301 Moved Permanently

---

nginx

In January, the U.S. Naval Research Laboratory (NRL) and Niles, Ill.- based MicroLink Devices Inc. announced a collaborative project to develop a triple-junction solar cell with the potential to reach 50% conversion efficiency. The goal of the research program, which also includes collaboration from Imperial College London, is to develop a lattice-matched triple-junction solar cell.

In October 2012, a few months before the NRL announcement, San Jose, Calif.-based Solar Junction announced it had broken its own record in cell efficiency with its newest multijunction solar cell that achieved 44% efficiency. In a press release, the company heralded its position as a leader in the multijunction cell space.

That multijunction solar space is currently a very small one, but it is also innovative and could contribute substantially to the broader PV module market, according to industry experts.

 

The research

Multijunction solar cells feature layers, each tuned to different wavelength bands in the solar spectrum. “We are effectively dividing up the spectrum of light,” explains Ray Chan, director of solar sheet products at MicroLink Devices. “It’s like three different solar cells stacked together.”

One of those three layers captures blue light, while the second captures visible light and the third captures infrared light.

For researchers, a key task is to find the right materials for these three different layers. “The novelty comes from the identification of the quaternary material InAlAsSb, which is lattice-matched to indium phosphide (InP) and, according to our modeling, can display a direct bandgap as high as 1.8 electron volts (eV),” explains Dr. Robert J. Walters, who heads the solid-state devices branch of the NRL.

“With this material, one can envision a three-junction cell made of materials all lattice-matched to InP to give bandgaps ideally matched to the terrestrial spectrum, under both one-sun [conditions] and concentration,” he says.

The term “bandgap” refers to how electrons respond when they are hit by a particle of light. Different materials have different bandgaps, measured in eV. The NRL noted in its research announcement that high-bandgap semiconductor material is used to absorb the short-wavelength radiation, with longer wavelength parts transmitted to subsequent semiconductors.

The goal for the NRL and MicroLink investigation of alternate materials is to enable a fully lattice-matched structure with optimized bandgaps. Walters points out that other existing multijunction solar cells are based on lattice-mismatched materials to attain the desired bandgaps.

“Our modeling has shown how these technologies cannot attain the high efficiencies that are achievable with lattice-matched materials,” he says. “The primary breakthrough was the identification of the InAlAsSb material.”

The researchers’ task now is to grow this material and demonstrate a high-quality junction within it. “We have just begun our research into this material development and hope to see demonstrable progress within the next year,” Walters says.

Currently, the research staff is only in the initial stage of the project. “We have produced the modeling and identified the materials, and we are now starting on the materials development,” he says. “There are challenges in the growth of InAlAsSb, so we are expecting some time spent in material development.”

 

Commercial uses

MicroLink grows cells using its proprietary epitaxial lift-off (ELO) process, which uses metal organic chemical vapor deposition. According to the company’s website, a release layer is deposited directly on top of a GaAs substrate, and multijunction solar cells are then deposited onto the release layer.

The solar cell epitaxial structure can be removed from the substrate without damage, and a metal backing layer provides mechanical support for the solar cell after lift-off. This ELO process is designed to reduce costs because the GaAs substrate can be reused for subsequent growths of additional solar cells, MicroLink says.

The company currently manufactures multijunction solar cells with a demonstrated efficiency of 31%. The cells are thin, less than 40 micrometers, and flexible, for use on curved surfaces.

Multijunction cells are used in concentrated photovoltaics (CPV) - mainly in military and government applications such as satellites, unmanned aerial vehicles and other uses. “These have been used in space for years,” Chan says. “But that community can spend a premium.”

Indeed, the higher expense is one reason why multijunction CPV modules generally do not attract as much attention in the traditional markets as the crystalline silicon and thin-film sectors.

 

 

Manufacturers try to tackle these high costs through increased power output by magnifying the amount of sun hitting the modules. The magnification is usually expressed as “suns.” For example, Solar Junction recently announced that the power conversion efficiency (verified by the National Renewable Energy Laboratory) of its newest cells was measured at 44% at 947 suns, beating the company’s April 2011 record of 43.5% at 418 suns.

Solar Junction also says it is currently producing high-efficiency multijunction solar cells based on its proprietary A-SLAM materials. A-SLAM refers to Adjustable Spectrum Lattice Matched architecture, which provides material bandgap tunability from 0.8 eV to 1.42 eV to maximize the absorbed sunlight within CPV modules.

What is unique about this triple-junction technology is that the material system is bandgap-tunable, but the atomic spacing of the materials remains the same, says Vijit Sabnis, vice president of technology at Solar Junction.

“It’s easy to find one material that can absorb one electron volt and another that can absorb two electron volts, but they are not compatible, so they would be lattice-mismatched to each other,” he explains.

Sabnis says the easiest, most reliable and lowest-cost path to high efficiency for multijunction cells is to use lattice-matched materials. “You need to have a manufacturing platform that is flexible enough to deposit the materials you need, and the materials need to be flexible enough to be lattice-matched,” he says. “That gives you manufacturing robustness and allows you the bandgap tunability you need.”

 

The future

According to Sabnis, recent innovations from the various multijunction technology developers will lift the entire CPV sector. However, it is important to keep in mind how cell efficiencies ultimately translate into module efficiencies. The sector must also keep in mind ambient conditions and solar irradiance.

“Fifty-percent efficient solar cells should give you [a] DC module efficiency of 42 to 45 percent,” Sabnis says. “The tricky thing about CPV is you only look at direct normal incidence light.

“The 42 to 45 percent might go down to 35 percent when taking into account diffuse radiation, but nonetheless, you put an inverter in, [and] you’re talking about AC efficiencies of 34 percent, and that is a really big deal,” he adds.

To put that in perspective, he says, traditional flat-panel modules may offer 18% efficiency.

Meanwhile, Heilbronn, Germany-based Azur Space Solar Power GmbH announced last November that it had achieved 43.3% efficiency with its new multijunction cell using 500-fold concentration.

The company reported at the time that this was the highest efficiency ever for a solar cell grown upright on metal-organic vapor phase epitaxy technology. The cells consist of gallium indium phosphide, gallium indium arsenide and germanium.

“The continuous small improvements, as well as new cell architectures, lead to the steady increase in efficiency,” says Dr. Andreas W. Bett, director of materials, solar cells and technology at the Fraunhofer Institute for Solar Energy Systems in Freiburg, Germany.

“Today, records still are made with triple-junction solar cells, but with different internal architectures,” he adds. “If we proceed well beyond 45 percent [efficiency], we will see more junctions than three.”

Bett adds that the industry is very promising and innovations happen quickly. In the PV technology sector, commercial-ready products often hit the market as soon as two years after record efficiencies are announced. This schedule is unique to the PV industry, he says.

In general, these new multijunction cells will be used in other applications besides space - namely, land-dwelling PV modules. “Indeed, the terrestrial CPV market is just as [large a] - if not a larger - goal as space,” says the NRL’s Walters.

Sabnis agrees that there is a positive future for multijunction cells. He says Solar Junction is working with several companies, including Semprius, Soitec, Amonix, Heliotrop and several Chinese firms.

“What’s great about CPV is it has a lot of headroom for improvement for performance and cost,” Sabnis says. “When you look at traditional PV, really huge volumes have been deployed, but those technologies are mature, and they are successful.

“CPV is now riding the coattails of traditional PV,” he adds. “In three to five years, you will see 50 percent efficiency multijunction cells.” S

 

Nora Caley is a Denver-based freelance writer.