Changes in the Grain Boundary Character and Energy Distributions Resulting from a Complexion Transition in Ca-Doped Yttria STEPHANIE A. BOJARSKI, SHUAILEI MA, WILLIAM LENTHE, MARTIN P. HARMER, and GREGORY S. ROHRER The grain boundary character distribution and the relative grain boundary energy of 100 ppm Ca-doped yttria were measured before and after a previously identified grain boundary com- plexion transition. The grain boundary character distribution of samples exhibiting normal grain growth (before the complexion transition) favored {111} planes, whereas those exhibiting abnormal grain growth (after the complexion transition) favored {001} planes. Additionally, the relative grain boundary-to-surface energy ratios in the sample exhibiting abnormal grain growth were 33 pct lower than in the sample exhibiting normal grain growth. The results also indicate that the complexion transition increased the anisotropy of the grain boundary energy, and this may be responsible for the increase in the anisotropy of the grain boundary character distribution. DOI: 10.1007/s11661-012-1172-y Ó The Minerals, Metals & Materials Society and ASM International 2012 I. INTRODUCTION CONTROLLING microstructural development to obtain a theoretically dense material has been an important objective of research on yttria ceramics. [1–7] This isotropic cubic material has a large range of transparency, high melting temperature, high thermal conductivity, low thermal expansion, and corrosion resistance. Transparent yttria, in particular, is widely investigated for use as a host material for lasers [8–10] and for military applications such as infrared windows in heat-seeking rockets. [11–13] Tailoring the grain size to obtain a dense, homogenous, fine-grained microstruc- ture is essential for optical and infrared transparency in polycrystalline ceramics. Understanding the grain boundaries in yttria will allow for more accurate control of the processes, such as grain growth [14] and sinter- ing, [1–3,15] that influence the microstructural develop- ment and, thus, the mechanical and optical properties of the bulk ceramic. The term ‘‘grain boundary complexion’’ is relatively new in microstructural science and is being used to refer to groups of grain boundaries, which are thermody- namically stable phases in their own right possessing distinct structures and compositions different from any bulk phases. [16–23] In at least some cases, grain bound- aries with different complexions can have very different properties that dominate microstructural evolution. For example, the coexistence of a high mobility and low mobility complexion in the same sample can lead to abnormal grain growth. [24] Previous work on doped aluminas has shown that a complexion transition can change both the grain boundary character distribution (GBCD) and the relative grain boundary energy. [25,26] Furthermore, it has recently been shown that the existence of a nanometer-thick intergranular film reduces the energy of the Au-alumina interface. [22] However, in the prior work, the GBCD was determined only as a function of the two grain boundary plane parameters; in this article, we examine changes in the five parameter grain boundary character distribution that are coupled to a complexion transition. Ma [27] recently conducted a comprehensive investiga- tion of grain growth kinetics in dense Ca and Si-doped yttria. In this study, abnormal grain growth occurred in 100 ppm Ca-doped yttria samples that were isother- mally annealed in a reducing atmosphere at tempera- tures above 1973 K (1700 °C) and held at the annealing temperature for times longer than 0 hours. Mobility measurements paired with high-resolution transmission electron microscopy (HRTEM) and high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) found that the boundaries around the high-mobility grains have an amorphous intergran- ular film, whereas the boundaries around the slow moving grains have a lower order grain boundary complexion, which was deduced to be bilayer of adsorbed Ca. With reference to the original study of doped aluminas, [18] the film would correspond to the complexion labeled V or VI and the bilayer to the complexion labeled III. Regardless of the labels, the STEPHANIE A. BOJARSKI, PhD Candidate, and GREGORY S. ROHRER, W.W. Mullins Professor and Head, are with the Department of Materials Science and Engineering, Carnegie Mellon University, Pittsburgh, PA 15213. Contact e-mail: gr20@andrew. cmu.edu SHUAILEI MA, formerly PhD Candidate, Lehigh Uni- versity, Bethlehem, PA 18015, is now Product Developer with the SABIC Innovative Plastics, Pittsfield, MA 01201. WILLIAM LENTHE, Undergraduate, and MARTIN P. HARMER, Alcoa Professor, are with the Department of Materials Science and Engineering, Lehigh University. Manuscript submitted January 11, 2012. Article published online April 13, 2012 3532—VOLUME 43A, OCTOBER 2012 METALLURGICAL AND MATERIALS TRANSACTIONS A