After illuminating an area of a 20-nanometer-thick silver-oxide film with blue light at the specific wavelength of 515 nanometers, chemist Robert Dickson observed that the film became photoactive. Subsequent illumination at a longer wavelength caused the exposed areas to fluoresce. The effect could be used to store data on the films, and since the fluorescent centers appear to comprise silver nanoclusters with as few as eight atoms, the effect might lead to extremely high data densities.

"Other work has shown that silver compounds will do this sort of thing at cryogenic temperatures and in rare-gas environments, so we were surprised to find this effect at room temperature," Dickson said. While the physical process underlying the effect is not fully understood, Dickson believes the blue light causes a few silver atoms to dissociate from their oxygen partners and coalesce into nanoclusters. The clusters, surrounded by the higher-dielectric-constant silver oxide, become trapping centers for electrons and holes. Subsequent illumination seems to activate the electrons and holes, producing photons, representing a data readout operation.

"One problem with nanocluster research in general is that you can't directly observe the particles, so models have to be constructed from observed data to explain what is going on," said Dickson. The process does bear some resemblance to photographic film. With film, silver atoms are combined in a salt — silver halide — and exposure to light causes the silver atoms to dissociate and form grains. The longer the exposure, the denser the distribution of grains, which appear as dark "pixels" in the image, leading to the typical negative image. The silver clusters, however, are much larger than the nanoclusters that occur on silver-oxide films.

In the silver-oxide films, pure silver nanoclusters seem to form by photo-induced dissociation from the silver oxide. The result is a silver-oxide cluster adjacent to a pure silver cluster.

The two seem to act in tandem as a chromophore, Dickson explained. The silver-oxide clusters appear to react strongly to blue light, which energizes the pair. At longer, green and red wavelengths, the companion silver nanocluster appears to be most strongly affected, giving off photons at a variety of wavelengths (yellow, green and red).

Nondestructive readout

Because the longer-wavelength radiation does not act on the system photochemically, it provides a means of nondestructive readout of data written with blue light. The method was tested by illuminating areas of the film with a small aperture of blue light; the areas reappeared under the illumination of green light. One unknown in the process is how long the data will be retained after initial exposure.

The theory of how the films operate was tested by growing 20-nm silver films and allowing them to oxidize in a normal atmosphere. On thicker films, oxidation produces a uniform top layer of silver oxide. On very thin films, however, the oxidation process cannot completely take place, leaving silver-oxide clusters interspersed with pure silver nanoclusters. As expected, the films showed the same type of fluorescent behavior.

The clusters not only create small data points but can also produce light at multiple wavelengths and thus could allow more than 1 bit to be stored at a single site. That would enable not only high-density data storage but also parallel read and write operations.

Dickson said achievement of a medium comprising silver films operating at room temperature bodes well for practical applications. Although work remains to be done, the versatile aspects of the effect appear to hold out some promising possibilities for optical storage technology.

"Of course, my motivation is the study of the basic effects," Dickson said, noting that in addition to possible technological applications, the fluorescence from the films also has esthetic value. "We were amazed at the beauty of the fluorescence from the sample," he said.