Scanning tunneling microscopy just got faster
Characterization
January 2, 2008
 |
| RF-STM image of an atomic-scale step on a highly oriented pyrolytic graphite surface. Scan area is 205 nm x 205 nm and the image was acquired in 2 s at a tunnel resistance of 4 megaohms. Available bandwidth in this measurement is 2 MHz. (Credit: Kemiktarak et al.) |
Scanning tunneling microscopy, one of the key tools in nanoscience, has just taken a significant step forward in its development.
The atomic-scale resolution technique depends upon the tunneling of electrons between a sharp tip and a conducting specimen. However, STM is limited by low temporal resolution – or the time taken to acquire data – because of the response of the tunnel readout circuitry.
Various ways of circumventing this issue have been attempted, but now researchers from Boston and Cornell Universities report a simple means of vastly improving the speed of surface topography measurements, as well as enabling other more delicate ones [Kemiktarak et al., Nature (2007) 450, 85].
“The key to our method is a simple measurement scheme called impedance matching,” explains Kamil L. Ekinci of Boston University. “We use a so-called resonant circuit, which in some sense nulls out the parasitic elements and opens up a useful bandwidth.”
The tunnel junction is embedded in a resonant inductor-capacitor circuit, the reflection from which is measured, leading to a ~100-fold bandwidth improvement.
In constant-height mode, this radio-frequency (RF) STM technique has a measurement time at each scan point of ~0.1 µs. “Our hope,” says Keith Schwab of Cornell, “is that we can produce video images, as opposed to a scan that takes forever.”
Intriguingly, shot noise in the tunnel current can also be used to carry out nanoscale, rapid thermal imaging. For example, the researchers suggest that RF-STM could produce a 100 pixel by 100 pixel thermal image with ~1 K resolution at a temperature of 300 K in ~100 s.
Such a technique could be very useful for imaging semiconductor devices where heat flow is a crucial factor in performance and future development.
The technique can also be used as an ultrasensitive displacement sensor comparable to the best current optical and electrical detection techniques. The researchers detect high-frequency mechanical motion at a sensitivity of ~15 fm Hz–½ – and believe that RF-STM may be capable of quantum-limited position measurements.
“I firmly believe that ten years from now there will be a lot of RF-STMs around,” says Schwab.
Cordelia Sealy