James Juen Hsu

Flash EPROM endurance simulation using physics-based models

A novel unified field-dependent oxide charge generation (FDG) model is introduced to consistently simulate oxide degradation due to Fowler Nordheim (FN) tunneling and hot carrier injection (HCI) stresses over a wide range of oxide field intensity. This model, combined with an interface charge generation model, is used to study effects of stress-induced interface and oxide charges on flash device erase and programming speeds, band-to-band tunneling leakage current, and threshold voltage shift. An efficient cycle-weighting method is introduced to simulate flash device programming/erase (P/E) cycle endurance. Excellent agreement has been achieved between the simulation predications and experimental data over various operation conditions without parameter fittings or preassumed interface and oxide charge distributions. Simulation results show that the endurance characteristics are mainly affected by both the P/E gate current reduction due to oxide charges and the flat-band voltage increase due to both oxide and interface charges.<<ETX>>

Short channel enhanced degradation during discharge of flash EEPROM memory cell

A new mode of channel length dependent degradation due to band-to-band tunneling current during the discharge of short channel flash memory device is discussed. The degradation is due to the holes created by band-to-band tunneling current and accelerated by the lateral electric field. The holes gain enough energy, are injected into the oxide and cause damage. The amount of degradation increases significantly as channel length decreases. This could be a fundamental limitation to the scaling of flash memory cells. A new discharge method is proposed to inhibit the hot hole injection and to obtain precise discharged Vt control.

Simulation of read-disturb lifetime reduction in submicron EPROM devices

An accurate and efficient physical simulation method, based on a 2-D energy transport model and Monte Carlo calculated hot electron energy distribution, has been used to simulate the read-disturb lifetime reduction of deep-submicron EPROM devices. This method has been validated by data of a 0.5 mu m EPROM technology and used to predict the read disturb lifetime for the 0.35 mu m technology generation.<<ETX>>