Abstract
The transport of electrons through assemblies of nanocrystals is critical for the performance of optoelectronic materials. While previous research has focused on single nanocrystals or pairs, there is a significant gap in understanding the collective behavior of large assemblies at the mesoscale. This study investigates variable range hopping (VRH) transport in disordered assemblies of touching, heavily doped ZnO nanocrystals. Using the Skinner, Chen, and Shklovskii (SCS) model alongside experimental conductive atomic force microscopy (CAFM) on ultrathin films, the researchers visualized current flow as a function of temperature. While experimental results largely aligned with the SCS model's prediction of a dominant "optimum band" for conduction, CAFM revealed additional dynamic subnetworks. These subnetworks connect and disconnect over time, representing a mesoscale manifestation of the "blinking" phenomenon.
Introduction: The Challenge of Collective Transport
Semiconductor nanocrystals (NCs) are foundational to next-generation technologies like solar cells and LEDs. However, bridging the gap between the physics of a single nanocrystal and the bulk properties of a material remains a challenge. In disordered assemblies, charge transport typically occurs via Efros-Shklovskii Variable Range Hopping (ES-VRH), where electrons "hop" between sites of varying energy and distance. Due to the structural randomness of these "piles" of particles, predicting and visualizing the actual pathways current takes is difficult. This work aims to map these pathways at the mesoscale—the intermediate regime where thousands of disordered particles interact.
Methods and Experimental Design
The researchers utilized a multi-pronged approach to study transport:
- Synthesis: ZnO nanocrystals were synthesized via plasma and deposited using supersonic impact to create dense, ultrathin films only a few particle layers thick.
- Control: Atomic Layer Deposition (ALD) was used to coat NCs with ZnO or $Al_2O_3$, allowing precise tuning of the contact radius and carrier concentration.
- Modeling: A 1000-particle assembly was simulated using Brownian dynamics and treated as a resistor network based on the SCS model.
- Visualization: Conductive Atomic Force Microscopy (CAFM) mapped local resistance and current flow from 38 K to 296 K with nanometer precision.
Key Insights: The "Blinking" Backbone
The study yielded several critical findings regarding how current traverses disordered networks:
- Beyond the Optimum Band: The SCS model predicts conduction is dominated by a single "optimum band" or backbone of low-resistance sites. Experiments confirmed this but also identified "subnetworks"—secondary peaks in resistance histograms representing higher-resistance regions that still contribute to flow.
- Mesoscale Blinking: A surprising discovery was the temporal instability of these subnetworks. Large regions of the film were observed joining and disconnecting from the main conduction network. This dynamic behavior is a collective, mesoscale version of the "blinking" often seen in individual quantum dots.
- Tunable Transport Regimes: By adding just a few cycles of ALD ZnO, the researchers could transition the material from an insulator-like hopping regime to a metallic-like transport regime, demonstrating significant control over electrical properties.
Results and Quantitative Agreement
The experimental data showed strong quantitative agreement with theory. The temperature dependence of the peak resistance in CAFM maps (slope = 10.6) closely matched the SCS model's prediction (slope = 13.9). Furthermore, the CAFM maps successfully visualized the "percolation pathways," providing a direct look at how charge navigates the disordered maze of particles.
Conclusion and Future Directions
While the SCS model provides an excellent framework for understanding mesoscale transport, it does not currently account for the dynamic "blinking" of subnetworks. These tenuous, high-resistance connections are highly sensitive to environmental changes or charge states. Future work will focus on incorporating these temporal dynamics into transport models. Understanding these shifting connections could pave the way for more stable nanocrystal devices or even neuromorphic hardware that mimics biological neural networks.
Summary updated