Breakdown in silicon dioxide films on silicon has been studied making use of various measurement techniques and statistical tools. The use of probability scales is advocated in order to distinguish intrinsic breakdown from defect‐related breakdown. In such a way defect densities can be determined and the yield of large‐area circuitry can be predicted when the breakdown distribution of small capacitors is known. Comparison of field‐dependent, time‐dependent, and current‐induced breakdown showed that intrinsic breakdown occurred at practically equivalent conditions of field and time. At high fields or large currents breakdown occurred within a short period of time, whereas at low fields or small currents this period can be relatively long. The critical parameter was found to be the time integral of the leakage current until breakdown (Qbd). This charge needed for breakdown is constant for similar capacitors. At not too high currents (Qbd) is found to be only a little dependent on stress conditions: field or time. Qbd appears to be even independent of the measurement technique, i.e., field‐, or time‐, or current‐induced breakdown technique. At higher currents or applied fields (Qbd) values tend to decrease, sometimes over orders of magnitude. It follows from our observations that this is due to space‐charge effects. Special attention is given to the energy needed for breakdown, which is different from that which is released at the breakdown itself.

This content is only available via PDF.
You do not currently have access to this content.