Share this story
Close X
Switch to Desktop Site

Energy from solar cells: clean, quiet, but not cheap - yet

Solar cells are the high-tech alternative energy source which almost everyone loves. Close cousin to the transistor and integrated circuit, these cooky-sized wafers, which transform light directly into electricity, may play an increasing role in the world energy picture.

Engineers admire them for the sophisticated science designed into their atomic structure. Alternative energy advocates and environmentalists favor them because they promise quiet, clean, economical, and decentralized electrical power. Major oil companies invest in them as potentially profitable new products. And energy analysts like to speculate how they could revolutionize the way energy is generated, distributed, and used if their costs can be brought low enough.

About these ads

The cost of a peak watt (a watt produced in full sunlight) from a solar cell has fallen 200-fold to $7 to $10. As a result, the domestic US market for photovoltaics has grown from almost zero to $25 million last year. US firms, which currently dominate the world market, sold another $25 million worth of solar cells overseas. And this year's market is expected to be 50 percent better. Photovoltaic or PV devices, as the solar cells are called, now are economically competitive with diesel-electric generators.

Today, largely due to an ambitious federal research and development effort, US companies are at the forefront of photovoltaics technology. Cutbacks in US government support, coupled with increased spending by Japan and European nations, raise the question of whether the US will retain this lead. However, US manufacturers generally believe that, with the market expanding as fast as it is , they can retain the lion's share of sales for at least the immediate future.

Still, to realize its full potential, the costs of solar-cell electricity must fall another factor of 10. According to researchers in the field, this is not a trivial challenge. However, recent developments have increased their confidence that the federal goal of producing 50 cent per peak watt electricity will be achieved by the end of the decade.

The Solar Energy Research Institute (SERI) is the lead federal agency in photovoltaic research. ''In the past few years, we have made enormous progress, '' says Satyen Deb, head of SERI's photovoltaics program. Raising the efficiency , increasing the stability, reducing the cost of the materials and the cost of manufacturing PV devices are the areas which are actively being pursued.

To generate 50 cent per watt solar electricity from cells of silicon (the traditional material), the Department of Energy has determined that the cost of solar cell grade silicon must be less than $14 per kilogram. Bulk silicon, made from sand, is both cheap and abundant. But to purify it to the point needed for efficient solar cells takes a number of steps. As a result, the cost of the highly purified material needed for traditional single crystal solar cells runs about $370 per kilogram, although fairly efficient cells have been made in the laboratory from $80 per kilogram silicon.

In recent laboratory work, however, SERI researcher Jerry Olson has developed and bench-tested a purification technique that may eventually produce PV-grade silicon for $3 to $10 per kilogram.

The traditional way to make a solar cell involves growing a large ingot or single crystal of silicon from a melt and slicing this into a series of wafers. These are called Czochralski cells, and they convert as much as 20 percent of incident sunlight into electricity. To make them, however, wastes about half of the high-cost material and the wafers do not pack efficiently into panels. To avoid these problems, a number of laboratories are working on ways to grow ribbons of silvery silicon, rather than ingots.

About these ads

A number of processes have been developed which produce ribbons which make 11 to 12 percent efficient photovoltaic devices. (Ten percent is generally considered the minimum practical efficiency.) The problem, until recently, has been the slowness with which these ribbons could be grown: a few centimeters per minute. Attempts to grow ribbons much faster have led to buckling, wavy edges, and similar problems. To overcome these obstacles, researchers are trying to pull the ribbons from the melt at a low angle, rather than straight up.

Using this general approach, Energy Materials Corporation of Massachusetts has demonstrated that they can pull ribbons of over 10 percent photovoltaic efficiency at 35 to 55 centimeters per minute. ''Although they are not out of the woods yet, this is a technology we have high hopes for,'' says Joe Milstein, a SERI photovoltaics researcher.

Meanwhile, Motorola and Shell have combined to commercialize a totally different approach. This starts by depositing silicon on another material, called a substrate. When the silicon is thick enough, it is separated from its backing. Then it is remelted with either a laser or an electron beam just enough so it solidifies in crystals large enough to produce 11 percent efficient cells. Although critics characterize the process as inherently expensive, the two companies believe that they should be able to produce PV devices at under $1 per watt.

Despite these projections, a number of experts in the field are doubtful that it will prove possible to achieve the 50 cent goal using silicon. Being an elemental substance, silicon has the advantage of stability. But, to get a decent efficiency, silicon cells must be several millimeters thick. To reach a really rock-bottom price, they argue, will require a ''thin film'' material that can convert a goodly fraction of the sunlight falling on it within a layer a few millionths of a meter thick.

One of the thin films showing considerable promise is an exotic alloy consisting of copper, indium, and selenium (CuInSe). The material was developed at the Boeing Company with federal funding, and cells made of it material have demonstrated a 10 percent efficiency. Other thin films have shown similar efficiencies but their performance has degraded rapidly over time. The CuInSe cells, on the other hand, have kept their efficiency through 2,000 hours of continuous exposure at elevated temperatures, John Gintz of Boeing reports. Although there are still a number of uncertainties in their economic analysis, Boeing appears optimistic that it will be able to build a plant and begin selling these cells for under $1 per peak watt in the late 1980s.

To get much higher photovoltaic efficiencies, scientists would like to sandwich different materials. Every photoelectric material works more efficiently at some wavelengths than at others. Higher efficiencies can theoretically be achieved by combining the right materials. Except for one or two special cases, however, distortions in the crystal structure caused by joining dissimilar materals have made these ''cascaded'' cells worthless. Now SERI researchers believe this problem has been solved, opening the door to cells of 35 percent or better efficiencies.''

This represents a sampling of the research and development currently being conducted in the PV field. Japan and a number of European nations are mounting major efforts in this area because they believe it will develop into a multibillion-dollar market by the end of the century. These results, as well as a number of others, suggest that the industry's optimism is fully justified. The American homeowner, the Japanese craftsman, and the East Indian villager all seem destined to have solar cells in their future.