Analysis of the mechanisms of dielectric breakdown in ceramic materials
Our experiments from the first project period showed that space charge limited conduction (SCLC) is the dominating conduction mechanism in 5µm-5mm thick ceramics at dielectric breakdown. This conduction mechanism is also the reason for the thickness-dependence of the breakdown strength. This means that all breakdown models which assume ohmic conduction at breakdown for 5µm-5mm thick ceramics are incorrect. In the first project period, we developed a breakdown model in analogy to the energy release rate concept known from fracture mechanics. This breakdown model is able to explain the experimentally found thickness-dependence of the breakdown strength as well as the dependence of the breakdown strength to the permittivity.In the following the scientific objectives of the renewal proposal are described:1. The developed breakdown model will be validated. The model predicts a dependence of the breakdown strength to the length of conducting filaments which initiate the breakdown process. These conducting filaments will be imitated experimentally by electrically conducting channels of different lengths which will be introduced by FIB (Focused Ion Beam) into defect-free substrates. The breakdown strength of these so prepared substrates will be correlated to the length of the FIBed channels and compared to the prediction of the breakdown model.2. The conduction mechanism of SCLC is thickness-dependent which means that the onset of SCLC in shifted to higher voltages in thicker samples. Therefore, there should be a limiting thickness above which the conduction mechanism will not be SCLC at breakdown anymore. In order to validate this assumption, experiments on thick samples are planned. The thickness is chosen that way that SCLC cannot establish before breakdown happens. In this third thickness range the breakdown strength should at least show a different thickness-dependence compared to what was already measured. The newly proposed breakdown model is not valid for this thickness range anymore and we can define an upper validity limit for the model in terms of sample thickness. 3. The breakdown strength of stacks should be evaluated. The dominating conduction mechanism at breakdown is SCLC. Therefore, it can be assumed that the breakdown of a layered sample cannot be explained by the linear ohmic model of a layered dielectric. Here, the influence of stacks with a thickness ranging in the thickness-dependent as well as in the thickness-independent breakdown strength range will be evaluated. Besides stacks of same material also stacks of different materials in terms of permittivity will be tested. We expect that the breakdown strength of a high-permittivity material can be enhanced by a thin layer of a low-permittivity material. This is of notable interest for technical application. In order to describe the breakdown strength of layered samples the linear model of a layered dielectric should be enhanced by SCLC.