Journal of Oleo Science

Copyright ©2014 by Japan Oil Chemists’ Society

doi : 10.5650/jos.ess14086
J. Oleo Sci. 63, (12) 1309-1322 (2014)

Physicochemical and Structural Properties of

Glycerin Gel Prepared using Glycyrrhizic Acid
Diethyl Ester
Kenjiro Koga1, Toshiyuki Kimura1, Kenichi Sakai2, Hiroshi Kushida3 and
Nobuji Yoshikawa3＊
1
Faculty of Pharmaceutical Sciences, Hokuriku University, Ho-3, Kanagawa-machi, Kanazawa 920-1181, Japan
2
Faculty of Science and Technology, Tokyo University of Science2641, Yamazaki, Noda 278-8510, Japan
3
Research and Development Division, Cokey Systems Co. Ltd., 3639-7 Kamikawa-machi, Matsusaka 515-0041, Japan

Abstract: Glycyrrhizic acid diethyl ester (GZ-DE) was developed as a prodrug of glycyrrhizic acid (GZ), a
hepatitis therapeutic drug. We fortuitously found that GZ-DE gels with glycerin selectively while searching
for a safe solvent with which to dissolve GZ-DE. Based on this gelation, the aim of this study was to
investigate the preparation of the gel and study the rheology, physicochemical and structural properties of
the glycerin gel by differential scanning calorimeter (DSC), capillary electrophoresis (CEP), nuclear
magnetic resonance (NMR), and small angle X-ray scattering (SAXS). The glycerin gel was prepared by the
addition of at least 2.0% w/w GZ-DE. This gel did not flow at room temperature. After mixing glycerin and
GZ-DE, a gel was formed after 2 days at 25℃ or 3 h at 60℃. Glycerin gel containing 2.4% w/w GZ-DE
provided the following results: 1) The glycerin gel exhibited creep at a constant stress of less than 10 Pa, but
it is a fragile gel, showing Newtonian flow at 10 Pa stress. 2) Dynamic viscoelastic measurements showed
that the elastic modulus (G’) exceeds the viscous modulus (G’’), indicating that glycerin gel has solid-like
properties. 3) DSC showed a significant difference between the glass transition temperature of glycerin and
glycerin gel. 4) CEP did not reveal a new compound in the glycerin gel. 5) NMR confirmed that glycerin gel
is a physical gel. 6) SAXS measurements revealed that the glycerin gel has an oval-shaped basic frame (119
nm long and 65 nm wide).

Key words: glycerin, glycyrrhizic acid, gelation, rheology, NMR

1 INTRODUCTION
Glycyrrhizic acid diethyl ester（GZ-DE） is a compound in
which two ethyl esters bind to the carboxyl group of two
glucuronic acids in the structure of glycyrrhizic acid（Fig.
1） （GZ）
. Glycyrrhizic acid is a drug that improves the func-
tion of the liver, but is poorly absorbed from the intestinal
tract1−3）. Therefore, GZ-DE was synthesized in 2012 as a
prodrug for the purpose of improving the absorption of GZ
from the intestinal tract（CAS registry number: 1413393-
38-0）. GZ-DE does exhibit improved absorption compared
to GZ4）, but there are several problems with the physico-
chemical properties of GZ-DE. First, GZ-DE is very poorly Fig. 1 Structures of glycyrrhizic acid monoammonium
soluble in water at room temperature （less than 1 mg/mL at (A) and glycyrrhizic acid diethyl ester (GZ-DE)
20℃）. Second, GZ-DE gradually decomposes in aqueous (B).
solution; for example, the degradation rate constant of
GZ-DE in 50 mM phosphate buffered solution （pH 7.4） and

＊
Correspondence to: Nobuji Yoshikawa, Research and Development Division, Cokey Systems Co. Ltd., 3639-7 Kamikawa-machi,
Matsusaka 515-0041, Japan
E-mail: yoshikawa@cokey.co.jp
Accepted August 22, 2014 (received for review April 29, 2014)
Journal of Oleo Science ISSN 1345-8957 print / ISSN 1347-3352 online
http://www.jstage.jst.go.jp/browse/jos/ http://mc.manusriptcentral.com/jjocs

