A reevaluation of the epimeric and anomeric relationship of glucosides and galactosides in thermotropic liquid crystal self-assemblies

Article abstract of DOI:10.1016/j.carres.2011.10.032

Anomers and epimers α- and β-gluco and -galactosides are expected to behave differently. However, recent results on a series of Guerbet glycosides have indicated similar liquid crystal clearing temperatures for pure β-glucosides and the corresponding α-galactosides. This observation has led to speculation on similarities in the self-assembly interactions between the two systems, attributed to the trans-configuration of the 4-OH group and the hydrophobic aglycon. Previous simulations on related bilayers systems support this hypothesis, by relating this clearing transition temperature to intralayer (sugar-sugar) hydrogen bonding. In order to confirm the hypothesis, the comparison was expanded to include the cis-configurated pair, that is, α-gluco/β-galactoside. A set of α-configurated Guerbet glucosides as well as octyl α-galactoside were prepared and their thermotropic phase behavior studied. The data obtained enabled a complete comparison of the isomers of interest. While the results in general are in line with a pairing of the stereo-isomers according to the indicated cis/trans-configuration, differences within the pairs can be explained based on the direction of hydrogen bonds from a simple modeling study.

Full text of DOI:10.1016/j.carres.2011.10.032

Carbohydrate Research 346 (2011) 2948–2956
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
A reevaluation of the epimeric and anomeric relationship of glucosides
and galactosides in thermotropic liquid crystal self-assemblies
Rauzah Hashim a, , Seyed M. Mirzadeh a, Thorsten Heidelberg a, Hiroyuki Minamikawa b,
⇑
Tanaka Yoshiaki c, Akhiko Sugimura c
a Department of Chemistry, University of Malaya, 50603 Kuala Lumpur, Malaysia
b Nanotube Research Center (NTRC), National Institute of Advanced Industrial Science and Technology, AIST, Tsukuba Central 5, 1-1-1 Higashi, Tsukuba, Ibaraki 305 8565, Japan
c Department of Information Systems Engineering, Faculty of Engineering, Osaka Sangyo University, 3-1-1 Nakagaito, Daito-shi, Osaka 574 8530, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2011
Received in revised form 18 October 2011
Accepted 20 October 2011
Available online 28 October 2011
Anomers and epimers
a- and b-gluco and -galactosides are expected to behave differently. However,
recent results on a series of Guerbet glycosides have indicated similar liquid crystal clearing tempera-
tures for pure b-glucosides and the corresponding -galactosides. This observation has led to speculation
on similarities in the self-assembly interactions between the two systems, attributed to the trans-config-
uration of the 4-OH group and the hydrophobic aglycon. Previous simulations on related bilayers systems
support this hypothesis, by relating this clearing transition temperature to intralayer (sugar–sugar)
hydrogen bonding. In order to confirm the hypothesis, the comparison was expanded to include the
a
Keywords:
Glycolipids
Branched chain lipid
Self-assembly
Amphitropic liquid crystals
Epimers
cis-configurated pair, that is,
a-gluco/b-galactoside. A set of a-configurated Guerbet glucosides as well
as octyl -galactoside were prepared and their thermotropic phase behavior studied. The data obtained
a
enabled a complete comparison of the isomers of interest. While the results in general are in line with a
pairing of the stereo-isomers according to the indicated cis/trans-configuration, differences within the
pairs can be explained based on the direction of hydrogen bonds from a simple modeling study.
Ó 2011 Elsevier Ltd. All rights reserved.
Anomers
1. Introduction
caused by the separation of the hydrophilic from the hydrophobic
region, creates a polar order, from which a non-linear response to
Glycolipids are amphiphiles with a hydrophilic sugar head
group and a hydrophobic alkyl chain. Some are amphitropic,1 that
is, capable of forming liquid crystals in dry state and when sol-
vated. They are usually associated with membranes and their func-
tions are related to lyotropic and surfactant properties.2 The cell
membrane is a fluid lamellar system, where the hydrophilic region
is based on an extensive hydrogen bonding network. It is distinctly
separated from the hydrophobic domain, which is dominated lar-
gely by repulsive, geometric driven interaction of the alkyl chains,
often assumed to be randomly arranged and giving it a fluid nat-
ure.3 The simplistic concept of the glycolipid self-assembly re-
quires a systematic examination of how the complex sugar
structure relates to the properties and to uncover some of the driv-
ing forces responsible for its formation.
