Applications of Shoda's reagent (DMC) and analogues for activation of the anomeric centre of unprotected carbohydrates

Article abstract of DOI:10.1016/j.carres.2020.108197

2-Chloro-1,3-dimethylimidazolinium chloride (DMC, herein also referred to as Shoda's reagent) and its derivatives are useful for numerous synthetic transformations in which the anomeric centre of unprotected reducing sugars is selectively activated in aqueous solution. As such unprotected sugars can undergo anomeric substitution with a range of added nucleophiles, providing highly efficient routes to a range of glycosides and glycoconjugates without the need for traditional protecting group manipulations. This mini-review summarizes the development of DMC and some of its derivatives/analogues, and highlights recent applications for protecting group-free synthesis.

Full text of DOI:10.1016/j.carres.2020.108197

Journal Pre-proof
Applications of Shoda’s reagent (DMC) and analogues for activation of the anomeric
centre of unprotected carbohydrates
Antony J. Fairbanks
PII:
S0008-6215(20)30568-1
DOI:
https://doi.org/10.1016/j.carres.2020.108197
CAR 108197
Reference:
To appear in: Carbohydrate Research
Received Date: 8 October 2020
Revised Date: 10 November 2020
Accepted Date: 11 November 2020
Please cite this article as: A.J Fairbanks, Applications of Shoda’s reagent (DMC) and analogues for
activation of the anomeric centre of unprotected carbohydrates, Carbohydrate Research, https://
doi.org/10.1016/j.carres.2020.108197.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

Applications of Shoda’s reagent (DMC) and analogues for
activation of the anomeric centre of unprotected
carbohydrates
Antony J Fairbanks a*
a School of Physical and Chemical Sciences, University of Canterbury, Private Bag 4800,
Christchurch, 8140, New Zealand
e-mail: antony.fairbanks@canterbury.ac.nz
Dedicated to Professor Shin-ichiro Shoda on the occasion of his 67th birthday
Abstract
2-Chloro-1,3-dimethylimidazolinium chloride (DMC, herein also referred to as Shoda’s reagent)
and its derivatives are useful for numerous synthetic transformations in which the anomeric
centre of unprotected reducing sugars is selectively activated in aqueous solution. As such
unprotected sugars can undergo anomeric substitution with a range of added nucleophiles,
providing highly efficient routes to a range of glycosides and glycoconjugates without the need
for traditional protecting group manipulations. This mini-review summarizes the development of
DMC and some of its derivatives/analogues, and highlights recent applications for protecting
group-free synthesis.
Keywords: carbohydrates, DMC, 2-chloro-1,3-dimethylimidazolinium chloride, ADMP,
glycosylation, glycosides, protecting group free, aqueous solution, nucleophilic substitution
1. Introduction
Glycosylation is the chemical transformation most central to glycoscience. In biological systems
glycosylation is achieved under aqueous conditions, usually by glycosyl transferases processing
activated sugar donor substrates, and attaching them to a myriad of glycosyl acceptors with
complete control of both regio- and stereochemistry. Glycosylation for the synthetic

carbohydrate chemist is usually much more difficult. The organic chemist must strive to define
reaction outcome; achieving control of regiochemistry usually only after multiple protecting
groups manipulations, and performing glycosylation reactions under strictly anhydrous
conditions. Clearly such processes are logistically highly inefficient, both in terms of the lengths
of reaction sequences, and also with respect to atom economy and the quantities of reagents,
solvents, etc required.
Emulation of Nature’s supreme efficiency, i.e. the direct chemical glycosylation of unprotected
carbohydrates, particularly in an aqueous environment, has therefore become an area of
significant interest. Fischer glycosylation, whist typically being performed in an alcohol as the
solvent, is an almost ubiquitous reaction of unprotected monosaccharides, performed at some
point in time by almost all synthetic carbohydrate chemists. However the harsh acidic conditions
involved means that it cannot generally be applied to di- or oligosaccharides. A range of other
chemoselective reactions of unprotected carbohydrates have therefore been developed, and the
reader is refered to several excellent recent reviews for further details of these processes.1,2
This specific focus of the mini-review is the development and application of the dehydrating
agent 2-chloro-1,3-dimethylimidazolinium chloride (DMC 1, ‘Shoda’s reagent’, Fig. 1) for
selective activation of the anomeric centre of unprotected carbohydrates in aqueous solution.
Applications of the azido-derivative 2-azido-1,3-dimethylimidazolinium hexafluorophosphate
(ADMP 3), and the benzimidazole analogue, 2-chloro-1,3-dimethyl-1H-benzimidazol-3-ium
chloride (CDMBI 4) will also be discussed (Fig. 1).
Cl
O
Cl
O
N3
PF6
Cl
Cl
N
N
N
N
N
N
N
N
N
N
ADMP
CDMBI
DMBI
DMI
DMC
3
4
5
2
1
Figure 1. DMC and analogues
DMC 1,3 although first used as a reagent for organic synthesis as far back as the 1972 for the
amidation of penicillanic acid,4 only found intermittent application5 until its development as a
versatile dehydrating agent by Isobe and Ishikawa6,7,8 the late 1990’s in their search for a
replacement for dicyclohexylcarbodiimide (DCC). However it was only some 10 years later that
Shoda realised the potential utility of DMC in the carbohydrate field for reactions in an aqueous

