Relative reactivities in the O-methylation of glucomannans: The influence of stereochemistry at C-2 and the solvent effect

Article abstract of DOI:10.1016/j.carres.2014.06.005

The main hemicellulose in softwood, glucomannan (GM), structurally resembles cellulose but has quite different physical and chemical properties. In addition to branching and original acetylation, the only other difference between these two β-1,4-linked gl

Full text of DOI:10.1016/j.carres.2014.06.005

Accepted Manuscript
Relative Reactivities in the O-Methylation of Glucomannans: The Influence of
Stereochemistry at C-2 and the Solvent Effect
Yujia Zhang, Jiebing Li, Mikael E. Lindström, Petra Mischnick
PII:
S0008-6215(14)00243-2
DOI:
http://dx.doi.org/10.1016/j.carres.2014.06.005
CAR 6766
Reference:
Carbohydrate Research
To appear in:
Received Date:
Revised Date:
Accepted Date:
17 April 2014
27 May 2014
6 June 2014
Please cite this article as: Zhang, Y., Li, J., Lindström, M.E., Mischnick, P., Relative Reactivities in the O-
Methylation of Glucomannans: The Influence of Stereochemistry at C-2 and the Solvent Effect, Carbohydrate
Research (2014), doi: http://dx.doi.org/10.1016/j.carres.2014.06.005
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

RELATIVE REACTIVITIES IN THE O-METHYLATION OF GLUCOMANNANS: THE INFLUENCE OF
STEREOCHEMISTRY AT C-2 AND THE SOLVENT EFFECT
Yujia Zhang1, Jiebing Li1,3, Mikael E. Lindström1,3, Petra Mischnick1,2*
1
Department of Fiber and Polymer Technology, Royal Institute of Technology, KTH, Teknikringen 56-58,
SE-100 44, Stockholm, Sweden
2
Department of Food Chemistry, Technische Universität Braunschweig, Schleinitzstraße 20,
D-38106 Braunschweig, Germany
3
Wallenberg Wood Science Center, The Royal Institute of Technology and Chalmers University of Technology,
SE-100 44 Stockholm, Sweden
* Corresponding author: p.mischnick@tu-braunschweig.de
Abstract
The main hemicellulose in softwood, glucomannan (GM), structurally resembles cellulose
but has quite different physical and chemical properties. In addition to branching and
original acetylation, the only other difference between these two β-1,4-linked glycans is the
configuration at C-2 in approximately 80% of the sugar residues. In contrast to glucose, the
2-OH in mannose has an axial orientation. The influence of this stereochemistry on the
relative reactivities of glucosyl compared to mannosyl units in methylation reactions are
studied in this work.
Glucomannan isolated from spruce (SGM) and commercially available konjac glucomannan
(KGM) were methylated in DMSO/Li-dimsyl/MeI and water/NaOH/MeI system, respectively.
In the early stage of the reaction, the glucose part of the SGM achieved slightly higher DS
values than the mannose residues, but the overall relative rate constants were close to 1:1.
The order of reactivities in glucose was k2 > k3 > k6 and k3 > k2 > k6 for mannose (in DMSO/Li-
dimsyl/MeI). The rate constants did not remain constant, but k3 decreased when k2
increased for both epimeric sugars. In water/NaOH/MeI, the methylation of the primary 6-
OH was much more pronounced with an order of reactivity of O-6 > O-2 > O-3 for mannose
and O-2 > O-6 > O-3 for glucose. The results are discussed with respect to the OH-acidity and
the stereoelectronic, sterical and solvent effects.
Keywords
Spruce glucomannan, Relative reactivities, Partial methylation, Solvent effect
1

INTRODUCTION
Norway spruce (Picea abies) grows abundantly in Scandinavian countries and is composed of
long fibers that are primarily made up of cellulose, hemicelluloses and lignin. As a typical
softwood species, it contains a substantial amount of hemicellulose, which is mainly
composed of mannans. Spruce mannans consist of β-1,4-D-Manp and β-1,4-D-Glcp residues
with D-Galp and acetyl side-groups. There are two types of mannans according to different
monosaccharide ratios: one is galactose-rich with a galactose to glucose to mannose ratio of
1:1:3, termed galactoglucomannan (GGM), and the other is galactose-poor with a ratio of
0.1:1:3, termed spruce glucomannan (SGM)1. Compared to Konjac glucomannan (KGM),
from the herb Amorphophallus konjac, SGM is of lower molecular weight and has a higher
mannose: glucose ratio ca. 3.8: 1 (KGM ca. 1.6: 1). Both SGM and KGM show a certain
degree of branching at O6 of the mannose-glucose backbone. Recently, KGM was shown to
be branched at the glucosyl units and bear mannose-rich oligomeric side chains2; however,
different opinions exist in the literature.
Esterification and etherification of hydroxyl groups are common methods to modify sugar
molecules to equip natural carbohydrates with new functions and characteristics.
Glucomannans and their derivatives, blends and composites have been used in the
pharmaceutical, food, cosmetics and painting industries as solubilizers, adhesives, or drug
delivery substrates.3, 4 By substitution of the hydroxyl groups, the properties of the native
polysaccharide can be varied over a wide range. Xu et al. 5 studied carboxymethylation of a
galactoglucomannan in aqueous NaOH/isopropanol. Based on the GC-MS of monosaccharide
derivatives and the NMR spectra, they conclude that the accessibility of O-6 is the highest
while O-2 and O-3 are shielded by carboxymethyl substituents at O-6. Another study from
the same group was aimed at enhancing the properties of spruce galactoglucomannan by
acetylation. When applying different reaction conditions, the migration of acetyl groups
from O-2 to O-3 to O-6 occurs, whether this happens directly or stepwise and within or
between different units is not known 6. An et al. reported the methylation of KGM to obtain
water-soluble derivatives using aqueous NaOH/MeI at pH 8-10 and studied methyl KGM
1
13
7
using H, C and 2D NMR spectroscopy in D2O at 80 °C . Degree of substitution (DS) values
of 0.09 - 0.18 were estimated, but the positions of the methyl groups on glucose and
mannose could not be differentiated. Because the assignment of OCH3 is contradictory, the
order of reactivity cannot be deduced from the signal. Vopel et al. developed cross-linked
hydrogels based on acetylated galactoglucomannan. The DS was tuned by varying the steric
requirements of the imidazole-activated alkenyl reagents 8.
2

