Using 'dirty silicon' to cut the cost of solar cells
Most everyone not vested in oil and gas agrees that renewable energies such as solar are a more sustainable option, but cost remains an issue. To make solar more competitive by addressing the high cost of solar cell production, researchers out of Norway have developed a method that could bring down the amount of silicon used per unit area by as much as 90 percent. The price of silicon is a major driver in the cost of solar panels.
Developed by Professor Ursula Gibson and PhD candidate Fredrik Martinsen from Norwegian University of Science and Technology (NTNU), the key is using solar cells made from low grade silicon – as much as 1000 times less pure than typical industry levels. The approach is also said to reduce the energy costs in manufacturing solar cells by using fewer production steps, and cheaper and fewer raw materials.
The process starts by preparing a glass preform (thick-walled tube) about 3 mm in diameter, then coating the inside with a layer of calcium oxide (CaO) and closing one end of the tube. After that, the coated preform is loaded with small rods of silicon and the closed end of the preform is heated to melt the silicon and soften the glass. Finally, the glass/silicon combination is stretched down to a thread-like diameter up to 100 times thinner. The silicon core has a diameter of around 100 micrometers, and is the active portion of the solar cell.
Anyone familiar with standard fiber optic cable production will recognize this as the same process. And it’s how Gibson says she came by her inspiration. "I became interested in making solar cells from fibers when the Clemson and Virginia Tech [university] groups published their work on the formation of silicon-core fibers using scalable techniques," she said.
The process results in a directional growth of crystals, which is a known technique for purifying silicon, and Gibson rightly believed they could use it to reduce the processing steps needed to make solar cells. At the same time, the design work by the Atwater research group at the California Institute of Technology on microwire solar cells suggested to Gibson that the size of the fibers would work well, and the combination inspired her to pursue the research. After some development work at NTNU, they partnered with Clemson University to produce large quantities of fiber, the first to make solar cells with these types of silicon-core fibers.
In a traditional solar cell, the distance can be relatively long from where a charge is generated to the surface where it’s captured; purified silicon facilitates the process. Otherwise, an un-captured charge dissipates and the energy heats up the cell instead of providing electrical energy. Gibson’s silicon/glass tubes shorten that distance, and the charge can be more easily captured even with lower grade silicon.
Unlike the manufacture of pure silicon wafers required for current solar cells, the NTNU process is less labor- and energy-intensive, and thus cheaper. The reason, according to Gibson, is their use of "dirty" silicon, which is purified naturally during the melting and re-hardening process.
How much expense will their process save over current silicon wafer manufacturing? "This is a tough question, because so much depends on the economies of scale (and because it is not my area of expertise)," she says. "But crude estimates suggest that we would reduce the energy input required to process the silicon by a factor of two to six, depending on what process is used to make the pure silicon wafers. We would also be replacing 50 to 90 percent of the silicon with glass, which would reduce the energy budget further."
Despite the savings, there is still plenty of work before the silicon/glass tubes are ready for a commercial market. The prototype created by NTNU clocks in at a 3.6 percent efficiency conversion rate, while the X-Series solar panel by SunPower – one of the market leaders in this respect – has a 21.5 percent efficiency rate.
Gibson adds that the anticipated form of composite wafer cells made from their technique could be handled in a regular solar cell assembly line, though it would represent a paradigm shift for wafer production. Until then, the group will work on improving efficiency, and assembling larger arrays of fibers to test their light gathering effectiveness.