So far, investigation and interest in the possible thermotropic
application of glycolipids is underexploited. For example, although
glycolipids are chiral and ferroelectric,4 possible tilted structures
(like SC) in the lipid organization have been reported only re-
cently,5,6 even though ferroelectricity phenomena in cell mem-
brane functions are known.7 The possible columnar structure,
polarization is expected.8 However, such investigations are recent
and preliminary; some of these synthetic glycolipids have shown
spontaneous macroscopic polarization along the columnar axis
and may display a reversible signal of second harmonic generation
(SHG) by heating.9
Understanding the amphiphilic surfactant self-assembly phe-
nomena, both in dry and hydrated forms, and the implications
for bio- and material technology require basic experiments on pure
materials, which are difficult to extract. A synthetic substitute is
therefore, always desirable. From the early works of Jeffery10,11
and Vill,12 many synthetic glycolipids—mostly monoalkylated,
have been prepared, but significantly few self-assembly related
properties (especially thermotropic liquid crystals) have been re-
ported.13 Previous studies on these, have demonstrated a direct
correlation between molecular structure and liquid crystals behav-
ior, where common phases observed included smectic A (SA),
columnar (Col) and bicontinuous cubic phases (Cub). See Figure 1
for examples.14–16 Moreover, molecular amphiphilic nature is the
driving force for the thermotropic phase formation, due to the
microscopic phase separation of a dichotomy, such as hydro-
philic/lipophilic.3 Minimum molecular size is important; the
dichotomic balance can be tuned to obtain an optimum lipid chain
⇑
Corresponding author. Tel.: +60 3 7967 4009; fax: +60 3 7967 4193.
E-mail address: rauzah@um.edu.my (R. Hashim).
length for
a given polar head group. Sugar stereochemistry
0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2011.10.032

R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
2949
Figure 1. Generic structures of different glycolipids (simple and complex sugars, single and double lipid chains) are shown in (a). The possible thermotropic phases for these
as reported previously are shown in (b).14–16,31 Both (thermotropic and lyotropic) naming conventions are used here; for example, smectic A (SA) for thermotropic (dry state)
phase or also known as lamellar L the fluid bilayer phase for lyotropic phase. Since glycolipids are amphitropic, the lyotropic naming convention is perhaps more appropriate
a
even through in the dry state, since there may be other complex dry phases to be observed such as the bicontinuous cubic or gel phases.
(epimers and anomers) influences the thermotropic behaviors; hy-
droxyl groups (OHs) increase the melting transition from hydrated
solid to the liquid crystal phase and widen the range of the ordered
phase before clearing into the isotropic phase.13 Different behav-
iors are observed for the sugar head-group of various oligosachar-
rides, even though the difference between them is very subtle.
except in the case of lactosides, whose anomalous behavior is due
to the presence of an end-galactose sugar unit.3 A few of these gly-
cosides even show thermotropic polymorphism.18 Therefore, the
Guerbet chain, which has high hydrophobicity, generally lowers
the melting point from the crystalline state of these glycosides into
their ordered phases—a behavior similar to their precursor alco-
hols, which are all liquids at room-temperature.19 Guerbet glyco-
sides show the same thermotropic clearing transition tendency
in homologous series as monoalkylated analogs, that is, the transi-
tion temperature into the isotropic phase increases initially with
increasing chain length until a maximum is reached. Beyond this
maximum, increasing the chain length will decrease the transition
temperature and promote non-lamellar or curved phases, such as
cubic and columnar phases.18
Malto-oligosaccharides with a (1?4-a) link for example, form a
tertiary structure of flexible helix, while the corresponding cello-
oligosaccharides with a (1?4-b) link exhibit an extended ribbon
structure. Due to the non-linear alignment of the glucose units in
maltose (1?4-a-linkage) the corresponding lipids have lower
melting transitions and are more soluble than the cellobiose b-iso-
mers, which are expected to pack more compactly, hence exhibit-
ing a higher melting temperature.17 Flexibility of the alkyl chain is
important and there is a simple dependency of the transition tem-
peratures on the chain length.3 For monoalkylated glycolipids, a
minimum of about seven carbons chain is required to exhibit a li-
quid crystal phase. The transition temperature into the isotropic
phase usually increases with increasing chain length until a maxi-
mum is reached, beyond which increasing chain length depresses
the phase transition due to the ease of chain randomization.
Although monoalkylated glycolipids are interesting, membrane
lipids (both phospholipids and glycolipids) are usually double
tailed, often involving two non-equivalent chains. Hence chain
branching plays an important role in membrane stability, and
many synthetic branched chain glycolipids were developed by var-
ious groups, for example 1,2-dialkyl/diacyl-glycerol-based16, 1,3-
glyco-glycero-lipids17 and Guerbet glycosides.18
An interesting phenomenon observed from these studies sug-
gests a possible structural connection between
a-galactosides
and b-glucosides. It was found that the behavior of a series of pure
a
-galactosides18 is similar to recently reported behavior of pure b-
glucosides.20 This has led to the speculation that the epimeric/ano-
meric pairs may have similar structural arrangements within their
hydrophilic region. It has been attributed to the relative configura-
tion of the anomeric linkage and the hydroxyl group at C-4, which
are trans to each other (see Fig. 2).
