Density of Compound Oxides

A method of predicting the density of compound oxides was given on the basis of the density of the single oxides which can be regarded to compose the compound oxide. A compound oxide is first expressed in the form of a combination of single oxides, such as 2MgO⋅SiO2 for forsterite (Mg2SiO4), being applicable to any compound oxide. The density, Da, of the mixture of single oxides whose composition is identical to that of the compound oxide, 2MgO and SiO2, by way of example, is then calculated from the molar weight (W) and the density (d) of the component single oxides; Da=(2WMgO+WSiO2)/{2(WMgO/dMgO)+(WSiO2/dSiO2)}. Da is approximated to be the density of the compound, Dc. The deviation, (Dc-Da)/Da(=δ), was calculated for 88 popular silicates and non-silicate oxides. δ was found to be within ±0.1 for about 80% of the oxides examined. Furthermore, δ was found to be positively large for oxides with a higher density and negatively large for oxides with a lower density. The silicates having a negatively large δ have a tendency to exhibit a negative thermal expansion, and those with a positively large δ tend to have high hardness. The relationship of δ to Dc leads to the following approximations; Dc=Da {1+0.33(Da-3.1)} for silicates, and Dc=Da {1+0.11(Da-5.3)} for non-silicate oxides. Both of these empirical expressions hold within an error of less than 10% for most oxides, though a more detailed examination with much more data will be necessary. The assumed density of 2.0 for Si (F, OH)4 and Si(OH)4 results in a good approximation for the densities of (F, OH)- and (OH)-containing silicates.