Fig. 1
Proposed structures for spruce glucomannan and its partially methylated derivatives. Approximately
20% of the units have the glucosyl-configuration, and 80% have the mannosyl configuration.
Fig. 1 shows the proposed chemical structure of SGM. Depending on the source, there are
2
variations in the degree of acetylation and branching 9-11. As reported by Albrecht et al. ,
little is known about the precise structure of glucomannans, especially the sequence of
glucose and mannose within the backbone. In the work from Katsuraya et al.12, the main
13
chain and the branched structure of KGM were investigated using C NMR. They concluded
that KGM was composed of β-1,4-linked D-glucosyl and D-mannosyl residues comprising the
backbone with branches through β-1,6-glucosyl units at approximately 8%, but the length of
the branches is not known and may be longer than one sugar unit.
The backbone could be considered a stereoisomer of cellulose with a partial inversion at C-2.
Due to the different orientation of OH at C-2, which is axial in mannose and equatorial in
glucose, the relative reactivities of the free hydroxyl groups at the 2, 3, and 6 positions of
glucose and mannose are expected to be different. A tunable regioselectivity during the
mono-etherification of the 2,3 diol of a mannose derivative has been recently published13.
In addition, the overall reactivities of glucosyl compared to mannosyl units are of interest.
The regioselectivities will not only influence the substituent distribution in the monomer
units but also the substituent pattern over the polymer chain and, thus, the physical
properties 14. Williamson-type etherification using methyl iodide (MeI) was chosen as the
model reaction to study the relative reactivities of OH in mannosyl and glucosyl residues of
GM because the steric requirements of the reagent are low, etherification is irreversible, and
quantitative analysis on various structural levels is well established for the corresponding
cellulose ethers 15, 16
.
For a better understanding of the mechanisms and directed synthesis, it is essential to study
the reactivity of the hydroxyl groups on different positions of the sugar units under various
reaction conditions. The relative reactivities of the free hydroxyl groups in cellulose, dextran,
amylose and cyclodextrin derivatives have been determined by studying their etherification
patterns 16-18. In contrast to polysaccharides composed of a single monosaccharide, the
study of the reactivity of glucomannans is more complicated. Therefore, the relative
reactivities of glucosyl compared to mannosyl units in etherification reactions need to be
discussed for further application of this material. To the best of our knowledge, no work
3

regarding the individual reactivities of glucose and mannose in glucomannan has been
reported so far.
A highly purified SGM 1 isolated from Norway spruce has been used in this study, and a
method for evaluating the methylated glucomannans has been established. With SGM, the
reactions were performed in homogeneous systems with DMSO as an aprotic polar solvent,
the corresponding anion Li-dimsyl as the base and MeI as the methylating agent 19, 20. KGM
was reacted in water with NaOH and MeI for comparison.
EXPERIMENTAL
General
1
SGM extracted from spruce wood chips was used . The acetyl groups originally present had
been cleaved off due to the strong alkaline extraction condition during isolation and
purification. The sugar composition of the pure material was determined to be arabinose
0.3%, galactose 2.3%, glucose 20.2%, xylose 1.4% and mannose 75.8% (all w/w). (M:G = 3.75).
Commercially available mannan (Ivory nut, Megazyme Co. Wicklow, Ireland.) and
methylcellulose (Sigma Aldrich) were used as standards for assigning partially methylated
alditol acetates of mannose and glucose. KGM with a much higher molecular mass than SGM
with M:G = 1.67 was purchased from Hubei Konson Konjac Gum Co., Ltd., China (food
additive grade).
Trace elemental analysis was performed using inductively coupled plasma atomic emission
spectroscopy (Thermo Scientific iCAP 6000 series ICP spectrometer, ICP-AES). The analyses
was performed at wavelengths of 213.5 and 219.9 nm for copper and at 182.5 and 182.6 nm
for bromium, using an ICP multi-element standard IV from Merck. The ICP elemental analysis
of SGM showed 2.549 ppm B corresponding to 0.0003 mol/mol(AGU) and 13.2 ppm Cu
corresponding to 0.0070 mol/mol(AGU).
All reagents were of high purity and purchased from Fluka, Aldrich or Merck. Analytical
reactions were carried out in 1 mL or 5 mL V-vials in a heating block and an evaporating unit
from Barkey, GmbH & Co. KG, Germany. For quantitative GLC evaluation using a flame
ionization detector (FID), the effective carbon response (ECR) concept was applied 21.
Instrumental
Gas chromatographic analysis was carried out using a GLC-FID instrument Shimadzu GC 2010,
equipped with a Phenomenex Zebron ZB-5HT Iferno column (30 m) and a retention gap
(methyl deactivated., 1.5 m), carrier gas H2 (65.4 kPa), split injection port 250 °C, split ratio 0,
carrier gas flow 40 cm/s (linear velocity mode). Temperature program: 60 °C (1 min), with
20 °C/min to 150 °C, with 1 °C/min to 180 °C and with 20 °C/min to 310 °C (10 min). Data
4