1309
 K. Koga, T. Kimura, K. Sakai et al.
saline（0.9％ NaCl）is 0.088 h−1 and 0.020 h−1, respective-
ly4）. Thus, a safe solvent was required to solubilize GZ-DE.
While investigating solvents, we fortuitously discovered
that GZ-DE gels with Japanese pharmacopoeia glycerin （JP
glycerin; glycerin/water（ 85:15, v/v））when left at room
temperature after heating at 60℃ for 3 h. We then investi-
gated whether or not GZ-DE gels with glycerin-like com-
pounds such as propylene glycol, 1,3-propanediol,
1,2,6-hexanetriol, 2-methyl-1-propanol, and 1-propanol, in
order to clarify the specific selectivity of GZ-DE for gela-
tion with glycerin. GZ-DE did not gel these glycerin-like
compounds, indicating that GZ-DE selectively gels with
glycerin. GZ-DE has very interesting characteristics as a
selective gelling agent. Previous studies have reported that
12-hydroxy stearic acid gels in many kinds of oil by forming
（2-aminoethyl）
fiber-like micelles in the oil phase5, 6）. N- -α-
［（1-oxoheptadecyl）amino］acetamide derivatives self-as-
semble and gel in various common solvents7）. For example,
bola-type peptide lipid can gelate water, organic solvents,
and water/organic-solvent mixtures8）. However, the com-
pound which gels with glycerin was not found from docu-
ments search except for the effect of glycerin on polymer
gel. Therefore, the gelation characteristics of GZ-DE are
unique as a gelator for glycerin. Our research aims to
clarify the mechanism of gelation of glycerin by GZ-DE. In
this paper, we investigated both the preparation of glycerin
gel and the physicochemical and structural properties of
the glycerin gel.
2 MATERIALS AND METHODS
2.1 Materials
GZ-DE was synthesized by Cokey Systems（Matsusaka,
Japan） （GLY0605, 18beta type, purity 100％）
. GZ standard
was purchased from National Institute of Health Sciences
（Tokyo, Japan）. Japanese Pharmacopoeia glycerin（ JP
glycerin; glycerin/water（85:15, v/v））was purchased from
Nikko Seiyaku Co. Ltd.（Gifu, Japan）. Pure glycerin（JIS
special grade, ＞99％ purity, 0.13％ water content）was
purchased from Wako Pure Chemicals Industries（Osaka,
Japan）. Other chemicals were of reagent grade.
2.2 Preparation of glycerin gel
The standard method for preparing glycerin gel was as
follows: GZ-DE（0.5％-2.5％ w/w final concentration）and
JP glycerin were added to a glass screw-top vial （50 mL, 30
mm inner diameter）. The mixture weight of GZ-DE and
glycerin was set to be 10 g. After sonicating in a bath-type
sonicator（ 1210, Yamato, Tokyo, Japan）for 10 min, the
mixture was heated at 60℃ for 3 h using a constant tem-
perature oven（DX-402, Yamato, Tokyo, Japan） . elastic modulus was calculated by:（980（dyn constant）×
Glycerin gel containing ethanol was prepared as follows:
0.10 g of GZ-DE, 0.735 g of ethanol（30％, 40％, 45％, or
1310
J. Oleo Sci. 63, (12) 1309-1322 (2014)
Fig. 2 Fluidity evaluation test of glycerin gel. A screw-
top glass vial containing glycerin gel (10 g) was
quickly placed on its side at 25℃, then the time
required for the edge of the gel to travel a dis-
tance of 4 cm from the bottom of the vial was
measured.
50％ v/v ethanol concentration prepared with distilled
water）, and 4.165 g of pure glycerin were added to a glass
screw-top vial. The mixture was sonicated in a bath-type
sonicator（1210, Yamato）for 10 min, heated at 60℃ for 3 h
using a constant temperature oven（DX-402, Yamato）.
To confirm if heating at 60℃ was necessary for gelation,
a mixture solution of JP glycerin containing 2.4％ w/w
GZ-DE was kept at 20℃ for at least 2 days. Moreover, to
confirm if water was necessary for gelation, pure glycerin
containing 2.4％ w/w GZ-DE was sonicated in a bath-type
sonicator（1210, Yamato）for 10 min, heated at 60℃ for 3 h
using a constant temperature oven（DX-402, Yamato）.
2.3 Fluidity of the glycerin gel
Five glycerin gels（10 g）containing GZ-DE（0.5％, 1.0％,
1.5％, 2.0％, and 2.5％ w/w） were prepared in 50 mL glass
screw-top vials（30 mm inner diameter）, based on the stan-
dard preparation of glycerin gel described above. The con-
tainer was quickly placed on its side at 25℃ and the time
required for the edge of the gel to flow to a mark 4 cm from
the bottom of the container was measured（Fig. 2） .
2.4 Effect of GZ-DE concentration on the rheology of
glycerin gel
Four glycerin gels（10 g）containing GZ-DE （0.5％, 1.0％,
2.0％, and 2.5％ w/w） were prepared in 50-mL glass screw-
top vials（30 mm inner diameter）as described above and
placed in a rheometer（CR-500DZ-S II, Sun Scientific Co.
Ltd., Tokyo, Japan） equipped with a plunger（15 mm diame-
ter）at 25℃. The compression speed was 60 mm/min and
the compression length was 5 mm. The following rheology
parameters of each glycerin gel were measured: maximum
load, gel strength, degree of hardness, and elastic modulus.
The maximum load shown on the y-axis was the directly
measured maximum load. Gel strength is the value ob-
tained by dividing the maximum load by the plunger area.
Hardness was calculated by the formula:（gel strength×
sample height）/distance travelled at maximum load. The
maximum load×sample height）/distance travelled at
maximum load.
 Physicochemical and structural properties of glycerin gel
Fig. 3 Rheology parameters: the relationship between
force and fracture distortion or fracture distance
in the creep curve of glycerin gel.
2.5 Effect of ethanol on the rheology of glycerin gel
Glycerin gels prepared as described in Section 2.2 were
kept at 25℃. These gel samples were tested using a creep-
meter（ RE2-33005C, Yamaden Co. Ltd., Tokyo, Japan）
equipped with a plunger（ 5 mm diameter）at 25℃. The
compression speed was 60 mm/min. The fracture force,
fracture strain, fracture distortion, and frailty force were
measured to study the rheology of glycerin gels containing
ethanol. These parameters were obtained from relation-
ships between force and fracture distortion（fracture dis-
tance） as shown in Fig. 3.
2.6 Rheology of glycerin gel
2.6.1 Creep recovery test of glycerin gel on steady shear-
ing gap
（10 g）containing 2.4％ w/w GZ-DE was pre-
Glycerin gel
pared as described above. This hard glycerin gel did not
flow even when the vial was held upside down. The gel was
put on the parallel plate of a rheometer（AR1000, TA In-
struments Japan, Tokyo, Japan） equipped with a 20 mm di-
ameter plunger. After leaving the gel on the parallel plate
for 2 min at 25℃, a shearing stress（2 Pa, 5 Pa, and 10 Pa）
was applied to retard the gel for 3 min at a 1000 μm gap
condition, then the shearing stress was removed and the
gel was left to recover for 5 min. The load strain ％ of the
gel was measured at 25℃ throughout the process.
2.6.2 Dynamic viscosity of glycerin gel
Glycerin gel containing 2.4％ w/w GZ-DE was prepared
as described above. A fixed amount of gel was put on the
plate of a rheometer （AR-G2, TA Instruments Japan, Tokyo,
Japan）equipped with a temperature control system.
Dynamic viscosity assays were carried out while varying
（0-500 Pa）
two factors, oscillation stress and angle frequen-
cy（0-100 rad/s）, at 25℃ and 50℃. Dynamic oscillation pa-
rameters, elastic modulus （G’ ）
, viscous modulus（G”） , and
J. Oleo Sci. 63, (12) 1309-1322 (2014)
complex viscosity（ η*）were measured as a function of
angular frequency, strain, and temperature to evaluate the
viscoelasticity of glycerin gel containing 2.4％ w/w GZ-DE.
2.7 Assay of reaction product after preparation of glycerin
gel
The gelling of glycerin by the addition of GZ-DE suggest-
