The new device, reported in the current issue of the journal Science, combines both of these approaches. A sharp tip is placed at the end of a tiny horizontal quartz tuning fork, similar to the ones used to regulate wristwatches. "This tuning fork vibrates at its own resonant frequency," says Markus Ternes, another member of the Almaden research team.

As the tip is brought into the proximity of an atom on a surface, forces acting between the two will alter the frequency at which the tuning fork vibrates. The new microscope measures changes in the piezoelectric currents generated by the vibrating quartz; like an STM, however, it also measures changes in a tunneling current flowing between the surface and the tip. The combination of techniques enables it to measure the forces exerted on individual atoms with unprecedented precision.

An atom hopping from one position to another on a surface is much like a sphere moving from one well in an egg carton to the next. By positioning the tip at various points around the atom, the IBM researchers take measurements of forces acting on it before it hops. It's a bit like holding one magnet close enough to another to feel a pull but not close enough to actually move it.

With the atoms, however, the IBM researchers keep moving the tip closer, says Heinrich. "At some point, the force is too large, and that is when the atom hops," he says. Although this happens too quickly to record, it is possible to calculate the force of the hop from measurements taken before it occurs.

The idea for the microscope's tuning-fork arrangement originally came from Giessibl. But it was only the collaboration of the IBM researchers that yielded a device with electronics and materials sensitive enough to detect the lateral forces on individual atoms. One of the reasons the microscope is so sensitive compared with regular AFMs is that the quartz material it uses is roughly 40 times stiffer than the silicon used in most AFM cantilevers.

The probe has already led to some unexpected results. For example, for some reason, moving a cobalt atom across a platinum surface requires more than 12 times as much force as moving it across copper surface. Moving a carbon monoxide molecule across copper, on the other hand, requires 10 times as much force as moving a single cobalt atom. Until now, scientists were unaware of any such nuances: "Before this, we were flying blind," says Heinrich.