1.3 Modeling BTI with Defects

According to the current understanding BTI is either due to the creation of interface states at the Si-SiO
   2   interface, generally known as P
  b   -centers, and/or to the trapping of positive charge inside the SiO2   , possibly at so-called   ′
E -centers. In order to link the observed degradation to these two types of defects as possible underlying physical origin of the BTI phenomenon, the definition

          ΔQot + ΔQit
ΔVTH  = − -------------
              Cox
(1.1)

with Qot   as oxide charge, Qit = q0Nitf(VG)  as the charge stored in the interface traps, and Nit   as the interface trap density and f  as the occupancy function at the interface is used. Unfortunately, these charges, whose sum is directly related to the change in VTH   , cannot be measured directly. Also, the relative contribution of Qot   and Qit   does not necessarily have to be constant over VG   . Nevertheless, till now the defects causing Qot   and Qit   are assumed to be the most likely explanation to why BTI happens.

So far, a large number of recent publications have tried to explain NBTI, and in recent years also PBTI. Numerous methods have been devised to classify the contributing defects, their number, spatial positions and energies, and time response. Also, the possible mechanisms of defect creation were thoroughly studied, but up to now, no consensus has been achieved and the debate on the underlying physics and its consequences continues.