[tt] efficient, cheap solar cells

[Eugen Leitl]

http://www.technologyreview.com/printer_friendly_article.aspx?id=21405&channel=biztech&section=

Tuesday, September 23, 2008

Efficient, Cheap Solar Cells

New materials for high-performance cells could make solar power affordable.

By Kevin Bullis

A cheap new way to attach mirrors to silicon yields very efficient solar
cells that don't cost much to manufacture. The technique could lead to solar
panels that produce electricity for the average price of electricity in the
United States.

Suniva, a startup based in Atlanta, has made solar cells that convert about
20 percent of the energy in the sunlight that falls on them into electricity.
That's up from 17 percent for its previous solar cells and close to the
efficiency of the best solar cells on the market. But unlike other
high-efficiency silicon solar cells, says Ajeet Rohatgi, the company's
founder and chief technology officer, Suniva's are made using low-cost
methods. One such method is screen printing, a relatively cheap process much
like the silk-screen process used to print T-shirts.

So far, the high cost of solar cells has limited them to a marginal role in
power production, accounting for less than 1 percent of electricity
worldwide. Rohatgi calculates that the company's low-cost manufacturing
techniques will make solar power competitive with conventional sources,
producing electricity for about 8 to 10 cents per kilowatt-hour--the average
cost of electricity in the United States and far less than prices in many
markets.

Suniva's cells are efficient largely because they can trap light, keeping
photons inside the active material of the solar cell until their energy can
be used to free electrons and generate an electrical current. The basic
concept of trapping light is not new. It relies on texturing the front
surface of the layer of silicon that forms the active material of the solar
cell. The texturing creates facets that redirect incoming light, refracting
it so that, instead of passing directly through the silicon, it travels along
the length of the silicon layer. The photons thus stay in the material longer
and have a better chance of being absorbed by atoms in the material. When
that happens, the energy in the photons can free electrons that are used to
generate current.

Light trapping can be enhanced by pairing the textured surface with a
reflective layer at the back of the silicon layer. The mirror keeps the light
in the solar cell still longer, further increasing the number of freed
electrons. As a consequence, the silicon can be half its ordinary thickness
while absorbing the same amount of light. Using less of an expensive material
reduces costs directly. But it also allows solar-cell makers to make do with
cheaper, less pure forms of silicon. In a conventional solar cell, which can
have a silicon layer 200 micrometers thick, impurities within the material
can easily trap electrons before they reach the surface and escape to
generate a current. In a layer of silicon just 100 micrometers thick,
however, the electrons have a shorter distance to travel, so they're less
likely to encounter an impurity before they escape. Lower-grade silicon is
much cheaper and easier to make than the highly refined silicon ordinarily
used in solar cells.

Some companies have already introduced products that pair a textured front
with a mirrored back, and the technique has been demonstrated to work well in
laboratories for years. But adding the reflective layer typically requires
expensive processing and lithography. Rohatgi has developed proprietary
materials that can be incorporated into the solar cells using screen
printing. This, along with other advances that simplify the manufacturing
process, allowed the company to produce highly efficient cells at a low cost.

Tonio Buonassisi, a professor of mechanical engineering at MIT, says that
Suniva's new solar cell is "exciting" because "it's a demonstration that some
of the high-efficiency technologies that have been worked on for years in the
laboratory can be applicable in the marketplace." He says that Suniva's
decision to use such technologies is a risk that most other solar-cell
companies have been avoiding. Now that Suniva has developed a way to apply
these techniques cheaply, he predicts that other solar-cell companies could
be forced to do likewise to compete.

To be sure, significant work remains before the goal of 8 to 10 cents per
kilowatt can be achieved. Suniva has demonstrated the crucial first step,
which is to show that it can make solar cells that are more than 20 percent
efficient using screen printing. The results have been confirmed by the
National Renewable Energy Laboratory, in Golden, CO. But for those tests,
Suniva used cells with 200-micrometer-thick silicon wafers, and reaching 8
cents a kilowatt will require 100-micrometer wafers. That this is technically
possible has been established. The challenge lies in acquiring large amounts
of such silicon, since wafers that thin aren't commercially available,
Rohatgi says. What's more, factories will need to be retooled to handle
100-micrometer cells, which machines designed to handle thicker wafers could
break.

The company's priority now is to scale up production of its highly efficient
200-micrometer cells, which could still lower the cost of solar power. Once
it has established high-volume manufacturing, the next step is to introduce
thinner wafers, bringing down the costs yet further.

Copyright Technology Review 2008.