCM (25 mL) under Ar atmosphere and
cooled down to the designated temperature with an acetone/dry ice bath.
The acceptor (4 mmol) was injected into the mixture, followed by a
dropwise addition of triflic anhydride (1.2 mmol). The resulting mixture was
stirred during 1 h, at which point TEA (0.1 mL) was added, and the
quenched reaction mixture was allowed to warm up to room temperature,
filtered and evaporated under vacuum. The resulting crude material was
purified through a silica gel column chromatography.
(m, 2H, Ar), 5.59 (s, 1H, O-CH-O), 4.97 (d, J = 1.5 Hz, 1H, H-1), 4.24 (dd,
J = 9.9, 4.1 Hz, 1H, H-6eq), 4.09 (t, J = 9.5 Hz, 1H, H-4), 4.00-3.92 (m, 2H,
CH2-CF3), 3.84 (t, J = 9.9 Hz, 1H, H-6ax), 3.81-3.76 (m, 1H, H-5), 3.73 (m,
2H, H-2, H-3), 3.57 (s, 3H, OCH3), 3.56 (s, 3H, OCH3); 13C-NMR (101 MHz,
CDCl3) δ 145.7 (C, Ar), 131.2 (CH, Ar), 130.0 (q, J = 280 Hz, CH2-CF3),
129.4 (2 CH, Ar), 124.9 (2 CH, Ar), 101.8 (O-CH-O), 99.1 (CH-1), 78.9
(CH-4), 78.3 (CH-2), 77.4 (CH-3), 68.6 (CH2-6), 64.8 (CH-5), 64.4 (q, J =
64.0 Hz, CH2-CF3), 60.1 (OCH3), 59.4 (OCH3). HRMS (ESI) m/z calcd for
1
C17H21F3NaO6 [M+Na]+: 401.11824, found: 401.1189. For 1b-: H-NMR
(500 MHz, CDCl3) δ 7.52-7.42 (m, 2H, Ar), 7.42-7.29 (m, 3H, Ar), 5.57 (s,
1H, O-CH-O), 4.64 (br.s, 1H, H-1), 4.32 (dd, J = 10.3, 4.9 Hz, 1H, H-6eq),
4.20 (dq, J = 12.6, 8.8 Hz, 1H, CH2-CF3), 4.05 (t, J = 9.9 Hz, 1H, H-4),
4.02-3.94 (m, 1H, CH2-CF3), 3.91 (t, J = 10.3 Hz, 1H, H-6ax), 3.82 (dd, J
= 3.1, 1.0 Hz, 1H, H-2), 3.67 (s, 3H, OCH3), 3.57 (s, 3H, OCH3), 3.43 (dd,
J = 9.9, 3.1 Hz, 1H, H-3), 3.35 (ddd, J = 10.3, 9.9, 4.9 Hz, 1H, H-5); 13C-
NMR (126 MHz, CDCl3) δ 137.4 (C, Ar), 129.1 (CH, Ar), 128.3 (2 CH, Ar),
126.2 (2 CH, Ar), 123.8 (q, J = 280.5 Hz, CF3), 101.80 (CH-1), 101.78 (O-
CH-O), 79.9 (CH-3), 78.5 (CH-2), 78.4 (CH-4), 68.4 (CH2-6), 67.7 (CH-5),
66.1 (q, J = 34.8 Hz, CH2-CF3), 62.3 (OCH3), 59.1 (OCH3). HRMS (ESI)
m/z calcd for C17H21F3NaO6 [M+Na]+: 401.11824, found: 401.11834.
Synthesis
mannopyranosyl-(1
of
4,6-O-benzylidene-2,3-di-O-methyl-/-D-
6)-1,2:3,4-di-O-isopropylidene--D-
→
galactopyranose (1a-/): Glycosyl sulfoxide 1 (153 mg, 0.38 mmol) and
1,2:3,4-di-O-isopropylidene--D-galactose (138 mg, 0.49 mmol) were
coupled following the general procedure I (-50 ºC, 1 h). After work up, the
residue was purified by flash silica gel chromatography (hexane/ethyl
acetate 8:2 to 7:3) to give glycoside 1a- (12 mg, 6 %) along with glycoside
1a- (79 mg, 39 %).
Synthesis of 2-fluoroethyl 4,6-O-benzylidene-2,3-di-O-methyl-/-D-
mannopyranoside (1c-/): Glycosyl sulfoxide 1 (182 mg, 0.45 mmol)
and 2-fluoroethanol (0.10 mL, 1.80 mmol) were coupled following the
general procedure II (-45 ºC, 1 h). After work up, the residue was purified
by flash silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
glycoside 1c- (23 mg, 15 %) along with glycoside 1c- (94 mg, 61 %).
Similarly, glycosyl sulfoxide 2 (174 mg, 0.43 mmol) and 2-fluoroethanol
(0.10 mL, 1.72 mmol) were also coupled following the general procedure
II (-50 ºC, 1 h). After work up, the residue was purified by flash silica gel
chromatography (hexane/ethyl acetate 9:1 to 7:3) to give glycoside 1c-
(10 mg, 7 %) along with glycoside 1c- (100 mg, 68 %). For 1c-: 1H-NMR
(500 MHz, CDCl3) δ 7.51-7.46 (m, 2H, Ar), 7.38-7.30 (m, 3H, Ar), 5.59 (s,
1H, O-CH-O), 4.94 (d, J = 1.6 Hz, 1H, H-1), 4.67-4.52 (m, 2H, CH2-F),
4.27-4.20 (m, 1H, H-6a), 4.08 (t, J = 10.1 Hz, 1H, H-4), 3.96– 3.81 (m, 1H,
CH2-CH2-F), 3.85-3.82 (m, 2H, H-6b, H-5), 3.79-3.71 (m, 1H, CH2-CH2-F ),
3.76 (dd, J = 10.1, 3.4 Hz, 1H, H-3), 3.71 (dd, J = 3.4, 1.6 Hz, 1H, H-2),
3.57 (s, 3H, OCH3), 3.56 (s, 3H, OCH3); 13C-NMR (126 MHz, CDCl3) δ
137.7 (C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 101.8
(O-CH-O), 98.8 (CH-1), 82.6 (d, J = 170.1 Hz, CH2-F), 79.2 (CH-4), 78.7
(CH-2), 77.7 (CH-3), 68.9 (CH2-6), 66.9 (d, J = 19.7 Hz, CH2-CH2-F), 64.3
(CH-5), 60.0 (OCH3), 59.3 (OCH3). HRMS (ESI) m/z calcd for C17H24FO6
[M+H]+: 343.15514, found: 343.15288. For 1c-: 1H-NMR (500 MHz,
