Synthesis and characterization of a small library of bisglucosides: Influence of the nature of the diol/diphenol used in O-glucosylation

Article abstract of DOI:10.1016/j.carres.2020.108217

In this paper, the synthesis of different bisglucosides is investigated through the reaction of two acetylated glucose units with a diol (or diphenol) in order to develop a versatile molecular platform for the future development of bio-based polymers. A p

Full text of DOI:10.1016/j.carres.2020.108217

Carbohydrate Research 500 (2021) 108217
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Synthesis and characterization of a small library of bisglucosides: Influence
of the nature of the diol/diphenol used in O-glucosylation
*
´
Stephane Patry, Mike Robitzer, Jean-Pierre Habas
ICGM, Univ Montpellier, CNRS, ENSCM, 34095 Montpellier, France
A R T I C L E I N F O
A B S T R A C T
Keywords:
In this paper, the synthesis of different bisglucosides is investigated through the reaction of two acetylated
glucose units with a diol (or diphenol) in order to develop a versatile molecular platform for the future devel-
opment of bio-based polymers. A panel of five diols and one diphenol is initially used in order to examine the
influence of their chemical skeleton on the reaction yield and both nature and proportion of formed species.
Reaction products are identified using 1H and 13C NMR spectroscopies completed by MALDI-TOF MS technique.
The nucleophilicity of these dihydroxy compounds is identified as being the main factor that governs the reaction
characteristics. In particular, the best selectivity is obtained with the use of hydroquinone. Inversely, by-products
(oligomers, deacetylated compounds) are observed with the diols defined by higher nucleophilicity despite the
choice of stereoselective pathway using acyl protecting groups.
Bisglucoside
Diol
Hydroquinone
Biobased monomer
Molecular spacer
O-glycosylation
Analytical techniques
classical peracetic pathway, these researchers produced epoxidized su-
crose polyesters of fatty acids (ESPEFAs) and reacted them with cyclic
anhydrides [10]. The resulting materials presented higher ultimate
performances compared to that of materials derived from native Epox-
1. Introduction
Over the last few decades, the development of biobased polymers
arises a great interest in both scientific and industrial communities [1,
2]. Different elements support researches in this area such the reduced
dependence on petroleum and the implementation of new environ-
mental and safety regulations. Then, the valorization of biobased
(macro)molecules appears as a possible solution for reducing the use of
toxic precursor compounds [3]. A more recent trend consists in creating
innovative molecular building blocks to achieve the creation of
custom-made polymers by controlling the chemical structure in order to
obtain precise properties for specific use [4]. Among the different
existing molecules, carbohydrates are worth of interest because they
combine a high degree of functionality with a non-toxic cyclic structure
[5]. The former characteristic offers a large range of possible reactions
while the latter insures higher physical properties than that observed
with aliphatic sequences. As proof of this, carbohydrate derivatives are
already present into polymer formulations when high glass transition
temperature Tg is required. For instance, they are used to produce
thermoset polymeric formulations such as alkyd resins [6], poly-
urethanes [7] or epoxy resins [8]. Concerning this latter family, Pan
et al. worked with commercial sucrose polyesters of fatty acids (SPEFAs)
named as Sefose® by Procter & Gamble (P&G) Chemicals [9]. Using a
idized Soybean Oil (ESO) cured with same hardeners. This hierarchy was
justified by the presence of cyclic sucrose in ESPEFAs whereas ESO is
made with aliphatic molecular segments. Unfortunately, this study was
limited to the only family of unsaturated SPEFAs previously developed
by P&G.
The boundaries of this concept are likely to be pushed back by the
creation of other cyclic molecular cores to achieve the emergence of new
performances. Such approach could be carried out through the prior
synthesis of a little library of bisglucosides. We studied more particularly
the production of this class of molecules through the reaction of a
dihydroxy compound with two glucose units. This O-glucosylation
process was investigated with various aliphatic diols and a diphenol.
Indeed, O-glycosylation is not easy to be performed in organic chemistry
because it is often accompanied by parasitic and complex mechanisms
[11]. For this reason, different strategies are described in literature
involving the pre-activation of the aglycone, Huisgen cycloaddition
between alkynes and azides [12–15], olefin cross-metathesis [16], and
Sonogashira coupling [17]. The use of protective participating groups
(OAc) or non-participating groups (OBn) seems to be essential for con-
ducting selective O-glycosylation. For instance, β-D-glucose pentaacetate
* Corresponding author.
E-mail addresses: stephane.patry@umontpellier.fr (S. Patry), mike.robitzer@enscm.fr (M. Robitzer), jean-pierre.habas@umontpellier.fr, jean-pierre.habas@
umontpellier.fr (J.-P. Habas).
https://doi.org/10.1016/j.carres.2020.108217
Received 15 July 2020; Received in revised form 26 November 2020; Accepted 1 December 2020
Available online 5 December 2020
0008-6215/© 2020 Elsevier Ltd. All rights reserved.

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
Abbreviations
Ionization)
FT-IR
Fourier Transform InfraRed
BF3⋅OEt2 Boron trifluoride diethyl etherate
MALDI-TOF MS Matrix Assisted Laser Desorption/Ionization - Time
of flight – Mass Spectrometry
BGP(OAc)8 BisGlucosidePropyl octaacetate
BGX
BisGlucosideX
MGP(OAc)5 MonoGlucosidePropyl pentaacetate
MGX(OAc)5 MonoGlucosideX pentaacetate
BGX(OAc)8 BisGlucosideX octaacetate
BnD: cis 2-butene-1,4-diol
NMR
OAc
1,3-PD
PDI
Nuclear Magnetic Resonance
Acetate group –O(CO)CH3
1,3-propanediol
BPA
ByD
CHD
DEG
DSC
Bisphenol A
2-butyne-1,4-diol
Cyclohexanediol
PolyDispersity Index
Diethylene glycol
pKa
Constant of acidity
Differencial Scanning Calorimetry
RWL:
Tdeg
Relative weight loss
Glc(OAc)5 Glucose pentaacetate (
α
or/and β)
Degradation temperature
Melting temperature
HD
HQ
Homopolymerization Degree
Tm
Hydroquinone
TGA
ThermoGravimetric Analysis
HRMS (ESI) High Resolution Mass Spectrometry (ElectroSpray
β-Glc(OAc)5 can be easily synthesized through a simple acetylation. It is
also suitable with a one-step glycosylation that can be catalyzed by a
Lewis-acid [18]. In literature, some works describe the bisglycosylation
reaction between glycosyl peracetates with various diols in dichloro-
methane (CH2Cl2) as solvent and catalyzed by BF3⋅OEt2 [19,20]. In fact,
this catalyst is frequently used for phenol O-glycosylation [21]. It effi-
ciently promotes the conversion of β-Glc(OAc)5 into the corresponding
aryl-O-glucosides with high yield and β-stereoselectivity [22]. Other
Lewis acids are also employed but they induce the formation of complex
mixtures [20] and even lead to lower reaction yield [23].
