Resistive Switching in CaF2-MoS2 Nanocomposite thin films for low energy memory and neuromorphic applications

Abstract

Cognitive computing paradigm is an innovative field of research contrary to the conventional methods, seeking inspiration from human brain. The limitations of von Neumann architecture can be overcome by replacing the conventional computer architecture with artificial synapse-based techniques that are inspired by human brain composed of neurons and synapses. In this study, we report the fundamental synaptic performance of CaF₂-MoS₂ nanocomposite film. The electronic conduction exhibited by the Al/CaF₂-MoS₂/Al device is attributed to via the charge trapping and detrapping at the interface. The device can be operated with low energy consumption/spike ($\sim$ nJ) that varies linearly with the voltage pulse width. Device exhibits excellent synaptic functionalities with stable potentiation and depression curves lasting multiple cycles. Each potentiation and depression curve can be fitted with a bi-exponential function indicating two dominant processes affecting synaptic current. Further the synaptic amplification can be tuned by using voltage pulses of different frequencies. Spike rate dependent plasticity (SRDP) shows a cut-off frequency beyond which the synaptic amplification occurs in the devices. This work proposes a new avenue for CaF₂ based artificial synaptic device which can be utilized for multifunctional applications in future. The work also details the importance of tuning various parameters for electrical stimulation such as pulse voltage and pulse width to get optimum performance of artificial synapse in such nanocomposite based devices.