environment. It is in the context of these significant applications in the carbohydrate field that the
epithet ‘Shoda’s reagent’ is applied to DMC. DMC 1 is a relatively cheap commercially
o
available,9 colourless odourless crystalline (m.p. 95-100 C) though moderately hygroscopic
solid, which is very soluble in protic (e.g. H2O, EtOH etc), chlorinated (e.g. CH2Cl2, CHCl3) and
some polar aprotic (e.g. MeCN, DMF) solvents. It is insoluble in non-polar and ethereal solvents
(e.g. Et2O, THF). The azido-analogue of DMC, ADMP 3, which is also commercially available10
but considerably more expensive than DMC, can be made from DMC via firstly a counter-ion
exchange to give the hexafluorophosphate salt (which reduces the hygroscopicity) and then
reaction with sodium azide in acetonitrile (Scheme 1). ADMP 3 is a reagent that was originally
developed and used extensively by Kitamura for diazo-transfers,11,12 migratory aminations13,14
and azide transfer reactions,15 and despite first appearances is reportedly non-explosive.16 Shoda
has also synthesised and used the benzo- analogue of DMC, 2-chloro-1,3-dimethyl-1H-
benzimidazol-3-ium chloride (CDMBI 4, Fig 1).17
The electrophilic nature of DMC 1, ADMP 3, and CDMBI 4 is clear in that they should be
readily attacked by nucleophiles at the carbon atom of the imidazolinium ring between the two
nitrogens. Their hydrolysis products are the ureas 1,3-dimethylimidazolidin-2-one (DMI, 2) and
1,3-dimethylbenzimidazol-2-one (DMBI, 5) respectively, along with 2 equivalents of HCl. DMI
2 and DMBI 5 are therefore side-products formed in all DMC/ADMP or CDMBI mediated
reactions. The corollary is that DMC 1 and CDMBI 4 can be made by chlorination of either DMI
2 or DMBI 5 respectively, and this can be achieved simply by heating with oxalyl chloride in
toluene (Scheme 1). The urea DMBI 5 can in turn be made by methylation of 1H-benzimidazol-
2-ol (itself readily available from benzene-1,2-diamine) under bi-phasic conditions
(toluene/aqueous sodium hydroxide) with a phase transfer catalyst.17
Cl
O
N3
PF6
Cl
(i) NaPF6, MeCN
rt, 99%
(COCl)2
N
N
N
N
N
N
toluene, 50 oC
90%
(ii) NaN3, MeCN
0 oC, quant.
DMC 1
DMI 2
ADMP 3
OH
O
Cl
Cl
MeI, Bu4NBr
HN
N
N
N
(COCl)2
N
N
toluene, 50 o
50%
C
NaOH(aq), toluene
60 oC, 70%
DMBI, 5
CDMBI, 4

Scheme 1. Preparation of DMC 1, ADMP 3, and CDMBI 4.
2. Discussion
2.1 General considerations
2.1.1 Mechanistic pathways
The application of a known dehydrating agent, reportedly unstable in protic solvents, to reactions
of unprotected sugars to be performed in aqueous solution at first seems somewhat counter-
intuitive. However as direct reaction of DMC with solvent water proceeds at only a moderate
rate, particularly at reduced temperature, then selective reaction with an unprotected
carbohydrate substrate becomes a possibility. Crucially the key to selective activation of the
anomeric centre by DMC is the enhanced acidity (pKa ~12.1-12.5)18,19,20,21 of the anomeric
hydroxyl group, which is presumably selectively de-protonated to a significant degree under the
typically basic reaction conditions used for DMC mediated reactions (e.g. pKa of Et3NH+
~10.75). As de-protonation increases nucleophilicity by several orders of magnitude, reaction
can occur via this minor amount of the de-protonated form, with a corresponding shift in
equilibrium. This key difference in pKa’s, in combination with a base of appropriate strength
(most commonly at least a two-fold excess of Et3N), therefore allows the anomeric hydroxyl
group the possibility of out-competing both solvent water and the other sugar hydroxyl groups
for attack on DMC.

OH
O
OH
O
OH
O
OH
O
a)
b)
mutarotation
H2O
mutarotation
H2O
OH
HO
OH
HO
HO
HO
AcHN
AcHN
H2O
OH
HO
OH
OH
DMC 1, base
DMC 1, base
DMC 1, base
H2O
DMC 1, base
H2O
H2O
?
SN2
SN2
?
OH
O
OH
O
OH
OH
O
O
O
N
O
N
HO
HO
HO
HO
AcHN
AcHN
HO
6α
OH
6β
O
N
N
O
?
10β
10α
N
N
O
O
N
N
NGP
O
O
Nu
SN2
OH
SN2
Nu
SN2
HO
HO
9
OH
O
13
NHAc
OH
O
OH
O
O
Nu
Nu
SN2
HO
HO
Nu
HO
Nu
HO
SN2
H+
O
AcHN
1,2-trans product 12
HO
1,2-trans product 8
O
N
1,2-anhydro sugar 7
11
Figure 2. Plausible mechanistic rationalisations of the DMC activation of unprotected sugars for:
a) 2-hydroxy sugars; b) 2-acetamido sugars.
A plausible mechanistic rationalisation for DMC activation of a reducing sugar in aqueous
solution is illustrated in Fig 2. A caveat is that Fig. 2 is based largely on observed reaction
outcomes rather than any detailed mechanistic studies. Different mechanistic pathways are
followed depending on whether the substrate has either a hydroxyl (Fig 2a) or acetamido group
(Fig 2b) at position-2 of the residue at the reducing terminus. For 2-hydroxy sugars, firstly the
anomeric hydroxyl group attacks DMC 1. In the case of the α-anomer this yields the 1,2-cis
activated intermediate 6α, while the β-anomer yields the 1,2-trans activated intermediate 6β.
Hydrolysis of either 6α or 6β by attack of water at the imidazolinium ring results in hydrolysis
back to the starting anomer, without any change of anomeric configuration. In an early paper
Shoda suggested that water could also perhaps cause hydrolysis of 6α back to the β-anomer of
the starting material (see: dotted arrow in Fig 2a), but there is no evidence of this pathway
occurring, and given the relatively low nucleophilicity of water it seems perhaps unlikely.
Intramolecular attack of the 2-hydroxyl of the β-activated intermediate 6β can then result in the
formation of the 1,2-anyhdro sugar 7, which has been observed as an intermediate by NMR (vide

