Ag/TiO <sub/>2</sub> NPs/TiO <sub/>2</sub> TF/Si Based Non-Volatile Memristor Device for Neuromorphic Computing Applications

摘要： Memristor device is a very promising emerging component for a revolution of the memory and computing applications in the recent years. It could be enhancing the field of artificial intelligence and helping the patients, suffering from various kinds of autism disorders, as well as in neuromorphic computing, neural networks, etc. This research article proposes fabricated non-volatile memristor device for neuromorphic computing applications. The demonstrate memory is based on Ag/TiO2 NPs/TiO2 TF/Si layers’ structure and achieves better conductivity and storage capacity, which could improve the performance of the neuromorphic computing as compared to conventional ones. The fabrication method of the proposed multi-layer memristor is examine with well precise techniques, which overcome the previous challenges. The surface morphology of the device is analysed by field emission gun scanning electron microscopy (FEGSEM) and Energy dispersive X-ray system. The rise time (Tr) of 2.5 s and fall time (Tf) of 3 s are demonstrated under ON/OFF white light illumination. While X-ray diffraction depicted that titanium dioxide nano particle, (TiO2 NPs) are crystalline in nature. Moreover, Photoluminescence and Raman analysis justify crystalline nature also and increasing oxygen vacancies concentration with the heating effect of TiO2 NPs. The electrical analysis reveals the driving mechanism under different sweeping voltages during SET and RESET resulting in low resistance state (‘ON’). Finally, capacitance-voltage characteristic of the proposed memory device shows excellent charge storage capacity within the dynamic range of operation.