ed the possibility that new compounds via chemical reac-
tion. The presence of new chemical products was examined
using a capillary electrophoresis apparatus（Agilent Tech-
nologies, HP3D, Tokyo, Japan）in order to test this hypothe-
sis.
The analytical conditions were as follows: mobile phase;
25 mM sodium dodecyl sulfate-10 mM sodium tetraborate
buffer solution（pH 9.4）, injection pressure; 50 mbar×4
sec, applied voltage; 17 kV, fused silica bubble cell capillary
（64.5 cm long, 75 μm inner diameter） , wavelength; 190-500
nm（248 nm principle wavelength）. The injection samples
were JP glycerin, 200 μg/mL GZ-DE ethanol solution, 20
mg/mL glycerin gel（with 2.4％ w/w GZ-DE）ethanol solu-
tion, and an ethanol solution of GZ-DE/glycerin gel （1:1, v/
v）.
2.8 Thermodynamic properties of glycerin gel by DSC
The thermodynamic properties of glycerin gel containing
2.4％ w/w GZ-DE were investigated using a differential
scanning calorimeter（DSC, RDC220, Hitachi High-Tech
Science Corporation, Tokyo, Japan）. The temperature
control program was as follows: 1）JP glycerin and glycerin
gel were heated at 30℃ for 20 min, 2）both samples were
cooled at the speed of 10℃/min to -150℃, 3）immediately
both samples were heated at the same speed（10℃/min）to
120℃. This operation was repeated two times. N2 gas infu-
sion was 40 mL/min, and the weight of the samples（glycer-
in gel and JP glycerin）was approximately 30 mg.
2.9 Molecular dynamics of glycerin gel by NMR
Molecular interactions between glycerin and GZ-DE
were investigated by 1H-NMR（EX-400 and ECA-400, JEOL
Ltd., Tokyo, Japan）and 13C-NMR（EX-400 and AL-400,
JEOL Ltd., Tokyo, Japan）. The samples were JP glycerin
and glycerin gel containing 2.4％ w/w GZ-DE. Both NMR
assays were carried out using single pulse methods. Each
NMR assay was started 30 min after the sample was pre-
pared in an NMR tube. Trichloromethane（7.26 ppm）and
tetramethylsilane（ 0 ppm）were used as standard com-
pounds in the 1H-NMR and 13C-NMR assays, respectively.
2.10 Structural characteristics of glycerin gel by SAXS
Structural analysis by small-angle X-ray scattering
（SAXS, SAXSSess mc2, Anton Paar Japan, Tokyo, Japan）
was carried out to clarify the basic unit of the glycerin gel
network. The assay conditions were as follows: radiation
source; Cu K α-ray, power; 40 kV 50 mA, focus; line focus,
1311
 K. Koga, T. Kimura, K. Sakai et al.
focus size（filament size）; 0.4×12 mm2, detector; imaging
plate, detection length; 261.2 mm, cell; liquid cell, determi-
nation range; 0.04̶2.8 nm−1, detection time; 1 h, determi-
nation temperature; 25℃.
3 RESULTS AND DISCUSSION
3.1 Basic conditions for preparation of glycerin gel
The strength of the glycerin gel increased as the concen-
tration of GZ-DE added to JP glycerin increased. For
example, glycerin gel containing 0.5％ w/w GZ-DE was
very soft, but glycerin gel containing 2.5％ w/w GZ-DE was
hard and did not flow even when the glass vial with the
glycerin gel was tipped over or laid sideways. These results
showed that a higher weight ratio of GZ-DE in JP glycerin
was important for a hard glycerin gel. JP glycerin contains
approximately 15％ water by volume. Therefore, the influ-
ence of water was studied to determine the conditions re-
quired for gel formation. Glycerin gel was prepared by
adding GZ-DE（2.5％ w/w）to pure glycerin（0.13％ water
content）in order to confirm the contribution of water to
gel formation. After heating at 60℃ for 3 h, a glycerin gel
was obtained that was similar to that obtained using a
mixture of JP glycerin and GZ-DE（2.5％ w/w）. However,
much of the GZ-DE was trapped as a solid in the formed
glycerin gel, suggesting that the solubility of GZ-DE in pure
glycerin is lower than in JP glycerin. These results indicate
that the 15％ water content in JP glycerin has little influ-
ence on gel formation.
The standard preparation of glycerin gel involves heating
at 60℃ for 3 h. However, the necessity for heating is not
based on detailed experiments, and may not be required
for the preparation of glycerin gel. Therefore, a mixture of
JP glycerin and GZ-DE（2.4％ w/w in the gel）was kept at
20℃ for at least 2 days. A gel formed, indicating that
heating accelerated gel formation, but the physical and/or
chemical interactions between glycerin and GZ-DE gradu-
ally progressed in the absence of added heat.
3.2 Fluidity of glycerin gel
Table 1 shows the relationship between GZ-DE concen-
tration and the time required for the glycerin gel to flow a
given distance （4 cm from the bottom of a screw-top glass
vial）.
Glycerin gel containing 0.5％ w/w GZ-DE flowed almost
like a liquid. The physicochemical properties of glycerin
containing GZ-DE changed from that of a soft gel to a hard
gel depending on the GZ-DE concentration, and the flow
time increased with an increase in GZ-DE concentration
（0.5％̶2.0％ w/w）. Glycerin gel containing 2.5％ w/w
GZ-DE did not flow. These results suggested that GZ-DE
shows interaction for glycerin molecules and inhibits the
fluidity of glycerin molecules depended on the GZ-DE con-
1312
J. Oleo Sci. 63, (12) 1309-1322 (2014)
Table 1 T
 he relationship between GZ-DE concen-
tration and the time required for the glycer-
in gel to flow a set distance.
GZ-DE concentration flow time
(% w/w) (sec)
0.5 0.67±0.06
1.0 1.63±0.06
1.5 3.50±0.10
2.0 9.33±0.47
2.5 not flow
Data represent the mean±SD from three experiments. The
distance was set at 4 cm from the bottom of a screw-top
glass vial.
centration.
3.3 Effect of GZ-DE concentration on the rheology of
glycerin gel
Rheological studies are important for structural charac-
terization of the gel. Recently, many researchers have re-
ported data obtained using a rheometer9−12）. The influence
of GZ-DE on fluidity of glycerin became clear, as described
above, and further detailed rheological evaluations were
carried out using a rheometer. Figure 4 and Table 2 show
the relationship between compression load and compres-
sion length obtained in compression and rheology tests, re-
spectively, for four glycerin gels containing different con-
centrations of GZ-DE.
At a GZ-DE concentration of 2.5％ w/w, the maximum
load, gel strength, hardness and elastic modulus increased
remarkably compared with the other glycerin gels, suggest-
ing that the critical concentration of GZ-DE which drasti-
cally changes the properties of the gel is approximately
2.5％. Furthermore, the maximum load tolerated by glyc-
erin gels prepared with 0.5, 1.0, and 2.0％ w/w GZ-DE was
in close agreement with the compression length（5 mm）.
On the other hand, maximum load tolerated by glycerin gel
prepared with 2.5％ w/w GZ-DE remained stable to a com-
pression length of 0.78 mm, then the load rapidly de-
creased from a compression length of 0.78 mm to 5 mm,
indicating that the glycerin gel was fragile and easily col-
lapsed.
3.4 Effect of ethanol concentration on the rheology of
glycerin gel
Ethanol was added to completely dissolve the GZ-DE
since the solubility of GZ-DE in JP glycerin or pure glycerin
is low: some of the GZ-DE did not dissolve in the glycerin
gel when the gel was prepared using more than a 2.0％ w/
w final concentration of GZ-DE. Indeed, when the GZ-DE
concentration was 2.5％ w/w for the preparation of glycerin
gel, the gel did not look homogeneous: some of the GZ-DE
solid-like powder was dispersed in the gel. This heteroge-
 Physicochemical and structural properties of glycerin gel

