Fine-Tuning of Single-Walled Carbon Nanotube Properties for Transparent Conductive Applications

Abstract

Single-walled carbon nanotubes (SWCNTs) are a unique material for next-generation electronics, thanks to their outstanding physical properties. In particular, SWCNT-based transparent conductive films (TCFs) are of great interest for replacing indium-tin-oxide in touch screens, displays, and solar cells with a technology development paradigm shifting to stretchable and wearable devices. However, recent advances in the field of SWCNT thin-film production, particularly material performance and synthesis productivity, are insufficient for pushing SWCNTs into the industry. This fact highlights the relevance of research on the efficient synthesis of SWCNTs with tailored characteristics, which is the main focus of the current thesis. Two different aerosol chemical vapor deposition (CVD) reactors based on carbon monoxide disproportionation and hybrid hydrocarbon pyrolysis were thoroughly investigated and used to tackle the challenge of highly conductive SWCNT film synthesis with desired productivity.

The systematic study of the carbon monoxide-based system allowed to associate SWCNT structural properties and optoelectrical performance of their films, providing a direction for further optimization. In particular, low defectiveness, thick nanotube diameter, and high length corresponded to the highest optoelectrical performance of the films. The engineering of gas flows in the reactor was carried out to enhance material performance and synthesis productivity. Thus, the investigation of catalyst injection strategy reinforced with CFD simulations helped to reach a 9-fold enhancement of synthesis yield, preserving SWCNT characteristics. Besides, the residence time adjustment was employed for lengthening the produced SWCNTs. Despite its positive effect on length and film sheet resistance, high residence time resulted in elevated diffusion losses on the reactor walls limiting the process productivity. For post-synthesis enhancement of SWCNT optoelectrical properties, a reversible doping technique based on electrochemical gating of the films was developed. The method was evaluated on the samples produced by the carbon monoxide-based reactor, which exhibited a 13-fold drop in sheet resistance at 90% transmittance down to 53 Ω/□.

A more complex and multiparametric hydrocarbon-based synthesis, implementing toluene and ethylene as feedstocks, was employed with a focus on overcoming performance-productivity trade-off. The adjustment of synthesis parameters resulted in sheet resistance (at 90% transmittance) value of 57 Ω/□ and yield of 0.24 cm2·L-1, outperforming previous advances. The multiparametric experimental data was also used in the comparative analysis of machine-learning algorithms and their applicability for predicting SWCNT properties. Among the tested algorithms, artificial neural networks were found to provide the lowest error comparable with experimental inaccuracy, coping well with the prediction of sheet resistance.