Synthetic Oxides Versus Minerals

Although there are many features common to synthetic oxides and minerals, fundamental studies of the charge-transfer processes in mixed-valence compounds can only be systematically carried out on synthetic oxides of controlled stoichiometry and impurity concentration. However, with the exception of Seebeck coefficients, transport measurements require single-crystal data if quantitative interpretations are to be made. Nevertheless, conductivity data for polycrystalline samples of cubic phases are useful if the samples are dense and care has been taken to eliminate any segregation of impurities into the grain boundaries. 

Although a particular mineral property may be due to iron impurities, it is generally easier to show this dependance by preparing synthetic samples of variable iron concentration and extrapolating back to the impurity levels of interest in a particular mineral. For example, the iron impurities in blue beryl are at so low a concentration that only weak Mossbauer signals can be obtained. 

Small monovalent cations such as Li or Cu may be mobile in iron oxides at relatively low temperatures. This fact can provide an opportunity to prepare unusual compounds at low temperatures by either chemical or electrochemical insertion or extraction of lithium or copper. However, care must be taken in high-temperature preparations to prevent, for example, loss of Li as Li20 at high temperatures or disproportionation at low temperatures - especially at grain boundaries - into lithium-rich and lithium-poor phases. 

In addition to the usual chemical characterization, the properties of iron oxides vary sensitively with oxidation state, and it is commonly necessary to control the partial pressure of oxygen during preparation and during measurement at elevated temperatures. 

---