infra). Reaction of 1,2-anhydro sugar 7 with an external nucleophile attacking the anomeric
centre in an SN2 sense then produces the 1,2-trans substituted product 8. Alternatively, an
external nucleophile may also attack the anomeric centre of 6α in an SN2 sense, to again produce
the 1,2-trans substituted product 8. In the absence of external nucleophiles, or in competition
with intermolecular processes, intramolecular attack of the 6-hydroxyl group of 1,2-anhydro
sugar 7 results in formation of the corresponding 1,6-anhydro sugar 9. Although it is possible
that activated intermediate 6α could also undergo this process directly (Fig. 2a dotted arrow), the
report that a 2-hydroxyl group is essential for 1,6-anhydro sugar formation would seem to imply
that this route is not generally followed.
For 2-acetamido sugars the pathways (Fig 2b) differ in so far as that neighbouring group
participation (NGP) of the 2-acetamido group can occur. Again, activation of the α- and β-
anomers of the starting material occurs by attack of the anomeric hydroxyl on DMC, yielding the
reactive intermediates 10α and 10β (Fig. 2b). Intramolecular attack of the 2-acetamido group at
the anomeric centre of the 1,2-trans configured 10β, followed by proton removal by the base,
then affords the oxazoline 11. Under certain circumstances, notably in the presence of added
acid, oxazoline 11 can be opened by attack of an external nucleophile to give the 1,2-trans
substitution product 12. In contrast the α-imidazolinium intermediate 10α cannot directly form
an oxazoline. In the presence of an external nucleophile 10α can undergo an SN2 type reaction to
directly yield the 1,2-trans substitution product 12. However, in the absence of external
nucleophiles the oxazoline is the sole reaction product, so the question therefore arises is how
precisely is 10α converted into the oxazoline. Shoda has suggested (vide supra) that 10α is
probably hydrolysed to regenerate the β-anomer of the free sugar, but again the required SN2
substitution by water has not been evidenced. Additionally Shoda has also stated that ‘the
intermediacy of glycosyl chlorides cannot be ruled out’. Indeed since glycosyl chlorides have
now been isolated following reaction of protected reducing sugars with DMC and Et3N in
CH2Cl2, as recently reported by Judeh and co-workers,22 their formation as transient
intermediates in DMC-mediated reactions in aqueous solution seems distinctly possible.
Therefore, plausibly SN2 attack at the anomeric centre of α-activated intermediates such as
6α and 10α by chloride would give the corresponding β-chlorides as ephemeral species that
could then undergo rapid conversion to either the 1,2-anhydro sugar 7 or oxazoline 11. Finally in

studies performed to date using 2-acetamido sugars as substrates there has been no reported
formation of the corresponding 1,6-anhydro sugar 13 under the reported reaction conditions
(crossed dotted arrow in Fig. 2b)
Several aspects of these mechanistic rationalisations are worthy of comment. Firstly, it is not
completely clear how fast mutarotation of the α- and β-anomers of the staring material is relative
to other processes, particularly as mutarotation is both pH and temperature dependent.
Importantly the reaction pH invariably changes with time as two equivalents of HCl are
produced for each molecule of DMC that reacts: though precise values depend on the identity of
the added nucleophile, the pH at the start of the reaction with Et3N as the base is typically ~12,23
and this decreases during the course of the reaction to ~11. Secondly, it should also be noted that
although some reactions are possible with weaker bases, such as 2,6-lutidine (pKa of conjugate
acid ~ 6.7), the required reaction times (36-48 h with lutidine vs 1 h with Et3N) and temperatures
(rt with lutidine vs 0 oC with Et3N) indicate that such reactions are very much slower.24 Thirdly,
early in the development of DMC Shoda postulated the possible intermediacy of 1,2-anyhydro
sugars,25 such as 7, in nucleophilic substitution processes involving DMC. Shoda actually
confirmed this theory in 2017 when a low T NMR study26 revealed that 1,2-anhydro glucose was
indeed an intermediate in DMC reactions. At the approximately the same time Hoeck also
provided NMR evidence27 for the intermediacy of 1,2-anhydro sugars in a related study on
nucleophilic substitution of 5-O-trityl ribofuranose. It can therefore be concluded that 1,2-
anhydro sugars are indeed formed under a variety of aqueous reaction conditions. Fourthly,
hydrolysis of intermediates such as 6a and 10a back to the starting material (as a mixture of
anomers) could possibly occur by an SN1 pathway. However, there is no direct evidence of
unimolecular processes occurring under the quite strongly basic reaction conditions, and no E1
elimination products have been reported. Likewise, competing SN1 processes may also be
expected to lead to measurable amounts of stereochemical leakage. Finally in the presence of
extremely good external nucleophiles, for example thiols, intermolecular substitution of
intermediates such as 6β or 10β may compete with the stereocontrolling intramolecular
processes, and so some amount of stereochemical leakage, namely the formation of minor
amounts of 1,2-cis products, may be observed.
2.1.2 Side reactions, solvents and scope

In general, the objective of DMC activation is the introduction of an external nucleophile, though
important examples of cyclisation reactions have been developed, pre-eminently oxazoline
formation. The side reactions that generally compete with ‘desired’ nucleophilic substitution are:
a) the hydrolysis of either DMC or activated intermediates (such as 6 and 10) by solvent water;
b) reaction of the added nucleophile directly with DMC; c) intramolecular nucleophilic
substitution reactions to give 1,6-anhydro sugars or oxazolines.
Direct reaction of DMC with the added nucleophile may in some cases be avoided by the pre-
activation process developed by Winssinger (see Section 2.2.3, Scheme 12).28 Herein activation
o
of the sugar can be undertaken at low temperature (e.g. -10 C) by reaction of with DMC and
Et3N in a mixed aqueous solvent system, before the nucleophile is then added after a certain time
period (e.g. 30 min).
Oxazoline formation is invariably a significant competing process when 2-acetamido sugars are
used as substrates, and it cannot be avoided altogether, though the use of a weaker base than
Et3N, for example 2,6-lutidine, has been reported to reduce competitive oxazoline formation
during glycosyl azide formation.24 In contrast 1,6-anhydro sugar formation can reportedly be
reduced significantly either by the use of a bulkier base, such as Hunig’s base (diisopropylamine,
DIPEA), or by using MeCN as a co-solvent with water.29 In some cases, particularly for the use
of ADMP 3, the use of an organic co-solvent such as MeCN is actually required in order to
solubilise the activating reagent. When using ADMP typically the organic co-solvent is the
minor component, e.g. water/MeCN ratios of 4:1 are used. Contrastingly the use of stronger
bases such as DBU (pKa of the conjugate acid, ~13.5) has been reported to increase the amount
of 1,6-anhydro sugar formation.30 Hydrolysis of reactive intermediates by solvent water can be
reduced by using D2O as the solvent; in many cases the use of D2O instead of H2O results in an
increase in product yield, typically in the order of ~5-10 percentage points (e.g. a 60% yield in
water becomes ~65-70% in D2O). This observation has been ascribed to the known solvent
kinetic isotope effect of water in hydrolysis reactions.31,32 Recent studies indicate that DMC
activation is probably limited to aldoses, and that uronic acids are not suitable substrates,
presumably as a result of the deactivating electronic withdrawing effect of the C-6 acid.23
Furthermore, it is notable that where processes have occasionally been applied to 2-deoxy
sugars, the product yields are invariably poorer, and the reaction outcome is invariably not
stereocontrolled. Finally, a comment on product purification may be useful. Typically the