were recorded with Shimadzu GC solution Chromatography Data SYSTEM Version 2.3. GLC-
MS was performed on a GC 5890A (Hewlett Packard) using electron impact ionization (EI) (70
eV) in split injection mode (1:20). The same column and temperature program as in GLC-FID
was used. Data evaluation was performed using Amdis Analysis (Version 2.68).
Methylation of GM
The polysaccharide (Mannan, SGM or KGM) was dissolved in DMSO or water. In the aqueous
system, the base and MeI were both added to the polysaccharide solution. When Li-dimsyl
was used, the base was added first to the polysaccharide solution, and after 30 min (called
the “incubation time” and not included in the "reaction time"), methyl iodide was added.
The reaction conditions for the individual samples are shown in Table 1. Methylation of
mannan and KGM were performed in H2O using NaOH/MeI, while SGM was methylated in
DMSO using Li-dimsyl/MeI. The reactions were carried out under nitrogen at room
-
temperature (r.t.). Next, 1.6 M Li-dimsyl (CH3SOCH2 Li+) was freshly prepared from methyl
lithium (5% in diethyl ether) and dry DMSO under nitrogen.
Table 1 Methylation conditions of mannan and glucomannans. Concentration of polysaccharide: 10 mg/mL
(KGM2: 25 mg/mL; mannan: 50 mg/mL); incubation time with Li-dimsyl: 30 min.; reactions were
performed at room temperature.
Base
[equiv./
glycosyl unit]
2.0
CH3I
[equiv./
glycoyl unit]
2.20
Sample
Name
Reagent
Reaction time
[h]
(Base/ Solvent)
MeMan
NaOH/H2O
48
24
2
MeKGM
NaOH/H2O
1.5
2.25
MeSGM1
MeSGM2
MeSGM3
MeSGM4
MeSGM5
MeSGM6
MeSGM7
MeSGM8
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
Li-dimsyl/DMSO
0.1
0.15
0.3
0.45
2
0.5
0.75
2
0.8
1.20
2
1.2
1.80
2
3.0
4.50
1
3.0
4.50
3
12.0
18.00
3
The partially methylated samples were diluted and isolated using dialysis (molecular weight
cut off, MWCO, 3500 Da), followed by freeze drying.
Preparation of Aldiol Acetates for Monomer analysis (AAM method)15
The partially methylated polysaccharide (ca. 2 mg) was hydrolyzed using 2 M trifluoroacetic
acid (TFA) for 120 min at 120 °C in 1-mL V-vials. After cooling to r.t., the aqueous acid was
removed under a stream of nitrogen until the residue was nearly dry. To neutralize any
remaining traces of acid, a droplet of NaOH solution was added.
5

Subsequently, the monosaccharides obtained via hydrolysis were reduced to alditols. A
solution of 0.5 mL of 0.25 M NaBD4 in 2 M NH3 (aq.) was added to the residue and heated at
60 °C for 120 min. After cooling to r.t., the residual NaBD4 was quenched using acetic acid,
and the borate was removed by repeated evaporation (5 times) with methanolic acetic acid
(15%) under a stream of nitrogen.
Alditols were acetylated with 200 μL acetic anhydride and 50 μL pyridine at 90 °C for 120
min. After cooling to r.t., the solution was transferred into a 5-mL V-vial. The 1 mL V-vial was
washed 3 times with dichloromethane (ca. 0.8 mL). The washing solutions were collected in
5-mL V-vials. The dichloromethane phase was washed once with saturated aq. NaHCO3 (ca.
1.6 mL). After phase separation, the upper phase was carefully removed using a Pasteur
pipette. The combined organic layers were washed once with 0.1 M HCl (1 mL), and
subsequently washed three times with 1 mL of water. The organic phase was dried using
CaCl2 and used for GLC-FID analysis.
RESULTS AND DISCUSSION
The methylation of glucomannans leads to eight possible substitution patterns for each
glucose and mannose moiety, which are classified as non-substituted (c0), mono-substituted
(c1), di-substituted (c2) and tri-substituted (c3) units. In addition, up to tetra-O-methylated
compounds can be obtained from the terminal residues, including the reaction of O-4. The
position of the methylation can be deduced from the mass spectra (GLC-MS) of the alditol
acetates prepared according to Fig. 2, but to distinguish stereoisomers, i.e., mannitols and
glucitols, standards are required. These were prepared from mannan after partial
methylation and from commercial methylcellulose (DS 2.02). GLC analysis of a partially
methylated alditol acetate (PMAA) mixture of these two standards provides the retention
times of all PMAAs of interest, which allowed unambiguous assignments of the peaks of
MeSGM and evaluation of the methyl patterns (see Fig. 3).
6

