Effect of Barrier layer thickness on device performance of AlInN/GaN Underlap DG MOSFET

We analyze the influence of Al 0.83 In 0.17 N barrier layer thickness (T B ) on device performance of 18nm gate length ultra thin body AlInN/GaN heterostructure underlap DG MOSFET, using 2D Sentaurus TCAD simulation. The device is designed according to the ITRS specifications and simulation is done...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Pardeshi, H., Sarkar, A., Mohankumar, N., Sarkar, C. K.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We analyze the influence of Al 0.83 In 0.17 N barrier layer thickness (T B ) on device performance of 18nm gate length ultra thin body AlInN/GaN heterostructure underlap DG MOSFET, using 2D Sentaurus TCAD simulation. The device is designed according to the ITRS specifications and simulation is done using the hydrodynamic model. The simulation is validated with previously published experimental results. Very high drain current density (~8.8 mA/μm) is achieved, due to high values of two-dimensional electron gas (2DEG) density and velocity. Simulation of major device performance parameters such as DIBL, SS, delay, threshold voltage (V t ), ON current, energy delay product and total gate capacitance C gg have been done for T B ranging from 0nm to 4nm. As T B is increased the drain current increases and delay decreases, but at the expense of loss of electrostatic control leading to increased short channel effect i.e. higher DIBL and SS. Also, negative shift in threshold voltage is observed for rising T B . Decrease in C gg is observed as T B increases, due to increase in separation between the gate and channel, leading to reduced gate control. There is tradeoff between achieved drain current and electrostatic control for varying T B Thus, the selection of appropriate T B is of vital significance as it determines the device performance.
DOI:10.1109/CODIS.2012.6422233