When researchers at Chicago's Northwestern University were trying to figure out how to maximize the amount of time light spent inside their organic polymer-based solar photovoltaic cells, they turned to an age-old technique to refine their design: evolution by natural selection.
Northwestern researchers working on making thin-film solar cells more efficient recently reported the resultes of their studies in Scientific Reports, a publication of Nature.
The team was looking at ways to increase efficiency of very thin organic polymer PV cells, with active layers one hundredth of a micron thick. (A typical silicon wafer in a standard PV cell is 20,000 times thicker than that.) Making active PV layers that thin allows much cheaper solar cells, but at .01 microns the PV layer would actually be much smaller than the wavelength of the light it's trying to absorb -- even with cladding a tenth of a micron thick laminated on both sides.
In order to boost the likelihood that a light photon will actually be captured by the photovoltaic layer and turned into electrical power, researchers decided to add a scattering layer to deflect incoming light in all directions, potentially bouncing the light back and forth inside the film as well. In order to do that, the team had to decide on a scattering pattern.
Which is where "evolution" comes in, or at least simulated evolution. "We wanted to determine the geometry for the scattering layer that would give us optimal performance," Cheng Sun, assistant professor of mechanical engineering in Northwestern's McCormick School of Engineering and Applied Science and co-author of the Scientific Reports paper, said in a Northwestern press release. "But with so many possibilities, it's difficult to know where to start, so we looked to laws of natural selection to guide us."
Sun's team selected a number of computer generated patterns, analyzed how those patterns would scatter light, then "mated" the patterns and generated likely "offspring." Second-generation patterns that scattered light more thoroughly were selected and "bred" with other high-performing patterns. The computing algorithms included simulations of possible "mutations" and "crossover."
"Due to the highly nonlinear and irregular behavior of the system, you must use an intelligent approach to find the optimal solution," said Wei Chen, Wilson-Cook Professor in Engineering Design at McCormick. "Our approach is based on the biologically evolutionary process of survival of the fittest."
The resulting pattern, which Sun and Chen's team will be fabricating with Argonne National Laboratory, traps light for three times the span predicted in older models of how light and semiconducting materials interact. That's good news for solar cell efficiency: the longer the light is in the cell, the more likely it will be trapped and converted by the PV layer.