In the present report, we try to confirm this phenomenon by
comparing three new
-glucopyranoside (
side Glc-C10C6) and 2-decyl-tetradecyl
Glc-C14C10), with previously reported b-galactosides.18 This epi-
a
a
-Guerbet glucosides, namely 2-ethyl-hexyl
Glc-C6C2), 2-hexyl-decyl -glucopyrano-
-glucopyranoside
a-
D
a-D
a-D
(a
(a
Unlike monoalkylated glycosides, Guerbet glycosides have a
simple asymmetric branched chain. They also exhibit a variety of
ordered phases (SA, Col, Cubic) from below the room–temperature,
meric/anomeric pair has a cis-configuration of the 4-OH and the
anomeric bond. In addition, we have also determined the thermo-
tropic phase behavior of octyl
a-D-galactopyranoside to complete

2950
R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
Figure 2. Correlation between axial/equatorial bond at 4th carbon position to the
a
/b-glycosidic 1–4 bond may influence intralayer hydrogen bonding.20
the epimer/anomer quartet. Although the latter has been previ-
ously synthesized,21,22 its liquid crystal behavior, to the best of
our knowledge, has not been reported.
than 2 kJ molÀ1 is needed for the corresponding liquid crystal to
isotropic transition.13,24 Thus, the observed data Table 1 reasonably
accounts for the liquid crystalline phase transitions of both SA to Iso
and Col to Iso. We should expect the former to be slightly higher
than the latter, as packing within the smectic layer is more than
within a columnar structure, hence greater hydrogen bonding ap-
plies to the former.
2. Materials and methods
Starting materials and reagents of purity 98% or above, and sol-
vents of AR or GR grade were applied without further purification.
The Guerbet glycosides were synthesized following a previously
reported procedure.4,18 Spectroscopic details of new products are
given in the appendix, although they were omitted for previously
reported compounds. 1H NMR analyses of the materials did not re-
veal any indication of anomeric impurities, thus suggesting a pur-
ity of 97% or above.
Unlike in previously described examples, the compounds with a
total chain length of 16 carbon atoms, that is,
aGlc-C10C6 and bGal-
C10C6 exhibit neither similar phases nor transition temperatures.
As these glycosides are of different purity (anomeric impurities)
and within a region (of chain length), where the hydrophilic and
lipophilic balance is highly sensitive, these two results are not
comparable. Moreover, in a recent study, bGlc-C10C6 gave an in-
verse micellar phase (no birefringence was observed), which was
not a liquid crystal, below the isotropic phase transition.20 Due to
this inconsistent phase behavior, an extensive investigation was
conducted, which revealed the presence of very minor amounts
of hydration water transformed the hexagonal assembly into the
micellar phase (details will be reported elsewhere). This phase sen-
sitivity has not been observed for any other chain length and is be-
lieved to be due to a critical balance of the oppositely directing
effects of the head group and the chain. It was argued that, due
The thermotropic liquid crystalline phase was identified by an
optical polarizing microscope (OPM), an Olympus BH polarising
microscope equipped with a Mettler FP82 Hotstage (heating stage).
Microphotographs were taken with
a digital resolution of
1300 Â 1030 pixel on a Soft Imaging System Color View XS camera.
Differential scanning calorimetry measurements were carried out
using a Mettler Toledo DSC at a heating rate of 5 °C per minute
(The error of measurement as given by the instrument specifica-
tion is 0.3 °C).
to packing frustration, the Guerbet derived b-glucosides and
a-
galactosides having chain lengths above 12 carbons in total, prefer
to curve and form a non-lamellar phase. However, a chain length of
12, which can be extended to 16 upon addition of hydration water,
is insufficiently long to stabilize the columnar phase. This results in
an inverse aggregated micellar phase instead, whose structure may
be confirmed by X-ray diffraction.20
3. Results and discussions
The OPM textures and DSC thermograms for
aGlc-C6C2, aGlc-
C
10C6, Glc-C14C10 Glc-C8 and Gal-C8 are given in Figure 3,
a
,
a
a
while Table 1 summarises and compares the transition tempera-
tures of these glycosides with previously reported results for the
corresponding b-4-epimers.24
aGlc-C6C2 and bGal-C6C2 show al-
Computer simulations of monoalkylated
a/b-octyl glucosides
and galactosides28 in the thermotropic smectic A phase were con-
ducted and these simple bilayer models were investigated to
understand the nature of bonding within the hydrophilic region
and the effect of sugar stereochemistry. Although numerous com-
puter simulations on lipid bilayers have been reported before,29,30
this investigation28 is the only example of the thermotropic system
devoid of any solvent interaction. Also, a fixed lipid chain was used
throughout the investigation, enabling a systematic comparison of
sugar configurations. Hydrogen bonding was grouped into inter-
and intralayer interactions, of which the latter dominated the
clearing temperature. Some of the results are quoted here in Ta-
ble 2.28 It was noted that a cis-configuration of the 4-OH and the
glycosidic linkage correlated with higher intralayer hydrogen
bonding compared to the trans-pair. This observation matches
our present results, as indicated in Table 2. It is even possible to
use the H-bonding strength for a relative prediction of clearing
temperatures.28
most the same transition temperature for the clearing into the iso-
tropic liquid, as shown in Table 1. However, the previously
reported transition enthalpy for bGal-C6C2 of about
24.2 kJ molÀ1 24
,
exceeds the range of typical liquid crystal transi-
tions. We have therefore repeated this measurement with an initial
heating scan rate of 10 °C per min and a slower rate (0.5 °C minÀ1
)
for cooling and reheating. The present result identifies the phase
transition at 105 °C upon heating and reheating as a melting pro-
cess from a crystal to the isotropic phase. The OPM texture in Fig-
ure 2e confirms this assignment. The liquid crystal phase transition
is only observed on cooling at 87 °C, indicating that this compound
displays a monotropic transition behavior. Like the lower homolog,
aGlc-C14C10 and bGal-C14C10 also gave similar transition tempera-
tures and enthalpies, but in this case the transformation was be-
tween the columnar and the isotropic phase (Col-Iso). In the case
24
of the previously reported bGal-C14C10 however, no crystal to
columnar transition was observed for
aGlc-C14C10. This may be
Table 2 indicates a reasonably good correlation of intralayer H-
bonding and clearing temperature, whereas the total H-bonding
cannot explain the relative clearing temperatures. An exception
due to strong hydrogen-bonding interactions for non-reducing gal-
actose units, as previously indicated for lactosides.