2. Results and discussion
2.1. Conditions retained for the synthesis of BisGlucosideX octaacetate
BGX(OAc)8
Considering the data already published in literature and summarized
hereabove, the synthesis of the BGX(OAc)8 was conducted by reaction of
β-Glc(OAc)5 1a with a diol or a diphenol (HO-X-OH) 2a-f in anhydrous
CH2Cl2 as solvent, catalyzed by BF3⋅OEt2 and at room temperature
(Fig. 1). A molar ratio of the sugar donor/diol (or diphenol) acceptor
was set to 1:0.5 in order to promote the formation of BGX(OAc)8.
We chose to use different diols and one diphenol to investigate the
effects and limitations of the selected glucosylation conditions (Fig. 2).
The diols 2a-e are based upon different aliphatic skeletons and are likely
to present discrepancies in terms of molecular mobility. Indeed,
compared to propane-1,3-diol (1,3-PD 2a), diethylene glycol (DEG 2c) is
characterized by a longer linear sequence including a flexible ether
hinge. 2-butyne-1,4-diol ByD (2d) and cis-2-butene-1,4-diol (BnD 2b)
have rigid unsaturations at the center of their molecular structure. A
cycloaliphatic spacer is represented by 1,4-cyclohexanediol (CHD 2e)
The nature of the solvent is another factor that has to be considered
for performing O-glycosylation reaction. For instance, Mabic et al.
described the coupling reaction between 3-methylcatechol and β-Glc
(OAc)5 in the presence of BF3⋅OEt2 as an activator [24]. No coupling is
observed when participating solvents are used such as diethyl ether,
acetonitrile or tetrahydrofuran (THF). This result is probably the reason
why CH2Cl2, a non-complexing solvent, is preferred. The configuration
of the anomeric center is also dependent on the temperature used for the
glucosylation. Temperatures lower than 30 ◦C favors β-glucoside for-
mation while both anomeric compounds coexist at higher temperatures
[25].
Despite these numerous researches, it is important to note that the
collection of bivalent glycosides synthesized up to now remains rather
limited. Most of the structures are derived from linear diols such as long-
chain alkyl “HO(CH2)12OH” [26], alkynes with 2-butyne-l,4-diol [27]
and polyethoxy with HO(CH2CH2O)5H [20]. Another researchers
investigated the use of phenols [28]. Reported reaction yields are
comprised between 11 and 81%. Within this context, we decided to
conduct a more global study by building with an only pathway, a library
of BisGlucosideX octaacetate with different spacers between two glucose
units. Each of them was represented as BGX(OAc)8 in which X would
correspond to the diol or diphenol spacer (Fig. 1). The synthesis results
were carefully discussed and interpreted on the basis of reliable mo-
lecular mechanisms.
Fig. 2. Panel of dihydroxy compounds (2a-f) used in this study for O-
glucosylation.
Fig. 1. General glucosylation of diols/diphenol (2a-f) with β-Glc(OAc)5 (1a) promoted by BF3⋅OEt2 for the production of BisGlucosideX octaacetate (BGX(OAc)8 with
X = aromatic chain, cycloaliphatic, aliphatic of variable length).
2

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
while hydroquinone (HQ 2f) is the only diphenol employed in this work
percentage of glucosides that can be calculated at different times using
the following formula (1):
section.
IH1βG + IH1
α
G
τ
=
*100
αG
1
2.2. Synthesis of BisGlucosidePropyl octaacetate BGP(OAc)8
(IH1βOAc + IH1 OAc + IH1βG + IH1
)
α
IH1 OAc and IH1βOAc are the intensities of the 1H NMR bands charac-
First experiments consisted in investigating the reaction kinetic of
β-Glc(OAc)5 1a with commercial 1,3-PD 2a in order to define the time
necessary for obtaining the optimal conversion rate of the chemicals
used.
α
teristic of the acetylated glucoside under the
α and β anomeric forms,
respectively. IH1 and IH1βG define the intensities of bands specific of
G
the glucoside unαder the
α and β anomeric forms, respectively.
The exploitation of 1H NMR spectra reveals a fast conversion of about
56% after 2 h and 77% after 8 h. After, the evolution of
τ is much
2.2.1. Kinetic study
reduced between 8 h and 23 h (+1%). In the same period, the anome-
rization from β-Glc(OAc)5 1a to -Glc(OAc)5 1b is more noticeable and
the
The composition of the reaction mixture was followed over time by
thin layer chromatography (TLC) (supporting information Figure S1).
These analyses show that the glycosylation of the 1,3-PD leads to a
mixture of compounds but can be roughly divided into four fractions. To
investigate their chemical composition, samples of the crude reaction
mixture were taken at different times t and analyzed by 1H NMR spec-
troscopy. As depicted in Fig. 3, 1H NMR spectroscopy can distinguish the
starting ester leaving group of β-Glc(OAc)5 1a with the characteristic H1
proton at 5.7 ppm. The formation of the final product mixture is detected
by the presence of different doublets between 4.4 and 4.7 ppm that
correspond to β-glucosidic links [29]. For information and in order to
remove any form of ambiguity, the peaks at 4.7 ppm are characteristic of
β-glucosidic bonds of homopolymerized glucose present in the F4
fraction.
α
α
/β ratio reaches a final value of 3:1. In this same time interval, the
anomerization of residual Glc(OAc)5 observed conjointly with the con-
stancy of the conversion rate (≈78%) indicates that the consumption of
1,3-PD 2a is total. As a result, an average reaction time of 8 h was
considered as sufficient for achieving 1,3-PD glucosylation.
