Portland, Ore. -- All natural materials bend electromagnetic radiation--from microwaves to visible light--in the same predictable direction: away from a line perpendicular to their surface, or away from "normal." On the other hand, metamaterials--exotic, artificially created materials with optical properties not found in nature--substitute periodic mechanical structures that can force a photon to travel a path where it bends toward normal, thereby enabling a flat lens to nevertheless focus light.
For the last few years, researchers have demonstrated that shorter and shorter wavelengths of light can be focused by metamaterials with more and more closely spaced periodic structures. Now, researchers at the Department of Energy's Ames (Iowa) Laboratory say they have demonstrated the world's first metamaterial for visible wavelengths.
"To achieve optical wavelengths, we had to reduce our design rules to 30 nanometers using e-beam lithography, which is roughly the limit today," said Ames Lab senior physicist Costas Soukoulis.
For the last few years, Soukoulis' research group has demonstrated what it calls a "fishnet" design, which casts matching periodic structures in two metal screens separated by an optically clear dielectric. To pass through, electromagnetic radiation must navigate as though it was bending toward the normal, thereby realizing a negative index of refraction (all natural materials have a positive index of refraction).
For its original gigahertz and terahertz demonstrations--extending from microwaves all the way up to the near-infrared--the team was able to use photolithography to form periodic structures up to 100x smaller than the wavelength to be transmitted, thereby mitigating losses and increasing efficiency. But to reach visible wavelengths--measured in nanometers--the researchers had to cut back their periodic structures to measure just three times smaller than the light's wavelength, thereby incurring significant losses.
"For the visible-wavelength demonstration at 780 nm, the periodicity of the fishnet pattern was around 200 nm--only about one-third the size of the wavelength being measured," said Soukoulis. "Our fishnet design worked very well at gigahertz, then we pushed it to terahertz, and now all the way to visible wavelengths where we have lots of losses. Next we want to improve our material before we try it in real-world applications."
Soukoulis worked with researchers Stefan Linden and Martin Wegener at the University of Karlsruhe in Germany.
Metamaterials exhibiting an effective negative index of refraction could theoretically focus perfectly even light for very small objects that are very nearby. Today, the necessary concave structures cause distortion, which gets increasingly worse as the object to be imaged gets smaller and closer. Metamaterials, on the other hand, can theoretically focus light onto or from objects of any size and at any distance without any aberration whatsoever.
Normal lenses use refraction that passes light rays obliquely from one transparent medium into another in which its wave velocity is different. Metamaterials rely instead on defraction: modulating wave fronts passing by the edge of an opaque periodic structure, thereby causing a redistribution of energy within the front. When done well, defraction with metamaterials can outperform refraction with natural materials, enabling aberration-free and nearly lossless metamaterial lenses.
"We did not measure diffraction for our material yet," said Soukoulis. The team's next demo, he said, will offer "a direct proof using refraction at optical wavelengths that everybody will believe has a negative index, rather than the indirect methods we had to use in this first demonstration."
The material was fabricated by growing a fishnet pattern on both the top and the bottom of a transparent dielectric substrate. By precisely aligning the two fishnet layers, 780-nm light was forced to bend around the periodic structures in the manner of a material with a negative index of refraction.
"We measured the transmission and reflection of incident electromagnetic waves, from which we inferred the effective index of refraction of this metamaterial, which ended up to be negative point six [–0.6]," said Soukoulis. "We also did some experiments with pulses, and by measuring how fast the pulse propagates you can indirectly find the index of refraction."
Metamaterials exhibit a negative permittivity and permeability, and thus enable a negative index of refraction. That quality enables metamaterials without an optical axis to focus waves, despite their planarity, by means of refraction. Combining normal materials with metamaterials could also enable exotic lenslike structures that bend light around an object, thereby cloaking it, or making it invisible (see www.eetimes.com, article ID: 191901472).
Positive or negative
Metamaterials were first hypothesized in 1968 by Russian theorist Victor Veselago, who theorized that light would interact with them in the opposite way from natural materials, which all have positive permeability and permittivity. One intriguing prediction was that such composite metamaterials could allow a flat lens to nevertheless focus electromagnetic waves.
Many groups have proved that metamaterials exhibit a negative index of refraction, but only for relatively long wavelengths such as microwaves and, more recently, the near infrared. Various research groups have fabricated metamaterials made from submicron metal rings and rods, thereby enabling exquisite focusing of nanoscale features. But until now none exhibited a negative index of refraction for visible light.
The next step for Soukoulis' group will be to first verify its results with a refraction demonstration of the left-handed rule for its material (Snell's Law says that all natural materials follow the right-handed rule--that current in a wire flows in the direction of the thumb of your right hand when your fingers are curled around the wire in the direction of its magnetic field).
Soukoulis then plans to try reducing propagation losses by using a substrate material that provides gain. The team is also hunting for methods to build three-dimensional isotropic structures so that waves can propagate parallel to the surface of the chip as well as perpendicular, as they do today. "We have some new 3-D designs that we think might help, but before we fabricate them for visible wavelengths, we will first try them out at gigahertz and then terahertz wavelengths," Soukoulis said.