Switching-time analysis of binary-oxide memristors via a nonlinear model

A simulation-based analysis is conducted of the ionic switching times for nanometer-scale binary-oxide "memristor" devices. This analysis is based upon a device model that incorporates nonlinear field-driven ionic transport within the bulk of the memristor. In contrast, prior models of charge transport in such devices have relied upon linear simplifications, or else they have included nonlinear effects only at the electrode interfaces. As shown here via simulation, the nonlinear model provides much closer quantitative agreement with experimentally observed device switching times. Also, this model predicts a distinct asymmetry between the "set" and "reset" switching behaviors of memristors that is not present in linear models. Thus, the model and the quantitative results derived using it suggest an experimental route by which the underlying device physics might be elucidated further.