In this paper, an L-shaped tunneling field effect transistor (LTFET) with ferroelectric gate oxide layer (Si: HfO₂) is proposed. The electric characteristic of NC-LTFET is analyzed using Synopsys Sentaurus TCAD. Compared with the conventional LTFET, a steeper subthreshold swing (SS = 18.4 mV/dec) of NC-LTFET is obtained by the mechanism of line tunneling at low gate voltage instead of diagonal tunneling, which is caused by the non-uniform voltage across the gate oxide layer. In addition, we report the polarization gradient effect in a negative capacitance TFET for the first time. It is noted that the polarization gradient effect should not be ignored in TFET. When the polarization gradient parameter g grows larger, the dominant tunneling mechanism that affects the SS is the diagonal tunneling. The on-state current (I_on) and SS of NC-LTFET become worse.

Fully transparent ITO/LaAlO₃/ITO structure RRAM (resistive random access memory) devices were fabricated on glass substrate, and ITO/Al₂O₃/ITO structure devices were set for comparison. The electrical characteristics of the devices were analyzed by Agilent B1500A semiconductor analyzer. Compared with the ITO/Al₂O₃/ITO RRAM devices, the current stability, SET/RESET voltage distribution, and retention characteristic of the ITO/LaAlO₃/ITO RRAM devices have been greatly improved. In the visible light range, the light transmittance of the device is about 80%, that of the LaAlO₃ layer is about 95%, the on-off ratio of the device is greater than 40, and the data retention time is longer than 10,000 s. The devices have great optical and electrical properties and have huge application potential as fully transparent RRAM devices.

In this paper, a novel ferroelectric-based electrostatic doping (Fe-ED) nanosheet tunneling field-effect transistor (TFET) is proposed and analyzed using technology computer-aided design (TCAD) Sentaurus simulation software. By inserting a ferroelectric film into the polarity gate, the electrons and holes are induced in an intrinsic silicon film to create the p-source and the n-drain regions, respectively. Device performance is largely independent of the chemical doping profile, potentially freeing it from issues related to abrupt junctions, dopant variability, and solid solubility. An improved ON-state current and I_ON/I_OFF ratio have been demonstrated in a 3D-calibrated simulation, and the Fe-ED NSTFET’s on-state current has increased significantly. According to our study, Fe-ED can be used in versatile reconfigurable nanoscale transistors as well as highly integrated circuits as an effective doping strategy.

This paper mainly studies the hardware implementation of a fully connected neural network based on the 1T1R (one-transistor-one-resistor) array and its application in handwritten digital image recognition. The 1T1R arrays are prepared by connecting the memristor and nMOSFET in series, and a single-layer and a double-layer fully connected neural network are established. The recognition accuracy of 8 × 8 handwritten digital images reaches 95.19%. By randomly replacing the devices with failed devices, it is found that the stuck-off devices have little effect on the accuracy of the network, but the stuck-on devices will cause a sharp reduction of accuracy. By using the measured conductivity adjustment range and precision data of the memristor, the relationship between the recognition accuracy of the network and the number of hidden neurons is simulated. The simulation results match the experimental results. Compared with the neural network based on the precision of 32-bit floating point, the difference is lower than 1%.