78
J. Knapik-Kowalczuk et al.
a series of samples with various concentrations, one is able to experimentally determine the concentration dependence of the glass transition temperatures (presented
on the example of FLU-KVA systems in Fig. 16b). This dependence will serve as
the basis for the identification of a saturated solution (solubility limit) instead of the
G-T prediction.
Consecutive part strictly corresponds to the second step of the Mahieu’s
approach—annealing of the sample at elevated temperature in order to re-crystallize
the excess amount of the drug from the supersaturated solution.
4.1 Isothermal Dielectric Measurements
In the following subsection of this chapter, step by step instruction, on how to determine the drug’s solubility within the polymer matrix utilizing Broadband Dielectric
Spectroscopy, will be provided. For this purpose, FLU-KVA composition was chosen
as the representative example.
Taking into account that demixing process (re-crystallization of the excess amount
of the drug from the supersaturated solution) is the crucial step in discussed solubility determination, annealing should be performed well above the glass transition temperature—in the supercooled liquid region (preferably higher or close to
the temperature onset of crystallization). Therefore, during the dielectric measurements the structural relaxation peak will be well visible in the experimental window.
Furthermore, as was already stated in this book, when the re-crystallization of the
neat amorphous component starts, the intensity of the α-process peak will decrease.
However, when observed re-crystallization refers to the amorphous drug from the
drug-polymer ASD system, then in addition to mentioned decrease, one can expect
the shift of the structural relaxation process towards lower (in case of low-T g drug
and high-T g polymer system) [96] or higher frequencies (in case of high-T g drug
and low-T g polymer). The representative example of discussed phenomenon can
be observed in Fig. 17a, in case of the re-crystallization from the FLU + 13 wt.%
KVA system. Observed shift (towards lower frequencies) is caused by the weakening
of the plasticization effect (i.e. an increase in the global mobility associated with a
decrease in T g of the binary system compared with the initial concentration) due to the
decreasing amount of amorphous drug in the mixture. When the re-crystallization of
the excess amount of the small molecule (herein FLU) from the supersaturated solution stops—leaving saturated (partially amorphous) solution, one can still observe,
on the dielectric spectrum one well-resolved loss peak—α
-process (see Fig. 17a).
When no further changes in molecular dynamics, such as drop in intensity or horizontal shifts, can be observed, the sample should be cooled down and re-measured
during heating up to the temperature of the performed isothermal crystallization, as
can be seen in Fig. 17b (in discussed example T = 353 K). Described above additional
measurement is necessary to cover wide temperature range of the relaxation time
related to the α
-process. Next, based on the HN fitting procedure performed on the
dielectric loss spectra obtained during mentioned procedure, one can determine the
J. Knapik-Kowalczuk et al.
a series of samples with various concentrations, one is able to experimentally determine the concentration dependence of the glass transition temperatures (presented
on the example of FLU-KVA systems in Fig. 16b). This dependence will serve as
the basis for the identification of a saturated solution (solubility limit) instead of the
G-T prediction.
Consecutive part strictly corresponds to the second step of the Mahieu’s
approach—annealing of the sample at elevated temperature in order to re-crystallize
the excess amount of the drug from the supersaturated solution.
4.1 Isothermal Dielectric Measurements
In the following subsection of this chapter, step by step instruction, on how to determine the drug’s solubility within the polymer matrix utilizing Broadband Dielectric
Spectroscopy, will be provided. For this purpose, FLU-KVA composition was chosen
as the representative example.
Taking into account that demixing process (re-crystallization of the excess amount
of the drug from the supersaturated solution) is the crucial step in discussed solubility determination, annealing should be performed well above the glass transition temperature—in the supercooled liquid region (preferably higher or close to
the temperature onset of crystallization). Therefore, during the dielectric measurements the structural relaxation peak will be well visible in the experimental window.
Furthermore, as was already stated in this book, when the re-crystallization of the
neat amorphous component starts, the intensity of the α-process peak will decrease.
However, when observed re-crystallization refers to the amorphous drug from the
drug-polymer ASD system, then in addition to mentioned decrease, one can expect
the shift of the structural relaxation process towards lower (in case of low-T g drug
and high-T g polymer system) [96] or higher frequencies (in case of high-T g drug
and low-T g polymer). The representative example of discussed phenomenon can
be observed in Fig. 17a, in case of the re-crystallization from the FLU + 13 wt.%
KVA system. Observed shift (towards lower frequencies) is caused by the weakening
of the plasticization effect (i.e. an increase in the global mobility associated with a
decrease in T g of the binary system compared with the initial concentration) due to the
decreasing amount of amorphous drug in the mixture. When the re-crystallization of
the excess amount of the small molecule (herein FLU) from the supersaturated solution stops—leaving saturated (partially amorphous) solution, one can still observe,
on the dielectric spectrum one well-resolved loss peak—α
-process (see Fig. 17a).
When no further changes in molecular dynamics, such as drop in intensity or horizontal shifts, can be observed, the sample should be cooled down and re-measured
during heating up to the temperature of the performed isothermal crystallization, as
can be seen in Fig. 17b (in discussed example T = 353 K). Described above additional
measurement is necessary to cover wide temperature range of the relaxation time
related to the α
-process. Next, based on the HN fitting procedure performed on the
dielectric loss spectra obtained during mentioned procedure, one can determine the