Princeton University

School of Engineering & Applied Science

Applications of Large-Area Nanopatterning to Energy Generation & Storage Devices

Eric Mills
Engineering Quadrangle J401
Tuesday, March 22, 2016 - 3:00pm to 4:30pm

Wind & solar power will never be allowed to generate a substantial portion of the nation’s electricity until uncertainty in their power outputs can be addressed. “Grid-level storage” using batteries is currently a leading option, but Li-ion batteries are still incredibly expensive. Many avenues exist: make batteries better and cheaper, allow fuel to be generated from renewable sources, or improve distributed solar generation, among others.
We sought to utilize ordered nanopatterning by Nanoimprint Lithography (NIL) to target one of these options. Nanopatterning fundamentally alters a material’s surface, so we fabricated devices which are dependent on chemical reactivity: Li-ion battery electrodes, “Photocatalysts” which generate H2 from water, and Dye-Sensitized Solar Cells (DSSCs), which incorporate chemically reactive electrolytes.
Li-ion battery performance depends on that of their charge-storing electrodes. Silicon can store ~10x the charge per gram of the current leading material, but it undergoes dramatic (>300%) volume expansion upon charging, pulverizing any structure with non-nanoscopic dimensions (>250nm). We overcome this by creating large-area arrays of <100nm-diameter Si “nanopillars” on commercially-available metal substrates using flexible-mold NIL. We also introduce techniques for creating large-area arrays of high aspect-ratio Si “nanowires”, and for removing them from their mother substrate while leaving their array structures intact.
Photocatalyst materials use absorbed light to drive chemical reactions. In our case, TiO2, a common food & paint additive, can be used in combination with other materials to split water into gaseous hydrogen and oxygen, allowing solar power to generate fuel for later use. However, TiO2 can only absorb UV light, meaning its solar energy conversion efficiency is extremely low. It can be improved by integrating nanoscopic gold particles, which allow large portions of the visible spectrum to be absorbed, and improved electron mobility in TiO2.
DSSCs have been researched for 25 years as a cheap alternative to crystalline Si solar cells. They can be constructed on flexible substrates, giving them a tremendous advantage for small-scale distributed solar power generation. By nanopatterning an array of nanoscopic holes in metal films, we aim replace the costly Indium Tin Oxide as the transparent front electrode material.