Silicon solar cells 20 times thinner than commercial ones have been produced using 95 percent less silicon, reducing production costs considerably. The trick? Using a back sheet peppered with microbeads that fool light into staying longer within the solar cell. Standard commercial photovoltaics are made of 200-µm-thick plates sliced from a large block of silicon. To make several billion solar panels per year, large amounts of silicon are consumed – around 5 grams of silicon per watt of electricity produced – and much of the silicon is wasted: Roughly half of the block is turned into sawdust by the slicing process. Although the price of solar cells is falling steadily, manufacturing silicon pure enough to make these cells is costly. Aasmund Sudbø and Erik Marstein of the University of Oslo are now developing next-generation solar cells to address this challenge. Their solution is to make very thin solar cell slices without increasing costs. “The thinner the solar cells become, the easier it is to extract the electricity,” said Marstein, head of the Norwegian Research Center for Solar Cell Technology, head of research for the solar cell unit at the Institute for Energy Technology at Kjeller and an associate professor in the university’s department of physics. “In principle, there will therefore be a higher voltage and more electricity in thinner cells.” The team now is developing solar cells that are just as efficient as current ones, but that can be made with just one-twentieth of the silicon, reducing consumption by 95 percent, Marstein said. Professors Aasmund Sudbø and Erik Marstein of the University of Oslo have used all kinds of tricks with light to reduce the thickness of solar cells by 95 percent. Reducing the thickness of solar cells makes a lot of sense from a commercial point of view, but it introduces a significant issue: As the wafer gets thinner, more light passes straight through the silicon, reducing the amount of electricity produced by the photovoltaic effect. This has to do with the wavelengths of light. Blue light, which has a short wavelength, can be captured by a very thin wafer of silicon; however, red light, with a longer wavelength, requires a slab of silicon 1 mm thick. For infrared light, the plate must be even thicker. The thickness of a solar cell can in effect be doubled by using a mirror, which, by reflecting the light, doubles the passage of light through the plate. However, with this approach, some of the longer wavelengths still will not be trapped. Moreover, current mirrors are far from perfect: They reflect only 70 to 80 percent of the light. To trap the longer wavelengths, the Oslo researchers used some tricks. “This is where the magic comes in,” Marstein said. “We are trying every possible wonderful trick with light. Our trick is to deceive the sunlight into staying longer in the solar cell.” Their first trick was to cover an entire back sheet with Uglestad microbeads – very small, uniformly sized plastic spheres – to direct light to move sideways through the cell. “We can increase the apparent thickness 25 times by forcing the light up and down all the time,” Marstein said. “We have calculated what this back sheet must look like and are currently studying which structures work.” They have tried forcing the microbeads to lie closely together on the silicon surface, in an almost perfect periodic pattern. The team is also experimenting with asymmetric microindentations on the back of the silicon wafer. “Cylinders, cones and hemispheres are symmetrical shapes. We have proposed a number of structures that break the symmetry,” Marstein said. “Our calculations show that asymmetrical microindentations can trap even more of the sunlight.” The researchers are now investigating whether these methods can be scaled up to industrial production, and are confident that their technology could come to market within five to seven years. “Our main goal has been to get the same amount of electricity from thinner cells,” Sudbø said. “We will be very satisfied even if our new solar cells are 30 µm.”