High-resolution sub-surface imaging of carbon nanotube (CNT) networks within polymer nanocomposites is demonstrated through electrical characterization techniques based on dynamic atomic force microscopy (AFM). We compare three techniques implemented in the single-pass configuration: DC-biased amplitude modulated AFM (AM-AFM), electrostatic force microscopy (EFM) and Kelvin probe force microscopy (KPFM) in terms of the physics of sub-surface image formation and experimental robustness. The methods were applied to study the dispersion of sub-surface networks of single-walled nanotubes (SWNTs) in a polyimide (PI) matrix. We conclude that among these methods, the KPFM channel, which measures the capacitance gradient (∂C/∂d) at the second harmonic of electrical excitation, is the best channel to obtain high-contrast images of the CNT network embedded in the polymer matrix, without the influence of surface conditions. Additionally, we propose an analysis of the ∂C/∂d images as a tool to characterize the dispersion and connectivity of the CNTs. Through the analysis we demonstrate that these AFM-based sub-surface methods probe sufficiently deep within the SWNT composites, to resolve clustered networks that likely play a role in conductivity percolation. This opens up the possibility of dynamic AFM-based characterization of sub-surface dispersion and connectivity in nanostructured composites, two critical parameters for nanocomposite applications in sensors and energy storage devices.

The morphological features of carbon nanotube (CNT) polymer composites and their influence on the effective modulus are evaluated. The considered features include bundle formation from the helical sub‐bundles made of individual CNTs. The formation of bundles is considered as a result of agglomeration of individual nanotubes above and below onset of percolation and is related to electrical conductivity. The proposed geometrical model yields a bundle diameter that agrees closely with that of the experimentally measured by voltage‐contrast method and scanning electron microscopy analysis of polyimide nanocomposites. The proposed micromechanical analytical model includes the helical structure of a bundle and provides close agreement of the effective Young's modulus of nanocomposite over a wide range of CNT content. It is shown that considering the helical structure of CNT bundles and its effect on bundle modulus is vital for predicting the effective modulus of CNT‐polymer nanocomposite.

Abstract