Chasing ghosts: How an SRAM detected the subtle impact of stray light

The ever-shrinking world of semiconductors has always challenged the interplay of tool capability, process integration, and characterization. The fine line between structural or electrical success and failure has steadily been redefined from microns to nanometers with leading edge technology using terms with the likes of angstroms and layers of atoms. Failure modes at these nodes become increasing difficult at times to even grasp, let alone “visualize” with electrical and physical analytical techniques. The complexity of these state-of-the-art problem sets demand adaptive and robust problem solving methodologies tailored to uncover and drive root cause understanding. Assembling and organizing diverse teams covering a broad range of expertise becomes paramount. In this case study, we were able to make use of a unique low voltage static random access memory fail mode and spatial fingerprint to chase down a 1 nanometer change in transistor gate line widths; the sleuthing effort concluded with the unlikely combination of mask chrome frame and stray light from lithography as the ultimate root cause.

[1]  David Clark,et al.  Tristate Inverter Array: A new test structure that compliments traditional SRAM arrays as a yield learning vehicle , 2014, 25th Annual SEMI Advanced Semiconductor Manufacturing Conference (ASMC 2014).

[2]  Jianhua Yin,et al.  Case studies of fault isolation for the global failing patterns on SRAM bitmap caused by the defects in peripheral logic regions , 2013, ASMC 2013 SEMI Advanced Semiconductor Manufacturing Conference.