The conversion of β-Glc(OAc)5 1a and the composition of glucoside
can be also determined by the exploitation of the intensities of signals
characteristic of anomeric carbons in the 13C NMR spectrum (supporting
information Figure S2) as described by Sokolov [22]. In particular, 13
C
NMR confirms in a clearer form the absence of α-glycosidic bonds that
are likely to have a chemical shift around 95 ppm in general [29].
The conversion degree
τ of the glucosylation can be evaluated by
integration of these different peaks. Indeed, it corresponds to the molar
Fig. 3. 1H NMR spectrum (400.1 MHz, CDCl3) for the O-glucosylation of 1,3-PD 2a (left) and table of the calculated conversion rates at times from 2 to 23 h (right).
3

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
2.2.2. Characterization and identification of the different species of the
mixture
The efficiency of MALDI-TOF MS analyses was already shown in
literature for the identification of high mass compounds in product
mixtures based on AlkylPolyGlycoside (APG) models [30]. As regards
our study, the MALDI-TOF MS spectrum characteristic of the mixture
after 23 h of reaction is shown in Fig. 5-a. Each ionized oligomer is
detected on the mass spectrum by a signal made of several peaks cor-
responding to the mass-to-charge ratios m/z characteristic of various
isotopic forms (Fig. 5-b).
The different products resulting from 1,3-PD glycosylation were
separated by silica gel chromatography into four different fractions, F1,
F2, F3 and F4. Then, the chemical structures of the compounds present
in each fraction were carefully investigated using different analytical
techniques.
• NMR analyses
The interest of the MALDI-TOF MS is given by comparing measured
monoisotopic (m/z)exp values for [Mi + Na]+ ions with calculated
values. Indeed, the mass spectrum of the crude mixture (Fig. 5-a) con-
tains five series of cationized adducts [Mi + Na]+ separated by Δm/z =
288 (hexose unit) corresponding to different degrees of glucose homo-
polymerization (HD). The molar mass Mi of each oligomer can be
calculated from the measured value (m/z)exp according to equation (2):
The NMR coupling made it possible the identification of desired
BisGlucosidePropyl octaacetate BGP(OAc)8 3a that is the only constit-
uent of fraction F3. This compound is produced in a low yield (21%)
with essentially β-isomeric form (Fig. 4). The fraction F2 contains a first
by-product identified as MonoGlucosidePropyl pentaacetate (MGP
(
)
(OAc)5) 4a (24%;
with both anomeric forms (
fact a mixture of different oligomers (33%;
α
/β ratio 0:1). F1 consists of residual Glc(OAc)5 (22%)
/β ratio 3:1) 1a,b. The last fraction F4 is in
/β ratio 0:1), the nature of
m
z
α
Mi =
× z ꢀ mcation
2
exp
α
which will be defined later in our study by MALDI-TOF MS techniques.
Most formed products mainly have the β-configuration (supporting in-
formation Figure S3). Such yield of BGP(OAc)8 3a was comparable to
those reported by Xue et al. for the synthesis of 1,12-dodecyl bisga-
lactoside [20].
where mcation is the molar mass of Na (23 g.mol-1)
The adduct [Mi + Na]+ = 759.2 is interpreted as being representative
of the desired bisglucoside BGP(OAc)8 3a of molar mass 736.6 g molꢀ 1
(Fig. 6). The adducts [Mi + Na]+ registered at 413.3 and 471.3 are
characteristic of residual Glc(OAc)5 1a,b (390.3 g molꢀ 1) and MGP
(OAc)5 4a (448.4 g molꢀ 1) respectively. The adducts [Mi + Na]+
=
• MALDI-TOF MS analysis
717.2 and 429.2 are attributed to the presence of mono- and BisGluco-
sideX but under deacetylated form. In our work, all calculated (m/z)theo
Fig. 4. 1H NMR spectrum (CDCl3) of the different fractions: F1:
α
,β-Glc(OAc)5 (1a,b); F2: MGP(OAc)5 (4a); F3: BGP(OAc)8 (3a) and F4: (A, B, C) mixture.
4

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
Fig. 5. MALDI-TOF MS spectrum of the crude mixture obtained after 23 h with DCTB as matrix and NaI as cationizing agent (a). Zoom on the isotopic mass of the
adduct [Mi + Na] + (m/z = 759.2) corresponding to BGP(OAc)8 3a (736.6 g molꢀ 1) (b).
values for [Mi + Na]+ ions including a possible 2-O deprotection are
gathered in supporting information (Table T1) to enable a further
comparison with the experimental data (m/z)exp obtained using MALDI-
TOF MS. On the basis of these calculations, two major populations,
namely, B and C are suspected to be formed during the O-glucosylation
of 1,3-PD 2a. Unfortunately, they cannot be distinguished separately
with MALDI-TOF MS experiments because their oligomers have equiv-
alent m/z values (Figures S3 and S4 in supporting information). The
further exploitation of the MALDI-TOF MS spectrum makes it possible
the identification of the population A that defines oligomers character-
ized by a sequence of peracetylated glucose units but without 1,3-PD
building block in their structure (Fig. 6).
conditions are fulfilled, the intensity of the adducts detected in the mass
spectrum reflects the number of oligomers actually present in the sam-
ple. The crude mixture of oligomers obtained in our research presents
molecular weights lower than 2000 g.molꢀ 1 and low degrees of oligo-
merization ranging from 1 to 6. In other words, they comply with the
criteria for the validity of this technique for quantitative evaluation.
Then, the BGP(OAc)8 that is characterized by a peak at 759.2 with the
greatest intensity, is the most abundant chemical specie in the mixture
by comparison with other oligomers. The relative proportion of Glc
(OAc)5 1a,b and MGP(OAc)5 4a cannot be accurately defined because
their corresponding peaks are located too close to the detection limit
(>500 g.mol-1).
It is usually accepted that MALDI-TOF MS is a quantitative analysis
of oligomers provided that certain requirements are satisfied such as a
polydispersity index PDI (= Mw/Mn) lower than 1.3 [31,32]. If these
• Proposed mechanisms
5

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
Fig. 6. Overview of the different products obtained by O-glucosylation of 1,3-PD 2a with β-Glc(OAc)5 1a and identified according different techniques.
