Vs. Urusov et Nn. Eremin, CHARGE-TRANSFER ENERGY IN COMPUTER MODELING OF STRUCTURE AND PROPERTIES OF MINERALS, Physics and chemistry of minerals, 24(5), 1997, pp. 374-383
As is known, a variable parameter for partial ionicity is often involv
ed in the techniques of modeling structure and properties of inorganic
crystals by energy-minimization methods. In such a case, to make a ca
lculated energy comparable to experimental estimate (atomization energ
y) there is a need to involve an intra-atomic energy term the so-calle
d charge-transfer energy (Urusov and Eremin 1995). For this purpose a
set of charge-transfer energies for valence states of many elements is
compiled. The procedure based on aspiration of equalizing the electro
nic levels electronegativities is proposed to estimate the total contr
ibution of the charge transfer energy in the cohesive energy of a crys
tal. The other prescribed quantities are repulsive potential parameter
s which are related to the energetic propertis of valence shells of bo
nded atoms. The final potential function consists of an effective Coul
omb attraction. repulsion and covalent Morse type term. properly weigh
ted. The energy-minimization procedure is performed for various values
of the ionicity degree parameter of the crystal under study. Thereaft
er the charge-transfer energy correction is introduced in order for th
e cohesive energy calculations to be made. The calculations of structu
re, energy and elastic properties were performed using the METAPOCS co
de with ionicity variation for halides (NaCl, CaF2), oxides (MgO, Al2O
3, SiO2 quartz and SnO2), sulphide ZnS and silicate CaSnSiO5. The resu
lting cohesive energies usually agree with experiment much better than
those calculated in pure ionic or covalent approximations. As to elas
tic properties of crystals under consideration, they are generally fun
ctions of the ionicity parameter too and in most instances the predict
ed properties at intermediate values of the ionicity degree are in a s
atisfactory accordance with experimental data.