Patterson Air Force Base, Ohio 45433, and University of Illinois at The Zr0,-Hf02 system was investigated by metallographic, X-ray diffraction, and microprobe analysis using arc-melted samples which had been annealed and quenched. T h e monoclinic-tetragonal inversion was investigated by DTA and high-temperature X-ray diffraction analysis. Results reveal that zirconia and hafnia a r e completely miscible in all proportions in the binary system. Within experimental error, the melting points of t h e compositions lie roughly on a straight line connecting t h e melting points of oxides. There is a continuous increase in the monoclinic-tetragonal inversion temperature as the composition moves from ZrOs to Hf02. T h e lattice parameters of the monoclinic phase follow smooth curves which have a negative deviation from Vegard's law.

Electrical Properties and Defect Structure of Zirconia, I 635 reasonable. If a diffusion couple is formed between Ti and ZrOy, one would expect to find a two-phase region with compositions varying between 100% liquid of composition 1 a t the end facing the Ti source, to 100% solid of composition 4 a t the other end. At any point between, the compositions of liquid and solid and the percentages of each are given by the appropriate tie line and the lever rule. Across the two-phase region, the liquid would vary in composition from that of point 1 to point 3. This would correspond to the black, well-sintered region in Fig. 1. In this work, however, titanium is in contact with vacuum-reduced zirconia, i.e. ZrOz-r. Reoxidation ofThe defect structure of monoclinic zirconia was investigated by measuring the oxygen partial pressure dependence of the electrical conductivity and sample weight. Techniques were developed to calculate the degree of nonstoichiometry and the mobility of the charge carriers from electrical and weight change data. ZrOz was found to be an amphoteric semiconductor at 1000°C with the transition from n-type to p-type conductivity occurring at 10-16 atm. The predominant defect in the oxygen excess region was shown to be completely ionized zirconium vacan-cies. The positive holes arising from the zirconium vacancies were found to have an extremely small mobility.

Thermogravimetric measurements were made on NiO from 800" to 1100°C over the oxygen pressure range lo-' to atm. On the basis of complementary conductivity measurements showing a P(0,)"' oxygen pressure dependence, it is proposed that the predominant defects are described by an electroneutrality condition involving doubly ionized metal vacancies, impurities, and electron holes, Z[V",,] = [F'nr] +p. It is shown that for this defect model, the weight change relative to a low oxygen pressure reference weight is a measure of the effective vacancy concentration, defined as [V"A,].ff~[V",,,] -l/[F'nr], and therefore has the same oxygen pressure dependence as the electron hole Concentration followed in the conductivity measurements. The expression [V",,],f,=0.168 P(0,)"' exp (-0.86 aO.lS/kT) is derived to express the effective vacancy concentration in NiO. The probable effective impurity content of the specimens used is calculated.

Guarded measurements of the electrical conductivity of high-purity, polycrystalline YZO3 in thermodynamic equilibrium with the gas phase were made under controlled temperature and oxygen partial pressure conditions. Data are presented as isobars from 1200O to 16OO0C, and as isotherms from oxygen partial pressures of 10-1 to 10-1' atm. The ionic contribution to the total conductivity, determined by the blocking electrode polarization technique, was less than 1% over the entire range of temperatures and oxygen partial pressures studied. Yttria is shown to be an amphoteric semiconductor with the region of predominant hole conduction shifting to higher pressures at higher temperatures. In the region of $-type conduction, the conductivity is represented by the expressionThe observed pressure dependence is attributed to the predominance of fully ionized yttrium vacancies. Yttria is shown to be a mixed conductor below 900°C. samples, fabricated by the methods described in the following, were prepared from Lindsay Code 1118 yttrium oxide powder having a spectrographic analysis (ppm) of: Fe < 5, Si 20, Mg 1, A1 < 10, Ca 2, and rare earths < l .( A ) Sam@le Y-2: This sample was hot-pressed in a graphite die a t 3000 psi and 125OoC for 2 hr, and then sintered in air at 18OO0C for G hr.This sample was hot-pressed in a graphite die at 3000 psi and 1250°C for 2 hr, reoxidized in air at 8OO0C for 12 hr, sintered in vacuum at 1800°C for 2 hr, and finally reoxidized in air at 8OO0C for 12 hr.(C) SumpZe Y-4: This sample was cold-pressed at 20,000 psi, sintered in vacuum at 180Ooc for 2 hr, and reoxidized in air a t 800°C for 12 hr.( D ) Samples Y-9 and Y-10: These samples were coldpressed a t 20,000 psi, sintered in vacuum a t 2300°C for 2 hr, and reoxidized in air at 1300°C for 48 hr. The density was 98% of theoretical X-ray density.Electrodes were applied to the disk faces of all samples using platinum paste (Engelhard No. 6082). One face was completely coated; a center electrode and guard ring were painted on the opposite face. The samples were then heated in air to 800°C to remove the organic binder in the paste.( B ) Sample Y-3: