Silicon is the second most abundant element in the Earth's crust and mantle, after oxygen, and since the Si–O bond is considerably stronger than that between any other element and oxygen, it is not surprising that silicate minerals make up the vast majority of rocks. 95% of the Earth's crust is composed of only a handful of mineral groups – feldspars, quartz, amphiboles, pyroxenes and micas, and in this chapter we will be concerned mainly with the structure, composition and behaviour of these groups. Within each group there can be considerable diversity in the composition and the related structural modifications, and in this context, ‘behaviour’ means the way in which these adaptations of the basic structure are dependent on temperature and pressure in the Earth. We will need to discuss the flexibility of the linkages between the various ions in an attempt to understand the way in which a mineral can respond to changes in its physical and chemical environment.

The description of silicate structures depends to some extent on the model for the Si–O bond. In a purely ionic model composed of Si4+ and O2- ions, held together by non-directional electrostatic forces, the oxygen ions would tend to be close-packed with charge balance provided by Si4+ and other cations in interstitial sites of the appropriate size. Few silicates however have a density anywhere near that of a close-packed oxygen array, nor are the cation: oxygen radius ratios often consistent with the observed coordinations.