The observation of MGP(OAc)5 4a was coherent with results ob-
tained by Murakami et al. [29]. Indeed, these researchers reported the
formation of bromoalkyl acetate as side-product during the
glucosylation of bromo-alkanols with the formation of 1,2-orthoester
intermediates. This latter topic is particularly well described in Kong’s
work [33]. By analogy, MGP(OAc)5 4a that was isolated is likely the
Fig. 7. Intermediate products of the reaction involved in the formation of MGP(OAc)5 4a.
6

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
result of the conversion of an orthoester 1 into MGP(OAc)3(OH)2 6a and
3-hydroxy-propyl acetate 2′a (not isolated). Such mechanism is sche-
matically drawn in Fig. 7.
peculiarity decreases the hydroxyl charge and affects the value of its
corresponding pKa (14.4). CHD 2e presents a pKa value close to that of
1,3-PD 2a (pKa ≈15) but both diols differ from each other with their
respective steric effects.
Please note that 3-hydroxy-propyl acetate 2′a can itself be poten-
tially glycosylated to give the MGP(OAc)5 4a in accordance with the
results of Banoub et al. and others [34,35]. This mechanism suggests the
formation of monoglucosides MGP(OAc)3(OH)2 6a and MGP
(OAc)3(OH)(OR) 7a that are deacetylated in position 2. It is very likely
that these intermediates are responsible of the formation of oligomers.
To summarize, the higher the pKa is, the higher diol reactivity is. But,
considering our own data, a high reactivity is not synonym of bisglu-
coside production with high yield. On the contrary, a lower diol reac-
tivity seems preferable for obtaining a better selectivity. When the pKa
value is low as encountered with ByD 2d, the hydroxyl charge density is
reduced. Then, the hydroxyl units are involved into a direct nucleophilic
attack on C-1 (Fig. 7 - pathway I), which favors the formation of MGX
(OAc)4(OH) (5a-f) and then BGX(OAc)8 (3a-f). Conversely, with a high
pKa value as found with 1,3-PD 2a, an increased charge density is
centered into hydroxyl units. Then, the attack of the carbonyl carbon
present in the acyloxonium ion (Fig. 7 - pathway II) is privileged with
the formation of intermediates orthoesters.
2.3. Synthesis of the other BisGlucosideX derived from (cyclo)aliphatic
diols
Other bisglucosides were synthesized from the respective reaction of
aliphatic diols 2b-2e (Fig. 2) with β-Glc(OAc)5 1a using the same
chemical pathway as described hereabove with 1,3-PD 2a. The results
issued from their chemical analyses (1H, 13C NMR and MALDI-TOF MS)
are available in supporting information for preserving the clarity of this
paper. In all cases, the formation of by-products is observed in addition
to the desired bisglucoside production. The respective reaction yields of
these compounds are summarized in Table 1. Their proportion in the
final mixture appear to be dependent on the nature of the diol used for
O-glucosylation. As concerns more particularly the bisglucoside BGX
(OAc)8 synthesis, the highest reaction yield value is observed with the
use of 2-butyne-1,4-diol 2d (67%) whereas the reverse case is observed
with propane-1,3-diol 2a (21%). A first possible hypothesis to explain
the discrepancies observed between diols used is based on their reac-
tivity, i.e. in our study, their ability to react with peracetylated gluco-
sidic units. This characteristic is directly related to the nucleophilicity of
the hydroxyl groups, which is itself influenced by the nature of the
chemical skeleton that bears them.
With a pKa value similar to that of 1,3-PD 2a, CHD 2e led to the
production of BisGlucosideCyclohexyl octaacetate (BGC(OAc)8) 3e with
a reaction yield close to that of BGP(OAc)8 3a (25% against 21%).
Nevertheless, the low content of MonoGlucosideCyclohexyl pentaace-
tate (MGC(OAc)5) 4e (12%) seems to indicate that 1,2-orthoesters
pathway is not the only one to be involved in this reaction.
Among our results, it is interesting to note that, despite of its
nucleophilicity value comparable to that of BnD 2b, DEG 2c led to the
production of bisglucoside 3c with a higher efficiency (yield of 45%)
than that of BGBn(OAc)8 3b (36%). This peculiarity is likely due to the
presence of an ether group in the chain (2c). This group is likely to have
a better interaction with BF3 than an aliphatic diol bearing unsaturation
such as found with 2b or 2d diols.
2.4. Synthesis of the BisGlucosideHydroquinone octaacetate (BGH
(OAc)8)
In the present study in which we compare the reactivity of the same
family of molecules, (i.e. dihydroxy compounds), the nucleophilicity can
be evaluated from the base strength with approximate acid dissociation
constant pKa [36–38]. The corresponding values range from 15.1 (for 1,
3-PD 2a) to 12.5 (for ByD 2d). It is well known that a sp hybridized
carbon as found in (ByD 2d) is less electron-donating than a sp2 (BnD
2b), which in turn is less electron-donating than a sp3 carbon (1,3-PD
2a). Such hierarchy is well illustrated by their respective pKa values
arranged in ascending order. DEG 2c presents an oxygen atom in its
aliphatic sequence and that has an electron-withdrawing effect. This
The interest of hydroquinone was explored for the synthesis of
diphenol-based bisglucosides by following the same experimental pro-
cedure as that used previously with (cyclo)aliphatic diols 2a-2e. The
products isolated from the glucosylation mixture were analyzed by
means of 1H NMR and MALDI-TOF MS (see supporting information).
In strong contrast with the analyses previously registered with
(cyclo)aliphatic diols, MALDI-TOF MS analysis of the crude product
obtained from O-glucosylation with hydroquinone 2f reveals the only
presence of three m/z peaks. The first one can be assigned to the residual
Glc(OAc)5 1a,b (390.3 g molꢀ 1) being ionized by a sodium cation. The
other two peaks can be attributed to the monoglucoside MGH
(OAc)4(OH) 5f (440.4 g molꢀ 1) and desired bisglucoside BGH(OAc)8 3f
(770.7 g molꢀ 1). Both latter compounds result from direct nucleophilic
attack of 2f on C-1.
