Crystal structure and substitution

Sites, not slots

A crystal structure offers cation sites of particular sizes. An ion whose charge and radius fit a site can occupy it; one that fits badly cannot. This is the whole basis of mineral chemistry, and later of geochemistry.^[1]

Solid solution

When two ions fit the same site well enough, they substitute for one another freely and the composition varies continuously between end members. Olivine runs from forsterite to fayalite by exchanging iron for magnesium.^[1]

Sometimes the exchange has to be coupled. Plagioclase runs from albite to anorthite by swapping sodium and silicon for calcium and aluminium together, because replacing Na+ with Ca2+ alone would leave the structure with the wrong charge.^[1]

The same pattern recurs: enstatite to ferrosilite, annite to phlogopite, almandine to pyrope. Substitution is the normal state of a natural mineral, not an impurity in it.^[1,2,3,4]

Polymorphism

The converse also happens: one composition, several structures. Microcline, orthoclase and sanidine all have the formula K(AlSi3O8) and differ only in how ordered their aluminium and silicon are. Sanidine’s disorder records rapid cooling.^[1]

Because different structures are stable under different conditions, a polymorph is a record of the pressure and temperature at which a rock formed.^[1]

The silicate hierarchy

Silicates are organised by how their SiO4 tetrahedra link. Isolated tetrahedra give olivine and garnet; single chains give pyroxene; double chains give amphibole; sheets give mica and clay; frameworks give quartz and feldspar.^[1]

The progression is legible in the formulas. As tetrahedra share more of their oxygens, the ratio of silicon to oxygen rises from 1:4 to 1:2 — and the physical properties follow, from the equant grains of olivine to the perfect cleavage of mica.^[1,2]