IEEE Transactions on Electron Devices · 2012 · 21 citations · 23 references
EngineeringElectrical EngineeringMaterials ScienceMicroelectronicsSemiconductor DevicesSemiconductor MaterialsCarrier TransportDopingMetal-oxide-semiconductor Field-effect TransistorsSemiconductor InterfacesSemiconductor PhysicsWide Band Gap SemiconductorsCharge NeutralityCharge TrapsConduction BandElectron MobilityInterface StatesInterface Trap DensityInterface Trap StatesInterface Traps
The effects of interface-trap states <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$(D_{\rm it})$</tex></formula> and the shift of the charge neutrality level (CNL) on the enhancement of the drain current in In-rich surface-channel enhancement-mode n-type InGaAs MOSFETs are investigated. In addition to the increase in the bulk mobility, the shift of the CNL toward the conduction band together with high densities of <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> is responsible for the experimentally observed remarkable enhancement of the on-state drain current with increasing In content in the channel. However, when <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> is low, current enhancement is weak, and the location of the CNL has little effect on the current enhancement. Acceptor-type interface-trap states above the conduction-band minima (CBM) play an important role in determining the inversion-layer electron mobility. Representing <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> distribution above the CBM by a constant equal to the <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> value at the CNL causes an overestimation of the drain current at higher gate voltages. It is also observed that the extraction of <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> from the low-frequency gate <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$C$</tex></formula> – <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$V$</tex></formula> data is independent of the location of the CNL. We further show that the subthreshold slope (SS) is doubled due to <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> . However, the location of the CNL or the magnitude of <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$D_{\rm it}$</tex></formula> above the CBM has little effect on the SS.
23