Creating the Building Blocks for Next-Generation Batteries
With over a trillion tons of carbon dioxide in the atmosphere and global temperatures expected to rise anywhere from 2 to 9.7 degrees Fahrenheit (1.1 to 5.4 degrees Celsius) over the next 80 years, the transition from fossil fuels to renewable energy is a pressing issue that requires immediate attention. Humanity will require whole new energy storage technologies to complete the change.
The current standard, lithium-ion batteries, use volatile electrolytes and can only be recharged around 1,000 times before their capacity is drastically decreased. Other prospective successors are dealing with their own problems.
Long needle-like abnormalities called dendrites emerge whenever electrons are shuttled between the anode and cathode of lithium metal batteries, for example, resulting in a limited lifespan.
Such perplexing chemistry, according to Chibueze Amanchukwu, Neubauer Family Assistant Professor of Molecular Engineering at the University of Chicago's Pritzker School of Molecular Engineering, boils down to one incorrect and generally neglected process: current electrolyte design.
“The current approach to battery design, specifically with electrolytes, works like this: I want a new property, I look for a new molecule, and I mix it together and hope that it works,” Amanchukwu explained. “But because battery chemistries are always changing, it becomes a nightmare to predict what new compound you should use out of the million possible options. We want to demystify the dark art of electrolyte design.”
The third key component within a battery is the electrolyte, which is a specific material, usually a liquid, that permits ions to pass from the anode to the cathode. An electrolyte, on the other hand, must meet a large list of specific requirements, such as adequate ionic conductivity and oxidative stability, which are made even more difficult by the millions of possible chemical combinations.
Amanchukwu and his colleagues intend to catalog as many electrolyte components as possible, making it easy for any researcher to develop, manufacture, and describe a multifunctional electrolyte that meets their demands. They compare the technique to a popular building toy.
“The beautiful thing about Legos, and the aspect we’re going to replicate, is the ability to build different structures out of individual pieces,” Amanchukwu said. “You can use the same 100 Lego pieces to build any number of structures because you know how each piece fits together—we want to do that with electrolytes.”
How to catalog a million components
Amanchukwu starts with the archives to make his electrolyte building blocks. Electrolytes have been studied by scientists for over a century, and their data is open to anybody willing to comb through it.
To scrape data from scientific publications, Amanchukwu and his team employ "natural language processing," a form of machine learning software. Researchers create and analyze promising chemicals using methods such as nuclear magnetic resonance (NMR), a relative of MRI, to better understand their characteristics and optimize them even more.
The chemicals are then placed in actual batteries and evaluated again, with the results being transmitted back into the system.
The ultimate result is a database of electrolyte components that may be mixed and matched to meet specific requirements. A system like this would significantly speed up the creation of new batteries, but its influence would be seen well beyond that.
Electrolytes are now used in two ways in carbon capture technologies. During the capture phase, one electrolyte works as a solvent to aid in the separation of carbon dioxide from the air, while a second electrolyte aids in the conversion of carbon dioxide into a useful product such as ethylene.
This procedure, however, consumes a lot of energy. Amanchukwu believes that an electrolyte with the correct properties may combine both procedures, collecting CO2 while also turning it into a beneficial product.