CDCl3) δ 7.52-7.43 (m, 2H, Ar), 7.40-7.30 (m, 3H, Ar), 5.56 (s, 1H, O-CH-
O), 4.71-4.50 (m, 2H, CH2-F), 4.59 (br.s, 1H, H-1), 4.30 (dd, J = 10.3, 4.9
Hz, 1H, H-6eq), 4.13-4.02 (m, 2H, CH2-CH2-F), 4.04 (t, J = 10.3 Hz, 1H,
H-4), 3.89 (t, J = 10.3 Hz, 1H, H-6ax), 3.82 (d, J = 3.1 Hz, 1H, H-2), 3.68
(s, 3H, OCH3), 3.56 (m, 3H, OCH3), 3.43 (dd, J = 10.3, 3.1 Hz, 1H, H-3),
Similarly, glycosyl sulfoxide 2 (140 mg, 0.34 mmol) and 1,2:3,4-di-O-
isopropylidene-D-galactose (126 mg, 0.45 mmol) were also coupled
following the general procedure I (-50 ºC, 1 h). After work up, the residue
was purified by flash silica gel chromatography (hexane/ethyl acetate 8:2
to 7:3) to give glycoside 1a- (17 mg, 9 %) along with glycoside 1a- (108
mg, 59 %). For 1a-: 1H-NMR (500 MHz, CDCl3) δ 7.51-7.44 (m, 2H, Ar),
7.38-7.30 (m, 3H, Ar), 5.58 (s, 1H, O-CH-O), 5.54 (d, J = 5.0 Hz, 1H, H-1),
4.97 (d, J = 1.6 Hz, 1H, H-1´), 4.64 (dd, J = 7.9, 2.4 Hz, 1H, H-3), 4.34 (dd,
J = 5.0, 2.4 Hz, 1H, H-2), 4.26-4.23 (m, 2H, H-4, H-6a´), 4.06 (t, J = 9.2 Hz,
1H, H-4´), 4.03-3.99 (m, 1H, H-5), 3.84-3.79 (m, 3H, H-6b´, H-5´, H-6a),
3.77-3.71 (m, 2H, H-6b, H-3´), 3.68 (dd, J = 3.3, 1.6 Hz, 1H, H-2´), 3.55 (s,
3H, OCH3), 3.54 (s, 3H, OCH3), 1.55 (s, 3H, CH3), 1.46 (s, 3H, CH3), 1.35
(s, 3H, CH3), 1.34 (s, 3H, CH3); 13C-NMR (126 MHz, CDCl3) δ 137.8 (C,
Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.3 (2 CH, Ar), 109.6 (O-C-O),
108.8 (O-C-O), 101.8 (O-CH-O), 98.2 (CH-1´), 96.5 (CH-1), 79.3 (CH-4´),
79.0 (CH-2´), 77.7 (CH-3´), 71.2 (CH-4), 70.9 (CH-3), 70.8 (CH-2), 68.9
(CH2-6´), 66.2 (CH2-6), 65.7 (CH-5), 64.3 (CH-5´), 59.8 (OCH3), 59.2
(OCH3), 26.3 (CH3), 26.1 (CH3), 25.1 (CH3), 24.7 (CH3). HRMS (ESI) m/z
calcd for C27H39O11 [M+H]+: 539.24869, found: 539.25069. For 1a- : 1H-
NMR (300 MHz, CDCl3) δ 7.50-7.41 (m, 2H, Ar), 7.39-7.29 (m, 3H, Ar),
5.54 (s, 1H, O-CH-O), 5.50 (d, J = 5.0 Hz, 1H, H-1), 4.62-4.58 (m, 2H, H-
1´, H-3), 4.33-4.30 (m, 1H, H-2), 4.29 (dd, J = 10.5, 4.9 Hz, 1H, H-6eq´),
4.19 (dd, J = 7.9, 1.8 Hz, 1H, H-4), 4.14 (dd, J = 11.1, 2.3 Hz, 1H, H-6a),
4.03 (t, J = 9.4 Hz, 1H, H-4´), 4.08-3.94 (m, 1H, H-5), 3.94-3.81 (m, 2H, H-
3´ H-6ax´), 3.73-3.61 (m, 1H, H-6b), 3.68 (s, 3H, OCH3), 3.54 (s, 3H,
OCH3), 3.41 (dd, J = 9.9, 3.2 Hz, 1H, H-2´), 3.32 (td, J = 9.7, 4.9 Hz, 1H,
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
3.35 (td, J = 10.3, 4.9 Hz, 1H, H-5); 13C-NMR (126 MHz, CDCl3) δ 137.5
(C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 102.2 (CH-1),
101.7 (O-CH-O), 82.9 (d, J = 169.7 Hz, CH2-F), 80.1 (CH-3), 78.8 (CH-2),
78.7 (CH-4), 68.9 (d, J = 19.7 Hz, CH2-CH2-F), 68.6 (CH2-6), 67.49 (CH-
5), 62.3 (OCH3), 59.0 (OCH3). HRMS (ESI) m/z calcd for C17H23FNaO6
[M+Na]+: 365.13709, found: 365.13753.
by flash silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
exclusively glycoside 3b- (110 mg, 58 %). 1H-NMR (300 MHz, CDCl3) δ
7.53-7.42 (m, 3H, Ar), 7.41-7.30 (m, 2H, Ar), 5.54 (s, 1H, O-CH-O), 5.05
(d, J = 3.8 Hz, 1H, H-1), 4.27 (dd, J = 9.8, 4.6 Hz, 1H, H-6eq), 4.06-3.95
(m, 2H, CH2-CF3), 3.89-3.81 (m, 1H, H-5), 3.72 (t, J = 9.8 Hz, 1H, H-6ax),
3.71 (t, J = 9.2, 1H, H-3) 3.65 (s, 3H, OCH3), 3.54 (s, 3H, OCH3), 3.33 (t,
J = 9.2 Hz, 1H, H-4), 3.33 (dd, J = 9.2, 3.8, Hz, 1H, H-2); 13C-NMR (101
MHz, CDCl3) δ 145.4 (C, Ar), 131.2 (CH, Ar), 129.5 (2 CH, Ar), 123.8 (q, J
= 277.1 Hz, CF3), 124.9 (2 CH, Ar), 101.5 (O-CH-O), 97.2 (CH-1), 81.9
(CH-4), 81.0 (CH-2), 79.3 (CH-3), 68.8 (CH2-6), 64.7 (q, J = 35.0 Hz, CH2-
CF3), 63.1 (CH-5), 61.2 (OCH3), 59.1 (OCH3). HRMS (ESI) m/z calcd for
C17H21F3NaO6 [M+Na]+: 401.11879, found: 401.12319.
Synthesis of 2-propyl 4,6-O-benzylidene-2,3-di-O-methyl-/-D-
mannopyranoside (1d-/): Glycosyl sulfoxide 1 (158 mg, 0.39 mmol)
and 2-propanol (0.12 mL, 1.56 mmol) were coupled following the general
procedure II (-45 ºC, 1 h). After work up, the residue was purified by flash
silica gel chromatography (hexane/ethyl acetate 9:1 to 8:2) to give
glycoside 1d- (5 mg, 4 %), along with glycoside 1d- (20 mg, 15 %).