The transition enthalpy quantifies the energy change due to the
structural changes within the self-assembly, when a phase transi-
tion occurs. For a crystal to a liquid crystalline phase transforma-
tion, the latent heat change is greater than that of a liquid crystal
to an isotropic phase. For example, the transition of a crystal into
to this apparently good correlation is that for the
aGal-C8. The
experimental clearing transition (98 °C) exceeds the prediction
(68 °C) quite considerably. The discrepancy is probably due to
some uncertainty in the measurement of intralayer H-bond in
the simulation. Intermolecular H-bonding effects have been previ-
ously applied to explain the different behavior of sugar isomers. In
a
columnar phase requires DH of more than 63 kcal/mol
(ꢀ263 kJ/mol) for a galactopyranosyl-glycerol,16 while only less

R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
2951
Figure 3. Thermotropic phase behaviors of the three
Glc-C6C2), (b) -hexyl-decyl-glucoside Glc-C10C6), (c)
galactoside. (a) Glc-C6C2, SA (L ), DSC shows typical liquid crystal to isotopic transition, (b)
Glc-C14C10,: Col (HII), DSC shows typical liquid crystal to isotopic transition, (d)
a
-glucosides characterized by OPM textures and the corresponding DSC thermograms of (a)
-decyl-tetradecyl-glucoside Glc-C14C10), (d) monoalkylated
Glc-C10C6, isotropic micellar phase (L2), DSC shows a weak phase transition, (c)
Gal-C8, SA at 95 °C (heating), DSC shows typical transitions for crystal to liquid crystal (non-
a
-ethyl-hexyl-glucoside
(
a
a
(a
a
(
a
a
a-octyl-galactoside and (e) b-ethyl-hexyl-
a
a
a
a
reversible) and liquid crystal to isotopic (reversible) and (e) bGal-C6C2, crystal phase texture on cooling at 101 °C. The initial DSC scan rate is 10 °C per min and shows a large
crystal melting peak. On cooling at 0.5 °C per min, the system initially forms a liquid crystalline phase, which is, however, closely followed by crystallization, thus only the
melting peak can be observed upon reheating.
a study reported by Kikuchi et al.21,22 the binding of alkyl glyco-
sides to a cyclic resorcinol tetramer was found to be largely
influenced by the intra-complex guest–guest (alkyl glycosides)
H-bonding. This implies that the alkyl glycosides form a network
of hydrogen-bonds within the host system. Hydrogen bonding
affects solvophobicity/solvophilicity due to the geometry, stereose-
lectivity and cooperativity, hence indirectly supporting the
simulation findings.

2952
R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
Fig. 3 (continued)
Table 1
Thermotropic phases, transitions temperatures and enthalpies of
a/b-gluco-/galactoside pairs; present results are bold
a-Glycosides (current work)
Phases (transition T in °C)
D
HM (kJ molÀ1
)
Epimeric b-glycosides
Phases (transition T in °C)
D )
HM (kJ molÀ1
23
a
a
a
a
a
Glc-C6C2
Cr? SA (105 4) Iso
Cr? Micellar (45 4) Iso
Cr? Col (118 4) Iso
Cr ꢀ40 SA (95 4) Iso
Cr 69 SA 116 Iso25
5.7
0.14
1.7
bGal-C6C2
bGal-C10C6
bGal-C14C10
bGlc-C8
(SA 87) Cr 111 Iso
Cr 52 Col 69 Iso24
Cr 80 Col 123 Iso24
Cr 69 SA 107 Iso25
Cr 96 SAd 127 Iso25
(1.3) 24.224
0.824
Glc-C10C6
Glc-C14C10
Gal-C8
1.224
1.4
0.4026
Glc-C8
0.4927
bGal-C8
Reference to the previous literature is given in superscript number.
