Courtesy of PhysicsWeb.

Stefan Hell and Marcus Dyba of the Max Planck Institute for Biophysical Chemistry used conventional optics to image clumps of bacteria just 33 nanometres across - equivalent to 1/23 of the wavelength of light used to illuminate them. The achievement shows that 'far-field' optical microscopes can operate well beyond the so-called diffraction limit without exploiting the quantum nature of light (M Dyba and S Hell 2002 Phys. Rev. Lett. 88 163901).

Scientists long believed that the maximum resolution of a microscope was about half the wavelength of the light used to illuminate an object - a constraint known as the diffraction limit. One way to improve the resolution of microscopes is to use radiation with shorter wavelengths - such as X-rays - but this method does not overcome the diffraction limit, and is unsuitable for some biological samples. 'Scanning probe' techniques - in which the sample is illuminated by a tiny light source - have recently reached high resolutions beyond the diffraction limit.

Now Hell and Dyba have combined two techniques to image bacteria labelled with an optically active dye in unprecedented detail. Both of these methods - which are known as stimulated emission depletion and 4Pi confocal microscopy - have shown that the diffraction limit can be beaten.

First, a laser pulse illuminates the smallest possible area - determined by the diffraction limit - of the bacteria sample, and excites the dye molecules. A second laser pulse then illuminates an area that partly overlaps with the patch of excited molecules. This forces the excited molecules in the overlapping region to emit their excess energy as light, and return to the ground state. Some time later, the remaining excited molecules relax naturally, emitting light from a region smaller than the diffraction limit would allow. Alone, this technique has achieved resolutions around one-tenth the wavelength of the light used.

But Hell and Dyba increased this resolution further by modifying the 'natural relaxation' step. When the dye molecules are about to emit light, two more laser pulses illuminate - but do not excite - the sample. These pulses are reflected by the sample and are then combined to produce a standing wave with a single central minimum. This standing wave acts as a filter and transmits only the light emitted by a tiny fraction of the dye molecules. In effect, the size of the light emitting area is reduced to just 1/23 of the wavelength of the laser pulse.

"This is the first time that a focusing light microscope has reached the tens of nanometres regime, which so far has been considered virtually impossible," Hell told PhysicsWeb.

Hell and Dyba used pulses of visible and near-infrared light, which were suitable for the energy gap in their dye. But they are optimistic that a resolution of around 17 nanometres could be reached in a system tuned to respond to ultraviolet light. According to Hell, the technique could be transformed into a practical device within two or three years. The researchers believe that such a device could be useful in microlithography and optical data storage, two fields in which physicists are trying to access ever-smaller length scales with visible light.