Comprehensive physics-based modeling of post-cycling long-term data retention in 176L 3-D NAND Flash Memories

The contributions to long-term data retention in Charge Trap (CT) 3-D NAND from (i) Vertical Loss components, including Trap Assisted Tunneling (TAT), charge De-Trapping (DT) and interface trap passivation (ITP), and (ii) Inter-Cell components, namely Lateral Migration (LM), are studied for various Program/Erase (P/E) cycling doses and program levels (L), at different bake temperatures (T
BAKE
) using a physics-based Activated Barrier Double Well Thermionic (ABDWT) model. The advantages and computational efficiency of the above model with respect to the empirical stretched exponential model is illustrated by comparing the parameters required to model the overall charge loss. It is shown that the ABDWT approach can model the total charge loss with a single parameter, which varies for each contributing mechanism with cycling dose only. The physical significance of the parameter variation of each mechanism with various conditions is explained.