reaction products are highly polar, and the crude reaction mixture contains a very large excess of
Et3NH+ salts, which are difficult to remove completely, either by ion exchange or by
chromatography. A useful method for facilitating product purification is therefore to co-
evaporate the crude reaction mixture twice with an added solution of ~35% w/w aqueous
NH4OH; this volatilises the Et3N, and simplifies subsequent chromatography.
2.2 Applications for nucleophilic substitution at the anomeric centre
2.2.1 Oxygen nucleophiles
Since nucleophilic substitution processes must out-compete reaction with solvent water, all of
the initial DMC-mediated processes that were developed involved intramolecular oxygen
nucleophiles. Firstly, and perhaps most importantly, Shoda developed a reliable method for the
production of glycosyl oxazolines from carbohydrates with a GlcNAc residue at the reducing
terminus (Scheme 2).33 The reaction was found to be compatible with either sulfation (Scheme
2b) or phosphorylation (Scheme 3a)34,35 of sugar hydroxyl groups, and could be applied to
oligosaccharides containing sialic acid.36 The particular importance of this simple procedure
arises since unprotected glycosyl oxazolines are important activated donor substrates for ENGase
enzymes,37 and can be used in the convergent enzymatic synthesis of glycopeptides and
glycoproteins.38,39 Shoda’s seminal contribution in this context was the development of a very
simple DMC mediated process for their direct synthesis in water, which has become the method
of choice for their production,40 superseding all previous methods.

OH
O
OH
O
DMC 1, Et3N
HO
HO
HO
HO
a)
b)
OH
OH
H2O, 0 oC
90%
AcHN
O
N
OSO3Na
O
OSO3Na
O
DMC 1, Et3N
HO
HO
HO
HO
H2O, 0 oC
84%
AcHN
O
N
HO
HO
OH
O
HO
HO
OH
O
OH
O
OH
O
DMC 1, Et3N
c)
O
O
HO
OH
H2O, 0 oC
90%
HO
HO
AcHN
HO
O
N
Scheme 2 Direct synthesis of glycosyl oxazolines from reducing sugars in water.
In particular the process can be readily applied to large N-glycan oligosaccharide structures, and
since it can be readily performed on a small scale it is ideal for application to substrates isolated
from natural sources (Scheme 3b).

Scheme 3 The DMC-mediated production of N-glycan oxazolines using; a) synthetic
oligosaccharides; b) N-glycans isolated from natural sources.

Shoda has reported that CDMBI 4 (Fig. 1) has advantages over DMC for oxazoline formation.17
Firstly, it is less hygroscopic and therefore easier to handle. Secondly that replacement of the
imidazolidine ring in DMC with the more electron-rich benzimidazole both reduces competitive
hydrolysis by direct attack by water on CDMBI, and also makes the reactive glycosyl-
imidazolidinium intermediate more stable. Finally, that the side product urea DMBI 5 (Fig. 1) is
not aqueous soluble and so easier to remove. However, despite these apparent advantages almost
all reports to date have used the more readily available DMC.
O
OH
O
OH
O
DMC 1 (10 equiv.), Et3N (30 equiv.)
a)
HO
HO
D2O, 0 oC, 15 min., 99%
OH
HO
OH OH
OH
O
OH
O
HO
HO
HO
HO
OH
O
OH
O
HO
HO
DMC 1 (6 equiv.)
Et3N (18 equiv.)
O
O
HO
OH
O
HO
OH
O
HO
b)
HO
O
O
D2O, 0 oC, 15 min.
HO
OH
O
HO
OH
O
HO
O
HO
O
87%
HO
O
OH
HO
HO
Scheme 4 DMC-mediated conversion of unprotected sugars directly into 1,6-anhydrosugars.
Following on from oxazoline formation, Shoda reported shortly afterwards41 that for 2-hydroxy
sugars (except mannose), activation with excess DMC in D2O the presence of excess Et3N led to
good conversion directly to the corresponding 1,6-anhydro sugar (Scheme 4). The reaction was
also applied to di- and oligosaccharides, but sugars with a substituent on the 3-hydroxyl group
(e.g. 3-methyl glucose, nigerose [Glcα(1-3)Glc], and laminaribiose [Glcβ1-3Glc]) were not good
substrates. DBU could be used as an alternative base to Et3N, but all the other bases tested
(Hunig’s, lutidine, pyridine, N-methyl morpholine and sodium bicarbonate) did not give any 1,6-
anhydro sugar. A mechanistic rationale was provided (see Fig 2a) which involved the
intermediacy of the 1,2-anyhdro sugar 7 as 2-deoxy glucose and 2-fluoro-2-deoxy glucose did
not react, highlighting the importance of the 2-hydroxyl group. No application of this process to
2-acetamido substrates was reported.

