The multi-bridge method is often used in the industry to measureelectrical linewidths of interconnect lines because it requires lesssensitive equipment than the standard cross-bridge method. The methodassumes that the linewidth process bias is independent of design width,which is shown here to be only approximately correct. This can result inlinewidth determinations that are significantly different from thevalues obtained by the cross-bridge method. This is observed most oftenat the narrower designed widths. A composite method is described thatminimizes these differences. The multi-bridge test structure can also beused to detect the linewidths at which chemical mechanical polishing(CMP) induces “dishing” effects in copper lines
展开▼