Similarly, glycosyl sulfoxide 2 (123 mg, 0.30 mmol) and 2-propanol (0.09
mL, 1.22 mmol) were also coupled following the general procedure II (-50
ºC, 1 h). After work up, the residue was purified by flash silica gel
chromatography (hexane/ethyl acetate 9:1 to 8:2) to give exclusively
glycoside 1d-(21 mg, 20 %). For 1d-: 1H-NMR (400 MHz, CDCl3) δ
7.54-7.44 (m, 2H, Ar), 7.41-7.29 (m, 3H, Ar), 5.59 (s, 1H, O-CH-O), 4.98
(d, J = 1.7 Hz, 1H, H-1), 4.27-4.17 (m, 1H, H-6a), 4.06 (t, J = 9.0 Hz, 1H,
H-4), 3.93 (hept, J = 6.2 Hz, 1H, CH3-CH-CH3), 3.86-3.80 (m, 2H, H-5, H-
6b), 3.74 (dd, J = 9.9, 3.3 Hz, 1H, H-3), 3.59 (dd, J = 3.3, 1.7 Hz, 1H, H-2),
3.57 (s, 3H, OCH3), 3.55 (s, 3H, OCH3), 1.23 (d, J = 6.2 Hz, 3H, CH3-CH-
CH3), 1.17 (d, J = 6.1 Hz, 3H, CH3-CH-CH3); 13C-NMR (101 MHz, CDCl3)
δ 137.7 (C, Ar), 128.9 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 101.6
(O-CH-O), 96.5 (CH-1), 79.53 (CH-2), 79.50 (CH-4), 77.8 (CH-3), 69.6
(CH3-CH-CH3), 69.0 (CH2-6), 64.1 (CH-5), 59.9 (OCH3), 59.2 (OCH3), 23.4
(CH3-CH-CH3), 21.4 (CH3-CH-CH3). HRMS (ESI) m/z calcd for
C18H26NaO6 [M+Na]+: 361.16092, found: 361.16216. For 1d-: 1H-NMR
(400 MHz,CDCl3) δ 7.56-7.38 (m, 3H, Ar), 7.37 (s, 2H, Ar), 5.55 (s, 1H, O-
CH-O), 4.59 (d, J = 1.0 Hz, 1H, H-1), 4.29 (dd, J = 10.5, 4.9 Hz, 1H, H-
6eq), 4.04 (t, J = 9.6 Hz, 1H, H-4), 4.01 (hept, J = 6.3 Hz, 1H, CH3-CH-
CH3), 3.90 (t, J = 10.3 Hz, 1H, H-6ax), 3.69 (dd, J = 3.2, 1.0 Hz, 1H, H-2),
3.67 (s, 3H, OCH3), 3.55 (s, 3H, OCH3), 3.42 (dd, J = 9.9, 3.2 Hz, 1H, H-
3), 3.33 (ddd, J = 10.0, 9.2, 4.9 Hz, 1H, H-5), 1.26 (d, J = 6.2 Hz, 3H, CH3-
CH-CH3), 1.19 (d, J = 6.1 Hz, 3H, CH3-CH-CH3); 13C-NMR (101 MHz,
CDCl3) δ 137.6 (C, Ar), 128.9 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar),
101.7 (O-CH-O), 100.1 (CH-1), 80.3 (CH-3), 79.7 (CH-2), 78.8 (CH-4),
71.4 (CH3-CH-CH3), 68.7 (CH2-6), 67.4 (CH-5), 62.2 (OCH3), 58.9 (OCH3),
23.6 (CH3-CH-CH3), 21.7 (CH3-CH-CH3). HRMS (ESI) m/z calcd for
C18H26NaO6 [M+Na]+: 361.16092, found: 361.16164.
Synthesis of 2-fluoroethyl 4,6-O-benzylidene-2,3-di-O-methyl-/-D-
glucopyranoside (3c-/): Glycosyl sulfoxide 3 (153 mg, 0.38 mmol) and
2-fluoroethanol (0.09 mL, 1.51 mmol) were coupled following the general
procedure II (-60 ºC, 1 h). After work up, the residue was purified by flash
silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
glycoside 3c- (26 mg, 20 %) along with glycoside 3c- (6 mg, 5 %). For
3c-:1H-NMR (400 MHz, CDCl3) δ 7.54 -7.41 (m, 2H, Ar), 7.39-7.32 (m,
3H, Ar), 5.54 (s, 1H, O-CH-O), 5.03 (d, J = 3.8 Hz, 1H, H-1), 4.62 (dt, J =
47.6, 4.4 Hz, 2H, CH2-F), 4.27 (dd, J = 10.2, 4.8 Hz, 1H, H-6eq), 3.99-3.76
(m, 3H, CH2-CH2-F, H-5), 3.72 (t, J = 9.3 Hz, 1H, H-4), 3.71 (t, J = 10.2 Hz,
1H, H-6ax), 3.65 (s, 3H, OCH3), 3.55 (s, 3H, OCH3), 3.54 (t, J = 9.3 Hz,
1H, H-3), 3.32 (dd, J = 9.3, 3.8 Hz, 1H, H-2); 13C-NMR (101 MHz, CDCl3)
δ 137.4 (C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 101.4
(O-CH-O), 97.6 (CH-1), 82.5 (d, J = 169.8 Hz, CH2-F), 82.3 (CH-3) 81.3
(CH-2), 79.8 (CH-4), 69.1 (CH2-6), 67.2 (d, J = 20.2 Hz, CH2-CH2-F), 62.5
(CH-5), 61.1 (OCH3), 59.1 (OCH3). HRMS (ESI) m/z calcd for
C17H23FNaO6 [M+Na]+: 365.13709, found: 365.13708. For 3c-1H-NMR
(400 MHz, CDCl3) δ 7.50-7.48 (m, 2H, Ar), 7.42-7.29 (m, 3H, Ar), 5.54 (s,
1H, O-CH-O), 4.70-4.49 (m, 2H, CH2-F), 4.45 (d, J = 7.6 Hz, 1H, H-1), 4.32
(dd, J = 10.3, 5.0 Hz, 1H, H-6eq), 4.07 (dddd, J = 31.3, 12.1, 4.6, 3.0 Hz,
1H, CH2-CH2-F), 3.86 (dddd, J = 26.5, 12.2, 5.9, 3.4 Hz, 1H, CH2-CH2-F),
3.76 (t, J = 10.3 Hz, 1H, H-6ax), 3.64 (s, 3H, OCH3), 3.62 (s, 3H, OCH3),
3.57 (t, J = 9.3 Hz, 1H, H-4), 3.39 (t, J = 9.2 Hz, 1H, H-3), 3.42-3.35 (m,
1H, H-5), 3.12 (dd, J = 9.2, 7.6 Hz, 1H, H-2); 13C-NMR (101 MHz, CDCl3)
δ 137.4 (C, Ar), 129.1 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 104.2
(CH-1), 101.3 (O-CH-O), 83.8 (CH-2), 84.3 (d, J = 170.4 Hz, CH2-F), 82.6
(CH-3), 81.3 (CH-4), 69.2 (d, J = 20.1 Hz, CH2-CH2-F), 68.8 (CH2-6), 66.1
(CH-5), 61.1 (OCH3), 61.0 (OCH3). HRMS (ESI) m/z calcd for C17H24FO6
[M+H]+: 343.15514, found: 343.15362.
