This article is reproduced with the permission of New Scientist for exclusive use by Nova users.

Solar power: The future's bright
08 December 2007
From New Scientist Print Edition.
Bennett Daviss
Building for the future
Boom times ahead
Super cell
Sun traps

In theory, solving the world's energy problems should be pretty straightforward. Locate a piece of sun-drenched land about half the size of Texas, find a way to capture just 20 per cent of the solar energy that falls there and bingo - problem solved. You have enough power to replace the world's entire energy needs using the cleanest, most renewable resource there is.

Can it really be that easy? For years, supporters of solar power have heralded every new technical breakthrough as a revolution in the making. Yet time and again it has failed to materialise, largely because the technology was too expensive and inefficient and, unlike alternatives such as nuclear and wind power, no substantial subsidies were available to kick-start a mass transition to solar energy. This time things are different. A confluence of political will, economic pressure and technological advances suggests that we are on the brink of an era of solar power.

The prospect of relying on the sun for all our power demands - conservatively estimated at 15 terawatts in 2005 - is finally becoming realistic thanks to the rising price of fossil fuels, an almost universal acceptance of the damage they cause, plus mushrooming investment in the development of solar cells and steady advances in their efficiency. The tried-and-tested method of using the heat of the sun to generate electricity is already hitting the big time (see "Hot Power"), but the really big breakthroughs are happening in photovoltaic (PV) cells.

Ever since the first PV cell was created by Bell Labs in 1954, the efficiency with which a cell can convert light into electricity has been the technology's Achilles' heel. The problem is rooted in the way PV cells work. At the heart of every PV cell is a semiconducting material, which when struck by a photon liberates an electron. This can be guided by a conductor into a circuit, leaving behind a "hole" which is filled by another electron from the other end of the circuit, creating an electric current (see Diagram).

Photons from the sun arrive at the semiconductor sporting many different energies, not all of which will liberate an electron. Each semiconducting material has a characteristic "band gap" - an energy value which photons must exceed if they are to dislodge the semiconductor's electrons. If the photons are too weak they pass through the material, and if they are too energetic then only part of their energy is converted into electricity, the rest into heat. Some are just right, and the closer the photons are to matching the band gap, the greater the efficiency of the PV cell.

Bell Labs discovered that silicon, which is cheap and easy to produce, has one of the best band gaps for the spectrum of photon energies in sunlight. Even so, their first cell had an efficiency of only 6 per cent. For a long time improvements were piecemeal, inching up to the mid-teens at best, and at a cost only military and space exploration programmes could afford. The past decade has seen a sea change as inexpensive cells with an efficiency of 20 per cent have become a commercial reality, while in the lab efficiencies are leaping forward still further.

Last year, Allen Barnett and colleagues at the University of Delaware, Newark, set a new record with a design that achieved 42.8 per cent energy conversion efficiency. Barnett says 50 per cent efficiency on a commercial scale is now within reach. Such designs, married to modern manufacturing techniques, mean costs are falling fast too (see Diagram).

As a result, in parts of Japan, California and Italy, where the retail price of electricity is among the world's highest, the cost of solar-generated electricity is now close to, and in some cases matches, that of electricity generated from natural gas and nuclear power, says Michael Rogol, a solar industry analyst with Photon Consulting, based in Aachen, Germany. For example, in the US the average price of conventionally generated electricity is around 10 cents per kilowatt-hour. The cost of solar-generated electricity has fallen to roughly double that. This has created a booming market for PV cells - now growing by around 35 per cent annually - and private investors are starting to take a serious interest. The value of stocks in companies whose business focuses primarily on solar power has grown from $40 billion in January 2006 to more than $140 billion today, making solar power the fastest-growing sector in the global marketplace.

George W. Bush has acknowledged this new dawn, setting aside $168 million of federal funds for the Solar America Initiative, a research programme that aims to make the cost of PV technology competitive with other energy technologies in the US by 2015. Rogol thinks Bush's target is achievable. He says the cost of manufacturing PV equipment has fallen to the point where, in some places, PV-generated electricity could already be produced for less than conventional electricity. Manufacture PV cells at $1 per watt of generating capacity and the cost should be competitive everywhere.

