The defect chemistry of donor-doped BaTiO3:: A rebuttal

It has recently been asserted that the donor charge in La+3-doped BaTiO3 is always compensated by Ti vacancies, and that electrons are never the primary compensating defect. It was also stated that the conductivity observed in reduced, donor-doped BaTiO3 results from the loss of a very small amount of oxygen not directly related to the donor content. However, the observed reproducible and reversable weight loss on reduction, or gain on oxidation, is exactly that expected for a change between ionic and electronic compensation. It corresponds to the loss or gain of the "excess" oxygen contained in the donor oxide, e.g. LaO1.5 vs. the BaO it replaces. The amount of this weight change is proportional to the donor concentration. This is in agreement with the observation that the equilibrium conductivity in the P(O-2)-independent region of electronic compensation is proportional to the donor concentration. Thus the conductivity observed in reduced samples is directly coupled to the donor concentration, and the carrier concentration is equal to the net donor content. In fact, the equilibrium conductivity of donor-doped BaTiO3 conforms to the behavior expected from classical defect chemistry, and exhibits regions of both ionic and electronic compensation of the donor charge, as expected. Phase studies by TEM have shown that donor-doped BaTiO3 sintered in air self-adjusts its composition, by splitting out a second phase if necessary, so that the appropriate amount of compensating Ti vacancies are formed. However, when sintered in a reducing atmosphere, the composition self-adjusts to accommodate charge compensation by electrons.