Effects of Yttrium Doping α‐Alumina: I, Microstructure and Microchemistry

Abstract

The influence of yttrium doping on the microstructure and microchemistry of hot-pressed ␣-alumina was investigated using a combination of electron microscopy techniques. The implications of microstructure and microchemistry on the improved creep behavior of the doped material are discussed. The samples doped only with yttrium had a bimodal grain-size distribution that was strongly correlated to the frequency and distribution of Y 3 Al 5 O 12 (YAG) precipitates in the microstructure. Yttrium segregated to most of the grain boundaries, with a normal excess concentration of ⌫= 3.3 ± 0.9 atoms/nm 2 at random boundaries. Two types of twin boundaries (⌺3 and ⌺7) accommodated no yttrium. None of the boundaries or triple-point junctions contained a glassy grain-boundary phase. Strong interaction of the grain boundaries and dislocations with YAG precipitates indicated a pinning mechanism by the precipitates. Yttrium doping did not appear to favor formation of special boundaries in ␣-alumina.

Atomistic modeling of dopant segregation in α-alumina ceramics: Coverage dependent energy of segregation and nominal dopant solubility

“…The broken lines correspond to various solubility limits reported in the literature, 37,75,76 whereas the grey and black data points correspond to experimental observations 41,[65][66][67][68] indicating whether precipitation was observed or not for a given grain size and dopant concentration.…”

“…This sample contains however a significant proportion of highly special low Σ boundaries, which in real ␣-alumina microstructures constitute only a very small fraction of the grain boundary population. [38][39][40][41] An average taken only over high-Σ, high-energy boundaries, which should have more general character, is therefore likely to result in a better description of experiment. The average value of 5.29 La/nm 2 over the Σ93 (1 1 · 0), Σ13 (1 1 · 3) and Σ43 (2 2 · 3) boundaries shows that indeed these more general boundaries can accommodate a higher dopant content, as suggested by their more open structure (see Supporting Information Table S2 for coordinative environment).…”

“…7, along with various experimental literature values. 41,[65][66][67][68] The Table 4 Minimum energy dopant ion concentration (Γ Emin ) and estimated solubility (Γ eq ) for different Mg doped ␣-alumina surfaces and mirror twin grain boundaries. For interfaces with a dash, a continuous increase in energy is observed, for those marked with a * the curve decreased and flattened out but no clear minimum in energy was discernable.…”

“…[38][39][40][41] Therefore for the calculation of nominal solubilities representative of real microstructures, only the three high-energy, high-Σ grain boundaries Σ93 (1 1 · 0), Σ13 (1 1 · 3) and Σ43 (2 2 · 3) were considered. To limit the number of possible dopant configurations per unit cell, segregation was assumed to be limited to the 12 lowest energy sites for Y and Mg, and the 18 lowest energy sites for oxygen vacancies.…”

“…It is possible however that the grain shape and size distribution changes as a function of grain size, leading to a different specific grain boundary area. As experimental data points were collected from several different sources, 41,[65][66][67][68] it is also possible that experimental parameters such as impurity level, porosity, sintering temperatures or cooling rates lead to slight variations in solubility.…”