This research has illustrated the complete thermodynamic
characterization of sorbed water for crystalline solids as well as
the amorphous polymer PVP. Although models such as the GAB
equation can fit water sorption profiles of crystalline solids well,
the average energy of H
2
O–solid interactions obtained from the
analysis obscures the detailed distribution of H
2
O–solid ener-
gies as demonstrated in the partial molar enthalpy of sorbed
water reported in this article. The results of this research have
demonstrated that the partial molar enthalpy of sorbed water
can range from positive to negative values relative to the molar
enthalpy of liquid water, depending on the solid and presum-
ably the surface chemistry of the crystals.
Crystalline solids were found to exhibit distinctive H
2
O–
solid interactions at the lowest levels of water uptake near
X
w
∼
0.002 (RH
∼
10%–40%). Water’s enthalpy was determined to
be endothermic (relative to liquid water) for some crystalline
solids (genistein, indomethacin, and lactose monohydrate) and
exothermic for others (griseofulvin). It was revealed that above
this lower RH regime, H
2
O–H
2
O interactions become apparent
as enthalpy and entropy values approach those of pure liquid
water. It was also demonstrated that the partial molar entropy
of water on crystalline solids (except channel hydrates) is posi-
tive at low sorption level, but can become negative with respect
to bulk water at higher RH because of structuring effects at
the water–solid interface and minimal available sites. The ery-
thromycin channel hydrate case is distinctive in that water’s
partial molar entropy is negative even at lowest water sorption
level and RH, presumably because of the strong structuring
effect in channels.
Inherent heterogeneity of the crystalline solid surfaces ex-
amined in this research, if present, appears to be at low surface
coverage of less than approximately 20%. The RH associated
with this surface coverage depends on the material’s surface
area and water content, and for the compounds examined in
this research is estimated to be approximately 10%–40%. It is
important to consider this finding when probing surface het-
erogeneity at ultralow levels by surface analysis techniques.
Although heterogeneity may be detected under very dry condi-
tions, the material at the normal RH range of 20%–75% may
have sufficient moisture to have completely covered higher en-
ergy sites. Thus, detected heterogeneity may not easily corre-
latewithprocessingorperformancechallengesforthematerial.
For the amorphous polymer PVP K29-32, water’s enthalpy
is significantly exothermic relative to liquid water with a rela-
tively constant value up to the RH at which the glass transition
temperature is observed. Also, PVP’s enthalpy is unchanged
from the pure polymer value throughout the glassy phase. The
near constancy of both water and PVP’s enthalpy is consis-
tent with the microstructural model of glasses in which water
penetrates into interfacial space around structurally coherent
domains. Also consistent with the microstructural model is wa-
ter and PVP’s partial molar entropy trends with increasing
RH, which indicate significant interaction of water with itself,
a result that would be unusual if the glass were a homoge-
neous solid with sorbed water molecularly mixing with polymer
chains.
DOI 10.1002/jps.23806 Sacchetti, JOURNAL OF PHARMACEUTICAL SCIENCES 103:2772–2783, 2014
2782
RESEARCH ARTICLE –
Pharmaceutics, Drug Delivery and Pharmaceutical Technology
Overall, the findings of this research are certainly not intu-
itive. There is a wealth of detail in the thermodynamic inter-
actions of sorbed water in pharmaceutical solids. It is expected
that an increased understanding of the partial molar enthalpy
and entropy of sorbed water may be important in the char-
acterization of crystalline and amorphous solids, especially in
understanding batch-to-batch differences in physicochemical
property factors that influence processing and performance of
capsules and tablets, as well as changes in performance on sta-
bili