Spotting Very Thin Graphite Layers

Hank Hogan

Researchers are focusing on graphite monolayers, called graphene, because of their electrical and optical properties. In their quest to use graphene to build devices, however, they confront a basic problem. The simplest technique for producing the material involves scraping graphite on a silicon substrate covered with SiO2, but the resulting flakes vary in thickness. Only monolayer flakes are suitable for experimentation, but researchers have not had a good way to quickly pick out such pieces from the mass of much thicker scrapings.

Researchers have developed the means to measure the varying thicknesses of graphite sheets on a silicon substrate covered with SiO₂. Shown are a gradient of color versus graphene thickness (a), an optical microscope image of very large graphite flakes with regions of many thicknesses and, consequently, many colors (b), and the calculated reflection spectrum for ΔSiO₂ of 230 nm (c). Reprinted with permission of the American Chemical Society.

Now a discovery by a group of scientists at the National Enterprise for nanoScience and nanoTechnology at Scuola Normale Superiore in Pisa, Italy, may have given them a method that can identify a one-atom-thick layer of graphite with a normal microscope. “Optical microscopy would provide an extremely handy tool for a quick survey and identification of ultrathin graphite flakes,” said team member Stefano Roddaro.

The technique could lead to a reliable way to evaluate graphite flake thickness. According to the investigators, the method works because of the SiO₂ that is found on the silicon substrate. If it is thin enough, such a coating creates a distinct color that is related to the thickness of the oxide. The color results from interference between light reflected off the top of the oxide and light reflected off the silicon below. Thickness differences of a fraction of a wavelength lead to color changes that can be seen readily.

Graphene is far too thin for such an effect, yet even very thin graphite flakes have been reported to be visible. The researchers investigated this phenomenon by using an optical microscope from Nikon, an atomic force microscope from NT-MDT Co. of Moscow and a scanning electron microscope from Zeiss. They imaged graphite flakes on a silicon substrate covered by a layer of SiO₂ that was 500 nm thick. They found that the atomic force microscope produced the most precise information about graphite thickness but that optical methods could highlight very thin graphite.

By using classical electromagnetic theory and applying it to the air/graphite/SiO₂/silicon structure of their samples, they showed that the graphite modulates the interference between the light reflected from the SiO₂ and from the silicon. Changes of even a few nanometers in graphite thickness led to large changes in the interference, making the graphite visible because the observed reflected color also changed.

Their analysis also showed that a smaller numerical aperture resulted in greater graphene visibility, with the trade-off being a decrease in image details. Roddaro indicated that it would be necessary to optimize imaging performance and graphite contrast to get the best results. Other aspects of the system also could be tweaked. “In principle, there is space for improvement also on the substrate side,” he said.

The hope is that the technique eventually will lead to a dependable method for optically evaluating the thickness of ultrathin graphite down to single monolayers. For that to happen, however, there must be an accurate comparison between optical and atomic force microscopy imaging to ensure correct calibration, Roddaro noted.

Nano Letters, September 2007, pp. 2707-2710.