Researchers have developed a new filtering system designed to reduce carbon dioxide emissions from electric power station smokestacks.

The design, presented at the 246th National Meeting and Exposition of the American Chemical Society, found its muse in an unlikely place: the lungs of birds and swim bladders of fish.

With climate change a growing concern, many power plants currently rely on CO2 capture and sequestration methods to reduce greenhouse gas emissions. In a lecture presented at the conference, Aaron P. Esser-Kahn said he envisions the new CO2-capture units as containing arrays of tubes fitted side-by-side in a way that reflects the blood vessels in a natural lung.

To capture the most CO2 possible, Esser-Kahn and a group from the University of California, Irvine, had to first determine the best pattern to fit two sets of tubes of different sizes (those for waste emissions and others for a CO2 absorption liquid) in the unit.

In search for inspiration, the team examined the arrangement of the blood vessels in the avian lung and fish swim bladder, both of which are built to accommodate gas exchange. They found that avian lungs contain a hexagonal pattern while the fish bladder boasts a squarer formation.

Using computer simulations, the group of researchers examined the efficiency of these two patterns, as well as seven others. They came down to four in particular that they predicted to be highly efficient. Among these finalists were the avian lung's hexagonal pattern and the fish swim bladder's square pattern.

In the end, the most efficient appeared to be one that doesn't exist in nature. Called the "double-squarer" pattern, it is similar to the fish swim bladder. However, whereas the bladder has a pattern in which a large and small tube alternate between vertices of a square, the "double squarer" contains two small tubes alternating with one large tube. All told, this new pattern outperforms the avian lung and fish bladder by almost 50 percent.

"Biological systems spent an incredible amount of time and effort moving towards optimization," Esser-Kahn said. "What we have is the first step in a longer process."