Synthesis of 4,6-O-benzylidene-2,3-di-O-methyl--D-glucopyranosyl-
(1
→
6)-1,2:3,4-di-O-isopropylidene--D-galactopyranose (3a-):
Synthesis of 2-propyl 4,6-O-benzylidene-2,3-di-O-methyl-/-D-
glucopyranoside (3d-/): Glycosyl sulfoxide 3 (105 mg, 0.26 mmol) and
2-propanol (0.08 mL, 1.04 mmol) were coupled following the general
procedure II (-60 ºC, 1 h). After work up, the residue was purified by flash
silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
glycoside 3d- (17 mg, 19 %) along with glycoside 3d- (7 mg, 8 %). For
Glycosyl sulfoxide 3 (142 mg, 0.35 mmol) and 1,2:3,4-di-O-isopropylidene-
-D-galactose (128 mg, 0.46 mmol) were coupled following the general
procedure I (-60 ºC, 1 h). After work up, the residue was purified by flash
silica gel chromatography (hexane/ethyl acetate 8:2 to 7:3) to give
exclusively glycoside 3a- (53 mg, 28 %). 1H-NMR (500 MHz, CDCl3) δ
7.54-7.46 (m, 2H, Ar), 7.39-7.30 (m, 3H, Ar), 5.54 (s, 1H, O-CH-O), 5.52
(d, J = 5.0 Hz, 1H, H-1), 5.05 (d, J = 3.7 Hz, 1H, H-1´), 4.62 (dd, J = 8.0,
2.4 Hz, 1H, H-3), 4.33 (dd, J = 8.0, 1.9 Hz, 1H, H-4), 4.31 (dd, J = 5.0, 2.4
Hz, 1H, H-2), 4.28 (dd, J = 10.3, 5.1 Hz, 1H, H-6eq´), 4.05 (td, J = 6.7, 1.9
Hz, 1H, H-5), 3.89 (td, J = 9.9, 4.9 Hz, 1H, H-5), 3.82 (dd, J = 10.5, 6.3 Hz,
1H, H-6a), 3.74 (dd, J = 10.4, 7.1 Hz, 1H, H-6b), 3.70 (t, J = 10.3,1H, H-
6ax´), 3.68 (t, J = 9.3 Hz, 1H, H-3´) 3.63 (s, 3H, OCH3), 3.53 (t, J = 9.3 Hz,
1H, H-4´) 3.51 (s, 3H, OCH3), 3.29 (dd, J = 9.3, 3.7 Hz, 1H, H-2´), 1.54 (s,
3H, CH3), 1.44 (s, 3H, CH3), 1.33 (s, 6H, 2 CH3); 13C-NMR (126 MHz,
CDCl3) δ 137.6 (C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar),
109.3 (O-C-O), 108.7 (O-C-O), 101.4 (O-CH-O), 97.6 (CH-1´), 96.4 (CH-
1), 82.3 (CH-4´), 81.3 (CH-2´), 79.4 (CH-3´), 70.9 (CH-2), 70.8 (CH-3),
70.7 (CH-4), 69.1 (CH2-6´), 67.0 (CH2-6), 65.9 (CH-5), 62.4 (CH-5´), 61.1
(OCH3), 58.4 (OCH3), 26.2 (CH3), 26.1 (CH3), 25.0 (CH3), 24.5 (CH3).
HRMS (ESI) m/z calcd for C27H42NO11 [M+NH4]+: 556.27524, found:
556.27532.
1
3d-α: H-NMR (400 MHz, CDCl3) δ 7.64-7.44 (m, 2H, Ar), 7.42-7.30 (m,
3H, Ar), 5.55 (s, 1H, O-CH-O), 5.06 (d, J = 3.8 Hz, 1H, H-1), 4.26 (dd, J =
10.3, 4.8 Hz, 1H, H-6eq), 3.93 (hept, J = 6.3 Hz, 1H, CH3-CH-CH3), 3.71
(t, J = 10.3 Hz, 1H, H-6ax), 3.70 (t, J = 9.3 Hz, 1H, H-3), 3.65 (s, 3H, OCH3),
3.60-3.57 (m, 1H, H-5), 3.53 (t, J = 9.3 Hz, 1H, H-4), 3.52 (s, 3H, OCH3),
3.28 (dd, J = 9.3, 3.8 Hz, 1H, H-2), 1.27 (d, J = 6.3 Hz, 3H, CH3-CH-CH3),
1.22 (d, J = 6.3 Hz, 3H, CH3-CH-CH3); 13C-NMR (101 MHz, CDCl3) δ 137.5
(C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 126.2 (2 CH, Ar), 101.4 (O-CH-
O), 95.3 (CH-1), 82.6 (CH-4), 81.4 (CH-2), 79.6 (CH-3), 69.8 (CH3-CH-
CH3), 69.2 (CH2-6), 62.4 (CH-5), 61.1 (OCH3), 58.8 (OCH3), 23.4 (CH3-
CH-CH3), 21.4 (CH3-CH-CH3). HRMS (ESI) m/z calcd for C18H26NaO6
[M+Na]+: 361.16216, found: 361.16278. For 3d-: 1H-NMR (400
MHz,CDCl3) δ 7.66-7.42 (m, 2H, Ar), 7.42-7.30 (m, 3H, Ar), 5.53 (s, 1H,
O-CH-O), 4.45 (d, J = 7.7 Hz, 1H, H-1), 4.31 (dd, J = 10.3, 5.0 Hz, 1H, H-
6eq), 3.98 (hept, J = 6.1 Hz, 1H, CH3-CH-CH3), 3.76 (t, J = 10.3 Hz, 1H,
H-6ax), 3.63 (s, 3H, OCH3), 3.61 (s, 3H, OCH3), 3.56 (t, J = 9.3 Hz, 1H, H-
4), 3.45-3.28 (m, 2H, H-5, H-3), 3.06 (dd, J = 8.8, 7.7 Hz, 1H, H-2), 1.26
(d, J = 6.1 Hz, 3H, CH3-CH-CH3), 1.23 (d, J = 6.1 Hz, 3H, CH3-CH-CH3);
13C-NMR (101 MHz, CDCl3) δ 137.5 (C, Ar), 129.0 (CH, Ar), 128.3 (2 CH,
Ar), 126.2 (2 CH, Ar), 102.7 (CH-1), 101.3 (O-CH-O), 84.0 (CH-2), 82.8
(CH-3), 81.4 (CH-4), 72.8 (CH3-CH-CH3), 67.0 (CH2-6), 66.0 (CH-5), 61.1
Synthesis of 2,2,2-trifluoroethyl 4,6-O-benzylidene-2,3-di-O-methyl--
D-glucopyranoside (3b-): Glycosyl sulfoxide 3 (202 mg, 0.50 mmol) and
2,2,2-trifluoroethanol (0.14 mL, 1.99 mmol) were coupled following the
general procedure II (-60 ºC, 1 h). After work up, the residue was purified
10
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
(OCH3), 61.0 (OCH3), 23.7 (CH3-CH-CH3), 22.1 (CH3-CH-CH3). HRMS
(ESI) m/z calcd for C18H26NaO6 [M+Na]+: 361.16216, found: 361.16319.
a 1:1.3 mixture of glycosides 4c- and 4c- respectively (46 mg, 67%).
