W. Richard Bowen and Nidal Hilal 4

178 6. NANOSCALE ANALySIS Of PHARMACEUTICALS by SCANNINg PRObE MICROSCOPy
These and other studies can broadly be divided into two types: those
that rank relative particulate interactions and those that attempt to make
a quantitative comparison of force per unit area of contact. Ranking studies
address, e.g., how drug-drug cohesion compares to drug-excipient
particle and drug-device adhesion. Since particulate interactions are dominated
by aspects such as surface morphology, surface roughness, exposed
chemical moieties and thermodynamic properties, all of which can vary
from particle to particle and indeed within a single particle, such ranking
comparisons are normally made using the same particle to challenge
all the possible combinations of interactions. In this case, ranking should
be consistent for each drug particle probe, but the absolute values of adhesion
force determined cannot be used to make comparisons from particle
to particle or between materials [18].
To quantify such measurements, particle variability and a lack of direct
knowledge of the contacting regions must be overcome to allow the
determination of factors such as surface energy and work of adhesion.
The use of AFM to determine such properties on model flat surfaces, such
as silicon, was established relatively early [25]. However, its application
to pharmaceuticals due to difficulties of rough and variable particle morphology
came later [19, 21, 22]. In these works, the principal variable that
is allowed for is contact area. Contact areas have, to date, been estimated
either via a direct imaging approach [22] or indirectly through imaging
indents made by the particle probe in a plastic polymer film [19]. The
subsequent use of adhesion models such as Johnson–Kendall–Roberts
and Derjaguin–Muller–Toporov can then allow the surface energy of the
particle (over the region of contact) to be determined. In this way, e.g.,
micronised (milled) salbutamol sulphate and a version prepared via a
novel supercritical fluid method have been compared [22].
An alternative to this approach of trying to model the variable contact
region between particles is to compare ratios of cohesive and adhesive
forces between different particles rather than actual separation forces.
This has allowed an assessment for relatively flat crystals of the affinity
of salbutamol sulphate to lactose and budesonide to lactose, showing
that salbutamol sulphate has a stronger affinity for lactose [26]. This
information was then related successfully to the likely blend uniformity
these materials would form.
In addition to monitoring force normal to a surface, AFM is also capable
of assessing frictional forces between a probe and a surface. This is
achieved by recording the twist of the cantilever in addition to its vertical
bend as it scans a surface in continuous contact with that surface
(Figure 6.3a). This approach has recently been used to obtain singleparticle
friction measurements on DPI formulations [27] and blister
packaging material (used in DPIs) [28] and provides an opportunity to
consider sliding as well as separation forces. Figure 6.3 shows examples