Hydrothermally synthesized and exfoliated WS2-PEO nanocomposite based nonvolatile resistive devices for energy-efficient neuromorphic applications #MMPMID41360011
Das NS; Das S; Mayanglambam F; Dutta T
Nanotechnology 2025[Dec]; ? (?): ? PMID41360011show ga
This work presents a facile, scalable nanocomposite-based resistive memory device incorporating a 2D hybrid of hydrothermally synthesised and exfoliated tungsten disulfide (E-WS2) nanosheets embedded in a poly (ethylene oxide) (PEO) matrix for energy efficient neuromorphic applications. WS2 was synthesized via a simple, cost-effective hydrothermal method and subsequently exfoliated via liquid phase exfoliation (LPE) to obtain few-layer nanosheets with improved surface uniformity and reduced defect density. These nanosheets were integrated into the active layer of an ITO/E-WS(2)+PEO/Cu device fabricated via spin coating and thermal evaporation. The device exhibits reliable bipolar resistive switching with low SET voltages, a high ON/OFF current ratio (~10(4)), excellent retention (>450 s), and endurance over 70 cycles. The transport mechanism is governed by Ohmic conduction at low voltages, followed by space charge limited current (SCLC) and trap-controlled SCLC (TC-SCLC) mechanisms near-threshold voltages. Energy band analysis indicates that charge trapping and de-trapping at the WS2/PEO interface plays a critical role in the switching process.Compared to bulk WS2, exfoliated WS2 offers enhanced interfacial contact, lower resistance pathways, and reduced variability in switching, resulting in improved device performance and stability. It also shows more analog like behaviour. Sulfur vacancies in E-WS(2) assist in forming conductive filaments, while the PEO matrix enhances ionic mobility and switching behaviour. This work offers a scalable, environmentally benign approach to fabricating 2D material-based resistive memory, establishing solution-processed E-WS(2) nanocomposites as strong candidates 2 for next-generation, scalable, energy-efficient non-volatile memory and neuromorphic technologies.