Sven van der Gijp

Ph.D. thesis

Summary

Doped barium titanate is used frequently in several types of electrical devices. The electrical behaviour of barium titanate ceramics strongly depend on the composition and the microstructure. In this thesis mainly the application of zirconium doped barium titanate as a pulse-generating device is described. Next to this also the application of barium titanate in multilayer ceramic capacitors is briefly discussed. Finally the preparation of homogeneously doped barium titanate powder is described as well as the microstructure and the related properties of the ceramic.

In chapter 1 the ferroelectric behaviour of barium titanate as well as the influence of dopants on the dielectric behaviour and more specific on the position of the Curie-temperature are described. The physical background on the use of barium titanate for application in lamp starters and in multilayer ceramic capacitors is also discussed. For the application in a lamp starter it is essential that the ceramic is chemically homogeneous and consists of large grains. For application in the multilayer capacitor the particle size is the determining factor.

In chapter 2 a survey of the preparation method for (doped) barium titanate which has been published previously in literature is given. A subdivision in preparation method is made in complexation, precipitation, sol-gel and dispersion techniques. For each technique a number of different preparation methods are given. In this survey emphasis is put on the chemical background of the process as well as the morphology and sinteractivity of the resulting powders. Several of these techniques are tested. Precipitation and complexation-precipitation seem the most suitable techniques for the commercial production of doped barium titanate, because cheap precursors can be used and moreover, because no or only a low amount of organic molecules are used and because the processes can be conducted in an aqueous environment without too many complicated processing steps.

The preparation of barium titanate with the oxalate and peroxide method is described extensively in literature, however the preparation of doped barium titanate is not. When oxalate is added to the peroxide process, the so-called peroxo-oxalate process is formed. In chapter 3 it is described that the addition of oxalate results in an improvement of the morphology of the powder as well as the prevention of second phase formation. The improved powder properties are likely to be due to the formation of BaTi_0.91Zr_0.09O₂(C₂O₄)·3H₂O during the complexation-precipitation stage of the process. The structure of this complex is confirmed by a study on the thermal decomposition behaviour.

In chapter 4 the peroxo-oxalate method is studied in closer detail. To prepare a single-phase powder with a small grain size it is required to start the peroxide process with a chloride salt as precursor. Due to the slow hydrolysis rate, the use of alkoxides as precursors results in the formation of undesired second phases. Next to the hydrolysis rate, the precipitation temperature is of great influence to the particle size also and second phase formation. A precipitation temperature of 40°C, combined with the use of the chloride precursor, will lead to the best results. It is further required to keep the acidity of the ammonium oxalate solution above pH = 9.

The particle size and the degree of aggregation of powders produced by the peroxo-oxalate method can be reduced further. A method that seems suitable to do so is the so-called homogeneous precipitation. In chapter 5 three different methods for homogeneous precipitation are discussed. First the esterification is described, in which by means of an esterification reaction water is generated, which reacts with barium and titanium precursors in an alcoholic environment. Due to difference in hydrolysis rate between the barium and titanium precursor this method results in the formation of undesired second phases. Secondly the complexation method is described, in which a Ba‑EDTA complex is thermally decomposed. This method results in a powder with spherical aggregates of approximately 1 µm in diameter. Finally the urea method is discussed, in which the pH of an aqueous solution is raised by a hydrolysis reaction of urea, which results in the formation of a precipitate. Unfortunately, it is not possible to precipitate barium and titanium simultaneously. All three powders have a reasonably well-defined morphology but still the degree of aggregation can be reduced further. For all three methods a large titanium excess is found with XRF measurements. This excess was the largest in case of the urea method. More development is required in order to make homogeneous precipitation a suitable method for the preparation of barium titanate.

As mentioned before, the homogeneity of the ceramic is important when barium titanate is applied as a pulse-generating device. Little is known in literature about the determination of homogeneity. In chapter 6, therefore, a number of techniques are discussed and compared, for example Auger, SEM-EDX and EPMA which can be used for the determination of the chemical homogeneity of the three powders and their derived ceramics. A hydrothermally prepared powder obtained from Sakai, a peroxo-oxalate prepared powder and a mixed oxide prepared powder are tested. The latter is prepared by a solid state reaction. No or only a small difference in chemical homogeneity between the various ceramics could be found and hence no relation could be found between the homogeneity of ceramics and the homogeneity of powders. This can easily be explained by the occurrence of diffusion at the high temperature used for sintering. SEM-EDX measurements indicate that the deviation in zirconia concentration in the commercial Sakai powder is the largest. To measure the homogeneity of ceramics, Auger spectroscopy seems the best method because, the concentration can be measured locally due to the small spot size.

In chapter 7 the effect of mechanical stress and chemical homogeneity on the dielectric properties of zirconia doped barium titanate is described. To reduce the level of stress, large grains which can be formed with slow heating rates are required. However, it seems that a high sintering temperature has more influence on the value of the maximum dielectric constant than on the grain size. Moreover, no relation between the value of the dielectric maximum and the grain size could be found. Nevertheless, a relation between the height of the maximum constant and the position of the Curie-temperature can be found.

Finally, in chapter 8 perspectives for the research on lamp starters and multilayer capacitor devices are given. For the use of multilayer capacitors, emulsion technology seems very promising, because the resulting fine particles in the solution can directly be used in a tape casting process. Next, a study for the preparation of multilayer capacitors without shaping techniques seems interesting. For the application in lamp starters, the control of composition and the powder morphology seem very important. It is necessary to find the main reason for the lowering of the maximum dielectric constant. Finally it is essential that a standard procedure for the analysis of homogeneity data is developed.