Researchers at Northwestern University (Evanston, IL) have used mathematics and machine learning to design an optimal material for light management in photovoltaic cells, then fabricated the nanostructured surfaces simultaneously with a new nanomanufacturing technique called wrinkle lithography.1
The fast, highly scalable, streamlined method could replace cumbersome trial-and-error nanomanufacturing and design methods.
Nanophotonic materials and surface treatments are especially useful for light absorption in ultrathin, flexible solar cells. (Other potential uses include anti-wet surfaces and dyeless color in clothing.) For solar cells, the ideal nanostructure surface has quasirandom structures, which appear random but do have a pattern. Designing these patterns can be difficult and time consuming, since there are thousands of geometric variables that must be optimized simultaneously to discover the optimal surface pattern to absorb the most light.
To bypass the issues of nanolithography, the researchers manufactured the quasirandom structures in amorphous silicon using wrinkle lithography, a new nanomanufacturing technique in which wrinkle patterns are rapidly transferred into different materials to realize a nearly unlimited number of quasirandom nanostructures. Formed by applying strain to a substrate, wrinkling is a simple method for the scalable fabrication of nanoscale surface structures.
"Importantly, the complex geometries can be described computationally with only three parameters -- instead of thousands typically required by other approaches," says Teri Odom, a professor of chemistry at Northwestern. "We then used the digital designs in an iterative search loop to determine the optimal nanowrinkles for a desired outcome."
The team demonstrated the concurrent design and manufacturing method to fabricate 3D photonic nanostructures on a silicon wafer for potential use as a solar cell. The resulting material absorbed 160% more light than other designs in the 800 to 1200 nm wavelength range -- a range in which current solar cells are less efficient.
Next, the team plans to apply its method to other materials, such as polymers, metals, and oxides, for other photonics applications.
Source: https://www.eurekalert.org/pub_releases/2017-08/nu-sda080417.php
REFERENCE:
1. Won-Kyu Lee et al., PNAS (2017); http://www.pnas.org/content/early/2017/07/26/1704711114.full