Dielectric properties of WS2-coated multiwalled carbon nanotubes studied by energy-loss spectroscopic profiling

Abstract

We investigate experimentally the electronic properties of the coating for multiwalled carbon nanotubes covered in tungsten disulfide ͑WS 2 ͒ of various thicknesses. Coatings of thicknesses between 2 and 8 monolayers ͑ML͒ are analyzed using energy-loss spectroscopic profiling ͑ELSP͒, by studying the variations in the plasmon excitations across the coated nanotube, as a function of the coating thickness. We find a change in the ELSP for coatings above 5 ML thickness, which we interpret in terms of a change in its dielectric properties. © 2005 American Institute of Physics. ͓DOI: 10.1063/1.1861985͔The discovery of carbon nanotubes 1 and fullerenes 2 has started a significant research effort into developing carbon electronics as a possible alternative to the end-of-the-road for silicon-based electronics, which is predicted to be reached in the next 10-15 years. Recent advances in controlling the growth of nanotubes 3 and the manipulation and welding of individual nanotubes 4 have brought the goal of producing working carbon devices within reach. Carbon nanotubes coated with functional layers are excellent candidates as starting blocks for these devices. The heterojunctions formed at multiwalled carbon nanotubes ͑MWCNTs͒ coated with WS 2 , a known bulk semiconductor, 5 could have applications in high-efficiency large-area photovoltaics and light-emitting structures. Their use, however, requires knowledge of the dielectric properties of the coating, as a function of its thickness, which we investigate here.This information can be acquired in a general-purpose transmission electron microscope ͑TEM͒ fitted with a Gatan imaging filter ͑GIF͒ using a technique known as energy-loss spectroscopic profiling ͑ELSP͒, 6,7 without the need for focusing the electron probe to subnanometer sizes. While the spatial sampling of the electronic structure information obtained in this manner is below that of a dedicated scanning TEM, it has significantly better spatial resolution than if the same TEM were used to form a small electron probe, whose diameter is not smaller than 1.5 nm in our case. 8MWCNTs coated with a varying number of ordered WS 2 layers were produced by pyrolysis; the production method and the characteristics of the nanotube and the coating are described elsewhere. 5 The tungsten disulfide coating appears to cover the nanotubes along their whole length. Small coating thickness variations ͓1-2 monolayers ͑ML͔͒ can occur along the length of the nanotube, so we only analyzed regions where the coating thickness was constant over lengths 5-10 times the nanotube's diameter. There was no detectable oxygen ͓using electron energy loss spectroscopy ͑EELS͔͒ in the coating. The samples were analyzed using a Philips CM200TEM with a LaB 6 filament, fitted with a GIF2000 spectrometer.When orienting a linear feature along the axis integrated by the quadrupoles of the GIF, the CCD records a twodimensional image where one coordinate is the energy loss and other is, to within a scaling factor, the spatial coordinate normal to the linear f...

Novel nanoscale architectures: coated nanotubes and other nanowires

“…In particular, Dr Vlad Stolojan at the University of Surrey has recently employed EELS to study the changes in the plasmon peak energy across the WS 2coated CNT as a function of coating thickness. Results showed that four monolayers of WS 2 are required in order that the coating approaches the WS 2 bulk dielectric state, for a WS 2 -coated MWCNT (Stolojan et al . 2004).…”

Energy loss spectroscopic profiling across linear interfaces: The example of amorphous carbon superlattices

“…It must be noted that here there is a physical limit on the thickness of the filler layer so that one can dielectrically define it as a layer with two surfaces, as opposed to a 'thick' interface between two media [38].…”