Published online 14 February 2010 | Nature | doi:10.1038/news.2010.69


Silicon whiskers catch rays well

New device could make solar cells cheaper.

solar cellsTiny rods of silicon (top) could lead to cheaper solar cells (bottom).Credit: T. TAKAHARA / SCIENCE PHOTO LIBRARY

Roll out the micro-carpet — a new solar-cell design based on a blanket of silicon rods could produce electricity at a fraction of the cost of conventional solar devices.

The carpets have yet to be made into a working solar cell, but preliminary measurements of their ability to absorb light and generate current suggest they could become a cheap replacement for existing technology.

The idea behind photovoltaic solar cells is straightforward: when sunlight strikes a material, it dislodges electrons, which start to flow in one direction. The electrons leave behind empty, positively-charged 'holes' that move the other way, effectively creating a current.

It's a simple idea, but it isn't cheap. The material of choice for cells is often thin silicon wafers, which are efficient at absorbing light but expensive to produce.

Now Harry Atwater and his colleagues at the California Institute of Technology in Pasadena have found an alternative that uses one hundredth of the material of current silicon technology. Using a well-established technique for assembling nanowires on a surface, the researchers grew a 'carpet' of micrometre-scale silicon rods lined up like hairs standing on end. They then embedded the rods in a transparent polymer.

Absorbing science

This configuration alone is not enough to absorb light efficiently. Although the array can absorb light coming in at shallow angles — the angle at which early morning sunlight might strike a solar roof panel, say — light beating down from directly overhead tends to pass between the rods and is lost. "Not being able to absorb light at noon isn't a great property for a solar cell," says Atwater.

To solve this problem, the team sprinkled aluminium oxide particles into the transparent polymer. These particles scatter incoming light so that it bounces around inside the array, increasing the light's chances of striking a silicon wire. As a result, up to 85% of usable incident sunlight can be absorbed effectively, says Atwater. The team reports its results in Nature Materials1.


Atwater is pleasantly surprised to find that the silicon carpet outperforms conventional wafers at absorbing infrared frequencies. The result, says Atwater, flies in the face of the "established dogma" that the most highly absorbing surface is a roughly textured wafer. "We've broken through what people thought was the absorption limit," he says.

"This is a beautiful demonstration and a striking set of results," says physicist Ken Durose, an expert on solar cells at Durham University, UK. "The fundamental issue with all solar cells is that they are too expensive, and reducing the material and the cost, but keeping the efficiency, has been the main thrust of research in this field."

Not so fast

The next step for Atwater's team will be to make a fully functional solar-cell device from their silicon array, says Durose. He notes that a group based in St Petersburg, Russia, recently built a solar cell based on a nanowire array made from gallium arsenide — a more expensive alternative to silicon2.

Atwater says that his team is now attempting to scale up their creation. If successful, their silicon-based technology has the advantage that it should be relatively easy to incorporate into current manufacturing processes. "The tools and facilities we need to build the wires are already in use," Atwater says.

However, Keith Barnham of the Quantum Photovoltaics Group at Imperial College London says that it is too early to say whether the little carpets really will work. Material defects are a big problem when building a solar cell because they prompt separated electrons and holes to 'recombine', reducing the cell's performance. Atwater's team must demonstrate that this is not a significant issue in devices made using silicon wires, says Barnham. "By demonstrating that they can get an efficient current, the group are halfway there though, and that's great." 

  • References

    1. Kelzenberg, M. D. et al. Nature Mater. doi:10.1038/NMAT2635 (2010).
    2. Cirlin, G. E. et al. Nanoscale Res. Lett. 5, 360-363 (2010). | Article | ChemPort |
Commenting is now closed.