Antenna structures are commonly used to study charging, and the resultant metal–oxide–semiconductor gate oxide damage, from plasma processing. Degradation of various device parameters, such as gate leakage and threshold voltage, have been correlated with high levels of charging. However, there has been little work done to correlate charging with reduced device reliability. In this study, we use antenna structures and accelerated life testing of application specific integrated circuit devices on the same wafers to show that oxide damage from charging measured by antennas does not necessarily correlate with reliability degradation. Depending on how the antenna is designed, the measured oxide damage results from a combination of process‐induced charging and initial gate oxide quality. When initial oxide quality is high, then charging which does not degrade oxide reliability can occur. However, for initially defective oxide, the same process‐induced charging can degrade reliability. As shown in this study, the key to distinguishing between the two is in the antenna design. Failures of simple antenna structures having a large antenna area ratio and small gate oxide area did not correlate with degraded oxide reliability. However, failures of antenna test structures having a relatively small antenna area ratio and a relatively large gate oxide area did correlate with reliability, probably by highlighting pre‐existing defects in the gate oxide which induced the early reliability failures. Together, these two types of antenna structures can help differentiate between excessive plasma damage and intrinsic weakness in the gate oxide.

This content is only available via PDF.
You do not currently have access to this content.