Table 2
Correlation of simulation results for hydrogen bonding and clearing temperature of thermotropic glycolipid assemblies
Lipid
Exp. clearing transition (°C) H-Bonding from simulation28 Calc. clearing transition (°C) (intralayer H-bond) Calc. clearing transition (°C) (total H-bond)
Intralayer
Total
Reference system used
Reference system used
a
Glc-C8
bGlc-C8
aGlc-C8
bGlc-C8
bGlc-C8 10725
2.7
2.4
2.8
2.9
3.5
3.9
3.8
3.5
106
68
—
—
69
117
135
85
121
—
—
a
a
Gal-C8 95⁄
145
140
106
Glc-C8 11625
bGal-C8 12725
134
85
Prediction results (in bold) agree with experimental data (from Sakya et al.25). The experimental clearing temperature (⁄) for
a
Gal-C8 is determined from the present work.
The reference system has to be defined when using equation 1 from Chong et al.28 Both
aGlc-C8 and bGlc-C8 are used as reference systems to obtain the predicted values.
Apparently this concept also applies to the columnar phase,
although it’s self-assembly structure differs from the lamellar sys-
tem. Therefore, the epimeric/anomeric pair may be grouped even
for the longer chained decyl-tetradecyl
see Table 3. To our knowledge, data for
is scarce in the literature, probably due to the more challenging
a
/b-gluco/galactoside pairs,
a-galactopyranosides

R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
2953
Table 3
Transition into the isotropic phase from computer simulation predicted based on intra-layer hydrogen bonding of octyl-glucosides and galactosides are given
Trans-configuration of 4-OH/aglycon
T(SA-Iso)/°C
Intra-layer H-bond
bGlc-C8
Cis-configuration of 4-OH/aglycon
T(SA-Iso)/°C
Intra-layer H-bond
bGlc-C8
Simulation28
Simulation28
Reference lipid
aGlc-C8
Reference lipid
aGlc-C8
bGlc-C8
106
68
Ref. (107)
69
a
Glc-C8
Ref. (116)
134
117
135
aGal-C8
bGal-C8
Experimental (straight chain)
T(SA-Iso) (°C)
Literature ref.
Experimental (straight chain)
T(SA-Iso) (°C)
Literature Ref.
25
21,22
bGlc-C8
107
98
146
145.9
a
Glc-C8
bGal-C8
Glc-C18
116
127
150
164
25
a
Gal-C8
bGlc-C18
Gal-C18
2
Present work
33
33
a
33
33
a
bGal-C18
Experimental (Guerbet chain)
T(SA/Col-Iso) (°C)
Literature ref.
Experimental (Guerbet chain)
T(SA/Col-Iso) (°C)
Literature Ref.
18
24
bGlc-C6C2
55
67
a
Glc-C6C2
105
87
3
Present work
Present work
24
aGal-C6C2
bGal-C6C2
2
111 2 (Cr-Iso)
18
24
bGlc-C14C10
Gal-C14C10
95
104
a
Glc-C14C10
118
3
Present work
24
a
bGal-C14C10
123
Experimental data on these thermotropic phase transitions are shown here for other glycosides, both monoalkylated as well as those of Guerbet types. These are grouped
according to the anomer/epimeric gluco/galactoside pairs. Present results are highlighted in bold.
preparation of the
only one complete set of anomer–epimers, that is, octadecyl
a
-anomers.31,32 For monoalkylated compounds
/b-
ing temperatures for the b-galactoside, which substantiates the
present experimental results (Table 3), and accounts for the high
value of intra-molecular H-bonding from the previous computer
simulation study (Table 2).28
The differences in hydroxyl group orientations for the remain-
ing three isomers with respect to inter- and intralayer H-bonding
a
D
-gluco-/galactopyranoside, has been prepared and investigated
by Hori et al.33 The clearing phase transition temperatures are gi-
ven in Table 3. Again the data confirms our suggestion regarding
the similar behavior of the epimeric/anomeric trans-pair and high-
er clearing temperatures of the corresponding cis-pair.
are less obvious. The a-anomeric glycosides seem to support inter-
layer H-bonding based on three hydroxyl groups, that is, OH-2, OH-
3 and OH-4 (Fig. 4c and d). However, the orientations for two of
these, that is, OH-2 and OH-4, are also appropriate for intralayer
interactions. On the other hand, the b-glucoside (Fig. 4b) exhibits
two hydroxyl groups that appear to be contributing to interlayer
interactions only, namely the hydroxyl groups at C-4 and C-3.
The quantification of the relative intralayer hydrogen bonding is
additionally complicated by a deviation of glycolipid assemblies
from a parallel alignment of the alkyl chain and the layer nor-
4. Rationalization of the epimeric–anomeric trend by modeling
Although the observed trend is preliminary and should be
investigated further, it appears that these can be grouped into
the trans- and cis-pairs as if each pair defines a boundary of similar
molecular interaction comprising the network of hydrogen bond.
