Here comes the sun

Talk to anyone involved in photovoltaics (PV) research and you will hear predictions that solar-electric technology is about to turn the corner. Photovoltaics, the direct conversion of sunlight to electricity, is gaining ground as a viable alternate energy source because it's clean, quiet, reliable, and unlimited. First used in the '60s to power satellites, PV's potential became more apparent during the oil embargoes of the '70s when consumers began looking for more control over their energy source. Currently, solar electricity is used to power loads such as lighting and small appliances, and in remote areas where high reliability and low maintenance are prerequisites.

PV's power comes from solar cells which use semiconductor material to directly convert sunlight into electricity. About 95% of today's solar cells are made from crystalline silicon, the other 5% use newer thin-film technology. A typical crystalline silicon cell is about 4 in. around and 0.010 in. thick and generates 2 A of direct current at 0.5 V in direct sunlight. Of late, silicon cells are about 12 to 14% efficient, while thin-film falls between 7 and 10%. But researchers from the Dept. of Energy's National Renewable Energy

Laboratory believe thin film is the way of the future. It uses significantly less semiconductor material than crystalline silicon and is easier and less expensive to make. While the PV market is growing at about 25 to 30%/year worldwide, solar technology still isn't cheap enough to compete with conventional energy sources. For that, manufacturing costs must go down and efficiency must rise, which is why the PV industry has developed a plan to catch up.

Mapping the Future
The DOE and the U.S. photovoltaic industry, through the National Center for Photovoltaics, have created a plan for domestic PV as far out as 2030. Their PV Roadmap calls out short, medium, and long-range goals as well as barriers to the technology's success.

The goal is to make PV a more significant source of electricity within the next 20 to 30 years. By 2030, the industry hopes PV will provide 10% of U.S. peak generation capacity. To make this projection reality, the Roadmap implores government to level the playing field with other powerproducing technologies via tax credits, manufacturing incentives, and by establishing standards, codes, and certification.

The plan hinges on overcoming a host of technical barriers. Chief among them is the high cost of manufacturing solar-electricpower components. Low-cost, highthroughput manufacturing technologies for high-efficiency thinfilm and crystalline-silicon cells are a must. The industry has established an 18 to 20% conversion efficiency goal with a cost of less than $0.50/W for each module technology. To produce crystalline silicon at these efficiency and cost models, solar-grade silicon must be available at less than $20/kg.

Another obstacle is the current lack of a manufacturing infrastructure to meet future demands. Throughput at this point is low, process controls are inadequate, and production lines lack the automation necessary for cost efficiency. PV systems as currently designed do not lend themselves to mass production. Stepped-up R&D efforts and a manufacturing infrastructure may be the answer. Researchers also need to figure out how to drive down the high cost of module materials and their encapsulation.

Though the task is daunting, researchers are making big strides.

The future's looking sunny
Recently, 11 universities and five businesses were awarded $6 million toward researching high-tech, nonconventional photovoltaic technologies. Some are testing exotic materials while others are developing entirely new ways of harvesting energy. For instance, ITN Energy Systems Inc., Littleton, Colo., is now developing optical rectenna solar cells. The idea is to replace the semiconductor with an incredibly tiny antenna and rectifier that would capture light frequencies and convert them into electricity. "In our concept, the antenna couples to the solar spectrum and a diode rectifies it," says Dr. Brian Berland, ITN researcher. "If you think of it as a sine wave, there is an alternating positive and negative voltage. Putting the diode in the middle creates a dc waveform. This is often referred to as rectenna technology." Rectennas have been used for converting ac to dc power from satellites and remote airplanes because power can be broadcast through the air. Previous concepts were designed for RF frequencies, but ITN is working to apply that technology at visible and infrared frequencies in the solar spectrum. "People have been able to do this in the RF world because the wavelengths are on the order of millimeters, says Berland. "An antenna can be made with a fairly simple printed-circuit board and components are relatively large, and commercially available. The wavelengths for infrared, on the other hand, are on the micron-length scale, 1,000 times smaller than regular rectenna technology. Diodes for those frequencies are just nanometers long."