Table 1
BF3⋅OEt2 catalyzed glucosylation of diols (2a-e) with β-Glc(OAc)5 (1a).
Entry
Diol
pKaa
15.1
14.5
14.4
12.5
15
Yieldb
Bisglucoside
3a (21%)
By-products
The selectivity of the HQ 2f glycosylation seems quite good without
the parasitic formation of orthoesters or 2-O–deacetylated by-products.
Once more, this peculiarity can be explained on the basis of nucleophilic
considerations. Indeed, HQ presents two hydroxyl groups directly linked
to the aromatic ring in para position. The charge delocalization is much
reduced than that encountered with the dihydroxy compounds previ-
ously investigated. In other words, HQ is less nucleophilic than (cyclo)
aliphatic diols 2a-e as shown by its pKa value close to 10.2.
1
2
3
4
5
2a
1 (22%)
4a (24%)
A, B, C (33%)
1 (15%)
2b
2c
3b (36%)
3c (45%)
3d (67%)c
3e (25%)
4b (13%)
A, B, C (36%)
1 (22%)
4c (15%)
A, B, C (18%)
1 (9%)
2d
2e
The extraction of each chemical specie from the mixture would make
their respective proportion in the mixture clear. In particular, the
4d (6%)
A, B, C (18%)
1 (32%)
desired BGH(OAc)8 3f was produced in moderate yield (53%; α/β ratio
0:1). The proportion of MonoGlucosideHydroquinone (MGH
(OAc)4(OH)) 5f was evaluated as being close to 22% ( /β ratio 0:1)
while that of residual Glc(OAc)5 1a,b was found close to 25% ( /β ratio
4e (12%)
α
A, B, C (31%)
α
Experimental conditions: molar ratio 1:0.5:1 (β-Glc(OAc)5/HO-X-OH/BF3⋅OEt2)
1:1). The yield obtained with BGH(OAc)8 3f can appear weak but it
remains in the same range of order of that observed by Bergeron-Brlek
[19] on the synthesis of a bisglucoside from bisphenol A BPA (49%).
This similarity is likely a consequence of the pKa values of these
diphenols that are evaluated as being rather close (10.2 for HQ against
in CH2Cl2, r.t. - 23 h.
a
Approximative pKa data.
b
Yields of products isolated after column chromatography based on crude
reaction mixture.
c
Reported yield: 34% [27].
7

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
9.6 for BPA).
Bruker Ultra-Flex III (2004) (Bremen, Germany) equipped with a ni-
trogen laser (337 nm, 30 ns) and a detector. Peptide mixtures were used
for external calibration. The ions were accelerated with a potential of 25
kV. The measurements were performed in positive mode POS. The
spectra were obtained with trans-2-[3-(4-tertbutylphenyl)-2-methyl-2-
propenylidene]malononitrile (DCTB) as matrix and NaI as ionizing salt.
High resolution electrospray ionization mass spectrometry (HRMS-
ESI) analyses were obtained using a Bruker MicroTOF QII mass spec-
trometer. The experiments were carried out in positive ion mode at high
resolution.
2.5. Overall hierarchy
Considering our results and their interpretation, the different diols
and phenol studied in this research can be ranked in decreasing order
considering the selectivity of their reactivity with glucose via a direct
attack on the anomeric carbon C-1 (previously defined as pathway I).
The corresponding hierarchy is built as following: aromatic > aliphatic
(sp) > ethoxylated > aliphatic (sp2) > cycloaliphatic > aliphatic (sp3)
(Fig. 8).
Thin layer chromatography analyses (TLC) of the reaction mixtures
of BGX(OAc)8 were carried out on silica gel plates with a mixture of
CH2Cl2/EA (7/3; v/v) as eluent. The various compounds were visualized
by spraying the TLC plates with a solution of sulfuric acid (10% by
volume in ethanol), followed by charring at 150 ◦C for a few minutes.
Spots were compared to standards.
The influence of the nucleophilicity of the dihydroxy compounds
upon the selectivity of their reaction with glucose units seems now clear.
In this present case, nucleophilicity can be correlated to the base
strength (pKa value). The best selectivity is obtained with the compound
(diphenol) characterized by the lowest pKa value, or in other words,
lowest nucleophilicity. This observation seems a direct consequence of
its lowest reactivity due to electronic charge delocalization that prevents
the formation of by-products.
Column chromatographic separations were carried out with a silica
gel from VWR (40–63 μm). The various products of the mixtures were
isolated with the same eluent as the TLC.
The thermo-oxidative stability of each bisglucoside was examined
using a Q50 thermogravimetric analyzer (TGA) from TA Instruments®
(New Castle, DE, USA). The experiments consisted in registering the
weight loss of the sample as a function of temperature from ambient up
to 600 ◦C with a heating rate of 10 ◦C.minꢀ 1 under air flow (25 mL
minꢀ 1). A relative weight loss (RWL) value of 5% was chosen as a
reasonable criterion to evaluate the thermal degradation temperature
3. Materials and methods
3.1. Materials
β-glucose pentaacetate (β-Glc(OAc)5), 1,3-propanediol (1,3-PD),
hydroquinone (HQ), diethylene glycol (DEG), 2-butyne-1,4-diol (ByD),
cis-2-butene-1,4-diol (BnD), 1,4-cyclohexanediol (CHD), boron tri-
fluoride diethyl etherate (BF3⋅OEt2; ≥46.5% BF3 basis) were purchased
from Aldrich (France). Anhydrous dichloromethane (CH2Cl2), ethyl ac-
etate (EtOAc or EA) were purchased from Analytic Lab. Deuterated
chloroform (CDCl3) were purchased from SDS. All these chemicals were
used without further purification.
Tdeg of each bisglucoside.
Differential scanning calorimetric experiment (DSC) were registered
using a Mettler Toledo DSC Star One (Greifensee, Switzerland), under air
and with a heating rate of 5 ◦C.minꢀ 1. Before analysis, the sample was
cooled down from ambient to 0 ◦C with a ramp of ꢀ 3 ◦C.minꢀ 1 and
maintained at this constant temperature during 15 min to insure thermal
equilibrium. The thermogram characteristic of each bisglucoside anal-
ysis presented an endotherm. Then, the value of the corresponding
melting temperature Tm was taken at the inception of the endothermic
peak.
