Search

This paper presents an algorithm based approach for TFET (Tunnel Field Effect Transistor) design. A numerous number of meta-heuristic algorithms have been used to procure the best device dimension for a Si–Ge (Silicon–Germanium) pocket n-TFET. The foremost important task is to find an alternative to hit and trails based device optimization and thereby improve the device performance by using those techniques. The algorithm based approach requires an objective function. The surface potential based models efficiently represents the device physical properties, thus surface potential based model is used as an objective function. The impact of the channel length, of the Si–Ge layer and device thickness, as well as of oxide thickness are studied by considering them as design variables. The design process involves simulating and validating the obtained dimensions in Technology Computer Aided Design (TCAD). State of art techniques are being outperformed by this algorithmic approach and out of all applied algorithms the Human Behavior Particle Swarm Optimization algorithm (HBPSO) is more accurate. An ON-current of 4.8 × 10–4 A and OFF-current of 4.8 × 10–12 A is achieved by optimizing the structure. [ABSTRACT FROM AUTHOR]

The complex process of semiconductor device design requires precise models and efficient optimizers. This article puts forward an Asymmetrical Hetero-Dielectric (AHD) Triple Material Gate (TMG) n-type Junctionless Tunnel Field Effect Transistor (JL-TFET). A higher gate control is achieved by using triple material in control gate and hetero-dielectric oxide, which results in high ON current and low leakage. The surface potential based model for the proposed structure is derived by analytically solving 2-D Poisson's equation with hetero-dielectric gate oxide. This work also adopts intelligent techniques for extraction of optimal model parameters by using the derived mathematical model for the proposed JLTFET structure. The optimization technique used in this work combines the advantage of Particle Swarm Optimization (PSO) algorithm and Differential Evolution (DE) algorithm. A comparison with the conventional design process reflects that the use of optimization technique provides a novel approach to tune the process parameters. This technique outperforms the state of art design techniques and provides best accuracy along with exceptional computational efficiency. A current ratio of 1.25 × 1010 A and Point Subthreshold Swing (SS) values of 9 mV/dec and average SS of 48 mV/dec is achieved by optimizing the proposed structure. [ABSTRACT FROM AUTHOR]