Mesoscopic Modeling of High-Density Carbon Nanotube Films for Memristive Device Applications

TL;DR

The paper develops a mesoscale, coarse-grained framework to model high-density CNT films with tunable chirality, length, density, and amorphous carbon content for memristive devices. It constructs bead-based CNT networks, compresses them to target densities, and analyzes conduction via a nodal resistor network, linking transport to a small set of structural descriptors. Key findings show that local curvature and buckling promote conduction, while bundling hinders it, with amorphous carbon modulating morphology and transport in a configuration-dependent manner; PCA reveals a dominant structural axis that drives current. The work demonstrates how mesoscale modeling can guide the design of CNT-based memristors and provides a quantitative framework to map structure to electrical performance in complex CNT networks.

Abstract

Carbon nanotube (CNTs) materials, which exhibit intrinsically high electrical conductivity, are promising candidates for energy-efficient electronic devices. Recently, high-density CNT films have also been successfully employed as switching elements in non-volatile memory cells. However, the mechanism of electrical conduction through such complex systems is still poorly understood. To identify structural parameters that govern the electrical current in CNT films, we employed coarse-grained molecular dynamics to construct dense mesoscale CNT film models, where we considered CNTs with different chiralities and lengths. The effects of CNT geometrical features on the film morphologies were quantified by devising a set of structural descriptors and analyzing their mutual correlations. The impact of varying the concentration of amorphous carbon (aC) inclusions on the film structure was assessed. Finally, we employed a nodal analysis framework to compute the electrical current across the networks and correlate the charge transport characteristics to the underlying structural descriptors. Transport is found to be enhanced in films that exhibit high curvature and buckling, low bundling, and strong connectivity, with amorphous carbon components playing a nontrivial configuration-dependent role. These findings provide a framework for the rational design of CNT-based memristor architectures and highlight the potential of mesoscale modeling to guide the engineering of advanced nanostructured materials.

Figures (13)

• Figure 1: Atomistic models of aC (a) and CNT (32,0) segment (b) used to compute the Lennard-Jones parameters describing their interactions.
• Figure 2: Structures of single-layer films with a density of 0.6 g$\cdot$cm$^{-3}$ composed of (16,0), 100 nm-long CNTs containing between 10% to 40% amorphous carbon. CNT diameter is $d^{(16)} \simeq1.25~\mathrm{nm}$. CNT segments are shown in cyan and aC particles in yellow. (see Section \ref{['method_CNT_films']} for details)
• Figure 3: Structures of single-layer films with a density of 0.6 g$\cdot$cm$^{-3}$ composed of (32,0), 100 nm-long CNTs containing between 10% to 40% amorphous carbon. CNT diameter is $d^{(32)} \simeq2.5~\mathrm{nm}$. CNT segments are shown in cyan and aC particles in yellow. (see Section \ref{['method_CNT_films']} for details)
• Figure 4: Correlation matrices between the different descriptors and the currents.
• Figure 5: Principal Component Analysis of structural descriptors. Each point represents one structure. The projection is shown in the plane of the first two principal components, PC1 and PC2. Arrows indicate the directions of correlation with transport properties. Cluster A includes only films with 100 nm-long tubes containing aC. Cluster B contains only films with 15 nm-long tubes. Cluster C mainly consists of films with either 15 or 40 nm-long tubes, along with a few low-density films with 100 nm-long tubes. Cluster D consists mostly of films with 100 nm-long tubes without aC, except for structures #30 and #31.