Fig. 4 Compression load versus compression length curves in glycerin gels with different concentrations of GZ-DE.
Line a: 0.5% w/w, line b: 1.0% w/w, line c: 2.0% w/w, line d: 2.5% w/w GZ-DE.

Table 2 ‌ heology parameters of glycerin gels containing different concentrations of

R
GZ-DE.
GZ-DE concentration max. load gel strength hardness elastic modulus
(% w/w) (N) (Pa) (Pa) (N)
0.5 0.070 396 79.3 0.0140
1.0 0.072 408 82.4 0.0146
2.0 0.112 634 126.8 0.0224
2.5 0.576 3261 4207.8 0.7432

Table 3 Rheology parameters of glycerin gels with different ethanol concentrations.

ethanol concentration fracture force fracture strain fracture distortion frailty force
(% w/w) (N) (mm) (%) (N)
0 n.d. n.d. n.d. n.d.
4.41 0.144 0.19 4.87 0.04
5.88 0.190 0.21 5.29 0.08
6.61 0.342 0.19 5.22 0.18
7.35 0.356 0.10 2.65 0.24
The GZ-DE concentration in the glycerin gels was 2.0% w/w.
n.d. indicates no detection when the ethanol concentration was 0%.

neity, with dispersed GZ-DE, made it difficult to precisely
evaluate the rheological characteristics of the glycerin gel.
Therefore glycerin gel was prepared by adding ethanol
（4.41％̶7.35％ w/w final concentration）to completely
solubilize GZ-DE during gel preparation. Table 3 shows the
effect of ethanol concentration on the rheology of glycerin
gel.
No undissolved particles of GZ-DE were observed in the
four glycerin gels, and the gels appeared homogeneous.
The fracture force and fracture distortion increased with
J. Oleo Sci. 63, (12) 1309-1322 (2014)
an increase in ethanol concentration（Table 3） , suggesting
that the gels changed in a solid-like state and that their fra-
gility depended on the ethanol concentration. Although
glycerin gel（2.0％ w/w GZ-DE）without ethanol did not
provide a clear fracture point, the four glycerin gels with
ethanol each provided a definite fracture point and frailty
point in the force versus fracture distortion curve. These
results indicate that the addition of ethanol enhances the
dissolution of GZ-DE in glycerin and ensures the formation
of a glycerin gel: the solubility of GZ-DE in glycerin clearly
1313
 K. Koga, T. Kimura, K. Sakai et al.

affects gel rheology, and the rheological properties of the
glycerin gel are controlled by the ethanol concentration
when the GZ-DE concentration is 2.0％ w/w.
3.5 Rheology of glycerin gel on a constant shearing gap
Figure 5 shows the relationship between the ％ stress
and total time of glycerin gel containing 2.4％ w/w GZ-DE
under a shearing stress of 2 Pa, 5 Pa, and 10 Pa. When the
shearing stress was 2 Pa and 5 Pa, creep and creep recov-
ery phenomena were observed at a stress time of 3 min
and at a recovery time of 5 min, respectively. Both results
suggested that the rheological properties of viscosity and
elasticity of glycerin gel are similar to that of the Voigt
model at a shear stress of 2 Pa and 5 Pa. On the other
hand, when the shear stress was 10 Pa, the shearing rate
increased linearly for 3 min and then remained constant
during a 5 min recovery time: the glycerin gel exhibited
Newtonian flow under a 10 Pa shear stress. These results
mean that elasticity in the glycerin gel disappears between
5 Pa and 10 Pa. This rheological property indicates that
glycerin gel containing 2.4％ w/w GZ-DE is similar to agar
gel, a polymer jelly with high elasticity.
Figure 6 shows the dynamic behavior of the elastic
modulus（G’ ）, viscous modulus（G” ） , and complex viscosi-
ty（η*）in glycerin gel （2.4％ w/w GZ-DE） at 25℃ and 50℃.
The G’ , G”, and η* values changed at the same point in the
range between 10 Pa to 100 Pa oscillation stress: when the
oscillation stress was 51.8 Pa at 25℃ and 37.3 Pa at 50℃,
these values decreased sharply, depending on an increase
in the oscillation stress. Furthermore, the G’value was
higher than the G”value at lower stress, rather than at the
change point. The results suggested that the elasticity of
glycerin gel changes in tandem with the viscosity until the
glycerin gel collapses. The glycerin gel becomes brittle due
to the changes in G’, G”, and η* after the collapse of the
gel. These data support the rheological properties of glyc-
erin gel shown in Figs. 4 and 5. On the other hand, the G’
and G”values, which depend on an increase in the angle
frequency, slowly increased between 25℃ and 50℃,
whereas η* decreased linearly. Because oscillation stress
levels were in the range 2.48̶8.51 Pa at 25℃ and 2.75̶
5.05 Pa at 50℃, the glycerin gel did not collapse. Similar to
the results shown in Figs. 6A and B, the G’value was
higher than the G”value at both temperatures.
3.6 Confirmation of compounds in the glycerin gel by
capillary electrophoresis
In the glycerin gel prepared by the addition of GZ-DE, it
was clear that warming at 60℃ increased the speed of gel
formation. In general, there are two types of gels: physical
gels and chemical gels. Physical gels show temperature de-
pendence such as sol-gel changes, whereas chemical gels
are generated by the formation of new chemical bonds
during a chemical reaction, and often do not show temper-
ature dependence.
Glycerin gel（2.4％ w/w GZ-DE） did not show a gel to sol
change in the temperature range of 60℃ to 100℃. There-
fore, it was necessary to clarify whether glycerin gel is a
physical gel or a chemical gel. If it is a chemical gel, new
compounds, such as a complex formed by the chemical re-

Fig. 5 Strain versus time curves on a steady shearing gap. A: 2 Pa, B: 5 Pa, C: 10 Pa. Stress time and recovery time
were 3 min and 5 min, respectively. Experiments were carried out at 25℃.
1314
J. Oleo Sci. 63, (12) 1309-1322 (2014)
 Physicochemical and structural properties of glycerin gel

