Field Acceleration for Oxide Breakdown - Can an Accurate Anode Hole Injection Resolve the E vs 1/E Controversy

Concerned by the inability of the Berkeley Anode Hole Injection (B-AHI) model to explain recent experimental breakdown data at low voltages, researchers have speculated on alternate mechanisms for oxide breakdown. The B-AHI predicts an 1/E scaling of the time-to-breakdown (in log scale) for SiO sub 2 films which is inconsistent with an approximately E-dependent breakdown behavior at low voltages. This new mechanism is expected to explain the linear field dependence (E-model) of oxide breakdown under low voltages conditions. The practical implication of this has been very pessimistic projection for dielectric reliability, and a scenario where oxide reliability concerns prevent future device scaling. In this paper, we establish that such pessimism may not be warranted, and is based on inadequate appreciation and implementation of the concept of Anode Hole Injection - tunneling electrons produce hot holes in the anode through impact ionization which can penetrate the oxide, create traps and lead to breakdown.