Researchers from the U. of New South Wales (Australia) have produced a multicell combination that tops any previous mark for solar-cell conversion efficiency in any type of cell.
August 31, 2009 – Researchers from the U. of New South Wales (Australia) have produced a multicell combination that tops any previous mark for solar-cell conversion efficiency in any type of cell.
The work, led by Martin Green, research director at UNSW’s ARC Photovoltaics Center of Excellence, combines a silicon cell optimized to capture red and infrared light, converting up to 46% of light. To this was added four other cells from the US National Renewable Energy Laboratory (NREL) and Emcore Corp., made of various other materials with different bandgaps (GaInP/GaAs, a two-terminal stack; GaInAsP/GaInAs, a three-terminal stack; GaInAsP; and GaInAs), optimized for other parts of the spectrum (UV to far-infrared). Together, the five-cell combination managed 43% efficiency, beating the previous mark of 42.7% held by the U. of Delaware (which also used the four nonsilicon cells in its multicell combination).
The research, to be published in the September issue of the journal Progress in Photovoltaics, also addresses problems in using cell filters with sharp cutoffs (GaAs for the silicon cell and “an idealized silicon cell filter” for the low-bandgap GaInAsP/ GaInP cell combination) which result in overlaps in certain spectrum ranges (871-890 wavelength and 1100-1200, respectively). These are characterized as a “limitation” but ultimately “not important,” translating to “about 10% relative reduction in performance,” or “0.5% absolute efficiency reduction,” within an estimated measurement certainty of ±2.5%.
Stuart Wenham, director of UNSW’s ARC Center, noted in a statement that the new record is “not directly comparable” to a 25% efficient individual solar cell that they developed a year ago — the five-cell structure is an “expensive combination” and the sunlight is focused to produce “much higher intensity.” Further tests showed cells provided for a nonconcentrating silicon modules were calculated with 22.9% efficiency vs. 20.5% for concentrated modules. The concentrator also only works in direct sunlight, they added. And there would be “other unavoidable losses” in concentrator cells due to the lens, light uniformity, and thermal conditions, they added.
Nevertheless, the work shows “what eventually may be practical,” Wenham noted. In the paper the researchers note that a split-spectrum’s increased system complexity and component cost “could more than offset” at the system level with “substantially higher efficiency” — likening the comparison vs. monolithic tandem cells to the practical advantages of silicon vs. III-V cells.