Fig. 2 Scheme of the methylation and preparation of the alditol acetates for the analysis of the methyl
pattern of the GMs
G236
G26
M
M6
M2
M26
M23 G23
G2
M236
G6
M3
G36 M36
G
G3
Standard
MeSGM 6
G2346
M2346
M234
G234
15.0 16.0 17.0 18.0 19.0 20.0 21.0 22.0 23.0 24.0 25.0 26.0 27.0 28.0 29.0 30.0 31.0 32.0 33.0 34.0 min
Fig. 3 Gas chromatograms of the standards (co-injection of alditol acetates from partially methylated
mannan and methyl cellulose), alditol acetates of MeSGM 6 (DS 2.09). The peaks from glucose (G)
and mannose (M) are assigned according to the methylated position.
The yield for methylated SGM was in the range of 87% to quantitative yield, while for
methylated KGM the yield was 70-75%. The peak areas of all of the derivatives were
corrected according to the effective carbon response concept 21. The methylation analysis of
SGM used in this study indicates that both mannose and glucose were involved in terminal
and branched positions, corresponding to their mol fractions present.
The results for MeSGM 1 to MeSGM 8 are shown in Table 2. Due to the approximately 7.4%
branching units in SGM, detailed values for the molar fractions for up to 20 constituents si,
7

including terminal 2,3,4,6- and 2,3,4-O-methylated units, are considered in the calculations.
For the evaluation of the relative reactivities, 4-O-methylation was neglected. Therefore, the
2,3,4,6-tetra-O-methylated units were summarized with 2,3,6-tri-O-methylated units, and
2,3,4-tri-O-methylated were added to the mol fraction of 2,3-di-O-methylated residues. The
molar fractions ci of i-fold substituted glycosyl units, partial DS values xi of the individual
hydroxyl groups in position i, and total DS are listed in Table 2 as well. In SGM, the M/G ratio
of the starting material was 3.8. For a better comparison, the molar compositions of
mannose and glucose were individually normalized to 100%.
8

Table 2 Relative molar composition (Mol %) of glucose and mannose of SGM substituted at position i (si), of molar fraction in % of i-fold substituted monomer units (ci),
partial DS-values xi for position i, and total DS. Data are independently normalized for both glucose and mannose. The ratio of mannose to glucose in the SGM
used was 3.8. Each value except for MeSGM 6 is the average of two independent analyses.
Composition in Mol%
Me in position
MeSGM 1
MeSGM 2
MeSGM 3
MeSGM 4
MeSGM 5
MeSGM 6
MeSGM 7
MeSGM 8
MeKGM
Glc
Man Glc Man
Glc Man
Glc Man
Glc Man
Glc Man
Glc Man
Glc Man
Glc Man
s0
s2
98.55 96.62 85.95 83.32 80.00 77.84 49.31 40.42 22.78 17.98 4.20 4.15 0.87 1.02 0.64 0.90 75.10 72.40
0.48 0.45 3.91 5.77 5.10 6.99 9.80 17.49 12.43 18.31 8.48 11.29 3.23 5.14 0.06 0.16 7.93 11.96
0.44 2.04 6.39 7.78 7.86 9.49 17.40 16.97 17.55 13.71 6.51 4.11 0.96 1.60 0.06 0.05 4.76 2.36
0.22 0.13 2.57 1.85 3.32 2.33 4.50 2.09 4.97 1.99 2.47 0.86 0.34 0.24 0.08 0.13 9.29 10.23
0.03 0.06 0.35 0.38 1.06 1.31 9.17 12.50 20.06 24.61 28.08 31.40 20.64 21.37 7.41 6.96 0.70 0.58
0.08 0.12 0.30 0.28 0.75 0.71 2.01 2.51 4.39 5.19 7.75 7.36 7.51 8.23 0.48 0.40 1.17 2.05
0.08 0.39 0.39 0.40 1.13 0.72 3.20 1.91 6.84 3.51 8.49 2.93 3.95 2.24 0.33 0.23 0.69 0.22
0.11 0.20 0.14 0.22 0.79 0.60 4.60 6.12 10.97 14.71 34.01 37.90 62.50 60.15 90.94 91.17 0.34 0.21
s3
s6
s23
s26
s36
s236
75.10 72.40
21.99 24.54
2.56 2.85
0.34 0.21
0.10 0.15
0.07 0.03
0.11 0.13
1.67
c0
98.55 96.62 85.95 83.32 80.00 77.84 49.31 40.42 22.78 17.98 4.20 4.15 0.87 1.02 0.64 0.90
1.15 2.62 12.87 15.40 16.27 18.81 31.71 36.55 34.96 34.01 17.46 16.26 4.53 6.99 0.20 0.34
0.19 0.57 1.04 1.06 2.94 2.74 14.38 16.92 31.29 33.31 44.32 41.69 32.10 31.84 8.22 7.59
0.11 0.20 0.14 0.22 0.79 0.60 4.60 6.12 10.97 14.71 34.01 37.90 62.50 60.15 90.94 91.17
0.01 0.01 0.05 0.07 0.08 0.10 0.26 0.39 0.48 0.63 0.78 0.88 0.94 0.95 0.99 0.99
0.01 0.03 0.07 0.09 0.11 0.12 0.34 0.37 0.55 0.57 0.77 0.76 0.88 0.85 0.99 0.98
0.00 0.01 0.03 0.03 0.06 0.04 0.14 0.13 0.27 0.25 0.53 0.49 0.74 0.71 0.92 0.92
c1
c2
c3
x2
x3
x6
M/G ratio
4.03
4.12
4.02
4.09
3.78
3.75
3.99
3.81
DS(M) and DS(G) 0.02 0.04 0.15 0.18 0.25 0.26 0.74 0.89 1.30 1.45 2.08 2.13 2.56 2.51 2.89 2.89 0.28 0.31
DS (SGM) Ø 0.02 0.16 0.25 0.77 1.33 2.09 2.55 2.89 0.29
9

