Essential Chemistry and Biochemistry Associated with Low Temperature and Freezing - Freeze-drying of Pharmaceuticals and Biopharmaceuticals - page 67

Table 5 A selection of salts that form stable eutectics (E), unstable glasses
(UG) or that remain amorphous (SG) during freezing

T g (1C) Salt T e (1C) Class

Na 2 HPO 4 .12H 2 O 1 E

Na 2 SO 4 2 E

Na 2 CO 3 3 E

KCl 11 E

Urea 13 E

(NH 4 ) 2 SO 4 19 E

Glutamic acid 0.1 33 UG

NaCl -23 ?? UG

K 2 HPO 4 .6H 2 O -14 -81 UG

Tris-HCl -13 -86 UG

Guanidine-HCl -24 -103 UG

Na citrate -42 SG

NaH 2 PO 4 -65 SG

K citrate -100 SG

component is prone to crystallisation from the anhydrous mixture in
real time. 78 It is interesting to note that the glass transition/composition
profile of sucrose does not resemble that displayed by an equimolar
mixture of glucose and fructose. For chemically dissimilar components,
and temperatures above T g , the molecular mobilities of the various
components in real time increase rapidly with a rise in temperature.
Phase separation then becomes a distinct possibility. Figure 9 shows
such behaviour for a freeze-dried solution, originally containing water,
sucrose and NaCl. After the almost complete removal of water, the
system became a solid solution of NaCl in sucrose solvent and had a
very low residual water content. The figure shows the state of the
solution after it had been heated to slightly above its T g and kept at
that temperature for several minutes, after which it was quenched back
to its original low temperature. 63 Phase separation, in the form of the
growth of NaCl (or NaCl.2H 2 O) as well-formed crystals of submicro-
metre dimensions, embedded in the glassy solvent (sucrose), is clearly
indicated. Depending on the temperature, the rate of crystal growth can
be closely controlled.
As regards structural properties of PHC glasses that find applications
in the formulation of drugs destined for freeze-drying, hard information
is limited. By ''hard information'' we understand diffraction data from
which atomic distances and molecular orientations in space can be
calculated. While there is an abundance of X-ray diffraction data for
crystalline PHCs, such techniques do not ''see'' hydrogen atom posi-
tions. No direct information about hydrogen bonds can therefore be
obtained, except by speculation, based on observed O-O distances.