IPF 2011: Cars, batteries, and thermoelectrics

‘Energy Frontiers’ is the title of the second of this year’s Industrial Physics Forums, which is being held in Nashville, Tennessee, in conjunction with the annual meeting of AVS: Science and Technology of Materials, Interfaces, and Processing. The session entitled “Materials for a sustainable future” included two talks about electricity in cars: one about improving the energy density of car batteries; the other about turning the heat from a car’s exhaust gases into useful electricity.

Toward a better car battery

Every year IBM challenges its researchers to nominate an exploratory project that would significantly advance science and engineering—or, as IBM’s slogan has it, build a better planet. The winner in 2009 was Battery 500, a quest to raise the energy density of electric car batteries by an order of magnitude.

The project’s leader, Sally Swanson of IBM’s research center in Almaden, California, began her IPF talk by pointing out that of the 4 million cubic meters of oil consumed each day in the US, 70% is used for transportation. Gasoline engines convert only 13% of oil’s chemical energy into motion. The rest is wasted as heat.

Electric cars are much more efficient than gasoline-powered cars. All but 10% of the energy in their batteries ends up propelling the car. But those batteries are so heavy that automobile manufacturers can’t simply add more of them to extend a car’s range. Of the current lineup of electric cars available in the US—the Tesla Roadster, the Nissan Leaf, and the Chevrolet Volt—the Tesla has the longest range, 250 miles. Its batteries weigh 450 kg.

Extending the range of electric cars would make them more attractive to consumers. Swanson and her colleagues, for example, have set a goal of 500 miles. Meeting that goal, however, entails building a new type of battery that holds more charge per kilogram.

Rather than invent a brand new battery technology, the IBM team chose to improve an existing one: the lithium–air battery. The LAB shares some properties with the lithium-ion batteries (LIBs) in the Tesla Roadster and other electric or hybrid cars. In both battery types, lithium ions from a dissolved salt flow between an anode and a carbon cathode.

In the Tesla’s LIB, the cathode is made from layers of lithium cobalt oxide; the anode is made from graphite. Electrons that originate in the cathode balance the positively charged lithium ions. In an LAB, the cathode is made from lithium and the anode is made from a porous form of carbon. The negative charge carriers are oxygen ions that diffuse into the anode from the surrounding air.

Thanks to using lithium in place of lithium cobalt oxide and porous carbon in place of graphite, LABs are considerably lighter than LIBs. But that weight advantage comes at a price. Lithium is highly reactive in air. Whereas an LIB can be completely and safely sealed, an LAB must be permeable to air. What’s more, the molecule, lithium peroxide, that mediates the flow of oxygen ions is also reactive.

Putting aside those drawbacks for now, Swanson’s team of 24 researchers have focused their efforts on improving the LAB’s performance. In particular, they are trying to identify the best anode material, lithium salt, and solvent. They also looked at using a catalyst.

Two years into the project, Swanson has concluded that

• 1. the cathode should be as porous as possible;
• 2. adding a catalyst doesn’t appear to improve performance;
• 3. the choice of salt and solvent is critical, but which salt and which solvent remains unresolved; and
• 4. some of the best salts and solvents turned out to be unstable.

She also realized that simply running through a selection of possible chemicals is not the best approach. Optimization requires understanding. To that end, her team has developed theoretical and experimental tools, including a new instrument, the differential electrochemical mass spectrometer.

The Battery 500 team is optimistic that it can achieve its goal of a 10-fold increase in energy density, but “five times would be OK,” Swanson said.

From waste heat to useful electricity

Like Swanson, Gregory Meisner of General Motors Global R&D began his IPF talk by decrying the inefficiency of gasoline engines. Forty percent of the fuel’s chemical energy is spent heating exhaust gases, which are expelled uselessly into the environment. To recover some of that lost energy, Meisner and his team are developing a thermoelectric generator (TEG).

The heart of a TEG is the thermoelectric itself, a material in which charge carriers (electrons or holes) flow in response to a thermal gradient. Two widely used thermoelectrics, bismuth telluride and lead telluride, are suitable for an automobile TEG, but they are not ideal because their performance falls off at the high temperatures (400–800 K) of exhaust gases.

Meisner and his team made their first TEG, TEG1, using the two tellurides, but their search for a better thermoelectric led them to the skutterudites. The family members consist of a transition metal (cobalt, nickel, or iron) arranged in an open cubic lattice with a metalloid (phosphor, antimony, or arsenic). Adding a rare earth (barium or ytterbium) into the lattice’s voids improves the thermal conductivity without affecting the electrical conductivity—that is, it makes the skutterudite into a good thermoelectric.

TEG1 and its successors share the same modular design. The basic element consists of pillars of n- and p-type thermoelectric that span an air-filled gap between the hot-side and cold-side connects. A sandwich of materials separates the pillars from the connects: a conducting layer to convey the current to the TEG’s electrodes and an insulating layer to prevent the current leaking out.

Making the skutterudite pillars turned out to be tough. Meisner and his team found that they achieved the best results using a technique called spark plasma sintering.

A fully assembled TEG contains 40 or so modules and is about the size of a small suitcase. To test its performance, the team installed it in a Chevrolet Suburban, whose bulk provided both the space for the installation and ample exhaust heat to work with.

On a test drive, TEG3 quickly warmed up and generated a more-or-less-steady 57 W. To reach the target of 425 W, Meisner and his team are working to further optimize the thermoelectric and to improve the thermal interfaces. In TEG1, the temperature gradient across the connects was just 100 K. If it could be boosted to 500 K, Meisner expects the TEG to generate 425 W, which would be enough to power a Suburban’s lights, entertainment systems, and air conditioning.

Charles Day