The DS values for O-2, O-3, and O-6 were between 0.02 and 2.89. MeSGM 8 is practically
permethylated (total DS including O-4 is 3.01), and the corresponding alditol acetates reflect
the branching pattern. From the data in Table 2, it is obvious that there is no difference in
the distribution of glucose and mannose with respect to the terminal units and branching in
SGM. Comparing the glucose and the mannose fractions, the individual DS of glucose is
slightly higher, up to a DS of approximately 1.5, but at about DS2, mannose is equal to
glucose. From the substitution patterns of the individual MeSGM (Table 2), it is evident that,
in contrast to the typical methyl celluloses, the reactivity of the primary OH at C-6 is
remarkably low. Substitution at the secondary OH groups is strongly favored with the
highest reactivity at 3-OH. In mannose, 2-OH is equal to 3-OH at a DS of approximately 2
(MeSGM 6), while for glucose, it equals 3-OH at DS 0.8 (DSglc 0.89, MeSGM 4). In commercial
methylcellulose prepared in aqueous systems with NaOH and methyl chloride, typically 20-
25% of methyl groups are located at O-3, 40% at O-2 and 35-40% at O-6. No direct influence
of primary substitution on the reactivity of the remaining OHs in the glucosyl unit is
observed for these derivatives 15.
The statistical model for a random substituent pattern, introduced by Spurlin for kinetically
controlled reactions of cellulose, 22 was applied to compare the analytical data with the
random pattern calculated for the partial DS values (xi).
Fig. 4 shows the un-, mono-, di- and trisubstituted mol fractions (ci) of the consecutive
methylation reaction of glucose and mannose in SGM. Deviations from the Spurlin model are
significant. As is typical for non-random distributions, the mol fractions of un- (c0) and
trisubstituted residues (c3) are still or already higher than expected, while c1 (mannose) or c1
and c2 (glucose) are lower. Obviously, methylation was favored in already methylated units,
which indicates fast fluctuation of the protons in the partially deprotonated chains, because
the electrostatic repulsion of anions suggests that the opposite effect would be expected. It
will be of interest to look into the methylation patterns at the polymer level of SGM, which
should be the result of the individual reactivities and the distribution of glucose and
mannose in the GM macromolecules.
10

100
90
80
70
60
50
40
30
20
10
0
100
90
80
70
60
50
40
30
20
10
0
(b)
(a)
0
0,5
1
1,5
2
2,5
3
0
0,5
1
1,5
2
2,5
3
DS Glc
DS Man
exp Glc c0
spu Glc c0
exp Glc c1
spu Glc c1
exp Glc c2
spu Glc c2
exp Glc c3
spu Glc c3
exp Man c0
spu Man c0
exp Man c1
spu Man c1
exp Man c2
spu Man c2
exp Man c3
spu Man c3
Fig. 4
Comparison of the experimentally determined mol fractions (ci) for (a) Man and (b) Glc of MeSGM1-
8 with model data for a random pattern
Methylation of carbohydrates can be performed in different ways. Typically, OH groups are
deprotonated to form good nucleophiles, which then attack methyl halide (or
dimethylsulfate) to form methyl ethers. This Williamson etherification is irreversible and,
thus, kinetically controlled. However, when sub-stoichiometric amounts of base are applied,
the degree of deprotonation will be different for the various OH groups at positions 2, 3 and
6 in the 1,4-glycan and between glucose and mannose. Because the acid-base reaction is
reversible, this step is thermodynamically controlled, and the equilibria between ROH/RO-
are determined by their pKa values. Thus, the regioselectivity of etherification will, to a
certain extent, also reflect the relative acidities of the OH groups. In addition to the acidity,
the stereoelectronic and steric effects have to be considered. Cyclodextrins, for example,
show a much higher regioselectivity in favor of O-2 compared to the corresponding open-
18, 23
chain malto-oligosaccharides and amylose due to the poor accessibility of O-3
.
Regarding the mechanism, the steric requirements of the transition state are higher for SN2
than SN1 reactions, as has been studied in reactions of selectively protected cycloinositols 24,
25. Consequently, the solvent is an influential factor, especially in the case of substrates with
reactive groups in close proximity. Because of the multifunctionality of carbohydrates,
derivatization comprises several consecutive steps. For the data on MeKGM in aqueous
reactions (Table 2), the overall reactivity is also slightly higher for glucose (MeKGM: DS(man)
0.28, DS (glc) 0.31). Different regioselectivities are found for glucose and mannose. For
mannose, a slight preference for OH at position 6 is observed, while glucose shows more
pronounced differences of reactivities following the order 2 ≥ 6 > 3, again an indication of
the higher acidity of the equatorial 2-OH compared to the axial one. Thus, methylation of the
primary 6-OH is much more competitive under these conditions and can be further
enhanced by increasing the amount of base. According to the above mentioned model of
Spurlin 22 for reactions of cellulose in aqueous systems, the relative reactivities of the
11

