REPOSITORY > RESULTS

Abstract

In this thesis we demonstrate a new, simple, and ecient method for the solid-state synthesis
of oxides with a complex chemical composition. We modify the traditional solidstate
synthesis method, which is used in the majority of ceramics laboratories around
the world, in order to prepare multi-metallic oxide powders from classic ferroelectrics
to very popular multiferroic and other materials. Firm principles are established in the
field of solid-state chemistry, i:e:, the particles of the starting compounds should be
small and should be homogeneously distributed in the reactant mixture. This is very
true for the simple binary systems; however it is not necessarily true for more complex
systems involving several starting compounds, which may show complex reaction
pathways during heating. It can happen that some undesirable intermediate products
form at lower temperatures, which at the end react to form the final product, however,
the phase purity of such products and their chemical homogeneity is questionable.
Therefore, in many cases a simple, solid-state synthesis does not give the desired results.
The problem described above is encountered in the synthesis of Pb(Mg1=3Nb2=3)O3
(PMN) based materials. In the mixture of PbO, Nb2O5, and MgO, the first two species
preferentially react to form Pb-Nb pyrochlore phases, which are stable and hard to eliminate
from the final product. The presence of these phases deteriorates the properties
of the ceramics. First we have to address the question how to avoid or slow down the
parasitic reactions and how to force the correct ones. In a solid-state synthesis there
are not many options to tailor the chemical reactivity. Our idea is to simply avoid or
limit the contacts between the reactant particles that lead to non-desirable products,
i:e:, the PbO-Nb2O5 contacts. Our next challenge is to create the desired contacts in
the reagent powder mixture. The solution to this problem is the formation of colloidal
agglomerates of reagent particles with the desired structure.
First, we measure the electrokinetic properties of our starting components, i:e:,
PbO, Nb2O5 and (MgCO3)4∙Mg(OH)2∙4H2O, (MHC). Using these data as the input
parameters we employ Monte Carlo simulations to predict the formation of agglomerates
in the suspensions of reagent particles in different pH conditions. Our simulations reveal
a large population of clusters with close contacts between the PbO and Nb2O5 at pH
= 11.4, which thus enable the pyroclore phase to form. Whereas, at pH = 12.5 the
competition between the repulsive and the attractive interactions changes in favor of
the assembly of the equilibrium clusters in which MHC particles effectively separate
PbO and Nb2O5. We also explore how the varying size of MHC particles affects the
cluster assembly in both relevant pH conditions. A detailed analysis of the contacts
between the particles in the system as a function of different conditions is presented.
Next we use the results from simulations to conduct the experiments. The aqueous
suspensions of reagent particles are prepared at pH = 10.0, pH = 11.4, pH = 12.5, and in
acetone. The suspensions are dried and heated at elevated temperatures. After heating
at 900 °C a pure PMN perovskite phase is obtained only from the sample with the pH
= 12.5. All the other samples contain, in addition to the perovskite, the detrimental
pyrochlore phase. The results are in agreement with the structure of the agglomerates
predicted by simulations. The properties of the ceramics prepared from the sample with
pH = 12.5 are superior to those of the other samples. Furthermore, dense ceramics are
obtained by sintering as low as 950 °C with the dielectric properties being comparable
to the values obtained for ceramics sintered at about 200 K higher temperatures.
Finally, we adopt the same approach for the synthesis of the 0.65Pb(Mg1=3Nb2=3)O3-
0.35PbTiO3. Here, one additional component, TiO2, is added to the reactant mixture.
The aqueous suspensions of the reactant particles are prepared at pH = 11.4 and pH
= 12.5. The two powders are calcined at 850 °C, where only the pH12.5 powder yields
a phase-pure perovskite. Similar to the PMN case, the pH11.4 powder is composed
of a mixture of perovskite and pyrochlore. After sintering at 950 °C the density of
the pH12.5 ceramic is equal to 96 %, while the pH11.4 sample reaches only 81 % of
theoretical density. The pH12.5 ceramic exhibits electrical properties comparable to
the ceramics prepared by the columbite method. On the other hand, the electrical
properties of the pH11.4 ceramic are heavily deteriorated due to the presence of the
pyrochlore phase and a high degree of porosity.