Fig. 6 Elastic modulus (G’), viscous modulus (G”), and complex viscosity (η*) of glycerin gel based on changes in
oscillation stress and angle frequency. A and C: 25℃, B and D: 50℃.

action of GZ-DE and glycerin will be produced in the glyc-
erin gel. Capillary electrophoresis was used to identify this
complex and thus confirm this hypothesis. Figure 7 shows
the chromatograms of JP glycerin ethanol solution, GZ-DE
dissolved in ethanol, glycerin gel ethanol solution, and an
ethanol solution of GZ-DE/glycerin gel （1:1, v/v） .
Ethanol was used as a solvent to prepare four samples.
The major peak from the GZ-DE sample was detected at 13
min（Figs. 7B, C and D） . Small peaks corresponding to im-
purities in GZ-DE were observed in Figs. 7B, C and D.
However, no large unique peak from the glycerin gel was
observed. These results indicate that GZ-DE and glycerin
do not form a new compound by a chemical reaction sug-
gesting that glycerin gel is likely a physical gel.
3.7 Calorimetric assay of glycerin gel by DSC
Figure 8 shows DSC profiles of JP glycerin and glycerin
gel（2.4％ w/w GZ-DE）in the decreasing and increasing
temperature processes as the interval between −120℃
and 120℃.
Although the freezing point of JP glycerin is −10.9℃13, 14）,
no phase transition peak was observed at −10.9℃. This
was expected, as in general glycerin does not show a phase
transition for the freezing of the crystalline form unless the
temperature change is extremely slow. Namely, the tem-
perature change of 10℃/min caused a state of supercooled
JP glycerin in temperature less than freezing point（−
10.9℃）. This shows that JP glycerin and glycerin gel re-
J. Oleo Sci. 63, (12) 1309-1322 (2014)
mained a liquid state without even freezing point or less
becoming the solid by not having regular arrangement of
glycerin molecules for this temperature change（10℃/min） .
In addition, it was reported that the condition to crystallize
glycerin at freezing point temperature was to heat glycerin
at the speed of 1℃/min after cooling to −80℃ at the
speed of 5℃/min15）. If there is little interaction between
GZ-DE and glycerin, the DSC curve of glycerin gel will be
similar to that of a simple physical mixture of GZ-DE and
glycerin. However, the DSC curves of glycerin gel were dif-
ferent from those of JP glycerin, with a pronounced transi-
tion at −101.2℃ for JP glycerin and −89.7℃ for glycerin
gel at the range of cooling temperature process, and at −
92.1℃ for JP glycerin and −81.2℃ for glycerin gel at the
range of heating temperature process. Moreover, no differ-
ence was observed for those calculated temperature data
in repeated experiments.
It is the phenomenon that a supercooled liquid loses ki-
neticism, and these temperatures are glass transition tem-
perature（Tg）16, 17）. In a supercooling state, it is shown that
the decrease in big capacity while releasing heat outside a
system at the time of the cooling. On the other hand, both
compounds with glass state absorb heat and thermal
motion becomes active, then the capacity expands sudden-
ly at the time of the heating. Therefore, Tg in cooling
process is different from that in heating process. In fact,
DSC curve of glycerin in heating process was approximate-
ly 10℃ high compared with that in cooling process based
1315
 K. Koga, T. Kimura, K. Sakai et al.
Fig. 7  hromatograms showing detected compound
C molecular motion is 400 MHz. Table 4 shows the relation-
peaks in four samples using capillary electro-
phoresis. A: JP glycerin ethanol solution, B: 200
µg/mL GZ-DE ethanol solution, C: 20 mg/mL
glycerin gel (2.4% w/w GZ-DE) ethanol solu-
tion, D: ethanol solution of GZ-DE/ glycerin gel
(1:1, v/v).
on excess heat capacity18）. Although the Tg at the time of
the heating was different from Tg in cooling based on
quantity of thermodynamics, the differences of Tg between
JP glycerin and glycerin gel at the time of cooling and
heating were approximately 11℃. From these results, it is
suggested that GZ-DE interacts with glycerin molecules.
The addition of GZ-DE to JP glycerin restricts the motion
of glycerin molecules motion, thus possibly raising the Tg
by approximately 11℃. Therefore, it seemed that GZ-DE
caused the formation of glycerin gel due to molecular in-
teractions.
3.8 Molecular dynamics of glycerin gel by NMR
3.8.1 Relaxation time by 1H-NMR
NMR measurements were carried out to clarify the
mechanism of gel formation by the interaction of glycerin
1316
J. Oleo Sci. 63, (12) 1309-1322 (2014)
and GZ-DE. It was expected that the molecular motion of
glycerin would be decreased in the presence of GZ-DE, de-
pending on the stage of gel formation. Spin-lattice relax-
ation time（ T1）measurements by 1H-NMR are useful to
evaluate this molecular motion19）. Thus, 1H-NMR measure-
ments of JP glycerin were carried out first to confirm the
relationship between glycerin molecular motion and T1.
The result is shown in Fig. 9.
Figure 10 shows the relationship between T1 of 1H and
temperature. Clearly, the molecular motion of glycerin in-
creases with an increase in temperature between 10̶
40℃. On the other hand, the T1 of the CH and CH2 posi-
tions of glycerin were at a minimum between 20̶25℃. In
general, this minimum depends on the strength of the NMR
magnetic field, and the rotation or vibration of the hydro-
gen is the same as the resonance frequency of the NMR.
Therefore, the molecular motion of glycerin based on the
gelation time-process can be evaluated from the molecular
motion at the T1 minimum（NMR resonance frequency）.
The T1 of CH2（ a）was shorter than that of CH（b）
because of the difference in atomic distance between H-H.
The dipolar interaction between the nuclear spin is
changed by molecular motion, resulting in relaxation due
to dipolar interactions. In general, dipolar interactions are
inversely proportional to the distance between the spin to
the sixth power20）. It was estimated that the atomic dis-
tance between the H-H of CH2（a）was shorter than that of
CH（b）because the efficiency of relaxation was improved,
resulting in a shortened T1.
In general, T1 is at a minimum when the frequency of
ship between temperature, molecular motion, and T1.
Even if the frequency of molecular motion is lower or
higher than 400 MHz, T1 is long. When the temperature of
JP glycerin was ca. 20℃, T1 was the shortest, whereas
when the temperature was lowered to 10℃ or raised to
40℃, the frequency of molecular motion was higher than
400 MHz or lower than 400 MHz, respectively.
From the relationship between the frequency of molecu-
lar motion and T1, it appears that the change in the fre-
quency of molecular motion reflects the time required for
glycerin gel formation, and that molecular motion, which is
dependent on the progress of glycerin gel formation, can
be evaluated by T1.
T1 of CH（b） and CH2（a） in JP glycerin with GZ-DE （2.4％
w/w GZ-DE） was monitored（Fig. 11） .
T1 of CH（b） was shortened in a time dependent manner
at the range of 280̶380 msec until approximately 160 h,
then remained constant（282-300 msec）. The shortening of
T1 of CH（b）indicated gel formation by the addition of
GZ-DE. Gel formation proceeded rapidly until approxi-
mately 160 h after mixing of JP glycerin and GZ-DE. The
suppression of molecular motion during gel formation
strongly contributed to the T1 of CH（b）compared with
 Physicochemical and structural properties of glycerin gel

