Sintering of nanostructured alumina : influence of various parameters

Most nanocrystalline oxide ceramic powders are metastable and this metastability may have a critical influence on the sintering behaviour. Studies on producing nanocrystalline ceramics from fine powders (< 100 nm) have highlighted the problem of achieving high densities without excessive grain growth. It is particularly true for transition alumina powders which are currently produced with very high specific surface area and ultrafine crystallite sizes; the transformations into the stable alpha-alumina are generally accompanied by vermicular microstructures consisting of a network of large pores which developes during the transformation. The final stages of sintering then require very high temperatures to achieve high densities. In literature, three main routes have been investigated to overcome this difficulty in elaborating dense nanograined alumina ceramics the alpha-alumina seeding of precursor gels (mostly boehmite) which results in the reduction of the transformation temperature and in a refinement of the final microstructure. Despite this breakthrough, and whereupon gel fragments can be sintered to almost full density beyond about 1,200 degreesC, processing difficulties still seem to hinder its application to the production of monolithic pieces. Moreover the sintering mechanism remains to be elucidated; the second approach consists of pressure sintering methods : uniaxial hot-pressing or sinter-forging. It has been suggested that during hot pressing of transition aluminas, pressure-induced particle rearrangement causes impingement of the growing a-alumina colonies, thus limiting the formation of a vermicular pore microstructure; - the last route involves the influence of doping elements on the gamma to alpha transformation. However in general no paper has reported an enhancement or retardation of the gamma to a phase transformation of a sufficient magnitude to lead to a high-density nanograined monolithic material at sintering temperatures lower than those used for fine grained alpha-Al2O3 powders. The ultimate goal of this study is the synthesis of alpha-Al O-2(3) nanoceramics starting from nanocrystalline gamma-Al2O3, doped or un-doped but without pressure assisted sintering. The densification of gamma-Al2O3 typically shows two steps of densification during a constant heating rate experiment. The first step, R1, is a rapid densification associated with the phase transition of gamma-Al2O3 to the stable a phase at around 1,100 degreesC. The second step R2 corresponds to a slower densification of the alpha-Al2O3 at higher temperatures. With the idea that a phase transformation could be used to enhance the densification of nanostructured alumina and working on readily available commercial transition alumina powders, we explore the effects of various parameters such as green density, heating rate and a-alumina initial content on the phase transformation gamma-alpha and on the main steps of densification. An enhanced densification, over and above that expected for the gamma to alpha phase transition is clearly observed. This enhanced relative density change is brought about by particle rearrangement during the transformation, the degree of particle rearrangement being influenced by the above mentioned parameters which also influence the subsequent sintering of the alpha-phase. By following the densification curves and following the corresponding microstructural evolution we hope to elucidate the key mechanism of such phase transformation assisted densification. Results concerning the effects of doping elements (magnesium, yttrium and zirconium) on the transformation-densification behaviour of the same gamma-alumina raw powder batch are also presented. Y and Zr shift the transformation to higher temperatures and Mg has a much less pronounced effect. An additional densification rate peak is observed for Y doping, its temperature depending on the Y-doping level. This temporary densification was correlated with the precipitation of a second phase when the Y segregation at grain boundaries goes beyond a saturation level. For each doping element, microstructural investigations have shown that above its solubility limit in the bulk, the dopant is segregated to grain boundaries and then precipitated as a second phase with increasing grain size.