Whenever Swiss chemist Michael Graetzel attends a conference or visits a laboratory, he brings a suitcase with him. In it he keeps a small fan that is powered by a prototype of the solar cell that has made him famous - the Graetzel cell. During presentations he demonstrates the cell by powering the fan using the light from an overhead projector.

Inspired by photosynthesis

Nine years ago, Graetzel developed a novel way of converting light into an electric current. He was inspired by the way in which plants transform sunlight into other forms of energy. The Graetzel cell had an unprecedented efficiency of 11% and could be manufactured from relatively cheap materialss. The cell seemed to have everything going for it and yet nowadays almost all solar current is generated in classic photovoltaic cells that are based on silicon, which is a very expensive material.

Albert Goossens, associate professor of chemistry at the University of Delft in the Netherlands said: "Conventional solar cells are expensive, therefore no-one buys them. Consequently, production volumes remain low and their unit price is high."

Goossens is one of many researchers working to improve on the original Graetzel design and make it a viable alternative to silicon solar cells. He is optimistic about the future. "In 10 years the price of electricity from solar power will be comparable to that from a conventional power plant," he said.

Since the early 1970s, scientists have been trying to build an improved solar cell by imitating nature (see box). "The idea behind the system is simple. You put a layer of strongly absorbing dye molecules on the surface of a semiconductor. Once the dye is excited, it delivers an electron into the conduction band. In this way you are no longer dependent on the absorption ability of the semiconductor itself," said Goossens.

The electrons move slowly through the layer of dye molecules, so this has to be very thin. However, the slow movement of the electrons implies that the dye's absorption decreases, so fewer electrons are produced. The energy efficiency of the first dye-sensitized solar cells was therefore never more than 0.01%. It was Graetzel who found a way out of this deadlock. Instead of using one big semiconductor crystal of titanium dioxide (TiO₂), he took a large number of small TiO₂ colloidal particles.

"Graetzel applied the TiO₂ paste in a thin layer over a glass plate" said Goossens. By heating the TiO₂ layer,a porous material is obtained, which is then soaked in a dye for several hours. The porous layer acts like a sponge and the sintered particles become coated in a 1 mm thick layer of dye.

These particles are small - no more than 20 nm - so there is a large surface area that is available to absorb light. A standard Graetzel cell consists of a liquid electrolyte that is sandwiched between two transparent conducting electrodes. This set-up gives the dye molecules a supply of electrons and results in a high energy efficiency. Graetzel recently improved on the cell's performance. He said: "I am very excited about a new dye that gives better results than the one that we used in 1991. We are currently working on a promising new generation of organic transfer dyes."

Organic solar cells are not without their problems. It can be difficult to manufacture the cells in a reproducible and reliable way

Reliable mass-productIon

Jan Kroon of the Energy Research Foundation (FCN) in Petten, the Netherlands, works on ways in which to improve the organic-cell manufacturing process. A few years ago he and his colleagues succeeded in making ultrathin Graetzel cells on a flexible polymer film. They have also produced large batches of cells using a screen-printing method on glass, thus making the first step towards industrial production. Kroon said: "The first low-power applications we have in mind are all indoors, in consumer electronics, for example. Flexible solar cells may be used in disposable electronics, such as electronic price tags."

The main problem with organic solar cells is their stability. A solar cell should be able to function for 10 years in full sunlight without failure. However, working temperatures can be high, which can induce mechanical stress. Also, organic solvents are generally not very stable lin light and the presence of so many electrons can initiate decomposition reactions.

Kroon explains that to some extent it is possible to circumvent these problems by improving the manufacturing process. "The transfer time between the various steps has to be fast, so it is best to work under a dry atmosphere where the temperature profile during sealing can be well controlled. Also, the chemical stability can be improved by using additives" he said.

For example, the addition of magnesium iodide effectively removes the cell's sensitivity towards ultraviolet light. "We have developed accelerated test procedures, from which we can conclude that a lifetime of 10 years is indeed possible, with only a minor reduction in power," said Koon.

