To compress light ultrasmall in metal nanostructures
To focus light ultrasmall is the key to a sharper microscopy, smaller structures in semiconductor industry and more efficient solar cells. Unfortunately, it has been known for 135 years that it is not able to focus light smaller than its wavelength. Researchers at the FOM-Institute for Atomic and Molecular Physics (AMOLF) in Amsterdam show that it is possible to break this barrier by using special metal nanostructures. They have published two articles in the leading scientific journal Nano Letters. Their research might be useful to solve two familiar problems. The researchers have just demonstrated that conversion of infrared light into visible light can be intensified by well-focusing. This process is also important to creating more efficient solar cells: sunlight contains a lot of infrared light, but solar cells absorb only visible light. By using the familiar photolithography technology from the computerchip industry, focusing makes it possible to realize structures that are ten times smaller than is now standard possible with light.
Ask what the smallest length scale is on which a beam of light can be focused and the traditional answer is: the wavelength of a photon. Thus in 1872, Ernst Abbe already proved that it is not possible for a microscope to see details that are smaller than the wavelength of light (about 500 nanometres), because a beam of light cannot be focused closely and infinitely. This means not only a restriction on producing sharp exposures of tiny objects under a microscope, but also on manipulating light in optical chips. Is focusing ultrasmall impossible, or rather?
Researchers at AMOLF are developing so-called 'plasmon structures'. The diffraction limit that applies for the usual 'transparant' optical materials such as glas, can eventually be circumvented by using these structures. The trick is to concentrate on the light with the aid of electrons in a metal. Part of the energy of the light is being transmitted to a back and forth movement of electrons. Such an oscillation of charge is also called plasmon. It is located closely to the interface of metal and air. So, it is possible to combine light with surfaces of metal by coupling light with the aid of a grating (see figure 1 on the left) to plasmons.
Structure of metal decreases wavelength of light
However, the plasmons will not be automatically combined with the surface of a metal film. In order to achieve this the Amsterdam researchers have made a narrow tip in the metal film that functions as a waveguide for plasmons. A broad plasmon wave that arises at the wide side of the waveguide, is compressed even closer as it approaches the tip, which also causes an increase of the energy density toward the tip. The researchers have demonstrated that indeed, the light intensity seriously increases toward the tip of the waveguide by examining a non-lineair optical process, called 'upconversion': four infrared photons from the beam of plasmon (wavelength of 1490 nanometres) are converted into one photon with a higher energy (wavelength of 550 nanometres, i.e.visible light). This process is strongest at the tip of the waveguide, because there the beam of plasmon has been focused most. The intensified conversion of infrared into visible light is, among other things, important to solar cells, which do not absorb the infrared part of sunlight, but indeed the visible photons.
Silver nanoparticles focus on light
Whereas the plasmon waveguide is a solution to focusing on a beam of plasmon in, for example, an optical chip, it will not solve all problems with the diffraction limit. In a second article the AMOLF-researchers, together with colleagues at the University of Amsterdam and Auburn University, United States of America, show that plasmons help to break up the diffraction limit in the process of photolithography, the process that has been improved by semiconductor tycoons like Intel, IBM, AMD and ASML in order to create even smaller electronic circuits.
In photolithography a mask (a kind of slide) is projected on a photosensitive layer (the 'resist'). Photolithography has two disadvantageous properties. The first disadvantage includes the fact that the smallest possible transistor has at most the wavelength of light. The second one is the fact that manufacturing the mask is very expensive and once made, it is not possible to alter it. This implies that the exposure cannot be changed when using an existing mask. In the article the researchers at AMOLF show that masks of silver particles are able to eliminate both restrictions. By pressing a mask of silver particles closely against a resist and illuminating this with an unfocused beam of light, the resist is just locally illuminated through the high light intensity closest to the particles. The smallest length scale now is the distance between the particles that are ten times smaller than the wavelength of the light at which the mask is being illuminated. The unexpected finding of the Amsterdam researchers is the fact that the exact exposure is not laid down by the mask, but that it is still programmable (see figure 2). The fact is that high intensities do not arise equally at all particles in the mask. Instead, all kinds of different possible patterns may arise, which can be selected by cleverly changing the colour and the scope of the light.
For more information, please contact Femius Koenderink and Professor Albert Polman, FOM Institute for Atomic and Molecular Physics (phone +31 (0)20 608 123 4; website www.erbium.nl and www.amolf.nl.
The article on the plasmon waveguide is entitled 'Enhanced Nonlinear Optical Effects with a Tapered Plasmonic Waveguide', the authors are Ewold Verhagen, Kobus Kuipers and Albert Polman. This article has been published in Nano Letters in Februari 2007. The described research has been executed as part of the Joint Solar Programme, a joint research programme by FOM, Shell and NWO Chemical sciences.
The article 'Programmable Nanolithography with Plasmon Nanoparticle Arrays' by Femius Koenderink, Jesus Hernandez (Auburn), Francis Robicheaux (Auburn), Bart Noordam (University of Amsterdam) and Albert Polman will be published in Nano Letters, March 2007.