Fig. 8 DSC curves for JP glycerin (A, B) and glycerin gel (2.4% w/w GZ-DE) (C, D) at the range of cooling (A, C)
and heating (B, D) temperature processes.

1
Fig. 10 R
 elationship between T1 of H and temperature
for the CH and CH2 groups of glycerin.
1
Fig. 9 H-NMR spectrum of JP glycerin.
‌ elationship between temperature,
Table 4 R
molecular motion, and T1.
that of CH2（a）. In general, when cross-linking is occurring
during gel formation, a rapid decrease in T1 is observed21, 22）. frequency of
temperature T1
Therefore, the results suggest that T1 of CH （b） up to 160 h molecular motion
was strongly influenced by the formation of a cross-linked low ＜400 MHz long
structure. middle 400 MHz short
On the other hand, the T1 of CH2（a）was the same as high ＞400 MHz long
that of JP glycerin（at 20℃ in Fig. 10） initially after mixing
JP glycerin and GZ-DE, then was moderately prolonged
depending on time, and finally became constant （268̶276
msec） at 260̶420 h. The contribution of CH2 （a） to gel for-
1317
J. Oleo Sci. 63, (12) 1309-1322 (2014)
 K. Koga, T. Kimura, K. Sakai et al.
1
Fig. 11 C
 hanges in H-T1 of CH (b) and CH2 (a) in JP
glycerin with 2.4% w/w GZ-DE from 0.5 h to
420 h after mixing the compounds together at
20℃.
mation may be weak, as suggested by the prolonged T1 of
CH2（a）between 0.5 h to 260 h. During this time, an in-
crease in T1 of CH2（a）was observed at a frequency＜400
MHz, as shown in Table 4, indicating a change in frequency
13
Fig. 12 C-NMR spectra of JP glycerin, and two glycerin gels (containing 2.4% w/w or 12% w/w GZ-DE).
1318
J. Oleo Sci. 63, (12) 1309-1322 (2014)
in molecular motion as soon as JP glycerin and GZ-DE
were mixed. Moreover, gel formation was essentially com-
plete after 160 h, since T1 of CH（b）and CH2（a）no longer
changed.
These results indicated that the site of interaction
between GZ-DE and glycerin is likely the CH （b） area of JP
glycerin.
3.8.2 13C-NMR chemical shift
As noted above, if the gelation of glycerin by the addition
of GZ-DE is due to a chemical reaction, then a change in
the hydrogen bonds of the resulting physical gel is expect-
ed. In general, a change due to a chemical reaction and the
corresponding change in hydrogen bond strength results in
an NMR peak shift23, 24）. For example, when the CH2-OH
and CH-OH groups of glycerin are changed to ether groups,
the 13C-NMR peaks of CH2（a）and CH（b）shift significantly
to higher frequency. In addition, when the O-H bond length
is increased due to hydrogen binding, the 1H NMR peak of
the -OH trends to shift to a higher frequency. Therefore,
monitoring the 13C chemical shift of CH2（a）and CH（b）
would provide information on the formation of a chemical
gel and the 1H chemical shift would provide information on
the formation of a physical gel.
Figure 12 shows the 13C-NMR spectra of JP glycerin and
two glycerin gels（2.4％ w/w and 12％ w/w GZ-DE）. CH2
 Physicochemical and structural properties of glycerin gel
1 1
Fig. 13 The relationship between H chemical shift dif-
ference and temperature for JP glycerin.
（a） and CH（b） peaks were observed at 63 ppm and 72 ppm,
respectively. No difference was observed in the 13C spectra
of the three samples（JP glycerin, glycerin gels with 2.4％
w/w and 12％ w/w GZ-DE） . was affected by hydrogen bond formation and spread from
High frequency shifts of more than 5 ppm will be expect-
ed if the ether groups in glycerin bind another compound
such as GZ-DE following a chemical reaction. Since no
chemical shift was observed, glycerin gel containing GZ-DE
（Fig. 12）
is not a chemical gel .
3.8.3 1H-NMR chemical shift
1
H-NMR chemical shifts are frequently used to estimate
the structure of a gel 25, 26）. A change in the hydrogen
bonding of glycerin was examined using 1H-NMR spectra in
order to confirm hypothesis that glycerin gel is a physical
gel. The 1H-NMR spectra of JP glycerin are shown in Fig. 9.
Figure 13 shows the relationship between the 1H chemical
shift difference and temperature in JP glycerin. The 1H
chemical shift difference was calculated as a difference of
CH2（a）peak chemical shift by the -OH chemical shift of
CH-OH and CH2-OH peaks. Both the CH-OH peaks and
CH2-OH peaks shifted to lower frequency with an increase
in temperature, suggesting that the O-H bond length short-
ened and the strength of the hydrogen bond decreased as
the temperature increased. This indicates an increase in
the disorder of the system at higher temperature, consis-
tent with theory regarding the relationship between molec-
ular motion and temperature. Therefore, the observed 1H
chemical shift in JP glycerin strongly suggests that glycerin
gel is a physical gel.
Figure 14 shows the relationship between the 1H chemi-
cal shift difference and the time after mixing JP glycerin
and GZ-DE. The GZ-DE concentration was 2.4％ w/w.
CH-OH peak is not distinct 100 h after mixing JP glycerin
and GZ-DE, so the chemical shift difference in the CH-OH
peak was not detected during glycerin gel formation. On
the other hand, the chemical shift of the CH2-OH peak
J. Oleo Sci. 63, (12) 1309-1322 (2014)
Fig. 14 T
 he relationship between H chemical shift dif-
ference and time for JP glycerin with 2.4% w/
w GZ-DE.
could be determined up to 417 h, and shifted to higher fre-
quency with time. Therefore, the chemical shift difference
1.701 ppm to 1.765 ppm. The shifts observed in JP glycerin
from high temperature to low temperature conditions（Fig.
13） by the addition of GZ-DE indicated that glycerin gel is
a physical gel.
3.9 Structural analysis of glycerin gel by SAXS
The NMR data described above support the suggestion
that GZ-DE and glycerin form a physical gel. We attempted
to identify the basic unit which constitutes the glycerin gel
using small-angle X-ray scattering（SAXS）. The particle
size of a dispersion, the particle shape, the internal struc-
ture, and the interaction between the particles can be de-
termined using SAXS27, 28）. When gel structure has a regular
unit structure, the gel structure and shape can measured
by SAXS.
Figure 15 shows the results obtained for glycerin gel
（2.4％ w/w GZ-DE）using SAXS. The x-axis provides the
scatter vector, q, in the range 0.04-28 nm−1. The gel struc-
ture and the basic frame distance of the gel are determined
from the relationship between scatter intensity（（q） I ）vs
scatter vector（q） . As shown by the linear broken line, the
fractal order was calculated to be approximately 4, sug-
gesting that the fundamental structural unit of the gel may
be a sphere. In general, q and the diameter of a structure
（d） are given by the following equations,
q＝（4π/λ）
sin（θ）
d＝2π/q
where λ is the incident wavelength and θ is the scatter
angle. A relatively sharp peak was observed at 1.51 nm−1,
indicating that d is 4.16 nm.
1319
 K. Koga, T. Kimura, K. Sakai et al.