For 4c-:1H-NMR (400 MHz, CDCl3) δ 7.54 – 7.46 (m, 2H, Ar), 7.41 – 7.32
(m, 3H, Ar), 5.59 (s, 1H, O-CH-O), 4.97 (d, J = 3.7 Hz, 1H, H-1), 4.58 (dt,
J = 47.6, 4.2 Hz, 2H, CH2F), 4.24 (dd, J = 10.0, 4.6 Hz, 1H, H-6eq), 3.93 –
3.78 (m, 3H, H-3, 5; CH2-CH2-F), 3.75 (t, J = 10.2 Hz, 1H, H-6ax), 3.72 –
3.64 (m, 1H, CH2-CH2-F), 3.61 (t, J = 9.3 Hz, 1H), 3.51 (s, 3H, OCH3), 2.35
(dd, J = 13.3, 5.1 Hz, 1H, H-2eq), 1.70 (ddd, J = 13.6, 11.2, 3.9 Hz, 1H, H-
2ax); 13C NMR (101 MHz, CDCl3) δ 137.5 (C, Ar), 128.9 (CH, Ar), 128.2 (2
CH, Ar), 126.1 (2 CH, Ar), 101.6 (O-CH-O), 98.2 (CH-1), 83.5 (CH-4), 82.6
(d, J = 169.8 Hz, CH2F), 74.2 (CH-3), 69.0 (CH2-6), 66.4 (d, J = 20.3 Hz,
CH2-CH2-F), 63.0 (CH-5), 58.5 (OCH3), 35.7 (CH2-2). HRMS (ESI) m/z
calcd for C16H21FNaO5 [M+Na]+: 335.12652, found: 335.12727. For 4c-
:1H NMR (400 MHz, CDCl3) δ 7.54 – 7.45 (m, 2H, Ar), 7.40 – 7.32 (m,
3H, Ar), 5.57 (s, 1H, O-CH-O), 4.67 (dd, J = 10.1, 2.2 Hz, 1H, H-1), 4.64
(dd, J = 6.9, 3.3 Hz, 1H, CH2-CH2-F), 4.56 – 4.48 (m, 1H, CH2-CH2-F), 4.32
(dd, J = 10.5, 4.9 Hz, 1H, H-6eq), 4.05 (dddd, J = 33.4, 12.3, 4.7, 2.4 Hz,
1H, CH2F), 3.83 (dddd, J = 24.8, 12.3, 7.1, 2.4 Hz, 1H, CH2F), 3.82 (t, J =
10.5 Hz, 1H, H-6ax), 3.60 (t, J = 8.7 Hz, 1H, H-4), 3.55 (td, J = 10.1, 8.7,
4.7 Hz, 1H, H-3), 3.50 (s, 3H, OCH3), 3.38 (td, J = 9.3, 4.9 Hz, 1H, H-5),
2.43 (ddd, J = 12.9, 4.7, 2.2 Hz, 1H, H-2eq), 1.65 (dt, J = 12.9, 10.3 Hz,
1H, H-2ax); 13C NMR (101 MHz, CDCl3) δ 137.3 (C, Ar), 129.0 (CH, Ar),
128.2 (2 CH, Ar), 126.1 (2 CH, Ar), 101.5 (O-CH-O), 100.5 (CH-1), 82.71
(d, J = 169.5 Hz, CH2F), 82.65 (CH-4), 76.5 (CH-3), 68.8 (CH2-6), 68.3 (d,
J = 19.8 Hz, CH2-CH2-F), 66.6 (CH-5), 58.2 (OCH3), 36.8 (CH2-2). HRMS
(ESI) m/z calcd for C16H22FO5 [M+H]+: 313.14458, found: 313.14441.
Synthesis
glucopyranosyl-(1
of
4,6-O-benzylidene-2-deoxy-3-O-methyl--D-
6)-1,2:3,4-di-O-isopropylidene--D-
→
galactopyranose (4a-): Glycosyl sulfoxide 4 (111 mg, 0.297 mmol)
and 1,2:3,4-di-O-isopropylidene--D-galactose (100 mg, 0.386 mmol)
were coupled following the general procedure I (-60 ºC, 1 h). After work up,
the residue was purified by flash silica gel chromatography (hexane/ethyl
acetate 9:1 to 8:2) to give a 1:2 mixture of glycosides 4a- and 4a-
respectively (62 mg, 41 %). For 4a-α: 1H NMR (400 MHz, CDCl3) (selected
signals) δ 7.54 – 7.44 (m, 2H, Ar), 7.40 – 7.30 (m, 3H, Ar), 5.56 (s, 1H, O-
CH-O), 5.54 (d, J = 5.5 Hz, 1H, H-1), 4.69 (dd, J = 9.6, 2.9 Hz, H-1’), 4.60
(dd, J = 8.1, 2.7 Hz, H-3), 3.48 (s, 3H, OCH3), 3.37 (td, J = 9.0, 4.9 Hz, H-
5’), 2.45 (ddd, J = 12.4, 4.8, 2.6 Hz, H-2’eq), 1.54 (s, 3H, CH3), 1.45 (s, 3H,
CH3), 1.33 (s, 6H, 2 CH3); 13C NMR (101 MHz, CDCl3) δ 137.4 (C, Ar),
128.9 (CH, Ar), 128.2 (2 CH, Ar), 126.2 (2 CH, Ar), 109.4 (O-C-O), 108.7
(O-C-O), 101.5 (O-CH-O), 101.0 (CH-1’), 96.3 (CH-1), 82.7 (CH-4’), 77.2,
76.6, 71.4, 70.7, 70.4, 69.1, 68.9, 67.8 (CH-5’), 58.0 (OCH3), 36.8 (CH2-
2’), 26.0 (CH3), 26.0 (CH3), 25.0 (CH3), 24.4 (CH3). HRMS (ESI) m/z calcd
for C26H37O10 [M+H]+: 509.23812, found: 509.23948. For 4a-: 1H NMR
(400 MHz, CDCl3) δ 7.54 – 7.44 (m, 2H, Ar), 7.40 – 7.30 (m, 3H, Ar), 5.58
(s, 1H, O-CH-O), 5.53 (d, J = 5.2 Hz, 1H, H-1), 4.97 (d, J = 3.7 Hz, 1H, H-
1’), 4.63 (dd, J = 8.1, 3.2 Hz, 1H, H-3), 4.32 (dd, J = 8.1, 5.2 Hz, 1H, H-4),
4.29 – 4.21 (m, 2H, H-2, 6’eq), 4.04 – 3.93 (m, 1H, H-5), 3.90 – 3.73 (m,
4H, H-6a, 6b, 3’, 5’), 3.67 (dd, J = 10.5, 6.9 Hz, 1H, H-6’ax), 3.59 (t, J = 9.5
Hz, 1H, H-4’), 3.50 (s, 3H, OCH3), 2.32 (dd, J = 13.4, 5.0 Hz, 1H, H-2’eq),
1.67 (tdd, J = 12.0, 3.9, 1.2 Hz, 1H, H-2’ax), 1.55 (s, 3H, CH3), 1.45 (s, 3H,
CH3), 1.35 (s, 3H, CH3), 1.34 (s, 3H, CH3); 13C NMR (101 MHz, CDCl3) δ
137.5 (C, Ar), 128.9 (CH, Ar), 128.2 (2 CH, Ar), 126.2 (2 CH, Ar), 109.3
(O-C-O), 108.6 (O-C-O), 101.5 (O-CH-O), 98.1 (CH-1’), 96.3 (CH-1), 83.6
(CH-4’), 74.3 (CH-3’), 71.0 (CH-2), 70.61 (CH-3), 70.59 (CH-4), 69.0 (CH2-
6’), 65.9 (CH2-6), 65.8 (CH-5), 63.0 (CH-5’), 58.4 (OCH3), 35.8 (CH2-2’),
26.1 (CH3), 26.0 (CH3), 24.9 (CH3), 24.5 (CH3). HRMS (ESI) m/z calcd for
C26H40NO10 [M+NH4]+: 526.26467, found: 526.26555.