Scheme 5 DMC-mediated production of pNP glycosides from unprotected sugars.
It was not until recently that good yields of reaction products were obtained using DMC
activation of unprotected sugars in the presence of intermolecular oxygen nucleophiles. In a

recent study Fairbanks and Qiu reported42 that the activation of sugars at -10 oC, in either D2O or
water, in the presence of para-nitrophenol (pNPOH) and excess Et3N gave the corresponding
pNP-glycosides (Scheme 5). This one step method of the production of these very useful and
widely used enzyme substrates is significantly more efficient than traditional multi-step
syntheses. The process worked well for 2-hydroxy sugars (Scheme 5a), but for 2-acetamido
substrates oxazoline formation out-competed pNP-glycoside formation. Therefore, for 2-
acetamido substrates a two-step one-pot process was developed in which the oxazoline was
firstly formed deliberately by DMC activation at low T, and then converted to the desired pNP-
glycoside by microwave irradiation. Iterations of DMC activation/oxazoline formation and then
microwave irradiation gave the pNP-glycoside in good yield (typically 60-70%, Scheme 5b). The
process was equally applicable to a complex biantennary N-glycan isolated from natural sources
(Scheme 5c).
In a follow-up study23 the scope of the reaction was investigated, and the authors found that the
process was applicable to both sulfated and phosphorylated sugars (Scheme 6a), but that it was
not applicable to either ketoses or uronic acids. They also reported that whilst the process worked
with other phenols, the product yield depended markedly on the phenol pKa (Scheme 6b). The
process for 2-acetamido sugars was simplified to a single operation involving a change of
solvent; thus in the first step the oxazoline was made by treatment of the reducing sugar with
DMC and Et3N in water, the whole reaction mixture was then freeze dried, the residue dissolved
in a 10:1 mixture of MeCN and DMF before the para-nitrophenol was added, and finally the
mixture irradiated (Scheme 6c). This process was advantageous in that no iteration was required,
saving time and meaning that considerably lower quantities of reagents were needed.

2-O3PO
OH
O
O
OH
+
2NH4
HO
HO
O
Na+
P O
DMC 1, Et3N,
a)
OH
O
H2O, -10 oC
OH
HO
HO
+
O
then co-evaporation
with NH4OHaq
OH
NO2
OH
62%
NO2
O
b)
c)
HO
HO
HO
OMe
HO
HO
HO
DMC 1, Et3N
O
O
+
OH
H2O, -10 oC
52%
OH
AcHN
OMe
NO2
DCM 1, Et3N, H2O, 0 oC
HO
HO
HO
HO
HO
HO
HO
HO
O
O
O
then freeze dry
O
O
O
O
HO
HO
OH
AcHN
AcHN
OH
then
AcHN
AcHN
MeCN / DMF (10:1)
MW 80 oC, 30 min.
NO2
49%
Scheme 6 DMC-mediated production of phenyl glycosides from unprotected sugars.
2.2.2 Nitrogen nucleophiles
Glycosyl azides are highly versatile synthetic intermediates,43 but their synthesis by traditional
methods requires multiple steps. Shoda was quick to realise the potential of DMC for
intermolecular nucleophilic substitution reactions of unprotected sugars with good nucleophiles,
such as azide, which should outcompete side reactions. In the first report DMC activation of
unprotected sugars in the presence of a large excess of azide and Et3N (typically 10 equiv. of
each) produced the corresponding 1,2-trans glycosyl azides directly using D2O as the solvent
(Scheme 7).24 However, significant side reactions were also discovered. Firstly, when the
reaction was applied to disaccharides with Et3N as the base, considerable amounts of 1,6-
anhydro sugars were formed as by-products. It was suggested that the amount of 1,6-anyhdro
sugar formation depended on the identity of the added base and a screen of alternatives revealed
that 1,6-anhydro sugar formation was reduced using Hunig’s base (diisopropylethylamine,
DIPEA). It was tentatively suggested that this bulkier base resulted in less intramolecular
reaction. Secondly for 2-acetamido substrates, oxazoline formation competed significantly with

glycosyl azide formation. In these cases it was found that the use of the weaker base 2,6-lutidine
(pKa of the conjugate acid ~6.7) reduced oxazoline formation. Interestingly it was reported that
reaction also worked on a 2-deoxy sugar (yield not given), but in this case the transformation
was not stereoselective, highlighting the importance of 1,2-anhydro sugar intermediates for
stereocontrolled reactions, but also revealing that the 2-hydroxyl group is not absolutely
required. Glycosyl azides formed in this way were used as substrates for subsequent click
reactions with acetylene terminated polylactides for the formation of 70-200 nm diameter
nanoparticles.44
Scheme 7 DMC-mediated protecting group free production of glycosyl azides.
Later Fairbanks and co-workers45 revisited glycosyl azide formation with the intention of
developing a one-pot method for glycosyl azide formation and subsequent click reaction by
Huisgen cycloaddition,46,47 to give a wide variety of glycosyl triazoles.48 However, the very large
excesses of azide and base required for the Shoda procedure made this difficult. Several
improvements were made. Firstly, realising that doubtless azide itself actually reacted with DMC
during the transformation, producing one equivalent of HCl in the process, an equivalent
compound to the product of that reaction (e.g. 2-azido-1,3-dimethylimidazolinium
hexafluorophosphate, ADMP 3, Fig. 1) was used as the activator instead of DMC. It should be
noted that a mixed solvent system of water and MeCN (4:1, v/v) is required to completely
solubilise the ADMP for its use as an activator. Unlike DMC 1, ADMP 3 is a non-hydroscopic

stable crystalline solid and is easily handled, and can easily be made from DMC as shown
previously (Scheme 1). Secondly in order to overcome any competitive oxazoline formation a
procedure first developed by Shoda for the synthesis of thiopyridyl glycosides (vide infra, was
applied to 2-acetamido substrates. Thus, following the completion of the reaction, 1 M aqueous
HCl was added to the crude reaction mixture until pH 2 was reached; this was followed by
immediate neutralisation by the addition of aqueous NaHCO3. This process resulted in oxazoline
ring opening by azide in solution, and the production of desired glycosyl azides as the sole
reaction products.
Scheme 8 ADMP mediated production of glycosyl triazoles.

