PowerSource

In general, you have two options with lithium ion battery technology. You can optimize the battery chemistry for high-energy storage in the kind of batteries you’ll find in today’s laptops and cell phones. These li-ion cells rely on cobalt cathodes which have an unfortunate although rarely experienced ability to go into a thermal runaway condition and catch fire. Or, you can optimize for power – that is, the ability to charge and discharge rapidly. These are the batteries that GM is pursuing for its extended-range electric car, the Chevy Volt, which uses lithium ion iron phosphate batteries from companies like A123 and LG Chem. It’s almost impossible to get an lithium iron phosphate  battery to go into thermal runaway, but at the trade-off of much lower energy storage. [Update: Thanks to the alert reader who corrected me -  LG Chem uses a mangaese-based cathode.]

This past week, Yi Cui, assistant professor of materials science and engineering at Stanford, announced success in expanding the energy storage capacity of lithium ion batteries by using silicon nanowires for the anode, which in current batteries are typically made of carbon. Silicon’s advantage over carbon is that it has a much higher energy capacity. However, silicon's drawback has been that as it absorbs positively charged lithium atoms during charging, it swells and then shrinks during discharge, resulting in cracks. Cui’s research shows that while the silicon nanowires still expand and contract, they don’t fracture. Use of silicon nanowires results in a 10x increase in energy storage, with the potential to last for over 1,000 cycles. (Actually laboratory cycles are currently in the tens, rather than thousands of cycles.)

There’s an excellent interview with Cui on the GM-Volt.com site that asks him if his research could transfer over to the lithium iron phosphate battery chemistry used by A123 and LG Chem. Cui says that combining a current battery’s cathode materials with a silicon nanowire anode will significantly improve the performance. Cui says in the interview that he thinks the technology could be commercialized within five years.