Synthesis of 2-propyl 4,6-O-benzylidene-2-deoxy-3-O-methyl-/-D-
glucopyranoside (4d-/): Glycosyl sulfoxide 4 (95 mg, 0.254 mmol) and
2-propanol (77 L, 1.016 mmol) were coupled following the general
procedure II (-60 ºC, 1 h). After work up, the residue was purified by flash
silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give a 1:2.2
mixture of glycosides 4d- and 4d- respectively (47 mg, 60%). For 4d-α:
1H-NMR (400 MHz, CDCl3) δ 7.54 – 7.47 (m, 2H, Ar), 7.41 – 7.31 (m, 3H,
Ar), 5.60 (s, 1H, O-CH-O), 5.03 (d, J = 3.7 Hz, 1H, H-1), 4.23 (dd, J = 10.2,
4.8 Hz, 1H, H-6eq), 3.93 – 3.78 (m, 2H, H-5, CH3-CH-CH3), 3.74 (t, J =
10.3 Hz, 1H, H-6ax), 3.59 (t, J = 8.7 Hz, 1H, H-4), 3.52 (s, 3H, OCH3), 2.22
(dd, J = 13.2, 5.2 Hz, 1H, H-2eq), 1.69 (ddd, J = 13.1, 11.4, 4.0 Hz, 1H, H-
2ax), 1.21 (d, J = 1.3 Hz, 3H, CH3-CH-CH3), 1.15 (d, J = 6.1 Hz, 3H, CH3-
CH-CH3); 13C NMR (101 MHz, CDCl3) δ 137.6 (C, Ar), 128.9 (CH, Ar),
128.2 (2 CH, Ar), 126.1 (2 CH, Ar), 101.4 (O-CH-O), 95.7 (CH-1), 83.9
(CH-4), 74.4 (CH-3), 69.1 (CH2-6), 68.73 (CH3-CH-CH3), 62.9 (CH-5), 58.5
(OCH3), 36.4 (CH2-2), 23.4 (CH3-CH-CH3), 21.3 (CH3-CH-CH3). HRMS
(ESI) m/z calcd for C17H25NaO5 [M+Na]+: 331.15159, found: 331.15347.
For 4d-: 1H-NMR (400 MHz,CDCl3) δ 7.55 – 7.47 (m, 2H, Ar), 7.40 – 7.33
(m, 3H, Ar), 5.57 (s, 1H, O-CH-O), 4.68 (dd, J = 9.8, 1.9 Hz, 1H, H-1), 4.30
(dd, J = 10.5, 4.9 Hz, 1H, H-6eq), 4.01 (hept, J = 6.2 Hz, 1H, CH3-CH-CH3),
3.82 (t, J = 10.3 Hz, 1H, H-6ax), 3.60 (dd, J = 9.4, 7.9 Hz, 1H, H-4), 3.57
– 3.52 (m, 1H, H-3), 3.50 (s, 3H, OCH3), 3.37 (ddd, J = 10.6, 9.0, 5.2 Hz,
1H, H-5), 2.31 (ddd, J = 12.8, 4.7, 2.0 Hz, 1H, H-2eq), 1.62 (td, J = 11.0,
1.3 Hz, 1H, H-2ax), 1.24 (d, J = 6.4 Hz, 3H, CH3-CH-CH3), 1.18 (d, J = 6.1
Hz, 3H, CH3-CH-CH3); 13C NMR (101 MHz, CDCl3) δ 137.4 (C, Ar), 128.9
(CH, Ar), 128.2 (2 CH, Ar), 126.1 (2 CH, Ar), 101.5 (O-CH-O), 98.4 (CH-
1), 82.7 (CH-4), 76.8 (CH-3), 71.1 (CH3-CH-CH3), 68.9 (CH2-6), 66.6 (CH-
5), 58.1 (OCH3), 37.5 (CH2-2), 23.5 (CH3-CH-CH3), 21.8 (CH3-CH-CH3).
HRMS (ESI) m/z calcd for C17H25O5 [M+H]+: 309.16965, found: 309.16993.