3.2. Techniques
Nuclear magnetic resonance (NMR) spectra were recorded on a
Bruker Avance I 300 MHz or a Bruker Avance III 400 MHz spectrometers
at room temperature with CDCl3 as the solvent. The chemical shifts
measured in part per million (ppm) were referenced to the CDCl3 signal
peak at 7.26 ppm (1H) and 77.16 ppm (13C). Coupling constants (J) are
reported in Hz.
3.3. Synthesis
3.3.1. Standard O-glucosylation protocol
β-Glc(OAc)5 1a (3.00 g, 7.68 mmol) and diol or diphenol (HO-X-OH)
2 (3.84 mmol) were mixed in anhydrous CH2Cl2 (50 mL). Then BF3⋅OEt2
MALDI-TOF MS analyses were performed using a MALDI-TOF MS
Fig. 8. Influence of diol/diphenol nucleophilicity on O-glucosylation mechanisms.
8

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
(0.950 mL, 7.68 mmol) was added dropwise. The molar ratio β-Glc
(OAc)5/HO-X-OH/BF3⋅OEt2 is 1:0.5:1. The mixture was stirred at room
temperature for 23 h under inert atmosphere (nitrogen), and then
quenched by saturated aq. NaHCO3. The organic layer was separated,
washed successively with aq. NaHCO3 and (once) with brine solution,
dried over MgSO4, and concentrated. BGX(OAc)8 and other products are
isolated by column chromatography (eluent CH2Cl2/EA, 7/3) or by
recrystallization (ethanol), depending on the nature of the BGX(OAc)8.
Remarks. The glucosylation reactions were also carried out for
different quantities up to 15 g. The products were obtained in similar
yields and selectivity.
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 221 ◦C.
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 5.66 (m, 2H, H-2); 5.19 (t, J
= 9.5, 2H, H-3′); 5.06 (t, J = 9.7, 2H, H-4′); 4.95 (t, J = 8.9, 2H, H-2′);
4.49 (d, J = 8.0, 2H, H-1′β); 4.32 (m, 2H, H-1b); 4.25 (dd, 2H, H-6′b);
4.18 (m, 2H, H-1a); 4.10 (dd, 2H, H-6′a); 3.68 (m, 2H, H-5′); 1.97–2.06
(4s, 24H, CH3CO).
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.61, 170.21, 169.39,
′
′
–
169.28 (O–C O); 128.71 (C-2); 99.43 (C-1 β); 72.76 (C-3 ); 71.80 (C-
–
5′); 71.20 (C-2′); 68.29 (C-4′); 64.47 (C-1); 61.79 (C-6′); 20.69 (CH3CO).
HRMS (ESI): m/z calculated for C32H44O20 [M+Na]+: 771.2324;
Found: 771.2311.
3.3.2. β-BisGlucosidePropyl octaacetate (β-BGP(OAc)8) (3a)
3.3.5. β-BisGlucosideDiEthyleneGlycol octaacetate (β-BGDEG(OAc)8)
1,5-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy)-3-oxa-
pentane (3c)
Compound
(3a)
1,3-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyr-
anosyloxy)propane.
The reaction was carried out as described in the standard protocol
using 0.292 g (3.84 mmol) 1,3-PD 2a. The product 3a was isolated after
purification by column chromatography.
The reaction was carried out as described in the standard protocol
using 0.408 g (3.84 mmol) DEG 2c. The product 3c was isolated after
purification by column chromatography.
Yield: 0.594 g (21%).
Yield: 1.326 g (45%).
Rf: 0.59 (CH2Cl2/EtOAc; 7/3)
Rf: 0.30 (CH2Cl2/EtOAc; 7/3)
DSC (5 ◦C.minꢀ 1), Tm: 146 ◦C.
DSC (5 ◦C.minꢀ 1), Tm: 117 ◦C.
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 234 ◦C.
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 267 ◦C.
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 5.18 (t, J = 9.5, 2H, H-3′);
5.06 (t, J = 9.7, 2H, H-4′); 4.96 (t, J = 9.6, 2H, H-2′); 4.45 (d, J = 7.9,
2H, H-1′β); 4.25 (dd, J = 4.8, 12.3, 2H, H-6′b); 4.11 (dd, J = 2.3, 12.3,
2H, H-6′a); 3.86 (dt, J = 5.6, 10.0, 2H, H-1b); 3.68 (m, 2H, H-5′); 3.58
(dt, J = 6.7, 9.7, 2H, H-1a); 1.99–2.07 (4s, 24H, CH3CO); 1.83 (m, 2H, H-
2).
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 5.18 (t, J = 9.5, 2H, H-3′);
5.05 (t, J = 9.7, 2H, H-4′); 4.95 (t, J = 9.6, 2H, H-2′); 4.56 (d, J = 8.0,
2H, H-1′β); 4.23 (dd, J = 4.7, 12.3, 2H, H-6′b); 4.11 (dd, J = 2.3, 12.3,
2H, H-6′a); 3.90 (dt, J = 4.2, 10.7, 2H, H-1b); 3.68 (m, 2H, H-5′); 3.67
(m, 2H, H-1a); 3.59 (m, 4H, H-2); 1.97–2.05 (4s, 24H, CH3CO).
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.74, 170.35, 169.51,
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.60, 170.21, 169.38,
169.41 (O–C O); 100.95 (C-1 β); 72.91 (C-3 ); 71.91 (C-5 ); 71.37 (C-
′
′
′
–
–
′
′
′
169.22 (O–C O); 101.03 (C-1 β); 72.82 (C-3 ); 71.84 (C-5 ); 71.43 (C-
2′); 70.46 (C-2); 69.23 (C-1); 68.51 (C-4′); 62.05 (C-6′); 20.70–20.84
–
–
2′); 68.47 (C-4′); 66.61 (C-1); 61.98 (C-6′); 29.90 (C-2); 20.57–20.70
(CH3CO).
(CH3CO).
HRMS (ESI): m/z calculated for C32H46O21 [M+Na]+: 789.2429;
HRMS (ESI): m/z calculated for C31H44O20 [M+Na]+: 759.2324;
Found: 789.2417.
