Tomas Raming

Ph.D. thesis

Thesis title:

The synthesis of nano-nano dual phase ceramic composites

[thesis in pdf format]




Prof. Dr. Ir. H. Verweij

Assistant promotor:

Dr. W.E. Van Zyl


Nano-nano dual-phase composites in which both phases are present as grains with typical dimensions <100 nm are expected to exhibit physical properties, deviating from those of coarser-grained composites. If one phase is electron conducting and the other electrically isolating, special properties can be expected due to Schottky-barrier space-charge formation in the isolator phase near dual-phase boundaries. The extent of the Schottky-barrier (Debije length) can be of the same order as the total grain dimension resulting in special electrical and optical properties. These expectations are the base of this research.

Dense nano-nano dual-phase composites hardly have been obtained, so far. Therefore the focus of this research was the synthesis of dense oxidic nano-nano dual-phase composites. Only wet-chemical powder preparation routes were used. The start of all preparation routes was the synthesis of nano-crystalline low-agglomerated powders, for the system zirconia-hematite (α-Fe2O3) and zirconia-ruthenia (RuO2). The zirconia in both cases was 3Y-TZP, ZrO2 doped with 3 mol% Y2O3.

For the system zirconia-hematite, two types of powder synthesis routes were chosen. Either a co-precipitation method was used, in which the three different types of metal ions (Fe3+, Zr4+ and Y3+) were precipitated simultaneously, or a sequential precipitation method was in used. In the latter case, first a low-agglomerated suspension of either nanocrystalline zirconia or hematite particles was made. The synthesis of hematite suspensions was studied in more detail. It was found that the amount of agglomeration of the hematite particles in suspension could be reduced by making small (40 nm) superparamagnetic crystallites. For the sequential precipitation methods the second phase was precipitated in an aqueous suspensions of the first phase. All precipitation methods resulted in (partly) amorphous gels that were calcined to produce fully oxidic nanocrystalline powders.

The co-precipitation method led to the most homogeneous distribution of Zr- and Fe-species. If the Fe2O3-content was £34 mol%, the co-precipitation method and subsequent calcination at 700°C produced single-phase solid solutions of ferric oxide in yttria-doped-zirconia. This mixed oxide had a cubic structure. When heating this powder at 800-900°C almost all Fe3+-ions left the zirconia phase to form a separate hematite phase. Concomitantly, the structure of the zirconia phase changed from cubic to tetragonal.

By developing a generally applicable EDX-SEM method the homogeneity of Fe-Zr distribution of the powders could semi-quantitatively be determined. The by sequential precipitation prepared powders were less homogeneous compared to the single-phase co-precipitated powders, but the powders that were made by precipitating ferric chloride in a zirconia suspension were quite homogeneous too.

Different methods were used to form the powders into compacts. Magnetic pulse compaction proved to be the most promising green compaction method. Most powders were compacted by using isostatical pressing, however. The compacts were sintered either pressureless in air at 1150°C, or by using sinterforging at 1000°C and 100MPa. Sinterforging in most cases produced dense (>95%) composites with smaller grain sizes compared to pressureless sintering. Sinterforging proved to be successful in producing for the first time dense zirconia-hematite composites with average grain size for both phases <100 nm. Pressureless sintering could not achieve this goal, probably due to an inhomogeneous pore distribution in the compacts.

Sequentially precipitated powders produced more homogeneous composites with smaller grains compared to co-precipitation after sintering. The co-precipitated powders showed strong grain growth and dehomogenisation during sintering, probably due to the higher amount of energy stored in these systems compared to powders prepared by sequential precipitation. Hydrothermal crystallisation showed to be a promising preparation method to make low-agglomerated nanopowders.

For the zirconia-ruthenia system, co-precipitation resulted in dual-phase zirconia-ruthenia powders, just as for the sequential precipitation method. Almost no ruthenia dissolved into the zirconia lattice. Co-precipitation still resulted in more homogeneous powders. All prepared zirconia-ruthenia powders showed strong dehomogenisation of the two phases when heating >400°C, due to the oxidative volatilisation of the RuO2. It was more difficult to densify the zirconia-ruthenia powders compared to the zirconia-hematite powders. Pressureless sintering was not successful, but sinterforging at 1150°C and 100MPa produced for the first time dense dual-phase zirconia-ruthenia composites. These composites showed dehomogenisation of the two phases on very large scale (100 micrometer). This phase separation during sintering could largely be prevented by using explosive compaction as powder compaction technique. This resulted in green compacts with the extremely high density of 78%. Explosive compaction resulted in brittle materials, however, that only could be densified to high density (93%) by using pressureless sintering. Still in the latter case dehomogenisation was much less compared to isostatical pressing and subsequent sinterforging.

The dense dual-phase zirconia-ruthenia composites showed high electron conductivity, especially at low RuO2-percentage (15 mol%), due to their inhomogeneous microstructure. The electrical conductivity was shown to be depending very much on the distribution of the ruthenia in the composite.