WITH ITS streamlined profile and gleaming skin, the Sunraycer looks more like a giant, high-tech cockroach than a road vehicle. But this experimental car recently raced 1,950 miles across the middle of Australia to win what was billed as the world's first international, transcontinental road race for solar-powered vehicles.
Fueled only by sunlight, the Sunraycer averaged 43 miles per hour over five and a half days (including overnight stopovers) and finished more than two days ahead of its nearest rival. The win was not only a triumph for Detroit-based GM Hughes Electronics, which designed and built the car. It was a dramatic success for solar-cell technology, which many scientists believe is approaching the point of widespread practical use.
The Sunraycer is powered by an array of 7,200 solar cells similar to those used in communications satellites. The cells have an efficiency of 16.5 percent in converting sunlight to electricity -- among the highest of all commercially available units. (The cell powering a handheld calculator typically has an efficiency of less than 8 percent.) Each is about twice the size of a postage stamp and roughly the thickness of a business card. Covering 90 square feet of the vehicle's surface, they provide 150 volts at a peak power of 1,000 watts -- about the same as a hair dryer. The car, 19.7 feet long and 6.6 feet wide, weighs a mere 360 pounds. On a sunny day, it can go as fast as 45 miles per hour on solar energy alone. It also normally uses a battery of rechargeable silver-zinc cells to provide extra power for acceleration and climbing hills, allowing the car to reach 60 mph.
But don't expect to see it on the road in the near future. Solar cells must become much more efficient before they are cost-effective. But that time is coming, owing to recent innovations -- from cells designed to operate in concentrated sunlight to units that work in the dark to others that imitate the solar-conversion process in green plants. Several industrial and government laboratories are betting that solar-cell technology will grow into a billion-dollar industry by the year 2000, especially if oil prices rise again.
The Power of Light
Formally called a photovoltaic cell, a solar cell acts like a pump. Sunlight striking its surface releases an electrical charge, which is forced to move in a particular direction through a circuit. The push comes from electrical differences built into the cell's sandwich-like structure.
The most common type of solar cell is made from crystals of the semiconductor silicon and consists of several layers. Atop a glass or plastic base is a thin metallic strip which conducts electricity and acts as an electrical contact. On top of that go two layers of silicon. Traces of different impurities added to the silicon give each layer a different electrical property. The uppermost layer is a metallic grid.
When sunlight shines on the cell's surface, it frees electrons from silicon atoms in the exposed parts of the upper silicon layer. Because these loose electrons are repelled by the underlying silicon layer, they are forced to make their way through any path -- in this case, a wire connecting the top grid to the lower contact -- that bypasses the junction between the silicon layers. The current that flows along the wire can drive a motoror other electrical device.
The path from sunlight to electricity, however, is fraught with obstacles. Some sunlight is reflected off the cell's surface; and even when it is absorbed, materials like silicon take in only a fraction of the light available. Electrons, freed by the sunlight, sometimes bounce around randomly for a while instead of heading directly for the circuit. Some electrons readily slip back into place among the silicon atoms where they started. All these losses lower the cell's efficiency. Researchers are looking now for new materials and structures that absorb light more effectively, and for ways to concentrate sunlight and reduce reflections.
The most efficient device yet produced -- the point-contact photovoltaic cell, developed at Stanford University -- has achieved an unprecedented 28.2 percent sunlight-to-electricity conversion efficiency in the laboratory. So far, the Electric Power Research Institute (EPRI) in Palo Alto, Calif., a utilities-sponsored research center has invested nearly $8 million to determine whether this kind of cell can be manufactured at a reasonable cost.
"This cell has come the closest in performance to what we feel needs to be achieved for photovoltaic cells used in utility systems," says Edgar DeMeo, EPRI program manager for solar power systems. The group is encouraged, he says, because "it really looks like the cell can be manufactured using techniques that are well established within the electronics industry." The idea is to use a system of tiny lenses to concentrate sunlight onto small photovoltaic cells specially designed to operate efficiently in high-intensity sunlight.
The point-contact cell has several features that make it particularly efficient. First, each single-crystal silicon chip, smaller than a fingernail and only a fraction of an inch thick, has a textured upper surface to spread out incoming light. A mirror-like lower surface helps trap light within the material so that more can be absorbed. Each of the surfaces has a thin, electrically insulating layer, except at the points where the current is conducted out of the cell. The insulation reduces the chance of light-ejected electrons recombining with the "holes" left behind by departed electrons in silicon atoms. In conventional solar cells, both the top and bottom surfaces must be coated with metal films or grids, which decrease a cell's efficiency. EPRI expects to spend as much as $20 million to demonstrate commercially viable mass production and to confirm that the cells can be packed into large arrays to generate the huge currents needed for utility power generation.
