Chapter 7
The Extended Nonradiative Multi-Phonon Model

The Kirton model and the TSM rely on the concept of NMP processes and can therefore give an explanation for the experimentally observed temperature activation of τcap  and τem  . However, the energy barriers used in those models are not calculated by the crossing point of two adiabatic potentials, as it would be necessary according to theoretical considerations. Palma et al.  [125] related the measured capture and emission time constants of RTN to the properties of oxide defects based on NMP theory. The success of their model was based on the effect of a Coulomb barrier, which was required to obtain the correct field dependence of τcap  and τem  . Irrespective of the physical correctness of this assumption, the Coulomb barrier was only successfully applied to MOSFETs with thick dielectrics [184125123]. In a few RTN studies [56] as well as in more recent TDDS measurements [53], interesting phenomena, such as aRTN and tRTN (both discussed in Section 1.3.4), have been observed. They are linked to quite complex trapping dynamics, which can only be explained by metastable defect states. Accordingly, a successful, physics-based NBTI model must incorporate additional metastable defect configurations in order to account for all experimental results. In a new modeling attempt, the theoretical framework of NMP charge transfer reactions has been combined with the concept of metastable states. The resultant model has been termed extended NMP (eNMP) and will be addressed in the following.

 7.1 Transition Rates according to the NMP Theory
 7.2 States of a Bistable Defect
 7.3 Model Evaluation
 7.4 Analytics Derivation of the Capture and Emission Time Constants
 7.5 Explanation for Noise in TDDS Measurements
 7.6 Discussion
 7.7 Conclusion