Using ADMP 3 activation (typically only 3 equivalents are required) a one-pot conversion of the
glycosyl azide formed in situ to a wide variety of functionalised glycosyl triazoles was possible,
simply by the addition of an alkyne, CuSO4, and L-ascorbic acid, and then heating at 50 °C. The
process could not only be used to make glycosyl triazoles in a single step (Scheme 8a), but also
as a ligation method by which sugars could be linked together (Scheme 8b), or by which sugars
could be linked to peptides via triazole rings. The process was applicable to both large
oligosaccharides and for peptides up to 20 amino acids long (Scheme 8c). It should be noted that
the use of D2O as the reaction solvent leads to some deuterium incorporation into the triazole
product.
This type of ADMP-mediated click chemistry has found useful application, particularly in the
glyco/biomaterials field. For example recently Zuckermann and co-workers used it for the
production of glycosylated peptoid nanosheets.49 In this work (Scheme 9a) mannose was
converted to the corresponding glycosyl azide using ADMP 3, and then immediately clicked to a
variety of peptoids (typically ~30 units long) containing alkynes by the addition of copper (I)
hexafluorophosphate (TACP) as the copper catalyst, 2,6-lutidine as the base, and
tris(hydroxypropyltriazolylmethyl)amine (THPTA) as stabilising agent. In another recent report
Pieters50 used ADMP/click chemistry for the assembly of divalent ligands for the adhesion
protein of Peudomonas aeruginosa (Scheme 9b). ADMP 3 has also been recently been used by
Pei and co-workers51 for the direct production of unprotected glycosyl azides and their
attachment to alkyne-functionalised surfaces by click chemistry, to make branched chain
carbohydrate chips for the study of protein/carbohydrate interactions. Godula and co-workers52
used ADMP 3 for the direct production of the glycosyl azide of the trisaccharide 6’-sialylactose
and its conjugation to a dibenzocyclooctyne containing lipid using a strain-promoted click
reaction, as part of the de novo assembly of a synthetic glycocalyx.
Fairbanks has demonstrated that ADMP 3 can be used for the protecting group free synthesis of
glycoconjugates by a double click approach, which involves the use of a bifunctional linker such
as allyl propargyl ether 13 (Scheme 9c). In the first step this is conjugated to the sugar, via
t
ADMP-mediated glycosyl azide formation in a BuOH/water solvent mixture (2:1, required
because of the low aqueous solubility of 13) and azide-alkyne click. Then in the second step the

product is attached to a thiol via a photochemical thiol-ene click reaction using 2,2-dimethoxy-2-
phenyl acetophenone (DMPA) as the initiator (Scheme 9c).53

Scheme 9 Some recent applications of ADMP/click chemistry.
2.2.3 Sulfur nucleophiles
Since sulfur is highly nucleophilic is may be expected that DMC activation of reducing sugars in
the presence of S-nucleophiles should result in efficient intermolecular substitution. However
with 2-acetamido sugars some oxazoline formation is also to be expected. Thioglycosides are
commonly used glycosyl donors for chemical oligosaccharide synthesis, but are normally
required in protected form, so there is perhaps only a limited advantage to be gained in their
direct production in aqueous solution. On the other hand glycosyl thiols find direct use in
unprotected form, particularly for their chemoselective attachment to peptides and proteins.
Therefore there are significant advantages in routes by which they are directly accessible from
unprotected sugars.
Shoda and co-workers reported29 that DMC-mediated activation of reducing sugars in aqueous
solution in the presence of Et3N and a range of aryl and pyridyl thiols led to the direct synthesis
of range of the corresponding aryl thioglycosides (Scheme 10). Two points are worthy of note.
Firstly in some cases when electron rich aryl thiols were used the process was not completely
stereoselective (Scheme 10a). This was presumably as a result of the high nucleophilicity of the
aryl thiol, which was able to compete with 1,2-anyhydro sugar formation as an intermolecular
nucleophile, leading to the formation of minor amounts of 1,2-cis thioglycoside product.
Secondly when water alone was used as the solvent some 1,6-anhydro sugar formation was
observed, but it was stated that this was suppressed when MeCN was used as a co-solvent (>35%
by volume of MeCN required). However, in contrast to this statement, the most commonly used
o
reaction conditions used a 4:1 (v/v) mixture of water and MeCN as the solvent at 0 C. The
reaction was extended to the production of more pyridyl thioglycosides using a range of mono-,
di-, and oligosaccharide substrates.30 Using this approach 2-deoxy glucose could be converted to
the corresponding pyridyl thioglycoside by the use of large excesses of reagents (6 equiv. of
DMC, and 16 equiv. of thiol and base), but the reaction was not stereoselective (β:α, 1.6:1).
Interestingly it was found that when GlcNAc was used as a substrate a considerable amount of
oxazoline was also formed (as observed by NMR), but this could be converted to the desired
product in situ simply by treating the crude reaction mixture with 1M HCl. This acidic treatment

method allowed the procedure to be applied to a range of oligosaccharides with a GlcNAc
residue at the reducing terminus (Scheme 10c).
SH
HO
HO
HO
Me
HO
HO
O
DMC 1, Et3N
a)
b)
O
+
S
HO
HO
OH
H2O / MeCN,1:1
0 oC, 93%
HO
β
:
α
, 4.5:1
Me
S
HO
HO
HO
HO
HO
HO
O
DMC 1, Et3N
O
HN
+
S
N
HO
OH
H2O / MeCN,4:1
0 oC, 91%
HO
S
OH
O
OH
O
OH
O
OH
O
HN
c)
d)
HO
HO
HO
HO
DMC 1, Et3N
O
O
OH
S
HO
HO
N
AcHN
AcHN
AcHN
AcHN
H2O / MeCN, 4:1, 0 oC
then acifify to pH 2
80%
SH
OH
O
OH
O
HO
HO
HO
O
HO
(i) DMC 1, Et3N
H
HO
+
N
HO
OH
H2O / MeCN,1:1, 63%
(ii) acryloyl chloride
S
RAFT
glycopolymer
NH2
Et3N, H2O, THF, 61%
polymerisation
Scheme 10 Direct synthesis of aryl and pyridyl thioglycosides using DMC.
Shoda and co-workers went on to apply this basic transformation to the synthesis54 of a range 4-
methyl-7-thioumbelliferone (MUS) glycosides of mono- and oligosaccharides, which comprised
a carbohydrate attached to a fluorescent label. Another interesting application by Shoda was the
direct DMC-mediated synthesis of 4-aminophenyl 1-thio-glycosides in water, and their
subsequent acrylamidation (Scheme 10d).55 The acylamide containing ‘glycomonomers’ then
underwent reversible addition-fragmentation chain transfer (RAFT) living radical
polymerisation, to give glycopolymers which were finally conjugated to gold nanoparticles to
investigate glycocluster effects.
In related work Winssinger56 used DMC activation and substitution with a thiopicolyl azide 14 to
access azide-containing thioglycosides (Scheme 11). Interestingly when direct DMC mediated of