Synthesis of 2,2,2-trifluoroethyl 4,6-O-benzylidene-2-deoxy-3-O-
methyl--D-glucopyranoside (4b-): Glycosyl sulfoxide 4 (78 mg,
0.208 mmol) and 2,2,2-trifluoroethanol (60 L, 0.834 mmol) were coupled
following the general procedure II (-60 ºC, 1 h). After work up, the residue
was purified by flash silica gel chromatography (hexane/ethyl acetate 9:1
to 7:3) to give glycosides 4b- (17 mg, 23%) and 4b- (45 mg, 62%). For
4b-: 1H NMR (400 MHz, CDCl3) δ 7.54 – 7.46 (m, 2H, Ar), 7.42 – 7.32 (m,
3H, Ar), 5.60 (s, 1H, O-CH-O), 5.01 (d, J = 3.7 Hz, 1H, H-1), 4.25 (dd, J =
9.7, 4.2 Hz, 1H, H-6eq), 4.03 – 3.78 (m, 5H, H-3, 4, 5, CH2-CF3), 3.76 (t, J
= 10.0 Hz, 1H, H-6ax), 3.62 (t, J = 9.0 Hz, 1H, H-4), 3.52 (s, 3H, OCH3),
2.38 (dd, J = 13.4, 5.1 Hz, 1H, H-2eq), 1.73 (ddd, J = 13.9, 11.2, 3.9 Hz,
1H, H-2ax); 13C NMR (101 MHz, CDCl3) δ 137.3 (C, Ar), 129.0 (CH, Ar),
128.2 (2 CH, Ar), 127.9 (2 CH, Ar), 123.7 (q, J = 277.5 Hz, CF3), 101.6 (O-
CH-O), 98.5 (CH-1), 83.2 (CH-4), 73.9 (CH-3), 68.8 (CH2-6), 64.3 (q, J =
34.8 Hz, CH2-CF3), 63.5 (CH-5), 58.6 (OCH3), 35.3 (CH2-2). HRMS (ESI)
m/z calcd for C16H20F3O5 [M+H]+: 349.12573, found: 349.12706. For 4b-:
1H NMR (400 MHz, CDCl3) δ 7.52 – 7.45 (m, 2H, Ar), 7.40 – 7.32 (m, 3H,
Ar), 5.58 (s, 1H, O-CH-O), 4.71 (dd, J = 9.8, 2.2 Hz, 1H, H-1), 4.32 (dd, J
= 10.5, 4.9 Hz, 1H, H-6eq), 4.13 (dq, J = 12.5, 8.8 Hz, 1H, CH2-CF3), 3.95
(dq, J = 12.4, 8.4 Hz, 1H, CH2-CF3), 3.82 (t, J = 10.3 Hz, 1H, H-6ax), 3.61
(t, J = 8.9 Hz, 1H, H-4), 3.55 (td, J = 10.2, 9.4, 5.6 Hz, 1H, H-3), 3.51 (s,
3H, OCH3), 3.38 (ddd, J = 9.7, 8.8, 4.9 Hz, 1H, H-5), 2.45 (ddd, J = 13.0,
4.8, 2.3 Hz, 1H, H-2eq), 1.65 (dt, J = 13.0, 10.2 Hz, 1H, H-2ax); 13C NMR
(101 MHz, CDCl3) δ 137.2 (C, Ar), 129.0 (CH, Ar), 128.3 (2 CH, Ar), 127.7
(2 CH, Ar), 123.6 (q, J = 280.4 Hz, CF3), 101.6 (O-CH-O), 100.2 (CH-1),
82.4 (CH-4), 76.3 (CH-3), 68.7 (CH2-6), 66.7 (CH-5), 65.6 (q, J = 34.8 Hz,
CH2-CF3), 58.3 (OCH3), 36.5 (CH2-2). HRMS (ESI) m/z calcd for
C16H20F3O5 [M+H]+: 349.12573, found: 349.12591.
Pre-activation glycosylations with donors 1-3 and acceptors a-d for
qualitative stereoselectivity assessment: The corresponding sulfoxide
donor (0.1 mmol) and 2,6-di-tert-butyl-methylpyridine (DTBMP) (0.3 mmol)
were co-evaporated twice with dry toluene and dissolved in dry DCM (25
mL/mmol) under Ar. Activated 4 Å molecular sieves were added and the
reaction mixture was stirred for 30 min at room temperature under Ar
atmosphere. The reaction mixture was cooled down to -78 ºC and then,
triflic anhydride (0.13 mmol) was added dropwise. The reaction mixture
was allowed to stir at -78 ºC for 20 min before adding the acceptor (0.2-
0.3 mmol). The resulting mixture was allowed to stir at the designated
temperature for 1 h, at which point TEA (100 L) was added. The
quenched reaction mixture was then filtered, and the filtrate was
evaporated under vacuum. The crude material was directly dissolved in
Synthesis of 2-fluoroethyl 4,6-O-benzylidene-2-deoxy-3-O-methyl-
/-D-glucopyranoside (4c-/): Glycosyl sulfoxide 4 (83 mg, 0.222
mmol) and 2-fluoroethanol (51 L, 0.89 mmol) were coupled following the
general procedure II (-60 ºC, 1 h). After work up, the residue was purified
by flash silica gel chromatography (hexane/ethyl acetate 9:1 to 7:3) to give
11
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
CDCl3 and subsequently analyzed by 1H-NMR to determine the final
glycosylation anomeric ratio.
[1]
C. S. Bennet in Selective Glycosylations: Synthetic Methods and
Catalysts, John Wiley & Sons, Weinheim, 2017.
Y. Yang, X. Zhang,B. Yu, Nat. Prod. Rep. 2015, 32, 1331-1355.
C. Chang, C. Wu, M. Lin, P. Liao, C. Chang, H. Chuang, S. Lin, S. Lam,
V. P. Verma, C. Hsu, C. Wang, Angew. Chem. Int. Ed. 2019, 58, 16775–
16779.
[2]
[3]
General method for the NMR tube glycosylation experiments with
sulfoxide donors: Considering the increased complexity of these muti-
component reactions, a minimum of three runs under identical reaction
conditions, were completed with each derivative. Additionally, competition
experiments were performed for selected equimolecular donor mixtures.
Thus, the internal standard 4,4,5,5-tetramethyl-2-(naphthalen-1-yl)-1,3-
dioxolane (25.6 mg, 0.1 mmol), DTBMP (61.5 mg, 0.3 mM), the
appropriate acceptor alcohol and activated molecular sieves 4 Å (100 mg)
were added to a septum-capped oven-dried vial. The vial was twice
evacuated and purged with Ar before the addition of CDCl3 (10 mL). The
resulting stock solution (550 L) was transferred to a 5 mm NMR tube fitted
with a septum and purged with Ar. A solution of the adequate donor in
CDCl3 (50 L, 100 mM) was then added to the tube and shaken to mix.
Afterwards, the NMR tubes were cooled down in a liquid nitrogen/acetone
bath to the selected reaction temperature before the treatment with triflic
anhydride and immediately transferred to the NMR spectrometer,
previously equilibrated at the proper working temperature. Glycosylation
reactions were followed by low-temperature NMR experiments, employing
a Bruker Avance 500 MHz spectrometer. 1D-1H spectra were acquired
with 300 excitation pulses and a relaxation delay of 1 s. The extra
sensitivity provided by 13C-labelled samples allowed us to acquire 2D-
HSQC data sets with just 2 scans per increment and 64 increments,
limiting the total experimental time to 2-3 minutes. These conditions were
found to be adequate to monitor relatively fast glycosylation reactions.
Transmitter offsets were set to 5 and 90 ppm, in the proton and carbon
dimensions, respectively, which allowed for an optimal excitation of the
anomeric CH groups. In addition, a delay corresponding to a J value of
165 Hz (between those values expected for anomeric - and -CH
fragments) was employed in all cases. Data processing was carried out
with MestReNova. The absence of well resolved peaks for some
glycosylation products in the 1D data sets led us to mainly rely on 2D-
HSQC experiments to derive relative populations at different reaction
times. To assess the validity of this approach, control HSQC spectra were
measured with increasing relaxation delays (in the 1-10 s range), showing
that this parameter had a relatively minor influence (<10%) on the
estimated volume ratios. When possible, values derived from 2D data sets
were also confirmed by 1D experiments for selected non-overlapped
signals. Relative concentrations at short reaction times (<10 min, when
<20% evolution had occurred) were taken as indicative of the relative
formation rates for the different products. Finally, relative formation rates
for the different products were also estimated, without any mechanistic
assumptions, from their relative conversions at short reaction times.
[4]
[5]
[6]
[7]
W. L. Leng, H. Yao, J. X. He, X. W. Liu, Acc. Chem. Res. 2018, 51, 628–
639.
M. H. Zhuo, D. J. Wilbur, E. E. Kwan, C. S. Bennett, J. Am. Chem. Soc.
2019, 141, 16743–16754.
