Scientists in the US have shown how to multiply the power output of photovoltaic (solar) cells by up to ten times using organic dyes to concentrate sunlight. They say that their work could be scaled up to make solar cells competitive with fossil-fuel power generation.

Conventional photovoltaic concentrators are designed to increase the power output from a solar cell by collecting sunlight over a large surface area and then focusing it down, using large moving mirrors to track the Sun. Unfortunately the mirrors are expensive to set up and maintain, while the solar cells themselves must be cooled.

An alternative approach is to use “luminescent solar concentrators”. These generally consist of a block of transparent plastic containing dye molecules with solar cells placed at the ends of the block. The dye molecules absorb incident light and then reemit the light at longer wavelengths, leaving most of the light trapped inside the block through total internal reflection (in this way the block acts as a waveguide). The light is then collected by the solar cells.

However, in devices built to date much of the extra light collected has been lost via absorption in the dye. The “geometric gain” of such devices is the ratio of the areas of the concentrator and solar cell — but according to Marc Baldo of the Massachusetts Institute of Technology in the US, researchers have struggled to achieve an overall gain (or “flux gain”) of 1, rendering the technology pretty much useless.

Thin films

Baldo and colleagues have therefore taken a slightly different approach, by depositing thin films of two different organic dyes — one called Pt(TPBP) and the other DCJTB — onto glass. Along with a number of other technical innovations, this has allowed them to achieve higher gains (Science 321 226).

“In previous devices the dye absorbed too much of its own radiation,” explained Baldo. “This means that there was a limit in the size of the concentrator — light simply could not propagate far enough to reach the edges of larger concentrators,” he said.

The team overcame this problem by borrowing an idea from lasers known as a ‘four level system’, in which the absorption of the concentrator is separated in energy from the emission.

These higher gains rely on higher concentrator efficiencies — the ratio of electrical power out to solar power in. The MIT researchers say that previous demonstrations of luminescent concentrators have yielded efficiencies of the order of 1%, whereas they calculate that their Pt(TPBP) device operates at 4.1% while the DCJTB runs at 5.9%. Furthermore, they believe these figures could be increased by creating “tandem” concentrators, which are able to use a wider spectral range of the incoming light. They calculate that Pt(TPBP) combined with a DCJTB would yield 6.8%, while concentrators coupled with additional solar cells could push the efficiency up to over 20%.

In the demonstrations thus far, Baldo’s group has built concentrators with a geometric gain of 45. Taking into account both the efficiency of the concentrator and that of the solar cell, they calculate that their DCJTB concentrator generates a flux gain of 11. However, this would still leave solar power far more expensive than fossil fuels. Baldo points out that to be competitive solar must come down to about a dollar per peak watt, but that the solar cells used in concentrators produce power at about $50 per peak watt. Concentrators must therefore yield flux gains of around 50.

Very cheap to make

Fortunately, the MIT researchers predict that Pt(TPBP) could yield flux gains of between 30 and 60 for geometric gains of 630. Baldo adds that the concentrators themselves will be very cheap to make, and therefore should not add significantly to the price. “The paint industry alone makes thousands of tons per year of these kinds of dyes at very low cost,” said Baldo.

However, Martin Green, who carries out research on solar cells at the University of New South Wales in Australia, believes that the MIT work is still some way from producing useful devices. “The projected 6.8% efficiency is not yet high enough for major commercial impact,” he said.