The aim of mineralogical research is to understand the relationship between
the structure of a mineral and its physical, elastic and thermodynamic
properties, so as to develop the same structure-function relationship that
underpins the modern structural approach to biology. The principles will be
illustrated in this talk by examining recent developments in understanding
the structure-function relationships in framework minerals.

Framework minerals are the dominant structure type from the crust to the
lower mantle, and include feldspars and perovskites. Their response to
pressure and temperature controls the geochemistry of the Earth through
changes in cation partitioning and volatile uptake, and the structure and
dynamics of the Earth through changes in elasticity. Framework minerals
respond to changes in pressure, temperature and extra-framework cations by
cooperative tilting of the polyhedra that comprise their structures. The
conventional view that the polyhedra are essentially rigid under
compression leads to the specific prediction that they should become more
tilted with increasing pressure, and that phase transition boundaries will
have positive slopes dP/dT. Recent experiments on some perovskites shows
that this is not always correct. We have shown instead that the structural
response and the form of the phase diagrams of perovskites depends on the
balance of bonding forces between the framework and the extra-framework
cation. This allows us to predict the evolution of framework minerals with
pressure, temperature and composition and will ultimately lead to a
structure-based model for their thermodynamic properties.