Speckle pattern analysis of PVK:rGO composite based memristor device

TL;DR

This paper introduces a speckle pattern analysis method to non-destructively characterize the resistive switching behavior of PVK:rGO memristors, providing insights into their conduction mechanisms during operation.

Contribution

It presents a novel optical speckle pattern technique for in situ, non-destructive analysis of memristor states, enhancing device characterization methods.

Findings

Speckle pattern variations correlate with device resistance states.

The method enables real-time, non-invasive monitoring of memristor switching.

Statistical analysis of speckle patterns reveals conduction mechanisms.

Abstract

The memristors are expected to be fundamental devices for neuromorphic systems and switching applications. For example, the device made of a sandwiched layer of poly(N-vinylcarbazole) and reduced graphene composite between asymmetric electrodes (ITO/PVK:rGO/Al) exhibits bistable resistive switching behavior. Depending on the resistance state of the (ON-state or OFF-state) at a constant applied voltage, it may show two different resistivities. The performance of the memristor can be optimized by controlling the doping amount of graphene oxide in the PVK polymer. To assess the performance of the device, when it switches between ON and OFF states, optical characterization approaches are highly promising due to their non-destructive and remote nature. Here, we characterize the memristor device by the use of speckle pattern (SP) analysis. The speckle pattern is the interference of multiple light waves with random relative phases, which is generated via different mechanisms such as scattering from diffusive materials. Therefore, SPs can be used to investigate such samples as they include a huge amount of information to be statistically elaborated. The experimental paradigm includes \textit{in situ} acquisition of SPs of the PVK:rGO in different states followed by statistical post-processing toward examining its conduction mechanism. The variations in these statistical parameters are attributed to the resistance state of the PVK:rGO samples under the applied voltage with regard to the physical switching mechanism of the device. The resistance/conduction state, in turn, depends on the activity and properties of PVK:rGO memristors as well as the additional non-uniformities induced through the variations of density of carriers. The present optical methodology can be potentially served as a bench-top device for characterization purposes of similar devices while they are operating.