Depth profiling of very thin HfO2/Al2O3 stacks by ellipsometry

Abstract

Abstract Ellipsometry (VASE and MAIE) with appropriate algorithms for experimental data interpretation was applied for quantitative characterization of thin Al2O3/HfO2 multilayers formed by atomic layer deposition (ALD) applicable to the fabrication of charge-trapping nonvolatile memories. A substantial benefit of the algorithms is the depth profiling of the stacks. In this work, the depth profiles were retrieved of the HfO2 constituent in very thin (Al2O3:HfO2) stacks embedded in thin Al2O3 surrounding layers. The peculiarities in the achieved depth profiles are used to determine the bi-layer blocks and their sub-layer thicknesses. The influence of a rapid thermal annealing in O2 on the depth profile and the sub-layer thicknesses is studied; substantial changes are thus revealed in the thickness and composition of the stacks. This information could be used in optimizing the dielectric and electrical properties of the stacks and multilayers.