Physical model of the impact of metal grain work function variability on emerging dual metal gate MOSFETs and its implication for SRAM reliability

A new model of work function variability (WFV) based on grain orientation differences of the polycrystalline metal gate is reported. Our model predicts that at the 22 nm technology node, among the three device variability sources: random dopant fluctuation (RDF), line edge roughness (LER) and WFV, WFV will cross over RDF and becomes the dominating factor. The SRAM circuit analysis shows that write/read failures are underestimated by 9 orders of magnitude by the area weighted averaged work function model.