Chemistry Reference
In-Depth Information
carboxylic acid group and the carbonyl oxygen in PVP (a hydrogen bond acceptor)
shows evidence of a short
2 Å contact consistent with some degree of hydrogen
bonding occurring in the simulation. In contrast, the distance between a
∼
fluorine atom in
the drug and the carbonyl oxygen in PVP shows no evidence of a short interaction. It is
important to note that the choice of force-
eld methods used in these calculations varies
in published studies and difference in force
field, charging schemes, cutoffs, and the use
of summation schemes such as the Ewald method can affect the results [175]. The results
of MD calculations can be combined with the analytical techniques discussed previously
to enhance the interpretation of results, particularly for techniques such as IR, Raman,
and SSNMR spectroscopy (e.g., by modeling hydrogen bonding and other interactions of
interest) and for diffraction techniques such as PXRD by assisting with models for PDF
analysis [158].
The estimation of important properties such as
T
g
can also be accomplished
computationally by MD methods [172,174]. For example, in a recent MD study of
amorphous sucrose with varying amounts of water added as a plasticizer,
T
g
values of
367, 352, and 343 K were obtained for amorphous sucrose models containing 0, 3, and
5% (w/w) water, respectively [174]. The MD calculation was in reasonable agreement
with experimental
T
g
values and also correctly predicted the in
uence of the plasticiza-
tion effect of water on
T
g
[174]. In this study, MD simulations were performed using the
COMPASS force
field and the
NPT
ensemble, and the speci
c volumes of amorphous
cells were computed in the temperature range of 265
-
440 K. The
T
g
was then de
ned
from a characteristic
temperature plots using regression
analysis. In addition to the analysis of
T
g
, PDF analysis of the MD simulation structures
identi
“
kink
”
obtained from volume
-
ed likely hydrogen bonding effects between sucrose hydroxyl oxygens and water
oxygens [174].
Simulation methods may also allow investigation of molecular mobility at slow
timescales that may be dif
cult to measure by the characterization techniques described
here. For example, it is possible to relate MD simulations of polymers to nuclear spin
relaxation measurements obtained by SSNMR, particularly when combined with
eld
cycling methods, which has great potential for obtaining detailed insight into molecular
mobility in amorphous solid dispersions [176]. The combination of MD simulations with
dielectric relaxation methods and other techniques may also provide useful interpretative
support for experimental findings.
4.12 CONCLUSIONS
Effective characterization plays a critical role in the development of pharmaceutical
amorphous solid dispersions. The complexities of the amorphous state, the potential for
heterogeneity over a wide range of spatial dimensions, and the multicomponent nature of
these materials all present unique challenges for analytical characterization. A number of
established techniques can be applied to obtain critical information about amorphous
solid dispersions. Over the past
cant progress has been
made in adapting recently developed analytical techniques for the role of characterization
of amorphous solid dispersions. This chapter has attempted to summarize both the basic
five years in particular, signi