Within each, although the trans/cis- pairs (e.g., bGlc- and
a-Gal-)
differ structurally, they may have similar cooperativity and this
probably affects the thermal properties in the same manner.
In order to visualize the differences in the hydrogen bonding
network that can explain the relative clearing temperatures for
epimeric–anomeric gluco/galacto-glycosides, a simple modeling
mal.28,34 The larger surface area of their head group affects
a-ano-
meric glycolipids more than b-analogs. The direction of the tilt is
not obvious. However, the effect on the surface area of the sugar
head group suggests that the tilting must involve the normal of
the carbohydrate ring. Figure 5 illustrates that the tilting of the
study was performed. The four different octyl glycosides (
aGal-
a
-glucoside reduces the potential interlayer directed hydroxyl
groups, as they face opposite sides of the tilting plane (Fig. 5a).
With the -galactoside, on the other hand (Fig. 5b), the hydroxyl
C8, bGal-C8, bGlc-C8 and Glc-C8) were constructed in Chem3D
a
and minimized with MOPAC using AM1. These molecules were
replicated without further minimization or molecular dynamics
simulation in a regular 2-D lattice based on a layer separation (of
the hydrophilic and the hydrophobic domain) and parallel align-
ment of the alkyl chains and the layer-normal. The images are dis-
played in Figures 4a–d. The examination of the images was
confined to the orientation of hydroxyl oxygen atoms rather than
to the interacting hydrogen atom, since the direction of the latter
is easily affected by intermolecular interactions. Therefore, the
structure optimization based on isolated molecules is likely to give
inaccurate impressions on the hydroxyl hydrogens.
Examining Figure 4, the structure (a) shows the lowest number
of interlayer directed hydroxyl groups for the b-galactoside. A com-
parison with the corresponding epimeric b-glucoside, (Fig. 4b) re-
veals that the axially oriented hydroxyl group at C-4 of the
galactoside contributes to intra- rather than interlayer H-bonding.
This stabilizes the assembly, subsequently leading to higher clear-
a
groups all face the same side of the tilting plane. This arrangement
leads to a higher interlayer H-bonding but a reduced intralayer H-
bonding consequently lowering the clearing temperatures for
these systems.
Based on these arguments, the intralayer H-bonding and subse-
quently the clearing temperature, are expected to decrease in the
following sequence:
bGal >
aGlc > bGlc ꢀ
aGal:
Differentiating the b-glucoside and the
a-galactoside is not
straightforward, given the fact that interlayer directed hydroxyl
groups may still contribute to intralayer H-bondings (Fig. 4d).
Moreover, the tilt angle of the alkyl chain and the layer normal af-
fects the H-bonding network (especially for the
a-galactoside);
hence the liquid crystal phase stability. Since the tilt angle is not

2954
R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
Figure 4. Modeling images for lamellar assemblies of anomeric octyl gluco- and galactosides; potential interlayer directed OH-groups are labeled on the left, while the right
side indicates possible intralayer OH-bonding.
the same for single and branched chain compounds, the relative
stability for these two is expectedly different. In single chain glyco-
sides the thermal phase stability, associated with the intralayer H-
the hydrophobic domain in branched (or double chained) glycolip-
ids enables the formation of a lamellar assembly with less tilting
than suitable for the single chain analog. It appears that the in-
crease in the tilt angle increases the interlayer orientation of the
hydroxyl groups, as can be anticipated from Figure 5b, assuming
a clockwise tilt. This reduces the intralayer hydrogen bonding
and, thus, the clearing temperature.
bonding, is higher for the b-glucoside than for the
branched analogs however, the reverse applies (Table 3). This can
be explained by the difference in the tilting angle of the -galacto-
sides. The reduced ratio of the surface areas for the hydrophilic and
a-galactoside. In
a

R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
2955
3
3
3
CH2), 0.90 (t, 6H, 2 CH3) ppm; J1,2 = 3.5, J2,3 = 10.0, J5,6a = 2.5,
3J5,6b = 5.0, 2J6 = 11.5 Hz.
Peracetate, 1H NMR (400 MHz, CDCl3): d = 5.44 (dd ꢀ t, H-3),
5.03 (dd ꢀ t, H-4), 5.01 (d, H-1), 4.81 (dd, H-2), 4.23 (dd, H-6a),
4.06 (dd, H-6b), 3.69 (2 dt, a-CH2a, 2 diastereomers due to racemic
Guerbet alcohol), 3.58 (ddd, H-5), 3.22 (dd,
a-CH2b), 2.06, 2.01,
2.00, 1.99 (4s, 4 Â 3H, 4 Ac), 1.55 (mc, b-CH), 1.33–1.19 (mc, 24H,
3
3
bulk-CH2), 0.84 (2t, 2 Â 3, CH3) ppm; J1,2 = 3.5, J2,3 = 10.0,
3J3,4 = 10.0,
3J4,5 = 10.0,
3J5,6a = 4.5,
3J5,6b = 2.0,
2J6 = 12.0,
2J = 10.0 Hz.
