By Kimberly Patch,
Technology Research News

One way to make solar cells more efficient is to find a material that will capture energy from a large portion of the spectrum of sunlight -- from infrared to visible light to ultraviolet.

Energy transfers from photons to a photovoltaic material when the material absorbs lightwaves that contain the same amount of energy as its bandgap. A bandgap is the energy required to push an electron from a material's valence band to the conduction band where electrons are free to flow.

The trouble is, most photovoltaic materials absorb a relatively narrow range of light energy. The most efficient silicon solar cells capture about 25 percent of the sun's energy. Multijunction solar cells combine several materials to capture multiple bands of photonic energy. Today's most efficient combination -- germanium, gallium arsenide and gallium indium phosphide -- boosts efficiency to 36 percent, but is relatively difficult to make and therefore expensive.

Researchers from Lawrence Berkeley National Laboratory, the University of California, and the Massachusetts Institute of Technology have engineered a single material that contains three bandgaps. The material is capable of capturing more than 50 percent of the sun's energy, said Wladek Walukiewicz, a senior staff scientist at the Lawrence Berkeley National Laboratory.

The material could lead to relatively inexpensive, highly-efficient solar cells. Such cells would be much simpler than today's high-end multijunction solar cells because the three bandgaps reside in a single material, said Walukiewicz.

The researchers have manufactured a prototype single-junction 3-band semiconductor from the material. Although the concept of a multiband material was proposed in 1960, the researchers' prototype is "we believe, the first realization of a multiband semiconductor," said Walukiewicz.

The researchers were working on making a three-junction photovoltaic cell when they accidentally made a material that had a split bandgap. Once they realized the nature of the material, they reverse-engineered their inadvertent discovery to figure out how it happened.

The key turned out to be replacing some of the atoms of a material that is strongly electronegative with atoms of a material that is even more electronegative to create a highly mismatched alloy. The introduced atoms split the conduction band.

The researchers found that in the case of zinc-manganese-tellurium, instead of splitting the conduction band, introduced oxygen molecules "formed their own band well separated from the original conduction band," said Walukiewicz. "As a result we had a semiconductor with three bands -- the valence band and two conduction bands.

The difference between the material's valence band and first conduction band, or the amount of energy needed to push an electron from one to the other, provided a bandgap that absorbs photons that contain 1.8 electron volt. The difference between the two splits made for a 0.7 electron volt bandgap. And the difference between the valence band and second conduction band made for a bandgap of 2.6 electron volts. These three gaps cover much of the solar spectrum. An electron volt is the work required to move an electron through a potential difference of one volt.

The researchers made the material by forcing oxygen into zinc-manganese-tellurium. They did so by heating a mix of zinc-manganese-tellurium and oxygen with a laser so that it melted and then recrystallized. The laser made this happen fast enough that oxygen could not escape, and so was forced to become part of the crystal, said Walukiewicz.

The researchers are working on making a practical solar cell from the material, according to Walukiewicz. This requires forming p-type and n-type versions of the material. The valence band of p-type, or positive-type material has missing electrons, or holes. The conduction band of n-type, or negative material, contains electrons. A junction between the two types of materials creates a place where, when photons are absorbed, electrons move to the p-type material and holes toward the n-type to create an electrical current.

It will take to three years to assess the technical feasibility of the multiband solar cell, according to Walukiewicz.

Walukiewicz's research colleagues were Kin Man Yu, J. Wu, W. Shan, and J. W. Beeman from Lawrence Berkeley National Laboratory, M. A. Scarpulla, and O. D. Dubon From Lawrence Berkeley National Laboratory and the University of California, and P. Becla from the Massachusetts Institute of Technology. The work appeared in the December 12, 2003 issue of Physical Review Letters. The research was funded by the Department of Energy (DOE).