More, Faster, Cheaper
Alternatively, some labs are examining less expensive, less efficient materials spread over a larger area in the form of films much thinner than paper. Such films soak up light more effectively than bulk silicon crystals, and offer less resistance to the passage of electrons. Most thin-film solar cells, including those that drive solar-powered watches and calculators, are made from a noncrystalline, or amorphous, form of silicon. This saves having to grow large, single crystals of silicon -- normally a time-consuming, expensive process. Although its conversion efficiency is generally less than 10 percent, amorphous silicon is cheap and easy to make. Such devices, however, tend to be unstable. Up to half of their efficiency is lost after a few weeks of exposure to sunlight, limiting their use to devices that draw very little power.
Researchers have tried mixing amorphous silicon with other materials to improve its performance. In one particularly promising approach, several thin-film cells are stacked one on top of the other. Each cell is tuned to different wavelengths of light so that the combination captures a larger portion of the sunlight than would a single cell by itself. Such stacked cells also seem to be more resistant to degradation than purely amorphous silicon cells. Another alternative is a material made from the elements copper, indium and selenium. Not only is the material extremely stable, but it's also the most light-absorbent photovoltaic material known, and its freed electrons tend to stay free for a long time. Unfortunately, that design generates a lower voltage than conventional solar cells, is difficult to make and hard to work with. Researchers are also exploring the use of cadmium telluride and other exotic materials, including the semiconductor gallium arsenide. One type of cell made with the latter has shown an efficiency of 26 percent under concentrated sunlight.
Chemistry and Current
The silicon-based photovoltaic cell, however, is not the only device that converts sunlight directly into electricity. So does the photoelectrochemical cell, which is part solid and part liquid. It can also convert such plentiful materials as water and carbon dioxide into fuels such as hydrogen and methane, and can store solar energy for later use.
A typical photoelectrochemical cell consists of slivers of a solid semiconducting material immersed in an electrically conducting chemical soup. A special combination of ingredients is needed to produce the right chemical products and to generate an electric current efficiently and for a long time. And there are several problems. One is a kind of light-induced rusting that causes the semiconducting material to decompose. In addition, low conversion efficiency results because the junction between the semiconductor and its liquid is much less efficient at separating electrons and "holes" than the junction between layers in a silicon solar cell.
Although the technology of such "liquid-junction" cells still lags substantially behind solid-state systems, scientists have made substantial progress. Recently, Stanford chemist Nathan Lewis and his collaborators reported a photoelectrochemical device with a solar efficiency of 15 percent employing gallium arsenide. But stability and efficiency must be improved still further if photoelectrochemical cells are to become a viable alternative to traditional energy sources.
Stuart Licht and his colleagues at the Weizmann Institute of Science in Rehovot, Israel, have constructed a novel photoelectrochemical solar cell that includes the equivalent of a built-in storage battery. Its light-absorbing material is a single crystal of the semiconductor cadmium selenide telluride. The photoelectrochemical half of the device produces more than a volt of electrical potential at a respectable solar conversion efficiency of 11.8 percent. At the same time, part of the generated current is used to convert electrically charged tin into neutral tin metal in the storage half of the device. In darkness or below a certain level of light, the storage half of the cell delivers power by converting metallic tin back into its charged form. The net result is that the cell continues to work regardless of the light level.
"It's a wonderful system in its simplicity," says Licht, currently at MIT. "There's no electronic switching. There's no computer control. It's just a chemical system that stores energy and spontaneously releases it when it's needed." However, a great deal more research is needed before this experimental cell nears commercial development.
So far, solar cells have found application as power sources for satellites and for communications devices, such as microwave receivers and transmitters in remote locations. Their major use is in consumer products, including calculators, portable radios, TV sets and watches.
Cell efficiencies have improved to the point where solar-cell modules are now a competitive energy source for users not connected to a utility grid. In underdeveloped countries, for example, they can provide the energy needed to run an irrigation pump, a battery charger or a radio transmitter. But prices must still be brought down dramatically. Some optimistic forecasters predict that by the end of the century, solar-cell systems will be a cheaper source of electricity than oil. Yoshihiro Hamakawa of Japan's Osaka University says that megawatt power plants based on solar cells could be in service within 15 years. But few now talk about wholesale replacement of electrical power grids.
If costs fall, some product designers foresee mounting solar panels on cars -- to supplement the energy provided by other fuels -- or glazing office towers with low-cost, thin-film photovoltaic sheets. In the near-term, however, the most likely applications will involve common electronic products. On the horizon are portable, solar-powered electrical generators to run navigation systems and pumping stations. Soon the nearly ubiquitous boom box may carry pop culture into the remotest sites on earth.
"Photovoltaics is a technology that eventually may rival any other energy technology in size," says Kenneth Zweibel of the Solar Energy Research Institute in Golden, Colo. "Indeed, photovoltaics may someday be our best energy option."