Found: 759.2315.
3.3.6. β-BisGlucosideButyne octaacetate (β-BGBy(OAc)8)
1,4-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy)-2-butyne
(3d)
3.3.3. β-BisGlucosideCyclohexane octaacetate (β-BGC(OAc)8)
1,4-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy)cyclo-
hexane (mixture of cis and trans) (3e)
The reaction was carried out as described in the standard protocol
using 0.331 g (3.84 mmol) ByD 2d. The product 3d was isolated after
purification by column chromatography.
The reaction was carried out as described in the standard protocol
using 0.446 g (3.84 mmol) CHD 2e. The product 3e was isolated after
purification by column chromatography.
Yield: 1.922 g (67%).
Yield: 0.746 g (25%).
Rf: 0.62 (CH2Cl2/EtOAc; 7/3)
Rf: 0.62 (CH2Cl2/EtOAc; 7/3)
DSC (5 ◦C.minꢀ 1), Tm: 121 ◦C.
DSC (5 ◦C.minꢀ 1), Tm: 180 ◦C.
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 267 ◦C.
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 5.21 (t, J = 9.4, 2H, H-3′);
5.08 (t, J = 9.7, 2H, H-4′); 4.97 (t, J = 9.5, 2H, H-2′); 4.70 (d, J = 7.9,
2H, H-1′β); 4.38 (m, 4H, H-1); 4.25 (dd, J = 4.5, 12.4, 2H, H-6′b); 4.12
(dd, J = 2.3, 12.4, 2H, H-6′a); 3.72 (m, 2H, H-5′); 1.98–2.06 (4s, 24H,
CH3CO).
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 283 ◦C.
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 5.17 (t, J = 10.0, 2H, H-3′);
5.03 (t, J = 9.0, 2H, H-4′); 4.93 (t, J = 9.3, 2H, H-2′); 4.53 (d, J = 9.3,
2H, H-1′β); 4.23 (dd, J = 4.0, 11.3, 2H, H-6′b); 4.08 (dd, J = 12.2, 2H, H-
6′a); 3.68 (m, 2H, H-1); 3.68 (m, 2H, H-5′); 1.30–1.94 (m, 8H, H-2);
1.98–2.05 (4s, 24H, CH3CO).
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.67, 170.27, 169.45,
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.72, 170.41, 169.50,
169.42 (O–C O); 98.39 (C-1 β); 81.84 (C-2);72.76 (C-3 ); 71.98 (C-5 );
′
′
′
–
–
′
′
169.27 (O–C O); 99.85, 98.32 (C-1 β); 76.50 (C-1); 72.96 (C-3 ); 71.84
71.15 (C-2′); 68.33 (C-4′); 61.83 (C-6′); 56.22 (C-1); 20.66–20.80
–
–
(C-5′); 71.64 (C-2′); 68.68 (C-4′); 62.21 (C-6′); 26.42, 27.34, 28.39,
(CH3CO).
28.87 (C-2); 20.73–20.84 (CH3CO).
HRMS (ESI): m/z calculated for C32H42O20 [M+Na]+: 769.2167;
HRMS (ESI): m/z calculated for C34H48O20 [M+Na]+: 799.2637;
Found: 799.2629.
Found: 769.2150.
3.3.7. β-BisGlucosideHydroquinone octaacetate (β-BGH(OAc)8)
1,4-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy)benzene
(3f)
3.3.4. β-BisGlucosideButene octaacetate (β-BGBn(OAc)8)
(Z)-1,4-bis(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy)-2-
butene (3b)
The reaction was carried out as described in the standard protocol
using 0.423 g (3.84 mmol) HQ 2f. The product 3f was isolated after
purification by column chromatography or by recrystallization
(ethanol).
The reaction was carried out as described in the standard protocol
using 0.339 g (3.84 mmol) BnD 2b. The product 3b was isolated after
purification by column chromatography.
Yield: 1.036 g (36%).
Yield: 1.570 g (53%).
Rf: 0.54 (CH2Cl2/EtOAc; 7/3)
Rf: 0.64 (CH2Cl2/EtOAc; 7/3)
DSC (5 ◦C.minꢀ 1), Tm: 127 ◦C.
DSC (5 ◦C.minꢀ 1), Tm: 185 ◦C.
9

S. Patry et al.
Carbohydrate Research 500 (2021) 108217
TGA (10 ◦C.minꢀ 1 – RWL = 5%), Tdeg: 311 ◦C.
this study by the French Ministry of Higher Education, Research and
1H NMR (CDCl3, 400.1 MHz): δ (ppm) = 6.92 (s, 4H, H-1); 5.27 (t, J
= 9.2, 2H, H-3′); 5.23 (t, J = 9.3, 2H, H-2′); 5.15 (t, J = 9.3, 2H, H-4′);
4.98 (d, J = 7.5, 2H, H-1′β); 4.28 (dd, J = 5.1, 12.3, 2H, H-6′b); 4.16 (dd,
J = 2.5, 12.3, 2H, H-6′a); 3.81 (m, 2H, H-5′); 2.02–2.07 (4s, 24H,
CH3CO).
Innovation through the award of a thesis scholarship to Stephane Patry.
´
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.carres.2020.108217.
13C NMR (CDCl3, 100.6 MHz): δ (ppm) = 170.64, 170.34, 169.49,
′
–
169.38 (O–C O); 152.97 (C-2); 118.57 (C-1); 99.96 (C-1 β); 72.82 (C-
–
3′); 72.18 (C-5′); 71.34 (C-2′); 68.41 (C-4′); 62.02 (C-6′); 20.73–20.82
References
(CH3CO).
HRMS (ESI): m/z calculated for C34H42O20 [M+Na]+: 793.2167;
[1] J.M. Raquez, M. Deleglise, M.F. Lacrampe, P. Krawczak, Progress in Polymer
Science 35 (2010) 487–509.
Found: 793.2153.
[2] A. Gandini, Green Chem 13 (2011) 1061–1083.
[3] J.C. Ronda, G. Lligadas, M. Galia, V. Cadiz, Reactive & Functional Polymers 73
(2013) 381–395.
