9
18
R. Fan et al. / Food Chemistry 190 (2016) 912–921
decreased at the moderate temperatures (140–180 °C). When the
reaction temperature was higher than 180 °C, the GA formation
ratio slightly decreased at the beginning of the hydrolytic reaction.
The GA formation ratio changed with temperature firstly increased
and then decreased in a shorter period (<50 min), whereas the GA
formation ratio was always decreased slightly over longer times
increased. When k
ature (180 °C), k sharply increased from 0.00182 to 0.00404 s
which indicated that high temperatures resulted in the hydrolysis
of GL. The value of k was considerably larger than k and k , which
implied that the direct formation of GA was predominant. This was
similar with literature, which stated that GL was metabolized to
GA as a main product and GAMG as a minor product by human
intestinal microflora (Kim et al., 2000). At lower temperatures of
1 2
and k started to decrease at a higher temper-
ꢀ1
3
,
3
1
2
(>50 min). This phenomenon was also reported in the literature
Vicente, Salagre, Cesteros, Medina, & Sueiras, 2010). It is well
(
known that the temperature was one of the principal factors in
controlling chemical reactions by directly influencing equilibrium
and reaction rate constants (Kruse & Dinjus, 2007). A high level
of GA formation ratio was obtained at higher temperature in a
shorter time, for instance, 220 °C for 10 min or at lower tempera-
ture in a longer time, for example, 120 °C for 80 min. A longer hold-
ing time at a higher reaction temperature caused some undesirable
reactions, such as carbonization, which was evident by the visible
brown color of the hydrolysate and emission of a scorched flavor, a
similar result was also reported by Wang et al. (2012).
120–180 °C, k
tion rate was faster than that of decomposition, which promoted
GAMG to accumulate. At a lower temperature, the value of k
was bigger than k and the GAMG yield was higher. When the dif-
ference between the k and k values appeared to be at the largest
value at 160 °C, the GAMG yield reached a maximum level.
However, with the rise of temperature, k was increased at a rate
faster than k , when the curves of k and k were intersected at
180 °C, k was larger than k due to GAMG decomposition.
The reaction times for attaining the maximum GAMG yield (t
1 2
was larger than k , implying that the GAMG forma-
1
2
1
2
2
1
1
2
2
1
m
)
at different temperatures and the corresponding yields are listed in
Table 1(a). It was observed that the highest yield of GAMG was
obtained by GL hydrolysis at 160 °C for 16 min; the yield was
insignificantly different from that obtained at 140 °C for 74 min.
Considering the cost, the parameters of 160 °C for 16 min were
preferred to produce GAMG. The time for producing GAMG in sub-
critical water was much shorter than that of other methods, such
as the biosynthesis in a water-miscible ionic liquid by immobiliz-
ing whole cells of P. purpurogenum Li-3, where the optimal time
was 62 h in ionic liquid co-solvent medium compared to 72 h in
buffer medium (Chen et al., 2012). It was concluded that subcritical
water technology for the preparation of GAMG and GA was the effi-
cient and green technology.
3
3
.4. The kinetic analysis of the GL hydrolytic reaction
.4.1. Reaction pathways and mechanism
The GL hydrolysis scheme is shown in Fig. 4(a), which is com-
posed of two parallel reactions. One is a consecutive irreversible
reaction via producing an intermediate product (GAMG) and then
further decomposing into GA. The other is the direct formation of
GA by cleaving of the b-1,3 glycosidic bond, removing two mole-
cules of glucuronic acid from GL. The main by-product of the
hydrolytic reaction was glucuronic acid. This pathway of reaction
was similar to the proposed metabolic pathway of GL by human
intestinal bacteria (Kim et al., 2000).
