Structure, bonding, and stereochemistry

The fact that almost all minerals are crystalline solids, indeed, were among the first crystalline solids to be systematically studied by x-ray 

Covalency and ionicity in solids from calculation and experiment 

There has been considerable effort directed towards the assessment of the covalency or ionicity of various solids. The output of standard ab initio (SCF) Hartree-Fock-Roothaan calculations contains Mulliken (1955) charge distribution analysis parameters such as the atomic-orbital populations, net atomic charges, and bond overlap populations deseribed earlier (Chapter 3), which are often used to discuss the relative covalency or ionicity of materials. Considerable caution is required in using such parameters, however, since net atomic charges and other such quantities are not quantum-mechanical observables that is, they cannot even in principle be measured, and are highly basis-set dependent, as noted by Hehre et al. (1986 pp. 336-41). This is illustrated for molecules more relevant to mineralogy in Table 7.1, in which a number of properties of CO2 and SiOj are shown calculated at various basis-set levels. It is clear 

Can an absolute description of covalency versus ionicity be obtained from observable properties First, bond type is generally defined in principle in terms of the total electron density, since the molecular energy is a unique functional of the electron density (Hohenberg and Kohn, 1964). From x-ray diffraction, total densities are available, but the problem is to determine which aspects of the data are most informative. We can inspect the topology of the total density in the way described by Bader and Ngu- 

Because of the continuous variation of the electron distribution in molecules or crystals, the counting of electron charge is hopelessly dependent on the nature of the distribution function over which the integration is taken. Clearly, the charge of an atom in a molecule or crystal is not an observable it is arbitrary. 

---