different OHs should remain constant over the course of the reaction as long as the glycosyl
units in the polysaccharide are equally accessible. Changes in the primary substitution do not
influence the residual OH groups, indicating that no neighboring groups influence the local
reactivities.
2,50
2,00
1,50
1,00
0,50
0,00
2,50
2,00
1,50
1,00
0,50
0,00
(b)
Glc-k2
(a)
Man-k2
Man-k3
Man-k6
Glc-k3
Glc-k6
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
-lnc0
-lnc0
Fig. 5 Relative rate constants k2, k3 and k6 calculated independently for both (a) Man and (b) Glc and obtained
from the data on MeSGM1-6
If these requirements are fulfilled, a logarithmic plot, as shown in Fig. 5, should give a linear
relationship with a slope indicating the relative reactivity constants, ki. Instead of time, the
decrease in unsubstituted glycosyl units, -ln c0, is taken to be the reaction coordinate. We
applied this evaluation to the MeSGM series. Figs. 5a and b show the results for mannose
and glucose. To allow comparison of glucose and mannose, the same reaction coordinate
must be used. Therefore, the -ln co of glucose and mannose was not used; instead, the co of
MeSGM was used. As mentioned above, O-3 was the most reactive in the early stages of the
reaction for both glucose and mannose. However, for glucose, the relative reactivities
followed the order 2 > 3 > 6, while for mannose, the reactivity of O-3 was the highest over a
wide portion of the reaction, closely followed by O-2. The reactivity of the primary 6-OH was
far behind for both sugars, which is a distinct difference to reactions in water, where
methylation of the primary OH, which has better accessible and is more acidic than the
secondary 3-OH, is usually much more pronounced, as mentioned above. However, constant
slopes were not observed for the individual OHs, but the data points could be fitted very
well with an exponential.
12

0,90
0,80
0,70
0,60
0,50
0,40
0,30
0,20
0,10
0,00
1,60
1,40
1,20
1,00
0,80
0,60
0,40
0,20
0,00
(a)
(b)
0,00
0,50
1,00
1,50
2,00
0,00
0,50
1,00
1,50
2,00
-ln(1-x2)
-ln(1-x2)
Fig. 6 Relationship of relative reactivities in mannose of MeSGM 1-6: (a) for O-2 and O-3 (b) O-2 and O-6.
1,60
1,40
1,20
1,00
0,80
0,60
0,40
0,20
0,00
0,80
0,70
0,60
0,50
0,40
0,30
0,20
0,10
0,00
(a)
(b)
0,00
0,50
1,00
1,50
2,00
2,50
0,00
0,50
1,00
1,50
2,00
2,50
-ln(1-x2)
-ln(1-x2)
Fig. 7 Relationship of the relative reactivities in glucose of MeSGM 1-6: (a) for O-2 and O-3 (b) O-2 and O-6.
The relative reactivity of O-3 decreased with increasing methylation of O-2 (see Fig. 6a for
mannose and Fig. 7a for glucose), while O-6 remained unaffected by the status of O-2
(Fig. 6b for mannose and Fig. 7b for glucose). This result is observed for both mannose and
glucose. MeSGM1-6 were included in these calculations because the higher partial DS xi
approaches the maximum value of 1, and the uncertainties in the -ln(1-xi) values are higher.
What is the reason for this completely different reactivity and change in the relative
reactivities in DMSO compared to water? Reactions with weak and strong bases, in aqueous
and aprotic systems, have to be differentiated. In the carbohydrate methylation reactions,
which take place in an aqueous system, the polysaccharides are dissolved or swollen and
activated in the NaOH solution and should thus be equally accessible 24. The presence of
base and methyl halide is possible because the HO- is not nucleophilic enough that it
consumes all of the alkylating agent. Due to very good solvation of the anions, the reactive
species are well shielded and not influenced by their neighboring groups. Therefore, they
react independently and show the distribution calculated for consecutive reactions with a
constant ratio of reactivities of 2-, 3- and 6-OH, which is also supported by the low
13