4. Conclusions
[4] V. Flaris, G. Singh, Journal of Vinyl & Additive Technology 15 (2009) 1–11.
[5] F.W. Lichtenthaler, S. Peters, Comptes Rendus Chimie 7 (2004) 65–90.
[6] H.J. van, E.A. Oostveen, F. Micciche, B.A.J. Noordover, C.E. Koning, B.R.A.T.
M. van, A.E. Frissen, J.G.J. Weijnen, J. Coat. Technol. Res. 4 (2007) 177–186.
[7] A.J. Thomas, in: Huntsman Ici Chemicals, Llc, USA, 2000, p. 31.
[8] N.D. Sachinvala, M.H. Litt, D. Winsor, in: USA, 2004, p. 37.
[9] X. Pan, P. Sengupta, D.C. Webster, Green Chem 13 (2011) 965–975.
[10] X. Pan, D.C. Webster, Macromol. Rapid Commun. 32 (2011) 1324–1330.
[11] L.K. Mydock, A.V. Demchenko, Org. Biomol. Chem. 8 (2010) 497–510.
[12] S. Hotha, S. Kashyap, J. Org. Chem. 71 (2006) 364–367.
[13] D. Crich, F. Yang, Angew. Chem., Int. Ed. 48 (2009) 8896–8899.
[14] V. Neto, R. Granet, P. Krausz, Tetrahedron 66 (2010) 4633–4646.
[15] C. Mukherjee, K. Ranta, J. Savolainen, R. Leino, Eur. J. Org. Chem. (2012)
2957–2968, 2012.
This article demonstrates that the production of bisglucosides
through O-glucosylation of a dihydroxy compound is not a trivial topic
at all. Indeed, depending on the nature of the diol/diphenol used,
different by-products can be formed in addition to the desired bisglu-
coside. On the basis of our results, the nucleophilicity of the dihydroxy
compound related to the value of its pKa seems an important factor to
control the reaction yield and the potential formation of side-products.
Among the different molecules investigated, the hydroquinone-based
monomer is identified as the most promising compound for the pro-
duction of bisglucosides. Indeed, due to its low nucleophilicity, its re-
action with acetylated glucose units is characterized by a good
selectivity without glucose homopolymerization. Even if the reaction
yield is limited, the purification process is easy to perform and the
overall production remains economical. Reminding that this research
was initially supported by the wish of producing biobased monomers,
that based on hydroquinone is undeniably promising. Indeed, β-BGH
(OAc)8 presents a high thermostability (Tdeg > 300 ◦C) through the
presence of an aromatic cycle set at the core of the bisglucoside chemical
skeleton. Moreover, this critical temperature is somewhat 120 ◦C higher
than the melting temperature. Thus, this temperature gap seems large
enough to make the conduction of further functionalization in the
molten state possible i.e. without the use of any solvent.
[16] R. Dominique, S.K. Das, B. Liu, J. Nahra, B. Schmor, Z. Gan, R. Roy, Methods
Enzymol 362 (2003) 17–28.
[17] R. Roy, S.K. Das, F. Santoyo-Gonzalez, F. Hernandez-Mateo, T.K. Dam, C.F. Brewer,
Chem. - Eur. J. 6 (2000) 1757–1762.
[18] P. Collins, R. Ferrier, Monosaccharides: Their Chemistry and Their Roles in Natural
Products, Wiley, 1995.
[19] M. Bergeron-Brlek, T.C. Shiao, M.C. Trono, R. Roy, Carbohydr. Res. 346 (2011)
1479–1489.
[20] J.L. Xue, S. Cecioni, L. He, S. Vidal, J.-P. Praly, Carbohydr. Res. 344 (2009)
1646–1653.
[21] Y.S. Lee, E.S. Rho, Y.K. Min, B.T. Kim, K.H. Kim, J. Carbohydr. Chem. 20 (2001)
503–506.
[22] V.M. Sokolov, V.I. Zakharov, E.P. Studentsov, Russ. J. Gen. Chem. 72 (2002)
806–811.
[23] K. Toshima, K. Tatsuta, Chem. Rev. 93 (1993) 1503–1531.
[24] S. Mabic, C. Benezra, J.P. Lepoittevin, Tetrahedron Lett 34 (1993) 4531–4534.
[25] R.U. Lemieux, W.P. Shyluk, Can. J. Chem. 31 (1953) 528–535.
[26] S. Gerber, V.M. Garamus, G. Milkereit, V. Vill, Langmuir 21 (2005) 6707–6711.
[27] G.B. Giovenzana, L. Lay, D. Monti, G. Palmisano, L. Panza, Tetrahedron 55 (1999)
14123–14136.
Author contributions
[28] M. Jacobsson, J. Malmberg, U. Ellervik, Carbohydr. Res. 341 (2006) 1266–1281.
[29] T. Murakami, R. Hirono, Y. Sato, K. Furusawa, Carbohydr. Res. 342 (2007)
1009–1020.
´
This research was led within the context of Stephane Patry’s Ph-D
thesis under the scientific direction of Prof. Jean-Pierre Habas. Dr
Mike Robitzer was helpful for the understanding the different reaction
mechanisms and contributed to the supervision of this work.
[30] A.V. Kuehn, R.H.H. Neubert, Pharm. Res. 21 (2004) 2347–2353.
[31] C. Ladaviere, P. Lacroix-Desmazes, F. Delolme, Macromolecules 42 (2009) 70–84.
[32] G. Montaudo, M.S. Montaudo, C. Puglisi, F. Samperi, Rapid Commun. Mass
Spectrom. 9 (1995) 453–460.
[33] F. Kong, Carbohydr. Res. 342 (2007) 345–373.
Declaration of competing interest
[34] J. Banoub, D.R. Bundle, Can. J. Chem. 57 (1979) 2091–2097.
[35] U. Aich, D. Loganathan, Carbohydr. Res. 341 (2005) 19–28.
[36] J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, Oxford
University Press, 2000.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[37] S. Zhang, J. Baker, P. Pulay, J. Phys. Chem. A 114 (2010) 432–442.
[38] R.D. Rossi, J. Chem. Educ. 90 (2013) 183–190.
Acknowledgments
The authors are very grateful for the financial support provided to
10