Superlensing Demonstrated in the Mid-IR

Hank Hogan

Researchers at the University of Texas at Austin and at Case Western Reserve University in Cleveland have created a functional superlens in the mid-infrared, achieving a resolution of better than λ/10 using an 11-µm source.

Using optical phonons, a research team at the University of Texas at Austin and at Case Western Reserve University in Cleveland have created a mid-infrared superlens, achieving a resolution of better than λ/10 at 11 µm. According to the group’s calculations, a square array of nanorods, perhaps fabricated of anodized aluminum, may demonstrate superlensing at near-IR and visible wavelengths. Courtesy of Gennady Shvets.

They did so by making use of phonons, quantized vibrations found in crystals. The work may enable the development of devices that resolve features far below the classical limit of the illumination wavelength, as well as smaller antennas than are practical today.

Research team leader Gennady Shvets, an assistant professor of physics at the University of Texas, explained that optical phonons arise whenever a crystal consists of several elements. The phonons lead to resonances at specific wavelengths, an important factor in achieving the dielectric characteristics required for superlensing.

“In the vicinity of those resonances, you can have negative dielectric permittivity,” Shvets said. He added that silicon carbide phonons suffer smaller losses than those of other crystals. As a result, the material is well suited for a superlens.

In constructing their device, the researchers grew a 400-nm SiC film on a silicon wafer. They then etched away the silicon in a 300 × 300-µm area, yielding a SiC membrane that stretched across the opening. They deposited a 200-nm-thick silicon-dioxide film on both sides of the SiC, creating a metamaterial with an effective dielectric permittivity of 0 at 11 µm.

In a series of tests involving 500-nm slits that were spaced 2.5 µm apart, the investigators demonstrated that the resolution of the superlens was better than 1 µm for an 11-µm source, much better than the λ/2 resolution possible with a conventional lens.

Although the demonstration of superlensing and its potential applications in imaging and lithography are the major motivations for the work, it also might enable the development of antennas that are much shorter than the l/2 needed now for practical use. An enhanced antenna might be built by depositing small metallic posts on a SiC film. The group is researching this approach.

Other regions

As for pushing superlensing down into the near-infrared and visible spectral regions, that will require a different tactic.

“It’s not going to involve silicon carbide, because silicon carbide has interesting properties only in the vicinity of 11 µm,” Shvets said.

One possibility, which the scientists are investigating, is the use of a square array of metallic nanorods spaced approximately 100 nm apart. This will produce a metamaterial with a negative refractive index, he said, and numerical calculations show that such a structure will exhibit superlensing. He said that this work is theoretical but that one approach might be to create the array using anodized aluminum.

The group reported on the work in October at the Optical Society of America’s 89th annual meeting in Tucson, Ariz.