a
6.3. 2-Decyl-tetradecyl a-D-glucopyranoside (aGlc-C14C10)
Yield: 9 %. 1H NMR (400 MHz, CD3OD): d = 4.74 (d, H-1), 3.78
(dd, H-6a), 3.69 (dd, H-6b), 3.65–3.52 (m, 3H), 3.38 (dd, H-2),
3.37–3.26 (m, 2H), 1.65 (mc, b-CH), 1.45–1.22 (m, 36H, bulk-
3
3
3
Figure 5. Orientation of interlayer directed hydroxyl groups in octyl
a-
CH2), 0.90 (t, 6H, 2 CH3) ppm; J1,2 = 3.5, J2,3 = 10.0, J5,6a = 2.5,
3J5,6b = 5.0, 2J6 = 11.5 Hz.
glycopyranoside relative to the plane involving the largest molecular axes, that is,
the head-tail connection and the connection between C-6 and O-2 of the sugar. The
OH groups face different sides of the plane in the case of the glucoside (a), but face
the same side in the case of the galactoside (b).
Peracetate, 1H NMR (400 MHz, CDCl3): d = 5.45 (dd ꢀ t, H-3),
5.03 (dd ꢀ t, H-4), 5.02 (d, H-1), 4.81 (dd, H-2), 4.23 (dd, H-6a),
4.06 (dd, H-6b), 3.70 (mc,
a-CH2a), 3.58 (ddd, H-5), 3.23 (dd, a-
CH2b), 2.07, 2.03, 2.02, 2.00 (4s, 4 Â 3H, 4Ac), 1.56 (mc, b-CH),
1.33–1.18 (mc, 36 H, bulk-CH2), 0.89 (t, 2 Â 3, CH3) ppm;
5. Conclusions
3
3
3
3
3
3J1,2 = 3.5, J2,3 = 10.0, J3,4 = 10.0, J4,5 = 10.0, J5,6a = 4.5, J5,6b = 2.0,
2J6 = 12.0, 2J = 10.0 Hz.
The present work aims to demonstrate a possible epimeric–
a
anomeric relationship in pyranoside lipids, namely
a-gluco/b-
Acknowledgements
galactoside and -galacto-/b-glucoside. Both single and double
a
chain glycolipids were investigated to confirm the hypothesis; thus
the behavior applies to lamellar as well as to columnar assemblies.
However, due to the formation of a micellar phase in between
these two assembly types, the correlation does not match for a lim-
ited chain range of branched glycosides, that is, the Guerbet C16
(C10C6). This may be because the micellar phase is not liquid crys-
talline with a definite structural arrangement, or it could simply be
due to a purity issue of the samples (anomeric impurities as well as
traces of water). Relative clearing temperatures of lamellar sys-
tems can be explained by intralayer hydrogen bonding, which is
highly affected by the relative orientation of the 4-OH group and
the anomerically bonded aglycon.
R.H. thanks both the University of Malaya and the Osaka Sangyo
University for supporting her sabbatical leave, during which this
work was written. The grant UM.C/625/1/HIR/MOHE/05 is grate-
fully acknowledged.
References
1. Baron, M. Recomm. Pure Appl. Chem. 2001, 73.
2. Holmberg, K. Curr. Opin. Colloid Interface Sci. 2001, 6, 148–159.
3. Vill, V.; Hashim, R. Curr. Opin. Colloid Interface Sci. 2002, 7, 395–409.
4. Vill, V.; Fischer, F.; Thiem, J. Z. Naturforsch. 1988, 43a.
5. Abeygunaratne, S.; Hashim, R.; Vill, V. Phys. Rev. E: Stat. Nonlin. Soft Matter Phys.
2006, 73, 011916.
6. Liao, G.; Zewe, S. K.; Hagerty, J.; Hashim, R.; Abeygunaratne, S.; Vill, V.; Jákli, A.
Liquid Crystals 2006, 33, 361–366.
7. Athenstaedt, H. Naturwissenschaften 1961, 48, 465–472.
8. Miyajima, D.; Araoka, F.; Takezoe, H.; Kim, J.; Kato, K.; Takata, M.; Aida, T. J. Am.
Chem. Soc. 2010, 132, 8530–8531.
6. Experimental
6.1. Synthesis of a-glycosides
9. Takezoe, H. Polar Order in Columnar Phase of Glycolipids (personal
communication), 2007–2010.
A solution of b-glucose/galactose pentaacetate (4.5 mmol) and
the Guerbet alcohol (1.4 equiv) in dichloromethane (100 mL) was
treated with tin tetrachloride (1 equiv) and stirred at rt for 2 days.
