Phase Evolution in the Transformation of Atomically Mixed Versus Ball-Milled Mixtures of Nanopowders in the Formation of Composite MO·3Al2O3 Spinels: Bottom-Up Processing is Not Always Optimal

• Published in: Journal of the American Ceramic Society 2014-11, Vol.97 (11), p.3442-3451
• Main Authors: Taylor, Nathan J., Pottebaum, Andrew J., Uz, Veli, Laine, Richard M.
• Format: Article
• Language: English
• ISSN: 0002-7820; 1551-2916
• DOI: 10.1111/jace.13160

Liquid‐feed flame spray pyrolysis (LF‐FSP) provides atomically homogeneous mixed metal powders with 30–40 nm average particle sizes, often producing kinetic phases due to the high quench rate As produced LF‐FSP Al2O3‐rich spinels, such as MgO·3Al2O3, form an Al2O3‐rich metastable single‐phase spinel. On heating, the powders phase separate to form MAl2O4 and α‐Al2O3. Compacts of MO·3Al2O3 (M = Co, Ni, Mg) were produced and sintered to evaluate the final duplex microstructure. The same composition was also approached from stoichiometric LF‐FSP MAl2O4 nanopowders ball‐milled with Al2O3 nanopowders in an attempt to evaluate how the initial length scale of mixing affected the final microstructure. Contrary to traditional sintering, we observe two distinct mechanisms. At 1000°C–1200°C, cation diffusion appears to control densification as a consequence of high vacancy concentrations and atomic mixing where traditionally expected site inversion plays less of a factor given the high quench rates. The second mechanism follows α‐Al2O3 exsolution and densification occurs via oxygen diffusion and α‐Al2O3 grain growth. When sintering the duplex MAl2O4/α‐Al2O3 compacts to at least 95% theoretical density, we find final microstructures that do not reflect the initial degrees of mixing. That is, the atomically mixed MgO·3Al2O3 does not does not offer an advantage over the submicron length scale of mixing in the ball‐milled samples.