IBM Research is beginning an ambitious project that it hopes will lead to the commercialization of batteries that store 10 times as much energy as today’s within the next five years. The company will partner with U.S. national labs to develop a promising but controversial technology that uses energy-dense but highly flammable lithium metal to react with oxygen in the air. The payoff, says the company, will be a lightweight, powerful, and rechargeable battery for the electrical grid and the electrification of transportation.

Lithium metal-air batteries can store a tremendous amount of energy–in theory, more than 5,000 watt-hours per kilogram. That’s more than ten-times as much as today’s high-performance lithium-ion batteries, and more than another class of energy-storage devices: fuel cells. Instead of containing a second reactant inside the cell, these batteries react with oxygen in the air that’s pulled in as needed, making them lightweight and compact.

IBM is pursuing the risky technology instead of lithium-ion batteries because it has the potential to reach high enough energy densities to change the transportation system, says Chandrasekhar Narayan, manager of science and technology at IBM’s Almaden Research Center, in San Jose, CA. “With all foreseeable developments, lithium-ion batteries are only going to get about two times better than they are today,” he says. “To really make an impact on transportation and on the grid, you need higher energy density than that.” One of the project’s goals, says Narayan, is a lightweight 500-mile battery for a family car. The Chevy Volt can go 40 miles before using the gas tank, and Tesla Motors’ Model S line can travel up to 300 miles without a recharge.

One of the main challenges in making lithium metal-air batteries is that “air isn’t just oxygen,” says Jeff Dahn, a professor of materials science at Dalhousie University, in Nova Scotia. Where there’s air there’s moisture, and “humidity is the death of lithium,” says Dahn. When lithium metal meets water, an explosive reaction ensues. These batteries will require protective membranes that exclude water but let in oxygen, and are stable over time.

IBM does not currently have battery research programs in place. However, Narayan says that IBM has the expertise needed to tackle the science problems. In addition to Oak Ridge, IBM will partner with Lawrence Berkeley, Lawrence Livermore, Argonne, and Pacific Northwest national labs. The company and its collaborators are currently working on a proposal for funding from the U.S. Department of Energy under the Advanced Research Projects Agency-Energy.

Research on lithium-metal batteries stalled about 20 years ago. In 1989, Canadian company Moli Energy recalled its rechargeable lithium-metal batteries, which used not air but a more traditional cathode, after one caught fire; the incident led to legal action, and the company declared bankruptcy. Soon after, Sony brought to market the first rechargeable lithium-ion batteries, which were safer, and research on lithium-metal electrodes slowed nearly to a halt. (After restructuring, Moli Energy refocused its research efforts and is now selling lithium-ion batteries under the name Molicel.) Only a handful of labs around the world, including those at PolyPlus Battery, in Berkeley, CA, Japan’s AIST, and St. Andrews University, in Scotland, are currently working on lithium-air batteries.

Safety problems with lithium-metal batteries can arise when they’re recharged. “When you charge and discharge, you have to electroplate and strip the metal over and over again,” says Dahn, who is not a contributor to the IBM project. Over time, just as in a lithium-ion battery, the lithium-metal surface becomes rough, which can lead to thermal runaway, when the battery literally burns until all the reactants inside are used up. But Narayan says that lithium-air batteries are inherently safer than previously developed lithium-metal batteries as well as today’s lithium-ion batteries because only one of the reactants is contained in the cell. “A lithium-air cell needs air from outside,” says Narayan. “You will never get a runaway reaction because air is limited.”

PolyPlus Battery has been working on lithium metal-air technology for about six years and has some dramatic evidence of the technology’s viability: floating among clownfish in an aquarium tank at the company’s headquarters, a lithium-metal battery pulls in oxygen from the salt water to power a green LED. The company has also developed a prototype battery that pulls oxygen from ambient air. But Steven Visco, founder and vice president of research at the company, says that lithium metal-air batteries are “still a young technology that’s not ready to be commercialized.”

IBM’s Narayan points to two remaining major problems with lithium metal-air technology. First, the design of the cathode needs to be optimized so that the lithium oxide that forms when oxygen is pulled inside the battery won’t block the oxygen intake channels. Second, better catalysts are needed to drive the reverse reaction that recharges the battery.

Narayan says that it won’t be clear how much money and how much time the project will take until about a year and half from now, after research has begun. He estimates that the company will devote about five years to the project. IBM will probably not make the batteries but will license the technology to manufacturers.