Device modelling of archimedean spiral graphene nanoscroll field-effect-transistor

Abstract

For the past decades, researchers indicate that persistent scaling of conventional silicon Metal-Oxide-Semiconductor Field Effect Transistor (MOSFET) reaching its physical limit at 10nm, resulted in its performance degradation as the search continues for a low-power and high speed, density and reliability devices. The frailty due to the Short Channel Effects (SCE) has limited the device scaling. In addition, the emergence of Carbon Nanotube (CNT) in the past two decades has been a remarkable breakthrough in solving for transistor SCE; but there has also been a problem in controlling its band gap energy. Graphene Nanoscroll (GNS) is one of the carbon-based materials that inherit most likely similar electrical properties as CNT. But GNS possesses an advantage to modulate its properties by varying its carbon layer overlapping region owing to the open edge spiral, resulting in band gap variations. This study is to investigate the GNS carrier statistic against its geometry variation and its performance as a MOSFET. The carrier statistic such as the energy band gap, density of states, carrier density and intrinsic velocity were modeled and the results show strong relation to the overlapping region of GNS. The energy band gap exhibits an inverse relation to the overlapping region and metallic properties was restored when the overlap has reached certain limit. The carrier density also increases with the overlapping region as a sign of gap narrowing. Moreover, the intrinsic velocity increases with overlap region and remains constant as it reaches graphene Fermi velocity, signifying ballistic transport near Fermi point. The charge distribution in GNSFET was characterized based on the Landauer Buttiker’s formalism. The output current shows good agreement with the experimental results at constant conductance and GNS structural parameters. Furthermore, the GNSFET demonstrated comparable performance to the CNTFET within ballistic limit. The GNSFET was also benchmarked with the latest 22nm MOSFET technology, which indicates faster switching capability due to enhancement in Subthreshold Swing (SS) and Drain Induced Barrier Lowering (DIBL).