Covalent Chemistry on Carbon Nanotubes: From Electronic Fundamentals to Sensor Applications

Abstract

Carbon nanotube devices are particularly well-suited to build chemical or biological electronic nanosensors due to their inherent nanoscale channel, exceptional electrical conductance and high sensitivity to charge transfer. This charge sensitivity is mostly unselective though, which means that functionality must be added to the nanotube sidewall in order to tailor its affinity to specific chemical species. Single-point functionalization is particularly desirable to allow probing molecules at the individual level. Among the various functionalization types available for carbon nanotubes, covalent chemistry provides the most robustness and reproducibility. However, its invasive nature is known to alter the electronic performance of the nanotubes, and achieving single-point covalent binding on pristine nanotubes is challenging. Here we present experimental work providing fundamental insight of the impact of covalent reactions on carbon nanotubes electronic properties, as well as recent advances on the use of covalently chemistry for assembling single-molecule nanosensors. First, electrical transport experiments are used to probe the electronic states of carbon nanotubes fully covered with covalent adducts. Results on numerous individual nanotube devices show that addition of monovalent groups such as aryl derivatives severely disrupt the nanotube electronic bands and also generate graft-induced localized states in the nanotube band gap [1]. Oppositely, divalent grafting using carbene-based addition reactions is found to leave the nanotube electronic properties unaltered [2]. We discuss the mechanisms behind these results based on symmetry and conjugation considerations. Second, high-resolution lithography patterning and aryldiazonium chemistry are used to add covalent adducts on various portions of carbon nanotubes devices, from several microns down to 20-nm segments. The intensity of conductance alteration is found to scale exponentially with the length of the exposed segment, with large variations in decay constants between devices. All devices nevertheless present a robust 20% current drop for the shortest exposed segments, which points to a consistent small number of binding sites. Finally, we demonstrate the ability of this approach to bind few molecules onto carbon nanotubes with controlled position and high yield over arrays of hundreds of devices, which opens a very promising route for assembling a variety of carbon-nanotube-based single-molecule electronic nanosensors. [1] D. Bouilly, J. Laflamme-Janssen, J. Cabana, M. Côté, R. Martel, under revision (2014) [2] D. Bouilly, J. Cabana, R. Martel. Appl. Phys. Lett. 101, 053116 (2012)