T. G. Frihed, M. Bols, C. M. Pedersen, Chem. Rev. 2015, 115, 4963–
5013.
P. O. Adero, H. Amarasekara, P. Wen, L. Bohé, D. Crich, Chem. Rev.
2018, 118, 8242–8284.
[8]
[9]
L. Bohé, D. Crich, C. R. Chim. 2011, 14, 3–16.
T. Hansen, H. Elferink, J. M. A. van Hengst, K. J. Houthuijs, W. A.
Remmerswaal, A. Kromm, G. Berden, S. van der Vorm, A. M. Rijs, H. S.
Overkleeft, D. V. Filippov, F. P. J. T. Rutjes, G. A. van der Marel, J.
Martens, J. Oomens, J. D. C. Codée, T. J. Boltje, Nat. Commun. 2020,
11, 2664.
[10] a) S. van der Vorm, T. Hansen, H. S. Overkleeft, G. A. van der Marel, J.
D. C. Codée, Chem. Sci. 2017, 8, 1867–1875. b) Y. Otsuka, T.
Yamamoto, K. Fukase, Chem. Asian J. 2019, 14, 2719–2723.
[11] S. van der Vorm, J. M. A. van Hengst, M. Bakker, H. S. Overkleeft, G. A.
van der Marel, J. D. C. Codée, Angew. Chem. Int. Ed. 2018, 57, 8240–
8244.
[12] D. Crich, L. B. L. Lim, Org. React. 2004, 64, 115–251.
[13] D. Crich, J. Carbohydr. Chem. 2002, 21, 667-690.
[14] B. Yang, W. Yang, S. Ramadan, X. Huang, Eur. J. Org. Chem. 2018, 17,
1075–1096.
[15] D. Crich, S. Sun, J. Org. Chem. 1997, 62, 1198–1199.
[16] a) D. Crich, S. Sun, Tetrahedron 1998, 54, 8321–8348. b) Y. Zhu, B. Yu,
Chem. Eur. J. 2015, 21, 8771–8780.
[17] a) D. Crich, S. Sun, J. Am. Chem. Soc. 1997, 119, 11217–11223. b) K.
Kowalska, C. M. Pedersen, Chem. Commun. 2017, 53, 2040–2043.
[18] J. Gildersleeve, R. A. Pascal, D. Kahne, J. Am. Chem. Soc. 1998, 120,
5961–5969.
[19] M. P. Mannino, A. V. Demchenko, Chem. Eur. J. 2020, 26, 2927–2937.
[20] M. Huang, G. E. Garrett, N. Birlirakis, L. Bohé, D. A. Pratt, D. Crich, Nat.
Chem. 2012, 4, 663–667.
[21] a) M. Huang, P. Retailleau, L. Bohé and D. Crich, J. Am. Chem. Soc.
2012, 134, 14746–14749. b) P. O. Adero, T. Furukawa, M. Huang, D.
Mukherjee, P. Retailleau, L. Bohé, D. Crich, J. Am. Chem. Soc. 2015,
137, 10336–10345.
[22] A. Martin, A. Arda, J. Désiré, A. Martin-Mingot, N. Probst, P. Sinaÿ, J.
Jiménez-Barbero, S. Thibaudeau, Y. Blériot, Nat. Chem. 2016, 8, 186–
191.
[23] B. Dhakal, L. Bohé, D. Crich, J. Org. Chem. 2017, 82, 9263–9269.
[24] A. G. Santana, L. Montalvillo-Jiménez, L. Díaz-Casado, F. Corzana, P.
Merino, F. J. Cañada, G. Jiménez-Osés, J. Jiménez-Barbero, A. M.
Gómez, J. L. Asensio, J. Am. Chem. Soc. 2020, 142, 12501–12514.
[25] Lowary et al, have previously observed that in the activation of furanosyl
sulfoxides at -78 ºC, multiple intermediate species were present. All of
these intermediates, however, equilibrated to a single species upon
warming to -40 ºC, a) C. S. Callam, R. R. Gadikota, D. M. Krein, .T. L.
Lowary J. Am. Chem. Soc. 2003, 125, 13112–13119; b) R. R. Gadikota,
Ch. S. Callam, T. Wagner, B. Del Fraino, T. L. Lowary , J. Am. Chem.
Soc. 2003, 125, 4155–4165. c) Y. Bai, T. L. Lowary J. Org. Chem. 2006,
71, 26, 9658–9671.
Acknowledgements
This investigation was supported by research grants of the
Spanish “Plan Nacional” (MINECO) CTQ2016-79255-P and
RTI2018-094862-B-100. A.G.S. thanks the EU commission for a
MSCA-IF postdoctoral fellowship. L.M.-J. and L.D.-C. thank the
Spanish Ministerio de Economía, Industria y Competitividad and
Ministerio de Ciencia, Innovación y Universidades for an FPI
fellowship each (BES-2014-070232 and BES-2017-080618).
[26] For an earlier precedent of an intramolecular trapping at the sulfur atom
in the activation of glycosyl sulfoxides to give a sulfonium adduct see : M.
Huang, T. Furukawa, P. Retailleau , D. Crich, L. Bohé Carbohydrate
Research 2016, 427, 21–28.
Conflict of interest
The authors declare no conflict of interest.
[27] T. Nokami, A. Shibuya, S. Manabe, Y. Ito, J. Yoshida, Chem. Eur. J. 2009,
15, 2252–2255.
Keywords: glycosylation
• NMR spectroscopy • reaction
[28] J. D. C. Codée, L. Kröck, B. Castagner, P. H. Seeberger, Chem. Eur. J.
2008, 14, 3987–3994.
mechanisms • carbohydrates • glycosyl triflate
[29] C. L. Perrin, Tetrahedron 1995, 51, 11901–11935.
12
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
[30] Derivatives 1 and 2 incorporated 13C at the anomeric position whereas
donors 3 and 4 incorporated 13C at all pyranose positions.
[31] Karthaus, S.-I. Shoda, S. Kobayashi, Tetrahedron Asymmetry 1994, 5,
2213–2216.
[32] N. Khiar, I Alonso, N. Rodriguez, A. Fernández-Mayoralas, J. Jiménez-
Barbero, O. Nieto, F. Cano, C. Foces-Foces C, M. Martin-Lomas M
Tetrahedron Lett 1997, 38, 8267–8270.
[33] V. Ferrières, J. Joutel , R. Boulch , M. Roussel , L.C. Toupet, D.
Plusquellec D Tetrahedron Lett 2000, 41, 5515–5519.
[34] M. A. Fascione, R. Brabham, W. B. Turnbull, Chem. Eur. J. 2016, 22,
3916–3928.O.
13
This article is protected by copyright. All rights reserved.

10.1002/chem.202003850
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
A systematic low-temperature NMR-based study of single-step glycosylations employing 13C-labelled glycosyl sulfoxide donors
provides new evidence for the key role of long-postulated glycosylation intermediates, allowing the detection of hitherto unreported
reactive species.
14
This article is protected by copyright. All rights reserved.