The human brain's efficiency inspires the development of artificial synapses using solid-state materials. However, limitations in mimicking neurotransmitter movement have led to interest in soft memory devices. These devices leverage ion concentration polarization for smoother neurotransmission. Resistive switching memory (RSM) devices are promising due to their simple fabrication, stability, and fast switching. Liquid-based RSMs, using electrolytes, liquid metals, or ionic liquids (ILs), offer flexibility, conductivity, low cost, and ease of fabrication. They operate on the principle of cation and anion transport, exhibiting electrochemical metallization and filamentary conduction. This paper focuses on an Ag@AgCl core-shell (liquid electrolyte) synapse device demonstrating multistate resistive switching behavior. The core-shell structure controls the movement of Cu²⁺ ions, leading to improved device performance and stability at low voltages.

The paper cites previous research on resistive switching memory (RSM) devices and their applications in neuromorphic computing. It mentions the work of Leon Chua and Hewlett Packard Laboratories in the development of RSMs. The literature review highlights the importance of mimicking the brain's synaptic processes, emphasizing the role of neurotransmitters (Na⁺ and K⁺ ions) and the electrical activities of neurons in learning processes. It also discusses the use of various liquid materials in RSMs to improve flexibility and conductivity and the electrochemical metallization behavior leading to filamentary conduction in these devices.

The study involved synthesizing an Ag@AgCl core-shell IL ink, characterized using FESEM, EDS, FTIR, and XRD. FESEM analysis showed an average Ag@AgCl size of 100–120 nm. EDS confirmed the presence of Ag, Fe, and Cl. FTIR identified various functional groups. XRD analysis showed the crystalline structure of Ag and Ag@AgCl. The resistive memory behavior was investigated using Cu/FeCl₃/Cu and Cu/Ag@AgCl/Cu devices. The ionic transport mechanism in FeCl₃-based ILs was analyzed, observing the movement of Fe²⁺ and Cu²⁺ ions. The memristive behavior was characterized by analyzing current-voltage (I-V) curves, charge-flux characteristics, and endurance cycles. Multistate resistive switching was achieved by applying dual voltage sweeps and pulse sequences of varying voltages, widths, and frequencies to the Cu/Ag@AgCl/Cu device. The device's synaptic behavior was evaluated by mimicking spike rate-dependent plasticity (SRDP).

The research successfully demonstrated a core-shell soft ionic liquid resistive memory device with multistate resistive switching behavior. The Ag@AgCl core-shell effectively controlled the movement of Cu²⁺ ions, leading to stable and repeatable multistate switching. The device exhibited a high OFF/ON resistance ratio (~10.5:1). The I-V characteristics showed a pinched hysteresis loop. Time-domain charge and flux characteristics were asymmetric, indicating non-ideal memristor properties. The device demonstrated stable repeatability over 100 cycles. The incorporation of Ag nanoparticles into the FeCl₃-based electrolyte resulted in the formation of the Ag@AgCl core-shell structure, crucial for achieving multistate switching. The device's behavior was consistent with the electrical properties of memristive devices used as synapses. Applying various pulse schemes mimicked the spike rate-dependent plasticity (SRDP) observed in biological synapses, further validating the device's potential as an artificial synapse. The synaptic weight could be modulated by changing the pulse width, with wider pulses resulting in greater current changes due to enhanced ionic diffusion.

The findings demonstrate the feasibility of using a soft, flexible core-shell ionic liquid-based device as an artificial synapse. The multistate resistive switching behavior closely mimics the synaptic plasticity observed in biological systems. The ability to modulate the synaptic weight by controlling pulse width further strengthens the device's potential for neuromorphic computing applications. The improved control over ion movement achieved through the core-shell structure addresses a key limitation of previous liquid-based RSM devices. This research contributes to the development of energy-efficient, flexible electronic synapses, paving the way for more sophisticated neuromorphic computing systems.

This study successfully fabricated and characterized a core-shell ionic liquid resistive memory device that exhibits multistate resistive switching and mimics the behavior of biological synapses. The Ag@AgCl core-shell structure plays a critical role in controlling ion movement and enabling the observed multistate switching. The device's ability to emulate spike rate-dependent plasticity opens exciting possibilities for the development of next-generation neuromorphic computing systems. Future work could focus on optimizing the device's performance and exploring its integration into more complex neuromorphic architectures.

The study focused on characterizing the device's fundamental behavior under specific experimental conditions. Further research is needed to assess the device's long-term stability, reliability, and performance under varying environmental conditions. The scalability and manufacturability of the device for practical applications also need to be investigated. A more comprehensive investigation of the device's performance across a wider range of stimuli and conditions is also warranted.