An Investigation Into the Depandance of Electrical Properties on Microstructure in Polycrystalline Barium Titanate

Abstract

It has been recognised for some considerable time that the intrinsic properties of particular materials determined from single crystals are not always reflected in the properties of polycrystalline materials of the same composition. In particular, it is known that samples of polycrystalline ceramics are invariably more complex than their single-crystal form. The differences which occur may be seen to depend upon the processing and thermal history of the material, in the size and distribution of grains, the number, size and distribution of voids, and the nature of the boundaries between the grains. Ceramic barium titanate is a typical example of this type of material. Various investigators have stated that the dielectric properties of the material are influenced by the microstructure, but relatively few direct correlations have been reported. A survey of the literature available revealed that, in every case, the investigators had based their work on test pieces formed by pressing powders of varying purity in a punch and die using organic additives as binders. Whilst this method of compaction can be used to good advantage when producing commercial grade bodies, it would appear to be quite unsatisfactory for the fabrication of laboratory test pieces, the inhomogeneous nature of the material, caused by differential compaction, producing warping and (in extreme cases) cracking of the fired ceramic. The strains present within this type of body would obviously contribute towards the electrical properties of the material and could, perhaps, produce effects which might mask the true nature of the test pieces, and so produce the discrepancies observed in the reported literature. In order to determine the true nature of the material, ceramic test pieces were produced by pressing powders of known composition in a water filled die, a method known to produce homogeneous powder compacts. These were then fired under closely controlled conditions together with samples of the same powders pressed in the conventional manner (with a punch and die), care being taken to remove all possible sources of impurity during this process. Metal electrodes were then applied to discs of the ceramics and electrical measurements were made over a wide temperature range (-70° to 150°C). Specimens from each batch of test pieces were examined in detail, using advanced physical methods, in order to determine their exact structure. A detailed correlation of physical and electrical properties then followed, together with a direct comparison with specimens pressed by the conventional method.