2003 · 28 citations · 11 references
EngineeringElectrical EngineeringMaterials ScienceElectronic MaterialsMicroelectronicsSemiconductor DevicesDevice CharacterizationDevice ModelingDevice ReliabilityMetal-oxide-semiconductor Field-effect TransistorsOxide SemiconductorsSemiconductor ReliabilityBreakdown MechanismsGate Oxide DegradationGate Oxide ReliabilityGate Oxide Thickness EffectsLifetime EstimationMetal-oxide-semiconductor Field-effect Transistor ModelingOxide BreakdownOxide Reliability
We have performed TDDB measurements at temperatures lower than 125/spl deg/C in an electric field (E/sub ox/) range of 7-13.5 MV/cm and have evaluated the intrinsic lifetime in a wide electric field range, using both area and temperature dependences of oxide lifetime. For positive gate bias, log(t/sub BD/) of 7.1 and 9.6 nm oxides is not proportional to the electric field but is proportional to 1/E/sub ox/. This suggests that the breakdown mechanism of 9.6 nm oxide is the same as that of 7.1 nm oxide and adheres to the anode hole injection model. However, the breakdown mechanism of 4.0 nm oxide is not the same as that of 7.1 and 9.6 nm oxides. The slope of the log(t/sub BD/) versus 1/E/sub ox/ plot in 4.0 nm oxide increases with a decrease in the oxide field. The intrinsic lifetime under positive gate bias decreases with increasing oxide thickness, while the lifetime under negative gate bias increases with increasing oxide thickness in the range of electric fields employed in this experiment.
11