Performance Analysis And Modeling Of Novel Nano-Devices

According to the Moore’s law. the size of semiconductor devices in integrated circuit is more and more small. There will result in serious short channel effects (SCEs) and the gate leakage current increases rapidly at the same time with the scaling of semiconductor technology. Therefore, researchers are exploring some new methods so that it can not only reduce the size of semiconductor device according to Moore’s law but also improve the performance of the device. Generally speaking, the metal-oxide-semiconductor field-effect-transistor (MOSFET) can be improved its performance based on four aspects:gate/gate dielectric engineering, channel engineering, source/drain engineering, novel devices. But the theoretical limit of the subthreshold swing (SS) is 60mV/dec at the room temperature, so it will restrict the power and leakage current decreasing continuously. Therefore, reasearchers explore tunnel field-effect-transistor (TFET) whose working principle is different from MOSFET. and its SS can achieve less than 60mV/dec, so it can widely be used in various low-power applications.In this paper, the performances and the models of these nano-devices are studied based on the double doping polysilicon gate (DDPG) MOSFET and double gate (DG) TFET. First, the transient characteristics of DDPG MOSFET are studied, and the radio frequency (RF) characteristics of DDPG MOSFET are studied by non-quasi static model. The previous model of DDPG MOSFET can be derived by solving the two-dimensional (2-D) Poisson’s equation and obtaining an analytic solution by generating a second-order (or higher) polynomial of the potential distribution function. However, it will produce physical inconsistencies because the 2-D nature of the gate insulator region is not considered, violating the boundary condition that the lateral field is continuous across two different dielectric boundaries. Therefore, one-dimensional treatment of the gate insulator region will yield a physical inconsistency. In this paper, a physical-based subthreshold potential model for DDPG MOSFET is proposed. The Steps of the 2-D semi-analytical potential model for DDPG MOSFET is:establish the coordinate system and divide the gate insulator area and the channel depleted region into three regions according to the structure of the gate. Two different rectangular sources and boundary conditions in the gate insulator area and the channel depleted region are defined. By using a semi-analytical method combined with an eigenfunction expansion method,2-D Poisson equations are solved. And then the expressions of the potentials are obtained, which are special function for the infinite series expressions. The potential distribution in the gate insulator layer obtained by our model is non-linear along the channel direction, and inclusion of two-dimensional treatment of the gate insulator region has resulted in a physical consistency across the dielectric boundary. From the surface potential profile, the current-voltage characteristics in the subthreshold region can be derived from both the drift-diffusion theory and the thermionic emission theory. According to the working principle of TFET, the 2-D semi-analytical potential model of DG TFET is also proposed. On the basis of potential profile, the shortest tunneling length and the average electrical field can be derived, and then, the drain current is constructed by using Kane’s model. DG TFET can surfer from hot carrier effects (HCEs) due to the presence of the strong electric field in the channel, especially in the tunneling region at the source end. Therefore, an analytical model for the surface potential of a DG TFET including the mobile charges has been developed in the presence of localized charges at the gate insulator and silicon film interface.The potential, the subthreshold drain current and threshold voltage of DDPG MOSFET, the potential and the drain current of DG TFET are simulated using SILVACO software. The calculated results are consistent with the simulated results. Device model can reflect well the characteristics of two nano-devices. The precision is relatively high, and the amount of computation is relatively small. The model can also avoid the shortcomings of the numerical analysis, i.e., the physical interpretation of the results obtained by numerical device simulator is not always clear.