Dissertation Defense Committee:

• Ph.D. Advisor: Mario Lanza
• Committee Chair: Khaled Salama
• External Examiner: Bowen Zhu
• Committee Member: Jawaher Almutlaq

Abstract:

Two-dimensional (2D) materials have emerged as promising candidates for next-generation electronic and memory technologies due to their atomic-scale thickness, diverse electronic properties, and compatibility with aggressive device scaling. Despite rapid progress in materials synthesis and device demonstrations, the field continues to face major challenges related to the reliability, reproducibility. In particular, discrepancies between reported device performance and statistically representative measurements, combined with ambiguities arising from poorly defined device geometries and measurement conditions, limit meaningful benchmarking and technological translation. The overall goal of this dissertation is to develop and apply robust, reproducible, and industry-relevant characterization strategies for thin film dielectrics—such as hexagonal boron nitride (hBN)—to fabricate state-of-the-art memristors and transistors.

First, we address one of the most widely used nanoscale electrical characterization techniques: conductive atomic force microscopy (CAFM). While CAFM offers high spatial resolution, its quantitative reliability is often compromised by uncertainties in the effective contact area, environmental conditions, tip-related parameters, and limited current sensitivity. In this dissertation, a new CAFM technique is introduced using inert platinum nanodot electrodes fabricated via anodic aluminum oxide (AAO) shadow masks. This approach enables electrical measurements over well-defined, industry-relevant areas (down to 500 nm²) while retaining the versatility of CAFM. The method significantly reduces sensitivity to relative humidity, tip radius, and contact force, resulting in improved reproducibility across measurements and samples. It also allows the extraction of technologically meaningful figures of merit, such as leakage current density (below 0.01 A/cm²) and breakdown behavior, that are directly comparable to state-of-the-art transistor and memristor technologies.

Second, we focus on resistive switching devices based on hBN, with an emphasis on identifying and addressing common pitfalls in data collection. This work demonstrates that many previously reported performance metrics on large devices (areas ≥ 4µm²) on unfunctional SiO₂/Si substrates can lead to an over exaggeration of the materials properties. We show exceptional device performance (endurance up to 9 million cycles) for devices with an area of 4 µm² on SiO₂/Si, but when we fabricate smaller devices (area ≈ 0.05 µm²) on a Si microchip, further device engineering is required. We achieve high on-chip device endurance (up to 4 million cycles) by engineering the interface between the dielectric (hBN) and the metallic electrode via introducing limited-volume (1 nm-thick) silver layer to assist in the formation of the conductive filament.

Finally, we extend the theme of robust characterization to solution-processed 2D materials, with a focus on liquid-phase exfoliated MoS₂, a material that has exhibited great potential for memristive technologies. A systematic workflow is developed, starting from flake-level evaluation to spray-coated thin films and field-effect transistors (FETs). Electrical measurements are correlated with material quality and processing conditions, and chemical surface doping strategies are explored to enhance device performance. By combining multiple characterization techniques within a unified framework, this work demonstrates how reliable conclusions can be drawn for materials produced using scalable, low-cost methods, which lay the groundwork for future investigations in printed, flexible, and memristive electronics.

Overall, this dissertation presents an ultra-reliable protocol for measuring and characterizing novel materials and device structures with competitive and industry-relevant geometries.