The two-junction solar cell has reached a new milestone with a new world record of 31.1% conversion efficiency at the Energy Department’s National Renewable Energy Lab (NREL). The previous record of 30.8% efficiency was held by Alta Devices.
The new record may not last long. NREL Principal Scientist Sarah Kurtz, who leads the F-PACE project in NREL’s National Center for Photovoltaics said, “This joint project with the University of California, Berkeley and Spectrolab has provided us the opportunity to look at these near-perfect cells in different ways. Myles Steiner, John Geisz, Iván García and the III-V multijunction PV group have implemented new approaches providing a substantial improvement over NREL’s previous results.”
NREL Scientist Myles Steiner announced the new record June 19 at the 39th IEEE Photovoltaic Specialists Conference in Tampa, Fla.
Meanwhile MIT researchers are opening another avenue for improvement, aiming to produce the thinnest and most lightweight solar panels possible. As just noted above most efforts at improving solar cells have focused on increasing the efficiency of their energy conversion, or on lowering the cost of manufacturing.
In an interesting metric The MIT panels have the potential to surpass any substance other than reactor-grade uranium in terms of energy produced per pound of material. The MIT cells could be made from stacked sheets of one-molecule-thick materials such as graphene or molybdenum disulfide.
Jeffrey Grossman, the Carl Richard Soderberg Associate Professor of Power Engineering at MIT, says the new approach “pushes towards the ultimate power conversion possible from a material” for solar power.
Scientists have devoted considerable attention in recent years to the potential of two-dimensional materials such as graphene, Grossman says, but there has been little study of their potential for solar applications. It turns out, he says, “they’re not only OK, but it’s amazing how well they do.”
Using two layers of such atom-thick materials, Grossman’s team has predicted solar cells with 1 to 2 percent efficiency in converting sunlight to electricity. That’s low compared to the 15 to 20 percent efficiency of standard silicon solar cells, he says, but it’s achieved using material that is thousands of times thinner and lighter than tissue paper. The two-layer solar cell is only 1 nanometer thick, while typical silicon solar cells can be hundreds of thousands of times that. The stacking of several of these two-dimensional layers could boost the efficiency significantly.
These are two very different takes on the photovoltaic future.
At the start of NREL’s F-PACE project, which aims to produce a 48%-efficient concentrator cell, NREL’s best single-junction gallium-arsenide solar cell was 25.7% efficient. This efficiency has been improved upon by other labs over the years: Alta Devices set a series of records, increasing the gallium-arsenide record efficiency from 26.4% in 2010 to 28.8% in 2012. Alta’s then-record two-junction 30.8% efficient cell was achieved just two months ago.
NREL’s Steiner said, “Historically, scientists have bumped up the performance of multijunction cells by gradually improving the material quality and the internal electrical properties of the junctions – and by optimizing variables such as the bandgaps and the layer thicknesses. But internal optics plays an under appreciated role in high-quality cells that use materials from the third and fifth columns of the periodic tables – the III-V cells. “The scientific goal of this project is to understand and harness the internal optics,” he said.
When an electron-hole pair recombines, a photon can be produced, and if that photon escapes the cell, luminescence is observed – that is the mechanism by which light emitting diodes work. In traditional single-junction gallium-arsenide cells, however, most of the photons are simply absorbed in the cell’s substrate and are lost. With a more optimal cell design, the photons can be re-absorbed within the solar cell to create new electron-hole pairs, leading to an increase in voltage and conversion efficiency. In a multijunction cell, the photons can also couple to a lower bandgap junction, generating additional current, a process known as luminescent coupling.
The NREL researchers improved the cell’s efficiency by enhancing the photon recycling in the lower, gallium-arsenide junction by using a gold back contact to reflect photons back into the cell, and by allowing a significant fraction of the luminescence from the upper, GaInP junction to couple into the GaAs junction. Both the open-circuit voltage and the short-circuit current were increased.
MIT’s Marco Bernardi, a postdoc in MIT’s Department of Materials Science who was the lead author of the paper said, “Stacking a few layers could allow for higher efficiency, one that competes with other well-established solar cell technologies.”
For applications where weight is a crucial factor — such as in spacecraft, aviation or for use in remote areas of the developing world where transportation costs are significant – such lightweight cells could already have great potential, Bernardi says.
Pound for pound, Bernardi says, the new solar cells produce up to 1,000 times more power than conventional photovoltaics. At about one nanometer (billionth of a meter) in thickness, “It’s 20 to 50 times thinner than the thinnest solar cell that can be made today,” Grossman adds. “You couldn’t make a solar cell any thinner.”
On the cost front MIT has an advantage. About half the cost of today’s panels is in support structures, installation, wiring and control systems, expenses that could be reduced through the use of lighter structures. Plus the material itself is much less expensive than the highly purified silicon used for standard solar cells and because the sheets are so thin, they require only minuscule amounts of the raw materials.
MIT has a third party comment from John Hart, an assistant professor of mechanical engineering, chemical engineering and art and design at the University of Michigan, saying, “This is an exciting new approach to designing solar cells, and moreover an impressive example of how complementary nanostructured materials can be engineered to create new energy devices.” Hart, who will be joining the MIT faculty this summer but had no involvement in this research, adds that, “I expect the mechanical flexibility and robustness of these thin layers would also be attractive.”
The MIT team’s work so far to demonstrate the potential of atom-thick materials for solar generation is “just the start,” Grossman says. For one thing, molybdenum disulfide and molybdenum diselenide, the materials used in this work, are just two of many 2-D materials whose potential could be studied, to say nothing of different combinations of materials sandwiched together. “There’s a whole zoo of these materials that can be explored,” Grossman says. “My hope is that this work sets the stage for people to think about these materials in a new way.”
An additional advantage of such materials is their long-term stability, even in open air. In comparison other solar-cell materials must be protected under heavy and expensive layers of glass. “It’s essentially stable in air, under ultraviolet light, and in moisture,” Grossman says. “It’s very robust.”
So far the work has been based on computer modeling of the materials, Grossman says, adding that his group is now trying to produce such devices. “I think this is the tip of the iceberg in terms of utilizing 2-D materials for clean energy” he says.
Today no large-scale methods of producing molybdenum disulfide and molybdenum diselenide exist and it is already an active area of research. Grossman said manufacturing is “an essential question, but I think it’s a solvable problem.”
These two technologies couldn’t be further apart in basic positioning. Both offer huge gains in market ability. Still, the early, but lesser developed MIT work should have the strongest legs. Costs, where MIT practically gleams, should push the technology along quickly.
For now silicon solar cells dominate the world PV market, but researchers see opportunities for new materials. NREL is looking to high-efficiency concentrator cells bolstered by lenses that magnify the power of the sun are attracting interest from utilities because the modules have demonstrated efficiencies well over 30%. And there may be commercial opportunities for one-sun or low-concentration III-V cells if growth rates can be increased and costs reduced.
NREL’s Steiner said same cell should work well when lenses are added to multiply the sun’s power. “We expect to observe similar enhancements of the solar cell characteristics when measured under concentrated illumination.”
NREL is closing in on 48%. The question will be cost. Those high efficiencies will need less area, but the best photovoltaic land is cheap.