Next: 1.3.4 Strong Localization Up: 1.3 Characteristic Features of HCD Previous: 1.3.2 HCD in High-Voltage Devices
In contrast to HV devices, the operating voltages of scaled devices are such that a single carrier is unlikely to reach energies sufficiently large to trigger an SP-process. The process of energy exchange between carriers is of a stochastic nature and therefore one may expect that a certain fraction of particles - however small - may still obtain a relatively large energy. Regardless, although particles which are able to launch the SP-mechanism are in principle present, their relative number is rather small and, hence, the MP-process becomes dominant [40,88]. Contrary to the SP-mechanism, the individual carriers contributing to the MP-mechanism require only a relatively low energy. However, a large number of those carriers is needed. Thus, the carrier flux rather than the single-carrier energy becomes important in this case. The maximum carrier flux is obtained at Vds = Vgs for both scaled n- and p-MOSFETs [89,90,91,92], which now becomes the region of maximum HCD.
It is worth mentioning that even in the case of ultra-short devices a certain fraction of "hot" carriers exists because the high-energy tail of the carrier distribution function is populated for instance by the electron-electron scattering process [29,41]. Therefore, the SP-mechanism will still contribute in these devices. Also, thermalized, that is, "cold", particles still exist even in the case of high-voltage devices, thereby also leading to HCD by the MP-process. To conclude, in a real device under real operating/stress conditions, the interplay between the SP- and MP-modes of bond-breakage must considered and is controlled by the way carriers are distributed over energy, that is, by the carrier DF.