No CrossRef data available.
Published online by Cambridge University Press: 01 February 2011
Transmission electron microscopy (TEM) is a key technique in the structural characterization of carbon nanotubes. For device applications, carbon nanotubes are typically grown by chemical vapor deposition (CVD) on silicon substrates. However, TEM requires very thin samples, which are electron transparent. Therefore, for TEM analysis, CVD grown nanotubes are typically deposited on commercial TEM grids by post-processing. This procedure has two problems: It can damage the nanotubes, and it does not work reliably if the nanotube density is too low. The ability to do TEM directly on as-grown nanotubes lying on the silicon substrate would solve these two problems. In this work, for this purpose, we have fabricated micromachined TEM grids from silicon substrates. In particular, we have wet-etched large membranes from the back side of silicon wafers with a thin layer of thermal oxide on them. We have then etched a large array of long and narrow open slits on these membranes from the top side using a deep silicon etcher. Subsequently, we have grown nanotubes on these micromachined TEM grids by CVD, and characterized the nanotubes by high resolution TEM (HRTEM), micro-Raman spectroscopy, scanning electron microscopy (SEM), and atomic force microscopy (AFM). Since the nanotubes grown on the micromachined substrates are completely suspended over the width of the open slits, these substrates form a natural TEM grid for direct imaging of CVD-grown nanotubes. Furthermore, the signal from the substrate is significantly reduced during micro-Raman spectroscopy, resulting in a better signal-to-noise ratio. In addition, the silicon membranes are strong enough to support AFM and SEM characterization. As a result, these substrates provide a low cost, mass producible, efficient, and reliable platform for direct TEM, Raman, AFM, and SEM analysis of as-grown nanotubes or other nanomaterials on the same substrate, eliminating the need for any post-processing after CVD growth.