An international team of researchers has added a small twist to typical light filters. They’ve fabricated a nanoscale device that divides incident light into its different component colors based on the direction the light is coming from. Their device that could drastically alter how color displays are designed and even improve the efficiency of solar cells.
To accomplish this trick, the researchers have moved beyond the periodic architectures of a generation of tunable optical devices. In other words, they’ve switched from making filters with equally spaced grooves to an aperiodic architecture wherein the grooves are not equally spaced.
Researchers at the U.S. National Institute of Standards and Technology (NIST), in collaboration with Syracuse University and Nanjing University in China, have discovered that when incident light strikes these aperiodic grooves (nanoscale trenches on an otherwise opaque metal film) it triggers the creation of surface waves, referred to as plasmons.
In research described in the journal Nature Communications, the scientists were able to engineer the structure so that plasmons generated by individual grooves meet at the device’s central through slit with the same phase (constructive interference), but at multiple user-defined angles and wavelengths. The result: an angle-dependent color filtering response.
Schematic shows two different ways that white light interacts with a newly developed device, a directional color filter ruled with grooves that are not uniformly spaced. When white light illuminates the patterned side of the compact metal device at three different angles—in this case, 0° degrees, 10° and 20°—the device transmits light at red, green and blue wavelengths, respectively. When white light incident at any angle illuminates the device from the non-patterned side, it separates the light into the same three colors, and sends off each color in different directions corresponding to the same respective angles.Gif: NIST
The angular response of an equivalent device with uniformly spaced grooves is limited by the grooves’ uniformity. In diffraction gratings, for example, the groove dimensions are restricted by the diffraction limit. But because the wavelength of plasmons is much smaller than the incident light, the researchers are able to sidestep the diffraction limit and shrink the groove dimensions so that they are much smaller than the wavelength of incident light.
“Plasmons, because of their shorter wavelengths, help us shrink the device footprint to the micron scale while at the same time allowing us to engineer light-matter interaction at the nanoscale,” explained NIST physicist and co-author of the paper Amit Agrawal. “By incorporating engineered aperiodicity into the device, we are able to get multi-functional response: angle-to-color, or color-to-angle mapping.”
Currently, displays and consumer cameras typically use what is referred to as Bayer color filters. One super-pixel in a display is in fact a combination of three sub-pixels, one for each RGB color. Any combination thereof will provide the desired color for that super-pixel. However, this comes at a loss of resolution by a factor of three.
Because the footprint of the device developed by Agrawal and his colleagues (Matthew Davis and Wenqi Zhu of NIST and the University of Maryland, along with NIST physicist Henri Lezecmatches) matches the size of individual pixels (10 micrometers or so), Agrawal envisions placing these color filtering devices on top of individual CCD pixels, and performing color filtering by changing the angle of incidence using a micromirror-array based device.
Another alternative Agrawal suggests is using the device in reverse. In this way, one can direct the RGB colors coming at different angles to three pixels forming a super-pixel (same as an LCD).
While this all sounds very appealing, Agrawal is quick to caution that the efficiency of the devices at this point is still pretty low because of scattering and absorption losses in the metal film. So, although there is some work to be done to improve it, Agrawal believes that the incorporation of such devices on top of an array of CCD pixels or on top of multi-junction solar cells, for example, should be pretty straightforward.
“There is already a precedent for it,” said Agrawal “Since these devices are angle sensitive (in the case of forward illumination), there will be some engineering required to illuminate them at various angles with micromirror arrays. However, at this point it is a matter of making an effort to increase the efficiency.”
The key to this efficiency-raising effort will be working with higher quality metal films than can be created using traditional physical vapor deposition techniques. It will also call for further optimization of the design process, according to Agrawal.
While the researchers grapple with those issues, Agrawal notes that the approach can also be easily extended to other regions of the electromagnetic spectrum, such as mid-infrared and the terahertz, where there are more choices for metal, and the losses are significantly smaller.
Agrawal adds: “In terms of efficiency, these devices should work remarkably well at those wavelengths and are especially suitable for applications such as sensing. Some biological and chemical molecules have spectral signatures at those wavelengths, so it makes logical sense for us to extend the operation of these devices to longer wavelengths.”
Dexter Johnson is a contributing editor at IEEE Spectrum, with a focus on nanotechnology.