temporary concentration of alcoholate groups. Deprotonated hydroxyls can immediately
react with the electrophile.
In contrast, the alkylation with the stronger base, Li-dimsyl (pKa DMSO 35), needs to be
26
performed in two steps . SGM was dissolved in DMSO. The calculated amount of Li-dimsyl
solution was added (see Table 1) and consumed for deprotonation of the polysaccharide
before the alkyl halide was added, which is necessary to avoid alkylation of DMSO.
The alcoholic OH of the carbohydrates is more acidic than common aliphatic mono-alcohols
with the hydroxyl at the C-2 next to the electron-withdrawing anomeric center being the
most acidic (pKa ca. 13). Axial OH are known to be less acidic then equatorial OH 27. They also
have less electron-withdrawing effects than the equatorial groups 28.
Because the anions formed in the deprotonation step are not solvated by the aprotic DMSO,
they "help themselves" by sharing the proton of the vicinal OH. The most acidic OH at C-2 is
preferentially deprotonated and forms an intramolecular H-bridge with 3-OH 27, see Fig. 8.
Fig. 8
The intramolecular H-bridge formed in (a) Man and (b) Glc after deprotonation at O-2 in an aprotic
solvent
When the alkylating agent is added, the deprotonation pattern is probably not simply
trapped by methyl iodide. The regioselectivity of etherification is also influenced by steric
factors. Thus, the more accessible equatorial 3-OH benefit from the interaction with the
neighboring, more acidic 2-OH. As illustrated in Fig. 8, in the aprotic solvent the naked RO-
ion at C-2 will interact with 3-OH, thus sharing the proton and activating O-3 for a
nucleophilic attack. This effect is more pronounced for mannose than for glucose because of
the axial orientation of 2-OH in the earlier. Due to the ecliptic position of H-4 and the
interference with the non-binding electron pairs at the glycosidic oxygens with the axial O-2,
the nucleophilic reaction of the equatorial O-3 is much more favored than is expected from
its acidity. As the methylation of O-2 increases, this "catalytic effect" disappears, and the
relative reactivity of O-3 decreases.
CONCLUSION
14

The relative reactivities of stereoisomeric glucose and mannose in glucomannan (SGM and
KGM) were studied. MeKGM with a DS of 0.3 from the reaction in aqueous solution with
NaOH/MeI showed the order O-6 > O-2 > O-3 with only slight differences, while for glucose,
the reactivities of O-2 and O-6 were both distinctly more reactive than O-3. A series of
MeSGM was prepared in DMSO using soluble Li-dimsyl/MeI to obtain the relative reactivities
of the OH groups in stereoisomeric glucose and mannose. The results from this
homogeneous reaction clearly showed the preference for the equatorial 2-OH in glucose,
followed by 3-OH and 6-OH for glucose, while for mannose, the reactivity of 3-OH was the
highest up to DS 2, and at this DS, it was matched by O-2. The reactivity of the primary 6-OH
was the lowest for both epimeric sugars. In the aprotic solvent, the less acidic 3-OH most
likely benefited from the interaction with the anion formed from the more acidic 2-OH. The
equatorial position of 2-OH in glucose favors its reactivity over the axial 2-OH in mannose by
a factor of 1.4 because it is more acidic and less sterically hindered. Comparing the relative
reactivities of 2-OH and 3-OH, the latter can compete with the 2-OH in mannose. In
conclusion, up to a DS of approximately 2, the order of the six OH groups in GM is G-k2 > M-
k3 ≥ M-k2 > G-k3 > M-k6 >G-k6.
Acknowledgements
Björns Stiftelse is acknowledged for supporting the research work of Yujia Zhang in TU-
Braunschweig. Silke Lehmann is thanked for technical support.
15