It was suggested that the hydrolysis mainly occurred by the
attacking of a proton ion dissociated from subcritical water. The
oxygen atoms in the glycosidic bond were protonized, the glyco-
sidic bond was broken and then formed the intermediates which
appeared in glycosyl positive ions form or half-chair structure,
3.5. Thermodynamic analysis
The Arrhenius equation is a common method for analyzing the
relationship between the reaction rate and temperature. Due to the
k and k being decreased at a higher temperature, the estimation
1 2
ꢀ
and then glucuronic acid was formed by bonding OH dissociated
from subcritical water, releasing the hydrion and acting as a cata-
lyst (Haghighat Khajavi, Ota, Kimura, & Adachi, 2006). Fig. 3(b)
shows the reaction mechanism, the breakage of the glycosidic
bond was carried out in three steps, listed as follows (Sasaki,
Furukawa, Minami, Adschiri, & Arai, 2002):
of the Arrhenius parameters was carried out below 180 °C. To eval-
uate the Arrhenius parameters, the logarithmic values of the rate
constants (k ) were calculated, and the results are showed in
i
Fig. 5(b). The estimated values are listed in Table 1(b).
In consecutive reactions, the activation energies (Ea) of the reac-
tions 1 and 2 were 61.112 kJ/mol and 129.521 kJ/mol, respectively.
This indicated that the energy barrier for the cleavage of the b-1,2
glucosidic bond would be lower than that of the b-1,3 glucosidic
bond. The higher activation energy in reaction 2 led a slower clearing
rate of the b-1,3 glucosidic bond compared to the b-1,2 glucosidic
bond. This phenomenon was due to stronger steric interactions
(intra- and intermolecular van der Waals repulsions), which could
(
1) The oxygen atom in the glycosidic bond was attacked by H+,
which led to its rapid protonation (procedure A).
(
2) The positive charge was transferred to C
and then a carbenium ion was formed due to the breaking of
the C–O bond, and a hydroxyl group was provided to the C
of another glucuronic acid or C of pentacyclic triterpene
procedures B and C).
1
of glucuronic acid,
2
3
+
(
slow attacks on the carbocation by H self-dissociated of SW (Ong
ꢀ
(
3) Subcritical water delivered OH to the carbenium ion, form-
et al., 2013). Compared with the hydrolysis of cellulose in SW
(Sasaki et al., 2002), the activation energy of GL hydrolysis was smal-
ler, which indicated the b-1,4- and b-1,6-glycosidic linkages in cellu-
lose exhibited a much higher stability compared to the b-1,3- and
b-1,2 glucosidic bond. The rate constants of the hydrolytic reactions
were determined for the resulting monomers, and the values were
found to strongly depend on the type of bonds (Rogalinski et al.,
2008). The pre-exponential factor (A) expresses how often the mole-
cules collide, the higher value of A meant a greater probability of a
successful collision, which might imply the occurrence of a chemical
reaction. When comparing the pre-exponential factor (A) between
reactions 1 and 2, it can be implied that cleaving the b-1,3 glucosidic
bond was more feasible than the b-1,2 glucosidic bond, however the
energy required for cleaving b-1,2 glucosidic bond was lower than
b-1,3 glucosidic bond. Therefore, once activated, reaction 2 pro-
ceeded quickly, even if the temperature was increased slightly.
+
ing glucuronic acid residues, releasing H (procedure D).
3.4.2. Kinetic analysis
The changes in GL, GAMG and GA content during the hydrolytic
reactions were measured for the kinetic analysis. The experimental
data were correlated well with the irreversible consecutive first
order theoretical models estimated using Eq. (18), the values of
2
R were 0.9606, 0.9803, 0.9718, 0.9868, 0.9951 and 0.9835, respec-
tively. It could be concluded that the GL hydrolysis was considered
to be first-order reaction. Similar results were mentioned in previ-
ous reports (Liu et al., 2008; Ruen-ngam et al., 2012; Wardhani
et al., 2008)
The rate constants (k
the results are shown in Fig. 5(a). With the rise in temperature,
and k initially increased, and then decreased, while k always
i
) were affected by the temperature, and
k
1
2
3
Fan, Rui
