Evolution of the absorbance spectra of a SWNT aqueous solution during the course of the reaction with methyl-oxycarbonylbenzene-diazoester. The S22 peak corresponding to absorption by semiconducting nanotubes is almost unaffected while the M11 transition peak drops.
Depending on their chirality (the angle between carbon atom chains and the tube axis at a given diameter), SWNTs can be either metallic, with a resistance as low as 6.5 kΩ for a diameter of 1 to 2 nm, or semiconducting, with a 0.4 to 1.2 eV bandgap depending on diameter. One major roadblock to the use of SWNTs in electronics relates to the fact that all synthesis methods produce both metallic and semiconducting chiralities as a mixture. Such inhomogeneity of properties at the material level has major deleterious effects since, for example, metallic nanotubes quench the luminescence of semiconducting ones in optoelectronic devices or form short circuits in electronic devices. It represents a well-identified pitfall for further development of SWNT-based industrial applications.
The LEM got interested in the SWNT-aryldiazonium coupling reaction years ago and studied its mechanism in details, in the aim of controlling its selectivity towards metallic nanotubes. For the first time, the reaction was proved to process through a complex radical chain reaction and the selectivity determining step was described in detail. Based on this work, we were then able to enhance significantly the selectivity of the coupling reaction (from 3 to 12 in terms of reaction rate ratio) by proposing a neighboring diazo reagent: diazoester. The selectivity was high enough to remove the metallic SWNT conductivity while preserving the semiconducting SWNT performances, so that separation was no longer necessary for most electronic applications. As a demonstration, we notably used the product of diazoester coupling directly for the fabrication of series of SWNT-network based field-effect transistors with improved performances.
Our process has been recently improved further by introducing another diazo reagent: the diazoether of ascorbic acid. The metallic vs. semiconducting differentiation in this case was efficient on all kinds of SWNTs of small and large diameters, and shows very good reproducibility with selectivity above 250. In addition, the preservation of the quality of sc-SWNTs is such that there luminescence (a very sensitive indicator of sc-SWNT quality) is almost unperturbed by the reaction.
This activity was primarily fundamental and focused on the detailed understanding of chemical reactions. Yet, we took great care to protect its outcomes through the filing of two international patents. First supported by the CEA Nanosciences Program, the initial results contributed to the success of an ANR proposal coordinated by the LEM since 2012. The exceptional carrier mobility in SWNTs makes this material appealing for the study of devices operating at high frequency. However, building HF-FETs from nanoscale objects with high impedance is extremely challenging. In the project NanotSConde, we push forward the chemistry-oriented studies and evaluate the potential of this reaction in the context of high-performance printed electronics.
G. Schmidt et al., "Mechanism of the Coupling of Diazonium to Single-Walled Carbon Nanotubes and Its Consequences." Chem. Eur. J. 15, 2101 (2009).
G. Schmidt et al., "Labile Diazo Chemistry for Efficient Silencing of Metallic Carbon Nanotubes." Chem. Eur. J. 17, 1415 (2011).
L. Darchy et al, "A highly selective non-radical diazo coupling provides low cost semi-conducting carbon nanotubes", Carbon (2013).
Maj : 10/12/2013 (2260)