REFERENCES
1.
2.
3.
Zhang, Y.; Li, J.; Lindström, M.E.; Stepan, A.; Gatenholm, P. Nord. Pulp Pap. Res. J., 2013, 28,
323-330.
Albrecht, S.; Van, M.G.C.J.; Xu, J.; Schols, H.A.; Voragen, A.G.J.; Gruppen, H.J. Agric. Food
Chem., 2011, 59, 12658-12666.
Alonso-Sande, M.; Teijeiro-Osorio, D.; Remuñán-López, C.; Alonso, M.J. Eur. J. Pharm.
Biopharm., 2009, 72, 453-462.
4.
5.
Zhang, Y.; Xie, B.; Gan, X. Carbohydr. Polym., 2005, 60, 27-31.
Xu, C.; Eckerman, C.; Smeds, A.; Reunanen, M.; Eklund, P.C.; Sjoholm, R.; Willfor, S. Nord.
Pulp Pap. Res. J., 2011, 26, 167-178.
6.
7.
Xu, C.; Leppänen, A.-S.; Eklund, P.; Holmlund, P.; Sjöholm, R.; Sundberg, K.; Willför, S.
Carbohydr. Res., 2010, 345, 810-816.
An, N.T.; Thien, D.T.; Dong, N.T.; Le Dung, P.; Hanh, P.T.B.; Nhi, T.T.Y.; Vu, D.A. Carbohydr.
Polym., 2011, 84, 173-179.
8.
Voepel, J.; Edlund, U.; Albertsson, A-C. J. Polym. Sci. A Polym. Chem., 2009, 47, 3595-3606.
Croon, I.; Lindberg, B. Acta Chem. Scand., 1958, 12, 453-8.
9.
10.
11.
Croon, I.; Lindberg, B.; Meier, H.L. Acta Chem. Scand, 1959, 13, 1299-1304.
Capek, P.; Kubackova, M.; Alfoldi, J.; Bilisics, L.; Liskova, D.; Kakoniova, D. Carbohydr. Res.,
2000, 329, 635-645.
12.
Katsuraya, K.; Okuyama, K.; Hatanaka, K.; Oshima, R.; Sato, T.; Matsuzaki, K. Carbohydr.
Polym., 2003, 53, 183-189.
13.
14.
15.
16.
17.
Petursson, S.; Scully, S.M.; Jonsdottir, S. Carbohydr. Res., 2014, 388, 37-43.
Mischnick, P.; Adden, R. Macromol. Symp., 2008, 262, 1-7.
Voiges, K.; Adden, R.; Rinken, M.; Mischnick, P. Cellulose, 2012, 19, 993-1004.
Mischnick, P.; Momcilovic, D. Adv. Carbohydr. Chem. Biochem., 2010, 64, 117-210.
Vollmer, A.; Voiges, K.; Bork, C.; Fiege, K.; Cuber, K.; Mischnick, P. Anal. Bioanal. Chem., 2009,
395, 1749-1768.
18.
19.
20.
Lee, C.K.; Kim, E.J.; Jun, J.H.Bull. Korean Chem. Soc., 1999, 20, 1153-1158.
Blakeney, A.B.; Stone, B. A.Carbohydr. Res., 1985, 140, 319-324.
Mischnick, P.; Kühn, G. Carbohydr. Res., 1996, 290, 199-207.
16

21.
22.
23.
24.
Sweet, D.P.; Shapiro, R.H.; Albersheim, P. Carbohydr. Res., 1975, 40, 217-225.
Spurlin, H.M. Ind. Eng. Chem., 1938, 30, 538-542.
Mischnick-Lübbecke, P.; Krebber, R. Carbohydr. Res., 1989, 187, 197-202.
Haines, A.H. Adv. Carbohydr. Chem. Biochem., R. S. Tipson; H. Derek, Editors. 1976, Academic
Press. 11-109.
25.
26.
Angyal, S.J.; Tate, M.E. J. Chem. Soc., 1965, 6949-6955.
Matthews, W.S.; Bares, J.E.; Bartmess, J.E.; Bordwell, F.; Cornforth, F.J.; Drucker, G.E.;
Margolin, Z.; McCallum, R.J.; McCollum, G.J.; Vanier, N.R. J. Am. Chem. Soc., 1975, 97, 7006-
7014.
27.
28.
Milliard, E.C.; Schroeder, L.R.; Thompson, N.S. Carbohydr. Res., 1977, 56, 259-276.
Jensen, H.H.; Lyngbye, L.; Bols, M. Angew. Chem. Int. Ed., 2001, 40, 3447-3449.
17

RELATIVE REACTIVITIES IN THE O-METHYLATION OF GLUCOMANNANS: THE INFLUENCE OF
STEREOCHEMISTRY AT C-2 AND THE SOLVENT EFFECT
Figure legends
1. Fig. 1 Proposed structures for spruce glucomannan and its partially methylated derivatives.
Approximately 20% of the units have the glucosyl-configuration, and 80% have the mannosyl
configuration.
2. Fig. 2 Scheme of the methylation and preparation of the alditol acetates for the analysis of the methyl
pattern of the GMs
3. Fig. 3 Gas chromatograms of the standards (co-injection of alditol acetates from partially methylated
mannan and methyl cellulose), alditol acetates of MeSGM 6 (DS 2.09). The peaks from glucose (G) and
mannose (M) are assigned according to the methylated position.
4. Fig. 4 Comparison of the experimentally determined mol fractions (ci) for (a) Man and (b) Glc of
MeSGM1-8 with model data for a random pattern
5. Fig. 5 Relative rate constants k2, k3 and k6 calculated independently for both (a)Man and (b)Glc and
obtained from the data on MeSGM1-6
6. Fig. 6 Relationship of relative reactivities in mannose of MeSGM 1-6: (a) for O-2 and O-3 (b) O-2 and O-
6.
7. Fig. 7 Relationship of the relative reactivities in glucose of MeSGM 1-6: (a) for O-2 and O-3 (b) O-2 and
O-6.
8. Fig. 8 The intramolecular H-bridge formed in (a) Man and (b) Glc after deprotonation at O-2 in an
aprotic solvent
18

Graphical abstract
28

Highlights
-
A set of O-methyl glucomannans (GM) in DS range 0.02-2.89 were prepared in DMSO with Li-
dimsyl/methyl iodide
-
-
-
Relative reactivities of stereoisomeric glucose and mannose in glucomannan were studied
Up to DS 2, the order of the six OH groups in GM is G-k2 > M-k3 ≥ M-k2 > G-k3 > M-k6 >G-k6.
Enhanced reactivity of O-3 in both glucose an mannose in the early stage of the reaction is
explained by a solvent effect.
29