Fig. 15 Scatter pattern of glycerin gel with 2.4% w/w GZ-DE.

Fig. 16 Scatter curves by GIFT analysis (A), PDDF profile (B) and S (q) profile (C). PDDF: pair distance distribution
function.

Reverse Fourier transform was used to convert from
reverse space information（scatter curve）to real space in-
formation. Figure 16 shows the spectra which were ana-
lyzed using GIFT （Generalized Indirect Fourier Transform） .
It was predicted that the fundamental structural unit of
glycerin gel is over 100 nm, since no change in （q）
I was ob-
served in the Guinier region. The influence of the structure
（q）
factor S was observed since scattering on the low q side
decreased. Since GIFT analysis can explain the structure
factor S（q）, the determined curve was almost identical to
1320
J. Oleo Sci. 63, (12) 1309-1322 (2014)
the theoretical, calculated line. These results allowed the
structure of glycerin gel to be estimated from the structure
factor S（q） in detail.
From the profiles shown in Figs. 16A and B, it appeared
that the glycerin gel structure was nearly spherical, but
was slight asymmetry, suggesting that the structural unit in
glycerin gel is an ellipse or core-shell form. The data indi-
cated that the elliptical form would be 119 nm long and ap-
proximately 65 nm wide, while the core-shell form would
have a core diameter of approximately 92 nm and a shell
 Physicochemical and structural properties of glycerin gel
depth of＞26 nm. Glycerin gel probably does not have a
core-shell structure because glycerin is not a polymer and
the water content in glycerin gel is small. The results in
Fig. 16C indicate that the distance between the center of
the particles in glycerin gel is approximately 119 nm, and S
（q＝0）indicates that the interference and exclusion
volume effect between the particles is 0.3, indicating that
glycerin gel is composed of minute structural units.
4 CONCLUSION
To clarify the mechanism by which glycerin forms a gel
following the addition of GZ-DE, the physicochemical
properties of glycerin gel were investigated by rheology
assays, confirmation of reactant generation, and DSC
assays. The fluidity of the glycerin gel depended on the
concentration of added GZ-DE. Glycerin gel with 2.5％ w/
w GZ-DE was robust and did not flow even if when the vial
containing the gel was turned on its side. Since GZ-DE is
hard to dissolve in JP glycerin and glycerin, ethanol was
used as to solubilize GZ-DE. Glycerin gel（2.0％ w/w
GZ-DE） containing 7.35％ w/w ethanol did not flow. Rheol-
ogy assays showed that glycerin gel with 2.4％ or 2.5％ w/
w GZ-DE crept under 5 Pa shearing stress, but showed
only Newtonian viscosity under 10 Pa shearing stress.
Therefore, glycerin gel is fragile. No new compound was
observed in glycerin gel by capillary electrophoresis, indi-
cating that glycerin gel is a physical gel formed by inter-
molecular interactions such as hydrogen bonds, and no
chemical reaction. A remarkable difference was observed
in the glass transition temperature of JP glycerin and glyc-
erin gel by DSC measurements, indicating that glycerin
molecules interact with GZ-DE and that GZ-DE is a key
compound for network formation in the glycerin gel.
NMR assays showed a decrease in glycerin molecular
motion following the addition of GZ-DE to JP glycerin and
confirmed that molecular motion during gel formation
could be estimated from the relaxation time（T1）. After
mixing JP glycerin and GZ-DE（2.4％ w/w final concentra-
tion）, T1 of CH in glycerin remarkably decreased up to 160
h, then remained constant after 160 h, indicating that glyc-
erin and GZ-DE interactions during the gel formation af-
fected CH more strongly than CH2 of the glycerin mole-
cule. Furthermore, it was clear that glycerin gel is not a
chemical gel because no 13C chemical shift in the glycerin
molecule was observed. That glycerin gel is a physical gel
was further supported by the observation that the chemical
shift of the CH2-OH peak in glycerin shifted to higher fre-
quency due to hydrogen bonding.
SAXS measurements indicated that the basic structural
unit of glycerin gel is an oval, long diameter 119 nm long
and 65 nm wide. Glycerin gel is composed of a gel network
of these minute structural units, as suggested by the result
J. Oleo Sci. 63, (12) 1309-1322 (2014)
that S（q＝0）indicates an excluded volume effect of 0.3.
In conclusion, glycerin and JP glycerin form a physical
gel following the addition of GZ-DE in a time dependent
behavior either with or without heating. Glycerin gel is
fragile, but showed thixotropic flow under 5 Pa. GZ-DE is
the first small molecule compound which gels glycerin se-
lectively and physically by a simple operation.
Acknowledgments
We thank Mr. Takayuki Ichi of YAMADEN Co. Ltd. and
Mr. Nao Yamada, product manager of Anton Paar Japan Co.
Ltd., for help with rheology and SAXS measurements and
analysis, respectively, of glycerin gel.
References
1）Sasaki, K.; Yonebayashi, S.; Yoshida, M.; Shimizu, K.;
Aotsuka, T.; Takayama, K. Improvement in the bio-
availability of poorly absorbed glycyrrhizin via various
non-vascular administration routes in rats. Int. J.
Pharm. 265, 95-102（2003）.
2）Higaki, K.; Yata, T.; Sone, M.; Ogawara, K.-I.; Kimura, T.
Estimation of absorption enhancement by medium-
chain fatty acids in rat large intestine. Res. Commun.
Mol. Pathol. Pharmacol. 109, 231-240（2001）.
3）Shibata, N.; Ohno, T.; Shimokawa, T.; Hu, Z.; Yoshika-
wa, Y.; Koga, K.; Murakami, M.; Takada, K. Application
of pressure-controlled colon delivery capsule to oral
administration of glycyrrhizin in dogs. J. Pharm.
Pharmacol. 53, 441-447（2001）.
4）Koga, K.; Kawamura, M.; Iwase, H.; Yoshikawa N. In-
testinal absorption and biliary elimination of glycyrrhi-
zic acid diethyl ester in rats. Drug Des. Devel. Ther. 7,
1235-1243（2013）.
5）Terech, P.; Friol, S. Rheometry of an androstanol ste-
roid derivative paramagnetic organogel. Methodology
for a comparison with a fatty acid organogel. Tetrahe-
dron 63, 7366-7374（2007）.
6）Burkhardt, M.; Kinzel, S.; Gradzielski, M. Macroscopic
properties and microstructure of HSA based organo-
gels: sensitivity to polar additives. J. Colloid Interface
Sci. 331, 514-21（2009）.
7）Delbecq, F.; Tsujimoto, K.; Ogue. Y.; Endo, H.; Kawai, T.
N-stearoyl amino acid derivatives: Potent biomimetic
hydro/organogelators as templates for preparation of
gold nanoparticles. J. Colloid Interface Sci. 390,
17-24（2013）.
8）Mukai, M.; Minamikawa, H.; Aoyagi, M.; Asakawa, M.;
Shimizu, T; Kogiso, M. A hydro/organo/hybrid gelator:
A peptide lipid with turning aspartame head groups. J.
Colloid Interface Sci. 395, 154-160（2013）.
1321
 K. Koga, T. Kimura, K. Sakai et al.