Perhaps surprisingly, given its less than cloudless skies, one of the countries leading the solar revolution is Germany. In November 2003, amid rising oil and gas prices and growing concern over global warming, its parliament agreed a "feed-in tariff" programme, which guarantees a market for solar power. Anyone who produces electricity from solar power can sell it to the national grid for between ¬0.45 and ¬0.57 per kilowatt-hour, which is almost three times what consumers pay for their electricity, roughly ¬0.19 per kilowatt-hour. Germany's power-generating companies are required by law to pay this premium, which is guaranteed until 2024. This guarantee has spurred enterprising individuals to invest in solar panels, confident of earning back the cost of their systems and possibly turning a profit. Today there are over 300,000 PV systems in Germany, mostly on the rooftops of homes and small businesses, and Germany is the world's fastest-growing PV market. It has 55 per cent of the world's installed base of PV panels and can generate around 3 gigawatts of electricity from solar energy, equivalent to between three and five conventional power stations.

Last year, following in Germany's footsteps, Italy and Spain launched their own tariff programmes, while the California Solar Initiative earmarked $2.8 billion for cash incentives that will subsidise new PV installations to the tune of up to $2.50 per watt, with the aim of creating 3 gigawatts of capacity by 2016. By the end of 2008, 20 nations will have similar tariff programmes for solar power, Rogol predicts.

The hope is that by spurring demand, these subsidies will also stimulate PV research and manufacturing technology, driving down costs. This may help speed the development of existing PV technologies, but could also drive the industry down a blind alley, as silicon PVs may soon reach their theoretical efficiency limit of about 30 per cent. Yet according to Martin Green at the University of New South Wales, Australia, it should be possible to create cells from other materials with a 74 per cent efficiency limit. And while subsidies go some way to stimulating the market, most analysts agree that the cost of existing PV cells is too high for the technology to hit the mainstream.

That's why researchers have been looking at alternative designs. One of the cheapest cells to manufacture is the thin-film cell, in which semiconductor compounds are sprayed onto a flexible substrate. Thin-film cells use as little as 1 per cent of the volume of materials that ordinary PV cells demand, and the band gap of the cells can be improved by adjusting the proportions of the ingredients that form the film. For example, cells that use a low-cost blend of copper, indium, a pinch of gallium, and selenium (CIGS), have already achieved an efficiency of around 19 per cent in lab tests. The material's efficiency is so high relative to its cost that researchers have shifted their attention from boosting CIGS's photon-collecting power to slashing the cost of producing the cells. This could enable the technology to deliver grid-competitive electricity within five years.

Grand designs

One innovation aimed at improving mainstream solar-cell design is the use of lenses to focus and amplify the amount of light hitting the PV material. Among the most successful designs to incorporate a concentrating lens is one created by Soliant Energy, a California start-up company staffed by scientists formerly at NASA's Jet Propulsion Laboratory. Its PV module is a box holding rows of half-pipes, like gutters facing skywards. The trough of each half-pipe is lined with a strip of PV material, while the open side of the pipe is covered with an acrylic lens that concentrates sunlight by a factor of 500. This slashes the quantity of PV material required for a given power output, and thus the cost of the cell.

The company's next generation of PV modules will couple concentrators with PV crystals made by Spectrolab, Boeing's subsidiary which engineers PV materials for NASA space probes. With efficiencies of up to 40 per cent, these alternative materials are twice as good as current silicon PV cells, says Brad Hines, Soliant's founder and chief technology officer. The downside is that they cost 100 times as much, but Soliant has found a way of using just a sliver of the amount used in spacecraft solar cells, keeping them affordable.

Barnett's record-breaking cell also uses a concentrator, but it only needs to intensify the light by a factor of 20. The real breakthrough in Barnett's design is to split the incoming light into separate beams, each containing a narrow range of wavelengths. These are each directed into materials optimised to convert those frequencies into electricity.

Light entering the cell first falls onto PV material that absorbs high-energy wavelengths of light up to 500 nanometres. Longer wavelengths slip through to a dichroic mirror - a material that reflects certain wavelengths while allowing others to pass through it. Here, light with wavelengths of between 500 and 900 nanometres is reflected onto one photovoltaic stack, while wavelengths longer than 900 nanometres pass through the mirror and fall onto another stack (see Diagram).

Barnett's group is stepping back to let engineers at DuPont and other companies take on the task of producing a prototype. The US Defense Advanced Research Projects Agency has committed $33 million to the project, while cash from private investors could bring the total investment to as much as $100 million. Barnett says PV modules based on the new design could be up to 50 per cent efficient and should go on sale within five years, costing under $2 per watt.

