182
S. Cerveny and J. Swenson
crystallization time, the water relaxation becomes progressively broader on the low
frequency side and for longer times, there is a slow extra peak in the spectra (f ~
65 Hz). The dielectric strength corresponding to the water relaxation decreases with
increasing crystallization time since part of the water turns into ice and therefore a
fraction of the water dipoles do not contribute anymore to the dielectric response of
water. On the other hand, the number of “ice” dipoles during crystallization increases
and therefore the dielectric strength corresponding to the slow process also increases.
The presence of this slow process does not depend on the crystallization temperature
and it is independent of the solute, i.e., this extra process is also observed during the
isothermal crystallization of PVP- or PVME–water solutions and it corresponds to
the dielectric response of ice, as we will discuss latter.
In Fig. 8, we can observe two different behaviors depending on the crystallization
temperature. For the sample in Fig. 8a, there is no change of the glass transition
temperature between the amorphous and the semi-crystalline solution whereas in the
second case (Fig. 8b), a change of 2.4° is produced in the T g value (see T g values in
Table 1). In addition, in Fig. 8, a the peak maximum of water is not changing whereas
in Fig. 8b, the relaxation time becomes slower after each cycle of crystallization.
The isothermal data during crystallization were fitted using the sum of three
functions, two of them for describing the relaxations occurring in the amorphous
fraction of the samples (a HN function for the α-relaxation and a CC function for
the water relaxation). The third CC function was used to fit the new process at low
frequencies (“ice relaxation” in Fig. 8b). When there is no change in the T g value
before and after crystallization, the relaxation times of the HN and CC functions of
the α-relaxation and water relaxation, respectively, were fixed using the values of
the amorphous sample. For these relaxations, ε were free during the fitting as well
as for the “ice relaxation” for which all the parameters were free. By contrast, for
samples in which T g changes are produced by the crystallization, all the parameters
were free during the fitting. An example of the fitting can be seen in Fig. 9 for the
amorphous (t cryst = 0) and semi-crystalline sample (t cryst = 8 h). Figure 8c and d show
the time evolution of the relaxation strength of the ice and water relaxations for 3PG
samples containing 50 and 40 wt% water, respectively. The lower the crystallization
temperature (Fig. 8c) the longer it takes for water to crystallize (slow kinetics),
and after 480 min, ε water of water has dropped to ~44%. At a higher crystallization
temperature (Fig. 8d), the water crystallizes much faster (ε water of water has dropped
~52% after 25 min), and thereafter there is no more water crystallization.
Another example during isothermal crystallization is shown in Fig. 10 for PVME
(c w = 50 wt%) and PVP (c w = 55 wt%). Both polymers were crystallized at high
temperatures and therefore the crystallization proceed fast. In both cases, the water
peak shown in the figure corresponds to the water relaxation for PVME and the
fast water relaxation for PVP. As in the previous case, the intensity decreases and a
shoulder is developed at lower frequencies. For PVME, the maximum of the water
peak becomes slower after each cycle of crystallization (as in the case of 3PG)
until ~1 decade is reached after 480 min. By contrast, for PVP, the water peak also
becomes slower, but only ~0.4 decades in spite of the fact that the crystallization
time is longer than for PVME and the glass transition value decreases more for PVP
S. Cerveny and J. Swenson
crystallization time, the water relaxation becomes progressively broader on the low
frequency side and for longer times, there is a slow extra peak in the spectra (f ~
65 Hz). The dielectric strength corresponding to the water relaxation decreases with
increasing crystallization time since part of the water turns into ice and therefore a
fraction of the water dipoles do not contribute anymore to the dielectric response of
water. On the other hand, the number of “ice” dipoles during crystallization increases
and therefore the dielectric strength corresponding to the slow process also increases.
The presence of this slow process does not depend on the crystallization temperature
and it is independent of the solute, i.e., this extra process is also observed during the
isothermal crystallization of PVP- or PVME–water solutions and it corresponds to
the dielectric response of ice, as we will discuss latter.
In Fig. 8, we can observe two different behaviors depending on the crystallization
temperature. For the sample in Fig. 8a, there is no change of the glass transition
temperature between the amorphous and the semi-crystalline solution whereas in the
second case (Fig. 8b), a change of 2.4° is produced in the T g value (see T g values in
Table 1). In addition, in Fig. 8, a the peak maximum of water is not changing whereas
in Fig. 8b, the relaxation time becomes slower after each cycle of crystallization.
The isothermal data during crystallization were fitted using the sum of three
functions, two of them for describing the relaxations occurring in the amorphous
fraction of the samples (a HN function for the α-relaxation and a CC function for
the water relaxation). The third CC function was used to fit the new process at low
frequencies (“ice relaxation” in Fig. 8b). When there is no change in the T g value
before and after crystallization, the relaxation times of the HN and CC functions of
the α-relaxation and water relaxation, respectively, were fixed using the values of
the amorphous sample. For these relaxations, ε were free during the fitting as well
as for the “ice relaxation” for which all the parameters were free. By contrast, for
samples in which T g changes are produced by the crystallization, all the parameters
were free during the fitting. An example of the fitting can be seen in Fig. 9 for the
amorphous (t cryst = 0) and semi-crystalline sample (t cryst = 8 h). Figure 8c and d show
the time evolution of the relaxation strength of the ice and water relaxations for 3PG
samples containing 50 and 40 wt% water, respectively. The lower the crystallization
temperature (Fig. 8c) the longer it takes for water to crystallize (slow kinetics),
and after 480 min, ε water of water has dropped to ~44%. At a higher crystallization
temperature (Fig. 8d), the water crystallizes much faster (ε water of water has dropped
~52% after 25 min), and thereafter there is no more water crystallization.
Another example during isothermal crystallization is shown in Fig. 10 for PVME
(c w = 50 wt%) and PVP (c w = 55 wt%). Both polymers were crystallized at high
temperatures and therefore the crystallization proceed fast. In both cases, the water
peak shown in the figure corresponds to the water relaxation for PVME and the
fast water relaxation for PVP. As in the previous case, the intensity decreases and a
shoulder is developed at lower frequencies. For PVME, the maximum of the water
peak becomes slower after each cycle of crystallization (as in the case of 3PG)
until ~1 decade is reached after 480 min. By contrast, for PVP, the water peak also
becomes slower, but only ~0.4 decades in spite of the fact that the crystallization
time is longer than for PVME and the glass transition value decreases more for PVP