Sandia LabNews

Photonic tungsten lattice developed

Tungsten photonic lattice developed at Sandia changes heat to light

Tungsten-filament bulbs — the most widely used light source in the world — are infamous for generating more heat than light.

That is, they radiate more energy in the infrared than in the visible spectrum.

Now a microscopic tungsten lattice — in effect, a tungsten filament fabricated with an internal crystalline pattern — developed at Sandia has been shown to have the potential to transmute the majority of this wasted infrared energy (commonly called heat) into the frequencies of visible light.

This would raise the efficiency of an incandescent electric bulb from five percent to greater than 60 percent and greatly reduce the world’s most vexing power problem — excess electrical generating capacity and costs to homeowners caused by inefficient lighting.

The advance — which shifts emphasis from a photonic lattice’s ability to guide light to its capability of stopping other frequencies from passing through it — also opens the possibility of increased efficiencies in thermal photovoltaic applications (TPV). Using a tungsten lattice as an emitter at desirable frequencies, model calculations showed that the TPV conversion efficiency reached 51 percent compared with 12.6 percent with a blackbody emitter.

The advance, achieved at Sandia by Shawn Lin (1746) and Jim Fleming (1749), is reported in the May 2 Nature.

The imaginative work seems logical in retrospect, though the theory for the effect — re-partitioning energy between heat and visible light — remains unexplained. "It’s not theoretically predicted," says Jim. "Possible explanations may involve the variation in the speed of light as it propagates through such structures."

The achievement was accomplished by an extension of well-known MEMS (microelectromechanical systems) technologies that themselves have been derived from mature semiconductor technologies. As a result, fabrication of such devices could be cheap and easy.

The most common use postulated for photonic lattices was based on their capability to transmit beams of light at selected frequencies and bend their paths without losing any energy. The structures, most often made out of silicon, consist of tiny bars fabricated to sit astride each other somewhat like Lincoln Logs at pre-set distances and angles that form in effect an artificial crystal. Spacing of the bars allows passage of only certain wavelengths; other wavelengths cannot pass through. Desirable wavelengths not only transmit but also can be changed in direction by creating defects in the artificial crystal that cause the light to follow the defect along like a car passing through a curving tunnel. This meant photonic crystals had potential in optical communications, in which light beams currently carrying telephone messages and data must be converted to electrons — an expensive process — to change direction.

That was where published conceptions and economic activity seemed to have stopped.

Meltdown? Apparently not

A further question considered by Shawn and Jim, with assistance from colleagues Ihab El-Kady, Rana Biswas, and Kai-Ming Ho at Ames Laboratories in Iowa, was what happens to other energies that enter the interior of a three-dimensional crystal. If the crystal were built of tungsten — fabricated by creating a structure of polysilicon, removing some silicon and using chemical vapor deposition to deposit tungsten as a kind of backfill in the mold — the metal could handle quite high temperatures and have a large and absolute photonic band gap in the visible range where it is already known to emit light. But what would happen to the other, lower-wavelength energies brought in by an electric current? Would the structure melt through the build-up of heat? Or, more desirably, would the thermally excited tungsten atoms reinforce emissions at higher wavelengths, such as in the visible frequency range?

An order-of-magnitude enhancement

Energy at the edge of the photonic band was observed to undergo an order-of-magnitude absorption increase, or enhancement. This meant that energy was being preferentially absorbed into a selected frequency band. Meanwhile periodic metallic-air boundaries led to an extraordinarily large transmission enhancement. Experimental results showed that a large photonic band gap for wavelengths from 8 to 20 microns proved ideally suited for suppressing broadband blackbody radation in the infrared and has the potential to redirect thermal excitation energy into the visible spectrum.

Thus, not only is a more efficient incandescent lamp shown to be possible, but photovoltaics also can be provided with energy from heat-generators that have transposed energy wavelengths into the most optimal frequencies.

All work was performed on commercially available, monitor-grade six-inch silicon wafers. These photonic devices were fabricated in Sandia’s Microelectronics Development Laboratory using modifications of the standard CMOS processes originally developed for Sandia’s radiation-hardened CMOS (complementary metal-oxide semiconductor) technologies

The work was funded by the Laboratory-Directed Research and Development program through project manager James Gee (6200). Co-principal investigator Jim Moreno (6216) modeled the thermovoltaic results.