A new symmetrical double gate nanoscale MOSFET with asymmetrical side gates for electrically induced source/drain
Introduction
The CMOS transistor gate length scaling is projected to continue through 2016 down to the incredible 9 nm [1]. Even if these dimensions can be realized using technological innovations, CMOS devices will suffer from a number of short channel effects, such as the threshold voltage roll-off, the drain induced barrier lowering (DIBL) and the subthreshold swing all of which degrade the MOSFET performance. A number of solutions have been proposed to overcome these problems [2]. Employing a double gate field effect transistor (DG MOSFET) structure instead of using bulk-Si transistors is one of these solutions. In addition to the inherent suppression of SCEs, DG MOSFETs offer high drive current and transconductance [3]. More importantly, the electrical coupling between the two gates results in high Ion/Ioff ratios when the threshold voltage is properly controlled [4], [5].
An ultra shallow extended source/drain is another effective method to suppress SCEs. But, it is very difficult to form shallow junctions by conventional fabrication techniques. However, it has been reported that SCEs can be suppressed by using an inversion layer as an ultra shallow extended S/D [6], [7], [8], [9], [10], [11], [12]. These devices are known as electrically variable shallow junction MOSFETs (EJ MOSFETs). In spite of the area penalty associated with these devices due to the additional side gates required for the inversion layer formation, EJ MOSFETs are expected to play a major role in the reduction of SCEs and hot carrier effects as discussed in our paper. To combine the advantages of both DG and EJ structures, in this paper we propose a novel symmetrical double gate nanoscale MOSFET with asymmetrical side gates for electrically induced source/drain regions. This structure is similar to that of a symmetrical DG MOSFET with the exception that the front gate consists of two asymmetrical side gates on both sides of the main gate. The aim of this paper is, therefore, to present the design and performance considerations of the proposed structure using two-dimensional simulation [13]. First, we have compared the performance of the EJ-DG structure with the symmetrical double gate (S-DG) structure in terms of threshold voltage, subthreshold swing, and electric field. Second, we have investigated the side gate design considerations in terms of threshold voltage, electric field, and hot carriers effects. Based on our simulation results, we demonstrate that the proposed EJ-DG structure is superior to the symmetrical double gate (S-DG) structure in controlling the SCEs.
Section snippets
EJ-DG structure
A schematic cross-sectional view of EJ-DG MOSFET implemented using the 2-D device simulator MEDICI is shown in Fig. 1. The front gate consists of two side gates using n+-poly and a main gate using p+-poly while the back gate is a p+-poly gate. The doping in the silicon thin film is kept at 1015 cm−3 (i.e., lightly doped with NA < 1016 cm−3 [14]) to avoid adverse effects associated with heavy doping, such as the mobility degradation [15] and the random microscopic fluctuations of dopant atoms [16].
Results and discussion
Fig. 2 shows a typical MEDICI simulated 2-D electron density distribution from source to drain of the EJ-DG structure for the main gate voltage VMGS = 0 V, the side gate voltage VSGS = 1.5 V and the drain to source voltage VDS = 0 V. As can be seen from the figure, the electron distribution in the channel region under the side gates decreases exponentially (linearly in the log scale) from the top of the silicon film to its bottom. Due to the difference between the work functions of the side gates and
Conclusions
In this paper, we have proposed a novel configuration for the symmetrical double gate SOI MOSFET using two side gates in order to investigate the influence of extremely shallow source and drain junctions on the short-channel effects. A constant voltage, independent of the main gate voltage, is applied to the side gates to form inversion layers acting as the extremely shallow virtual source and drain. Based on our simulation results we demonstrate that the combination of extremely shallow
References (21)
- et al.
Solid-State Electron.
(1994) - International SEMATECH, Intl. Technology Roadmap for Semicoductor (ITRS), Available at:...
- et al.
IEEE Trans. Dev. Mater. Reliab.
(2004) - et al.
IEDM Tech. Dig.
(1998) - et al.
IEEE Electron Dev. Lett.
(1994) - J.G. Fossum, Y. Chong, Proc. IEEE Int. SOI Conf., October 1998, pp....
- et al.
IEEE Trans. Electron Dev.
(2001) - et al.
IEDM Tech. Dig.
(1993) - W. Xusheng, Z. Shengdong, M. Chan, P. Chan, IEEE Conf. Electron Dev. Solid-State Circuits, 2003, pp....
- et al.
IEEE Electron Dev. Lett.
(1998)
Cited by (21)
Investigation of veritcal graded channel doping in nanoscale fully-depleted SOI-MOSFET
2016, Superlattices and MicrostructuresCitation Excerpt :Furthermore, SCEs leads to gate voltage controllability degradation to drain current which causes off current increment [5–7,9]. In the weak inversion regime, at source to channel junction is a potential barrier that control of its height is an important issue for leakage current improvement [4–7]. The potential barrier height for channel carriers should ideally be controlled by the gate voltage.
Analytical study of Dual Material Surrounding Gate MOSFET to suppress short-channel effects (SCEs)
2014, Engineering Science and Technology, an International JournalA 2D Transconductance and Sub-threshold behavior model for triple material surrounding gate (TMSG) MOSFETs
2013, Microelectronics JournalSOI MOSFET with an insulator region (IR-SOI): A novel device for reliable nanoscale CMOS circuits
2013, Materials Science and Engineering: BAnalytical threshold voltage model for cylindrical surrounding-gate MOSFET with electrically induced source/drain extensions
2011, Microelectronics ReliabilityCitation Excerpt :For the symmetry of the structure, the electrostatic potential and the electric field have no variation with angular in plane of the radial direction, hence a 2D analysis is sufficient (only channel radius direction r and channel-length direction z are employed). It was further found that when L1 = L2 = L3, the device can be operated in optimal condition [12–14]. Unless otherwise stated, this optimal configuration is applied in whole paper.
Cylindrical surrounding-gate MOSFETs with electrically induced source/drain extension
2011, Microelectronics JournalCitation Excerpt :Due to local carrier heating, carrier temperatures significantly differ from the lattice temperature resulting in inhomogeneous carrier temperature distributions in the channel region of the device. As a result, the carrier temperature gradients can make the thermal diffusion currents to be significant and together with the local carrier heating may influence the spatial carrier distributions and hence the thermal currents [13]. For this reason, The velocity saturation effects and the local carrier heating in the channel region must be considered.