A study of clustering using microelectronic defect monitors

Clustering of defects has received extensive attention for its impact on the predicted yields of large chips and on-chip redundancy. Defect density, defined for the separation of variables, permits product-independent cross-comparisons between fabricators and technologies. Defect densities are best tracked by inline measurements on microelectronic test structures, with product identity in their sensitivity to defects. It is reported that the defects monitored via the microelectronic test structures may not have the extent of clustering that has been noted in the literature.<
>