Role of Point Defects\nand Ion Intercalation in Two-Dimensional\nMultilayer Transition Metal Dichalcogenide Memristors
Two-dimensional materials, in particular transition metal\ndichalcogenides\n(TMDs), have attracted a nascent interest in the implementation of\nmemristive architectures. In addition to being functionally similar\nto synapses, their nanoscale footprint promises to achieve the high\ndensity of a biological neural network in the context of neuromorphic\ncomputing. However, in order to advance from the current exploratory\nphase and reach reliable and sound memristive performances, an understanding\nof the underlying physical mechanisms in TMD memristors seems imperative.\nDespite the distinctive transport medium inherent to multilayer TMDs,\nthe memristance is routinely attributed to defects or metal atoms\npresent in the system, with their precise contribution remaining elusive.\nSpecifically, the role of intrinsic point defects in the formation\nof conductive channels, although shown for monolayer TMDs, is not\nconclusively studied for multilayer samples. In this work, using density\nfunctional theory (DFT) and nonequilibrium Green’s function\n(NEGF) formalism, a systematic study is carried out to analyze the\nimpact that defects and metal atoms produce on the out-of-plane conductivity\nof multilayer TMDs. MoS<sub>2</sub>, a representative of the 2H structural\nconfiguration, and PtS<sub>2</sub>, a representative of the 1T structure,\nthe most common crystal arrangements among TMDs, are used for this\nanalysis. It is found that the intrinsic sulfur vacancies, which are\nthe dominant defects present in both TMDs, appear to be insufficient\nin causing resistive switching on the application of an external bias.\nThe claim that the intrinsic point defects on their own can realize\na valence change memory-type device by providing a controllable conductive\nchannel through the van der Waals structure seems, according to our\nstudy, improbable. The presence of metallic atoms is demonstrated\nto be essential to trigger the memristive mechanism, emphasizing the\nproper choice of a metal electrode as being critical in the fabrication\nand optimization of memristors using TMDs.
Publication Details
- Journal
- Figshare
- Published
- 2024-10-24
- DOI
- https://doi.org/10.1021/acsanm.4c04769.s001
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00