The reaction mixture was filtered through water treated Celite be-
fore washing with saturated sodium bicarbonate solution. After
drying over magnesium sulfate the solvent was evaporated. The
residue was taken up in acetonitrile and the solution was extracted
with hexane. The acetonitrile phase was concentrated and the
crude acetylated glucosides were subjected to chromatography
10. Jeffrey, G. A. Acc. Chem. Res. 1986, 19, 168–173.
11. Jeffrey, G. A. Acta Crystallogr., Sect. B 1990, 46, 89–103.
12. Vill, V.; Böcker, T.; Thiem, J.; Fischer, F. Liq. Cryst. 1989, 6, 349–356.
13. Vill, V. LiqCryst 4.2, Fujitsu FQS, Fukuoka, 2002.
14. Vill, V.; von Minden, H. M.; Koch, M. H.; Seydel, U.; Brandenburg, K. Chem. Phys.
Lipids 2000, 104, 75–91.
15. von Minden, H. M.; Brandenburg, K.; Seydel, U.; Koch, M. H.; Garamus, V.;
Willumeit, R.; Vill, V. Chem. Phys. Lipids 2000, 106, 157–179.
16. Mannock, D. A.; Collins, M. D.; Kreichbaum, M.; Harper, P. E.; Gruner, S. M.;
McElhaney, R. N. Chem. Phys. Lipids 2007, 148, 26–50.
17. Hato, M.; Minamikawa, H.; Tamada, K.; Baba, T.; Tanabe, Y. Adv. Colloid
Interface Sci. 1999, 80, 233–270.
using hexane and ethyl acetate. The pure
a-fraction was collected
18. Hashim, R.; Hashim, H. H. A.; Rodzi, N. Z. M.; Hussen, R. S. D.; Heidelberg, T.
Thin Solid Films 2006, 509, 27–35.
and subsequently deacetylated in methanol with
a
catalytic
amount of sodium methoxide. The catalyst was removed by treat-
ment with Amberlite IR-120 and the solvent was evaporated to
provide the glucoside, which was finally dried at 50 °C in high vac-
uum over phosphorus pentoxide overnight.
19. O’Lenick, A. J.
A Review of Guerbet Chemistry. http://www.zenitech.com/
documents/guerbet_chemistry.pdf (November, 5).
20. Brooks, N. J.; Hamid, H. A. A.; Hashim, R.; Heidelberg, T.; Seddon, J. M.; Conn, C.
E.; Husseini, S. M. M.; Zahid, N. I. M.; Hussen, R. S. D. Liq. Cryst., 2011.
doi:10.1080/02678292.2011.625689.
21. Kikuchi, Y.; Tanaka, Y.; Sutarto, S.; Kobayashi, K.; Toi, H.; Aoyama, Y. J. Am.
Chem. Soc. 1992, 114, 10302–10306.
22. Kikuchi, Y.; Toi, H.; Aoyama, Y. B. Chem. Soc. Jpn. 1993, 66, 1856–1858.
23. Nilsson, F.; Söderman, O.; Johansson, I. J. Colloid Interface Sci. 1998, 203, 131–
139.
24. Rodzi, N. B. M. Branched Chain Galactosides and Melibiosides: Synthesis and
Mesomorphic Properties; University of Malaya: Kuala Lumpur, 2006 (M.Sc.).
25. Sakya, P.; Seddon, J. M.; Vill, V. Liq. Cryst. 1997, 23, 409–424.
6.2. 2-Hexyl-decyl a-D-glucopyranoside (aGlc-C10C6)
Yield: 12%. 1H NMR (400 MHz, CD3OD): d = 4.74 (d, H-1), 3.78
(dd, H-6a), 3.68 (dd, H-6b), 3.66–56 (m, 3H), 3.39 (dd, H-2),
3.36–3.27 (m, 2H), 1.65 (mc, b-CH), 1.45–1.22 (m, 24H, bulk-

2956
R. Hashim et al. / Carbohydrate Research 346 (2011) 2948–2956
26. Goodby, J. W. Mol. Cryst. Liq. Cryst. 1984, 110, 205–219.
27. Dorset, D. L.; Rosenbusch, J. P. Chem. Phys. Lipids 1981, 29, 299–307.
28. Chong, T. T.; Heidelberg, T.; Hashim, R.; Gary, S. Liq. Cryst. 2007, 34, 251–265.
29. Mombelli, E.; Morris, R.; Taylor, W.; Fraternali, F. Biophys. J. 2003, 84, 1507–
1517.
31. Carmona, A. T.; Moreno-Vargas, A. J.; Robina, I. Curr. Org. Synth. 2008, 5, 33–60.
81–116.
32. Ferrieres, V.; Bertho, J.; Plusquellec, D. Tetrahedron Lett. 1995, 36, 2749–2752.
33. Hori, R.; Ikegami, Y. Yakugaku Zasshi (J. Pharm. Soc. Japan) 1959, 79, 80.
34. Nguan, H. S.; Heidelberg, T.; Hashim, R.; Tiddy, G. J. T. Liq. Cryst. 2010, 37, 1205–
1213.
30. Róg, T.; Vattulainen, I.; Bunker, A.; Karttunen, M. J. Phys. Chem. B 2007, 111,
10146–10154.