Despite these promising results, a lot of research is geared towards finding solid-state substitutes for the liquid electrolyte. It was Graetzel himself who first achieved this two years ago, but his amorphous spirobifluorene compound has now almost been abandoned for a whole class of conductive polymers, such as poly(phenylene vinylene) (PPV), which are also successfully applied in organic light-emitting diodes.

An organic solar cell is made up of several layers.

These polymers absorb light efficiently and, at the same time, function as electrolytes that conduct charge carriers. The idea sounds attractive, but it is difficult to achieve in practice. Goossens said: "The pores in the sintered layer are the same size as the TiO2 particles. However, the polymer molecules are much larger. To try to get the polymer molecules into the TiO2 pores, we have to drop a fine mist of TiO2 particles onto a fast-rotating disc, ou which we have put a small drop of polymer solution."

So far there has been little progress. Other scientists are trying to eliminate the use of TiO2 completely. A group of chemists from the Uuiversity of California in Santa Barbara in the US, have made a solar cell using a blend of a conductive polymer and the new carbon allotrope, Buckminster fullerene. They managed to bring these two compounds into close contact ou a molecular scale. However, this proved not to be of any advantage: the electrous that are produced have to be collected on one of the electrodes so that they do not get lost in the chaotic network structure that is formed by the fullerene molecules.

Chemists are now trying to find ways of ordering the fullerene molecules within the polymer matrix, by fixing them onto a polymer chain like beads on a string, for example. The energy efficiency of these kinds of cells varies between 2 and 3%.

Researchers from Toshiba Laboratories in Japan have achieved higher efficiencies. Researchers at the Netherlands' Enrgy Research Centre (ECN) have produced large batches of cells using a screen-printing method an glass, bringing this technology a step clser to industrial production.

They have made a solid-state cell based on a polymer-gel electrolyte. Graetzel said: "The cell does not contain organic solvents anymore. To achieve a yield of more than 7% with a solid electrolyte is amazing and constitutes a real breakthrough."

Other groups investigating organic solar cells include the Bell Labs' group in the US, which recently developed the first electrically pumped solid-state organic laser (OLE September p6). At the beginning of this year, Bell Labs developed a solar cell that was based on the organic semiconductor pentacene and had an efficiency of more than 4%. One of the researchers, Jan Hendrik Schön, believes that it is still too early to think of potential applications. He said: "We do not know enough about the stability of our cells: for example, the effect of ultraviolet light on pentacene."

The advanced high-vacuum techniques that are required for purifying the materials might be a problem for widespread applications. Schön disagrees: "Recently, we demonstrated thin-film devices on flexible plastic substrates with efficiencies that exceeded 2%. The electrical properties of these optimized films are similar to those of good single crystals. The scale-up of this deposition technique is possible, but needs a significant amount of engineering."
en cours
After 10 years of intensive research there is still a lot to be done. This may be the reason why the scientists that are involved are
reluctant to make predictions. Graetzel told OLE: "Both the power-generation modules and the cells for indoor applications are making rapid progress. What is crucial at this stage is to develop cheap production facilities that allow for a real cut in price. A module for USD 2 per peak watt would be a marvellous goal to achieve."

Experience shows that the time span between an initial idea and its realization into a commercial product is often 10 to 15 years. The next few years will decide on the fate of the Graetzel celi. In the meantime, Graetzel will continue to carry around his prototype cell in his suitcase.

Box: Imitating nature At the heart of every solar cell is a semiconductor. Absorption of photons effectively frees electrons from atoms by promoting the electrons from the valence band into the conduction band. In order for this to happend, photons require a certain minimum energy. Any energy that is in excess of this minimum is lost. Owing to these two effects, even in the best material, 50% of the light is wasted. In principle, many species of bacteria and all green plants suffer from this same problem because they depend on the Sun as their source of energy. However, after millions of years of evolution, plants have come up with an ingenious solution.  In order to prevent unnecessary losses, the processes of the absorption of light and the conversion of the absorbed light energy into electrons have been separated. Light is harvested by chlorophyll molecules that have an absorption region that fits the solar-emission spectrum. Once the energy is absorbed, it is transferred to a protein complex in which charge separation takes place. This ultimately drives the chemical reactions that convert carbon dioxide and water into carbohydrates and oxygen.