SLD-MOSCNT: A new MOSCNT with step–linear doping profile in the source and drain regions

Abstract

In this paper, a new structure, step–linear doping MOSCNT (SLD-MOSCNT), is proposed to improve the performance of basic MOSCNTs. The basic structure suffers from band to band tunneling (BTBT). We show that using SLD profile for source and drain regions increases the horizontal distance between valence and conduction bands at gate to source/drain junction which reduces BTBT probability. SLD performance is compared with other similar structures which have recently been proposed to reduce BTBT such as MOSCNT with lightly-doped drain and source (LDDS), and with double-light doping in source and drain regions (DLD). The obtained results using a nonequilibrium Green’s function (NEGF) method show that the SLD-MOSCNT has the lowest leakage current, power consumption and delay time, and the highest current ratio and voltage gain. The ambipolar conduction in the proposed structure is very low and can be neglected. In addition, these structures can improve short-channel effects. Also, the investigation of cutoff frequency of the different structures shows that the SLD has the highest cutoff frequency. Device performance has been investigated for gate length from 8 to 20 nm which demonstrates all discussions regarding the superiority of the proposed structure are also valid for different channel lengths. This improvement is more significant especially for channel length less than 12 nm. Therefore, the SLD can be considered as a candidate to be used in the applications with high speed and low power consumption.

An efficient structure for T-CNTFETs with intrinsic-n-doped impurity distribution pattern in drain region

“…Ambipolar behavior and BTBT which occur at gate reverse voltages (i.e., at high leakage current) are among the important issues for MOS-CNTFETs [12,14], and should be well addressed in simulations. Some researchers examined these issues and suggested several methods [11,12,[15][16][17][18][19][20][21][22][23][24]. These methods include the change in impurity concentration [11,[15][16][17], bandgap engineering [12], use of linearly and lightly doped impurity injection in source and drain regions [11], underlap engineering [18,20], and use of gate insulators with asymmetrical thicknesses [19], and so on.…”

“…Some researchers examined these issues and suggested several methods [11,12,[15][16][17][18][19][20][21][22][23][24]. These methods include the change in impurity concentration [11,[15][16][17], bandgap engineering [12], use of linearly and lightly doped impurity injection in source and drain regions [11], underlap engineering [18,20], and use of gate insulators with asymmetrical thicknesses [19], and so on. To improve power consumption and save energy, the T-CNTFETs were proposed [14,24,25].…”

Design and Optimization of Lightly Doped CNTFET Architectures Based on NEGF Method and PSO Algorithm

“…20 Recently, a combination of the linear doping and the uniform doping in the lightly doped regions has been proposed in Ref. 21. Although many researches have worked on the improvement of the CNTFETs characteristics by using appropriate lightly doped region length and the doping level, but the optimization of this type of transistors incorporating the optimization algorithm has not yet been done.…”

Electrical Characteristics of T-CNTFET: Partially-Gated Channel vs Doping Engineering

“…Their 1D structure, as well as their promising characteristics, makes the CNTFETs have a pronounced interest for future nano-scaled device applications. [1][2][3][4][5][6][7] Therefore, it is expected that T-CNTFETs become an alternative to the conventional MOSFET CNTFETs. On the other hand, the conduction of current, for both high positive and negative gate voltages, which is called ambipolarity, is the major drawback of the p-channel and n-channel CNTFETs, respectively.…”