Granular description of charging kinetics in silicon nanocrystals memories

In this paper we present a new model of charging kinetics in Silicon nanocrystals memories based on a granular charging of a floating gate. In 2000, a previous model, due to De Salvo and co-workers, was based on a continuous charging of a floating gate, borrowed from the conventional of non-volatile floating gate memories model. We will summarise the main equations of the latter and try to compare the results with current–voltage and current–time measurements. We will show that this approach only allows a qualitative description of the charging kinetics but is not accurate enough to perfectly fit the experimental data. Its main limitation lies in a lack of describing the granular nature of the storage nodes: a fraction of an electron—or of a hole—can be stored inside a nanocrystal. This, of course, is not really correct. Instead, our new model admits only two available states: a nanocrystal is neutral or is charged with one electron. This simple logic will enable an accurate description of nanocrystals memories with granular effects unpredictable with De Salvo’s approach. The charging kinetics in our model is more accurately described in all regimes and better incorporates the effect of the nanocrystals size distribution permitting direct extraction of this one from current–voltage measurements. An advantage of our model is enhanced by the analytical leading equations approach easy to incorporate in a simulator and directly applicable to memory devices. It is fast and easy to handle.