glucose was attempted with 14 only poor product yields were obtained because of azide
reduction by the thiol under the basic reaction conditions (Scheme 11a). Thiol 14 was therefore
deliberately reacted with DMC first, to produce imidazolinium species 15 as a stable crystalline
solid (Scheme 11b). Reaction of 15 with a range of mono-, and small oligosaccharides in the
presence of Hunig’s base in a water / DMF (3:1) solvent mixture then proceeded smoothly to
give the corresponding thioglycosides. The reactions were completely 1,2-trans stereoselective
except in the cases of galactose and fucose where some of the α-anomers (10-16%) were also
isolated. The presence of azide in the products then allowed chelation-assisted Cu-catalysed click
reactions in the presence of the ligand tris-hydroxypropyltriazolylmethylamine (THPTA) for the
attachment of sugars to either peptides (Scheme 11c) or oligonucleotides containing alkyne side
chains. The approach was also extended to HeLa cells that had been incubated with
homopropargylglycine to incorporate an alkyne into protein biosynthesis; herein a fluorescent
labelled sugar could be conjugated to live cells and subsequently visualised by fluorescence
microscopy.56

Scheme 11 Synthesis of azido-picolyl thioglycosides for chelation-assisted click chemistry.
Winssinger was the first to investigate DMC-activation as a means of linking sugars to peptides
by reaction of activated intermediates with cysteine residues (Scheme 12).28 Key to the
development of a successful reaction here was pre-activation of the sugar by DMC, by stirring
the sugar and Et3N at -10 ˚C in a H2O/dioxane solvent mixture and adding the DMC, before then
adding a cysteine containing amino acid or peptide 30 min. later. An NMR study indicated the
stability of an intermediate species at this temperature (T1/2 >100 min. at -10 oC), and so provided
a window of opportunity for activation before addition of the nucleophile. Although this
intermediate was originally assigned as the oxazolinium ion 6α,28 with hindsight and comparison
with the subsequent work of Shoda, this species is probably the 1,2-anhydrosugar 7. Regardless
of the identity of the intermediate, the use of a pre-activation step is required to avoid direct
reaction of the thiol with DMC, which is a major side reaction, and in the case of a non-aromatic

thiol is irreversible. Using this low T pre-activation method Winssinger and co-workers firstly
glycosylated a polymer-linked peptide containing an aryl thiol (Scheme 12a). Subsequently they
obtained moderate to good yields (30%-95%) of glycosylation of the cysteine residue in a
polymer bound PhePheCys tripeptide (Scheme 12b) with a range of mono- and disaccharides.
Importantly direct reaction of the cysteine containing peptide with the sugar and DMC led only
to formation of the side product 16 resulting from direct reaction of the thiol with DMC. Finally
by using three iterations of sugar pre-activation and then reaction with the peptide, conversions
could be increased to >99%. This high yielding iterative approach allowed application of the
methodology to the glycosylation of two cysteine residues in a 20-mer peptide, which was
synthesised as a MUC1 analogue. Glycosylation could be performed either during the solid-
supported peptide synthesis, allowing the attachment of different mono- or disaccharides to the
two cysteines, of following completion of the peptide allowing installation of the same sugar at
both positions.
OH
O
OH
O
DMC 1, Et3N
OH
O
a)
HO
HO
HO
HO
H2O / dioxane, 1:1 -10 oC
HO
HO
OH
HO
HO
O
O
30 min.
7
Ph
N
N
O
O
H
6α
N
Then add:
N
N
H
Ph
H
O
SH
Ph
O
O
= NovaPEG Rink resin
H
N
N
N
H
H
OH
OH
O
Ph
S
HO
O
OH
DMC 1, Et3N
H2O / dioxane, 1:1 -10 oC
30 min.
OH
O
Phe Phe
NHFmoc
b)
HO
HO
OH
OH
OH
HO
Then add:
S
O
HO
Phe Phe
NHFmoc
SH
OH
Phe Phe
NHFmoc
S
N
N
16

o
Scheme 12 The Winssinger pre-activation method with DMC at -10 C allows: a) glycosylation
of a polymer bound thiol; b) glycosylation of cysteine containing peptides on the solid phase.
Glycosyl thiols have proven to be extremely useful species for the production of neo-
glyconjugates, using a variety of strategies to link the thiol to peptides or proteins.57,58 However,
typically their synthesis has required multistep reaction sequences and protecting group
manipulations. Rademann59 and Fairbanks60 both reported the use of DMC for the protecting
group free synthesis of glycosyl thiols (Scheme 13). The method, which is limited in application
to substrates (mono-, di-, or oligosaccharides) with a 2-acetamido residue at the reducing
terminus, involves DMC mediated formation of an oxazoline, which is then opened in situ by the
addition of a thiocarboxylic (e.g. thioacetic or thiobenzoic) acid to the reaction mixture once
oxazoline formation is complete. The anomeric thioester is then hydrolysed in a subsequent step.
The presence of a thiol and DMC together in the reaction mixture simply leads to reaction
between the two (vide supra), whilst attempted application of the method to 2-hydroxy sugars
led to extensive competitive 1,6-anhydro sugar formation. Rademann used this approach for the
production of glycosyl thiols via both glycosyl thioacetates and thiobenzoates (Scheme 13b), and
exemplified their application in numerous derivatisation reactions.59 Impressively this included
the conjugation of a hexasaccharide fragment of hyaluronic acid to an engineered lipase
Thermoanaerobacter thermohydrosulfuricus, which contains an N-pentenoyl lysine residue at
position 221, via thiol-ene click chemistry. Fairbanks demonstrated application of the method for
the protecting group free conjugation of a naturally derived oligosaccharide to a synthetic 16-mer
peptide containing an allyl glycine (Agl) residue via thiol-ene click chemistry (Scheme 13c).60

