Chemistry Reference
In-Depth Information
different from the monoexponential relaxation pro
les of the individual amorphous
materials, whereas a phase-separated system will be biexponential. Using nifedipine as
the model drug, Aso et al. found that the
T
1
and
T
1
ρ
relaxation times were mono-
exponential for HMPC and PVP dispersions [105]. In contrast, dispersions with
-poly
(
N
-5-hydroxypentyl)-
L
-aspartamide were monoexponential for the
T
1
relaxation times,
but biexponential for
T
1
ρ
relaxation measurements. From these results, it was concluded
that the latter systemwas phase separated into two amorphous phases whereby the size of
the phase-separated domains was between 5 and 50 nm.
SSNMR has been used to evaluate drug-polymer hydrogen bonding interac-
tions [106]. For a dispersion containing 30% acetaminophen blended with PVP, it
was concluded that the phenol group in acetaminophen formed a hydrogen bond with the
carbonyl group of the polymer, with a hydrogen bond distance of approximately
1.75
α
,
β
0.04 Å [106]. Based on additional measurements, it was further demonstrated
that acetaminophen and PVP are mixed on a molecular level.
SSNMR can be a very sensitive technique for the detection of crystallinity in
amorphous solid dispersions. Using
19
F SSNMR, a detection limit of around 3%
crystalline material in a solid dispersion of an experimental drug with HPMCAS was
observed [110]. Using this method, it was found that incomplete crystallization of the
drug occurred from the amorphous dispersion. It was postulated that there is a
composition of drug and polymer for which crystallization does not occur because
the drug is below its solubility limit in the polymer.
±
5.5.4 Calorimetric Methods
Calorimetric methods, in particular DSC, are widely used for the characterization of
amorphous solid dispersions. Due to the nonisothermal character of DSC and hence the
potential for sample changes during heating, some care must be exercised when
interpreting results. One of the most common applications of DSC is to determine
the
T
g
or
T
g
values of the solid dispersion. It is of interest to know the
T
g
since
crystallization is known to be faster above
T
g
than below
T
g
[14]. In addition, samples
above
T
g
become sticky and may be dif
cult to handle. Therefore, for amorphous
dispersions, it is desirable to have a dispersion
T
g
that is above the temperature range of
storage and processing conditions. While it would be nice to be able to predict the
likelihood of crystallization of the dispersion from a knowledge of the
T
g
, it appears that
crystallization does not correlate well with
T
g
for systems that contain more than one
component [77,103].
Another major application of DSC is to determine if crystallization occurs during
heating, or if a melting endotherm (indicative of the presence of crystalline material
arising from processing or storage) is present. Experiments can be performed in
different ways. For example, the solid dispersion can be held at a certain temperature
using the DSC in isothermal mode to evaluate the crystallization kinetics at this
temperature. Yoshihashi et al. [115] used this approach to characterize the crystalli-
zation kinetics in an amorphous solid dispersion. They
first heated the solid disper-
sions to above the melting point of the drug to eliminate any residual crystalline
material. This heating stage was followed by rapid cooling and then the samples were
