2003 IRW High-k discussion summary

Summary form only given. In our discussion there were widely divergent opinions with regard to most important issues (some consensus) Can we assess high-k reliability now? Probably too early to draw conclusions, for three reasons: 1. probably high densities of extrinsic traps 2. processing techniques still in infancy 3. standard SiOz/Si methodologies may not be valid when applied to high-k systems. Much of our discussion revolved around the interfacial layers between the high-k dielectrics and the silicon of the channel (no consensus). It??s not clear what the chemical compositions are. Are they pure SO?? Silicon-rich SiOz? Hafnium (or other) metal silicates? A phase separated mixture of Si02 and HfOz (or other metals)? Although most attendees seemed to agree that the interfacial layer is extremely important, there was little consensus with regard to interfacial chemistry and structure.