There are yet more ambitious plans to build cheap, efficient PV cells. Several groups worldwide are now working with nanocrystals called quantum dots (see New Scientist, 27 May 2006, p 44) with the aim of developing low-cost PV cells with an efficiency of 42 per cent. The nanocrystal's special properties mean one photon of light will release up to four electrons.

Martin Green, one of the leaders in the field, is designing quantum dots to match specific light spectra and so make them more energy efficient. He wants to address a specific problem with conventional solar cells: some of the energy supplied by an incoming photon is lost as heat. Green is designing "hot carrier" cells that should transfer more of the energy from the photon to the electron, producing a higher output voltage. "In principle, a hot carrier cell would have an efficiency quite close to the 74 per cent limit," he says.

The challenge in creating a hot carrier cell is collecting the electrons quickly, before they move around the semiconducting material and lose their energy. That would mean creating a semiconductor shot through with nanowires or other collectors that would gather up electrons as soon as they are liberated from atoms - a requirement that could send manufacturing costs skyrocketing. "So far," Green acknowledges, "we don't know how to do that."

Allen Heeger, at the University of California, Santa Barbara, is trying another approach. The co-inventor of plastics that can conduct electricity, Heeger has created a semiconducting plastic which allows incoming photons to liberate electrons, just as in silicon and other photovoltaic materials. In July, Heeger unveiled a two-layer polymer PV stack that reached 6.5 per cent efficiency, a record for plastic solar cells.

The promise of plastic PV cells, Heeger points out, is that they could be manufactured using a kind of printing process "similar to the way newspapers are printed", because all the materials they are made of are soluble. "We have a goal of getting to 10 per cent efficiency and eventually well beyond that," he says, but he's not too bothered about efficiency. "The critical comparison is dollars per watt," he says. "Even if our efficiency is lower than silicon, the cost per watt could still be better because this is such a low-cost manufacturing process."

And perhaps that's where the true promise of solar power lies, not in expensive high-efficiency cells, but in clever new designs that are dirt cheap to produce. It has been a long slow revolution, but finally years of diligent research and investment by a group of true believers is beginning to pay off. Solar power has finally come of age.

Bennett Daviss is a science writer in New Hampshire

From issue 2633 of New Scientist magazine, 08 December 2007, page 32-37

Hot power

Exotic photovoltaic materials aren't the only way to convert sunlight into electricity. Over the past few decades, techniques have been developed that use mirrors to concentrate the sun's rays and convert the heat this generates into electricity on a commercial scale.

The most widely used so far is the solar trough concentrator system. Rows of parabolic mirrors track the sun, focusing its energy onto tubes filled with a fluid such as oil or water. The liquid is heated to around 400 °C and circulated to a conventional steam turbine to generate electricity. Solar trough systems convert roughly 20 per cent of the sun's heat they capture into electricity, comparable with some commercial photovoltaic cells, but at a fraction of the cost.

In the late 1980s, nine parabolic-trough energy farms were built in the Nevada desert. Together covering just over 1 square kilometre, they produced 354 megawatts of electricity. Plans to expand the farm were abandoned in the early 1990s when the price of fossil fuels slumped, but they kept supplying power to the grid. Now, with energy prices on the rise, plans are being drawn up to revive the technology. In December 2005 the first trough system built in the US since 1988 was switched on in Seguaro, Arizona. It is capable of generating 1 megawatt of power.

An alternative to the trough system is the solar tower, pictured above, in which a field of flat mirrors track the sun and reflect its rays onto water pipes. The water boils, generating steam that drives a turbine. This approach achieves an efficiency of around 15 per cent. The first commercial solar tower, with a capacity of 11 megawatts, was completed in 2005 near Seville in Spain. Construction of a second tower, capable of generating 20 megawatts, began last year.

New solar tower designs replace the water with molten salts (for example, a mix of sodium nitrate and potassium nitrate) which can be heated to 600 °C. The heat can be used immediately to produce steam to drive a turbine, or stored and used overnight or when clouds block the sun.

Another method uses dish-shaped mirrors around 10 metres in diameter to focus solar energy onto a Stirling engine, which contains a gas that expands under heating and so drives a generator. At 24 per cent, its efficiency beats all other solar concentrator systems. In 2005, the California Public Utilities Commission, the state body responsible for regulating private power stations, gave the go-ahead for the world's biggest solar dish concentrator farm to be built in the Mojave desert, north-east of Los Angeles. When completed in 2010, its 20,000-dish array will generate 500 megawatts.

For the latest from New Scientiist visit

Academy disclaimer: We cannot guarantee the accuracy of information in external sites.