Barium titanate, a leadfree ferroelectric material with numerous application capabilities, is among the more popular oxides flash sintering was applied to. While some studies report very promising densification behavior of
>95%, many others have observed hotspot-caused damage to the green bodies during conventional flash experiments long before proper densification of the samples was possible, in violation to the production map established by previous data.
This seeming contradiction is resolved when geometry of the green bodies is considered. In our work, we use in-situ infrared imaging and dilatometry to investigate the temperature distributions on the sample surfaces during the flash events in both conventional flash and current-ramped experiments. We vary geometrical parameters of the samples and setups to investigate changes to homogeneity, onset criteria and density. We explore how far the violent hotspots in barium titanate can be avoided through minimization of temperature gradients through changes to the setup and demonstrate that, in barium titanate, two separate processes take place: proper flash sintering incubating and densifying the sample, and a dielectric breakdown which is not directly related to flash sintering behavior.
Additionally, the lower boundary of the onset temperature is investigated and does not coincide with the Debye-temperature of barium titanate.