Negative-Bias Temperature Instability: Measurement and Degradation Mechanisms

摘要

This paper addresses two important aspects of negative-bias temperature instability (NBTI), namely measurement methodology and degradation mechanisms. An improved ultra-fast switching measurement method with sub-100 ns delay is presented. The method is based on customary pulsed current-voltage equipment and thus provides an attractive standard for NBTI measurement. Both linear and subthreshold drain current measurements are discussed and the relevant limitations addressed. In the discussion, it is shown that the gate voltage shift extracted through the computation of ∫ partial derivI_(d,lin)/g_m grossly overestimates the threshold voltage shift |ΔV_t| (I.e. the fraction of gate voltage shift due to interfacial trapped charge) because of mobility degradation. Neglecting this effect would result in an unduly pessimistic assessment of interfacial trapped charge generation. We demonstrate a better method that could accurately delineate both |ΔV_t| and low-field mobility degradation. In addition, results which indicate two distinct degradation mechanisms are presented. Experimental evidence that shows a "bipolar" (I.e. both acceptor- and donor-like) characteristic of stress induced interfacial traps is discussed. Finally, features of dynamic NBTI which do not conform to the reaction-diffusion model are highlighted.