A figure of merit for optimization of nanocrystal flash memory design

Abstract

Nanocrystals can be used as storage media for carriers in flash memories. The performance of a nanocrystal flash memory depends critically on the choice of nanocrystal size and density as well as on the choice of tunnel dielectric properties. The performance of a nanocrystal memory device can be expressed in terms of write/erase speed, carrier retention time and cycling durability. We present a model that describes the charge/discharge dynamics of nanocrystal flash memories and calculate the effect of nanocrystal, gate, tunnel dielectric and substrate properties on device performance. The model assumes charge storage in quantized energy levels of nanocrystals. Effect of temperature is included implicitly in the model through perturbation of the substrate minority carrier concentration and Fermi level. Because a large number of variables affect these performance measures, in order to compare various designs, a figure of merit that measures the device performance in terms of design parameters is defined as a function of write/erase/discharge times which are calculated using the theoretical model. The effects of nanocrystal size and density, gate work function, substrate doping, control and tunnel dielectric properties and device geometry on the device performance are evaluated through the figure of merit. Experimental data showing agreement of the theoretical model with the measurement results are presented for devices that has PECVD grown germanium nanocrystals as the storage media. Copyright