Berland admits ITN's approach is challenging, requiring sophisticated nonpatterning techniques and work with advanced materials properties to get antennas behaving the way they are supposed to.

Currently, the devices are fabricated using electron beam lithography, which is similar to state-ofthe-art processing used to make computer chips. An organic resist is exposed to an electron gun that creates a mask to selectively deposit material in the desired pattern across the substrate.

If ITN succeeds, the pay off would be huge: perhaps a 90%-efficient solar cell. "Our strategy has been to work at the nanometer-length scale," he says, "In collaborations with NIST and the University of Colorado, we have fabricated devices in this size that respond at 30 THz, which is 10-micron radiation in the infrared spectrum. If we're successful, it will be better than the solar cell could ever be."

Ready, set, concentrate
Another interesting approach, called concentrating photovoltaics, focuses sunlight through mirrors or lenses onto a very small solar cell with the idea that more illumination equals more power per solar cell area. One company, United Innovations Inc., San Marcos, Calif., is developing and proving its novel approach to this familiar technology. Its mission: develop what are called rugate filters for a superhigh-efficiency solar-concentrator system. The design combines a first-stage parabolic concentrator with circular heliostat facets, a second-stage nonimaging concentrator, and a spherical cavity lined with multiple solar cells that collect light and convert it to electricity. In this design, all sunlight focuses on the entrance of a sphere. The interior wall of the cavity is lined with sets of solar cells tuned to particular frequency bands in the solar spectrum. Each cell has a specially designed rugate filter on top that lets in only the portion of

the solar spectrum that matches the spectral response of the cell beneath. This concentrator concept could ultimately convert 50% of sunlight into electricity.

A bit further up the coast in Torrance, Calif., Amonix Inc. is already developing high-concentration PV systems. In fact, the company holds the world record for efficiency of a silicon solar cell manufactured in a commercial environment at 26.5%. The Amonix concept substitutes an inexpensive plastic concentrator lens for costly silicon. In a typical flat-plate solar module the entire sun-receiving surface is covered with silicon solar cells positioned at the fixed tilt to the sun.

Amonix's system, on the other hand, uses flat, plastic Fresnel lenses between the sun and the cell to focus and concentrate sunlight by a factor of 250 onto a relatively small cell area. Because less cell material is needed to generate a given amount of electricity, system costs are lower. The company also developed a hydraulically driven tracker that follows the sun throughout the day, concentrating sunlight onto the cell and boosting energy production. The result? Amonix's system captures as much as 30% more energy than traditional fixed systems, says the company.

A recent deal between Amonix and the Arizona Public Service (APS) Co. will put the largest field of high-concentration PV systems in operation. The distributed, multisite system will be rated at more than 500 kW and will produce enough energy to power more than 165 homes.

When fully operational, the system will generate more than 1,000 MW-h/year. Power will feed into APS' electrical grid and displace the same amount of electricity that would have been produced using traditional methods, say company officials, without the pollution and fuel costs inherent in fossil-fueled generators.

Focusing on materials
Additional dollars are going toward new-materials research. Crystalline silicon has been PV's workhorse for years but it is expensive to produce. Some of the most promising alternatives are thinfilm materials in which layers of different electricity-producing materials are applied sequentially to glass, plastic, or steel backing.

To get there, however, researchers have several hurdles to overcome. Currently thin-film modules are only 7 to 10% efficient as compared to silicon, which is nearly double that, according to NREL's National Center for Photovoltaics. Scientists are working overtime to better understand semiconductors — how to boost efficiencies and deposit these materials onto substrates.

Researchers from the DOE National Renewable Energy Laboratory (NREL) recently made big strides by surpassing a record for electricity produced by cadmium telluride (CdTe) solar cells. Previously, a cadmium telluride cell was 15.8% efficient, but now hits 16.4%. The process uses new materials that interact chemically with the CdTe to improve adhesion, light collection, and electronic properties, say NREL researchers. Of the several materials that can be used for thin-film panels, CdTe yields higher W/ft², at a lower price/W of capacity. Two large U.S. photovoltaic plants are now producing CdTe thin-film panels.