9）Hilliou, L.; Sereno, A. M.; Goncalves, M. P. Gel setting Polymer Sci. 36, 1184-1253（2011）.
of hybrid carrageenan solutions under steady shear. 20）Ihara, T.; Saito, T.; Sugimoto, N. Expansion of organic
Food Hydrocolloids 35, 531-538 （2014） . reference materials for the analysis of hazardous sub-
10）Wang, Z.; Yang, K.; Brenner, T.; Kikuzaki, H.; Nishinari, stances in food and the envirnment. Synthesiol. 2,
K. The influence of agar gel texture on sucrose re- 12-22（2009）.
lease. Food Hydrocolloids 36, 196-203 （2014） . 21）Studebaker, M. L.; Beatty, J. R. Science and technolo-
11）Tyapkova, O.; Bader-Mittermaier, S.; Schweiggert- gy of rubber, edited by Eirich, F. R. Academic Press,
Weisz, U.; Wurzinger, S.; Beauchamp, J.; Buettner, A. 367（1978）.
Characterisation of flavour texture interactions in sug- 22）Kron, T.; Jonas, D.; Pope, J.M. Fast T1 imaging of dual
ar-free and sugar-containing pectin gels. Food Res. gel samples for diffusion measurements in NMR do-
Int. 55, 336-346（2014） . simetry gels. Magn. Reson. Imaging. 15, 211-221
12）Gentile, G.; Greco, F.; Larobina, D. Stress-relaxation （1997）.
behavior of a physical gel: Evidence of co-occurrence 23）Weingarth, M.; Masuda, Y.; Takegoshi, K.; Bodenhau-
of structural relaxation and water diffusion in ionic al- sen, G.; Tekely, P. Sensitive 13C-13C correlation spectra
ginate gels. Eur. Polymer J. 49, 3929-3936 （2013） . of amyloid fibrils at very high spinning frequencies and
13）Glycerine Producers’Association, Physical properties magnetic fields. J. Biomol. NMR 50, 129-136（2011）.
of glycerine and its solutions. Glycerine Producers’ 24）Casieri, C.; Testa, C.; Carpinelli, G.; Canese, R.; Podo,
Association, New York, pp.1-27（1963） . F.; De Luca, F. Proton MRI of 13C distribution by J and
14）Lane, L. B. Freezing points of glycerol and its aqueous chemical shift editing. J. Magn. Reson. 153, 117-123
solutions. Ind. Eng. Chem. 17, 924 （1925） . （2001）.
15）Mobius, M. E.; Xia, T.; van Saarloos, W.; Orrit, M.; van 25）Tang, J.-K.; Sun, Z.-Y.; An, L.-J. Physical gelation and
Hecke, M. Aging and solidification of supercooled glyc- structures of P 123 in D2O solution. Gaodeng Xuexiao
erol. J. Phys. Chem. B 114, 7439-7444 （2010） . Huaxue Xuebao 31, 172-176（2010）.
16）Gibson, G. E.; Giauque, W. F. The third law of thermo- 26）Oestlund, A.; Bernin, D.; Nordstierna, L.; Nyden, M.
dynamics. evidence from the specific heats of glycerol Chemical shift imaging NMR to track gel formation. J.
that the entropy of a glass exceeds that of a crystal at Colloid Interface Sci. 344, 238-240（2010）.
the absolute zero. J. Am. Chem. Soc. 45, 93-104 27）Khonina, T. G.; Safronov, A. P.; Shadrina, E. V.; Ivanen-
（1923） . ko, M. V.; Suvorova, A. I.; Chupakhin, O. N. Mechanism
17）Kauzmann, W. The nature of the glass state and the of structural networking in hydrogels based on silicon
behavior of liquids at low temperatures. Chem. Rev. and titanium glycerolates. J. Colloid Interface Sci.
43, 219-256 （1948） . 365, 81-89（2012）.
18）Claudy, P.; Commercon, J. C.; Letoffe, J. M. Quasi-stat- 28）Himmelein, S.; Lewe, V.; Stuart, M. C. A.; Ravoo, B. J.
ic study of the glass transition of glycerol by DSC. A carbohydrate-based hydrogel containing vesicles as
Thermochim. Acta 128, 251-260 （1988） . responsive non-covalent cross-linkers. Chem. Sci. 5,
19）Shapiro Y. E. Structure and dynamics of hydrogels and 1054-1058（2014）.
organogels: An NMR spectroscopy approach. Progress

1322
J. Oleo Sci. 63, (12) 1309-1322 (2014)