Deep surface trenches less than 10 nm across are difficult to image faithfully because they are smaller than most probes and because of the sharp step edges that disturb the AFM. Indeed, the images obtained are distorted without the user even being aware of it. Now, Solares suggests that these image distortions could potentially be eliminated by combining frequency-force modulation AFM, intelligent scanning algorithms and fine nanotube probes.

"This is important because most routine AFM users rely on the accuracy of their images to draw scientific conclusions," he told "If the accuracy of these images is questionable, then the conclusions derived from them may also be questionable." Although crucial for characterizing ever-smaller semiconductor features, enhancing AFMs in this manner may also be useful for imaging biomaterials with molecular resolution for solving problems in fields like cancer research, genomic and cell biology, he adds.

In the new method, a carbon nanotube is attached to a cantilever tip oscillating in "intermittent-contact frequency-force modulation AFM". In this technique, the effective frequency of the oscillating cantilever increases or decreases depending on whether the tip is tapping too hard, or not hard enough, on the sample.

The user selects a frequency slightly above the cantilever's natural frequency, which signals the AFM to adjust the excitation signal so that the cantilever only gently taps down on the surface. The tip repeatedly slow downs and re-approaches the sample to make sure that the step edges are properly imaged. As many researchers have already shown, carbon nanotubes are ideal AFM tips thanks to their small diameters and high aspect ratios, which means that they can easily image surface crevices or pores.

If demonstrated in an experiment, the technique could create accurate, high-resolution images of sharp surface features, like trenches, pores and thin protrusions on surfaces.

Solares would now like to do just this. "I also plan to develop quantitative methods to estimate the nanotube-sample interaction forces as a function of the deformation induced during measurement – from which sample mechanical properties can be obtained. I am being helped here by graduate students Joshua Crone and Gaurav Chawla."

The work was reported in Meas. Sci. Technol.