Perovskite Quantum Dot-Based Memory Technologies: Insights from Emerging Trends

Abstract

Perovskite quantum dots (PVK QDs) are gaining significant attention as potential materials for next-generation memory devices leveraged by their ion dynamics, quantum confinement, optoelectronic synergy, bandgap tunability, and solution-processable fabrication. In this review paper, we explore the fundamental characteristics of organic/inorganic halide PVK QDs and their role in resistive switching memory architectures. We provide an overview of halide PVK QDs synthesis techniques, switching mechanisms, and recent advancements in memristive applications. Special emphasis is placed on the ionic migration and charge trapping phenomena governing resistive switching, along with the prospects of photonic memory devices that leverage the intrinsic photosensitivity of PVK QDs. Despite their advantages, challenges such as stability, scalability, and environmental concerns remain critical hurdles. We conclude this review with insights into potential strategies for enhancing the reliability and commercial viability of PVK QD-based memory technologies.

Keywords: CMOS process; halide materials; memristor; perovskite quantum dots; switching mechanism.

Figure 2

Schematic representation for the PVK QDs synthesis methods: (a) hot injection; (b) LARP; (c) solvothermal; (d) microwave-assisted method [89]; (e) ultrasonication method [90]; (f) microfluidic flow synthesis [91]; (g) laser in situ technique [92]; (h) spray synthetic procedure [93].

Figure 4

(a) Schematic demonstrating VCM mechanism by illustrating the iodide vacancies migration across the periphery of the perovskite octahedra [113]; (b) the ECM illustration exemplifying Ag filament formation and dissolution [118]; (c) double-filament model featuring Ag/MAPbI₃/FTO device having a thicker (upper) and thinner (lower) MAPbI₃ layers, showcasing the synchronicity of VCM and ECM within a single RS architecture [119].

Figure 8

(a) Diagrammatic representation of the memristive device and its corresponding top-view SEM image, inset: I–V curve for 50 cycles; (b) I–V behavior of the flexible device; (c) I–V characteristic of the device unsealed at 60% humidity and as a function of storage time; (d) I–V characterization conducted at various temperatures under normal conditions [142]; (e) optical photograph of the crossbar array architecture; (f) schematics of the device showing the materials involved; (g) I–V plots for Al/Cs_1−xFA_xPbBr₃ QD/ITO (x = 0.00 and 0.11) and Al/V₂O_5–y/PVK/Cs_1−xFA_xPbBr₃ QD/b-PEI/ZnO NCs/ITO (x = 0.11); (h) rewritten 28 × 28 contour images corresponding to the digit “3” concerning the number of learning epochs as a function of FA concentration (0.00 ≤ x ≤ 0.15) [152]; (i) cross-sectional FESCM; (j) retention observed at various states of the device; (k) schematics of the signal transmission process in biological neurons and LTP and LTD emulation [153].

Figure 1

(a) Crystal structure and nano morphologies representations [45]; (b) the TRPL decay spectra of CsPbX3 QDs (excluding CsPbCl₃ QDs; (c,d) the TEM images representing the nanocrystals of CsPbBr₃ [46]; (e) PL spectral survey and mixed UV and daylight image of PVK nanocrystals [47].

Figure 3

Schematics of memristors: (a) single cell configuration, (b) lateral configuration, (c) cross-point single device, and (d) crossbar array architecture.

Figure 5

(a) Perovskite-based random-access memory schematics. (b) Electrical switching mechanism. (c) A photo-assisted mechanistic approach is detailed, featuring four distinct states: (I) the initial state representing HRS: characterized by the presence of hole capturing reservoirs situated at the surface of the perovskite material; (II) the SET process: where corresponding traps become filled, shifting the Fermi level near to the valence band; (III) elimination of the light electricity: leading to the formation of a reduced barrier and quasi-ohmic (the LRS); and (IV) resetting process (electrically): extracting holes as of their trapping reservoirs, transitioning back to HRS [122].

Figure 6

(a) Diagrammatic representation of PVK QDs flexible memory architecture, inset: energy level presentation; (b) devices’ results showing erase–read–write–read; (c) I–V curve attained before and after 10th bending, inset: photographic illustration in bending state [137]; (d) graphics of the CsPbBr₃ QD-based RRAM, fabricated via solution processing; (e) the I–V characterization: inset (right) device photograph and (left) image showing PVK QDs solution under daylight and UV light illumination, respectively (f) pictorial representation of the CsPbBr₃ QD-based logic OR device (g) retention characteristics with and without light exposure [139]; (h) the TEM image of CsPbBr₃ QDs, inset: HRTEM and device structure; (i) RS behavior of the device, inset: corresponding image of the memory device [140].

Figure 7

(a) Schematic representation of the anatomy of the vision system featuring the LGMD, showing the photoreceptors arranged in packed hexagons within the locust’s compound eye, capturing light stimuli and delivering the visual signal through an electrical impulse through retinotropic layers comprised of lamina/medulla/lobula. Field A of LGMD correspond to receiving the feedforward excitation, and fields B and C correspond to the feedforward inhibition; (b) diagrammatic illustration of the TSM; (c) a unipolar TS behavior of 100 TSM cells; (d) illustration of a bird approach in the direction of the locust and images of the visual stimuli perceived by the locust, illustrating a steady rise in optical intensity from position A1 to A4; (e) the excitatory and the inhibitory responses received from the device to a looming light stimulus while concurrently receive the programmed electronic pulses (characterized at 0.2 V, a duration of 50 μs with and interval of 50 μs; (f) the schematic of a model car testing configuration, inset: provide a rear view of TSM on a printed circuit board; (g) illustration of decision making for the robot vehicle equipped with optical signal processability; (h) the upper part demonstrates the time evolution of f(t) functioned LGMD response, while the lower part statistically reveals the firing pulses at variable power supplies, ranging from 0.00 to 2.5 mW [59].