OH
O
OH
O
DMC 1, Et3N
H2O, 0 oC, 30 min.
a)
b)
HO
HO
HO
HO
SAc
AcHN
AcHN
OH
then add AcSH,
90%
OH
O
OH
OH
O
HO
HO
O
O
HO
O
HO
AcHN
AcHN
OH
AcHN
DMC 1, Et3N
then add BzSH
in MeCN, 65%
D2O/MeCN (2:1)
0 oC to RT, 90 min.
OH
O
OH
O
OH
HO
O
HO
O
O
HO
HO
AcHN
OH
SBz
AcHN
AcHN
HO
CO2H
OH
c)
OH
HO
AcHN
O
O
O
O
O
HO
O
HO
HO
OH
AcHN
HO
O
HO
HO
OH
O
OH
O
O
HO
O
HO
O
OH
AcHN
HO
HO
HO
HO
OH
OH
CO2H
OH
O
O
HO
AcHN
O
O
O
O
O
HO
HO
HO
OH
AcHN
(i) DMC 1, Et3N, D2O, 0 o
C
(iii) DMPA, DMF/H2O
then add AcSH in MeCN, 60%
(ii) 1M NaOH (aq), RT, quant.
hν, 4h RT, >95%
HO
OH
HO
CO2H
OH
OH
O
HO
AcHN
DVAglTSQAVLPDDFPRY
O
O
O
O
HO
O
HO
OH
AcHN
HO
O
HO
HO
OH
OH
O
O
O
HO
O
HO
O
S
AcHN
HO
HO
HO
HO
OH
OH
HO
CO2H
O
OH
HO
AcHN
DVAglTSQAVLPDDFPRY
O
O
O
O
O
O
HO
HO
OH
AcHN
Scheme 13 Protecting group free synthesis of glycosyl thiols of 2-acetamido sugars using DMC
and application for thiol-ene click chemistry.

OH
O
OH
O
OH
O
DMC 1, Et3N
NH
OMe
NaOMe, MeOH
a)
b)
HO
HO
HO
HO
HO
HO
HSCH2CN
S
CN
S
RT, 64 h
H2O, 0 oC
30 min., 83%
HO
HO
HO
OH
35% conversion.
OH
O
OH
O
HO
HO
HO
HO
OH
O
OH
O
DMC 1, Et3N
HO
HO
O
O
HSCH2CN
HO
HO
OH
O
OH
HO
HO
O
H2O, 0 oC, 30 min.
90%
O
O
S
CN
HO
HO
OH
HO
HO
NaOMe, MeOH, rt, 64 h,
60% conversion.
OH
O
OH
O
HO
HO
OH
O
HO
HO
HO
O
OH
O
HO
OH
O
HO
NH
O
HO
Lysozyme
HO
O
OH
O
S
NH
OMe
HO
HO
N
H
O
borate buffer
pH 9, H2O
HO
S
HO
n
HO
n = 1-3 (mean = 2.0)
Scheme 14 Protecting group free synthesis of cyanoethyl thioglycosides of 2-hydroxy sugars
using DMC, and applications for neoglycoprotein synthesis
Fairbanks reported the protecting group free production61 of cyanoethyl thioglycosides by
reaction of unprotected sugars with DMC and mercaptoacetonitrile (HSCH2CN) in the presence
of Et3N (Scheme 14). Since direct reaction of a thiol with DMC is a competing side reaction, this
must be reversible in this case, in line with the low pKa of HSCH2CN (pKa ~6), meaning that
pre-activation is not required. However the correct order of the addition of reagents was
important to maximise product yield, and reduce the amount of 1,6-anhydro sugar by-product
formed. The reaction was applicable to a range of mono-, di- and oligosaccharides with a 2-
hydroxy sugar at the reducing terminus, but the instability of both the cyanoethyl glycoside
product and mercaptoacetonitrile towards acid meant that good yields could not be obtained from
2-acetamido sugars because of competing oxazoline formation. Cyanoethyl thioglycosides can
be readily converted into the corresponding imidates (2-imino-2-methoxyethyl-1-thioglycosides,
or IME-thioglycosides), by reaction with methoxide as first reported by Lee;62 this equilibrium
activation process typically produces the corresponding imidate with ~50% conversion. These
imidates are useful reagents for the conjugation of carbohydrates to other entities,63 and in
particular for protein modification by selective reaction with surface exposed lysines, for

example with lysozyme (Scheme 14b).64 Access to them without recourse to any protecting
group manipulations opens up the production of diverse new IME-thioglycosides using small
quantities of oligosaccharides derived from natural sources.
Scheme 15 Protecting group free synthesis of glycosyl dithiocarbamates and their conversion to
1-deoxy sugars.
In 2016 Shoda first reported the DMC mediated production of a range of glycosyl
dithiocarbamates (glycosyl DTCs, Scheme 15), by reaction of unprotected sugars with DMC,
Et3N and several different DTC salts in a water/MeCN solvent mixture (usually 1:1) at low
temperature (ranging from -15 to 0 oC).65 The process could be applied to a range of mono-, di-
and oligosaccharides. When GlcNAc was used as a substrate more that 40% of the oxazoline was
formed as a side-product, but the yield of the desired product was significantly increased by the
addition of 1M aqueous trifluoracetic acid, which caused oxazoline opening by the remaining
dithiocarbamate in solution (Scheme 15b). Interestingly 2-deoxy glucose could be converted to
the corresponding glycosyl dithiocarbamate, but the yield was poor (42%) and the reaction was
not stereoselective (α:β, 45:55). The reaction was expanded to a one-pot three-component
process involving the in situ synthesis of the DTC salt by reaction of CS2 with a dialkylamine in
the presence of the Et3N, before the addition of an aqueous solution of the sugar, and finally the