Scientists at Sandia National Laboratories (Albuquerque, NM) believe the semiconductor alloy indium gallium arsenide nitride (InGaAsN) may have a future role as a junction layer in a photovoltaic power source for communications satellites. The reason is the material's potential efficiency rating of 40% in a multilayer solar cellwhich is nearly twice that of a standard solar cell.
"Adding 1%-2% nitrogen to an indium gallium arsenide (InGaAs) structure dramatically alters the alloy's optical and electrical properties," said Sandia physicist Eric Jones. "Nitrogen, a small atom with high electronegativity, has a large effect on gallium arsenide's bandgap structure, reducing it by almost a third (as much as 0.4 eV). The InGaAsN alloy is of interest to industry because it allows designers to tailor properties for maximum current production."
The material is produced with a metal-organic chemical vapor deposition (MOCVD) process, which uses an Emcore (Somerset, NJ) reactor to heat a GaAs wafer to 500°C-800°C. This heat causes the source chemicals containing indium, gallium, arsenic, and nitrogen to decompose and the elements to form a crystal on the wafer. Annealing may then be beneficial to boost quantum efficiencies (see figure).
According to Jones, the InGaAsN solar cell designed to power a satellite will probably have four layers. The top would be InGaP, the second layer GaAs, the third layer 2% nitrogen with indium in gallium arsenide, and the fourth layer germanium. A 1-2-µm-thick InGaAsN layer, lattice-matched to the GaAs substrate with n- and p-type doping, would be desirable.
Each layer will basically absorb light at different wavelengths in the spectrum. The first layer will absorb yellow and green light, and the second one green to deep red. The arsenide layer will absorb from deep red to infrared, and the germanium will absorb infrared and far infrared. The absorbed light would create electron pairs, with electrons drawn to one terminal and holes to the other to produce electrical current.
Existing satellite systems use either silicon for solar cells or a two-layered solar panel with an InGaP layer and a GaAs layer. The silicon cells have a maximum theoretical efficiency approaching 23%, while the dual-layer solar cells achieve roughly 30%. Jones and colleagues believe that commercial applications become interesting if the addition of an InGaAsN junction can make the overall efficiency surpass the best devices available today by at least 4%-5%. "By incorporating this semiconductor material," said Jones. "the size of the solar collecting package decreases, so the satellite weighs less and is cheaper to launch."
Before InGaAsN can realistically be applied to a photovoltaic system, though, more work is needed to better understand the material and improve its quality. "We are doing a lot of tweaking to try to make the material viable," said Sandia researcher Andy Ellerman. "This process includes changing parameters of the growth process, such as temperature, and then measuring the impact on the alloy grown. We also require a better understanding of both the alloy's optical and electrical properties."
One complex question the researchers are still fleshing out an answer for involves determining whether the nitrogen-induced states are extended (band-like) or localized (impurity-like). They also are working to understand the physical (or chemical) origin of the large bandgap reduction.
Paula Noaker Powell