Philip PietersSome of you may know the spectacular sight of solar farms in desert areas with hundreds of specialized PV panels that turn to the sun just like sunflowers do. These are concentrator PV panels: high-efficiency solar panels consisting of very small solar cells (typically only 1cm²) with integrated lens system to focus the sun on this small piece of high-tech PV technology. It’s an expensive piece of technology though, mainly because of the materials used. Concentrator PV cells consist typically of Ge, InGaP and GaAs, monolithically grown on top of each other. Each material covers a specific piece of the spectrum.

But while this multiple junction principle is the strength of the concentrator cell (the whole covers the major part of the spectrum of the sun light), it’s also a weakness. During a day the light conditions vary – for example at sunrise, sunset, a cloudy passes by – and in such circumstances one of the stacked layers will not work optimally and become a resistive burden for the others. The global efficiency of the concentrator cell will consequently drop. To minimize these effects, concentrator PV is currently limited to areas on earth where sun conditions are as stable as possible, e.g. in desert areas. But even for these regions spectral variations reduce the yearly energy output.
What if we could make concentrator PV less dependent on these fluctuations in the spectrum? Not only would their yearly output would increase in desert areas, they would also become economically viable in Southern Europe, California etc.

The solution came when scientists mechanically stacked the GaAs, InGaP and Ge layers instead of growing them monolithically on top of each other. In this way the layers (solar cell junctions) don’t need to be connected in series anymore, and because of the mechanically stacking the contacts of each cell can be coupled out separately. In this way the less performing cell in the stack does anymore not limit the current generation in the other cells. Simulations show that (e.g. in Las Vegas weather conditions) up to 20 percent more electricity generation over a year time may be possible.

With this new concept, it also becomes possible to rethink the material stack. In today’s triple junction monolithical stacks the spectral coverage is not optimal. The materials (GaAs, InGaP and Ge) were chosen because of the possibility to grow the layers on top of each other. Compromises were worked out to fit the crystal lattices of the materials. But when the layers are separated, it becomes possible to optimize them individually to cover a larger part of the sunlight’s spectrum with less overlapping between the layers. Again, this would mean an increase in overall efficiency for concentrator PV. One could even think of developing and combining cells of completely different materials in the future, further optimizing spectral coverage.

This alternative concept for concentrator PV and the excellent results open the door towards the next-generation concentrator panels. If the industry wants to take on the challenge to adopt the technology and complete it with the appropriate optics, a broader application of concentrator PV will become possible. For example it could be used in sunny areas to set up small energy plants where a significant amount of energy can be generated on a small footprint. Or maybe you can even put it in your garden along with some beautiful sun flowers.

Philip Pieters, PhD, business development director Energy, imec (
Philip Pieters received a masters degree and Ph. D. in electrical engineering from the Katholieke Universiteit Leuven in Leuven, Belgium. He joined imec in 1994 doing pioneering R&D work for innovative heterogeneous integration and RF-SIP technologies. Today, he is business development director Energy, creating the bridge between imec’s hightech research on PV technologies and the market needs.