A pair of physicists has found a way around a fundamental optical limit that blinds light microscopes to details smaller than a few ten-millionths of a meter. The new technique may eventually enable biologists to see the innermost workings of living cells.
It has been an unavoidable consequence of basic optics for more than a century: A light microscope can't reveal objects much smaller than half the wavelength of the light passing through its lens. As a result, optical microscopes blur details such as the structure of a cell's nucleus. Researchers can capture these nanometer-sized features with an x-ray or electron microscope, but using either device requires killing the organism.
Now researchers have shown that it's possible to improve the resolution of an optical technique called confocal microscopy more than 10-fold. In ordinary confocal microscopy, a lens focuses a pulse of laser light into a small spot in a fluorescent sample. Light from the fluorescence then passes back through the lens to a detector. As a sample is moved in front of the lens, the variations in the fluorescence produce a three-dimensional image of it with slightly better resolution than the diffraction limit.
The key advance by physicists Stefan Hell and Marcus Dyba of the Max Planck Institute for Biophysical Chemistry in Göttingen, Germany, was to drastically shrink the fluorescing spot. They achieved this by installing a second lens directly behind the sample, and as soon as the sample was excited by the first pulse, they sent a pair of intense pulses through the two lenses. These essentially squeezed the light out of the excited molecules through a process called stimulated emission. The pulses cancelled one another in the middle of the sample, leaving only a very narrow slice of the sample to fluoresce into the detector. The narrowness sharpens the resolution of the resulting picture. To prove their technique worked, Dyba and Hell took an image of the 30-nanometer-thick membrane of a bacterium, as they report in the 22 April issue of Physical Review Letters.
Although other researchers have begun to push under the diffraction limit, Hell and Dyba's technique offers the best resolution so far, says Peter So, a biological engineer at the Massachusetts Institute of Technology. Improving resolution by 30%, as the original confocal microscopy technique did, "is kind of nice," So says, but "improving by 1000% will allow you to do things you couldn't do before."